GOLDMAN-CECIL MEDICINE - 2 VOLUME SET [25th Edition][2016][UnitedVRG] [PDF] | Online Book Share (2024)

4,177 Pages • 906,424 Words • PDF • 158.6 MB

+ Volume+ Medicine

Uploaded at 2021-07-02 12:05

Report DMCA



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


ISBN: 978-1-4557-5017-7 Volume 1 Part Number: 9996096564 Volume 2 Part Number: 9996096629

International Edition (IE):

ISBN: 978-0-323-28800-2 IE Volume 1 Part Number: 9996118347 IE Volume 2 Part Number: 9996118282

Copyright © 2016, 2012, 2008, 2004, 2000, 1996, 1991, 1988, 1982, 1979, 1975, 1971, 1963, 1959, 1955 by Saunders, an imprint of Elsevier Inc. Copyright 1951, 1947, 1943, 1940, 1937, 1933, 1930, 1927 by Saunders, an imprint of Elsevier Inc. Copyright renewed 1991 by Paul Beeson. Copyright renewed 1979 by Russell L. Cecil and Robert F. Loeb. Copyright renewed 1987, 1975, 1971, 1965, 1961, 1958, 1955 by Elsevier Inc. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the Publisher. Details on how to seek permission, further information about the Publisher’s permissions policies and our arrangements with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency, can be found at our website: This book and the individual contributions contained in it are protected under copyright by the Publisher (other than as may be noted herein).

Notices Knowledge and best practice in this field are constantly changing. As new research and experience broaden our understanding, changes in research methods, professional practices, or medical treatment may become necessary. Practitioners and researchers must always rely on their own experience and knowledge in evaluating and using any information, methods, compounds, or experiments described herein. In using such information or methods they should be mindful of their own safety and the safety of others, including parties for whom they have a professional responsibility. With respect to any drug or pharmaceutical products identified, readers are advised to check the most current information provided (i) on procedures featured or (ii) by the manufacturer of each product to be administered, to verify the recommended dose or formula, the method and duration of administration, and contraindications. It is the responsibility of practitioners, relying on their own experience and knowledge of their patients, to make diagnoses, to determine dosages and the best treatment for each individual patient, and to take all appropriate safety precautions. To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors assume any liability for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions, or ideas contained in the material herein. Library of Congress Cataloging-in-Publication Data Goldman’s Cecil medicine.   Goldman-Cecil medicine / [edited by] Lee Goldman, Andrew I. Schafer.—25th edition.    p. ; cm.   Cecil medicine   Preceded by Goldman’s Cecil medicine / [edited by] Lee Goldman, Andrew I. Schafer. 24th ed. c2012.   Includes bibliographical references.   ISBN 978-1-4557-5017-7 (hardcover, 2 vol set : alk. paper)—ISBN 978-0-323-28800-2 (international edition : alk. paper)—ISBN 978-9996096563 (volume 1 : alk. paper)—ISBN 9996096564 (volume 1 : alk. paper)—ISBN 978-9996096624 (volume 2 : alk. paper)—ISBN 9996096629 (volume 2 : alk. paper)   I.  Goldman, Lee (Physician), editor.  II.  Schafer, Andrew I., editor.  III.  Title.  IV.  Title: Cecil medicine.   [DNLM:  1.  Medicine.  WB 100]   RC46   616—dc23 2014049904 Executive Content Strategist: Kate Dimock Senior Content Development Manager: Maureen Iannuzzi Publishing Services Manager: Anne Altepeter Senior Project Manager: Cindy Thoms Design Specialist: Paula Catalano Printed in the United States of America Last digit is the print number:  9  8  7  6  5  4  3  2  1


Joseph P. Routh Professor of Rheumatic Diseases in Medicine Weill Cornell Medical College Physician-in-Chief and Benjamin M. Rosen Chair in Immunology and Inflammation Research Hospital for Special Surgery New York, New York

James H. Doroshow, MD Bethesda, Maryland

Jeffrey M. Drazen, MD

Distinguished Parker B. Francis Professor of Medicine Harvard Medical School Senior Physician Brigham and Women’s Hospital Boston, Massachusetts

Robert C. Griggs, MD

Professor of Neurology, Medicine, Pediatrics, and Pathology and Laboratory Medicine University of Rochester School of Medicine and Dentistry Rochester, New York

Donald W. Landry, MD, PhD

Samuel Bard Professor of Medicine Chair, Department of Medicine Physician-in-Chief Columbia University Medical Center New York, New York

Wendy Levinson, MD Professor of Medicine Chair Emeritus Department of Medicine University of Toronto Toronto, Ontario, Canada

Anil K. Rustgi, MD

T. Grier Miller Professor of Medicine and Genetics Chief of Gastroenterology American Cancer Society Professor University of Pennsylvania Perelman School of Medicine Philadelphia, Pennsylvania

W. Michael Scheld, MD

Bayer-Gerald L. Mandell Professor of Infectious Diseases Professor of Medicine Clinical Professor of Neurosurgery Director, Pfizer Initiative in International Health University of Virginia Health System Charlottesville, Virginia

Allen M. Spiegel, MD

Dean Albert Einstein College of Medicine Bronx, New York

PREFACE In the 90 years since the first edition of the Cecil Textbook of Medicine was published, almost everything we know about internal medicine has changed. Progress in medical science is now occurring at an ever-accelerating pace, and it is doing so within the framework of transformational changes in clinical practice and the delivery of health care at individual, social, and global levels. This textbook and its associated electronic products incorporate the latest medical knowledge in multiple formats that should appeal to students and seasoned practitioners regardless of how they prefer to access this rapidly changing information. Even as Cecil’s specific information has changed, however, we have remained true to the tradition of a comprehensive textbook of medicine that carefully explains the why (the underlying pathophysiology of disease) and the how (now expected to be evidence-based from randomized controlled trials and meta-analyses). Descriptions of physiology and pathophysiology include the latest genetic advances in a practical format that strives to be useful to the nonexpert. Medicine has entered an era when the acuity of illness and the limited time available to evaluate a patient have diminished the ability of physicians to satisfy their intellectual curiosity. As a result, the acquisition of information, quite easily achieved in this era, is often confused with knowledge. We have attempted to address this dilemma with a textbook that not only informs but also stimulates new questions and gives a glimpse of the future path to new knowledge. Grade A evidence is specifically highlighted in the text and referenced at the end of each chapter. In addition to the information provided in the textbook, the Cecil website supplies expanded content and functionality. In many cases, the full articles referenced in each chapter can be accessed from the Cecil website. The website is also continuously updated to incorporate subsequent Grade A information, other evidence, and new discoveries. The sections for each organ system begin with a chapter that summarizes an approach to patients with key symptoms, signs, or laboratory abnormalities associated with dysfunction of that organ system. As summarized in E-Table 1-1, the text specifically provides clear, concise information regarding how a physician should approach more than 100 common symptoms, signs, and laboratory abnormalities, usually with a flow diagram, a table, or both for easy reference. In this way, Cecil remains a comprehensive text to guide diagnosis and therapy, not only for patients with suspected or known diseases but also for patients who may have undiagnosed abnormalities that require an initial evaluation. Just as each edition brings new authors, it also reminds us of our gratitude to past editors and authors. Previous editors of Cecil include a short but remarkably distinguished group of leaders of American medicine: Russell Cecil, Paul Beeson, Walsh McDermott, James Wyngaarden, Lloyd H. Smith,

Jr., Fred Plum, J. Claude Bennett, and Dennis Ausiello. As we welcome new associate editors—Mary K. Crow, James H. Doroshow, and Allen M. Spiegel—we also express our appreciation to William P. Arend, James O. Armitage, David R. Clemmons, and other associate editors from the previous editions on whose foundation we have built. Our returning associate editors—Jeffrey M. Drazen, Robert C. Griggs, Donald W. Landry, Wendy Levinson, Anil K. Rustgi, and W. Michael Scheld—continue to make critical contributions to the selection of authors and the review and approval of all manuscripts. The editors, however, are fully responsible for the book as well as the integration among chapters. The tradition of Cecil is that all chapters are written by distinguished experts in each field. We are also most grateful for the editorial assistance in New York of Maribel Lim and Silva Sergenian. These individuals and others in our offices have shown extraordinary dedication and equanimity in working with authors and editors to manage the unending flow of manuscripts, figures, and permissions. We also thank Cassondra Andreychik, Ved Bhushan Arya, Cameron Harrison, Karen Krok, Robert J. Mentz, Gaétane Nocturne, Patrice Savard, Senthil Senniappan, Tejpratap Tiwari, and Sangeetha Venkatarajan, who contributed to various chapters, and we mourn the passing of Morton N. Swartz, MD, co-author of the chapter on “Meningitis: Bacterial, Viral, and Other” and Donald E. Low, MD, author of the chapter “Nonpneumococcal Streptococcal Infections, Rheumatic Fever.” At Elsevier, we are most indebted to Kate Dimock and Maureen Iannuzzi, and also thank Maria Holman, Gabriela Benner, Cindy Thoms, Anne Altepeter, Linda McKinley, Paula Catalano, and Kristin Koehler, who have been critical to the planning and production process under the guidance of Mary Gatsch. Many of the clinical photographs were supplied by Charles D. Forbes and William F. Jackson, authors of Color Atlas and Text of Clinical Medicine, Third Edition, published in 2003 by Elsevier Science Ltd. We thank them for graciously permitting us to include their pictures in our book. We have been exposed to remarkable physicians in our lifetimes and would like to acknowledge the mentorship and support of several of those who exemplify this paradigm— Eugene Braunwald, Lloyd H. Smith, Jr., Frank Gardner, and William Castle. Finally, we would like to thank the Goldman family—Jill, Jeff, Abigail, Mira, Samuel, Daniel, Robyn, Tobin, and Dashel—and the Schafer family— Pauline, Eric, Melissa, Nathaniel, Pam, John, Evan, Samantha, Kate, and Sean, for their understanding of the time and focus required to edit a book that attempts to sustain the tradition of our predecessors and to meet the needs of today’s physician. LEE GOLDMAN, MD ANDREW I. SCHAFER, MD

CONTRIBUTORS Charles S. Abrams, MD Professor of Medicine, Pathology, and Laboratory Medicine, University of Pennsylvania School of Medicine; Director, PENN-Chop Blood Center for Patient Care & Discovery, Hospital of the University of Pennsylvania, Philadelphia, Pennsylvania Thrombocytopenia Frank J. Accurso, MD Professor of Pediatrics, University of Colorado School of Medicine; Attending Physician, Children’s Hospital Colorado, Aurora, Colorado Cystic Fibrosis Ronald S. Adler, MD, PhD Professor of Radiology, New York University School of Medicine; Department of Radiology, NYU Langone Medical Center, New York, New York Imaging Studies in the Rheumatic Diseases Cem Akin, MD, PhD Associate Professor, Harvard Medical School; Attending Physician, Director, Mastocytosis Center, Brigham and Women’s Hospital, Department of Medicine, Division of Rheumatology, Immunology, and Allergy, Boston, Massachusetts Mastocytosis Allen J. Aksamit, Jr., MD Professor of Neurology, Mayo Clinic College of Medicine, Consultant in Neurology, Mayo Clinic, Rochester, Minnesota Acute Viral Encephalitis Qais Al-Awqati, MB ChB Robert F. Loeb Professor of Medicine, Jay I. Meltzer Professor of Nephrology and Hypertension, Professor of Physiology and Cellular Biophysics, Division of Nephrology, Columbia University, College of Physicians and Surgeons, New York, New York Structure and Function of the Kidneys Ban Mishu Allos, MD Associate Professor of Medicine, Division of Infectious Diseases, Associate Professor, Preventive Medicine, Vanderbilt University School of Medicine, Nashville, Tennessee Campylobacter Infections David Altshuler, MD, PhD Professor of Genetics and of Medicine, Harvard Medical School, Massachusetts General Hospital; Professor of Biology (Adjunct), Massachusetts Institute of Technology, Boston and Cambridge, Massachusetts The Inherited Basis of Common Diseases

Larry J. Anderson, MD Professor, Division of Infectious Disease, Department of Pediatrics, Emory University School of Medicine and Children’s Healthcare of Atlanta, Atlanta, Georgia Coronaviruses Aśok C. Antony, MD Chancellor’s Professor of Medicine, Indiana University School of Medicine; Attending Physician, Indiana University Health Affiliated Hospitals and Richard L. Roudebush Veterans Affairs Medical Center, Indianapolis, Indiana Megaloblastic Anemias Gerald B. Appel, MD Professor of Medicine, Division of Nephrology, Department of Medicine, Columbia University College of Physicians and Surgeons, New York, New York Glomerular Disorders and Nephrotic Syndromes Frederick R. Appelbaum, MD Executive Vice President and Deputy Director, Fred Hutchinson Cancer Research Center; President, Seattle Cancer Care Alliance; Professor, Division of Medical Oncology, University of Washington School of Medicine, Seattle Washington The Acute Leukemias Suneel S. Apte, MBBS, DPhil Staff, Cleveland Clinic Lerner College of Medicine at Case Western Reserve University, Cleveland, Ohio Connective Tissue Structure and Function James O. Armitage, MD The Joe Shapiro Professor of Internal Medicine, University of Nebraska Medical Center, Omaha, Nebraska Approach to the Patient with Lymphadenopathy and Splenomegaly; Non-Hodgkin Lymphomas M. Amin Arnaout, MD Professor of Medicine, Departments of Medicine and Developmental and Regenerative Biology, Harvard Medical School; Physician and Chief Emeritus, Division of Nephrology, Massachusetts General Hospital, Boston, Massachusetts Cystic Kidney Diseases Robert M. Arnold, MD Leo H. Criep Professor of Clinical Care, Chief, Section of Palliative Care and Medical Ethics, University of Pittsburgh; Medical Director, UPMC Palliative and Supportive Care Institute, Pittsburgh, Pennsylvania Care of Dying Patients and Their Families

Michael Aminoff, MD, DSc Professor, Department of Neurology, University of California San Francisco, San Francisco, California Approach to the Patient with Neurologic Disease

David Atkins, MD, MPH Director, Health Services Research and Development, Veterans Health Administration, Washington, D.C. The Periodic Health Examination

Jeffrey L. Anderson, MD Professor of Internal Medicine, University of Utah School of Medicine; Vice-Chair for Research, Department of Internal Medicine, Associate Chief of Cardiology and Director of Cardiovascular Research, Intermountain Medical Center, Intermountain Healthcare, Salt Lake City, Utah ST Segment Elevation Acute Myocardial Infarction and Complications of Myocardial Infarction

John P. Atkinson, MD Chief, Division of Rheumatology, Internal Medicine, Washington University School of Medicine in St. Louis, St. Louis, Missouri Complement System in Disease



Bruce R. Bacon, MD Endowed Chair in Gastroenterology, Professor of Internal Medicine, Co-Director, Saint Louis University Liver Center; Director, Saint Louis University Abdominal Transplant Center, Saint Louis University School of Medicine, St. Louis, Missouri Iron Overload (Hemochromatosis) Larry M. Baddour, MD Professor of Medicine, Chair, Division of Infectious Diseases, Mayo Clinic, Rochester, Minnesota Infective Endocarditis Grover C. Bagby, MD Professor of Medicine and Molecular and Medical Genetics, Knight Cancer Institute at Oregon Health and Science University and Portland VA Medical Center, Portland, Oregon Aplastic Anemia and Related Bone Marrow Failure States Barbara J. Bain, MBBS Professor in Diagnostic Haematology, Imperial College London; Honorary Consultant Haematologist, St. Mary’s Hospital, London, United Kingdom The Peripheral Blood Smear Dean F. Bajorin, MD Attending Physician and Member, Medicine, Memorial Hospital, Memorial Sloan Kettering Cancer Center; Professor of Medicine, Weill Cornell Medical College, New York, New York Tumors of the Kidney, Bladder, Ureters, and Renal Pelvis

Stephen G. Baum, MD Chairman of Medicine, Mount Sinai Beth Israel Hospital; Professor of Medicine and of Microbiology and Immunology, Albert Einstein College of Medicine, New York, New York Mycoplasma Infections Daniel G. Bausch, MD, MPH&TM Associate Professor, Department of Tropical Medicine, Tulane University Health Sciences Center, New Orleans, Louisiana Viral Hemorrhagic Fevers Arnold S. Bayer, MD Professor of Medicine, David Geffen School of Medicine at University of California Los Angeles; LA Biomedical Research Institute; Vice Chair for Academic Affairs, Department of Medicine, Harbor-UCLA Medical Center, Los Angeles, California Infective Endocarditis Hasan Bazari, MD Associate Professor of Medicine, Harvard Medical School, Department of Medicine, Clinical Director, Nephrology, Program Director, Internal Medicine Residency Program, Massachusetts General Hospital, Boston, Massachusetts Approach to the Patient with Renal Disease John H. Beigel, MD National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, Maryland Antiviral Therapy (Non-HIV)

Robert W. Baloh, MD Professor of Neurology, University of California Los Angeles School of Medicine, Los Angeles, California Neuro-Ophthalmology; Smell and Taste; Hearing and Equilibrium

George A. Beller, MD Professor of Medicine, University of Virginia Health System, Charlottesville, Virginia Noninvasive Cardiac Imaging

Jonathan Barasch, MD, PhD Professor of Medicine and Pathology and Cell Biology, Department of Medicine, Division of Nephrology, Columbia University College of Physicians & Surgeons, New York, New York Structure and Function of the Kidneys

Robert M. Bennett, MD Professor of Medicine, Oregon Health and Science University, Portland, Oregon Fibromyalgia, Chronic Fatigue Syndrome, and Myofascial Pain

Richard L. Barbano, MD, PhD Professor of Neurology, University of Rochester, Rochester, New York Mechanical and Other Lesions of the Spine, Nerve Roots, and Spinal Cord Elizabeth Barrett-Connor, MD Professor of Community and Family Medicine, University of California San Diego, San Diego, California Menopause John R. Bartholomew, MD Section Head, Vascular Medicine, Cardiovascular Medicine, Cleveland Clinic, Professor of Medicine, Cleveland Clinic Lerner College of Medicine of Case Western Reserve University, Cleveland, Ohio Other Peripheral Arterial Diseases Mary Barton, MD, MPP Vice President, Performance Measurement, National Committee for Quality Assurance, Washington, D.C. The Periodic Health Examination Robert C. Basner, MD Professor of Medicine, Columbia University Medical Center; Director, Columbia University Cardiopulmonary Sleep and Ventilatory Disorders Center, Columbia University College of Physicians and Surgeons, New York, New York Obstructive Sleep Apnea

Joseph R. Berger, MD Professor of Neurology, Chief of the Multiple Sclerosis Division, Department of Neurology, Perelman School of Medicine, University of Pennsylvania, Philadelphia, Pennsylvania Cytomegalovirus, Epstein-Barr Virus, and Slow Virus Infections of the Central Nervous System; Neurologic Complications of Human Immunodeficiency Virus Infection; Brain Abscess and Parameningeal Infections Paul D. Berk, MD Professor of Medicine, Department of Medicine, Columbia University College of Physicians and Surgeons, New York, New York Approach to the Patient with Jaundice or Abnormal Liver Tests Nancy Berliner, MD Professor of Medicine, Harvard Medical School; Chief, Division of Hematology, Brigham and Women’s Hospital, Boston, Massachusetts Leukocytosis and Leukopenia James L. Bernat, MD Louis and Ruth Frank Professor of Neuroscience, Professor of Neurology and Medicine, Geisel School of Medicine at Dartmouth, Hanover, New Hampshire; Department of Neurology, Dartmouth-Hitchco*ck Medical Center, Lebanon, New Hampshire Coma, Vegetative State, and Brain Death Philip J. Bierman, MD Professor, Department of Internal Medicine, University of Nebraska Medical Center, Omaha, Nebraska Approach to the Patient with Lymphadenopathy and Splenomegaly; Non-Hodgkin Lymphomas



Michael R. Bishop, MD Professor of Medicine, Director, Hematopoietic Cellular Therapy Program, Section of Hematology and Oncology, Department of Medicine, University of Chicago, Chicago, Illinois Hematopoietic Stem Cell Transplantation

William E. Boden, MD Professor of Medicine, Albany Medical College; Chief of Medicine, Albany Stratton VA Medical Center; Vice-Chairman, Department of Medicine, Albany Medical Center, Albany, New York Angina Pectoris and Stable Ischemic Heart Disease

Bruce R. Bistrian, MD, PhD, MPH Professor of Medicine, Beth Israel Deaconess Medical Center; Professor of Medicine, Harvard Medical School, Boston, Massachusetts Nutritional Assessment

Jean Bolognia, MD Professor of Dermatology, Yale Medical School; Attending Physician, Yale-New Haven Hospital, New Haven, Connecticut Infections, Hyperpigmentation and Hypopigmentation, Regional Dermatology, and Distinctive Lesions in Black Skin

Joseph J. Biundo, MD Clinical Professor of Medicine, Tulane Medical Center, New Orleans, Louisiana Bursitis, Tendinitis, and Other Periarticular Disorders and Sports Medicine Adrian R. Black, PhD Assistant Professor, Director of Tissue Sciences for the Eppley Institute, The Eppley Institute for Research in Cancer and Allied Diseases, University of Nebraska Medical Center, Omaha, Nebraska Cancer Biology and Genetics Charles D. Blanke, MD Professor of Medicine, Oregon Health and Science University, Portland, Oregon Neoplasms of the Small and Large Intestine Joel N. Blankson, MD, PhD Associate Professor, Johns Hopkins University School of Medicine, Baltimore, Maryland Immunopathogenesis of Human Immunodeficiency Virus Infection Martin J. Blaser, MD Muriel and George Singer Professor of Medicine, Professor of Microbiology, Director, Human Microbiome Program, New York University Langone Medical Center, New York, New York Acid Peptic Disease; Human Microbiome William A. Blattner, MD Professor and Associate Director, Institute of Human Virology, School of Medicine, University of Maryland; Professor of Medicine, School of Medicine, University of Maryland; Professor and Head, Division of Cancer Epidemiology, Department of Epidemiology and Public Health, School of Medicine, University of Maryland, Baltimore, Maryland Retroviruses Other Than Human Immunodeficiency Virus Thomas P. Bleck, MD Professor of Neurological Sciences, Neurosurgery, Internal Medicine, and Anesthesiology, Associate Chief Medical Officer (Critical Care), Rush Medical College, Chicago, Illinois Arboviruses Affecting the Central Nervous System Joel A. Block, MD The Willard L. Wood MD Professor and Director, Division of Rheumatology, Rush University Medical Center, Chicago, Illinois Osteoarthritis Henk Blom, MD Laboratory of Clinical Biochemistry and Metabolism, Department of General Pediatrics, Adolescent Medicine and Neonatology, University Medical Centre Freiburg, Head of Laboratory/Clinical Biochemical Geneticist, Freiburg, Germany hom*ocystinuria and Hyperhom*ocysteinemia Olaf A. Bodamer, MD Medical Genetics, University of Miami Hospital, Miami, Florida Approach to Inborn Errors of Metabolism

Robert A. Bonomo, MD Chief, Medical Service, Louis Stokes Cleveland VA Medical Center; Professor of Medicine, Pharmacology, Biochemistry, Molecular Biology, and Microbiology, Case Western Reserve University School of Medicine, Cleveland, Ohio Diseases Caused by Acinetobacter and Stenotrophom*onas Species Larry Borish, MD Professor of Medicine, Allergy, and Clinical Immunology, University of Virginia Health System, Charlottesville, Virgina Allergic Rhinitis and Chronic Sinusitis Patrick J. Bosque, MD Associate Professor of Neurology, University of Colorado Denver School of Medicine; Neurologist, Denver Health Medical Center, Denver, Colorado Prion Diseases David J. Brenner, PhD, DSc Higgins Professor of Radiation Biophysics, Center for Radiological Research, Columbia University Medical Center, New York, New York Radiation Injury Itzhak Brook, MD, MSc Professor of Pediatrics and Medicine, Georgetown University, Georgetown University Medical Center, Washington, D.C. Diseases Caused by Non–Spore-Forming Anaerobic Bacteria; Actinomycosis Enrico Brunetti, MD Assistant Professor of Infectious Diseases, University of Pavia; Attending Physician, Division of Infectious and Tropical Diseases, IRCCS San Matteo Hospital Foundation; Co-Director, WHO Collaborating Centre for Clinical Management of Cystic Echinococcosis, Pavia, Italy Cestodes David M. Buchner, MD, MPH Shahid and Ann Carlson Khan Professor in Applied Health Sciences, Department of Kinesiology and Community Health, University of Illinois at Urbana-Champaign, Champaign, Illinois Physical Activity Pierre A. Buffet, MD, PhD Research Unit Head, Erythrocyte Parasite Pathogenesis Research Team INSERM–University Paris 6, CIMI–Paris Research Center, University Pierre and Marie Curie; Associate Professor of Parasitology, Faculty of Medicine, University Pierre and Marie Curie, Pitié-Salpêtrière Hospital, Paris, France Leishmaniasis H. Franklin Bunn, MD Professor of Medicine, Harvard Medical School; Physician, Brigham and Women’s Hospital, Boston, Massachusetts Approach to the Anemias David A. Bushinsky, MD John J. Kuiper Distinguished Professor of Medicine, Chief, Nephrology Division, University of Rochester School of Medicine; Associate Chair for Academic Affairs in Medicine, University of Rochester Medical Center, Rochester, New York Nephrolithiasis



Vivian P. Bykerk, MD Associate Professor of Medicine, Weill Cornell Medical College; Associate Attending Physician, Hospital for Special Surgery, New York, New York Approach to the Patient with Rheumatic Disease Peter A. Calabresi, MD Professor of Neurology and Director of the Richard T. Johnson Division of Neuroimmunology and Neuroinfectious Diseases, Johns Hopkins University; Director of the Multiple Sclerosis Center, Johns Hopkins Hospital, Baltimore, Maryland Multiple Sclerosis and Demyelinating Conditions of the Central Nervous System David P. Calfee, MD, MS Associate Professor of Medicine and Healthcare Policy and Research, Weill Cornell Medical College; Chief Hospital Epidemiologist, New YorkPresbyterian Hospital/Weill Cornell Medical Center, New York, New York Prevention and Control of Health Care–Associated Infections Douglas Cameron, MD, MBA Professor of Ophthalmology and Visual Neurosciences, University of Minnesota, Minneapolis, Minnesota Diseases of the Visual System Michael Camilleri, MD Atherton and Winifred W. Bean Professor, Professor of Medicine, Pharmacology, and Physiology, College of Medicine, Mayo Clinic, Consultant, Division of Gastroenterology and Hepatology, Mayo Clinic, Rochester, Minnesota Disorders of Gastrointestinal Motility Grant W. Cannon, MD Thomas E. and Rebecca D. Jeremy Presidential Endowed Chair for Arthritis Research, Associate Chief of Staff for Academic Affiliations, George E. Wahlen VA Medical Center, Salt Lake City, Utah Immunosuppressing Drugs Including Corticosteroids Maria Domenica Cappellini, MD Professor of Internal Medicine, University of Milan, Fondazione IRCCS Ca’ Granda Ospedale Maggiore Policlinico, University of Milan, Milan, Italy The Thalassemias Blase A. Carabello, MD Professor of Medicine, Chairman, Department of Cardiology, Mount Sinai Beth Israel Heart Institute, New York, New York Valvular Heart Disease Edgar M. Carvalho, MD Professor of Medicine and Clinical Immunology, Faculdade de Medicina da Bahia, Universidade Federal da Bahia and Escola Bahiana de Medicina e Saúde Pública, Salvador, Bahia, Brazil Schistosomiasis (Bilharziasis) William H. Catherino, MD, PhD Professor and Research Head, Department of Obstetrics and Gynecology, Uniformed Services University of the Health Sciences Division of Reproductive Endocrinology and Infertility; Program in Reproductive and Adult Endocrinology, Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, Maryland Ovaries and Development; Reproductive Endocrinology and Infertility Jane A. Cauley, DrPH Professor of Epidemiology, University of Pittsburgh Graduate School of Public Health, Vice Chair of the Department of Epidemiology, Pittsburgh, Pennsylvania Epidemiology of Aging: Implications of the Aging of Society

Naga P. Chalasani, MD David W. Crabb Professor and Director, Division of Gastroenterology and Hepatology, Indiana University School of Medicine, Indianapolis, Indiana Alcoholic and Nonalcoholic Steatohepatitis Henry F. Chambers, MD Professor of Medicine, University of California San Francisco School of Medicine; Director, Clinical Research Services, Clinical and Translational Sciences Institute, San Francisco, California Staphylococcal Infections William P. Cheshire, Jr., MD Professor of Neurology, Mayo Clinic, Jacksonville, Florida Autonomic Disorders and Their Management Ilseung Cho, MD, MS Assistant Professor of Medicine, Division of Gastroenterology, Department of Medicine, New York University, New York, New York Human Microbiome Arun Chockalingam, PhD Professor of Epidemiology and Global Health, Director, Office of Global Health Education and Training; Dalla Lana Faculty of Public Health, University of Toronto, Toronto, Ontario, Canada Global Health David C. Christiani, MD Professor of Medicine, Harvard Medical School; Physician, Pulmonary and Critical Care, Massachusetts General Hospital; Elkan Blout Professor of Environmental Genetics, Environmental Health, Harvard School of Public Health, Boston, Massachusetts Physical and Chemical Injuries of the Lung David H. Chu, MD, PhD Director, Contact Dermatitis, Division of Dermatology and Cutaneous Surgery, Scripps Clinic Medical Group, La Jolla, California Structure and Function of the Skin Theodore J. Cieslak, MD Pediatric Infectious Diseases, Clinical Professor of Pediatrics, University of Texas Health Science Center at San Antonio; Department of Pediatrics, Fort Sam Houston, Texas Bioterrorism Carolyn Clancy, MD Interim Under Secretary for Health, Veterans Administration, Washington, D.C. Measuring Health and Health Care David R. Clemmons, MD Kenan Professor of Medicine, University of North Carolina School of Medicine; Attending Physician, Medicine, UNC Hospitals, Chapel Hill, North Carolina Approach to the Patient with Endocrine Disease David Cohen, MD Professor of Medicine, Division of Nephrology; Medical Director, Kidney and Pancreas Transplantation, Columbia University Medical Center, New York, New York Treatment of Irreversible Renal Failure Jeffrey Cohen, MD Chief, Laboratory of Infectious Diseases, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, Maryland Varicella-Zoster Virus (Chickenpox, Shingles)



Myron S. Cohen, MD Associate Vice Chancellor for Global Health, Director, UNC Institute for Global Health and Infectious Diseases, Chief, Division of Infectious Diseases, Yeargan-Bate Eminent Professor of Medicine, Microbiology, and Immunology and Epidemiology, Chapel Hill, North Carolina Approach to the Patient with a Sexually Transmitted Infection; Prevention of Human Immunodeficiency Virus Infection

Mary K. Crow, MD Joseph P. Routh Professor of Rheumatic Diseases in Medicine, Weill Cornell Medical College; Physician in Chief and Benjamin M. Rosen Chair in Immunology and Inflammation Research, Hospital for Special Surgery, New York, New York The Innate Immune Systems; Approach to the Patient with Rheumatic Disease; Systemic Lupus Erythematosus

Steven P. Cohen, MD Professor of Anesthesiology and Critical Care Medicine and Physical Medicine and Rehabilitation, Johns Hopkins School of Medicine, Baltimore, Maryland, and Uniformed Services University of the Health Sciences, Bethesda, Maryland; Director, Pain Research, Walter Reed National Military Medical Center, Bethesda, Maryland Pain

John A. Crump, MB ChB, MD, DTM&H McKinlay Professor of Global Health, Centre for International Health, University of Otago, Dunedin, New Zealand Salmonella Infections (Including Enteric Fever)

Steven L. Cohn, MD Professor of Clinical Medicine, University of Miami Miller School of Medicine; Medical Director, UHealth Preoperative Assessment Center; Director, Medical Consultation Service, University of Miami Hospital, Miami, Florida Preoperative Evaluation Robert Colebunders, MD Emeritus Professor, Institute of Tropical Medicine, Antwerp, Belgium Immune Reconstitution Inflammatory Syndrome in HIV/AIDS Joseph M. Connors, MD Clinical Professor, University of British Columbia; Clinical Director, BC Cancer Agency Centre for Lymphoid Cancer, Vancouver, British Columbia, Canada Hodgkin Lymphoma Deborah J. Cook, MD, MSc Professor of Medicine, Clinical Epidemiology and Biostatistics, McMaster University, Hamilton, Ontario, Canada Approach to the Patient in a Critical Care Setting Kenneth H. Cowan, MD, PhD Director, Fred & Pamela Buffett Cancer Center; Director, The Eppley Institute for Research in Cancer and Allied Diseases; Professor of Medicine, University of Nebraska Medical Center, Omaha, Nebraska Cancer Biology and Genetics Joseph Craft, MD Paul B. Beeson Professor of Medicine and Immunobiology, Section Chief, Rheumatology, Program Director, Investigative Medicine, Department of Internal Medicine, Yale University School of Medicine, New Haven, Connecticut The Adaptive Immune Systems Jill Patricia Crandall, MD Professor of Clinical Medicine, Division of Endocrinology and Diabetes Research Center, Albert Einstein College of Medicine, Bronx, New York Diabetes Mellitus Simon L. Croft, BSc, PhD Professor of Parasitology, Faculty of Infectious and Tropical Diseases, London School of Hygiene and Tropical Medicine, London, United Kingdom Leishmaniasis Kristina Crothers, MD Associate Professor, Department of Medicine, Division of Pulmonary and Critical Care, University of Washington School of Medicine, Seattle, Washington Pulmonary Manifestations of Human Immunodeficiency Virus and Acquired Immunodeficiency Syndrome

Mark R. Cullen, MD Professor of Medicine, Department of Medicine, Stanford University School of Medicine, Stanford, California Principles of Occupational and Environmental Medicine Charlotte Cunningham-Rundles, MD, PhD Professor of Medicine and Pediatrics, Icahn School of Medicine at Mount Sinai, New York, New York Primary Immunodeficiency Diseases Inger K. Damon, MD, PhD Director, Division of High Consequence Pathogens and Pathology, Centers for Disease Control and Prevention, Atlanta, Georgia Smallpox, Monkeypox, and Other Poxvirus Infections Troy E. Daniels, DDS, MS Professor Emeritus of Oral Pathology and Pathology, University of California San Francisco, San Francisco, California Diseases of the Mouth and Salivary Glands Nancy E. Davidson, MD Hillman Professor of Oncology, University of Pittsburgh; Director, University of Pittsburgh Cancer Institute and UPMC CancerCenter, Pittsburgh, Pennsylvania Breast Cancer and Benign Breast Disorders Lisa M. DeAngelis, MD Chair, Department of Neurology, Memorial Sloan-Kettering Cancer Center; Professor of Neurology, Weill Cornell Medical College, New York, New York Tumors of the Central Nervous System Malcolm M. DeCamp, MD Fowler McCormick Professor of Surgery, Feinberg School of Medicine, Northwestern University; Chief, Division of Thoracic Surgery, Northwestern Memorial Hospital, Chicago, Illinois Interventional and Surgical Approaches to Lung Disease Carlos del Rio, MD Hubert Professor and Chair and Professor of Medicine, Hubert Department of Global Health, Rollins School of Public Health and Department of Medicine, Emory University School of Medicine, Atlanta, Georgia Prevention of Human Immunodeficiency Virus Infection Patricia A. Deuster, PhD, MPH Professor and Director, Consortium for Health and Military Performance, Department of Military and Emergency Medicine, Uniformed Services University of the Health Sciences, Bethesda, Maryland Rhabdomyolysis Robert B. Diasio, MD William J. and Charles H. Mayo Professor, Molecular Pharmacology and Experimental Therapeutics and Oncology, Mayo Clinic, Rochester, Minnesota Principles of Drug Therapy



David J. Diemert, MD Associate Professor, Department of Microbiology, Immunology, and Tropical Medicine, School of Medicine and Health Sciences, The George Washington University, Washington, D.C. Intestinal Nematode Infections; Tissue Nematode Infections Kathleen B. Digre, MD Professor of Neurology, Ophthalmology, Director, Division of Headache and Neuro-Ophthalmology, University of Utah, Salt Lake City, Utah Headaches and Other Head Pain James H. Doroshow, MD Bethesda, Maryland Approach to the Patient with Cancer; Malignant Tumors of Bone, Sarcomas, and Other Soft Tissue Neoplasms John M. Douglas, Jr., MD Executive Director, Tri-County Health Department, Greenwood Village, Colorado Papillomavirus Jeffrey M. Drazen, MD Distinguished Parker B. Francis Professor of Medicine, Harvard Medical School; Senior Physician, Brigham and Women’s Hospital, Boston, Massachusetts Asthma Stephen C. Dreskin, MD, PhD Professor of Medicine and Immunology, Division of Allergy and Clinical Immunology, Department of Medicine, University of Colorado Denver, School of Medicine, Aurora, Colorado Urticaria and Angioedema W. Lawrence Drew, MD, PhD Professor Emeritus, Laboratory Medicine and Medicine, University of California San Francisco, San Francisco, California Cytomegalovirus George L. Drusano, MD Professor and Director, Institute for Therapeutic Innovation, College of Medicine, University of Florida, Lake Nona, Florida Antibacterial Chemotherapy Thomas D. DuBose, Jr., MD Emeritus Professor of Internal Medicine and Nephrology, Wake Forest School of Medicine, Winston-Salem, North Carolina Vascular Disorders of the Kidney F. Daniel Duffy, MD Professor of Internal Medicine and Steve Landgarten Chair in Medical Leadership, School of Community Medicine, University of Oklahoma College of Medicine, Tulsa, Oklahoma Counseling for Behavior Change Herbert L. DuPont, MD, MACP Mary W. Kelsey Chair and Director, Center for Infectious Diseases, University of Texas School of Public Health; H. Irving Schweppe Chair of Internal Medicine and Vice Chairman, Department of Medicine, Baylor College of Medicine; Chief of Internal Medicine, St. Luke’s Hospital System, Houston, Texas Approach to the Patient with Suspected Enteric Infection Madeleine Duvic, MD Professor and Deputy Chairman, Department of Dermatology, The University of Texas MD Anderson Cancer Center, Houston, Texas Urticaria, Drug Hypersensitivity Rashes, Nodules and Tumors, and Atrophic Diseases

Kathryn M. Edwards, MD Sarah H. Sell and Cornelius Vanderbilt Chair in Pediatrics, Vanderbilt University School of Medicine; Director, Vanderbilt Vaccine Research Program, Monroe Carrell Jr. Children’s Hospital at Vanderbilt, Nashville, Tennessee Parainfluenza Viral Disease N. Lawrence Edwards, MD Professor of Medicine, Vice Chairman, Department of Medicine, University of Florida; Chief, Section of Rheumatology, Medical Service, Malcom Randall Veterans Affairs Medical Center, Gainesville, Florida Crystal Deposition Diseases Lawrence H. Einhorn, MD Distinguished Professor, Department of Medicine, Division of Hematology/Oncology, Livestrong Foundation Professor of Oncology, Indiana University School of Medicine, Indianapolis, Indiana Testicular Cancer Ronald J. Elin, MD, PhD A.J. Miller Professor and Chairman, Department of Pathology and Laboratory Medicine, University of Louisville School of Medicine, Louisville, Kentucky Reference Intervals and Laboratory Values George M. Eliopoulos, MD Professor of Medicine, Harvard Medical School; Physician, Division of Infectious Diseases, Beth Israel Deaconess Medical Center, Boston, Massachusetts Principles of Anti-Infective Therapy Perry Elliott, MD Professor in Inherited Cardiovascular Disease, Institute of Cardiovascular Science, University College London, London, United Kingdom Diseases of the Myocardium and Endocardium Jerrold J. Ellner, MD Professor of Medicine, Boston University School of Medicine; Chief, Section of Infectious Diseases, Boston Medical Center, Boston, Massachusetts Tuberculosis Dirk M. Elston, MD Director, Ackerman Academy of Dermatopathology, New York, New York Arthropods and Leeches Ezekiel J. Emanuel, MD, PhD Vice Provost for Global Initiatives, Diane V.S. Levy and Robert M. Levy University Professor, Chair, Department of Medical Ethics and Health Policy, University of Pennsylvania, Philadelphia, Pennsylvania Bioethics in the Practice of Medicine Joel D. Ernst, MD Director, Division of Infectious Diseases and Immunology, Jeffrey Bergstein Professor of Medicine, Professor of Medicine, Pathology, and Microbiology, New York University School of Medicine; Attending Physician, New York University Langone Medical Center, New York, New York Leprosy (Hansen Disease) Gregory T. Everson, MD Professor of Medicine, Director of Hepatology, University of Colorado School of Medicine, Aurora, Colorado Hepatic Failure and Liver Transplantation Amelia Evoli, MD Associate Professor of Neurology, Catholic University, Agostino Gemelli University Hospital, Rome, Italy Disorders of Neuromuscular Transmission

Contributors Douglas O. Faigel, MD Professor of Medicine, Mayo Clinic, Chair, Division of Gastroenterology and Hepatology, Scottsdale, Arizona Neoplasms of the Small and Large Intestine

Manuel A. Franco, MD, PhD Director of Postgraduate Programs, School of Sciences, Pontificia Universidad Javeriana, Bogota, Colombia Rotaviruses, Noroviruses, and Other Gastrointestinal Viruses

Matthew E. Falagas, MD, MSc, DSc Director, Alfa Institute of Biomedical Sciences, Athens, Greece; Adjunct Associate Professor of Medicine, Tufts University School of Medicine, Boston, Massachusetts; Chief, Department of Medicine and Infectious Diseases, Iaso General Hospital, Iaso Group, Athens, Greece Pseudomonas and Related Gram-Negative Bacillary Infections

David O. Freedman, MD Professor of Medicine and Microbiology, University of Alabama at Birmingham; Director, Gorgas Center for Geographic Medicine, Birmingham, Alabama Approach to the Patient before and after Travel

Gary W. Falk, MD, MS Professor of Medicine, Division of Gastroenterology, University of Pennsylvania Perelman School of Medicine, Philadelphia, Pennsylvania Diseases of the Esophagus Gene Feder, MBBS, MD Professor, Centre for Academic Primary Care, School of Social and Community Medicine, University of Bristol; General Practitioner, Helios Medical Centre, Bristol, United Kingdom Intimate Partner Violence David J. Feller-Kopman, MD Director, Bronchoscopy and Interventional Pulmonology, Associate Professor of Medicine, The Johns Hopkins University, Baltimore, Maryland Interventional and Surgical Approaches to Lung Disease Gary S. Firestein, MD Dean and Associate Vice Chancellor of Translational Medicine, University of California San Diego School of Medicine, La Jolla, California Mechanisms of Inflammation and Tissue Repair Glenn I. Fishman, MD Director, Leon H. Charney Division of Cardiology, Vice-Chair for Research, Department of Medicine, William Goldring Professor of Medicine, New York University School of Medicine, New York, New York Principles of Electrophysiology Lee A. Fleisher, MD Robert D. Dripps Professor and Chair, Anesthesiology and Critical Care, Professor of Medicine, University of Pennsylvania Perelman School of Medicine, Philadelphia, Pennsylvania Overview of Anesthesia Paul W. Flint, MD Professor and Chair, Otolaryngology, Head and Neck Surgery, Oregon Health and Science University, Portland, Oregon Throat Disorders Evan L. Fogel, MD, MSc Professor of Clinical Medicine, Indiana University School of Medicine, Indianapolis, Indiana Diseases of the Gallbladder and Bile Ducts Marsha D. Ford, MD Adjunct Professor of Emergency Medicine, School of Medicine, University of North Carolina-Chapel Hill; Director, Carolinas Poison Center, Carolinas HealthCare System, Charlotte, North Carolina Acute Poisoning Chris E. Forsmark, MD Professor of Medicine, Chief, Division of Gastroenterology, Hepatology, and Nutrition, University of Florida, Gainesville, Florida Pancreatitis Vance G. Fowler, Jr., MD, MHS Professor of Medicine, Duke University Medical Center, Durham, North Carolina Infective Endocarditis


Martyn A. French, MD Professor in Clinical Immunology, School of Pathology and Laboratory Medicine, University of Western Australia, Perth, Australia Immune Reconstitution Inflammatory Syndrome in HIV/AIDS Karen Freund, MD, MPH Professor of Medicine, Associate Director, Tufts Clinical and Translational Science Institute, Tufts University School of Medicine, Tufts Medical Center, Boston, Massachusetts Approach to Women’s Health Cem Gabay, MD Professor of Medicine, Head, Division of Rheumatology, University Hospitals of Geneva, Geneva, Switzerland Biologic Agents Kenneth L. Gage, PhD Chief, Entomology and Ecology Activity, Centers for Disease Control and Prevention, Division of Vector-Borne Diseases, Bacterial Diseases Branch, Fort Collins, Colorado Plague and Other Yersinia Infections John N. Galgiani, MD Professor of Medicine, Valley Fever Center for Excellence, University of Arizona, Tucson, Arizona Coccidioidomycosis Patrick G. Gallagher, MD Professor of Pediatrics, Pathology, and Genetics, Yale University School of Medicine; Attending Physician, Yale–New Haven Hospital, New Haven, Connecticut Hemolytic Anemias: Red Blood Cell Membrane and Metabolic Defects Leonard Ganz, MD Director of Cardiac Electrophysiology, Heritage Valley Health System, Beaver, Pennsylvania Electrocardiography Hasan Garan, MD Director, Cardiac Electrophysiology, Dickinson W. Richards, Jr. Professor of Medicine, Columbia University Medical Center, New York, New York Ventricular Arrhythmias Guadalupe Garcia-Tsao, MD Professor of Medicine, Yale University School of Medicine; Chief, Digestive Diseases, VA Connecticut Healthcare System, West Haven, Connecticut Cirrhosis and Its Sequelae William M. Geisler, MD, MPH Professor of Medicine, University of Alabama at Birmingham, Birmingham, Alabama Diseases Caused by Chlamydiae Tony P. George, MD Division of Brain and Therapeutics, Department of Psychiatry, University of Toronto; Schizophrenia Division, The Centre for Addiction and Mental Health, Toronto, Ontario, Canada Nicotine and Tobacco



Lior Gepstein, MD, PhD Edna and Jonathan Sohnis Professor in Medicine and Physiology, Rappaport Faculty of Medicine and Research Institute, Technion–Israel Institute of Technology, Rambam Health Care Campus, Haifa, Israel Gene and Cell Therapy

Larry B. Goldstein, MD Professor of Neurology, Director, Duke Stroke Center, Neurology, Duke University; Staff Neurologist, Durham VA Medical Center, Durham, North Carolina Approach to Cerebrovascular Diseases; Ischemic Cerebrovascular Disease

Susan I. Gerber, MD Team Lead, Respiratory Viruses/Picornaviruses, Division of Viral Diseases/Epidemiology Branch, National Center for Immunization and Respiratory Diseases, Centers for Disease Control and Prevention, Atlanta, Georgia Coronaviruses

Lawrence T. Goodnough, MD Professor of Pathology and Medicine, Stanford University; Director, Transfusion Service, Stanford University Medical Center, Stanford, California Transfusion Medicine

Dale N. Gerding, MD Professor of Medicine, Loyola University Chicago Stritch School of Medicine, Research Physician, Edward Hines, Jr. VA Hospital, Hines, Illinois Clostridial Infections Morie A. Gertz, MD Consultant, Division of Hematology, Mayo Clinic, Rochester, Minnesota; Roland Seidler, Jr. Professor of the Art of Medicine in Honor of Michael D. Brennan, MD, Professor of Medicine, Mayo Clinic, College of Medicine, Rochester, Minnesota Amyloidosis Gordon D. Ginder, MD Professor, Internal Medicine, Director, Massey Cancer Center, Virginia Commonwealth University, Richmond, Virginia Microcytic and Hypochromic Anemias Jeffrey S. Ginsberg, MD Professor of Medicine, McMaster University, Member of Thrombosis and Atherosclerosis Research Institute, St. Joseph’s Healthcare Hamilton, Hamilton, Ontario, Canada Peripheral Venous Disease Geoffrey S. Ginsburg, MD, PhD Director, Duke Center for Applied Genomics and Precision Medicine; Professor of Medicine, Pathology and Biomedical Engineering, Duke University, Durham, North Carolina Applications of Molecular Technologies to Clinical Medicine Michael Glogauer, DDS, PhD Professor, Faculty of Dentistry, University of Toronto, Toronto, Ontario, Canada Disorders of Phagocyte Function John W. Gnann, Jr., MD Professor of Medicine, Department of Medicine, Division of Infectious Diseases, Medical University of South Carolina, Charleston, South Carolina Mumps Matthew R. Golden, MD, MPH Professor of Medicine, University of Washington, Director, HIV/STD Program, Public Health–Seattle & King County, Seattle, Washington Neisseria Gonorrhoeae Infections Lee Goldman, MD Harold and Margaret Hatch Professor, Executive Vice President and Dean of the Faculties of Health Sciences and Medicine, Chief Executive, Columbia University Medical Center, Columbia University, New York, New York Approach to Medicine, the Patient, and the Medical Profession: Medicine as a Learned and Humane Profession; Approach to the Patient with Possible Cardiovascular Disease Ellie J.C. Goldstein, MD Clinical Professor of Medicine, David Geffen School of Medicine at University of California Los Angeles, Los Angeles, California; Director, R.M. Alden Research Laboratory, Santa Monica, California Diseases Caused by Non–Spore-Forming Anaerobic Bacteria

Eduardo H. Gotuzzo, MD Professor of Medicine, Director, Alexander von Humboldt Tropical Medicine Institute, Universidad Peruana Cayetano Heredia; Chief Physician, Department of Infectious, Tropical, and Dermatologic Diseases, National Hospital Cayetano Heredia, Lima, Peru Cholera and Other Vibrio Infections; Liver, Intestinal, and Lung Fluke Infections Deborah Grady, MD, MPH Professor of Medicine, University of California San Francisco, San Francisco, California Menopause Leslie C. Grammer, MD Professor of Medicine, Northwestern University Feinberg School of Medicine; Attending Physician, Northwestern Memorial Hospital, Chicago, Illinois Drug Allergy F. Anthony Greco, MD Medical Director, Sarah Cannon Cancer Center, Nashville, Tennessee Cancer of Unknown Primary Origin Harry B. Greenberg, MD Professor, Departments of Medicine and Microbiology and Immunology, Stanford University School of Medicine, Stanford, California Rotaviruses, Noroviruses, and Other Gastrointestinal Viruses Steven A. Greenberg, MD Associate Professor of Neurology, Harvard Medical School; Associate Neurologist, Brigham and Women’s Hospital, Boston, Massachusetts Inflammatory Myopathies Robert C. Griggs, MD Professor of Neurology, Medicine, Pediatrics, and Pathology and Laboratory Medicine, University of Rochester School of Medicine and Dentistry, Rochester, New York Approach to the Patient with Neurologic Disease Lev M. Grinberg, MD, PhD Professor, Chief, Department of Pathology, Ural Medical University; Chief Researcher of the Ural Scientific Research Institute of Phthisiopulmonology, Chief Pathologist of Ekaterinburg, Ekaterinburg, Russia Anthrax Daniel Grossman, MD Vice President for Research, Ibis Reproductive Health, Oakland, California; Assistant Clinical Professor, Bixby Center for Global Reproductive Health, Department of Obstetrics, Gynecology and Reproductive Sciences, University of California San Francisco, San Francisco, California Contraception Lisa M. Guay-Woodford, MD Hudson Professor of Pediatrics, The George Washington University; Director, Center for Translational Science, Director, Clinical and Translational Institute at Children’s National, Children’s National Health System, Washington, D.C. Hereditary Nephropathies and Developmental Abnormalities of the Urinary Tract

Contributors Richard L. Guerrant, MD Thomas H. Hunter Professor of International Medicine, Founding Director, Center for Global Health, Division of Infectious Diseases and International Health, University of Virginia School of Medicine, University of Virginia Health Sciences Center, Charlottesville, Virginia Cryptosporidiosis Roy M. Gulick, MD, MPH Gladys and Roland Harrison Professor of Medicine, Medicine/Infectious Diseases, Weill Cornell Medical College; Attending Physician, New York– Presbyterian Hospital, New York, New York Antiretrovial Therapy of HIV/AIDS Klaus D. Hagspiel, MD Professor of Radiology, Medicine, and Pediatrics, Chief, Noninvasive Cardiovascular Imaging, University of Virginia Health System, Charlottesville, Virginia Noninvasive Cardiac Imaging John D. Hainsworth, MD Chief Scientific Officer, Sarah Cannon Research Institute, Nashville, Tennessee Cancer of Unknown Primary Origin Anders Hamsten, MD, PhD Professor of Cardiovascular Diseases, Center for Molecular Medicine and Department of Cardiology, Karolinska University Hospital, Department of Medicine, Karolinska Institute, Stockholm, Sweden Atherosclerosis, Thrombosis, and Vascular Biology Kenneth R. Hande, MD Professor of Medicine and Pharmacology, Vanderbilt/Ingram Cancer Center, Vanderbilt University School of Medicine, Nashville, Tennessee Neuroendocrine Tumors and the Carcinoid Syndrome H. Hunter Handsfield, MD Professor Emeritus of Medicine, University of Washington Center for AIDS and STD, Seattle, Washington Neisseria Gonorrhoeae Infections Göran K. Hansson, MD, PhD Professor of Cardiovascular Research, Center for Molecular Medicine at Karolinska University Hospital, Department of Medicine, Karolinska Institute, Stockholm, Sweden Atherosclerosis, Thrombosis, and Vascular Biology Raymond C. Harris, MD Professor of Medicine, Ann and Roscoe R. Robinson Chair in Nephrology, Chief, Division of Nephrology, Vanderbilt University School of Medicine, Nashville, Tennessee Diabetes and the Kidney Stephen Crane Hauser, MD Associate Professor of Medicine, Internal Medicine, Division of Gastroenterology and Hepatology, Mayo Clinic College of Medicine, Rochester, Minnesota Vascular Diseases of the Gastrointestinal Tract Frederick G. Hayden, MD Stuart S. Richardson Professor of Clinical Virology and Professor of Medicine, University of Virginia School of Medicine; Staff Physician, University of Virginia Health System, Charlottesville, Virginia Influenza Douglas C. Heimburger, MD, MS Professor of Medicine, Associate Director for Education and Training, Vanderbilt University School of Medicine, Vanderbilt Institute for Global Health, Nashville, Tennessee Nutrition’s Interface with Health and Disease


Erik L. Hewlett, MD Professor of Medicine and of Microbiology, Immunology, and Cancer Biology, University of Virginia School of Medicine, University of Virginia Health System, Charlottesville, Virginia Whooping Cough and Other Bordetella Infections Richard J. Hift, PhD, MMed School of Clinical Medicine, University of KwaZulu-Natal, Durban, South Africa The Porphyrias David R. Hill, MD, DTM&H Professor of Medical Sciences, Director of Global Public Health, Frank H. Netter MD School of Medicine at Quinnipiac University, Hamden, Connecticut Giardiasis Nicholas S. Hill, MD Professor of Medicine, Tufts University School of Medicine; Chief, Division of Pulmonary, Critical Care, and Sleep Medicine, Tufts Medical Center, Boston, Massachusetts Respiratory Monitoring in Critical Care L. David Hillis, MD Professor and Chair, Department of Medicine, University of Texas Health Science Center at San Antonio, San Antonio, Texas Acute Coronary Syndrome: Unstable Angina and Non-ST Elevation Myocardial Infarction Jack Hirsh, CM, MD, DSc Professor Emeritus, McMaster University, Hamilton, Ontario, Canada Antithrombotic Therapy Steven M. Holland, MD Chief, Laboratory of Clinical Infectious Diseases, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, Maryland The Nontuberculous Mycobacteria Steven M. Hollenberg, MD Professor of Medicine, Cooper Medical School of Rowan University; Director, Coronary Care Unit, Cooper University Hospital, Camden, New Jersey Cardiogenic Shock Edward W. Hook III, MD Professor and Director, Division of Infectious Diseases, University of Alabama at Birmingham, Birmingham, Alabama Granuloma Inguinale (Donovanosis); Syphilis; Nonsyphilitic Treponematoses David J. Hunter, MBBS, MPH, ScD Vincent L. Gregory Professor of Cancer Prevention, Harvard School of Public Health; Professor of Medicine, Harvard Medical School, Brigham and Women’s Hospital, Boston, Massachusetts The Epidemiology of Cancer Khalid Hussain, MBChB, MD, MSc Developmental Endocrinology Research Group, Clinical and Molecular Genetics Unit, Institute of Child Health, University College London, Department of Paediatric Endocrinology, Great Ormond Street Hospital for Children, London, United Kingdom Hypoglycemia/Pancreatic Islet Cell Disorders Steven E. Hyman, MD Director, Stanley Center for Psychiatric Research, Broad Institute, Distinguished Service Professor of Stem Cell and Regenerative Biology, Harvard University, Cambridge, Massachusetts Biology of Addiction



Michael C. Iannuzzi, MD, MBA Chairman, Department of Internal Medicine, State University of New York Upstate Medical University, Syracuse, New York Sarcoidosis

Richard C. Jordan, DDS, PhD Professor of Oral Pathology, Pathology and Radiation Oncology, University of California San Francisco, San Francisco, California Diseases of the Mouth and Salivary Glands

Robert D. Inman, MD Professor of Medicine and Immunology, University of Toronto; Staff Rheumatologist, University Health Network, Toronto, Ontario, Canada The Spondyloarthropathies

Ralph F. Józefowicz, MD Professor, Neurology and Medicine, University of Rochester, Rochester, New York Approach to the Patient with Neurologic Disease

Sharon K. Inouye, MD, MPH Professor of Medicine, Harvard Medical School; Director, Aging Brain Center, Institute for Aging Research, Hebrew SeniorLife, Boston, Massachusetts Neuropsychiatric Aspects of Aging; Delirium or Acute Mental Status Change in the Older Patient

Stephen G. Kaler, MD Senior Investigator and Head, Section on Translational Neuroscience, Molecular Medicine Program, Eunice Kennedy Shriver National Institute of Child Health and Human Development, Bethesda, Maryland Wilson Disease

Geoffrey K. Isbister, MD, BSc Associate Professor, Clinical Toxicologist, Calvary Mater Newcastle, Callaghan, Senior Research Academic, School of Medicine and Public Health, University of Newcastle, New South Wales, Australia Envenomation Michael G. Ison, MD, MS Associate Professor in Medicine-Infectious Diseases and Surgery-Organ Transplantation, Northwestern University Feinberg School of Medicine, Chicago, Illinois Adenovirus Diseases Elias Jabbour, MD Associate Professor, Department of Leukemia, Division of Medicine, The University of Texas MD Anderson Cancer Center, Houston, Texas The Chronic Leukemias Michael R. Jaff, DO Professor of Medicine, Harvard Medical School, Chair, Institute for Heart, Vascular, and Stroke Care, Massachusetts General Hospital, Boston, Massachusetts Other Peripheral Arterial Diseases Joanna C. Jen, MD, PhD Professor of Neurology, University of California Los Angeles School of Medicine, Los Angeles, California Neuro-Ophthalmology; Smell and Taste; Hearing and Equilibrium Dennis M. Jensen, MD Professor of Medicine, David Geffen School of Medicine at University of California Los Angeles; Staff Physician, Medicine-GI, VA Greater Los Angeles Healthcare System; Key Investigator, Director, Human Studies Core & GI Hemostasis Research Unit, CURE Digestive Diseases Research Center, Los Angeles, California Gastrointestinal Hemorrhage Michael D. Jensen, MD Professor of Medicine, Endocrine Research Unit, Director, Obesity Treatment Research Program, Mayo Clinic, Rochester, Minnesota Obesity Robert T. Jensen, MD Chief, Cell Biology Section, Digestive Disease Branch, National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, Clinical Center, Bethesda, Maryland Pancreatic Neuroendocrine Tumors Stuart Johnson, MD Professor of Medicine, Loyola University Chicago Stritch School of Medicine; Associate Chief of Staff for Research, Edward Hines, Jr. VA Hospital, Hines, Illinois Clostridial Infections

Moses R. Kamya, MB ChB, MMed, MPH, PhD Chairman, Department of Medicine, Makerere University College of Health Sciences, Kampala, Uganda Malaria Louise W. Kao, MD Associate Professor of Emergency Medicine, Department of Emergency Medicine, Indiana University School of Medicine, Indianapolis, Indiana Chronic Poisoning: Trace Metals and Others Steven A. Kaplan, MD E. Darracott Vaughan, Jr. Professor of Urology, Chief, Institute for Bladder and Prostate Health, Weill Cornell Medical College; Director, Iris Cantor Men’s Health Center, NewYork–Presbyterian Hospital, New York, New York Benign Prostatic Hyperplasia and Prostatitis Daniel L. Kastner, MD, PhD Scientific Director, National Human Genome Research Institute, National Institutes of Health, Bethesda, Maryland The Systemic Autoinflammatory Diseases Sekar Kathiresan, MD Associate Professor in Medicine, Harvard Medical School; Director, Preventive Cardiology, Massachusetts General Hospital, Boston, Massachusetts The Inherited Basis of Common Diseases David A. Katzka, MD Professor of and Consultant in Medicine, Gastroenterology, Mayo Clinic, Rochester, Minnesota Diseases of the Esophagus Debra K. Katzman, MD Professor of Pediatrics, Senior Associate Scientist, The Research Institute, The Hospital for Sick Children and University of Toronto, Toronto, Ontario, Canada Adolescent Medicine Carol A. Kauffman, MD Professor of Internal Medicine, University of Michigan Medical School; Chief, Infectious Diseases Section, Veterans Affairs Ann Arbor Healthcare System, Ann Arbor, Michigan Histoplasmosis; Blastomycosis; Paracoccidioidomycosis; Cryptococcosis; Sporotrichosis; Candidiasis Kenneth Kaushansky, MD Senior Vice President for Health Sciences, Dean, School of Medicine, Stony Brook University, Stony Brook, New York Hematopoiesis and Hematopoietic Growth Factors Keith S. Kaye, MD, MPH Professor of Medicine, Division of Infectious Diseases, Wayne State University School of Medicine, Detroit, Michigan Diseases Caused by Acinetobacter and Stenotrophom*onas Species

Contributors Armand Keating, MD Professor of Medicine, Director, Division of Hematology, Epstein Chair in Cell Therapy and Transplantation, Professor, Institute of Biomaterials and Biomedical Engineering, University of Toronto, Toronto, Ontario, Canada Hematopoietic Stem Cell Transplantation Robin K. Kelley, MD Assistant Professor of Medicine, University of California San Francisco, Helen Diller Family Comprehensive Cancer Center, San Francisco, California Liver and Biliary Tract Cancers Morton Kern, MD Chief of Medicine, VA Long Beach Health Care System School of Medicine; Professor of Medicine, Associate Chief, Cardiology, University of California–Irvine, Irvine, California Catheterization and Angiography Gerald T. Keusch, MD Professor of Medicine and International Health and Public Health, Boston University School of Medicine, Boston, Massachusetts Shigellosis Fadlo R. Khuri, MD Professor and Chair, Hematology and Medical Oncology, Deputy Director, Winship Cancer Institute, Emory University, Atlanta, Georgia Lung Cancer and Other Pulmonary Neoplasms David H. Kim, MD Vice Chair of Education, Professor of Radiology, Section of Abdominal Imaging, University of Wisconsin School of Medicine and Public Health, Madison, Wisconsin Diagnostic Imaging Procedures in Gastroenterology


Kevin M. Korenblat, MD Associate Professor of Medicine, Department of Medicine, Washington University School of Medicine, St. Louis, Missouri Approach to the Patient with Jaundice or Abnormal Liver Tests Bruce R. Korf, MD, PhD Wayne H. and Sara Crews Finley Chair in Medical Genetics, Professor and Chair, Department of Genetics, University of Alabama at Birmingham, Birmingham, Alabama Principles of Genetics Neil J. Korman, MD, PhD Professor, Dermatology, Case Western Reserve University School of Medicine, University Hospitals Case Medical Center, Cleveland, Ohio Macular, Papular, Vesiculobullous, and Pustular Diseases Mark G. Kortepeter, MD, MPH Associate Dean for Research, Associate Professor of Preventive Medicine and Medicine, Consultant to the Army Surgeon General for Biodefense; Office of the Dean, Edward Hébert School of Medicine, Uniformed Services University of the Health Sciences, Bethesda, Maryland Bioterrorism Joseph A. Kovacs, MD Senior Investigator and Head, AIDS Section, Critical Care Medicine Department, National Institutes of Health, Bethesda, Maryland Pneumocystis Pneumonia Thomas O. Kovacs, MD Professor of Medicine, Division of Digestive Diseases, David Geffen School of Medicine at University of California Los Angeles, Los Angeles, California Gastrointestinal Hemorrhage

Matthew Kim, MD Instructor of Medicine, Harvard Medical School; Associate Physician, Brigham and Women’s Hospital, Boston, Massachusetts Thyroid

Monica Kraft, MD Professor of Medicine, Duke University School of Medicine; Chief, Division of Pulmonary, Allergy, and Critical Care Medicine, Duke University Medical Center, Durham, North Carolina Approach to the Patient with Respiratory Disease

Louis V. Kirchhoff, MD, MPH Professor, Departments of Internal Medicine (Infectious Diseases) and Epidemiology, University of Iowa Health Care; Staff Physician, Medical Service, Department of Veterans Affairs Medical Center, Iowa City, Iowa Chagas Disease

Christopher M. Kramer, MD Ruth C. Heede Professor of Cardiology, Professor of Radiology, Director, Cardiovascular Imaging Center, University of Virginia Health System, Charlottesville, Virginia Noninvasive Cardiac Imaging

David S. Knopman, MD Professor of Neurology, Mayo Clinic College of Medicine, Rochester, Minnesota Regional Cerebral Dysfunction: Higher Mental Function; Alzheimer Disease and Other Dementias

Donna M. Krasnewich, MD, PhD Program Director, National Institute of General Medical Sciences, National Institutes of Health, Bethesda, Maryland The Lysosomal Storage Diseases

Tamsin A. Knox, MD, MPH Associate Professor of Medicine, Nutrition/Infection Unit, Tufts University School of Medicine, Boston, Massachusetts Gastrointestinal Manifestions of HIV and AIDS D.P. Kontoyiannis, MD, ScD Professor, Department of Infectious Diseases, Infection Control and Employee Health, The University of Texas MD Anderson Cancer Center, Houston, Texas Mucormycosis; Mycetoma Barbara S. Koppel, MD Professor of Clinical Neurology, New York Medical College, Chief of Neurology, Metropolitan Hospital Center, New York City Health and Hospital Corporation, New York, New York Nutritional and Alcohol-Related Neurologic Disorders

Peter J. Krause, MD Senior Research Scientist in Epidemiology, Medicine, and Pediatrics, Yale School of Public Health and Yale School of Medicine, New Haven, Connecticut Babesiosis and Other Protozoan Diseases John F. Kuemmerle, MD Chair, Division of Gastroenterology, Hepatology, and Nutrition, Professor of Medicine, and Physiology and Biophysics, Center for Digestive Health, Virginia Commonwealth University, Richmond, Virginia Inflammatory and Anatomic Diseases of the Intestine, Peritoneum, Mesentery, and Omentum Ernst J. Kuipers, MD, PhD Professor of Medicine, Department of Gastroenterology and Hepatology, Chief Executive Officer, Erasmus MC University Medical Center, Rotterdam, The Netherlands Acid Peptic Disease



Paul W. Ladenson, MD Professor of Medicine, Pathology, Oncology, and Radiology and Radiological Sciences, John Eager Howard Professor of Endocrinology and Metabolism, University Distinguished Service Professor, The Johns Hopkins University School of Medicine; Physician and Division Director, The Johns Hopkins Hospital, Baltimore, Maryland Thyroid Daniel Laheru, MD Ian T. MacMillan Professorship in Clinical Pancreatic Research, Medical Oncology, Johns Hopkins University School of Medicine, Baltimore, Maryland Pancreatic Cancer Donald W. Landry, MD, PhD Samuel Bard Professor of Medicine, Chair, Department of Medicine, Physician-in-Chief, NewYork-Presbyterian Hospital/Columbia University Medical Center, New York, New York Approach to the Patient with Renal Disease Anthony E. Lang, MD Director, Division of Neurology, Jack Clark Chair for Research in Parkinson’s Disease, University of Toronto; Director, Morton and Gloria Shulman Movement Disorders Clinic and the Edmond J. Safra Program in Parkinson’s Disease and the Lily Safra Chair in Movement Disorders, Toronto Western Hospital, Toronto, Ontario, Canada Parkinsonism; Other Movement Disorders

Gary R. Lichtenstein, MD Professor of Medicine, Perelman School of Medicine at the University of Pennsylvania, Director, Center for Inflammatory Bowel Disease, University of Pennsylvania, Philadelphia, Pennsylvania Inflammatory Bowel Disease Henry W. Lim, MD Chairman and C.S. Livingood Chair, Department of Dermatology, Henry Ford Hospital; Senior Vice President for Academic Affairs, Henry Ford Health System, Detroit, Michigan Eczemas, Photodermatoses, Papulosquamous (Including Fungal) Diseases, and Figurate Erythemas Aldo A.M. Lima, MD, PhD Professor of Medicine and Pharmacology, School of Medicine, Federal University of Ceará, Fortaleza, Ceará, Brazil Cryptosporidiosis; Amebiasis Geoffrey S.F. Ling, MD, PhD Professor of Neurology, Uniformed Services University of the Health Sciences, Bethesda, Maryland Traumatic Brain Injury and Spinal Cord Injury William C. Little, MD Patrick Lehan Professor of Cardiovascular Medicine, Chair, Department of Medicine, University of Mississippi Medical Center, Jackson, Mississippi Pericardial Diseases

Richard A. Lange, MD, MBA President and Dean, Paul L. Foster School of Medicine, Texas Tech University Health Sciences Center El Paso, El Paso, Texas Acute Coronary Syndrome: Unstable Angina and Non-ST Elevation Myocardial Infarction

Donald M. Lloyd-Jones, MD, ScM Senior Associate Dean, Chair, Department of Preventive Medicine, Eileen M. Foell Professor of Preventive Medicine and Medicine, Northwestern University Feinberg School of Medicine, Chicago, Illinois Epidemiology of Cardiovascular Disease

Frank A. Lederle, MD Core Investigator, Center for Chronic Disease Outcomes Research, Minneapolis VA Medical Center; Professor of Medicine, University of Minnesota School of Medicine, Minneapolis, Minnesota Diseases of the Aorta

Bennett Lorber, MD Thomas M. Durant Professor of Medicine and Professor of Microbiology and Immunology, Temple University School of Medicine, Philadelphia, Pennsylvania Listeriosis

Thomas H. Lee, MD, MSc Senior Physician, Department of Medicine, Brigham and Women’s Hospital; Chief Medical Officer, Press Ganey, Boston, Massachusetts Using Data for Clinical Decisions

Donald E. Low, MD† Nonpneumococcal Streptococcal Infections, Rheumatic Fever

William M. Lee, MD Professor of Internal Medicine, University of Texas Southwestern Medical Center, Dallas, Texas Toxin- and Drug-Induced Liver Disease James E. Leggett, MD Associate Professor, Department of Medicine, Oregon Health and Science University; Infectious Diseases, Department of Medical Education, Providence Portland Medical Center, Portland, Oregon Approach to Fever or Suspected Infection in the Normal Host Stuart Levin, MD Professor of Medicine, Emeritus Chairman, Department of Medicine, Rush University Medical Center, Chicago, Illinois Zoonoses Stephanie M. Levine, MD Professor of Medicine, Division of Pulmonary Diseases and Critical Care Medicine, The University of Texas Health Science Center San Antonio, South Texas Veterans Health Care System, San Antonio, Texas Alveolar Filling Disorders

Daniel R. Lucey, MD, MPH Adjunct Professor, Microbiology and Immunology, Georgetown University Medical Center, Washington, D.C. Anthrax James R. Lupski, MD, PhD Cullen Professor of Molecular and Human Genetics, Professor of Pediatrics, Baylor College of Medicine and Texas Children’s Hospital, Houston, Texas Gene, Genomic, and Chromosomal Disorders Jeffrey M. Lyness, MD Senior Associate Dean for Academic Affairs, Professor of Psychiatry and Neurology, University of Rochester School of Medicine and Dentistry, Rochester, New York Psychiatric Disorders in Medical Practice Bruce W. Lytle, MD Chair, Heart and Vascular Institute, Professor of Surgery, Thoracic and Cardiovascular Surgery, Cleveland Clinic, Cleveland, Ohio Interventional and Surgical Treatment of Coronary Artery Disease


Contributors C. Ronald MacKenzie, MD Assistant Attending Physician, Department of Medicine-Rheumatology, C. Ronald MacKenzie Chair in Ethics and Medicine, Hospital for Special Surgery, Associate Professor of Clinical Medicine and Medical Ethics, Weill Cornell Medical College of Cornell University, New York, New York Surgical Treatment of Joint Disease Harriet L. MacMillan, MD, MSc Professor, Departments of Psychiatry and Behavioural Neurosciences, and Pediatrics, Chedoke Health Chair in Child Psychiatry, Offord Centre for Child Studies, McMaster University, Hamilton, Ontario, Canada Intimate Partner Violence Robert D. Madoff, MD Professor of Surgery, Stanley M. Goldberg, MD, Chair, Colon and Rectal Surgery, University of Minnesota, Minneapolis, Minnesota Diseases of the Rectum and Anus Frank Maldarelli, MD, PhD Head, Clinical Retrovirology Section, HIV Drug Resistance Program, National Cancer Institute, National Institutes of Health, Bethesda, Maryland Biology of Human Immunodeficiency Viruses Atul Malhotra, MD Chief of Pulmonary and Critical Care, Kenneth M. Moser Professor of Medicine, Director of Sleep Medicine, University of California San Diego, La Jolla, California Disorders of Ventilatory Control Mark J. Manary, MD Helene B. Roberson Professor of Pediatrics, Washington University School of Medicine; Attending Physician, St. Louis Children’s Hospital, St. Louis, Missouri; Adjunct Professor, Children’s Nutrition Research Center, Baylor College of Medicine, Houston, Texas; Senior Lecturer in Community Health, University of Malawi College of Medicine, Blantyre, Malawi Protein-Energy Malnutrition Donna Mancini, MD Professor of Medicine, Department of Medicine, Division of Cardiology, Columbia University College of Physicians and Surgeons, Center for Advanced Cardiac Care, Columbia University Medical Center, New York, New York Cardiac Transplantation Lionel A. Mandell, MD Professor of Medicine, Faculty of Health Sciences, McMaster University, Hamilton, Ontario, Canada Streptococcus Pneumoniae Infections Peter Manu, MD Professor of Medicine and Psychiatry, Hofstra North Shore–LIJ School of Medicine at Hofstra University, Hempstead, New York; Adjunct Professor of Clinical Medicine, Psychiatry and Behavioral Sciences, Albert Einstein College of Medicine, Bronx, New York; Director of Medical Services, Zucker Hillside Hospital, Glen Oaks, New York Medical Consultation in Psychiatry Ariane Marelli, MD, MPH Professor of Medicine, McGill University, Director, McGill Adult Unit for Congenital Heart Disease, Associate Director, Academic Affairs and Research, Cardiology, McGill University Health Centre, Montreal, Québec, Canada Congenital Heart Disease in Adults Xavier Mariette, MD, PhD Professor, Rheumatology, Université Paris-Sud, AP-HP, Le Kremlin Bicêtre, France Sjögren Syndrome


Andrew R. Marks, MD Wu Professor and Chair, Department of Physiology and Cellular Biophysics, Founding Director, Helen and Clyde Wu Center for Molecular Cardiology, Columbia University College of Physicians and Surgeons, New York, New York Cardiac Function and Circulatory Control Kieren A. Marr, MD Professor of Medicine and Oncology, The Johns Hopkins University, Director, Transplant and Oncology Infectious Diseases, Baltimore, Maryland Approach to Fever and Suspected Infection in the Compromised Host Thomas J. Marrie, MD Dean, Faculty of Medicine, Dalhousie University; Professor of Medicine, Capital District Health Authority, Halifax, Nova Scotia, Canada Legionella Infections Paul Martin, MD Professor of Medicine and Chief, Division of Hepatology, Miller School of Medicine, University of Miami, Miami, Florida Approach to the Patient with Liver Disease Joel B. Mason, MD Professor of Medicine and Nutrition, Tufts University; Staff Physician, Divisions of Gastroenterology and Clinical Nutrition, Tufts Medical Center, Boston, Massachusetts Vitamins, Trace Minerals, and Other Micronutrients Henry Masur, MD Chief, Critical Care Medicine Department, Clinical Center, National Institutes of Health, Bethesda, Maryland Infectious and Metabolic Complications of HIV and AIDS Eric L. Matteson, MD, MPH Professor of Medicine, Mayo Clinic College of Medicine, Consultant, Divisions of Rheumatology and Epidemiology, Mayo Clinic, Rochester, Minnesota Infections of Bursae, Joints, and Bones Michael A. Matthay, MD Professor, Departments of Medicine and Anesthesia, University of California San Francisco, San Francisco, California Acute Respiratory Failure Toby A. Maurer, MD Professor of Dermatology, University of California San Francisco; Chief of Dermatology, San Francisco General Hospital, San Francisco, California Skin Manifestations in Patients with Human Immunodeficiency Virus Infection Emeran A. Mayer, MD, PhD Professor of Medicine, Physiology, and Psychiatry, Division of Digestive Diseases, Department of Medicine, University of California Los Angeles, Los Angeles, California Functional Gastrointestinal Disorders: Irritable Bowel Syndrome, Dyspepsia, Chest Pain of Presumed Esophageal Origin, and Heartburn Stephan A. Mayer, MD Director, Institute for Critical Care Medicine, Icahn School of Medicine at Mount Sinai, New York, New York Hemorrhagic Cerebrovascular Disease Stephen A. McClave, MD Professor of Medicine, Director of Clinical Nutrition, University of Louisville School of Medicine, Louisville, Kentucky Enteral Nutrition F. Dennis McCool, MD Professor of Medicine, The Warren Alpert Medical School of Brown University; Medical Director of Sleep Center, Memorial Hospital of Rhode Island, Pawtucket, Rhode Island Diseases of the Diaphragm, Chest Wall, Pleura, and Mediastinum



Charles E. McCulloch, PhD Professor of Biostatistics, Department of Epidemiology and Biostatistics, University of California San Francisco, San Francisco, California Statistical Interpretation of Data

Ernest Moy, MD, MPH Medical Officer, Center for Quality Improvement and Patient Safety Agency for Healthcare Research and Quality, Rockville, Maryland Measuring Health and Health Care

William J. McKenna, MD Professor of Cardiology, Institute of Cardiovascular Science, University College London, London, United Kingdom Diseases of the Myocardium and Endocardium

Atis Muehlenbachs, MD, PhD Infectious Diseases Pathology Branch, Centers for Disease Control and Prevention, Atlanta, Georgia Leptospirosis

Vallerie McLaughlin, MD Kim A. Eagle, MD, Endowed Professor of Cardiovascular Medicine, Director, Pulmonary Hypertension Program, University of Michigan, Ann Arbor, Michigan Pulmonary Hypertension

Andrew H. Murr, MD Professor and Chairman, Roger Boles, MD Endowed Chair in Otolaryngology Education, Department of Otolaryngology-Head and Neck Surgery, University of California San Francisco School of Medicine, San Francisco, California Approach to the Patient with Nose, Sinus, and Ear Disorders

John J.V. McMurray, MB, MD Professor of Cardiology, Institute of Cardiovascular and Medical Sciences, University of Glasgow, Glasgow, Scotland, United Kingdom Heart Failure: Management and Prognosis Kenneth R. McQuaid, MD Professor of Clinical Medicine, Marvin H. Sleisenger Endowed Chair, Vice Chairman, University of California San Francisco; Chief, Medical Services and Gastroenterology, San Francisco VA Medical Center, San Francisco, California Approach to the Patient with Gastrointestinal Disease Marc Michel, MD Professor of Internal Medicine, Head of the Unit of Internal Medicine at Henri Mondor University Hospital, National Referral Center for Adult’s Immune Cytopenias, Creteil, France Autoimmune and Intravascular Hemolytic Anemias Jonathan W. Mink, MD, PhD Frederick A. Horner, MD Endowed Professor in Pediatric Neurology, Professor of Neurology, Neurobiology & Anatomy, Brain & Cognitive Sciences, and Pediatrics, Chief, Division of Child Neurology, Vice Chair, Department of Neurology, University of Rochester, Rochester, New York Congenital, Developmental, and Neurocutaneous Disorders William E. Mitch, MD Gordon A. Cain Chair in Nephrology, Director of Nephrology, Baylor College of Medicine, Houston, Texas Chronic Kidney Disease Mark E. Molitch, MD Martha Leland Sherwin Professor of Endocrinology, Division of Endocrinology, Metabolism, and Molecular Medicine, Northwestern University Feinberg School of Medicine, Chicago, Illinois Neuroendocrinology and the Neuroendocrine System; Anterior Pituitary Bruce A. Molitoris, MD Professor of Medicine, and Cellular and Integrative Physiology Director, Indiana Center for Biological Microscopy, Indiana University, Indianapolis, Indiana Acute Kidney Injury Jose G. Montoya, MD Professor of Medicine, Division of Infectious Disease and Geographic Medicine, Stanford University School of Medicine, Stanford, California; Director, Palo Alto Medical Foundation Toxoplasma Serology Laboratory, National Reference Center for the Study and Diagnosis of Toxoplasmosis, Palo Alto, California Toxoplasmosis Alison Morris, MD, MS Associate Professor of Medicine, Clinical Translational Science, and Immunology, Division of Pulmonary, Allergy, and Critical Care Medicine, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania Pulmonary Manifestations of Human Immunodeficiency Virus and Acquired Immunodeficiency Syndrome

Daniel M. Musher, MD Professor of Medicine, Molecular Virology, and Microbiology, Distinguished Service Professor, Baylor College of Medicine, Infectious Disease Section, Michael E. DeBakey Veterans Affairs Medical Center, Houston, Texas Overview of Pneumonia Robert J. Myerburg, MD Professor of Medicine and Physiology, Division of Cardiology, Department of Medicine, American Heart Association Chair in Cardiovascular Research, University of Miami Miller School of Medicine, Miami, Florida Approach to Cardiac Arrest and Life-Threatening Arrhythmias Sandesh C.S. Nagamani, MD Assistant Professor, Department of Molecular and Human Genetics, Director, Clinic for Metabolic and Genetic Disorders of Bone, Baylor College of Medicine and Texas Children’s Hospital, Houston, Texas Gene, Genomic, and Chromosomal Disorders Stanley J. Naides, MD Medical Director and Interim Scientific Director, Immunology, Quest Diagnostics Nichols Institute, San Juan Capistrano, California Arboviruses Causing Fever and Rash Syndromes Yoshifumi Naka, MD, PhD Professor of Surgery, Department of Surgery, Columbia University College of Physicians and Surgeons, New York, New York Cardiac Transplantation Theodore E. Nash, MD Principal Investigator, Clinical Parasitology Section, Laboratory of Parasitic Diseases, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, Maryland Giardiasis Avindra Nath, MD Chief, Section of Infections of the Nervous System, National Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda, Maryland Cytomegalovirus, Epstein-Barr Virus, and Slow Virus Infections of the Central Nervous System; Neurologic Complications of Human Immunodeficiency Virus Infection; Meningitis: Bacterial, Viral, and Other; Brain Abscess and Parameningeal Infections Eric G. Neilson, MD Vice President for Medical Affairs and Lewis Landsberg Dean, Northwestern University Feinberg School of Medicine, Northwestern Memorial Hospital, Chicago, Illinois Tubulointerstitial Nephritis

Contributors Lawrence S. Neinstein, MD Professor of Pediatrics and Medicine, Keck School of Medicine of USC; Executive Director, Engemann Student Health Center, Division Head of College Health, Assistant Provost, Student Health and Wellness, University of Southern California, Los Angeles, California Adolescent Medicine Lewis S. Nelson, MD Professor of Emergency Medicine, Director, Fellowship in Medical Toxicology, New York University School of Medicine; Attending Physician, New York University Langone Medical Center and Bellevue Hospital Center, New York, New York Acute Poisoning Eric J. Nestler, MD, PhD Nash Family Professor and Chair, Department of Neuroscience, Director, The Friedman Brain Institute, Icahn School of Medicine at Mount Sinai, New York, New York Biology of Addiction Anne B. Newman, MD, MPH Professor of Epidemiology, The University of Pittsburgh Graduate School of Public Health; Chair, Department of Epidemiology, Director, University of Pittsburgh Center for Aging and Population Health, Pittsburgh, Pennsylvania Epidemiology of Aging: Implications of the Aging of Society Thomas B. Newman, MD, MPH Professor, Epidemiology & Biostatistics and Pediatrics, University of California San Francisco, San Francisco, California Statistical Interpretation of Data William L. Nichols, MD Associate Professor, Medicine and Laboratory Medicine, Mayo Clinic College of Medicine; Staff Physician, Special Coagulation Laboratory, Comprehensive Hemophilia Center, and Coagulation Clinic, Mayo Clinic, Rochester, Minnesota Von Willebrand Disease and Hemorrhagic Abnormalities of Platelet and Vascular Function Lindsay E. Nicolle, MD Professor of Internal Medicine and Medical Microbiology, University of Manitoba, Health Sciences Centre, Winnipeg, Manitoba, Canada Approach to the Patient with Urinary Tract Infection Lynnette K. Nieman, MD Senior Investigator, Program on Reproductive and Adult Endocrinology, Eunice Kennedy Shriver National Institute of Child Health and Human Development, Bethesda, Maryland Approach to the Patient with Endocrine Disease; Adrenal Cortex; Polyglandular Disorders Dennis E. Niewoehner, MD Professor of Medicine, University of Minnesota; Staff Physician, Minneapolis Veterans Affairs Health Care System, Minneapolis, Minnesota Chronic Obstructive Pulmonary Disease S. Ragnar Norrby, MD, PhD Director General, Swedish Institute for Infectious Disease Control, Solna, Sweden Approach to the Patient with Urinary Tract Infection Susan O’Brien, MD Professor, Department of Leukemia, Division of Medicine, The University of Texas MD Anderson Cancer Center, Houston, Texas The Chronic Leukemias Christopher M. O’Connor, MD Professor of Medicine and Chief, Division of Cardiology, Director, Duke Heart Center, Durham, North Carolina Heart Failure: Pathophysiology and Diagnosis


Francis G. O’Connor, MD, MPH Professor and Chair, Military and Emergency Medicine, Medical Director, Uniformed Services University Consortium for Health and Military Performance, Bethesda, Maryland Disorders Due to Heat and Cold; Rhabdomyolysis Patrick G. O’Connor, MD, MPH Professor and Chief, General Internal Medicine, Yale University School of Medicine, New Haven, Connecticut Alcohol Abuse and Dependence James R. O’Dell, MD Bruce Professor and Vice Chair of Internal Medicine, Chief, Division of Rheumatology, University of Nebraska Medical Center and Omaha VA Nebraska–Western Iowa Health Care System, Omaha, Nebraska Rheumatoid Arthritis Anne E. O’Donnell, MD Professor of Medicine, Chief, Division of Pulmonary, Critical Care, and Sleep Medicine, Georgetown University Medical Center, Washington, D.C. Bronchiectasis, Atelectasis, Cysts, and Localized Lung Disorders Jae K. Oh, MD Professor of Medicine, Director, Echocardiography Core Laboratory and Pericardial Clinic, Division of Cardiovascular Diseases, Co-Director, Integrated Cardiac Imaging, Division of Cardiovascular Diseases, Mayo Clinic, Rochester, Minnesota Pericardial Diseases Jeffrey E. Olgin, MD Gallo-Chatterjee Distinguished Professor of Medicine, Chief, Division of Cardiology, Co-Director, Heart and Vascular Center, University of California San Francisco, San Francisco, California Approach to the Patient with Suspected Arrhythmia Walter A. Orenstein, MD Professor of Medicine, Pediatrics, and Global Health, Emory University School of Medicine, Atlanta, Georgia Immunization Douglas R. Osmon, MD, MPH Professor of Medicine, Mayo Clinic College of Medicine; Consultant, Division of Infectious Diseases, Mayo Clinic, Rochester, Minnesota Infections of Bursae, Joints, and Bones Catherine M. Otto, MD J. Ward Kennedy-Hamilton Endowed Chair in Cardiology, Professor of Medicine, University of Washington School of Medicine; Director, Heart Valve Clinic, University of Washington Medical Center, Seattle, Washington Echocardiography Mark Papania, MD, MPH Medical Epidemiologist, Division of Viral Diseases, Measles, Mumps, Rubella, and Herpes Virus Laboratory Branch, Centers for Disease Control and Prevention, Atlanta, Georgia Measles Peter G. Pappas, MD Professor of Medicine, University of Alabama at Birmingham, Birmingham, Alabama Dematiaceous Fungal Infections Pankaj Jay Pasricha, MD Director, The Johns Hopkins Center for Neurogastroenterology; Professor of Medicine and Neurosciences, The Johns Hopkins School of Medicine; Professor of Innovation Management, Johns Hopkins Carey Business School, Baltimore, Maryland Gastrointestinal Endoscopy



David L. Paterson, MD Professor of Medicine, University of Queensland Centre for Clinical Research, Royal Brisbane and Women’s Hospital Campus, Brisbane, Queensland, Australia Infections Due to Other Members of the Enterobacteriaceae, Including Management of Multidrug Resistant Strains Carlo Patrono, MD Professor and Chair of Pharmacology, Department of Pharmacology, Catholic University School of Medicine, Rome, Italy Prostaglandin, Aspirin, and Related Compounds Jean-Michel Pawlotsky, MD, PhD Professor of Medicine, The University of Paris-Est; Director, National Reference Center for Viral Hepatitis B, C, and Delta and Department of Virology, Henri Mondor University Hospital; Director, Department of Molecular Virology and Immunology, Institut Mondor de Recherche Biomédicale, Créteil, France Acute Viral Hepatitis; Chronic Viral and Autoimmune Hepatitis Richard D. Pearson, MD Professor of Medicine and Pathology, University of Virginia School of Medicine and University of Virginia Health System, Charlottesville, Virginia Antiparasitic Therapy Trish M. Perl, MD, MSc Professor of Medicine and Pathology, The Johns Hopkins School of Medicine; Professor of Epidemiology, Johns Hopkins Bloomberg School of Public Health; Infectious Diseases Specialist and Senior Epidemiologist, The Johns Hopkins Hospital and Health System, Baltimore, Maryland Enterococcal Infections Adam Perlman, MD, MPH Associate Professor, Department of Medicine, Duke University Medical Center; Executive Director, Duke Integrative Medicine, Duke University Health System, Durham, North Carolina Complementary and Alternative Medicine William A. Petri, Jr., MD, PhD Wade Hampton Frost Professor, Departments of Medicine, Pathology, Microbiology, Immunology, and Cancer Biology, School of Medicine, University of Virginia; Chief, Division of Infectious Diseases and International Health, University of Virginia Hospitals, Charlottesville, Virginia Relapsing Fever and Other Borrelia Infections; African Sleeping Sickness; Amebiasis Marc A. Pfeffer, MD, PhD Dzau Professor of Medicine, Harvard Medical School; Senior Physician, Cardiovascular Division, Brigham and Women’s Hospital, Boston, Massachusetts Heart Failure: Management and Prognosis Perry J. Pickhardt, MD Professor of Radiology and Chief, Gastrointestinal Imaging, Section of Abdominal Imaging, University of Wisconsin School of Medicine and Public Health, Madison, Wisconsin Diagnostic Imaging Procedures in Gastroenterology David S. Pisetsky, MD, PhD Chief of Rheumatology, Medical Research Service, Durham VA Medical Center; Professor of Medicine and Immunology, Department of Medicine, Duke University Medical Center, Durham, North Carolina Laboratory Testing in the Rheumatic Diseases Marshall R. Posner, MD Professor of Medicine, Director of Head and Neck Medical Oncology, Director of the Office of Cancer Clinical Trials, The Tisch Cancer Institute, Icahn School of Medicine at Mount Sinai, New York, New York Head and Neck Cancer

Frank Powell, PhD Professor of Medicine, Chief of Physiology, University of California San Diego, La Jolla, California Disorders of Ventilatory Control Reed E. Pyeritz, MD, PhD William Smilow Professor of Medicine and Genetics and Vice Chair for Academic Affairs, Perelman School of Medicine at the University of Pennsylvania, Philadelphia, Pennsylvania Inherited Diseases of Connective Tissue Thomas C. Quinn, MD, MSc Associate Director for International Research, Head, Section of International HIV/AIDS Research, Division of Intramural Research, National Institute of Allergy and Infectious Diseases, National Institutes of Health; Professor of Medicine, Pathology, International Health, Molecular Microbiology and Immunology, and Epidemiology, The Johns Hopkins Medical Institutions, Baltimore, Maryland Epidemiology and Diagnosis of Human Immunodeficiency Virus Infection and Acquired Immunodeficiency Syndrome Jai Radhakrishnan, MD, MS Professor of Medicine, Division of Nephrology, Department of Medicine, Columbia University Medical Center; Associate Division Chief for Clinical Affairs, Division of Nephrology, New York-Presbyterian Hospital, New York, New York Glomerular Disorders and Nephrotic Syndromes Petros I. Rafailidis, MD, PhD, MSc Senior Researcher, Alfa Institute of Biomedical Sciences, Attending Physician, Department of Medicine and Hematology, Athens Medical Center, Athens Medical Group, Athens, Greece Pseudomonas and Related Gram-Negative Bacillary Infections Ganesh Raghu, MD Adjunct Professor of Medicine and Laboratory Medicine, University of Washington, Director, CENTER for Interstitial Lung Diseases at the University of Washington; Co-Director, Scleroderma Clinic, University of Washington Medical Center, Seattle, Washington Interstitial Lung Disease Margaret Ragni, MD, MPH Professor of Medicine and Clinical Translational Science, Department of Hematology/Oncology, University of Pittsburgh Medical Center; Director, Hemophilia Center of Western Pennsylvania, Pittsburgh, Pennsylvania Hemorrhagic Disorders: Coagulation Factor Deficiencies Srinivasa N. Raja, MD Professor of Anesthesiology and Neurology, Director, Division of Pain Medicine, The Johns Hopkins University School of Medicine, Baltimore, Maryland Pain S. Vincent Rajkumar, MD Professor of Medicine, Division of Hematology, Mayo Clinic, Rochester, Minnesota Plasma Cell Disorders Stuart H. Ralston, MB ChB, MD Professor of Rheumatology, Institute of Genetics and Molecular Medicine, Western General Hospital, The University of Edinburgh, Edinburgh, United Kingdom Paget Disease of Bone Didier Raoult, MD, PhD Professor, Aix Marseille Université, Faculté de Médecine; Chief, Hôpital de la Timone, Fédération de Microbiologie Clinique, Marseille, France Bartonella Infections; Rickettsial Infections

Contributors Robert W. Rebar, MD Professor, Department of Obstetrics and Gynecology, Western Michigan University Homer Stryker MD School of Medicine, Kalamazoo, Michigan Ovaries and Development; Reproductive Endocrinology and Infertility Annette C. Reboli, MD Founding Vice Dean, Professor of Medicine, Cooper Medical School of Rowan University, Cooper University Healthcare, Department of Medicine, Division of Infectious Diseases, Camden, New Jersey Erysipelothrix Infections


Karen Rosene-Montella, MD Professor and Vice Chair of Medicine, Director of Obstetric Medicine, The Warren Alpert Medical School of Brown University; Senior Vice President, Women’s Services and Clinical Integration, Lifespan Health System, Providence, Rhode Island Common Medical Problems in Pregnancy Philip J. Rosenthal, MD Professor, Department of Medicine, University of California San Francisco, San Francisco, California Malaria

K. Rajender Reddy, MD Professor of Medicine, Professor of Medicine in Surgery, Perelman School of Medicine at the University of Pennsylvania; Director of Hepatology, Director, Viral Hepatitis Center, Hospital of the University of Pennsylvania, Philadelphia, Pennsylvania Bacterial, Parasitic, Fungal, and Granulomatous Liver Diseases

Marc E. Rothenberg, MD, PhD Director, Division of Allergy and Immunology, Director, Cincinnati Center for Eosinophilic Disorders; Professor of Pediatrics, Cincinnati Children’s Hospital Medical Center, University of Cincinnati College of Medicine, Cincinnati, Ohio Eosinophilic Syndromes

Donald A. Redelmeier, MD Professor of Medicine, University of Toronto; Senior Scientist and Staff Physician, Sunnybrook Health Sciences Centre, Toronto, Ontario, Canada Postoperative Care and Complications

James A. Russell, MD Professor of Medicine, University of British Columbia; Associate Director, Intensive Care Unit, St. Paul’s Hospital, Vancouver, British Columbia, Canada Shock Syndromes Related to Sepsis

Susan E. Reef, MD Centers for Disease Control and Prevention, Atlanta, Georgia Rubella (German Measles)

Anil K. Rustgi, MD T. Grier Miller Professor of Medicine and Genetics, Chief of Gastroenterology, American Cancer Society; Professor, Perelman School of Medicine at the University of Pennsylvania, Philadelphia, Pennsylvania Neoplasms of the Esophagus and Stomach

Neil M. Resnick, MD Thomas P. Detre Endowed Chair in Gerontology and Geriatric Medicine, Professor of Medicine and Division Chief, Geriatrics, Associate Director, University of Pittsburgh Institute on Aging, University of Pittsburgh; Chief, Division of Geriatric Medicine and Gerontology, University of Pittsburgh Medical Center, Pittsburgh, Pennsylvania Incontinence David B. Reuben, MD Director, Multicampus Program in Geriatric Medicine and Gerontology; Chief, Division of Geriatrics, Archstone Professor of Medicine, David Geffen School of Medicine at University of California Los Angeles, Los Angeles, California Geriatric Assessment Emanuel P. Rivers, MD, MPH Professor and Vice Chairman of Emergency Medicine, Wayne State University; Senior Staff Attending, Critical Care and Emergency Medicine, Henry Ford Hospital, Detroit, Michigan Approach to the Patient with Shock Joseph G. Rogers, MD Professor of Medicine, Senior Vice Chief for Clinical Affairs, Division of Cardiology, Durham, North Carolina Heart Failure: Pathophysiology and Diagnosis Jean-Marc Rolain, PharmD, PhD Professor, Institut Hospitalo-Universitaire Méditerranée-Infection, Aix-Marseille Université, Marseille, France Bartonella Infections José R. Romero, MD Professor of Pediatrics, University of Arkansas for Medical Sciences, Horace C. Cabe Professor of Infectious Diseases; Director, Section of Pediatric Infectious Diseases, Arkansas Children’s Hospital, Little Rock, Arkansas Enteroviruses

Daniel E. Rusyniak, MD Professor of Emergency Medicine, Adjunct Professor of Neurology and Pharmacology and Toxicology, Department of Emergency Medicine, Indiana University School of Medicine, Indianapolis, Indiana Chronic Poisoning: Trace Metals and Others Robert A. Salata, MD Professor and Executive Vice Chair, Department of Medicine, Chief, Division of Infectious Diseases and HIV Medicine, Case Western Reserve University, University Hospitals Case Medical Center, Cleveland, Ohio Brucellosis Jane E. Salmon, MD Collette Kean Research Chair, Hospital for Special Surgery, Professor of Medicine, Weill Cornell Medical College, New York, New York Mechanisms of Immune-Mediated Tissue Injury Edsel Maurice T. Salvana, MD, DTM&H Associate Professor of Medicine, Section of Infectious Diseases, Department of Medicine, Philippine General Hospital; Director, Institute of Molecular Biology and Biotechnology, National Institutes of Health, University of the Philippines Manila, Manila, Philippines Brucellosis Renato M. Santos, MD Associate Professor, Cardiology, Wake Forest School of Medicine, Winston-Salem, North Carolina Vascular Disorders of the Kidney Michael N. Sawka, PhD Professor, School of Applied Physiology, Georgia Institute of Technology, Atlanta, Georgia Disorders Due to Heat and Cold Paul D. Scanlon, MD Professor of Medicine, Division of Pulmonary and Critical Care Medicine, Mayo Clinic, Rochester, Minnesota Respiratory Function: Mechanisms and Testing



Carla Scanzello, MD, PhD Assistant Professor of Medicine, Division of Rheumatology, Perelman School of Medicine at the University of Pennsylvania and Translational Musculoskeletal Research Center, Philadelphia Veterans Affairs Medical Center, Philadelphia, Pennsylvania Osteoarthritis Andrew I. Schafer, MD Professor of Medicine, Director, Richard T. Silver Center for Myeloproliferative Neoplasms, Weill Cornell Medical College, New York, New York Approach to Medicine, the Patient, and the Medical Profession: Medicine as a Learned and Humane Profession; Approach to the Patient with Bleeding and Thrombosis; Hemorrhagic Disorders: Disseminated Intravascular Coagulation, Liver Failure, and Vitamin K Deficiency; Thrombotic Disorders: Hypercoagulable States William Schaffner, MD Professor and Chair, Department of Preventive Medicine, Department of Health Policy; Professor of Medicine (Infectious Diseases), Vanderbilt University School of Medicine, Nashville, Tennessee Tularemia and Other Francisella Infections W. Michael Scheld, MD Bayer-Gerald L. Mandell Professor of Infectious Diseases, Professor of Medicine, Clinical Professor of Neurosurgery, Director, Pfizer Initiative in International Health, University of Virginia Health System, Charlottesville, Virginia Introduction to Microbial Disease: Host-Pathogen Interactions Manuel Schiff, MD Professor, Université Paris 7 Denis Diderot, Sorbonne Paris Cité, Head of Metabolic Unit/Reference Center for Inborn Errors of Metabolism, Robert Debré University Hospital, APHP, Paris, France hom*ocystinuria and Hyperhom*ocysteinemia Michael L. Schilsky, MD Associate Professor, Medicine and Surgery, Yale University School of Medicine, New Haven, Connecticut Wilson Disease Robert T. Schooley, MD Professor and Head, Division of Infectious Diseases, Executive Vice Chair for Academic Affairs, Department of Medicine, University of California San Diego, La Jolla, California Epstein-Barr Virus Infection David L. Schriger, MD, MPH Professor, Department of Emergency Medicine, University of California Los Angeles, Los Angeles, California Approach to the Patient with Abnormal Vital Signs Steven A. Schroeder, MD Distinguished Professor of Health and Healthcare and of Medicine, University of California San Francisco, San Francisco, California Socioeconomic Issues in Medicine Lynn M. Schuchter, MD Professor of Medicine, University of Pennsylvania; Chief, Hematology/ Oncology Division, Program Leader, Melanoma and Cutaneous Malignancies Program, Abramson Cancer Center, Hospital of the University of Pennsylvania, Philadelphia, Pennsylvania Melanoma and Nonmelanoma Skin Cancers Sam Schulman, MD, PhD Professor, Division of Hematology and Thromboembolism, Director of Clinical Thromboembolism Program, Department of Medicine, McMaster University, Hamilton, Ontario, Canada Antithrombotic Therapy

Lawrence B. Schwartz, MD, PhD Charles and Evelyn Thomas Professor of Medicine, Internal Medicine, Virginia Commonwealth University, Richmond, Virginia Systemic Anaphylaxis, Food Allergy, and Insect Sting Allergy Carlos Seas, MD Associate Professor of Medicine, Vice Director, Alexander von Humboldt Tropical Medicine Institute, Universidad Peruana Cayetano Heredia; Attending Physician, Department of Infectious, Tropical, and Dermatologic Diseases, National Hospital Cayetano Heredia, Lima, Peru Cholera and Other Vibrio Infections Steven A. Seifert, MD Professor of Emergency Medicine, University of New Mexico School of Medicine, Medical Director, New Mexico Poison and Drug Information Center, University of New Mexico Health Sciences Center, Albuquerque, New Mexico Envenomation Julian L. Seifter, MD Associate Professor of Medicine, Harvard Medical School; Senior Physician, Brigham and Women’s Hospital, Boston, Massachusetts Potassium Disorders; Acid-Base Disorders Duygu Selcen, MD Associate Professor of Neurology and Pediatrics, Department of Neurology, Mayo Clinic, Rochester, Minnesota Muscle Diseases Clay F. sem*nkovich, MD Herbert S. Gasser Professor and Chief, Division of Endocrinology, Metabolism and Lipid Research, Washington University School of Medicine, St. Louis, Missouri Disorders of Lipid Metabolism Carol E. Semrad, MD Professor of Medicine, The University of Chicago Medicine, GI Section, Chicago, Illinois Approach to the Patient with Diarrhea and Malabsorption Harry Shamoon, MD Professor of Medicine and Associate Dean for Clinical and Translational Research, Albert Einstein College of Medicine; Director, Harold and Muriel Block Institute for Clinical and Translational Research at Einstein and Montefiore, Bronx, New York Diabetes Mellitus James C. Shaw, MD Associate Professor, Department of Medicine, University of Toronto; Head, Division of Dermatology, Department of Medicine, Women’s College Hospital, Toronto, Ontario, Canada Examination of the Skin and an Approach to Diagnosing Skin Diseases Pamela J. Shaw, DBE, MBBS, MD Professor of Neurology, University of Sheffield, Consultant Neurologist, Royal Hallamshire Hospital, Sheffield, United Kingdom Amyotrophic Lateral Sclerosis and Other Motor Neuron Diseases Robert L. Sheridan, MD Associate Professor of Surgery, Burn Service Medical Director, Boston Shriners Hospital for Children, Massachusetts General Hospital, Division of Burns, Harvard Medical School, Boston, Massachusetts Medical Aspects of Injuries and Burns Stuart Sherman, MD Professor of Medicine and Radiology, Director of ERCP, Indiana University School of Medicine, Indianapolis, Indiana Diseases of the Gallbladder and Bile Ducts

Contributors Michael E. Shy, MD Professor of Neurology, Pediatrics, and Physiology, University of Iowa, Iowa City, Iowa Peripheral Neuropathies

Frederick S. Southwick, MD Professor of Medicine, Division of Infectious Diseases, University of Florida and VF Health, Gainesville, Florida Nocardiosis

Ellen Sidransky, MD Chief, Section on Molecular Neurogenetics, Medical Genetics Branch, National Human Genome Research Institute, National Institutes of Health, Bethesda, Maryland The Lysosomal Storage Diseases

Allen M. Spiegel, MD Dean, Albert Einstein College of Medicine, Bronx, New York Principles of Endocrinology; Polyglandular Disorders

Richard M. Siegel, MD, PhD Clinical Director, National Institute of Arthritis, Musculoskeletal, and Skin Diseases, National Institutes of Health, Bethesda, Maryland The Systemic Autoinflammatory Diseases Robert F. Siliciano, MD, PhD Professor, The Johns Hopkins University School of Medicine, Howard Hughes Medical Institute, Baltimore, Maryland Immunopathogenesis of Human Immunodeficiency Virus Infection Michael S. Simberkoff, MD Chief of Staff, VA New York Harbor Healthcare System; Professor of Medicine, NYU School of Medicine, New York, New York Haemophilus and Moraxella Infections David L. Simel, MD, MHS Professor of Medicine, Duke University; Chief, Medical Service, Durham Veterans Affairs Medical Center, Durham, North Carolina Approach to the Patient: History and Physical Examination Kamaljit Singh, MD Associate Professor of Medicine, Attending Physician, Infectious Diseases, Rush University Medical Center, Chicago, Illinois Zoonoses Karl Skorecki, MD Annie Chutick Professor in Medicine, Rappaport Faculty of Medicine and Research Institute, Technion–Israel Institute of Technology; Director, Medical and Research Development, Rambam Health Care Campus, Haifa, Israel Gene and Cell Therapy; Disorders of Sodium and Water Homeostasis Itzchak Slotki, MD Associate Professor of Medicine, Hebrew University, Hadassah Medical School; Director, Division of Adult Nephrology, Shaare Zedek Medical Center, Jerusalem, Israel Disorders of Sodium and Water Homeostasis Arthur S. slu*tsky, MD Professor of Medicine, Surgery, and Biomedical Engineering, University of Toronto; Vice President (Research), St. Michael’s Hospital, Keenan Research Centre, Li Ka Shing Knowledge Institute, Toronto, Ontario, Canada Acute Respiratory Failure; Mechanical Ventilation Eric J. Small, MD Professor of Medicine and Urology, Deputy Director and Director of Clinical Sciences, Helen Diller Family Comprehensive Cancer Center; Chief, Division of Hematology and Oncology, University of California San Francisco School of Medicine, San Francisco, California Prostate Cancer Gerald W. Smetana, MD Professor of Medicine, Harvard Medical School; Division of General Medicine and Primary Care, Beth Israel Deaconess Medical Center, Boston, Massachusetts Principles of Medical Consultation


Robert F. Spiera, MD Professor of Clinical Medicine, Weill Cornell Medical College; Director, Scleroderma, Vasculitis, and Myositis Center, The Hospital for Special Surgery, New York, New York Polymyalgia Rheumatica and Temporal Arteritis Stanley M. Spinola, MD Professor and Chair, Department of Microbiology and Immunology, Professor of Medicine, Microbiology and Immunology, and Pathology and Laboratory Medicine, Indiana University School of Medicine, Indianapolis, Indiana Chancroid David Spriggs, MD Head, Division of Solid Tumor Oncology, Department of Medicine, Memorial Sloan Kettering Cancer Center; Professor of Medicine, Department of Medicine, Weill Cornell Medical College, New York, New York Gynecologic Cancers Paweł Stankiewicz, MD, PhD Department of Molecular and Human Genetics, Baylor College of Medicine, Houston, Texas Gene, Genomic, and Chromosomal Disorders Paul Stark, MD Professor Emeritus, University of California San Diego; Chief of Cardiothoracic Radiology, VA San Diego Healthcare System, San Diego, California Imaging in Pulmonary Disease David P. Steensma, MD Professor of Medicine, Harvard Medical School, Adult Leukemia Program, Dana-Farber Cancer Institute, Boston, Massachusetts Myelodysplastic Syndrome Martin H. Steinberg, MD Professor of Medicine, Pediatrics, and Pathology and Laboratory Medicine, Boston University School of Medicine; Director, Center of Excellence in Sickle Cell Disease, Boston Medical Center, Boston, Massachusetts Sickle Cell Disease and Other Hemoglobinopathies Theodore S. Steiner, MD Associate Professor, University of British Columbia; Associate Head, Division of Infectious Diseases, Vancouver General Hospital, Vancouver, British Columbia, Canada Escherichia Coli Enteric Infections David S. Stephens, MD Stephen W. Schwarzmann Distinguished Professor of Medicine, Emory University School of Medicine and Woodruff Health Sciences Center, Atlanta, Georgia Neisseria Meningitidis Infections David A. Stevens, MD Professor of Medicine, Stanford University Medical School; President, Principal Investigator, Infectious Diseases Research Laboratory, California Institute for Medical Research, San Jose and Stanford, California Systemic Antifungal Agents



James K. Stoller, MD, MS Chairman, Education Institute, Jean Wall Bennett Professor of Medicine, Cleveland Clinic Lerner College of Medicine; Staff, Respiratory Institute, Cleveland Clinic, Cleveland, Ohio Respiratory Monitoring in Critical Care John H. Stone, MD, MPH Professor of Medicine, Director, Clinical Rheumatology, Harvard Medical School, Massachusetts General Hospital, Boston, Massachusetts The Systemic Vasculitides Richard M. Stone, MD Professor of Medicine, Harvard Medical School, Clinical Director, Adult Leukemia Program, Dana-Farber Cancer Institute, Boston, Massachusetts Myelodysplastic Syndrome Raymond A. Strikas, MD, MPH Education Team Lead, Immunization Services Division, National Center for Immunization and Respiratory Diseases, Centers for Disease Control and Prevention, Atlanta, Georgia Immunization Edwin P. Su, MD Associate Professor of Clinical Orthopaedics, Orthopaedic Surgery, Weill Cornell University Medical College; Associate Attending Orthopaedic Surgeon, Adult Reconstruction and Joint Replacement, Hospital for Special Surgery, New York, New York Surgical Treatment of Joint Disease Roland W. Sutter, MD, MPH&TM Coordinator, Research, Policy and Product Development, Polio Operations and Research Department, World Health Organization, Geneva, Switzerland Diphtheria and Other Corynebacteria Infections Ronald S. Swerdloff, MD Professor of Medicine, David Geffen School of Medicine at University of California Los Angeles; Chief, Division of Endocrinology, Department of Medicine, Harbor-UCLA Medical Center, Torrance, California The Testis and Male Hypogonadism, Infertility, and Sexual Dysfunction Heidi Swygard, MD, MPH Associate Professor of Medicine, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina Approach to the Patient with a Sexually Transmitted Infection Megan Sykes, MD Michael J. Friedlander Professor of Medicine, Director, Columbia Center for Translational Immunology, Columbia University Medical Center, New York, New York Transplantation Immunology Marian Tanofsky-Kraff, PhD Associate Professor, Department of Medical and Clinical Psychology, Uniformed Services University of Health Sciences, Bethesda, Maryland Eating Disorders Susan M. Tarlo, MBBS Professor of Medicine, Department of Medicine and Dalla Lana School of Public Health, University of Toronto, Respiratory Physician, University Health Network, Toronto Western Hospital and St. Michael’s Hospital, Toronto, Ontario, Canada Occupational Lung Disease Victoria M. Taylor, MD, MPH Professor of Medicine, University of Washington, Fred Hutchinson Cancer Research Center, Seattle, Washington Cultural Context of Medicine

Ayalew Tefferi, MD Professor of Medicine, Department of Hematology, Mayo Clinic, Rochester, Minnesota Polycythemia Vera, Essential Thrombocythemia, and Primary Myelofibrosis Paul S. Teirstein, MD Chief of Cardiology, Department of Medicine, Scripps Clinic, La Jolla, California Interventional and Surgical Treatment of Coronary Artery Disease Sam R. Telford III, ScD Professor, Tufts University Cummings School of Veterinary Medicine, North Grafton, Massachusetts Babesiosis and Other Protozoan Diseases Rajesh V. Thakker, MD May Professor of Medicine, University of Oxford; Radcliffe Department of Clinical Medicine, OCDEM, Churchill Hospital, Headington, Oxford, United Kingdom The Parathyroid Glands, Hypercalcemia, and Hypocalcemia Antonella Tosti, MD Professor of Clinical Dermatology, Department of Dermatology and Cutaneous Surgery, University of Miami, Miami, Florida Diseases of Hair and Nails Indi Trehan, MD, MPH, DTM&H Assistant Professor of Pediatrics, Washington University School of Medicine; Attending Physician, St. Louis Children’s Hospital, BarnesJewish Hospital, St. Louis, Missouri; Visiting Honorary Lecturer in Paediatrics and Child Health, University of Malawi College of Medicine; Consultant Paediatrician, Queen Elizabeth Central Hospital, Blantyre, Malawi Protein-Energy Malnutrition Ronald B. Turner, MD Professor of Pediatrics, University of Virginia School of Medicine, Charlottesville, Virginia The Common Cold Thomas S. Uldrick, MD Staff Clinician, HIV and AIDS Malignancy Branch, National Cancer Institute, Bethesda, Maryland Hematology and Oncology in Patients with Human Immunodeficiency Virus Infection Anthony M. Valeri, MD Professor of Medicine, Columbia University Medical Center; Director, Hemodialysis, Medical Director, Kidney and Pancreas Transplantation, New York-Presbyterian Hospital (CUMC); Director, Hemodialysis, Columbia University Dialysis Center, New York, New York Treatment of Irreversible Renal Failure John Varga, MD John and Nancy Hughes Professor of Medicine, Northwestern University Feinberg School of Medicine, Chicago, Illinois Systemic Sclerosis (Scleroderma) Bradley V. Vaughn, MD Professor of Neurology, Department of Neurology, University of North Carolina, Chapel Hill, North Carolina Disorders of Sleep Alan P. Venook, MD Professor of Medicine, University of California San Francisco, Helen Diller Family Comprehensive Cancer Center, San Francisco, California Liver and Biliary Tract Cancers Joseph G. Verbalis, MD Professor of Medicine, Georgetown University; Chief, Endocrinology and Metabolism, Georgetown University Hospital, Washington, D.C. Posterior Pituitary

Contributors Ronald G. Victor, MD Professor of Medicine, Burns and Allen Chair in Cardiology Research, Director, Hypertension Center, Associate Director, The Heart Institute, Cedars-Sinai Medical Center, Los Angeles, California Arterial Hypertension Angela Vincent, MBBS Professor of Neuroimmunology, University of Oxford; Honorary Consultant in Immunology, Oxford University Hospital Trust, Oxford, United Kingdom Disorders of Neuromuscular Transmission Robert M. Wachter, MD Professor and Associate Chairman, Department of Medicine, University of California San Francisco, San Francisco, California Quality of Care and Patient Safety Edward H. Wagner, MD, MPH Director Emeritus, MacColl Center for Health Care Innovation, Group Health Research Institute, Seattle, Washington Comprehensive Chronic Disease Management Edward E. Walsh, MD Professor of Medicine, University of Rochester School of Medicine and Dentistry; Head, Infectious Diseases, Rochester General Hospital, Rochester, New York Respiratory Syncytial Virus Thomas J. Walsh, MD Director, Transplantation-Oncology Infectious Diseases Program, Chief, Infectious Diseases Translational Research Laboratory, Professor of Medicine, Pediatrics, and Microbiology and Immunology, Weill Cornell Medical Center; Henry Schueler Foundation Scholar, Sharp Family Foundation Scholar in Pediatric Infectious Diseases, Adjunct Professor of Pathology, The Johns Hopkins University School of Medicine; Adjunct Professor of Medicine, The University of Maryland School of Medicine, Baltimore, Maryland Aspergillosis Jeremy D. Walston, MD Raymond and Anna Lublin Professor of Geriatric Medicine and Gerontology, The Johns Hopkins University School of Medicine, Baltimore, Maryland Common Clinical Sequelae of Aging Christina Wang, MD Professor of Medicine, David Geffen School of Medicine at University of California Los Angeles; Associate Director, UCLA Clinical and Translational Research Institute, Harbor-UCLA Medical Center, Torrance, California The Testis and Male Hypogonadism, Infertility, and Sexual Dysfunction Christine Wanke, MD Professor of Medicine and Public Health, Director, Division of Nutrition and Infection, Associate Chair, Department of Public Health, Tufts University School of Medicine, Boston, Massachusetts Gastrointestinal Manifestions of HIV and AIDS Stephen I. Wasserman, MD Professor of Medicine, University of California San Diego, La Jolla, California Approach to the Patient with Allergic or Immunologic Disease Thomas J. Weber, MD Associate Professor, Medicine/Endocrinology, Duke University, Durham, North Carolina Approach to the Patient with Metabolic Bone Disease; Osteoporosis Heiner Wedemeyer, MD Professor, Department of Gastroenterology, Hepatology, and Endocrinology, Hannover Medical School, Hannover, Germany Acute Viral Hepatitis


Geoffrey A. Weinberg, MD Professor of Pediatrics, University of Rochester School of Medicine and Dentistry; Director, Pediatric HIV Program, Golisano Children’s Hospital at University of Rochester Medical Center, Rochester, New York Parainfluenza Viral Disease David A. Weinstein, MD, MMSc Professor of Pediatric Endocrinology, Director, Glycogen Storage Disease Program, Division of Pediatric Endocrinology, University of Florida College of Medicine, Gainesville, Florida Glycogen Storage Diseases Robert S. Weinstein, MD Professor of Medicine, Department of Medicine, University of Arkansas for Medical Sciences; Staff Endocrinologist, Department of Medicine, Central Arkansas Veterans Health Care System, Little Rock, Arkansas Osteomalacia and Rickets Roger D. Weiss, MD Professor of Psychiatry, Harvard Medical School, Boston, Massachusetts; Chief, Division of Alcohol and Drug Abuse, McLean Hospital, Belmont, Massachusetts Drug Abuse and Dependence Martin Weisse, MD Chair, Pediatrics, Tripler Army Medical Center, Honolulu, Hawaii; Professor, Pediatrics, Uniformed Services University of the Health Sciences, Bethesda, Maryland Measles Jeffrey I. Weitz, MD Professor of Medicine and Biochemistry, McMaster University; Executive Director, Thrombosis and Atherosclerosis Research Institute, Hamilton, Ontario, Canada Pulmonary Embolism Samuel A. Wells, Jr., MD Medical Oncology Branch, National Cancer Institute, National Institutes of Health, Bethesda, Maryland Medullary Thyroid Carcinoma Richard P. Wenzel, MD, MSc Professor and Former Chairman, Internal Medicine, Virginia Commonwealth University, Richmond, Virginia Acute Bronchitis and Tracheitis Victoria P. Werth, MD Professor of Dermatology and Medicine, Hospital of the University of Pennsylvania and Philadelphia Veterans Administration Medical Center; Chief, Dermatology Division, Philadelphia Veterans Administration Medical Center, Philadelphia, Pennsylvania Principles of Therapy of Skin Diseases Sterling G. West, MD, MACP Professor of Medicine, University of Colorado School of Medicine; Associate Division Head for Clinical and Educational Affairs, University of Colorado Division of Rheumatology, Aurora, Colorado Systemic Diseases in Which Arthritis Is a Feature A. Clinton White, Jr., MD Paul R. Stalnaker Distinguished Professor and Director, Infectious Disease Division, Department of Internal Medicine, University of Texas Medical Branch, Galveston, Texas Cestodes Christopher J. White, MD Professor of Medicine, Ochsner Clinical School, University of Queensland School of Medicine; System Chairman of Cardiovascular Diseases, Ochsner Medical Center, New Orleans, Louisiana Atherosclerotic Peripheral Arterial Disease; Electrophysiologic Interventional Procedures and Surgery



Perrin C. White, MD Professor of Pediatrics, The Audry Newman Rapoport Distinguished Chair in Pediatric Endocrinology, University of Texas Southwestern Medical Center, Chief of Endocrinology, Children’s Medical Center Dallas, Dallas, Texas Disorders of Sexual Development Richard J. Whitley, MD Distinguished Professor of Pediatrics, Loeb Eminent Scholar Chair in Pediatrics, Professor of Pediatrics, Microbiology, Medicine, and Neurosurgery, The University of Alabama at Birmingham, Birmingham, Alabama Herpes Simplex Virus Infections Michael P. Whyte, MD Professor of Medicine, Pediatrics, and Genetics, Division of Bone and Mineral Diseases, Washington University School of Medicine; MedicalScientific Director, Center for Metabolic Bone Disease and Molecular Research, Shriners Hospital for Children, St. Louis, Missouri Osteonecrosis, Osteosclerosis/Hyperostosis, and Other Disorders of Bone Samuel Wiebe, MD, MSc Professor of Clinical Neurosciences, University of Calgary; Co-Director, Calgary Epilepsy Program, Alberta Health Services, Foothills Medical Centre, Calgary, Alberta, Canada The Epilepsies Jeanine P. Wiener-Kronish, MD Henry Isaiah Dorr Professor of Research and Teaching in Anaesthesia and Anesthestist-in-Chief, Department of Anesthesia, Critical Care and Pain Medicine, Massachusetts General Hospital/Harvard Medical School, Boston, Massachusetts Overview of Anesthesia Eelco F.M. Wijdicks, MD, PhD Professor of Neurology, Division of Critical Care Neurology, Department of Neurology, Mayo Clinic, Rochester, Minnesota Coma, Vegetative State, and Brain Death David J. Wilber, MD George M. Eisenberg Professor of Medicine, Loyola Stritch School of Medicine; Director, Division of Cardiology, Director, Clinical Electrophysiology, Loyola University Medical Center, Maywood, Illinois Electrophysiologic Interventional Procedures and Surgery Beverly Winikoff, MD, MPH President, Gynuity Health Projects; Professor of Clinical Population and Family Health, Mailman School of Public Health, Columbia University, New York, New York Contraception Gary P. Wormser, MD Professor of Medicine and Chief, Division of Infectious Diseases, Department of Medicine, New York Medical College, Valhalla, New York Lyme Disease

Myron Yanoff, MD Professor and Chair, Ophthalmology, Drexel University College of Medicine, Philadelphia, Pennsylvania Diseases of the Visual System Robert Yarchoan, MD Branch Chief, HIV and AIDS Malignancy Branch, National Cancer Institute, Bethesda, Maryland Hematology and Oncology in Patients with Human Immunodeficiency Virus Infection Neal S. Young, MD Chief, Hematology Branch, NHLBI and Director, Trans-NIH Center for Human Immunology, Autoimmunity, and Inflammation, National Institutes of Health, Bethesda, Maryland Parvovirus William F. Young, Jr., MD, MSc Professor of Medicine, Mayo Clinic College of Medicine; Chair, Division of Endocrinology, Diabetes, Metabolism, and Nutrition, Mayo Clinic, Rochester, Minnesota Adrenal Medulla, Catecholamines, and Pheochromocytoma Alan S.L. Yu, MB, BChir Harry Statland and Solon Summerfield Professor of Medicine, Director, Division of Nephrology and Hypertension and the Kidney Institute, University of Kansas Medical Center, Kansas City, Kansas Disorders of Magnesium and Phosphorus Sherif R. Zaki, MD, PhD Chief, Infectious Diseases Pathology Branch, Centers for Disease Control and Prevention, Atlanta, Georgia Leptospirosis Mark L. Zeidel, MD Herman L. Blumgart Professor of Medicine, Harvard Medical School; Physician-in-Chief and Chairman, Department of Medicine, Beth Israel Deaconess Medical Center, Boston, Massachusetts Obstructive Uropathy Thomas R. Ziegler, MD Professor, Department of Medicine, Division of Endocrinology, Metabolism, and Lipids, Emory University School of Medicine, Atlanta, Georgia Malnutrition, Nutritional Assessment, and Nutritional Support in Adult Hospitalized Patients Peter Zimetbaum, MD Associate Professor of Medicine, Harvard Medical School; Director of Clinical Cardiology, Beth Israel Deaconess Medical Center, Boston, Massachusetts Cardiac Arrhythmias with Supraventricular Origin

VIDEO CONTENTS This icon appears throughout the book to indicate chapters with accompanying video available on For quick viewing, use your smartphone to scan the QR codes in the front of the book. Aging and Geriatric Medicine Confusion Assessment Method (CAM) Video 28-1 – MARCOS MIALNEZ, JORGE G. RUIZ, AND ROSANNE M. LEIPZIG

Clinical Pharmacology Interlaminar Epidural Steroid Injection Video 30-1 – ALI TURABI

Cardiovascular Disease Standard Echocardiographic Views: Long Axis Image Plane Video 55-1A – CATHERINE M. OTTO Standard Echocardiographic Views: Short Axis Image Plane Video 55-1B – CATHERINE M. OTTO Standard Echocardiographic Views: Short Axis Image Plane Video 55-1C – CATHERINE M. OTTO Standard Echocardiographic Views: Four-Chamber Image Plane Video 55-1D – CATHERINE M. OTTO Dilated Cardiomyopathy: Long Axis View Video 55-2A – CATHERINE M. OTTO Dilated Cardiomyopathy: Short Axis View Video 55-2B – CATHERINE M. OTTO Dilated Cardiomyopathy: Apical Four-Chamber View Video 55-2C – CATHERINE M. OTTO Three-Dimensional Echocardiography Video 55-3 – CATHERINE M. OTTO Stress Echocardiography: Normal Reaction Video 55-4A – CATHERINE M. OTTO Stress Echocardiography: Normal Reaction Video 55-4B – CATHERINE M. OTTO Stress Echocardiography: Proximal Stenosis of the Left Anterior Descending Coronary Artery Video 55-4C – CATHERINE M. OTTO Stress Echocardiography: Proximal Stenosis of the Left Anterior Descending Coronary Artery Video 55-4D – CATHERINE M. OTTO Pericardial Effusion: Parasternal Long Axis Video 55-5A – CATHERINE M. OTTO Pericardial Effusion: Parasternal Short Axis Video 55-5B – CATHERINE M. OTTO Pericardial Effusion: Apical Four-Chamber Views Video 55-5C – CATHERINE M. OTTO Secundum Atrial Septal Defect Video 69-1 – ARIANE J. MARELLI Perimembranous Ventricular Septal Defect Video 69-2 – ARIANE J. MARELLI Coronary Stent Placement Video 74-1 – PAUL S. TEIRSTEIN Guidewire Passage Video 74-2 – PAUL S. TEIRSTEIN Delivering the Stent Video 74-3 – PAUL S. TEIRSTEIN Inflating the Stent Video 74-4 – PAUL S. TEIRSTEIN

Final Result Video 74-5 – PAUL S. TEIRSTEIN Superficial Femoral Artery (SFA) Stent Procedure Video 79-1 – CHRISTOPHER J. WHITE Orthotopic Bicaval Cardiac Transplantation Video 82-1 – Y. JOSEPH WOO

Respiratory Diseases Wheezing Video 87-1 – JEFFREY M. DRAZEN Technique for Use of a Metered-Dose Inhaler Video 87-2 – LESLIE HENDELES and the New England Journal of Medicine VATS Wedge Resection Video 101-1 – MALCOLM M. DeCAMP

Critical Care Medicine Ventilation of an Ex Vivo Rat Lung Video 105-1 – ARTHUR S. slu*tSKY, GEORGE VOLGYESI, AND TOM WHITEHEAD

Renal and Genitourinary Diseases Renal Artery Stent Video 125-1 – RENATO M. SANTOS AND THOMAS D. DUBOSE, JR. Interpretation of a Computed Tomographic Colonography Video 133-1 – DAVID H. KIM Donor Liver Transplantation—Donor and Recipient Video 154-1 – IGAL KAM, THOMAS BAK, AND MICHAEL WACHS

Oncology Snare Polypectomy of a Colon Adenoma Video 193-1 – DOUGLAS O. FAIGEL Laparoscopic-Assisted Double Balloon Enteroscopy with Polypectomy of a Jejunal Adenoma Followed by Surgical Oversew of the Polypectomy Site Video 193-2 – DOUGLAS O. FAIGEL Endoscopic Mucosal Resection Using Saline Lift Polypectomy of a Colon Adenoma Followed by Closure of the Mucosal Defect with Clips Video 193-3 – DOUGLAS O. FAIGEL Endoscopic View of Rectal Cancer Video 193-4 – DOUGLAS O. FAIGEL Endoscopic Ultrasound Video 193-5 – DOUGLAS O. FAIGEL

Nutritional Diseases Laparoscopic Roux-en-Y Gastric Bypass Video 220-1 – JAMES M. SWAIN

Endocrine Diseases Pituitary Surgery Video 224-1 – IVAN CIRIC

Diseases of Allergy and Clinical Immunology Skin Testing Video 251-1 – LARRY BORISH Nasal Endoscopy Video 251-2 – LARRY BORISH


Video Contents

Rheumatic Diseases Hip Arthroscopy Osteochondroplasty Video 276-1 – BRYAN T. KELLY

Neurology Cervical Provocation Video 400-1 – RICHARD L. BARBANO Spurling Maneuver Video 400-2 – RICHARD L. BARBANO Cervical Distraction Test Video 400-3 – RICHARD L. BARBANO Straight Leg Raise Video 400-4 – RICHARD L. BARBANO Contralateral Straight Leg Raise Video 400-5 – RICHARD L. BARBANO Seated Straight Leg Raise Video 400-6 – RICHARD L. BARBANO Discectomy Video 400-7 – JASON H. HUANG Absence Seizure Video 403-1 – SAMUEL WIEBE Left Rolandic Seizure Video 403-2 – SAMUEL WIEBE Left Temporal Complex Partial Seizure Video 403-3 – SAMUEL WIEBE Left Temporal Complex Partial Seizure Postictal Confusion Video 403-4 – SAMUEL WIEBE Left Temporal Complex Partial Seizure Video 403-5 – SAMUEL WIEBE Supplementary Sensory-Motor Seizure Video 403-6 – SAMUEL WIEBE Right Posterior Temporal Seizure-Dramatic Frontal Semiology Video 403-7 – SAMUEL WIEBE Right Mesial Frontal Seizure Video 403-8 – SAMUEL WIEBE Nonconvulsive Status Epilepticus Video 403-9 – SAMUEL WIEBE GTC Seizure Tonic Phase Video 403-10 – SAMUEL WIEBE GTC Seizure Clonic Phase Video 403-11 – SAMUEL WIEBE Myoclonic Facial Seizure Video 403-12 – SAMUEL WIEBE Tonic Seizure Lennox Gastaut Video 403-13 – SAMUEL WIEBE Atonic Seizure Lennox Gastaut Video 403-14 – SAMUEL WIEBE Reflex Auditory Seizure Video 403-15 – SAMUEL WIEBE Four Score Video 404-1 – JAMES L. BERNAT AND EELCO F.M. WIJDICKS Persistent Vegetative State Video 404-2 – JAMES L. BERNAT AND EELCO F.M. WIJDICKS Minimally Conscious State Video 404-3 – JAMES L. BERNAT AND EELCO F.M. WIJDICKS Akinetic Mutism Video 404-4 – JAMES L. BERNAT AND EELCO F.M. WIJDICKS Early Parkinson’s Disease Video 409-1 – ANTHONY E. LANG Freezing of Gait in Parkinson’s Disease Video 409-2 – ANTHONY E. LANG

Gunslinger Gait in Progressive Supranuclear Palsy Video 409-3 – ANTHONY E. LANG Supranuclear Gaze Palsy in Progressive Supranuclear Palsy Video 409-4 – ANTHONY E. LANG Applause Sign in Progressive Supranuclear Palsy Video 409-5 – ANTHONY E. LANG Apraxia of Eyelid Opening in Progressive Supranuclear Palsy Video 409-6 – ANTHONY E. LANG Cranial Dystonia in Multiple System Atrophy Video 409-7 – ANTHONY E. LANG Anterocollis in Multiple System Atrophy Video 409-8 – ANTHONY E. LANG Stridor in Multiple System Atrophy Video 409-9 – ANTHONY E. LANG Alien Limb Phenomenon in Corticobasal Syndrome Video 409-10 – ANTHONY E. LANG Myoclonus in Corticobasal Syndrome Video 409-11 – ANTHONY E. LANG Levodopa-Induced Dyskinesia in Parkinson’s Disease Video 409-12 – ANTHONY E. LANG Essential Tremor Video 410-1 – ANTHONY E. LANG Huntington’s Disease Video 410-2 – ANTHONY E. LANG Hemiballism Video 410-3 – ANTHONY E. LANG Blepharospasm Video 410-4 – ANTHONY E. LANG Oromandibular Dystonia Video 410-5 – ANTHONY E. LANG Cervical Dystonia Video 410-6 – ANTHONY E. LANG Writer’s Cramp Video 410-7 – ANTHONY E. LANG Embouchure Dystonia Video 410-8 – ANTHONY E. LANG Sensory Trick in Cervical Dystonia Video 410-9 – ANTHONY E. LANG Generalized Dystonia Video 410-10 – ANTHONY E. LANG Tics Video 410-11 – ANTHONY E. LANG Tardive Dyskinesia Video 410-12 – ANTHONY E. LANG Hemifacial Spasm Video 410-13 – ANTHONY E. LANG Wernicke Encephalopathy Eye Movements: Before Thiamine Video 416-1 – BARBARA S. KOPPEL Wernicke Encephalopathy Eye Movements: After Thiamine Video 416-2 – BARBARA S. KOPPEL Limb Symptoms and Signs Video 419-1 – PAMELA J. SHAW Bulbar Symptoms and Signs Video 419-2 – PAMELA J. SHAW Normal Swallowing Video 419-3 – PAMELA J. SHAW Charcot-Marie-Tooth Disease Exam and Walk Video 420-1 – MICHAEL E. SHY

QUICK REFERENCE (QR) VIDEO ACCESS The images below are QR codes. Each code corresponds to a video from the Goldman-Cecil Medicine 25 collection. For fast and easy video access, right from your mobile device, follow these instructions. The videos are also available on

What You Need

• A mobile device, such as a smartphone or tablet, equipped with a camera and Internet access • A QR code reader application (If you do not already have a reader installed on your mobile device, look for free versions in your app store.)

How It Works

• Open the QR code reader application on your mobile device. • Point the device’s camera at the code and scan. • Each code opens an individual video player for instant viewing—no log-on required.

Confusion Assessment Method (CAM) Chapter 28, Video 28-1 – Marcos Mialnez, Jorge G. Ruiz, and Rosanne M. Leipzig

Standard Echocardiographic Views: Four-Chamber Image Plane Chapter 55, Video 55-1D – Catherine M. Otto

Interlaminar Epidural Steroid Injection Chapter 30, Video 30-1 – Ali Turabi

Dilated Cardiomyopathy: Long Axis View Chapter 55, Video 55-2A – Catherine M. Otto

Standard Echocardiographic Views: Long Axis Image Plane Chapter 55, Video 55-1A – Catherine M. Otto

Dilated Cardiomyopathy: Short Axis View Chapter 55, Video 55-2B – Catherine M. Otto

Standard Echocardiographic Views: Short Axis Image Plane Chapter 55, Video 55-1B – Catherine M. Otto

Dilated Cardiomyopathy: Apical Four-Chamber View Chapter 55, Video 55-2C – Catherine M. Otto

Standard Echocardiographic Views: Short Axis Image Plane Chapter 55, Video 55-1C – Catherine M. Otto

Three-Dimensional Echocardiography Chapter 55, Video 55-3 – Catherine M. Otto


Quick Reference (QR) Video Access

Stress Echocardiography: Normal Reaction Chapter 55, Video 55-4A – Catherine M. Otto

Perimembranous Ventricular Septal Defect Chapter 69, Video 69-2 – Ariane J. Marelli

Stress Echocardiography: Normal Reaction Chapter 55, Video 55-4B – Catherine M. Otto

Coronary Stent Placement Chapter 74, Video 74-1 – Paul S. Teirstein

Stress Echocardiography: Proximal Stenosis of the Left Anterior Descending Coronary Artery Chapter 55, Video 55-4C – Catherine M. Otto

Guidewire Passage Chapter 74, Video 74-2 – Paul S. Teirstein

Stress Echocardiography: Proximal Stenosis of the Left Anterior Descending Coronary Artery Chapter 55, Video 55-4D – Catherine M. Otto

Delivering the Stent Chapter 74, Video 74-3 – Paul S. Teirstein

Pericardial Effusion: Parasternal Long Axis Chapter 55, Video 55-5A – Catherine M. Otto

Inflating the Stent Chapter 74, Video 74-4 – Paul S. Teirstein

Pericardial Effusion: Parasternal Short Axis Chapter 55, Video 55-5B – Catherine M. Otto

Final Result Chapter 74, Video 74-5 – Paul S. Teirstein

Pericardial Effusion: Apical Four-Chamber Views Chapter 55, Video 55-5C – Catherine M. Otto

Superficial Femoral Artery (SFA) Stent Procedure Chapter 79, Video 79-1 – Christopher J. White

Secundum Atrial Septal Defect Chapter 69, Video 69-1 – Ariane J. Marelli

Orthotopic Bicaval Cardiac Transplantation Chapter 82, Video 82-1 – Y. Joseph Woo

Quick Reference (QR) Video Access

Wheezing Chapter 87, Video 87-1 – Jeffrey M. Drazen

Endoscopic Mucosal Resection Using Saline Lift Polypectomy of a Colon Adenoma Followed by Closure of the Mucosal Defect with Clips Chapter 193, Video 193-3 – Douglas O. Faigel

VATS Wedge Resection Chapter 101, Video 101-1 – Malcolm M. DeCamp

Endoscopic View of Rectal Cancer Chapter 193, Video 193-4 – Douglas O. Faigel

Ventilation of an Ex Vivo Rat Lung Chapter 105, Video 105-1 – Arthur S. slu*tsky, George Volgyesi, and Tom Whitehead

Endoscopic Ultrasound Chapter 193, Video 193-5 – Douglas O. Faigel

Renal Artery Stent Chapter 125, Video 125-1 – Renato M. Santos and Thomas D. DuBose, Jr.

Laparoscopic Roux-en-Y Gastric Bypass Chapter 220, Video 220-1 – James M. Swain

Interpretation of a Computed Tomographic Colonography Chapter 133, Video 133-1 – David H. Kim

Pituitary Surgery Chapter 224, Video 224-1 – Ivan Ciric

Donor Liver Transportation–Donor and Recipient Chapter 154, Video 154-1 – Igal Kam, Thomas Bak, and Michael Wachs

Skin Testing Chapter 251, Video 251-1 – Larry Borish

Snare Polypectomy of a Colon Adenoma Chapter 193, Video 193-1 – Douglas O. Faigel

Nasal Endoscopy Chapter 251, Video 251-2 – Larry Borish

Laparascopic-Assisted Double Balloon Enteroscopy with Polypectomy of a Jejunal Adenoma Followed by Surgical Oversew of the Polypectomy Site Chapter 193, Video 193-2 – Douglas O. Faigel

Hip Arthroscopy Osteochondroplasty Chapter 276, Video 276-1 – Bryan T. Kelly



Quick Reference (QR) Video Access

Cervical Provocation Chapter 400, Video 400-1 – Richard L. Barbano

Left Rolandic Seizure Chapter 403, Video 403-2 – Samuel Wiebe

Spurling Maneuver Chapter 400, Video 400-2 – Richard L. Barbano

Left Temporal Complex Partial Seizure Chapter 403, Video 403-3 – Samuel Wiebe

Cervical Distraction Test Chapter 400, Video 400-3 – Richard L. Barbano

Left Temporal Complex Partial Seizure Postictal Confusion Chapter 403, Video 403-4 – Samuel Wiebe

Straight Leg Raise Chapter 400, Video 400-4 – Richard L. Barbano

Left Temporal Complex Partial Seizure Chapter 403, Video 403-5 – Samuel Wiebe

Contralateral Straight Leg Raise Chapter 400, Video 400-5 – Richard L. Barbano

Supplementary Sensory-Motor Seizure Chapter 403, Video 403-6 – Samuel Wiebe

Seated Straight Leg Raise Chapter 400, Video 400-6 – Richard L. Barbano

Right Posterior Temporal Seizure - Dramatic Frontal Semiology Chapter 403, Video 403-7 – Samuel Wiebe

Discectomy Chapter 400, Video 400-7 – Jason H. Huang

Right Mesial Frontal Seizure Chapter 403, Video 403-8 – Samuel Wiebe

Absence Seizure Chapter 403, Video 403-1 – Samuel Wiebe

Nonconvulsive Status Epilepticus Chapter 403, Video 403-9 – Samuel Wiebe

Quick Reference (QR) Video Access

GTC Seizure Tonic Phase Chapter 403, Video 403-10 – Samuel Wiebe

Minimally Conscious State Chapter 404, Video 404-3 – James L. Bernat and Eelco F. M. Wijdicks

GTC Seizure Clonic Phase Chapter 403, Video 403-11 – Samuel Wiebe

Akinetic Mutism Chapter 404, Video 404-4 – James L. Bernat and Eelco F. M. Wijdicks

Myoclonic Facial Seizure Chapter 403, Video 403-12 – Samuel Wiebe

Early Parkinson’s Disease Chapter 409, Video 409-1 – Anthony E. Lang

Tonic Seizure Lennox Gastaut Chapter 403, Video 403-13 – Samuel Wiebe

Freezing of Gait in Parkinson’s Disease Chapter 409, Video 409-2 – Anthony E. Lang

Atonic Seizure Lennox Gastaut Chapter 403, Video 403-14 – Samuel Wiebe

Gunslinger Gait in Progressive Supranuclear Palsy Chapter 409, Video 409-3 – Anthony E. Lang

Reflex Auditory Seizure Chapter 403, Video 403-15 – Samuel Wiebe

Supranuclear Gaze Palsy in Progressive Supranuclear Palsy Chapter 409, Video 409-4 – Anthony E. Lang

Four Score Chapter 404, Video 404-1 – James L. Bernat and Eelco F. M. Wijdicks

Applause Sign in Progressive Supranuclear Palsy Chapter 409, Video 409-5 – Anthony E. Lang

Persistent Vegetative State Chapter 404, Video 404-2 – James L. Bernat and Eelco F. M. Wijdicks

Apraxia of Eyelid Opening in Progressive Supranuclear Palsy Chapter 409, Video 409-6 – Anthony E. Lang



Quick Reference (QR) Video Access

Cranial Dystonia in Multiple System Atrophy Chapter 409, Video 409-7 – Anthony E. Lang

Hemiballism Chapter 410, Video 410-3 – Anthony E. Lang

Anterocollis in Multiple System Atrophy Chapter 409, Video 409-8 – Anthony E. Lang

Blepharospasm Chapter 410, Video 410-4 – Anthony E. Lang

Stridor in Multiple System Atrophy Chapter 409, Video 409-9 – Anthony E. Lang

Oromandibular Dystonia Chapter 410, Video 410-5 – Anthony E. Lang

Alien Limb Phenomenon in Corticobasal Syndrome Chapter 409, Video 409-10 – Anthony E. Lang

Cervical Dystonia Chapter 410, Video 410-6 – Anthony E. Lang

Myoclonus in Corticobasal Syndrome Chapter 409, Video 409-11 – Anthony E. Lang

Writer’s Cramp Chapter 410, Video 410-7 – Anthony E. Lang

Levodopa-Induced Dyskinesia in Parkinson’s Disease Chapter 409, Video 409-12 – Anthony E. Lang

Embouchure Dystonia Chapter 410, Video 410-8 – Anthony E. Lang

Essential Tremor Chapter 410, Video 410-1 – Anthony E. Lang

Sensory Trick in Cervical Dystonia Chapter 410, Video 410-9 – Anthony E. Lang

Huntington’s Disease Chapter 410, Video 410-2 – Anthony E. Lang

Generalized Dystonia Chapter 410, Video 410-10 – Anthony E. Lang

Quick Reference (QR) Video Access

Tics Chapter 410, Video 410-11 – Anthony E. Lang

Limb Symptoms and Signs Chapter 419, Video 419-1 – Pamela J. Shaw

Tardive Dyskinesia Chapter 410, Video 410-12 – Anthony E. Lang

Bulbar Symptoms and Signs Chapter 419, Video 419-2 – Pamela J. Shaw

Hemifacial Spasm Chapter 410, Video 410-13 – Anthony E. Lang

Normal Swallowing Chapter 419, Video 419-3 – Pamela J. Shaw

Wernickes Encephalopathy Eye Movements: Before Thiamine Chapter 416, Video 416-1 – Barbara S. Koppel

Charcot-Marie-Tooth Disease Exam and Walk Chapter 420, Video 420-1 – Michael E. Shy

Wernickes Encephalopathy Eye Movements: After Thiamine Chapter 416, Video 416-2 – Barbara S. Koppel





Harold and Margaret Hatch Professor Executive Vice President and Dean of the Faculties of Health Sciences and Medicine Chief Executive, Columbia University Medical Center Columbia University New York, New York


Professor of Medicine Director, Richard T. Silver Center for Myeloproliferative Neoplasms Weill Cornell Medical College New York, New York





Medicine is a profession that incorporates science and the scientific method with the art of being a physician. The art of tending to the sick is as old as humanity itself. Even in modern times, the art of caring and comforting, guided by millennia of common sense as well as a more recent, systematic approach to medical ethics (Chapter 2), remains the cornerstone of medicine. Without these humanistic qualities, the application of the modern science of medicine is suboptimal, ineffective, or even detrimental. The caregivers of ancient times and premodern cultures tried a variety of interventions to help the afflicted. Some of their potions contained what are now known to be active ingredients that form the basis for proven medications (Chapter 29). Others (Chapter 39) have persisted into the present era despite a lack of convincing evidence. Modern medicine should not dismiss the possibility that these unproven approaches may be helpful; instead, it should adopt a guiding principle that all interventions, whether traditional or newly developed, can be tested vigorously, with the expectation that any beneficial effects can be explored further to determine their scientific basis. When compared with its long and generally distinguished history of caring and comforting, the scientific basis of medicine is remarkably recent. Other than an understanding of human anatomy and the later description, albeit widely contested at this time, of the normal physiology of the circulatory system, almost all of modern medicine is based on discoveries made within the past 150 years. Until the late 19th century, the paucity of medical knowledge was perhaps exemplified best by hospitals and hospital care. Although hospitals provided caring that all but well-to-do people might not be able to obtain elsewhere, there is little if any evidence that hospitals improved health outcomes. The term hospitalism referred not to expertise in hospital care but rather to the aggregate of iatrogenic afflictions that were induced by the hospital stay itself. The essential humanistic qualities of caring and comforting can achieve full benefit only if they are coupled with an understanding of how medical science can and should be applied to patients with known or suspected diseases. Without this knowledge, comforting may be inappropriate or misleading, and caring may be ineffective or counterproductive if it inhibits a sick person from obtaining appropriate, scientific medical care. Goldman-Cecil Medicine focuses on the discipline of internal medicine, from which neurology and dermatology, which are also covered in substantial detail in this text, are relatively recent evolutionary branches. The term internal medicine, which is often misunderstood by the lay public, was developed in 19th-century Germany. Inneren medizin was to be distinguished from clinical medicine because it emphasized the physiology and chemistry of disease, not just the patterns or progression of clinical manifestations. Goldman-Cecil Medicine follows this tradition by showing how pathophysiologic abnormalities cause symptoms and signs and by emphasizing how therapies can modify the underlying pathophysiology and improve the patient’s well-being. Modern medicine has moved rapidly past organ physiology to an increasingly detailed understanding of cellular, subcellular, and genetic mechanisms. For example, the understanding of microbial pathogenesis and many inflammatory diseases (Chapter 256) is now guided by a detailed understanding of the human immune system and its response to foreign antigens (Chapters 45 to 49). Advances in our understanding of the human microbiome raise the possibility that our complex interactions with microbes, which outnumber our cells by a factor of 10, will help explain conditions ranging from inflammatory bowel disease (Chapter 141) to obesity (Chapter 220).1 Health, disease, and an individual’s interaction with the environment are also substantially determined by genetics. In addition to many conditions

that may be determined by a single gene (Chapter 41), medical science increasingly understands the complex interactions that underlie multigenic traits (Chapter 42). The decoding of the human genome holds the promise that personalized health care increasingly can be targeted according to an individual’s genetic profile, in terms of screening and presymptomatic disease management, as well as in terms of specific medications and their adjusted dosing schedules.2 Although gene therapy has been approved for only one disease, lipoprotein lipase deficiency (Chapter 206), and only in Europe, it has shown promise in other conditions, such as Leber congenital amaurosis (Chapter 423). Cell therapy is now beginning to provide vehicles for the delivery of genes, gene products, and vaccines. It has also opened the way for “regenerative medicine” by facilitating the regeneration of injured or diseased organs and tissues. Such advances and others, such as nanomedicine, have already led to targeted and personalized therapies for a variety of cancers.3 Knowledge of the structure and physical forms of proteins helps explain abnormalities as diverse as sickle cell anemia (Chapter 163) and prion-related diseases (Chapter 415). Proteomics, which is the normal and abnormal protein expression of genes, also holds extraordinary promise for developing drug targets for more specific and effective therapies. Concurrent with these advances in fundamental human biology has been a dramatic shift in methods for evaluating the application of scientific advances to the individual patient and to populations. The randomized controlled trial, sometimes with thousands of patients at multiple institutions, has replaced anecdote as the preferred method for measuring the benefits and optimal uses of diagnostic and therapeutic interventions (Chapter 10). As studies progress from those that show biologic effect, to those that elucidate dosing schedules and toxicity, and finally to those that assess true clinical benefit, the metrics of measuring outcome has also improved from subjective impressions of physicians or patients to reliable and valid measures of morbidity, quality of life, functional status, and other patient-oriented outcomes (Chapter 11). These marked improvements in the scientific methodology of clinical investigation have expedited extraordinary changes in clinical practice, such as recanalization therapy for acute myocardial infarction (Chapter 73), and have shown that reliance on intermediate outcomes, such as a reduction in asymptomatic ventricular arrhythmias with certain drugs, may unexpectedly increase rather than decrease mortality. Just as physicians in the 21st century must understand advances in fundamental biology, similar understanding of the fundamentals of clinical study design as it applies to diagnostic and therapeutic interventions is needed. An understanding of human genetics will also help stratify and refine the approach to clinical trials by helping researchers select fewer patients with a more hom*ogeneous disease pattern to study the efficacy of an intervention. This explosion in medical knowledge has led to increasing specialization and subspecialization, defined initially by organ system and more recently by locus of principal activity (inpatient vs. outpatient), reliance on manual skills (proceduralist vs. nonproceduralist), or participation in research. Nevertheless, it is becoming increasingly clear that the same fundamental molecular and genetic mechanisms are broadly applicable across all organ systems and that the scientific methodologies of randomized trials and careful clinical observation span all aspects of medicine. The advent of modern approaches to managing data now provides the rationale for the use of health information technology. Computerized health records, oftentimes shared with patients in a portable format, can avoid duplication of tests and assure that care is coordinated among the patient’s various health care providers. Extraordinary advances in the science and practice of medicine, which have continued to accelerate with each recent edition of this textbook, have transformed the global burden of disease.4 Life expectancies for men and women are increasing, a greater proportion of deaths are occurring among people older than age 70 years, and far fewer children are dying before the age of 5 years. Nevertheless, huge regional disparities remain, and disability from conditions such as substance abuse, mental health disorders, injuries, diabetes, musculoskeletal disease, and chronic respiratory disease have become increasingly important issues for all health systems.


Patients commonly have complaints (symptoms). These symptoms may or may not be accompanied by abnormalities on examination (signs) or on laboratory testing. Conversely, asymptomatic patients may have signs or laboratory abnormalities, and laboratory abnormalities can occur in the absence of symptoms or signs.


Symptoms and signs commonly define syndromes, which may be the common final pathway of a wide range of pathophysiologic alterations. The fundamental basis of internal medicine is that diagnosis should elucidate the pathophysiologic explanation for symptoms and signs so that therapy may improve the underlying abnormality, not just attempt to suppress the abnormal symptoms or signs. When patients seek care from physicians, they may have manifestations or exacerbations of known conditions, or they may have symptoms and signs that suggest malfunction of a particular organ system. Sometimes the pattern of symptoms and signs is highly suggestive or even pathognomonic for a particular disease process. In these situations, in which the physician is focusing on a particular disease, Goldman-Cecil Medicine provides scholarly yet practical approaches to the epidemiology, pathobiology, clinical manifestations, diagnosis, treatment, prevention, and prognosis of entities such as acute myocardial infarction (Chapter 73), chronic obstructive lung disease (Chapter 88), obstructive uropathy (Chapter 123), inflammatory bowel disease (Chapter 141), gallstones (Chapter 155), rheumatoid arthritis (Chapter 264), hypothyroidism (Chapter 226), tuberculosis (Chapter 324), and virtually any known medical condition in adults. Many patients, however, have undiagnosed symptoms, signs, or laboratory abnormalities that cannot be immediately ascribed to a particular disease or cause. Whether the initial manifestation is chest pain (Chapter 51), diarrhea (Chapter 140), neck or back pain (Chapter 400), or a variety of more than 100 common symptoms, signs, or laboratory abnormalities, Goldman-Cecil Medicine provides tables, figures, and entire chapters to guide the approach to diagnosis and therapy (see E-Table 1-1 or table on inside back cover). By virtue of this dual approach to known disease as well as to undiagnosed abnormalities, this textbook, similar to the modern practice of medicine, applies directly to patients regardless of their mode of manifestation or degree of previous evaluation. The patient-physician interaction proceeds through many phases of clinical reasoning and decision making. The interaction begins with an elucidation of complaints or concerns, followed by inquiries or evaluations to address these concerns in increasingly precise ways. The process commonly requires a careful history or physical examination, ordering of diagnostic tests, integration of clinical findings with test results, understanding of the risks and benefits of the possible courses of action, and careful consultation with the patient and family to develop future plans. Physicians can increasingly call on a growing literature of evidence-based medicine to guide the process so that benefit is maximized while respecting individual variations in different patients. Throughout Goldman-Cecil Medicine, the best current evidence is highlighted with specific grade A references that can be accessed directly in the electronic version. The increasing availability of evidence from randomized trials to guide the approach to diagnosis and therapy should not be equated with “cookbook” medicine. Evidence and the guidelines that are derived from it emphasize proven approaches for patients with specific characteristics. Substantial clinical judgment is required to determine whether the evidence and guidelines apply to individual patients and to recognize the occasional exceptions. Even more judgment is required in the many situations in which evidence is absent or inconclusive. Evidence must also be tempered by patients’ preferences, although it is a physician’s responsibility to emphasize evidence when presenting alternative options to the patient. The adherence of a patient to a specific regimen is likely to be enhanced if the patient also understands the rationale and evidence behind the recommended option. To care for a patient as an individual, the physician must understand the patient as a person. This fundamental precept of doctoring includes an understanding of the patient’s social situation, family issues, financial concerns, and preferences for different types of care and outcomes, ranging from maximum prolongation of life to the relief of pain and suffering (Chapters 2 and 3). If the physician does not appreciate and address these issues, the science of medicine cannot be applied appropriately, and even the most knowledgeable physician will fail to achieve the desired outcomes. Even as physicians become increasingly aware of new discoveries, patients can obtain their own information from a variety of sources, some of which are of questionable reliability. The increasing use of alternative and complementary therapies (Chapter 39) is an example of patients’ frequent dissatisfaction with prescribed medical therapy. Physicians should keep an open mind regarding unproven options but must advise their patients carefully if such options may carry any degree of potential risk, including the risk that they may be relied on to substitute for proven approaches. It is crucial for the


physician to have an open dialogue with the patient and family regarding the full range of options that either may consider. The physician does not exist in a vacuum, but rather as part of a complicated and extensive system of medical care and public health. In premodern times and even today in some developing countries, basic hygiene, clean water, and adequate nutrition have been the most important ways to promote health and reduce disease. In developed countries, adoption of healthy lifestyles, including better diet (Chapter 213) and appropriate exercise (Chapter 16), is the cornerstone to reducing the epidemics of obesity (Chapter 220), coronary disease (Chapter 52), and diabetes (Chapter 229). Public health interventions to provide immunizations (Chapter 18) and to reduce injuries and the use of tobacco (Chapter 32), illicit drugs (Chapter 34), and excess alcohol (Chapter 33) can collectively produce more health benefits than nearly any other imaginable health intervention.


In a profession, practitioners put the welfare of clients or patients above their own welfare.5 Professionals have a duty that may be thought of as a contract with society. The American Board of Internal Medicine and the European Federation of Internal Medicine have jointly proposed that medical professionalism should emphasize three fundamental principles: the primacy of patient welfare, patient autonomy, and social justice.6 As modern medicine brings a plethora of diagnostic and therapeutic options, the interactions of the physician with the patient and society become more complex and potentially fraught with ethical dilemmas (Chapter 2). To help provide a moral compass that is not only grounded in tradition but also adaptable to modern times, the primacy of patient welfare emphasizes the fundamental principle of a profession. The physician’s altruism, which begets the patient’s trust, must be impervious to the economic, bureaucratic, and political challenges that are faced by the physician and the patient (Chapter 5). The principle of patient autonomy asserts that physicians make recommendations but patients make the final decisions. The physician is an expert advisor who must inform and empower the patient to base decisions on scientific data and how these data can and should be integrated with a patient’s preferences. The importance of social justice symbolizes that the patient-physician interaction does not exist in a vacuum. The physician has a responsibility to the individual patient and to broader society to promote access and to eliminate disparities in health and health care. To promote these fundamental principles, a series of professional responsibilities has been suggested (Table 1-1). These specific responsibilities represent practical, daily traits that benefit the physician’s own patients and society as a whole. Physicians who use these and other attributes to improve their patients’ satisfaction with care are not only promoting professionalism but also reducing their own risk for liability and malpractice. An interesting new aspect of professionalism is the increasing reliance on team approaches to medical care, as exemplified by physicians whose roles are defined by the location of their practice—historically in the intensive care unit or emergency department and more recently on the inpatient general hospital floor. Quality care requires coordination and effective communication across inpatient and outpatient sites among physicians who themselves now typically work defined hours.7 This transition from reliance on a single, always available physician to a team, ideally with a designated coordinator, places new challenges on physicians, the medical care system, and the medical profession.

TABLE 1-1  PROFESSIONAL RESPONSIBILITIES Commitment to: Professional competence Honesty with patients Patient confidentiality Maintaining appropriate relations with patients Improving the quality of care Improving access to care Just distribution of finite resources Scientific knowledge Maintaining trust by managing conflicts of interest Professional responsibilities From Brennan T, Blank L, Cohen J, et al. Medical professionalism in the new millennium: a physician charter. Ann Intern Med. 2002;1136:243-246.




SYMPTOMS Constitutional Fever Fatigue Poor appetite Weight loss Obesity Snoring, sleep disturbances

280 274 132 132, 219 220 100, 405

Tables 280-1 to 280-8 E-Table 274-1 Table 132-1 Figure 132-4; Tables 132-4, 219-1, 219-2 Figure 220-1 Table 405-6

Head, Eyes, Ears, Nose, Throat Headache Visual loss, transient Ear pain Hearing loss Ringing in ears (tinnitus) Vertigo Nasal congestion, rhinitis, or sneezing Loss of smell or taste Dry mouth Sore throat Hoarseness

398 423, 424 426 428 428 428 251, 426 427 425 429 429

Tables 398-1, 398-2 Tables 423-2, 424-1 Table 426-3 Figure 428-1 Figure 428-2 Figure 428-3 Figure 251-1; Table 251-2 Table 427-1 Table 425-7 Figure 429-2; Table 429-1

Cardiopulmonary Chest pain Bronchitis Shortness of breath Palpitations Dizziness Syncope Cardiac arrest Cough Hemoptysis

51, 137 96 51, 83 51, 62 51, 62, 428 62 63 83 83

Tables 51-2, 137-5, 137-6

132 132, 138 135, 153 132, 137, 138, 139

Figure 132-5; Table 132-5 Table 132-1 Figure 135-3; Table 135-1 Figures 132-6, 138-2; Tables 137-3, 137-4, 139-1

132, 142 132, 137 137, 140 135 136, 137 145 145

Figures 132-1, 132-2; Tables 132-2, 132-3, 142-1 Figure 132-3; Tables 132-2, 137-1 Figures 137-1, 140-1 to 140-4 Figures 135-3, 135-4, 135-6; Table 135-4 Figures 136-3, 136-5, 137-1; Table 136-2 Figure 145-5

284, 285 284 26 123 126 285 236 236 240 234 234 200 285

Tables 284-3, 284-5, 285-2 Table 284-3 Tables 26-1 to 26-3 Tables 123-1 to 123-3 Figure 126-1

Gastrointestinal Nausea and vomiting Dysphagia, odynophagia Hematemesis Heartburn/dyspepsia Abdominal pain   Acute   Chronic Diarrhea Melena, blood in stool Constipation Fecal incontinence Anal pain Genitourinary Dysuria Frequency Incontinence Urinary obstruction Renal colic vagin*l discharge Menstrual irregularities Female infertility Hot flushes Erectile dysfunction Male infertility Scrotal mass Genital ulcers or warts

Figure 83-3 Figure 62-1; Tables 51-4, 62-5 Figure 62-1; Table 428-1 Figure 62-1; Tables 62-1, 62-2, 62-4 Figures 63-2, 63-3 Figure 83-1; Tables 83-2, 83-3 Tables 83-6, 83-7

Figure 236-3; Tables 236-3, 236-4 Table 236-5 Table 240-1 Figure 234-10 Figures 234-8, 234-9; Table 234-7 Figure 200-1 Table 285-1






Musculoskeletal Neck or back pain


Figures 400-4, 400-5, 400-6; Tables 400-3 to 400-5

Painful joints


Figure 256-1; Tables 256-1, 256-3

Swollen feet, ankles, or legs   Bilateral   Unilateral

51 81

Figure 51-8 Figure 81-2; Table 81-2



Table 79-3

Acute limb ischemia


Figure 79-5; Table 79-1


396, 420, 421, 422

Tables 396-1, 420-2, 421-2, 421-4

Sensory loss

396, 420

Figure 420-1; Tables 420-1, 420-3 to 420-5

Memory loss


Figures 402-1, 402-2; Tables 402-1 to 402-6

Abnormal gait


Table 396-2



Tables 403-1 to 403-6

Abnormal bleeding


Table 171-1


436, 441

Figure 436-1; Tables 436-1 to 436-6, 441-5


252, 440

Figure 252-2; Tables 252-1, 440-1, 440-2

Abnormal pigmentation


Table 441-2

Alopecia and hirsutism


Tables 442-1, 442-3

Nail disorders


Table 442-4


280, 281

Figure 281-1; Tables 280-1 to 280-8, 281-2


8, 109

Table 109-4


8, 62, 64, 65

Figures 62-2, 62-3; Tables 64-4, 65-2



Table 67-5


8, 106

Figures 106-3, 108-1; Tables 106-1, 107-1, 107-2

Altered respiration

8, 86, 104

Tables 86-1, 86-2, 104-2

Eye pain


Table 423-3

Red eye


Tables 423-4, 423-6

Dilated pupil


Figure 424-4



Table 424-5



Table 424-2



Figure 424-6



Figure 147-2; Tables 147-1 to 147-3



Table 426-3


251, 426

Tables 251-3, 426-1, 426-2

Oral ulcers and discolorations


Tables 425-1 to 425-4

Salivary gland enlargement


Table 425-6

Neck mass


Figure 190-3



Tables 168-1 to 168-6

Thyroid nodule


Figure 226-4



Figures 226-1, 226-3




SIGNS Vital Signs

Head, Eyes, Ears, Nose, Throat


Breast Breast mass


Lungs Wheezes


Table 83-4

Heart murmur or extra sounds


Figure 51-6; Tables 51-7, 51-8

Jugular venous distention


Table 51-6

Carotid pulse abnormalities


Figure 51-5






Abdomen Hepatomegaly


Figure 146-5



Tables 168-7, 168-9

Acute abdomen

142, 143

Figure 143-1; Table 142-1

Abdominal swelling/ascites

142, 153

Table 153-3

Rectal bleeding/positive stool

135, 193

Figures 135-3, 135-4, 135-6; Table 135-4



Table 145-1



Figure 256-1



Figure 51-8





Figure 51-10



Figure 28-1; Tables 28-1, 28-2

Psychiatric disturbances


Tables 397-1 to 397-4, 397-6 to 397-8, 397-10, 397-11, 397-13, 397-14



Tables 404-1 to 404-4


407, 408

Figure 407-1; Tables 407-2, 407-3, 407-5, 407-6, 408-5, 408-6

Movement disorders

409, 410

Tables 409-4, 410-1 to 410-9



Figure 420-1; Tables 420-1 to 420-5, E-Table 420-1

Suspicious mole


Table 203-1

Nail diseases


Table 442-4



Tables 158-2 to 158-6



Figure 166-2; Table 166-4



Figure 167-4; Table 167-1



Table 167-3



Table 167-2



Figure 170-2; Table 170-1

Neutropenia   With fever

167 281

Figure 167-7; Tables 167-4 and 167-5 Figure 281-1



Figure 166-6; Table 166-6



Figure 172-1; Tables 172-1, 172-3

Prolonged PT or PTT


Figure 171-4


114, 120

Tables 114-2, 120-6

Abnormal liver enzymes


Figures 147-2 to 147-4

Elevated BUN/creatinine   Acute   Chronic

120 130

Figure 120-1; Tables 120-1 to 120-5 Table 130-1



Tables 229-1, 229-2



Tables 230-1, 230-2

Electrolyte abnormalities

116, 117

Figure 116-4; Tables 116-6, 116-7, 117-2, 117-3

Acid-base disturbances


Figures 118-1, 118-2; Tables 118-1 to 118-6



Figure 245-3; Tables 245-2 to 245-4



Figure 245-4; Table 245-6

Hypo- and hyperphosphatemia


Tables 119-2, 119-3

Magnesium deficiency


Table 119-1

Elevated Pco2


Figure 86-2

Solitary pulmonary nodule


Figure 191-2

Pleural effusion


Tables 99-4 to 99-6

ECG abnormalities


Tables 54-2 to 54-5



Skin and Nails



Chest Radiograph/ECG

BUN = blood urea nitrogen; ECG = electrocardiogram; PT = prothrombin time; PTT = partial thromboplastin time.

The changing medical care environment is placing increasing emphasis on standards, outcomes, and accountability. As purchasers of insurance become more cognizant of value rather than just cost (Chapter 12), outcomes ranging from rates of screening mammography (Chapter 198) to mortality rates with coronary artery bypass graft surgery (Chapter 74) become metrics by which rational choices can be made. Clinical guidelines and critical pathways derived from randomized controlled trials and evidence-based medicine can potentially lead to more cost-effective care and better outcomes. These major changes in many Western health care systems bring with them many major risks and concerns. If the concept of limited choice among physicians and health care providers is based on objective measures of quality and outcome, channeling of patients to better providers is one reasonable definition of better selection and enlightened competition. If the limiting of options is based overwhelmingly on cost rather than measures of quality, outcomes, and patient satisfaction, it is likely that the historical relationship between the patient and the truly professional physician will be fundamentally compromised. Another risk is that the same genetic information that could lead to more effective, personalized medicine will be used against the very people whom it is supposed to benefit—by creating a stigma, raising health insurance costs, or even making someone uninsurable. The ethical approach to medicine (Chapter 2), genetics (Chapter 40), and genetic counseling provides means to protect against this adverse effect of scientific progress. In this new environment, the physician often has a dual responsibility: to the health care system as an expert who helps create standards, measures of outcome, clinical guidelines, and mechanisms to ensure high-quality, costeffective care; and to individual patients who entrust their well-being to that physician to promote their best interests within the reasonable limits of the system. A health insurance system that emphasizes cost-effective care, that gives physicians and health care providers responsibility for the health of a population and the resources required to achieve these goals, that must exist in a competitive environment in which patients can choose alternatives if they are not satisfied with their care, and that places increasing emphasis on health education and prevention can have many positive effects. In this environment, however, physicians must beware of overt and subtle pressures that could entice them to underserve patients and abrogate their professional responsibilities by putting personal financial reward ahead of their patients’ welfare. The physician’s responsibility to represent the patient’s best interests and avoid financial conflicts by doing too little in the newer systems of capitated care provides different specific challenges but an analogous moral dilemma to the historical American system in which the physician could be rewarded financially for doing too much. In the current health care environment, all physicians and trainees must redouble their commitment to professionalism. At the same time, the challenge to the individual physician to retain and expand the scientific knowledge base and process the vast array of new information is daunting. In this spirit of a profession based on science and caring, Goldman-Cecil Medicine seeks to be a comprehensive approach to modern internal medicine. GENERAL REFERENCES For the General References and other additional features, please visit Expert Consult at


GENERAL REFERENCES 1. Martin R, Miquel S, Langella P, et al. The role of metagenomics in understanding the human microbiome in health and disease. Virulence. 2014;5:413-423. 2. Ganesh SK, Arnett DK, Assimes TL, et al. Genetics and genomics for the prevention and treatment of cardiovascular disease: update: a scientific statement from the American Heart Association. Circulation. 2013;128:2813-2851. 3. Paoletti C, Hayes DF. Molecular testing in breast cancer. Annu Rev Med. 2014;65:95-110.


4. The Global Burden of Disease Study 2010. Lancet. 2012-2013;380:2053-2260. 5. Walton M, Kerridge I. Do no harm: is it time to rethink the Hippocratic Oath? Med Educ. 2014;48:17-27. 6. Snyder L, for the American College of Physicians Ethics, Professionalism, and Human Rights Committee. American College of Physicians ethics manual: sixth edition. Ann Intern Med. 2012; 156:73-104. 7. O’Malley AS, Reschovsky JD. Referral and consultation communication between primary care and specialist physicians: finding common ground. Arch Intern Med. 2011;171:56-65.


CHAPTER 2  Bioethics in the Practice of Medicine  

2  BIOETHICS IN THE PRACTICE OF MEDICINE EZEKIEL J. EMANUEL It commonly is argued that modern advances in medical technology, antibiotics, dialysis, transplantation, and intensive care units have created the bioethical dilemmas that confront physicians in the 21st century. In reality, however, concerns about ethical issues are as old as the practice of medicine itself. The Hippocratic Oath, composed sometime around 400 bc, attests to the need

of ancient Greek physicians for advice on how to address the many bioethical dilemmas that they confronted. The Oath addresses issues of confidentiality, abortion, euthanasia, sexual relations between physician and patient, divided loyalties, and, at least implicitly, charity care and executions. Other Hippocratic works address issues such as termination of treatments to dying patients and telling the truth. Whether we agree with the advice dispensed or not, the important point is that many bioethical issues are not created by technology but instead are inherent in medical practice. Technology may make these issues more common and may change the context in which they arise, but many, if not most, bioethical issues that regularly confront physicians are timeless and inherent in the practice of medicine. Many physicians have been educated that four main principles can be invoked to address bioethical dilemmas: autonomy, nonmaleficence, beneficence, and justice. Autonomy is the idea that people should have the right and freedom to choose, pursue, and revise their own life plans. Nonmaleficence is the idea that people should not be harmed or injured knowingly; this principle is encapsulated in the frequently repeated phrase that a physician has an obligation to “first do no harm”—primum non nocere. This phrase is not found either in the Hippocratic Oath or in other Hippocratic writing; the only related, but not identical, Hippocratic phrase is “at least, do not harm.” Whereas nonmaleficence is about avoiding harm, beneficence is about the positive actions that the physician should undertake to promote the wellbeing of his or her patients. In clinical practice, this obligation usually arises from the implicit and explicit commitments and promises surrounding the physician-patient relationship. Finally, there is the principle of justice as the fair distribution of benefits and burdens. Although helpful in providing an initial framework, these principles have limited value because they are broad and open to diverse and conflicting interpretations. In addition, as is clear with the principle of justice, they frequently are underdeveloped. In any difficult case, the principles are likely to conflict. Conflicting ethical principles are precisely why there are bioethical dilemmas. The principles themselves do not offer guidance on how they should be balanced or specified to resolve the dilemma. These principles, which are focused on the individual physician-patient context, are not particularly helpful when the bioethical issues are institutional and systemic, such as allocating scarce vaccines or organs for transplantation or balancing the risks and benefits of mammograms for women younger than 50 years. Finally, these four principles are not comprehensive. Other fundamental ethical principles and values, such as communal solidarity, duties to future generations, trust, and professional integrity, are important in bioethics but not encapsulated except by deformation in these four principles. There is no formula or small set of ethical principles that mechanically or magically gives answers to bioethical dilemmas. Instead, medical practitioners should follow an orderly analytic process. First, practitioners need to obtain the facts relevant to the situation. Second, they must delineate the basic bioethical issue. Third, it is important to identify all the crucial principles and values that relate to the case and how they might conflict. Fourth, because many ethical dilemmas have been analyzed previously and subjected frequently to empirical study, practitioners should examine the relevant literature, whether it is commentaries or studies in medical journals, legal cases, or books. With these analyses, the particular dilemma should be reexamined; this process might lead to reformulation of the issue and identification of new values or new understandings of existing values. Fifth, with this information, it is important to distinguish clearly unethical practices from a range of ethically permissible actions. Finally, it is important not only to come to some resolution of the case but also to state clearly the reasons behind the decisions, that is, the interpretation of the principles used and how values were balanced. Although unanimity and consensus may be desirable ideals, reasonable people frequently disagree about how to resolve ethical dilemmas without being unethical or malevolent. A multitude of bioethical dilemmas arise in medical practice, including issues of genetics, reproductive choices, and termination of care. In clinical practice, the most common issues revolve around informed consent, termination of life-sustaining treatments, euthanasia and physician-assisted suicide, and conflicts of interest.



It commonly is thought that the requirement for informed consent is a relatively recent phenomenon. Suggestions about the need for a patient’s informed consent can be found as far back as Plato, however. The first

CHAPTER 2  Bioethics in the Practice of Medicine  

recorded legal case involving informed consent is the 1767 English case of Slater v. Baker and Stapleton, in which two surgeons refractured a patient’s leg after it had healed improperly. The patient claimed they had not obtained consent. The court ruled: [I]t appears from the evidence of the surgeon that it was improper to disunite the callous without consent; this is the usage and law of surgeons: then it was ignorance and unskillfulness in that very particular, to do contrary to the rule of the profession, what no surgeon ought to have done. Although there may be some skepticism about the extent of the information disclosed or the precise nature of the consent obtained, the notable fact is that an 18th-century court declared that obtaining prior consent of the patient is not only the usual practice but also the ethical and legal obligation of surgeons. Failure to obtain consent is incompetent and inexcusable. In contemporary times, the 1957 case of Salgo v. Leland Stanford Junior University Board of Trustees constitutes a landmark by stating that physicians have a positive legal obligation to disclose information about risks, benefits, and alternatives to patients; this decision popularized the term informed consent.

Definition and Justification

Informed consent is a person’s autonomous authorization of a physician to undertake diagnostic or therapeutic interventions for himself or herself. In this view, the patient understands that he or she is taking responsibility for the decision while empowering someone else, the physician, to implement it. However, agreement to a course of medical treatment does not necessarily qualify as informed consent. There are four fundamental requirements for valid informed consent: mental capacity, disclosure, understanding, and voluntariness. Informed consent assumes that people have the mental capacity to make decisions; disease, development, or medications can compromise patients’ mental capacity to provide informed consent. Adults are presumed to have the legal competence to make medical decisions, and whether an adult is incompetent to make medical decisions is a legal determination. Practically, physicians usually decide whether patients are competent on the basis of whether patients can understand the information disclosed, appreciate its significance for their own situation, and use logical and consistent thought processes in decision making. Incompetence in medical decision making does not mean a person is incompetent in all types of decision making and vice versa. Crucial information relevant to the decision must be disclosed, usually by the physician, to the patient. The patient should understand the information and its implications for his or her interests and life goals. Finally, the patient must make a voluntary decision (i.e., one without coercion or manipulation by the physician). It is a mistake to view informed consent as an event, such as the signing of a form. Informed consent is viewed more accurately as a process that evolves during the course of diagnosis and treatment. Typically, the patient’s autonomy is the value invoked to justify informed consent. Other values, such as bodily integrity and beneficence, have also been cited, especially in early legal rulings.

Empirical Data

Fairly extensive research has been done on informed consent. In general, studies show that in clinical situations, physicians frequently do not communicate all relevant information for informed decision making. In a study of audiotapes from 1057 outpatient encounters, physicians mentioned alternatives in only 11.3% of cases, provided pros and cons of interventions in only 7.8% of situations, and assessed the patient’s understanding of the information in only 1.5% of decisions. The more complex the medical decisions, the more likely it was that the elements of informed consent would be fulfilled. Importantly, data suggest that disclosure is better in research settings, both in the informed consent documents and in the discussions. For instance, in recorded interactions between researchers and prospective participants, the major elements of research, such as that the treatment was investigational and the risks and benefits of treatment, were disclosed in more than 80% of interactions. Greater disclosure in the research setting may be the consequence of requiring a written informed consent document that has been reviewed by an independent committee, such as an institutional review board or a research ethics committee. Some have suggested that for common medical interventions, such as elective surgery, standardized informed consent documents should include the risks and benefits as quantified in randomized controlled trials, relevant data on the surgeon, the institution’s clinical outcomes for the procedure, and a list of acceptable alternatives.1


Patients frequently fail to recall crucial information disclosed, although they usually think they have sufficient information for decision making. Whether patients fail to recall key information because they are overwhelmed by the information or because they do not find much of it salient to their decision is unclear. The issue is what patients understand at the point of decision making, not what they recall later. Studies aimed at improving informed consent in the clinical setting suggest that interactive media, such as videos and interactive computer software, can improve understanding by patients. A1  Conversely, data on shared decision making show that interactive media do not improve participants’ understanding, whereas more personal interaction, whether as an additional telephone call by a research nurse or as an additional face-to-face meeting, does enhance understanding.2 One of the most important results of empirical research on informed consent is the gap between information and decision making. Many studies show that most patients want information, but far fewer prefer decisionmaking authority. One study showed that most patients wanted information, but only about one third desired decision-making authority, and patients’ decision-making preferences were not correlated with their informationseeking preferences. Several investigators found that patients’ preference for decision-making authority increases with higher educational levels and declines with advancing age. Most important, the more serious the illness, the more likely patients are to prefer that physicians make the decisions. Several studies suggest that patients who have less of a desire to make their own decisions generally are more satisfied with how the decisions were made.

Practical Considerations

Implementing informed consent raises concerns about the extent of information to be disclosed and exceptions to the general requirement. A major area of ethical and legal disagreement has been what information to disclose and how to disclose it. As a practical matter, physicians should disclose at least six fundamental elements of information to patients: (1) diagnosis and prognosis; (2) nature of the proposed intervention; (3) alternative interventions, including no treatment; (4) risks associated with each alternative; (5) benefits of each alternative; and (6) likely outcomes of these alternatives (Table 2-1). Because risk is usually the key worry of physicians, it generally is recommended that physicians disclose (1) the nature of the risks, (2) their magnitude, (3) the probability that each risk will occur, and (4) when the consequence might occur.3 Increasingly, these disclosures should include data both from clinical trials as well as the actual data from the institution and physician performing the test and treatments. Some argue that minor risks need not be disclosed. In general, all serious risks, such as death, paralysis, stroke, infections, or chronic pain, even if rare, should be disclosed, as should common risks. The central problem is that the physician should provide this detailed information within reasonable time constraints and yet not overwhelm patients with complex information in technical language. The historical constraint of office time is no longer tenable. Interactive electronic media, which patients can view at home on their own time, can facilitate the transfer of information outside of the physician’s office. Different states have adopted two contrasting legal standards defining how much information should be disclosed. The physician or customary standard, adapted from malpractice law, states that the physician should disclose information “which a reasonable medical practitioner would make under the same or similar circ*mstances.” Conversely, the reasonable person or lay-oriented standard states that physicians should disclose all information that a “reasonable person in the patient’s circ*mstances would find material to” the medical decision. The physician standard is factual and can be determined empirically, but the patientoriented standard, which is meant to engage physicians with patients, is hypothetical. Currently, each standard is used by about half the states.

TABLE 2-1  FUNDAMENTAL ELEMENTS FOR DISCLOSURE TO PATIENTS Diagnosis and prognosis Nature of proposed intervention Reasonable alternative interventions Risks associated with each alternative intervention Benefits associated with each alternative intervention Probable outcomes of each alternative intervention


CHAPTER 2  Bioethics in the Practice of Medicine  

There are exceptions to the requirements of informed consent. In emergency situations, consent can be assumed because patients’ interests concentrate on survival and retaining maximal mental and physical functioning; as a result, reasonable persons would want treatment. In some circ*mstances, physicians may believe the process of informed consent could pose a serious psychological threat. In rare cases, the “therapeutic privilege” promoting a patient’s well-being trumps autonomy, but physicians should be wary of invoking this exception too readily. If patients are deemed incompetent, family members—beginning with spouse, children, parents, siblings, then more distant relatives—usually are selected as surrogates or proxies, although there may be concerns about conflicting interests or knowledge of the patient’s wishes. In the relatively rare circ*mstance in which a patient formally designated a proxy, that person has decision-making authority. The substituted judgment standard states that the proxy should choose what the patient would choose if he or she were competent. The best interests standard states that the proxy should choose what is best for the patient. Frequently, it is not clear how the patient would have decided because the situation was not discussed with the patient and he or she left no living will. Similarly, what is best for a patient is controversial because there are usually tradeoffs between quality of life and survival. These problems are exacerbated because a proxy’s predictions about a patient’s quality of life are poor; proxies tend to underestimate patients’ functional status and satisfaction. Similarly, proxy predictions are inaccurate regarding life-sustaining preferences when the patient is mentally incapacitated. Families tend to agree with patients about two thirds of the time in deciding whether to provide life-sustaining treatments if the patient became demented, when chance alone would generate agreement in 50% of the cases. Such confusion about how to decide for incapacitated patients can create conflicts among family members or between the family and medical providers. In such circ*mstances, an ethics consultation may be helpful.



Since the start of medicine, it has been viewed as ethical to withhold medical treatments from the terminally ill and “let nature take its course,” while keeping the patient as comfortable as possible.4 Hippocrates argued that physicians should “refuse to treat those [patients] who are overmastered by their disease.” In the 19th century, prominent American physicians advocated withholding of cathartic and emetic “treatments” from the terminally ill and using ether to ease pain at the end of life. John Collins Warren, who wrote Etherization: with Surgical Remarks in 1848, included a chapter on using ether to ease the pain of a cancer patient’s death. The editors of The Lancet, in 1900, argued that physicians should intervene to ease the pain of death and that they did not have an obligation to prolong a clearly terminal life. The contemporary debate on terminating care began in 1976 with the Quinlan case, in which the New Jersey Supreme Court ruled that patients had a right to refuse life-sustaining interventions on the basis of a right of privacy and that the family could exercise the right for a patient in a persistent vegetative state.

Definition and Justification

It generally is agreed that all patients have a right to refuse medical interventions. Ethically, this right is based on the patient’s autonomy and is implied by the doctrine of informed consent. Legally, state courts have cited the right to privacy, right to bodily integrity, or common law to justify the right to refuse medical treatment. In the 1990 Cruzan case and in the subsequent physician-assisted suicide cases, the U.S. Supreme Court affirmed that there is a “constitutionally protected right to refuse lifesaving hydration and nutrition.” The Court stated that “[A] liberty interest [based on the 14th Amendment] in refusing unwanted medical treatment may be inferred from our prior decisions.” All patients have a constitutional and an ethical right to refuse medical interventions. These rulings were the basis of the consistent state and federal court rulings to permit the husband to terminate artificial nutrition and hydration in the Schiavo case.

Empirical Data

Data show that termination of medical treatments is now the norm, and the trend has been to stop medical interventions more frequently based on the preferences of patients and their surrogate decision makers.5 More than 85% of Americans die without cardiopulmonary resuscitation, and more than 90% of decedents in intensive care units do not receive cardiopulmonary resuscitation. Of decedents in intensive care units, more than 85% die after the

withholding or withdrawal of medical treatments, with an average of 2.6 interventions being withheld or withdrawn per decedent. Despite extensive public support for use of advance care directives and the passage of the Patient Self-Determination Act mandating that health care institutions inform patients of their right to complete such documents, less than 30% of Americans have completed one.6 Even among severely or terminally ill patients, less than 50% have an advance directive in their medical record. Data suggest that more than 40% of patients required active decisionmaking about terminating medical treatments in their final days, but more than 70% lacked decision-making capacity, thereby emphasizing the importance of advance directives. Efforts to improve completion of advance care directives have generated mixed results. In La Crosse County, Wisconsin, for example, after health care organizations in the county added an “Advance Directive” section to their electronic medical records, 90% of decedents had some type of advance directive. Unfortunately, even successful pilot efforts like La Crosse County’s have not been adopted or easily scaled. A persistent problem has been that even when patients complete advance care directives, the documents frequently are not available, physicians do not know they exist, or they tend to be too general or vague to guide decisions. The increasing use of electronic health records should make it possible for advance directives to be available whenever and wherever the patient presents to a health care provider. Although electronic health records will help in making existing advance directives available, they will not solve the problem of actually having a conversation between the physician and the patient about advance care planning. Starting that conversation still seems to be a persistent barrier. Just as proxies are poor at predicting patients’ wishes, data show that physicians are probably even worse at determining patients’ preferences for lifesustaining treatments. In many cases, life-sustaining treatments are continued even when patients or their proxies desire them to be stopped. Conversely, many physicians discontinue or never begin interventions unilaterally without the knowledge or consent of patients or their surrogate decision makers. These discrepancies emphasize the importance of engaging patients early in their care about treatment preferences.

Practical Considerations

There are many practical considerations in enacting this right (Table 2-2). First, patients have a right to refuse any and all medical interventions, from blood transfusions and antibiotics to respirators, artificial hydration, and nutrition. Although initiation of cardiopulmonary resuscitation was the focus of the early court cases, this issue is viewed best as addressing just one of the many medical interventions that can be stopped or withheld. The question of what medical interventions can be terminated—or not started—is a recurrent topic of debate among physicians and other health care providers. The fact is that any treatment prescribed by a physician and administered by a health care provider can be stopped. The issue is not whether the treatment is ordinary, extraordinary, or heroic, or whether it is high technology or low technology. Treatments that can be stopped include not only ventilators, artificial nutrition, and hydration but also dialysis, pacemakers, ventricular assist devices, antibiotics, and any medications. Second, there is no ethical or legal difference between withholding an intervention and withdrawing it. If a respirator or other treatment is started because physicians are uncertain whether a patient would have wanted it, they always can stop it later when information clarifies the patient’s wishes. Although physicians and nurses might find stopping a treatment to be more difficult psychologically, withdrawal is ethically and legally permitted—and required—when it is consonant with the patient’s wishes. Third, competent patients have the exclusive right to decide about terminating their own care.7 If there is a conflict between a competent patient and his or her family, the patient’s wishes are to be followed. It is the patient’s right to refuse treatment, not the family’s right. For incompetent patients, the situation is more complex; if the patients left clear indications of their wishes, whether as explicit oral statements or as written advance care directives, these wishes should be followed. Physicians should not be overly concerned about the precise form patients use to express their wishes; because patients have a constitutional right to refuse treatment, the real concern is whether the wishes are clear and relevant to the situation. If an incompetent patient did not leave explicit indications of his or her wishes or designate a proxy decision maker, the physician should identify a surrogate decision maker and rely on the decision maker’s wishes while being cognizant of the potential problems noted. There is a potential problem in terminating life-sustaining care to patients who are permanently incompetent but still conscious. Some state courts have restricted what treatments a proxy decision maker can terminate,

CHAPTER 2  Bioethics in the Practice of Medicine  






Is there a legal right to refuse medical interventions?

Yes. The U.S. Supreme Court declared that competent people have a constitutionally protected right to refuse unwanted medical treatments based on the 14th Amendment.

What interventions can be legally and ethically terminated?

Any and all interventions (including respirators, antibiotics, pacemakers, ventricular assist devices, intravenous or enteral nutrition and hydration) can be legally and ethically terminated.

Is there a difference between withholding life-sustaining interventions and withdrawing them?

No. The consensus is that there is no important legal or ethical difference between withholding and withdrawing medical interventions. Stopping a treatment once begun is just as ethical as never having started it.

Whose view about terminating The views of a competent adult patient prevail. life-sustaining interventions It is the patient’s body and life. prevails if there is a conflict between the patient and family? Who decides about terminating life-sustaining interventions if the patient is incompetent?

Are advance care directives legally enforceable?

If the patient appointed a proxy or surrogate decision maker when competent, that person is legally empowered to make decisions about terminating care. If no proxy was appointed, there is a legally designated hierarchy, usually (1) spouse, (2) adult children, (3) parents, (4) siblings, and (5) available relatives. Yes. As a clear expression of the patient’s wishes, they are a constitutionally protected method for patients to exercise their right to refuse medical treatments. In almost all states, clear and explicit oral statements are legally and ethically sufficient for decisions about withholding or withdrawing medical interventions.

thereby requiring the incompetent patient to have given very specific instructions about the particular treatments he or she does not want to receive and the conditions under which care should be withheld or withdrawn. This requirement severely limits the authority and power of proxy decision makers in these cases. Fourth, the right to refuse medical treatment does not translate into a right to demand any treatment, especially treatments that have no pathophysiologic rationale, have already failed, or are known to be harmful. Futility has become a justification to permit physicians unilaterally to withhold or withdraw treatments despite the family’s requests for treatment. Some states, such as Texas, have enacted futility laws, which prescribe procedures by which physicians can invoke futility either to transfer a patient or to terminate interventions. However, the principle of futility is not easy to implement in medical practice. Initially, some commentators advocated that an intervention was futile when the probability of success was 1% or lower. Although this threshold seems to be based on empirical data, it is a covert value judgment. Because the declaration of futility is meant to justify unilateral determinations by physicians, it generally has been viewed as an inappropriate assertion that undermines physician-patient communication and violates the principle of shared decision making. Similar to the distinction between ordinary and extraordinary, futility is viewed increasingly as more obfuscating than clarifying, and it is being invoked much less often.



Since Hippocrates, euthanasia and physician-assisted suicide have been controversial issues. In 1905, a bill was introduced into the Ohio legislature to legalize euthanasia; it was defeated. In the mid-1930s, similar bills were introduced and defeated in the British Parliament and the Nebraska legislature. As of January 2014, physician-assisted suicide is legal in Oregon and Washington State, based on statewide public referenda, and in Vermont, based on legislation passed in May 2013. Both euthanasia and physician-assisted



Voluntary active euthanasia

Intentional administration of medications or other interventions to cause the patient’s death with the patient’s informed consent

Involuntary active euthanasia

Intentional administration of medications or other interventions to cause the patient’s death when the patient was competent to consent but did not consent (e.g., the patient may not have been asked)

Nonvoluntary active Intentional administration of medications or other euthanasia interventions to cause the patient’s death when the patient was incompetent and was mentally incapable of consenting (e.g., the patient might have been in a coma) Passive euthanasia

Withholding or withdrawal of life-sustaining medical treatments from a patient to let him or her die (termination of life-sustaining treatments)—a poor term that should not be used

Indirect euthanasia

Administration of narcotics or other medications to relieve pain with the incidental consequence of causing sufficient respiratory depression to result in the patient’s death

Physician-assisted suicide

A physician provides prescription medications or other interventions to a patient with the understanding that the patient can use them to commit suicide

suicide are legal in the Netherlands, Belgium, and Luxembourg, and physician-assisted suicide is legal in Switzerland. The Montana Supreme Court did not recognize a constitutional right to physician-assisted suicide, but it ruled that the law permitting the termination of life-sustaining treatment protected physicians from prosecution if they helped hasten the death of a consenting, rational, terminally ill patient.

Definition and Justification

The terms euthanasia and physician-assisted suicide8 require careful definition (Table 2-3). So-called passive and indirect euthanasia are misnomers and are not instances of euthanasia, and both are deemed ethical and legal. There are four arguments against permitting euthanasia and physicianassisted suicide. First, Kant and Mill thought that autonomy did not permit the voluntary ending of the conditions necessary for autonomy, and as a result, both philosophers were against voluntary enslavement and suicide. Consequently, the exercise of autonomy cannot include the ending of life because that would mean ending the possibility of exercising autonomy. Second, many dying patients may have pain and suffering because they are not receiving appropriate care, and it is possible that adequate care would relieve much pain and suffering (Chapter 3). Although a few patients still may experience uncontrolled pain and suffering despite optimal end-of-life care, it is unwise to use the condition of these few patients as a justification to permit euthanasia or physician-assisted suicide for any dying patient. Third, there is a clear ethical distinction between intentional ending of a life and termination of life-sustaining treatments. The actual acts are different— injecting a life-ending medication, such as a muscle relaxant, or providing a prescription for one is not the same as removing or refraining from introducing an invasive medical intervention. Finally, adverse consequences of permitting euthanasia and physician-assisted suicide must be considered. There are disturbing reports of involuntary euthanasia in the Netherlands and Belgium, and many worry about coercion of expensive or burdensome patients to accept euthanasia or physician-assisted suicide. Permitting euthanasia and physician-assisted suicide is likely to lead to further intrusions of lawyers, courts, and legislatures into the physician-patient relationship. There are four parallel arguments for permitting euthanasia and physicianassisted suicide. First, it is argued that autonomy justifies euthanasia and physician-assisted suicide. To respect autonomy requires permitting individuals to decide when it is better to end their lives by euthanasia or physicianassisted suicide. Second, beneficence—furthering the well-being of individuals—supports permitting euthanasia and physician-assisted suicide. In some cases, living can create more pain and suffering than death; ending a painful life relieves more suffering and produces more good. Just the reassurance of having the option of euthanasia or physician-assisted suicide, even if people do not use it, can provide “psychological insurance” and be


CHAPTER 2  Bioethics in the Practice of Medicine  

beneficial to people. Third, euthanasia and physician-assisted suicide are no different from termination of life-sustaining treatments that are recognized as ethically justified. In both cases, the patient consents to die; in both cases, the physician intends to end the patient’s life and takes some action to end the patient’s life; and in both cases, the final result is the same: the patient’s death. With no difference in the patient’s consent, the physician’s intention, or the final result, there can be no difference in the ethical justification. Fourth, the supposed slippery slope that would result from permitting euthanasia and physician-assisted suicide is not likely. The idea that permitting euthanasia and physician-assisted suicide would undermine the physicianpatient relationship or lead to forced euthanasia is completely speculative and not borne out by the available data. In its 1997 decisions, the U.S. Supreme Court stated that there is no constitutional right to euthanasia and physician-assisted suicide but that there also is no constitutional prohibition against states legalizing these interventions. Consequently, the legalization of physician-assisted suicide in Oregon, Vermont, and Washington State was constitutional.

Empirical Data

Attitudes and practices related to euthanasia and physician-assisted suicide have been studied extensively. First, surveys consistently indicate that between 50 and 80% of the American and British public support legalizing euthanasia and physician-assisted suicide for terminally ill patients who are suffering intractable pain.9 However, public support declines significantly for euthanasia and physician-assisted suicide in other circ*mstances, such as for psychological reasons.10 Physicians tend to be much less supportive of euthanasia and physician-assisted suicide, with oncologists, palliative care physicians, and geriatricians among the least supportive. Among American and British physicians, the majority opposes legalizing either practice. Second, approximately 25% of American physicians have received requests for euthanasia or physician-assisted suicide, including about 50% of oncologists. Third, multiple studies indicate that less than 5% of American physicians have performed euthanasia or physician-assisted suicide. Among oncologists, 4% have performed euthanasia and 11% have performed physician-assisted suicide during their careers. Fourth, in many cases, the safeguards are violated. One study found that in 54% of euthanasia cases, it was the family who made the request; in 39% of euthanasia and 19% of physician-assisted suicide cases, the patient was depressed; in only half of the cases was the request repeated. In the Netherlands and Belgium, where euthanasia and physician-assisted suicide are legal, less than 2% of all deaths are by these measures, with 0.4 to 1.8% of all deaths as the result of euthanasia without the patient’s consent.11 Since the practice of assisted suicide was legalized in Oregon in 1997, a cumulative 0.2% of all deaths are by physician-assisted suicide. Counterintuitively, data indicate that it is not pain that primarily motivates requests for euthanasia or physician-assisted suicide but rather psychological distress, especially depression and hopelessness. Interviews with physicians and with patients with amyotrophic lateral sclerosis, cancer, or infection with human immunodeficiency virus show that pain is not associated with interest in euthanasia or physician-assisted suicide; instead, depression and hopelessness are the strongest predictors of interest. Studies of patients in Australia and the Netherlands confirm the importance of depression in motivating requests for euthanasia. The desire to avoid dependence and loss of dignity are key motivations. Finally, data from the Netherlands and the United States suggest that there are significant problems in performing euthanasia and physician-assisted suicide. Dutch researchers reported that physician-assisted suicide causes complications in 7% of cases, and in 15% of cases, the patients did not die, awoke from coma, or vomited up the medication. Ultimately, in nearly 20% of physician-assisted suicide cases, the physician ended up injecting the patient with life-ending medication, converting physician-assisted suicide to euthanasia. These data raise serious questions about how to address complications of physician-assisted suicide when euthanasia is illegal or unacceptable.

interventions; (3) there should be a waiting period to ensure that the patient’s desire for euthanasia or physician-assisted suicide is stable and sincere; and (4) the physician should obtain a second opinion from an independent physician. Oregon and Washington State require patients to be terminally ill, whereas the Netherlands, Belgium, and Switzerland have no such requirement. Although there have been some prosecutions in the United States, there have been no convictions—except for Dr. Kevorkian—when physicians and others have participated in euthanasia and physician-assisted suicide.



Worrying about how payment and fees affect medical decisions is not new. In 1899, a physician reported that more than 60% of surgeons in Chicago were willing to provide a 50% commission to physicians for referring cases. He subsequently argued that in some cases, this fee splitting led to unnecessary surgical procedures. A 1912 study by the American Medical Association confirmed that fee splitting was a common practice. Selling patent medicines and patenting surgical instruments were other forms of financial conflicts of interest thought to discredit physicians a century ago. In the 1990s, the ethics of capitation for physician services and pharmaceutical prescriptions and payments by pharmaceutical and biotechnology companies to clinical researchers and practitioners raised the issue of financial conflicts of interest.

Definition and Justification

It commonly is argued that physicians have certain primary interests: (1) to promote the well-being of their patients, (2) to advance biomedical research, (3) to educate future physicians, and, more controversially, (4) to promote public health (Table 2-4). Physicians also have other, secondary interests, such as earning income, raising a family, contributing to the profession, and pursuing avocational interests, such as hobbies. These secondary interests are not evil; typically, they are legitimate, even admirable. A conflict of interest occurs when one of these secondary interests compromises pursuit of a primary interest, especially the patient’s well-being. Conflicts of interest are problematic because they can or appear to compromise the integrity of physicians’ judgment, compromising the patient’s well-being or research integrity. Conflict of interest can induce a physician to do something—perform a procedure, fail to order a test, or distort data—that would not be in a patient’s best interest. These conflicts can undermine the trust of patients and the public, not only in an individual physician but also in the entire medical profession. Even the appearance of conflicts of interest can be damaging because it is difficult for patients and the public “to determine what motives have influenced a professional decision.” The focus is on financial conflicts of interest, not because they are worse than other types of conflicts, but rather because they are more pervasive and more easily identified and regulated compared with other conflicts. Since ancient times, the ethical norm on conflicts has been clear: the physician’s primary obligation is to patients’ well-being, and a physician’s personal financial well-being should not compromise this duty.

Empirical Data

Financial conflicts are not rare but are frequently under-reported.12 The increased use of medical services and escalating health care spending, sometimes without clear benefit to patients, have been linked, at least statistically, to ownership of imaging facilities and referral to specialty hospitals owned by physicians. In Florida, it was estimated that nearly 40% of physicians were involved as owners of freestanding facilities to which they referred patients. In one study, 4 to 4.5 times more imaging examinations were ordered by self-referring physicians than by physicians who referred patients to radiologists. Similarly, patients referred to joint-venture physical therapy facilities have an average of 16 visits compared with 11 at non–joint-venture facilities. A recent study of urologists found that those who had integrated radiation

Practical Considerations

There is widespread agreement that if euthanasia and physician-assisted suicide are used, they should be considered only after all reasonable attempts at physical and psychological palliation have failed. A series of safeguards have been developed and embodied in the Oregon and the Dutch procedures, as follows: (1) the patient must be competent and must request euthanasia or physician-assisted suicide repeatedly and voluntarily; (2) the patient must have pain or other suffering that cannot be relieved by optimal palliative

TABLE 2-4  PRIMARY INTERESTS OF PHYSICIANS Promotion of the health and well-being of their patients Advancement of biomedical knowledge through research Education of future physicians and health care providers Promotion of the public health

facilities into their practices increased their use of the radiation by 2.5 times compared with urologists who did not have financial relationships with radiation facilities.13 There are no comparable data on the influence of capitation on physicians’ judgment. Similarly, multiple studies have shown that interaction with pharmaceutical representatives can lead to prescribing of new drugs, nonrational prescribing, and decreased use of generic drugs by physicians. Industry funding for continuing medical education payment for travel to educational symposia increases prescribing of the sponsor’s drug. Regarding researcher conflicts of interest, the available data suggest that corporate funding does not compromise the design and methodology of clinical research; in fact, commercially funded research may be methodologically more rigorous than government- or foundation-supported research. Conversely, data suggest that financial interests do distort researchers’ interpretation of data. The most important impact of financial interests, however, appears to be on dissemination of research studies. Growing evidence suggests the suppression or selective publication of data unfavorable to corporate sponsors but the repeated publication of favorable results.

bioethical issues. Because these tests have serious implications for the patient and others, scrupulous attention to informed consent must occur. The bioethical issues raised by genetic tests for somatic cell changes, such as tests that occur commonly in cancer diagnosis and risk stratification, are no different from the issues raised with the use of any laboratory or radiographic test. In some cases, ethics consultation services may be of assistance in resolving bioethical dilemmas, although current data suggest that consultation services are used mainly for problems that arise in individual cases and are not used for more institutional or policy problems.

Practical Considerations

For the General References and other additional features, please visit Expert Consult at

First, financial conflicts of interest are inherent in any profession when the professional earns income from rendering a service. Second, conflicts come in many different forms, from legitimate payment for services rendered to investments in medical laboratories and facilities, drug company dinners and payment for attendance at meetings, payment for enrolling patients in clinical research trials, and consultation with companies. Third, in considering how to manage conflicts, it is important to note that people are poor judges of their own potential conflicts. Individuals often cannot distinguish the various influences that guide their judgments, do not think of themselves as bad, and do not imagine that payment shapes their judgments. Physicians tend to be defensive about charges of conflicts of interest. In addition, conflicts tend to act insidiously, subtly changing practice patterns so that they then become what appear to be justifiable norms. Fourth, rules—whether laws, regulations, or professional standards—to regulate conflicts of interest are based on two considerations: (1) the likelihood that payment or other secondary interests would create a conflict and (2) the magnitude of the potential harm if there is compromised judgment. Rules tend to be of three types: (1) disclosure of conflicts, (2) management of conflicts, and (3) outright prohibition. Federal law bans certain types of self-referral of physicians in the Medicare program. The American Medical Association and the Pharmaceutical Research and Manufacturers of America have established joint rules that permit physicians to accept gifts of minimal value but “refuse substantial gifts from drug companies, such as the costs of travel, lodging, or other personal expenses . . . for attending conferences or meetings.” Additionally, the Physician Payment Sunshine Act, which was passed in 2010 as part of the Affordable Care Act and went into effect in August 2013, requires that drug and device manufacturers report all payments and transfers of value given to physicians to the Centers for Medicare and Medicaid Services so that information can be published on a searchable public website. Fifth, there is much emphasis on disclosure of conflicts, with the implicit idea being that sunshine is the best disinfectant. Disclosure may be useful in publications, but it is unclear whether this is a suitable safeguard in the clinical setting. Disclosure just may make patients worry more. Patients may have no context in which to place the disclosure or to evaluate the physician’s clinical recommendation, and patients may have few other options in selecting a physician or getting care, especially in an acute situation. Furthermore, self-disclosure often is incomplete, even when required. Finally, some conflicts can be avoided by a physician’s own action. Physicians can refuse to engage in personal investments in medical facilities or to accept gifts from pharmaceutical companies at relatively little personal cost. In other circ*mstances, the conflicts may be institutionalized, and minimizing them can occur only by changing the way organizations structure reimbursem*nt incentives. Capitation encourages physicians to limit medical services, and its potentially adverse effects are likely to be managed by institutional rules rather than by personal decisions.


In the near future, as genetics moves from the research to the clinical setting, practicing physicians are likely to encounter issues surrounding genetic testing, counseling, and treatment. The use of genetic tests without the extensive counseling so common in research studies would alter the nature of the

Grade A Reference A1. Stacey D, Légaré F, Bennett CL, et al. Decision aids for people facing health treatment or screening decisions. Cochrane Database Syst Rev. 2014;1:CD001431.


CHAPTER 2  Bioethics in the Practice of Medicine  

GENERAL REFERENCES 1. Krumholz HM. Informed consent to promote patient-centered care. JAMA. 2010;303:1190-1191. 2. Flory J, Emanuel E. Interventions to improve research participants’ understanding in informed consent for research: a systematic review. JAMA. 2004;292:1593-1601. 3. Caring Connections. Accessed February 9, 2015. 4. Education in Palliative and End-of-life Care. Accessed February 9, 2015. 5. Silveira MJ, Kim SY, Langa KM. Advance directives and outcomes of surrogate decision making before death. N Engl J Med. 2010;362:1211-1218. 6. Rao JK, Anderson LA, Lin FC, et al. Completion of advance directives among U.S. consumers. Am J Prev Med. 2014;46:65-70. 7. Loggers ET, Starks H, Shannon-Dudley M, et al. Implementing a Death with Dignity program at a comprehensive cancer center. N Engl J Med. 2013;368:1417-1424.


8. Boudreau JD, Somerville MA, Biller-Andorno N. Clinical decisions. Physician-assisted suicide. N Engl J Med. 2013;368:1450-1452. 9. Seale C. Legalisation of euthanasia or physician-assisted suicide: survey of doctors’ attitudes. Palliat Med. 2009;23:205-212. 10. Emanuel EJ. Euthanasia and physician-assisted suicide: a review of the empirical data from the United States. Arch Intern Med. 2002;162:142-152. 11. van der Heide A, Onwuteaka-Philipsen BD, Rurup ML, et al. End-of-life practices in the Netherlands under the Euthanasia Act. N Engl J Med. 2007;356:1957-1965. 12. Okike K, Kocher MS, Wei EX, et al. Accuracy of conflict-of-interest disclosures reported by physicians. N Engl J Med. 2009;361:1466-1474. 13. Mitchell JM. Urologists’ use of intensity-modulated radiation therapy for prostate cancer. N Engl J Med. 2013;369:1629-1637.


CHAPTER 3  Care of Dying Patients and Their Families  

3  CARE OF DYING PATIENTS AND THEIR FAMILIES ROBERT ARNOLD By 2030, 20% of the U.S. population will be older than 65 years, and people older than 85 years constitute the fastest growing segment of the population. Owing to successes in public health and medicine, many of these people will live the last years of their lives with chronic medical conditions such as cirrhosis, end-stage kidney disease, heart failure, and dementia. Even human immunodeficiency virus (HIV) and many cancers, once considered terminal, have turned into chronic diseases. The burden associated with these illnesses and their treatments is high. Chronically ill patients report multiple physical and psychological symptoms that lower their quality of life. The economic pressures associated with medical care adversely affect patients’ socioeconomic status and cause family stress, especially among caregivers, who spend 20 or more hours a week helping their loved ones. Palliative care, which was developed to decrease the burden associated with chronic illness, emphasizes patient- and family-centered care that optimizes quality of life by anticipating, preventing, and treating suffering. Palliative care throughout the continuum of illness addresses physical, intellectual, emotional, social, and spiritual needs while facilitating the patient’s autonomy, access to information, and choice. Palliative care and services, which are coordinated by an interdisciplinary team, are available concurrently with or independent of curative or life-prolonging care. Palliative and nonpalliative health care providers should collaborate and communicate about care needs while focusing on peace and dignity throughout the course of illness, during the dying process, and after death. Five points deserve special emphasis. First, palliative care can be delivered at any time during the course of an illness and is often provided concomitantly with disease-focused, life-prolonging therapy. Waiting until a patient is dying to provide palliative care is a serious error. For example, most elderly patients with chronic incurable illnesses, who might benefit from palliative care, are in the last 10 years of their lives but do not consider themselves to be dying. If palliative care is to have an impact on patients’ lives, it should be provided earlier in a patient’s illness, in tandem with other treatments. A1  Second, prediction is an inexact science. Although many cancers have a predictable trajectory in the last 3 to 6 months of life, for most illnesses, doctors rarely can accurately predict whether a patient is in the last 6 months


CHAPTER 3  Care of Dying Patients and Their Families  


of life (E-Fig. 3-1). Third, palliative care primarily focuses on the illness’s burden rather than treating the illness itself. Because these burdens can be physical, psychological, spiritual, or social, good palliative care requires a multidisciplinary approach. Fourth, palliative care takes the family unit as the central focus of care. Treatment plans must be developed for both the patient and the family. Fifth, palliative care recognizes that medical treatments are not uniformly successful and that patients die. At some point in a patient’s illness, the treatments may cause more burden than benefit. Palliative care recognizes this reality and starts with a discussion of the patient’s goals and the development of an individualized treatment plan. Many people confuse palliative care with hospice—an understandable confusion because hospices epitomize the palliative care philosophy. The two, however, are different. In the United States, hospice provides palliative care, primarily at home, for patients who have a life expectancy of 6 months or less and who are willing to forgo life-prolonging treatments. However, the requirement that patients must have a life expectancy of 6 months or less limits hospice’s availability, as does the requirement that patients give up expensive and potentially life-prolonging treatments. Moreover, because doctors and patients often are unwilling to cease these treatments until very late in the disease course, so are most patients. Palliative care is both a subspeciality and a domain of good internal medicine.2 Given the need for palliative care, every clinician must be able to provide basic palliative care, and subspecialties such as oncology need special expertise.


Palliative care is a holistic discipline with physical, psychological, spiritual, existential, social, and ethical domains. When caring for patients with chronic life-limiting illness, good palliative care requires that the following questions be addressed:

Is the Patient Physically Comfortable?

Across many chronic conditions, patients have a large number of inadequately treated physical symptoms (Table 3-1). The reasons are multifactorial and range from inadequate physician education, to societal beliefs regarding the inevitability of suffering in chronic illness, to public concerns regarding opioids, to the lack of evidence-based treatments in noncancer patients. The first step to improve symptom management is a thorough assessment. Standardized instruments such as the Brief Pain Inventory (Fig. 3-1) measure both the patient’s symptoms and the effect of those symptoms on the patient’s life. Use of standardized instruments assures that physicians will identify overlooked or underreported symptoms and, as a result, will enhance the satisfaction of both the patient and family. The evidence for the treatment of end-stage symptoms continues to improve. The use of nonsteroidal anti-inflammatory agents and opioids A2  can result in effective pain management in more than 75% of patients with cancer. Advances such as intrathecal pumps and neurolytic blocks are helpful in the remaining 25% (Chapter 30). The use of oxygen is not helpful for refractory dyspnea except when hypoxia has been documented A3 , whereas use of medications for depression often can be helpful A4  (Chapter 397).

Is the Patient Psychologically Suffering?

Patients may be physically comfortable but still suffering. Psychological symptoms and syndromes such as depression, delirium, and anxiety are common in patients with life-limiting or chronic illnesses. It may be difficult to determine whether increased morbidity and mortality are caused by the physical effects of the illness or by the psychological effects of depression and anxiety on energy, appetite, or sleep. Screening questions focusing on mood (e.g., “Have you felt down, depressed, and hopeless most of the time for the past 2 weeks?”) and anhedonism (e.g., “Have you found that little brings you pleasure or joy in the past 2 weeks?”) have been shown to help in diagnosing depression in this population. Increasing data show that treatment of depression in chronic illness is possible and improves both morbidity and mortality. A4-A6  For patients and families facing mortality, existential and spiritual concerns are common. Progressive illness often raises questions of love, legacy, loss, and meaning. A physician’s role is not to answer these questions or to provide reassurance, but rather to understand concerns of the patient and family, how they are coping, and what resources might help. Spirituality often is a source of comfort, and physicians can ascertain a patient’s beliefs using a brief instrument such as the FICA Spiritual Assessment Tool (Table 3-2). A single

screening question such as “Are you at peace?” may identify patients who are in spiritual distress and facilitate referrals to chaplains.

Is the Family Suffering?

Families, defined broadly as those individuals who care most for the patient, are an important source of support for most patients. Families provide informal caregiving, often at the expense of their own physical, economic, and psychological health. Good palliative care requires an understanding of how the family is coping and a search for ways to provide family members with the social or clinical resources they need to improve their well-being. Comprehensive and individually targeted interventions can reduce caregivers’ burdens, although the absolute benefits are relatively small. Because patients in palliative care often die, the palliative care team must address bereavement and postdeath family suffering. Good communication and informational brochures in an intensive care unit can decrease family members’ adverse psychological outcomes after death. A7  A letter of condolence or a follow-up phone call to the next of kin after a patient’s death is respectful and offers the opportunity to clarify questions about the patient’s care. Some family members suffer from complicated grief— a recently described syndrome associated with separation and traumatic distress, with symptoms persisting for more than 6 months. Primary care physicians, who have ongoing relationships with the loved one, and hospices, which provide bereavement services for a year after the patient’s death, have the opportunity to assess whether the grief symptoms persist or worsen.

Is the Patient’s Care Consistent with the Patient’s Goals?

The sine qua non for palliative care is ensuring that the treatment plan is consistent with the patient’s values. In one European cohort of elderly patients, most preferred longevity over quality of life, and half wanted resuscitation if necessary.3 However, a large proportion of elderly, seriously ill patients are not focused on living as long as possible. Instead, they want to maintain a sense of control, relieve their symptoms, improve their quality of life, avoid being a burden on their families, and have a closer relationship with their loved ones. Ensuring that treatment is consistent with a patient’s goals requires good communication skills (Table 3-3). The approaches to giving bad news, discussing goals of care, and talking about forgoing life-sustaining treatment have similar structures (Table 3-4). First, the patient needs to understand the basic facts about the diagnosis, possible treatments, and prognosis. The communication skill that helps physicians communicate information is Ask-TellAsk—exploring what the patient knows or wants to know, then explaining or answering questions, and then providing an opportunity for the patient to ask more. In the hospital, where discontinuity of care is common and misunderstandings frequent, it is important to determine what the patient knows before providing information so as to keep everyone well coordinated. When giving bad news, knowing what the patient knows allows the physician to anticipate the patient’s reaction. Finally, information must be titrated based on the patient’s preferences. Although most patients want to hear everything about their disease, a minority do not. There is no foolproof way to ascertain what any patient wants to know other than by asking. When giving patients information, it is important to give small pieces of information, not use jargon, and check the patient’s understanding.4 Giving information is like dosing a medication: one gives information, checks understanding, and then gives more information based on what the patient has heard. After ensuring that the doctor and the patient have a shared understanding of the medical facts, the physician should engage in an open-ended conversation about the patient’s goals as the disease progresses. This strategy requires that the patient be asked about both hopes and fears. One might ask: “What makes life worth living for you?” “If your time is limited, what are the things that are most important to achieve?” “What are your biggest fears or concerns?” “What would you consider to be a fate worse than death?” The clinician can use an understanding of these goals to make recommendations about which treatments to provide and which treatments would not be helpful. As a result, early palliative care can improve quality of life, mood, and even survival. Physicians find talking about prognosis particularly difficult for two reasons: first, it is hard to foretell the future accurately; and second, they fear this information will “take away patients’ hope.” Thus, they often avoid talking to patients about these issues unless specifically asked. Although some patients do not want to hear prognostic information, for many patients, this

CHAPTER 3  Care of Dying Patients and Their Families  

Mostly cancer




Death Low

Time Short period of evident decline

High Chronic, consistent with usual role

Mostly heart and lung failure




Time Long-term limitations with intermittent serious episodes

Chronic, progressive, eventually fatal illness

Mostly frailty and dementia




Time Prolonged dwindling

E-FIGURE 3-1.  Different disease trajectories for different illnesses. Permission obtained from RAND Corporation © Lynn J. Perspectives on care at the close of life. Serving patients who may die soon and their families: the role of hospice and other services. JAMA. 2001;285:925-932.

CHAPTER 3  Care of Dying Patients and Their Families  






How severe is the symptom (as assessed with the use of validated instruments) and how does it interfere with the patient’s life? What is the etiology of the pain? Is the pain assumed to be neuropathic or somatic? What has the patient used in the past (calculate previous days’ equal analgesic dose)?

Prescribe medications to be administered on a standing or regular basis if pain is frequent. For mild pain: use acetaminophen or a nonsteroidal anti-inflammatory agent (see Table 30-3). For moderate pain: titrate short-acting opioids (see Table 30-4). For severe pain: rapidly titrate short-acting opioids until pain is relieved or intolerable side effects develop; start long-acting opiates once pain is controlled. Rescue doses: prescribe immediate-release opioids—10% of the 24-hour total opiate every hour (orally) or every 30 minutes (parenterally) as needed. Concomitant analgesics (e.g., corticosteroids, anticonvulsants, tricyclic antidepressants, and bisphosphonates) should be used when applicable (particularly for neuropathic pain). Consider alternative medicine and interventional treatments for pain.


Is the patient taking opioids? Does the patient have a fecal impaction?

Prescribe laxatives for all patients on opiates. If ineffective, add drugs from multiple classes (e.g., stimulant, osmotic laxatives, and enemas). Prescribe methylnaltrexone if still constipated.

Shortness of breath

Ask the patient to assess the severity of the shortness of breath. Does the symptom have reversible causes?

Prescribe oxygen to treat hypoxia-induced dyspnea, but not if the patient is not hypoxic. Opioids relieve breathlessness without measurable reductions in respiratory rate or oxygen saturation; effective doses are often lower than those used to treat pain. Aerosolized opiates do not work. Fans or cool air may work through a branch of the trigeminal nerve. Consider anxiolytics (e.g., low-dose benzodiazepines) and use reassurance, relaxation, distraction, and massage therapy.


Is the patient too tired to do activities of daily living? Is the fatigue secondary to depression? Is a disease process causing the symptom or is it secondary to reversible causes?

Provide cognitive education about conserving energy use. Treat underlying conditions appropriately.


Which mechanism is causing the symptom (e.g., stimulation of the chemoreceptor trigger zone, gastric stimulation, delayed gastric emptying or “squashed stomach” syndrome, bowel obstruction, intracranial processes, or vestibular vertigo)? Is the patient constipated?

Prescribe an agent directed at the underlying cause (Chapter 132). If persistent, give antiemetic around the clock. Multiple agents directed at various receptors or mechanisms may be required.

Anorexia and cachexia

Is a disease process causing the symptom, or is it secondary to other symptoms (e.g., nausea and constipation) that can be treated? Is the patient troubled by the symptom or is the family worried about what not eating means?

A nutritionist may help find foods that are more appetizing (Chapter 213). Provide counseling about the prognostic implications of anorexia (Chapter 219).


Is the confusion acute, over hours to days? Does consciousness wax and wane? Are there behavioral disturbances, marked by a reduced clarity in the patient’s awareness of the environment, e.g., a problem of attention? Does the patient have disorganized thinking? Does the patient have an altered level of consciousness—either agitated or drowsy? Is there a reversible reason for the delirium? D: Drugs (opioids, anticholinergics, sedatives, benzodiazepines, steroids, chemotherapies and immunotherapies, some antibiotics) E: Eyes and Ears (poor vision and hearing, isolation) L: Low-flow states (hypoxia, myocardial infarction, congestive heart failure, chronic obstructive pulmonary disease, shock) I: Infections R: Retention (urine/stool), Restraints I: Intracranial (central nervous system metastases, seizures, subdural, cerebrovascular accident, hypertensive encephalopathy) U: Underhydration, Undernutrition, Undersleep M: Metabolic disorders (sodium, glucose, thyroid, hepatic, deficiencies of vitamin B12, folate, niacin, and thiamine) and toxic (lead, manganese, mercury, alcohol)

Identify underlying causes and manage symptoms (Chapter 28). Recommend behavioral therapies, including avoidance of excess stimulation, frequent reorientation, and reassurance. Ensure presence of family caregivers and explain delirium to them. Prescribe haloperidol, risperidone, or olanzapine.


Have you felt down, depressed, or hopeless most of the time during the past 2 weeks? Have you found that little brings you pleasure or joy during the past 2 weeks? (Somatic symptoms are not reliable indicators of depression in this population.)

Recommend supportive psychotherapy, cognitive approaches, behavioral techniques, pharmacologic therapies (see Table 397-5), or a combination of these interventions; prescribe psychostimulants for rapid treatment of symptoms (within days) or selective serotonin reuptake inhibitors, which may require 3 to 4 weeks to take effect; tricyclic antidepressants are relatively contraindicated because of their side effects.

Anxiety (applicable also Does the patient exhibit restlessness, agitation, insomnia, for family members) hyperventilation, tachycardia, or excessive worry? Is the patient depressed? Is there a spiritual or existential concern underlying the anxiety?

Recommend supportive counseling and consider prescribing benzodiazepines.

Spiritual distress

Inquire about spiritual support.

Are you at peace?

Modified from Morrison RS, Meier DE. Palliative care. N Engl J Med. 2004;350:2582-2590.


CHAPTER 3  Care of Dying Patients and Their Families  

information helps them plan their lives. Patients who are told that their disease is generally terminal are more likely to spend a longer period of time in hospice and to avoid aggressive technology at the end of life, without adverse psychological consequences. Furthermore, their families usually have fewer postdeath adverse psychological outcomes. Given that one cannot guess how much information to provide, a physician can start these conversations by asking, “Are you the kind of person who wants to hear about what might happen in the future with your illness or

would you rather take it day by day?” If the patient requests the latter, the physician can follow up by asking if there is someone else with whom he or she can talk about the prognosis. Second, before giving prognostic information, it is useful to inquire about the patient’s concerns in order to provide information in the most useful manner. Finally, it is appropriate when discussing prognostic information to acknowledge uncertainty: “The course of this cancer can be quite unpredictable, and physicians don’t have a crystal ball. I think you should be aware of the possibility that your health may



Brief Pain Inventory (Short Form) Time:

Date: Name: Last


Middle Initial

1. Throughout our lives, most of us have had pain from time to time (such as minor headaches, sprains, and toothaches). Have you had pain other than these everyday kinds of pain today? 1. Yes

2. No

2. On the diagram, shade in the areas where you feel pain. Put an X on the area that hurts the most.





3. Please rate your pain by circling the one number that best describes your pain at its worst in the last 24 hours.

0 No pain










10 Pain as bad as you can imagine

4. Please rate your pain by circling the one number that best describes your pain at its least in the last 24 hours.

0 No pain










10 Pain as bad as you can imagine


10 Pain as bad as you can imagine


10 Pain as bad as you can imagine

5. Please rate your pain by circling the one number that best describes your pain on the average.

0 No pain









6. Please rate your pain by circling the one number that tells how much pain you have right now.

0 No pain









FIGURE 3-1.  Brief Pain Inventory (short form). (Copyright 1991. Charles S. Cleeland, PhD, Pain Research Group. All rights reserved.)

CHAPTER 3  Care of Dying Patients and Their Families  


7. What treatments or medications are you receiving for your pain?

8. In the last 24 hours, how much relief have pain treatments or medications provided? Please circle the one percentage that most shows how much relief you have received. 0% No pain










100% Complete relief

9. Circle one number that describes how, during the past 24 hours, pain has interfered with your: A. General Activity 0 1 Does not interfere









10 Completely interferes









10 Completely interferes









10 Completely interferes

B. Mood 0 1 Does not interfere C. Walking Ability 0 1 Does not interfere

D. Normal Work (includes both work outside the home and housework) 0 1 Does not interfere









10 Completely interferes









10 Completely interferes









10 Completely interferes









10 Completely interferes

E. Relations with Other People 0 1 Does not interfere F. Sleep 0 1 Does not interfere G. Enjoyment of Life 0 1 Does not interfere FIGURE 3-1, cont’d.

deteriorate quickly, and you should plan accordingly. We probably are dealing with weeks to months, although some patients do better, and some do worse. Over time, the course may become clearer, and if you wish, I may be able to be a little more precise about what we are facing.” The physician must discuss these topics in an empathic way. Palliative care conversations are as much about emotions as facts.5 Talking about disease progression or death may elicit negative emotions such as anxiety, sadness, or frustration. These emotions decrease a patient’s quality of life and interfere with the ability to hear factual information. Empathic responses strengthen the patient-physician relationship, increase the patient’s satisfaction, and make the patient more likely to disclose other concerns. The first step is recognizing when the patient is expressing emotions. Once the physician recognizes the emotion being expressed, he or she can respond empathically.

It is also important for physicians to recognize their own emotional reactions to these conversations. The physician’s emotional reactions color impressions of the patient’s prognosis, thereby making it hard to listen to the patient, and may influence the physician to hedge bad news. The physician should become aware of her or his own emotional reactions to ensure that the conversation focuses on the patient rather than the health care provider’s needs. In addition to good communication skills, palliative care requires a basic knowledge of medical ethics and the law. For example, patients have the moral and legal right to refuse any treatment, even if refusal results in their death. There is no legal difference between withholding and withdrawing life-sustaining treatment. When confronted with areas of ambiguity, the physician should know how to obtain either a palliative care or ethics consultation.


CHAPTER 3  Care of Dying Patients and Their Families  



F—What is your faith/religion? Do you consider yourself a religious or spiritual person? What do you believe in that gives meaning/importance to life? I—Importance and influence of faith. Is your faith/religion important to you? How do your beliefs influence how you take care of yourself? What are your most important hopes? What role do your beliefs play in regaining your health? What makes life most worth living for you? How might your disease affect this? C—Are you part of a religious or spiritual community? Is this of support to you, and how? Is there a person you really love or is very important to you? How is your family handling your illness? What are their reactions/expectations? A—How would you like me to address these issues in your health care? What might be left undone if you were to die today? Given the severity or chronicity of your illness, what is most important for you to achieve? Would you like me to talk to someone about religious/spiritual matters?


From Puchalski C, Romer A. Taking a spiritual history. J Palliat Med. 2000;3:129-137.

• Plan what to say. Create the right setting, allow adequate time, and determine who else should be present at the meeting. • Listen carefully. Be prepared for strong emotions, respond empathetically, encourage description of feelings, and allow time for silence and response. ESTABLISHING GOALS OF MEDICAL CARE • Determine what the patient knows. Clarify any uncertainties or misconceptions. • Understand what the patient is hoping to accomplish as well as any fears and worries. • Repeat the goals back to the patient to make sure they are heard. • Suggest treatments to meet these goals and clarify what will not be done because it will not help achieve the goals. Focus on the goals that you think you can achieve. Plan follow-up, review and revise plan as needed. COMMUNICATING BAD NEWS



A.  IDENTIFYING CONCERNS AND RECOGNIZING CUES Elicit Concerns Open-ended questions “Is there anything you wanted to talk to me about today?” Active listening

Allowing patient to speak without interruption; allowing pauses to encourage patient to speak

Recognize Cues Informational concerns Patient: “I’m not sure about the treatment options” Emotional concerns

Patient: “I’m worried about that”

• • • • • •

Determine what the patient knows, wants to know, and can comprehend. Share information, recognizing that people handle information in different ways. Avoid jargon, pause frequently, check for understanding, and use silence. Recognize and support the patient’s emotional reaction. Assess the patient’s safety. Agree to a plan that enlists potential sources of support.

WITHDRAWING TREATMENT • Discuss the context of the current discussion and what has changed to precipitate it. • Review prior treatment goals and reassess their virtues. • Discuss alternative treatments based on the new goals. • Document a plan for forgoing treatment and share with the patient, the patient’s family, and the health care team. Adapted from Morrison RS, Meier DE. Clinical practice. Palliative care. N Engl J Med. 2004;350:2582-2590.


Topic: communicating information about cancer stage


“Have any of the other doctors talked about what stage this cancer is?”


“That’s right, this is a stage IV cancer, which is also called metastatic cancer…”


“Do you have questions about the staging?”


Face the patient Squarely


Adopt an Open body posture


Lean toward the patient


Use Eye contact


Maintain a Relaxed body posture

Verbal Empathy: N-U-R-S-E N

Name the emotion: “You seem worried”


Understand the emotion: “I see why you are concerned about this”


Respect the emotion: “You have shown a lot of strength”


Support the patient: “I want you to know that I will still be your doctor whether you have chemotherapy or not”


Explore the emotion: “Tell me more about what is worrying you”

From Back AL, Arnold RM, Tulsky JA. Discussing Prognosis. Alexandria, VA: American Society of Clinical Oncology; 2008.

During the past 10 years, there has been a societal push to encourage patients to designate health care proxies and to create advance care planning documents, typified by the use of living wills. These documents are meant to protect patients against unwanted treatments and to ensure that as they are dying, their wishes are followed.6 Unfortunately, there are few empirical data showing that these documents actually change practice. Still, discussions of the documents with health professionals and family members generally provoke important conversations about end-of-life care decisions and may help families confronted with difficult situations know they are respecting their loved one’s wishes.

Is the Patient Going to Die in the Location of Choice?

Most patients say that they want to die at home. Unfortunately, most patients die in institutions—either hospitals or nursing homes. Burdensome transitions decrease quality in end-of-life care. Good palliative care requires establishing a regular system of communication to minimize transitional errors. A social worker who knows about community resources is important in the development of a dispositional plan that respects the patient’s goals. Hospice programs are an important way to allow patients to die at home. In the United States, hospice refers to a specific, government-regulated form of end-of-life care, available under Medicare since 1982 but subsequently adopted by Medicaid and many other third-party insurers. Hospice care typically is given at home, a nursing home, or specialized acute care unit. Care is provided by an interdisciplinary team, which usually includes a physician, nurse, social worker, chaplain, volunteers, bereavement coordinator, and home health aides, all of whom collaborate with the primary care physician, patient, and family. Bereavement services are offered to the family for a year after the death. Hospices are paid on a per diem rate and are required to cover all the costs related to the patient’s life-limiting illness. Because of this and the fact that their focus is on comfort rather than life prolongation, many hospices will not cover expensive treatments such as inotropic agents in heart failure or chemotherapy in cancer, even if they have a palliative effect. Many hospices are experimenting with different service models in an attempt to enroll patients earlier in the course of their illness and increase access to their services.

Grade A References A1. Temel JS, Greer JA, Muzikansky A, et al. Early palliative care for patients with metastatic non-smallcell lung cancer. N Engl J Med. 2010;363:733-742. A2. Michna E, Cheng WY, Korves C, et al. Systematic literature review and meta-analysis of the efficacy and safety of prescription opioids, including abuse-deterrent formulations, in non-cancer pain management. Pain Med. 2014;15:79-92. A3. Abernethy AP, McDonald CF, Frith PA, et al. Effect of palliative oxygen versus room air in relief of breathlessness in patients with refractory dyspnoea: a double-blind, randomised controlled trial. Lancet. 2010;376:784-793. A4. Laoutidis ZG, Mathiak K. Antidepressants in the treatment of depression/depressive symptoms in cancer patients: a systematic review and meta-analysis. BMC Psychiatry. 2013;13:140. A5. Gallo JJ, Morales KH, Bogner HR, et al. Long term effect of depression care management on mortality in older adults: follow-up of cluster randomized clinical trial in primary care. BMJ. 2013;346:f2570.

A6. Jiang W, Krishnan R, Kuchibhatla M, et al. Characteristics of depression remission and its relation with cardiovascular outcome among patients with chronic heart failure (from the SADHART-CHF Study). Am J Cardiol. 2011;107:545-551. A7. Lautrette A, Darmon M, Megarbane B, et al. A communication strategy and brochure for relatives of patients dying in the ICU. N Engl J Med. 2007;356:469-478.

GENERAL REFERENCES For the General References and other additional features, please visit Expert Consult at

CHAPTER 3  Care of Dying Patients and Their Families  

GENERAL REFERENCES 1. Gill TM, Gahbauer EA, Han L, et al. Trajectories of disability in the last year of life. N Engl J Med. 2010;362:1173-1180. 2. Quill TE, Abernethy AP. Generalist plus specialist palliative care: creating a more sustainable model. N Engl J Med. 2013;368:1173-1175. 3. Brunner-La Rocca HP, Rickenbacher P, Muzzarelli S, et al. End-of-life preferences of elderly patients with chronic heart failure. Eur Heart J. 2012;33:752-759.


4. Center to Advance Palliatrive Care. Accessed February 9, 2015. 5. National Consensus Project for Quality Palliative Care. Clinical Practice Guidelines for Quality Palliative Care, 2013. Accessed February 9, 2015. 6. Silveira MJ, Kim SY, Langa KM. Advance directives and outcomes of surrogate decision making before death. N Engl J Med. 2010;362:1211-1218.


CHAPTER 3  Care of Dying Patients and Their Families  

REVIEW QUESTIONS 1. A 75-year-old man with lung cancer is admitted to the hospital with severe shortness of breath. Work-up reveals no other cause of his shortness of breath other than lymphogenic spread of his cancer. His oxygen saturation is 94%. Which of the following treatments should be instituted for his dyspnea? A. Morphine B. Benzodiazepines C. Oxygen D. A and C E. All the above Answer: A  In randomized controlled data, opioids have been shown to decrease dyspnea both in lung cancer patients and in patients with COPD. Oxygen is helpful only if the patient has hypoxia. Benzodiazepines have not been shown to decrease breathlessness. 2. Which of the following is NOT required for a patient to be in hospice? A. The patient must be DNR. B. The patient must have a life-limiting illness, which is likely to cause her death in 6 months. C. The patient wishes to focus on quality of life rather than longevity of life. D. If the patient lives at home, she must have a primary caregiver. Answer: A  The patient does not have to be DNR to be in hospice. The others are requirements of hospice.

3. Which of the following is true of depression in life-limiting illnesses? A. It is a normal reaction when people have a life-limiting illness, and it should not be treated. B. It cannot be improved because the treatments take too long to work in patients with serious illness. C. Treatment of depression decreases both morbidity and mortality. D. It requires a psychiatric consult because treatment is very complicated. Answer: C  Data show that the treatment of depression improves both quality of life and mortality. 4. Which of the following is true? A. Telling patients that they have a terminal illness will result in their losing hope. B. Telling patients they have a terminal illness has no impact on their desire for future treatment. C. Telling patients that they have terminal illnesses is associated with their choosing hospice more frequently. D. Patients have clearly stated that they do not want to be told that they have a terminal illness. Answer: C  Data suggest that telling patients that they have a life-limiting illness is associated with a lower likelihood of choosing aggressive care at the end of life and is not associated with poorer psychiatric outcomes.


CHAPTER 4  Cultural Context of Medicine  


Components of health care access include the ability to get into the health care system as well as to obtain appropriate care once in the system. The availability of health care providers who meet an individual patient’s needs is another key component of access to care. Quality care is based on scientific evidence (i.e., is effective), avoids injury to the patient (i.e., is safe), minimizes harmful delays (i.e., is timely), is responsive to the individual patient’s needs (i.e., is patient centered), promotes communication among providers (i.e., is coordinated), does not vary because of personal characteristics (i.e., is equitable), and avoids waste (i.e., is efficient).

4  CULTURAL CONTEXT OF MEDICINE VICTORIA M. TAYLOR The 2010 U.S. Census counted about 39 million blacks or African Americans (13% of the population), nearly 15 million Asian Americans (5% of the population), about 3 million American Indians and Alaska Natives, and more than 500,000 Native Hawaiians and other Pacific Islanders. It also counted more than 50 million individuals of Hispanic or Latino origin (16% of the population). Approximately 40 million Americans (13% of the population) were foreign born. One in 2 immigrants to the United States have limited English proficiency (i.e., they do not speak English very well or fluently), and 1 in 10 immigrants do not speak English at all (Fig. 4-1). During the past two decades, a large body of literature has documented substantial disparities in health status. Although some of these disparities are based on socioeconomic status, many are based on race, ethnicity, or other characteristics. Black men have a substantially higher age-adjusted incidence of prostate cancer than do white men (236 per 100,000 versus 147 per 100,000). American Indians/Alaska Natives are more than twice as likely as non-Latino whites of a similar age to have diabetes. More than half of the Americans who are living with chronic hepatitis B infection are Asians or Pacific Islanders. Lesbian, gay, bisexual, and transgender individuals have higher rates of suicidal behavior compared with heterosexual individuals. A major goal of Healthy People 2020 is to eliminate health disparities for preventable and treatable conditions such as cancer, diabetes, and human immunodeficiency virus infection. Culture can be defined as a shared system of values, beliefs, and patterns of behavior, and it is not simply defined by race and ethnicity. Culture can also be shaped by factors such as country and region of origin, acculturation, language, religion, and sexual orientation. For instance, the black population of the northeastern United States includes individuals who moved from southern states decades ago as well as recent immigrants from Ethiopia. As the United States population becomes increasingly diverse and as pronounced differences in health status continue to be documented, consideration of the cultural context of medicine is becoming a national priority.

Access to Health Care

Racial and ethnic minority groups, particularly immigrants, disproportionately have problems accessing health care. Before the implementation of the Affordable Care Act, the proportions of Latinos and Native Americans/ Alaska Natives who lacked health insurance was more than twice the proportion among non-Latino whites, and less than two thirds of Americans with limited English proficiency were insured. About 1 in 3 Korean American and Vietnamese American adults had no regular source of medical care compared with about 1 in 10 non-Latino white adults. Blacks and Latinos are far less likely than are whites and Asians to have access to physicians of their own race and ethnicity. This imbalance is important because racial concordance between physicians and patients can improve the processes of care. For example, patients with race-concordant physicians are more likely to use needed health services, are less likely to postpone or delay seeking care, and are more satisfied with their care than are patients in race-discordant relationships. Whether these differences translate into different health outcomes, however, is less clear.1

Quality of Health Care

National surveys confirm population-level disparities in the quality of preventive care. Recent immigrants have far lower levels of interval screening for breast, cervical, and colorectal cancer than do individuals who were born in the United States (Fig. 4-2). The proportion of Native Hawaiians and other Pacific Islanders whose serum cholesterol levels are measured at least once every 5 years is significantly lower than among whites. In 2011, only 40% of Asians aged 65 years and older had ever received the pneumococcal vaccine compared with 67% of non-Latino whites. Racial and ethnic disparities have been documented for a number of specific clinical situations. For example, Latino women with breast cancer are less likely to receive radiation therapy within a year of breast-conserving surgery than are white women, Native Americans and Alaska Natives are less likely than whites to receive recommended care such as initial antibiotics within 6 hours of hospital arrival, and blacks with end-stage renal disease


US-born Foreign-born, in US ≥ 10 years

Cervical Latin America

Foreign-born, in US < 10 years

Asia Africa Colorectal Europe Oceania 0%

Total 0%








FIGURE 4-1.  Proportion of immigrants aged 5 years and older with limited English proficiency by region of origin. (From Grieco EM, Acosta YD, de la Cruz P, et al. The foreign-born population in the United States: 2010. Washington DC: U.S. Department of Commerce; 2012.)






FIGURE 4-2.  Adherence to cancer screening guidelines by immigration status. Breast = mammography during last 2 years among women aged 50 to 74 years. Cervical = Papanicolaou test during last 3 years among women aged 21 to 65 years. Colorectal = among individuals aged 50 to 75 years, fecal occult blood test last year; sigmoidoscopy last 5 years and fecal occult blood test last 3 years; or colonoscopy last 10 years. (From Centers for Disease Control and Prevention. Cancer screening—United States, 2010. MMWR Morb Mortal Wkly Rep. 2012; 61:41-45.)


CHAPTER 4  Cultural Context of Medicine  

are less likely to be entered to a transplant list than are whites. Moreover, disparities in the quality of care are found even when variations in insurance status, income, and comorbid conditions are taken into account. Disparities in health care quality exist even in systems that are generally believed to provide equal access.2 For example, in the Veterans Affairs Health System, disparities between blacks and whites have been documented for blood pressure control among patients with hypertension, cholesterol control among patients with coronary heart disease, and glucose control among patients with diabetes. Moreover, these disparities persist even after adjusting for location and socioeconomic status. Similar disparities have been documented in Medicare managed care programs between elderly blacks and whites with diabetes and cardiovascular disorders.


Health disparities can be reduced or perhaps even eliminated by maintaining culturally competent health care systems. Cultural competence may be defined as a set of congruent attitudes, behaviors, and policies that come together both among professionals and within systems to enable effective work in cross-cultural situations (Fig. 4-3). Ongoing efforts to improve cultural competence in the health care system target organizational, structural, and clinical barriers. These initiatives aim to close gaps in health status, to decrease differences in the quality of care, to enhance patients’ satisfaction, and to increase patients’ trust.

Organizational Barriers and Interventions

Diversity among health care professionals is associated with better access to care for disadvantaged populations. Black and Latino physicians are more likely than their white colleagues to work in medically underserved communities and to have a better understanding of barriers to health care. Because less than 10% of practicing physicians are black or Latino, and only about 15% of medical school students are from one of these groups, many U.S. medical schools have implemented comprehensive programs to infuse diversity among their students, resident physicians, and faculty. About two thirds of the patients who receive care at federally funded community health centers in medically underserved areas are members of racial and ethnic minority groups. In these health centers, patients are three times more likely to have limited proficiency in English compared with the general population. The community health center model has proved effective not

only in increasing access to care but also in improving continuity of care and health outcomes. For example, medically underserved communities with community health centers have fewer preventable hospitalizations and uninsured emergency department visits than do similar communities without health centers. Compared with national rates, community health centers report minimal racial and ethnic disparities in clinical outcomes such as the control of diabetes and hypertension.3

Structural Barriers and Interventions

Accumulating evidence suggests that trained professional interpreters can improve the clinical care received by individuals with limited English proficiency. A1  However, interpreter services often remain ad hoc, with family members and untrained nonclinical employees acting as interpreters.4 Use of ad hoc services has potentially negative clinical consequences, including breach of the patient’s confidentiality and inaccurate communication. One major obstacle to the implementation of professional interpreter programs is a lack of reimbursem*nt; Medicare and most private insurers do not pay for interpretation and related services, and most states do not pay for interpretation under Medicaid. Assistance with navigation represents a promising model to enable racial and ethnic minority patients to move through the health system effectively and to be actively involved in decision making about their medical care.5 Guides may be nurses, social workers, or volunteers who are familiar with the health care system. They help patients and their families navigate the treatment process, steering them around obstacles that may limit their access to quality care, choice of doctors, and access to treatment options. For example, an American Cancer Society navigation program is effective in reducing the time to diagnostic resolution after abnormal cancer screening tests in medically underserved patients. A2  Another option for closing the gap in health care among various minority populations is community health workers.6 In general, community health workers live locally and share the language and culture of the patients being served. Lay community health workers provide cultural mediation between communities and the health care system; culturally appropriate and accessible health education and information; help in obtaining needed medical services, informal counseling, and social support; and advocacy within the health care system. The effectiveness of community health workers is documented by a study in which Mexican American women randomized to

Health Care Interventions



Programs to increase the diversity of health care providers Culturally specific health care settings

Programs to recruit and retain staff who reflect cultural diversity of the community Use of interpreter services or bilingual providers Patient navigator and community health worker programs Use of linguistically and culturally appropriate health education materials


Cultural competency training for health care providers

Intermediate Outcomes Health care providers reflect diversity of communities served

Increase cultural relevance and acceptability of health information

Less miscommunication due to language differences or cultural understanding of health events

Increase accuracy of diagnosis and use of appropriate interventions

Greater provider knowledge of variation in health beliefs, practices, and conditions

Increase patient understanding of and adherence to treatment recommendations

More provider sensitivity to their own beliefs and behaviors that marginalize ethnic groups

Improve access to quality health care services by diverse populations

Patient Outcomes Increase patient satisfaction with health care system Increase patient confidence in health care system

Health Outcomes Decrease inappropriate differences in the characteristics and quality of care provided Close gaps in health status across diverse populations

FIGURE 4-3.  Analytic framework for evaluating the effectiveness of health care interventions to increase cultural competence. (From Anderson LM, Scrimshaw SC, Fullilove MT, et al., for the Task Force on Community Preventive Services. Culturally competent healthcare systems: a systemic review. Am J Prev Med. 2003; 24[suppl]:68-79.)

receive health worker education were significantly more likely to obtain Papanicolaou tests than were women randomized to receive usual care. A3  The largest formal system of community health workers is the Indian Health Service, which currently has about 1400 community health representatives.

Clinical Barriers and Interventions

Patients who are members of racial and ethnic minority groups often understand health and disease (i.e., explanatory model) differently than the general population. For example, many Vietnamese people believe that disease is caused by an imbalance of the humoral forces of yin and yang. When ill, they commonly use Chinese herbal medicine as well as indigenous folk practices known as Southern medicine in an effort to restore the balance of humoral forces. In addition, Vietnamese patients may think that Western medicine is too strong and will upset the internal balance. Consequently, a hypertensive Vietnamese patient may, for example, use Chinese herbal medicines instead of prescribed antihypertensive medication. Alternatively, the patient may take a lower dose of medication than prescribed by his or her physician. Cultural competency training for health care providers generally includes teaching cross-cultural knowledge and communication skills, while avoiding stereotypes.7 Examples include the effect of prejudice on gays and lesbians and how this prejudice shapes their interactions with the health care system, and common spiritual practices that might interfere with prescribed therapies (such as Ramadan fasting practices, when observed by diabetic Muslim patients). Communication skills that can be addressed in cultural competence training include approaches to eliciting patients’ explanatory models and use of traditional treatments, as well as methods for negotiating different styles of communication and levels of family participation in decisionmaking. Cultural competency training improves the attitudes and skills of health professionals as well as patient satisfaction, but there is less evidence that it improves clinical outcomes. A4 


Individual clinical practices should regularly assess their current organizational climate, policies, and training related to diversity. Practices can address health disparities by hiring clinical and office staff who are representative of the communities they serve, by routinely using professional interpreters during clinical encounters with patients who have limited proficiency with English, by offering cultural competency education and training to physicians and staff, and by providing educational and informational materials that are culturally and linguistically appropriate for their patient populations.8 National and state efforts to improve cultural competence in health care, whether used alone or in conjunction with socioeconomic initiatives, are likely to play a significant role in reducing health disparities across population subgroups. An important goal of the Affordable Care Act is to reduce health disparities by expanding health insurance coverage, addressing diversity in the health care workforce, increasing the capacity of community health centers, and promoting the use of patient navigators and community health workers.

Grade A References A1. Bagchi AD, Dale S, Verbitsky-Savitz N, et al. Examining effectiveness of medical interpreters in emergency departments for Spanish-speaking patients with limited English proficiency: results of a randomized controlled trial. Ann Emerg Med. 2011;57:248-256. A2. Paskett ED, Katz ML, Post DM, et al. The Ohio Patient Navigation Research Program: does the American Cancer Society patient navigation model improve time to resolution in patients with abnormal screening tests? Cancer Epidemiol Biomarkers Prev. 2012;21:1620-1628. A3. Byrd TL, Wilson KM, Smith JL, et al. AMIGAS: a multicity, multicomponent cervical cancer prevention trial among Mexican American women. Cancer. 2013;119:1365-1372. A4. Sequist TD, Fitzmaurice GM, Marshall R, et al. Cultural competency training and performance reports to improve diabetes care for black patients: a cluster randomized, controlled trial. Ann Intern Med. 2010;152:40-46.

GENERAL REFERENCES For the General References and other additional features, please visit Expert Consult at

CHAPTER 4  Cultural Context of Medicine  

GENERAL REFERENCES 1. Meghani SH, Brooks JM, Gipson-Jones T, et al. Patient-provider race-concordance: does it matter in improving minority patients’ health outcomes. Eth Health. 2009;14:107-130. 2. Trivedi AN, Grebla RC, Wright SM, et al. Despite improved quality of care in the Veterans Affairs Health Care System, racial disparity persists for some clinical outcomes. Health Aff. 2011;4: 707-715. 3. Lebrun LA, Shi L, Zhu J, et al. Racial/ethnic differences in clinical quality performance among health centers. J Ambul Care Manage. 2013;36:24-34. 4. VanderWielen LM, Enurah AS, Rho HY, et al. Medical interpreters: improvements to address access, equity, and quality of care for limited-English-proficient patients. Acad Med. 2014;89:1324-1327.


5. Natale-Pereira A, Enard KR, Nevarez L, et al. The role of patient navigators in eliminating health disparities. Cancer. 2011;117(suppl):3543-3552. 6. Brownstein JN, Hirsch GR, Rosenthal EL, et al. Community health workers 101 for primary care providers and other stakeholders in health care systems. J Ambul Care Manage. 2011;34:210-220. 7. Betancourt JR, Cervantes MC. Cross-cultural medical education in the United States: key principles and experiences. Kaohsiung J Med Sci. 2009;25:472-478. 8. Chin MH, Clarke AR, Nocon RS, et al. A roadmap and best practices for organizations to reduce racial and ethnic disparities in health care. J Gen Intern Med. 2012;27:992-1000.


CHAPTER 4  Cultural Context of Medicine  

REVIEW QUESTIONS 1. What percentage of immigrants to the United States have limited English proficiency? A. 20% B. 30% C. 40% D. 50% E. 60% Answer: D  The 2010 Census found that one in two immigrants have limited English proficiency. (Grieco EM, Acosta YD, de la Cruz GP, et al. The foreign born population in the United States: 2010. Washington, DC: U.S. Department of Commerce, 2012.) 2. A 19-year-old Vietnamese man presents for a routine physical examination. He is a recent immigrant, has no symptoms or significant medical history, and has not previously had a physical examination or any blood testing in the United States. Which of the following blood tests should you perform? A. HIV B. Hepatitis B C. Glucose D. Cholesterol E. All of the above Answer: B  More than half of the Americans who have chronic hepatitis B infection are Asians or Pacific Islanders. Therefore, all immigrants from Asia should be tested for hepatitis B. (Pollack H, Wang S, Wyatt Le, et al. A comprehensive screening and treatment model for reducing disparities in hepatitis B. Health Aff. 2011;30:1974-1983.) 3. Which of the following statements about health insurance is incorrect? A. Whites are more likely to have insurance than American Indians/ Alaska Natives. B. Latinos are less likely to have insurance than non-Latino whites. C. Gay/lesbian/hom*osexual individuals are less likely to have insurance than heterosexual individuals. D. Citizens are more likely to have insurance than noncitizens. E. Individuals who are proficient in English are more likely to have insurance than individuals who are not. Answer: C  The California Health Interview Survey provides information about health insurance coverage among population subgroups and shows differences by race/ethnicity, citizenship status, and level of English proficiency but not by sexual orientation. (University of California Los Angeles. Ask CHIS 2009, Accessed March 22, 2014.)

4. Which of the following are core community health worker functions? A. Cultural mediation B. Health education C. Informal counseling D. Social support E. All of the above Answer: E  Cultural mediation, health education, informal counseling, and social support are all core community health worker functions. (Brownstein JN, Hirsch GR, Rosenthal EL, Rush CH. Community health workers 101 for primary care providers and other stakeholders in health care systems. J Ambul Care Manage. 2011;34:210-220.) 5. Your practice has recently started seeing a large number of Hispanic immigrant patients with limited English proficiency. You have decided to make some changes to your practice to accommodate the specific needs of these patients. Which of the following approaches are appropriate? A. Hire Hispanic staff B. Ask limited English proficiency patients to bring a family member with them to provide C. Medical interpretation D. Provide cultural competency training to providers E. A and C F. All of the above Answer: D  Clinical practices can address health disparities in multiple ways, including by hiring staff who are representative of the practice population, providing professional interpreter services, offering cultural competency training to providers, and providing linguistically appropriate patient education materials. Family members should not be asked to provide medical interpretation. (Washington DL, Bowles J, Saha S. Transforming clinical practice to eliminate racial-ethnic disparities in healthcare. J Gen Intern Med. 2008;23:685-691.)


CHAPTER 5  Socioeconomic Issues in Medicine  


All nations—rich and poor—struggle with how to improve the health of the public, obtain the most value from medical services, and restrain rising health care expenditures. Many developed countries also wrestle with the paradox that their citizens have never been so healthy or so unhappy with their medical care. Despite the reality that only about 10% of premature deaths result from inadequate medical care, the bulk of professional and political attention focuses on how to obtain and pay for state-of-the-art medical care. By comparison, 40% of premature deaths stem from unhealthy behaviors— including smoking (about 44%; Chapter 32), excessive or unwise drinking (about 11%; Chapter 33), obesity and insufficient physical activity (about 15% but estimated to rise substantially in the years to come; Chapters 16 and 220), illicit drug use (about 2%; Chapter 34), and imprudent sexual behavior (about 3%; Chapter 285). Genetics (Chapter 40) account for an additional 30%; social factors—discussed next—account for 15%, and environmental factors (Chapter 19) account for 5%. Of the major behavioral causes of premature deaths, tobacco use (Chapter 32) is by far the most important, although recent increases in obesity (Chapter 220) and physical inactivity (Chapter 16) are also alarming. Health is influenced by genetic predisposition, behavioral patterns, environmental exposures, social circ*mstances, and health care.


Socioeconomic status, or class, is a composite of many different factors, including income, net wealth, education, occupation, and neighborhood. In general, people in lower classes are less healthy and die earlier than people at higher socioeconomic levels, a pattern that holds true in a stepwise fashion from the poorest to the richest. In the United States, the association between health and class is usually discussed in terms of racial and ethnic disparities; but in fact, race and class are independently associated with health status, and it can be argued that class is the more important factor. For example, U.S. racial disparities in the prevalence of adult smoking are relatively small among whites, blacks, and Hispanic Americans, whereas there are huge differences among smoking rates by educational level (Fig. 5-1).1 U.S. physicians have reduced their smoking prevalence to a record low of only 1%. Although both smoking rates and the numbers of cigarettes smoked by those who continue to smoke are gradually declining (Fig. 5-2), more than 43 million Americans and millions more elsewhere continue to smoke.2 Because people of higher socioeconomic status adopt health-promoting behaviors at a faster rate than people of lower socioeconomic status, overall population health can increase while health disparities also widen (Fig. 5-3).

No high school diploma


GED diploma


High school graduate


Some college


Undergraduate degree


Graduate degree

5.0% 0%






FIGURE 5-1.  Prevalence of adult smoking, by education, United States, 2011. GED = General Education Development. (From Centers for Disease Control and Prevention. Current cigarette smoking among adults: United States, 2011. MMWR Morb Mortal Wkly Rep. 2012;61:889-894.)


CHAPTER 5  Socioeconomic Issues in Medicine  

45 40

Smoking prevalence (%)


Average number cigarettes smoked/day per smoker

30 25 20 15 10 5 0

1965 1970 1976 1978 1980 1985 1988 1991 1993 1995 1998 2000 2002 2004 2006 2008 2010 2012

Percent/number of cigarettes smoked daily

In part, the relationship between class and health is mediated by higher rates of unhealthy behaviors among the poor, such as the inverse relationship between educational attainment and cigarette smoking, but unhealthy behaviors do not fully explain the poor health of those in the lower socioeconomic classes. Even when such behaviors are held constant, people in lower socioeconomic classes are much more likely to die prematurely than are people of higher classes. Of interest is that first-generation immigrants to the United States appear to be more protected from the adverse health consequences of low socioeconomic status than are subsequent generations. It is unclear which of the components of class—education, wealth (either absolute wealth or the extent of the gap between rich and poor), occupation, or neighborhood—makes the greatest impact on a person’s health. Most likely, it is a combination of all of them. For example, the constant stress of a lower class existence—lack of control over one’s life circ*mstances, social isolation, and the anxiety derived from the feeling of having low status—is linked to poor health. This stress may trigger a variety of neuroendocrinologic responses that are useful for short-term adaptation but bring long-term adverse health consequences. What can clinicians do with this knowledge? Clearly, it is difficult to write prescriptions for more income, a better education, good neighborhoods, or high-paying jobs. Physicians can, however, encourage healthy behavior. At key times of transition, such as during discharge planning for hospitalized patients, clinicians should be attentive to social circ*mstances. For patients who are likely to be socially isolated, clinicians should encourage or arrange interactions with family, neighbors, religious organizations, or community agencies to improve the likelihood of optimal outcomes. Access points to vital social services, such as child care, disability insurance, and food supplementation, can be provided in clinical settings.3 In addition, physicians should seek to identify and eliminate any aspects of racism in health care institutions (Chapter 4). Finally, in their role as social advocates, physicians

FIGURE 5-2.  Smoking prevalence and average number of cigarettes smoked per day per current smoker. (Data based on Centers for Disease Control and Prevention (CDC). Smoking prevalence, 1965-2010. MMWR Morb Mortal Wkly. 2011;60:109-113; Current cigarette smoking in the United States: current estimate. CDC; tobacco/data_statistics/fact_sheets/adult_data/cig_smoking. Accessed February 10, 2015; National Health Interview Survery. CDC; data_related_1997_forward.htm. Accessed February 10, 2015; Jamal A, Agaku IT, O’Connor E, et al. Current cigarette smoking among adults—United States, 2005-2013. MMWR Morb Mortal Wkly. 2014;63:1108-1112.)



Upper SES Lower SES Time FIGURE 5-3.  Health improves while disparities widen. SES = socioeconomic status.

can promote such goals as safe neighborhoods, improved schools, and access to quality health care.


Medical care today is on a collision course. On the one hand, an everexpanding science base continuously generates new technologies and drugs that promise a longer and healthier life. Add a public eager to obtain the latest breakthroughs touted in the media and over the Internet, plus a well-stocked medical industry eager to meet that demand, and it is easy to understand why expenditures continue to soar. On the other hand, payers for medical care— health insurance companies, government (federal, state, and local), and employers—increasingly bridle at medical care costs. The United States continues to lead the world in health care expenditures.4 In 2011, it spent more than $2.7 trillion, amounting to 17.9% of the gross domestic product. Most policy analysts contend that this rate of increase in medical care expenditures is unsustainable, but this claim has been made for many years. A potent combination of supply and demand factors explains why the United States spends so much.5 On the supply side, the United States far exceeds other countries in the availability and use of expensive diagnostic technologies, such as magnetic resonance imaging and computed tomography. For example, the United States has four times as many magnetic resonance imaging machines per capita as does Canada. Similar patterns exist for therapeutic technologies, whether coronary angioplasty, cancer chemotherapy, or joint prostheses. The differences are especially dramatic in older patients. Other supply factors that drive high medical expenditures in the United States include a fee-for-service payment system that compensates physicians much more when they use expensive technologies than when they do not6; a medical professional work force that earns much higher incomes relative to the population than in other nations and that emphasizes specialist rather than generalist practice; accelerated development of new and costly medications that are directly marketed to consumers; much higher administrative costs; higher rates of fraud and abuse; and a high rate of defensive medicine in response to pervasive fears about medical malpractice suits. Supply factors that do not appear to be unique to the United States are the number of physicians or hospitals. Many other developed countries have a much larger physician work force relative to their population, as well as a much higher ratio of primary care physicians to specialists. The number of hospitals and hospital beds, the frequency of hospitalizations, and the length of hospital stay are relatively low in the United States, although it does have a much greater proportion of intensive care beds. Finally, recent analyses suggest that a principal driver of high expenditures on health care in the United States is the much greater price charged per unit of service compared with other developed countries. Demand factors also drive medical expenditures. The extent to which the media and the medical profession feature medical “breakthroughs” is extensive and one-sided. New promising treatments merit front-page stories and commercial advertisem*nts, whereas subsequent disappointing results are buried or ignored. The cumulative result is to whet patients’ appetite for more and to leave the impression that good health depends only on finding the right treatment. This same quest explains the popularity of alternative medicine, for which patients are willing to spend $34 billion annually out of their own pockets (Chapter 39). The cumulative impact of these supply and demand drivers is that there are incentives to do more at every step of the American medical system.5 It could be argued that rising expenditures for medical care are not a bad thing. What could be more important than ensuring maximal health? There are several rebuttals to that argument. First, it is not clear that money spent on medical care brings appropriate value in the United States, given that its health statistics are worse than those of virtually every other developed country. Second, there are substantial regional differences in the supply and use of medical care, such as a two-fold difference in the supply of acute hospital beds and a four-fold difference in the risk of being hospitalized in an intensive care unit at the end of life. Similar regional differences exist for procedures such as transurethral prostatectomy, hysterectomy, and coronary artery bypass surgery. Yet there is no evidence that “more is better” on a regional basis. Consequently, rising health care expenditures are stressing public programs such as Medicare, Medicaid, the Veterans Administration health system, and municipal hospitals, with budget requests outstripping the tax base to pay for them. Medical debt is by far the most important cause of bankruptcy. Finally, as health care becomes less affordable for businesses and

government, the number of people without health insurance will continue to increase.

Cost-Containment Strategies

Since the mid-1970s, a variety of strategies to contain rising medical expenditures have yielded limited success.5 These attempts have tried to restrict the supply of costly medical technologies as well as the production of physicians, especially specialists; to promote health maintenance organizations that have incentives to spend less on medical care; to ration indirectly by limiting health insurance coverage; to institute prospective payment for hospital care; to use capitation payments or discounted fee schedules for physician reimbursem*nt; to introduce gatekeeper mechanisms to reduce access to costly care; to put patients at more financial risk for their own medical care; to reform malpractice procedures; to reduce administrative costs; and to encourage less aggressive care at the end of life. The most recent suggestions— comparative effectiveness research to curtail the use of unnecessary technology, electronic medical records to avoid duplication of tests, payment for performance, accountable care organizations that change payment incentives—all hold promise to improve quality, but their potential for substantial cost reduction is only theoretical at present. Recently, however, the rate of increase in health care expenditures has slowed relative to the gross domestic product.7 Two basic hypotheses have been offered: the recession that began in 2008, and heightened cost consciousness among hospitals, health insurers, and some physician groups. Payment for medical care varies by country. In the United States, health insurance coverage is an incomplete patchwork, consisting of governmentsponsored programs for elderly people (Medicare), poor people (Medicaid), and veterans, plus employer-based coverage for workers and their families. Medicare covers acute care services in the hospital and in physicians’ offices but has limited coverage for prescription drugs and long-term care. More than half of all Medicare subscribers also buy supplemental insurance. Medicaid covers more services than Medicare does, but Medicaid payments to physicians and hospitals are so low in many states that patients have restricted access to care. At any given time, more than 44 million Americans have lacked health insurance, and 70 million have been without insurance at some point during the year. In addition, millions of immigrant workers are also uninsured. The lack of health insurance contributes to poor health, such as delayed diagnosis and undertreatment of asthma, diabetes, hypertension, and cancer. The 2010 Patient Protection and Affordable Care Act (ACA) contains numerous insurance reform features that took effect in 2010 and 2011, as well as coverage expansions that began in 2014.8 The ACA was originally expected to cover 32 million previously uninsured Americans, with about half enrolling in subsidized private insurance plans and half in expanded state Medicaid programs. However, a 2012 Supreme Court decision gave states the choice of opting out of the Medicaid expansion. As a result, only 27 states plus the District of Columbia accepted that expansion. States that opted out of Medicaid expansion, such as Texas, tend to be those with the highest proportion of uninsured—mainly poor—people. In addition, various coverage components, especially regarding contraception, continue to be litigated. Revenuegenerating provisions of the ACA are split about evenly between spending reductions and cost containment. In contrast to what happened after the passage of Medicare and Medicaid, the ACA continues to be highly controversial politically, and thus subject to potential changes, depending on election results. Because medical care is both so valued and so expensive, physicians everywhere will inevitably become more involved in issues of medical economics. As cost-containment pressures force patients to assume more of their medical expenses, patients will become more aware of costs and more demanding about the price and value of care. In addition, knowledge will continue to accumulate about the real and potential harm from unnecessary or marginally useful medical services. Thus, informed clinical decision making will require that physicians have accurate information about the risks, benefits, and costs of medical care and better ways to communicate what is known and what is not. GENERAL REFERENCES For the General References and other additional features, please visit Expert Consult at

CHAPTER 5  Socioeconomic Issues in Medicine  

GENERAL REFERENCES 1. Centers for Disease Control and Prevention. Current cigarette smoking among adults: United States, 2011. MMWR Morb Mortal Wkly Rep. 2012;61:889-894. 2. Jha P, Ramasundarahettige C, Landsman V, et al. 21st Century hazards of smoking and benefits of cessation in the United States. N Engl J Med. 2013;368:341-350. 3. Gottlieb L, Sandel M, Adler NE. Collecting and applying data on social determinants of health in health care settings. JAMA Intern Med. 2013;173:1017-1020. 4. Lorenzoni L, Belloni A, Sassi F. Health-care expenditure and health policy in the USA versus other high-spending OECD countries. Lancet. 2014;384:83-92.


5. Schroeder SA. Personal reflections on the high cost of American medical care. Arch Intern Med. 2011;171:722-727. 6. Schroeder SA, Frist W. Phasing out fee-for-service payment. N Engl J Med. 2013;368:2929-2932. 7. Fuchs VR. The gross domestic product and health care spending. N Engl J Med. 2013;369: 107-109. 8. Shaw FE, Asomugha CN, Conway PH, et al. The Patient Protection and Affordable Care Act: opportunities for prevention and public health. Lancet. 2014;384:75-82.


CHAPTER 5  Socioeconomic Issues in Medicine  

REVIEW QUESTIONS 1. Preventing premature mortality is a prime goal for all clinicians. Of the following statements regarding premature mortality, which one is incorrect? A. About 10% of premature deaths could be prevented by assuring highquality medical care to all. B. Among the various general causes of premature deaths, behavioral factors such as risky sexual behavior, alcohol and drug abuse, smoking, and obesity and physical inactivity are the most important and also offer opportunities for remediation. C. Among behavioral factors, the most important cause of premature death is smoking cigarettes. D. As smoking prevalence gradually decreases, the remaining smokers are smoking more cigarettes per day. E. Based on current trends, obesity and physical inactivity will likely become more important causes of premature deaths. Answer: D  Statement A is correct. As important as good medical care is, the lack of such services only accounts for about 10% of premature deaths. Statements B, C, and E are correct. As regards statement D, at the same time as overall smoking prevalence has declined, the number of daily cigarettes consumed by those who continue to smoke has also declined. This decline is probably of function of several factors, including rising tobacco taxation that makes smoking more expensive, the spread of clean indoor air laws, and the increasing stigma attached to smoking. 2. Increasing evidence points to the important health impacts of social economic status (SES). Which one of the following statements regarding health and social class is correct? A. Among the various components of SES, racial and ethnic characteristics contribute the most to health disparities. B. Virtually all of the worst health among low SES populations can be explained by personal behaviors such as cigarette smoking and drug and alcohol abuse. C. The relationship between SES and health is concentrated among those with low SES. In other words, for people above 400% of the poverty level, SES does not contribute to health status. D. Because so many of the determinants of low SES (e.g., education, income, housing, net wealth) are outside the purview of clinicians, they should not be concerned about them. E. Co-location of access to important social services—such as food stamps or disability payments—at medical sites can improve the social status of selected patients. Answer: E  Answer A is incorrect. Class, as measured by income, net wealth, educational status, and neighborhood, is the most important factor leading to health disparities. Regarding answer B, although personal behaviors such as smoking and physical activity are important, many other factors associated with low social class contribute to poor health. Regarding answer C, there is a stepwise association between SES and health, even at the higher levels of SES. For example, those in the top decile of SES enjoy better health than those in the ninth decile, even though both deciles have high SES. For answer D, it is true that clinicians would have a hard time improving these factors. Nevertheless, they should be conscious of them because they often influence treatment strategies (e.g., the ability to obtain nutritious food or to buy medications). Finally, answer E is correct. For example, convenient provision of food stamps for those with food insecurity would help stabilize diabetic patients. 3. Regarding per capita medical expenditures, population health, and access to medical care, which statement best expresses the performance of the United States versus other developed nations? A. It leads the world in medical expenditures but trails badly in health outcomes and access to health care services. B. It trails the world in medical expenditures, health outcomes, and access to care. C. It is about in the middle for expenditures, health, and access. D. It leads the world in expenditures, health, and access. E. It leads the world in expenditures and health but lags in access.

Answer: A  Answer A is correct as written. Of the three statements in answer B, only the second is correct. Of the three answers in C, none are correct. Of those in D, only the first is correct. And of those in E, only the first is correct. 4. Regarding the causes of rising medical expenditures in the United States, which answer is correct? A. Fee-for-service payment to physicians is no longer a major driver of cost escalation because it will soon be replaced by bundled or capitated payment. B. The United States is an outlier in that it has more hospital beds per capita and a higher length of stay. C. The United States is an outlier in that it features more physicians per capita. D. There are multiple factors responsible for the patterns of expenditures in the United States, and it is unlikely that any one bears major responsibility for the high expenditure profile. E. The threat and reality of malpractice suits, in a country with the highest number of lawyers per capita, is the major reason for high U.S. medical expenditures. Answer: D  For answer A, although there is much current talk about the imminent demise of fee-for-service payment, it remains the dominant way of paying physicians. Regarding answer B, the United States actually has fewer hospital beds per capita than other developed countries as well as a shorter length of stay. The cost of a day in the hospital, however, is far higher in the United States. Similarly, for answer C, the United States has fewer physicians per capita than most developed nations but has a much higher proportion of specialists. Answer D, the correct answer, reflects the multiple reasons that medical care is more costly in the United States. Regarding answer E, malpractice suits are indeed more of a factor in the United States than other nations. Nevertheless, if all malpractice costs were to vanish, the United States would still have the most expensive health care system by far. 5. Regarding responses to high medical expenditures in the United States, which of the following statements is most correct? A. Because expenditures on medical care are so essential and because wealthy countries characteristically spend more on health care as wealth increases, there is little interest in curtailing rising medical expenditures. B. The cost-containment measures found in the Patient Protection and Affordable Care Act will be sufficient to rein in rising medical costs. C. The combination of pressures on personal and governmental health care spending and crowding out of other social expenditures makes it likely that more intense cost-containment activity will occur. D. The combination of widespread electronic medical record use and the spread of Accountable Care Organizations will be sufficient to curb rising medical expenditures. E. Political obstacles to cost containment, such as the fear of rationing, will not be a problem. Answer: C  Answer A is incorrect. There is emerging consensus that something must be done to curtail rising medical expenditures (as phrased in answer C), but not on how to accomplish that. Although there are some costcontainment features in the Patient Protection and Affordable Care Act (PPACA) (answer B), it is unproved whether they will reduce medical expenditures. Answer C is correct. Regarding answer D, there is no good evidence that either the spread of electronic medical records or Accountable Care Organizations will curb rising medical expenditures, despite the enthusiasm of advocates for those programs. Finally, political obstacles (answer E) remain potent barriers to medical cost containment, such as the assertion during the debate over the PPACA that access to palliative care services would be tantamount to creating death panels.


CHAPTER 6  Global Health  


Health is a human right, but more than 2 billion people live with a daily income of less than $2 and have no access to good health care. Health is determined by the context of people’s lives. Individuals are unable to control many of the social determinants of health (Chapter 5), such as income and social status, education, physical environment, social support network, genetics, health services, and gender.1 In the process of modernization from a less developed to a more developed nation, the epidemiologic transition of modern sanitation, medications, and health care has drastically reduced infant and maternal mortality rates and extended average life expectancy. As a result, the world has progressed from the age of pestilence and famine, with a life expectancy between 20 and 40 years, to the age of receding pandemics, with a life expectancy of 30 to 50 years, and now to the current age of degenerative and man-made diseases, with a life expectancy of 60 years or more. These trends, coupled with subsequent declines in fertility rates, have driven a demographic transition in which the major causes of death change from infectious diseases to chronic and degenerative diseases.2 As many countries around the world have undergone globalization, owing to their internal urbanization, modernization, and economic development, an increased proportion of their burden of morbidity and mortality is now due to chronic noncommunicable diseases, including cardiovascular, cerebrovascular, and renovascular diseases as well as cancer, diabetes, chronic respiratory diseases, and mental disorders (Table 6-1).


The term global health is sometimes confused with public health, international health, tropical medicine, and population health. Global health, which is defined as the health of populations in a global context, transcends the perspectives and concerns of individual nations and crosses national borders. Global health depends on the public health efforts and institutions of all countries, including their strategies for improving health, both populationwide and for individuals. Global health depends on multiple factors, including social, political, environmental, and economic determinants of health. Although global health often focuses on improving the health of people who live in low- and middle-income countries, it also includes the health of any marginalized population in any country. Global health requires use of a wide range of institutions that collaborate in addressing all health issues. Global health also depends on the constructive use of evidence-based information to provide health and health equity, in part by strengthening primary health care and the health care delivery system.

Millennium Development Goals

In an attempt to address global inequity, the United Nations advanced eight millennium development goals with the objective of achieving these goals between 2000 and 2015. These eight goals incorporate 21 targets (Table 6-2), with a series of measurable health and economic indicators for each target.3 Although many of the targets have not yet been achieved, substantial progress has been made toward all targets. The millennium development goals emphasize that health and development are interconnected. To address global inequity, fundamental issues include reducing poverty, improving education, and empowering people. In addition to specific goals for reducing infant and child mortality, maternal mortality, and mortality due to infectious diseases such as human immunodeficiency virus infection/acquired immunodeficiency syndrome (HIV/ AIDS), malaria, and tuberculosis, the millennium development goals strongly encourage environmental sustainability and global partnership.


The global burden of disease is measured in terms of total and cause-specific mortality and morbidity as well as the national economic burden for health care. The Global Burden of Diseases, Injuries and Risk Factors Study 20104 shows that an estimated 53 million people died from all causes in 2010, with


CHAPTER 6  Global Health  





1. Age of pestilence and famine

20-40 years


2. Age of receding pandemics

30-50 years

3. Age of degenerative and man-made diseases

50->60 years

3A. Age of delayed degenerative diseases

>60 years

3B. Age of health regression and social upheaval

50-60 years



Infections, rheumatic heart disease, and nutritional cardiomyopathies

Rural India, sub-Saharan Africa, South America


As above plus hypertensive heart disease and hemorrhagic strokes



All forms of strokes; ischemic heart disease at young ages; increasing obesity and diabetes

Aboriginal communities, urban India, former socialist economies

Stroke and ischemic heart disease at old age

Western Europe, North America, Australia, New Zealand

Re-emergence of deaths from rheumatic heart disease, infections, increased alcoholism and violence; increase in ischemic and hypertensive diseases in the young


6.5) (Chapter 425). Patients who use smokeless tobacco products are at significantly increased risk for premalignant and malignant oral lesions (Chapter 32). Bimanual palpation of the cheeks and floor of the mouth facilitates identification of potentially malignant lesions (Chapter 425).


The eye examination begins with simple visual inspection to look for symmetry in the lids, extraocular movements, pupil size and reaction, and the presence of redness (Chapters 423 and 424). Abnormalities in extraocular movements should be grouped into nonparalytic (usually chronic with onset in childhood) or paralytic causes (third, fourth, or sixth cranial nerve palsy). Pupillary abnormalities may be symmetrical or asymmetrical (anisocoria). Red eyes should be categorized by the pattern of ciliary injection, presence of

pain, effect on vision, and papillary abnormalities. When the eye examination is approached systematically, the generalist physician can evaluate the likelihood of conjunctivitis, episcleritis or scleritis, iritis, and acute glaucoma. Routine determination of visual acuity can confirm a patient’s report of diminished vision but does not replace the need for formal ophthalmologic evaluation in patients with visual complaints (Chapter 423). Aging patients often experience acute flashes and floaters, especially with posterior vitreal detachments. If acute flashes and floaters are associated with visual loss, the patient should be urgently referred for an ophthalmologic examination for the evaluation of a possible acute retinal detachment.9 Cataracts can be detected with direct ophthalmoscopy, but the generalist’s proficiency in this evaluation is uncertain. After identifying the optic disc by ophthalmoscopy, the examiner should note the border of the disc for clarity, color, and the size of the central cup in relation to the total diameter (usually less than half the diameter of the disc). A careful observer usually can see spontaneous venous pulsations that indicate normal intracranial pressure, but about 10% of patients with normal intracranial pressure will not have spontaneous pulsations. Abnormalities of the optic disc include optic atrophy (a white disc), papilledema (see Fig. 423-27) (blurry margins with a pink, hyperemic disc), and glaucoma (a large, pale cup with retinal vessels that dive underneath and that may be displaced toward the nasal side). The generalist’s examination inadequately detects early glaucomatous changes, so high-risk patients should undergo routine ophthalmologic examination for glaucoma.10 After inspecting the disc, the upper and lower nasal quadrants should be examined for the appearance of vessels and the presence of any retinal hemorrhages (see Fig. 423-24) or lesions. Proceeding from the nasal quadrants to the temporal quadrants decreases the risk for papillary constriction from the bright light focused on the fovea. Dilating the pupils leads to an improved examination. Patients with diabetes (Chapter 229) should undergo routine examination by eye care experts because the sensitivity of a generalist’s examination is not adequate to exclude diabetic retinopathy or monitor it over time.

Neck Carotid Pulses

The carotid pulses should be palpated for contour and timing in relation to the cardiac impulse. Abnormalities in the carotid pulse contour reflect underlying cardiac abnormalities (e.g., aortic stenosis) but are generally appreciated only after detecting an abnormal cardiac impulse or murmur (Chapter 51). Many physicians listen for bruits over the carotid arteries because asymptomatic carotid bruits are associated with an increased incidence of cerebrovascular and cardiac events in older patients (Chapters 406 and 407). In asymptomatic patients, the presence of a carotid bruit increases the likelihood of a 70 to 90% stenotic lesion (LR 4 to 10), but the absence of a bruit is of uncertain value. Unfortunately, clinical data do not provide adequate data for judging the importance of detecting bruits in asymptomatic patients.

Jugular Veins

The examination of the neck veins is an interesting but often unreliable indicator of central venous pressure or fluid responsiveness in hospitalized sick patients (Chapter 51).11 Inspection of the waveforms may facilitate the interpretation of the cardiac examination for right heart valvular lesions. The waves are seen best by shining a penlight obliquely on the vein while the examiner looks for the dynamic changes of the projected shadow on the bed linen.


The thyroid gland is felt best when standing behind the patient and using both hands to palpate the thyroid gland gently (Chapter 226). Palpation is enhanced when the patient swallows sips of water to allow the thyroid to glide underneath the fingers. When viewed from the side, lateral prominence of the thyroid between the cricoid cartilage and the suprasternal notch indicates thyromegaly. The generalist physician should estimate the size of the thyroid gland as normal or enlarged; the impression of an enlarged thyroid gland by a generalist physician has an LR of almost 4, whereas assessment of normal size makes thyromegaly less likely (LR 0.4). The volume of a normal thyroid gland is no greater than the volume of the patient’s distal thumb phalanx.

Lymphatic System

While palpating the thyroid, the examiner may also identify enlarged cervical lymph nodes (Chapter 168). Lymph nodes can also be palpated in the supraclavicular area, axilla, epitrochlear area, and inguinofemoral region. Simple

CHAPTER 7  Approach to the Patient: History and Physical Examination  

lymph node enlargement confined to one region is common and does not usually represent an important underlying disorder. Unexpected gross lymph node enlargement in a single area or diffuse lymph node enlargement is more important. Patients with febrile illnesses, underlying malignancy, or inflammatory diseases should routinely undergo an examination of each of the aforementioned areas for lymph node enlargement.


Inspection of the patient’s posture may reveal lateral curves in the back (scoliosis) or kyphosis that may be associated with loss of vertebral height from osteoporosis (Chapter 243). When patients have back pain, the spine and paravertebral muscles should be palpated for spasm and tenderness (Chapter 400). The patient may be placed through maneuvers to assess loss of mobility associated with ankylosing spondylitis (Chapter 265), but a history of loss of lateral mobility may be just as efficient in the early stages of spondylitis.


The incremental value of palpation and percussion of the chest to supplement the history, auscultation, and eventual chest radiograph is unknown. Normal vesicular sounds, which approximate a 3:1 inspiratory:expiratory ratio with no pause between phases, are heard throughout most of the normal posterior chest during quiet breathing. Auscultated wheezes are continuous adventitial sounds. Crackles (formerly called rales) are discontinuous sounds heard in conditions that stiffen the lung (heart failure, pulmonary fibrosis, and obstructive lung disease). The best piece of information for increasing the likelihood of chronic obstructive pulmonary disease is a history of more than 40 pack years of smoking (LR 19). The presence of wheezing or downward displacement of the larynx to within 4 cm of the sternum (distance between the top of the thyroid cartilage and the suprasternal notch) increases the likelihood of obstructive pulmonary disease (LR of 4 for either).


The patient should be examined in the sitting and lying positions (Chapter 51). Palpation of the apical impulse in the left lateral decubitus position helps detect a displaced apical impulse and can reveal a palpable S3 gallop. When the apical impulse is lateral to the midclavicular line, radiographic cardiomegaly (LR 3.5) and an ejection fraction of less than 50% (LR 6) are more likely. Most examiners auscultate in sequence the second right then the second left intercostal spaces, the left sternal border, and then the apex. The examiner should concentrate on the timing, intensity, and splitting of sounds with respiration. The first and second heart sounds are heard best with the diaphragm, as are pericardial rubs. Gallops (S3 and S4) are heard best with the stethoscope bell. High-pitched versus low-pitched murmurs are detected by switching from the diaphragm to the bell. The location, timing, intensity, radiation patterns, and respiratory variation of murmurs should be noted. Special maneuvers during auscultation (e.g., Valsalva, auscultation during sudden squatting or standing) do not usually need to be performed if the results of routine precordial examination are entirely normal. The presence of an S3 gallop is useful for detecting left ventricular systolic dysfunction (LR > 4 for identifying patients with an ejection fraction of 4 cm in diameter). However, palpation misses a substantial proportion of small to medium aneurysms (Chapter 78). The presence of bowel sounds in patients with acute symptoms can be falsely reassuring because the sounds can be present despite an ileus and may be increased early in an obstruction. For patients without gastrointestinal symptoms or abnormalities on palpation, auscultation for bruits is important primarily to detect renal bruits in patients with hypertension (Chapters 67 and 125). The presence of an abdominal bruit in a hypertensive patient, if heard in systole and diastole, strongly suggests renovascular hypertension (LR ≈ 40).


Detection of liver disease depends mostly on the history and laboratory evaluations (Chapter 146). By the time that signs are present on physical examination, the patient usually has advanced liver disease. The first abnormalities on physical examination associated with liver disease are extrahepatic. The clinician should assess the patient for ascites, peripheral edema, jaundice, or splenomegaly. In patients with an enlarged liver, palpation should begin at the liver edge, but palpation of the edge below the costal margin increases the likelihood of hepatomegaly only slightly (LR 1.7). The upper border of the liver may be detected by percussion, and a span of less than 12 cm reduces the likelihood of hepatomegaly. In the absence of a known diagnosis (e.g., a hepatoma, which may cause a hepatic bruit), auscultation of the liver rarely is helpful.


Examination for splenomegaly in patients without findings suggestive of a disorder associated with splenomegaly almost always reveals nothing (Chapter 168). Approximately 3% of healthy teenagers may have a palpable spleen. The examination for an enlarged spleen begins first with percussion in the left upper quadrant to detect dullness. Palpation can be performed by any of the following three approaches (κ ≈ 0.2 to 0.4): palpating with the right hand while providing counterpressure with the left hand behind the spleen, palpating with one hand without counterpressure (with the patient in the right lateral decubitus position for both techniques), or placing the patient supine with the left fist under the left costovertebral angle while the examiner tries to hook the spleen with the hands.

Musculoskeletal System

The musculoskeletal examination in adult patients is almost always driven by symptoms (Chapters 256 and 263). Most patients have back pain at some point during their life (Chapter 400). The patient’s history helps assess the likelihood of an underlying systemic disease (age, history of systemic malignancy, unexplained weight loss, duration of pain, responsiveness to previous therapy, intravenous drug use, urinary infection, or fever). The most important physical examination findings for lumbar disc herniation in patients with sciatica all have excellent reliability, including ipsilateral straight leg raising causing pain, contralateral straight leg raising causing pain, and ankle or great toe dorsiflexion weakness. The generalist physician should evaluate an adult patient with knee discomfort for torn menisci or ligaments. The best maneuver for demonstrating a tear in the anterior cruciate ligament is the anterior drawer or Lachman maneuver, in which the examiner detects the lack of a discrete end point as the tibia is pulled toward the examiner while the femur is stabilized. A variety of maneuvers that assess for pain, popping, or grinding along the joint line between the femur and tibia are used to evaluate for meniscal tears. As with many musculoskeletal disorders, no single finding has the accuracy of the orthopedist’s examination, which factors in the history and a variety of clinical findings. The shoulder examination is directed toward determining range of motion, maneuvers that cause discomfort, and assessment of functional disability.

Hip osteoarthritis is detected by evidence of restriction of internal rotation and abduction of the affected hip. Generalist physicians often rely on radiographs to determine the need for referral to orthopedic physicians, but routine radiographs are not needed early in the course of shoulder or hip disorders. The degree of pain and disability experienced by the patient may prompt confirmation of the diagnosis and referral. The hands and feet may show evidence of osteoarthritis (local or as part of a systemic process) (Chapter 262), rheumatoid arthritis (Chapter 264), gout (Chapter 273), or other connective tissue diseases. In addition to regional musculoskeletal disorders, such as carpal tunnel syndrome, a variety of medical and neurologic conditions should prompt routine examination of the distal ends of the extremities to prevent complications (e.g., diabetes [neuropathy or ulcers] or hereditary sensorimotor neuropathy [claw toe deformity]).


The skin should be examined under good lighting (Chapter 436). It is best to ask the patient to point out any spots on the skin of concern. Examiner agreement on some of the most important features of melanoma (asymmetry, haphazard color, border irregularity) is fair to moderate (Chapter 203). A lesion that is symmetrical, has regular borders, is only one color, is 6 mm or smaller, or has not enlarged in size is unlikely to represent a melanoma (LR 0.07). However, an increasing number of findings greatly enhance the likelihood of melanoma (LR 2.6 for two or more findings and LR 98 for the presence of all five findings) (Chapter 203). Basal cell carcinoma and squamous cell carcinoma occur more frequently than melanoma (Chapter 203). These lesions can be detected during routine examination by paying careful attention to sun-exposed areas of the nose, face, forearms, and hands.

Neurologic Examination

Full details of the neurologic examination are given in Chapter 396.

Psychiatric Evaluation

During the general examination, much of the psychiatric assessment (including cognition) is accomplished while eliciting the routine history and performing the review of systems (Chapter 397). Observation of the patient’s mannerisms, affect, facial expression, and behavior may suggest underlying psychiatric disturbances. When a screening survey and review of systems are obtained by a questionnaire completed by the patient, the clinician should review the responses carefully to determine whether the patient exhibits symptoms of depression. Specific questioning for symptoms of depression is appropriate for all adult patients. Military veterans should be screened for post-traumatic stress disorder and possible prior traumatic brain injuries that may affect their behaviors. Delirium (Chapter 28) is common in both medical and surgical inpatients and is recognized by fluctuating mental status. Delirium should be suspected when the patient has trouble carrying on a normal conversation during bedside rounds; but the patient’s nurse and visitors may detect delirium before the physician, so their report may contribute to diagnosis.12

Genitalia and Rectum Pelvic Examination

A complete examination includes a description of the external genitalia, appearance of the vagin* and cervix as seen through a speculum, and bimanual palpation of the uterus and ovaries (Chapters 199 and 237). About 10 to 15% of asymptomatic women have some abnormality on examination, and 1.5% have abnormal ovaries. However, screening for ovarian cancer is limited by the low sensitivity of the physical examination for detecting early-stage ovarian carcinoma (Chapter 199). In the emergency setting, all women of reproductive age with vagin*l bleeding and pelvic pain should have a pregnancy test and an ultrasound to evaluate them for a possible ectopic pregnancy.13

Male Genitalia

Examination of the male genitalia should begin with a description of whether the penis is circumcised and whether there are any visible skin lesions (e.g., ulcers or warts). Palpation should confirm the presence of bilateral testes in the scrotum. The epididymis and testes should be palpated for nodules. The low incidence of testicular carcinoma means that most nodules are benign (Chapter 200). The prostate should be examined in all quadrants, with attention focused on surface irregularities or differences in consistency throughout the prostate

(Chapter 201). An estimate of prostate size may be confounded by the size of the examiner’s fingers. It may be best to estimate the size of the prostate in centimeters of width and height.


Patients can be examined while lying on their side, although this approach may place the examiner in an awkward stance (Chapters 132 and 145). The rectal examination in women can be performed as part of a bimanual examination, with the index finger in the vagin* and the third finger in the rectum to permit palpation of the rectovagin*l vault. Men may be asked to stand and lean over the examining table; alternatively, they may be examined while on their back with their hips and knees flexed. This latter maneuver is not used often, although it may facilitate examination of the prostate, which falls into the finger in this position. The rectal examination begins with inspection of the perianal area for skin lesions. A well-lubricated, gloved finger is placed on the anus, and while applying gentle pressure, the examiner asks that the patient bear down as though having a bowel movement. This maneuver facilitates entry of the finger into the rectum. A normal rectal response includes tightening of the anal sphincter around the finger. The examiner should palpate circumferentially around the length of the fully inserted finger for masses. On withdrawing the gloved finger, the finger should be wiped on a stool guaiac card for fecal blood testing to assess for acute blood loss. As a screening test for colorectal carcinoma (Chapter 193), digital examination does not replace the need for testing stool samples collected by the patient (or using alternative screening strategies, such as flexible sigmoidoscopy or colonoscopy).


The physician should summarize the pertinent positive and negative findings for the patient and be willing to express uncertainty to the patient, provided that it is accompanied by a plan of action (e.g., “I will reexamine you on your next visit”). The rationale for subsequent laboratory, imaging, or other tests should be explained. A plan should be established for providing further feedback and results to the patient, especially when there is a possibility that bad news may need to be delivered. Some physicians ask the patient if there is “anything else” to be covered. Patients who express additional new concerns at the end of the visit may have been fearful to address them earlier (e.g., “by the way, doctor, I’m getting a lot of chest pain”); when the problems seem non-urgent, it is acceptable to reassure the patient and offer the promise of evaluating the patient in a follow-up phone call or at the next visit.


The common assumption that physicians’ diagnostic skills are deteriorating is not supported by evidence. There is considerable evidence that the scientific approach to understanding what is worthwhile and what is not worthwhile during the clinical examination identifies a core set of skills for clinical diagnosticians. Because good patient outcomes at good value are driven primarily by the quality of the information obtained during the clinical examination, continued application of scientific principles to the history and physical examination should improve diagnostic skills. GENERAL REFERENCES For the General References and other additional features, please visit Expert Consult at

CHAPTER 7  Approach to the Patient: History and Physical Examination  

GENERAL REFERENCES 1. Krogsboll LT, Jorgensen KJ, Larsen CG, et al. General health checks in adults for reducing morbidity and mortality from disease: Cochrane systematic review and meta-analysis. BMJ. 2012;345:e7191. 2. Verghese A, Brady E, Kapur CC, et al. The bedside evaluation: ritual and reason. Ann Intern Med. 2011;155:550-553. 3. Powers BJ, Trinh JV, Bosworth HB. Can this patient read and understand written health information? JAMA. 2010;304:76-84. 4. Makadon HJ. Ending LGBT invisibility in health care: the first step in ensuring equitable care. Cleve Clin J Med. 2011;78:220-224. 5. Department of Veterans Affairs. Military Health History Pocket Card for Clinicians. http://www. Accessed February 9, 2015. 6. Murff HJ, Spigel DR, Syngal S. Does this patient have a family history of cancer? An evidence-based analysis of the accuracy of family cancer history. JAMA. 2004;292:1480-1489. 7. Brenner S, Guder G. The patient with dyspnea. Rational diagnostic evaluation. Herz. 2014; 39:8-14.


8. Bagai A, Thavendiranathan P, Detsky AS. Does this patient have hearing impairment? JAMA. 2006;295:416-428. 9. Hollands H, Johnson D, Brox AC, et al. Acute-onset floaters and flashes: is this patient at risk for retinal detachment? JAMA. 2009;302:2243-2249. 10. Hollands H, Johnson D, Hollands S, et al. Do findings on routine examination identify patients at risk for primary open-angle glaucoma? The rational clinical examination systematic review. JAMA. 2013;309:2035-2042. 11. Marik PE, Cavallazzi R. Does the central venous pressure predict fluid responsiveness? An updated meta-analysis and a plea for some common sense. Crit Care Med. 2013;41:1774-1781. 12. Wong CL, Holroyd-Leduc J, Simel DL, Straus SE. Does this patient have delirium? Value of bedside instruments. JAMA. 2010;304:779-786. 13. Crochet JR, Bastian LA, Chireau MV. Does this woman have an ectopic pregnancy? The rational clinical examination systematic review. JAMA. 2013;309:1722-1729.


CHAPTER 7  Approach to the Patient: History and Physical Examination  

REVIEW QUESTIONS 1. A 24-year-old-woman with right lower quadrant abdominal pain and vagin*l bleeding has a positive home pregnancy. Before a pelvic ultrasound is ordered, the pelvic examination is performed and reveals cervical motion tenderness. Cervical motion tenderness has the following diagnostic characteristics for an ectopic pregnancy: sensitivity 45%, specificity 91%, likelihood ratio positive (LR+) 4.9, LR negative (LR−) 0.62, positive predictive value (PPV) 46%. Which of these values is most helpful for assessing the probability that she has an ectopic pregnancy? A. Sensitivity B. Specificity C. Likelihood ratio positive D. Likelihood ratio negative E. Positive predictive value Answer: C  The sensitivity is the percentage of patients who have cervical motion tenderness among women with an ectopic pregnancy. The specificity is the percentage of patients who do not have cervical motion tenderness among women without an ectopic pregnancy. These individual values are not helpful for this particular patient (Crochet JR, Bastian LA, Chireau MV. does this woman have an ectopic pregnancy? The rational clinical examination systematic review. JAMA. 2013;309:1722-1729) because without knowing whether or not she has an ectopic pregnancy, we do not know which result applies. The PPV describes the probability of ectopic pregnancy when there is cervical motion tenderness. To use the PPV for this patient requires knowing the prevalence of disease from which the PPV was derived. Without knowing that value, you cannot be certain whether the PPV is appropriate for your patient. The LR+ quantifies the increase in odds of disease among those with cervical motion tenderness, and the LR− quantifies the decrease in odds of disease when cervical motion tenderness is absent. Based on your assessment of the prior probability of ectopic pregnancy, which is 10 to 20% among all pregnant women with abdominal pain and/or vagin*l bleeding, the LR+ is the most relevant value because it can be used to determine the increase in likelihood of ectopic pregnancy (Chapter 10). 2. The primary role of completing a review of systems in helping with clinical diagnosis during the clinical evaluation is which of the following? A. To review symptoms associated with the presenting problem B. To pick up the presence of concerns that were uncomfortable to address during the history of the present illness C. To complete a record that enhances patient billing D. To focus on the presence or absence of findings during a constrained time period E. To allow open-ended questioning that enhances the patient’s relationship with the physician Answer: D  A complete medical history should reveal the most important symptoms experienced by the patient that are pertinent to the presenting problems. Open-ended questioning should occur during elicitation of the medical history. Although completing a review of systems may be required for electronic medical records and billing, simply fulfilling that function does not help with diagnosis. However, about 10% of the time, the review of systems might pick up on an important finding that was not addressed during the rest of the evaluation. The main purpose of the review of systems is to use direct questioning to pick up on a limited set of symptoms, during a specified recent time interval (e.g., over the past week, or the past month). This approach helps the patient focus on a well-defined and recent time period that should enhance the reliability of their answers that pertain to their current condition.

3. While conducting your new patient evaluation on a 35-year-old male veteran who served in combat duty in Afghanistan, you note that he seems anxious and that he is questioning you frequently about a variety of difficulties in making his appointment and the attitudes of your office staff. You recognize the need to screen for post-traumatic stress disorder using a four-item screening instrument. You start by asking, “In your life, have you ever had an experience so horrible, frightening, or upsetting that, in the past month you …” All of the items below are part of the four-item screening instrument for post-traumatic stress disorder that follows this introduction except: A. Have little interest or pleasure in doing things? B. Have had nightmares about it or thought about it when you did not want to? C. Tried hard not to think about it or went out of your way to avoid situations that reminded you of it? D. Were constantly on guard, watchful, or easily startled? E. Felt numb or detached from others, activities, or your surroundings? Answer: A  The loss of interest or pleasure in doing things is a symptom that suggests depression. The other symptom that is elicited in a two-item screener for depression asks the patient whether they feel down, depressed, or hopeless (Kroenke K, Spitzer RL, Williams JB. The Patient Health Questionnaire-2: validity of a two-item depression screener. Med Care. 2003;41:1284-1292). Patients who answer “yes” to at least one of the items have an LR+ of 2.7 for depression, whereas answering “no” to both questions makes depression much less likely with an LR of 0.14 (Williams JW Jr, Noël P, Cordes JA, et al. Is this patient clinically depressed? JAMA. 2002;287:11601170). Although it is important to screen for depression among patients with post-traumatic stress disorder (military related or not) (Campbell DG, Felker BL, Liu CF, et al. Prevalence of depression-PTSD comorbidity: implications for clinical practice guidelines and primary care-based interventions. J Gen Intern Med. 2007;22:711-718), screening for depression alone may not detect the associated post-traumatic stress disorder.


CHAPTER 8  Approach to the Patient with Abnormal Vital Signs  

8  APPROACH TO THE PATIENT WITH ABNORMAL VITAL SIGNS DAVID L. SCHRIGER Care of the patient is guided by integration of the chief complaint, history, vital signs, and physical examination findings (Chapter 7). Physicians should be keenly aware of a patient’s vital signs but should seldom make them the centerpiece of the evaluation.

CHAPTER 8  Approach to the Patient with Abnormal Vital Signs  




36°-38° C (96.8°-100.4° F)

40° C (104° F)


60-100 beats/min

130 beats/min


12-20 breaths/min

26 breaths/min

Oxygen saturation


115 mm Hg) should stimulate an evaluation for hypertensive urgencies (Chapter 67). Note that hypertension in the absence of signs of acute end-organ damage does not require acute treatment, which can reduce intracranial perfusion pressure and cause stroke. Patients with elevated blood pressure should be offered standard evaluation and treatment for chronic hypertension (Chapter 67). Markedly low pulse or blood pressure in patients receiving cardioactive medications should lead to a confirmation that the patient is truly asymptomatic, an inquiry into the dosing of these medications, and a reconsideration of the regimen. Markedly low pulse in elderly patients who are not receiving rate-controlling drugs should trigger an evaluation of the patient’s cardiac conduction system. Oxygen saturation below 93% in the absence of known pulmonary problems should prompt an evaluation of the patient’s pulmonary status.

measurement of an elevated blood pressure leading to a diagnosis of hypertension is the classic example of the value of vital signs in such patients.

Patients Who Complain of Systemic Illness but Do Not Appear to Be Very Ill

Vital signs serve two additional roles in symptomatic patients who do not appear particularly ill. First, abnormalities in vital signs provide information that may suggest or support a diagnosis. The presence of elevated temperature in a patient with productive cough, shortness of breath, and localized rales and egophony supports a diagnosis of infectious pneumonia. Vital signs may also play a role in defining therapy and triage. For example, guidelines for patients with community-acquired pneumonia (Chapter 97) formally incorporate vital signs. The second role of vital signs in the stable symptomatic patient is to provide warning that the patient is sicker than he or she appears. For example, the presence of hypotension in a well-appearing patient thought to have pyelonephritis may be an indication of sepsis or hypovolemia. For vital signs to be of use, the physician must be aware of them and must incorporate them explicitly into a thought process that considers the dangerous diagnoses associated with the abnormal vital sign. The physician then must decide whether the likelihood of each potentially dangerous diagnosis is high enough to warrant specific evaluation. Unfortunately, no quick or easy rules differentiate spurious abnormalities that can be ignored from those that should trigger additional testing or treatment. What can be said is that the well-trained physician who is aware of abnormal vital signs and is willing to contemplate a change in treatment or disposition in response to them is less likely to make mistakes. A few specific points bear mention. First, for most vital signs, “normal” is relative. Blood pressure must be interpreted in the context of the patient. For example, a blood pressure of 88/64 mm Hg may be reasonable for an otherwise healthy, young 50-kg woman but should cause concern in a 90-kg middle-aged man. Similarly, a blood pressure of 128/80 mm Hg would be fine in a 60-year-old man but worrisome in a 34-week pregnant woman. Second, because vital signs are insensitive measures of disease, normal vital signs should not dissuade the physician from pursuing potentially critical diagnoses. For example, young, well-conditioned adults may maintain normal vital signs well into the course of shock.

Use of Vital Signs in Patients Who Appear to Be Ill

For some patients, abnormal vital signs are expected on the basis of their appearance and their symptoms. For patients in extremis, care should proceed according to established guidelines such as Advanced Cardiac Life Support (Chapter 63), Advanced Trauma Life Support, and algorithms for the treatment of shock (Chapters 107 and 108). For other ill-appearing patients, two processes must occur. In one, the physician, armed with knowledge of the differential diagnosis of each abnormal vital sign and the ability to take a thorough history and to perform an appropriate physical examination, narrows the list of potential diagnoses and decides which are of sufficient probability to warrant evaluation. Simultaneously, the physician considers

the list of treatment options for all diagnoses associated with the abnormal vital sign and, before establishing a diagnosis, initiates those treatments for which the potential benefit of prompt administration exceeds potential harms. For example, antibiotics for febrile patients at risk for bacterial infection, hydrocortisone for hypotensive patients at risk for hypoadrenalism, and thiamine for hypothermic patients at risk for Wernicke encephalopathy may improve outcome and are unlikely to cause harm even if the patient does not have the suspected condition. Although early presumptive treatment can be life-saving in selected patients, it should not be abused; physicians must avoid knee-jerk responses that can cause harm.

Differential Diagnosis and Treatment Options Single Abnormal Vital Signs

Because vital signs can be abnormal in virtually any disease process, no differential diagnosis can be encyclopedic. The physician should focus initially on common diseases and diseases that require specific treatment. The thought process should begin with the chief complaint and history and then incorporate information about the vital signs and the remainder of the physical examination.

Multiple Abnormal Vital Signs

Patients who are acutely ill are likely to have several abnormal vital signs. Although certain patterns of abnormal vital signs predominate in specific conditions (e.g., hypotension, tachycardia, and hypothermia in profound sepsis), no pattern can be considered pathognomonic. The physician’s goal is to work toward a diagnosis while simultaneously providing treatments whose benefits outweigh potential harms. Fever is generally accompanied by tachycardia, with the general rule of thumb that the heart rate will increase by 10 beats per minute for every 1° C increase in temperature. The absence of tachycardia with fever is known as pulse-temperature dissociation and has been reported in typhoid fever (Chapter 308), legionnaires disease (Chapter 314), babesiosis (Chapter 353), Q fever (Chapter 327), infection with Rickettsia spp (Chapter 327), malaria (Chapter 345), leptospirosis (Chapter 323), pneumonia caused by Chlamydia spp (Chapter 318), and viral infections such as dengue fever (Chapter 382), yellow fever (Chapter 381), and other viral hemorrhagic fevers (Chapter 381), although the predictive value of this finding is unknown. Much can be learned by comparing the respiratory rate with pulse oximetry. Hyperventilation in the presence of high oxygen saturation suggests a central nervous system process or metabolic acidosis rather than a cardiopulmonary process. Low respiratory rates in the presence of low levels of oxygen saturation suggest central hypoventilation, which may respond to narcotic antagonists. Hypertension and bradycardia in the obtunded or comatose patient are known as the Cushing reflex, a relatively late sign of elevated intracranial pressure. Physicians should strive to diagnose and treat this condition before the Cushing reflex develops.

Approach to Abnormalities of Specific Vital Signs Elevated Temperature

Normal temperature is often cited as 37° C (98.6° F), but there is considerable diurnal variation and variation among individuals, so 38° C is the most commonly cited threshold for fever. Fever thought to be due to infection should be treated with antipyretics and appropriate antimicrobials (Chapter 280). The importance of early administration of antibiotics to potentially septic patients cannot be overstated (Chapters 280 and 281). Hyperthermia (temperature above 40° C) should be treated with cooling measures such as ice packs, cool misting in front of fans, cold gastric lavage, and, for medicationrelated syndromes, medications such as dantrolene (Chapter 109). Most hospital anesthesia departments will have a designated kit for the treatment of malignant hyperthermia (Chapters 432 and 434).

Low Temperature

The treatment of hypothermia is guided by its cause (Chapter 109). The body’s temperature decreases when heat loss exceeds heat production. Every logically possible mechanism for this phenomenon has been observed. Decreased heat production can result from endocrine hypofunction (e.g., Addison disease [Chapter 227], hypopituitarism [Chapter 224], hypothyroidism [Chapter 226]) and loss of the ability to shiver (e.g., drug-induced or neurologic paralysis or neuromuscular disorders). Malfunction of the hypothalamic regulatory system can be due to hypoglycemia (Chapter 229) and a variety of central nervous system disorders (Wernicke encephalopathy

CHAPTER 8  Approach to the Patient with Abnormal Vital Signs  

[Chapter 416], stroke [Chapter 407], tumor [Chapter 189], and trauma [Chapter 399]). Resetting of the temperature set point can occur with sepsis. Increased heat loss can be due to exposure, behavioral and physical disorders that prevent the patient from sensing or responding to cold, skin disorders that decrease its ability to retain heat, and vasodilators (including ethanol). A careful history and physical examination should illuminate which of these possibilities is most likely. Several considerations are worthy of emphasis. The spine of an obtunded hypothermic patient who is “found down” must be protected and evaluated because paralysis from a fall may have prevented the patient from seeking shelter and may have diminished the ability to produce heat. The physician should not forget to administer antibiotics to patients who may be septic (Chapter 108), thiamine to those who may have Wernicke encephalopathy (Chapter 416), hydrocortisone to those who may be hypoadrenal (Chapter 227), and thyroid hormone to those who may have myxedema coma (Chapter 226). Severely hypothermic patients (Chapter 109) should be treated gently because any stimulation may trigger ventricular dysrhythmias; even in the absence of pulses, cardiopulmonary resuscitation should be used only in patients with ventricular fibrillation or asystole.

Elevated Heart Rate

The rate, rhythm, and electrocardiogram differentiate sinus tachycardia from tachyarrhythmias (Chapters 62 to 65). Tachyarrhythmias can be instigated by conditions that may require specific treatment (e.g., sepsis [Chapter 108], electrolyte disorders [Chapters 116, 117, and 118], endocrine disorders [Chapter 221], and poisonings [Chapters 22 and 110]) before the arrhythmia is likely to resolve. For sinus tachycardia, treatment of the underlying cause is always paramount. Treatments may include antipyretics (for fever); anxiolytics; oral or intravenous fluids (for hypovolemia); nitrates, angiotensinconverting enzyme inhibitors, and diuretics (for heart failure and fluid overload [Chapter 59]); oxygen (for hypoxemia); α-blockers (for stimulant overdose); β-blockers (for acute coronary syndromes [Chapters 72 and 73] or thyroid storm [Chapter 226]); and anticoagulation (for pulmonary embolism [Chapter 98]). Tachycardia is often an appropriate response to a clinical condition and should not be treated routinely unless it is causing or is likely to cause secondary problems.

Low Pulse

Bradycardia can be physiologic (athletes and others with increased vagal tone), due to prescribed cardiac medications (e.g., β-blockers, calciumchannel blockers, digoxin), overdoses (e.g., cholinergics, negative chronotropic agents), disease of the cardiac conducting system, electrolyte abnormalities (severe hyperkalemia), and inferior wall myocardial infarction (Chapters 64 and 73). Asymptomatic patients do not require immediate treatment. The goal of therapy is to produce a heart rate sufficient to perfuse the tissues and alleviate the symptoms (Chapter 63). Overdoses should be treated with specific antidotes (Chapter 110). Endocrine disorders should be treated with replacement therapy. In patients with acute coronary syndrome (Chapter 72), the goal is to restore perfusion and alleviate the ischemia. Patients with profound bradycardia or hypotension may require chronotropic drugs to increase perfusion even if these agents increase myocardial oxygen demand. In normotensive patients with milder bradycardia, chronotropic agents should be used only if symptoms and ischemia cannot be resolved by other means. Atropine is the primary therapy for bradycardia; isoproterenol and cardiac pacing are reserved for those who do not respond (Chapter 63).

Elevated Blood Pressure

Elevated blood pressure does not require acute treatment in the absence of symptoms or signs of end-organ damage (Chapter 67). In patients whose blood pressure is markedly above their baseline, the history and physical examination should assess for the conditions that define “hypertensive emergency”: evidence of encephalopathy, intracranial hemorrhage, ischemic stroke, heart failure, pulmonary edema, acute coronary syndrome, aortic dissection, renal failure, and preeclampsia. In the absence of these conditions, treatment should consist of restarting or adjusting the medications of patients with known hypertension and initiating a program of blood pressure checks and appropriate evaluation for those with no prior history of hypertension (Chapter 67). The patient with a true hypertensive emergency should be treated with agents appropriate for the specific condition. Because rapid decreases in blood pressure can be as deleterious as the hypertensive state itself, intrave-


nous agents with short half-lives, such as nitroprusside, labetalol, nitroglycerin, and esmolol, are preferred (Chapter 67).

Low Blood Pressure

Low blood pressure must be evaluated in the context of the patient’s symptoms, general appearance, and physical examination findings. Treatment depends on context. The same blood pressure value may necessitate intravenous inotropic agents in one patient and no treatment in another. In tachycardic hypotensive patients, the physician must rapidly integrate all available evidence to determine the patient’s volume state, cardiac function, vascular capacitance, and primary etiology (Chapter 106). Not all patients with hypotension and tachycardia are in shock, and not all patients in shock will have hypotension and tachycardia. Patients in shock should be treated on the basis of the cause (Chapters 106 to 108). Symptomatic hypotensive patients thought to be intravascularly volume depleted should receive intravenous fluid resuscitation with crystalloid or blood, depending on their hemoglobin level (Chapter 106). In patients with known heart disease, patients who are frail or elderly, and patients whose volume status is uncertain, small boluses of fluid (e.g., 250 mL of normal saline), each followed by reassessment, are preferred so that iatrogenic heart failure may be avoided. Inotropic support should be reserved for patients who do not respond to fluid resuscitation. High-output heart failure should be kept in mind in patients with possible thyroid storm or stimulant overdose.

Increased Respiratory Rate

Tachypnea is a normal response to hypoxemia (see later). Treatment of tachypnea in the absence of hypoxemia is directed at the underlying cause, which often is pain (Chapter 30). Anxiolytics (e.g., diazepam, 5 to 10 mg PO or IV; lorazepam, 1 to 2 mg PO, IM, or IV) or reassurance can calm patients with behavioral causes of hyperventilation. Breathing into a paper bag has been shown to be an ineffective treatment. Pulmonary embolism (Chapter 98) does not necessarily reduce the oxygen saturation or cause a low Po2 and should always be considered in at-risk patients with unexplained tachypnea.

Decreased Respiratory Rate

Any perturbation of the respiratory center in the central nervous system can slow the respiratory drive (Chapter 86). Narcotics and other sedatives and neurologic conditions are common causes of a decreased respiratory rate. The primary treatment of apnea is mechanical ventilation (Chapter 105), but narcotic antagonists can be tried in patients with a history or physical examination findings (miosis, track marks, opiate patch) suggestive of narcotic use or abuse (Chapter 34). In nonapneic patients, mechanical ventilation is indicated for patients who are breathing too slowly to maintain an acceptable oxygen saturation and for patients who are retaining carbon dioxide in quantities sufficient to depress mental function. Patients who are unable to protect their airway should be intubated. Oxygen should be administered to all hypopneic patients who are hypoxemic (see earlier). Patients with chronic hypoventilation (Chapter 86) may have retained HCO3– to compensate for an elevated Pco2 and so may depend on hypoxia to maintain respiratory drive; in these patients, overaggressive administration of oxygen can decrease the respiratory rate, increase the Pco2, and increase obtundation (Chapter 104).

Decreased Oxygen Saturation

In hypopneic patients, initial efforts should try to increase the respiratory rate (see earlier) and tidal volume. Regardless of etiology, oxygen, in amounts adequate to restore adequate oxygen saturation (Po2 > 60 mm Hg, oxygen saturation >90%), is the mainstay of therapy. When oxygen alone fails, noninvasive methods for improving ventilation or tracheal intubation are required (Chapter 104). Oxygen should increase the Po2 in all patients except those who have severe right-to-left shunting (Chapter 69). Treatment of conditions that cause hypoxemia includes antibiotics (pneumonia), bronchodilators (asthma, chronic obstructive pulmonary disease), diuretics and vasodilators (pulmonary edema), anticoagulants (pulmonary embolism), hyperbaric oxygen (carbon monoxide poisoning), methylene blue (methemoglobinemia, sulfhemoglobinemia), and transfusion (anemia). GENERAL REFERENCES For the General References and other additional features, please visit Expert Consult at

CHAPTER 8  Approach to the Patient with Abnormal Vital Signs  

GENERAL REFERENCES 1. Straede M, Brabrand M. External validation of the simple clinical score and the HOTEL score, two scores for predicting short-term mortality after admission to an acute medical unit. PLoS ONE. 2014;9:e105695. 2. Lamantia MA, Stewart PW, Platts-Mills TF, et al. Predictive value of initial triage vital signs for critically ill older adults. West J Emerg Med. 2013;14:453-460.


3. Lighthall GK, Markar S, Hsiung R. Abnormal vital signs are associated with an increased risk for critical events in US veteran inpatients. Resuscitation. 2009;80:1264-1269. 4. Gabayan GZ, Sun BC, Asch SM, et al. Qualitative factors in patients who die shortly after emergency department discharge. Acad Emerg Med. 2013;20:778-785. 5. Bleyer AJ, Vidya S, Russell GB, et al. Longitudinal analysis of one million vital signs in patients in an academic medical center. Resuscitation. 2011;82:1387-1392.


CHAPTER 8  Approach to the Patient with Abnormal Vital Signs  

REVIEW QUESTIONS 1. A patient presents with malaise, cough, and shortness of breath. Vital signs include temperature 40° C, blood pressure 120/74 mm Hg, respiratory rate 18 breaths per minute, pulse 70 beats per minute, and oxygen saturation 97%. This presentation could be consistent with: A. Streptococcal pneumonia B. Pyelonephritis due to Escherichia coli C. Legionella pneumonia D. Influenza-like illness E. Mycoplasma pneumonia Answer: C  This patient is exhibiting a pulse-temperature dissociation because the pulse (70) is far lower than one would expect given that the patient is febrile to 40° C. This phenomenon is seen in a number of conditions, including typhoid fever and legionella infection. The other conditions would all be expected to produce tachycardia unless the patient could not become tachycardic because of medications (e.g., β-blockers) or cardiac conduction problems. 2. An 88-year-old man presents from a nursing home with slight agitation and vital signs that include temperature 38.7° C, blood pressure 96/64 mm Hg, respiratory rate 22 breaths per minute, pulse 94 beats per minute, and oxygen saturation 96%. Physical examination reveals dry mucous membranes, clear lungs, a soft abdomen, an indwelling Foley catheter, and slightly cool but noncyanotic extremities. The patient should be given: A. Antipyretics (e.g., acetaminophen) B. Intravenous normal saline, 500 mL with additional boluses as tolerated C. Intravenous antibiotics D. All of the above E. Only A and B until urine culture results are available Answer: D  This case is an example of how vital signs can guide treatment in the absence of a firm diagnosis. The patient meets all three of the physical examination criteria for the systemic inflammatory response syndrome (SIRS) and is likely septic. The physician should not wait for his white blood cell count or other laboratory results to initiate antibiotic treatment because evidence suggests that early antibiotics are a crucial step in preventing morbidity and mortality. Although antibiotics should not be overused, the early provision of appropriate broad-spectrum antibiotics before the confirmation of a specific diagnosis is prudent and may be life-saving for this patient. 3. An intern is awakened at 3 am by the ward nurse regarding a patient who is postoperative day 2 from a hip replacement and is newly tachycardic. Vital signs include temperature 36° C, blood pressure 146/82 mm Hg, respiratory rate 18 breaths per minute, pulse 112 beats per minute, and oxygen saturation 97% on room air. The intern drowsily orders a 1000-mL normal saline fluid challenge for dehydration. Later that morning, the patient is acutely intubated for respiratory distress. What most likely went wrong? A. The intern failed to consider pulmonary embolism as a possible cause for the tachycardia. B. The intern failed to consider fat embolism in a patient who had recently undergone hip surgery. C. The intern failed to consider sepsis in the differential diagnosis. D. The intern failed to consider failure of the patient controlled anesthesia (PCA) pump in the differential diagnosis. E. The intern failed to realize that tachycardia can be present in both dehydration and heart failure. Answer: E  First and foremost, the intern’s main mistake was not getting out of bed to evaluate the patient in person. Vital signs alone are not sufficient data on which to base an important clinical decision. Although choices A and B are certainly possible in a postoperative orthopedic patient, heart failure is a more likely diagnosis. There is little clinical support for the other choices.

4. A patient arrives in the emergency department comatose with decreased respiratory rate in the winter. Vital signs are temperature 36° C, blood pressure 128/68 mm Hg, respiratory rate 10 breaths per minute, pulse 100 beats per minute, and oxygen saturation 100% on room air. Pupils are 6 mm and reactive, and lungs are clear. What is the single most important initial treatment? A. High-flow O2 administered by non-rebreather mask B. Intravenous normal saline, 1000 mL with additional boluses as tolerated C. Intravenous antibiotics D. Naloxone, 0.8 mg IV E. Immediate endotracheal intubation Answer: A  This patient may have carbon monoxide poisoning. It is winter, a time when people use heating devices that may have incomplete combustion. The pupillary examination is not suggestive of opiate intoxication (D), and no other diagnosis is apparent. Because oxygen is the best treatment for this condition and is generally harmless in adults, it makes sense to initiate this therapy while efforts (e.g., blood gas analysis with co-oximetry) are made to confirm the diagnosis. There is no basis for thinking this patient is dehydrated (B) or infected (C), and intubation would be premature (E). Remember that pulse oximetry is falsely elevated in carbon monoxide poisoning, so the 100% oxygen saturation means nothing.


CHAPTER 9  Statistical Interpretation of Data  



Much of medicine is inherently probabilistic. Not everyone with hypercholesterolemia who is treated with a statin is prevented from having a myocardial infarction, and not everyone not treated does have one, but statins reduce the probability of a myocardial infarction in such patients. Because so much of medicine is based on probabilities, studies must be performed on groups of people to estimate these probabilities. Three component tasks of statistics are: selecting a sample of subjects for study, describing the data from that sample, and drawing inferences from that sample to a larger population of interest.1


The goal of research is to produce generalizable knowledge, so that measurements made by researchers on samples of individuals will eventually help draw inferences to a larger group of people than was studied. The ability to draw such inferences depends on how the subjects for the study (the sample) were selected. To understand the process of selection, it is helpful to begin by identifying the group to which the results are to be generalized and then work backward to the sample of subjects to be studied.

Target Population

The target population is the population to which it is hoped the results of the study will be generalizable. For example, to study the efficacy of a new drug to treat obesity, the target population might be all people with a certain level of obesity (e.g., body mass index [BMI] of ≥30 kg/m2) who might be candidates for the drug.


The intended sample is the group of people who are eligible to be in the study based on meeting inclusion criteria, which specify the demographic, clinical, and temporal characteristics of the intended subjects, and not meeting exclusion criteria, which specify the characteristics of subjects whom the investigator does not wish to study. For example, for the study of a new obesity drug, the intended sample (inclusion criteria) might be men and women 18 years or older who live in one of four metropolitan areas, who have a BMI of 30 kg/m2 or higher, and who have failed an attempt at weight loss with a standard diet. Exclusion criteria might include an inability to speak English or Spanish, known alcohol abuse, plans to leave the area in the next 6 months, and being pregnant or planning to become pregnant in the next 6 months. In some cases, particularly large population health surveys such as the National Health and Nutrition Examination Survey (NHANES), the intended sample is a random sample of the target population. A simple random sample is a sample in which every member of the target population has an equal chance of being selected. Simple random samples are the easiest to handle statistically but are often impractical. For example, if the target population is the entire population of the United States (as is the case for NHANES), a simple random sample would include subjects from all over the country. Getting subjects from thousands of distinct geographic areas to examination sites would be logistically difficult. An alternative, used in NHANES, is cluster sampling, in which investigators take a random sample of “clusters” (e.g., specific census tracts or geographic areas) and then try to study all or a sample of the subjects in each cluster. Knowledge of the cluster sampling process must then be used during analysis of the study (see later) to draw inferences correctly back to the target population. Regardless of the method used to select the intended sample, the actual sample will almost always differ in important ways because not all intended subjects will be willing to enroll in the study and not all who begin a study will complete it. In a study on treatment of obesity, for example, those who consent to be in the study probably differ in important, but difficult-toquantify ways from those who do not (and may be more likely to do well with treatment). Furthermore, subjects who respond poorly to treatment

may drop out, thus making the group that completes the study even less representative. Statistical methods address only some of the issues involved in making inferences from a sample to a target population. Specifically, most statistical methods address only the effect of random variation on the inference from the intended sample to the target population. Estimating the effects of differences between the intended sample and the actual sample depends on the quantities being estimated and content knowledge about whether factors associated with being in the actual sample are related to those quantities. One rule of thumb about generalizability is that associations between variables are more often generalizable than measurements of single variables. For instance, subjects who consent to be in a study of obesity may be more motivated than average, but this motivation would be expected to have less effect on the difference in weight loss between groups than on the average weight loss in either group.


Types of Variables

A key use of statistics is to describe sample data. Methods of description depend on the type of variable (E-Table 9-1). Numerical variables include continuous variables (those that have a wide range of possible values), count variables (e.g., the number of times a woman has been pregnant), and timeto-event variables (e.g., the time from initial treatment to recurrence of breast cancer). Whereas numerical variables describe the data with numbers, categorical variables consist of named characteristics. Categorical variables can be further divided into dichotomous variables, which can take on only two possible values (e.g., alive/dead); nominal variables, which can take on more than two values but have no intrinsic ordering (e.g., race); and ordinal variables, which have more than two values and an intrinsic ordering of the values (e.g., tumor stage). Numerical variables are also ordinal by nature and can be made binary by breaking the values into two disjointed categories (e.g., systolic blood pressure >140 mm Hg or not), and thus sometimes methods designed for ordinal or binary data are used with numerical variable types, often for ease of interpretation.

Univariate Statistics for Numerical Variables: Mean, Standard Deviation, Median, and Percentiles

When describing data in a sample, it is a good idea to begin with univariate (one variable at a time) statistics. For numerical variables, univariate statistics typically measure central tendency and variability. The most common measures of central tendency are the mean (or average, i.e., the sum of the observations divided by the number of observations) and the median (the 50th percentile, i.e., the value that has equal numbers of observations above and below it). One of the most commonly used measures of variability is the standard deviation (SD). The SD is defined as the square root of the variance, which is calculated by subtracting each value in the sample from the mean, squaring that difference, totaling all of the squared differences, and dividing by the number of observations minus 1. Although this definition is far from intuitive, the SD has some useful mathematical properties, namely, that if the distribution of the variable is the familiar bell-shaped, normal, or Gaussian distribution, about 68% of the observations will be within 1 SD of the mean, about 95% within 2 SD, and about 99.7% within 3 SD. Even when the distribution is not normal, these rules are often approximately true. For variables that are not normally distributed, including most count and time-to-event variables, the mean and SD are not as useful for summarizing the data. In that case, the median may be a better measure of central tendency because it is not influenced by observations far below or far above the center. Similarly, the range and pairs of percentiles, such as the 25th and 75th percentiles or the 15th and 85th percentiles, will provide a better description of the spread of the data than the SD will. The 15th and 85th percentiles are particularly attractive because they correspond, in the Gaussian distribution, to about −1 and +1 SD from the mean, thus making reporting of the 50th, 15th, and 85th percentiles roughly equivalent to reporting the mean and SD.

Univariate Statistics for Categorical Variables: Proportions, Rates, and Ratios

For categorical variables, the main univariate statistic is the proportion of subjects with each value of the variable. For dichotomous variables, only one proportion is needed (e.g., the proportion female); for nominal variables and ordinal variables with few categories, the proportion in each group can be provided. Ordinal variables with many categories can be summarized by

CHAPTER 9  Statistical Interpretation of Data  





Categorical (dichotomous)

Alive; readmission to 2 × 2 table, the hospital chi-square within 30 days analysis

Logistic regression

Categorical (nominal)

Race; cancer, tumor type

Nominal logistic regression

Categorical (ordinal)

Glasgow Coma Scale Mann-WhitneyWilcoxon, Kruskal-Wallis

Numerical (continuous)

Cholesterol; SF-36 scales*

Chi-square analysis

t Test, analysis of variance

Ordinal logistic regression Linear regression

Numerical (count) Number of times Mann-Whitneypregnant; number Wilcoxon, of mental health Kruskal-Wallis visits in a year

Poisson regression, linear models

Time to event regression

Cox proportional hazards

Time to breast cancer; time to viral rebound in HIV-positive subjects

Log rank

*Numerical scores with many values are often treated as though they were continuous. HIV = human immunodeficiency virus; SF-36 = short-form 36-item health survey.


CHAPTER 9  Statistical Interpretation of Data  


using proportions or by using medians and percentiles, as with continuous data that are not normally distributed. It is worth distinguishing among proportions, rates, and ratios because these terms are often confused. Proportions are unitless, always between 0 and 1 inclusive, and express what fraction of the subjects have or develop a particular characteristic or outcome. Strictly speaking, rates have units of inverse time; they express the proportion of subjects in whom a particular characteristic or outcome develops over a specific time period. The term is frequently misused, however. For example, the term false-positive rate is widely used for the proportion of subjects without a disease who test positive, even though it is a proportion, not a rate. Ratios are the quotients of two numbers; they can range between zero and infinity. For example, the male-to-female ratio of people with a disease might be 3 : 1. As a rule, if a ratio can be expressed as a proportion instead (e.g., 75% male), it is more concise and easier to understand.

conventionally the risk is for something bad, and the risk in the group of interest is subtracted from the risk in a comparison group, so the ARR will be positive for effective interventions. In this case, the ARR = 0.06% per year, or 6 in 10,000 per year.

Incidence and Prevalence

Two terms commonly used (and misused) in medicine and public health are incidence and prevalence. Incidence describes the number of subjects who contract a disease over time divided by the population at risk. Incidence is usually expressed as a rate (e.g., 7 per 1000 per year), but it may sometimes be a proportion if the time variable is otherwise understood or clear, as in the lifetime incidence of breast cancer or the incidence of diabetes during pregnancy. Prevalence describes the number of subjects who have a disease at one point in time divided by the population at risk; it is always a proportion. At any point in time, the prevalence of disease depends on how many people contract it and how long it lasts: prevalence = incidence × duration.

When the treatment increases the risk for a bad outcome, the difference in risk between treated and untreated patients should still be calculated, but it is usually just called the risk difference rather than an ARR (because the “reduction” would be negative). In that case, the NNT is sometimes called the number needed to harm. This term is a bit of a misnomer. The reciprocal of the risk difference is still a number needed to treat; it is just a number needed to treat per person harmed rather than a number needed to treat per person who benefits. In the WHI, treatment with estrogens was estimated to cause about 12 additional strokes per 10,000 women per year, so the number needed to be treated for 1 year to cause a stroke was about 10,000/12, or 833.

Bivariate Statistics

Odds Ratio

Bivariate statistics summarize the relationship between two variables. In clinical research, it is often desirable to distinguish between predictor and outcome variables. Predictor variables include treatments received, demographic variables, and test results that are thought possibly to predict or cause the outcome variable, which is the disease or (generally bad) event or outcome that the test should predict or treatment prevent. For example, to see whether a bone mineral density measurement (the predictor) predicts time to vertebral fracture (the outcome), the choice of bivariate statistic to assess the association of outcome with predictor depends on the types of predictor and outcome variables being compared.

Dichotomous Predictor and Outcome Variables

A common and straightforward case is when both predictor and outcome variables are dichotomous, and the results can thus be summarized in a 2 × 2 table. Bivariate statistics are also called measures of association (E-Table 9-2).

Relative Risk

The relative risk or risk ratio (RR) is the ratio of the proportion of subjects in one group in whom the outcome develops divided by the proportion in the other group in whom it develops. It is a general (but not universal) convention to have the outcome be something bad and to have the numerator be the risk for those who have a particular factor or were exposed to an intervention. When this convention is followed, an RR greater than 1 means that exposure to the factor was (on average) bad for the study subjects (with respect to the outcome being studied), whereas an RR less than 1 means that it was good. That is, risk factors that cause diseases will have RR values greater than 1, and effective treatments will have an RR less than 1. For example, in the Women’s Health Initiative (WHI) randomized trial, conjugated equine estrogen use was associated with an increased risk for stroke (RR = 1.37) and decreased risk for hip fracture (RR = 0.61).

Relative Risk Reduction

The relative risk reduction (RRR) is 1 − RR. The RRR is generally used only for effective interventions, that is, interventions in which the RR is less than 1, so the RRR is generally greater than 0. In the aforementioned WHI example, estrogen had an RR of 0.61 for hip fracture, so the RRR would be 1 − 0.61 = 0.39, or 39%. The RRR is commonly expressed as a percentage and used only when it is positive.

Absolute Risk Reduction

The risk difference or absolute risk reduction (ARR) is the difference in risk between the groups, defined as earlier. In the WHI, the risk for hip fracture was 0.11% per year with estrogen and 0.17% per year with placebo. Again,

Number Needed to Treat

The number needed to treat (NNT) is 1/ARR. To see why this is the case, consider the WHI placebo group and imagine treating 10,000 patients for a year. All but 17 would not have had a hip fracture anyway because the fracture rate in the placebo group was 0.17% per year, and 11 subjects would sustain a fracture despite treatment because the fracture rate in the estrogen group was 0.11% per year. Thus, with treatment of 10,000 patients for a year, 17 − 11 = 6 fractures prevented, or 1 fracture prevented for each 1667 patients treated for 1 year. This calculation is equivalent to 1/0.06% per year.

Risk Difference

Another commonly used measure of association is the odds ratio (OR). The OR is the ratio of the odds of the outcome in the two groups, where the definition of the odds of an outcome is p/(1 − p), with p being the probability of the outcome. From this definition it is apparent that when p is very small, 1 − p will be close to 1, so p/(1 − p) will be close to p, and the OR will closely approximate the RR. In the WHI, the ORs for stroke (1.37) and fracture (0.61) were virtually identical to the RRs because both stroke and fracture were rare. When p is not small, however, the odds and probability will be quite different, and ORs and RRs will not be interchangeable.

Absolute versus Relative Measures

RRRs are usually more generalizable than ARRs. For example, the use of statin drugs is associated with about a 30% decrease in coronary events in a wide variety of patient populations (Chapter 206). The ARR, however, will usually vary with the baseline risk, that is, the risk for a coronary event in the absence of treatment. For high-risk men who have already had a myocardial infarction, the baseline 5-year risk might be 20%, which could be reduced to 14% with treatment, an ARR of 6%, and an NNT of about 17 for approximately 5 years. Conversely, for a 45-year-old woman with a high low-density lipoprotein cholesterol level but no history of heart disease, in whom the 5-year risk might be closer to 1%, the same RRR would give a 0.7% risk with treatment, a risk difference of 0.3%, and an NNT of 333 for 5 years. The choice of absolute versus relative measures of association depends on the intended use of the measure. As noted earlier, RRs are more useful as summary measures of effect because they are more often generalizable across a wide variety of populations. RRs are also more helpful for understanding causality. However, absolute risks are more important for questions about clinical decision making because they relate directly to the tradeoffs between risks and benefits—specifically, the NNT, as well as the costs and side effects that need to be balanced against potential benefits. RRRs are often used in advertising because they are generally more impressive than ARRs. Unfortunately, the distinction between relative and absolute risks may not be appreciated by clinicians, thereby leading to higher estimates of the potential benefits of treatments when RRs or RRRs are used.

Risk Ratios versus Odds Ratios

The choice between RRs and ORs is easier: RRs are preferred because they are easier to understand. Because ORs that are not equal to 1 are always farther from 1 than the corresponding RR, they may falsely inflate the perceived importance of a factor. ORs are, however, typically used in two circ*mstances. First, in case-control studies (Chapter 11), in which subjects with and without the disease are sampled separately, the RR cannot be

CHAPTER 9  Statistical Interpretation of Data  














Risk ratio or relative risk (RR)

a c ÷ (a + b) (c + d )

Relative risk reduction (RRR)

1 – RR

Risk difference or absolute risk reduction (ARR)

a c − (a + b) (c + d)

Number needed to treat (NNT) Odds ratio (OR)




1 ARR ad bc

*The numbers of subjects in each of the cells are represented by a, b, c, and d. Case-control studies allow calculation of only the odds ratio.



CHAPTER 9  Statistical Interpretation of Data  

calculated directly. This situation does not usually cause a problem, however, because case-control studies are generally performed to assess rare outcomes, for which the OR will closely approximate the RR. Second, in observational studies that use a type of multivariate analysis called logistic regression (see later), use of the OR is convenient because it is the parameter that is modeled in the analysis.

Dichotomous Predictor Variable, Continuous Outcome Variable

Many outcome variables are naturally continuous rather than dichotomous. For example, in a study of a new treatment of obesity, the outcome might be change in weight or BMI. For a new diuretic, the outcome might be change in blood pressure. For a palliative treatment, the outcome might be a qualityof-life score calculated from a multi-item questionnaire. Because of the many possible values for the score, it may be analyzed as a continuous variable. In these cases, dichotomizing the outcome leads to loss of information. Instead, the mean difference between the two groups is an appropriate measure of the effect size. When the outcome is itself a difference (e.g., change in blood pressure over time), the effect is measured by the difference in the withingroup differences between the groups. Most measurements have units (e.g., kg, mm Hg), so differences between groups will have the same units and be meaningless without them. If the units of measurement are familiar (e.g., kg or mm Hg), the difference between groups will be meaningful without further manipulation. For measurements in unfamiliar units, such as a score on a new quality-of-life instrument, some benchmark is useful to help judge whether the difference between groups is large or small. In that case, authors typically express the difference in relation to the spread of values in the study by calculating the standardized mean difference (SMD), which is the difference between the two means divided by the SD of the measurement. It is thus expressed as the number of SDs by which the two groups are apart. To help provide a rough feel for this difference, a 1-SD difference between means (SMD = 1) would be a 15-point difference in IQ scores, a 600-g difference in birthweight, or a 40-mg/dL difference in total cholesterol levels.

Continuous Predictor Variable

When predictor variables are continuous, the investigator can either group the values into two or more categories and calculate mean differences or SMDs between the groups as discussed earlier or use a model to summarize the degree to which changes in the predictor variable are associated with changes in the outcome variable. Use of a model may more compactly describe the effects of interest but involves assumptions about the way the predictor and outcome variables are related. Perhaps the simplest model is to assume a linear relationship between the outcome and predictor. For example, one could assume that the relationship between systolic blood pressure (mm Hg) and salt intake (g/day) was linear over the range studied: SBPi = a + (b × SALTi ) + ε i where SBPi is the systolic blood pressure for study subject i, SALTi is that subject’s salt intake, and εi is an error term that the model specifies must average out to zero across all of the subjects in the study. In this model, a is a constant, the intercept, and the strength of the relationship between the outcome and predictor can be summarized by the slope b, which has units equal to the units of SBP divided by the units of SALT, or mm Hg per gram of salt per day in this case. Note that without the units, such a number is meaningless. For example, if salt intake were measured in grams per week instead of grams per day, the slope would only be one seventh as large. Thus, when reading an article in which the association between two variables is summarized, it is critical to note the units of the variables. As discussed earlier, when units are unfamiliar, they are sometimes standardized by dividing by the SDs of one or both variables. It is important to keep in mind that use of a model to summarize a relationship between two variables may not be appropriate if the model does not fit. In the preceding example, the assumption is that salt intake and blood pressure have a linear relationship, with the slope equal to b mm Hg/g salt per day. The value of b is about 1 mm Hg/g salt per day for hypertensive patients. If the range of salt intake of interest is from 1 to 10 g/day, the model predicts that blood pressure will increase 1 mm Hg as a result of a 1-g/day increase in salt intake whether that increase is from 1 to 2 g/day or from 9 to 10 g/day. If the effect of a 1-g/day change in salt intake differed in subjects ingesting low- and high-salt diets, the model would not fit, and misleading conclusions could result.

When the outcome variable is dichotomous, the relationship between the probability of the outcome and a continuous predictor variable is often modeled with a logistic model: 1 Pr{Yi = 1} = 1 + e −(a+bxi ) where the outcome Yi is coded 0 or 1 for study subject i, and xi is that subject’s value of the predictor variable. Once again, a is a constant, in this case related to the probability of the disease when the predictor is equal to zero, and b summarizes the strength of the association; in this case, it is the natural logarithm of the OR rather than the slope. The OR is the OR per unit change in the predictor variable. For example, in a study of lung cancer, an OR of 1.06 for pack years of smoking would indicate that the odds of lung cancer increase by 6% for each pack year increase in smoking. Because the outcome variable is dichotomous, it has no units, and “standardizing” it by dividing by its SD is unnecessary and counterproductive. On the other hand, continuous predictor variables do have units, and the OR for the logistic model will be per unit change in the predictor variable or, if standardized, per SD change in the predictor variable. Re-expressing predictors in standardized or at least more sensible units is often necessary. For example, suppose 10-year mortality risk decreases by 20% (i.e., RR = 0.8) for each increase in gross income of $10,000. The RR associated with an increase in gross income of $1 (which is what a computer program would report if the predictor were entered in dollars) would be 0.99998, apparently no effect at all because a change of $1 in gross income is negligible and associated with a negligible change in risk. To derive the coefficient associated with a $1 change, the coefficient for a $10,000 change is raised to the 110 ,000 power: 0.8(1/10,000) = 0.99998.

Multivariable Statistics

In many cases, researchers are interested in the effects of multiple predictor variables on an outcome. Particularly in observational studies, in which investigators cannot assign values of a predictor variable experimentally, it will be of interest to estimate the effects of a predictor variable of interest independent of the effects of other variables. For example, in studying whether breastfeeding reduces the mother’s risk for subsequent breast cancer, investigators would try to take differences in age, race, family history, and parity into account. Trying to stratify by all these variables would require a massive data set. Instead, models are used because they enable the information about individual predictors to be summarized by using the full data set. In this way, the estimated coefficients from the model are powerful descriptive statistics that allow a sense of the data in situations in which simpler methods fail. These models are similar to those described earlier but include terms for the additional variables.

Multiple Linear Regression

The multiple linear regression model for an outcome variable Y as function or predictor variables x1, x2, and so forth is as follows: Yi = a + (b1 × x 1i ) + (b2 × x 2 i ) + …+ (bk × x ki ) + ε i , where the subscripts 1, 2, …, k are for the first, second, … kth variables of the model, and the i subscripts are for each individual. As before, the relationships between each of these predictor variables and the outcome variable are summarized by coefficients, or slopes, which have units of Y divided by the units of the associated predictor. In addition, the linear combination of predictor variables adds a major simplifying constraint (and assumption) to the model: it specifies that the effects of each variable on the outcome variable are the same regardless of the values of other variables in the model. Thus, for example, if x1 is the variable for salt intake and x2 is a variable for sex (e.g., 0 for females and 1 for males), this model assumes that the average effect of a 1-g increase in daily salt intake on blood pressure is the same in men and women. If such is not believed to be the case, either based on previous information or from examining the data, the model should include interaction terms, or separate models should be used for men and women.

Multiple Logistic Regression

The logistic model expands to include multiple variables in much the same way as the linear model: 1 Pr{Yi = 1} = − ( a+b1i +b2 x2 i +…+bk xki ) 1+ e Again, the additional assumption when more than one predictor is included in the model is that in the absence of included interaction terms,

CHAPTER 9  Statistical Interpretation of Data  

the effect of each variable on the odds of the outcome is the same regardless of the values of other variables in the model. Because the logistic model is multiplicative, however, the effects of different predictors on the odds of the outcome are multiplied, not added. Thus, for example, if male sex is associated with a doubling of the odds for heart disease, this doubling will occur in both smokers and nonsmokers; if smoking triples the odds, this tripling will be true in both men and women, so smoking men would be predicted to have 2 × 3 = 6 times higher odds of heart disease than nonsmoking women.

Recursive Partitioning

Recursive partitioning, or “classification and regression trees,” is a prediction method often used with dichotomous outcomes that avoids the assumptions of linearity. This technique creates prediction rules by repeatedly dividing the sample into subgroups, with each subdivision being formed by further separating the sample on the value of one of the predictor variables. The optimal choice of variables and cut points may depend on the relative costs of falsepositive and false-negative predictions, as set by the investigator. The end result is a set of branching questions that forms a treelike structure in which each final branch provides a yes/no prediction of the outcome. The methods of fitting the tree to data (e.g., cross-validation) help reduce overfitting (inclusion of unnecessary predictor variables), especially in cases with many potential predictors.

Proportional Hazards (Cox) Model

A multivariate model often used in studies in which subjects are monitored over time for development of the outcome is the Cox or proportional hazards model. Like the logistic model, the Cox model is used for continuous or dichotomous predictor variables, but in this case with a time-to-event outcome (e.g., time to a stroke). This approach models the rate at which the outcome occurs over time by taking into account the number of people still at risk at any given time. The coefficients in the Cox model are logarithms of hazard ratios rather than ORs, interpretable (when exponentiated) as the effect of a unit change in predictors on the hazard (risk in the next short time period) of the outcome developing. Like the logistic model, the Cox model is multiplicative; that is, it assumes that changes in risk factors multiply the hazard by a fixed amount regardless of the levels of other risk factors. A key feature of the Cox model and other survival analysis techniques is that they accommodate censored data (when the time to event is known only to exceed a certain value). For example, if the outcome is time to stroke, the study will end with many subjects who have not had a stroke, so their time to stroke is known only to exceed the time to their last follow-up visit.


The next step after describing the data is drawing inferences from a sample to the population from which the sample was drawn. Statistics mainly quantify random error, which arises by chance because even a sample randomly selected from a population may not be exactly like the population from which it was drawn. Samples that were not randomly selected from populations may be unrepresentative because of bias, and statistics cannot help with this type of systematic (nonrandom) error.

Inferences from Sample Means: Standard Deviation versus Standard Error

The simplest case of inference from a sample to a population involves estimating a population mean from a sample mean. Intuitively, the larger the sample size, N, the more likely it will be that the sample mean will be close to the population mean, that is, close to the mean that would be calculated if every member of the population were studied. The more variability there is in the population (and hence the sample), the less accurate the sample estimate of the population mean is likely to be. Thus, the precision with which a population mean can be estimated is related to both the size of the sample and the SD of the sample. To make inferences about a population mean from a sample mean, the standard error of the mean (SEM), which takes both of these factors into account, is as follows: SD SEM = N To understand the meaning of the SEM, imagine that instead of taking a single sample of N subjects from the population, many such samples were taken. The mean of each sample could be calculated, as could the mean of those sample means and the SD of these means. The SEM is the best estimate from a single sample of what that SD of sample means would be.


Confidence Intervals

The SEM expresses variability of sample means in the same way that the SD expresses variability of individual observations. Just as about 95% of observations in a population are expected to be within ±1.96 SD of the mean, 95% of sample means are expected to be within 1.96 SEM of the population mean, thereby providing the 95% confidence interval (CI), which is the range of values for the population mean consistent with what was observed from the sample. CIs can similarly be calculated for other quantities estimated from samples, including proportions, ORs, RRs, regression coefficients, and hazard ratios. In each case, they provide a range of values for the parameter in the target population consistent with what was observed in the study sample.

Significance Testing and P Values

Many papers in the medical literature include P values, but the meaning of P values is widely misunderstood and mistaught. P values start with calculation of a test statistic from the sample that has a known distribution under certain assumptions, most commonly the null hypothesis, which states that there is no association between variables. P values provide the answer to the question, “If the null hypothesis were true, what would be the probability of obtaining, by chance alone, a value of the test statistic this large or larger (suggesting an association between groups of this strength or stronger)?” When the P value is small, there are two possible explanations. First, something with a small possibility of occurring actually happened; or second, the null hypothesis is false, and there is a true association. Values of P less than 0.05 are customarily described as “statistically significant.” There are a number of common pitfalls in interpreting P values. The first is that because P values less than .05 are customarily described as being “statistically significant,” the description of results with P values less than .05 sometimes gets shortened to “significant” when in fact the results may not be clinically significant (i.e., important) at all. A lack of congruence between clinical and statistical significance most commonly arises when studies have a large sample size and the measurement is of a continuous or frequently occurring outcome. A second pitfall is concluding that no association exists simply because the P value is greater than .05. However, it is possible that a real association exists, but that it simply was not found in the study. This problem is particularly likely if the sample size is small because small studies have low power, defined as the probability of obtaining statistically significant results if there really is a given magnitude of difference between groups in the population. One approach to interpreting a study with a nonsignificant P value is to examine the power that the study had to find a difference. A better approach is to look at the 95% CI. If the 95% CI excludes all clinically significant levels of the strength of an association, the study probably had an adequate sample size to find an association if there had been one. If not, a clinically significant effect may have been missed. In “negative” studies, the use of CIs is more helpful than power analyses because CIs incorporate information from the study’s results. Finally, a common misconception about P values is that they indicate the probability that the null hypothesis is true (e.g., that there is no association between variables). Thus, it is not uncommon to hear or read that a P value less than .05 implies at least a 95% probability that the observed association is not due to chance. This statement represents a fundamental misunderstanding of P values. Calculation of P values is based on the assumption that the null hypothesis is true. The probability that an association is real depends not just on the probability of its occurrence under the null hypothesis but also on the probability of another basis for the association (see later)—an assessment that depends on information from outside the study, sometimes called the prior probability of an association (of a certain magnitude) estimated before the study results were known and requiring a different approach to statistical inference. Similarly, CIs do not take into account previous information on the probable range of the parameter being estimated. Bayesian methods, which explicitly combine prior knowledge with new information, are beginning to enter the mainstream medical literature.2 Appropriate test statistics and methods for calculating P values depend on the type of variable, just as with descriptive statistics (see E-Table 9-1). For example, to test the hypothesis that the mean values of a continuous variable are equal in two groups, a t test would be used; to compare the mean values across multiple groups, analysis of variance would be used. Because there are many different ways for the null hypothesis to be false (i.e., many different ways that two variables might be associated) and many test statistics that could be calculated, there are many different ways of calculating a P value for


CHAPTER 9  Statistical Interpretation of Data  

the association of the same two variables in a data set, and they may not all give the same answer.


Statistical techniques for inferring population values from a sample are not restricted to samples of individuals. Meta-analysis is a statistical method for drawing inferences from a sample of studies to derive a summary estimate and confidence interval for a parameter measured by the included studies, such as a risk ratio for a treatment effect.3 Meta-analysis allows the formal combination of results while estimating and accommodating both the within-study and between-study variations. Meta-analysis is most often done when raw data from the studies are not available, as is typically the case when synthesizing information from multiple published results. For example, the previously cited estimate that a 1-g/day change in salt intake is associated with a 1-mm Hg change in blood pressure was obtained from a meta-analysis of randomized trials of low-salt diets in adults.


In many cases, a goal of clinical research is not just to identify associations but also to determine whether they are causal, that is, whether the predictor causes the outcome. Thus, if people who take vitamin E live longer than those who do not, it is important to know whether it is because they took the vitamin or for some other reason. Determination of causality is based on considering alternative explanations for an association between two variables and trying to exclude or confirm these alternative explanations. The alternatives to a causal relationship between predictor and outcome variables are chance, bias, effect-cause, and confounding. P values and CIs help assess the likelihood of chance as the basis for an association. Bias occurs when systematic errors in sampling or measurements can lead to distorted estimates of an association. For example, if those making measurements of the outcome variable are not blinded to values of the predictor variable, they may measure the outcome variable differently in subjects with different values of the predictor variable, thereby distorting the association between outcome and predictor. Effect-cause is a particular problem in cross-sectional studies, in which (in contrast to longitudinal studies) all measurements are made at a single point in time, thereby precluding demonstration that the predictor variable preceded the outcome—an important part of demonstrating causality. Sometimes biology provides clear guidance about the direction of causality. For example, in a cross-sectional study relating levels of urinary cotinine (a measure of exposure to tobacco smoke) to decreases in pulmonary function, it is hard to imagine that poor pulmonary function caused people to be exposed to smoke. Conversely, sometimes inferring causality is more difficult: are people overweight because they exercise less, or do they exercise less because they are overweight (or both)?


Confounding can occur when one or more extraneous variables are associated with both the predictor of interest and the outcome. For example, observational studies suggested that high doses of vitamin E might decrease the risk for heart disease. However, this association seems to have been largely due to confounding: people who took vitamin E were different in other ways from those who did not, including differences in factors causally related to coronary heart disease. If such factors are known and can be measured accurately, one way to reduce confounding is to stratify or match on these variables. The idea is to assemble groups of people who did and did not take vitamin E but who were similar in other ways. Multivariate analysis can accomplish the same goal—other measured variables are held constant statistically, and the effect of the variable of interest (in this case the use of vitamin E) can be examined. Multivariate analysis has the advantage that it can control simultaneously for more potentially confounding variables than can be considered with stratification or matching, but it has the disadvantage that a model must be created (see earlier), and this model may not fit the data well. A new technique that is less dependent on model fit but still requires accurate measurements of confounding variables is the use of propensity scores. Propensity scores are used to assemble comparable groups in the same way as stratification or matching, but in this case the comparability is achieved on the basis of the propensity to be exposed to or be treated with the predictor variable of primary interest. Although propensity scores can adjust only for known confounders and are more subject to manipulation by investigators, systematic reviews suggest that they usually give answers that are generally similar to those of randomized trials addressing the same question.4

A major limitation of these methods of controlling for confounding is that the confounders must be known to the investigators and accurately measured. In the case of vitamin E, apparent favorable effects persisted after controlling for known confounding variables. It is for this reason that randomized trials provide the strongest evidence for causality. If the predictor variable of interest can be randomly assigned, confounding variables, both known and unknown, should be approximately equally distributed between the subjects who are and are not exposed to the predictor variable, and it is reasonable to infer that any significant differences in outcome that remain in these now comparable groups would be due to differences in the predictor variable of interest. In the case of vitamin E, a recent meta-analysis of randomized trials found no benefit and in fact suggested harm from high doses.


Missing Data

Research on human subjects is challenging. People drop out of studies, refuse to answer questions, miss study visits, and die of diseases that are not being studied directly in the protocol. Consequently, missing or incomplete data are a fact of medical research. When the particular data that are missing are unrelated to the outcome being studied (which might be true, for example, if the files storing the data got partially corrupted), analyses using only the data present (sometimes called a complete case analysis) are unlikely to be misleading. Unfortunately, such is rarely the case. Subjects refusing to divulge family income probably have atypical values, patients not coming for scheduled visits in a study of depression may be more or less depressed, and patients in an osteoporosis study who die of heart disease probably differ in many ways from those who do not. Whenever a sizable fraction of the data is missing (certainly if it is above 10 or 15%), there is the danger of substantial bias from an analysis that uses only the complete data. This is the gap noted earlier between the intended and actual samples. Any study with substantial missing data should be clear about how many missing data there were and what was done to assess or alleviate the impact; otherwise, the critical consumer of such information should be suspicious. Multiple imputation is a technique of using observations with nonmissing data to estimate missing values; it can produce less biased estimates than simply excluding the observations with missing data.5 In a randomized trial, the general rule is that the primary analysis should include all subjects who were randomized, regardless of whether they followed the study protocol, in an intention-to-treat analysis.

Clustered or Hierarchical Data

Data are often collected in a clustered (also called hierarchical) manner; for example, NHANES used a cluster sample survey, and a study of patient outcomes might be conducted at five hospitals, each with multiple admission teams. The cluster sample or the clustering of patients within teams within hospitals leads to correlated data. Said another way, and other things being equal, data collected on the same patient, by the same admission team, or in the same cluster are likely to be more similar than data from different patients, teams, or clusters. Failure to use statistical methods that accommodate correlated data can seriously misstate standard errors, widths of CIs, and P values, most often leading to overly optimistic estimates, that is, standard errors and P values that are incorrectly too small and CIs that are incorrectly too narrow. Statistical methods for dealing with correlated data include generalized estimating equations and the use of robust standard errors and frailty models (for time-to-event data). Studies with obvious hierarchical structure that fail to use such methods may be in serious error.

Multiple Hypothesis Testing

The “multiple hypothesis testing” or “multiple comparisons” issue refers to the idea that if multiple statistical tests are conducted, each at a significance level of .05, the chance that at least one of them will achieve a P value of less than .05 is considerably larger than .05, even when all the null hypotheses are true. For example, when comparing the mean value of a continuous variable across many different groups, analysis of variance is a time-tested method of performing an overall test of equality and avoiding making a large number of pairwise comparisons. Because most medical studies collect data on a large number of predictor variables, performing a test on the association of each one with the outcome may generate false-positive results. The risk for falsely positive results is especially high with genomic studies, in which a researcher may test a million single-nucleotide polymorphisms for association with a disease.

A typical method for dealing with the problem of multiple testing is the Bonferroni correction, which specifies that the P value at which the null hypothesis will be rejected (e.g., .05) should be divided by the number of hypothesis tests performed. Although simple to use, a problem with this approach is that it is overly conservative. Studies with many listed or apparent outcomes or predictors (or both) are subject to inflation of the error rate to well above the nominal .05. Automated stepwise regression methods for choosing predictors in regression models typically do not alleviate and may exacerbate this problem. If no adjustment or method for dealing with multiple comparisons is used, the high probability of false-positive results should be kept in mind. GENERAL REFERENCES For the General References and other additional features, please visit Expert Consult at

CHAPTER 9  Statistical Interpretation of Data  

GENERAL REFERENCES 1. Newman TB, Kohn MA. Evidence-Based Diagnosis. New York: Cambridge University Press; 2009. A practical text for clinicians. 2. Goodman SN. Bayesian methods for evidence evaluation: are we there yet? Circulation. 2013;127: 2367-2369.


3. Murad MH, Montori VM, Ioannidis JP, et al. How to read a systematic review and meta-analysis and apply the results to patient care: users’ guides to the medical literature. JAMA. 2014;312:171-179. 4. Lonjon G, Boutron I, Trinquart L, et al. Comparison of treatment effect estimates from prospective nonrandomized studies with propensity score analysis and randomized controlled trials of surgical procedures. Ann Surg. 2014;259:18-25. 5. Cummings P. Missing data and multiple imputation. JAMA Pediatr. 2013;167:656-661.


CHAPTER 9  Statistical Interpretation of Data  

REVIEW QUESTIONS 1. A study of 105 vegan Buddhist nuns randomly sampled from monasteries around Ho Chi Minh City found that the average femoral neck bone mineral density was 0.62 g/cm2, with a standard deviation of 0.11 g/cm2 (Ho-Pham LT, Nguyen PL, Le TT, et al. Veganism, bone mineral density, and body composition: a study in Buddhist nuns. Osteoporos Int. 2009;20:2087-2093). Which of the following statements about this result is correct? A. The 95% confidence interval for the mean bone mineral density in these women is about 0.4 to 0.84 g/cm2. B. If bone mineral density is normally distributed, we would expect about 10% of the women in the sample to have bone mineral density outside of the interval: 0.4 to 0.84 g/cm2. C. The 95% confidence interval for the mean bone mineral density in these women is about 0.60 to 0.64 g/cm2. D. Because the women were sampled randomly, there is a 95% chance that a randomly selected woman from the population would have a bone mineral density between 0.60 and 0.64 g/cm2. E. Because the women were sampled randomly, there is a 95% chance that a randomly selected woman from the sample would have a bone mineral density between 0.60 and 0.64 g/cm2. Answer: C  We would expect about 95% of observations to be within 2 standard deviations of the sample mean, leaving 5% out, so choice B is incorrect. The 95% confidence interval for a sample mean is about mean ± 2 standard SD errors of the mean (SEM), where the SEM = . In this case, the SD is N 2 0.11 g/cm , and the N is about 100, so the SEM will be about 0.11/10 = 0.01 g/cm2, and the 95% CI will be about 0.60 to 0.64 g/cm2, as indicated in choice C. Choices D and E are incorrect because the range given is too narrow: it is ± 2 SEM when it should be ± 2 SD. 2. A study of data collected through the Get with the Guidelines-Stroke Program examined the time from onset of stroke symptoms to treatment with tissue-type plasminogen activator (tPA) among 58,353 patients with acute ischemic stroke treated within 4.5 hours of the onset of symptoms (Saver JL, Fonarow GC, Smith EE, et al. Time to treatment with intravenous tissue plasminogen activator and outcome from acute ischemic stroke. JAMA. 2013;309:2480-2488). The authors reported that “faster onset-to treatment time, in 15-minute increments, was associated with … increased achievement of independent ambulation at discharge (OR, 1.04; 95% CI 1.03-1.05; P < 0.001) … .” Which of the following is a correct interpretation of these findings? A. In this study, the effect of a 15-minute reduction in onset-to treatment time was associated with a 4 % (relative) increase in the odds of independent ambulation at discharge. B. Because the odds ratio is very close to 1.0, the results are not statistically significant. C. Because the odds ratio is very close to 1.0, the results, although highly statistically significant, are not clinically significant. D. The 4% increase in odds of independent ambulation translates into a number-needed-to treat (NNT) of 25. E. None of the above is correct. Answer: A  Choice A exactly expresses the meaning of the odds ratio for this study. Choices B and C are incorrect because the proximity of the odds ratio to 1 is in this case based partly on the choice of the authors to express it per 15 minutes onset-to-treatment time. This illustrates the importance of knowing the units of the predictor variable when it is not dichotomous. If the authors had expressed the difference per hour instead of per 15 minutes, the odds ratios would have been taken to the fourth power, that is, the odds ratio for independent ambulation at discharge would have been about 1.044 = 1.17. Choice D is incorrect because estimation of the NNT requires knowing the absolute risk reduction, which was not provided in this case.

3. Assume a study of both smoking and nonsmoking mothers reports that the effect of smoking on birthweight is about a 32-g decrease in birthweight per cigarette smoked per day during pregnancy (similar to what is reported by Juarez SP, Merlo J. Revisiting the effect of maternal smoking during pregnancy on offspring birthweight: a quasi-experimental sibling analysis in Sweden. PLoS One. 2013;8:e61734.). Which of the following statements about this finding is NOT correct? A. The result is based on a model, in which the predicted birthweight is linearly related to the number of cigarettes smoked per day. B. This model predicts that the difference in birthweight between a baby whose mother did not smoke and one who smoked 5 cigarettes per day is the same as the difference between babies of mothers who smoked 20 and 25 cigarettes per day. C. The model predicts the same effect of smoking on birthweight, regardless of the mother’s age and prepregnancy weight. D. This model predicts that cutting cigarette smoking in half will lead to a 64-g expected weight increase in the baby. E. The model could include additional terms that would reflect the effect of mother’s age and prepregnancy weight. Answer: D  Choices A, B and C accurately describe characteristics of a linear model for the effect of smoking on birthweight. Choice D is not consistent with a linear model. Choice E reflects that the effects of other variables can be taken into account, while still maintaining a linear model for the effect of cigarettes smoked on birthweight. 4. A case-control study of the relationship of breast cancer and the use of COX-2 inhibitors found that the odds ratio and confidence interval relating cancer to use of baby aspirin were OR = 0.77 and 95% CI (0.42-1.41) (Harris RE, Beebe-Donk J, Alshafie GA. Reduction in the risk of human breast cancer by selective cyclooxygenase-2 [COX-2] inhibitors. BMC Cancer. 2006;6:27). The authors then stated, “Neither acetaminophen nor baby aspirin had any effect on the relative risk of breast cancer.” This is incorrect because of which of the following? A. The confidence interval crosses 1. B. The authors do not give the P value. C. The confidence interval has a lower limit of 0.42. D. Odds ratios are inappropriate for this study. E. The authors should have used a higher level of confidence. Answer: C  C is correct because the confidence interval allows for a 58% reduction (from [1 to 0.42]*100%) in the chance of breast cancer associated with the use of baby aspirin, a potentially important effect. Choice A is incorrect because it merely indicates a lack of a statistically significant result and does not bear on the size of the effect. Choice B is incorrect because CIs are more useful for ruling out important effects. Odds ratios are especially useful in case-control studies, so D is incorrect. And E is incorrect because a higher level of confidence would make the CI even wider. 5. In a randomized trial of an intervention to reduce hypertension, researchers selected subjects from a single antihypertensive patient club in Shanghai (Xue F, Yao W, Lewin RJ. A randomised trial of a 5 week, manual based, self-management programme for hypertension delivered in a cardiac patient club in Shanghai. BMC Cardiovasc Disord. 2008;8:10). About one third of those approached agreed to be randomized. The results of this study can be safely generalized to which of the following? A. All hypertensives B. All hypertensives in China C. All hypertensives in Shanghai D. All hypertensives belonging to the single club in Shanghai E. None of the above. Answer: E  Because two thirds of the participants refused to participate, the extrapolation of the effect of the intervention might not even apply to the particular club in which the study was conducted, much less a broader population.

CHAPTER 10  Using Data for Clinical Decisions  


TABLE 10-1  KEY DEFINITIONS* Probability

A number between 0 and 1 that expresses an estimate of the likelihood of an event


The ratio of [the probability of an event] to [the probability of the event’s not occurring]


Percentage of patients with disease who have an abnormal test result


Percentage of patients without disease who have a normal test result

Positive predictive value

Percentage of patients with an abnormal test result who have disease

Negative predictive value

Percentage of patients with a normal test result who do not have disease

BAYESIAN ANALYSIS Pretest (or prior) probability

10  USING DATA FOR CLINICAL DECISIONS THOMAS H. LEE Key functions in the professional lives of all physicians are the collection and analysis of clinical data. Decisions must be made on the basis of these data, including which therapeutic strategy is most appropriate for the patient and whether further information should be gathered before the best strategy can be chosen. This decision-making process is a blend of science and art in which the physician must synthesize a variety of concerns, including the patient’s most likely outcome with various management strategies, the patient’s worst possible outcome, and the patient’s preferences among these strategies. Only rarely does the physician enjoy true certainty regarding any of these issues, so a natural inclination for physicians is to seek as much information as possible before making a decision. This approach ignores the dangers inherent in the collection of information. Some of these dangers are immediate, such as the risk of cerebrovascular accident associated with coronary angiography. Some dangers are delayed, such as the risk of a malignant neoplasm due to radiation exposure from diagnostic tests. And some dangers are subtle, such as the risk of unnecessary anguish for patients due to delays, uncertainty, and confusion. An additional concern is the cost of information gathering, including the direct costs of the tests themselves and the indirect costs that flow from decisions made on the basis of the test results. Substantial data demonstrate marked variation in use of tests among physicians located in different regions and even within the same group practice. Standards of medical professionalism endorse the need for physicians to exert their influence to minimize inefficiency, but this challenge grows increasingly complex as medical progress leads to proliferation of alternative testing strategies. For the physician, there are three key questions in this sequence: Should I order a test to improve my assessment of diagnosis or prognosis? Which test is best? Which therapeutic strategy is most appropriate for this patient?


The decision of whether to order a test depends on the physician’s and the patient’s willingness to pursue a management strategy with the current degree of uncertainty.1 This decision is influenced by several factors, including the patient’s attitudes toward diagnostic and therapeutic interventions (e.g., a patient with claustrophobia might prefer an ultrasound to magnetic resonance imaging) and the information provided by the test itself. The personal tolerance of the patient and physician for uncertainty also frequently influences test-ordering approaches. A decision to watch and wait rather than to obtain a specific test also should be considered an information-gathering alternative because the information obtained while a patient is being observed often reduces uncertainty about the diagnosis and outcome. In other words,

The probability of a disease before the information is acquired

Post-test (or posterior) probability The probability of a disease after new information is acquired Pretest (or prior) odds

(Pretest probability of disease)/(1 − pretest probability of disease)

Likelihood ratio

(Probability of result in diseased persons)/ (Probability of result in nondiseased persons)

*Disease can mean a condition, such as coronary artery disease, or an outcome, such as postoperative cardiac complications.

the “test of time” should be recognized as one of the most useful tests available when this tactic does not seem inappropriately risky. Most tests do not provide a definitive answer about diagnosis or prognosis but instead reduce uncertainty. Accordingly, the impact of information from tests often is expressed as probabilities (Table 10-1). A probability of 1.0 implies that an event is certain to occur, whereas a probability of 0 implies that the event is impossible. When all the possible events for a patient are assigned probabilities, these estimates should sum to 1.0. It is often useful to use odds to quantify uncertainty instead of probability. Odds of 1 : 2 suggest that the likelihood of an event is only half the likelihood that the event will not occur, or a probability of 0.33. The relationship between odds and probability is expressed in the following formula: Odds = P /(1 − P) where P is the probability of an event.

Performance Characteristics

Sensitivity and specificity are key terms for the description of test performance. These parameters describe the test and are in theory true regardless of the population of patients to which the test is applied. Research studies that describe test performance often are based, however, on highly selected populations of patients; test performance may deteriorate when tests are applied in clinical practice. The result of a test for coronary artery disease, such as an electron beam computed tomography scan, rarely may be abnormal if it is evaluated in a low-risk population, such as high-school students. Falsepositive abnormal results secondary to coronary calcification in the absence of obstructive coronary disease are common when the test is performed in middle-aged and elderly people. Another increasingly appreciated factor that can distort the performance of screening tests is the phenomenon of “overdiagnosis,” in which the “disease” that is detected would not have led to clinical harm if it had not been found.2 Although researchers are interested in the performance of tests, the true focus of medical decision making is the patient. Physicians are more interested in the implications of a test result on the probability that a patient has a specific disease or outcome, that is, the predictive values of abnormal or normal test results. These predictive values are extremely sensitive to the population from which they are derived (Table 10-2; see also Table 10-1). An abnormal lung scan result in an asymptomatic patient has a much lower positive predictive value than that same test result in a patient with dyspnea and diminished oxygen saturation. Bayes theorem (see later) provides a


CHAPTER 10  Using Data for Clinical Decisions  


Question: What is the probability of coronary disease for a patient with a 50% pretest probability of coronary disease who undergoes an exercise test if that patient develops (a) no ST segment changes, (b) 1 mm of ST segment depression, or (c) 2 mm of ST segment depression? Step 1. Calculate the pretest odds of disease: P /(1 − P) = 0.5/(1 − 0.5) = 0.5/0.5 =1

Post-test probability


Step 2. Calculate the likelihood ratios for the various test results, using the formula LR = sensitivity/(1 − specificity). (Data from pooled literature.) TEST RESULT No ST segment changes 1-mm ST segment depression 2-mm ST segment depression




0.34 0.66 0.33

0.15 0.85 0.97

0.4 4.4 11

Step 3. Calculate the post-test odds of disease and convert those odds to post-test probabilities. TEST RESULT No ST segment changes 1-mm ST segment depression 2-mm ST segment depression

















framework for analyzing the interaction between test results and a patient’s pretest probability of a disease. As useful as the performance characteristics may be, they are limited by the fact that few tests truly provide dichotomous (i.e., positive or negative) results. Tests such as exercise tests have several parameters (e.g., ST segment deviation, exercise duration, hemodynamic response) that provide insight into the patient’s condition, and the normal range for many blood tests (e.g., a serum troponin level) varies markedly according one’s willingness to “miss” patients with disease. Tests that require human interpretation (e.g., radiologic studies) are particularly subject to variability in the reported results.

Bayes Theorem

The impact of a test result on a patient’s probability of disease was first quantified by Bayes, an 18th century English clergyman who developed a formula that describes the probability of disease in the presence of an abnormal test result. The classic presentation of Bayes theorem is complex and difficult to use. A more simple form of this theorem is known as the odds ratio form, which describes the impact of a test result on the pretest odds (see Table 10-1) of a diagnosis or outcome for a specific patient. To calculate the post-test odds of disease, the pretest odds are multiplied by the likelihood ratio (LR) for a specific test result. The mathematical presentation of this form of Bayes theorem is as follows: Post-test odds = (Pretest odds) × (LR ) The LR is the probability of a particular test result in patients with the disease divided by the probability of that same test result in patients without disease. In other words, the LR is the test result’s sensitivity divided by the false-positive rate. A test of no value (e.g., flipping a coin and calling “heads” an abnormal result) would have an LR of 1.0 because half of patients with disease would have abnormal test results, as would half of patients without disease. This test would have no impact on a patient’s odds of disease. The further an LR is above 1.0, the more that test result raises a patient’s probability of disease. For LRs less than 1.0, the closer the LR is to 0, the more it lowers a patient’s probability of disease. When it is displayed graphically (Fig. 10-1), a test of no value (dotted line) does not change the pretest probability, whereas an abnormal or normal result from a useful test moves the probability up or down. For a patient with a high pretest probability of disease, an abnormal test result changes the

0 0

Pre-test probability


FIGURE 10-1.  Impact of various test results on the patient’s probability of disease. The x-axis depicts a patient’s probability of disease before a test. If the test is of no value, the post-test probability (dotted line) is no different from the pretest probability. An abnormal test result raises the post-test probability of disease, as depicted by the concave downward arc, whereas a normal test result lowers the probability.

patient’s probability only slightly, but a normal test result leads to a marked reduction in the probability of disease. Similarly, for a patient with a low pretest probability of disease, a normal test result has little impact, but an abnormal test result markedly raises the probability of disease. Consider how various exercise test results influence a patient’s probability of coronary disease (see Table 10-2). For a patient whose clinical history, physical examination, and electrocardiographic findings suggest a 50% probability of disease, the pretest odds of disease are 1.0. LRs for various test results are developed by pooling data from published literature. The sensitivity of an exercise test with any amount of ST segment changes is the rate of such test results in patients with coronary disease, and the specificity is the percentage of patients without coronary disease who do not have this test result. The LR for no ST change is less than 1, whereas the LRs for patients with ST changes are greater than 1 (see Table 10-2). Therefore, when the LRs for various test results are multiplied by the pretest odds to calculate post-test odds, the odds decrease for patients without ST segment changes but increase for patients with 1 or 2 mm of ST segment change. Post-test odds can be converted to post-test probabilities according to the following formula: Probability = Odds/(1 + odds) The calculations quantify how the absence of ST segment changes reduces a patient’s probability of disease, whereas ST segment depression raises the probability of disease. This form of Bayes theorem is useful for showing how the post-test probability of disease is influenced by the patient’s pretest probability of disease. If a patient’s clinical data suggest a probability of coronary disease of only 0.1, the pretest odds of disease would be only 0.11. For such a low-risk patient, an exercise test with no ST segment changes would lead to post-test probability of coronary disease of 4%, whereas 1-mm or 2-mm ST segment changes would lead to a post-test probability of disease of 33% or 55%, respectively. Even if clinicians rarely perform the calculations that are described in Bayes theorem, there are important lessons from this theorem that are relevant to principles of test ordering (Table 10-3). The most crucial of these lessons is that the interpretation of test results must incorporate information about the patient. An abnormal test result in a low-risk patient may not be a true indicator of disease. Similarly, a normal test result in a high-risk patient should not be taken as evidence that disease is not present. Figure 10-2 provides an example of the post-test probabilities for positive and negative results for a test with a sensitivity of 85% and a specificity of 90% (e.g., radionuclide scintigraphy for diagnosis of coronary artery disease). In a high-risk population with a 90% prevalence of disease, the positive predictive value of an abnormal result is 0.99 compared with 0.31 for the same test result obtained in a low-risk population with a 5% prevalence of disease. Similarly, the negative predictive value of a normal test result is greater in the low-risk population than in the high-risk population.

CHAPTER 10  Using Data for Clinical Decisions  

TABLE 10-3  PRINCIPLES OF TEST ORDERING AND INTERPRETATION The interpretation of test results depends on what is already known about the patient. No test is perfect; clinicians should be familiar with its diagnostic performance (see Table 10-1) and never believe that a test “forces” them to pursue a specific management strategy. Tests should be ordered if they may provide additional information beyond that already available. Tests should be ordered if there is a reasonable chance that the data will influence the patient’s care. Two tests that provide similar information should not be ordered. In choosing between two tests that provide similar data, use the test that has lower costs or causes less discomfort and inconvenience to the patient. Clinicians should seek all of the information provided by a test, not just an abnormal or normal result. The cost-effectiveness of strategies using noninvasive tests should be considered in a manner similar to that of therapeutic strategies.

1000 Patients

900 with disease

765 truepositive results


100 without disease

135 falsenegative results

Test with: Sensitivity = 85% Specificity = 90%

10 falsepositive results

90 truenegative results

pathophysiology, such as ventilation-perfusion scintigraphy and pulmonary angiography. Regardless of whether tests are independent, the performance of multiple tests increases the likelihood that an abnormal test result will be obtained in a patient without disease. If a chemistry battery includes 20 tests and the normal range for each test has been developed to include 95% of healthy individuals, the chance that a healthy patient will have a normal result for any specific test is 0.95. However, the probability that all 20 tests will be normal is (0.95)20, or 0.36. Most healthy people can be expected to have at least one abnormal result. Unless screening test profiles are used thoughtfully, falsepositive results can subject patients to unnecessary tests and procedures.

Threshold Approach to Decision Making

Even if a test provides information, that information may not change management for an individual patient. Lumbar spine radiographs of a patient who is not willing to undergo surgery may reveal the severity of disease but expose the patient to needless radiation. Similarly, a test that merely confirms a diagnosis that already is recognized is a waste of resources (see Table 10-3). Before ordering a test, clinicians should consider whether that test result could change the choice of management strategies. This approach is called the threshold approach to medical decision making, and it requires the physician to be able to estimate the threshold probability at which one strategy will be chosen over another. The management of a clinically stable patient with a high probability of coronary disease might not be changed by any of the posttest probabilities shown in Table 10-2. If that patient had no ST segment changes, the post-test probability of 0.29 still would be too high for a clinician to consider that patient free of disease. An abnormal test result that strengthened the diagnosis of coronary disease might not change management unless it suggested a greater severity of disease that might warrant another management strategy.

Testing for Peace of Mind Positive predictive value: 765/775 = 0.99 Negative predictive value: 90/225 = 0.40

1000 Patients

950 without disease

50 with disease

Physicians frequently order tests even when there is little chance that the outcomes will provide qualitatively new insights into a patient’s diagnosis or prognosis or alter a patient’s management. In such cases, the cited goal for testing may be to improve a patient’s peace of mind. Although a decrease in uncertainty can improve quality of life for many patients, individuals with hypochondriasis and somatization disorders rarely obtain comfort from normal test results; instead, their complaints shift to a new organ system, and their demands focus on other tests. For such patients, management strategies using frequent visits and cognitive tactics are recommended.


42.5 truepositive results


7.5 falsenegative results

Test with: Sensitivity = 85% Specificity = 90%

95 falsepositive results

855 truenegative results

Positive predictive value: 42.5 /137.5 = 0.31 Negative predictive value: 855/862.5 = 0.99

FIGURE 10-2.  Interpretation of test results in high-risk and low-risk patients. A,

High-risk population (90% prevalence of disease). B, Low-risk population (5% prevalence of disease).

Multiple Testing


Clinicians frequently obtain more than one test aimed at addressing the same issue and at times are confronted with conflicting results. If these tests are truly independent (i.e., the tests do not have the same basis in pathophysiology), it may be appropriate to use the post-test probability obtained through performance of one test as the pretest probability for the analysis of the impact of the second test result. If the tests are not independent, this strategy for interpretation of serial test results can be misleading. Suppose a patient with chronic obstructive pulmonary disease and a history vaguely suggestive of pulmonary embolism is found to have an abnormal lung ventilation-perfusion scan. Obtaining that same test result over and over would not raise that patient’s probability of pulmonary embolism further and further. In this extreme case, the tests are identical; serial testing adds no information. More commonly, clinicians are faced with results from tests with related but not identical bases in

If the clinician decides that more information is needed to reduce uncertainty, and if it appears possible that tests might lead to a change in management strategies, the question arises as to which test is most appropriate. Note that just because guideline development committees have concluded that a specific test is “appropriate” in a given clinical context, it does not mean that this test is the most appropriate option. Several factors influence the choice among diagnostic strategies, including the patient’s preferences, the costs and risks associated with the tests, and the diagnostic performance of alternative tests. Diagnostic performance of a test often is summarized in terms of sensitivity and specificity,3 but as shown in the example in Table 10-2, these parameters depend on which threshold (e.g., 1 mm vs. 2 mm of ST segment change) is used. A low threshold for calling a test result abnormal might lead to excellent sensitivity for detecting disease but at the expense of a high false-positive rate. Conversely, a threshold that led to few false-positive results might cause a clinician to miss many cases of true disease. The receiver operating characteristic (ROC) curve is a graphic form of describing this tradeoff and providing a method for comparing test performance (Fig. 10-3). Each point on the ROC curve describes the sensitivity and the false-positive rate for a different threshold for abnormality for a test. A test of no value would lead to an ROC curve with the course of the dotted line, whereas a misleading test would be described by a curve that was concave upward (not shown). The more accurate the test, the closer its ROC curve comes to the upper left corner of the graph, which would indicate a test threshold that has excellent sensitivity and a low false-positive rate. The closer an ROC curve comes to the upper left corner, the greater the area under the curve. The area under ROC curves can be used to compare the information provided by two tests.


CHAPTER 10  Using Data for Clinical Decisions  




Frame the question. Create the decision tree. Identify the alternative strategies. List the possible outcomes for each of the alternative strategies. Describe the sequence of events as a series of decision nodes and chance nodes. Choose a time horizon for the analysis. Determine the probability for each chance outcome. Assign a value to each outcome. Calculate the expected utility for each strategy. Perform sensitivity analysis.

0 0

False-positive rate


FIGURE 10-3.  Receiver operating characteristic curve. The points on the curve reflect the sensitivity and false-positive (1 − specificity) rates of a test at various thresholds. As the threshold is changed to yield greater sensitivity for detecting the outcome of interest, the false-positive rate rises. The better the test, the closer the curve comes to the upper left corner. A test of no value (e.g., flipping a coin) would lead to a curve with the course of the dotted line. The area under the curve is used often to compare alternative testing strategies.

Even if one test is superior to another as shown by a greater area under its ROC curve, the question still remains as to what value of that test should be considered abnormal. The choice of threshold depends on the purpose of testing and on the consequences of a false-positive or false-negative diagnosis. If the goal is to screen the population for a disease that is potentially fatal and potentially curable, a threshold with excellent sensitivity is appropriate even if it leads to frequent false-positive results. In contrast, if a test is used to confirm a diagnosis that is likely to be treated with a high-risk invasive procedure, a threshold with high specificity is preferred. Only 1 mm of ST segment depression might be the appropriate threshold when exercise electrocardiography is used to evaluate the possibility of coronary disease in a patient with chest pain. If the question is whether to perform coronary angiography in a patient with stable angina in search of severe coronary disease that might benefit from revascularization, a threshold of 2 mm or more would be more appropriate.


Physicians and patients ultimately must use clinical information to make decisions. These choices usually are made after consideration of a variety of factors, including information from the clinical evaluation, patients’ preferences, and expected outcomes with various management strategies. Insight into the impact of these considerations can be improved through the performance of decision analysis (Table 10-4). The first step in a decision analysis is to define the problem clearly; this step often requires writing out a statement of the issue so that it can be scrutinized for any ambiguity. After the problem is defined, the next step is to define the alternative strategies. Consider the question of which test is most appropriate to screen patients for breast cancer: mammography with or without breast magnetic resonance imaging—a technology that is highly sensitive for detecting breast cancer but is more costly and less specific. The expected outcomes for these strategies depend on each test’s sensitivity and specificity for detecting breast cancer, which is influenced in turn by other factors, such as the frequency with which the test is performed. Patients’ outcomes also are influenced by their underlying risk for breast cancer and the likelihood that earlier detection of tumors reduces the risk for death. Each of these variables must be known or estimated for calculations to be made of each strategy’s predicted life expectancy and direct medical costs. These outcomes differ for patients according to age, medical history, family history, and presence or absence of genetic markers such as BRCA mutations. Optimal strategies for an elderly patient with a short life expectancy and low clinical risk of cancer are unlikely to be the same as those for a younger patient with inherited mutations of the BRCA1 or BRCA2 gene, indicating a cumulative lifetime risk of breast cancer of 50 to 85% (Chapter 198).

The credibility of the decision analysis depends on the credibility of these estimates. Published reports often do not provide information on the outcomes of interest for specific subsets of patients, or there may not have been sufficient statistical power within subsets of patients for the findings to be statistically significant. Randomized trial data are relevant to the populations included in the trial; the extension of the findings to other genders, races, and age groups requires assumptions by individuals performing the analysis. For many issues, expert opinion must be used to derive a reasonable estimate of the outcome. For many diseases, the potential outcomes are more complex than perfect health or death. With chronic diseases, patients may live many years in a condition somewhere between these two, and the goal of medical interventions may be to improve quality of life rather than to extend survival. The value of life in imperfect health must be reflected in decision analyses. These values by convention are expressed on a scale of 0 to 100, where 0 indicates the worst outcome and 100 indicates the best outcome. Life-expectancy and quality-of-life estimates are combined in many decision analyses to calculate quality-adjusted life years. A strategy that leads to a 10-year life expectancy with such severe disability that utility of the state of health is only half that of perfect health would have a quality-adjusted life expectancy of 5 years. With such adjustments to life-expectancy data, the impact of interventions that improve quality of life but do not extend life can be compared with interventions that extend life but do not improve its quality or perhaps even worsen it.4 After the value and the probability of the various outcomes have been estimated, the expected utility of each strategy can be calculated. In comparing the different strategies available at a decision node, the analysis generally selects the option with the highest expected utility. At chance nodes, the expected utility is the weighted average of the utility of the various possible branches. After the analysis has been performed with the baseline assumptions, sensitivity analyses should be performed in which these assumptions are varied over a reasonable range. These analyses can reveal which assumptions have the most influence over the conclusions and identify threshold probabilities at which the conclusions would change. For example, the threshold at which breast magnetic resonance imaging should be added to mammography is likely to be influenced by the cost of the magnetic resonance imaging and the accuracy of the radiologists who interpret the images.

Cost-Benefit and Cost-Effectiveness Analyses

For clinicians and health care policymakers, the choices that must be addressed go beyond the choices within any single decision analysis. Because resources available for health care are limited, policymakers may have to choose among many competing options for “investments” in health. Although such decisions frequently are made on the basis of political considerations, cost-benefit and cost-effectiveness analyses can be informative in making the choices. The methodology of these techniques is similar to that of decision analysis except that costs for the various possible outcomes and strategies also are calculated. Discounting is used to adjust the value of future benefits and costs because resources saved or spent currently are worth more than resources saved or expended in the future. In cost-benefit analyses, all benefits are expressed in terms of economic impact. Extensions in life expectancy are translated into dollars by estimating societal worth or economic productivity. Because of the ethical discomfort associated with expressing health benefits in financial terms, cost-effectiveness analyses are used more commonly



Treating rheumatoid arthritis with drugs that slow disease progression

Saves money and improves health

Using warfarin for 70-year-olds with atrial fibrillation

$3,000 per QALY

Daily dialysis for 60-year old critically ill men with kidney injury

$6,000 per QALY

Using an implantable cardioverterdefibrillator to prevent sudden cardiac death in high-risk patients

$38,000 per QALY

Treating spinal stenosis and leg pain with spine surgery

$90,000 per QALY

Screening 60-year-old heavy smokers with $140,000 per QALY annual CT scans Annual HIV screening for people with a low to moderate risk

Increases costs and makes health worse

Modified from the CEA Registry.6

than cost-benefit analyses. In these analyses, the ratio of costs to health benefits is calculated; one frequently used method for evaluating a strategy is calculation of cost per quality-adjusted life year.5 These estimates can be used to identify strategies that are both cost-saving and health-improving, to compare strategies by which the health care system can “purchase” additional quality-adjusted life years, and even to caution about strategies that increase costs while worsening health (Table 10-5).6 Cost-effectiveness analyses can provide important insights into the relative attractiveness of different management strategies and can help guide policymakers in decisions about which technologies to make available on a routine basis. No medical intervention can have an attractive cost-effectiveness if its effectiveness has not been proved. The cost-effectiveness of an intervention depends heavily on the population of patients in which it is applied. An inexpensive intervention would have a poor cost-effectiveness ratio if it were used in a low-risk population unlikely to benefit from it. In contrast, an expensive technology can have an attractive cost-effectiveness ratio if it is used in patients with a high probability of benefiting from it. Table 10-5 shows cost-effectiveness estimates from published literature for some selected medical interventions. Such estimates should be used only with understanding of the population for which they are relevant. GENERAL REFERENCES For the General References and other additional features, please visit Expert Consult at

CHAPTER 10  Using Data for Clinical Decisions  

GENERAL REFERENCES 1. Laine C. High-testing begins with a few simple questions. Ann Intern Med. 2012;156:162-163. 2. Etzioni R, Gulati R, Mallinger L, et al. Influence of study features and methods on overdiagnosis estimates in breast and prostate cancer screening. Ann Intern Med. 2013;158:831-838. 3. Otero HJ, Fang CH, Sekar M, et al. Accuracy, risk and the intrinsic value of diagnostic imaging: a review of the cost-utility literature. Acad Radiol. 2012;19:599-606.


4. Heijnsdijk EAM, Wever EM, Auvinen A, et al. Quality-of-life effects of prostate-specific antigen screening. N Engl J Med. 2012;367:595-605. 5. Ryen L, Svensson M. The willingness to pay for a quality adjusted life year: a review of the empirical literature. Health Econ. 2014. [Epub ahead of print]. 6. Cost-Effectiveness Analysis Registry. Accessed February 10, 2015.


CHAPTER 10  Using Data for Clinical Decisions  

REVIEW QUESTIONS 1. A 53-year-old man presents to his primary care physician with a chief complaint of chest pain for the past 2 weeks. The pain is described as intermittent, usually exertional, midline, and aching in nature. His electrocardiogram is completely normal. Which one of the following is the most appropriate next step? A. Watch and wait B. Exercise electrocardiography C. B-type natriuretic peptide D. Exercise nuclear scintigraphy E. Coronary angiography Answer: B The immediate question is whether this patient’s symptoms represent new ischemic heart disease. His clinical presentation suggests a moderate probability of coronary artery disease, so watch-and-wait is probably too risky a strategy. At the same time, he does not appear to have unstable ischemic disease, so there is no need to proceed immediately to coronary angiography in preparation for coronary revascularization. Measurement of B-type natriuretic peptide would not alter management. Of the two noninvasive tests for ischemic heart disease, exercise electrocardiography is the least expensive, is the most convenient, and carries no radiation exposure. Guidelines would thus suggest that answer B is the most appropriate next step. 2. A patient undergoes an exercise test and has 2 mm of ST-segment depression on electrocardiogram. In which one of these patients would this finding be most likely to change management? A. A healthy 19-year-old volunteer in a research study B. A 62-year-old woman who is completely asymptomatic C. A 62-year-old woman with frequent nonexertional aching chest pain D. A 62-year-old woman with recent myocardial infarction and chest pain at rest E. A 62-year-old woman who is completely symptom free 4 months after coronary artery bypass graft surgery

Answer: C Patients A and B have a sufficiently low probability of coronary disease that the exercise test abnormality is highly likely to be a false-positive result. Patient D has an extremely high probability of coronary disease and likely needs coronary angiography and revascularization as next steps; she does not need exercise electrocardiography because the test is unlikely to change this management plan. Patient E’s care is also not likely to be influenced by an exercise test result because she is asymptomatic after major coronary revascularization surgery. Patient C has a low to moderate probability of coronary disease, and this abnormal exercise test result moves her into a mid-range probability. Thus, she is likely to undergo either further testing, initiation of antianginal therapy, or both as a result of this exercise test result. 3. Which one of the following is NOT an important consideration when weighing whether to order a test for a patient? A. Test may influence decision making for patient’s care B. Test is less expensive than alternative strategies C. Test is safer than alternative strategies D. Test reduces patient’s or clinician’s uncertainty E. Test is expected to be abnormal in presence of patient’s already confirmed diagnosis Answer: E Tests should be ordered when they are expected to change care and should be chosen on basis of safety, cost, and impact. They should not be ordered simply because they can confirm an already known diagnosis.

CHAPTER 11  Measuring Health and Health Care  


professionals and accredited facilities is sufficient to ensure consistent highquality care. A second major trend is attributable to the successes of biomedical science: the major challenge in health care today is the management of chronic disease for a population with increased life expectancy. For chronic conditions, health benefits are increasingly measured in improvements in functional status or quality of life, rather than simply using mortality rates or life expectancy. A third trend relates directly to how the increasing costs of health care are now threatening public budgets and investments in other social goals, such as education. Although the United States spends more per capita on health care than any other developed nation (Chapter 5), the outcomes achieved lag far behind. Finally, advances in communication and information technologies have inspired more people to play an active role in their health and health care. These innovations have accelerated demands for transparency and shared decision making. As health insurance and health care regulation have expanded, requirements to track and justify health care services have grown. Intensifying urgency to improve the quality of health care, reduce disparities, control costs, and enhance transparency will likely lead patients and insurers to demand more data and to link quality measures to payments for services. Fortunately, modern technology can help to meet the demand for data. Patients can record and submit their health parameters using hand-held devices connected to personal health records. Automated billing programs can track health care services, while electronic health records can assess the quality of physician care. Ultimately, fully integrated health information systems will allow patient information to be retrieved instantly and seamlessly wherever and whenever it is needed. In addition to assessing care quality today, these tools offer enormous promise for learning as a byproduct of care delivery.


11  MEASURING HEALTH AND HEALTH CARE CAROLYN M. CLANCY AND ERNEST MOY Physicians routinely quantify a variety of health measures, including symptoms, vital signs, and findings on physical examination, to improve diagnosis, treatment, and prognostication. Similarly, the efficacy and quality of health care also can and should be measured for several reasons. First, the quality of care delivered is often suboptimal.1 Persistent variations in practice for patients with the same diagnosis reflect a combination of clinical uncertainty, individualized practice styles, patients’ preferences and characteristics (age, race, ethnicity, education, income), and other factors. Both suboptimal care and varied care for the same condition undermine the historical assumption that a combination of highly trained health

Three types of measures typically assess health and health care. Measures of health quantify the sickness or well-being of a person. Measures of health care quality quantify the extent to which a patient receives needed care and does not receive unnecessary care. Health care quality is assessed using measures of structure (e.g., education and credentialing of clinicians), process (adherence to professional standards and evidence-based recommendations), and outcomes (or end results of care, including how patients experience their care and their self-reported health and function). Measures of health care resources quantify the resources used (e.g., radiographs, surgery, medication, intensive care) to improve the health of a patient.2 All measures can be summed up across populations within a practice or community (Table 11-1). Measures of health and health care often overlap (E-Fig. 11-1). Health measures that can be improved by health care, such as blood pressure or blood glucose levels, are often used as health care quality outcome measures. The delivery of quality health care requires the use of resources and the generation of direct health care costs, which may or may not improve health care at the margins of spending. Impaired health that reduces the ability to do work and earn wages but that could have been prevented by the delivery of health care contributes to the indirect costs of health care. At the intersection of health, health care quality, and health care resources are measures of health care value. These measures compare the health benefits of specific health care services with their costs.2


Most researchers, provider groups, insurers, regulators, and credentialing organizations that develop measures consult with physicians to ensure that their metrics are consistent with professional standards. Insurers may create measures to allocate health care resources, to plan for future needs, and to identify efficient physicians for inclusion on panels or to be rewarded with performance bonuses. Regulators may develop measures to establish licensure requirements and to identify physicians who might benefit from remedial instruction. Credentialing organizations may construct measures to demonstrate the superior performance of physicians who meet their high standards. For example, the National Committee for Quality Assurance maintains the Healthcare Effectiveness Data and Information Set that is widely used to accredit health plans, and the American Board of Internal Medicine and other specialty boards include measures of practice performance as well as measures of medical knowledge for the maintenance of certification.

CHAPTER 11  Measuring Health and Health Care  


Health care outcomes

Indirect cost Health care value

Health care quality

Direct cost

Health care resources

E-FIGURE 11-1.  Type of measures of health and health care.



CHAPTER 11  Measuring Health and Health Care  

TABLE 11-1  MEASURES OF HEALTH AND HEALTH CARE MEASURES OF HEALTH • Mortality: rates of death typically adjusted for age and sex • Morbidity: incidence and prevalence rates of diseases and their sequelae • Functional status: assessments of a patient’s ability to perform various actions such as activities of daily living or instrumental activities of daily living as observed by a provider or reported by the patient • Self-reported health status: a patients’ assessment of their health and well-being MEASURES OF HEALTH CARE QUALITY • Health care outcomes: the end results or health benefits derived from good health care or the health loss attributable to poor health care • Health care processes: assessments of whether the right care was delivered at the right time and in the right way • Health care infrastructure: the availability of resources needed to deliver good health care • Patient perceptions of health care: a patient’s assessment of health care received, usually emphasizing patient-provider communication and shared decision making • Access to health care: the ability of patients to gain entry into health care and navigate to needed resources MEASURES OF HEALTH CARE RESOURCES • Health care utilization: the quantity of health care services that are used • Direct costs: the costs of providers, supplies, and equipment needed to deliver health care • Indirect costs: the costs of lost wages and decreased productivity due to illness or injury that could have been prevented by appropriate health care • Nonmedical costs: the costs of health care not related to the delivery of services, such as administration, advertising, research, and profits earned by health industries TYPE OF MEASURE


HEALTH Mortality

Deaths due to colorectal cancer per 100,000 population


New AIDS cases per 100,000 population

Functional status

% of people unable to perform one or more activities of daily living

Self-reported health status

% of people reporting that their overall health is excellent

HEALTH CARE QUALITY Health care outcomes

Death per 1000 hospitalizations with pneumonia

Intermediate outcomes

% of adults with diabetes whose blood pressure is 5 L), whereas glimepiride has a relatively small VD (0.18 L). As discussed later, VD is a useful pharmaco*kinetic tool for calculating the loading dose and appreciating how various changes can affect a drug’s half-life.



Absorption refers to the transfer of a drug from the site of administration to the systemic circulation. Many drugs cross a membrane barrier by passive diffusion and enter the systemic circulation. Because passive diffusion in this setting depends on the concentration of the solute at the membrane surface, the rate of drug absorption is affected by the concentration of free drug at the absorbing surface. Factors that influence the availability of free drug thus affect drug absorption from the administration site; this effect can be exploited to design medications that release a drug slowly into the circulation by prolonging drug absorption. With certain sustained-released oral prepa­ rations, the rate of dissolution of the drug in the gastrointestinal tract determines the rate at which the drug is absorbed (e.g., timed-release antihistamines). Similarly, a prolonged drug effect can be obtained by the use of transdermal medications (e.g., nitroglycerin) or intramuscular depot preparations (e.g., benzathine penicillin G).

First-Pass Effect

Some drugs that are administered orally are absorbed relatively efficiently into the portal circulation but are metabolized by the liver before they reach the systemic circulation. Because of this “first-pass” or “presystemic” effect, the oral route may be less suitable than other routes of administration for such drugs. A good example is nitroglycerin, which is well absorbed but efficiently metabolized during the first pass through the liver. However, the same drug can achieve adequate systemic levels when it is given sublingually or transdermally.


The extent of absorption of a drug into the systemic circulation may be incomplete. The bioavailability of a particular drug is the fraction (F) of the total drug dose that ultimately reaches the systemic circulation from the site of administration. This fraction is calculated by dividing the amount of the drug dose that reaches the circulation from the administration site by the amount of the drug dose that would enter the systemic circulation after direct intravenous injection into the circulation (essentially the total dose). Bioavailability, or F, can range from 0, in which no drug reaches the systemic

Drugs are removed from the body by two major mechanisms: hepatic elimination, in which drugs are metabolized in the liver and excreted through the biliary tract; and renal elimination, in which drugs are removed from the circulation by either glomerular filtration or tubular secretion. For most drugs, the rates of hepatic and renal elimination are proportional to the plasma concentration of the drug. This relationship is often described as a “first-order” process. Two measurements, clearance and half-life, are used to evaluate elimination.


The efficiency of elimination can be assessed by quantifying how fast the drug is cleared from the circulation. Drug clearance is a measure of the volume of plasma cleared of drug per unit of time. It is similar to the clinical measurement used to assess renal function—creatinine clearance, which is the volume of plasma from which creatinine is removed per minute. Total drug clearance (Cltot) is the rate of elimination by all processes (Eltot) divided by the plasma concentration of the drug (Cp):

Cl tot = El tot /C p


Drugs may be cleared by several organs, but as noted earlier, renal clearance and hepatic clearance are the two major mechanisms. Total drug clearance (Cltot) can best be described as the sum of clearances by each organ. For most drugs, this is essentially the sum of renal clearance and hepatic clearance:

Cl tot = El Ren + Cl Hep


Table 29-1 shows the wide variation in clearance values among commonly used medications; some drugs (e.g., phenobarbital) have relatively low clearances (500 mL/minute). Tobramycin is cleared almost entirely by the kidneys, whereas aspirin, carbamazepine, and phenytoin are cleared less than 5% by the kidneys. Drug clearance is affected by several factors, including blood flow through the organ of clearance, protein binding to the drug, and activity of the clearance processes in the organs of elimination (e.g., glomerular filtration rate and

CHAPTER 29  Principles of Drug Therapy  


Site of drug administration Absorption Distribution

Bound drug

Distribution Systemic circulation Free drug

Sites of action and/or toxicity

Effect and/or toxicity

Metabolism Metabolites Distribution Excretion

Urine, bile, feces, and other routes Pharmaco*kinetics


FIGURE 29-1.  Schematic of a drug’s movement through the body, from the site of administration to production of a drug effect. The relationship between pharmaco*kinetics and

pharmacodynamics is shown.

The time needed to eliminate the drug is best described by its half-life (t1/2), which is the time required during the elimination phase (see Fig. 29-2) for the plasma concentration of the drug to be decreased by half. Mathematically, the half-life is equal to the natural logarithm of 2 (representing a reduction of drug concentration to half) divided by Ke. Substituting for Ke from Equation 4 and calculating the natural logarithm of 2, the half-life can be represented by the following equation:

Log concentration of drug X in plasma (mg/L)


Distribution phase 10 Cp0


Elimination phase



0.1 0




t 1 2 = 0.693 VD/Cl

1/2 C a

120 150 180 210 240 270 300

Time after dosing (min) FIGURE 29-2.  Representative drug concentration versus time plot used in pharmaco*kinetic studies. Concentration of drug is plotted with a logarithmic scale on the ordinate, and time is plotted with a linear scale on the abscissa. The resultant curve has two phases: the distribution phase, which is the initial portion of the plotted line when the concentration of drug decreases rapidly; and the later elimination phase, during which there is an exponential disappearance of drug from the plasma over time. The dotted line extrapolated from the elimination phase back to time zero is used to calculate plasma concentration at time zero (Cp0). During the elimination phase, the half-life (t½) can be calculated as the time it takes to decrease the concentration by half (shown here as the time needed to decrease from concentration Ca to ½ Ca).


From this equation, one can predict that at a given clearance, as the VD increases, the half-life increases. Similarly, at a given VD, as the clearance increases, the half-life decreases. Clinically, many disease states (see later) can affect VD and clearance. Because disease affects VD and clearance differently, the half-life may increase, decrease, or not change much at all. Therefore, the half-life by itself is not a good indicator of the extent of abnormality in elimination. The half-life is useful to predict how long it takes for a drug to be eliminated from the body. For any drug that has a first-order elimination, one would expect that by the end of the first half-life, the drug would be reduced to 50%; by the end of the second half-life, to 25%; by the end of the third half-life, to 12.5%; by the end of the fourth half-life, to 6.25%; and by the end of the fifth half-life, to 3.125%. In general, a drug can be considered essentially eliminated after three to five half-lives, when less than 10% of the effective concentration remains. Table 29-1 shows the wide variation in half-life for several commonly used drugs.


Using a Loading Dose tubular secretion in the kidney, enzyme activity in the liver). Drug clearance is not affected by the distribution of drug throughout the body (VD) because clearance mechanisms act only on drug in the circulation.


The amount of time needed to eliminate a drug from the body depends on the clearance and the VD. The first-order elimination constant (Ke) represents the proportion of the apparent VD that is cleared of drug per unit of time during the drug’s exponential disappearance from the plasma over time (elimination phase):

K e = Cl / VD


The value of this constant for a particular drug can be determined by plotting drug concentration versus time on a log-linear plot (see Fig. 29-2) and measuring the slope of the straight line obtained during the exponential (elimination) phase.

To attain a desired therapeutic concentration rapidly, a loading dose is often used. In determining the amount of drug to be given, the clinician must consider the “volume” within the body into which the drug will be distributed. This volume is best described by the apparent VD. The loading dose can be calculated by multiplying the desired concentration by the VD:

Loading dose = desired concentration × VD


Rapid administration of the entire loading dose may produce an initially high peak concentration that results in toxicity. This problem can be avoided either by administering the loading dose as a divided dose or by varying the rate of access to the circulation, such as by administering the drug as an infusion (with an intravenous drug) or by taking advantage of the slower access to the circulation from various other routes (e.g., oral dosing). This approach is illustrated by phenytoin (see Table 29-1), which may need to be administered with a loading dose to achieve a therapeutic level (10 to 20 mg/L) rapidly. Because the VD for phenytoin is approximately 0.6 L/kg, the loading dose calculated from Equation 6 is 420 mg/L to attain a minimally


CHAPTER 29  Principles of Drug Therapy  


VD (L/kg)










0.62 ± 0.26

25 to 75%) in daily smoking consumption, there is little if any decrease in cardiovascular disease and lung or other smoking-related cancer risk, further substantiating the merits of quitting versus reducing smoking.


Nicotine is the primary reinforcer in tobacco smoke, with contributions from more than 4000 components to the sensory (non-nicotine) aspects of cigarette smoking. The primary site of action of nicotine is the α4β2 nicotinic acetylcholine receptor (nAChR), and the endogenous neurotransmitter acting on nAChRs is acetylcholine. nAChRs in the central nervous system (CNS) are pentameric ion channel complexes comprising two α- and three β-subunits; the seven α-subunits are designated α2 to α9 and the three β-subunits are designated β2 to β4. This produces considerable diversity in subunit combinations, which may explain the region-specific and functional selectivity of nicotinic effects in the CNS.3 Activation of nAChRs leads to Na+/Ca2+ ion channel fluxes and neuronal membrane depolarization. nAChRs are located presynaptically on several neurotransmitter-secreting neuron types in the CNS, including mesolimbic dopaminergic (DA) neurons that project from the ventral tegmental area (VTA) to the nucleus accumbens (NAc). Activation of nAChRs on mesolimbic DA neurons leads to DA secretion in the nucleus accumbens. At low concentrations of nicotine, α4β2 nAChR stimulation of afferent GABAergic projections onto mesoaccumbal DA neurons predominates, leading to reduced mesolimbic DA neuron firing and DA release. At higher nicotine concentrations, α4β2 nAChRs desensitize, and predominant activation of α7 nAChRs on glutamatergic projections occurs, leading to increased mesolimbic DA neuron firing and release. Within milliseconds of activation by nicotine, nAChRs desensitize. After overnight abstinence, nAChRs resensitize; this may explain why most smokers report that the first cigarette in the morning is the most satisfying. Interestingly, positron emission tomography (PET) neuroimaging studies have shown that smoking 2 or 3 puffs from a cigarette produces saturation of nAChRs in the brain reward system, suggesting that although binding to central nAChRs is an important first step in the effects of nicotine, it is not a complete explanation for continued smoking behaviors.

tar. Besides positive reinforcement, withdrawal, and craving, there are several secondary effects of nicotine and tobacco use that may contribute to both maintenance of smoking and smoking relapse, including mood modulation (e.g., reduction of negative affect), stress reduction, and weight control. In addition, conditioned cues can elicit the urge to smoke even after prolonged periods of abstinence. Specific effects might be most relevant to smokers wishing to lose weight and to those with psychiatric presentations (mood modulation, cognitive enhancement, stress reduction). These secondary effects may present additional targets for pharmacologic intervention in certain subgroups of smokers (e.g., those with schizophrenia or depression, or those concerned about their weight).


The Diagnostic and Statistical Manual, 5th edition (DSM-5),4 which was released in 2013 by the American Psychiatric Association, has changed the diagnostic terminology for nicotine and tobacco, eliminating the term dependence and instead using the term tobacco use disorder. Tobacco use disorder is established clinically by historical documentation of 2 of the following 11 criteria: 1. Tobacco often taken in larger amounts or over a longer period than was intended 2. Persistent desire or unsuccessful efforts to cut down or control tobacco use 3. A great deal of time spent in activities necessary to obtain or use tobacco 4. Presence of craving, or a strong desire or urge to use tobacco 5. Recurrent tobacco use resulting in failure to fulfill major obligations at work, school, or home 6. Continued tobacco use despite persistent or recurrent social or interpersonal problems caused or exacerbated by the effects of tobacco 7. Important social, occupational, or recreational activities given up or reduced because of tobacco use 8. Recurrent tobacco use in situations in which it is physically hazardous (e.g., smoking in bed) 9. Continued tobacco use despite persistent or recurrent physical or psychological problems that are caused or exacerbated by tobacco use 10. Tolerance, as defined by either a need for markedly increased amounts of tobacco to achieve desired effects, or markedly diminished effects with continued use of the same amount of tobacco 11. Withdrawal, manifested by the presence of the characteristic tobacco abstinence syndrome (e.g., four of the following: irritability, anxiety, difficulty concentrating, increased appetite, restlessness, dysphoric mood, insomnia), or tobacco (or nicotine) taken to relieve or avoid tobacco withdrawal symptoms. For abstinent smokers, remission is classified as early (between 3 and 12 months of abstinence) or sustained (>12 months of abstinence). Moreover, current severity of tobacco use disorder is coded as mild (2 or 3 symptoms), moderate (4 or 5 symptoms) or severe (6 or more symptoms). In addition, most physiologically dependent tobacco smokers state that they smoke their first cigarette of the day within the first 5 minutes of awakening (e.g., time to first cigarette 4 drinks/day Women: >7 drinks/week or >3 drinks/day ALCOHOL USE DISORDER CRITERIA* Tolerance Withdrawal More use than intended Craving Unsuccessful attempts to cut down Excessive time acquiring alcohol Activities given up because of use Use despite negative effects Failure to fulfill major role obligations Recurrent use in hazardous situations Continued use despite social or intrapersonal problems


Alcohol dehydrogenase

NADH NAD+ CH3COH CH3COO– Acetaldehyde Acetate Aldehyde Microsomal ethanol dehydrogenase oxidizing system

CH3CH2OH Ethanol

NADPH + H+ + 2O2


FIGURE 33-1.  Ethanol metabolism. Alcohol dehydrogenase predominates at low to moderate ethanol doses. The microsomal ethanol-oxidizing system is induced at high ethanol levels of chronic exposure and by certain drugs. Aldehyde dehydrogenase inhibition (genetic or drug induced) leads to acetaldehyde accumulation, particularly in the latter group.

*Mild = 2-3 criteria, moderate = 4-5 criteria, severe = 6 or more criteria. (From American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. 5th ed. Arlington, VA: American Psychiatric Publishing; 2013.)


At-risk drinking is defined differently for men younger than 65 years than for women of all ages because of generally lower body weights and lower rates of metabolism of alcohol in women; the definition in men older than 65 years is the same as in women because of the age-related increased risk for alcohol problems, in part owing to changes in alcohol metabolism in older individuals. Binge drinking or heavy drinking is the episodic consumption of large amounts of alcohol, usually five or more drinks per occasion for men and four or more drinks per occasion for women. One standard drink contains 12 g of pure alcohol, an amount equivalent to that contained in 5 ounces of wine, 12 ounces of beer, or 1.5 ounces of 80-proof spirits. The recently published Diagnostic and Statistical Manual of Mental Disorders, 5th edition (DSM-5) replaced the previous terminology of alcohol abuse and alcohol dependence with the term alcohol use disorders (see Table 33-1) in order to more clearly describe the spectrum of symptoms experienced by patients. Patients who meet 2 or 3 criteria are considered to have mild, 4 or 5 criteria moderate, and 6 to 11 criteria severe alcohol use disorder.1


In national surveys, 52% of American adults reported that they use alcoholic beverages (liquor, wine, or beer), whereas 23% reported binge drinking, and 6.5% reported heavy drinking in the past 30 days.2 Among individuals who use alcohol, many experience problems because of their drinking. It has been estimated that more than $100 billion is spent by American society each year to treat alcohol use disorders and to recover the costs of alcohol-related economic losses. Excessive alcohol consumption ranks as the third leading preventable cause of death in the United States after cigarette smoking and obesity. More than 100,000 deaths per year in the United States are attributed to alcohol use disorders. Population-based epidemiologic studies have shown that alcohol use disorders are among the most prevalent medical, behavioral, or psychiatric disorders in the general population. An epidemiologic survey of the general population in the United States estimated a prevalence of alcohol abuse and dependence (using the older DSM-IV criteria) to be between 7.4 and 9.7%. The lifetime prevalence of abuse and dependence is estimated to be even higher. Despite higher thresholds and tolerance, men are at least twice as likely as women to meet criteria for alcohol abuse and dependence by standard diagnostic survey techniques. Although sociodemographic features, such as young age, low income, and low education level, have been associated with an increased risk for problem drinking, alcohol use disorders are prevalent throughout all sociodemographic groups, and all individuals should be screened carefully. The “skid row” stereotype of the alcohol-dependent patient is much more the exception than the rule. The prevalence of alcohol use disorders is higher in most health care settings than it is in the general population because alcohol problems often result in treatment-seeking behaviors. The prevalence of problem drinking in general outpatient and inpatient medical settings has been estimated between 15 and 40%. These data strongly support the need for physicians to screen all patients for alcohol use disorders.

Beverage alcohol contains ethanol, which acts as a sedative-hypnotic drug. Alcohol is absorbed rapidly into the blood stream from the stomach and intestinal tract. Because women have lower levels of gastric alcohol dehydrogenase, the enzyme primarily responsible for metabolizing alcohol, they experience higher blood alcohol concentrations than do men who consume similar amounts of ethanol per kilogram of body weight. The absorption of alcohol can be affected by other factors, including the presence of food in the stomach and the rate of alcohol consumption. By means of metabolism in the liver, alcohol is converted to acetaldehyde and acetate (Fig. 33-1). Metabolism is proportional to an individual’s body weight, but a variety of other factors can affect how alcohol is metabolized. A genetic variation in a significant proportion of the Asian population alters the structure of an aldehyde hydrogenase isoenzyme, resulting in the development of an alcohol flush reaction, which includes facial flushing, hot sensations, tachycardia, and hypotension. In the brain, alcohol seems to affect a variety of receptors, including γ-aminobutyric acid (GABA), N-methyl-d-aspartate, and opioid receptors. Glycinuric and serotoninergic receptors also are thought to be involved in the interaction between alcohol and the brain. The phenomena of reinforcement and cellular adaptation are thought, at least in part, to influence alcoholdependent behaviors. Alcohol is known to be reinforcing because withdrawal from ethanol and ingestion of ethanol itself are known to promote further alcohol consumption. After chronic exposure to alcohol, some brain neurons seem to adapt to this exposure by adjusting their response to normal stimuli. This adaptation is thought to be responsible for the phenomenon of tolerance, whereby increasing amounts of alcohol are needed over time to achieve desired effects. Although much has been learned about the variety of effects alcohol can have on various brain receptors, no single receptor site has been identified. A variety of neuropsychological disorders are seen in association with chronic ethanol use, including impaired short-term memory, cognitive dysfunction, and perceptual difficulties. Although the brain is the primary target of alcohol’s actions, a variety of other tissues have a major role in how alcohol affects the human body. Direct liver toxicity may be among the most important consequences of acute and chronic alcohol use (Chapter 152). A variety of histologic abnormalities ranging from inflammation to scarring and cirrhosis have been described. The pathophysiologic mechanism of these effects is thought to include the direct release of toxins and the formation of free radicals, which can interact negatively with liver proteins, lipids, and DNA. Alcohol also has substantial negative effects on the heart and cardiovascular system. Direct toxicity to myocardial cells frequently results in heart failure (Chapter 58), and chronic heavy alcohol consumption is considered to be a major contributor to hypertension (Chapter 67). Other organ systems that experience significant direct toxicity from alcohol include the gastrointestinal tract (esophagus, stomach), immune system (bone marrow, immune cell function), and endocrine system (pancreas, gonads).


Alcohol has a variety of specific acute and chronic effects. The acute effects seen most commonly are alcohol intoxication and alcohol withdrawal. Chronic clinical effects of alcohol include almost every organ system.

CHAPTER 33  Alcohol Use Disorders  

Acute Effects Alcohol Intoxication

After entering the blood stream, alcohol rapidly passes through the bloodbrain barrier. The clinical manifestations of alcohol intoxication are related directly to the blood level of alcohol. Because of tolerance, individuals chronically exposed to alcohol generally experience less severe effects at a given blood alcohol level than do individuals who are not chronically exposed to alcohol. The symptoms of mild alcohol intoxication in nontolerant individuals typically occur at blood alcohol levels of 20 to 100 mg/dL and include euphoria, mild muscle incoordination, and mild cognitive impairment. At higher blood alcohol levels (100 to 200 mg/dL), more substantial neurologic dysfunction occurs, including more severe mental impairment, ataxia, and prolonged reaction time. Individuals with blood alcohol levels in these ranges can be obviously intoxicated, with slurred speech and lack of coordination. These effects progress as the blood alcohol level rises to higher levels, to the point at which stupor, coma, and death can occur at levels equal to or greater than 300 to 400 mg/dL, especially in individuals who are not tolerant to the effects of alcohol. The usual cause of death in individuals with very high blood levels of alcohol is respiratory depression and hypotension.

Alcohol Withdrawal Syndrome

Alcohol withdrawal can occur when individuals decrease their alcohol use or stop using alcohol altogether. The severity of symptoms can vary greatly. Many individuals experience alcohol withdrawal without seeking medical attention, whereas others require hospitalization for severe illness. Because ethanol is a central nervous system depressant, the body’s natural response to withdrawal of the substance is a hyperexcitable neurologic state. This state is thought to be the result of adaptive neurologic mechanisms being unrestrained by alcohol, with an ensuing release of a variety of neurohumoral substances, including norepinephrine. In addition, chronic exposure to alcohol results in a decrease in the number of GABA receptors and impairs their function. The clinical manifestations of alcohol withdrawal include hyperactivity resulting in tachycardia and diaphoresis. Patients also experience tremulousness, anxiety, and insomnia. More severe alcohol withdrawal can result in nausea and vomiting, which can exacerbate metabolic disturbances. Perceptual abnormalities, including visual and auditory hallucinations and psychom*otor agitation, are common manifestations of more moderate to severe alcohol withdrawal. Grand mal seizures commonly occur during alcohol withdrawal, although they do not generally require treatment beyond the acute withdrawal phase. The time course of the alcohol withdrawal syndrome can vary within an individual and by symptom complex, and the overall duration of symptoms can be a few to several days (Fig. 33-2). Tremor is typically among the earliest symptoms and can occur within 8 hours of the last drink. Symptoms of tremulousness and motor hyperactivity typically peak within 24 to 48 hours. Although mild tremor typically involves the hands, more severe tremors can involve the entire body and greatly impair a variety of basic motor functions. Perceptual abnormalities typically begin within 24 to 36 hours after the last drink and resolve within a few days. When withdrawal seizures occur, they are typically generalized tonic-clonic seizures and most often occur within 12 to 24 hours after reduction of alcohol intake. Seizures can occur, however, at later time periods as well.

Tremor, anxiety, insomnia

The most severe manifestation of the alcohol withdrawal syndrome is delirium tremens. This symptom complex includes disorientation, confusion, hallucination, diaphoresis, fever, and tachycardia. Delirium tremens typically begins after 2 to 4 days of abstinence, and the most severe form can result in death.3

Chronic Effects

Acute manifestations, including intoxication and withdrawal, are generally stereotypical in their appearance and time course, but chronic manifestations tend to be more varied. Many patients with alcohol dependence may be without evidence of any chronic medical manifestations for many years. As time goes on, however, the likelihood that one or more of these manifestations will occur increases considerably. All major organ systems can be affected, but the primary organ systems involved are the nervous system, cardiovascular system, liver, gastrointestinal system, pancreas, hematopoietic system, and endocrine system (Table 33-2). Patients who drink are at risk for a variety of malignant neoplasms, such as head and neck, esophageal, colorectal, breast, and liver cancers (see individual chapters on those cancers). Excessive alcohol use often causes significant psychiatric and social morbidity that can be more common and more severe than the direct medical effects, especially earlier in the course of problem drinking.

Nervous System

In addition to the acute neurologic manifestations of intoxication and withdrawal, alcohol has major chronic neurologic effects. About 10 million Americans have identifiable nervous system impairment from chronic alcohol use. Individual predisposition to these disorders is highly variable and is related to genetics, environment, sociodemographic features, and gender; the relative contribution of these factors is unclear. In the central nervous system, the major effect is cognitive impairment. Patients may present with mild to moderate short-term or long-term memory problems or may have severe dementia resembling Alzheimer disease (Chapter 402). The degree to which the direct toxic effect of alcohol is responsible for these problems or the impact of alcohol-related nutritional


Delirium tremens 0






Days since last drink FIGURE 33-2.  Time course of alcohol withdrawal.





Nervous system

Intoxication Withdrawal Cognitive impairment Cerebellar degeneration Peripheral neuropathy

Cardiovascular system

Cardiac arrhythmias Chronic cardiomyopathy Hypertension


Fatty liver Alcoholic hepatitis Cirrhosis

Gastrointestinal tract, esophagus

Chronic inflammation Malignant neoplasms Mallory-Weiss tears Esophageal varices


Gastritis Peptic ulcer disease


Acute pancreatitis Chronic pancreatitis

Other medical problems

Cancers: mouth, oropharynx, esophagus, colorectal, breast, hepatocellular carcinoma Pneumonia Tuberculosis


Depression Anxiety Suicide

Behavioral and psychosocial

Injuries Violence Crime Child or partner abuse Tobacco, other drug abuse Unemployment Legal problems





CHAPTER 33  Alcohol Use Disorders  

deficiencies is uncertain (Chapter 416). The deficiency of vitamins such as thiamine may plays a major role in promoting alcoholic dementia and severe cognitive dysfunction, as is seen in Wernicke encephalopathy and Korsakoff syndrome (Chapter 416). Alcohol also causes a polyneuropathy that can present with paresthesias, numbness, weakness, and chronic pain (Chapters 416 and 420). As with the central nervous system, peripheral nervous system effects are thought to be caused by a combination of the direct toxicity of alcohol and nutritional deficiencies. A small proportion (1%


Fraction of all human heterozygosity attributable to variants with a frequency of >1%


contains essentially all common sequence variants in the human population (with frequency >1%). At the time of this writing, this public database contains more than 44 million human genetic variants ( SNP/index.html). Not all these entries represent common variants (some are rare), and a small fraction may represent technical false-positive findings. The major contribution of common variation in human sequence diversity is explained by the unique demographic history of the human population. Despite the global distribution of the current human population, it is now clear that all humans are the descendants of a single population that lived in Africa only 10,000 to 40,000 years ago. The ancestral population was small (with an effective size of perhaps 10,000 individuals), lived a hunter-gatherer existence at low population densities (relative to other humans and later domesticated animals), and had evolved in Africa during millions of years. Most human genetic variation arose in this phase of human history, before the more recent migrations, expansions, and invention of technologies (e.g., farming) that resulted in widespread population of the globe. Most common human genetic variation predates the Diaspora and is shared by all populations on earth. A second factor is the slow rate of change in human DNA. Mutation and recombination occur at very low rates, on the order of 10−8 per base pair per generation; and yet, any pair of human genes traces a lineage back to a shared ancestor who lived on the order of 103 to 104 generations ago (if a generation is 20 years, then 104 generations is 200,000 years). In other words, considering the typical nucleotide in two unrelated humans, it is more likely that they trace back to a shared ancestor without any mutation having occurred than it is that a mutation has arisen in the intervening time. This explains why 99.9% of base pairs are identical when any two copies of the human genome are compared. Another aspect of human variation is explained by these simple mathematical and population genetic relationships: the extent of human DNA sequence diversity attributable to rare and common variants. Each of us inherits from our parents some 3 million common polymorphisms (classically defined as those with frequency of >1%). We inherit genetic variants that are shared by apparently unrelated individuals but are at frequencies less than 1%, and we inherit thousands of variants that are limited only to the individual and the individual’s closest relatives. The shared ancestry of human populations explains another aspect of human genetic variation: the correlations among nearby variants known as linkage disequilibrium. Empirically, individuals who carry a particular common variant at one site in the genome are observed to be more likely than chance to carry a particular set of variants at nearby positions along the chromosome. That is, not all combinations of nearby variants are observed in the population but rather only a small subset of the possible combinations. These correlations reflect the fact that most variants in our genomes arose once in human history (typically long ago) and did so on an arbitrary but unique copy carried by the individual in whom the mutation first arose. This unique ancestral copy can be recognized in the current population by the stretch of


particular alleles (known as a haplotype). These ancestral haplotypes, passed down from shared prehistoric ancestors in Africa, offer a practical tool in association studies of human disease because it is not necessary to measure directly each nucleotide to capture much of the information.


The genetic architecture of a disease refers to the number and magnitude of genetic risk factors that exist in each patient and their frequencies and interactions in the population. Diseases can be due to a single gene in each family (monogenic) or to multiple genes (polygenic). It is easiest to identify genetic risk factors when only a single gene is involved and this gene has a large impact on disease in that family. In cases in which a single gene is necessary and sufficient to cause disease, the condition is termed a mendelian disorder because the disease tracks perfectly with a mutation (in the family) that obeys Mendel’s simple laws of inheritance. Some single-gene disorders are caused by the same gene in all affected families; for example, cystic fibrosis is always caused by mutations in CFTR. Although many individuals with cystic fibrosis carry the same founder mutation (δ-508), others carry any pair of a wide variety of different mutations in CFTR. The existence of many different mutations at a given disease gene is known as allelic heterogeneity. A mendelian disorder can be due to a single genetic lesion in any given family but in different families can be due to mutations in a variety of genes. This phenomenon, termed locus heterogeneity, is illustrated by retinitis pigmentosa. Although mutation in a single gene is typically necessary and sufficient to cause retinitis pigmentosa, there are dozens of different genes in which retinitis pigmentosa mutations have been found (Online Mendelian Inheritance in Man #268000). In each family, however, only one such gene is mutated to cause disease. Most single-gene disorders are rare (present in 1%) B. Rare mutations (500 conditions


Information on potentially clinically actionable gene-drug associations and genotypephenotype relationships; currently lists 186 well-known pharmacogenomic associations and provides 46 summaries for very important genes

CHAPTER 43  Application Of Molecular Technologies to Clinical Medicine  

REVIEW QUESTIONS 1. Genome sequencing of DNA can be used for all of the following except which one? A. Diagnosis of rare diseases B. Detection of pharmacogenetic variants C. Newborn screening D. Prognosis of cancer E. Diagnosis of infection Answer: D  Various types of genome sequencing have served special clinical functions. Exome sequencing is being increasingly used to diagnose rare genetic diseases and patients with diagnostic dilemmas. Germline genotyping has allowed prediction of adverse events and dosing of many commonly used drugs; although these discoveries have been “actionable,” most have failed to diffuse widely into clinical practice at this time. Screening newborns, prenatal diagnosis, and preconception carrier testing are feasible for identified mendelian disorders. In infectious diseases, diagnosis of causative pathogens can be performed by next-generation sequencing, which may in the future supplant the need to first grow microorganisms in culture. Genome sequencing has not been demonstrated to be useful for prognosis of cancer at this time. 2. Transcriptional (RNA) profiling may be useful for all of the following except which one? A. Susceptibility B. Diagnosis C. Prognosis D. Pharmacogenetics E. Monitoring Answer: A  Next-generation sequencing has allowed sequencing of RNA transcripts in cells. The so-called transcriptome, unlike the static genome, is continually changing. It allows examination, at any given time, of alternatively gene-spliced transcripts, post-transcriptional changes, gene fusion, and changes in gene expression. Transcriptional (RNA) profiling has been applied to diagnosis, prognosis, pharmacogenomics, and monitoring, but not susceptibility analysis.


3. Among the great challenges to implementing genomic diagnostics in the clinic is which one of the following? A. The precision of the results B. The evidence to support use C. The ability of the laboratories to perform the test D. Their integration into electronic medical records E. Lack of patient understanding Answer: B  Surmountable challenges to implementing genomic diagnostics at the bedside include increasing the precision of results, expanding the availability of expert laboratories to perform the tests, the ability to integrate the information into electronic medical records, and lack of understanding by patients who are, however, now becoming increasingly savvy about this aspect of their own health care. Evidence to support the utility and costeffectiveness of wide application of genomic diagnostics in clinical practice is only now being taken up by investigators in comparative effectiveness and health services research. 4. Which of the following describes the microbiome? A. A small genome B. A community of microbes colonizing humans C. A tool used to make thinly sliced materials (e.g., paraffin blocks) D. The sequence of a virus or bacterium E. An organelle of the cell Answer: B  The microbiome is the community of microbes that colonize a human host. It is the ecological community of commensal, symbiotic, and pathogenic microorganisms that actually inhabit our body space and outnumber our own native human cells by a ratio of 10 : 1. 5. Which of the following methods may be used for detection of microbial pathogens? A. Transcriptional profiling B. Metabolomics C. Sequencing D. DNA methylation E. Proteomics Answer: C  At this time, sequencing is used to detect and identify a microbial pathogen.

CHAPTER 44  Regenerative Medicine, Cell, and Gene Therapies  




Introduction and Definitions

A remarkable clinical need exists for the development and clinical assessment of various methods to facilitate the regeneration of injured or diseased tissues and organs. This need derives from the unrelenting prevalence of trauma, congenital disorders, ischemia, and degenerative processes, which becomes increasingly urgent as the global population expands and ages. Cancer is tied to this field both directly (e.g., replacement of lost vital organ function as a result of cancer invasion or treatment modalities, cell-based delivery of cancer immune and gene therapies) and indirectly (e.g., role of stem cells in cancer


CHAPTER 44  Regenerative Medicine, Cell, and Gene Therapies  

pathogenesis, risk for tumorigenesis in stem cell−based therapies). The recent developments in stem cell biology, molecular interventions, biopolymers, and other related biologic and engineering disciplines have paved the way to the emerging research and clinical discipline of regenerative medicine.1

Regenerative Medicine

Regenerative medicine seeks to harness methods for the replacement or repair of dysfunctional cells, tissue, or organs in an attempt to restore normal function. It therefore draws on therapies from the three conventional pillars of medical therapeutics (pharmaceuticals, biologics, and medical devices) as well as from the newest platform technology, namely, cell therapy. The longterm goal of regenerative medicine is to cure disease by replacing the lost functions of tissues and organs, and thus it truly represents a paradigm shift from conventional therapies aiming to alter the natural course of disease or to provide symptomatic control. Consequently, regenerative medicine aims to develop curative strategies for unmet clinical needs such as diabetes, heart failure, and neurodegenerative disorders, among others.

Cell Therapy

Cell therapy involves the application of cells to achieve a therapeutic benefit, regardless of the cell type or clinical indication. Although achieving tissue and organ regeneration through cell replacement represents an important goal of cell therapy technology, its applications may reach far beyond the field of regenerative medicine. Hence, the spectrum of cell therapy approaches may range from permanent cell replacement strategies (attempting to replace lost or dysfunctional cells) to more transient cell therapies aiming to modulate disease progression or to protect tissues at risk, to achieve immunomodulatory effects (e.g., for prevention of graft-versus-host disease), to act as vehicles for the delivery of genes or gene products (cell-based gene therapy strategies), and even to act as cell-based cancer vaccines. This chapter focuses on the use of cell therapy for regenerative medicine and specifically concentrates on the potential role of different stem cell types to meet this challenge.

Stem Cells

Stem cells possess two defining properties: (1) the capacity for self-renewal and (2) the ability to differentiate into cell types with specialized cellular functions (Fig. 44-1). This may occur at the individual stem cell level through the process of asymmetrical cell division or at the cell population level wherein a subset of cells differentiate and the remaining stem cells remain dormant or replicate themselves as stem cells. After asymmetrical cell division, non–stem cell derivatives may either generate a pool of organ system–restricted, transitamplifying cells with enhanced proliferative capacity or continue to differentiate by epigenetic and gene expression profile changes until reaching the terminally differentiated state. This conceptual framework was developed after the discovery of bone marrow cells that were capable of reconstituting the adult hematopoietic system. These hematopoietic stem cells constitute the basis for hematopoietic stem cell transplantation, the only form of stem cell therapy currently routinely well established in clinical practice (Chapter 178).

Stem cell

Stem cell

Specialized cell

(e.g., neuron) FIGURE 44-1.  Asymmetrical cell division. Although this first characteristic was considered a required characteristic for stem cells based on their original description in the adult hematopoietic system, not all cell types currently named as stem cells necessarily display this property. For instance, human embryonic stem cells divide by symmetrical cell division.

The different stem cells types are routinely classified based on the protein or transcription factors they express, but also according to three basic additional attributes. These include replicative capacity (limited vs. unlimited), the scope or potency of differentiation (e.g., pluripotent, multipotent, oligopotent, unipotent), and their place in the life history of the organism (developmental or postdevelopmental). Thus, more recent terminology has broadened use of the term stem cells to cover a wider array of cell types that contribute to organ development or have the capacity to repopulate tissues and organ systems. The term stem cells, together with the formulations noted previously, has also recently been extrapolated to describe certain cellular subpopulations that may be principally responsible for the growth of malignant tumors. However, because cancer stem cells have no role in tissue regeneration, they are considered in Chapter 181. Adult (Postnatal) Stem Cells

After birth, many tissues are thought to contain a subpopulation of cells with the capacity for extended self-renewal, combined with the ability to differentiate into more mature cell types with specialized functions (Fig. 44-2). Adult stem cells, thought to represent less than 0.01% of the total number of cells, are located in specialized supportive niche compartments at various sites within the hematopoietic system and elsewhere, and respond to cues in their local microenvironment. As a result of the success of hematopoietic stem cell transplantation in the treatment of bone marrow failure or in conjunction with myeloablative therapy in malignancy, scientists have been motivated to find adult stem cells in other organs. Adult tissues and organ systems reported to contain putative stem cells include bone marrow (hematopoietic and mesenchymal compartments) and peripheral blood, blood vessel endothelium, dental pulp, epithelia of the skin, adipose tissue, digestive system, cornea, retina, testis, and liver. Similar stem/progenitor cells were also reported in organs historically not thought to contain such cells, such as the central nervous system, the heart, and the kidney. Whether adult stem cells represent remnants of developmental stem cells that persist into adulthood for purposes of organ maintenance and repair or represent a distinct cell type dedicated for this latter purpose is not clear. Importantly, in many organs, despite the presence of such tissue-specific stem cells, their regenerative capacity is still inadequate to deal with massive cell loss such as occurs, for example, after a large myocardial infarction or after ischemic brain injury. Embryonic and Induced Pluripotent Stem Cells

In contrast to adult stem cells that have relatively limited differentiation potency, cells in the developing preimplantation embryo still retain the capacity to differentiate into derivatives of all three germ layers (ectoderm, mesoderm, and endoderm), eventually contributing to all tissues in the body (Fig. 44-3). In normal development, however, such cells do not persist beyond the blastocyst stage. When isolated from unused preimplantation blastocysts generated for in vitro fertilization, the inner cell mass cells isolated can be used to generate human embryonic stem cell (hESC) lines (see Fig. 44-3). The generated hESCs exhibit unlimited self-renewal in cell culture in the undifferentiated state, while retaining the capacity to differentiate into cell derivatives of all three germ layers, essentially giving rise to any cell type in the body. Taking advantage of lessons learned from embryology, scientists were able to utilize the sequential application of different combinations of growth factors to achieve efficient differentiation systems from hESCs, yielding purified populations of different types of neurons, glial cells, cardiomyocytes, vascular endothelial and smooth muscle cells, pancreatic β cells, hepatocytes, different blood cells (platelets, red blood cells), and several other cell lineages. One of the limitations of the hESC technology is the inability to derive such cells from an adult individual, preventing their utilization in a patientspecific manner. These limitations can be overcome with the introduction of induced pluripotent stem cell (iPSC) technology.2 This approach allows adult somatic cells (e.g., fibroblasts) to be reprogrammed into pluripotent stem cells by introduction of a set of transcription factors linked to pluripotency (the originally reported combination of factors included OCT3/4, SOX2, c-MYC, and KLF4). The human iPSCs (hiPSCs) generated in this manner can then be coaxed to differentiate into a variety of cell types, using differentiation protocols similar to those already in place for hESC (Fig. 44-4). Importantly, because the hiPSCs can be generated in a patient-specific manner, this technology can potentially be used to develop autologous cellreplacement strategies that can evade the immune system, to generate patient- and disease-specific models of different genetic disorders, and to establish screens for drug testing and drug discovery.


CHAPTER 44  Regenerative Medicine, Cell, and Gene Therapies  


Natural killer (NK) cell T lymphocytes


Bone Lymphoid progenitor cell Eosinophil

B lymphocyte

Hematopoietic stem cell

Multipotent stem cell


Myeloid progenitor cell Red blood cells

Bone matrix

Bone (or cartilage) Stromal cell


Hematopoietic supportive stroma

Marrow adipocyte

Osteoblast Stromal stem cell

Lining cell Osteocyte

Blood vessel

Pre-osteoblast Skeletal muscle stem cell? Adipocyte

Hepatocyte muscle stem cell?

Hematopoietic stem cell

Osteoclast FIGURE 44-2.  Adult stem cells. Adult stem cells can be multipotent and have the capacity to differentiate into a limited number of different cell types, often restricted to a given tissue or organ system, as in the case of adult hematopoietic or epidermal stem cells. Two stem cell types have been isolated from adult bone marrow—the hematopoietic stem cell and the mesenchymal stem cell. Adult mesenchymal stem cells of bone marrow origin, although their range of differentiation has been shown to be broader than that of any other adult stem cell type, do not reach pluripotency. It is thought that in some organ systems, such as the gastrointestinal epithelium, a unipotent pool of progenitors exists for repopulating a rapid population turnover of only one type of cell—although it is difficult to be certain whether such progenitors can be distinguished from the overall population of fully differentiated cells in tissues with high cellular turnover.

Cell Therapy Approaches to Regenerative Medicine

Historically, the field of cell therapy can be traced to the transfusion of blood and blood products (Chapter 177), solid organ transplantation (Chapter 49), in vitro fertilization, and bone marrow transplantation (Chapter 178). Nevertheless, beyond the aforementioned therapies, which have become the mainstay treatments in several medical fields, additional cell therapy approaches are considered highly experimental and are still at different stages of preclinical and clinical development. These ongoing efforts can be conceptually grouped into six different approaches (Fig. 44-5).

Delivery of Bone Marrow− and Blood-Derived Stem/ Progenitor Cells

A flurry of studies during the past decade evaluated the ability of bone marrow−derived hematopoietic or mesenchymal stem cells to achieve tissue repair after delivery to a variety of organs. These studies were based initially on the assumption that these types of adult stem cells may display some degree of plasticity, allowing them to transdifferentiate into the relevant cell types (e.g., heart cells, nerve cells, and liver cells) after transplantation into the appropriate tissue environment. Although mounting evidence suggests that such transdifferentiation probably does not occur to a significant extent, many of these studies appeared to result in some degree of functional improvement after stem cell delivery to different organs. This clinical benefit may stem from the secretion of different growth factors by the engrafted cells (“paracrine hypothesis”); these factors in turn are thought to augment endogenous tissue repair mechanisms, improve tissue vascularization, modulate inflammation, and protect tissues at risk.

Delivery or Activation of Tissue-Specific Stem/Progenitor Cells or Induction of Cell Proliferation

In contrast to the conventional dogma, recent evidence suggests that a number of organs previously believed to lack any regenerative capacity (e.g., the brain, pancreas, kidney, and heart) in fact do possess such ability, albeit at a limited capacity. Whether this capability is due to the presence of tissue-specific stem/progenitor cells or due to some replication capa-

bility of terminally differentiated cells is still a matter of debate for each organ. Significant efforts have been made in recent years to isolate such putative tissue-specific stem/progenitor cells based on the expression of general or specific stem cell markers or based on their unique culturing properties. These studies also highlighted the potential of such cells to be cultured in a clonal manner and to give rise to one or more cell types relevant to the organs from which they were isolated. Current efforts to utilize the aforementioned findings for regenerative medicine are focused either on the isolation, ex vivo expansion, and transplantation of such putative stem/progenitor cells back to their respective native organs or on the augmentation of their endogenous reparative potential in vivo. The former strategy can be exemplified in the central nervous system where progenitor cells are harvested, cultivated in culture (as neurospheres), and give rise to different types of neurons and supporting glial cells. Similar efforts have followed in other organs. In the heart, for example, such efforts have already reached early clinical trials, in which autologous cardiac stem cells were harvested from the heart, expanded ex vivo, and then engrafted back to the heart. The latter approach, in contrast, aims to influence putative stem cell niches within damaged organs to enhance the endogenous reparative properties of those stem/progenitor cells. Such an effect may underlie the potential therapeutic benefit of bone marrow−derived stem cells after their delivery to different organs. The final strategy aims to boost endogenous organ repair through the replication of terminally differentiated tissue-specific cells. Such strategies can either augment the inherent physiologic capability of a given organ (e.g., insulin secretagogues for pancreatic β cells) or attempt to induce replication in cells that have already withdrawn from the cell cycle. Caution is warranted with respect to the latter approach because induction of uncontrolled proliferation (e.g., by genetic manipulation) may increase the risk for tumorigenesis.

Engraftment of Fetal Tissue

The most straightforward approach to organ repair would be to replace the missing cells with identical counterparts. Harvesting and expanding adult


CHAPTER 44  Regenerative Medicine, Cell, and Gene Therapies  

In vitro fertilization Day 0

Totipotent cells Day 3

Blastocyst Day 5 Origin: Derived from preimplantation or peri-implantation embryo

Self-renewal: The cells can divide to make copies of themselves for a prolonged period of time without differentiating.

Stem cell

Pluripotency: Embryonic stem cells can give rise to cells from all three embryonic germ layers even after being grown in culture for a long time.

The three germ layers and one example of a cell type derived from each layer: Ectoderm


Neuron Ectoderm gives rise to: brain, spinal cord, nerve cells, hair, skin, teeth, sensory cells of eyes, ears, nose, and mouth, and pigment cells.

Blood cells

Mesoderm gives rise to: muscles, blood, blood vessels, connective tissues, and the heart.


Liver cell Endoderm gives rise to: the gut (pancreas, stomach, liver, etc.), lungs, bladder, and germ cells (egg or sperm).

FIGURE 44-3.  Embryonic stem cells. Totipotency refers to the capacity to differentiate into all cell types in an organism, including extraembryonic tissues, placenta, and umbilical cord, a property confined to the fertilized egg itself, including the cells derived from the first few cell divisions after fertilization. Pluripotency refers to the capacity to differentiate into all the specialized cell types derived from the three germ layers (ectoderm, mesoderm, endoderm) of the developing embryo and is a hallmark feature of embryonic stem and germ cells.

human cells for transplantation, however, may not be possible in the case of several organs with limited regenerative capacity. During prenatal human development, cells of fetal origin often show enhanced proliferative capacity as well as the ability to differentiate into more than one type of mature or specialized cell. Moreover, animal studies have demonstrated that transplantation of tissues harvested from developing organs (harvested within a specific time window during embryonic development) may give rise to entire functioning organs such as kidneys, lungs, and pancreas. Nevertheless, to date, the only fetal-derived cells that have been used in human clinical applications are the dopaminergic cells derived from the developing fetal nervous system for the treatment of Parkinson disease (Chapter 409). The broader use of fetal tissues for regenerative medicine may be hampered by the limited

access to such cells for both technical and ethical reasons, the allogeneic nature of such procedures (requiring immune suppression), and the potential for tumor formation as already described in some case reports.

Transplantation of Ex Vivo Differentiation of Pluripotent Stem Cells

Unlike fetal tissues, hESCs are truly pluripotent (can give rise to advanced cell derivatives of all three germ layers). Importantly, hESCs can be propagated in the undifferentiated state and then coaxed to differentiate into a variety of cell types, giving rise to a potentially unlimited number of specialized cell types for transplantation. Consequentially, numerous preclinical studies have demonstrated the ability of hESC derivatives to engraft, survive,


CHAPTER 44  Regenerative Medicine, Cell, and Gene Therapies  




Human iPS cells



Somatic cells: (Fibroblasts)

Guided differentiation

Cell therapy and tissue engineering


Gene correction through gene editing




















Patient-specific therapies (personalized medicine)


Disease Modeling

Drug Screening and Discovery

FIGURE 44-4.  Application of the induced pluripotent stem cells (iPSC) technology. Patient-specific human iPSC can be generated by reprogramming of adult somatic cells (fibroblasts) with a set of transcription factors and then coaxed to differentiate into a variety of cell lineages. The patient-specific human iPSCs can then be transplanted back to the patient in an autologous manner for regenerative medicine applications. In a similar manner disease- and patient-specific human iPSC models of inherited disorders could be generated (“disease-in-a-dish models”) and used for better understanding of genetic disorders, for drug development, and for optimizing patient-specific therapies. Gene editing techniques can be used for mutation correction and for transplantation of healthy cells. CM = cardiomyocytes; iPS = induced pluripotent stem cells.

and improve organ performance in a wide spectrum of relevant animal disease models (e.g., heart failure, Parkinson disease and other neurodegenerative disorders, diabetes). Early clinical studies using hESC derivatives are just emerging and have been focused so far on the retina (transplantation of retinal pigment epithelium [RPE] cells) and spinal cord injury (using oligodendrocyte progenitors). Despite the significant achievements made with hESCs, the inability to create patient-specific hESCs from adult individuals, the ethical issues arising from destructive use of human embryos, and the anticipated immune rejection associated with such allogeneic cell transplantation impose important hurdles for their clinical utilization. The hiPSC technology provides a potential solution to these challenges. As noted, the patient’s own somatic cells (fibroblasts, hair follicles, urine epithelial cells, or blood cells) could be reprogrammed by a set of transcription and chemical factors to yield pluripotent stem cells. The patient-specific hiPSCs could then be coaxed to differentiate to a variety of cell lineages, using protocols similar to those already in place for hESCs. In turn, these differentiated derivatives could then be transplanted either in an autologous or allogeneic manner. Clinical trials using hiPSC-derived cell lineages are expected to be initiated in the coming few years, with the initial targets being macular degeneration (RPE cells), Parkinson disease (dopaminergic neurons), blood product transfusion (hiPSCderived platelets and red blood cells), and heart failure (cardiomyocytes). One of the concerns in translating hESCs and hiPSCs into a therapeutic platform is the oncogenic risk. This concern stems from the potential for remaining undifferentiated cells within the cell grafts to form teratomas, from

the use of oncogenic reprogramming factors, from the random integration of the viral vectors used in cellular reprogramming (“insertional oncogenesis”), and from genetic instability, potentially leading to both chromosomal aberrations and mutations. Progress to clinical trials requires definitive clarification of this key concern.

Direct Reprogramming

In contrast to the iPSC approach, which seeks to initially reprogram somatic cells to a pluripotent state followed by differentiation of the generated iPSCs to specific cell lineages, recently described direct reprogramming strategies aim to convert the phenotype of one mature cell type (fibroblasts) directly to another. The prototype for such a strategy was the demonstration that MyoD, a master regulator of skeletal muscle formation, can convert fibroblasts directly to skeletal muscle. Progress to derive other cell types after this report was delayed for many years because, unlike skeletal muscle, a single master developmental regulatory gene does not exist for most cell lineages. Based on the experimental approach used to identify the combination of transcription factors that can reprogram somatic cells into iPSCs, researchers evaluated the ability to achieve analogous transcription factor reprogramming strategies to convert the cell fate of somatic cells directly. Consequentially, using a combination of lineage-specific developmental transcription factors, scientists were able to convert terminally differentiated fibroblasts or other somatic cells directly to neurons, β cells, different hematopoietic cell lineages, and cardiomyocyte-like cells. Recent studies have taken this concept a further step forward by demonstrating that transcription factor−based


CHAPTER 44  Regenerative Medicine, Cell, and Gene Therapies  

Induction of Endogenous Regeneration/Repair

Cell/Tissue Transplantation Blastocyst Fetal tissue (fetal ventral mesencephalon)

Activation of tissue-specific stem/progenitor cells hESCs, hiPSCs

Oct4, Sox2, Klf4, cMyc Fibroblasts

Neurons Fibroblasts Induction of cell replication



Direct TF reprogramming


In vivo TF direct reprogramming

Tissue engineering (biopolymers and cells)


Additional mechanisms: • Promotion of angiogenesis • Modulation of inflammation • Cell protection • Trophic effects

Bone marrow Hematopoietic/ mesenchymal stem cells

FIGURE 44-5.  Conceptual framework for regenerative medicine approaches. These strategies can be divided into those attempting to augment endogenous regeneration (left side) and those focusing on transplantation of cells (right side). The former could be achieved through the activation of putative tissue-specific stem/progenitor cells, through induction of cell replication, by in vivo transcription factor (TF)-based direct reprogramming (directly converting one somatic cell [fibroblast] into another), and by several other indirect means (e.g., modulation of inflammation, induction of angiogenesis, trophic effect, protection of tissue at risk). Cell sources that can be used for cell transplantation include fetal tissues (e.g., dopaminergic-rich fetal ventral mesencephalon for Parkinson disease), pluripotent stem cells (human embryonic stem cells [hESCs] and human induced pluripotent stem cells [hIPSCs]), derived cell-lineages, and somatic cells that can be generated ex vivo by direct transcription factor−based reprogramming of fibroblasts. CMs = cardiomyocytes.

transdifferentiation can also be achieved in vivo, suggesting a method whereby resident cells (fibroblasts, hepatic cells, or other cells) could be converted to the appropriate cell types for organ repair. Development of the latter approach for clinical application may be considered more analogous to gene therapy, with the associated advantages, shortcomings, and challenges of this discipline (see later).

Tissue Engineering

Tissue engineering is an interdisciplinary technology combining principles from life sciences and engineering with the goal of developing functional substitutes for damaged tissues and organs.3,4 Rather than simply introducing cells into a diseased area, in tissue engineering, cells are embedded or seeded onto three-dimensional scaffolds (derived from different biomaterials) before

CHAPTER 44  Regenerative Medicine, Cell, and Gene Therapies  

transplantation. Regardless of the specific clinical application, tissueengineering strategies usually involve the utilization of combinations of biomaterials, cells, and biologically active factors. The scaffold serves many purposes, including the control of the shape and size of the engrafted tissue, the delivery of biologic signals and adequate biomechanical support to the cells, the induction of vascularization of the graft, and the protection of the cells from physical damage. Scaffolds used in tissue engineering approaches are commonly divided into two general categories: (1) cellular scaffolds that are seeded ex vivo with cells before their in vivo transplantation; and (2) acellular scaffolds that depend on cells in the recipient to repopulate the scaffold with subsequent reconstitution after transplantation. Such tissueengineered efforts have already reached proof-of-concept clinical trials. These efforts have mainly concentrated on the musculoskeletal system (bone and cartilage repair) but have also targeted other organs such as the heart and even complex organ structures such as the esophagus, trachea, and urinary bladder.

Specific Disease Applications in Cell Therapy

Although a growing number of experimental cell therapies have reached various stages of clinical trials, as yet none has become an established or approved treatment, with the aforementioned exception of hematopoietic stem cells and solid organ transplantation. Nonetheless, with the expectation of significant advances on the horizon, examples of some of the current cell therapy efforts being made in the fields of neurodegenerative disorders, heart failure, and diabetes are provided.

Neurodegenerative Disorders

The central nervous system has limited capacity for regenerating lost tissue both in slowly progressive degenerative neurologic conditions such as Parkinson disease, Alzheimer disease, and amyotrophic lateral sclerosis (ALS) and in acute injuries leading to rapid cell loss (ischemic stroke or traumatic spinal cord injury). Stem cell−based therapies are being explored as potential novel therapeutic paradigms for both acute and chronic neurodegenerative disorders.5 Consistent with the spectrum of cell therapy−related mechanistic actions described above, these procedures could potentially act through the following mechanisms: (1) cell replacement, whereby cells (precommitted to specific neuronal or glial lineages) are transplanted to replace the specific subtypes of cells that were lost (i.e., dopaminergic neurons in Parkinson disease, motor neurons in ALS, or a mixture of different neuronal and glial subtypes in other disorders); (2) trophic support, whereby the engrafted cells are used to promote the survival of affected neurons or glia or stimulate endogenous repair of the diseased central nervous system through the secretion of neurotrophic factors; and (3) modulation of the inflammatory process thought to contribute to the pathogenesis of many neurodegenerative processes. Achieving the first mechanistic goal, despite being the most attractive, is probably also the most challenging because one would need not only to derive a clinically relevant number of the specific glial or neuronal subtypes or a combination of these cells but also to deliver them to the appropriate site (either focally or diffusely throughout the brain), as well as to assure cellgraft survival, its continuous and appropriate function, and importantly, its integration with the host neuronal network. Parkinson disease (Chapter 409) involves loss of melanin-containing dopaminergic neurons within the substantia nigra pars compacta of the midbrain, coupled with accompanying depletion of striatal dopamine. This cellular loss is responsible for the major motor features of the disease. In the search for a more definitive therapy than pharmacology, early reports of cell replacement therapy suggested significant improvement in motor function after intrastriatal implantation of mesencephalic dopamine-rich tissue, obtained from aborted human fetuses aged 6 to 9 weeks. Long-term immunosuppressive treatment is essential to allow transplanted dopaminergic neurons to develop into their full functional potential despite the notion of an immunologic sanctuary within the brain. Clinical assessment standards have provided evidence of long-lived graft survival, morphologic and functional integration, and clinical benefit after therapy with cells of fetal origin that have now lasted up to 10 years or longer in some patients. Further progress, however, has been limited by lack of sufficient source tissue to treat a large number of affected patients, prohibitive variability in functional outcome, reports of serious dyskinesias in a subset of treated patients, and ethical considerations. Given the aforementioned limitations of fetal tissue engraftment, stem cell derivatives could offer a viable alternative for the treatment of Parkinson disease by either replacing the dopaminergic neurons or slowing the degeneration process and restoring the integrity of the nigrostriatal pathway


through the release of trophic factors. Importantly, dopaminergic neuroblastlike cells have been generated ex vivo from different stem cell sources, including pluripotent stem cells (hESCs and hiPSCs) by direct fibroblast reprogramming, neural stem cells and progenitors from the embryonic ventral mesencephalon, and adult neural stem cells from the subventricular zone. Preclinical engraftment studies demonstrated that such cells could survive in animal models of Parkinson disease and exert beneficial functional effects after cell maturation. Nevertheless, some properties that are fundamental for successful clinical translation have not been fully met in animal transplantation trials employing human stem cell–derived dopaminergic neurons. Additional challenges that should be addressed include development of methods to prevent the disease process from also destroying the grafted neurons (e.g., engineering the cells to secrete neurotrophic factors) and limiting graft-induced dyskinesia (e.g., by minimizing the number of serotonergic neuroblasts in the grafted tissue). Investigations of stem cell–based approaches for the treatment of other neurodegenerative diseases, including ALS, Alzheimer disease, Batten disease, stroke, and brain and spinal cord injury, are now moving from experimental animal model studies to planning of clinical trials. Recent reports have shown a major clinical benefit in animal models of directed differentiation and transplantation of hESCs and hiPSCs toward retinal pigment epithelum. Human studies with these cells were recently initiated for patients suffering from macular degeneration.

Heart Disease

Although recent studies have challenged the dogma of the heart being a completely terminally differentiating organ, the endogenous repair mechanisms of the adult heart are usually inadequate in dealing with an extensive myocardial infarction. The resulting decrease in the contractile mass, which is associated with the loss of approximately 1 billion cardiomyocytes, may lead to the development of clinical heart failure (Chapters 58 and 59). With heart failure being the leading cause of hospitalization and with the paucity of donor organs limiting the number of heart transplantations worldwide, it is not surprising that the heart has become the focus of various regenerative medicine efforts.6 The first cells that reached clinical trials for heart failure were skeletal myoblasts. Such cells could be harvested (satellite cells) in an autologous manner, expanded ex vivo, and transplanted to the heart. However, skeletal myoblasts display different physiologic properties than cardiomyocytes and cannot form electromechanical connections with host cardiac tissue. Consequentially, these clinical efforts have largely been abandoned because of lack of efficacy as well as evidence suggesting increasing arrhythmogenicity in some patients. The largest clinical experience in myocardial cell therapy comes from the use of bone marrow−derived stem cells (primarily hematopoietic stem cells and more recently also mesenchymal stem cells). The effects of delivery of such cells (mainly through the coronary circulation) were studied in thousands of patients, primarily in the setting of acute or recent myocardial infarction. These studies revealed either a neutral effect on myocardial performance or mild functional improvement. Although a recent meta-analysis of 33 randomized controlled trials studying transplantation of adult bone marrow−derived cells revealed a statistically significant improvement in left ventricular ejection fraction, this improvement was not associated with a change in mortality. A1  Bone marrow−derived stem cells are thought to exert their beneficial effects through the secretion of different growth factors rather than transforming to become new heart cells. Consequentially, a cell source that could truly re-muscularize the heart is direly needed. A potential candidate for such a task could be the recently described cardiac progenitor cells. Several reports have described cardiac progenitor cells as multipotent clonogenic cells that could be isolated based on different markers or culturing properties and potentially differentiate into cardiomyocytes and vascular cells. Such cells can be harvested from the heart (during surgery or a percutaneous cardiac catheterization biopsy approach), expanded ex vivo, and then transplanted back to the left ventricle in an autologous manner. In contrast to the aforementioned cell types, human pluripotent stem cell lines (hESCs and hiPSCs) can undoubtedly become cardiomyocytes during ex vivo differentiation. Research efforts in recent years established efficient directed differentiation systems that could give rise to clinically relevant numbers of cardiomyocytes and demonstrated the ability of the generated cells to engraft, functionally integrate with host cardiac tissue, and improve myocardial performance in animal models of myocardial infarction. Nevertheless, issues related to ethics and the allogeneic nature of the graft (hESCs),


CHAPTER 44  Regenerative Medicine, Cell, and Gene Therapies  

to the inefficient and incomplete reprogramming process (iPSCs), to the heterogeneous and relatively immature properties of the generated cardiomyocytes, to the tumorigenic risk, and to the complex regulatory and financial issues have hindered clinical development of these cells to date. Most recent efforts in the field have focused on attempting to induce mature cardiomyocytes to reenter the cell cycle (directly or after an initial dedifferentiation phase) or to convert the phenotype of nonmyocytes (fibroblasts) into cardiomyocytes. In the latter approach, recent studies have demonstrated the ability to convert the phenotype of murine fibroblasts both in vitro and in vivo into cardiomyocyte-like cells by the expression of a combination of cardiomyocyte-specific transcription factors (GATA-4, MEF-2C, and TBX-5 in one study). Although these and other efforts have the potential to augment the number of cardiomyocytes and consequentially improve contraction of the failing heart, they are still in the early phase of discovery.

Diabetes Mellitus

Successful pancreatic transplantation and improved glucocorticoid-free protocols for transplantation of islets of Langerhans have been shown not only to restore glucose control in patients with diabetes mellitus but also to prevent or even reverse some of the disease’s complications (Chapters 229). However, whole organ or islet-based transplantation approaches are limited both by immunologic rejection and by limitation of an available source of transplantable tissues. This has motivated the search for cell types that can replace (type 1 diabetes mellitus) or augment (type 2 diabetes mellitus) deficient β-cell function.7 The development of hESCs and hiPSCs, coupled with improved understanding of β-cell development, has provided a potentially unique cell source to derive β cells for transplantation therapy. β cells make an especially attractive case for cell replacement strategies because only a single cell type is missing, cell replacement does not necessarily need to be performed in the native environment (pancreas), and, theoretically, such cells could even be engrafted subcutaneously. Harnessing lessons from embryology, efficient protocols were developed to promote differentiation of pluripotent cells in vitro into precursor or early-stage β-cell phenotype. More recent efforts have moved the field even closer to clinical application by tackling the challenge of creating more mature and functional β cells. One of the problems with using β cells for cell replacement therapy is that the autoimmune destruction of endogenous β cells, which underlies the pathogenesis of type 1 diabetes, will probably also result in the destruction of the pluripotent cell−derived β cells, even when derived from an autologous (hiPSCs) source. Consequentially, significant efforts are being made to develop the biotechnologic means (encapsulation technologies) to deliver the cells in an immunoprotective environment that will prevent cell rejection but will retain the capacity of the engrafted β cells to sense glucose and to secrete insulin. Beyond the derivation of new β cells from pluripotent stem cells, progress has also been made in reprogramming closely related cell types to β cells by the overexpressing of master regulatory transcription factors. Early studies focused on the conversion of hepatocytes to β-like cells through the overexpression of PDX1, the transcription factor MAFA, and NeuroD. In vivo transdifferentiation of mouse acinar cells to β cells has been achieved by transient viral overexpression of three transcription factors (PDX1, NGN3, and MAFA), whereas overexpression of a single transcription factor, PAX4, has successfully converted murine α cells to β cells. Regenerative strategy focuses on increasing pancreatic β-cell mass by inducing the replication of existing β cells. This therapeutic approach would probably mainly target type 2 diabetic patients by decreasing the burden on existing overworked β cells but may also be beneficial for some patients with type 1 diabetes who still retain some β-cell mass. Whereas several tissues are regenerated by differentiation of tissue-specific stem cells, new pancreatic β cells are derived from the replication of existing β cells. Promising candidates for augmenting β-cell replication were recently identified and include the use of gluco*kinase activators or betatrophin, a protein secreted by the liver.

Stem Cell–Derived Platforms for Disease Modeling, Personalized Medicine, and Drug Discovery

In addition to the generation of cells for regenerative applications, the ability to grow a wide variety of different specialized cell types of human origin in culture provides unparalleled opportunities for gene and drug discovery and testing. For example, the ability to grow human cardiomyocytes in culture provides a preclinical human cellular-based experimental platform for

screening newly developed drugs in terms of their potential to cause QT-interval prolongation and hence the risk for arrhythmia in the clinical setting. Other examples include the creation of an experimental tissue microenvironment of human origin for studying the stromal response to tumor growth and testing anticancer drugs that target tumorigenic responses such as angiogenesis.8 The hiPSC technology has further revolutionized this field because it allows for the first time the generation of disease/genotype- and patientspecific hiPSC models of a wide array of inherited disorders. Initial studies focused on diseases with monogenic inheritance, but more recent studies have included diseases with more complex inheritance patterns.9 Consequently, different types of patient-specific hiPSC-derived neurons, cardiomyocytes, skeletal muscle, blood cells, hepatocytes, and other cell types were demonstrated to recapitulate in a culture dish the abnormal phenotype of a wide array of genetic disorders, including neurodegenerative disorders (e.g., spinal muscular atrophy, familial dysautonomia, ALS, schizophrenia, and even late-onset diseases such as Parkinson and Alzheimer disease), different cardiomyopathies and arrhythmogenic syndromes, a wide array of blood disorders, and several other genetic disorders. These models have already yielded important insights into the mechanisms underlying these disease states and have established unique experimental platforms that will enable the testing of existing therapies in a patient-specific manner (personalized medicine) to evaluate evolving therapies (“clinical studies in the culture dish”) and to develop new therapeutic strategies.


Gene therapy can be broadly defined as the transfer of genetic material into cells to restore or correct a cellular dysfunction or to provide a new cellular function in an attempt to cure a disease or at least to improve the clinical status of a patient. The use of genes as therapeutic platforms emerged during the mid-20th century, and in the 1990s, the first regulated registered studies were performed in the United States. In the first clinical study, a 4-year-old girl with adenosine deaminase (ADA) deficiency was treated by transfecting the ADA gene into her white blood cells, resulting in improvements in her immune system. Since then, more than 10,000 patients have been involved in more than 1700 gene therapy clinical studies performed throughout the world. The most common patient populations targeted in these studies have been cancer patients (more than 1000 studies), with another important category being monogenic inherited disorders (more than 100 studies). Although gene therapy initially was conceived as a way to treat life-threatening disorders (inborn errors, cancers) refractory to conventional treatment, it is now being explored for non-life-threatening conditions that adversely affect a patient’s quality of life. Although early clinical failures and a number of reported deaths (only two of which were actually attributed directly to gene therapy) and cases of gene therapy−related leukemic transformation led many to dismiss gene therapy as hazardous and premature, recent clinical successes have bolstered new optimism in the promise of this discipline. These include entirely novel initiatives in treating primary immunodeficiency syndromes, the improvement of vision in patients with the retinal disease (e.g., Leber congenital amaurosis), the successful treatment of X-linked adrenoleukodystrophy,10 and the encouragement of experimental results in treating different forms of cancer. Despite these success stories, only a few gene therapy agents are currently approved and available. Fomivirsen (Vitravene) is used for the treatment of cytomegalovirus retinitis in patients with AIDS. In 2012, Glybera became the first gene therapy treatment to be approved for clinical use in either Europe or the United States. Glybera uses a virus injected into a patient to deliver a working copy of a gene for producing lipoprotein lipase (LPL) to treat the rare inherited disorder of LPL deficiency. Finally, the p53 tumor suppressor coding sequence in an adenovirus vector is used for the treatment of head and neck cancer but is registered only in China.

Classifications and Mechanisms of Action

In general, somatic gene therapy applications can be divided into those aiming to treat or correct various genetic disorders and those targeting nongenetic diseases by attempting to alter cell, tissue, and organ function in a favorable manner. According to the World Health Organization, there are more than 10,000 disorders with monogenic inheritance described in humans (, but only a small fraction of these may be amenable to gene therapy. Traditional gene therapy efforts for inherited disorders have mainly focused on the

CHAPTER 44  Regenerative Medicine, Cell, and Gene Therapies  

exogenous expression of genes encoding the missing or abnormal proteins and to a lesser extent also on altering the abnormal gene expression patterns. Future efforts are expected to shift the focus from uncontrolled overexpression of the missing protein to directly correcting the mutation at the DNA level (gene-editing strategies) in affected cells, using the newly emerging technologies known as the TALEN and CRISPR11 approaches described in greater detail later under Gene Editing. With successful widespread use in research studies and proven applications in editing of gene sequence in stem cells, the transition to clinical application is sure to follow. For nongenetic disorders, gene therapy efforts are aimed at overexpressing a specific protein in an attempt to alter cellular function favorably (e.g., to increase contractility in heart failure by overexpression of the sarcoplasmic reticulum calcium ATPase SERCA2a), protect tissue at risk (e.g., in acute kidney injury), exert paracrine effects through local secretion of specific proteins by the engineered cells (e.g., promote angiogenesis in ischemic tissues or induce neurotrophic effects in neurodegenerative disorders), and even secrete proteins systemically (e.g., in gene therapy trials attempting to correct bleeding disorders by secretion of coagulation factors or for systemic delivery of hormones such as erythropoietin for the treatment of anemia). Major efforts in the gene therapy arena to date have been in designing various methods to treat cancer (see later). Progress in the field of gene therapy has developed into two different strategies: ex vivo and in vivo gene therapy. The ex vivo gene therapy approach (combined cell and gene therapy strategy) involves the initial harvesting of cells from a given patient followed by genetic modification of these cells in the laboratory. The genetically modified cells can then be selected, amplified in numbers, and returned to the same patient in an attempt to achieve the desired therapeutic effect. This strategy is particularly attractive for the genetic modification of stem cells that could reconstitute the relevant tissues, organs, and organ systems after transplantation. The most prominent example is using hematopoietic stem cell grafts in gene therapy trials for hematopoietic disorders. The in vivo gene therapy approach, in contrast, involves the delivery of the relevant transgene (through the use of various vectors) directly to the targeted tissue, followed by the stable or transient expression of the transgene in the relevant cells. The expression of the transgene only in the relevant cells/ tissues can be achieved by a combination of localized delivery (injection), a particular tropism of the vector used for the tissue of interest, and the expression of the transgene under the control of a cell/tissue-specific promoter.

Gene Therapy Delivery Methods

Gene therapy agents are often composed of two elements: the genetic material itself (i.e., the DNA expression cassette [the most common therapeutic payload used], short interfering RNA, or an antisense molecule) and the vector delivery system. The latter is usually the more complex and limiting component, and it is important to select the most efficient delivery method for any genetic therapy as well as to be aware of the potential adverse effects of each vector type, thus tailoring the therapy to specific clinical considerations. There are formidable barriers to successful gene transfer, such as crossing the cellular membrane, escaping from the endosome, moving through the nuclear membrane, and integrating into the host genome. Vectors that have been developed to try to overcome these obstacles fall into two broad categories: nonviral and viral vectors. Gene therapy mediated by nonviral vectors is referred to as transfection and consists of the direct delivery of naked DNA by injection, the use of liposomes (cationic lipids mixed with nucleic acids), nanoparticles, and other means. Although nonviral vectors can be produced in relatively large amounts and are likely to present minimal toxic or immunologic problems, their major shortcoming is inefficient gene transfer. In addition, expression of the foreign gene tends to be transient, precluding the application of nonviral vectors to many disease states in which sustained and high-level expression of the transgene is required. The efficiency of nonviral vector delivery could be enhanced by the use of different physical methods that have evolved, such as electroporation (for well-circ*mscribed body compartments or masses such as muscle, skin, and tumors), gene gun (for DNA vaccination), and ultrasound delivery (for cardiovascular and tumor-related applications). Gene therapy mediated by viral vectors is referred to as transduction, and this approach has been the main conduit for transferring genes to human cells in most gene therapy trials. The basic concept of viral vectors is to harness the innate ability of viruses to deliver genetic material into the infected cell. Viruses used in gene therapy have been modified to enhance safety, increase


specific uptake, and improve efficiency. However, for each specific virusbased gene therapy vector, there have been major disadvantages that should be balanced against potential therapeutic benefits. For example, in cancer gene therapy, the immune response to the delivery vehicle carrying the anticancer genetic material can be used to advantage by serving as an adjuvant. In contrast, the system for delivery of a gene to be expressed for a prolonged period to replace or supplement a missing gene product in monogenic disease states should preferably be ignored by the immune system. Viral vectors are derived from viruses with either RNA (retroviruses and lentiviruses) or DNA (adenovirus, adeno-associated virus [AAV], herpes simplex virus [HSV], and poxvirus [vaccine virus]) genomes. Viral vectors also fall into one of two main categories: integrating vectors, which insert themselves into the recipient’s genome, and nonintegrating vectors, which often (although not always) form an extrachromosomal genetic element. Integrating vectors, such as γ-retroviral vectors and lentiviral vectors, are generally used to transfect actively dividing cells because they are stably inherited. Integrating vectors, however, may carry the risk for insertional mutagenesis (with clinical oncogenic transformations reported with the use of retroviruses). Nonintegrating vectors, such as adenoviral vectors and AAV vectors, can be used to transfect quiescent or slowly dividing cells, but they are quickly (in the case of adenoviral vectors) lost from cells that divide rapidly. Finally, efficient gene transduction can also be achieved using vectors that are maintained as episomes, especially in nondividing cells. Adenoviral vectors and retroviral vectors based on Moloney murine leukemia virus featured prominently in early gene therapy trials. There has been a movement away from both, however, after the case of a fatality, which was linked to the toxicity of the adenoviral vector (used to introduce the ornithine transcarbamylase gene in that specific study) and the leukemia cases in SCID-X1 patients (which were linked with activation of LMO2, an oncogene on chromosome 11, due to insertional mutagenesis associated with the murine leukemia viral vector). Consequentially, these vectors have been largely been replaced with AAV and lentiviral vectors, respectively, which have become the most common vectors used in clinical trials today. Other viral vectors may have applications in specific settings. For example, in gene therapy applications being developed for pain management, a replicationdefective HSV vector is being used because of its tropism for nerve tissues. Also, different oncolytic viruses with a preferential tropism to cancer cells are being used for gene therapy applications in cancer.

Diseases Treated by Gene Therapy Inherited Immunodeficiency

More than 30 patients reported to date worldwide have undergone treatment with different retroviral vectors for inherited immunodeficiencies (Chapter 250). Patients with one of the following three diseases are included in this group: two types of severe combined immunodeficiency (SCID), both of which are characterized by dysregulation of lymphocyte development, and X-linked chronic granulomatous disease (X-CGD), an inherited immune deficiency with absent phagocyte reduced nicotinamide adenine diphosphate oxidase activity caused by mutations in the gp91 (phox) gene. Individuals with adenosine deaminase (ADA) SCID suffer from premature death of T, B, and natural killer (NK) cells as a result of the accumulation of purine metabolites; patients with this condition have been treated with vectors expressing the ADA gene. In the first patients with ADA SCID, transduced T cells expressing transgenic ADA have been shown to persist for longer than 10 years; however, the therapeutic effect of gene therapy resulted in incomplete correction of the metabolic defect. More recently, an improved gene transfer protocol of bone marrow CD34-positive cells, combined with lowdose busulfan, resulted in multilineage, stable engraftment of transduced progenitors at substantial levels, restoration of immune function, correction of the ADA metabolic defect, and proven clinical benefit.12 Overall, no adverse effect or toxicity has been observed in patients treated with ADA gene transfer in mature lymphocytes or hematopoietic progenitors. The X-linked type (X-SCID group), in which there is defective cytokinedependent survival signaling in T and NK cells, was shown to be corrected by introduction of the wild-type sequence of the common γ-C chain, which is an essential component of five cytokine receptors. In one clinical study, hematologic malignancies developed in four patients. One of the four died of this complication. Ten patients were successfully treated with a different viral transduction protocol, with one reported malignancy in up to 8 years of follow-up. Two adult X-CGD patients who suffered recurrent bacterial infections have been treated with CD34-positive cells transduced with a γ-retroviral vector expressing gp91 phox, with significant clinical improvement in the


CHAPTER 44  Regenerative Medicine, Cell, and Gene Therapies  

short term. However, in both these patients, there was an expansion of genetransduced cells caused by the transcriptional activation of growth-promoting genes leading to myelodysplasia and gradual loss of efficacy. In summary, of the nearly 30 patients worldwide treated with gene therapy for immunodeficiency disorders, significant clinical improvement has been observed in many. However, severe and even life-endangering adverse consequences have been encountered with certain viral vectors and protocols. Additional clinical information from long-term observation and new clinical studies will be important for a clearer assessment of clinical benefit.13

that deliver a copy of the CFTR gene to the airway of CF patients have been developed. Several placebo-controlled clinical trials of liposome-mediated CFTR gene transfer to the nasal epithelium have confirmed its safety and demonstrated variable degrees of functional correction. In addition, several clinical studies have assessed the potential of retrovectors, adenovectors, and AAV vectors for gene therapy for CF. With both nonviral and viral delivery systems, there were only mild side effects. However, the long-term clinical benefit has been marginal. Improved vectors are being assessed in preclinical studies.

Visual Loss


Both cell- and gene-based therapy approaches are leading areas for promising inroads in retinal disease. Although early trials of stem cell−based retinal cell therapy have not yet achieved proof of efficacy, at the level of gene therapy, clinical scientists have used gene augmentation therapy with direct subretinal injection of a recombinant AAV expressing RPE65 complementary DNA in adults and children with Leber congenital amaurosis. This rare inherited eye disease destroys photoreceptors (Chapter 424), and the gene therapy results have shown medical evidence of visual preservation despite continued retinal degeneration.14

Cardiovascular and Pulmonary Conditions

Gene therapy efforts in the cardiovascular field have focused on achieving therapeutic angiogenesis in patients suffering from chronic ischemic heart disease A2  A3  or from critical limb ischemia (CLI) and for improving cardiac function in heart failure patients. The use of genes to revascularize the ischemic myocardium due to coronary artery disease and CLI due to peripheral artery disease has been the focus of two decades of preclinical research with a variety of angiogenic mediators, including vascular endothelial growth factor, fibroblast growth factor, hepatocyte growth factor, and others, encoded by DNA plasmids or adenovirus vectors. Overall, these gene therapy studies in animal experimental models of ischemia were very encouraging, leading eventually to several clinical trials. Despite the established proof of concept and reasonable safety, however, results of the latest clinical trials on therapeutic angiogenesis for myocardial ischemia and CLI have provided inconsistent results, and the definite means of inducing clinically useful therapeutic angiogenesis remain elusive. These less than optimal results may stem from a number of reasons, including the application of a single growth factor that may not be sufficient to meet the multifaceted challenge for developing efficient induction of collateral vessels, the need for more sustained growth factor delivery in order to establish more stable vessels, and the need to target arteriogenesis rather than angiogenesis to achieve a more significant increase in perfusion. Therefore, efforts in the field are moving toward the use of different cell therapies for these ischemic conditions, as well as using combined cell and gene delivery strategies to achieve better outcomes. For example, a recent trial has used combined delivery of endothelial and smooth muscle cells (each cell type modified to secrete a different angiogenic growth factor) in CLI patients. For heart failure, gene therapy trials have focused on restoring the abnormal calcium handling characteristic of failing human cardiomyocytes.15 Because a reduction in levels of the sarcoplasmic reticulum calcium ATPase (SERCA2a), the sarcoplasmic reticulum calcium pump, was found to be a key factor in the alteration of calcium cycling in heart failure, this protein became an attractive clinical target for gene delivery purposes. Overexpression of SERCA2a levels by cardiomyocyte gene delivery has led to the restoration of previously abnormal calcium transients and to improved cardiac contractility, reduction of the frequency of arrhythmias, and improved oxygen utilization in animal models of heart failure. More recently, the clinical benefits of overexpressing SERCA2a have been demonstrated in phase I and II of the Calcium Upregulation by Percutaneous Administration of Gene Therapy in Cardiac Disease (CUPID) trials. A4  These studies demonstrated that AAV delivery of the SERCA2a transgene by intracoronary delivery is feasible and safe, results in persistent expression of the transgene, and is associated with a significant improvement in associated biochemical alterations and clinical symptoms of heart failure in the treated patients. ,

Cystic Fibrosis

Experimental protocols for gene therapy for cystic fibrosis (CF) (Chapter 89) have been implemented since 1990. The cystic fibrosis transmembrane conductance regulator (CFTR) protein is mutated in patients with CF. Transducing the epithelium of the nasal and bronchial tree is potentially feasible through nonsystemic approaches. Nonviral gene therapy methods

One of the most exciting opportunities for gene therapy lies in the cancer arena. Gene therapy strategies targeting cancer can be grouped according to their proposed mechanisms of action and include gene therapies aiming to directly induce cytotoxic effects in cancer cells (through the use of oncolytic viruses or by the delivery of apoptotic inducers and suicide genes), gene therapies aiming to boost the immune response to tumor antigens, and gene therapies targeting the tumor microenvironment.

Direct Cytotoxic Effects

An interesting approach for cancer gene therapy is to harness the action of oncolytic viruses.16 Oncolytic viruses are therapeutically useful anticancer viruses that will selectively infect, amplify, and then damage cancerous tissues without causing harm to normal tissues. Cancer selectivity of the different oncolytic viruses takes advantage of defects commonly found across many tumor types, such as lack of antiviral responses, activation of Ras pathways, loss of tumor suppressors, and defective apoptosis. Oncolytic viruses can kill infected cancer cells in many different ways, ranging from direct virus-mediated cytotoxicity through a variety of cytotoxic immune effector mechanisms. Several viruses such as the Newcastle disease virus (which activates the innate or adaptive immune response), reovirus (which activates host protein kinases to shut down protein production), and mumps virus have an inherent ability to specifically target cancer cells and, upon virus replication, cause significant cell death and tumor regression. Other viruses (HSV, adenovirus, vaccinia virus, vesicular stomatitis virus, and poliovirus) need to be genetically engineered to engender oncolytic activity. Genetically engineered viruses and inherently antitumor-selective viruses are being tested in early and late clinical conditions to determine their effectiveness in specific types of cancer (e.g., metastatic melanoma and different brain tumors). Beyond the direct viral cytopathic effect, viral vectors can be used to deliver genes to cancer cells that will result in tumor cell death. The relevant transgenes encode for cellular proteins that are involved in apoptosis or prevent proliferation. The selectivity for the activation of such genes only in tumor cells is achieved either through the use of the aforementioned oncolytic viruses or by the expression of the transgenes under the control of promoters that are activated only in cancer cells, either as a general property of cancer (e.g., human telomerase or survivin) or in specific types of tumors (probasin in prostate cancer, ceruloplasmin in ovarian cancer, HER2 in breast cancer, and carcinoembryonic antigen in colon cancer). The most clinically advanced gene therapy drug against cancer is the replication-deficient adenovector expressing the human p53 gene. This therapy (Gendicine) is approved in China for the treatment of patients with head and neck squamous cell carcinoma by direct administration into the tumor bed. Another attractive approach is the use of suicide genes. Suicide gene therapy involves delivery of a pro-drug activating enzyme (suicide gene) that converts nontoxic pro-drugs to cytotoxic metabolites. The prototype for such a suicide gene/pro-drug combination is HSV thymidine kinase (TK)/ganciclovir (GCV). The TK gene is selectively expressed only in cancer cells (by one of the methods described previously), and after application of GCV, it converts it to the cytotoxic agent phosphorylated GCV. Interestingly, phosphorylated GCV is only toxic to dividing cells, further increasing the selectivity to the cancer cells. Other cytotoxic strategies are to express secreted pro-apoptotic proteins, such as tumor necrosis factor−related apoptosisinducing ligand (TRAIL) or cytotoxins such as Pseudomonas exotoxin.

Immunomodulatory Cell and Gene Therapy for Cancer and Autoimmune Disease

In recent years, the focus of gene- and cell-based therapy for cancer has shifted away from directly manipulating or targeting the cancer cells toward modulation of the immune system itself. Cytotoxic T-lymphocyte antigen 4 (CTLA-4) and programmed death 1 (PD-1) are two T-lymphocyte

CHAPTER 44  Regenerative Medicine, Cell, and Gene Therapies  

proteins that have long been known to attenuate immune destruction of cancer cells. Blocking monoclonal antibodies to circumvent this attenuation have been shown to induce limited remissions in several forms of previously intractable metastatic tumors, including malignant melanoma. These partial successes have now revived still more sophisticated therapeutic approaches, based on the ex vivo personalized genetic engineering of cytotoxic T lymphocytes of cancer patients to enable the immune system to target tumor cells. Chimeric antigen receptor (CAR) therapy releases the encumbrance of major histocompatibility complex restriction in cancer antigen recognition by combining the antigen-binding site of a monoclonal antibody with the signal-activating machinery of the cytotoxic T lymphocytes. This enables combining a high level of target specificity typical of monoclonal antibodies with in vivo expansion and the potential for a durable response, as has been demonstrated in clinical treatment protocols in leukemia and other malignancies.17 It can be anticipated that CAR-modified T lymphocytes might also prove useful as a combined genetic engineering/cell therapy approach to the management of autoimmune disease.

Disrupting Tumor Microenvironment

Targeting the tumor microenvironment is another attractive approach for cancer gene therapy because it consists of normal cells that should not develop resistance to the therapy. The most obvious target is the tumor neovascularization process. The use of antiangiogenic drugs such as bevacizumab (Avastin), an anti−vascular endothelial growth factor monoclonal antibody, has shown success in clinical trials for some cancer cell types, but the effect may be transient or negligible in others. This may be because the angiogenesis process is complex, and inhibiting just one aspect may not be sufficient. Developing alternative strategies such as combination therapies, including targeting multiple angiogenic pathways, might be a better strategy, especially because inhibiting angiogenesis is cytostatic and not cytotoxic. A number of antiangiogenic factors (e.g., angiostatin) have been expressed in viral vectors and have been used in preclinical studies but have not reached the clinic yet.

Other Forms of Molecular Therapies: RNA Interference and Gene Editing RNA Interference

RNA interference (RNAi) regulates gene expression by a highly precise mechanism of sequence-directed gene silencing at the stage of translation by degrading specific messenger RNAs or by blocking their translation into protein. Research on the use of RNAi for therapeutic applications has gained considerable momentum. It has been suggested that many of the novel disease-associated targets that have been identified are amenable to conventional small molecule drug blockade and can potentially be targeted with RNAi. In the coming years, the concept of RNAi will be actively translated into a therapeutic option, with numerous early-phase trials already underway.

Gene Editing

The center of gravity for gene therapy may be shifting from gene restoration (where a whole new gene is pasted into the genome) to genome editing, whereby the pathogenic mutation is corrected in its natural gene location with zinc finger nucleases, transcription activator−like effector nucleases (TALENs), or clustered regulatory interspaced short palindromic repeats


(CRISPRs).18,19 These hybrid molecules act as highly specific “molecular scissors,” which are engineered to target a specific location in the genome and introduce a double-strand break in the DNA proximal to the targeted mutation. The cleavage in the DNA is then resolved by hom*ologous recombination between the endogenous genes and an exogenously introduced donor fragment containing the normal sequence. In this fashion, the pathogenic mutation is permanently changed back to the normal sequence. This also preserves the architecture of the genome and maintains gene control under the normal cellular regulatory elements. Consequentially, gene editing represents a paradigm shift in the way gene therapy could be performed. To date, gene editing techniques have been used to correct the disease-causing mutations associated with X-linked SCID, hemophilia B, sickle cell disease,20 and α1-antitrypsin deficiency and to repair Parkinson disease−associated mutations (SNCA gene) in patient-derived hiPSCs or in preclinical mouse models. Targeted gene knockout through similar technologies promises to be a potentially powerful strategy for combating HIV/AIDs. Zinc finger nucleases have been used to confer HIV-1 resistance by disabling the HIV coreceptor C-C chemokine receptor type 5 (CCR5) in primary T cells and hematopoietic stem/progenitor cells. This approach is currently used in clinical trials. Additionally, zinc finger nucleases have been used to improve the performance of T-cell-based immunotherapies by inactivating the expression of endogenous T-cell-receptor genes, thereby enabling the generation of tumor-specific T cells with improved efficacy profiles. Finally, site-specific nucleases may also bring a unique value to the conventional gene-adding approach by enabling insertion of therapeutic transgenes into specific “safe harbor” locations in the human genome, ensuring longterm expression of the transgene as well as reducing the potential for random insertional mutagenesis. It is important to mention that the use of site-specific nuclease technology at its current state requires the presence of proliferating cells, and its utility is therefore still relatively limited for nonproliferating somatic cells and for direct in vivo applications. Continued progress in stem cell research, including the production and manipulation of hiPSCs cells, will ultimately open countless new directions for gene therapy, including treatments based on autologous stem cell transplantation.

Grade A References A1. Clifford DM, Fisher SA, Brunskill SJ, et al. Stem cell treatment for acute myocardial infarction. Cochrane Database Syst Rev. 2012;2:CD006536. A2. Fisher SA, Brunskill SJ, Doree C, et al. Stem cell therapy for chronic ischaemic heart disease and congestive heart failure. Cochrane Database Syst Rev. 2014;4:CD007888. A3. Wang ZX, Li D, Cao JX, et al. Efficacy of autologous bone marrow mononuclear cell therapy in patients with peripheral arterial disease. J Atheroscler Thromb. 2014;21:1183-1196. A4. Zsebo K, Yaroshinsky A, Rudy JJ, et al. Long-term effects of AAV1/SERCA2a gene transfer in patients with severe heart failure: analysis of recurrent cardiovascular events and mortality. Circ Res. 2014;114:101-108.

GENERAL REFERENCES For the General References and other additional features, please visit Expert Consult at

CHAPTER 44  Regenerative Medicine, Cell, and Gene Therapies  

GENERAL REFERENCES 1. Daley GQ. The promise and perils of stem cell therapeutics. Cell Stem Cell. 2012;10:740-749. 2. Takahashi K, Yamanaka S. Induced pluripotent stem cells in medicine and biology. Development. 2013;140:2457-2461. 3. Atala A, Kasper FK, Mikos AG. Engineering complex tissues. Sci Transl Med. 2012;4:160rv112. 4. Doulatov S, Daley GQ. Development. A stem cell perspective on cellular engineering. Science. 2013;342:700-702. 5. Lindvall O, Barker RA, Brustle O, et al. Clinical translation of stem cells in neurodegenerative disorders. Cell Stem Cell. 2012;10:151-155. 6. Xin M, Olson EN, Bassel-Duby R. Mending broken hearts: cardiac development as a basis for adult heart regeneration and repair. Nat Rev Mol Cell Biol. 2013;14:529-541. 7. Pagliuca FW, Melton DA. How to make a functional beta-cell. Development. 2013;140:2472-2483. 8. Abelson S, Shamai Y, Berger L, et al. Intratumoral heterogeneity in the self-renewal and tumorigenic differentiation of ovarian cancer. Stem Cells. 2012;30:415-424. 9. Imaizumi Y, Okano H. Modeling human neurological disorders with induced pluripotent stem cells. J Neurochem. 2014;129:388-399. 10. Cartier N, Hacein-Bey-Abina S, Bartholomae CC, et al. Hematopoietic stem cell gene therapy with a lentiviral vector in X-linked adrenoleukodystrophy. Science. 2009;326:818-823. 11. Sheridan C. Gene therapy finds its niche. Nat Biotechnol. 2011;29:121-128.


12. Candotti F, Shaw KL, Muul L, et al. Gene therapy for adenosine deaminase-deficient severe combined immune deficiency: clinical comparison of retroviral vectors and treatment plans. Blood. 2012;120:3635-3646. 13. Zhang L, Thrasher AJ, Gaspar HB. Current progress on gene therapy for primary immunodeficiencies. Gene Ther. 2013;20:963-969. 14. Cideciyan AV, Jacobson SG, Beltran WA, et al. Human retinal gene therapy for Leber congenital amaurosis shows advancing retinal degeneration despite enduring visual improvement. Proc Natl Acad Sci U S A. 2013;110:E517-E525. 15. Tilemann L, Ishikawa K, Weber T, et al. Gene therapy for heart failure. Circ Res. 2012;110: 777-793. 16. Russell SJ, Peng KW, Bell JC. Oncolytic virotherapy. Nat Biotechnol. 2012;30:658-670. 17. Porter DL, Levine BL, Kalos M, et al. Chimeric antigen receptor-modified T cells in chronic lymphoid leukemia. N Engl J Med. 2011;365:725-733. 18. Gaj T, Gersbach CA, Barbas CF 3rd. ZFN, TALEN, and CRISPR/Cas-based methods for genome engineering. Trends Biotechnol. 2013;31:397-405. 19. Tebas P, Stein D, Tang WW, et al. Gene editing of CCR5 in autologous CD4 T cells of persons infected with HIV. N Engl J Med. 2014;370:901-910. 20. Romero Z, Urbinati F, Geiger S, et al. beta-globin gene transfer to human bone marrow for sickle cell disease. J Clin Invest. 2013;123:3317-3330.


CHAPTER 44  Regenerative Medicine, Cell, and Gene Therapies  

REVIEW QUESTIONS 1. Meniscus repair by constructing an implantable scaffold seeded with mesenchymal stem cells that differentiate into chondrocytes would be an example of which of the following? A. Ex vivo gene therapy B. Autologous tissue meniscus implantation C. Tissue engineering D. Direct stem cell reprogramming E. Somatic gene therapy Answer: C  Rather than simply introducing cells into a diseased area, in tissue engineering, cells are embedded or seeded onto three-dimensional scaffolds (derived from different biomaterials) before transplantation. This can involve either (i) cellular scaffolds that are seeded ex vivo with cells before their in vivo transplantation, or (ii) acellular scaffolds that require the recipient’s cells to repopulate the scaffold to reconstitute it after transplantation. To date, these efforts have mainly concentrated on the musculoskeletal system, as in the example presented. Ex vivo and somatic gene therapies by definition involve the transfer of specific genetic material into cells to correct or restore a cellular defect. Direct stem cell programming is the direct conversion of the phenotype of one cell type (e.g., fibroblasts) to another (e.g., chondrocytes). Autologous tissue meniscus implantation does not involve any type of cell and gene therapy. (See Tissue Engineering.) 2. To date, the only established clinical application of fetal-derived cell therapy has been in patients with which of the following? A. Parkinson disease B. Myocardial infarction C. Heart failure D. Blood product transfusion E. Cartilage repair Answer: A  Cells of fetal origin show enhanced proliferative capacity and enhanced ability to differentiate into mature or specialized cells. Although this form of cell therapy represents a form of regenerative medicine with great potential, the only fetal-derived cells that have been used in clinical applications to date are the dopaminergic cells derived from the developing fetal nervous system for the treatment of Parkinson disease. (See Engraftment of Fetal Tissue.)

3. Which one of the following in not a form of gene therapy? A. Transfection B. Transduction C. RNA interference D. Gene editing E. DNA electroporation Answer: E  Electroporation is a strictly in vitro research method in cell biology that electropermeabilizes cell membranes by an externally applied electrical field to introduce a piece of DNA (or other agents) into a cell. All the other choices are methods of gene therapy. Transfection is the direct delivery of naked DNA by injection, the use of liposomes, nanoparticles, and other means. Transduction is gene therapy that is mediated by viral vectors. RNA interference is a highly precise mechanism of regulating gene expression by sequence-directed silencing (by degrading specific messenger RNAs or by blocking their translation into protein). Gene editing is a way of correcting a pathogenic mutation in its natural gene location (rather than the more conventional gene therapy method of gene restoration, in which a whole gene is inserted into the genome). (See Gene Therapy Delivery Methods and Other Forms of Molecular Therapies: RNA Interference and Gene Editing under the main heading of Gene Therapy.)


CHAPTER 45  The Innate Immune System  



The immune system, comprising cells, the molecules they produce, and the organs that organize those components, evolved over millions of years in response to infections with pathogenic microorganisms.1 Its essential role in maintaining health is based on its recognition and elimination or control of those foreign microbes. Central to the success of the protective role of the immune system is its capacity to distinguish foreign and dangerous invaders from self-components.2,3 In addition to its contributions to host defense, the immune system is involved in the prevention of malignancy by surveying and recognizing self-cells that express novel antigens,4 and it also plays a role in resolution and repair of tissue damage. The immune system is generally described as including an innate immune system and an adaptive immune system. The former provides the first and rapid line of defense and cellular response to a foreign stimulus. The latter, dependent on activation by the innate immune response, develops a more specific response targeted to the offending organism and generates memory for that stimulus that can be elicited rapidly should that organism be encountered again on a later occasion. Immune system cells derive from precursor cells of the hematopoietic lineage and populate discrete lymphoid organs, including lymph nodes, spleen, and thymus, as well as skin and intestine. Cells of the innate immune system serve as sentinels at locations that are likely to encounter foreign organisms, and after activation they will often travel to a local lymphoid organ. The induction of the adaptive immune response occurs in the context of structured aggregates of innate and adaptive immune cells in the lymphoid organs. Once activated and differentiated to produce effector molecules, immune system cells can be sampled in blood as they travel to sites of infection or tissue damage. There they can interact directly with target cells to mediate cell death or, alternatively, provide activating signals to expand or regulate a response, or secrete high local levels of immunomodulatory substances called cytokines. Cytokines are small soluble proteins that communicate among cells within the immune system or between immune system cells and cells in other tissues.5 The cells and products of the immune system function as an exquisitely regulated complex system.6 Inherited variations in hundreds of genes have evolved, under pressure of microbial challenge, to ensure adequate defense against pathogenic organisms across the human population.1 However, in any one individual, the composite genetic profile can generate predisposition to infection or, alternatively, autoimmune or inflammatory disease. The innate immune response was traditionally viewed as mediating nonspecific protection through the production of preformed effector molecules. However, important advances in characterization of the cell surface and intracellular pattern recognition receptors (PRR), particularly the toll-like receptor (TLR) family, and signaling pathways used by innate immune cells to implement a defensive response are now understood to have relative specificity for pathogen-associated molecular patterns (PAMPs) that are characteristic of categories of microbes.7,8 In contrast to those receptor systems that initiate an innate immune response, the protein products that implement the response, whether to expand the reaction to additional cells, promote trafficking to the most relevant location, or shape the differentiation programs of adaptive immune system cells, do not show specificity based on the initial triggering stimulus. The products of the innate immune response can be highly effective at ablating or limiting the extent of infection and can generate a tissue repair program that establishes a satisfactory resolution of the episode of infection. However, when sustained or poorly regulated, they can represent an important pathophysiologic mechanism for many autoimmune and inflammatory diseases.

Cells of the Innate Immune System Monocytes and Macrophages

Monocytes circulate in the peripheral blood with a half-life of 1 to 3 days. Macrophages arise from monocytes that have migrated out of the circulation

and have proliferated and differentiated in tissue. Tissue macrophages include alveolar macrophages in the lung, Kupffer cells in the liver, osteoclasts in bone, microglia in the central nervous system, and type A synoviocytes in the synovial membrane. Macrophages secrete myriad products, including hydrolytic enzymes, reactive oxygen species, cytokines, and chemokines. Macrophages engulf microorganisms and foreign particles directly or are activated by protein complexes containing antibodies that bind to cell surface receptors for the Fc portion of immunoglobulin molecules (Fc receptors, or FcRs). These encounters activate intracellular signaling pathways that induce transcription of target genes, primarily those encoding mediators that promote inflammation or enzyme-mediated death of the microbe. Cytokines from other immune system cells, including interferon (IFN)-γ or interleukin (IL)-4, can drive macrophage differentiation toward the production of mediators that are primarily pro-inflammatory or to a wound healing functional profile. Researchers have characterized those functional phenotypes as M1 or M2, although it is recognized that the context of an innate immune response will determine the functional response, with composite profiles common.9 In addition to responding to foreign microbes, macrophages contribute to the elimination of senescent or apoptotic cells in a manner that avoids induction of an inflammatory response. Macrophages also interact with other cell types through complementary cell surface adhesion or costimulatory receptors. After capturing antigen, they can function as antigen-presenting cells for T lymphocytes, and they can interact with non−immune system cells such as endothelial cells or fibroblasts.

Dendritic Cells

Dendritic cells (DCs) comprise a complex family of cells that perform essential functions in the innate immune response and serve as a bridge to activation of an adaptive immune response. Myeloid dendritic cells can incorporate antigens derived from invading microbes, travel to nearby lymph nodes, and present processed antigenic peptides to T lymphocytes (T cells) in the form of peptide−major histocompatibility complex (MHC) molecule complexes. They are the most effective antigen-presenting cells based on expression of cell surface costimulatory molecules, and they produce cytokines, including IL-12 and IL-23, after interaction with PAMPs. They thereby contribute to the shaping of the T-cell differentiation program to generate effector cell functions. Plasmacytoid dendritic cells (pDCs) have been identified as highly effective producers of type I interferon, a key mediator of host defense against viral infections.

Natural Killer Cells

Natural killer (NK) and NK T cells provide early defense against viral infections and other intracellular pathogens while adaptive responses are developing.10 NK cells are sensitized by cytokines, including type I interferons, released from pDCs and macrophages, and secrete abundant IFN-γ, which activates macrophages and other cells. They also are poised to kill virusinfected cells by injecting pore-forming enzymes and granzymes. Activation of NK cells is inhibited by interaction with self-MHC class I molecules on target cells. When those self-histocompatibility antigens are not present, NK cell−mediated killing is implemented. NK cells are important in tumor surveillance because they are able to kill MHC class I–deficient tumor cells that are no longer susceptible to adaptive immune responses. In addition to NK cells, a type of lymphocytes, so-called innate lymphoid cells, which participate early in innate immune responses but do not express rearranged receptors, is a focus of current study.11


Neutrophils are the most abundant circulating white blood cells. They are recruited rapidly to inflammatory sites and can phagocytose and digest microbes (Chapters 167 and 169). Activation of neutrophils and phagocytosis is facilitated through the triggering of FcRs or complement receptors. Microbe-containing phagosomes fuse with lysosomes, which contain enzymes, proteins, and peptides that inactivate and digest microbes. Beyond their phagocytic capability, neutrophils produce a variety of toxic products. The release of toxic products is known as the respiratory burst because it is accompanied by an increase in oxygen consumption. During the respiratory burst, oxygen radicals are generated by nicotinamide adenine dinucleotide phosphate (NADPH) oxidases. Neutrophils also contribute to host defense through extrusion of DNA and associated proteins in the form of neutrophil extracellular traps, or NETs, to which bacteria can stick, facilitating their clearance. Despite their effective contributions to the innate immune

CHAPTER 45  The Innate Immune System  

response and microbial host defense, neutrophils can generate considerable collateral damage. NETs have the capacity to induce production of cytokines by pDCs and may damage vascular endothelial cells. Secretion of neutrophil granule contents, particularly their enzymes (myeloperoxidase, elastase, collagenase, and lysozyme), causes direct cellular injury and damages macromolecules at inflamed sites.


In contrast to macrophages and neutrophils, eosinophils are only weakly phagocytic but are potent cytotoxic effector cells against parasites. Their major effector mechanism is the secretion of cationic proteins (major basic protein, eosinophil cationic protein, and eosinophil-derived neurotoxin). These proteins are released into the extracellular space, where they directly destroy the invading microorganism but can also damage host tissue (Chapter 170).

Basophils and Mast Cells

Basophils and tissue mast cells secrete inflammatory mediators such as histamine, prostaglandins, leukotrienes, and some cytokines.12 Release of these substances is triggered when cell surface immunoglobulin E (IgE) receptors encounter monomeric IgE. They play a role in atopic allergies, in which allergens bind immunoglobulin (IgE) and cross-link FcεRs. Mast cells have been observed in rheumatoid arthritis synovial tissue and have been implicated in local inflammatory responses (Chapter 255). Like pDCs and macrophages, mast cells express TLRs and FcRs and produce cytokines after encountering immune complexes composed of TLR ligands.

Recognition Receptors and Triggers of an Innate Immune Response Toll-like Receptors

The innate immune system utilizes both cell surface and intracellular PRRs to recognize conserved structures on microbes (PAMPs). Examples of PAMPs are bacterial lipopolysaccharides, peptidoglycans, mannans, bacterial DNA, double-stranded RNA, and glucans. The discovery and characterization of the TLR family of receptors and their relevant ligands has focused attention on the mechanisms that allow an innate immune response to shape the nature of the resulting inflammatory or repair programs, as well as the T-cell effector cell functions that follow recognition of antigens from the relevant pathogen. The TLRs have in common leucine-rich domains and bind PAMPs common to classes of pathogenic organisms.7,8 For example, TLR-4, a cell surface−expressed PRR, binds lipopolysaccharide of gram-negative bacteria, and TLR-2 recognizes bacterial peptidoglycans and lipoproteins, often based on dimerization with other TLR family members. Important advances in understanding systemic autoimmune diseases have followed the characterization of endosomal TLRs with relative specificity for single-stranded RNA (TLR-7 and TLR-8), demethylated CpG-enriched DNA (TLR-9), and double-stranded RNA (TLR-3, which has both cell surface and endosomal forms). The distribution of particular TLRs among cells of the innate immune system varies, and additional members of the TLR family may still be discovered and characterized. The TLRs play central roles in alerting the immune system that a microbe, typically a bacteria in the case of TLR-2 and TLR-4 or a virus in the case of TLR-3, TLR-7, TLR-,8 and TLR-9, is threatening the host. But in some cases, when an immune complex with self-nucleic acid gains access to an endosomal TLR, a self-directed innate immune response can be initiated or amplified.

Cytoplasmic Nucleic Acid Sensors

Following the description of the TLR family and the capacity of the endosomal TLRs to recognize microbial and self-nucleic acids, a second category of intracellular innate immune system receptors was defined that recognize RNA or DNA from microbes, primarily viruses, that gain access to the cell cytoplasm. The DExD/H-box family of helicases include retinoic acid−inducible gene I (RIG-I) and melanoma differentiation−associated protein 5 (MDA5), described as members of the RIG-I-like receptor (RLR) family that recognizes viral RNAs with particular structural characteristics that distinguish the viral RNA from most host RNAs (Fig. 45-1).13,14 Cytoplasmic DNA receptors have also been defined, with cyclic guanosine monophosphate−adenosine monophosphate synthase (cGAS) recently identified as an important sensor of cytoplasmic DNA that triggers an innate immune response after interacting with the stimulator of interferon genes (STING).15 Whether RNA or DNA triggers these cytoplasmic sensors, the result is transcription and production of interferon-β and other pro-


inflammatory cytokines that orchestrate the early phase of an antiviral immune response.

NOD Receptors

Another category of intracellular receptors is proving important in antimicrobial defense as well as contributing to activation of inflammatory states. The nucleotide-binding oligomerization domain (NOD)-like receptor (NLR) family comprises components of an intracellular structure called the inflammasome, a signaling platform that organizes innate immune system activation in response to some stimuli.16 The inflammasome can activate caspase 1, an enzyme important for maturation of the pro-inflammatory cytokines IL-1β and IL-18. The NLRP3-containing inflammasome has been best studied and implicated in the inflammatory response to monosodium urate crystals, the triggers of gout attacks (Chapter 273). Mutations in the NLRP3 gene are the basis of chronic autoinflammatory syndromes that are associated with exaggerated production of IL-1 (reviewed in Chapter 261).

C-Type Lectin Receptors

Members of the C-type lectin receptor family have a carbohydrate recognition domain and a calcium-binding domain that promotes signaling after interaction with carbohydrate-expressing microbes as well as self-molecules. DC-SIGN (DC-specific intracellular adhesion molecule-3 grabbing nonintegrin) is an example of a family member that recognizes high-mannosecontaining structures on foreign antigens and supports DC activation. Mannose receptors on macrophages, dendritic cells, and other cell types, such as renal mesangial cells, participate in clearance of microbes as well as antigen trapping for presentation to adaptive immune system cells. The selectin family of proteins have a lectin domain, bind to carbohydrate ligands, and mediate the first steps of leukocyte migration. L-selectin is present on virtually all leukocytes; P-selectin and E-selectin are expressed on activated endothelial cells, and P-selectin is also stored in platelets. Selectins capture floating leukocytes and initiate their attachment and rolling on activated endothelial cells.

Scavenger Receptors

Scavenger receptors comprise a diverse family of receptors with the common functional role of binding various ligands and transporting or removing nonself or altered-self targets.17 They can participate in clearance of microorganisms and cholesterol transport but can also contribute to disease pathology. For example, among the scavenger receptors is the receptor for oxidized low-density lipoproteins, which can promote generation of lipid-laden macrophages and atherosclerosis when accumulated in excess, and receptors for relatively inert substances such as silicon, which can drive an inflammatory response once taken into phagocytic cells. Scavenger receptors can also participate in activation of the inflammasome, as can occur after binding serum amyloid A protein.

Inhibitory Natural Killer Cell Receptors

The immunoglobulin-like killer inhibitory receptor (KIR) family of receptors participates in distinguishing self-cells from cells of foreign origin or tumor cells expressing modified-self-molecules. NK cells are ready to produce their toxic mediators, but they are held in check by inhibitory receptors that recognize MHC class I or MHC class I–like molecules.10 Recognition of MHC class I molecules provides a negative signal that suppresses cell activity. The observation that NK cells kill target cells lacking MHC class I molecules recognized as self led to the missing-self hypothesis. By screening cell surfaces for the expression of MHC class I molecules, the innate immune system collects information about the intactness of tissues, emphasizing the crucial role of MHC class I molecules as markers of tissue integrity.

Fc and Complement Receptors

Most cells of the innate immune system possess receptors (FcRs) that specifically interact with the constant region (Fc portion) of immunoglobulins and can bind antibodies attached to antigens. The isotype of the antibody determines which cell type is activated in a given response. Triggering of most FcRs transmits activating signals; however, inhibitory FcRs on B lymphocytes (B cells) and macrophages can limit responses. Ligation of an FcγR on macrophages or neutrophils triggers phagocytosis of the antigen, activation of respiratory burst, and induction of cytotoxicity. On NK cells, FcγRs initiate antibody-dependent cell-mediated cytotoxicity. FcRs on pDCs are important for bringing immune complexes into intracellular compartments containing endosomal TLRs. FcRs on mast cells, basophils, and activated


CHAPTER 45  The Innate Immune System  


IFNα/β IFNα/β IFNα/β

IFNα/β receptor















IKKε Mitochondria









Nucleus P

P CBP/ IRF-3 NF κB p300 IRF-3 P





IRF-9 IFN-β IFN-α FIGURE 45-1.  Induction of antiviral type I interferon response. Cytoplasmic sensors of RNA, including RIG-I and MDA5, trigger a signaling cascade that results in translocation of IRF-3 to the nucleus and transcription of interferons. Those cytokines promote an antiviral immune response after binding to their receptor and activating the JAK-STAT pathway. CBP/p300 = CREB binding protein; NEMO = NF-κB essential modulator; IFN = interferon; IKK = inhibitor of nuclear factor κB kinase subunit; IPS-1 = interferon-β promoter stimulator-1; IRF = interferon response factor; ISG = interferon stimulated gene; ISGF3 = interferon-stimulated gene factor 3; JAK = Janus kinase; MDA5 = melanoma differentiation-associated protein 5; RIG-1 = retinoic acid−inducible gene 1; STAT = signal transducer and activator of transcription; TRAF3 = TNF (tumor necrosis factor) receptor−associated factor; Tyk = tyrosine kinase. (From Wilkins C, Gale M Jr. Recognition of viruses by cytoplasmic sensors. Curr Opin Immunol. 2010;22:41-47.)

eosinophils bind monomeric IgE with extremely high affinity. Cross-linking of the constitutively cell surface–bound IgE induces cell activation and the release of cytoplasmic granules. Some immunoglobulin isotypes fix complement, and complement receptors on monocytes amplify cell activation induced by antigen-antibody-complement immune complexes18 (Chapter 50). Complement receptor 1 (CR1) binds C3b and C4b, initial degradation products of complement activation, and when activated promotes phagocytosis of a complement-bearing immune complex. CR3 and CR4 are β2integrins and bind the degradation product iC3b.

of activated macrophages but also binds to those cells through its specific receptor, expanding an inflammatory response. Innate immune cells also express receptors for IL-6, which induces acute phase reactants and type I interferon, which orchestrates a broad host defense program in response to virus infection (see Fig. 45-1). Chemokine receptors include many family members that are differentially distributed among immune system cells and sense the gradient generated by soluble chemokines, resulting in attraction of cells to sites where they are needed to implement inflammatory or immune functions.

Cytokine and Chemokine Receptors

Signaling Pathways and Effector Mediators of the Innate Immune System

Cells of the innate immune system express receptors for many cytokines, soluble, low-molecular-weight glycoproteins that derive from many cellular sources.5 Binding of IFN-γ, produced by NK or type 1 helper T cells (TH1 cells), by its receptor on monocytes activates a differentiation program that expands an inflammatory response. Receptors for IL-4 on monocytes induce a gene transcription program that is more supportive of a wound healing and repair program. Tumor necrosis factor-α (TNF-α) is a product

Each family of innate immune system receptors utilizes a complex network of molecules to transmit information from the cell surface or its cytoplasm to the nucleus, resulting in induction of a broad gene transcription and protein synthesis program that implements the next phase of the response. The contributions of each of the signal transduction pathways to the overall innate immune response will depend on the proteins produced and will determine whether the resulting cell products focus the overall immune

CHAPTER 45  The Innate Immune System  

function on ablating the damaging effects of virus infection on the host, limiting the inflammation and tissue damage that follow a bacterial or fungal infection, or healing a tissue wound through the production of scar tissue.

Receptor-Mediated Signaling Pathways

Certain common cell signaling systems are utilized by many cells and receptor systems.6,7 Arguably the most important is the nuclear factor κ light-chain enhancer of activated B cells (NF-κB) pathway. NF-κB is a rapid-acting transcription factor because it is preformed in cells of the innate immune system and does not require new protein synthesis to take action. Its activity is induced by ligation of TLRs and many cytokine receptors. Its component transcription factors translocate to the cell nucleus after degradation of an inhibitory component, inhibitor of κB (IκB), and bind to promoter regions of genes encoding mediators of inflammation and cell proliferation. Another important pathway is mediated by the interferon regulatory factor (IRF) family, including transcription factors that are activated by endosomal TLRs in response to ligation by DNA or RNA, or by cytoplasmic nucleic acid sensors, usually from viral sources. IRF-3 is particularly important for promoting transcription of interferon-β, typically produced early in an antivirus innate immune response. IRF-7 is particularly supportive of interferon-α production induced by endosomal TLRs and is constitutively present in pDCs, the most active producers of IFN-α. The Janus kinase ( JAK)-signal transducer and activator of transcription (STAT) pathway is utilized by many cytokine receptors and involves sequential enzymatic reactions by kinases that eventuate in translocation of STAT proteins to the nucleus, where they bind to gene promoters and induce transcription and production of products important in implementing immunoregulation and inflammation. TNF receptor family members activate a complex signaling pathway that involves proteins called TNF receptor−associated death domain (TRADD) proteins and TNF receptor−associated factors (TRAFs), ultimately activating the NF-κB and the mitogen-activated protein (MAP) kinase pathways. The TGF-β receptor is a serine/threonine receptor kinase that phosphorylates cytoplasmic proteins of the SMAD family, which act as transcription factors after receptor engagement by TGF-β. TGF-β signaling can play an important role in terminating an innate immune response and initiating a wound healing or tissue repair program. It is apparent that common intracellular signaling strategies are used by many of the receptor systems that activate and regulate the innate immune system, with ligand-receptor engagement triggering the activation of kinases that phosphorylate downstream pathway proteins, and result in translocation of important transcription factors from cytoplasm to nucleus where new gene transcription takes place.

Soluble Products of the Innate Immune Response

Cells of the innate immune system are the principal producers of many proinflammatory and regulatory cytokines already mentioned, and are also their targets. In addition to the cytokines described, cells of the innate immune system produce chemokines that attract immune system cells to sites of tissue damage or infection, and they produce cell survival and differentiation factors that help to develop an adaptive immune response. Macrophages and dendritic cells produce IL-12 and IL-23 to support development of effector T-cell programs, and they produce B-cell-activating factor (BAFF), a soluble mediator of the TNF family. BAFF supports B-cell survival and can provide costimulatory signals to B cells that have received antigen-specific activation signals through their surface B-cell antigen receptors, promoting differentiation to antibody-producing plasma cells. A particularly important set of products includes components of the complement system, a group of plasma enzymes and regulatory proteins that are converted from inactive pro-enzymes to active enzymes in a controlled and systematic cascade, which is crucial in linking microbial recognition to cellular effector function (Chapter 50). Mannose-binding lectin circulates in the plasma, functioning as an opsonin, and is involved in activation of the complement pathway. C-reactive protein, an acute phase protein, participates in opsonization by binding to bacterial phospholipids. Macrophages and neutrophils are important in the initiation phase of an innate immune response through their production of antimicrobial defensins, cysteine-rich cationic proteins, and cathelicidin peptides, such as LL37.19 Both categories of mediators can assist in killing of microbes in phagosomes. Neutrophils extrude stimulatory DNA in the form of NETs or release mitochondrial DNA, along with DNA-associated proteins like high mobility group box 1 (HMGB1) that amplifies TLR responses in pDCs or macrophages.


Role of the Innate Immune System in Localization, Extension, and Resolution of a Host Defense Reaction Localization of Innate Immune System Cells

Most cells of the innate immune response are free agents, moving through blood or lymph in transit from one site to another. Mobility of the cellular constituents of the innate immune system is required for effective initiation of a response to invading microbes. Cells use a multistep process of adherence and activation. Initially, leukocytes roll on activated endothelial cells, activate chemokine receptors, increase adhesiveness, and eventually migrate through the endothelial layer across a chemokine gradient. The selectin family of proteins mediates the first steps of leukocyte migration. P-selectin and E-selectin are expressed on activated endothelial cells, and P-selectin is also stored in platelets. Selectins capture floating leukocytes and initiate their attachment and rolling on activated endothelial cells. To transform attachment and rolling into firm adhesion, the concerted action of chemokines, chemokine receptors, and integrins is necessary. Integrins are heterodimers formed of many different α chains and β chains; different α/β combinations are expressed on different cell subsets. Only after activation can integrins interact with ligands on endothelial cells. Activation involves modification of the cytoplasmic domain of the β chain, which leads to a structural change of the extracellular domains. This process is termed inside-out signaling. The last step of homing is transendothelial migration. Here, the firmly attached leukocytes migrate through the endothelial cell monolayer and the basem*nt membrane of the vessel wall.

Transition to an Adaptive Immune Response

Movement of innate immune system cells is also required to transition a host response from primarily one depending on cells of the innate immune system to one that engages T and B lymphocytes. Dendritic cells resident in the skin and gut serve as sentinels and a first line of defense against invading organisms. When those cells are activated following sensing of PAMPs by PRRs and following uptake of microbial components by those cells, the DCs migrate to local lymph nodes where their contents, by now expressed on their surface in association with MHC class I or II molecules, can be sampled by T cells. As noted, activated macrophages, DCs, and pDCs produce cytokines that shape the differentiation program of T cells. In addition, cell surface costimulatory molecules induced after TLR-mediated activation, such as CD80 and CD86, provide essential accessory activation signals to T cells to ensure their effective activation. Macrophages and DCs also support the development of an adaptive immune response through their production of survival and differentiation factors. Chapter 46 provides a full description of the adaptive immune system and its implementation.

Role of Innate Immune System Cells in Resolution of an Immune Response and Wound Repair

Macrophages are particularly important in resolving an immune response and organizing the repair of damaged tissue. A classic paradigm describing pro-inflammatory/classically activated (M1) and anti-inflammatory/alternatively activated (M2) macrophages (see earlier under Monocytes and Macrophages) is likely to be overly simplistic. Yet it is clear that in the course of a chronic infection, macrophages can shift their functional profile from M1 to M2, in some cases promoted by the T-cell cytokines IL-4 and IL-13, to develop a gene expression program that includes production of TGF-β, supportive of a fibrotic response, and IL-10, a cytokine that inhibits antigenpresenting cell function.9 Although an M1-like profile driven by IFN-γ is highly productive in achieving initial control over a pathogenic invading microbe, and M2-derived mediators promote wound healing, it should be recognized that either macrophage phenotype, and complex in-between profiles, can also be associated with pathologic states (Fig. 45-2). Current research is unraveling the innate immune mechanisms that account for such diverse diseases as atherosclerosis (Chapter 70), viewed as associated with M1 macrophages, and idiopathic pulmonary fibrosis (Chapter 92), possibly involving M2-like macrophages.

Contribution of the Innate Immune Response to Pathogenesis of Autoimmune Disease

Among the most significant insights of the past decade is the essential contribution of the innate immune system to the pathogenesis of autoimmune and inflammatory diseases. As described, the cells of the innate immune system are integral players in the early recognition of invading pathogenic microbes, and when the functions of this complex system are carefully

Chronic phase





Breast cancer



Atopic dermatitis


Wound healing






Airway inflammation (asthma)

FIGURE 45-2.  Schematic representation of macrophage plasticity and polarization in pathology. Dynamic changes occur over time with evolution of pathology: for instance, a switch from M1 to M2 macrophage polarization characterizes the transition from early to chronic phases of infection. Moreover, mixed phenotypes or populations with different phenotypes can coexist. (From Sica A, Mantovani A. Macrophage plasticity and polarization: in vivo veritas. J Clin Invest. 2012;122:787-795.)

orchestrated and balanced, the result is efficient ablation, or at least isolation, of the microbe. However, if the microbe is not effectively cleared from the system and persists, a chronic state of infection associated with immune activation and tissue damage is the result. Interestingly, many parallels can be seen between the immune alterations observed in the setting of chronic viral infection and the impaired immunoregulation characteristic of the prototypic autoimmune disease systemic lupus erythematosus. Excessive production of interferon-α is a feature of most patients with that disease, and it is now understood that activation of the endosomal TLRs by nucleic acid−containing immune complexes amplifies the activity of the innate immune response and drives production of interferon-α and other proinflammatory cytokines. Neutrophils are now recognized to contribute to the induction of that response through their production of HMGB1, cathelicidins, and extrusion of stimulatory DNA aggregates. TLR activation is proposed to contribute to many additional autoimmune and inflammatory diseases; as endogenous TLR ligands can act as effective TLR stimuli in the setting of a pro-inflammatory environment associated with oxidative cell damage. The inflammasome and its component proteins, including the NOD-like receptors, are recognized as mediators of inflammatory responses induced by urate crystals that result in gout attacks (Chapter 273), and they are targets of mutations that define dramatic autoinflammatory syndromes (Chapter 261), particularly seen in children.


The cells and products of the innate immune response, for many years viewed as less sophisticated and important than the highly specific T and B lymphocytes of the adaptive immune response, have taken their place as essential defenders against pathogenic microbes. Through the recognition of common molecular patterns characteristic of microbes by members of receptor families, some still being discovered, the cells of the innate immune response orchestrate the effector programs that are fine-tuned to target the vulnerabilities of each pathogen and kill, or at least limit the expansion of, that microbe. Advances in understanding the mechanisms utilized by the innate immune

response and the clinical syndromes that result when components of that system are genetically altered, have elucidated the central role that receptors and products of the innate immune system play in the pathogenesis of autoimmune and inflammatory diseases. These insights are guiding efforts to develop targeted therapies that will leverage the new knowledge to control or even prevent human diseases in which the innate immune system plays an important pathogenic role. GENERAL REFERENCES For the General References and other additional features, please visit Expert Consult at

CHAPTER 45  The Innate Immune System  

GENERAL REFERENCES 1. Quintana-Murci L, Clark AG. Population genetic tools for dissecting innate immunity in humans. Nat Rev Immunol. 2013;13:280-293. 2. Iwasaki A, Pillai PS. Innate immunity to influenza virus infection. Nat Rev Immunol. 2014; 14:315-328. 3. Busca A, Kumar A. Innate immune responses in hepatitis B virus (HBV) infection. Virol J. 2014;11:22. 4. Marcus A, Gowen BG, Thompson TW, et al. Recognition of tumors by the innate immune system and natural killer cells. Adv Immunol. 2014;122:91-128. 5. Torrado E, Cooper AM. Cytokines in the balance of protection and pathology during mycobacterial infections. Adv Exp Med Biol. 2013;783:121-140. 6. Zak DE, Tam VC, Aderem A. Systems-level analysis of innate immunity. Annu Rev Immunol. 2014;32:547-577. 7. Broz P, Monack DM. Newly described pattern recognition receptors team up against intracellular pathogens. Nat Rev Immunol. 2013;13:551-565. 8. O’Neill LA, Golenbock D, Bowie AG. The history of toll-like receptors: redefining innate immunity. Nat Rev Immunol. 2013;13:453-460. 9. Sica A, Mantovani A. Macrophage plasticity and polarization: in vivo veritas. J Clin Invest. 2012;122:787-795.


10. Terabe M, Berzofsky JA. The immunoregulatory role of type I and type II NKT cells in cancer and other diseases. Cancer Immunol Immunother. 2014;63:199-213. 11. Hazenberg MD, Spits H. Human innate lymphoid cells. Blood. 2014;124:700-709. 12. Cromheecke JL, Nguyen KT, Huston DP. Emerging role of human basophil biology in health and disease. Curr Allergy Asthma Rep. 2014;14:408. 13. Schlee M. Master sensors of pathogenic RNA - RIG-I like receptors. Immunobiology. 2013;218: 1322-1335. 14. Wu J, Chen ZJ. Innate immune sensing and signaling of cytosolic nucleic acids. Annu Rev Immunol. 2014;32:461-488. 15. Cai X, Chiu YH, Chen ZJ. The cGAS-cGAMP-STING pathway of cytosolic DNA sensing and signaling. Mol Cell. 2014;54:289-296. 16. Latz E, Xiao TS, Stutz A. Activation and regulation of the inflammasomes. Nat Rev Immunol. 2013;13:397-411. 17. Canton J, Neculai D, Grinstein S. Scavenger receptors in homeostasis and immunity. Nat Rev Immunol. 2013;13:621-634. 18. Holers VM. Complement and its receptors: new insights into human disease. Annu Rev Immunol. 2014;32:433-459. 19. Silva PM, Gonçalves S, Santos NC. Defensins: antifungal lessons from eukaryotes. Front Microbiol. 2014;5:97.


CHAPTER 45  The Innate Immune System  

REVIEW QUESTIONS 1. Which of the following cell types is not considered to be a component of the innate immune system? A. Macrophages B. Neutrophils C. T lymphocytes D. Plasmacytoid dendritic cells E. Eosinophils Answer: C  T lymphocytes are important components of the adaptive immune system. T and B lymphocytes use mechanisms that rearrange DNA to form novel specific antigen-binding receptors. Cells of the innate immune system express pattern recognition receptors but do not express specific antigen-binding receptors. 2. Endosomal toll-like receptors (TLRs) recognize which of the following pathogen-associated molecular patterns (PAMPs)? A. Flagellin B. Lipopolysaccharide C. Antigenic peptides D. Nucleic acids Answer: D  TLRs are important innate immune system receptors that recognize patterns expressed by pathogenic microbes and some endogenous molecules. Cell surface−expressed TLRs recognize molecules that are typically expressed on the surface of microbes. Intracellular endosomal TLRs, such as TLR-3, -7, -8, and -9, recognize RNA or DNA. The sequestering of those nucleic acid−responsive TLRs protects the immune system from inadvertent activation by self-nucleic acids. However, in diseases such as systemic lupus erythematosus, nucleic acid−containing immune complexes can gain access to the endosomal TLRs and induce an innate immune response. 3. Which of the following innate immune system stimuli utilizes the inflammasome to trigger an inflammatory disease? A. Peptide−major histocompatibility class (MHC) class II complex B. Monosodium urate crystals C. Interleukin-6 (IL-6) D. Immunoglobulin E E. Complement Answer: B  Urate crystals access the components of the NOD-like receptors of the inflammasome, activate caspase I, and induce the formation of IL-1, a pro-inflammatory mediator that can amplify an innate immune response. In some patients, this response leads to the acute inflammatory arthritis known as gout. Peptide−MHC class II complexes are stimuli for activation of an adaptive immune response. IL-6 is a broadly active cytokine, and immunogloblulin E is a component of an allergic response.

4. M2 macrophages participate in which of the following? A. Wound healing responses B. Antigen presentation C. Production of IL-12 D. Recognition of oxidized low-density lipoprotein (LDL) E. Complement activation Answer: A  Although the designation of M1 and M2 macrophages is overly simplistic, macrophages do shift their functional program as the course of an immune response progresses toward a more chronic state, with M2 macrophages expanding and producing mediators, such as transforming growth factor-β and IL-10, that contribute to resolution of responses and repair of damaged tissue. Antigen presentation, production of IL-12, and recognition of oxidized LDL are more likely to be features of M1 macrophages. 5. Cells of the innate immune system produce soluble mediators that contribute to activation and expansion of an adaptive immune response. Among those mediators are which of the following? A. BAFF B. IL-23 C. LL37 D. All of the above E. A and B Answer: E  Macrophages and dendritic cells produce mediators that influence both the T and B cell arms of the adaptive immune response. IL-23 can promote generation of T-cell effector programs. BAFF is a B-cell survival and differentiation factor. LL37 is a cathelicidin that is produced by neutrophils and participates in the killing of phagocytosed microbes.


CHAPTER 46  The Adaptive Immune System  



Structure of Antigen-Specific Receptors

The innate immune system recognizes structural patterns that are common in the microbial world, whereas the adaptive immune system is designed to respond to the entire continuum of antigens. This goal is achieved through two principal types of antigen recognition receptors: antibodies and T-cell receptors (TCRs). Antibodies, or immunoglobulins, are expressed as cell surface receptors on B cells or are secreted, both of which have the same


CHAPTER 46  The Adaptive Immune System  

specificity for antigen. They recognize conformational structures formed by the tertiary configuration of proteins. In contrast, α/β TCRs, the most abundant class of TCRs, fit specifically to epitopes formed by a small linear peptide embedded into major histocompatibility complex (MHC) molecules on the surface of antigen-presenting cells.


Antibodies consist of two identical heavy chains and two identical light chains, which are covalently linked by disulfide bonds. The amino (N)-terminal domain of each chain is variable and represents the recognition structure that interacts with the antigen. Each antibody has two binding arms of identical specificity. The carboxy (C)-terminal ends of the heavy and light chains form the constant region, which defines the subclass of the antibody (κ or λ for light chains; immunoglobulin M (IgM), IgA, IgD, IgE, or IgG for heavy chains). Additional subclasses can be distinguished for IgG and IgA. The constant region of antibodies includes the Fc region. Fc regions can polymerize (IgA) or pentamerize in the presence of a J (joining) chain (IgM). Fc regions are also the ligand for Fc receptors (FcRs) on cells of the innate immune system.

T-Cell Receptors

TCRs are dimers of α chains and β chains or of γ chains and δ chains, each of which contains three complementary-determining binding sites in the N-terminal domain. These complementary-determining sites define the specificity. α/β TCRs recognize peptide fragments in the context of MHC molecules, although certain ones bind glycolipid antigens, for example from mycobacteria, displayed by molecules with structural similarity to MHC. γ/δ TCRs are more variable and can recognize peptides or certain glycolipid antigens in the context of MHC-like molecules, or even unprocessed antigens, functioning similar to antibodies; the latter is a reflection of their structural similarity.

antigen presentation. The two classes of MHC molecules are used as restriction elements by two different subsets of T cells. CD4+ T cells recognize antigenic peptides embedded into MHC class II molecules, whereas CD8+ T cells bind peptides complexed with MHC class I molecules. Generally, MHC class II molecules are expressed only on specialized, so-called professional, antigen-presenting cells, such as dendritic cells, monocytes, macrophages, and B cells, whereas class I proteins are displayed by virtually all nucleated cells, facilitating recognition by CD8+ T cells of peptides from viruses that often have a broad range of target tissues. Peptides bound to MHC class II molecules typically derive from extracellular antigens that are captured and internalized into endosomes to be digested by proteinases, notably cathepsin. Occasionally, however, intracellular proteins or membrane proteins are also funneled into this pathway. MHC class II molecules are assembled in the endoplasmic reticulum in association with a protein called the invariant chain (Fig. 46-1). The molecules are transported to the endosome, where the invariant chain is removed from the peptide-binding cleft, making the cleft accessible to peptides derived from extracellular proteins. MHC class II molecules, stabilized by peptides of 10 to 30 amino acids in length, are displayed on the cell surface, where they are recognized by CD4+ T cells. MHC class I–associated peptides are produced in the cytosol by the proteasome, a large cytoplasmic multiprotein enzyme complex (see Fig. 46-1). Specialized transporter proteins, called transporter in antigen processing (TAP), facilitate translocation of peptides from the cytosolic proteasome to the endoplasmic reticulum. There, the peptides bind to newly formed MHC class I molecules and are transported to the cell surface, where they are recognized by antigen-specific CD8+ T cells. MHC class I−associated peptides may also originate in the extracellular environment and be presented to T cells through the appropriately named cross-presentation pathway. This enables CD8+ T cells to recognize foreign peptides, for example, from viruses, that

Specificities of Antibodies and T-Cell Receptors

The repertoires, or total number of specificities, of antibodies and TCRs are extremely diverse and have been estimated in the human to account for up to 1011 or higher, and 1018, respectively, combinations. This enormous diversity reflects the anticipatory nature of adaptive immune receptors and must be acquired; it cannot be genetically encoded in contrast to that of innate receptors. Its foundation consists of fewer than 400 genes that are recombined and modified. Immunoglobulin heavy chains are formed from four gene segments—the variable, diversity, joining, and constant region gene segments. Also, TCR β chains and δ chains are assembled by the recombination of variable, diversity, joining, and constant region segments of TCR genes. Immunoglobulin light chains and TCR α chains and γ chains lack the diversity segment and are composed of three gene segments. During antibody or TCR rearrangement, gene segments are cut out by nucleases and recombined at the DNA level to form linear coding units for each receptor gene. Through the combination of several different mechanisms, an enormous diversity of receptors is generated. First, the genome contains multiple forms of gene segments; each receptor or antibody uses a different combination of these gene segments. Second, the splicing process is imprecise, introducing nucleotide variations at the variable-diversity, diversity-joining, and variable-joining junctions. These inaccuracies lead to frame shifts and result in completely different amino acid sequences. Finally, random nucleotides can be inserted at the junctional region by an enzyme, deoxyribonucleotidyl transferase. Once generated, TCR sequences remain unchanged. This rule does not apply to immunoglobulins, which undergo modification. Immunoglobulin modification includes (1) replacement of an entire variable region, or receptor editing, typically occurring in the bone marrow during B-cell development to modify those immunoglobulin receptors that inadvertently bind self-antigens on initial recombination of gene segments; (2) class switching, in which the variable-diversity-joining unit combines with different constant region genes (isotype switching); or (3) somatic hypermutation, in which the antigen-contact areas of the antibody undergo mutations during an immune response to improve the affinity (affinity maturation). The latter two events occur in secondary lymphoid tissues, such as the spleen, lymph nodes, and mucosal lymphoid tissue, where immune responses to antigens are initiated.

Antigen Processing

T cells bearing α/β TCRs recognize peptide fragments that are displayed in the context of MHC class I and class II molecules through a process named

Endoplasmic reticulum MHC I Antigen TAP

MHC II Invariant chain Antigen Golgi complex

Proteosome Endosome

CD8 T cell

CD4 T cell FIGURE 46-1.  Pathways of antigen processing and delivery to major histocompatibility complex (MHC) molecules. Cytosolic proteins are broken down by the proteosome to generate peptide fragments, which are transported into the endoplasmic reticulum by specialized peptide transporters (TAP). After peptides are bound to MHC class I molecules, MHC-peptide complexes are released from the endoplasmic reticulum and travel to the cell surface, where they are ligands for CD8+ T-cell receptors (TCRs). Extracellular foreign antigens are taken into intracellular vesicles, called endosomes. As the pH in the endosomes gradually decreases, proteases are activated that digest antigens into peptide fragments. After fusing with vesicles that contain MHC class II molecules, antigenic peptides are placed in the antigen-binding groove. Loaded MHC class II–peptide complexes are transported to the cell surface, where they are recognized by the TCRs of CD4+ T cells.


CHAPTER 46  The Adaptive Immune System  

are derived from infected and dying cells that are ingested by myeloid cells and then presented by MHC class I molecules. The nature of the antigen-processing pathway determines the sequence of events in immune responses. Extracellular antigens, in general, enter the endosomal pool and associate with MHC class II molecules to stimulate CD4+ T cells. Cytosolic antigens, including antigens from intracellular infectious agents, are degraded and displayed in the context of MHC class I molecules to initiate CD8+ T-cell responses.


T Cells T-Cell Development

T precursor cells are derived from hematopoietic stem cells that migrate to the thymus, a primary lymphoid tissue, where all the subsequent stages of T-cell maturation occur (Fig. 46-2). Pre-T cells express two enzymes, recombinase and terminal deoxynucleotidyl transferase, enabling them to recombine TCR genes. The β chain of the TCR is rearranged first and is expressed together with a pre-TCR α chain. Signals from the immature TCR complex inhibit rearrangement of the second β-chain allele and induce thymocyte proliferation and expression of both CD4 and CD8 molecules, so-called double positive thymocytes. Subsequently, the TCR α chain is recombined, with formation of a mature TCR. From here, the thymocyte undergoes many

CD4−CD8− pre-T cells

Nurse cell Cortex

differentiation and selection steps modulated by the thymic microenvironment, with the end result being formation of a T cell that is ready to migrate to secondary lymphoid tissues and to be poised to recognize antigenic peptides. Early-stage thymocytes reside in the thymic cortex, where they mostly interact with epithelial cells. They then migrate toward the medulla, encountering dendritic cells and macrophages at the corticomedullary junction. Thymic stromal cells regulate T-cell proliferation by secreting lymphopoietic growth factors, such as interleukin-7 (IL-7). Interactions of the TCR with MHC molecules expressed on epithelial cells and on dendritic cells or macrophages determine the fate of the thymocyte.1 Low-avidity recognition of peptide-MHC complexes on thymic epithelial cells by the TCR results in positive selection.2 This recognition event rescues cells from apoptotic cell death and ensures that only T cells with functional receptors that can recognize MHC molecules, critical for T-cell activation on subsequent residence in the spleen and lymph nodes, survive. Thymocytes that express a receptor not fitting any MHC antigen complex die by neglect. High-affinity interaction between the TCR and peptide-MHC complex induces apoptotic death of the recognizing T cell. This process of negative selection eliminates T cells with specificity for self-antigens and is responsible for central tolerance to many autoantigens. It has been estimated that approximately 1% of thymocytes survive the stringent selection process. While undergoing selection, T cells continue to differentiate, with orderly expression of cell surface molecules. Double-positive thymocytes expressing both CD4 and CD8 molecules downregulate one or the other, developing into single-positive CD4+ helper T cells that have been selected on MHC class II complexes or CD8+ cytotoxic T cells that are restricted to MHC class I complexes. These single-positive cells are now mature T cells that are ready for exit and migration through the circulation to secondary lymphoid organs, including the spleen, lymph nodes, and mucosal lymphoid tissues, following chemokine cues and using adhesion molecules to enter. They exist in these tissues as inactivated, or naïve, cells until receiving the appropriate antigenic signal for activation and subsequent effector function.

T-Cell Stimulation and Accessory Molecules CD4+CD8+ T cells


MHC I+ self-antigen


MHC II+ self-antigen Positive selection

Cortical epithelial cells


Negative selection


Dendritic cell

High-avidity self-recognition

CD4 Low-avidity self-recognition

Self-tolerant cells MHC class I MHC class II restricted restricted CD8 CD4 Apoptosis Medulla



Periphery FIGURE 46-2.  Maturation of T cells in the thymus. Precursors committed to the T-cell lineage arrive in the thymus and begin to rearrange their T-cell receptor (TCR) genes. Immature T cells with receptors binding to self–major histocompatibility complex (MHC) on cortical epithelial cells receive signals for survival (positive selection). At the corticomedullary junction, surviving T cells probe self-antigens presented by dendritic cells and macrophages. T cells reacting strongly to self-antigens are deleted by apoptosis (negative selection). T cells released into the periphery are tolerant toward self and recognize foreign antigens in the context of self-MHC.

T-cell activation is initiated when TCR complexes recognize antigenic peptides in the context of the appropriate MHC molecule on the surface of an antigen-presenting cell in secondary lymphoid organs. The principal antigenpresenting cells for activation of naïve T cells are dendritic cells. MHCpeptide recognition by the TCR, the first signal for T-cell activation, leads to receptor clustering and phosphorylation of the intracellular portion of the CD3 protein complex, the signaling component of the TCR, by receptorassociated tyrosine kinases. These events transmit signals to the nucleus of the T cell and initiate its activation. The coreceptors CD4 and CD8 are also critical for the initial events in T-cell activation, through their interaction with MHC class II and class I molecules, respectively, supporting CD3mediated signals. Yet, this first activation signal delivered by the TCR and coreceptors is not alone sufficient for robust T-cell survival and differentiation. It needs to be complemented by the interaction of accessory molecules on the T cell and their ligands on the antigen-presenting cell. A spectrum of accessory molecules is known, of which the best known is CD28, which are engaged by CD80 and CD86 (also known as B7.1 and B7.2, respectively) on antigen-presenting cells (E-Table 46-1). Engagement of CD28 provides to the T cells a second, or costimulatory, signal to the T cell.3 This second signal, delivered by the antigen-presenting dendritic cell, ensures T-cell survival and expansion. CD28-mediated signals are mandatory for the expression of many activation markers on the responding T cells and, in particular, for the secretion of IL-2. In the absence of such a second signal, T cells are rendered nonresponsive and anergic or undergo apoptosis. Finally, adhesion molecules (integrins) stabilize the interactions between T cells and antigenpresenting cells. Signals from the TCR result in the activation of many genes and entry of the T cell into the cell cycle. The signals are transmitted by a cascade of cytoplasmic events. Cross-linking of the TCR and associated CD3 molecules results in the recruitment and activation of phosphotyrosine kinases and the phosphorylation of molecular constituents of the TCR and various adapter molecules. Signals mediated through the TCR then activate several biochemical pathways, which collectively lead to the activation of transcription factors that regulate gene expression Three major variables determine the outcome of TCR stimulation: the duration and affinity of the TCR-antigen interaction, the maturation stage of the responding T cell, and the nature of the antigen-presenting cell. Antigenpresenting cells are gatekeepers in the initiation of T-cell responses. They can

CHAPTER 46  The Adaptive Immune System  






T cells T cells T cells, mast cells T cells, mast cells Macrophages, endothelial cells Bone marrow, thymic epithelium T cells Stromal fibroblasts T cells Fibroblasts and monocytes Non-T cells Macrophages, T cells

Proliferation of T cells, B cells, and NK cells Early hematopoiesis B-cell activation, IgE switch, inhibition of TH1 cells Eosinophil growth and differentiation T-cell and B-cell growth and differentiation, induction of acute phase proteins Growth of pre-B cells and pre-T cells Stimulation of mast cells and TH2 cells Hematopoiesis B-cell growth and differentiation, inhibition of TH1 cells and macrophages Neutrophil development and differentiation Growth of T cells and NK cells Growth and differentiation of myelomonocytic lineage cells

Leukocytes Fibroblasts T cells, NK cells

Antiviral, increases MHC class I expression Antiviral, increases MHC class I expression Macrophage activation, increases expression of MHC molecules, Ig class switching, inhibition of TH2 cells



Macrophages, NK cells, T cells T cells, B cells T cells, B cells

Induction of pro-inflammatory cytokines, endothelial cell activation, apoptosis Cell death, endothelial activation, lymphoid organ development Cell death, lymphoid organ development

Monocytes, T cells Macrophages, endothelial cells T cells, macrophages Macrophages, dendritic cells T cells, mast cells, eosinophils CD4 memory cells Macrophages Macrophages, dendritic cells

Anti-inflammatory, inhibits cell growth, induces IgA secretion Acute phase response, fever, macrophage activation, costimulation Suppression of macrophage functions NK cell activation, TH1 cell differentiation Chemoattractant for CD4 T cells, monocytes, and eosinophils Cytokine production by epithelia, endothelial cells, and fibroblasts IFN-γ production by T cells and NK cells TH17 cell differentiation

OTHERS TGF-β IL-1α, IL-1β IL-10 IL-12 IL-16 IL-17 IL-18 IL-23

CD = cluster of differentiation; G-CSF = granulocyte colony-stimulating factor; GM-CSF = granulocyte-macrophage colony-stimulating factor; IFN = interferon; Ig = immunoglobulin; IL = interleukin; LT = lymphotoxin; MHC = major histocompatibility complex; NK = natural killer; TGF = transforming growth factor; TH = helper T lymphocyte; TNF = tumor necrosis factor.

CHAPTER 46  The Adaptive Immune System  

upregulate the expression of accessory molecules that provide costimulatory signals. MHC-peptide complexes are particularly dense on dendritic cells, enabling them to activate naïve T cells. In contrast, memory and effector cells have a lower threshold for activation and can react to antigens presented on peripheral tissue cells.

T-Cell Differentiation and Effector Functions

T-cell activation induces T-cell proliferation, with the goal of clonally selecting and expanding antigen-specific T cells. The extent of clonal proliferation is impressive. Antigen-specific CD8+ T cells expand several thousand−fold; CD4+ T cells expand somewhat less. During the phase of rapid growth, T cells differentiate from naïve T cells that are essentially devoid of effector functions into effector T cells that are needed for clearance of infectious organisms, or pathogens. The transition into effector cells is associated with a fundamental shift in functional profiles. First, effector T cells have a lower activation threshold; they do not require costimulation and can scan tissues that lack professional antigen-presenting cells. Second, they switch the expression of chemokine receptors and adhesion molecules to gain access to peripheral tissues. Finally, they gain effector functions. The principal effector function of CD8+ T cells is to lyse infected, antigenbearing target cells. This commitment to eventual cytotoxic function is made during development in the thymus. Upon emigration from the thymus in the naïve, or inactivated, state, CD8+ T cells circulate through secondary lymphoid tissues, surveying antigen-presenting dendritic cells for the appropriate MHC class I−peptide complex that can engage the TCR and that can supply costimulatory signals. On activation, CD8+ T cells acquire cytotoxic functions and, using a variety of receptors and adhesion molecules, can emigrate from secondary lymphoid organs to peripheral tissue sites seeking cells infected by viruses or intracellular bacteria displaying pathogen-derived peptides on MHC class I molecules. On recognizing the appropriate MHC class I–peptide complex, CD8+ T cells induce apoptosis of target cells. The T cell polarizes toward the area of antigen contact; specialized lytic granules are clustered in the contact area. A pore-forming protein, perforin, is released from the lytic granules and inserted into the target cell membrane. Proteases (granzymes) are injected into the target cells to initiate the apoptotic process by activating enzyme cascades. Mechanisms deployed by CD8+ T cells are essentially identical to those of natural killer (NK) cells. CD4+ T cells can also induce apoptosis but by a different mechanism than CD8+ T cells. On activation, they express cell surface molecules such as Fas ligand (CD178) and TRAIL, which initiate the apoptotic cascade selectively in cells expressing the respective ligands Fas (CD95) or the death receptors DR4 and DR5. Compared with CD8+ T cells, the spectrum of options for CD4+ T cells is larger. They are generally characterized as helper T, or TH cells, because they produce cytokines and express cell surface molecules that promote the effector function of other lymphocytes and phagocytes. Like CD8+ T cells, they are initially activated in secondary lymphoid tissues on contact with dendritic cells displaying the MHC−peptide complex (MHC class II, compared with MHC class I for CD8+ T-cell activation) bound by a specific TCR along with the proper costimulatory signals. On activation, different subsets of CD4+ effector T cells can be distinguished based on the preferential production of certain cytokines (see E-Table 46-1). TH1 T cells predominantly produce interferon-γ (IFN-γ) and tumor necrosis factor-α (TNF-α) and are involved in cell-mediated immunity, such as delayed-type hypersensitivity reactions. These cytokines, among other actions, promote macrophage activation that is critical for protective responses against intracellular pathogens such as mycobacteria and listeria. TH2 T cells preferentially produce IL-4, IL-5, and IL-13, cytokines that promote eosinophil maintenance, expansion, and tissue accumulation, as well as macrophage function; these are all important for host protection following infection with helminths, such as schistosomes and other worms. TH17 T cells produce IL-17, critical for neutrophil expansion and function, with killing of extracellular bacteria, such as streptococci, and pathogenic fungi.4 These cells may also produce IL-22 that promotes hostprotective function at barrier surfaces, such as the skin and gut. Follicular helper T (TFH) cells home to lymphoid follicles, where B cells congregate, where they express CD40 ligand (CD154) and other surface proteins along with cytokines, including IL-21, IL-4, and IFN-γ, that are critical for B-cell maturation to plasma cells and memory B cells. The decision as to which differentiation pathway to take is made during the early stages of naïve T-cell activation by antigen-presenting cells in secondary lymphoid organs. Pathway differentiation depends on several factors, including (1) the cytokines produced by the activating antigen-presenting cell and other innate cells in the


microenvironment, (2) the nature of costimulatory signals, and (3) the avidity of the TCR–MHC antigen interaction. CD4+ T-cell subset, or lineage, development is generally correlated with the expression of specific transcription factors (T-bet for TH1, GATA3 for TH2, RORγt for TH17, and Bcl6 for TFH cells). However, lineage commitment among differentiated CD4+ T cells is not absolute and is not terminal, and transition between different effector types is possible.

Regulatory T Cells

Depending on their cytokine profile, CD4+ T cells have the ability to crossregulate each other, influence T-cell differentiation, and suppress T-cell effector activity. Classic examples of T cells with regulatory activity generated during the normal immune response are IL-10– and transforming growth factor-β (TGF-β)–producing cells. In addition, specialized subsets of regulatory T (Treg) cells are characterized by expression of the transcription factor forkhead box P3 (Foxp3). Naturally occurring Foxp3+ Treg cells are generated during T-cell development in the thymus and recognize self-antigens. Foxp3+ Treg cells can also arise from conventional CD4+ T cells in the periphery. Natural and inducible Treg cells are in many ways indistinguishable, particularly because their development and function depend on Foxp3, and they are able to suppress T-cell expansion and constitutively express several cell surface markers, albeit markers that are not necessarily specific for Treg because activated T cells can also express them. Treg cells are important in peripheral tolerance, controlling the expansion of autoreactive T cells. They also play a role in immune responses to pathogens by virtue of their ability to suppress T-cell effector function and consequently downmodulate the inflammatory response incited by the former, a natural consequence of pathogen elimination. A principal difference between natural and induced Treg cells is that the latter largely survey mucosal and other environmentally exposed surfaces. Despite extensive studies in various models, the mechanism by which Treg cells function in vivo remains incompletely understood, although it is certainly a consequence of secretion of regulatory cytokines, like IL-10 and TGF-β, that can dampen inflammatory responses. Tregs may also express the T-cell molecule cytotoxic T-lymphocyte antigen (CTLA)-4 (CD152) that, like CD28, engages CD80 and CD86 on antigen-presenting cells. In contrast to CD28, which receives a positive signal from CD80 and CD86, leading to robust T-cell activation, engagement of these molecules on antigen-presenting cells by CTLA-4 on Tregs suppresses the ability of antigen-presenting cells to activate naïve T cells.

T-Cell Homeostasis

Effective immunity depends on the ability of the immune system to generate large numbers of antigen-specific T cells rapidly, yet the space in the T-cell compartment is limited. To avoid competition for space and resources and to prevent perturbation of T-cell diversity by lifelong exposure to antigens, the adaptive immune system employs several counterbalancing mechanisms. In the later stages of the activation process, a strong negative signal derives from interaction of CTLA-4 with CD80/CD86 on antigen-presenting cells. In addition, T cells undergo activation-induced cell death. Activated CD4+ T cells begin to secrete Fas ligand and acquire sensitivity to Fas-mediated death, inducing apoptotic suicide and fratricide in neighboring T cells. These mechanisms impose constraints in the early stages of the T-cell antigen response. Other mechanisms control the rapid decline of expanded antigen-specific T cells when elimination of the antigen has been achieved. Removal of the driving antigen causes a deprivation of cytokines and costimulatory molecules, and growth factor–deprived T cells die from apoptosis. It has been estimated that only 5% of the antigen-expanded population survives after antigen clearance, becoming memory cells that are poised to respond if the host is again challenged by the same offending pathogen.

B Lymphocytes B-Cell Development

B cells are generated in the bone marrow, like the thymus, a primary lymphoid organ. Lymphoid stem cells differentiate into distinctive B-lineage cells in the marrow, supported by a specialized microenvironment of nonlymphoid stromal cells supplying necessary chemokines, including stromal cell– derived factor 1 and cytokines (IL-7). Precursor B cells enter a process of tightly controlled sequential rearrangements of heavy chain and light chain immunoglobulin genes. On pre-B cells, the membrane µ chain is associated with a surrogate light chain to form a pre-B-cell receptor (BCR). Signals provided through this receptor induce proliferation of progeny that subsequently rearrange different light chain gene segments.


CHAPTER 46  The Adaptive Immune System  

Bone marrow

Lymph node


Pre-B cell


Stem cell


Bone marrow stromal cell

NaÏve mature B cell

Germinal center

Mantle zone Follicular dendritic cell

Memory cells Plasma cells Isotype switching

T-cell zone

B-cell proliferation Somatic hypermutation

Positive selection

FIGURE 46-3.  B-cell development and differentiation. The early stages of B-cell development occur in the bone marrow, with cells progressing through a developmental program determined by the rearrangement and expression of immunoglobulin (Ig) genes. Immature B cells with receptors for multivalent self-antigens die in the bone marrow. Surviving B cells coexpress IgD and IgM surface receptors. They are seeded into peripheral lymphoid organs, where they home to selected locations and receive signals to survive and become longer-lived naïve B cells. Antigen-binding B cells and antigen-presenting B cells that receive help from antigen-specific T cells are activated through membrane-bound and secreted molecules. Activated B cells migrate into the follicles, leading to the formation of germinal centers. B cells in germinal centers undergo somatic hypermutation of immunoglobulin genes; cells with high affinity for antigens presented on the surface of follicular dendritic cells are selected to differentiate into either memory B cells or plasma cells.

It is estimated that only 10% of B cells generated in the bone marrow reach the circulating pool. Losses are mostly due to negative selection and clonal deletion of immature B cells that express receptors directed against selfantigens. Cross-linking of surface IgM by multivalent self-antigens causes immature B cells to die. Such self-reactive B cells can be rescued from death by replacing the light chain with a newly rearranged light chain that is no longer self-reactive, a process named receptor editing.5 On maturation, B cells begin to express surface IgD. B cells positive for IgD and IgM are exported from the bone marrow and migrate to peripheral lymphoid tissues following a chemokine gradient, in a process analogous to the migration of naïve T cells from the thymus to the same tissues (Fig. 46-3). There, colocalization of both types of lymphocytes facilitates their interaction following pathogen challenge. This enables B cells to receive T-cell help for the former’s activation and subsequent function, including memory development and antibody secretion, required for responses to protein antigens.

B-Cell Stimulation

Mature, but naïve, B cells in secondary lymphoid organs are activated by soluble and cell-bound antigens to develop into antibody-secreting effector cells. B cells respond to a large variety of antigens, including proteins, polysaccharides, and lipids. Binding of antigen to cell surface IgM molecules induces BCR clustering, the initial step in B-cell activation. In addition to the antigenbinding immunoglobulin, the BCR comprises two proteins, Ig-α and Ig-β. The Ig-α/Ig-β heterodimer functions to transduce a signal and initiates the intracellular signaling cascade, analogous to the CD3 molecule of the TCR. Thus, the composition of the BCR, with ligand-binding and signal-transducing units, and the signaling events that lead to gene induction, are similar to those of the TCR. BCR triggering is enhanced by coreceptors, as for the TCR. The BCR-coreceptor complex is composed of CD81, CD19, and CD21, analogous to the TCR coreceptors CD4 and CD8. CD21 binds to complement fragments on opsonized antigens that are bound by the BCR, resulting in phosphorylation of the intracellular tail of CD19 by tyrosine kinases and augmentation of the BCR-mediated signal. Like naïve T cells, naïve B cells require accessory signals in addition to triggering of their antigen-binding receptor. They receive second signals either from follicular helper T cells or from microbial components. Microbial

constituents, such as bacterial polysaccharides, can induce antibody production in the absence of helper T cells, comprising thymus-independent, or T-independent, antigens.6 In contrast, in the case of protein antigens, which are thymus- or T-dependent, the initial BCR stimulation prepares the cell for subsequent interaction with follicular helper T cells. These activated B cells start to enter the cell cycle; upregulate cell surface molecules, such as CD80 and CD86, that provide costimulatory signals to T cells; and upregulate certain cytokine receptors. As such, these B cells are prepared to activate helper T cells and to respond to cytokines secreted by those T cells, but they cannot differentiate into antibody-producing cells in the absence of T-cell help. Survival and differentiation factors produced by myeloid cells, such as B-cell-activating factor (BAFF), also stimulate B cells and help to maintain the B-cell pool.7

B-Cell Differentiation

Differentiation of B cells activated by protein antigens depends on interaction with helper T cells. B cells use their antigen receptor not only to recognize antigens but also to internalize them. After processing endocytosed antigens, MHC class II–peptide complexes appear on the cell surface, where antigenspecific CD4+ T cells detect them. Also, B cells express costimulatory molecules and provide optimal conditions for T-cell activation. On activation, CD4+ T cells express CD154, also known as CD40 ligand, on their surface and are able to stimulate the CD40 molecule on their B-cell partner. CD40CD154 interaction is essential for subsequent B-cell proliferation and differentiation. Cytokines secreted by the helper T cells act in concert with CD154 to amplify B-cell differentiation and to determine the antibody type by controlling isotype switching. Isotypes greatly influence the versatility of antibodies as effector molecules, and cytokines drive isotype switching by stimulating the transcriptional activation of heavy chain constant region genes and enabling switching from transcription of the IgM heavy chain gene to that of IgG, IgA, or IgE. T-cell-dependent B-cell differentiation and maturation take place in germinal centers, specialized areas in secondary lymphoid tissues where B cells rapidly proliferate, with mutation of the variable, or antigen-binding portion, of their immunoglobulin surface receptors (BCRs) (see Fig. 46-3). Those B cells bearing receptors with the highest affinity for antigen are selected for

CHAPTER 46  The Adaptive Immune System  

survival with the help of specific signals delivered by follicular helper T cells, whereas those with lesser affinity die by apoptosis. This process enables affinity maturation of B cells that most efficiently bind antigen and thereby facilitate its removal. As somatic hypermutation and affinity maturation proceed in the germinal center, isotype class switching of the immunoglobulin receptors is also occurring.8

Lymphocytes and Lymphoid Tissue

The initiation of adaptive immune responses depends on rare antigen-specific T cells and B cells meeting antigen-presenting cells and their relevant antigen. The recognition of a specific antigen in the tissue by uncommon T cells has a low probability, and it is unlikely that sufficient numbers of antigenpresenting cells and lymphocytes can be brought together to provide crucial momentum. The immune system uses specialized lymphoid microstructures to bring antigens to the site of lymphocyte traffic and accumulation. Secondary lymphoid organs include the spleen for blood-borne antigens, the lymph nodes for antigens encountered in peripheral tissues, and the mucosaassociated, bronchial-associated, and gut-associated lymphoid tissues, where antigens from epithelial surfaces are collected. Lymphocytes circulate through secondary lymphoid organs, constantly searching for their antigen. Their homing to lymph nodes is facilitated by specialized microvessels, called high endothelial venules, which provide the proper structure for them to leave the circulation and enter the tissue. Secondary lymphoid tissues have developed several strategies to sequester the relevant antigen. Antigens in peripheral tissue are encountered first by dendritic cells that, after activation, are mobilized to transport antigens into the local lymph nodes by the draining lymph. These antigen-bearing dendritic cells enter the lymph nodes through the afferent lymphatic vessel and settle in the T-cell-rich zones to present processed antigens to T cells. The net result of this process is an accumulation and concentration of the antigen in an environment that can be readily screened by infrequent antigen-specific T cells. B cells are segregated from T cells in the lymph nodes and are localized in follicles. If, on antigen engagement, B cells find their cooperating (cognate) T cells at the borders of the T-cell-rich areas and the follicle, they receive cues to enter germinal centers along with their cognate follicular helper T cells. Germinal centers contain a network of follicular dendritic cells that capture particulate antigen or immune complexes on the cell surface. This unprocessed antigen is taken up by antigen-specific B cells, processed and presented, and recognized by antigen-specific TFH cells. These T cells provide cytokines and cell-cell contact signals to support the germinal center reaction, a process that includes somatic hypermutation, affinity selection, and isotype switching (see Fig. 46-3). Germinal centers are essential for generating long-lived antibody-secreting plasma cells and memory B cells. Lymphoid organ development is highly dependent on environmental cues. The symbiotic relationship between the host immune system and microorganisms is best exemplified in the gastrointestinal tract. Development of gutassociated lymphoid tissue is absolutely dependent on bacterial colonization. Increasing evidence suggests that host-symbiont interactions regulate adaptive immune functions throughout life. Disturbances in the bacterial microbiota and failure to maintain intestinal homeostasis are important in diverse diseases, including inflammatory bowel disease (Chapter 143) and HIVassociated immune defects.


An important consequence of adaptive immunity is the generation of immunologic memory, the basis for long-lived protection after a primary infection. Memory induction by vaccination is one of the landmark successes in medicine. Immunologic memory is defined as the ability to respond more rapidly and effectively to pathogens that have been encountered previously. The bases of immunologic memory are qualitative and quantitative changes in antigen-specific T cells and B cells. As a direct result of clonal expansion and selection in antigen-driven responses, the frequencies of antigen-specific memory B cells and memory T cells are increased 10-fold to 1000-fold compared with the naïve repertoires. The mechanisms through which memory T cells and B cells escape clonal downsizing in the terminal stages of the primary immune response are consequences of upregulation of a selected group of transcription factors that ensure survival. The enrichment of antigen-specific B cells and T cells enhances the sensitivity of the system to renewed challenges and provides a head start of 4 to 10 cell divisions. In addition to increased frequencies, memory T cells and B cells are functionally different from their naïve counterparts. Memory cells are long-lived and survive in the presence of certain cytokines without the need for continuous antigenic


stimulation, guaranteeing immunologic memory for the life expectancy of the individual cell. Memory B cells produce predominantly IgG and IgA antibodies with evidence of somatic hypermutation and high affinity for the antigen. Cell surface expression of high-affinity antibodies allows more efficient antigen uptake, which enhances the crucial interaction with T cells. On antigen encounter, memory B cells change to antibody-secreting plasma cells, or re-enter the germinal center, where the high affinity of their immunoglobulin receptor gives them a competitive advantage over naïve B cells in antigen binding, leading to progressive affinity maturation of somatically mutated antibody molecules. Because the TCR does not undergo isotype switching or affinity maturation, memory T cells are more difficult to distinguish from naïve or effector T cells. In contrast to effector cells, memory T cells lack activation markers and need antigen stimulation to resume effector functions. In contrast to naïve T cells, memory T cells have a lower activation threshold and are less dependent on costimulatory signals. In essence, their requirements for antigen stimulation are fewer, and their clonal size is larger, permitting fast, efficient responses to secondary antigen encounters. Also, memory T cells resume effector functions without having to undergo cell divisions.

Immunologic Tolerance and Autoimmunity

Unresponsiveness to self is a fundamental property of the immune system and is a condition, sine qua non, to maintain tissue integrity of the host. Self/ nonself distinction is relatively straightforward for the innate immune system, in which receptors to nonself molecules are genetically encoded and evolutionarily selected. Self/nonself discrimination is much more complex for the adaptive immune system, in which antigen-specific receptors are generated randomly and the entire spectrum of antigens can be recognized. Thus, the adaptive immune system must acquire the ability to distinguish between self and nonself. Several different mechanisms are used, collectively called tolerance. Tolerance is antigen specific; its induction requires the recognition of antigen by lymphocytes in a defined setting. Failure of self-tolerance results in immune responses against self-antigens. Such reactions are called autoimmunity and may give rise to chronic inflammatory autoimmune disease. Central and peripheral tolerance mechanisms can be distinguished. In central tolerance, self-reactive lymphocytes are deleted during development. This process of negative selection is particularly important for T cells. During thymic development, T cells that recognize antigen with high affinity, in particular antigens that are constitutively expressed on antigen-presenting cells, are deleted. Central tolerance for B cells follows the same principles. Recognition of antigen by developing B cells in the bone marrow induces apoptosis, or receptor editing that replaces the self-reactive receptor with one containing the product of a newly rearranged light chain gene. Negative selection is particularly important for B cells that recognize multivalent antigens because they do not depend on T-cell help and cannot be controlled peripherally. Not all self-reactive T cells are centrally purged from the repertoire; certain antigens are not encountered at sufficient densities in the thymus. Also, all T cells have some degree of self-reactivity, which is necessary for positive selection in the thymus and for peripheral survival. Mechanisms of peripheral T-cell tolerance include (1) anergy, (2) peripheral deletion, (3) clonal ignorance, and (4) suppression of immune responses by regulatory T cells. T-cell anergy is transient and is actively maintained. It is induced if CD4+ T cells recognize antigens presented by MHC class II molecules without receiving costimulatory signals. In general, costimulatory molecules such as CD80 and CD86 are restricted to antigen-presenting cells, and their expression is dependent on microbial recognition, leading to activation of the antigenpresenting cells. MHC-peptide presentation to T cells by immature or inactivated, resting antigen-presenting cells or on any cell other than peripheral antigen-presenting cells results in anergy because these cells typically lack expression of costimulatory molecules. Tissue-residing immature dendritic cells need to be activated by cytokines or recognition of pathogen-associated molecular patterns (PAMPs) to stimulate and not to anergize T cells. A second tolerance mechanism, peripheral deletion, is induced as a consequence of hyperstimulation. Hyperstimulation of T cells (e.g., by high doses of antigen and high concentrations of IL-2) preferentially activates proapoptotic pathways and causes elimination of the responding T-cell specificity. This mechanism may be responsible for the elimination of T cells specific for plentiful peripheral self-antigens and for foreign antigens abundantly present during infection. Whereas induction of anergy and activationinduced cell death are active consequences of antigen recognition, the third tolerance mechanism, clonal ignorance, is less well understood. Clonal

ignorance is defined as the presence of self-reactive lymphocytes that fail to recognize or to respond to peripheral antigens. These cells remain responsive to antigenic challenge if given in the right setting. An example of clonal ignorance is nonresponsiveness to sequestered antigens that are not accessible to the immune system. Other mechanisms must exist, however, because clonal ignorance has also been shown for accessible antigens. Fourth, Treg cells play a pivotal role in maintaining peripheral tolerance. During an immune response, T cells can acquire the ability to produce regulatory cytokines, such as TGF-β, IL-10, or IL-4, that dampen or suppress immune responses. A dedicated subset of Treg cells, Foxp3 CD4+ T cells, has been identified and characterized. Harnessing the frequencies and function of these cells may offer a promising approach to restoring peripheral tolerance in treating autoimmune diseases or facilitating transplantation tolerance; their elimination or functional suppression may potentiate cancer immunotherapy. A critically important mechanism of peripheral tolerance of B cells is maintained through the absence of T-cell help. B cells require signals from T cells to differentiate into effector cells. B lymphocytes that recognize self-antigens in the periphery in the absence of T-cell help are rendered anergic or are unable to enter lymphoid follicles, where they could receive T-cell help, effectively excluding them from immune responses. Generation and maintenance of self-tolerance can fail, in which case autoimmune responses are generated. Overall, chronic inflammatory diseases induced by tolerance failure occur in about 5% of the general population. Given the complexity of regulation, it is surprising that autoimmune diseases are not more frequent. It is thought that most autoimmune diseases result from dysfunction of the adaptive immune system, although activation of the innate immune system can set the stage for a self-reactive adaptive immune response. Many models of autoimmunity rely on the hypothesis that peripheral anergy is broken. Aberrant expression of costimulatory molecules on nonprofessional antigen-presenting cells or inappropriate activation of tissue-residing dendritic cells sets the stage for the induction of “forbidden” T-cell responses. Also, autoreactive B cells that recognize self-antigen complexed with foreign antigen may engulf this complex and receive help from T cells specific for the foreign antigen. Autoimmunity also may emerge if antigen ignorance is broken. This could happen if tissue barriers break down and antigens that are usually sequestered from the immune system, such as antigens from the central nervous system or the eye, become accessible. Tolerance mechanisms of anergy or clonal ignorance can also fail if a foreign antigen is sufficiently different from a self-antigen to initiate an immune response but sufficiently similar for activated T cells to elicit T-cell and B-cell effector functions (molecular mimicry). GENERAL REFERENCES For the General References and other additional features, please visit Expert Consult at

CHAPTER 46  The Adaptive Immune System  

GENERAL REFERENCES 1. Klein L, Kyewski B, Allen PM, et al. Positive and negative selection of the T cell repertoire: what thymocytes see (and don’t see). Nat Rev Immunol. 2014;14:377-391. 2. Fu G, Rybakin V, Brzostek J, et al. Fine-tuning T cell receptor signaling to control T cell development. Trends Immunol. 2014;35:311-318. 3. Hubo M, Trinschek B, Kryczanowsky F, et al. Costimulatory molecules on immunogenic versus tolerogenic human dendritic cells. Front Immunol. 2013;4:82. 4. Peters A, Yosef N. Understanding Th17 cells through systematic genomic analyses. Curr Opin Immunol. 2014;28:42-48.


5. Luning Prak ET, Monestier M, Eisenberg RA. B cell receptor editing in tolerance and autoimmunity. Ann N Y Acad Sci. 2011;1217:96-121. 6. Bortnick A, Allman D. What is and what should always have been: long-lived plasma cells induced by T cell-independent antigens. J Immunol. 2013;190:5913-5918. 7. Rickert RC, Jellusova J, Miletic AV. Signaling by the tumor necrosis factor receptor superfamily in B-cell biology and disease. Immunol Rev. 2011;244:115-133. 8. Matthews AJ, Zheng S, DiMenna LJ, et al. Regulation of immunoglobulin class-switch recombination: choreography of noncoding transcription, targeted DNA deamination, and long-range DNA repair. Adv Immunol. 2014;122:1-57.


CHAPTER 46  The Adaptive Immune System  

REVIEW QUESTIONS 1. IgG antibodies are characterized by all of the following except which one? A. Two heavy chains and two light chains B. A common γ chain C. Disulfide bonds D. Subclasses Answer: B  IgG molecules consist of two identical heavy and two identical light chains, linked by disulfide bonds. IgG and IgA molecules include distinct subclasses encoded by distinct genomic DNA sequences. A common γ chain is a signaling component of certain cytokine and Fc receptors but is not a component of IgG antibodies. 2. CD4+ T cells recognize antigenic peptides presented on which of the following? A. Epithelial cells B. Major histocompatibility complex (MHC) class I molecules C. MHC class II molecules D. Costimulatory molecules E. CD80 Answer: C  Antigenic peptides processed by specialized antigen-presenting cells associate with the peptide-binding cleft of MHC class II molecules and are recognized by the T-cell receptor expressed on CD4+ T cells. CD8+ T cells recognized antigenic peptides associated with MHC class I molecules. CD80 is an example of a costimulatory molecule expressed on antigen-presenting cells that supports the activation of T cells that have interacted with peptideMHC complexes. Epithelial cells do not typically activate CD4+ T cells. 3. Which of the following functions is not provided by differentiated T cells? A. Production of interleukin-17 (IL-17) B. Lysis of virus-infected target cells C. Help for B-cell activation and differentiation D. Production of BAFF E. Production of interferon-γ Answer: D  The cytokine milieu of T cells contributes to differentiation of those cells to provide various effector functions. These include support for B-cell activation and differentiation (through CD154-CD40 interactions and production of IL-21), secretion of cytokines (interferon-γ) that support macrophage activation and delayed-type hypersensitivity reactions, induction of inflammatory responses (IL-17), and mediating the killing of virus-infected cells. Myeloid cells, rather than T cells, are the major producers of BAFF.

4. B cells minimize self-reactivity through a process called which of the following? A. Isotype switching B. Costimulation C. Somatic hypermutation D. Recombination E. Receptor editing Answer: E  Isotype switching, somatic hypermutation, and recombination are processes that promote the diversity of the B-cell repertoire. Costimulation of T-cell activation supports amplification of an adaptive immune response. Receptor editing modifies the sequence of the B-cell receptor to avoid production of autoreactive B cells. 5. Immunologic tolerance is mediated by all of the following except which of the following? A. Thymic deletion of T cells that recognize antigen with high affinity B. Regulatory T cells C. Toll-like receptor activation D. T-cell activation in the absence of costimulation E. It is mediated by all of the above. Answer: C  Central and peripheral tolerance mechanisms include thymic deletion of self-reactive T cells, T-cell activation without costimulation, and control of self-reactive cells by regulatory T cells. Toll-like receptor activation is an important feature of innate immune system activation.


CHAPTER 47  Mechanisms of Immune-Mediated Tissue Injury  

Effector mechanisms that eliminate pathogens in adaptive immune responses are essentially identical to those of innate immunity. The specific antigen recognition feature of the adaptive immune response seems to have been appended to the preexisting innate defense system. As a result, the inflammatory cells and molecules of the innate immune system are essential for the effector functions of B and T lymphocytes. In addition to initiating protective responses, they mediate tissue injury in allergy, hypersensitivity, and autoimmunity.

Effector Mechanisms




The adaptive immune response is a crucial component of host defense against infection. Its distinguishing and unique feature is the ability to recognize pathogens specifically, based on clonal selection of lymphocytes bearing antigen-specific receptors. Antigens unassociated with infectious agents also may elicit adaptive immune responses. Many clinically important diseases are characterized by normal immune responses directed against an inappropriate antigen, typically in the absence of infection. Immune responses directed at noninfectious antigens occur in allergy, in which the antigen is an innocuous foreign substance, and in autoimmunity, in which the response is to a self-antigen.

Effector actions of antibodies depend on recruiting cells and molecules of the innate immune system. Antibodies are adapters that bind antigens to nonspecific inflammatory cells and direct their destructive effector responses. Antibodies also activate the complement system, which enhances opsonization of antigens, recruits phagocytic cells, and amplifies (or “complements”) antibody-triggered damage. The isotype or class of antibodies produced determines which effector mechanisms are engaged. Cell-bound receptors for immunoglobulin (Ig) constitute the link between humoral and cellular aspects of the immune cascade and play an integral part in the process by which foreign and endogenous opsonized material is identified and destroyed. These cell-based binding sites for antibodies, termed Fc receptors, interact with the constant region (Fc portion) of the immunoglobulin heavy chain of a particular antibody class regardless of its antigen specificity. Accessory cells that lack intrinsic specificity, such as neutrophils, macrophages, and mast cells, are recruited to participate in inflammatory responses through the interaction of their Fc receptors with antigen-specific antibodies. Distinct receptors for different immunoglobulin isotypes are expressed on different effector cells. Receptors for IgG (FcγRs) are a diverse group of receptors expressed as hematopoietic cell surface molecules on phagocytes (macrophages, monocytes, neutrophils), platelets, mast cells, eosinophils, and natural killer (NK) cells. FcγRs often are expressed as stimulatory and inhibitory pairs.1 Triggering of stimulatory FcγRs initiates a series of events, including phagocytosis; antibody-dependent, cell-mediated cytotoxicity; secretion of granules; and release of inflammatory mediators, such as cytokines, reactive oxidants, and proteases. Extensive structural diversity among FcγR family members leads to differences in binding capacity, signal transduction pathways, and cell type–specific expression patterns. This diversity allows IgG complexes to activate a broad program of cell functions relevant to inflammation, host defense, and autoimmunity. Phagocyte activation is triggered by stimulatory FcγRs, facilitating the recognition, uptake, and destruction of antibodycoated targets, whereas multivalent IgG binding to FcγRs on platelets leads to platelet aggregation and thrombosis, and binding to FcγRs on NK cells mediates cytotoxicity of antibody-coated targets. IgE binds to high-affinity FcεRs on mast cells, basophils, and activated eosinophils.2 In contrast to FcγRs, which are low affinity and bind to multivalent IgG rather than circulating individual IgG molecules, FcεRs can bind monomeric IgE. A single mast cell may be armed with IgE molecules specific for different antigens, all bound to surface FcεRs. Mast cells, localized beneath the mucosa of the gastrointestinal and respiratory tracts and the dermis of the skin, await exposure to multivalent antigens, which cross-link surface IgE bound to FcεRs and cause release of histamine-containing granules and generation of cytokines and other inflammatory mediators. IgEmediated activation of eosinophils, cells normally present in the connective tissue of underlying respiratory, urogenital, and gut epithelium, leads to the release of highly toxic granule proteins, free radicals, and chemical mediators such as prostaglandins, cytokines, and chemokines. These amplify local inflammatory responses by activating endothelial cells and recruiting and activating more eosinophils and leukocytes. Prepackaged granules and highaffinity FcεRs that bind to free monomeric IgE enable an immediate response to pathogens or allergens at the first site of entry, a location where FcεRbearing cells reside. Inhibitory FcγRs, which modulate activation thresholds and terminate stimulating signals, are key elements in the regulation of effector function. Given that inhibitory and stimulatory Fc receptors are often coexpressed on the same cells, the effector response to a specific stimulus in a particular cell represents the balance between stimulatory and inhibitory signals. Inhibitory FcγRs can dampen responses triggered by FcεRs on mast cells and FcγRmediated inflammation at sites of immune complex deposition. Effector activities targeted by IgG and IgM also may be mediated by components of the complement system (Chapter 50). Antigen-bound multimeric immunoglobulin can initiate activation of the classic pathway of


CHAPTER 47  Mechanisms of Immune-Mediated Tissue Injury  

complement, causing enhanced phagocytosis of antigen-antibody complexes, increased local vascular permeability, and recruitment and activation of inflammatory cells. The target of injury is specified by the antibody, and the extent of damage is determined by the synergistic activities of immunoglobulin and complement. Antigen-specific effector T cells also may initiate tissue injury. On exposure to an appropriate antigen, memory T cells are stimulated to release cytokines and chemokines that activate local endothelial cells and recruit and activate macrophages and other inflammatory cells. The effector cells directed by T-cell-derived cytokines, or cytolytic T cells themselves, mediate tissue damage. T helper 1 (TH1) cells produce interferon-γ (IFN-γ) and activate macrophages to cause injury, whereas TH2 cells produce interleukin-4 (IL-4), IL-5, and eotaxin (an eosinophil-specific chemokine) and trigger inflammatory responses in which eosinophils predominate. TH17 cells secrete several effector molecules, including IL-17, which act on both immune and nonimmune cells to trigger differentiation; release of antimicrobial molecules, cytokines, and chemokines; and recruitment to sites of inflammation.3 New TH effector subsets have recently been identified, including follicular T helper cells (TFH), which provide help to B cells in germinal centers and thus are key regulators of humoral responses and antibody production.


In predisposed individuals, innocuous environmental antigens may stimulate an adaptive immune response, immunologic memory, and, on subsequent

exposure to the antigen, inflammation. These “overreactions” of the immune system to harmless environmental antigens (allergens), called hypersensitivity or allergic reactions, produce tissue injury and can cause serious disease. Hypersensitivity reactions are grouped into four types according to the effector mechanisms by which they are produced (Table 47-1). The effectors for types I, II, and III hypersensitivity reactions are antibody molecules, whereas type IV reactions are mediated by antigen-specific effector T cells.4 Autoimmune disease is characterized by the presence of antibodies and T cells specific for self-antigens expressed on target tissues. The mechanisms of antigen recognition and effector function that lead to tissue damage in autoimmune disease are similar to the mechanisms elicited in response to pathogens and environmental antigens. These mechanisms resemble certain hypersensitivity reactions and may be classified accordingly (Table 47-2). Autoimmune disease caused by antibodies directed against cell surface or extracellular matrix antigens corresponds to type II hypersensitivity reactions; disease caused by formation of soluble immune complexes that subsequently are deposited in tissue corresponds to type III hypersensitivity; and disease caused by effector T cells corresponds to type IV hypersensitivity. Typically, several of these pathogenic mechanisms are operative in autoimmune disease. However, IgE responses are not associated with damage in autoimmunity.

Type I Hypersensitivity Reactions

Type I hypersensitivity reactions (Fig. 47-1) are triggered by the interaction of antigen with antigen-specific IgE bound to FcεRs on mast cells, which




TYPE III (IgG ANTIBODY) Soluble antigen


TH2 Cells

TH17 Cells

T Cells


Soluble antigen allergen

Cell- or matrixassociated antigen

Soluble antigen

Soluble antigen

Soluble antigen

Cell-associated antigen

Effector mechanism

FcεRI- or FcγRIIIdependent mast cell activation, with release of mediators/ cytokines

FcγR+ cells FcγR+ cells, (phagocytes, NK complement cells), complement

Macrophage activation

Eosinophil activation

Macrophage activation Neutrophil activation

Direct cytotoxicity


Systemic anaphylaxis, Certain drug Arthus reaction and Contact dermatitis, Chronic allergic Contact dermatitis, Contact dermatitis asthma, allergic reactions and other immune tuberculin inflammation atopic dermatitis, (e.g., poison ivy), rhinitis, urticaria, reactions to complex–mediated reaction (e.g., chronic asthma, reactions to angioedema incompatible reactions (e.g., asthma, rheumatoid certain blood transfusions serum sickness, chronic allergic arthritis virus-infected subacute bacterial rhinitis) cells, some endocarditis) instances of graft rejection

*Hypersensitivity reactions were classified into four types by Coombs and Gell (1963) and modified by Janeway and colleagues (2001). FcγR = Fc receptor for immunoglobulin G; FcεR = Fc receptor for immunoglobulin E; NK = natural killer. From Coombs RRA, Gell PGH: Classification of allergic reactions responsible for clinical hypersensitivity and disease. In: Gell PGH, Coombs RA, eds. Clinical Aspects of Immunology. Oxford, UK: Blackwell; 1963; and Janeway C, Travers P, Walport M, Shlomchick M: Immunobiology: The Immune System in Health and Disease. 5th ed. New York: Garland Publishing; 2001.




TYPE II Antibody against cell surface antigens Antibody against receptors Antibody against matrix antigens

Autoimmune hemolytic anemia Autoimmune thrombocytopenic purpura Graves disease Myasthenia gravis Goodpasture syndrome

Rh blood group antigens, I antigen Platelet integrin glycoprotein IIb/IIIa Thyroid-stimulating hormone receptor (agonistic antibodies) Acetylcholine receptor (antagonistic antibodies) Basem*nt membrane collagen (α3-chain of type IV collagen)

Pemphigus vulgaris

Epidermal cadherin (desmoglein)

Mixed essential cryoglobulinemia Systemic lupus erythematosus

Rheumatoid factor IgG complexes (with or without hepatitis C antigens) DNA, histones, ribosomes, binuclear proteins

Insulin-dependent diabetes mellitus Rheumatoid arthritis Multiple sclerosis

Pancreatic B-cell antigen Unknown synovial joint antigen Myelin basic protein, proteolipid protein

TYPE III Immune complex diseases TYPE IV T-cell-mediated diseases


CHAPTER 47  Mechanisms of Immune-Mediated Tissue Injury  

Cell-associated antigen

IgE FcεRI Soluble antigen

Mast cell Erythrocyte Histamine Proteolytic enzymes Cytokines (IL-4, IL-5, TNF-α) Leukotrienes Chemokines

IgG FcγR

MΦ Tissue damage FIGURE 47-1.  Type I hypersensitivity. Type I responses are mediated by immunoglobulin E (IgE), which induces mast cell activation. Cross-linking of the Fc receptor for IgE (FcεR) on mast cells, triggered by the interaction of multivalent antigen with antigenspecific IgE bound to FcεR, causes the release of preformed granules containing histamine and proteases. Cytokines, chemokines, and lipid mediators are synthesized after cell activation. IL = interleukin; TNF = tumor necrosis factor.

causes mast cell activation. Proteolytic enzymes and toxic mediators, such as histamine, are released immediately from preformed granules, and chemokines, cytokines, and leukotrienes are synthesized after activation. Together, these mediators increase vascular permeability, break down tissue matrix proteins, promote eosinophil production and activation (IL-3, IL-5, and granulocyte-macrophage colony-stimulating factor [GM-CSF]), and cause influx of effector leukocytes (tumor necrosis factor-α [TNF-α], plateletactivating factor, and macrophage inflammatory protein [MIP-1]), constriction of smooth muscle, stimulation of mucus secretion, and amplification of TH2 cell responses (IL-4 and IL-13). Eosinophils and basophils, activated through cell surface FcεRs, rapidly release highly toxic granular proteins (major basic protein, eosinophil peroxidase, and collagenase) and, over a longer period, produce cytokines (IL-3, IL-5, and GM-CSF), chemokines (IL-8), prostaglandins, and leukotrienes that activate epithelial cells, leukocytes, and eosinophils to augment local inflammation and tissue damage. FcεR-bearing effectors act in a coordinated fashion. The immediate allergic inflammatory reaction initiated by mast cell products is followed by a latephase response that involves recruitment and activation of eosinophils, basophils, and TH2 lymphocytes.5 The manifestations of IgE-mediated reactions depend on the site of mast cell activation. Mast cells reside in vascular and epithelial tissue throughout the body. In a sensitized host (an individual with IgE responses to antigens), re-exposure to antigen leads to type I hypersensitivity responses only in the mast cells exposed to the antigen. Inhalation of antigens produces bronchoconstriction and increased mucus secretion (asthma and allergic rhinitis); ingestion of antigens causes increased peristalsis and secretion (diarrhea and vomiting); and the presence of subcutaneous antigens initiates increased vascular permeability and swelling (urticaria and angioedema). Blood-borne antigens cause systemic mast cell activation, increased capillary permeability, hypotension, tissue swelling, and smooth muscle contraction—the characteristics of systemic anaphylaxis.

Type II Hypersensitivity Reactions

Type II hypersensitivity reactions (Fig. 47-2) are caused by chemical modification of cell surface or matrix-associated antigens that generates “foreign” epitopes to which the immune system is not tolerant. B cells respond to this antigenic challenge by producing IgG, which binds to these modified cells and renders them susceptible to destruction through complement activation, phagocytosis, and antibody-dependent cytotoxicity. This phenomenon is seen clinically when drugs interact with blood constituents and alter their cellular antigens. Hemolytic anemia caused by immune-mediated destruction of erythrocytes (Chapter 160) and thrombocytopenia caused by destruction of platelets (Chapter 172), both type II hypersensitivity reactions, are adverse effects of certain drugs. Chemically reactive drug molecules bind covalently to the surface of red cells or platelets creating new epitopes that in a small subset of individuals are recognized as foreign antigens by the immune system and stimulate production of IgM and IgG antibodies reactive with the conjugate of drug and cell surface protein. Penicillin-specific IgG binds to penicillin-modified proteins on red blood cells and triggers activation of the complement cascade. Activation of complement components C1 through C3 results in covalent binding of C3b to

FIGURE 47-2.  Type II hypersensitivity. Type II responses are mediated by immunoglobulin G (IgG) directed against cell surface or matrix antigens, which initiates effector responses through the Fc receptor for IgG (FcγR) and complement. The relative contributions of these pathways vary with the IgG subclass and the nature of the antigen. Only FcγR-mediated phagocytosis by macrophages (MΦ) is depicted in this figure. Activation of complement components would result in binding of C3b to the red blood cell membrane, rendering red blood cells susceptible to phagocytosis and leading to formation of the membrane attack complex and cell lysis.

the red cell membrane and renders circulating red cells susceptible to phagocytosis by FcγR and complement receptor–bearing macrophages in the spleen or liver. Activation of complement components C1 through C9 and formation of the membrane attack complex cause intravascular lysis of red cells. The factors that predispose only some people to drug-induced type II hypersensitivity reactions are unknown. Penicillin, quinidine, and methyldopa have been associated with hemolytic anemia and thrombocytopenia through this mechanism. Another example is heparin-induced thrombocytopenia or thrombosis, a severe, life-threatening complication that occurs in 1 to 3% of patients exposed to heparin (Chapter 172). Interactions among heparin, human platelet factor 4, antibodies to the human platelet factor 4– heparin complex, platelet FcγRIIA, and splenic FcγRs (which remove opsonized platelets) are involved in the pathogenesis of this disease. Autoantibodies directed at antigens on the cell surface or extracellular matrix cause tissue damage by mechanisms similar to type II hypersensitivity reactions. IgG or IgM antibodies against erythrocytes lead to cell destruction in autoimmune hemolytic anemia because opsonized cells (coated with IgG or IgM and complement) are removed from the circulation by phagocytes in the liver and spleen or are lysed by formation of the membrane attack complex. Platelet destruction in autoimmune thrombocytopenic purpura occurs through a similar process. Because nucleated cells express membrane-bound complement regulatory proteins, they are less sensitive to lysis through the membrane attack complex, but when coated with antibody, they become targets for phagocytosis or antibody-dependent cytotoxicity. This mechanism is responsible for autoimmune and alloimmune neutropenia (Chapter 167). IgM and IgG antibodies recognizing antigens within tissue or binding to extracellular antigens cause local inflammatory damage through FcγR and complement mechanisms. Pemphigus vulgaris (Chapter 439) is a serious blistering disease that results from a loss of adhesion between keratinocytes caused by autoantibodies against the extracellular portions of desmoglein 3, an intercellular adhesion structure of epidermal keratinocytes. Another example of a type II hypersensitivity reaction is Goodpasture disease (Chapter 121), in which antibodies against the α3-chain of type IV collagen (the collagen in basem*nt membranes) are deposited in glomerular and lung basem*nt membrane. Tissue-bound autoantibodies activate monocytes, neutrophils, and basophils through FcγRs, initiating release of proteases, reactive oxidants, cytokines, and prostaglandins. Local activation of complement, particularly C5a, recruits and activates inflammatory cells and amplifies tissue injury. Neighboring cells are lysed by assembly of the membrane attack complex or by FcγR-initiated, antibody-dependent cytotoxicity. Autoantibodies against cell surface receptors produce disease by stimulating or blocking receptor function. In myasthenia gravis (Chapter 422), autoantibodies against the acetylcholine receptors on skeletal muscle cells bind the receptor and induce its internalization and degradation in lysosomes, reducing the efficiency of neuromuscular transmission and causing pro­ gressive muscle weakness. In contrast, Graves disease (Chapter 226) is


CHAPTER 47  Mechanisms of Immune-Mediated Tissue Injury  

Proteolytic enzymes Reactive oxidants Cytokines Chemokines Leukotrienes

Ag Immune complex

Complement activation C3a C5a Chemotaxis

FcγR activation Monocyte


Eosinophil activation Ag


Mediator release Oxidants Proteases Chemokines Cytokines

Tissue damage



Complement receptors

T H1

Mast cell

Proteolytic enzymes Cytokines Chemokines Leukotrienes

TH2 Toxic proteins

IL-4 IL-5 Eotaxin

FIGURE 47-3.  Type III hypersensitivity. Type III responses are mediated by immunoglobulin G (IgG) directed against soluble antigens. Localized deposition of immune complexes activates mast cells, monocytes, neutrophils, and platelets bearing the Fc receptor for IgG (FcγR), and initiates the complement cascade, all effectors of tissue damage. Generation of complement components C3a and C5a recruits and stimulates inflammatory cells and amplifies effector functions. PMN = polymorphonuclear leukocyte (also called neutrophil).


Tissue damage

Direct cytotoxicity

Cell-associated antigen

characterized by autoantibodies that act as agonists. Autoantibodies to thyroid-stimulating hormone receptors bind the receptor, mimicking the natural ligand, inducing thyroid hormone overproduction, disrupting feedback regulation, and causing hyperthyroidism.

Type III Hypersensitivity Reactions

Type III hypersensitivity reactions (Fig. 47-3) are caused by tissue deposition of small soluble immune complexes that contain antigens and high-affinity IgG antibodies directed at these antigens. Localized deposition of immune complexes activates FcγR-bearing mast cells and phagocytes and initiates the complement cascade, all effectors of tissue damage.6 Immune complexes are generated in all antibody responses. The formation and the fate of immune complexes depend on the biophysical and immunologic properties of the antigen and the antibody. These properties include the size, net charge, and valence of the antigen; the class and subclass of the antibody; the affinity of the antibody-antigen interaction; the net charge and concentration of antibody; the molar ratio of available antigen and antibody; and the ability of the immune complex to interact with the proteins of the complement system. The lattice size of the immune complex is influenced strongly by the physical size and valence of the antigen, the association constant of antibody for that antigen, the molar ratio of antigen and antibody, and the absolute concentrations of the reactants. Larger aggregates fix complement more efficiently, present a broader multivalent array of ligands for complement and FcγRs to bind, and are taken up more readily by mononuclear phagocytes in the liver and spleen and thereby removed from the circulation. Smaller immune complexes, which form in antigen excess—as occurs early in an immune response—circulate in the blood and are deposited in blood vessels, where they initiate inflammatory reactions and tissue damage through interactions with FcγRs and complement receptors. Serum sickness is a systemic type III hypersensitivity reaction, historically described in patients injected with therapeutic horse antiserum for the treatment of bacterial infections. In general, serum sickness occurs after the injection of large quantities of a soluble antigen. Clinical features include chills, fever, rash, urticaria, arthritis, and glomerulonephritis. Disease manifestations become evident 7 to 10 days after exposure to the antigen, when antibodies are generated against the foreign protein and form immune complexes with these circulating antigens. Immune complexes are deposited in blood vessels, where they activate phagocytes and complement, producing widespread tissue injury and clinical symptoms. The effects are transient, however, and resolve after the antigen is cleared. A syndrome similar to serum sickness occurs in chronic infections in which pathogens persist in the face of continued immune response. In subacute bacterial endocarditis (Chapter 76), antibody production continues

FIGURE 47-4.  Type IV hypersensitivity. Type IV responses are mediated by T cells through three different pathways. In the first, type 1 helper T (TH1) cells recognize soluble antigens (Ag) and release interferon-γ (IFN-γ) to activate effector cells, in this case macrophages (MΦ), and cause tissue injury. In TH2-mediated responses, eosinophils predominate. TH2 cells produce cytokines to recruit and activate eosinophils, leading to their degranulation and tissue injury. In the third pathway, damage is caused directly by cytolytic T lymphocytes (CTL). IL = interleukin.

but fails to eliminate the infecting microbes. As the pathogens multiply, generating new antigens, immune complexes form in the circulation and are deposited in small blood vessels, where they lead to inflammatory damage of skin, kidney, and nerve. Hepatitis B virus infection (Chapters 148 and 149) may be associated with immune complex deposition early in its course, during a period of antigen excess, because antibody production in response to hepatitis B surface antigen is as yet relatively insufficient; some anicteric patients may present with acute arthritis. Mixed essential cryoglobulinemia, which may be associated with hepatitis C viral infection, is an immune complex–mediated vasculitis in which deposition of complexes containing IgG, IgM, and hepatitis C antigens causes inflammation in peripheral nerves, kidneys, and skin. Serum sickness also can develop in transplant recipients who are treated with mouse monoclonal antibodies specific for human T cells to prevent rejection, and in patients with myocardial infarction who are treated with the bacterial enzyme streptokinase to effect thrombolysis. Systemic lupus erythematosus (Chapter 266), the prototypical immune complex–mediated autoimmune disease, is characterized by circulating IgG directed against common cellular constituents, typically DNA and DNAbinding proteins. Small immune complexes are deposited in skin, joints, and glomeruli and initiate local tissue damage.

Type IV Hypersensitivity Reactions

Type IV hypersensitivity reactions (Fig. 47-4), also known as delayed-type hypersensitivity reactions, are mediated by antigen-specific effector T cells. They are distinguished from other hypersensitivity reactions by the lag time from exposure to the antigen until the response is evident (1 to 3 days). Antigen is taken up, processed, and presented by macrophages or dendritic cells. TH1 effector cells that recognize the specific antigen (these are scarce and take time to arrive) are stimulated to release chemokines, which recruit macrophages to the site, release cytokines that mediate tissue injury and growth factors that stimulate monocyte production. IFN-γ activates macrophages and enhances their release of inflammatory mediators, whereas TNF-α and TNF-β activate endothelial cells, enhance vascular permeability, and damage local tissue. The prototypical type IV hypersensitivity reaction

is the tuberculin test, but similar reactions can occur after contact with sensitizing antigens (e.g., poison ivy, certain metals) and lead to epidermal reactions characterized by erythema, cellular infiltration, and vesicles. CD8+ T cells also may mediate damage by direct toxicity. In contrast to TH1-mediated hypersensitivity reactions, in which the effectors are macrophages, eosinophils predominate in TH2-mediated responses. TH2 effector T cells are associated with tissue damage in chronic asthma (Chapter 87). TH2 cells produce cytokines to recruit and activate eosinophils (IL-5 and eotaxin), leading to degranulation, further tissue injury, and chronic, irreversible airway damage. Additional TH effector cells, such as TH17 cells, mediate tissue damage. TH17 cells produce IL-17 family cytokines, as well as IL-21, IL-22, and GM-CSF, that regulate innate effectors and orchestrate local inflammation by inducing release of proinflammatory cytokines and chemokines, proliferation and activation of effector cells and other target cells, recruitment of neutrophils, and enhanced TH2-mediated inflammation, all of which amplify allergic and autoimmune responses.3,7 TH17 cells have been implicated in allergic disorders (atopic dermatitis, asthma) and autoimmune and inflammatory diseases (psoriasis, inflammatory bowel disease, rheumatoid arthritis, systemic lupus erythematosus, multiple sclerosis). In some autoimmune diseases, effector T cells specifically recognize selfantigens to cause tissue damage, either by direct cytotoxicity or by inflammatory responses mediated by activated macrophages. In type 1 insulin-dependent diabetes mellitus, T cells mediate destruction of β cells of the pancreatic islets. IFN-γ-producing T cells specific for myelin basic proteins have been implicated in multiple sclerosis. Rheumatoid arthritis is another autoimmune disease caused, at least in part, by activated TH1 cells. GENERAL REFERENCES For the General References and other additional features, please visit Expert Consult at

CHAPTER 47  Mechanisms of Immune-Mediated Tissue Injury  

GENERAL REFERENCES 1. Hogarth PM, Anania JC, Wines BD. The FcγR of humans and non-human primates and their interaction with IgG: implications for induction of inflammation, resistance to infection and the use of therapeutic monoclonal antibodies. Curr Top Microbiol Immunol. 2014;382:321-352. 2. Salazar F, Ghaemmaghami AM. Allergen recognition by innate immune cells: critical role of dendritic and epithelial cells. Front Immunol. 2013;4:356. 3. Singh RP, Hasan S, Sharma S, et al. Th17 cells in inflammation and autoimmunity. Autoimmun Rev. 2014;13:1174-1181.


4. Shah A. The pathologic and clinical intersection of atopic and autoimmune disease. Curr Allergy Asthma Rep. 2012;12:520-529. 5. Voehringer D. Protective and pathological roles of mast cells and basophils. Nat Rev Immunol. 2013;13:362-375. 6. Karsten CM, Kohl J. The immunoglobulin, IgG Fc receptor and complement triangle in autoimmune diseases. Immunobiology. 2012;217:1067-1079. 7. Maddur MS, Miossec P, Kaveri SV, et al. Th17 cells: biology, pathogenesis of autoimmune and inflammatory diseases, and therapeutic strategies. Am J Pathol. 2012;181:8-18.


CHAPTER 47  Mechanisms of Immune-Mediated Tissue Injury  

REVIEW QUESTIONS 1. Type III hypersensitivity reactions are caused by tissue deposition of immune complexes which activate local effectors and lead to injury. Which of the following mediators are NOT characteristic of this pathway of inflammation? A. Fc receptors for IgE B. Complement C. Neutrophils D. Mononuclear phagocytes Answer: A  Type III responses are mediated by IgG directed against soluble antigens. Localized deposition of immune complexes activates mast cells, monocytes, neutrophils, and platelets bearing the Fc receptor for IgG (FcγR) and initiates the complement cascade, all effectors of tissue damage. Generation of complement components C3a and C5a recruits and stimulates inflammatory cells and amplifies effector functions. IgE does not participate in the inflammatory response. Of note, mast cells may be activated in type III responses through receptors for IgG or complement. 2. Type IV hypersensitivity reactions, as exemplified by the tuberculin test, does NOT require which of the following elements? A. IgG B. Macrophages C. TH1 effector cells D. IFN-γ Answer: A  The prototypic type IV hypersensitivity reaction is the tuberculin test, in which antigen is taken up, processed, and presented by macrophages. Type 1 helper T (TH1) effector cells that recognize the specific antigen are stimulated to release chemokines, which recruit macrophages to the site, and release cytokines including IFN-γ, which activate macrophages and enhance their release of inflammatory mediators. IgG does not participate in type IV hypersensitivity reactions. 3. In which of the following clinical situations are autoantibodies NOT critical triggers of tissue damage? A. Asthma B. Pemphigus vulgaris C. Graves disease D. Systemic lupus erythematosus Answer: A  IgE antibodies against innocuous environmental antigens initiate asthma, not autoantibodies. Autoantibodies against TSH receptors act as agonists in Graves disease. Autoantibodies against desmoglein-3 effect adhesion between epidermal keratinocytes and cause blister formation. In systemic lupus erythematosus, inflammation is initiated by deposition of immune complexes containing autoantibodies directed against common cellular constituents (e.g., DNA and ribonucleolar proteins).

4. A syndrome similar to serum sickness occurs in which of the following infections? A. Hepatitis B B. Tuberculosis C. Pneumococcal pneumonia D. Influenza Answer: A  A syndrome similar to serum sickness occurs in chronic infections in which pathogens persist in the face of continued immune response. Hepatitis B virus infection may be associated with immune complex deposition early in its course, during a period of antigen excess, because antibody production in response to hepatitis B surface antigen is as yet relatively insufficient. Efficiency of clearance of immune complexes depends on their size and valence. Smaller immune complexes, which form in antigen excess—as occurs early in an immune response—circulate in the blood and are deposited in blood vessels, where they initiate inflammatory reactions and tissue damage through interactions with FcγRs and complement receptors. In influenza and pneumococcal pneumonia, there is no clinical evidence of systemic immune complex deposition. Tuberculosis generates T-cell responses. 5. The effector mechanisms that lead to tissue damage in autoimmune diseases are similar to those elicited in response to environmental antigens that result in allergy. Which hypersensitivity reaction is correctly matched to an autoimmune condition that occurs through a similar mechanism? A. Systemic lupus erythematosus and serum sickness B. Idiopathic thrombocytopenia purpura and asthma C. Multiple sclerosis and heparin-induced thrombocytopenia D. Myasthenia gravis and atopic dermatitis Answer: A  Systemic lupus erythematosus, the prototypical immune complex–mediated autoimmune disease, is characterized by circulating IgG directed against common cellular constituents, typically DNA and DNAbinding proteins. Like in the case of serum sickness, small immune complexes are deposited in skin, joints, and glomeruli and initiate local tissue damage. Idiopathic thrombocytopenia purpura, like type II hypersensitivity reactions, is mediated by autoantibodies that are directed to platelet surface antigens, whereas asthma is a triggered by IgE against innocuous environmental antigens (type I hypersensitivity reaction). T cells are key effectors in multiple sclerosis, whereas heparin-induced thrombocytopenia is caused by IgG autoantibodies. Myasthenia gravis is caused by IgG autoantibodies that recognize acetylcholine receptors and impair neuromuscular signaling, whereas atopic dermatitis is mediated by T cells.


CHAPTER 48  Mechanisms of Inflammation and Tissue Repair  

complex somatic mutations and gene rearrangements. This provides defense tailored for each member of the species; its complexity and beauty permit specificity but also provide opportunities for error such as responses against self-antigens in autoimmunity.

Pathogen-Associated Molecular Pattern Recognition

48  MECHANISMS OF INFLAMMATION AND TISSUE REPAIR GARY S. FIRESTEIN Host defense mechanisms have evolved to recognize pathogens rapidly, render them harmless, and repair the damaged tissue. This complex and highly regulated sequence of events can also be triggered by environmental stimuli such as noxious mechanical and chemical agents. Under normal circ*mstances, tightly controlled responses protect against further injury and clear damaged tissue. In disease states, however, pathologic inflammation can lead to marked destruction of the extracellular matrix (ECM) and organ dysfunction.


When normal tissue encounters a pathogen, resident cells are stimulated by engagement of pattern recognition receptors that activate an ancient arm of host defense known as innate immunity. In contrast to adaptive immunity, which provides exquisite antigen specificity, innate immune responses recognize common motifs on pathogens (Chapter 45). Additional cytoplasmic receptors can sense “danger” signals from a toxic environment or cellular stress, such as urate or adenosine triphosphate (ATP). Innate mechanisms are designed for rapid responses (minutes to hours) compared with the more leisurely adaptive system that can take days to weeks to develop. In addition to orchestrating early events that are critical to host defense, cells of the innate system like dendritic cells orchestrate the subsequent adaptive cascade through the generation of chemokines that organize lymphoid tissue and presentation of antigens to lymphocytes. Innate immunity provides intergenerational continuity in that the receptors are encoded in the germline and are passed unchanged to progeny to protect the species. In contrast, each individual must generate his or her own adaptive immune system through

The toll-like receptor (TLR) family of proteins binds common patterns of molecular structures on microbial pathogens that normally are not found in mammalian cells. The TLRs are critical members of the innate immune system and serve as sentinels that initiate a rapid response.1 Some are expressed on the cell surface, such as TLR2, which is activated primarily by bacterial peptidoglycan and lipoproteins, and TLR4, which is activated by lipopolysaccharide (LPS, or endotoxin). Others are expressed mainly on the inner leaflet of cytoplasmic vesicles, like TLR9, which is activated by unmethylated bacterial sequences that are enriched for CpG motifs (regions of DNA where cytosine and guanine nucleotides in the linear sequence of bases along its length are separated by one phosphate), or TLR3 and TLR7, which are important for antiviral defense because they bind double-stranded and single-stranded viral RNA, respectively. In addition to exogenous molecules, some endogenous structures can bind to TLRs, including heat shock proteins and oxidized low-density lipoproteins (oxLDLs). The latter might be especially important in the pathogenesis of atherosclerosis, in which LDL activates TLR4 within vascular plaques. Local endothelial cell– and macrophage-derived chemotactic factors can then recruit activated T cells into the atheroma. Signaling by TLR2 and TLR4 progresses through adaptor proteins and often converges on a kinase known as MyD88, which orchestrates several downstream cascades. By directing the phosphorylation of IκB kinase-β (IKKβ), MyD88 activates nuclear factor-κB (NF-κB), a master switch for inflammatory genes.2 Translocation of NF-κB to the cell nucleus stimulates the production of cytokines (e.g., interleukin-6 [IL-6], IL-8, and tumor necrosis factor [TNF]), the machinery for prostaglandin release (e.g., cyclooxygenase 2 [COX2]), and genes that regulate the ECM (e.g., metalloproteinases). This rapid response is normally transient, although it can persist in pathogenic states. MyD88-independent pathways that stimulate innate immunity also exist. For instance, TLR3 stimulation by RNA viruses uses a separate pathway involving IKKε and interferon regulating factor-3 (IRF-3). IRF-3, in combination with several other transcription factors, induces the expression of genes such as interferon-β (IFN-β) to establish an antiviral state. These genes primarily offer protection against pathogens by initiating key defense mechanisms. However, these same pathways can create a hazardous milieu that is toxic to normal cells through the production of oxygen radicals, nitric oxide, and other reactive intermediaries. These molecules can damage DNA and harm bystander cells, or even lead to neoplasia (E-Table 48-1). For instance, long-standing inflammation in the colon, as in ulcerative colitis, is associated with adenocarcinoma. Increased COX2 expression as a result of NF-κB translocation is another mechanism that contributes to the development of tumors at inflammatory sites. An unanticipated finding is that NF-κB itself can also directly augment carcinogenesis by serving as a survival signal for damaged cells that would normally be deleted by apoptosis. The TLR signal transduction mechanisms integrate the environmental stimuli and generate a broadly antipathogen response. Fine-tuning of host defenses against unique pathogen structures to provide long-lived immunity requires the slower, more precise adaptive immune system. Although it is more cumbersome and primitive, innate immunity provides signals that activate adaptive responses. For instance, TLRs can direct dendritic cells (Chapter 45), which have internalized and processed antigen, to migrate from peripheral tissues to central lymphoid organs. The dendritic cells can also produce cytokines and, after maturation, present antigens to T cells in the context of class II major histocompatibility molecules and surface costimulatory proteins. The activated T cells can then migrate to the tissue to enhance and amplify the host response. T cells also provide help to B cells, thereby stimulating antibody production and activating other components of innate immunity (e.g., the complement system, Chapter 50). Other non-TLR cytoplasmic sensors also serve a similar purpose in the environment. For instance, retinoic acid−inducible gene 1 (RIG-1) and melanoma differentiation−associated gene 5 (MDA5) can detect RNA viruses and initiate an inflammatory response. These are, in some cases, partially redundant with TLR3 and TLR7 and can activate similar signaling mechanisms, such as NF-κB through the IKKß and IRFs through a distinct pathway involving IKKε and TBK1.

CHAPTER 48  Mechanisms of Inflammation and Tissue Repair  




Toll-like receptor activation (e.g., oxLDL) Chemokine-mediated leukocyte recruitment (e.g., MCP-1)


Reactive oxygen and nitrogen intermediate-induced mutations Cyclooxygenase 2–mediated neoplasia (e.g., colon, breast) NF-κB prolonging survival of damaged cells


IgE-mediated mast cell activation TH2 cytokine-mediated leukocyte activation Leukotriene-induced bronchospasm Protease-induced airway remodeling

Rheumatoid arthritis

Toll-like receptor activation (e.g., peptidoglycan) Macrophage/fibroblast cytokine production, including IL-1, TNF, and IL-6 Cyclooxygenase 2 induction Protease-mediated cartilage destruction Synovial complement activation

Systemic lupus erythematosus

Complement activation in multiple organs α-Interferon production and interferon signature

Autoinflammatory diseases, including psoriasis

Inflammasome activation, including production of IL-1, IL-18, and IL-33 TH17 cell activation IL-17A-, IL-12-, and IL-23-mediated inflammation

IgE = immunoglobulin E; MCP-1 = monocyte chemoattractant protein 1; NF-κB = nuclear factor-κB; oxLDL = oxidized low-density lipoprotein; TH17 = helper T lymphocyte type 17; TH2 = helper T lymphocyte type 2; TNF = tumor necrosis factor.


CHAPTER 48  Mechanisms of Inflammation and Tissue Repair  

Environmental Stress and Danger-Associated Molecular Patterns Danger-associated molecular pattern molecules serve as a mechanism to detect and respond to damage to the microenvironment. Tissue injury due to direct trauma or a noxious stimulus initiates an inflammatory response and is associated with microvascular damage, extravasation of leukocytes through vascular walls, and leakage of plasma and proteins into the tissue. Endogenous proteins, including ATP receptors, S100, heat shock proteins, and high mobility-group box 1 (HMGB1), mediate release of molecules that reflect cellular toxicity and induce a cellular response. Acid-sensitive ion channels (ASICs) on the cell surface can also detect the environmental stress caused by a decrease in tissue pH. ASICs can mediate a variety of cellular functions, including cell death through apoptosis or pain responses that can lead to adaptive pain behaviors that limit further exposure to noxious stimuli.

Proteases, Coagulation, and Inflammation Although the coagulation system’s primary function is to maintain vascular integrity (Chapter 171), the proteases that regulate its functions also play an important role in the early responses to tissue damage and inflammation. For example, plasminogen is a circulating proenzyme that can be cleaved to plasmin by enzymes in the coagulation pathway, including factors XIa and XIIa. Tissue plasminogen activating factor and kallikrein also have this capacity. When activated, the serine protease plasmin can digest fibrin, fibronectin, thrombospondin, and laminin as well as activating pro-matrix metalloproteinases like collagenase (MMP1). By remodeling the extracellular matrix, this system can ultimately regulate cell recruitment and tissue damage. Thrombus formation at the site of vascular damage can begin the inflammatory cascade through the release of vasoactive amines (e.g., serotonin), release of lysosomal proteases, and formation of eicosanoid products. The platelets can also later regulate healing with release of growth factors such as platelet-derived growth factor (PDGF) and transforming growth factor-β (TGFβ).

Inflammasome The inflammasome3 is among the best characterized mechanisms for sensing danger and includes the 22-member human Nod-like receptor (NLR) family of cytoplasmic proteins. The activated NLR proteins recruit additional proteins to form a complex with caspase-1 and adaptor molecule apoptosisassociated specklike protein (ASC). Activation of caspase-1 is a key function of inflammasomes, with resultant cleavage and activation of IL-1, IL-18, and IL-33. The latter molecule is also known as an “alarmin” because of its rapid release in the presence of tissue damage or a pathogen. Alarmins are often preformed in cells, such as mast cells, and can be either released directly into the microenvironment or quickly processed and secreted. Other alarmins include products of cell destruction, such as ATP or uric acid. Disorders of the inflammasome are associated with a group of conditions known as autoinflammatory diseases (Chapter 261). The prototypic syndromes known as familial cold autoinflammatory disease, Muckle-Wells disease, and neonatal-onset multisystem inflammatory disease (NOMID) are due to nonconserved mutations in the NLR gene that encodes cryopyrin (also known as NALP3). These rare diseases are characterized by abnormal inflammasome activation with aberrant release of processed IL-1β. The clinical manifestations, including fever, rash, hearing impairment, and arthritis, depend on the specific amino acid substitution as well as other less welldefined genetic influences. The critical role of IL-1 has been proved by studies using treatment with IL-1 inhibitors, which prevent flares and can reverse end-organ damage. The inflammasome also participates in some common diseases, such as gout (Chapter 273), in which urate crystals can activate the inflammasome.

Immune Complexes and Complement The complement system (Chapter 50) is another ancient defense mechanism that links innate immunity and the humoral arm of adaptive immunity. Both the classical complement pathway, activated by immunoglobulin G (IgG)and IgM-containing immune complexes, and the alternative pathway, activated by bacterial products, converge at the third component of complement, C3, with proteolytic release of fragments that amplify the inflammatory response and mediate tissue injury. C3a and C5a directly increase vascular permeability and contraction of smooth muscle. C5a induces mast cell release of histamine, thereby indirectly mediating increased vascular permeability. C5a also activates leukocytes and enhances their chemotaxis,


adhesion, and degranulation, with release of proteases and toxic metabolites. C5b attaches to the surface of cells and microorganisms and is the first component in the assembly of the C5b-9 membrane attack complex. Individuals with abnormalities of the early complement components, especially C1q, C2, and C4, usually have a minimally increased incidence of infection but demonstrate an enhanced risk for developing autoimmune diseases such as systemic lupus erythematosus (SLE) (Chapter 266). The mechanism of increased disease susceptibility is probably related to inefficient clearance of immune complexes. Enhanced activation and consumption of complement proteins can also occur in SLE accompanied by low plasma C3 and C4 levels, especially in association with disease exacerbations. C3 or C5 deficiency increases susceptibility to bacterial infections, whereas defects in the late components that form the membrane attack complex result in an increased incidence of Neisseria sp bacteremia (Chapter 298).


Activation of innate immunity quickly leads to the robust influx of inflammatory cells. Resident cells, such as vascular endothelial cells, mast cells, dendritic cells, and interstitial fibroblasts, respond by releasing soluble mediators, including eicosanoids and pro-inflammatory cytokines (E-Table 48-2). These mediators amplify the inflammatory response and recruit additional leukocytes. Locally stimulated cells, along with the newly arrived inflammatory cells, release toxic reactive intermediates of nitrogen and oxygen as well as a myriad of proteases, principally matrix metalloproteinases (MMPs), serine proteases, and cysteine proteases. These molecules help destroy infectious agents and remove damaged cells, thus clearing the injured site for tissue repair. In most situations, the normal physiologic response is an exquisitely coordinated program that uses proteolytic enzymes to remodel the ECM and promote a supportive environment for wound healing rather than tissue damage.

Cellular Response Inflammatory cell infiltration at the site of initial tissue damage typically begins with release of chemokines and soluble mediators from resident cells, including interstitial fibroblasts, mast cells, and vascular endothelial cells. Signaling from these events alters the local adhesion molecule profile and creates a chemotactic gradient that recruits cells from the blood stream. Mast cells, in particular, act as sentinels that degranulate within seconds after ligation of immunoreceptors and activation of the signaling molecule spleen tyrosine kinase (Syk) to release vasoactive amines. In most acute responses, polymorphonuclear leukocytes (PMNs) are the first inflammatory cells to arrive at the site of injury, followed later by mononuclear cells. Most tissue fibroblasts and vascular endothelial cells are generally quiescent before migration of PMNs into the tissue. However, these resident cells can be triggered to proliferate and migrate toward the site of injury as well as to synthesize cytokines, proteases, and ECM components. Growth factors are released, such as basic fibroblast growth factor (bFGF) and vascular endothelial growth factor (VEGF), stimulating new blood vessel formation. Together with granulocyte-macrophage colony-stimulating factor (GMCSF), these locally released growth factors contribute to cellular proliferation and amplification of the inflammatory response and also induce maturation of dendritic cells that process antigens. In addition, fibroblasts and endothelial cells secrete new ECM proteins, MMPs, and other ECM-digesting enzymes. Initially, the response favors proteolytic activity to clear damaged infrastructure. This is followed by a shift to increased production of new ECM to allow tissue repair and wound healing. Increased vascular permeability, caused by disruption of endothelial cell tight junctions, allows blood-borne proteins such as fibrinogen, fibronectin, and vitronectin to extravasate into the perivascular ECM. Interaction with preexisting ECM allows the assembly of new ligands for a subset of adhesion molecules (e.g., integrins α5β1 and αvβ3). This increased vascular permeability and change in the profiles of adhesion molecules and ligands, in conjunction with release of chemoattractant molecules, leads to the recruitment of leukocytes to sites of inflammation. Some of the chemokines involved are IL-8 (for neutrophils), macrophage chemoattractant protein-1 (MCP-1) for monocytes, RANTES (regulated on activation, T-cell expressed and secreted) for monocytes and eosinophils, and IL-16 (for CD4+ T cells). Chemokines have the capacity to recruit specific subsets of cells by binding to G protein−coupled chemokine receptors. Directly targeting chemokines, either with biologics or with small molecules, has met with limited success in clinical trials, perhaps because the system is quite complex and highly

CHAPTER 48  Mechanisms of Inflammation and Tissue Repair  





IL-1 family (IL-1, IL-18, IL-33)


IL-6 family (IL-6, IL-11, LIF, osteopontin)


IL-4, IL-13




IL-17 family (IL-17A-F)


IL-12 family (IL-12, IL-23, IL-27)

Soluble TNF-R


IL-18 binding protein

Chemokines HMBG1 PROTEASES Matrix metalloproteinases   Collagenases   Gelatinases   Stromelysins   Matrilysins


Serine proteases   Trypsin   Chymotrypsin   Kallikrein   Plasmin

SERPINs, α2-macroglobulin

Cysteine proteases ADAMTS family   Aggrecanases SMALL MOLECULE MEDIATORS Prostaglandins (especially PGE2) Leukotrienes (especially LTB4) C3a and C5a

Lipoxins Cyclopentenone Antioxidants

Histamine Bradykinin Reactive oxygen Reactive nitrogen APOPTOSIS REGULATORS Soluble Fas ligand

Fas TRAIL Reactive oxygen Reactive nitrogen

ADAMTS = a disintegrin and metalloproteinase family; FGF = fibroblast growth factor; IL = interleukin; LIF = leukemia inhibitory factor; R = receptor; Ra = receptor antagonist; SERPINs = serine protease inhibitors; TGF = transforming growth factor; TIMPs = tissue inhibitors of metalloproteinase; TNF = tumor necrosis factor; TRAIL, TNF-related apoptosis-inducing ligand; VEGF = vascular endothelial growth factor; HMBG1= high mobility-group box 1.



CHAPTER 48  Mechanisms of Inflammation and Tissue Repair  

redundant. An alternative approach might be to target intracellular mechanisms distal to receptor ligation. Chemokine receptors generally signal through the phosphoinositide-3 kinase (PI3K) system, especially the gamma isoform. PI3Kγ is mainly expressed in bone marrow−derived cells and is the convergence point for multiple chemotactic factors. Preclinical studies suggest that blocking this pathway decreases inflammatory cell recruitment in models of lupus and rheumatoid arthritis. The precise combination of chemokines and vascular adhesion molecules present in an inflammatory lesion determines the timing for recruitment of individual inflammatory cell types. Ligation of integrins on leukocytes also prolongs cell survival after they have moved into the tissue, by preventing apoptosis. The central role of certain specific adhesion molecule−ligand pairs has been confirmed in human diseases. For instance, α4β1 plays a key role in the recruitment of lymphocytes to the central nervous system in multiple sclerosis, and blocking this interaction suppresses disease activity (Chapter 411). Eosinophils use the same adhesion receptors to migrate into the lung in allergen-induced asthma (Chapter 87). Increased expression of intracellular adhesion molecule 1 (ICAM-1) and vascular cell adhesion molecule 1 (VCAM-1), as well as increased chemokine expression, is evident in other cell types, such as the airway epithelium after allergen challenge in asthma. Rapid transient influx of neutrophils occurs in allergic airway disease, along with activation of the local T cells and mast cells. These neutrophils produce lipid mediators, reactive oxygen intermediates, and proteases such as elastase, which may contribute to airflow obstruction, epithelial damage, and remodeling. Neutrophil elastase, together with chemokines released by both recruited and allergen-activated T cells and mast cells, serves to recruit eosinophils.

Soluble Mediators


Pro-inflammatory cytokines, often derived from macrophages and fibroblasts, are mediators that activate the immune system. The pro-inflammatory members of the IL-1 family (e.g., IL-α, IL-1β, IL-18, and IL-33) and TNF have pleiotropic activities and can enhance adhesion molecule expression on endothelial cells, induce proliferation of endogenous cells, and stimulate antigen presentation. IL-1 and TNF also increase expression of matrixdegrading enzymes, such as collagenase and stromelysin. In addition, they stimulate synthesis of other inflammatory mediators such as prostaglandins from fibroblasts. TNF inhibitors (Chapter 36) are effective in inflammatory diseases such as psoriasis, rheumatoid arthritis, and inflammatory bowel disease, and IL-1 inhibitors (Chapter 36) are beneficial in genetic diseases such as Muckle-Wells syndrome and familial cold autoinflammatory syndrome. IL-1 and TNF comprise only a small fraction of the acute cytokine response. Many other factors also participate, including IL-6 and its related cytokines (IL-11, osteopontin, and leukemia inhibitory factor), which can both induce acute phase reactants and bias an immune response toward a helper T type 1 (TH1) or TH2 phenotype (Chapter 47). GM-CSF can regulate dendritic cell maturation, increase expression of human leukocyte antigen (HLA-DR) on these cells, and enhance antigen presentation. The TH1 lymphokine IFN-γ, although often considered part of the secondary wave that ensues after T-cell activation, can also induce expression of HLA-DR, increase expression of endothelial cell adhesion molecules, and inhibit collagen production. IL-1, IL-6, and IL-23 can coordinate differentiation toward TH17 cells, a phenotype that is thought to play a major role in inflammation and autoimmunity owing to the production of IL-17 family members (IL-17A through F). Of these, IL-17A and perhaps IL-17F are especially important because they can synergize with IL-1 and TNF. The growth factor TGF-β biases cells toward the regulatory T cell (Treg) phenotype, which can suppress antigen-specific responses of other T cells (see later). The benefit of individual cytokine inhibitors varies depending on the disease. For instance, IL-6 blockade is effective in rheumatoid arthritis, whereas IL-12/23 and IL-17A inhibition suppresses skin inflammation in psoriasis. Clinical trials now clearly show that IL-17A antibodies are effective in psoriasis. A1  Many cytokines activate cells by ligating their receptors and engaging the Janus kinase ( JAK) family of signaling molecules, including JAK1, JAK2, JAK3, and Tyk2. These kinases, in turn, phosphorylate the signal transducer and activator of transcription (STAT) proteins. The STATs serve as transcription factors that initiate expression of many other cytokines and mediators of the inflammation and amplify the response. JAK inhibition represents an alternative approach to abrogating the inflammatory response.

Cytokines play a key role in the establishment and perpetuation immunemediated diseases. As noted earlier, autocrine and paracrine cytokine networks play a critical role in the perpetuation of inflammation in rheumatoid arthritis4 (Chapter 264). MCP-1 recruits and activates macrophages into atheromas containing oxLDLs and foam cells. In allergic asthma (Chapter 87), IL-13 is emerging as a central inflammatory cytokine. IL-13 functions through binding to cell surface IL-4 receptors, and IL-4R–deficient mice are relatively resistant to the development of asthma.


In addition to cytokines and immune complexes, local inflammatory responses lead to the release of eicosanoids, which are lipid-derived molecules. Because lipids are present in the cell membrane, they are readily available substrates for the synthesis of mediators. These molecules are produced adjacent to sites of injury, and their half-lives range from seconds to minutes. Eicosanoids are not stored but are produced de novo from membrane lipids when cell activation by mechanical trauma, cytokines, growth factors, or other stimuli leads to release of arachidonic acid. Cytosolic phospholipase A2 (cPLA2) is the key enzyme in eicosanoid production. Cell-specific and agonist-dependent events coordinate the translocation of cPLA2 to the nuclear envelope, endoplasmic reticulum, and Golgi apparatus, where interaction with COX (in the case of prostaglandin synthesis) or 5-lipoxygenase (in the case of leukotriene synthesis) can occur.


Prostanoids5 are produced when arachidonic acid is released from the plasma membrane of injured cells by phospholipases and metabolized by cyclooxygenases and specific isomerases (Chapter 37). These molecules act both at peripheral sensory neurons and at central sites within the spinal cord and brain to evoke pain and hyperalgesia. Their production is increased in most acute inflammatory conditions, including arthritis and inflammatory bowel disease. In response to exogenous and endogenous pyrogens, prostaglandin E2 (PGE2) derived from COX2 mediates a central febrile response. In addition, prostaglandins synergize with bradykinin and histamine to enhance vascular permeability and edema. The levels of prostaglandins are usually very low in normal tissues and increase rapidly with acute inflammation, well before leukocyte recruitment. COX2 induction with inflammatory stimuli most likely accounts for the high levels of prostanoids in chronic inflammation. COX2 also plays a key role in platelet−endothelial cell interactions by increasing the production of prostacyclin (PGI2) in endothelial cells (Chapter 37). Increased risk for myocardial infarction associated with the use of selective COX2 inhibitors may be related to unopposed production of thromboxane A2 by COX1 in platelets. Prostacyclin also protects against atherosclerosis in mice, and COX2 blockade abrogates this beneficial effect. Thus, COX inhibitors can potentially increase thrombotic events.


A distinct set of enzymes direct arachidonic acid metabolites toward the synthesis of leukotrienes (Chapter 87). Their relative importance depends on the specific target organ of an inflammatory response. For instance, leukotriene receptor antagonists are effective in asthma, whereas similar approaches have been less impressive in rheumatoid arthritis. Unlike prostaglandins, leukotrienes are primarily produced by inflammatory cells such as neutrophils, macrophages, and mast cells. 5-Lipoxygenase is the key enzyme in this cascade, transforming released arachidonic acid to the epoxide leukotriene A4 (LTA4) in concert with 5-lipoxygenase-activating protein (FLAP). LTA4 can be hydrolyzed by cytosolic LTA4 hydrolase to LTB4, a potent neutrophil chemoattractant and stimulator of leukocyte adhesion to endothelial cells. LTA4 can also conjugate with glutathione to form LTC4 by LTC4 synthase at the nuclear envelope. LTC4 can be metabolized extracellularly to LTD4 and LTE4. These three cysteinyl leukotrienes promote plasma leakage from postcapillary venules, upregulation of expression of cell surface adhesion molecules, and bronchoconstriction.


Histamine is a vasoactive amine produced by basophils and mast cells that markedly increases capillary leakage. In basophils, histamine is released in response to bacterial formylmethionyl-leucyl-phenylalanine (f-MLP) sequences, complement fragments C3a and C5a, and IgE. The resultant edema can be readily observed clinically in urticaria (Chapters 252 and 440) and allergic rhinitis (Chapter 251). The stimulus for release of histamine from


CHAPTER 48  Mechanisms of Inflammation and Tissue Repair  

mast cell granules is the same as in basophils, except for the absence of f-MLP receptors in this cell type. Histamine can also synergize with locally produced LTB4 and LTC4. In addition, histamine enhances leukocyte rolling and firm adhesion, and induces gaps in the endothelial cell lining, enhancing leukocyte extravasation. Despite the production of histamine in asthma and in acute synovitis, currently available histamine blockers have minimal therapeutic effect in these conditions. Targeting the more recently described histamine type 4 receptor (HR4), which has a variety of immunomodulatory effects on bone marrow−derived cells, suggests that more precise inhibition of this novel histamine pathway might have greater success.6


Kinins induce vasodilation, edema, and smooth muscle contraction, as well as pain and hyperalgesia, through stimulation of C fibers. They are formed from high- and low-molecular-weight kininogens by the action of serine protease kallikreins in plasma and peripheral tissues. The primary products of kininogen digestion are bradykinin and lysyl-bradykinin. These products have high affinity for the B2 receptor, which is widely expressed and is responsible for the most common effects of kinins. The peptides desArg-BK and Lys-desArg-BK are generated by carboxypeptidases and bind the kinin B1 receptor subtype, which is not expressed in normal tissues but is rapidly upregulated by TLR ligands and cytokines. The kinin B2 receptor is internalized rapidly and desensitized, whereas the B1 receptor remains highly responsive. Kinin actions are associated with the secondary production of other mediators of inflammation, including nitric oxide, mast cell–derived products, and the pro-inflammatory cytokines IL-6 and IL-8. In addition, kinins can increase IL-1 production through initial stimulation of TNF and can increase prostanoid production through activation of phospholipase A2 and release of arachidonic acid.


Neural outflow also can rapidly activate inflammatory mechanisms and alter vascular permeability at sites of tissue damage. Pain receptors can activate type δ fibers and carry information to the spinal cord about noxious stimuli where cytokines like IL-1 or TNF are produced. Spinal cytokines lead to phosphorylation of signal molecules in the central nervous system like mitogen activated protein kinases (MAPKs). Reflex neural loops, including sympathetic and parasympathetic nerves, release mediators like substance P, acetylcholine, epinephrine or norepinephrine into the immediate location as well as surrounding tissue. Vascular permeability and activation of resident cells like macrophages can help recruit additional cells to the affected region.


Nitric oxide synthases (NOS) convert l-arginine and molecular oxygen to l-citrulline and nitric oxide (NO). There are three known isoforms of NOS: neuronal NOS (ncNOS or NOS1) and endothelial cell NOS (ecNOS or NOS3) are both constitutively expressed, whereas macrophage NOS (macNOS, iNOS, or NOS2) is induced by inflammatory cytokines such as TNF and IFN-γ, as well as by products of viruses, bacteria, protozoa, and fungi and by low oxygen tension and low environmental pH. Together with prostaglandins, the production of NO by NOS2 and ROIs by NADPH oxidase is a key mechanism by which macrophages paradoxically impair T-cell proliferation. This might control inflammatory processes or delete autoreactive T cells and partially accounts for the immunosuppression observed in certain infections and malignancies.

Proteases and Matrix Damage Production of enzymes that degrade the ECM regulates tissue turnover in inflammation. Reconfiguring of the matrix remodels damaged tissue, releases matrix-bound growth factors and cytokines, prepares the tissue for the ingrowth of new blood vessels, and alters the local milieu to permit adherence and retention of newly recruited cells. The MMPs are a family of more than 20 extracellular endopeptidases that participate in degradation and remodeling of the ECM matrix (Table 48-1). They are produced as pro-enzymes and require limited proteolysis or partial denaturation to expose the catalytic site. Their name is derived from their dependence on metal ions (zinc/metzincin superfamily) for activity and from their potent ability to degrade structural ECM proteins. MMPs can also cleave cell surface molecules and other pericellular nonmatrix proteins, thereby regulating cell behavior. For instance, MMPs can alter cell growth by digesting matrix proteins associated with growth factors. FGF and TGF-β have high affinities for matrix molecules that serve as depots for storage of these cytokines. Matrix proteolysis releases some growth factors and can make them available to cell surface receptors. In addition, MMPs can directly cleave and activate growth factors. MMPs affect cell migration by altering cell-matrix or cell-cell receptor sites. The adhesion molecule β4 integrin is



Collagen I, II, III, VII, and X Pro-MMP-1, -2, -8, -9, and -13 Aggrecan Pro-TNF


α1-Proteinase inhibitors Gelatin Tenascin


Aggrecan Denatured collagen Elastin Fibronectin Laminin Vitronectin

Pro-MMP-1, -2, and -13 Pro-TNF Pro-IL-1β Latent TGF-β


Proteoglycans Denatured collagens Entactin Fibrin, fibrinogen Fibronectin Gelatin Laminin Tenascin Vitronectin

Pro-MMP-2 and -7 Pro-TNF Membrane-bound Fas ligand (FasL) Plasminogen β4 Integrins


Proteoglycans Aggrecan Collagen III, IV, V, IX, X, and XI Pro-IL-1β Entactin Fibrin, fibrinogen Fibronectin Gelatin Laminin Tenascin Vitronectin

Pro-MMP-1, -3, -7, -8, -9, -10, and -13 Pro-TNF Plasminogen α1-Proteinase inhibitors

Reactive Oxygen and Nitrogen

Macrophages, neutrophils, and other phagocytic cells can generate large amounts of toxic reactive oxygen intermediates (ROIs) and reactive nitrogen intermediates (RNIs) that can directly kill pathogens. ROIs and RNIs also serve as critical signal transduction molecules that regulate expression of inflammatory genes. These molecules can also have deleterious effects on normal tissue by damaging DNA, oxidizing membrane lipids, and nitrosylating proteins. Release of reactive intermediates can be initiated by microbial products such as LPS and lipoproteins, by cytokines such as IFN-γ and IL-8, and by engagement of Fc receptors by IgG. These events cause translocation of several cytosolic proteins, including Rac2 and Rho-family guanosine triphosphatase (GTPase) to the membrane-bound complex carrying cytochrome c, with subsequent activation of reduced nicotinamide adenine dinucleotide phosphate (NADPH) oxidase. The reaction catalyzed by NADPH oxidase leads to superoxide production, which, in turn, increases hydrogen peroxide, hydroxyl radicals and anions, hypochlorous acid, and chloramines. In some cases, ROIs can contribute directly to the initiation of chronic disease. Lipid oxidation produces aldehydes that substitute lysine residues in apolipoprotein B-100. This altered moiety either binds to TLR2 to induce cytokine production or is internalized by macrophages, leading to the production of foam cells and fatty streaks, the primary lesions of atherosclerosis (Chapter 70). Subsequently, altered epitopes in damaged host proteins can be presented to T cells to initiate an adaptive immune response that amplifies the inflammatory vascular lesion.



IL = interleukin; MMP = matrix metalloproteinase; TGF = transforming growth factor; TNF = tumor necrosis factor.


CHAPTER 48  Mechanisms of Inflammation and Tissue Repair  

cleaved by MMP-7. MMP-3 and MMP-7 digest E-cadherin and not only disrupt endothelial cell junctions but also stimulate cell migration. Degradation of the ECM is usually initiated by collagenases, which cleave native collagen. Denatured collagen is then recognized and further degraded by gelatinases and stromelysins. Unlike the collagenases, stromelysins demonstrate broad substrate specificity and act on many ECM proteins, such as proteoglycan, fibronectin, laminin, and many cartilage proteins. Stromelysins can also amplify the remodeling process by activating collagenase through limited proteolysis. MMP gene expression can be induced by many pro-inflammatory cytokines, including TNF, IL-1, IL-17A, and IL-18. One common element in MMP promoters that regulates transcription is activator protein-1 (AP-1). AP-1 is a dimer that includes members of the Jun and Fos families. Cytokines can regulate the MMP gene by activating MAPKs, especially c-Jun amino terminal kinase ( JNK), which, in turn, phosphorylates c-Jun and markedly enhances MMP production. NF-κB and NF-κB-like binding sites also can contribute to protease transcription. Several other classes of proteases remodel the matrix, including serine proteases and cysteine proteases. High levels of active serine proteases, such as trypsin, chymotrypsin, and elastase, are released by infiltrating PMNs at sites of inflammation and can directly digest the ECM or activate the proenzyme forms of secreted MMPs. The ADAM (a disintegrin and metalloproteinase) family can cleave the extracellular domain of cytokine receptors. These ECM proteases include two members of the aggrecanase family. One of the aggrecanases (aggrecanase 2, or ADAMTS5) has been implicated in osteoarthritis because mice deficient in this enzyme have decreased cartilage destruction in models of osteoarthritis (Chapter 262).


Inflammation is a normal physiologic response but can cause serious host injury if allowed to persist. Additional mechanisms are required to reestablish homeostasis after this response is initiated. Suppression of acute inflammation by removal or deactivation of mediators and effector cells permits the host to repair damaged tissues through elaboration of appropriate growth factors and cytokines (Fig. 48-1). As in the initial generation of an inflammatory response, components of resolution include a cellular response (apoptosis and necrosis), formation of soluble mediators (such as anti-inflammatory cytokines and antioxidants), and production of direct effectors (such as protease inhibitors).

Deletion of Inflammatory Cells Cells can be removed from an inflammatory site by several mechanisms. First, the influx of cells can be decreased by suppressing chemotactic factor produc-

tion and vascular adhesion molecule expression. Second, cells, especially lymphocytes, can be released from the tissue and return to the circulation through lymphatics. Third, stressed cells can undergo necrosis with the release of their contents into the local environment. A fourth mechanism, known as autophagy,7 can lead to digestion of internal organelles and ultimately to cell death. Perhaps the most critical and effective method for clearing cells from an inflammatory site is programmed cell death, or apoptosis. Apoptosis is a highly regulated process in eukaryotic cells that leads to cell death and marks the surface membrane for rapid removal by phagocytes. This clearance process does not elicit an inflammatory response, in contrast to cell death by necrosis. PMN phagocytes have a very short half-life in the tissue, and the persistence or release of their contents into the microenvironment after death can be deleterious. In some pathologic conditions, such as leukocytoclastic vasculitis (Chapter 270), abundant neutrophil apoptosis is readily apparent on histopathologic examination. Other cells, including T lymphocytes, undergo postactivation apoptosis to prevent an overwhelming persistent host response. Defective apoptosis or even persistence of apoptotic cells that escape clearance may contribute to chronic inflammatory and autoimmune diseases. For instance, loss of tolerance to self-antigens might participate in autoimmune responses in SLE. Commitment of a cell to apoptosis can be initiated by a number of factors, including the ROIs in the cellular microenvironment as well as signaling through several death receptor pathways (e.g., FasL/Fas and TNF-related apoptosis-inducing ligand [TRAIL]). The former can damage DNA, which is a common byproduct of the genotoxic environment created by inflammation. If DNA damage is excessive, repair by tightly regulated mismatch repair mechanisms is terminated, and programmed cell death can be initiated by genes such as the p53 tumor suppressor. The burden of mutations induced by ROIs or RNIs in chronic inflammation can potentially accumulate over time and eventually lead to amino acid substitutions in key regulatory proteins. Ultimately, as has been observed in ulcerative colitis, neoplastic disease can ensue. Removal of apoptotic bodies, or the remnants of packaged apoptotic cells, is rapid and can be accomplished by macrophages, fibroblasts, epithelial and endothelial cells, muscle cells, and dendritic cells. The surface receptors used in recognition and engulfment of apoptotic cells include integrins (e.g., αvβ3), lectins, scavenger receptors, ATP-binding cassette transporter 1, LPS receptor, CD14, and complement receptors CR3 and CR4. However, some of these membrane molecules can be used in both pro-inflammatory and apoptotic pathways, the divergence of which may be based on differing ligands and accessory molecules. Apoptotic cells display a series of membraneassociated molecular patterns that interact with receptors on phagocytes. A general feature of apoptotic cells is loss of phospholipid asymmetry, with external presentation of phosphatidylserine. Externalized phosphatidylserine




Deletion of innate and adaptive immune response cells



Anti-inflammatory cytokines, IL-10, TGF-β

TIMPs Collagen α2-Macroglobulin SERPINs Metalloproteinases


Cytokine antagonists

Soluble receptors IL-1, TNF-α

Intracellular signals

Phosphatases Signaling inhibitors

Binding proteins IL-18

Natural antagonists IL-1Ra

Cell deactivation

Decreased adaptive immune response

FIGURE 48-1.  Anti-inflammatory mechanisms that resolve inflammation and lead to repair of the extracellular matrix. IL = interleukin; SERPINs = serine protease inhibitors; TGF = transforming growth factor; TIMPs = tissue inhibitor of metalloproteinases; TNF = tumor necrosis factor.

CHAPTER 48  Mechanisms of Inflammation and Tissue Repair  

may be sufficient to trigger phagocytosis, but other apoptotic cell surface structures exist. Although some inflammatory and immune cells are being deleted, other cell lineages expand during the resolution phase. Mesenchymal cells, especially fibroblasts, proliferate and produce new matrix that can contract to form a fibrotic scar. Locally produced growth factors such as PDGF induce DNA synthesis of these stromal cells through activation of PI3Ks. TGF-β8 also stimulates fibroblast proliferation and converts cell phenotype to matrix formation rather than matrix destruction by increasing collagen production and suppressing MMP expression. In addition, mesenchymal stem cells that either reside in the tissue or migrate from the peripheral blood can differentiate into the appropriate organ-specific lineage. The pluripotential cells, in the presence of the appropriate milieu, can become adipocytes, chondrocytes, bone cells, or other terminally differentiated stromal cells.

Soluble Mediators


A variety of anti-inflammatory cytokines are released by resident and infiltrating cells. TGF-β and IL-10 are examples that are produced by macrophages, interstitial fibroblasts, or T cells. Some T-cell cytokines, including IL-4, IL-10, and IL-13, suppress the expression of MMP by cells stimulated by IL-1 or TNF. In addition to increasing fibroblast proliferation, TGF-β suppresses collagenase production, increases collagen deposition, and decreases MMP activity by inducing production of the tissue inhibitors of metalloproteinases (TIMPs). The repair phase is abnormal in diseases in which tissue fibrosis represents a major pathologic manifestation. For example, scleroderma (Chapter 267) is marked by diffuse fibrosis and is accompanied by high levels of TGF-β and increased production of ECM. Cytokine decoy receptors can also downregulate the inflammatory response. Receptors can also be shed from the cell surface after proteolytic cleavage and can absorb cytokines, thereby preventing them from ligating functional receptors on cell membranes. These cytokine inhibitors can be released as a coordinated attempt to prevent unregulated inflammation, as in septic shock (Chapter 108), in which endotoxin induces production of soluble receptors after initial massive production of TNF and IL-1. Other types of cytokine-binding proteins are also produced as counter-regulatory mechanisms, including IL-18-binding protein (IL-18BP), which is an Ig superfamily−related receptor that captures IL-18. In bone remodeling (Chapter 243), interactions of receptor activator of NF-κB (RANK) with RANK ligand are required for osteoclast-mediated resorption. The competitive antagonist osteoprotegerin is a member of the TNF receptor family that binds to RANK ligand and inhibits osteoclast activation. At least two distinct mechanisms contribute to natural IL-1 inhibition. An IL-1 decoy receptor (type II IL-1R) has both cell membrane and soluble forms that neutralize IL-1 activity. In addition, a natural IL-1 antagonist, IL-1Ra, can bind to functional IL-1 receptors and compete with IL-1α or IL-1β. However, IL-1Ra does not transduce a signal to the cell and blocks the biologic functions of ambient IL-1. The balance of IL-1 and IL-1Ra production depends on many influences. For instance, monocytes produce more IL-1, whereas mature macrophages produce IL-1Ra.


The signaling pathways described previously that initiate an inflammatory response have intracellular mechanisms to ensure that the process is selflimited. Many kinases, such as the MAPKs, require post-translational modification through phosphorylation to increase enzyme activity. A system of phosphatases that remove these phosphates can return the kinase to its resting form. For example, dual specificity phosphatase 1 (DUSP1) is an enzyme that dephosphorylates p38 MAPK as well as other MAPKs. DUSP1 expression is increased by p38 MAPK; thus, the very process of activating the cell through p38 is responsible for its own counter-regulatory mechanism. NF-κB activation is typically initiated by phosphorylation of the inhibitor of kB (IkB), which targets it for proteolysis. IkB expression later increases dramatically and stops the signaling through this pathway. JAK-STAT signaling is inhibited by the suppressor of cytokine stimulation (SOCS) proteins. Thus, cellular defense mechanisms have evolved to prevent persistent cell activation.


none prostaglandins (CyPG). The prostanoids can serve as ligands for peroxisome proliferator-activated receptors (PPARs) (Chapter 206). There are three main classes of PPAR receptors—PPARα, PPARβ/δ, and PPARγ— all of which bind to DNA as heterodimers in association with the retinoid X receptor. Activation of PPARγ by CyPG is associated with the suppression of AP-1 and STAT transcriptional pathways in macrophages. A variety of natural and synthetic PPAR agonists have demonstrated efficacy in models of ischemia-reperfusion injury, arthritis, and inflammatory airway disease.

Inhibitors of Direct Effectors


Antioxidant enzymes that can inactivate the toxic intermediates and protect normal tissues include catalase and superoxide dismutase. Catalase is a peroxisomal enzyme that catalyzes the conversion of hydrogen peroxide to water and oxygen. Superoxide dismutases (SODs) catalyze the dismutation of superoxide to hydrogen peroxide, which is then removed by catalase or glutathione peroxidase. Glutathione peroxidases and glutathione reductase are additional mechanisms for maintaining redox balance and removal of toxic metabolites. Insufficient production of intracellular antioxidants such as glutathione can suppress lymphocyte responses and could account for defective T-cell receptor signaling and blunted immunity in T cells derived from rheumatoid arthritis synovium (Chapter 264). Interactions of free radicals with surrounding molecules can generate secondary radical species in a self-propagating chain reaction. Chain-breaking antioxidants are small molecules that can receive or donate an electron and thereby form a stable byproduct with a radical. These antioxidant molecules are categorized as either aqueous phase (vitamin C, albumin, reduced glutathione) or lipid phase (vitamin E, ubiquinol-10, carotenoids, and flavonoids). In addition, transition metal-binding proteins (ceruloplasmin, ferritin, transferrin, and lactoferrin) can serve as antioxidants by sequestering cationic iron and copper and thereby inhibiting the propagation of hydroxyl radicals.


Protease inhibitors regulate the function of endogenous proteases and reduce the likelihood of collateral damage to tissues. These proteins form two functional classes, active site inhibitors and α2-macroglobulin (α2M). The latter class of protease inhibitors acts by covalently linking the protease to the α2M chain and thereby blocking access to substrates. α2M binds to all classes of proteases and, after forming a covalent bond, conveys them to cells through receptor-mediated endocytosis with subsequent enzymatic inactivation. The family of inhibitors of serine proteases (SERPINs) are the most abundant members of the former class of protease inhibitors and play a major role in regulation of blood clot resolution and inflammation, as indicated by many of their names: antithrombin III, plasminogen activator inhibitors 1 and 2, α2-antiplasmin, α1-antitrypsin, and kallistatin. The TIMP family blocks the function of most MMPs. The TIMPs bind to activated MMPs and irreversibly block their catalytic sites. Examples of disease states with an unfavorable balance between TIMPs and MMPs include loss of cartilage in arthritis and regulation of tumor metastasis. TIMP-MMP imbalance in destructive forms of arthritis appears be caused by the limited production capacity for protease inhibitors, which is overwhelmed by the prodigious expression of MMPs. Whereas IL-1 and TNF induce MMPs, IL-6 and TGF-β suppress production of MMPs and increase levels of TIMPs. Therefore, the cytokine profile has a profound influence on the status of remodeling. When pro-inflammatory cytokines predominate, the balance favors matrix destruction; in the presence of pro-inflammatory cytokine inhibitors and growth factors, matrix protein production increases, and MMPs are inhibited by TIMPs.

Grade A Reference A1. Langley RG, Elewski BE, Lebwohl M, et al. Secukinumab in plaque psoriasis—results of two phase 3 trials. N Engl J Med. 2014;371:326-338.


COX2 induced by pro-inflammatory mediators appears early and can contribute to inflammatory responses. However, COX2 expression late in the process has led to speculation that it also functions in the resolution of inflammation. This regulation might occur through formation of the cyclopente-

GENERAL REFERENCES For the General References and other additional features, please visit Expert Consult at

CHAPTER 48  Mechanisms of Inflammation and Tissue Repair  

GENERAL REFERENCES 1. Bryant CE, Symmons M, Gay NJ. Toll-like receptor signalling through macromolecular protein complexes. Mol Immunol. 2015;63:162-165. 2. Pal S, Bhattacharjee A, Ali A, et al. Chronic inflammation and cancer: potential chemoprevention through nuclear factor kappa B and p53 mutual antagonism. J Inflamm (Lond). 2014;11:23. 3. Strowig T, Henao-Mejia J, Elinav E, Flavell R. Inflammasomes in health and disease. Nature. 2012;481:278-286. 4. Arend W, Firestein GS. Pre-rheumatoid arthritis: predisposition and transition to chronic synovitis. Nature Rev Rheumatol. 2012;8:573-586.


5. Aoki T, Narumiya S. Prostaglandins and chronic inflammation. Trends Pharmacol Sci. 2012;33:304-311. 6. Yamaura K, Shigemori A, Suwa E, et al. Expression of the histamine H4 receptor in dermal and articular tissues. Life Sci. 2013;92:108-113. 7. Choi AM, Ryter SW, Levine B. Autophagy in human health and disease. N Engl J Med. 2013;368:651-662. 8. Samarakoon R, Overstreet JM, Higgins PJ. TGF-β signaling in tissue fibrosis: redox controls, target genes and therapeutic opportunities. Cell Signal. 2013;25:264-268.


CHAPTER 49  Transplantation Immunology  



Clinical transplantation encompasses transplantation of organs and islets of Langerhans containing insulin-producing β cells, in which it is necessary to overcome the host-versus-graft (HVG) immune response to avoid rejection, as well as hematopoietic cell transplantation (HCT) (Chapter 178), in which it is necessary to contend with not only the HVG but also the graft-versushost (GVH) immune response. Because preparations of bone marrow or mobilized peripheral blood stem cells (mPBSCs) contain mature T cells, their administration to conditioned and consequently immunoincompetent recipients is associated with the risk for GVH disease. Organs transplanted include corneas, kidneys, livers, hearts, lungs, small intestines, pancreases, and composite tissue allografts such as hands and faces. The list of transplanted allogeneic cells is likely to expand in the future to include other cell types, such as hepatocytes, myoblasts, and stem cell–derived replacement cells. Transplants originating from a member of the same species are referred to as allotransplants. However, transplants from other species, termed xenografts, are believed by many to be a promising solution to the severely inadequate supply of allogeneic organs and tissues, and such grafts may be used in the future. Transplants of tissues or cells originating from the recipient, either by processing of cells from the recipient’s own organ (e.g., islets of Langerhans following pancreatectomy for chronic pancreatitis) or cell populations (e.g., CD34+ hematopoietic progenitor and stem cells collected from leukapheresis products following mobilization from the bone marrow before high-dose radiation or chemotherapy to treat cancer) are referred to as autologous. In the future, these transplants may include stem cell−derived autologous cells used for therapeutic purposes.


The major antigens recognized during graft rejection and the cell types targeting them are summarized in Table 49-1.

Major Histocompatibility Antigens

The major histocompatibility complex (MHC; human leukocyte antigens [HLAs] in the human) controls adaptive and some innate immune responses and is of central importance in many immune-mediated diseases. The MHC also presents the strongest immunologic obstacle to all types of allografts. The HLA molecules include two major isoforms, termed class I and class II, and are all encoded in the MHC complex of chromosome 6. Although all HLA molecules have a similar general structure, class I and class II molecules show different expression patterns, with class I MHC expressed on most cells of the body, whereas class II antigens are expressed mainly on antigen-

presenting cell (APC) populations, such as dendritic cells, macrophages, and B cells, as well as thymic epithelial cells involved in T-cell selection. Class II MHC can also be expressed on vascular endothelial cells and activated T cells of some species, including humans. The specialized function of both classes of MHC molecules is the presentation of peptide antigens to T-cell receptors (TCRs), allowing adaptive immune responses to occur. In general terms, the peptides presented by class I molecules are 8- to 9-amino acid peptides derived from cytosolic proteins (e.g., viral proteins) that are transported into the endoplasmic reticulum, where they are processed and loaded onto class I molecules during their synthesis. CD8 molecules interact with the α3 domain of the class I heavy chain, thereby strengthening the interaction of CD8+ T cells that recognize class I−peptide complexes. Peptides presented by class II MHC molecules, on the other hand, are mostly 10 to 20mers derived from exogenous proteins (e.g., phagocytosed bacteria) that are processed through the endosomal processing pathway, and these are recognized by TCRs of T cells whose CD4 molecules strengthen the overall T cell−APC interaction. The class I presentation pathway is of particular importance in allowing destruction of virally infected cells, consistent with the expression of class I MHC on almost all cell types in the body. However, there are exceptions to this paradigm that account for the phenomenon of cross-priming and cross-presentation, wherein peptides from exogenous proteins are presented by class I molecules, a phenomenon that may have significance for transplantation. Class II MHC presentation of exogenous antigens takes place primarily on professional APCs and B cells, consistent with the role of CD4+ T cells in initiating immune responses by activating APCs, providing direct and indirect (through activated APCs that also present peptides on class I molecules) “help” for CD8+ cytotoxic T lymphocytes (CTLs), and providing help for antibodyproducing B cells. B cells are able to focus the antigens recognized by their specific surface immunoglobulin receptors by binding and internalizing these antigens, which thereby predominate in the endosomal antigen-processing pathway and become presented by a high proportion of class II molecules on that B cell. This ability of B cells to preferentially present peptides derived from their cognate antigens to CD4 T cells recognizing those alloantigens is very important in driving alloantibody production. A number of MHC molecules have been crystallized, both alone and with TCRs that recognize them. The TCR binding structure of class I and II MHC molecules is similar overall and includes both the peptide binding cleft formed by a β-pleated sheet and two α-helices forming the sides of the cleft (E-Fig. 49-1). However, class I and II MHC molecules also have significant structural differences, as summarized in E-Table 49-1. Although class I molecules are formed by the combination of a highly variable heavy (45-kD) chain (α chain) noncovalently linked to a nonpolymorphic, smaller (12-kD) light chain (β2-microglobulin), class II molecules are heterodimers of two polymorphic chains, a 32-kD α chain and a noncovalently bound 28-kD β chain. TCRs interact physically with both the α-helices of the MHC molecules and side chains of the peptide that is bound in the groove, representing a trimolecular MHC-peptide-TCR interaction (see E-Fig. 49-1). It is the most variable (“hypervariable”) portion of the TCR, produced by V-D-J somatic rearrangements and N insertions in the TCR α and β chains, known





CD4+ T cells

Allogeneic class II MHC (+ peptide) Self class II MHC + donor peptide

Antigen-presenting cell activation Help (cytokines and costimulation) Proinflammatory cytokine production Cytotoxicity Regulatory function

Organ allografts Cellular allografts Xenografts GVHD

CD8+ T cells

Allogeneic class I MCH (+ peptide) Self class I MHC + donor peptide

Cytotoxicity Cytokine production Regulatory function

Organ allografts Cellular allografts Xenografts GVHD

NK cells

Class I MHC (activates or inhibits NK cell function) Other activating ligands

Cytotoxicity Cytokine production

? Organ allografts Cellular allografts Xenografts

B cells

Class I and class II MHC blood group antigens Xenogeneic carbohydrates

Antibody-mediated rejection (hyperacute, acute humoral, and chronic rejection)

Organ allografts Cellular allografts Xenografts

CTL = cytotoxic T lymphocyte; GVHD = graft-versus-host disease; MHC = major histocompatibility complex; NK = natural killer.

CHAPTER 49  Transplantation Immunology  

Peptidebinding cleft




N N Peptide-binding cleft


C β2m




E-FIGURE 49-1.  Two views of an HLA class I molecule. A, Ribbon diagram showing the x-ray crystallographic structure of an HLA class I molecule (side view). The β-strand structures are indicated by thick green arrows (oriented in an amino to carboxy direction), whereas connecting loops are indicated as thin lines. The α-helices are shown flanking a peptide-binding cleft at the top (membrane distal portion) of the molecule. The base (membrane proximal portion) of the molecule is formed by the noncovalent association between the α3 domain of the class I α chain and β2-microglobulin (β2m). B, View from the top of the molecule emphasizing that the base of the peptide-binding cleft consists of β-pleated sheets flanked by α-helical structures. C = C terminal; N = N terminal. (Adapted from Bjorkman PJ, Saper MA, Samraoi B, et al. Structure of the class I histocompatibility antigen HLA-A2. Nature. 1987;329:506-512.)




Chain structure of heterodimer

45-kD α chain 12-kD β2-microglobulin

34-kD α chain 28-kD β chain

Tissue distribution

All nucleated cells

Antigen-presenting cells (monocytes, B cells, dendritic cells, Langerhans cells), thymic epithelium, and some T cells; inducible on other cell types by interferon-γ

Size of bound peptides

8-9 amino acids

10-20 amino acids

Source peptides




Presentation of antigenic peptides to CD8+ T cells; ligands for natural killer cell receptors

Presentation of antigenic peptides to CD4+ T cells

CHAPTER 49  Transplantation Immunology  

as complementarity-determining region 3 (CDR3), that recognizes specific MHC-peptide complexes. The HLA molecules are all encoded within a 3.6-million base-pair region that encodes more than 200 genes, including complement and tumor necrosis factor (TNF) genes and many others in addition to MHC that have immunologic functions. The organization of the HLA region is illustrated in E-Figure 49-2, which shows that the heavy chains for “classic” class I HLA-A, B, and C and “nonclassic” class I molecules are encoded in a region that is telomeric to the “central MHC” region that includes complement and TNF genes among others, and lies between the class II and class I HLA regions. The class II region contains two α- and β-chain genes, only one of which is functional, for each of HLA-DQ and DP. However, the DR locus contains different numbers of β chains for different HLA alleles. Some of these DR β chains are pseudogenes, but various HLA-DR alleles contain either one or two functional β-chain genes. One of the striking features of HLA molecules (and the MHC of most mammalian species) is their extensive polymorphism. There are thousands of defined HLA alleles in the class I and class II regions. Because the primary function of antigen presentation to T cells is to permit responsiveness to and clearance of pathogenic microorganisms, this polymorphism may have evolved to maintain the diversity of immune responsiveness to various pathogens within a population, avoiding annihilation of that population by a single microorganism that might not be presented well by a particular MHC. HLA alleles were originally distinguished by panels of highly sensitized human sera containing multiple alloantibodies. Although it effectively identified structurally related HLA alleles, this method failed to distinguish many allelic differences that are of functional importance for antigen binding and T-cell recognition. It was only with the development of molecular methods to distinguish alleles at the genomic level, eventually through specific genomic sequences, that the full extent of the polymorphism in this region was revealed. In association with this knowledge, it has been necessary to continually revise and refine the system of nomenclature defining these alleles. According to the most recently accepted nomenclature,1 HLA alleles are identified by the locus (e.g., HLA-A), followed by an asterisk, and then a unique number with up to four sets of digits separated by colons. The first set describes the allele group (e.g., HLA-A*02), which usually corresponds to a serologically defined antigen, and the second set indicates the specific allele (e.g., HLA-A*02:101). The third and fourth set of digits are of less practical importance because they identify silent nucleotide substitutions in different alleles and variations in the nontranslated regions of the gene, respectively. Within certain populations, however, the level of diversity within allele groups may be quite limited because of the common genetic origin of the allele. For example, for the originally serologically defined HLA-DR3 allele group, there is little diversity among Northern Europeans, such that most carry the DRB1*0301 allele. Thus, for this population, it is reasonable to refer to the serologic HLA-DR3 type as defining this allele. There are certain alleles that predominate within racial groups. For example, as few as five DRB1 alleles predominate among Northern Europeans, with each allele represented in 10 to 30% of this population. E-Table 49-2 summarizes the major DRB1 allelic groups defined initially at the serologic level and later at the level of genomic sequencing. Most organ transplantations are performed across HLA disparities, and the strong immunosuppressive regimens used in transplant recipients are designed to prevent rejection by this exceptionally strong immune response. In contrast to T-cell responses to peptide antigens derived from foreign proteins, which are recognized by a very small fraction of naïve T cells (in the range of 1 in 105), a very high proportion, estimated at 1 to 10% of the T-cell repertoire, recognizes MHC alloantigens. The strong immunogenicity of allogeneic MHC molecules relates to the manner in which T cells are selected in the thymus; developing thymocytes do not survive unless they can weakly recognize a self MHC/peptide complex on a thymic stromal cell. This process is termed positive selection. Thymocytes whose receptors have high affinity for self/MHC complexes are deleted, however, so strongly autoreactive T cells rarely make it into the peripheral T-cell pool. Allogeneic antigens are not part of this negative selection process. The net result of these two selection steps is that the human T-cell “repertoire” is strongly biased to have cross-reactivity to allogeneic MHC molecules, providing a barrier to organ and hematopoietic cell transplantation. In the case of organ transplantation, in which long-term pharmacotherapy with powerful immunosuppressive drugs is used in an effort to prevent graft rejection, this can translate into improved results with matched organs in some situations. However, for unrelated, deceased donor transplantation, the benefits of HLA matching may be counterbalanced by


the disadvantages associated with prolonged graft ischemia when attempts are made to transport organs to the most closely matched recipient.2 For hematopoietic cell transplantation (Chapter 178), the risks for GVH disease and marrow graft failure are so greatly amplified in the presence of extensive HLA mismatches that such transplantations have been avoided whenever possible; if a sufficiently matched, related donor cannot be found, a search is conducted through large registries containing millions of volunteer unrelated donors. Because of its extensive polymorphism, truly MHCidentical, unrelated donors can be difficult to find in the human population at large. For individuals with common HLA genotypes, the likelihood of finding a matched unrelated donor is markedly greater than that for individuals with rare genotypes. This situation relates in part to the phenomenon of linkage disequilibrium, wherein alleles at nearby loci are found together on the same chromosomal segment, or haplotype, more frequently than would be predicted by chance. The pattern of linkage disequilibrium is different in different racial groups, so the chance of finding a truly genotypically identical haplotype is greatest within the same population. For example, among whites, the DRB1*0301 allele is in linkage disequilibrium with DQB1*0201, which is located several hundred thousand base pairs away on chromosome 6; this complex, in addition to the DR4 alleles that are in linkage disequilibrium with DQB1*0302, confers the greatest genetic component of risk for the development of type 1 diabetes. Many autoimmune diseases demonstrate similarly strong HLA associations. Although there are data to indicate that HLAspecific autoantigen presentation plays a major role in determining disease susceptibility, non-HLA genes in linkage disequilibrium likely account for a significant component of these genetic risk factors. The use of alternative donors has also increased the availability of hematopoietic cell transplantation (HCT) in individuals for whom an HLAidentical related or unrelated donor cannot be identified (Chapter 178). The use of cord blood transplantation, which has reduced GVH disease−inducing activity compared with adult stem cell products, as well as advances in avoiding GVH disease in haploidentical related donor HCT, has recently increased the safety and use of HLA-mismatched HCT.3

Minor Histocompatibility Antigens

“Minor” histocompatibility antigens are peptides derived from polymorphic peptides presented by an MHC molecule. Even genotypically HLA-identical siblings have different minor histocompatibility antigens. These are sufficient to induce graft rejection if immunosuppressive pharmacotherapy is not used. In the case of HCT, significant GVH disease frequently (about 30 to 50% of the time) complicates transplantation between HLA-identical siblings, even with the use of pharmacologic immunoprophylaxis.

Other Antigens

The major blood group (ABO) antigens can be the targets of a dramatic “hyperacute” rejection process that occurs when mismatched vascularized grafts are transplanted. Recognition of blood group antigens on the endothelial surface of the graft vessels by recipient “natural” antibodies (antibodies that are present without known sensitization to the antigens) activates the complement and coagulation cascades, resulting in rapid graft thrombosis and ischemia. A similar outcome can occur after transplantation to an individual with preformed anti–donor HLA antibodies resulting from presensitization by prior transplantations, transfusions, or pregnancies. Antibodies against other polymorphic antigens, such as MHC class I−related chain A (MICA), have been associated with graft rejection. In the past, transplantation could not be successfully performed in the presence of a positive antidonor crossmatch. However, considerable success has been achieved in the transplantation of ABO-mismatched kidneys, livers, and hearts (the latter in the neonatal period only), and in transplantation of kidneys to highly presensitized patients.4,5 In the case of kidney and liver transplantation, initial removal of the antibody and sometimes depletion of B cells, as well as the infusion of intravenous immunoglobulin (IVIG), has led to these successes. ABO-mismatched neonatal heart transplantation has succeeded because the transplantations are performed before the recipient has developed high levels of anti–blood group antigen antibodies, and the B cells seem to be rendered tolerant to the donor blood group antigen by the grafting process. Recognition of blood group antigens can also be of significance in HCT, in which ABO barriers are routinely crossed in both directions. This can cause hemolysis of recipient erythrocytes if the mismatch is in the GVH direction, but this complication can be avoided by washing the cellular product before infusion. Mismatches in the HVG direction can cause more persistent problems due to ongoing destruction of donor erythropoietic cells, resulting in

CHAPTER 49  Transplantation Immunology  

HLA class II region DP

Chromosome 6p21

"Central" MHC

HLA class I region

DQ DR Telomere MICB B C E AG MICA TNF-α / β







fB C2




DRB subregion gene organization DRB1 ψDRB DR1,10 DRB1 ψDRB


DR15,16 DRB1 ψDRB DRB3 DR3,5 DRB1 ψDRB ψDRB DRB4 DR4,7,9 DRB1 DR8 E-FIGURE 49-2.  Map of the human major histocompatibility complex (MHC) spanning approximately 3.5 million base pairs on the short arm of chromosome 6. The HLA class I and class II molecules are encoded in distinct regions of the MHC. The HLA class II region contains three subregions: DR, DQ, and DP. Each of these subregions contains a variable number of α- and β-chain genes. HLA class II loci with known functional protein products are labeled in bold. In the case of DR, different numbers of DRB genes are present in different haplotypes, some of which are nonfunctional pseudogenes (ψ). A summary of the most common of these is shown in the box. The DQ and DP subregions each contain one pair of functional α- and β-chain genes. The HLA class I region contains the three “classic” class I genes—HLA-A, HLA-B, and HLA-C—as well as other related “nonclassic” class I molecules such as MICA, MICB, HLA-E, and HLA-G. The gene for familial hemochromatosis (HFE) is found just telomeric to the HLA class I region, about 3 million base pairs distant from HLA-A. The “central” MHC also contains a number of genes related to immune function, including the complement components (C4A, C4B, C2, and factor B), as well as tumor necrosis factor (TNF)-α and -β. Not shown in the figure are more than 100 additional genes, many of which are located in the central MHC. A complete listing of MHC-encoded genes can be found in Horton R, Wilming L, Rand V, et al. Gene map of the extended human MHC. Nat Rev Genet. 2004;5:889-899.


Serologic “Splits”



DR15 DR16


DRB1*1501, 1502 DRB1*1601 DRB1*0301


DRB1*0401, 0402, 0403, 0404, 0405, 0406, 0407, 0408


DR11 DR12

DRB1*1101, 1102, 1103, 1104 DRB1*1201


DR13 DR14

DRB1*1301, 1302, 1303 DRB1*1401




DRB1*0801, 0802, 0803, 0804, 0806





*Alleles in bold are found in at least 10% of individuals in the population. From Williams F, Meenagh A, Single R, et al. High resolution HLA-DRB1 identification of a Caucasian population. Hum Immunol. 2004;65:66-77.


CHAPTER 49  Transplantation Immunology  

pure red cell aplasia. More often, however, donor erythropoiesis is successfully established, and antidonor isohemagglutinins disappear from the circulation. A and B blood group antigens are the consequence of the presence or absence of specific glycosylation enzymes in different individuals. Likewise, an antigenic specificity of the utmost importance in xenotransplantation is a carbohydrate epitope, Galα1–3Galβ1–4GlcNAc (αGal), which is produced by a specific galactosyl transferase. Humans and Old World monkeys lack a functional αGal transferase and produce high levels of natural antibodies against the ubiquitous αGal epitope. Because animals of interest as xenograft sources (e.g., pigs) express αGal at high levels on their vascular endothelium, transplantation of vascularized organs from pigs results in hyperacute rejection unless something is done to absorb the antibodies or inactivate complement. The development of αGal-knockout pigs, therefore, was an important milestone, and encouraging results have been obtained in pig-to-primate transplantation in initial studies. In another type of transplant reaction, recognition as foreign results not from the presence of an antigen, but paradoxically from the absence of a self MHC molecule. Natural killer (NK) cells express a series of surface inhibitory and activating receptors that, collectively, determine whether the NK cell does or does not kill a potential target cell. The ligands for the inhibitory receptors are MHC class I molecules, and the receptors recognize specific groups of alleles. An NK cell may kill an allogeneic target that lacks a self MHC inhibitory ligand. This phenomenon has been shown in animal models to result in rapid bone marrow rejection when the donor marrow cells are not given in excess numbers or when a fraction of them are destroyed by an incompletely suppressed T-cell response. A similar phenomenon has not been clearly demonstrated in clinical HCT. The possibility that NK cells play a role in organ allograft rejection has long been an area of controversy. NK cells may be of particular importance in xenotransplantation, where they appear early in infiltrates of organ xenografts undergoing acute vascular rejection. NK cells clearly play a strong role in rejection of xenogeneic bone marrow, an observation that is relevant in one approach to inducing tolerance (see later discussion).


Cellular Mediators

Many different cell types participate in rejection responses, and there is considerable redundancy. T cells are key players in most forms of rejection, with the exception of rejection that can be induced by antibodies in the absence of T-cell help. These include hyperacute and acute vascular rejection processes that may be induced by natural antibodies, as described earlier, or by antibodies that are present due to presensitization. The possible role of NK cells has already been discussed.

Direct and Indirect Allorecognition

T-cell responses are induced by APCs that present alloantigens. There are two forms of alloantigen recognition, termed direct and indirect (Fig. 49-1). Direct allorecognition denotes recognition of donor antigens on donor APCs provided by the graft. The extraordinarily high frequency of T cells with alloreactivity is caused by direct recognition of allogeneic MHC. Indirect recognition is the recognition of donor antigens that are picked up and presented on recipient MHC molecules on recipient APCs. The indirect response is more similar to “normal” T-cell responses, in which professional APCs present peptide antigens to T cells that are present at relatively low frequency in the naïve repertoire. In organ transplantation, direct alloreactivity is particularly important in the early post-transplantation period, when APCs within the transplanted organ are still present; many of these cells migrate to the lymphoid tissues, where they initiate the alloresponse. However, the APC supply that comes with the donor graft is not renewable, so if the direct response is not maintained by recognition of donor antigens on endothelial cells or other cells in the graft, it may recede in importance. The indirect response, on the other hand, can be maintained by the constantly renewed pool of recipient APCs. The indirect response is of particular importance in inducing antibody responses.

Effector Mechanisms of Rejection

T cells can promote graft rejection through several effector mechanisms. One is the antibody-dependent processes that have already been discussed, which can be induced by CD4+ helper T cells that promote differentiation and

Direct Allorecognition T-cell receptor

Donor Class II MHC

Recipient T cell Donor APC


Indirect Allorecognition T-cell receptor Recipient T cell

Host Class II MHC

Donor peptide

Recipient APC

CD4 FIGURE 49-1.  Direct and indirect allorecognition. Direct allorecognition involves the recognition by a T-cell receptor of major histocompatibility complex (MHC) molecules (with or without a peptide) on a donor antigen-presenting cell (APC). Indirect allorecognition involves recognition by the T-cell receptor of a donor peptide presented on a recipient APC that has picked up and processed donor antigens.

immunoglobulin (Ig) class switching of B cells that recognize other specificities on the same alloantigens. T cells provide cognate help to B cells when the TCRs recognize complexes of self MHC with donor MHC−derived peptide antigens (produced by B cells whose surface Ig receptors recognize and pick up the donor MHC antigen). If antidonor antibody is not present before transplantation but is induced afterward, the response can lead to the pathologic picture of acute humoral rejection. Antibodies may also participate in a slower, poorly understood process of chronic rejection, which, in the case of kidney and heart, is characterized by unique vascular lesions with intimal thickening and loss of the vessel space, and in the case of lung transplantation, by obliterative bronchiolitis. The mechanisms underlying these chronic rejection lesions are not well understood, and several different immune processes may in fact lead to similar lesions. Another major effector pathway leading to graft rejection involves CTLs, which are predominantly members of the CD8+ T-cell subset but also include CD4+ T cells. Several effector mechanisms lead to killing of target cells by CTLs, and these include the granzyme/perforin-mediated pathway and the pathways involving Fas/Fas ligand (FasL) and other members of the TNF receptor family and their ligands (Chapter 47). Because CD8+ cells recognize class I MHC molecules, which are widely expressed, it is not difficult to envision graft destruction by CD8+ CTLs. CD8+ CTLs may be activated through an APC that is stimulated initially through contact with an alloreactive CD4+ cell. This is one form of CD4 “help” for CD8+ cells. In addition, CD8+ cells may be dependent on cytokines such as interleukin-2 (IL-2) from CD4+ cells for their expansion and cytotoxic differentiation. However, there are also many examples of CD8+ cell–mediated rejection that is independent of “help” from CD4+ cells. Class II MHC, which is recognized by CD4+ T cells, is less widely expressed on graft tissues than is class I MHC, although it may be induced on endothelial cells and graft parenchymal cells in the presence of inflammatory cytokines such as interferon-γ (IFN-γ). In addition to cytotoxic mechanisms resulting from direct allorecognition, CD4+ and CD8+ T cells with indirect specificity seem also to be capable of causing graft destruction under some circ*mstances. Cytokines such as IFN-γ have been implicated in some instances, but in general, the pathways of indirect graft destruction are not well understood. A CD8+ cell–mediated form of skin graft rejection that is dependent on donor antigens crosspresented on recipient MHC molecules (a form of indirect allorecognition for CD8+ cells) has been described in an animal model. This form of graft rejection may be directed at antigen presented on endothelial cells of recipient vessels that revascularize the graft. This mechanism would not apply to primarily vascularized organ allografts.

The Role of T-Cell Trafficking

All the rejection processes described require trafficking of T cells into the graft. This process is made possible after the initial activation of naïve T cells

CHAPTER 49  Transplantation Immunology  

in the lymphoid tissues. Naïve T cells can migrate into lymph nodes because of their expression of the CCR7 chemokine receptor and the adhesion molecule L-selectin. These T cells are activated by migratory graft APCs that also enter the lymph nodes. T-cell activation is associated with loss of CCR7 and L-selectin expression and acquisition of a new set of chemokine receptors and adhesion molecules that allow rolling and adhesion on the graft endothelium and entry into the graft parenchyma (Chapter 47). Inflammation in the graft, such as that induced by ischemia-reperfusion injury and the transplantation procedure, as well as that induced by initially responding T cells, is associated with upregulation of chemokines and adhesion ligands that promote entry of lymphocytes into the graft. Nevertheless, well healed-in grafts can be slowly rejected by adoptively transferred memory T cells, demonstrating that acute graft injury and inflammation are not essential for rejection in the presence of an established memory T-cell response. Rejection of hematopoietic cell grafts may involve many of the same mechanisms as those discussed for solid organs, although less detailed work has been done in this area.

Mechanisms of Graft-versus-Host Disease

Initiation of GVH disease (Chapter 178) requires that donor T cells recognize host alloantigens. The disease involves attacks on a variety of recipient epithelial tissues, namely skin, the intestine, and liver. Animal models have demonstrated clear roles for both CD4+ and CD8+ cells in initiating GVH disease, and each subset is able to do so independently of the other. The mechanisms of GVH disease include activation of alloreactive donor T cells by recipient APCs, leading to the differentiation of effector cells with direct cytotoxic activity and cytokine production in response to host antigens. A prominent role is played by TNF-α, whose production is induced in part by the translocation of bacteria across the intestinal wall, promoting innate immune system activation through toll-like receptors (Chapter 45). An intensely pro-inflammatory environment is produced by the combination of conditioning-induced tissue injury and disruption of mucosal barriers, bacterial activation of the innate immune system, and the GVH alloresponse. An important role is now appreciated for the inflamed microenvironment in target tissues in promoting the trafficking of GVH-reactive T cells into these tissues.6


In view of the critical role of donor T cells in inducing GVH disease, an obvious strategy for preventing this complication is to remove mature T cells from the marrow graft. This approach has indeed been shown in both animal models and clinical studies to prevent GVH disease effectively. However, there are several disadvantages to this approach. One is that adult humans, particularly those who have undergone prior chemotherapy and radiotherapy, have little remaining thymic tissue and therefore demonstrate sluggish T-cell recovery, leading to serious opportunistic infections. The second disadvantage applies to the most common indication for allogeneic HCT, namely the treatment of hematologic malignancies (Chapter 178). In this setting, T-cell depletion is often associated with an increased relapse rate due to loss of a graft-versus-tumor (GVT) effect, which is in large part mediated by GVH alloreactivity. Separation of GVH disease from GVT effects is a major goal of research in HCT, and some promising strategies are being explored (E-Table 49-3). These include control of T-cell trafficking so that the GVH alloresponse is confined to the lymphohematopoietic tissues where the tumor resides and a number of other approaches.6,7 The third disadvantage of donor T-cell depletion in HCT is that it increases the rate of engraftment failure. GVH alloreactivity and a “veto” effect of donor T cells help to overcome host resistance to donor engraftment. A veto cell, which may be a T cell or an NK cell, kills a CTL that attacks it. Although the phenomenon has been well established in animal models, its mechanisms are not clearly established, and its potential role in humans is uncertain. NK-cell recognition in the GVH direction resulting from the absence in the recipient of a class I MHC ligand (E-Fig. 49-3) that can trigger a donor NK-cell inhibitory receptor (KIR) may promote donor marrow engraftment and antitumor effects against acute myeloid leukemias in the setting of T-celldepleted, HLA-mismatched HCT. Clinically, pharmacologic immunosuppressive prophylaxis is usually used in at least the first 6 months after HCT to minimize the complication of GVH disease. Additionally, HLA-matched or closely matched donors are chosen whenever possible because GVH disease increases in frequency and severity as increased HLA barriers are transgressed. These measures, nevertheless, are


insufficient, and GVH disease remains a major complication of HCT. Therefore, many of the new strategies being explored in organ transplantation and other fields are also being examined for the prevention of GVH disease in experimental models. It should be borne in mind, however, that tolerance of donor T cells to recipient alloantigens (see later discussion) might not be entirely beneficial in the HCT setting for the treatment of malignant disease because loss of GVH alloreactivity is likely to come with loss of antitumor effects.


Nonspecific Immunosuppression

Immunosuppressive drugs are the mainstay of clinical organ transplantation, and improvements in these drugs following the discovery of cyclosporine have extended organ transplantation to include hearts, lungs, pancreases, livers, and other organs and tissues in the past 30 years. The mechanisms of action of these agents are discussed in Chapter 35. However, it is noteworthy that, despite these improvements and their enormous impact on early graft survival, these agents have been less effective in attenuating late graft loss. Because chronic immunologic rejection processes and side effects of the immunosuppressive drugs themselves are responsible for much of this late graft loss, improved immunosuppressive agents and induction of immune tolerance (see later discussion) are major research goals in transplantation.

Costimulatory Blockade

As understanding of immune responses has increased, recent years have seen the exploration of numerous biologic agents, including antibodies and small molecules targeting receptors of the immune system as well as cell-based therapies, in efforts to improve allograft survival. Because of the central role played by T cells in the immune response, considerable attention has been focused on blockers of T-cell costimulation. When a naïve T cell recognizes antigen through its unique TCR, additional “costimulatory” signals are required to allow full activation, expansion, and differentiation to occur. These signals are often provided by APCs in the form of ligands (e.g., B7-1, B7-2) for costimulatory receptors (e.g., CD28) on the T cell. Cross-talk between the T cell and the APC (e.g., due to CD40 activation by CD154 upregulation on the activated T cell) further amplifies the costimulatory activity of the APC, allowing it to effectively activate other T cells as well. The CD154 (T cell)–CD40 (B cell) interaction also promotes Ig class switching and functioning of B cells as APCs. Blockade of these processes (e.g., by CTLA4Ig and anti-CD154 monoclonal antibodies [mAbs]) has led to marked prolongation of allograft survival in stringent rodent and largeanimal models. Robust, systemic tolerance to donor antigens has been achieved in rodents receiving bone marrow transplantation with costimulatory blockade and little or no additional conditioning. Some of these agents have joined the armamentarium of immunosuppressive agents in clinical trials in transplantation and autoimmune diseases.8 Although anti-CD154 antibodies have been associated with thromboembolic complications, precluding further evaluation in transplantation trials, recently developed anti-CD40 antibodies have shown promise in animal studies. Numerous additional costimulatory and inhibitory pathways that affect T-cell responses have been described, and these all are potential targets for further manipulation of the alloresponse.

Immune Tolerance

Immune tolerance denotes a state in which the immune system is specifically unreactive to the donor graft (or recipient in the case of GVH reactivity) while remaining normally responsive to other antigens.9-11 Tolerance is distinct from the state produced by nonspecific immunosuppressive agents, which increase risks for infection and malignancy. Numerous approaches to tolerance induction have been described in rodent models, largely owing to the strong tolerogenicity of primarily vascularized heart, liver, and kidney grafts in these animals. Because such grafts are less tolerogenic in humans, none of these strategies has been effectively applied clinically to date. Therefore, tolerance strategies that are appropriate for clinical evaluation must first be tested in “stringent” models, including relatively nontolerogenic grafts such as MHC-mismatched skin in rodents and vascularized organ graft models in large animals. In most of the models, only a superficial understanding of the mechanisms leading to tolerance is currently available. The three major mechanisms of T-cell tolerance are deletion, anergy, and suppression (often referred to as “regulation”). Deletion denotes the

CHAPTER 49  Transplantation Immunology  





Donor T-cell TH2 polarization (e.g., conditioning with ATG and TLI; in vitro stimulation with cytokine exposure)

May preserve GVL

May limit GVL; TH2 can contribute to acute and chronic GVHD

Tolerance induction of donor T cells (e.g., costimulatory blockade; regulatory cells)

Some strategies may selectively tolerize GVH-reactive T cells (e.g., in vitro antigen exposure with costimulatory blockade)

Global immunosuppression may limit GVL and antiinfectious immunity; tolerance (i.e., GVH protection) may be incomplete

Donor T-cell depletion plus NK-cell infusion with class I mismatched transplantation

NK cells do not cause GVHD but may mediate antitumor effects; donor NK cells may eliminate host APCs that trigger GVHD

Antitumor effect against only certain types of malignancies; requires appropriate MHC disparity and expression of polymorphic NK-cell receptors; insufficient T-cell immunity to infection

Donor T-cell depletion followed by delayed donor lymphocyte infusion (DLI)

Preserves high level of GVL due to GVH reactivity. GVHD does not occur if host inflammation from conditioning has subsided and initial HCT was devoid of donor T cells

Antitumor effect delayed until time of DLI; most applicable for indolent lymphohematopoietic tumors. GVHD more difficult to control in humans than animal models, probably owing to occult or overt infection resulting from T-cell deficiency before DLI

Depletion of donor T cells recognizing host alloantigens by in vitro or in vivo activation/ depletion (i.e., “allodepletion”)

Preserves anti-infectious immunity and tumor antigen−specific responses while limiting GVHD

Loss of GVH reactivity will limit GVL and engraftment; highly efficient allodepletion methods not yet available. Residual T cells may cause GVHD

Donor T-cell depletion with infusion of expanded infection-specific T cells (e.g., CMV or EBV specific)

Reduces GVHD potential while protecting against significant infectious organisms

Lack of GVL effect; lack of broad anti-infectious immunity; expense and inefficiency of in vitro T-cell expansion; loss of survival/homing potential of cultured T cells

Donor T-cell depletion with infusion of expanded tumor antigen-specific T cells (expanded from natural repertoire or transduced with a T-cell receptor or chimeric antigen receptor)

GVL without GVHD

Lack of anti-infectious immunity; expense and inefficiency of in vitro expansion of tumor-specific T cells; loss of survival/homing potential of cultured T cells

Insertion of suicide gene (e.g., thymidine kinase) into donor T cells

Drug targeting inserted gene (e.g., ganciclovir) kills donor T cells to treat GVHD after GVL initiated.

Expense and inefficiency of in vitro transduction of T cells; loss of function/survival/homing potential of cultured T cells; risk for GVHD if transduction incomplete; curtailment of GVL when donor T cells killed in vivo

Block T-cell trafficking to epithelial GVHD target tissues (e.g., blockade of adhesion molecules or chemokines, sphingosine 1 phosphate agonists)

Permits lymphohematopoietic GVH reactions to occur, with associated GVL effects

Redundancy of trafficking pathways in inflammatory environment may limit efficacy; tumors outside of lymphohematopoietic system not targeted

Block injury/promote repair in epithelial target tissues (e.g., keratinocyte growth factor)

Permits lymphohematopoietic GVH reactions to occur, with associated GVL effects

Efficacy may be limited

APC = antigen-presenting cell; ATG = antithymocyte globulin; CMV = cytomegalovirus; DLI = donor lymphocyte infusion; EBV = Epstein-Barr virus; GVHD = graft-versus-host disease; GVL = graft-versus-leukemia effects; HCT = hematopoietic cell transplantation; MHC = major histocompatibility complex; NK = natural killer; TH2 = helper T lymphocytes type 2; TLI = total lymphoid irradiation.

Receptor 1 A

HLA group 1


Receptor 2 B Receptor 3


HLA group 2 HLA group 3 HLA group 4

Receptor 4 Autologous cell A C B

HLA group 1 HLA group 2

D Receptor 4

No inhibitory HLA ligand—cytotoxicity

Allogeneic cell E-FIGURE 49-3.  Killing of allogeneic targets by natural killer (NK) cells due to “missing self.” NK cells express clonally distributed inhibitory receptors (KIRs) with specificity for different groups of major histocompatibility complex (MHC) class I alleles, referred to in the figure as human leukocyte antigen (HLA) groups 1, 2, 3, and 4. Four different NK cells (A, B, C, and D) are shown, each with a different set of KIRs (referred to as receptors 1, 2, 3, and 4). Examples of HLA allele groups in the human are the HLA-Cw4, HLA-Cw3, and HLA-Bw4 groups; examples of KIRs are the ligands for these allele groups—namely, KIR2DL1, KIR2DL2/3, and KIR3DL1, respectively. Each functional NK cell has one or more inhibitory receptors that recognize a “self” (autologous) HLA molecule. Although some of the NK cells (e.g., cells A and B in the figure) will also find an HLA ligand to which their receptors bind on allogeneic cells, others (e.g., cells C and D) will not. The latter cells therefore will not receive inhibitory signals from the allogeneic cells and will kill them due to recognition by other (activating) receptors.

destruction of T cells with receptors that recognize donor antigens; it can be achieved during T-cell development in the thymus, for example, by induction of mixed chimerism in T-cell-depleted hosts. Deletion can also be applied to mature T cells in the periphery, for example, by transplantation of a tolerogenic organ or marrow graft in combination with blockade of costimulatory molecules. Anergy denotes the inability of T cells to respond fully to antigens they recognize, and it can be induced by antigen presentation without costimulation. Suppression has attracted considerable interest since the discovery that constitutively CD25+ T cells of the CD4+ subset have suppressive activity that is dependent on expression of the transcription factor Forkhead Box Protein 3 (FoxP3). These and other types of suppressive T cells (e.g., NKT cells, regulatory CD8+ cells and B cells, myeloid-derived suppressor cells) have been implicated in rodent transplantation tolerance models and in prevention of autoimmunity. The use of expanded regulatory cells has recently entered clinical trials, and both the ultimate practicality of the approach and the relative advantages of antigen-specific versus nonspecific regulatory cell therapy remain to be determined. There is also interest in strategies for activating or expanding regulatory T cells in vivo, thereby favoring the suppressive immune response over destructive alloimmunity.12 The developments in animal models and understanding of immune mechanisms described here have provided impetus for efforts to achieve immune tolerance in clinical transplantation. Every transplantation center has anecdotal cases of patients who have removed themselves from chronic immunosuppression without experiencing graft rejection. However, for every such patient, there are dozens more who have experienced rejection on dose reduction or removal of immunosuppressive drugs. Although trials of minimization and slow withdrawal of nonspecific immunosuppressive therapy are underway in organ transplant recipients, a major current limitation is the absence of good predictors of success. It remains to be seen whether recently identified molecular “tolerance signatures” will provide markers with sufficient predictive value to allow such withdrawal to be safely undertaken. One approach developed in animal models has been successfully applied to the induction of immune tolerance in a small group of patients receiving renal allografts. This approach, involving bone marrow transplantation after nonmyeloablative conditioning, which is much less toxic than standard HCT conditioning, was shown to be effective in the most stringent rodent and large-animal models before being evaluated clinically. Initial success using combined kidney and bone marrow transplantation in patients with renal failure due to multiple myeloma led to pilot studies in patients with renal failure without malignant disease, with encouraging preliminary results. This approach and others that have emerged from ongoing investigations provide hope that, in the future, transplantation might be routinely performed without the need for chronic immunosuppressive therapy, with its attendant

complications and limited ability to control chronic rejection.11 Because autoimmune diseases are major contributors to end-stage renal disease, diabetes, and other types of organ failure, the potential for tolerance strategies to reverse autoimmunity while inducing allograft tolerance is also a source of hope. All these approaches must, however, be undertaken with the caution that successful regimens could also lead to immune tolerance to active infectious organisms. GENERAL REFERENCES For the General References and other additional features, please visit Expert Consult at

CHAPTER 49  Transplantation Immunology  

GENERAL REFERENCES 1. Marsh SG, Albert ED, Bodmer WF, et al. Nomenclature for factors of the HLA system, 2010. Tissue Antigens. 2010;75:291-455. 2. Susal C, Opelz G. Current role of human leukocyte antigen matching in kidney transplantation. Curr Opin Organ Transplant. 2013;18:438-444. 3. Brunstein CG, Fuchs EJ, Carter SL, et al. Alternative donor transplantation after reduced intensity conditioning: results of parallel phase 2 trials using partially HLA-mismatched related bone marrow or unrelated double umbilical cord blood grafts. Blood. 2011;118:282-288. 4. Montgomery JR, Berger JC, Warren DS, et al. Outcomes of ABO-incompatible kidney transplantation in the United States. Transplantation. 2012;93:603-609. 5. Montgomery RA, Lonze BE, King KE, et al. Desensitization in HLA-incompatible kidney recipients and survival. N Engl J Med. 2011;365:318-326.


6. Li HW, Sykes M. Emerging concepts in haematopoietic cell transplantation. Nat Rev Immunol. 2012;12:403-416. 7. Blazar BR, Murphy WJ, Abedi M. Advances in graft-versus-host disease biology and therapy. Nat Rev Immunol. 2012;12:443-458. 8. Pilat N, Schwarz C, Wekerle T. Modulating T-cell costimulation as new immunosuppressive concept in organ transplantation. Curr Opin Organ Transplant. 2012;17:368-375. 9. Ferrer IR, Hester J, Bushell Wood KJ. Induction of transplantation tolerance through regulatory cells: from mice to men. Immunol Rev. 2014;258:102-116. 10. Fuchs EJ. Transplantation tolerance: from theory to clinic. Immunol Rev. 2014;258:64-79. 11. Zachary AA, Leffell MS. Desensitization for solid organ and hematopoietic stem cell transplantation. Immunol Rev. 2014;258:183-207. 12. Issa F, Robb RJ, Wood KJ. The where and when of T cell regulation in transplantation. Trends Immunol. 2013;34:107-113.


CHAPTER 50  Complement System in Disease  

50  COMPLEMENT SYSTEM IN DISEASE JOHN P. ATKINSON The complement system consists of plasma and membrane proteins that participate in host defense against infections and in clearance of cellular and extracellular debris, as well as in a wide variety of autoimmune and inflammatory states (Fig. 50-1).1,2 Complement is essential in innate immunity and a potent effector arm of adaptive (humoral) immunity. It is a first responder, especially in blood, to bacterial and viral invasion (Table 50-1). It helps to maintain sterility (“guardian of the intravascular space”) by depositing within seconds its opsonic and membrane-perturbing fragments on a pathogen’s surface. A second major activity of complement is to promote the inflammatory response via the release of soluble fragments (anaphylatoxins). They bind to their receptors, leading to cellular activation, including chemokinesis and chemotaxis by phagocytic cells, and thereby enhance protection against infections. Furthermore, the deposition of complement fragments on immune complexes keeps them from precipitating and promotes their adherence to red blood cells (RBCs) for a hand-off to monocytes and dendritic cells in the liver and spleen. Through these interactions, complement also instructs the adaptive immune response. Antigens decorated by complement proteins are taken up Complement

Membrane modification

Inflammation (anaphylatoxins)




Degranulation of mast cell

Induction of neutrophil chemotaxis



Nature’s adjuvant FIGURE 50-1.  Function of the complement system. The most important function of the complement system is to alter the membrane of the pathogen by coating its surface with clusters of activation fragments. In one case, they facilitate the key process of opsonization in which C4b and C3b interact with complement receptors. In the other case, as with certain gram-negative bacteria and viruses, the membrane attack complex lyses the organism. The second critical function of complement is to activate cells and thus promote inflammatory and immune responses. The complement fragments C3a and C5a (known as anaphylatoxins) stimulate many cell types such as mast cells to release their contents and stimulate phagocytic cells to migrate to sites of inflammation (chemotaxis). Through these phenomena of opsonization and cell activation, complement serves as nature’s adjuvant to prepare, facilitate, and instruct the host’s adaptive immune response. Because complement activation occurs in a few seconds, this innate immune system initially engages most pathogens, especially those that try to enter the vascular space. As will be illustrated, these basic functions are also required to handle immune complexes and prevent autoimmunity. (Modified from Arthritis Foundation. Primer on the Rheumatic Diseases. 12th ed. Arthritis Foundation; Atlanta, Ga 2001.)

CHAPTER 50  Complement System in Disease  





Ancient innate system of immunity predominantly found in blood (the “guardian of the intravascular space”) Capable of rapidly opsonizing and lysing bacteria and viruses (millions of active fragments can be deposited on a target) Works in seconds! Most proteins are synthesized by the liver Constantly turning over (AP protein C3 “ticks over” at a rate of 1% to 2% per hr) The AP also features a feedback or amplification loop, which requires tight control Effector arm of the humoral immune system (IgM and IgG) Critical for clearance of self-debris (garbage removal) After immunoglobulins and albumin, complement proteins are among the most abundant in blood Nature’s adjuvant (almost all foreign antigens are coated with complement fragments); instructs the adaptive immune response A deficiency of an activator leads to bacterial infections or autoimmunity (SLE) A deficiency of a regulator leads to undesirable cellular and tissue damage at sites of injury or degeneration (excessive activation)



(C3b > C4b, C1q, MBL)*

Membrane perturbation including lysis (the membrane attack complex)


Proinflammatory via cellular activation (the anaphylatoxins and their receptors)

(C3a, C5a)

*C3b is the major opsonin of the complement system. C1q and MBL (mannose- or mannanbinding lectin) both participate in classical and lectin pathway activation, respectively, but also bind to their specific receptors upon attachment to a target.

by monocytes, follicular-dendritic cells, B lymphocytes, and other antigenpresenting cells, resulting in an adaptive immune response. (The complement system is often called “nature’s adjuvant.”) Thus, complement activation is required for an optimal antibody response to most foreign antigens. Individuals lacking a functional complement system are predisposed to bacterial infections, predominantly by encapsulated organisms, including streptococcus, staphylococcus, Haemophilus spp., and Neisseria spp.3 Surprisingly, a complete deficiency in an early component of the classical complement pathway predisposes to autoimmune diseases, particularly systemic lupus erythematosus (SLE).4 This association suggests that complement is required not only for host defense against foreign agents but also to identify and safely clear self-materials (debris removal), particularly RNA and DNA species. A remarkable feature of the complement system is that it reacts within seconds (Table 50-2). In less than 2 minutes, it can coat an encapsulated gram-positive bacterium with several million C3b opsonic fragments and lyse gram-negative bacteria by insertion of its terminal components (the membrane attack complex [MAC]). It works even more efficiently if driven by IgM or IgG binding to an antigen on a microbial membrane to activate the cascade. Antibodies and lectins direct the activation process to the pathogen’s surface. Overall, the complement cascade is designed to become engaged on the surface of a pathogen, particularly bacteria. Plasma and membrane regulators of complement activation inhibit formation on normal “self ” cells. Much of the complement-mediated pathology revolves around the alternative pathway’s (AP’s) amplification loop. This feedback amplification loop is key in triggering activation early in an immune response; however, it must be rigorously regulated to prevent activation on normal self and excessive activation on injured self.5 Approximately half of the proteins associated with the complement system are dedicated to the control of its activation and effector functions, especially to maintain homeostasis of the AP’s amplification loop. In clinical medicine (Table 50-3), the complement system participates in three pathologic processes (Table 50-4): (1) an inherited decrease in functional activity leading to increased susceptibility to bacterial infections and to autoimmunity, (2) mediating undesirable tissue damage upon activation by autoantibodies and immune complexes, and (3) excessive activation at sites of tissue injury in individuals carrying genetic variants in regulators. Knowledge of how complement is activated and how it can be controlled points to opportunities for the development of therapeutic agents such as anti-C5 monoclonal antibody (mAb) therapy, which has been recently approved to treat several complement-dependent hemolytic disorders.


Classical Pathway

The binding of IgM or IgG to a target antigen activates this exceptionally powerful and quick acting pathway to destroy microbes (Figs. 50-2 and 50-3). The classical pathway (CP) reaction cascade is designed to opsonize and perturb the surface membrane of microorganisms. Of course, autoantibodies also trigger this highly efficient CP. Complement action mediated by immune complexes may then lead to cellular and tissue damage. Instructive examples of autoantibodies and complement-mediated diseases are immune hemolytic anemias, myasthenia gravis, and bullous pemphigoid. The basic problem or pathologic defect in this type of human disease is, of course, the formation of the autoantibody. A misidentification of self that has occurred because of a breaking of tolerance. In this pathologic situation, the complement system is working at the behest of the autoantibody.

AP = alternative pathway; SLE = systemic lupus erythematosus.

TABLE 50-3  PARTICIPATION OF THE COMPLEMENT SYSTEM IN HUMAN DISEASE Activation by autoantibody (formation of immune complexes) Engagement with modified self (clearance of debris or garbage) • Degenerative processes (diseases of aging such as age-related macular degeneration) • Cell and tissue damage (ischemia-reperfusion injury; atypical hemolytic uremic syndrome)

TABLE 50-4  PATHOLOGIC CONDITIONS ASSOCIATED WITH COMPLEMENT ACTIVATION Examples of diseases in which complement activation contributes to the immunopathology:   Atypical hemolytic uremic syndrome*†   Paroxysmal nocturnal hemoglobinuria†   Age-related macular degeneration*†   Membranoproliferative glomerulonephritis (types 1, 2 and 3)†‡   Myasthenia gravis‡   Bullous pemphigoid‡   Systemic lupus erythematosus/antiphospholipid syndrome‡   Rheumatoid arthritis‡   Immune hemolytic anemias‡   Immune vasculitis (the ANCA-positive syndromes)‡   Ischemia reperfusion injury*†   Allotransplantation‡   Serum sickness‡   Exposures to foreign materials (e.g., membranes, nanoparticles)* *Injury, ischemia, trauma, degeneration, or foreign body is the trigger (innate immune activation). † Lack of adequate regulation contributes to disease pathogenesis. ‡ Antibody dependent activation of the complement system (adaptive humoral immune activation).

The CP is also activated by means other than the formation of IgM- and IgG-bearing immune complexes. β-Amyloid in the neuritic plaques of patients with Alzheimer disease directly engages the CP via an interaction with C1q. Likewise, C-reactive protein (CRP) and serum amyloid protein (SAP) bind to chromatin and other ribonucleoprotein complexes released from apoptotic cells, and these types of complexes activate the CP. As noted, the CP plays a key role in the opsonization and removal of nuclear debris. Approximately 80% of patients with hereditary absence of C1q or C4 develop SLE. Deposits of CRP and activated C1 have been demonstrated in ischemic tissue such as infarcted human myocardium. These observations indicate that CP activation via these antibody-independent means is critical in protecting against autoimmune responses by facilitating debris clearance. Regulation of the CP activation occurs at two levels. First, the serine protease inhibitor (serpin) known as the C1-inhibitor (C1-INH) blocks the activity of many proteases, including factor XIIa, kallikrein, and factor XIa of


CHAPTER 50  Complement System in Disease  

Complement Activation CLASSICAL PATHWAY



Antibody binds to specific antigen on pathogen surface

Mannose-binding lectin binds to pathogen surface

Pathogen surface creates local environment conducive to complement activation

FIGURE 50-2. The three pathways of complement activation.

TABLE 50-5  TISSUE INJURY OR DEGENERATION AND COMPLEMENT ACTIVATION* Age-related macular degeneration Osteoarthritis (degenerative joint disease) Ischemic stroke Myocardial infarction Traumatic brain injury (e.g., liver, kidney, gut) Ischemia-reperfusion injury Burns Acute respiratory distress syndrome Septic shock Multiorgan failure syndromes Alzheimer disease

Alternative pathway Continuous and spontaneous tickover of C3 in blood and on cells C3a C3 (H2O) or C3b + C3b + FB C3bB + FD Ba Feedback loop

*In these conditions, complement activation leads to deposition of fragments at the site of injury; however, how much of the tissue injury is attributable to complement system is unknown. In many cases, animal models support a pathologic role for the complement system. Only in age-related macular degeneration do we also have powerful genetic evidence in humans to indicate a key role for the complement system.

the clotting system as well as C1r, C1s, and MASP2 of the complement system. The importance of C1-INH is exemplified by its role in hereditary angioedema (Table 50-6). In this dominantly inherited disease, a deficiency of C1-INH allows uncontrolled proteolysis of C4 and C2 and generation of bradykinin, leading to recurrent swelling episodes. This serpin prevents chronic activation of the CP cascade and, after a few minutes, helps to shut down the system. CP activation is also regulated by multiple inhibitors at the key step of C3 activation. These plasma and membrane proteins inhibit C3 convertase formation on healthy self. Membrane regulators are highly expressed on most cell types, where they prevent activation on normal self and overexuberant activation on altered and nonself.

Lectin Pathway

The protein mannose-binding lectin (MBL) is a member of the collectin family that also includes pulmonary surfactants A and D.6 MBL has a structure similar to C1q in that it consists of several subunits; namely, a globular recognition head domain for carbohydrates and a collagen-like tail that interacts with serine proteases. In the case of MBL, the globular domain is a lectin (protein) that binds to repeating mannose and N-acetylglucosamine residues on the surface of pathogens (see Figs. 50-2 and 50-3). Many microorganisms are recognized by MBL, including gram-positive and gram-negative bacteria, mycobacteria, fungi, parasites, and viruses (including human immunodeficiency virus 1 [HIV-1]). In general, as would be expected, mammalian glycoproteins and glycolipids are not readily recognized by MBL and the related lectins (ficolins and collectins) that activate the lectin pathway. Three serine proteases, MASP-1, MASP-2, and MASP-3, associate with MBL (and the ficolins and collectins) through their collagen-like domain. This is analogous to the association of C1r and C1s with C1q. Activation of MASP-2, with some help from MASP-1, results in cleavage of C2 and C4, leading to formation of the classical/lectin pathway C3 convertase (C4b2a). Genetic variations in the structural and regulatory portions of the MBL gene lead to wide differences in serum levels. A low level of MBL is associated with recurrent infections in children and adults and is a risk factor for the development of SLE. More striking is the association of low levels of MBL with infections in the setting of the treatment of SLE. For example, heterozygous MBL deficiency has been associated with a fourfold increase in the risk of bacterial pneumonia and hom*ozygous deficiency with a more than 100-fold increase.

C3bBb + P (properdin) C3bBbP +

Classical pathway

C3b on targets Lectin pathway

Enzymes Proconvertase

C3 convertase

Stabilized C3 convertase



(C3b)2 BbP +

C5 convertase

C5 C5a C5b + C6, C7, C8, C9 C5b-C9 (MAC)

FIGURE 50-3.  Complement activation pathways. In the reaction cascade shown, C3b or C3 (H2O) binds the proenzyme factor B (FB), and the C36B complex then cleaved by the protease factor D (FD). The addition of properdin (P) to the enzyme complex increases the half-life of the enzyme complex approximately 10-fold. Although the source of the C3b can be from spontaneous turnover or via lectin pathway (LP) and classical pathway (CP) activation, the alternate pathway (AP) feedback loop commonly takes over to generate most of the C3b that binds to a target. The alternate pathway is continuously turning over. If activated C3b or C3 (H2O) remains in the fluid phase, it is rapidly inhibited by the plasma regulator factor H. if activated C3 binds to normal or healthy self, it is prevented from forming a convertase by the ubiquitously expresses membrane cofactor protein (MCP [CD46]) and decay-accelerating factor (DAF [CD55]). DAF “kicks out” the catalytic Bb domain (a temporary stop), but MCP is a permanent stop because, upon its binding, the C3b is proteolytically cleaved to inactive C3b (iC3b) by a serine protease known as factor I. The feedback loop is a powerful amplification system. A single Escherichia coli organism in blood can be coated with several million C3bs in a couple of minutes!

Alternative Pathway

The AP takes advantage of the fact that C3 undergoes spontaneous, chronic, low-grade activation (Figs. 50-2 to 50-4). This C3b may covalently attach to any cell; however, on normal self, amplification of the cascade is blocked by inhibitors. In contrast, deposition on polysaccharides of bacterial membranes and to other targets, such as endotoxin and virally infected cells, leads to a rapid engagement of this pathway. These sites, similar to immune complexes and almost any type of biomaterial (cardiopulmonary bypass and hemodialysis membranes, nanoparticles, and so on), lack regulators, so rapid, massive activation may occur. During spontaneous activation, called tickover, small amounts of activated C3 are continuously generated (C3 turns over in blood at 1% to 2%/hr). It can initiate a feedback loop and cleave more C3 to C3b. Also, the initial C3b


CHAPTER 50  Complement System in Disease  




C1r, C1s, MASP-2

Binds to and displaces C1r and C1s from C1q and MASP-2 from MBL



C4bp* (C4 binding protein)

C4b, GAGs

Displaces C2a (DAA); cofactor for C4b cleavage by factor I (CA)

No clinical syndrome clearly defined

CPN-1 (carboxypeptidase-N)

C3a, C5a

Inactivates C3a and C5a

Urticaria and angioedema

Factor H*†

C3b, C3d, GAGs

Displaces Bb from AP C3 and C5 convertases (DAA) and is a cofactor for factor I to cleave C3b (CA)

AMD, aHUS, C3 glomerulopathies; bacterial infections secondary to low C3

Factor I†

C3b, C4b

Serine protease; cleaves C3b and C4b, requires a cofactor protein (CA)

AMD, aHUS; bacterial infections secondary to low C3

Protein S (vitronectin)


Inhibits membrane attachment by C5b67

None defined



DAF (CD55)

C3 and C5 convertases

Displaces Bb from AP convertase and C2a from CP or LP convertases, respectively


Membrane cofactor protein (MCP, CD46)

C3b, C4b

Cofactor for factor I (CA)


Protectin (CD59)

C8, C9

Inhibits MAC formation or insertion


CR1 (CD35) (immune adherence or C4b/C3b receptor)

C3b, C4b, C3, and C5 convertases

Cofactor for factor I to cleave C4b and C3b (CA); displaces Bb from C3b and C2a from C4b to inhibit convertases (DAA)

No complete deficiency described; decreased levels in immune complex– mediated diseases such as lupus


C3b, iC3b, C3c

Inhibits activation of AP

None defined



*Factor H and C4bp also bind to surfaces, particularly at sites of tissue and cellular injury, where they also carry out regulatory activity. † If heterozygous deficient, individual is predisposed to AMD and aHUS. If hom*ozygous deficient, the AP turns over excessively, resulting in kidney disease (C3 glomerulopathies) and bacterial infections (secondary to the very low C3). aHUS, atypical hemolytic uremic syndrome; AMD, age-related macular degeneration; AP, alternative pathway; (C3b)2 Bb, alternative pathway C5 convertase; C3bC4bC2a, classical and lectin pathway C5 convertase; C4bC2a, classical and lectin pathway C3 convertase; CA, cofactor activity; CR1, complement receptor type 1; CRIg, complement receptor of the Ig superfamily; DAA, decay-accelerating activity; GAG, glycosaminoglycan; HAE, hereditary angioedema; MAC, membrane attack complex; MASP, mannan-binding lectin-associated serine protease; MBL, mannan or mannose binding lectin; PNH, paroxysmal nocturnal hemoglobinuria.

Vitronectin, clusterin, protectin CP/LP C5 con










AP C5 con


Increasing lipophilicity Topography of C5b-9 assembly

C6 β α'

β C7 α'




β α' C6

α' C6

γ α C8

C6 C7

C7 α β



β α' C6

C9 site

C7 α

γ C9 C9C9

β C9



C5b-6 C5b-7



FIGURE 50-4.  Activation of C5 and the membrane attack complex (MAC). A, The C5 convertases (“con”) are the same as C3 convertases except a C3b has been attached to C4bC2a or a second C3b in the case of AP C5 convertase. B, Schematic representation of the assembly of the assembly of the MAC on a cell membrane. C5b (composed of two chains) binds C6 and then C7. The C5b-7 complex can insert into a membrane and then bind C8 (composed of three chains) and multiple C9s to form a pore or channel in the membrane. (Modified from Liszewski MK, et al. The Human Complement System in Health and Disease. Marcel Dekker; New York, NY 1998.)

may be derived from either the classical or lectin pathway. Thus, activation of complement by any one of the three pathways has the potential to be rapidly magnified.7 The AP C3 convertase is negatively controlled (to maintain homeostasis) both in the fluid phase and on host cells by two abundant plasma proteins and two widely expressed membrane proteins.8

The central role of the AP as an amplifier of complement activation is borne out by its association with a number of clinicopathologic states in the setting of deficient regulation (Tables 50-6 and 50-7). For example, multiple forms of membranoproliferative glomerulonephritis are associated with excessive C3 fragment deposition in the kidney because of either the


CHAPTER 50  Complement System in Disease  







Lyse RBCs

Acquired hemopoietic somatic stem cell mutation in gene required for synthesis of GPI anchor

mAb to C5



Damage to endothelial cells

Inherited loss of function variants in AP regulators or gain of function variants in AP activators

mAb to C5



Bradykinin generation

Autosomal dominant variants in the C1-inhibitor gene

C1 inhibitor replacement Bradykinin receptor blockage Kallikrein inhibitor

Yes Yes Yes


Degeneration of the retina

Inherited variants in a regulator (FH or FI) or gain of function in an alternative pathway component (C3 or FB)

Clinical trials in progress


aHUS= atypical hemolytic uremic syndrome; AMD = age-related macular degeneration; FDA = Food and Drug Administration; HAE= hereditary angioedema; GPI = glycosyl phosphatidylinositol; mAb = monoclonal antibody; PNH= paroxysmal nocturnal hemoglobinuria; RBC = red blood cell.

presence of autoantibodies (C3 or C4 nephritic factors) that stabilize C3 convertases or a genetic deficiency in complement regulatory protein (factor H or factor I).9 Likewise, atypical hemolytic uremic syndrome (i.e., not associated with a preceding enteropathic infection featuring a Shiga-like toxin) occurs in individuals who harbor heterozygous missense mutations in factor H or I or have gain-of-function mutations in factor B or C3.10 Genome-wide association and targeted deep sequencing studies have also linked age-related macular degeneration (AMD) to functional coding mutations in factor H and factor I and more uncommonly in factor B and C3.11 Finally, rodent models of rheumatoid arthritis, SLE, and ANCA-positive vasculitic syndromes are ameliorated if the AP is disrupted.

C3 and C5 Convertases

The three activation pathways converge at C3. A remarkable feature of C3 is the presence of a thioester bond. Buried within the three-dimensional structure of the C3 protein lies a γ-carboxy group of a reactive glutamic acid residue linked to a cysteine in an “internal thioester.” Upon its cleavage, for a few microseconds, a covalent attachment can occur via an ester or amide linkage to any nearby hydroxyl or amino group. Most of the cleaved thioester bonds are hydrolyzed by water to produce a form of C3 (known as C3 [H2O]); however, a substantial percentage forms an amide or ester bond to amino groups or carbohydrates thereby covalently attaching C3b to a target’s surface. Also, the addition of a C3b to C4b2a (classical/lectin pathway C3 convertase) or to C3bBb (AP C3 convertase) then forms a convertase for C5 (C3bBbC3b for the AP and C4bC2aC3b for the CP/LP).

Regulators of Complement Activation at the C3 and C5 steps

The regulators of complement activation (RCA) (see Table 50-2) limit the production of C3b, primarily by the AP C3 convertases. Because the addition of C3b to a C3 convertase makes it a C5 convertase, regulation of the two enzyme complexes is linked. Modulation of their activity on host cells limits tissue destruction and the production of inflammatory mediators. The RCA proteins control complement activation by two processes. Decayaccelerating activity refers to when the inhibitor transiently binds to C3b or C4b in the convertase and thereby dissociates the other members of the complex, rendering it enzymatically inactive (as the component released is the catalytic domain of the protease). The second is cofactor activity, which requires recognition of C3b or C4b by a plasma cofactor protein. Upon this interaction, the protease, factor I, cleaves C3b or C4b. Cleavage of C3b by factor I renders the convertase irreversibly inactive (generates iC3b which cannot participate in convertase formation).

Membrane Attack Complex

The cleavage of C5 generates C5a, the most potent of the complement anaphylatoxins, and C5b. C5b associates with C6 and C7 to create a lipophilic trimer as the initial part of the MAC (Fig. 50-4). The C5b67 trimer inserts into the lipid bilayer and serves as a binding site for C8 and C9. C9 selfpolymerizes, leading to 12 to 18 C9 molecules that form a ring structure (completing the MAC). The MAC resembles a doughnut with a 10-nm pore running through the center. This pore allows water and ions to enter the cells, ultimately leading to osmotic lysis. Many pathogens such as gram-positive bacteria possess a capsule that makes them resistant to lysis.12 Opsonization leading to phagocytosis is thus the major means of eliminating such organisms.



C5aR (CD88) Neutrophils Eosinophils Basophils Mast cells Monocytes Hepatocytes Pulmonary epithelium Neuronal cells Endothelial cells Renal epithelial/mesangial cells

Chemotaxis Enzyme release Generation of reactive oxygen species Upregulation of adhesion molecules Increased synthesis of IL-1, IL-6, and IL-8 Prostaglandin and leukotriene synthesis Increased synthesis of acute phase reactants Increased IL-8 Cellular activation Increased expression of P-selectin Proliferation Synthesis of growth factors

C3aR Eosinophils Mast cells Platelets Epithelial, endothelial, etc.

Chemotaxis Enzyme release Generation of reactive oxygen species Upregulation of adhesion molecules Cellular activation

CNS = central nervous system; IL = interleukin.

The MAC appears to be essential only for elimination of Neisseria spp. Individuals completely deficient in C5, C6, C7, C8, or C9 are at an increased risk only for meningococcal and gonococcal infections. C9 deficiency is a common immunodeficiency in Japan, with a heterozygote frequency of 3% to 5%. Thus, heterozygous deficiency seems to not be deleterious to the population in general but may have a selective advantage. Extensive complement activation during an inflammatory response can result in sufficient MAC deposition to produce host cell lysis. Host cells, however, have mechanisms in place to resist the osmotic changes caused by the MAC and to block assembly of the MAC as it is formed (the protein is known as protectin or CD59). Rather, the nonlethal effects of sublytic MAC deposition are more likely to contribute to pathology. In most cells, this occurs through a general activation of multiple cell signaling pathways. The response to MAC deposition at sites of complement activation depends on the cell type (Table 50-8). In phagocytic cells, such as neutrophils or macrophages, sublytic MAC insertion leads to the production of reactive oxygen species (e.g., superoxide, hydrogen peroxide) as well as release of prostaglandins and leukotrienes. Platelets undergoing a “MAC attack” incorporate phosphatidylserine on their outer membrane, facilitating formation of blood coagulation enzyme complexes with a potentially procoagulant effect. On endothelial cells, MAC deposition induces the synthesis of interleukin-1α (IL-1α), which leads to further autocrine and paracrine endothelial cell activation. It stimulates a procoagulant state by (1) altering the phospholipid composition of the endothelial membrane; (2) inducing the synthesis of tissue factor and upregulating the synthesis of plasminogen activator inhibitor; (3) upregulating the expression of adhesion molecules, including intercellular adhesion molecule 1 (ICAM-1) and E-selectin; and (4) stimulating endothelial cells to proliferate through growth factor production. In summary,

CHAPTER 50  Complement System in Disease  





Eculizumab C3





C3b Bb


C3b B





C3 con

C5 con FIGURE 50-5.  Complement activation and the mechanism of action of eculizumab (monoclonal antibody [mAb] to C5). The alternate pathway (AP) constantly undergoes “tickover” but can also be primed by the classical pathway (CP) and lectin pathway (LP). The C3b that is formed interacts with factor B (B), which is then cleaved by factor D to form a C3 convertase (C3Bb). As more C3b is generated, some binds to the C3 convertase to form a C5 convertase. mAb to C5 (eculizumab) prevents the cleavage of C5 by the C5 convertase. Not shown is properdin that binds to both the C3 and C5 convertases to increase their half-lives approximately five- to 10-fold (from ≈30 seconds to several minutes).7 (Modified from Wong EK, Goodship TH, Kavanagh D. Complement therapy in atypical haemolytic uraemic syndrome (aHUS). Mol Immunol. 2013;56(3):199-212.)

although cell death does not usually occur, deposition of the sublytic levels of MAC leads to a potentially dangerous situation, with increased inflammation, a procoagulant state, and cellular proliferation. Of course, part of this response is necessary at sites of injury to eliminate pathogens and debris and to facilitate wound repair. The short duration of complement activation and the presence of inhibitors help to maintain homeostasis. Regulation of MAC formation is important clinically (Fig. 50-5). Two plasma proteins, clusterin and S-protein (vitronectin), bind the C5b-7 complex and prevent its association with the lipid membrane. C8 and multiple C9 molecules adhere to this soluble complex, termed soluble C5b-9, which is lytically inactive. CD59 (protectin) is a membrane-bound inhibitor of MAC formation. This small glycoprotein is attached to the cell membrane through a glycosyl phosphatidylinositol tail (GPI anchor). It binds to C5b-8, inserted in the cell membrane, to prevent binding to and polymerization of C9. The expression of CD59 is defective in patients with paroxysmal nocturnal hemoglobinuria (PNH), owing to the failure to synthesize the GPI anchor used by this and many other membrane proteins (including decay-accelerating factor [DAF]) to insert on the cell. The clinical features of PNH are primarily chronic hemolysis and intermittent thrombosis. Hemolysis is caused by complement activation on RBCs because of a lack of DAF and particularly CD59. Thrombosis is likely secondary to intravascular complement activation, leading to endothelial cell activation. The primary defect is an acquired hematopoietic stem cell mutation of a gene on the X chromosome responsible for encoding the first enzyme in the pathway to synthesize a GPI-anchor.


The anaphylatoxins serve a key early role in initiating a local inflammatory response as they trigger pathways to prepare a cell to face a pathogen or injury (Table 50-9). Similar to the MAC, anaphylatoxins are another major source of potential pathologic damage to self that results from complement activation. These peptides, C3a and C5a, are cleaved from their respective proteins during complement activation. They were named in 1910 to describe their toxic effects, including shock after the transfer of complement-activated serum into laboratory animals. They are 77 (C3a) or 74 (C5a) amino acids long and contain a key carboxy (C)-terminal arginine. They interact with the anaphylatoxin receptors. In plasma, the C-terminal arginine is removed by carboxypeptidase-N from anaphylatoxins not bound to their receptors. Depending on the response studied, this removal totally inactivates the anaphylatoxin or reduces its potency by about 1000-fold. The C5a receptor (C5aR [CD88]) is a seven-transmembrane-spanning protein that couples ligand binding to G-protein signaling. Expressed on myeloid cells, particularly neutrophils and eosinophils, it mediates the potent


Most cells

Increased intracellular calcium flux Activation of G proteins Activation of protein kinases Activation of transcription factors Proliferation

Neutrophils and macrophages

Release of reactive oxygen species Activation of phospholipase A2 Release of prostaglandins, thromboxane, and leukotrienes


Release of ATP Increased P-selectin expression Procoagulant membrane changes

Endothelial cells

Increased synthesis of IL-1α Increased release of tissue factor Increased release of von Willebrand factor Increased synthesis of basic fibroblast and plateletderived growth factors


Increased synthesis of prostaglandin Increased synthesis of IL-6 Increased production of matrix metalloproteinase

Glomerular epithelium

Activation of phospholipase A2 Synthesis of prostaglandin Increased synthesis of collagen and fibronectin


Increased synthesis of myelin basic protein and proteolipids Increased proliferation

ATP = adenosine triphosphate; IL = interleukin.

chemoattractant property of C5a for both of these cell types. Signaling through CD88 leads to rapid secretion of all granule contents. These include lipases and proteases as well as lactoferrin from neutrophils, and peroxidase, major basic protein, and cationic protein from eosinophils. C5a also induces the release of cytokines, such as tumor necrosis factor (TNF), IL-1, IL-6, IL-8, and adhesion molecules, promoting the inflammatory response. The C5aR is expressed by numerous other tissues, including hepatocytes, bronchial and alveolar epithelium, vascular endothelium, renal mesangial and tubular epithelial cells, and brain neuronal cells. These cells are activated by receptor engagement, leading to production and release of cytokines, chemokines, and prostaglandins and to cellular proliferation. The C3a receptor is also a seven-transmembrane-domain protein. It is expressed on almost all myeloid cells, including mast cells, where it mediates the release of allergic mediators. The C3aR also has been detected on many tissues, including in the brain and lung. The anaphylatoxins have multiple biologic effects. In general, they cause smooth muscle contraction and recruitment of granulocytes, monocytes, and mast cells. In theory, they can contribute to the pathophysiology of any inflammatory condition. In disease models, C3a and C5a have been shown to play a role in diseases such as acute respiratory distress syndrome (ARDS), multisystem organ failure, septic shock, myocardial ischemia-reperfusion injury, asthma, rheumatoid arthritis, SLE, and inflammatory bowel disease. The anaphylatoxin peptides also are responsible for the “postpump” syndrome seen in patients undergoing cardiopulmonary bypass or hemodialysis. Exposure of blood to dialysis or perfusion membranes leads to complement activation. Within minutes of starting bypass, there is a sharp increase in the levels of C3a and C5a in the extracorporeal circuit being returned to the patient. This increase can be associated with respiratory distress, pulmonary hypertension, and pulmonary edema. It has been shown that the length of time that patients stay on the ventilator after bypass surgery correlates with the level of C3a generated during reperfusion. C3a and C5a have been implicated in the initiation and prolongation of ARDS and multisystem organ failure. After severe trauma, levels of C3a have been measured that suggest activation of the entire circulating C3 pool. This activation leads to bronchoconstriction, increased vascular permeability, hypotension, and vascular plugging with leukocytes. The activation of white blood cells continues the cycle of tissue damage with further complement activation. Continued elevation of C3a in shock or ARDS is a poor prognostic sign. C3a and C5a also appear to play a major role in the pathogenesis of asthma.


CHAPTER 50  Complement System in Disease  







x7 MCP

Other concerns about complement inhibition include whether it is shortor long term and whether it is systemic or local. Long-term inhibition of complement, particularly at one of the early steps, is likely to predispose to infection and possibly autoimmunity. Short-term (hours to days) inhibition at any step is unlikely to cause problems. Given that inflammation is usually a local phenomenon, several mechanisms are being tested to target complement inhibitors to these sites. In this way, higher levels of inhibition can be achieved when needed and with lower doses of inhibitor.

Natural Complement Inhibitors C3b Factor H


FIGURE 50-6.  Model of the complement regulators required to inhibit complement activation on self at the steps of C3 and C5 cleavage. Circles represent individual complement control repeats (~60 amino acids each), and shading indicates higher organizational units composed of several repeats. The approximate locations for binding of C3b and C4b fragments are indicated. Factor H and C4bp are plasma proteins. C4bp = C4-binding protein; CR = complement receptor; DAF = decay-accelerating factor; MCP = membrane cofactor protein.

Complement Receptors

Opsonization of target by C4b and C3b is effective in preventing infections because these two complement fragments (and the further cleavage products in the case of C3b) are ligands for complement receptors (Fig. 50-6). After covalent attachment of C4b and C3b, immune adherence occurs between the opsonized microbe and immune cells, predominantly neutrophils, monocytes, and macrophages. Complement opsonins are highly effective mediators of immune adherence. On phagocytic cells, this is the prelude to the ingestion and destruction of the target antigen. On RBCs, immune adherence is followed by transfer of the C4b/C3b-coated cargo to monocytes and macrophages in the liver and spleen. CR1 is particularly efficient at immune adherence. Proteolytic modification of C3b leads to iC3b, which is a ligand for the highly phagocytic CR3 and CR4. A further degradation of iC3b to C3d leads to an interaction with CR2 to lower the threshold of B cell activation. Overall, the process is designed with two goals in mind: first is destruction by phagocytosis of the microbe and second is to coat microbial antigens for an adaptive immune response. For example, follicular dendritic cells and B lymphocytes express CR1, CR2, CR3, and CR4 that facilitate complementcoated antigens to be bound, internalized, and presented to other immune cells. CR3 and CR4 facilitate phagocytosis, and CR2 on follicular dendritic cells facilitates immunologic memory generation.


Given the many disease states in which complement is one of the central mediators of pathology, it is no surprise that complement inhibitors are in preclinical or clinical development for treatment of human diseases (see Table 50-2 and Fig. 50-6). These inhibitors take several different forms. Whereas some are variations of physiologic inhibitors, others are the products of molecular biologic searches for novel compounds. It is important to consider where in the complement pathway to design an inhibitor to act. Inhibition of the activation pathways limits the production of biologically active peptides. Inhibiting the activation of C3 not only prevents the generation of the C3a anaphylatoxin but also may leave the patient susceptible to infection by limiting the deposition of C3b on targets as an opsonin. Inhibition of C3b deposition would decrease the patient’s ability to clear immune complexes, potentially resulting in renal, pulmonary, and vascular damage. It also might promote the development of antibodies to self-antigens. Inhibition of the C5 convertases is an attractive goal because it would prevent the generation of the C5a anaphylatoxin and the MAC (see Fig. 50-5). This strategy would inhibit complement activation without limiting C3b deposition. Inhibitors based on this concept have been successful; the mAb to C5 is approved by the Food and Drug Administration (FDA) to treat PNH and atypical hemolytic-uremic syndrome (aHUS).

Naturally occurring compounds that control complement activation include products or extracts of plants, fungi, insects, bacteria, viruses, and venoms.12 The mechanism of complement inhibition by some of these natural products is known and is of clinical and experimental importance. In particular, to protect themselves from the host’s complement system, poxviruses, herpesviruses, and flaviviruses produce either mimics of the human regulators that they at one time hijacked from their hosts or proteins that bind the hosts’ regulators such as the plasma protein factor H. Bacteria also express a wide variety of inhibitors of the human complement system. Staphylococcus aureus, for example, synthesizes up to 10 distinct proteins that inhibit at almost every key step of the complement cascade. Cobra venom factor (CVF) is a modified form of cobra C3b secreted by venom glands in the oral cavity. It is an 144,000-dalton glycoprotein that forms an AP convertase in association with host factor B. Upon injection, CVF leads to massive activation of the AP, leading to shock and pulmonary microvascular injury in experimental animals. It is resistant to the host’s inhibitors because the site for that interaction is altered on CVF. Perhaps the most widely used natural inhibitor of complement activation is heparin. It decreases activation of the CP and AP. In clinical practice, the anticomplementary effect of heparin has been used to prevent complement activation during cardiopulmonary bypass. Measurement of complement activation products such as C3a or soluble C5b-9 after bypass showed decreases of 35% to 70% for adult and pediatric patients when heparin-coated extracorporeal circuits were used. Although numerous studies have looked at the decrease in complement activation by heparin-coated bypass circuits, there have been few attempts to correlate this with clinical outcome.


The complement inhibitor that has achieved the widest attention as a therapeutic agent to stop complement activation is a mAb to C5 that prevents its cleavage to C5a (potent anaphylatoxin) and C5b (initiator of the MAC) (see Fig. 50-5). The generation of the C3b and C4b still occurs, allowing opsonization of pathogens and formation of immune complexes. Because activation of early complement components is also important for the maintenance of tolerance to self-antigens, inhibition of C5 activation is less worrisome than inhibition of C3 activation. The one consequence of C5 deficiency in humans is an increased risk of Neisseria infections that can largely be mitigated through vaccination. The anti-C5 mAb eculizumab has been approved for use in patients with PNH and aHUS. The complement system is undergoing a renaissance. There are several reasons but probably the foremost is the discovery of mutations in complement regulators leading to aHUS and AMD. Second is the therapeutic success of a mAb to C5 in the treatment of aHUS and PNH. Third is the introduction of purified C1-Inhibitor, kallikrein inhibitors, and a bradykinin receptor antagonist to prevent and treat swelling attacks in hereditary angioedema. Last, intriguing recent data implicate the complement system in the pathophysiology of multiple disorders, including AMD, ischemia/reperfusion injury, organ regeneration, brain development (pruning of undesirable synapses), obesity, asthma, T-cell activation phenomena associated with allergic and rheumatic diseases,13 and more. GENERAL REFERENCES For the General References and other additional features, please visit Expert Consult at

CHAPTER 50  Complement System in Disease  

GENERAL REFERENCES 1. Ricklin D, Hajishengallis G, Yank K, et al. Complement: a key system for immune survelliance and homeostasis. Nat Immunol. 2010;11:785-797. 2. Frank MM. Complement disorders and hereditary angioedema. J Allergy Clin Immunol. 2010; 125:5262-5271. 3. Clarke EV, Tenner AJ. Complement modulation of T cell immune responses during homeostasis and disease. J Leukoc Biol. 2014;96:745-756. 4. Cozzani E, Drosera M, Gasparini G, et al. Serology of lupus erythematosus: Correlation between immunopathological features and clinical aspects. Autoimmune Dis. 2014;2014:321359. 5. Sethi S, Fervenza FC. Pathology of renal diseases associated with dysfunction of the alternative pathway of complement: C3 glomerulopathy and atypical hemolytic uremic syndrome (aHUS). Semin Thromb Hemost. 2014;40:416-421. 6. Genster N, Takahashi M, Sekine H, et al. Lessons learned from mice deficient in lectin complement pathway molecules. Mol Immunol. 2014;61:59-68.


7. Kemper C, Atkinson JP, Hourcade DE. Properdin: emerging roles of a pattern-recognition molecule. Annu Rev Immunol. 2010;28:131-155. 8. Nesargikar PN, Spiller B, Chavez R. The complement system: history, pathways, cascade and inhibitors. Eur J Microbiol Immunol (Bp). 2012;2:103-111. 9. Sethi S, Fervenza FC. Membranoproliferative glomerulonephritis—a new look at an old entity. N Engl J Med. 2012;366:1119-1131. 10. Joseph C, Gattineni J. Complement disorders and hemolytic uremic syndrome. Curr Opin Pediatr. 2013;25:209-215. 11. Seddon JM, Yu Y, Miller EC, et al. Rare variants in CFI, C3 and C9 are associated with high risk of advanced age-related macular degeneration. Nat Genet. 2013;45:1366-1370. 12. Okumura CY, Nizet V. Subterfuge and sabotage: evasion of host innate defenses by invasive grampositive bacterial pathogens. Annu Rev Microbiol. 2014;68:439-458. 13. Liszewski MK, Kolev M, Le Friec G, et al. Intracellular complement activation sustains T cell homeostasis and mediates effector differentiation. Immunity. 2013;39:1143-1157.


CHAPTER 50  Complement System in Disease  

REVIEW QUESTIONS 1. What statement is false about the alternate pathway of complement activation? A. It serves as a feedback or amplification loop for each pathway. B. It mediates lysis and cell damage in several types of hemolytic disorders. C. It is involved in debris removal and wound repair. D. It requires lectins and antibodies to be initiated. Answer: D  The alternate pathway is the original complement system. For example, it is present in insects and echinoderms and functions in hemolymph similar to its role in blood vertebrates. It is often called the “guardian of the intravascular space,” being particular designed to prevent invasion by bacteria. It is continuously “turning over” at a low rate. Antibody and lectins are more evolutionary recent means to specifically target complement activation. 2. Which statement is not true of the classical pathway? A. In conjunction with IgG and IgM, the classical pathway mediates tissue damage in autoimmune diseases. B. A complete deficiency of an early component predisposes to SLE. C. A total complement titer (CH50 or THC) measures the quantity of hemolytic activity in this pathway. D. Proteins such as CRP and serum amyloid protein (SAP) also can activate the classical pathway. E. All are true. Answer: E  It requires only a single IgM or two IgGs in close proximity to activate the classical complement pathway. IgA and IgE do not activate the classical complement pathway. 3. A deficiency of a complement inhibitor at the steps of C3 and C5 activation leads to all of the following except: A. excessive production of the C3a and C5a anaphylatoxins. B. excessive production of opsonins and the membrane attack complex. C. systemic lupus erythematosus and closely related rheumatic diseases. D. renal disease (atypical hemolytic uremic syndrome, membrano­ proliferative glomerulonephritis, C3 glomerulonephritis, and C3 glomerulopathies. Answer: C  Excessive activation at sties of injury or degeneration is a hallmark of inadequately regulated C3 and C5 convertases. The diseases for which this type of pathogenesis has been most studies are atypical hemolytic uremic syndrome and age-related macular degeneration. The other commonly observed phenotype is renal disease as described in Answer D.

4. Which statement is not true? A. CP is activated by the Fc portion of IgG or IgM engaging C1q, which is part of C1 complex. B. Complement receptor one (CR1, CD35) on RBCs serves as a taxi or ferry to take C4b and C3b bearing immune complexes to liver and spleen for disposal. C. Only a limited number of specific foreign antigens can be coated with complement fragments. D. Complement proteins in blood are synthesized by the liver, but most cell types also express complement proteins locally. Answer: C  Almost all foreign materials become coated with complement C3 fragments. Of note, a role in host defense and response to injury for local synthesis of complement components remains to be definitely established. 5. The C3a and C5a anaphylatoxins accomplish all but which one of the following? A. Engage their specific receptors to activate many cell types. B. If liberated in substantial amounts, they can lead to shock and death. C. They are responsible for opsonic activity of the complement system. D. Both are produced by all complement pathways. Answer: C  C3a and C5a are liberated in substantial amounts and this occurs in seconds. It is an early warning system to the host cell to develop a proinflammatory environment. The major opsonin of the complement system is C3b and its subsequent limited proteolytic degradation fragments. C4b is also an opsonin if the classical pathway or lectin pathway is activated.


CHAPTER 51  Approach to the Patient with Possible Cardiovascular Disease  

51  APPROACH TO THE PATIENT WITH POSSIBLE CARDIOVASCULAR DISEASE LEE GOLDMAN Patients with cardiovascular disease may present with a wide range of symptoms and signs, each of which may be caused by noncardiovascular conditions. Conversely, patients with substantial cardiovascular disease may be asymptomatic. Because cardiovascular disease is a leading cause of death in the United States and other developed countries, it is crucial that patients be evaluated carefully to detect early cardiovascular disease, that symptoms or signs of cardiovascular disease be evaluated in detail, and that appropriate therapy be instituted. Improvements in diagnosis, therapy, and prevention have contributed to a 70% or so decline in age-adjusted cardiovascular death rates in the United States since the 1960s. Furthermore, among people age 65 years and older, regular visits to a primary care physician are associated with a 25 to 30% reduction in overall mortality. However, the absolute number of deaths from cardiovascular disease in the United States has not declined proportionately because of the increase in the population older than 40 years as well as the aging of the population in general. In evaluating a patient with known or suspected heart disease, the phy­ sician must determine quickly whether a potentially life-threatening condition exists. In these situations, the evaluation must focus on the specific issue at hand and be accompanied by the rapid performance of appropriately directed additional tests. Examples of potentially life-threatening conditions include acute myocardial infarction (MI) (Chapter 73), unstable angina (Chapter 72), suspected aortic dissection (Chapter 78), pulmonary edema (Chapter 59), and pulmonary embolism (Chapter 98).


Patients may complain spontaneously of a variety of cardiovascular symptoms (Table 51-1), but sometimes these symptoms are elicited only by obtaining a careful, complete medical history. In patients with known or suspected cardiovascular disease, questions about cardiovascular symptoms are key components of the history of present illness; in other patients, these issues are a fundamental part of the review of systems.

Chest Pain Chest discomfort or pain is the cardinal manifestation of myocardial ischemia resulting from coronary artery disease or any condition that causes myocardial ischemia by an imbalance of myocardial oxygen demand compared with myocardial oxygen supply (Chapter 71). New, acute, often ongoing pain may indicate an acute MI, unstable angina, or aortic dissection; a pulmonary cause, such as acute pulmonary embolism or pleural irritation; a musculoskeletal condition of the chest wall, thorax, or shoulder; or a gastrointestinal abnormality, such as esophageal reflux or spasm, peptic ulcer disease, or cholecystitis (Table 51-2). The chest discomfort of MI commonly occurs without an immediate or obvious precipitating clinical cause and builds in intensity for at least several minutes; the sensation can range from annoying discomfort to severe pain (Chapter 73). Although a variety of adjectives may be used by patients to describe the sensation, physicians must be suspicious of any discomfort, especially if it radiates to the neck, shoulder, or arms. The probability of an acute MI can be estimated by integrating information from the history, physical examination, and electrocardiogram (Fig. 51-1). The chest discomfort of unstable angina is clinically indistinguishable from that of MI except that the former may be precipitated more clearly by activity and may be more rapidly responsive to antianginal therapy (Chapter 72). Aortic dissection (Chapter 78) classically presents with the sudden onset of severe pain in the chest and radiating to the back; the location of the pain often provides clues to the location of the dissection. Ascending aortic dissections commonly present with chest discomfort radiating to the back, whereas dissections of the descending aorta commonly present with back pain radiating to the abdomen. The presence of back pain or a history of hypertension or other predisposing factors, such as Marfan syndrome, should prompt a careful assessment of peripheral pulses to determine whether the great vessels are affected by the dissection and of the chest radiograph to

evaluate the size of the aorta. If this initial evaluation is suggestive, further testing with transesophageal echocardiography, computed tomography (CT), or magnetic resonance imaging (MRI) is indicated. The pain of pericarditis (Chapter 77) may simulate that of an acute MI, may be primarily pleuritic, or may be continuous; a key physical finding is a pericardial rub. The pain of pulmonary embolism (Chapter 98) is commonly pleuritic in nature and is associated with dyspnea; hemoptysis also may be present. Pulmonary hypertension (Chapter 68) of any cause may be associated with chest discomfort with exertion; it commonly is associated with severe dyspnea and often is associated with cyanosis. Recurrent, episodic chest discomfort may be noted with angina pectoris and with many cardiac and noncardiac causes (Chapter 71). A variety of stress tests (Table 51-3) can be used to provoke reversible myocardial ischemia in susceptible individuals and to help determine whether ischemia is the pathophysiologic explanation for the chest discomfort (Chapter 71).

Dyspnea Dyspnea, which is an uncomfortable awareness of breathing, is commonly caused by cardiovascular or pulmonary disease. A systematic approach (see Fig. 83-3) with selected tests nearly always reveals the cause. Acute dyspnea can be caused by myocardial ischemia, heart failure, severe hypertension, pericardial tamponade, pulmonary embolism, pneumothorax, upper airway obstruction, acute bronchitis or pneumonia, or some drug overdoses (e.g., salicylates). Subacute or chronic dyspnea is also a common presenting or accompanying symptom in patients with pulmonary disease (Chapter 83). Dyspnea also can be caused by severe anemia (Chapter 158) and can be confused with the fatigue that often is noted in patients with systemic and neurologic diseases (Chapters 256 and 396). In heart failure, dyspnea typically is noted as a hunger for air and a need or an urge to breathe. The feeling that breathing requires increased work or effort is more typical of airway obstruction or neuromuscular disease. A feeling of chest tightness or constriction during breathing is typical of bronchoconstriction, which is commonly caused by obstructive airway disease (Chapters 87 and 88) but also may be seen in pulmonary edema. A feeling of heavy breathing, a feeling of rapid breathing, or a need to breathe more is classically associated with deconditioning. In cardiovascular conditions, chronic dyspnea usually is caused by increases in pulmonary venous pressure as a result of left ventricular failure (Chapters 58 and 59) or valvular heart disease (Chapter 75). Orthopnea, which is an exacerbation of dyspnea when the patient is recumbent, is caused by increased work of breathing because of either increased venous return to the pulmonary vasculature or loss of gravitational assistance in diaphragmatic effort. Paroxysmal nocturnal dyspnea is severe dyspnea that awakens a patient at night and forces the assumption of a sitting or standing position to achieve gravitational redistribution of fluid.

Palpitations Palpitations (Chapter 62) describe a subjective sensation of an irregular or abnormal heartbeat. Palpitations may be caused by any arrhythmia (Chapters 64 and 65) with or without important underlying structural heart disease. Palpitations should be defined in terms of the duration and frequency of the episodes; the precipitating and related factors; and any associated symptoms of chest pain, dyspnea, lightheadedness, or syncope. It is crucial to use the history to determine whether the palpitations are caused by an irregular or a regular heartbeat. The feeling associated with a premature atrial or ventricular contraction, often described as a “skipped beat” or a “flip-flopping of the heart,” must be distinguished from the irregularly irregular rhythm of atrial fibrillation and the rapid but regular rhythm of supraventricular tachycardia. Associated symptoms of chest pain, dyspnea, lightheadedness, dizziness, or diaphoresis suggest an important effect on cardiac output and mandate further evaluation. In general, evaluation begins with ambulatory electrocardiography (ECG) (Table 51-4), which is indicated in patients who have palpitations in the presence of structural heart disease or substantial accompanying symptoms. Depending on the series, 9 to 43% of patients have important underlying heart disease. In such patients, more detailed evaluation is warranted (see Fig. 62-1). Lightheadedness or syncope (Chapter 62) can be caused by any condition that decreases cardiac output (e.g., bradyarrhythmia, tachyarrhythmia, obstruction of the left ventricular or right ventricular inflow or outflow, cardiac tamponade, aortic dissection, or severe pump failure), by reflexmediated vasomotor instability (e.g., vasovagal, situational, or carotid sinus syncope), or by orthostatic hypotension (see Table 62-1). Neurologic

CHAPTER 51  Approach to the Patient with Possible Cardiovascular Disease  

TABLE 51-1  CARDINAL SYMPTOMS OF CARDIOVASCULAR DISEASE Chest pain or discomfort Dyspnea, orthopnea, paroxysmal nocturnal dyspnea, wheezing Palpitations, dizziness, syncope Cough, hemoptysis Fatigue, weakness Pain in extremities with exertion (claudication)


diseases (e.g., migraine headaches, transient ischemic attacks, or seizures) also can cause transient loss of consciousness. The history, physical examination, and ECG are often diagnostic of the cause of syncope (see Table 62-2). Syncope caused by a cardiac arrhythmia usually occurs with little warning. Syncope with exertion or just after conclusion of exertion is typical of aortic stenosis and hypertrophic obstructive cardiomyopathy. In many patients, additional testing is required to document central nervous system disease, the cause of reduced cardiac output, or carotid sinus syncope. When the history, physical examination, and ECG do not provide helpful diagnostic information that points toward a specific cause of syncope, it is imperative that patients with heart disease or an abnormal ECG be tested with continuous ambulatory ECG monitoring to diagnose a possible arrhythmia (see Fig. 62-1); in selected patients, formal electrophysiologic testing may be indicated








Retrosternal region; radiates to or occasionally isolated to neck, jaw, epigastrium, shoulder, or arms (left common)

Pressure, burning, squeezing, heaviness, indigestion

1 flight in normal conditions

Patient can perform to completion any activity requiring ≥2 metabolic equivalents but cannot and does not perform to completion any activities requiring ≥5 metabolic equivalents, e.g., shower without stopping, strip and make bed, clean windows, walk 2.5 mph, bowl, play golf, dress without stopping


Patients with cardiac disease resulting in inability to carry on any physical activity without discomfort Symptoms of cardiac insufficiency or of the anginal syndrome may be present even at rest. If any physical activity is undertaken, discomfort is increased.

Inability to carry on any physical activity without discomfort—anginal syndrome may be present at rest

Patient cannot or does not perform to completion activities requiring ≥2 metabolic equivalents; cannot carry out activities listed above (Specific Activity Scale, class III)

From Goldman L, Hashimoto B, Cook EF, et al. Comparative reproducibility and validity of systems for assessing cardiovascular functional class: advantages of a new specific activity scale. Circulation. 1981;64:1227-1234. Reproduced by permission of the American Heart Association.

(Chapter 62). In patients with no evident heart disease, tilt testing (Chapter 62) can help detect reflex-mediated vasomotor instability.

Other Symptoms Nonproductive cough (Chapter 83), especially a persistent cough (see Fig. 83-1), can be an early manifestation of elevated pulmonary venous pressure and otherwise unsuspected heart failure. Fatigue and weakness are common accompaniments of advanced cardiac disease and reflect an inability to perform normal activities. A variety of approaches have been used to classify the severity of cardiac limitations, ranging from class I (little or no limitation)

to class IV (severe limitation) (Table 51-5). Hemoptysis (Chapter 83) is a classic presenting finding in patients with pulmonary embolism, but it is also common in patients with mitral stenosis, pulmonary edema, pulmonary infections, and malignant neoplasms (see Table 83-6). Claudication, which is pain in the extremities with exertion, should alert the physician to possible peripheral arterial disease (Chapters 79 and 80).

Complete Medical History The complete medical history should include a thorough review of systems, family history, social history, and past medical history (Chapter 15). The


CHAPTER 51  Approach to the Patient with Possible Cardiovascular Disease  

review of systems may reveal other symptoms that suggest a systemic disease as the cause of any cardiovascular problems. The family history should focus on premature atherosclerosis or evidence of familial abnormalities, such as may be found with various causes of the long QT syndrome (Chapter 65) or hypertrophic cardiomyopathy (Chapter 60). The social history should include specific questioning about cigarette smoking, alcohol intake, and use of illicit drugs. The past medical history may reveal prior conditions or medications that suggest systemic diseases, ranging from chronic obstructive pulmonary disease, which may explain a complaint of dyspnea, to hemochromatosis, which may be a cause of restrictive cardiomyopathy. A careful history to inquire about recent dental work or other procedures is crucial if bacterial endocarditis is part of the differential diagnosis.


Jugular venous distention Sternal angle


FIGURE 51-2.  Jugular venous distention is defined by engorgement of the internal jugular vein more than 5 cm above the sternal angle at 45 degrees. The central venous pressure is the observed venous distention above the sternal angle plus 5 cm.

The cardiovascular physical examination, which is a subset of the complete physical examination, provides important clues to the diagnosis of asymptomatic and symptomatic cardiac disease and may reveal cardiovascular manifestations of noncardiovascular diseases. The cardiovascular physical examination begins with careful measurement of the pulse and blood pressure (Chapter 8). If aortic dissection (Chapter 78) is a consideration, blood pressure should be measured in both arms and, preferably, in at least one leg. When coarctation of the aorta is suspected (Chapter 69), blood pressure must be measured in at least one leg and in the arms. Discrepancies in blood pressure between the two arms also can be caused by atherosclerotic disease of the great vessels. Pulsus paradoxus, which is more than the usual 10-mm Hg drop in systolic blood pressure during inspiration, is typical of pericardial tamponade (Chapter 77).

General Appearance The respiratory rate may be increased in patients with heart failure. Patients with pulmonary edema are usually markedly tachypneic and may have labored breathing. Patients with advanced heart failure may have CheyneStokes respirations. Systemic diseases, such as hyperthyroidism (Chapter 226), hypothyroidism (Chapter 226), rheumatoid arthritis (Chapter 264), scleroderma (Chapter 267), and hemochromatosis (Chapter 212), may be suspected from the patient’s general appearance. Marfan syndrome (Chapter 260), Turner syndrome (Chapter 235), Down syndrome (Chapter 41), and a variety of congenital anomalies also may be readily apparent.

FIGURE 51-3. Typical distention of the internal jugular vein. (From http://


Ophthalmologic Examination Examination of the fundi may show diabetic (see Fig. 423-24) or hypertensive retinopathy (see Fig. 67-8) or Roth spots (see Fig. 423-28) typical of infectious endocarditis. Beading of the retinal arteries is typical of severe hypercholesterolemia. Osteogenesis imperfecta, which is associated with blue sclerae, also is associated with aortic dilation and mitral valve prolapse. Retinal artery occlusion (see Fig. 423-29) may be caused by an embolus from clot in the left atrium or left ventricle, a left atrial myxoma, or atherosclerotic debris from the great vessels. Hyperthyroidism may present with exophthalmos and typical stare (see Fig. 423-6), whereas myotonic dystrophy, which is associated with atrioventricular block and arrhythmia, often is associated with ptosis and an expressionless face (see Fig. 421-2).


Phono LSB




Jugular Veins The external jugular veins help in assessment of mean right atrial pressure, which normally varies between 5 and 10 cm H2O; the height (in centimeters) of the central venous pressure is measured by adding 5 cm to the height of the observed jugular venous distention above the sternal angle of Louis (Fig. 51-2). The normal jugular venous pulse, best seen in the internal jugular vein (and not seen in the external jugular vein unless insufficiency of the jugular venous valves is present), includes an a wave, caused by right atrial contraction; a c wave, reflecting carotid artery pulsation; an x descent; a v wave, which corresponds to isovolumetric right ventricular contraction and is more marked in the presence of tricuspid insufficiency; and a y descent, which occurs as the tricuspid valve opens and ventricular filling begins (Fig. 51-3). Abnormalities of the jugular venous pressure (Fig. 51-4) are useful in detecting heart failure, and they correlate well with brain natriuretic peptide levels (Chapter 58) and echocardiographic evidence of an elevated pulmonary artery pressure (Chapter 55).1 The jugular venous pressure also helps in the diagnosis of pericardial disease, tricuspid valve disease, and pulmonary hypertension (Table 51-6).


0.1 sec

FIGURE 51-4.  Normal jugular venous pulse. ECG = electrocardiogram; JUG = jugular vein; LSB = left sternal border; phono = phonocardiogram; S1 = first heart sound; S2 = second heart sound.

Carotid Pulse The carotid pulse should be examined in terms of its volume and contour. The carotid pulse (Fig. 51-5) may be increased in frequency and may be more intense than normal in patients with a higher stroke volume secondary to aortic regurgitation, arteriovenous fistula, hyperthyroidism, fever, or anemia. In aortic regurgitation or arteriovenous fistula, the pulse may have a bisferious quality. The carotid upstroke is delayed in patients with valvular aortic

CHAPTER 51  Approach to the Patient with Possible Cardiovascular Disease  

stenosis (Chapter 75) and has a normal contour but diminished amplitude in any cause of reduced stroke volume.

Cardiac Inspection and Palpation Inspection of the precordium may reveal the hyperinflation of obstructive lung disease or unilateral asymmetry of the left side of the chest because of right ventricular hypertrophy before puberty. Palpation may be performed with the patient either supine or in the left lateral decubitus position; the latter position moves the left ventricular apex closer to the chest wall and increases the ability to palpate the point of maximal impulse and other

Positive hepatojugular reflux

Suspect heart failure, particularly left ventricular systolic dysfunction (echocardiography recommended)

Elevated systemic venous pressure without obvious x or y descent, quiet precordium, and pulsus paradoxus

Suspect cardiac tamponade (echocardiography recommended)

Elevated systemic venous pressure with sharp y descent, Kussmaul sign, and quiet precordium

Suspect constrictive pericarditis (cardiac catheterization and MRI or CT recommended)

Elevated systemic venous pressure with a sharp brief y descent, Kussmaul sign, and evidence of pulmonary hypertension and tricuspid regurgitation

Suspect restrictive cardiomyopathy (cardiac catheterization and MRI or CT recommended)

A prominent a wave with or without elevation of mean systemic venous pressure

Exclude tricuspid stenosis, right ventricular hypertrophy caused by pulmonary stenosis, and pulmonary hypertension (echo-Doppler study recommended)

A prominent v wave with a sharp y descent

Suspect tricuspid regurgitation (echo-Doppler or cardiac catheterization to determine etiology)




P A2 2




P A2 2




P A2 2

Dicrotic notch



P A2 2


P A2 2

Dicrotic notch

Dicrotic notch


The first heart sound (Fig. 51-6), which is largely produced by closure of the mitral and—to a lesser extent—the tricuspid valves, may be louder in patients with mitral valve stenosis and intact valve leaflet movement and less audible in patients with poor closure caused by mitral regurgitation (Chapter 75). The second heart sound is caused primarily by closure of the aortic valve, but closure of the pulmonic valve is also commonly audible. In normal individuals, the louder aortic closure sound occurs first, followed by pulmonic closure. With expiration, the two sounds are virtually superimposed. With inspiration, by comparison, the increased stroke volume of the right ventricle commonly leads to a discernible splitting of the second sound. This splitting may be fixed in patients with an atrial septal defect (Chapter 69) or a right bundle branch block. The split may be paradoxical in patients with left bundle branch block or other causes of delayed left ventricular emptying. The aortic component of the second sound is increased in intensity in the presence of systemic hypertension and decreased in intensity in patients with aortic stenosis. The pulmonic second sound is increased in the presence of pulmonary hypertension. Early systolic ejection sounds are related to forceful opening of the aortic or pulmonic valve. These sounds are common in congenital aortic stenosis, with a mobile valve; in hypertension, with forceful opening of the aortic valve; and in healthy young individuals, especially when cardiac output is increased. Midsystolic or late systolic clicks are caused most commonly by mitral valve prolapse (Chapter 75). Clicks are relatively high-frequency sounds that are heard best with the diaphragm of the stethoscope. An S3 corresponds to rapid ventricular filling during early diastole. It may occur in normal children and young adults, especially if stroke volume is

Dicrotic notch

Dicrotic notch


phenomena. Low-frequency phenomena, such as systolic heaves or lifts from the left ventricle (at the cardiac apex) or right ventricle (parasternal in the third or fourth intercostal space), are felt best with the heel of the palm. With the patient in the left lateral decubitus position, this technique also may allow palpation of an S3 gallop in cases of advanced heart failure or an S4 gallop in cases of poor left ventricular distensibility during diastole. The left ventricular apex is more diffuse and sometimes may be frankly dyskinetic in patients with advanced heart disease. The distal palm is best for feeling thrills, which are the tactile equivalent of cardiac murmurs. By definition, a thrill denotes a murmur of grade 4/6 or louder. Higher-frequency events may be felt best with the fingertips; examples include the opening snap of mitral stenosis or the loud pulmonic second sound of pulmonary hypertension.



CT = computed tomography; MRI = magnetic resonance imaging. From Braunwald E, ed. Heart Disease: A Textbook of Cardiovascular Medicine. 5th ed. Philadelphia: WB Saunders; 1997.



FIGURE 51-5.  Schematic diagrams of the configurational changes in the carotid pulse and their differential diagnosis. Heart sounds also are illustrated. A, Normal. B, Anacrotic pulse with slow initial upstroke. The peak is close to the second heart sound. These features suggest fixed left ventricular outflow obstruction, such as valvular aortic stenosis.  C, Pulsus bisferiens, with percussion and tidal waves occurring during systole. This type of carotid pulse contour is observed most frequently in patients with hemodynamically significant aortic regurgitation or combined aortic stenosis and regurgitation with dominant regurgitation. It rarely is observed in patients with mitral valve prolapse or in normal individuals. D, Pulsus bisferiens in hypertrophic obstructive cardiomyopathy. This finding rarely is appreciated at the bedside by palpation. E, Dicrotic pulse results from an accentuated dicrotic wave and tends to occur in sepsis, severe heart failure, hypovolemic shock, and cardiac tamponade and after aortic valve replacement. A2 = aortic component of the second heart sound; P2 = pulmonary component of the second heart sound); S1 = first heart sound; S4 = atrial sounds. (From Chatterjee K. Bedside evaluation of the heart: the physical examination. In: Chatterjee K, Chetlin MD, Karliner J, et al, eds. Cardiology: An Illustrated Text/Reference. Philadelphia: JB Lippincott; 1991:3.11-3.51.)


CHAPTER 51  Approach to the Patient with Possible Cardiovascular Disease  


Atrial or presystolic gallop (S4)

A M1 T1 Split first heart sound

B EC Aortic or pulmonary systolic ejection click (EC)

C A2 P2 Split second heart sound

D OS Opening snap of mitral stenosis (OS)

E S3 Third heart sound (S3)

supravalvular, or infravalvular aortic stenosis and pulmonic stenosis. The murmur of hypertrophic obstructive cardiomyopathy has a similar ejection quality, although its peak may be later in systole when dynamic obstruction is maximal (Chapter 60). Pansystolic murmurs are characteristic of mitral or tricuspid regurgitation or with a left-to-right shunt from conditions such as a ventricular septal defect (left ventricle to right ventricle). A late systolic murmur is characteristic of mitral valve prolapse (Chapter 75) or ischemic papillary muscle dysfunction. Ejection quality murmurs also may be heard in patients with normal valves but increased flow, such as occurs with marked anemia, fever, or bradycardia secondary to congenital complete heart block; they also may be heard across a valve that is downstream from increased flow because of an intracardiac shunt. Maneuvers such as inspiration, expiration, standing, squatting, and hand gripping can be especially useful in the differential diagnosis of a murmur; however, echocardiography commonly is required to make a definitive diagnosis of cause and severity (Table 51-8). High-frequency, early diastolic murmurs are typical of aortic regurgitation and pulmonic regurgitation from a variety of causes. The murmurs of mitral and tricuspid stenosis begin in early to mid diastole and tend to diminish in intensity later in diastole in the absence of effective atrial contraction, but they tend to increase in intensity in later diastole if effective atrial contraction is present. Continuous murmurs may be caused by any abnormality that is associated with a pressure gradient in systole and diastole. Examples include a patent ductus arteriosus, ruptured sinus of Valsalva aneurysm, arteriovenous fistula (of the coronary artery, pulmonary artery, or thoracic artery), and a mammary soufflé. In some situations, murmurs of two coexistent conditions (e.g., aortic stenosis and regurgitation, atrial septal defect with a large shunt and resulting flow murmurs of relative mitral and pulmonic stenosis) may mimic a continuous murmur. Unfortunately, the physical examination is limited for detecting meaningful valvular heart disease.2 As a result, echocardiography (Chapter 55) is critical to the evaluation of patients with suspected structural heart disease.



S2 Midsystolic click (SC)

G FIGURE 51-6. Timing of the different heart sounds and added sounds. (Modified from Wood P. Diseases of the Heart and Circulation. 3rd ed. Philadelphia: JB Lippincott; 1968.)

increased. After about 40 years of age, however, an S3 should be considered abnormal; it is caused by conditions that increase the volume of ventricular filling during early diastole (e.g., mitral regurgitation) or that increase pressure in early diastole (e.g., advanced heart failure). A left ventricular S3 gallop is heard best at the apex, whereas the right ventricular S3 gallop is heard best at the fourth intercostal space at the left parasternal border; both are heard best with the bell of the stethoscope. An S4 is heard rarely in young individuals but is common in adults older than 40 or 50 years because of reduced ventricular compliance during atrial contraction; it is a nearly ubiquitous finding in patients with hypertension, heart failure, or ischemic heart disease. The opening snap of mitral and, less commonly, tricuspid stenosis (Chapter 75) occurs at the beginning of mechanical diastole, before the onset of the rapid phase of ventricular filling. An opening snap is high pitched and is heard best with the diaphragm; this differential frequency should help distinguish an opening snap from an S3 on physical examination. An opening snap commonly can be distinguished from a loud pulmonic component of the second heart sound by the differential location (mitral opening snap at the apex, tricuspid opening snap at the left third or fourth intercostal space, pulmonic second sound at the left second intercostal space) and by the longer interval between S2 and the opening snap. Heart murmurs may be classified as systolic, diastolic, or continuous (Table 51-7). Murmurs are graded by intensity on a scale of 1 to 6. Grade 1 is faint and appreciated only by careful auscultation; grade 2, readily audible; grade 3, moderately loud; grade 4, loud and associated with a palpable thrill; grade 5, loud and audible with the stethoscope only partially placed on the chest; and grade 6, loud enough to be heard without the stethoscope on the chest. Systolic ejection murmurs usually peak in early to mid systole when left ventricular ejection is maximal; examples include fixed valvular,

The most common cause of hepatomegaly in patients with heart disease is hepatic engorgement from elevated right-sided pressures associated with right ventricular failure of any cause. Hepatojugular reflux is elicited by pressing on the liver and showing an increase in the jugular venous pressure; it indicates advanced right ventricular failure or obstruction to right ventricular filling. Evaluation of the abdomen also may reveal an enlarged liver caused by a systemic disease, such as hemochromatosis (Chapter 212) or sarcoidosis (Chapter 95), which also may affect the heart. In more severe cases, splenomegaly and ascites also may be noted. Large, palpable, polycystic kidneys (Chapter 127) commonly are associated with hypertension. A systolic bruit suggestive of renal artery stenosis (Chapter 125) or an enlarged abdominal aorta (Chapter 78) is a clue of atherosclerosis.

Extremities Extremities should be evaluated for peripheral pulses, edema, cyanosis, and clubbing. Diminished peripheral pulses suggest peripheral arterial disease (Chapters 79 and 80). Delayed pulses in the legs are consistent with coarctation of the aorta and are seen after aortic dissection. Edema (Fig. 51-7) is a cardinal manifestation of right-sided heart failure.3 When it is caused by heart failure, pericardial disease, or pulmonary hypertension, the edema is usually symmetrical and progresses upward from the ankles; each of these causes of cardiac edema commonly is associated with jugular venous distention and often with hepatic congestion. Unilateral edema suggests thrombophlebitis or proximal venous or lymphatic obstruction (Fig. 51-8). Edema in the absence of evidence of right-sided or left-sided heart failure suggests renal disease, hypoalbuminemia, myxedema, or other noncardiac causes. Among unselected patients with bilateral edema, about 40% have an underlying cardiac disease, about 40% have an elevated pulmonary blood pressure, about 20% have bilateral venous disease, about 20% have renal disease, and about 25% have idiopathic edema. Cyanosis (Fig. 51-9) is a bluish discoloration caused by reduced hemoglobin exceeding about 5 g/dL in the capillary bed. Central cyanosis is seen in patients with poor oxygen saturation resulting from a reduced inspired oxygen concentration or inability to oxygenate the blood in the lungs (e.g., as a result of advanced pulmonary disease, pulmonary edema, pulmonary arteriovenous fistula, or right-to-left shunting); it also may be seen in patients with marked erythrocytosis. Methemoglobinemia (Chapter 158) also can present with cyanosis. Peripheral cyanosis may be caused by reduced blood flow to the extremities secondary to vasoconstriction, heart failure, or shock.

CHAPTER 51  Approach to the Patient with Possible Cardiovascular Disease  




SYSTOLIC Holosystolic   Mitral regurgitation   Tricuspid regurgitation   Ventricular septal defect

Apex → axilla LLSB LLSB → RLSB

Early to mid systolic   Aortic valvular stenosis    Fixed supravalvular or subvalvular    Dynamic infravalvular

↑ with handgrip; S3 if marked mitral regurgitation; left ventricular dilation common ↑ with inspiration; right ventricular dilation common Often with thrill


  Pulmonic valvular stenosis    Infravalvular (infundibular)    Supravalvular   “Flow murmurs”


Ejection click if mobile valve; soft or absent A2 if valve immobile; later peak associated with more severe stenosis Hypertrophic obstructive cardiomyopathy; murmur louder if left ventricular volume lower or contractility increased, softer if left ventricular volume increased†; can be later in systole if obstruction delayed ↑ with inspiration ↑ with inspiration ↑ with inspiration Anemia, fever, increased flow of any cause‡

Mid to late systolic   Mitral valve prolapse   Papillary muscle dysfunction

LLSB or apex → axilla Apex → axilla

Preceded by click; murmur lengthens with maneuvers that decrease left ventricular volume† Ischemic heart disease

Early diastolic   Aortic regurgitation


High-pitched, blowing quality; endocarditis, diseases of the aorta, associated aortic valvular stenosis; signs of low peripheral vascular resistance Pulmonary hypertension as a causative factor

LLSB → apex + axilla


  Pulmonic valve regurgitation


Mid to late diastolic   Mitral stenosis, tricuspid stenosis

Apex, LLSB

  Atrial myxomas

Apex (L), LLSB (R)

Continuous   Venous hum   Patent ductus arteriosus   Arteriovenous fistula    Coronary    Pulmonary, bronchial, chest wall   Ruptured sinus of Valsalva aneurysm

Low pitched; in rheumatic heart disease, opening snap commonly precedes murmur; can be caused by increased flow across normal valve‡ “Tumor plop”

Over jugular or hepatic vein or breast LUSB LUSB Over fistula RUSB

Sudden onset

*See also Chapters 69 and 75. † Left ventricular volume is decreased by standing or during prolonged, forced expiration against a closed glottis (Valsalva maneuver); it is increased by squatting or by elevation of the legs; contractility is increased by adrenergic stimulation or in the beat after an extrasystolic beat. ‡ Including a left-to-right shunt through an atrial septal defect for tricuspid or pulmonic flow murmurs, and a ventricular septal defect for pulmonic or mitral flow murmurs. LLSB = left lower sternal border (fourth intercostal space); LUSB = left upper sternal border (second and third intercostal spaces); RLSB = right lower sternal border (fourth intercostal space); RUSB = right upper sternal border (second and third intercostal spaces).














Valsalva maneuver




Squat to stand




Stand to squat




Leg elevation









MR and VSD



Transient arterial occlusion

MR and VSD



HC = hypertrophic cardiomyopathy; MR = mitral regurgitation; RS = right sided; VSD = ventricular septal defect. Modified with permission from Lembo NJ, Dell’Italia IJ, Crawford MH, et al. Bedside diagnosis of systolic murmurs. N Engl J Med. 1988;318:1572-1578. Copyright 1988 Massachusetts Medical Society. All rights reserved.

Clubbing (Fig. 51-10), which is loss of the normal concave configuration of the nail as it emerges from the distal phalanx, is seen in patients with pulmonary abnormalities such as lung cancer (Chapter 191) and in patients with cyanotic congenital heart disease (Chapter 69).4

Examination of the Skin Examination of the skin may reveal bronze pigmentation typical of hemochromatosis (Chapter 212); jaundice (see Fig. 146-1) characteristic of severe right-sided heart failure or hemochromatosis; or capillary hemangiomas typical of Osler-Weber-Rendu disease (see Fig. 173-1), which also is associated with pulmonary arteriovenous fistulas and cyanosis. Infectious endocarditis may be associated with Osler nodes (see Fig. 76-2), Janeway lesions, or splinter hemorrhages (Fig. 51-11) (Chapter 76). Xanthomas (Fig. 51-12) are subcutaneous deposits of cholesterol seen on the extensor surfaces of the extremities or on the palms and digital creases; they are found in patients with severe hypercholesterolemia.

Laboratory Studies All patients with known or suspected cardiac disease should have an ECG and chest radiograph. The ECG (Chapter 54) helps identify rate, rhythm, conduction abnormalities, and possible myocardial ischemia. The chest radiograph (Chapter 56) yields important information on chamber enlargement, pulmonary vasculature, and the great vessels. Blood testing in patients with known or suspected cardiac disease should be targeted to the conditions in question. In general, a complete blood cell count, thyroid indices, and lipid levels are part of the standard evaluation. Point-of-care biomarker measurements in the emergency department can


CHAPTER 51  Approach to the Patient with Possible Cardiovascular Disease  

decrease unnecessary admissions and reduce median length-of-stay. For example, among patients who are being evaluated for an acute MI, an undetectable high-sensitivity troponin level at presentation reduces the probability of acute MI to less than 1%.5 A protocol in which the ECG and troponin level are repeated in 2 hours is as good as longer observation periods for

evaluating patients with acute chest pain and suspected MI. A1  However, the advent of high-sensitivity troponin assays has also greatly increased the risk for a false-positive diagnosis of MI,5 especially because of chronic troponin elevations in many cardiac conditions and in elderly patients (Chapter 72).6 Echocardiography (Chapter 55) is the most useful test to analyze valvular and ventricular function. By use of Doppler flow methods, stenotic and regurgitant lesions can be quantified. Hand-held ultrasonography performed by generalists can improve the assessment of left ventricular function, cardiomegaly, and pericardial effusion. Transesophageal echocardiography is the preferred method to evaluate possible aortic dissection and to identify clot in the cardiac chambers. Radionuclide studies (Chapter 56) can measure left ventricular function, assess myocardial ischemia, and determine whether ischemic myocardium is viable. CT can detect coronary calcium, which is a risk factor for symptomatic coronary disease (Chapter 56). In the setting



FIGURE 51-7.  Pitting edema in a patient with cardiac failure. A depression (“pit”) remains in the edema for some minutes after firm fingertip pressure is applied. (From Forbes CD, Jackson WD. Color Atlas and Text of Clinical Medicine. 3rd ed. London: Mosby; 2003.)

FIGURE 51-9.  Arterial embolism causing acute ischemia and cyanosis of the leg. Initial pallor of the leg and foot was followed by cyanosis. (From Forbes CD, Jackson WD. Color Atlas and Text of Clinical Medicine. 3rd ed. London: Mosby; 2003.)

Unilateral or bilateral? Bilateral


Detailed history Physical exam






Urine dipstick



Fever or increased WBC?

Postphlebitic syndrome?


Obvious findings of CHF?

R/O concurrent cardiac and hepatic disease







Cellulitis or other infection?

Characteristic physical signs of popliteal cyst or gastrocnemius rupture

Continue anticoagulation

R/O malignancy Detailed history Pelvic exam Rectal exam

Initiate appropriate therapy

Creatinine Electrolytes Albumin Cholesterol Prothrombin time Liver enzymes TSH Chest x-ray Cardiac echo

Antibiotic treatment



Initiate symptomatic therapy

Consider MRI

Pursue further diagnostic work-up as appropriate

Renal disease

Occult CHF


Consider renal biopsy

Initiate appropriate therapy


Other or idiopathic

Follow-up abnormalities Initiate appropriate therapy FIGURE 51-8.  Diagnostic approach to patients with edema. CHF = congestive heart failure; DVT = deep vein thrombosis; MRI = magnetic resonance imaging; R/O = rule out; TSH = thyroid-stimulating hormone; WBC = white blood cell count. (From Chertow G. Approach to the patient with edema. In: Braunwald E, Goldman L, eds. Primary Cardiology. 2nd ed. Philadelphia: WB Saunders; 2003.)

FIGURE 51-10.  Severe finger clubbing in a patient with cyanotic congenital heart disease. (From Forbes CD, Jackson WD. Color Atlas and Text of Clinical Medicine. 3rd ed. London: Mosby; 2003.)

on echocardiography.7,8 These tests are often crucial in diagnosis of possible myocardial ischemia (Chapter 71) and in establishment of prognosis in patients with known ischemic heart disease. However, they are not recommended for the screening of asymptomatic individuals9 or prior to participation in sports.10 Cardiac catheterization (Chapter 57) can measure precise gradients across stenotic cardiac valves, judge the severity of intracardiac shunts, and determine intracardiac pressures. Coronary angiography provides a definitive diagnosis of coronary disease and is a necessary prelude to coronary revascularization with a percutaneous coronary intervention or coronary artery bypass graft surgery (Chapter 74). Continuous ambulatory ECG monitoring can help diagnose arrhythmias. A variety of newer technologies allow longer-term monitoring in patients with important but infrequently occurring symptoms (Chapter 62). Formal invasive electrophysiologic testing can be useful in the diagnosis of ventricular or supraventricular wide-complex tachycardia, and it is crucial for guiding a wide array of new invasive electrophysiologic therapies (Chapter 66).


The history, physical examination, and laboratory evaluation should help the physician establish the cause of any cardiovascular problem; identify and quantify any anatomic abnormalities; determine the physiologic status of the valves, myocardium, and conduction system; determine functional capacity; estimate prognosis; and provide primary or secondary prevention. Key preventive strategies, including diet modification, recognition and treatment of hyperlipidemia, cessation of cigarette smoking, and adequate physical exercise, should be part of the approach to every patient, with or without heart disease.

Grade A References

FIGURE 51-11.  Splinter hemorrhage (solid arrow) and Janeway lesions (open arrow). These findings should stimulate a work-up for endocarditis. (Courtesy of Daniel L.  Stulberg, MD.)

A1. Than M, Aldous S, Lord SJ, et al. A 2-hour diagnostic protocol for possible cardiac chest pain in the emergency department: a randomized clinical trial. JAMA Intern Med. 2014;174:51-58. A2. Goodacre SW, Bradburn M, Cross E, et al. The randomised Assessment of Treatment using Panel Assay of Cardiac Markers (RATPAC) trial: a randomised controlled trial of point-of-care cardiac markers in the emergency department. Heart. 2011;97:190-196. A3. Litt HI, Gatsonis C, Snyder B, et al. CT angiography for safe discharge of patients with possible acute coronary syndromes. N Engl J Med. 2012;366:1393-1403. A4. Hoffmann U, Truong QA, Schoenfeld DA, et al. Coronary CT angiography versus standard evaluation in acute chest pain. N Engl J Med. 2012;367:299-308.

GENERAL REFERENCES For the General References and other additional features, please visit Expert Consult at

FIGURE 51-12.  Eruptive xanthomas of the extensor surfaces of the lower extremities. This patient had marked hypertriglyceridemia. (From Massengale WT, Nesbitt LT Jr. Xanthomas. In: Bolognia JL, Jorizzo JL, Rapini RP, eds. Dermatology. Philadelphia: Mosby; 2003:1449.)

of acute chest pain, multislice CT is effective in diagnosing coronary disease. A2  In a randomized trial of emergency department patients at low to intermediate risk for a possible acute coronary syndrome, coronary CT angiography resulted in a higher rate of discharge from the emergency department (50% vs. 23), a shorter length of stay (median, 18 vs. 24.8 hours), and a higher rate of detection of coronary disease (9% vs. 3.5%) without any change in the rate of serious adverse events. A3  However, in a subsequent randomized trial of emergency department patients with symptoms suggestive of acute coronary syndromes but without ischemic ECG changes or an initially positive troponin test, incorporating coronary CT angiography into the triage strategy did not decrease overall costs of care. A4  Stress testing by exercise or pharmacologic stress is useful to precipitate myocardial ischemia that may be detected by ECG abnormalities, perfusion abnormalities on radionuclide studies, or transient wall motion abnormalities

CHAPTER 51  Approach to the Patient with Possible Cardiovascular Disease  

GENERAL REFERENCES 1. Pellicori P, Kallvikbacka-Bennett A, Zhang J, et al. Revisiting a classical clinical sign: jugular venous ultrasound. Int J Cardiol. 2014;170:364-370. 2. Roberts KV, Brown AD, Maguire GP, et al. Utility of auscultatory screening for detecting rheumatic heart disease in high-risk children in Australia’s Northern Territory. Med J Aust. 2013;199: 196-199. 3. Clark AL, Cleland JG. Causes and treatment of oedema in patients with heart failure. Nat Rev Cardiol. 2013;10:156-170. 4. Rutherford JD. Digital clubbing. Circulation. 2013;127:1997-1999. 5. Storrow AB, Christenson RH, Nowak RM, et al. Diagnostic performance of cardiac troponin I for early rule-in and rule-out of acute myocardial infarction: Results of a prospective multicenter trial. Clin Biochem. 2014; [Epub ahead of print].


6. Korley FK, Jaffe AS. Preparing the United States for high-sensitivity cardiac troponin assays. J Am Coll Cardiol. 2013;61:1753-1758. 7. Mancini GB, Gosselin G, Chow B, et al. Canadian Cardiovascular Society guidelines for the diagnosis and management of stable ischemic heart disease. Can J Cardiol. 2014;30:837-849. 8. Mieres JH, Gulati M, Bairey Merz N, et al. Role of noninvasive testing in the clinical evaluation of women with suspected ischemic heart disease: a consensus statement from the American Heart Association. Circulation. 2014;130:350-379. 9. Chou R, Arora B, Dana T, et al. Screening asymptomatic adults with resting or exercise electrocardiography: a review of the evidence for the U.S. Preventive services task force. Ann Intern Med. 2011;155:375-385. 10. Sharma S, Estes NA 3rd, Vetter VL, et al. Clinical decisions: cardiac screening before participation in sports. N Engl J Med. 2013;369:2049-2053.

CHAPTER 52  Epidemiology of Cardiovascular Disease  


52  EPIDEMIOLOGY OF CARDIOVASCULAR DISEASE DONALD M. LLOYD-JONES Cardiovascular diseases are the leading cause of death, disability, and medical costs in the world, and they are expected to remain so for the foreseeable future. Cardiovascular disease manifests in a number of different ways, including congenital heart and vascular malformations (Chapter 69); coronary heart disease (Chapters 70, 71, 72, 73, and 74); heart failure (Chapter 59); cardiomyopathies (Chapter 60); valvular heart disease (Chapter 75); dysrhythmias (Chapters 62, 63, 64, and 65); pericardial diseases (Chapter 77); aortic (Chapter 78), peripheral (Chapter 79), and cerebrovascular diseases (Chapter 406); systemic hypertension (Chapter 67); vasculitides (Chapter 270); venous thromboembolic disease (Chapter 81); and pulmonary vascular hypertension (Chapter 68). Of these, coronary heart disease, stroke, and heart failure, which share many common underlying risk factors, have by far the largest impact on the population in terms of incidence, prevalence, quality of life, and medical costs.


CHAPTER 52  Epidemiology of Cardiovascular Disease  


Cardiovascular diseases have been the leading cause of death in the United States in every year of the 20th and 21st century except for 1918, when the influenza epidemic surpassed them. Cardiovascular diseases account for 1 in 3 deaths in America annually, or about 790,000 deaths, including about 400,000 in women and about 390,000 in men.1 The overall rate of death due to cardiovascular disease in the United States is about 230 per 100,000 persons, with higher rates in men than in women and in blacks than in whites. Because of secular trends over the past 40 to 50 years, coronary heart disease alone may soon fall below all cancers combined, but all cardiovascular diseases combined are expected to remain the leading causes of death in the United States and globally for the foreseeable future. Cardiovascular diseases also are the leading cause of hospitalizations and medical costs in the United States. Each year, about 5.8 million Americans are hospitalized for cardiovascular disease, more than 1.3 million cases of which are due to coronary heart disease and another 1 million or more due to heart failure. The United States currently spends more than $300 billion annually on direct and indirect costs for cardiovascular diseases, and these total costs are projected roughly to triple to more than $1 trillion annually by 2030. In the United States, about 15.4 million adults have coronary heart disease, roughly half of whom have had a myocardial infarction. Each year, Americans suffer more than 900,000 new and recurrent myocardial infarctions, with about 380,000 deaths due to coronary heart disease, a large percentage of which are sudden cardiac deaths. There are about 6.8 million stroke survivors in the United States, with 800,000 new or recurrent strokes occurring every year. Strokes are especially prominent in the so-called “stroke belt” in the southeastern United States, where many African Americans live. With aging, the risks for stroke and heart failure tend to increase earlier in women and African Americans than in white men, whose coronary risk increases earlier. At present, more than 5 million Americans suffer from chronic heart failure, with approximately equal numbers of men and women affected. However, the prevalence of heart failure is about twice as high in blacks as in whites.


Cardiovascular diseases, including coronary heart disease and stroke, became the leading cause of death and disability globally in the early 21st century.2 About 80% of cardiovascular deaths and events now occur in low- and middle-income countries, and the onset of cardiovascular disease tends to be at an earlier age in these countries. For example, about 50% of coronary deaths occur before age 70 years in India, whereas only 25% occur by that age in high-income countries. Unfavorable global trends in eating patterns, high rates of smoking, and increasing burdens of obesity, diabetes, and hypertension are driving the burden of cardiovascular disease.3 Whereas stroke was the dominant cause of death and disability in East Asian countries for decades owing to high sodium intake and resulting hypertension, recent changes in diet, activity levels, and smoking have made coronary heart disease an equivalent or greater health burden in this area of the world.


Established Risk Factors

A number of factors have been established for cardiovascular disease based on their strength and consistency of associations, specificity, temporality, and biologic plausibility.4,5 Furthermore, these established risk factors explain the vast majority of risk for incident myocardial infarction. Longitudinal cohort studies demonstrate that 90% of individuals who suffer a myocardial infarction have at least one established clinical risk factor before their first event, and adverse levels of nine risk factors and behaviors collectively account for 90% or more of the risk for myocardial infarction in men and women, in older and younger individuals, and in all regions of the world. These nine risk factors and behaviors include smoking (Chapter 32), elevated apolipoprotein B−to−apolipoprotein A1 ratio (Chapter 206), hypertension (Chapter 67), diabetes (Chapter 229), abdominal obesity (Chapter 220), psychosocial factors, lower consumption of fruits and vegetables (Chapter 213), alcohol intake (Chapter 33), and physical inactivity (Chapter 16). Many of the established risk factors tend to cluster in a metabolic syndrome, which is characterized by abdominal obesity, insulin resistance, hyperglycemia, elevated blood pressure, elevated triglyceride levels, and lower high-density lipoprotein (HDL) cholesterol levels. Age is the most powerful risk factor for the development of most cardiovascular diseases, especially stroke (Chapter 407), heart failure (Chapters 58 and 59), and atrial fibrillation (Chapter 64). Chronologic age represents a

person’s aggregate exposure to multiple physiologic and environmental effects on the cardiovascular system. The incidence of cardiovascular disease at least doubles with each additional decade of age in adulthood until the oldest ages, when the heavy burden of competing causes of mortality (Chapter 23) limits further progression. The impact of a person’s sex on cardiovascular disease is important. More women than men die of cardiovascular diseases annually. However, women tend to develop risk factors later in life than do men, and women’s incidence rates lag men’s by approximately 10 years. The precise contributions of sex hormones to these age trends are uncertain, but many women develop worsening risk factor levels, particularly with regard to lipids, blood pressure, weight, and insulin resistance, during and after the menopausal transition (Chapter 240). Race per se is not thought to be an independent risk factor for cardiovascular disease, and the established causal risk factors have broadly similar effects in all race and ethnic groups. Nevertheless, hypertension tends to be more prevalent in individuals of African ancestry, especially in environments with higher sodium intake, and to have a somewhat stronger association with cardiovascular events, especially heart failure and stroke. Compared with whites, individuals of East Asian and South Asian descent have a greater risk for developing the metabolic syndrome, insulin resistance, and diabetes at a lower overall body mass index. However, some of the cardiovascular risk differences observed across race and ethnic groups can be attributed to differences in socioeconomic status, rather than race or ethnicity. Blood lipid levels (Chapter 206), including the total serum cholesterol level and its subfractions, particularly low-density lipoprotein (LDL) cholesterol, have significant, continuous, and graded associations with the risk for coronary heart disease and peripheral arterial atherothrombotic disease. By comparison, independent associations of blood lipids with stroke and heart failure events are much weaker, indicating a potentially lesser role in the pathogenesis of these diseases when they occur independently of their relationship to coexisting coronary heart disease. Apolipoprotein B−containing particles make up the subpopulation of circulating cholesterol-containing particles that represent the atherogenic lipoprotein fractions. These particles are considered to be the central actors in the initiation and promotion of atherogenesis on the basis of a substantial body of epidemiologic, clinical, and basic science evidence. Among U.S. adults aged 20 years and older, 43% (or nearly 100 million) have total cholesterol levels above the desirable range of less than 200 mg/dL, and 14% (31 million) have elevated levels of 240 mg/ dL or higher. Mean total cholesterol levels have been falling sharply in recent decades, mostly because of changes in dietary composition but also because of more widespread use of lipid-lowering medications. In the 1970s, mean total cholesterol concentrations were approximately 220 mg/dL, whereas currently they are just under 200 mg/dL. These improvements have been a major contributor to the decline in coronary death rates over the same time period. Randomized clinical trials have unequivocally established LDL cholesterol as a causal agent for coronary heart disease, and statins are effective at reducing rates of both coronary heart disease and stroke, significantly and substantially. A1  By comparison, niacin is of no apparent added value A2  and other medications are being actively investigated (Chapter 206). Blood pressure (Chapter 67) has a continuous, graded association with incident coronary heart disease, stroke, and heart failure events. In worldwide studies of nearly 1 million individuals, the risk at every age for all types of cardiovascular disease death doubled with each 20-mm Hg higher systolic blood pressure and each 10-mm Hg higher diastolic blood pressure, beginning at a blood pressure of 115/75 mm Hg.6 Although the relationship with outcomes is linear, hypertension is typically defined by blood pressures of 140 mm Hg or higher systolic or 90 mm Hg diastolic (Chapter 67). Using this definition, hypertension is the most prevalent modifiable cardiovascular risk factor worldwide. Among people who are normotensive at age 55 years, the remaining lifetime risk for development of hypertension is 90%. Approximately one third of all American adults currently have hypertension, and its prevalence has been increasing owing to the obesity epidemic. Hypertension has stronger relative associations with stroke and heart failure than with coronary heart disease, in part because of its effects on myocardial and cerebrovascular remodeling. In the United States, rates of treatment and control for hypertension have been gradually increasing. The effective treatment of hypertension reduces the risk for stroke, heart failure, and coronary heart disease events. A3  Cigarette smoking (Chapter 32) is one of the strongest risk factors for cardiovascular disease events. After adjustment for other risk factors, smoking confers two- to three-fold higher risk for all manifestations of cardiovascular

CHAPTER 52  Epidemiology of Cardiovascular Disease  

disease, especially coronary heart disease and peripheral arterial disease. Fortunately, consistent public health efforts have reduced the prevalence of smoking in the United States from about 45% in the 1960s to just under 20% currently. The prevalence of smoking remains higher in many European and Asian countries, and its continued increase in some parts of the world drives unfavorable trends in cardiovascular morbidity and mortality. A large body of evidence indicates that environmental exposure to tobacco smoke in nonsmokers (“second-hand” or “passive” smoking) also increases risk for cardiovascular events substantially (Chapter 32) and contributes to the population burden of disease. Substantial data also support the benefits of smoking cessation for reducing the risks for a subsequent coronary event and death.7 Overweight and obesity have been increasing in the United States and worldwide. Before 1985, fewer than 10% of Americans were obese, defined as having a body mass index of 30 kg/m2 or higher. Now, however, about 35% of Americans are obese, and another 35% are overweight (Chapter 220). Major societal changes in the availability of food and in dietary content, coupled with reductions in physical activity, have produced this unprecedented epidemic. Although overweight and obesity themselves tend to be weak independent predictors of cardiovascular events in the short term, they are major drivers of elevated blood pressure, elevated blood glucose levels, and adverse lipid profiles that are themselves major contributors to the incidence of cardiovascular disease.8 Blood glucose and its surrogate marker, hemoglobin A1c, have a continuous and graded association with cardiovascular events. People with diabetes (Chapter 229), whether diagnosed or undiagnosed, have two- to three-fold higher adjusted risk for cardiovascular events compared with persons without diabetes, and they also have substantially higher risks for developing chronic renal disease (Chapter 130). Whereas diabetes was relatively uncommon before the 1980s, the obesity epidemic has led to a dramatic increase in the prevalence of type 2 diabetes and of impaired fasting glucose levels, termed pre-diabetes. At present in the United States, nearly 20 million people, representing more than 8% of all adults, have diagnosed diabetes, and another 8 million (about 3.5% of adults) have undiagnosed diabetes. Fully 87 million more adults, or about 38% of the adult U.S. population, currently have prediabetes. If current trends continue, an estimated 77% of men and 53% of women in the U.S. could have pre-diabetes by 2020. Diabetes affects nonwhite racial and ethnic groups, such as American Indians, African Americans, South Asians, East Asians, and Latinos, who appear to have greater sensitivity to insulin resistance at lower body mass index, in much greater proportions than whites. Unfortunately, tight control of glucose levels in persons with diabetes has not been associated with significant reductions in risk for macrovascular cardiovascular disease. A4  A5  Adverse diet (Chapter 213) is a major contributor to obesity, diabetes, hypertension, and hyperlipidemia. Healthy eating patterns emphasize a lower caloric intake and focus on fruits and vegetables, healthy fats from nuts and olive oil, lean sources of protein such as fish, whole grains, a reduced sodium intake, and limiting the intake of processed foods, unhealthy fats, and simple sugars. This eating pattern is typical of the “Mediterranean diet,” which has been shown to be associated with a lower incidence of cardiovascular disease. A6  By comparison, no vitamin or mineral supplement has been shown conclusively to reduce cardiovascular risk.9 Alcohol (Chapter 33) has a complex association with cardiovascular events. Moderate intake of one serving of alcohol per day is associated with a modestly lower risk for cardiovascular disease. At higher levels of intake, however, risks for total mortality, hypertension, stroke, and heart failure tend to increase. Physical inactivity (Chapter 16) and a sedentary lifestyle are also significant risk factors for cardiovascular disease. Individuals who participate in no physical activity are at highest risk for events. The risk is significantly lower for people who participate in even minimal physical activity, and risks decrease further with greater activity levels, particularly to the extent that they contribute to improvement in objective physical fitness. The biology and risks of sedentary time may be more than just the absence of physical activity because sedentary lifestyle, measured best by the hours of time spent in front of a television or computer screen, seems to have an adverse effect independent of time spent doing physical activity. Family history is clearly an important cardiovascular risk factor, independent of other measurable risk factors. However, ideal levels of cardiovascular health do not appear to be genetically programmed nor inexorably compromised as a consequence of aging. Data indicate that the heritability of ideal cardiovascular health is less than 20%, thereby suggesting strong environmental and behavioral influences on this trait. ,


Novel Risk Markers

Blood markers of inflammation, thrombosis, and target organ damage also appear to characterize the atherosclerotic process (Chapter 70). Serum biomarkers such as C-reactive protein, fibrinogen, plasminogen activator inhibitor-1, interleukin-6, and lipoprotein-associated phospholipase A2 have significant associations with incident cardiovascular events that are independent of established risk factors.10 However, because of their lack of specificity and their relatively weak independent associations with incident disease, none of these markers has yet proved useful for routine screening or for incorporation into risk assessment algorithms in primary or secondary prevention. To date, none has provided meaningful reclassification of risk in individuals after quantitative assessment using traditional established risk factors. Newer biomarkers that indicate the presence of existing target organ damage, such as high-sensitivity troponin or natriuretic peptide levels, hold promise for screening and targeting of prevention efforts in older, asymptomatic individuals (Chapter 23). Noninvasive cardiac testing and imaging holds the potential for detecting preclinical disease and potentially guiding early intervention. For example, electrocardiographic evidence of left ventricular hypertrophy confers significant excess risk for coronary heart disease over and above the presence of hypertension and other risk factors. High levels of coronary calcification on computed tomography (CT) imaging of the heart (Chapter 56) or greater carotid intima-media thickness measured by B-mode ultrasound of the carotid arteries portends a higher risk for future cardiovascular events. Because these imaging markers detect evidence of the actual underlying diseases of interest (i.e., left ventricular hypertrophy or atherosclerosis), rather than nonspecific risk factors, they are more effective at identifying individuals at high risk for incident clinical events, such as heart failure, stroke, and myocardial infarction. Of the available modalities, CT screening for coronary artery calcification appears to be the best widely available means for detecting individuals at near-term risk. For example, in the Multi-Ethnic Study of Atherosclerosis, asymptomatic individuals with coronary artery calcium scores of more than 100 Agatston units had relative hazards for a coronary event that were 7- to 10-fold higher than in individuals without any coronary calcification, even after adjustment for major established risk factors.11 Coronary calcium scoring also has been shown to be the most effective and reliable means for reclassifying risk after a quantitative risk assessment using established risk factors, with the ability to identify otherwise low-risk individuals who nonetheless will have a cardiovascular event. Although noninvasive screening for cardiovascular disease holds much promise for the future, its precise role remains uncertain at the present time (Chapter 56).

Assessment of Risk for Cardiovascular Disease Estimation of Short-Term Risk

Adverse levels of any single risk factor or risk marker are associated with elevated risk for incident cardiovascular events. However, combinations of adverse risk factors are additive and sometimes synergistic for increasing risk. To improve the prediction of cardiovascular events and provide quantitative risk assessment, a number of multivariable risk equations or scores, such as the Framingham equations (E-Tables 52-1 and 52-2), have been developed. The vast majority of risk scores available have focused on predicting 10-year absolute risk, and essentially all include age, sex, smoking status, cholesterol, and blood pressure, with some also including diabetes, family history, body mass index, socioeconomic status, or novel biomarkers. The end points of interest for diverse risk equations have varied widely, from the prediction of cardiovascular death alone to the prediction of fatal and nonfatal major coronary events, major atherosclerotic events (coronary disease and stroke), and a broader range of cardiovascular events (including heart failure, coronary revascularization, angina, or claudication). For example, the 10-year risk for incident atherosclerotic cardiovascular disease can be predicted in 50-yearold men and women according to sex, race, and different levels of risk factors (Fig. 52-1), and the risks are dramatically higher with a greater risk factor burden.

Lifetime Risk Estimation

Despite the widespread use of 10-year risk estimates to guide prevention strategies, this approach has important limitations. For example, one consequence of the substantial weighting of age in 10-year risk equations is that younger men and women, even those with substantial risk factor burden, do not tend to have a high short-term risk. When treatment thresholds are applied to quantitative risk estimates for clinical guidelines, men younger


CHAPTER 52  Epidemiology of Cardiovascular Disease  


AGE (yr)

HDL-C (mg/dL)








WOMEN 65 Hypertension Diabetes mellitus Congestive heart failure Prior stroke or TIA Yes

Drug dosages Dabigatran 150 mg bid, 75 mg bid if creatinine clearance < 30 Rivaroxaban 20 mg qd, 15 mg qd if creatinine clearance 15-50 Apixaban 5 mg bid, 2.5 mg bid if 2 or more (age > 80, body weight ≤ 60 kg, creatinine ≥ 1.5 mg/dL)


as the advent of reliable ablative therapies for AF may necessitate a reevaluation of this question. If a strategy of rate control is chosen, it is important to confirm a heart rate of 80 to 110 beats per minute at rest and less than 140 beats per minute with exercise, preferably by monitoring the heart rate during exercise on an exercise treadmill test or with an ambulatory monitor. More strict rate control is not beneficial. A5  Failure to confirm rate control can result in the development of tachycardia-induced cardiomyopathy. First-line therapy for rate control includes β-blockers or calcium-channel blockers; digoxin can also be used but is generally less effective. Patients commonly require a combination of mediations to achieve goal heart rates.7 If a strategy of rhythm control is chosen, many patients will first require cardioversion, either pharmacologic or electrical (Chapter 66). The risk for clot formation must be mitigated before cardioversion in all patients with AF of more than 48 hours’ duration. The first step generally is to perform transesophageal echocardiography (TEE) (Chapter 55). If TEE shows no evidence  of a left atrial clot, cardioversion can be undertaken without systemic anticoagulation; if the patient has risk factors for stroke in association with AF, however, most clinicians administer anticoagulation during the cardioversion and for the subsequent 4 weeks. If the TEE shows evidence of clot, 4 con­ secutive weeks of warfarin anticoagulation with an INR of at least 2, or equivalent anticoagulation with therapeutic doses of dabigatran, rivaroxaban, or apixaban is required; anticoagulation must be maintained for at least 3 to 4 weeks after cardioversion. Electrical cardioversion, which should be performed with a minimum of 200 joules, is successful in more than 90% of cases. Pharmacologic cardioversion can be performed with intravenous drugs such as ibutilide, which is more successful for atrial flutter (60% efficacy) than AF (50%). Oral medications can also be used as a “pill in the pocket” strategy. Patients can take a single dose of propafenone (600 mg) or flecainide (300 mg) with a conversion rate for recent-onset AF ( 48 hr duration

Warfarin, dabigatran, apixaban, or rivaroxaban for 3-4 weeks

FIGURE 64-23.  Management of recent-onset atrial fibrillation (AF). CV = cardioversion; INR = international normalized ratio; TEE = transesophageal echocardiography; TIA, transient ischemic attack.


CHAPTER 64  Cardiac Arrhythmias with Supraventricular Origin  

weeks) can also be used for cardioversion and is successful in approximately 50% of patients with both recent and more prolonged AF.

Long-Term Management

AVNRT, AVRT, Atrial Tachycardias, and Atrial Flutter

Chronic therapy for AVNRT is guided by the frequency and severity of symptoms.8 Many patients are able to live with this rhythm with infrequent recurrences, which terminate spontaneously or with adenosine. If chronic therapy is required, β-blockers or calcium-channel blockers and less commonly digoxin are used. Ablation of AVNRT (Chapter 66) is highly effective and should be considered before using sodium- or potassium-channel blocking drugs. Most patients with symptomatic AVRT are treated with catheter ablation (Chapter 66). Ablation of an accessory pathway located near the AV node or His bundle carries a 1% risk for complete heart block, whereas ablation of accessory pathways on the left side of the heart and distant from the AV node and His bundle region is not associated with a risk for heart block but carries a small risk for stroke. At present, it is not standard of care to ablate accessory pathways in patients without symptomatic arrhythmias.9,10 The long-term management of atrial tachycardia depends on symptoms. If the rhythm is highly symptomatic, it is generally managed with a β-blocker or calcium-channel blocker. If these medications are unsuccessful or not tolerated, ablation is frequently recommended, but antiarrhythmic medications are an alternative. In patients with atrial flutter, ventricular rate control is possible by achieving AV nodal block with β-blockers, calcium-channel blockers, and digitalis. However, radiofrequency ablation, which is curative, is now the preferred choice for most patients with atrial flutter (Chapter 66), especially recurrent atrial flutter. Because atrial flutter carries a 3% per year risk for thromboembolism, patients with atrial flutter should also receive long-term anticoagulation similar to what is recommended for AF (see later). If atrial flutter is successful, the risk for recurrence is very small, and long-term anticoagulation is not necessary.

Atrial Fibrillation

Therapies for the chronic maintenance of sinus rhythm in patients with AF include pharmacologic and procedural approaches. The procedural approaches include catheter-based ablation inside the left atrium with the goal of electrically isolating the pulmonary veins from the left atrium. Similarly, a minimally invasive surgical approach can electrically isolate the pulmonary veins from the external surface of the heart with the additional resection of the left atrial appendage. Both these procedures have become standard options for AF, especially in patients who have recurrent AF despite at least one antiarrhythmic drug. A6  The catheter approach carries a small risk for cardiac perforation, including pericardial tamponade and atrioesophageal fistula formation, and a 1% risk for stroke. There is also a small risk for pulmonary vein stenosis, which has been reduced by newer technologies. A second procedure typically is offered to patients who have recurrent AF following a first catheter-based procedure. In randomized trials of patients with paroxysmal AF, the cumulative burden of AF over a period of 2 years appeared to be slightly lower with initial radiofrequency catheter ablation therapy compared with antiarrhythmic medications, but at the expense of procedural risks and without any differences  in patient-reported quality of life. A7  A8  Catheter ablation may, however, be preferred as initial therapy in patients with persistent AF and symptomatic heart failure. A9  For patients with long-standing persistent AF, 5-year success rates are 20% for a single ablation procedure and 45% for multiple ablation procedures. The surgical approach carries a higher risk for cardiac bleeding, particularly during the resection of the left atrial appendage, and is associated with a significantly longer recovery time than the percutaneous approach. However, there should be no stroke risk associated with the surgical procedure because it is performed completely from the epicardial surface of the heart. A more extensive surgical operation, called the maze procedure, requires a full thoracotomy and is most often performed concomitantly as part of open coronary artery bypass surgery or an open valve operation. In this procedure, electrical lines of block are created in the left atrium to interrupt the perpetuation of AF, the pulmonary veins are isolated, and the left atrial appendage is resected. Success rates for this procedure, which should be reserved for refractory, symptomatic AF, exceed 80%. The pharmacologic options for the treatment of AF work by blocking sodium, potassium, or a combination of cardiac channels. Blockade of these channels results in slowing of cardiac conduction (sodium channels) and  prolongation in cardiac repolarization (potassium channels) as well as additional effects from modulation of the autonomic nervous system. The choice of antiarrhythmic drug is based on the patient’s underlying clinical condition (Table 64-6). Amiodarone is the most widely used medication for AF with an efficacy of 60 to 70% at 1 year. It is associated with a number of drug interactions, most notably with warfarin and digoxin. Its associated risk for thyroid, liver, and lung toxicities, related in part to the iodine moieties on this compound, necessitate ,



No structural heart disease

First line: flecainide, propafenone, dronedarone, sotalol Second line: amiodarone, dofetilide

Depressed left ventricular ejection fraction with heart failure

First line: amiodarone, dofetilide Avoid: dronedarone, flecainide, propafenone

Coronary artery disease without congestive heart failure

First line: sotalol, dronedarone, dofetilide, amiodarone Avoid: flecainide, propafenone

Hypertrophic cardiomyopathy

First line: amiodarone, sotalol, dronedarone Second line: disopyramide

TABLE 64-7  CURRENT RECOMMENDATIONS FOR THROMBOEMBOLIC PROPHYLAXIS FOR PATIENTS WITH ATRIAL FIBRILLATION BASED ON RISK FACTORS FOR STROKE RISK FACTORS* Heart failure (1 point) Hypertension (1 point) Age ≥65 (1 point), ≥75 (2 points) Diabetes (1 point) Stroke/TIA (2 points) Vascular disease (1 point) Female gender (1 point)

RECOMMENDATIONS 2 or more points: anticoagulation with warfarin or a new oral anticoagulant 1 point: anticoagulation or no therapy depending on the preference of the patient and treating physician 0 points: no therapy

*Based on the CHA2DS2-VASc risk stratification scoring system. TIA = transient ischemic attack. Data from January CT, Wann S, Alpert JS, et al. 2014 AHA/ACC/HRS Guideline for the Management of Patients with Atrial Fibrillation: A Report of the American College of Cardiology/ American Heart Association Task Force on Practice Guidelines and the Heart Rhythm Society. J Am Coll Cardiol 2014;64:e1-e76.

careful follow-up. For prevention of recurrent AF, oral amiodarone is significantly more effective than propafenone, flecainide, dofetilide, or sotalol, which are the recommended alternatives. Dronedarone (400 mg twice daily), which is related to amiodarone but has no iodine and a 24-hour half-life, is well tolerated in terms of noncardiovascular side effects but has been associated with an increased risk for heart failure, stroke, and death in patients with permanent AF. As a result, it should be discontinued in patients in whom sinus rhythm is not well maintained. Quinidine, procainamide, and disopyramide are predominantly sodiumchannel blocking drugs that also block potassium channels at slow heart rates. Each of these drugs is moderately successful in AF, with about 50% of treated patients in sinus rhythm at 1 year, but each also has idiosyncratic noncardiovascular toxicities that can significantly limit their utility (see Table 64-5). Propafenone and flecainide are also sodium-channel blockers that are widely used for the maintenance of sinus rhythm. These drugs are moderately effective, with a 50% rate of sinus rhythm at 1 year, and are generally well tolerated but must be avoided in patients with structural heart disease, particularly with a history of prior myocardial infarction and impaired left ventricular function, because of a risk for drug-induced ventricular arrhythmia. Dofetilide is a potassium-channel blocking medication that is moderately effective for suppressing AF but carries a dose-dependent risk for QT prolongation and torsades de pointes.


The presence or absence of associated conditions helps determine which patients with AF require chronic anticoagulation with warfarin or other systematic coagulants (Table 64-7). Long-term anticoagulation therapy with warfarin, dabigatran, rivaroxaban, or apixaban is generally recommended in all patients who have persistent or paroxysmal AF, who are older than 65 years, and who have no contraindications to anticoagulation.4 Anticoagulation also should be maintained for 6 months after both catheter and surgical procedures in patients without clinical risk factors for stroke and chronically in patients with risk factors. Catheter-based procedures directed at excluding the left atrial appendage from the systemic blood stream may become options in patients who have a high risk for stroke and who cannot tolerate systemic anticoagulation owing to an excessive risk for bleeding. A10  Warfarin alone is superior to aspirin or the combination of clopidogrel  and aspirin, with meta-analysis showing that adjusted-dose warfarin and 

antiplatelet agents reduce stroke by approximately 60% and 20%, respectively. A11  Although there is some protective effect at an INR as low as 1.8, the target INR for chronic anticoagulation with warfarin should be 2 to 3 to avoid INRs less than 1.8. Guidelines no longer recommend aspirin or other antiplatelet agents in patients without an indication for warfarin or the newer anticoagulants. New oral anticoagulant medications have the potential to replace warfarin as more effective and safer (except for gastrointestinal bleeding) A12  primary therapy to prevent systemic emboli in patients with AF. In a randomized trial of patients with nonvalvular atrial fibrillation, rivaroxaban (an oral factor Xa inhibitor at 20 mg per day) was better than warfarin at preventing stroke  or systemic embolization, with significantly less intracranial and fatal bleeding. A13  In another randomized trial of patients with atrial fibrillation, apixaban (an oral factor Xa inhibitor at 5 mg twice daily) prevented more strokes and systemic emboli than warfarin, with less bleeding from all causes and fewer deaths. A14  Dabigatran, a direct thrombin inhibitor (150 mg twice daily), is superior to warfarin for preventing thromboembolism, with a lower risk for intracranial bleeding but a slightly higher risk for extracranial bleeding. A15  All three drugs are eliminated by the kidney (apixaban 25%, rivaroxaban 65%, and dabigatran 85%), so they are not recommended in patients with substantial renal dysfunction, and the doses should be reduced in patients with moderate renal dysfunction (Chapter 38). A reasonable approach is to use apixaban in patients at the highest risk for bleeding, to use rivaroxaban in patients who prefer once-daily dosing, and to avoid dabigatran in patients older than 80 years because of increased bleeding risk. The addition of aspirin to moderateintensity warfarin (INR 2 to 3) or to dabigatran, rivaroxaban, or apixaban can decrease vascular events and is recommended, despite its increased risk of causing bleeding, in some AF patients with concomitant risk factors, such as coronary artery disease or a prior stroke that is attributed to vascular disease rather than to AF.

Grade A References A1. Curtis AB, Worley SJ, Adamson PB, et al. Biventricular pacing for atrioventricular block and systolic dysfunction. N Engl J Med. 2013;368:1585-1593. A2. Cappato R, Castelvecchio S, Ricci C, et al. Clinical efficacy of ivabradine in patients with inappropriate sinus tachycardia: a prospective, randomized, placebo-controlled, double-blind, crossover evaluation. J Am Coll Cardiol. 2012;60:1323-1329. A3. Al-Khatib SM, Allen LaPointe NM, Chatterjee R, et al. Rate- and rhythm-control therapies in patients with atrial fibrillation: a systematic review. Ann Intern Med. 2014;160:760-773. A4. Roy D, Talajic M, Nattel S, et al. Rhythm control versus rate control for atrial fibrillation and heart failure. N Engl J Med. 2008;358:2667-2677. A5. Van Gelder I, Groenveld H, Crijns H. Lenient versus strict rate control in patients with atrial fibrillation. N Engl J Med. 2010;362:1363-1373. A6. Wilber DJ, Pappone C, Neuzil P, et al. Comparison of antiarrhythmic drug therapy and radiofrequency catheter ablation in patients with paroxysmal atrial fibrillation: a randomized controlled trial. JAMA. 2010;303:333-340. A7. Morillo CA, Verma A, Connolly SJ, et al. Radiofrequency ablation vs antiarrhythmic drugs as first-line treatment of paroxysmal atrial fibrillation (RAAFT-2): a randomized trial. JAMA. 2014;311:692-700. A8. Cosedis Nielsen J, Johannessen A, Raatikainen P, et al. Radiofrequency ablation as initial therapy in paroxysmal atrial fibrillation. N Engl J Med. 2012;367:1587-1595. A9. Jones DG, Haldar SK, Hussain W, et al. A randomized trial to assess catheter ablation versus rate control in the management of persistent atrial fibrillation in heart failure. J Am Coll Cardiol. 2013;61:1894-1903. A10. Reddy VY, Möbius-Winkler S, Miller MA, et al. Left atrial appendage closure with the Watchman device in patients with a contraindication for oral anticoagulation: the ASAP study (ASA Plavix Feasibility Study With Watchman Left Atrial Appendage Closure Technology). J Am Coll Cardiol. 2013;61:2551-2556. A11. Hart RG, Pearce LA, Aguilar MI, et al. Meta-analysis: antithrombotic therapy to prevent stroke in patients who have nonvalvular atrial fibrillation. Ann Intern Med. 2007;146:857-867. A12. Ruff CT, Giugliano RP, Braunwald E, et al. Comparison of the efficacy and safety of new oral anticoagulants with warfarin in patients with atrial fibrillation: a meta-analysis of randomised trials. Lancet. 2014;383:955-962. A13. Patel MR, Mahaffey KW, Garg J, et al. Rivaroxaban versus warfarin in nonvalvular atrial fibrillation. N Engl J Med. 2011;365:883-891. A14. Granger CB, Alexander JH, McMurray JJ, et al. Apixaban versus warfarin in patients with atrial fibrillation. N Engl J Med. 2011;365:981-992. A15. Connolly SJ, Ezekowitz MD, Yusuf S, et al. Dabigatran versus warfarin in patients with atrial fibrillation. N Engl J Med. 2009;361:1139-1151.

GENERAL REFERENCES For the General References and other additional features, please visit Expert Consult at

CHAPTER 64  Cardiac Arrhythmias with Supraventricular Origin  

GENERAL REFERENCES 1. Brignole M, Auricchio A, Baron-Esquivias G, et al. 2013 ESC guidelines on cardiac pacing and cardiac resynchronization therapy: the task force on cardiac pacing and resynchronization therapy of the European Society of Cardiology (ESC). Developed in collaboration with the European Heart Rhythm Association (EHRA). Europace. 2013;15:1070-1118. 2. Epstein AE, DiMarco JP, Ellenbogen KA, et al. 2012 ACCF/AHA/HRS focused update incorporated into the ACCF/AHA/HRS 2008 guidelines for device-based therapy of cardiac rhythm abnormalities: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines and the Heart Rhythm Society. J Am Coll Cardiol. 2013;61: e6-e75. 3. Zimetbaum P. Antiarrhythmic drug therapy for atrial fibrillation. Circulation. 2012;125:381-389. 4. Wann LS, Curtis AB, Ellenbogen KA, et al. Management of patients with atrial fibrillation (compilation of 2006 ACCF/AHA/ESC and 2011 ACCF/AHA/HRS recommendations): a report of the American College of Cardiology/American Heart Association Task Force on practice guidelines. Circulation. 2013;127:1916-1926. 5. January CT, Wann LS, Alpert JS, et al. 2014 AHA/ACC/HRS Guideline for the Management of Patients With Atrial Fibrillation: Executive Summary: A Report of the American College of

6. 7. 8. 9.



Cardiology/American Heart Association Task Force on Practice Guidelines and the Heart Rhythm Society. J Am Coll Cardiol. 2014;64:e1-e76. Heidbuchel H, Verhamme P, Alings M, et al. EHRA practical guide on the use of new oral anticoagulants in patients with non-valvular atrial fibrillation: executive summary. Eur Heart J. 2013;34: 2094-2106. Verma A, Cairns JA, Mitchell LB, et al. 2014 focused update of the Canadian Cardiovascular Society guidelines for the management of atrial fibrillation. Can J Cardiol. 2014;30:1114-1130. Link MS. Evaluation and initial treatment of supraventricular tachycardia. N Engl J Med. 2012;367: 1438-1448. Cohen MI, Triedman JK, Cannon BC, et al. PACES/HRS expert consensus statement on the management of the asymptomatic young patient with a Wolff-Parkinson-White (WPW, ventricular preexcitation) electrocardiographic pattern: developed in partnership between the Pediatric and Congenital Electrophysiology Society (PACES) and the Heart Rhythm Society (HRS). Endorsed by the governing bodies of PACES, HRS, the American College of Cardiology Foundation (ACCF), the American Heart Association (AHA), the American Academy of Pediatrics (AAP), and the Canadian Heart Rhythm Society (CHRS). Heart Rhythm. 2012;9:1006-1024. Pappone C, Vicedomini G, Manguso F, et al. Wolff-Parkinson-White syndrome in the era of catheter ablation: insights from a registry study of 2169 patients. Circulation. 2014;130:811-819.


CHAPTER 64  Cardiac Arrhythmias with Supraventricular Origin  

REVIEW QUESTIONS 1. Supraventricular tachyarrhythmias. A 25-year-old woman presents to the emergency department with palpitations and the attached electrocardiogram. Which of the following statements is false? I













A. The rhythm can be treated with intravenous diltiazem. B. The rhythm is best treated with adenosine. C. The rhythm is curable with radiofrequency ablation. D. The rhythm is most consistent with atrial flutter. E. The rhythm is most consistent with atrioventricular (AV) nodal re-entrant tachycardia. Answer: D  The rhythm is AV nodal re-entrant tachycardia given the absence of evident P waves. This rhythm is typical in women in this age group. It can be treated acutely with adenosine or diltiazem, and long-term treatment can include a curative ablation procedure. The absence of flutter waves makes atrial flutter unlikely. 2. Supraventricular tachyarrhythmias. A 54-year-old man presents to the emergency department with the new onset of palpitations and shortness of breath. The accompanying electrocardiogram is obtained. What is the diagnosis? aVR















A. AV nodal re-entrant tachycardia B. AV re-entrant tachycardia C. Atrial flutter D. Atrial fibrillation E. Atrial tachycardia Answer: C  The electrocardiogram demonstrates negative (“sawtooth”) waves in leads 2, 3, and aVF, positive flutter waves in V1 and negative in V6. There are two flutter waves for every QRS complex, consistent with typical atrial flutter with 2 : 1 conduction.

CHAPTER 64  Cardiac Arrhythmias with Supraventricular Origin  


3. Atrioventricular conduction disturbances. An 18-year-old man presents to the hospital with fatigue for the prior week. It is late August and he has spent the summer on Nantucket as a life guard. His pulse is 40 beats per minute, and the accompanying electrocardiogram is obtained. What is the most likely diagnosis? I












A. Sinus bradycardia B. AV Wenckebach C. Complete heart block D. 2 : 1 AV block E. Junctional rhythm Answer: D  The rhythm shows 2 : 1 AV block with a long PR interval associated with the conducted P wave. The clinical history is consistent AV block secondary to Lyme disease. It should be treated with antibiotics, and in most cases there will be resolution of conduction disease. 4. Supraventricular tachycardia. An 18-year-old man presents with a syncopal episode while playing basketball. He is brought to the emergency department where the following electrocardiogram is recorded. His blood pressure is stable. What is the diagnosis? I













A. Ventricular tachycardia B. AV nodal re-entrant tachycardia with aberration C. Atrioventricular tachycardia with aberration D. Atrioventricular tachycardia without aberration E. Atrial fibrillation with conduction over an accessory pathway (bypass tract) Answer: E  The rhythm is irregularly irregular, consistent with atrial fibrillation, and the variable QRS durations are consistent with varying amounts of conduction over an accessory pathway. This classic fast, broad, and irregular pattern is consistent with atrial fibrillation conducted over an accessory pathway.


CHAPTER 64  Cardiac Arrhythmias with Supraventricular Origin  

5. Bradyarrhythmias. A 45-year-old man complains of daytime somnolence. Evaluation is notable for morbid obesity and hypertension. He takes 20 mg of lisinopril for hypertension. The following telemetry strip occurred during sleep and was not associated with symptoms. During daytime hours, his heart rate never dropped below 70 beats per minute. Which statement is true? N II 1 mV





PLETA 2 : 04 : 10




5/2/07 2 : 04 : 16



22 : 04 : 21

***BRADY 39 < 40


2 : 04 : 22

A. The telemetry strip shows complete heart block. B. The telemetry strip shows sinus node dysfunction, which requires a pacemaker. C. The telemetry strip shows sinus bradycardia progressing to sinus arrest, consistent with a vagal mechanism. D. Sinus bradycardia is due to metoprolol. E. The telemetry strip shows sinus arrhythmia. Answer: C  The telemetry strip shows sinus bradycardia progressing to a sinus pause. This finding is most consistent with a vagal mechanism, probably associated with obstructive sleep apnea. The best approach is to treat the sleep apnea. In the absence of symptoms such as syncope, a pacemaker is not indicated.

CHAPTER 65  Ventricular Arrhythmias  




Ventricular arrhythmias are cardiac rhythms that originate in the ventricular myocardium or in the His-Purkinje tissue. They include a wide spectrum of arrhythmias, from the most innocuous isolated premature ventricular contraction (PVC) to the most malignant and life-threatening ventricular arrhythmia (Fig. 65-1). Two consecutive PVCs are termed a couplet, whereas ventricular tachycardia (VT) is arbitrarily defined as three or more ventricular contractions in a row at a rate faster than 100 beats per minute. The definition of sustained VT—a continuous ventricular rhythm, at a rate faster than 100 beats per minute, with no interruption for 30 seconds or longer—is equally arbitrary. However, most if not all sustained VTs are much faster than 100 beats per minute, persist for more than 30 seconds, and cause a substantial decrease in ventricular function and cardiac output, especially in patients with underlying organic heart disease. These abrupt physiologic changes may result in acute heart failure, hypotension, syncope, or even circulatory collapse within several seconds to minutes after the onset of VT. Monomorphic VT is electrocardiographically defined as a wide-complex tachycardia with no change in QRS configuration, frontal axis, or horizontal axis from one beat to the next (see Fig. 65-1C). Monomorphic ventricular tachycardia (Fig. 65-2A) at a very rapid (>250 beats per minute) rate is sometimes called ventricular flutter, but there is no consensus for a definite rate cutoff, and it is not possible to separate the QRS clearly from the T waves when the rate exceeds 250 beats per minute. Polymorphic VT is characterized by beat-to-beat changes in the QRS morphology and axis, and very fast polymorphic VT may be difficult to distinguish from ventricular fibrillation (VF) (Fig. 65-2B). VF is a grossly irregular ventricular rhythm, usually at a rate faster than 300 beats per minute and with markedly variable low amplitude in the QRS morphology, during which there is no cardiac output. Torsades de pointes and bidirectional polymorphic VT are two distinct subtypes of polymorphic VT. To avoid confusion, the term pleomorphic VT should be used rather than the term polymorphic VT to describe the phenomenon of multiple clinical monomorphic VTs, each with distinct QRS configurations and axis observed at different times in the same patient.


The prevalence of PVCs is a function of sampling method and duration, and PVCs may be seen in 50% of apparently healthy individuals if the monitoring time is 24 hours or longer. Nonsustained VT may be recorded in up to 3% of apparently healthy individuals with no identifiable heart disease. The prevalence of PVCs and nonsustained VT increases with age, but also with the presence and severity of an underlying heart disease. Therefore, the finding of nonsustained VT often leads to a cardiac evaluation to exclude organic heart disease, even if it is incidentally discovered in an asymptomatic patient. The prevalence of nonsustained VT rises to 7 to 12% in the late phase of myocardial infarction (MI) and may be as high as 80% in patients with heart failure owing to dilated cardiomyopathy (Chapter 60). Approximately 10% of patients with documented sustained VT have no identifiable heart disease, in which case idiopathic VT is diagnosed. Idiopathic VF is exceedingly rare. Sudden cardiac death (Chapter 63) owing to ventricular arrhythmias accounts for an estimated 50% of all annual cardiovascular deaths in the United States.1 The nature of the underlying heart disease in patients dying of VT or VF is age dependent. Before 30 years of age, the organic heart disease most commonly associated with VT and VF is genetic cardiomyopathy (Chapter 60), whereas acute MI and chronic ischemic cardiomyopathy are the most common underlying heart diseases in individuals older than 40 years. In about one third of cases of sudden cardiac death without obvious underlying organic heart disease at autopsy, post-mortem genetic analysis may identify a deleterious mutation in an ion channel—a so-called channelopathy that predisposes to VT and VF.


CHAPTER 65  Ventricular Arrhythmias  



C FIGURE 65-1.  Ventricular arrhythmias. A, Multifocal premature ventricular beats. B, Nonsustained monomorphic ventricular tachycardia. Note dissociated P waves indicated by arrows. C, Sustained monomorphic ventricular tachycardia. Dissociated P waves are indicated by arrows.














B FIGURE 65-2.  A, Monomorphic ventricular tachycardia (VT) in a patient with chronic myocardial infarction. The arrows identify P waves in lead V1, showing atrioventricular dissociation. No R wave is recorded in any of the precordial leads V1 to V6 during VT. B, Polymorphic VT in a patient with chronic ischemic cardiomyopathy and marked first-degree atrioventricular block. There is no QT prolongation before the onset of the polymorphic VT.


Based on their underlying mechanisms, ventricular arrhythmias are classified as re-entrant, triggered, or automatic (Chapter 61). Re-entry, which results from activation in pathways sharing a common isthmus, is initiated by the simultaneous presence of conduction block in one limb and abnormally slow conduction in an adjacent limb, thereby allowing recovery of excitability in the former (E-Fig. 65-1A). One type of triggered activity results from early afterdepolarizations, which are oscillatory depolarizations occurring during the late phase of the action potential (E-Fig. 65-1B). Another type of triggered activity results from delayed afterdepolarizations, which are transient

depolarizations that occur immediately after the termination of the action potential and may reach activation threshold. Automatic arrhythmias arise from accelerated pacemaker activity (E-Fig. 65-1C). Sustained re-entrant activation in the myocardium, which is the most common cause of monomorphic VT, usually arises from subendocardial scarring, which is the result of prior ischemic injury and which creates an electrophysiologically abnormal substrate that results in re-entry. Other pathologic conditions capable of creating a substrate for re-entry include inflammation, granuloma (e.g., cardiac sarcoidosis), fibrofatty infiltration (e.g. arrhythmogenic right ventricular cardiomyopathy [ARVC]), genetically caused sarcomeric disarray (e.g., hypertrophic cardiomyopathy), and iatrogenic scar or

CHAPTER 65  Ventricular Arrhythmias  


LAD Base

Apex Left lateral


EAD Shorter cyle length, abnormal automaticity

0 mV



Faster phase 4 depolarization


E-FIGURE 65-1.  A, Re-entry within the myocardial infarction zone in an experimental canine model of ventricular tachycardia (VT). A central region of slow abnormal conduction, commonly referred to as the isthmus, is characterized by narrow crowded isochrones flanked by arcs of bidirectional conduction block depicted by dark lines, isolating the isthmus and enabling the maintenance of re-entry. The arrows indicate the spread of the wave of depolarization outside the central isthmus, in the shape of a figure of 8, with the red zone as the early breakthrough of activation and the dark blue area as the late activation in the VT cycle, which is also the point of re-entry into the isthmus. LAD = left anterior descending artery. B, Schematic depiction of the cardiac action potential with early afterdepolarizations (EAD) during phase 3 of a prolonged action potential (dotted lines) and delayed afterdepolarizations (DAD) reaching threshold and resulting in a premature action potential at the end of the phase 3 and the very start of the phase 4. Em = membrane potential. C, Schematic depiction of cardiac action potential with an increased slope of depolarization toward the threshold during phase 4, at a site of automatic tachycardia.

CHAPTER 65  Ventricular Arrhythmias  

patch (e.g., surgical repair of tetralogy of Fallot). These substrates may also result in polymorphic VT and VF by more than one mechanism. The mechanism of ventricular arrhythmias in Brugada syndrome is not completely understood. One proposed mechanism is based on intraventricular phase 2 re-entry owing to an exaggerated endocardial-to-epicardial gradient in membrane potential due to differences in transient outward current. Other evidence suggests abnormal conduction in the epicardium of the right ventricular outflow tract. Triggered activity, which results from adenosine-sensitive delayed afterdepolarizations rather than re-entry, is thought to be the underlying mechanism for idiopathic monomorphic VT of outflow tract origin. Idiopathic VT from re-entry in the fascicles of the left bundle branch has a relatively narrow QRS complex that always manifests right bundle branch block mimicry, most commonly with left, but rarely with right, frontal axis deviation. Torsades de pointes is caused by early afterdepolarizations that arise during an abnormally prolonged action potential owing to a delayed repolarization process in the setting of genetic long QT syndromes or acquired long QT during therapy with QT-prolonging drugs. The cause may be either diminished outflowing potassium currents or enhanced inflowing sodium or calcium currents. Although many episodes terminate spontaneously, the rates are usually very fast, and a torsade episode, if long enough, can transform into VF. Bundle branch re-entry, which results from re-entrant activation incorporating the right and the left bundle branches distally joined by the slowly conducting septal myocardium, may cause one or two nonsustained ventricular beats in a normal heart. However, sustained bundle branch re-entry occurs when myocardial disease causes chamber enlargement and bundle branch elongation and/or disease in the conduction system causes abnormal slow conduction, thereby creating the scenario for sustained bundle branch re-entry. The common type of bundle branch re-entry has anterograde activation over the right bundle and uses the left bundle retrogradely, thereby resulting in a left bunch branch block (LBBB) pattern on surface electrocardiogram (ECG), but the reverse direction with right bundle branch block (RBBB) may also occur rarely. Accelerated pacemaker activity in an ectopic location, with rates exceeding the underlying sinus rhythm rate, may arise in settings such as transient inflammation, excess digoxin levels, intracellular calcium loading, electrolyte imbalance, and coronary reperfusion following thrombotic occlusion. Bidirectional VT is thought to result from calcium overload of the myocytes owing to congenitally acquired abnormal calcium release from the ryanodine receptor or digitalis toxicity. Finally, there is no consensus regarding the mechanisms underlying VF. Theoretically, VF may be initiated when early or delayed afterdepolar­ izations fall in the vulnerable period of the action potential, thereby pre­ cipitating a re-entrant wave that breaks into sister wavelets and results in high-frequency electrical activity. In fact, VF may be regarded as an end stage for a variety of severe electrophysiologic abnormalities that result in chaotic activation.


Ventricular arrhythmias can present in a variety of clinical settings (Table 65-1). Often, ventricular arrhythmias are asymptomatic and are detected by an irregular pulse on a physical examination, on a routine ECG, on an exercise test, or on routine inpatient monitoring. In other patients, symptomatic ventricular arrhythmias can present as palpitations, dizziness, syncope (Chapters 51 and 62), shortness of breath, or sudden cardiac arrest (Chapter 63). The diagnosis usually can be confirmed on an ECG, but ambulatory monitoring (Chapter 62) is often needed because the arrhythmia may be intermittent. Ambulatory monitoring can also help correlate arrhythmias with any potentially related symptoms. In some patients, exercise testing can be helpful, especially in patients with exercise-induced symptoms. On the ECG, the QRS complex duration will typically be more than 0.12 seconds. In monomorphic VT (Fig. 65-3), the QRS complexes are the same from beat to beat, whereas polymorphic VT has multiple and changing QRS morphologies (Fig. 65-4). In VF, the ECG shows continuous irregular activation without any discrete QRS complexes (Fig. 65-5). Although underlying structural heart disease is usually present, these arrhythmias do not require a fixed structural substrate.

Acute Myocardial Infarction

VT and VF may arise as early as minutes to hours after the onset of symptoms during acute myocardial infarction (MI), and prehospital VT and VF during


TABLE 65-1  VENTRICULAR TACHYCARDIA AND CARDIAC DIAGNOSIS STRUCTURAL HEART DISEASE Acquired heart disease Acute myocardial infarction Chronic myocardial infarction, ischemic heart disease Nonischemic dilated cardiomyopathy Hypertensive heart disease Valvar heart disease Cardiac sarcoidosis Cardiac amyloidosis Other infiltrative diseases (e.g., Chagas disease) Cardiac tumors Congenital heart disease Arrhythmogenic right ventricular cardiomyopathy Hypertrophic cardiomyopathy Genetic dilated cardiomyopathies Iatrogenic Surgically repaired congenital heart disease Left ventricular assist devices NO STRUCTURAL HEART DISEASE Idiopathic ventricular tachycardia Right and left ventricular outflow tract tachycardias Left intrafascicular re-entry Papillary muscle tachycardias Idiopathic ventricular fibrillation Ion channel mutations Long QT syndromes Catecholaminergic polymorphic ventricular tachycardia Short QT syndrome Mixed etiology Brugada syndrome

acute MI are responsible for a large proportion of out-of-hospital sudden cardiac deaths (Chapter 63). The incidence of peri-infarction VF has declined over the past two decades, presumably related to the widespread practice of coronary revascularization (Chapter 74) during acute MI. Among patients with ST elevation MI who now reach the hospital, about 3 to 4% develop VT, mostly during the acute phase. The incidence of VT in patients with non-ST elevation MI (Chapter 72) is lower, about 1%. Accelerated idioventricular rhythm (AIVR) is an automatic ventricular rhythm that is faster than the sinus rate but usually less than 120 beats per minute. It may occur in the setting of acute MI and is commonly observed immediately after coronary reperfusion. AIVR rates are slower than those of the fast and malignant VT and VF of acute MI, and this arrhythmia typically terminates spontaneously without causing hemodynamic instability.


Not every wide-complex tachycardia is VT. The diagnosis is straightforward from the His bundle electrogram recorded at the time of a wide-complex tachycardia during a cardiac electrophysiology study, but diagnosis on a standard 12-lead ECG may be challenging (Table 65-2). The differential diagnosis of a sustained regular-rate wide-complex tachycardia includes any type of supraventricular tachycardia with aberrant conduction (Chapter 64), supraventricular tachycardia with ventricular preexcitation, bundle branch re-entry (which is a specific type of VT), and myocardial VT. The clinical setting and the patient’s background (e.g., history of previous MI or cardiomyopathy) play a major role in making an accurate diagnosis. New-onset wide-complex tachycardia in a young and otherwise healthy individual with no structural heart disease is most likely supraventricular tachycardia (SVT) with aberration, an SVT with preexcitation, or idiopathic VT. The most reliable observation in favor of VT is evidence of AV dissociation, that is, absence of any relationship between the atrial and ventricular rate, with the ventricular rate faster than the atrial (see Fig. 65-2A), or a regular wide-complex tachycardia with the atria fibrillating. However, the absence of atrioventricular (AV) dissociation does not exclude VT because ventriculoatrial conduction is present in about 25% of VTs. Fusion beats (which occur when an occasional sinus beat conducts through the AV node and reaches the His-Purkinje system at the same time as the VT source activates the myocardium, thereby resulting in a beat with a morphology that is the hybrid of a conducted QRS complex and the VT complex) confirm AV


CHAPTER 65  Ventricular Arrhythmias  








V5 220 ms




FIGURE 65-3.  Monomorphic ventricular tachycardia in a patient with chronic ischemic cardiomyopathy. In lead V2, the duration from the onset of the R wave to the nadir of the S wave is more than 200 msec. See text for further explanation.


V5 FIGURE 65-4. Torsades de pointes (TdP) in a patient with a markedly prolonged QT interval. A premature ventricular beat just after the peak of the T wave initiates TdP. As the tachycardia progresses, the rotation or the “twist” in the QRS axis is clearly observed in lead V1, with the polarity of the signal changing gradually from negative to positive.

V6 FIGURE 65-5. This electrocardiogram in a patient with idiopathic ventricular fibrillation (VF) shows recurrent closely coupled premature ventricular contractions (PVCs) and the initiation of VF by one of these closely coupled PVCs.

TABLE 65-2  DISTINGUISHING VENTRICULAR TACHYCARDIA FROM SUPRAVENTRICULAR TACHYCARDIA WITH ABERRANT CONDUCTION VENTRICULAR TACHYCARDIA AV dissociation aVR: initial R > S or initial R or Q > 40 msec Absence of any R wave in V1 to V6 V1 to V6: onset of R to S > 100 msec in any lead QRS duration >160 msec Initial R wave in aVR

SUPRAVENTRICULAR TACHYCARDIA Same QRS morphology as preexisting bundle branch block in sinus rhythm V1: rsR′

AV = atrioventricular.

dissociation but are observed only when VT rates are relatively slow. Other findings that favor VT include a QRS duration longer than 160 msec, or longer than 140 msec with an RBBB pattern. One approach, based on the QRS configuration on the ECG, uses the absence of RS complex in all precordial leads or an interval of more than 100 msec from the onset of R to the nadir of S wave as observations strongly favoring VT (see Figs. 65-2A and 65-3). The absence of any R waves in the QRS complexes recorded from all six precordial leads, described as negative concordance, strongly suggests VT, but unfortunately is not a common finding. Prominent R waves observed in all six precordial ECG leads, termed positive concordance, may be seen in SVT with left ventricular preexcitation but otherwise also suggests VT with a basal site of origin. In the absence of preexcitation, a slow rate of rise in the voltage

during the first 40 to 60 msec of the QRS onset suggests VT, as does the presence of initial R wave in lead aVR. A wide-complex tachycardia with a QRS morphology identical to that of aberrantly conducted beats manifesting bundle branch block (BBB) on a previously recorded ECG in the same patient should raise suspicion of bundle branch re-entry VT if AV dissociation is present. If AV dissociation is not present, the differential diagnosis includes SVT with aberrant conduction, but the rare condition of preexcitation with an atriofascicular accessory pathway should also be considered in a patient with LBBB aberration. A monomorphic wide-complex tachycardia with an irregular rate, manifested by more than 60-msec difference in cycle length from one beat to the next, is likely to be atrial fibrillation (AF) or atrial flutter, with variable AV block and aberrant conduction or with preexcitation. It is important to emphasize that electrolyte imbalances or the use of antiarrhythmic drugs diminishes the predictive accuracy of all of these diagnostic clues. Sustained polymorphic wide-complex tachycardia with marked beat-tobeat changes in the QRS morphology is always ventricular and either terminates spontaneously or transforms into VF. Torsades de pointes, a specific type of polymorphic VT, derives its name from the “twisting” or rotating of the QRS axis as the tachycardia progresses. It occurs in genetic or acquired long QT syndrome and is frequently pause dependent—typically starting when a premature beat falls on the prolonged T wave of the beat following a long RR interval (see Fig. 65-4). Finally, bidirectional VT manifesting a unique feature of beat-by-beat axis alternans may occur with digitalis toxicity or in the congenital catecholaminergic polymorphic ventricular tachycardia syndrome. Several different algorithms based on the configurations of the QRS complexes have high sensitivity, high specificity, and acceptable predictive accuracy for distinguishing epicardial VT from endocardial VT (Table 65-3).2 All

CHAPTER 65  Ventricular Arrhythmias  














FIGURE 65-6.  Monomorphic epicardial ventricular tachycardia in a patient with nonischemic dilated cardiomyopathy. The positive polarity pseudo-delta wave is prominent in the right precordial leads and the negative polarity pseudo-delta wave is prominent in the inferior limb leads.



Pseudo-delta wave

>75 msec favors epicardial site

Intrinsicoid deflection time

>85 msec favors epicardial site

Onset of R to nadir of S in precordial leads

>120 msec favors epicardial site

QRS duration

Epicardial longer

Q waves during VT in lead I

Favors epicardial site

Q waves during VT in II-III-aVF

Favors endocardial site

aVR/aVL amplitude ratio

Epicardial higher

VT = ventricular tachycardia.

are based on ECG criteria for whether the initial activation likely starts at an epicardial site. If so, the rapidly conducting His-Purkinje system is not available immediately, and the intramyocardial conduction delay produces a slurred initial component of the QRS complex, often called a pseudo-delta wave, which is manifested as a slow rate of rise of voltage before it reaches the intrinsicoid deflection (Fig. 65-6). Early recognition of ECG findings suggesting an epicardial origin of VT is important in planning and preparing a patient before a catheter ablation procedure (Chapter 66) because the epicardial approach requires a special technique in the cardiac electrophysiology laboratory. Cardiac electrophysiology testing (Chapter 62) may be indicated in patients who have organic heart disease and recurrent syncope but in whom the history, physical examination, ECG, echocardiogram, and ambulatory cardiac rhythm monitoring fail to clarify the cause, especially if the patient has a history of myocardial infarction or cardiomyopathy, either of which increases the probability that VT may be the cause of syncope. A second diagnostic indication is to identify the mechanism underlying a documented wide-complex tachycardia before the consideration of catheter ablation therapy (Chapter 66).

Identifying the Underlying Cause of Ventricular Arrhythmias

In patients with a diagnosed ventricular arrhythmia, the next step is to conduct a careful evaluation to exclude any underlying structural heart disease. This evaluation must include a comprehensive history and physical examination (Chapter 51), echocardiography (Chapter 55), and stress testing (Chapter 71). The family history may provide clues to guide genetic testing for an inherited cardiomyopathy (Chapter 60). Cardiac magnetic resonance imaging (Chapter 56) is indicated in selected patients to exclude conditions such as sarcoidosis and ARVC. Despite a comprehensive evaluation, about 10 to 15% of patients will have PVCs or VT with no identifiable structural or genetically identifiable cause. Most of the idiopathic monomorphic VTs are in one of two categories,

defined by ECG morphology. VTs that arise in the right or left ventricular outflow tract typically manifest an inferiorly directed frontal axis and are markedly positive in inferior leads (E-Fig. 65-2); the QRS configuration observed in the right precordial leads may further discriminate the sites of origin as the right or the left ventricular outflow tract or one of the sinuses of Valsalva. By comparison, idiopathic left ventricular tachycardia usually manifests RBBB mimicry and left axis deviation, but there may also be right axis deviation. The QRS complexes typically are not very wide because the involved region is His-Purkinje tissue adjacent to the interventricular septum. The differential diagnosis includes idiopathic VT arising in one of the left ventricular papillary muscles (E-Fig. 65-3). When either of these typical patterns is observed in a patient with no structural heart disease, the physician should suspect idiopathic VT. Conversely, sustained VT that does not fall into either of these two broad categories should always raise a high index of suspicion that organic heart disease may be present.

Chronic Ischemic Heart Disease and Post−Myocardial Infarction Ventricular Tachycardia

In survivors of ST elevation MI, the prevalence of sustained VT by 6 weeks is about 1%, and VT may occur as late as 15 to 20 years after the acute MI without any intervening event. VT commonly, but not invariably, reflects poor left ventricular function, especially a dyskinetic left ventricular wall segment. The electrophysiologic substrate is the surviving but electrophysiologically abnormal tissue embedded in the infarcted zone, which creates the conditions for re-entry. The areas that harbor pathways underlying re-entry can be identified by low-amplitude fractionated local electrograms recorded from the endocardium. Up to 16% of the patients have VT of epicardial origin. The same substrate may cause polymorphic VT and VF, which do not depend on a long QT interval and are different than torsades de pointes seen with repolarization abnormalities.

Nonischemic Dilated Cardiomyopathy

The most common cause of sustained monomorphic VT in nonischemic cardiomyopathy (Chapter 60) is also re-entry within the myocardium, but it differs from the post-infarction VT of chronic ischemic heart disease. The pathologic substrate, such as fibrosis, may be hard to identify. The abnormal, low-voltage, fractionated local electrograms tend to be located in basal, lateral, and often perivalvar left ventricular areas, which may correlate with the location of intramyocardial or subepicardial scarring identified by cardiac magnetic resonance imaging. The proportion of monomorphic VTs due to bundle branch re-entry is higher in nonischemic dilated cardiomyopathy compared with chronic ischemic heart disease, and VT with a focal rather than re-entrant mechanism rarely may be observed. Also, VT of nonischemic dilated cardiomyopathy is more likely to have an epicardial origin—as high as 22 to 35% in many series—and reaching 70% in Chagas disease.3 Ventricular tachycardia resulting from bundle branch re-entry also is more common in nonischemic dilated cardiomyopathy.

Heart Failure

The failing heart from any underlying cause (Chapter 58) is highly vulnerable to ventricular arrhythmias, and 40 to 60% of the deaths in patients with

CHAPTER 65  Ventricular Arrhythmias  














E-FIGURE 65-2. Electrocardiogram recorded during idiopathic ventricular tachycardia originating in the right ventricular outflow tract and manifesting a deeply inferior frontal axis and left bundle branch block mimicry in the precordial leads.















V5 E-FIGURE 65-3.  Ventricular tachycardia (VT) originating in the anterolateral papillary muscle. Note the right bundle branch mimicry of the QRS and the right axis deviation, similar to the electrocardiographic configuration of an interfascicular re-entrant VT.


CHAPTER 65  Ventricular Arrhythmias  

heart failure are sudden and commonly from VT and VF. Re-entrant VT is common especially in patients whose heart failure is due to advanced ischemic heart disease, but triggered activity resulting from derangements of calcium homeostasis may also play a prominent role. In addition, hormonal factors, electrolyte abnormalities, and changes in autonomic nervous system activity also increase the vulnerability of the failing heart to ventricular arrhythmias.

Inflammatory and Infiltrative Disease

Among patients with sarcoidosis (Chapter 95), about 40 to 50% have cardiac involvement, which may first manifest as progressive AV block and VT. Although the true prevalence of VT in sarcoidosis is not known, in the selected patients who have received implantable cardiac defibrillators (ICDs) for cardiac sarcoidosis diagnosed by endomyocardial biopsy, cardiac magnetic resonance imaging, or cardiac positron emission tomographic scans, about 15% per year have appropriate ICD discharges for sustained VT.4 Patients with other infiltrative heart diseases such as amyloidosis (Chapter 188) also have an elevated risk for VT and life-threatening ventricular arrhythmias.5

Adult Congenital Heart Disease

VT may occur in the setting of any adult congenital heart disease when there is a ventricular surgical scar or patch, as is seen after repair of tetralogy of Fallot or a ventricular septal defect closure, or a failing ventricle such as after a Mustard or Senning procedure to palliate transposition of great arteries (Chapter 69). In patients with surgically repaired tetralogy of Fallot, the prevalence of VT is about 5%, and about 2% have sudden cardiac death.

Genetically Inherited Cardiomyopathies

Hypertrophic cardiomyopathy (Chapter 60) is responsible for more than one third of sudden cardiac deaths in patients younger than age 25 years (Chapter 63), and mortality in young hypertrophic cardiomyopathy patients is almost exclusively due to VT and VF. Neither genetic testing nor a cardiac EP study can definitively identify patients at high risk for VT and VF, and the risk is determined based on findings such as a history of syncope, documented nonsustained VT especially in a young patient, a markedly thickened (>3 cm) interventricular septum, and a paradoxical decrease in blood pressure during exercise.6 ARVC is a congenital cardiomyopathy (Chapter 60), usually with an autosomal dominant inheritance. The fibrofatty infiltration of the right ventricular myocardium, which may also involve the interventricular septum and the left ventricle, results in progressive histologic change and marked electrophysiologic abnormalities, which may be manifest on the surface ECG as an epsilon wave (Fig. 65-7). The markedly altered conduction characteristics are conducive to re-entry. The incidence of VT in ARVC is related to the severity of the pathologic myocardial changes and ranges from 25 to 100%, depending on the penetrance and the expressivity of the disease. VT typically is initiated by exercise and demonstrates LBBB mimicry in the precordial ECG leads. However, unlike idiopathic right ventricular outflow tract VT, the frontal axis may be variable and not always inferiorly directed, and the site of origin may be epicardial in about 40% of cases.

Genetically Inherited “Channelopathies”

Several genetically acquired syndromes, including the long QT syndromes, Brugada syndrome, and catecholaminergic polymorphic VT increase the risk for sudden card