Principles and Practice of Pediatric Infectious Disease Revised Reprint, Third Edition

Principles and Practice of

Pediatric Infectious Diseases T H I R D E D I T I O N R e v i s e d R e p r i n t

Editor Sarah S. Long,

M.D.

Professor of Pediatrics Drexel University College of Medicine Chief, Section of Infectious Diseases St. Christopher’s Hospital for Children Philadelphia, Pennsylvania

Associate Editors Larry K. Pickering,

M.D.

Senior Advisor to the Director National Center for Immunization and Respiratory Diseases Centers for Disease Control and Prevention Professor of Pediatrics Emory University School of Medicine Atlanta, Georgia

Charles G. Prober,

M.D.

Professor of Pediatrics, Microbiology & Immunology Senior Associate Dean for Student Education Stanford University School of Medicine Stanford, California

First edition 1997, Churchill Livingstone Second edition 2003, Churchill Livingstone, an imprint of Elsevier Science ç 2008, Elsevier Inc. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior permission of the Publishers. Permissions may be sought directly from Elsevier’s Health Sciences Rights Department, 1600 John F. Kennedy Boulevard, Suite 1800, Philadelphia, PA 19103-2899, USA: phone: (+1) 215 239 3804; fax: (+1) 215 239 3805; or, e-mail: [email protected]. You may also complete your request on-line via the Elsevier homepage (http://www.elsevier.com), by selecting ‘Support and contact’ and then ‘Copyright and Permission’. ISBN: 978-0-7020-3468-8 British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging in Publication Data A catalog record for this book is available from the Library of Congress Notice Medical knowledge is constantly changing. Standard safety precautions must be followed, but as new research and clinical experience broaden our knowledge, changes in treatment and drug therapy may become necessary or appropriate. Readers are advised to check the most current product information provided by the manufacturer of each drug to be administered to verify the recommended dose, the method and duration of administration, and contraindications. It is the responsibility of the practitioner, relying on experience and knowledge of the patient, to determine dosages and the best treatment for each individual patient. Neither the Publisher nor the author assume any liability for any injury and/or damage to persons or property arising from this publication. The Publisher Material in chapters 1,3,7,12,54,56,59,60,63,139,158,159,168,169,170,178,179,213,215,216,217,218,220,230, 231,238,240,241,242,265,271,274,278,280,281 and 282 is in the public domain and may be used and reprinted without special permission; citation of the source, however, is appreciated. Cover figure. Colored transmission electron micrograph of influenza viruses budding from a host cell.

Printed in China Last digit is the print number: 9

8

7

6

5

4

3

2

1

Commissioning Editors: Karen Bowler/Thu Nguyen Development Editor: Ann Ruzycka Anderson Project Manager: Kathryn Mason Design: Gene Harris Marketing Managers (UK/USA): Clara Toombs, Paul Leese

Preface

The field of infectious diseases is ever changing—with emerging pathogens, globalization, escalating antimicrobial resistance, novel diagnostic methods, expanding therapeutic options, and continuous development of vaccines. The landscape is increasingly complex— with deepening recognition of the role of infectious agents in cancer, cardiac and gastrointestinal diseases. Our goal is to provide a comprehensive, reliable, up-to-date reference focused on evidence-based, practical information that is required to care for the neonate, infant, child, or adolescent with any infectious disease. Entries always address the four imperatives of pediatric infectious diseases: understand the problem, diagnose the etiology correctly, manage the patient to optimize outcome, and prevent a first occurrence or recurrence. The scope also includes epidemiology, control, and prevention of infectious diseases with guidance for establishing policy as well as managing individual patients. Features permeating the third edition include web-based resources, and important telephone numbers and links to primary literature to aid easy access to expanded, up-to-the-minute information as well as to obtain restricted therapeutic agents or access to experts for management of rare diseases. New tables, figures, illustrated cases, scan- and slide-ready graphics and algorithms have been added. We are very grateful to two individuals for special contributions that enhance diagnostic and educational value of the textbook. Dr. James H. Brien from Scott & White Memorial Hospital in Temple, Texas has shared multiple clinical photographs. Dr. Eric N. Faerber and the radiologic staff at St. Christopher’s Hospital for Children in Philadelphia have contributed and annotated numerous radiologic images. We have worked with authors and have edited all chapters to reflect, a prescribed, predictable and focused format that will reward the reader with answers to “What should I do next?” We have increased the number of authors from the Centers for Disease Control and Prevention, the American Academy of Pediatrics’ Red Book and Section, the American Board of Pediatrics Subboard, and the Pediatric Infectious Diseases Society to affect consistent recommendations and to build a compendium of best practices. A few examples of new content are highlighted here, within the context of the four major sections of the book. Part I. Understanding, Controlling, and Preventing Infectious Diseases: expanded primer in biostatistics; expanded use of immunoglobulin products; latest vaccines, schedule of immunizations, adverse event reporting, listings of resources in electronic, telephone, and paper media; newest recommendations in infection control for hospitals and offices; special considerations for children who are in out-of-home care, and traveling, or are immigrating.

Part II. Clinical Syndromes and Cardinal Features of Infectious Diseases: Approach to Diagnosis and Initial Management: new content on conditions that mimic infectious diseases (such as hemophagocytic lymphohistiocytosis and macrophage activation syndrome); developmental stages of innate and adaptive immunity; mechanisms, clinical manifestations and evidence for interventions for systemic inflammatory response syndrome; understanding and managing infectious risks in uniquely susceptible hosts — with cancer, solid organ or hematopoietic transplants, congenital and acquired immunodeficiency, or chronic disorders such as cystic fibrosis and sickle cell anemia; expanded content on central nervous system infectious and parainfectious conditions; new morbidities and evidence-based approaches to controlling and preventing healthcareassociated infections. Part III. Etiologic Agents of Infectious Diseases: significant new entries related to antimicrobial resistance and therapies for bacterial infections, especially due to staphylococci, enterococci, pneumococci, mycobacteria, gram-negative bacilli; recently discovered viruses, new antiviral therapies and new vaccines, such as hepatitis viruses, human metapneumovirus, influenza viruses, rotavirus, human papillomavirus, arboviruses and herpesviruses; evidence, and guidance where evidence is incomplete, for treatment of fungal infections; comprehensive management and outcomes of congenital toxoplasmosis; latest guidance for management and prevention of malaria. Part IV. Laboratory Diagnosis and Therapy of Infectious Diseases: extensively rewritten, “must haves” of best tests for laboratory identification of infectious agents; best reference for clinical use and differential diagnoses relevant to abnormal laboratory test results such as neutropenia, monocytosis, elevated C-reactive protein, procalcitonin, extreme elevation of sedimentation rate; new insights into principles of use of anti-infective agents to allow choice of best and most judicious therapies; expanded best primer on the pharmacodynamic basis of optimal use of antimicrobial agents; mechanisms and best laboratory techniques to detect antimicrobial resistance; new antimicrobial agents for bacterial, fungal, viral and parasitic infections as well as specific use of topical antimicrobial agents. The primary audience for our textbook is the subspecialist in infectious diseases who provides care for or advises on policy regarding infants, children and adolescents. We hope that our book also will serve as a daily “consultant” for pediatricians and family physicians and a valuable resource for surgeons, clinical microbiologists, infection control practitioners, health policy makers and other health professionals who care for children. Sarah S. Long Larry K. Pickering Charles G. Prober

v

D E D I C AT I O N With our spouses (Bob, Mimi, and Laura), our offspring (Stephen, Suzanne, and Caroline; Maggie and Andy; Meghan and Andrew), their loved ones, and our parents, whose patience and endurance are our inspiration We share the achievement of this book

For our mentors and colleagues, who share knowledge and stimulate learning We are profoundly grateful

To physicians and others who use the information in this book in the practice of medicine as an art based on science, and to the children they serve We dedicate this work

Acknowledgements

With special contributions by James H. Brien, D.O., Department of Pediatrics, Texas A & M University College of Medicine, Scott & White Memorial Hospital, Temple, Texas; and Eric N. Faerber, M.D., Department of Radiology, Drexel University College of Medicine, St. Christopher’s Hospital for Children, Philadelphia, Pennsylvania.

vii

Contributors

Elisabeth E. Adderson, MD Associate Professor of Pediatrics and Molecular Sciences, University of Tennessee Health Sciences Center; Associate Member, Department of Infectious Diseases, St. Jude Children’s Research Hospital, Memphis, Tennessee Infectious Complications of Antibody Deficiency

Felice C. Adler-Shohet, MD Assistant Clinical Professor of Pediatrics, University of California, Irvine, Irvine, California; Associate Director, Department of Pediatric Infectious Diseases, Miller Children’s Hospital, Long Beach, California Infection Following Trauma

Manuel R. Amieva, MD, PhD Assistant Professor of Pediatrics, Microbiology & Immunology, Stanford University School of Medicine; Attending Physician, Division of Infectious Diseases, Lucile-Packard Children’s Hospital, Stanford, California Campylobacter jejuni and Campylobacter coli; Other Campylobacter Species

Gregory L. Armstrong, MD Medical officer, Division of Global Migration and Quarantine, National Center for Preparedness, Detection, and Control of Infectious Diseases. Centers for Disease Control and Prevention, Atlanta, Georgia Hepatitis C Virus

Wences Arvelo, MD, MSc Epidemic Intelligence Officer, Enteric Diseases Epidemiology Branch, Division of Foodborne, Bacterial and Mycotic Diseases, National Center for Zoonotic, Vector-Borne, and Enteric Diseases, Centers for Disease Control and Prevention, Atlanta, Georgia Foodborne and Waterborne Disease

Ann M. Arvin, MD Lucile Packard Professor of Pediatrics and Professor of Microbiology & Immunology; Stanford University School of Medicine;

Chief, Division of Infectious Diseases, Lucile Packard Children’s Hospital, Stanford, California Varicella-Zoster Virus

David M. Asher, MD Chief, Laboratory of Bacterial, Parasitic and Unconventional Agents, Division of Emerging and Transfusion-Transmitted Diseases, Office of Blood Research and Review, Center for Biologics Evaluation and Research, United States Food and Drug Administration, Kensington, Maryland Transmissible Spongiform Encephalopathies; Slow Infections of the Nervous System

Shai Ashkenazi, MD, MSc Professor of Pediatrics, Sackler School of Medicine, Tel Aviv University, Tel Aviv, Israel; Director, Department of Pediatrics, Schneider Children’s Hospital, Petah Tikva, Israel Shigella Species

Kevin A. Ault, MD Associate Professor of Gynecology and Obstetrics Emory University School of Medicine, Atlanta, Georgia Human Papillomaviruses

Carol J. Baker, MD Professor, of Pediatrics, Molecular Virology and Microbiology, Baylor College of Medicine; Attending Physician, Infectious Diseases, Texas Children’s Hospital; Medical Director of Infection Control, Woman’s Hospital of Texas, Houston, Texas Bacterial Infections in the Neonate; Streptococcus agalactiae (Group B Streptococcus)

William J. Barson, MD Professor of Clinical Pediatrics, The Ohio State University College of Medicine and Public Health; Attending Physician, Section of Infectious Diseases, Children’s Hospital, Columbus, Ohio Klebsiella and Raoultella Species; Proteus, Providencia, and Morganella Species

Beth P. Bell, MD, MPH Chief, Epidemiology Branch, Division of Viral Hepatitis, National Center for HIV/AIDS, Viral Hepatitis, STD, and TB Prevention, Centers for Disease Control and Prevention, Atlanta, Georgia Hepatitis A Virus

Michael J. Bell, MD Associate Director for Infection Control, Division of Healthcare Quality Promotion, National Center for Infectious Diseases, Centers for Disease Control and Prevention, Atlanta, Georgia Ebola and Marburg Hemorrhagic Fever Viruses; New-World Arenaviruses, Lassa Virus, and Lymphocytic Choriomeningitis Virus

Daniel K. Benjamin, Jr, MD, MPH, PhD Associate Professor of Pediatrics, Duke University Medical Center; Associate Professor, Duke Clinical Research Institute, Duke University, Durham, North Carolina Necrotizing Enterocolitis; Clinical Approach to the Infected Neonate

Stephanie R. Bialek, MD, MPH Medical Epidemiologist, Epidemiology Branch, Division of Viral Hepatitis, National Center for HIV/AIDS, Viral Hepatitis, STD, and TB Prevention, Centers for Disease Control and Prevention, Atlanta, Georgia Hepatitis C Virus

Margaret J. Blythe, MD Professor of Pediatrics, Indiana University School of Medicine; Director, Adolescent Clinical Services, James Whitcomb Riley Hospital for Children; Director, Adolescent Clinical Services, Wishard Memorial Hospital, Indianapolis, Indiana Sexually Transmitted Infection Syndromes

Joseph A. Bocchini, Jr, MD Professor and Chairman, Department of Pediatrics, Louisiana State University Health Science Center–Shreveport; Medical Director, Children’s Hospital, Shreveport, Louisiana Human Papillomaviruses ix

x

Contributor

Michael Boeckh, MD

Joseph S. Bresee, MD

Jane L. Burns, MD

Associate Professor of Medicine, University of Washington, School of Medicine; Associate Member, Division of Infectious Diseases, Fred Hutchinson Cancer Research Center, Seattle, Washington Human Polyomaviruses

Chief, Epidemiology and Prevention Branch, Influenza Division, National Center for Immunizations and Respiratory Diseases, Centers for Disease Control and Prevention; Assistant Professor, Department of Family and Preventive Medicine, Emory University School of Medicine, Atlanta, Georgia Viral Gastroenteritis; Rotaviruses; Astroviruses

Professor of Pediatrics, University of Washington School of Medicine; Attending Physician Division of Infectious Diseases, Immunology and Rheumatology, Children’s Hospital and Regional Medical Center, Seattle, Washington Infectious Complications in Special Hosts; Pseudomonas Species and Related Organisms; Burkholderia cepacia Complex and Other Burkholderia Species

William A. Bower, MD Medical Officer, Epidemiology Surveillance and Response Branch, Division of Bioterrorism Preparedness and Response, National Center for Preparedness, Detection and Control of Infectious Diseases, Centers for Disease Control and Prevention, Atlanta, Georgia Hepatitis B and Hepatitis D Viruses; Hepatitis E Virus and Other Newly Identified Viruses

Itzhak Brook, MD, MSc

Woman’s Board Professor of Pediatrics, Chairman, Department of Pediatrics, Rush Medical College, Rush University Medical Center, Chicago, Illinois Toxoplasma gondii (Toxoplasmosis)

Professor of Pediatrics, Georgetown University; Attending Physician, Department of Pediatrics, Georgetown University Hospital, Washington, DC Anaerobic Bacteria; Classification, Normal Flora, and Clinical Concepts; Clostridium tetani (Tetanus); Clostridium difficile; Other Clostridium Species; Bacteroides and Prevotella Species and Other Anaerobic Gram-Negative Bacilli; Fusobacterium Species; Anaerobic Cocci; Anaerobic GramPositive Nonsporulating Bacilli (Including Actinomycosis)

Christopher R. Braden, MD

Kevin E. Brown, MD

Kenneth M. Boyer, MD

Associate Director for Science, Division of Parasitic Diseases, National Center for Zoonotic, Vector-Borne, and Enteric Diseases, Centers for Disease Control and Prevention, Atlanta, Georgia Foodborne and Waterborne Disease

John S. Bradley, MD Director, Division of Infectious Diseases, Children’s Hospital San Diego, San Diego, California Abdominal and Retroperitoneal Lymphadenopathy; Principles of AntiInfective Therapy; Antimicrobial Agents

Michael T. Brady, MD Professor and Interim Chair, Department of Pediatrics, The Ohio State University College of Medicine,; Physician-In-Chief, Columbus Children’s Hospital, Columbus, Ohio Less Commonly Encountered Nonenteric Gram-Negative Bacilli; Eikenella, Pasteurella, and Chromobacterium Species

Denise Bratcher, DO Associate Professor of Pediatrics, University of Missouri–Kansas City School of Medicine; Attending Physician, Section of Infectious Diseases, Children’s Mercy Hospitals and Clinics, Kansas City, Missouri Archanobacterium haemolyticum; Bacillus Species (Anthrax); Other Corynebacteria; Other Gram-Positive Bacilli

Paula K. Braverman, MD Professor of Clinical Pediatrics, University of Cincinnati College of Medicine; Attending Physician Division of Adolescent Medicine, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio Urethritis, Vulvovaginitis, and Cervicitis

Consultant Virologist, Virus Reference Department, Centre for Infections, Health Protection Agency, London, England; Hematology Branch, National Heart, Lung and Blood Institute, National Institutes of Health, Bethesda, Maryland Human Parvoviruses

John C. Browning, MD Resident, Department of Dermatology, Baylor College of Medicine; Department of Dermatology, Baylor College of Medicine Affiliated Hospitals, Houston, Texas Cellulitis and Superficial Skin Infections; Erythematous Macules and Papules; Vesicles and Bullae; Purpura; Urticaria and Erythema Multiforme; Papules, Nodules and Ulcers; Subcutaneous Tissue Infections and Abscesses; Dermatophytes and Other Superficial Fungi

Steven C. Buckingham, MD Associate Professor of Pediatrics, University of Tennessee Health Science Center; Attending Physician, Department of Pediatrics, Le Bonheur Children’s Medical Center, Memphis, Tennessee Other Haemophilus Species

E. Stephen Buescher, MD Professor of Pediatrics, Eastern Virginia Medical School; Attending Physician, Division of Infectious Diseases, Children’s Hospital of The King’s Daughters, Norfolk, Virginia Nosocomial Infections in the Neonate; Evaluation of the Child with Suspected Immunodeficiency; Infectious Complications of Dysfunction or Deficiency of Polymorphonuclear and Mononuclear Phagocytes

Michael Cappello, MD Professor of Pediatrics, Microbial Pathogenesis, and Epidemiology and Public Health, Yale University School of Medicine; Director, Yale Program in International Child Health, New Haven, Connecticut Nematodes; Taenia solium and Taenia saginata (Taeniasis and Cysticercosis); Taenia (Multiceps) multiceps and Taenia serialis (Coenurosis)

Bryan D. Carter, PhD Professor of Child and Adolescent Psychiatry, Associate Professor of Pediatrics, University of Louisville School of Medicine; Director, Pediatric Consultation–Liaison Service, Kosair Children’s Hospital, Louisville, Kentucky Chronic Fatigue Syndrome

Ellen Gould Chadwick, MD Professor, Associate Chair for Education, Department of Pediatrics, Feinberg School of Medicine, Northwestern University; Associate Director, Section of Pediatric and Maternal HIV Infection, Division of Infectious Diseases, Children’s Memorial Hospital, Chicago, Illinois Nocardia Species

Patricia Joan Chesney, MD, CM Professor of Pediatrics, University of Tennessee Health Sciences Center; Staff Physician, Pediatric Infectious Disease Section, Le Bonheur Children’s Medical Center; Director of Academic Programs, Department of Pediatric Infectious Disease, St. Jude Children’s Research Hospital, Memphis, Tennessee Lymphatic System and Generalized Lymphadenopathy; Cervical Lymphadenitis and Neck Infections; Mediastinal and Hilar Lymphadenopathy

James E. Childs, ScD Senior Research Scientist, Department of Epidemiology and Public Health, Yale University School of Medicine, New Haven, Connecticut Ehrlichia and Anaplasma Species; Rickettsia rickettsii (Rocky Mountain Spotted Fever)

John C. Christenson, MD Professor of Clinical Pediatrics, Indiana University School of Medicine; Director,

Contributor

Pediatric Travel Medicine Clinic, Center for International Adoption and Geographic Medicine, Ryan White Center for Infectious Diseases, Riley Hospital for Children, Indianapolis, Indiana Laboratory Diagnosis of Infection Due to Bacteria, Fungi, Parasites, and Rickettsiae

Thomas G. Cleary, MD Professor of Pediatrics, Director, Pediatric Infectious Diseases Division, Department of Pediatrics, The University of Texas Health Science Center at Houston, Houston, Texas; Department of Pediatrics, Memorial Hermann Children’s Hospital, Houston, Texas Plesiomonas shigelloides; Shigella Species; Yersinia Species; Aeromonas Species

Susan E. Coffin, MD, MPH Associate Professor of Pediatrics, University of Pennsylvania School of Medicine; Hospital Epidemiologist and Medical Director, Infection Prevention and Control; Attending Physician, Division of Infectious Diseases, Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania Healthcare-Associated Infections; Clinical Syndromes of Device-Associated Infections

Beverly L. Connelly, MD Professor of Pediatrics, University of Cincinnati College of Medicine; Director, Infection Control Program; Director, Infectious Diseases Training Program, Division of Infectious Diseases, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio Granulomatous Hepatitis; Acute Pancreatitis; Cholecystitis and Cholangitis

C. Michael Cotton, MD Assistant Clinical Professor, Department of Pediatrics, Duke University School of Medicine; Director, Neonatology Clinical Research, Duke University Medical Center, Durham, North Carolina Necrotizing Enterocolitis

Elaine Cox, MD Assistant Clinical Professor of Pediatrics, Indiana University School of Medicine; Attending Physician, Section of Infectious Diseases, James Whitcomb Riley Hospital for Children, Indianapolis, Indiana Agents of Eumycotic Mycetoma; Pseudallescheria boydii (anamorph Scedosporium apiospermum)

Robert Andrew Cramer, Jr, PhD Assistant Professor Fungal Pathogenesis, Veterinary Molecular Biology, Montana State University, Bozeman, Montana; formerly Molecular Mycology and Pathogenesis Post-Doctoral Fellow, Department of Molecular Genetics and

Molecular Biology, Duke University Medical Center, Durham, North Carolina Aspergillus Species

Maryanne E. Crockett, MD, MPH, FRCPC, DTM&H

xi

Infectious Diseases; Director of Diagnostic Virology Laboratory, Texas Children’s Hospital, Houston, Texas Adenoviruses

Dickson D. Despommier, PhD

Faculty of Medicine, University of Toronto; Research Fellow, Division of Infectious Diseases, Department of Paediatrics, The Hospital for Sick Children, Toronto, Ontario, Canada Protection of Travelers

Professor of Public Health and Microbiology, Departments of Environmental Health Sciences and Microbiology, The Mailman School of Public Health, Columbia University, New York, New York Tissue Nematodes

James E. Crowe, Jr, MD

Karen A. Diefenbach, MD

Ingram Professor Department of Pediatrics, Microbiology and Immunology, and the Vanderbilt Vaccine Center, Vanderbilt University Medical Center, Nashville, Tennessee Respiratory Syncytial Virus; Human Metapneumovirus

John Seashore Pediatric Surgery Research Fellow, Department of Surgery, Yale University School of Medicine, New Haven, Connecticut Intra-Abdominal, Visceral, and Retroperitoneal Abscesses

Elidia Dominguez, MD Dennis J. Cunningham, MD Assistant Professor of Pediatrics, The Ohio State University College of Medicine; Physician Director, Epidemiology and Infection Control; Attending Physician, Section of Infectious Diseases, Columbus Children’s Hospital, Columbus, Ohio Enterobacter and Pantoea Species

Coordinator, Clinical Research Unit, Institute of Advanced Scientific Investigations & High Technology Services (INDICASAT); National Council for Science, Technology and Innovation, Panama City, Panama Babesia Species (Babesiosis); Balantidium coli; Sarcocystis Species

Toni Darville, MD

Stephen M. Downs, MD, MS

Professor of Pediatrics, Microbiology & Immunology, University of Arkansas for Medical Sciences; Attending Physician, Division of Infectious Diseases, Arkansas Children’s Hospital, Little Rock, Arkansas Chlamydia trachomatis

Jean and Jerry Bepko Scholar of Pediatrics, Indiana University School of Medicine; Director of General and Community Pediatrics and Children’s Health Services Research, James Whitcomb Riley Hospital for Children, Indianapolis, Indiana Urinary Tract Infections

Gregory A. Dasch, PhD Rickettsial Team Leader, Rickettsial Zoonoses Branch, Division of Viral and Rickettsial Diseases, Centers for Disease Control and Prevention, Atlanta, Georgia Other Rickettsia Species

Robert S. Daum, MD Professor of Pediatrics, Microbiology and Molecular Medicine, University of Chicago; Attending Physician, Department of Pediatrics, University of Chicago Comer Children’s Hospital, Chicago, Illinois Staphylococcus aureus

Maite de la Morena, MD Associate Professor of Internal Medicine and Pediatrics, University of Texas Southwestern Medical Center at Dallas; Attending Physician, Division of Allergy and Immunology, Children’s Medical Center of Dallas, Dallas, Texas Immunologic Development and Susceptibility to Infection

Gail J. Demmler, MD Professor of Pediatrics, Baylor College of Medicine; Attending Physician, Division of

Christopher C. Dvorak, MD Clinical Instructor of Pediatrics, Stanford University, Palo Alto, California; Division Pediatric Stem Cell Transplantation, Lucile Packard Children’s Hospital, Stanford, California Antifungal Agents

Kathryn Edwards, MD Professor and Vice Chair Pediatric Clinical Research Office, Department of Pediatrics, Vanderbilt University; Attending Physician, Division of Infections Diseases, Vanderbilt Children’s Hospital, Nashville, Tennessee Prolonged, Recurrent, and Periodic Fever Syndromes; Bordetella pertussis (Pertussis) and Other Bordella Species

Morven S. Edwards, MD Professor of Pediatrics, Baylor College of Medicine; Attending Physician, Division of Infectious Diseases, Texas Children’s Hospital, Houston, Texas Bacterial Infections in the Neonate; Streptococcus agalactiae (Group B Streptococcus)

xii

Contributor

Janet A. Englund, MD

Patricia M. Flynn, MD, MS

Peter Gilligan, PhD

Associate Professor of Pediatrics, University of Washington, School of Medicine; Clinical Associate, Division of Clinical Research, Fred Hutchinson Cancer Research Center, Seattle, Washington Infectious Complications in Special Hosts

Professor of Pediatrics and Preventive Medicine, University of Tennessee Health Science Center; Arthur Ashe Chair of Pediatric AIDS Research; Attending Physician, Section of Infectious Diseases, St. Jude Children’s Research Hospital, Memphis, Tennessee Cryptosporidium Species; Isospora and Cyclospora Species; Microsporidia

Professor of Microbiology-Immunology and Pathology-Laboratory Medicine, University of North Carolina School of Medicine; Director, Clinical Microbiology-Immunology Laboratories, University of North Carolina Hospitals, Chapel Hill, North Carolina Mechanisms and Detection of Antibiotic Resistance

Véronique Erard Staff Scientist, Division of Infectious Diseases, Fred Hutchinson Cancer Research Center, Seattle, Washington Human Polyomaviruses

Marina E. Eremeeva, MD, PhD, ScD Senior Service Fellow, Rickettsial Zoonoses Branch, Division of Viral and Rickettsial Diseases, National Center for Zoonotic, Vector-Borne and Enteric Diseases, Centers for Disease Control and Prevention, Atlanta, Georgia Other Rickettsia Species

Lyn Finelli, DrPH, MS Epidemiologist, Epidemiology and Prevention Branch, Influenza Division, National Center for Immunization and Respiratory Diseases, Centers for Disease Control and Prevention, Atlanta, Georgia Hepatitis B and Hepatitis D Viruses

Adam Finn, BM, BCh, MA, PhD, FRCP, FRCPCH David Baum Professor of Paediatrics, Department of Clinical Science at South Bristol, University of Bristol; Honorary Consultant Paediatrician, Department of Immunology and Infectious Diseases, Bristol Royal Hospital for Children, Bristol, United Kingdom Neisseria meningitidis

Anthony E. Fiore, MD, MPH Senior Medical Officer, Epidemiology Branch, Division of Viral Hepatitis, National Center for HIV/AIDS, Viral Hepatitis, STD, and TB Prevention, Centers for Disease Control and Prevention, Atlanta, Georgia Hepatitis A Virus

Marc Fischer, MD, MPH Medical Epidemiologist, Arboviral Diseases Branch, Division of Vector-Borne Infectious Diseases, National Center for Zoonotic, Vector-Borne, and Enteric Diseases, Centers for Disease Control and Prevention (CDC), Fort Collins, Colorado Coltivirus (Colorado Tick Fever)

Sarah J. Fitch, MD Attending Physician, Department of Radiology, Virginia Commonwealth University; Main Hospital of the Medical College of Virginia, Richmond, Virginia Mediastinal and Hilar Lymphadenopathy

J. Dennis Fortenberry, MD, MS Professor of Pediatrics, Indiana University School of Medicine, Indianapolis, Indiana Sexually Transmitted Infection Syndromes

LeAnne M. Fox, MD, MPH Instructor of Pediatrics, Harvard Medical School; Assistant in Medicine, Division of Infectious Diseases, Children’s Hospital, Boston, Massachusetts; Assistant Professor of International Health, Boston University, Boston, Massachusetts; Division of Parasitic Diseases, Centers for Disease Control and Prevention, Atlanta, Georgia Blood and Tissue Nematodes (Filarial Worms)

David O. Freedman, MD Professor of Medicine and Epidemiology, University of Alabama at Birmingham, Division of Infectious Diseases, Center for Geographic Medicine; Director, University of Alabama at Birmingham Travelers Health Clinic, University of Alabama Birmingham Health System, Birmingham, Alabama Antiparasitic Agents

Hayley A. Gans, MD Assistant Professor of Pediatrics, Stanford University School of Medicine; Attending Physician, Division of Infections Diseases, Lucile Packard Children’s Hospital, Stanford, California Hemophagocytic Lymphohistiocytosis and Macrophage Activation Syndrome

Michael A. Gerber, MD Professor of Pediatrics, University of Cincinnati College of Medicine; Attending Physician, Division of Infectious Diseases, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio Pharyngitis; Streptococcus pyogenes (Group A Streptococcus)

Francis Gigliotti, MD Professor of Pediatrics, Microbiology and Immunology, Associate Chair for Academic Affairs; Chief, Division of Infections Diseases, Department of Pediatrics, University of Rochester School of Medicine and Dentistry; Chief, Division of Infectious Diseases, Golisano Children’s Hospital at Strong, Rochester, New York Pneumocystis jirovecii (P. carinii)

Benjamin D. Gold, MD Marcus Professor of Pediatric Gastroenterology, Hepatology and Nutrition, Emory University School of Medicine; Director, Division of Pediatric Gastroenterology and Medical Director, GI Diagnostics and Endoscopy Laboratory, Children’s Healthcare of Atlanta, Egleston Children’s Hospital Campus, Atlanta, Georgia Helicobacter pylori; Other Gastric and Enterohepatic Helicobacter Species

David L. Goldman, MD Associate Professor of Pediatrics, Assistant Professor of Microbiology and Immunology, Albert Einstein College of Medicine; Attending Physician, Department of Pediatrics, Children’s Hospital at Montefiore, Bronx, New York Passive Immunization

Brahm Goldstein, MD, FCCM Medical Director, Clinical Research, Novo Nordisk Inc., Princeton, New Jersey The Systemic Inflammatory Response Syndrome (SIRS), Sepsis, and Septic Shock

Susan T. Goldstein, MD Associate Director for Science, Division of Viral Diseases, National Center for Immunizations and Respiratory Diseases, Centers for Disease Control and Prevention, Atlanta, Georgia Hepatitis B and Hepatitis D Viruses

Jane M. Gould, MD Assistant Professor of Pediatrics, Drexel College of Medicine; Attending Physician, Section of Infectious Diseases, St. Christopher’s Hospital for Children, Philadelphia, Pennsylvania Infection Following Burns; Topical Antimicrobial Agents

Michael Green, MD, MPH Professor of Pediatrics and Surgery, University of Pittsburgh School of Medicine; Attending Physician, Division of Infectious Diseases, Children’s Hospital of Pittsburgh, Pittsburgh, Pennsylvania Infections in Solid-Organ Transplant Recipients

Sharon K. Greene, PhD, MPH Epidemic Intelligence Service Officer, Enteric Diseases Epidemiology Branch, Division of Foodborne, Bacterial and Mycotic Diseases, National Center for Zoonotic, Vector-Borne, and Enteric

Contributor

Diseases, Centers for Disease Control and Prevention, Atlanta, Georgia Other Vibrio Species

Mark J. Greenwald, MD Professor of Ophthalmology & Visual Science and Pediatrics, University of Chicago; Director, Pediatric Ophthalmology and Adult Strabismus Service, University of Chicago Medical Center Chicago, Illinois Endophthalmitis

Alexei A. Grom, MD Associate Professor of Pediatrics, University of Cincinnati College of Medicine; Attending Physician, Division of Rheumatology, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio Fever and the Inflammatory Response

Leigh B. Grossman, MD Vice Provost for International Affairs, University of Virginia School of Medicine; Chief, Pediatric Infectious Disease Division of University of Virginia Health System, Charlottesville, VA Pediatric Infection Prevention and Control

Marta A. Guerra, DVM, MPH, PhD Senior Staff Epidemiologist, Bacterial Zoonoses Branch, Division of Foodborne, Bacterial and Mycotic Diseases, National Center for Zoonotic, Vector-Borne and Enteric Diseases, Centers for Disease Control and Prevention, Atlanta, Georgia Rickettsia rickettsii (Rocky Mountain Spotted Fever)

Kathleen Gutierrez, MD Assistant Professor of Pediatrics, Stanford University School of Medicine; Attending Physician, Division of Infectious Diseases, Lucile Packard Children’s Hospital, Stanford, California Musculoskeletal Symptom Complexes; Osteomyelitis; Infections and Inflammatory Arthritis; Diskitis; Transient Synovitis; Mumps Virus

Judith A. Guzman-Cottrill, DO Assistant Professor of Pediatrics, Oregon Health and Science University; Attending Physician, Division of Infectious Diseases, Doernbecher Children’s Hospital, Portland, Oregon The Systemic Inflammatory Response Syndrome (SIRS), Sepsis, and Septic Shock

Caroline Breese Hall, MD Professor of Pediatrics and Medicine, University of Rochester, School of Medicine and Dentistry; Attending Physician, Division of Infectious Diseases, Golisano Children’s Hospital at Strong, Rochester, New York Human Herpesviruses 6 and 7 (Roseola, Exanthem Subitum); Human Herpesvirus 8

Marvin B. Harper, MD Assistant Professor of Pediatrics, Harvard Medical School; Attending Physician, Division of Infectious Diseases and Division of Emergency Medicine, Children’s Hospital Boston, Boston, Massachusetts Pneumonia in the Immunocompromised Host; Infection Following Bites

David B. Haslam, MD Associate Professor of Pediatrics and Molecular Microbiology, Washington University School of Medicine; Attending Physician, Division of Infectious Diseases, St. Louis Children’s Hospital, St. Louis, Missouri Classification of Streptococci; Enterococcus Species; Viridans Streptococci, Abiotrophia and Granulicatella Species, and Streptococcus bovis; Groups C and G Streptococci; Other Gram-Positive, CatalaseNegative Cocci; Leuconostocs and Pediococci

Edward B. Hayes, MD Chief, Surveillance and Epidemiology Activity, Arboviral Diseases Branch, Division of Vector-Borne Infectious Diseases, National Center for Zoonotic, Vector-borne, and Enteric Diseases, Centers for Disease Control and Prevention, Fort Collins, Colorado Togaviridae; Alphaviruses; Flaviviridae: Flaviviruses

J. Owen Hendley, MD Professor of Pediatrics, University of Virginia School of Medicine; Attending Physician, Division of Pediatric Infectious Disease, University of Virginia Health System, Charlottesville, Virginia The Common Cold; Rhinoviruses

Kelly J. Henrickson, MD Professor of Pediatrics and Microbiology, Medical College of Wisconsin; Attending Physician, Division of Pediatric Infectious Diseases, Children’s Hospital of Wisconsin, Milwaukee, Wisconsin Parainfluenza Viruses

Marion C. W. Henry, MD, MPH Resident, Department of Surgery, Yale University School of Medicine, New Haven, Connecticut Appendicitis

Joseph A. Hilinski, MD Assistant Professor of Pediatrics, Emory University School of Medicine; Division of Pediatric Infectious Diseases, Epidemiology, and Immunology, Emory Children’s Center, Atlanta, Georgia Myocarditis; Pericarditis

Peter J. Hotez, MD, PhD, FAAP Professor and Chair, Department of Microbiology, Immunology and Tropical

xiii

Medicine, The George Washington University, Washington, DC Classification of Parasites; Nematodes

David L. Ingram, MD Professor of Pediatrics, University of North Carolina School of Medicine, Chapel Hill, North Carolina; Director of Pediatric Education, Wakemed Faculty Physicians, Raleigh, North Carolina Infectious Diseases of Child Abuse

Mary Anne Jackson, MD Professor of Pediatrics, University of Missouri – Kansas City School of Medicine; Chief, Section of Infectious Diseases, Children’s Mercy Hospitals and Clinics, Kansas City, Missouri Lymphatic System and Generalized Lymphadenopathy; Mediastinal and Hilar Lymphadenopathy; Abdominal and Retroperitoneal Lymphadenopathy; Localized Lymphadenitis, Lymphadenopathy, and Lymphangitis

Richard F. Jacobs, MD, FAAP Horace C. Cabe Professor and Chairman, Department of Pediatrics, University of Arkansas for Medical Sciences; President, Arkansas Children’s Hospital Research Institute, Little Rock, Arkansas Mycobacterium tuberculosis; Bartonella Species (Cat-Scratch Disease)

M. Gary Karlowicz, MD Professor of Pediatrics, Eastern Virginia Medical School; Attending Physician, Division of Neonatal-Perinatal Medicine, Children’s Hospital of The King’s Daughters, Norfolk, Virginia Nosocomial Infections in the Neonate

Ben Z. Katz, MD Professor of Pediatrics, Northwestern University Feinberg School of Medicine; Attending Physician, Division of Infectious Diseases, Children’s Memorial Hospital, Chicago, Illinois Epstein–Barr Virus (Mononucleosis and Lymphoproliferative Disorders)

Jay S. Keystone, MD, FRCPC, MSc Professor of Medicine, University of Toronto; Staff Physician, Tropical Disease Unit, Department of Medicine, Division of Infectious Disease, Toronto General Hospital, University Health Network, Toronto, Ontario, Canada Protection of Travelers

David W. Kimberlin, MD Associate Professor of Pediatrics, University of Alabama at Birmingham; Attending Physician, Division of Pediatric Infectious Diseases, University of Alabama Health Systems, Birmingham, Alabama Antiviral Agents

xiv

Contributor

Martin B. Kleiman, MD

Moise L. Levy, MD

Sarah S. Long, MD

Ryan White Professor of Pediatrics, Indiana University School of Medicine; Director, Pediatric Infectious Diseases, Director, Infection Control, James Whitcomb Riley Hospital for Children, Indianapolis, Indiana Classification of Fungi; Histoplasma capsulatum (Histoplasmosis); Blastomyces dermatitidis (Blastomycosis); Coccidioides immitis and Coccidioides posadasii (Coccidiomycosis); Agents of Eumycotic Mycetoma; Pseudallescheria boydii (anamorph Scedosporium apiospermum)

Professor of Dermatology and Pediatrics, Baylor College of Medicine; Chief, Dermatology Service, Texas Children’s Hospital, Houston, Texas Cellulitis and Superficial Skin Infections; Erythematous Macules and Papules; Vesicles and Bullae; Purpura; Urticaria and Erythema Multiforme; Papules, Modules and Ulcers; Subcutaneous Tissue Infections and Abscesses; Dermatophytes and Other Superficial Fungi

Mucocutaneous Symptom Complexes; Prolonged, Recurrent, and Periodic Fever Syndromes; Respiratory Tract Symptom Complexes; Encephalitis, Meningoencephalitis, Acute Disseminated Encephalomyelitis, and Acute Necrotizing Encephalopathy; Bordetella pertussis (Pertussis) and Other Bordella Species; Anaerobic Bacteria; Classification, Normal Flora, and Clinical Concepts; Clostridium botulinum (Botulism); Laboratory Manifestations of Infectious Diseases; Principles of Anti-Infective Therapy

Jerome O. Klein, MD Professor of Pediatrics, Boston University School of Medicine; Member, Division of Pediatric Infectious Diseases, Boston Medical Center, Boston, Massachusetts Streptococcus pneumoniae

Mark W. Kline, MD Professor of Pediatrics, Baylor College of Medicine; Chief of Retrovirology, Texas Children’s Hospital, Houston, Texas Diagnosis and Clinical Manifestations of HIV Infection; Infectious Complications of HIV Infection; Management of HIV Infection

Andrew Y. Koh, MD Instructor, Department of Pediatrics, Harvard Medical School; Assistant in Medicine, Department of Infectious Diseases and Hematology/Oncology, Children’s Hospital Boston; Assistant in Medicine, Department of Pediatric Oncology, Dana Farber Cancer Institute, Boston, Massachusetts Fever and Granulocytopenia; Infections in Children with Cancer

Katalin I. Koranyi, MD Professor of Pediatrics, Department of Pediatrics, The Ohio State University College of Medicine and Public Health; Attending Physician, Division of Infectious Diseases, Children’s Hospital, Columbus, Ohio Less Commonly Encountered Enterobacteriaceae; Moraxella and Psychrobacter Species

E. Kent Korgenski, MS, MT Microbiology Technical Supervisor, Microbiology Laboratory, Primary Children’s Medical Center, Salt Lake City, Utah Laboratory Diagnosis of Infection Due to Bacteria, Fungi, Parasites, and Rickettsiae

Robert J. Leggiadro, MD Professor of Clinical Pediatrics, Weill Medical College of Cornell University, New York, New York; Chairman and Program Director, Department of Pediatrics, Lincoln Medical and Mental Health Center, Bronx, New York Infections of the Oral Cavity

David B. Lewis, MD Professor of Pediatrics, Stanford University School of Medicine; Attending Physician in Immunology and Infectious Diseases, Department of Pediatrics, Lucile Packard Children’s Hospital; Member, Interdepartmental Program in Immunology, Stanford University, Stanford, California Hemophagocytic Lymphohistiocytosis and Macrophage Activation Syndrome; Infectious Complications of Cell-Mediated Immunity Other Than AIDS: Primary Immunodeficiencies

Jay M. Lieberman, MD Professor of Clinical Pediatrics, University of California, Irvine School of Medicine, Irvine, California; Chief, Pediatric Infectious Diseases, Miller Children’s Hospital, Long Beach, California Infection Following Trauma; Neisseria gonnorrhoeae; Other Neisseria Species

Abhijit Limaye, MD Associate Professor of Laboratory Medicine and Medicine, University of Washington School of Medicine; Co-Director of Microbiology, Medicine, Division of Infectious Diseases, University of Washington Medical Center, Seattle, Washington Human Polyomaviruses

Jacob A. Lohr, MD Lohr Distinguished Professor of General Pediatrics and Associate Chair, Department of Pediatrics, University of North Carolina at Chapel Hill; Attending Physician, North Carolina Children’s Hospital, Chapel Hill, North Carolina Urinary Tract Infections; Renal (Intrarenal and Perinephric) Abscesses

Bennett Lorber, MD, DSc (hon) Thomas M. Durant Professor of Medicine, Temple University School of Medicine; Attending Physician, Section of Infectious Diseases, Temple University Hospital; Consultant, Fox Chase Cancer Center, Philadelphia, Pennsylvania Listeria monocytogenes

Donald E. Low, MD, FRCPC Professor, Department of Laboratory Medicine and Pathobiology and Department of Medicine, University of Toronto; Microbiologist-in-Chief, University Health Network/Mount Sinai Hospital, Toronto, Ontario, Canada Myositis, Pyomyositis, and Necrotizing Fasciitis

Gina Lowell, MD Clinical Instructor, Department of Pediatrics, University of Illinois at Chicago, Chicago, Illinois Staphylococcus aureus

Elizabeth Lowenthal, MD Instructor, Department of Pediatrics, Baylor College of Medicine; Associate Director, Botswana Baylor Children’s Clinical Center of Excellence, Gaborone, Botswana, Africa Management of HIV Infection

Jorge Lujan-Zilbermann, MD Assistant Professor of Pediatrics, Division of Infectious Diseases, University of South Florida College of Medicine, Tampa, Florida; Attending Physician, All Children’s Hospital, St. Petersburg, Florida; Attending Physician, Tampa General Hospital, Tampa, Florida Infections in Hematopoietic Stem Cell Transplant Recipients

Katherine Luzuriaga, MD Professor of Pediatrics, Division of Immunology and Infectious Diseases, University of Massachusetts Medical School, Worcester, Massachusetts Immunology of HIV-1 Infection; Introduction to Retroviridae; Human T-Cell Lymphotropic Viruses; Human Immunodeficiency Viruses

Noni E. MacDonald, Msc, MD, FRCP Professor of Pediatrics, Dalhousie University; Attending Physician, Division of Infectious Diseases, IWK Health Center, Halifax, Nova Scotia, Canada Epididymitis, Orchitis, and Prostatitis

Contributor

Yvonne A. Maldonado, MD

Robert F. Massung, PhD

Rima McLeod, MD

Associate Professor of Pediatrics, Stanford University School of Medicine; Attending Physician, Division of Infectious Diseases, Lucile Packard Children’s Hospital, Stanford, California Epidemiology and Prevention of HIV Infection in Children and Adolescents; Rubella Virus; Rubeola Virus (Measles and Subacute Sclerosing Panencephalitis); Polioviruses

Chief, Rickettsial Zoonoses Branch, Division of Viral and Rickettsial Diseases, National Center for Zoonotic, Vector-Borne, and Enteric Diseases, Centers for Disease Control and Prevention, Atlanta, Georgia Ehrlichia and Anaplasma Species

Professor, Departments of Ophthalmology and Visual Sciences and Pediatrics (Infectious Diseases), Committees on Molecular Medicine, Genetics and The College and Attending Physician, The University of Chicago; Attending Physician, Michael Reese Hospital, Chicago, Illinois Toxoplasma gondii (Toxoplasmosis)

Chitra S. Mani, MBBS, DCH, FAAP Associate Professor of Pediatrics, Medical College of Georgia; Attending Physician, Division of Infectious Diseases, Associate Hospital Epidemiologist, Children’s Medical Center, Augusta, Georgia Acute Pneumonia and its Complications; Persistent and Recurrent Pneumonia

John F. Marcinak, MD Associate Professor, Department of Pediatrics (Infectious Diseases), Associate Director, University of Chicago Infection Control Program, The University of Chicago, Chicago, Illinois Toxoplasma gondii (Toxoplasmosis)

Mario J. Marcon, PhD Clinical Associate Professor of Pathology and Laboratory Medicine, Department of Pathology, The Ohio State University College of Medicine and Public Health; Director, Clinical/Molecular Microbiology and Immunodiagnostics, Department of Laboratory Medicine, Children’s Hospital, Columbus, Ohio Klebsiella and Raoultella Species; Enterobacter and Pantoea Species; Less Commonly Encountered Enterobacteriaceae; Proteus, Providencia, and Morganella Species; Acinetobacter Species; Less Commonly Encountered Nonenteric GramNegative Bacilli; Eikenella, Pasteurella, and Chromobacterium Species; Moraxella and Psychrobacter Species

Gary S. Marshall, MD Professor of Pediatrics, University of Louisville School of Medicine; Chief, Division of Pediatric Infectious Diseases, Kosair Children’s Hospital, Louisville, Kentucky Chronic Fatigue Syndrome

Stacey W. Martin, MSc Epidemiologist and Statistican, Meningitis and Vaccine Preventable Diseases Branch, Division of Bacterial Diseases, National Center for Immunization and Respiratory Diseases, Centers for Disease Control and Prevention, Atlanta, Georgia Principles of Epidemiology and Public Health

Eric E. Mast, MD, MPH Chief, Prevention Branch, Division of Viral Hepatitis, National Center for HIV/AIDS, Viral Hepatitis, STD, and TB Prevention, Centers for Disease Control and Prevention, Atlanta, Georgia Hepatitis E Virus and Other Newly Identified Viruses

xv

Julia A. McMillan, MD Professor of Pediatrics, Vice Chair for Pediatric Education, Department of Pediatrics, Johns Hopkins University School of Medicine, Baltimore, Maryland Bioterrorism

Tony Mazzulli, MD, FRCPC Professor, Department of Laboratory Medicine and Pathobiology, University of Toronto; Deputy Chief Microbiologist, Mount Sinai Hospital, Toronto, Ontario, Canada Laboratory Diagnosis of Infection due to Viruses, Chlamydia, Chlamydophila, and Mycoplasma

George H. McCracken, MD Professor of Pediatrics, Charles E and Sarah M Seay Chair in Pediatric Infectious Disease, GlaxoSmithKline Distinguished Professor of Pediatric Infectious Disease, Department of Pediatrics, University of Texas Southwestern Medical Center at Dallas, Dallas, Texas; Attending Physician, Children’s Medical Center of Dallas, Dallas, Texas Acute Bacterial Meningitis Beyond the Neonatal Period

Robert S. McGregor, MD Professor of Pediatrics, Associate Chair, Department of Pediatrics, Drexel University College of Medicine; Residency Director and Attending Physician, St. Christopher’s Hospital for Children, Philadelphia, Pennsylvania Abdominal Symptom Complexes

Kenneth McIntosh, MD Professor of Pediatrics, Department of Pediatrics, Harvard Medical School; Emeritus Chief, Division of Infectious Diseases, Children’s Hospital Boston, Boston, Massachusetts Pneumonia in the Immunocompromised Host

Catherine A. McLean, MD Senior Medical Advisor, HIV Prevention Branch, Global AIDS Program, National Center for HIV/AIDS, Viral Hepatitis, STD and TB Prevention, Centers for Disease Control and Prevention; Assistant Professor of Medicine, Department of Medicine, Emory University School of Medicine, Atlanta, Georgia Skin and Mucous Membrane Infections and Inguinal Lymphadenopathy; Klebsiella (Calymmatobacterium) granulomatis (Granuloma Inguinale)

Jennifer H. McQuiston, DVM, MS Zoonoses Team Leader, Geographic Medicine and Health Promotion Branch, Division of Global Migration and Quarantine, National Center for Preparedness, Detection, and Control of Infectious Diseases, Centers for Disease Control and Prevention, Atlanta, Georgia Coxiella burnetii (Q Fever)

H. Cody Meissner, MD Professor of Pediatrics, Tufts University School of Medicine; Chief, Pediatric Infectious Disease, Department of Pediatrics, Tufts–New England Medical Center, Boston, Massachusetts Bronchiolitis

Manoj P. Menon, MD, MPH Medical Officer, Enteric Diseases Epidemiology Branch, Division of Foodborne, Bacterial, and Mycotic Diseases, National Center for Zoonotic Vector-Borne and Enteric Diseases, Centers for Disease Control and Prevention, Atlanta, Georgia Vibrio cholerae (Cholera)

Marian G. Michaels, MD, MPH Professor of Pediatrics and Surgery, University of Pittsburgh School of Medicine; Attending Physician, Division of Infectious Diseases, Children’s Hospital of Pittsburgh, Pittsburgh, Pennsylvania Eosinophilic Meningitis; Infections in SolidOrgan Transplant Recipients

Melissa B. Miller, PhD Assistant Professor, Department of Pathology and Laboratory Medicine, University of North Carolina School of Medicine; Director, Molecular Microbiology, Associate Director, Microbiology-Immunology Laboratory, McLendon Clinical Laboratories, University of North Carolina Hospitals, Chapel Hill, North Carolina Mechanisms and Detection of Antibiotic Resistance

xvi

Contributor

Juan Carlos Millon, MD

Simon Nadel, FRCP

Sara M. O’Hara, MD

Instructor of Pediatrics, Baylor College of Medicine; Attending Physician, Section of Retrovirology, Texas Children’s Hospital, Houston, Texas Management of HIV Infection

Consultant in Paediatric Intensive Care, Department of Paediatrics, Imperial College London; Department of Paediatrics, St. Mary’s Hospital, London, United Kingdom The Systemic Inflammatory Response Syndrome (SIRS), Sepsis, and Septic Shock

Associate Professor of Radiology and Pediatrics, University of Cincinnati College of Medicine; Chief, Division of Ultrasound, Department of Radiology, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio Renal (Intrarenal and Perinephric) Abscesses

John F. Modlin, MD Professor and Chairman, Department of Pediatrics, Dartmouth Medical School; Medical Director, Children’s Hospital at Dartmouth, Dartmouth-Hitchcock Medical Center, Lebanon, New Hampshire Introduction to Picornaviridae; Enteroviruses; Coxsackieviruses, Echoviruses, and Newer Enteroviruses

James P. Nataro, MD, PhD

Matthew R. Moore, MD, MPH

Michael N. Neely, MD

Medical Epidemiologist, Respiratory Diseases Branch, Division of Bacterial Diseases, National Center for Immunization and Respiratory Diseases, Centers for Disease Control and Prevention, Atlanta, Georgia Chlamydophila (Chlamydia) psittaci (Psittacosis)

Assistant Professor of Pediatrics, Keck School of Medicine, University of Southern California; Physician Specialist, Department of Pediatrics, Los Angeles County and University of Southern California Medical Center and Children’s Hospital Los Angeles, Los Angeles, California Pharmacokinetic–Pharmacodynamic Basis of Optimal Antibiotic Therapy

Zack S. Moore, MD, MPH Medical Epidemiologist, Communicable Disease Branch, North Carolina Division of Public Health, Raleigh, North Carolina Poxviridae

Mary M. Moran, MD Associate Professor of Pediatrics, Associate Dean for Faculty Development, Department of Pediatrics, Drexel University College of Medicine; Attending Physician, Department of Pediatrics, St. Christopher’s Hospital for Children, Philadelphia, Pennsylvania Neurologic Symptom Complexes

Pedro L. Moro, MD, MPH Research Analyst, Immunization Safety Office, Office of the Chief Science Officer, Centers for Disease Control and Prevention, Atlanta, Georgia Echinococcus Species (Agents of Cystic, Alveolar, and Polycystic Echinococcosis)

R. Lawrence Moss, MD Robert Pritzker Professor and Chair of Pediatric Surgery, Yale University School of Medicine; Surgeon-in-Chief, Yale New Haven Children’s Hospital, New Haven, Connecticut Peritonitis; Appendicitis; Intra-Abdominal, Visceral, and Retroperitoneal Abscesses

Dennis L. Murray, MD Professor of Pediatrics, Medical College of Georgia; Chief, Infectious Diseases, Associate Medical Director, Children’s Medical Center, Augusta, Georgia Acute Pneumonia and its Complications; Persistent and Recurrent Pneumonia; Pneumonia in the Immunocompromised Host

Professor of Pediatrics, Medicine, Microbiology & Immunology, and Biochemistry and Molecular Biology, Center for Vaccine Development, University of Maryland School of Medicine, Baltimore, Maryland Inflammatory Enteritis; Escherichia coli

Victor Nizet, MD Professor of Pediatrics and Pharmacy, University of California, San Diego School of Medicine, Skaggs School of Pharmacy and Pharmaceutical Sciences, La Jolla, California; Chief, Division of Pediatric Pharmacology and Drug Discovery, Department of Pediatrics, Rady Children’s Hospital San Diego, San Diego, California Localized Lymphadenitis, Lymphadenopathy, and Lymphangitis

Anna Norrby-Teglund, PhD Karolinska University Hospital, Center for Infectious Medicine, Karolinska Institute, Stockholm, Sweden Myositis, Pyomyositis, and Necrotizing Fasciitis

Ann-Christine Nyquist, MD, MSPH Associate Professor, Department of Pediatrics and Preventive Medicine/Biometrics, University of Colorado School of Medicine; Medical Director, Infection Control; Attending Physician, Pediatric Infectious Diseases and Epidemiology, The Children’s Hospital, Denver, Colorado Laboratory Manifestations of Infectious Diseases

Theresa J. Ochoa, MD Assistant Professor of Pediatrics, Universidad Peruana Cayetano Heredia, Lima, Peru; Post Doctoral Fellow, Department of Medicine, Baylor College of Medicine, Houston, Texas Yersinia Species

Walter A. Orenstein, MD Professor of Medicine and Pediatrics, Division of Infectious Diseases, Emory University School of Medicine; Associate Director, Emory Vaccine Center, Atlanta, Georgia Active Immunization

Eduardo Ortega-Barria, MD Head of Clinical Research and Development & Medical Affairs, GlaxoSmithKline Biologicals, Latin America & Caribbean; Adjunct Professor, Department of Microbiology and Tropical Medicine, The George Washington University; Associate Investigator, Institute of Advanced Scientific Investigations & High Technology Services (INDICASAT); National Council for Science, Technology and Innovation, Panama City, Panama; Associate Scientist, Smithsonian Tropical Research Institute, Panama City, Panama Babesia Species (Babesiosis); Balantidium coli; Blastocystis hominis; Endolimax nana; Leishmania Species (Leishmaniasis); Sarcocystis Species; Trypanosoma Species (Trypanosomiasis)

Gary D. Overturf, MD Professor of Pediatrics and Pathology, University of New Mexico School of Medicine; Senior Staff Consultant, Department of Pediatrics, Division of Infectious Diseases, Children’s Hospital of New Mexico; Medical Director, Infectious Diseases, TriCore Reference Laboratories, Albuquerque, New Mexico Chemoprophylaxis; Corynebacterium diphtheriae; Bunyaviridae

Christopher D. Paddock, MD, MPHTM Staff Pathologist and Research Medical Officer, Infectious Disease Pathology Branch, Division of Viral and Rickettsial Diseases, National Center for Zoonotic, Vector-Borne, and Enteric Diseases, Centers for Disease Control and Prevention, Atlanta, Georgia Ehrlichia and Anaplasma Species; Rickettsia rickettsii (Rocky Mountain Spotted Fever)

John A. Painter, DVM, MS Epidemiologist, Enteric Diseases Epidemiology Branch, Division of Foodborne, Bacterial, and Mycotic Diseases, National Center for Zoonotic, Vector-Borne,

Contributor

and Enteric Diseases, Centers for Disease Control and Prevention, Atlanta, Georgia Other Vibrio Species

Diane E. Pappas, MD, JD Associate Professor of Clinical Pediatrics, University of Virginia School of Medicine; University of Virginia Health System, Charlottesville, Virginia The Common Cold; Rhinoviruses

Monica E. Parise, MD Chief, Parasitic Diseases Branch, Division of Parasitic Diseases, National Center for Zoonotic, Vector-Borne, and Enteric Diseases, Centers for Disease Control and Prevention, Atlanta, Georgia Plasmodium Species (Malaria)

Robert F. Pass, MD Professor of Pediatrics and Microbiology, University of Alabama, Birmingham School of Medicine; Division of Infectious Diseases, Children’s Hospital of Alabama, Birmingham, Alabama Viral Infections in the Fetus and Neonate; Cytomegalovirus

Thomas F. Patterson, MD Professor of Medicine; Director, Infectious Diseases Fellowship Training Program; Attending Physician, Division of Infectious Diseases, University of Texas Health Science Center at San Antonio, San Antonio, Texas Agents of Hyalohyphomycosis and Phaeohyphomycosis; Agents of Zygomycosis (Mucormycosis); Malassezia Species; Sporothrix schenckii (Sporotrichosis); Cryptococcus Species

Andrew T. Pavia, MD George and Esther Gross Presidential Professor of Pediatrics, University of Utah; Chief, Division of Infectious Diseases, Primary Children’s Medical Center, Salt Lake City, Utah Foodborne and Waterborne Disease

Stephen I. Pelton, MD Professor of Pediatrics and Epidemiology, Boston University Schools of Medicine and Public Health; Chief, Section of Pediatric Infectious Diseases, Boston Medical Center, Boston, Massachusetts Otitis Media

Georges Peter, MD Professor Emeritus of Pediatrics, Warren Alpert Medical School at Brown University, Providence, Rhode Island; Senior Lecturer in Pediatrics, Harvard Medical School, Boston, Massachusetts Streptococcus pneumoniae

Timothy R. Peters, MD Assistant Professor of Pediatrics, Wake Forest University of Health Sciences;

Attending Physician, Section of Infectious Diseases, Brenner Children’s Hospital, Winston-Salem, North Carolina Respiratory Syncytial Virus

Charles G. Prober, MD

William A. Petri, Jr, MD, PhD

Shawn J. Rangel, MD, MSCE

Wade Hampton Frost Professor of Epidemiology, Professor of Medicine, Microbiology and Pathology, University of Virginia School of Medicine, Charlottesville, Virginia Entamoeba histolytica (Amebiasis); Other Entameba, Amebas, and Intestinal Flagellates; Naegleria fowleri; Acanthamoeba Species

xvii

Classification of Viruses; Introduction to Herpesviridae; Herpes Simplex Virus; Human T-Cell Lymphotropic Viruses

Clinical Fellow, Department of Surgery, University of Cincinnati, Cincinnati, Ohio; Clinical Fellow, Department of Pediatric Surgery, Cincinnati Children’s Hospital and Medical Center, Cincinnati, Ohio Peritonitis

Sarah Anne Rawstron, MB, BS

Infections Associated with Group Childcare; Active Immunization; Approach to the Diagnosis and Management of Gastrointestinal Tract Infections; Giardia lamblia (Giardiasis)

Associate Professor of Clinical Pediatrics, Weill Medical College of Cornell University, New York, New York; Pediatric Residency Program Director, Department of Pediatrics, The Brooklyn Hospital Center, Brooklyn, New York Treponema pallidum (Syphilis); Other Treponema Species

Philip A. Pizzo, MD

Michael D. Reed, PharmD

Larry K. Pickering, MD

Carl and Elizabeth Naumann Dean, Professor of Pediatrics and of Microbiology and Immunology, Stanford University School of Medicine; Medical Staff, Department of Pediatrics-Hematology/Oncology, Stanford Hospital and Clinics, Stanford, California Fever and Granulocytopenia; Infections in Children with Cancer

Andrew J. Pollard, FRCPCH, PhD Reader in Paediatric Infection and Immunity, University of Oxford; Honorary Consultant Paediatrician, Children’s Hospital; Consultant in Charge of the Oxford Vaccine Group, Oxford, United Kingdom Neisseria meningitidis

Susan M. Poutanen, MD, MPH, FRCPC Assistant Professor, Departments of Laboratory Medicine and Pathobiology and Medicine, University of Toronto; Medical Microbiologist and Infectious Disease Consultant, University Health Network and Mount Sinai Hospital, Toronto, Ontario, Canada Human Coronaviruses

Professor of Pediatrics, Case Western Reserve University School of Medicine; Director, Pediatric Clinical Pharmacology and Toxicology, Rainbow Babies and Children’s Hospital, Cleveland, Ohio Pharmacokinetic–Pharmacodynamic Basis of Optimal Antibiotic Therapy

Megan E. Reller, MDCM, MPH, DTM&H Instructor in Pediatrics, Department of Pediatrics, Harvard Medical School, Boston, Massachusetts; Assistant in Medicine, Infectious Diseases, Department of Medicine, Children’s Hospital Boston; Clinical Assistant, Infectious Diseases, Department of Medicine, Massachusetts General Hospital, Boston, Massachusetts Salmonella Species; Vibrio cholerae (Cholera)

Frank O. Richards Jr, MD

Professor of Pediatrics, The Ohio State University College of Medicine and Public Health; Chief, Section of Infectious Diseases, Children’s Hospital, Columbus, Ohio Citrobacter Species; Serratia Species; Acinetobacter Species

Director, Malaria, River Blindness, Lymphatic Filariasis, Schistomsomiasis, The Carter Center, Atlanta, Georgia; Assistant Adjunct Professor of Pediatrics, Associate Adjunct Professor of Global Health, Emory University, Atlanta, Georgia; Professional Staff, Children’s Health Care Atlanta, Atlanta, Georgia Diphyllobothrium, Dipylidium, and Hymenolepsis Species; Intestinal Trematodes; Clonorchis, Opisthorchis, Fasciola, and Paragonimus Species; Blood Trematodes (Schistosomiasis)

Alice S. Prince, MD

Gail L. Rodgers, MD

Professor of Pediatrics, Columbia University College of Physicians and Surgeons; Attending Physician, Infectious Diseases, Morgan Stanley Children’s Hospital of New York, New York, New York Infectious Complications in Special Hosts; Pseudomonas aeruginosa

Director, Clinical Affairs, Vaccines, Global Medical Affairs, Wyeth, Collegeville, Pennsylvania Infection Following Burns; Topical Antimicrobial Agents

Dwight A. Powell, MD

xviii

Contributor

Luz I. Romero, MD Director, Institute of Advanced Scientific Investigations & High Technology Services (INDICASAT); National Council for Science, Technology and Innovation, Panama City, Panama Leishmania Species (Leishmaniasis)

Harley A. Rotbart, MD Professor and Vice Chairman of Pediatrics, Professor of Microbiology, University of Colorado School of Medicine; Attending Physician, Division of Infectious Diseases, The Children’s Hospital, Denver, Colorado Aseptic and Viral Meningitis

Anne H. Rowley, MD Professor of Pediatrics and of Microbiology/Immunology, Northwestern University Feinberg School of Medicine; Attending Physician, Division of Infectious Diseases, The Children’s Memorial Hospital, Chicago, Illinois Kawasaki Disease

Lorry G. Rubin, MD Professor of Pediatrics, Albert Einstein College of Medicine, Bronx, New York; Chief, Division of Infectious Diseases, Schneider Children’s Hospital of the North Shore-Long Island Jewish Health System, New Hyde Park, New York Capnocytophaga Species; Francisella tularensis (Tularemia); Kingella Species; Legionella Species; Streptobacillus moniliformis (Rat-Bite Fever); Other GramNegative Coccobacilli

Guillermo M. Ruiz-Palacios, MD Professor and Head, Department of Infectious Diseases, National Institute of Medical Sciences and Nutrition, Mexico City, Mexico Campylobacter jejuni and Campylobacter coli; Other Campylobacter Species

Xavier Sáez-Llorens, MD Professor of Pediatrics, University of Panama School of Medicine; Head, Infectious Disease Department, Hospital del Niño, Panama, Republic of Panama Acute Bacterial Meningitis Beyond the Neonatal Period

Lisa Saiman, MD, MPH Professor of Clinical Pediatrics, Columbia University College of Physicians and Surgeons; Attending Physician, Section of Infectious Diseases, Morgan Stanley Children’s Hospital of New York, New York Presbyterian Hospital, New York, New York Endocarditis and Intravascular Infections

Jason B. Sauberan, PharmD Assistant Clinical Professor, University of California, San Diego Skaggs School of Pharmacy and Pharmaceutical Sciences;

Department of Pharmacy, University of California San Diego Medical Center, San Diego, California Antimicrobial Agents

Mark H. Sawyer, MD Professor of Clinical Pediatrics, University of California, San Diego, California; Senior Staff, Children’s Hospital San Diego, San Diego, California Aseptic and Viral Meningitis

Peter M. Schantz, DVM, PhD Epidemiologist, Parasitic Diseases Branch, Division of Parasitic Diseases, National Center for Zoonotic, Vector-Borne, and Enteric Diseases, Centers for Disease Control and Prevention, Atlanta, Georgia; Adjunct Professor, Department of Epidemiology, Rollins School of Public Health, Emory University, Atlanta, Georgia Taenia solium and Taenia saginata (Taeniasis and Cysticercosis); Echinococcus Species (Agents of Cystic, Alveolar, and Polycystic Echinococcosis); Taenia (Multiceps) multiceps and Taenia serialis (Coenurosis)

Theresa A. Schlager, MD Professor of Pediatrics and Emergency Medicine, University of Virginia School of Medicine; Attending Physician, University of Virginia Health System, Charlottesville, Virginia Urinary Tract Infections

Gordon E. Schutze, MD Professor of Pediatrics, Baylor College of Medicine; Attending Physician, Texas Children’s Hospital, Houston, Texas; Vice President, International Medical Services, Baylor College of Medicine International Pediatric AIDS Initiative, Houston, Texas Bartonella Species (Cat-Scratch Disease)

Children’s Health Center for International Adoption, Department of Pediatrics, Texas Children’s Hospital, Houston, Texas Diagnosis and Clinical Manifestations of HIV Infection

Samir S. Shah, MD Assistant Professor of Pediatrics, University of Pennsylvania School of Medicine; Attending Physician, Division of Infectious Diseases and Division of General Pediatrics, The Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania Chlamydophila (Chlamydia) pneumoniae; Mycoplasma pneumoniae; Other Mycoplasma Species; Ureaplasma urealyticum

Andi L. Shane, MD, MPH Fellow, Division of Pediatric Infectious Diseases, Department of Pediatrics, University of California at San Francisco, San Francisco, California; Assistant Professor, Division of Infectious Diseases, Department of Pediatrics, Emory University School of Medicine, Atlanta, Georgia Infections Associated with Group Childcare

Eugene D. Shapiro, MD Professor of Pediatrics, Epidemiology and Public Health and Investigative Medicine, Yale University School of Medicine; Attending Pediatrician, Department of Pediatrics, Children’s Hospital at Yale–New Haven, New Haven, Connecticut Fever Without Localizing Signs; Leptospira Species (Leptospirosis); Borrelia burgdorferi (Lyme Disease); Other Borrelia Species and Spirillum minus

Avinash K. Shetty, MD

Senior Science Advisor, National Vaccine Program Office, U.S. Department of Health and Human Services, Washington, DC Principles of Epidemiology and Public Health

Associate Professor of Pediatrics, Wake Forest University of Health Sciences, Winston-Salem, North Carolina; Attending Physician, Section of Infectious Diseases, Brenner Children’s Hospital, Winston-Salem, North Carolina Epidemiology and Prevention of HIV Infection in Children and Adolescents

Richard H. Schwartz, MD

Jane D. Siegel, MD

Professor of Pediatrics, Virginia Commonwealth University Medical School, (Medical College of Virginia); Clinical Professor of Pediatrics, University of Virginia School of Medicine; Clinical Professor of Pediatrics, George Washington University School of Medicine, Washington, DC; Attending Physician, Department of Pediatrics, Inova Fairfax Hospital for Children, Falls Church, Virginia Infections Related to the Upper and Middle Airways

Professor of Pediatrics, University of Texas Southwestern Medical Center at Dallas; Attending Physician, Medical Director, Infection Control, Children’s Medical Center of Dallas; Attending Physician, Parkland Health and Hospital System, Dallas, Texas Pediatric Infection Prevention and Control

Benjamin Schwartz, MD

Heidi Schwarzwald, MD, MPH Assistant Professor of Pediatrics, Baylor College of Medicine; Director, Texas

Robert D. Siegel, MD, PhD Associate Professor (Teaching), Department of Microbiology and Immunology, Program in Human Biology, and Center for African Studies, Stanford University, Stanford, California Classification of Viruses

Contributor

Walter E. B. Sipe, MD

William J. Steinbach, MD

Fellow, Department of Pediatrics, University of California–San Francisco, San Francisco, California Chronic Hepatitis

Assistant Professor of Pediatrics, Molecular Genetics and Microbiology, Duke University School of Medicine, Durham, North Carolina Candida Species; Aspergillus Species; Antifungal Agents

Jacek Skarbinski, MD Epidemic Intelligence Service Officer, Malaria Branch, Division of Parasitic Diseases, National Center for Zoonotic, Vector-Borne, and Enteric Diseases, Centers for Disease Control and Prevention, Atlanta, Georgia Plasmodium Species (Malaria)

P. Brian Smith, MD Fellow, Department of Molecular Genetics and Microbiology, Duke Clinical Research Institute, Duke University; Fellow, Department of Pediatrics, Duke University Medical Center, Durham, North Carolina Clinical Approach to the Infected Neonate; Candida Species

John D. Snyder, MD Professor, Department of Pediatrics, University of California–San Francisco, San Francisco, California Acute Hepatitis; Chronic Hepatitis

Shahram SolaymaniMohammadi, BVSc, MSPH, PhD Candidate Research Fellow, Department of Medical Parasitology and Mycology, School of Public Health and Institute of Public Health Research, Tehran University of Medical Sciences, Tehran, Iran; Visiting Researcher, Division of Infectious Diseases and International Health, Department of Internal Medicine, University of Virginia School of Medicine, Charlottesville, Virginia Entamoeba histolytica (Amebiasis); Other Entameba, Amebas, and Intestinal Flagellates; Naegleria fowleri

Mary Allen Staat, MD, MPH Associate Professor of Pediatrics, University of Cincinnati College of Medicine; Associate Director, International Adoption Center, Department of Pediatrics, Division of Infectious Diseases, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio Infectious Diseases in Refugee and Internationally Adopted Children

Jeffrey R. Starke, MD Professor and Vice Chairman, Department of Pediatrics, Baylor College of Medicine; Chief, Department of Pediatrics, Ben Taub General Hospital; Infection Control Officer, Texas Children’s Hospital, Houston, Texas Mycobacterium tuberculosis

Ina Stephens, MD Pediatric Residency Program Director, Sinai Hospital of Baltimore, Baltimore, Maryland Inflammatory Enteritis

xix

Georgia Rickettsia rickettsii (Rocky Mountain Spotted Fever)

Robert V. Tauxe, MD, MPH Deputy Director, Division of Foodborne, Bacterial and Mycotic Diseases, National Center for Zoonotic, Vector-Borne and Enteric Diseases, Centers for Disease Control and Prevention, Atlanta, Georgia Vibrio cholerae (Cholera)

Herbert A. Thompson, PhD Joseph W. St. Geme, III, MD Professor of Pediatrics and Molecular Genetics and Microbiology, Chairman of Pediatrics, Duke University Medical Center; Attending Physician, Department of Pediatrics, Duke Children’s Hospital, Durham, North Carolina Classification of Bacteria; Classification of Streptococci; Enterococcus Species; Viridans Streptococci, Abiotrophia and Granulicatella Species, and Streptococcus bovis; Group C and G Streptococci; Other Gram-Positive, Catalase-Negative Cocci; Leuconostocs and Pediococci; Haemophilus influenzae

Kanta Subbarao, MBBS, MPH Senior Investigator, Laboratory of Infectious Diseases, National Institutes of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, Maryland Influenza Viruses

John L. Sullivan, MD Professor of Pediatrics, Vice Chancellor, Office of Research, University of Massachusetts Medical School, Worcester, Massachusetts Immunology of HIV-1 Infection; Introduction to Retroviridae; Human Immunodeficiency Viruses

Deanna A. Sutton, PhD Assistant Professor/Research, Department of Pathology, University of Texas Health Science Center at San Antonio, San Antonio, Texas Malassezia Species

Madeline Y. Sutton, MD, MPH Epidemiology Branch, Division of HIV/AIDS Prevention, Surveillance, and Epidemiology, National Center for HIV/AIDS, Viral Hepatitis, STD, and TB Prevention, Centers for Disease Control and Prevention, Atlanta, Georgia Pelvic Inflammatory Disease

David L. Swerdlow, MD Team Leader, Disease Assessment and Epidemiology Team, Rickettsial Zoonoses Branch, Division of Viral and Rickettsial Diseases, National Center for Zoonotic, Vector-Borne and Enteric Diseases, Centers for Disease Control and Prevention, Atlanta,

Chief, Viral and Rickettsial Zoonoses Branch, Division of Viral and Rickettsial Diseases, National Center for Infectious Diseases, Centers for Disease Control and Prevention, Atlanta, Georgia Coxiella burnetii (Q Fever)

Richard B. Thomson, Jr, PhD Professor of Pathology, Northwestern University Feinberg School of Medicine, Chicago, Illinois; Director, Microbiology and Virology Laboratories, Department of Pathology and Laboratory Medicine, Evanston Northwestern Healthcare, Evanston, Illinois Nocardia Species

Emily A. Thorell, MD Instructor of Pediatrics, University of Utah Health Sciences Center; Attending Physician, Divisions of Pediatric Infectious Diseases and Hospital Medicine, Primary Children’s Medical Center, Salt Lake City, Utah Cervical Lymphadenitis and Neck Infections

James K. Todd, MD Professor of Pediatrics, Microbiology and Preventive Medicine/Biometrics, University of Colorado School of Medicine; Director of Epidemiology, Clinical Microbiology, and Clinical Outcomes, Department of Epidemiology, The Children’s Hospital, Denver, Colorado Toxic Shock Syndrome

Philip Toltzis, MD Associate Professor of Pediatrics, Case Western Reserve University School of Medicine; Attending Physician, Divisions of Pharmacology of Critical Care, Rainbow Babies and Children’s Hospital, Cleveland, Ohio Staphylococcus epidermidis and Other Coagulase-Negative Staphylococci

Theodore F. Tsai, MD, MPH Medical Director, North America, Novartis Vaccines and Diagnostics, Philadelphia, Pennsylvania Coltivirus (Colorado Tick Fever); Togaviridae; Alphaviruses

xx

Contributor

Ellen R. Wald, MD

Ian T. Williams, PhD, MS

Alfred Dorrance Daniels Professor and Chair, Department of Pediatrics, University of Wisconsin; Physician-in-Chief, University of Wisconsin Children’s Hospital, Madison, Wisconsin Mastoiditis; Sinusitis; Periorbital and Orbital Infections

Acting Branch Chief, Division of Viral Hepatitis, Epidemiology Branch, Division of Viral Hepatitis, National Center for HIV/AIDS, Viral Hepatitis, STD, and TB Prevention, Centers for Disease Control and Prevention, Atlanta, Georgia Hepatitis C Virus

Richard J. Wallace, Jr, MD

John V. Williams, MD

Professor of Medicine and Microbiology, University of Texas Health Center at Tyler, Tyler, Texas Mycobacterium Species Non-tuberculosis

Assistant Professor of Pediatrics, Division of Pediatric Infectious Diseases, Vanderbilt University Medical Center, Nashville, Tennessee Human Metapneumovirus

Geoffrey A. Weinberg, MD Associate Professor of Pediatrics, University of Rochester School of Medicine and Dentistry; Director, Pediatric HIV Program, Golisano Children’s Hospital at Strong, Strong Memorial Hospital, Rochester, New York Neurologic Symptom Complexes

Avery H. Weiss, MD Associate Professor of Ophthalmology, University of Washington; Chief, Division of Ophthalmology, Department of Surgery, Children’s Hospital and Regional Medical Center; Affiliate Professor, Department of Pediatrics, Children’s Hospital and Regional Medical Center, Seattle, Washington Conjunctivitis in the Neonatal Period (Ophthalmia Neonatorum); Conjunctivitis Beyond the Neonatal Period; Infective Keratitis; Uveitis, Retinitis, and Chorioretinitis

Rodney E. Willoughby, Jr, MD Associate Professor of Pediatrics, Medical College of Wisconsin; Attending Physician, Division of Infectious Diseases, Children’s Hospital of Wisconsin, Milwaukee, Wisconsin Encephalitis, Meningoencephalitis, Acute Disseminated Encephalomyelitis, and Acute Necrotizing Encephalopathy; Cerebellar Ataxia, Transverse Myelitis and Myelopathy, Guillain–Barré Syndrome, Neuritis, and Neuropathy; Rabies Virus

Craig M. Wilson, MD Professor of Pediatrics, Medicine and Epidemiology; Director, Division of Geographic Medicine, University of Alabama at Birmingham, Birmingham, Alabama Plasmodium Species (Malaria); Antiparasitic Agents

A. Clinton White, Jr

Jerry A. Winkelstein, MD

The Paul R. Stalnaker, MD Distinguished Professor of Internal Medicine; Director, Infectious Disease Division, Department of Internal Medicine, University of Texas Medical Branch, Galveston, Texas

Eudowood Professor of Pediatrics, Professor of Medicine and Pathology, Johns Hopkins University School of Medicine; Attending Physician, Department of Pediatrics, Johns Hopkins Hospital, Baltimore, Maryland Infectious Complications of Complement Deficiencies

Marc-Alain Widdowson, MA, MSc, VetMB Medical Epidemiologist, Epidemiology Branch, Division of Viral Diseases, National Center for Immunization and Respiratory Diseases, Centers for Disease Control and Prevention, Atlanta, Georgia Caliciviruses

Kimberly Workowski, MD Chief, Guidelines Unit, Epidemiology and Prevention Branch, Division of Sexually Transmitted Diseases Prevention, National Center for HIV/AIDS, Viral Hepatitis, STD, and TB Prevention, Centers for Disease Control and Prevention; Associate Professor

of Medicine, Division of Infectious Diseases, Emory University, Atlanta, Georgia Skin and Mucous Membrane Infections and Inguinal Lymphadenopathy; Trichomonas vaginalis

Terry W. Wright, PhD Associate Professor of Pediatrics and Microbiology and Immunology, Department of Pediatrics, University of Rochester School of Medicine, Rochester, New York Pneumocystis jirovecii (P. carinii)

Nada Yazigi, MD Assistant Professor of Clinical Pediatrics, University of Cincinnati School of Medicine; Attending Physician, Division of Gastroenterology, Hepatology and Nutrition, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio Granulomatous Hepatitis; Acute Pancreatitis; Cholecystitis and Cholangitis

Ram Yogev, MD Professor of Pediatrics, Northwestern University School of Medicine; Director, Section of Maternal, Adolescent and Pediatric HIV/AIDS, Children’s Memorial Hospital; Director, Experimental Therapeutics Program, Children’s Memorial Research Center, Chicago, Illinois Chronic Meningitis; Recurrent Meningitis; Focal Suppurative Infections of the Central Nervous System

Edward J. Young, MD Professor of Medicine and Molecular Virology and Microbiology, Baylor College of Medicine; Attending Physician, Michael E. DeBakey Veterans Affairs Medical Center, Houston, Texas Brucella Species (Brucellosis)

Theoklis E. Zaoutis, MD, MSCE Assistant Professor of Pediatrics, University of Pennsylvania School of Medicine; Associate Hospital Epidemiologist, Attending Physician, Department of Pediatrics, Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania Healthcare-Associated Infections; Clinical Syndromes of Device-Associated Infections

PA RT

I

Understanding, Controlling & Preventing Infectious Diseases

SECTION

A

Epidemiology & Control of Infectious Diseases CHAPTER

1

Principles of Epidemiology and Public Health Benjamin Schwartz and Stacey W. Martin

Epidemiology is the study of the distribution and determinants of disease or other health-related states or events in specified populations and the application of this study to control of health problems.1 The key component of this definition is the link between epidemiology and populations, which distinguishes it from clinical case studies that focus on individuals. Health events can be characterized by their distribution (descriptive epidemiology) and by factors that influence their occurrence (analytic epidemiology). In both descriptive and analytic epidemiology, healthrelated questions are addressed using quantitative methods to identify patterns or associations from which inferences can be drawn and interventions developed, applied, and assessed.

DESCRIPTIVE EPIDEMIOLOGY Surveillance The goals of descriptive epidemiology are to define the frequency of health-related events and their distribution by person, place, and time. The foundation of descriptive epidemiology is surveillance, or case detection. Retrospective surveillance identifies health events from existing data, such as clinical or laboratory records, hospital discharge data, and death certificates. Prospective surveillance identifies and collects information about cases as they occur, for example, through ongoing laboratory-based reporting. With passive surveillance, case reports are supplied voluntarily by clinicians, laboratories, health departments, or other sources. The completeness and accuracy of passive reporting are affected by whether reporting is legally mandated, whether a definitive diagnosis

can be established, illness severity, interest of the public and the medical community in a condition, and whether a report will elicit a public health response. Because more severe illness is more likely to be diagnosed and reported, the severity and clinical spectrum of passively reported cases are likely to differ from that of all cases of an illness. Passively collected reports of nationally notifiable diseases are tabulated in the Morbidity and Mortality Weekly Report (http:// www.cdc.gov/mmwr/). In active surveillance, effort is made to ascertain all cases of a condition occurring in a defined population. Active case finding can be prospective, involving routine contacts with reporting sources; retrospective, through record audits; or both. Population-based active surveillance, in which all cases in a defined geographic area are identified and reported, provides the most complete and unbiased ascertainment of disease and is optimal for describing the rate of a disease and its clinical spectrum. By contrast, active surveillance conducted at only one or several participating facilities can yield biased information on disease frequency or spectrum based on the representativeness of the patient population and the size of the sample obtained.

Case Definition Establishing a standard case definition is an important first step for surveillance and description of the epidemiology of a disease or health event.2 Formulation of a case definition is particularly important where laboratory diagnostic testing is not definitive. More restrictive case definitions minimize misclassification but may exclude true cases and are most useful when investigating a newly recognized condition, in which the ability to determine etiology, pathogenesis, or risk factors is decreased by inclusion of noncases in the study population. A more inclusive definition might be important in an outbreak setting where cases are being detected for further investigation or where preventive interventions can be applied. Multiple research or public health objectives can be addressed by developing a tiered case definition that incorporates varying degrees of diagnostic certainty for definite and probable cases.

Sensitivity, Specificity, and Predictive Value Sensitivity, specificity, and predictive values can be used to quantify the performance of a case definition or the results of a diagnostic test 1

2

SECTION

A Epidemiology & Control of Infectious Diseases

or algorithm (Table 1-1). Unlike sensitivity and specificity, predictive values vary with the prevalence of a condition within a population. Even with a highly specific diagnostic test, if a disease is uncommon among those tested, a large proportion of positive test results will be false positives, and the positive predictive value will be low (Table 1-2). If the test is applied more selectively, where the proportion of people tested who truly have disease is greater, the test’s predictive value will be improved. Thus, sensitivity and specificity are characteristics of the test, whereas predictive values depend on the disease prevalence in the population in which the test is applied. Often, the sensitivity and specificity of a test are inversely related. Selecting the optimal balance of sensitivity and specificity depends on the purpose for which the test is used: a screening test should be highly sensitive, whereas a follow-up confirmatory test should be highly specific.

Incidence and Prevalence Characterizing disease frequency is one of the most important aspects of descriptive epidemiology. Frequency measures typically include a count of new or existing cases of disease as the numerator and a quantification of the population at risk as the denominator. Cumulative incidence is expressed as a proportion and describes the number of new cases of an illness occurring in a fixed at-risk population over a specified period of time, generally 1 year, unless otherwise stated. Incidence density is defined as the rate of new cases of disease in a dynamic at-risk population; for this measure the denominator is typically expressed as the population-time at-risk (e.g., person-time). Because the occurrence of many infections varies with season, extrapolating annual incidence from cases detected during a short observation period can be inaccurate. In describing the risk of acquiring illness during a disease outbreak, the attack rate, defined as the number of new cases of disease occurring in a specified population and time period, is a useful measure.

Prevalence refers to the proportion of the population having a condition at a specific point in time. As such, it is a better measure of disease burden for chronic conditions than is incidence or attack rate, which identify only new (incident) cases. Prevalent cases of disease can be ascertained in a cross-sectional survey, whereas determining incidence requires longitudinal surveillance. When disease prevalence is low and incidence and duration are stable, prevalence is a function of disease incidence multiplied by its average duration.

Describing Illness by Person, Place, and Time Characterizing disease by person, place, and time is often useful. Demographic variables, including age, sex, socioeconomic status, and race or ethnicity are often associated with the risk of disease. Describing a disease by place can help define risk groups; for example, when an illness is caused by an environmental exposure or is vector-borne, or in an outbreak with a point source exposure. Time, also, is a useful descriptor of disease occurrence. Evaluating long-term (secular) trends provides information that can be used by policymakers or clinicians to identify emerging health problems or to assess the impact of prevention programs. The timing of illness in outbreaks

TABLE 1-2. Positive and Negative Predictive Values for a Hypothetical Diagnostic Test Having a Sensitivity of 90% and Specificity of 90% Proportion with Condition

Positive Predictive Value

Negative Predictive Value

1%

8%

> 99%

10%

50%

99%

20%

69%

97%

50%

90%

90%

TABLE 1-1. Definitions and Formulae for the Calculation of Important Epidemiologic Parameters Measures of test accuracy

Sensitivity: Proportion of true positives (diseased) with a positive test

A/(A + C)

Specificity: Proportion of true negatives (nondiseased) with a negative test

D/(B + D)

Positive predictive value (PPV): Proportion of positive tests that are true positives

A/(A + B)

Negative predictive value (NPV): Proportion of negative tests that are true negatives

D/(C + D)

Measures of data dispersion Variance: Statistic describing variability of individuals in and precision a population Standard error: Statistic describing the variability of sample-based point estimates (Po) around the true population value being estimated

[1/(n – 1)][(x1 – x–)2 + (x2 – x–)2 + … + (xn – –x)2] Variance

Confidence interval: A range of values that is believed to contain the true value within a defined level of certainty (usually 95%, as shown) Measures of association and risk

Relative risk or risk ratio (RR): Risk (probability) of a health event in those with a given exposure divided by the risk in those without the exposure

A/(A + B) C/(C + D)

Odds ratio (OR): Odds of a given exposure among those with a health event divided by odds of exposure among those without the health event

AD/BC

Population attributable fraction: The proportion of disease [Pe (RR – 1)]/[1 + Pe (RR – 1)] in a population due to the specific exposure (Proportion exposed, Pe = (A + B)/(A + B + C + D)) PART I Understanding, Controlling, and Preventing Infectious Diseases

Principles of Epidemiology and Public Health

can be displayed in an epidemic curve and can be useful in defining the mode of transmission of an infection or its incubation period and in assessing the effectiveness of control measures.

ANALYTIC EPIDEMIOLOGY Study Design The goal of analytic studies is to identify predictors of an outcome. This goal can be addressed in experimental or epidemiologic (observational) studies. Ecological or trend studies can also be used to assess predictors when the frequency or distribution of an outcome has changed over time or differs between populations. In contrast to experimental or observational studies which analyze information about individuals, ecological studies draw inferences from data on a population level. Inferences from ecological studies must be made with caution because populations differ in multiple ways and because relationships observed for groups do not necessarily apply to individuals (known as the “ecological fallacy”). Because of these drawbacks, ecological studies are best suited to generating hypotheses that can be tested using other study methods. In experimental studies, hypotheses are tested by systematically allocating an exposure of interest to subjects in separate groups to achieve the desired comparison. Such studies include randomized, controlled, double-blinded treatment trials, and laboratory experiments. By carefully controlling study variables, investigators can restrict differences among groups, thereby increasing the likelihood that the observed differences are a consequence of the specific factor being studied. Because experiments are prospective, the temporal sequence of exposure and outcome can be clearly established, making it easy to define cause and effect. By contrast, epidemiologic studies test hypotheses using observational methods to assess exposures and outcomes among individuals in populations and identifying statistical associations from which inferences regarding causation are drawn. Although observational studies cannot be controlled to the same degree as experiments, they

CHAPTER

1

are practical in circumstances in which exposures or behaviors cannot be assigned. Moreover, the results are more generalizable to a real population having a wide range of attributes. The three basic types of observational studies are cohort studies, cross-sectional studies, and case-control studies (Table 1-3). Hybrid study designs, incorporating components of these three, have also been developed.3 In planning observational studies care must be taken in the selection of participants to minimize the possibility of bias. Selection bias results when study subjects have differing probabilities of being selected that are related to the risk factors or outcomes under evaluation.

Cohort Studies In a cohort study, subjects are categorized based on their exposure to a suspected risk factor and are observed for the development of disease. Associations between exposure and disease are expressed by the relative risk of disease in exposed and unexposed groups (see Table 1-1). Cohort studies are typically prospective, with exposure being defined before disease occurs. In retrospective cohort studies, in which the cohort is selected after the outcome has occurred, exposures are determined from existing records that preceded the outcome, and, thus, the directionality of the exposure–disease relationship is still forward. By characterizing exposures before development of disease, selection bias is minimized and inference of cause and effect is easier. A major disadvantage of cohort studies is that they are impractical for studying rare diseases or conditions with a long latent period between exposure and the onset of clinical illness. Moreover, whereas cohort studies can assess multiple potential outcomes resulting from an exposure, it is difficult to investigate multiple exposures as risk factors for a single outcome. In general, cohort studies are unsuited for investigating risk factors for new or rare diseases or for generating new hypotheses about possible exposure–disease relationships. Cohort studies not only provide data on whether an outcome occurs but, for those experiencing the outcome, on when it occurs. Analysis of time-to-event data for outcomes such as death or illness is a powerful approach to assess or compare the impacts of preventive or

TABLE 1-3. Types of Observational Studies and Their Advantages and Disadvantages Type of Study

Design, Characteristics

Advantages

Disadvantages

Cohort

Prospective or retrospective

Ideal for outbreak investigations in defined populations

Unsuited for rare diseases or those with long latency

Select study group

Prospective design ensures that exposure preceded disease

Expensive

Observe for exposures and disease

Selection of study group is unbiased by knowledge of disease status

May require long follow-up periods

Calculate relative risk (RR) or hazard ratio (HR) of disease given exposure

RR and HR accurately describe risk given exposure

Difficult to investigate multiple exposures

Nondirectional

Rapid, easy to perform, and inexpensive

Timing of exposure and disease may be difficult to determine

Select study group

Ideal to determine knowledge, attitudes, and behaviors

Biases may affect recall of past exposures

Retrospective

Rapid, easy to perform, and inexpensive

Timing of exposure and disease may be difficult to determine

Identify cases with disease

Ideal for studying rare diseases, those with long latency, new diseases

Biases can occur in selecting cases and controls and determining exposures

Cross-sectional

Determine exposure and disease status Calculate RR for disease given exposure Case-control

Identify controls without disease Determine exposures in cases and controls Calculate odds ratio (OR) for an exposure given disease

3

OR only provides an estimate of the RR if disease is rare

4

SECTION

A Epidemiology & Control of Infectious Diseases

0.8

0.4

0.6

0.3

Cumulative hazard

Probability of survival

1.0

0.4

0.2

0.0 0

7

14

21

28

35

0.2

0.1

0.0 0

10

Days post blood culture

A

Combination therapy (n = 47) Monotherapy (n = 47)

20

30

40

50

60

Age (months)

B

HBV PncCRM

Figure 1-1. Example of Kaplan–Meier and cumulative hazard curves. (A) Survival plot for critically ill patients with Streptococcus pneumoniae bacteremia treated with monotherapy or combination therapy. (Redrawn from Baddour LM, Yu VL, Klugman KP, et al. Combination antibiotic therapy lowers mortality among severely ill patients with pneumococcal bacteremia. Am J Respir Crit Care Med 2004;170:440–444.) (B) Cumulative hazard of tympanostomy tube placement from 2 months until 4–5 years of age in children who received pneumococcal conjugate vaccine (PncCRM) or a control vaccine (HBV). (Redrawn from Palmu AAI, Verho J, Jokinen J, et al. The seven-valent pneumococcal conjugate vaccine reduced tympanostomy tube placement in children. Pediatr Infect Dis J 2004;23:732–738.)

therapeutic interventions. The probability of remaining event-free over time can be expressed in a Kaplan–Meier survival curve where, initially, the event-free probability is 1 and declines in a step-function as the outcomes of interest occur (Figure 1-1A). Time-to-event data can also be displayed as the cumulative hazard of an event occurring among members of a cohort, increasing from 0 at enrollment (Figure 1-1B). These two approaches are related in that the hazard reflects the incident event rate while survival reflects the cumulative nonoccurrence of that outcome.4,5 With time-to-event analysis, the association between exposure and disease is often expressed as a hazard ratio. Like a relative risk, the hazard ratio is a comparative measure of risk between exposed and unexposed groups, the primary difference being the hazard ratio compares event experience over the entire time period, whereas the relative risk compares event occurrence only at the study endpoint.6

Cross-Sectional Studies In a cross-sectional study, or survey, a sample is selected and at a single point in time exposures and outcome are determined. Outcomes may include disease status or behaviors and beliefs. Unlike cohort studies, multiple exposures can be evaluated as explanations for the outcome. Associations are characterized by the relative risk, as in cohort studies. Because neither exposures nor outcomes are used in selection of the study group, their prevalence is an estimate of that in the overall population from which the sample was drawn. National survey data characterizing health status, behaviors, and medical care are available from the National Center for Health Statistics (http://www.cdc.gov/nchs/express.htm).

Case-Control Studies In a case-control study, the investigator identifies a group of people with a disease or outcome of interest (cases) and compares their exposures with those in a selected group of people who do not have disease (controls). Differences between the groups are expressed by an odds ratio, which compares the odds of an exposure in case and

control groups (see Table 1-1). Case-control studies are retrospective in that disease status is known and serves as the basis for selecting the two comparison groups; exposures are then determined by reviewing available records or by interview. A major advantage of case-control studies is their efficiency in studying uncommon diseases or those with a long latency. Casecontrol studies can also evaluate multiple exposures that may contribute to a single outcome. They tend to be less costly and more time-efficient than cohort studies because study subjects can be identified from existing sources (such as hospital or laboratory records, disease registries, or surveillance reports) and, after identification of suitable control subjects, data on prior exposures can be collected rapidly. Case-control studies also have several drawbacks: bias can be introduced during selection of cases and controls and in retrospectively determining exposures, and inferring causation from statistically significant associations can be complicated by difficulty in determining the temporal sequence of exposure and disease in a retrospective study.

Causal Inference and the Impact of Bias Care must be taken in the design of all analytic studies, whether experimental or observational; however, concerns about validity and the impact of potential biases are particularly important for observational studies. The validity of a study is the degree to which inferences drawn from a study are warranted. Internal validity refers to the correctness of study conclusions for the population from which the study sample was drawn, whereas external validity refers to the extent to which the study results can be generalized beyond the population sampled. The validity of a study can be affected by bias, or systematic error, in selecting the study participants (sampling), in ascertaining their exposures, or in analyzing and interpreting study data. For errors to result in bias, they must be systematic, or directional. Nonsystematic error (random misclassification) decreases the ability of a study to identify a true association but does not result in detection of a spurious association.

PART I Understanding, Controlling, and Preventing Infectious Diseases

Principles of Epidemiology and Public Health

BOX 1-1. Potential Sources of Bias in Observational Studiesa BIAS IN CASE ASCERTAINMENT AND CASE/CONTROL SELECTION Surveillance bias: differential surveillance or reporting for exposed and unexposed Diagnosis bias: differential use of diagnostic tests in exposed and unexposed Referral bias: differential admission to hospital based on an exposure or a variable associated with exposure Selection bias: differential sampling of cases based on an exposure or a variable associated with exposure Nonresponse bias: differential outcome or exposures of responders and nonresponders Survival bias: differential exposures between those who survive to be included in a study and those who die following an illness Misclassification bias: systematic error in classification of disease status BIAS IN ESTIMATION OF EXPOSURE Recall bias: differential recall of exposures based on disease status Interviewer bias: differential ascertainment of exposures based on disease status Misclassification bias: systematic errors in measurement or classification of exposure a

A more complete listing is provided by Sackett.7

Several sources of bias can occur in selection of study participants (Box 1-1). Diagnosis bias results when persons with a given exposure are more likely to be diagnosed as having disease than are people without the exposure; this can occur because diagnostic testing is more likely to be done or because the interpretation of a test may be affected by knowledge of exposure status. For hospital-based studies, differential referral can also bias selection of a study sample. This bias would occur if, for example, the frequency of an exposure varied with socioeconomic status and a hospital predominantly admitted persons from either a high- or low-income group. Bias can also occur when eligible subjects refuse to participate in a study as cases or controls. However, nonresponse, even at high rates, does not result in bias if responders and nonresponders are similar with respect to the exposures of interest. Determining exposures can be affected by several types of bias. Recall of exposures may be different for persons who have had an illness compared with people who were well. This bias occurs in either direction: cases may be more likely to remember an exposure that they associate with their illness (e.g., what was eaten before an episode of diarrhea) or less likely to recall an exposure if a severe illness affected memory. Interviewers can introduce bias by questioning cases and controls differently about their exposures. Misclassification of exposures can result from errors in measurement such as might occur with the use of an inaccurate laboratory test; these errors are often random rather than systematic. Even a carefully designed study that minimizes potential biases can make erroneous causal inferences. An exposure can appear falsely to be associated with disease because it is closely linked to the true, but undetermined, risk factor. Race is often found to be associated with the risk of a disease, whereas the true risk factor may be an unmeasured variable that is linked to race, such as socioeconomic status. The risk of making incorrect inferences can be minimized by considering certain general criteria for establishing causation. These include the strength of an association, the presence of a dose–response effect, a clear temporal sequence of exposure to disease, the consistency of findings with those of other studies, and the biologic plausibility of the hypothesis.8

Statistical Analysis Characteristics of Populations and Samples Whereas epidemiologic analysis seeks to make valid conclusions about populations, the entire population is rarely included in a study.

CHAPTER

1

5

An assumption underlying statistical analysis is that the sample evaluated was selected randomly from the population. Often, this criterion is not met and calls into question the appropriateness and interpretation of statistical analyses. The mean, median, and mode describe a central value for samples and populations. The arithmetic mean is the average, determined by summing individual values and dividing by the sample size. When data are not normally distributed (skewed), calculation of a geometric mean can limit the impact of outlying values. The geometric mean is calculated by taking the nth root of the product of all the individual values, where n is the total number of individual values. For example, immunogenicity of vaccines is usually expressed by the geometric mean titer. The median, or middle value, is another way to describe nonnormally distributed data. The mode, or most commonly occurring value in a sample, is rarely used. Several measures can be used to describe the variability in a sample. The range describes the difference between the highest and lowest value, whereas the interquartile range defines the difference between the 25th and 75th percentiles. Variation among individuals is most often characterized by the variance or standard deviation. The variance is defined as the mean of the squared deviation of each observation from the sample’s mean. The standard deviation is the square root of the variance (see Table 1-1). For a normally distributed population, 68% of values fall within 1 standard deviation of the mean and 95% of values within 1.96 standard deviations. The standard error represents a type of standard deviation and is used to describe the standard deviation of an estimate (e.g., mean, odds ratio, relative risk). When analyzing a sample, the mean or other statistics describing the sample represent a point estimate of that parameter for the entire population. If another random sample were drawn from the same population, the point estimate for the parameter of interest would likely be different, depending on the variability in the population and the sample size selected. A confidence interval defines a range of values that includes the true population value, within a defined level of certainty. Most often, the 95% confidence interval is presented (see Table 1-1).

Measures of Association and Statistical Significance In a cohort study or survey, the relative risk or risk ratio compares the risk of disease for those with versus those without an exposure (see Table 1-1). In case-control studies, association is assessed by the odds ratio, which compares the odds of exposure among those with and without a disease or health outcome; when disease is uncommon (< 10%) in both exposed and unexposed groups, the odds ratio approximates the relative risk. For time-to-event analyses, the comparative risk is expressed as the hazard ratio. Odds ratios, relative risks, and hazard ratios greater than 1 signify increased risk given exposure and values less than 1 suggest that exposure decreases the risk of an outcome. Because observational studies generally do not include all members of a population, the odds ratios, relative risks, and hazard ratios represent an estimate of the true value within the entire population. Statistical analyses can help guide investigators in making causal inferences based on a point estimate of these measures of association.

Statistical Significance Statistical tests are applied to assess the likelihood that the study results were obtained by chance alone rather than representing a true difference within the population. Most investigators consider a Pvalue < 0.05 as being statistically significant, indicating a less than 5% risk that the observed association is the result of chance alone (designated a type I error, the probability of which is the alpha level). Although use of this cutoff for significance testing has become conventional, ignoring higher P-values can cause the researcher to miss a real and important association, whereas blind faith in the significance of lower P-values can lead to erroneous conclusions. Statistical testing should contribute to, but not replace, criteria for

6

SECTION

A Epidemiology & Control of Infectious Diseases

evaluating possible causation. Statistical significance can also be defined based on 95% confidence intervals, which approximately correspond to a P-value of 0.05. An odds ratio, relative risk, or hazard ratio is considered statistically significant if the 95% confidence interval does not include 1. An advantage of using confidence intervals to define statistical significance is that they provide information on whether a finding is statistically significant and on the possible range of values for the point estimate in the population, with 95% certainty. Another pitfall in interpreting statistical significance is ignoring the magnitude of an effect in favor of its “significance.” A very large study can identify small, perhaps trivial, differences between study groups as significant. Some epidemiologists have proposed that, despite statistical significance, odds ratios less than 2 or 3 in an observational study should not be interpreted because unidentified bias or confounding could have accounted for a difference of this magnitude.9 Conversely, the relative risk or odds ratio associating an exposure and outcome may be large but, if the exposure is uncommon in both groups, will not explain most cases of illness. The public health importance of an exposure can be described by the populationattributable fraction, or the proportion of the disease in a population that is related to the exposure of interest. Attributable risk can also be calculated among those with a given exposure, quantifying the proportion of their risk of disease due to that specific exposure.

Sample Size One reason that a study may fail to identify a true risk factor as statistically significant is that the sample size was too small (designated a type II error, the probability of which is the beta level). Statistical power is defined as 1 – b and is the complement of type II error, that is, the probability of correctly identifying a difference of specified size between groups, if such a difference truly exists. The problem of inadequate sample size in clinical studies was highlighted in an analysis of “negative” randomized controlled trials reported in three leading medical journals between 1975 and 1990. Of 70 reports, only 16% and 36% had sufficient statistical power (80%) to detect a true 25% or 50% relative difference, respectively, between treatment groups.10 In calculating sample sizes for testing hypotheses, investigators must select type I and type II errors and define the magnitude of the difference that is deemed clinically important. Often, the type II error is set at 0.2, indicating acceptance of a 20% likelihood that a true difference exists but would not be identified by the study. Sample size calculations can be performed using a range of computer software. The program Epi-Info can be used to perform sample size calculations as well as other statistical functions and is available at no charge from the Centers for Disease Control and Prevention (www.cdc.gov/ epiinfo/). Ensuring an adequate sample size is particularly important for studies attempting to prove equivalence or noninferiority of a new treatment compared with standard therapy. Food and Drug Administration guidance recommends that noninferiority trials adopt a null hypothesis that a difference exists between treatments; this hypothesis is rejected if the lower 95% confidence limit for the new treatment is within a specified margin of the point estimate for standard therapy. Because the null hypotheses can never be proven or accepted, the failure to reject a null hypothesis of no difference between treatments or exposure does not prove equivalence. The importance of this distinction is illustrated by an analysis of 25 studies claiming equivalence of therapies for pediatric bacterial meningitis. Twenty-three studies claimed equivalence based on a failure to detect a significant difference between treatment groups. However, only 3 of these trials could exclude a 10% difference in mortality, potentially missing a clinically significant difference.11 In some situations, an investigator would want to detect a significant difference among study groups as soon as possible, for example, when a therapeutic or preventive intervention could be applied once a risk group is identified or when there are concerns about the safety of a

drug or vaccine. One approach to this situation is to include in the study design interim analyses after a specified number of subjects are evaluated. Because the likelihood of identifying chance differences as significant increases with the number of analyses, it is recommended that the threshold for defining statistical significance should become more stringent as the number of planned analyses increases.12 If each interim analysis can lead to stopping the trial, this study design is considered a group sequential method.13 Another example of a group sequential design is when concordance or discordance in outcome is tabulated for each matched set exposed to alternate treatments. Results for each set are plotted on a graph and data collection continues until a preset threshold for a significant difference between study groups is crossed or no significant difference is detected at a given power.12

Statistical Inference Statistical testing is used to determine the significance of differences between study groups, and, thus, it provides guidance on whether to accept or reject the null hypothesis. Although providing details of specific statistical tests is outside the scope of this chapter, Table 1-4 provides examples of statistical tests that can be applied in analyzing different types of exposure and outcome variables. Using appropriate analytic and statistical methods is important in identifying significant predictors of an outcome (i.e., risk factors) correctly. Confounding variables are associated with the disease of interest and with other exposure variables that are associated with the disease and are not part of the causal pathway (Figure 1-2). For example, in a study evaluating the link between childcare attendance and pharyngeal carriage of penicillin-resistant pneumococci, recent antimicrobial use would be a confounder because it is associated with both childcare attendance and carriage of resistant pneumococci. Failure to adjust for recent antimicrobial use as a confounding variable would result in overestimating the relationship between childcare and resistant pneumococcal carriage. Effect modifiers interact with risk factors to affect their impact on outcome but may or may not themselves affect outcome. Frequently, age is an effect modifier, with an exposure associated significantly with an outcome in one age group but not in another. There are several approaches to control for confounding variables and effect modifiers. In study design, an extraneous variable can be controlled for by randomization, restricting sampling to one category of the variable or by frequency-matching to obtain similar proportions of cases and controls in each stratum. A more extreme form of

TABLE 1-4. Types of Statistical Tests Used to Evaluate the Significance of Associations Among Categorical and Continuous Variables Dependent Variable (Disease, Outcome) Independent Variable (Exposure, Risk Factor)

Categorical and Dichotomous

Continuous

Chi-squared test

Student-test (parametric)

Fisher exact test

Wilcoxon rank sum test (nonparametric)

Chi-squared test

Analysis of variance (parametric)

Categorical Dichotomous

> 2 categories

Kruskal–Wallis test (nonparametric) Continuous

Logistic regression

PART I Understanding, Controlling, and Preventing Infectious Diseases

Linear regression Correlation (Pearson: parametric; Spearman: nonparametric)

Principles of Epidemiology and Public Health F

D Figure 1-2. Path diagram illustrating association between a confounding variable (C), another risk factor (F), and the disease outcome (D).

C

matching is to select control subjects who are similar to individual cases for extraneous variables (e.g., age, sex, underlying disease) and to analyze whether exposures are concordant or discordant within matched sets. A newer approach to study design is the casecrossover14 or case series15 analysis. In this method, the investigator compares exposures occurring in a defined risk period before the outcome with exposures occurring outside the risk window. This approach has been adapted to the study of adverse events after vaccination. If the vaccine causes the event, the rate of the event would be greater within a defined risk window than would be predicted by chance alone based on the expected distribution of the event.16 The strength of this approach is that each case serves as his or her own control, decreasing confounding. At the analysis stage, the impact of confounding variables and effect modifiers can be limited by performing a stratified analysis or using a multivariable model. In a stratified analysis, the possible association between a risk factor and an outcome is determined separately within different categories, or strata, of the extraneous variable. Stratum-specific estimates can be combined into a single estimate using an appropriate statistical test (e.g., a Mantel–Haentzel odds ratio). If a stratification variable is an effect modifier, the relative risk or odds ratio will differ substantially between the strata; for example, an exposure may be a strong risk factor in one age group but not another. In this setting, a summary statistic should not be presented and results for each stratum should be presented separately. When the extraneous variable is a confounder, an unstratified analysis can identify an exposure as a significant risk factor; when the analysis is stratified by the confounder, however, the apparent association with outcome is abolished in each stratum, indicating that the exposure is not an independent risk factor for disease. Because stratified analyses rapidly become confusing as the number of strata increases, techniques of mathematical modeling have been developed that permit the simultaneous control of multiple variables. Significant risk factors determined in a multivariable model are interpreted as each contributing independently and significantly to the outcome, thus controlling for confounding. Effect modification can be taken into account by including terms expressing the interaction between a risk factor and effect modifier in the model. Various multivariable models are appropriate for discrete, continuous, and time-dependent outcomes. A limitation of multivariable modeling is multicollinearity, which occurs when two or more explanatory variables of interest are highly correlated, and can result in inaccurate measures of association and decreased statistical power. The risk of multicollinearity can be reduced by assessing correlations between potential risk factors and selecting which variables to include in the model. Various methods to identify and minimize multicollinearity have been developed.17

VACCINE EFFICACY STUDIES With advances in vaccine development and the licensure of new vaccines, the ability to interpret results of vaccine efficacy studies becomes increasingly important. Most prelicensure efficacy studies are experimental randomized, double-blind, controlled trials in which vaccine efficacy (VE) is calculated by comparing the attack rates (AR) for disease in the vaccinated and unvaccinated groups (VE (%) = ((AR

CHAPTER

1

7

unvaccinated – AR vaccinated)/AR unvaccinated) μ 100; or (1 – RR) μ 100). After licensure, conducting controlled studies, which requires withholding vaccine from a control group, is no longer ethical. Therefore, further studies of efficacy must be observational rather than experimental, comparing persons who have chosen to be immunized with those who have not. In case-control efficacy studies, investigators identify persons with disease and compare their vaccination status with the vaccination status of healthy controls. The number of vaccinated and unvaccinated cases and controls is included in a twoby-two table, and vaccine efficacy is calculated as 1 minus the odds ratio (VE (%) = (1 – OR) μ 100). When the proportion of cases who have been vaccinated is less than the proportion of vaccinated controls, the odds ratio is < 1 and the point estimate for efficacy indicates that immunization is protective. The precision of the estimate is expressed by the 95% confidence interval. A lower 95% confidence limit that is greater than 0% indicates statistically significant protection; often investigators set power of vaccine efficacy studies so that the lower confidence limit is much greater than zero and consistent with meaningful levels of protection. The most important component of a case-control efficacy study is selecting controls who have the same opportunity for immunization as do cases. If cases had less opportunity to be immunized, results will be biased toward showing protection. Factors such as low socioeconomic status, which may increase the risk of disease and decrease the chance of being immunized, are potential confounding variables and can be controlled for by matching controls to cases for those factors. Cohort studies can also be used to determine vaccine efficacy after licensure. A study design called the indirect-cohort method was developed by researchers at the Centers for Disease Control and Prevention to evaluate the efficacy of the pneumococcal polysaccharide vaccine using data collected by disease surveillance.18 The study cohort included persons identified with invasive pneumococcal infections. The study hypothesis was that, if vaccine was protective, the proportion of vaccinated persons infected with pneumococcal serotypes that are included in the vaccine formulation would be less than the proportion of unvaccinated persons infected with vaccine-type strains. Vaccine efficacy was calculated from the relative serotype distributions overall and for each individual serotype. In a study of pneumococcal polysaccharide vaccine efficacy that utilized this approach, the point estimate of efficacy for preventing invasive infection was 57% (95% confidence interval from 45% to 66%);19 this estimate is similar to that obtained in a case-control efficacy study.20

DISEASE CONTROL AND PUBLIC HEALTH POLICY Outbreak Investigations Outbreak investigations require knowledge of disease transmission and use of descriptive and analytic epidemiologic tools. Possible outbreaks can be identified from surveillance data showing an increased rate of an infection or an unusual clustering in person, place, and time. Comparing the incidence rate of disease with a baseline rate from a previous period is helpful in validating the occurrence of an outbreak. Other explanations for changes in the apparent rate of disease occurrence, such as diagnostic error, seasonal variations, and changes in reporting, must be considered. Using sensitive molecular methods to assess similarity between isolates from cases may be helpful in documenting that an apparent cluster of cases represents an outbreak, because most outbreaks are caused by a single strain. After establishing the presence of an outbreak, the next steps of an investigation are to develop a case definition, identify cases, and characterize the descriptive epidemiology. An epidemic curve depicts number of cases over time and can provide information on possible transmission. In an outbreak with a point source exposure, an index case may be identified, with other cases occurring after an incubation period or at multiples of an incubation period (Figures 1-3 and 1-4). Plotting the location of cases on a spot map may be helpful in

8

SECTION

A Epidemiology & Control of Infectious Diseases Figure 1-3. Example of an epidemic curve for a common source outbreak with continuous exposure. Cases of hepatitis A by date of onset, Fayetteville, Arkansas, November to December 1979. (Centers for Disease Control and Prevention, unpublished data.)

Case Food handler case Number of cases

Presumed index case

Secondary case

10

5

1

7

13

19

25

31

6

12

October

18

24

30

6

12

November

18

24

December

Date of onset

20 Bloody cases Boil water order

Nonbloody cases Number of cases

15

10

Water main breaks

Chlorination

5

0 15

17

19

21

23

25

27

29

31

2

4

6

8

10

December

12

14

16

18

20

January Date of onset

Figure 1-4. Example of an epidemic curve for a common source outbreak with continuous exposure. Cases of diarrheal illness in city residents by date of onset and character of stool, Cabool, Missouri, December 1989 to January 1990. (Centers for Disease Control and Prevention, unpublished data.)

determining possible exposures. Describing host characteristics can be important in identifying at-risk populations for further investigation or targeting control measures, and for developing hypotheses that can be investigated in an analytic study. Cohort studies are optimal for investigating outbreaks that occur in small, well-defined populations. These include outbreaks in school or childcare, social gatherings, and hospitals. In populations that are not well defined, a case-control study is the most feasible approach. It is important to select controls who had an opportunity equal to that of cases for exposure to potential risk factors and developing disease. Neighbors of cases or patients from the same medical-provider practice or hospital as the case are commonly selected as controls. Friends of cases have been used as controls in some investigations, but concerns have been raised about possible “overmatching,” because friends may be more likely than others to have similar exposures. When the number of cases is relatively small, enrolling multiple controls per case increases the power of the study to find significant risk factors. After a standard questionnaire is administered, significant risk factors are determined by comparing exposures of cases and controls. The results of analyses may lead to inferences of causation and development of prevention and control strategies or to further hypotheses that can be evaluated later. The impact of intervention can

be determined by ongoing surveillance and continuing to plot additional cases on the epidemic curve (Figure 1-4).

Impact and Economic Analysis of Disease Prevention Assessing health and economic impacts of public health interventions is important in developing or supporting policy decisions. Health impacts can be expressed directly as cases of disease, deaths, and sequelae prevented. Vaccine efficacy is a specific example of the prevented fraction (PF), where PF = P (1 – RR), with P representing the proportion exposed to an intervention. Secondary measures of health impact include years of potential life lost (YPLLs) or qualityadjusted life-years (QALYs) lost, which quantify the impact of death, or death and disability, respectively, based on the age at which these events occur.21 A measure of the efficiency of an intervention is the number needed to treat (NNT), which describes how many persons must be exposed to an intervention to avoid one case of adverse health outcome; for example, the number of persons who would need to be immunized to prevent a case or a death, or to be given antibiotic prophylaxis to prevent one case of infection. Because public health resources are limited, it is becoming increasingly important to calculate impact in economic as well as in

PART I Understanding, Controlling, and Preventing Infectious Diseases

Pediatric Infection Prevention and Control

health terms and to compare an intervention with other potential uses of available resources. Cost-effectiveness analyses determine the cost per health outcome achieved, such as the cost per death or complication averted. In a cost-effectiveness formula, costs appear in the numerator and health benefits appear in the denominator. The numerator includes expenditures for the prevention program from which cost savings occurring with disease prevention are subtracted. In addition to direct costs averted (e.g., savings from decreased medical care), indirect costs savings occur from increased productivity of people who do not become ill or miss time from work while receiving care or caring for ill family members. Cost–utility calculations are similar to cost-effectiveness but assess cost per QALY saved or YPLL averted. Cost–benefit analyses differ from cost-effectiveness analyses in that the calculation is made entirely in economic terms. Health benefits are assigned an economic value and expenditures are compared with savings. One problem with this approach lies in the difficulty assigning an economic value to a health effect. For example, the value of a life saved may be quantified as the estimated value of a person’s earnings over his or her lifetime, foregone earnings due to premature death, or by a standard amount; both economic and ethical issues may be raised by the choice of approach. Because the parameters used in economic analyses are often uncertain or based on limited data, and because choices made by the investigator (e.g., regarding the value of life) may be influential to the analysis, sensitivity analyses are often performed where parameters are varied across a range of potential values. In addition to defining a range of possible economic outcomes, sensitivity analyses can identify the factors that most influence the results, elucidating where further studies may be important.

EVALUATING THE MEDICAL LITERATURE Basic epidemiologic knowledge is important not only in performing research but also in evaluating studies reported in the medical literature. Steps in reviewing published medical research are shown in Box 1-2. The ability to assess published studies carefully is often limited by the information presented in the report. To improve reporting of randomized controlled trials, a group of investigators and editors developed a Consolidated Standards of Reporting Trials (CONSORT)23 and later extended these recommendations to reporting of noninferiority and equivalence randomized trials.24 Although these standards have been adopted by many journals and editorial groups, reporting often does not adhere to the quality standards proposed.25, 26 Although the guidelines refer to experimental rather than observational studies, most criteria still apply. Assessing the research hypothesis allows readers to determine the relevance of the study to their practice and to judge whether the

2

9

analyses were done to test the hypothesis or to identify other interesting associations. The ability to make causal inferences from a confirmatory study that tests a single hypothesis is greater than from an exploratory study in which multiple exposures are considered as potential explanations for an outcome. Several components of study design are important to consider. Details should be presented regarding the criteria for selecting a cohort or cases and controls. Exposure and outcome variables should be clearly defined, and the potential for misclassification and its impact should be considered. Quantifying exposure may be important to establish a dose–response relationship. Finally, sample size estimates should be presented, making clear the magnitude of difference between study groups considered clinically meaningful and the type I and type II error levels. In the analysis, it is important that outcomes for all study subjects are reported, even if that outcome is “lost to follow-up.” Intent-to-treat analyses consider outcomes for all enrolled subjects, whether or not they completed the therapy (e.g., those who were nonadherent with therapy or who received only part of a vaccination series). The appropriateness of the statistical tests should be assessed; for example, if data are not normally distributed, they can be transformed to a scale that is more normally distributed (e.g., geometric mean titers) or nonparametric statistical tests should be used. In assessing a multivariable model, the reader should critically evaluate the type of model chosen, the variables included, and whether interaction terms were considered. Missing data pose a particular problem in modeling, in that study subjects can only be included if data are available for each variable in the model; thus, the power of a multivariate model may be much less than that predicted in a sample size calculation. Bias can have an important impact on study results and must be carefully considered. Approaches to minimize bias should be clearly described. The direction and potential magnitude of remaining bias should be estimated and its impact on results considered. Potential confounding, the presence of important unmeasured variables, and possible effect modification can have a major impact on the results. Investigators should openly discuss the potential limitations of the investigation and describe the strategies they applied to overcome those limitations. Finally, interpretation of study results includes assessing the magnitude of the associations, their relevance to practice, and the likelihood that the relationships observed are causal. The importance of an exposure in explaining an outcome can be expressed by the attributable proportion. The external validity of the results, however, and the potential impact on one’s own patient population must still be assessed.

ACKNOWLEDGMENT Thanks to Paul M. Gargiullo, PhD, for assistance.

BOX 1-2. Steps in the Critical Evaluation of Epidemiologic Literature24 1. Consider the research hypothesis 2. Consider the study design • Type of study • Selection of study participants • Selection and definition of outcome variables • Selection and definition of exposure (predictor) variables • Sample size and power 3. Consider the analysis • Complete accounting of study subjects and outcomes • Appropriateness of statistical tests • Potential sources and impact of bias • Potential impact of confounding and effect modification 4. Consider the interpretation of results • Magnitude and importance of associations • Study limitations • Ability to make causal inferences

CHAPTER

CHAPTER

2

Pediatric Infection Prevention and Control Jane D. Siegel and Leigh Grossman Developing effective prevention strategies for healthcare-associated infections (HAIs) in pediatric patients is a unique science that requires consideration of various host factors, sources of infection, routes of transmission, behaviors associated with care of infants and children, pathogens and their virulence factors, treatments, preventive therapies,

10

SECTION

A Epidemiology & Control of Infectious Diseases

and behavioral theory. Although the term nosocomial still applies to infections that are acquired in acute care hospitals, a more general term, healthcare-associated infections (HAIs), is now used since much care of high-risk patients, including those with medical devices (e.g., central venous catheters, ventilators, peritoneal dialysis catheters), has shifted to ambulatory settings, rehabilitation or chronic care facilities, and to the home; thus, often the geographic location of acquisition of the infection cannot be determined. A true nosocomial infection is defined as an infection that was not incubating or present at the time of hospital admission, and that develops 48 hours or more after hospital admission or within 10 days of hospital discharge. In neonates, a transplacental infection is not considered a nosocomial infection. An infection is nosocomial, however, if a mother is not infected at the time of admission but delivers an infected infant more than 48 hours after her admission. The principles of transmission of infectious agents in healthcare settings and prevention are reviewed in the Guideline for Isolation Precautions: Preventing Transmission of Infectious Agents in Healthcare Settings 2007.1 The pediatric host is highly susceptible to common respiratory and gastrointestinal tract viruses (e.g., respiratory syncytial virus (RSV), influenza virus, rotavirus) that may be transmitted in healthcare settings in addition to the usual healthcare-associated bacteria and fungi. HAIs can result in the serious morbidity and mortality that occur in adult patients and in lifetime physical, neurologic, and developmental disabilities. Infection rates between 2% and 13% of admissions or discharges from pediatric units are typical.2–5 Intensive care units, oncology services, and gastroenterology services that care for patients with short gut who are dependent on total parenteral nutrition (TPN) have the highest rates of bacterial and fungal infection associated with central venous catheters. Those children who have complex underlying diseases are at greatest risk for prolonged hospitalization, complications, and mortality associated with acquisition of new infections in the hospital.3–8 Severely immunosuppressed patients (e.g., allogeneic hematopoietic stem cell transplant (HSCT) recipients, children with leukemia undergoing intensive chemotherapy, solid-organ transplant recipients during the periods of most intense immunosuppression), are at increased risk for invasive aspergillosis and other environmental fungal infections, especially during periods of facility renovation, construction, and water leaks.9

UNIQUE ASPECTS OF HEALTHCARE-ASSOCIATED INFECTION IN CHILDREN Unique aspects of HAIs in children have been reviewed in detail5 and are summarized below. Specific risks and pathogens are addressed in multiple other chapters in this textbook.

Host or Intrinsic Factors Rates of all HAIs as high as 7% to 25% are reported in neonatal intensive care units (NICUs) and are inversely proportional to birthweight.4,7,10 Host, or intrinsic, factors that make children particularly vulnerable to infection are immaturity of the immune system, congenital abnormalities, and congenital or acquired immunodeficiencies. The populations of immunosuppressed children have expanded with the advent of more intense immunosuppressive therapeutic regimens used for oncologic conditions, HSCTs, solid-organ transplants, and rheumatologic conditions and inflammatory bowel disease for which immunosuppressive agents and tumor necrosis factor-a inhibitors (infliximab) and other immune modulators are used. Fortunately, the population of children with perinatally acquired human immunodeficiency virus (HIV) infection and acquired immunodeficiency syndrome (AIDS) has dramatically decreased since 1994, but new cases of sexually transmitted HIV infection are diagnosed increasingly in teens who are cared for in children’s hospitals. Innate deficiencies of the immune system in prematurely born infants, who may be hospitalized for prolonged periods of time and exposed to intensive monitoring, supportive therapies and invasive procedures, contribute

to the high rates of infection in the NICU. All components of the immune system are deficient in neonates and the degree of deficiency is inversely proportional to the gestational age (see Chapter 10, Immunologic Development and Susceptibility to Infection). Additionally, the underdeveloped skin of the very-low-birthweight infant (< 1000 grams) provides another mode of entry for pathogens. Children with congenital anomalies have a high risk of HAI because they require prolonged and repeated hospitalizations, undergo many complex surgical procedures, and have extended exposure to invasive supportive and monitoring equipment. For example, at the University of Virginia Medical Center, children with myelomeningocele have had an average of 9 hospitalizations (range, 3 to 50) and 6 surgical procedures (range, 2 to 30) by 15 years of age. The source of many HAIs may be the endogenous flora of the patient.11,12 An asymptomatic colonizing pathogen can invade an individual patient or be transmitted on the hands of healthcare personnel to other patients. As the rates of colonization with community-associated methicillin-resistant Staphylococcus aureus (CA-MRSA) at the time of hospital admission have increased, so has transmission of community strains, most often USA 300, within the hospital,13 making prevention especially challenging. Finally, young infants who have not yet been immunized, or immunosuppressed children who do not respond to vaccines or lose their antibody during treatment (e.g., patients with nephrotic syndrome), have increased susceptibility to infections that would be prevented by vaccines.

Sources or Extrinsic Factors Important sources of HAIs in infants and children include the mother, invasive monitoring and supportive equipment, blood products, infant formula and human milk, healthcare personnel, and other contacts, including adult and sibling visitors. Maternal infection with Neisseria gonorrhoeae, Treponema pallidum, HIV, hepatitis B virus, parvovirus B19, Mycobacterium tuberculosis, herpes simplex virus, group B streptococcus, or the emerging CA-MRSA pose substantial threats to the neonate. During perinatal care, procedures such as fetal monitoring with scalp electrodes, fetal transfusion and surgery, umbilical cannulation, and circumcision are risk factors for infection. Intrinsically contaminated powdered formulas and infant formulas prepared in contaminated blenders or improperly stored or handled, or both, have resulted in sporadic and epidemic infections in the nursery (e.g., Enterobacter sakazakii).14 Human milk that has been contaminated by maternal flora or by organisms transmitted through breast pumps has caused isolated serious infections and epidemics. The risks of neonatal hepatitis, cytomegalovirus infection, and HIV infection from contaminated human milk warrant further caution for handling. Rates of central vascular line-associated bloodstream infections (CLA-BSIs) in the pediatric intensive care units (PICUs) and high-risk nurseries (HRN) in the National Nosocomial Infection Surveillance (NNIS) system (now the National Healthcare Safety Network (NHSN)) from 1/2002 to 6/2004 are among the highest for all reporting ICUs, with a mean of 6.6 CLA-BSIs per 1000 catheter-days in the PICUs; this rate is only surpassed in trauma and burn units, with a mean of 7.4 and 7.0 CLA-BSIs per 1000 catheter-days, respectively.10 Rates of umbilical and CLA infections vary by birthweight category from 3.5 per 1000 catheter-days in the > 2500 gram birthweight group to 9.1 per 1000 catheter-days in the < 1000 gram birthweight group (Tables 2-1 and 2-2). Medical device-related infections (e.g., CLA-BSIs, ventilator-associated pneumonia (VAP), and surgical site infections (SSIs)) can be prevented by implementing 3 to 5 sets or “bundles” of evidence-based practices, as defined in the Institute for Healthcare Improvement (IHI) 100,000 lives campaign (www.ihi.org/IHI/Programs/Campaign). Although most work has been done in adult populations, there are modifications for pediatrics and the efficacy of CLA-BSI preventive practices was demonstrated in a 1-year collaborative of children’s hospitals sponsored by the Child Health Corporation of America (CHCA) in 2005. Thus, it is likely that rates of device-related infections have been reduced even further since the NNIS report

PART I Understanding, Controlling, and Preventing Infectious Diseases

Pediatric Infection Prevention and Control

TABLE 2-1. National Nosocomial Infection Surveillance Central Line ‘Infection’ (CLI)-Associated Bloodstream Infection (CLA-BSI) Rates: Intensive Care Units (ICUs) January, 2002 to June, 2004a ICU Type

No. of ICUs Reporting

Trauma

22

7.4 (5.2, 1.9–11.9)

Burn

14

7.0 (NA)

Pediatric

54

6.6 (5.2, 0.9–11.2)

Medical

94

5.0 (3.9, 0.5–8.8)

Respiratory

Rate/1000 Catheter-Days: Pooled Mean (Median, Range)

6

4.8 (NA)

Surgical

99

4.6 (3.4, 0–8.7)

Neurosurg

30

4.6 (3.1, 0–10.6)

Coronary

60

3.5 (3.2, 1.0–9.0)

100 109

4.0 (3.4, 1.7–7.6) 3.2 (3.1, 0.8–6.1)

Medical-surgical Major teaching All others Cardiothoracic

48

2.7 (1.8, 0–4.9)

a

No. of central catheter-associated BSIs μ 1000 No. of central catheter-days

TABLE 2-2. National Nosocomial Infection Surveillance Central Line (CLA-BSI)-Associated Bloodstream Infection Rates: Umbilical and Central Catheter Bloodstream Infection Rates: High-Risk Nursery, January, 2002 to June, 2004a Birthweight Group ≤ 1000 grams

No. of Nurseries Reporting 104

Rate/1000 Catheter Days: Pooled Mean (Median, Range) 9.1 (8.5, 1.6–16.1)

1001–1500 grams

98

5.4 (4.0, 0–12.2)

1501–2500 grams

97

4.1 (3.2, 0–8.9)

> 2500 grams

94

3.5 (1.9, 0–7.4)

a

No. of central catheter-associated BSIs μ 1000 No. of central catheter-days Adapted from the National Nosocomial Infections Surveillance (NNIS) System Report, data summary from January 1992 through June 2004, issued October 2004. Am J Infect Control 2004;32:470–485.

published in December, 2004. Use of more specialized life-saving technologies, such as extracorporeal membrane oxygenation (ECMO), hemodialysis/hemofiltration, pacemakers, and implantable ventricular assist devices (VADs), further increases the risk of infection in the sickest children who require the most intense, invasive support. Many standard infection control procedures for prevention of device-related infections in adults cannot be followed routinely for children. In adults, for example, peripheral intravascular catheters are changed routinely every 3 to 4 days to reduce the risk of catheter colonization and subsequent infection of the bloodstream. Infants, however, may have such limited vascular access that catheters remain in place until they become unnecessary, nonfunctional, or contaminated. Additionally, the specific indications for deep-vein thrombosis and peptic ulcer disease prophylaxis have not been defined for children requiring mechanical ventilator support and there is some evidence suggesting that peptic ulcer disease prophylaxis is associated with an increased risk of necrotizing enterocolitis and candidemia in low-birthweight infants.15 There are theoretical concerns that infection risk will also increase in association with the innovative practices of co-bedding and

CHAPTER

2

11

kangaroo care in the NICU because of increased opportunity for skinto-skin exposure of multiple-gestation infants to each other and to their mothers, respectively. However, the infection risk is reduced with kangaroo care.16 Finally, exposure to vancomycin and to thirdgeneration cephalosporins contributes substantially to the increase in infections caused by vancomycin-resistant enterococcus (VRE) and multidrug-resistant gram-negative bacilli, including extendedspectrum beta-lactamase (ESBL)-producing organisms, respectively. Exposure to third-generation cephalosporins is also a risk factor for the development of invasive candidiasis in low-birthweight infants in the NICU.17,18

Transmission The principal modes of transmission of infectious agents are direct and indirect contact, droplet, and airborne. Most infectious agents are transmitted by the contact route via hands of healthcare personnel, but many pathogens can be transmitted by more than one route. Viruses, bacteria, and Candida spp. can be transmitted horizontally. Although the source of most Candida HAIs is the patient’s endogenous flora, horizontal transmission, most likely via healthcare personnel hands, has been demonstrated in studies using DNA fingerprinting in the NICU and in a pediatric oncology unit. Transmission of infectious agents by the droplet route requires exposure of mucous membranes to large respiratory droplets (> 5 μm) within 1 to 2 meters (3 to 6 feet) of the infected individual, who may be coughing or sneezing. Large respiratory droplets do not remain suspended in the air. Adenovirus, influenza virus, and rhinovirus are primarily transmitted by the droplet route whereas other respiratory viruses (e.g., RSV, parainfluenza) are primarily transmitted by the contact route. Although influenza virus can be transmitted via the airborne route under unusual conditions of reduced air circulation or relative humidity, there is ample evidence that transmission of influenza is prevented by droplet precautions and, in the care of infants, the addition of contact precautions under usual conditions.19 Some agents (e.g., severe acute respiratory syndrome–coronavirus (SARS-CoV)) can be transmitted as small-particle aerosols under special circumstances of aerosol-producing procedures (e.g., endotracheal intubation, bronchoscopy); therefore, an N95 or higher respirator is indicated for those in the same airspace when these procedures are performed, but an airborne infection isolation room (AIIR) may not always be required. Roy & Milton proposed a new classification for aerosol transmission when evaluating routes of SARS transmission20: (1) obligate: under natural conditions, disease occurs following transmission of the agent only through small-particle aerosols (e.g., tuberculosis); (2) preferential: natural infection results from transmission through multiple routes, but small-particle aerosols are the predominant route (e.g., measles, varicella); and (3) opportunistic: agents naturally cause disease through other routes, but under certain environmental conditions can be transmitted via fineparticle aerosols. This conceptual framework may explain rare occurrences of airborne transmission of agents that are transmitted most frequently by other routes (e.g., smallpox, SARS, influenza, noroviruses). Concerns about unknown or possible routes of transmission of agents that can cause severe disease and have no known treatment often result in more extreme prevention strategies than may be necessary; therefore, recommended precautions could change as the epidemiology of emerging agents is defined and these controversial issues are resolved. Although transmission of M. tuberculosis can occur rarely from a child with active tuberculosis, the more frequent source is the adult visitor who has not been diagnosed with active pulmonary tuberculosis; thus screening of visiting family members is an important component of control of tuberculosis in pediatric healthcare facilities.21 Transmission of microbes among children and between children and healthcare personnel is a frequent risk due to the very close contact that occurs during care of infants and young children. Traditionally, multi-bed rooms are crowded with children, parents, and

12

SECTION

A Epidemiology & Control of Infectious Diseases

healthcare personnel. However, with the increasing evidence that single-patient rooms provide improved environments for patients that includes reduced risk of transmission of infectious agents and reduced medical errors, the American Institute of Architects’ 2006 Guidelines for Design and Construction of Health Care Facilities recommends single-patient rooms for acute medical/surgical and postpartum patients as the standard for all new construction (www.aia.org/aah_gd_hospcons). Although there are insufficient data at this time to support a definitive recommendation for single-patient rooms in NICUs, there is increasing experience that suggests a benefit to reduce the risk of infection and to improve neurosensory development.22 Toddlers often share waiting rooms, playrooms, toys, books, and other items and therefore have the potential of transmitting pathogens directly and indirectly to one another. Contaminated bath toys were implicated in an outbreak of multidrug-resistant Pseudomonas aeruginosa in a pediatric oncology unit.23 Before effective preventive measures were established,24 17% of preschool children hospitalized for more than 1 week had a nosocomial viral respiratory tract illness.25 Infection of pediatric healthcare workers was also common. Since routine care of infants and younger children involves holding, cuddling, wiping noses, feeding, and changing diapers, it is easy to see how RSV and other respiratory tract viral agents can be transmitted in secretions that are then inoculated into the eyes and noses of healthcare workers. RSV infections were more likely in healthy volunteers who held or cuddled infants or handled items that the infants had touched and did not occur in those who were in the patients’ rooms but had no direct patient contact and did not touch any items or surfaces.26 A source of further concern involves healthcare workers with mild symptoms of infection who unknowingly become intermediary hosts and who transmit organisms to susceptible children. Several published studies have shown that infected pediatric healthcare personnel, including resident physicians, transmitted Bordetella pertussis to other patients.27 Healthcare personnel have been implicated as the source of outbreaks of rotavirus28 and influenza.29 Transmission of infectious agents is further facilitated by overcrowding and understaffing. Several studies demonstrated the association of understaffing and overcrowding with increased rates of HAIs in NICUs, PICUs, and general pediatrics units30–33 and contributed substantially to the evidence base that supports recommendations to consider staffing levels and composition as important components of an effective infection control program in the 2007 revision of the Healthcare Infection Control Practices Advisory Committee (HICPAC)/Centers for Disease Control and Prevention (CDC) guideline for isolation precautions in healthcare settings.1 Healthcare personnel are rarely the source of outbreaks of HAIs caused by bacteria and fungi, but when they are, there are usually factors present that increase the risk transmission of infectious agents to others (e.g., sinusitis, draining otitis externa, respiratory tract infections, dermatitis, onychomycosis, wearing of artificial nails).34 Those individuals with direct patient contact wearing artificial nails have been implicated in outbreaks of Pseudomonas aeruginosa and ESBL-producing Klebsiella pneumoniae in NICUs.35,36 These studies contributed to the recommendation to prohibit use of artificial nails or extenders when having direct contact with high-risk patients.1,37

Pathogens While there is no agreed-upon definition for what constitutes an “epidemiologically important organism,” the following characteristics apply and are presented for guidance to infection control staff in the 2007 revision of the HICPAC/CDC Guideline for Isolation Precautions in Healthcare Settings (www.cdc.gov/ncidod/dhqp/ pdf/guidelines/isolation2007.pdf): 1. A propensity for transmission within healthcare facilities based on published reports and the occurrence of temporal or geographic clusters of > 2 patients (e.g., VRE, MRSA, and methicillinsusceptible Staphylococcus aureus (MSSA), Clostridium difficile,

norovirus, RSV, influenza, rotavirus, Enterobacter spp., Serratia spp., group A streptococcus). A single case of healthcareassociated invasive disease caused by certain pathogens (e.g., group A streptococcus postoperatively or in burn units; Legionella sp.; Aspergillus sp.) should trigger an investigation. 2. Antimicrobial resistance (e.g., MRSA, VRE, ESBL-producing gram-negative bacilli, Burkholderia cepacia, Ralstonia spp., Stenotrophomonas maltophilia, and Acinetobacter. Many of the intrinsically resistant gram-negative bacilli also suggest possible water or medication contamination. 3. Association with serious clinical disease, increased morbidity and mortality (e.g., MRSA and MSSA, group A streptococcus). 4. A newly discovered or reemerging pathogen (e.g., vancomycininsensitive or resistant Staphylococcus aureus (VISA, VRSA), C. difficile). Pathogens associated with HAIs in hospitalized children differ from those in adults. Viral agents and other respiratory tract pathogens (e.g., Bordetella pertussis) have heightened potential for transmission in pediatric facilities. Gram-negative bacilli, including ESBL and other multidrug-resistant isolates, may be more frequent than MRSA and VRE in many PICUs and NICUs. Patients who are transferred from chronic care facilities may be colonized with resistant gram-negative bacilli at the time of admission to the PICU.11 Trends in targeted multidrug-resistant pathogens that have been tracked in the NNIS (now NHS) ICUs are summarized in Figure 2-1. Continued increases in MRSA, VRE, and certain resistant gram-negative bacilli are a “call to action” for all healthcare facilities. The CDC Campaign to prevent antimicrobial resistance and the Guideline for Management of Multi-Drug Resistant Organisms (MDRO) in Healthcare Settings 2006 can be accessed on the following websites, respectively: www.cdc.gov/drugresistance/healthcare; www.cdc.gov/nciod/dhqp/ index.html. Of note, in 2004, rates of healthcare-associated MRSA and VRE appear to have reached a plateau, whereas there are steep increases in the incidence of Klebsiella pneumoniae resistant to thirdgeneration cephalosporins in the ICUs reporting to the current CDC surveillance system, NHSN (formerly NNIS) (www.cdc.gov/ncidod/ dhqp/ar_mrsa_data.html). HAIs caused by MDROs are associated with increased length of stay, increased morbidity and mortality, and increased cost, in part due to the delay in initiating antimicrobial therapy that will be active against the infecting agent.38 While there is lower prevalence of specific MDROs in pediatric institutions, the same principles of target MDRO identification and control interventions apply to all settings. The emergence of CA-MRSA isolates characterized by the unique Scc mec type IV element was first observed among infants and children and is now being transmitted in hospitals, notably in NICUs,13 making prevention more complex. The viruses most frequently associated with transmission in a pediatric healthcare facility are RSV, rotavirus, and influenza. However, other respiratory viruses (e.g., parainfluenza, adenovirus) have been implicated in outbreaks in high-risk units. Outbreaks of varicella and measles in pediatric healthcare facilities are rare events now due to consistent uptake during the past decade of vaccines in children and in healthcare personnel. Clinical manifestations with certain pathogens are more severe in infants and young children. RSV and Bordetella pertussis usually cause mild upper respiratory tract infections and cough, respectively, in older children and adults, yet cause severe disease with substantial morbidity and mortality in infants and children, especially those who are immunocompromised or who have underlying cardiac or pulmonary disease. An excessive burden of disease and mortality associated with influenza in infants and young children is also recognized.39,40 Candida sp. had been increasing in incidence in most PICUs and NICUs during the 1990s. There is considerable center-to-center variability in both the incidence of invasive candidiasis and the proportion of Candida infections caused by Candida non-albicans sp., most of which are resistant to fluconazole. Risk factors for Candida infections include prolonged length of stay in an ICU, use of CVCs, intralipids, H2-blocking agents, and exposure to third-generation cephalosporins. Gram-negative bacilli and Candida sp. are especially

PART I Understanding, Controlling, and Preventing Infectious Diseases

Pediatric Infection Prevention and Control

Vancomycin/enterococci

28.5%

Methicillin/S. aureus

59.5%

Methicillin/CNS

89.1%

3rd Ceph/E. coli **

5.8%

3rd Ceph/K. pneumoniae **

20.6%

Imipenem/P. aeruginosa

21.1%

Quinolone/P. aeruginosa

29.5%

3rd Ceph/P. aeruginosa

31.9%

3rd Ceph/Enterobacter spp.

31.1% 0

10

20

30

40

50

60

70

80

CHAPTER

2

Jan–Dec 2003 No. of isolates

% increase in resistance (2003 vs 98–02*)

2048

12%

4100

11%

3336

1%

1355

0%

1068

47%

1392

15%

1825

9%

2119

20%

1411

–6%

13

90

% resistance January through December 2003 1998 through 2002 (+/- standard deviation)* Figure 2-1. Selected antimicrobial-resistant pathogens associated with nosocomial infections in intensive care unit patients; comparison of resistance rates from January through December 2003 with 1998 through 2002, National Nosocomial Infections Surveillance (NNIS) system. CNS, Coagulasenegative staphylococci; 3rd Ceph, resistance to third-generation cephalosporins (ceftriaxone, cefotaxime, or ceftazidime); Quinolone, resistance to either ciprofloxacin or ofloxacin. *Percent (%) increase in resistance rate of current year (January–December 2003) compared with mean rate of resistance over previous 5 years (1998–2003): [(2003 rate—previous 5-year mean rate)/previous 5-year mean rate] μ 100. **“Resistance” for Escherichia coli or Klebsiella pneumoniae is the rate of nonsusceptibility of these organisms to either 3rd Ceph group or aztreonam. Redrawn from the National Nosocomial Infections Surveillance (NNIS) System Report, data summary from January 1992 through June 2004, issued October 2004. Am J Infect Control 2004;32:470–485.

important pathogens for HAIs in patients with short gut who are TPNdependent and can cause repeated episodes of sepsis.41,42 Finally, environmental fungi (e.g., Aspergillus, Fusarium, Scedosporium, Bipolaris, Zygomycetes), are important sources of infection for severely immunocompromised patients, demanding meticulous attention to the conditions of the internal environment of any facility that provides care for severely immunocompromised patients and prevention of possible exposure to construction dust in and around healthcare facilities.9,43 With the advent of more effective and less toxic antifungal agents, it is important to identify the infecting agent by obtaining tissue samples and to determine susceptibility to candidate antifungal agents.43,44

PREVENTION Prevention remains the mainstay of infection control and requires special considerations in children. The goals of infection control and prevention are to prevent the transmission of infectious agents among individual patients or groups of patients, visitors, and healthcare personnel who care for them. If prevention cannot always be achieved, the next best strategy is early diagnosis, treatment, and prevention of continued transmission. An effective infection control program should improve patient and healthcare personnel safety and decrease short-

and long-term morbidity, mortality, and healthcare costs. This chapter describes the unique principles and practice of infection control for the care of children. Specific pathogens and diseases are discussed in detail in chapters dedicated to those topics. Recommended isolation precautions by infectious agent can be found in the Red Book Report of the Committee on Infectious Diseases (COID) of the American Academy of Pediatrics (AAP) and in the Guideline for Isolation Precautions: Preventing Transmission of Infectious Agents in Healthcare Settings 2007.1 In addition, textbooks on healthcare epidemiology and infection control contain chapters devoted to pediatric-specific programs. A series of infection control prevention and control guidelines have been developed and updated by HICPAC/CDC and others to provide evidence-based, rated recommendations for practices that are associated with reduced rates of HAIs, especially those associated with the use of medical devices and surgical procedures (Box 2-1). Bundled practices are groups of three to five evidence-based “best practices” with respect to a disease process that individually improve care, but when applied together result in substantially greater reduction in infection rates. Adherence to the individual measures within a bundle is readily measured. Bundles for the reduction of CLA-BSIs, surgical site infections (SSIs), and ventilator assocated pneumonia (VAP) established for adults have been adapted to pediatrics (www.ihi.org/IHI/Programs/Campaign).

14

SECTION

A Epidemiology & Control of Infectious Diseases

BOX 2-1. Resources for Infection Control Recommendations CENTERS FOR DISEASE CONTROL AND PREVENTION/ HEALTHCARE INFECTION CONTROL PRACTICES COMMITTEE (www.cdc.gov/ncidod/dhqp/index.html) Influenza vaccination of health-care personnel: recommendations of the Healthcare Infection Control Practices Committee (HICPAC) and the Advisory Committee on Immunization Practices (ACIP). MMWR 2006;55:RR-2 Guideline for disinfection and sterilization in health-care facilities (in revision) Guideline for prevention of healthcare-associated pneumonia, 2003. MMWR 2004;53(RR-3) Guideline for environmental infection control in health-care facilities, 2003. MMWR 2003;52(RR-10) Guideline for hand hygiene in health-care settings, 2002. MMWR 2002;51(RR-16) Guideline for prevention of intravascular catheter-related infections, 2002. MMWR 2002;51(RR-10) Recommendations for preventing transmission of infections among chronic hemodialysis patients, 2001. MMWR 2001;50(RR-5) Guidelines for preventing opportunistic infections among hematopoietic stem cell transplant recipients, 1999. MMWR 2000;49(RR-10) Guideline for the prevention of surgical site infection, 1999. Infect Control Hosp Epidemiol 1999;20:247–278 Infection control in health care personnel, 1998. Infect Control Hosp Epidemiol 1998;19:407–463 Guideline for isolation precautions in hospitals in healthcare settings, 2007. www.cdc.gov/neidod/dhqp/pdf/guidelines/isolation2007.pdf Management of multi-drug resistant organisms (MDROs) in healthcare settings, 2006. www.cdc.gov/nciod/dhqp/index.html: posted 10/19/06 Guidelines for preventing the transmission of Mycobacterium tuberculosis in health-care facilities, 2005. MMWR 2005;54(RR-17) Guideline for prevention of catheter-associated urinary tract infections. Am J Infect Control 1983;11:28–33 AMERICAN ACADEMY OF PEDIATRICS Committee on Infectious Diseases. 2006 Report of the Committee on Infectious Diseases. In: Pickering LK, Baker CJ, Long SS (eds). Red Book, 27th ed. Illinois, American Academy of Pediatries, 2006 OTHER Society for Healthcare Epidemiology of America (SHEA) Position Papers (www.shea-online.org/PositionPapers.html) Infectious Diseases Society of America (IDSA) Clinical Practice Guidelines (www.journals.uchicago.edu/IDSA/guidelines) Association for Professionals in Infection Control and Epidemiology (APIC) Practice. Guidelines and State of the Art Reports (www.apic.org) Carrico R (ed.) APIC Text of Infection Control and Epidemiology, 2nd ed. Washington, DC, Association for Professionals in Infection, Control and Epidermiology 2005 Saiman L, Siegel JD, and the Cystic Fibrosis Foundation Consensus. Infection control recommendations for patients with cystic fibrosis: microbiology, important pathogens, and infection control practices to prevent patient-to-patient transmission. Infect Control Hosp Epidermial 2003;24(Suppl. 1–62).

Administrative Factors The importance of certain administrative measures for a successful infection control program has been demonstrated. There is now an adequate evidence base to designate infection control as one component of the institutional culture of safety and to obtain support from the senior leadership of the healthcare organization in order to provide necessary fiscal and human resources for a proactive, successful infection control program. Critical elements requiring administrative support include access to appropriately trained healthcare epidemiology and infection control personnel; access to clinical microbiology laboratory services needed to support infection control outbreak investigations and multidisciplinary programs to assure judicious use of antimicrobial agents and control of antimicrobial resistance; delivery of effective educational information to healthcare personnel, patients,

families, and visitors; and provision of adequate numbers of welltrained infection control, and bedside nursing staff.1,30–33

The Infection Control and Prevention Team The goals of infection control and prevention are to prevent the transmission of infectious agents among individual patients or groups of patients, visitors, and healthcare personnel who care for them. If prevention cannot always be achieved, the next best strategy is early diagnosis, treatment, and prevention of continued transmission. An effective infection control program should improve patient and healthcare personnel safety and decrease short- and long-term morbidity, mortality, and healthcare costs.45 This chapter describes the unique principles and practice of infection control for the care of children. Recommended isolation precautions by infectious agent may be found in the Red Book Report of the COID (AAP) and in the Guideline for Isolation Precautions: Preventing Transmission of Infectious Agents in Healthcare Settings 2006.1 The infection control committee establishes policies and procedures to prevent or reduce the incidence and costs associated with HAIs. The infection control committee should be one of the strongest and most accessible committees in the hospital; committee composition should be carefully considered and limited to active, authoritative participants who have well-defined responsibilities on the committee and who represent major groups within the hospital. The chairperson should be a good communicator with expertise in infection control issues, healthcare epidemiology, and clinical pediatric infectious diseases. An important function of the infection control committee is the regular review of infection control policy and the development of new infection control policies as needed. Annual review is required by the Joint Commission on Accreditation of Healthcare Organizations and can be optimally accomplished by careful review of a few policies each month. With the advent of unannounced inspections, a constant state of readiness is required. If a facility chooses not to have an infection control committee, an alternative strategy is needed to accomplish the above tasks. The infection control and prevention division or team is the working group (including physicians, nurses, microbiologists, and administrators) that performs and coordinates all infection control activities. The hospital epidemiologist or medical director of the infection control division is usually a physician with training in pediatric infectious diseases and a dedicated expertise and interest in healthcare epidemiology. In multidisciplinary medical centers, pediatric infectious disease experts should be consulted for management of pediatric infection control and report to the broader infection control leadership. Infection control and prevention professionals (ICPs) are specialized professionals with advanced training and preferably certification in infection control. Although the majority of ICPs are registered nurses, others, including microbiologists, medical technologists, pharmacists, and epidemiologists, are successful in this position. Pediatric patients should have ICP services provided by someone with expertise and training in the care of children. In a large, general hospital, at least one ICP should be dedicated to infection control services for children. The responsibilities of ICPs have expanded greatly in the last decade and include the following: (1) surveillance and infection prevention in facilities affiliated with the primary acute care hospitals (e.g., ambulatory clinics, day-surgery centers, long-term care facilities, rehabilitation centers, home care) in addition to the primary hospital; (2) oversight of employee health services related to infection prevention, (e.g., assessment of risk and administration of recommended treatment following exposure to infectious agents, tuberculosis screening, influenza and pertussis vaccination, respiratory protection fit testing, administration of other vaccines as indicated during infectious disease crises such as pre-exposure smallpox vaccine in 2003); (3) preparedness planning for annual influenza outbreaks, pandemic influenza, SARS, bioweapons attacks; (4) adherence monitoring for selected infection control practices; (5) oversight of risk assessment and

PART I Understanding, Controlling, and Preventing Infectious Diseases

Pediatric Infection Prevention and Control

implementation of prevention measures associated with construction, renovation, and other environmental conditions associated with increased infection risk; (6) prevention of transmission of MDROs; (7) evaluation of new products that could be associated with increased infection risk (e.g., intravenous infusion materials), for introduction and assessment of performance after implementation; (8) mandatory public reporting of HAI rates in states as legislation is enacted; (9) increased communication with the public and with local public health departments concerning infection control-related issues; and (10) participation in local and multicenter research projects. Infection control programs must be adequately staffed to perform all of these activities. Thus, the ratio of 1 ICP per 250 beds that was associated with a 30% reduction in the rates of nosocomial infection in the Study on Efficacy of Nosocomial Infection Control (SENIC) study performed in the 1970s46 is no longer sufficient, as the complexity of patient populations and the responsibilities of infection control professionals have increased. Many experts recommend that a ratio of 1 ICP per 100 beds is more appropriate for the current workload, but no study has been performed to confirm the effectiveness of that ratio. There is no information on the number of individuals required outside acute care, but it is clear that individuals well trained in infection control must be available for all sites where healthcare is delivered.1

Surveillance Surveillance for HAIs consists of a systematic method of determining the incidence and distribution of infections acquired by hospitalized patients. The CDC recommends the following: (1) prospective surveillance on a regular basis by trained infection control professionals, using standardized definitions; (2) analysis of infection rates using established epidemiologic and statistical methods (e.g., calculation of rates using appropriate denominators that reflect duration of exposure; use of statistical process control charts for trending rates); (3) regular use of data in decision-making; and (4) employment of an effective and trained healthcare epidemiologist who develops infection control strategies and policies and serves as a liaison with the medical community and administration.47–50 The CDC has established a set of standard definitions of HAIs that have been validated and accepted widely51 with updates posted on the CDC website or published in HICPAC/CDC guidelines. Standardization of surveillance methodology has become especially important with the advent of state legislation for mandatory reporting to the public of HAI infection rates.52 Although various surveillance methods are used, the basic goals and elements are similar and include using standardized definitions of infection, finding and collecting cases of HAIs, tabulating data, using appropriate denominators that reflect duration of risk, analyzing and interpreting the data, reporting important deviations from endemic rates (epidemic, outbreaks) to the bedside care providers and to the facility administrators, implementing appropriate control measures, auditing adherence rates for recommended measures, and assessing efficacy of the control measures. Medical centers can utilize different methods of surveillance, as outlined in Box 2-2. Most experts agree that a combination of methods enhances surveillance and data reliability and that some combination of clinical chart review and database retrieval is important.47–50 Administrative databases created for the purposes of billing should not be used as the sole source to identify HAIs because of both the overestimates and underestimates that result from inaccurate coding of HAIs.52 Use of software designed specifically for infection control data entry and analysis facilitates real-time tracking of trends and timely intervention when clusters are identified. The microbiology laboratory can provide online culture information about individual patients, outbreaks of infection, antibiotic susceptibility patterns of pathogens in periodic antibiotic susceptibility summary reports, and employee infection data. This laboratory can also assist with surveillance cultures and facilitation of molecular

CHAPTER

2

15

BOX 2-2. Sources of Data for Surveillance • Clinical rounds with physicians and/or nurses • Review of: Patient orders Radiology reports/databases Pharmacy reports/databases Operating room diagnoses and procedures Microbiology: bacteriology, virology, mycology, acid-fast bacilli, serology reports autopsy reports data-mining reports • Postdischarge surveillance, especially for surgical site infections • Public health surveillance • Review of: Employee health reports Admission diagnoses Outpatient diagnoses Administrative databases, but should not be used as sole source due to inaccurate coding of healthcare-associated infections

typing of isolates during outbreak investigations. Rapid diagnostic testing of clinical specimens for identification of viruses and Bordetella pertussis is especially important for pediatric facilities. The infection control division and the microbiology laboratory must communicate daily, because even requests for cultures from physicians (e.g., Mycobacterium tuberculosis, Neisseria meningitidis, Clostridium difficile) can be an early marker for identifying patients who are infected, are at high risk of infection, or require isolation. If microbiology laboratory work is outsourced, it is important to assure that the services needed to support an effective infection control program will be available, as described in a policy statement of the Infectious Diseases Society of America on this matter.53 The pharmacy is an important collaborative member of any multidisciplinary team working on strategies to prevent antimicrobial resistance. Antimicrobial utilization in the hospital should be assessed for appropriateness, efficacy, cost, and association with emergence of resistant organisms. For surveillance purposes, use of specific antimicrobial agents can alert the ICP to potentially infected patients (e.g., tuberculosis). The need to restrict use of antimicrobial agents is a collaborative decision based on review of all of these data. Restriction of new, potent antimicrobial agents is advised to prevent emergence of resistance that occurs with increased exposure to most antimicrobial agents (e.g., extended-spectrum cephalosporins, quinolones, linezolid, daptomycin).54–56 Control of unusual infections or outbreaks in the community is generally the responsibility of the local or state public health department; however, the individual facility must be responsible for preventing transmission within that facility. Public health agencies can be particularly helpful in alerting hospitals of community outbreaks so that outpatient and inpatient diagnosis, treatment, necessary isolation, and other preventive measures begin promptly to avoid further spread. Conversely, the responsibility of designated individuals in the hospital is to notify public health department personnel of reportable infections so as to facilitate early diagnosis, treatment, and infection control in the community. Benefits of community or regional collaboratives of individual healthcare facilities and local public health departments for prevention of HAIs, especially those caused by MDROs, have been demonstrated and should be encouraged.1

ISOLATION PRECAUTIONS Isolation of patients with potentially transmissible infectious diseases is a proven strategy for reducing transmission of infectious agents in healthcare settings. During the past decade, many published studies, including those performed in pediatric settings, have provided a strong evidence base for most recommendations for isolation precautions. However, many controversies still exist concerning the most clinically

16

SECTION

A Epidemiology & Control of Infectious Diseases

and cost-effective measures for preventing certain HAIs, especially those associated with MDROs. Since 1970, the guidelines for isolation developed by CDC have responded to the needs of the evolving healthcare systems in the United States. For example, universal precautions became a required standard in response to the HIV epidemic and the need to prevent transmission of bloodborne pathogens (e.g., HIV, hepatitis B and C viruses, rapidly fatal infections such as the viral hemorrhagic fevers). The Occupational Safety and Health Administration (OSHA) published specific requirements57 in 1991 for universal precautions (now called Standard Precautions) for healthcare personnel who, as a result of their required duties, are at increased risk for skin, eye, mucous membrane, or parenteral contact with blood or other potentially infectious materials. Although all requirements may not have been proven to be clinically or cost-effective, healthcare facilities must enforce these measures. The federal Needlestick Safety and Prevention Act, signed into law in November, 2000, authorized OSHA’s revision of its Bloodborne Pathogens Standard more explicitly to require the use of safety-engineered sharp devices (www.osha.gov/SLTC/bloodbornepathogens/index.html). The 1996 CDC/Hospital Infection Control Advisory Committee Guideline for Isolation Precautions in Hospitals58 has been updated and published in 2007 as the Guideline for Isolation Precautions: Preventing Transmission of Infectious Agents in Healthcare Settings.1 This guideline affirms standard precautions, a combination of universal precautions and body substance isolation as the foundation of transmission prevention measures. Healthcare personnel are provided with guidance to recognize the importance of body fluids, excretions, and secretions in the transmission of infectious pathogens and to take appropriate protective precautions by using personal protective equipment (e.g., masks, gowns, gloves, face shields, or

goggles) and safety devices even if an infection is not suspected or known. In addition, this updated guideline provides recommendations for all settings where healthcare is delivered (acute care hospitals, ambulatory surgical and medical centers, long-term care facilities, and home health agencies). Extensive background discussion and recommendations for the prevention of transmission of multidrug-resistant organisms (e.g., MRSA, VRE, VISA, VRSA, and gram-negative bacilli) in various settings are also included and were pre-released on the CDC website in October, 2006 (www.cdc.gov/ncidod/dhqp/ pdf/ar/mdroGuidelines2006.pdf. The categories of Transmissionbased Precautions described previously have been retained: Contact, Droplet, and Airborne Precautions. The characteristics of a protective environment for prevention of environmental fungal infections in HSCT recipients that were introduced in previously published guidelines are summarized. Finally, discussion and recommendations with evidence-based ratings for administrative measures that are necessary for effective prevention of infection in healthcare settings are provided. The isolation information presented in this chapter is based on these 2006 to 2007 isolation recommendations.

Standard Precautions The term Standard Precautions replaced Universal Precautions and Body Substance Isolation in 1996. Standard Precautions should be used when there is likely to be exposure to: (1) blood; (2) all other body fluids, secretions, and excretions, whether or not they contain visible blood, except sweat; (3) nonintact skin; and (4) mucous membranes. Standard Precautions strategy is designed to reduce the risk of transmission of microorganisms from both identified and unidentified sources of infection. The components of Standard

TABLE 2-3. Recommendations for Application of Standard Precautions for the Care of all Patients in all Healthcare Settings Component

Recommendations for Performance

Hand hygiene

After touching blood, body fluids, secretions, excretions, contaminated items; immediately after removing gloves; between patient contacts. Alcohol-containing antiseptic handrubs preferred except when hands are visibly soiled with blood or other proteinaceous materials or if exposure to spores (e.g., Clostridium difficile, Bacillus anthracis) is likely to have occurred

Gloves

For touching blood, body fluids, secretions, excretions, contaminated items; for touching mucous membranes and nonintact skin

Gown

During procedures and patient care activities when contact of clothing/exposed skin with blood/body fluids, secretions, and excretions is anticipated

Mask,a eye protection (goggles), face shield

During procedures and patient care activities likely to generate splashes or sprays of blood, body fluids, secretions, especially suctioning, endotracheal intubation to protect healthcare personnel. For patient protection, use of a mask by the individual inserting an epidural anesthesia needle or performing myelograms when prolonged exposure of the puncture site is likely to occur

Soiled patient-care equipment

Handle in a manner that prevents transfer of microorganisms to others and to the environment; wear gloves if visibly contaminated; perform hand hygiene

Environmental control

Develop procedures for routine care, cleaning, and disinfection of environmental surfaces, especially frequently touched surfaces in patient care areas

Textiles and laundry

Handle in a manner that prevents transfer of microorganisms to others and to the environment

Injection practices (use of needles and other sharps)

Do not recap, bend, break, or hand-manipulate used needles; if recapping is required, use a one-handed scoop technique only; use needle-free safety devices when available; place used sharps in a puncture-resistant container. Use a sterile, single-use, disposable needle and syringe for each injection given. Single-dose medication vials are preferred when medications are administered to > 1 patient

Patient resuscitation

Use mouthpiece, resuscitation bag, or other ventilation devices to prevent contact with mouth and oral secretions

Patient placement

Prioritize for single-patient room if the patient is at increased risk of transmission, is likely to contaminate the environment, does not maintain appropriate hygiene, or is at increased risk of acquiring infection or developing adverse outcome following infection

Respiratory hygiene/cough etiquetteb

Instruct symptomatic persons to cover mouth/nose when sneezing/coughing; use tissues and dispose in no-touch receptacle; observe hand hygiene after soiling of hands with respiratory secretions; wear surgical mask if tolerated or maintain spatial separation, > 1 meter (3 feet) if possible

a

During aerosol-generating procedures on patients with suspected or proven infections transmitted by aerosols (e.g., severe acute respiratory syndrome), wear a fit-tested N95 or higher respirator in addition to gloves, gown, and face/eye protection. b Source containment of infectious respiratory secretions in symptomatic patients, beginning at initial point of encounter (e.g., triage and reception areas in emergency departments and physician offices).

PART I Understanding, Controlling, and Preventing Infectious Diseases

Pediatric Infection Prevention and Control

Precautions are summarized in Table 2-3. In the updated isolation guideline, safe injection practices are included as a component of standard precautions, not because new practices are recommended but because recent outbreaks of hepatitis B and C virus infection in ambulatory care settings as a result of failure to follow recommended practices indicate a need to reiterate the established effective practices.59 There are two new additions to Standard Precautions: (1) respiratory hygiene/cough etiquette for source containment by patients with signs and symptoms of respiratory tract infection; and (2) use of a mask by the individual inserting an epidural anesthesia needle or performing myelograms when prolonged exposure of the puncture site is likely to occur to prevent introduction of respiratory tract microorganisms from the person performing the procedure into the cerebrospinal fluid of the patient and therefore prevent meninigitis. Both new components have a strong evidence base. Implementation of Standard Precautions requires critical thinking from all healthcare personnel providing direct patient care and the availability of personal protective equipment in proximity to all patient beds. Healthcare personnel with exudative lesions or weeping dermatitis must avoid direct patient care and handling of patient care equipment. Individuals having direct patient contact should be able to anticipate an exposure to blood or other potentially infectious material and to take proper protective precautions. Individuals should also know what to do if a high-risk exposure does occur. Exposures of concern are exposures to blood or other potentially infectious material defined as an injury with a contaminated sharp object (e.g., needlestick, scalpel cut); a spill or splash of blood or other potentially infectious material on to nonintact skin (e.g., cuts, hangnails, dermatitis, abrasions, chapped skin) or on to a mucous membrane (e.g., mouth, nose, eye); or a blood exposure covering a large area of normal skin. Handling food trays or furniture, pushing wheelchairs or stretchers, using restrooms or phones, having personal contact with patients (e.g., giving information, touching intact skin, bathing, giving a back rub, shaking hands), or doing clerical or administrative duties for a patient do not constitute high-risk exposures. If hands or other skin surfaces are exposed to blood or other potentially infectious material, the healthcare worker should immediately wash the area with soap and water for at least 10 seconds and rinse with running water for at least 10 seconds. If an eye, the nose, or mouth is splashed with blood or body fluids, the area should be immediately irrigated with a large volume of water. If a skin cut, puncture, or lesion is exposed to blood or other potentially infectious material, the area should be immediately washed with soap and water for at least 10 seconds and rinsed with 70% isopropyl alcohol. Any exposure incident should be immediately reported to the occupational health department and a determination must be made if blood samples are required from the source patient and the exposed individual and if immediate prophylaxis is indicated. All healthcare personnel should know where to find the exposure control plan that is specific for their place of employment, whom to contact, where to go, and what to do if inadvertently exposed to blood or body fluids. Important resources include the occupational health department, the emergency department, and the infection control/ hospital epidemiology division. The most important recommendation in any accidental exposure is to seek advice and intervention immediately, because the efficacy of recommended prophylaxis regimens is improved with shorter intervals after exposure, such as for hepatitis B immune globulin administration after exposure to the hepatitis B virus or for antiretroviral therapy after percutaneous exposure to HIV. Chemoprophylaxis following exposure to HIV-infected material is most effective if initiated within 4 hours of exposure.60 Additionally, reporting a work-related exposure is required for subsequent medical care and workers’ compensation.

Transmission-Based Precautions Transmission-based Precautions are designed for patients with documented or suspected infection with pathogens for which

CHAPTER

2

17

additional precautions beyond Standard Precautions are needed to prevent transmission. The three categories of Transmission-based Precautions are Contact Precautions, Droplet Precautions, and Airborne Precautions and are based on the likely routes of transmission of specific infectious agents. They may be combined for infectious agents that have more than one route of transmission. Whether used singly or in combination, they are always used in addition to Standard Precautions. Since the infectious agent is often not known at the time of admission, Transmission-based Precautions are applied based on the clinical presentation and the most likely pathogens – so-called Empiric or Syndromic precautions. This approach is especially useful for emerging agents (e.g., SARS-CoV, avian influenza, pandemic influenza), for which information concerning routes of transmission is still evolving. The categories of clinical presentation are as follows: diarrhea, central nervous system, generalized rash/exanthem, respiratory, skin or wound infection. Singlepatient rooms are always preferred for children needing Transmissionbased Precautions. If unavailable, cohorting of patients, and in some cases of staff, according to clinical diagnosis is recommended. Table 2-4 lists the three categories of isolation based on routes of transmission and the necessary components. Table 2-5 lists precautions by syndromes, to be used when a patient has an infectious disease and the agent is not yet identified. It should be noted that for infectious agents that are more likely to be transmitted by the droplet route except during an aerosol-producing procedure (e.g., pandemic influenza), N95 or higher respirators are indicated during the procedure, but an AIIR is not necessarily needed (www.pandemicflu.gov/ plan/healthcare/maskguidancehc.html).

ENVIRONMENTAL MEASURES Contaminated environmental surfaces and noncritical medical items have been implicated in transmission of several healthcare-associated pathogens, including VRE, C. difficile, Acinetobacter sp., MRSA, and RSV.1,9,61 Pathogens on surfaces are transferred to the hands of healthcare personnel and then transferred to other patients or items. Pathogens with a gastrointestinal tract reservoir, including MRSA, are especially likely to contaminate surfaces when the patient has diarrhea; surfaces surrounding such patients may need to be cleaned and disinfected repeatedly. Frequently touched surfaces and those closest to the patient are most likely to be contaminated (e.g., bedrails, bedside tables, commodes, doorknobs, sinks, surfaces, and equipment in close proximity to the patient). Most often, the failure to follow recommended procedures for cleaning and disinfection contributes more than the specific agent to the environmental reservoir of pathogens during outbreaks. In an educational and observational intervention that targeted a defined group of housekeeping personnel, there was a persistent decrease in the acquisition of VRE in a medical ICU; therefore, monitoring for adherence to recommended environmental cleaning practices is an important determinant of success in controlling transmission of MDROs and other environmental pathogens.62 Certain infectious agents (e.g., rotavirus, noroviruses, C. difficile) may be resistant to some routinely used hospital disinfectants; thus, when there is ongoing transmission and cleaning procedures have been observed to be appropriate, a 1:10 dilution of 5.25% sodium hypochlorite (household bleach) or other special disinfectants may be indicated.9 Pediatric facilities should use disinfectants active against rotavirus.63

VISITATION POLICIES Since acquisition of a seemingly innocuous viral infection in neonates and in children with underlying diseases can result in unnecessary evaluation and empirical therapy for septicemia as well as serious lifethreatening disease, special visitation policies are required in pediatric units, especially the high-risk units. All visitors with signs or symptoms of respiratory or gastrointestinal tract infection should be

18

SECTION

A Epidemiology & Control of Infectious Diseases

TABLE 2-4. Transmission-Based Precautionsa Component

Contact

Droplet

Airborne

Hand hygiene

Per Standard Precautions Soap and water preferred over alcohol handrub for Clostridium difficile, Bacillus anthracis spores

Per Standard Precautions

Per Standard Precautions

Gown

Yes. Don before room entry

Per Standard Precautions

Per Standard Precautions and, if infectious, draining skin lesions present

Gloves

Yes. Don before room entry

Per Standard Precautions

Per Standard Precautions

Mask

Per Standard Precautions

Yes. Don before room entry

N95 particulate respirator or higher

Goggles/face shield

Per Standard Precautions

Per Standard Precautions. Always for SARS, avian influenza

Per Standard Precautions Always for SARS, avian influenza

N95 or higher respirator

When aerosol-producing procedures performed for influenza, SARS, VHF

When aerosol-producing procedures performed for influenza, SARS, VHF

Yes. Don upon entry

Room placement

Single-patient room preferred. Cohort like-infections if single-patient rooms unavailable

Single-patient room preferred. Cohort like-infections if single-patient rooms unavailable

Single-patient room. Negative air pressure; 12 air changes/hour for new construction, 6 air changes/hour for existing rooms

Environmental measures

Increased frequency, especially in the presence of diarrhea. transmission of Clostridium difficile, norovirus

Routine

Routine

Transport

Mask patient if coughing. Cover infectious skin lesions

Mask patient

Mask patient Cover infectious skin lesions

SARS, severe acute respiratory syndrome; VHF, viral hemorrhagic fever. a An addition to Standard precautions, use Transmission-based Precautions, use Transmission-based Precautions for patients with highly transmisible or epidemiologically important pathogens for which additional precautions are needed.

restricted from visiting patients in healthcare facilities. During the influenza season, it is preferred for all visitors to have received influenza vaccine. Increased restrictions may be needed in the midst of a community outbreak (e.g., SARS, influenza). For patients requiring Contact Precautions, the use of personal protective equipment by visitors is determined by the nature of the interaction with the patient and the likelihood that the visitor will frequent common areas on the patient unit or interact with other patients and their family members. Although most pediatricians encourage visits by siblings in inpatient areas, the medical risk must not outweigh the psychosocial benefit. Studies demonstrate that parents favorably regard sibling visitation64 and that bacterial colonization65,66 or subsequent infection67 does not increase in the neonate or older child who has been visited by siblings, but these studies are limited by small numbers. Strict guidelines for sibling visitation should be established and enforced in an effort to maximize visitation opportunities and minimize risks of transmission of infectious agents. The following recommendations regarding visitation may guide policy development: 1. Sibling visitation is encouraged in the well-child nursery and NICU, as well as in areas for care of older children. 2. Before visitation, parents should be interviewed by a trained staff nurse concerning the current health status of the sibling. Siblings who are visiting should have received all vaccines recommended for age. Children with fever or symptoms of an acute illness such as upper respiratory tract infection, gastroenteritis, or dermatitis should not be allowed to visit. Siblings who have been exposed to a known infectious disease and are still within the incubation period should not be allowed to visit. After the interview, the physician or nurse should place a written consent for sibling visitation in the permanent patient record and a name tag indicating that the sibling has been approved for visitation for that day. 3. Asymptomatic siblings who have been recently exposed to varicella but have been previously immunized can be assumed to be immune.

4. The visiting sibling should visit only his or her sibling and not be allowed in playrooms with groups of patients. 5. Visitation should be limited to periods of time that ensure adequate screening, observation, and monitoring of visitors by medical and nursing staffs. 6. Children should observe hand hygiene before and after contact with the patient. 7. During the entire visit, sibling activity should be supervised by parents or another responsible adult.

PETS Many zoonoses and infections are attributable to animal exposure. Most of these infections result from inoculation of animal flora through a bite or scratch or self-inoculation after contact with the animal, animal secretions or excretions, or contaminated environment. No universal guidelines for hospital pet visitation have been developed because there are no controlled experiences upon which evidencebased recommendations can be made. This topic is reviewed in the Guidelines for Environmental Infection Control in Health-Care Facilities and recommendations are provided to guide institutional policies.9 Pets can be of significant clinical benefit to the child hospitalized for prolonged periods, and many centers have created their own pet visitation guidelines. Prudent visitation policies should include limiting visitation to animals that meet the following requirements: (1) they are domesticated; (2) they do not require a water environment; (3) they do not bite or scratch; (4) they can be brought to the hospital in a carrier or easily walked on a leash; (5) they are trained to defecate and urinate outside or in appropriate litter boxes; (6) they can be bathed before visitation; and (7) they are known to be free of respiratory, dermatologic, and gastrointestinal tract disease. Reptiles should be excluded due to the risk of transmission of Salmonella sp. and development of severe invasive disease in young infants68 and exotic animals that are

PART I Understanding, Controlling, and Preventing Infectious Diseases

Pediatric Infection Prevention and Control

CHAPTER

2

19

TABLE 2-5. Clinical Syndromes or Conditions Warranting Empiric Transmission-Based Precautions in Addition to Standard Precautions Pending Confirmation of Diagnosisa Clinical Syndrome or Conditionb

Potential Pathogensc

Empiric Precautions (Always Includes Standard Precautions)

Acute diarrhea with a likely infectious cause in an incontinent or diapered patient

Enteric pathogensd

Contact Precautions (pediatrics and adult)

MENINGITIS

Neisseria meningitidis

Droplet Precautions for first 24 hours of antimicrobial therapy; mask and face protection for intubation Contact Precautions for infants and children Airborne Precautions if pulmonary infiltrate Airborne Precautions plus Contact Precautions if potentially infectious draining body fluid present

DIARRHEA

Enteroviruses Mycobacterium tuberculosis

RASH OR EXANTHEMS, GENERALIZED, ETIOLOGY UNKNOWN

Petechial/ecchymotic with fever (general) If traveled in an area with an ongoing outbreak of VHF in the 10 days before onset of fever

Neisseria meningitidis Ebola, Lassa, Marburg viruses

Droplet Precautions for first 24 hours of antimicrobial therapy Droplet Precautions plus Contact Precautions, with face/eye protection, emphasizing safety sharps and barrier precautions when blood exposure likely. Use N95 or higher respiratory protection when aerosol-generating procedure performed

Vesicular

Varicella-zoster, herpes simplex, variola (smallpox), vaccinia viruses

Airborne plus Contact Precautions. Contact Precautions only if herpes simplex, localized zoster in an immunocompetent host, or vaccinia viruses most likely

Maculopapular with cough, coryza, and fever

Rubeola (measles) virus

Airborne Precautions

Cough/fever/upper-lobe pulmonary infiltrate in an HIV-negative patient or a patient at low risk for HIV infection

Mycobacterium tuberculosis, respiratory viruses, Streptococcus pneumoniae, Staphylococcus aureus (MSSA or MRSA)

Airborne Precautions plus Contact Precautions

Cough/fever/pulmonary infiltrate in any lung location in an HIV-infected patient or a patient at high risk for HIV infection

Mycobacterium tuberculosis, respiratory viruses, Streptococcus pneumoniae, Staphylococcus aureus (MSSA or MRSA)

Airborne Precautions plus Contact Precautions Use eye/face protection if aerosol-generating procedure performed or contact with respiratory secretions anticipated. If tuberculosis is unlikely and there are no AIIRs and/or respirators available, use Droplet Precautions instead of airborne precautions. Tuberculosis more likely in HIV-infected than in HIV-negative individuals

Cough/fever/pulmonary infiltrate in any lung location in a patient with a history of recent travel (10–21 days) to country with outbreak of SARS, avian influenza

Mycobacterium tuberculosis, severe acute respiratory syndrome virus–coronavirus (SARS-CoV), avian influenza

Airborne plus Contact Precautions plus eye protection. If SARS and tuberculosis unlikely, use Droplet Precautions instead of Airborne Precautions

RESPIRATORY INFECTIONS

Respiratory infections, particularly bronchiolitis and Respiratory syncytial virus, pneumonia, in infants and young children parainfluenza virus, adenovirus, influenza virus, human metapneumovirus

Contact Precautions plus Droplet Precautions; Droplet Precautions may be discontinued when adenovirus and influenza have been ruled out.

SKIN OR WOUND INFECTION

Abscess or draining wound that cannot be covered

Staphylococcus aureus (MSSA or MRSA), group A streptococcus

Contact Precautions Add droplet precautions for the first 24 hours of appropriate antimicrobial therapy if invasive group A streptococcal disease is suspected

AIIR, airborne infection isolation room; HIV, human immunodeficiency virus; MRSA, methicillin-resistant Staphylococcus aureus; MSSA, methicillin-susceptible Staphylococcus aureus; VHF, viral hemorrhagic fever. a Infection control professionals should modify or adapt this table according to local conditions. To ensure that appropriate empiric precautions are always implemented, hospitals must have systems in place to evaluate patients routinely according to these criteria as part of their preadmission and admission care. b Patients with the syndromes or conditions listed may have atypical signs or symptoms (e.g., neonates and adults with pertussis may not have paroxysmal or severe cough). The clinician’s index of suspicion should be guided by the prevalence of specific conditions in the community, as well as clinical judgment. c The organisms listed under the column “Potential Pathogens” are not intended to represent the complete, or even most likely, diagnoses, but rather possible etiologic agents that require additional precautions beyond standard precautions until they can be ruled out. d These pathogens include enterohemorrhagic Escherichia coli O157:H7, Shigella spp., hepatitis A virus, noroviruses, rotavirus, Clostridium difficile.

imported should be excluded because of unpredictable behavior and the potential for transmission of unusual pathogens (e.g., monkeypox).69 Visitation should be limited to short periods of time and confined to designated areas. Visiting pets should have a certificate of immunization from a licensed veterinarian. Children should observe hand hygiene after contact with pets. Most pediatric facilities restrict pet interaction with severely immunosuppressed patients and those in ICUs.

DISINFECTION, STERILIZATION, AND REMOVAL OF INFECTIOUS WASTE The topic of disinfection and sterilization as it relates to infection prevention and control has been reviewed.70,71 Cleaning is the removal of all foreign material from surfaces and objects. This process is accomplished using soap and enzymatic products. Failure to remove all organic material from items before disinfection and sterilization

20

SECTION

A Epidemiology & Control of Infectious Diseases

will reduce the effectiveness of those processes. Disinfection is a process that eliminates all forms of microbial life except the endospore. Disinfection usually requires liquid chemicals. The ability of an inanimate surface or object to be disinfected can be adversely affected by the presence of organic matter; a high level of microbial contamination; too dilute germicide; inadequate disinfection time; an object that can harbor microbes in protected cracks, crevices, and hinges; and pH and temperature. Sterilization is the eradication of all forms of microbial life, including fungal and bacterial spores. Sterilization is achieved by physical and chemical processes such as steam under pressure, dry heat, ethylene oxide, and liquid chemicals. Patient care equipment was originally categorized by Spaulding72 and subsequently by the CDC71 as critical, semicritical, and noncritical items with regard to sterilization and disinfection. Critical items require sterilization because they enter sterile body tissues and would have a high risk of causing infection if contaminated; semicritical objects require disinfection because they may contact mucous membranes and nonintact skin; and noncritical items require routine cleaning because they only come in contact with intact skin. If noncritical items used on patients requiring Transmission-based Precautions, especially Contact Precautions, must be shared, these items should be disinfected after use on a patient who is under isolation precautions. Guidelines for specific objects and specific disinfectants are published and updated by the CDC. Multiple published reports and manufacturers similarly recommend the use and reuse of objects with appropriate sterilization, disinfection, or cleaning recommendations. Recommendations in guidelines for reprocessing endoscopes focus on training of personnel, meticulous manual cleaning, high-level disinfection followed by rinsing, air-drying, and proper storage to avoid contamination.73 Medical devices that are designed for single use (e.g., specialized catheters, electrodes, biopsy needles) must be reprocessed by third parties or hospitals according to the guidance issued by the Food and Drug Administration (FDA) in August, 2000 with amendments in September, 2006; such reprocessors will be considered “manufacturers” and will be regulated in the same manner. Available data show that single-use devices reprocessed according to the FDA regulatory requirements are as safe and effective as new devices (www.fda.gov/cdrh/reprocessing). Healthcare facility waste is all biologic or nonbiologic waste that is discarded and not intended for further use. Medical waste is the material generated as a result of use with a patient, such as for diagnosis, immunization, or treatment, and includes soiled dressings and intravenous tubing. Infectious waste is that portion of medical waste that could potentially transmit an infectious disease. Microbiologic waste, pathologic waste, contaminated animal carcasses, blood, and sharps are all examples of the estimated infectious waste discarded in the United States each day. Methods of effective disposal of infectious waste include incineration, steam sterilization, drainage to a sanitary sewer, mechanical disinfection, chemical disinfection, and microwave. State regulations guide the treatment and disposal of regulated medical waste. Recommendations for developing and maintaining a program within a facility for safe management of medical waste can be found in the Guidelines for Environmental Infection Control in Health-Care Facilities.9

OCCUPATIONAL HEALTH Occupational health and student health collaboration with the infection control division of a hospital is required by OSHA57 and is important for a successful infection control program. The occupational health program is of paramount importance in hospitals caring for children because healthcare personnel are at increased risk of infection for various reasons, including the following: (1) children have a high incidence of infectious diseases; (2) personnel may be susceptible to many of the infecting pathogens; (3) pediatric care requires close contact; (4) children lack good personal hygiene; (5) infected children may be asymptomatic; and (6) healthcare personnel are exposed to multiple family members who may also be infected.

The occupational health department should be an educational resource for information on infectious pathogens in the healthcare workplace. Occupational health, in concert with the infection control service, should provide pre-employment education and respirator fit testing; annual retraining for all employees regarding routine health maintenance, available vaccines, standard precautions and isolation categories, and exposure plans; and screening for tuberculosis at regular intervals, as determined by the facility’s risk assessment.74 With new pathogens being isolated, new diseases and their transmission described, and new prophylactic regimens and treatment available, it is mandatory that personnel have an up-to-date working knowledge of infection control and know where and what the available services, equipment, and therapies are for the healthcare worker. Many educational resources are available to assist with employee education.75 All healthcare personnel should be screened by history or serologic testing, or both, to document their immune status to specific agents, and immunization should be provided for the following for all employees who are nonimmune and who do not have contraindications to receiving the vaccine: diphtheria, tetanus, hepatitis B virus, influenza (yearly), mumps, poliomyelitis, rubella, rubeola, and varicella. Providing vaccines at no cost to healthcare personnel increases acceptance. Recognition of the importance of raising influenza vaccination rates among healthcare personnel to protect the individuals, their patients, and their family members led to the publication of evidence-based recommendations in 2006.76 After licensure of the adolescent/adult pertussis (Tdap) vaccine in 2005, the Advisory Committee on Immunization Practices and the AAP recommend administration of a single dose of Tdap to all healthcare personnel in whom > 2 years have elapsed since the most recent Td booster in order to prevent pertussis in the vaccine recipient and to prevent transmission of pertussis to high-risk patients.77

Special Concerns of Healthcare Personnel Healthcare personnel who have underlying medical conditions (e.g., hypertension, diabetes, obesity, tobacco use) should be able to obtain general information on wellness and screening when needed from the occupational health service. Healthcare providers with direct patient contact who have infants younger than 1 year of age at home are concerned about acquiring infectious agents from patients and transmitting them to their susceptible children. An immune healthcare worker who is exposed to varicella does not become a silent “carrier” of this pathogen. However, pathogens to which the healthcare worker is partially immune or nonimmune can cause a severe, mild, or asymptomatic infection in the employee that can be transmitted to family members. Examples include influenza, pertussis, RSV, rotavirus, and tuberculosis. Important preventive procedures for healthcare workers with infants at home are: (1) consistent observance of Standard Precautions, Transmission-based Precautions, and hand hygiene according to published recommendations1,37; (2) annual influenza immunization; (3) routine tuberculosis screening; (4) assurance of immunity or immunization against poliomyelitis, measles, mumps, hepatitis B, rubella, and pertussis (Tdap); (5) early medical evaluation for infectious illnesses; (6) routine, on-time immunization of infants; and (7) prompt initiation of prophylaxis or therapy following exposure or certain infections. The healthcare worker who is, could be, or anticipates becoming pregnant should feel comfortable working in the healthcare workplace. In fact, with Standard Precautions and appropriate adherence to environmental cleaning and isolation precautions, the vigilant healthcare worker can be at less risk than a preschool teacher, childcare provider, or mother of children with many playmates in the home. Pathogens of potential concern to the pregnant healthcare worker include cytomegalovirus, hepatitis B virus, influenza, measles, mumps, parvovirus B19, rubella, varicella-zoster virus, and Mycobacterium tuberculosis. Important preventive procedures include documentation of immunity or immunization before pregnancy for rubella, mumps, measles, poliomyelitis, and hepatitis B virus; annual influenza

PART I Understanding, Controlling, and Preventing Infectious Diseases

Pediatric Infection Prevention and Control

vaccine; routine tuberculosis screening; early medical evaluation for infectious illnesses; and prompt prophylaxis or therapy if exposed to or infected with certain pathogens. It is important to note that pregnancy is an indication for influenza vaccine to prevent the increased risk of serious disease and hospitalization that occurs in second- and third-trimester women who develop influenza infection. Pregnant workers should assume that all patients are potentially infected with cytomegalovirus and other “silent” pathogens and should use gloves (followed by hand hygiene) when handling body fluids, secretions, and excretions. Table 2-6 summarizes the information about infectious agents that are relevant to the pregnant woman and a comprehensive review has been published.78

INFECTION CONTROL IN THE AMBULATORY SETTING Because most patient visits are in the ambulatory setting and more patients who were formerly admitted to acute care hospitals are being cared for in these settings, it becomes important to establish and

CHAPTER

2

maintain rigorous infection control practices in the outpatient environment. The risk of HAIs in ambulatory settings has been reviewed79,80 and has been associated with lack of adherence to routine infection control practices and procedures, especially recommended safe injection practices.59 Respiratory viral agents and M. tuberculosis are among the infectious agents transmitted in ambulatory settings. Crowded waiting rooms, toys, furniture, lack of isolation of children undiagnosed and waiting to be seen, contaminated hands, contaminated secretions, and susceptible healthcare workers are only some of the factors that result in sporadic and epidemic illness in outpatient settings. Transmission of MRSA and VRE in outpatient settings has not been reported, but the association of CA-MRSA in healthcare personnel working in an outpatient HIV clinic with environmental CA-MRSA contamination of that clinic indicates the potential for transmission in this setting.81 Patient-to-patient transmission of Burkholderia species and Pseudomonas aeruginosa in outpatient clinics for adults and children with cystic fibrosis has been confirmed and prevented by implementing recommended infection control practices.82,83 Outpatient infection control guidelines and policies for

TABLE 2-6. The Pregnant Healthcare Worker: Guide to Management of Occupational Exposure to Selected Infectious Agentsa Agent

In-Hospital Source

Bioweapons Agents, Category A Smallpox (vaccinia) Respiratory secretions, contents of pustulovesicular lesions

21

Potential Effect on the Fetus

Rate of Perinatal Transmission

Fetal vaccinia, premature delivery, spontaneous abortion, and perinatal death

Maternal Screening

Prevention

Limited data

History of successful vaccination with “take” within previous 5 years

Pre-event vaccination contraindicated during pregnancy. Vaccine and vaccinia-immune globulin (VIG) after exposure; pre-exposure vaccine only if smallpox present in the community and exposure to patients with smallpox likely. Airborne plus Contact Precautions

Primary infection (25–50%); recurrent infection (52%); symptomatic (< 5–15%)

Routine screening not recommended; antibody is incompletely protective

Efficacy of CMV immune globulin not established. No vaccine available. Standard Precautions

Cytomegalovirus (CMV)

Urine, blood, semen, vaginal secretion, immunosuppressed, transplant, dialysis, day care

Classic diseaseb (5–10%); hearing loss (10–15%)

Hepatitis A (HAV)

Feces (most common), blood (rare)

No fetal transmission Unknown described; transmission can occur at the time of delivery if mother still in the infectious phase and can cause hepatitis in the infant

Routine screening not recommended

Vaccine is a killed viral vaccine and can safely be used in pregnancy. Contact Precautions during acute phase

Hepatitis B (HBV)

Blood, bodily fluids, vaginal secretions, semen

Hepatitis, early-onset HBeAg– and HBsAg+ hepatocellular carcinoma (10%) HBeAg+ and HbsAg+ (90%)

Routine HBsAg testing advised

HBV vaccine during pregnancy if indications exist. Neonate: HBIG plus vaccine at birth. Standard Precautions

Hepatitis C (HCV)

Blood, vaginal secretions, semen

Hepatitis

5% (0–25%)

Routine screening not recommended

No vaccine or immune globulin available; postexposure treatment with antiviral agents investigational. Standard Precautions

Herpes simplex virus (HSV)

Vesicular fluid, oropharyngeal and vaginal secretions

Sepsis, encephalitis, meningitis, mucocutaneous lesions, congenital malformation (rare)

Primary genital (33–50%) Recurrent genital (1–2%)

Antibody testing minimally useful. Genital inspection for lesions if in labor

Chemoprophylaxis at 36 weeks decreases shedding. Standard precautions. Contact Precautions for patients with mucocutaneous lesions

22

SECTION

A Epidemiology & Control of Infectious Diseases

TABLE 2-6. The Pregnant Healthcare Worker: Guide to Management of Occupational Exposure to Selected Infectious Agentsa—Continued Agent

In-Hospital Source

Human immunodeficiency virus (HIV)

Blood, bodily fluids, vaginal secretions, semen

Influenza

Potential Effect on the Fetus

Rate of Perinatal Transmission

Maternal Screening

Prevention

No congenital syndrome. Depends on HIV viral If fetus infected, AIDS load and use of in 2–4 years antiretroviral agents during pregnancy, labor and postnatally in the infant If viral load < 1000 (rate 2%) If viral load ≥10 000 (rate up to 25%)

Routine maternal screening advised. If exposed, testing every 3 months

Antiretroviral chemoprophylaxis for exposures; intrapartum postnatal chemoprophylaxis for HIV+ mothers and their infants indicated to prevent perienatal transmission. Standard Precautions

Sneezing and coughing, respiratory tract secretions

No congenital syndrome Rare (influenza in mother could cause hypoxia in fetus)

None

Trivalent inactivated vaccine (TIV) for all pregnant women during influenza season to decrease risk of hospitalizations for cardiopulmonary complications in mother. No risk if exposed to individuals who received live attenuated influenza vaccine (LAIV). Droplet Precautions. Add Contact Precautions for young infants

Rubeola (measles)

Respiratory secretions, coughing

Prematurity, Rare spontaneous abortion; no congenital syndrome

Antibody test, MDdocumented disease, or 2 doses of measles containing vaccine at or > 12 months of age

Vaccine. Airborne Precautions

Parvovirus B19

Respiratory secretion, blood, immunocompromised patients

Fetal hydrops, stillbirth; Approximately 25%; no congenital syndrome fetal death < 10%

No routine screening. B19 DNA can be detected in serum, leukocytes, respiratory secretions, urine, tissue specimens

No vaccine. Defer care of immunocompromised patients with chronic anemia when possible. Droplet Precautions

Rubella

Respiratory secretions

Congenital syndrome

90% in first trimester; 40–50% overall

Routine rubella IgG testing in pregnancy. Preconceptional screening recommended

Vacccine. No congenital rubella syndrome described for vaccine. Droplet Precautions Contact Precautions for patients with congenital rubella

Treponemia pallidum (syphilis)

Blood, lesion, fluid, amniotic fluid

Congenital syndrome

Variable 10–90%; depends upon stage of maternal disease and trimester of the infection

VDRL, RPR FTA-ABS

Postexposure prophylaxis with penicillin. Standard Precautions; wear gloves when handling infant or caring for patients with primary syphilis with mucocutaneous lesions until completion of 24 hours of treatment.

Mycobacterium tuberculosis

Sputum, skin lesions

Neonatal tuberculosis; liver most frequently infected

Rare

Skin test: PPD Chest radiograph

Varies with PPD reaction size and chest radiograph result; therapy for active disease during pregnancy. Airborne Precautions Contact Precautions if draining skin lesions

PART I Understanding, Controlling, and Preventing Infectious Diseases

Infections Associated with Group Childcare

CHAPTER

3

23

TABLE 2-6. The Pregnant Healthcare Worker: Guide to Management of Occupational Exposure to Selected Infectious Agentsa—Continued Agent

In-Hospital Source

Varicella-zoster virus

Respiratory secretion, vesicle fluid

Potential Effect on the Fetus

Rate of Perinatal Transmission

Maternal Screening

Prevention

Malformations, skin, limb, central nervous system, eye. Disseminated or localized disease

Congenital syndrome (2%)

Varicella IgG serology; history 90% correct

Vaccinec; VariZIG within 96 hours of exposure if susceptible. Airborne plus Contact Precautions

AIDS, acquired immunodeficiency syndrome; FTA-ABS, fluorescent treponemal antigen-antibody test; HBeAg, hepatitis B e antigen; HBIG, hepatitis B immune globulin; HBsAg, hepatitis B surface antigen; IgG, immunoglobulin G; PPD, purified protein derivative; RPR, rapid plasma reagin test; VDRL, Venereal Disease Research Laboratory test. a Employment, prepregnancy screening/vaccination is primary prevention for certain agents. Annual immunization for influenza is primary prevention. b Congenital syndrome: varying combinations of jaundice, hepatosplenomegaly, microcephaly, thrombocytopenia, anemia, retinopathy, skin, and bone lesions. c Live virus vaccine given before or after pregnancy. d See Chapter 205, Varicella-zoster Virus.

pediatricians’ offices have been published and are being updated.84 Prevention strategies include definition of policies, education, and strict adherence to guidelines.

CHAPTER

3

Infections Associated with Group Childcare Andi L. Shane and Larry K. Pickering

On average 11.6 million (63%) children 5 years of age or younger and 53% of children 5 to 14 years of age were in some form of group childcare on a regular basis in the winter of 2002 (National Child Care Information Center Child Care Bureau. http://nccic.org).1 Aggregation of young children potentiates transmission of organisms that can produce disease in other children, adult care providers, parents, and community contacts. Group childcare settings may potentiate increased frequency of certain diseases, the occurrence of outbreaks of illness (Table 3-1), greater severity of illness, an increase in antibiotic use to permit earlier return to care, which results in the potential for emergence of resistant organisms, and an increased economic burden to individuals and society.2–4 The extent of illness resulting from interaction of children and adults in group childcare depends on the age and immune status of children and adults involved, season, environmental characteristics of the childcare facility, and inoculum size and virulence potential of microbes. Children newly entered into a childcare program are at especially high risk of enteric and respiratory tract infections,5–10 but as a consequence of these infections may be protected against respiratory tract viral infections and reactive airway diseases during subsequent years.11 Children who are exposed to infectious pathogens of siblings and contacts in group care and often manifest clinical symptoms of frequent infections early in life may be protected against developing atopic disease in later childhood.12

TYPES OF GROUP CHILDCARE The United States Census Bureau classifies regular preschool childcare arrangements by provider (relative of a child in care versus nonrelative) and location of care. Of the 63% of preschool children in a regular childcare arrangement in the winter of 2002, 40% regularly

received care by a relative, 37% by nonrelatives, and 11% by both relatives and nonrelatives. The remaining 12% were not classified as receiving a form of childcare regularly. Nonrelative care can be further divided into provision of care in an organized care facility or childcare center (23%), by a nonrelative in the child’s home (4%) or in the provider’s home (10%).1 Types of facilities can also be classified by size of enrollment, age of enrollees, and environmental characteristics of the facility. Grouping of children by age varies by setting but in organized care facilities usually children are separated as infants (6 weeks to 12 months), toddlers (13 to 35 months), preschool (36 months to 59 months), and school-aged children (5 to 12 years). The classification of group childcare settings has relevance to infectious disease epidemiology with regard to regulation and monitoring. Most nonrelative care provided in an organized care facility is subject to state licensing and regulation, whereas relative care in a child or provider’s home may not be subject to state regulations and monitoring.

EPIDEMIOLOGY AND ETIOLOGY OF INFECTIONS Although almost any infectious disease has the propensity to propagate in the childcare setting, diseases shown in Table 3-1 are commonly associated with outbreaks. Organisms that infect enrollees and providers may do so with a predilection for nonimmune persons of specific ages.

Enteric Infections Outbreaks of diarrhea occur at a rate of approximately 3 per year per childcare center and are most frequently associated with organisms that result in infection after ingestion of a low inoculum. These organisms generally are transmitted from person to person13,14 and include rotavirus, sapovirus, norovirus, astrovirus, enteric adenovirus, Giardia lamblia, Cryptosporidium, Aeromonas, Shigella, Escherichia coli O157:H7, E. coli O114, enteropathogenic E. coli, and Clostridium difficile.14–26 These fecal coliforms27,28 and enteric viruses contaminate the environment;29 contamination rates are highest during outbreaks of diarrhea. The attack rates and frequency of asymptomatic excretion of these organisms in children in childcare are shown in Table 3-2. Reported attack rates depend on several factors, including methods used for organism detection.22,23 Enteric viruses are the predominant etiology of diarrheal syndromes among children in group care, with impact by season. In a prospective study of children enrolled in childcare in Denmark during 6 months of winter, rotavirus was the predominant organism identified in 40% of cases with a confirmed etiology, sapoviruses in 18%, and astroviruses in 7%.30 Organisms generally associated with foodborne outbreaks, such as Salmonella and Campylobacter jejuni, are infrequently associated with diarrhea in the childcare setting. How-

24

SECTION

A Epidemiology & Control of Infectious Diseases

TABLE 3-1. Association of Infectious Diseases with Group Childcare Settings Disease or Infection

Risk Factors and Association with Outbreaks

Enteric

Close person-to-person contact, fecal-oral contact, food preparation practices, and suboptimal hand hygiene

Viral Rotaviruses, enteric adenoviruses, astroviruses, noroviruses, hepatitis A virus (HAV)

Commonly associated with outbreaks HAV and rotavirus are vaccine-preventable

Bacterial Shigella, Escherichia coli O157:H7 Campylobacter spp., Salmonella spp., Clostridium difficile

Commonly associated with outbreaks Less commonly associated with outbreaks

Parasitic Giardia lamblia, Cryptosporidium parvum

Commonly associated with outbreaks

Respiratory tract (acute upper and lower respiratory tract infections and invasive disease) Bacterial Haemophilus influenzae type b (Hib) Streptococcus pneumoniae

Aerosolization and respiratory droplets, person-to-person contact, suboptimal hand hygiene

Few outbreaks; Hib is vaccine-preventable Few outbreaks; invasive Streptococcus pneumoniae caused by serotypes in vaccine is vaccinepreventable Few outbreaks and low risk of secondary cases Few outbreaks; N. meningitidis caused by serogroups in vaccine is vaccine-preventable in persons over 2 years of age Increasingly associated with outbreaks in childcare centers and schools; vaccine-preventable Occasional outbreaks, usually as a result of contact with an infectious adult care provider Outbreaks rare; oropharynx usual habitat; usually manifest as arthritis and osteomyelitis

Group A streptococcus Neisseria meningitidis Bordetella pertussis Mycobacterium tuberculosis Kingella kingae Viral Rhinoviruses, parainfluenza, influenza, respiratory syncytial virus (RSV), respiratory adenoviruses, influenza, metapneumoviruses

Disease usually caused by same organisms circulating in the community; influenza is vaccinepreventable in children ≥ 6 months of age

Multiple organ systems Cytomegalovirus

Prevalent asymptomatic excretion with transmission from children to providers

Parvovirus B19

Outbreaks reported; risk to susceptible pregnant women and immunocompromised

Varicella-zoster virus (VZV)

Outbreaks in childcare centers occur. VZV is vaccine-preventable in children ≥ 12 months of age. Zoster lesions present low risk of infection

Herpes simplex virus (HSV)

Low risk of transmission from active lesions and oral secretions

Hepatitis B virus

Rarely occurs in childcare centers; vaccine-preventable

Hepatitis C virus

No documented cases of transmission in the childcare setting

Human immunodeficiency virus (HIV)

No documented cases of transmission in the childcare setting

Skin

Close person-to-person contact

Staphylococcal and streptococcal impetigo

Transmission increased by close person-to-person contact with lesions; outbreaks less likely with decreased incidence of varicella infections; methicillin-resistant Staphylococcus aureus (MRSA) disease increasing

Scabies

Outbreaks in group childcare reported

Pediculosis

Common in children attending group childcare

Ringworm

Tinea corporis and T. capitis outbreaks associated with childcare

Conjunctiva

Outbreaks in group childcare reported with both bacterial and viral etiologies

ever, a report of an outbreak of diarrhea in 14 of 67 (21%) exposed children and adult care providers associated with ingestion of fried rice contaminated with Bacillus cereus31 highlights the fact that foodborne outbreaks can occur in the childcare setting, especially when food is prepared and served at the center. Bacterial pathogens that have the potential to cause severe systemic infections, including E. coli O157:H7, have been associated with fecal–oral transmission in group childcare settings. An outbreak of this pathogen in a childcare center in Alberta, Canada in June 2002 likely began following introduction of the organism by a 3-year-old enrollee with farm animal contact who developed hemolytic–uremic syndrome. A diarrheal attack rate of 23% was noted among enrollees, which is comparable with attack rates of E. coli O157:H7 reported

during previous childcare-associated outbreaks. Prolonged asymptomatic shedding and subclinical cases in concert with poor hygiene and toileting practices likely contributed to propagation of the outbreak.32 Shigella sonnei has been responsible for periodic multicommunity outbreaks in group childcare. A multicommunity outbreak of over 1600 culture-confirmed cases in the greater metropolitan area of Cincinnati, Ohio from May to September, 2001 had an overall mean attack rate of 10% among childcare center enrollees, with highest attack rates occurring among newly or incompletely toilet-trained enrollees and lowest attack rates occurring among diapered children. Attack rate was 6% among staff. An epidemiologic investigation revealed that a single negative stool culture may be sufficient to confirm the clearance of S. sonnei in convalescent, treated enrollees.

PART I Understanding, Controlling, and Preventing Infectious Diseases

Infections Associated with Group Childcare

TABLE 3-2. Outbreaks of Diarrhea by Organism Organism

Attack Rate (Enrollees) (%)

Secondary Attack Rate (Family Members) (%)

Asymptomatic Excretion (Enrollees)

Rotavirus

50

15–80

Common

Enteric adenovirus

40

Unknown

Common

Astrovirus

50–90

Unknown

Common

Calicivirus

50

Unknown

Common

Giardia lamblia

17–54

15–50

Common

Cryptosporidium

33–74

25–60

Common

Shigella

33–73

25–50

Uncommon

O157:H7

29, 34

Unknown

Uncommon

O114:NM

67

Unknown

Uncommon

O111:K58

56, 94

Unknown

Uncommon

Unknown

Common

Escherichia coli

Clostridium difficile 32

Secondary transmission was facilitated by poor hygiene practices, including inaccessible handwashing supplies and incomplete diaper disposal practices, as well as recreational activities involving water.33 A prolonged multistate increase in shigellosis with similar biochemical and genetic profiles occurred in the south and mid-Atlantic areas from June 2001 to March 2003. A significant proportion of cases were associated with group childcare, emphasizing the ongoing public health challenge of management and control.34 Spread of microbes that cause diarrhea from the childcare setting into families has been reported for many enteropathogens (see Table 3-2). The secondary attack rates range from 15% to 80% depending on the enteropathogen, mode of transmission, and length of time in the household. Children in group childcare are generally the index cases within households. A retrospective evaluation of transmission of infectious gastroenteritis (80% due to rotavirus) in 936 households in northern California revealed a secondary household attack rate of 8.8% (95% confidence interval (CI) 7.9–9.7). Older children in the households had a two- to eightfold greater risk of secondary infection than adults in the household. Clinical illness in secondary household cases usually was less severe and of decreased duration, compared with illness in the index case.35 During outbreaks of diarrhea in childcare centers, asymptomatic excretion of enteropathogens is frequent14,22,23,36–39 (see Table 3-2). During outbreaks associated with enteric viruses and G. lamblia in children younger than 3 years of age, asymptomatic infection occurs in up to 50% of infected children. In one longitudinal study of diarrhea in 82 children younger than 2 years of age in a childcare center, more than 2700 stool specimens were collected on a weekly basis.39 Using enzyme immunoassays, 21 of 27 (78%) children infected with G. lamblia were found to be asymptomatic and 19 of 37 (51%) children infected with rotavirus were asymptomatic. The role that asymptomatic excretion of enteropathogens plays in spread of disease is unknown. Acute infectious diarrhea is two to three times more common in children in childcare than in age-matched children cared for in their homes.5,40,41 Approximately 20% of clinic visits for acute diarrheal illness among children younger than 3 years of age have been shown to be attributable to childcare attendance.5 In addition, diarrheal illness is threefold higher among children in their first month in out-of-home childcare than in children cared for at home.5,6 Several factors have been reported to be associated with occurrence of diarrhea among children in group settings. Although diarrhea occurs 17 times more frequently in diapered children than in children not wearing diapers,42 it is unclear whether diapering is a confounding factor, risk factor, or protective factor in group childcare infections.

CHAPTER

3

25

Children who are diapered are more likely to be younger than children who are not; therefore, higher attack rates may merely represent exposure of a younger, nonimmune cohort. In a multicommunity group childcare outbreak of Shigella sonnei, the highest attack rates were noted in rooms where both toilet-trained and diapered children were combined (14%) compared with rooms with toilet-trained children only (9%) and rooms with only diapered children (5%), despite comparable availability of sinks and toilets.33 A study evaluating costs associated with office visits for diarrhea in children younger than 36 months of age showed that the average cost for each episode of diarrhea was $289 in 1991, with 21% of the total cost for diarrhea in this 1-year study attributable to rotavirus diarrhea.43 A 3-year study analyzing medical claims data for the period from 1993 to 1996 showed that the median cost (in 1998 constant dollars) of a diarrhea-associated hospitalization was $2307, and the median cost of a diarrhea-associated outpatient visit was $47.44

Rotavirus Rotaviruses are the most common etiology of significant symptomatic diarrhea in children less than 2 years of age. Most symptomatic infections occur in infants and children between 4 and 24 months of age and are manifest by profuse, watery diarrhea, preceded by emesis and fever. Infections are primarily transmitted from person to person by the fecal–oral route, and are facilitated by interpersonal contact. Rotavirus can be isolated from human stools for approximately 21 days after illness begins and rotavirus RNA has been detected on toys and surfaces in childcare centers.29 The highest attack rates of rotaviral infections occur in infants and children who may be enrolled in group childcare, with notable transmission rates from infected contacts and significant rates of hospitalization. In one study, children in childcare centers developed predominantly homotypic antibody responses after infection with rotavirus, but as the number of rotavirus infections increased, children developed heterotypic antibody responses to G types at levels that correlate with broad protection against rotavirus infection and illness.45 In another study, infections with rotavirus were associated with lower concentrations of antirotavirus-specific fecal immunoglobulin A (IgA), indicating a protective role for higher titers of antirotavirus-specific fecal IgA.46 Prevention of transmission of rotavirus infections in persons involved in group childcare includes meticulous hand hygiene and disinfection of potentially contaminated surfaces, with processing of soiled diapers and clothing in areas that are inaccessible to mobile children. Primary prevention of rotavirus may be accomplished with administration of two rotaviral vaccines licensed in 2005: a pentavalent bovine–human reassortant vaccine licensed in the United States, or a monovalent human G1 rotavirus vaccine licensed in several countries outside the United States. Large trials of both products have demonstrated both efficacy and safety, and neither vaccine appears to have an association with intussusception in vaccine recipients but this will be monitored closely47,48,48a (see Chapter 216, Rotaviruses).

Hepatitis A Virus Hepatitis A virus (HAV) infections usually are mild or asymptomatic in children. Less than 5% of children younger than 3 years of age and less than 10% of children between 4 and 6 years of age with HAV infection develop jaundice. The first outbreak of HAV in a childcare center was reported in 1973 in North Carolina;49 since then, outbreaks have been recognized throughout the United States.50 Peak viral titers in stool and greatest infectivity occur during the 2 weeks before onset of symptoms. Outbreaks in childcare centers generally are not recognized until illness becomes apparent in older children or adults.50 Prior to availability of hepatitis A vaccine in the United States, approximately 15% of episodes of HAV infection were estimated to be associated with childcare centers. HAV infections are transmitted in

26

SECTION

A Epidemiology & Control of Infectious Diseases

the childcare setting by the fecal–oral route and occur more frequently in settings that include diapered children, although large size and long hours of operation are also risk factors for outbreaks of HAV infection.51 The mainstays for prevention of HAV infection include general measures such as maintenance of personal hygiene, hand hygiene, and disinfecting procedures. Universal administration of two doses of hepatitis A vaccine to all children, beginning at 1 year (12 to 23 months) of age, the two doses administered at least 6 months apart,52 is recommended by the Advisory Committee on Immunization Practices (ACIP) and the American Academy of Pediatrics (AAP).53 Administration of immune globulin during outbreaks for postexposure prophylaxis to unimmunized contacts may be indicated54 (see Chapter 64, Acute Hepatitis; Chapter 238, Hepatitis A Virus). A case-control study to evaluate the effectiveness of a hepatitis A vaccination program targeted at childcare attendees between 2 and 5 years of age found that individuals with direct contact with a childcare center were protected against disease. Furthermore, the 6 times greater risk of hepatitis A that occurred in persons who had contact with a childcare center prior to implementation of the hepatitis A immunization program in Maricopa County, Arizona was not found in the postvaccination case-control study.55 Education and training of staff regarding appropriate hygienic practices, as well as modes of transmission of HAV and other enteric diseases, and frequent monitoring of hygienic practices by center directors are essential components of any preventive plan.

Respiratory Tract Infections Children younger than 2 years of age attending childcare centers have an increased number of upper and lower respiratory tract illnesses compared with age-matched children cared for at home.5,7,56,57 Studies have shown that approximately 10% to 17% of respiratory tract illnesses in United States children younger than 5 years of age are attributable to childcare attendance.58,59 Another prospective cohort study found that 89% of disease episodes among children attending a childcare center are respiratory tract infections.60 In a retrospective cohort study of 2568 children from 1 to 7 years of age, 1-year-old children in childcare centers had an increased risk of the common cold (relative risk (RR), 1.7; 95% CI, 1.4 to 2.0), otitis media (RR, 2.0; 95% CI, 1.6 to 2.5), and pneumonia (RR, 9.7; 95% CI, 2.3 to 40.6).56 Attendance in family childcare did not increase risk. In a prospective cohort study in France comparing the risk of upper respiratory tract illnesses (URTIs) in children in three settings – family, small center and large center – demonstrated increased risk of ≥ 5 URTIs in small center enrollers (OR2.2; 95% CI, 1.4–3.4) and modestly increased risk in large center enrollers (OR1.2; 95% CI, 0.8–1.8). In this study, the risk for children attending large childcare centers was intermediate between children in family childcare homes and smaller childcare centers, probably as a result of segregation of children in large centers into small classrooms. Respiratory tract infections that have been studied in the childcare setting include pharyngitis, sinusitis, otitis media, common cold, bronchiolitis, and pneumonia.7,56–58,60 Organisms responsible for illness in children in childcare settings are similar to organisms that circulate in the community and include respiratory syncytial virus, parainfluenza viruses, adenovirus, rhinovirus, coronavirus, influenza viruses, parvovirus B19, and Streptococcus pneumoniae. Infections due to Bordetella pertussis in the United States have experienced a dramatic increase, with 8296 cases of pertussis reported in 2002 and 25,827 cases reported in 2004.61–63 Incompletely immunized infants under 12 months of age experience significant clinical disease, whereas adolescents and adults remain mildly to moderately symptomatic and infectious, accounting for a significant proportion of cases. In many group childcare arrangements adolescents and adults may be the index case for pertussis infections. In 2005, the AAP and ACIP recommended use of the two Food and Drug Administration (FDA) licensed Tdap vaccines in people, one for adolescents 10

through 18 years of age, the other for people 11 through 64 years of age.62,63 These vaccines may impact disease in people who receive them as well as in susceptible contacts. An adult or adolescent source may also be the index case for Mycobacterium tuberculosis infections in a group childcare setting, with child-to-child transmission occurring infrequently.64,65 An outbreak of tuberculosis (TB) associated with a private-home childcare facility in San Francisco, California occurred between 2002 and 2004. Of 11 outbreak cases, 9 (82%) occurred in children less than 7 years of age; all had extensive contact with the private-home childcare facility, where the adult index patient spent significant time. Two children presented with clinical illness, 3 were identified by contact investigation, and 4 were identified by primary care providers during routine TB screening evaluations. Isolates from 4 of the pediatric patients and 2 of the adult patients shared identical molecular patterns. Thirty-six additional children and adult contacts had latent TB infections.66 High transmissibility of TB among residents of adult daycare centers has also been demonstrated.67 Person-to-person transmission of Chlamydophila pneumoniae among children in the childcare setting has been reported without occurrence of disease.68 Kingella kingae colonizes the oropharynx and respiratory tracts of young children and has been associated with invasive disease.69 The first reported outbreak of invasive K. kingae osteomyelitis/septic arthritis occurred in a childcare center in 2003. Fifteen (13%) children older than 16 months of age were found to be colonized, with 9 children (45%) in the same class as the 2 children with invasive disease. Matching pulse field gel electrophoresis (PFGE) patterns supported child-to-child transmission.70 A retrospective review of osteoarticular infectious etiologies from 1999 to 2002 of 406 hospitalized children in Paris, France demonstrated that K. kingae was isolated from 14% of clinical specimens. This pathogen was isolated more frequently from children younger than 36 months of age and was the second most common bacterial isolate in this population. Awareness of the clinical manifestations, laboratory requirements for growth, and risk factors for acquisition in childcare may account for the increasing incidence of K. kingae infections.71 Group A streptococcal infection among children and adult staff in the childcare setting is not a common problem, but outbreaks have been reported.72–74 In a study of prevalence of group A streptococcus conducted in a childcare center after a fatal case of invasive disease, 25% of 258 children and 8% of 25 providers had group A streptococci isolated from throat cultures.72 Risk of carriage was increased in children who shared the room of the index case (odds ratio (OR), 2.7; 95% CI, 0.8 to 9.4). In a study in childcare centers in Israel, the prevalence of group A streptococcus was 3% in infants and 8% in toddlers. Pharyngeal carriage was not associated with respiratory tract symptoms.75 Group A streptococcal perianal infection and infection associated with varicella have also been reported.69,70 The risk of acute otitis media is significantly increased in children in childcare, especially in children younger than 2 years of age.60,58,76–78 In one study, the incidence rate ratio for otitis media was 1.5 in children in childcare compared with that in children in home care.58 Otitis media is responsible for most antibiotic use in children younger than 3 years of age in the childcare setting. However, implementation of the 2004 acute otitis media treatment guidelines developed by the AAP79 may reduce use of prescription antibiotics for acute otitis media in children enrolled in group childcare. Childcare attendance has also been associated with risk of developing recurrent otitis media (more than 6 episodes in 1 year), as well as chronic otitis media with effusion persisting for more than 6 months.80 The size of the childcare center was an important variable in the occurrence of frequent otitis media in children younger than 12 months of age, varying from 16% in small care groups to 36% in large care groups.77 Genotypically similar strains of nontypable Haemophilus influenzae were isolated from throats of 127 children attending 16 childcare centers in Michigan. Rates of colonization were greater among attendees of childcare centers with > 5 classrooms, and when suboptimal hand hygiene was performed by staff and children. Colonized children who were recipients of a course of antibiotics at

PART I Understanding, Controlling, and Preventing Infectious Diseases

Infections Associated with Group Childcare

the time of culturing were more likely to be colonized with a betalactamase-producing nontypable H. influenzae strain.81 Handwashing decreases the frequency of acute respiratory tract diseases in childcare.82,83 A cluster, randomized, controlled trial of an infection control intervention including training of childcare staff regarding handwashing, transmission modes of infection, and aseptic techniques related to nose-wiping demonstrated a significant reduction in respiratory tract illnesses among enrollees less than 24 months of age over 311 child-years of surveillance.84 Because most infectious agents are communicable for a few days before and after clinical illness, exclusion from childcare of children with symptoms of upper respiratory tract infections will probably not decrease spread. Exclusion should occur when illness limits the child’s participation in activities or when the child’s needs exceed the capacity for provision of care.

CHAPTER

3

27

alcohol-based hand rubs, by both childcare enrollees and providers. Respiratory etiquette with disposal of tissues and cleansing of hands after contact with secretions should be observed. Frequently touched surfaces, toys, and commonly shared items should be cleaned at least daily and when visibly soiled. Children with signs and symptoms of respiratory tract illness should be cohorted, if possible, and excluded from group childcare if their symptoms prevent participation in activities or if their illness requires a level of care that exceeds a level that can be provided by the care provider. Ill childcare providers should be discouraged from providing care or having contact with children in group childcare. Vaccination of both child attendees and adult providers should be encouraged and both children and providers should receive frequent reminders regarding hand hygiene and respiratory etiquette to reduce influenza infections in group childcare settings.

Influenza Although influenza is responsible for disease among persons of all ages, rates of infection are highest among children less than 2 years of age and rates of complications of influenza infection are greatest among children of all ages with predisposing or underlying medical conditions. Influenza viruses are spread from person to person primarily through transmission of large respiratory tract droplets, either directly or by secondary contact with objects that are contaminated with infectious droplets. Children can shed virus for several days prior to onset of clinical symptoms and may be considered to be infectious for > 10 days following symptom onset. Transmission of infections may be increased by close contact among children who are not able to contain their secretions. Complications of influenza, including febrile seizures, encephalopathy, transverse myelitis, Reye syndrome, myositis, myocarditis, pericarditis, and death, can occur in children of preschool age. Among preschool-aged children with influenza infections, hospitalization rates range from 100 to 500/100,000 children, with highest hospitalization rates among children aged 0 to 1 year of age.85 Deaths from influenza uncommonly occur among both children with and without predisposing medical conditions. Reports of 153 laboratory-confirmed influenza-related pediatric deaths from 40 states during the 2003 to 2004 influenza season indicated that 61 (40%) were < 2 years of age and, of 92 children 2 to 17 years of age, 64 (70%) did not have an underlying medical condition traditionally considered to place a person at risk for influenza-related complications.86 Annual vaccination against influenza is the primary method for preventing influenza infection, and reducing transmission of infection among children in the childcare setting and among childcare providers. Influenza vaccine is recommended for all children 6 to 59 months of age, care providers of children 0 to 59 months of age in the childcare setting, and children and adolescents > 59 months of age with underlying medical conditions predisposing them to complications from influenza infection.85 A single-blind randomized controlled trial conducted during the 1996 to 1997 influenza season in 10 childcare centers in San Diego, California revealed that vaccinating children against influenza reduced influenza-related illness among their household contacts.87 In addition to preventing respiratory tract illness, several studies have shown the effectiveness of influenza vaccine in preventing otitis media among children in childcare.83,84,87a In one study of children 6 to 30 months of age in childcare centers, OR for acute otitis media was 0.69 and the 95% CI was 0.49 to 0.98 for those who received influenza immunization.88 Routine use of intranasal influenza vaccine among healthy children may be cost-effective and may be maximized by using group-based vaccination approaches. A prospective 2-year efficacy trial of intranasal influenza vaccine in healthy children 15 to 71 months of age demonstrated clinical efficacy as well as economic efficacy associated with focusing vaccination efforts on children in group settings.89 Vaccinating children has been associated with protection of older persons as well.89,90 Effective secondary prevention of transmission of influenza can be achieved with frequent hand hygiene using either soap and water or

Invasive Bacterial Infection Studies conducted before routine use of Haemophilus influenzae type b (Hib) vaccine in the United States have shown that the risk of developing primary invasive infection due to H. influenzae type b was higher among children attending childcare centers than in children cared for at home, independent of other possible risk factors.91,92 Risk of subsequent or secondary H. influenzae type b disease in the childcare setting was less convincing.91 Incorporation of conjugated Hib vaccines into the routine immunization schedule of children in the United States has dramatically reduced the frequency of invasive disease due to H. influenzae type b. Risk of disease due to Neisseria meningitidis may be increased in children in group childcare. Using space–time cluster analysis of invasive infections during 9 years of surveillance, from 1993 to 2001, in the Netherlands, researchers noted that clustering beyond chance occurred at a rate of 3% (95% CI 2% to 4%), and concluded that this rate was likely the result of direct transmission. Childcare center attendance was reported as the likely exposure for 8/40 (20%) of clusters, accounting for 13/82 (16%) cases of invasive disease with multiple serosubtypes.93 Childcare attendees who develop clinical disease while enrolled in group care prompt heightened community awareness and often result in distribution of prophylaxis to family and childcare contacts.94 The risks of developing primary invasive disease due to Streptococcus pneumoniae, of nasopharyngeal carriage of S. pneumoniae, and carriage of antibiotic-resistant strains are increased for children in childcare centers95–102 and childcare homes.97 In Finland, an increased risk of invasive pneumococcal disease in children younger than 2 years of age was associated with childcare attendance (OR, 36; 95% CI, 5.7 to 233), family childcare (OR, 4.4; 95% CI, 1.7 to 112), and history of frequent otitis media (OR, 8.8; 95% CI, 2.5 to 31).97 Resistance significantly decreased with a reduction in antibiotic use. Acute otitis media is the most common manifestation of pneumococcal infection and the source of most antibiotic prescriptions for children.103 Secondary spread of S. pneumoniae in the childcare setting has been reported, but the exact risks are not known.102,104–106 Colonization with S. pneumoniae in a childcare center was found in 32 of 54 (59%) children 2 to 24 months of age; 75% of the strains were penicillin-nonsusceptible.98 In an evaluation of the childcare cohort of an 11-month enrollee with multidrug-resistant S. pneumoniae in southwest Georgia, S. pneumoniae was isolated from 19 (90%) of the 21 nasopharyngeal cultures; 10 (53%) were serotype 14 and matched the susceptibility pattern of the strain from the index child; 4 of the 10 children with index-strain carriage had shared a childcare room with the index child, suggesting person-to-person transmission.107 Incorporation of a conjugated pneumococcal vaccine into the routine childhood immunization schedule of children in the United States in August of 2000 has resulted in a dramatic reduction in the frequency of invasive disease.108 The impact of vaccination on acute otitis media and reduction of penicillin-nonsusceptible pneumococcal infection is less dramatic and more variable, with evidence of

28

SECTION

A Epidemiology & Control of Infectious Diseases

increasing nasopharyngeal colonization and respiratory tract infections caused by nonvaccine serotypes and nontypable strains of pneumococcus.108–111

Methicillin-Resistant Staphylococcus aureus Infections due to methicillin-resistant Staphylococcus aureus (MRSA) were reported infrequently in the group childcare setting before 2000.112–114 However, with emergence of a community-acquired MRSA, the incidence of infection and its predisposition for affecting individuals in crowded conditions, where sharing of fomites exists, where skin-to-skin contact occurs, and hygiene is compromised place children in group care at risk.

TABLE 3-3. Acquisition of Cytomegalovirus by Childcare Center Providers and Others

Study

Number of Seronegative Persons

Annual Rate of Seroconversion (%)

Childcare providers (Alabama)122

202

11

Hospital employees (Alabama)122

229

2

Childcare providers (Virginia)118

82

20

Hospital employees (Virginia)118

300

2

82

8

68

13

Childcare providers (Iowa)

123

Childcare providers (Toronto)121

Echovirus During an outbreak of echovirus 30 infection in children and care providers in a childcare center, and in exposed parents, infection occurred in 75% of children and 60% of adults, but aseptic meningitis was more frequent in infected adults (12 in 65, 18%) than in children (2 in 79, 3%).115 A retrospective cohort study of childcare center attendees, employees, and household contacts in Germany revealed that 42% of childcare attendees, 13% of their household contacts, 5% of childcare center employees, and 2% of their household contacts were ill over a 31-day period. Thirteen percent (12/92) of childcare attendees had meningitis. This outbreak likely began among children enrolled in group childcare centers, with secondary cases occurring among their household contacts.116

Cytomegalovirus Young childcare attendees shed cytomegalovirus (CMV) chronically after acquisition and often transmit virus to other children and adults with whom they have close daily contact.117–119 Transmission is thought to occur through direct person-to-person contact and from contaminated toys, hands of childcare providers, or classroom surfaces.120 Prevalence studies have shown that 10% to 70% of children younger than 3 years of age (peak, 13 to 24 months) in childcare settings have CMV detected in urine or saliva.117,119,121 CMV-infected children can transmit the virus to women, with rates from 8% to 20% for their childcare providers and 20% of their mothers per year (Table 3-3)118,121–123 compared with rates of 1% to 3% per year in women whose toddlers are not infected.

Bloodborne Viral Pathogens Concern has arisen about the potential for spread of bloodborne organisms in the childcare setting: hepatitis B virus (HBV), hepatitis C virus (HCV), and human immunodeficiency virus (HIV).124–127 The highest concentrations of HBV in infected persons are found in blood and blood-derived body fluids. The most common and efficient routes of transmission are percutaneous blood exposure, sexual exposure, and, perinatally, from mothers to offspring at the time of delivery (see Chapter 109, Epidemiology and Prevention of HIV Infection in Children and Adolescents; Chapter 111, Diagnosis and Clinical Manifestations of HIV Infection). Other recognized but less efficient modes of transmission include bites and mucous membrane exposure to blood or other body fluids.128,129 Two case reports and a larger study have demonstrated possible transmission of HBV among children in the childcare setting.129–131 Other investigators have failed to demonstrate transmission in childcare, despite long-term exposure to children positive for hepatitis B surface antigen (HBsAg).132 Because of the small number of studies, the risk of HBV transmission in childcare cannot be quantified precisely. If a known HBsAg carrier bites and breaks the skin of an unimmunized child, hepatitis B

immune globulin and the HBV vaccine series should be administered.125 With implementation of universal immunization of infants with HBV vaccine beginning in 1991, horizontal HBV transmission in the childcare setting has been reduced to negligible. As the seroprevalence of HCV infection in children under 12 years of age is estimated to be 0.2% and most acute infections are asymptomatic, the transmission risk of HCV infection in childcare settings is unknown. The general risk of HCV infection from percutaneous exposure to infected blood is estimated to be 10 times greater than HIV but less than HBV. Areas of concern regarding attendance of an HIV-infected child in group care include the child’s potential risk of transmitting HIV and of acquisition of infectious agents.124 No cases of HIV infection are known to have resulted from transmission of the virus in out-of-home childcare. Children with HIV infection in the childcare setting should be monitored for exposure to infectious diseases, and their health and immune status should be evaluated frequently. The risk of transmission of HIV by percutaneous body fluid exposure, such as biting, is low. Complete evaluation of the source and extent of exposure should be undertaken to assess the risks and benefits of postexposure prophylaxis.125 An infectious disease physician with expertise in HIV care can be contacted for guidance. Precautions for the prevention of HBV, HCV, and HIV infection should be directed toward preventing transfer of blood or exudate fluids from person to person. Childcare providers should be educated about modes of transmission of bloodborne diseases and their prevention, and each center should have written policies for managing illnesses and common injuries such as bite wounds. Standard precautions for handling blood and blood-containing body fluids should be practiced in all childcare settings.125 Children infected with HIV or HCV or children who are HBsAg carriers should not be excluded from childcare. Decisions regarding attendance at childcare and the optimal type of childcare must be made by parents and the child’s physician after considering the possible risks and benefits.

Skin Infection and Infestation The magnitude of skin infections or infestations and the rates of occurrence in children in group childcare compared with rates in agematched children not in group childcare are not known. The most frequently recognized nonvaccine-preventable conditions are impetigo or cellulitis (due to Staphylococcus aureus or group A streptococcus), pediculosis, and scabies.73,83,133 Other conditions with skin manifestations that occur in children in childcare include herpes simplex virus (HSV) infection, varicella, ringworm, and molluscum contagiosum.73,134–137 Unimmunized children in childcare facilities are susceptible to varicella infection; most reported cases occur in children younger than 10 years of age.73,135 An outbreak of varicella was reported when a child with zoster attended a childcare center.134 Although the lesions

PART I Understanding, Controlling, and Preventing Infectious Diseases

Infections Associated with Group Childcare

were covered, the child continuously scratched and showed others the lesions, indicating the potential difficulties with this policy. Infection with group A streptococcus is a known complication of varicella.73 Although universal immunization with varicella vaccine has reduced cases of both varicella infection and its associated streptococcal complications among children in group childcare,52 several outbreaks of varicella have been reported among childcare attendees in the postlicensure era.74,138139 An outbreak of varicella in a childcare center in New Hampshire in 2000 with a vaccination coverage rate of 66% resulted in clinical disease in 25 of 88 (28%) children. This outbreak demonstrated poor protection against overall varicella disease. However vaccination was shown to be protective against moderate or severe clinical varicella.74 An active surveillance evaluation for vaccine effectiveness in Israel revealed 8 childcare-associated outbreaks in a 6-month time period from January to June, 2003 involving 116 children with clinical disease from 3 to 6 years of age. A vaccine coverage rate of 37% was noted among this cohort. In concordance with findings from vaccine postlicensure outbreaks in the United States, 94% of children with breakthrough varicella and 14% with natural varicella had mild disease.138 A varicella outbreak occurring among elementary school attendees in Maine in December to January 2003 was due to failure to vaccinate. The vaccination rates were notable for a decrease from 90% of kindergarten attendees to 60% of third-grade enrollees. Vaccine effectiveness in this cohort of 296 students was 89% against all varicella disease and 96% against moderate to severe disease. This outbreak illustrates the importance of vaccination of susceptible older children and adolescents to decrease the incidence of severe disease in unvaccinated children.139 As evidenced by these outbreaks, varicella incidence is highest in children 1 to 6 years of age. Therefore implementation of varicella vaccination requirements for childcare and elementary school attendees without evidence of immunity as recommended by the ACIP140 would reduce the susceptible population, consequently reducing the frequency of varicella outbreaks in group care settings. Primary HSV infection results in gingivostomatitis, most often in children 1 to 4 years of age.136,137 In two studies, clusters of primary infections occurred in children in childcare, most frequently manifesting as gingivostomatitis.136,137 In one study, restriction endonuclease analysis of DNA of isolated HSV revealed that a single strain of HSV-1 had been transmitted among children.137 Molluscum contagiosum is a benign, usually asymptomatic viral infection of the skin; humans are the only source. Virus is spread by direct contact or by fomites. Infectivity is low, but outbreaks have been reported. The frequency of occurrence in the childcare setting is unknown. People with atopic dermatitis or immunocompromised hosts, including people with acquired immunodeficiency syndrome (AIDS), have increased risk for acquiring infection or for having more extensive clinical manifestations. The incidence of pediculosis capitis (head lice) among children in childcare facilities in Seattle was 0.02/100 child-weeks141 and 0.03/child-year in San Diego.142 Treatment of infested children and their contacts with pediculicides that are used as directed may be considered as control measures.

Parvovirus B19 Parvovirus B19, the agent of erythema infectiosum (fifth disease), can cause arthropathy, transient aplastic crisis, persistent anemia in immunocompromised hosts, and nonimmune fetal hydrops (see Chapter 214, Human Parvoviruses). Serologic evidence of past infection has been reported to be 30% to 60% in adults, 15% to 60% in school-aged children, and 2% to 15% in preschool children.143 The virus is endemic among young children and has caused outbreaks of disease in the childcare setting.144,145 Parvovirus B19 spreads by the respiratory route or through contact with oropharyngeal secretions. In an outbreak during which more than 571 school and childcare personnel were tested serologically, the overall attack rate among susceptible individuals was 19%, with the highest rate (31%) occurring in childcare personnel.144 A cross-sectional study of 477

CHAPTER

3

29

childcare staff revealed a seroprevalence for parvovirus B19 IgG antibodies of 70%. Seropositivity was associated with age, and among staff less than 40 years of age, with length of group childcare contact.146 The greatest concern is that an infected pregnant woman could transmit the virus transplacentally, leading to fetal hydrops; neonatal illness and congenital malformations have not been linked to prenatal parvovirus B19 infection. Estimates of the risk of fetal loss when a pregnant woman of unknown antibody status is exposed are 2.5% for fetal death after household exposure and 1.5% after occupational exposure in a school.147

VACCINE-PREVENTABLE DISEASES In the United States, there are 16 diseases against which all children should be immunized, unless there are contraindications: (1) diphtheria; (2) tetanus; (3) pertussis; (4) Haemophilus influenzae type b; (5) measles; (6) mumps; (7) rubella; (8) poliomyelitis; (9) HBV; (10) varicella; (11) Streptococcus pneumoniae; (12) HAV; (13) influenza; (14) rotavirus; (15) human papillomavirus; and (16) meningococcal disease.52 Immunization of children and their care providers should be high priority (Table 3-4) and immunization is especially likely to benefit children in childcare settings.148 High levels of immunization exist among children in licensed childcare facilities,149 partially because laws requiring age-appropriate immunizations of children attending licensed childcare programs exist in almost all United States states, including vaccine mandates for childcare for HBV in 37 (74%) states as of April 2004, hepatitis A in 9 (18%) states as of August 2005, and varicella in 42 (84%) states as of August 2005.150 In a study of exemptions to immunizations, children of childcare age (3 to 5 years) with exemptions to immunizations were 66 times more likely to acquire measles and 17 times more likely to acquire pertussis than were age-matched immunized children.151

INFECTIONS ASSOCIATED WITH ANIMALS Human interactions with animals may be beneficial components of an educational or developmental curriculum in group childcare. However, animal exposure has been associated with sporadic zoonotic infections as well as outbreaks, injuries, and allergies, most notably in children less than 5 years of age. The increased prevalence of infections in this age group is likely due to compromised hand hygiene resulting in transmission of pathogens from animal to child. Animal interaction may occur in locations where childcare is provided, with a resident pet or visiting animal display, or in public venues where children visit, including petting zoos, aquariums, county fairs, parks, carnivals, circuses, or farms. Guidelines to reduce opportunities for transmission and infection have been developed to prevent disease transmission in many of these settings.152 Infections with enteric organisms pose the greatest risk for human disease from animals. A retrospective review of clinical and agricultural databases from 1966 through 2000 identified 11 published outbreaks of zoonotic disease associated with human–animal contact. A concomitant survey of state public health veterinarians revealed 16 additional outbreaks as well as a paucity of formal guidelines for the prevention of disease and injury from animal contact in the public setting.153 A subsequent review of the years 1991 to 2005 yielded reports of more than 55 outbreaks of infectious diseases among visitors to public animal exhibits.153a The predominant infection was enteric, resulting from direct or indirect fecal–oral contact. Inadequate hand hygiene, suboptimal supervision of children’s activities following animal contact, and hand-to-mouth activities following animal contact were risk factors for infection. Human–animal contact on public farms with inadequate hand hygiene was responsible for two Escherichia coli O157:H7 outbreaks in 2000. The median age of the 51 ill persons in one of these outbreaks that occurred in Pennsylvania was 4 years; 8 (16%) children developed hemolytic–uremic syndrome.

30

SECTION

A Epidemiology & Control of Infectious Diseases

TABLE 3-4. Vaccine-Preventable Infections Immunization Indicated Organism

Childcare Attendee

Childcare Provider

Diphtheria, pertussis, tetanus

As part of the 5-dose DTaP series

Tdap booster as adolescent/young adult; then Tdap every 10 years

Haemophilus influenzae type b (Hib)

As part of the 3–4-dose series, depending on vaccine used

Not indicated

Hepatitis A

2-dose series beginning at 1 year (12 to 23 months) of age

2-dose series recommended for adults at high risk for hepatitis A virus infection; not routinely recommended for childcare providers

Hepatitis B

3-dose series beginning at birth

3-dose series recommended for hepatitis B virus infection; not routinely recommended for childcare providers

Influenza A and B

Annual immunization for all children 6 to 59 months of age and high-risk children ≥ 59 months of age; 2 doses if first influenza immunization and ≤ 8 years of age

Annual immunization with trivalent inactivated or live attenuated influenza vaccine

Measles, mumps, rubella

2-dose series starting at 12 months of age

Booster immunization if only one dose received

Meningococcal disease

Polysaccharide vaccine for children 2 to 10 years of age in high-risk groups. Conjugate vaccine if 11 to 19 years of age

Conjugate vaccine recommended for adults 20 to 55 years of age at increased risk (polysaccharide is an acceptable alternative)

Pneumococcal disease

4 doses of heptavalent conjugate vaccine for all children 2 to 23 months of age; 1 dose of conjugate vaccine for certain children 24 to 59 months of age. Polysaccharide vaccine in addition to conjugate vaccine for certain high-risk groups 2 to 18 years of age

Pneumococcal polysaccharide vaccine for high-risk groups

Poliomyelitis

4-dose series

Most adults are immune; inactivated poliovirus vaccine may be indicated in select populations

Rotavirus

3-dose series beginning at 2 months of age and completed by 32 weeks of age

Not indicated

Varicella

2 doses, one at 12 to 18 months of age and the second at 4 to 6 years of age

2 doses for susceptible persons ≥13 years of age

Human papillomavirus

3 doses for females 9 years through 26 years of age

3 doses for females through 26 years of age

Identical E. coli O157:H7 isolates were noted in case patients, farm animals, and the farm environment,154 suggesting transmission of organisms to children resulting in clinical disease. During 2004 to 2005, three outbreaks of E. coli O157:H7 infections occurred among petting zoo visitors in North Carolina, Florida, and Arizona. A total of 173 cases, including 22 cases of hemolytic–uremic syndrome, were reported from the three states; children who visited petting zoos were predominantly affected. Both direct and indirect animal contact, including exposure in a play area contaminated with petting zoo drainage, was associated with infections. Restriction of entry into open-interaction areas of petting zoos by young children was proposed to reduce disease transmission and prevent additional outbreaks.155 Salmonella enteritica serotype Typhimurium, Cryptosporidium parvum, Campylobacter jejuni, Shiga toxin-producing E. coli (STEC), and Giardia have also been associated with infections with direct and indirect contact with zoo exhibits, farm day camps, and petting zoos.152 In addition to enteric infections, animal exposure can result in transmission of ecto- and endoparasites, Mycobacterium tuberculosis in certain settings, and local or systemic infections as a consequence of bites, scratches, stings, and other injuries. Food products produced by farm animals as demonstrations should not be consumed by children unless the food has undergone appropriate pasteurization and sterilization. Contact with animals within the childcare environment should occur where controls are established to reduce the risk of injuries and disease. Guidelines to reduce disease associated with animal contact outside and within the childcare facility have been developed and include education of staff, operators, and visitors; specialized design of exhibits where humans and animals will interface; guide to hand

hygiene instructions, agents, and stations; cleansing of facilities; and use of facilities for nonanimal events. Specific recommendations for group childcare settings include close supervision of children during animal contact, strict hand hygiene after direct animal contact or contact with animal products or environment, designation of areas for animal contact that are separate from areas in which food or drink are consumed, disinfection and cleaning of all animal areas with supervision of children over 5 years of age who may be participating in this task. Animals that visit or live in childcare facilities should be certified by veterinarians and should receive routine preventive health maintenance, including appropriate rabies immunization. Human– animal contact, especially for children less than 5 years of age, should always be supervised. Amphibians, reptiles, and weasels (ferrets and mink) should be housed in a cage and not handled by children. Wild or exotic animals, nonhuman primates, mammals with a high risk of transmitting rabies, wolf–dog hybirds, aggressive wild or domestic animals, stray animals, venomous or toxin-producing spiders, and insects should not be permitted in the group childcare setting.152,156

ANTIBIOTIC USE AND RESISTANCE PATTERNS Several studies have demonstrated the more frequent use of antimicrobial agents in children in childcare centers.157 During an 8-week period of observation of 270 children, antimicrobial agents were used by 36% of children in childcare centers compared with 7% and 8% of children in childcare homes or in home care, respectively (P < 0.001). The mean duration of antibiotic therapy prescribed for children in childcare centers (20 days) differed significantly

PART I Understanding, Controlling, and Preventing Infectious Diseases

Infections Associated with Group Childcare

(P < 0.001) from children in childcare homes (4 days) and children in home care (5 days).108 The estimated annual rates of antibiotic treatment ranged from 2.4 to 3.6 times higher for children in group care when compared with children in home care.157 Multiple studies have documented an association of childcare center attendance and colonization or infection due to resistant bacteria, including outbreaks of illness due to resistant Streptococcus pneumoniae95,97,99–101,106 and Shigella sonnei,21,158 as well as colonization due to resistant H. influenzae, 159 E. coli,160–161 and MRSA.113,114

INFECTIOUS DISEASES IN ADULTS Parents of children who attend a childcare facility and persons who provide care to these children have increased risk of acquiring infections such as CMV,118,121–123,147,159 parvovirus B19,144,145 HAV,147,162 and diarrhea.33,42 Childcare providers experience annual rates of CMV seroconversion ranging between 8 and 20%, compared with hospital employees who experience annual rates of seroconversion of 2%.118,121–123 During community outbreaks of erythema infectiosum, childcare providers were found to be among the most affected occupational groups, with seroconversion rates ranging from 9% to 31%.144,145 In a prevalence study of hepatitis A antibodies among childcare providers employed in 37 randomly selected childcare centers in Israel during 1997, 90% (402 of 446) of the childcare providers had antibodies to hepatitis A; the authors postulated a twofold risk of acquiring hepatitis A among providers.162 During outbreaks of diarrhea in childcare centers, 40% of care providers developed diarrhea.42 During a multicommunity outbreak of shigellosis, the overall median attack rate among employed staff of childcare centers was 6%, with a range of 0% to 17%.33 In outbreaks of group A streptococcal infection and echovirus 30 infection115 in childcare centers, adult providers and parents were affected. Childcare providers compared with nonproviders have a significantly higher annual risk of at least one infectious disease and lose more work days due to infectious diseases.147,163 Childcare providers should have all immunizations routinely recommended for adults, as shown on the adult immunization schedule (see Table 3-4) (www.cdc.gov/nip).

ECONOMIC IMPACT OF GROUP CHILDCARE ILLNESS The economic burden of illness associated with group childcare was estimated at $1.5 billion annually adjusted to 2005 United States dollars.164 Precise mechanisms for estimating illness burden and for evaluating effectiveness of infection control interventions are rare due to multiple challenges associated with performing such assessments.165 Attributing an outbreak to group childcare is challenging, because although these settings may promote transmission of infection, childcare attendees and staff interact with household contacts external to the childcare arrangement, thus facilitating secondary spread. For example, an economic assessment of an Escherichia coli O157:H7 outbreak in 1994 in rural Edinburgh, Scotland involved 71 persons with a median age range of 5 years and 7 months. In all cases children had consumed milk with increased coliform counts from a local dairy in the 2 weeks prior to illness onset. Although there was not a specific group childcare association to this outbreak, children of group childcare age were affected disproportionately. Investigating and containing the outbreak cost the community the equivalent of $296,660.166 In the United States, hepatitis A infections in children less than 18 years of age were estimated to range between $433 and $1492 per case. Between 11% and 16% of hepatitis A infections have been linked to the group childcare setting, although this estimate did not require a strict epidemiologic link of a case patient to the group childcare setting.167 However, since many hepatitis A infections in young children are asymptomatic, estimates of illness burden are primarily extrapolated from the fewer children who experience more significant complications from hepatitis A infections. In addition to

CHAPTER

3

31

economic analyses of vaccine-preventable infections, an economic analysis of a childcare-associated outbreak of Shigella sonnei in southwestern Ohio in 2001 incurred an overall cost of $821,725 to contain the outbreak of over 1600 infections, which was the equivalent of $514 per culture-confirmed case.168 A prospective evaluation of 208 families with at least one childcare enrollee, conducted from November 2000 to May 2001 in the Boston area, documented 2072 viral illnesses over 105,352 person-days. Among the 834 subjects, 1683 upper respiratory infections (URI) and 389 gastrointestinal (GI) illnesses were reported during the study period, with a total mean cost of $49 per URI and $56 per GI episode. Decreased parental productivity during missed days of work to care for a child who was not in childcare accounted for a significant proportion of the nonmedical costs.164 Future investigations of outbreaks of illness associated with group childcare could utilize recently developed computerized models and paradigms to assess the economic impact of outbreaks. In an era of limited funding, an understanding of expenses and allocation of resources will be important information to justify utility of interventions.

PREVENTION Specific standards should be established for personal hygiene, especially hand hygiene, maintenance of current immunization records of children and providers, exclusion policies, targeting frequently contaminated areas for environmental cleaning, and appropriate handling of food and medication. In studies in which improved infection control measures were implemented and monitored, both upper respiratory tract illness and diarrhea were reduced in intervention centers.82,169 In addition, in children at intervention centers, 24% fewer antibiotic prescriptions were given and fewer absences from work on the part of parents occurred.83 Educational sessions on health topics by healthcare professionals was found to be the most efficacious means of promoting health education in simultaneous surveys of licensed childcare center directors, parents, and health providers in Boston, Massachusetts.170 A cross-sectional survey conducted in 2000 of childcare providers, parents, and pediatricians in Baltimore, Maryland revealed deficits of knowledge among all groups. Compared with national guidelines on exclusion for 12 symptoms, childcare providers and parents were overexclusive and pediatricians were underexclusive. More childcare providers and parents than pediatricians felt that exclusion would reduce transmission of disease.171 As asymptomatic excretion and potential for transmission precede the onset of clinical symptoms in many childcare-associated infectious diseases, strategies that involve prevention would likely be most efficacious in reducing incidence. In addition to traditional handwashing, the use of alcohol-based hand-sanitizing hand gels in healthcare and other settings is an efficacious means of achieving hand hygiene.171 In support of this preventive strategy, a cluster randomized, controlled trial was conducted in the homes of 292 families with children who were enrolled in out-of-home childcare centers. A multifactorial intervention emphasizing alcohol-based hand sanitizer use in the home reduced transmission of GI illnesses within families. The effect on reduction of respiratory tract illness transmission in this evaluation was less pronounced and may relate to the use of handsanitizing gel following toileting activities but not following sneezing, coughing, or blowing/wiping of nasal secretions.172 Molecular techniques, including DNA probes, could be used as surrogate markers to study transmission of enteric pathogens in childcare centers and from centers to children’s homes.173 Further evaluations of molecular techniques during outbreak investigations, hand hygiene strategies, and educational interventions could assist with allocation of resources to the most effective prevention regimens. Written policies should be available, followed, and reviewed regularly for the following areas: managing child and employee illness, including exclusion policies; maintaining health, including

32

SECTION

A Epidemiology & Control of Infectious Diseases

immunizations; diaper-changing procedures; hand hygiene; personal hygiene policies for staff and children; environmental sanitation policies and procedures; handling, serving, and preparation of food; dissemination of information about illness; and handling of animals. Local health authorities should be notified about cases of communicable diseases involving children or care providers in the childcare setting. The AAP and the American Public Health Association jointly published the National Health and Safety Performance Standards: Guidelines for Out-of-Home Childcare Programs.174 This comprehensive manual provides guidelines regarding infectious diseases and other health-related matters pertinent to out-of-home childcare.

CHAPTER

4

Infectious Diseases in Refugee and Internationally Adopted Children Mary Allen Staat Each year thousands of immigrant children come to the United States to begin a new life. In this chapter, the infectious disease issues of two groups of immigrants will be discussed: refugees and internationally adopted children. Refugees are noncitizen immigrants who are unable or unwilling to return to their country of origin because of persecution or a fear of persecution.1 In addition to refugees, there are other categories of noncitizen people in the United States. These noncitizens may be immigrants or nonimmigrants. Immigrants include refugees, licensed permanent residents, asylees, and parolees. Nonimmigrants include people who are undocumented, students, tourists, or visitors on business. Internationally adopted children are immigrants that are classified as orphans. Most of these children however are not truly orphaned, but instead have been abandoned by or separated from both parents. In most cases these children will be given United States citizenship as they arrive in the United States with their new families. In 2005, there were a total of 53 813 refugees who arrived in the United States.1 Of these, 38% were < 18 years of age. Nearly 80% of these refugees came from just eight countries: Somalia (19%), Laos (16%), Cuba (12%), Russia (11%), Liberia (8%), the Ukraine (5%), the Sudan (4%), and Vietnam (4%). Refugees come to the United States from all around the world, with the exception of northern Europe, Australia, New Zealand, and Canada.2 Over the past decade, international adoption has become an increasingly popular way to build families. More than 175,000 children have been internationally adopted in the United States since 1996.3 In 2005 alone, 22,728 children were adopted coming from more than 20 countries, with 83% of these children coming from just five countries: China (35%), Russia (20%), Guatemala (17%), Korea (7%) and the Ukraine (4%).3 Although there has been little variation in the countries of origin over the past 10 years, in 1990, there were very few children arriving from the current top three countries of China, Russia, and Guatemala.3 The majority of children are adopted as infants and toddlers. With the exception of Korea and Guatemala, where most children are in foster care, children from the other countries have generally resided in orphanages prior to coming to the United States. Because refugees and internationally adopted children come from resource-poor countries, physicians and other healthcare providers should be cognizant of the global prevalence of other infectious diseases that are seen less commonly in native-born North Americans. Both groups are at increased risk for common infectious diseases such as tuberculosis, intestinal parasites, dermatologic infections, and infestations. Hepatitis B, hepatitis C, human immunodeficiency virus

(HIV), and syphilis, although seen in the United States, are far more prevalent in the countries of immigrants where there are few resources for screening and prevention. Although there are a number of similarities in refugees and internationally adopted children, there are also important differences. Refugee children and internationally adopted children differ in terms of the general medical screening they receive before arrival in the United States. Most refugees are subjected to organized screening evaluations before emigration visas are issued.4 For children > 15 years of age, predeparture screening includes serologic testing for HIV and syphilis and a chest radiograph to assess for evidence of tuberculosis. A physical examination is performed on children of all ages.4 In contrast, no organized screening procedure is required for internationally adopted children. Second, because medical screening for refugees usually is sponsored by responsible medical organizations, results of such testing are typically accurate. In internationally adopted children, medical testing is often incomplete or done shortly after birth. Generally, HIV and syphilis serology and hepatitis B surface antigen test results are provided with the referral information. Although the reliability of this testing has been a concern in the past, in recent years testing done in the child’s country of origin has proven to be accurate when repeated in the United States. Third, preventive measures such as immunizations, vitamin supplementation, and dental care are undertaken in most refugee children while still in the camps; in adoptees, there are inconsistencies in the receipt of these measures. Last, differences in the types of infectious diseases may also distinguish refugees from adopted children. Although both populations are susceptible to a variety of infectious agents, because the countries of origin and the living conditions differ, refugee children are more likely to have been exposed to infections such as typhoid fever, malaria, filariasis, flukes, or schistosomiasis, which occur uncommonly in internationally adopted children. As these immigrants join our community, healthcare professionals will inevitably have the opportunity to provide care for these children and should therefore be knowledgeable of the infectious disease issues they may encounter and the need for screening for infectious diseases in these populations.

GUIDELINES FOR EVALUATION Because of the predominance of infectious diseases in developing nations, recommendations for screening tests are weighted toward infectious disease processes, but aspects of general health, including vision, hearing, dental, and developmental examinations, also should be included.5–7 Despite the healthy appearance of many immigrant children, children should be evaluated by a healthcare professional within 2 weeks after arrival to assure that they are screened properly and receive preventive healthcare services. Table 4-1 outlines the recommended infectious disease screening for refugees and internationally adopted children.

Hepatitis A Virtually all inhabitants of resource-poor countries have contracted hepatitis A by early adulthood and have immunoglobulin (Ig) G antibodies. In the past, screening for hepatitis A was recommended only in children with chronic hepatitis B infection to determine the need for hepatitis A immunization. Healthcare providers in some states where hepatitis A vaccine was recommended for routine use also may have been screened for hepatitis A. Now that hepatitis A vaccine is recommended for all children ≥ 12 months of age,8 IgG antibody testing in children > 1 year of age may be useful for all immigrants to determine who should be immunized. Age- and country-specific prevalence data are needed for both internationally adopted and refugee children to develop cost-effective screening strategies. Testing will likely be most cost-effective in older children. Screening for IgM antibodies is only useful in the diagnosis of acute infection and therefore is not used

PART I Understanding, Controlling, and Preventing Infectious Diseases

Infectious Diseases in Refugee and Internationally Adopted Children TABLE 4-1. Recommended Tests for Refugee and Internationally Adopted Children Test

Refugee Children

Internationally Adopted Children

Tuberculin skin test (TST)

All

Alla

HEPATITIS B VIRUS SEROLOGIC TESTING

All

Alla

Hepatitis C virus serologic testing

Some

Somea

Human immunodefiency virus 1 and 2 serologic testing

> 15 years of age

Alla

Hepatitis B surface antigen (HBsAg) Hepatitis B surface antibody (anti-HBs) Hepatitis B core antibody (anti-HBc)

SYPHILIS SEROLOGIC TESTING

Nontreponemal test (RPR, VDRL, or ART) Treponemal test (TPPA, MHA-TP, or FTA-ABS)

> 15 years All of age Not All recommended

STOOL EXAMINATION

Microscopic evaluation for ova and parasites (3 specimens) Giardia lamblia and Cryptosporidium antigen (1 specimen)

All

All

All

All

Complete blood count

All

All

Urinalysis

All

Not recommended

ART, automated reagin test; FTA-ABS, fluorescent treponemal antibody absorption; MHA-TP, microhemagglutination-Treponema pallidum; RPR, rapid plasma reagin; TPPA, Treponema pallidum particle agglutination; VDRL, Venereal Disease Research Laboratory. a Consider reassessing 6 months after arrival.

routinely for screening. Infection with hepatitis A is anicteric in up to 90% of children < 6 years of age, 50% to 60% in older children and 20% to 30% in adults.8 Symptoms such as vomiting and loss of appetite, when accompanied by fever and an enlarged, tender liver, should prompt evaluation for viral hepatitis in recently arrived immigrants or adoptees, even in the absence of jaundice.8

Hepatitis B In contrast to hepatitis A testing, routine screening for hepatitis B virus infection is recommended for all refugees and internationally adopted children.5–7 Early identification of hepatitis B infection is important so that appropriate management can be initiated and household contacts and caregivers can be vaccinated. The prevalence of hepatitis B infection is 2% to 7% in Asia, Africa, and Central and South America, in contrast to 0.5% in the United States.9 In many published studies, refugee and internationally adopted children from these countries mirror these rates of infection and typically acquire the virus by vertical transmission, although bloodborne infection and horizontal transmission also have been implicated.10–18 In international adoptees, rates were the highest in Romanian adoptees from the early 1990s, where 53% had evidence of past or present infection and 20% had active infection.15 Studies consistently have shown a prevalence of 3% to 5%.16–18 Hepatitis B screening tests, including the hepatitis B surface antigen (HBsAg) and antibodies to surface (anti-HBs) and core (antiHBc) antigens, are recommended in the medical evaluation of internationally adopted children and refugees.5–7 All three tests should be done to determine whether the child is immune due to immunization,

CHAPTER

4

33

recovered from infection, or has acute or chronic hepatitis B. Common patterns of hepatitis B serology profile are shown in Chapter 213, Hepatitis B and Delta Virus. In a child who is found to be HBsAgpositive, repeat testing should be done 6 months later; if HBsAg persists for > 6 months, the child has chronic hepatitis B infection. If the child no longer has HBsAg and has developed anti-HBs, then the child had acute infection and has recovered and is no longer able to transmit the virus to others. Children with acute or chronic hepatitis B (positive HBsAg) can transmit hepatitis B to others. Children with acute or chronic hepatitis B should have additional testing done, including testing for HBeAg, HBeAb, and serum hepatic enzyme levels. HBsAg-positive children should be followed carefully by a hepatologist for management. In children with hepatitis B infection who are from countries in which delta hepatitis coexists with hepatitis B, such as in southern Italy, parts of Eastern Europe, South America, Africa, and the Middle East, testing for antibodies to the delta hepatitis virus could be done, but generally is not recommended because a positive test would not change clinical management.5 In children who initially test negative, some experts recommend repeat testing for hepatitis B approximately 6 months after arrival to insure the child was not infected just prior to arrival to the United States.5 Vaccination of household contacts of children with acute or chronic hepatitis B (HBsAg-positive) must occur promptly. Epidemiologic studies have demonstrated that up to 20% of unvaccinated household contacts become HBsAg-positive within > 5 years or more of exposure within the home19,20 and transmission of hepatitis B from newly adopted children to their parents has been documented.19–21

Hepatitis C Until more is known regarding the incidence and prevalence of hepatitis C in refugee and internationally adopted children, screening for this virus is not recommended routinely. However, in both populations, if there is a history of specific risk factors such as current or former injection drug use, receipt of blood products, or residence in settings with a documented high prevalence of hepatitis C, then testing should be done.22 These risk factors are difficult to ascertain in either group of children. In one study, the prevalence of hepatitis C antibody in internationally adopted children was shown to be < 1%.18 Until additional data are available, it is recommended that children from China, Russia, Eastern Europe, and Southeast Asia should be screened for hepatitis C infection.5 In addition, children from other countries should be screened if there is a history of receipt of blood products, maternal drug use, or a high prevalence of infection in the child’s birth country.5 Antibody testing is done initially and if the antibody is positive, polymerase chain reaction (PCR) is performed to confirm infection. Children found to have hepatitis C infection should be immunized for hepatitis A and B and should be referred to a hepatologist for further management.

Human Immunodeficiency Virus-1 and Human Immunodeficiency Virus-2 Infection Refugees ≥ 15 years of age are tested routinely for HIV prior to coming to the United States.4 Children < 15 years of age are not tested unless they are in a high-risk situation. Otherwise, for children not tested abroad, routine testing is not recommended for refugees unless indicated clinically. In contrast, most internationally adopted children have been tested for antibodies to HIV-1 by enzyme immunoassay (EIA) in their birth country and routine screening by EIA is recommended for all internationally adopted children upon arrival to the United States.5 Reports of HIV infection in internationally adopted children are rare. In the early 1990s there were two case reports of Romanian adoptees with HIV infection.23 In recent case series, there have been no reports of HIV infection in adoptees.14,17,18 Positive or indeterminate results should be resolved with use of HIV DNA PCR to detect HIV-1 virus (see Chapter 111, Diagnosis and Clinical

34

SECTION

A Epidemiology & Control of Infectious Diseases

Manifestations of HIV-1 Infection). If there is a clinical suspicion for HIV infection, HIV DNA PCR testing should be considered if the antibody testing by EIA is negative. Retesting > 6 months after arrival to the United States should be considered. Preadoptive testing for HIV-2 infection is not performed routinely. HIV-2 infection is prevalent in some African nations and is now recognized on several other continents; perinatal transmission appears to be limited. Symptoms suggestive of HIV infection with negative EIA results for HIV-1 should prompt testing for HIV-2.

Tuberculosis Tuberculosis is highly prevalent in the countries of origin for both refugees and internationally adopted children and therefore screening upon arrival to the United States is recommended for both groups.5–7 In 2003, 53% of all tuberculosis cases were among foreign-born people in the United States.24 Even though the overall incidence of tuberculosis decreased in the United States from 1993 to 2003, there was no change in the incidence in foreign-born people. In a study examining trends in the epidemiology of tuberculosis disease in children in the United States, 22% of cases were foreign-born, with case rates of 12.2/100,000 in foreign-born compared with 1.1/100,000 in United States-born children.25 Refugees ≥ 15 years of age and children < 15 years of age with concern about tuberculosis exposure or disease are screened with a chest radiograph prior to arriving in the United States. If the chest radiograph is abnormal, microscopic evaluation of sputum or gastric aspirate for acid-fast bacilli is performed.4 Sputum-positive people are banned from entry until sputum smears are negative. Several studies in the past have found a high proportion of immigrants and refugees with latent tuberculosis infection (LTBI) in Minnesota (49%),26 San Francisco (40%),27 Buffalo (20%),28 and Maine (35%).29 A study in San Diego county found that 7% of immigrants screened had active pulmonary tuberculosis and 76% had LTBI.30 In adoptees, screening in the country of origin is inconsistently done and generally is documented only in older children. After arrival, the proportion of internationally adopted children with LTBI appears to be lower than that seen in refugees. This is likely due to the younger age and lower risk of living conditions of internationally adopted children. LTBI is still prevalent in adoptees. In the most recent study in which all tuberculin skin tests were evaluated by a healthcare professional, 19% of international adoptees had LTBI, with the highest proportion in Russian and Eastern European adoptees (29%) and the lowest in Chinese adoptees (10%).18 Tuberculosis disease in international adoptees is reported rarely.31,32 In the first series of Korean adoptees, although the overall prevalence was < 1%, the two children with tuberculosis disease had nonreactive skin tests. One child had pneumonia due to Mycobacterium tuberculosis and recovered.31 The other child died from tuberculosis meningitis.31 These cases illustrate the importance of considering tuberculosis in the differential diagnosis even with negative tuberculin skin tests. In the second case report,32 an adoptee from the Marshall Islands with tuberculosis infected his female guardian and 20% of the contacts identified. Upon arrival in the United States, he had a TST done which was never read. It wasn’t until his guardian developed tuberculosis that he was diagnosed with cavitary tuberculosis. For both refugees and internationally adopted children, screening for tuberculosis should be done at the initial assessment by placement of a tuberculin skin test (5 TU purified protein derivative).5–7,24 The test should be read 48 to 72 hours later by a healthcare professional. A reading of ≥ 10 mm of induration is considered positive for both refugees and internationally adopted children. A reading of 5 mm or more is considered positive if there is a known contact with a person with active tuberculosis, an abnormal chest radiograph finding, signs or symptoms suggestive of tuberculosis, or evidence of immunosuppression. Children with positive TST should have a thorough physical examination and chest radiograph performed to assess for M. tuberculosis disease. If the chest radiograph and physical examin-

ation are normal and the skin test is ≥ 10 mm, the diagnosis of LTBI is made and treatment with a 9-month course of isoniazid is begun. Similarly, a child with a 5 mm to 9 mm reaction who has known exposure to someone with active tuberculosis or who is receiving immunosuppressive therapy or has an immunosuppressive condition, including HIV, with no evidence of disease, should receive isoniazid therapy. Retesting of skin test-negative children ≥ 6 months after arrival should be considered, especially if there is concern that the child was malnourished or may have been exposed just prior to coming to the United States. While most refugees and internationally adopted children with M. tuberculosis infection have LTBI, disease should be considered in children with pneumonia or nonspecific symptoms such as fever, malaise, growth delay, weight loss, cough, night sweats, and chills. Pulmonary tuberculosis is the most common site of infection in children, accounting for 77% of cases, followed by lymphatic tuberculosis in 16% of children.25 Chest radiographs may reveal adenopathy, segmental or lobar infiltrates, or pleural effusion. Cavitary lesions and miliary disease are less common. Extrapulmonary manifestations of the central nervous system, middle ear and mastoids, lymph nodes, bone, joints, and skin can be seen. Recovery of the responsible organism is paramount, because many children arrive from countries in which drug-resistant M. tuberculosis is common. Gastric aspirates or bronchoscopy, or both, can be useful adjuncts in a child too young or too ill to expectorate sputum. Initial management of tuberculous disease in refugees or internationally adopted children should include treatment with isoniazid, rifampin, pyrazinamide, and at least one other agent to ensure bactericidal coverage while culture results and susceptibility testing are pending (see Chapters 134, Mycobacterium tuberculosis and 292, Antimicrobial Agents). It is important to consider the most recent resistant patterns from the child’s country of origin when making treatment decisions. Vaccination with Calmette-Guérin bacillus (BCG) is common in most resource-poor countries, but is not a contraindication to the placement of a tuberculin skin test.5 Previous studies have demonstrated that placement of a tuberculin skin test after BCG vaccination fails to elicit induration of 10 mm or greater, and thus history of BCG does not affect the interpretation of purified protein derivative.33,34 BCG vaccination can be recognized by a 2- to 4-mm scarification, typically on the left deltoid. Occasional complications include enlargement of the regional lymph nodes or nodularity or ulceration at the vaccination site. The granuloma typically is located in the deltoid region and may or may not be draining. Culture of the drainage yields Mycobacterium bovis. There is no consensus regarding the management of this condition. Some lesions resolve without treatment, whereas other lesions require excision or treatment with isoniazid, erythromycin, or clarithromycin, or can be associated with regional or distant sites of infection.35,36

Intestinal Pathogens and Enteric Infections Intestinal parasites are common in both refugee and internationally adopted children but rates of infection vary by age and country of origin. Most children will not have had testing prior to coming to the United States. A wide variety of parasites, both pathogenic and nonpathogenic (Table 4-2), can be seen in both groups of immigrants. However, refugees are more likely to have nematodes and trematodes compared with internationally adopted children. In internationally adopted children, the prevalence of intestinal parasites varies from 9% to 51%, depending on the age of the child and country of origin, with an increased risk in older children and children originating from countries other than Korea.14–18 Giardia lamblia is the most commonly identified pathogen in all studies, with as many as 19% of children infected.18 Helminthic infections are identified less frequently and include Hymenolepsis species, Ascaris lumbricoides, and Trichuris trichiura.18 In contrast, in a study of refugees, 17% were found to have Giardia, and 19% had helminths, mainly hookworms.37

PART I Understanding, Controlling, and Preventing Infectious Diseases

Infectious Diseases in Refugee and Internationally Adopted Children

TABLE 4-2. Pathogenic and Nonpathogenic Intestinal Parasites Parasite Type

Pathogens

Nonpathogens

Protozoa

Giardia lamblia Entamoeba histolytica Dientamoeba fragilis Balantidium coli Blastocystis hominisa Isospora belli Cryptosporidium parvum Cyclospora cayentensis Microsporidium species

Endolimax nana Entamoeba coli Entamoeba gingivalis Entamoeba hartmanni Entamoeba polecki Iodamoeba butschlii Chilomastix mesnili Enteromonas hominis Retortamonas intestinalis Trichomonas hominis Trichomonas tenax

HELMINTHS

Nematodes (roundworms)

Ascaris lumbricoides (roundworm) Trichuris trichiura (whipworm) Strongyloides stercoralis (threadworm) Enterobius vermicularis (pinworm) Necator americanus (hookworm) Ancyclostoma duodenale (hookworm)

Cestodes (tapeworms)

Hymenolepsis species Taenia saginata (beef tapeworm) Taenia solium (pork tapeworm) Schistosoma species

a

Controversy exists regarding the pathogenicity of this organism.

Both refugees and adoptees should be screened for intestinal parasites upon arrival to the United States.5–7 The diagnosis of most intestinal parasites is done by examination of stool preserved in formalin and polyvinyl alcohol or sodium-acetate formalin for ova and parasite testing. Testing multiple stool specimens increases the sensitivity of microscopic examination for ova and parasites.38,39 In one study, the sensitivity was 76% using one specimen and 92% in two specimens and a third specimen identified parasites in an additional 8% of children.39 Examination by an experienced technologist is critical in identifying intestinal parasites. One stool specimen should also be evaluated for G. lamblia and Cryptosporidium parvum using antigen testing.5–7 Repeat stool samples are essential to assess the effectiveness of treatment for the identified parasite or to determine the presence of new (or newly found) pathogens. If the child remains persistently symptomatic, additional stool sample examinations are warranted and testing for other parasites such as Cyclospora and Strongyloides stercoralis should be considered. In adoptees, screening for parasitosis with eosinophil counts is not indicated because most infections are due to protozoa (e.g., G. lamblia) and nonmigrating nematodes (pinworms) and therefore do not elicit peripheral blood eosinophilia.7 However, in refugees, eosinophilia may be a useful screening tool since refugees are more likely to have migrating nematodes. In contrast to intestinal parasites, bacterial enteric pathogens appear to be less common in immigrants and adoptees; however, most studies have not assessed systematically for bacterial causes. Children with bloody diarrhea or diarrhea associated with high fever should have stool examined for bacterial enteropathogens (e.g., Salmonella, Shigella, Yersinia, Campylobacter) as part of the initial evaluation.5

CHAPTER

4

35

Syphilis Refugees ≥ 15 years of age routinely are screened for syphilis before resettlement.4 Although refugees are not screened routinely upon arrival to the United States, refugees of any age should be tested if there is clinical suspicion for syphilis or there is evidence or concern for sexual exposure. Most internationally adopted children will have documentation of syphilis screening in their birth country. Serologic status for syphilis is determined accurately in many foreign countries; as a result, most children with positive tests are identified before emigration. Syphilis rarely has been reported in case series of adoptees.14–18 Unfortunately, in those children with a history of syphilis, it is difficult to determine when the birth mother contracted syphilis and if she was treated and followed appropriately. In addition, in children with suspected syphilis exposure, details of the evaluation and prescribed treatment regimens are often incomplete or incorrect. Most frequently, documentation of treatment for congenital syphilis is insufficient, often only stating that penicillin has been given. For internationally adopted children, screening with both a nontreponemal test (rapid plasma reagin, Venereal Disease Research Laboratory, automated reagin test) and a treponemal test (microhemagglutination-Treponema pallidum, fluorescent treponemal antibody absorption, or Treponema pallidum particle agglutination) is recommended5 (see Chapter 182, Treponema pallidum (Syphilis)). A positive treponemal test result warrants a complete and thorough evaluation to document the extent of the disease and to insure adequate treatment.

Other Testing A complete blood count is recommended for refugees and internationally adopted children.5–7 The hemoglobin and red blood cell indices may identify children with anemia, iron deficiency, hemoglobinopathies, or malaria. Low white blood cell counts and lymphopenia may indicate HIV infection or malnutrition. Eosinophilia is common among refugees and suggests parasitic infection.7 For refugees, a urinalysis is recommended to assess for hematuria and pyuria. Hematuria may indicate schistosomiasis and pyuria may reflect a urinary tract infection or renal tuberculosis. In refugees, routine screening for malaria is not recommended, but malaria should always be considered in children with fever who come from endemic areas.7 Elevated blood lead levels have been reported in internationally adopted children. All international adoptees are recommended to have lead testing done as part of their initial evaluation.40 Exposure is felt to be from leaded gasoline engine exhaust, ceramic ware, lead-based paint, and industrial waste. There is no formal recommendation for refugees; however, it is likely that they could benefit from this screening as well.

Other Common Infections Dermatologic Infections Lice, scabies, and molluscum contagiosum are common in both refugees and internationally adopted children. The incidence of these conditions has not been well documented. The physical examination should include careful evaluation for these entities so that appropriate treatment is given (see Chapter 202, Poxviridae (Molluscum Contagiosum)).

Upper Respiratory Tract Infections Upper respiratory tract infections, including fever, coryza, and otitis media, occur in a large percentage of adopted children within the first month of arrival; one retrospective study placed their incidence at 40%.41 Since the past history of otitis media is often unknown and since many children have language delays, referral to an otolaryn-

36

SECTION

A Epidemiology & Control of Infectious Diseases

TABLE 4-3. Serologic Testing Available for Verifying Immunization Status Children ≥ 5 Months of Age

Children ≥ 12 Months of Age

Diphtheria IgG EIA

Diphtheria IgG EIA

Tetanus IgG EIA

Tetanus IgG EIA

Poliovirus serotypes 1–3 neutralizing antibody

Poliovirus serotypes 1–3 neutralizing antibody

Hepatitis B surface antibody

Hepatitis B surface antibody Rubeola (measles) antibody Mumps antibody Rubella antibody

Providing care for new immigrants can be rewarding. Application of directed screening tests upon arrival yields incalculable dividends for prevention and for the welfare of adoptees, refugees, and their families.

CHAPTER

5

Bioterrorism Julia A. McMillan

Varicella antibody Hepatitis A antibody EIA, enzyme immunoassay; IgG, immunoglobulin G.

gologist for hearing test evaluation should be considered for children with otitis media and language delays.

PREVENTIVE MEASURES Immunizations Making immunization recommendations can be a challenge for clinicians providing care for immigrant children. Most refugees arrive in the United States without immunization records whereas most international adoptees have some documentation of receipt of immunizations. The recommended approach for these two groups of immigrants differs. For refugees, if immunization documents exist, they should be reviewed and may be accepted as valid if the immunizations were provided at ages and intervals acceptable to United States standards.42 A disease history is only acceptable as proof of immunity for varicella. Refugees who have no records or have incomplete immunizations should receive vaccines at their first visit. Guidelines for routine and catch-up immunizations should be followed.5,43 There are few data on verifying immunization status by antibody testing in refugees. In one study in refugees, screening for varicella was more cost-effective than providing immediate vaccine for individuals ≥ 5 years of age.44 There have been several studies examining antibody testing to verify the immunization status of internationally adopted children.45–49 Unfortunately, results of studies differ; some studies show inadequate protection whereas others show good protection. Differences in study design and laboratory methods likely account for the different results. With a lack of consensus, there currently are two acceptable approaches recommended to insure that internationally adopted children are protected against vaccine-preventable diseases.5,43 The first approach is to reinitiate immunizations regardless of documentation; the second approach is to use serologic testing to determine which immunizations are needed. Because antibody testing for one vaccine may not be predictive for others, a combination of reimmunization and antibody testing could be utilized. Serologic testing is available for most vaccine antigens (Table 4-3). For children ≥ 5 months of age, testing for diphtheria and tetanus toxoid, polio, and hepatitis B antibodies can be done. Testing for measles, mumps, rubella, varicella, and hepatitis A should only be done in children ≥ 12 months of age because of the possible presence of maternal antibody. For varicella, testing can be done to verify whether the child had varicella disease since varicella immunization is not available in countries from which children are adopted.

This chapter provides information concerning agents that might be used in a biological attack by groups or individuals with two objectives: causing illness and death, and inciting widespread fear and the public chaos that ensues. The illness that results from biological agents introduced through terrorism does not differ significantly from the signs and symptoms of naturally acquired infection with the same agent. Those clinical aspects of infection are discussed elsewhere in this text. Rather, it is the epidemiology of disease due to bioterrorism that warrants special attention. And, although it is appropriate to consider epidemiologic characteristics of biological terrorism in a pediatric infectious diseases textbook, many of those characteristics are also typical of illness brought on by acts of chemical terrorism. Airborne release and contamination of water or food are the most likely methods for initial dispersal of biological or chemical agents. Medical providers, public health agencies, and first responders (fire fighters, police, and emergency medical services) must be prepared to suspect disease due to acts of terrorism, whatever the causal agent. Infectious diseases that can be spread person to person, e.g., smallpox or salmonellosis, enhance the likelihood of widespread disease.

BACKGROUND Infectious agents have been used as agents of terrorism for centuries. Plague-infested cadavers were used as weapons by the Tartars during the Middle Ages, and clothing contaminated by smallpox was used by Spaniards in their conquest of South American peoples. During the past century a variety of microbes were used as weapons, both as agents of war and by small groups whose goal was apparently to terrorize, rather than to kill. Clostridium botulinum and plague were employed by the Japanese and the Chinese, respectively, during World War II. In domestic attacks in the United States, followers of the Bhagwan Shree Rajneesh used Salmonella typhimurium to contaminate 10 restaurant salad bars in 1984, resulting in 751 cases of gastroenteritis, and Aum Shinrikyo followers in Tokyo are reported to have attempted bioterrorism using anthrax and botulinum toxin before their successful sarin gas attack on the Tokyo subway in 1995. During World War II the United States, as well as the United Kingdom, Canada, Japan, and the USSR had active bioweapons research programs involving many of the infectious agents that are now feared as possible weapons of terrorism, including Bacillus anthracis, Francisella tularensis, botulinum toxin, and staphylococcal enterotoxin. By 1969 international efforts at bioweapons disarmament had begun, and the United States had destroyed its offensive bioweapons arsenals by 1973. The Biological Weapons Convention (BWC), more formally known as The Convention on the Prohibition of the Development, Production, and Stockpiling of Bacteriological (Biological) and Toxin Weapons and on Their Destruction, became effective in 1975 and is the first multilateral disarmament treaty involving bioweapons. Though 150 nations have signed and ratified this treaty, there is no established mechanism for monitoring

PART I Understanding, Controlling, and Preventing Infectious Diseases

Bioterrorism

compliance. Noncompliance on the part of at least one signatory nation, the USSR, was detected during investigation of the accidental release of B. anthracis spores from a Soviet military facility in Sverdlovsk, Russia, in 1979, which resulted in the death of 69 people. Public concern for the risk of bioterrorism in the United States was raised following the September, 2001 attacks on the World Trade Center and the Pentagon. These attacks were closely followed by covert terrorism employing the United States Postal Service for distribution of anthrax, resulting in 22 proven or suspected infections and 5 deaths.

CHAPTER

5

37

difficult to disseminate and cause moderate morbidity and low mortality, but their identification requires enhanced diagnostic capacity and disease surveillance. Finally, emerging agents that could be engineered for mass dissemination and may be attractive as weapons because of their associated high morbidity, mortality, and potential public health impact are listed in category C. These include Nipah virus, hantavirus, tickborne hemorrhagic fever viruses, tickborne encephalitis viruses, yellow fever virus, and multidrugresistant tuberculosis.

EPIDEMIOLOGY OF DISEASE CAUSED BY TERRORISM LIKELY AGENTS OF BIOTERRORISM Bioterrorism involves the intentional, malevolent use of diseasecausing biological agent(s) to cause sickness, death, destruction, or panic. The harm caused by bioterrorism may involve human beings, animals, or agriculture, and its goal is political or social gain. Theoretically, there are many biological agents that could be used for these purposes. The United States Centers for Disease Control and Prevention (CDC) has identified and categorized the biological agents most likely to be used in an attack by terrorists (Table 5-1).1 Category A agents have been assigned the highest priority for defensive preparation because they can be easily disseminated from person to person, they have the potential to cause high mortality, public panic and social disruption, and they require special public health preparedness. The Category A agents include the organisms that cause anthrax (B. anthracis), smallpox, plague (Yersinia pestis), botulism (toxin of C. botulinum), and viral hemorrhagic fevers (Ebola, Marburg, Lassa, and others). Category B includes Coxiella burnetii (Q fever), Brucella species (brucellosis), Burkholderia mallei (glanders), alphaviruses (Venezuelan encephalomyelitis, eastern and western equine encephalomyelitis), ricin toxin (from Ricinus communis, found in castor beans), epsilon toxin (from Clostridium perfringens), and enterotoxin B from Staphylococcus, as well as food- and water-borne pathogens, including Salmonella species, Shigella dysenteriae, enterohemorrhagic Escherichia coli, Vibrio cholerae, and Cryptosporidium parvum. The category B agents are somewhat more

In some instances a single case of recognizable illness should alert the clinician that terrorism is likely. For instance, if a single case of smallpox, a disease that was declared to have been eradicated in 1980, were suspected anywhere in the world, terrorism would be the likely explanation. Similarly, a diagnosis of plague made in the northeastern part of the United States should raise concern for terrorism. Disease presenting in unusual numbers, in an age group that is uncharacteristic for a given pathogen, in a more severe form than usually expected, or among individuals with a common opportunity for exposure, e.g., who attended the same function, should provoke suspicion for terrorism. A more complete list of epidemiologic clues that should suggest bioterrorism is provided in Box 5-1.

CHILDREN AND BIOTERRORISM There is scant historical evidence on which to base predictions regarding particular risks of infants and children in the face of a biological attack, but there are some physiological and behavioral characteristics of children that suggest their disproportionate vulnerability2 (Box 5-2). The respiratory rate of children is more rapid than that of adults, their intake of air on a per-weight basis is greater, and their skin is more permeable. These characteristics may allow enhanced absorption of toxins. The higher skin-to-mass ratio and

BOX 5-1. Epidemiologic Clues That Should Suggest Bioterrorism

TABLE 5-1. Categories of Potential Agents of Bioterrorism as Determined by the Centers for Disease Control and Prevention Category A Agents

Category B Agents

Category C Agents

Bacillus anthracis (anthrax) Smallpox Yersinia pestis (plague) Francisella tularensis (tuleremia) Clostridium botulinum toxin (botulism) Viral hemorrhagic fevers (Ebola, Marburg, Lassa, others)

Coxiella burnetii (Q fever) Brucella species (brucellosis) Burkholderia mallei (glanders) Alphaviruses (including Venezuelan encephalomyelitis, eastern and western equine encephalomyelitis) Ricinus communis toxin (ricin) (from castor beans) Clostridium perfringens epsilon toxin Staphylococcus enterotoxin B Salmonella species Shigella dysenteriae Enterohemorrhagic Escherichia coli Vibrio cholerae Cryptosporidium parvum

Nipah virus Hantavirus Tickborne hemorrhagic fever viruses Tickborne encephalitis viruses Yellow fever virus Multidrug-resistant Mycobacterium tuberculosis

• • • • • • • • • • • • • • • • • •

Single case of disease caused by an uncommon agent Unusual illness in a particular age group Disease with an unusual geographic or seasonal distribution Large number of ill persons with similar disease or syndrome Unusual route of exposure for a particular pathogen, e.g., inhalational anthrax Disease usually transmitted by a vector not found in the region, e.g., plague in an individual who lives in the northeastern United States Higher morbidity and mortality than expected with a common disease or syndrome Failure of a common disease to respond to usual therapy Multiple unusual or unexplained disease entities coexisting in the same patient without other explanation Multiple atypical presentations of disease agents Similar genetic type, organism variant, or antimicrobial pattern among agents isolated from temporally or spatially distinct sources Unusual, atypical, genetically engineered, or antiquated strain of agent Endemic disease with unexplained increase in incidence Simultaneous clusters of similar illness in noncontiguous areas, domestic or foreign Atypical aerosol, food, or water transmission Ill people presenting near the same time Death or illness among animals that precedes or accompanies illness or death in humans No illness in people not exposed to common ventilation systems, but illness among those people in proximity to the systems

Adapted from Buehler JW, Berkelman RL, Hartley DM, et al. Syndromic surveillance and bioterrorism-related epidemics. Emerg Infect Dis 2003 October. Available from URL: http://www.cdc.gov/ncidod/EID/vol9no10/03-0231.htm.1

38

SECTION

A Epidemiology & Control of Infectious Diseases

reduced fluid reserves of children increase their susceptibility to the dehydrating effects of gastrointestinal losses. The immaturity of the immune system in infants and children may enhance their susceptibility to severe disease if infected with some microbes or toxins. Children may be less physically able to escape a biological threat, and their dependence upon adult caretakers, who may suffer illness or death or be quarantined, could allow them to become secondary victims, even if they were not directly affected by infection or intoxication. In addition, facilities, equipment, pharmaceutical products, and training of emergency response personnel designed to

BOX 5-2. Physiological and Behavioral Vulnerabilities of Infants and Children in the Face of a Bioterrorist Attack • Increased inhalation of air per weight • Increased skin permeability • Higher skin-to-mass ratio and reduced fluid reserve, leading to greater susceptibility to dehydration • Immature immune system • Dependence on adult caretakers who may become ill, die, or be quarantined • Inability to verbalize exposure or symptoms • Lack of availability of equipment, expertise, medications, and personnel appropriate for infants and children • Inability to express postevent anxiety and distress Adapted from American Academy of Pediatrics Committee on Environmental Health and Committee on Infectious Diseases. Chemical-biological terrorism and its impact on children. Pediatrics 2006;118:1267–1278.2

accommodate the needs of infants and children are often not readily available. Finally, children are less able and less likely to verbalize the postevent distress they may feel after witnessing or being a victim of an act of terrorism.

RECOGNITION OF DISEASE DUE TO BIOTERRORISM Recognition of a bioterrorist event depends upon both system-wide detection of increased numbers of patients presenting with similar manifestations of illness, and upon individual clinicians who are knowledgeable about the signs and symptoms produced by agents likely to be used in such an event. Development of systems to enhance detection of outbreaks, termed syndromic surveillance, has been emphasized as an important component of the national effort to recognize public health emergencies, whether due to terrorism or to naturally occurring disease. Automated syndromic surveillance mechanisms are now being used in some parts of the country. Equally important is the role of the astute clinician who recognizes that something about the presentation of an individual patient suggests disease potentially caused by intentional exposure to an agent of bioterrorism. Table 5-2 lists the likely manifestations of illness following exposure to the agents identified by the CDC as likely weapons of bioterrorism. Because of the similarity of these signs and symptoms to those associated with common community-acquired illnesses, and because many of the agents anticipated to be used by terrorists are also causes of community-acquired infection, recognition that an individual patient has been exposed to an agent of biological terrorism early in the course of the illness is unlikely. More

TABLE 5-2. Clinical Manifestations Associated with Likely Agents of Bioterrorism Clinical Manifestations

Agents/Disease

RESPIRATORY SYMPTOMS

Influenza-like illness with or without atypical pneumonia and cough Purulent conjunctivitis with preauricular or cervical lymphadenopathy Exudative pharyngitis and cervical lymphadenopathy

Tularemia, brucellosis, Q fever, alphavirus (Venezuelan, eastern, and western encephalitis), inhalational anthrax, pneumonic plague, inhalational tularemia, ricin, aerosol exposure to staphylococcal enterotoxin B, hantavirus Oculoglandular tularemia Oropharyngeal tularemia

DERMATOLOGIC DISEASE

Vesicular rash Painless ulceration progressing to black eschar Ulcer plus painful regional lymphadenopathy Petechiae Shock

Smallpox Cutaneous anthrax Ulceroglandular tularemia Viral hemorrhagic fever Inhalational anthrax, ricin, viral hemorrhagic fever

HEMATOLOGIC DISEASE

Thrombocytopenia Neutropenia Hemorrhage Disseminated intravascular coagulation

Brucellosis, viral hemorrhagic fever, hantavirus Viral hemorrhagic fever, alphavirus (Venezuelan, eastern, and western encephalitis) Viral hemorrhagic fever Viral hemorrhagic fever

NEUROLOGIC FINDINGS

Flaccid paralysis Encephalitis Meningitis

Botulism Alphavirus (Venezuelan, eastern, and western encephalitis) Inhalational anthrax, septicemic and pneumonic plague, alphavirus (Venezuelan, eastern, and western encephalitis)

GASTROINTESTINAL DISEASE

Diarrhea Vomiting, abdominal pain, gastrointestinal bleeding

Salmonella spp., Shigella dysenteriae, Escherichia coli O15:H7, Vibrio cholerae, Cryptosporidium parvum Gastrointestinal anthrax

RENAL DISEASE

Hemolytic–uremic syndrome, thrombotic thrombocytopenic purpura Renal failure Painful lymphadenopathy

Escherichia coli O157:H7 Viral hemorrhagic fever, hantavirus Bubonic plague

Adapted from American Academy of Pediatrics. Recommendations for care of children in special circumstances: biological terrorism. In: Pickering LK (ed) Red Book: 2006 Report of the Committee on Infectious Diseases, 26th ed. Elk Grove Village, IL, American Academy of Pediatrics;2003:100–101.

PART I Understanding, Controlling, and Preventing Infectious Diseases

Bioterrorism

pathognomonic findings associated with likely agents of bioterrorism may become apparent later in the course. Those findings that are more likely to raise suspicion for a specific agent are listed in Table 5-3.

DIAGNOSIS AND MANAGEMENT OF SUSPECTED BIOTERRORISM EVENT Accurate diagnosis is important in the management of any infectious disease; if bioterrorism is suspected, identification and isolation of the causative organism are critical to further investigation and to subsequent public health efforts. Collection of appropriate diagnostic specimens, prior to treatment with antimicrobial agents if possible, will allow targeted treatment, appropriate preventive measures for contacts, effective infection control, and a more rapid and effective public health effort. Table 5-4 lists recommended microbiologic

CHAPTER

5

39

diagnostic procedures, treatment options, postexposure prophylaxis recommendations, and appropriate isolation precautions for the CDC’s list of likely agents of bioterrorism.3 If bioterrorism is suspected, however, affected patients should be isolated using contact precautions and airborne infection isolation, as well as standard precautions, until test results are available and transmissibility of the illness can be determined.

NOTIFICATION OF PUBLIC HEALTH AUTHORITIES Local public health authorities should be notified as soon as an illness associated with terrorism is suspected. Local authorities may initiate investigation, but they also will likely involve the state agency and the CDC. State health department and state epidemiologists are listed on the websites provided below.

TABLE 5-3. Distinctive Findings Associated with Some Likely Agents of Bioterrorism Biological Agent

Pathognomonic (Classical) Findings

Inhalational anthrax

Hemoptysis, with chest radiograph demonstrating widened mediastinum

Cutaneous anthrax

Vesicular lesion that develops into necrotic ulceration with black eschar

Botulism

Descending flaccid paralysis that begins with dysphagia, dysarthria, diplopia, dysphonia, and ptosis

Plague, inhalational exposure

Fever, cough, hemoptysis, chest pain, with chest radiograph showing bronchopneumonia

Plague, cutaneous exposure

Swollen regional lymph nodes

Q fever

Fever, chills, headache, weakness, and sweating with elevated serum hepatic enzyme levels

Smallpox

Fever, headache, backache, abdominal pain, malaise, followed by a rash that begins on the face and progresses in a centrifugal fashion to involve extremities and then trunk. The lesions evolve from papules to firm vesicles, and then deep-seated, hard pustules

Staphyloccal enterotoxin, ingestion

Abrupt onset of nausea, abdominal cramps, vomiting, and prostration, often accompanied by diarrhea

Staphyloccal enterotoxin, inhalation

Nonproductive cough, retrosternal chest pain, dyspnea, and fever, usually without evidence of pulmonary involvement on chest radiograph

Tularemia, cutaneous exposure

Systemic illness, with painful regional lymphadenopathy, with or without cutaneous ulcers

Tularemia, ocular exposure

Purulent conjunctivitis, with chemosis, periorbital edema, and conjunctival nodules or ulceration, and accompanying preauricular or cervical lymphadenopathy

Ricin, ingestion

Vomiting, hemorrhagic gastroenteritis, shock, and cardiovascular collapse

Ricin, inhalation

Respiratory distress, with necrotizing pneumonitis

Ricin, injection

Rapid onset of shock and cardiovascular collapse

Viral hemorrhagic fevers

Fever, myalgia, prostration, petechiae, progressing to shock, mucous membrane hemorrhage, with or without renal involvement

Adapted from American Academy of Pediatrics. Recommendations for care of children in special circumstances: biological terrorism. In: Pickering LK (ed) Red Book: 2006 Report of the Committee on Infectious Diseases, 27th ed. Elk Grove Village, IL, American Academy of Pediatrics, 2006:107–111.

40

SECTION

A Epidemiology & Control of Infectious Diseases

TABLE 5-4. Biological Weapons: Recommended Diagnostic Procedures, Isolation, and Treatment of Children Agent

Diagnostic Sample(s)

Treatment Options

Prophylaxisa

Isolation Precautions

Alphaviruses (Venezuelan encephalomyelitis, eastern and western equine encephalitis)

CSF for virus isolation and antibody testing; acute and convalescent serum for antibody testing

Supportive

Protection from mosquito vectors

Standard; respiratory precautions for western equine encephalitis virus

Anthrax

Gram stain of buffy coat, CSF, pleural fluid, swab of skin lesion; culture of blood, CSF, pleural fluid, skin biopsy

Ciprofloxacinb or doxycyclinec; combine with 1 or 2 additional antimicrobial agents for inhalational, gastrointestinal, or oropharyngeal diseased

Ciprofloxacin or Standard; contact for doxycyclinec or skin lesions e amoxicillin ; anthrax vaccine

Botulism

Serum, stool, enema fluid, Supportive care; gastric fluid or vomitus, mechanical suspected food samples ventilation and for toxin detection; parenteral nutrition culture of stool or gastric may be required. secretions; nerve Equine botulism conduction testing antitoxin given as soon as possible (CDC)f

Brucellosis

Culture of blood or bone marrow; acute and convalescent serum for antibody testing

Doxycyclinec and rifampin; if < 8 years use trimethoprimsulfamethoxazole (TMP-SMX)

Plague

Culture or FA staining of blood, sputum, lymph node aspirateg

Q fever

Comments

Additional antimicrobial agents to be used for inhalational, gastrointestinal, or oropharyngeal disease include rifampin, vancomycin, penicillin, ampicillin, chloramphenicol, imipenem, clindamycin, and clarithromycin

Standard

Type-specific antitoxin should be administered when possible; antitoxin prevents additional nerve damage but does not reverse existing paralysis

Doxycyclinec and rifampin

Standard; contact for draining skin lesions

TMP-SMX; TMP-SMX may substitute for rifampin with doxycycline

Streptomycin or gentamicin; doxycyclinec or tetracyclinec

Doxycyclinec; tetracyclinec

Droplet

TMP-SMX is alternative to chloramphenicol for meningitis

Acute and convalescent serum for antibody testing

Doxycyclinec or tetracyclinec

Doxycyclinec or tetracyclinec

Standard

Chloramphenicol is an alternative for treatment/prophylaxis

Ricin

Serum and/or respiratory secretions for EIA detection

Supportive care; gastric lavage and cathartics if toxin is ingested

Protective mask

Standard

Smallpox

Swab of pharyngeal Supportive care secretions or skin lesions. to be sent to the CDC for isolation of virusg

Vaccine if administered Airborne, contact within 4 days

Staphylococcal enterotoxin B

Serum, urine, and respiratory secretions for toxin testing; acute and convalescent serum for antibody testing

Supportive care

None available

Standard

Tularemia

Gastric aspirate, sputum, pharyngeal exudates, conjunctival exudates, lymph node aspirate, swab from ulcer, and blood for cultureg, direct fluorescent antibody staining, and PCR testing. Blood for antibody testing

Gentamicin

Doxycyclinec

Standard

PART I Understanding, Controlling, and Preventing Infectious Diseases

Doxycyclinec, ciprofloxacinb, and chloramphenicol are alternatives for treatment

Passive Immunization

CHAPTER

6

41

TABLE 5-4. Biological Weapons: Recommended Diagnostic Procedures, Isolation, and Treatment of Children—Continued Agent

Diagnostic Sample(s)

Treatment Options

Viral hemorrhagic fevers

Culture and/or antigen detection of blood and other body tissuesh; serum for acute and convalescent antibody testing

IV ribavirin for Lassa fever; plasma from convalescent patients for Argentine hemorrhagic fever; supportive care

Prophylaxisa

Isolation Precautions

Comments

Standard, droplet, and contacti

CDC, Centers for Disease Control and Prevention; CSF, cerebrospinal fluid; EIA, enzyme immunoassay; FA, fluorescent antibody; IV, intravenous; PCR, polymerase chain reaction. a Prophylaxis should only be administered after consultation with public health officials in situations in which exposure is highly likely. The duration of prophylaxis has not been determined for most agents. b If susceptibility testing is unknown or indicates resistance to other agents. Ciprofloxacin is licensed by the United States Food and Drug Administration (FDA) for use in people younger than 18 years of age for treatment of or prophylaxis against anthrax, but it is not licensed for those younger than 18 years for treatment of tularemia. Treatment with ciprofloxacin is warranted for those less than 18 years for selected serious infections. c Tetracyclines, including doxycycline, are not FDA-approved and are usually contraindicated for children younger than 8 years of age, but treatment is warranted for selected serious infections. d Treatment should initially be administered parenterally but may be changed to oral therapy for cutaneous infection without dissemination. e Amoxicillin may only be used as prophylaxis if the organism is known to be susceptible. f Botulism antitoxin must be obtained from the Centers for Disease Control and Prevention Drug Service, 404/639-3670 (weekdays, 8–4:30) or 404/639-2888 (weekends, nights, holidays). g The microbiology laboratory should be notified of suspected pathogen so that appropriate safety precautions can be undertaken. h Isolation should only be attempted under biosafety level-4 conditions. i Because of the risk of nosocomial transmission the state health department and the Centers for Disease Control and Prevention should be contacted for specific advice about management and diagnosis of suspected cases. Adapted from American Academy of Pediatrics. Recommendations for care of children in special circumstances: biological terrorism. In: Pickering LK (ed). Red Book: 2006 Report of the Committee on Infectious Diseases, 27th ed. Elk Grove Village, IL, American Academy of Pediatrics, 2006:107–111.

SECTION

B

Prevention of Infectious Diseases

CHAPTER

6

Passive Immunization David C. Goldman

Passive immunization provides individuals with preformed antibodies that can prevent or treat infectious diseases. Over a century ago, hyperimmune animal sera were produced to treat specific infections. After human plasma fractionation was developed (during World War II), immune globulin (human) (IG) became available for passive immunization. Although this was an enormous breakthrough, intravenous (IV) infusion of the product was found to evoke a variety of severe adverse reactions. Therefore, this route of administration was precluded, and IG was largely limited to intramuscular (IM) injection. For some indications, this was not a disadvantage. However, for treatment of primary immunodeficiency, it became clear that the volume (and hence the amount of immunoglobulin G (IgG)) that could be administered was suboptimal. In 1981, the first (United States-

licensed) (human) IGIV was approved by the Food and Drug Administration. It dramatically changed immunoglobulin therapy. Using IGIV enabled clinicians to give large doses of IgG with minimal discomfort and to produce an immediate rise in both total plasma IgG and titers of specific antibodies. The products available for passive immunization can be grouped as follows: (1) IG; (2) specific IGs for IM use; (3) IGIV; and (4) specific IGs for IV administration (Table 6-1). In addition, two antitoxins of animal origin are still available for limited distribution, and there is now a licensed “humanized” monoclonal antibody. Passive immunization can be used for a variety of clinical indications, including: (1) treatment of primary and, in certain cases, secondary immunodeficiency or its sequelae; and (2) prophylaxis against infections due to specific organisms. In general, treatment of established infections has been less successful, even when the specific organism or toxin can be identified. However, the use of IGIV for the treatment of various conditions that involve immune dysregulation, such as immune thrombocytopenic purpura (ITP) and Kawasaki disease, has become routine and has stimulated much investigation of immunomodulation. In the following sections, the various products and their uses are discussed. An effort has been made to use nonproprietary product names exactly as they appear in the labeling, but for brevity, there is liberal use of abbreviations. Emphasis is placed on the approved

42

SECTION

B Prevention of Infectious Diseases

TABLE 6-1. Immune Globulins Prepared from Human Plasma

TABLE 6-2. Available Immune Globulin Intravenous Products in the United States

Nonproprietary Name

Abbreviation

FOR INTRAMUSCULAR ADMINISTRATION

Immune globulin (human)

IG

Hepatitis B immune globulin (human)

HBIG

Rabies immune globulin (human)a

RIG

Tetanus immune globulin (human)

TIG

Varicella-zoster immune globulin (human)b

VZIG/VariZIG

FOR INTRAVENOUS ADMINISTRATION

Product Registered Name

Manufacturer

FDA-Approved Indication

Carimune NF

ZLB Behring

ITP, PID

Vivaglobulin

ZLB Behring (subcutaneous)

PID

Gammar-P IV

ZLB Behring

PID

Flebogamma (previously Griflos USA Venoglobulin-S)

PID

Immune globulin intravenous (human)

IGIV

Gammagard Liquid

Baxter

PID

Cytomegalovirus immune globulin intravenous (human)

CMV-IGIV

Gammagard S/D (powder: phasing out)

Baxter

CLL, ITP, KD, PID

Botulism immune globulin intravenous (human)

BIG-IGIV

Iveegam EN

Baxter

KD, PID

Vaccinia immune globulin intravenous (human)

VIG-IGIV

Gamunex

Talecris Biotherapeutics (previously Bayer)

ITP, PID

Octagam

Octapharma USA

WinRho SDF

Baxter

a

As much of the dose as possible should be instilled around the wound. b See text.

indications (i.e., those that appear in the product package inserts). In the case of IGIV, not all products have been approved or studied for all indications (Table 6-2). Moreover, in contrast to IGs for IM administration, IGIV products undergo a variety of manufacturing processes, utilize a number of different stabilizers and excipients, are formulated in various concentrations or physical states, and may differ in subtle ways (e.g., IgA content). Compounding this variety is the fact that licensed manufacturers are constantly studying process changes in order to improve the products, and additional manufacturers have products undergoing clinical trials in the hope of bringing them to market. The clinician should always consult the current package insert for the specific IGIV product being used in addition to more general sources of information.

IMMUNE GLOBULIN (HUMAN) IG is prepared by cold alcohol fractionation of pooled human plasma. It is formulated as a 16.5% protein solution. At least 96% of the total protein is IgG; small quantities of IgM and IgA are present. Each lot of product must represent the pooled plasma of at least 1000 donors so as to provide a wide diversity of antibodies. In practice, however, each lot represents many more donors (up to 60,000). Individual donations of the plasma, like those used as the source for all human plasma derivatives, are screened for markers of a variety of viruses (hepatitis B, hepatitis C, human immunodeficiency) to minimize the potential for transmission of infections. Other steps taken to minimize such transmission are noted at the end of this chapter. A major use of IG is for postexposure prophylaxis of hepatitis A. When injected deep into a large muscle mass within 14 days of exposure, IG can prevent symptomatic infection. The usual dose is 0.02 mL/kg, given as soon as possible after exposure. If re-exposure is likely and the patient is at least 2 years old, hepatitis A vaccine should be administered concomitantly, in a separate site with a separate syringe. IG is also effective for pre-exposure prophylaxis of hepatitis A in travelers to areas where hepatitis A is prevalent. However, for travelers who are at least 1 year old and whose departure is not imminent, hepatitis A vaccine has largely replaced IG. If IG is to be used, the dose depends on the circumstances. For a child younger than 1 year, the dose is 0.02 mL/kg if the anticipated stay is 3 months or less and 0.06 mL/kg if it is longer. In older persons whose departure is imminent, the dose of IG is 0.02 mL/kg, concomitant with the vaccine, for a stay of up to 5 months; and 0.06 mL/kg, concomitant with vaccine, for a longer stay. As in the case of postexposure prophylaxis, if IG and hepatitis A vaccine are given simultaneously, separate sites and separate syringes should be used.

PID Chronic or acute ITP and HIV-immune related conditions

CLL, chronic lymphocytic leukemia; FDA, Food and Drug Administration; HIV, human immunodeficiency virus; ITP, immune thrombocytopenia purpura; KD, Kawasaki disease; PID, primary immunodeficiency.

Vaccination is, by far, the major strategy for achieving protection against measles. It is highly effective. Nonetheless, prophylaxis for measles remains an important indication for IG in children younger than 1 year, in older children who have not been vaccinated, and in immunocompromised children who are not receiving routine immunoglobulin replacement therapy. Prophylaxis is indicated for susceptible household or hospital contacts. A single dose of 0.25 mL/kg given as soon as possible (and within 6 days) after exposure in high-risk, susceptible individuals prevents or modifies infection. Immunocompromised children should receive 0.50 mL/kg. (In either case, no more than 15 mL should be injected, and small children should be given no more than 3 mL in any single site.) A child who is receiving routine IgG replacement therapy is already protected, and no further dosage is indicated. Once an appropriate period has elapsed, every immunocompetent child who is at least 1 year old should be vaccinated.

SPECIFIC IMMUNE GLOBULINS Specific IGs for IM use are IgG preparations produced from plasma that is selected by screening donations or collected from donors who have been deliberately immunized or given immune booster therapy. Either approach can ensure the presence of high levels of antibody directed against one or more specific antigens. The manufacturing process is essentially the same as that used to prepare IG. The protein concentration differs for individual products; it can be found in the “Description” section of the package insert. Specific products available for IM administration are listed in Table 6-1. They are used to prevent hepatitis B, rabies, tetanus, and varicella-zoster.

Hepatitis B Immune Globulin Used in conjunction with hepatitis B vaccine, hepatitis B IG (human) (HBIG) has proved to be very effective in preventing disease and subsequent chronic infection in infants of mothers who are chronic carriers of hepatitis B virus (HBV). All pregnant women should be tested for circulating hepatitis B surface antigen (HBsAg). If the

PART I Understanding, Controlling, and Preventing Infectious Diseases

Passive Immunization

mother is positive for HBsAg, the infant should receive 0.5 mL of HBIG as soon as possible after birth. Active immunization with the vaccine should begin at once (with the use of a separate site and a separate syringe), and subsequent doses of vaccine should be given according to the recommended schedule. In general, HBIG is not recommended if the mother’s HBsAg status is not known, but the vaccination series should begin at once. Ideally, in such a case, the mother should be tested as soon as possible after delivery. If the test results are positive, the infant should receive 0.5 mL of HBIG as soon as possible and within 7 days of birth, and the vaccination series should be completed according to schedule. Preterm infants (< 2 kg) constitute a special class. If the mother’s HBsAg status is unknown and cannot be determined rapidly, the child should be given 0.5 mL of HBIG as well as vaccine, because the premature infant’s immune system may have a suboptimal response to the vaccine. HBIG is also used (concomitantly with vaccine) if an unimmunized or inadequately immunized child is exposed to blood or another body fluid from a person known or suspected to be acutely infected with hepatitis B. Such exposure could be percutaneous or permucosal; it could occur, for example, through a bite or by sexual contact. In such cases, the dose is 0.06 mL/kg.

Rabies Immune Globulin Rabies IG (human) (RIG) is always used in association with rabies vaccine. Its sole indication is for postexposure prophylaxis. In general, such prophylaxis is recommended whenever the skin is broken and the child is exposed by bite, scratch, or other means to fluid from a wild animal or a domestic animal known or thought to be rabid. However, the decision to immunize should be made after consultation with local health authorities. If immunization is indicated, RIG should be given concomitantly with the first dose of vaccine, and dosage of the latter should be continued according to the postexposure schedule. The dose of RIG is 20 IU/kg. The current recommendation is that as much of the product as possible should be instilled around the cleaned wound. (If, for example, there are multiple bite sites, the dose should be divided among them.) Any of the dose that cannot be administered by this route should be given IM with a new syringe. Likewise, the first dose of vaccine should be given in a different part of the body and with a separate syringe. RIG is supplied in 2-mL (300IU) and 10-mL (1500-IU) vials. Thus, the contents of a 2-mL vial can be used to treat a child of up to 15 kg; for larger children, additional vials are required. A 10-mL vial is sufficient to treat a 75-kg adult, making it less suitable for pediatric use. However, if the smaller vials are not available, an appropriate portion can be withdrawn to treat a child who is in need of immediate prophylaxis. The product contains no preservative; once a vial has been entered, the contents should be administered promptly, and the remainder discarded.

Tetanus Immune Globulin Tetanus IG (human) (TIG) is the oldest of the specific IGs; it is still recommended for the treatment of tetanus and for prophylaxis in certain circumstances. Tetanus is so rare in persons whose immunization history includes at least three doses of tetanus toxoid that TIG is not recommended for such individuals. On the other hand, if a child who has received fewer than three doses of the toxoid (or whose tetanus immunization history is unknown) sustains a wound other than a clean, minor wound, TIG is indicated. The dose is 250 U, that is, the entire contents of a single container (vial or prefilled syringe), given IM. At the same time, the tetanus toxoid-containing vaccine that is appropriate for the age of the child should be administered at a separate site with a separate syringe. Arrangements should be made to complete the active immunization series. TIG is also recommended as part of the therapeutic regimen (including wound cleaning and debridement, antibiotics, and supportive care) for treatment of tetanus. The recommended dose for

CHAPTER

6

43

treatment is 3000 to 6000 U, that is, the contents of 12 to 24 containers. Little information is available that permits a choice between these extremes. Moreover, doses of TIG as low as 500 U (the contents of two containers) have been reported to be effective, for example, in tetanus neonatorum.2

Varicella-Zoster Immune Globulin Postexposure prophylaxis should be targeted toward exposed, susceptible individuals (i.e., without a history of disease or ageappropriate immunization) who are at high risk for the development of severe varicella. Such individuals include: (1) immunocompromised children; (2) pregnant women; (3) neonates whose mothers have signs and symptoms of varicella within 5 days before delivery or within 48 hours after delivery; (4) premature infants > 28 weeks’ gestation who are exposed in the neonatal period and whose mothers do not have evidence of immunity; (5) premature infants born at < 28 weeks’ gestation or who weigh < 1000 g at birth and are exposed during the neonatal period regardless of maternal history of varicella disease or vaccination. Significant exposure can occur through residence in the same household, close contact with a playmate, proximity to a contagious patient, or contact with a contagious visitor or hospital staff member. In the past, varicella-zoster IG (human) (VZIG) was used routinely in the postexposure prophylaxis following varicella exposure in certain situations. The only manufacturer of VZIG discontinued production in 2004 and supplies were depleted by 2006. Physicians should access www.cdc.gov/nip/home-hcp.htm for relevant recommendations. In February 2006, an investigational VZIG product, VariZIG (Cangene Corporation, Winnipeg, Canada) became available under an investigational new drug application to the Food and Drug Administration (FDA).3 Similar to VZIG, VariZIG is a purified human IG preparation made from plasma containing high levels of antivaricella IgG antibodies. This product is lyophilized, and after reconstitution is administered IM within 96 hours of exposure. The dose is 125 U (1 vial)/10 kg body weight up to a maximum of 625 U (5 vials). The minimum dose is 125 U. If obtainable in < 96 hours postexposure, VariZIG is preferred to IGIV or acyclovir. It is distributed by FFF Enterprises (Temecula, California: 24-hour telephone, 800-843-7477) under expanded access protocol. Pharmacists and healthcare providers who expect to need VariZIG can participate in a program that permits acquisition of an inventory in advance. If VariZIG cannot be obtained within 96 hours postexposure, IGIV can be given at a dose of 400 mg/kg. Although licensed IGIV preparations are known to contain antivaricella antibody, titer in any specific lot is uncertain. An alternative to postexposure prophylaxis with IGIV is oral acyclovir, which should be started 7 to 10 days after exposure and continued for a total of 7 days. This approach can be immunocompetent considered in immunocompetent, seronegative adults with significant exposure. The dose for adults is 800 mg given 4 times a day. If a child requires acyclovir pre-emptively, the dose is 40 to 80 mg/kg per day, divided in four doses. (See Chapter 205, Varicella-Zoster Virus.) Generally, exposed immunocompetent children should be given varicella vaccine within 96 hours of exposure if they have not previously received vaccine and have no contraindication.

IMMUNE GLOBULIN INTRAVENOUS (HUMAN) Like IG, IGIV is prepared from large pools of plasma that can represent as many as 60,000 donors. The early steps in its manufacture are similar to those used for preparing IG. Further processing of individual products is done by a variety of techniques, including steps such as polyethylene glycol precipitation, ion exchange, and exposure to low pH. This variety reflects, in part, the lack of consensus on the optimal procedure for preparing IgG on a commercial scale in a form that is safe for IV administration. There is a diversity of final formulations (e.g., freeze-dried or in solution, different protein concentrations

44

SECTION

B Prevention of Infectious Diseases

of the latter, different pH, and various stabilizers and other excipients that can be used in a number of different combinations). This variety and the dynamic state of the industry render any compilation of manufacturing methods or formulations rapidly obsolete. The clinician who intends to use a particular product should therefore consult the “Description” section of the package insert for a brief synopsis of the product’s preparation and properties. Although each lot of IGIV must contain at least minimal levels of antibodies to certain infectious organisms or toxoids (measles, polio, and diphtheria), these levels were established to ensure lot-to-lot consistency rather than to match the clinical indications for individual products. Even though the use of large plasma pools enhances this consistency, IGIV products differ in sources of plasma as well as processing methods. As a result, antibody titers to other bacterial and viral pathogens can vary signiÀcantly among preparations and lots.4,5 In addition, processing steps can alter functional capabilities of IgG or change the relative distribution of immunoglobulin subclasses. These changes may or may not be reflected in antibody titers measured by routine laboratory tests. Moreover, most individual products have undergone clinical trials for only a certain subset of indications. In addition, even when many IGIV products carry the same indication, direct comparisons of the efÀcacy of multiple products in a single clinical trial are rare. For all of these reasons, IGIV cannot be considered a uniform generic product. Available products in the United States (2006) are shown in Table 6-2. Certain physiologic properties intrinsic to IGIV preparations (and variable among products) should be considered in all individuals for whom administration is considered and especially in those who have underlying conditions. These include high volume load, sodium content and osmolality, especially in neonates and young children as well as those with cardiac disease, renal impairment or thromboembolic risk; 2% to 10% carbohydrate content, in those with diabetes or renal impairment; low pH in neonates; and average IgA content (< 2 to 720 mg/mL) in those with anti-IgA antibodies.

Approved Indications There are currently 6 FDA approved indications for IGIV: (1) replacement therapy for primary immunodeÀciency; (2) treatment of idiopathic thrombocytopenic purpura; (3) prevention of infection and graft-versus-host disease in adult bone marrow transplant recipients; (4) treatment of Kawasaki disease; (5) prevention of infection in children with HIV; and (6) replacement therapy in individuals with chronic B-cell lymphocytic leukemia (Table 6-3).

Primary Immune Deficiencies During the past 20 years, IGIV has become the drug of choice for replacement therapy in primary immunodeÀciency and all IGIV products carry this indication.6 Although most of the subjects enrolled in IGIV efÀcacy trials have had the more common primary immunodeÀciencies, clinical experience with IGIV in a wide variety of these conditions has been favorable. Patients with the following primary immunodeÀciencies have associated hypogammaglobulinemia and may beneÀt from replacement IGIV therapy: X-linked agammaglobulinemia, common variable immunodeÀciency, severe combined immunodeÀciency and hyper-IgM syndrome. As clinical experience has grown, physicians have learned to tailor dosage to the individual patient. The objective is to determine the dose and regimen that will achieve and maintain freedom from serious infections. Monitoring the trough levels of IgG in parallel with the patient’s clinical course has been helpful in accomplishing this objective. As could be expected, a range of effective doses, dosing schedules, and trough levels has been observed. However, the usual schedule for IV infusion is once every 3 or 4 weeks, and typical doses have been approximately 400 mg/kg with a dosing range of 300 to 800 mg/kg. Trough concentrations of IgG should be maintained in the

TABLE 6-3. Approved Indications for Immune Globulin Intravenous Therapy Indications

Dosage

Comments

REPLACEMENT THERAPY

Primary immunodeficiency Chronic lymphocytic leukemia Bone marrow transplantation Human immunodeficiency virus infection in children

~400 mg/kg q4 weeks Adjust dose according (average dose) to individual response 400 mg/kg q3–4 weeks Cost-effectiveness questioned 500 mg/kg q1 week Other anti-infective therapies also effective 400 mg/kg q2–4 Useful in selected weeks symptomatic patients

IMMUNE MODULATION

Idiopathic thrombocytopenic purpura (ITP)

400 mg/kg daily for 5 days or 1 g/kg, single dose

Kawasaki disease

2 g/kg, single dose

Useful in acute and chronic immune thrombocytopenia purpura Treat before 10 days of fever

range of 400 to 500 mg/dL.7 In 2006, the United States FDA approved IG subcutaneous (Vivaglobulin, ZLB Behring) for people with primary immunodeÀciency. It is the Àrst subcutaneous IG approved that can be self-administered weekly (158 mg/kg) using a portable pump.

Secondary Immune Deficiencies Chronic lymphocytic leukemia (CLL) and the immune suppression that occurs in patients undergoing bone marrow transplantation (BMT) result in quantitative and/or qualitative humoral immunodeÀciency. One IGIV product has been studied and approved for treatment of CLL, and another for patients undergoing BMT. However, although IGIV reduces infections in these conditions, selection of patients most likely to beneÀt, optimal timing and duration of administration, and the relative effectiveness of IGIV and other anti-infective therapies are unresolved issues. Infection with human immunodeÀciency virus (HIV) causes dysregulation of humoral immunity with impaired functional antibody activity. One IGIV product is approved for use in HIV-infected children. Such children were shown to have fewer respiratory infections while receiving IGIV; hence, it may beneÀt selected patients.8,9 In these studies, a beneÀt was observed, for children with CD4 counts higher than 200/mm3. Antimicrobial prophylaxis (e.g., trimethoprimsulfamethoxazole) is an alternative approach to IGIV for preventing bacterial infections and is effective in preventing Pneumocystis carinii pneumonia.10

Immunomodulation Mechanism of immunomodulation by IGIV and its use in treatment of inflammatory and autoimmune diseases have been reviewed recently.11 Potential anti-inflammatory mechanisms include: neutralization of pathologic autoantibodies, enhanced clearance of autoantibodies, inhibition of phagocytosis by effector cells, altered complement deposition, altered cytokine expression, neutralization of superantigens and modulation of lymphocyte function (reviewed in Knezevic-Maramica & Kruskall12). Because “immunomodulation” does not appear as a separate category in the labeling of IGIV, the labeled indications that belong to this category are considered individually in this discussion. Immune Thrombocytopenic Purpura. ITP often follows viral infections. Antibodies that are speciÀc for or that crossreact with platelet surface antigens coat the cell surface and promote greater

PART I Understanding, Controlling, and Preventing Infectious Diseases

Passive Immunization

clearance of platelets by the spleen. In patients with ITP, IGIV can affect a rapid increase in platelet counts (often within 24 hours of infusion) and thereby can avert life-threatening bleeding. The prevailing hypothesis is that IgG in IGIV interacts with phagocytic Fc receptors, blocking clearance of platelets and allowing them to enter or remain in the circulation.13 However, the mechanism may be more complex, involving induction of receptor expression.14 Many different IGIV products have been shown to be effective for treatment of ITP, although few direct comparisons of effectiveness have been made. Compounding the lack of data is the fact that no laboratory tests can predict efficacy in ITP, either of individual products or in individual patients. IGIV has been studied in acute and chronic ITP in both adults and children. Not all cases respond; however, in general, children show response more frequently than do adults, and the increase in platelet count is more prolonged in children treated for acute ITP. (A B-lymphocyte-suppressive effect may account for prolonged decrease in antiplatelet antibodies.) Because acute ITP can resolve spontaneously, especially in young children, it is important to obtain the consultation of an experienced hematologist during the planning of the therapeutic approach.15 Initially, the dosage of IGIV employed for treatment of ITP was 400 mg/kg daily for 5 consecutive days.13 This regimen is still in use, but monitoring of the platelet count may indicate that dosage can be stopped after the second, third, or fourth day. On the other hand, some IGIV products have been shown to raise the platelet count adequately when a single dose of 1 g/kg is infused. If the rise is insufficient, this dose can be repeated the next day or on alternate days, depending on the product used. In all cases, it is important: (1) not to exceed the rate of administration recommended in the package insert for the particular IGIV product; and (2) to avoid the higher dose in patients with expanded fluid volume. Kawasaki Disease. Immune dysregulation may play a role in diseases for which a specific etiology has not been identified. Kawasaki disease is an acute inflammatory condition with multisystem vasculitis that can result in arterial aneurysms, particularly of the coronary vessels. Several IGIV preparations have been studied for the treatment of this condition. IGIV given concurrently with oral aspirin rapidly downregulates the inflammatory response. Its use is associated with a lower incidence of coronary artery aneurysms.16,17 Treatment with IGIV is effective when begun early (preferably within 10 days); a single dose of 2 g/kg infused over 10 to 12 hours zis recommended (http://aappolicy.aappublications.org/).16 The initial dosage of aspirin is usually 80 to 100 mg/kg daily, divided among 4 doses, although lower dosage may be equally effective.18,19 When fever has subsided and remains under control, 3 to 5 mg/kg given as a single daily dose is adequate. Children who have persistent or recrudescent symptoms can be treated with a second IGIV dose of 2 g/kg. Inasmuch as the etiology of Kawasaki disease is unknown, it is possible that not all IGIV preparations (or all lots) are equally effective, even among products that carry this indication. Moreover, some clinicians recommend oral prednisolone for retreatment when the response to IGIV is inadequate.20 (See Chapter 199, Kawasaki Disease, for specific recommendations.) Bone Marrow Transplantation. As noted previously, IGIV can reduce the risk of infection in recipients of bone marrow transplants. In addition, several studies have reported IVIG to lower the incidence of acute graft-versus-host disease, but in one of the studies this result was only observed in patients who were older than 20 years.21,22

Other Uses The number of clinical conditions for which IGIV has been tried greatly exceeds that of its approved indications. The reasons for this are many. The simplest is the obvious fact that clinical research necessarily precedes the compilation of data by the manufacturer, submission of the data to regulatory authorities, and review and approval by the latter. Thus, in principle, a manufacturer might simply choose not to request approval of a particular indication if it offers no

CHAPTER

6

45

market advantage. Other reasons are more complex. A well-received, concisely written paper published in the medical literature might present positive results but might not include all of the less common adverse effects or might not address statistical nuances or subtleties of trial design that raise concerns. Alternatively, widespread use (indeed, commonly accepted treatment) for a particular condition can be generated by the enthusiastic endorsement of a respected clinician. In practice, such use may or may not be confirmed by appropriately controlled and adequately powered studies. The growing off-label use of IGIV has been implicated in shortages of IGIV (see below). This section includes some of the better-known off-label uses of IGIV, but no attempt has been made to include all or even most of them. IGIV has been used an adjuvant for the treatment of toxic shock syndrome associated with staphylococci and invasive streptococcal disease. IGIV typically contains antibodies against superantigen toxins produced by both these organisms. Furthermore, the antiinflammatory properties of IGIV may help ameliorate the exaggerated cytokine response induced by superantigens. Both animal studies and case reports suggest the utility of IGIV as adjuvant therapy for the treatment of toxic shock syndrome secondary to staphylococcal toxin and streptococcal infection.23,24 Typical dose of IGIV for this purpose is 1 g/kg, though a range of doses has been used. Because of the high morbidity and mortality rates in preterm infants, the effects of IGIV in this population have been studied extensively.25 Despite enthusiasm and the apparent logic of replacing missing antibodies, many studies have shown minimal benefit.25–27 One prospective, randomized trial was conducted with preterm (gestational age < 33 weeks) infants whose IgG levels at birth were 400 mg/dL. Those who received IGIV had no fewer infectious episodes and no lower mortality rate than those who received an albumin placebo.28 Parvovirus B19 infection can be prolonged in patients with a variety of immunodeficiencies, including: primary immunodeficiencies, HIV infection, sickle-cell disease, organ transplantation, and cytotoxic therapies. In such patients, parvovirus B19 causes chronic, severe anemia. IGIV (400 mg/kg daily for 5 days) has been reported to reduce the viral load and to restore erythropoiesis in selected patients.29,30 Anemia, hydrops, and persistent neonatal infection can also occur with intrauterine infection. In conjunction with intrauterine transfusions of red blood cells, IGIV may be helpful in ameliorating anemia associated with congenital parvovirus B19 infection.31 Patients with a variety of immune-mediated cytopenias, including anemia and neutropenia, have been treated with IGIV, apparently with some success.32 Published reports support the use of IGIV in HIVinfected individuals with ITP, although none of the licensed IGIV products carries this specific indication.33,34 Guillain–Barré syndrome, an acute immune-mediated inflammatory polyneuropathy, may respond to IGIV therapy. Plasma exchange was first shown to decrease morbidity and improve outcome of affected patients.35,36 High-dose IGIV (1 g/kg on 2 consecutive days) is apparently as effective as plasma exchange and has been reported to have fewer complications.37 The use of IGIV for management of acute disseminated encephalomyelitis is not established.

SPECIFIC IMMUNE GLOBULINS FOR INTRAVENOUS ADMINISTRATION Currently four specific, plasma-derived immunoglobulin products for IV administration are approved for prophylaxis or therapy of infectious disease (see Table 6-1).

Cytomegalovirus Immune Globulin Cytomegalovirus (CMV) IGIV (human) is indicated for the prophylaxis of CMV disease associated with organ transplantation.38–40 Dosage is initiated at a level of 150 mg/kg within 72 hours of organ

46

SECTION

B Prevention of Infectious Diseases

transplantation and then tapered over 6 additional biweekly doses to either 50 mg/kg (for kidney transplants) or 100 mg/kg (for liver, lung, pancreas, or heart transplants). The use of CMV-IGIV for the prophylaxis of CMV disease varies among transplant centers. Factors that influence the susceptibility of transplant recipients to CMV disease include: organ type, immunosuppressive regimen, and donor/recipient CMV status. CMV-negative transplant recipients who receive an organ from a CMV-positive donor are at highest risk for CMV disease and typically receive some form of CMV prophylaxis. This can include ganciclovir (using a prophylactic or pre-emptive strategy) alone, CMV-IGIV alone, or CMV-IGIV in combination with ganciclovir therapy.

Respiratory Syncytial Virus Immune Globulin Respiratory syncytial virus (RSV) IGIV (human), shown to be beneficial in reducing RSV infections in high-risk infants,41 is no longer available. Palivizumab, a humanized (95% human, 5% murine) monoclonal antibody to the A epitope of the F protein of RSV, has replaced RSV-IGIV. In one study, monthly doses of 15 mg/kg decreased the number of hospital admissions as well as the rate of admission to intensive care, both in infants with chronic lung disease (bronchopulmonary dysplasia) and in premature infants without this condition.42 Palivizumab is given IM and does not interfere with immunization by live-virus vaccines.43 Palivizumab can be given to neonates with hemodynamically significant congenital heart disease (RSV-IGIV could not be).44 Because of the high cost of palivizumab, guidelines for use have been recommended (see Chapter 225, Respiratory Syncytial Virus).45

Botulism Immune Globulin Botulism IGIV (human) (BIG-IV) was approved by the FDA in 2003 under the orphan drug program. It is indicated in infants less than 1 year of age for the treatment of botulism suspected to be caused by either toxin type A or B. It is available under the trade name BabyBIG, only through the California Department of Health Services (510-2317600, all days/hours). BIG-IV has been shown to reduce the hospitalization time of affected infants significantly if used within 3 days of admission.46, 47 The recommended dose is 50 mg/kg. A trivalent equine antitoxin is indicated for use in foodborne botulism (see Chapter 189, Clostridium botulinum (Botulism)).

Vaccinia Immune Globulin Vaccinia IG (human) (VIG) was developed in the 1960s for the purpose of ameliorating side effects associated with vaccinia immunization, including eczema vaccinatum, generalized, and progressive vaccinia.48 The original preparation contained a high proportion of protein aggregates and thus was administered IM. The use of VIG has become extremely limited since the eradication of smallpox. It is considered valuable “insurance,” to be held in reserve if a patient is receiving an experimental vaccine that involves a vaccinia carrier virus, or to prevent or manage complications of smallpox vaccination should such be required for a bioterrorism threat. Recently, a preparation of VIG suitable for IV use (VIG-IGIV) has gained FDA approval. In the event that VIG is required, it can be obtained through the Centers for Disease Control and Prevention (CDC: www.bt.cdc.gov/agent/smallpox/).

IMMUNOGLOBULIN PRODUCTS PREPARED FROM ANIMAL PLASMA In the past, there existed an array of approved animal-derived immunoglobulin products specific for various infectious agents or

their toxins. Few are in use currently, most having been supplanted by analogues of human origin (e.g., RIG, TIG) or by other therapeutic modalities. (There are approved animal-derived products for the treatment of venomous bites and digoxin intoxication and for prevention of organ rejection; these are beyond the scope of this chapter.) The only available products in this category are botulism antitoxin (see Chapter 189, Clostridium botulinum (Botulism)) and diphtheria antitoxin (see Chapter 130, Corynebacterium diphtheriae (Diphtheria)). They can be obtained from the CDC for emergency use in the treatment of foodborne botulism and diphtheria, respectively. These equine immunoglobulin products are only provided when specifically indicated. Because immediate anaphylaxis and delayed serum sickness are possible adverse reactions, these products should only be used when their life-saving potential clearly outweighs the risk. Procedures for skin testing and desensitization are readily available and the physician should always be prepared to treat any adverse event.49

ADVERSE REACTIONS TO IMMUNE GLOBULINS PREPARED FROM HUMAN PLASMA The most common adverse reaction to IM-administered IGs is pain at the injection site. Local or facial flushing can occur, as can headache, chills, or nausea, but these symptoms are less common. Severe systemic reactions are rare.

Infusion-Related Reactions Adverse events after IV administration of IGs are common (Box 6-1). Severe reactions to IGIV occur infrequently, but mild side effects have been associated with up to 20% of infusions.50 The “Adverse Reactions” section of the package insert for an individual product provides the rates of adverse reactions obtained during studies of individual products. However, in view of the fact that few comparative trials of different products have been conducted, these numbers cannot be compared directly and should only be used for general guidance. The bases for these side effects are poorly understood. As knowledge of the products has grown, it has become possible to minimize some sources of adverse reactions. For example, although the precise mechanisms remain controversial, the association between adverse reactions and aggregated forms of IgG was repeatedly observed. Accordingly, manufacturers sought procedures to eliminate such aggregates, minimize their formation, or both. Similarly, intermediates of the contact activation system, such as prekallikrein activator, could elicit a variety of reactions but largely could be avoided by suitable manufacturing strategies. Nonetheless, even with these advances, the classic constellation of side effects continues to be observed.

BOX 6-1. Adverse Effects of Immune Globulin Intravenous Therapy MINOR SYSTEMIC REACTIONS Headache, back or hip pain, fever, dizziness or lightheadedness, nausea, flushing PYROGENIC REACTIONS High fever and systemic symptoms VASOMOTOR/CARDIOVASCULAR MANIFESTATIONS Changes in blood pressure and heart rate INFREQUENT, SERIOUS REACTIONS Aseptic meningitisa Acute renal failure Hypersensitivity reactionsb Anaphylaxis a Aseptic meningitis may be more common in adults or children with specific underlying conditions. b Reaction reports are more common for products that contain sucrose.

PART I Understanding, Controlling, and Preventing Infectious Diseases

Passive Immunization

Apparently, patients with certain conditions (e.g., primary immunodeficiency) are more vulnerable to the side effects of IGIV. Furthermore, among patients who receive IGIV repeatedly, some respond adversely more frequently than do others. The incidence of adverse events is particularly affected by the rate of infusion. For this reason, the instructions in the “Dosage and Administration” sections of package inserts for individual products should be carefully followed. These generally provide a starting rate (usually as flow rate per kilogram of body mass), a schedule for increasing the rate, and a maximum rate. Experience with the individual patient and product is of paramount importance. Most reactions subside when the rate of infusion is decreased. However, if reactions occur repeatedly, corticosteroid given IV, 30 min prior to the infusion, may alleviate symptoms. Some clinicians prefer to use oral antihistamine or analgesic medications. Severe, migrainelike headaches and aseptic meningitis have been reported in children.51,52 In one clinical trial in which adults received high-dose IGIV (2 g/kg), aseptic meningitis developed in 4 of 8 patients with a history of migraine headaches.53 By contrast, the incidence in patients without such a history was 4% (2 of 46). Several large controlled trials in children with Kawasaki disease and preterm neonates have not documented IGIV-induced aseptic meningitis. However, this condition has been observed in children as well as adults being treated for ITP.54,55 Symptoms of meningitis (fever, headache, photophobia, and nuchal rigidity) can appear during the infusion or up to 48 hours later; cerebrospinal fluid pleocytosis in case reports ranges from < 100 to > 1000 neutrophils/mm3. Although the mechanism is not fully understood, cerebrospinal fluid levels of protein, including IgG, are often elevated.53,56 Severe hypersensitivity reactions and anaphylaxis are rare; if they occur, IgA deficiency should be considered.57 For individuals with IgA deficiency and hypersensitivity reactions, IGIV with an extremely low IgA content is available (Baxter Healthcare Corporation). Routine screening for IgA deficiency is not recommended. Other rare events reported to be associated with IGIV infusion are Coombs-positive hemolytic anemia and accelerated serum sickness (upon repeated exposure) as well as thromboembolic events.50,58 Two relatively rare adverse reactions to IGIV, that is, acute renal failure and renal insufficiency, deserve special attention. The mechanism underlying these reactions is unclear, but strong associations with both the nature of products and conditions of the patient have been observed.50,59 Patients 65 years or older, patients receiving concomitant treatment with nephrotoxic agents, and patients with diabetes mellitus, pre-existing renal disease, hypovolemia, or sepsis appear particularly vulnerable.59 Most, but not all, reports of adverse renal events have involved IGIV products that contain sucrose. Patients with renal impairment (clinical or subclinical) should be observed during and after IGIV infusion for exacerbation of the condition. Adequate hydration (but not overhydration) should be maintained, and the rates of administration recommended in the package insert for the individual product should be carefully followed.

Interference with Active Immunization Antibodies present in IGs can interfere with the response to the corresponding vaccines. This is rarely a problem when the vaccine is a substance such as a toxoid, a polysaccharide, or a killed virus preparation, even when it is given simultaneously with the IG, provided that the two products are administered with separate syringes at different sites. However, these antibodies can interfere with replication after administration of live-virus vaccines and, thus, prevent successful immunization. The effect on measles immunization has been studied in detail, but in principle, similar interference could occur with any live-virus vaccine. Measles, mumps, and rubella vaccine should not be given in the 2 weeks before or 3 months after an individual receives any IgG preparation. Because high doses of IGIV can inhibit the response to measles vaccine for extended periods, longer intervals are required as the dose is increased (up to 11 months after 2 g/kg

CHAPTER

6

47

IGIV as for Kawasaki disease) to provide sufficient time for passively acquired antibodies to dissipate.1,18

Transmission of Infectious Agents Although IGs prepared from human plasma are biologic products and, hence, must always be considered to carry a theoretical risk for transmission of viruses, they have a remarkable record of safety. Even IGIV, which is administered directly into the bloodstream, often in large quantities, has been extremely safe. The many investigations conducted have revealed no documented case of transmission of hepatitis by United States-licensed IGs for IM use. Similarly, there is no evidence of transmission of HIV by United States-licensed IGs for either IM or IV administration.60 Several cases of non-A, non-B hepatitis were associated with IGIV administration between 1983 and 1987.61,62 These cases involved sources of IGIV that were not commercial lots available in the United States. In February 1994, however, a worldwide recall of United States-licensed IGIV products manufactured by Baxter Healthcare Corporation was initiated because of hepatitis C virus (HCV) transmissions.63 Cases of hepatitis C occurred in several countries, including the United States, among individuals who received these products between April 1, 1993, and February 23, 1994 (when they were removed from the market). The reason for these transmissions of hepatitis illustrates the complex relationships that underlie the production of IGs and all human plasma derivatives. These products had initially been licensed in February 1986 and, like other United States-licensed IGIV products, had an impeccable safety record. In 1990, the first test for antibody to HCV – a single-antigen anti-HCV test – became commercially available. In 1992, a more sensitive test (a multiantigen anti-HCV test) came into use. Although using these tests to screen donors of blood and blood components (e.g., red blood cells, platelets) was of obvious benefit, application of the more sensitive test for screening of donors of plasma for fractionation had a paradoxical effect; that is, although withholding units of plasma that gave a positive test result meant that fewer infectious donors were represented in the plasma pools, these pools were also depleted with respect to beneficial anti-HCV antibodies that had cocirculated with the antibodies detected by the test. These beneficial antibodies had served two functions: (1) neutralization of viruses; and (2) complexing with viruses so as to partition them away from the protein fraction that became the final product. As a consequence, many lots of IGIV manufactured by Baxter Healthcare Corporation from plasma collected from donors who had been tested by the more sensitive test contained infectious (i.e., nonneutralized, noncomplexed) virus.64,65 By contrast, IGIV made by other manufacturers from plasma collected from donors who had undergone the same type of testing did not transmit hepatitis C.63–65 This finding showed that the margin of safety had differed among individual IGIV products. Concerns about the potential transmission of prion disease by IG products have been raised. These concerns are highlighted by the lack of serologic assays to screen blood products for prions. The findings of several studies suggest that the likelihood of prion transmission via plasma products is exceedingly small. This includes the observations that very low levels of prions are found in the plasma of affected individuals and that the processing of plasma for the production of IGs typically inactivates prions.66–69 To date, no case of prion disease has been linked to IG therapy. Studies performed on a laboratory scale have shown that many manufacturing steps that are intrinsic to the process for preparing IgG greatly lower viral and prion burden. For example, alcohol fractionation of plasma partitions virus and prions away from the IgG fraction.67,70 If the alcohol concentration at a particular manufacturing step is sufficiently high and the virus is sufficiently labile (e.g., HIV), alcohol is viricidal.70 Other precipitating agents, such as polyethylene glycol, can achieve similar partitioning of viruses. Incubation at low pH, either in the presence or in the absence of enzymes, has a strong antiviral effect.71,72

48

SECTION

B Prevention of Infectious Diseases

Solvent-detergent treatment is one step that has been deliberately introduced to achieve viral inactivation in various IGs. This treatment destroys enveloped viruses such as HBV, HCV, and HIV.73 It is used in the manufacturing of numerous IGs for IM and IV use (including current IGIV products made by Baxter Healthcare Corporation). Another deliberate viral inactivation used for some products is heat treatment, which can be performed in the presence of stabilizers to prevent denaturation of the IgG.74 The process for making some IG products now includes so-called nanofiltration, that is, filtration of the product through membranes with pore dimensions that permit the passage of proteins but not viruses. Brief summaries of these procedures as well as of the experiments demonstrating inactivation and removal of specific viruses are generally provided in the “Description” sections of the package inserts for the individual products.

ACKNOWLEDGMENT

The author thanks John S. Finlayson for significant use of material from the second edition of Principles and Practices of Pediatric Infectious Diseases.

CHAPTER

7

Active Immunization

PRODUCT SHORTAGES There have been periodic shortages of certain IG products. The reasons for this are multiple. In some cases, there have been specific production problems. In others, demand is a major factor. For example, use of IGIV in the United States has steadily increased by approximately 10% each year, and production has not always kept pace. Off-label use of IGIV contributes importantly to shortage. A study of 12 academic health centers revealed that 52% of IGIV prescribed was used for off-label indications.75 In response to these shortages, some hospitals have developed committees to review and restrict usage of IGIV. It is incumbent on clinicians to employ IGs appropriately so that products will be available for use in conditions for which they have proven benefit.

THE FUTURE In the immediate future, we may anticipate additional IG products for IV administration. Some of these are likely to be specific IGs directed against known infectious agents. In the future, additional specific monoclonal antibody (mAb) products also may become available. mAb technology enables the production of large amounts of purified antibody with defined specificity and isotype that may not be present in standard IG preparations. This approach makes it possible rapidly to provide patients with a defined amount of protective antibody. Furthermore, advances in molecular biology have made possible the humanization of murine mAb, whereby the murine sequences are limited to the variable region of the antibody. This process leads to decreased antigenicity, improved pharmacokinetics, and presumably produces fewer side effects. In the last 10 years, a number of mAbs have been licensed for use, including those for: (1) organ transplantation (via inhibition of cytokine activity or depletion of lymphocyte subsets); (2) treatment of malignancies (i.e., lymphoma, colon and breast cancer); and (3) immunologic modulation (i.e., asthma and Crohn disease). Early attempts to utilize mAb against endotoxin in the treatment of gram-negative bacillary septicemia were not successful.76,77 A variety of mAbs for the treatment and prevention of infectious diseases are in clinical development related to the following pathogens: Cryptococcus neoformans, Staphylococcus aureus, HIV, RSV and Candida albicans. The advantages provided by mAbs have led some to hope that these reagents will be helpful in dealing with the problems of drug-resistant pathogens and potential bioterrorism attacks.78 In parallel with continuing product development, it is reasonable to expect improvements in the design and conduct of clinical trials. One example might be streamlining the design of trials involving wellknown products (e.g., IGIV) used to treat well-understood conditions such as primary immunodeficiency. Another, one hopes, is the conduct of trials that are sufficiently powered to permit major “off-label” uses to be either approved or abandoned.

Larry K. Pickering and Walter O. Orenstein

Vaccines are among the most effective means of preventing disease, disability, and death. The use of vaccines, initiated by Jenner in 1796 with the demonstration that inoculation of material from cowpox lesions could prevent smallpox, predates the germ theory of disease. Use of conventional viral and bacterial culture techniques led to development of vaccines to prevent 7 diseases in 1985. Subsequently, advances in understanding the immunologic basis of immunity and new molecular biologic techniques, including genome sequencing, have facilitated the definition of the precise composition and structure of antigens, the development and extensive use of new vaccines, and an expansion of their potential uses beyond prevention of childhood diseases. The profound impact of vaccines on disease incidence is a result not only of availability of safe and effective vaccines but also of strategies for disease control and programs to deliver vaccines to target groups. Eradication of smallpox in 1977 is among the greatest public health achievements of its age and was based on competent disease surveillance and containment of spread of disease with a highly effective vaccine.1 Global efforts under way to eradicate poliomyelitis and to certify eradication of polio involve more comprehensive disease control strategies. National and multicountry mass vaccination campaigns, intensive surveillance for wild poliovirus, and houseto-house vaccination programs have succeeded in terminating transmission of wild poliovirus in the Americas, the Western Pacific, and Europe, and gains are being made in the remaining endemic countries in the Indian subcontinent and Africa.2–4 Efforts to reduce or even eliminate measles and neonatal tetanus, two of the major causes of child fatality worldwide, also are progressing.5 The Global Alliance for Vaccines and Immunization, an alliance between organizations in the public and private sectors, provides support for introduction of vaccines in countries that previously could not support them (www.gavialliance.org). Benefits of successful vaccines include not only reductions in disease incidence, disability, pain, and suffering, but also savings in healthcare costs in both the economically developed and developing world. Studies in the United States have reaffirmed benefits of childhood immunization,6–8 indicating that: (1) each of the traditional vaccines – diphtheria and tetanus toxoids, and acellular pertussis (DTaP), measles, mumps, and rubella (MMR), and oral polio vaccine (OPV) – is cost-saving in terms of direct medical costs alone. A cost–benefit analysis covering 10 vaccine-preventable diseases of childhood, including diphtheria, tetanus, pertussis, polio, hepatitis B, Haemophilus influenzae type b (Hib), measles, mumps, rubella, and varicella (MMRV), estimated that for every dollar spent on childhood immunization in 2001, there were $5.3 dollars saved in direct costs and $16.5 saved by society.8

PART I Understanding, Controlling, and Preventing Infectious Diseases

Active Immunization

IMMUNIZATION AND VACCINES Immunization is the process of artificially inducing immunity or providing protection from disease. Active immunization is the process of stimulating the body to produce antibody and other immune responses (e.g., cell-mediated immunity) through administration of a vaccine or toxoid. Passive immunization is provision of temporary immunity by administration of preformed antibodies derived from humans or animals (see Chapter 6, Passive Immunization).

Vaccine Content Biologic agents used to induce active immunization include vaccines and toxoids. Traditionally, a vaccine is defined as a suspension of live (usually attenuated) or inactivated microorganisms, or fractions thereof, which is administered to induce immunity and prevent infectious disease or its sequelae; efforts to develop vaccines to increase immune response to cancers or to treat diseases like diabetes mellitus necessitate rethinking of this definition. Live-attenuated vaccines traditionally have been developed by means of serial passage (in culture or animals) of an initially pathogenic bacteria or virus strain with selection for strains that are less pathogenic for humans but that induce protective immunity (MMRV). Live-attenuated vaccines can also be developed with the use of reassortants of attenuated animal or human virus strains with virus coat antigens from pathogenic strains (cold-adapted influenza, rotavirus). Inactivated vaccines can consist of: (1) whole organisms inactivated by heat, formalin, or other agents (polio, hepatitis A, rabies); (2) purified protein (acellular pertussis and influenza) or polysaccharide antigens (pneumococcal, meningococcal, and im typhoid) from whole organisms; and (3) purified antigens produced by genetically altered organisms (hepatitis B and human papillomavirus (HPV) vaccines produced by yeast); or (4) chemically modified antigens, such as polysaccharides conjugated to carrier proteins to increase immune response (conjugated Hib, pneumococcal and meningococcal vaccines). Toxoids are modified bacterial toxins produced in bacterial culture. These toxoids have been rendered nontoxic but retain the ability to stimulate formation of antitoxin. Vaccine and toxoid preparations also contain other constituents that are intended to enhance immunogenicity and stability but that can also be responsible for adverse reactions.9 Such constituents include: (1) suspending fluid, which can be saline or complex fluids containing constituents derived from the biologic system or medium in which the vaccine is produced (e.g., egg or serum proteins); (2) preservatives, stabilizers, and antimicrobial agents, which are used to inhibit bacterial growth in viral cultures or the final product or to stabilize antigens (e.g., mercurials, phenols, albumin, glycine, neomycin); and (3) adjuvants, which enhance response to inactivated antigens (aluminum hydroxide or phosphate). Concern about the possibility of adverse effects from cumulative exposure to mercury in the environment has led to the removal of thimerosal as a preservative from United States vaccines recommended for infants.10 Only some formulations of influenza vaccine that can be administered to infants contain thimerosal as a preservative. Physicians should be knowledgeable about the constituents of each vaccine, which are described in package inserts.

TYPES OF VACCINES Immunologic Basis of Response to Vaccines The two major approaches to active immunization are use of: (1) liveattenuated vaccines and (2) inactivated or detoxified agents or their purified components. For some diseases, such as poliomyelitis and influenza, both approaches have been used to develop vaccines. Live-attenuated vaccines have the advantage of producing a complex immunologic response simulating natural infection. Because replication of the organism and processing of antigens mimic those of

CHAPTER

7

49

the natural organism, both humoral and cell-mediated responses can be generated to a variety of antigens. Generally, immunity induced by one dose of a live-attenuated vaccine is long-lasting, possibly lifelong. However, the strength of response, particularly the humoral response, is usually less than the response following natural infection, and detectable antibodies can wane with time, resulting in some loss of protection. Induction of immunity by live vaccines can be inhibited by passive antibody, whether from transplacental acquisition from the mother or receipt of immunoglobulin-containing blood products; thus, optimal response depends on ensuring that this level of passive antibody has declined (e.g., primary measles vaccination at 12 months of age instead of earlier, delay of measles vaccination after administration of blood products). In addition, because response may be only 90% to 95% after a single dose, a two-dose or multiple-dose regimen may be necessary to induce higher levels of protection in the community and prevent spread of disease if the population is exposed (herd immunity). Inactivated or purified antigen vaccines induce response only to components present in the vaccine. Generally, multiple doses, usually three or more, are necessary to induce satisfactory antibody levels that persist for long periods of time; booster doses at longer intervals (e.g., 10 or more years for tetanus and diphtheria toxoids (Td)) are sometimes needed to ensure lasting protection. The nature of response depends on antigen type. Protein (and glycoprotein) antigens usually induce both humoral immunity and memory (T-helper lymphocytes) after multiple doses, evidenced as more rapid and intense response to successive doses. Polysaccharide antigens by themselves induce only humoral antibody without T-lymphocyte stimulation and fail to induce anamnestic response with repeated antigenic challenge. This shortcoming can be overcome by conjugation of polysaccharides to protein carriers (e.g., Hib polysaccharide conjugated to whole or modified diphtheria or tetanus toxoids or to outer-membrane protein (OMP) complex of Neisseria meningitidis) to induce stronger immune response in younger children as well as immunologic memory.

Immune Response to Active Immunization Development of an immune response generally requires interaction of T lymphocytes with antigen-processing and antigen-presenting cells (dendritic cells or macrophages).11,12 Certain types of antigens (thymus-independent, e.g., polysaccharide) can initiate B-lymphocyte antibody production without help of T lymphocytes but fail to induce immunologic memory. T-lymphocyte immunity is induced after uptake of antigen by mononuclear phagocytes or dendritic cells, which can be enhanced by use of an adjuvant, followed by processing and presentation of the antigen, in association with major histocompatibility complex (MHC) antigens, to helper T lymphocytes. T lymphocytes recognize polypeptide antigens of approximately 8 amino acids in size, presented in association with specific MHC molecules; the type of MHC molecule with which the antigen is presented by antigen-processing cells depends on the source and processing of the polypeptide. Inactivated antigens, absorbed into vacuoles, are processed and presented with MHC2 antigens; antigens from live-attenuated vaccines or vectored vaccines, produced within the cell, are processed in microtubules and presented with MHC1 antigens. These antigen–MHC complexes determine the primary type of T-lymphocyte response, either cytotoxic or helper. Presentation to helper T lymphocytes results in secretion of immune mediators (cytokines) that can stimulate the maturation of naive T lymphocytes and communicate between leukocytes (interleukins) to regulate the immune response. Depending on the antigen and its MHC presentation, T lymphocytes differentiate into either type 1 T-helper lymphocytes (Th1, stimulated by MHC1-associated antigen), which mediate cellular immune response, or type 2 T-helper lymphocytes (Th2, stimulated by MHC2-associated antigen), which assist B lymphocytes in developing antibody production.13 Each of these subsets produces different interleukins and other immune mediators responsible for modulating the immune response.

50

SECTION

B Prevention of Infectious Diseases

Antibodies produced in response to immunization can function in any of several ways: (1) direct neutralization of toxin; (2) opsonization (neutralization) of organisms; (3) initiation of or combination with the complement pathway to lyse organisms or promote phagocytosis (pneumococcus); (4) reaction with nonsensitized lymphocytes to promote phagocytosis; or (5) sensitization of macrophages to promote phagocytosis. These mechanisms can occur simultaneously with cellmediated responses. Rarely, vaccination can result in immune responses that alter the course of natural infection detrimentally. For example, killed measles vaccine, a formalin-inactivated vaccine administered to some children in the United States between 1963 and 1967, sensitized some vaccine recipients so that when they were exposed to wild virus they developed an atypical infection with enhanced severity of disease. The prevailing theory has been failure of formalin-inactivated vaccine to produce response to the measles virus fusion protein, leading to altered immune response and atypical disease upon subsequent challenge. More recent theories suggest killed measles virus failed to induce high-avidity antibody, and immune complex deposition is the major cause of atypical measles.14 After immunization with inactivated antigens, antibody response to initial doses develops in 2 to 6 weeks but may be incomplete after even two doses; after effective priming, booster responses occur within 4 to 14 days. The initial response is usually immunoglobulin M (IgM) antibodies, followed within weeks by IgG antibodies. Response to live vaccines requires one incubation period, followed by several weeks to months for development of a strong immune response. Response to measles vaccination is usually maximal by 6 weeks, but in younger children, antibody levels can continue to rise for several months.

Determinants of Response Determinants of vaccine immunogenicity and response involve characteristics of the vaccine and of the host. Vaccine dose, presence of an adjuvant, route and site of administration, timing of doses, and vaccine handling can affect response. Vaccine doses are adjusted before licensure to ensure a high level of response (generally > 90%); adjuvants permit a better response with a lower dose of inactivated antigen. The routes of administration (e.g., intradermal, subcutaneous, intramuscular, and mucosal) can determine the strength and nature of the immune response. Mucosal administration (intranasal or oral) stimulates higher levels of mucosal immunity (IgA antibodies) that may inhibit disease transmission with greater effectiveness than parenteral administration, which induces limited or no mucosal response.15 Intradermal vaccination with low doses can induce antibody responses similar to responses induced by intramuscular or subcutaneous administration of recommended doses, but intradermal vaccine is more difficult to deliver precisely and, in practice, achieves less predictable responses.

Sites of Administration Intramuscular injections should be given in the anterior thigh (infants) or deltoid (toddlers, children, and adults); injection into the buttocks may produce lower antibody response, which has been documented for hepatitis B and rabies vaccines in adults, probably owing to delivery of vaccine into adipose tissue.16,17 Vaccines with adjuvants should be given in deep muscle, because subcutaneous or intradermal injection can induce local inflammation, granuloma formation, or necrosis.18 Timing of doses of killed vaccines is important; a minimal interval of 1 month between primary doses is usual.7 Delay of a third or reinforcing dose for 6 months or longer after the first dose enhances response and duration of antibody persistence and is recommended unless high disease risk necessitates shorter intervals. The recommended routes and sites of administration and timing of doses are devised to ensure optimal effectiveness in disease prevention and should be used.

Host Factors Intrinsic factors in the host that affect immune response include genetic factors,19,20 age, nutritional or disease status, primary or secondary immunodeficiency, gender, pregnancy, and smoking. Although genetic factors such as MHC polymorphism are known to affect both cellular and humoral immune response at a molecular level for some vaccines, the precise mechanisms for these genetic influences are often unknown.19–21 Age is an important factor in response to immunization. With killed vaccines, neonates generally do not develop as strong a response as older infants or children (i.e., hepatitis B), and with certain vaccines, too early immunization may result in poor response or development of tolerance (DTaP, inactivated poliovirus vaccine (IPV), Hib conjugates). For live (and some killed) vaccines, inhibition of response by maternal antibodies determines the optimal timing for vaccination in early childhood (measles, hepatitis A). Generally, response to all vaccines is excellent in young children, adolescents, and young adults but diminishes with increasing age. In adults, male sex and pregnancy have minor dampening effects on antibody response that have limited significance; smoking decreases response to many antigens and may raise the risk of nonresponse to vaccination when other negative factors are present.16,18 Extreme debilitation, primary or secondary immunodeficiency disorders, including diseases or treatments that cause immunosuppression, and some chronic diseases (renal disease, diabetes mellitus) can diminish immune response. For people with such conditions, inactivated vaccines may be recommended despite their lower effectiveness, although higher or more frequent doses may be required; live vaccines are often contraindicated because of the risk of disseminated disease and possible death due to the vaccine organism.18,22

Measurement of Response Ideally, reliable laboratory tests should be available to measure the presence and strength of each of the major effectors of protection against the disease for which the vaccination is administered. In practice, a wide variety of tests for presence of antibody are available, and include radioimmunoassay (RIA), enzyme immunoassay (EIA), complement fixation, polymerase chain reaction, and immunofluorescent techniques, but these tests often do not measure the presence of functional (neutralizing or opsonizing) antibody. Tests for cell-mediated immunity are available in reference laboratories and research facilities; different tests are required to determine cytotoxic and helper T-lymphocyte (memory) functions. For certain diseases, such as hepatitis B, poliovirus, and measles, reliable tests exist and antibody levels that correlate with protection are known; however, inexpensive tests are only widely available for hepatitis B. For other diseases, such as rubella, commercial tests are available, often using EIA methods, but their specificity is often less well defined, and their sensitivity is lower than sensitivity of neutralization assays. For some diseases, such as pertussis, no serologic correlate of protection has been defined. Development of better laboratory methods to measure protection and to permit rapid diagnosis of acute disease continues to be a priority of vaccine-preventable disease control programs.

Vaccine Licensure and Approval Before Food and Drug Administration (FDA) licensure, a new vaccine generally requires 10 to 15 years of preclinical testing and clinical trials. Prior to testing a vaccine in humans, a manufacturer files an investigational new drug (IND) application with the FDA, followed by three phases of clinical trials that are performed to study vaccine safety, immunogenicity, and efficacy.23,24 Following completion of the prelicensure clinical trials, the following steps are required: the manufacturer must apply for a Biologics Licensure Application (BLA) with the United States FDA; the FDA must license the vaccine;

PART I Understanding, Controlling, and Preventing Infectious Diseases

Active Immunization

Vaccine development

Advisory Committee on Immunization Practices (ACIP)

FDA licensure

CDC consideration

COID/AAP consideration

Recommendations for use

State laws

7

51

and public health groups and industry representatives participate in ACIP discussions. To formulate recommendations, the ACIP establishes subject-specific work groups to review and synthesize data months to years before presentation to the ACIP, vote, and a recommendation is released. ACIP recommendations are subject to the approval of the director of the Centers for Disease Control and Prevention (CDC) (www.cdc.gov/nip/ACIP/charter). The COID of the AAP also develops recommendations for vaccine use which are approved by the Board of Directors of the AAP. These recommendations are usually the same as, or similar to, recommendations of the ACIP. After vaccine licensure, monitoring for rare adverse events continues for some vaccines through formal phase IV trials conducted by the FDA and the manufacturer. In addition, postmarketing surveillance for adverse events permits detection of new or unanticipated adverse events; reporting of adverse events observed after vaccines is required by the National Childhood Vaccine Injury Act for vaccines covered under the Vaccine Injury Compensation Program.25 The importance of postmarketing surveillance was demonstrated following licensure and wide use of tetravalent rhesus rotavirus vaccine (RRV) in United States infants during 1999. Surveillance of adverse events detected cases of intussusception within 1 week after receipt of the first or second doses of RRV. Followup studies determined that risk of intussusception was 1 case per 10 000 doses of vaccine administered.26 Subsequently, the vaccine was withdrawn from distribution, and recommendation for universal use in infants in the United States was withdrawn.27

the Advisory Committee on Immunization Practices (ACIP), the American Academy of Family Physicians (AAFP), and the Committee on Infectious Diseases (COID) of the American Academy of Pediatrics (AAP) must recommend the vaccine for use; and financing for the vaccine must be secured for people in the public and private sectors (Figure 7-1). Following licensure of a new vaccine by the FDA, information about the vaccine is reviewed by the ACIP. The ACIP is comprised of 15 voting members appointed by the Secretary of the Department of Health and Human Services. In addition, many professional medical

Vaccines and Related Biological Products (VRBPAC)

CHAPTER

PRINCIPLES OF IMMUNIZATION PROGRAMS

Uptake and financing

Disease Reduction Public sector

Childhood immunization programs have reduced substantially the occurrence of vaccine-preventable diseases in the United States from the representative annual morbidity during the 20th century (Table 7-1).28–30 Declines exceed 95% for all diseases for which universal vaccination has been well implemented, with the exception of hepatitis B, which remains a common clinical disease in adults not reached by universal vaccine programs31 and pertussis.32 Smallpox has been eradicated, poliomyelitis due to indigenous wild poliovirus has not occurred since 1979, and endemic rubella has been declared

Private sector

Figure 7-1. Development of pediatric vaccine recommendations and policies. FDA, Food and Drug Administration; CDC, Centers for Disease Control and Prevention; COID, Committee on Infectious Diseases; AAP, American Academy of Pediatrics BoD, Board of Directors. (Redrawn from Pickering LK, Orenstein WA. Development of pediatric vaccine recommendations and policies. Semin Pediatr Infect Dis 2002;13:148–154, with permission.)

TABLE 7-1. Reported Morbidity of Selected Vaccine-Preventable Diseases and Vaccine Coverage Levels – United States, 20th Century and 2004 Disease

United States, 20th-Century Annual Morbiditya

United States, 2006 Morbidityb

Vaccine Coverage Levels, 2005 %c

Healthy People 2010 Coverage Level Goals

Diphtheria

175,885

0

86% (≥ 4 doses)

90%

Tetanus

1314

38

86% (≥ 4 doses)

90%

Pertussis

147,271

5826

86% (≥ 4 doses)

90%

d

Poliomyelitis (paralytic)

16,316

0

92% (≥ 3 doses)

90%

Measles

503,282

73

92% (≥ 1 dose)

90%

Mumps

152,209

6585

92% (≥ 1 dose)

90%

Congenital rubella

823

1

92% (≥ 1 dose)

90%

Haemophilus influenzae type b and unknown; < 5 years of age

20,000

242

94% (≥ 3 doses)

90%

Varicella

Unknown

26,683

88% (≥ 1 dose)

90%

a

MMWR 2004;53:687–696, number of reported cases. b MMWR 2007; 54:1–92, number of reported cases. c MMWR 2006; 55:988–93 – reference 30. d Vaccine-associated paralytic poliomyelitis. e Inactivated poliovirus vaccine.

52

SECTION

B Prevention of Infectious Diseases

eliminated in the United States.33,34 Fewer than 10 cases each of diphtheria and tetanus in children are now reported each year, and indigenous transmission of measles has been interrupted. Wide use of Hib conjugate vaccines has reduced Hib disease by more than 98% in the United States and in some European countries.35,36 Use of pneumococcal conjugate vaccine has led to marked reductions in invasive pneumococcal disease among vaccinated children as well as unvaccinated very young infants, adolescents, and adults.37–40 Building on the success of the program to date, the United States Public Health Service has established 2010 goals to eliminate indigenous transmission of measles, rubella and congenital rubella syndrome, mumps, diphtheria, poliomyelitis, Hib disease in children younger than 5 years of age, and tetanus in persons younger than 35 years of age, and to reduce hepatitis B in people 2 to 18 years of age by 90%.41

Immunization Coverage Immunization coverage among preschool children has increased steadily after the measles resurgence that began in 1989 and stimulated unprecedented efforts to improve delivery of immunization. In 2005, coverage with three doses of DTaP, Hib conjugate, polio, and hepatitis B vaccines, and one dose of MMR vaccine among children 19 to 35 months of age each reached or exceeded 92%, whereas coverage with one or more doses of varicella vaccine was 88%.42 Coverage with the recommended series of four DTaP, three OPV, one MMR, three Hib, and three hepatitis B vaccines was 81%. Among school-aged children as well as attendees of childcare centers and Head Start programs, coverage with recommended vaccines has been more than 95% since the early 1980s, as a result of enforcement of comprehensive state immunization laws requiring receipt of specified vaccines for school attendance. School laws are now being expanded to include newly recommended vaccines.43

that will most effectively and efficiently increase the proportion of adolescents who receive newly recommended vaccines and develop ways to integrate these approaches into other adolescent health, education, and development programs.46

Vaccine Spacing Minimal spacing of vaccine doses is generally 1 month for the initial doses of killed vaccines; longer intervals are needed for booster doses to provide optimal boosting.7 The minimal spacing for MMR vaccines is 28 days. For children with delayed initiation of immunization (after 6 months of age), an accelerated schedule is recommended (Figure 7-2).7,47 To optimize adherence to the schedule in this circumstance, visits should be scheduled at 1-month intervals, and all recommended vaccines should be given at each visit. There is no need to restart any of the vaccine series among people with long delays between doses.

Simultaneous Administration All childhood vaccines can be administered simultaneously. This practice is based on extrapolation of data from multiple studies showing that most vaccines can be administered at the same time without compromising safety or immunogenicity.48 Thus, DTaP, Hib, IPV, hepatitis B, heptavalent pneumococcal conjugate vaccine (PCV7), MMR, varicella, and rotavirus vaccines can be administered simultaneously or within any interval of one another when appropriate.7,47 Interference between live virus vaccines other than OPV, rotavirus, and cold-adapted influenza vaccine (e.g., MMR and varicella) theoretically can occur if they are given within a short interval; live virus vaccines should be given either simultaneously or at least 1 month apart. Vaccines should not be mixed in the same syringe unless specifically licensed for such use. Interference has been found between certain vaccines (cholera and yellow fever).

Vaccine Administration

Spacing of Antibody-Containing Products and Vaccines

Vaccine Schedules

Immunoglobulins or blood products containing immunoglobulins inhibit response to certain live-virus vaccines (MMR and possibly varicella). The duration of inhibition of response is related to the dose of immunoglobulin delivered, and algorithms for calculating appropriate delays of MMR or measles vaccination after receipt of such products are available.18,47,49 In general, MMR vaccines should be delayed 3 months or longer after administration of usual doses of immunoglobulin (e.g., to prevent hepatitis A) or blood products, and for longer periods of time after higher doses (e.g., 10 months or more after 2 g/kg immune globulin intravenous such as for treatment of Kawasaki disease).

The CDC, the AAFP, and AAP annually publish harmonized childhood and adolescent immunization schedules. The ACIP and AAFP also publish an annual adult immunization schedule (www.cdc.gov/ vaccines). The ACIP, with input from many liaison organizations, periodically reviews the schedules to ensure consistency with new vaccine developments and policies. The first harmonized childhood immunization schedule was published in 1995 and recommended six vaccines containing antigens against nine infectious diseases44: diphtheria and tetanus toxoids and whole-cell pertussis vaccine (DTP); Td; MMR; Hib; OPV; and hepatitis B virus (HBV) vaccine. In January 2007 there were 12 vaccines against 16 infectious diseases in the childhood and adolescent immunization schedule (Figure 7-2). The harmonized schedule specifies both timing and the acceptable range of timing recommended for each dose of universally recommended vaccine and for vaccines recommended for children and adolescents in selected high-risk populations. Since inception, the major focus of the United States immunization program has been on immunizing infants and young children. In 1996, following growing concern about morbidity associated with vaccinepreventable diseases in the hard-to-reach adolescent population, the ACIP recommended expanding efforts to immunize adolescents (11 to 18 years of age) by establishing a routine vaccination visit at 11 to 12 years of age.45 In addition to providing Td and previously missed vaccinations, the report emphasized that this visit should be used to provide other important preventive health services. The addition of several new vaccines for adolescents (tetravalent meningococcal conjugate vaccine (MCV4), tetanus toxoid and reduced-content diphtheria toxoid and acellular pertussis vaccine (Tdap) and HPV) to the recommended schedule has stimulated a reappraisal of approaches

Interchangeability of Vaccines Available data support interchangeability of most vaccines produced by different manufacturers to prevent the same disease (tetanus, diphtheria, hepatitis B, and hepatitis A). Studies indicate that response to a three-dose series using different Hib conjugate vaccines equals or exceeds that when the same vaccine is used for all doses.50 The ACIP and AAP recommend that, when feasible, the same vaccine should be used for the primary series but that three doses of any vaccine are sufficient.7,47 Data are limited regarding safety, immunogenicity, and efficacy of using acellular pertussis (as DTaP) vaccines from different manufacturers for successive doses of the pertussis series. Data suggest that two of the current DTaP preparations may be used interchangeably for the first three doses of the DTaP series without affecting safety or immunogenicity.51 Whenever feasible, use of the same DTaP product for the entire series is recommended. When the specific product is not known or available, any DTaP vaccine should be used to continue or complete the series.

PART I Understanding, Controlling, and Preventing Infectious Diseases

Active Immunization

CHAPTER

7

53

DEPARTMENT OF HEALTH AND HUMAN SERVICES • CENTERS FOR DISEASE CONTROL AND PREVENTION

Recommended Immunization Schedule for Persons Aged 0–6 Years —UNITED STATES•2007 Vaccine

Age

Hepatitis B1

Birth HepB

1 2 4 6 12 15 18 19–23 2–3 month months months months months months months months years see footnote 1

HepB

HepB

Rotavirus2

Rota

Rota

Rota

Diphtheria, Tetanus, Pertussis3

DTaP

DTaP

DTaP

Haemophilus influenzae type b4

Hib

Hib

Hib4

Hib

Pneumococcal5

PCV

PCV

PCV

PCV

IPV

IPV

Inactivated Poliovirus Influenza6 Measles, Mumps, Rubella7 Varicella8 Hepatitis

A9

Meningococcal10 This schedule indicates the recommended ages for routine administration of currently licensed childhood vaccines, as of December 1, 2006, for children aged 0–6 years. Additional information is available at http://www.cdc.gov/nip/recs/child-schedule.htm. Any dose not administered at the recommended age should be administered at any subsequent visit, when indicated and feasible. Additional vaccines may be licensed and recommended during the year. Licensed combination vaccines may be used whenever any 1.Hepatitis B vaccine (HepB). (Minimum age: birth) At birth: • Administer monovalent HepB to all newborns before hospital discharge. • If mother is hepatitis surface antigen (HBsAg)-positive, administer HepB and 0.5 mL of hepatitis B immune globulin (HBIG) within 12 hours of birth. • If mother’s HBsAg status is unknown, administer HepB within 12 hours of birth. Determine the HBsAg status as soon as possible and if HBsAg-positive, administer HBIG (no later than age 1 week). • If mother is HBsAg-negative, the birth dose can only be delayed with physician’s order and mother’s negative HBsAg laboratory report documented in the infant’s medical record. After the birth dose: • The HepB series should be completed with either monovalent HepB or a combination vaccine containing HepB. The second dose should be administered at age 1–2 months. The final dose should be administered at age ≥24 weeks. Infants born to HBsAg-positive mothers should be tested for HBsAg and antibody to HBsAg after completion of ≥3 doses of a licensed HepB series, at age 9–18 months (generally at the next well-child visit). 4-month dose: • It is permissible to administer 4 doses of HepB when combination vaccines are administered after the birth dose. If monovalent HepB is used for doses after the birth dose, a dose at age 4 months is not needed. 2. Rotavirus vaccine (Rota). (Minimum age: 6 weeks) • Administer the first dose at age 6–12 weeks. Do not start the series later than age 12 weeks. • Administer the final dose in the series by age 32 weeks. Do not administer a dose later than age 32 weeks. • Data on safety and efficacy outside of these age ranges are insufficient. 3. Diphtheria and tetanus toxoids and acellular pertussis vaccine (DTaP). (Minimum age: 6 weeks) • The fourth dose of DTaP may be administered as early as age 12 months, provided 6 months have elapsed since the third dose. • Administer the final dose in the series at age 4–6 years. 4. Haemophilus influenzae type b conjugate vaccine (Hib). (Minimum age: 6 weeks) • If PRP-OMP (PedvaxHIB® or ComVax® [Merck]) is administered at ages 2 and 4 months, a dose at age 6 months is not required. • TriHiBit®(DTaP/Hib) combination products should not be used for primary immunization but can be used as boosters following any Hib vaccine in children aged ≥ 12 months.

4–6 years

HepB Series

DTaP

DTaP Hib

IPV

Range of recommended ages

PCV PPV IPV

Catch-up immunization

Influenza (Yearly) MMR Varicella HepA (2 doses)

MMR

Certain high-risk Varicella groups HepA Series MPSV4

other components of the vaccine are not contraindicated and if approved by the Food and Drug Administration for that dose of the series. Providers should consult the respective Advisory Committee on Immunization Practices statement for detailed recommendations. Clinically significant adverse events that follow immunization should be reported to the Vaccine Adverse Event Reporting System (VAERS). Guidance about how to obtain and complete a VAERS form is available at http://www.vaers.hhs.gov or by telephone, 800-822-7967. 5. Pneumococcal vaccine. (Minimum age: 6 weeks for pneumococcal conjugate vaccine [PCV]; 2 years for pneumococcal polysaccharide vaccine [PPV]) • Administer PCV at ages 24–59 months in certain high-risk groups. Administer PPV to children aged ≥2 years in certain high-risk groups. See MMWR 2000;49 (No. RR-9):1–35. 6.Influenza vaccine. (Minimum age: 6 months for trivalent inactivated influenza vaccine [TIV]; 5 years for live, attenuated influenza vaccine [LAIV]) • All children aged 6–59 months and close contacts of all children aged 0–59 months are recommended to receive influenza vaccine. • Influenza vaccine is recommended annually for children aged ≥59 months with certain risk factors, health-care workers, and other persons (including household members) in close contact with persons in groups at high risk. See MMWR 2006;55 (No. RR-10):1–41. • For healthy persons aged 5–49 years, LAIV may be used as an alternative to TIV. • Children receiving TIV should receive 0.25 mL if aged 6–35 months or 0.5 mL if aged ≥3 years. • Children aged 40 years) for those with carditis and residual heart disease. Recommendations for cardiac prophylaxis of the American Academy of Pediatrics and American Heart Association are shown in Table 8-3.1, 46 Recommendations for other infections associated with Streptococcus pyogenes are less certain. Most experts recommend prophylaxis for children with poststreptococcal chorea similar to that for rheumatic fever. Information for the cost–benefit, safety, and efficacy of treatment of suspected poststreptococcal sequelae (pediatric autoimmune neuropsychiatric disorders associated with streptococcal infection, PANDAS) has not been established; currently, routine prophylaxis is not recommended. Some experts recommend prophylaxis for several years in children who have had an episode of poststreptococcal reactive arthritis. In general in those exposed to cases of severe invasive group A streptococcal disease, including toxic shock syndrome and necrotizing fasciitis, although they are at modestly increased risk of disease, the risk is not sufficient to recommend prophylaxis, nor have studies been conducted to establish the efficacy of such treatment.

Asplenia Children with functional or anatomic asplenic states, particularly sickle-cell anemia, have been shown to benefit from continuous penicillin prophylaxis to prevent pneumococcal and other bacterial infections (see Chapter 108, Infectious Complications in Special Hosts). A single prospective controlled study showed an 84% reduction in rate of infection in children younger than 3 years who received daily oral penicillin V (125 mg twice daily for children < 5 years; 250 mg twice daily for children ≥ 5 years) compared with those receiving placebo.47 A follow-up placebo-controlled trial of penicillin prophylaxis beyond 5 years of age in children receiving com-

PART I Understanding, Controlling, and Preventing Infectious Diseases

Chemoprophylaxis

CHAPTER

8

77

TABLE 8-3. Chemoprophylaxis in Special Situations Disorder 46

Prior rheumatic fever

Prophylactic Agents

Dosing

Duration

Benzathine penicillin G

1.2 million units every 4 weeks (every 3 weeks if high-risk)

RF without carditis: 5 years after episode, or 21 years of age, whichever is longer

or Penicillin V or Sulfisoxazole or Erythromycin

250 mg twice daily

RF with carditis, without regurgitation (clinical or echocardiograph): 10 years after episode, or 500 mg if ≤ 27 kg (60 lb); 1 g if > 27 kg; well into adulthood, whichever is longer once daily 250 mg twice daily

RF with carditis, with regurgitation: 10 years after episode, or at least 40 years of age, sometimes lifelong

Asplenic statesa 47,49

Penicillin V

125 mg twice daily if < 5 years 250 mg twice daily if ≤ 5 years

At least until 5 years of age or 5 years after surgical removal, whichever is longer; see text

Pneumocystis jirovecii infection

TMP-SMX

5 mg/kg TMP (maximum 320 mg TMP) once daily or 3 consecutive days/week

Duration chemotherapy, significant immunosuppressione

Sulfisoxazole alt Amoxicillin alt Amoxicillin

50 mg/kg at bedtime

3–6 months

20 mg/kg at bedtime

3–6 months

40 mg/kg per day divided q8 hours at onset of upper respiratory tract infection

3–5 days

TMP-SMX

2 mg/kg TMP once daily

Variable

or Nitrofurantoin

1–2 mg/kg once daily

In children with cancer, organ transplantation, HIV infection51 Recurrent otitis media53,56

Recurrent urinary tract infection57,62

Endocarditis3 Dental, oral, or upper respiratory tract procedures

Standard Amoxicillin (PO) Standard (penicillin-allergic) or Clindamycin (PO) or Oral medication impossible Ampicillin (IM or IV) or Azithromycin or Clarithomycin Cefazolin or Ceftriaxone

50 mg/kg (maximum 2 g) procedure; then 25 mg/kg 6 hours after initial dose

Once 30 to 60 minutes before the procedure

20 mg/kg (maximum 600 mg)

50 mg/kg (maximum 2 g) 15 mg/kg (maximum 500 mg)

Once 30 to 60 minutes before the procedure

50 mg/kg (maximum 1 g)

Once 30 to 60 minutes before the procedure

Gentamicin (IM or IV) Alternate for low-risk patient Amoxicillin (PO)

2 mg/kg (maximum 80 mg) before procedure; can be repeated 8 hours after initial dose 50 mg/kg (maximum 3 g) 1 hour before procedure; then 25 mg/kg 6 hours after initial dose

alt, alternative; C, carditis; HIV, human immunodeficiency virus; IM, intramuscularly; IV, intravenously; PO, by mouth; R, residual valvular disease; RF, rheumatic fever; SMX, sulfamethoxazole; TMP, trimethoprim. a Asplenic states include congenital and surgical asplenia, and splenic dysfunction associated with hemoglobinopathies; also considered for individuals with complement disorders. Vaccinations are also very important. c Efficacy of prophylaxis is not proven in all circumstances. Dosing is listed only as guideline. b See Chapter 525, Pneumocystis jirovecii.

prehensive care for sickle-cell disease showed a low incidence of infection and no significant benefit for prophylaxis in preventing pneumococcal bacteremia or meningitis (2% in penicillin group versus 1% in placebo group).48 It is recommended that prophylaxis be discontinued in children older than 5 years, provided that they have not previously had invasive pneumococcal infection and are appropriately immunized for age. Asplenia from other causes (congenital, surgical, functional) is also associated with increased risk of fulminant septicemia. In com-

parison with the incidence in healthy children, the incidence of mortality from septicemia is 50-fold higher after splenectomy for trauma (compared with > 350-fold higher in children with sickle-cell disease). Although the risk is higher in younger children and in the first 5 years after splenectomy, fulminant septicemia has been reported in adults more than 25 years after splenectomy.1 The general consensus is that asplenic children with malignancy, thalassemia, congenital anomalies, or other diseases with high risk of fulminant infection should receive daily chemoprophylaxis. There is

78

SECTION

B Prevention of Infectious Diseases

less certainty about the need for prophylaxis in children who undergo splenectomy for trauma. In general, prophylaxis (in addition to appropriate immunization) should be strongly considered for asplenic children younger than 5 years and should be considered for older children. Asplenic children should receive conjugate pneumococcal, meningococcal, and Haemophilus vaccines in the recommended series; pneumococcal polysaccharide multivalent vaccines should be added after 2 years of age; revaccination with pneumococcal polysaccharide vaccine should be strongly considered after 3 to 5 years for children older than 5 years who are at high risk of severe pneumococcal infection.1

Other Underlying Conditions Individuals lacking terminal components of complement, properidin deficiency, or other abnormalities of complement pathways are at risk of recurrent meningococcal and other disseminated neisseria infections and may benefit from penicillin prophylaxis, but immunization with protein-polysaccharide or conjugate vaccines in children older than 2 years of age with these conditions is likely to be more protective (see Chapter 105, Infectious Complications of Complement Deficiencies).49 Trimethoprim-sulfamethoxazole (TMP-SMX) can prevent bacterial infections in children with chronic granulomatous disease, HIV infection, or acute lymphoblastic leukemia, although emergence of resistant bacteria is a potential problem.50 TMP-SMX is also beneficial in the prevention of Pneumocystis carinii pneumonia in many immunosuppressive conditions, including bone marrow and organ transplantation and acquired immunodeficiency syndrome (AIDS).51 Strategies to prevent serious clinical disease due to cytomegalovirus and herpes simplex virus after bone marrow and solid organ transplantation have been published.52

Recurrent Otitis Media Both sulfisoxazole and amoxicillin are effective for prophylaxis in otitis-prone children, and TMP-SMX has also been used with apparent efficacy.53,54 Children are considered candidates for prophylaxis if they have experienced 3 episodes of acute otitis media within the previous 6 months or 4 episodes within the previous 12 months. Continuous prophylaxis may be more effective than intermittent prophylaxis.54 Increase in b-lactam resistance of Streptococcus pneumoniae and in b-lactamase-producing organisms during prophylaxis with amoxicillin compared with sulfisoxazole favors choice of the latter agent.55 Most authorities recommend a trial of prophylactic antimicrobial therapy before consideration of placement of tympanostomy tubes.56

Urinary Tract Infection Antibiotic prophylaxis is effective in preventing recurrent urinary tract infection and is indicated in children with underlying anatomic or neurologic lesions leading to a higher risk of infection, especially those with obstructive lesions or vesicoureteral reflex. Children without identifiable risk factors who suffer recurrent infections may also benefit from prophylaxis. In the latter group, documentation of 3 or more urinary tract infections within a 1-year period is a recommended indication for consideration of prophylaxis.57 Nitrofurantoin (1 to 2 mg/kg per day), TMP-sulfadiazine, and TMP-SMX (1 to 2 mg/kg per day of TMP component) are the best-studied regimens in children.58–62 The benefit of prophylaxis against asymptomatic bacteriuria in an unobstructed urinary tract without ureteral reflux is unproven.

Cardiac Abnormalities The American Heart Association (AHA) periodically publishes guidelines for endocarditis prophylaxis. In 2007, major changes were made based on the Committee’s conclusions that 1) extremely small numbers of cases of infective endocarditis might be prevented by antibiotic prophylaxis for dental procedures even if prophylaxis were 100% protective; 2) prophylaxis should not be recommended solely on the basis of an increased lifetime risk of acquisition of endocarditis; 3) infective endocarditis prophylaxis for dental procedures should be recommended only for people with underlying conditions associated with the highest risk of adverse outcome from infective endocarditis; 4) for patients with these underlying conditions, prophylaxis should be given for all dental procedures that involve manipulation of gingival tissue or the periapical region of teeth or perforation of the oral mucosa.3 Recommendations also focus emphasis on good oral hygiene, access to routine dental care and eradication of dental disease to decrease the frequency of bacteremia from routine daily activities. Cardiac conditions associated with the highest risk of adverse outcome from endocarditis for which prophylaxis with dental procedures is recommended are shown in Box 8-1. Antibiotic prophylaxis is no longer recommended for any other form of congenital heart disease, including mitral valve prolapse with regurgitation, or acquired valvulopathy, hypertrophic cardiomyopathy or presence of cardiac pacemakers or implanted defibrillators – unless there is a history of previous endocarditis or placement of a prosthetic valve. Antibiotic prophylaxis is not recommended for shedding of deciduous teeth or bleeding from trauma to the lips or oral mucosa. Antibiotics recommended when prophylaxis is appropriate according to Box 8-1 are shown in Table 8-3. A single dose of antibiotic should be administered before the procedure. If the dose was inadvertently omitted, it could be administered up to 2 hours after the procedure. Antibiotic prophylaxis is no longer recommended for routine respiratory tract procedures (e.g., intubation, extubation, bronchoscopy).3 Antibiotic prophylaxis may be considered for patients listed in Box 8-1 who undergo an invasive procedure that involves incision or biopsy of the respiratory tract mucosa (e.g., tonsillectomy, adenoidectomy, bronchoscopy with incision of the respiratory mucosa) or when performed to treat an established infection (e.g., to drain an abscess or empyema). Antibiotic regimens should always include an agent active against viridans group streptococcus. Suspected or proven etiology of an abscess may dictate broader spectrum or an additional antibiotic. Antibiotic prophylaxis is no longer recommended solely to prevent endocarditis in patients who undergo genitourinary (GU) or gastrointestinal (GI) tract procedures, including diagnostic esophagogastroduodenoscopy or colonoscopy.3 For patients listed in Box 8-1 who have an established GU or GI tract infection or for those who receive antibiotic therapy to prevent wound infection or septicemia associated with a GU or GI procedure, it may be reasonable that the antibiotic regimen includes an agent active against Enterococcus species.3 For patients listed in Box 8-1 scheduled for elective cystoscopy or other urinary tract manipulation who have enterococcal urinary tract infection or colonization, antibiotic therapy to eradicate/suppress enterococci in the urine before the procedure may be reasonable. If the procedure is not elective, it may be reasonable that the empiric or specific antimicrobial regimen administered include an agent active against enterococci.3 Antibiotic prophylaxis is recommended for procedures on infected skin, skin structures, or musculoskeletal tissue only for patients listed in Box 8-1. In 2007, the AHA reaffirmed previous recommendations of settings for which endocarditis prophylaxis is not indicated (including cardiac catheterization and Cesarean delivery)63 and added the following procedures: ear and body piercing, tattooing, vaginal delivery and hysterectomy.3

PART I Understanding, Controlling, and Preventing Infectious Diseases

Protection of Travelers

CHAPTER

9

79

BOX 8-1. Prophylaxis for Infective Endocarditis According to Cardiac Conditions and Dental Procedures3 Cardiac Conditions Associated with Highest Risk of Adverse Outcome from Endocarditis for Which Prophylaxis with Dental Procedures is Recommended • Prosthetic cardiac valve • Previous infective endocarditis • Congenital heart disease (CHD)a — Unrepaired cyanotic CHD including palliative shunts and conduits — Completely repaired congenital heart defect with prosthetic material or device, whether placed by surgery or by catheter intervention, during the first 6 months after the procedureb — Repaired CHD with residual defects at the site or adjacent to the site of a prosthetic patch or prosthetic device (which inhibit endothelialization) • Cardiac transplantation in which cardiac valvulopathy has developed Dental Procedures for which Endocarditis Prophylaxis is Recommended for Patients Above • All dental procedures that involve manipulation of gingival tissue or the periapical region of teeth or perforation of the oral mucosa — Professional cleaning with gingival probing, biopsies, suture removal, placement of orthodontic bands Dental Procedures for which Endocarditis Prophylaxis is not Recommended even for Patients Above • Routine anesthetic injections through noninfected tissue • Taking of dental radiographs • Drilling of carious teeth • Orthodontic/prosthodontic procedures — placement or removal of appliances — placement of orthodontic brackets — adjustment of orthodontic appliances a

Except for conditions listed, antibiotic prophylaxis is no longer recommended for any other form of CHD. Prophylaxis is recommended because endothelialization of prosthetic material occurs within 6 months after the procedure.

b

CHAPTER

9

Protection of Travelers Maryanne E. Crockett and Jay S. Keystone Increasing numbers of people travel internationally each year: more than 750 million travelers crossed international borders in 2004.1 An estimated 4% of these people are children; consequently, more than 30 million children travel internationally each year.2 Annually up to 8% of travelers to the developing regions of the world are ill enough to seek medical healthcare while abroad or upon returning home.3,4 Although travel may expose children to certain risks, the beneÀts are many. Therefore, a careful pretravel evaluation to provide appropriate guidance and preparation is critical to protect pediatric travelers and their families and allow them to enjoy their time abroad.

PREPARATION FOR TRAVEL General Advice A pretravel evaluation should be performed at least 6 to 10 weeks prior to travel. The entire itinerary for the trip should be reviewed, including destinations, time and duration of travel, types of accommodation, activities, and potential exposure to insects and animals. The evaluation should also review the medical and particularly the immunization history of the child in order to ensure that appropriate advice is given regarding preventive measures, including necessary vaccines. This evaluation can be accomplished by providing a form for the parents to complete and bring to the initial pretravel assessment visit. Particular attention should be given to children of immigrants who are returning to their home countries to visit friends and relatives because these children have been shown to be at increased risk of many infectious diseases and may be less likely to seek pretravel advice.5,6 There are many excellent resources available that provide

pretravel advice for pediatricians. The majority of these resources are accessible online (Box 9-1). Guidance regarding travel health should be provided regarding safety issues and infectious diseases.7–9 Motor vehicle crashes are the most common cause of death among travelers; therefore, particular attention must be given to use of seat belts and car seats as recommended according to the age and size of the child. Car seats may not be readily available at the destination and therefore should accompany the family. Other injury concerns for children include drowning, falls from unprotected balconies or windows, and electrical injuries from unprotected outlets. A parent traveling alone with children should have notarized documentation authorizing him or her to travel with the children. Advice regarding food and water precautions and insect avoidance should be thoroughly reviewed. Skin protection is an important topic and includes both risk of serious sunburn and avoidance of infectious diseases. For sunblock, 30 is the minimum sun protection factor (SPF) recommended for children. Sunblock should be applied 30 minutes before exposure and always before insect repellent is applied where both are needed. Adolescent travelers should be counseled regarding safer sex practices and risks of body piercing and tattooing in less developed countries. Fresh water exposure of any kind should be avoided in areas that are endemic for schistosomiasis or where Leptospira organisms may contaminate the water. Exposure to infected stool of animals or humans can result in several types of parasitic infection either directly (e.g., hookworm) or through fecal–oral exposure (e.g., Toxocara spp.). Shoes provide more protection than sandals for children exposed to contaminated environments. Animal bites may result in injury, bacterial infection at the site, or rabies; therefore, children should be cautioned to avoid unknown animals, particularly dogs, while traveling. Since disposable diapers may not be available in some countries, parents should be aware that cloth diapers must be ironed after washing to kill eggs and larvae deposited on clothing by the tumbu fly, the vector of myiasis, in parts of Africa. A travel medical kit should be assembled prior to travel and carried with the family at all times (Box 9-2). As at home, medications should be stored in childproof containers out of reach of children. A discussion of travel health insurance and what to do in the event of illness should be included in the evaluation. Written material summarizing the pretravel advice also may be helpful for families.

80

SECTION

B Prevention of Infectious Diseases

BOX 9-1. Resources and Additional Information for Travelers • International Travel and Health, print version updated biannually, online version updated regularly by the World Health Organization (WHO). Available online at www.who.int/ith/ • WHO vaccine summaries: www.who.int/vaccines/globalsummary/immunization/countryprofileselect.cfm • Centers for Disease Control and Prevention (CDC) Health Information for International Travel, updated approximately every 2 years by the CDC, Atlanta, USA: US Department of Health and Human Services (The Yellow Book). Available online at www.cdc.gov/travel/yb/index.htm • CDC travel information section: www.cdc.gov/travel/ • CDC Morbidity and Mortality Weekly Report (MMWR): http://www.cdc.gov/mmwr/ • CDC Emerging Infectious Diseases Journal: http://www.cdc.gov/ncidod/EID/index.htm • CDC Malaria Hotline: 770-488-7788 • CDC Travelers’ Health Automated Information Line (toll-free): 1-877-FYI-TRIP • GIDEON (Global Infectious Diseases and EpidemiOlogy Network), available online at www.gideononline.com/ • Pickering LK, Baker CJ, Long SS, et al. (eds) Red Book: 2006 Report of the Committee on Infectious Diseases, 27th ed. Elk Grove Village, IL, American Academy of Pediatrics. (1-888-227-1770 Publications) – a new edition is published every 3 years • The Pan American Health Organization, the regional office of the WHO: www.paho.org/ • Immunization Action Coalition: www.immunize.org/izpractices/p5120.pdf • United States State Department Hotline for American Travelers (202-647-5225) • United States State Department: http://travel.state.gov/ • International Association for Medical Assistance to Travellers: www.iamat.org • Program for Monitoring Emerging Diseases (Pro-MED-mail): www.promedmail.org • Committee to Advise on Tropical Medicine and Travel (CATMAT): www.travelhealth.gc.ca • Travax: www.travax.scot.nhs.uk/ • United States: American Society for Tropical Medicine and Hygiene travel health: www.astmh.org • The International Society for Travel Medicine: www.istm.org • United Kingdom: www.travelhealth.co.uk/diseases/travelclinics.htm • Canada: www.travelhealth.gc.ca

BOX 9-2. Pediatric Travel Medical Kit NONPRESCRIPTION ITEMS • Personal information card: name, birth date, chronic medical conditions, regular medications, allergies, blood type, vaccination record, emergency contact information • First-aid supplies: bandages, adhesive tape, gauze, antiseptic cleaning solution, commercial suture/syringe kit (with letter from physician) • Thermometer • Analgesics/antipyretics: acetaminophen, ibuprofen • Skin care products: barrier ointment/cream, topical corticosteroid cream, disinfectant solution (e.g., chlorhexidine) • Antihistamine (e.g., diphenhydramine) • Insect repellent (diethyltoluamide: DEET), insecticide (permethrin) • Water purification system • Oral rehydration packets • Antimotility agent (e.g., loperamide) if older child • Extra pair of prescription glasses PRESCRIPTION ITEMS • Currently prescribed medications • Antimalarial prophylaxis • Antibiotic for travelers’ diarrhea (see text) • Topical antibacterial ointment/cream • Topical antifungal ointment/cream • Topical ophthalmic/otic antibiotic solution

Immunizations Although immunization rates have been increasing over the last few years in the United States, there remain a significant number of children who are underimmunized.10 Many countries with low immunization rates have ongoing transmission of vaccine-preventable illnesses that rarely are seen in North America. Consequently, children who travel must have up-to-date immunization coverage to minimize their risk of contracting vaccine-preventable diseases if they travel to countries where these diseases are prevalent. Country-specific vaccine-preventable disease statistics and immunization schedules can be found on the World Health Organization (WHO) website and a listing of international vaccine names also is available online.11,12 Travel vaccines are divided into the categories of routine, required, and recommended. Required travel vaccines are needed by travelers

to cross international borders, according to health regulations at destination. Proof of yellow fever vaccination may be required for entry into or travel from endemic countries. Vaccination against meningococcus and polio are required for travelers to the Hajj in Saudi Arabia.13 Recommended travel vaccines include vaccines that should be considered according to the risk of infection during travel. During the pretravel evaluation, some children may need to receive vaccines in the recommended childhood and adolescent immunization schedule administered in an accelerated manner to complete their primary series, catch-up with routine vaccinations, or complete the recommended pretravel vaccine series prior to departure13–16 (Table 9-1). The routine or catch-up schedule for immunizations should be continued when the child returns from traveling. Two or more inactivated vaccines may be administered simultaneously or with any interval between doses, as can inactivated and live vaccines. Two parenterally administered live vaccines, if not given at the same time, should be administered at least 28 days apart.17 Caution must be used when scheduling live vaccine administration following immune globulin (IG) administration because decreased immunogenicity of the vaccines may result.15 This is particularly true of measles and varicella-containing vaccines. IG should not be given less than 14 days prior to administration of a live vaccine, and measles and varicella-containing vaccines should be deferred from 3 to 11 months after IG administration depending on the indication and dose of IG required (see Chapter 6, Passive Immunization). Although the effect of IG administration on the immunogenicity of varicella vaccine is unknown, the current recommendation is to use the same guidelines for varicella vaccine and IG as are used for measles-containing vaccines.18 IG administration does not interfere with the immune response to yellow fever, oral polio virus (OPV), rotavirus vaccines or any inactivated vaccines.

Routine Childhood Immunizations Many vaccine-preventable diseases are endemic in most of the world; therefore, a child’s routine vaccine schedule should be brought up-todate prior to travel.16 In particular, the primary series of vaccines, including at least 3 doses of the diphtheria and tetanus toxoids and the acellular pertussis (DTaP) vaccine, should be administered and may be given according to accelerated dosing schedules as required (see Table

PART I Understanding, Controlling, and Preventing Infectious Diseases

Protection of Travelers

TABLE 9-1. Acceleration of Routine Vaccine Schedule for Travel Earliest Age for First Dose

Minimum Interval Between Doses

Combined hepatitis A and Ba

1 year

1 week, 2 weeks between 2nd and 3rd doses (booster after 1 year)

Hepatitis A

1 year

6 monthsb

DTaP

6 weeks

4 weeks, 6 months between 3rd and 4th doses

IPV

6 weeks

4 weeks

OPV

Birth

4 weeks

Hib (conjugate)

6 weeks

4 weeks (booster after 12 months of age)

Hepatitis B

Birth

4 weeks, 8 weeks between 2nd and 3rd doses (3rd dose should be given ≥ 16 weeks after 1st dose)

PCV7

6 weeks

4 weeks, 8 weeks between 3rd and 4th doses (after 12 months of age)

Measles

6 months followed by MMR at 12 months and at 4 to 6 years of age

4 weeks

MMR

12 months

4 weeks

Varicella

12 months

4 weeks if ≥ 13 years of age 3 months if < 13 years of age

Vaccine

DTaP, diphtheria, tetanus, acellular pertussis; Hib, Haemophilus influenzae b; IPV, inactivated polio virus; MMR, measles, mumps, rubella; OPV, oral polio virus; PCV7, pneumococcal conjugate. Regular immunization schedule should be reinstituted upon return from the endemic area. a Combined hepatitis A and B accelerated schedule is an off-label use for children. b Hepatitis A booster does not need to be given as an accelerated schedule as seroconversion rate following the first dose is high. The second dose can be given any time after 6 months to induce long-lasting immunity.

9-1). The Tdap adolescent preparation with acellular pertussis vaccine should be used as the adolescent booster beginning at 11 years of age.19 Children under 6 years of age should also receive the conjugate Haemophilus influenzae type b (Hib) vaccine prior to travel. Although global polio eradication previously had been targeted for 2005, 21 previosuly polio-free countries documented polio infection between 2002 and 2005, and polio remains endemic in a few countries in Asia and Africa. (An up-to-date listing of polio cases can be found at www.polioeradication.org).20 OPV, although widely used in the WHO Expanded Programme on Immunization – Plus (EPI-PLUS), is not available in the United States. An accelerated schedule for inactivated poliovirus vaccine (IPV) may be initiated if required, with the first dose being given at 6 weeks of age and subsequent doses being given at least 4 weeks apart.16 If a child is traveling in the first few weeks of life and OPV is available, vaccination with OPV may be initiated at birth, with subsequent doses at 4-week intervals.13 A booster dose of IPV should be given at 4 to 6 years of age. More than half a million children die of measles annually, with children less than 1 year of age having the highest risk of severe disease. The risk of subacute sclerosing panencephalitis also is related to acquisition of measles virus at a young age. Maternal antibodies generally protect infants for less than 6 months. Children between 6 and 12 months of age who are traveling to countries where measles is endemic (including all countries where measles vaccination is not universal) should receive one dose of monovalent measles vaccine prior to travel. The measles, mumps, rubella (MMR) vaccine may be

CHAPTER

9

81

used if monovalent measles vaccine is unavailable; however, only doses given at or after 12 months of age count as part of the routine immunization schedule. Children older than 12 months of age should receive two doses of MMR given at least 28 days apart prior to travel. Hepatitis B is part of the routine immunization schedule in the United States.21 Children who have not completed their hepatitis B series should receive hepatitis B vaccine prior to travel to highly endemic areas. The hepatitis B series may be accelerated with an interval of 4 weeks after the first dose and 8 weeks between the second and third doses (with at least 16 weeks between the first and third doses). There is also an accelerated schedule with doses given at 0, 1, and 2 months, followed by a fourth dose at 12 months. A hyperaccelerated schedule of 0, 7, and 21 days with a fourth dose at 12 months can be used if necessary, but this schedule is not licensed by the Food and Drug Administration. A 2-dose schedule of adult Recombivax at 0 and 4 to 6 months is licensed in the United States for adolescents 11 to 15 years of age.21 Hepatitis A vaccine is universally recommended for children in the United States and should be given as a 2-dose schedule beginning at 12 to 24 months of age with the second dose 6 to 18 months later.22 Children who have not received their hepatitis A vaccine series should be vaccinated prior to travel to developing countries. The majority of hepatitis A cases imported into the United States by travelers are related to travel to Mexico and Central America.22 Although hepatitis A generally causes asymptomatic or mild infection in young children, such children may shed the virus for prolonged periods; consequently, vaccination of young travelers is recommended to protect both the recipient and any contacts. Children from birth to younger than 1 year of age who are at high risk of exposure to hepatitis A may be given 0.02 mL/kg of IG intramuscularly as passive hepatitis A prophylaxis.22 For travel lasting longer than 3 months, a larger dose of 0.06 mL/kg should be used. If a child is traveling within 2 weeks of receiving the first dose of vaccine, the concomitant administration of IG may be considered; however, most travel medicine advisors do not recommend IG in this situation, even for travelers leaving the day after vaccination. Twinrix (GlaxoSmithKline) is a combined hepatitis A and B vaccine that is licensed for individuals 18 years of age and older.13,21 Twinrix-Junior is not licensed in the United States but is widely available in Europe and Canada for children between 1 and 15 years of age. These vaccines are given in a 3-dose schedule at 0, 1, and 6 months. For last-minute travel they can be accelerated in a schedule of 0, 7, and 21 days with a booster given at 1 year.23 Recently, in Canada and parts of Europe, two adult doses of the vaccine 6 months apart have been approved for children 1 to 15 years of age.24 Varicella vaccine is recommended for all susceptible children and is given in the United States 2 as doses to children from 12 months through 12 years of age. For children less than 13 years of age, the second dose should be given 3 months after the first. For adolescents 13 years of age and older, 2 doses are required with an interval of at least 4 weeks between doses.16 For children with unknown varicella status, a cost analysis suggests that serotesting before immunization is cost-effective for children 5 years of age and older if follow-up for immunization is assured, whereas immunization without assessing antibody status is cost-effective up to 4 years of age.25 The conjugate pneumococcal vaccine is part of the routine childhood immunization schedule and should be given as a 4-dose series at 2, 4, 6, and 12 to 15 months of age, although it also can be accelerated as needed (see Chapter 123, Streptococcus pneumoniae). A quadrivalent conjugate meningococcal vaccine for serogroups A/C/Y/W-135 was licensed in 2005 in the United States for children 11 years and older. It is recommended for use in all children 11 to 12 years of age and unvaccinated adolescents at high-school entry (15 years) (see Chapter 125, Neisseria meningitidis).26,27 This vaccine has been shown to be safe and to produce an excellent immune response in children between 2 and 10 years of age, although it is not yet approved for use in this age group.28 Guillain–Barré syndrome (GBS) was reported in adolescents vaccinated with the quadrivalent conjugate meningococcal vaccine;29 rate of GBS among vaccine

82

SECTION

B Prevention of Infectious Diseases

recipients is slightly higher than that seen in unvaccinated people. Surveillance for additional cases is ongoing. Influenza vaccine is recommended for children 6 months of age and older who are at risk of developing complications, such as children with chronic diseases. Influenza vaccine also is recommended for healthy infants and children from 6 to 59 months of age and close contacts of infants and children from 0 to 59 months of age.30 It is noteworthy for children who are traveling that the influenza season occurs from April to September in the southern hemisphere and yearround in the tropics.31 Influenza outbreaks have occurred on cruise ships and on organized group tours in any latitude and season.30

every 2 years for the polysaccharide vaccine and every 5 years for the oral Ty21a vaccine. The Ty21a vaccine is only available in capsules in the United States, which limits usefulness in younger children. The Ty21a vaccine must be refrigerated and taken with cool liquids approximately 1 hour before eating. The Ty21a vaccine should not be taken concurrently with the antimalarial proguanil, and antibiotics should not be used from the day before the first capsule until 7 days after completing the vaccine course. Clinical trials of a Vi conjugate vaccine demonstrating safety, efficacy, and immunogenicity in children 2 years of age and older are ongoing.36,37

Required and Recommended Vaccines for Travel

Yellow Fever Vaccine

Table 9-2 provides details regarding travel vaccines recommended for children.

Yellow fever vaccine is a live attenuated vaccine that may be required or recommended for travel to central South America and sub-Saharan Africa. Some countries in Africa require an international certificate of vaccination (or physician waiver letter) against yellow fever of all entering travelers; other countries may require evidence of vaccination from travelers coming from or traveling through endemic or infected areas. The vaccine is recommended for all children 9 children of age and older traveling to endemic areas. Yellow fever vaccine is effective 10 days after administration of the first dose and a booster is required every 10 years for travelers at ongoing risk. Risks and benefits of yellow fever vaccination and likelihood of infection must be considered carefully in pregnant women and people who are immunocompromised.38 Yellow fever vaccine contains egg protein; therefore, people with previous anaphylaxis to eggs should not receive the vaccine. The vaccine is only available in the United States from providers certified by state health departments.39 A vaccine-associated encephalitis syndrome has been reported in young infants at a rate of 0.5 to 4 per 1000 infants vaccinated.13 Neurologic symptoms occur 7 to 21 days after immunization; disease is related to reversion of vaccine virus to wild-type neurotropic virus. Consequently, the vaccine is contraindicated in infants less than 6 months of age. For infants 6 to 9 months of age who cannot avoid travel to a yellow fever-endemic area, consultation with an expert in the field is recommended. Yellow fever vaccine-associated viscerotropic disease, a severe systemic illness that can result in fatal organ failure, rarely has been reported.

Cholera Vaccine The risk of cholera is low for travelers. Cholera vaccines are not available in the United States. Cholera vaccines are licensed in some countries: WC/rBS (inactivated), variant WC/rBS (inactivated), and CVD 103-HgR (live attenuated).32 Cholera vaccine is not required for entry into any country. The WHO recommends use of cholera vaccine only for travelers who plan to work in refugee camps or as healthcare providers in endemic areas.33

Typhoid Vaccine Typhoid vaccine is recommended for pediatric travelers to the Indian subcontinent and other developing countries in Central and South America, the Caribbean, Africa, and Asia.34 Children are particularly at risk of developing the disease and of becoming chronic carriers. Two vaccines are available for prevention of typhoid: a live attenuated oral vaccine (Ty21a), which can be used in children 6 years of age and older, and a purified Vi capsular polysaccharide vaccine that is delivered intramuscularly to children 2 years of age and older. The efficacy of both vaccines is approximately 70%; therefore, receipt of the vaccine does not eliminate the need for food and water precautions.35 If exposure continues, revaccination is recommended

TABLE 9-2. Schedule and Dosing for Travel Vaccines Vaccine

Schedule

Minimum age

Dose (mL)

Route

BCG (live attenuated)

1 dose

Birth

< 30 days: 0.3 mL Intradermal preferred (dilute to half concentration) but subcutaneous > 30 days: 0.3 mL acceptable

None

Hepatitis A/B, combined (inactivated/recombinant)

3 doses: 0, 1, and 6 months

1 year

0.5 mL

Intramuscular

None

Japanese encephalitis (inactivated)

3 doses: 0, 7, and 14 or 30 days

1 year

1–3 year: 0.5 mL > 3 years: 1.0 mL

Subcutaneous

3 years

Meningococcal – A/C/Y/W-135 (polysaccharide)

1 dose

3 months (see text)

0.5 mL

Subcutaneous

< 4 years: 2–3 years ≥ 4 years: 3–5 years

Meningococcal – A/C/Y/W-135 (conjugated polysaccharide)

1 dose

11 years

0.5 mL

Intramuscular

Unknown

Rabies (inactivated cell culture)

3 doses: 0, 7, 21, or 28 days

Birth

1.0 mL

Intramuscular

Consider at 2 years if high-risk

Typhoid, Ty21a (live attenuated)

4 doses: alternate days

6 years

1 capsule

Oral

5 years

Typhoid, Vi (capsular polysaccharide)

1 dose

2 years

0.5 mL

Intramuscular

2 years

Yellow fever (live attenuated)

1 dose

9 months

0.5 mL

Subcutaneous

10 years

BCG, bacille Calmette-Guérin.

PART I Understanding, Controlling, and Preventing Infectious Diseases

Booster Dose

Protection of Travelers

Rabies Vaccine Rabies is highly endemic in Africa, Asia (particularly India), and parts of Latin America, but the risk to travelers is low. Pre-exposure rabies immunization is recommended for travelers with an occupational risk of exposure, for people planning extended stays in endemic areas where medical care is limited, and for outdoor travelers.40 Given that children are more likely to interact with animals and not report an animal bite, rabies pre-exposure vaccination should be considered for children traveling to endemic countries for at least 1 month. The pre-exposure vaccine series involves 3 doses of 1.0 mL given intramuscularly at 0, 7, and 21 or 28 days.40 The series can be administered using either of the two licensed vaccines in the United States: human diploid cell vaccine (HDCV), and purified chick embryo cell (PCEC) vaccine. If a child is bitten or sustains a skin-penetrating scratch by a potentially rabid animal, 2 additional doses must be completed as soon as possible, but rabies IG is not required. Without pre-exposure immunization, treatment requires rabies IG and 5 doses over 28 days of an approved vaccine. (Note: rabies IG is often not available in many developing countries.)

Japanese Encephalitis Virus Vaccine Japanese encephalitis, an arboviral infection transmitted by Culex mosquitoes, is endemic in rural areas of Asia although occasional epidemics occur in periurban areas. In temperate regions, transmission occurs from April to November, but disease occurs year-round in tropical and subtropical areas. The disease is uncommon in travelers.41 Although the majority of cases are subclinical, half of patients with clinical disease have persistent neurologic abnormalities and the case fatality rate is close to 25%.42 Vaccine is recommended for all travelers older than 12 months of age who are traveling in rural endemic areas for at least 1 month. Three doses of the inactivated vaccine that is available in the United States are given over 2 to 4 weeks. The vaccine has been associated with both immediate and delayed hypersensitivity reactions; therefore, travelers should receive their last dose of vaccine at least 10 days prior to travel and be observed for 30 minutes after vaccine administration. The duration of immunity is unknown. A booster can be administered after 36 months.

Meningococcal Vaccine Five serogroups of Neisseria meningitidis (A, B, C, Y, and W135) are responsible for the vast majority of meningococcal disease. The epidemiology of serogroups responsible for disease is changing worldwide; B, C, and Y are most prevalent in the United States, whereas A, C, and more recently W135 cause the majority of epidemic disease in sub-Saharan Africa where the incidence of meningococcal disease can be as high as 30 cases per 100,000 annually.26,27 Meningococcal vaccine is required for travelers to the Hajj and is also recommended for people traveling to the “meningitis belt” in equatorial Africa during the dry season from December to June. The quadrivalent conjugate vaccine for serogroups A/C/Y/W-135 should be given to children 11 years of age and older. For children 2 years of age and older who are traveling to areas where epidemics are occurring, the polysaccharide quadrivalent A/C/Y/W-135 vaccine is recommended. Although there is little response to polysaccharide vaccines in children less than 2 years of age, some short-term protection to serogroup A may be provided by two doses of the vaccine given 3 months apart; consequently, this is advised for infants from 3 to 24 months of age who are traveling to high-risk areas. Children who received the polysaccharide meningococcal vaccine before 4 years of age should be revaccinated within 2 to 3 years if they remain at risk.27 Conjugate vaccines for serogroups A, C, and A/C are available in a number of countries other than the United States for use in infants and older children. Seventeen cases of GBS have been reported in adolescents who received conjugated A/C/Y/W-135 meningococcal vaccine in the United States during 2005 and 2006; an association between the two events has been shown, with an excess risk of

CHAPTER

9

83

1.25/million doses.29 A vaccine for group B meningococcus has proven elusive, although development is ongoing and an epidemic strain-specific vaccine (MeNZB) has been licensed in New Zealand and is undergoing postlicensure evaluation.43,44

Tickborne Encephalitis Virus Vaccine Tickborne encephalitis is transmitted by Ixodes rincinus ticks in the forests of central and eastern Europe during the summer months.42 Although 2 vaccines are licensed in some countries, including Canada, for use in children, neither is available in the United States.

BCG Bacille Calmette-Guérin (BCG) vaccine is part of the routine vaccination schedule in many developing countries where tuberculosis (TB) is highly endemic. BCG does not prevent TB infection but has been shown to decrease the incidence of severe TB disease such as miliary TB and TB meningitis. Vaccination with BCG can be considered for a young human immunodeficiency virus (HIV)-negative traveler (under 5 years of age) who will be spending a substantial period of time in a country that is highly endemic for TB when contact with people with active TB is likely.13,45 In addition, children who do not receive BCG and who have traveled to a country with a high TB burden should have a tuberculin skin test prior to and 3 months following their travel.31

Malaria Prophylaxis Malaria is caused by infection with Plasmodium species, most commonly through the bite of an infected female Anopheles mosquito. Malaria is one of the leading causes of death among children under 5 years of age worldwide, causing more than half a billion infections and 1 million deaths each year. Young children, pregnant women, and people who previously or recently have not been exposed to malaria have the highest risk of severe disease. Although malaria is endemic throughout the tropics, the highest risk for malaria infection in travelers occurs in sub-Saharan Africa, Papua New Guinea, the Solomon Islands, and Vanuatu.46 There is no vaccine available for prevention of malaria infection; therefore, families traveling with children must be given advice regarding personal protective measures and malaria chemoprophylaxis if they are traveling to endemic areas.

Chemoprophylaxis The type of chemoprophylaxis recommended depends on the likelihood of drug resistance, potential adverse reactions, cost, and convenience. In addition, characteristics of the individual traveler, including age, ability to swallow tablets, and any specific contraindications, are relevant.47 Breastfeeding infants require prophylaxis since antimalarial drugs do not reach high enough levels in human milk. Several medications are recommended for prevention of malaria in children: chloroquine, mefloquine, doxycycline, and atovaquone/ proguanil (AP, Malarone). Primaquine, a second-line drug for prophylaxis, may be useful when other antimalarial drugs cannot be used (see Chapter 271, Plasmodium Species (Malaria)). Chloroquine and mefloquine should be initiated 1 to 2 weeks prior to travel although doxycycline, AP, and primaquine may be started 1 day before exposure. All chemoprophylaxic agents must be continued for 4 weeks after departure from malaria-endemic areas, except for AP and primaquine which need be continued for only 1 week after exposure.

Protective Measures Because no malaria chemoprophylaxis is 100% effective, personal protective measures, such as barrier and chemical protection and

84

SECTION

B Prevention of Infectious Diseases

exposure avoidance, should be used to minimize risk of contact with mosquitoes. These protective measures also can decrease risk of other insectborne diseases, such as dengue and other arboviruses. Since Anopheles mosquitoes that transmit malaria bite from dusk to dawn children must have adequate protection during these hours. The Aedes mosquito that transmits yellow fever, chikungunya, and dengue virus bites primarily in the early morning and late afternoon. The vector of Japanese encephalitis, the Culex mosquito, bites between dusk and dawn. When there is a risk of insect exposure, children should be dressed in light-colored clothing that covers their arms and legs. Other measures to avoid insect bites include staying in air-conditioned or well-screened accommodation or using insecticidetreated bed nets. Chemical protection provides additional defense against insectborne diseases. The safest and best studied is N,N-diethyl-metatoluamide (DEET).47 Although adverse reactions, such as encephalopathy and rashes, have been described with use of high concentrations of DEET in children, this compound is considered safe when used appropriately according to product label instructions46, 48 (Box 9-3). The concentration of DEET correlates with duration of protection; therefore, products with lower concentrations need to be reapplied. DEET is approved by the Environmental Protection Agency and the American Academy of Pediatrics in a concentration of 30% down to 2 months of age; in standard preparations, this concentration will provide 4 to 6 hours of protection. Picaridin (7%), recently approved as Bayrepel and Cutter Advanced in the United States, appears to be a safe and well-tolerated repellent that provides protection for only 2 to 3 hours. Citronella oil is impractical since its duration of action is less than 1 hour.49 Permethrin (a safe chrysanthemum derivative) is a contact insecticide that may be used for treatment of bed nets and clothing.50 Permethrin-treated fabric has a duration of efficacy between 2 weeks and 6 months depending on the method of treatment. The best chemical protection against mosquito bites is the use of a combination of permethrin-treated clothing and DEET on exposed skin.

TRAVELERS’ DIARRHEA Risk Travelers’ diarrhea is one of the most common illnesses among travelers, affecting 9% to 40% of children who travel.50 Both the incidence and severity of travelers’ diarrhea are age-dependent, with the highest rates, longest duration, and greatest severity occurring in infants and children under 3 years of age.51 Children’s stools may normally be quite variable; consequently, travelers’ diarrhea is defined as a twofold or greater increase in the frequency of unformed stools lasting at least 2 to 3 days. The infectious causes of travelers’ diarrhea in children and adults are predominantly bacterial and include enterotoxigenic Escherichia coli (ETEC), which is the most common

BOX 9-3. Precautions for Use of Diethyltoluamide (DEET) • • • • • • • • •

Use repellents containing ≥ 30% DEET only Apply sparingly to exposed skin Apply only to intact skin Apply to face by wiping; avoid eyes and mouth; do not spray directly on face Wash off with soap and water when coming indoors Do not inhale or ingest repellent Do not apply on hands or other areas that are likely to come in contact with the eyes or mouth Do not allow children under 10 years to apply DEET themselves. Apply to your own hands then apply to the child Do not use on children less than 2 months of age

cause, enteroaggregative Escherichia coli (EAEC), Salmonella, Campylobacter, Shigella, enteropathogenic Escherichia coli (EPEC), and, rarely, enterohemorrhagic Escherichia coli (EHEC). Viral and parasitic infections are less common causes of pediatric travelers’ diarrhea, although rotavirus, Cryptosporidium parvum, Giardia lamblia, and Entamoeba histolytica also account for a small proportion of diarrhea in young travelers. The risk of developing travelers’ diarrhea depends on the travel destination, with rates as high as 73% among children traveling to North Africa and 61% among children visiting India.51 Travel to Southeast Asia, Latin America, and other African countries has been associated with rates of approximately 40%. Although travelers’ diarrhea is generally a self-limited infection, it can cause significant morbidity, particularly if it results in moderate to severe dehydration. Parents must be counseled regarding the symptoms and signs of dehydration as well as the approach to oral rehydration and when to seek medical attention.

Preventive Measures Because there are no vaccines licensed in the United States for prevention of travelers’ diarrhea in children, counseling regarding food and water precautions is the most important preventive measure. Vaccines are in development in preclinical and clinical phases against ETEC, Shigella spp., and Campylobacter jejuni; a combined cholera and ETEC oral vaccine is licensed in Canada for children 2 years of age and older.52 General rules regarding food and water precautions when traveling apply to both children and adults; however, young children are more likely to explore the environment with their hands and mouths, thus creating opportunities for infection. Frequent handwashing with soap and water is critical, particularly before eating, although alcohol-based handwashes may be used no water is not available. Children must be reminded to use safe water sources for all drinking, toothbrushing, and food preparation. Safe water sources include bottled water from a trusted source or water that has been boiled, chemically treated, or filtered. Combination chemical and filter pumps may provide the best protection in filtered water as filters vary in the size of microbes which are removed.2 Water should be boiled for at least 1 minute at altitudes less than 2000 meters and 3 minutes at greater than 2000 meters.31 Carbonated drinks also are considered safe for drinking, but water used to make ice may be contaminated. For infants, breastfeeding is the safest form of nutrition. In addition to its many health benefits, breastfeeding does not require a source of clean water, unlike the use of formula, both in its preparation and the cleaning of bottles. The selection and preparation of foods are important during travel to minimize the risk of travelers’ diarrhea. Although the advice to “boil it, cook it, peel it, or forget it” frequently is given, it is often not practical to follow. If possible, only steaming-hot freshly made food should be consumed. Families traveling with children should have a ready supply of snacks and avoid buying food from street vendors (Box 9-4). Additional food and water precautions can decrease risk of other infectious diseases while traveling. These include avoidance of unpasteurized dairy products to eliminate risk of brucellosis and other bacterial infections. Raw or undercooked meat and fish should not be consumed due to risk of parasitic infections. Avoiding undercooked seafood can decrease risk of hepatitis A. In developing countries raw vegetables and fruit that cannot be self-peeled should be avoided. Chemoprophylaxis for travelers’ diarrhea generally is not advised in children.51 However, short-term prophylaxis (< 3 weeks) could be considered for children with increased susceptibility to travelers’ diarrhea, such as children with achlorhydria, or children in whom travelers’ diarrhea might have significant medical consequences (e.g., children with chronic renal failure, congestive heart failure, diabetes mellitus, or inflammatory bowel disease).53

PART I Understanding, Controlling, and Preventing Infectious Diseases

Protection of Travelers

CHAPTER

9

BOX 9-4. Prevention of Travelers’ Diarrhea in Children

TABLE 9-3. Formulation of Oral Rehydration Solution (ORS)

DO • Eat only thoroughly cooked food served hot • Peel fruit • Drink only bottled, carbonated, boiled, chemically treated, or filtered water • Prepare all beverages and icecubes with boiled or bottled water • Wash hands before eating or preparing foods • Continue breastfeeding throughout travel period DON’T • Eat raw vegetables or unpeeled fruit • Eat raw seafood or shellfish or undercooked meat • Eat food from street vendors • Drink tap water • Consume milk or dairy products unless labeled as pasteurized or irradiated

World Health Organization

Home Formula

• Sodium chloride 2.6 g/L (75 mmol/L sodium) • Potassium chloride 1.5 g/L (20 mmol/L potassium) • Trisodium citrate, dihydrate 2.9 g/L (10 mmol/L citrate) • Glucose, anhydrous 13.5 g/L (75 mmol/L glucose)

• 3.5 g NaCl (3/4-teaspoon table salt)

Treatment Treatment of travelers’ diarrhea in children must include close attention to hydration status, and parents should be counseled regarding early signs of dehydration. Oral rehydration therapy (ORT) using a homemade or commercially prepared oral rehydration solution (ORS) can be used to prevent dehydration associated with diarrheal disease. Commercial ORS should be used to treat mild to moderate dehydration; severe dehydration may require intravenous fluid resuscitation.54,55 ORS packets should be part of a family’s travel medical kit. Locally made preparations can be used early in therapy, although they differ in composition from the reduced-osmolarity ORS recommended by WHO (Table 9-3).54,55 Breastfeeding should be continued in infants, and solid food intake should be maintained along with rehydration with ORT throughout the diarrheal episode, although foods high in simple sugars should be avoided because the increased osmotic load may worsen fluid losses. Loperamide generally is used in combination with antibiotics for treatment of travelers’ diarrhea in adults; however, the role of loperamide in pediatric travelers’ diarrhea remains controversial, despite being licensed for use in children 2 years of age and older. Although loperamide has been shown to decrease duration and severity of acute diarrhea in children, this drug has been associated with significant side effects in children and is not recommended for younger children.54,56 Racecadotril is an enkephalinase inhibitor that has been associated with decreased stool output in clinical trials; however, further studies are required. Zinc supplementation has been associated with improved outcomes in diarrheal disease in children in developing countries, but zinc supplementation is not recommended in treatment of travelers’ diarrhea.54 There is little evidence for use of antimicrobial agents in pediatric travelers’ diarrhea.51 Fluoroquinolones for 1 to 3 days are the drug of choice for adults with travelers’ diarrhea that is moderate to severe, persistent (> 3 days), or associated with fever or bloody stools. Although there are concerns regarding the potential for development of arthropathy and antimicrobial resistance with fluoroquinolone use in children, the Food and Drug Administration has approved ciprofloxacin for anthrax and as a second-line agent for the treatment of urinary tract infections in children from 1 to 17 years of age.57,58 Therefore, fluoroquinolones could be considered safe in children for the short course required for travelers’ diarrhea. A 3-day course of ciprofloxacin at a dose of 20 to 30 mg/kg per day divided twice daily with a maximum dose of 500 mg bid is recommended for children with moderate to severe or bloody diarrhea.51 Azithromycin is often used as the first choice for treatment of pediatric travelers’ diarrhea, especially in areas with a high prevalence of fluoroquinolone-resistant Campylobacter species such as India and Thailand because it is given once a day and has a known safety profile in children. A dose of 10 mg/kg once daily for 3 days (maximum dose

85

• 1.5 g KCl (1 cup orange juice) • 2.5 g NaHCO3 (1 teaspoon baking soda) • 20 g glucose (4 tablespoons sugar) Water to final volume of 1 L (33 oz)

of 500 mg) is appropriate.51 In adults a single dose of antibiotic has been shown to be as effective as 3 days’ treatment; therefore, in children a full 3-day course may not be necessary.59,60 Rifaximin (Xifaxan), a nonabsorbed rifamycin derivative, has been approved in the United States for treatment and prevention of travelers’ diarrhea for people 12 years of age and older.61 A liquid preparation is available in some countries for pediatric use. If travelers’ diarrhea does not respond to a course of antimicrobial therapy, medical attention should be sought to investigate other possible causes of the diarrhea.

EMERGING INFECTIOUS DISEASES Over the past few years, several infectious agents, such as severe acute respiratory syndrome (SARS) coronavirus, and the H5N1 strain of avian influenza, have emerged as potentially widespread health threats. Although the SARS coronavirus does not appear currently to be of concern, pediatricians who are advising families regarding travel health must keep informed of the current status of emerging infectious diseases that may pose a threat to the traveler. Several websites provide up-to-date information regarding such infections, including that of the WHO and the Centers for Disease Control and Prevention (see Box 9-1). A highly pathogenic strain of avian influenza (H5N1) has caused outbreaks in poultry in several countries in Asia, Africa, the middle East, and eastern Europe. Human cases of H5N1 also have been documented in Cambodia, China, Indonesia, Iran, Thailand, Turkey, and Vietnam. Although there have been rare confirmed human-tohuman transmissions of H5N1, which have high case-fatality rates have also been documented in a number of these countries. (An up-todate listing of confirmed human cases can be found. Although there have been rare confirmed human-to-human transmission of H5N1 in these outbreaks, there is concern that further mutations in the virus may result in a pandemic strain of influenza. A number of recommendations for travelers have been made to decrease their risk of acquiring H5N1 infection62 (Box 9-5). Although oseltamivir has been used in treatment of and prophylaxis against H5N1, it is not recommended that a prescription for oseltamivir be given to travelers.

THE IMMUNOCOMPROMISED TRAVELER Children with immunodeficiencies require special consideration at their pretravel evaluation because of increased risk of travel-related illness.63 Most patients with altered immune systems, particularly those with decreased T-lymphocyte immunity, should not receive live vaccines because of risk of developing clinical illness from the vaccine strain.64 IPV should be given instead of OPV to all members in the family of an immunocompromised person, and vi typhoid vaccine should be administered instead of the Ty21a vaccine to an immunocompromised child, although there is no risk to the patient if family members receive the live oral vaccine.15,21 However, MMR, varicella, and yellow fever vaccines should be considered for HIV-seropositive

86

SECTION

B Prevention of Infectious Diseases

BOX 9-5. Precautions to Decrease Risk of H5N1 Infection • Avoid all direct contact with poultry and ducks, including poultry farms and bird markets • Wash hands frequently with soap and water (alcohol-based handwashes can be used if hands are not visibly soiled) • Cook all poultry-based foods, including eggs, thoroughly

children who are not severely immunocompromised (see Chapter 227, Rubeola Virus (Measles and Subacute Sclerosing Panencephalitis); Chapter 205, Varicella-Zoster Virus). Killed or subunit vaccines may be administered to children with altered immunity, although responses to the vaccines may be diminished.64 Asplenic patients may respond poorly to polysaccharide vaccines in particular. Patients with certain B-lymphocyte deficiencies, such as X-linked and common variable agammaglobulinemia, should avoid OPV, vaccinia, and live bacterial vaccines, although other patients with humoral deficiencies, including selective immunoglobulin A (IgA) and IgG subclass deficiency, need only avoid OPV; other live vaccines can be considered. Some travel-associated illnesses may be more severe in immunocompromised travelers. Asplenic travelers are at greater risk of severe babesiosis and malaria, and organ and stem cell transplant recipients are more likely to develop bacteremia associated with gastroenteritis due to Salmonella or Campylobacter spp.65 HIV-seropositive travelers with low CD4 lymphocyte counts must be particularly conscious of risk factors associated with opportunistic infections such as Toxoplasma gondii, Isospora belli, Salmonella spp. and Cryptosporidium parvum,65 and, therefore, must be particularly cautious regarding food, water, and animal exposures.

RETURN FROM TRAVEL Routine posttravel screening generally is not required for asymptomatic, short-term travelers, although screening may be considered for long-term travelers, expatriates, adventure travelers, and people who have experienced significant illness while traveling.4,66 If posttravel screening is indicated, the tests required should be determined by the potential exposures associated with the travel itinerary and any symptoms, if present. Children who develop symptoms after travel should seek immediate medical attention, and parents must inform the physicians caring for them of their travel itinerary. This is particularly critical if the itinerary has included a malaria-endemic area, since chemoprophylaxis cannot prevent all cases of malaria. Because malaria can present with nonspecific symptoms in children, any symptoms of fever, rigors, headache, malaise, abdominal pain, vomiting, diarrhea, poor feeding, or cough following travel to an endemic country should be evaluated promptly by a physician.67 Travel-relatetd illness has been shown to be highly dependent on itinerary. In a report of disease and relationship to place of exposure among ill returned travelers, significant regional differences in proportionate morbidity were reported.4 Typhoid fever was seen most frequently in travellers returning from South Asia. Malaria was one of the three most frequent causes of systemic febrile illness among travelers from every region, especially sub-Saharan Africa, although travelers from every region except sub-Saharan Africa and Central America had confirmed or probable dengue more frequently than malaria. Rickettsial infection, primarily tickborne spotted fever, occurred more frequently than malaria or dengue among travelers returning from southern Africa.4

SECTION

C

Host Defenses Against Infectious Diseases

CHAPTER

10

Immunologic Development and Susceptibility to Infection Maite de la Morena

The human immune system has evolved to protect the individual from infectious microbes. It does this by utilizing a complex interactive network of cells, proteins, and organs. This response is both innate and adaptive, each with unique characteristics. The innate response to a pathogen occurs immediately (within hours), lacks clonal specificity for a particular pathogen, and does not confer long-lasting protection, i.e., immunologic memory. The adaptive immune response, although triggered by components of the innate immune response, takes days

to evolve, requires processing and presentation of antigens derived from the pathogen, is specific to the particular pathogen and most importantly confers immunologic memory, i.e., the organism “remembers” the signature of a pathogen upon subsequent encounter. Experiments of nature in humans, such as those recognized as the inherited disorders of immune function,1 have taught us that despite the apparent redundancy of the system, quantitative and qualitative defects in individual components and/or pathways result in abnormal function and susceptibility to particular infections. This chapter provides a general overview of the development of innate and adaptive immune responses, addresses some of the immunologic developmental characteristics unique to the fetus and newborn, and addresses pathogen susceptibility in general terms, which can serve as an introduction to Sections M, N, and R of this textbook.

THE INNATE IMMUNE RESPONSE The innate immune system offers a first line of defense against invading pathogens. Both cellular and humoral factors constitute its major components. These include: (1) antimicrobial products and phy-

PART I Understanding, Controlling, and Preventing Infectious Diseases

Immunologic Development and Susceptibility to Infection

sical barriers such as skin and mucosal surfaces; (2) receptors for pathogen molecules, including the family of Toll-like receptors (TLRs); (3) phagocytic cells: neutrophils and macrophages; (4) dendritic cells (DC); (5) the complement system; and (6) natural killer (NK) cells.

Antimicrobial Products, the Skin, and Mucosal Barriers In mammals, epithelial cells are capable of secreting two classes of antimicrobial peptides: defensins and cathelicidins. Defensins can be further categorized into a- or b-defensins and contribute to host defense by disrupting the cytoplasmic membranes of microbes. aDefensins are produced by neutrophils, monocytes, and Paneth cells of the gut whereas b-defensins are produced by epithelial cells. A human cathelicidin, hCap18/LL-37, has been found in epithelial cells, mast cells, monocytes, and lymphocytes and has neutralizing capability against lipopolysaccharide (LPS), stimulates angiogenesis, and acts as a chemoatractant for neutrophils, monocytes, and T lymphocytes.2–5 The skin is the most important barrier to pathogen entry. Tight junctions between epithelial cells, skin thickness, and a dry environment offer a shield against microbes. Loss of skin integrity, as seen in wounds, burns, and inflammation, allows the entry of pathogens through this barrier. Both psoriasis and atopic dermatitis (AD) are known inflammatory skin conditions associated with skin disruption. However, although infection is rarely associated with psoriasis, patients with AD are commonly infected with Staphylococcus aureus. Human b-defensin 2 (HBD2) and the cathelicidin LL37 appear to be strongly expressed in psoriasis and not in eczematous skin. Interleukin (IL)-13, produced under atopic conditions, suppresses the induction of these antimicrobial peptides.6 Interestingly, LL-37 has also been identified in the ductal epithelium of salivary and sweat glands, suggesting a role in the protection of the gland itself from microbial invasion and providing protection to the epithelial surface via secreted products.7 During the third trimester of pregnancy, the fetus becomes covered by the vernix caseosa, which contains antimicrobial peptides, including a-defensins, LL-37, and psoriasin, a calcium-binding protein that is upregulated in psoriasis. Vernix extracts exhibited both antibacterial activity against gram-negative bacteria and antifungal properties against Candida albicans, whereas amniotic fluid derived proteins and peptides showed only the former activity.8 The more common entry pathway for pathogens is through the mucosal barrier. Mucosal epithelial cells secrete mucus that contains many antimicrobial peptides, including defensins and cathelicidins. Mucus acts dually. First, it coats the pathogen, allowing the antimicrobial peptides to exert their action; and second, it acts as a vehicle for particles and pathogens to be cleared by the action of cilia. Within the respiratory tract, cilia move the mucus towards the upper airways where it is either expelled through cough mechanism or swallowed. Ineffective clearance, as seen in patients with immotile cilia syndrome or after lung transplantation, may further contribute to colonization with pathogens such as Pseudomonas aeruginosa (see Chapter 155, Pseudomonas aeruginosa). The balance between the fluid composition of the mucus and antimicrobial properties is disrupted in patients with cystic fibrosis (CF) due to mutations in the CF transmembrane conductance regulator (CFTR) gene, leading to bacterial overgrowth and chronic inflammation. Recently, LL-37 gene transfer experiments preformed in C57B mice conferred protection against intratracheal injection of Pseudomonas aeruginosa, suggesting a potential therapeutic approach for these antimicrobial peptides.9 Furthermore, HDB2 and LL37 appear to be consistently elevated in states of infection and may contribute to pathogen clearance, as knockout mice lacking b-defensin cannot clear Haemophilus influenzae.10 Surfactant-associated proteins, specifically surfactant protein A (SP-A) and surfactant protein D (Sp-D), contribute to the innate immune responses in the lung. Produced by type II pneumocytes and nonciliated respiratory epithelial cells, they belong to the family of proteins called collectins. SP-A and SP-D can interact with

CHAPTER

10

87

microorganisms, modulate local inflammatory responses, modulate neutrophil responses in vitro, and participate in clearance of pollens and other complex organic antigens.11 Recently SP-D has been shown to limit the intracellular growth of bacilli in macrophages by increasing phagosome–lysosome fusion, but not by generating a respiratory burst.12 Lysozyme, lactoferrin, and phospholipase A2 in tears and saliva, and histatins in saliva, are potent antibacterial enzymes as well.13,14 The gastrointestinal tract is protected by digestive enzymes, bile salts, fatty acids, and lysolipids. Paneth cells in the human gut secrete adefensins and influence the virulence of orally ingested bacteria. Thus, children and adults with infections due to Shigella or virulent Salmonella strains have demonstrated decreased synthesis by colonic enterocyte of HBD1 and cathelicidin LL-37. HBD2 expression is also reduced in enterocytes of patients with Crohn disease and gastric mucosa-derived b-defensins are seen in Helicobacter pylori infections. Interestingly, in vitro, this microbe is susceptible to HBD2. Commensal bacteria are resistant to endogenous antimicrobial peptides but induce epithelial defensins. Porphyromonas gingivalis does not induce HBD2 and behaves as a silent invader (for an excellent review, see reference 14).

Pathogen Receptors and Toll-like Receptors Although the innate immune response is not specific in terms of immunologic memory, features that may be common to different pathogens can be recognized by the cells of the innate immune system. These unique features of microbe are known as pathogen-associated molecular patterns (PAMP) and include carbohydrates and lipoproteins or nucleic acids expressed as part of their life cycle within the host. For example, bacterial DNA as unmethylated repeats of dinucleotide CpG and double-stranded (ds) or single-stranded (ss) RNA are known PAMPs. Pathogen recognition receptors such as mannose-binding lectin (MBL), which is a circulating soluble protein, can bind mannose or fucose residues of a certain spatial orientation, and allows the bacteria to become susceptible to phagocytosis and complement activation. Macrophages can carry a C-type lectin called macrophage mannose receptor (MMR) which not only binds carbohydrate moieties found on the surface of bacteria but can also recognize viruses such as the human immunodeficiency virus (HIV).16 Toll receptors are an important group of signaling molecules capable of recognizing PAMPs. Their importance lies in their ability to link innate and adaptive effector functions (for a review see Chapter 11, Fever and the Inflammatory Response). Toll receptors were first identified in Drosophila melanogaster and discovered in humans due to similarities to the mammalian IL-1 receptor (IL-1R), thus the name Toll-like receptor.17,18 They are present on many cells, including airway and gut epithelial cells, antigen-presenting cells (APC: B lymphocytes, macrophages, DC, monocytes), mast cells, regulatory T lymphocytes, NK lymphocytes, and endothelial cells.19 A total of 10 different TLRs have been identified in humans: TLR-1, -2, -4, -5, TLR-6 and TLR-10 are found on cell surfaces, whereas TLR-3, -7, -8, and -9 are localized within the endosomes. TLR-2 is involved in responses to gram-positive bacteria (peptidoglycans and lipoproteins) and yeast.20 TLR-4 mediates the interaction of gram-negative bacteria by transducing signals derived from LPS. A model for TLR-4 mutations renders mice resistant to endotoxin but highly susceptible to gram-negative organisms.21 All TLRs are capable of interacting with different ligands (see Table 11-1). RSV F protein, LPS, and Pseudomonas exoenzyme S have been shown to interact with TLR-4 whereas flagellin is recognized by TLR-5.22 TLR-2 recognizes envelope proteins of herpes simplex virus (HSV) whereas TLR-9 identifies CpG motifs within the viral genome.23 Binding of the microbial components to the TLR triggers the activation of two downstream signaling pathways where myeloid differentiation factor 88 (MyD88) and/or Toll-IL-1 receptor domain containing adaptor-inducing interferon (IFN)-b (TRIF) lead to

88

SECTION

C Host Defenses Against Infectious Diseases

activation of NF-kappaB and subsequent transcription of proinflammatory cytokines: tumor necrosis factor-a (TNF-a, IL-1, and IL-6. MyD88 recruits the IL-1R-associated kinase (IRAK) family of proteins: IRAK-1 and IRAK-4. Humans and mice with IRAK-4 deficiency have severe impairment of IL-1 and TLR downstream signaling and are susceptible to recurrent bacterial infections.24,25 TRIF signaling results in the activation of IFN-regulatory factor (IRF) 3 and induction of type 1 IFN genes such as IFN-a and IFN-b,26 helpful for viral clearance. TLR polymorphism may be linked to diseases such as asthma27 and atherosclerosis,19 TLR-2 has been linked to different responses to ischemia and reperfusion injury after solid-organ transplantation27; and a deletion of the signaling domain of TLR-5 has been found to increase susceptibility to legionnaire disease.29 Finally, Toll signaling pathways have been implicated in the pathogenesis of sepsis and shock30–32 (see Chapter 12, The Systemic Inflammatory Response Syndrome (SIRS), Sepsis and Septic Shock, for review).

Phagocytes Major phagocytic cells are neutrophils and macrophages. In humans, myeloid precursors are found in the yolk sac by day 19 of development, 2 days before the onset of blood circulation. Hematopoiesis then shifts to the fetal liver and finally to the bone marrow. In the bone marrow phagocyte development is under the control of multiple growth factors, including IL-3, granulocyte–macrophage colonystimulating factor (GM-CSF), granulocyte colony-stimulating factor (G-CSF), and macrophage colony-stimulating factor (M-CSF). The marrow pool of neutrophils in adults is 20 times the number of neutrophils in circulation. The mechanisms responsible for the release of neutrophils from the marrow are not well understood. Once in the circulation neutrophils circulate for a few hours and then move into tissues where they are active for 2 to 6 days. At term, the neonatal peripheral neutrophil count is higher than that of adults, but there appears to be little reserved capacity to respond with an outpouring of phagocytic cells which are often immature, during infection. Newborn infants with septicemia often have severe neutropenia and depletion of phagocytic storage pools, a finding associated with a high mortality rate.34–35 The cause of depletion

remains unknown; it often occurs in the presence of increased numbers of neutrophil precursors and elevated levels of cerebrospinal fluid in blood.35 Perhaps a decreased number of neutrophils at the site of infection contribute to the susceptibility of neonates to pneumonia and skin infection and to the development of multiple sites of infection after bacterial or fungal bloodstream infection.36 This lack of adequate numbers of cells at the site of infection may cause or result in functional deficiencies. Monocytes also move from the circulation to tissue spaces, where they develop into macrophages and live for 2 to 3 months, assuming specialized characteristics most determined by their location (e.g., lung, liver, or spleen). Circulating monocytes also have chemotactic and phagocytic activities and have receptors for immunoglobulin G (IgG) Fc receptor domains (FcR) and the complement complex iC3b.37 The function of phagocytes (which are particularly important in defense against bacteria and fungi) requires not only sufficient numbers of cells but adequate ability to sense and migrate toward the site of infection (chemotaxis) and to ingest and kill (phagocytosis) microorganisms. These processes are mediated by the expression of adhesion molecules, opsonins (complement and antibodies), and release of toxic substances (Figure 10-1).

Chemotaxis As a result of a local inflammatory response, endothelial cells within the local vessels express adhesion molecules called selectins (CD62E, CD62P). These molecules reversibly bind to ligands on neutrophils (sialyl-Lewis X and PSGL1) and consequently make the neutrophil slow down and “roll” along the endothelium. Subsequently, another group of adhesion molecules, called integrins, are upregulated on the surface of neutrophils. Integrins are composed of one of three different alpha chains: CD11a, CD11b (CR3), or CD11c (CR4); they are noncovalently linked to a beta chain, CD18, thus forming CD11a/CD18 or LFX1, CD11b/CD18 or MAC-1, and CD11c/CD18 or p150,95 integrin complex. Integrin molecules “stop” the neutrophil, which then undergoes skeletal changes and migrates through the vascular lumen into the extravascular space by adhering to intracellular adhesion molecules (ICAMs). Interestingly, the sialic acid constituent of the group B streptococcus (GBS) capsular polysaccharide mimics the

Blood vessel

Bone marrow

Tissue

Endothelium FcR CD11a,b,c/CD18 Integrins

Site

Selectins

Ig

Sialyl Lewis X

Function

Effectors

Phagocyte production *(reserve) Phagocyte maturation

G-CSF Stem cell factor M-CSF IL-3 IL5

CR1 C3b

C5a IL-8 Leukotrienes

Adhesion molecules and ligands Phagocytes: Endothelium: Sialyl-Lewis X E–Selectin; P-Selectin CD11a/CD18 (LFA1) ICAM-1, -2, -3 CD11b/CD18 (Mac1) ICAM-1 (CD54); iC3b CD11c/CD18 (CR4; p150/95) Fibrinogen

Bacteria

O– H2O2

Chemoattractants

*Opsonization *Chemotaxis *Phagocytosis Killing

*C5a *IL-8 Leukotrienes

*Antibody *Complement Superoxide Hydrogen peroxide

*Fibronectin *Actin polimerization *CR3

Figure 10-1. Aspects of immunologic function. *Indicates aspects that are immature or defective in the neonatal period. C, complement; CD, cluster of differentiation; CR, complement receptor; G-CSF, granulocyte colony-stimulating factor; Ig, immunoglobulin; IL, interleukin; M-CSF, macrophage colony-stimulating factor. PART I Understanding, Controlling, and Preventing Infectious Diseases

Immunologic Development and Susceptibility to Infection

human Lewis X antigen, making this a poor immunogen and perhaps rendering the neonate more susceptible to this organism.38 Defects in the expression of adhesion molecules have been described in humans. Leukocyte adhesion defects (LADs) include lack of integrin expression (LAD I), lack of sialyl-Lewis X expression (LAD II) and defects of integrin activation (LAD III). Affected children have persistent leukocytosis, delayed separation of the umbilical cord, skin ulcers, periodontitis, and delayed wound healing.39 L-selectin (CD62L) levels on fetal and immature infant neutrophils are comparable with those of adults. However, their expression is downregulated in the term neonate and is further diminished during acute bacterial infection in vivo.40 Other defects in chemotaxis have been described.39,41–44 In the newborn infant defects in chemotaxis have been linked to decreased expression of Rac2, a signaling molecule, on cord blood neutrophils.36

Phagocytosis Phagocytosis is an active process by which a previously bound pathogen is engulfed by a phagocyte in a membrane-bound vesicle called a phagosome. Both macrophages and neutrophils contain lysosomes, which are membrane-bound acidic organelles containing proteolytic enzymes, and are capable of producing toxic products: nitric oxide (NO), superoxide anion (O2–) and hydrogen peroxide (H2O2). Fusion of the lysosome and phagosome membranes is necessary for killing of the organism. Within azurophilic granules, a bactericidal/permeability-increasing protein (BPI) binds to bacterial LPS and kills gram-negative bacteria. Once bacteria are killed, neutrophils die whereas macrophages are capable of generating new lysosomes. Abnormal BPI function has been implicated in both neonatal sepsis45 and chronic Pseudomonas infection in CF patients.46 Superoxide and H2O2 production is dependent on the NADPH oxidase enzyme complex (see Chapter 106, Infectious Complications of Dysfunction or Deficiency of Polymorphonuclear and Mononuclear Phagocytes). Defects in the different components of this enzyme result in the immunodeficiency, chronic granulomatous disease. Affected patients are susceptible to infections with catalase-positive organisms, Aspergillus and Nocardia species. Because phagocytes are unable to kill the microbes, the host tries to contain the infection by calling in more macrophages and lymphocytes, resulting in granuloma formation. There are no well-described phagocytic defects in the developing human embryo. The capability of the newborn for nonopsonic adherence to organisms and phagocytosis is nearly equal to that of adult cells. However, deficiencies in chemotaxis and superoxide production have been described.38,47–49 Bacterial killing by cord blood phagocytes is effective against Escherichia coli and Streptococcus pyogenes and is similar to adults, but killing appears abrogated for GBS.50 Abnormalities in chemoattractants (IL-8, complement fragment C5a, fibronectin)51,52 and defective expression of complement receptors, such as C5a receptor, caused by C5a-mediated exocytosis of myeloperoxidase,53 are also described. Defects in membrane fluidity and cytoskeletal changes may also contribute to defects in neutrophil motility.54 Intrapartum administration of magnesium sulfate has been reported to decrease neutrophil motility and phagocytosis of cord blood neutrophils, as measured by chemotaxis, random motility, and chemiluminescence.55

Dendritic Cells In humans, CD34+ hematopoietic stem cells (HSC) capable of generating DC are detected in the fetal liver at 20 weeks’ gestation, after which they are mainly found in the bone marrow. After birth, 1% to 3% of cord blood cells express CD34. During differentiation these CD34+ cells lose CD34 expression and express CD4, CD45RA, IL-3R, and major histocompatibility complex (MHC) class II antigens.56,57 A class of DC called Langerhans cells (LC) were first described by Paul Langerhans in 1868. It is difficult to identify lineage ontogeny of DC

CHAPTER

10

89

in humans. Bone marrow differentiation studies of DC suggest a dual origin of DC in myeloid and lymphoid cells, but debate remains.58,59 LC can derive from blood DC.60 Although lineage-specific markers are still being defined, three transcription factors have been shown to regulate their development: PU.1, RelB, and Ikaros. PU.1 is important for myeloid-derived DC,61 RelB is associated with DC activation,62 and ikaros proteins are transcriptional activators and influence chromatin remodeling and histone deacetylation.63 DC are prototypic APCs and are capable of regulating both innate and adaptive immune responses. When activated, they have unique morphologic characteristics. Several pathogen receptors have been identified: TLRs, which appear to be involved in DC maturation, and scavenger receptors, which mediate bacterial internalization. MAC-1 (CD11b/CD18) or the CR3 complement receptor have demonstrable phagocytosis of complement-coated bacteria.64 In the skin LC are localized to the basal and suprabasal layers of the epidermis; in the murine gut, DC are found in the Peyer patches; and in the human lungs, they can be found within the airway epithelium, alveolar septa, visceral pleura, and vascular wall.65–67 While surveying the tissue environment, DC can be recognized by an “immature” phenotype (CD11bbright, CD11cmod, CD86low, class IIlow, CD4–). Upon uptake of antigen/microbial products by different mechanisms of phagocytosis, they migrate via the afferent lymphatics to the regional lymph node where they arrive as mature nonphagocytic DC (CD40high). These DC cells can produce inflammatory cytokines and chemokines. Bacterial uptake of Mycobacterium tuberculosis, bacille Calmette-Guerin (BCG), Saccharomyces cerevisiae, Corynebacterium parvum, Staphylococcus aureus, Leishmania spp. and Borrelia burgdorferi has been demonstrated in vitro68–71 (reviewed in reference 63). Once in the lymph nodes, DC move around from the marginal zones to the T zones until they encounter a naive T lymphocyte.72 When the naive T lymphocyte and the DC meet, they remain stuck together for a while, forming the important immunologic functional unit called the immunologic synapse.73

Complement The complement system comprises a series of serum proteins that function in host defense as an enzymatic cascade (see Chapter 105, Infectious Complications of the Complement System). When microorganisms invade the host, activation of complement occurs locally by one of three pathways that converge at the stage of the formation of an enzyme called C3 convertase. This enzyme cleaves complement component C3 into C3b and C3a. C3b, the major effector molecule of the complement system, binds to the bacterial cell membrane. This important molecule functions as an opsonin (to facilitate phagocytosis) and also helps cleave C5 into C5a and C5b. C5a is a potent chemoattractant, and C5b is an integral part of the membrane attack complex along with other terminal components: C6, C7, and C9. One pathway (classical) is activated by antibody–antigen complexes and thus depends on and enhances specific humoral immunity. A second pathway (alternative) can be activated by direct binding to the surface of some microorganisms and thus functions to provide innate (nonspecific) immunity. A third pathway (MBL pathway) is initiated by the binding of MBL on mannose and fucose-containing surfaces of bacteria and viruses favoring their phagocytosis. Until 18 months of age, the concentration of most complement proteins is lower than that of adults, with the exception of C7. Between 28 and 33 weeks of gestation, there appears to be little development of the complement system. Levels of C8 and C9 are the most markedly reduced at all gestational ages. Levels correlate with gestational age, but not with birthweight, type of delivery, or sex.74 Deficiencies of complement activation in both classical and alternative pathways have been described (see Chapter 105, Infectious Complications of the Complement System).75,76 Low levels of total hemolytic complement activity are a significant predictor of mortality in neonates with septicemia.76 The molecular basis for defects in complement function in neonates and the details of the consequences are only partially under-

90

SECTION

C Host Defenses Against Infectious Diseases

stood. For example, a possible defect has been described in formation of a reactive thioester bond on C3 that is essential for opsonic and covalent binding of C3b to bacteria.77 Inefficient killing of E. coli by neonatal sera appears to correlate with low concentrations of C9.78 and can be overcome by adding C9 to ampicillin-treated serum from neonates in vitro.79 Finally, deficient formation of C5a may also increase the newborn infant’s risk of infection. Although levels of C5 are similar in adult and neonatal sera, neonatal sera form significantly less C5a on exposure to type III GBS.80 This deficiency was apparent in newborn sera with antibody levels similar to those of adults and could be corrected by in vitro addition of C3. Hemolytic–uremic syndrome (HUS) occurs in childhood and is frequently preceded by a diarrheal illness caused by E. coli O157:H7. Plasma protein factor H and plasma serine protease I, regulatory proteins of the alternative complement pathway, have been associated with the atypical form of HUS (aHUS).81,82 A study of 120 patients with aHUS found that 10% of patients had mutations in the membrane cofactor protein (MCP; CD46). The onset was typically in early childhood; most did not develop endstage renal failure.83

Natural Killer Cells NK cells are a subgroup of lymphocytes that exhibit cytolytic activity against tumor cells or cells infected with viruses (Table 10-1). In humans, NK cells are similar to T lymphocytes in their effector function, but lack the T-lymphocyte receptor/CD3 complex and express the low-affinity Fc receptor for IgG (CD16, FcgRIII). They comprise up to 10% of peripheral blood lymphocytes (PBL) in adults. NK cells appear in substantial numbers by 6 to 9 weeks of gestation in the embryonic liver and later in the fetal liver, thymus, and spleen. After birth, NK cells primarily develop from HSC in the marrow and are driven to maturity by cytokines, in particular IL-15. IL-15 has the same intracellular signaling molecule, the common gamma chain (gC), as other cytokines (IL-2, IL-4, IL-7, IL9, and IL21). Mutations in gC are responsible for a form of severe combined immunodeficiency (SCID) that lacks both T and NK cells (see Chapter 11, Fever and the Inflammatory Response). Unlike cytotoxic T lymphocytes (CTL), NK cells do not require MHC class I antigens to recognize their targets, do not recognize particular viral antigens,84 and can be activated by cytokines without previous exposure to the antigen – making this cell an important contributor to innate responses. However, NK cells can function with some degree of antigenic specificity because they can lyse cells that

are coated with specific antibody molecules. This process is called antibody-dependent cell-mediated cytotoxicity (ADCC). Both NK and CTLs mediate cytolysis in a similar manner as these two cell populations contain granules composed of cytolytic proteins called perforin and enzymes called granzymes. Perforin creates pores in target cell membranes; granzyme introduced through these pores induces target cell apoptosis.85 Defects in the vesicle membrane fusion, perforin, and granzyme have been described in patients with hemophagocytosis (see Chapter 14, Hemophagocytic Lymphohistiocytosis and Macrophage Activation Syndrome). A subset of T lymphocytes that coexpress markers associated with NK cells are termed NKT cells. They are defined by surface markers and functionality, express CD1d, use a limited T-cell receptor (TCR) repertoire, and upon stimulation secrete large amounts of IFN-g and IL-4. When deficient, development of both autoimmunity and tumors is enhanced (reviewed in reference 86). NK cell activity in cord blood from infants born between 32 and 36 weeks of gestation is low compared with activity in infants born after 36 weeks. Antenatal glucocorticoid therapy for preterm labor can accelerate maturation of NK cells.87 Compared with adult NK cells, neonatal NK cells have decreased cytotoxic activity88 and diminished ADCC87 until at least 6 months of age. On the other hand, when IFNs, IL-2, IL-12, and IL-15 are added to cultured NK cells, they are capable of responding similarly to adult cells.84 Neonatal HSV infection and severe recurrent herpesvirus infections in adults provide evidence of the importance of NK cells in host defense.89,90 Human umbilical cord blood cells demonstrate defective NK cell cytotoxicity and ADCC against HSV-infected targets.91–93 Evidence for the relevance of ADCC in protecting the infant from HSV is provided by the association of high levels of maternal- or neonatal-specific ADCC with HSV, or high levels of neonatal HSV-neutralizing antibodies with absence of disseminated HSV infection in infants.94 Term infants infected perinatally with HIV are deficient in ADCC, whereas preterm infants are deficient in both NK cell cytotoxicity and ADCC against HIV-infected targets. These observations may relate to an increased risk of HIV transmission in preterm neonates. Low NK cell cytotoxicity and ADCC of newborn lymphocytes to HIV-infected cells may further explain the newborn’s inability to reduce plasma levels of HIV after acute infection.95 Recently, NK defects have been recognized as either part of a broad immunodeficiency syndrome or as isolated defects within NK cell populations.96

THE ADAPTIVE IMMUNE RESPONSE TABLE 10-1. Comparison of Natural Killer Cells and Cytolytic T Lymphocytes Characteristic

Natural Killer Cell

Cytolytic T Lymphocyte

Identification of target for kill

Nonspecific killing of virus-infected cells or Binding to antibodycoated cells via CD16 (ADCC)

TCR specifically identifies viral peptide complexed with MHC class I molecule on surface of infected cell

Surface markers

CD16 and/or CD56

CD3/CD8

Signal(s) for activation

IFN-a, IFN-b, IL-12

IL-2 and antigen (on surface of antigenpresenting cell)

Onset of function

1–6 days after infection

5–10 days after infection

Mechanism of killing

Granule exocytosis and secretion of toxin

Granule exocytosis and secretion of toxin

ADCC, antibody-dependent cell-mediated cytotoxicity; IFN, interferon; IL, interleukin; MHC, major histocompatibility complex; TCR, T-cell receptor.

The adaptive immune response is essential for host defense, as shown by the primary immunodeficiency syndromes outlined in Chapter 103, Evaluation of the Child with Suspected Immunodeficiency. Specificity and immunologic memory are the two most important consequences of adaptive immunity. Specificity is determined by the vast range of molecular diversity of the antigen receptor. Immunologic memory is the ability to respond rapidly and effectively to pathogens previously encountered. This mechanism implies the pre-existence of clonally expanded populations of antigen-specific lymphocytes. The effector cells of the adaptive immune response are T and B lymphocytes. These cells derive from a common HSC (Figure 10-2). It is generally accepted that HSC differentiate into a common lymphoid progenitor (CLP) that will give rise to T, B, and NK cells. Lymphopoiesis is a tightly regulated sequence of events that leads to the expression of a functional antigen receptor on the surface of the lymphocyte. For the B lymphocyte it is the immunoglobulin molecule and for the T lymphocyte, the TCR complex. Cellular microenvironment, growth factors, cytokines, and chemokines, along with silencing or activation of certain genes at different stages of lineage commitment, are some of the multiple factors contributing to a successful and mature lymphocyte. Although mouse lymphoid development has been well elucidated, human lymphoid development has been fragmented, as our knowledge is patched together from in vitro

PART I Understanding, Controlling, and Preventing Infectious Diseases

Immunologic Development and Susceptibility to Infection

CHAPTER

10

91

Figure 10-2. Developmental stages of T, B, and natural killer (NK) lymphocytes. B, B lymphocyte; CD, cluster of differentiation; CLP, common lymphoid precursor; CSR, class switch recombination; HSC, hematopoietic stem cell; IG, immunoglobulin; MHC, major histocompatibility complex; T, T lymphocyte.

HSC CD34+

CLP

T lineage

B lineage

NK lineage

Cortex

Thymus

Pro B CD34+ CD45+ MHCII CD10 CD19

Double Positive CD4+/CD8+

Pre B MHCII CD10 CD19 CD20

μ

Medulla

TCR

Double negative CD4–/CD8–

IL-15

sIgM CD4+

Single Positive

CD8+

IgM CD4+

CD8+ IgD CSR Memory B IgG IgA IgE

Inmature B CD45 MHCII CD10 CD19 CD20 Mature B CD45 MHCII CD19 CD20 CD21 CD22

Mature NK

Plasma cell CD135 CD38

analysis of bone marrow-derived cells, analysis of patients, and comparisons with mouse models (reviewed in reference 96).

B Lymphocytes B lymphocytes play a critical role in pathogen-specific immunity by producing antibodies. B lymphocytes recognize soluble antigens via immunoglobulins anchored on their surface and differentiate into antibody-producing cells, called plasma cells, capable of secreting immunoglobulins. These proteins function alone (neutralization) or

with complement or phagocytes to inactivate microorganisms (see Chapter 104, Infectious Complications of Antibody Deficiency). The B-lymphocyte system is fully developed at birth. The origin of the human B lymphocyte is not well defined but fetal B lymphocyte can be recognized in the yolk sac, omentum, and fetal liver.97,98 After birth B-lymphocyte development takes place in the bone marrow. The ordered steps of B-lymphocyte development are marked by a rearrangement of the heavy chain first and then the light chain variable region genes of the Ig molecule. From a lymphoid progenitor, B lymphocytes mature following a sequence from pro-B ˜ pre-B ˜ immature B lymphocyte ˜ mature B lymphocyte ˜ plasma cell (see

92

SECTION

C Host Defenses Against Infectious Diseases

Figure 10-2). Transcriptional factor PU.1, E2A, and early B lymphocyte factor (EBF) are essential for the early stages. Intracellular proteins such as recombinase-activating gene (RAG) 1 and 2 are identified during the pro and pre-B lymphocyte stage. Surface IgM appears at the immature B stage and the mature B lymphocyte carries both IgD and IgM. Surface markers can further identify these cells using flow cytometry analysis and are commonly used in clinical practice. A mature B lymphocyte can be identified by CD19, CD20, CD21 (the Epstein–Barr virus receptor) and CD40 (ligand for CD154, defective in hyper-IgM syndrome), amongst others. All B-lymphocyte surface markers are lost when the cell reaches the plasma cell state. Thus the monoclonal antibody antiCD20 used for management of B-cell CD20+ lymphomas will not interfere with the plasma cell pool.

Passively Acquired Antibodies Neonates have a limited serum antibody repertoire (at least 1011) of actively formed (self-produced) antibodies as they have not had the opportunity to encounter pathogens. This is alleviated by transplacental transfer of maternal antibody, which occurs by a selective, active process. Only IgG antibodies appear in umbilical cord blood, and some IgG subclasses are transferred better than others.99,100 This mechanism may be related to differential binding of FcRII isoforms in the placenta.101,102 At birth, full-term infants have serum IgG levels that are equal to the maternal level or exceed it by 5% to 10%.103 Most antibody transfer occurs during the third trimester of pregnancy and, therefore, preterm infants may have substantially low levels. For example, most infants born at 32 weeks of gestation or earlier have serum IgG levels below 400 mg/dL.104 Transplacentally acquired antibody disappears rapidly after birth, reaching a nadir at 3 months (Table 10-2). Average concentrations at this time are 60 mg/dL for infants born at 25 to 28 weeks of gestation, 104 mg/dL for infants born at 29 to 32 weeks, and 430 mg/dL for infants born at term.105.106 However, these low levels of maternal IgG antibody appear to be associated with less risk of infection than might be expected. A longitudinal study of the ability of preterm infants to form specific antibodies provides a partial explanation for this low risk.107 By about 6 months of age, infants have formed specific IgG antibodies in response to immunization with diphtheria, tetanus toxoids, and pertussis vaccine. Also by about this age, their antibodies have functional opsonic activity against E. coli and coagulase-negative staphylococci. The maternal repertoire of specific antibodies is critical for protection of the newborn infant from commonly encountered pathogens.108 Baker & Kasper109 demonstrated that a concentration of 2 μg/mL or less of serum antibody to type III GBS in cord blood correlated with susceptibility to invasive disease. Evidence of the protective effect of maternal antibodies against HSV94,104 and varicella-zoster virus infection is well established.111 Many gram-negative

TABLE 10-2. Concentration of Serum Immunoglobulin G (IgG) in Term and Premature Infants Gestational Age: Mean Serum IgG (mg/dL) Postnatal Age 1 week

25–28 weeksa

29–32 weeksa

Termb

organisms require IgM antibodies and complement for efficient opsonization.112,113 Since IgM does not cross the placenta, neonatal sera opsonize these organisms poorly.107 The apparent protective role of transplacental IgG antibodies against GBS infection has led to extensive attempts to prevent or treat such infection with passively administered antibody or by immunization of women with GBS vaccines114 (see Chapter 119, Streptococcus agalactiae (Group B Streptococcus)).

Active Production of Antibodies The ability of newborn infants to produce an active antibody response to antigenic stimulation develops in an orderly fashion. An adult pattern of antibody responses is not acquired until 4 to 5 years of age. Analysis of the factors responsible for this developmental pattern is complex because production of antibody depends not only on Blymphocyte maturity but also on interactions with other cells that mature at different rates (Table 10-3). After birth, active production of IgG slowly increases, with differences among IgG subclasses; IgG1 and IgG3 production matures before IgG2 and IgG4.99 The last isotype to achieve adult concentrations is IgA.115–117 Postmortem studies of fetuses suggest that secretory IgA (sIgA)-containing epithelial cells appear at 20 to 21 weeks of gestation and their number increases from 2.5 cells per 10 000 mm at 23 to 26 weeks, and to 8 cells per 10 000 mm at 36 to 40 weeks. In fetuses with pneumonia or sepsis, the number of sIgA-containing epithelial cells in the trachea, bronchi, and intrahepatic duct decreases and, at times, completely disappears. This suggests that sIgA is an important component of mucosal immunity at as early as 20 weeks of gestation.117,118 Fetuses can produce serum antibodies, IgM, predominantly in response to intrauterine infection.119 Preterm infants respond nearly as well as term infants to immunization beginning at 2 months with diphtheria and tetanus toxoids and the acellular pertussis (DTaP), poliovirus, and hepatitis B vaccines.120–122 Term infants immunized or infected during the first few days of life usually produce protective antibody responses, although at somewhat lower levels than adults.123–125 The presence of fetal bone marrow B-lymphocyte pools similar in size to those of adults and with comparable isotype diversity suggests that their functional deficits may reflect a developing memory repertoire with increased specificity to antigen or the need to generate T lymphocytes capable of producing strong B-lymphocyte signals, rather than inherent B-lymphocyte immaturity as a sole factor. Studies of B-lymphocyte activation have revealed much about the

TABLE 10-3. Developmental Milestones of Humoral Immunity Event

Age

Surface-positive B lymphocytes of all isotypes present in liver

16 weeks of gestation

Surface-positive B lymphocytes of all isotypes present in bone marrow

22 weeks of gestation

Stimulated B lymphocytes secrete primarily IgM

Fetus–newborn

Production of antibody in response to protein antigens

Fetus–newborn

Serum IgG reaches 60% of adult levels

1 year

251

368

1031

3 months

60

104

430

Production of antibody in response to polysaccharide antigens

2–3 years

6 months

159

179

427

Stimulated B lymphocytes secrete all isotypes

2–5 years

Serum IgA reaches 60% of adult levels

6–8 years

a

Data for premature infants based on samples at 1 week, 3 months, and 6 months of age.105 b Data for term infants based on cord blood and samples at 1–3 months and 4–6 months of age.106

Ig, immunoglobulin.

PART I Understanding, Controlling, and Preventing Infectious Diseases

Immunologic Development and Susceptibility to Infection

complexities of B-lymphocyte signaling126 and suggest possible mechanisms for deficits of B-lymphocyte function that can be examined in healthy neonates.127 In contrast to the early development of antibody responses to most protein antigens, responses to thymus (T)-independent antigens, such as polysaccharides, develop much later. The basis for this remains unclear.128,129 Cord blood B lymphocytes are capable of activation in vitro with T-independent activators. This suggests that signal transduction pathways within the B lymphocyte are normal.130 However, the mechanism of B-lymphocyte activation by polysaccharides differs from protein antigens in that it involves co-stimulation through a Blymphocyte surface molecule called CD21 and may be influenced by the expression of yet another B-lymphocyte surface molecule, CD22.128 Both CD21 and CD22 were noted to be reduced on neonatal B lymphocytes upon stimulation, suggesting a unique role for these molecules in antibody production to polysaccharide antigens.128,131 Although the extent to which neonatal deficiencies of neutrophil function, complement, or antibody contribute to the increased risk of infection is unknown, these factors are important in vitro to the opsonophagocytic killing of E. coli, GBS, and Candida species. Thus the combined effect of these deficiencies no doubt contributes to the increased risk of serious infection with these pathogens in this group of children.108

T Lymphocytes T lymphocytes play a central role in the regulation of antigen-specific immune responses, modulating the function of APCs, B lymphocytes, and other T lymphocytes both through contact with receptor binding and secretion of cytokines.126 T lymphocytes are effector cells of cellmediated immune response and function as cytotoxic cells (CTLs), able to kill target cells that express foreign antigens. Most mature T lymphocytes (95%) recognize antigen through a TCR. In contrast to B lymphocytes, T lymphocytes can only recognize antigens that are displayed on cell surfaces. These antigens can be derived from pathogens that replicate within the cell, such as viruses or intracellular bacteria, or products internalized from the extracellular space. The reason T lymphocytes can recognize intracellular pathogens is that these infected cells display on its surface fragments/peptides derived from the pathogens’ proteins. The molecules responsible for holding these peptides within their groove are the MHC molecules. Two classes of MHC molecules are recognized: class I, which include human leukocyte antigen (HLA)-A, -B, and –C, and MHC class II molecules, HLA-DR, -DQ, and -DP. MHC class I are present on all nucleated cells, whereas MCH class II are present on APCs and other specialized cells. Peptide MCH class I complexes (MHC I) present antigen to CTL (CD8+). Peptide MHC class II complexes (MHC II) present to helper T lymphocytes (Th1/CD4+). Cytosolic peptides derived from vaccinia virus, influenza virus, rabies virus and Listeria are copled to MHC I molecules on the surface of the infected cell. These peptide-MCH I complexes can be recognized by cytotoxic CD8+ cells, which then kill the targeted infected cell. Mycobacterium tuberculosis, M. leprae, Leishmania donovani, and Pneumocystis carinii are localized within macrophage vesicles. Peptide–MHC II complex as on the infected macrophage result in IL-12 production and subsequent activation of CD4+ cells (Th1). These then produce IFN-g which recruits more macrophages, a granuloma is formed, and the activated macrophage kills their intracellular pathogen by releasing the antimicrobial products within its vesicles. Clostridium tetani, Staphylococcus aureus, Streptococcus pneumoniae, polio viruses, Pneumocystis carinii, and Trichinella spiralis require both a humoral and a cellular immune response for effective killing.16 It is therefore understandable that defects in antigen processing, presentation, and both T- and B-effector functions make the individual susceptible to pathogens that require a particular pathway for clearance (see Section R). Sir Peter Medawar first recognized that antigen presentation in fetal life leads to a form of immunologic tolerance that was different

CHAPTER

10

93

than in adult life.132 This mechanism was postulated to be due to a state of immunologic naiveté of the fetus. However, an alternative hypothesis suggests that perhaps certain adaptive immune responses may be actively suppressed in the developing fetus, mediated by an important population of T lymphocytes called regulatory T lymphocytes.133 This theory is favored by the fact that that fetuses are capable of generating specific adaptive responses after transplacental spread of infectious agents.134–136 HSC progenitors migrate to the thymus where, similar to B lymphocytes, T lymphocytes rearrange their TCR genes by somatic recombination. In contrast to B cell receptors, TCR do not diversify their variable (V) region genes through somatic hypermutation. Also, unlike the B lymphocytes, T lymphocytes must develop into two populations – ab T lymphocytes (the most abundant, 95% of the circulating pool) or gd T lymphocytes (approximately 5% of the circulating pool). Maturing T lymphocytes move within the thymus in a directed manner from the cortex to the medulla. During this maturational process, T lymphocytes express CD4 and CD8 molecules on their surface (see Figure 10-2). The expression of these molecules can further serve to identify their state of maturity. Thus T lymphocytes start as the double-negative (CD4–/CD8–) stage ˜ doublepositive (CD4+/CD8+)˜ single-positive CD4+ or CD8+ just before being released to the periphery. During maturation in the thymus, T lymphocytes are selected to proceed to the next developmental stage thanks to the interactions of a successfully assembled TCR with selfMHC–self-peptide complexes. As in B lymphocytes, RAG1/2 genes and their products are identified during early stages of development and the expression of transcription factors and signaling events drives clonal commitment.137–140 The ontogeny of T-lymphocyte immunity in the neonate has been reviewed (Table 10-4).141,142 Cord blood has an increased absolute number of T lymphocytes compared with the peripheral blood of older children and adults. The mean T-lymphocyte counts in newborn infants, children, and adults are 3100/mm3, 2500/mm3, and 1400/mm3, respectively. The ratio of CD4+ to CD8+ cells in cord blood is the same as that in adults (1.2), but it is increased to 1.9 from birth through 11 months of life.143,144 Although the absolute number of T lymphocytes decreases beyond the neonatal period, the percentage of T lymphocytes among the total lymphocyte population increases.

Proliferative Responses Neonatal T lymphocytes proliferate normally in response to phytohemagglutinin and allogenic cell stimulation but have a limited ability to develop immunologic memory.145–147 Cord blood cell populations

TABLE 10-4. Comparison of Newborn and Adult T Lymphocytes Characteristics

Newborn

Repertoire of TCR-binding specificity

Unknown

Mean T-lymphocyte count

3100/mm

Adult

3

Broad 1400/mm3

CD4+/CD8+ (ratio)

1.2

1.2

Proliferation (mitogen-stimulated)

Good

Good

Proliferation (antigen-stimulated)

Poor

Good

Ability to provide help to B lymphocytes

Poor

Good

CD45RA (naive CD4 T lymphocytes)

90%

48%

Production of cytokines

Decreased IFN-g, Multiple IL-4, G-CSF, GM-CSF, IL-3

+

+

G-CSF, granulocyte colony-stimulating factor; GM-CSF, granulocyte–macrophage colony-stimulating factor; IFN-g, interferon-g; IL-4, interleukin-4; TCR, T-cell receptor.

94

SECTION

C Host Defenses Against Infectious Diseases

have been shown to contain large numbers of naive T lymphocytes (CD45RA+) versus memory T lymphocytes (CD45RO+).147,148 This proportion declines slowly to 61% by 7 years of age, as naive lymphocytes are replaced by memory lymphocytes, which probably represents an intrinsic maturation of T lymphocytes as a consequence of antigenic stimulation.140,149

CHAPTER

11

Fever and the Inflammatory Response

Cytokine Production The predominant naive phenotype of newborn T lymphocytes may further account for differences in cytokine production. Neonatal T lymphocytes produce less INF-g, IL-2, IL-4, IL-10, and TNF-a than adult T lymphocytes in response to various stimuli. A modest reduction of GM-CSF and decreased G-CSF and IL-3 production and gene expression have also been described.150–154 Dysregulation of various immunoregulatory and cytokine genes in cord blood mononuclear cells could explain the apparent immaturity of neonatal cell-mediated immunity.155 A group of soluble polypeptide mediators, called chemokines, are also important in immune surveillance. Chemokines regulate leukocyte chemotaxis, inflammation, antitumor activity, and HIV infection in humans. Placental cord blood mononuclear cells, in comparison to adult peripheral blood mononuclear cells, produce smaller amounts of the chemokine called RANTES (regulated upon activation, T lymphocyte expressed). Since RANTES is a known ligand for CCR5, it suggests a role of this chemokine in HIV pathogenesis.156,157 T lymphocytes influence the functional activity of many other cell types responsible for both natural and specific immunity. Decreased Tlymphocyte function in neonates is likely to increase their susceptibility to infection by many pathogens. For example, decreased T-lymphocyte help in antibody production and isotype switching via CD40L–CD40 interaction,158 along with decreased phagocytic function, probably contribute to increased susceptibility to bacterial infection. Susceptibility to viral infection and other intracellular pathogens, such as Toxoplasma gondii and Listeria, probably results from decreased cytolytic activity of T lymphocytes and decreased IFN-g.150 The specific role of T-lymphocyte immaturity in severe clinical HSV infection in neonates is suggested by the observation that neonates infected with HSV showed decreased antigen-specific cellular responses (decreased proliferation and IFN-g production) compared with adults who had primary HSV infection.159 Finally, NK cell cytotoxicity and ADCC, along with defective chemokine production, may be contributors to perinatally acquired HIV infection.

INTERRELATIONSHIPS AND FUTURE Humans resist infection in several ways. The innate defense mechanisms act first and may be capable of eliminating the pathogen completely. If not, adaptive responses are initiated and put in place, releasing clonally expanded effector T and B cells to the sites of infection. The mechanisms that regulate the final clearance are dependent on the pathogen itself. For certain pathogens, an effective adaptive immune response leads to a state of protective immunity. However, many pathogens evolve mechanisms that permit their evasion from an effective immune response. Since Edward Jenner’s studies of cowpox 200 years ago, vaccination has become a successful application of our interpretation of nature’s experiments. Furthermore, the study of patients with primary immunodeficiency provides not only a better understanding of biologic systems as they relate to humans but potentially leads to therapies not only for these groups of patients but for others as well.

Alexei A. Grom

FEVER Genesis Fever has been recognized as an important manifestation of childhood infections since ancient times. Fever is often the first symptom noted by parents signaling that their child is ill. Fever is defined as a centrally mediated rise of body temperature above the normal daily variation in response to many different pathologic insults.1 A variety of microbial products, including endotoxins and exotoxins, are exogenous pyrogens. Although these molecules can act directly to induce fever, evidence indicates that they stimulate host cells to secrete mediators known as endogenous pyrogens. The pivotal endogenous pyrogens are cytokines produced during the inflammatory response, most notably tumor necrosis factor-a (TNF-a), interleukin1 (IL-1), IL-6, and, to a lesser degree, the interferons (IFNs).2–4 Fever is more likely to be caused by infection, but any inflammatory, neoplastic, immunologic, or traumatic event can also generate fever. Once these pyrogenic cytokines are produced, they enter the systemic circulation and stimulate the rich vascular network surrounding the preoptic area of the hypothalamus (thermoregulatory center). Here they activate phospholipase A2, liberating plasma membrane arachidonic acid as a substrate for the cyclooxygenase pathway.5,6 Some cytokines can increase cyclooxygenase expression directly, causing liberation of the arachidonate metabolite prostaglandin (PG) E2. Because this small lipid molecule easily diffuses across the blood–brain barrier, it is believed by some to be the local mediator that activates thermoregulatory neurons which in turn raise the thermostat set point (Figure 11-1). Although the blood–brain barrier prevents migration of large proteins such as circulating cytokines, at certain sites the presence of the pyrogenic substances has been demonstrated; these are known as circumventricular organs, and lack a blood–brain barrier.6 Subsequently, peripheral mechanisms are activated that stimulate the sympathetic chain and terminal adrenergic efferent nerves leading to vasoconstriction (heat conservation) and muscle contraction (heat production), which generate fever. Additionally, autonomic (decreased sweating) and endocrine (decreased vasopressin secretion to reduce amount of body fluid to be heated) pathways contribute to thermoregulation.6 Conservation and production of heat continue until the temperature of the blood bathing the hypothalamic neurons matches the new setting. When cytokine stimulation ceases, the hypothalamic set point is revised downward and the processes of heat loss through vasodilatation and sweating are initiated. In addition to these thermoregulatory mechanisms, certain areas in the cerebral cortex are stimulated to promote behavioral changes designed to help control temperature, such as pulling on extra blankets during a shaking chill to save heat or removing clothing to dissipate heat. Fever must be distinguished from hyperthermia, which is an uncontrolled increase in the body temperature. Hyperthermia typically develops when exogenous heat exposure or endogenous heat production exceeds the body’s ability to lose heat. This occurs despite a normal hypothalamic set point.

PART I Understanding, Controlling, and Preventing Infectious Diseases

Fever and the Inflammatory Response

PAMPs

Pathogens Microbial toxins

TLRs on Monocytes/Macrophages Neutrophils Endothelial cells

Endogenous pyrogens IL-1, TNF-α, IL-6, IFN

Hypothalamic endothelium

PG E2

↑ Thermoregulatory set point

↑ Heat production and conservation

Fever Figure 11-1. Central and peripheral mechanisms of induction of fever. IFN, interferon; IL-1, interleukin-1; PAMPs, pathogen-associated molecular patterns; PG E2, prostaglandin E2; TLRs, Toll-like receptors; TNF-a, tumor necrosis factor-a.

Clinical Aspects Body temperature varies with age, physical activity, and at various times of the day; it usually fluctuates from values less than 37°C in the early morning to values near 38ºC in the late afternoon. Normal diurnal fluctuation of temperature is also exhibited in febrile patients. In general, values higher than 37.8°C are considered to be fever, although single elevations do not always infer a pathologic process. Young infants tend to have blunted fever rises more often than older children do in response to the same antigenic stimulus. In clinical practice, core temperature is measured best by use of a rectal thermometer; oral temperatures can be influenced by prior ingestion of hot or cold foods and are reduced in the presence of open-mouthbreathing in patients with tachypnea. Axillary readings are less reliable and are typically 0.5°C lower than oral values and 1°C lower than rectal readings. In theory, the tympanic membrane is an ideal site for measuring core body temperature because it is perfused by a tributary artery supplying the body’s thermoregulatory center. Unfortunately, numerous studies of many different tympanic membrane thermometers have shown that, although convenient, such instruments tend to give highly variable readings that correlate poorly with simultaneously obtained oral or rectal readings.7,8

Antipyretic Therapy Controversy continues about whether febrile episodes should be treated.9 Substantial evidence suggests that fever is more beneficial than harmful to the host. High temperatures interfere with the replication and virulence of certain pathogens and may speed the recovery of infected patients. In addition, some immunologic responses (e.g., leukocyte migration and phagocytosis, as well as IFN production) are enhanced by temperature elevation. Moreover, fever is likely to represent a regulatory mechanism to reduce cytokine activation of the acute inflammatory response through a negative biofeedback process. Finally, nonspecific suppression of fever may

CHAPTER

11

95

eliminate an important clue for the need for further diagnostic investigations or for changes in therapy. Short courses of approved doses of standard antipyretic drugs carry a low risk of toxicity, and most of these drugs have analgesic as well as antipyretic properties. Their delivery reduces the intensity of the symptoms of an illness, but it must be remembered that the illness may be prolonged by their administration. An individual approach is suggested to determine which patients should be given antipyretic medications.

Antipyretic Agents Because fever is generated after local hypothalamic stimulation by PGs, inhibitors of the cyclooxygenase enzyme system are potent antipyretics. Although acetaminophen is a poor cyclooxygenase inhibitor in peripheral tissue, this agent is oxidized in the brain, and the resulting compound inhibits cyclooxygenase activity. Aspirin inhibits PG synthetase in a wide variety of tissues, and its antipyretic effect is equivalent to that of acetaminophen. Because these drugs are broad cyclooxygenase inhibitors, they can cause many side effects. The potential association between aspirin therapy in children with influenza or varicella and the development of Reye syndrome precludes its general use in children. Several nonsteroidal anti-inflammatory drugs (e.g., ibuprofen, naproxen) have antipyretic effects similar to those of aspirin and acetaminophen. Because these agents are associated with more adverse effects than is acetaminophen, their use for treatment of simple fever is ill advised and should be restricted to those conditions requiring an anti-inflammatory agent. Anti-inflammatory effect could have an adverse effect on clinical course of infection. Corticosteroids are among the most potent antipyretic drugs because not only do they inhibit the activity of phospholipase A2, thereby interfering with arachidonic acid metabolism and PG synthesis, but they also block the production of pyrogenic cytokines (i.e., TNF-a, IL-1, and IL-6) at a proximal step in the genesis of a febrile response. Although corticosteroids are excellent antiinflammatory agents, they should not be used for the management of simple febrile episodes and should be used with caution for noninfectious indications in the presence of infection. The use of a cooling blanket or water–alcohol bathing to accelerate peripheral heat losses is advised by some physicians. Both treatments are uncomfortable, and the latter can lead to alcohol toxicity. Peripheral cooling, in the absence of pharmacologic downregulation of the hypothalamic set point, can be counterproductive because cold receptors in the skin send signals to the spinal cord and brain for reactive vasoconstriction and shivering, thus increasing heat conservation, and eliciting oxygen consumption and heat production.

THE INFLAMMATORY RESPONSE The body has a system of sentinels that maintain immunologic surveillance to avoid pathogen-induced derangements of homeostasis. Once this background activity is circumvented, a host inflammatory response ensues. The inflammatory response is a complex reaction that involves the migration of elements of the immune system into sites of tissue injury or microbial invasion.10 In most clinical situations, the inflammatory response, with or without the aid of antimicrobial therapy, is effective in resolving the infection and contributes to tissue remodeling. If the infection is not brought under control, however, the infectious stimulus gains access to the circulation and stimulates the release of a cascade of systemic and local effector molecules. This host reaction, now called the systemic inflammatory response syndrome (SIRS), can be caused by a variety of immunologic, traumatic, surgical, or drug-induced insults; most often, however, an infectious agent is the trigger (see Chapter 12, The Systemic Inflammatory Response Syndrome (SIRS), Sepsis, and Septic Shock).

96

SECTION

C Host Defenses Against Infectious Diseases small-molecule mediators of inflammation such as arachidonic acid derivatives (Figure 11-2). The contribution of the adaptive immune system becomes important at later stages of the immune response. The adaptive immune response is mediated by T and B lymphocytes and is characterized by high specificity and long-lasting memory. The adaptive immunity is influenced by the generation of helper T-cell subsets (Th) and the subsequent production of “effector” cytokines by these cells. Thus, naive T lymphocytes that recognize the antigen presented by antigen-presenting cells (APC) undergo activation and expansion. At this stage, they can differentiate into two subsets: Th1 or Th2.12 Th1 cells secrete IFN-g and primarily promote cellular immunity, whereas Th2 cells produce IL-4, IL-5, IL-10, and IL-13 and primarily promote humoral immunity. The predominant pathway of T-lymphocyte differentiation is determined by the cytokine milieu at this step. The IFNs and IL-12 drive Th1 differentiation, whereas IL-4 induces Th2 differentiation. Infection of intracellular pathogens induces primarily a Th1-dominated response that protects against the

Innate and Adaptive Immunity The immune response to infection and cellular injury can be broadly divided into two categories: the innate and adaptive immunity.11 The immediate response, associated with the production of the pyrogenic cytokines IL-1, IL-6, and TNF-a, is mounted by the innate immune system that involves neutrophils, monocytes/macrophages, dendritic cells, and natural killer (NK) lymphocytes. The main receptors of the innate immune system recognize broad patterns of conserved and often integral structural components of microbes that are not present in or on human cells. These highly conserved microbial structural components are often called pathogen-associated molecular patterns, or PAMPs, while the host receptors are called pattern recognition receptors, or PRRs. The binding of PAMPs to PRRs results in rapid changes in expression of genes, including those encoding pyrogenic cytokines such as IL-1, Il-6, and TNF-a. In addition to inflammatory cytokines, many other pathways are activated, including synthesis of degradative enzymes and enzymes responsible for the production of

Innate immunity

Adaptive immunity

Microbial pathogens

PAMP Recognition

TLR

Phagocytosis Th2

NFκB

Antigen presentation

IL- 4, IL -13

TCR

MHC

Clonal expansion

Naïve T cell CD86/80

CD28 IL-2R

Costimulatory molecules

IL-2

Phospholipase A2

IFNγ

Th1

Arachidonic acid G-CSF

TLR

PGE2

Vascular endothelium

TF

IL-1β IL-6 IL-8 TNF-α G-CSF

Adhesion molecules

↑ Coagulation ↑ Leukocyte recruitment

Hypothalamus

Liver

Bone marrow

Acute ↑ Neutrophils phase response Figure 11-2. Induction of the innate and adaptive immune responses. The recognition of pathogen-associated molecular patterns (PAMPs) by Toll-like receptors (TLR) results in the activation of intracellular signaling pathways leading to the activation of the transcription factor NFkB. The translocation of NFkB into the nucleus leads to upregulation of expression of genes encoding proinflammatory cytokines interleukin-1 (IL-1), interleukin6 (IL-6), tumor necrosis factor-a (TNF-a), and colony-stimulating factors (CSF). It also leads to increased expression of costimulatory molecules involved in antigen presentation and induction of the adaptive immune responses. The cytokines IL-1, IL-6, TNF, and granulocyte CSF (G-CSF) initiate the inflammatory cascade through their effects on hypothalamus, bone marrow, liver, and vascular endothelial cells. IFN, interferon; MHC, major histocompatibility complex; T-cell receptor (TCR) hemostatic tissue factor, (TF), Fever

PART I Understanding, Controlling, and Preventing Infectious Diseases

Fever and the Inflammatory Response

majority of microorganisms, whereas some parasitic infections induce a Th2 response that is associated with resistance to helminths. In addition to instructive cytokines, APCs use several costimulatory molecules, including CD80 and CD86, to signal T lymphocytes and to induce clonal expansion of antigen-specific T lymphocytes. Antigen presentation in the absence of costimulatory molecules leads to cell anergy. Since the engagement of the innate receptors on APC leads to upregulation of expression of the costimulatory molecules and enhances antigen presentation, the interaction between the innate and adaptive immunity at this step becomes very important.11

Receptors of the Innate System The initial recognition of infectious agents by the innate immune system is mediated by the PRRs. The PRRs recognize PAMPs which are shared by large groups of microorganisms but are not present in mammalian cells.11 Probably the best-known examples of PAMPs are lipopolysaccharides (LPS). LPS are the mixture of fragments of the outer membrane of gram-negative bacteria. LPS administration induces local and systemic inflammation and tissue damage that is very similar to that observed in infection caused by gram-negative bacteria. For this reason, the LPS-induced response has been used as a model of inflammation for several decades. It was only in 1998, however, when the principal component of the human receptor system recognizing LPS was identified as Toll-like receptor 4 (TLR 4). The term is based on structural similarities with the Drosophila protein “Toll” which is required for the normal development of the flies and protection against Aspergillus species.13,14 At least 9 other human TLRs have been identified. Each appears to recognize a distinct set of PAMPs (Table 11-1). It is becoming clear that, collectively, TLRs can detect most, if not all, microbes,11,15,16 including protozoa, bacteria, viruses, and fungi. Some TLRs may also recognize endogenous ligands such as degradation products of various extracellular matrix components released as a result of tissue damage. Such recognition may contribute to the development and perpetuation of inflammation as well. TLRs are expressed in different cell types that are important components of the innate immune system, including monocytes/ macrophages, dendritic cells, neutrophils, vascular endothelial cells, and epithelial cells lining mucosal surfaces. The binding of microbial molecules to TLRs leads to the activation of intracellular signaling pathways. The predominant pathway used by TLRs leads to the activation of NFkB, a nuclear transcription factor that initiates rapid changes in expression of several groups of genes (see Figure 11-2). Most important are those encoding: ●

cytokines and chemokines (including the pyrogenic cytokines TNF-a, IL-1, IL-6, as well as IL-8, colony-stimulating factors (CSFs), platelet-activating factor (PAF), among others)17

● ● ●

CHAPTER

11

97

enzymes involved in the degradation of extracellular matrix proteins enzymes responsible for the production of small-molecule mediators of inflammation such as arachidonic acid derivatives proteins involved in microbial killing mechanisms

The engagement and activation of TLRs expressed on vascular endothelial cells also lead to increased expression of chemokines and adhesion molecules (selectins and integrin ligands).18 Combined with the simultaneous upregulation of selectin ligands and integrins on leukocytes (also stimulated by TLR engagement), this results in increased adhesion of leukocytes to the vascular endothelium. Once these cells are firmly attached, they begin transmigration across endothelial surfaces (i.e., extravasation) to the sites of microbial invasion and tissue damage.18 During extravasation, degranulation of polymorphonuclear cells can occur. Such degranulation is associated with the release of several proteolytic enzymes and toxic oxygen radicals that contribute to increased vascular permeability. Furthermore, activation of the complement system (classic and alternate pathways) and coagulation cascades (intrinsic and extrinsic pathways) occurs concurrently with cytokine stimulation. Release of the anaphylatoxins C3a and C5a promotes vascular abnormalities and neutrophil activation. The cytokines and chemokines whose production is stimulated by TLR engagement in turn produce multiple effects that further amplify the inflammatory response:17 ●

●

●

●

The cytokines IL-1, TNF-a, and IL-6 further stimulate metabolism of arachidonic acid to form leukotrienes, thromboxane A2, and prostaglandins (especially PGE2 and PGI2) IL-1, TNF-a, and IL-6 elevate PGE2 synthesis in the vascular and perivascular cells of the hypothalamus, leading to generation of fever2–4 Granulocyte colony-stimulating factor (G-CSF) increases production and release of neutrophils from the bone marrow to replace those consumed by inflammation, often leading to an increase in absolute neutrophil counts IL-1 and TNF-a further stimulate endothelial cells to synthesize a number of secondary factors that contribute to the inflammatory process. These include endothelial-cell adhesion molecules, chemokines (IL-8, MIP-1a), the hemostatic tissue factor, as well as IL-1, PGI2, and granulocyte–macrophage colony-stimulating factor (GM-CSF).

The release of the hemostatic tissue factor by endothelial cells and macrophages appears to be pivotal in the activation of the coagulation cascade. In turn, the activation of the coagulation cascade can trigger activation of the complement system. Combined, these pathways can lead to the development of disseminated intravascular coagulation, often seen in SIRS.19 Platelets are also primed to interact with endothelium and to aggregate in dense masses that interfere with blood supply to tissues. This generalized endothelial cell activation is critical in the pathogenesis of vascular injury and capillary permeability associated with acute inflammation that may progress to shock and multiorgan dysfunction in some patients.19

TABLE 11-1. Toll-like Receptors (TLR) and their Microbial Ligands TLR

Ligand

Microbial Source

TLR 2

Lipoproteins Peptidoglycan Lipopolysaccharide Zymosan

Bacteria Gram-positive bacteria Leptospira Yeast

TLR 3

Double-stranded RNA

Viruses

TLR 4

Lipopolysaccharide

Gram-negative bacteria

TLR 5

Flagellin

Bacteria

TLR 7/8

Single-stranded RNA

Viruses

TLR 9

CpG DNA

Bacteria, protozoa

Acute-Phase Response In response primarily to IL-1 and IL-6, a structurally and functionally heterogeneous group of proteins change their concentration in peripheral blood. These proteins are known as acute-phase response proteins.1,20 Some of these are known as positive acute-phase proteins because they increase in concentration after antigenic stimuli. These proteins are mainly synthesized by the liver; the most important examples are C-reactive protein (CRP), serum amyloid A, a1-acid glycoprotein, a1-antitrypsin, haptoglobin, ceruloplasmin, and fibrinogen. Some other acute-phase proteins are referred to as negative acutephase proteins (e.g., albumin, prealbumin, retinol-binding protein, transferrin) because their plasma concentrations are reduced.1 The

98

SECTION

C Host Defenses Against Infectious Diseases

exact functions of the acute-phase proteins are not completely understood. CRP appears to be a pattern recognition molecule.21 CRP typically binds to molecular configurations that are either exposed during cell death or found on the surface of certain pathogens such as Streptococcus pneumoniae. Ligand-bound or aggregated CRP efficiently activates the classic complement pathway through a direct interaction with C1q, and stimulates phagocytosis. This raises the possibility that CRP is involved in clearing the cellular debris from necrotic and apoptotic cells, but it may also provide additional protection against certain microbes. The magnitude of acute-phase responses provides a guide to the intensity of the inflammation or the extent of tissue involvement. For instance, the rise in fibrinogen causes erythrocytes to form stacks (rouleaux) that sediment more rapidly than do individual erythrocytes. This is the basis for measuring erythrocyte sedimentation rate as a simple test for the magnitude of systemic inflammatory response regardless of the initiating stimuli. Mineral changes are uniformly present during the acute-phase host response. The best-documented alterations are decreased serum concentrations of iron and zinc caused by the uptake in hepatocytes and phagocytes. Conversely, serum copper levels are elevated as a consequence of increased synthesis of ceruloplasmin, the copper carrier protein. Because of hypoferremia, reduction of red blood cell synthesis, and decreased red blood cell lifespan, mild but reversible anemia usually accompanies a significant acute inflammatory response. Finally, profound alterations in utilization of carbohydrates, proteins, and lipids occur during acute inflammatory responses.1,22,23 The hypermetabolic state typically seen in patients with sepsis requires massive use of carbohydrate and lipid stores to meet energy needs. In addition, amino acids from muscle tissue are used by the liver to produce glucose (gluconeogenesis). These changes are presumably provoked by cytokine-induced production of cortisol, insulin, glucagon, and growth hormone. Clinically, these abnormalities manifest as reactive hyperglycemia, substantial fluctuation in plasma concentrations of free fatty acids, hypertriglyceridemia, and negative nitrogen balance resulting from catabolism of amino acids. Many of these effects are mediated by TNF-a.

The Regulatory Anti-Inflammatory Pathways The overproduction of proinflammatory substances can lead to extensive tissue damage and vascular injury, disseminated intravascular coagulation, and shock. To attenuate potential damage to the host, the inflammatory cascade involves participation of regulatory anti-inflammatory pathways. Among these modulating pathways are those that involve anti-inflammatory cytokines such as transforming growth factor-b (TGF-b), IL-4, and IL-10. TGF-b also has the ability

to stimulate production of extracellular matrix proteins and thus contributes to tissue remodeling after the resolution of inflammation. Other examples of anti-inflammatory mechanisms are natural inhibitors of inflammatory cytokines.24 For instance, many inflammatory cells, including neutrophils and macrophages, have the ability to shed their TNF receptors. These soluble TNF receptors bind free TNF and neutralize its proinflammatory activity. Also, the IL-1 system has unique pathways of negative regulation that involves natural IL-1 receptor antagonists and the type II decoy receptor. Interestingly, some anti-inflammatory cytokines (e.g., IL-4, IL-10) can simultaneously inhibit the production of IL-1 and stimulate production of receptor antagonists and the decoy receptor. Thus, an intricate network of positive and negative biofeedback loops operates during the course of a systemic inflammatory response and the net balance determines clinical expression and severity of disease.

Integration and Homeostasis In summary, the immediate inflammatory response to microbial invasion or tissue injury is mainly controlled by the innate immune system. This innate response is triggered by the binding of certain microbial molecules known as PAMPs to the pattern recognition receptors (i.e., TLRs) expressed in host cells. The subsequent intracellular signaling in host cells leads to increased expression of several groups of genes, including those encoding pyrogenic and proinflammatory cytokines TNF-a, IL-1, and IL-6. TLR activation also leads to upregulation of expression of the costimulatory molecules involved in antigen presentation (i.e., CD80/CD86). This leads to the enhancement of the adaptive immunity which becomes more important in the later stages of the inflammatory response. Anti-inflammatory pathways are activated simultaneously, and the net balance between pro- and anti-inflammatory stimuli determines the magnitude and the outcome of the inflammatory response. Basic mechanisms of the inflammatory response are similar to those described regardless of whether the response is systemic or local.25 For instance, in bacterial meningitis, local production of cytokines within the subarachnoidal space in response to presence of bacteria, or their cell wall components, induces the disruption of the blood–brain barrier, increased vascular permeability, and chemotactic attraction of polymorphonuclear cells to the meningeal site of microbial invasion. The effective cooperation between the adaptive and innate immunity eventually leads to the elimination of the invading microbes. The absence of microbial stimulation leads to the cessation of the proinflammatory stimuli and resolution of inflammation. At this stage, several cytokines and growth factors (including TGF-b) contribute to the repair of the damaged tissues. This allows the host to return to a relative state of homeostasis until the next microbial challenge.

PART I Understanding, Controlling, and Preventing Infectious Diseases

PA RT

II

Clinical Syndromes and Cardinal Features of Infectious Diseases: Approach to Diagnosis & Initial Management SECTION

A

Septicemia, Toxin and InflammationMediated Syndromes CHAPTER

12

The Systemic Inflammatory Response Syndrome (SIRS), Sepsis, and Septic Shock Judith Guzman-Cottrill, Simon Nadel, and Brahm Goldstein

Sepsis remains a major cause of morbidity and mortality among children.1–4 Sepsis-associated mortality in children has decreased from 97% in 19665 to 9% among infants in the early 1990s.6 A recent population-based study of United States children with severe septicemia (bacterial or fungal infection with at least one acute organ dysfunction) reported a mortality rate of 10.3%.7 Although this represents a significant improvement over past decades, severe sepsis remains one of the leading causes of death in children, with over 4300 deaths annually (7% of all deaths among children) and estimated annual total costs of $1.97 billion.8 In a seminal study, Watson et al.8 analyzed the impact of age, sex, birthweight, underlying disease, and microbiologic etiology on the incidence, mortality, and hospital costs of children who develop septicemia using 1995 hospital discharge and population data from seven states. Table 12-1 shows the annual incidence, case fatality, and national estimates of severe sepsis by age. The incidence is highest in infants (5.16 per 1000), falls significantly in older children (0.20 per 1000 in 10 to 14-year-olds), and also exhibits a sex difference, being 15% higher in boys than in girls (0.60 versus 0.52 per 1000, P < 0.001).8 Overall hospital mortality was 10.3%, or 4383 deaths nationally (6.2 per 100,000 population).8 Of interest, about 50% of the cases had an underlying disease and over 20% were lowbirthweight neonates. The most common infections were respiratory tract (37%) and primary bloodstream infections (BSIs) (25%).8 The mean length of stay was 31 days, and the cost was $40,600 per admission.8

DEFINITIONS An international panel of experts in the fields of adult and pediatric septicemia and clinical research proposed the first set of specific definitions and criteria for the components of the sepsis continuum that can be applied consistently in the pediatric population.6 The consensus definitions for systemic inflammatory response syndrome (SIRS), infection, sepsis, severe sepsis, septic shock, and multiple organ dysfunction syndrome in children are listed in Box 12-1. It is important to recognize that these definitions were meant for use in the design, conduct, and analysis of large, multicenter, international therapeutic trials rather than as a clinical tool at the bedside.

TABLE 12-1. Annual Incidence, Case Fatality, and National Estimates of Severe Sepsis by Age

Age 38.5°C or < 36°C • Tachycardia, defined as a mean heart rate > 2 SD above normal for age in the absence of external stimulus, chronic drugs, or painful stimuli; or otherwise unexplained persistent elevation over a 0.5- to 4-hour time period or for children < 1 year old: Bradycardia, defined as a mean heart rate < 10th percentile for age in the absence of external vagal stimulus, beta-blocker drugs, or congenital heart disease; or otherwise unexplained persistent depression over a 0.5-hour time period • Mean respiratory rate > 2 SD above normal for age or mechanical ventilation for an acute process not related to underlying neuromuscular disease or the receipt of general anesthesia • Leukocyte count elevated or depressed for age (not secondary to chemotherapy-induced leukopenia) or > 10% immature neutrophils INFECTION A suspected or proven (by positive culture, tissue stain, or polymerase chain reaction test) infection caused by any pathogen or a clinical syndrome associated with a high probability of infection. Evidence of infection includes positive findings on clinical exam, imaging, or laboratory tests (e.g., white blood cells in a normally sterile body fluid, perforated viscus, chest X-ray consistent with pneumonia, petechial or purpuric rash, or purpura fulminans) SEPSIS SIRS in the presence of or as a result of suspected or proven infection SEVERE SEPSIS Sepsis plus the following: cardiovascular organ dysfunction, acute respiratory distress syndrome (ARDS), or two or more other organ dysfunctions SEPTIC SHOCK Sepsis and cardiovascular organ dysfunction a

Core temperature must be measured by rectal, oral, or central catheter probe. From Goldstein B, Giroir B, Randolph A. International pediatric sepsis consensus conference: definitions for sepsis and organ dysfunction in pediatrics. Pediatr Crit Care Med 2005;6:2–8.

Candida spp. Viruses can also cause sepsis in the neonate. Viral sepsis can be identical clinically to bacterial disease, including infection due to herpes simplex virus, enteroviruses, respiratory syncytial virus, and influenza virus. Beyond the neonatal period, S. pneumoniae, Neisseria meningitidis, and H. influenzae type b (Hib; in nonimmunized children) are common pathogens in otherwise healthy children. Widespread Hib immunization since 1988 has virtually eradicated invasive Hib disease in developed countries; in 2005, only 7 cases of invasive Hib disease in children < 5 years of age were reported in the United States.11 Since the routine administration of the heptavalent pneumococcal conjugate vaccine in 2000, the overall incidence of invasive pneumococcal disease due to vaccine serotypes in children < 5 years of age has declined by 94%.12 Other organisms include Staphylococcus aureus, group A streptococcus, Salmonella spp., and rickettsiae in specific geographic regions (Rocky Mountain spotted fever and ehrlichiosis). In hospitalized infants and children, coagulase-negative staphylococci (if a foreign body is present, such as an indwelling vascular catheter), Enterobacteriaceae, S. aureus, and fungi (especially Candida spp.) are important pathogens. Children with an underlying immunodeficiency can develop infections with the same pathogens as healthy children; however, certain conditions predispose them to septicemia due to additional organisms. Neutropenic patients are at risk of BSI due to gram-negative bacteria (including Pseudomonas aeruginosa), alpha-hemolytic streptococci, and cytomegalovirus. Viridans group streptococci are an important cause of sepsis in the neutropenic oncology population, oftentimes with associated acute respiratory distress syndrome (ARDS).13,14 Patients with acquired immunodeficiency virus (AIDS) or severe immunosuppression due to human immunodeficiency virus (HIV) are at higher risk of bacteremia due to Streptococcus pneumoniae, P. aeruginosa, Staphylococcus aureus, and Hib. Patients with anatomic or functional asplenia (including sickle-cell disease) are at increased risk for developing BSI caused by encapsulated organisms, including Streptococcus pneumoniae, Salmonella spp., Hib, and N. meningitidis.

PATHOPHYSIOLOGY The diagnosis and thus the definition of septic shock in children can be challenging. Children often maintain blood pressure until severely ill;9 there is no requirement for systemic hypotension in order to make the diagnosis of septic shock as there is in adults. Shock can occur long before hypotension occurs in children. Hypotension is a sign of late and decompensated shock in children and is confirmatory of shock state if present in a child with suspected or proven infection.10 Although there are distinct clinical presentations and classifications of shock in children (e.g., warm and cold shock; fluid-refractory and catecholamine-resistant shock), septic shock is defined as septicemia in the presence of cardiovascular dysfunction (i.e., severe sepsis with cardiovascular dysfunction).

ETIOLOGY Several factors influence the various pathogens causing septicemia in children, including age and host immune status. In addition, community-acquired pathogens differ from those in a hospital setting. During the neonatal period, the predominant bacterial causes include group B streptococci and enteric bacilli, such as Escherichia coli. Other less common pathogens include enterococci, Listeria monocytogenes, Haemophilus influenzae, and Streptococcus pneumoniae. Advances in neonatal management and survival of very-low-birthweight infants have led to a shift in etiologic agents to hospitalassociated infection. The use of intravascular access devices and other foreign bodies further predisposes these already compromised neonates to nosocomial infection, such as due to coagulase-negative staphylococci, Staphylococcus aureus, gram-negative bacilli, and

If a microbe gains access to the intravascular compartment, the host activates defensive mechanisms. Transient BSI without significant clinical consequences occurs commonly in children. Depending on the age and immunocompetence of the patient, the virulence and number of pathogens in the blood, and the timing and nature of a therapeutic intervention, a SIRS ensues that can progress independently despite successful eradication of the microbe. Although infection is a major cause of SIRS, a number of other entities, including trauma, ARDS, neoplasm, burn injury, and pancreatitis, are recognized causes. Most pathophysiologic consequences of the sepsis syndrome result from an imbalance between pro- and anti-inflammatory mediators in combination with microbial toxins.15 In children, severe sepsis arises from coordinated activation of the innate immune response.16 This response, triggered by diverse pathogens, is multifaceted.15–17 Once triggered, the response leads to secretion of pro- and anti-inflammatory cytokines, activation and mobilization of leukocytes, activation of coagulation and inhibition of fibrinolysis,18,19 and increased apoptosis.20 As a result of coagulation activation, thrombin generated promotes fibrin deposition in the microvasculature and also exacerbates ongoing inflammation by direct and indirect mechanisms.15 Although evolutionarily designed to limit microbial dissemination, innate inflammatory processes can also be detrimental, resulting in cardiac dysfunction, vasodilation, capillary injury, and micro- and macrovascular thromboses. Despite antimicrobial therapy and intensive care, these processes frequently lead to organ dysfunction, gangrenous extremities, long-term neurologic morbidity, or death (Figure 12-1).1,21 The clinical manifestations are the result of systemic inflammation and include abnormal temperature regulation, flushed warm skin, widened pulse pressure, tachycardia, tachypnea, metabolic acidosis (elevated serum lactate, decreased base excess), renal and/or hepatic

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

The Systemic Inflammatory Response Syndrome (SIRS), Sepsis, and Septic Shock

CHAPTER

12

101

TABLE 12-2. Partial List of Clinical Signs and Biologic Markers of Systemic inflammation

Sepsis in Children

Coagulation

Classification

Increased

Endotoxin Enterobacterial common antigen Candida antigen Bacterial DNA Physiologic indicators Temperature Heart rate Cardiac index Respiratory rate

Decreased

Microbial products Sepsis

Impaired fibrinolysis

Acute organ dysfunction

Death

Figure 12.1. The pathophysiologic consequences of the sepsis syndrome result from an imbalance between pro- and anti-inflammatory mediators, activation of the coagulation cascade, and inhibition of fibrinolysis. The syndrome can progress to acute multiorgan dysfunction and death.

dysfunction, thrombocytosis, and leukocytosis. As the syndrome progresses, multiorgan failure, including acute respiratory failure, hypotension, myocardial failure, decreased neurologic function, oliguric or anuric renal failure, hepatic failure, leukopenia, anemia, and thrombocytopenia, can ensue and can lead to death.

CLINICAL AND LABORATORY FINDINGS Fever, tachycardia, and tachypnea are the most common physiologic abnormalities associated with sepsis, even though they are insensitive and nonspecific. Other clinical signs include decreased tone, diminished activity, pale or gray skin color, prolonged capillary refill time, and poor feeding or sucking.22 Biochemical markers of inflammation may one day prove to be more objective and reliable than physiologic findings; however, no biochemical marker has been confirmed to be robust enough to use for the definitive diagnosis of sepsis or for tracking response to therapy and disease progression. A rapid and reliable clinical or biologic marker would be invaluable as early treatment with antibiotics and fluid resuscitation have been clearly demonstrated to reduce both morbidity and mortality.23–31

Temperature Heart rate Blood pressure Systemic vascular resistance Urine output Level of consciousness Neutrophils Monocytes Platelets

Hematopoietic cells

Neutrophils Monocytes

Soluble receptors

sTNF-RI sTNF-RII IL-1 IL-6 IL-8 IL-10 IL-18 TNF C-reactive protein Albumin LipopolysaccharidePrealbumin binding protein Fibrinogen Fibrin degradation Antithrombin III products Activated protein C von Willebrand’s factor PAI-I Thrombin–antithrombin complexes D-dimers Thrombomodulin

Cytokines

Acute-phase reactants

Mediators of coagulation

IL, interleukin; PAI-I, plasminogen-activator inhibitor type 1; sTNF, soluble TNF-a; TNF, tumor necrosis factor. Modified from Carcillo JA, Planquois J-M, Goldstein B. Early markers of infection and sepsis in newborns and children. Adv Sepsis 2006;5:118–125

Laboratory Findings Clinical Signs The earliest clinical sign of clinical infection is age-dependent changes in body temperature.32 In immune-competent children the earliest sign is fever. In immune-compromised children and premature infants the earliest sign can be hypothermia, or fever.32 Fever in association with changes in a child’s behavior, such as an infant’s loss of smiling or playfulness (especially after fever has been controlled with antipyretic therapy), are signs of serious infection which can benefit from antibacterial, -viral, or -fungal therapy.33–36 Tachycardia is a useful sign of sepsis in the neonate born at term,37 as is tachycardia and/or tachypnea in older children.32 Fever can account for some tachycardia, as each 1°C increase can result in an increase in heart rate of 10%; however, the heart rate and respiratory rate should become normal for age when fever is controlled with antipyretic therapy or falls spontaneously.32 Heart rate >150 beats/ minute in children and >160 beats/minute in infants, and respiratory rates >50 beats/minute in children, and >60 beats/minute in infants are associated with increased mortality risk and commonly presage the development of septic shock.6

Table 12-2 lists some of the more clinically relevant and studied biologic markers of sepsis in children.32,38 Among markers that have been extensively studied are the total peripheral white blood cell (WBC) count,39,40 procalcitonin (PCT),41–44 C-reactive protein (CRP),45–47 lipopolysaccharide-binding protein (LBP),44 interleukin-6 (IL-6),44,48,49 IL-8,50 protein C,15,51–62 and endocan.63 None is completely sensitive or specific, or has proven reliability in clinical decisionmaking or clinical trials. There are many types of biologic markers currently under development that may prove to be more accurate in screening, diagnosing, and determining the response to therapy during infection and sepsis. These include specific rapid antigen assays (e.g., for Streptococcus pneumoniae antigen),64 polymerase chain reaction,65–71 genomic testing (for guiding therapy and determining host response),72,73 and proteomic testing (for identification of differentially expressed proteins and peptides).74–76 Combinations of markers may improve sensitivity and specificity. Examples include increase in IL-6 in combination with CRP;48 PCT and IL-8 but not IL-6 in bacterial infections;50 increases in tumor necrosis factor-alpha (TNF-a), soluble TNF-a receptor (sTNFR), and IL-1 receptor antagonist (IL-1ra).77

102

SECTION

A Fever, Bacteremia, & Septicemia

MANAGEMENT Antimicrobial Therapy Empiric therapy for suspected BSI should be administered promptly, targeted towards the likely causative pathogens (Table 12-3). Important considerations when selecting a regimen include: the patient’s age, community versus nosocomial acquisition, the host immune status, and penetration into tissues and compartments (such as central nervous system). Knowledge of organism susceptibility patterns in the community and institution is essential; in some areas of the United States, rates of community-acquired methicillin-resistant Staphylococcus aureus (CA-MRSA) are as high as 76%.78 In such areas, vancomycin should be included in the empiric regimen if this pathogen is suspected. Once the causative organism is isolated and antibiotic susceptibilities are available, antimicrobial therapy is adjusted appropriately. If a gram-positive organism is susceptible to an antibiotic other than vancomycin, vancomycin should be discontinued to minimize the emergence of highly resistant pathogens. If Escherichia coli or Klebsiella spp. is isolated (or additional gram-negative bacilli in certain hospital settings), the organism should be tested for extendedspectrum b-lactamase (ESBL) production. If ESBL production is confirmed, then extended-spectrum cephalosporins should be discontinued. A carbapenem is considered the treatment of choice for serious infections with ESBL-producing organisms.79

Supportive Care Effective treatment of sepsis and septic shock is dependent on prompt recognition and initiation of supportive as well as specific therapy. The basic principles of initial critical care include ensuring adequate airway patency, gas exchange (breathing), and circulation (ABCs). The interventions required to achieve these goals depend on the specific physiologic state of the patient at the time of presentation. Shock that occurs during sepsis can result from decreased intravascular volume, maldistribution of intravascular volume, and/or impaired myocardial function, all of which can occur at different times during the course of septic shock.80 Children with sepsis who receive early aggressive fluid resuscitation (> 40 mL/kg in the first hour with isotonic intravenous fluids) demonstrate improved survival and no increased risk of noncardiogenic pulmonary edema or ARDS.31,81 Determination of when, what type, and how much pharmacologic support is needed in a patient with septic shock requires careful consideration of many factors. These factors include the patient’s clinical state (e.g., capillary refill time, urine output, peripheral versus core temperature gradient), information obtained from monitoring devices (heart rate, blood pressure, central venous pressure, pulmonary artery pressure, cardiac output, stroke volume, and systemic vascular resistance), and knowledge of basic drug effects (including

TABLE 12-3. Suggested Antimicrobial Choices for Empiric Therapy in Infants and Children with Suspected Sepsis Age or Clinical Situation

Antimicrobial(s)

Neonate

Ampicillin + gentamicin or cefotaxime

Neonate (nosocomial)

Vancomycin + gentamicin or ceftazidime

Child

Cefotaxime or ceftriaxone + vancomycin

Child (nosocomial)

Vancomycin + aminoglycoside or antipseudomonal penicillin or ceftazidime or carbapenem

Skin or soft-tissue involvement

Semisynthetic penicillin or vancomycin + clindamycin

Herpes simplex virus

Acyclovir

dopamine, norepinephrine, epinephrine, and phenylephrine) in the setting of septic shock. Septic shock causes multisystem organ dysfunction, and it is important to evaluate and treat abnormalities in other organ systems, including the kidney and gastrointestinal tract. Maintaining adequate intravascular volume and the use of low-dose dopamine (2 to 5 μg/kg per min), furosemide, or mannitol may improve renal blood flow and be beneficial in preventing renal failure. Patients with acute renal failure may require dialysis or continuous venovenous hemofiltration. The gastrointestinal tract is vulnerable to disturbances such as hemorrhage, ileus, brush border atrophy, and translocation of enteric organisms into the blood. Antacid therapy or treatment with an H2-receptor blocker or sucralfate is standard. Additionally, early institution of nutritional support, particularly enteral feedings, may ameliorate gastrointestinal atrophy and improve multiorgan function.82 Finally, fever may cause an in increase in oxygen consumption and contribute to decreased oxygen delivery to organs. Therefore, control of fever with a standard antipyretic agent is recommended. Some studies suggest that controlled hypothermia may be of benefit in diminishing oxygen consumption and metabolic demands of tissues.

Endotoxin Physiology and Antiendotoxin Therapy Endotoxin Physiology Endotoxin is probably one of the most important bacterial components contributing to the inflammatory process. Levels of endotoxin correlate directly with severity of meningococcal disease and other forms of sepsis, and with elaboration and release of inflammatory mediators, including the cytokines. Endotoxin upregulates TNF-a, IL1 and IL-6, complement and coagulation pathways. Endotoxin can also be found in the presence of critical illness, not related to gramnegative sepsis,83,84 where its presence appears to be related to severity of disease and outcome. It is postulated that the presence of endotoxin in the blood in these circumstances is related to the altered gut permeability. The assumption that the inflammatory process is related to the presence of endotoxin in the bloodstream is based on the finding that the pathophysiology of gram-negative sepsis can be reproduced by the administration of purified endotoxin or a variety of endotoxin-free inflammatory mediators, which are upregulated by endotoxin. In addition, many effects can be blocked, in vitro and in vivo, by agents that neutralize the effects of endotoxin or the elaborated inflammatory mediators. A variety of antiendotoxin strategies have been proposed, including agents that bind to and neutralize endotoxin, enhance endotoxin clearance, or inhibit the interaction of endotoxin with its receptors.

Antiendotoxin Antibodies Since the 1960s investigators have attempted to produce neutralizing antibodies to the highly conserved elements in the core region of endotoxin (such as Lipid A). Early studies indicated that passively administered antisera raised to vaccines generated from rough mutant bacteria (such as E. coli J5) protected against challenge from heterologous gram-negative bacteria.85–87 Two cross-protective monoclonal antibodies apparently directed against Lipid A were developed; two antibodies that went into mass production and were found to weakly bind were subjected to larger studies in adults and children.88–90 A murine monoclonal antibody of the immunoglobulin M (IgM) class, E5, was studied in a large multicenter, placebo-controlled trial in adult patients with severe sepsis due to probable or confirmed gramnegative infection.89 The study was stopped after the second interim analysis. A total of 1090 patients received study medication and 915 had gram-negative infection confirmed by culture. There were no statistically significant differences in mortality between the E5 and placebo groups at either day 14 (29.7% versus 31.1%; P = 0.67) or day

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

The Systemic Inflammatory Response Syndrome (SIRS), Sepsis, and Septic Shock

28 (38.5% versus 40.3%; P = 0.56). Patients presenting without shock had a slightly lower mortality when treated with E5, but the difference was not significant (28.9% versus 33.0% for the E5 and placebo groups, respectively, at day 28; P = 0.32). Another antibody, HA-1A, a humanized monoclonal IgG antibody against the lipid A moiety of endotoxin, was studied in two multicenter, randomized, double-blind, placebo-controlled studies in adults and one trial in children with meningococcal septic shock. The two adult studies showed conflicting results, the first showing a significant benefit in adults with gram-negative septic shock.90 However, due to methodologic problems, this study was repeated in a larger number of patients. This study of 621 patients with presumed gram-negative infection and shock failed to repeat the positive results of the first study.91 Mortality rates in this group were as follows: placebo, 32% (95 of 293) and HA-1A, 33% (109 of 328) (P = 0.864). Mortality rates in the patients treated with HA-1A without gramnegative bacteremia were higher: placebo, 37% (292 of 793) and HA1A, 41% (318 of 785) (P = 0.073), and following this study HA-1A was abandoned as a potentially useful agent. The study in children with meningococcal disease was completed before the second adult study was reported. The pediatric study demonstrated a 33% absolute reduction in mortality in the treatment group,92 which was not statistically significant. The authors concluded that findings could indicate: 1. a genuine beneficial effect of HA-1A that was dampened by nonoptimal timing of intervention. The median time from initiation of antibiotic therapy to administration of study medication was 6.4 hours, with 25% of patients receiving study medication > 9.7 hours following antibiotic therapy 2. a chance trend of decreased mortality in treated children. In vitro data have suggested that HA-1A is not efficient in binding and neutralizing meningococcal endotoxin93 3. a lack of efficacy of HA-1A

Other Antiendotoxin Therapies There are several promising antiendotoxin therapies under development. These are derived from the existence of peptides and proteins in insects and animal species, which have evolved for the purpose of binding and neutralizing endotoxin. Endotoxin, which is present in the circulation, forms complexes with circulating proteins and lipoproteins, such as the acute-phase reactant LBP which facilitates the transfer of endotoxin to its receptors Toll-like receptor (TLR4), CD14, and the lipoproteins.94 Endotoxin is also bound and neutralized by several neutrophil granule proteins, including the bactericidal/ permeability-increasing protein (BPI) and the cationic antibacterial protein hCAP-18 (a cathelicidin).95,96 Only one of these compounds has been the subject of controlled clinical trials. A recombinant form of BPI (rBPI21), consisting of 21 amino acids of the N-terminal fragment of naturally occurring BPI, has been shown to act synergistically with antibiotics in the killing of many bacteria, and to bind to and neutralize endotoxin. Use of this recombinant protein in an uncontrolled clinical study in children with severe meningococcal septicemia was associated with reduction in mortality compared with historical controls.97 A phase III randomized, double-blind, placebo-controlled study of rBPI21 was then performed in 393 children with meningococcal disease (190 children received rBPI21 and 203, placebo). The primary endpoint, mortality rate, was similar in the two groups: 14 (7.4%) in the rBPI21 group and 20 (9.9%) in the placebo group (P = 0.48). Trend of positive drug effects was seen in secondary endpoints. Fewer patients treated with rBPI21 had multiple severe amputations (6 of 190 (3.2%) versus 15 of 203 (7.4%), odds ratio 2.47 (0.94 to 6.51); P = 0.067), and more had a functional outcome at day 60 similar to that before illness (136 of 176 (77.3%) versus 126 of 190 (66.3%); P = 0.019).98 A large number of children who died before starting or completing the infusion of rBPI21, as well as late administration of drug after antibiotic (mean 5.9 hours), may have confounded potential beneficial effect. No further phase III studies have been performed.

CHAPTER

12

103

Statin Therapy Endotoxin is known to bind to lipid-containing particles such as highdensity lipoprotein (HDL), low-density lipoprotein (LDL), and verylow-density lipoprotein (VLDL). These particles appear to be involved in detoxification and clearance of circulating endotoxin through the reticuloendothelium.99 Preparations of HDL reconstituted from plasma are able to neutralize endotoxin more potently than natural lipoproteins100 and have shown promising results in adult volunteers challenged with endotoxin.101 Although the mechanism by which HDLs modulate the inflammatory response are complex, the results so far clearly indicate that HDLs, and some pharmacologic agents, such as fibrates, niacin, and statins (which have been shown to elevate HDL levels significantly), have the potential to become valuable therapeutic agents in the prevention and treatment of sepsis and septic shock.102 Preliminary studies seem to warrant study in children. Statins exert anti-inflammatory effects by modifying leukocyte– endothelial cell interactions and by altering the responses of monocytes/macrophages and T lymphocytes. Statins also modulate inflammatory cell signaling and inflammatory gene expression, thereby reducing the release of inflammatory cytokines. Statins also have important antioxidant effects through a variety of mechanisms, demonstrate antithrombotic properties, and appear to improve endothelial function by enhancing the expression of endothelial constitutive nitric oxide (NO) and by inhibiting apoptosis. Statins also directly inhibit major histocompatibility complex (MHC) class II expression induced by interferon-gamma (IFN-g), thereby modifying T-lymphocyte activity. These multiple anti-inflammatory properties may be responsible for the beneficial effects of statins observed in clinical trials in other inflammatory, immune-mediated diseases such as multiple sclerosis and rheumatoid arthritis.103 No data from randomized trials of statins and sepsis are available, but observational studies lend support to a potentially important preventive, and possibly a treatment, effect. The largest study to date is a population-based cohort study involving the linked administrative databases in Ontario, Canada, and included a matched cohort of 69,168 patients.104 The incidence of sepsis was substantially lower among patients receiving statins (hazard ratio (HR) 0.81; 95% CI 0.72 to 0.91). The protective association between statins and sepsis persisted in high-risk subgroups, including patients with diabetes mellitus, malignancy, and those receiving oral corticosteroids. Significant reductions in severe sepsis (HR 0.83; 95% CI 0.70 to 0.97) and fatal sepsis (HR 0.75; 95% CI 0.61 to 0.93) were also observed. Retrospective cohort data from two United States teaching hospitals provided information on 787 patients with a discharge diagnosis of pneumonia.105 Patients were deemed to be on a statin if it was noted to be prescribed at admission. Mortality at 30 and 90 days was 9.2% and 13.6%, respectively. After adjusting for confounding factors, including using propensity analysis (a statistical technique to ensure that patients in both groups are equally likely to be prescribed a particular intervention, in this instance a statin), its use at presentation was associated with an odds ratio for death at 30 days of 0.36 (95% CI 0.14 to 0.92). However, not all the data regarding statin use point towards a benefit. In one study, investigators assessed the effects of statin treatment before or during intensive care unit admission, in 438 patients ventilated for more than 96 hours.106 Although statin-treated patients showed a trend toward lower rates of infection and delay in onset, they had higher hospital mortality (61% versus 42%, P = 0.03).

Endotoxin Removal Enhancement of the clearance of endotoxin to reduce plasma levels has been proposed as a mechanism behind the anecdotal use of extracorporeal methods of endotoxin removal. These methods include plasmapheresis, exchange transfusion, hemofiltration, and hemadsorption. Despite many anecdotal reports of the use of these methods (particularly plasmapheresis or blood exchange) in patients with sepsis, there have been no well-controlled studies of their use, and

104

SECTION

A Fever, Bacteremia, & Septicemia

only a small number of patients have been recruited to these studies.107–110 Properly conducted studies are required to prove clinical efficacy. High-volume hemofiltration may be beneficial in critically ill adults. A large study comparing volumes of filtration per hour in acute renal failure patients (only about 10% of whom had sepsis) treated with venovenous hemofiltration demonstrated a significantly higher survival rate in patients treated with at least 35 mL/kg per hour of filtration.111 But, another study of 30 patients (including 8 children) with septic shock, 14 of whom were randomized to treatment by plasma exchange, showed no difference between the two groups.112 A study of plasma or whole blood exchange in meningococcal septicemia showed transient reduction in soluble TNF-a receptors, which rebounded following exchange. There was no influence on mortality.113 Intermittent high-permeability hemofiltration (a renal replacement modality) studied in patients with septic shock with multiple organ failure induced by septic shock showed exceptional removal of IL-6 but not TNF-a.114 A different approach using extracorporeal adsorption apheresis for elimination of endotoxin from plasma is under investigation. Use of polymyxin B or human albumin-coated cartridges has been applied to human subjects, demonstrating significant reduction in circulating endotoxin levels.115,116 In addition, other adsorbents such as DEAEcellulose mats have been studied in clinical trials. In one such study in 15 critically ill intensive care patients with sepsis, a significant reduction in plasma endotoxin levels from a median of 0.61 to 0.39 EU/mL (–35%) was achieved (P < 0.001).117 Long-term comparison of the initial and posttreatment levels after a series of 5 to 6 individual apheresis treatments also showed a statistically significant decline in circulating endotoxin, IL-6, CRP, fibrinogen, and an increase in cholesterol levels. Thus extracorporeal endotoxin removal may prove a promising therapeutic tool for patients suffering from bacterial sepsis and proven endotoxemia.

In these models TNF-a is the first cytokine to appear in the circulation. Infusion of either a relatively low dose of endotoxin in healthy people, or a lethal dose of live Escherichia coli in baboons, results in transient release of TNF-a, peaking after 90 minutes. There is a close positive correlation between the size of the bacterial challenge and the extent of TNF-a release. Other proinflammatory cytokines, including IL-1, are released shortly after the release of TNF-a, in conjunction with several anti-inflammatory mediators, in particular IL-10, IL-1ra, and soluble TNF receptors.119 In 1985, it was first recognized that TNF-a production has a pivotal role in the systemic toxicity elicited by high-dose endotoxin. Pretreatment with antiserum to TNF-a was seen to protect mice against the lethal effect of intravenous endotoxin.120 Two years later, a monoclonal anti-TNF antibody was reported to protect baboons against lethal gram-negative bacteremia.121 Since that time anti-TNF strategies have proved to be protective in several sepsis models in which bacteria or bacterial products were administered systemically as a bolus or a brief infusion. Neutralization of IL-1 activity, by administration of recombinant IL-1ra, was also seen to reduce lethality induced by endotoxin or living bacteria in various species. Importantly, administration of recombinant TNF-a or IL-1 to laboratory animals can reproduce most of the characteristics of the sepsis syndrome, and TNF-a and IL-1 have synergistic toxicity in experimental models.122 These landmark studies formed the basis of the design of several clinical trials that use TNF-a and IL-1 neutralizing agents. IL-10 represents an important autoregulatory mechanism that controls the production of proinflammatory cytokines and endotoxin toxicity in vivo. Indeed, neutralization of endogenously produced IL10 in endotoxemic mice was associated with an increased production of several proinflammatory cytokines, including TNF-a, and was associated with increased mortality. IL-10 gene-deficient mice showed increased mortality after endotoxin administration, together with high concentrations of TNF-a, IL-1, and other inflammatory mediators.123

Anticytokine Therapy Cytokine Physiology and Anticytokine Therapy Cytokine Physiology Cytokines have a central role in the pathogenesis of bacterial infection and sepsis. Cytokines coordinate a wide variety of inflammatory reactions at tissue level. They interact in a complex network in which they influence each other’s production and activity. The cytokine network can be roughly divided into a proinflammatory arm and an anti-inflammatory arm. Prominent proinflammatory cytokines are TNF-a and IL-1. Anti-inflammatory cytokines, of which IL-10 is a well-studied example, inhibit the synthesis of proinflammatory cytokines and exert several direct anti-inflammatory effects on different cell types. The action of proinflammatory cytokines can be further inhibited by naturally occurring soluble inhibitors, such as soluble TNF receptors type I and type II which inhibit TNF activity, soluble IL-1 receptor type II, and IL-1ra, which both inhibit IL-1 activity. The plasma concentrations of cytokines vary greatly in patients with sepsis. In general, proinflammatory cytokines can be detected in only a subset of patients, whereas anti-inflammatory cytokines and soluble inhibitors can be seen in virtually all patients with sepsis and even in healthy individuals. It has been argued that those patients who fulfill the clinical criteria for SIRS may not have detectable levels of proinflammatory cytokines in their circulation because they are studied late in the septic process.118 This may explain why the cytokines TNF-a, IL-1b, IL-12, and IFN-b, which according to animal models play a central role in the pathogenesis of septic shock, are not consistently correlated with disease severity or outcome in patients with septic shock. The kinetics of cytokine release has been studied in models of sepsis and systemic inflammation, induced by intravenous administration of either live bacteria or bacterial products such as endotoxin.

TNF-a. Recognition of the central role of TNF-a in the development of multiple organ failure and death in lethal systemic challenge models in animals led to strategies first to inhibit TNF-a activity. Most investigations studied the efficacy of anti-TNF antibodies, but several trials studied the use of recombinant soluble TNF receptors. Marshall pooled data from clinical trials that examined monoclonal antibodies directed against TNF-a and showed a significant 3.5% reduction in mortality.124 No such effect is seen in investigations that use TNF receptor constructs. In fact, in one study, patients who received a high dose of a dimeric type II TNF receptor had a higher mortality rate than placebo-treated patients.125 IL-6. In comparison with other cytokines, IL-6 (a mixed pro- and antiinflammatory cytokine) has been reported most consistently in patients with sepsis, although actual levels of IL-6 show considerable variation.126 Since the release of IL-6 can be triggered by TNF-a, elevated IL-6 levels may be useful in identifying those patients most likely to benefit from an anti-TNF-a approach. Afelimomab, the F(ab„)2 fragment of a murine monoclonal antibody to human TNF-a, effectively neutralizes TNF-a both in vitro and in vivo. In a large multicenter, placebo-controlled study of patients with severe sepsis and documented infection, in the group of patients with elevated IL-6 levels the mortality rate was 243 of 510 (47.6%) in the placebo group and 213 of 488 (43.6%) in the Afelimomab group. Using a logistic regression analysis, treatment with Afelimomab was associated with an adjusted reduction in the risk of death of 5.8% (P = 0.041) and a corresponding reduction of relative risk (RR) of death of 11.9%. Afelimomab resulted in significant reductions in TNF-a and IL-6 levels, and a more rapid improvement in organ failure scores compared with placebo.126 In a large study of children with meningococcal disease, IL-6 was the cytokine thought to be most closely correlated with severity of cardiovascular dysfunction, as evidenced by inotrope requirements,

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

The Systemic Inflammatory Response Syndrome (SIRS), Sepsis, and Septic Shock

and is postulated to be the most potent serum-derived myocardial depressant factor in meningococcal septicemia.127 IL-1. Efforts to inhibit IL-1 activity by continuous infusion of recombinant IL-1ra did not prove to reduce mortality significantly in several clinical trials. In one study of recombinant human IL-1 receptor antagonist (rhIL-1ra) in the treatment of 696 patients with severe sepsis and septic shock, the study was terminated after an interim analysis found that it was unlikely that the primary efficacy endpoints would be met. The 28-day, all-cause mortality rate was 33.1% (116/350) in the rhIL-1ra treatment group, whereas the mortality rate in the placebo group was 36.4% (126/346), yielding a 9% reduction in mortality (P = 0.36).128 This study confirmed the findings of an earlier study of 893 patients with sepsis syndrome which also failed to demonstrate any benefit of IL-1ra in this group of patients.129

Cytokines and Antibacterial Defense Mechanisms What is the reason for the remarkable discrepancy between the strongly protective effects of anti-inflammatory therapies in preclinical studies of sepsis and the absence of an effect in clinical trials with patients with sepsis? Although the detrimental role of proinflammatory cytokines in models of systemic inflammation induced by direct infusion of bacteria or bacterial products is not debated, it should be acknowledged that these models are associated with an acute syndrome, unlike many cases of clinical sepsis. Furthermore, these models do not have a localized infectious source, i.e., an infected organ or cavity, from which the infection disseminates. Most frequently, the primary site of infection in patients with sepsis is the lung, followed by the abdomen, and the urinary tract. Cytokine production takes place primarily at the site of the infection. Infection models that use an initially localized source of infection such as pneumonia and peritonitis have suggested that proinflammatory cytokines have a crucial role in host defense against bacterial infection. Neutralization of endogenous TNF-a during murine pneumonia caused by either gram-positive or gram-negative bacteria resulted in an accelerated course of the infection, and was associated with greater outgrowth of bacteria in the lungs, and decreased survival.130 Conversely, the elimination of IL-10 improved survival of murine pneumonia and reduced the bacterial load within the pulmonary compartment.131 Endogenous TNF-a also has a protective role in the pathogenesis of gram-negative peritonitis. This model provides insight into the seemingly paradoxical role of certain cytokines in severe bacterial infection: intraperitoneal administration of E. coli in mice results in a subacute syndrome, and eventually in systemic inflammation, multiple organ failure, and death. IL-10 gene-deficient mice given intraperitoneal E. coli showed an accelerated bacterial clearance from the peritoneal cavity and a reduced dissemination of the infection to distant organs. Nonetheless, systemic inflammation and multiple organ failure were more prominent and lethality was increased.132 Hence, the role of cytokines in acute models (bolus administration of endotoxin, or of bacteria) and more realistic infection models (e.g., pneumonia or peritonitis) can be completely different. The severity of the bacterial challenge seems to be a critical determinant of the effect of certain cytokines on survival. The action of proinflammatory cytokines appears important for an adequate antibacterial host response at the site of infection, whereas their systemic action can harm the host and cause tissue damage. Anti-inflammatory cytokines, such as IL-10, impair local antibacterial defenses, yet can diminish systemic toxicity produced by bacteria.

Immunoparalysis Induction of anti-inflammatory pathways to inhibit excessive proinflammatory activity can be demonstrated in most patients with sepsis. This has led to the concept of “compensatory antiinflammatory response,” following SIRS in time course.133 In addition,

CHAPTER

12

105

shortly after the onset of a septic event, a refractory state develops that is characterized by a relative inability of host inflammatory cells to respond to usually proinflammatory stimuli (such as endotoxin challenge).134 This adaptation or hyporesponsiveness of immune cells is not unique to sepsis, but is seen in other stressful conditions such as following trauma and surgery. It has also been referred to as “immunoparalysis” and anergy. The diminished responsiveness involves monocytes, granulocytes, and lymphocytes. Although the mechanisms that underlie immunoparalysis have not been completely explained, it is conceivable that anti-inflammatory cytokines, particularly IL-10 and transforming growth factor-b (TGF-b) are involved. Indeed, plasma from patients with sepsis greatly diminished the capacity of normal monocytes to secrete TNF-a.135 It has been proposed that immunoparalysis could contribute to the increased susceptibility to nosocomial infections and late mortality of patients who survive the acute phase of sepsis syndrome. As a result, strategies aiming to restore immune function have been developed and partially evaluated in patients with sepsis. Cytokines able to reverse monocyte deactivation in vitro and in animals, IFN-g and granulocyte–macrophage colony-stimulating factor (GM-CSF), are being studied.136,137

Arachidonic Acid Metabolism and Inhibitor Therapy Products of the cyclooxygenase (COX) and lipooxygenase pathways of arachidonic acid metabolism include leukotrienes, prostaglandins, and thromboxane. These products appear to play a major role in diminishing systemic vascular resistance and causing platelet aggregation, membrane lysis, and increased capillary permeability, which are the hallmarks of SIRS with shock. Drugs that interfere with these pathways have been tested as treatments for sepsis.

Ibuprofen In a large randomized study using ibuprofen in 455 patients with sepsis, treatment with ibuprofen reduced levels of prostacyclin (PGI2) and thromboxane and decreased fever, tachycardia, oxygen consumption, and lactic acidosis, but it did not prevent the development of shock or ARDS or improve survival.138 A later study that examined the effects of ibuprofen on the physiology and survival of hypothermic patients with sepsis revealed a significant reduction in the 30-day mortality from 90% (18 of 20 placebo-treated patients) to 54% (13 of 24 ibuprofen-treated patients). Compared with febrile patients, the hypothermic group had exaggerated response of cytokines TNF-a and IL-6 and of the lipid mediators thomboxane and PGI2.139 There is increasing evidence that COX-2 is involved in the endothelial dysregulation that occurs in sepsis.140 There is some recent evidence that COX inhibition may be beneficial in animal models of sepsis.141

Pentoxifylline Pentoxifylline, a xanthine derivative and a phosphodiesterase inhibitor, has been shown to have numerous potential beneficial effects in human and animal models of sepsis. Pentoxifylline suppresses the production of inflammatory mediators such as TNF-a in vitro. In adults and neonates it has been shown to decrease TNF-a, IL-1, and IL-10, but not IL-6 or IL-8.142,143 Pentoxifylline delays the release of endothelin-1, abolishes TNF burst and suppresses IL-6 and lactate, while improving survival from abdominal sepsis in rats. Pentoxifylline also augments hemodynamic performance during sepsis, improving renal blood flow during bacteremia and preventing transition from hyperdynamic to hypodynamic response during sepsis. Pentoxifylline has been shown to prevent endothelial cell dysfunction in sepsis, preserve endothelial thrombomodulin, protein C, and the protein S anticoagulant system, stimulate fibrinolysis associated with the increased release of tissue plasminogen activator (tPA) in sepsis, enhance PGI2 release, and attenuate the release of thromboxane.

106

SECTION

A Fever, Bacteremia, & Septicemia

There are few human controlled studies, 1 in surgical patients with severe sepsis, and 2 in neonates with confirmed sepsis. The surgical study included 51 adults with severe sepsis.144 This study showed an improvement in organ dysfunction score in patients receiving pentoxifylline compared with placebo, but was too small to assess change in mortality. The neonatal studies included 107 infants.145,146 There was a significant reduction in all-cause mortality during hospital stay in neonates with sepsis who received pentoxifylline as an adjunct to antibiotics compared with those who had placebo (typical RR 0.14; 95% CI 0.03 to 0.76). The results from these two studies showed a statistically significant reduction in mortality and a trend towards earlier correction of metabolic and hemodynamic derangements in preterm neonates with confirmed late-onset sepsis. However, because of methodological weaknesses of these studies, routine use of pentoxifylline cannot be recommended.

Intravenous Immunoglobulin Therapy Intravenous immunoglobulins (IGIV), like IFN-g and GM-CSF, can be regarded as a treatment method that seeks to improve the host defense. Although plasma immunoglobulin concentration may be reduced in patients with sepsis, the use of IGIV as therapy is not supported by large randomized clinical trials. Indeed, no individual well-designed trial has been undertaken in adults with sepsis. A small nonblinded study in 21 patients with streptococcal toxic shock syndrome showed a reduced mortality (6% versus 34%, P = 0.02).147 Additionally, a recent meta-analysis has suggested that IGIV could be beneficial in sepsis, although even in this analysis patient numbers were low.148

tation improves the hemodynamic condition of vasopressor-dependent septic shock.157 Shock reversal is hastened, vasopressor needs are reduced more rapidly, and this benefit is observed regardless of the duration of shock before hydrocortisone is instituted. What remains more controversial is the definition of adrenal insufficiency, the optimal dose and timing of corticosteroid supplementation, whether this should then be tapered slowly, and the impact of corticosteroid supplementation on outcome. An ongoing phase III, randomized, placebo-controlled, doubleblind trial is aimed at confirming the benefit from cortisol replacement in less severely affected patients in Europe and at clarifying treatment effects in patients who have presumably normal adrenal function Corticosteroid Therapy of Septic Shock (Corticus). A strategy that would maximize benefit while minimizing potential toxicity or unnecessary drugs would be the use of low-dose corticosteroids in patients with septic shock. Corticosteroids should probably be given for 1 week from onset of shock defined by the need of vasopressors. Corticosteroid administration should be preceded by an acute ACTH test. When the results of the test are available (preferably within a few hours) treatment should only be continued in patients with corticosteroid insufficiency defined by a random cortisol of 15 μg/dL or less, a peak cortisol of 20 μg/dL or less, or a cortisol increment of 9 μg/dL or less. Although no study has evaluated the efficacy of corticosteroids in children with sepsis, several well-designed trials conducted in children with bacterial meningitis, most of whom had bacteremia when enrolled, have shown that early administration of dexamethasone was associated with significant reduction in hemodynamic instability in the 6 hours after initiation of antibiotic therapy.158

Corticosteroids Since the 1960s investigators have attempted to modulate the inflammatory response to sepsis with short courses of corticosteroids, given at doses much higher than normal physiologic concentrations. These studies failed to show a beneficial effect of glucocorticoids in patients with sepsis; this lack of efficacy was confirmed in several meta-analyses.149,150 However, recent investigations involving a more limited number of patients indicate that glucocorticoids in more physiologic doses could be of benefit to patients with septic shock. These studies examined the effects of glucocorticoids given at doses from 10- to 100-fold less than doses administered in the older trials (i.e., at doses inducing fewer or no immunosuppressive effects) and were limited to patients with septic shock requiring vasopressor agents.151 Although the number of patients studied was too low (only 40 were included) to assess effect on survival, glucocorticoid use was associated with restoration of hemodynamic stability. One controlled study investigated the efficacy of stress-dose hydrocortisone (50 mg every 6 hours) plus 50 mg of fludrocortisone daily for 7 days in 299 patients who had septic shock for no longer than 8 hours. The target population included patients with an inadequate increase in serum cortisol after adrenocorticotropic hormone (ACTH) stimulation. A 30% relative reduction in mortality was seen in patients treated with corticosteroids.152 Adrenal failure is common in critical illness and in particular in vasopressor-dependent septic shock. High baseline total serum cortisol together with a low response to a corticotropin stimulation test (< 9 μg/dL cortisol increase 60 minutes after 0.25 mg of corticotropin) is correlated with a poor outcome in sepsis.153 Several studies in children and adults with septic shock have demonstrated abnormalities of control of adrenal corticosteroid secretion over the course of illness.154–156 Although it is rare to have severe adrenal insufficiency on admission, a relative deficiency of adrenal steroid secretory function has been demonstrated in sepsis, often associated with resistance to high doses of inotropic agents, suggesting that replacement doses of corticosteroids may be beneficial in some patients with refractory shock. Various randomized controlled trials comparing hydrocortisone to placebo have been performed in septic shock. Despite variable inclusion criteria, dose regimens and endpoints, there is general agreement that clinical studies show that hydrocortisone supplemen-

Anticoagulant Therapies Virtually all patients with sepsis have coagulation abnormalities. These abnormalities can vary from subclinical alterations in clotting times, to full-blown disseminated intravascular coagulation (DIC).

Tissue Factor Pathway Inhibitor (TFPI) Tissue factor (thromboplastin) is a major initiator of coagulation. It is a transmembrane cell surface receptor for plasma clotting factor VII, and exhibits homology with the cytokine receptor family. TFPI is an endogenous serine protease inhibitor, synthesized and secreted by endothelial cells, which inhibits factor Xa directly, and the factor VIIa–tissue factor catalytic complex in a Xa-dependent fashion. A significant portion of endogenous TFPI is bound to the microvasculature through low-affinity binding to glycosaminoglycans. This pool of TFPI is releasable into the circulation by exposure to heparin. A small pool of TFPI is stored in platelets, and secreted on activation and degranulation. Most of the circulating TFPI is bound to lipoproteins. The circulating concentrations of TFPI vary widely in healthy volunteers and in patients with sepsis. The functional properties of circulating TFPI are not well delineated. Endothelial damage is common in severe sepsis, as shown by the presence of coagulation abnormalities, including prolongation of prothrombin time, in most septic patients. It has been hypothesized that TFPI may protect the endothelium from coagulation and sepsisinduced injury. This is supported by preclinical studies in which exogenous TFPI improved outcome in septic animals.159 Importantly, inhibition of tissue factor by either tissue factor antibody treatment or infusion of recombinant TFPI, not only abrogated DIC in primates with severe bacteremia, but also prevented death. However, in the same sepsis model, intervention further downstream in the coagulation cascade by infusion of site-inactivated factor Xa did not affect survival despite complete protection against DIC.160 Thus, it has been suggested that tissue factor exerts effects on inflammatory mechanisms distinct from its effect on coagulation. There have been several phase I and phase II studies that have examined recombinant TFPI in patients with sepsis. In the first study, a greater than expected

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

The Systemic Inflammatory Response Syndrome (SIRS), Sepsis, and Septic Shock

effect on prothrombin time was seen in 3 patients receiving TFPI and this effect was associated with an increase in serious bleeding adverse events. As a result, in subsequent studies lower TFPI doses were given. A phase II study comparing placebo with 0.025 or 0.05 mg/kg per hour TFPI in patients with severe sepsis showed no difference with respect to adverse effects between treatment arms, and a trend towards a reduced mortality in TFPI-treated patients.161 In view of these data, a phase III randomized, double-blind, placebo-controlled, multicenter study evaluated the safety and efficacy of recombinant TFPI in patients with severe sepsis. The main outcome measure was all-cause 28-day mortality. The primary efficacy population consisted of 1754 adult patients with severe sepsis. There was no overall effect on 28-day mortality (recombinant TFPI group, 34.2% versus placebo group, 33.9%, P = 0.88).162

Antithrombin Antithrombin is another anticoagulant protein that inhibits a number of clotting factors, including thrombin and factors IXa and Xa. Antithrombin also inhibits the activity of products of the contact system, such as factors XIa and XIIa, and kallikrein. Apart from its anticoagulant effects, antithrombin has anti-inflammatory properties. It appears to modulate the inflammatory response by binding to the endothelium via cell surface heparin sulfate proteoglycans and may promote the release of PGI2. The anti-inflammatory effects of antithrombin are only seen at supraphysiologic concentrations and in the absence of heparin. In animal models of sepsis, antithrombin therapy is protective against lethality and organ failure.163 There have been several studies of the use of antithrombin in human sepsis that suggested that antithrombin therapy could be beneficial. In one investigation, continuous long-term antithrombin infusion was seen to attenuate SIRS, as indicated by reduction in the plasma concentrations of IL-6, soluble endothelial adhesion molecules (soluble E-selectin and soluble intercellular adhesion molecule-1) and a diminished acute-phase protein response (CRP).164 A randomized, prospective, placebo-controlled phase III multicenter clinical trial (KyberSept) was performed to test the efficacy of high-dose antithrombin therapy in patients with severe sepsis, and specifically examined patients concomitantly treated with heparin for prophylaxis against deep-vein thrombosis.165 From 2314 patients with severe sepsis (1157 placebo and 1157 antithrombin subjects, each), 1616 patients (811 placebo and 805 antithrombin subjects) received heparin concomitantly with study drug (antithrombin 30,000 IU) over 4 days, whereas 698 patients (346 placebo and 352 antithrombin) did not. In patients who did not receive concomitant heparin, 28-day mortality was reduced in the antithrombin group compared with placebo (37.8% versus 43.6%; absolute reduction: 5.8%; RR, 0.860 (0.725 to 1.019), which increased to day 90 (44.9% versus 52.5%; absolute reduction: 7.6%; endpoint: 0.851 (0.735 to 0.987)). In patients with concomitant heparin, no effect of antithrombin on mortality was seen (28-day mortality: 39.4% versus 36.6%; absolute increase: 2.8%; RR, 1.08 (0.96 to 1.22)). Treatment with high-dose antithrombin III may increase survival time up to 90 days in patients with severe sepsis and high risk of death. This benefit may even be stronger when concomitant heparin is avoided. Despite this large multicenter phase III trials failing to show a beneficial effect of antithrombin treatment overall in septic patients, it is conceivable that antithrombin may be of benefit to subsets of patients with sepsis, in particular those with severe sepsis who are not concurrently receiving heparin, which appears to neutralize the effect of antithrombin.

CHAPTER

12

107

during sepsis as a result of increased consumption of protein S and protein C, decreased activation of protein C by downregulation of thrombomodulin on endothelial cells and reduction of the endothelial protein C receptor (EPCR). Furthermore, protein S can be bound by the acute-phase protein C4b-binding protein, which reduces the bioavailability of protein S. The hypothesis that aPC could be beneficial in sepsis is based on a number of preclinical observations. Infusion of aPC into septic baboons prevented hypercoagulability and death, whereas inhibition of activation of endogenous protein C by a monoclonal antibody exacerbated the response to a lethal E. coli infusion, and converted a sublethal model into a severe-shock response associated with DIC and death.166 The monoclonal antibody prevented protein C from binding to the EPCR, thereby reducing protein C activation by the thrombin–thrombomodulin complex. Hence, like the tissue factor VIIa-mediated pathway, the protein C pathway appears to have other effects on host responses apart from its role in coagulation. One mechanism that could contribute to anti-inflammatory properties of the protein C pathway system is the capacity of protein C and protein S to inhibit endotoxin-induced production of TNF-a, IL-1, and IL-6 by monocytes in vitro, and the ability of aPC to reduce TNF-a release during endotoxemia in rats.167 Other anti-inflammatory effects of aPC include inhibition of rolling of monocytes and neutrophils on injured endothelium by inhibition of adhesion molecule binding, and an increase in the fibrinolytic response by inhibiting the acute-phase protein inhibitor of plasmin generation, plasminogenactivator inhibitor type 1 (PAI-1).168 In 2001, a large multicenter placebo-controlled trial reported on the efficacy of recombinant human aPC in 1690 patients with severe sepsis.52 This was the first study in more than 20 years of clinical sepsis trials to show a positive effect on 28-day all-cause mortality. The trial was designed to enroll 2280 patients, but was stopped prematurely by the data-monitoring committee because it met efficacy criteria. aPC was shown to reduce mortality significantly from 30.8% in the placebo group to 24.7% in the treatment group (P = 0.005), an absolute reduction in the risk of death of 6.1%. The incidence of serious bleeding was higher in aPC-treated patients (3.5% versus 2.0%, P = 0.06). aPC infusion was associated with a reduction in plasma D-dimer concentrations, evidence that aPC attenuated the procoagulant response. aPC also reduced plasma IL-6 concentrations, indicating inhibition of inflammatory responses. On the basis of this trial, aPC therapy is now recommended by the Surviving Sepsis campaign for adults with severe sepsis and organ failure who meet the inclusion criteria of the phase III aPC trial within 24 hours of admission and who do not have increased risk for bleeding.54 The efficacy of aPC has not yet been shown in patients with more moderate sepsis, and the safety of aPC in patients who are at high risk of bleeding is being assessed in further studies. Following the report of its benefit in adults, a large phase III randomized, placebo-controlled study of aPC was performed in children with septic shock. This study was stopped following the second interim analysis of the data. Despite enrolling 477 children with severe sepsis, this study failed to demonstrate any benefit in the primary endpoint (composite time to organ failure resolution) or any of the secondary endpoints (including 28-day mortality). In addition, there were significant concerns regarding potentially deleterious effects in children under 60 days of age who appeared to have an increased risk of hemorrhagic complications.169 So while there is clear evidence of benefit of aPC in adults with septic shock, until subgroups of children who may benefit from aPC can be identified clearly, the use of aPC in children is not recommended.

Activated Protein C (aPC) Activated protein C (aPC) is a natural anticoagulant due to its actions that proteolytically inactivate clotting factors Va and VIIIa. aPC is generated after an interaction of protein C with thrombin, bound to the endothelial cell surface protein thrombomodulin. aPC function is dependent on its cofactor, protein S. The activity of aPC is impaired

Plasminogen Activator Inhibitor (PAI) and Tissue Plasminogen Activator (tPA) The finding of a direct relationship between PAI-1 levels and mortality in meningococcal disease has led to the proposal of the use of fibri-

108

SECTION

A Fever, Bacteremia, & Septicemia

nolytic therapy for meningococcal disease. There have been anecdotal reports of the successful use of tPA in children with meningococcal disease.170 Understandably there are concerns regarding the potential for catastrophic bleeding in patients with a hemorrhagic diathesis. A review of the use of tPA as rescue therapy in children with meningococcal disease has demonstrated an unacceptable level of adverse events, including fatal intracranial hemorrhage.171 A recent study has demonstrated beneficial effects in mice treated with an inhibitor of PAI-1.172 This may be a future direction for research as it is likely not to be associated with the hemorrhagic complications of fibrinolytic therapy. Because of the recognized interactions between inflammation and coagulation, manipulation of the coagulation cascade would appear to be an attractive target for new therapies.

Therapies Targeting the Endothelium Endothelial dysfunction appears to be pivotal as the primary pathologic feature of severe sepsis. This is the main target organ of injury affected by the release of multiple inflammatory mediators defined above, and the processes causing upregulation of the coagulation and complement systems. Agents to restore endothelial function by interventions to reduce endothelial cell injury and dysfunction are being developed and studied in sepsis.

Platelet Activating Factor (PAF) PAF is a phospholipid produced by macrophages, neutrophils, platelets, and endothelial cells. PAF increases cell adhesion, activates endothelial cells, and amplifies release of cytokine mediators. A phase II placebo-controlled study of the PAF receptor antagonist, BB-882, was carried out in 152 patients with severe infection.173 There was no effect on hemodynamic, respiratory, or oxygen transport variables or mortality in each group. However, another study of the use of TCV309, a different PAF antagonist, in 98 patients with septic shock revealed a substantial reduction in organ dysfunction and morbidity, although there was no change in overall mortality.174 In a further study of a different PAF receptor-blocking agent (BN52021) in 609 patients with severe sepsis, mortality was 50% in the placebo group and 44% in the treated group (P = 0.29).175 PAF acetylhydrolase (PAF-AH) is a member of the phospholipase A2 family of enzymes. The extracellular form of PAF-AH is a secreted plasma protein that serves to inactivate PAF and other oxidized phospholipids that can produce PAF-like effects.176 The therapeutic rationale for the administration of recombinant PAF-AH (rPAF-AH) in severe sepsis is to increase PAF-AH activity in the presence of generalized inflammation and coagulation. The therapeutic potential for this strategy was supported by the results from a phase II trial of rPAF-AH in 127 patients with severe sepsis.177 The results from this study showed that 28-day all-cause mortality was 21% in the 1.0 mg/kg rPAF-AH group, 28% in the 5.0 mg/kg rPAF-AH group, and 44% in the placebo group (overall c2 P = 0.07; 1.0 mg/kg rPAFAH versus placebo, P = 0.03). A trend toward reduced multiple organ dysfunction was also observed in the 1.0 mg/kg rPAF-AH group compared with the placebo group (P = 0.11). Following this a phase III trial was undertaken to confirm these results in patients at risk for ARDS and mortality from severe sepsis;178 2522 patients were planned to be enrolled in a prospective, randomized, double-blind, placebo-controlled, multicenter, international trial. Eligible patients were randomized to receive either rPAF-AH 1.0 mg/kg or placebo. The study was terminated after the second planned interim analysis, and the enrollment of 1261 patients (618 placebo and 643 rPAF-AH subjects). The study showed no improvement in 28-day all-cause mortality in the rPAF-AH group compared with placebo (25% for rPAF-AH versus 24% for placebo; RR, 1.03; 95% confidence interval (CI), 0.85 to 1.25; P = 0.80). There were no statistically significant differences between treatment groups in any of the secondary efficacy endpoints. One finding in this study was that plasma levels of PAF-AH changed during the course of disease, with higher levels in survivors

without organ failure compared with those who die, and higher levels still in those with severe septic shock and multiple organ failure, suggesting that higher levels may not be good, but that dynamic changes may be more important than absolute levels.179

Neutrophil/Endothelial Cell Interactions Many proinflammatory mediators stimulate adhesion molecule expression on leukocytes, platelets, and endothelial cells. There is much evidence to implicate the adhesion of neutrophils to endothelial cells in the tissue injury and multiple organ dysfunction that occurs during sepsis.180 Patients with inherited abnormalities of adhesion molecules have recurrent, severe infections typically characterized by a marked leukocytosis, and may develop systemic sepsis and septic shock.181 This highlights the important role these molecules have in host defense. While there are in vitro data suggesting the importance of these interactions in sepsis, the inhibition of adhesion molecules in animal and human studies of sepsis shows results which are consistent with their important role in host defense. That is, these studies have shown either no benefit, or have demonstrated a worse outcome in the treatment groups.182,183

Nitric Oxide Balance Activation of the inflammatory response results in elaboration of a number of mediators with direct effects on vasomotor tone. NO, bradykinin, histamine, and PGI2 can all decrease vascular tone and cause hypotension. NO is formed by the enzymatic action of NO synthase (NOS) on the guanidino group of the amino acid L-arginine.184 The inducible isoform of NOS (iNOS) is produced in response to endotoxin, PAF, IL-1b, and TNF-a. Glucocorticoids, IL-1ra, PAF antagonists, TNF-a, tyrosine kinase inhibitors, and dihydropyridine calcium-channel blockers inhibit iNOS induction.185 NO is a highly diffusible compound that activates soluble guanylate cyclase in smooth-muscle cells. This converts guanosine triphosphate (GTP) to cyclic guanosine monophosphate (cGMP) which relaxes the smooth-muscle cell via a protein kinase, by promoting calcium entry into the sarcoplasmic reticulum.186 It appears that iNOS is the predominant source of the excessive NO production responsible for the hypotension and profound and refractory vasodilatation frequently observed in septic shock. Normally, vasomotor tone is tightly regulated through the combination of NO generation by vascular endothelial cells being rapidly inhibited by binding to circulating hemoglobin in red blood cells. NO is highly soluble in water and may react with the superoxide ion to form peroxynitrite which is highly toxic and relatively stable. This moiety then reacts with water to form nitrites and nitrates which can be measured as a surrogate indicator of NO production. High levels of nitrates and nitrites have been detected in sepsis models and patients with septic shock.187 The inflammatory response in sepsis, including increased NO production by iNOS, may result in endothelial cell dysfunction affecting vascular smooth muscle. The resulting effects on organ perfusion may be instrumental in the pathogenesis of the multiple organ dysfunction syndrome seen in sepsis and septic shock, which is associated with increased morbidity and mortality. The role of iNOS and cGMP in the vasculopathy of septic shock have been supported by the finding that competitive NOS inhibitors, such as L-N monomethyl arginine (LNMMA) and N(G)-nitro-L-arginine methyl ester (L-NAME), act as vasopressors when administered to patients with septic shock.188,189 The increased NO resulting from iNOS induction may contribute to the myocardial depression and a-adrenergic hyporesponsiveness associated with sepsis. The NO-induced production of cGMP in cardiac myocytes inhibits the a-adrenergic-stimulated increase in the slow calcium channel and decreases the affinity of calcium for the contractile apparatus. This results in a negative inotropic effect and increases the relaxation phase of the cardiac cycle.190 The implication of NO in the vascular hyporesponsiveness and cardiac depression of

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

The Systemic Inflammatory Response Syndrome (SIRS), Sepsis, and Septic Shock

sepsis support the hypothesis that blockage or reduction of NO production will produce clinical benefit in patients with sepsis. There are many animal models of sepsis in which various inhibitors of NO production have demonstrated potential benefit as well as potentially harmful effects. It has become clear however that nonspecific NOS inhibitors cause detrimental effects secondary to reduced organ perfusion, elevation of pulmonary artery pressures, and increased renal vascular resistance.191,192 This is likely to be due to inhibition of baseline NO production which is essential for control of organ perfusion under normal circumstances. In addition, there is evidence of increased capillary permeability and intestinal damage associated with L-NMMA after endotoxin challenge, together with a decrease in cardiac index and tissue oxygen delivery.185 This leads to an increase in lactic acidosis and hepatic toxicity. Therefore, animal studies have determined that reduction of NO activity is associated with the potential benefit of improvement of hypotension and vasodilatation, but at the expense of reduction of cardiac output and tissue oxygen delivery and with an increase in pulmonary vascular resistance and subsequent mortality. Despite these major concerns, several human studies have been carried out. All of these have shown similar effects to those demonstrated in animal models.189,193 In addition, concerns have been raised over activation of intravascular coagulation.194 A phase II multicentered, randomized, placebo-controlled, safety, and efficacy of the NO synthase inhibitor 546C88 (N,G-methyl-L-arginine hydrochloride) was performed in 312 adults with septic shock.195 The conclusion of this study was that this nonselective NOS inhibitor can reduce the elevated plasma nitrate concentrations observed in patients with septic shock. But treatment was also associated with an increase in vascular tone and a reduction in both cardiac index and oxygen delivery. There were no substantive untoward consequences accompanying these hemodynamic effects.195 Following this study an international, randomized, double-blind, placebo-controlled phase III study was conducted in patients with septic shock.196 The objective was to assess the safety and efficacy of 546C88 in patients with septic shock. The predefined primary efficacy objective was survival at day 28. A total of 797 patients with septic shock diagnosed for < 24 hours were recruited. The trial was stopped early after review by the independent data safety-monitoring board. The 28-day mortality was 59% (259/439) in the 546C88 group and 49% (174/358) in the placebo group (P < 0.001), with a higher proportion of cardiovascular deaths and a lower incidence of deaths caused by multiple organ failure in the 546C88 group. The recent development of selective iNOS inhibitors such as S-methylisothiourea (SMT) and TGF-b which inhibit iNOS mRNA, and their application in animal models of septic shock, suggest that these agents may offer the benefits of reduced NO production due to iNOS inhibition, without the adverse effects of nonselective NOS inhibition.197,198

INNATE IMMUNE RESPONSES AND TOLL-LIKE RECEPTORS (TLRS) The most exciting new development in sepsis research in the past years is the discovery of TLRs as signal-transducing elements of multiple antigens and the rapidly unfolding picture of TLRs as essential players in the innate immune response to infection.199 On first encounter with a pathogen, the innate immune system can distinguish between different classes of pathogenic bacteria, viruses, and fungi. Additionally, the innate immune response is vital for activating the slower-acting adaptive immune system. The innate immune system can recognize conserved motifs on pathogens that are not seen on higher eukaryotes. These motifs have been referred to as “pathogen-associated molecular patterns” or PAMPs, whereas their binding partners on immunocompetent cells have been termed “pattern recognition receptors.” Endotoxin, for example, interacts with cells via the pattern recognition receptor CD14. Spontaneous binding of endotoxin to

CHAPTER

12

109

CD14 happens at very slow rates. Lipopolysaccharide CD14-binding is greatly accelerated in the presence of LBP, an acute-phase reactant mainly derived from the liver. CD14 does not have an intracellular domain; cells respond to endotoxin via signaling through TLR4, which requires the presence of a secreted protein, MD-2. TLR2 in turn is essential for signaling the proinflammatory effects of the bacterial lipoproteins, peptidoglycan and zymosan, whereas TLR5 mediates cellular effects induced by bacterial flagellin, and TLR9 mediates effects induced by unmethylated CpG-containing oligonucleotides present in bacterial (but not eukaryotic) DNA. Different members of the TLR family can act together in activating cells in response to pathogens, e.g., TLR2 and TLR6 cooperate in detecting certain bacterial components, including peptidoglycan.200 The in vivo relevance of induction of an effective innate immune response to infection has been shown with specific-TLR-deficient mice. TLR2 knockout mice are highly susceptible to gram-positive infection, whereas TLR4 knockout mice have reduced resistance to gramnegative infection.201 Designing methods to neutralize microbial products or block their interaction with specific receptor on immune cells is an attractive concept. Potential targets include LBP, CD14, TLR4, and MD-2 for gram-negative sepsis, and CD14, TLR2, and TLR6 for Gram positive sepsis. Monoclonal antibodies against CD14 have been evaluated in a phase I study;202 16 healthy subjects received an intravenous injection of LPS (4 ng/kg) preceded by IC14, a recombinant chimeric monoclonal antibody against human CD14. IC14 attenuated LPSinduced clinical symptoms and strongly inhibited LPS-induced proinflammatory cytokine release, while delaying the release of the anti-inflammatory cytokines, soluble TNF receptor type I, and IL-1 receptor antagonist. IC14 also inhibited leukocyte activation, but more modestly reduced endothelial cell activation and the acute-phase response. The capacity of circulating monocytes and granulocytes to phagocytose E. coli was only marginally reduced after infusion of IC14. These data provided the first proof of principle that blockade of CD14 is associated with reduced LPS responsiveness in humans in vivo. A further phase I study was performed in patients with septic shock.203 This study was performed to evaluate the safety, pharmacokinetics, pharmacodynamics, and clinical pharmacology of IC14 in a randomized, double-blind, placebo-controlled, dose-ranging study in 46 patients with severe sepsis. IC14 did not induce antibody formation or increase the incidence of secondary bacterial infection. The pattern of pro- and anti-inflammatory cytokines, chemokine, soluble receptor, soluble E-selectin, and acute-phase proteins in response to treatment was highly variable by patient and IC14 treatment group. The results suggest that CD14 blockade with IC14 warrants further clinical investigation to determine its ability to attenuate the proinflammatory response due to infection. Several intracellular signaling molecules, such as MyD88 and the mitogen-activated protein kinase, are other possible therapeutic targets. However, inactivating molecules that are pivotal to innate immunity can be harmful, as shown by the increased sensitivity to bacterial sepsis in mice with mutations of the Tlr4 gene.204 Careful selection of patients with severe infections associated with a high probability of death will therefore be essential. Another recent discovery of interest is high-mobility group (HMG)-1, a protein previously known as DNA-binding protein, which regulates gene transcription and stabilizes nucleosome formation. HMG-1 has recently been described as a “late” mediator of endotoxin toxicity.205 Importantly, postponed administration of antibodies against HMG-1 reduced endotoxin-induced lethality, whereas administration of HMG-1 was lethal.206 Furthermore, patients with sepsis have raised concentrations of HMG-1 in their circulation. These first data are promising and warrant further investigation into HMG-1 as a therapeutic target. Macrophage migration inhibitory factor (MIF) is a cytokine that has been shown to be important in innate immunity and sepsis.207 It is constitutively expressed in large amounts by immune, endocrine, and epithelial cells and is released after exposure to microbial products and

110

SECTION

A Fever, Bacteremia, & Septicemia

proinflammatory cytokines. Macrophage MIF has been shown to regulate innate immune responses to endotoxin and gram-negative bacteria by modulating the expression of TLR4.208 High levels of macrophage MIF have been detected in patients with severe sepsis and septic shock.209 Immunoneutralization of macrophage MIF or deletion of the Mif gene protects mice against lethal endotoxemia, gram-positive toxic shock syndromes, and experimental bacterial peritonitis. Conversely, mice injected with macrophage MIF together with live bacteria or microbial toxins have increased death rates.209,210 MIF has been described as a good predictor of poor outcome in sepsis, and efforts to modulate its production or action may be important as therapeutic modalities for sepsis.211

FUTURE CONSIDERATIONS The publication of the human genome will lead to massive advances in genomics and proteomics in the coming decade. The possibilities for individualized drug treatment of patients with sepsis, related to their genotype, will become reality. New technology may soon allow bedside testing of patients’ genotypes or determination of protein or peptide biomarkers associated with poor outcome, to allow targeted therapy of even the sickest patients. It is probable that many new agents will be developed based on the unraveling of the host–pathogen interaction. However, until this time we must utilize currently available therapies to the best of our knowledge. Despite huge advances, our treatment of sepsis is still dependent upon administration of appropriate antibiotics, intravenous fluid support, and relatively crude methods of organ support. We can only improve upon current treatment of pediatric sepsis after there is agreement that properly conducted multicenter clinical trials can and must be carried out in critically ill children in order to test new therapies. To reach this goal, we should model pediatric sepsis trials after the successful clinical trial program that has so greatly improved survival of childhood cancer. There have only been three large properly controlled phase III studies in children with sepsis, none of which has recruited adequate numbers to determine efficacy definitively. Although these and all the many adult studies except one have failed to demonstrate a significant survival advantage, there is much that can be learned from these unsuccessful studies that is relevant to the design of future sepsis trials. Children with severe sepsis and shock should be enrolled in double-blind, placebo-controlled studies to evaluate new treatments. These studies should be large enough to minimize random error and avoid type II error (or false-negative results). Definitions for the target population should be explicit, reproducible, and include illness severity scores. Protocols for both the use of the investigational agent and conventional treatment should be standardized. Outcomes should be clinically relevant and predefined, and should include measures of both benefit and harm.6 In addition, the analysis of results should be carried out, both on evaluable patients and on the intent-to-treat population. Finally, a health economic evaluation of the implications of the introduction of ever-increasingly expensive therapies should be mandatory. Only in this way will we be likely to influence further the unacceptably high mortality rate of severe sepsis in children, with the added advantage of limiting the widespread use of extremely expensive new therapies that have been insufficiently evaluated.

CHAPTER

13

Toxic Shock Syndrome James K. Todd

Toxic shock syndrome (TSS) was originally described in 1978 in seven children (four girls and three boys) with Staphylococcus aureus infections.1 TSS is an acute, febrile, exanthematous illness involving multiple systems with potential complications that include shock, renal failure, myocardial failure, and the adult respiratory distress syndrome (ARDS).1,2 Although there is a greater risk of the syndrome occurring in menstruating females who are using tampons,3 TSS can occur in nonmenstruating females and in males of any age, including children.1,4 Despite its severity, TSS can be treated effectively if recognized early.5 A similar toxic shock-like syndrome has been seen in conjunction with group A streptococcal infection.6,7 Differentiating features of these conditions and complications are compared with shock caused by gram-negative organisms in Table 13-1. The clinical presentation and principles of management for shock caused by the various gram-positive organisms are, for the most part, the same as shown in Box 13-1.

STAPHYLOCOCCAL TOXIC SHOCK Etiology and Pathogenesis TSS, as originally described, is caused by certain strains of S. aureus that produce one or more exotoxins (e.g., TSS toxin 1 (TSST-1), staphylococcal enterotoxin; see Chapter 115, Staphylococcus aureus).8 Data suggest that these toxins act as superantigens that stimulate lymphocytes and endothelial cells to produce endogenous mediators (e.g., tumor necrosis factor (TNF) and interleukin 1 (IL1)).9–12 Inflammatory mediators cause extensive capillary leakage, leading to loss of intravascular volume and hypotension secondary to low peripheral resistance, despite high cardiac output.13 The severity of associated multiorgan system dysfunction (i.e., cerebral, hepatic, renal, and pulmonary) is directly related to the degree of hypotension.2 A better understanding of the conditions (i.e., increased level of protein, aerobic PO2, neutral pH, increased level of CO2) of S. aureus growth in focal sites of infection in patients with TSS led to observations in vitro that strains of S. aureus capable of producing TSST-1 do so only in the presence of these specific growth conditions.14 Such conditions are found classically in focal staphylococcal infections, such as abscesses, or they can occur in the vagina during menstruation with the insertion of a tampon or other device, such as a contraceptive sponge or diaphragm, which adds oxygen to an otherwise anaerobic environment.15 Presumably, individuals who have previously been exposed to lesser concentrations of TSS toxins (e.g., from nasal carriage) may develop antibody that protects from overt disease: 90% of healthy adults have antibodies to TSST-1.16,17

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Toxic Shock Syndrome

CHAPTER

13

111

TABLE 13-1. Comparison of Common Bacterial Shock Syndromes Gram-Positive Bacteria

Gram-Negative Bacteria

GENERAL FEATURES

Organism Toxin Sex Age Epidemiologic associationsa

Staphylococcus aureus TSST-1, enterotoxin(s) Males and females All Menstruation

Streptococcus pyogenes Pyrogenic exotoxins Males and females All Varicella infection

Neisseria meningitidis and others Endotoxin Males and females All Complement deficiency

Uncommon Scarlatiniform “Strawberry” Mild Usually present (abscess, menses with tampon use)

Common Scarlatiniform “Strawberry” Mild In many cases (cellulitis, necrotizing fasciitis)

Common Purpuric Normal appearance Severe Often absent

Initial antibiotic

Vancomycin plus clindamycin

Vancomycin plus clindamycin

Fluid management Surgical intervention

Aggressive Drain any focus

IGIV Corticosteroids

Considered in severe cases Considered in severe cases

Aggressive Repeated debridement (fasciitis, necrotic lesions) Considered in severe cases Considered in severe cases

Third-generation cephalosporin, aminoglycoside Aggressive Late (if necrosis)

CLINICAL FEATURES

Bloodstream infection Rash Tongue Coagulopathy Focal infection TREATMENT

Efficacy unknown Efficacy unknown

IGIV, immune globulin intravenous; TSST-1, toxic shock syndrome toxin 1. a In some cases.

Epidemiology Although early case-control studies demonstrated a strong correlation of TSS with menstruation, and use of a particular highly absorbent tampon, ultimately leading to the withdrawal of Rely tampons from the market in 1980,3,18,19 other investigators subsequently pointed out inherent study design flaws that may have resulted in biased results.20 Additional studies suggested that high absorbency, rather than tampon composition, may be the critical factor.21,22 In the first epidemiologic study of TSS that used active case ascertainment, 60% of cases occurred in young women (aged 14 to 25 years) who were menstruating, and 40% occurred in males and in females who were not menstruating.23 Although subsequent passivereporting surveys suggested a dramatic decrease in menstruationassociated TSS since the early 1980s,24 active case ascertainment studies show little overall change in incidence but suggest that TSS is being recognized earlier and treated more effectively.25,26 Our current understanding of the pathogenesis of TSS also explains its epidemiology.15 TSS is more likely to occur in younger individuals who have not encountered the toxins previously and, thus, have no neutralizing antibody. Young women (often still in the pediatric age group) who have vaginal colonization with toxin-producing S. aureus but no antibody to TSST-1 are at higher risk of developing TSS during menstruation (presumably because menstrual blood adds protein and neutralizes the normally acid pH of the vagina), especially with tampon use.15 Similarly, individuals (either male or female) with focal infection or surgical wound infection secondary to S. aureus can develop TSS. Sometimes the focal infection, such as sinusitis, is occult.

Clinical Manifestations TSS should be considered in the differential diagnosis of fever and erythroderma or fever and signs of hypotension.27 Symptoms begin with acute onset of fever, sore throat, and myalgia; profuse diarrhea is common, and vomiting may occur.1,28 Erythroderma or scarlatiniform rash that is more prominent on the trunk than the extremities is common except in the presence of severe hypotension. Symptoms of hypotension include orthostatic dizziness, fainting, or overt shock; orthostatic changes in blood pressure can be measured before overt hypotension occurs. Nonpurulent conjunctival

BOX 13-1. Management of Toxic Shock Syndrome PRIMARY DISEASE • Intravenous fluids to maintain adequate cardiac preloada • Drainage of focal infection(s), debridement of necrotic tissue • Antistaphylococcal antibiotic(s) COMPLICATIONSa • Adult respiratory distress syndrome: intubation, positive airway pressure • Myocardial failure: dopamine, dobutamine • Renal failure: judicious fluid restriction,a dialysis SEVERE OR UNRESPONSIVE ILLNESS • Consider intravenous immune globulin • Consider methylprednisolone a Avoid restriction of fluids unless central venous or pulmonary wedge pressure is high.

hyperemia, pharyngeal inflammation, and “strawberry tongue” are expected findings on physical examination. Hypotensive patients can have nonfocal central nervous system abnormalities, especially altered consciousness. Complications of TSS are usually related to the severity of hypotension and include renal failure, ARDS, and ischemia of the extremities, sometimes resulting in gangrene. A “stunned myocardium,” which is, presumably, a transient toxic cardiomyopathy, can occur during or following correction of hypotension.29 Seven to 21 days after onset of illness, desquamation of the digits, palms, and soles occurs, commonly in a full-thickness, sheetlike manner. Hair loss and nail pitting or grooving can also occur. Laboratory test results are commonly abnormal in patients with TSS and mirror the severity of hypotension.2 Most patients have an elevated white blood cell count with an increase in mature and immature neutrophils. Abnormalities of hepatic and renal function are prominent. Hepatic dysfunction manifests with mild elevations of serum transaminase levels and disproportionately elevated conjugated bilirubin levels; hydrops of the gallbladder is not uncommon. Elevation of the creatine kinase level can be extreme. Consumption coagulopathy can occur, but isolated thrombocytopenia is more common. The collaborative definition of TSS was initially devised by investigators for epidemiologic purposes (Box 13-2).3 Cases of lesser severity that do not meet the full definition probably occur frequently,

112

SECTION

A Fever, Bacteremia, & Septicemia

BOX 13-2. Definition of Staphylococcal Toxic Shock Syndrome MAJOR INCLUSIONARY CRITERIA (ALL REQUIRED) • Acute fever ≥ 38.8°C • Hypotension (orthostatic hypotension, syncope, or shock) • Rash (macular erythroderma ˜ late desquamation) MINOR INCLUSIONARY CRITERIA (ANY THREE REQUIRED) • Mucous membrane inflammation (conjunctival, pharyngeal) • Gastrointestinal abnormalities (vomiting, diarrhea) • Muscle abnormalities (myalgia, elevated creatine kinasea) level • CNS abnormalities (nonfocal: coma, obtundation) • Hepatic abnormalities (elevated bilirubin, transaminasea) • Renal abnormalities (urinalysis with WBC count > 5/hpf, elevated blood urea nitrogena) level • Decreased platelet count (< 100,000/mm3) EXCLUSIONARY CRITERIA • Absence of other explanation • Negative blood cultures (except for Staphylococcus aureus) CNS, central nervous system; hpf, high-power field; WBC, white blood cell. a More than twice the upper limit for age.

especially with earlier recognition and treatment.26 The clinical definition requires both inclusionary and exclusionary criteria. Signs of multiorgan system dysfunction are usually proportionate to the degree of hypotension.2 The isolation of S. aureus is not required for the diagnosis of TSS, although every effort should be made to identify and drain any focus of infection. Toxin testing of S. aureus isolates has limited value in establishing the diagnosis of TSS, because multiple toxins have been associated with disease, and many unaffected individuals are colonized with toxin-producing strains. Acute and convalescent antibody titers to TSST-1 also have limited value, because TSS can suppress the formation of antibody.

tibilities of the causative organism are known. Clindamycin is useful adjunctive therapy to inhibit toxin production in severe disease. With aggressive fluid management, drainage of purulent collections, and appropriate antistaphylococcal antibiotic therapy, most patients with TSS improve promptly.5 The pathogenesis and treatment of toxic shock caused by gram-positive organisms differ from that of endotoxic shock caused by gram-negative organisms. In severe cases of gram-positive toxic shock, adjunctive corticosteroid therapy with methylprednisolone or administration of immune globulin intravenous (IGIV) may be of benefit.5,30 In severe or unresponsive TSS, complications related to organ failure require specific interventions. Initial renal failure usually results from decreased perfusion and resolves when vascular volume is replenished. Rarely, acute tubular necrosis develops, necessitating fluid restriction and dialysis. Myocardial dysfunction is evidenced by decreased fractional contraction as measured by means of echocardiography and increasing central venous or pulmonary wedge pressure; these symptoms are frequently responsive to inotropic therapies, such as dopamine or dobutamine.13,29 CA-MRSA infection can be complicated additionally by septic embolization to the lung and necrotizing pneumonia. ARDS usually develops several days after hospitalization, often while other manifestations of TSS improve. Diminishing oxygenation or pulmonary infiltrates on chest radiograph are often not the result of excessive fluid administration; pulmonary capillary leakage is the more likely mechanism. Intubation and use of positive end-expiratory pressure often decrease leakage and improve oxygenation. Restricting fluids in patients with presumed cardiac or respiratory “failure” is often counterproductive and should only be undertaken with vigilant monitoring of vascular volume because ongoing capillary leakage resulting in decreased venous return can further compromise cardiac output and tissue perfusion.

Outcome Differential Diagnosis The differential diagnosis of TSS includes exclusion of septic shock, staphylococcal exfoliative syndromes, streptococcal toxic shock, scarlet fever, Rocky Mountain spotted fever, viral hemorrhagic shock, Stevens–Johnson syndrome, leptospirosis, and measles.27 (See Chapter 15, Mucocutaneous Symptom Complexes, for an approach to the distinction of these diseases.) Although Kawasaki syndrome has similar physical findings, including desquamation of the palms and soles during convalescence, it is usually differentiated by a prolonged course of fever, absence of diarrhea, and lack of hypotension.

Management Capillary leakage with decreased vascular volume is the primary pathogenic mechanism in TSS. Patients often present with orthostatic hypotension that is responsive to vigorous administration of intravenous fluids. Rapid fluid resuscitation is followed by a volume of maintenance fluid in excess of calculated requirements because capillary leakage is likely to persist5 (see Box 13-1). Loss of intravascular fluid into the soft tissues, resulting in edema, is expected and should not be taken as a priori evidence of “overhydration”; intravascular volume can still be depleted. Central venous pressure monitoring may be helpful for severely affected patients. Because TSS is directly related to the effect of toxin produced by S. aureus, any focus of infection should be identified and drained promptly. In the case of a menstruating female, a tampon or other vaginal foreign body should be removed and the vagina irrigated. Many toxin-producing strains of S. aureus are susceptible to b-lactamase-resistant penicillins, such as nafcillin, and cephalosporins, such as cefazolin (methicillinsusceptible S. aureus, MSSA). However, in areas where methicillinresistant S. aureus (MRSA) is increasingly community-associated (CA), vancomycin should be given until the antimicrobial suscep-

Most pediatric patients who are treated promptly and aggressively survive TSS; the mortality rate of MSSA-associated disease is < 5%. Because the primary immune response to TSST-1 can be blunted in the presence of large concentrations of toxin, recurrence is possible within the next several menstrual cycles in young women who continue to use tampons or who are not treated with antistaphylococcal antibiotics.2 Menstruating women should not use tampons for the next five menstrual cycles and, thereafter, are encouraged to change tampons frequently and use pads instead of tampons at night. Recurrent episodes are usually less severe. Anticipatory guidance in general healthcare in prepubertal and pubertal years should include instruction to parents and girls that inserted vaginal foreign bodies should be removed immediately and medical attention sought if fever, especially with rash or dizziness, occurs during menstruation.

STREPTOCOCCAL TOXIC SHOCK-LIKE SYNDROME In the early 1990s, there was an increase in the incidence of severe invasive disease caused by Streptococcus pyogenes, characterized by a resurgence of severe scarlet fever, necrotizing cellulitis and fasciitis, and toxic shock-like syndrome (see Chapter 118, Streptococcus pyogenes).6,7,31,32 Shock caused by staphylococcal or streptococcal organisms now approach the incidence of shock caused by gramnegative organisms in many hospitals.26 Most cases resulting from infection with group A streptococcus (GAS) have been due to strains of M types 1 or 3.33 These strains have been shown to produce pyrogenic exotoxins that are epidemiologically associated with scarlet fever as well as toxic shock-like syndrome, and a protease that is associated with necrotizing cellulitis and fasciitis.34 The streptococcal pyrogenic exotoxins share some DNA homology with staphylococcal enterotoxins. They act similarly as superantigens that stimulate host cells to produce TNF and IL-1, causing capillary leakage and shock.11

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Hemophagocytic Lymphohistiocytosis and Macrophage Activation Syndrome

BOX 13-3. Definition of Streptococcal Toxic Shock-Like Syndrome Hypotension or shock, plus two or more of the following: • Renal impairment • Disseminated intravascular coagulation • Hepatic abnormalities • Adult respiratory distress syndrome • Scarlet fever rash • Soft-tissue necrosis A definite case meets the above requirements, and additionally group A streptococcus is isolated from a sterile body site. A probable case meets the above requirements, and group A streptococcus is isolated from a nonsterile body site.

Box 13-3 shows the clinical definition of the streptococcal toxic shock-like syndrome.35 The isolation of GAS from a usually sterile body site (e.g., blood- or cellulitis-infected) is required to confirm this diagnosis. Clinical pharyngitis often does not precede streptococcal toxic shock-like illness; throat culture is positive in only 50% of patients.7 More common infections that precede the illness include wound infection (often seemingly minor), cellulitis, and pneumonia. Further differentiating features of staphylococcal and streptococcal toxic shock are shown in Table 13-1; however, the clinical syndromes overlap sufficiently so that the bacterial cause cannot reliably be assigned at the time of presentation. Fortunately, treatment is essentially the same as that for classic staphylococcal TSS, including aggressive fluid management, antimicrobial treatment, and management of focal infection, including early exploration and debridement of infected soft tissue.31 A b-lactamase-resistant antibiotic, usually nafcillin or vancomycin, should be given empirically until the causative microbe and its antimicrobial susceptibility are specifically identified. In addition, clindamycin should be added for its potential superior efficacy in the highinoculum, low-replication setting of tissue-invasive GAS infection and its inhibitory effect on toxin production.36 A 1999 report suggests a cure rate of < 50% for severe streptococcal disease using a b-lactam antibiotic alone and a marked improvement with the addition of clindamycin.37 IGIV or corticosteroids, or both, may be useful in severe disease or in patients who do not respond to conventional management. Report of a small, matched case-control study showed a decreased incidence of mortality in patients who received IGIV (500 mg/kg per day for 5 days).38 Another study supports the use of IGIV for severe invasive streptococcal disease.39 Data are still considered limited. As in staphylococcal TSS, early diagnosis and treatment are critical.

CHAPTER

14

Hemophagocytic Lymphohistiocytosis and Macrophage Activation Syndrome Hayley A. Gans and David B. Lewis

HEMOPHAGOCYTIC LYMPHOHISTIOCYTOSIS Hemophagocytic lymphohistiocytosis (HLH) is an uncommon disorder that falls into the broader category of histiocytosis syndromes, sharing in common abnormal activation, proliferation, and accumulation of lymphocytes, mononuclear phagocytes, and dendritic cells. This potentially life-threatening syndrome is characterized by pro-

CHAPTER

14

113

longed fever, hepatosplenomagaly, cytopenias, and hemophagocytosis. The main defect described in HLH is deregulated activation of T lymphocytes and macrophages, resulting in sustained production of proinflammatory cytokines leading to systemic inflammation and rapidly progressive multiorgan failure. Hemophagocytosis, in which macrophages in the liver, spleen, bone marrow, and cerebrospinal fluid (CSF) phagocytose intact red blood cells, is a classic feature. Two forms of HLH have been deÀned, familial HLH (FHL) and acquired HLH (AHL). These types are often clinically indistinguishable from each other and from a number of other clinical syndromes, posing major diagnostic and therapeutic challenges. Infections are the most common precipitating events for both forms of HLH. Therefore, it is important to consider HLH as a potential complication of many infections, particularly because HLH carries a high morbidity and mortality rate if untreated. Macrophage activation syndrome (MAS) shares many clinical features with HLH, occurring in patients with an underlying rheumatologic disorder.

ETIOLOGY AND INCIDENCE HLH was Àrst recognized in 1952. Initial descriptions were of a lethal hereditary syndrome referred to as familial hemophagocytic reticulosis,1 and subsequently called recessive FHL,2 familial erythrophagocytic lymphohistiocytosis,3 and virus-associated hemophagocytic syndrome.4 It is now clear that two types of HLH exist. FHL, or primary HLH, has an autosomal-recessive pattern of inheritance and accounts for approximately 25% of HLH cases. Some experts place FHL into a broader categorization including other immunodeÀciencies that carry a high incidence of HLH as part of the syndrome (Table 14-1).5 AHL or secondary HLH is a nonfamilial disease associated with infectious agents, transplantation, malignancy, or rheumatologic disorders and treatment of a variety of conditions including malignancy, transplantation, and HIV/AIDS (see Table 14-1).5 The incidence of FHL is approximately 0.12/100,000 children per year based on retrospective studies from Sweden and the United Kingdom, affecting about 1 in 50 000 live births.2,6 These analyses likely underestimate the true incidence of FHL since many individuals die before a diagnosis is established. FHL cases are reported worldwide and include diverse ethnic groups.7–9 As is true of other autosomal disorders, consanguinity is an important risk factor.7,8,10 In Asia, where consanguinity is less common, FHL is a less frequent form of HLH.11 Approximately 70% of cases manifest during the Àrst year of life,5 and 85% manifest before 2 years of age.7 Presentation later in childhood and adult-onset disease has been reported.7,12 If untreated, FHL is almost always fatal. The treatment of choice is hematopoietic stem cell transplantation.13 There are limited data regarding the incidence of AHL, but it is likely substantially more common than FHL. Any inciting event that leads to a highly stimulated but ineffective immune response can trigger AHL. Historically, Epstein–Barr virus (EBV) infection, especially in an immunocompromised host, was a commonly reported precipitant of AHL.4,14–18 AHL can also occur in previously immunocompetent individuals in response to severe infections caused by a wide range of pathogens (see Table 14-1). Infectious triggers include herpesviruses19–21; other viruses such as influenza and parvovirus B1916,22; bacteria such as Mycobacterium tuberculosis23–25; fungi such as Histoplasma capsulatum,26 and during parasitic diseases such as visceral leishmaniasis.27 The common feature is infection causing sustained elevation of proinflammatory cytokines. Noninfectious etiologies include oncologic disorders28–31 and autoimmune diseases.32–34 There are also reports of AHL associated with primary35,36 and acquired immunodeÀciencies,37 with an infectious trigger in these immunocompromised hosts likely to be important in pathogenesis. Onset of disease temporally associated with initiation of chemotherapy,38 transplantation39–41 and highly active antiretroviral therapy for human immunodeÀciency virus (HIV)42 is also reported (see Table 14-1). In contrast to FHL, AHL tends to occur in older children and adolescents.7,11 AHL can be a self-limited process with recovery after

114

SECTION

A Fever, Bacteremia, & Septicemia

TABLE 14-1. Classification of Hemophagocytic Lymphohistiocytosis Genetic HLH

Familial HLH

Chromosome

Gene or Product

FHL1

9q21.3-22

Unknown

FHL2

10q21.22

Encoding Perforin (PRF1)

FHL3

17q25

Encoding Munc 13-4

FHL4

6q24

Encoding Syntaxin 11

CHS-1

1q42.1-q42.2

LYST

GS-2

15q21

RAB27A

XLP

Xq25

SH2D1A

IMMUNODEFICIENCY DISEASE

Acquired HLH

Associated Exogenous Event

Agent

Infections Viral

EBV, CMV, HSV, HHV-6, HHV-8, VZV, parvovirus B19, adenovirus, echovirus, Q-fever, influenza measles, HIV

Bacterial

Mycobacterium tuberculosis, Brucella

Parasitic

Plasmodium spp., Leishmania spp.

Fungal

Histoplasma capsulatum

Treatment

Chemotherapy, BMT, renal transplant, liver transplant, HAART

Associated Underlying Disease

Disease

Oncologic

ALL, B-cell lymphoma, T-cell lymphoma, Hodgkin disease, MM

Rheumatologic

JIA, SLE

ALL, acute lymphoblastic leukemia; BMT, bone marrow transplantation; CHS-1, Chédiak–Higashi syndrome; CMV, cytomegalovirus; EBV, Epstein–Barr virus; FHL, familial hemophagocytic lymphohistiocytosis; GS-2, Griscelli syndrome; HAART, highly active antiretroviral therapy; HIV, human immunodeÀciency virus; HHV-6, human herpesvirus 6; HHV-8, human herpesvirus 8; HLH, hemophagocytic lymphohistiocytosis; HSV, herpes simplex virus; JIA, juvenile idiopathic arthritis; MM, multiple myeloma; SLE, systemic lupus erythematosus; XLP, X-linked lymphoproliferative syndrome; VZV, varicella-zoster virus. Adapted from Janka G, Zur Stadt U. Familial and acquired hemophagocytic lymphohistiocytosis. Hematol (Am Soc Hematol Educ Program) 2005;82–88 and Janka G, Imashuku S, Elinder G, et al. Infection- and malignancy-associated hemophagocytic syndromes. Secondary hemophagocytic lymphohistiocytosis. Hematol Oncol Clin North Am 1998;12:435–444.

only supportive measures, but long-term remission is uncommon in individuals > 30 years of age and in those with central nervous system (CNS) disease.43 There is a growing awareness of an overall high mortality in patients with AHL.15,44

PATHOGENESIS The signs and symptoms of HLH are due to elevated and prolonged levels of circulating cytokines and chemokines produced from uncontrolled activation of T lymphocytes and mononuclear cells, particularly macrophages.45,46 High levels of circulating tumor necrosis factor a (TNF-a) interleukin (IL)-1b, IL-6, IL-10, IL-12, IL-18, interferongamma (IFN-g), and soluble IL-2 receptor alpha chain (CD25) have been found in the plasma of HLH patients.45–48 These proinflammatory cytokines contribute to the clinical Àndings of fever, hyperlipidemia,

endothelial activation, coagulopathy, inÀltration of lymphocytes and histiocytes into tissues, CNS vasculitis, marrow hyperplasia, or aplasia. Hemophagocytosis is a hallmark of overstimulated macrophages.46,49 Most studies of the pathogenesis of HLH favor a defect in the regulatory and effector lymphocyte pathways resulting in a dysregulated immune response. Genetic defects in perforin-mediated cytotoxicity have been identiÀed in HLH.50,51 Perforin is a poreforming protein secreted by activated cytotoxic T lymphocytes and natural killer (NK) cells that is critical for inducing the apoptosis of infected target cells.46,50,52–54 Defects in perforin-dependent cellmediated cytotoxicity may not only allow the persistence of intracellular pathogens and immune stimulation, but may also impair lymphocyte homeostasis, thereby preventing apoptosis of T lymphocytes after the immune stimulus has been eliminated. Patients with defective cell-mediated cytotoxicity but normal expression of perforin and other proteins involved in cytotoxicity can also manifest HLH.55 In contrast to defective lymphocyte-mediated cytotoxicity in patients with FHL, EBV-associated AHL may have a different sequence of pathogenic events. EBV-induced T-lymphocyte activation may result in clonal expansion of T lymphocytes and inducing cytokine production that potently activate mononuclear phagocytes to release proinflammatory cytokines and chemokines, which in turn act in a positive-feedback loop to activate T lymphocytes further.46 For unclear reasons, this proinflammatory milieu results in a transient NK-cell deÀciency perpetuating HLH.

GENETICS OF FAMILIAL CASES Four genetic loci, FHL1 through 4, have been identiÀed in association with FHL. The FHL1 locus is in the chromosome 9q21.3–q22 region and was Àrst described in kindred families in Pakistan56; the speciÀc gene defect remains unknown. It is unclear if patients with FHL1 locus mutation(s) have defects in T-lymphocyte or NK-cell-mediated cytotoxic activity or both, but it is speculated that the FHL1 gene plays a role in the negative regulation of the cell cycle. Defects on chromosome 9q22 are associated with malignancies, supporting the importance of the FHL1 locus in regulation of hematopoietic cell proliferation or differentiation.57,58 The FHL2 genetic locus in the chromosome 10q21 region59,60 encodes perforin (PRF1).50 Genetic defects in PRF1 account for approximately 20% to 50% of FHL cases,61 and can include microdeletions, missense or nonsense mutations.50,59,61–63 Genetic perforin deÀciency results in severe impairment of the in vitro cytotoxic activity of T and NK cells. The level of perforin expression is inversely related to the age of onset of HLH. Patients with null mutations come to medical attention in the Àrst 3 months of life,64 whereas those with missense mutations present between 1 and 27 months of age, most likely because of retaining low levels of perforin expression with residual function.65 There is evidence that exogenous triggers, in addition to genetic predisposition, are necessary for HLH to develop.65 The FHL3 locus in the chromosome 17q25.1 region66 is the UNC13D gene, which encodes the Munc13-4 protein. FHL3 patients with UNC13D mutations account for approximately 30% of FHL cases and have disease that is indistinguishable from those with genetic perforin defects.61,66 Although perforin expression is normal, defective Munc 13-4 protein impairs the secretion of cytolytic granules, resulting in impaired cell-mediated cytotoxicity.66 FHL4 was Àrst identiÀed in a Kurdish family with deletions in chromosome 6 (6q24).67,68 FHL4 appears to be due to defects in the gene encoding syntaxin 11 (STX11), an intracellular protein that is involved in the transport of cytotoxin-containing granules to the cell surface, a key event in cell-mediated cytotoxicity. In the remainder of patients with FHL, a particular genetic mutation has not been detected. Candidate genes being investigated for HLH include proteins involved in regulating cytotoxin (e.g., perforin and granzyme) expression or function or both, and the formation and secretion of intracellular vesicles containing cytotoxin molecules.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Hemophagocytic Lymphohistiocytosis and Macrophage Activation Syndrome

CLINICAL AND RADIOLOGIC FEATURES FHL and AHL cannot be differentiated reliably on the basis of clinical, laboratory, or histopathologic features. Clinical features are variable, but prolonged high fever, hepatosplenomegaly, and cytopenias are hallmarks of disease.2 Rash, lymphadenopathy, respiratory and gastrointestinal symptoms, failure to thrive, and irritability can also be present (Table 14-2).7,8,11,46,69,70 Clinical and laboratory abnormalities appear to peak at days 6 to 10 of illness, although there is substantial variability in the clinical course.70

Systemic Manifestations Fever can fluctuate but is present in all patients to some degree,11,70 and is commonly prolonged, typically lasting up to 4 to 6 weeks.8,11,70 Organomegaly, usually pronounced, is found in more than 90% of cases, and can be progressive.2,7,11,69 A diffuse erythematous maculopapular or petechial rash is often present transiently and only associated with high fevers.8,70 Respiratory symptoms range in severity from mild cough to acute respiratory distress syndrome requiring ventilatory support.11,70 Radiographic pulmonary abnormalities including inÀltrates6,70 are common, even in the absence of respiratory symptoms.70 Although respiratory failure is the most common form of organ failure in patients with HLH, cardiovascular collapse and renal failure can also occur.70,71 Lymph node enlargement is present in approximately 50% to 70% of patients, and is often prominent.8,70 Gastrointestinal symptoms, including vomiting, diarrhea, and abdominal pain, occur in about 40% of patients.11

Central Nervous System Manifestations Neurologic symptoms can dominate the clinical course. Although in some children neurologic symptoms and signs appear early in the course, Àndings may not manifest for several weeks, especially if deÀnitive treatment is delayed.6,72–74 Signs and symptoms can include irritability, seizures, cranial nerve palsies, ataxia, nystagmus, disturb-

CHAPTER

14

115

ances of gait and vision, hemiplegia or tetraplegia, delayed psychomotor development, and signs of increased intracranial pressure.72 Examination of CSF in patients with HLH usually reveals pleocytosis with predominance of lymphocytes and monocytes, and elevated protein level. Hemophagocytosis may be observed in the cell pellet of centrifuged CSF, a Ànding that is more speciÀc for HLH than hemophagocytosis observed in some non-CNS sites.46,74 In the few reports of neuroradiographic Àndings in HLH patients, lesions were present in the white matter of the cortex and cerebellum and often in the cortical structures.75 Necrotic areas with parenchymal volume loss are also described.74 Cerebral atrophy may be present and tends to become more pronounced with prolonged corticosteroid treatment. Magnetic resonance imaging is more sensitive than computed tomography in deÀning the characteristic inflammation and demyelination in patients with HLH.46 DeÀnitive diagnosis of CNS disease due to HLH may require a biopsy of the meninges and white matter.75 Neurologic abnormalities, especially those due to meningitis, may be reversible with appropriate and aggressive therapy initiated promptly.46 Neurologic symptoms are a common feature of disease relapse. The severity of the CNS disease does not necessarily correlate with the severity of other organ involvement, but commonly CNS disease is associated with marked systemic disease.74,75

LABORATORY AND PATHOLOGIC FINDINGS General Characteristic laboratory abnormalities in patients with HLH include pancytopenia, hypertriglyceridemia, hyperferrritinemia, hyperbilirubinemia, elevated transaminase levels, and hypoÀbrinogemia (Table 14-3).7,8,11,46,69,70 Thrombocytes are the most consistently depressed blood lineage, followed by both red and white blood cells. Two depressed cell lineages are evident at presentation in 80% to 90% of patients.11,70 Platelet number may be a useful indicator of disease activity, because they increase early during disease remission and decrease during relapse.8 Coagulopathy occurs in the majority of patients with HLH. In one study, 94% of affected children required blood product transfusions.70 Hepatic transaminase concentrations are

TABLE 14-2. Clinical Manifestations of Hemophagocytic Lymphohistiocytosis Symptoms

TABLE 14-3. Laboratory Findings in Hemophagocytic Present at Time of Diagnosis (%)

Lymphohistiocytosis

91–100

Laboratory Finding

Hepatomegaly

89–97

Anemia

89–94

Splenomegaly

61–98

82–100

Lymphadenopathy

17–52

Thrombocytopenia (< 100 000 platelets/mm3)

Skin rash

6–65

Neutropenia (< 1000 cells/mm3)

58–100

Respiratory distress

33–88

Leukopenia

39–87

Hypotension

85

Jaundice

72

Gastrointestinal Neurologic

Fever

Present at Time of Diagnosis (%)

Hypertriglyceridemia

80–100

HypoÀbrinogenemia

65–85

44

Elevated serum alanine aminotransferase level

33–92

20–47

Hyperbilirubinemia

33–74

Hyponatremia

79

Cerebrospinal fluid pleocytosis

52–91

Adapted from Chen CJ, Huang YC, Jaing TH, et al. Hemophagocytic syndrome: a review of 18 pediatric cases. J Microbiol Immunol Infect 2004;37:157–163; Arico M, Janka G, Fischer A, et al. Hemophagocytic lymphohistiocytosis. Report of 122 children from the International Registry. FHL Study Group of the Histiocyte Society. Leukemia 1996;10:197–203; Henter JI, Elinder G, Soder O, et al. Incidence in Sweden and clinical features of familial hemophagocytic lymphohistiocytosis. Acta Paediatr Scand 1991;80:428–435; Janka GE. Familial hemophagocytic lymphohistiocytosis. Eur J Pediatr 1983;140:221–230; Palazzi DL, McClain KL, Kaplan SL. Hemophagocytic syndrome in children: an important diagnostic consideration in fever of unknown origin. Clin Infect Dis 2003;36:306–312.

Adapted from Arico M, Janka G, Fischer A, et al. Hemophagocytic lymphohistiocytosis. Report of 122 children from the International Registry. FHL Study Group of the Histiocyte Society. Leukemia 1996;10:197–203; Henter JI, Elinder G, Soder O, et al. Incidence in Sweden and clinical features of familial hemophagocytic lymphohistiocytosis. Acta Paediatr Scand 1991;80:428–435; Janka GE. Familial hemophagocytic lymphohistiocytosis. Eur J Pediatr 1983;140:221–230.

116

SECTION

A Fever, Bacteremia, & Septicemia

usually at least threefold greater than normal,70 and the extent of hepatic abnormalities may signify stage of disease and risk of death.11 All patients with HLH have some elevation in serum ferritin and lactate dehydrogenase levels. Ferritin levels can be markedly elevated (i.e., > 10 000 mg/L); more than 90% of individuals have levels > 4000 mg/L. Ferritin concentrations greater than 500 mg/L are 80% speciÀc for the diagnosis of HLH.76 Hyponatremia and low protein and albumin concentrations are common Àndings in patients with HLH.8,11,70

Immunology HLH is typically associated with highly elevated levels of many proinflammatory cytokines and chemokines77; however, these elevations are not speciÀc for this disorder and testing is typically not readily available commercially. Cytolytic activity of NK cells and CD8 T lymphocytes is reduced or absent in patients with FHL regardless of the gene mutation, and in most cases of AHL, during the acute phase.52,78 Quantitative immunoglobulin levels show no consistent abnormalities in patients with HLH. Immunophenotyping of circulating cells shows decreased numbers of B lymphocytes but Tlymphocyte subset numbers are usually normal. T lymphocytes often show depressed responses to mitogenic stimulation.52

BOX 14-1. Diagnostic Criteria for Hemophagocytic Lymphohistiocytosis (HLH) The diagnosis of HLH can be established if either 1 or 2 is fulÀlled: 1. A molecular diagnosis consistent with HLH 2. Diagnostic criteria are fulÀlled (5 of 8) a) Original diagnostic criteria Clinical: Fever Splenomegaly Laboratory: Cytopenia (≥ 2 cell lineages in the peripheral blood) Hemoglobin < 9.0 g/dL (less than 4 weeks < 10 g/dl) Platelets < 100 000/mm3 Neutrophils < 1000/mm3 Hypertriglyceridemia and/or hypoÀbrinogenemia Fasting triglycerides ≥ 265 mg/dL Fibrinogen ≤ 1.5g/L Histopathologic: Hemophagocytosis in bone marrow, spleen, or lymph nodes without evidence of malignancy b) New diagnostic criteria Laboratory: Low or absent natural killer cell activity Ferritin ≥ 500 mg/L Soluble interleukin-2-receptor ≥ 2400 U/mL Adapted from Henter JI, Samuelsson-Horne A, Arico M, et al. Treatment of hemophagocytic lymphohistiocytosis with HLH-94 immunochemotherapy and bone marrow transplantation. Blood 2002;100:2367–2373 and HLH-2004 protocol: www.histio.org/society/protocols/trails-protocols.html.

Pathology The characteristic histopathology of HLH is diffuse inÀltration of the liver, spleen, lymph nodes, lungs, and brain with activated nonLangerhans histiocytes actively phagocytosing blood cells. However, this Ànding is often not present at disease onset.46 In most cases the expected histopathologic abnormalities develop as the disease progresses, and repeated biopsies from multiple tissue sources may be necessary for diagnosis.7 Hemophagocytosis is less often seen in the liver than other tissues such as lymph nodes and bone marrow. Autopsy specimens from patients with FHL can demonstrate focal CNS lesions and lymphocytes and histiocytes inÀltrating into the parenchyma and perivascular space, and hemophagocytosis in the meninges.73

Evaluation for Infectious Diseases Infection is a common inciting event in patients with HLH.79 Bloodstream infections caused by Pseudomonas aeruginosa, Candida tropicalis, and Staphylococcus aureus can occur, particularly in infants and those with prolonged neutropenia.11 Initial evaluations should include routine cultures of blood, urine, and, if warranted, CSF. A chest radiograph should be performed. Throat and rectal swabs for virus culture and serologic evaluation for EBV, cytomegalovirus, parvovirus B19, human herpesvirus-6 (HHV-6) and HIV infection should be considered. Nucleic acid-based tests may be warranted because antibody-based diagnosis may not be reliable if the patient is immunocompromised or has received recent transfusions. Evaluation for fungal infections should include blood culture, antigen testing, and serologic assays. SpeciÀc epidemiologic factors, such as exposure to tuberculosis, travel, and animal exposure guide further testing.

DIAGNOSIS The Histiocyte Society has developed guidelines80 for the diagnosis of HLH (Box 14-1). According to these guidelines, the diagnosis of HLH is based upon the presence of Àve criteria, including the clinical features of fever and splenomegaly, the laboratory criteria of cytopenias affecting two or more lineages and hypertriglyceridemia or hypoÀbrinogenemia, and histopathologic evidence of hemophagocytosis in bone marrow, spleen, or lymph nodes. Since hemophagocytosis may not be evident in bone marrow biopsies obtained early in

the disease, other tissue should be evaluated and serial bone marrow aspirates should be considered.5,7 If all other criteria are fulÀlled many experts would not delay treatment based on the absence of hemophagocytosis.46 The diagnosis of FHL is made in a patient meeting the criteria for HLH who has a positive family history or parental consanguinity or both.80 However, family history is negative in the majority of cases of FHL, due to the recessive inheritance pattern. In patients in whom a molecular diagnosis is established, other diagnostic criteria do not need to be fulÀlled to diagnose FHL.76 Three additional criteria have been added to the original diagnostic guidelines of HLH (see Box 14-1). These criteria include low or absent NK-cell cytolytic activity, serum ferritin concentration of > 500 mg/L, and soluble IL-2 receptor (IL-2R) alpha chain (sIL-2R) > 2400 U/mL. Either of the Àrst two additional criteria or a combination of the second and third may substitute for one major original criteria.76 Levels of soluble IL-2R may have prognostic implications. In one study of 74 patients, level of IL-2R > 10 000 U/mL before treatment was associated with signiÀcantly lower survival rate compared with patients with lower levels, despite more aggressive chemotherapy administered to those with higher concentrations.81 Immunologic testing can aid in the diagnosis of HLH and differentiate FHL from AHL. NK-cell function should be evaluated, and in most cases of FHL is undetectable or markedly decreased, although the number of circulating NK cells (CD56+/16+) are usually within normal limits.50 In FHL, the NK-cell function is permanently decreased. In contrast, the NK-cell activity returns to normal once the circulating NK-cell numbers normalize in patients with AHL, particularly when associated with EBV infection.46,78 A rapid-flow cytometric screening assay has been developed to evaluate the expression of perforin by NK and T lymphocytes.62 Agematched controls have been established, and it appears that > 85% of NK cells at all ages are normally perforin-positive, potentially allowing the rapid diagnosis of cases of perforin deÀciency (FHL2).76 Commercial genetic testing is available for mutations of the genes encoding perforin and Munc 13-4, which are the causes of FHL2 and FHL3, respectively.66,82

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Hemophagocytic Lymphohistiocytosis and Macrophage Activation Syndrome

Therapy of HLH, prior to definitive diagnosis, may be necessary because many patients have atypical manifestations and delay in therapy may have adverse consequences.83

DIFFERENTIAL DIAGNOSES The differential diagnosis of HLH is broad because of the variable and multisystem abnormalities of the disorder and the nonspecific nature of the most prominent features of fever, hepatosplenomegaly, and cytopenias. For example, the hematologic abnormalities can suggest leukemia, lymphoma, aplastic anemia, or myelodysplastic syndromes.15 Other histiocyte disorders, such as Langerhans cell histiocytosis, should be considered, especially in cases in which neurologic manifestations are prominent.75,80 Features of HLH can be found in association with a number of other immunodeficiency syndromes, including X-linked lymphoproliferative syndrome,84 severe combined immunodeficiency,85 Chédiak–Higashi syndrome,86 and Griscelli syndrome,87 as well as with rheumatologic disorders.88

CHAPTER

14

117

Cytotoxic chemotherapy is not indicated for MAS, HIV-associated HLH, and iatrogenically immunosuppressed patients. Studies evaluating treatment regimens for EBV-associated HLH have supported the early use of combination of etoposide and CSA for improved survival.96,97 In addition to its cytotoxic effects, etoposide may inhibit the synthesis of EBV nuclear antigen,98 thereby preventing the clonal expansion of EBV-infected T lymphocytes. The use of CSA is also associated with correction of neutropenia, thus reducing the risk of infection in patients with HLH. Hematopoietic stem cell transplantation has been needed for a minority of cases, which are recurrent.43 Treatment for HLH requires aggressive, supportive care with transfusions, antibiotics, and nutrition.83,91 Close monitoring for secondary bacterial and fungal infections is important, as both contribute to morbidity and mortality. If a patient has an identifiable infection, the infection should be treated. Acyclovir does not appear to be useful in patients who have documented EBV-induced HLH.79 HLH has reportedly resolved in occasional immunocompromised hosts, treated for documented adenovirus99 and HHV-8100 infections.

PROGNOSIS TREATMENT Chemotherapy and Immunosuppression The type of treatment for children with HLH is determined by disease severity, age of onset, and presence of familial disease.89 The natural course of HLH is typically progressive with multiorgan failure, cerebral or meningeal symptoms, and a mean survival of only 2 months if familial in nature.8 AHL also has a high mortality,15,44,80 especially if EBV-induced.5,14,90 There are no controlled treatment trials for HLH. However, the HLH-2004 protocol developed by the Histiocyte Society (available at www.histio.org/society/protocols/trails-protocols.html) serves as a treatment guide for all cases of FHL and severe, persistent, or recurrent ALH. These protocols combine chemotherapy and immunotherapy.91 The main goal of treatment for HLH is to suppress inflammation aggressively and to induce remission. Current initial treatment of patients with FHL and severe AHL includes administration of etoposide, glucocorticoid, and cyclosporine A (CSA) for 8 weeks. For all cases of FHL and for AHL that does not respond to initial therapy, etoposide, glucocorticoid pulse therapy, and CSA are continued and intrathecal methotrexate is considered. These agents help control HLH, by inducing apoptosis of lymphocytes and inhibiting cytokine and chemokine production. Hematopoietic stem cell transplantation is indicated for children with persistent or reactivated disease.83,91 Patients with FHL ultimately require allogeneic hematopoietic stem cell transplantation for cure. In the registry, 60% of children with HLH who received a matched sibling bone marrow transplant survived 5 years, compared with 10% of those receiving chemotherapy alone.7 Promising results have also been obtained in some cases with matched unrelated donors and haploid-identical family-member donors.92–94 Variable responses to treatment of HLH using intravenous immunoglobulin, glucocorticoids, and antithymocyte globulin as single or combined agents have been observed.11,95 These treatment regimens are only considered as initial therapy for mild cases of ALH.

HLH carries a high mortality rate even with early therapy. Increased survival is associated with prompt response to initial therapy and with older age at presentation; 2-year survival was 100% in these populations in a study of 113 HLH cases from 21 countries.83 In this study, 62% of patients requiring hematopoietic stem cell transplantation survived 3 years and all were disease free.46 Of the 32% of patients who presented with neurologic symptoms, 67% responded to initial therapy.

MACROPAHGE ACTIVATION SYNDROME First described as a complication of juvenile arthritis in 1985,101 MAS refers to a clinical condition caused by the excessive activation and proliferation of well-differentiated macrophages and T lymphocytes in individuals with an underlying rheumatologic disorder.102 Sharing many of the clinical and laboratory features of HLH, MAS is considered by some as an acquired form of HLH, although no formal criteria have been established for MAS.88,103 MAS is most commonly a complication associated with juvenile idiopathic arthritis (JIA), but can also occur in patients with systemic lupus erythematosus (SLE) and other rheumatologic disorders; MAS can also be the initial manifestation of an autoimmune disease.104–106 The pathophysiology shares common features with HLH, and patients with systemic-onset JIA have been shown to have defective NK-cell function, low perforin expression, and elevated levels of hemoglobin scavenger receptor.106,107 Clinical and laboratory findings are characterized by persistent fever, hepatosplenomegaly, neurologic abnormalities, marked cytopenias, and hypofibrinoginemia. There may be a decrease in the erythrocyte sedimentation rate compared with baseline, signaling a paradoxical improvement in the underlying disease state.102 Although successful treatment of MAS has been reported with CSA and glucocorticoids,108 if patients do not respond to initial therapy within a few days, implementation of the HLH-2004 protocol should be considered.

SECTION

B

Cardinal Symptom Complexes

CHAPTER

15

Mucocutaneous Symptom Complexes Sarah S. Long Generally, in differentiating between bacterial and viral causes of febrile illnesses in children, the more mucous membranes involved in the patient’s illness (e.g., conjunctiva, throat, respiratory, gastrointestinal tract), the more likely the cause is viral. When multiple mucous membranes are involved and an exanthem (i.e., mucocutaneous complex) is present, a self-limited viral cause is likely, but other important diagnoses must be considered. They commonly include inflammatory or immunologically mediated conditions, such as Kawasaki disease, Stevens–Johnson syndrome, and drug hypersensitivity, and bacterial toxin-mediated diseases, including staphylococcal and streptococcal toxic shock, streptococcal scarlatiniform disorders, and staphylococcal exfoliative toxin syndromes (toxic epidermal necrolysis and staphylococcal scalded-skin syndrome). A “best,” if not the definitive, diagnosis can be deduced through careful assessment of: (1) the dominant features of the illness; (2) prodromal events and exposures; (3) specific characteristics of the exanthem and abnormality at each affected mucous membrane; and (4) the cadence of the developing constellation. Laboratory features are of secondary importance, only adding weight to or removing weight from the clinical assessment. Table 15-1 lists the useful differentiating features of commonly considered causes of mucocutaneous symptom complexes. Features distinguishing streptococcal toxic shock due to streptococcal pyrogenic exotoxin (SPE) A or B from staphylococcal toxic shock associated with toxic shock syndrome toxin-1 (TSST-1) are shown in Table 15-2, and are described in the discussion where pertinent. Less common conditions share certain clinical features of the mucocutaneous syndromes shown in Table 15-1 but can usually be distinguished on the basis of circumstances of occurrence. Periodic fever with aphthous stomatitis, pharyngitis, and adenitis (the PFAPA syndrome) is distinguished by the multiplicity and periodicity of febrile episodes as well as the absence of exanthem and conjunctival abnormalities1 (see Chapter 17, Prolonged, Recurrent and Periodic Fever Syndromes). Behçet syndrome displays chronicity and recurrences and has the following characteristic findings: large oral and anogenital ulcers, erythema nodosum, and uveitis.2 Although the ulcerative gingivitis and pseudomembrane of mucositis associated with neutropenia and anticancer chemotherapeutic agents3 can be indistinguishable from the confluent denuding ulcers of Stevens– Johnson syndrome;4–6 the clinical setting of the former as well as the absence of conjunctivitis and rash distinguishes mucositis from Stevens–Johnson syndrome. Paraneoplastic vasculitis complicating myeloproliferative disorders can cause urticaria, erythema multiforme, or palpable purpura. This diagnosis is suspected because of the associated malignancy and is confirmed by skin biopsy.7 Rocky Mountain spotted fever, an infective vasculitis caused by Rickettsia rickettsii, shares many features of other mucocutaneous syndromes (i.e., high fever, progressive serious illness, conjunctival hyperemia and suffusion, peripheral edema despite hypovolemia, and 118

hyponatremia), but the presence of unremitting headache, peripheral petechial rash (if present), season, and exposure to ticks usually set it apart.8 Ehrlichiosis and anaplasmosis occur in a similar setting; leukopenia may be a clue.9 Treatment for Rocky Mountain spotted fever or ehrlichiosis is sometimes required because the diagnosis cannot be reasonably excluded. Leptospirosis is an infectious as well as immunemediated disease, indistinguishable from Rocky Mountain spotted fever except possibly for a biphasic illness and disproportionate organ involvement of the kidney or liver in leptospirosis.10 Drug hypersensitivity reactions can cause a variety of mucocutaneous abnormalities.11 For the previously healthy child or adolescent who comes to medical attention in shock, septic shock from unrecognized ruptured appendix, urosepsis, invasive meningococcal or pneumococcal infection (especially if petechiae, purpura, or purpura fulminans is present), and disseminated staphylococcal and group A streptococcal infection must also be considered. Streptococcus pneumoniae and group B Streptococcus can also cause toxic shock-like manifestations, which are probably related to cytokine stimulation.12

SPECIFIC DISTINGUISHING CHARACTERISTICS Fever and Prodrome The sequence of events and duration of fever when the child comes to medical attention provide useful clues to diagnosis. Although children with Kawasaki disease and bacterial toxin-mediated syndromes ultimately share many mucocutaneous features, the cadence of the prodrome is distinct. The child with staphylococcal toxic shock usually has profuse diarrhea and severe prostration within hours of onset of fever, whereas the child with Kawasaki disease begins with fever and crankiness, sometimes with unilateral cervical lymphadenitis initially, not unlike the beginning of a common viral illness.13,14 Concern about the clinical state of the child with Kawasaki disease does not heighten for several days, when fever and crankiness persist and the symptom complex evolves rather than abates, as might be expected in self-limited viral infection.15,16 The younger the child with a staphylococcal exfoliative toxin syndrome, the more rapidly progressive and dramatic the skin manifestations, with constitutional illness usually secondary.17,18 Initial manifestations of Stevens–Johnson syndrome can be urticaria, with subsequent progression to the fixed tissue lesions of erythema multiforme, and evolution of mucous membrane inflammation and systemic illness.4,19 Enteroviruses and respiratory viruses that cause mucocutaneous findings evolve over 3 to 5 days; nasal symptomatology, rhinorrhea, hoarseness, or cough is present in > 75% of cases, distinguishing these viral infections from Kawasaki disease. High fever (39.2ºC or more) or persistent fever (5 days or more) does not eliminate viral etiology, because viruses commonly considered in such patients (adenovirus, influenza A and B viruses) typically cause high fever, and at least a third of affected children have fever beyond 5 days.20–22

Conjunctiva To ascribe conjunctival findings accurately, the examiner must pay particular attention to: (1) presence of inflammation and exudate versus erythema alone; (2) relative involvement of bulbar versus

Mucocutaneous Symptom Complexes

CHAPTER

15

119

TABLE 15-1. Differentiating Among Causes of Mucocutaneous Symptom Complexes Kawasaki Disease

Toxic Shocka

Staphylococcal Exfoliative Toxin Syndromes

Streptococcal Scarlatina

Stevens–Johnson Syndrome

Viral Infection

CLINICAL FEATURES

Fever ≥5 days Conjunctiva

+++ 2 days +++ Bilateral hyperemia (bulbar > palpebral); anterior uveitis

+++ 3 days +++ Bilateral hyperemia (bulbar > palpebral)

Lips

+++ Erythema, fissures

++ Erythema

Oropharynx

+++ Mucosal erythema; strawberry tongue

++ Mucosal erythema; strawberry tongue

Exanthem

+++ Polymorphous, vasoactive and changing; morbilliform, symmetric; exaggerated or solely in groin ++ Symmetric, indurative edema distally; painful erythema, palms and soles; stocking/ glove distribution; occasional digital cyanosis +++ Unremitting crankiness; cervical lymphadenopathy; arthralgia/arthritis; meningismus, cranial nerve palsy; abdominal pain, distension, tenderness; hydropic gallbladder Persistent fever and unremitting crankiness

++ Erythroderma

Extremities

Other

Predominant feature(s)

Convalescent clinical features

Desquamation periungually to palms and soles (full-thickness); minimal desquamation elsewhere; hair loss, nail abnormalities; coronary artery aneurysms/thrombosis

+ + 2–3 days 2–3 days ++ – Unilateral or bilateral Normal purulent conjunctivitis (palpebral > bulbar); or normal

++ 5 days ++ Bilateral purulent conjunctivitis, chemosis; keratitis, panophthalmitis

– Normal, or edema contiguous with exanthem

– Normal

– Normal, or edema contiguous with exanthem

– Normal, or palmar vesicles (enteroviruses, herpesviruses), other exanthem

+++ + Profuse prodromal Infected tissue site diarrhea; dizziness, headache, confusion; hypotension, shock; hydropic gallbladder

+ Sore throat, odynophagia; cervical lymphadenitis; malaise, vomiting

++ Malaise, arthralgia; urethral/anal ulceration and symptoms; abdominal pain, diarrhea

++ Headache, malaise, myalgia, cough, rhinorrhea; pneumonitis; lymphadenopathy, splenomegaly; each dependent on specific virus

Fever and prostration Desquamating skin lesions

Exanthem and sore throat

Edematous bloody lips and oral pseudomembrane

Variable; respiratory tract and/or constitutional symptomatology

Desquamation hands and feet (fullthickness); mild desquamation elsewhere; hair loss, nail abnormalities

Desquamation hands and feet (fullthickness); desquamation extensive elsewhere

Desquamation at sites of exanthem, lips, perineum; recurrences; serious ophthalmologic sequelae

Desquamation at sites of exanthem (mild)

++ Symmetric distally; erythema on palms and soles

++ 3–5 days +/+++ Unilateral or bilateral purulent conjunctivitis (palpebral > bulbar); cobblestone lymphoid hyperplasia – – +++ – Normal or Normal, with Erythema and edema, Normal desquamation circumoral pallor fissures, denudation; bleeding, black eschar – ++ ++ +/+++ Normal, or mucosal Tonsillar erythema, Panmucosal Erythema; anterior or erythema exudate; palatal erythema, confluent posterior discrete petechiae; strawberry ulceration, ulceration; tonsillar tongue denudation; exudate or follicular pseudomembrane hyperplasia; palatal petechiae; each dependent on specific virus +++ +++ +++ +/++ Indurative or papular Papular erythroderma Erythema multiforme; Maculopapular erythroderma; tender; (sandpaper); polymorphous, fixed; (discrete or confluent), Nikolsky sign and Pastia sign bullae vesicular, or petechial; desquamation during each dependent on acute phase specific virus

Desquamation extensive during acute and convalescent periods

120

SECTION

B Cardinal Symptom Complexes

TABLE 15-1. Differentiating Among Causes of Mucocutaneous Symptom Complexes—Continued Kawasaki Disease LABORATORY FEATURES

DIAGNOSIS

TREATMENT

Toxic Shocka

Elevated peripheral neutrophils, platelets, sedimentation rate, ± IgE; pyruria ± electro- or echocardiographic abnormalities

Clinical, with exclusion of others and supportive laboratory Àndings; response to IGIV; ectasia, aneurysm, or thrombosis of coronary arteries IGIV 2 g/kg

Thrombocytopenia, left shift of neutrophils; coagulopathy; hyponatremia; multiorgan dysfunction related to hypoperfusion; adult respiratory distress syndrome TSST-1 producing Staphylococcus aureus recovered (from mucosa/ infected site); response to aggressive fluid support, antibiotic Crystalloid, colloid, then pressor agents; penicillinaseresistant antibiotic (or vancomycin) plus clindamycin

Staphylococcal Exfoliative Toxin Syndromes

Streptococcal Scarlatina

Stevens–Johnson Syndrome

Left shift of neutrophil counts; hypovolemia in young infant; bacteremia/ septicemia related to infected site

Elevated peripheral neutrophils

Elevated peripheral neutrophils, sedimentation rate

Exfoliatin producing S. aureus recovered (usually from skin); response to antibiotic

Erythrogenic toxin- Clinical; slow producing improvement; Streptococcus recurrences pyogenes recovered (usually from throat); response to antibiotic

Pencillinase-resistant Penicillin antibiotic (consider vancomycin or clindamycin); supportive fluids

Viral Infection Normal or low peripheral neutrophils, or elevated lymphocytes; normal or low platelets

Clinical; rapid improvement

Supportive; Supportive; occasional discontinue inciting speciÀc antiviral agent drug; corticosteroid, IGIV not proven (see text)

IgE, immunoglobulin E; IGIV, immune globulin intravenous; TSST, toxic shock syndrome toxin; +++, prominent, expected Ànding; ++, frequent Ànding; +, variable Ànding; –, not present. a Features of staphylococcal toxic shock (i.e., related to TSST-1); see text and Table 15–2 for differences in streptococcal toxic shock.

TABLE 15-2. Comparative Features of Toxic Shock Syndromea Feature

Staphylococcal Toxic Shock

Streptococcal Toxic Shock

Primary toxin Prodrome Duration prodrome Severity of prodrome Focal infection Extreme pain/ hyperesthesia at focal site Rash Shock

TSST-1 Vomiting, diarrhea Hours +++ +

SPEA, B Flu-like illness Hours–days + ++

– Erythroderma Predictable

+++ Scarlatina/none Unpredictable, related to clotting Sometimes untreatable Unpredictable

Treatable Predictable, related to blood pressure Treatable Positive blood culture – Coagulopathy + Complicated + hospitalization Gangrene ± Mortality + Multiorgan failure

Untreatable ++ +++ +++ +++ +++

SPE, streptococcal pyrogenic exotoxin; TSST, toxic shock syndrome toxin; ±, not an expected occurrence; +, occurs, but infrequently; ++, occurs with some frequency; +++, occurs commonly; –, does not occur. a Differences represent general comparisons rather than clinical Àndings or outcomes in individual patients.

palpebral and tarsal sites; (3) presence or absence of photophobia, pain on movement, and itching; and (4) presence of uveitis, destructive keratitis, and panophthalmitis versus superÀcial inflammation. Ocular manifestations of many infectious, inflammatory, allergic, and toxin-mediated conditions begin with bilateral conjunctival

erythema.23 The earliest ocular Àndings of Kawasaki disease, measles, Stevens–Johnson syndrome, Rocky Mountain spotted fever, and leptospirosis, for example, cannot be distinguished. However, in Kawasaki disease and toxic shock syndromes (as well as in Rocky Mountain spotted fever and leptospirosis), nonexudative erythema of bulbar conjunctivae is the complete evolution. Photophobia and pain on movement are absent.24 Conjunctival blood vessels are merely inflamed or dilated. In sharp contrast, adenoviral conjunctivitis is unilateral in 65% of cases, is predominantly palpebral (frequently with follicular lymphoid hyperplasia), and is associated with keratitis, purulent exudate, and photophobia in approximately half of cases and with concurrent pharyngitis in 55%.25 Bacterial conjunctivitis (especially that due to nontypable Haemophilus influenzae, S. pneumoniae, or Staphylococcus aureus) is also purulent, and usually bilateral; other mucocutaneous symptoms are not present (except with exfoliative toxin-producing S. aureus infection). Concurrent acute otitis media is common when nontypable H. influenzae is causative.26 The “red eye” of allergic disorders is easily distinguished when hallmarks of ocular itching, tearing, and photophobia are present; papillary hypertrophy of the palpebral conjunctivae is characteristic. Normal conjunctivae (as well as normal lips and circumoral skin) are somewhat distinctive in streptococcal scarlet fever. The Àndings of light sensitivity, ocular pain, and decreased vision suggest uveitis, which is characteristic of juvenile idiopathic arthritis (JIA) and Behçet syndrome. Such forms of JIA are not generally associated with fever and rash. Anterior uveitis is common in Kawasaki disease, being present in 83% of children examined in the Àrst week and in 66% examined after the Àrst week in one study.27 Unlike in other causes of uveitis, the inflammation in Kawasaki disease is mild and transient and is associated with minimal photophobic behavior or pain. Stevens–Johnson syndrome is frequently associated with severe anterior-segment involvement, which comprises exudative conjunctival discharge and sloughing of epithelium as well as pseudomembrane formation, and has the potential to cause subsequent severe corneal and eyelid scarring.28

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Mucocutaneous Symptom Complexes

Anterior uveal inflammation can occur with measles and leptospirosis but is not expected with other mucocutaneous syndromes (e.g., streptococcal or staphylococcal toxin-mediated diseases, enteroviral or adenoviral infection, rickettsial diseases). Ophthalmologic evaluation with slit-lamp examination is important in children with evolving conjunctival findings and a mucocutaneous syndrome to uncover intraocular disease and aid in the diagnosis and management of Stevens–Johnson syndrome.

Lips In infants, the lips (and frequently the ears) are brightly colored when fever is high, returning to normal with even transient defervescence. The lips are infrequently involved in mucocutaneous syndromes due to viruses, except during primary herpes simplex stomatitis, when diagnosis is apparent. Measles and influenza may be other exceptions, in which lips can be red, edematous, and cracked several days after persistent high fever.29 The lips are noticeably red and are sometimes cracked and fissured in Kawasaki disease and toxic shock, however, providing important clues to diagnosis. The swollen, denuded, bleeding lips with black eschar of Stevens–Johnson syndrome are pathognomonic (Figure 15-1), the mucocutaneous junctions of lips, conjunctivae, urethra, and anus being primary target sites of the pathologic process.

Oropharynx Examination is performed to detect: (1) diffuse erythema, sometimes with uvulitis (Kawasaki disease and toxic shock); (2) hypertrophied or dilated papillae on the tongue to give the appearance of “strawberry tongue” (Kawasaki disease, toxic shock, streptococcal scarlatina); (3) confluent buccal and gingival ulceration with pseudomembrane formation (Stevens–Johnson syndrome and mucositis of neutropenia and antineoplastic drug therapy); (4) palatal petechiae (Streptococcus pyogenes and Epstein–Barr virus); (5) discrete ulcers (enteroviruses, herpes simplex virus); (6) Koplik spots (rubeola); and (7) exudative enanthem (S. pyogenes, respiratory tract viruses, Epstein–Barr virus). The singular (or few) deep, large ulcerative lesions of aphthous

Figure 15-1. Ten-year-old girl with Stevens–Johnson syndrome and characteristic denudation of mucocutaneous junctions of the lips.

CHAPTER

15

121

stomatitis and Behçet disease are found characteristically on the buccal or lingual mucosa or the lateral tongue.

Exanthem Even though exanthematous lesions can seem nonspecific and variable, their exact characteristics, distribution, pattern of evolution, and timing and manner of resolution are extremely helpful in differentiating among causes of mucocutaneous syndromes.

Lesions Vesicular skin eruptions are almost unique to herpesviruses and enteroviruses.30 Bullae are the typical initial lesions of staphylococcal exfoliative toxin syndrome, as is sloughing of sheets of epidermis with gentle pressure or minor trauma in some patients – Nikolsky sign. Localized hemorrhagic bullae can form in the skin overlying necrotizing cellulitis, myositis, or fasciitis due to toxin-producing S. pyogenes.31 Bullous erythema multiforme, the hallmark of Stevens–Johnson syndrome, occurs first in localized areas, evolves at the site over days, and commonly progresses to other sites. Usually, the lesions are symmetric, on extensor surfaces of extremities as well as the trunk, with a predilection for sun-exposed areas. Typical target or iris lesions of erythema multiforme are expected in Stevens– Johnson syndrome, with urticarial lesions as well. Vesicular, bullous, or petechial lesions are distinctly rare in Kawasaki disease. The rash of Kawasaki disease is polymorphous (patchy macular, maculopapular, infrequently papular) and is indistinguishable in individual cases from the exanthem of viral infection, drug, or food allergy, Stevens–Johnson syndrome, or toxinmediated disease; however, characteristically, the rash in Kawasaki disease is fiery red, morbilliform (confluent), symmetric (involving both hands, feet, knees, and elbows), and changeable from hour to hour, rather than fixed as in viral infection or Stevens–Johnson syndrome. Although urticarial and erythema multiforme-like lesions can occur in Kawasaki disease, they are not fixed. Rashes in Kawasaki disease characteristically spare the head, whereas exanthems in many viral infections, Stevens–Johnson syndrome, and staphylococcal toxin diseases commonly begin on, or involve, the face. Exaggeration or confinement of exanthem at the groin is characteristic of Kawasaki disease,32 although this feature can occur in streptococcal scarlatiniform eruptions, staphylococcal exfoliative toxin syndrome, influenza,29 and bacterial cellulitis in the neonate (in which primary umbilical colonization is usual). Erythema, induration of the scrotum, and pain in testicles can occur in Kawasaki disease (and Rocky Mountain spotted fever); sometimes, a hydrocele appears acutely.33 Viral exanthems are not painful or tender. The skin of children with diffuse staphylococcal exfoliative toxin syndrome, however, is erythematous and edematous, sometimes has a sandpaper quality, and is usually painful and tender.34 In Kawasaki disease, specific sites of exanthem are not painful except on the hands and feet, where edema (and probably vasculitis) causes pain, frequently leading to refusal to bear weight. An important clue to soft-tissue infection due to streptococci (usually b-hemolytic group A organisms, but also group B, C, and G organisms on occasion, and S. pneumoniae, the last especially in individuals with connective tissue diseases)35 is that the patient complains of exquisite pain, sometimes with hyperesthesia or hypoesthesia, at the site of infection or deeper, which is out of proportion to objective abnormality (or tenderness to palpation, initially).31,36 Epidemiologic features, such as season, exposure, incubation period, and associated findings, are frequently more helpful in distinguishing viral infections than is the exanthem itself. Enteroviruses are the leading cause of exanthematous diseases, with more than 30 types associated with rash illnesses, some of which also cause mucosal lesions.30,37,38 Although neonates and young infants can have diffuse macular or blotchy rashes, exanthem at peak ages for enteroviral exanthematous diseases (especially due to echoviruses 4, 9, and 16 as well as coxsackieviruses A9 and B5) usually consists of rubelliform

122

SECTION

B Cardinal Symptom Complexes

Figure 15-2. Four-year-old boy with staphylococcal exfoliative toxin disease. Note characteristic paper-thin desquamation of skin on eyelid and impetiginous infection around the nose and chin.

(maculopapular discrete, nonconfluent) lesions beginning on the face and upper trunk and spreading down to the extremities.38 An important exception is hand–foot–mouth disease (usually due to coxsackievirus A16, followed by A5, A9, A10, B1, and B3, and enterovirus 71), in which peripheral distribution of vesicular or rubelliform lesions is characteristic. Rashes during viral infections due to influenza A and B, parainfluenza, and respiratory syncytial virus (especially in young children) probably occur more than occasionally, are always rubelliform, are frequently present for less than 24 hours, and spread from the head to the trunk and extremities.30 Rash occurs in 2% to 8% of adenoviral infections, and adenovirus types 1, 2, 3, 4, 7, and 7a have been isolated most frequently;38 lesions are distributed as with influenza and are usually rubelliform, occasionally morbilliform (maculopapular confluent patches), and rarely erythema multiforme-like. Mucocutaneous and systemic manifestations of measles can simulate Kawasaki disease initially, the rash being least differentiating; prominent cough, coryza, and eventual purulent conjunctivitis distinguish measles. Similarly, the prominent respiratory tract symptoms caused by Mycoplasma pneumoniae distinguish this infection; maculopapular rashes (5% to 15% of cases), consisting of vesicular or bullous, papular, petechial, or urticarial lesions, or erythema multiforme-like, have been described.39 Classically, staphylococcal toxic shock is associated with diffuse erythroderma, a flushed, sunburned appearance (without discrete lesions, tenderness, or induration) especially on the face, trunk, and proximal extremities. Streptococcal scarlet fever of scarlatina is associated with a diffuse, fine maculopapular rash, which is palpable, sandpaper-like, and prominent on the trunk and proximal extremities; exaggeration of erythema in skinfolds is characteristic (Pastia sign). Frequently, toxin-mediated exanthems overlap (presumably because more than one toxin is encoded), and manifestations of staphylococcal exfoliative toxin syndrome can occur in addition to those of staphylococcal toxic shock, or those of scarlet fever can occur in addition to those of streptococcal toxic shock. Children with staphylococcal exfoliative toxin syndrome commonly have crusted, exudative, infective conjunctivitis, or paranasal lesions, in addition to generalized nonexudative exfoliation (Figure 15-2).

Extremity Changes Indurative erythroderma with frank edema of the extremities is probably the most helpful differentiating physical finding in some mucocutaneous syndromes, reflecting inflammatory vasculitis

Figure 15-3. Six-week-old boy with the infantile polyarteritis nodosa form of Kawasaki disease. Note the dusky fingers with distal gangrene.

(Kawasaki disease), vascular dilatation (toxic shock syndromes), or infective vasculitis (Rocky Mountain spotted fever). It does not occur in viral infections or Stevens–Johnson syndrome, except contiguous to sites of erythema multiforme in the latter. The demarcated, stocking or glove distribution of erythema and induration in Kawasaki disease is dramatic on occasion, as are digital cyanosis and gangrene (Figure 15-3). The latter are easily distinguished from the peripheral ischemia of toxic or septic shock by low blood pressure, weak pulse, and pale, cool extremities in shock.

Evolution and Resolution The exanthem of Kawasaki disease evolves, waning, waxing, and migrating over days; that of toxic shock and exfoliative syndromes is rapid in onset and is progressive; viral exanthems characteristically appear days after onset of systemic illness and evolve at one site while progressing to others. Desquamation occurs during the crescendo phase of staphylococcal exfoliative toxin syndrome (as maximal disease manifestations evolve) but during the decrescendo phase of Kawasaki disease (typically days 10 to 20 after onset, beginning in the periungual region), and only during the convalescent phases of staphylococcal and streptococcal toxic shock and scarlet fever. Desquamation of palms and soles is likely to be a full-thickness loss (as if molted) after Kawasaki disease, streptococcal scarlet fever, and toxic shock. In Kawasaki disease, unlike bacterial toxin-mediated diseases, total body desquamation does not occur or is fine and superficial, occurring in only about 10% of patients, particularly in the groin and perineal area. Hair loss and nail bed deformities (i.e., grooves (Beau lines), pits, transverse (red) lines) are well-described occurrences in the weeks to months after Kawasaki disease40 but can occur after other highly febrile, hypotensive, ischemic illnesses, such as in invasive bacterial infection, toxic shock, and Rocky Mountain spotted fever.

Other Clinical Features and Cardinal Feature Clinical features other than mucocutaneous signs or symptoms, alone or in combination, can provide valuable clues to diagnosis. A partial listing is given in Table 15-1. Inflammation at multiple sites is typical of Kawasaki disease, anterior uveitis being quite specific among diagnoses considered. Transient hydrops of the gallbladder, manifesting as right upper quadrant abdominal pain, tenderness, and

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Mucocutaneous Symptom Complexes

fullness, is a common but nonspecific finding in Kawasaki disease that also occurs in staphylococcal toxic shock,13 streptococcal scarlet fever,41 and other systemic infectious and inflammatory diseases, such as enteric fever,42 leptospirosis, Epstein–Barr virus mononucleosis,43 and neonatal group B streptococcal septicemia, and after asphyxia and parenteral alimentation.44 Infants younger than 6 months with Kawasaki disease are especially prone to a rapidly progressive course and severe vascular complications.15,45,46 They frequently lack classic diagnostic findings, such as exanthem and erythematous conjunctivae. Extremely high fever, anxious appearance, unconsolable irritability, respiratory distress, and cardiac gallop murmur with hepatomegaly can be clues. Aneurysmal dilation of multiple vessels can occur; pulsatile masses in the axillae or groin can sometimes be appreciated acutely, within days of onset of fever (personal experiences). Delineating the predominant complaint, or cardinal feature, of the child’s illness aids in establishment of a correct diagnosis (see Table 15-1). Unremitting crankiness is almost universal in Kawasaki disease, probably related to the multiple sites of vasculitis and painful edema or ischemia; unilateral, remarkable cervical lymphadenitis is sometimes a dominant feature.47 Prodromal fever, diarrhea, and then profound prostration and hypotension are predominant in staphylococcal toxic shock but can also occur in streptococcal toxic shock. The prodrome is frequently nonspecific in streptococcal toxic shock; the clinical picture is dominated by extreme painfulness of apparent minor soft-tissue infection (if present), followed by sudden inexorable shock and coagulopathy. Children with staphylococcal exfoliative toxin syndrome have variable degrees of illness, depending on age and presence of bloodstream or focal infection; the rapidly evolving exanthem with loss of skin in sheets (as well as purulence around eyelids and nose) dominates the clinical picture. Edematous, bleeding, black eschared lips and oral mucosal denudation are the dramatic events in Stevens–Johnson syndrome; in some cases, inflammatory necrosis of respiratory or intestinal tract mucosa causes wheezing, odynophagia, and other symptoms.5 In viral illnesses, the mucocutaneous features broaden the differential diagnosis, but other symptoms (such as cough, hoarseness, sore throat) commonly dominate the patient’s or parent’s complaints; general debilitation is usually modest, and the epidemiologic setting heightens the likelihood of a specific diagnosis. Table 15-2 summarizes the differentiating features of classic staphylococcal toxic shock (associated with TSST-1 and enterotoxin B) and streptococcal toxic shock (associated with SPE A or B and, possibly, proteases). Although these syndromes are superficially similar, the evolution of clinical symptoms, the complications, and the outcomes are frequently distinctive. Staphylococcal toxic shock usually has a dramatic onset and rapid progression. Focal infection is present in most nonmenstrual pediatric cases but can be relatively minor (e.g., sinusitis, surgical wound).48,49 Streptococcal toxic shock usually has a less abrupt onset of symptoms, with initial malaise, myalgia but complaints of severe pain, hyperesthesia, or hypoesthesia at an apparently minor soft-tissue site of infection.35,36,50 Nonexudative pharyngitis is the occasional primary site of streptococcal infection, with a complaint of sore throat out of proportion to the findings.51 Many children with streptococcal toxic shock have been brought for medical attention one or more times before the onset of catastrophic multiorgan system failure and shock.52 A possible potentiating role of nonsteroidal anti-inflammatory drugs in invasive streptococcal disease has been questioned53 and not excluded by case-control studies.54,55

DIAGNOSIS AND EMPIRIC THERAPY The best working diagnosis, barring the presence of pathognomonic features, is made on clinical grounds. Laboratory findings infrequently confirm a diagnosis at the time therapy must be given. Brisk elevations of acute-phase reactants are supportive laboratory findings in Kawasaki disease (and are variably present in Stevens–Johnson syndrome); white blood cell count < 15 000/mm3, erythrocyte sedimentation rate < 40 mm/

CHAPTER

15

123

hour, C-reactive protein < 3 mg/dL, or platelet count < 200 000/mm3 are negative indicators for Kawasaki disease.15 Leukemoid reaction with neutrophil count exceeding 30 000/mm3 frequently occurs in very young infants with the infantile polyarteritis presentation of Kawasaki disease; anemia is universal.45 These can be important clues to the diagnosis when mucocutaneous features are absent. Leukocytosis is not expected in enteroviral infections. Leukocyte counts are < 15 000/mm3 in 90% of children hospitalized with influenza; leukopenia occurs in approximately 25%.21,29 Adenoviral infection can be associated with a brisk inflammatory response, probably owing to cytokine stimulation.56 In a study of 105 children hospitalized with adenoviral infection, the mean peripheral leukocyte count was 13 300/mm3 (10% had counts > 20 000/mm3), and almost 30% had an erythrocyte sedimentation rate > 40 mm/hour. Thrombocytosis is universal in Kawasaki disease but not until after the first week of illness. Thrombocytopenia (probably immunologically mediated) is described in rare cases of Kawasaki disease, usually unassociated with consumption coagulopathy.57 Thrombocytopenia is common in invasive bacterial infection and toxic shock syndromes and occurs in the early stages of Rocky Mountain spotted fever58 and ehrlichiosis;9 modest thrombocytopenia can occur with uncomplicated enteroviral infections. A normal or low, rather than high, neutrophil count with extreme left shift is a typical effect of bacterial toxins;59 thrombocytopenia and consumption coagulopathy (with extremely low erythrocyte sedimentation rate) are also common. Pyuria without hematuria is common in Kawasaki disease but infrequent in other conditions except Stevens–Johnson syndrome. Empiric therapies are shown in Table 15-1, and conditions other than Stevens–Johnson syndrome are discussed in depth in specific chapters. Prompt and aggressive antibiotic and supportive therapies are lifesaving in invasive bacterial infections and in staphylococcal exfoliative toxin syndrome in young infants. Antibiotic therapy may be beneficial in staphylococcal toxic shock, especially to treat focal infection and to prevent recurrence, but aggressive reversal of hypovolemia and cardiovascular shock is paramount. The important challenges in streptococcal toxic shock are: (1) halting microbial replication and toxin production; (2) reversing tissue ischemia, acidosis, hypovolemia, and hypotension; and (3) supporting multiorgan failure related to diffuse thromboemboli. Data from animal models of streptococcal necrotizing myositis, as well as clinical observations suggest that clindamycin is superior to b-lactam agents for treatment.50,60 Corticosteroid therapy is contraindicated as initial treatment of Kawasaki disease, is not indicated in staphylococcal exfoliative toxin syndrome, or staphylococcal or streptococcal toxic shock, and is controversial (with no valuable published studies) in Stevens–Johnson syndrome. Aggressive care in burn units for children with Stevens– Johnson syndrome or toxic epidermal necrolysis with more than 10% skin loss improves outcome.61 Many experts would give a 5-day trial of corticosteroids to a subset of children with Stevens–Johnson syndrome who are toxic and have mucosal involvement in addition to the mouth, skin eruption for 3 days or less, and less than 20% skin denudation.19 Excess rate of infections62 and mortality in drug-induced toxic epidermal necrolysis61 have been reported with use of corticosteroids in uncontrolled studies. Therapy with immune globulin intravenous (IGIV) is critical for optimal outcome in patients with Kawasaki disease15,16 and is lifesaving for very young infants.45 No randomized prospective study with adequate power has been performed to evaluate IGIV therapy in staphylococcal or streptococcal toxic shock. Case reports and retrospective, matched case-control studies in adults suggest a potential benefit. Dosage used was 500 mg/kg per day μ 5 days for staphylococcal disease and 2 g/kg once for streptococcal disease.63 A European randomized, double-blind, placebo-controlled trial of IGIV (1g/kg on day 1 followed by 500 mg/kg on days 2 and 3) for streptococcal toxic shock64 showed trend of benefit but did not answer the question soundly.65 Anecdotal reports of use of IGIV for severe Stevens–Johnson syndrome suggest possible beneficial effect.66,67

124

SECTION

CHAPTER

B Cardinal Symptom Complexes TABLE 16-1. Age-Related Causes of Serious Bacterial Infections in

16

Very Young Infantsa BACTEREMIA/MENINGITIS

Fever without Localizing Signs

< 1 month

Group B streptococcus Escherichia coli (and other enteric gramnegative bacilli) Listeria monocytogenes Streptococcus pneumoniae Haemophilus influenzae Staphylococcus aureus Neisseria meningitidis Salmonella spp.

1–3 months

Streptococcus pneumoniae Group B streptococcus Neisseria meningitidis Salmonella spp. Haemophilus influenzae Listeria monocytogenes

Eugene D. Shapiro

The vast majority of young children with fever and no apparent focus of infection have self-limited viral infections that resolve without treatment and are not associated with significant sequelae. However, a small proportion of young children with fever who do not appear to be seriously ill may be seen early in the course of a serious bacterial illnesses or may have occult bacteremia. A very small proportion of these children may subsequently develop a serious illness such as meningitis. Despite numerous studies that attempted to identify the febrile child who appears well but who actually has a serious infection and to assess potential interventions, no clear answers have emerged.1–4 Studies show that parents are generally more willing than are physicians to assume the small risk of serious adverse outcomes in exchange for avoiding the short-term adverse effects of invasive diagnostic tests and antimicrobial treatment.5,6 The best approach to the management of the febrile child combines informed estimates of risks, careful clinical evaluation and follow-up of the child, and judicious use of diagnostic tests.

OSTEOARTICULAR INFECTIONS

< 1 month

Group B streptococcus Staphylococcus aureus

1–3 months

Staphylococcus aureus Group B streptococcus Streptococcus pneumoniae

URINARY TRACT INFECTION

0–3 months

ETIOLOGIC AGENTS The list of microbes that cause fever in children is extensive. Relative importance of specific agents varies with age, season, and associated symptoms. The focus of this chapter is the febrile child with occult bacterial infection. Table 16-1 shows the most common causes of serious bacterial infection in children younger than 3 months.7,8 The division at 1 month is not absolute; considerable overlap exists. It is also important to remember that certain viruses, notably herpes simplex and enteroviruses, can cause serious infections in neonates, mimicking septicemia, and beginning as fever with no apparent focus of infection. In children older than 3 months, most bacterial infections with no apparent focus are caused by Streptococcus pneumoniae (in unimmunized children), Neisseria meningitidis, or Salmonella spp. (the latter often occurring in association with symptoms of gastroenteritis). Haemophilus influenzae type b, formerly an important cause of occult bacteremia, has become rare and the incidence of infection with Streptococcus pneumoniae has fallen substantially since the universal administration of effective vaccines began.9,10 Other common causes of invasive bacterial infections in children, such as Staphylococcus aureus, are usually associated with identifiable focal infections.

EPIDEMIOLOGY Children Younger than 3 Months The risk of serious bacterial infection varies with age. Although longitudinal studies have shown that only 1% to 2% of all children are brought to medical attention for fever (temperature = 38.3°C) during the first 3 months of life, a greater proportion of such febrile infants has serious bacterial infections compared with older children.10–14 Risk is greatest during the immediate neonatal period and through the first month of life (and is heightened in the infant born prematurely). In a prospective study conducted at the University of Rochester, researchers identified factors associated with a low risk of serious bacterial infection in febrile infants younger than 3 months.15 Among 233 infants who were born at term with no perinatal complications or underlying diseases, who had not received antibiotics, and who were

Escherichia coli Other enteric gram-negative bacilli Group D streptococcus (including Enterococcus species)

a

In decreasing order of frequency.

hospitalized for fever and possible septicemia, 144 (62%) were considered unlikely to have a serious bacterial infection and fulfilled all of the following criteria: (1) no clinical evidence of infection of the ear, skin, bones, or joints; (2) white blood cell (WBC) count between 5000 and 15,000/mm3; (3) < 1500 band cells/mm3; and (4) normal results of urinalysis. Only 1 of these 144 infants (0.7%) had a “serious” bacterial infection (Salmonella gastroenteritis), and none had bacteremia. By contrast, among the 89 infants who did not meet one or more of these criteria, 22 (25%) had a serious bacterial infection (P < 0.0001) and 9 (10%) had bacteremia (P < 0.0005). Subsequent investigators have corroborated results of the Rochester study.16–19 Although investigators have used slightly different criteria to define young febrile infants at low risk of serious bacterial infection (and some investigators excluded children younger than 1 month), all found that the risk of a serious bacterial illness in the group defined as being at low risk is, indeed, very low. In a metaanalysis of studies of febrile children younger than 3 months, the risks of “serious bacterial illness,” bacteremia, and meningitis were 24.3%, 12.8%, and 3.9%, respectively, in “high-risk” infants and 2.6%, 1.3%, and 0.6%, respectively, in “low-risk” infants.14 The negative predictive value for serious bacterial illnesses of infants fulfilling low-risk criteria ranged from 95% to 99% (and was 99% for bacteremia and 99.5% for meningitis).19 Thus, although the risk of serious bacterial infection is high in febrile infants younger than 3 months of age with no apparent focus, clinical and laboratory assessment can be used to identify the slightly more than 50% of infants at very low risk. An observational study of more than 3000 infants < 3 months of age with fever > 38°C treated by practitioners and reported as part of the Pediatric Research in Office Settings network found that the majority (64%) were not hospitalized.4,20 Practitioners individualized management and relied on clinical judgment; current guidelines were followed in only 42% of episodes.20,21 Outcomes of the children were excellent. If the guidelines had been followed, outcomes would not have improved but there would have been both substantially more laboratory tests performed and more hospitalizations.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Fever without Localizing Signs

Children Older than 3 Months During the 1970s, reports of occult pneumococcal bacteremia began to appear.22–24 It became apparent that some children aged 3 months or older with fever who did not appear to be toxic and who had no apparent focus of infection had bacteremia, most often due to Streptococcus pneumoniae but occasionally due to H. influenzae type b or N. meningitidis.25–32 Moreover, in some instances, serious focal infections such as meningitis developed in children with occult bacteremia. The overall rates of bacteremia reported in studies of unselected febrile children range from 3% to 8%.33 High fever alone is not associated with an excessive risk of bacteremia; the two largest studies of children 3 to 36 months of age with fever ≥ 39°C and no apparent focus of infection documented bacteremia in 2.8% (27 of 955) and 2.9% (195 of 6733) of children, respectively.34,35 Frequency of occult bacteremia in disadvantaged urban populations and in suburban populations served by private practitioners is similar.36,37 Risk of bacteremia is greater when very high fever is associated with high total WBC count.21,27,33,38,39 Most children with occult bacteremia have transient infection and recover (without antimicrobial therapy) without having a serious complication such as meningitis or septic shock develop.21,40–45 Risk of meningitis complicating occult bacteremia varies with bacterial species. Compared with the risk of developing meningitis with occult pneumococcal bacteremia (4 of 225; 1.8%), the odds of developing meningitis was 15 times greater for children with occult H. influenzae type b bacteremia and 81 times greater for children with occult N. meningitidis bacteremia.32 The risk of meningitis among children with occult bacteremia has decreased substantially because occult bacteremia due to H. influenzae type b has been virtually eliminated since licensure of the conjugate vaccine.9After the virtual elimination of bacteremia due to H. influenzae type b due to vaccination, it was estimated that among children aged 3 to 36 months who are evaluated for high fever without a focus, bacterial meningitis would develop subsequently in approximately 1 of 1000 to 1500 untreated children.6 Consequently, even if “expectant” antimicrobial treatment of febrile children were 100% effective, it would have been necessary to treat 1000 to 1500 children to prevent 1 case of meningitis. Following licensure in the United States in 2000 of the polysaccharide–protein conjugate vaccine against 7 serotypes of pneumococci,46,47 substantial reduction in incidence of invasive disease has been documented in vaccinated children,10,48 thereby markedly reducing the problem of occult bacteremia and its consequences.

LABORATORY FINDINGS AND DIAGNOSIS Various diagnostic tests to quantify the risk of bacteremia and its complications have been assessed and include the WBC count and differential, microscopic examination of buffy coat of blood, erythrocyte sedimentation rate, C-reactive protein, morphologic changes in peripheral blood neutrophils, and quantitative cultures of blood.21,38,48–54 In addition, clinical scales have been developed to help identify the febrile child with a serious illness.55 Unfortunately, no test has sufficient sensitivity and positive or negative predictive value to be clinically useful for an individual patient. For example, in one prospective study of children with a temperature > 40°C, those with a WBC count of ≥15,000/mm3 had a risk of bacteremia three times greater than did those with a WBC count of < 15,000/mm3.27 However, the positive predictive value of this test for bacteremia was only 14%; thus, more than 85% of highly febrile children with a WBC count of ≥15,000/mm3 did not have bacteremia. Others have reported similar results.38,56 Subsequent to these studies, the prevalence of occult bacteremia has substantially diminished because of implementation of vaccination programs, so the positive predictive value of such test results is now even lower. Consequently, such testing as routine can no longer be justified.57

CHAPTER

16

125

The outcome of primary concern is not occult bacteremia but meningitis. An ideal diagnostic test would specifically identify those febrile children at risk of a serious complication, because many focal infections after bacteremia (e.g., pneumonia or cellulitis) can be treated when they become apparent and are not usually associated with serious sequelae. Unfortunately, there is no such test.

MANAGEMENT Although there is no single correct approach to the management of febrile infants without localizing signs who appear well, studies have provided data based on which informed decisions can be made. There is general agreement that febrile children who are “very young” (variably considered to be younger than 3, 2, or 1 month of age) should be managed differently from the way in which older children are managed.

Children Younger than 3 Months Because of the substantially greater risk of serious infections in very young infants with fever and the difficulty in assessing degree of wellness accurately, pediatricians have approached the management of such infants conservatively. Some clinicians adhere to a protocol of treating all young infants with fever and no apparent focus of infection with broad-spectrum antimicrobial agents administered intravenously in the hospital until the results of cultures of the blood, urine, and cerebrospinal fluid (CSF) are known.58 Although perceived as the “safe” approach, such management incurs considerable financial cost and risk of iatrogenic complications and of diagnostic misadventures associated with hospitalization.59–61 These risks include errors in the type and dosage of drugs, complications of venous cannulation (such as phlebitis and sloughing of the skin), and nosocomial infections. In addition, hospitalization of a young infant is a major disruption for the family and may potentiate the development of the “vulnerable child” syndrome.62 Investigators have found that selected young infants with fever can do well without hospitalization.15–18,20,63 Consequently, many experts believe that febrile infants from 2 to 3 months of age with no apparent focus of infection who appear well can be managed without laboratory tests or hospitalization, provided that careful follow-up is ensured. Others require laboratory criteria predictive of low risk (some include normal CSF analysis in the criteria). Some would simply observe the patient very closely without giving antimicrobial therapy; others would treat all such infants for 2 days with a single daily dose of ceftriaxone (50 mg/kg), administered parenterally, while awaiting the results of the cultures. Either approach can be defended. If an antimicrobial agent is to be administered, cultures of the blood, urine, and CSF should be obtained first. Febrile infants at low risk of serious bacterial infection for whom adequate home observation and follow-up cannot be ensured should be hospitalized and can be observed without antimicrobial treatment. Doing so (if the child appears well) is reasonable and avoids the adverse side effects of antimicrobial agents and intravenous cannulation, shortens the duration of hospitalization, and saves money without placing the child at significant risk of complications.2,17,20,62 Most infants with fever who are younger than 1 month should be hospitalized and treated with antimicrobial therapy, although, in selected instances, hospitalization without antimicrobial treatment, or management as an outpatient (after laboratory evaluations, including analysis of CSF), may be reasonable. If a decision is made to administer antimicrobial agents intravenously, ampicillin (100 to 200 mg/kg per day q6 hours) plus gentamicin (7.5 mg/kg per day q8 hours) provides a suitable spectrum of activity until results of cultures permit discontinuation or alteration of treatment. Ampicillin plus a third-generation cephalosporin (ceftriaxone, 50 mg/kg per day in 1 dose; or cefotaxime, 150 mg/kg per day q8 hours) could be chosen, but there is no theoretical or proven benefit in children without meningitis.

126

SECTION

B Cardinal Symptom Complexes

Before initiating antimicrobial treatment, cultures of the blood, urine (obtained by either urethral catheterization or suprapubic aspiration of the bladder), and CSF should be obtained.

Children Older than 3 Months Children older than 24 months of age who appear well and have no apparent focus of infection can be followed clinically without laboratory tests or treatment with antimicrobial agents; risk of occult or serious bacterial infection is extremely low. For febrile children (i.e., those with a temperature ≥ 39°C) aged 3 to 24 months, there has been controversy about whether and which diagnostic tests should be performed and whether “expectant” antimicrobial treatment should be initiated.6,21 Although results of a complete WBC count and differential may help to identify children at increased risk of occult bacterial infection, these tests have no direct therapeutic impact, and the positive predictive value is poor. Substantial evidence suggests that obtaining blood cultures routinely in these children has little impact on outcome (although false-positive blood culture results lead to substantial unnecessary costs).64,65 The authors of a carefully conducted decision analysis concluded that a strategy of obtaining blood cultures in all such febrile children did more harm than good, in part because many children in whom bacteremia spontaneously clears are hospitalized and treated unnecessarily.2 It is not clear that “expectant” therapy of febrile children prevents serious complications such as meningitis. Two large randomized clinical trials were conducted of the efficacy of “expectant” antimicrobial treatment in preventing focal complications in all febrile (temperature of ≥39°C) children 3 to 36 months of age with no apparent focus of infection. In the first trial, 955 children were randomized to receive either amoxicillin or placebo in a double-blind manner; no statistically significant difference in outcomes was observed.34 However, because of the rarity of focal complications of bacteremia (2 of 10 (10%) patients in the amoxicillin group and l of 8 (12.5%) patients in the placebo group), there was insufficient statistical power to exclude the possibility that amoxicillin is effective. In the other clinical trial, 6733 children aged 3 to 36 months with a temperature of ≥39°C and no apparent focus of infection (or with otitis media) were randomized to receive 1 dose of ceftriaxone (50 mg/kg) or amoxicillin (20 mg/kg per dose) tid for 2 days.35 Among children with occult bacteremia, no statistically significant difference was observed in the frequency of definite and probable complications (ceftriaxone, 3 of 101 (3.0%) patients; amoxicillin, 6 of 91 (6.6%) patients). Although the investigators seemed to endorse routine use of ceftriaxone, their methods and conclusions have been criticized.6,66 Criticisms have included biased definition of the outcomes (“definite” infections required a positive culture at follow-up, which is less likely in patients treated with ceftriaxone), incomplete follow-up, and inappropriate statistical analyses (because the analysis was only of the children who had bacteremia and not of all who were randomized and treated). Furthermore, 4 of 5 children in whom meningitis developed were infected with H. influenzae type b, a problem that has been virtually eliminated since the introduction of effective vaccines. Consequently, even if one accepted the investigators’ conclusions, these data are no longer relevant. Routine antimicrobial treatment of febrile children for possible occult bacteremia is not without risk.6,64–67 In addition to substantial financial costs, antimicrobial agents have predictable as well as idiopathic adverse side effects. Widespread use of antibiotics selects for resistant organisms. In addition, loss of clinical improvement as a marker of natural history of infection in a partially treated child, difficulty in interpreting mildly abnormal CSF at follow-up, and frequent contaminated blood cultures all lead to increased frequency of unnecessary hospitalization and increased use of laboratory tests and of antimicrobial therapy. Perhaps most important, thoughtful assessment, individualized management, and close follow-up of the febrile child may be forgotten.

In view of current data, including the virtual elimination of infections with H. influenzae type b and the marked reduction in the incidence of invasive infections with Streptococcus pneumoniae in children, the following approach seems appropriate: The febrile child should be carefully assessed for foci of infection and, if foci are found, should be treated according to likely pathogens. If the child appears toxic, appropriate cultures and diagnostic tests should be performed and antimicrobial treatment (usually with cefotaxime, 150 mg/kg per day in divided doses q8 hours or ceftriaxone, 50 mg/kg once a day) should be initiated (some would add vancomycin, 40 mg/kg per day in divided doses q6 to 8 hours); most such children should be hospitalized. If no focus is found and the child does not appear toxic, no diagnostic tests are indicated routinely. Parents should be instructed to look for signs that a more serious problem is developing (e.g., persistent irritability or lethargy, inattentiveness to the environment). Serial observations should be planned that will permit subsequent clinical and laboratory evaluation and antimicrobial treatment as indicated.

Other Considerations This chapter focuses on invasive bacterial infections (particularly bacteremia) as a cause of fever without apparent focus. It should not be forgotten that urinary tract infection is another important cause of fever in young children.68 Indeed, urinalysis may be a more appropriate diagnostic test in the febrile infant than complete blood count and blood culture.6,69 In addition, viral infections are the major cause of fever in infants and toddlers. Human herpesvirus 6 (and, to a lesser extent, human herpesvirus 7) has been implicated as a common cause of fever in young children.70–72 Although other serious illnesses, such as autoimmune diseases and inflammatory bowel disease, can manifest as fever without a focus, they are rare and come to attention because of persistence or recurrence of fever (see Chapter 17, Prolonged, Recurrent, and Periodic Fever Syndromes).

CHAPTER

17

Prolonged, Recurrent, and Periodic Fever Syndromes Sarah S. Long and Kathryn M. Edwards

Diagnosing and managing patients with prolonged, recurring, or periodic fever requires extensive review of symptoms and systems to establish onset and cardinal feature(s) of illness, to define the exact fever pattern and to understand the context of illness within the patient’s family and past medical history. Temperature is interpreted with the use of norms for age and sex. Figure 17-1 depicts maturational changes in temperature.1 Normal mean rectal temperature of infants for the first 3 years of life is significantly above 37°C. At 18 months of age, mean rectal temperature for males is 37.7°C with a standard deviation (SD) of 0.38°C; thus, a rectal temperature of 38.5°C is < 2 SD above the mean. Defining fever patterns (such as shown in Box 17-1) is useful to prioritize differential diagnosis and investigation.2 A disciplined physical examination is aimed at identifying target organ abnormalities of potential infectious agents and noninfectious diseases such as malignancies and autoimmune, autoinflammatory, endocrine, and metabolic disorders. In children with prolonged fever, laboratory testing should include simple screening and then is targeted to specific

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Prolonged, Recurrent, and Periodic Fever Syndromes 100.0

17

127

BOX 17-2. Physical Examination and Laboratory Testing in Children with Prolonged, Recurrent, or Periodic Fever

99.8 Mean temperature (°F)

CHAPTER

99.6 99.4 99.2 Boys Girls

99.0 98.8 98.6 0

3

6

9

12

15 18 21

24 27 30 33 36

Age in months Figure 17-1. Normal mean rectal temperature is shown for boys and girls through 36 months of age. 100°F (37.8°C). Data from Bayley M, Stolz HR. Maturational changes in rectal temperature of 61 infants from 1 to 36 months. Child Dev 1937;8:195.

BOX 17-1. Fever Patterns of Illness (for Purpose of Defining an Approach) Prolonged fever: A single illness in which duration of fever exceeds that expected for the clinical diagnosis (e.g., > 10 days for viral upper respiratory tract infections; > 3weeks for mononucleosis) Or A single illness in which fever was an initial major symptom and subsequently is low-grade or only a perceived problem Fever of unknown origin: A single illness of at least 3 weeks’ duration in which fever > 38.3°C is present on most days, and diagnosis remains uncertain after 1 week of intense evaluation Recurrent fever: A single illness in which fever and other signs and symptoms wax and wane (sometimes in relationship to discontinuation of antimicrobial therapy) Or Repeated unrelated febrile infections of the same organ system (e.g., sinopulmonary, urinary tract) Or Multiple illnesses occurring at irregular intervals, involving different organ systems in which fever is one variable component Periodic fever: Recurring episodes of illness for which fever is the cardinal feature, other associated symptoms are similar and predictable, duration of episodes is days to weeks, with intervening weeks to months of complete wellbeing. Episodes can have either “clockwork” or irregular periodicity

organs as identified by history, physical examination, or prioritized clinical differential diagnosis (Box 17-2).2 In children with periodic fever, the main goal of performing simple laboratory tests is to confirm normal organ system function, to lead the clinician toward a specific disorder (e.g., recurrent urinary tract infection), or to support the diagnosis of a noninfectious periodic fever syndrome. In children with any fever pattern, casting a broad net of antibody testing for unusual infectious agents is rarely fruitful in the absence of a specific exposure history or clinical finding.

FEVER OF UNKNOWN ORIGIN (FUO) Definition and Approach Petersdorf & Beeson’s3 prospective evaluation from the 1950s of 100 adults who had well-defined entry criteria for prolonged unexplained fever significantly advanced the understanding of FUO. Criteria for FUO in Petersdorf & Beeson’s study in adults were: (1) illness of more than 3 weeks’ duration; (2) fever higher than 38.3°C on several

PHYSICAL EXAMINATION • Growth chart • Thorough general examination • Careful organ-specific examination • Notation of mouth ulcers, exanthem, joint abnormalities, lymph nodes TESTS • Complete blood count with manual differential count of white blood cellsa • Erythrocyte sedimentation rate and C-reactive proteina • Screening serum chemistry tests (and uric acid level if prolonged fever) • Serum quantitative immunoglobulin levels • Urinalysis • Urine culture • Chest plain radiograph (if prolonged or recurrent fever) • Other imaging only as directed by examination • Blood culture (if prolonged fever) a

Perform during episode and interval if periodic fever.

occasions; and (3) uncertain diagnosis after 1 week of study in the hospital. A high incidence of serious disease was noted. Infections accounted for 36% of cases, neoplastic diseases for 19%, and collagen vascular diseases for 13%; no diagnosis was made in 7% of cases. In a separate assessment of individuals with “fever” that did not exceed 38.3°C (even if protracted), diagnosis of significant illness was rare. A prospective, population-based study in the 1990s was performed in 167 adults in The Netherlands, with the use of fixed epidemiologic entry criteria and diagnostic protocol.4,5 The classic definition of FUO was adjusted to modern times; immunocompromised patients were excluded, and the third criterion, 1 week of hospitalization, was replaced with 1 week of intensive evaluation (usually including computed tomography [CT] of the abdomen) and pursuit of potentially diagnostic clues without diagnosis. Infections accounted for 26% of cases, neoplasms for 13%, and noninfectious inflammatory diseases for 24%. No diagnosis was made in 30% of patients, three-quarters of whom recovered spontaneously. At least nine studies and reviews of FUO in children have been reported since 1960.6–14 Small numbers, variability in definition of FUO and study methodologies, and subspecialty referral biases preclude meaningful meta-analysis or generalization of the results. Table 17-1 shows the categories of diagnoses made in studies of children with FUO.6–12,15 Infections are most common. Decreasing stringency of entry criteria correlates with increased likelihood of self-limited conditions. When fever approaches 4 weeks’ duration without a definable source, the rate of life-altering diagnoses, such as malignancy, collagen vascular disease, and inflammatory bowel disease, rises to 40% in some series.6,7,9,10 For purposes of management, FUO in children is only considered after a minimum of 14 days of daily documented temperature of 38.3°C or greater without apparent cause after performance of repeated physical examinations and screening laboratory tests. This definition assumes exclusion of: (1) protracted but waning symptoms from acute, self-limited respiratory tract infections; (2) welldocumented periodic fever; and (3) repeated but unrelated episodes of fever with identifiable causes. Almost one-half of children referred for FUO do not meet fever and exclusion criteria. In those who do, careful reassessment of history and physical examination reveals potentially diagnostic clues that should be pursued with targeted testing before application of a nonfocused diagnostic strategy. FUO is more likely an unusual presentation of a common disease than an uncommon disease.16 A chest radiograph and CT of the sinuses are viewed as “early studies” in the workup. Chest CT may be indicated when granulomatous or embolic disease is suspected. Abdominal CT is a major advance in detecting disease and is a rewarding early investigation for certain patients. In an evaluation of 28 children with prolonged fever

128

SECTION

B Cardinal Symptom Complexes

TABLE 17-1. Diagnoses in Published Series of Children with Fever of Unknown Origin Brewis (1965)6

Dechovitz & Moffet (1968)9

McClung (1972)8

Pizzo et al. (1975)7

Feigin & Shearer Lohr & Hendley (1976)10 (1977)11

Jacobs & Schutz (1998)12

5–7 days No 165

2 weeks No 8

3 weeks Yes, 1 week 99

2 weeks No 100

2 weeks Yes, 1 week 20

5 weeks Yes, 1 week 54

2 weeks No 146

63 (38%) 54 9 9 (5%) 0 3 (2%) 18 (11%) 9 (5%)

2 (25%) 0 2 6 (75%) 0 0 0 0

29 (28%) 14 15 11 (11%) 3 (3%) 8 (8%) 16 (16%) 11 (11%)

52 (52%) 31 21 20 (20%) 0 6 (6%) 10 (10%) 12 (12%)

7 (35%) 1 6 3 (15%) 1 (5%) 1 (5%) 2 (10%) 6 (30%)

18 (33%) 2 16 8 (15%) 3 (6%) 7 (13%) 8 (15%) 10 (19%)

64 (44%) 0 64 9 (6%) 2 (1%) 4 (3%) 5 (3%) 62 (42%)

0 1

21 (20%) 1

– 9

– 2

– 9

62 (42%) 0

STUDY CRITERIA

Daily temperature > 38.0°C Inpatient evaluation Total cases DIAGNOSIS

Infection Respiratory tracta Other Collagen disease Inflammatory bowel disease Malignancy Miscellaneous No diagnosis OUTCOME

Resolved during investigation 35 (21%) Death(s) 1 a

Upper and lower respiratory tract infections.

by means of abdominal CT, 86% of children in whom clinical features directed suspicion to the abdomen had positive CT Àndings; the yield of the procedure in others was low.17 In another study of 109 children with FUO, multiple modalities, including radioisotope scans, rarely led to an unsuspected diagnosis.18 “DeÀnite normal” interpretation of imaging studies is unusual, and any other interpretation leads to further studies, all of which can divert focus from a correct diagnosis. In a rigorous prospective study of FUO in adults in whom no potentially diagnostic clues were gleaned from the history or physical examination, the diagnostic yield of scintigraphy, other imaging modalities, liver or bone marrow biopsy, and screen for endocrine abnormalities was nil.5 Biopsy of enlarged lymph nodes conÀned to cervical and inguinal areas (in a patient with normal Àndings on chest radiograph and abdominal ultrasonography) was also fruitless. The investigators of this study recommend repeating a thorough exploration of the history, physical examination, and “Àrst-level” laboratory tests and waiting for potential diagnostic clues to appear.5This approach is appropriate in children as well, unless there is progression of weight loss, or of an abnormality on examination or simple screening tests.

infection, vertebral and pelvic osteomyelitis, bartonellosis, and complicated urinary tract infection. None of the children had respiratory tract infection, undoubtedly because the organ system involved was identiÀed at preliminary evaluations.12 Agents that produce granulomatous infections cause FUO disproportionately and frequently involve the visceral organs, reticuloendothelial system, and bone marrow.21 Such agents include Bartonella, Mycobacterium, Salmonella, Brucella, and Francisella species.

Salmonellosis Salmonella typhi and nontyphoid Salmonella species were responsible for all cases of enteric infection causing FUO in the Àve series. Salmonellosis is a major cause of FUO in tropical areas. Although fever from salmonellosis is usually of shorter duration, occasionally high spiking fevers can persist for weeks.22,23 Abdominal pain, often present, can be mild enough to go unnoticed. Relative bradycardia is not usually seen in children with nontyphoid Salmonella infection.24 Cultures of blood and stool, or occasionally only of bone marrow, conÀrm the diagnosis (see Chapter 146, Salmonella Species).

Yersiniosis

Etiology A comprehensive list of reported causes of FUO provides a checklist. It cues but never supplants continuous rethinking of diagnoses based on the patient’s history, course, and examination; it should not be used as a standard for serologic testing to “rule out” diagnoses.

Infectious Causes Infections account for more than one-third of cases of FUO. Generally, the agents, site of infection, and clinical manifestations are typical but subtle and overlooked, or are associated with an insidious or prolonged course.2,19,20 Table 17-2 is a compilation of infections identiÀed in children with fever of > 2 weeks’ duration from FUO case series and case reports. Undoubtedly, some had potentially diagnostic clues at the time of evaluation.21 In older series, almost half the patients with infections had upper or lower respiratory tract infections exclusive of granulomatous disease. Pneumonia, sinusitis, and otitis media accounted for the majority. In the latest prospective series of children, the most common infections were Epstein–Barr virus

Yersinia enterocolitica can be associated with prolonged fever in infants25 and children.26 Although most patients have short duration of fever, diarrhea or abdominal pain, or both, gastrointestinal symptoms occasionally occur during the second to third week of fever. The pathogen can be isolated from culture of stool, sometimes requiring cold enrichment (see Chapter 148, Yersinia Species).

Endocarditis History of fever alone for 2 weeks or longer (frequently with abatement during oral antibiotic therapy) in a child with congenital or acquired heart disease should prompt investigation for infective endocarditis. Endocarditis also occasionally occurs in children without predisposing heart disease, after spontaneous bacteremia or septicemia, or after bacteremia from focal infection or an indwelling intravascular catheter. Although acquisition of a prominent murmur is the hallmark of endocarditis, it is not an early or constant Ànding. Multiple samples of large volumes of blood for culture conÀrm the diagnosis except in rare circumstances of chronic or right-sided

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Prolonged, Recurrent, and Periodic Fever Syndromes

CHAPTER

17

129

TABLE 17-2. Infectious Causes of Fever of Unknown Origin (FUO) in 160 Children in FUO Series Percent of Cases in Compiled Seriesa Case Reports

< 1%

1–5%

6–10%

Intra-abdominal/ retroperitoneal abscess Visceral abscess Hepatitis Odontogenic infection Nontuberculous mycobacteria Q fever Syphilis Tickborne typhus Leptospirosis Chronic meningococcemia Histoplasmosis Toxocara canis and T. cati Inflammatory pseudotumors

Malaria Rocky Mountain spotted fever Cystic fibrosis Brucellosis Blastomycosis HSV generalized

Bartonellosis Tonsillopharyngitis/ Otitis media/sinusitis peritonsillar abscess Tuberculosis Lower respiratory tract Bacterial meningitis/ infection parameningeal abscess Endocarditis Streptococcosis Enteric infection Tularemia Ehrlichiosis

11–15%

6–20%

Systemic viral syndrome Infectious mononucleosis (not specified) (EBV and CMV) Urinary tract infection

CMV, cytomegalovirus; EBV, Epstein–Barr virus; HSV, herpes simplex virus. a Series data collated from references 6–12.

infections or uncultivatable pathogens such as Brucella, Coxiella, and Rickettsia species. In any patient in whom multiple blood cultures are positive or bacteremia persists after initiation of antibiotic therapy, an intravascular focus of infection should be investigated (see Chapter 39, Endocarditis and Other Intravascular Infections). Viridans streptococci and Staphylococcus aureus are prominent causes in community-acquired endocarditis, the latter being responsible for most cases without underlying cardiac defects. Recovery of a socalled HACEK organism (any member of the following microbes, whose names yield the acronym: Haemophilus aphrophilus, Actinobacillus actinomycetemcomitans, Cardiobacterium hominis, Eikenella corrodens, Kingella kingae) in a patient with prolonged fever, with or without underlying heart disease, suggests endocarditis.

Tuberculosis

Mononucleosis Epstein–Barr virus infection frequently causes prolonged fever and was the most common diagnosis in a 1998 study of children with FUO.12 “Typhoidal” cases without other classic features of mononucleosis, such as lymphadenopathy, pharyngitis, and malaise, or with atypical features, such as pneumonia, can come to medical attention as FUO; serologic testing is definitive (see Chapter 207, Epstein–Barr Virus).31 Systemic cytomegalovirus (CMV) infection can cause a mononucleosis syndrome, with prominent fever and malaise and little or no pharyngitis symptoms. Splenomegaly, atypical lymphocytosis, and mild hepatitis are clues; serologic testing plus culture of urine in the infant or of the throat in the older child, or a molecular test for viremia may be required to confirm the diagnosis (see Chapter 206, Cytomegalovirus). Hepatitis virus A, B, C, and HIV can cause mononucleosis-like FUO.

Tuberculosis is a cause of pediatric FUO in studies separated by decades, the relative incidence reflecting prevalence in adults. Tuberculosis is considered in every patient with FUO, because prolonged fever alone is one manifestation. Extrapulmonary or miliary disease more commonly causes FUO than does tubercular pneumonia.28 Diagnosis in such cases can be difficult and depends on one or more of the following features: history of exposure, positive Mantoux skin test result, suggestive symptoms, signs or radiographic findings, and isolation of the organism (usually from gastric aspirate, blood, bone marrow, or another extrapulmonary site) (see Chapter 134, Mycobacterium tuberculosis). Skin test frequently is negative in disseminated disease, immunosuppressed patients, and patients with concurrent human immunodeficiency virus (HIV) infection.28,30 The diagnosis of tuberculosis should prompt testing for HIV infection.

Bartonella henselae, the agent of cat-scratch disease, is an increasingly identified cause of FUO.12 Other symptoms, such as listlessness, anorexia, and headache, are common but nonspecific. History of intimate contact with a cat (usually a kitten) is often overlooked in patients with this presentation if they do not also have regional lymphadenopathy. FUO due to B. henselae, with and without hepatosplenic involvement or with or without bone or bone marrow involvement, occurs in healthy children as well as in immunosuppressed patients.12,32–35 Diagnosis is confirmed by serology, with the B. henselae titer expected to be at least 1:128 (see Chapter 160, Bartonella Species).32

Nontuberculous Mycobacterial Infection

Brucellosis

Several case reports have documented Mycobacterium avium complex endobronchial infection and mediastinal and hilar lymphadenopathy in immunocompetent children with prolonged fever.30 Most patients have respiratory tract symptomatology. Organisms are usually isolated from gastric aspirates. Diagnosis is confirmed by bronchoscopy, mediastinoscopy, or open biopsy. Concurrent malignancy, HIV infection, or another cause of immunodeficiency should be excluded when pulmonary or disseminated infection due to M. avium complex is diagnosed (see Chapter 135, Mycobacterium sp. Nontuberculosis).

The primary manifestation of brucellosis is FUO with spiking fevers and lethargy lasting for up to 4 weeks.36,37 Its importance as a cause of FUO is related to exposure to Brucella spp. through consumption of unpasteurized dairy products from an infected cow, sheep, or goat or through contact with infected farm animals. Some patients have tender hepatomegaly, splenomegaly, epistaxis, arthralgia, myalgia, and weight loss. Leukopenia or leukocytosis can occur. Elevation of serum hepatic transaminase levels is a constant feature. Serologic test confirms the diagnosis (see Chapter 161, Brucella Species [Brucellosis]).

Bartonellosis

130

SECTION

B Cardinal Symptom Complexes

Malaria

Tularemia

Malaria is the leading cause of FUO in countries where Plasmodium infections are prevalent.22 Prolonged or recurrent fever is the most common manifestation of infection. Eliciting a history of travel to areas where malaria is indigenous or receipt of blood transfusion is the most important clue. Rare cases have occurred after mosquito transmission within the United States. Although fever is usually paroxysmal, is abrupt in onset, and lasts only hours, such patterns may not be established in the early weeks of infection, and fever can be persistent. Diagnosis is made through examination of Wright-stained peripheral blood smears from multiple specimens (see Chapter 271, Plasmodium Species [Malaria]).

Humans are highly susceptible to Francisella tularensis, becoming infected through the bites of ticks, fleas, mites, or deer flies or through airborne or hand contact with infected animals.44,45 Although ulceroglandular presentation is distinctive, suggesting tularemia, typhoidal or pleuropulmonic tularemia can manifest predominantly as FUO. A history of exposure is the key to suspicion; serologic testing is diagnostic (see Chapter 171, Francisella tularensis).

Intra-Abdominal and Retroperitoneal Abscesses Intra-abdominal, visceral organ or pelvic abscesses (or osteomyelitis or pyomyositis) can be cryptogenic, causing prolonged, subtle symptoms and FUO. Usually a complication of subclinically perforated appendix or surgery in children, polymicrobial abscesses due to gut organisms occur at a periappendicular site, another intraperitoneal site, or in a visceral organ, especially the liver.38 Hepatic, splenic, nephric, or pelvic abscess also can occur as the metastatic focus of bacteremia in normal hosts; Staphylococcus aureus is the predominant pathogen. Visceral abscess as a complication of intestinal amebiasis can cause prolonged fever. Nephric or perinephric abscess and lobar nephronia from ascending infection can cause FUO; most, but not all, affected patients have urinary tract anomalies; urine culture can be negative when obstruction is present. A consistent complaint of localized pain or the finding of subtle organ enlargement or tenderness can be a clue to intra-abdominal, retroperitoneal, or pelvic infection. Abdominal CT is the most sensitive and specific diagnostic test and is justified in such cases, as well as when FUO approaches 3 weeks’ duration without localization. Findings of a mass, lymphadenopathy, enlarged kidney, ascites, or bowel wall abnormalities informs further management39,40 (see Chapter 71, Intra-abdominal, Visceral, and Retroperitoneal Abscesses; Chapter 21, Abdominal and Retroperitoneal Lymphadenopathy).

Odontogenic Infection Occult dental infection can cause persistent fever. In a review of adults with prolonged fever related to dental infection, only 19% had dental symptoms.41 Repeated examination, direct questioning about subtle dental symptoms (painful or loose teeth, discomfort with chewing or yawning, fever after eating), and performance of dental radiography lead to the diagnosis. Anaerobic lung abscess or pleural empyema, as a complication of dental infection, sinusitis, or obstructing foreign body, can cause fever and putrid breath without respiratory tract symptoms (see Chapter 27, Infections of the Oral Cavity; Chapter 187, Anaerobic Bacteria: Classification, Normal Flora, and Clinical Concepts).

Q Fever Coxiella burnetii, the cause of Q fever, is acquired by inhaling aerosolized dust contaminated by parturient cats or by consuming unpasteurized contaminated milk products. Fever, headache, and mild hepatic dysfunction are almost universal; fever occurs as a single episode of 5 to 10 days’ duration or as relapsing episodes over months.42 Serologic testing is diagnostic (see Chapter 169, Coxiella burnetii [Q Fever]).

Congenital Syphilis Prolonged fever can be the presenting sign of congenital syphilis in an infant.43 Hepatomegaly and monocytosis are additional clues. Results of serum reagin and antitreponemal antibody tests are positive (see Chapter 182, Treponema pallidum [Syphilis]).

Leptospirosis Although leptospirosis can cause FUO, diagnosis was not made once in published pediatric series. Manifestations can occur during the acute bacteremic phase, in a subsequent immunologic phase, or both. The immunologic phase can overlap or follow bacteremia after an asymptomatic afebrile period of 1 to 14 days. Diagnosis is considered for prolonged or biphasic acute febrile illnesses, especially when aseptic meningitis and hepatitis are evident. Rats and other wild rodents are reservoirs of Leptospira spp. Domestic dogs are infected and excrete organisms in urine (frequently asymptomatically when immunized). Contact with environment subject to the traffic of infected animals, especially after a rainy period, is a consistent part of the history. Serologic testing is diagnostic (see Chapter 184, Leptospira Species [Leptospirosis]).

Other Infections and Disorders of Unknown Etiology Rickettsia rickettsii (Rocky Mountain spotted fever) and Ehrlichia and Anaplasma species (ehrlichiosis and anaplasmosis) can cause FUO in areas where the infections are prevalent,12 especially when rash is absent or atypical; headache is a prominent clinical feature. Chronic Neisseria meningitidis infection is suspected when a typical rash is present. Borrelia burgdorferi (Lyme disease) rarely manifests as FUO alone. Toxocara canis and T. cati, the dog and cat ascarids, respectively, cause visceral larva migrans, which sometimes manifests as fever alone or with visceral organomegaly. Dramatic eosinophilia is the clue; serologic testing is diagnostic. HIV infection (usually with hepatosplenomegaly), permissive infection with other microbes in HIV-infected children, or HIV-induced immunologic dysregulation can cause FUO (see Chapter 111, Diagnosis and Clinical Manifestations of HIV Infection; Chapter 112, Infectious Complications of HIV Infection). Inflammatory pseudotumors in children can produce nonspecific symptoms and signs, including prolonged fever, malaise, weight loss, anemia, and elevated erythrocyte sedimentation rate (ESR).46–48 Pseudotumors occur most commonly in the lung in adults but occurred intra-abdominally in 8 children reported with FUO presentation.46 Use of CT, which reveals the mass, undoubtedly has increased identification. Histology shows mixed inflammatory response and fibrosis. The cause is unknown but many probably represent a hostcontrolled infection. If more than a single site is identified, recurrence following resection may be more common and differentiation from malignancy can be difficult.48 Other conditions of unknown etiology, such as Kikuchi disease (histiocytic necrotizing lymphadenitis)49 and cholesterol granuloma,50 can manifest as FUO, as can granulomatous hepatitis.

Kawasaki Disease Incomplete Kawasaki disease, lacking classic features or manifesting with asynchronous or delayed features, can lead to FUO. Incomplete Kawasaki disease should be considered if findings incompatible with Kawasaki disease (e.g., generalized lymphadenopathy, pneumonia) are absent and ESR is > 40 mm/h or CRP is > 3.0 mg/dL. An echocardiogram should be performed in such children; FUO presentation of Kawasaki disease is associated with increased risk of coronary artery damage (see Chapter 199, Kawasaki Disease).51,52

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Prolonged, Recurrent, and Periodic Fever Syndromes

CHAPTER

17

131

Box 17-3 summarizes the noninfectious causes of FUO.

suggestive of the diagnosis, and the presence of the autoantibody classic form of antineutrophil cytoplasmic antibody (C-ANCA) is a sensitive and specific laboratory finding.

Autoimmune and Autoinflammatory Conditions

Sarcoidosis

Autoimmune and autoinflammatory conditions are the second most common cause of FUO in children, accounting for approximately 15% of diagnoses in the published pediatric series, with a higher percentage in children referred for FUO to pediatric rheumatologists.13 Systemiconset juvenile idiopathic arthritis (SoJIA, Still disease) is diagnosed in three-quarters of such children in older series. Associated features include hepatosplenomegaly, adenopathy, leukocytosis with a left shift, elevated ESR, and microcytic anemia. Still disease may be an autoinflammatory rather than an autoimmune disease. There is controversy whether it should be classified as a presentation of JIA.53 Systemic lupus erythematosus (SLE) was diagnosed in < 5% of reported children with FUO, but in approximately 20% of children who had FUO due to collagen vascular disease. Multisystem symptomatology is typical, especially with involvement of kidneys, joints, and skin. Occasionally, patients have generalized symptoms, with malaise, fever, and weight loss for weeks before more distinctive symptoms arise. Testing for high-titer antinuclear antibody (positive in > 90% of patients with SLE), anti-double-stranded DNA and antismooth-muscle antibody is usually diagnostic. Acute rheumatic fever is currently a rare cause of FUO in children. Occasionally, fever can predate the other signs and symptoms necessary for specific diagnosis. Repeated examinations are helpful, because mitral valve regurgitant murmur is often the first specific clinical feature. Elevated antistreptococcal antibody values or isolation of group A streptococcus from the throat culture supports the diagnosis. Virus-associated and familial hemophagocytic lymphohistiocytosis as well as macrophage activation syndrome have fever as a prominent symptom and evolve over weeks before diagnoses are evident. Natural killer cell and cytotoxic CD8+ T-lymphocyte dysfunction may underlie these dysregulatory conditions.54 Severity of illness, organomegaly, lymphadenopathy, rashes, cytopenias, and extreme elevations of acutephase reactants and serum ferritin levels are clues (see Chapter 14, Hemophagocytic Lymphohistiocytosis and Macrophage Activation Syndrome).

Sarcoidosis in children is uncommon and predominantly is restricted geographically to North and South Carolina, Virginia, and Arkansas, and is more common in African Americans. Features include fever of > 2 weeks’ duration, fatigue, leg pain, weight loss, anemia, increased ESR, and elevated serum immunoglobulin levels and angiotensinconverting enzyme values; negative antinuclear antibody and rheumatoid factor test results are typical. In one study, magnetic resonance imaging of 3 children with normal nuclear bone scans showed multifocal nodular lesions within the bone marrow of the lower extremities.55 Bone biopsy revealed noncaseating epithelioid granuloma. The cause of sarcoidosis is unknown.

Noninfectious Causes

Vasculitis Syndromes Several vasculitis syndromes can occasionally cause FUO. Presentation varies according to the specific entity. Hypertension, renal involvement, and mononeuritis multiplex are associated with polyarteritis nodosa. Aphthous stomatitis, uveitis, central nervous system involvement, and gastrointestinal symptoms, including perianal ulcers, are hallmarks of autoinflammatory Behçet disease. Wegener granulomatosis infrequently manifests as prolonged fever; the triad of renal, upper respiratory tract, and lower respiratory tract symptoms is

BOX 17-3. Noninfectious Causes of Fever of Unknown Origin • • • • • • • • • • • • •

Autoimmune diseases Autoinflammatory disorders Inflammatory bowel disease Malignancy Drugs, other medicinal and nutritional products Munchausen syndrome by proxy Dysautonomia Central thermoregulatory disorder Diabetes insipidus Anhidrotic ectodermal dysplasia Hyperthyroidism Hematoma in a closed space Pulmonary embolus

Inflammatory Bowel Disease Inflammatory bowel disease is an uncommon but consistent cause of FUO in children. Its relative importance appears to be growing, possibly as advances in diagnostic testing more easily eliminate other causes. Crohn disease accounts for 90% of cases, because the more recognizable signs of ulcerative colitis permit its earlier diagnosis. Linear growth failure is a clue to the diagnosis, which, like fever, can antedate gastrointestinal symptoms by months to years. Besides subtle gastrointestinal symptoms or tenderness, important clues are finger-clubbing, arthritis, stomatitis, perirectal skin tags and fistulas, pyoderma gangrenosum, heme-positive stools, iron-deficiency anemia, and elevated ESR.

Malignancy Malignancy is diagnosed in at least 10% of cases of FUO; relative importance appears to be growing. Lymphoreticular malignancy (especially leukemia and lymphoma) is most likely, accounting for approximately 80% of all malignancies in reported series. Because of musculoskeletal symptoms, JIA can be the anticipated diagnosis before bone marrow is examined. Nonarticular bone pain and evolving hematologic abnormalities are clues to malignancy.56 Hepatoma, neuroblastoma, and various soft-tissue sarcomas account for most other malignancies causing prolonged fever.57 In a study to assess the value of routine bone marrow biopsy in children, Hayani and colleagues58 concluded that the practice is only helpful if clinical or laboratory evidence suggests malignancy or aberration of hematopoiesis; culture of bone marrow for infectious agents also appears to be of value, predominantly in immunocompromised hosts.

Drug Fever Drug fever is rarely named as a cause of FUO in children in reported series (1.5%) but is a common cause of prolonged fever in clinical practice. Typical onset of fever is 7 to 10 days after commencement of therapy, but the onset is variable and can occur years into therapy. A preserved sense of wellbeing despite hectic fever (sometimes with rigors) and the presence of symmetric morbilliform rash or eosinophilia in approximately 20% of cases support the diagnosis. The diagnosis is established when: (1) resolution of the fever occurs within 48 hours, or 3 to 5 drug half-lives, whichever is longer, of discontinuation of the drug; and (2) recrudescent fever occurs within a few hours after the drug is restarted. Fever associated with drugrelated serum sickness-like reactions, erythema multiforme, or desquamating exfoliative dermatitis resolves more slowly after discontinuation of the offending agent. Any drug can cause fever, especially those used to mediate neurologic disorders, cardiovascular conditions, and neoplasia.59 Nonprescription products, aspirin, nonsteroidal anti-inflammatory

132

SECTION

B Cardinal Symptom Complexes

agents, and nutritional supplements can be associated with drug fever. Among antimicrobial agents, vancomycin, b-lactam agents, isoniazid, and sulfa drugs are most likely to cause fever, whereas aminoglycosidic agents and macrolides are least likely.

Factitious Fever, Munchausen Syndrome by Proxy Factitious fever can rarely be perpetrated by the patient, but more frequently is a manifestation of Munchausen syndrome by proxy.60 Temperature elevations are discordant with the patient’s normal appearance, physical examination, vital signs, and absence of laboratory signs of inflammatory response. Validation of parentreported temperature elevation, or performance of concurrent urinary temperature (such as on a specimen allegedly obtained for urinalysis) as the parent obtains the core temperature can uncover this diagnosis.

Noninflammatory Conditions Dysautonomia, anhidrotic ectodermal dysplasia, and dehydration are occasional causes of recurring fevers in which fever is not mediated through the acute-phase response and thus there is no sweating.1 Fever due to hypothalamic dysfunction is invariably seen in the setting of profound injury to the central nervous system, third ventricular tumor, or more widespread thalamic failure with loss of appetite and control of thirst.61 Diabetes insipidus, due to either a renal defect or to central dysfunction, can manifest as FUO. Dehydration without apparent cause, polyuria, polydipsia, and hypernatremia in the absence of sweating, glucosuria, or elevated acute-phase reactant values are clues to the diagnosis. Periods of fever as well as hypothermia have been described in association with absence of the corpus callosum.62 Fever in these conditions is not mediated via the acute-phase response. Normal ESR, C-reactive protein, and failure to respond to nonsteroidal anti-inflammatory agents (which block completion of the acute-phase response) aid in these diagnoses.

PROLONGED INSIGNIFICANT FEVER, PROLONGED ILLNESS WITH RESOLVED FEVER, FATIGUE OF DECONDITIONING One of the most frequent referrals to pediatric infectious disease subspecialists for “prolonged fever” is an adolescent with low-grade or falsely perceived fever, or whose fever has resolved who generally feels unwell and is unable to attend school and social activities. Such patients require the same disciplined performance of history and examination as those with true FUO. For the subspecialist to offer a deÀnitive opinion, the family must perceive that a thorough and thoughtful consultation has occurred. All laboratory test results, actual imaging studies, and biopsy slides should be reviewed. The Àndings listed in Box 17-4 and evaluation of complaints and review of systems in Table 17-3 taken together suggest that no cryptogenic infection or serious medical condition is present.2 The discordance of lengthy list of complaints with the normal Àndings of examination and laboratory tests is typical. A family modus operandi centers around the patient despite lack of objective abnormalities. Prodded recitation of a typical 24-hour period of activity reveals a feeling of tiredness but no prolongation of actual time asleep. There is frequently a parental model of chronic illness, or a recent change in extended-family dynamics (e.g., divorce, serious medical diagnosis, death). Deconditioning (i.e., diminution of physical strength, stamina, and vitality), loss of self-esteem, fear of failure to perform at previous expectation, and secondary gain may all play into the clinical state. The patient should be queried privately about potential abuse and other insights. Depression should be considered. Clues in depressed individuals compared with fatigued individuals include for the depressed state a feeling of sadness all the time, lack of self-worth, dislike of or wanting to hurt oneself, anger, marginalization, and feeling of being disliked by others.63Fatigued individuals have increased somatic complaints and tendency to

BOX 17-4. Typical Findings in Patients with Fatigue of Deconditioning • • • • • • • • • • • • •

Age > 12 years Preillness achievement high Family expectations high Acute febrile illness with onset easily dated Family and outside attention high Multiple but vague complaints Odd complaints (e.g., 10-second “shooting” pains at multiple sites; 30second “blindness”; stereotypic sporadic, brief unilateral tremors, jerks, or “paralysis” lasting < 1 minute) Tiredness, but no daytime sleep (or reversal of daytime and nighttime sleep) Model of chronic illness in family, recent loss of important person, change in family dynamic Unusual cooperation and interest during interview and examination (or unusual fearfulness and dependency on parent) Preserved or increased weight Normal physical and neurologic examination Normal results of screening laboratory tests

TABLE 17-3. Frequent Complaints and Reassuring Findings on Specific Questioning in Patients with Fatigue of Deconditioning Complaint

Further Findings

Sore throat

Prominent morning only; not worse when swallowing

Respiratory complaints

“Stuffy,” “big head”; no purulence; no cough

Weakness

Not quantiÀable or determinable as proximal, distal, or truncal; slumped posture but able to hop, squeeze, resist force when asked

Fatigue, tiredness

“Exhausted” but asleep < 12 hours per night, not asleep during day; or reversal of day and night sleep

Dizziness, faintness

“Foggy,” “dazed”; no loss of stature; no maneuvers precipitate or used to resolve episode

Poor color

“Pale,” “pasty”; dependent veins prominent, hands cool, and nails blue

Poor appetite

“Not hungry,” “stomach feels weird”; no pain, vomiting, diarrhea

Skin rashes

Fleeting blotches; extremities, insigniÀcant papules, mottled dependent

Heart palpitations

“Heart jumps”; no predictability, loss of stature, or chest pain; lasts < 30 seconds

internalize stress.64 In the last decade, neurally mediated hypotension and postural tachycardia have been postulated as mechanisms of persistent fatigue. Two studies in adults published in 2005 suggest that orthostatic instability is a common Ànding in healthy as well as in fatigued individuals,65 and that among military recruits with orthostatic intolerance, endurance exercise training signiÀcantly improved the physiologic Àndings and symptoms66 (see Chapter 200, Chronic Fatigue Syndrome). Management of patients with fatigue of deconditioning begins with a precise review of Àndings and conclusions of what the diagnosis is and is not.2 Management is outlined in Box 17-5. It is a disservice to diagnose “chronic fatigue syndrome,” support home tutoring, or refer the patient to additional subspecialists for minor Àndings. The primary care pediatrician should set the tempo for reconditioning and monitoring.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Prolonged, Recurrent, and Periodic Fever Syndromes

RECURRENT, PERIODIC, AND HEREDITARY PERIODIC FEVER SYNDROMES Recurrent Fever When history is carefully obtained, most children who are referred to subspecialists for recurrent fever have multiple self-limited infections. Children can normally have up to 10 self-limited viral diseases per year for the first 2 to 3 years of life. Children attending out-of-home childcare can have more, and the history accrues of “too many” febrile illnesses, especially in acute otitis media-prone children. A defect in host immune response or cryptogenic infection should not be pursued if illnesses are self-limited (or otitis media responds promptly to antibiotic therapy), if they predominantly involve the respiratory and gastrointestinal tract, there is no history of serious or unusual infection, illnesses are separated by times of wellness, and growth is unimpeded. Relapsing infections is an unusual cause of recurring fevers.67 Fever, in the absence of respiratory tract symptoms, which abates repeatedly during antibiotic therapy can occasionally be due to endocarditis or renal infection in predisposed hosts. Borrelia recurrentis is a rare cause of tickborne relapsing fever distinguished by history of travel through forested areas (see Chapter 185, Borrelia burgdorferi [Lyme disease]).

Periodic Fever Syndromes Periodic fever is defined as recurrent episodes of illness in which fever is the cardinal feature and is associated with a predictable and similar set of symptoms that last days to weeks and recur at regular intervals. Each episode is separated by symptom-free periods, ranging from weeks to months. In some instances, the episodes have consistent, clockwork periodicity, whereas in others, they do not. Differentiating features of periodic fever syndromes likely encountered by an infectious diseases subspecialist are shown in Table 17-4. Periodic fever, aphthous stomatitis, pharyngitis, and cervical adenopathy (PFAPA) is the most common disorder with periodic fever; it has no known pattern of inheritance, no defined etiology, and no confirmatory laboratory tests are available (see next section). Cyclic neutropenia is an inherited defect of neutrophil elastase.68–70 Periodonitis and cellulitis are common infectious complications, and Clostridium septicum septicemia is a unique catastrophic complication.71 Treat-

BOX 17-5. Management of the Patient with Fatigue of Deconditioning PATIENT • Validate symptoms as accurate • Explain that there is no serious disease • Prescribe incremental, forced return to school • Prescribe exercise • Predict increased symptoms in the process of reconditioning • Disallow school absence without examination that day by pediatrician • Reassure that there will be no expectation of achievement from performance FAMILY • Validate their concern • Validate that symptoms could have been clues to serious conditions • Validate their medical pursuit • Justify your diagnosis with facts of organ-based findings in history, examination, and screening laboratory test • Advise change of family focus from illness to health • Advise family to avoid asking, “How do you feel?” • Elicit promise to patient of no expectation of achievement from performance PRIMARY PHYSICIAN • Will set pace of reconditioning and monitoring • Will continue to perform examination to assure normal findings

CHAPTER

17

133

ment with granulocyte colony-stimulating factor greatly reduces infectious complications. The hereditary periodic fever syndromes, including familial Mediterranean fever,72–75 the hyperimmunoglobulinemia D with periodic fever syndrome (HIDS),76–80 and tumor necrosis factor receptor-associated periodic syndromes (TRAPS)81–84 are termed “autoinflammatory” diseases. They are related to defects in gene families and pathways in the regulation of innate immunity, which distinguishes them from autoimmune disorders such as SLE and rheumatoid arthritis, which reflect abnormalities in the adaptive immune system and are characterized by high-titer autoantibodies or antigen-specific T lymphocytes. Three other hereditary autoinflammatory disorders caused by mutations in C1AS1, which encodes a protein denoted cryopyrin or NALP3, are familial cold autoinflammatory syndrome, Muckle–Wells syndrome, and neonatal-onset multisystem inflammatory disease/chronic infantile neurologic cutaneous and articular syndrome. With C1AS1 genetic defects, recurrent urticaria, arthritis, and multiorgan dysfunction are prominent features that overshadow fever. Excellent recent reviews of the hereditary periodic fever and other autoinflammatory syndromes have been published.84–87 With genetic testing becoming available (http://www.genedx.com), disorders previously believed to be rare are being diagnosed increasingly, and spectra of genetic and clinical findings are broadening.

Periodic Fever, Aphthous Stomatitis, Pharyngitis, and Adenitis In 1987, a chronic syndrome was described in 12 children characterized by periodic episodes of high fever lasting 3 to 6 days and recurring every 3 to 8 weeks, accompanied by aphthous stomatitis, pharyngitis, and cervical adenitis.88 Additional reports followed.89,90 In 1989, the acronym PFAPA (for periodic fever, aphthous stomatitis, pharyngitis, and [cervical] adenopathy) was coined to describe the entity.91 Additional reports of more than 100 cases have appeared.92–95 Cardinal features are shown in Table 17-4. A registry initiated at Vanderbilt University provides some aggregate data over 10 years on cases of PFAPA at presentation and follow-up.94 Table 17-5 shows characteristics of cases at diagnosis. Additional features have been noted.94–96 Few patients of African American or Hispanic heritage have been described. Clockwork cycle and abrupt onset after 1 to 2 hours of “glassy eyes” and clinging behavior followed by high fever > 39°C (which persists at that height for the duration of the fever) are cardinal features, overshadowing others. Congestion, conjunctivitis, cough, and wheezing are distinctly absent. Mouth ulcers are solitary or scattered, sometimes not mentioned until queried, and are not thought by parents to be the cause of the child’s poor appetite. Lymph nodes are not tender or erythematous, and are not usually prominent. Tonsils are of normal size or modestly enlarged and without exudates. Children are completely well between episodes. Parents consider the child “healthy” and energetic (between episodes), and subject to fewer childhood illnesses compared with siblings. They do not have invasive pyogenic infections. Family members are not ill subsequent to the patient’s febrile episode. There is sometimes a parental history of similar unexplained fevers in childhood. Growth and development are not affected. Pursuit of infectious agents by culture, antigen, or antibiotic is negative. Hemogloblin, urinalysis, and serum hepatic enzymes, albumin and immunoglobulin concentrations are normal. PFAPA is a common cause of strictly defined periodic fever in children. The etiology is unknown. It has some features of an infectious disease and some of an autoinflammatory disorder.96 In single case reports a child with Epstein–Barr virus infection and a child with Mycobacterium chelonae infection had typical PFAPA.97,98 Genetic analyses of patients diagnosed with PFAPA have occasionally, but not commonly, identified hereditary autoinflammatory disorders, leading one to conclude that most cases of PFAPA are not due to the known underlying defects of innate immunity.99 Original symptoms in

134

SECTION

B Cardinal Symptom Complexes

TABLE 17-4. Differentiating Features of Periodic Fever Syndromes PFAPA

Cyclic Neutropenia

Onset < 5 years

Expected

Length of fever episode

4 days

Familial Mediterranean Fever

HIDS

TRAPS

Usual; often < 1 year old Common; peak onset middle of first decade

Expected; often < 1 year old

Variable

5–7 days

1–3 days

3–7 days

2 days–weeks

Periodicity of episodes q3–6 weeks; typically q28 days

q2–8 weeks; q21 days in > 90%

Irregular intervals: weekly, q3–4 months or less often

q4–8 weeks or irregular

Irregular intervals; varies weeks to years

Associated symptoms/ Pharyngitis 65–70%; signs aphthous stomatitis 65–70%; cervical adenopathy 75–85%

Ulcers, lingering gingivitis and periodontitis; recurrent otitis media and sinusitis; rare peritonitis; rare gramnegative bacillary or Clostridium septicum septicemia

Severe abdominal pain; erysipelas-like rash; scrotal pain and swelling; polyserositis

Abdominal pain, diarrhea in young; headache; arthralgia; diffuse maculopapular rashes; aphthous ulcers; splenomegaly; mood swings; immunizations trigger

Abdominal pain; migratory pseudocellulitis and myalgia; periorbital edema; scrotal pain; polyserositis

Ethnicity/geography

None; rare in siblings

No ethnicity

Most common among Jewish, Armenian, Turkish, Arab, Italian, but cases from worldwide ancestry

Predominantly northern European ancestry

Variable ancestry

Inheritance

None; parent may have Autosomal dominant history of excessive high fevers as child

Autosomal recessive

Autosomal recessive

Autosomal dominant

Laboratory findings

Mild neutrophilia; ESR elevated < 60 mm/ hour during episode only

Absolute neutrophil count < 200 cells/mm3 for 3–5 days

Elevated acute-phase reactants

Elevated acute-phase reactants; variable Œ serum cholesterol; variable Ø IgA and IgD (> 100 IU/mL) or may not be Ø (especially < 3 years)

Elevated acute-phase reactants

Etiology/diagnosis

Unknown; clinical diagnosis

Chromosome 19p; ELA2 mutations leading to mutant neutrophil elastase; apoptosis marrow myeloid cells

Chromosome 16p; MEFV missense mutations; PYRIN domain; dysregulation inflammation and apoptosis

Chromosome 12q; MVK mutations leading to Πmevalonate kinase and isoprenoids; mevalonic aciduria during attacks

Chromosome 12p; TNFRSF1A mutations; ŒTNFR-1; complex TNF pathophysiology in binding, intracellular trafficking and leukocyte apoptosis

Treatment

None established (see text)

Recombinant G-CSF; aggressive periodontal care; aggressive treatment suspected septicemia

Colchicine

Simvastatin (investigational); etanercept (investigational); allogeneic bone marrow transplant (single case)103

Corticosteroid; etanercept (investigational)

Sequelae

None established

Dental problems; infection

Amyloidosis

Amyloidosis case reports

Amyloidosis 10%

ESR, erythrocyte sedimentation rate; G-CSF, granulocyte colony-stimulating factor; HIV, human immunodeficiency virus; Ig, immunoglobulin; PFAPA, periodic fever, aphthous stomatitis, pharyngitis, and cervical adenopathy; TNF, tumor necrosis factor; TRAPS, tumor necrosis factor receptor-associated periodic syndromes. Modified from Long.2

1 patient confirmed to have TRAPS lacked clockwork periodic fever when “PFAPA” was diagnosed and the episodes sometimes lasted several weeks.100 Another adopted child confirmed to have HIDS had onset of clockwork periodic fever before 1 year of age, had normal immunoglobulin D levels in preschool years, but had developed increasingly belligerent behavior during episodes, which is not expected in PFAPA.101 The diagnosis of PFAPA is a working diagnosis based on a typical constellation. History of episodes and the diagnosis should be reassessed over time. Patients with “extra” symptoms or unusual immunoglobulin levels should be pursued further.

The treatment of PFAPA has been the subject of only one randomized clinical trial;102 case reports and retrospective reviews have suggested several beneficial treatments.92,93–97 Patients with PFAPA given corticosteroids (total of 1 to 2 mg/kg prednisone or prednisolone as one or two doses) have a dramatic resolution of symptoms. Response does not confirm the diagnosis, but failure to respond makes PFAPA unlikely. Some families report that the cycles of fever became more closely spaced after successful treatment with corticosteroids. Cimetidine used for either treatment or prophylaxis was judged by parents to be “somewhat-to-very effective” in 43% of

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Lymphatic System and Generalized Lymphadenopathy

TABLE 17-5. Characteristics of Cases at Diagnosis of Periodic Fever, Aphthous Stomatitis, Pharyngitis, and Cervical Adenopathy (PFAPA) Demographic and Clinical Features Male: Female

1.2

Onset of PFAPA

2.8 years (2.4–3.3 years)b

Maximum temperatures

40.5°C (40.4–40.6°C)

Duration of fever

3.8 days (3.5–4.1 days)

Duration of episode

4.8 days (4.5–5.1 days)

Episodes/ year

11.5 (10.5–12.5)

Wellness interval

28.2 days (26.0–30.4 days)

Mean white blood cell count at episode

13,000/mm3 (range 2–37 μ 103)

Percent with Signs and Symptoms at Episodesa Fever + >1 cardinal feature

97%

Cervical lymphadenopathy (77%)

CHAPTER

CHAPTER

18

135

18

Lymphatic System and Generalized Lymphadenopathy Mary Anne Jackson and P. Joan Chesney

ANATOMY AND FUNCTION OF LYMPHOID TISSUE

Aphthous stomatitis (67%) Pharyngitis (65%) Chills

80%

Headache

65%

Nausea

52%

Abdominal pain

45%

Diarrhea

30%

Coryza

18%

Rash

15%

Mean 41 (range 5– erythrocyte 190 mm/hour) sedimentation rate at episode Modified from Thomas et al.94 a Although 94 children were included in the original registry, parents did not always answer all queries (frequently citing inability to assess because of age). b Mean (95% confidence interval).

cases in the registry,94 and has been reported to induce total remission in nearly 25%.92,94 Tonsillectomy, performed in 47 of the 94 children,94 was judged to be somewhat-to-very effective in 86%. In a randomized controlled trial of 26 children in Finland, with PFAPA, all randomized to tonsillectomy were symptom-free at 6 months’ follow up compared with 6 of 12 randomized to no surgery.102 The duration of PFAPA is variable. Spontaneous resolution without sequelae is the rule. In one cohort study with mean follow-up of 3.3 years (range < 1 month to 9.4 years), mean duration of illness before resolution was 4.5 years. Constellation of symptoms and length of fever episodes that occurred during follow-up were identical to those at diagnosis; however, symptomfree interval was significantly longer (28.2 days; 95% confidence interval 26.0 to 30.4).94 Lengthening interval frequently heralded resolution of PFAPA.

The lymphoid system is composed of an extensive capillary network that collects lymph into an elaborate system of collecting vessels. These collecting vessels merge to empty lymph into the bloodstream by way of the thoracic duct, at its entry into the left subclavian vein, or the right lymphatic duct, which empties into the right subclavian vein. Interspersed along the collecting vessels are specialized lymphatic structures, including the tonsillar tissues of the Waldeyer ring, mucosa-associated lymphoid nodules, spleen, thymus, and lymph nodes (Table 18-1). The Waldeyer ring of lymphoid tissue that surrounds the oropharyngeal isthmus and the opening of the nasopharynx into the oropharynx is uniquely positioned to interact with foreign material entering the nose or mouth. The ring is formed superiorly by the midline pharyngeal tonsil, which is located in the roof of the nasopharynx (adenoid), and inferiorly by the lingual tonsil in the posterior third of the tongue. On either side of the pharynx, the lateral pharyngeal bands of lymphoid tissue connect the adenoid to the tubal tonsils of Gerlach at the openings of the eustachian tubes and to the faucial (palatine) tonsils. Smaller aggregates of lymphoid tissue in this area include the posterior pharyngeal granulations and the lymphoid tissue within the laryngeal ventricle. Small submucosal lymphoid nodules, located throughout the respiratory, gastrointestinal, and genitourinary tracts, are composed of phagocytic and lymphoid cell collections without a connective tissue capsule. These nodules are ideally situated to respond to mucosal antigens. The thymus, which is located over the superior vena cava in the anterior mediastinum, is relatively protected from antigens. Surrounded by a thin connective tissue capsule, the thymus is uniquely composed of both epithelial and lymphatic elements. The spleen is the largest lymphatic organ in the body and the only lymphatic tissue specialized to filter blood. Similar to the lymph nodes, the spleen is a component of the peripheral lymphoid system and is composed of red pulp (red blood cells) and the interior white pulp, which contains lymphoid nodules with germinal follicles. Lymph nodes are normally small oval or bean-shaped bodies that are strategically located along the course of lymphatic vessels to filter lymph on its way to the bloodstream. Lymphatic vessels enter around the periphery of the nodes. Lymph filters through the cortex to the medulla of the node and exits through the hilum. Blood vessels enter and leave through the hilum, connected to capillaries that course through the node. During this process, lymphocytes can leave the blood and re-enter the lymphatic circulation. Nodes are densely packed with lymphocytes that are loosely organized into cortical nodules and medullary cords by connective tissue trabeculae and lymphatic sinuses. The juxtaposition of phagocytic cells, antigen-processing cells, and lymphocytes in an area of sluggish blood flow is ideally suited to provide the first line of defense against pathogens. As lymph slowly

136

SECTION

B Cardinal Symptom Complexes

TABLE 18-1. Anatomic Types and Locations of Lymphoid Tissue

TABLE 18-2. Frequency and Location of Palpable Peripheral Lymph Nodes in Healthy Children

Type of Tissue

Location

Distinguishing Features

Discrete lymph Occipital, preauricular, node groups postauricular, submandibular, submental, facial, parotid, cervical, supraclavicular, para-aortic, axillary, epitrochlear, inguinal, iliac, popliteal, mediastinal, hilar, pelvic, mesenteric, celiac

Nodes have discrete capsules; afferent lymphatic flow enters from periphery; efferent lymphatic flow exits and blood vessels enter and exit through hilum of node

Waldeyer ring

Aggregates of lymphoid nodules are partially encapsulated; there are no afferent lymphatics and no lymphoid sinuses; efferent lymphatic flow is not as structured as for lymph nodes

Pharyngeal (adenoid), palatine (faucial), lingual and tubal (Gerlach) tonsils; lateral pharyngeal bands; posterior pharyngeal granulations

Lymphoid nodules

Small, submucosal lymphoid collections throughout the intestinal (Peyer patches), respiratory, and genitourinary tracts (mucosal-associated lymphoid tissue, or MALT)

Lymphatic flow is not encapsulated or structured; tissue responds to mucosal antigens with phagocytosis and immunoglobulin A production

Thymus

Anterior mediastinum

Organ is composed of lymphoid and epithelial cells; no afferent lymphatic vessels; protected from antigen; essential for development and maturation of peripheral lymphoid tissues

Spleen

Abdomen

Lymphatic tissue is uniquely specialized to filter blood; largest lymphatic organ in the body; no afferent lymphatic vessels and no lymphatic vessels within spleen; sinusoid structure is similar to lymph node but lymph empties into splenic vein

filters through the rich reticular network, organisms are trapped and can be ingested by phagocytic cells, thus stimulating cytokine release, which in turn recruits lymphocytes into immunologic responses. The lymph node groups in the body can be divided into the superficial and peripheral nodes, which are generally easily palpable, and the deeper groups adjacent to major vessels and viscera (see Table 18-1).

DEVELOPMENTAL CHANGES Lymphoid tissue, including the thymus, forms a significantly larger percentage of total bodyweight in infants and children than in adults. Considerable lymphoid activity is present at birth, and continuing exposure to environmental antigens results in an increase in lymphoid mass that reaches a peak between ages 8 and 12. Atrophy of lymphoid tissue begins during adolescence. In children, the thymus can weigh 40 g; in adults, it is replaced by fibrous and fatty tissue and may weigh only 10 g. By adult standards, almost all children have “lymph-

Palpable Node

Neonatea

Age < 2 yearsb

Age > 2 yearsb

Cervical

+

++

++

Postauricular

–

+

–

Occipital

–

++

+

Submandibular

–

+

++

Supraclavicular

–

–

–

Axillary

+

+++

+++

Epitrochlear

–

–

–

Inguinal

+

+++

+++

Popliteal

–

–

–

None

++

++

++

+++, normally present in > 50% of children; ++, normally present in 25% to 50%; +, normally present in 5% to 25%; –, normally present in < 5%. a Data from Bamji M, Stone RK, Kaul A, et al. Palpable lymph nodes in healthy newborns and infants. Pediatrics 1986;78:573. b Data from Herzog LW. Prevalence of lymphadenopathy of the head and neck in infants and children. Clin Pediatr 1983;22:485.

adenopathy,” because palpable nodes, particularly in the cervical, axillary, and inguinal areas, are common in children of all ages, including neonates (Table 18-2). Prominent palatine tonsils are common in preadolescent children, whereas in infants less than 1 year old and adults, the palatine tonsils are not normally visible. Aside from, or because of, their normally hyperplastic nodes, the response of children to antigenic, infectious, or neoplastic stimuli is much more rapid, prolific, and exaggerated than in adults. Lymph nodes can increase in size as much as 15-fold within 5 to 10 days of antigen exposure. Lymphadenopathy can be particularly pronounced in mediastinal and mesenteric nodes.

CHARACTERISTICS OF LYMPHADENOPATHY Lymphadenopathy, or enlarged lymph nodes, can be characterized by size, location, consistency, rate of growth, tissue inflammation, and fixation. In all ages and in all lymph node groups, a node is considered enlarged if it measures more than 10 mm in its longest diameter. There are two exceptions to this rule. In the epitrochlear region, nodes larger than 5 mm in diameter are abnormal, and in the inguinal region, only nodes larger than 15 mm are abnormal. In healthy neonates, lymph nodes ranging from 3 to 12 mm in diameter can be found in the cervical, axillary, and inguinal regions.1 Most children examined when healthy have palpable lymph nodes.2 In the child less than 2 years old, palpable nodes may be present at any peripheral location, except in the epitrochlear, supraclavicular, and popliteal areas, where palpable lymph nodes are always abnormal. In the child older than 2 years palpable lymph nodes in these areas and in the posterior auricular and suboccipital areas are considered abnormal. Characteristics of lymph nodes are determined by palpation (Table 18-3). Soft, discrete, nontender, small (< 2 cm) nodes that are found bilaterally or generally with no periadenitis, cellulitis, or abscess formation usually result from hyperplasia secondary to viral infection. Unilateral, large (> 2 cm), warm, tender, poorly defined nodes with surrounding edema, erythema, or abscess formation are usually infected with pyogenic bacteria. Moderately large, unilateral nodes, with discrete margins and minimal inflammation, that slowly progress to become erythematous (but not warm), fluctuant, and adherent to overlying skin are characteristic of chronic, usually bacterial, infections.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Lymphatic System and Generalized Lymphadenopathy

TABLE 18-3. General Characteristics of Enlarged Lymph Nodes

CHAPTER

18

137

BOX 18-1. Mechanisms of Lymph Node Enlargement

According to Cause Causes of Enlargement Characteristic

Acute Bacterial

Chronic Bacterial

Acute Viral

Malignant

Large size

+++

+++

+

++/+++

Erythema

+++

++

–

–

Tenderness

+++

++

+

++

Consistency

Soft/firm

Firm

Soft

Firm/rubbery

Discrete

++

+++

+++

+++

Matted

++

++

–

++

Fixed

+++

+

–

–

Fluctuant

+++

+++

–

–

Associated with cellulitis

+++

+

–

–

Unilateral

+++

+++

–

+

+++, common finding; ++, less common; +, occasional; –, rare.

Enlarged nodes resulting from lymphoma are generally firm, discrete, freely movable, nontender, and rubbery and have no surrounding inflammation. They increase in size over time and adjacent nodes may become matted together and lose their individual character. Suppuration and fixation to skin or deeper structures, as seen in inflammatory adenopathy, are not expected in lymphoma. Lymphomatous nodes can wax and wane in size over weeks to months, and lymphomatous changes can develop (or become more apparent) in nodes for which a biopsy specimen initially showed only hyperplasia.1–3 Enlarged nodes that result from metastatic tumors are described as being hard and bound to each other and to surrounding tissues. Enlarged nodes may be single or multiple and contiguous when they are located in one region, in which the condition is referred to as regional lymphadenopathy. Enlargement of a single node or regional nodes can be the result of localized disease or the first manifestation of generalized lymphadenopathy. Generalized lymphadenopathy is defined as the simultaneous presence of two or more enlarged nodes in noncontiguous groups; enlargement need not be present in every group of lymph nodes. Thus, the simultaneous presence of mesenteric and hilar adenopathy is considered generalized lymphadenopathy. Splenomegaly can also be present but is not included in the definition.

PATHOGENESIS OF LYMPHADENOPATHY AND LYMPHADENITIS Microorganisms reach lymph nodes directly by lymphatic flow from the inoculation site or by lymphatic spread from adjacent nodes. If initial involvement of regional nodes does not adequately contain infection, organisms may reach noncontiguous nodes by hematogenous spread. Lymph node enlargement can result from a variety of mechanisms (Box 18-1). In acute pyogenic lymphadenitis, the initial inflammatory response in the node, including complement activation and cytokine release, causes recruitment of neutrophils and mononuclear phagocytes. Vascular engorgement and intranodal edema, as well as cellular replication in response to the antigenic stimulus, lead to rapid enlargement of the node. Involvement of adjacent lymph nodes and surrounding soft tissues, including skin, may result in cellulitis, suppuration, necrosis, and fixation to adjacent tissues. Once purulence occurs, lymph node architecture (and antibiotic access) is destroyed. Healing occurs by fibrosis.

• Cells within the node replicate in response to antigenic stimuli or as a result of malignant transformation • Cells exogenous to the node, such as neutrophils or metastatic neoplastic cells, enter the node in large numbers • Foreign material is deposited within histiocytic cells of the node (e.g., lipid storage diseases) • Local cytokine release leads to vascular engorgement and edema • Tissue necrosis leads to suppuration

For microorganisms that cause subacute or chronic granulomatous changes, the increase in node size and tenderness and adjacent inflammatory response are usually less impressive. Formation of granulomas and, in some cases, caseating necrosis also destroys nodal architecture; drainage may be required to relieve pain or hasten resolution. With local or generalized viral infections, the nodal response is primarily one of hyperplasia without necrosis; this resolves without sequelae as infection abates. Enlarged lymph nodes, lymphoid nodules, or lymphoid tissue, such as the Waldeyer ring, can result in obstruction, compression, or erosion of important structures, rupture of nodal contents, or inflammation of adjacent structures. Generalized hyperplasia of the Waldeyer ring can result in obstruction of the posterior nares, eustachian tube, or oropharynx. Extensive mediastinal lymphadenopathy can lead to obstruction or erosion of the airways, esophagus, superior vena cava, recurrent laryngeal nerve, or lymphatics (see Chapter 20, Mediastinal and Hilar Lymphadenopathy, Table 20-2). Inflammation of Peyer patches (e.g., Salmonella typhi infection) can lead to intestinal perforation and hemorrhage or may serve as lead point for intussusception. Mesenteric infectious lymphadenitis can lead to an intra-abdominal abscess. Perinodal inflammation of muscles in the neck results in torticollis, and spread of infection in the neck can lead to deep fascial space infections.

HISTOPATHOLOGY OF LYMPHADENITIS Nonspecific hyperplasia is the most common histopathologic finding in enlarged lymph node biopsy specimens. The etiology is not determined in most children, and the condition usually resolves. For 17% to 25% of cases, however, a pathologic process ultimately develops, most often a lymphoreticular disease.3–5 In acute pyogenic infection, lymph nodes are filled with neutrophils, microorganisms, edema, and necrotic debris. Granuloma formation along with caseating necrosis is typical of infections caused by Mycobacterium tuberculosis, Histoplasma capsulatum, Coccidioides immitis, and some nontuberculous mycobacteria. Stellate abscesses surrounded by palisading epithelioid cells are typical for lymphogranuloma venereum (caused by Chlamydia trachomatis) and infections caused by Bartonella henselae and Francisella tularensis (with most extensive granuloma formation in the latter). Toxoplasmosis produces characteristic nodal histologic results that show reactive follicular hyperplasia with scattered clusters of epithelioid histiocytes in cortical and paracortical zones, blurring of margins of germinal centers, and focal distention of subcapsular and trabecular sinuses by monocytoid cells. Yersinia spp. can cause necrotizing lymphadenitis in cervical, mediastinal, and mesenteric nodes. Brucellosis is characterized by noncaseating granulomas that are indistinguishable from those of sarcoidosis.

LYMPHOCUTANEOUS AND OCULOGLANDULAR SYNDROME For several infections resulting in regional and, with extension, generalized lymphadenopathy, a characteristic (often granulomatous) inflammatory reaction develops at the site of inoculation. Regional adenopathy often develops from lymphatic spread of organisms before the inoculation site has healed. When the initial inoculation site is in the conjunctiva, the constellation is called Parinaud oculo-

138

SECTION

B Cardinal Symptom Complexes

glandular syndrome (see Chapter 19, Cervical Lymphadenitis and Neck Infections) and, when in the skin, the lymphocutaneous syndrome. Organisms that cause these syndromes are shown in Table 19-3.

INFECTIOUS CAUSES OF GENERALIZED LYMPHADENOPATHY Because generalized or local lymph node enlargement can herald serious disease, such as a neoplasm, histiocytic proliferation, or autoimmune disease, benign, self-limited entities must be from distinguished enlargement resulting from potentially life-threatening disorders.6,7 Such distinctions can usually be made on the basis of history, lymph node examination (see Table 18-3), the presence of other systemic manifestations, and a limited number of laboratory tests (Box 18-2). Systemic infections are the most common causes of generalized lymphadenopathy. Many conditions characterized by generalized adenopathy also cause hepatic or splenic enlargement or, initially, regional adenopathy. Differentiating features of these syndromes and the lymphatic involvement characteristic of each syndrome are presented in Table 18-4.

Spirochetal Infection Rash and generalized, painless lymphadenopathy are present in 90% of patients with secondary syphilis. Enlargement of the epitrochlear nodes is a unique and common finding. In 70% of patients, rash and lymphadenopathy are accompanied by constitutional symptoms of fever, malaise, anorexia, and weight loss. In leptospirosis, generalized lymphadenopathy, and hepatosplenomegaly, in addition to constitutional symptoms of muscle tenderness, conjunctival injection, and skin rashes, are present in the first septicemic stage. The first clinical manifestation of Lyme disease is the typical annular rash, erythema migrans. Untreated, nearly 25% have dissemination that leads to secondary lesions of erythema migrans. Patients also complain of fever, headache, myalgia, malaise, and arthralgia; nontender regional or generalized lymphadenopathy can be present on physical examination.

Bartonella Infections In the mountain valleys of Peru, Ecuador, and southwest Colombia, between the altitudes of 600 and 2760 meters, Bartonella bacilliformis infection can be transmitted by the sand fly vector (or by blood transfusion); it causes Oroya fever. In the acute stage, cells of the BOX 18-2. Steps in Evaluating the Child with Generalized Lymphadenopathy • Take a careful history, paying attention to epidemiologic features, including travel, animal exposures, tick bite, unpasteurized milk, immunization status, blood transfusion, drug exposures; review of HIV risk factors for mother • Consider whether the child appears ill or well to determine pace of evaluation and differential diagnosis • Identify associated signs and symptoms, including fever and weight loss; perform thorough review of symptoms • Identify location and characteristics of lymph nodes, associated organomegaly, musculoskeletal findings, skin rash • Perform laboratory evaluation, including CBC, CRP, ESR, serum hepatic enzymes, LDH, ferritin, blood culture, chest radiograph • Serologic testing based on clinical risk; consider EBV, CMV, HIV, Mycoplasma, Treponema, Brucella, Bartonella, Histoplasma, Francisella, HHV-6, HHV-8; perform tuberculin skin testing • Consider biopsy of lymph node and plan specific testing of specimen CBC, complete blood count; CMV, cytomegalovirus; CRP, C-reactive protein; EBV, Epstein–Barr virus; ESR, erythrocyte sedimentation rate; HHV, human herpesvirus; HIV, human immunodeficiency virus; LDH, lactate dehydrogenase.

reticuloendothelial system contain many organisms after phagocytosis and destruction of infected, deformed red cells occurs. Fever, headache, and muscle and joint pain are accompanied by anemia and generalized, painless lymphadenopathy. Splenomegaly is only present in patients with intercurrent infection. B. henselae infection (cat-scratch disease) on occasion manifests with generalized lymphadenopathy, splenomegaly, and granulomatous hepatitis (especially in the immunocompromised host). In most instances of generalized lymphadenopathy, only two or three sites are involved, suggesting that, in some cases, separate inoculations occurred.

Enteric Infections Enteric fever caused by Salmonella spp. primarily affects the lymphoid tissue in Peyer patches in the ileum, the lymphoid follicles of the cecum, and the mesenteric nodes. About 50% of patients have hepatosplenomegaly. Occasionally, generalized adenopathy, particularly cervical adenopathy, is present. Lymphoid tissue undergoes consecutive stages of hyperplasia, necrosis, ulceration (after sloughing), and healing. Suppurative lymphadenitis has been described. Yersinia pseudotuberculosis and Y. enterocolitica infection has been associated with terminal ileitis and mesenteric adenitis, in which mesenteric lymph nodes are enlarged and can become necrotic and intensely suppurative. Enlarged inguinal and, rarely, cervical and hilar nodes have been described in cases of mesenteric adenitis.

Pulmonary Infections Although rare, legionellosis in children can manifest with nonspecific symptoms and occasionally with findings of rash, splenomegaly, and adenopathy. Pneumonia caused by Mycoplasma pneumoniae infection is accompanied by lymphadenopathy, particularly cervical, in about 25% of cases; mediastinal and hilar adenopathy have also been described. Tuberculosis is usually characterized by mediastinal and occasionally cervical adenopathy. Protracted hematogenous disease can cause high fever, hepatosplenomegaly, and generalized lymphadenopathy. Symptomatic, disseminated coccidioidomycosis occurs, rarely, within weeks or months following the initial localized pulmonary infection; most frequently, extrapulmonary spread is to bone, soft tissue, lymph nodes, and meninges. Tissue reaction is primarily granulomatous but may be accompanied by acute inflammation. In acute disseminated histoplasmosis in infants, the reticuloendothelial system has a high density of yeast forms compared with mycelial forms. Hepatosplenomegaly and intra-abdominal lymphadenopathy are common, and one-third of patients have peripheral lymphadenopathy. In acute pulmonary histoplasmosis, lymphohematogenous spread involves lymph nodes in the neck (where suppuration of supraclavicular or cervical nodes can occur), mediastinum, liver, and spleen. Paracoccidioidomycosis is endemic in most countries of Latin America, particularly Brazil. The disease is almost always disseminated, with tropism for the reticuloendothelial system. Peripheral lymphadenopathy is present in 75% of children with the acute or subacute forms. Cervical, inguinal, mesenteric, or mediastinal lymphadenopathy is almost universally present. Lymph nodes vary in size, number, and consistency and, over time, form abscesses and fistulas. Masses of lymph nodes (frequently with caseation) may be present; hepatosplenomegaly is usually present.

Other Bacterial Infections Acute onset of symptoms occurs in 50% of patients with brucellosis. As organisms are ingested by mononuclear phagocytes, disease is primarily localized to the lymph nodes, liver, spleen, and bone marrow. Noncaseating granulomas that are indistinguishable from sarcoidosis are the usual finding in liver biopsy specimens.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Lymphatic System and Generalized Lymphadenopathy

CHAPTER

18

TABLE 18-4. Diseases and Organisms That Cause Generalized Lymphadenopathy and Accompanying Predominant Regional Involvement Lymphadenopathy Regional Disease

Organism

Generalized

Hepatosplenomegaly

Mediastinal

Cervical

Other

Treponema pallidum Leptospira spp. Borrelia burgdorferi

++++ ++++ +

– ++++ +(H)

+ – –

++ ++ –

++ ++ ++

Bartonella bacilliformis

++++

–

–

–

–

Bartonella henselae

+

+(S)

++

+++

++++

Salmonella typhi Yersinia enterocolitica

++++ +

++++(S) –

+ +

+ +

+ +++

Mycoplasma pneumoniae Legionella pneumophila Mycoplasma tuberculosis Histoplasma capsulatum Coccidioides immitis

+ + ++ + +

– +(S) ++++ ++++ –

++ – ++++ +++ ++

++ – ++ ++ –

– – + ++ +

Paracoccidiodes brasiliensis

++++

+++

++++

Streptococcus pyogenes Brucella melitensis Francisella tularensis

+ +++ ++

+ +++ ++

– – –

++++ ++ +++

+++ + ++

Measles Rubella Chickenpox

Measles virus Rubella virus Varicella-zoster virus

+++ ++ ++

++(S) +(S) –

+++(A) – –

+++ ++++ –

+ + –

MONONUCLEOSIS SYNDROMES

Epstein–Barr virus Cytomegalovirus HIV HHV-6 Parvovirus B1919 Hepatitis A virus Toxoplasma gondii

++++ +++

+++ +++

+ –

++++ +++

++

++++ ++ + ++

+++ ++ +++(H) +

– – – –

++ – +++ ++++

+ +++ + ++

HHV-8

++++

+++

+++

+++

+++

Adenovirus Enterovirus

++ +

+++ +

– –

++++ +

++ +

Rickettsia tsutsugamushi Chlamydia trachomatis Ehrlichia sennetsu Ehrlichia chaffeensis

++++ + ++++ +++

+++ – – –

– – – –

+++ – – –

++++ ++++ – –

Trypanosoma cruzi

++++

++

–

+

+

Trypanosoma brucei

++

++

–

++++

+

++ +++

++++ –

– –

+ –

+ +++

+++ ++ +++ +++ ++

+++ + – ++ –

– – –

+ + ++++

+ + –

SPIROCHETAL SYNDROMES

Syphilis, secondary Leptospirosis Lyme disease BARTONELLA SYNDROMES

Bartonellosis (Oroya fever; verruga peruana) Cat-scratch disease

MESENTERIC LYMPHADENOPATHY SYNDROMES

Typhoid fever Yersiniosis PULMONARY SYNDROMES

Mycoplasma infection Legionnaire disease Primary tuberculosis Histoplasmosis (disseminated) Coccidioidomycosis (disseminated) Paracoccidioidomycosis MISCELLANEOUS BACTERIAL SYNDROMES

Scarlet fever Brucellosis Tularemia EXANTHEMATOUS SYNDROMES

CASTLEMAN DISEASE MISCELLANEOUS VIRAL SYNDROMES

Pharyngoconjunctival fever Nonspecific febrile illness RICKETTSIA/CHLAMYDIA

Scrub typhus Lymphogranuloma venereum Ehrlichiosis TROPICAL SYNDROMES

Chagas disease (American trypanosomiasis) African sleeping sickness (African trypanosomiasis) Kala-azar (leishmaniasis) Filariasis

Leishmania spp. Wuchereria bancrofti Brugia spp. Schistosomiasis (Katayama fever) Schistosoma spp. Dengue fever Dengue virus Chikungunya disease Chikungunya virus Lassa/Ebola fever Lassa and Ebola viruses West Nile fever West Nile virus

139

140

SECTION

B Cardinal Symptom Complexes

TABLE 18-4. Diseases and Organisms That Cause Generalized Lymphadenopathy and Accompanying Predominant Regional Involvement —Continued Lymphadenopathy Regional Disease

Organism

Generalized

Hepatosplenomegaly

Mediastinal

Cervical

Other

CONGENITAL SYNDROMES

HIV Rubella virus Cytomegalovirus Toxoplasma gondii Treponema pallidum Trypanosoma cruzi

++++ +++ + +++ +++ +++

+++ ++++ ++++ ++++ ++++ ++++

– – – – –

+++ – – + +++

+ + – + +++

Varied – – – – Salmonella typhi (inactivated)

++++ ++++ ++++ +++ ++ ++

++++ +++ ++++

– +++ – ++ – –

+++ +++ + +++ +++ ++

+++ + + +++ +++ ++

MISCELLANEOUS SYNDROMES

IAHS Sarcoidosis Gianotti–Crosti syndrome Chronic granulomatous disease Chronic atopic eczema Typhoid immunization

+ –

A, atypical; H, predominantly hepatomegaly; HHV, human herpesvirus; HIV, human immunodeficiency virus; IAHS, infection-associated hemophagocytic syndrome; S, predominantly splenomegaly; ++++, characteristic association; +++, frequent association; ++, occasional association; +, rare association.

Lymphadenitis is a common manifestation of tularemia. Although regional involvement (with the site, depending on whether acquisition was oropharyngeal or tick-related) is the most common presentation, involvement of multiple sites may occur. Hepatosplenomegaly was present in 35% of infected children in one series.8 Generalized lymphadenopathy is rarely described as a feature of scarlet fever. It may be more typical of the severe toxic form of the disease.

Exanthematous Syndromes Studies of rubella virus inoculation in adolescents and young adults with volunteers demonstrated that the most characteristic sites for lymphadenopathy are the suboccipital and posterior auricular nodes, but generalized involvement may also occur.9 Pharyngitis and cervical lymphadenopathy are common in measles during the exanthematous period. Generalized lymphadenopathy with suboccipital, posterior auricular, and mediastinal involvement and splenomegaly are common. Generalized lymphadenopathy is an uncommon manifestation of varicella-zoster viral infection and is more often related to secondary bacterial infections. Parvovirus B19 infection is most often seen in conjunction with erythema infectiosum. A mononucleosis-like syndrome, along with generalized lymphadenopathy and hepatosplenomegaly, is also described.

Mononucleosis Syndromes Lymphadenopathy in Epstein–Barr virus (EBV)-associated mononucleosis most prominently involves anterior and posterior cervical nodes, but diffuse lymphadenopathy is often present and involves the occipital, supraclavicular, axillary, epitrochlear, inguinal, and mediastinal lymphatic chains. Enlarged nodes are usually nontender (or minimally tender) and most prominent during the second to fourth week of illness, with no overlying erythema. Splenomegaly is present in 50% and hepatomegaly in 30% to 50% of cases. In children, cytomegalovirus (CMV) is a less common cause of mononucleosis than is EBV, but clinical manifestations are similar. In one study of 124 children with the mononucleosis syndromes, fever,

hepatosplenomegaly, rashes, and upper-airway obstruction were found with equal frequency in infection caused by EBV or CMV. Cervical lymphadenopathy was more common with EBV (93% of cases) than CMV (75%). Clinical disease caused by human immunodeficiency virus (HIV) in infants and young children is characterized by generalized lymphadenopathy, hepatosplenomegaly, failure to thrive, intermittent fever, chronic or recurrent diarrhea, parotitis, chronic dermatitis, and recurrent infections. In older children and adults, a mononucleosislike syndrome (so-called seroconversion syndrome) can occur weeks after HIV infection and includes generalized lymphadenopathy, fever, malaise, myalgia, headache, sore throat, diarrhea, and rash. Lymphadenopathy can involve several sites and persist for months. Lymph nodes are discrete, nontender, and nonsuppurative; biopsy results reveal follicular hyperplasia. With progression to acquired immunodeficiency syndrome (AIDS), lymphocyte depletion occurs; a biopsy specimen of rapidly enlarging nodes should be examined to rule out malignancy. Patients with anicteric hepatitis A or B can manifest a mononucleosis-like syndrome, commonly characterized by posterior cervical lymphadenopathy. Generalized lymphadenopathy is unusual; splenomegaly is present in 15% of patients. The most common findings in symptomatic acquired toxoplasmosis are fatigue and lymphadenopathy without fever. Nodes are discrete, may or may not be tender, and do not suppurate. The cervical, suboccipital, supraclavicular, axillary, and inguinal nodes are most often involved. Lymphadenopathy can be localized or generalized, including involvement of the retroperitoneal and mesenteric nodes. Lymphadenopathy can simulate lymphoma.

Miscellaneous Viral Infections The usual onset of pharyngoconjunctival fever resulting from an adenovirus infection is abrupt, with sore throat, headache, generalized aches and pains, eye irritation or pain, and fever. Some degree of anterior and posterior cervical lymphadenopathy occurs in most patients. Preauricular lymphadenopathy is surprisingly uncommon, and generalized lymphadenopathy is found in 10% to 20% of patients. Hepatosplenomegaly is common. Generalized lymphadenopathy occasionally occurs in nonspecific illnesses caused by enteroviruses.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Lymphatic System and Generalized Lymphadenopathy

Rickettsia, Ehrlichia, and Anaplasma In patients with scrub typhus, after an incubation period of 1 to 2 weeks, the initial mite bite lesion and necrotic eschar are noted. This coincides with the onset of the main characteristic features of the disease, which are fever, headache, rash, and generalized lymphadenopathy. Rare or absent in all other rickettsial diseases, lymphadenopathy is particularly prominent in patients with scrub typhus in the axilla, neck, and inguinal areas. Hepatosplenomegaly and conjunctival injection are common. Sennetsu fever caused by Ehrlichia sennetsu appears to be confined to western Japan. Abrupt onset of fever, chills, headache, malaise, sore throat, and muscle and joint pains is accompanied by generalized, tender lymphadenopathy. Posterior auricular and posterior cervical adenopathy are particularly prominent. Generalized lymphadenopathy is not as consistently found in patients with Ehrlichia or Anaplasma infections.

Tropical Syndromes During the acute phase of symptomatic Chagas disease (Trypanosoma cruzi), generalized lymphadenopathy, moderate hepatosplenomegaly, rash, vomiting, diarrhea, and neurologic and cardiac changes are present. At various stages of African trypanosomiasis, the hemoflagellate trypanosomal protozoans (T.b. rhodesiense and T.b. gambiense) can be found in lymphatics and lymph nodes. Fever and posterior cervical adenopathy may be present, along with the local chancre. In visceral leishmaniasis, or kala-azar, the principal histopathologic lesions are the result of infected macrophages and reticuloendothelial hyperplasia. Massive splenomegaly and generalized lymphadenopathy with pancytopenia are present. In filarial infections, lymphangitis with involvement of regional nodes is more common than generalized lymphadenopathy. Hematogenous dissemination by way of the thoracic duct, however, can lead to generalized adenopathy. The acute schistosomiasis syndrome, or Katayama fever, is seen with severe schistosomal infestation. Generalized, nontender lymphadenopathy, splenomegaly, and hepatomegaly with tenderness are common, in addition to fever, headache, myalgia, weakness, and gastrointestinal symptoms. Generalized lymphadenopathy is an integral part of several tropical viral hemorrhagic fever syndromes (see Table 18-4).

Congenital Syndromes One-third of infants who have congenital toxoplasmosis show signs and symptoms of acute infection. In these infants, splenomegaly (90%), hepatomegaly (70%), and generalized adenopathy (68%) are present. Hepatosplenomegaly is present in 50% to 75% and generalized lymphadenopathy in 20% to 50% of infants with congenital rubella. These manifestations usually clear over a few weeks. Generalized lymphadenopathy is uncommon in congenital herpes simplex and CMV infections. Infants with intrauterine HIV infection have generalized painless lymphadenopathy; lymph node biopsy results can demonstrate a variety of patterns, including follicular hyperplasia, angioimmunoblastic changes, or atrophy. Hepatosplenomegaly is present in almost all infants with early congenital syphilis. Generalized lymphadenopathy is described in 50% of patients. Nodes may be as large as 1 cm and are usually nontender. Enlarged epitrochlear nodes are relatively unique to this syndrome.

OTHER CAUSES OF GENERALIZED LYMPHADENOPATHY Hodgkin and Non-Hodgkin Lymphoma Lymphadenopathy can result from malignant transformation of cells intrinsic to the node or from metastatic spread of cells extrinsic to the

CHAPTER

18

141

node (Box 18-3). Primary neoplasms of lymph nodes include the lymphomas and histiocytosis.10 Data from the American Cancer Society for 2005 show that lymphomas account for approximately 8% of cancers in children less than 15 years old, and leukemia and central nervous system tumors account for 30% and 21% respectively. Of the lymphomas, both Hodgkin lymphoma (HL) and nonHodgkin lymphoma (NHL) occur. HL in children under age 16 years accounts for 10% to 15% of the 7880 total HL cases diagnosed in the United States. In United States children, NHL make up about 5% of the 53 370 cases of the total cases of NHL diagnosed each year. Among childhood NHL, lymphoblastic lymphomas (30%), Burkitt lymphoma (30% to 40%), anaplastic large-cell lymphoma (10%) and large B-lymphocyte lymphoma (20%) are most common. The cure rate for lymphoma in children is as high as 90% for certain categories; lymphoma is among the most curable of all cancers – this makes early diagnosis important. The most common first clinical manifestation of both HL and NHL is painless lymph node enlargement, most often in the cervical (in 60% to 90% of cases) or supraclavicular chains. Onset is typically subacute and prolonged in HL, but it occurs rapidly over a few days or weeks in NHL. HL is primarily a disease of adolescents and young adults. Of childhood cases, 60% occur in children aged 10 to 16 years, with fewer than 3% of cases in children under age 5 years. Constitutional symptoms of fever, weight loss, drenching night sweats, generalized pruritus, and pain are present in only 30% of children at the time of presentation. Ninety percent of patients with HL come to medical attention with an unusual mass or swelling and enlarged painless supraclavicular or cervical nodes. Up to 70% of patients have mediastinal lymphadenopathy at the time of presentation, making a chest radiograph valuable if the diagnosis is suspected. Involvement of the Waldeyer ring should be considered if high cervical nodes are enlarged. Hepatosplenomegaly occurs in patients with advanced disease. Left-sided or bilateral cervical or supraclavicular involvement is more common in extramediastinal paraaortic and splenic spread. Mediastinal disease more often accompanies right-sided cervical or supraclavicular disease.

Other Lymphoproliferative Disorders Children with congenital or acquired immunodeficiency syndromes (including HIV infection and immunosuppression following organ transplantation) have risk of developing lymphoproliferative disorders that is 100 to 10,000 times that of age-matched controls. The lymphoproliferative disorders are a heterogeneous group of B-lymphocyte proliferations that range from polyclonal hyperplasia to true monoclonal malignant lymphoma. The child’s underlying condition (inherited, iatrogenic, or acquired) permits expansion of lymphoid

BOX 18-3. Noninfectious Causes of Generalized Lymphadenopathy • Inherited immunodeficiency syndromes Chronic granulomatous disease Wiskott–Aldrich syndrome Chédiak–Higashi syndrome • Papular acrodermatitis (Gianotti–Crosti syndrome) • Sarcoidosis • Hyperthyroidism • Drug-induced hyperplasia/hypersensitivity • Autoinflammatory and autoimmune disorders • Lipid storage diseases • Other lymphoid hyperplasia syndromes Infection-associated hemophagocytic syndrome Rosai–Dorfman disease Kikuchi–Fujimoto disease Multicentric Castleman disease • Cancers Primary lymphomas and metastatic neoplasms Non-Hodgkin lymphoproliferative disorders Childhood histiocytoses

142

SECTION

B Cardinal Symptom Complexes

cell populations that would be more strictly regulated in the healthy child. Defects in immune regulation and surveillance result in unchecked proliferation of subpopulations of cells with new, irreversible cytogenetic aberrations, eventuating in NHL (Box 18-4). A mononucleosis-like syndrome with fever, pharyngitis, and generalized lymphadenopathy resulting from lymphoproliferation develops. In early stages, this syndrome mimics mononucleosis, and EBV infection plays an etiologic role (see Chapter 207, Epstein–Barr Virus).

Drug-Induced Hyperplasia Related to Hypersensitivity Localized or generalized lymphadenopathy can appear 1 to 2 weeks after beginning phenytoin or other anticonvulsant medications such as carbamazepine. A severe, pruritic, maculopapular rash, along with fever, hepatosplenomegaly, jaundice, anemia, leukopenia, and plasmacytosis in blood and bone marrow occurs concurrently with or after lymphadenopathy. Other drugs that have been associated with generalized lymphadenopathy include pyrimethamine, antileprosy and antithyroid drugs, isoniazid, aspirin, barbiturates, penicillin, tetracycline, iodides, sulfonamides, allopurinol, and phenylbutazone.

Metastatic Neoplasms Metastatic involvement of lymph nodes occurs in a variety of nonlymphomatous cancers, including neuroblastoma, the leukemias, and malignancies of the head and neck (rhabdomyosarcoma, lymphosarcoma, and sarcoma of the parotid gland, thyroid, and nasopharynx).

Histiocytosis Childhood histiocytoses are a rare and diverse group of disorders that can cause generalized lymphadenopathy. Diagnosis and treatment are difficult. The histiocytoses are characterized by infiltration and accumulation of macrophage cells formed from histiocytic stem cells. Cells of the histiocytic system consist of antigen-processing (phagocytic cells) and antigen-presenting (dendritic cells). Both are found in normal and reactive lymph nodes. Langerhans dendritic cells are primarily located in the skin but can be found in lymph nodes. Sinus histiocytes are the principal cells involved in phagocytosis of foreign particulate matter and are primarily located in lymph node sinuses. The pathophysiology of histiocytoses is thought to be uncontrolled immunologic stimulation of normal antigen-processing cells, rather than malignant transformation. Lymphadenopathy is a frequent manifestation.

Infection-Associated Hemophagocytic Syndrome Infection-associated hemophagocytic syndrome (IAHS) is characterized by reactive histiocytic hyperplasia with leukoerythrophagocytosis in a variety of organs. Children with IAHS are usually critically ill, with fever, pancytopenia, generalized lymphadenopathy, and hepatosplenomegaly.11,12 Laboratory abnormalities include

cytopenia, disseminated intravascular coagulation, elevated levels of tissue enzymes and triglycerides, and extremely high levels of cytokines.13 Bone marrow examination reveals excessive proliferation of benign-looking histiocytes that are phagocytosing white blood cells, red blood cells, and platelets. IAHS has been associated with viruses, bacteria, fungi, and parasites, although the pathophysiology is unknown. IAHS is more commonly associated with immunodeficiency syndromes but can occur in otherwise healthy children. The mortality rate is high (20% to 40%). IAHS should be distinguished from malignant histiocytosis and familial erythrophagocytic lymphohistiocytosis (see Chapter 14, Systemic Inflammatory Syndromes Mimicking Infectious Disease).

Inherited Immunodeficiency Syndromes Generalized lymphadenopathy is often characteristic of inherited immunodeficiency syndromes (e.g., Wiskott–Aldrich syndrome and common variable immunodeficiency) (see Box 18-4). Detection of dysregulated lymphoproliferation in these patients with underlying chronic generalized lymphadenopathy can be difficult. Over 85% of patients with Chédiak–Higashi syndrome have an accelerated phase of disease with generalized lymphadenopathy and hepatosplenomegaly. Lymphadenopathy, with or without dermatitis, is a common early feature of chronic granulomatous disease and occurs in almost all patients. Although cervical lymphadenopathy is most common, femoral, inguinal, hilar, mediastinal, and generalized adenopathy are also observed. Typically, early signs of inflammation, such as fever, pain, and local inflammation, are lacking and affected nodes suppurate chronically (cold abscesses). Such nodes demonstrate caseating or noncaseating granulomas with central necrosis, a reaction to inadequate killing of intracellular and extracellular organisms.

Sinus Histiocytosis with Massive Lymphadenopathy (Rosai–Dorfman Disease) Sinus histiocytosis with massive lymphadenopathy (Rosai–Dorfman disease) is characterized by generalized proliferation of sinusoidal histiocytes.14 A rare disease that occurs in the first two decades of life, the cause of Rosai–Dorfman disease is unknown but is thought to be the result of dysregulated immune system or response to a presumed infectious agent or both. All patients have a history of bilateral cervical lymphadenopathy, often of several months’ duration. Lymphadenopathy can be asymmetric, and in 80% of patients, involves other nodal groups, including the axillary, inguinal, hilar, and mediastinal nodes. In 30% of patients, extranodal disease occurs, involving the nasal and oral cavities, salivary glands, pharynx, tonsils, paranasal sinuses, trachea, orbit, bone, or skin. Systemic manifestations usually include fever, anemia, leukocytosis, elevated sedimentation rate, and hypergammaglobulinemia. The course is usually indolent with spontaneous regression, but immune-modulating therapy may be required for extensive or progressive disease.

Histiocytic Necrotizing Lymphadenitis (Kikuchi–Fujimoto Disease) BOX 18-4. Lymphoproliferative Disorders Associated with Immunodeficiency • Inherited immunodeficiency syndromes Ataxia-telangiectasia syndrome Wiskott–Aldrich syndrome Combined immunodeficiency syndromes X-linked lymphoproliferative syndrome • Immunodeficiency resulting from solid organ and stem-cell transplant therapy • Viral infections causing immunodeficiency syndromes Epstein–Barr virus infection Retrovirus infection

Histiocytic necrotizing lymphadenitis (Kikuchi–Fujimoto disease) is a benign cause of lymphadenopathy, which is accompanied in 20% of patients by systemic symptoms such as fever, nausea, weight loss, night sweats, arthralgia or hepatosplenomegaly, and leukopenia.15 The most common manifestation is regional or generalized lymphadenopathy, with or without fever. The median age of onset is 30 years, but cases in children 2 to 18 years of age have been described. Lymph node histologic examination reveals reactive histiocytes and large immunoblastic lymphoid cells. Focal necrotic lesions are present in the paracortical areas of lymph nodes. The cause is thought to be dysregulation of the disordered immune system. The prognosis is excellent, with spontaneous resolution of lymphadenopathy within 4 months.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Cervical Lymphadenitis and Neck Infections

Multicentric Castleman Disease Castleman disease is an atypical lymphoproliferative disorder characterized by hyperplasia of lymph nodes in which plasma cell predominance and marked capillary proliferation are noted histologically. It may be identified as a localized, indolent disease that can be cured by local excision or, particularly in HIV-infected individuals, as a rapidly progressive disease with a poor prognosis. The multicentric form has been strongly associated with human herpesvirus 8 (HHV-8).16,17 Symptoms include fever, myalgia, and weight loss; patients have diffuse lymphadenopathy, splenomegaly and rhabdomyolysis. An HIV-infected adult with Castleman disease (with HHV-8 identified in lymph nodes, peripheral blood, and mononuclear cells) responded well to foscarnet therapy with concomitant antiretroviral therapy.18

Autoinflammatory and Autoimmune Disorders Forty percent of children with juvenile idiopathic arthritis of systemic onset have generalized lymphadenopathy, often preceding joint involvement. Splenomegaly and, less often, hepatomegaly are present. Seventy percent of children with systemic lupus erythematosus have lymphadenopathy, which is generalized in one-third of these children. Hepatosplenomegaly is common. Serum sickness (a hypersensitivity reaction to foreign proteins, often drugs) is characterized by fever, urticaria, edema, polyarthralgia, and generalized lymphadenopathy. In Sjögren syndrome, a rare disease in children, lymphadenopathy and splenomegaly may be present along with parotid and lacrimal gland involvement. Lymphadenopathy in Kawasaki disease is not usually generalized.

Papular Acrodermatitis (Gianotti–Crosti Syndrome) Papular acrodermatitis of childhood is a distinctive eruption in children aged 6 months to 12 years; it is associated with malaise, lowgrade fever, generalized lymphadenopathy, and hepatomegaly (when accompanied by hepatitis B viremia). The discrete, 1- to 5-mm, flattopped, firm papules that appear in groups on the face, buttocks, limbs, palms, and soles resolve spontaneously within 3 weeks, whereas lymphadenopathy and hepatomegaly can persist for months. This eruption has been associated with a number of viral syndromes, including hepatitis B and C infection.

Sarcoidosis Sarcoidosis is a multisystem granulomatous disease of unknown cause. It affects young adults, most commonly manifesting as asymptomatic bilateral hilar and paratracheal adenopathy, often with parenchymal lung involvement. Generalized lymphadenopathy with prominent cervical involvement is the most common finding in children. Nodes are discrete, painless, and freely movable. Characteristic findings on histologic examination of nodes are epithelioid cell tubercles with little or no necrosis. Other findings result from local granuloma formation and include changes in the eyes, skin, liver, spleen, and parotid glands. Biopsy results on a supraclavicular node are diagnostic in 85% of cases.

Lipid Storage Diseases In Niemann–Pick, Gaucher, Wolman, and Farber diseases, lipid-laden histiocytes accumulate in lymph nodes, liver, or spleen, resulting in detectable enlargement. Diagnosis is made on the basis of bone marrow examination or lymph node biopsy.

Miscellaneous Disorders Hyperthyroidism is associated with generalized lymphoid hyperplasia. Beryllium exposure can result in granulomatous generalized lymphadenopathy.

CHAPTER

CHAPTER

19

143

19

Cervical Lymphadenitis and Neck Infections Emily A. Thorell and P. Joan Chesney

EPIDEMIOLOGY Neck masses in children, unlike those in adults, seldom represent ominous disease. Most (95%) masses are enlarged or inflamed lymph nodes and are acute in nature. Other masses are congenital cysts and sinuses (3%), vascular malformations, salivary and thyroid anomalies, benign and malignant neoplasms, traumatic injuries, and nonlymphatic infections. Of those children admitted to the hospital for evaluation of a persistent neck mass, malignancy is present in about 15%.1,2 Although cervical lymphadenopathy (enlargement) and lymphadenitis (inflammation) are very common, only a few original studies are recent and represent a minority of cases. Most were published decades ago. Attempts to delineate the causes and causative agents of neck masses and lymphadenitis have been based on analysis of specimens obtained through needle aspiration and surgical biopsy, or by serologic testing. In 40% to 80% of cases, aspiration of acutely inflamed, enlarged unilateral nodes reveals infection by Staphylococcus aureus or Streptococcus pyogenes.3,4 Recent studies of suppurative adenitis show the predominance of S. aureus and the emergence of community-acquired methicillin-resistant Staphylococcus aureus (CA-MRSA).5,6 Anaerobes also play an important role. In a study from 1980, anaerobic bacteria were isolated in 38% of cases, and the anaerobe was the only isolate in one-half.7 As anaerobic organisms outnumber aerobic and facultative organisms in normal oropharyngeal flora by 10 to 1, this finding is not surprising. Depending on many factors, especially pretreatment with antimicrobial therapy, no pathogen is isolated in at least 25% of cases of acutely inflamed nodes.4,7 Other organisms identified in one series of aspiration of acutely inflamed nodes were: coagulase-negative staphylococci, gram-negative bacilli, group B and group C streptococci, viridans group streptococci, Francisella tularensis, nontuberculous Mycobacterium, and Histoplasma capsulatum. Many other organisms have been isolated sporadically from enlarged cervical nodes. One large study from 1963 examined the results of surgical biopsy of persistent cervical masses in 267 children.2 Congenital cyst or cystic hygroma accounted for 60% of masses; malignant tumors accounted for 15.7%, and benign tumors for 7%. Nonspecific lymphoid hyperplasia was found in 10%, and tuberculous granuloma in 7%. Of the 46 malignant tumors, 20 were lymphosarcoma or Hodgkin disease; 11, neurogenic tumors (neuroblastoma, neurofibroma, neurofibrosarcoma); 11, thyroid; and 4, parotid tumor. In a second series of excisional lymph node biopsy in 75 children from 1978, malignant neoplasm was present in 17%, nondiagnostic hyperplasia in 55%, noncaseating granuloma in 21%, and caseating granuloma in 7%.1 In a third review of peripheral lymph node excisional biopsy in 239 children, a specific cause was found in only 41% of cases.8 Of all nodes undergoing biopsy, 182 were located in the neck. Reactive hyperplasia accounted for 52% of cases, granulomatous diseases for 32%, neoplasia for 13%, and chronic lymphadenitis for 3%. Of the 31 malignancies, Hodgkin lymphoma and non-Hodgkin lymphoma (NHL) accounted for 24, neuroblastoma for 4, and rhabdomyosarcoma for 3. In this series, generalized lymphadenopathy was most often associated with reactive hyperplasia. Close follow-up of patients with a diagnosis of nonspecific reactive hyperplasia is important, because up to 25% of such patients ultimately have severe lymphoreticular disease.1,5,9,10

144

SECTION

B Cardinal Symptom Complexes

TABLE 19-1. Age-Specific Causes of Cervical Lymphadenitis

TABLE 19-2. Physiology of Lymph Flow in the Head and Neck

Patient Age Organism

2 Months– Neonate 1 Year

Lymph Node Group

Areas Drained

HEAD

1–4 Years

5–18 Years

Group A streptococcus

–

–

+

++

Group B streptococcus

++

+

–

–

Staphylococcus aureus

+

++

++

++

Nontuberculous Mycobacterium

–

–

++

+

Bartonella henselae

–

+

++

++

Toxoplasma gondii

–

–

+

+

Anaerobic bacteria

–

–

+

++

Postauricular (mastoid)

Temporal and parietal scalp; posterior wall of ear canal; upper half of pinna

Occipital

Posterior scalp; skin of upper, posterior side of neck

Preauricular

Anterior and temporal regions of scalp; anterior ear canal and pinna; lateral conjunctivae

Parotid

Root of nose; eyelids; temporal scalp; exterior auditory meatus; parotid glands; middle ear; floor nasal activity; posterior palate

Facial

Eyelids, conjunctivae; skin and mucous membranes of nose and cheek; nasopharynx

++, Frequent cause; +, occasional cause; –, rare cause or never a cause. NECK

Surgical biopsy results from children in developing countries show a slightly different spectrum of disease. A 2003 review of 1332 patients younger than 15 years of age over a 23-year span in South Africa shows 48% reactive nodes, 25% Mycobacterium tuberculosis, 11.6% neoplasm, and 11.5% chronic granuloma. Miscellaneous etiologies included syphilis, yaws, toxoplasmosis, sinus histiocytosis, and Kaposi sarcoma.11 Age is useful in predicting etiology of infection (Table 19-1). Group B streptococci commonly cause lymphadenitis in young infants; Staphylococcus aureus infects infants and children; and Streptococcus pyogenes and Staphylococcus aureus cause lymphadenitis in children 1 to 4 years old, as does nontuberculous Mycobacterium. Acute cervical lymphadenitis may be more common in the young child because of an inability to localize the organism at the initial nasal or pharyngeal site of attachment. In children 5 to 15 years old and adults, anaerobic bacteria, toxoplasmosis, cat-scratch disease, and tuberculosis are important considerations.

Submental

Central lower lip; floor of mouth; skin of chin; tongue tip

Submandibular (submaxillary)

Buccal mucosa; side of nose; medial palpebral commissure; upper lip; lateral part of lower lip; gums, anterior part of tongue margin, teeth

Superficial cervical

Anterior: superficial anterior neck tissues including skin, lower larynx, thyroid, cranial trachea Posterior: lower ear canal; parotid region

Superior deep cervical

Tonsil, adenoid; posterior scalp and neck; tongue, larynx, palate; thyroid; nose, nasopharynx; esophagus; paranasal sinuses; all nodes of head and neck except inferior deep cervical

Inferior deep cervical (scalene, supraclavicular)

Dorsal scalp and neck; superficial pectoral region of arm; superior deep cervical nodes; larynx, trachea; thyroid Right: left lower lobe, lingula, right lung and pleura Left: left upper lobe; entire abdomen

LYMPHATIC FLOW IN THE HEAD AND NECK A complex and efficient lymphatic system has evolved to defend against microbial invasion of the head, neck, nasopharynx, and oropharynx, as shown in Table 19-2 and Figure 19-1. There are three interrelated lines of defense. The ring of Waldeyer is composed of a circle of adenoidal, tonsillar, and lingual lymphoid tissue (see Chapter 18, Lymphatic System and Generalized Lymphadenopathy). A collar of superficial satellite lymph nodes runs along the lower margins of the jaw and encircles this ring. This outlying collar of nodes of the head consists of the occipital, postauricular, preauricular, parotid, and facial groups. Finally, the nodes of the neck are the salivary glandassociated submaxillary and submental nodes and the vertically oriented superficial and deep cervical chains. Occipital nodes are often enlarged when generalized lymphadenopathy is present; regional enlargement is almost always infectious and is due to tinea capitis, pediculosis capitis, seborrheic dermatitis, or scalp cellulitis or abscess. Enlargement of preauricular nodes reflects local skin or conjunctival infection. Asymptomatically enlarged parotid glands raise the possibility of malignancy.

Specific Lymph Node Groups Superficial Cervical Nodes The superficial cervical nodes in the neck are a disparate group composed roughly of three vertical chains. The deep lateral or spinal accessory chain runs behind the posterior border of the sternomastoid

muscle and along the spinal accessory nerve. The superficial cervical chain follows the external jugular vein, which runs obliquely across the surface of the sternomastoid muscle to empty into the subclavian vein in the supraclavicular triangle. The superficial anterior chain runs in the midline from the chin to the suprasternal notch and comprises, in descending order, the infrahyoid, prelaryngeal, pretracheal, and anterior cervical nodes. “Posterior cervical nodes” is a nonspecific term referring to nodes of the spinal accessory chain and those of the superficial cervical chain that lie over and behind the sternomastoid muscle. Lymph nodes in and anterior to the sternomastoid muscle are designated as lying in the anterior triangle. Those behind the muscle lie in the posterior triangle.

Deep Cervical Nodes The deep cervical chain of lymph nodes runs from the base of the skull to the root of the neck in close approximation to the internal jugular vein on the carotid sheath, and under the sternomastoid muscle. This chain contains numerous large nodes and is divided into superior and inferior deep cervical groups. The superior group is above, and the inferior group below, the point low in the neck where the omohyoid muscle crosses the internal jugular vein. The lymphatic channels and nodes in the head and neck are linked, and ultimately all empty into the thoracic duct on the left or the lymphatic duct on the right, each of which in turn immediately empties into the respective subclavian vein.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Cervical Lymphadenitis and Neck Infections Pre-parotid, intra-parotid, or superficial parotid Parotid gland Facial

Preauricular

Sub or mid-mandibular

Postauricular (mastoid)

Submental

Subparotid Jugulodigastric (tonsilar)

Superficial jugular

Deep lateral (spinal accessory) Posterior cervical Trapezius muscle

Internal jugular Internal jugular vein Sternomastoid muscle

Occipital

Transverse cervical (supraclavicular) Drainage from thoracic duct

CHAPTER

19

145

BOX 19-1. Organisms Associated with Lymphocutaneous and Oculoglandular (Parinaud) Syndromes VIRUSES Herpes simplex BACTERIA Treponema pallidum Mycobacterium tuberculosis Bartonella henselae Francisella tularensis Corynebacterium diphtheriae Spirillum minor Chlamydia trachomatis Bacillus anthracis Nocardia brasiliensis Yersinia enterocolitica Listeria monocytogenes FUNGI Histoplasma capsulatum Coccidioides immitis Paracoccidioides spp. PARASITES Trypanosoma spp. RICKETTSIA Rickettsialpox

OCULOGLANDULAR SYNDROME OF PARINAUD

Figure 19-1. Lymphatic flow in the head and neck.

The superior deep cervical group of nodes consists of the tonsillar or jugulodigastric node located at the angle of the mandible just below the posterior belly of the digastric muscle. The lymphoid tissue of the palatine tonsil drains into this gland. Other nodes of this group, which lie under the sternomastoid muscle along the length of the internal jugular vein, drain the adenoid, larynx, trachea, thyroid, palate, esophagus, paranasal sinuses, nasopharynx, occipital scalp, back of the neck, pinna, and much of the tongue. The large jugulo-omohyoid node that drains the tongue lies just above the omohyoid muscle, at the point where it crosses the internal jugular vein, which separates the superior and inferior deep cervical nodes. Because the majority of lymphatics of the head and neck drain to the submaxillary and deep cervical nodes, it is not surprising that these nodes are involved in more than 80% of young children with cervical lymphadenitis. The inferior deep cervical nodes lie low in the neck, below the omohyoid muscle and under and posterior to the anterior clavicular insertion of the sternomastoid muscle. They lie in close approximation to the brachial plexus and to the entrance sites of the thoracic duct and right lymphatic duct into the left and right subclavian veins. All lymph from the head and neck, arms, superficial thorax, lungs, mediastinum, and abdomen passes through these nodes. The left supraclavicular nodes (Virchow–Troisier nodes) drain the left upper lobe of the lung, left mediastinum, stomach, small intestine, kidney, and pancreas. Enlargement of these nodes in the absence of cervical adenopathy suggests intra-abdominal tumor or inflammation (Troisier sign) or intrathoracic disease. The right supraclavicular nodes drain the left lingula and lower lobes as well as the entire right lung, pleura, and right mediastinum. Enlargement of these nodes most often indicates thoracic lesions. Both Hodgkin lymphoma and NHL are common causes of enlargement of these nodes, and prompt biopsy is indicated in the absence of easily documented pulmonary or cervical infection.

The oculoglandular syndrome of Parinaud consists of unilateral, chronic granulomas or ulcers of the conjunctivae associated with preauricular and submaxillary lymphadenitis.12 Causes are listed in Box 19-1. The most common cause in children is Bartonella henselae infection; primary conjunctival inoculation causes gray or yellow granulomatous nodules or areas of focal necrosis, often surrounded by significant conjunctival chemosis and palpebral inflammation. Recovery without sequelae (except for occasional mild conjunctival scarring) occurs within 2 to 3 months. Antimicrobial therapy has not been proven to alter the course of the oculoglandular syndrome of Parinaud.13

INFECTIOUS CAUSES OF LYMPHADENITIS Most cases of infectious cervical lymphadenitis can be divided into the following three categories (with some overlap): (1) acute bilateral lymphadenitis; (2) acute unilateral lymphadenitis; and (3) subacute (chronic) lymphadenitis.

Acute Bilateral Lymphadenitis Acute bilateral lymphadenitis is most often a localized response to acute pharyngitis or part of a generalized lymphoreticular response to systemic infection in older children; viral upper respiratory tract infection is the most common cause, followed by pharyngitis due to Streptococcus pyogenes and Mycoplasma pneumoniae. Usually, lymph nodes are small and soft, may or may not be tender, and are not associated with erythema or warmth of the overlying skin. Viruses are the cause of the majority of cases.

Epstein–Barr Virus Although generalized lymphadenopathy is common in Epstein–Barr virus (EBV) infection, cervical adenopathy is most prominent and is present in 93% of children with this infection.14 Enlargement is almost always bilateral and most prominent in the posterior cervical chain, followed by the anterior cervical chain. Lymph nodes are enlarged

146

SECTION

B Cardinal Symptom Complexes

singly or in groups, vary from 5 to 25 mm in longest diameter, and are firm, discrete, and minimally tender. Moderate splenomegaly occurs in 75% of patients. Significant enlargement of the lymphoid tissue in Waldeyer ring can result in nasopharyngeal and oropharyngeal airway obstruction.

Cytomegalovirus Mononucleosis due to cytomegalovirus (CMV) is similar to that due to EBV, although more frequent in children < 4 years of age.14 Cervical adenopathy is less common in CMV infection (75%) than in EBV (95%) infection. Tonsillopharyngitis and sore throat are more common in EBV infection, whereas hepatosplenomegaly, upperairway obstruction, and rashes are more common in CMV infection.

Herpes Simplex Virus Infections Cervical, submaxillary, and submental nodes are frequently enlarged and tender during the course of primary gingivostomatitis due to herpes simplex virus. Primary infections of the conjunctivae and lids are accompanied by preauricular adenopathy. Localized or regional necrotizing lymphadenitis from primary HSV infection has also been reported as a rare occurrence.15

Rubella Virus In patients with experimentally induced rubella, lymph node enlargement begins as early as 7 days before the onset of the rash. Although generalized lymphadenopathy occurs, the nodes most commonly involved are the posterior auricular, suboccipital, and cervical nodes. Tenderness and swelling are most severe on the first day of the rash. Although tenderness subsides within 1 to 2 days, enlargement persists for weeks.

Parvovirus B19 A mononucleosis-like syndrome due to parvovirus B19 infection can be associated with bilateral cervical and intraparotid lymphadenopathy. Generalized lymphadenopathy and hepatosplenomegaly can also be present. Facial palsy and parotitis can accompany the intraparotid lymphadenopathy.20

Mycoplasma pneumoniae Physical examination of children and adults with pneumonia due to M. pneumoniae reveals cervical adenopathy in 25%, pharyngitis in 50%, and auscultatory findings in 75%. In patients in whom pharyngitis is the major manifestation, the pharyngitis is exudative in 43%, and cervical adenopathy is found in 50%.

Adenoviral Syndromes Pharyngoconjunctival fever, caused by adenovirus, is characterized by abrupt onset of fever, pharyngitis, and conjunctivitis; hyperplasia of the tonsillar, adenoidal, and pharyngeal lymphoid tissue is present. Granular conjunctivitis is present in one or both eyes. A study from Spain showed that 32% of children with positive pharyngeal cultures for adenovirus had cervical node enlargement.16 Preauricular adenopathy is surprisingly infrequent. Generalized adenopathy is present in 10% to 20% of patients with pharyngoconjunctival fever, and hepatosplenomegaly is common. Preauricular adenitis occurs in 90% of patients with epidemic keratoconjunctivitis. Half of patients with the usually unilateral, acute, follicular conjunctivitis also have pharyngitis and rhinitis. Acute respiratory disease due to adenoviruses generally has significant constitutional as well as localized respiratory symptoms. Bilateral cervical adenopathy is almost always present.

Enteroviruses Nonspecific febrile illness due to coxsackieviruses and echoviruses generally has an abrupt onset. Manifestations include fever, malaise, sore throat, nausea, vomiting, and abdominal pain. Minimal conjunctival and pharyngeal erythema with bilateral cervical adenopathy is present.

Human Herpesvirus 6 and 7 (Roseola) In one reported series of human herpesvirus 6 (HHV-6) infection in 688 children, mild bilateral cervical adenopathy was present in 31% of cases. Typically, the occipital, posterior auricular, and posterior cervical nodes are involved, but enlargement is modest.17HHV-7 has been reported to cause roseola; however the complete clinical picture of disease has not been defined.18

Human Herpesvirus 8 HHV-8 is generally thought of as the causative agent of Kaposi sarcoma in patients with acquired immunodeficiency syndrome (AIDS). However, HHV-8 was associated with atypical lymphocytosis/ mononucleosis in 3 children reported from China.19 One child was previously healthy and had 8 days of fever and sore throat, morbilliform rash, and cervical lymphadenopathy.

Corynebacterium diphtheriae Cervical adenitis is variable in tonsillar and pharyngeal diphtheria. In some cases, it is associated with edema of the soft tissues of the neck (“erasive edema”) so severe as to appear as to cause a bull-neck appearance. In other cases, lymphadenitis is minimal.

Acute Unilateral Lymphadenitis Table 19-3 summarizes clinical clues to and diagnostic tests for selected causes of unilateral cervical lymphadenitis.

Staphylococcus aureus and Streptococcus pyogenes Staphylococcus aureus or Streptococcus pyogenes infection accounts for 40% to 80% of cases of acute unilateral cervical lymphadenitis.3,4 Typically, the patient is a 1- to 4-year-old child with a history of recent upper respiratory tract symptoms (sore throat, earache, coryza) or impetigo and signs of pharyngitis, tonsillitis, or acute otitis media. Few clinical findings distinguish streptococcal from staphylococcal infection. The infected nodes, usually the submaxillary or superior deep cervical, are 2.5 to 6 cm in diameter and moderately to intensely tender, and often, the overlying skin is erythematous and warm. Systemic symptoms may be associated with cellulitis. Bacteremia or metastatic foci of infection can occur. Occasionally, suppuration and periadenitis are so severe that torticollis is present and individual nodes cannot be palpated. Systemic symptoms such as high fever, toxicity, tachycardia, and flushed facies may be present. In up to a third of cases of staphylococcal or streptococcal acute unilateral lymphadenitis that come to medical attention, lymph node suppuration and fluctuation develop. Infection caused by Staphylococcus aureus tends to have a longer duration of disease before diagnosis, a higher likelihood of suppuration, and slower resolution (Figure 19-2). The role of S. aureus as a sole cause of acute unilateral lymphadenitis has been questioned. In a California study from 1979, node aspirates from 65% of the patients yielded pure growth of S. aureus; 41% of these patients exhibited an immune response to one or more of the extracellular antigens of Streptococcus pyogenes.21 Additionally, it has been observed that many children improve clinically with penicillin or ampicillin therapy alone, which would not be effective against b-lactamase-producing Staphylococcus aureus. The resurgence of S. aureus soft-tissue disease that has been seen in the

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Cervical Lymphadenitis and Neck Infections

CHAPTER

19

147

TABLE 19-3. Clinical Clues and Diagnostic Test(s) for Selected Causes of Unilateral Cervical Lymphadenitis Organisms

Clinical Clues

Diagnostic Test(s)

Staphylococcus aureus Streptococcus pyogenes

Acute unilateral adenitis; ages 1–4 years; 2–6-cm nodes, frequent associated cellulitis; 25–33% become fluctuant For Staphylococcus aureus, suppuration is rapid; without therapy, 85% of nodes suppurate

Throat culture for Streptococcus pyogenes; needle aspirate of node for Gram stain and culture

Group B streptococcus

Ages 2–6 weeks; cellulitis–adenitis syndrome; facial or submandibular cellulitis; ipsilateral otitis media

Blood culture and aspirate of node or soft tissue for Gram stain and culture CSF evaluation if cellulitis–adenitis syndrome

Anaerobic bacteria

Older children; caries; periapical or periodontal disease present

Needle aspiration of node for Gram stain and culture; blood culture

Bartonella henselae

Most common cause of chronic unilateral adenopathy; contact with kittens; oculoglandular syndrome; inoculation papule present

Serologic analysis; culture of tissue with granulomas with stellate abscesses on biopsy

Nontuberculous Mycobacterium

Age 1–5 years; no systemic symptoms; no exposure to TB; unilateral submandibular node: normal chest radiograph and ESR; PPD usually < 15 mm

Characteristic granulomatous reaction; culture of excised node

Mycobacterium tuberculosis

Age > 5 years; systemic symptoms with history of exposure to TB; bilateral lower cervical node involvement; elevated ESR

Positive PPD and chest radiograph; culture of gastric aspirate

Toxoplasma gondii

Fever, fatigue, myalgia, sore throat; discrete < 3 cm localized anterior cervical, suboccipital, or supraclavicular nodes; ± mediastinal adenopathy

Serologic analysis

CSF, cerebrospinal fluid; ESR, erythrocyte sedimentation rate; PPD, purified protein derivative; TB, tuberculosis.

last few years would argue for the organism as a sole cause. Recent literature reports a predominance of S. aureus over Streptococcus pyogenes when etiology is confirmed; however recent serologic studies have not been performed.5,6,22

Streptococcus agalactiae (Group B Streptococcus) Typically, the young infant with group B streptococcal cellulites– adenitis (submandibular/cervical) syndrome has abrupt onset of fever, poor feeding, and irritability associated with unilateral nondiscrete facial or submandibular swelling that is erythematous and tender. Male predominance has been described.23 Bacteremia is usually present, and the organism is isolated from the aspirate of cellulitis or a lymph node. Ipsilateral otitis media is common. Meningitis is present in as many as 24% of infants with cellulites–adenitis.24 Suppurative submandibular lymphadenitis caused by Staphylococcus aureus is sometimes distinguishable by manifestation as a discrete mass and propensity for suppuration.

Anaerobic Bacteria In older children with the acute onset of unilateral adenitis, an anaerobic infection secondary to periodontal or dental abscesses should be considered. Such infection can lead to septic thrombophlebitis of jugular veins, septic pulmonary emboli, and central nervous system infections (Lemierre syndrome; see Chapter 27, Infections of the Oral Cavity; Chapter 30, Infections Related to the Upper and Middle Airways). Clues are dental disease, bull neck, severe inflammatory response, systemic toxicity, and several positive blood cultures with isolation of Fusobacterium spp. or viridans streptococci. In one study utilizing optimal culture techniques for anaerobic bacteria, 38% of aspirates from acutely inflamed single neck nodes in children 2 to 16 years of age contained anaerobic bacteria.7 Such findings were associated with dental cavities and dental abscesses. Most studies report anaerobes in < 5% of cases.

Figure 19-2. Staphylococcus aureus neck abscess in a toddler.

148

SECTION

B Cardinal Symptom Complexes

Francisella tularensis

Subacute or Chronic Unilateral Lymphadenitis

In one series of children with tularemia, lymphadenopathy (especially of multiple cervical nodes) was the most common manifestation, being present in 27 of 28 patients; in 4 children, cervical nodes were the only enlarged modes.25 Cervical lymphadenopathy alone or concurrent with other regional adenopathy was common. One-third of the children had late suppuration after antibiotic therapy. In untreated cases, the skin over the involved nodes is inflamed, and about 50% of nodes suppurate and drain. In the remainder of cases, the nodes remain firm, enlarged, and tender for several months. An outbreak occurred in Kosovo in 2000 in which 327 cases of tularemia pharyngitis and cervical lymphadenitis were confirmed by serology. The source was thought to be contaminated food or water with a rodent source as environmental conditions were poor.26 In oculoglandular tularemia, the eyelids become edematous, and the conjunctivae are inflamed and painful, with nodules and ulcers (Figure 19-3). Preauricular, submaxillary, and cervical nodes are enlarged, tender, and painful.

Mycobacterial infections, cat-scratch disease, and toxoplasmosis are most commonly associated with subacute lymphadenitis. Typical findings are painless or minimally tender, unilateral, cervical or submaxillary node swelling. If the lymph nodes enlarge and the disease progresses, the overlying skin becomes taut and effaced, with pinkish discoloration (although there is no increase in skin temperature). Skin becomes attached to the underlying mass. If the infection is untreated, fluctuation and spontaneous drainage (with or without formation of sinus tracts) usually follow. Suppuration from toxoplasmosis is rare.

Pasteurella multocida After an animal bite or scratch on the head, neck, or upper chest, acute unilateral cellulitis due to P. multocida can be associated with tender cervical lymphadenitis.

Yersinia pestis Fleas from wild mammals, cats, and dogs in the western United States serve as the vectors for Y. pestis (bubonic plague). Cervical adenopathy is the third most common site of involvement, because bites occur more commonly on the extremities than around the head and neck. Organisms are carried in lymphatics to the nearest node, usually causing acutely enlarged, edematous, exquisitely tender, unilateral lymphadenitis with overlying erythema (bubo). Fever and other systemic manifestations accompany the appearance of buboes. Untreated, the nodes suppurate and ulcerate.

Unusual Organisms Rarely, gram-negative bacilli, Streptococcus pneumoniae, group C streptococci, Nocardia brasiliensis, Yersinia enterolitica, Staphylococcus epidermidis, or a-hemolytic streptococci are isolated from a node. Immunodeficiencies such as Job syndrome and chronic granulomatous disease should be considered if a catalase-producing grampositive or gram-negative organism is causative.

Figure 19-3. Ulceroglandular tularemia. Note adenopathy relative to papular lesion at the site of the tick bite.

Cat-Scratch Disease Cat-scratch disease is regional lymphadenitis that follows inoculation of B. henselae into damaged skin or mucosal membrane. The most common sites of lymphadenopathy are the axilla (52%) and neck (28%), presumably from scratches on the extremities and cuddling an animal, respectively. Classically, adenopathy begins 5 days to 2 months after inoculation (bite, scratch, exposure of mucous membrane). Generally, the affected node is solitary, often > 4 cm in diameter, and tender. Constitutional symptoms are usually mild and include fever in up to 25% of cases. Overlying skin is not red or warm, but suppuration occurs in 30% to 50% of patients brought to medical attention. The oculoglandular syndrome of Parinaud was initially described in catscratch disease.12 Supraclavicular nodes can be involved from scratches on the neck or upper chest. Nonsuppurative nodes diminish in size after 4 to 6 weeks. Treatment probably aborts suppuration of larger nodes (see Chapter 160, Bartonella Species).

Toxoplasmosis In the 10% of patients with acquired toxoplasmosis who are symptomatic, lymphadenopathy and fatigue without fever are the most common manifestations.27 The nodes most commonly involved are the cervical, suboccipital, supraclavicular, axillary, and inguinal. The involved nodes are discrete, may or may not be tender, and do not suppurate. Adenopathy is localized or involves multiple areas.

Nontuberculous Mycobacterium (NTM) Infection due to NTM occurs in children 1 to 4 years old. The risk is higher in whites, in rural areas, and in the upper midwest and southeast United States, but the disease is not uncommon outside these settings. Organisms are ubiquitous in soil and are probably ingested; infection is localized to a submandibular or single tonsillar node. Bilateral involvement is rare. A recognizably enlarged node is usually the superficial marker of a large, deeper cluster. Initial appearance can be rapid (over 24 hours), with gradual increase in node size over 2 to 3 weeks. Most enlarged nodes are 3 cm or less in diameter, although enlargement to > 5 cm can occur. Pain, tenderness, and constitutional illness are minimal. Approximately 50% of children with recognized lymphadenitis have fluctuant lesions, and in 10%, spontaneous drainage and sinus tract formation occur. The skin changes from pink to a distinctive lilac red, with the overlying skin developing a very thin, parchment-like quality. In some cases, fluctuation without skin changes develops. Signs and symptoms of Mycobacterium tuberculosis adenitis and nontuberculous mycobacterial adenitis are identical. Other clinical and epidemiologic features are distinctive. Chest radiograph is normal in children with nontuberculous mycobacterial infection, and an intermediate Mantoux tuberculin skin test results in < 15 mm induration (usually 5 to 9 mm). NTM lymphadenitis in children in the United States is most often caused by the M. avium complex. Less commonly, it is caused by other mycobacterial species, including M. scrofulaceum, M. kansasii, and M. haemophilum. M. bovis adenitis is seen in children who consume unpasteurized dairy products from Mexican cattle. M. bovis can be misdiagnosed in the laboratory as M. tuberculosis as they are both

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Cervical Lymphadenitis and Neck Infections

considered in the M. tuberculosis complex. This can be a pitfall for optimal therapy as resistance patterns are different.28 Therapy with antimicrobial agents active against the M. avium complex may change the management and outcome of this infection, which otherwise is surgical excision.29–31

Mycobacterium tuberculosis Striking enlargement of the superficial regional lymph nodes is an integral part of the primary tuberculous complex. Involvement of the cervical nodes is most often the result of extension from the paratracheal nodes to the tonsillar and submaxillary nodes or from the apical pleurae and upper lung fields by direct spread to the inferior deep cervical (supraclavicular) nodes. Rarely, superficial lymph nodes are enlarged secondary to generalized adenopathy during the course of lymphohematogenous spread of disease (Figure 19-4). Reactivation of quiescent tuberculous infection can manifest initially as localized or generalized adenopathy. When superficial nodes are involved early in the infection, enlargement is usually discrete and painless, and the node is rubbery.32 Bilateral enlargement is the rule, but right-sided involvement can predominate. Acute nontuberculous respiratory tract infections can precipitate or aggravate tuberculous lymphadenitis, resulting in local pain and perilymphadenitis. Rarely, the patient is seen with a fluctuant mass and shiny, erythematous overlying skin. These can rupture and drain chronically. Secondary infection with other pyogenic bacteria can occur. In general, clinical features do not distinguish tuberculous and nontuberculous mycobacterial infections and diagnosis may be difficult if exposure to tuberculosis is suspected.32 Surgical excision is frequently avoided with infection due to M. tuberculosis as it usually extends into the mediastinum and its response to antituberculous therapy is good.

Actinomyces and Nocardia In cervicofacial actinomycosis, Actinomyces israelii causes a chronic, granulomatous suppurative infection of the soft tissues in and around the mandible. Although lymph nodes are not usually involved, chronic

CHAPTER

19

149

tissue induration can mimic chronic lymphadenopathy. Tissue distruction can be considerable without proper therapy. The lymphocutaneous syndrome due to Nocardia spp., involving an initial facial papular lesion followed by fever and an enlarged, tender, unilateral, submaxillary gland, is well described in children.33 N. brasiliensis is the most common species causing the skin and lymph node cervicofacial syndrome. Lymphadenitis was associated with the skin lesions in 58% of children in Texas with nocardiosis.34 This disease can occur in healthy children, and an exhaustive search for an immunodeficiency is not necessary if there is a prompt response to trimethoprim-sulfamethoxazole therapy.

BCG Vaccination In countries where the vaccination against the Calmette-Guérin bacillus (formerly called bacille Calmette-Guérin; hence the abbreviation BCG) is given routinely, painless enlargement of regional nodes associated with vaccination (axillary, rarely cervical) can occur and is presumably due to multiplication of bacilli and formation of granulomas. Calcification and abscess formation with breakdown can occur, particularly in infants.35 Rarely, regional adenitis can develop years after vaccination, particularly if the individual becomes immunocompromised. Treatment is controversial and spontaneous resolution of nonsuppurative lesions generally occurs without antimicrobial therapy in young children.36

NONINFECTIOUS CAUSES OF LYMPHADENOPATHY Kawasaki Disease Lymph node swelling is the least common of the principal diagnostic criteria of Kawasaki disease, occurring in 50% to 75% of patients. Lymphadenitis, one of the earliest manifestations, is usually unilateral and confined to the anterior triangle; the node is only moderately tender. The enlarged node or mass of nodes is usually > 1.5 cm in diameter and nonfluctuant, and there may be overlying erythema. The nodes of Kawasaki disease appear different by ultrasound than those of presumed bacterial lymphadenitis, so this may be a useful tool in differentiation.37 Suppuration does not occur, and resolution usually appears early in the course of disease.

Periodic Fever, Aphthous Stomatitis, Pharyngitis, and Cervical Adenitis A unique syndrome consisting of periodic fever, aphthous stomatitis, pharyngitis, and cervical adenitis (PFAPA) was described in 1987.38 It usually affects children younger than 5 years. Symptoms appear abruptly every 2 to 9 weeks, with spontaneous resolution in 4 to 5 days. Bilateral modest enlargement and tenderness of cervical nodes are typical; other lymph nodes are not affected (see Chapter 17, Prolonged, Recurrent, and Periodic Fever Syndromes).

Sarcoidosis The most common physical finding in children with sarcoidosis is peripheral lymphadenopathy. The involved cervical nodes are bilateral, discrete, firm, and rubbery. More than 80% of children have involvement of the supraclavicular nodes; bilateral hilar adenopathy is almost always present, along with nonspecific symptoms and other multisystem manifestations.

Sinus Histiocytosis with Massive Lymphadenopathy (Rosai–Dorfman Disease)

Figure 19-4. Draining lymph node secondary to Mycobacterium tuberculosis.

Rosai–Dorfman disease is a rare disorder that manifests in the first decade of life with mobile, discretely and asymmetrically enlarged lymph nodes, initially located in the neck.39 With progression, the nodes become massively enlarged bilaterally and are painless and adherent. Other nodal groups (e.g., axillary, inguinal, and hilar nodes)

150

SECTION

B Cardinal Symptom Complexes

and extranodal sites may be involved. Systemic manifestations of fever, leukocytosis, neutrophilia, anemia, an elevated erythrocyte sedimentation rate, and hypergammaglobulinemia may be present. Histopathologic analysis demonstrates florid follicular hyperplasia, marked histiocytic proliferation, and prominent plasmacytosis, in keeping with the hypergammaglobulinemia. Resolution occurs spontaneously after 6 to 9 months, but therapy may be required for extension or progression of disease.

Histiocytic Necrotizing Lymphadenitis (Kikuchi–Fujimoto Disease) Kikuchi–Fujimoto disease is a benign, rare entity of unknown etiology that commonly manifests in older children as bilateral, enlarged, firm, painful cervical nodes, most often in the posterior cervical triangle, that are unresponsive to antibiotic therapy. Perinodal inflammation is common. Associated findings in 50% of patients include skin lesions, fever, leukopenia with atypical lymphocytosis, and an elevated erythrocyte sedimentation rate. Those with prolonged fever are more likely to develop leukopenia.40 Splenomegaly, nausea, weight loss, and night sweats may be present. Nodal histology is characteristic but can be confused with that of malignant lymphoma. No other diagnostic test is available. Most patients resolve symptoms a few days after excisional biopsy. Viral etiology is suspected based on a few reports of serologic studies suggesting EBV, CMV, human T lymphotropic virus (HTLV), HHV-6 and parvovirus B19 infection.41 Kikuchi–Fujimoto disease has also been shown to occur in association with or preceding systemic lupus erythematosus in some cases. The prognosis for spontaneous resolution of Kikuchi–Fujimoto disease in a majority of cases is excellent.42,43

Kimura Disease Kimura disease is a chronic, benign, but potentially disfiguring unilateral, localized cervical lymphadenopathy of Asians, most often males of Chinese or Japanese descent. Insidious onset of painless, subcutaneous nodules overlying the affected lymph nodes occurs. Eosinophilia and a marked increase in serum immunoglobulin E are present. Renal involvement may occur, with proteinuria present in 12% to 16% of cases.44 Resection, irradiation, or corticosteroid therapy can be associated with recurrences.45,46

NECK MASSES NOT INVOLVING LYMPH NODES Tumors of the Head and Neck The differential diagnosis of neck masses includes congenital cysts and sinuses, vascular malformations, salivary and thyroid anomalies, and benign and malignant neoplasms of nonsquamous origin. Malignancy should be considered, particularly in older children with a painless, firm, cervical mass > 3 cm in diameter that is fixed to the deep cervical tissues, has grown rapidly, or is located in the supraclavicular or posterior triangle areas. Such masses may be multiple and may extend across the anterior and posterior triangles. The incidence of all pediatric cancer has increased in the last few decades; however, the incidence of pediatric neck masses over the same timeframe has increased at a greater rate. Teenagers 15 to 18 years old are the most frequently affected (39%), followed by children aged 4 years and younger (27%).47 Fifty percent of masses in the posterior triangle are malignancies, most of lymphoid origin. Fifty percent of all malignant neck masses in children are Hodgkin lymphoma or NHL; neuroblastoma is the next most common, accounting for 15%, followed by thyroid tumors. Softtissue sarcomas, including primary rhabdomyosarcoma and carcinoma of salivary glands and the nasopharynx, also occur. Age is predictive of diagnosis; in children younger than 6 years, the most common malignancies, in order of decreasing frequency, are neuroblastoma, NHL, rhabdomyosarcoma, and Hodgkin lymphoma. In patients 7 to 15 years old, Hodgkin lymphoma and lymphosarcoma occur with

equal frequency, followed by thyroid carcinoma, rhabdomyosarcoma, and parotid adenocarcinoma. Soft-tissue sarcomas manifesting in the neck are primary rhabdomyosarcoma and undifferentiated sarcomas. Rhabdomyosarcoma accounts for 18% of all solid carcinomas in children; of these, 36% manifest in the head and neck region, usually as an asymptomatic mass or as symptoms related to location. Hodgkin lymphoma manifests in 80% to 90% of cases as an asymptomatic cervical or supraclavicular mass. Upper cervical nodes are three to four times more likely to be involved than supraclavicular nodes. NHL is more likely to be associated with systemic symptoms, including fever and leukocytosis with pronounced neutrophilia, and cervical lymph node involvement, which can be unilateral and associated with diffuse swelling. Cervical adenopathy is a primary symptom in only 11% to 35% of patients with Hodgkin lymphoma, and extranodal head and neck masses are unusual.

Congenital Cysts and Sinuses Multiple congenital cysts and sinuses related to embryologic development can manifest as infection or as a sudden mass as if infected.48 Recurrence of a suppurative mass may be an indication of an underlying congenital lesion and should be evaluated.49 Table 19-4 delineates noninfectious causes of neck masses in children.

Branchial Cleft Cysts and Sinuses During the fifth week of embryologic development, four endodermal pharyngeal pouches fuse with four ectodermally derived clefts and are obliterated. Rarely, they are not completely obliterated, and persistence of any one of the four clefts can result in a cystic mass or sinus tract that is prone to infection. Persistence of the first pharyngeal cleft results in a sinus tract opening just behind the ramus of the mandible. Cystic components of the tract, which extends to the external ear canal, are often more extensive than is clinically apparent. Persistence of the second cleft is most common, the opening of which is just anterior to the sternal attachment of the sternomastoid muscle. The unseen tract extends the length of the sternomastoid muscle, beginning at the tonsillar bed. Cysts, with potential for acute enlargement, can occur anywhere along the length of the tract. Anomalies of the third cleft are indistinguishable from those of the second cleft. Because the thymus is also derived from the third cleft, thymic duct remnants can be seen as soft, fluctuant, mobile paratracheal masses low in the neck. Surgical intervention is usually required to remove the sinus tract or cyst.

Midline Sinuses and Cysts Dermoid or epidermoid cysts are embryologic defects in fusion. They occur as painless, soft, fluid-filled midline masses. Thyroid tissue originates embryologically at the base of the tongue and descends the thyroglossal duct to its usual site anterior to the proximal end of the trachea. Although the tract is usually obliterated, cysts can form along its course, appearing as 1-cm midline nodules. Retaining attachment to the base of tongue, such cysts elevate on tongue protrusion. Although elevation on tongue protrusion is not a pathognomonic finding, it is important to determine the presence and position of thyroid tissue, because cysts may contain the only functional gland. Nodules within the thyroid gland have a high incidence of malignancy in children. A thyroid mass in the neonate is most often a congenital goiter or a teratoma.

Lymphangiomas In the sixth embryologic week, clefts develop in the cervical mesenchyma, which subsequently fuse to form lymph channels and lymph nodes. Sequestration, or failure to communicate with the rest of the lymphatic system, leads to tumor-like masses of lymph channels most commonly found along the external jugular chain of lymphatics in the lateral cervical region. Lymphatic masses vary in size from a few

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Cervical Lymphadenitis and Neck Infections

centimeters to massive collarlike lesions with extension into the mediastinum. In the larger lesions, lymphatic channels dilate into cystic spaces and are called cystic hygroma. Most are present posterior

Salivary and Thyroid Gland Masses

Condition

Parotid Gland Masses

Comments

Thyroglossal duct cyst

Most common congenital neck mass; discrete 1-cm midline nodule that may elevate with tongue protrusion; elective excision best; may contain only existing thyroid tissue

Second pharyngeal (branchial) cleft anomaly

Second most common congenital neck mass; anterior to upper or middle one-third of sternomastoid muscle; external sinus opens anterior to sternal head of muscle; tract needs to be excised; cyst associated with lymphatic tissue

Cystic hygroma

Third most common congenital neck mass; occurs along jugular lymphatic chain in posterior supraclavicular fossa; failure of mesenchymal clefts to fuse; varies from few centimeters to massive collar-like lesions; may extend into mediastinum

Dermoid or epidermoid cysts

Midline, deep to mylohyoid muscle; soft, fluid-filled midline mass; may elevate with tongue protrusion; contain caseous material or epithelial debris

NONLYMPHOMATOUS MALIGNANCIES

Neuroblastoma

Second most common malignant neck mass in children (first in younger children)

Thyroid cancer

Third most common neck malignancy (first in 11–18 years); high incidence of malignancy in thyroid nodules; prompt biopsy required

Rhabdomyosarcoma

Most common nasopharyngeal malignancy in children

Nasopharyngeal carcinoma Cervical adenopathy often initial and only symptom; arises most often in fossa of Rosenmüller Parotid tumors

19

Parotid tumors are rare. The most common parotid masses are vascular anomalies, usually capillary hemangiomas and rarely cavernous hemangiomas. Large cavernous hemangiomas can disfigure and destroy tissue, whereas capillary hemangiomas usually resolve without any permanent findings by the age of 4 to 5 years. Parotid gland enlargement in childhood can result from calculus formation or recurrent sialectasia, in which case the parotid gland becomes rubbery, firm, and tender but has no signs of inflammation. Parotid infections are discussed in Chapter 27, Infections of the Oral Cavity.

Thyroiditis Infections of the thyroid gland are rare but potentially life-threatening. Infection can arise via the hematogenous route or directly from an adjacent fascial space infection, from a patent foramen cecum and infected thyroglossal duct cyst, or from anterior esophageal perforation. Suppuration usually results from pyriform sinus fistula or thyroglossal duct remnant. The most common aerobic pathogens are Staphylococcus aureus, group A streptococci, pneumococci, and viridans streptococci. Anaerobic bacteria are also isolated, with the most common being gram-negative bacilli and Peptostreptococcus spp. Other pathogens are Haemophilus influenzae, Eikenella corrodens, Actinomyces spp., Bacteroides and Prevotella spp., with rare reports of enteric gram-negative bacilli, mycobacteria, and fungi.50,51 History and physical exam identify acute onset of fever, chills, sore throat, dysphagia, hoarseness, and an extremely tender, red, warm, and enlarged gland. Involvement can be unilateral or bilateral; results of thyroid function tests are usually normal. Pain can be referred to the ear or chest. Laboratory investigation of thyroiditis should include ultrasonography, radionuclide scan, or computed tomography to detect paratracheal extension; needle aspiration is performed for diagnostic

Vascular anomalies most common; asymptomatic node may be malignant and should be excised

MISCELLANEOUS

Sternocleidomastoid tumor Fibrous mass within muscle; detected at 2–4 weeks of age in 0.4% of infants; head turned away from mass; ipsilateral facial hypoplasia Sinus histiocytosis with Bilateral, painless cervical nodes; generalized massive lymphadenopathy lymphadenopathy may develop; systemic (Rosai–Dorfman disease) manifestations include fever and hypergammaglobulinemia; self-limited Giant lymph node hyperplasia (Castleman disease)

Asymptomatic lymphadenopathy in mediastinum or neck; systemic symptoms include fever and hypergammaglobulinemia; surgical removal curative (see Ch. 20)

Histiocytic necrotizing lymphadenitis (Kikuchi– Fujimoto disease)

Asymptomatic cervical or generalized adenopathy; skin lesions, fever, and leukopenia; spontaneous resolution; recurrences possible

Kimura disease

Benign, chronic, unilateral lymphadenopathy; adjacent subcutaneous nodules, peripheral eosinophilia and increased serum immunoglobulin E, proteinuria and renal involvement possible

a For review of lymphomas, see Chapter 18, Lymphatic System and Generalized Lymphadenopathy; Chapter 20, Mediastinal and Hilar Lymphadenopathy.

151

to the sternomastoid muscle in the supraclavicular fossa. Sixty-five percent are present at birth and gradually increase in size as a result of obstructed flow or bleeding; the remainder almost always appear before age 2 years (Figure 19-5).

TABLE 19-4. Noninfectious Causes of Neck Massesa

CONGENITAL ANOMALIES

CHAPTER

Figure 19-5. Cystic hygroma in a 2-month-old child, infected secondarily with group B streptococcus.

152

SECTION

B Cardinal Symptom Complexes

microbiology. With prompt and aggressive therapy, complete resolution without abscess formation is expected. Subacute (de Quervain) thyroiditis is a poorly defined, self-limited condition characterized by an insidious onset with tenderness and marked induration of the gland. The sedimentation rate is high, and biopsy reveals acute and chronic inflammation with granulomatous changes. In most cases, the cause is a virus, such as mumps virus, influenza virus, enterovirus, adenovirus, echovirus, EBV, or, rarely, St. Louis encephalitis virus.51 Mild and transient manifestations of hypermetabolism and elevated thyroid hormone levels can be present. Differentiation from bacterial thyroiditis can be difficult.

DIAGNOSTIC APPROACH TO CERVICAL LYMPHADENOPATHY The assessment and specific evaluation and management of cervical lymphadenopathy and lymphadenitis are directed by integration of the history and initial examination (Table 19-5). In young infants with acute unilateral involvement and systemic manifestations, causative agents can be expected to be Staphylococcus aureus, group B streptococcus, or gram-negative organisms. An aspirate of the node or cellulitis is obtained for culture in addition to blood. Examination of cerebrospinal fluid should be considered in young infants if the etiology is group B streptococcus. Discrete small cervical nodes in a healthy infant are likely to be normal (see Table 18-2). Other patients may have manifestations that lead to evaluation for congenital infection. The older child with bilateral, soft, discrete, minimally tender, and minimally enlarged nodes high in the neck who has respiratory tract symptoms and fever may have a viral syndrome, streptococcal pharyn-

gitis, or M. pneumoniae infection. In roseola, rubella, adenoviral, or enteroviral infections, cervical adenopathy is frequently in the suboccipital, posterior auricular, or superficial posterior cervical areas (Table 19-6). If nodes are large and minimally tender, with no overlying cellulitis, mononucleosis due to EBV or CMV is likely. For protracted or unusual cases, sarcoidosis and Kikuchi–Fujimoto and Rosai–Dorfman diseases are considered. When acute unilateral or bilateral adenopathy is associated with an obvious site of infection, diagnostic studies are focused on the site. Obvious sites of infection include herpes simplex ulcerations, infected abrasions, scalp infections, animal bites or scratches, and conjunctival granulomas. Acutely inflamed, unilateral, tender large nodes with associated periadenitis and cellulitis, fever, and constitutional symptoms, often with poor node definition and torticollis, are usually due to S. aureus or Streptococcus pyogenes infection in the young child (see Tables 19-3 and 19-5). In older children and adolescents, a dental focus and anaerobic infection should be considered. Failure of response to antibiotic therapy after 48 hours should result in evaluation for an abscess or more unusual cause or causative agent, such as Kawasaki disease, and infected or uninfected congenital cysts and sinuses. A recent study of factors predictive of need for surgical drainage analyzed 284 children admitted to a tertiary care hospital with unilateral adenitis.52 Inclusion criteria were 10 days of unilateral lymph node swelling 2.5 cm. Cases were excluded if noninfectious etiology was found, or if nodes were present at > 1 site, or if the patient had prior history of node inflammation, presence of immune deficiency, cancer, or chronic underlying disease. Surgical drainage was required in 21% of patients for which abscesses were confirmed.

TABLE 19-5. Mode of Onset and Characteristics of Lymph Nodes Infected by Selected Organisms Mode of Onset Pathogen

Characteristics

Acute

Subacute/Chronic

Suppuration

Cellulitis

Generalized Adenopathy

++++ ++++ +++ – +++ ++ +++ ++++ + + +++ + + ++ –

– – – ++++ ++ ++ – – ++++ ++++ – ++ + ++ ++

++++ +++ +++ +++ +++ +++ + ++++ +++ +++ + ++ – ++ ++

++++ ++++ ++ + +++ + +++ ++ + + +++ – – – –

– + – + – + – – – +++ – – – – –

+++ ++ + ++++ – +++ +++ +++ +++ ++ ++ +++

++ ++ + – ++ – – – – – – –

– – – – – – – – – – – –

– – – ++ – – – – – – – –

++++ +++ + – +++ + +++ +++

++

+++

–

–

+

BACTERIA

Staphylococcus aureus Streptococcus pyogenes Anaerobic bacteria Bartonella henselae Francisella tularensis Yersinia enterocolitica Pasteurella multocida Yersinia pestis Nontuberculosus Mycobacterium Mycobacterium tuberculosis Group B streptococcus Calmette-Guérin bacillus Mycoplasma pneumoniae Nocardia brasiliensis Actinomyces israelii VIRUSES

Epstein–Barr Cytomegalovirus Hepatitis A, B Herpes simplex Human immunodeficiency Adenovirus Rubella Rubeola Influenza Human herpesvirus 6 Enterovirus Parvovirus B19 Protozoa Toxoplasma gondii

++++, characteristic; +++, frequently associated; ++, occasionally associated; +, rarely associated; –, not known to be associated.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

– + +++

Cervical Lymphadenitis and Neck Infections

CHAPTER

19

153

TABLE 19-6. Associations of Selected Organisms With Involvement of Specific Lymph Node Groups in the Neck Lymph Node Group Preauricular

Postauricular

Occipital

Submaxillary

Tonsillar

Anterior Cervical

Posterior Cervical

Superior Deep Cervical

Supraclavicular

– – – – – ++ – ++ ++++ –

– – – – – – – – – –

++ ++ – – – – – – – –

++ ++ ++++ + ++ ++ +++ +++ + ++++

++++ ++++ ++ +++ ++ + +++ +++ +++ +

+++ +++ ++ +++ ++ ++ – – +++ –

– – – – ++ ++ – – ++ –

+++ +++ – ++ – + – + + –

+ + – – – – +++ – + –

– – – – – ++++ ± ++

– ++ +++ ++ – – – –

– ++ +++ ++ ++ – – –

+++ – – – ++ – – +

+++ – – – ++ – – +

+ – – ++ ++ – ++ +

+ – ++ ++ ++ – ++ –

+ – – – + – – +

– – – – – – – –

– +

– –

– –

– –

– –

– –

+++ +++

– –

– –

– – –

+ – –

++++ – –

– – –

– – –

– ++ –

+ – –

– + –

– +++ +++

Pathogen BACTERIA

Staphylococcus aureus Streptococcus pyogenes Anaerobic bacteria Arcanobacterium haemolyticum Mycoplasma pneumoniae Francisella tularensis Mycobacterium tuberculosis Nontuberculous Mycobacterium Bartonella henselae Nocardia brasiliensis VIRUSES

Herpes simplex Rubeola Rubella Human herpesvirus 6 Epstein–Barr Keratoconjunctival fever Pharyngoconjunctival fever Parvovirus B19 Protozoa Toxoplasma gondii Trypanosoma cruzi FUNGI

Tinea capitis Histoplasma capsulatum Coccidioides immitis

++++, characteristic site of adenopathy; +++, additional site, often associated; ++, occasional site; +, rare site; –, not described.

Staphylococcus aureus was most commonly identified (70%). Factors that were predictive of surgical drainage using multivariate analysis included age < 1 year and node abnormality > 48 hours prior to admission. The unilateral, asymptomatic, initially noninflamed, nontender, firm node of acute or subacute onset is the most challenging problem. Some such nodes become fluctuant and develop the characteristic findings of cat-scratch disease or nontuberculous mycobacterial disease. Many children, however, remain asymptomatic, and the nodes may or may not change in size over time. It is important to identify toxoplasmosis, tuberculosis, and the lymphoma or other malignancies as soon as possible to initiate appropriate therapy if indicated. Accurate measurements of the size and consistency of such nodes should be performed once or twice a week in order to determine whether malignancy should be considered. The presence of systemic symptoms, hepatosplenomegaly, pneumonia, or mediastinal adenopathy (see Chapter 20, Mediastinal and Hilar Lymphadenopathy) is suggestive of tuberculosis or malignancy. Guidelines for lymph node biopsy to rule out a malignancy are provided in Box 19-2. If an excisional biopsy is performed and nonspecific hyperplasia is reported, the patient must be followed carefully, because 25% of patients with such findings are ultimately found to have a lymphoreticular malignancy.

Needle Aspiration Needle aspiration of the infected node is a valuable diagnostic tool for acute and subacute lymphadenitis and can also be therapeutic. In 60% to 90% of patients with acute cervical lymphadenitis in whom needle aspiration is performed for bacterial and mycobacterial culture, an etiologic agent is recovered.3,4,9,53 The procedure is safe and easy to

BOX 19-2. Settings that Prompt Early Consideration of Biopsy of Enlarged Cervical Lymph Node to Rule out Malignancy SITE Supraclavicular Posterior cervical (particularly with extension across sternomastoid muscle to involve anterior and posterior triangles) Deep to fascia SIZE > 2 cm diameter Continued enlargement after 2 weeks No significant decrease in size after 4–6 weeks Not normal in size after 8–12 weeks CONSISTENCY No inflammation Nontender (unless rapidly enlarging) Firm, rubbery, or matted (may be fixed to underlying structures) Ulceration POTENTIAL LOCAL OR SYSTEMIC SPREAD EVIDENCED BY Mediastinal or generalized adenopathy Bone marrow involvement Fever, weight loss, hepatosplenomegaly Recurrent epistaxis, progressive nasal obstruction, facial paralysis, or otorrhea NO EVIDENCE TO SUPPORT AN INFECTIOUS ETIOLOGY Negative results of evaluation for infectious causes No response to empirically chosen antibiotics

perform and has no serious complications. Controversy exists as to whether a persistently draining fistula is precipitated by aspiration if the lymphadenitis is due to mycobacterial infection. The preponderance of data suggests that aspiration is not contraindicated and, in

154

SECTION

B Cardinal Symptom Complexes

fact, can direct definitive therapy. As tuberculous and nontuberculous diseases are difficult to distinguish clinically, needle aspiration is a useful diagnostic tool for these cases if M. tuberculosis exposure is possible. Laboratory polymerase chain reaction diagnosis is also being evaluated as a diagnostic tool using needle aspiration samples.54,55 The specimen is obtained from the largest and most fluctuant node. If no material is aspirated, saline is injected into the node and repeat aspiration performed. Gram and acid-fast stain should be performed, and the specimen should be inoculated on to media for growth and isolation of anaerobic and aerobic bacteria, fungi, and mycobacteria. If no visible specimen is obtained, the needle and syringe can be flushed into a blood culture flask.

Lymph Node Biopsy If the diagnosis remains in doubt and the lymph node fails to regress, enlarges, is hard, or is fixed to adjacent structures, biopsy should be performed. An accelerated schedule for biopsy is appropriate in the presence of persistent fever or weight loss (in the absence of confirmed diagnosis) or if firm lymph nodes are found in the posterior cervical triangle or supraclavicular area (if M. tuberculosis and B. henselae are excluded). These settings are associated with greater possibility of malignancy. Excisional biopsy is preferred to incisional biopsy of lymph nodes, although there are a few retrospective studies that support the use of fine-needle aspiration as an initial diagnostic mechanism.56,57 Yield from biopsy is maximized through careful choice of tissue site and proper handling of resected tissue. Generally, the largest, firmest nodes should be excised. Whenever possible, more than one node should be obtained. When several groups of nodes are involved, biopsy specimens taken from the lower neck and supraclavicular area (or a region other than the neck) have higher diagnostic yield than those taken from high cervical nodes. If EBV or toxoplasmosis is expected, biopsy is avoided if possible because of the difficulty of discrimination of such diseases from malignancy and the ability to confirm either diagnosis by serologic testing. If lymphoma is suspected, needle biopsy or frozen section is unreliable for excluding the diagnosis. When nontuberculous mycobacterial infection is suspected and surgery is required for diagnosis and therapy, excisional dissection of deep nodes may be necessary. The individual situation dictates specific evaluations of excised tissue, but evaluation commonly consists of special stains, including Giemsa, periodic acid–Schiff (PAS), methenamine silver, and Warthin–Starry silver stains and, in selected cases, in vitro cultures or electron microscopy. Culture for facultative and anaerobic bacteria, mycobacteria, and fungi is appropriate. Many reactive processes simulate malignant histology and make definitive diagnosis impossible. Examples are rheumatoid arthritis, toxoplasmosis, phenytoin-induced adenopathy, dermatopathic adenitis, and EBV mononucleosis. A thorough history and a second series of evaluations, including serologic tests, are essential for making a correct diagnosis of these reactive processes.

therapy. Antimicrobial resistance, especially CA-MRSA, should be considered if empiric antimicrobial therapy fails. A culture should be obtained from suppurative material to aid in therapy whenever possible. If a primary focus of infection is identified elsewhere, the mode and duration of therapy are directed in reference to that site as well. Impetigo and other skin infections are usually caused by Staphylococcus aureus and Streptococcus pyogenes. Dental or periodontal disease leads to lymphadenitis due to anaerobic bacteria. Aspiration of the inflamed site is performed to guide therapy if moderate to severe adenitis with cellulitis and constitutional symptoms is present or if no improvement occurs after 48 to 72 hours of therapy. In the latter instance, additional diagnoses are considered, including viral infection, Kawasaki disease, nontuberculous mycobacterial infection, and more unusual causes of adenitis. If cat-scratch disease is suspected, azithromycin, trimethoprim-sulfamethoxazole, or rifampin is given if a node is > 3 cm, in an attempt to avoid abscess formation. In one placebo-controlled study, azithromycin decreased the volume of the lymph node faster than placebo, but there was no difference in the two groups after 1 month.58 Oral ciprofloxacin and parenteral gentamicin have also been associated with resolution of superficial lymphadenitis or visceral cat-scratch disease. Routine antituberculous therapy is generally ineffective for nontuberculous mycobacterial infections. In cases in which removal of the node is not advised (for example, because of size or proximity to the facial nerve), antituberculous therapy (clarithromycin or a combination of drugs), which has occasionally been effective, can be given. A study from Dallas shows that antibiotic therapy alone was curative in 67% of NTM lesions. Of the remaining patients, 40% had some response to antibiotics prior to surgery.59 Patients with marked lymph node enlargement, moderate to severe systemic symptoms, and cellulitis frequently require parenteral therapy for the first 3 days. This route provides a higher concentration of drug in inflamed tissue, halts bacteremia if present, and may promote more rapid localization. Although a decrease in fever, inflammation, and tenderness is expected within 48 to 72 hours after appropriate treatment of bacterial lymphadenitis, lymph node regression may be slow, requiring 4 to 6 weeks, and suppuration despite appropriate therapy is not unusual.

Abscess Drainage Ultrasonography or computed tomography can be helpful in cases in which abscess is suspected in the absence of fluctuation.53 For abscesses due to Staphylococcus aureus or Streptococcus pyogenes infection, incision and drainage are the procedures of choice. Needle aspiration may be therapeutic. For abscesses from suspected mycobacterial or Bartonella infection, aspiration (with an 18- or 19gauge needle) is initially preferred to avoid fistula formation. Installation of saline may facilitate aspiration when pus is very thick. Surgical excision is required infrequently for cat-scratch disease.

Surgical Excision MANAGEMENT Initial Therapy Children with bilaterally enlarged cervical lymph nodes (< 3 cm in size) that are not erythematous or exquisitely tender are observed without evaluation or therapy. If the initial findings are typical of acute bacterial lymphadenitis (unilateral node, > 2 to 3 cm, erythematous, and tender) without systemic symptoms, empiric therapy is given for Staphylococcus aureus and Streptococcus pyogenes infection unless the course or findings are typical for S. pyogenes (Table 19-7). Therapy is usually given for 10 days or at least 5 days beyond resolution of acute signs and symptoms, whichever is longer. It is useful to know local antiobiogram data when choosing empiric

Surgical excision is the most effective treatment for nontuberculous mycobacterial cervical adenitis.60,61 Excision of the largest and necrotic mass or masses is usually adequate, because the remaining adenopathy resolves spontaneously over ensuing months. Excision of all affected nodes (especially as documented by imaging studies) requires extensive and difficult surgery, and is not warranted as a first procedure and antimicrobial therapy may be considered. When infection has ruptured into soft tissue, resection of inflamed overlying skin and sinus tracts, followed by primary closure without leaving a drain, is advocated. A decision to begin therapy for M. tuberculosis infection is based on the clinical setting and tuberculin skin test results. The response of M. tuberculosis infection to antituberculous therapy is usually rapid, with resolution of symptoms and marked regression of the lymph nodes within 3 months.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Mediastinal and Hilar Lymphadenopathy

CHAPTER

20

155

TABLE 19-7. Treatment of Acute Bacterial Cervical Lymphadenitis Presumed Pathogen and Setting

Empiric Therapy

Route

No prominent systemic symptoms, cellulitis, or suppuration

Amoxicillin-clavulanate, cephalexin, or clindamycin

PO

With cellulitis, marked enlargement, moderate or severe symptoms

Nafcillin, cefazolin, or clindamycin; needle aspiration if no improvement after 48 hours; consider MRSA

IV

With suppuration

Parenteral antibiotics as above and incision and drainage

IV

Nafcillin or oxacillin

IV

Clindamycin or penicillin plus metronidazole

PO or IV, depending on severity

Needle aspiration of node; consider use of azithromycin, rifampin, TMP-SMX, or gentamicin

PO, IV

TB ruled out; node not draining

Excision of node

–

MYCOBACTERIUM TUBERCULOSIS

See Ch. 134

STAPHYLOCOCCUS AUREUS OR STREPTOCOCCUS PYOGENES

GROUP B STREPTOCOCCUS OR STAPHYLOCOCCUS AUREUS

Infant < 2 months with cellulitis with or without systemic symptoms or adenitis ANAEROBIC BACTERIA

Associated dental or periodontal disease, bull neck BARTONELLA HENSELAE

Fluctuant or draining node; systemic manifestations NONTUBERCULOUS MYCOBACTERIUM

FRANCISELLA TULARENSIS

Acute onset, fever; diagnosis reasonably certain

Streptomycin (gentamicin)

IM, IV

Amoxicillin-clavulanate or nafcillin until diagnosis confirmed; then penicillin

PO, IV

PASTEURELLA MULTOCIDA

Onset within 24 h of animal bite

IM, intramuscular; IV, intravenous; MRSA, methicillin-resistant Staphylococcus aureus; PO, oral; TB, tuberculosis; TMP-SMX, trimethoprim–sulfamethoxazole.

COMPLICATIONS Complications of pyogenic neck infections, such as mediastinal abscess, purulent pericarditis, thrombosis of the internal jugular vein, and pulmonary emboli or disseminated septic emboli, have become uncommon since the availability of effective antimicrobial therapy. Rapid progression or failure to treat adenitis due to Staphylococcus aureus or Streptococcus pyogenes can result in cellulitis, bacteremia, septicemia, or toxin-related symptoms. Poststreptococcal acute glomerulonephritis occurs occasionally.

CHAPTER

20

Mediastinal and Hilar Lymphadenopathy Mary Anne Jackson, P. Joan Chesney, and Sarah J. Fitch

ANATOMY OF THE MEDIASTINUM The mediastinum is the space between the pleural cavities that contains the heart and all chest viscera except the lungs. Consisting of loose areolar tissue and organs, it is more a potential space than an actual body cavity. It is bounded laterally by the parietal pleurae, anteriorly by the sternum, posteriorly by the ribs and paravertebral gutters, superiorly by the thoracic inlet, and inferiorly by the diaphragm. The anterior mediastinum contains everything anterior and

superior to the heart, including the thymus, aortic arch and its major branches, innominate veins, and lymphatic tissue (nodes and vessels). Most masses here are of thymic origin, teratoma, lymphoma, or angiomatous tumor. The middle mediastinum is triangular with the apex at the fourth thoracic vertebra. It contains the heart, pericardium, trachea, pulmonary hilar and mediastinal lymph nodes and vessels, and phrenic and vagus nerves. Masses in this compartment are usually infectious or malignant lesions of lymph nodes. The posterior mediastinum extends from the first rib to the diaphragm and behind the heart and lung roots. It contains the esophagus, descending aorta, paravertebral lymph nodes, lower portion of the vagus nerve, and sympathetic nerve chains. Neurogenic tumors and duplication cysts are the most common lesions encountered. Although the delicacy of mediastinal tissue planes offers little resistance to the spread of disease between compartments, tumors and inflammation tend to extend within compartments. Infections within the mediastinum are relatively uncommon, but their proximity to many vital structures makes accurate diagnosis essential.

LYMPHATIC DRAINAGE OF THE LUNGS AND PLEURA As shown in Figure 20-1, lymph from the thoracic viscera (heart, pericardium, lungs, pleura, thymus, and esophagus) traverses one of the three possible sets of nodes before entering the thoracic duct or right lymphatic duct. Anterior mediastinal nodes are located anterior to the aortic arch, innominate veins, and large arterial trunks leading from the aorta. They receive afferents from the thymus and pericardium, the sternal nodes, and the thyroid gland. Posterior mediastinal nodes lie dorsal to the pericardium and adjacent to the esophagus and descending aorta. They receive afferents from the esophagus, dorsal pericardium, diaphragm, and convex surface of the liver. Middle or mediastinal nodes drain the lungs and pleura. Lymphatic drainage of the lungs is composed of superficial and deep plexuses.

156

SECTION

B Cardinal Symptom Complexes TABLE 20-1. Lymphatic Drainage of the Lung and Pleura Lymph Node Group

Areas Drained

Intrapulmonary

Pulmonary vessels, bronchi, and parenchyma

Hilar (bronchopulmonary)

Intrapulmonary nodes and superficial plexus of visceral pleural lymphatics

Tracheobronchial: superior and inferior (subcarinal) Right side All hilar nodes of right lung and hilar nodes of left lower lobe and lingula Lymph drainage patterns Inferior cervical

Deep scalene

Right jugular lymph trunk Rt. subclavian vein

Thoracic duct

Left upper lobe hilar nodes

Paratracheal nodes Right side

Right tracheobronchial nodes

Left side

Left tracheobronchial nodes

Supraclavicular nodes Right side Left side (Virchow node)

Aorta

Paratracheal Superior tracheobronchial

Lt. subclavian vein

Pulmonary veins Hilar Pulmonary

Left side

Subcarinal or tracheobronchial

Pulmonary veins

Bronchopulmonary or hilar Figure 20-1. Lymphatic drainage of the thorax.

The superficial plexus lies beneath the visceral pleura. Lymph flows around the border of the lung to enter the bronchopulmonary (hilar) nodes. The deep plexus accompanies branches of the pulmonary vessels and ramifications of the bronchi throughout the lungs. Lymphatic drainage of the lung passes through four sets of lymph nodes (Table 20-1). Intrapulmonary lymph nodes are located within the lung, chiefly at the bifurcations of the larger bronchi. Bronchopulmonary or hilar nodes are located at the pulmonary hilus at the site of entry of the main bronchi and vessels. Tracheobronchial nodes are divided into superior and inferior groups. The superior group lies in the obtuse angle between the trachea and bronchi on both sides. The inferior, or subcarinal, group lies under the carina at the tracheal bifurcation. The fourth group, the tracheal or paratracheal nodes, lies beside and somewhat anterior to the trachea throughout its course. A fifth group of lymph nodes of importance in the drainage of the lungs is the inferior deep cervical (scalene or supraclavicular) chain, which is located over the lower portion of the internal jugular vein, just above the clavicle and usually under the scalenus anterior muscle. The apical pleurae drain directly to these deep cervical nodes, as do the paratracheal chains. A finding of supraclavicular lymphadenopathy should lead to investigation for intrathoracic or intra-abdominal pathology. Ultimately, all lymph from the lungs and pleurae reaches the tracheobronchial and paratracheal lymph nodes. As a general rule, lymph from the lungs flows from left to right, a probable explanation for the pre-eminence of right upper paratracheal and supraclavicular lymphadenopathy in infectious pulmonary processes, particularly tuberculosis. Lymph from the left lower lobe (and usually also the lingula) flows from the hilar nodes to the lower tracheobronchial nodes, and then to the right paratracheal nodes. Lymph from the right hilar nodes travels to the right paratracheal nodes (see Table 20-1).

Right apical pleura, right paratracheal nodes, head, neck, arms, and upper thorax Left apical pleura, left paratracheal nodes, intra-abdominal nodes, head, neck, arms, and upper thorax

Lymph vessels from the paratracheal nodes join with lymph trunks from the anterior mediastinum to form the right and left bronchomediastinal trunks. These trunks then join with the lymphatic trunks from the supraclavicular nodes to form the right lymphatic duct and left thoracic duct.

EPIDEMIOLOGY At least one-third of all mediastinal masses occur in children younger than 15 years; one-half of such masses are symptomatic. In children, 50% of those who present with symptoms of airway compression have malignant tumors, whereas 90% of those with noncompressive symptoms associated with mediastinal masses are associated with nonmalignant etiology. The greater proportion of symptomatic masses in children compared with adults may be due to smaller thoracic size, resulting in symptoms of compression, or to a higher frequency of malignant lesions.1 The overall incidence of tumors (excluding metastatic disease) and cysts in the mediastinum is 1 per 100 000 people. Whereas mediastinal lymphadenitis is generally thought of infrequently as associated with an infectious etiology, in one large series of biopsies of only the anterior and middle mediastinal compartments in children (thus excluding neurogenic tumors), from 1% to 57% of masses are reported to be due to histoplasmosis. This extreme variance in rate is due to prevalences of Histoplasma capsulatum in the geographic areas in which studies are done. For example, at St. Jude Children’s Research Hospital in Memphis, histoplasmosis was a relatively common complication in children with cancer, and a reason for referral of children without cancer.2 Fever and pulmonary infiltrates (frequently nodular) and hilar or mediastinal mass/lymphadenopathy were the usual presentation. Blastomycosis, coccidiodomycosis, and tuberculosis are also associated with infectious mediastinal disease. A variety of other pathogens, including Pneumocystis carinii (P. jirovecii) and mycoplasma, can produce mediastinal disease in association with pulmonary infiltrates. Cysts and tumors predominate as the cause of posterior mediastinal lymphadenopathy (Table 20-2) and, overall, neurogenic tumors are the most common cause of mediastinal masses; lymphomas are second and germ cell tumors are third in frequency.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Mediastinal and Hilar Lymphadenopathy

CHAPTER

20

TABLE 20-2. Relative Frequencies of Noninfectious Mediastinal

TABLE 20-3. Symptoms of Compression Resulting from Mediastinal

Masses in Childrena

Adenopathy Study

157

Structure

Symptom or Sign

Airway

Cough, wheezing; recurrent respiratory infections, bronchitis, atelectasis, unresolved pneumonia; hemoptysis; chest pain, sudden death

Esophagus

Dysphagia (interruption of peristalsis); hematemesis (fistula formation)

Superior vena cava

Dilation of collateral veins of the neck and upper thorax; chemosis of conjunctiva, edema of face, neck, upper chest, and arm; cyanosis; headaches, visual disturbances; epistaxis, tinnitus

Lymphatic channels

Pleural effusion

Silverman & Saleiston27

Filler et al.28

Woods et al.15

Gaebler et al.29

437

429

68

37

Neurogenic tumor

40

33

–

–

Lymphoma

18

14

68

43

Leukemia

–

–

17

–

Germ cell tumor

11

9.8

–

–

Mesenchymal tumor

7

6.8

–

–

Bronchogenic cyst

7

7.5

–

–

Recurrent laryngeal nerve

Hoarseness, inspiratory stridor (paralysis of vocal cord)

9

57

Phrenic nerve

Paralysis of left diaphragm

6

–

Sympathetic ganglia

Horner syndrome

Vertebrae, ribs

Pain secondary to bony erosion; symptoms of spinal cord compression

Number of children FREQUENCY OF MASS (%)

Lymph node infection

–

Other cysts or malignancy

14

4.4 25

a

Anterior and middle mediastinal masses only.

CHARACTERISTICS OF LYMPHADENOPATHY Lymph nodes are readily identifiable on computed tomography (CT) and can be categorized according to size, shape, coalescence, replacement by tumor mass, presence of calcium, abscess cavities, and parenchymal lung involvement. Most authorities use 10 mm as the upper limit of normal for lymph node size.3,4 Mediastinal lymph nodes > 20 mm in diameter are virtually always abnormal; this statement may not be true for hilar nodes. Mediastinal nodes have the potential to become much larger than any other nodes in the body. Densely calcified bilateral nodes are typical of histoplasmosis and, occasionally, of blastomycosis, coccidioidomycosis, or sarcoidosis. Unilateral densely calcified nodes are typical of prior or latediagnosed tuberculosis. Benign mediastinal tumors can also be calcified (i.e., teratoma, cystic thymoma, and thyroid adenoma). Childhood lymphomas can cause bilateral hilar adenopathy but do not usually become calcified until after therapy. In general, mediastinal mass lesions require a tissue diagnosis. Situations in which a biopsy may not be necessary include: (1) certain granulomatous lesions that show dense calcification and are not increasing in size, as shown by consecutive imaging studies; (2) lymphomatous lesions diagnosed by biopsy of tissue outside the thorax; and, most importantly, (3) tuberculosis or histoplasmosis, which should be confirmed with other diagnostic testing.

CLINICAL MANIFESTATIONS Mediastinal lymphadenopathy (adenopathy) is frequently asymptomatic until compression or erosion through a mediastinal structure occurs. Table 20-3 delineates findings associated with mediastinal disease. Respiratory symptoms result from airway obstruction or erosion. If the obstruction is significant enough, distal obstructive emphysema, atelectasis, pneumonitis, or chronic, recurrent respiratory tract infections can result. Obstruction of the superior vena cava is a rare complication most commonly associated with rapidly growing malignant mediastinal tumors. The superior vena cava is particularly vulnerable to obstruction because of its thin wall, low intravascular pressure, and confinement by lymph nodes and other rigid structures. Older children with mediastinal masses may describe feeling intrathoracic discomfort or pain, which is thought to be related to pressure on intercostal nerves or pleura. Vertebral erosion secondary to posterior mediastinal tumors can cause a boring, interscapular pain.

Most children with mediastinal disease caused by histoplasmosis present with chest pain, or cough. The finding of mediastinal widening is often a surprise to the clinician and leads to evaluation by CT. Often, the diagnosis of lymphoma is considered first. If granulomatous lung disease is present, the diagnosis of tuberculous disease is considered and, in most endemic regions, histoplasmosis is added to the differential diagnosis. History of exposures, such as foreign birth, ingestion of unpasteurized milk, or farm/bird exposure, is helpful.

DIAGNOSIS Traditionally, the differential diagnosis of mediastinal masses evident on chest radiograph was developed according to location in one of the four mediastinal compartments. Use of CT has largely supplanted the need for this classification because distinct tissue groups are now defined: fat, lymph nodes, vessels, airways, thymus, esophagus, and paraspinal tissues.

Imaging Studies Unenhanced CT augments recognition of calcification. CT with administration of contrast material is useful to define anatomy better and distinguish vessels from lymph nodes. Identification of a solid, homogeneously enhancing soft-tissue mass on CT distinguishes lymphoma from cavitating lesions of histoplasmosis and tuberculosis, which are peripherally enhancing with low attenuation centrally. With the widespread use of multidetector helical CT, multiplanar imaging is now routine and allows definition of mediastinal structures in all planes. Magnetic resonance imaging (MRI) is rarely required to evaluate the mediastinum, since most lesions are well seen on multidetector helical CT. In addition, MRI presents unique challenges in children, including the need for sedation (which can be dangerous in the presence of intrathoracic mass) and the inability of the child to hold the breath (which causes motion artifact). In children, MRI can occasionally be used for specific problem-solving, such as evaluation of soft-tissue masses of the chest wall or evaluation of extension of neurogenic tumors into the spinal canal. Currently, CT is the imaging modality of choice for mediastinal and lung evaluation beyond plain radiography, which is an excellent initial study. Distinguishing between hilar masses and adjacent

158

SECTION

B Cardinal Symptom Complexes

collapsed or consolidated lung is not usually problematic. Although CT and MRI are highly sensitive in detecting enlarged lymph nodes, neither reliably distinguishes between benign and malignant causes. Both can show displacement or compression of trachea or esophagus, but neither distinguishes invasion. MRI has the limitation of not being able to demonstrate calcium and is highly sensitive to motion (breathing or heart contraction) artifact. CT is best used for evaluation of mediastinal masses detected on plain radiographs, and MRI can be performed as an adjunct if there is suspicion of a neurogenic lesion.

Tissue Diagnosis and Biopsy In general, mediastinal mass lesions require a tissue diagnosis because of likely causes and their proximity to vital structures. In some cases, an accurate diagnosis can be made without obtaining mediastinal tissue. Supraclavicular nodes (whether enlarged or not) or other extrathoracic adenopathy can provide diagnostic tissue. In one-half of children with mediastinal malignancy, at least one lymph node in the cervical, supraclavicular, or infraclavicular or axillary area is larger than 2 cm, and can be biopsied. Tissue biopsy from other sites can confirm sarcoidosis. Bronchoscopy with bronchoalveolar lavage or gastric aspirates may identify an infectious cause.

INFECTIOUS CAUSES Table 20-4 shows differentiating features of the causes of mediastinal lymphadenopathy. Mycobacteria and the endemic fungal organisms, Histoplasma and Coccidioides, commonly cause hilar and mediastinal adenopathy. Histoplasmosis and tuberculosis are the two most common infections affecting mediastinal nodes. Although mediastinal infection without significant pulmonary infection is rare, the degree of adenopathy is commonly disproportionate to the level of parenchymal involvement. Pneumonia due to Mycoplasma pneumoniae, Bartonella henselae, Yersinia enterocolitica, or Francisella tularensis can be associated with hilar adenopathy, as can bronchiectasis and cystic fibrosis. Adenopathy is rare in other acute bacterial and viral pneumonias.

Bacterial Causes

Mycobacterium tuberculosis Primary pulmonary infection with M. tuberculosis has the following three elements: the primary parenchymal focus, intraparenchymal lymphangitis, and regional lymphadenitis (Figures 20-2 and 20-3). At

TABLE 20-4. Causes of Mediastinal Lymphadenopathy Frequency

Associated Pneumonia

Node Abscess

Node Calcification

Bacteria Mycobacterium tuberculosis

++

++

++

++

Nontuberculous mycobacteria Mycoplasma pneumoniae Bartonella henselae Yersinia enterocolitica Actinomyces israelii Melioidosis

± + ± ± ± +

– ++ – – – –

+ – + + + +

– – – – – –

Fungus Histoplasma Coccidioides Blastomyces Paracoccidioides Cryptococcus

++ + + ± ±

+ + ++ + ±

+ ± + ++ –

++ + + –

Serology; antigen detection; culture Serology; culture Culture Culture Culture

Viruses Epstein–Barr virus

±

–

–

–

Serology

Protozoa Toxoplasma

±

–

–

–

Serology

Chronic infection Cystic fibrosis Bronchiectasis Lung abscess

++ + ±

++ + +

± ± –

– – –

Sweat test; gene identification Radiograph; culture Radiograph

Malignancy Hodgkin lymphoma Non-Hodgkin lymphoma Leukemia

++ ++ +

± – –

– – –

±a ±a

Biopsy Biopsy Bone marrow

Other Chronic granulomatous disease Sarcoidosis Rosai–Dorfman disease Castleman disease Lymphoproliferative syndromes

+ ++ + + ++

+ – – – –

++ – – – –

– – – – –

Neutrophil function assay Biopsy Biopsy; histopathology Biopsy; histopathology Epstein–Barr virus serology; biopsy

Etiology

Method of Diagnosis

INFECTIOUS CAUSES

Tuberculin skin test; chest radiograph; culture of gastric aspirate or bronchial washing Culture and histopathologic analysis of tissue Serology; cold agglutinin Serology; culture; histopathology Culture; serology Culture; histopathology Culture

NONINFECTIOUS CAUSES

++, frequent cause; +, occasional cause; ±, rare cause, consider under special circumstances. a After treatment.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Mediastinal and Hilar Lymphadenopathy

A

CHAPTER

20

159

affected) and deep cervical (supraclavicular) nodes.5 Additionally, apical subpleural primary infections can drain directly to the supraclavicular nodes. In one series of patients, 14 of 54 patients with a primary lesion in the right upper lobe had ipsilateral enlargement of deep cervical nodes.6 Enlargement of hilar nodes adjacent to bronchi can cause bronchial obstruction with the collapse or consolidation or fan-shaped segmental lesion typical of childhood tuberculosis. Infection and inflammation of the bronchial wall can occur, and obstruction of the lumen can rarely result in sudden death or in obstructive hyperaeration, segmental atelectasis, or secondary pneumonia. Multiple segmental lesions can occur, usually in the same lung. Calcification follows caseous necrosis and occurs more often in children and in the regional lymph nodes than in the primary pulmonary focus. Extensive calcification is uncommon with early treatment, which prevents caseation. In one series, CT of 23 young adults presenting with tuberculous mediastinal or hilar lymphadenitis demonstrated pulmonary involvement in 14 patients.7 There was remarkable preponderance of involvement of the right paratracheal and tracheobronchial nodes. Nodes > 2 cm in diameter invariably showed central areas of relatively low density and peripheral rim enhancement that was irregular in thickness.

Nontuberculous Mycobacteria In immunocompromised children (and occasionally in normal young children), nontuberculous mycobacteria can cause: (1) extensive hilar and paratracheal lymphadenopathy with or without parenchymal disease; and (2) endobronchial disease with atelectasis.8 In all reported patients in one series, the organism was isolated from lymph node, lung, or bronchoalveolar lavage. Lymphadenopathy can be unilateral or bilateral, and nodes undergo extensive caseating necrosis.

Mycoplasma pneumoniae A careful evaluation of the radiographic findings in 56 patients with pneumonia (21 younger than 20 years) due to M. pneumoniae revealed that 22% of patients had hilar or paratracheal node enlargement, and 14% had pleural effusions.9,10

Bartonella henselae B Figure 20-2. (A) Chest radiograph of a child with tuberculosis shows right paratracheal adenopathy (closed arrow) and right tracheobronchial adenopathy (arrowhead), causing effacement of the right lateral aspect of trachea (open arrows). (B) Chest radiograph of the same patient 4 months later (without therapy) shows right upper lobe pneumonia with subtle effacement of the right lateral aspect of the trachea and right mainstem bronchus, indicating right hilar and right paratracheal adenopathy (arrows). There is splaying of the carina (arrowheads) and double density (D) of the lower mediastinum centrally, indicating subcarinal adenopathy.

A few patients with cat-scratch disease have been described with mediastinal and peripheral lymphadenopathy. Mediastinal node biopsy in one patient revealed granuloma with microabscesses.11

Yersinia enterocolitica Significant bilateral self-limited hilar adenopathy without pneumonia due to Y. enterocolitica has been described in several adults. All patients recovered without therapy. The diagnosis was based on results of culture or serum agglutination titers.12

Lung Abscess least 70% of primary pulmonary foci are subpleural, with spread through lymphatics to the regional lymph nodes. After several weeks, hypersensitivity develops, with regional node enlargement and the potential for caseating necrosis. Caseating lesions have a high density of actively multiplying bacilli, which spread rapidly to adjacent lymphatics. The hallmark of early tuberculosis is excessive unilateral mediastinal lymphadenitis compared with the relatively insignificant focus in the lung. Hilar adenopathy is unilateral in 80% to 90% of patients with tuberculosis. Infection can spread beyond the hilar and tracheobronchial nodes to the more distant right upper paratracheal nodes (the ones most often

Lung abscess is frequently associated with anaerobic infections but can occur in the course of any necrotizing pneumonia. Mediastinal lymphadenopathy is not unusual.

Bronchiectasis In children and adults, bronchiectasis of any etiology (e.g., associated with congenital anomaly, foreign body), and especially that associated with cystic fibrosis in children, can cause mediastinal lymphadenopathy. CT in patients with bronchiectasis demonstrated mediastinal lymphadenopathy (nodes > 10 mm in diameter) in 29% of patients in one study.13

160

SECTION

B Cardinal Symptom Complexes

B

A

Figure 20-3. (A and B) Chest radiographs of a child with reactivated tuberculosis show cystic changes in the apex of the right upper lobe (open circle), and pleural thickening in the superior hemithorax (long arrows). There is a right pleural effusion (e); right paratracheal, tracheobronchial, and hilar adenopathy (short arrows); and atelectasis or pneumonia in the right middle lobe (a), right upper lobe (u), and right lower lobe (l).

generally occurs sporadically, manifesting as a prolonged febrile illness with subacute pulmonary symptoms.17

Fungal Causes Histoplasmosis Infection due to Histoplasma capsulatum has multiple clinical forms (Figures 20-4 and 20-5). In the immunocompetent host, it occurs most frequently as subclinical infection, recognized incidentally. Most commonly patients complain of cough or chest pain. Mediastinal lymphadenopathy, an uncommon finding in disseminated infection, is the most common manifestation in healthy children with selflimited infection. Of 35 children hospitalized with histoplasmosis between 1968 and 1988, 29 (83%) had pulmonary infection, mediastinal infection, or both. Thirty-one of the chest radiographs (91%) were abnormal, revealing adenopathy in 25 (74%) and infiltrates in 19 (56%).14 Isolated mediastinal adenopathy can be difficult to distinguish from lymphoma.15,16 Rarely, massive inhalation of fungus causes disseminated, symptomatic primary infection with fever, cough, and massively enlarged mediastinal lymph nodes. The differentiation is important as, in most cases, histoplasmosis is mild and self-limited in most healthy individuals. Antifungal therapy is reserved for those with acute diffuse pulmonary infection, chronic pulmonary histoplasmosis, progressive disseminated disease, and mediastinal adenitis associated with obstructive symptoms. In the healthy host, marked enlargement of mediastinal lymph nodes can cause obstruction of bronchi, the superior vena cava, or the esophagus. Additionally, a group of large nodes can become necrotic and matted to form a large single mediastinal mass known as an acute mediastinal granuloma. Most such masses resolve spontaneously; the effect of antifungal therapy has not been studied. In the transplant population, pulmonary histoplasmosis can be a difficult diagnosis, mainly because it is not often considered. It

Coccidioidomycosis Coccidioidomycosis occurs primarily in the southwestern United States. It is subclinical in children in 60% of cases. An insignificant flulike illness or severe respiratory infection with lobar pneumonia and pleural effusions can occur. Radiographic findings are not specific; bronchopneumonic infiltrates are often associated with hilar node enlargement. Cavitation, nodule formation, or calcification occurs in a minority of patients. Parenchymal infiltrates are commonly associated with hilar lymphadenopathy. More extensive mediastinal adenopathy is not considered part of the primary complex and may represent dissemination. Rarely, bilateral hilar adenopathy occurs without apparent pulmonary disease.

Paracoccidioides brasiliensis A progressive chronic disease due to P. brasiliensis, paracoccidioidomycosis preferentially involves the lungs and reticuloendothelial system. It occurs predominantly in a few Latin American countries. Only about 2% of cases occur in children. Pulmonary lesions can be infiltrative, fibrotic, fibrocaseous, or cavitary. Enlargement of hilar and mediastinal lymph nodes is observed. Enlargement of liver, spleen, and other lymph nodes (e.g., cervical, inguinal, and mesenteric) also occurs. Complications include suppuration and fistula formation and scarring of affected nodes with residual pulmonary fibrosis.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Mediastinal and Hilar Lymphadenopathy

A

CHAPTER

20

161

B

Figure 20-4. (A and B) Chest radiographs of a child with healed histoplasmosis shows granuloma in right upper lobe (open arrow), calcified paratracheal nodes (straight arrow), tracheobronchial nodes (arrowhead), and calcified right hilar nodes (curved arrow).

Pneumocystis jirovecii (P. carinii ) A patient with acquired immunodeficiency syndrome (AIDS) and P. jirovecii pneumonia with calcified hilar and mediastinal nodes has been reported.18

Viral Causes Epstein–Barr Virus Generalized lymphadenopathy secondary to extensive lymphocytic hyperplasia can include the mediastinum.19,20 In transplant recipients undergoing immunosuppression, lymphoproliferative disease related to Epstein–Barr virus can manifest as generalized lymphadenopathy, including involvement of the mediastinal nodes.

Rubeola Pulmonary infiltrates and hilar adenopathy appearing early in the course of measles are well described. In one series of 130 children with measles, 55% had pulmonary infiltrates, and 74% had hilar adenopathy.21

NONINFECTIOUS CAUSES Malignancy Lymphomas are a common cause of mediastinal masses in both children and adults (Figures 20-6 and 20-7). They characteristically occur in the anterior and middle mediastinal compartments and are the third most common cause of mediastinal masses in children. The first most common manifestation of Hodgkin lymphoma or non-Hodgkin lymphoma is painless lymph node enlargement, most often in the cervical or supraclavicular chains. Fifty percent to 60% of patients

with Hodgkin lymphoma have mediastinal lymphadenopathy (usually bilateral) and evidence of tracheobronchial compression at the time of diagnosis. Involved nodes are hilar (25% or greater), subcarinal (22%), posterior mediastinal (5%), and cardiophrenic angle (8%). Ipsilateral pulmonary parenchymal involvement is present in 10% of patients. Childhood non-Hodgkin lymphoma is a rapidly proliferating malignancy, usually affecting adolescent males and often with a duration of symptoms of only 4 to 6 weeks. Lymphoblastic lymphoma and the closely related acute lymphoblastic leukemia often manifest indistinguishably, with mediastinal adenopathy and compression. Dissemination is present in 70% of children at the time of diagnosis. Non-Hodgkin lymphoma can arise in any lymphoid tissue and in numerous extralymphatic sites. Painless, rapidly progressive enlargement of cervical, supraclavicular, and axillary nodes is a common presenting complaint. With involvement of these nodes, an anterior mediastinal tumor is common (50% to 70%), as are pleural effusions.

Other Causes Sarcoidosis Sarcoidosis is a multisystem granulomatous disease of unknown cause that is relatively uncommon in children; it has two age group incidence peaks, between 9 and 16 years of age and in young adults. Sarcoidosis most commonly involves the lungs, lymph nodes, eyes, skin, liver, and spleen. Nonspecific symptoms consist of fever, weight loss, cough, fatigue, lethargy, lymphadenitis, visual disturbances, and rashes. In large series of children and adults, 90% of patients have bilateral involvement of the lungs and hilar lymph nodes.22,23 In children, bilateral hilar lymphadenopathy, occasionally involving the paratracheal nodes, was associated with parenchymal involvement in 50% of cases.23 The patient should be evaluated for ocular involvement as well as other visceral sites of disease.

162

SECTION

B Cardinal Symptom Complexes

A

Figure 20-6. Chest radiograph of a child with mixed-cell Hodgkin disease shows right paratracheal (arrows) and tracheobronchial lymphadenopathy (arrows). The superior mediastinum is widened on the left as well, indicating enlargement of anterior mediastinal nodes (arrowheads).

B ovoid nodes 3 to 8 cm long are characterized by lymphoid follicular hyperplasia with and without germinal center formation and marked capillary proliferation with endothelial hyperplasia. Human herpesvirus 8 (HHV-8) has been found in the nodes of affected patients. Surgical removal of localized nodes for diagnosis and therapy may relieve systemic symptoms. No recurrences have been reported in children. In a patient infected with human immunodeficiency virus, foscarnet and antiretroviral agents resulted in dramatic improvement.26

Wegener Granulomatosis Mediastinal adenopathy with tracheal and bronchial stenosis associated with Wegener granulomatosis has been reported rarely in children. In these cases, disease was limited to the airways and sinuses.

C Figure 20-5. (A through C) Enhanced computed tomography scan of the patient in Figure 20-4 shows extensive right tracheobronchial, right hilar, and subcarinal lymphadenopathy. Particulate calcification within the nodes is characteristic of healed histoplasmosis. Arrow indicates calcified granuloma in right upper lobe. AA, ascending aorta; DA, descending aorta; H, calcified right hilar node; LB, left mainstem bronchus; RB, right mainstem bronchus; SC, calcified subcarinal nodes; SV, superior vena cava; T, trachea; TB, calcified tracheobronchial nodes; TH, thymus.

Giant Lymph Node Hyperplasia (Castleman Disease) Castleman disease, a rare disease in children younger than 14 years, can manifest as an asymptomatic mass in the mediastinum or neck, along the aorta, or in the abdomen.24,25 Some patients have prolonged fever, anemia, weight loss, splenomegaly, hypergammaglobulinemia, elevated erythrocyte sedimentation rate, bone marrow plasmacytosis, and a poor outcome. Microscopically, the large, well-circumscribed

Chronic Granulomatous Disease Hilar adenopathy in association with pneumonia raises suspicion of chronic infection or granuloma formation. Patients with chronic granulomatous disease frequently have mediastinal adenopathy in conjunction with pneumonia.

DIAGNOSTIC APPROACH The diagnostic approach depends on the age of the patient, exposures, presence and progression of symptoms, associated physical findings, and specific abnormalities seen on imaging studies (see Table 20-4). Results of serologic tests (e.g., for Bartonella, Mycoplasma, Histoplasma infection) and the Mantoux skin test frequently confirm the diagnosis, averting the need for tissue diagnosis. H. capsulatum antigen detection in serum, urine, or bronchoalveolar lavage fluid may confirm the diagnosis. It is most sensitive as an indicator of disseminated disease and can be negative in the setting of pulmonary histoplasmosis. Rapidly enlarging masses, in the absence of a confirmed,

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Abdominal and Retroperitoneal Lymphadenopathy

CHAPTER

CHAPTER

21

163

21

Abdominal and Retroperitoneal Lymphadenopathy John S. Bradley and Mary Anne Jackson

A

B Figure 20-7. (A and B) Enhanced computed tomography scans of a child with mixed-cell Hodgkin disease show extensive mediastinal adenopathy. Arrow indicates carina, and arrowhead indicates esophagus containing air. AM, anterior mediastinal nodes; AO, aortic arch; DA, descending aorta; SC, subcarinal nodes; SV, superior vena cava; T, trachea.

compatible diagnosis, require prompt tissue biopsy. Optimal therapy of the variety of causes depends on rapid and accurate diagnosis.27,28 Subacute infections associated with hilar lymphadenopathy can be difficult to distinguish from lymphoreticular malignancy. In one study of 37 children with anterior and middle mediastinal masses, 16 had lymphomas and 21 had histoplasmosis.29 The two entities could not be distinguished on the basis of patient age or sex, fever, weight loss, duration of illness, anemia, erythrocyte sedimentation rate, or lung infiltrates and calcifications. Anterior mediastinal masses were due to lymphoma in 81% of cases and histoplasmosis in 5%. Of the patients with lymphomas, only 7% had Histoplasma complement fixation titers of 1:8; of patients with histoplasmosis, 67% had titers of 1:32 or greater. The authors of this study concluded that anterior mediastinal masses should be biopsied. Immunodiffusion testing for H. capsulatum, however, is highly specific. The finding of H bands is highly suggestive of acute infection (see Chapter 251, Histoplasma capsulatum (Histoplasmosis)).

Lymphadenopathy, the abnormal enlargement of lymph nodes, can result from a wide variety of infectious and noninfectious causes. Acute enlargement of superficial lymph nodes caused by infection can occur over a period of days, accompanied by pain and tenderness to palpation, more often referred to as lymphadenitis. Alternatively, the enlargement of nodes can occur over weeks to months, with little tenderness, and this represents a more chronic inflammatory process. Abdominal and retroperitoneal lymphadenopathy is a radiologic diagnosis made by imaging techniques such as ultrasonography (US), computed tomography (CT), or magnetic resonance imaging (MRI). Without sequential imaging studies, the physician does not know if the enlargement represents a new, rapidly evolving process or a more prolonged, chronic, relatively stable one. Without the ability to examine the lymph nodes directly, the clinician must rely on clues from the history and physical examination to formulate a diagnosis. Inflammation in lymph nodes may be the direct result of infection in the tissues (or other lymph nodes) that they drain or a consequence of bloodstream dissemination of organisms. Tissues and organs of the abdomen and retroperitoneal space can be infected by a variety of pathogens through various routes (see Chapter 70, Appendicitis and Mesenteric Lymphadenitis; Chapter 71, Intra-abdominal, Visceral, and Retroperitoneal Abscesses).

EPIDEMIOLOGY AND DIFFERENTIAL DIAGNOSIS Reviews of lymphadenopathy in children delineate extensive infectious and noninfectious causes. Malignant processes and serious infection are the primary concerns; malignant tumors may account for up to 25% of cases of peripheral or hilar adenopathy in children.1–6 A systematic approach to evaluation for infection requires consideration of selected factors from the history and clinical examination (Box 21-1). The regional lymph nodes that drain the organs and tissues from the abdomen, retroperitoneal space, and lower extremities are shown in Figure 21-1.4,7,8 Focal, deep lymphadenopathy suggests infection in the adjacent draining structures or lymph node groups; generalized adenopathy suggests a disseminated process that is likely to involve other lymph nodes, which may be palpable and enlarged, as well as the liver and spleen. Infectious causes of enlarged superior or inferior mesenteric lymph nodes include most of the bacterial, viral, and parasitic pathogens that cause gastroenteritis. Mesenteric lymph node enlargement in association with an acute, symptomatic febrile illness that most often manifests as abdominal pain is usually referred to as mesenteric adenitis. Mesenteric adenitis may be caused by many pathogens: Yersinia spp., Salmonella, Shigella, Escherichia coli, adenoviruses, enteroviruses, and Entamoeba histolytica are important causes and each has a characteristic constellation of clinical and laboratory findings (see Chapter 70, Appendicitis and Mesenteric Lymphadenitis).4,9–13 Association of inflammation of the terminal ileum and ileocecal junction with mesenteric adenitis occurs with infectious causes, such as Yersinia,10 Mycobacterium tuberculosis or M. bovis, as well as with Crohn disease.12,14It is important to consider M. tuberculosis infection in such cases if the child was born in an endemic region, has traveled to tuberculosis-endemic areas, or may have ingested contaminated food products (e.g., homemade unpasteurized cheese).14–16

164

SECTION

B Cardinal Symptom Complexes

BOX 21-1. Evaluation of Patients with Lymphadenopathy of Suspected Infectious Etiology SYMPTOMS Duration of symptoms: acute (days) or chronic (weeks to months) Presence of weight loss, fever Presence of organ system-specific symptoms: diarrhea, abdominal pain, jaundice, dysuria, flank pain, hip pain EXPOSURES Exposure to animal bites or scratches: cats, rodents, farm animals Travel to areas endemic for enteric bacterial or parasitic pathogens, fungal pathogens: Histoplasma, Coccidioides, Mycobacterium, or Plasmodium Ingestion of unpasteurized dairy products Risk factors for HIV infection UNDERLYING DISORDERS Concomitant infection elsewhere in the patient History of previous or chronic infection of liver or gallbladder, spleen, pancreas, bowel, kidney or ureters, bladder, ovaries or uterus, iliopsoas muscle, or vertebral bodies Presence of underlying immunocompromising disorders: malignant tumors, chemotherapy, corticosteroid therapy, HIV infection Congenital immunodeficiencies: chronic granulomatous disease, severe combined immunodeficiency Transplantation: immunodeficiency and posttransplant lymphoproliferative disorders Crohn disease Celiac disease PHYSICAL EXAMINATION Lymphadenopathy: focal versus regional or systemic Evidence of infection in tissues or organs drained by lymph nodes Evidence of systemic infection: Epstein–Barr virus infection, cat-scratch disease, tuberculosis HIV, human immunodeficiency virus.

Epstein–Barr virus (EBV) and Bartonella henselae infection are the most commonly reported infectious causes of abdominal lymphadenopathy in otherwise healthy children. In mononucleosis, the accompanying malaise, pharyngitis, systemic lymphadenopathy, and hepatosplenomegaly often lead to the diagnosis.17 Peripheral findings of a papule at the site of inoculation and of regional peripheral adenopathy are frequently present in patients with cat-scratch disease. However, fever, with or without abdominal pain, in the absence of peripheral findings, can be the presentation of B. henselae infection of the liver, spleen, or abdominal lymph nodes.18,19 Less common causes

Celiac nodes drain stomach, spleen, pancreas, liver, proximal duodenum

of adenopathy associated with other systemic infections include malaria20 and toxoplasmosis.21 Tuberculosis in children can manifest with abdominal lymphadenopathy as the sole anatomic manifestation of systemic infection. Adenopathy can occur during primary infection or during reactivation of a previous infection. Both M. tuberculosis and M. bovis can cause symptomatic infection of the intestines, lymph nodes, peritoneum, omentum, liver, or spleen.15,16,23 Children can have symptoms of weight loss, fatigue, and intermittent abdominal pain in the absence of abdominal physical findings. The clinical presentation can mimic inflammatory bowel disease. If infection is caused by M. tuberculosis, hilar adenopathy may also be detected on chest radiograph but, in most cases, abdominal tuberculosis occurs in the absence of pulmonary disease. Lymphoproliferative syndromes caused by EBV and characterized by abdominal lymphadenopathy have been well described.17,24 Pickhardt and colleagues cite an incidence of approximately 8% for lymphoproliferative syndrome in pediatric transplant recipients at a single center.24 Evidence of EBV infection of B lymphocytes is usually present. Approximately one-half of children with an EBVrelated abdominal lymphoproliferative disorder have abdominal pain; one-quarter have symptoms from other enlarged lymphoid tissue within the liver or spleen or present with an abdominal mass. In children with symptomatic human immunodeficiency virus (HIV) infection, lymphadenopathy can signify the underlying HIV infection, an HIV-related cancer, or secondary infection.2,25 In particular, M. avium complex (MAC) infection can cause massive systemic lymphadenopathy, including the mesenteric and retroperitoneal lymph nodes.26 Congenital immunodeficiencies, such as chronic granulomatous disease, may also rarely be characterized by mesenteric lymphadenopathy.27 Mesenteric lymphadenopathy can be a reaction to noninfectious inflammation of the bowel, such as Crohn disease12 or celiac disease.28 Abdominal lymphadenopathy may be one of the systemic manifestations of a number of disorders of unknown cause, including sarcoidosis,29 Castleman disease,30 Rosai–Dorfmann disease,3 or familial Mediterranean fever.31

DIAGNOSTIC TESTS Imaging Studies The diagnosis of abdominal and retroperitoneal lymphadenopathy is based on imaging studies. US has been used to assess the significance and degree of abdominal adenopathy in both infectious and non-

Aorta

Figure 21-1. Location and drainage patterns of deep abdominal lymph nodes.

Celiac artery Superior mesenteric artery Lumbar lateral aortic nodes drain kidney, adrenals and superior ureter

Superior mesenteric nodes drain spleen, pancreas, distal duodenum to transverse colon Inferior mesenteric nodes drain descending colon to sigmoid colon

Inferior mesenteric artery Common iliac artery

External iliac nodes drain external genitalia and lower extremities Internal iliac artery

External iliac artery

Lumbar aortic and common iliac nodes drain pelvic organs PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Localized Lymphadenitis, Lymphadenopathy, and Lymphangitis

infectious conditions,32,33 including cat-scratch disease34 and abdominal tuberculosis.35,36 There may be subtle differences detectable by US that can help differentiate between infectious causes of enlarged nodes. In studies in which US and CT have been compared, CT is more sensitive for detecting abdominal lymphadenopathy.35 CT aids in the diagnosis of both abdominal and retroperitoneal adenopathy.12,37 Similarly, MRI has been used to evaluate lymphadenopathy in both the abdominal and retroperitoneal spaces and has advantages over other imaging techniques for the detection of vascular lesions and tumors.16,38,39

Clinical Findings Clues to the cause of lymphadenopathy are based on physical findings that can include systemic lymphadenopathy or hepatosplenomegaly. Other physical findings of infection of the respiratory tract, gastrointestinal tract, skin, skeletal system, cardiovascular system, or genitourinary system can also provide clues to the primary site of infection. In the immunocompetent child, a Mantoux intradermal tuberculin skin test remains one of the most sensitive and specific techniques for diagnosis of infections caused by M. tuberculosis or M. bovis. Falsenegative test results have been seen in conjunction with disseminated tuberculosis and immunocompromising conditions. Diagnostic techniques for tuberculosis based on production of interferon by cultured peripheral blood lymphocytes following in vitro exposure to purified proteins frem M. tuberculosis are becoming available for use in adults, but are not yet well studied in children.40 These tests are more specific for latent or active infection caused by M. tuberculosis and thus may be able to differentiate M. tuberculosis infection from that caused by M. bovis or non-tuberculous mycobacteria.

Biopsy Techniques When tissue samples are required for accurate diagnosis, particularly in the case of tumors, the techniques discussed in Chapter 71 (Intraabdominal, Visceral, and Retroperitoneal Abscesses) are useful not only for the diagnosis of organ disease, but also of lymph node disease. Options include open surgical biopsy and laparoscopic biopsy or fine-needle biopsy, directed by US or CT. Considerations for choosing the optimal technique in individual cases include the volume of tissue required for diagnosis, the urgency for arriving at a diagnosis, the skill of the clinician responsible for performing the procedure, and the ancillary support services available. In general, histologic diagnosis of specific tumors requires larger amounts of tissue than are required for diagnosis of infection. Special stains and cultures can be used to identify particular bacteria, fungi, mycobacteria, or viruses. Biopsy samples should be examined histologically for acute or chronic inflammation, abscess or granuloma formation, and evidence of primary or metastatic tumors. Each pathogen generally produces a characteristic histopathologic profile. For example, caseating granuloma characterizes tuberculosis, but stellate granuloma is more suggestive of infection due to B. henselae.

Laboratory Tests Many infections that lead to abdominal lymphadenopathy can be detected by specific serologic tests, including infections caused by B. henselae, EBV, cytomegalovirus, Toxoplasma sp., and HIV. The types of cultures obtained from blood, urine, stool, or biopsy specimens should be directed toward the suspected pathogens, noting the particular lymph node group affected, pertinent history and clinical findings, and associated or underlying disease(s).

THERAPY Although many infectious causes of abdominal and retroperitoneal lymphadenopathy are benign and self-limiting, antimicrobial therapy can be beneficial depending on the implicated pathogen. The chapters that discuss each specific pathogen in detail should be consulted.

CHAPTER

CHAPTER

22

165

22

Localized Lymphadenitis, Lymphadenopathy, and Lymphangitis Victor Nizet and Mary Anne Jackson

LYMPHADENOPATHY AND LYMPHADENITIS Lymphadenopathy is defined as disease of the lymph nodes, but the term is more commonly used to denote lymph node enlargement. Enlarged lymph nodes can arise in association with a wide variety of infectious, inflammatory, or neoplastic disease processes. Lymphadenitis refers to a localized inflammatory process within a given lymph node or group of nodes, usually of bacterial etiology. Lymphadenitis can develop acutely or chronically and can be pyogenic or granulomatous in nature. Other chapters in this textbook are specifically devoted to the evaluation and management of generalized lymphadenopathy (see Chapter 18, Lymphatic System and Generalized Lymphadenopathy) and lymphadenopathy of the cervical nodes (see Chapter 19, Cervical Lymphadenitis and Neck Infections), hilar nodes (see Chapter 20, Mediastinal and Hilar Lymphadenopathy), inguinal nodes (see Chapter 54, Skin and Mucous Membrane Infections and Inguinal Lymphadenopathy), and abdominal nodes (see Chapter 21, Abdominal and Retroperitoneal Lymphadenopathy) groups. The current discussion focuses on regional lymph node disease encountered in the remaining superficial locations.

Pathogenesis Lymph nodes can enlarge as a result of: (1) intrinsic proliferation of lymphocytes or reticuloendothelial cells; or (2) infiltration by cells from an extrinsic source. Lymphocytes or lymphoblasts proliferate upon recognition of antigenic stimuli, producing lymph node enlargement, which recedes upon antigen clearance. Infectious organisms able to survive intracellularly can represent persistent stimuli and can be associated with chronic lymphatic cell hyperplasia. Extrinsic invasion of lymph nodes occurs with neutrophils in response to bacteria and bacterial toxins, with histiocytes in histiocytosis and certain storage diseases, and with malignant cells in leukemia, lymphoma, and metastatic solid tumors.

Etiologic Agents The pyogenic bacteria Staphylococcus aureus and Streptococcus pyogenes account for greater than 95% of acute bacterial lymphadenitis. Lymph node infection can usually be attributed to spread from an adjacent skin infection or inoculation site; occasionally, no clear origin is evident. Acute localized adenitis with overlying cellulitis can develop in an infant as a manifestation of late-onset bacteremic group B streptococcal disease.1 Subacute development of localized lymph node enlargement in healthy children is found with cat-scratch disease (Bartonella henselae infection) and with nontuberculous mycobacterial infection. Adenitis with fungal pathogens such as Aspergillus or Candida can occur in immunosuppressed patients, such as those with leukemia who are undergoing chemotherapy.2 Unusual pathogens can produce lymphadenitis in otherwise healthy persons after specific environmental exposures (e.g., Francisella tularensis after tick bite/animal skinning or Corynebacterium pseudotuberculosis in sheep handlers).

166

SECTION

B Cardinal Symptom Complexes

History and Clinical Findings Certain clinical features may be useful in guiding the evaluation of localized lymph node infections (Table 22-1). Small inguinal, cervical, or axillary lymph nodes can be palpated in about a third of healthy neonates.3 Subsequent antigenic stimulation leads to steady enlargement of lymphoid tissue from infancy through puberty. As a result, the vast majority of healthy children have palpable cervical, inguinal, and axillary nodes.4 In contrast, other peripheral node groups (e.g., posterior auricular, supraclavicular, epitrochlear, iliac, and popliteal) are always considered abnormal if they can be palpated on examination.5 A careful history must be obtained regarding the time of onset and rate of lymph node enlargement (Box 22-1). Most local lymph node disease relates to a regionally inoculated infection or an adjacent focus. The skin and soft-tissue areas drained by the enlarged nodes should be examined for signs of infection and disruption. History of exposure to animals should be detailed. Bites including tick bites, or traumatic scratches or other papular or ulcerative skin lesions, must be elicited. Weight loss, protracted fever, rash, generalized lymphadenopathy, or hepatosplenomegaly suggests a systemic disease. Lymph nodes associated with viral infection tend to be small, bilateral, freely mobile, and variably tender.6 The presence of erythema, warmth, and tenderness overlying a node typically indicates the acute inflammatory response of a pyogenic process related to a bacterial or mycobacterial pathogen. Most acutely infected lymph nodes in children are rubbery or firm in consistency and freely mobile. Fixation of the node to underlying tissue should raise a concern about a neoplastic process. Fluctuation of the lymph node is suggestive of acute pyogenic bacterial adenitis but can also be seen with more indolent pathogens, including nontuberculous mycobacteria.

in the child with localized lymphadenopathy. The regions drained by specific lymph node groups are listed in Table 22-1, along with the associated infections encountered in each case.

Occipital Nodes The occipital lymph nodes drain the posterior scalp and can become enlarged in association with impetigo, pediculosis capitis, scalp ringworm, or inflamed seborrheic dermatitis. Tularemia often involves occipital nodes because a tick bite on the parieto-occipital scalp is common in children. The tick bite itself may be unknown or forgotten as often the bite itself occurred weeks prior. The history of exposure to a tick-infested area should raise suspicion. The scalp should be inspected carefully for a papule or an ulcer though glandular tularemia is more common than ulceroglandular disease. Less often, rubella infection or toxoplasmosis can produce isolated occipital lymphadenopathy.

Differential Diagnosis Knowledge of lymphatic drainage patterns is essential to identification of the primary focus of infection and the most likely etiologic agents

BOX 22-1. Clinical Clues to Etiology of Lymphadenopathy HISTORY Associated symptoms and duration of illness Ingestion of unpasteurized milk or undercooked meat Dental problems Skin lesions or trauma Animal exposures; flea or tick bites Contact with tuberculosis Drug usage (especially Dilantin) PHYSICAL EXAMINATION Dental disease Ocular, otic, or oropharyngeal lesions Skin lesions Noncervical adenopathy Hepatomegaly or splenomegaly

TABLE 22-1. Infections Associated with Localized Lymphadenopathy Infectious Etiologies Lymph Node Group

Area of Drainage

Palpable Nodes

Common

Less Common

Occipital

Back of scalp and neck

5% of healthy children

Impetigo Pediculosis (head lice) Tinea capitis Seborrheic dermatitis

Rubella Toxoplasmosis

Preauricular

Lateral portion of eyelids Lateral conjunctivae Skin above cheek, temple

Only with disease

Adenoviral conjunctivitis Parinaud syndrome secondary to cat-scratch disease

Chlamydia conjunctivitis Parinaud syndrome secondary to tularemia or herpes simplex virus infection

Axillary

Hand and arm

70–90% of healthy children

Local pyogenic infection

Calmette-Guérin bacillus vaccine, fever, tuberculosis (scrofuloderma), hidradenitis suppurativa

Chest wall, breast Upper lateral abdominal wall

Cat-scratch disease

Epitrochlear

Ulnar aspect of hand and forearm

Only with dsease

Local pyogenic infection

Tularemia, cat-scratch disease, secondary or congenital syphilis

Iliac

Lower abdominal viscera Urinary and genital tract Lower extremities

Only with disease

Abdominal infection such as appendicitis, following abdominal trauma, urinary tract infection

Pyogenic infection of lower extremity

Popliteal

Skin of lateral foot and lower leg, knee joint

Only with disease

Severe local pyogenic infection

–

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Localized Lymphadenitis, Lymphadenopathy, and Lymphangitis

CHAPTER

22

167

Preauricular Nodes Preauricular nodes drain the lateral eyelid and conjunctivae, along with skin overlying the cheek and temple. Severe conjunctivitis associated with ipsilateral preauricular lymphadenopathy is known as Parinaud or oculoglandular syndrome. Oculoglandular syndrome is a common complication of Bartonella henselae infection (cat-scratch disease), affecting approximately 5% of symptomatic patients (Figure 22-1).7The differential also includes tuberculosis, tularemia, and nontuberculous lymph node infection. Chlamydial neonatal inclusion conjunctivitis can appear 5 to 10 days after birth, and the presence of lymph node enlargement helps distinguish it from gonococcal ophthalmia neonatorum.8 Oculoglandular syndrome is often seen in trachoma, a serious and potentially blinding form of chlamydial eye infection endemic in many developing countries. Epidemic keratoconjunctivitis9 and pharyngoconjunctival fever10 are adenoviral infections that can be associated with preauricular adenopathy.

Axillary Nodes The axillary nodes drain the upper extremity, chest wall, and upper lateral abdominal wall. Localized axillary lymphadenopathy most often represents a response to a pyogenic infection of the upper extremity. B. henselae infection is commonly associated with localized axillary lymphadenopathy after a cat scratch on the arm, and rat-bite fever due to Spirillum minor can cause axillary node enlargement and tenderness.11 Regional cutaneous tuberculosis (scrofuloderma) can be associated with axillary node enlargement.12 The most common complication of Calmette-Guérin bacillus vaccination is granulomatous lymphadenitis of the ipsilateral axillary region, which sometimes heals spontaneously but may have a protracted course. Disease responds poorly to isoniazid therapy and can require surgical excision.13,14 This diagnosis should be kept in mind when evaluating internationally adopted children. Hidradenitis suppurativa is a defect of terminal follicular epithelium, often affecting obese adolescent women, which leads to recurrent polymicrobial axillary node abscesses, fistulas, and scarring.15,16 Noninfectious causes of axillary lymphadenopathy must be considered, because steadily enlarging nodes in the absence of an obvious focus are worrisome for lymphoma or other neoplasms. Rheumatologic diseases with inflammation of the wrists or finger joints also can produce axillary lymph node swelling, though this is not generally isolated to axillary nodes.

A

Epitrochlear Nodes The epitrochlear nodes drain the distal arm and the middle, ring, and fifth fingers. Isolated enlargement is typically a result of a local skin lesion infected with Staphylococcus or Streptococcus. Epitrochlear lymphadenopathy can also follow cutaneous inoculation of B. henselae, F. tularensis, or (rarely) Sporothrix schenckii. Chronic enlargement of epitrochlear nodes is known to occur in cases of secondary syphilis and can be a clue to congenital syphilis. Reports exist of Hodgkin disease presenting as epitrochlear adenopathy.

Iliac and Femoral Nodes Enlarged iliac nodes can be detected by deep palpation over the inguinal ligament, and enlarged femoral nodes by deep palpation below the inguinal ligament, on the leg in the fossa between the adductor and rectus muscles. Suppurative iliac adenitis is an important cause of retroperitoneal abscess and can result from abdominal trauma, appendicitis, urinary or genital tract infection, or infections of the lower extremity.17 Femoral adenitis results from superficial or deep infection of the lower extremity. S. aureus and S. pyogenes are most commonly implicated, but B. henselae can also cause prominent femoral or iliac lymphadenitis. Additionally, tickborne tularemia can involve the inguinal nodes. The child with iliac adenitis can manifest fever and limp, abdominal or hip pain, and spasm of the psoas muscle. The patient

B Figure 22-1. Two school-aged girls with granulomatous conjunctivitis and lymphadenopathy due to Bartonella henselae infection. One girl has associated preauricular lymphadenopathy (A) and the other has facial and anterior cervical lymphadenopathy (B). (Courtesy of Sarah Long, M.D.)

typically prefers to lie with the hip flexed and the thigh abducted, because hip extension is extremely painful. The distinction of iliac adenitis from appendicitis is facilitated by the lack of nausea and vomiting, and from pyogenic arthritis by demonstration of fuller range of motion of the hip on examination.18 Computed tomography is the favored imaging modality when more extensive evaluation of iliac lymph node abnormalities is required.

168

SECTION

B Cardinal Symptom Complexes

Popliteal Nodes The popliteal lymph nodes are difficult to palpate unless they are substantially enlarged. Consequently, popliteal adenopathy is only appreciated in association with severe pyogenic infections of the distal lower extremity or knee joint, or noninfectious diseases of the reticuloendothelial system.

Management and Therapy Definitive therapy of localized lymphadenopathy or lymphangitis requires identification of the most likely pathogen involved. The history of antecedent trauma, presence of a skin lesion or infection in the region drained by the involved node or nodes, along with rapid development of a large, warm, tender mass, is highly suggestive of acute pyogenic lymphadenitis. Needle aspiration of a fluctuant node can provide therapeutic benefit, rapid diagnosis (with use of Gram stain), and a sample for culture and antibiotic susceptibility testing. Children with nonfluctuant nodes whose symptoms are otherwise consistent with acute bacterial infection can be treated empirically and closely monitored for clinical response. Timely needle aspiration is increasingly important with the rise in community-associated methicillin-resistant S. aureus (CA-MRSA) infection. The child with significant fever or systemic symptoms should be further evaluated with a blood culture, complete blood cell count, and measurement of inflammatory marker (e.g., C-reactive protein). Treatment of acute lymphadenitis should be based on etiologic agents (Box 22-2). Because acute pyogenic lymphadenitis can be staphylococcal or streptococcal in origin, the use of a b-lactamaseresistant penicillin (e.g., dicloxacillin, 25 mg/kg per day), cephalosporin (e.g., cephalexin, 25 to 50 mg/kg per day) or clindamycin (25 to 40 mg/kg per day) is appropriate initial therapy for an outpatient. It is imperative to have knowledge of one’s local antibiogram data as rates of CA-MRSA and clindamycin-resistant MRSA and methicillinsusceptible S. aureus (MSSA) are increasing. Linezolid, a newer oxazolindinone with excellent bioavailability, could be considered but should be reserved for MRSA infections that are clindamycinresistant. Erythromycin (40 mg/kg per day) is a less attractive potential therapy given the high rate of resistance and lack of data supporting its use for deeper infections. However, it can be an option if cat-scratch infection is considered; most experts would use azithromycin as data on efficacy exist. In patients who appear systemically ill and in young infants at higher risk for bacterema, initiation of parenteral therapy with similar agents (e.g., nafcillin, cefazolin, clindamycin) is indicated. Nodes with significant suppurative changes are likely to respond more promptly to therapy after percutaneous aspiration or incision and drainage. A total antibiotic (parenteral plus oral) course of 2 to 3 weeks’ duration is often required for complete resolution of bacterial lymphadenitis. In the setting of bacteremia or for life-threatening disease, vancomycin should be given (40 mg/kg per day) in addition to an antistaphylococcal beta-lactam and clindamycin. Localized lymph node enlargement lacking the characteristic features of acute pyogenic lymphadenitis and enlarged nodes that do not respond to antibiotic therapy merit careful further evaluation. Any clue regarding possible exposure to unusual pathogens should be pursued diagnostically. A Mantoux test for Mycobacterium BOX 22-2. Treatment of Common Infections Associated with Isolated Lymphadenitis EMPIRIC THERAPY Dicloxacillin, cephalexin, cefadroxil, clindamycin (consider local Staphylococcus aureus antibiogram) PERIODONTAL ABSCESS Penicillin V or clindamycin CAT-SCRATCH DISEASE Rifampin, macrolides, trimethoprim-sulfamethoxazole (TMP-SMX) TULAREMIA Gentamicin, doxycycline, ciprofloxacin

tuberculosis should be performed, realizing that both M. tuberculosis and nontuberculous strains of mycobacteria can be associated with a positive test result, the latter typically an intermediate response. B. henselae infection is particularly prevalent in the pediatric population after cat exposure and can be evaluated by serologic testing. Catscratch adenitis is typically a self-limited illness requiring no antibiotic therapy; in some cases, however, antibiotic therapy is beneficial. Aspiration of suppurative nodes may be required, and careful examination and follow-up are essential to exclude unusual systemic complications, such as hepatic involvement and encephalitis (see Chapter 160, Bartonella Species). Tularemia can be diagnosed by serologic testing. Although the organism can be grown in culture, this should be requested rarely, as it is a hazard to the laboratory workers, who should be notified that the diagnosis is suspected so that appropriate precautions can be taken. Hidradenitis suppurativa can be temporized by clindamycin and isotretinoin therapy, although wide surgical excision with healing by granulation is considered most efficacious.15 In patients in whom history, associated physical findings, and screening laboratory tests fail to lead to a diagnosis, and one suspects but cannot confirm a diagnosis of neoplasm or an unusual infection, lymph node biopsy is indicated. In a series of 75 children who underwent excisional biopsy of a peripheral lymph node, nonspecific reactive hyperplasia was found in 55%, noncaseating granulomatous adenitis (e.g., cat-scratch disease) in 21%, lymphoreticular malignancies in 17%, and caseating granulomas in 7%.19 Studies also suggest that fine-needle aspiration, perhaps guided by ultrasound, is a safe and effective alternative to excisional biopsy for discriminating infection from malignancy in lymph nodes of children,20,21 although tissue histology is frequently nondiagnostic and malignancy cannot be excluded. Any specimen from an excised node or needle aspirate should be examined by Gram stain and acid-fast stain, and should be placed in culture for bacterial (unless tularemia is suspected), mycobacterial, and fungal pathogens; staining with Warthin–Starry reagents for Bartonella is sometimes performed but a positive result is not specific.

LYMPHANGITIS Lymphangitis refers to inflammation of subcutaneous lymphatic channels, typically in an extremity, and can represent an acute bacterial infection or a more chronic, indolent process secondary to fungal, mycobacterial, or parasitic pathogens.

Etiologic Agents Infections that cause lymphangitis are summarized in Table 22-2. Streptococcus pyogenes is the leading cause of acute lymphangitis. Uncommonly, other streptococci or Staphylococcus aureus can cause lymphangitis, as can Pasteurella species after cat or dog bites and Spirillum minor after a rat bite. When inoculated cutaneously, a variety of organisms are capable of producing chronic, nodular lymphangitic infection.22,23 They include the dimorphic fungus Sporothrix schenckii, Nocardia spp., Mycobacterium marinum and certain other mycobacteria, and Leishmania spp. In the case of Nocardia or nontuberculous mycobacteria, the antecedent history of trauma is often elicited; a foreign body, often wood, can accompany such presentation. Arthropod-borne filariasis due to Wuchereria bancrofti or Brugia spp., which is a problem in the tropics and subtropics, manifests as acute lymphangitic inflammation or chronic lymphatic obstruction with lymphedema.

Clinical Manifestations Acute bacterial lymphangitis can accompany cellulitis or can occur in association with minor or inapparent skin infection. The disease is

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Localized Lymphadenitis, Lymphadenopathy, and Lymphangitis

CHAPTER

22

169

TABLE 22-2. Causes of Lymphangitis Etiologic Agent

Exposure

Onset

Clinical Features

Streptococcus pyogenes (less commonly, Staphylococcus aureus)

Localized skin infection

Acute (< 24–48 hours)

Red streaking, tender regional nodes, fever, chills, malaise

Pasteurella multocida

Dog or cat bite

Acute (< 24–48 hours)

Red streaking, tender regional nodes, fever, chills, malaise

Spirillum minor

Rat bite

Acute, after 1–3-week incubation period

Red streaking, tender regional nodes, fever, chills, malaise, headache

ACUTE BACTERIAL LYMPHANGITIS

NODULAR LYMPHANGITIS (LYMPHOCUTANEOUS SYNDROME)

Sporothrix schenckii (sporotrichosis)

Soil, peat moss, thorned flowers

Insidious, 1–12-week incubation period

Inoculation site papule, proximal spread of reddish nodules, systemic symptoms rare

Mycobacterium marinum (occasionally Mycobacterium chelonae)

Fish, shellfish, aquarium, ponds

Insidious, 2–4-week incubation period

Nontender nodules at inoculation site and spreading proximally, systemic symptoms rare

Nocardia brasiliensis

Soil botanicals

Insidious, 2–4-week incubation period

Localized chronic granuloma with nodular spread, drainage with “sulfur granules,” regional adenopathy

Leishmania brasiliensis (also Leishmania mexicana)

Travel to endemic area, sandfly bite

Insidious, 2–8-week incubation period

Primary ulcer with surrounding nontender nodules, no adenopathy or systemic symptoms

Wuchereria bancrofti (also Brugia spp.)

Travel to endemic area, mosquito bite

Acute or chronic

Acute: red streaking, adenopathy, fever rare Chronic: lymphatic obstruction, elephantiasis

recognized from the rapid appearance of tender, red, linear streaks proceeding from the site of cutaneous infection toward the regional lymph nodes. Tender lymphadenopathy is typically present, as are systemic symptoms such as fever, chills, and malaise. Bacteremia frequently develops. Nodular lymphangitis, also known as lymphocutaneous syndrome, most often manifests present on the hands and upper extremities; sporotrichosis is the prototypical infection of this type (Figure 22-2). The disease often begins with a shallow ulcer at the site of inoculation. Subcutaneous nodules varying from 2 to 20 mm in size appear indolently as the infection advances along the course of the lymphatic channel. These nodules are either freely mobile or adherent to the superficial skin. Whereas larger nodules often exhibit overlying erythema, they are rarely painful. Lesions sometimes ulcerate, with release of serosanguineous fluid. In contrast to acute bacterial lymphangitis, systemic symptoms of infection and regional adenopathy are typically absent in chronic nodular lymphangitis. As this is a chronic process, most patients do not present until several weeks into the clinical illness. Biopsy of such lesions may establish the diagnosis.

Differential Diagnosis Acute lymphangitis is a clinical diagnosis made in the febrile patient with tender, linear red streaking that extends proximally from a site of peripheral infection or traumatic inoculation. Thrombophlebitis is the major differential diagnostic consideration, but this condition lacks the characteristic inciting lesion (unless it is associated with an intravascular cannula) and the tender regional adenopathy associated with acute lymphangitis. The precise etiologic agent (usually Streptococcus pyogenes) can be identified with Gram stain and culture of a specimen from the cutaneous lesion or, not uncommonly, by culture of the blood. Acute lymphangitis develops in about 20% of patients with animal bites infected with Pasteurella canis (dogs) or P. multocida (cats).24,25 Finally, Spirillum minor infection should be considered in a child living in a crowded urban dwelling that is prone to rat infestation, because the superficial bite wound often completely heals during the 1- to 3-week incubation period before development of acute lymphangitis.

The origin of nodular lymphangitis is usually established through careful history of potential exposure to causative pathogens. The incubation period between inoculation and development of lymphangitic nodules can vary from 1 to 8 weeks, depending on the infecting organism.23,26 Sporothrix schenckii is found in soil and botanical debris, and infection follows contact with thorned material such as rose bushes or sphagnum moss.27 Mycobacterium marinum is a ubiquitous organism in marine and freshwater environments as well as aquariums and swimming pools. A reddish blue primary lesion develops at the site of trauma (e.g., the fingers and hands in “fishhandlers’ granuloma,” or elbows or knees in “swimming-pool granuloma”). Ulceration with purulent drainage can occur, and the disease can spread centripetally to produce sporotrichoid-like nodular lymphangitis.28 The rapidly growing soil and water mycobacteria M. chelonae can produce a similar constellation.29 Nocardia brasiliensis can gain access to lymphatics through minor wounds contaminated with soil and can produce nodular lymphangitis with suppurative ulceration, occasionally accompanied by regional adenopathy and systemic symptoms.30 New World cutaneous leishmaniasis is a protozoal disease seen in travelers to rural Central or South America who encounter the sandfly vector. After development of a shallow ulcer at the location of the insect bite, nodular lymphatic spread with superficial scale, discharge, and crusting is not uncommon.31 The histopathologic changes associated with chronic nodular lymphangitis are typically granuloma formation with epithelioid and giant cells and varying degrees of neutrophilic infiltration. Diagnosis of sporotrichosis is made through identification of fungal spores surrounded by eosinophilic material (“asteroid bodies”) or inoculation of a specimen from tissue drainage on Sabouraud agar. M. marinum is often not detected on acid-fast stain of lymphatic biopsy specimens, but culture on appropriate media grown at 30°C is highly sensitive. Nocardia species appear as delicate, beaded, branching, gram-positive bacilli that can be grown on simple media but may take several days to display characteristic colonial morphology.32 Pathognomonic sulfur granules are sometimes seen in exudate from lesions. Diagnosis of cutaneous leishmaniasis is made by direct visualization of amastigotes within histiocytes from a biopsy specimen or scraping.

170

SECTION

B Cardinal Symptom Complexes leishmaniasis often heals spontaneously with topical care, but therapy with pentavalent antimony should be used if lesions evolve to the mucocutaneous form.

CHAPTER

23

Respiratory Tract Symptom Complexes Sarah S. Long A

MUCOPURULENT RHINORRHEA

B Figure 22-2. One adolescent girl with a 4-month history of an ulcerating skin lesion on her wrist (A) and nodular lymophangitis (B). Sporothrix schenkii was isolated. Her only exposure to roses was from a florist. (Courtesy of Sarah Long, M.D.)

Therapy

Mucopurulent rhinorrhea, or purulent nasal discharge, denotes nasal discharge that is thick, opaque, and colored. It occurs at any age, usually as a manifestation of self-limited, uncomplicated viral upper respiratory tract infection (URI). Mucopurulent rhinorrhea is most problematic in children younger than 3 years because of: (1) protracted course and frequent recurrence, especially in those in outof-home child care;1 (2) parental concern about and misperception of etiology; and (3) overprescription of antibiotics by healthcare providers.2–5 Occasionally, this symptom is a clue to diagnosis of a treatable bacterial infection or underlying condition. Acute, sporadic mucopurulent rhinorrhea has an infectious cause and almost always is the manifestation of the uncomplicated “common cold” due to rhinovirus, coronavirus, or other circulating viruses.6 When the problem is chronic or recurrent, or persistent and unilateral, broader underlying anatomic, obstructive, immunologic, and allergic disorders are considered (Table 23-1).7–10 Onset in an infant younger than 3 months heightens suspicion of anatomic anomaly, ciliary dyskinesia, or cystic fibrosis. Accompanying sinusitis, otitis media, or pneumonia raises consideration of an immunologic deficiency (especially immunoglobulin deficiency or dysfunction, as in

TABLE 23-1. Causes of Mucopurulent Rhinorrhea

Because Streptococcus pyogenes is the predominant cause of acute lymphangitis, penicillin is the preferred initial treatment. Children with mild disease can be treated with oral penicillin V (25 to 50 mg/kg per day). Those with prominent systemic symptoms have a high risk of concurrent bacteremia and should receive intravenous therapy (penicillin G, 100,000 to 250,000 U/kg per day). Penicillin is also the drug of choice for Pasteurella lymphangitis and Spirillum minor ratbite fever. Lack of familiarity with the syndrome of nodular lymphangitis often leads to delays in correct diagnosis and inappropriate antibiotic therapy directed at pyogenic bacteria. Conservative measures, such as local application of a heating pad, may contribute significantly to resolution of lesions associated with sporotrichosis, M. marinum infection, or cutaneous leishmaniasis. Itraconazole (100 to 200 mg/day) has become the drug of choice for lymphocutaneous sporotrichosis, supplanting saturated solution of potassium iodide (SSKI) because of a lower toxicity.27 Treatment should be continued for 4 weeks beyond resolution of lesions (2 to 3 months total). Antimicrobial agents (e.g., trimethoprim-sulfamethoxazole, minocycline, rifampin plus ethambutol) are variably effective against M. marinum lymphangitis, and some excisional surgical debridement is often required. Nocardia infection generally responds readily to a sulfa drug; amoxicillinclavulanate is an option for patients allergic to sulfa drugs. Cutaneous

Chronic or Recurrent Acute Viral nasopharyngitis Bacterial sinusitis Acute otitis media Streptococcal nasopharyngitis Anaerobic bacterial nasopharyngitis (nasal foreign body) Adenoiditis Syphilis Pertussis

Underlying Conditions a

Allergy Medicationsa (antihypertensives, oral estrogens, aspirin and nonsteroidal antiinflammatory drugs) Pregnancya Hypothyroidisma Rhinitis medicamentosaa (a1-adrenergic agonists) Immunoglobulin deficiency Human immunodeficiency virus infection Cystic fibrosis Ciliary dyskinesia

a

Obstructing Lesions Polyps Congenital nasal anomalies (choanal atresia or stenosis, Tornwaldt cyst, deviated septum) Neuroembryonal mass (dermoid, encephalocele, glioma, teratoma) Tumor (hemangioma, angiofibroma, neurofibroma, lipoma, craniopharyngioma) Neoplasm (lymphoma, rhabdomyosarcoma, nasopharyngeal carcinoma)

Rhinorrhea is characteristically clear, but opaque white discharge is not unusual.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Respiratory Tract Symptom Complexes

hypogammaglobulinemia or human immunodeficiency virus (HIV) infection), neutrophil defect, cystic fibrosis, or ciliary dyskinesia. URIs are conspicuously severe in such instances, with recrudescence almost immediately after discontinuation of antibiotic therapy. Unilateral nasal discharge and obstruction should prompt investigation for a foreign body, mass lesion, or unilateral posterior choanal atresia. Table 23-2 shows differentiating features of important or common causes of acute mucopurulent rhinorrhea; allergic rhinitis is included because it is frequently part of the differential diagnosis in older children and adolescents.

Causes of Acute Mucopurulent Rhinorrhea Viral Nasopharyngitis In uncomplicated viral nasopharyngitis or rhinitis, nasal discharge is initially clear but can become white, yellow, or green (related to mucous secretions, dryness, blood, exfoliation of damaged epithelial cells and cilia, and leukocytic inflammatory response). Presence of

CHAPTER

23

171

high fever and persistence of discharge depend on the specific viral cause but are more common in uncomplicated infection than generally perceived. In a study of hospitalized children, more than 50% of those with uncomplicated adenovirus, influenza, parainfluenza, or respiratory syncytial virus infection had temperatures > 39°C, and 12% had temperatures > 40°C; height of fever in these children was not different from that in children with serious bacterial infection.11 Fever persisted for 5 days or longer in 37% of the children in the study; 20% to 30% of those with adenovirus or influenza A infection had fever for 7 days or longer. In another study, nasal discharge or congestion associated with uncomplicated URI persisted for 6.6 days in 1- to 2year-old children who were in home care and for 8.9 days in children younger than 1 year in daycare centers.1 In this study, 13% of 2- to 3year-old children in out-of-home childcare had symptoms for more than 15 days. The bacteriology of nasopharyngeal flora in children with uncomplicated viral respiratory illnesses, mucopurulent rhinorrhea, acute otitis media, and sinusitis has been evaluated and compared

TABLE 23-2. Differentiating Among Causes of Nasal Dischargea Viral Nasopharyngitis1,11–15

Acute Bacterial Sinusitis16,17,24

Streptococcal Nasopharyngitis25

Foreign Body-Related Rhinitis (Bacterial)18

Allergic Rhinitis10

HISTORY

Peak age

Peak in first 2 years after Any “new recruitment” into childcare or school

< 3 years

< 3 years

> 2 years; peak in adolescence

Onset

Dryness, burning in nose Insidious, with cough or nasopharynx day and night; occasionally, acute, febrile, toxic

Insidious; occasional acute, febrile, toxic

Insidious

Seasonal; precipitants

Associated symptoms

Nasal congestion, sneezing malaise

Malodorous breath; head or facial pain, edema

Malodorous breath ± hyponasal voice

Sneezing; nasal or palatal pruritus; tearing; snoring

Fever

Yes/no

No/yes

Low/high

No

No

Duration of discharge

3–8 days

≥ 10 days

> 5 days

Chronic

Chronic, recurrent

Red, excoriated nares; sometimes, acute otitis media

Periorbital swelling, facial tenderness; mucopurulent postnasal discharge

Anterior cervical lymphadenitis; impetiginous lesions below nose

Mouth-breathing

Transverse nasal or lower eyelid crease; periorbital hyperpigmentation; cobblestone conjunctivae or posterior pharynx

Character of discharge Clear or colored, watery Thick, colored or thick

Thick, colored

Unilateral, purulent, putrid blood-stained

Watery, clear, or white

Rhinoscopy

Hyperemic mucosa; dry or glazed early, edematous later; crusted discharge

Normal mucosa; discharge from middle meatus

Normal, hyperemic, or excoriated mucosa

Identifiable object (button, pit, nut), boggy mass (vegetable), or rhinolith

Pale or blue, edematous turbinates

DIAGNOSTIC TESTS

None; nasal smear shows polynuclear and mononuclear cells ± inclusion bodies, pyknotic epithelial cells

None; sinus radiograph (> 6 years of age)

Nasopharyngeal culture Rhinoscopy for streptococcus only

CAUSE

Multiple agents, depending on age and season

Streptococcus Streptococcus pyogenes pneumoniae, Haemophilus influenzae, Moraxella catarrhalis

Normal nasopharyngeal Allergens in predisposed facultative and anaerobic individual bacteria

THERAPY

Saline nasal drops, humidification; amoxicillin if acute otitis media

Amoxicillin; b-lactamase Penicillin V stable agent

Removal of obstruction; amoxicillin-clavulanate or clindamycin if tissue or sinus complication

PHYSICAL EXAMINATION

Associated findings

a

Superscript numbers indicate references.

Nasal smear shows goblet cells and eosinophils; skin test or radioallergosorbent test (RAST)

Avoidance; oral antihistamine/ decongestant; or topical corticosteroid; cromolyn

172

SECTION

B Cardinal Symptom Complexes

with that in normal children.6,12–23 Viral infection is associated with acquisition of new serotypes of Streptococcus pneumoniae and with temporally increased risk of acute otitis media.21 Quantitative, and some qualitative, differences in nasopharyngeal flora have been found in children with purulent nasopharyngitis (and uncomplicated viral upper respiratory illnesses), with excessive isolation rates reported for S. pneumoniae and Haemophilus influenzae,13,18 Peptostreptococcus spp., Fusobacterium spp., and Prevotella melaninogenica.18,19 The significance of such findings is unclear; isolation of such organisms may reflect exuberant proliferation in virus-induced inflammatory mucus or acquisition of a more robust specimen than is collected in healthy subjects. Furthermore, “high” rates of isolation of S. pneumoniae in 25% to 46% of subjects do not exceed those in normal young children when fastidious technique is used.22 Only two systematically performed studies on the course of mucopurulent rhinorrhea have been published. In one study, prospective evaluation showed that there was no difference in duration of illness or complications in children with clear or purulent nasal discharge.14 In a placebo-controlled, blinded study of 142 children 3 months to 3 years old with mucopurulent rhinorrhea of any duration, antibiotic therapy (cephalexin), systemic use of an antihistaminedecongestant, or both had no effect on the course or complications of mucopurulent rhinorrhea.12 In a small pilot study of 13 children younger than 2 years whose purulent nasal discharge had persisted for at least 10 days without improvement, amoxicillin-clavulanate (40 mg/kg per day divided into 3 doses for 10 days) was significantly associated with resolution of symptoms in comparison with placebo.15 Response to antimicrobial therapy does not necessarily validate an entity of bacterial nasopharyngitis, however; it seems more likely that children with such responses have an incomplete symptom complex of ethmoid sinusitis. Acute bacterial adenoiditis is postulated to be another cause of purulent nasal discharge when: (1) tympanic membranes are normal; (2) S. pyogenes is not found in culture specimens; and (3) radiographs show an enlarged adenoid shadow but no sinus abnormality.23,24 Critical study has not been performed to validate this entity. A comparison of clinical and radiographic assessments of adenoidal enlargement may be an important first step.25

Bacterial Sinusitis Mucopurulent rhinorrhea of 10 or more days’ duration without improvement (or recrudescence after improvement) that is associated with daytime cough (which is frequently worse at night), or malodorous breath, facial pain, edema, headache, or fever is highly suggestive of paranasal sinusitis.17,26 Sinus radiographs show significant abnormalities in nearly 90% of children 2 to 6 years old with such findings (see Chapter 34, Sinusitis), and thus support the validity of clinical diagnosis without need for imaging.

Streptococcal Nasopharyngitis In children younger than 3 years, S. pyogenes has been associated with high fever, toxicity, and clear rhinorrhea or indolent infection with irregular fever and purulent nasal discharge, sometimes with associated excoriation of nares or tender anterior cervical lymphadenitis.13,18,25 In a streptococcal outbreak studied in a childcare facility for school-aged and young children, 26% of children younger than 3 years were affected, but pharyngitis was predominant, with no case of nasal streptococcosis.27

discharge is putrid and sanguineous and contains pieces of pseudomembrane.

Allergic Rhinitis Allergic rhinitis typically begins in the second decade of life, is uncommon before age 3 years, and may be rising in incidence in children between these ages. Diagnosis is suspected from the season, environmental precipitants, personal and family history of allergy, other associated symptoms and physical findings, and the response to specific interventions of avoidance or pharmacotherapy (see Table 232). Nasal secretions are usually clear or whitish. Diagnostic usefulness of nasal cytologic analysis is controversial.10,28 Relative eosinophilia (above 20%) is suggestive but not diagnostic of allergic rhinitis. The findings in vasomotor rhinorrhea, which is thought to be due to increased parasympathetic tone of the nasal mucosa, are similar to those in allergic rhinitis, except that symptoms of allergy and nasal eosinophils are absent. In severe allergic rhinitis, the inflammatory phase of response can cause accumulation of neutrophils and mononuclear cells.10

Management of Acute Mucopurulent Rhinorrhea In the vast majority of children with purulent nasal discharge (even if thick and green) of up to 1 week in duration, history and setting of illness, associated symptoms, and physical findings suggest uncomplicated viral URI. Antimicrobial therapy is inappropriate unless acute otitis media or sinusitis is diagnosed from additional findings (see Chapter 34, Sinusitis). Symptomatic therapy with saline nose drops or lavage facilitates expulsion of secretions and provides humidification. Its effectiveness reduces parental pressure to prescribe an antibiotic.29 If mucopurulent rhinorrhea persists for more than 5 days, and especially if some findings (e.g., anterior cervical lymphadenitis, scarlatiniform rash, excoriation around nostrils) or the epidemiology heightens the likelihood of group A streptococcal disease, nasopharyngeal specimens should be obtained for culture of S. pyogenes only. If findings are positive, penicillin V is given for 10 days. Routine culture for, or recovery of, S. pneumoniae, H. influenzae, Moraxella catarrhalis, or Staphylococcus aureus has no meaning and is an opportunity for misinterpretation. If mucopurulent rhinorrhea persists for more than 10 days without diminution, and especially if other symptoms are present, paranasal sinusitis is likely. Nasal mucosa is examined after use of single or second (5 minutes after the first) application of a topical vasoconstrictor such as oxymetazoline.15 If purulent secretions flow from the middle meatus, the diagnosis of acute sinusitis is confirmed. Signs of allergic rhinitis can also be confirmed. Radiographs may be helpful in patients older than 6 years to confirm sinusitis (or possibly to suggest adenoiditis). Pending definitive efficacy studies, many clinicians would treat children who have purulent nasal discharge of greater than 10 days’ duration as for acute sinusitis, usually with amoxicillin initially. When antimicrobial therapy is effective, substantial improvement of symptoms is expected within 48 to 72 hours. Therapy is continued for 1 week beyond complete resolution of respiratory symptoms.

STRIDOR

Other Infectious Causes Bacterial nasopharyngitis associated with nasal foreign body is typified by the young age of the patient and putrid, commonly bloodstained unilateral nasal discharge. Fever is unusual unless infection has spread to contiguous sinuses or distant sites. Prevotella, Fusobacterium, and Peptostreptococcus spp. as well as facultative flora are responsible. Nasal discharge can be the first manifestation of congenital syphilis and a later finding in nasal diphtheria, in which

Characteristics Stridor is a rough, crowing sound caused by passage of air through a narrowed upper airway, which includes the extrathoracic trachea, larynx, and hypopharynx. Because the extrathoracic airway normally narrows during the inspiratory phase of respiration, stridor due to upper-airway disease occurs during inspiration (or is more pronounced during inspiration if severe narrowing causes obstruction during

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Respiratory Tract Symptom Complexes

inspiration and expiration). Because the intrathoracic trachea normally narrows during expiration, obstruction of the intrathoracic trachea, such as that due to extrinsic compression of vascular ring or intraluminal obstruction of foreign body, inflammation, or tracheomalacia, causes a loud noise, acoustically like stridor, heard during both phases of respiration but more pronounced on expiration. Extrathoracic obstruction (inspiratory stridor) is associated with prolonged inspiration and underaeration of the chest, whereas intrathoracic obstruction (expiratory stridor or wheezing) is associated with prolonged expiration and overinflated chest. Stridor can be associated with mild tachypnea, but a respiratory rate > 50 breaths/minute should not be ascribed to upper-airway obstruction alone. The timbre of the stridulous sound provides a clue to etiology; for example, (1) the high-pitched, fixed, dry sound of congenital subglottic stenosis; (2) the wet, rhonchal changing sound of inflammatory laryngotracheitis; and (3) the low-pitched, vibratory, somewhat positional sound of laryngomalacia. Associated voice changes are useful in specifying disease as well. Vocal cord paralysis causes a weak, dysphonic cry; supraglottic obstruction, a muffled voice; and laryngotracheitis, hoarseness or aphonia, frequently with a barking cough.

Etiology Categorization of the setting and duration of stridor as acute, persistent, or recurrent or episodic provides a framework for considering likely causes (Table 23-3).30–34 Infectious agents cause most acute upper-airway obstruction, from intraluminal, epithelial inflammation or by encroachment on the airway by reactive or infected lymphoid tissue in parapharyngeal or paratracheal spaces. Fungal or viral tracheobronchitis must be considered when stridor occurs in an immunocompromised child; odynophagia and dysphagia are also commonly present.31 Congenital anatomic abnormalities are considered, especially in infants whose persistent stridor began neonatally. Acquired obstruction can have abrupt onset and an obvious cause (such as foreign-body aspiration or necrotizing tracheobronchitis in ventilated neonates) or more insidious onset and inapparent cause (such as expanding laryngotracheal papillomas or hemangioma or an extrinsic compressing mass). The younger the infant, the more likely that sudden obstruction, apnea, or feeding difficulties overshadow a singular complaint of stridor.

TABLE 23-3. Causes of Upper-Airway Obstruction and Stridora Acute

Persistent30

INFECTIOUS

CONGENITAL

Viral laryngotracheitis (croup) Bacterial tracheitis Epiglottitis, supraglottitis Peritonsillar, retropharyngeal, or parapharyngeal abscess Tracheobronchitis associated with immunodeficiency31

Laryngotracheal web, cleft, cyst, hemangioma Tracheal stenosis Vascular ring Laryngotracheal malacia Neuromuscular disorder Cystic hygroma

NONINFECTIOUS

ACQUIRED

Angioedema Foreign body Necrotizing tracheobronchitis in neonates32,33 Recurrent/episodic Spasmodic croup Gastroesophageal reflux34

Posttraumatic tracheal stenosis Foreign-body aspiration Mediastinal mass (tumor, lymphatic, vascular) Papilloma (perinatally acquired) Posttraumatic spinal cord, vagal or glossopharyngeal nerve, or vocal cord damage Bulbar neuropathy (infectious, postinfectious, malignant)

a

Superscript numbers indicate references.

CHAPTER

23

173

Clinical Features of Acute Infectious Causes Recognition, care to avoid precipitating sudden airway occlusion, and urgent, expert intervention to establish an airway when indicated are paramount to avert disastrous outcomes of acute upper-airway obstruction. Table 23-4 shows characteristic features of infectious causes of stridor and acute airway obstruction.35–45 Viral laryngotracheitis (infectious croup) or laryngotracheobronchitis due to parainfluenza viruses is by far the most common.35,36 Influenza viruses, respiratory syncytial virus, adenoviruses, and other viruses typically cause symptomatic disease elsewhere in the respiratory tract, but during epidemic seasons, stridor is the predominant feature in a minority of infected children. Bacterial tracheitis is usually a complication of viral laryngotracheitis (with concordant peak age and season) but can occur at any age or as a complication of oropharyngeal surgery.46 Staphylococcus aureus is the most common cause, followed by Streptococcus pyogenes; the role of anaerobic bacteria is less clear.43,46 With the universal use of H. influenzae b vaccine, epiglottitis is a rare cause of stridor; current cases of supraglottitis are more likely to affect the aryepiglottic region and to be caused by streptococci. Parapharyngeal and retropharyngeal infections in young children must also be considered; their incidence is increasing45,47,48 (see Chapter 30, Infections Related to the Upper and Middle Airways). The history surrounding the onset of stridor and the patient’s age and demeanor are the most helpful clues to the likely site and cause of infection. The child with viral laryngotracheitis usually has had 2 to 3 days of typical upper respiratory tract illness when cough worsens and stridor begins. The child with bacterial tracheitis has usually had a similar background illness and then has sudden high fever, toxicity, and rapid progression of airway obstruction. The young child with retropharyngeal abscess or adolescent with peritonsillar abscess has less stridor but refuses to swallow, has a muffled voice, and a guarded posture to maximize the oropharyngeal airway. Trismus is an expected and useful finding in patients with peritonsillar abscess as well as in some with lateral pharyngeal space infections of odontogenic origin.38 Epiglottitis and supraglottitis cause the patient to guard anxiously in a sitting posture with arms back, jaw forward, and chin raised (“sniffing dog”) to maximize “lift” of the epiglottis away from the airway. In contrast, subglottic, tracheal obstruction cannot be lessened by position; patients with laryngotracheitis or bacterial tracheitis thrash about with the anxiety of suffocation. The expected course and sequelae of acute infectious airway obstruction are shown in Table 23-5.49,50 Children with viral laryngotracheitis are less prone to sudden complete obstruction; hourly course is predictable by degree of stridor and adequacy of aeration; response to racemic epinephrine and corticosteroid therapy usually averts intubation. Establishment of an artificial airway is urgently required for almost all patients with stridor due to acute supraglottic and bacterial tracheal infection, and for many with retropharyngeal infection. The course of disease in children with bacterial tracheitis can be further complicated, because infection (and obstructive consequences) commonly extends for the length of the trachea and below.

COUGH Cough is a critical protective mechanism to expel particulate matter from the larynx and trachea as well as a cardinal sign of infectious and noninfectious respiratory tract and nonrespiratory tract disorders. Although the vast majority of coughs are related to self-limited infections, occasional life-threatening infectious and noninfectious causes may be overlooked unless the clinician adopts a disciplined approach. Careful assessment of a pathologic cough – its onset, duration, clinical context, and association with other findings as well as its specific timbre, pattern, and productivity – frequently predicts the site of pathophysiology and narrows the differential diagnosis to a limited number of entities. Table 23-6 provides a framework for assessment and lists the differentiating features of the various causes of cough.

174

SECTION

B Cardinal Symptom Complexes

TABLE 23-4. Differentiating Among Infectious Causes of Upper-Airway Obstructiona Viral Laryngotracheitis35–37

Supraglottitis38,39

Bacterial Tracheitis40–43

Retropharyngeal Abscess38,40,44,45

Peak age

1–2 years

3–6 years, any

2–4 years, any

< 3 years

Peak season

Late fall, late spring

Any

Late fall, late spring; any

Any

Prodrome

Viral illness

Uncommon

Viral illness

Uncommon

Onset of stridor

Gradual

Abrupt

Abrupt

Abrupt

Peak temperature (°C)

38–39

> 39

> 39

> 39

Predominant findings

Brassy cough, stridor

Toxicity, stridor

Toxicity, stridor

Toxicity, stridor

Associated findings

Bark, rhinorrhea

Sore throat, odynophagia, dysphagia, anxiety, drooling

Brassy cough, anxiety

Lethargy

Voice

Hoarse, raspy

Normal, muffled, mute

Hoarse, raspy

Muffled, mute

Position

Any; thrashing

“Sniffing dog”; still

Any; thrashing

“Sniffing dog”; still

Airway occlusion

Predictable from degree of stridor

Sudden

Sudden

Sudden

Response to racemic epinephrine?

Yes, with rebound

No

No or partial

No

Normal or low

High

Immature

Immature

Hypopharynx

Distended

Distended

Distended

Anteriorly displaced

Airway

Subglottic narrowing; edema cords

Swollen epiglottitis, aryepiglottic folds

Subglottic narrowing; irregular trachea ± intraluminal mass

Prevertebral soft-tissue mass with anterior displacement of airway (not valid sign if expiratory film, flexed neck)

Chest

Underaerated ± cardiomegaly

Underaerated ± cardiomegaly

Patchy parenchymal peribronchial infiltrate

Underaerated ± cardiomegaly

ENDOSCOPY

Red, edematous subglottis; crusting pseudomembrane

Red, edematous supraglottic structures

Red, edematous, eroded Bulging mass in posterior trachea and bronchi; pharyngeal wall; purulence purulence, pseudomembrane

CAUSE

Parainfluenza viruses (epidemic); other viruses (sporadic)

Streptococcus pyogenes, Streptococcus pneumoniae, Haemophilus influenzae b

Staphylococcus aureus, Streptococcus pyogenes, Streptococcus pneumoniae

HISTORY

PHYSICAL EXAMINATION

LABORATORY TESTS

Peripheral neutrophils RADIOGRAPH

Streptococcus pyogenes; Staphylococcus aureus; rare Streptococcus pneumoniae

a

Superscript numbers indicate references.

TABLE 23-5. Expected Course and Sequelae of Acute Infectious Upper-Airway Obstruction Viral Laryngotracheitis37,49,50

Supraglottitis, Epiglottitis

Bacterial Tracheitis41,42

Retropharyngeal Abscess

Artificial airway (% of cases)

< 20

> 90

> 75

≥ 75

Median intubation period

4 days

2 days

6 days

2 days

Airway occlusion after intubation

Rare

No

Yes

No

Death during hospitalization

No

No

Yes

No

Airway sequelae (% of cases)

degree of illness Diffuse crackles

Ill general appearance > respiratory illness Crackles, ± wheezes

Cough Well between paroxysms Ill only during cough

Hyperaeration, subsegmental atelectasis

Hyperaeration, diffuse alveolar and interstitial infiltrates

Diffuse interstitial infiltrates

Normal or perihilar infiltrate

Eosinophilia

Normal, eosinophilia, lymphocytosis, neutropenia Increases in IgG, IgA, IgM; thrombocytopenia Throat, bronchoscopy, lung biopsy, or urine culture

Lymphocytosis; eosinophilia unusual

HISTORY

Season Onset Illness in others Fever Cough Associated features PHYSICAL EXAMINATION

Clear

LABORATORY STUDIES

Chest radiograph

Hyperaeration, ± peribronchial thickening, ± diffuse interstitial infiltrates White blood cell count Normal or lymphocytosis Normal, lymphocytosis, neutropenia Other findings

Hypoxemia

Diagnostic tests

Nasal wash EIA, DFA, culture

Nasal wash EIA, DFA, culture; throat culture

Increases in IgG, IgA, IgM Conjunctival, NP DFA, EIA

NP DFA, culture, PCR

DFA, direct fluorescent antibody (test); EIA, enzyme immunoassay; Ig, immunoglobulin; NP, nasopharyngeal specimen; PCR, polymerase chain reaction; URI, upper respiratory tract infection. a Pertussis is included in this table because it should be considered in young infants with cough and respiratory distress, although pneumonia is characteristically absent.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Respiratory Tract Symptom Complexes

stridor, or choking spells, without regurgitation, can be a manifestation of gastroesophageal reflux in infants.34,56 Clues to this diagnosis are: (1) typical age of onset at 6 weeks to 6 months; (2) postprandial occurrence of cough; and (3) history of pneumonia. Diagnosis is best confirmed by esophageal pH study. Cough can be the result of irritation of normal airways by mucus or purulent secretions (e.g., postnasal discharge or sinusitis) or of hyperreactive airways by secretions, infection, environmental stimuli, or smoke. Dry cough and frequent throat-clearing are clues to irritation of postnasal secretions. Sinusitis and cough-variant asthma are the most common causes of chronic cough in children, even for those younger than 2 years (with a normal chest radiograph).57 Pertussis and tracheal anomalies are frequently missed diagnoses when protracted cough is incorrectly ascribed to sinusitis or coughvariant asthma. Sinusitis as a cause can usually be uncovered by noting its association with infectious prodrome, the occurrence of cough day and night, or associated symptomatology; radiographs or limited computed tomographic study frequently clarify a confusing situation (see Chapter 34, Sinusitis). Cough of asthma is suspected when there is family or patient history of allergies and symptoms are recurrent, are not associated with acute illness, are exaggerated at nighttime, or are provoked by exercise, cold, smoke, specific allergens, or pollutants. A diagnostic trial of bronchodilator therapy can be used in young children if the clinical history is compelling and other diagnoses are excluded or highly unlikely. Respiratory function is tested in older children, with a diagnostic trial of bronchodilator therapy if the result is abnormal, or methacholine challenge if the result is normal.58 Additional considerations in children whose cough is not explained by acute infection or allergic process include compression of the trachea by tumor mass, lymph nodes, or enlarged vessels; irritation of the diaphragm by an abdominal disease process; and irritation of the pleura or phrenic nerve by tumor, inflammatory fluid, mass, or blood. Cough is usually irritative (dry, occurring at end of inspiration, diminished by voluntarily decreased inspiratory excursion). Neuropathy, myopathy, and bulbar involvement in infectious, metabolic, and immunologic disorders or malignancy (especially neuroblastoma and rhabdomyosarcoma) are considered when symptoms are otherwise unexplained. Cough is almost invariably associated with other signs, such as gurgling, weak cry, hoarse or quiet voice, and stridor. Habit cough has a classic presentation, usually after an uncomplicated “starter” URI in an adolescent (commonly a girl). The cough is loud, rattling, resonant, and low-pitched, occasionally with canine or seal-like bark. It never awakens the patient from sleep. Others, but not the patient, are bothered by the cough. This diagnosis is not tenable in the presence of weight loss or systemic illness. Invasive diagnostic procedures and narcotic cough suppressants are inappropriate and ineffective, and further foster the family’s misplaced focus. Reassurance, redirection of focus, and frequent visits to the primary care provider for examination and caring support are curative.

TACHYPNEA AND OTHER SIGNS OF LOWER RESPIRATORY TRACT DISORDERS Tachypnea can be a voluntary or involuntary response to anxiety, fright, or pain; an abnormal breathing pattern related to central nervous system dysfunction; or the physiologic response to increased temperature or metabolic state. It is most usually the response to respiratory acidosis or hypoxemia of acute infection or the attempt to restore pH balance during metabolic acidosis (e.g., diabetes, salicylate poisoning, dehydration). Metabolic causes should not be forgotten, while the clinician pursues the much more likely primary pulmonary causes. Additionally, tachypnea can result from primary cardiac abnormalities (congestive heart failure, cyanotic congenital heart disease), pulmonary vascular abnormalities (cardiac shunts, capillary dilatation, hemorrhage, obstructed return to the heart, or infarction), impaired lymphatic flow (congenital lymphangiectasia, tumor) or

CHAPTER

23

177

pleural fluid collections (hemorrhagic, purulent, transudative, or lymphatic fluid or a misplaced infusion from a vascular catheter). Tachypnea is thought to be the best clinical predictor of lower respiratory tract infection in children. Reference values for normal respiratory rates have been reconfirmed in healthy and febrile infants and young children.59–62 Roughly, respiratory rates > 60 breaths/minute in infants younger than 6 months, > 50 breaths/minute in infants 6 to 11 months old, and > 40 breaths/minute in children 12 to 59 months old have a sensitivity of 50% to 85% for diagnosis of lower respiratory tract infection with specificity of 70% to 97%. A useful cutoff respiratory rate for febrile children 5 years of age and older might be 30 breaths/minute. For infants younger than 24 months, the younger the patient, the less likely that pneumonia is present if tachypnea is absent. Performance of a chest radiograph in febrile infants without an apparent focus of infection to exclude pneumonia “missed” by physical examination has extremely low yield in the absence of tachypnea.63,64 In one study, for infants younger than 2 months, respiratory rate of 60 breaths/minute, retractions, or nasal flaring had sensitivity for diagnosis of pneumonia of 91%.62 Other symptoms and signs associated with pneumonia, such as cough, are more sensitive but are nonspecific; nasal flaring, intercostal retractions, and cyanosis have less sensitivity (25%, 9%, and 9%, respectively) but high specificity (87%, 93%, and 94%, respectively).61 Grunting is an expiratory sound produced in the larynx when vocal cords are adducted to generate positive end-expiratory pressure (selfinduced PEEP) and increased resting volume of the lung. Its causes are myriad but never trivial. Grunting can be a sign of surfactant deficiency in the neonate, or of pulmonary edema, foreign-body aspiration, severe pneumonia, mediastinal mass or severe mediastinal shift from any cause, pleuritic or musculoskeletal chest pain, or myopericarditis or other cardiac abnormalities at any age.65 Care must be taken with sedation, positioning, or intubation of such patients; the sudden removal of the self-induced PEEP can cause hypoxemia and respiratory arrest. Adventitial respiratory sounds usually indicate lower respiratory tract disease, pulmonary edema, or hemorrhage. Wheezes are musical continuous sounds present predominantly on expiration and are a sign of airway obstruction. Widespread bronchiolar narrowing, as most commonly occurs with the inflammation of virus-associated lower respiratory tract infection, produces heterophonous high-pitched, sibilant wheezes of variable pitch and presence in different lung fields. Fixed obstruction in a larger airway, as from foreign body or anomaly, produces homophonous, monotonous wheeze. Rhonchi, sometimes also termed low-pitched wheezes, or coarse crackles, are nonrepetitive, nonmusical, low-pitched sounds frequently present on early inspiration and expiration; they are usually a sign of turbulent airflow through secretions in large airways. Fine crackles (the term preferred by pulmonologists for rales, which has a variety of meanings across languages)66 are high-pitched, low-amplitude, end-inspiratory, discontinuous popping sounds indicative of the opening of peripheral air–fluid interfaces. Fine crackle is the auscultatory finding suggestive of the diagnosis of pneumonia. Auscultatory abnormalities of crackles and wheezing have disparate diagnostic usefulness among various studies, depending on the categorization of bronchiolitis. Tachypnea is a more sensitive finding than crackles for bacterial pneumonia; wheezing is more sensitive than tachypnea for bronchiolitis. Diminished or distant breath sounds, dullness to percussion, and decreased vocal fremitus indicate peripheral pulmonary consolidation, pleural mass, or fluid collection. Tubular breath sounds (low-pitched sound of similar intensity throughout inspiration and expiration, as normally heard in the intrascapular area), dullness to percussion, and increased vocal fremitus indicate parenchymal consolidation, atelectasis, or the presence of another continuous tissue or fluid density abutting both a bronchus and the chest wall.

DIFFERENTIATING FEATURES OF PNEUMONIA Table 23-7 shows symptoms and signs of pneumonia in infants and children. Although fever, cough, and tachypnea are cardinal features,

178

SECTION

B Cardinal Symptom Complexes

any or all of them can be overshadowed or overlooked in patients who come to medical attention for pneumonia-associated stiff neck, abdominal pain, or chest pain or for nonspecific symptoms of illness as well as in infants with feeding difficulty. Classic symptoms of pneumonia reported in adolescents and adults are fever, chills, pleuritic chest pain, and cough productive of purulent sputum, with less noticeable tachypnea.67

referred to as causing “afebrile pneumonia,” this is a misnomer, because Bordetella pertussis infrequently causes lower respiratory tract abnormalities,70,71 and respiratory syncytial virus and especially other respiratory viruses frequently cause fever.11,68,72,73 A causal role for Ureaplasma urealyticum is not completely defined, because the situation is confounded by the asymptomatic presence of this organism in women and young infants. Pneumonia due to Pneumocystis carinii is probably confined to infants with severe debilitation or immune defects.

Pneumonia in Young Infants In young infants, acute infection with bacterial and nonbacterial respiratory tract pathogens frequently leads to lower respiratory tract infection. Except in the first few days of life, when pneumonia is due predominantly to bacteria acquired from the mother’s genital tract or to organisms acquired transplacentally, nonbacterial pathogens are overwhelmingly predominant.68 As perinatally acquired agents persist, community exposures increase, and maternally derived antibody protection wanes, the infant between 3 weeks and 3 months old is vulnerable to a unique array of lower respiratory tract pathogens.69 Clinical setting, specific symptom complex, and severity of illness in proportion to findings on physical examination aid distinction of likely causes and guide the diagnostic and therapeutic approach (see Table 23-8). Although the pathogens listed in Table 23-8 are frequently

Pneumonia in Older Infants, Children, and Adolescents A number of studies using complex diagnostic methodologies have confirmed the specific cause of pneumonia in 45% to 85% of cases.74–78 Viral etiologies predominate, and, currently, most are amenable to diagnosis. Table 23-9 categorizes the features of acute pneumonia in older infants, children, and adolescents by etiology. No single fact in history or finding on examination is unique for any agent, but when they are taken together, a working diagnosis emerges and guides intervention or further diagnostic testing. Chest radiography and laboratory tests are reserved for patients who are ill or whose clinical picture is not compelling for a category of etiologic agents. The efficacy trial and postmarketing studies of heptavalent

TABLE 23-9. Clinical Features of Acute Pneumonia in Children and Adolescents Bacteria

Virus

Mycoplasma

Tuberculosis

Age

Any; infants especially

Any

School age

Temperature (°C) Onset Others in home ill

Most ≥ 39 Abrupt No

Most < 39 Gradual Yes, concurrent; upper respiratory tract infection, rash, conjunctivitis Myalgia, rash, mucous membrane involvement

Most < 39 Worsening cough Yes, weeks apart; pharyngitis, “flu,” cough

Any; < 4 years and 15–19 years especially Most < 39 (unless empyema) Insidious cough Yes, persistent cough

HISTORY

Associated signs, symptoms Toxicity, rigors Cough

Headache, sore throat, chills, myalgia, rash, pharyngitis, myringitis Hacking, paroxysmal, usually nonproductive

Weight loss, night sweats (late)

Persistent cough Well ˜ no findings (± cough); ill ˜ findings No/occasional Most normal; or unilateral crackles ± dullness

Wet, productive

Nonproductive

Toxicity, respiratory distress Degree of illness > findings

Respiratory distress Degree of illness ≥ findings

Cough Degree of illness < findings

No/yes Unilateral, anatomically confined or no crackles; dullness, diminished or tubular sounds

No Diffuse, bilateral crackles, wheezes

No Unilateral, anatomically confined crackles; ± wheezes

Chest radiograph

Hyperaeration, patchy alveolar infiltrate or consolidation in lobe, segment, subsegment

Hyperaeration, interstitial infiltrate in diffuse or perihilar distribution; “wandering” atelectasis

Pleural fluid Peripheral white blood cell count (cells per mm3) Sedimentation rate > 40 mm/hour Sputum

No/yes ˜ large Majority > 15,000; neutrophils ± bands Usual

No/yes ˜ small Majority < 15,000; lymphocytes Infrequent

Patchy alveolar and/or interstitial infiltrate in single or contiguous, usually lower lobe(s), unilaterally; perihilar adenopathy No/yes ˜ small Majority < 15,000; neutrophils

Copious, purulent; neutrophils, abundant bacteria Sputum Gram stain, culture; blood culture

Scant mucoid; epithelial, Mononuclear cells Nasal wash, throat, bronchoscopy specimen for antigen detection, culture; acute and convalescent serology

Irritative or productive

PHYSICAL EXAMINATION

Predominant feature Degree of illness: respiratory finding Pleuritic chest pain Auscultation

LABORATORY STUDIES

Diagnostic tests

Infrequent Scant mucoid; mixed mononuclear cells/neutrophils Cold agglutinin; acute and convalescent specific serology; throat culture, antigen detection, DNA techniques

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Patchy alveolar infiltrate in single or contiguous lobes with disproportionate hilar adenopathy; or miliary or lobar consolidation No/yes ˜ small, large Majority < 15,000; neutrophils, monocytes Frequent Scant ˜ copious; neutrophils (if copious) Gastric aspirate; sputum stain and culture

Abdominal Symptom Complexes

conjugate pneumococcal vaccine infers Streptococcus pneumoniae as a relatively common cause of pneumonia with patchy or consolidative infiltrates.79,80 Urine antigen detection test in children with lobar pneumonia also supports the important role of S. pneumoniae.81 Currently, ascribing a causal role to S. pneumoniae is confounded by the findings of prolonged asymptomatic carriage and inconsistent serologic results among studies.82

CHAPTER

CHAPTER

24

179

24

Abdominal Symptom Complexes Robert S. McGregor

HEMOPTYSIS Hemoptysis, defined as coughing up of blood that originated below the larynx, is uncommon in children; most commonly, supposed episodes are due to a posteriorly draining nosebleed. Mechanisms of hemoptysis include bleeding from: (1) congenital or acquired abnormal bronchial or pulmonary blood flow, venous obstruction, or vascular abnormalities; (2) immune-mediated endothelial damage; or (3) infectious or traumatic erosion of tracheal, bronchial, or bronchiolar epithelium. Hemorrhage can be mild (tracheitis, tracheobronchitis) or massive (congenital malformations, foreign body, bronchiectasis, pulmonary hemosiderosis). Causes of hemoptysis in children are listed in Table 23-10. Infection is the most common cause of mild hemoptysis. Panton–Valentine leukocidinproducing Staphylococcus aureus pneumonia is specifically associated with hemoptysis.83 Epstein–Barr virus was implicated in a single case.84 For more severe hemoptysis, bronchiectasis associated with cystic fibrosis accounts for as many cases as all other causes combined.85 Rigid bronchoscopy, computed tomography, and magnetic resonance imaging are useful diagnostic modalities in most cases of hemoptysis. Digital subtraction angiography and, occasionally, cardiac catheterization or arteriography are required.

TABLE 23-10. Causes of Hemoptysis in Children Epithelial Damage

Vascular Abnormality/Damage

Acute infection Bronchiectasis (cystic fibrosis, immunodeficiency, retained foreign body) Trauma (airway or chest) Foreign body Tumor (primary airway or pulmonary, metastatic)

Congenital heart disease or pulmonary vascular anomalies (venous obstruction, arteriovenous fistulae) Congenital malformation (pulmonary sequestration) Autoimmune vasculitis (systemic lupus erythematosus, Wegener granulomatosis, inflammatory bowel disease, Goodpasture syndrome) Sickle-cell disease Pulmonary hemosiderosis Nonspecific endothelial damage (chemical, drug)

To simplify the clinical approach to abdominal symptom complexes, abdominal pain is usually classified as acute or recurrent abdominal pain (RAP). Acute abdominal pain demands rapid diagnosis and appropriate intervention so that catastrophic outcomes can be avoided.

ACUTE ABDOMINAL PAIN Signs and symptoms of medical and surgical conditions that cause acute abdominal pain have considerable overlap. Even though Scholer and associates1 determined that only 1.5% of 1141 nonscheduled healthcare visits for acute abdominal pain resulted in a surgical diagnosis, rapid diagnosis and intervention should always be a primary goal to avoid an adverse outcome. Cope,2 in a classic monograph, pointed out that the first principle in approaching the patient with acute abdominal pain is the necessity of coming to a “best,” albeit not “certain,” diagnosis because severe abdominal pain of 6 hours’ duration occurring in a previously well child is frequently caused by a condition of surgical importance.

History The history and character of the patient’s acute abdominal pain are elicited with specific consideration of anatomy, embryology, and physiology. Diaphragmatic irritation, for example, causes shoulder pain, because the diaphragm, a high thoracic structure embryologically, shares cervical nerve innervation with the shoulder. History of therapies already provided is elicited, and potential effects integrated. Anti-inflammatory agents, especially corticosteroids, can substantially alter expected clinical findings, and potent analgesics or pretreatment with antimicrobial agents can mask otherwise clarifying symptoms. Regimentation in history-taking is essential. The three features of pain of particular importance are location, migration, and radiation sites.

Location of Pain Pain over the entire abdomen suggests a diffuse peritoneal process. Pain relative to disease in the small intestine is chiefly felt in the epigastric and umbilical areas, and because innervation of the appendix is similarly derived embryologically, the initial pain of acute appendicitis is located periumbilically. Pain relative to disease in the large intestine is usually felt in the hypogastrium or over the site of colonic abnormality. Pain of pelvic structures is also appreciated in the hypogastrium.

Migration of Pain and Radiation Sites Migration of pain and sites of radiation are useful clues.3 The early epigastric pain of appendiceal obstruction is carried by visceral pain fibers. Once the inflamed appendix irritates or adheres to the abdominal wall, somatic pain fibers in the parietal peritoneum cause migration of pain to the right lower quadrant. Similarly, biliary colic begins with epigastric pain but moves to the right upper quadrant when the inflamed gallbladder contacts parietal peritoneum. Because the eighth thoracic nerve innervates both the bile ducts and the infrascapular area of the posterior thorax, pain of biliary colic is often perceived just inferior to the right scapula. Renal and ureteral colic radiates to the ipsilateral testicle. Vertebral pain, as in osteomyelitis, radiates to the corresponding site of abdominal innervation.

180

SECTION

B Cardinal Symptom Complexes

Associated Symptoms The presence, timing, and nature of associated symptoms, especially vomiting, provide important clues to diagnosis. Pancreatitis can cause severe, repeated vomiting, because the inflamed pancreas directly irritates the celiac nerve plexus. Bowel obstruction causes vomiting; the more proximal the obstruction, the more severe the vomiting. High small-bowel obstruction causes intractable bilious emesis early in the course, whereas distal small-bowel obstruction allows delayed onset of emesis with longer periods between episodes (which can progress to feculent emesis). Large-bowel obstruction is associated with lateonset emesis or no emesis. Nausea and loss of appetite can replace vomiting as a symptom in patients with less sensitive triggering of emesis. In children with acute abdominal pain, sudden loss of appetite heightens concern, whereas preserved hunger lessens concern.

Character and Relative Severity of Symptoms The younger the patient, the less helpful the character of the pain. A sense of well-being between waves of pain is helpful, however. The child with crampy pain from gastroenteritis is playful and active intermittently, whereas children with appendicitis, obstruction, or intussusception do not experience complete relief. Factors that either exacerbate or alleviate pain can be useful clues. The relative importance of symptoms is also helpful. Nausea, vomiting, and diarrhea are cardinal features of gastroenteritis, and abdominal pain is secondary. Abdominal pain is the initial and unremitting feature of acute appendicitis or peritonitis, with other symptoms being variable and less significant.

to overcome localized or generalized abdominal rigidity, elicits greater pain and rigidity unless disease is in the thorax. Rectal examination should be included in the examination for acute abdominal symptomatology. The rectal examination may identify heme-positive stool or mucus, a mass, fullness, or localization of tenderness. It may also implicate gynecologic disease. Percussion of the abdomen is generally not helpful in differentiating among abdominal processes; however, dullness suggests the presence of peritoneal fluid or helps delineate edges of the liver or spleen. Auscultation distinguishes ileus but rarely adds diagnostic specificity.

Other Findings Physical findings of importance outside the abdomen include: (1) evidence of respiratory distress (nasal flaring, retractions, and adventitial auscultatory findings); (2) rashes (HSP, gonococcemia); (3) vaginal discharge; (4) pelvic girdle tenderness; and (5) hip pain on testing range of motion. The patient with abdominal pain who holds one hip flexed and externally rotated is likely to have acute appendicitis or primary inflammation of the iliopsoas muscle.

Specific Causes and Approach Conditions in which acute abdominal pain is the cardinal feature are discussed here, with focus on early clinical approach and intervention. Abdominal pain as part of fever of unknown origin and malignancy is discussed in Chapter 16, Fever Without Localizing Signs. Specific diagnostic studies are discussed elsewhere.

Appendicitis Physical Examination The physical examination must not be supplanted by imaging studies or laboratory tests.

Vital Signs and Habitus Vital signs are often normal until the pathologic condition causing abdominal pain is advanced; however, an elevated respiratory rate (out of proportion to temperature) is a clue to thoracic causes of pain referred to the abdomen. Abdominal processes that cause splinting of the diaphragm lead to shallow tachypnea and the appearance of a respiratory tract condition. Acidosis related to compromised bowel or infection causes increases in both respiratory rate and tidal volume (Kussmaul breathing). Fever is not a discriminating feature, although high temperatures (39.5°C or higher) at the onset of abdominal pain in the absence of vomiting and diarrhea suggest a renal or pulmonary process or primary peritonitis. When the abdominal pain and high temperature are associated with vomiting, diarrhea, or both, primary infectious gastrointestinal disease (e.g., salmonellosis, shigellosis) is a primary consideration. The patient’s preferred position and degree of movement provide diagnostic information. The child with intussusception lies anxiously, anticipating a paroxysm of pain that causes the child to writhe. The patient with pain due to Henoch–Schönlein purpura (HSP) tosses and turns, trying to find comfort, whereas the child with appendicitis flexes at the waist, and the child with peritonitis lies motionless.

Examination of Abdomen Careful observation may reveal abdominal distention, swelling, or masses. Limited diaphragmatic movement implicates an upper abdominal process, including pancreatic or hepatobiliary disease. While distracting the child, the physician should palpate the abdomen with warmed hands and a light touch, beginning at the site most distant from the reported location of pain. Palpation can reveal a mass (intussusception, tubal pregnancy, ovarian cyst, malignancy, hydronephrosis) or a site of tenderness, or can confirm abdominal rigidity (sign of parietal peritoneal inflammation). Persistent palpation, as an attempt

Acute appendicitis in its classic, most common form can be diagnosed readily before appendiceal rupture. Wagner et al.4 described the characteristic sequence of symptoms (rather than the presence of any particular symptom) in acute appendicitis and recognized that the sequence reflects pathophysiologic events. Interruption of this order should increase suspicion of an alternative diagnosis. The expected sequential events or findings in appendicitis are shown in Box 24-1. Atypical localization of tenderness and pain can occur, depending on the position of the appendix. If the organ lies retrocecally and cephalad, the serosa abuts the iliopsoas muscle, causing pain that results in a preferred position of flexion at the waist with flexion and external rotation of the right hip. If the appendiceal tip is directed inferiorly, then pelvic, left lower quadrant, or urinary symptoms can predominate. Occasionally, pelvic appendicitis can inflame rectal tissue and cause painful defecation, spasms of diarrhea, or rectal obstruction. Presence of a mass or tenderness on palpation of the right or anterior rectal wall during rectal examination can clarify the diagnosis. Left-sided appendicitis, sometimes associated with thoracic situs inversus, can cause all of the presentations seen in right-sided appendicitis, but in mirror image.

BOX 24-1. Expected Sequence of Events or Findings in Appendicitis 1. Pain, usually epigastric or umbilical: Appendiceal obstruction and distention stimulate visceral afferent nerves of T8 to T10, referring pain to the epigastrium and periumbilical area 2. Anorexia, nausea, and vomiting: Further obstruction and distention of the appendix lead to colicky pain and reverse peristalsis 3. Abdominal tenderness: Serosal inflammation follows, with irritation of the parietal peritoneum. Pain shifts to somatic fibers, resulting in the localization of tenderness to deep palpation and, later, localization of pain, to the right lower quadrant 4. Fever: Arterial supply of the appendix is compromised, leading to gangrene and rupture. Fever precedes rupture and frank peritonitis and is usually mild (39°C or less) 5. Leukocytosis: Development of localized peritonitis triggers neutrophilia (a relatively late event)

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Abdominal Symptom Complexes

Appendicitis generally evolves from first symptoms to rupture in less than 24 hours.5 The diagnosis remains a clinical one, although increasingly, ultrasonography and helical computed tomography are gaining support as diagnostic adjuncts.6 Except in cases of altered anatomy, diagnosis and appendectomy are expected to be achieved before rupture. Infants and toddlers present a diagnostic challenge; preoperative ruptures occur in > 50% of cases in such patients because of their inability to communicate classic signs or symptoms and because of the relatively high incidence of acute gastroenteritis in this age group.7

Mesenteric Adenitis A diagnosis often made by surgeons during an otherwise normal laparotomy performed for suspected appendicitis, mesenteric adenitis has been illuminated with the use of improved imaging techniques. Typically, the inflamed nodes, often > 1 cm, cluster in the mesentery in groups of 5 to 8. The mesentery itself can also be inflamed and thickened. Mesenteric adenitis manifests clinically as symptoms of severe localized abdominal pain that mimics pain of other regional pathologic processes. Mesenteric adenitis has been associated with positive culture of sampled lymph node for Yersinia spp. and has preceded classic Epstein–Barr virus mononucleosis. Unfortunately, the diagnosis of mesenteric adenitis and of acute abdominal processes are not mutually exclusive, often making the diagnosis of mesenteric adenitis one of exclusion despite visualization of nodes on ultrasonography or computed tomography.7,8 Improper order of symptoms for acute appendicitis, milder pain, disproportionately high fever, and preserved appetite may permit judicious avoidance of surgery. Mesenteric adenitis occurs with appendicitis in up to 49% of cases, has been described in inflammatory bowel disease and acute pancreatitis, and occurs with multiple enteric and systemic viral infections.9

Pneumonia When abdominal pain is associated with high fever, cough, nasal flaring, tachypnea, respiratory distress, and abnormal auscultatory findings, the diagnosis of pneumonia is obvious. More subtle manifestations make differentiation more challenging, but careful attention to severity of fever and initial signs (especially respiratory rate, retractions, and nasal flaring), and meticulous auscultation over all lung fields permits the differentiation. The theory of referred abdominal pain secondary to diaphragmatic irritation has been refuted by the reports of isolated upper and middle lobe pneumonia causing similar pain syndromes.10 Typical bacterial causes of communityacquired pneumonia are expected.

Pyelonephritis Specificity of symptoms of urinary tract infection (UTI) is agerelated;11 in the neonate and infant, the disease can cause decreased appetite alone or intermittent fussiness (perhaps due to dysuria); the toddler often has nonspecific abdominal pain (presence of dysuria being variable). The older child and adolescent are most likely to have dysuria, flank pain, tenderness at the costovertebral angle, and hypogastric pain. Pyuria is not uncommon during acute appendicitis. Pelvic appendicitis, or pelvic abscess from any source, can cause dysuria and pyuria, mimicking UTI.

CHAPTER

24

181

Anatomic relationships predict the symptomatology of acute pancreatitis. The gland lies close to the celiac plexus and semilunar ganglion; consequently, pancreatic inflammation causes nausea, intractable vomiting (never feculent), and severe epigastric pain. The head of the pancreas is surrounded by duodenum, whereas the body overlies lumbar vertebrae, and the tail reaches the left flank; inflammation occasionally causes pain only in the left hypochondrium or flank. Because of the organ’s proximity to the diaphragm, pain can occasionally be referred simultaneously to the left scapula and the left supraspinous fossa (phrenic pain). Fever is commonly present, typically 38.5°C to 39.2°C; its presence does not presuppose bacterial superinfection. Epigastric tenderness is expected, but muscular rigidity is variable. Jaundice is common and is usually due to swelling of the head of the pancreas rather than to obstructing gallstones. Hemorrhagic pancreatitis, a life-threatening form of pancreatitis, can have either of two pathognomonic signs, bluish discoloration in the flank (Grey Turner sign) or around the navel (Cullen sign). Carbohydrate intolerance can occur as pancreatic islet cells are destroyed. Acute pancreatitis is a clinical diagnosis supported by elevated serum levels of pancreatic amylase and lipase and abnormal findings on ultrasonography or computed tomography. Edema and parapancreatic fluid collections are typical and should not be overinterpreted as abscess. Serum amylase concentration alone is neither sensitive nor specific. Combining the use of amylase and lipase tests increases sensitivity and specificity.13

Gallbladder Disease Acute cholecystitis and infection of the gallbladder are rarely seen in children, except when predisposing conditions exist (e.g., portoenterostomy for biliary atresia, obstructing anomalies). Cholelithiasis, the usual antecedent of cholecystitis in adults, occurs without cholecystitis in children. In two pediatric series consisting of 85 patients with gallstones, only 2 patients developed cholecystitis.14,15 Children with cholelithiasis typically have an identifiable precipitating cause, such as hemolysis, total parenteral nutrition, or adolescent pregnancy. The usual sequence of symptoms is: (1) fever; (2) colicky epigastric abdominal pain that shifts to the right upper quadrant; and (3) tenderness over the right upper quadrant. An elevated serum concentration of g-glutamyltranspeptidase and bilirubin (out of proportion to elevation of aminotransferase) is expected. Enterococci, Escherichia coli, and other Enterobacteriaceae and oropharyngeal flora are occasionally isolated from blood, hepatic biopsy, or as ascitic fluid specimen. Biliary dyskinesia is increasingly described in the pediatric literature.16,17 In one series of consecutive pediatric patients with cholecystectomies, biliary dyskinesia was the most frequent indication for surgery (58%), with gallbladder calculi being the indication in only 27%.16 Biliary dyskinesia symptoms mimic those of cholelithiasis, with right upper quadrant pain and fatty food intolerance. The diagnosis is considered when ultrasonography fails to identify gallstones despite a high clinical suspicion of gallbladder disease. Untrasonography may demonstrate gallbladder wall-thickening or sludge. The diagnosis is confirmed with hepatobiliary nuclear imaging scans using cholecystokinin stimulation. The spectrum of disorders of functional motility can blur the separation of acute from RAP.

Pelvic Inflammatory Disease Pancreatitis In six pediatric studies involving 277 patients, pancreatitis occurred in association with trauma (20%), infection (15%), biliary tract disease (14%), drugs (13%), and congenital anomalies of the pancreatobiliary duct (5%); 22% of cases were idiopathic. Pancreatitis frequently also occurs in association with endocrinopathies, with multiple organ failure, and with metabolic disorders.12 Alcohol consumption is rarely implicated in pediatric cases.

Adolescent women are at higher risk of pelvic inflammatory disease (PID) than adult women. Higher infection rates with Chlamydia trachomatis and Neisseria gonorrhoeae in general, combined with biologic factors such as immaturity of the menstrual cycle, lack of antibodies to sexually transmitted infectious agents, and extent of cervical ectropion, contribute to the explanation.18 Some women with PID are asymptomatic, and others have only mild or nonspecific symptoms or signs (abnormal bleeding,

182

SECTION

B Cardinal Symptom Complexes

dyspareunia, or vaginal discharge). Because of this wide variability, PID must be considered with even mild lower abdominal pain in a sexually active adolescent. Because of the difficulty of confirming PID and the possible consequence of long-term decreased fertility if the diagnosis is missed, a low threshold for diagnosis and treatment of PID is recommended. Diagnostic criteria are listed in Box 24-2.19 Classic physical findings are fever, lower abdominal pain, pelvic tenderness, or both, and vaginal discharge. The constellation of abnormal vaginal discharge, a tender mass on bimanual examination, and elevated erythrocyte sedimentation rate predicts laparoscopically proven PID; however, only 20% of patients with PID have the triad.

Henoch–Schönlein Purpura In its most common presentation, the following four features distinguish HSP: (1) purpura due to leukocytoclastic vasculitis, classically palpable and present below the waist (in young children, especially nonambulatory children, the distribution can be atypical – often involving face and scalp); (2) nephritis, with a spectrum from microscopic hematuria, with or without proteinuria, to renal failure with severe hypertension; (3) arthritis affecting larger joints and with pain out of proportion to synovial fluid accumulation; and (4) abdominal pain, caused by the leukocytoclastic vasculitis. The abdominal pain in HSP is colicky and usually midline; intestinal mucosal purpura causes gastrointestinal bleeding or, rarely, initiates intussusception. Abdominal pain can precede purpura by days (rarely weeks) in 14% to 36% of patients.20 Because many patients with HSP have fever and abdominal pain (which can be severe) at onset, intra-abdominal infections and appendicitis are often considered before the typical purpuric rash appears.

Enteric and Other Infections Enteric infections rarely cause abdominal pain and fever as cardinal features. In the relative absence of gastrointestinal symptoms, Salmonella, Shigella, Escherichia coli O157:H7, and Clostridium difficile infections can cause severe pain due to intestinal spasm before diarrhea; dysenteric stool with mucus and blood clarifies the pathophysiology. Campylobacter jejuni infection can mimic inflammatory bowel disease with myalgia and arthralgia in association with bloody stools. Infections due to Yersinia enterocolitica and Y. pseudotuberculosis most closely mimic acute appendicitis when they

cause mesenteric adenitis. Appendicitis has rarely been associated with shigellosis.21 Typhoid fever, visceral abscesses,22 intestinal granuloma in chronic granulomatous disease, yersiniosis, visceral Bartonella henselae infection, cryptococcal infection, Toxocara infection, and brucellosis can each manifest as fever with a predominant complaint of abdominal pain. Acute hepatitis can cause abdominal pain, but disproportionate nausea, right upper quadrant tenderness, and hepatomegaly provide clues to the correct diagnosis. Epstein–Barr virus infection can cause predominant abdominal symptomatology with severe splenic enlargement. Enterobius vermicularis is occasionally found in the appendix of a patient with acute symptomatology. Streptococcus pyogenes causing acute pharyngitis can also cause abdominal symptoms ranging from pain to nausea and vomiting. Pathophysiology is thought to be related to extracellular enzymes produced by the organism. One prospective study failed to identify abdominal pain as a positive predictor of S. pyogenes infection as a singular symptom or combined with other predictive factors.23 S. pyogenes can cause retroperitoneal abscess, necrotizing fasciitis, and tubo-ovarian abscess, all of which manifest as severe abdominal pain.

Intussusception Intussusception in childhood is most often idiopathic. The overall peak incidence of intussusception occurs in children 5 to 9 months of age. Typically, infants have no demonstrable intestinal lead point. Hypertrophy of Peyer patches or mesenteric lymphadenopathy may initiate intussusception. Adenovirus has been recovered in stool or mesenteric node cultures in idiopathic cases.24 Historically, the rotavirus vaccine of the late 1990s was implicated as causal.25 Older children are more likely to have a pathologic lead point, identified in 8% of cases in a 1998 review.26 Regardless of cause, the clinical manifestations of intussusception are similar. Midline abdominal pain occurs in paroxysms as each peristaltic wave advances the intussusception. An apathetic presentation is seen in approximately 5% of patients; children appear “drugged” and inattentive rather than writhing in pain. Primary intracranial disease or intoxication can be incorrectly pursued before a palpable abdominal mass or “currant jelly stool” focuses appropriate attention on the intestinal tract. It has been speculated that this peculiar presentation may be due to high levels of endogenous opiates released in response to the painful process of intussusception.

Volvulus BOX 24-2. Criteria for Clinical Diagnosis of Pelvic Inflammatory Disease Minimum Criteria (if no Other Cause of Illness is Identified) Include One of the Following Uterine/adnexal tenderness or Tenderness on motion of the cervix Additional Supportive Criteria (Enhancing Specificity of Minimum Criteria) Oral temperature >38.3°C Abnormal cervical or vaginal mucopurulent discharge Presence of white blood cells (WBCs) on saline microscopy of vaginal secretions Elevated erythrocyte sedimentation rate or C-reactive protein Laboratory documentation of chlamydial or gonococcal infection Specific Criteria Endometrial biopsy showing endometritis Imaging techniques demonstrating thickened, fluid-filled fallopian tubes with or without free pelvic fluid or tubo-ovarian complex Laparoscopic abnormalities characteristic of pelvic inflammatory disease From Centers for Disease Control and Prevention. Sexually transmitted diseases treatment guidelines 2002. MMWR 2002;51:48–49.

Volvulus, generally occurring in infants with congenital malrotation, is a life-threatening event. Abdominal pain is unusual without associated emesis, which becomes bilious in the acute presentation. Because vascular compromise is present, infection is a common secondary event. Treatment for septicemia is indicated but should never be considered as sole treatment in the ill infant with bilious emesis. Occasionally, postprandial pain can be a prominent feature of the subacute or chronic presentation of volvulus. Volvulus can occasionally occur in infants with apparently normal intestinal anatomy and in older children. Volvulus and internal hernia with strangulation should also be considered in children with a history of prior abdominal surgery, because adhesions can predispose to these entities.

CHRONIC OR RECURRENT ABDOMINAL PAIN Unlike acute abdominal pain, chronic or RAP requires an inclusive, thorough consideration of the medical, psychosocial, and family history. The approach can involve days to weeks of data collection with only selective use of the laboratory. RAP, as defined by Apley,27 required the occurrence of 3 episodes of pain severe enough to affect activities, over a period of 3 months and occurring during the year before investigation. Peak incidence of RAP is in 9- and 10-year-old girls. The pain is characterized as

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Abdominal Symptom Complexes

paroxysmal, periumbilical, and lasting less than 60 seconds. Its character is often vague but has been described as dull, crampy, or sharp but not temporally related to activity, meals, stress, or bowel habits. Prevalence of RAP in 1000 unselected schoolchildren has been reported to be as high as 25% in girls. Family histories of children with RAP showed a higher incidence of migraine, psychiatric disorders, and peptic ulcer disease than those of controls.28 In uncontrolled observations from a wide range of practice settings, socioeconomic status correlated directly with the prevalence of RAP.

Differentiating Causes Diagnostic studies of choice are a thorough history, physical examination (both between and during episodes of pain), and simple laboratory tests. Extensive use of laboratory testing usually fails to make a diagnosis and heightens the family’s fear and pursuit of missed disease. Distinguishing features of RAP of nonorganic origin are characteristic enough that the clinical diagnosis is not simply a diagnosis of exclusion (Table 24-1). In fewer than 10% of children whose RAP meets Apley criteria for RAP an organic disease is identified, which, if treated, eliminates the symptoms.29 In the largest follow-up study, tracking 161 patients over 5 years, only 3 of 161 patients were eventually found to have organic disease, which was Crohn disease in all 3.30

TABLE 24-1. Features of Nonorganic Versus Organic Causes of Abdominal Pain

24

183

Nonorganic Causes The pain of RAP is central in location, most often periumbilical, and vague in character. Associated phenomena such as headache, diarrhea, nausea, pallor, and sleepiness after attacks are common and do not predict organic disease. Fever with the first painful episode is not uncommon, and a few children have recurrent low-grade fever with each episode. Persistent fever or fevers > 38.3°C raise suspicion of possible infection or inflammatory process. Vomiting is common, but only rarely occurs with each episode. Recurrent vomiting with painful episodes, particularly if postprandial and associated with bloating, prompts concern about intermittent volvulus. The child must be evaluated at the time of an attack; an imaging study performed at this time may be the only means of clarifying the diagnosis. Children with RAP are “high-strung,” fussy, excitable, anxious, timid, or apprehensive compared with controls. Patients and their parents are described as overly conscientious. The child is indrawn and is more likely to express features of emotional disturbance.28 With open-ended history-taking, precipitating events can be identified at the onset of RAP in one-third of cases. Anxiety is prominent even in adults with a past history of RAP. In a study by Campo et al.,31 28 former RAP patients were significantly more likely than controls to describe anxiety symptoms and disorders, to demonstrate hypochondriacal beliefs, to have greater perceived vulnerability to physical impairment, to exhibitpoor social functioning, to be receiving current treatment with psychoactive drugs, and to have a family history of generalized anxiety. Within this study, as well as the Alfven study quoted earlier,28 there were trends suggesting associations between childhood RAP and lifetime psychiatric disorder, depression, familyhistory of depression, and migraine.

Organic Causes

Findings

Nonorganic Causes

Organic Causes

Pain Location Character

Periumbilical Dull, crampy

Peripheral Colicky, penetrating, burning, boring Progressive Associated with meals, or fluid bolus Nocturnal symptoms, daily, persistent

Pattern

CHAPTER

Not progressive Follows precipitating event in one-third of cases, normal between events, better on weekends

Associated signs

Normal abdominal examination, little objective findings

Retching, writhing Distended abdomen, abdominal tenderness Mouth ulcers, digital clubbing

Associated symptoms

Multiple, vague, often unrelated Headache, “dizzy,” fatigued Fever absent

Focused, one or two related symptoms, including fever, weight loss, poor growth, arthritis, diarrhea, vomiting, dysuria

Family history

Can be positive for depression, anxiety, migraine

Can be positive for pancreatitis, peptic ulcer, inflammatory bowel disease

Social history

Frequent school absence; high socioeconomic status

Screening laboratory tests

Normal ESR, CRP, complete blood count, and urinalysis

Anemia, high ESR and CRP Stool positive for occult blood Abnormal urinalysis

CRP, C-reactive protein; ESR, erythrocyte sedimentation rate.

Table 24-2 lists the organic disorders to be considered in the evaluation of patients with RAP. UTI is the most common organic cause. Obstructive uropathy is less frequent. When pain in the flank and an abdominal mass are present, abdominal ultrasonography should be performed; hydronephrosis is most likely. Gastroesophageal reflux has been demonstrated in a few studies to be more prominent than previously thought. A study of patients with atypical RAP demonstrated that 56% of consecutively evaluated patients had significant gastroesophageal reflux as detected by pH probe; 71% of cases were responsive to histamine receptor type 2blocking agents and prokinetic agents.32 Similarly, a Norwegian study prospectively used pH probe as part of a diagnostic scheme to evaluate children with RAP and demonstrated gastroesophageal reflux in 21%. This study did not describe effects of treatment of the gastroesophageal reflux on RAP symptoms.33 Inflammatory bowel disease can also manifest as RAP. Weight loss, poor linear growth, pubertal delay, digital clubbing, perirectal abnormalities (fistulae or skin tags), joint symptoms, and nocturnal bowel movement raise suspicion of inflammatory bowel disease. Interruption of linear growth can precede overt gastrointestinal symptoms and signs by years. Peptic ulcer disease can also cause RAP. A history of recurrent emesis and a pattern of nocturnal pain that awakens the child suggest peptic ulcer disease.34 Helicobacter pylori is the primary cause of peptic ulcer disease in adults35,36 and in children.37 Macarthur’s 1999 review points out the continuing lack of evidence linking RAP with H. pylori.38 However in some select populations, eradication of H. pylori is associated with resolution of RAP.39 Evaluation for H. pylori infection is currently limited to children with symptoms of peptic ulcer disease. Abdominal migraine, although rare and challenging to diagnose, is worth addressing specifically because of all the evidence-based medicine reviews regarding pharmacologic intervention, treatment with an antimigraine medication: pizotifen is the only intervention demonstrated to reduce days of abdominal pain. A small trial, but perhaps this will lead to further studies.40

184

SECTION

B Cardinal Symptom Complexes

TABLE 24-2. Chronic or Recurrent Abdominal Pain – Expanded Diagnosis Common Causes Psychophysiologic Recurrent abdominal pain Irritable-bowel syndrome Conversion reaction Task-induced phobia Gastrointestinal Nonulcer dyspepsia Gastroesophageal reflux Constipation

Less Common Causes Genitourinary Dysmenorrhea Pelvic inflammatory disease Mittelschmerz Gastrointestinal Inflammatory bowel disease

Genitourinary Urinary tract infection

Rare Causes Genitourinary Urolithiasis Tumors (ovarian, renal) Endometriosis Hematocolpos Gastrointestinal Angioedema Malrotation Cystic fibrosis Mesenteric cyst Recurrent pancreatitis Cholelithiasis Gallbladder dysmotility Recurrent intussusception Meckel diverticulum Chronic appendicitis Abdominal wall hernia

Neurologic Intracranial mass Radiculopathy Spinal cord tumor/injury Cardiovascular Chronic dysrhythmias Familial dysautonomia Superior mesenteric artery syndrome Miscellaneous Abdominal epilepsy Abdominal malignancy Abdominal migraine Acute intermittent porphyria Addison disease Collagen vascular disease Familial Mediterranean fever Heavy-metal intoxication Hyperthyroidism Wegener granulomatosis

Adapted from McGregor RS. Chronic complaints in adolescence. Chest pain, chronic fatigue, headaches, abdominal pain. Adolesc Med 1997;8:15–31.

Chronic pancreatitis, although uncommon, can mimic RAP. Measurement of serum amylase and lipase concentrations is justifiable in the evaluation of RAP. Celiac disease is on the list of organic causes of RAP; however, incidence of positive antibodies is no more frequent among children with RAP than age-matched controls.41

CHAPTER

25

Neurologic Symptom Complexes Geoffrey A. Weinberg and Mary M. Moran

Approach to Diagnosis and Management As seen in Table 24-1, the clinical features of RAP of nonorganic origin are sufficiently characteristic that after an open-ended history (including impact of the illness) has been obtained and a thorough physical examination has been performed, the diagnosis is often established. Screening laboratory tests to eliminate more common organic causes (UTI, inflammatory bowel disease, pancreatitis) can contribute to physician and family confidence in the diagnosis. Screening tests include urinalysis, urine culture, erythrocyte sedimentation rate, C-reactive protein (CRP) level, complete blood count and differential leukocyte count, stool test for occult blood, and serum albumin, amylase, and lipase measurements. Evaluation for gastroesophageal reflux may be moving to first-line evaluation. Endoscopy and radiographic evaluation are performed to follow up specific leads from the history or physical examination or abnormal screening laboratory findings, especially if pain is atypical (e.g., right upper quadrant pain and fatty food intolerance or atypical location of pain). Management of RAP must involve: (1) confirmation of the legitimacy of the patient’s complaints; (2) expression of empathy for the family’s concerns; (3) resetting of the family’s focus on wellness; and (4) normalization of the patient’s activities and the family’s response to the pain syndrome. Confident reassurance of the patient and family is the treatment for most children with RAP. The physician must speak confidently and insightfully (pointing out the implausibility of certain feared pathogens) and must refocus the family’s attention on the child’s good health. Assisting the family to gain insight into psychosocial influences on the patient’s symptoms is very important. Reassurance is coupled with suggestions for modifying stresses in the child’s environment and redirecting any excessive emotional demands on or unrealistic goals for the child. Thoughtful assessment of each patient is paramount; nonselective use of multiple consultants may be detrimental. Poor outcome has been associated with use of more than three consultants.42

The child manifesting symptoms referable to the nervous system is worrisome and requires a thoughtful approach and evaluation. Often the presence of an infectious disease is not readily apparent. This chapter focuses on the most common presenting neurologic symptoms and the features of the history and physical examination that are characteristic of infections and those that distinguish infectious from noninfectious causes.

HEADACHE Headache occurs in 35% to 70% of school-aged children and adolescents.1 It is severe enough to be brought to medical attention in a small fraction of cases. Most children evaluated in primary care settings or in a pediatric emergency department with an acute headache and a normal neurologic examination have an acute viral illness, sinusitis, or migraine.2–4 Chronic headaches that gradually but progressively worsen in frequency and severity over time are more ominous than acute single or acute recurrent headaches separated by periods of normalcy.2–4

History Important characteristics that aid in differentiating causes of headache are: (1) pattern, duration, severity, location, and frequency of pain; (2) associated symptoms, such as fever, sinus pain, and rhinorrhea; (3) family history, triggering events, and efficacy of medications; and (4) presence of any underlying condition that predisposes to infection (Table 25-1). Children who have headache coincident with other complaints in which headache is not the cardinal feature rarely have intracranial disease. Although fever is perhaps the most helpful clue in ascribing an infectious origin, its absence does not preclude serious infection. Fever is present in 95% of children with meningitis, but in only 30% of children with brain abscess.5

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Neurologic Symptom Complexes

CHAPTER

25

185

TABLE 25-1. Differentiating Features of Causes of Headaches in Children Meningoencephalitis and Meningitis

Brain Abscess

Sinusitis

Brain Tumor

Migraine and Tension Pseudotumor Cerebri

Fever Onset Location

+++ Acute Diffuse

++ Subacute Localized to site

++ Subacute Frontal or diffuse

+ Subacute Localized to site

– Subacute Diffuse

Frequency Duration

Single Hours

Daily Weeks

Daily Days

Daily Weeks

– Acute Unilateral, diffuse, or occipital Variably recurrent Hours

PREDISPOSING

Preceding illness; epidemic disease; seasonality

Mastoiditis, otitis, sinusitis; facial cellulitis; cyanotic heart disease; empyema; immunodeficiency; gramnegative bacterial meningitis

Allergic rhinitis

Morning severity; forceful emesis; awakens from sleep; increased severity with change of position

Family history

Obesity; menses; vitamin A; corticosteroids; Lyme disease

++a

+

–

+

–b

–

++a +++

++ ++

– +

++ –

–b –

+c –

Feature SYMPTOMS

CONDITIONS AND ASSOCIATED FEATURES

Daily Days–weeks

PHYSICAL EXAMINATION

Altered mental status Focal deficits Nuchal rigidity

+++, expected; ++, frequent; +, occasional; –, rare or not associated. a In meningoencephalitis. b Can occur with basilar artery or complicated migraine. c Especially abducens nerve palsy.

Persisting focalization of head pain in children is unusual and is commonly associated with primary intracranial disease. Additional clues to intracranial disease (in approximate order of importance) include: (1) abnormal neurologic examination (especially abnormalities in eye movement or gait); (2) papilledema; (3) headache worsened by cough, Valsalva maneuver, or change in position; (4) forceful vomiting after prolonged period of recumbency; (5) headache that awakens the child from sleep; (6) change in prior headache pattern, especially headaches of recent onset with progressive severity and frequency; (7) lack of family history of migraine; and (8) in the younger child, preference for lying in the knee–chest position.2–4,6 Questions regarding appetite, activity level, hydration, and mental status may also help identify the seriously ill child.

Physical Examination Physical examination is directed to exclude life-threatening intracranial disease and begins with evaluation of mental status. Nonspecific terms, such as lethargy and fussiness, should be replaced by notation of interaction with the environment, consolable irritability, verbalization, and sense of wellbeing. Specific responses to verbal stimulation should be observed and recorded. If the examiner can elicit a smile from an infant or engage an older child in conversation, serious disease is unlikely. Meticulous attention to abnormalities in vital signs provides critical assessment of diagnosis, pace of illness, and need for immediate intervention. Examples are tachycardia, hypotension, or orthostatic hypotension in the child with dehydration; toxic or septic shock; and bradycardia, systolic hypertension (wide pulse pressure), and slow, deep, respirations (Cushing triad) in the more extreme case. Overt signs of increased intracranial pressure and impending herniation, such as hyperventilation, Cheyne–Stokes respiration (pattern of progressive increase in depth and sometimes rate of breaths followed by apnea) or ataxic respiration (chaotic gasping and apnea), bulging fontanel, papilledema, miosis, mydriasis or anisocoria, must be identified quickly. Cutaneous flushing has been described as a presumed centrally mediated response to sudden elevation in intracranial

pressure.7 Presence of papilledema is a specific but insensitive sign of intracranial disease. It is rare in children with bacterial meningitis, given the relative rapidity of disease, but is notable in 60% of children with brain abscess.5 Abnormalities of cranial nerve function, asymmetry in strength, tone, or reflexes, gait changes, and papilledema help define an existing lesion; the vast majority of children with a brain tumor have some abnormality demonstrable on careful neurologic examination.2,4 Nuchal rigidity is present in a third of patients with brain abscess8 and in > 95% of children beyond the neonatal period with meningitis,9 but can occur with posterior fossa tumors as well. Fever, headache, and nuchal rigidity can occur with acute bacterial pneumonia, especially that involving the upper lobes; tachypnea is almost invariably present but may have been overlooked. Sinusitis is a relatively uncommon cause of headache in children (16% in one large series4), but in up to 25% of children with sinusitis, headache is the chief complaint.10 The constellation of nasopharyngitis, frontal location, and sinus tenderness usually identifies these patients.

Evaluation If examination reveals overt signs of increased intracranial pressure or focal deficits, brain imaging (computed tomography and/or magnetic resonance imaging) is urgently warranted as the first study. Electroencephalogram is not recommended in the routine evaluation of headaches in children.4 Much discussion has centered on whether brain imaging study must be performed prior to lumbar puncture in the child with probable bacterial meningitis. Reports in the literature of herniation after lumbar puncture in bacterial meningitis are retrospective in design or anecdotal, and often involve a study population including the most critically ill patients, who sometimes have had herniation even without lumbar puncture.11–16 The published incidence of herniation is 4–6%, although again, these estimates may be biased by study design and publication bias.11,12,17 It is clear that increased intracranial pressure accompanies bacterial meningitis16 and is a risk factor for herniation.

186

SECTION

B Cardinal Symptom Complexes

However the greater likelihood of herniation with meningitis may be a function of the disease process rather than the lumbar puncture. It is also worth noting that a normal result of brain imaging study does not exclude the possibility of herniation.14,17 Recent data suggests that brain imaging is overutilized in adults with suspected meningitis, and clinical feature may sufÀce to choose who may proceed to lmbar puncture without imaging.17a,17b Similar prospective pediatric data are lacking. Thus, in general, if meningitis is suspected in a child without papilledema or focal neurologic findings, a lumbar puncture can be performed without obtaining brain imaging. Additional important but less common contraindications to lumbar puncture include clinically important cardiorespiratory compromise in a neonate or older child; infection in the skin, soft tissue, or epidural area overlying lumbar puncture site; and severe bleeding diathesis.9 In the uncommon child with suspected bacterial meningitis and signs of impending herniation or focal neurologic signs, blood cultures are obtained, antibiotics are administered, and an urgent imaging study of the brain is performed. Lumbar puncture is postponed until signs of herniation (as judged by both imaging studies and physical examination) have resolved. A written and videographic review of lumbar puncture methods is available.17c

ALTERED MENTAL STATUS

Physical Examination The degree of alteration in consciousness should be measured specifically. Coma scales, particularly the Glasgow Coma Scale, at times fall short in assessing consciousness in children, because the scales require an adult level of neurodevelopment and often have a high degree of interobserver variability.18 Delineation of the patient’s neurologic examination is most useful. Identifying signs of increased intracranial pressure or focal neurologic deficit are of utmost importance. A record of normal blood pressure is important and excludes hypertensive encephalopathy. The remainder of the physical examination focuses on recognition of a toxic syndrome or identification of a systemic illness. Toxic ingestion is supported by abnormalities in vital signs, skin (flushing or sweating), mucous membranes (sparse or excessive saliva), and pupillary size and reactivity. Physical examination of the child with hepatic encephalopathy usually detects other obvious signs, most notably icterus and decreased liver span, and, in the case of Reye syndrome, hepatomegaly without icterus. The clinical presentation of intussusception may lack the familiar features of severe, paroxysmal, colicky abdominal pain and bloody stools, instead consisting only of progressive lethargy and obtundation, perhaps secondary to release of neuroactive endotoxins or central endorphins.19,20 A sausage-like mass in the right upper quadrant is sought. An infectious etiology of altered mental status is likely if there is fever, involvement of multiple mucous membranes (such as

Diagnostic Evaluation of the child with altered mental status focuses on quickly identifying the likely cause or causes and initiating appropriate treatment, which is usually determined empirically. The causes of altered mental status in children are shown in Table 25-2, and differentiating features in Table 25-3. Diagnostic Evaluation in the emergency department typically involves rapid elimination of hypoglycemia, abnormalities of electrolytes or renal function, and, perhaps, intoxications. The differential diagnosis at the time of consultation has usually been narrowed.

History Important features in the history include age, past medical history, presence of fever, convulsion, or antecedent illness, immunizations, length and progression of illness, medications in the home, length of time the child or adolescent was unobserved prior to the change in mental status, and, for the adolescent, any recent change in disposition at home or school.

TABLE 25-2. Causes of Altered Mental Status Intracranial

Systemic

INFECTION

METABOLIC ENCEPHALOPATHY

Meningitis Meningoencephalitis (enterovirus, HSV, WNV, other arboviruses) Postinfectious encephalitis (ADEM) Brain abscess Bartonella encephalopathy Mycoplasma pneumoniae encepha lopathy

Endogenous Hypoglycemia Hyperammonemia Hypercarbia Hypoxia Uremia Exogenous Acute poison ingestion Chronic heavy metal exposure

TRAUMA POSTICTAL STATE COMPLEX PARTIAL STATUS EPILEPTICUS INTRACRANIAL MASS CNS VASCULITIS

INTUSSUSCEPTION HYPOTENSION

ADEM, acute disseminated encephalomyelitis; CNS, central nervous system; HSV, herpes simplex virus; WNV, West Nile virus.

TABLE 25-3. Differentiating Features of Causes of Altered Mental Status in Children Feature

Encephalitis

Toxic Ingestion

Space-Occupying Lesion

Reye Syndrome

Predominant age

Variable

Toddlers, adolescents

Variable

4–12 years

Fever

++

+

+

–

Prodrome

Headache, fever, irritability, personality change

–

Headache

Upper respiratory tract infection days before protracted vomiting, stupor, coma

Seizures

++

+

++

–a

Focal neurologic signs

++

–

++

–

Other associated findings

Rhinorrhea, rash

Altered pupils or vital signs Œ, Ø sweating Œ, Ø saliva

++, frequent; +, occasional; –, rare; Œ, increased; Ø, decreased. a Can occur with markedly increased intracranial pressure.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Hepatomegaly

Neurologic Symptom Complexes

conjunctivitis, pharyngitis or other exanthem, diarrhea), abnormal lung findings, viral or rickettsial exanthem (see Table 25-2). Mycoplasma pneumoniae has rarely been reported to cause severe impairment of mental status with or without meningoencephalitis; most often in affected patients, there are also physical findings referable to the respiratory tract.21 The patient with central nervous system vasculitis may have no other signs of systemic involvement. Patients with Bartonella henselae-associated encephalopathy frequently have isolated lymphadenopathy or papular lesion at site of cat scratch or both; nearly half have an acute onset of seizures.22 Patients with toxic shock are expected to have global abnormalities only commensurate with hypotension, and no focal abnormalities. Patients with septic shock can have central nervous system abnormalities out of proportion to hypotension, presumably related to metabolic derangements.

CHAPTER

25

187

ATAXIA Ataxia is a disorder of impaired balance and poor coordination of intentional movement. It results from cerebellar dysfunction due to: (1) damage to either of the two afferent pathways (from the vestibular apparatus to the caudal vermis or from the spinal cord or brainstem to the cerebellum); or (2) damage to the motor fibers to or from the cerebellum. The differential diagnosis of ataxia in children is broad and includes infectious etiologies only when onset is acute, with the possible exception of brainstem encephalitis due to Listeria monocytogenes (see Chapter 132, Listeria monocytogenes).23 Rarely, Borrelia burgdorferi has been reported to cause acute onset of brainstem dysfunction followed, in some cases, by progressive myelitis.24,25 Differentiating features of acute ataxia are shown in Table 25-4.

History Evaluation Unless steps taken to this point determine a likely extracranial cause of the altered mental status, a brain imaging study is performed to identify focal lesions or evidence of intracranial hypertension and assess the size of ventricles. Lumbar puncture is performed if no mass lesion is identified. Opening and closing pressures are measured, and a specimen of fluid is sent for multiple tests (see Chapter 46, Encephalitis, Meningoencephalitis, and Postinfectious Encephalomyelitis). Not infrequently, these study findings are also normal, or show only a mildly elevated cerebrospinal fluid protein concentration and a mononuclear pleocytosis; the remaining differential diagnosis includes toxic encephalopathy and encephalitis. Serum hepatic enzymes and ammonia are measured; normal values exclude Reye syndrome. Electroencephalography may be helpful in determining a seizure focus and identifying periodic lateralizing epileptiform discharges (which increase the likelihood of herpes simplex virus encephalitis, but are not pathognomonic for it), or focal slowing; however, absence of abnormality does not narrow the differential diagnosis significantly. It is noteworthy that complex partial status epilepticus can manifest with only alteration in mental status. Close monitoring of neurologic status, so as to recognize the necessity to intervene to manage intracranial hypertension, and monitoring of cardiac function, for dysrhythmia associated with intoxications, are indicated. Empiric therapy for herpes simplex encephalitis is considered if no other diagnosis is likely, after appropriate specimens have been collected for culture and molecular identification.

History for a patient with ataxia focuses on: (1) the child’s age; (2) presence of fever, recent illness, immunization, trauma, medications in the home and ingestion or intoxication; (3) progression, duration, and frequency of the ataxia if intermittent; (4) presence of constitutional symptoms; and (5) family history of migraine headache. Fever and ataxia have rarely been reported as sole findings in patients with acute bacterial meningitis.26

Physical Examination In considering the differential diagnosis, the clinician should first distinguish cerebellar dysfunction and ataxia from other disorders of gait (muscular or neuromuscular disorders, primary hip or limb disease) and incoordination (neuromuscular or neurosensory disorders, chorea). The verbal child with middle-ear or inner-ear disease leading to disequilibrium complains of dizziness; this is not usually the case in the child with ataxia due to cerebellar or posterior column disease. The ataxic child is not weak. In the child with cerebellar ataxia, the Romberg test is positive with eyes both opened and closed. Standing with feet together and arms outstretched, the child falls toward the affected cerebellar hemisphere. Ataxia due to posterior column disease is associated with a positive Romberg test result only when the eyes are closed. The site of cerebellar disease can be localized to the cerebellum or part of the cerebellum involved. Damage to the cerebellar vermis

TABLE 25-4. Differentiating Features of Causes of Acute Ataxia in Children Feature

Acute Cerebellitis

Postinfectious Cerebellitis

Toxic Ingestion

Neuroblastoma

Basilar Artery Migraine

Benign Paroxysmal Vertigo

Predominant age

< 10 years

< 10 years

Toddler

< 5 years

Any

< 4 years

Fever

++

–

–

+

–

–

Onset

Acute

Subacute

Acute

Subacute

Acute

Acute

Frequency

Single

Single

Single

Single

Recurrent

Recurrent

Duration

Days–weeks

Days–weeks

Hours

Months

Minutes–hours

Seconds–minutes

Associated symptoms and signs

Vomiting; viral illness Resolving antecedent Altered mental status Opsoclonus, illness (e.g., myoclonus chickenpox)

Disturbed vision; vomiting; headache; vertigo; ataxia; dysarthria

Suddenly reaches for support; refuses to walk

Results of evaluation

CSF pleocytosis

Normal or modest ˜ Abnormal serum, protein, cells in CSF urine toxin screen

˜ Urine HVA, VMA; abnormal imaging of abdomen and chest

CSF, cerebrospinal fluid; HVA, homovanillic acid; VMA, vanillylmandelic acid; ++, frequently; +, occasionally; –, rare; ˜, increased.

188

SECTION

B Cardinal Symptom Complexes

results in ataxia of the trunk and gait, whereas damage to the cerebellar hemispheres results in ipsilateral limb ataxia and nystagmus. Hypotonia and dysarthria can occur when cerebellar injury is diffuse. Once disease is localized to the cerebellum, initial evaluation targets life-threatening causes. Central nervous system tumor, particularly brainstem glioma, is usually associated with chronic, intermittent ataxia but can rarely cause acute ataxia. Signs of increased intracranial pressure can be present, but are not invariably so. On the other hand, increased intracranial pressure from whatever cause can manifest initially as ataxia. The general physical examination should be thorough and should be directed at distinguishing etiologies (see Table 25-4). Fever and new or healing rash (e.g., chickenpox) can be clues to acute infectious or postinfectious cerebellitis. Careful examination of skin and scalp for ticks (wood or dog ticks) is warranted, because the initial signs of tick paralysis are ataxia and nystagmus; motor weakness and ascending paralysis follow within 48 hours.27,28 Middle-ear disease can result in vestibular dysfunction and ataxia. An abdominal mass may be palpable in the child with neuroblastoma; a paraneoplastic syndrome of opsoclonus-ataxia can occur in such a child and is usually acute in onset. Chronic ataxia and severe neurodevelopmental delay usually persist in these patients. It is noteworthy that a significant percentage of children with the opsoclonus-ataxia presentation of neuroblastoma have pleocytosis at presentation, which could easily be confused with acute infectious cerebellitis.29

Unless the ataxia is very short-lived, as with basilar artery migraine syndrome and benign paroxysmal vertigo, further evaluation is necessary, as summarized in Table 25-4 (see also Chapter 47, Cerebellar Ataxia, Transverse Myelitis and Myelopathy, Guillain–Barré Syndrome, Neuritis, and Neuropathy).

HYPOTONIA AND WEAKNESS Tone and strength are distinctly different, but intimately related, properties of skeletal muscle. Muscle tone is the least resistance generated against passive movement of the muscle and is best evaluated in the awake, relaxed state. Strength is the greatest force that a muscle can generate actively against an opposing force and requires the patient’s maximal effort. Hypotonia can occur without accompanying weakness when the upper motor neuron is the site of disease. Weakness, however, is almost invariably accompanied by a decrease in muscle tone. In children, hypotonia is a cardinal feature of disease almost exclusively in infancy, whereas weakness is a complaint in the toddler and older child. The spectrum of disease causing diminished tone and strength varies within these age groups, including infectious,30–32 postinfectious,33–40 psychologic, metabolic, malignant, and traumatic causes. For these reasons, the features and causes of hypotonia and weakness are considered separately in Tables 25-5 and 25-6.

TABLE 25-5. Differentiating Features of Causes of Hypotonia in Infancy Feature

Infant Botulism

Myasthenia Gravis

SITE OF DISEASE

Neuromuscular junction

Neuromuscular junction

Spinal Muscular Atrophy

CNS Injury/Genetic Syndrome

Peripheral Neuropathy

Myopathy

Anterior horn cell

Central nervous system

Peripheral nerve

Muscle

HISTORY

Prenatal, perinatal Normal

Decreased fetal movement, polyhydramnios; short umbilical cord, abnormal fetal lie, low Apgar scores can occur in all

Onset

1–12 months

Birth

Course

Birth; more obvious with time

Birth

Birth

Birth; more obvious with time

Deficits peak at Deficits peak at birth Death in infancy 1–5 days Recovery 3–12 weeks Recovery in passive form

Not progressive

Not progressive

Not progressive

Family history

–

Myasthenic mother in passive form

Autosomal-recessive

Variable

Variable

Variable

Other

Breastfed; constipation; loss of facial expression; poor suck; weak cry

Fatigability; loss of facial expression; poor suck

Preserved facial expression

–

–

–

–

–a

PHYSICAL EXAMINATION

General

Normal–flushed skin; – ptosis; poor suck; weak cry

Paucity of movement; Abnormal head size; frog-legged position; dysmorphic features expressive face

Cognition

Normal

Normal

Normal

Abnormal

Normal

Normal

Strength

Œ

Œ

Œ

Normal

Œ

Œ

Reflexes

Normal–Œ

Normal–Œ

Œ–Absent

Ø–Normal

Œ–Absent

Normal–Œ

Muscle bulk

Normal

Normal

Proximal atrophy

Normal

Distal atrophy

Proximal atrophy

Fasciculations

–

–

++

–

+

–

Sensation

Normal

Normal

Normal

Normal

Normal–Œ

Normal

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Neurologic Symptom Complexes

CHAPTER

25

189

TABLE 25-5. Differentiating Features of Causes of Hypotonia in Infancy—Cont’d CNS Injury/Genetic Syndrome

Peripheral Neuropathy

Myopathy

Oculomotor Œ; bulbar Oculomotor Œ; bulbar Facial muscles, findings; descending findings; fatigability diaphragm spared symmetrical paralysis, always including cranial nerves; autonomic signs

Apnea; seizures; abnormal fundi

–

Facial diplegia

Stool toxin assay

Brain imaging Chromosomes

EMG Nerve conduction study

EMG muscle biopsy

Feature

Infant Botulism

Other

EVALUATION

Myasthenia Gravis

EMG Edrophonium chloride

Spinal Muscular Atrophy

EMG Muscle biopsy

CNS, central nervous system; EMG, electromyography; ++, frequent; +, occasional; –, absent; Ø, increased; Œ, decreased. a Hepatomegaly in Pompes disease (glycogen storage disease IIa).

TABLE 25-6. Differentiating Features of Causes of Weakness in Older Children and Adults Guillain–Barré Syndrome

Transverse Myelitis

Paralytic Poliomyelitisb

West Nile Virus Acute Flaccid Paralysis

Conversion Spinal Cord Tumor Disorder

Peripheral nerve root

Spinal cord

Anterior horn cell

Anterior horn cell

Spinal cord

Psyche

Peripheral nerve and neuromuscular junction

Onset

Subacute

Acute or subacute Acute

Insidious

Acute

Subacute

Progression

Ascending

Subacute (50 h to maximum deficit) Paresthesias, leg weakness, then sensory deficit

Weeks–months

Variable

Ataxia, then ascending paralysis

Duration Recovery

Weeks–months +++

Weeks–months 64% of cases

Biphasic paralysis Acutely 3–5 days after progressive with fever febrile meningoencephalitis Days–weeks Weeks–months Dependent on + degree/extent of paralysis

Variable ++++

Days Recovery if tick is removed

Predominant age

Adolescence

Adolescence

Any

Predominantly adults

Variable Dependent on tumor type, location, intervention Any

Paresthesias

++++

+++

++

–

Back pain

+

+++

20%

–

++, if dorsal columns involved ++ Dull, aching

60%

++

20%

–

+ Ascending, symmetrica

+++ +++ Usually symmetric; Patchy motor/sensory level 95% 20%

Feature SITE OF DISEASE

Tick Paralysis

HISTORY

School age– 2–5 years adolescence; rarely < 5 years ++ + ++

–

+

–

–

+++ Asymmetric

+ Below level of tumor

– Varies moment to moment

– Ascending, symmetric

+++

+++

–

–

PHYSICAL EXAMINATION

Autonomic dysfunction Fever Distribution

Bowel, bladder dysfunction Sensation abnormality Reflexes ASSOCIATED FEATURES

CSF

+ + (proprioception) Absent

95% (pain and temperature) 30% Ø initially 65% Œ overall

+

–

++++

++

++

Πor absent on affected side

Œ

Πif corticospinal tract involved

Normal

Absent

Antecedent URI or diarrheal illness (e.g., Campylobacter)

Abdominal pain; antecedent trauma 30%; URI 30%

Intense muscle pain; URI symptoms; nuchal rigidity; headache

Headache; Change in posture, Subconscious Dermacentor dyskinesia (e.g., Valsava accentuate stress; model for andersoni and D. parkinsonism); pain; abnormal gait symptom; variabilis ticks meningoencephalitis secondary gain

Ø Protein 0–few cells

Ø Protein Ø cells

Ø Protein Ø cells

Ø Protein Ø cells

Not applicable

Normal

CSF, cerebrospinal fluid; URI, upper respiratory tract infection; ++++, expected; +++, frequent; ++, occasional; +, rare; –, absent; Ø, increased; Œ, decreased. a The less common Miller–Fisher variant consists of ophthalmoplegia, ataxia, areflexia, and weakness that is less pronounced than in the classic form. b Paralytic poliomyelitis-like syndrome can also occur with other enteroviruses, especially enterovirus 71.

Normal

190

SECTION

CHAPTER

B Cardinal Symptom Complexes

26

Musculoskeletal Symptom Complexes Kathleen M. Gutierrez Musculoskeletal pain is a common presenting complaint in young children and adolescents. The usual cause of musculoskeletal pain is trauma or strain to joint, bone, or muscle. Often, the etiology is never determined and symptoms resolve with symptomatic treatment. The challenge for the clinician is to rule out serious disease, including infection, as a cause of symptoms. This chapter will review the illnesses responsible for three different categories of musculoskeletal symptoms – extremity pain, back pain, and chest pain – and the distinguishing features of each of these illnesses. Evaluation of the patient with musculoskeletal complaints begins with a careful history and thorough physical examination.

HISTORY Because the differential diagnosis of musculoskeletal pain is extensive, a comprehensive history must be obtained. Important information includes the acuity of onset of illness and presence or absence of fever. Symptoms that are acute, persistent, and accompanied by fever are more likely to be caused by infection. Chronic or subacute symptoms are often the result of noninfectious illnesses. Important information obtained in the history of the present illness includes the date of onset, location, description of the progression and severity of the musculoskeletal pain, and whether pain medication has been used and has been helpful. The clinician should elicit information regarding gait, ability to bear weight or move the affected area, joint or soft-tissue swelling, and presence of induration, erythema, and warmth. A history of trauma, previous or recent illnesses, growth and development, family history, travel, animal exposures, and medication use is important. A complete review of systems, including history of fever, skin or mucous membrane lesions, pharyngitis, eye inflammation, lymphadenoapthy, respiratory or cardiac symptoms, abdominal pain, vomiting or diarrhea, recent sexually transmitted diseases, or neurologic symptoms is crucial in determining whether a systemic disease is associated with the musculoskeletal abnormalities. Pertinent historical findings are discussed with each of the entities described below.

PHYSICAL EXAMINATION The primary focus of the physical examination is to pinpoint the location of symptoms in order to generate an accurate differential diagnosis and facilitate diagnostic testing. Examination of an infant or young child can by difficult, as they are unable to verbalize the exact location of discomfort. Since musculoskeletal pain is a result of disease affecting a variety of structures, including joints, bones, muscles, nerves, or blood vessels, careful examination of each of these systems is necessary. In addition, a full general physical examination should be performed in order to identify signs of systemic disease associated with the localized pain. Vital signs document the presence or absence of fever, hypothermia, or cardiovascular instability. Height, weight, head circumference, and general appearance of the child should be noted. Poor growth or weight loss could be consistent with inflammatory bowel disease (IBD) or other chronic inflammatory processes, malignancy, or chronic infections such as tuberculosis. Examination of the eyes may reveal conjunctivitis consistent with Kawasaki disease,

which may cause arthralgia and extremity pain. Mucous membranes should be examined for the presence of ulcers (IBD, Behçet, systemic lupus erythematosus (SLE)), an enanthem (group A streptococcal infection, enterovirus infection) or hyperemia (Kawasaki disease). Lymphadenopathy can be seen in juvenile idiopathic arthritis (JIA) and malignancy and is a feature of Kawasaki disease. The chest should be palpated for bone pain (malignancy, multifocal osteomyelitis, pleurodynia). Tactile fremitus is decreased with pleural effusion and increased over consolidated lung. Dullness to percussion is noted with pleural effusion or consolidation. Auscultation can reveal decreased breath sounds, rales or rhonchi or wheezing (pneumonia associated with hematogenous dissemination of bacteria to bones, joints, or muscles), or decreased breath sounds consistent with pleural effusion (occasionally seen with JIA and SLE). Cardiac examination may reveal a new murmur, raising concerns for endocarditis, or acute rheumatic fever (ARF). Muffled heart sounds are consistent with pericarditis. The abdomen should be assessed for tenderness, masses (psoas abscess, tumor), hepatomegaly, or splenomegaly (JIA, malignancy, endocarditis). Skin should be evaluated for rash. Careful examination of bones, joints, muscles, and nervous system is particularly important. The examiner should note how the patient holds the affected limb and whether areas of skin or soft tissue appear swollen or red. Gait and stance should be carefully observed. The bones and muscles in the affected limb should be carefully palpated. Passive and active range of motion of joints must be evaluated. Pain in the lower extremities can be referred from the back, pelvis, abdomen, or hip. A careful examination of the back and pelvis is necessary to rule out diskitis, vertebral osteomyelitis, pelvic osteomyelitis, or abscess. Bones of the spine and pelvis should be palpated and the spine assessed for abnormalities in flexion or extension. Skin overlying the vertebral body should be examined for the presence of a hemangioma or hair, associated with spinal dysraphism. Infection or structural abnormalities of the hip may result in pain referred to the knee. Therefore, careful examination of the hip joint should be performed in all children with lower-extremity pain. Neurological examination includes evaluation for presence and symmetry of deep tendon reflexes, evaluation for clonus, and sensory examination and estimation of muscle strength. Foot deformities may suggest underlying neurologic conditions.

LABORATORY EVALUATION AND IMAGING The laboratory and radiologic evaluation of musculoskeletal pain is guided by findings on history and physical examination. If infection is suspected, attempts should be made to obtain blood cultures and joint, bone, or tissue culture (depending on location of disease) prior to initiation of antibiotic therapy. Synovial fluid cell count and cultures are recommended, particularly if the patient presents with monoarticular arthritis, fever, and no other signs of systemic disease. Inflammatory markers (erythrocyte sedimentation rate (ESR) and Creactive protein (CRP)) and white blood cell (WBC) counts can be abnormal with infection, inflammatory disease, and hematologic malignancy. If a bone or joint disease is suspected, a plain radiograph of the affected area is generally the first type of image indicated. In general, both anteroposterior and lateral views should be ordered since with a single view abnormalities can sometimes be missed. Plain radiographs can identify fracture, chronic bone infection, joint effusion, soft-tissue swelling, tumor, and structural abnormalities of bone. Ultrasound is useful in identifying joint effusions, particularly in the hip. Magnetic resonance imaging (MRI) or radionuclide scans may be necessary to diagnose bone infections, particularly in the acute phase. These scans are also useful in establishing the diagnosis and extent of bone involvement with malignancies. Radionuclide bone scans are useful if multifocal disease is suspected. Use of other diagnostic tests depends on the differential diagnosis generated on the basis of the history and physical examination.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Musculoskeletal Symptom Complexes

LIMB PAIN The causes of limb pain in children are extensive and include the broad categories of infection, inflammatory disease, malignancy, trauma, and “orthopedic” problems (Box 26-1). These diseases may involve the joints, bones, muscles, or soft tissue of the affected limb.

Disease Involving the Joints Arthritis is a common cause of limb pain in children (Table 26-1). Infection of the joints can be caused by bacterial, and, less often, viral or fungal infection.

Pyogenic Arthritis Pyogenic arthritis is usually the result of hematogenous dissemination of bacteria to the joint space, and less likely the result of direct extension from infected bone, soft tissue, or muscle. The usual bacterial causes vary depending on age (see Chapter 81, Infectious and Inflammatory Arthritis); Staphylococcus aureus is the most common cause of pyogenic arthritis in all age groups. Pathogens typically causing infection in neonates should be considered in infants less than 2 months old: Neisseria gonorrheae should be considered in sexually active adolescents. Any joint in the body can be affected, but the joints of the lower extremity are most commonly involved. Pyogenic arthritis is generally monoarticular, but N. gonorrheae and occasionally Staphylococcus aureus can infect multiple joints. Infants are particularly at risk for infection of multiple joints. Pyogenic arthritis in children is characterized by acute onset of fever, joint pain, redness, and swelling. Physical examination shows decreased range of motion of the affected joint. Pyogenic arthritis of the hip may be difficult to diagnose, since redness and swelling of the hip joint are not readily apparent. Pain is

BOX 26-1. Causes of Limb Pain and/or Limp In Children INFECTIOUS DISEASES Bacterial arthritis Viral arthritis Mycoplasma arthritis Mycobacterial arthritis Fungal arthritis Lyme disease Disseminated Neisseria gonorrhoeae Hepatitis B infection Subacute bacterial endocarditis Psoas/obturator internus muscle abscess Spinal/paraspinal infection Osteomyelitis Skin/soft-tissue infection POSTINFECTIOUS/REACTIVE DISEASES Acute rheumatic fever Enteric/sexually transmitted infection Neisseria meningitidis Haemophilus influenzae b Immunization Transient (toxic) synovitis NEOPLASTIC DISEASES Bone/soft-tissue tumors Leukemia/lymphoma Metastatic neuroblastoma HEMATOLOGIC DISEASES Sickle-cell anemia Hemophilia

RHEUMATIC DISEASES Juvenile idiopathic arthritis (JIA) Systemic lupus erythematosus (SLE) Spondyloarthropathy Dermatomyositis VASCULITIS Henoch–Schönlein purpura Serum sickness Kawasaki disease ORTHOPEDIC CONDITIONS Trauma Congenital hip dislocation/ dysplasia Slipped capital femoral epiphysis (SCFE) Legg–Calvé–Perthes disease Osgood–Schlatter disease Chondromalacia patella Osteochondritis dissecans Foreign-body synovitis Physical overuse syndromes MISCELLANEOUS CONDITIONS Inflammatory bowel disease (IBD) Reflex sympathetic dystrophy Familial Mediterranean fever Behçet disease Sarcoidosis Neurologic/neuromuscular disease Fibromyalgia syndrome Psychogenic disorders

CHAPTER

26

191

often referred to the groin, thigh, or knee. To decrease intracapsular hip joint pressure, the child prefers to hold the affected leg in a flexed, abducted, and externally rotated position. Examination for prone internal rotation is helpful in diagnosing intra-articular hip pathology.1 The child is placed in a prone position with the pelvis flat on the table and with knees and ankles flexed and falling away from the body. The examiner looks for differences in internal rotation between the hips. Pyogenic arthritis of the hip must be diagnosed rapidly in order to prevent vascular compromise to the femoral head due to increased intracapsular pressure. Joint effusion may be detected by ultrasound. If pyogenic hip arthritis is suspected, prompt surgical drainage of infected fluid is indicated.

Transient Synovitis Transient synovitis is a self-limited inflammation of the synovium of the joint and typically involves a hip joint unilaterally (see Chapter 83, Transient Synovitis). Often it is seen in association with a viral respiratory or gastrointestinal illness and may be the result of either direct viral infection of the synovium or a postinfectious response. It is a common cause of limp and hip pain in children 18 months to 12 years of age (mean age 5 to 6 years). Although illness is self-limited and resolves with supportive therapy, differentiation of transient synovitis from pyogenic arthritis can be challenging. Typically, children with pyogenic arthritis are more likely to have a history of fever, be unable to bear weight, and have an elevated ESR, CRP, WBC count or have an increased hip joint space (> 2 mm) on plain radiographs.2,3 There is, unfortunately, overlap in clinical and laboratory findings between transient synovitis and pyogenic arthritis and therefore careful clinical follow-up is necessary until the diagnosis is clear.

Other Noninfectious Causes Reactive arthritis is inflammation of the joint space that occurs in response to an infection elsewhere in the body. The usual organisms associated with reactive arthritis include gastrointestinal pathogens such as Salmonella spp., Campylobacter spp., Shigella spp., Yersinia spp., Clostridium difficile, sexually transmitted infections such as Chlamydia trachomatis and Neisseria gonorrhoeae, and occasionally Streptococcus pyogenes infection. Noninfectious inflammatory causes of arthritis are generally more subacute or chronic in presentation. JIA is considered in children less than 16 years of age when symptoms of joint inflammation have been present for 6 weeks or more and other causes have been excluded.4 Categories of JIA include systemic arthritis, oligoarthritis, and polyarthritis.5 Systemic JIA typically presents with high fever, an evanescent rash, hepatosplenomagaly, generalized lymphadenopathy and occasionally pleuritis, and pericarditis. These symptoms can precede joint involvement by several weeks to months. Symmetrical involvement of large or small joints is typical. The WBC count, ESR, and CRP are usually elevated. Rheumatoid factor is negative and antinuclear antibodies are generally not seen. Oligoarthritis is characterized by involvement of one to four joints within the first 6 months of disease. Iridocyclitis is often associated with this form of disease. Polyarthritis is inflammation of five or more joints during the first 6 months of illness and is further divided into rheumatoidnegative or positive forms. SLE is a multisystem autoimmune disease that is associated with arthralgia and arthritis. Although children of any age can develop SLE, it most commonly develops in females around the time of puberty or during pregnancy. Children and adolescents can manifest with constitutional symptoms of fatigue, weight loss, hair loss, lymphadenopathy, and hepatosplenomegaly. Nonerosive painful, symmetric arthritis involving two or more joints is commonly seen. Criteria for diagnosis of SLE include the presence of at least four of the following: (1) malar rash; (2) discoid rash; (3) photosensitivity; (4) oral ulcers; (5) arthritis; (6) serositis (pleural or pericardial effusion); (7) renal disorder (proteinuria or cellular casts); (8) neurologic

192

SECTION

B Cardinal Symptom Complexes

TABLE 26-1. Diseases Associated with Joint Pain Etiology

Joints Involved

Associated Findings

Laboratory Studies

Ancillary Studies

Pyogenic arthritis

Any joint, more often joints of lower extremities, usually monoarticular

Fever, joint pain, redness, swelling, decreased range of motion

Elevated WBC, CRP, ESR. Synovial fluid WBC > 50 000 /mm3, predominantly PMN Blood or joint fluid cultures may be positive

Joint space widening. US shows fluid in joint space

Transient synovitis

Unilateral hip

Recent viral infection, afebrile CBC < 12 000/mm3 or low-grade fever, decreased ESR < 40 mm/hour range of motion of hip joint CRP < 1 mg/dL

US shows fluid in joint space, cannot differentiate infected versus noninfected fluid

Reactive arthritis

Usually large joints of lower extremities, occasionally small joints, wrists, and elbows. Sacroiliac joint involvement in adults

Enteric infection, sexually transmitted diseases, occasionally Streptococcus pyogenes, Neisseria meningitidis. Urethritis, mucous membrane ulcers, conjunctivitis may be present

Elevated ESR, CRP. Synovial fluid WBC < 50 000 cells/mm3

HLA-B27-positive

Large or small joints, symmetrically

High fever, evanescent rash, hepatosplenomegaly, lymphadenopathy

Elevated WBC, ESR and CRP. RF and ANA are negative

Oligoarthritis

Involvement of ≤ 4 joints within the first 6 months of illness

Iridocyclitis can be present

Polyarthritis

Involvement of > 4 joints within the first 6 months of illness

Systemic lupus erythematosus

Painful symmetric arthritis involving 2 or more joints, may be migratory

Fever, malar or discoid rash, photosensitivity, oral ulcers, serositis, neurologic abnormalities

Abnormal urinalysis (casts, proteinuria), leukopenia, anemia, thrombocytopenia. Positive anti-DNA or anti-Sm antibody

Radiograph can show pleural effusion. ECG can show pericardial effusion

Acute rheumatic fever

Painful, migratory polyarthritis

Fever, carditis, erythema marginatum, chorea, subcutaneous nodules

Elevated ESR, CRP Evidence of Streptococcus pyogenes infection (positive culture, ASO titer or antiDNase)

Abnormal ECG

Juvenile idiopathic arthritis (JIA): Systemic

Subdivided into RF-positive and RF-negative forms

ANA, antinuclear antibodies; anti-Sm, antismooth muscle; CRP, C-reactive protein; ECG, electrocardiogram ESR, erythrocyte sedimentation rate; HLA, human leukocyte antigen; PMN, polymorphonuclear cells; RF, rheumatoid factor; US, ultrasound; WBC, white blood cell count;

derangement; (9) hematologic disorder (anemia, leukopenia, lymphopenia, or thrombocytopenia); (10) presence of anti-DNA antibody or anti-Sm antigen; and (11) abnormal antinuclear antibody titer.6,7 ARF is a nonsuppurative complication of Streptococcus pyogenes infection.8,9 Diagnosis is based on clinical and laboratory findings. Exquisitely painful migratory polyarthritis typically involving the elbows, knees, and ankles is a major criterion for diagnosis of ARF. Arthritis occurs early in the course of illness and does not result in chronic joint disease. Other major criteria of ARF include carditis, erythema marginatum, chorea, and subcutaneous nodules. Minor critera of ARF include fever, arthralgia, elevated ESR, or CRP, and prolonged P-R interval on electrocardiogram. Two major criteria of ARF or one major and two minor criteria plus laboratory evidence of a preceding S. pyogenes infection are necessary to establish the diagnosis of ARF. Chorea alone can be used to diagnose ARF if other causes have been ruled out. Hip and knee pain can also be the result of noninfectious and noninflammatory structural abnormalities. Hip pain in boys 2 to 12

years old (mean 7 years) can be secondary to Legg–Calve–Perthes (LCP) disease. LCP occurs when blood supply to the proximal femoral epiphyses is disrupted, resulting in necrosis and infarction of the femoral epiphyses. The reason this occurs is not clearly understood. Some investigators have proposed that patients with risk factors for thrombophilia, such as those with factor V Leiden mutation or anticardiolipin antibodies, are more likely to develop LCP.10 The most common clinical complaint is limp or vague chronic pain in the groin or anterior thigh. Antalgic gait (shortened stance and prolonged swing phase), muscle spasm, and mild limitation of hip range of motion are noted on physical examination. Diagnosis is made by plain radiographs of the hip and pelvis. Slipped capital femoral epiphysis (SCFE) must also be considered.11 SCFE is posterior and inferior slippage of the proximal femoral epiphyses on the femoral neck. Sometimes there is a history of associated trauma or chronic pain. Pain is often referred to the anterior thigh or knee. Affected adolescents are often obese, but slips can occur in children of normal weight. Both hips are involved in a substantial number of cases. Physical signs of an acute slip include

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Musculoskeletal Symptom Complexes

severe pain when range of motion is attempted. A child with a chronic stable slipped epiphysis has decreased internal and increased external range of motion of the affected hip. Obligate external rotation of the hip is noted when the hip is passively flexed. Diagnosis is made by plain radiographs of the hip and pelvis. Knee pain can be caused by a variety of processes, including Osgood–Schlatter disease, patellofemoral pain syndrome, and osteochondritis dissecans. Osgood–Schlatter is an inflammatory disorder of the proximal tibial physis where the patellar tendon inserts on the tibia. Pain is most pronounced over the tibial tubercle. Osteochondritis dissecans of the knee occurs when an area of bone adjacent to cartilage separates. The cause is unclear but is possibly related to repetitive trauma.12 Children complain of nonspecific aching knee pain exacerbated by exercise. Tenderness of the anterior medial aspect of the knee can be noted. Plain radiography shows lesions most often in the medial distal femoral condyle. Patellofemoral knee syndrome is one of the most common causes of knee pain. Pain may be the result of trauma to the patella or peripatellar tissues, overuse or abnormal patellar tracking. Risk factors include quadriceps weakness and tightness of patellar soft tissue.13 Symptoms include dull aching pain during activity and with prolonged sitting, joint stiffness, and crepitus. Diagnosis is based upon physical examination. Other causes of joint pain in children include hematologic malignancies and trauma.

Diseases Involving Bone Infectious Causes Osteomyelitis is an important cause of limb pain in children (see Chapter 80, Osteomyelitis). Osteomyelitis in children is generally caused by hematogenous dissemination of bacteria to the bony metaphysis. Infection is usually well localized. Multifocal bone involvement is reported in infants. Most cases of osteomyelitis involve the long bones. However, a substantial number of bone infections involve bones in the pelvis. Pelvic osteomyelitis can be difficult to diagnose, since pain is often poorly localized and may be referred to the hip, groin, or buttock.14 Onset of fever and bone pain is generally acute, although in some cases fever can be low-grade with intermittent bone pain evolving over a few weeks. Children can come to medical attention with limp or extremity pain. Physical examination can reveal tenderness, redness, warmth, and swelling over the affected bone. Children with pelvic osteomyelitis have pain upon compression of the bones of the pelvis. A draining fistula may be seen in a patient with chronic osteomyelitis. Plain radiographs show soft-tissue swelling a few days after onset of symptoms. Changes consistent with bone destruction typically are not visible until 2 or 3 weeks after onset of symptoms. Bone scan or MRI is useful for early diagnosis of osteomyelitis. WBC count, CRP, and ESR are frequently elevated and blood culture is often positive.

Noninfectious Causes Noninfectious causes of bone pain include trauma (accidental and nonaccidental), stress fractures, “growing pains,” malignancies, and hemoglobinopathies. Trauma is generally diagnosed by history. “Growing pains” are characterized by bilateral leg pain, pain only at night, and no symptoms during the day. Malignant tumors associated with bone pain include osteosarcoma, Ewing sarcoma, and hematologic malignancies such as lymphoma and leukemia.15,16 Night pain severe enough to awaken a child from sleep, diffuse or migratory bone pain, and pain out of proportion to findings on physical examination raise concern for a malignant process. Vaso-occlusive crises in patients with sickle-cell disease manifest with fever and bone pain and can be difficult to differentiate from osteomyelitis. Finally, chronic recurrent multifocal osteomyelitis (CRMO) can be a cause of limb pain. CRMO is characterized by periodic episodes of bone pain and fever. It is associated with febrile neutrophilic dermatosis and pustulosis palmaris and plantaris.17

CHAPTER

26

193

Muscle Diseases Infectious Causes Pyomyositis is a bacterial infection of skeletal muscle that can cause limb pain, particularly in the lower extremities.18,19 Muscle infection probably occurs when bacteria are seeded hematogenously to injured muscle. The most common bacterial causes of pyomositis are Staphylococcus aureus and Streptococcus pyogenes, although a number of other gram-negative and anaerobic organisms also infect muscle. Presentation can be subacute and symptoms include fever, pain, swelling of the affected extremity, and limp, if the lower extremity is involved. Initial symptoms of pyomyositis include malaise and cramping muscle pain. The overlying skin can appear normal; the muscle involved feels firm on palpation. As the infection progresses, abscesses form in the muscle and the affected area becomes more tender, red, and swollen. Untreated, pyomyositis can progress to septicemia, shock, and, rarely, death. The WBC and ESR are generally elevated. The causative bacteria are usually isolated from abscess culture; blood cultures can also be positive. Ultrasound, CT, and MRI have all been used to diagnose and determine the extent of infection. The differential diagnosis of pyomyositis includes cellulitis, fasciitis, osteomyelitis, pyogenic arthritis, thrombophlebitis or deepvein thrombosis, muscle strain, malignancy, or compartment syndrome. Viral infection of muscles can also cause severe pain. Viruses most commonly reported to cause muscle inflammation include influenza viruses, enteroviruses, and human immunodeficiency virus. Myositis associated with influenza infection is manifest as acute pain and tenderness of the gastrocnemius and soleus muscles, making it difficult for the patient to walk.20,21 Symptoms appear as influenza infection is resolving. On physical examination localized tenderness and swelling of calf muscles are noted. Serum creatinine kinase levels are temporarily elevated. Rhabdomyolysis occurs in rare circumstances and is more common in girls and with influenza A infection. Parasitic causes of myositis include Trichinella spp, Taenia solium (csyticercosis), and Toxoplasma gondii22 (see Chapter 79, Myositis, Pyomyositis, and Necrotizing Fasciitis).

Noninfectious Causes Noninfectious illnesses that are associated with muscle inflammation generally present less acutely than infections of muscles. Dermatomyositis is nonsuppurative inflammation of striated muscle and is the most frequently diagnosed inflammatory myopathy in children.23,24 Onset of illness is insidious and symmetrical weakness of proximal muscle groups, rather than pain, is usually the presenting symptom. The illness is associated with a characteristic malar rash and/or blue purple discoloration of the upper eyelids. The skin over extensor surfaces of joints may be scaly and erythematous. Affected muscles are often stiff and tender to palpation. Associated clinical findings include joint contractures, dysphagia, and cardiac arrhythmias. Serum creatinine kinase levels are elevated. Diagnosis is confirmed by electromyography and muscle biopsy typically reveals perivascular inflammation. Fibromyalgia syndrome is a chronic pain syndrome characterized by musculoskeletal pain of 3 months’ duration in the absence of any other underlying medical condition in patients with normal laboratory studies.25 Criteria for diagnosing fibromyalgia syndrome are not well defined in children or adolescents. In adults, symptoms include diffuse musculoskeletal pain in at least three areas of the body for at least 3 months, and a physical examination showing pain over at least 11 of 18 previously defined tender points. Associated symptoms include nonrestorative sleep, fatigue, chronic anxiety, chronic headaches, the feeling of soft-tissue swelling, irritable bowel syndrome, extremity numbness, and modulation of pain by stress, weather, or physical activity.

194

SECTION

B Cardinal Symptom Complexes

Diseases of Soft Tissue and Fascia

BACK PAIN

Limb pain can be the result of cellulitis or infections involving the deeper levels of the dermis and superficial muscle fascia. The most serious of these infections is necrotizing fasciitis.26–29 Necrotizing fasciitis is a bacterial infection that spreads rapidly along fascial planes, causing necrosis, thrombosis, and extensive tissue damage. Bacterial causes of necrotizing fasciitis include methicillin-susceptible and -resistant Staphylococcus aureus, Streptococcus pyogenes, gramnegative bacteria, and anaerobic organisms, including Clostridium spp. and Bacteroides spp. Infection can be polymicrobial. Predisposing factors include trauma, burns, omphalitis, varicella infection, and immunosuppression. Symptoms are rapid in onset and the patient often appears to be in pain out of proportion to findings on physical examination. The patient can have high fever and tachycardia and appear very uncomfortable. Hypotension is seen as disease progresses. Blisters or bullae may be present over the affected area. Skin may appear dusky and the underlying tissue is exquisitely tender. Crepitus is present in some cases. Untreated, infection spreads rapidly, progressing to shock and multisystem organ failure. Diagnosis requires a high index of suspicion. WBC counts > 15 000 cells/mm3 and serum sodium less than 135 mmol/L may help distinguish necrotizing fasciitis from cellulitis.30 One group of investigators has developed a laboratory risk indicator scoring system for necrotizing fasciitis using CRP, WBC, hemoglobin, serum sodium levels, serum creatinine, and glucose to estimate risk of necrotizing fasciitis.31 This scoring system was useful in their patient population for detection of early cases of necrotizing fasciitis but needs prospective validation. MRI is the preferred radiologic modality for ascertaining involvement of the fascia. Biopsy of the affected area is useful in determining depth of infection. The differential diagnosis includes pyomyositis and cellulitis.

Back pain is a common complaint in children, particularly in adolescents,32,33 and can have multiple etiologies (Box 26-2). As with limb pain, the cause is often mild trauma or never determined. More serious causes of back pain include infection of the disk space or vertebral body, epidural abscess, or psoas muscle abscess (Table 26-2). Upper-

BOX 26-2. Causes of Back Pain in Children INFECTIOUS DISEASES Diskitis Vertebral osteomyelitis Spinal epidural abscess Sacroiliac joint infection Paraspinal myositis/abscess RHEUMATIC DISEASES Juvenile ankylosing spondylitis MECHANICAL DISORDERS Muscle strain Spondylolysis/spondylolisthesis Scheuermann disease Herniated nucleus pulposus HAMARTOMATOUS AND NEOPLASTIC DISEASES Osteoid osteoma Osteoblastoma Aneurysmal bone cyst Hemangioma Eosinophilic granuloma Ewing sarcoma Lymphoma, leukemia Glioma Neurofibroma Ganglioneuroma Neuroblastoma

TABLE 26.2. Diseases Associated with Back Pain Clinical Features

Clinical Findings

Laboratory Studies

Ancillary Studies

Diskitis

Children < 6 years. Subacute presentation. Nonspecific signs: refusal to walk, difficulty sitting, abdominal pain

Low-grade fever, limp, pain with hip flexion, loss of normal spine lordosis

Elevated ESR and CRP Blood cultures or disk space cultures are rarely positive

After 2–4 weeks plain radiographs show narrowing of disk space and irregularity of vertebral endplate

Vertebral osteomyelitis

Older children. Signs may be subacute

Fever, pain in back, chest, abdomen. Tenderness over involved vertebral body, paraspinous muscle spasm, loss of normal spine curvature. Neurologic deficits in 15–20%

Elevated ESR and CRP. Blood cultures positive in 30%

Plain radiographs show narrowing of affected disk space and lucency of the affected vertebral bodies. MRI delineates extent of disease

Spinal epidural abscess

Anterior abscess results from spread of vertebral body or disk space infection Posterior abscess results from hematogenous source. History can include surgery to spine, trauma, immune dysfunction, congenital anomaly of spine

Fever, back pain, malaise, limp. As infection progresses: radicular nerve pain, muscle weakness, bowel and bladder dysfunction. Soft-tissue swelling over back

Elevated WBC, ESR and CRP. MRI delineates extent of Blood and abscess cultures abscess, bone, or muscle positive involvement

Psoas muscle abscess

Primary abscess results from hematogenous source. Secondary abscess results from underlying bowel pathology

Fever, abdominal pain, hip pain, limp, and back pain. “Psoas” sign results from inflammation of psoas muscle (see text). If abscess is large, a mass may be palpated on exam of the abdomen

Elevated WBC, ESR, and CRP Abdominal US or CT may well confirm the diagnosis

CRP, C-reactive protein; CT, computed tomography; ESR, erythrocyte sedimentation rate; MRI, magnetic resonance imaging; US, ultrasound; WBC, white blood cell count.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Musculoskeletal Symptom Complexes

back or neck pain may be associated with meningitis, retropharyngeal abscess, or Ludwig’s angina.

Infectious Causes Diskitis Diskitis is inflammation of the intervertebral disk and the vertebral body endplates (see Chapter 82, Diskitis). Children under the age of 6 years are most often affected.34 The rich vascular anastomosis between vertebral bodies in the young child may account for the higher incidence of inflammation localized to the disk space and vertebral endplates with relative sparing of vertebral body. Vertebral body infection and destruction are more frequent in older patients. Diskitis usually involves the lower thoracic or lumbar spine. Staphylococcus aureus is the organism most frequently isolated when cultures are positive, although Kingella kingae, enteric gram-negative bacteria and Streptococcus pneumoniae have occasionally been isolated from blood or disk space cultures. Clinical findings in children with diskitis vary depending on age. Symptoms are subacute and increase in severity over several days. The young child is likely to present with nonspecific signs of irritability or refusal to walk.35 Older children may be able to localize pain to their back, or complain of abdominal symptoms, leg pain, or hip pain. On physical examination, the child may walk with a limp or refuse to walk. Hip flexion causes pain. Loss of normal lordosis of the spine may be noted. Plain radiographs of the back are normal initially, but after approximately 2 to 4 weeks, narrowing of the affected disk and irregularity of the vertebral endplates are observed. ESR and CRP are usually elevated. Blood cultures are often negative, but should be obtained as a positive culture will help guide antibiotic therapy.

Vertebral Osteomyelitis Infection of the vertebral bodies is relatively rare in children compared with adults.36 Infection is the result of hematogenous spread of bacteria or contiguous infection from adjacent muscle or soft tissue.37 The organism most commonly isolated is Staphylococcus aureus, although infections with gram-negative enteric bacteria are reported. If the process appears chronic, infection with Mycobacterium tuberculosis must be considered. Children with vertebral osteomyelitis generally have fever. Pain can be localized to the affected area of the back or referred to the chest, abdomen, and legs. Spasm of paraspinous muscles is often present. Loss of normal curvature of the spine can be noted and the neurologic examination is abnormal in some cases. Blood culture is positive in 30% of cases and WBC, ESR, and CRP are often elevated. Plain radiographs or MRI show involvement of the disk space and also destruction of vertebral bodies. If the blood culture is negative, biopsy of the affected vertebral body should be considered.

Spinal Epidural Abscess Spinal epidural abscess is a rare but potentially catastrophic infection in children.38 Most epidural abscesses are located posteriorly in the thoracic and lumbar spine and are thought to originate when bacteria from a distant site of infection reach the spine hematogenously.39 Anterior epidural spinal abscesses are often the result of extension of a vertebral body or disk space infection, or from extension of retropharyngeal or retroperitoneal abscesses. A patient often has a history of trauma prior to development of the abscess. Other predisposing factors include previous spinal surgery and immunodeficiency. Epidural abscesses have been reported in young children with congenital abnormalities such as a neuroenteric fistula40 or Currarino triad (congenital sacral abnormality, anorectal malformation, and a presacral mass).41 S. aureus is most often isolated but other gram-positive, gram-negative, and anaerobic organisms have been identified from culture of abscess material.

CHAPTER

26

195

Initial clinical findings include back pain, malaise, fever, and difficulty walking. As the abscess increases in size, sensory and motor abnormalities typically appear. Untreated, radicular nerve pain, muscle weakness, and bladder or bowel dysfunction may develop. Local tenderness may be noted and, in young children, soft-tissue swelling over the lower back may be observed. Blood cultures are positive in 60% of cases and the WBC count is usually elevated. Diagnosis is made by gadolinium-enhanced MRI. Patients should be managed with a pediatric neurosurgeon since surgical decompression is necessary in most cases.

Muscle Abscess Psoas muscle abscess can cause back pain, limp, hip pain, or abdominal pain.42,43 Psoas muscle abscess is classified as primary or secondary. Primary abscess is more common in children compared with adults and occurs without an obvious intra-abdominal focus of infection. Most primary psoas muscle abscesses are caused by S. aureus, although older reports emphasize the importance of M. tuberculosis. Secondary psoas abscesses arise as a complication of appendicitis or IBD. Inflammation of the psoas muscle results in the physical exam finding of the “psoas sign,” which is pain elicited on extension of the right thigh. However, some clinicians have not found this sign consistently useful, particularly in the very young, frightened, irritable child.42 If the abscess is large, an abdominal mass may be palpated. Inflammatory markers and WBC count are generally elevated. Diagnosis can by confirmed by ultrasound or CT.

Noninfectious Causes Spondylosis is a defect in the lamina of the vertebra between the posterior articular facets, usually at L4 or L5.44,45 This defect may occur as a result of trauma or genetic predisposition. When the defect is bilateral, the vertebral body can slip forward (spondylolisthesis). Patients complain of low-back pain and leg pain. Lumbar flexion and extension are limited and hyperextension of the lumbar spine exacerbates pain. Muscle spasm of the hamstring and paravertebral muscles may be present. Standing lateral and supine oblique radiographs of the lower back are diagnostic.45 Scheuermann’s kyphosis presents in late childhood.46 Etiology is unclear but hypotheses have included defects in ossification of the anterior portion of the vertebral body or vertebral body wedging due to abnormal biomechanical stress. Progressive wedging of the vertebral body produces kyphosis. Back pain is more common in adults than in adolescents. On physical examination an angular deformity of the back is noted when the patient bends forward. With postural kyphosis, the back deformity is more rounded and less angular. Diagnosis is made by anteroposterior and lateral radiographs of the back. Children with malignant spinal cord tumors present with back pain that is typically worse at night,35 abnormal gait, and scoliosis. Tumors are classified by location. Intramedullary tumors are generally located in the cervical region of the spinal cord. Lymphoma, neuroblastoma, and sarcomas may metastasize to extramedullary sites, either in intradural or extradural locations. Neurologic abnormalities, including muscle weakness and sensory deficits, are commonly seen.35,47 Deep tendon reflexes may be diminished at the level of tumor infiltration and increased below the level of the lesion. Bowel and bladder function may be impaired. Plain radiographs will detect vertebral body destruction. MRI is generally used to assist in diagnosis.

CHEST PAIN Chest pain in children is rarely caused by cardiac disease. The majority of children who present with chest pain have either a musculoskeletal cause of pain or the etiology is never determined (Box 26-3).48–50 Musculoskeletal symptoms are often a result of

196

SECTION

B Cardinal Symptom Complexes

trauma or strain. Infection involving the muscles or bones of the chest wall, lungs, heart, or mediastinum can be a cause of chest pain. Less common causes of chest pain include other types of pulmonary disease, gastrointestinal disease and, in fewer than 10% of cases, cardiac disease (Table 26-3).

Bony and Cartilaginous Causes Osteomyelitis of the ribs or sternum is not common but does cause chest pain. Extension of infection from the lung with organisms such as Aspergillus spp. or Actinomyces spp. can occasionally cause infection of the contiguous bones and soft tissue. Bacterial pyomyositis and necrotizing fasciitis occasionally involve muscle and fascial tissue of the chest wall. Enterovirus infection (usually coxsackieviruses or echoviruses) can cause inflammation of the intercostal muscles, upper abdominal muscles, and pleura. The patient often has nonspecific symptoms of a viral illness, fever, and acute onset of sharp chest or abdominal pain. Pain is worse with inspiration and cough. Diagnosis is made by isolating virus from respiratory secretions or stool.51,52 Costochondritis is an inflammatory process involving the costochondral or costosternal joints. Rarely, biopsy and culture of cartilage reveal a bacterial pathogen. Patients present with anterior chest wall pain and tenderness over the costochondral junction of affected ribs. In contrast to Tietze syndrome, swelling, erythema, and warmth generally are not noted. The sedimentation rate can be

BOX 26-3. Causes of Chest Pain in Children MUSCULOSKELETAL DISORDERS Costochondritis Muscle strain/trauma Slipping-rib syndrome Tietze syndrome Precordial catch Gynecomastia CARDIAC DISEASES Myocarditis Pericarditis Endocarditis Dysrhythmia Mitral valve prolapse Hypertrophic cardiomyopathy Aortic stenosis Coronary arteritis Coronary artery anomaly Coronary artery atherosclerosis RESPIRATORY DISORDERS Pneumonia Pleuritis Pleural effusion Asthma Pneumomediastinum Pneumothorax Cough GASTROINTESTINAL DISORDERS Gastroesophageal reflux Esophagitis Gastritis Esophageal spasm Esophageal foreign body Caustic ingestion MISCELLANEOUS CONDITIONS Shingles Pleurodynia Sickle-cell crisis Hyperventilation Cocaine abuse Psychogenic Idiopathic

elevated. Plain radiographs are normal or show soft-tissue swelling, and occasionally cartilage calcification and destruction. Tietze syndrome is distinguished from costochondritis by the presence of swelling in the area of pain.53 It is not often reported in children, being more common in the second to fourth decade of life. The etiology is unknown and symptoms resolve over time with supportive care. Usually only one site of swelling and tenderness is noted. Because of the rarity of this diagnosis in children, other causes of anterior chest pain and swelling, such as tumor, soft-tissue infection, or osteomyelitis, must be excluded.

Pulmonary Causes Pulmonary causes of chest pain include pneumonia54 and reactive airways disease. Patients with pneumonia are usually febrile, tachypneic, and may have pleural effusion, pneumothorax, or pneumomediastinum. Signs and symptoms of pulmonary disease include decreased breath sounds, dullness to percussion, rales, rhonchi, wheezes, and, rarely, chest wall crepitus. Pulmonary embolus is rare but should be suspected in patients with a central venous catheter in place, those who have methicillin-resistant Staphylococcus aureus infections elsewhere (due to Panton–Valentine leukocidin), those who are obese, have had surgery or trauma, have cancer, who are pregnant or on oral contraceptives, or have an inherited hypercoagulable state.55,56 History of previous venous thrombosis is also a risk factor. Symptoms of pulmonary embolus include sudden onset of dyspnea, pleuritic chest pain, cough, and hemoptysis. Findings on physical examination include tachypnea and tachycardia, rales, an audible S4, and a loud S2 heart sound. Chest radiograph findings are nonspecific. A number of tests have been utilized to diagnose pulmonary embolus, including serum D-dimer levels, ventilation/perfusion scanning, pulmonary angiography, helical CT and MR angiography.

Cardiac Causes Pericarditis with pericardial effusion caused by bacterial, viral, fungal, or inflammatory disease is often associated with chest pain. Pain can be located over the precordium or referred to the neck, shoulder, or left arm. Pain is often relieved if the child bends forward. Physical examination reveals distant heart sounds (if a large effusion is present) and a pericardial friction rub. Pulsus paradoxus is seen with cardiac tamponade. Echocardiogram is used to assess pericardial disease. Myocardial infarction is rare in childhood.57 Mycocardial ischemia may occur as a result of anatomic or acquired heart lesions or arrhythmias. Anatomic lesions are often associated with a heart murmur. Myocardial ischemia can be seen with acquired heart disease caused by ARF, endocarditis, or viral myocarditis, trauma, or cocaine abuse. Children with Kawasaki disease and coronary artery aneurysms are at risk for coronary artery thrombosis with subsequent myocardial infarction. Symptoms of myocardial infarction in children may be nonspecific, with young children unable to describe the character of pain. Presenting symptoms include shock, vomiting, irritability, dyspnea, or abdominal pain. Older children are more likely to complain specifically of squeezing chest pain that may or may not radiate and shortness of breath. Diagnosis is made by electrocardiogram changes which include ventricular arrhythmias, ST segment changes, and wide Q waves. Elevation of cardiac enzymes is noted in most cases, but is not as reliable an indicator of myocardial ischemia in children as in adults.

Other Causes Gastrointestinal causes of chest pain include gastrointestinal reflux. A foreign body in the esophagus may cause discomfort and pain. Marfan disease is an inherited disorder of connective tissue. A child with Marfan disease is at risk for dissection of the aorta.58 Associated clinical findings of Marfan disease include a history of subluxation of the lens of the eye, increased height and arm span,

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Musculoskeletal Symptom Complexes

CHAPTER

26

197

TABLE 26-3. Diseases Associated with Chest Pain Clinical Features

Clinical Findings

Laboratory Studies

Ancillary Studies

Osteomyelitis

Infection of the ribs or sternum uncommon; can result from contiguous infection from lungs (fungal or actinomycoses), or previous surgery or trauma

Fever, tenderness, erythema, and warmth over affected bone, sinus tract or poorly healing surgical site

Elevated WBC, ESR, CRP. Blood or tissue cultures can be positive

Plain radiographs show bone destruction. Bone scan can be useful if multifocal disease suspected

Pleurodynia

Acute onset of sharp chest or upper abdominal pain. Preceding respiratory or enteric illness

Fever, pain worse on inspiration, mild chest wall tenderness

Enterovirus can be isolated from respiratory secretions or stool

Costochondritis

Acute or subacute onset of anterior chest wall pain

Tenderness over the costochondral or costosternal junctions (usually 2nd to 5th costal cartilages); unilateral

Elevated ESR in some patients

Radiograph normal; occasionally soft-tissue swelling at costochondral junction or cartilage calcification

Pneumonia

Acute onset. Preceding URI, cough

Fever, hypoxia, tachypnea, decreased breath sounds, dullness to percussion, rales, rhonchi

Elevated WBC, ESR, CRP. Blood and sputum cultures can be positive

Radiograph abnormal

Pulmonary embolus

Risk factors include the presence of a central venous catheter, immobility (and other, discussed in text). Acute onset of dyspnea, pleuritic chest pain, cough

Tachypnea, tachycardia, rales, loud S2 and S4

Positive D-dimers

Radiograph nonspecifically abnormal

Pericarditis

Precordial pain, can be referred to neck or arm

Fever, distant heart sounds, pericaridial friction rub. Relief of pain if patient bends forward

Myocardial infarction

Rare in children, usually underlying anatomic heart disease or acquired heart disease (Kawasaki disease)

Hypotension, shortness of breath, cyanosis, anxiety

Aortic dissection

Patient may have clinical features of Marfan disease. Acute onset of pain which can radiate to the back

Tachycardia, soft murmur over the aorta, hypotension

ST segment elevation, PR segment depression, T wave changes, low voltage with effusion. Radiograph can show increased heart size Elevated cardiac enzymes (reliability in children not well evaluated)

ECG changes: arrhythmias, ST segment changes, wide Q waves Echocardiogram confirms diagnosis

CRP, C-reactive protein; ECG, electrocardiogram; ESR, erythrocyte sedimentation rate; URI, upper respiratory infection; WBC, white blood cell count.

narrow fingers and toes, hypermobility of the joints, anterior chest wall deformities, and scoliosis. The most common cardiovascular complication is progressive enlargement of the aortic root predisposing to aortic aneurysm, dissection, or rupture. Symptoms of a dissecting aortic aneurysm include sudden onset of substernal chest pain that radiates to the back or shoulders. Findings on physical examination may include a soft murmur over the aorta, tachycardia, hypotension, and poor perfusion. Echocardiography establishes the

presence of dilatation of the aortic root and ascending aorta and aortic or mitral valve insufficiency. Spontaneous pneumothorax can also be a feature of Marfan disease and an additional cause of chest pain. Esophageal caustic ulceration can occur when oral medications (especially tetracyclines and nonsteroidal anti-inflammatory drugs) are not taken with sufficient fluid or when taken just before recumbency, or both.

SECTION

C

Oral Infections and Upper and Middle Respiratory Tract Infections CHAPTER

27

Infections of the Oral Cavity

cellulitis or fascial plane infection occurs. Excision of the gum flap or extraction of the involved tooth can be considered.

Pulpitis and Periapical Abscess

The oral cavity, excluding the pharynx, can be divided into the following five anatomic and functional areas: (1) teeth and supporting structures (periodontium); (2) oral mucosa; (3) salivary glands; (4) bones; and (5) deep fascial spaces. The periodontium supporting the tooth consists of the gingiva, alveolar bone, periodontal ligament, and root cementum. Most infections involving the tooth are polymicrobial, with anaerobic bacteria playing a leading role. More serious infections, which spread to the orofacial tissues, tend to include anaerobic and aerobic organisms. Aerobic organisms are more often associated with soft-tissue and fascial space infections.

Destruction of enamel and dentin by caries results in bacterial invasion of the pulp to produce pulpitis. Suppurative odontogenic infections, including periapical abscesses or deep fascial space infections, are typically polymicrobic; Fusobacterium nucleatum, Bacteroides and Prevotella spp., Peptostreptococcus, Actinomyces spp., and Streptococcus spp. are the predominant isolates.2 If drainage from the pulp is obstructed, the infection progresses rapidly, causing pulpal necrosis and proliferation of endodontic microorganisms that either invade the periapical areas to create a periapical abscess or spread to alveolar bone to create an acute alveolar abscess. Periapical abscess can extend to mucous membranes, skin, and the deep fascial spaces. Treatment consists of prompt drainage of the infected pulp, endodontic (root canal) therapy, or extraction of the affected tooth, with drainage of the abscess. Antibiotic therapy is indicated if drainage cannot be adequately established or when infection has perforated the cortex and spread into surrounding soft tissue.

ODONTOGENIC INFECTIONS

Periodontal Infections

Tooth Infections

Periodontal infections involve inflammation of the gingiva, or periodontium, or both. In gingivitis, dental plaque in the gingival crevice generates inflammation of the gingival tissue without affecting the underlying alveolar bone or periodontal ligament. In periodontitis, the attachment between the involved tooth and gingiva is lost. Subgingival plaque can form on the root surfaces, resulting in bone and tooth loss.3,4

Robert J. Leggiadro

Dental Caries The tooth surface is primarily colonized by Streptococcus sanguis, S. mutans, S. mitis, and Actinomyces viscosus. S. mutans is consistently the only organism isolated from decayed dental fissures and is found in greater numbers in carious teeth than in noncarious teeth.1 Sticky dextrans produced by sucrose metabolism and activity of the enzyme glucosyltransferase of S. mutans bind bacteria, cells, and debris (plaque) to the tooth. Acidic byproducts of carbohydrate metabolism of S. mutans are held next to the tooth surface by the plaque and cause demineralization of the dentin and enamel. When enough structure is destroyed, the bacteria penetrate the pulp and have access to periapical tissue. In children, 85% of carious lesions occur on the fissured surfaces of the teeth, and only 15% on the lingual or buccal surfaces. Topical or systemic fluoride is effective in reducing caries. Fluoride: (1) forms a complex with the apatite crystals in enamel by replacing the hydroxyl group, lending strength and acid resistance to the entire structure; and (2) promotes remineralization of carious lesions while exerting a modest bacteriostatic effect.

Pericoronitis Pericoronitis is an acute localized infection of the gum flap (operculum) that overlies a partially erupted or impacted third molar. Entrapment of food and microorganisms under the affected gingival tissue results in swelling and sudden onset of severe pain if drainage is interrupted. Because the masticator spaces are often involved, marked trismus secondary to irritation of the masseter or medial pterygoid muscle is a prominent feature. Fever, tender ipsilateral lymph node involvement, and dysphagia may be present. Gentle debridement and irrigation under the tissue flap as well as systemically administered penicillin are therapeutic. Incision and drainage may be necessary if 198

Gingivitis Healthy gingiva is a pink, keratinized mucosa that is firmly attached to the teeth and alveolar bone and extends between the teeth to form interdental papilla. A thin cuff of unattached gingiva surrounds each tooth, and the resulting crevice between the free gingiva and the tooth is normally about 1 to 2 mm in depth. Accumulation of bacteria and plaque in the gingival crevice results in gingivitis, a localized inflammation of the free gingiva that manifests as an erythematous nonpainful swelling of the buccal and lingual gingiva and interdental papillae. Mild fetor oris and gums that bleed easily can be noted. Most children have gingivitis at some time, and in volunteer studies, gingivitis develops in most subjects after 14 to 21 days of not brushing the teeth. Gingivitis can slowly proceed to periodontitis.

Acute Necrotizing Ulcerative Gingivitis (Vincent Disease, Trench Mouth) Acute necrotizing ulcerative gingivitis (ANUG) is a severe form of gingivitis that most commonly affects adolescents and young adults. It is characterized by the sudden onset of pain, severe halitosis, and gingival bleeding that interferes with normal mastication. Systemic symptoms such as fever, malaise, and lymphadenopathy can be present. Unlike simple gingivitis, necrosis of the intradental papillae resulting in a marginated, punched-out and eroded appearance is observed. Treatment with local debridement and efficient plaque con-

Infections of the Oral Cavity

trol is usually effective, but antibiotic therapy systemically with penicillin or metronidazole may be indicated if fever or lymphadenopathy is present. In children with cancer, especially those with poor nutrition, neutropenia, and poor oral hygiene, ANUG can result in loss of teeth.5,6

Periodontitis Periodontitis is a progressively severe infection characterized by inflammation of the gingiva, increased tooth mobility due to destruction of the periodontal ligament, and subsequent resorption of alveolar bone; in some cases, purulent exudate is present. Periodontal bacteria form a biofilm (subgingival plaque) and can elicit an inflammatory response that ultimately destroys the connective tissue (periodontium).7 Focal or diffuse periodontal abscesses can form, manifesting as red, fluctuant swelling of the gingiva or oral mucosa that is extremely tender to palpation. Later, the underlying supporting tissues are affected, ultimately leading to complete destruction of the periodontium and bone, causing permanent loss of teeth. The destructive process can proceed insidiously, and patients experience little or no discomfort. Treatment is surgical, aimed at drainage of pus and debridement of the root surface.7 Periodontitis, Gingivitis, and Mucositis in Immunodeficient Individuals. A large fraction of neutrophil turnover in normal individuals is due to clearance of these motile cells through the oral cavity. The predominantly neutrophilic infiltrates in normal gingival tissues are important in maintaining periodontal integrity. Thus, periodontal inflammation or degeneration is an important early clinical feature of decreased or defective neutrophil movement (particularly neutrophils with abnormal motility); neutropenia, in particular, is an important factor in mucositis in children with cancer. Mucositis associated with cancer chemotherapy may explain in part the significantly higher risk for infection of the cancer patient compared with that of the patient with congenital or acquired chronic neutropenia.8 The oral microflora of human immunodeficiency virus (HIV)infected adults with periodontal disease is qualitatively similar to that observed in nonimmunocompromised individuals with periodontal disease.9 Rapid progression of periodontal disease in HIV-infected individuals is believed to result from direct tissue damage by oral microflora. Prepubertal Periodontitis. Prepubertal periodontitis is a rare condition observed immediately after the eruption of deciduous teeth; it manifests in both localized and generalized forms. The localized form involves only the primary dentition, with rapid bone loss and minimal gingival inflammation. The generalized form is characterized by acute gingival inflammation, often with proliferation and cleft formation of the gingival tissue and rapid destruction of alveolar bone; it is associated with other skin and upper respiratory tract infections. Most cases of the generalized form are associated with defects in neutrophil adherence that cause inadequate margination. The localized form is amenable to local debridement and systemic antibiotic therapy, but the generalized form seems unresponsive to therapy. Juvenile Periodontitis. Juvenile periodontitis is a rare, highly destructive form of periodontitis in adolescents characterized by rapid bone loss. In the localized form, the first permanent molar and central incisor teeth are affected. In the generalized form, other teeth are affected, and the course is rapidly progressive. The gingiva may bleed easily but otherwise initially appear healthy. The amount of plaque does not correspond to the amount of bone loss. Specific defects of neutrophil chemotaxis and phagocytosis have been demonstrated in 75% of patients. A familial tendency suggests a possible genetic component of the disease. Excellent therapeutic results have been achieved with systemic tetracycline or metronidazole therapy combined with local periodontal treatment involving root debridement and surgical resection of inflamed periodontal tissues.10

CHAPTER

27

199

Complications of Odontogenic Infections Complications occur by direct extension or hematogenous spread. Direct extension of odontogenic infections can result in mediastinitis, cavernous sinus thrombosis, suppurative jugular thrombophlebitis (Lemierre syndrome), maxillary sinusitis, carotid artery erosion, and osteomyelitis.11 Extension to the potential fascial spaces of the lower part of the head and upper portion of the neck is reviewed here.12

Ludwig Angina The term Ludwig angina has been applied to a heterogeneous array of infections involving the sublingual, submaxillary, and submandibular spaces. It is important to restrict the diagnosis to cases conforming to the classic definition of bilateral sublingual and submandibular space involvement. This definition includes infection beginning in the floor of the mouth with rapidly spreading indurated cellulitis without abscess formation or lymphatic involvement. In most cases, a dental source of infection is present. Lacerations of the floor of the mouth or mandibular fractures are also inciting causes. Patients at special risk include those with immunodeficiency or malignancy, and, particularly, those who have undergone bone marrow or organ transplantation.13,14 The sublingual and submandibular spaces are separated by the mylohyoid muscle – the sublingual space lying above and medial to the muscle, the submandibular space inferior and lateral to the mylohyoid. The submandibular space is further subdivided into the submaxillary and submental spaces. A dental source of infection is found in 50% to 90% of cases, most commonly involving the second and third mandibular molars because the root apices are inferior to the mylohyoid muscle. Infections of the sublingual space generally arise from the mandibular incisors because the root apices lie above the mylohyoid muscle. The clinical finding is brawny, boardlike swelling in the submandibular spaces that does not pit with pressure. The mouth is held open, and the floor of the mouth is elevated, pushing the tongue to the roof of the mouth. Eating, swallowing, and talking are difficult, and respiration can be impaired by obstruction from the tongue. Rapid progression of the infection results in edema or a bull-neck appearance; edema of the glottis can result in asphyxia. Fever and extreme systemic toxicity are usually present, with drooling and neck stiffness. In deep fascial space infections secondary to polymicrobic odontogenic infections, F. nucleatum, Bacteroides and Prevotella spp., Peptostreptococcus spp., Actinomyces spp., and Streptococcus spp. are predominant isolates. Except in selected patients with serious underlying illnesses, facultative gram-negative bacilli and Staphylococcus aureus are isolated uncommonly. Treatment requires high doses of antibiotics, monitoring of the airway, early intubation or tracheostomy when required, soft-tissue decompression, and surgical drainage. Complications of Ludwig angina include progressive and direct extension of the infection to the neck and mediastinum.15

Suppurative Jugular Thrombophlebitis (Lemierre Disease) In 1936, Andre Lemierre, a Parisian bacteriologist, described 20 cases of anaerobic thrombophlebitis of the internal jugular vein with metastatic infection.16 Most cases of Lemierre syndrome occur in adolescents and young adults with tonsillopharyngitis, or, less commonly, those with odontogenic infection, mastoiditis, or sinusitis. Fusobacterium necrophorum, an anaerobic gram-negative rod, is identified in 82% of cases.17 The infection causes pain and swelling of the anterior border of the sternocleidomastoid muscle, neck stiffness, torticollis, spiking temperature with diaphoresis, headache, and increased intracranial pressure. Cephalad extension can result in cavernous sinus thrombosis. Cervical computed tomography (CT) demonstrates an enlarged,

200

SECTION

C Upper Respiratory Tract and Oral Infections

thrombosed vein that does not opacify with adjacent swelling of the pharyngeal soft tissues. From the neck, the infection spreads hematogenously, most commonly to the lungs, but also to the joints, liver, spleen, bones, kidneys, and meninges. Signs of metastatic infection include pulmonary infiltrates, hepatosplenomegaly, hyperbilirubinemia, elevated liver enzyme levels, hematuria, and disseminated intravascular coagulation.16 The diagnosis may first be suggested when F. necrophorum is isolated from blood culture. Contrast-enhanced CT imaging of the neck can confirm the diagnosis (see Chapter 30, Infections Related to the Upper and Middle Airways, Figure 30-X). Antimicrobial therapy effective against anaerobes is indicated. Ligation of the internal jugular vein is rarely required. Mortality rate approaches 18% despite appropriate therapy.17

MUCOSAL INFECTIONS The tongue and buccal mucosa are predominantly colonized by Streptococcus salivarius and Veillonella spp. The clinical characteristics of infections of the oral mucosa are summarized in Table 27-1. Brief additional details for selected infections are provided here.18

secretions. By the age of 3 to 4 weeks, 80% of infants have oral or intestinal colonization with the Candida spp. Locally invasive disease of the buccal and gingival mucosa, and sometimes tongue and soft palate, results in adherent patches of pearly white, curd-like material. Plaques are tenacious, resulting in mild erosions when scraped. A superficial pseudomembrane forms that contains both blastopheres and pseudohyphae. Normal immune function prevents oral candidiasis in most individuals; decrease in immune function or alteration in microflora results in disease. Variants of oral candidiasis, including glossitis with erosion of the lingual mucosa and loss of normal papillation with areas of leukoplakia, can occur in individuals with severe underlying immunodeficiency. Mucosal candidiasis after the age of 6 months (and in the absence of receipt of antimicrobial agents or underlying disease) should prompt an investigation for immunologic disorder. Candidiasis is the most common oral manifestation of pediatric HIV infection, with reported prevalences ranging between 20% and 72%.9 Treatment of oral candidiasis is nystatin. For infants, the dose is 1 mL (100,000 units) in each cheek q6 hours after feeding for 7 to 10 days. For older children, the dose is 1 to 2 million units/day in divided doses every 6 hours until resolution. Oral candidiasis in immunocompetent hosts can also be treated with clotrimazole troches. Fluconazole or itraconazole may be beneficial for immunocompromised patients with oropharyngeal candidiasis.21

Herpes Simplex Virus Stomatitis Herpes simplex virus (HSV) infections are the most frequently encountered viral disorders of the oral cavity and perioral tissues. Although HSV type 1 (HSV-1) is responsible for most infections, HSV2 can also cause oral lesions. Primary infection with HSV-1 occurs throughout childhood, most commonly between the ages of 6 months and 5 years old, and during adolescence, resulting in the presence of antibodies in up to 90% of all adults.19 Most HSV-1 infections are localized to the mouth and oropharynx. Of all perioral HSV-1 infections, only 10% to 30% are symptomatic. The incubation period ranges from 2 days to 2 weeks. Clinical findings in children include high fever and extensive oral mucosal eruption (predominantly the buccal and lingual mucosa and tongue) with tender submandibular lymphadenopathy and irritability. Intraoral lesions can begin as vesicular or ulcerative lesions on the lips, gingivae, tongue, buccal mucosa, and palate. Autoinoculation during primary infection can result in lesions on the face, conjunctivae, and hands. As the vesicles evolve to shallow ulcers on an erythematous base, intraoral edema and pain can prevent drinking and result in dehydration. Symptomatic infection lasts 2 to 3 weeks, with gradual re-epithelialization. One-third of individuals with clinical orolabial HSV-1 infection have a recurrence. Recurrent disease is usually mild and is preceded in many patients by 1 to 2 days of prodromal pain, burning, and tingling. Pain is maximal at the time of onset of the vesicles. The most common site for recurrences is the outer edge of the vermilion border of the lips, but lesions inside the mouth or on the face are not uncommon. Recurrence consists of 3 to 5 lesions located in a cluster < 10 mm3 in diameter. Over 1 to 2 days, the vesicles progress to pustules and ulcers, resolving in 8 to 10 days. Systemic acyclovir may be effective in severe primary gingivostomatitis. Minimal therapeutic benefit of oral acyclovir therapy has been demonstrated among adults with recurrent herpes labialis. Topical acyclovir is ineffective.20 Oral prophylaxis with acyclovir may be useful for frequent recurrences, especially those that involve the face.

Rare Causes of Stomatitis Neutropenia should be excluded when ulcerative or necrotic lesions are seen on the gingiva. Histiocytoses are rare causes of necrotizing gingivitis, and infectious mononucleosis is associated with stomatitis on rare occasions. Trauma to the buccal mucosa or gingiva can produce lesions resembling the ulcerations of stomatitis.

Mucositis in Immunocompromised Patients Mucositis and stomatitis in the immunocompromised patient appear to be due to a breakdown of the mucosal epithelium, which can become secondarily infected with bacteria, fungi, or reactivated latent viruses. Neutropenia is associated with inability to limit superficial invasion of microflora. Four to 7 days after irradiation or chemotherapy, nonkeratinized oral epithelium (including the buccal and labial mucosa, soft palate, oropharynx, floor of the mouth, and ventral and lateral surfaces of the tongue) can develop ulceration and pseudomembrane formation. Pain or tenderness is frequently the only abnormal finding, because chemotherapy or irradiation suppresses inflammatory reaction. Morbid viridans streptococcal septicemia can be associated. Specific microbiologic diagnosis of mucositis or stomatitis in an immunocompromised patient is important, especially to identify fungi, HSV, and gram-negative organisms. Topical and systemic antimicrobial agents may be indicated, along with antiseptic and anesthetic applications. Frequent saline rinses can reduce mucosal irritation, remove thickened secretions, and increase moisture in the mouth. Coating agents such as milk of magnesia and aluminum hydroxide gel have been useful for symptomatic relief, as have topical or oral cytoprotective agents (i.e., sucralfate). One study in adults showed benefit of palifermin in reducing the occurrence and severity of radiochemically induced mucositis,22 but further study is required.23 Meticulous oral and dental hygiene, effective management of xerostomia, and early control of latent viral reactivation are necessary to limit morbidity.

Pseudomembranous Candidiasis Oral candidiasis, the most common oral fungal disease, is usually caused by Candida albicans, which has the ability to adhere to mucosal surfaces. This organism is a common inhabitant of the oral cavity. On day 1 of life, 4% to 18% of full-term infants have C. albicans in the oropharynx as a result of exposure to maternal

SALIVARY GLAND INFECTIONS Painful enlargement of a salivary gland most commonly involves the parotid gland, less often the submandibular and sublingual glands. Parotitis results from bacterial or viral infection and can be acute or chronic.24–27 Infection is characterized by fever and glandular enlarge-

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Infections of the Oral Cavity

CHAPTER

27

201

TABLE 27-1. Oral Mucosal Manifestations of Local and Systemic Infection Cause

Lesion

Site

Disease Course/Comment

Herpes simplex virus (HSV)

Irregular, shallow, painful ulcers

Gingiva, labial and buccal mucosa, tongue, floor of mouth; uncommonly, lips and pharynx

Varicella-zoster

Shallow, 2–3 mm, nonpainful ulcers

Palate

Human herpesvirus 6 (HHV-6) Rubeola

Erythematous papules

Soft palate (base of uvula)

White papules, 1–2 mm Erythematous maculopapules

Buccal mucosa opposite lower molars Palate

Herpangina (coxsackievirus A and B; other enteroviruses)

Punctate vesicles and ulcers 2–3 mm on 10-mm erythematous base; not extremely painful

Posterior soft palate, anterior tonsillar pillars, posterior pharyngeal wall; rarely, dorsum or tip of tongue

Hand, foot, mouth disease (coxsackievirus A-16)

Ulcers 4–8 mm Papulovesicules; nontender, discrete

Hard palate, tongue, buccal mucosa Palms, soles, buttocks, arms, and legs

Intense inflammation of gingiva and oral mucosa precedes ulcers; recurrent lesions located on mucocutaneous border of lips or on keratinized tissue over bone and hard palate Mucosal lesions follow course of skin lesions; vesicular stage often missed owing to rapid rupture of vesicles; recurrent lesions accompanied by zoster of adjacent skin Nagayama spots present in majority of infected children by day 4 Koplik spots coalesce into larger number and size of lesions within 12 hours; background mucosal surface always bright red and granular; enanthema sloughs by 2nd day of rash Lesions usually discrete; average number 5 (1–14); enlarge from 1–4 mm over 2–3day period; ulcers remain discrete and do not coalesce Oral lesions present in all patients

VIRUS

BACTERIA

Scarlet fever Strawberry tongue (Streptococcus pyogenes)

Tongue

Reddened edematous papillae project through white coating; coating peels off to leave red glistening tongue with prominent papillae Intraoral erythema; uvula and free margin of soft palate red and edematous; lips uninvolved Initial labiitis and generalized erythema rapidly progress to painful desquamation Primary chancre heals over 5–10 days; followed by systemic manifestations if untreated Subgingival plaque always present; common in children, particularly during puberty Necrosis of interdental papillae results in marginated, punched-out, eroded appearance; gray necrotic pseudomembranes; fusiform bacilli and spirochetes consistently found Infants < 10 months with pneumococcal bacteremia Streptococcus pyogenes, Staphylococcus aureus, Haemophilus influenzae b Streptococcus pyogenes, Haemophilus influenzae b; can cause respiratory compromise

Punctiform and petechial lesions

Palate

Toxic shock syndrome (Staphylococcus aureus) Treponema pallidum

Intraoral erythema, edema, and desquamation Painless ulcer with rolled edges

Lips, all intraoral mucous membranes

Gingivitis

Edema, thickening, erythema, and easy bleeding; usually painless Pseudomembranes (necrosis of papillae)

Gingiva and interdental papillae

Cystic gingival lesions

Red, cystic lesion, 4–10 mm

Gingival ridge

Glossitis

Swollen, red, painful

Tongue

Uvulitis

Red, edematous

Uvula

Pseudomembranes of creamy white plaques on erythematous mucosa Ulcers

Tongue, gingiva, buccal mucosa; rarely, pharynx, larynx, esophagus; cheilitis Oropharynx

Lesions can be painless, can burn, or can feel sore or dry Occurs with subacute or chronic disseminated disease

Strawberry tongue; mucosal erythema; erythema, cracking peeling, bleeding lips Shallow, circular painful ulcer < 5 mm on erythematous base with pseudomembrane

Tongue, oropharynx, lips

Mucositis present in all but atypical forms

Nonkeratinized movable labial and buccal mucosa; tongue and floor of mouth

Deep ulcers of > 5 mm diameter; painful Small cluster of vesicles; extremely painful; no erythematous border

All areas of oral cavity

Healing occurs after 4 days; recurrences 12 times/year; etiology unknown; symptomatic therapy; common (incidence 20%) Lasts 6 weeks–3 months; heals with scarring (Sutton scarring aphthae) Herpetiform ulcers with no relationship to HSV

Acute necrotizing ulcerative gingivitis (ANUG)

Anywhere on oral mucous membranes

Gingiva

FUNGUS

Candida species Histoplasma capsulatum UNKNOWN

Kawasaki disease Recurrent oral ulcerations (aphthous stomatitis)

Tip and lateral margins of tongue

202

SECTION

C Upper Respiratory Tract and Oral Infections

TABLE 27-1. Oral Mucosal Manifestations of Local and Systemic Infection—Continued Cause

Lesion

Site

Disease Course/Comment

PFAPA syndrome

Small ulcers

Oral mucosa

Gangrenous stomatitis (noma)

Focal, destructive lesions

Gingiva and deeper structures

Syndrome of recurrent fever, aphthous stomatitis, pharyngitis, and cervical adenitis; affects children < 5 years; recurrence every 2–9 weeks Rare, acute, fulminating in severely debilitated and malnourished children; spirochetes and other anaerobic bacteria may be involved

Cyclic neutropenia and agranulocytosis Drug- or radiationinduced stomatitis

Small ulcers

Oral mucosa

Ulcers and pseudomembrane formation; painful

Behçet syndrome

Ulcers

Nonkeratinized labial and buccal mucosae soft palate; floor of mouth; ventral and lateral tongue surfaces Oral mucosa

Stevens–Johnson syndrome

Macules, papules, vesicles, bullae

Buccal mucosa and vermilion border of the lips; relative sparing of gingivae

Inflammatory bowel disease

Ulcers (aphthous-like)

Oral mucosa

Microbiologic diagnosis important to rule out viral reactivation or bacterial superinfection Multisystem disease that can involve many mucous membranes Erythema multiforme exudativum; 25% of cases confined to oral mucosa; lesions appear in crops; complete healing 4–6 weeks; can cause recurrent stomatitis and labiitis Manifestation of multisystem disease

PFAPA, periodic fever with aphthous stomatitis, pharyngitis, and adenitis.

ment. Swelling is equally distributed above and below the angle of the jaw, except that edema may lower the center of the swelling. Enlargement occurs rapidly if major ducts are obstructed. Parotitis is usually painful because of stretching of the capsule but is not usually exquisitely tender unless there is suppuration. Pain is made worse by foods that stimulate production of saliva. The opening of Stensen duct, which is easily seen in the buccal mucosa, is often red and swollen and may exude pus if the condition is suppurative.

Suppurative Parotitis Suppurative parotitis is an uncommon infection in infants and children: it occurs as an acute single episode or with multiple recurrences. It is often associated with dehydration, immunosuppression, or ductal obstruction. Most reported cases have occurred in infants younger than 2 months, and especially in prematurely born infants.28 Infection results from hematogenous spread (particularly in the neonatal period) or from organisms in the oral cavity ascending Stensen duct. More than half of patients are afebrile, but high fever, chills, and marked toxicity can occur. The organism recovered most often in suppurative parotitis is Staphylococcus aureus.28 Gramnegative bacilli are also an important cause in the neonate. Streptococci (Streptococcus pyogenes, S. pneumoniae, viridans streptococci) and anaerobic bacteria have also been isolated. Management involves accurate bacteriologic diagnosis based on culture and Gram stain of purulent drainage from the Stensen duct. If necessary, cannulation and lavage of the duct can be performed to obtain material for culture. Blood culture should also be obtained. In the neonate, parotitis is sometimes confused with the facial, buccal, or submandibular cellulitis typically caused by group B streptococcus. Management consists of maintaining adequate hydration and administering appropriate antimicrobial agents. Drainage and rehydration alone have been curative in some cases. Surgical drainage is limited to refractory cases. The dense fascia and septation of the parotid cause multiple small loculations of pus and prevent fluctuance in all but the most advanced stages of suppuration. Thus, complete surgical exposure of the gland is generally recommended, particularly if extension of infection into the parotid space is suspected. Evaluation by ultrasonography or CT may be helpful.

Viral Parotitis Viral parotitis in children has been uncommon since the introduction of universal immunization with mumps virus vaccine, with remarkable exceptions, such as in the midwest United States in 2006, when > 2600 cases of mumps occurred, predominantly in college students and young adults.29 Other viruses associated with parotitis are HIV, enteroviruses, Epstein–Barr virus, parainfluenza virus, influenza virus, cytomegalovirus, and lymphocytic choriomeningitis virus. Both Bartonella henselae and HIV infection can result in a chronic, lowgrade, relatively nontender infection of the parotid gland. Sjögren syndrome, which can be confused with chronic parotitis, is associated with xerostomia, conjunctivitis, and systemic autoimmune disease; the presence of temporomandibular arthritis or arthralgia should suggest this syndrome. Treatment is symptomatic. Prolonged swelling and tenderness should lead to evaluation for other causes of enlargement, including tumor or cyst, adverse drug effect, metabolic or immunologic disorders, and ductal obstruction.

Recurrent Acute Parotitis/Sialadenitis Sialadenitis can recur (most often affecting the parotid gland and called juvenile recurrent parotitis) once every few years to several times a year in occasional cases.27,30 S. pneumoniae, nontypable Haemophilus influenzae and oral streptococci are sometimes isolated from saliva and expressed pus.31 No other organism has been consistently isolated, and the pathogenesis is unknown. Predisposition due to local autoimmune factors, allergy, infection, and genetic inheritance have been suggested; immunoglobulin A deficiency was detected in 22% of patients in one series.32 Patients with HIV infection can manifest recurrent or chronic sialadenitis; cytomegalovirus and S. pneumoniae have been associated.33 The condition usually resolves at adolescence. Oral administration of penicillin, good hydration, parotid massage, and use of sialagogues such as chewing gum usually result in prompt remission. Endoscopic irrigation and ductal dilation have been reported to be beneficial.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

The Common Cold

OSTEOMYELITIS Although osteomyelitis of the jaw usually results from odontogenic (especially periapical) infection, open fracture of the jaw with delayed treatment also represents a significant inciting event. Predisposing conditions that affect host resistance and may play a role in the development of osteomyelitis of the jaw include diabetes mellitus, agranulocytosis, leukemia, sickle-cell disease, fibrous dysplasia, and Paget disease.34,35 Most instances of osteomyelitis of the jaw are caused by microaerophilic viridans streptococci, anaerobic streptococci, and other anaerobes, including Peptostreptococcus, Fusobacterium, and Bacteroides species. Only occasional cases are caused by Staphylococcus aureus, which is generally associated with breaks in the skin. Gram-negative organisms can also be involved, including Salmonella in patients with sickle-cell anemia.35 Specific forms of disease are caused by Actinomyces israelii, Treponema pallidum, and nontuberculous mycobacteria.

Mandibular Osteomyelitis The mandible is involved more often by osteomyelitis than the maxilla because of the relatively poorer blood supply, which is limited to one major vessel and the periosteal circulation. Clinical features include fever, facial swelling, jaw pain, and generally a carious or discolored tooth. Mental nerve paresthesia may be present early in the course. Management of acute suppurative osteomyelitis of the mandible is similar, in most respects, to that of acute osteomyelitis of any bone. If the infection began as a periapical dental abscess, the involved tooth should be removed as soon as possible to allow for drainage and to provide material for culture. Treatment is given for a minimum of 4 weeks. Actinomycosis is particularly challenging to manage. Amoxicillin (chosen for pharmacodynamic and antimicrobial activity) plus probenecid is usually given for 1 year or more. Relapses can occur year(s) later. Surgical debridement may be necessary to affect cure if sequestrum or involucrum is present.36

Infantile Maxillary Osteomyelitis Osteomyelitis of the jaw in the neonate is rare but can have serious sequelae. It most commonly involves the maxilla and is thought to arise from neonatal trauma to oral tissues, from hematogenous spread from skin, middle ear, or mastoid, or from an infected maternal nipple. The infant manifests facial cellulitis centered on the orbit. Fever, irritability, anorexia, and dehydration in association with palpebral edema and conjunctivitis can be seen in association with purulent discharge from the nose or inner canthus. Examination of the mouth reveals swelling of the maxilla extending to both the buccal and palatal regions, with fluctuance and fistula formation often present. Staphylococcus aureus is the most common cause. Aggressive treatment is necessary to prevent permanent optic damage, neurologic complication, loss of tooth buds and bone, and extension to the dural sinuses.

Garré Osteomyelitis Garré sclerosing osteomyelitis is a chronic, nonsuppurative, sclerosing osteomyelitis.37 It is characterized by a localized, hard, nontender swelling of the mandible. It is commonly associated with a carious tooth, usually a lower first molar. Frequently, there is history of a past toothache, recent dental extraction, or infection of a flap of tissue over an erupting tooth. Radiographs show a focal area of calcified proliferation of bone that is smooth and often has a laminated or onionpeel appearance (a periosteal response to low-grade stimulus such as a dental infection). It resembles infantile cortical hyperostosis (Caffey disease), osteosarcoma, and Ewing sarcoma. Biopsy is performed to exclude neoplasm if regression does not occur after extraction of the involved tooth. Antibiotic therapy is not necessary.

CHAPTER

CHAPTER

28

203

28

The Common Cold Diane E. Pappas and J. Owen Hendley The common cold, also known as upper respiratory tract infection (URI), is an acute, self-limited viral infection of the upper airway that may involve the lower respiratory tract as well. The characteristic symptom complex consisting of rhinorrhea, nasal congestion, and sore or scratchy throat is familiar to all adults. Colds are the most common cause of human illness and are responsible for significant absenteeism from school and work. Children are especially susceptible because: (1) they have not yet acquired immunity to many of the viruses; (2) they have poor personal hygiene practices; and (3) they have frequent close contact with other children who are excreting virus.

ETIOLOGY Colds are common because some of the causative viruses do not produce lasting immunity after infection and some viruses have numerous serotypes (Table 28-1). Cold viruses that do not produce lasting immunity include respiratory syncytial virus (RSV), parainfluenza viruses, and coronaviruses. Cold viruses that have numerous serotypes but produce lasting serotype-specific immunity after infection include rhinoviruses, adenoviruses, influenza viruses, and enteroviruses.1 Rhinoviruses (rhino, nose), with at least 100 serotypes, are the most common cause of URIs in children and adults. At least 50% of colds in adults are caused by rhinovirus. Other viruses that cause URIs are coronaviruses (corona, crown), RSV, human metapneumovirus, influenza virus, parainfluenza virus, adenovirus, echoviruses, and coxsackieviruses A and B. Some of these viruses cause characteristic syndromes; for example, RSV causes bronchiolitis in children 2 years or younger, influenza viruses cause febrile respiratory illness with severe lower respiratory tract involvement, adenoviruses cause pharyngoconjunctival fever, parainfluenza viruses cause croup in young children, and enteroviruses cause a variety of illnesses, including aseptic meningitis and herpangina.

TABLE 28-1. Immunity to Common Cold Viruses Virus

No of Serotypes

LONG-LASTING IMMUNITY NOT PRODUCED BY INFECTION (REPEATED INFECTION WITH SAME SEROTYPE USUAL)

Respiratory syncytial virus (RSV) Parainfluenza virus Coronavirus

1 4 2

IMMUNITY PRODUCED BY INFECTION (REINFECTION WITH SEROTYPE UNCOMMON)

SAME

Rhinovirus Adenovirus Influenza Echovirus Coxsackievirus group A Coxsackievirus group B

> 100 ≥ 33 3a 31 3 6

a Type A subtypes change. Modified from Hendley JO. Immunology of viral colds. In: Veldman JE, McCabe BF, Huizing EH, et al. (eds) Immunobiology, Autoimmunity, Transplantation in Otorhinolaryngology. Amsterdam, Kugler Publications, 1985, pp 257–260.

204

SECTION

C Upper Respiratory Tract and Oral Infections

EPIDEMIOLOGY

Incidence of colds

In temperate climates in the northern hemisphere, the predictable yearly epidemic of colds begins in September and continues unabated until spring. This sustained epidemic curve is a result of successive waves of different respiratory viruses moving through the community (Figure 28-1). The epidemic begins with a sharp rise in the frequency of rhinovirus infections in September (after children return to school), which is followed by parainfluenza viruses in October and November. RSV and coronaviruses circulate during the winter months, whereas infection due to influenza virus peaks in the late winter. The epidemic finally ends with a small resurgence of rhinovirus infections in the spring. Adenovirus infection occurs at a constant rate throughout the cold season.2 The frequency of colds varies with age. A 10-year study of families with children who did not attend a childcare facility showed that the peak incidence of colds occurs in preschool children 1 to 5 years old, with a frequency of 7.4 to 8.3 colds per year. Infants younger than 1 year averaged 6.7 colds per year, and teenagers averaged about 4.5 colds per year. Mothers and fathers experienced about 4 colds per year.3 With the greater exposure of children to other preschool children in childcare facilities, the frequency of colds in children younger than 6 years has increased. Thus, the typical preschool child experiences at least one URI per month throughout the cold season. Viral transmission occurs primarily in the home setting, although the exact mechanism of spread has not been clearly established. Colds can be spread by: (1) small-particle (< 5 mm in diameter) aerosol, which infects when inhaled; or (2) large-particle (> 10 mm in diameter) droplets, which infect by landing on nasal or conjunctival mucosa; or (3) direct transfer via hand-to-hand contact.4 Small-particle aerosol is an effective method of transfer for influenza virus5 and coronavirus6

Rhinovirus

Parainfluenza virus

Coronavirus

RSV

Influenza

Adenovirus

July Aug Sept Oct Nov Dec Jan Feb Mar Apr May June Month Figure 28-1. Schematic diagram of the incidence of colds and frequency of causative viruses. RSV, respiratory syncytial virus. (Redrawn from Hendley JO. The common cold. In: Goldman L, Bennett JC (eds) Cecil Textbook of Medicine, 21st ed. Philadelphia, WB Saunders, 2000, pp 1790–1793.)

but not for RSV.7 Rhinoviruses are most likely spread by large-particle droplets or direct transfer. Rhinoviruses can survive as long as 2 hours on human hands and up to several days on other surfaces. Studies in young adults have shown that infected individuals commonly have rhinovirus on their hands, which can be efficiently transferred to the hands of uninfected individuals during brief contact; infection then results when the uninfected person transfers the virus from the hands on to his or her nasal or conjunctival mucosa. Sneezing and coughing are ineffective modes of rhinovirus transmission.8 Inoculation of the oral mucosa with rhinovirus9 or RSV10 does not result in infection.

PATHOGENESIS Symptoms of the common cold do not appear to result from destruction of nasal mucosa, because nasal biopsy specimens from young adults with both natural and experimentally induced colds show intact nasal epithelium during symptomatic illness.11,12 Study by in situ hybridization of nasal biopsy specimens obtained during rhinovirus infection indicates that replication occurs in only a small number of epithelial cells.13,14 Furthermore, in vitro studies have shown that rhinovirus and coronavirus produce no detectable cytopathic effect when replicating in a cultured monolayer of nasal epithelial cells, whereas influenza virus A and adenovirus produce obvious damage.15 The symptoms of the common cold appear to result from release of cytokines and other mediators from infected nasal epithelial cells as well as from an influx of polymorphonuclear cells (PMNs). Nasal washings of volunteers experimentally infected with rhinovirus showed a 100-fold increase in PMN concentration 1 to 2 days after inoculation.16 This influx of PMNs coincides with onset of symptoms and correlates with the presence of a colored nasal discharge.17 A yellow or white nasal discharge may result from the higher number of PMNs, whereas the enzymatic activity of PMNs (due to myeloperoxidase and other enzymes) may cause a green nasal discharge. A potent chemoattractant for PMNs is produced by cells in culture infected with rhinovirus.18 This chemoattractant has been identified as interleukin 8 (IL-8).19 Elevated levels of IL-8 and other cytokines (IL1b, IL-6) have also been demonstrated in the nasal secretions of infected individuals.20,21 Furthermore, elevated levels of albumin and kinins (predominantly bradykinin) in nasal secretions have been shown to coincide with the onset of symptoms in experimental rhinovirus infection.16 The elevated concentration of albumin and kinins likely results from exudation of plasma proteins due to greater vascular permeability in the nasal submucosa. The method by which viral infection initiates this vascular leak has not yet been determined. The release of kinins resulting from plasma exudation may augment the symptoms of the cold; bradykinin alone can cause rhinitis and sore throat when sprayed into the noses of uninfected individuals.22 The paranasal sinuses are usually involved during an uncomplicated cold caused by respiratory viruses. In one study, computed tomographic (CT) scans obtained during the acute phase of illness revealed abnormalities of one or more sinuses in 27 (87%) of 31 young adults.23 Without antibiotic therapy, there was complete resolution or marked improvement of the sinus abnormalities in 11 (79%) of the 14 subjects in whom second CT scans were obtained 2 weeks later. It is not known whether these sinus abnormalities result from viral infection of the sinus mucosa or from impaired sinus drainage secondary to viral rhinitis. Nose-blowing can generate enough pressure to force fluid from the nasopharynx into the paranasal sinuses, suggesting that nose-blowing may force mucus containing viruses, bacteria, and inflammatory mediators into the paranasal sinuses during a cold.24 The middle ear can also be involved during uncomplicated colds. Studies in school-aged children have shown that two-thirds will develop abnormal middle-ear pressures within 2 weeks after onset of a cold.25 Otitis media was not diagnosed during the study. It is not known whether the abnormal middle-ear pressures result from viral infection of the mucosa of the middle ear and eustachian tube or from viral nasopharyngitis with secondary eustachian tube dysfunction.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

The Common Cold

CLINICAL MANIFESTATIONS Symptoms of the common cold do not vary by specific cause. In older children and adults, rhinorrhea, nasal obstruction, and sore or scratchy throat are typical. The rhinorrhea is initially clear but may become colored as the illness proceeds. Cough or sneeze may be present. Fever (> 38°C) is uncommon in adults. Other symptoms are malaise, sinus fullness, and hoarseness. Objective findings are minimal except for mild erythema of the nasal mucosa or pharynx. Symptoms resolve in 5 to 7 days. Compared with adults, infants and preschool children with colds are more likely to have fever (≥ 38°C) and moderate enlargement of the anterior cervical nodes (Table 28-2).1 Rhinorrhea may not be noticed until the nasal discharge becomes colored. Nasal congestion may disrupt sleep and may lead to fatigue and irritability. The illness often persists in infants and preschool children for 10 to 14 days.26

DIFFERENTIAL DIAGNOSIS The differential diagnosis of a cold includes allergic rhinitis, vasomotor rhinitis, intranasal foreign body, and sinusitis. A diagnosis of allergic rhinitis is suggested by a seasonal pattern of clear rhinorrhea, absence of associated fever, and family history of allergy. Possible associated conditions are asthma and eczema. Physical findings consistent with allergic rhinitis include allergic “shiners” and “nasal salute.” The detection of numerous eosinophils upon microscopic examination of the nasal mucus using Hansel stain confirms the diagnosis of allergic rhinitis. A diagnosis of vasomotor rhinitis is suggested by a chronic course without fever or sore throat. The diagnosis of bacterial sinusitis is suggested by persistent rhinorrhea or cough or both for greater than 10 days.27

CLINICAL APPROACH The diagnosis of a cold is based on history and physical examination; generally, laboratory tests are not useful. The rapid test for detecting RSV, influenza, parainfluenza, and adenovirus antigens in nasal secretions can be used to confirm the diagnosis. RSV, rhinovirus, influenza viruses, parainfluenza viruses, and adenoviruses can also be isolated in cell culture. Coronavirus cannot be detected reliably in cell culture, so serologic titer rise can be used for diagnosis, if necessary. Polymerase chain reaction assays for diagnosis of all the respiratory viruses are available in research laboratories and increasingly in clinical laboratories. Other methods of detection can be used but are rarely needed.

MANAGEMENT At present, no antiviral agents are available that are effective for treatment of the common cold. Although an array of medications may

TABLE 28-2. Characteristics of Viral Colds in Adults and Young

CHAPTER

28

205

be used to relieve symptoms, there is little scientific evidence to support the use of symptomatic treatments in children. Because the common cold is a self-limited illness with symptomatology that is largely subjective, a substantial placebo effect can suggest that various treatments have some efficacy. Inadequate blinding of placebo recipients in a study can make an ineffective treatment appear effective. In adults with colds, first-generation antihistamines (i.e., chlorpheniramine) have been shown to provide modest symptomatic relief, with decreases in nasal discharge, sneezing, nose-blowing, and duration of symptoms.28 This effect is presumably due to the anticholinergic effects of these medications. A randomized, double-blind, placebo-controlled study in preschool children with URIs showed that treatment with an antihistamine–decongestant combination (brompheniramine maleate and phenylpropanolamine hydrochloride) produced no improvement in cough, rhinorrhea, or nasal congestion, although a larger proportion of the treated children (47% versus 26%) were asleep 2 hours after treatment.29 Numerous decongestants, antitussives, and expectorants are available over the counter, but there is no evidence to support their use in children. A study of phenylephrine, a topical decongestant, in children 6 to 18 months old showed no decrease in nasal obstruction with its use during a URI.30 In a study comparing placebo, dextromethorphan, and codeine for cough suppression in children 18 months to 12 years old, cough decreased in all patients within 3 days, but there was no difference in cough reduction among the three treatment groups.31 Guaifenesin, an expectorant, has not been shown to change the volume or quality of sputum or the frequency of cough in young adults with colds.32 Echinacea preparations, commonly believed to be effective in the treatment of the common cold, have been shown to have no effect on the prevention or treatment of rhinovirus infection.33 Antibiotics have no role in the treatment of uncomplicated URIs in children. Antibiotic therapy does not hasten resolution of the viral infection or reduce the likelihood of occurrence of secondary bacterial infection.34 Antibiotics are only indicated in cases of secondary bacterial infection, such as sinusitis and acute otitis media. Thus, supportive measures remain the mainstay of treatment of the common cold in children. Bulb suction with saline drops (about 1 teaspoon salt in 2 cups of water) may help relieve nasal congestion and remove secretions.

COMPLICATIONS The common cold usually resolves in about 10 to 14 days in infants and children. New-onset fever and earache during this period may herald the development of bacterial otitis media, which occurs in about 5% of colds in preschool children. Persistence of nasal symptoms for longer than 10 days may signify the development of a secondary bacterial sinusitis. Bacterial pneumonia is an uncommon secondary infection. For children with underlying reactive airways disease, wheezing is common during the course of a viral URI; at least 50% of asthma exacerbations in children are associated with viral infection. Children who experience more than one lower respiratory tract infection (such as croup or bronchiolitis) during their first year of life have an increased risk of asthma thereafter.35 Other complications are epistaxis, eustachian tube dysfunction, conjunctivitis, and pharyngitis.

Children Characteristic

Adults

Children < 6 years

Frequency

2–4 per year

One per month, September–April

Fever

Rare

Common during first 3 days

Nasal manifestations

Congestion

Colored nasal discharge

Duration of illness

5–7 days

14 days

Modified from Hendley JO. Epidemiology, pathogenesis, and treatment of the common cold. Semin Pediatr Infect Dis 1998;9:50–55.

RECENT ADVANCES Research now suggests that the symptoms of the common cold result from effects of inflammatory mediators released in response to the viral infection of the respiratory tract. As the determinants of this process are further elucidated, treatments may be developed that can interrupt or ameliorate release of inflammatory mediators and thus prevent or reduce the symptoms of the common cold. Vaccines are unlikely to be useful for prevention, given the large number of serotypes of some cold viruses as well as the lack of lasting immunity to others. The use of alcohol-based hand gels has been suggested as a

206

SECTION

C Upper Respiratory Tract and Oral Infections

means of reducing secondary transmission of respiratory illnesses in the home,36 but in one field trial, this was not shown to be effective.37 Also, virucidal tissues have been shown to be effective in preventing viral passage and transmission, and may reduce secondary transmission by about 30%.38,39 Until new methods are developed, prevention of the common cold is limited to avoiding self-inoculation (transfer of virus from contaminated fingers to nasal or conjunctival mucosa) by removing virus through handwashing or by killing virus with application of a virucide to the hands.

TABLE 29-1. Etiology of Acute Pharyngitis Etiologic Agent

Associated Disorder(s) or Clinical Findings(s)

Bacterial Streptococci Group A

Scarlet fever

Groups C and G Mixed anaerobes

CHAPTER

Vincent angina

Neisseria gonorrhoeae

29

Pharyngitis Michael A. Gerber

Corynebacterium diphtheriae

Diphtheria

Arcanobacterium haemolyticum

Scarlatiniform rash

Yersinia enterocolitica

Enterocolitis

Yersinia pestis

Plague

Francisella tularensis

Tularemia (oropharyngeal form)

Viral

Pharyngitis is an inflammation of the mucous membranes and underlying structures of the throat. Acute pharyngitis is one of the most common illnesses for which children in the United States visit primary care physicians; pediatricians make the diagnosis of acute pharyngitis, acute tonsillitis, or streptococcal sore throat more than 7 million times annually.1 Many viruses and bacteria can cause acute pharyngitis, either as a separate entity or as part of a more generalized illness. A partial list of the more common microorganisms that can cause acute pharyngitis is presented in Table 29-1. Most cases of acute pharyngitis in children and adolescents are caused by viruses and are benign and self-limited. Group A beta-hemolytic streptococci (GAS) (Streptococcus pyogenes) is the most important of the bacterial causes of acute pharyngitis. Strategies for the diagnosis and treatment of pharyngitis in children and adolescents are directed at distinguishing the large group of patients with viral pharyngitis who would not benefit from antimicrobial therapy from the much smaller group of patients with GAS pharyngitis for whom antimicrobial therapy would be beneficial. Making this distinction is extremely important in attempting to minimize the unnecessary use of antimicrobial agents in children and adolescents.

Rhinovirus

Common cold

Coronavirus

Common cold

Adenovirus

Pharyngoconjunctival fever; acute respiratory disease

Herpes simplex virus types 1 and 2 Gingivostomatitis Parainfluenza virus

Common cold; croup

Coxsackievirus A

Herpangina; hand, foot, and mouth disease

Epstein–Barr virus

Infectious mononucleosis

Cytomegalovirus

Cytomegalovirus mononucleosis

HIV

Primary HIV infection

Influenza A and B viruses

Influenza

Mycoplasmal Mycoplasma pneumoniae Chlamydial Chlamydophila psittaci

Acute respiratory disease; pneumonia

Chlamydophila pneumoniae

Pneumonia

ETIOLOGY Viruses are the most common cause of acute pharyngitis in children and adolescents. Respiratory viruses (e.g., influenza virus, parainfluenza virus, rhinovirus, coronavirus, adenovirus, and respiratory syncytial virus) are frequent causes of acute pharyngitis. Other viruses that frequently cause acute pharyngitis include coxsackievirus, echovirus, herpes simplex virus (HSV), and Epstein–Barr virus (EBV). The acute pharyngitis produced by EBV is often accompanied by other clinical findings of infectious mononucleosis (e.g., splenomegaly, generalized lymphadenopathy). Systemic infections with other viruses (e.g., cytomegalovirus, rubella virus, and measles virus) can be associated with acute pharyngitis. GAS is the most common bacterial cause of acute pharyngitis, accounting for 15% to 30% of the cases of acute pharyngitis in children. Other bacteria that can cause acute pharyngitis include groups C and G beta-hemolytic streptococci and Corynebacterium diphtheriae. Arcanobacterium haemolyticum is a rare cause of acute pharyngitis in adolescents and Neisseria gonorrhoeae can cause acute pharyngitis in sexually active adolescents. Other bacteria such as Francisella tularensis and Yersinia enterocolitica as well as mixed infections with anaerobic bacteria (e.g., Vincent angina) are rare causes of acute pharyngitis. Chlamydophila pneumoniae and Mycoplasma pneumoniae have been

Acute respiratory disease; pneumonia

HIV, human immunodeficiency virus. Modified from Bisno AL, Gerber MA, Gwaltney JM, et al. Practice guideline for the diagnosis and management of group A streptococcal pharyngitis. Clin Infect Dis 2002;35:113–125, with permission.

implicated as rare causes of acute pharyngitis, particularly in adults. Although other bacteria such as Staphylococcus aureus, Haemophilus influenzae, and Streptococcus pneumoniae are frequently isolated from throat cultures of children and adolescents with acute pharyngitis, their etiologic role in this illness has not been clearly established.

EPIDEMIOLOGY Most cases of acute pharyngitis occur during the colder months of the year when respiratory viruses (e.g., rhinovirus, coronavirus, influenza virus, and adenovirus) are prevalent. Spread among family members in the home is a prominent feature of the epidemiology of most of these

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Pharyngitis

agents, with children being the major reservoir of infection. GAS pharyngitis is primarily a disease of children 5 to 15 years of age, and, in temperate climates, its prevalence is highest in the winter and early spring. The incidence of gonococcal pharyngitis is highest among sexually active adolescents and young adults. The usual route of infection is orogenital sexual contact with an infected sexual partner. Sexual abuse must be strongly considered when N. gonorrhoeae is isolated from the pharynx of a prepubertal child. Widespread immunization with diphtheria toxoid has made diphtheria a rare disease in the United States, with fewer than 5 cases reported annually in recent years. Both group C and group G beta-hemolytic streptococci can cause acute pharyngitis with clinical features similar to those of GAS pharyngitis. Group C streptococcus is a relatively common cause of acute pharyngitis among college students and among adults who come to an emergency department.2,3 Group C streptococci can also cause epidemic, foodborne pharyngitis. Outbreaks of group C streptococcal pharyngitis related to ingestion of contaminated food products such as unpasteurized cow milk have been reported in families and schools.4 Although there have been several well-documented foodborne outbreaks of group G streptococcal pharyngitis, the etiologic role of group G streptococci in acute, endemic pharyngitis remains unclear. A community-wide, respiratory outbreak of group G streptococcal pharyngitis in a pediatric population was described in which group G streptococcus was isolated from 56 of 222 (25%) consecutive children with acute pharyngitis seen in a private pediatric office.5 Results of DNA fingerprinting of the group G streptococcal isolates suggested that 75% of them were the same strain. The role of group C and group G streptococci in acute pharyngitis may be underestimated for several reasons. In the clinical laboratory, anaerobic incubation increases the yield of these organisms, but many laboratories do not routinely use anaerobic incubation for throat cultures. In addition, because laboratories may only report bacitracinsusceptible streptococci (consistent with GAS) and many group C and group G streptococci are bacitracin-resistant, group C and group G streptococci would be missed. Finally, many clinicians are no longer performing throat cultures but relying solely on rapid antigen detection tests (RADTs), and group C and group G streptococci would not be identified by an RADT for GAS.6

CLINICAL MANIFESTATION The presence of certain clinical and epidemiological findings suggests GAS as the cause of an episode of acute pharyngitis (Box 29-1). Patients with GAS pharyngitis commonly present with sore throat (usually of sudden onset), severe pain on swallowing, and fever. Headache, nausea, vomiting, and abdominal pain can also be present. Examination typically reveals tonsillopharyngeal erythema with or without exudates, and tender, enlarged anterior cervical lymph nodes. Other findings may include a beefy, red, swollen uvula; petechiae on the palate; excoriated nares (especially in infants); and a scarlitiniform rash. However, none of these findings is specific for GAS pharyngitis. Many patients with GAS pharyngitis exhibit signs and symptoms that are milder than a “classic” case of this illness. Some of these patients have bona fide GAS pharyngitis (i.e., have a rise in antistreptococcal antibodies), whereas others are merely colonized with GAS and have pharyngitis due to an intercurrent viral infection. Scarlet fever is an upper respiratory tract infection associated with a characteristic rash, that is caused by a pyrogenic exotoxin (erythrogenic toxin)-producing GAS in individuals who do not have antitoxin antibodies. Scarlet fever is encountered less often and is less virulent than in the past, but its incidence is cyclical, depending on both the prevalence of toxin-producing strains of GAS and the immune status of the population. The modes of transmission, age distribution, and other epidemiologic features are otherwise similar to those of GAS pharyngitis. The rash of scarlet fever appears within 24 to 48 hours of the onset of signs and symptoms, although it can appear with the first signs of

CHAPTER

29

207

BOX 29-1. Clinical and Epidemiologic Characteristics of Group A Beta-Hemolytic Streptococci (GAS) and Viral Pharyngitis FEATURES SUGGESTIVE OF GAS ETIOLOGY Sudden onset Sore throat Fever Scarlet fever rash Headache Nausea, vomiting, and abdominal pain Inflammation of pharynx and tonsils Patchy discrete exudates Tender, enlarged anterior cervical nodes Patient aged 5–15 years Presentation in winter or early spring History of exposure FEATURES SUGGESTIVE OF VIRAL ETIOLOGY Conjunctivitis Coryza Cough Hoarseness Myalgia Diarrhea Characteristic exanthems Characteristic enanthems Modified from Bisno AL, Gerber MA, Gwaltney JM, et al. Practice guideline for the diagnosis and management of group A streptococcal pharyngitis. Clin Infect Dis 2002;35:113–125, with permission.

illness. The rash often begins around the neck and then spreads over the trunk and extremities. It is a diffuse, finely papular, erythematous eruption producing a bright red discoloration of the skin that blanches with pressure. Involvement is often more intense along the creases in the antecubital area, axillae, and groin. The involved skin has a goosepimple appearance and feels rough. The face is usually spared, although the cheeks may be erythematous with pallor around the mouth (Figure 29-1). After 3 to 4 days, the rash begins to fade and is followed by desquamation, first on the face, progressing downward, and often resembling that following mild sunburn. Occasionally, sheet-like desquamation may occur around the free margins of the fingernails, the palms, and the soles. Examination of the pharynx of a patient with scarlet fever reveals essentially the same findings as with GAS pharyngitis. In addition, the tongue is usually coated and the papillae are swollen. After desquamation, the reddened papillae are prominent, giving the tongue a strawberry appearance. In contrast, the presence of certain clinical findings (e.g., conjunctivitis, cough, hoarseness, coryza, anterior stomatitis, discrete ulcerative lesions, viral exanthema, myalgia and diarrhea) suggests a virus rather than GAS as the cause of an episode of acute pharyngitis (see Box 29-1). Acute pharyngitis caused by adenovirus is typically associated with fever, erythema of the pharynx, enlarged tonsils with exudate, and enlarged cervical lymph nodes. Adenoviral pharyngitis can be associated with conjunctivitis, and, when it is, it is referred to as pharyngoconjunctival fever. The pharyngitis of pharyngoconjunctival fever can persist for up to 7 days, the conjunctivitis for up to 14 days, and both resolve spontaneously. Outbreaks of pharyngoconjunctival fever have been associated with transmission in swimming pools; widespread epidemics and sporadic cases also occur. Enteroviruses (coxsackievirus, echovirus, and newer enteroviruses) can cause acute pharyngitis, especially during the summer and early fall. The pharynx may be erythematous but tonsillar exudate and cervical adenopathy are unusual. Fever may be prominent. Resolution usually occurs within a few days. Herpangina is a specific syndrome caused by coxsackie A or B virus or echoviruses and is characterized by fever and painful, discrete, gray-white papulovesicular lesions on an erythematous base in the posterior oropharynx. These lesions become ulcerative and usually resolve within 7 days. Hand, foot, and mouth disease is a specific syndrome caused by coxsackievirus A16

208

SECTION

C Upper Respiratory Tract and Oral Infections

A

Figure 29-1. Child has group A streptococcal pharyngitis and scarlatiniform rash, with characteristic circumoral pallor. (Courtesy of J.H. Brien.©)

B Figure 29-2. Epstein–Barr virus mononucleosis in an adolescent girl. Tonsillar erythema and exudates (A) and periorbital edema (B) are clues to the diagnosis. (Courtesy of J.H. Brien.©)

virus. It is characterized by painful vesicles and ulcers throughout the oropharynx associated with vesicles on the palms, soles, and sometimes on the trunk or extremities. These lesions usually resolve within 7 days. Primary oral HSV infections usually occur in young children and typically produce acute gingivostomatitis associated with ulcerating vesicular lesions throughout the anterior mouth including the lips, but sparing the posterior pharynx. The gingivostomatitis can last up to 2 weeks and is often associated with high fever. The pain may be intense and the oral intake of fluids may be impaired, leading to dehydration. In adolescents and adults HSV can also produce a mild pharyngitis that may or may not be associated with typical vesicular, ulcerating lesions. Acute pharyngitis is a common finding in adolescents and young adults with infectious mononucleosis caused by EBV. The pharyngitis of infectious mononucleosis can be severe with clinical findings identical to those of GAS pharyngitis (Figure 29-2A). However, generalized lymphadenopathy and hepatosplenomegaly can also be present. Posterior cervical lymphadenopathy, presternal and periorbital edema, and palatal petechiae are distinctive if present (Figure 29-2B). If amoxicillin has been given, an intense maculopapular rash is expected (Figure 29-3). Fever and pharyngitis typically last 1 to 3 weeks, whereas the lymphadenopathy and hepatosplenomegaly resolve over 3 to 6 weeks. Laboratory findings include the presence of

atypical lymphocytosis (Figure 29-4), heterophile antibodies, and specific antibodies to EBV antigens. The acute pharyngitis caused by Arcanobacterium haemolyticum can resemble GAS pharyngitis, including the presence of a scarlitiniform rash. In rare cases, A. haemolyticum can produce a membranous pharyngitis that can be confused with diphtheria. Pharyngeal diphtheria is characterized by a grayish brown pseudomembrane that can be limited to one or both tonsils or can extend widely to involve the nares, uvula, soft palate, pharynx, larynx, and tracheobronchial tree. Involvement of the tracheobronchial tree may lead to life-threatening respiratory obstruction. Soft-tissue edema and prominent cervical and submental lymphadenopathy may create a bull-neck appearance.

DIAGNOSIS When seeing a child or adolescent with acute pharyngitis, physicians in the United States must primarily distinguish between GAS and viral pharyngitis. Efforts have been made to incorporate clinical and epidemiological features of acute pharyngitis into scoring systems that attempt to predict the probability that a particular illness is caused by GAS.7–9 These clinical scoring systems are helpful in identifying

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Pharyngitis

CHAPTER

29

209

Figure 29-4. Peripheral blood smear showing atypical lymphocytes (arrows) in a patient with Epstein–Barr virus mononucleosis. Note the abundant cytoplasm with vacuoles, and deformation of cell by surrounding cells. (Courtesy of J.H. Brien.©)

A

B Figure 29-3. Adolescents (A, B) with Epstein–Barr virus mononucleosis who received amoxicillin and developed diffuse erythematous raised rashes. Note predominance on trunk, and coalescence (B). (Courtesy of J.H. Brien.©)

patients at such low risk of GAS infection that a throat culture or RADT is usually unnecessary. However, the signs and symptoms of GAS and non-GAS pharyngitis overlap broadly, and the clinical diagnosis of GAS pharyngitis cannot be made with accuracy, even by the most experienced physicians. Therefore, recent guidelines from the Infectious Diseases Society of America (IDSA),10 as well as guidelines from the American Academy of Pediatrics (AAP)11 and the American Heart Association (AHA),12 indicate that microbiologic

conÀrmation (either with a throat culture or RADT) is required for the diagnosis of GAS pharyngitis. The decision to perform a microbiologic test on a child or adolescent with acute pharyngitis should be based on the clinical and epidemiologic characteristics of the illness (see Box 29-1). A history of close contact with a documented case of GAS pharyngitis or a high prevalence of GAS infections in the community can also be helpful. Testing usually need not be performed on patients with acute pharyngitis whose clinical and epidemiologic features do not suggest GAS as the etiology. Selective use of diagnostic studies for GAS will not only increase the proportion of positive test results, but also the percentage of patients with positive tests who are truly infected rather than merely GAS carriers. Recently, new practice guidelines from the Centers for Disease Control and Prevention (CDC), the American Academy of Family Physicians (AAFP), and the American College of Physicians– American Society of Internal Medicine (ACP–ASIM) recommended the use of a clinical algorithm without microbiologic conÀrmation as an acceptable approach to the diagnosis of GAS pharyngitis in adults only.13 Although the goal of this algorithm-based strategy was to reduce the inappropriate use of antibiotics in adults with pharyngitis, there is concern that use would result in the administration of antimicrobial treatment to an unacceptably large number of adults with non-GAS pharyngitis.14 A study was recently performed to assess the impact of six different guidelines (including the IDSA and CDC/AAFP/ACP–ASIM guidelines) on the identiÀcation and treatment of GAS pharyngitis in children and adults.15 Guidelines that recommended selective use of RADTs and/or throat culture and treatment based only on positive test results signiÀcantly reduced the inappropriate use of antibiotics in adults. In contrast, the empiric strategy proposed in the CDC/ AAFP/ACP–ASIM guidelines resulted in the administration of unnecessary antibiotics to an unacceptably large number of adults. Before abandoning the concept of treatment only after laboratory conÀrmation of GAS in adults with pharyngitis (IDSA guideline), additional studies need to be performed to compare empiric and laboratory-based strategies in terms of relevant patient outcomes and cost.

THROAT CULTURES Culture on sheep blood agar of a specimen obtained by throat swab is the standard laboratory procedure for the microbiologic conÀrmation of GAS pharyngitis.16 If performed correctly, a throat culture has a

210

SECTION

C Upper Respiratory Tract and Oral Infections

sensitivity of 90% to 95% in detecting the presence of GAS in the pharynx.17 Several variables impact on the accuracy of throat culture results. One of the most important is the manner in which the swab is obtained.18,19 Throat swab specimens should be obtained from the surface of both tonsils (or tonsillar fossae) and the posterior pharyngeal wall. Other areas of the pharynx and mouth are not acceptable sampling sites and should not be touched during the culturing procedure. Even with an appropriately collected specimen, false-negative results can occur if the patient has received antimicrobials before the throat swab is taken. Anaerobic incubation and the use of selective culture media have been reported to increase the sensitivity of throat cultures.20,21 However, data regarding the impact of the atmosphere of incubation and the culture media are conflicting, and, in the absence of definite benefit, the increased cost and effort associated with anaerobic incubation and selective culture media are difficult to justify.21–24 Duration of incubation can impact the yield of throat cultures. Once plated, cultures should be incubated at 35 to 37°C for 18 to 24 hours prior to reading. An additional overnight incubation at room temperature, however, identifies a considerable number of positive cultures that would not otherwise have been identified. Recently, Armengol and coworkers found that more than 40% of the positive confirmatory throat cultures performed on patients with pharyngitis and negative RADT results were negative after 24 hours of incubation but positive after 48 hours.25 Therefore, although initial therapeutic decisions can be guided by negative result at 24 hours, it is advisable to wait 48 hours for definitive results. The clinical significance of the number of colonies of GAS present on the throat culture plate is controversial. Density of bacteria is likely to be greater in patients with bona fide acute GAS pharyngitis than in GAS carriers. However, there is too much overlap in the colony counts between those acutely infected with GAS and GAS carriers to permit differentiation on the basis of degree of positivity alone.24 The bacitracin disk test is the most widely used method in physicians’ offices for the differentiation of GAS from other betahemolytic streptococci on a sheep blood agar plate. This test provides a presumptive identification based on the observation that > 95% of GAS demonstrate a zone of inhibition around a disk containing 0.04 units of bacitracin, whereas 83% to 97% of non-GAS are not inhibited by bacitracin.24 An alternative and highly specific method for the differentiation of beta-hemolytic streptococci is the performance of group-specific cell wall carbohydrate antigen detection test directly on isolated bacterial colonies. Commercial kits employing group-specific antisera are available for this purpose. Additional expense for the minimal improvement in accuracy may not be justified.24

RAPID ANTIGEN DETECTION TESTS RADTs have been developed for the identification of GAS directly from throat swabs. Although these RADTs are more expensive than blood agar cultures, the advantage they offer over the traditional procedure is the speed with which they can provide results. Rapid identification and treatment of patients with GAS pharyngitis can reduce the risk of the spread of GAS, allow the patient to return to school or work sooner, and speed clinical improvement.17,26 In addition, in certain environments (e.g., emergency departments) the use of RADTs compared with throat cultures has significantly increased the number of patients appropriately treated for GAS pharyngitis.27 The majority of currently available RADTs have specificities of 95% or greater compared with blood agar cultures.28 Therapeutic decisions, therefore, can be made with confidence on the basis of a positive RADT result. However, the sensitivity of RADTs is between 70% and 90%.28 Although it has been suggested that many of the falsely negative RADT results occur in patients who are merely GAS carriers, it has been demonstrated that a large proportion of patients with falsely negative RADT results truly are infected with GAS.29 The first RADTs utilized latex agglutination methodology, were relatively insensitive, and had unclear endpoints.28 Subsequent tests

based on enzyme immunoassay techniques had a more sharply defined endpoint and increased sensitivity. More recently, RADTs using optical immunoassay (OIA) and chemiluminescent DNA probes have been developed.28 These tests may be more sensitive than other RADTs and perhaps even as sensitive as blood agar plate cultures.28 However, because of conflicting and limited data about the OIA and other commercially available RADTs, advisory groups recommend that physicians continue to perform a confirmatory blood agar culture on children and adolescents suspected on clinical grounds of having GAS pharyngitis who have a negative RADT result. Physicians electing to use any RADT in children and adolescents without culture backup of negative results should do so only after demonstrating with adequate sample size calculation that the RADT is as sensitive as throat culture in their own practice.10,11 Because of considerable variability in study designs and culture techniques, it is difficult to compare the sensitivity of an RADT as determined in one study to the sensitivity of another RADT as determined in a different study.28 The relative sensitivities of different RADTs can only be determined by direct comparisons. There have been only four reports of direct comparisons of different RADTs reported in the English literature.30–33 Few studies have investigated the performance of RADTs in office practice.25,31–35 In one study,25 performed over three winter periods and using the on-site physician office laboratory at the pediatric group practice, RADT had a sensitivity of approximately 85% compared with a single blood agar plate culture. The investigators concluded that the sensitivity of this particular RADT was too low to consider abandoning the confirmatory throat culture in their practice. Investigators in a different pediatric group practice reviewed their experience with 11 427 RADTs performed between 1996 and 1999.36 Only 2.4% of specimens negative by RADT were positive by culture. Investigators concluded that culture confirmation of specimens negative for RADT was costly and may not be necessary in all circumstances.36 Neither blood agar culture nor RADT accurately differentiates individuals with bona fide GAS pharyngitis from GAS carriers. However, they facilitate withholding antimicrobial therapy from the great majority of patients with sore throats whose cultures or RADTs are negative. There are an estimated 6.7 million visits to primary care providers by adults who complain of sore throat each year in the United States, and antimicrobial therapy is prescribed at 73% of these visits.37 Recent trends show a modest decline in the use of antibiotics in children and adolescents diagnosed with pharyngitis to 68.6% in one study in 1999 to 2000,38 and to 54% in another study in 2003.39 Antistreptococcal antibody titers have no value in the diagnosis of acute GAS pharyngitis. They are useful in prospective epidemiologic studies to differentiate true GAS infections from GAS carriage. Antistreptococcal antibodies are valuable for confirmation of prior GAS infections in patients suspected of having acute rheumatic fever or poststreptococcal acute glomerulonephritis.

REPEAT DIAGNOSTIC TESTING The majority of asymptomatic persons who have a positive throat culture or RADTs after completing a course of appropriate antimicrobial therapy for GAS pharyngitis are GAS carriers.40 Follow-up throat cultures (or RADTs) are not indicated for such patients. Followup throat culture (or RADTs) for an asymptomatic individual should be performed in those with a history of rheumatic fever, and should be considered in patients who develop acute pharyngitis during outbreaks of acute rheumatic fever or poststreptococcal acute glomerulonephritis, and in individuals in closed or semiclosed communities during outbreaks of GAS pharyngitis.40

TREATMENT Antimicrobial therapy is indicated for individuals with symptomatic pharyngitis after the presence of GAS in the throat has been confirmed by either throat culture or RADT. In situations in which the clinical

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Pharyngitis

and epidemiologic findings are highly suggestive of GAS, antimicrobial therapy can be initiated while awaiting microbiologic confirmation, provided that such therapy is discontinued if culture or RADT is negative. Antimicrobial therapy for GAS pharyngitis shortens the clinical course of the illness.26 However, GAS pharyngitis is usually a self-limited disease, and most signs and symptoms resolve spontaneously within 3 or 4 days of onset without antimicrobial therapy.41 In addition, the initiation of antimicrobial therapy can be delayed for up to 9 days after the onset of GAS pharyngitis and still prevent the occurrence of acute rheumatic fever.42 Therefore, there can be flexibility in initiating antimicrobial therapy during the evaluation of an individual patient with presumed GAS pharyngitis. Children and adolescents with viral pharyngitis should be treated symptomatically; antimicrobial therapy is not indicated and is ineffective. Penicillin and its congeners (such as ampicillin and amoxicillin), as well as numerous cephalosporins, macrolides, and clindamycin, are effective treatment for GAS pharyngitis. Several advisory groups have recommend penicillin as the treatment of choice for this infection.10–12 Group A streptococcus has remained exquisitely susceptible to betalactam agents over five decades.43 Amoxicillin is often used because of acceptable taste of suspension; efficacy appears to equal penicillin. Orally administered erythromycin is indicated for patients allergic to penicillin. Other macrolides, such as clarithromycin or azithromycin, are also effective. First-generation cephalosporins are acceptable in penicillin-allergic patients who do not manifest immediate-type hypersensitivity to beta-lactam antibiotics. Sulfa drugs, including trimethoprim-sulfamethoxazole, and tetracyclines are not effective and should not be used for GAS pharyngitis. Some investigators concluded that cephalosporins should be treatments of choice for GAS tonsillopharyngitis44 following a metaanalysis performed of 35 clinical trials completed between 1970 and 1999 in which a cephalosporin was compared with penicillin for the treatment. However, others concluded that several major flaws make it impossible to accept the validity of this conclusion.45 Although the use of cephalosporins for GAS pharyngitis could reduce the number of persons (most merely chronic carriers) who harbor GAS in their throats after completing therapy, their use would be associated with substantial economic and ecologic costs. Penicillin has stood the test of time satisfactorily for five decades, and there are compelling reasons (e.g., its narrow antimicrobial spectrum, inexpensive cost, and impressive safety profile) to continue to recommend it as the drug of choice for GAS pharyngitis. Oral penicillin must be administered multiple times a day for 10 days in order to achieve maximal rates of eradication of GAS. Reduced frequency of dosing and shorter treatment courses (< 10 days) may result in better patient adherence to therapy than is seen with 10 days of oral penicillin. It has been reported that several antimicrobial agents, including clarithromycin, cefuroxime, cefixime, ceftibuten, cefdinir, and cefpodoxime, are effective in eradication of GAS from the pharynx when administered for ≤ 5 days.46–49 However, many of the studies of short-course therapy have serious methodologic flaws that cloud validity of conclusions. In addition, the spectra of these antibiotics are much broader than that of penicillin, and, even if administered for short courses, they are more expensive.46 Therefore, additional studies are needed before these short-course regimens can be recommended.10 Attempts to treat GAS pharyngitis with a single daily dose of penicillin have been unsuccessful.52 In recent years, investigators have demonstrated that several antimicrobial agents, including azithromycin, cefadroxil, cefixime, ceftibuten, cefpodoxime, cefprozil, and cefdinir, are effective in eradicating pharyngeal streptococci when given as a single daily dose.10,46,49,53 However, these agents are expensive and have broad spectra of activity compared with penicillin. Preliminary investigations have demonstrated that once-daily amoxicillin therapy is effective in the treatment of GAS pharyngitis.54,55 If confirmed by additional investigations, once-daily amoxicillin therapy, because of its low cost and relatively narrow spectrum, could become an alternative regimen for the treatment of GAS pharyngitis. Table 29-2

CHAPTER

29

211

TABLE 29-2. Antimicrobial Therapy for Group A Beta-Hemolytic Streptococci (GAS) Pharyngitis Route of Administration, Antimicrobial Agent

Dosage

Duration

Children: 250 mg bid or tid

10 days

Adolescents and adults: 250 mg tid or qid

10 days

ORAL

Penicillin

Adolescents and adults: 500 mg bid 10 days INTRAMUSCULAR

Benzathine penicillin G

6.0 μ 105 U (for patients ≤ 27 kg)

1 dose

1.2 μ 10 U (for patients > 27 kg)

1 dose

6

Mixtures of benzathine Varies with formulation and procaine penicillin G

a

ORAL, FOR PATIENTS ALLERGIC TO PENICILLIN

Erythromycin

Varies with formulation

10 days

First-generation cephalosporinsb

Varies with agent

10 days

a

Dose should be determined on basis of benzathine component. These agents should not be used to treat patients with immediate-type hypersensitivity to b-lactam antibiotics. Modified from Bisno AL, Gerber MA, Gwaltney JM, et al. Practice guideline for the diagnosis and management of group A streptococcal pharyngitis. Clin Infect Dis 2002;35:113–125, with permission. b

gives recommendations for several agents proven to be effective for the treatment of GAS pharyngitis.10 Intramuscular benzathine penicillin G is preferred in those patients unlikely to complete a full 10-day course of oral therapy. Although penicillin resistance has not occurred in GAS anywhere in the world,56 there have been geographic areas with relatively high levels of resistance to macrolide antibiotics.50,51 The rate of macrolide resistance among isolates of GAS in the United States has generally remained < 5%. In an investigation of 245 pharyngeal isolates and 56 invasive isolates of GAS obtained between 1994 and 1997 from 24 states and the District of Columbia, only 8 (2.6%) of the 301 isolates were determined to be macrolide-resistant.43 However, higher resistance rates have occasionally been reported. For example, 9% of pharyngeal and 32% of invasive GAS strains collected in San Francisco during 1994 to 1995 were reported to be macrolide-resistant.57 During a longitudinal investigation of GAS disease in a single elementary school in Pittsburgh, investigators found that 48% of isolates of GAS collected between 2000 and 2001 were resistant to erythromycin; none was resistant to clindamycin.58 Molecular typing indicated that this outbreak was due to a single strain of GAS. In addition, of 100 randomly selected isolates of GAS obtained from the community in 2001, 38 (38%) were resistant to erythromycin.58 The results of a prospective, multicenter, United States communitybased surveillance of pharyngeal isolates of GAS recovered from children 3 to 18 years of age during three successive respiratory seasons between 2000 and 2003 have been reported.59 Macrolide resistance was < 5%, clindamycin resistance was 1.04%, and rates were stable. In addition, there was no evidence of wide dissemination of specific macrolide-resistance clones, increasing clindamycin resistance, or increasing erythromycin minimum inhibitory concentrations over the 3-year study period. There was, however, considerable geographic variability in macrolide resistance rates in each study year, as well as year-to-year variability at individual study sites.59 Clinicians should be aware of local resistance rates. Acute rheumatic fever has not been described as a complication of either group C or group G streptococcal pharyngitis. Although acute glomerulonephritis has been reported as a complication of group C

212

SECTION

C Upper Respiratory Tract and Oral Infections

streptococcal pharyngitis, it is extremely unusual and a causal relationship has not been established. Therefore, the primary reason to identify either group C or group G streptococcus as the cause of acute pharyngitis is to initiate antimicrobial therapy that may mitigate the clinical course of the infection. However, there is currently no convincing evidence from controlled studies of clinical response to antimicrobial therapy in patients with acute pharyngitis and either group C or group G streptococcus isolated from their pharynx. If one elects to treat either group C or group G streptococcal pharyngitis, the treatment should be similar to that for GAS pharyngitis with penicillin as the antimicrobial agent of choice.6

COMPLICATIONS GAS pharyngitis can be associated with suppurative and nonsuppurative complications. Suppurative complications result from the spread of GAS to adjacent structures and include peritonsillar abscess, retropharyngeal abscess, cervical lymphadenitis, sinusitis, otitis media, and mastoiditis. Before antimicrobial agents were available, suppurative complications of GAS pharyngitis were common; however, antimicrobial therapy has greatly reduced the frequency of such complications. Acute rheumatic fever, acute poststreptococcal glomerulonephritis, and poststreptococcal reactive arthritis are recognized nonsuppurative sequelae of GAS pharyngitis (see Chapter 118, Streptococcus pyogenes). Acute rheumatic fever occurs after an episode of GAS pharyngitis (usually after a 2- to 4-week latent period) and not after GAS infections of the skin. Appropriate antimicrobial therapy begun within 9 days of the onset of pharyngitis can prevent this complication. In contrast to acute rheumatic fever, acute poststreptococcal glomerulonephritis can occur after a GAS infection of either the pharynx or skin and does not appear to be prevented by antimicrobial therapy of the antecedent GAS infection. The latent period for glomerulonephritis is about 3 weeks following skin infection and 10 days following upper respiratory tract infection. Poststreptococcal reactive arthritis is similar to other postinfectious arthritides. The relationship of this entity to acute rheumatic fever is still unclear.

TREATMENT FAILURES, CHRONIC CARRIAGE, AND RECURRENCES Antimicrobial treatment failures with GAS pharyngitis have traditionally been classified as either clinical or bacteriologic failures. However, the significance of clinical treatment failures (usually defined as persistent or recurrent signs or symptoms suggestive of GAS pharyngitis) is difficult to determine because GAS pharyngitis is a self-limited illness even without antimicrobial therapy.41 In addition, without the repeated isolation of the infecting strain of GAS (i.e., true bacteriologic treatment failure), it is particularly difficult to determine the clinical significance of persistent or recurrent signs or symptoms suggestive of GAS pharyngitis. Bacteriologic treatment failures can be classified as either true or apparent failures. True bacteriologic failure refers to the inability to eradicate the specific strain of GAS causing an acute episode of pharyngitis with a complete course of appropriate antimicrobial therapy. Apparent bacteriologic treatment failure can reflect a variety of circumstances. Apparent bacteriologic failures can occur when newly acquired GAS isolates are mistaken for the original infecting strain of GAS, when the infecting strain of GAS is eradicated but then rapidly reacquired, or when adherence to antimicrobial therapy is poor – the majority of apparent bacteriologic treatment failures are manifestations of the GAS carrier state. These chronic carriers have GAS in their pharynx but no immunologic response to the organism. They may be colonized for 6 to 12 months or more, during which time they may experience episodes of intercurrent viral pharyngitis and be identified to have a positive test for GAS. Carriers are unlikely to

spread GAS to their close contacts and are at very low, if any risk, for developing suppurative or nonsuppurative complications.60,61 During the winter and spring in temperate climates, as many as 20% of asymptomatic school-aged children carry GAS.60 GAS carriers should not be given antimicrobial therapy; the primary approach to the suspected or confirmed carrier is reassurance. A throat culture or RADT should be performed whenever the patient has symptoms and signs suggestive of GAS pharyngitis, but should be avoided when symptoms are more typical of viral illnesses (see Box 29-1). Each clinical episode confirmed with a positive throat culture or RADT should be treated. Identification and eradication of the streptococcal carrier state are desirable in certain specific situations (Box 29-2). When antimicrobial therapy is employed, oral clindamycin (20 mg/kg per day up to 450 mg, divided into three equal doses) for 10 days is preferred,43 but intramuscular benzathine penicillin (alone or in combination with procaine penicillin) plus oral rifampin (20 mg/kg per day divided into two equal doses (maximum dose 300 mg) for 4 days beginning on the day of the penicillin injection)32 is also effective. Chronic carriage can recur upon re-exposure to GAS. Several explanations have been proposed for true bacteriologic treatment failures. No penicillin-resistant strains of GAS have been identified. The role of penicillin tolerance (i.e., a discordance between the concentration of penicillin required to inhibit and to kill the organisms) in true bacteriologic treatment failures has never been established.62,63 The precise role that bacteria present in the normal pharyngeal flora contribute to true bacteriologic failures either by enhancing the colonization and growth of GAS in the upper respiratory tract or by producing beta-lactamases that inactive penicillin has not been determined.56 Routine throat culture (or RADT) is not indicated for asymptomatic individuals following completion of antimicrobial therapy for GAS pharyngitis. In a symptomatic patient, a throat culture (or RADT) is usually performed and, if positive, many clinicians would elect to administer a second course of antimicrobial therapy. When repeated episodes of spread of GAS occurs among family members, some physicians would perform simultaneous cultures of all family contacts and treat those whose cultures are positive. There is no credible evidence that family pets are reservoirs for GAS, nor do they contribute to spread. The patient with repeated episodes of acute pharyngitis associated with a positive throat culture (or RADT) is a common and difficult problem for the practicing physician. The fundamental question that must be addressed is whether this patient is experiencing repeated episodes of bona fide GAS pharyngitis or is a GAS carrier experiencing repeated episodes of viral pharyngitis. The latter situation is considerably more common than the former. Such a patient is likely to be a GAS carrier if: (1) the clinical and epidemiologic findings suggest a viral etiology; (2) there is little clinical response to appropriate antimicrobial therapy; (3) throat culture (or RADT) is positive between episodes of pharyngitis; and (4) there is no serologic response to GAS extracellular antigen (e.g., antistreptolysin O, antideoxyribonucleases B). In contrast, the patient with repeated

BOX 29-2. Indications for Identification and Eradication of the Streptococcal Carrier State • A personal or family history of rheumatic fever • During outbreaks of acute rheumatic fever or acute poststreptococcal glomerulonephritis • Outbreaks of GAS in closed communities, such as military barracks or prisons, or semiclosed communities, such as college campuses or boarding schools • Family with repeated episodes of spread of GAS among members • Case in which tonsillectomy is only being considered because of chronic carriage of GAS • An inordinate amount of anxiety about GAS among family members despite education and reassurance GAS, group A beta-hemolytic streptococci.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Infections Related to the Upper and Middle Airways

episodes of acute pharyngitis associated with positive throat cultures (or RADTs) for GAS is likely to be experiencing repeated episodes of bona fide GAS pharyngitis if: (1) the clinical and epidemiologic findings suggest GAS pharyngitis; (2) there is a demonstrable clinical response to appropriate antimicrobial therapy; (3) throat culture (or RADT) is negative between episodes of pharyngitis; and (4) there is a serologic response to GAS extracellar antigens. If determined that the patient is experiencing repeated episodes of bona fide GAS pharyngitis, some physicians have suggested use of prophylactic oral penicillin V. However, the efficacy of this regimen has not been proven, and antimicrobial prophylaxis is not recommended except to prevent recurrences of rheumatic fever in patients who have experienced a previous episode of rheumatic fever. Tonsillectomy may be considered in the rare patient whose symptomatic episodes do not diminish in frequency over time and in whom no alternative explanation for the recurrent GAS pharyngitis is evident. However, tonsillectomy has been demonstrated to be beneficial for a relatively small group of these patients, and any benefit can be expected to be relatively short-lived.64–66

CHAPTER

30

CHAPTER

30

213

tonsillar space.1–5 Retropharyngeal and lateral pharyngeal abscesses are infections of one of the deep cervical spaces; these infections are generally secondary to contiguous spread of infection from the oropharynx but occasionally can be secondary to penetrating injuries or dental infections.6–12 These three supraglottic infections share some clinical features, such as progression from cellulitis to organized phlegmon (a preabscess stage of organized deep inflammation) and finally to mature abscess. Acute epiglottitis, a rare disease since the widespread use of the conjugate Haemophilus influenzae type b vaccine, is a bacterial infection of the epiglottis and supraglottic structures. Acute epiglottitis is so rapidly progressive and life-threatening at the stage of cellulitis that abscess is rare. Airway obstruction and extreme toxicity can occur in all four of these supraglottic infections. The patient appears to be undergoing a toxic reaction and is weak and in pain. Leukocytosis with absolute elevation of the number of neutrophils and bands is the usual finding. The causative bacteria are similar for all sites of supraglottic infection. These infections tend to have distinctive manifestations and complications and to affect different age groups (Table 30-1). If not recognized early and managed appropriately, they can all be lifethreatening. Differentiation from other airway infections is discussed in Chapter 23, Respiratory Tract Symptom Complexes, Table 23-4.

ANATOMIC RELATIONSHIPS

Infections Related to the Upper and Middle Airways Richard H. Schwartz Supraglottic infections include peritonsillar abscess, retropharyngeal abscess, lateral pharyngeal (parapharyngeal) abscess, and acute epiglottitis. Classic Lemierre disease, a complication of the above deep cervical abscesses, comprises postanginal septic thrombophlebitis of the internal jugular vein with septic emboli to the lungs. Cases of retropharyngeal abscess and Lemierre disease are increasing while acute epiglottitis has virtually disappeared in children. Peritonsillar abscess (quinsy), the most common of supraglottic infections, is the mature phase of a bacterial infection of the peri-

There is a potential space between the capsule of the palatine tonsil and the pillars of the fauces enfolding the tonsil. There are also three clinically important spaces between the deep cervical fascial layers: the peritonsillar area surrounding the tonsils, the retropharyngeal space, and the lateral pharyngeal space. Each of the infections described in this chapter begins in one of these spaces. The retropharyngeal space extends longitudinally downward from the base of the skull to the posterior mediastinum. The posterior boundary of the space is the prevertebral fascia and the anterior boundary is the posterior portion of the pretracheal fascia. The retropharyngeal space communicates with the lateral pharyngeal space, where its lateral boundary is the carotid sheath, which contains the carotid artery and jugular vein. The lateral pharyngeal space is in the upper neck, above the hyoid bone, between the pretracheal fascia of the visceral compartment medially and the superficial fascia, which invests the parotid gland, internal pterygoid muscle, and mandible, laterally. Lateral pharyngeal

TABLE 30-1. Clinical Features of Peritonsillar, Retropharyngeal, and Lateral Pharyngeal Abscesses Disease

Typical Age of Patient

Original Infection

Location

Peritonsillar abscess

Adolescents, adults

Tonsillitis

Tonsillar space

Retropharyngeal abscess

Infancy to age 14 years

Pharyngitis, Between posterior Hyperextension of the Rupture and aspiration; Antibiotic therapy,a pharyngeal pharynx and neck, puddling of saliva, contiguous spread to surgical drainage,b trauma, dental prevertebral fascia gurgling respirations posterior mediastinum, airway monitoring infection lateral pharyngeal space

Lateral pharyngeal abscess

Older children Tonsillitis, ages 5–12; otitis media, adolescents mastoiditis, ages 13–19; parotitis, young adults dental manipulation

a

For the preabscess stage. For the abscess stage, in addition to antibiotic therapy.

b

Clinical Findings

Site of Extension and Complications

Definitive Management

Tonsillar fullness, deviation of the uvula, muffled voice, inability to open mouth widely

Rupture and aspiration; Antibiotic therapya contiguous spread to and surgical drainageb pterygomaxillary space

Anterior and In the anterior Carotid erosion; Antibiotic therapy,a posterior compartment: swelling septicemia; surgical drainage,b pharyngomaxillary of parotid area, trismus, airway obstruction; airway monitoring space prolapse of the tonsil or distant spread to tonsillar fossa; in the intracranial sites, lung; posterior compartment: contiguous spread to septicemia, mild pain, mediastinum or trismus

214

SECTION

C Upper Respiratory Tract and Oral Infections

inflammation occurs lateral to the great vessel sheath. Because the lateral pharyngeal space is contiguous with the retropharyngeal space, inflammation or mature abscesses of both fascial spaces can occur simultaneously.

ETIOLOGIC AGENTS AND GENERAL APPROACH The similarity in the polymicrobial microbiology (which consists of aerobic and anaerobic bacteria) of these infections reflects the host’s oropharyngeal (in peritonsillar and lateral pharyngeal abscess) or nasopharyngeal (in retropharyngeal abscess) flora.1,13,14 Predominant aerobic bacterial pathogens are Streptococcus pyogenes and bhemolytic group C streptococci, which are isolated from one-third of cultures of these infections, and methicillin-susceptible or -resistant Staphylococcus aureus (MSSA, MRSA). Peritonsillar or retropharyngeal cellulitis or abscess in young children is more likely to be caused by Streptococcus pyogenes, nongroup A streptococcus, or Staphylococcus aureus, alone or mixed. In older children, especially adolescents, S. aureus, with or without b-lactamase-producing anaerobic bacteria, is isolated from cultures of more than two-thirds of these infections. Anaerobic organisms isolated in peritonsillar,1 lateral pharyngeal, and retropharyngeal13,14 abscesses include anaerobic streptococci and Prevotella, Bacteroides, and Peptostreptococcus species. Fusobacterium necrophorum is uniquely characteristic of infections that cause septic thrombophlebitis of the jugular vein, often with metastatic abscesses to the lungs (Lemierre disease, or postanginal sepsis).15 Rarely, nontuberculous Mycobacterium species are recovered from retropharyngeal and lateral pharyngeal abscesses.16 Adequate clinical specimens must be obtained from deep pharyngeal infections for culture, and the specimens must be transported properly and handled promptly. Surface culture specimens, with the possible exceptions of specimens from patients with peritonsillar cellulitis or abscess and acute epiglottitis, are often inadequate because they can be contaminated by oropharyngeal flora. These specimens are inappropriate for isolation of anaerobic bacteria. Specimens should be collected at the time of surgical drainage of a mature abscess. Of the four supraglottic infections, only acute epiglottitis carries a high probability of concomitant bacteremia. In areas that have greater than 20% community-acquired (CA) MRSA, antimicrobial choices, usually administered intravenously, include clindamycin, vancomycin, and linezolid, individually, or in combination. In some areas, up to 30% of MRSA also are resistant to clindamycin. Where S. aureus remains susceptible to methicillin, a semisynthetic penicillin plus metronidazole, cefoxitin, the fixed combination of a penicillin or aminopenicillin and a b-lactamase inhibitor (i.e., ticarcillin- or ampicillin-clavulanate), or clindamycin can be prescribed. These antibiotics are effective against MSSA, Streptococcus pyogenes, and anaerobic bacteria. If given at cellulitis or early phlegmon stage of the infection, antimicrobial therapy alone can often halt progression to a mature abscess. Documentation of success includes resolution of signs of toxicity and localized findings, a fall in the serum C-reactive protein (CRP) level or sedimentation rate, and regression of findings on computed tomographic (CT) scan. However, when a mature abscess is clinically probable or is seen on contrast-enhanced CT scan (i.e., complete circumferential rim enhancement with or without scalloping of the rim), surgical drainage should be performed urgently.

PERITONSILLAR ABSCESS (QUINSY) Peritonsillar cellulitis or abscess is the most common deep oropharyngeal infection. Although it generally occurs in preadolescents, adolescents, and young adults, peritonsillar abscess can occur at any age; it is more difficult to manage as an outpatient in young children. Occasionally, it is a complication of mononucleosis caused by Epstein–Barr virus, but it more frequently follows streptococcal and viral infections. The infection penetrates the tonsillar capsule by means of the lymphatic vessels and enters the space between the faucial pillars and

the tonsillar capsule. Inflammation progresses from peritonsillar cellulitis to phlegmon and abscess. The most common location is at the superior pole of the tonsil. Ten percent of cases of peritonsillar abscess are bilateral; diagnosis of these cases is particularly difficult.1–3 Recurrence of peritonsillar abscess is estimated at 10%, but it is > 10% if there is a history of recurrent acute tonsillitis in the previous 12 months.

Clinical Manifestations Preadolescents or adolescents have increasingly severe odynophagia and complain of the “worst sore throat” they ever had. Muffling of the voice occurs and difficulty in swallowing progresses, first with solids, then liquids, and finally, saliva. Severe pain interferes with eating, drinking, and sleeping. Occasionally, the patient presents at the early stage of peritonsillar cellulitis, appearing deceptively well, often afebrile, with nonspecific erythema of the posterior pharynx and tonsillar surface. Tonsillar exudate and submandibular lymphadenopathy may be present. In patients with more advanced disease, classic diagnostic features are: (1) trismus, inability to open the mouth fully because of spasm of the affected pterygoid muscle; (2) fullness or frank bulging of the superior pole of the palatine tonsil; and (3) deviation of the uvula toward the contralateral tonsil. The last is least likely to be present. The patient has a notably muffled voice, halitosis, pooling of saliva in the floor of the mouth, and expresses severe pain, yet may still not appear to have a severe infection. Gentle palpation of the tonsillar fullness with a gloved finger may reveal fluctuance. Ipsilateral cervical lymph nodes are usually enlarged and tender.

Treatment Prompt aspiration or incision and drainage of a mature peritonsillar abscess is required and must be performed in a controlled environment by a skilled and knowledgeable physician, with adequate suction apparatus. Sometimes, aspiration can be performed under conscious sedation or with topical anesthetic spray in the otolaryngologist’s office or in the emergency medicine department.17,18 Only a scant amount of bloody pus (< 1 to 10 mL) may be found. Gram stain, culture, and susceptibility tests are performed. Clindamycin or ampicillin-sulbactam is usually given, initially intravenously. Antibiotic therapy alone, without drainage, may suffice for certain patients with peritonsillar cellulitis or phlegmon. Emergency tonsillectomy is sometimes performed in cases of peritonsillar abscess. Because of the increased risk of recurrence, patients with peritonsillar abscess and a history of recurrent tonsillitis should have tonsillectomy performed after the acute episode subsides.

Complications If the mature abscess is not treated by surgical drainage, it may leak slowly or burst suddenly after several days, leading to upper-airway obstruction, respiratory distress, or aspiration pneumonia. Infection can be invasive locally, leading to thrombophlebitis of the jugular vein, a dissecting deep-neck infection, mediastinitis, or osteomyelitis of cervical vertebrae.

RETROPHARYNGEAL ABSCESS The potential space between the posterior pharyngeal wall and prevertebral fascia contains two paramedial chains of lymph glands that disappear by puberty; therefore, retropharyngeal abscesses are most common in infancy and early childhood. These nodes filter lymph from the nasopharynx and paranasal sinuses. In young children, retropharyngeal cellulitis generally develops after a mild bacterial nasopharyngitis or pharyngitis. Cellulitis progresses to an organized phlegmon and then to a mature abscess. The incidence of retropharyngeal infections appears to be increasing, and the infection is now more common in children than is epiglottitis.16 In older children and adults, retropharyngeal infection

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Infections Related to the Upper and Middle Airways

occurs more frequently after penetrating injury of the posterior pharynx; for example, a school-aged child may have fallen while holding a sharp pencil or stick in the mouth or have had a fishbone stuck in the pharynx.19 Retropharyngeal infection can also occur as the ventral extension of vertebral osteomyelitis,20 as a complication of traumatic endoscopy, a complication of a dental abscess or dental procedure, or after other medical or surgical trauma.

Clinical Manifestations and Diagnosis A preschool-age child with fever, with or without acute cervical lymphadenitis who has neck stiffness, torticollis, retrocollis, or a neck mass, is suspect for having a retropharyngeal abscess.12 The infant or young child generally appears ill, with a moderate or high fever, sore throat, an aversion to fluid intake, odynophagia, a change in vocal quality (muffled), or a peculiar gurgling sound or stertor. Fever is absent in 16%.7 Progression can lead to stridor and drooling, signs similar to those of acute epiglottitis. However, because retropharyngeal infection progresses less rapidly, the signs are more subtle, and affected children can appear less ill (less toxic) than those with acute epiglottitis16 (see Chapter 23, Respiratory Tract Symptom

CHAPTER

30

215

Complexes, Tables 23-3 to 23-5). Often, there is tender swelling on one side of the child’s neck. Classically, the child maintains an abnormal head position, i.e., lateral flexion (torticollis), limitation of neck motion, and resistance to movement up or down, and strongly opposes performance of an intraoral examination. An intraoral finding of anterior bulging of the posterior pharyngeal wall, just lateral to the midline, is a classic sign, but this finding is present in less than 50% of infants and young children. Digital examination is not performed. Breathing difficulty can dominate the clinical picture (see Chapter 23, Respiratory Tract Symptom Complexes, Table 23-4). In a cooperative older child, a skilled examiner who is experienced in airway management may use gentle digital palpation with a gloved finger to detect posterior pharyngeal fluctuation. The patient should optimally be examined in the head-down (Trendelenburg) position. There must be immediate access to large-caliber bedside suction equipment in the event that the abscess ruptures. A lateral radiograph of the nasopharynx and neck in complete hyperextension may identify the retropharyngeal mass (Figure 30-1A). On the plain roentgenogram, the normal diameter of the retropharyngeal space is 3 to 6 mm (less than that of one vertebral body), measured from the most anterior aspect of C2 to the soft tissues

A

C Figure 30-1. A, Lateral neck radiograph of an 18-month-old toddler with retropharyngeal abscess from Staphylococcus aureus infection. Note marked retropharyngeal soft-tissue density (arrow) with anterior displacement of the hypopharynx and laryngotracheal airway. Note the normal sharp appearance of the epiglottis, glottis, and subglottic airway. B, Chest radiograph. Note extension of infection into the mediastinum (arrow). C, Computed tomography scan of upper cervical region without injection of contrast material. Note abscess in the retropharyngeal space (arrow) with anterior displacement and compression of the airway and lateral displacement of the great vessels. Bony structures are the mandible (top), hyoid bone, and the cervical vertebrae.

B

216

SECTION

C Upper Respiratory Tract and Oral Infections

of the posterior pharyngeal wall. With the child’s head extended fully and after breathing in, retropharyngeal soft-tissue space more than one-half the width of C2 is abnormal. A retropharyngeal space of more than 7 mm, or a retrotracheal space of more than 14 mm in a child, also suggests a mass effect.7,21,22 The “mass” can be caused by a phlegmon, abscess, blood, or, rarely, a malignant neoplasm. The radiograph can also show an opaque foreign body or an air or fluid level in the mass, suggesting an anaerobic infection. Chest and pharyngeal roentgenograms are usually performed simultaneously to identify possible extension of the infection into the mediastinum or concurrent lung disease (Figure 30-1B). Contrast-enhanced CT is superior to plain radiograph to differentiate retropharyngeal cellulitis from mature abscess, to locate a foreign body, or to detect dissection of the infection into the mediastinum (Figure 30-1C).21–24 CT has important limitations, however. Compared with surgical findings, at least 10% to 15% of contrastenhanced CT scans result in false-positive or false-negative interpretations of a mature abscess.21–24 In one retrospective study, CT results for 68 children with retropharyngeal abscess had sensitivity of 43% and a specificity of 63%.9 In that study, all patients were treated with clindamycin without urgent surgery, even with CT evidence of frank abscess. Only 25% of the group underwent surgical drainage of the abscess.9 In another study of 80 children with retropharyngeal and lateral pharyngeal infections, the sensitivity, specificity, and positive and negative predictive value of CT in differentiating abscess versus cellulitis was 68%, 56%, 71%, and 53%, respectively.10 The differential diagnosis of retropharyngeal abscess includes other supraglottic infections, penetrating pharyngeal foreign body, cervical osteomyelitis, Pott disease (tuberculous abscess of the cervical spine), laryngopharyngeal diphtheria, angioedema of the epiglottis, caustic burns of the posterior pharynx, and tumors, such as lymphangioma (cystic hygroma) and hemangioma.

Management Suspected retropharyngeal infection requires urgent medical or surgical intervention, or both, and continuous monitoring. Management includes maintenance of the airway, insertion of a secure intravenous catheter, an order for nothing by mouth, and constant observation and documented serial examinations to detect the onset of airway compromise or increased pain or severity of illness. When the risk of airway obstruction is high, surgical drainage must be performed immediately. Appropriate antibiotic therapy (empiric clindamycin, vancomycin, or linezolid) is administered intravenously in all cases, and changed to more traditional therapy if a beta-lactamsusceptible pathogen is isolated. A suboptimal response to antibiotics without initial surgical drainage mandates careful clinical reassessment and repeated contrast-enhanced CT. Surgical drainage of a mature abscess is always necessary.25–27 Most mature retropharyngeal abscesses can be drained by intraoral incision under controlled conditions in the operating room. An external incision of a neck mass may be required when an abscess cavity is lateral to the sheath of the great vessels (i.e., a combined retropharyngeal and lateral pharyngeal abscess). The need for placement of an artificial airway is case-specific.

Complications Complications of retropharyngeal abscess can have a life-threatening mass effect on the airway. Abscess can rupture, leading to aspiration pneumonia or asphyxiation. There can be lateral spread of infection to the carotid sheath or caudal spread to the mediastinum.6,8–10,19 Death can occur from aspiration, airway obstruction, and erosion into the jugular vein causing septic thrombophlebitis (Lemierre disease),15 or extension to the mediastinum.

LATERAL PHARYNGEAL (PARAPHARYNGEAL) ABSCESS The lateral pharyngeal (pharyngomaxillary) space is divided into two compartments by the styloid process of the mastoid bone. The anterior portion is close to the tonsillar fossa medially and to the internal pterygoid muscle laterally. The posterior compartment contains the carotid sheath. Involvement of these structures determines the clinical manifestations and complications of infections in these spaces.20 Lateral pharyngeal abscess is the third most common supraglottic infection, after peritonsillar and retropharyngeal infections. Infection of the compartments of the lateral pharyngeal space may be the result of extension of suppurative local tonsillopharyngitis, retropharyngeal abscess, mastoiditis (Bezold abscess), parotitis, or dental abscess.

Clinical Manifestations and Diagnosis The hallmarks of infection in the anterior compartment are tender swelling below the angle of the mandible, induration and erythema of the side of the neck, and trismus. Most patients are febrile, appear to be acutely ill, and complain of severe odynophagia and neck pain. A bulge in the lateral pharyngeal wall can be observed, but the tonsil is normal in size and relatively uninflamed, thus distinguishing this infection from a peritonsillar abscess. Torticollis, with head tilted toward the side of involvement, is often present, as is cervical lymphadenitis. The classic triad that indicates pharyngomaxillary abscess, an anterior compartment syndrome, includes: (1) tonsillar and tonsillar fossa prolapse; (2) trismus; and (3) swelling of the parotid area or lateral neck, or both. Infection in the posterior compartment is ill defined clinically, and characterized by signs of septicemia, with minimal pain or trismus. Swelling can often be overlooked when it is deep to the palatopharyngeal arch. A tender, high cervical mass, with ill-defined boundaries initially, can be palpated during the phlegmon or frank abscess stages. CT or magnetic resonance imaging (MRI) is necessary for diagnosis and management, as they can delineate involvement of critical fascial planes, vital structures, and complications.

Management Treatment of the posterior compartment of a pharyngeal abscess requires drainage of the lateral neck in conjunction with high dosages of appropriate antimicrobial therapy, administered intravenously. An external excision below the angle of the jaw is preferred because it provides access to the carotid artery, which should be ligated in cases of arterial erosion. Surgical drainage is best performed after localization of infection and during the course of intravenous antibiotic therapy (using agents as for retropharyngeal infection) unless hemorrhage or respiratory obstruction necessitates earlier intervention. The progress of disease must be closely monitored; establishment of an artificial airway may be required pre-emptively because airway obstruction can develop abruptly.

Complications Complications occur frequently, especially from infection in the hidden posterior compartment of the lateral pharyngeal space, and may include respiratory embarrassment, laryngeal edema, airway obstruction, septicemia, pneumonia, septic thrombosis of the internal jugular vein (Lemierre disease), suppurative intracranial complications (meningitis, brain abscess, and thrombosis of the cavernous and lateral sinus), and erosion of the carotid artery. Erosion into the carotid artery sheath can cause life-threatening hemorrhage or thrombosis. Extension of infection inferiorly along the carotid sheath or posteriorly into the retropharyngeal space can lead to mediastinitis.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Infections Related to the Upper and Middle Airways

LEMIERRE DISEASE Classically, Lemierre disease has four findings: (1) necrotizing invasive pharyngitis or tonsillitis; followed by (2) Fusobacterium necrophorum bloodstream infection (postanginal sepsis); leading to (3) internal jugular vein thrombophlebitis; and subsequent (4) metastatic abscesses, usually in the lungs, but occasionally elsewhere. Atypical Lemierre disease or Lemierre-like disease begins with suppurative acute otitis media instead of necrotizing pharyngitis, bloodstream infection with non-Fusobacterium organisms, thrombophlebitis of the fascial vein or dural sinus thrombosis instead of the internal jugular vein, and abscess in brain, heart, or liver instead of the lungs. Following the original report by Courmont & Cade in 1900,28 Lemierre reported 20 defining cases in 1936.29 A review of 40 cases of this syndrome was published in 1995.30 Although F. necrophorum is causative in 90% of cases (because of its propensity to cause infectious phlebitis), Lemierre disease has also been reported with other bacteria, including Staphylococcus aureus and Streptococcus pyogenes.31,32 Previously rare in pediatrics, cases are increasingly reported in the past decade.33–35 Lemierre disease affects previously healthy adolescents and young adults, similar in age to those affected by peritonsillar abscess. Although potentially life-threatening, fatality of < 10% has recently been reported.

Clinical Manifestations This is a difficult disease to diagnose and requires a high degree of clinical suspicion. Children and adolescents typically complain of increasingly severe sore throat (acute tonsillopharyngitis) finally interfering with swallowing liquids, persistent fever, and possibly chills. Some cases begin with acute otitis media.36 With progression, there is localized lateral neck pain and tender swelling along the course of the internal jugular vein. If undiagnosed, the next stage is “metastasis” of septic thrombi to the lungs, intra-abdominal organs, brain, or heart. Pulmonary involvement is characterized by putrid productive cough and chest pain. Uniquely, girls may complain of abdominal pain and have icterus and hepatomegaly on physical examination, indicating occult hepatic abscesses.

Diagnosis After obtaining cultures of the blood, complete blood count, and tests for acute-phase reactants, the key to the diagnosis of Lemierre disease is contrast-enhanced CT of the neck and chest, and Doppler ultrasonography or MR venography or both of the internal jugular vein. Contrast-enhanced CT shows a dilated internal jugular vein with low attenuation of intraluminal contents and enhancement of the vessel wall. Follow-up contrast-enhanced CT may be necessary if the disease worsens or is protracted during treatment. Evaluation of an underlying hypercoaguable state should be investigated as soon as the diagnosis of septic thrombophlebitis is made.

Management Antibiotics parenterally are the primary therapy for Lemierre disease. Because some Fusobacterium species produce beta-lactamase enzymes, treatment should include antibiotics with activity against oral anaerobes that are resistant to beta-lactam antibiotics. Such choices include ampicillin-sulbactam, clindamycin plus ceftriaxone with or without metronidazole. Therapy is continued intravenously for 4 to 6 weeks. Vigilance must be directed to monitoring for propagation of the septic thrombus or for worsening of pulmonary function. Some authors recommend heparin anticoagulation for Lemierre disease.33,37,38 In a series of 9 cases of Lemierre disease published in 2005, thrombophilia was documented in all 7 children tested. The

CHAPTER

30

217

prothrombotic state consisted principally of antiphospholipid antibodies and elevated factor VIII activity. Anticoagulation was given for a median duration of 3 months. After a median follow-up of 1 year, all children had survived without recurrent venous thrombophlebitis. However, vascular occlusion of the internal jugular vein, documented in the acute phase of illness, persisted in 38% of the anticoagulated children.

ACUTE EPIGLOTTITIS Acute epiglottitis (supraglottitis) is a locally invasive bacterial infection of the supraglottic area – the epiglottis, arytenoid cartilages, aryepiglottic folds, and false vocal cords (ventricular bands).39–41 The oropharyngeal structures, true vocal cords, and trachea are rarely, if ever, involved. A report on epiglottitis cases in New South Wales, Australia, before and after the introduction of vaccination against Haemophilus influenzae type b disease, concluded that in young children, infection is characterized by rapid (within hours), relentless progression from severe sore throat to severe respiratory obstruction. In the past, acute epiglottitis was often accompanied by bacteremia and in unimmunized children this should be anticipated. Epiglottitis is generally caused by infection with virulent, often encapsulated bacteria, which invokes an inflammatory response that causes rapid development of obstructing edema and spreads quickly through mucous membranes, in a consecutive, linear manner, from the epiglottis to the false vocal cords. Obstructive signs predominate clinically and the disease rarely progresses to phlegmon or abscess stages. Epiglottitis is now rare following universal immunization against H. influenzae type b. The disease now occurs predominantly in adults: it is rarely recovered from blood or epiglottis cultures.41 The differential diagnosis includes angioneuropathic edema of the supraglottic structures, secondary to an allergic reaction or C1-esterase deficiency and caustic ingestion, or thermal burns of the epiglottis.42

Etiology and Epidemiology Respiratory viral infections of the supraglottic areas commonly precede superinfection. Currently, Streptococcus pneumoniae, group A or C streptococci,43,44 and, less commonly, Staphylococcus aureus and Haemophilus species account for most cases of acute epiglottitis in children. Acute streptococcal epiglottitis has been reported as a complication during varicella infection in a 9-year-old.44 The disease in adults differs somewhat in etiology and pathogenesis. The progression in adults is often slower than in children, similar to that of peritonsillitis, which manifests as an increasingly severe sore throat and muffled voice. Dyspnea is not usually present. The age of children with epiglottitis has changed since introduction of H. influenzae type b conjugate vaccine, from the median age of 34 months (range of 2 to 6 years) to 6 to 7 years.40

Clinical Manifestations and Diagnosis In the young child, acute epiglottitis is usually a fulminant illness characterized by abrupt onset of high fever, illness, and tachycardia out of proportion to the fever’s severity. The child is often flushed, apprehensive, worried, and subdued. Although stertor and a gurgling noise on inspiration are present, cough is infrequent or absent because motion of the intensely inflamed epiglottis intensifies the pain. The voice is muffled (“hot-potato voice”) but not hoarse. Aphonia, drooling, and characteristic posture are seen as infection progresses. The child sits upright, hands extended behind the body in a tripod position, with the jaw thrust forward, mouth open, and the tongue protruding. Saliva collects in puddles in the sublingual gutters with continuous overflow drooling. Progression after this point is likely to be abrupt, without additional warning, and manifests as pallor, fluctuating levels of consciousness, central cyanosis, and respiratory arrest. Acute epiglottitis can occasionally be confirmed by judicious passive visualization of the twice-normal-sized “cherry-red” epiglottis

218

SECTION

C Upper Respiratory Tract and Oral Infections

protruding above the base of the tongue. Intraoral inspection should only be performed if the child willingly opens his or her mouth widely. Use of a tongue depressor is contraindicated because it may cause precipitous closure of the airway by means of laryngospasm or pressure against the massively enlarged epiglottis or arytenoid cartilage.45 Oral inspection often does not reveal a swollen epiglottis.46

Management If the diagnosis of acute epiglottitis is highly likely, it should be confirmed and an artificial airway should promptly be placed in the operating room under controlled conditions, with a skilled anesthesiologist and otolaryngologist present to manage the airway. On infrequent occasions, the epiglottis itself is only minimally inflamed and the disease is localized to the other supraglottic structures, such as the aryepiglottic folds. This variant of supraglottitis is difficult to detect without CT or controlled endoscopy. In earlier stages of the disease, before the development of marked toxicity, stertor, aphonia, or refusal to drink, a plain lateral view radiograph of the neck may be indicated. The study should be performed at the bedside or the child should be accompanied by the comforting parent and a physician who is skilled in airway management and who carries a properly maintained laryngoscope, endotracheal tubes, ventilation bag, and appropriately sized mask.34 In a true emergency situation, a 13- to 15-gauge needle can be inserted through the cricothyroid membrane. The child should be disturbed as little as possible and kept upright, and the film should be taken with the head in full extension. Diagnostic findings include swelling of the epiglottis (thumb sign), edema of the aryepiglottic folds, and ballooning of the pharyngeal airway, indicating inspiratory obstruction. The subglottic airway is usually normal (Figure 30-2). Once the airway is secured, antibiotic therapy is administered intravenously after a blood culture specimen is obtained. Ceftriaxone or cefotaxime intravenously is an appropriate choice; alternatives are ampicillin-sulbactam or cefuroxime. The airway must be secured and the child constantly observed to prevent accidental extubation.47 Extubation is usually performed within 72 hours. Duration and route of administration of antibiotic therapy depend on clinical course, causative pathogen, and presence of infection in the bloodstream or other sites.

LARYNGOTRACHEITIS AND LARYNGOTRACHEOBRONCHITIS (CROUP) The most common upper-airway infections are viral laryngotracheitis (LT) or laryngotracheobronchitis (LTB), also known as viral croup, and viral bronchitis.48–52 Serious subglottic infections include bacterial tracheitis and severe viral croup. Viral LT/LTB usually spreads downward from the nasopharynx into the larynx, trachea, and the smaller airways. The source of obstruction in viral croup is below the glottis and usually results from inflammatory edema and mucus production of the subglottic tracheal submucosa, without direct viral invasion. Exfoliation of the damaged mucosal lining of the trachea can intensify the obstruction. The level of glottic and subglottic infection – namely, the larynx, trachea, or bronchi – determines and is predicted by the clinical features of the disease, particularly the quality of the voice or cry and the frequency and quality of cough.

Epidemiology Croup is the most common infectious cause of obstruction of the upper airway in young children. It occurs most often in children between the ages of 6 months and 2 years, and more often in boys than in girls.48–52 Croup is traditionally divided into spasmodic and infectious (viral) types; however, this distinction is not necessary for practical purposes, since both are managed in the same way. Epidemics begin in October or November and peak in early winter; in some years, there is a

Figure 30-2. Lateral neck radiograph of a 4-year-old child with acute epiglottitis. Note characteristically distended hypopharynx and “thumbprint” edematous epiglottis and aryepiglottic folds (arrow).

smaller outbreak in mid-August. The virus is transmitted by contact with unwashed hands or sneezing, coughing, or breathing virus-laden microdroplets. The incubation period is 2 to 3 days, and the typical course of croup is about 3 days, with a range of 1 to 6 days.

Etiologic Agents Parainfluenza virus types 1, 2, and 3 are the most common causes of croup; together, they account for about 60% of cases.53 Type 1 virus, the most common, characteristically causes fall or winter epidemics; type 3 virus is less common but can cause more severe illness. Respiratory syncytial virus, influenza virus A and B, human adenoviruses that affect the respiratory tract (types 1, 2, 3, 4, 7, 8, 11, 14, and 21), human metapneumovirus, certain echoviruses, coxsackieviruses, and other viruses can also cause croup, and the inflammatory changes and clinical symptoms are similar to those of the prototypical parainfluenza-associated croup. Measles is sometimes associated with severe croup.54 The most common causes of outbreaks of croup in closed communities are influenza virus and adenovirus infections.

Clinical Manifestations Croup is heralded by symptoms characteristic of a nonspecific upper respiratory tract illness. These are followed by a seal-like barking or brassy cough, dysphonia that may progress to aphonia, intermittent

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Infections Related to the Upper and Middle Airways

inspiratory stridor, and suprasternal retractions. The onset is usually unexpected and dramatic and can become manifest after the child has been sleeping. NonspeciÀc antecedent upper respiratory tract symptoms are not always present; in this case, the condition is called spasmodic croup. Ninety percent of children with croup have mild symptoms. Some children develop signiÀcant symptoms during daylight hours, with or without a high fever. They are often evaluated by primary care physicians, who may prescribe antibiotic agents injudiciously.55,56 Moderate or severe croup can lead to airway occlusion and is potentially life-threatening. Stridor and barking cough are the predominant symptoms (see Chapter 23, Respiratory Tract Symptom Complexes). The leukocyte count is generally normal or mildly elevated, with lymphocytes predominating. Radiographs demonstrate the classic “steeple sign” of subglottic edema, with a 5- to 10-mm segment of narrowed subglottic air column. The lateral radiographic view of the neck can show widening of the hypopharyngeal air space resulting from a distal obstruction (Figure 30-3).

Differential Diagnosis Unusual infectious causes of croup syndrome include Corynebacterium diphtheriae,57 Mycobacterium tuberculosis, endotracheal infection and Candida albicans secondary to overuse of steroid inhalations. Laryngeal diphtheria can present as severe croup. In an unimmunized child, a degree of toxicity and tachycardia that is out of proportion to fever may help focus the clinician’s attention on this rare disease. The history can help differentiate true LTB from other causes of a croupy cough. A history of recent aspiration or choking on a foreign body, the presence of croup in the Àrst 90 days of life, or more than two or three recurrences of croup in a year suggest that the obstruction may have an anatomic cause. Infectious and noninfectious causes and clinical features of upper-airway obstruction, including croup, are delineated in Chapter 23, Respiratory Tract Symptom Complexes, and highlighted in Tables 23-3 to 23-5. Noninfectious causes of recurrent croup include foreign-body aspiration, tracheomalacia, gastroesophageal reflux,58 paradoxical vocal cord dysfunction,59 and subglottic

CHAPTER

30

219

stenosis from prolonged tracheal intubation in the neonatal period. Anatomic causes of airway compromise include a vascular ring. Infrequent causes of a croupy cough and stridor include laryngeal papillomatosis, laryngeal webs or cysts, vascular ring, allergic or hypocalcemic laryngospasm, H-type tracheoesophageal Àstula, and laryngeal trauma.60,61 CT, MRI, or MR angiography can help differentiate noninfectious from infectious causes of obstruction.60,61 Croup is not expected as a complication of viral infection in children older than age 5 because of the greater diameter of the lumen of the trachea and because previous infections with parainfluenza virus and other typical agents of croup have led to immunity. Another diagnosis or predisposition must be sought, particularly aspiration of a foreign body. Older children and adolescents with parainfluenza laryngotracheitis usually have symptoms of simple laryngitis, i.e., dysphonia, without a barking cough or stridor.

Management Maintenance of an adequate airway, humidiÀcation of room air, and oxygenation are tenets of management.62,63 Six percent of young children with viral croup are hospitalized for severe obstructive symptoms, hypoxia, dehydration, or fatigue.64,65 Endotracheal intubation is necessary for less than 1% of such patients. The condition is rarely fatal. Recovery from LTB occurs in 3 to 7 days. The most common complication of LTB is extension of infection to other regions of the respiratory tract, such as the middle ear, the bronchioles, and the parenchyma of the lung (bronchopneumonia). Recurrences can be expected in about one-third of children. Bacterial tracheitis sometimes complicates LT/LTB and can be life-threatening (see below).

General Care and Management of Airway Decisions concerning where the child with croup will be managed depend on the degree of parental or physician uneasiness with the clinical, geographic, and social setting; the course of previous episodes; the rapidity and degree of progression of upper-airway obstruction and respiratory distress (i.e., croup score65); progression of weakness and fatigue; and symptoms or signs of moderate-to-severe dehydration or poor intake of fluids. As long as a responsible caretaker remains with the child at all times, a telephone is in the home, immediate transportation is available, and an emergency medicine facility is close by, most children with mild or moderate croup do not require hospitalization. All children, however, require periodic reassessment by a caretaker during the hyperacute phase of the disease. HumidiÀcation of air may reduce edema of the airway and facilitate clearing of secretions. The relative therapeutic merit of hot steam over cold mist vaporizer is not supported by evidence. Exposure to a steam-Àlled bathroom or to night air (via an open window or a ride in an automobile with the windows down) is traditional advice of unproven merit. Ipecac syrup, once a popular treatment, has no role in modern treatment.

Racemic Epinephrine

Figure 30-3. Lateral neck radiograph of an 18-month-old toddler with acute laryngotracheobronchitis caused by parainfluenza virus 1 showing characteristic subglottic edema. Note normal sharp appearance of the epiglottis and distinct laryngeal ventricle (airspace between the false and true vocal cords) (arrows).

Nebulized racemic epinephrine can be useful in the treatment of moderate or severe cases of LTB.66,67 The adrenergic effects of racemic epinephrine induce vasoconstriction, which decreases subglottic edema.67 Racemic epinephrine 2.25% diluted 1:8 in saline is administered via nebulizer at a dose of 0.25 mL (4 drops) for children younger than 6 months and 0.5 mL for older children. The onset of drug action occurs in less than 10 minutes, and the effects last 60 to 90 minutes. The dose can be repeated in 1 hour. Continuous nebulization over 30 to 60 minutes may also be effective. Close observation in the emergency department or ofÀce is standard practice after this treatment to reassess in the period after waning of drug effect. Discharge is appropriate when there is minimal stridor at rest, and air entry, color, and level of consciousness are normal. If the child’s upper respiratory distress worsens following

220

SECTION

C Upper Respiratory Tract and Oral Infections

discharge, the patient should be immediately transported to a medical facility for re-evaluation.

Corticosteroid Therapy Corticosteroid treatment reduces symptoms in children with mild, moderate, or severe croup.68–71 In double-blind, placebo-controlled studies, children who receive corticosteroids have significant reductions in duration of severe symptoms, shorter observation times in an emergency medical facility, less need for return visits, less need for hospitalization and critical care observation when hospitalized, and less need for endotracheal intubation. A Cochrane review of benefit highlighted the efficacy of corticosteroids in the management of croup.71 Dosage options are dexamethasone phosphate (0.6 to 1 mg/ kg) once intramuscularly; dexamethasone syrup or crushed tablets (0.6 to 1 mg/kg) once orally; or prednisolone (2 mg/kg per day) orally in two or three divided doses.72 There is no significant difference in efficacy between single dexamethasone doses administered orally or intramuscularly.73 No adverse effect from single-dose dexamethasone has been noted. It seems prudent to administer dexamethasone to any child with croup whose airway obstruction is severe enough to warrant treatment with racemic epinephrine. Nebulized budesonide is nearly as effective as dexamethasone but is more expensive, takes longer to administer, and is less available.74,75 The addition of inhaled budesonide to oral dexamethasone offers no advantage in the treatment of children hospitalized with croup.76 Supplemental oxygen is often prescribed for hospitalized children with LT/LTB, particularly when the resting pulse oximetry reading in room air is less than 92%. The hypoxic child must be carefully monitored for impending respiratory failure. A small study has documented the benefits of heliox for children with LTB.77

ACUTE LARYNGITIS AND BRONCHITIS Clinical Manifestations and Etiology Acute laryngitis is an infection of older adolescents and adults. It is usually preceded by a simple upper respiratory infection that is caused by the agents that cause croup in younger children. Laryngitis is characterized by a sore throat, hoarse voice, and harsh cough. Specific additional symptoms depend on causative agents. Acute laryngitis caused by infection with influenza virus or adenovirus is sometimes accompanied by fever, headache, and malaise. Parainfluenza virus or respiratory syncytial virus tends to produce milder extralaryngeal symptoms. Acute bronchitis is also an infection of older adolescents and adults. The onset can be gradual or abrupt, with nasal or nasopharyngeal symptoms predominating in the first few days of illness. Cough is the major symptom. It begins as a dry, harsh, and occasionally brassy sound; 4 to 6 days later, the cough becomes loose and productive. It may last several weeks. Rhonchi and coarse rales can be heard on auscultation of the chest. Fever is variable. In addition to the viral agents of the LT/LTB spectrum, acute bronchitis can be caused by Mycoplasma pneumoniae, Chlamydophila pneumoniae, Bordetella pertussis, B. parapertussis, and human metapneumovirus. The clinical course and causes of acute bronchitis are different in patients with underlying chronic bronchopulmonary diseases, such as chronic, persistent bronchial asthma; cystic fibrosis; bronchiectasis; and chronic lung disease of prematurity, than they are in patients without these underlying conditions. Bacterial superinfection by organisms colonizing the upper respiratory tract is more likely in these patients. Smoking is a risk factor for recurrent acute bronchitis of uncertain cause.

Management Treatment of laryngitis and bronchitis is symptomatic and palliative. Antibiotics are not indicated in patients without an underlying

disease.78 The efficacy of expectorants and antihistamines is not proven. Antiviral therapy is useful for influenza infection if given early in the course of the disease. In an adolescent bronchitis presumed to be due to Mycoplasma pneumoniae or Chlamydophila pneumoniae can be treated with a macrolide antibiotic or doxycycline but therapeutic benefit is not proven for bronchitis (see Chapter 166, Chlamydophila (Chlamydia) pneumoniae, and Chapter 196, Mycoplasma pneumoniae). The diagnosis of pertussis should be considered, and the patient treated with a macrolide agent if pertussis is likely (see Chapter 162, Bordetella pertussis (Pertussis) and Other Species).

BACTERIAL TRACHEITIS Pathophysiology First described in 1979, bacterial tracheitis is distinct from viral croup and acute epiglottitis.79 It can be difficult to differentiate from severe viral LT and LTB, especially when caused by influenza viruses or adenovirus.64 Bacterial tracheitis is currently more prevalent than acute epiglottitis. The major site of disease in bacterial tracheitis is the cricoid cartilage, the narrowest part of the trachea. Marked subglottic edema and copious purulent secretions in the trachea and upper bronchi are typical, but there is no inflammation of the supraglottic structures. The hallmark of the disease is the development of a pseudomembrane composed of mucosal lining, bacteria, and inflammatory products that causes severe, even fatal, obstruction of the airway.80,81 Bacterial tracheitis can also present hyperacutely without a prodromal phase. In this presentation, it mimics acute epiglottitis or retropharyngeal abscess. Imaging of the upper airway or bronchoscopy may be necessary to establish the diagnosis. Toxic shock can occur if exotoxin-producing strains of Staphylococcus aureus or Streptococcus pyogenes cause bacterial tracheitis.82 Bacterial tracheitis occurs at any age and in any season. Debilitated individuals with neuromuscular disorders and children with an endotracheal tube in place are at risk for bacterial tracheitis. Most often it is caused by a gram-positive bacteria superinfecting influenza or parainfluenza (or measles) virus infection of the upper respiratory tract. The most common causative organisms are Staphylococcus aureus, Streptococcus pyogenes, Moraxella catarrhalis, Haemophilus influenzae, and S. pneumoniae.81 Infrequent or rare causes of bacterial tracheitis include respiratory anaerobic bacteria, Mycobacterium pneumoniae, M. tuberculosis (endobronchial disease), and Corynebacterium diphtheriae (laryngeal diphtheria; see Chapter 130, Corynebacterium diphtheriae (Diphtheria)). Gram stain and culture of purulent secretions confirm the specific pathogen. Blood culture is positive in less than 50% of cases.

Clinical Manifestations and Diagnosis The patient appears ill, agitated, or subdued; the voice is not muffled.79 Fever can be high or low. Coughing is frequent and not particularly painful; in the patient with acute epiglottitis, by contrast, coughing is infrequent but painful. Signs of severe stridor and obstruction to air exchange predominate, in contrast to the hoarseness and croupy cough typical of viral LTB. Symptoms progress and peak, sometimes to complete obstruction, within minutes to hours. The child may become anxious, pale, or cyanotic. Administration of nebulized epinephrine or humidified vaporized room air provides minimal or no relief. In a 10-year series of 94 cases of bacterial tracheitis ending in 2001,83 the mean age was 8 years; only 40% of the children were febrile on presentation. The mean leukocyte count was 10 800 cells/mm3. Staphylococcus aureus was the most commonly isolated bacteria, although recovery of M. catarrhalis was associated with a higher rate of endotracheal intubation. Fifty-three percent of children required intubation, which is considerably lower than that reported

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Otitis Media

previously. Nine of the children were described as toxic, and six presented in respiratory extremis. There were no deaths. On the basis of this report, one may conclude that a signiÀcant subset of older children with bacterial tracheitis do not have severe clinical symptoms. These patients still require aggressive medical management, vigilance, and debridement suctioning of desquamated tracheal mucosa. A radiograph may show a narrowed tracheal airspace. The presence of ragged border, especially a partially adherent pseudomembrane or mass (inflammatory debris and exfoliation of necrotic tracheal mucosa), is highly suggestive of bacterial tracheitis (Figure 30-4).

Management Suspected bacterial tracheitis is a medical emergency. Vital signs, level of consciousness and distress, and blood oxygen saturation must be monitored constantly. Endotracheal intubation is frequently required. Tracheotomy may be necessary to help with frequent tracheal suctioning. Rigid bronchoscopy with a ventilation port for

CHAPTER

31

221

oxygen administration, performed urgently in the operating room, is required to establish the diagnosis82 and to secure the airway. Aspiration of tracheal secretions is performed for Gram stain, bacterial culture, and virus identiÀcation. Suctioning of necrotic debris and inspissated secretions during bronchoscopy can be life-saving. Antimicrobial agents that are effective against S. aureus and streptococci, administered intravenously, are required. If Gram stain of tracheal secretions reveals neutrophils and gram-positive cocci only, clindamycin or vancomycin or combination therapy should be prescribed. When culture of tracheal secretions reveals the pathogen and susceptibility tests are performed, a speciÀc therapy can be selected. Frequent suctioning is necessary to prevent sudden obstruction of the endotracheal tube. Vital signs, pulse oximetry, degree of obstruction, and alertness must be monitored continuously, even after intubation.80 Acute obstruction due to pathology distal to the endotracheal tube can be fatal (see Chapter 23, Respiratory Tract Symptom Complexes). Extubation can be accomplished when mucosal edema and purulence decrease and the infection responds to the intravenous antibiotic agent, usually within 5 days. Complications of bacterial tracheitis include atelectasis, pulmonary edema, septicemia, and respiratory failure. Antibiotic therapy is usually continued for at least 10 days. Oral administration is appropriate after the condition improves and when an endotracheal tube or oxygen supplementation is no longer required.

CHAPTER

31

Otitis Media Stephen I. Pelton A

Otitis media is a disease of early childhood. Its relevance to child health has evolved from an association with suppurative complications, such as mastoiditis and brain abscess, to the current concern of prolonged conductive hearing loss as well as language or cognitive delays. Furthermore, the current frequency of diagnosis and treatment of acute otitis media (AOM) results in more than 20 million antibiotic prescriptions annually, as well as signiÀcant family disruption.1 Recent trends in the management of AOM have been influenced by the selective pressure from universal administration of pneumococcal conjugate vaccine on colonization with otopathogens and concern over the widespread use of antibiotics for treatment of AOM that has contributed to the emergence of antibiotic resistance resulting in a reevaluation of the criteria for diagnosis and treatment of this disorder.

ACUTE OTITIS MEDIA Pathogenesis

B Figure 30-4. A, Lateral neck radiograph of a 22-month-old boy with bacterial tracheitis caused by Staphylococcus aureus showing subglottic haziness (similar to laryngotracheobronchitis, or croup). B, Endoscopic view of trachea shows mucosal denudation, intraluminal debris, and purulent laryngotracheal secretions.

Important clues to the physiologic changes that result in AOM are provided by clinical observations, such as: (1) a seasonal peak during winter months that parallels the incidence of upper respiratory tract infection (URI);2 (2) an attack rate that is highest in children younger than 2 years and greatest in group childcare attendees;3,4 (3) nearuniversal occurrence in children with cleft palate;5 and (4) higher frequency in populations with immune dysfunction, such as those with human immunodeÀciency virus (HIV),6 immunoglobulin (Ig) G or IgG subclass deÀciency.7 AOM is generally considered a bacterial infection because bacterial otopathogens are isolated from middle-ear culture in 70% of

222

SECTION

C Upper Respiratory Tract and Oral Infections

cases,8 but frequently it is a coinfection in which a viral URI enhances the ability of bacterial otopathogens to ascend from the nasopharynx to the middle ear. Respiratory tract viruses alone can, on occasion, elicit signs and symptoms of AOM and are recovered from a small proportion of tympanocenteses (2% to 20%).9 Eustachian tube dysfunction (as indicated by the presence of negative pressure with tympanometry) has been demonstrated to occur in 75% of children with viral URI and is a major contributing factor to the development of bacterial AOM.10 Experimental intranasal challenge in adults with influenza A virus or rhinovirus has conÀrmed that viral URI frequently results in negative middle-ear pressure.11,12 Additionally, influenza A infection, in both humans and chinchillas, suppresses neutrophil function,13 which may contribute to the high attack rate for AOM observed in children with influenza A infection.14 Viral URI in animal models and as observed in children enhances the frequency and density of nasopharyngeal colonization in children with otopathogens.15–17 Proposed mechanisms are viral effect to increase expression of receptors for bacteria and to alter mucociliary barrier and clearance mechanisms.18,19 Figure 31-1 represents these important pathophysiologic mechanisms.

dies of symptomatic middle-ear disease. Jacobs and associates8 reported on otopathogens recovered from 917 episodes of AOM in Eastern European, Israeli, and United States children. Pneumococcus TABLE 31-1. Virus Detected Alone or in Combination with Bacterial Pathogens from Middle-Ear Fluid in 128 Children with Acute Otitis Media Virus

No. of Children

%

Respiratory syncytial

72

49.0

Parainfluenza (types 1, 2, and 3)

20

13.6

Influenza (A and B)

19

12.9

Enterovirus

10

6.8

Rhinovirus

10

6.8

Cytomegalovirus

8

5.0

Adenovirus

7

4.8

Herpes simplex

1

0.7

Total

A limited number of viral and bacterial pathogens are recovered from culture of the middle ear. Table 31-1 and Figure 31-2 provide data on the relative frequency of recovery of viral9 and bacterial pathogens. Four bacterial otopathogens – Streptococcus pneumoniae (Sp), nontypable Haemophilus influenzae (NTHi), Moraxella catarrhalis, and Streptococcus pyogenes (group A streptococcus: GAS) – are consistently recovered from middle-ear cultures. Although the proportion of disease due to each pathogen varies with geography and age, remarkable consistency has been observed in microbiologic stu-

147

100

Adapted from Chonmaitree T. Viral and bacterial interaction in acute otitis media. Pediatr Infect Dis J 2000;19:S24–S30.

80% % of cases

Microbiology and Antimicrobial Susceptibility

60% 40% 20% 0% Eastern and Central Europe

Viral URI Viral vaccines

Antimicrobial prophylaxis Anti-adherence molecules Bacterial vaccines Enhanced frequency and density of bacterial otopathogens Decreased mucociliary clearance

United States

Region

A

Eustachian tube 80% Surface agent to improve ET function anti-inflammatory agents

% of cases

Nasopharyngeal colonization

Israel

Eustachian tube dysfunction leads to negative middle ear pressure Bacterial invasion of middle ear

60% 40% 20% 0% 3–11 (60)

12–35 (107)

36–59 (85)

>60 (29)

Total (281)

Age in months (no. of pts.) Host defenses

Specific IgG antibody resulting in lysis or enhanced phagocytosis

B Streptococcus pneumoniae

Haemophilus influenzae

Moraxella catarrhalis

Streptococcus pyogenes

Mixed (S. pneumoniae and H. influenzae)

Decreased PMN function Purulent AOM Figure 31-1. Schema of events in viral–bacterial pathogenesis of acute otitis media (AOM), with potential intervention strategies (shown in boxes). ET, endotracheal tube; Ig, immunoglobulin; PMNL, polymorphonuclear leukocyte; URI, upper respiratory tract infection.

Figure 31-2. (A) Microbiology of acute otitis media in the United States, Eastern and Central Europe, and Israel. (B) Age distribution of middle-ear pathogens in United States children. (Redrawn from Jacobs MR, Dagan R, Applebaum PC, et al. Prevalence of antimicrobialresistant pathogens in middle ear fluid. Multinational study of 917 children with acute otitis media. Antimicrob Agents Chemother 1998;42:589–595.)

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Otitis Media

was recovered more frequently from United States children younger than 5 years (~30%) than from those older than 5 years (~10%) (Figure 31-2A). Moraxella was most commonly found in United States children, and NTHi was more common among Israeli children. GAS was seen almost exclusively in children older than 5 years (Figure 31-2B). In the United States, where universal administration of 7-valent pneumococcal conjugate vaccine (PCV7) is given, the microbiology of AOM is shifting, with increasing importance of nonvaccine serotypes of S. pneumoniae (NV-Sp), and emergence of NTHi as the most common otopathogen in children failing initial therapy.20–22

Streptococcus pneumoniae In the United States prior to the introduction of PCV7, 70% of pneumococcal otitis media was due to seven serotypes (Figure 31-3).23 The capsular polysaccharides from these serotypes are, in general, poor immunogens in children.24 Two studies have noted that S. pneumoniae is a common cause of AOM in young infants (< 6 months of age).25,26 Two studies have demonstrated that nasopharyngeal colonization with S. pneumoniae can occur early in life and early acquisition is associated with increased risk of AOM.27,28 A shift in serotypes recovered from middle-ear cultures of children with AOM has been observed in studies performed after 2000, consistent with the initial Finnish report that increased disease due to NV-Sp was observed among vaccine recipients. McEllistrem et al.20 observed that disease due to V-Sp declined over the 4-year time period 2000 to 2004 and the emergence of NV-Sp was greatest in children who received 3 doses of PCV7. This is consistent with the fact that differences in capacity of Sp serotypes to produce AOM are small;29 NV-Sp is as likely as V-Sp to cause AOM. PCV7 thus results overall in a marginal decline in episodes of AOM, in contrast to the dramatic decline in invasive pneumococcal disease (NV-Sp being less invasive).30 Signs and symptoms of AOM due to NV-Sp and V-Sp are similar.31 From 1985 through 1999 increasing penicillin and multidrug resistance emerged among isolates of Sp. Resistance was most common among 5–7 serotypes,32 which were included in PCV7. Vaccination has been associated with decline in carriage of V-Sp, decrease in disease due to penicillin-resistant and multidrug-resistant Sp, and decrease in treatment failures in postvaccine years.22 However, reduced efÀcacy of PCV7 against Sp19F, lack of cross-protection against Sp19A, absolute increase in Sp19A and emergence of resistance in several NV-Sp (especially Sp19A)33,34 suggest that antibiotic resistance remains a challenge in therapy of AOM.

CHAPTER

31

223

Haemophilus influenzae NTHi causes a substantial proportion of cases of middle-ear disease in children.35 A well-recognized clinical syndrome, otitis-conjunctivitis, is associated with recovery of NTHi from both the middle ear and the conjunctiva.36 Several investigators have reported an increased proportion of disease due to NTHi, speciÀcally among immunized children with persistent or recurrent disease after initial treatment.21,22 The majority of these isolates of NTHi are b-lactamase-producing and amoxicillin-resistant. In Japan, isolates of NTHi with high-level resistance to amoxicillin due to alterations in the penicillin-binding proteins have been recovered from children with AOM.37 Such isolates remain uncommon in the United States; however increased detection has been reported.

Moraxella catarrhalis M. catarrhalis has been recognized only recently as a bona Àde pathogen in AOM, with isolation from middle-ear cultures of children with AOM and development of a humoral immune response.38 The organism appears to be more prevalent in certain geographic areas as well as in the Àrst year of life.

Groups A and Group B Streptococcus Otitis media due to GAS was common in the preantibiotic era. Currently, these organisms are most often seen in school-aged children (> 5 years of age). Jacobs and associates8 reported a greater proportion of cases in Eastern European children than in United States children. Group B streptococcus (GBS) is isolated primarily from the middle ear of neonates and young infants.39 Bacteremia can be associated with AOM in this age group.

Staphylococcus aureus and Staphylococcus epidermidis Staphylococcus aureus is isolated in a small proportion (< 3%) of middle-ear cultures from children with AOM and intact tympanic membranes. The organism is more important as a potential pathogen in children with acute otorrhea associated with a tympanostomy tube.40 Community-associated methicillin-resistant S. aureus (CAMRSA) has emerged as an important pathogen in acute otorrhea in children with tympanostomy tubes.40,41 S. epidermidis is infrequently isolated as the sole pathogen in children with AOM, and its role is uncertain.42

Middle ear fluid isolates (%)

30 Vaccine groups 25 20 15

Nonvaccine groups

10 5 0

100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0%

Cumulative (%)

Epidemiology

19 6 23 14 18 9 4 3 15 1 7 11 22 8 10 35 Serogroups Figure 31-3. Serogroups of Streptococcus pneumoniae causing acute otitis media in North American children. (Redrawn from Hausdorff WP, Bryant J, Kloek C, et al. The contribution of specific pneumococcal serogroups to different disease manifestations: implications for conjugate vaccine formulation and use. Part II. Clin Infect Dis 2000;30:122–140.)

Otitis media is virtually universal in childhood. Ninety percent of children have at least one symptomatic or asymptomatic episode by 2 years of age.43 The age-speciÀc incidence peaks between 6 and 18 months in the United States and somewhat later in Europe. Risk factors for otitis media are: young age,44 exposure to young children (group childcare attendance or siblings in household),45 and family history. Studies evaluating features such as race,46 breastfeeding,47 use of paciÀer,48 and exposure to cigarette smoke49 have identiÀed associations with AOM less consistently. These Àndings may reflect issues in study design (passive reporting versus active surveillance), difÀculty in categorizing complex variables (breastfeeding: all versus partial versus none), or the inability to control multiple interrelated features (socioeconomic factors, smoking, access to care). Most important in the development of strategies for reducing the burden of AOM and otitis media with effusion (OME) is the recognition of the otitis-prone child. Onset of disease in the Àrst few months of life is a sentinel event, either as a summation of risk features or possibly as an event that itself creates enhanced susceptibility. Unfortunately, trends in the United States, as reported by Block et al.,50 suggest that early onset of AOM and the proportion of children with ≥3 or ≥6 episodes before their Àrst birthday are increasing compared with observations by Teele and colleagues3 in 1989 (Figure 31-4).

Proportion of children

224

SECTION

100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0%

C Upper Respiratory Tract and Oral Infections TABLE 31-2. Clinical Outcome by Treatment Assignment and Age

90 82

Age Cohort

62

59

Failure (Day 0–12)

Recurrence (Day 13–33)

Cure

No.

%

No.

%

No.

%

4

6

11

17

50

77

12 16

24 14

10 21

20 18

28 78

56 68

1

2

9

21

34

77

9 10

18 14

3 12

6 16

38 72

76 76

< 2 YEARS

42

Antibiotic immediately Watchful waiting Total

34 27 17

17 8

10 1 3

1 6

12

24

Age (months) *

†

Boston

1 episode

Bardstown

3 episodes 6 episodes

Figure 31-4. Changing epidemiology of acute otitis media. Incidence of the disease by year of life, 1970s (Boston, 1975 to 1979) versus 1990s (Bardstown; 1989 to 199337). *Boston; †Bardstown.

Treatment Multiple issues require evaluation in order to determine the appropriate strategy for treatment of AOM: (1) What is the natural history of AOM, and does antibiotic therapy alter the natural history? (2) Should the outcome be measured by clinical or microbiologic endpoints? (3) Does the emergence of multiple-drug-resistant Streptococcus pneumoniae affect the outcome and choice of therapy for AOM? (4) What guidelines have been established for the treatment of AOM?

Natural History In the preantibiotic era, mastoiditis was a common complication of AOM; the use of antimicrobial therapy for the treatment of AOM has resulted in a dramatic decline in incidence. Similarly, in special populations such as Alaska Eskimos, chronic otorrhea occurred in ~30% of children prior to the routine use of antimicrobial agents.51 In 1999, Baxter52 reported a substantial decline in chronic otorrhea over a 30-year period that was associated with the introduction of routine antimicrobial therapy for the treatment of AOM. During this same period, socioeconomic conditions also improved, blurring the speciÀc contribution of medical management in reducing chronic otorrhea. In general, an episode of AOM resolves in the majority of children with or without antimicrobial therapy. Antibiotic treatment does not beneÀt the 20% to 30% of episodes of middle-ear disease with negative culture results. Additionally, some children with bacterial AOM clear the pathogen spontaneously, which is predictable by pathogen: 20% for S. pneumoniae, 50% for NTHi, and 75% for M. catarrhalis.53 Older children are more likely to clear infections spontaneously than children younger than 2 years. Therapeutic trials that include placebo groups consistently report excess failure rates in children who receive only symptomatic treatment. Engelhard and associates54 reported > 70% failure in children who received myringotomy alone. Kaleida and colleagues55 observed a twofold higher failure rate among children with temperature > 39.4°C treated with myringotomy plus placebo compared with antibiotics (23.5% versus 11.5%). They also observed an almost twofold greater failure rate in children with nonsevere AOM who were treated with placebo compared with those who received amoxicillin (7.7%

> 2 YEARS

Antibiotic immediately Watchful waiting Total

From McCormick DP, Chonmaitree T, Pittman C, et al. Nonsevere acute otitis media: a clinical trial comparing outcomes of watchful waiting versus immediate antibiotic treatment. Pediatrics 2005;115:1455–1465.

versus 3.9%). Little and coworkers56 compared the outcome of AOM in children initially treated with amoxicillin and in those given a prescription to be Àlled only if symptoms persisted for 72 hours. Children treated with antibiotics improved more quickly. McCormick and colleagues evaluated the strategy of “watchful waiting” for children with perceived mild disease as deÀned by a structured assessment.57 Children assigned to the delayed treatment group had increased treatment failures and persistent symptoms (Table 31-2). Failure was deÀned as an abnormal tympanic membrane and an AOM symptom score higher than at enrollment. Even with the delayed resolution in the “watchful waiting” cohort, parent satisfaction was reported as equivalent in both the early treatment and the initial observation cohorts. More mild adverse events as well as the emergence of multidrug-resistanct isolates of Sp in the nasopharynx occurred in the early-treatment group. Evaluating outcomes within the Àrst 3 to 5 days discriminates best between antibiotic-treated and placebo-treated cohorts as well as between antibiotic regimens.58,59 Effective antimicrobial therapy sterilizes the middle ear, resulting in a more rapid resolution of clinical signs (bulging and erythema) and symptoms (fever, earache, irritability). If outcomes such as resolution of signs and symptoms by day 7 to 10, persistence of middle-ear fluid at day 14 or 28, or recurrences within the Àrst 30 days are used, no difference between treatment strategies can be consistently established. A substantial proportion of children (30% to 50%), especially younger children, have persistent middle-ear effusion (OME) 30 days after an acute episode, long after resolution of acute signs and symptoms.60 In a study by Teele and associates61 of resolution of OME, 40% of children had OME at 1 month, 20% at 2 months, and 10% at 3 months. OME is part of the morbidity of otitis media because of its association with conductive hearing loss of up to 50 db.62

Pharmacodynamic Principles of Antimicrobial Selection In 1992, the Infectious Disease Society of America established guidelines for evaluation of new agents for AOM.63 Guidelines require both clinical and microbiologic studies as well as outcome measures 3 to 5 days after initiation of antimicrobial therapy. They are consistent with the principles established by Carlin and colleagues,64 and conÀrmed by Dagan and associates,65 that sterilization of the middle ear has a high correlation with clinical success (Table 31-3) and that if only clinical endpoints are used, large numbers of study subjects are required to determine the superiority or equivalence of two treatment strategies. The selection of antimicrobial therapy should be based on pharmacodynamic principles and results of clinical trials using microbiologic endpoints. Although 18 antibiotics are currently approved by the

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Otitis Media

TABLE 31-3. Correlation Between Microbiologic Outcome at Day 4/5 and Clinical Outcome in Acute Otitis Media Clinical Outcome

Success Failure

Culture Negative at Day 4/5 %

No.

%

236 17

93 7

25 15

60 40

Adapted from Carlin SA, Marchant CD, Shurin PA, et al. Host factors and early therapeutic response in acute otitis media. J Pediatr 1991;118:178–183.

United States Food and Drug Administration (FDA) for treatment of AOM, the emergence of otopathogens with reduced susceptibility to b-lactam antibiotics (S. pneumoniae and NTHi) results in the failure of several antimicrobial agents to achieve sufÀcient middle-ear concentrations to eradicate otopathogens. b-lactam agents and trimethoprim-sulfamethoxazole (TMP-SMX) depend on time that concentration of drug is above the minimum inhibitory concentration (MIC) for efÀciency.66 In Craig’s studies,67 efÀcacy (sterilization of middle-ear fluid) for b-lactam agents is predictable when the drug concentration exceeds the MIC for ~40% of the dosing interval; for azolides (azithromycin) and fluoroquinolones, ratios of AUC24:MIC or peak concentration to MIC; and for azithromycin, by a AUC24:MIC ratio of 25. Although these principles are well established, controversy persists regarding whether extracellular or intracellular concentrations of antibiotics are necessary. For b-lactam agents, there is no enhanced intracellular accumulation; however, for azolides and ketolides, intracellular concentrations of drug exceed extracellular (or serum) concentrations by up to 100-fold. Results from animal models and clinical trials suggest that extracellular concentrations are the critical factor.68 Table 31-4 lists selected antibiotics and the time above MIC or AUC24:MIC ratio achieved for susceptible and resistant (MIC ≥ 2.0 mg/mL for penicillin) S. pneumoniae and b-lactamase-negative and b-lactamase-positive NTHi. The acceptance of these principles led the Clinical and Laboratory Standards Institute (CLSI) to identify speciÀc breakpoints for susceptibility versus resistance relative to sites of disease. Table 31-5 summarizes both CLSI-speciÀc breakpoints and those derived from pharmacodynamic analysis for S. pneumoniae and NTHi.69 Table 31-6 provides data on the proportion of isolates susceptible to a given antibiotic based on pharmacokinetic and pharmacodynamic (PK/PD) factors.70 In the majority of cases of AOM, the speciÀc pathogen is unknown, and presumptive therapy is based on the likely pathogens and their in vitro susceptibility (Figure 31-5). Although disease due to S. pneumoniae is frequently associated with higher temperatures and greater intensity of pain, the overlap in signs and symptoms is sufÀciently large that clinical differentiation is not possible.71 The proportion of isolates of NTHi-producing b-lactamase has risen to 40% over a 25-year period.72 A very limited number of b-lactamase-producing, amoxicillin-clavulanate-resistant isolates have also emerged. The spread of pneumococci with altered penicillinbinding proteins and reduced susceptibility to b-lactam agents occurred rapidly in the United States (Figure 31-6).73–75 A 2001 United States study identiÀed 42% of 500 middle-ear Sp isolates as penicillinresistant (MIC ≥ 2.0 mg/mL) and 17% as amoxicillin-resistant.32 Resistance to macrolides and azolides has also become common and is mediated either by an efflux pump mechanism that decreases intracellular accumulation of antibiotic or by a mutation in ribosomal methylase, which affects binding between drug and target site, or both.76,77 There are no clinical differences between AOM caused by resistant or susceptible pathogens. Risk factors for the presence of isolates of S. pneumoniae with reduced susceptibility are age < 2 years, recent treatment with antibiotics, season of the year (late winter to early

31

225

TABLE 31-4. Percentage of Time (%) Above Minimal Inhibitory Concentration (MIC) for b-Lactam Agents or AUC24: MIC Ratio for Azithromycin for Otopathogens

Culture Positive at Day 4/5

No.

CHAPTER

Streptococcus pneumoniae Antibiotic (mg/kg per day)

Haemophilus influenzae b-Lactamase– b-Lactamase– Negative Positive

PenicillinSusceptible

PenicillinResistant

Amoxicillin 45 mg

~85

~40

~60

~10

Amoxicillin 90 mg

~85

~60

~75

~10

Amoxicillinclavulanate 90 mg

~85

~60

~75

~75

Cefuroxime 30 mg

~60

~30

~30

~30

Cefprozil 30 mg

~80

~20

~10

~10

CeÀxime 8 mg

~80

~20

~80

~80

50

95% of Sp; clinical trials demonstrate rapid sterilization and clinical resolution of middle-ear infection due to both pathogens.83,84 This class of antimicrobial agent is not licensed for use in children for treatment of AOM. Relapse and recurrence have been deÀned as occurrence of episodes of disease after an initial symptomatic response to antibiotic treatment, either during treatment or within some deÀned period after completion of therapy. Discriminating relapse from recurrence requires deÀning the microbiology of both episodes. Two studies demonstrate that the majority of such events are due to new pathogens or different serotypes of the same pathogen.85,86 Two other studies found that, after treatment of AOM with ceftriaxone or other b-lactam antibiotics, susceptible isolates of S. pneumoniae are eliminated from the nasopharynx but there is little effect on resistant strains.87,88 Newly acquired S. pneumoniae in children who are receiving antimicrobial therapy frequently demonstrate resistance.89 These events appear to begin as early as 3 to 4 days into the course of treatment.88 Thus, the likelihood of nasopharyngeal colonization with a resistant otopathogen increases during and after antibiotic therapy if resistant isolates are present in the community;57 recurrent episodes of AOM are more likely due to resistant pathogens. Figure 31-7 demonstrates that the chance of recovering penicillin-nonsusceptible pneumococcus correlates with time since last receipt of antibiotics.89 The data support the recommendation from the American Academy of Pediatrics (AAP) that treatment of recurrent episodes within 30 days of a prior event may necessitate use of antibiotic agents (such as ceftriaxone or high-dose amoxicillin-clavulanate) that target resistant otopathogens.90

Percentage of Susceptible Isolates Antimicrobial Agent

Streptococcus pneumoniae

Haemophilus influenzae

Moraxella catarrhalis

Amoxicillinclavulanate

90

97

100

Amoxicillin

90

61

14

Cefaclor

27

2

5

CeÀxime

57

99

100

Cefpodoxime

63

99

100

Cefprozil

64

18

6

Cefuroxime

64

79

37

Macrolides

67

0

N/A

Clindamycin

89

N/A

N/A

TMP-SMX

57

75

9

N/A, not applicable; TMP-SMX, trimethoprim-sulfamethoxazole. From Sinus and Allergy Health Partnership. Antimicrobial treatment guidelines for acute bacterial rhinosinusitis. Otolaryngol Head Neck Surg 2000;123:5–31.

Initial AOM

Treatment failure after Amoxicillin

Recurrent AOM within 30 days

PSS BL-NTH BL+M

BL+NTH DRS PSS BI-NTH

DRS BL+NTH PSS BL-NTH

50

Intermediate (0.12 to 1.0 mg/mL)

40

33% 29%

Resistant (2.0 mg/gmL)

30 20

18% 16%

10 1998

1997

1994–95

1992–93

1990–91

1988–89

1987

1986

1985

1984

1983

1982

1981

1980

0 1979

% penicillin resistant (mg/mL)

Figure 31-5. Presumptive pathogens in acute otitis media (AOM). BL (+) or (–) NTHi, b-lactamase-positive or -negative nontypable Haemophilus influenzae; DRSP, drug-resistant Streptococcus pneumoniae; Mc, Moraxella catarrhalis; PSSp, penicillin-susceptible Streptococcus pneumoniae.

Year Figure 31-6. Proportion of Streptococcus pneumoniae isolates with reduced susceptibility to penicillin in United States (1979 to 1998). MIC, minimal inhibitory concentration. (Data from Doern74 and Jacobs et al.73,75)

also be effective against DRSP;81 linezoid has no efÀcacy against other otopathogens. Table 31-6 summarizes the proportion of pneumococcal isolates calculated to be susceptible to antimicrobial agents based on PK/PD principles. The table permits selection of antimicrobial regimens for children, especially those in whom initial therapy has failed and a resistant pathogen is more likely.

AAP Guidelines for the Diagnosis and Treatment of AOM The AAP guidelines, published in 2004, provide principles for improving the diagnosis of AOM.90 The presence of middle-ear effusion as detected by physical examination or tympanometry is the critical criterion. Diagnosis also requires recent onset of signs and symptoms of acute inflammation such as earache, ear-tugging, or a bulging tympanic membrane. In one study, approximately 85% of children with earache, ear-tugging, bulging or extreme redness of the tympanic membrane had positive tympanocentesis cultures, whereas only 50% of children with nonspeciÀc symptoms such as fever and

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Proportion of pts.

Otitis Media 100% 75% 50% 25% 0% No antibiotics Antibiotics Antibiotics Nonresponsive in last 3 in prior in AOM (on months 2–3 months prior month antibiotics) MIC 1 mg/mL Figure 31-7. Correlation between recent antibiotic therapy and likelihood of drug-resistant Streptococcus pneumoniae in acute otititis media (AOM). MIC, minimal inhibitory concentration.150

irritability had positive cultures.91 If bacterial AOM is less likely, it is less likely that the child will beneÀt from antimicrobial therapy as nonbacterial AOM is cured or improved within 10 days in over 96% of cases with only supportive therapy (Dagan, personal communication). The AAP guidelines recommend amoxicillin at 90 mg/kg per day administered twice daily for initial therapy in most children with AOM (Table 31-7).90 Among b-lactam antibiotics, only oral “highdose” amoxicillin or intramuscular ceftriaxone achieves middle-ear concentrations that fulÀll PK/PD requirements for penicillinnonsusceptible Sp and b-lactamase-nonproducing NTHi. Cefuroxime axetil, cefprozil, and cefpodoxime represent alternatives to high-dose amoxicillin; however, each only achieves sufÀcient middle-ear concentration to be effective against < 50% of penicillin-nonsusceptible Sp. Also, cefprozil has limited activity against NTHi.92 Macrolide efÀcacy at currently recommended dosage is limited to disease due to penicillin-susceptible Sp. Extracellular middle-ear fluid concentrations of macrolides are below the MIC for almost all NTHi

CHAPTER

31

227

and Sp isolates with efflux or ribosomal mechanisms of resistance. Amoxicillin is ineffective against b-lactamase-producing isolates of NTHi. Because amoxicillin-clavulanate resists destruction by the betalactamase, it effectively eradicates middle-ear infection caused by NTHi.30,42 Isolates of NTHi with altered penicillin-binding proteins are common in Japan and have been reported in the United States.93 Should such isolates continue to increase, the efÀcacy of high-dose amoxicillin-clavulanate would be expected to decline. Although the AAP guidelines recommend the consideration of amoxicillinclavulanate as initial therapy only for children with severe disease (temperature > 39°C) and substantial otalgia), the increasing prevalence of AOM due to NTHi associated with universal PCV7 immunization may warrant broader use of amoxicillin-clavulanate as Àrst-line therapy. Ongoing evaluation of the rate of treatment failure following therapy with amoxicillin will be required. Initial therapy for the child with type I allergy to penicillin (urticaria, laryngeal spasm, wheezing, or anaphylaxis) remains challenging. The choice of alternatives to b-lactams is limited by a substantial prevalence of resistance (Table 31-8). Macrolides, including azithromycin and clarithromycin, demonstrate in vitro activity against most SpI isolates; however macrolides would be ineffective for 25% to 40% of Sp.94 Resistance to TMP-SMX among Sp and NTHi is also substantial.95 The AAP guidelines acknowledge the potential limitations of these agents but recommend their use as best alternative. For the child with “severe” disease, a combination of agents such as clindamycin (for SP) and sulfasoxazole (for NTHi) may be effective. Unlike other macrolides, clindamycin may be effective against isolates of Sp with the efflux mechanism of resistance and therefore maintains activity against approximately 80% to 90% of SP isolates in the United States. However, the proportion of isolates of SP with ribosomal mechanisms of resistance is increasing and localcommunity antimicrobial susceptibility patterns provide the best information about the predicted efÀcacy of clindamycin.96 The AAP guidelines emphasize that “watchful waiting” includes providing analgesia for children suffering from AOM. A limited number of studies suggest that ibuprofen or acetaminophen is effective.97 Topical agents such as auralgan may also offer temporary symptomatic relief.98 For children with severe pain, myringotomy is an effective method to attain relief.

TABLE 31-7. Recommendations for Treatment of Acute Otitis Media in the Child without Penicillin Allergy Temperature > 39°C and/or severe otalgia

Diagnosis, Day 0 (when initial management includes antibiotic rx)

Clinical Failure, Day 3 (when initial management was observation)

Clinical Failure, Day 3 (when initial management was antibiotic)

No

High dose-amoxicillin, (80–90 mg/kg per day)

High-dose amoxicillin (80–90 mg/kg per day)

High-dose amoxicillin-clavulanate (90 mg/kg per day amoxicillin; 6.4 mg/kg per day clavulanate)

Yes

High-dose amoxicillin-clavulanate (90 mg/kg per day amoxicillin; 6.4 mg/kg per day clavulanate)

High dose amoxicillin- clavunate; or ceftraixone IM, 1 or 3 days

Ceftraixone IM, 3 days

IM, intramuscularly. From Lieberthal AS, Ganiates TG, Cox EO, Culpepper L, Mahony M, Miler D, et al. Clinical practice guidelines: diagnosis and treatment of acute otitis media. Pediatrics 2004;13:1451–1465.

TABLE 31-8. Recommendations for Treatment of Acute Otitis Media in Children with Penicillin Allergy Temperature > 39°C and/or severe otalgia

Diagnosis, Day 0 (when initial management includes antibiotic)

Clinical Failure, Day 3 (when management was observation)

Clinical Failure, Day 3 (when management was antibiotic)

No

Nontype I: cefdinir, cefuroxime, or cefpodoxime Type I: azithromcyin, clarithromycin

Nontype I: cefdinir, cefuroxime, cefpodoxime Type I: azithromcyin, clarithromycin

Nontype I: ceftriaxone IM, 3 days Type I: clindamycin

Yes

Nontype I: ceftriaxone IM, 1 or 3 days Type I: clindamycin

Nontype I: ceftiaxone IM, 1–3 days Type I: clindamycin

Tympanocentesis: clindamycin

IM, intramuscularly.

228

SECTION

C Upper Respiratory Tract and Oral Infections

Evaluation of the Child with Recurrent Otitis Media In clinical practice, most children with recurrent otitis media (ROM) do not have a quantiÀable immunologic abnormality.99,100 When an immune defect is detected, the abnormality is usually limited to IgG subclass or IgA deÀciency.99,101 These disorders are often maturational and therefore when laboratory values support such a diagnosis in infants, repeat testing after the age of 4 years is indicated. In children with recurrent serious infections in addition to ROM, more serious immunologic defects must be considered. The majority of children with severe combined immunodeÀciency syndrome (SCIDS) present with otolaryngologic signs and symptoms.102 Children with the DiGeorge anomalad occasionally have immune defects as severe in SCIDS. Common variable immunoglobulin deficiency (also referred to as common variable hypogammaglobulinemia) also is characterized by recurrent upper and lower respiratory tract infection, chronic diarrhea and bronchiectasis.103 These children have panhypogammaglobulinemia and can be differentiated from X-linked agammaglobulinemia by the presence of circulating B lymphocytes. In addition, autoimmune disorders such as arthalgia and hemolytic anemia may be present. Defects in phagocytic function may also predispose to recurrent respiratory tract infection and recurrent AOM.100,104,105 Such children frequently manifest cutaneous abscesses (pyoderma), gingivitis, candidal diaper dermatitis or perianal abscesses. The diagnosis should be suspected when pus is absent at the site of infection. Case reports of recurrent AOM and chronic diarrhea in association with defects of neutrophil granulocyte chemotaxis or with chronic granulomatious disease support a link between defective leukocyte function and enhanced susceptibility to recurrent AOM.104,105 Children with Job syndrome, hyper IgE and neutrophil dysfunction, may also suffer recurrent episodes of AOM. Recurrent AOM is very common in children with HIV infection.106–109 In general, additional signs of immunodeÀciency such as recurrent thrush, lymphadenopathy, hepatomegaly or splenomegaly, failure to thrive, or chronic diarrhea are present, although ROM may be the presenting feature and/or dominant feature early in the course of disease. Marchisio et al. deÀned the bacterial etiology of AOM in 60 episodes in 21 HIV-infected children.110 S. pneumoniae, NTHi and GAS were recovered from the middle ear in 56.5% of cases, a

proportion similar to that seen in immunocompetent children. Staphylococcus aureus was identiÀed in HIV-infected children with severe immunosuppression. In general, an immune evaluation for the child whose manifestations are limited to ROM, in the absence of additional concerns for enhanced susceptibility to infection or signs or symptoms such as failure to thrive, chronic diarrhea, lymphadenopathy, or organomegaly, is unlikely to discover a serious immune deÀciency. Even when the diagnosis of an IgA or IgG subclass deÀciency is suggested from the measurement of immunoglobulins and immunoglobulin subclasses, it is often maturational and transient. For the two most commonly considered management strategies, prophylactic antibiotics and intravenous immunoglobulin treatment, it is not necessary to demonstrate a speciÀc immune defect as the otitis-prone child is likely to beneÀt from either. In children with a family history of immune deÀciency, or an at-risk proÀle for HIV, or additional concerns, as detailed in Box 31-1, an extensive immune workup is indicated. Table 31-9 identiÀes the spectrum of evaluation that should be considered as well as the abnormality that would potentially be observed.

Prevention of Recurrent Acute Otitis Media Decreasing the morbidity of otitis media has potential implications for decreasing acute febrile illness in children and associated discomfort, BOX 31-1 Clinical Indications for Immunologic Evaluation in Children with Recurrent Otitis Media • • • • • • • •

Recurrent pneumonia or bronchiectasis Recurrent invasive bacterial disease Absent tonsils, lymph nodes, or thymus (on chest radiograph) Infection with opportunistic pathogen Persistent dermatitis Recurrent or chronic diarrhea Failure to thrive Family history of immune deÀciency or risk features for human immunodeÀciency virus (HIV) • Recurrent thrush and/or gingivitis • Hepatosplenomegaly

TABLE 31–9. Laboratory Evaluation of Children with Recurrent Otitis Media (ROM)a Laboratory Test

Observation

Potential Dx

Complete blood count

Lymphopenia

SCID HIV DiGeorge anomalad Leukocyte adhesion defect (LAD)

Leukocytosis Quantitative immunoglobulins (Ig)

Œ IgG Œ IgA Ø IgE

X-linked agammaglobulinemia Combined variable immunoglobulin deÀciency Transient hypogammaglobulinemia of infancy IgA deÀciency Ataxia telengiectasia Job syndrome

IgG subclasses

ΠIgG1, ΠIgG2, ΠIgG3, ΠIgG4

Rebuck skin window test

ΠNeutrophil migration

LAD

Boyden chamber

ΠChemotaxis

LAD

Chemiluminescence or nitro-blue tetrazolium (NBT) test

Œ Oxidative burst ±

Chronic granulomatous disease (CGD) LAD

Flow cytometry analysis of neutrophils HIV antibody

ΠCD11, ΠCD18 EIA and Western blota

LAD HIVb

IgG subclass deÀciency

SCID, severe combined immunodeÀciency; EIA, enzyme immunoassay; HIV, human immunodeÀciency virus; Ig, immunoglobulin; a Antibody can be present in the Àrst 18 months of life due to passive transfer across placenta without HIV infection. b For children less than 18 months, demonstration of HIV by polymerase chain reaction or culture is necessary for diagnosis. ModiÀed from Pelton SI, In Alper CM, Bluestone CD, Dohar JE et al. (eds) Advanced Therapy in Otitis Media. BC Dekker, Toronto, 2003.149

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Otitis Media

CHAPTER

Complications of Acute Otitis Media Perforation of the tympanic membrane is the most common complication of AOM and is most frequently observed in younger children. Certain ethnic groups, such as Alaska Eskimos and Native Americans, have a higher proportion of spontaneous perforation with AOM. Differentiation between AOM with perforation and acute otitis externa can be difÀcult. In general, the history of increasing pain with relief when otorrhea occurs is found with AOM, whereas increasing pain without relief in the face of otorrhea is seen with otitis externa (see Chapter 32, Otitis Externa and Malignant Otitis Externa). The microbiology of AOM in children with acute perforation demonstrates a higher proportion of episodes due to GAS and Staphylococcus aureus.40 However, Streptococcus pneumoniae, NTHi, and Moraxella catarrhalis remain predominant. The natural history of AOM with perforation is usually complete resolution with healing of the tympanic membrane. A small proportion of patients have persistent dry perforation or chronic suppurative otitis media (persisting for more than 6 to 12 weeks). S. pneumoniae and NTHi are the most common pathogens in infants and toddlers, whereas Staphylococcus aureus and Pseudomonas aeruginosa are frequent pathogens in older children and during the summer months.126 An increasing proportion of the S. aureus isolates are CA-MRSA. Amoxicillin is generally effective for the therapy of acute otorrhea through a tympanostomy tube and, compared with placebo, results in rapid clearing of bacterial pathogens and a shortened duration of otorrhea.127 An alternative to oral antimicrobial therapy is topical otic suspensions, either ofloxacin or ciprofloxacin.128,129 When MRSA is the pathogen, oral amoxicillin or topical quinolone preparations have largely failed and drainage persists.130 Limitied reports of success with topical vancomycin (25 mg/mL) drops or the use of TMP-SMX orally in combination with gentamicin otic solution suggest otorrhea due to MRSA can be cured without removal of the tympanostomy tube.131,132 Caution with both of these regimens is necessary as safety has not been established. Facial palsy as a complication of AOM is uncommon.133 Facial weakness and earache are the predominant symptoms. Antibiotic therapy and myringotomy (with or without tube insertion) are usually sufÀcient to achieve complete resolution. Mastoiditis, once commonplace, has dramatically decreased since the routine use of antimicrobial therapy. However, it still remains the most common suppurative complication of AOM.134 Although the potential for re-emergence of mastoiditis when antimicrobial agents are withheld for AOM has been a concern, there are few convincing

Chemoprophylaxis Chemoprophylaxis offers short-term beneÀts only. Most otitis-prone children continue to have recurrent episodes once prophylaxis is discontinued, until their immune systems and eustachian tube function have matured.113 A Àve-carbon sugar alcohol, xylitol, has been demonstrated to reduce episodes of AOM in group childcare attendees (3.03 versus 2.01 episodes per child year).115 The hypothesized mechanism is reduction in adherence and inhibition of growth of S. pneumoniae; however, these potential mechanisms have only been demonstrated in vitro.116,117 Viral respiratory tract illness is a cofactor in the pathogenesis of AOM. Influenza vaccine has been associated with a reduction in AOM episodes, febrile AOM and in myringotomy and insertion of tympanostomy tubes over a winter season.14,118 Studies of respiratory syncytial virus immune globulin (RSV-IGIV) but not RSV monoclonal antibody showed a reduction in number of AOM episodes.119,120 RSV-IGIV (no longer available) likely provided passive antibody against bacterial otopathogens.

Vaccination Two 7-valent pneumococcal conjugate vaccines (PCVCRM and PCVOMP), administered at 2, 4, and 6 months with a booster at 12 to 15 months of either 7-valent PCV or 23-valent pneumococcal polysaccharide vaccine, have been shown to reduce AOM due to vaccine serotypes of S. pneumoniae by approximately 60%, and all episodes of pneumococcal otitis media by one-third (Table 31-10).121,122 However, the overall reduction in episodes of AOM was more modest (6% to 10%). Postmarketing studies have conÀrmed a shift in the proportion

TABLE 31-10. Efficacy of 7-Valent Pneumococcal Conjugate Vaccine (PCV7) for the Prevention of Pneumococcal Otitis Media No. of Episodes PCV7

229

of disease due to NV-Sp and NTHi. The effect of these increases on long-term efÀcacy of the vaccines remains to be determined fully. Follow-up studies of cohorts of the original clinical trials of infant immunization with PCV7 in California and Finland have identiÀed signiÀcant reductions in tympanostomy tube insertions in immunized children.123,124 Studies of PCV7 immunization in children with frequent ROM have failed to demonstrate a signiÀcant reduction in episodes. Veenhoven et al. observed that ROM due to V-Sp was only a small proportion of overall episodes; reducing such episodes had no impact on total recurrences.125

reducing healthcare costs and surgical procedures, and preventing the language and cognitive delays that occur in some children with recurrent AOM and prolonged conductive hearing loss. Freid & Makuc111 identiÀed middle-ear disease as the most common reason for ambulatory healthcare visits. Paradise and colleagues43 reported that > 45% of urban children in the Àrst year of life, and 30% in the second year spend more than 3 months of each year with middle-ear effusion; 10% of children spent > 50% of each year with middle-ear effusions. Prevention of recurrent bacterial AOM can be achieved by preventing nasopharyngeal colonization with otopathogens, preventing viral respiratory infection, and providing speciÀc antibacterial immunity. Insertion of tympanostomy tubes does not reduce the frequency of acute episodes substantially; however, the presence of such tubes shortens the duration of middle-ear effusion.112 Antimicrobial prophylaxis lowers the frequency of colonization with respiratory otopathogens and decreases the number of acute episodes.113 Multiple studies have demonstrated this beneÀt. Mandel and colleagues114 found a decrease in acute episodes from 1.04 per child year in the placebo group to 0.28 in the amoxicillin group.114 They also reported a reduction in episodes of OME from 2.15 episodes to 1.53 episodes per child year. The greatest beneÀt occurs in children at highest risk for recurrent AOM (age 6 to 24 months) and in otitisprone children who have multiple episodes per year and in whom the problem does not abate with increasing age.

Etiology of Otitis Media

31

Control

Difference (%)

All pneumococcal episodes

269

420

Œ36

Vaccine serotypes

107

254

Œ57

Vaccine serogroups

143

325

Œ56

Nonvaccine serogroups

126

95

Ø34

Data from Eskola J, Kilpi T, Palmu A, et al. For the Finnish Otitis Media Study. EfÀcacy of a pneumococcal conjugate vaccine against otitis media. N Engl J Med 2001;344:403–409.

230

SECTION

C Upper Respiratory Tract and Oral Infections

data available that this is occurring.135 Streptococcus pneumoniae, GAS, and NTHi are most common pathogens of acute mastoiditis but Pseudomonas aeruginosa was found in 29% and Staphylococcus epidermidis in 31% of cases in one large series.134 These pathogens should be suspected when a history of otorrhea precedes development of acute mastoiditis (see Chapter 33, Mastoiditis). Labyrinthitis develops when AOM spreads (through the round window) into the cochlear space. The process can be suppurative or serous (due to toxins).136 The onset of labyrinthitis is often sudden, with vertigo and hearing loss being characteristic. Acute surgical intervention (myringotomy with tube insertion) with antimicrobial therapy is the treatment of choice. Additional rare complications of AOM are brain abscess, epidural abscess, and otitic hydrocephalus, which can result from transverse, lateral, or sigmoid sinus thrombosis. Magnetic resonance imaging and magnetic resonance venography aid differentiation of complications and optimal management.

OTITIS MEDIA WITH EFFUSION OME is the relatively asymptomatic presence of middle-ear effusion that is often associated with conductive hearing loss and can persist for months. The pathogenesis most often reflects a postinfectious event that results from the presence of cytokines and other inflammatory mediators. An alternative theory suggests that effusion results from hydrops ex vacuo due to eustachian tube obstruction, negative middleear pressure, and transudation of fluid. Bacterial otopathogens are recovered in 20% to 30% of middle-ear effusions in children with OME; however, the precise role of otopathogens in pathogenesis is unclear.137 Several investigators have suggested bioÀlm formation and persistence may be responsible for the ongoing inflammatory response that provokes persistence of middle-ear effusion.138,139 In most children, OME resolves without speciÀc intervention within a few months.140 When OME persists beyond this period and there is a substantial conductive hearing loss (> 20 db), intervention is considered. A recent meta-analysis failed to demonstrate signiÀcant language delay or cognitive impairment in otherwise healthy children with persistent middle-ear effusion.141–143 Results of a large longitudial prospective study in which otherwise healthy young children with persistent middle ear effusion were randomized to present insertion of tympanostomy tubes or placement up to 9 months later if effusion persisted showed no difference in developmental outcomes up to 9 to 11 years of age.144 Furthermore, the management of OME with tympanostomy tube placement may not be benign. Consequences of tube placement may include recurrent otorrhea, persistent perforation, development of granulation tissue, chronic otitis media, and cholesteatoma.145 Repeated insertions of tympanostomy tubes have also been associated with mild hearing loss, including a sensorineural component.146 Guidelines from the AAP and American Academy of Otolaryngology145 stress the need to identify children at risk for speech, language, and cognitive impairment early in the course of disease. The guidelines focus on children with sensory, physical, cognitive, or behavioral features that are likely to increase the risk of BOX 31-2 Comorbid Conditions Associated with Increased Risk for Developmental Difficulties in Children with Otitis Media with Effusion (OME) • • • • • •

Permanent hearing loss separate from OME Suspicion of speech language delay or disorder Autism spectrum disorders Uncorrectable visual disorders or blindness Cleft palate Congenital syndromes associated with cognitive, speech, or language delays • Documented developmental delay of unknown etiology From Rosenfeld RM, Culpepper L, Doyle KJ, et al. Clinical practice guideline: otitis media with effusion. Otolaryngol-Head Neck Surg 2004;130:s95–s118.

developmental difÀculties (Box 31-2) and therefore warrant earlier hearing and language assessment and tympanostomy tube insertion. For otherwise healthy children, watchful waiting for up to 6 months, with hearing evaluation at 3 months, is suggested. The primary intervention for OME is surgical; antihistamines and decongestants are ineffective and therapy with antibiotics or corticosteroids does not result in lasting beneÀt.145 Tympanostomy tube insertion for otherwise healthy children is recommended when persistent hearing loss > 40 db is documented or symptomatic disease is present (difÀculties in balance or sleep disturbance associated with intermittent ear pain, fullness, or popping). Tympanostomy tube insertion results in improved hearing, a decrease in days spent with effusion, but no decline in recurrent episodes of AOM. Adenoidectomy performed in children less than 2 years of age undergoing tube reinsertion after spontaneous extrusion145 appears to reduce the need for subsequent tube insertions. Tonsillectomy alone or myringotomy alone is not an effective intervention for OME. The beneÀt of antibiotic therapy for OME is controversial, despite a meta-analysis that demonstrated a modest beneÀcial effect of antibiotics.147 There is a high likelihood of recurrence of effusion in children who demonstrate clearance after a course of antimicrobial therapy.148 This observation has limited enthusiasm for antibiotic therapy, although most clinicians prescribe a short course of antimicrobial therapy prior to recommending tube insertion. A short course of corticosteroid also has limited efÀcacy in hastening the resolution of OME, with a substantial rate of relapse rate in children who demonstrate initial resolution of middle-ear effusion.149,150

CHAPTER

32

Otitis Externa and Malignant Otitis Externa Richard H. Schwartz

OTITIS EXTERNA Epidemiology and Clinical Manifestations Otitis externa is inflammation or infection of the skin of the hairy and glabrous parts of the ear canal (Box 32-1). This condition affects children, adolescents, and adults. Acute diffuse otitis externa is usually unilateral and is associated with head immersion in swimming pools, lakes, and the sea, but can occur after prolonged showering or shampooing. The disease peaks in the summer and during winter or spring vacations when children are taken to warm beach resorts. In humid tropical climates, acute otitis externa occurs throughout the year. When the skin of the ear canal is wet for prolonged periods, local defense mechanisms are impaired. Protective cerumen is washed away, the thickness of the keratin layer is reduced by excessive desquamation, the pH of the canal changes, and microscopic Àssures

BOX 32-1 Predisposing Factors for Otitis Externa • • • • • •

Swimming or shampooing with long hair Trauma (e.g., vigorous use of cotton-tipped swab) Dermatologic disorders (e.g., eczema) Chronic otorrhea Use of hearing aid(s) Immunocompromise of host (e.g., diabetes mellitus and mucocutaneous candidiasis) • Histiocytosis

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Otitis Externa and Malignant Otitis Externa

develop between the squamous epithelial cells. This can cause itching of the ear canal that may predispose to further trauma from cottontipped buds in an attempt to relieve the itching. Gram-negative bacteria, most commonly Pseudomonas spp. that inhabit swimming pools, lakes, and ocean water, flourish in the moist environment and invade superficial layers of skin.1–3 Localized and sometimes diffuse otitis externa also can occur after disruption of the skin of the canal, as from lodgment of foreign objects, an insect bite, vesicular lesions, scratching, eczema, or introduction of a cotton-tipped swab or bobby pin.4 The ensuing inflammatory dermatitis progresses through three stages.5,6 Each is associated with pain and tenderness exaggerated by movement of the tragus, the hallmark of acute otitis externa. The first stage, the most common, is mild otitis externa. Pain is mild to moderate, easily ameliorated by simple analgesics. Skin of the hairy part of the ear canal is erythematous but not very edematous. The second stage (moderate acute otitis externa) is characterized by increasing pain, often severe enough to interfere with restful sleep. The patient actively resists insertion of anything into the ear canal. The skin of the hairy part of the ear canal is intensely red under the macerated pieces of desquamated skin. The canal lumen is narrowed to less than 50% of its normal diameter. View of the tympanic membrane is often obscured by debris. After aural lavage with warm water or saline, the visualized lateral borders of the tympanic membrane appear red, with adherent plaques of gray-white desquamation. Mobility of the tympanic membrane is preserved, excluding the diagnosis of acute otitis media. The third stage (severe otitis externa) is less common and is characterized by: (1) intense pain, often requiring narcotics to permit normal daily activities; (2) inflammatory edema, which narrows the external auditory meatus to barely admit a nasopharyngeal swab; and (3) desquamated debris that precludes visualization of the tympanic membrane. The auricle can be swollen and painful, because the infection extends from the interior of the ear canal outward.7 In advanced cases, tender preauricular or postauricular lymphadenopathy, protrusion of the auricle, submandibular cervical lymphadenopathy, and contiguous cellulitis of the skin overlying the mastoid area can be present. In otitis externa, unlike acute mastoiditis, the entire auricle is not only protruding but is edematous and the posterior auricular sulcus is preserved. Needed lavage and suctioning of the ear canal is difficult to perform without adequate analgesia or even general anesthesia. Such cases are difficult to differentiate from malignant otitis externa (perichondritis of the cartilaginous outer third of the ear canal).8–10 It can also be difficult to distinguish intractable otitis externa from noninfectious disorders involving the ear canal, such as rhabdomyosarcoma, Langerhans cell histiocytosis, and Wegener granulomatosis. Chronic otits externa is defined as a single episode lasting longer than 4 weeks or four or more episodes in 1 year.1

Etiologic Agents In healthy individuals, the ear canal microflora includes coagulasenegative staphylococci, corynebacteria (diphtheroids), and micrococci. With prolonged exposure to water, the microflora changes to predominantly gram-negative bacteria.1 Pseudomonas aeruginosa is responsible for more than 80% of cases of otitis externa associated with swimming. Other gram-negative bacteria, such as Escherichia spp. or Proteus spp., occur less commonly. Staphylococcus aureus and Streptococcus pyogenes are usual causes of acute otitis externa that occurs as an extension of a contiguous focal infection from the site of an insect bite, ear piercing, traumatized skin, or persistent otorrhea from a patent tympanostomy tube.11 Anaerobic bacteria are not typically associated with otitis externa. Fungi, predominantly Aspergillus niger, cause fewer than 5% of cases of acute otitis externa;12 a diagnostic clue is that debris in the canal resembles wet blotting paper, with the “peppered” appearance of black mycelial elements. In tropical climates, fungi may be more common etiologic agents. Candida spp. can also cause otitis externa.

CHAPTER

32

231

Viruses, particularly varicella-zoster virus, can cause otitis externa and lead to Ramsay Hunt syndrome. Otitis externa associated with chronic draining suppurative otitis media is due to continuous irritation and maceration of the epithelium of the external canal by exudate from the middle ear. Pseudomonas aeruginosa and staphylococci do not necessarily reflect the microbiology of the middle-ear infection.

Management The most important preliminary steps in treatment of otitis externa are to: (1) alleviate pain and discomfort; (2) perform mechanical debridement of the ear canal, when there is excessive debris, by using gentle curettage, or lavage and suctioning; and (3) eliminate pathogenic organisms. Mild cases can be treated with instillation of 2% white vinegar solution (alone or with hydrocortisone suspension) or topical otic antibiotic drops. The patient with mild otitis externa can enter the swimming pool as long as prolonged submersion is avoided; excess water must be drained from the ear canal and the ear canal dried thoroughly when the patient leaves the pool area, and possibly again at home with the aid of a hair-dryer. Moderate otitis externa usually requires gentle aural lavage (warmed water, saline, or Burow solution) or suctioning of the ear canal to remove masses of skin and bacteria and thus permit contact of the antibiotic drops with underlying infected skin.13 The cleansing procedure may have to be performed daily in severe cases. After aural lavage, the canal is dried using a wisp of absorbent cotton. It may be necessary to insert a wick into the canal to allow topical antibiotic drops to remain in contact with the walls of the ear canal. Wicks are available commercially. Some otolaryngologists suggest that, after the wick is inserted, 2% dilute white vinegar (Burow solution) be instilled for 24 hours to reduce the intense inflammation and to acidify the squamous cell lining, thus creating an inhospitable microenvironment for Pseudomonas.14 Available topical otic antibiotic drops contain fluoroquinolone, aminoglycoside, or polymyxin. Combination agents include ciprofloxacin–dexamethasone, polymyxin–neosporin–hydrocortisone suspension, and trimethoprim–polymyxin–bacitracin solution. The use of hydrocortisone or dexamethasone in the otic drop mixture is empiric. The drops are applied twice or three times daily for 5 to 7 days for mild or moderate otitis externa. Head-to-head comparative studies confirm that fluoroquinolones are as effective as those containing aminoglycosides. The primary advantages of the fluoroquinolones, the current otic drops of choice for this condition, are that they are not ototoxic, do not initiate contact dermatitis as may the neomycincontaining otic drops, are effective against pseudomonal and staphylococcal pathogens, and can be administered twice daily, in contrast to polymyxin-containing otic drops that are administered four times a day.15 Culture of the debris from the ear should only be performed for severe or unresponsive cases. Severe otitis externa often requires daily meticulous aural suction or lavage (sometimes with the use of general anesthesia), administration of moderate or potent analgesics, placement of a wick, and instillation of 2% white vinegar drops for 1 or 2 days, followed by the use of 0.3% fluoroquinolone drops for 7 to 10 days. In severe cases, it may be necessary to prescribe a short course of a fluoroquinolone orally (not approved by the Food and Drug Administration for this purpose in children). Individuals with moderate or severe otitis externa should avoid swimming until inflammation subsides. When cleansing and topical otic antibiotic drops fail to improve signs and symptoms, fungal infections should be considered. A Gram stain of exudate may reveal the hyphae of A. niger or the budding yeast of Candida spp. If fungal otitis externa is diagnosed, topical otic drops containing an imidazole agent, amphotericin B, m-cresyl acetate (25%), or aqueous gentian violet (1% or 2%) are usually effective.13 Furunculosis and local cellulitis respond best to the application of warm compresses and an orally administered antibiotic directed against Staphylococcus aureus. If an oral antibiotic is prescribed for

232

SECTION

C Upper Respiratory Tract and Oral Infections

gram-positive aural skin infection, culture should be performed and community-acquired methicillin-resistent S. aureus should be considered.

Prevention Swimmers who are prone to recurrent otitis externa may benefit from thorough drying of the ear canals after bathing, such as by using a hairdryer. The prophylactic instillation of 2% white vinegar solution, prescription ear preparations (e.g., Domeboro Otic, VoSol Otic), or over-the-counter swimmer’s eardrops is considered; 2% acetic acid acidifies the environment of the canal and reduces Pseudomonas colonization.13 Effectiveness of earplugs is controversial.

MALIGNANT OTITIS EXTERNA Malignant otitis externa is the result of invasive infection of the cartilage and bone of the external ear canal. It is uncommon in children and occurs characteristically in association with acquired immunodeficiency syndrome (AIDS), neutropenia, leukemia, immunosuppression related to organ transplantation, or insulindependent diabetes mellitus.15–18 Malignant otitis externa can be the presenting abnormality of a malignancy, such as rhabdomyosarcoma or leukemia.15 The diagnosis is suspected when otitis externa progresses to involve soft tissues of the face and scalp with necrosis, or when neurologic abnormalities manifest. Aggressive operative debridement and retrieval of deep specimens for microbiologic evaluation are essential. Complications include facial nerve palsy, osteomyelitis of the temporal bone, extension to the central nervous system, and death. The 80% mortality rate in the 1950s has been dramatically reduced to 5% with aggressive diagnosis and surgical debridement, antimicrobial therapy, and management of underlying conditions. Empiric parenteral antibiotic therapy is directed against P. aeruginosa and invasive fungi, especially Aspergillus spp. in certain circumstances;16 therapy is subsequently altered by results of culture of tissue specimens. Irrigation and daily meticulous debridement are critical for control. Fluoroquinolone agent or ceftazidime, administered intravenously at least initially, is usually given for 4 to 6 weeks.17 Some adults have been treated successfully with a fluoroquinolone orally.18 Inadequate debridement therapy has been especially associated with relapse of infection. Aspergillus infection has excessive mortality; voriconazole, amphotericin B therapy, and caspofungin are drugs of choice; prolonged course and combination therapy are required.19

CHAPTER

33

Mastoiditis Ellen R. Wald

Pathogenesis A knowledge of the anatomy and physiology of the middle ear and mastoid is essential to an understanding of the clinical presentation of mastoiditis and its complications. Figure 33-1 demonstrates the relationship between the eustachian tube, middle ear, and mastoid. At birth, the mastoid consists of a single cell, the antrum, connected to the middle ear by a narrow channel, the aditus ad antrum.6 As the child grows, the mastoid bone becomes extensively pneumatized, resulting in a series of interconnected air cells. The whole system is lined by modified respiratory epithelium. When AOM develops as a result of eustachian tube dysfunction, there is an acute inflammatory response of the mucosa lining the middle ear and, in many cases, the mucosa lining the mastoid as well.7 Almost all episodes of AOM respond to antibiotic therapy; eustachian tube dysfunction resolves, and the respiratory mucosa of the middle ear and mastoid recovers. In rare cases of newly diagnosed AOM or in occasional cases of inadequately or inappropriately treated AOM, inflammation of the middle ear and mastoid persists. There is an accumulation of serous and then purulent material within the mastoid cavities. As the pressure increases, the thin bony septa between air cells may be destroyed – so-called acute coalescent mastoiditis.6 This may be followed by the formation of abscess cavities and ultimately by the dissection of pus into adjacent areas. The direction in which purulent material dissects determines the clinical presentation and complications associated with acute mastoiditis (Box 33-1). Pus traversing the aditus ad antrum is delivered to the middle ear and empties either through the eustachian tube (with resolution of the process) or through the tympanic Aditus Antrum Tegmen

Epitympanum Eustachian tube Middle ear (mesotympanum) Hypotympanic air cells

gular bulb Ju Post. canal wall

Bone covering sigmoid sinus Figure 33-1. Schematic representation of the anatomy of the middle ear and mastoid air cell system. The aditus ad antrum is the narrow connection between the two; it may be a site of obstruction inhibiting drainage into the middle ear. (From Bluestone CD, Klein JO. Intratemporal complications and sequelae of otitis media. In: Bluestone CD, Stool SE, Kenna M (eds) Pediatric Otolaryngology, 3rd ed. Philadelphia, WB Saunders, 1996, pp 583–647.) Mastoid air cells

BOX 33-1 Complications of Mastoiditis

ACUTE MASTOIDITIS Acute mastoiditis is exclusively a complication of acute otitis media (AOM). Previously, only a third of cases occurred in the context of a first episode of otitis media,1 but in later reports, acute mastoiditis has been the first evidence of otitis media in at least 50% of children.2,3 The incidence of mastoiditis declined remarkably after the introduction of antibiotics, with one or two cases occurring annually in a children’s hospital or in an otolaryngology practice. However, frequent intracranial and extracranial complications are observed in the cases of acute mastoiditis that do occur.3–5

EXTRACRANIAL Subperiosteal abscess Bezold abscess Facial nerve paralysis Osteomyelitis Deafness Labyrinthitis INTRACRANIAL Meningitis Temporal lobe or cerebellar abscess Epidural empyema Subdural empyema Venous sinus thrombosis

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Mastoid tip

Mastoiditis

membrane by perforation. If the pus erodes the lateral cortex of the mastoid, a subperiosteal abscess is produced. The abscess results in swelling or fluctuation either above the superior portion of the auricle in infants (when erosion comes from the zygomatic mast cells) or behind the lower portion of the earlobe over the mastoid process in older children.4 Rarely, erosion occurs through the medial aspect of the mastoid tip, resulting in a neck abscess beneath the attachment of the sternocleidomastoid and digastric muscles, referred to as a Bezold abscess.8 Pus can dissect medially to the petrous air cells, resulting in petrositis, or posteriorly to the occipital bone, resulting in osteomyelitis of the calvarium or a Citilli abscess.6 Purulent material can spread to the labyrinth and facial nerve. Finally, pus within the mastoid can dissect toward the inner cortical bone, causing one or more suppurative complications in the central nervous system, such as meningitis, epidural abscess, subdural abscess, temporal lobe or cerebellar abscess, and venous sinus thrombosis.

33

233

Clinical Presentation The clinical presentation of the patient with mastoiditis depends on two features: (1) the age of the patient; and (2) the stage of the osteitis (i.e., whether it is uncomplicated or has already evolved to a subperiosteal abscess). Uncomplicated infection in a child younger than 2 years manifests as fever, otalgia (often manifesting as irritability), retroauricular pain, swelling, erythema, and a downward and outward deviation of the auricle, as shown in Figure 33-2A and B.4 In most cases, otorrhea or a bulging, immobile, opaque tympanic membrane is observed (Figure 33-2C). In addition, there may be sagging of the posterosuperior wall of the external auditory canal. In rare instances, with obstruction at the aditus ad antrum, middle-ear infection clears via the eustachian tube but the mastoid, unable to drain, continues to suppurate.

B

A

C

CHAPTER

D

Figure 33-2. A 13-month-old toddler with a 10-day history of acute otitis media that was unresponsive to amoxicillin therapy. (A and B) There is erythema and edema above the left ear with downward, outward, and anterior displacement of the pinna. (C) Tympanic membrane is opaque and bulging. (D) Axial computed tomographic scan without contrast medium shows fluid surrounding ossicles and destruction of the cortex of the temporal bone, with soft-tissue abscess and marked soft-tissue swelling. Peptostreptococcus and Fusobacterium spp. were isolated from the abscess. (Courtesy of E. Deutsch, MD, St. Christopher’s Hospital for Children, Philadelphia, PA.)

234

SECTION

C Upper Respiratory Tract and Oral Infections

In children older than 2 years, the pinna is usually deviated upward and outward, because the inflammatory process frequently concentrates over the mastoid process. In a patient of any age, when subperiosteal pus has accumulated, a fluctuant, erythematous, and tender mass can be found overlying the mastoid bone. Considerable attention has been given in the past two decades to the entity “masked mastoiditis.”9 In this presentation, the patient typically has had persistent middle-ear effusion or recurrent episodes of AOM without sufÀcient antimicrobial therapy. In either case, there is low-grade but persistent infection in the middle ear and mastoid with osteitis. Masked mastoiditis can manifest as fever, otalgia, and an abnormal tympanic membrane or, occasionally, without classic signs of AOM or mastoiditis but with an extracranial or intracranial complication.

Microbiology The only relatively recent prospective study of the microbiology of acute mastoiditis in children was published in 1987.10 In part, precise delineation of the microbiology has been hampered by the common circumstance of antibiotic therapy before clinical presentation. Most of the existing series published in the antibiotic era are retrospective reviews.1–4,9,11–17 Clinical specimens should be obtained from the middle ear either by tympanocentesis through an intact eardrum or by aspiration through a tympanostomy tube or perforation after careful sterilization of the surrounding structures. Cultures obtained from the external canal without careful sterilization may be contaminated with Pseudomonas aeruginosa or Staphylococcus aureus. Culture of the cerebrospinal fluid is valuable when bacterial meningitis complicates the clinical picture. Blood cultures are rarely positive. Percutaneous aspiration of a subperiosteal abscess should be undertaken, particularly if deÀnitive surgery will be delayed. Any material from abscesses (brain, epidural, or subdural) that are drained or aspirated should be submitted for Gram stain and aerobic and anaerobic cultures. In contrast to the usual pathogens that cause AOM (i.e., Streptococcus pneumoniae, Haemophilus influenzae, and Moraxella catarrhalis),18 the bacterial species most often implicated in acute mastoiditis are S. pneumoniae, S. pyogenes, P. aeruginosa, and

TABLE 33-1. Bacteriology of 831 Cases of Acute Mastoiditis in Children (1955–2005)a Bacterial Species Streptococcus pneumoniae

Number

Percent

123

36

Streptococcus pyogenes

68

20

Staphylococcus aureus

37

11

Other gram-positive cocci b

15

4

Haemophilus influenzae

22

6

Pseudomonas aeruginosa

48

14

Other gram-negative bacilli c

20

6

7

2

Anaerobic bacteria d Mycobacterium tuberculosis Total isolates a

3

1

343

100

Data from references 1–4, 11–17; many patients received antimicrobial therapy before specimens were obtained for culture. Isolates were obtained from blood, myringotomy aspirate, external auricular drainage, subperiosteal aspirate, cerebrospinal fluid, or surgical specimens. b Viridans streptococci (3), Enterococcus (1). c Citrobacter freundii, Escherichia coli, Enterobacter sp., Acinetobacter sp., Klebsiella sp. d Bacteroides spp. (4), Fusobacterium nucleatum (2), microaerophilic streptococci (1).

Staphylococcus aureus (Table 33-1). P. aeruginosa is probably a true pathogen in acute mastoiditis less often than it is isolated in some clinical series. Exaggerated rates of recovery may reflect the method of retrieving the sample for culture from the external canal.17 Except for a single prospective study in which anaerobic bacteria predominated,10 H. influenzae, Enterobacteriaceae, and anaerobic organisms are only recovered occasionally. Although all the patients in that study were described as having acute mastoiditis, 37% had a history of otorrhea, which implies a chronic etiology.10

Diagnosis The diagnosis of mastoiditis is usually made clinically, without need for imaging studies (Box 33-2). If plain radiographs are obtained, they most often show clouding of the mastoid or coalescence of mastoid air cells (i.e., dissolution of the thin bony septa from increased pressure and ischemia). Clouding of the mastoid is not diagnostic of acute mastoiditis; it is observed in at least 50% of patients with uncomplicated AOM. Although coalescent mastoiditis is a diagnostic radiographic Ànding, it is seen in the minority of patients. There are well-documented cases of acute and even complicated mastoiditis in which plain radiographs are reported to be normal.1,19 Computed tomography (CT) can be helpful in the diagnosis of mastoiditis, particularly when there are intracranial complications or in patients suspected of having masked mastoiditis.8,17 Evidence of mastoiditis on CT scan consists of: (1) haziness or destruction of the mastoid outline; and (2) loss of or decrease in the sharpness of the bony septa that deÀne the mastoid air cells (Figure 33-3).6 Lytic lesions of the temporal bone and soft-tissue abscess can sometimes be seen (Figure 33-2D). Cloudiness in areas normally pneumatized, which may be seen in uncomplicated AOM, is not diagnostic (Figure 33-4). In order for bony destruction to be appreciated radiographically, 30% to 50% of bone must be demineralized. Accordingly, in cases in which the CT scan shows cloudiness, a technetium-99 bone scan, which is more sensitive to osteolytic changes, may be useful.20 CT with enhancement or magnetic resonance imaging (MRI) is recommended in patients with suspected vascular thrombosis as a complication of mastoiditis. MRI with gadolinium enhancement conÀrms the diagnosis and delineates the extent of suspected disease21 because of its higher sensitivity for detection of extraaxial fluid collections and associated vascular problems.22

Complications Complications associated with mastoiditis are those related to the spread of infection or inflammation from the middle ear or mastoid to contiguous structures and occur in 7% to 16% of cases.13,17 Suppurative labyrinthitis (resulting in deafness), Bezold abscess, and cranial osteomyelitis are infectious complications of mastoiditis. Facial paralysis can result from infection or from compression by inflammation of the facial nerve as it traverses the narrow canal in the petrous bone. If infection spreads to the petrous bone, Gradenigo syndrome can follow, as characterized by sixth-nerve palsy, eye pain, and otorrhea. Intracranial complications include epidural or subdural empyema, cerebellar abscess, temporal lobe abscess, meningitis, and venous thrombosis.

BOX 33-2 Diagnosis of Acute Mastoiditis • Fever, otalgia, postauricular swelling + redness Older child: ear up and out Infant: ear down and out • Tympanic membrane: acute otitis media • Radiograph: mastoid air cells coalescent or clouded • Computed tomography, magnetic resonance imaging, or bone scan as needed

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Mastoiditis

Figure 33-3. Computed tomographic scan of the temporal bones showing acute coalescent mastoiditis (destruction of septa between the mastoid air cells) on the left and normal mastoid on the right.37

CHAPTER

33

235

there is an intact tympanic membrane in order to ascertain the etiologic agent and its sensitivity pattern.6) The same treatment is appropriate when facial paralysis occurs as an isolated complication of AOM or mastoiditis.6 If the child does not improve in 48 hours (as evidenced by diminution in systemic and local findings), a simple mastoidectomy is required. Simple mastoidectomy in combination with antimicrobial therapy and tympanostomy tube placement is usually indicated as initial management when a subperiosteal abscess is noted at presentation. Radical mastoidectomy is only performed when there is no clinical response to simple mastoidectomy, as evidenced by continued otorrhea.6 Several different agents are appropriate for intravenous treatment of acute mastoiditis. Ceftazidime, 150 mg/kg per day in divided doses every 8 hours, or piperacillin-tazobactam, 200 mg/kg per day piperacillin component in divided doses every 6 hours (not approved for use for such infections in patients younger than 12 years), is suitable. In a penicillin-allergic patient, clindamycin, at a dose of 40 mg/kg per day in divided doses every 6 hours plus aztreonam (120 mg/kg per day in divided doses every 8 hours) will provide coverage for both gram-positive and gram-negative pathogens. If Gram stain of aspirated material demonstrates an unexpected finding, additional antimicrobial agents are considered (such as for brain abscess). When results of culture and susceptibility tests are available, therapy is adjusted precisely. Intravenous treatment should be maintained for at least 7 to 10 days. If the clinical response to treatment has been satisfactory, oral antimicrobial therapy can be substituted for an additional 3 weeks to complete a 4-week course.

CHRONIC MASTOIDITIS Pathogenesis Chronic mastoiditis is almost always a result of chronic suppurative otitis media (CSOM)6 and, less frequently, the result of inadequate management of acute mastoiditis. CSOM is a clinical syndrome characterized by long-standing (> 3 weeks), painless otorrhea through a nonintact tympanic membrane (either a spontaneous perforation or a patent tympanostomy tube) that is not responsive to the antibiotics customarily prescribed for patients with AOM (Box 33-3).6,23 (See Chapter 32, Otitis Externa and Malignant Otitis Externa.) A nonintact state of the tympanic membrane is essential in pathogenesis, because it gives microbial species that usually colonize the external auditory canal access to the middle ear and, ultimately, the mastoid. These organisms cause a low-grade, persistent inflammatory response that is typically unaltered by conventional therapeutic agents for AOM. Although the inflammatory process is painless and does not initiate a febrile response, it can lead to chronic osteitis of the mastoid and complications such as hearing loss, cholesteatoma, facial nerve paralysis, meningitis, and brain abscess. The microbiology of CSOM and, hence, chronic mastoiditis has been amply documented during the past decade (Table 33-2).6,24–28 P. aeruginosa is most common, followed by gram-negative enteric

BOX 33-3 Diagnosis of Acute Mastoiditis Figure 33-4. Computed tomographic scan of the temporal bone of a child with acute otitis media, showing clouding of the mastoid on the left. The patient had no clinical evidence of mastoiditis.37

Management Treatment of uncomplicated mastoiditis or masked mastoiditis without intracranial complications involves intravenous antimicrobial therapy and myringotomy with or without placement of a tympanostomy tube. (At an absolute minimum, tympanocentesis should be performed if

• Otorrhea ≥ 3 weeks • Nonintact tympanic membrane (perforation or patent tympanostomy tube) • Nonresponsive to usual antibiotics • Etiology Pseudomonads Staphylococci Other gram-negative bacilli • Complications Cholesteatoma Progressive mastoiditis Deafness (8th nerve or ossicular)

236

SECTION

C Upper Respiratory Tract and Oral Infections

TABLE 33-2. Bacteriology of Chronic Suppurative Otitis Media in 200 Children (1983–1992) Bacterial Species Pseudomonas spp. Pseudomonas aeruginosa Other species Enteric gram-negative rods Escherichia coli Enterobacter spp. Proteus spp. Others Staphylococcus spp. Staphylococcus aureus Staphylococcus epidermidis Streptococcus pneumoniae Haemophilus influenzae Others

Number 176 171 5 50 8 9 17 16 41 38 3 10 10 19

Percent 58 16

13 3 3 6

Data from references 24 to 28.

bacilli and Staphylococcus aureus. P. aeruginosa and S. aureus occur together frequently.6,24,29 With the exception of a single study, anaerobic flora such as anaerobic gram-positive cocci (Peptococcus and Peptostreptococcus spp.), Fusobacterium spp., and Bacteroides spp. are isolated only occasionally.30 Rarely, Streptococcus pneumoniae or H. influenzae is causative. Mycobacterium tuberculosis was a common cause of mastoiditis, manifesting as a chronically draining ear, in the first half of the 20th century. More recently, this organism has been recovered only rarely from children or adults with long-standing otorrhea.31 Nontuberculous mycobacteria and M. bovis can produce an identical clinical syndrome. The latter species, not easily distinguished from M. tuberculosis, should be considered in patients who ingest unpasteurized dairy products.32

Diagnosis The diagnosis of CSOM is usually based on the typical clinical presentation of painless otorrhea unresponsive to conventional antibiotic therapy. The patient with CSOM is readily distinguished from the patient with a spontaneous perforation due to AOM, in whom constitutional and local symptoms are prominent and response to antimicrobial management is brisk. Distinction must also be made from the child with otorrhea due to otitis externa, in whom pressure on or movement of the tragus causes great discomfort. Also, in patients with otitis externa, the tympanic membrane is normal, whereas in patients with AOM or acute mastoiditis, it is acutely inflamed or perforated. An otomicroscopic examination of the tympanic membrane should be undertaken by the otolaryngologist in the initial evaluation of the patient with chronic otorrhea.29 This procedure, which may require general anesthesia, provides an opportunity to detect cholesteatoma, retraction pocket, granulation tissue, polyp, or foreign body. A specimen should be obtained from the middle-ear cavity without contamination from the external auditory canal and sent to the laboratory for Gram and acid-fast stains and aerobic, anaerobic, and mycobacterial cultures. A biopsy should be performed of any suspicious tissue seen protruding through the perforation or tympanostomy tube, because rhabdomyosarcoma and neuroblastoma can manifest as CSOM or chronic mastoiditis, usually with associated cranial nerve palsies. 5-Tuberculin units (TU) purified protein derivative (PPD) tuberculin skin test should be performed on patients whose CSOM does not respond to standard antimicrobial therapy.

Management

Failure of response to oral antibiotics is important in defining the CSOM syndrome and indicating the likelihood of P. aeruginosa as the probable cause. Treatment of patients with CSOM is often initiated with topical antimicrobial agents and vigorous aural toilet. The latter consists of daily suctioning of the external auditory canal to permit delivery of topical therapy. When the otorrhea is copious, ototopical therapy cannot be effective unless the external canal is cleared of purulent material. Topical therapy is chosen because the bacteria cultured in specimens from patients with CSOM are not usually susceptible to commonly prescribed oral antibiotics. Ofloxacin otic solution (0.3%), a topical quinolone approved for use in children by the United States Food and Drug Administration (FDA) in June 1999, has excellent activity against virtually all microbial species likely to be found in patients with CSOM.23,33 Importantly, it has shown no ototoxic effects when administered to adults or children.34 A Cochrane database of systematic reviews has recently been performed to assess topical antibiotics without steroids for patients with chronically discharging ears with underlying eardrum perforations. Topical quinolones were significantly better for cure of CSOM than antiseptics and no drug. Studies were inconclusive regarding any differences between quinolone and nonquinolone antibiotics.35 The aural discharge frequently clears within several days to 1 week after appropriate topical therapy is started. If the CSOM does not respond to topical therapy, a course of antimicrobial therapy intravenously is appropriate. Comparison of regimens for the treatment of CSOM is confounded by differences in patient selection, distribution of causative agents, use of topical agents in conjunction with systemic therapy, frequency of aural toilet, and use of prophylactic therapy after completion of initial therapy. An appropriate choice for intravenous therapy is an antipseudomonal penicillin with or without a b-lactamase inhibitor (e.g., piperacillin and tazobactam at 240 mg/kg per day in divided doses every 6 hours; not approved for use in children younger than 12 years) or a third- or fourth-generation cephalosporin with antipseudomonal activity (e.g., ceftazidime at 150 mg/kg per day in divided doses every 8 hours or cefipime at 50 mg/kg in divided doses every 12 hours, respectively).23 Daily aural toilet is instituted for debridement and to assess cessation of drainage. In general, intravenous therapy should be continued at least 1 week after drainage has ceased, because the middle-ear mucosa is still likely to be inflamed, and immediate relapse is possible.36 At the end of intravenous therapy, antimicrobial prophylaxis with amoxicillin for several months should be considered.23 If the otorrhea persists despite parenteral therapy or recurs shortly after the cessation of intravenous therapy, a simple tympanomastoidectomy is required. Several investigators have reported treating CSOM in children with intravenous or intramuscular antibiotics on an outpatient basis. Ceftazidime at 100 mg/kg per day, divided into 2 doses every 12 hours, with daily return of the patient to the otolaryngologist for aural toilet, was effective.28

CHAPTER

34

Sinusitis Ellen R. Wald

Upper respiratory tract infections (URIs) are the most common medical condition evaluated by the primary practitioner who cares for children. It has been estimated that approximately 5% to 10% of URIs in early childhood are complicated by acute bacterial sinusitis (ABS).1 Children average 6 to 8 colds per year; accordingly, sinusitis is a very common problem in clinical practice.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Sinusitis

CHAPTER

34

PATHOGENESIS

TABLE 34-1. Factors Predisposing to Sinus Ostial Obstruction

The respiratory mucosa that lines the nose is continuous with the mucosa that lines the paranasal sinuses. Secretions produced within the sinus cavities are delivered via the sinus ostia to the nose by normal mucociliary function. The maxillary, anterior ethmoid, and frontal sinuses drain to the middle meatus; the sphenoid and posterior ethmoid sinuses drain to the superior meatus. The three key elements important to the normal physiology of the paranasal sinuses are: (1) patency of the ostia; (2) function of the ciliary apparatus; and, integral to the latter (3) quality of the secretions.2 Retention of secretions in the paranasal sinuses is usually due to one or more of the following conditions: obstruction of the ostia, reduction in number of or impairment in the function of the cilia, and overproduction of or change in the viscosity of secretions. Most cases of ABS are thought to be bacterial complications of viral URIs. The viral infection affects the mucosa of the nose, causing rhinitis, and often the mucosa of the sinuses as well. In most instances, the inflammatory response subsides spontaneously. In some cases, however, mucositis results in obstruction of the sinus ostia, impairment of the mucociliary apparatus, and alteration in the volume and quality of secretions. The narrow caliber of the individual ostia that drain the maxillary and ethmoid sinuses sets the stage for obstruction to occur easily and often during the course of a viral URI. When obstruction does occur, there is a transient increase in intrasinal pressure. This is quickly followed by the development of a negative pressure within the sinus cavities as the oxygen component of the intrasinus air is rapidly absorbed by a metabolically active mucosa. When the pressure in the sinuses is negative relative to normal atmospheric pressure in the nose, conditions favor aspiration of mucus heavily laden with bacteria from the nose or nasopharynx into the presumably sterile paranasal sinuses. Under ordinary circumstances, these contaminating bacteria would be swept out again by normal ciliary function. However, when ciliary function is impaired and sinus ostia are obstructed, bacteria multiply to high density and initiate an intense inflammatory response. Alternatively, sneezing, sniffling, and nose-blowing with altered intrasinus pressure also facilitate bacterial contamination of the paranasal sinuses. Intranasal pressures are extremely high during nose-blowing, and nasal fluid has been shown to enter the maxillary sinus at that time.3 The factors predisposing to ostial obstruction can be divided into those that cause mucosal swelling, consequent to either systemic illness or local insults, and those due to mechanical obstruction (Table 34-1). Although many conditions can lead to ostial closure, viral URI and allergic inflammation are by far the most common and most important.

Mucosal Swelling

CLINICAL MANIFESTATIONS Viral Upper Respiratory Tract Infections To develop criteria to be used in distinguishing episodes of ABS from other common respiratory infections, it is helpful to describe an uncomplicated viral URI (depicted schematically in Figure 34-1). The course of most uncomplicated viral URIs is 5 to 10 days.4 Although the patient may not be free of symptoms on the 10th day, almost always the respiratory symptoms have peaked in severity on days 3 to 6 and begun to improve. Most patients with uncomplicated viral URIs do not have fever. However, if fever is present, it tends to be present early in the illness, often in concert with other constitutional symptoms such as headache and myalgias. Typically, the fever and constitutional symptoms disappear in the first 24 to 48 hours and the respiratory symptoms become more prominent. Viral URIs are usually characterized by nasal symptoms (discharge and congestion/obstruction) or cough, or both. Patients may also complain of a scratchy throat. Usually the nasal discharge begins as clear and watery. Often, however, the quality of nasal discharge changes during the course of the illness. Most typically, the nasal dis-

237

Mechanical Obstruction

SYSTEMIC DISORDER

Viral upper respiratory tract infection Allergic inflammation Cystic fibrosis Immune disorders Immotile cilia

Choanal atresia Deviated septum Nasal polyps Foreign body Tumor Ethmoid bulla

LOCAL INSULT

Facial trauma Swimming, diving Drug-induced rhinitis Gastroesophageal reflux

Severity

Respiratory symptoms Fever

0

1

2

3

4

5

6

7

8

9

10

11

12

Days Figure 34-1.

charge becomes thicker and more mucoid and may become purulent (thick, colored, and opaque) for several days. Then the situation reverses, with the purulent discharge becoming mucoid and then clear again, or simply drying. The transition from clear to purulent to clear again occurs in uncomplicated viral URIs without the use of antimicrobial therapy.

Clinical Presentations of Acute Bacterial Sinusitis With the clinical picture of an uncomplicated viral URI in mind, three clinical presentations of ABS can be described (Box 34-1).5 The most common clinical presentation is onset with persistent symptoms.6,7 The cardinal clinical features are nasal symptoms (anterior or posterior nasal discharge/nasal obstruction/congestion) or cough or both for more than 10 but fewer than 30 days which is not improving. This last qualifier is extremely important. Some individuals with uncomplicated viral URIs will still have residual respiratory symptoms at the 10-day mark. To be considered a sign of ABS, these respiratory symptoms must be persistent without improvement. The nasal discharge in patients with persistent symptoms may be of any quality, thick or thin, serous, mucoid, or purulent, and the cough, which may be wet or dry, must be present during the daytime, although it is often described as worse at night. Malodorous breath is often reported by parents of preschool children. Complaints of facial pain and headache are rare, although painless morning eye-swelling occurs occasionally. The child may not appear very ill, and usually, if fever is present, it is low-grade. It is not the severity of the clinical symptoms but their persistence that calls for attention. In clinical practice, strict observance of this presentation results in a diagnosis of acute sinusitis in 6.7% of children with upper respiratory tract symptoms.8

238

SECTION

C Upper Respiratory Tract and Oral Infections

BOX 34-1 Clinical Manifestations of Acute Sinusitis

BOX 34-2 Acute Sinusitis: Diagnostic Confirmation

• PERSISTENT SYMPTOMS Nasal discharge, cough, or both present > 10 days and not improving • SEVERE SYMPTOMS High fever (temperature ≥ 39°C) and purulent nasal discharge together for > 3 days • WORSENING SYMPTOMS Resolving upper respiratory symptoms Worsening on day 6 or 7 with new or recurrent fever and exacerbation of nasal symptoms and/or cough

PLAIN RADIOGRAPHS: FINDINGSa Complete opacification Mucosal thickening (≥ 4 mm) Air–fluid level COMPUTED TOMOGRAPHY IMAGES: INDICATIONS Recurrent problems Persistent infection Complicated infection (orbital or central nervous system) SINUS ASPIRATION: INDICATIONS Clinical failure Complicated infection Immunosuppressed patient

The second presentation is characterized as onset with severe symptoms. Severity is described as a combination of high fever, a temperature of at least 38.5oC and a particular quality of nasal discharge, a purulent nasal discharge, concurrently for at least 3 to 4 consecutive days.7 The presence of persistent fever for at last 3 to 4 days distinguishes this presentation from an uncomplicated viral URI (in which fever is present for less than 48 hours). The third presentation is described as worsening symptoms or presentation with a biphasic illness (in the Scandinavian literature, this is referred to as double sickening).5,9 This illness begins similarly to an uncomplicated viral URI from which the patient seems to be recovering. On the 6th or 7th day of illness, the patient becomes substantially worse again. The worsening symptoms may be manifest as an increase in respiratory symptoms (exacerbation of nasal discharge or nasal congestion or daytime cough) or a new onset of fever or a recurrence of fever if it had been present at the outset. Patients with subacute or chronic sinusitis have a history of protracted (> 30 days and not improving) respiratory symptoms. Nasal congestion (obstruction) and cough (day and night) are most common. Sore throat is a frequent complaint resulting from mouth-breathing secondary to nasal obstruction. Nasal discharge (of any quality) and headache are less common; fever is rare. On physical examination, the patient with ABS can have mucopurulent discharge in the nose or posterior pharynx. The nasal mucosa is usually erythematous but can occasionally be pale and boggy; the throat can show moderate injection. Examination of the tympanic membranes can show evidence of acute otitis media or otitis media with effusion; this occurs more often in chronic than acute sinusitis. The cervical lymph nodes are not usually signiÀcantly enlarged or tender. Occasionally, there is either tenderness, as the examiner palpates over or percusses the paranasal sinuses, or appreciable periorbital edema, with soft, nontender swelling of the upper and lower eyelid and discoloration of the overlying skin. Unfortunately, facial tenderness is neither a sensitive nor a speciÀc sign of sinusitis. Malodorous breath (in the absence of pharyngitis, poor dental hygiene, or a nasal foreign body) can suggest bacterial sinusitis. None of these characteristics, however, differentiates rhinitis from sinusitis.

a

Views are anteroposterior, lateral, and occipitomental.

Figure 34-2. An occipitomental radiograph delineating the maxillary sinuses. Substantial mucosal thickening (distance in millimeters between the bony border and the air–mucosa interface) is seen in the right maxillary sinus. An air–fluid level is seen in the left maxillary sinus.

DIAGNOSTIC METHODS When the clinical history suggests a diagnosis of ABS, the following procedures can help conÀrm the diagnosis (Box 34-2).

Imaging The use of radiographic imaging in the diagnosis of ABS in children is controversial.10 Standard radiographic projections are anteroposterior, lateral, and, for the maxillary sinuses, occipitomental views. Radiographic Àndings in patients with ABS are diffuse opaciÀcation, mucosal thickening of at least 4 mm, and presence of an air–fluid level (see Figure 34-2). Although these radiographic Àndings are not speciÀc for ABS, they are helpful in conÀrming the diagnosis of ABS in patients with suggestive signs and symptoms. Chronic sinusitis causes an osteoblastic response in the affected sinus walls. Accordingly, in addition to mucosal thickening and

complete opaciÀcation, sinus size can actually decrease as a result of greater thickness of the bony wall. Foci of irregular bone thinning can also be seen.11 Much has been written about the frequency of abnormal sinus radiographs in asymptomatic populations of children; however, most studies have been flawed by either inattention to the presence of symptoms and signs of respiratory inflammation or failure to classify abnormal radiographic Àndings as major or minor. When children older than 1 year do not have respiratory signs or symptoms, their sinus radiographs are almost always normal.12 On the other hand, in children with persistent or severe respiratory symptoms whose sinus radiographs demonstrate the presence of an air–fluid level, complete opaciÀcation of the sinus cavities, or mucosal thickening of at least 4 to 5 mm, 75% of maxillary sinus aspirates yield growth of bacteria in high density.13

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Sinusitis

When children < 6 years of age suspected to have ABS (on the basis of presentation with persistent symptoms) have sinus radiographs performed, radiographs are found to be significantly abnormal 88% of the time.14 Accordingly, for children < 6 years of age, since a positive history of persistent respiratory symptoms predicts the finding of abnormal sinus radiographs so frequently (nearly 90% of the time), and since history plus abnormal radiographs results in a positive sinus aspirate in 75% of cases, radiographs can be safely omitted and a diagnosis of ABS can be made on clinical criteria alone.7 In contrast to the general agreement that radiographs are not necessary in children < 6 years of age with persistent symptoms, the need for radiographs as a confirmatory test of ABS in children over 6 years of age with persistent symptoms and for all children (regardless of age) with severe or worsening symptoms is unsettled. The American College of Radiology has taken the position that the diagnosis of acute uncomplicated sinusitis should be made on clinical grounds alone.15 They support this position by noting that plain radiographs of the paranasal sinuses are technically difficult to perform, particularly in very young children. Correct positioning may be difficult to achieve and therefore the radiographic images may both over- and underestimate the presence of abnormalities within the paranasal sinuses. Similarly, a recent set of guidelines generated by the Sinus and Allergy Health partnership (representing numerous constituencies from within otolaryngology) does not recommend radiographs or computed tomographic (CT) or magnetic resonance imaging scans to diagnose uncomplicated cases of ABS.2 However, they do not provide evidence to support their position. To summarize, images appear to be unnecessary in children less than 6 years of age with persistent symptoms; however, despite their inconvenience and cost they may be necessary (to enhance accuracy of diagnosis) with all other presentations and all other age groups. Several studies have examined the frequency of incidental paranasal sinus abnormalities on CT scans of pediatric patients.16,17 One investigation has shown frequent abnormalities on CT scans in adult patients with a “fresh common cold.”18 These abnormalities indicate inflammation but not necessarily bacterial infection in many patients with respiratory symptoms.19 The ostiomeatal complex is the site of disease in patients with recurrent ABS or chronic sinusitis and is best imaged with CT. However, CT scans are not necessary for the management of the child with uncomplicated ABS; they should be reserved for the evaluation of: (1) complicated sinus disease (either orbital or central nervous system complications); (2) patients who experience numerous recurrences; and (3) protracted or nonresponsive symptoms (i.e., circumstances in which sinus surgery is contemplated).7

Sinus Aspiration Maxillary sinus aspiration can safely be performed by a skilled otolaryngologist in the ambulatory setting using a transnasal approach. Sedation or general anesthesia may be required for adequate immobilization in the young child. Current indications for maxillary sinus aspiration are: (1) lack of response of sinusitis to multiple courses of antibiotics; (2) severe facial pain; (3) orbital or intracranial complications; and (4) evaluation of an immunocompromised host. There must be careful decontamination and anesthesia of the area beneath the inferior turbinate through which the trocar is passed. Material aspirated from the maxillary sinus should be sent for Gram stain and quantitative aerobic and anaerobic cultures. The recovery of bacteria in a density of at least 104 colony-forming units (CFU) per milliliter is considered to represent true infection.20 The finding of at least one organism per high-power field on Gram stain of sinus secretions correlates with the recovery of bacteria in a density of 105 CFU/mL.

MICROBIOLOGY Data on the microbiology of sinusitis in pediatric patients are best organized according to the duration of clinical symptoms (Table 34-2).

CHAPTER

34

239

TABLE 34-2. Bacteriology of Sinusitis Bacterial Species

Acute (10–29 days)

Subacute (30–120 days)

Chronic (> 120 days)

Streptococcus pneumoniae

+++

++

+

Haemophilus influenzae

++

++

+

Moraxella catarrhalis

++

++

+

Staphylococcus aureus

+

Anaerobic bacteriaa

+

a

Respiratory anaerobic cocci, Bacteroides spp., Prevotella spp.,Veillonella spp. +++, most common; ++, common; +, less common.

However, literature review is complicated by varying definitions of acute, subacute, and chronic sinusitis. Several studies of ambulatory patients with acute (duration, 10 to 30 days) and subacute (30 to 120 days)14,21 illnesses have highlighted the important bacterial pathogens as Streptococcus pneumoniae, Haemophilus influenzae, and Moraxella catarrhalis. S. pneumoniae is most common in all age groups, accounting for 30% to 40% of isolates. H. influenzae and M. catarrhalis are similar in prevalence and each accounts for approximately 20% of cases. In the last decade there has been an increasing prevalence of penicillin-resistant S. pneumoniae and many of the H. influenzae (35% to 50%) and M. catarrhalis (55% to 100%) are beta-lactamaseproducing and also resistant to penicillin.22,23 Other less frequently recovered bacterial species are group A streptococcus, group C streptococcus, viridans streptococci, Peptostreptococcus spp., other Moraxella spp., and Eikenella corrodens.13 Neither staphylococci nor anaerobic respiratory flora are commonly recovered from patients with acute or subacute sinusitis. Respiratory viruses, including adenovirus, parainfluenza virus, influenza virus, and rhinovirus, are identified in approximately 10% of patients (both with and without bacterial species). This percentage might be higher if diagnostic aspirates were obtained earlier in the course of respiratory symptoms. Unfortunately, there are no data that have been generated regarding the microbiology of ABS in children since 1986.14 However, because of the similarity of the pathogenesis and microbiology of acute otitis media and ABS, it is acceptable to regard recent data generated from cultures of middle-ear fluid, obtained by tympanocentesis, from children with acute otitis media as a surrogate for cultures of the paranasal sinuses.24 Perhaps attributable in part to the near-universal use of pneumococcal conjugate vaccine in the United States, several recent reports have highlighted a slight decrease in isolates of S. pneumoniae and an increase in isolates of H. influenzae recovered from middle-ear aspirates.25,26 Presumably, these changes are occurring in the paranasal sinuses as well. In patients with very protracted (years) or severe sinus symptoms (requiring surgical intervention), Staphylococcus aureus and anaerobic organisms are recovered more frequently. Commonly recovered anaerobic bacteria are anaerobic gram-positive cocci (such as Peptococcus and Peptostreptococcus spp.) and Bacteroides or Prevotella spp.27 In addition, viridans streptococci and H. influenzae are frequently recovered.

MEDICAL MANAGEMENT Antimicrobial agents are mainstays in the medical management of sinusitis.7 Table 34-3 shows a list of agents potentially useful in patients with ABS. Amoxicillin is the treatment of choice for many cases of uncomplicated sinusitis in children. It is effective most of the time as well as inexpensive and safe. The last characteristic is particularly important when one is treating a condition such as sinusitis, that has a spontaneous cure rate of approximately 40%.14 Concern about penicillin-resistant Streptococcus pneumoniae (in children < 2

240

SECTION

C Upper Respiratory Tract and Oral Infections

TABLE 34-3. Antimicrobial Agents and Dosage Schedules for the

BOX 34-3 Major Complications of Sinusitis

Treatment of Sinusitis in Children Antimicrobial Agent

Dosage

Amoxicillin

45–90 mg/kg per day in 2 divided doses

Amoxicillin/potassium clavulanate

45/10 mg/kg per day in 2 divided doses

Amoxicillin/potassium clavulanate (high dose)

90/6.4 mg/kg per day in 2 divided doses

Cefdinir

14 mg/kg per day in 1 or 2 daily doses

Cefuroxime axetil

30 mg/kg per day in 2 divided doses

Cefprozil

30 mg/kg per day in 2 divided doses

Cefpodoxime proxetil

10 mg/kg once daily

Azithromycin

10 mg/kg per day on day 1; 5 mg/kg per day on days 2–5 in single daily doses

Clarithromycin

15 mg/kg per day in 2 divided doses

years, those attending group childcare, and those who have recently been treated with antimicrobial agent) should prompt the use of highdose amoxicillin (90 mg/kg per day).7 A broader-spectrum regimen is appropriate in the following clinical situations: (1) lack of symptom improvement with amoxicillin therapy; (2) residence in a geographic area with a high prevalence of b-lactamase-producing H. influenzae; (3) the occurrence of frontal or sphenoidal sinusitis; (4) the occurrence of complicated ethmoidal sinusitis; and (5) the presence of protracted (> 30 days’) symptoms. Antimicrobial agents with the most comprehensive coverage for patients with sinusitis are “high-dose” amoxicillin-potassium clavulanate (amoxicillin 90 mg/kg per day and clavulanate 6.4 mg/kg per day in 2 divided doses), cefuroxime axetil (30 mg/kg in 2 divided doses), and cefpodoxime proxetil (10 mg/kg once daily). For patients with chronic sinusitis, amoxicillin-potassium clavulanate is an attractive choice, especially because the mode of amoxicillin resistance in the pathogens causing chronic sinusitis is b-lactamase production. Other antimicrobial agents are available for the management of respiratory infections: cefdinir, cefprozil, clarithromycin, and azithromycin. Studies in patients with acute otitis media have shown these agents, in general, to be comparable in efficacy to older, more commonly used antimicrobial drugs. In general, clarithromycin and azithromycin should be reserved for children who have lower respiratory tract disease or are allergic to b-lactams. Infections due to penicillin-resistant pneumococci are a persistent problem. Organisms are classified as susceptible when the minimum inhibitory concentration (MIC) of the drug is < 0.1 μg/mL; moderately resistant when the MIC is between 0.1 and 1.0 μg/mL; and resistant when the MIC is >1 μg/mL. The frequency of penicillin-resistant pneumococci varies geographically, and many isolates of pneumococci are resistant to other commonly used antimicrobial agents, such as trimethoprim-sulfamethoxazole and erythromycin-sulfisoxazole. Therapeutic options include high-dose amoxicillin, an advanced-generation cephalosporin (for some pneumococci with moderate resistance), clindamycin, and rifampin. Selection of an agent should be guided by susceptibility results when available. Patients with ABS may require hospitalization because of systemic toxicity or inability to take antibiotics orally. These patients can be treated with intravenous cefotaxime, at 200 mg/kg per day in four divided doses, intravenous ceftriaxone at 50 mg/kg per day given once daily, or intravenous ampicillin-sulbactam, at 200 mg/kg per day of ampicillin component in 4 divided doses (not approved for such indications in children younger than 12 years). Clinical improvement is prompt in nearly all children treated with an appropriate antimicrobial agent. Patients febrile at the initial encounter become afebrile, and there is a remarkable reduction of nasal

ORBITAL Inflammatory edemaa Subperiosteal abscess Orbital abscess Orbital cellulitis Optic neuritis INTRACRANIAL Epidural empyema Subdural empyema Cavernous or sagittal sinus thrombosis Meningitis Brain abscess OSTEITIS Frontal (Pott puffy tumor) Maxillary a Inflammatory edema is not a true orbital complication of sinusitis. Infection is confined to the paranasal sinuses; periorbital swelling is due to impedance of venous blood flow.

discharge and cough within 48 hours. If the symptoms either do not improve or worsen in 48 hours, clinical re-evaluation is appropriate. If the diagnosis is unchanged, sinus aspiration is considered for precise bacteriologic identification. Alternatively, an antimicrobial agent effective against b-lactamase-producing bacterial species and penicillin-resistant pneumococci should be prescribed. The appropriate duration of antimicrobial therapy for patients with ABS has not been investigated systematically. For patients whose respiratory symptoms improve dramatically within 3 to 4 days of commencement of therapy, 10 days of treatment is adequate. For cases that respond more slowly, it is reasonable to treat until the patient is symptom free plus an additional 7 days. Adjuvant therapies, such as antihistamines, decongestants, and anti-inflammatory agents, have received little systematic evaluation. The overall impact of intranasal budesonide as an adjunct to oral antibiotic therapy for children with acute sinusitis is extremely modest and cannot be recommended.28 Some children experience recurrent or chronic episodes of sinusitis. The most common cause of recurrent sinusitis is recurrent viral URI, often a consequence of attendance at out-of-home childcare or the presence of a school-aged child in the household. Other predisposing conditions are allergic and nonallergic rhinitis, cystic fibrosis, an immunodeficiency disorder (insufficient or dysfunctional immunoglobulins), ciliary dyskinesia, gastroesophageal reflux, and an anatomic abnormality (see Table 34-1). Evaluation of children with recurrent or chronic sinusitis should include consideration of consultation with an allergist, performance of a sweat test, quantitative measurements of serum immunoglobulin levels, and a mucosal biopsy to assess ciliary function and structure. If specific allergens are identified or an allergic diathesis is documented, therapy might include desensitization, antihistamine, or topical intranasal steroid therapy. If a treatable immunodeficiency is identified, specific immunoglobulin therapy is initiated. Antimicrobial prophylaxis has not been studied in patients with recurrent acute sinusitis, although it has proved to be a useful strategy in reducing symptomatic episodes of acute otitis media in patients with frequent recurrences. Concerns regarding antibiotic resistance have discouraged recent use of antibiotic prophylaxis. If the sinusitis does not respond to maximal medical therapy, surgical intervention may be appropriate. The major complications of acute sinusitis are shown in Box 34-3. Subperiosteal abscess of the orbit and intracranial abscess are the most common. Signs of orbital infection are eyelid-swelling, proptosis, and impaired extraocular eye movements; with intracranial spread of infection; there may be signs of increased intracranial pressure, meningeal irritation, and focal neurologic signs. CT is essential for diagnosis. Antibiotic therapy and surgical drainage are usually required for successful treatment.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Bronchiolitis

CHAPTER

35

241

SURGICAL MANAGEMENT 4

Patients with acute sinusitis rarely require surgical intervention in the absence of orbital or central nervous system complications. An appreciation for the fact that a large group of pediatric patients with chronic sinusitis have underlying allergic rhinitis or other medical problems such as gastroesophageal reflux has led to more aggressive medical management and less enthusiasm for surgical solutions.29 Occasionally, sinus aspiration is required to ventilate a sinus in a patient with no response to aggressive antimicrobial therapy. In the rare child who has failed conventional treatment with oral antibiotics and an array of local and systemic adjuvants to therapy, antibiotic therapy delivered via a percutaneous intravenous catheter may be a worthwhile trial.30 This can be combined with adenoidectomy, or adenoidectomy may be undertaken either as a solo intervention or performed in conjunction with functional endoscopic surgery.31,32 When functional endoscopic sinus surgery is performed, the focus is on the ostiomeatal complex, highlighted in Figure 34-3. This is the area between the middle and inferior turbinates that represents the confluence of drainage areas of the frontal, ethmoid, and maxillary sinuses. In the ostiomeatal complex, there are several areas where two mucosal layers come into contact, predisposing to local impairment of mucociliary clearance. Using an endoscope, the natural meatus of the maxillary outflow tract is enlarged by excising the uncinate process and the ethmoid bullae and performing an anterior ethmoidectomy. Ramodan recently reported a group of 66 children undergoing either adenoidectomy or endoscopic surgery for chronic rhinosinusitis.31 The group undergoing endoscopic sinus surgery had the greatest improvement in overall symptom scores. As our understanding of the

3 2

6

9

5 1

8 7

Figure 34-3. Coronal section of the nose and paranasal sinuses. 1, Maxillary sinus; 2, ethmoidal bursa; 3, ethmoidal cells; 4, frontal sinus; 5, uncinate process; 6, middle turbinate; 7, inferior turbinate; 8, nasal septum; 9 (stippled area), osteomeatal complex. (From Wald ER. Sinusitis in children. N Engl J Med 1992;326:319–323.)

pathogenetic factors predisposing to sinusitis expands and our therapy of these mucosal diseases becomes more effective, the pediatric candidates for surgical management of their sinus disease will continue to decrease. Judicious use of endoscopic surgery in children younger than 5 years is strongly advised.33

SECTION

D

Lower Respiratory Tract Infections

CHAPTER

35

Bronchiolitis H. Cody Meissner

Bronchiolitis, a disease primarily of the first 2 years of life characterized by signs and symptoms of obstructive airway disease, is caused most commonly by viruses. Approximately 3% of infants in the first 12 months of life are hospitalized with bronchiolitis, accounting for more than 125,000 annual hospitalizations in the United States. Data from the Centers for Disease Control and Prevention (CDC) indicate that the number of yearly hospital admissions attributable to bronchiolitis increased more than twofold between 1980 and 1996. Increasing survival rates for premature infants as well as infants with compromised cardiac, pulmonary, and immune status increase the number of children at risk for severe bronchiolitis.

ETIOLOGIC AGENTS Many viruses may cause bronchiolitis, although respiratory syncytial virus (RSV), human metapneumovirus, and parainfluenza virus type 3 are the most common etiologic agents.1–5 Other viruses are implicated less frequently (Table 35-1).6–10 During the winter months, RSV is identified as the etiologic agent by cell culture or antigen detection assays in up to 80% of children hospitalized with bronchiolitis or pneumonia. Epidemics of bronchiolitis in early spring and fall are often caused by parainfluenza virus type 3.11,12 The yearly cycles of these respiratory viruses are depicted in Figure 35-1. Bordetella pertussis, Mycoplasma pneumoniae, measles, influenza, and adenovirus have been associated with a severe form of bronchiolitis, bronchiolitis obliterans.13–18 This uncommon obstructive pulmonary disease is characterized histologically by the progression of acute airway inflammation to necrosis of the cells lining the lumen with severe obliterative fibrosis in the final stages. The pathogenesis of bronchiolitis obliterans probably differs from that of simple viral bronchiolitis.

242

SECTION

D Lower Respiratory Tract Infections

EPIDEMIOLOGY Bronchiolitis may be defined as an episode of obstructive lower airway disease that is precipitated by a viral infection in infants younger than

TABLE 35-1. Infectious Causes of Bronchiolitis Infectious Agent

Occurrences

Respiratory syncytial virus

++++

Human metapneumovirus

++

Parainfluenza virus 3

++

Parainfluenza virus 1

+

Parainfluenza virus 2

+

Adenovirus

+

Influenza virus (A or B)

+

Mycoplasma pneumoniae

+

Enterovirus

+

Rhinovirus

+

Number of infections

++++, most common cause; ++, causes substantial percentage of cases in some studies; +, occasional cause. Relative importance varies with season and epidemic disease (see text). Data from references 7, 8, 11, 13, 15.

20 15 10 5

Respiratory syncytial virus

15 10 5

Parainfluenza virus type 1

10 5

Parainfluenza virus type 2

15 10 5

Parainfluenza virus type 3

24 months of age. The peak incidence of severe disease occurs between 2 and 6 months of age.2,4,19,20 Rates of hospitalization are higher in boys and among infants living in industrialized urban settings rather than in rural settings.21,22 Hospitalization rates are about 5 times higher among infants and children in high-risk groups than among nonhigh-risk infants. These groups include premature infants (< 35 weeks’ gestation), infants born with hemodynamically significant congenital heart disease, as well as infants with chronic lung disease of prematurity (previously called bronchopulmonary dysplasia).23–28 Although mortality has been reduced in recent years (a CDC study estimated about 500 deaths per year in the United States among children in the first year of life due to RSV infection), morbidity among high-risk patients can be high, with average hospital length of stay and intensity of care several times that of previously healthy infants.29 Occurrence of the respiratory virus season is predictable, even though the severity of the season, the date of onset, the peak of activity, and the end of the season cannot be predicted with precision. There can be variation in timing of community outbreaks of disease due to RSV from year to year in the same community and among neighboring communities, even in the same season. In the United States communities in the south tend to experience the earliest onset of RSV activity and the midwest tends to experience the latest onset.30 The duration of the season for the west and the northeast is typically between that in the south and the midwest. Nevertheless, these variations occur within the overall pattern of RSV outbreaks, usually beginning in November or December, peaking in January or February, and ending by the end of March or April.

Number of patients positive

A/Victoria 300

200

A/Texas

A/Port Chalmers B/Hong Kong

100 A/Brazil

50

10

20

30

40 50

10 20 30 40 50

10 20 30 40 50

10 20 30 40 50

B/Singapore

10 20 30 40 50

10

20

Week no. 1975

1976

1977

1978

1979

1980

Figure 35-1. Patterns of occurrence of respiratory syncytial virus and parainfluenza virus virus in Houston, Texas. (From Couch RB. Viral respiratory diseases. In: Stringfellow DA (ed) Virology. Kalamazoo, MI, Scope, 1983, p 65.) PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Bronchiolitis

Limited numbers of cases of bronchiolitis occur during summer and early fall, and they are more likely to be caused by viruses other than RSV, such as rhinovirus and parainfluenza virus. These cases are generally milder than are RSV-related cases. In tropical countries, the annual epidemic of RSV coincides with the rainy season or “winter,” although sporadic cases can occur throughout the year.9 RSV can be divided into A and B strains: A strains may be associated more commonly with epidemics, severe disease, and a higher hospitalization rate than B strains, although not all studies are consistent with regard to differences in severity.31–34 Both strains may circulate during the same season, and infants may be infected by both within the same year. A progressive increase in hospitalization rates for bronchiolitis in the United States has occurred since the late 1980s.1,22 This increase may be related to a greater ability to identify hypoxic infants through the use of pulse oximetry. Alternatively, the increase in severity of bronchiolitis may be related in some way to the national increase in severity of asthma.1 Household crowding is an important risk factor for severe viral lower respiratory tract illness due to RSV as well as other respiratory viruses.35 Generally it is recognized that as the number of household members increases, the risk of exposure to infectious respiratory secretions also increases. Childcare attendance has been correlated with an increased risk of bronchiolitis in some studies. Exposure to passive household tobacco smoke has not been associated with an increased risk of RSV hospitalization on a consistent basis. In contrast to the well-documented beneficial effect of breastfeeding against some viral illnesses, the existing data are conflicting regarding the specific protective effect of breastfeeding against RSV infection.36–40 Parental history of bronchiolitis or asthma is associated with a higher risk for the development of lower respiratory illness in offspring.41,42 Young age at the beginning of the RSV season is a consistent risk factor for RSV hospitalization. Several reasons may account for this increase in risk. Most severe RSV disease occurs in the first 6 months of life so that birth shortly before or early after the onset of the RSV season will result in a longer period of exposure to RSV earlier in life. Second, maternal antibody concentrations to RSV show seasonal variation and infants born early in the RSV season are more likely to be born to mothers with low serum antibody concentrations to the F (fusion) protein of RSV.43,44 Low concentrations of RSV antibody correlate with susceptibility to severe RSV disease in infants.

PATHOGENESIS AND PATHOLOGIC FINDINGS Acute bronchiolitis generally implies disease of infectious etiology, usually due to viruses with specific tropism for bronchiolar epithelium. Because most healthy infants recover from bronchiolitis without incident, information regarding the pathologic changes caused by infection is inferred from animal studies and from biopsy or autopsy materials in severe cases. Viral infection causes profound alterations in the epithelial cell and mucosal surfaces of the human respiratory tract. The earliest lesions, seen 18 to 24 hours after the onset of disease, consist of bronchiolar cell necrosis, ciliary disruption, and peribronchiolar infiltration with lymphocytes.45,46 Terminal bronchiolar epithelial cells are targets for viral infection and are damaged by direct viral invasion and the resulting host inflammatory response. Edema of the small airways and mucus secretion, coupled with sloughed epithelial cells, lead to airway obstruction with atelectasis.46 Functional residual capacity increases, and dynamic compliance falls.47–79 The physiology of airflow in the infant lung is an important factor predisposing to severe clinical illness during virus infection of the bronchioles. Small airways of young infants result in up to sevenfold greater total airway resistance compared with adults.50 Thus, airflow in infants becomes critically diminished with minimal inflammatory changes of the bronchioles. Narrow or poorly functioning airways early in life appear to predispose to wheezing. The presence of high serum concentrations of immunoglobulin (Ig)G antibodies to RSV (whether transplacentally acquired or intra-

CHAPTER

35

243

venously administered) ameliorates RSV illness.51–54 Severe obstructive illness may be related to stimulation of virus-specific IgEmediated hypersensitivity responses or altered cell-mediated immune responses.55–60 In contrast to atopic asthma, which is characterized by increased expression in the airway of the cytokines interleukin-4 and interleukin-5, virus-induced wheezing may be related to greater expression of interferon-g.61,62

CLINICAL MANIFESTATIONS Bronchiolitis represents the late stage of a respiratory disease that progresses over several days. Upper respiratory tract symptoms consisting of nasal discharge and mild cough begin about 3 to 5 days after exposure to infection. Approximately 30% to 40% of RSVinfected infants have progression of disease to involve the lower respiratory tract. Spread to the lower airway occurs either by aspiration of RSV-infected epithelial cells or by cell-to-cell spread. Lower-airway involvement is marked by a sudden increase in the work of breathing, cough, tachypnea, wheezing, crackles, use of accessory muscles, and nasal flaring.63,64 The respiratory rate often exceeds 60 to 70 breaths/minute in young infants, and expiration is prolonged. Intercostal and subcostal retractions with wheezing are evident. Initially, wheezing occurs during the expiratory phase only and is only audible through a stethoscope. As wheezing progresses, it can be heard without a stethoscope. The chest becomes hyperexpanded and hyperresonant, respirations more labored, and retractions more severe. Hypoxemia out of proportion to clinical distress is typical of RSV infection. Mild hypoxemia occurs even in otherwise well-appearing infants, the so-called happy wheezers. Respiratory failure can be due to hypoxemia (an early and sometimes sudden occurrence) or progressive hypercapnia due to fatigue. The small airways of young infants can become so narrowed that wheezing is inaudible. In this setting disease severity is recognized by the absence of audible air exchange, flaring of the alae nasae, expiratory grunting, severe subcostal, supraclavicular, and intercostal retractions, and hypoxemia. Progressive illness is often accompanied by a rapid fall in oxygen saturation after minimal manipulation. A child with these findings usually requires intubation and ventilatory support. Apnea can be an early manifestation of RSV infection, at times resulting in respiratory failure.65 RSV-related apnea is mediated by the central nervous system, occurring more commonly in very young, often prematurely born infants. Because the severity of bronchiolitis often waxes and wanes prior to consistent improvement, assessment of respiratory status can vary markedly over a short period. The ability of the young infant to breastor bottlefeed without distress over time often provides a practical guide to disease severity and management. An infant who has substantial difficulty feeding as a result of respiratory distress has moderate or severe illness and usually requires hospitalization. Otherwise healthy infants younger than 2 months of age, infants born prematurely (less than 35 weeks’ gestation), and infants with chronic lung disease of prematurity (previously called bronchopulmonary dysplasia) or infants born with congenital heart disease have the highest morbidity and mortality rates due to bronchiolitis.66,67 Infants born with congenital heart disease who are at greatest risk of hospitalization due to bronchiolitis include those with moderate to severe pulmonary hypertension and infants with cyanotic heart disease. Infants and children with the following hemodynamically insignificant heart disease are at increased risk of hospitalization: secundum atrial septal defect, small ventricular septal defects, pulmonic stenosis, uncomplicated aortic stenosis, mild coarctation of the aorta, and patent ductus arteriosus, as well as infants with lesions adequately corrected by surgery (unless they continue to require medication for management of congestive heart failure).68,69 Severe respiratory distress with bronchiolitis can be the presenting manifestation of previously unrecognized congenital heart disease. Once hospitalized, the infant may have a highly variable course of illness.70–72 Among otherwise healthy infants, intensive care unit admission because of respiratory deterioration is uncommon.73 A

244

SECTION

D Lower Respiratory Tract Infections

decision to admit to the intensive care unit is based on the possible need for intubation because of progressive hypercapnia, increasing hypoxemia despite supplemental oxygen, or apnea. The typical course for a previously healthy infant older than 6 months is one of improvement over 2 to 5 days, as evidenced by decreases in respiratory rate, retractions, duration of expiration, and oxygen requirement. Wheezing commonly persists for a week or more after discharge. Pulmonary function abnormalities and evidence of mild desaturation (oxygen saturations in the range of 93% to 95%) may persist for several weeks.74 The differential diagnosis of bronchiolitis includes airway hypersensitivity to environmental irritants, anatomic abnormality of the airway, cardiac disease with pulmonary edema, cystic fibrosis, foreign-body aspiration, and gastroesophageal reflux.

DIAGNOSIS The diagnosis of bronchiolitis is based on clinical criteria with supporting radiographic findings. Typical chest radiographic findings are hyperinflation, with flattening of the diaphragms and hyperlucency of the lungs, and patchy atelectasis, especially involving the right upper lobe (Figures 35-2 and 35-3).75,76 Atelectasis is due to airway narrowing or mucous plugging and is associated with volume loss; it may be confused with lobar consolidation or aspiration pneumonia, both of which are generally volume-expanding lesions. Bacterial pneumonia infrequently occurs as a complication of bronchiolitis but should be suspected in the infant with fever persisting for more than 2 to 3 days and lack of response to supportive management. Establishing a specific etiologic diagnosis is helpful in predicting the clinical course, in cohorting in the hospital, and may become increasingly useful as more antiviral agents effective against respiratory viruses become available. Although viral culture of respiratory secretions has been the “gold standard” for diagnosis of RSV infection, it often is too slow a method to be clinically useful. Enzyme immunoassays and direct fluorescent antibody (DFA) techniques for the identification of RSV, influenza virus, parainfluenza viruses, and adenoviruses permit rapid and accurate diagnoses.77–79 Nasal wash is the preferred method of specimen collection. The DFA test permits evaluation of adequacy of the specimen’s number of epithelial cells for antigen detection. A respiratory screening of nasal

Figure 35-2. Typical radiographic appearance of bronchiolitis. Hyperinflation, hyperlucency, and flattened diaphragms are most characteristic. Peribronchiolar infiltrates are common, and parenchymal infiltrates are less common. (Courtesy of Richard Heller, MD, Vanderbilt University Children’s Hospital, Nashville, TN.)

secretions using pooled monoclonal antibodies to the common agents of bronchiolitis, followed by specific identification for a positive reaction, is a cost- and time-saving procedure compared with standard tissue culture isolation. Amplification of virus using the shell vial method, followed by use of specific monoclonal agents, and amplification of viral genome by the polymerase chain reaction offer the promise of improved sensitivity for rapid detection but are not as widely available nor as rapid as enzyme immunoassay and fluorescent antibody techniques.80–83 A multitest system for quantitative reverse transcription–polymerase chain reaction–enzyme hybridization assay (Hexaplex) is available to test a single nasopharyngeal sample for RSV, influenza A and B viruses, parainfluenza virus, and adenoviruses.84 Antigen detection tests are useful in diagnosing certain viral infections, but, as with all tests, the positive predictive value decreases as disease incidence goes down. Sensitivity and specificity of antigen detection assays are lowest at both the onset and the end of the respiratory virus season.

MANAGEMENT General Measures Most infants with bronchiolitis can be managed at home with supportive care, but hypoxia or inability to feed adequately necessitate hospitalization. Once hospitalized, most infants respond to administration of supplemental oxygen and replacement of fluid deficits.84a The value of mist inhalation by vaporizer or tent is not proven; its use can provoke reflex bronchoconstriction. The specific treatment strategies used differ widely across children’s medical centers.72 Fewer than 10% of previously healthy infants hospitalized for bronchiolitis require intubation and mechanical ventilation because of respiratory failure or apnea; the percentage is higher for prematurely born infants and infants with chronic lung disease or congenital heart malformations.

Bronchodilator Therapy The therapeutic role of bronchodilator agents in bronchiolitis is controversial.85 Occasionally, a single administration of an aerosolized bronchodilator elicits a response, but this improvement is not seen in most infants with bronchiolitis and is not generally reproducible with subsequent doses.86–90 Modest improvement in clinical scores

Figure 35-3. Atelectasis, particularly of the right upper lobe, is a not uncommon feature of bronchiolitis. This should not be confused with pneumonia, which manifests as volume expansion, not volume loss. (Courtesy of Richard Heller, MD, Vanderbilt University Children’s Hospital, Nashville, TN.)

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Acute Pneumonia

and in tests of pulmonary function have been reported with use of inhaled racemic epinephrine91,92 and b-adrenergic agents, principally salbutamol and albuterol.93,94 However, clinical improvement following repeated doses of epinephrine is not sustained and favorable response to b-adrenergic agents, as measured by clinical score and oxygenation, is inconsistent.95–100 Flores & Horwitz101 performed a meta-analysis of eight studies with similar designs. Overall, their analysis supported a beneficial effect in certain infants, but identifying those infants could not be consistently accomplished at the time of initial presentation. On balance, an initial trial of bronchodilator therapy for the hospitalized infant with bronchiolitis is reasonable, although brief episodes of hypoxia can be precipitated by adrenergic agents. Bronchodilator therapy should only be continued if consistent improvement in respiratory distress or oxygen saturation is observed. Racemic epinephrine should not be continued beyond one or two doses.

Corticosteroid Therapy Although corticosteroids reduce the inflammatory changes observed with bronchiolitis, they may increase viral replication and prolong shedding. One small controlled clinical trial concluded that the combination of albuterol plus prednisolone was more effective than albuterol alone in accelerating recovery from bronchiolitis.102 Most studies examining the role of corticosteroids in bronchiolitis, however, have not demonstrated a consistent clinical benefit.103–111 Although one meta-analysis of previously published reports of corticosteroid use in bronchiolitis concluded that there may be slight improvements in duration of symptoms, length of hospital stay, and clinical scores, these benefits appear to be limited.112 Therefore, the routine use of corticosteroids in bronchiolitis is not recommended.

Antiviral Therapy Ribavirin is a nucleoside analogue with in vitro activity against RSV, adenovirus, influenza A and B viruses, and parainfluenza viruses. Early trials indicated that ribavirin therapy was associated with modest improvement in clinical scores, oxygenation, and duration of mechanical ventilation for infants with severe bronchiolitis due to RSV infection. These studies were challenged on the basis that control groups received water aerosols, which may produce bronchospasm in individuals with hyperreactive airways. The high cost of ribavirin prohibits its use in cases of mild illness, but the drug might be useful in carefully selected patients with life-threatening infection.113–115 Guidelines for the use of ribavirin in RSV disease are presented in Chapter 225, Respiratory Syncytial Virus. Several treatment options are now available for infection due to influenza (see Chapter 229, Influenza Viruses). Amantadine and rimantadine are active against influenza A and have shortened the duration of fever, clinical symptoms, and duration of viral shedding in children and adults with upper respiratory tract disease.116,117 Amantadine is recommended for children 1 year of age and older with severe influenza A illness, including bronchiolitis. The neuraminidase inhibitors oseltamivir and zanamivir have similar activity in the treatment of both influenza A and B infections. Oseltamivir is approved for treatment of influenza in children 1 year of age and older.118–121 Zanamivir is approved for children 7 years of age and older. Effectiveness of these agents in the treatment of bronchiolitis due to influenza virus is not known.

Immune Globulins and Other Therapies Antibody preparations containing high titers of neutralizing antibody against RSV as well as a preparation of monoclonal antibodies directed against one of the two major RSV surface glycoproteins (fusion glycoprotein) reduce the risk of hospitalization due to RSV infection.51–54 Used therapeutically, they result in more rapid clearing of virus from the respiratory tract but do not alter the course of illness and should not be used for the treatment of RSV infection.122–126 Although vitamin A levels have been demonstrated to be low in infants

CHAPTER

36

245

with RSV bronchiolitis, a therapeutic benefit of vitamin A therapy has not been demonstrated.127–129

PROGNOSIS, COMPLICATIONS, AND SEQUELAE Most otherwise healthy infants recover completely from acute bronchiolitis, although subtle pulmonary abnormalities may persist for weeks.38 An important question is whether bronchiolitis in infancy increases the likelihood of childhood asthma. Numerous studies have defined a higher risk of recurrent wheezing throughout childhood after bronchiolitis in infancy, and abnormalities of small-airway function have been identified in school-aged children with a history of bronchiolitis in infancy. However, each of these findings may simply be a reflection of hereditary tendencies that are expressed both at the time of bronchiolitis and upon allergen exposure in later childhood.130–134 Moreover, by adolescence, the rate of recurrent wheezing in subjects who had bronchiolitis in infancy appears to fall to the rate observed in subjects without a history of bronchiolitis.134 Thus, it is uncertain whether bronchiolitis is causally associated with long-term respiratory morbidity.

PREVENTION Strategies that reduce contact of vulnerable infants with individuals with respiratory tract infections, minimizing passive exposure to cigarette smoke, and limiting nosocomial transmission of causative agents offer immediate opportunities to reduce bronchiolitis morbidity. Monthly administration of monoclonal anti-F antibody (palivizumab) throughout the RSV season reduces the incidence of hospitalization due to RSV infection in infants with bronchopulmonary dysplasia, congenital heart disease, and prematurity by about 50% (see Chapter 225, Respiratory Syncytial Virus). The high cost and modest effect of palivizumab limit its use for passive immunoprophylaxis to the most medically fragile infants. No vaccine to prevent infection with RSV or parainfluenza viruses, the most common causes of bronchiolitis, is licensed or near licensure. Trivalent influenza vaccine is recommended for all infants older than 6 months and less than 24 months of age during the influenza season. Because this is not approved for use in infants younger than 6 months, routine influenza vaccination is important for family members and caregivers of these young patients. Potential RSV vaccine candidates currently being evaluated include inactivated preparations of the purified fusion protein of RSV, DNA vaccines coding for the major immunogenic proteins of the virus, and replicating mutants of the virus that replicate in the upper respiratory tract but are inactivated at the higher temperatures of the lung.135

CHAPTER

36

Acute Pneumonia and its Complications Chitra S. Mani and Dennis L. Murray

“Pneumonia” is a Greek word meaning “inflammation of the lungs.” This illness is frequently acquired from exposure in the community and is therefore called community-associated pneumonia (CAP); rarely, it is acquired during/or after hospitalization, when it is referred to as nosocomial or hospital-associated pneumonia (HAP). The onset can be acute or chronic. This chapter focuses on acute, communityacquired pneumonia and its complications in infants and children.

246

SECTION

D Lower Respiratory Tract Infections

ACUTE PNEUMONIA Acute pneumonia is a common infection affecting infants and children, capable of causing significant morbidity and mortality.1 The clinical presentations and agents of pneumonia in children and infants are different compared with adults. An accurate measurement of the worldwide incidence of childhood pneumonia is difficult because of the many ways of defining pneumonia. In this chapter, acute pneumonia is defined as an acute lower respiratory tract infection (LRTI) with fever (> 37.8°C), presence of lower respiratory tract signs, and radiologic evidence of a lung abnormality in one or both lungs.2 The estimated annual incidence of pneumonia in North America and Europe is 36 to 40 cases/1000 population in children younger than 5 years and is 11 to 16 cases/1000 in those 5 to 14 years of age.2 Every year, 1 to 4/1000 children are hospitalized in the United States with LRTI.3 The incidence of pneumonia is almost twofold higher in American Indian and Alaska native infants and children,4 and is 10- to 12-fold higher in the developing countries.5 In the developing world, pneumonia is the third leading cause of death in children, accounting for over 1.9 million deaths annually in children younger than 5 years.5,6 In developed countries, acute pneumonia is associated with morbidity, with low mortality.

Etiologic Agents and Epidemiology Multiple microbes cause LRTI in infants and children and establishing a microbial diagnosis is difficult. In two recent studies of immunocompetent hospitalized patients etiologic agents of pneumonia were confirmed in 79% to 85%.2,7 However, these investigations to detect the etiology involved performing multiple laboratory tests, some only available in research laboratories. Other studies have confirmed etiologic agents of pneumonia in a minority of children.8 For some organisms, particularly viruses, Mycoplasma, and Chlamydiaceae, microbial etiology is inferred by detection of microorganisms in the upper respiratory tract. For others, serologic analysis or nucleic acid assay (polymerase chain reaction (PCR)) is the preferred method. Pyogenic bacteria present the most difficult challenge because some pathogens frequently coexist with or are normal upper respiratory tract flora. Bacteremia confirms the cause but is present in only 1% to 10% of hospitalized children with bacterial pneumonia.2,7,9 In most cases of acute pneumonia, extensive or invasive testing is not warranted. Epidemiologic information is frequently useful in guiding differential diagnosis and management (see Chapter 23, Respiratory Tract Symptom Complexes). Certain pathogens, particularly respiratory syncytial virus (RSV), rhinoviruses, influenza viruses, and Mycoplasma, are seasonal. In other instances, the pattern of family illness provides a clue to the causative agent. For purposes of management, the relative importance of etiologic agents in series of patients who have been extensively evaluated is extrapolated to patients with similar clinical syndromes, physical findings, and laboratory results. Table 36-1 lists the common etiologic agents of acute pneumonia in children.

Neonates and Young Infants Pneumonia in neonates can manifest as early-onset disease (within 5 to 7 days of life), or late-onset disease after 7 days of life. Most infections in the first week of life are caused by organisms acquired from the maternal genital tract through aspiration of either infected amniotic fluid or genital secretions. Group B streptococcus is the most frequent cause of early-onset pneumonia.10 Group B streptococcus, Listeria monocytogenes, Escherichia coli, and other gram-negative bacilli can cause severe respiratory distress resembling hyaline membrane disease, usually as a part of a widespread systemic infection. Prenatal and perinatal risk factors, including preterm delivery, maternal chorioamnionitis, and prolonged rupture of membranes, increase the risk for development of neonatal pneumonia. Hematogenous dissemination can also occur from an infected mother.

Pneumonia due to Chlamydia trachomatis, which becomes symptomatic > 2 to 3 weeks after birth, occurs in about 10% of infants born to women who carry this organism in their genital tract. Bordetella pertussis infection can lead to pulmonary hypertension (simulating pneumonia) or secondary bacterial pneumonia. Viruses are a less common cause of pneumonia in neonates compared with older infants. Severe pneumonia can be the result of congenital or perinatal infection with cytomegalovirus (CMV), herpes simplex virus (HSV), or Treponema pallidum. Genital Mycoplasma species and Ureaplasma urealyticum can cause LRTI in very-low-birthweight infants.

Infants, Children, and Adolescents Traditionally, viruses have been considered to be the most common cause of acute LRTI in children between 1 and 36 months of age. However, in a recent study of acute pneumonia in immunocompetent, hospitalized children between 2 months to 17 years of age, bacteria were identified in 60%, viruses in 45%, Mycoplasma species in 14%, Chlamydophila pneumoniae in 9%, and mixed bacterial–viral infections in 23% of the cases.7

Viruses Overall, viruses account for approximately 14% to 35% of CAP in childhood.11 However, when categorized by age, they accounted for 80% of CAP in children < 2 years compared with 49% in those > 2 years of age.12 RSV is the predominant respiratory tract viral pathogen. Other viruses include human metapneumovirus (hMPV),13 parainfluenza viruses (types 1, 2, and 3), influenza viruses (A and B), adenoviruses, rhinoviruses, and enteroviruses. Rhinoviruses have been recovered by culture in 2% to 24% cases of childhood pneumonia.2,14,15 Varicella-zoster virus (VZV), CMV, and HSV typically cause LRTI in immunocompromised children. Recently, human parechovirus 1 (HPeV-1), a picornavirus, was identified to cause LRTI in young children.16 Coronavirus became a global concern in 2003, causing the severe acute respiratory syndrome (SARS)17; although children were infected, the clinical course was mild with no documented death.18,19 Infections with RSV, hMPV, and influenza viruses occur during the winter season whereas infections with parainfluenza viruses and rhinoviruses are more common in spring and autumn; adenovirus infections can occur throughout the year.

Mycoplasma pneumoniae and Chlamydophila pneumoniae In one study, Mycoplasma pneumoniae was detected in 30% of children with CAP.20 Harris et al.21 found that children > 5 years of age had a higher rate of Mycoplasma infection (42%) compared with children < 5 years of age (15%). Coinfections with either Streptococcus pneumoniae (30%) or Chlamydophila pneumoniae (15%) are common.22 Infections due to M. pneumoniae occur in 2- to 4-year epidemic cycles.23 Unlike respiratory viruses that cause rapid transmission among family members, transmission of M. pneumoniae is slow with a median interval of 3 weeks between cases in family members.24 Between 9% and 20% of cases of CAP in children of all ages are associated with recovery of C. pneumoniae;11 the median age is 35 months.7 Asymptomatic carriage of C. pneumoniae is well documented and confounds assessment of pathogenicity.

Bacterial Pathogens Bacterial pneumonia is more common in children living in developing countries, presumably due to chronic malnutrition, crowding, and chronic injury to the respiratory tract epithelium from exposure to cooking and heating with biomass fuels without adequate ventilation.25 Various tests that determine bacterial products in blood, respiratory tract secretions, and urine have been used in an attempt to ascribe a causative role of bacteria but are positive in fewer than 10% of

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Acute Pneumonia

CHAPTER

36

247

TABLE 36-1. Microbial Causes of Community-Acquired Pneumonia in Childhood Age

Etiologic Agentsa

Clinical Features

Birth–3 weeks

Group B streptococcus

Part of early-onset septicemia; usually severe

Gram-negative enteric bacilli

Frequently nosocomial; occurs infrequently within 1 week of birth

Cytomegalovirus

Part of systemic cytomegalovirus infection

Listeria monocytogenes

Part of early-onset septicemia

Herpes simplex virus

Part of disseminated infection

Treponema pallidum

Part of congenital syndrome

Genital Mycoplasma or Ureaplasma

From maternal genital infection; afebrile pneumonia

Chlamydia trachomatis

From maternal genital infection; afebrile, subacute, interstitial pneumonia

Respiratory syncytial virus (RSV)

Peak incidence at 2–7 months of age; usually wheezing illness (bronchiolitis/pneumonia)

Parainfluenza viruses (PIV), especially type 3

Similar to RSV, but in slightly older infants and not epidemic in the winter

Streptococcus pneumoniae

The most common cause of bacterial pneumonia

Bordetella pertussis

Primarily causes bronchitis; secondary bacterial pneumonia and pulmonary hypertension can complicate severe cases

RSV, PIV, influenza, hMPV, adenovirus, rhinovirus

Most common causes of pneumonia

Streptococcus pneumoniae

Most likely cause of lobar pneumonia; incidence may be decreasing after vaccine use

Haemophilus influenzae

Type b uncommon with vaccine use; nontypable stains cause pneumonia in immunocompromised hosts and in developing countries

Staphylococcus aureus

Uncommon, although CA-MRSA is becoming more prevalent

Mycoplasma pneumoniae

Causes pneumonia primarily in children over 4 years of age

Mycobacterium tuberculosis

Major concern in areas of high prevalence and in children with HIV

Mycoplasma pneumoniae

Major cause of pneumonia; radiographic appearance variable

Chlamydophila pneumoniae

Controversial, but probably an important cause in older children in this age group

3 weeks–3 months

3 months–5 years

5–15 years

CA-MRSA, community-acquired methicillin-resistant Staphylococcus aureus; HIV, human immunodeÀciency virus; hMPV, human metapneumovirus. a Ranked roughly in order of frequency. Uncommon causes with no age preference: enteroviruses (echovirus, coxsackievirus), mumps virus, Epstein–Barr virus, Hantavirus, Neisseria meningitidis (often group Y), anaerobic bacteria, Klebsiella pneumoniae, Francisella tularensis, Coxiella burnetii, Chlamydophila psittaci. Streptococcus pyogenes occurs sporadically or especially associated with varicella-zoster virus infection.

cases.2,7,9,26–30 Evidence from multiple sources indicates that S. pneumoniae is the single most common cause of bacterial pneumonia beyond the Àrst few weeks of life.2,11,31 The serotypes that cause uncomplicated pneumonia in the United States are generally similar to those that cause bacteremia or acute otitis media (see Chapter 123, Streptococcus pneumoniae). Pneumococcal pneumonia occurs in all age groups.2,26,32 In the United States and Europe, the frequency of Haemophilus influenzae type b (Hib) infection, including pneumonia, has been markedly reduced in the past decade because of widespread immunization with the Hib conjugate vaccine.33 Pneumonia due to nontypable H. influenzae is also uncommon in the United States except in children with underlying chronic lung disease, immunodeÀciencies, or aspiration. Recently, a virulent strain of communityassociated, methicillin-resistant Staphylococcus aureus (CA-MRSA) carrying virulence factors including the Panton–Valentine leukocidin has emerged as an important agent of pneumonia, including, the United States,34 causing life-threatening necrotizing pneumonia. Streptococcus pyogenes (group A streptococcus) is not a frequent cause of acute pneumonia. However, both staphylococcal and streptococcal pneumonia are rapidly progressive and severe, frequently leading to hypoxemia and pleural effusion within hours.

Other bacteria, especially Gram-negative organisms, are rare causes of pneumonia in previously healthy children.

Occasional Pathogens A variety of epidemiologic and host factors prompt consideration of speciÀc organisms (Table 36-2). The most important of these is Mycobacterium tuberculosis (MTB), which should always be suspected if there is a history of exposure, in the presence of hilar adenopathy, or when pneumonia does not respond in a typical fashion to therapy or with passage of time. In North America and Europe, primary MTB in children is most common among those born to recent immigrants from countries with a high prevalence of infection, after contact with infected adults, or in HIV-infected individuals. Residence in or travel to certain geographic areas suggests consideration of certain pathogens. Coccidioides immitis is endemic in the southwestern United States, northern Mexico, and parts of Central and South America. Histoplasma capsulatum is endemic in the eastern and central United States and Canada. Some other pathogens such as Chlamydophila psittaci and Coxiella burnetii are transmitted from infected birds, animals, or humans. Pneumocystis jirovecii (P. carinii)

248

SECTION

D Lower Respiratory Tract Infections

TABLE 36-2. Occasional Causes of Pneumonia in Special Circumstances Organism

Risk Factors

Diagnostic Methods

Histoplasma capsulatum

Exposure in certain geographic areas (Ohio and Mississippi river valley, Caribbean)

Culture of respiratory tract secretions; urine antigen; serum immunodiffusion antibody test; and serum histoplasma complement fixation antibody test

Coccidioides immitis

Exposure in certain geographic areas (southwestern United States, Mexico, and Central America)

Culture of respiratory tract secretions; serum immunodiffusion antibody test

Blastomyces dermatitidis

Exposure in certain geographic areas (Ohio, Mississippi, St. Lawrence river valleys)

Culture of respiratory tract secretions; serum immunodiffusion antibody test

Legionella pneumophila

Exposure to contaminated water supply

Culture or direct fluorescent assay of respiratory tract secretions; antigen test on urine

Francisella tularensis

Exposure to infected animals, usually rabbits

Acute and convalescent serology

Pseudomonas pseudomallei (melioidosis)

Travel to rural areas of Southeast Asia

Culture of respiratory tract secretions; acute and convalescent serology

Brucella abortus

Exposure to infected goats, cattle, or their products of conception; ingestion of unpasteurized milk

Acute and convalescent serology

Leptospira spp.

Exposure to urine of infected dogs, rats, or swine, or to water contaminated by their urine

Culture of urine; acute and convalescent serology

Chlamydophila psittaci

Exposure to certain infected birds (often parakeets)

Acute and convalescent serology

Coxiella burnetii

Exposure to infected sheep

Acute and convalescent serology

Hantavirus

Exposure to dried mouse dung in a closed structure (opening cabins after winter closure)

Acute and convalescent serology; PCR test on the respiratory tract secretions

PCR, polymerase chain reaction.

causes pneumonia in HIV-infected infants at 3 to 6 months of age, in severely malnourished children, and in other immunocompromised or immunosuppressed hosts. Legionella pneumophila is a rare cause of pneumonia in children but is considered with certain environmental exposures and in immunocompromised individuals.

macrophages. Of these, the alveolar macrophages are the pre-eminent phagocytic cells that ingest and kill bacteria. Viral infection (especially due to influenza virus), high oxygen concentration, uremia, and use of alcohol and/or drugs can impair the function of the alveolar macrophages, predisposing to pneumonia. Cell-mediated immunity plays an important role in certain pulmonary infections such as those caused by M. tuberculosis and Legionella species.

Pathogenesis and Pathology Pneumonia occurs in a child who lacks systemic or secretory immunity to a pathogenic organism. Invasion of the lower respiratory tract or lung usually occurs at a time when normal defense mechanisms are impaired, such as after a viral infection, during chronic malnutrition, or with exposure to environmental pollutants. High density of aerosal exposure or hematogenous spread can occasionally cause bacterial pneumonia. The pulmonary defense mechanisms against LRTI consist of: (1) physical and physiologic barriers; (2) humoral and cell-mediated immunity; and (3) phagocytic activity. Physical barriers of the respiratory tract include the presence of hairs in the anterior nares that can trap particles > 10 μm in size, configuration of the nasal turbinates, and acute branching of the respiratory tract. Filtration and humidification capacities of the upper airways, mucus production, and protection of the airway by the epiglottis and cough reflex (eliminating particles between 2 and 10 μm) are protective physiologic functions. Mucociliary transport moves microscopic amounts of normally aspirated oropharyngeal flora and particulate matter up the tracheobronchial tree, minimizing the presence of bacteria below the carina. However, particles less than 1 μm can escape into the lower airways. Immunoglobulin A (IgA), which has good antibacterial and antiviral activities, is the major antibody secreted by the upper airways; IgG and IgM primarily protect the lower airways. In addition, other substances found in alveolar fluid – including surfactant, fibronectin, complement, lysozyme, and iron-binding proteins – have antimicrobial activity. The LRT has four distinct populations of

Viral Pneumonia Three pathologic patterns are seen with viral pulmonary infections: bronchiolitis, interstitial pneumonia, and parenchymal infection. The first two patterns often overlap.35,36 Viral pneumonia is characterized by neutrophilic infiltration of the lumen of the airway with lymphocytic infiltration of the interstitium and parenchyma of the lungs.37 Giant cell formation can be seen in infections due to measles or CMV, or in children with immune deficiency. Viral inclusions within the nucleus of respiratory cells can be seen in adenoviral pneumonia.38,39 Air trapping with resultant disturbances in ventilation–perfusion ratio can occur from obstructed or obliterated small airways and thickened septa impeding oxygen diffusion. Necrosis of bronchial or bronchiolar epithelium can be seen in severe, sometimes fatal, viral infections (e.g., adenovirus infection).

Bacterial Pneumonia Five pathologic patterns are seen with bacterial pneumonia: (1) parenchymal infection/inflammation/consolidation of a lobe or a segment of a lobe (lobar pneumonia, the classic pattern of pneumococcal pneumonia); (2) primary infection of the airways and surrounding interstitium (bronchopneumonia, often seen with Streptococcus pyogenes and Staphylococcus aureus); (3) necrotizing parenchymal pneumonia that occurs after aspiration; (4) caseating

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Acute Pneumonia

granulomatous disease, as seen with tuberculous pneumonia; and (5) peribronchial and interstitial disease with secondary parenchymal infiltration, as seen when viral pneumonia (usually due to influenza or measles) is complicated by bacterial infection.39 Bacterial pneumonia is associated with diffuse polymorphonuclear infiltration. The airspaces become filled with transudates or exudates, impairing oxygen diffusion. The proximity of alveoli and a rich pulmonary vascular bed increase the risk for complications, such as bacteremia, septicemia, or shock.

CHAPTER

36

249

infection at other sites, including the meninges, can dominate the clinical presentation. Pneumonia in children younger than 2 months of age is usually characterized by tachypnea (respiratory rate > 60 breaths/min), intercostal retractions, or both.52 In very young infants, particularly those who are born prematurely, fever may be absent and apneic spells may be the most prominent initial finding of LRTI from any cause.53 Infants with C. trachomatis pneumonia present insidiously between 3 weeks and 3 months of age with staccato cough, tachypnea, crackles on pulmonary auscultation, and absence of fever. Significant laboratory findings include eosinophilia and elevated total serum IgM concentration.54–56

Clinical Manifestations Use of clinical symptoms and signs to differentiate among sites and causes of LRTI is also discussed in Chapter 23, Respiratory Tract Symptom Complexes. The symptoms of pneumonia are varied and nonspecific. Acute onset of fever, rapid breathing (tachypnea), and cough are the classic symptom complex of pneumonia.40 Fever may be absent in very young infants and in infections due to Chlamydia trachomatis, B. pertussis, and Ureaplasma. Some children have a prodrome of low-grade fever and rhinorrhea prior to developing lower respiratory tract symptoms. There is no single sign that is pathognomonic for pneumonia. Tachypnea, nasal flaring, decreased breath sounds, and auscultatory crackles (crepitations or rales) are specific signs of LRTI. Crackles can be absent in a dehydrated patient. Guidelines developed by the World Health Organization (WHO) for the clinical diagnosis of pneumonia in developing countries highlight tachypnea and retractions as the two best indicators of LRTI.41 Tachypnea is defined as > 50 breaths/minute (min) in infants < 12 months of age, > 40 breaths/min in those between 1 and 5 years of age, and > 30 breaths/min in children > 5 years of age. Palafox et al. observed that, in children < 5 years of age, of all the clinical signs of pneumonia, tachypnea (as defined by WHO) had the highest sensitivity (74%) and specificity (67%) for radiologically confirmed pneumonia but it was less sensitive and specific in early disease.42 Tachypnea can occur in other conditions such as asthma, cardiac disease, and metabolic acidosis. Crackles and bronchial breathing were reported to have sensitivity of 75% but specificity of only 57%43 for pneumonia. Isolated wheezing or prolonged expiration is associated with bronchiolitis and is uncommon in bacterial pneumonia.44 The sensitivity and specificity of clinical findings for predicting the presence of radiographically evident pneumonia have been evaluated in a number of studies.44–47 In one study, the combination of a respiratory rate > 50 breaths/min, oxygen saturation 39°C and a white blood cell (WBC) count > 20,000/mm3 without an alternative source of major infection and with no symptoms or signs of lower respiratory tract disease have radiographically confirmed pneumonia.47,49 A Medline search from 1982 to 1995 of studies that considered observer agreement of clinical examination suggested that observed clinical signs were better than auscultatory signs50; interobserver agreement was low in recognizing crackles, retractions, and wheezing, but high in determining respiratory rate and cyanosis. However, neither respiratory rate nor cyanosis is a specific or sensitive indicator of hypoxia. Oxygen saturation should be measured in any child with respiratory distress, especially if the child has retractions or decreased level of activity.51

Neonates and Young Infants The neonate with bacterial infection due to group B streptococcus, Listeria monocytogenes, or gram-negative bacilli usually manifests respiratory distress in the first few hours of life. Septicemia and

Infants, Children, and Adolescents Viral Pneumonia The onset of viral pneumonia is usually gradual and occurs in the context of a preceding upper respiratory tract illness (URI) (rhinorrhea, low-grade fever, and decreased appetite) in the patient or family members. There is then increase in irritability, respiratory congestion, cough, posttussive emesis, and fever. The patient may not appear toxic although hypoxia can be marked, particularly in a young infant, whose initial presentation can be apnea. The auscultatory findings are not anatomically confined but diffuse and bilateral, consisting of wheezing and crackles. Adenovirus usually produces signs and symptoms similar to other viral infections but it can also cause severe pneumonia similar to a bacterial infection, especially in immunocompromised hosts.

Bacterial Pneumonia The onset of bacterial pneumonia is usually abrupt but may follow several days of mild URI. The patient with pyogenic bacterial pneumonia is usually ill and toxic-appearing with high fever, rigors, and tachypnea. Respiratory distress and hypoxemia, however, can be absent or mild unless there is widespread disease or a large pleural effusion. Cough occurs later in the course of the illness when the debris from the involved lung is swept into the upper airway. Unilateral pleuritic chest pain or abdominal pain in the presence of radiographically demonstrated infiltrate is a specific sign of bacterial pneumonia. Unless there is a parapneumonic effusion, auscultatory findings are usually few (especially in infants) and are focal and limited to an anatomic segment. These include decreased tactile and vocal fremitus on palpation, diminished air entry with rales, and dullness to percussion over the involved area of the lung. Presence of wheezing, in an otherwise healthy child, usually excludes pyogenic bacterial pneumonia.

Other Pathogens The major symptoms of LRTI due to Mycoplasma pneumoniae, Chlamydophila pneumoniae, and Coxiella burnetii (Q fever) are fever and cough that persist for more than 7 to 10 days. The onset of pneumonia caused by M. pneumoniae is not usually well demarcated but malaise, headache, sore throat, fever, and photophobia occur early, and sometimes subside when gradually worsening, nonproductive cough ensues. Although coryza is an unusual symptom, ear infection with or without bullous myringitis can occur with M. pneumoniae infections. Findings on physical examination and auscultation can be minimal, most commonly dry or musical crackles. The presence of Stevens–Johnson syndrome or hemolytic anemia suggests M. pneumoniae infection. M. pneumoniae can cause severe disease in persons with sickle-cell disease in whom acute chest syndrome is common. Chlamydophila pneumoniae infection usually causes bronchospasm and can cause an acute exacerbation of asthma. Q fever, caused by Coxiella burnetii, has an acute onset with intractable headache, fever, and cough with round parenchymal opacities on chest radiograph.

250

SECTION

D Lower Respiratory Tract Infections

Differential Diagnosis Pneumonia is highly probable in children with fever, cough, tachypnea, and shortness of breath (dyspnea) in whom chest radiograph demonstrates pulmonary infiltrates. There are many alternative diagnoses, particularly in the absence of fever or with chronic or relapsing symptoms and signs. These include foreign-body aspiration, asthma, gastroesophageal reflux, cystic fibrosis, congestive cardiac failure, systemic vasculitis, and bronchiolitis obliterans. Children who develop chemical pneumonia after ingestion of volatile hydrocarbons can have severe necrotizing pneumonia with high fever and peripheral neutrophil counts exceeding 15 000/mm3.

Laboratory Findings and Diagnosis Radiograph Routine use of chest radiography did not change the clinical outcome in most cases in a study evaluating ambulatory children > 2 months of age with acute LRTI.57,58 Prescription of an antibiotic was more frequent in those who underwent radiography (61% versus 53%). However, chest radiography is necessary to confirm the presence and determine the location of pneumonia in the following patients: those < 12 months of age with acute LRTI who are hospitalized or severely ill; those who have recurrent disease or chronic medical conditions; those who develop complications; and those in whom the diagnosis is uncertain. The chest radiograph can appear falsely normal in children examined early in the course of pneumonia or in dehydrated patients.44 Chest radiography is insensitive in differentiating bacterial from nonbacterial pneumonia; however, combined with clinical findings, it is accurate in excluding most cases of bacterial pneumonia.59,60 Bilateral diffuse infiltrates are seen with pneumonia caused by viruses, Pneumocystis jirovecii, Legionella pneumophila and occasionally M. pneumoniae. Both Chlamydophila pneumoniae and M. pneumoniae (Figure 36-1) cause focal radiographic abnormalities, which are out of proportion to clinical findings. Distinctly confined lobar or segmental abnormality or a large pleural effusion suggests bacterial infection (Figure 36-2). Rarely, M. pneumoniae or adenovirus can manifest with these findings.61–63 Round appearance of infiltrate is common in children younger than 8 years of age and is most often due to Streptococcus pneumoniae. Enlarged or calcified hilar lymph nodes suggest tuberculosis or a fungal infection such as histoplasmosis, and can also occur in Mycoplasma pneumonia and in patients with cystic fibrosis. Tuberculosis is highly likely in an adolescent with epidemiologic risk factors and apical disease or cavitation. Pneumatoceles (thin-walled air–fluidfilled cavities) resulting from alveolar rupture are usually associated with infection due to Staphylococcus aureus but can be seen in infections due to Streptococcus pneumoniae, S. pyogenes, Hib, other gram-negative bacteria, or anaerobes. Involvement of the lower lobes, particularly with recurrent infections, suggests aspiration pneumonia, or if confined to the same site, pulmonary sequestration. Recurrent bacterial pneumonia involving the same anatomic area suggests congenital anomaly or foreign body whereas recurrences in different areas suggest an abnormality of host defense, cystic fibrosis, or other causes (see Chapter 37, Persistent and Recurrent Pneumonia). A chest radiograph is rarely useful in following the clinical course of a child with acute pneumonia who is recovering as expected. Radiographic improvement lags clinical changes significantly; complete resolution is expected in children 4 to 6 weeks after onset. Follow-up radiography is indicated for children with lobar collapse, complicated pneumonia, recurrent pneumonia, and round pneumonia (to exclude tumor as the cause).64,65

(ESR), and C-reactive protein (CRP) level, best detect invasive infections, particularly those caused by bacteria. Viral pneumonia is associated with a less brisk rise of acute-phase reactants than bacteria pneumonia. However, some viral agents, especially adenovirus, influenza, and measles virus, can induce a host response similar to that of invasive bacterial infection. In a prospective study examining the utility of routinely obtaining the acute-phase reactants in children with pneumonia, the authors concluded that these tests do not stand alone as indicators of bacterial versus viral pneumonia.66,67

Laboratory Tests

Diagnosis of Specific Agents

Indices of host response (acute-phase reactants), including peripheral WBC, white blood cell differential, erythrocyte sedimentation rate

The rigor of an investigation for specific causative agents in pneumonia depends on the severity of illness, the presence of underlying

A

B Figure 36-1. (A) Posteroanterior and (B) lateral plain radiographs of a 2-year-old child with dextrocardia, complex congenital heart disease, and Mycoplasma pneumoniae pneumonia. Note bilateral patchy alveolar infiltrates in both lower lobes. (Courtesy of E.N. Faerber and S.S. Long, St. Christopher’s Hospital for Children, Philadelphia, PA.)

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Acute Pneumonia

CHAPTER

36

251

serum IgM and IgA antibodies to M. pneumoniae is positive in 80% of cases during the early convalescent period70; specificity and reproducibility may be suboptimal. Examining paired sera is the most definitive test. Chlamydophila pneumoniae infection is identifiable serologically, by isolation of the organism in tissue culture, or by PCR, although none of these tests may be readily available.71 Serology is also effective in detecting infections with other agents that cause atypical pneumonia, namely C. psittaci and Coxiella burnetii. When tuberculosis is considered, a Mantoux skin test (a 5 TU intradermal test, purified protein derivative (PPD)) should be placed on the patient as well as on all the immediate family members and other significant contacts. In acutely ill patients, the PPD skin test can be nonreactive because of general or specific anergy to MTB antigen. When the suspicion of tuberculosis is strong, multiple respiratory tract specimens should be obtained; specimens include sputum (spontaneous or induced), gastric aspirate, and/or bronchoalveolar lavage. Gastric aspirates are superior to bronchoscopic specimens in infants with tuberculosis72 and should be obtained in patients with suspected primary pulmonary infection or miliary disease without cough.

Bacterial Pathogens

Figure 36-2. Plain radiograph showing consolidative pneumonia in the right upper lobe, typical of acute bacterial pneumonia.

disease, and clinical manifestations. Mildly to moderately ill ambulatory patients can often be managed empirically without specific diagnostic tests. Under circumstances in which the identity of a specific etiologic diagnosis is desired, a number of investigations may be warranted. Some of these circumstances include all patients admitted to the hospital with pneumonia, patients with underlying medical conditions, and when there is a community outbreak caused by an organism that has recently been recognized to be contagious (i.e., an emerging infection).

Viruses A viral pathogen is best identified by recovering the organism in tissue culture or by detection of viral products (antigens or nucleic acid) in respiratory tract secretions. Combined real-time PCR can rapidly detect common viral and atypical bacterial agents of CAP.68 However, both false-positive and false-negative results can occur when the specimen is improperly obtained or transported and/or the tests are suboptimally performed. A nasopharyngeal wash or aspirate is the most sensitive specimen because it contains infected epithelial cells. The presence of a viral agent in the upper respiratory tract does not exclude the presence of secondary bacterial pneumonia. Clinical correlation is necessary. Obtaining acute and convalescent sera to assess rising antibodies to various viruses is usually confined to research settings.

The diagnosis of most bacterial causes of pneumonia is problematic. Young children do not effectively cough up sputum, resulting in a specimen contaminated with saliva. In older children, a sputum sample is considered appropriate for microbiologic evaluation when Gram stain reveals < 10 squamous epithelial cells and > 25 neutrophils per low-power field, and a predominant organism. Nasopharyngeal cultures are not usually reliable specimens because many bacterial pathogens are also common commensals. Further, noncommensal organisms residing in the upper airway may not be the causative agent of LRTI. Tracheal aspiration is useful for culture if performed with direct laryngoscopy. However, culture samples obtained via a catheter directly passed through a tracheostomy, endotracheal tube, or deep nasotracheal tube have limitations due to frequent contamination with upper respiratory tract organisms. (They could be evaluated as for a sputum sample.) Quantitative culture performed on a bronchoalveolar lavage specimen is useful, with isolate colony count > 104/mL considered significant. Blood culture is specific but insensitive. A recent study demonstrated that transthoracic needle aspiration (lung tap) in hospitalized children with clinical pneumonia had a high microbiologic yield and was relatively safe.31 However, this procedure is not widely performed in the United States. A diagnostic thoracocentesis should be considered in patients with more than a minimal pleural effusion, when the etiology is obscure, or when mechanical removal is indicated. A lung biopsy and/or a bronchoalveolar lavage may be necessary to confirm the diagnosis in patients who are seriously ill, immunocompromised, intubated, or who are not responding to empiric therapy.

Management Indications for Hospitalization

Other Pathogens M. pneumoniae can be detected most effectively by PCR methodology but the test may not be available in most hospital or commercial laboratories. Mycoplasma culture is available in some commercial and hospital laboratories but can take 3 weeks to complete. Cold agglutinins (IgM antibodies that agglutinate human red cells) are found in 30% to 75% of individuals with M. pneumoniae pneumonia during the acute phase of the disease.69 The titer correlates with the severity of disease; a titer of 1:64 or greater has a high predictive value for M. pneumoniae infection. The cold agglutinin test result can be positive in lower titers in infections due to adenovirus, influenza virus, Epstein–Barr virus, and CMV, as well as in lymphoma. The test can be negative in young children and in those with mild disease. Testing for

Hypoxemia with SaO2 < 92% is the single most important indication for hospitalization because a hypoxemic child is at greater risk of death than an adequately oxygenated child.43 Other indications include cyanosis, rapid respiratory rate (RR > 70 breaths/min in an infant or > 50 breaths/min in a child), apnea, dyspnea, expiratory grunting, dehydration, toxic appearance, poor oral intake, recurrent pneumonia, underlying medical condition, or uncertain observation at home.

Supportive Therapy Oxygen Therapy and Ventilatory Support. Oxygenation is assessed continuously by measuring oxygen saturation or arterial PO2. Hypoxic infants and children may not appear cyanotic until they are

252

SECTION

D Lower Respiratory Tract Infections

terminally ill. Mental agitation, clinically evident as increased irritability, may be an indication of hypoxemia. Therefore, supplemental oxygen therapy is indicated in any patient whose oxygen saturation is persistently 92% or less. Sole reliance on pulse oximetry values is hazardous in ill patients because hypercarbia is an important sign of impending respiratory failure, especially in the tiring young infant who may have relatively preserved oxygenation. Blood gas should be evaluated to detect impending respiratory failure and provide ventilatory support when indicated. Fluid Therapy. Rapid breathing, fever, and fatigue increase the fluid requirements in a child with acute LRTI. Most patients can be hydrated orally if they are given small volumes of fluids frequently. Intravenous hydration may be necessary for intubated or seriously ill children with very rapid breathing because of increased likelihood for pulmonary aspiration. The syndrome of inappropriate secretion of antidiuretic hormone (SIADH) can be seen in approximately one-third of patients hospitalized with pneumonia.73 Nutrition. Malnutrition has been associated with a worse prognosis of pneumonia. Infants and small children fare better if fed in small quantities and more frequently to prevent pulmonary aspiration.74 Seriously ill or intubated children may require placement of an enteral feeding tube or parenteral nutrition. Fever and Pain Management. Persistent and high fever increases the basal metabolic rate and oxygen consumption. Similarly, pain interferes with the depth of breathing and with the ability to cough effectively. Therefore, it is important that the patient is kept comfortable by using age-appropriate antipyretic or analgesic agents or both.

Antimicrobial Therapy Treatment of pneumonia in infants and children is often empiric because of the difficulty in proving etiology. Because viral LRTI is most frequent in previously healthy children, antibiotics are only administered when findings suggest bacterial infection. Optimal antibiotic treatment of pneumonia in infants and children has not been determined by randomized controlled clinical trials. Recommendations are based on the most likely etiologic agents at different ages and in various settings. Because pathogens of pneumonia in neonates are similar to those of sepsis, broad-spectrum antibiotics like ampicillin and gentamicin are appropriate in this age group. A macrolide antibiotic is appropriate for Chlamydia trachomatis and Ureaplasma. In infants < 1 month of age the preferred macrolide is azithromycin because azithromycin is not known to cause hypertrophic pyloric stenosis.75 For pertussis, the dose is 10 mg/kg per day on each of 5 days. Amoxicillin (80 to 90 mg/kg per day) is effective empiric therapy for febrile children with pneumonia; alternatives include high-dose amoxicillin-clavulanate (14:1 preparation), cefuroxime axetil, or cefdinir.76 In children > 5 years of age in whom atypical organisms (Mycoplasma or Chlamydophila pneumoniae) are suspected, initial treatment with a macrolide or doxycycline (recommended only after 7 years of age) may be appropriate if pyogenic bacterial pneumonia is not likely. For a hospitalized child beyond the neonatal period with uncomplicated pneumonia, initial parenteral (intravenous) therapy with ampicillin is appropriate, even in areas with penicillin-nonsusceptible Streptococcus pneumoniae. Some experts recommend cefuroxime, ceftriaxone, cefotaxime, or ampicillin-sulbactam, and use higher doses of beta-lactam agents than in the era prior to penicillin-nonsuspectibility.77–79 While the use of vancomycin, clindamycin, or linezolid is not recommended for initial treatment of uncomplicated CAP, these drugs are considered if infection due to CA-MRSA is suspected, if pneumonia is unresponsive to initial antibiotics, or in those patients allergic to betalactam agents.80 Other antimicrobial agents may be chosen if a likely pathogen is identified, the case has clinical or epidemiologic features strongly suggestive of a particular infection, or the evolution of the disease suggests a more specific cause. Opinions differ about the frequency with which viral pneumonia is

complicated by bacterial superinfection.81 There is a good deal of evidence, however, that withholding antibiotics from hospitalized children with pneumonia clinically compatible or proven to be of viral origin is safe and is preferable to empiric antibiotic treatment.82 Use of specific antiviral therapy depends on the pathogen, the severity of the clinical course, and availability of effective nontoxic therapy. Use of ribavirin for the treatment of RSV and acute LRTI is guided by recommendations from the American Academy of Pediatrics,83,84 although the value of treatment has been questioned.85,86

Prognosis and Sequelae Mortality due to CAP is uncommon beyond infancy in Europe and North America because of improved and enhanced immunization rates, early access to medical care and availability of antimicrobial therapy. Most healthy children with acute LRTIs recover without sequelae. However, in some patients, especially premature infants, immunocompromised hosts, or in children with chronic lung, neuromuscular, or cardiovascular diseases, complications can develop from an acute LRTI. In the late 1990s, there was a significant increase in the relative incidence of complications from bacterial pneumonia in infants and children living in North America,87,88 the exact reason for which is still obscure.89 Since 2000, universal immunization with pneumococcal conjugate vaccine in children less than 2 years of age has reduced the frequency of complications due to pneumococcal pneumonia,90 and the overall complications due to presumed bacterial pneumonia. The complications of bacterial pneumonia include: necrotizing pneumonia, parapneumonic effusion, empyema, pneumatocele formation, and lung abscess. Several epidemiologic studies have linked asthma and other respiratory problems occurring later in childhood to viral bronchiolitis or atypical pneumonia in infancy.91–95 Although many etiologic agents, including RSV and Chlamydia in particular, have been implicated, bacterial pneumonia has not been specifically found to have long-term sequelae. One study suggested that chronic cough can follow C. trachomatis pneumonia in infancy96 and another has suggested that asthma can follow Chlamydophila pneumoniae infection.97 A study of 35-year-old adults, whose parents had been interviewed 18 years previously, showed that those reported to have had pneumonia before age 7 years demonstrated a significant reduction of forced expiratory volume and forced vital capacity,98 confirming findings of prior studies.99,100 Several longitudinal studies of lung function in children with bronchiolitis have suggested that lung function abnormalities may have preceded the acute infectious illness.101–103 Thus, it remains unclear whether childhood pneumonia causes subsequent pulmonary abnormalities.

Prevention Most respiratory viral infections are transmitted by direct inoculation from hands contaminated with respiratory secretions on to conjunctival and nasal mucosa. Spread by airborne droplets also occurs occasionally. Hand hygiene by caregivers or medical personnel, before and after contact with patients, is the single most important method of preventing hospital-associated infections. Spread of infection by small droplets can be reduced by placing the patient in a negative-pressure room. All caregivers should wear facemasks and goggles. The development of vaccines for the prevention of pneumonia is complicated by the large number of etiologic agents. Universal use of Hib conjugate vaccine has eliminated invasive Hib disease. Universal immunization with pneumococcal conjugate vaccine has significantly reduced the incidence of pneumococcal pneumonia in children,104,105 and in elderly contacts through herd immunity.106,107 RSV bronchiolitis and pneumonia can be reduced in high-risk infants by passive immunoprophylaxis using palivizumab, a monoclonal antibody directed at the fusion protein of RSV, which can

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Acute Pneumonia

be administered intramuscularly every month to infants and young children at high risk for infection (see Chapter 225, Respiratory Syncytial Virus).108,109 Annual vaccination against influenza is recommended for individuals > 6 months of age with underlying medical conditions such as chronic lung diseases (including mild asthma), neuromuscular disorders, congenital cardiac conditions, and diabetes because of high risk of complications or more severe respiratory disease. Annual influenza vaccine is also recommended for all healthy children between 6 and 59 months of age to reduce morbid disease in them and the burden of disease in the community.110 It is anticipated that the incidence of bacterial pneumonia in children, especially that caused by Staphylococcus aureus and Streptococcus pyogenes, will decrease substantially in the future with increased immunization rates for influenza and varicella viruses in young children.

PLEURAL EFFUSION, PARAPNEUMONIC EFFUSION, AND EMPYEMA Pleural effusion is the presence of demonstrable fluid of any character between the visceral and parietal pleurae. Pleural effusions are classified as being a transudate or an exudate based on the biochemical characteristics of the fluid. The relative concentration of pleural fluid protein to serum protein is > 0.5 in an exudate versus < 0.5 in a transudate (Table 36-3). Exudates can have an infectious or noninfectious cause whereas transudates are less often caused by infections. Noninfectious causes of pleural effusions are listed in Table 36-4. Several drugs, including hydralazine, nitrofurantoin, dantrolene, amiodarone, methysergide, procarbazine, bromocriptine, methotrexate, and agents associated with a lupus-like reaction, are associated with pleural effusions. Parapneumonic effusions are inflammatory fluid collections adjacent to a pneumonic process, seen in about 40% of cases of bacterial pneumonia.84 They can be classified as complicated or uncomplicated on the basis of various characteristics, particularly pH, glucose, and lactate dehydrogenase (LDH) concentrations (see Table 36-3).111 The term empyema is used when a parapneumonic fluid becomes purulent or seropurulent. Parapneumonic effusion (PPE) usually occurs as a complication of pyogenic bacterial pneumonia but can occasionally occur secondary to other etiologic agents (e.g., Mycoplasma) or as a consequence of an infection from another contiguous site. PPE can be complicated

TABLE 36-3. Biochemical Characteristics of Parapneumonic Pleural Effusions Laboratory Value pH Glucose level Lactate dehydrogenase concentration Pleural protein:serum protein

Uncomplicated Effusion transudate

Complicated Effusion exudate

> 7.2 > 40 mg/dL < 1000 IU/mL

< 7.1 < 40 mg/dL > 1000 IU/mL

< 0.5

> 0.5

CHAPTER

Hypoalbuminemia Congestive heart failure Cirrhosis with ascites Myxedema Peritoneal dialysis Central venous catheter leak Fluid mismanagement Adult respiratory distress syndrome

253

(CPPE) or uncomplicated (UPPE). CPPE and empyema represent a different spectrum of the same disease.112 The estimated incidence of empyema in children is approximately 3.3 per 100,000.84 Both CPPE and empyema are serious illnesses, often associated with significant morbidity113,114 but with low mortality rates. Seventy percent of complicated pneumonia occurs in children < 4 years of age; pneumatoceles (defined as air–fluid-filled alveoli) occur predominantly in children < 3 years of age.115

Etiologic Agents During the latter 1990s Streptococcus pneumoniae, especially serotype 1, emerged as the most common isolate from children with complicated parapneumonic effusion.116 With the introduction of universal conjugated pneumococcal vaccination in the United States, the incidence of CPPE due to vaccine-serotype S. pneumoniae has decreased,87 although serotype 1 and other nonvaccine serotypes appear to be emerging.117 Likewise, universal childhood immunization against Hib has also significantly reduced the incidence of CPPE due to this agent. CA-MRSA is an important cause of pneumonia and CPPE in children.104,118 In South Asia, Staphylococcus aureus is the most common cause of CPPE or empyema.119 Less frequently, group A streptococcus, Pseudomonas aeruginosa, mixed anaerobic pathogens, Mycobacterium species and, rarely, fungi can be etiologic agents.120 Although effusions have been described in pneumonia due to M. pneumoniae and viruses,121–123 they are rarely large enough to require intervention. In several large studies, PPEs are found to be sterile in 22% to 58% of cases.116,124–128

Pathogenesis and Pathologic Findings Under normal circumstances the pleural space contains 0.3 mL of fluid per kilogram body weight. The pleural circulation is maintained by a delicate balance between secretion and absorption of pleural fluid by lymphatic vessels in the pleura. When this balance is disturbed, fluid accumulates. Various infectious agents induce pleural effusion by different mechanisms. Effusion can result from a sympathetic response to a bacterial infection (by elaboration of cytokines), extension of infection, an immune-complex phenomenon (e.g., pneumococcal infections) or as a hypersensitivity reaction (e.g., rupture of tuberculous granulomas). Replication of microorganisms in the subpleural alveoli precipitates an inflammatory response resulting in endothelial injury, capillary permeability, and extravasation of pulmonary interstitial fluid into the pleural space. The pleural fluid is readily infected because it lacks opsonins and complement. Bacteria also interfere with the host defense mechanism by production of endotoxins and other toxic substances. Anaerobic glycolysis results from further accumulation of neutrophils and bacterial debris. This in turn causes pleural fluid to become purulent and acidic (i.e., empyema). The acidic environment of the pleural fluid suppresses bacterial growth and interferes with antibiotic activity. With disease progression, more inflammatory cytokines are released and there is activation of coagulation leading to deposition of fibrin. The American Thoracic Society has divided empyema into three stages: (1) exudative phase, in which the pleural fluid has low cellular

TABLE 36-4. Noninfectious Causes of Pleural Effusion in Children Transudate

36

Exudate Spontaneous chylothorax Posttrauma or postsurgical Postoperative chylothorax Pulmonary lymphangiectasia Uremic pleuritis Sarcoidosis Dressler syndrome (postmyocardial infarction)

Malignancy Collagen vascular disease Pancreatitis Subphrenic or other intraabdominal abscess Drug reaction Meig syndrome (pelvic tumor)

254

SECTION

D Lower Respiratory Tract Infections

content; (2) fibrinopurulent phase, in which frank pus containing increased neutrophils and fibrin is formed. This fibrinous pus coats the inner surfaces of the pleura, interfering with lung expansion. Fibrin also leads to loculations within the pleural space; and (3) organizational phase (late stage), in which fibroblasts migrate into the exudate from visceral and parietal pleurae, producing a nonelastic membrane called the pleural peel. Before the availability of antibiotics, spontaneous drainage sometimes occurred by rupture through the chest wall (empyema necessitans) or into the bronchus (bronchopleural fistula). At present, such events are rare and are usually seen due to antibioticresistant, nosocomially acquired bacteria or insidiously progressive actinomycosis. The acidic pleural pH is the basis for biochemical tests that differentiate uncomplicated effusions (which usually resolve with antibiotic therapy alone) from complicated effusions (that benefit from more aggressive drainage and other interventions); see Table 36-3.

Clinical and Radiographic Manifestations Necrotizing pneumonia or effusion should be suspected when the response of a lobar, lobular, or alveolar pneumonia to appropriate antibiotic therapy is slow, or if there is clinical deterioration during treatment. Early in the disease, the symptoms of pleural effusion can be nonspecific and include malaise, lethargy, and fever. This is followed by cough and rapid breathing. Pain in the chest and/or in the abdomen develops on the involved side, associated with high fever, chills, and rigors. The child may guard (splint) or lie on the involved side of the chest in an attempt to minimize pain. As the effusion progresses, so does difficulty in breathing (dyspnea). On physical examination, the patient is usually ill and toxic-appearing, febrile, with significant tachypnea and shallow respirations (to minimize pain). Scoliosis may be noted on the involved side and the affected side may be tender to palpation. Breath sounds are usually grossly diminished on auscultation. Crackles from an associated pneumonia may be audible. Pleural rub is usually audible when there is a small fibrinous exudate within the pleural space. The percussion note on the involved side is dull when the effusion is free-flowing; by contrast, dullness can disappear as the effusion organizes. Chest radiography is more sensitive than physical examination, especially in detecting small pleural effusions. Blunting of the costophrenic angle, thickening of the normally paper-thin pleural shadow, and a subpulmonic density all suggest pleural effusion (Figure 36-3). Movement and layering of fluid on lateral decubitus films differentiate free effusions from loculated collections, pulmonary consolidation, and pleural thickening. In large effusions containing more than 1000 mL of fluid, compression of the lung and shift of the trachea away from the effusion occur.129 Computed tomography (CT) is not routinely necessary to differentiate simple parapneumonic effusions from empyema or for recognizing loculations within an empyema.130 However, CT often gives diagnostic information regarding the parenchymal component of the process not discernible by conventional radiographs.131 Ultrasonography is usually more useful in the management than in the detection of effusions; it helps to localize and estimate the size of an effusion precisely.

Laboratory Findings and Diagnosis A sample of pleural fluid is obtained. The fluid should be centrifuged if it is cloudy on gross inspection; persistent cloudiness is suggestive of a chylothorax. Putrid odor is pathognomonic of anaerobic infection but is present in only half of such infections. Pleural fluid is analyzed by biochemical (pH, glucose, and LDH levels), hematologic (total WBC count and differential), special staining techniques, and culture. If the diagnosis of pneumonia is not clear, other tests may be helpful, such as cellular histology and flow cytometry for malignancy. The usefulness of pH and glucose level in guiding the management of effusions is derived from data in

Figure 36-3. Plain radiograph showing left lower lobe pneumonia and a parapneumonic effusion, typical of acute bacterial pneumonia.

adults.132,133 The results of one small pediatric series support a similar interpretation in children.124 To enhance the accuracy of tests for pH and glucose, the fluid must be obtained in a heparinized syringe with the exclusion of air, placed on ice, and tested promptly. If immediate performance of such tests is not possible, the fluid should be stored at 0°C for no more than 2 hours. The pH can be lowered if the fluid is stored at room temperature and raised upon exposure to air. Effusions that are likely to require aggressive management (CPPEs) can be differentiated from those that do not require such management by gross appearance of pus, or by the pleural fluid having a pH < 7.1, a glucose level < 40 mg/dL, and a LDH concentration > 1000 IU/mL.133 Neutrophil predominance in pleural fluid suggests an acute process, such as pneumonia, pancreatitis, pulmonary embolism, or intra-abdominal abscess.134 Lymphocytic predominance suggests tuberculosis, fungal, or Mycoplasma infection but can also be due to chylothorax; mixed mononuclear cell types suggest a more chronic process, such as malignancy, uremia, or collagen vascular disease. Pleural fluid eosinophilia is not usually helpful in determining the cause of an effusion in children,135 although eosinophilia is consistently found in effusions accompanying pneumonia due to various forms of paragonimiasis.136 Bacteriologic examination of pleural fluid always includes Gram stain and culture for aerobic and anaerobic bacteria. Acid-fast stain and culture for Mycobacterium tuberculosis, and fungal stain and culture, should be obtained from both the pleural fluid and sputum when clinically indicated. The possibility of isolating a pathogen from either the pleural fluid or blood varies widely, from 8% to 76%.124–126,137–140 Direct antigen testing is not considered useful because of technical problems associated with all the current tests available. A Mantoux (PPD) skin test should be performed in patients with epidemiologic risk factors or clinical manifestations possibly suggesting tuberculosis. Anergy is unusual in the presence of pleural effusion.141 If organisms are not seen on acid-fast stains of the sputum or pleural fluid, a pleural biopsy (and thoracoscopy when possible) should be performed because granulomas are observed in 50% to 80% of pleural biopsy specimens compared with a 10% rate of isolation of Mycobacterium in cultures of pleural fluid in patients with tuberculous effusions.141,142 Pleural biopsy is also useful in diagnosing other granulomatous infections (histoplasmosis, other fungal infections, and Francisella tularensis) and malignancies.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Acute Pneumonia

Cytokine levels, especially interleukin-8, are significantly higher in empyema. They may be useful to differentiate CPPE from UPPE.143,144

Management The optimal management of PPEs in children is controversial. Data underlying many of the decision points come from studies in adults, although effusions in children differ in the following ways from those in adults: the spectrum of causative organisms is not the same; underlying airway and lung diseases are less common and are different in children; and the capacity of the lung and pleural space to recover fully without aggressive intervention is greater in children. There have been no prospective pediatric trials comparing management schemes and, therefore, no optimal evidence-based management has been established. The traditional approach to the management of PPEs is initially to obtain fluid by needle aspiration, primarily to establish a microbial etiology; this is followed by administration of antibiotic therapy and observation of the child’s clinical course. In most cases, symptoms resolve. A more aggressive approach is taken if any of the following is present: persistence of fever or toxicity; rapid reaccumulation of the pleural effusion; respiratory compromise because of the size of the effusion. Under these circumstances, a drainage tube is initially inserted using ultrasonographic guidance. Benefit of fibrinolytic agents is controversial. Surgical intervention is considered in moderately severe cases. In severely affected individuals, more aggressive approaches may be appropriate. Some experts advocate immediate thoracoscopic examination of the pleural space, lysis of the adhesions, and placement of a large-diameter drainage tube.127 With the availability of less invasive video-assisted thoracoscopic surgery (VATS), there is increasing opinion that VATS is the best treatment for patients with CPPE and empyema. In one study, treatment failure and mortality rates were higher in patients who had nonoperative therapies (chest tube placement, antibiotics, and/or fibrinolytic therapy) compared with primary operative therapy (VATS and thoracotomy).145 In addition, it was noted in this study that fibrinolytic therapy, using streptokinase or urokinase, was associated with more complications as systemic absorption of streptokinase after intrapleural instillation resulted in coagulopathy.146 Another study concluded that VATS was both safe and effective as well as superior to chest tube drainage for large loculated empyemas147 when used early in the course of the illness. Bronchoscopy may be indicated in anaerobic infections to hasten drainage or the removal of a foreign body. Pleural decortication is reserved for patients who have entrapment of lung with persistent and severe restrictive lung disease.

Antimicrobial Therapy When choosing an antibiotic for the treatment of CPPE or empyema, consideration must be given to the following: probable pathogens predicted from patient’s age; clinical circumstances; Gram stain of the pleural fluid; and radiographic appearance. The clinical circumstances include whether the infection was community- or hospital-acquired, the host is immunocompetent or immunocompromised, and/or has an underlying medical condition. In most cases, empiric therapy should adequately treat Streptococcus pneumoniae, CA-MRSA, and group A streptococcus. In addition, therapy should include adequate coverage for anaerobic bacteria in patients at risk for aspiration. In cases where atypical pathogens are suspected, a macrolide (for children < 7 years) or doxycycline (for children > 7 years) may be added. When the specific causative pathogen is determined, the antibiotic spectrum should be narrowed. Duration of parenteral therapy and total treatment is individualized on the basis of clinical response and adequacy of drainage. A minimal duration of 4 weeks is usual for CPPE; treatment should be prolonged when drainage is delayed and systemic manifestations are protracted. In circumstances in which the effusion persists and the microbial etiology is unknown, it is important to remember that fever, anorexia,

CHAPTER

36

255

and toxicity can be prolonged, even when the choice of antibiotics is correct – due to inflammatory response within the pleural space. Therefore, additions or changes in appropriately selected antibiotic therapy should be avoided.

Prognosis The mortality rate for CPPE in previously healthy children is between 0% and 3%.145 Historically, mortality was highest in small infants and with infection due to Staphylococcus aureus. Long-term follow-up data of patients with PPEs are limited. The capacity of the child’s lung to recover is great and even patients with prolonged morbidity rarely require decortication. Patients are usually asymptomatic at follow-up but radiographs may show pleural thickening,125,128 which regresses over months. Mild abnormalities occur with equal frequency in children treated with and without chest tube drainage.148 Re-expansion of the affected segment of the lung is an important consideration regarding need for surgical intervention, even if clinical improvement has ensued.

NECROTIZING PNEUMONIA AND LUNG ABSCESS Necrotizing pneumonia usually occurs as a consequence of a localized lung infection by particularly virulent, pyogenic bacteria. Necrotizing pneumonia in an otherwise healthy child can resolve without further complications after antimicrobial treatment, or it can be followed by the formation of a lung abscess or a pneumatocele (blebs in the lung parenchyma created by coalescence of alveolar spaces following rupture of septa) or bronchopleural fistula. Lung abscesses can be the outcome of widely variable pathogenic processes, such as: (1) a consequence of necrotizing pneumonia; (2) localized infection after aspiration of heavily infected mouth secretions (sometimes along with a foreign body); (3) focal infection of the lung that occurs during highgrade bacteremia or as a consequence of septic emboli; and (4) complication of a subacute or chronic airway infection seen as a late consequence of cystic fibrosis, after prolonged intubation, or after an infection with hospital-associated (nosocomial) bacteria. In the immunocompromised host, necrotizing pneumonia can be caused by bacteria and/or fungi that invade vessel walls, with subsequent lung infarction (see Chapter 38, Pneumonia in the Immunocompromised Host).

Etiologic Agents Community-acquired bacterial pneumonia can result in necrosis of the lung parenchyma, which is often discovered on chest radiography or CT in a child with prolonged fever and ill appearance.90,149 As in most cases of bacterial pneumonia, the causative pathogen may not be identified. When a pathogen is isolated, most often it is Streptococcus pneumoniae, or, less commonly, Staphylococcus aureus, (especially CA-MRSA) or Streptococcus pyogenes. Hemoptysis can be a manifestation of necrotizing pneumonia. Pneumococcal necrotizing pneumonia, if not accompanied by parapneumonic effusion, usually resolves with antimicrobial treatment alone. Infection with either S. pneumoniae or Staphylococcus aureus can cause pneumatoceles. Occasionally, abscesses can form from pneumatoceles or as a complication of staphylococcal pneumonia.150–153 Rarely, severe Mycoplasma pneumoniae pneumonia is accompanied by abscess formation.154 Lung abscess is frequently accompanied by PPE.155 Pneumonia associated with aspiration of bacteria from the oropharynx, or after aspiration of regurgitated stomach contents, is particularly likely to cause necrosis and abscess formation. Anaerobic bacteria are frequently recovered from lung abscesses, accounting for 30% to 70% of isolates.153,155 Peptostreptococcus spp., Bacteroides spp., Prevotella spp., and Veillonella spp. are most common. Facultative aerobic pathogens include b-hemolytic streptococci (Lancefield groups C and G). Single or multiple lung abscesses can result from right-sided

256

SECTION

D Lower Respiratory Tract Infections

endocarditis, severe septicemia (usually with S. aureus), or infarction/infection following endovascular infection of the large veins in the neck (Lemierre disease).156 The bacteriology reflects the primary pathogens, which can be S. aureus (common pathogen), members of the Streptococcus anginosus group, or Fusobacterium necrophorum (Lemierre disease). Abscesses in intubated infants and children are usually due to hospital-associated pathogens.157 Abscesses can develop in the later stages of cystic fibrosis secondary to chronic bronchiectasis. In such cases, Staphylococcus aureus, Pseudomonas aeruginosa, and mycobacteria are considered.153,158 Table 36-5 shows the common microorganisms isolated in several series of children with lung abscesses.153,155,157,159,160

Pathogenesis Necrotizing pneumonia appears as a consequence of severe lobar or alveolar pneumonia in which the confined infection results in parenchymal damage. Such necrosis can result in abscess formation in some instances when treatment is adequate, and more frequently when it is delayed or inadequate. Aspiration and obstruction of the airways predispose to lung abscess. Risk factors for aspiration include: decreased level of consciousness due to neurologic disease, anesthesia, alcohol, or drugs; neuromuscular disorders depressing the gag reflex; esophageal abnormalities; gastroesophageal reflux; and prolonged endotracheal intubation. Periodontal disease predisposes to bacterial hypercontamination of aspirated material.155 The mechanism of abscess formation in otherwise healthy children is most often from an obstruction due to an aspirated foreign body, with growth of bacteria distal to the obstruction. Abnormal drainage, as seen in congenital pulmonary sequestration, lobar emphysema, and pneumatocele formation, predisposes to abscess formation through the same mechanism. Any high-grade bacteremia or heavy seeding of bacteria into the pulmonary circulation via the systemic venous system can also lead to lung abscess formation. Pulmonary infarction associated with septic embolization increases the likelihood of abscess formation. Chronic airway disease, cystic fibrosis, congenital ciliary dysfunction, or bronchiectasis increases susceptibility to lung abscess. Instrumentation, either during surgery or in intensive care, raises the likelihood of pulmonary

infection and abscess formation, probably through a similar pathogenic pathway.157 In addition, impairment of humoral or cellular immune responses results in greater susceptibility to lung abscess.

Clinical Manifestations Clinical manifestations of necrotizing pneumonia are similar to, but usually more severe than, those of nonnecrotizing pneumonia due to the same bacteria. Prolonged fever and a toxic appearance or persistent hypoxia, despite appropriate antimicrobial therapy, is characteristic. The evolution from necrotizing pneumonia to abscess is frequently insidious.161 Lung abscess typically develops approximately 1 to 2 weeks after aspiration of oropharyngeal or gastric material. The site of involvement is usually the lobe that was dependent at the time of aspiration. Fever is the most common sign in patients with lung abscess.153 Cough, dyspnea, and sputum production are present in approximately half of patients.153,162 Chest pain and hemoptysis can also occur. Putrid breath or a mass effect is occasionally the sole or predominant manifestation (Figure 36-4).163 The differential diagnosis of lung abscess includes other necrotizing infections such as tuberculosis, nocardiosis, fungal infections, melioidosis, paragonimiasis, and amebic abscess. Certain noninfectious diseases, such as malignancy, sarcoidosis, and pulmonary infarction, can produce lesions that mimic abscess on chest radiographs.

Diagnosis Necrotizing pneumonia is suspected in a child when the symptoms do not respond to appropriate antibiotic treatment of a pneumonic consolidation. A chest radiograph can reveal a radiolucent lesion, but CT is more discerning. Decreased parenchymal contrast enhancement on CT correlates with impending necrosis and cavitation.130 The radiographic diagnosis of lung abscess is based on finding an air–fluid level in a cavity at least 2 cm in diameter, with a well-defined wall.157 In about 20% of cases, a chest radiograph may not be

TABLE 36-5. Microbiology of Lung Abscesses in Childrena Organisms

Percent Cases

Aerobic and facultative bacteria

Staphylococcus aureus Streptococcus pneumoniae Other streptococci Haemophilus influenzae Pseudomonas aeruginosa Escherichia coli Other gram-positive organisms Other gram-negative organisms

19 10 32 6 13 9 7 6

Anaerobic bacteria

Bacteroides speciesb Prevotella melaninogenica Peptostreptococcus species Fusobacterium species Veillonella species Other gram-positive organisms Other gram-negative organisms

25 9 21 5 8 8 3

Fungi Mycobacteria a

10 1

Note: more than one organism can be isolated from a lung abscess b Includes some Prevotella melaninogenica (formerly Bacteroides melaninogenica). Data compiled from references 153, 155, 157, 159, and 160.

Figure 36-4. Anaerobic pleural empyema in a 5-year-old girl who came to medical attention because of a 1-month history of abdominal pain, tiredness, and constipation, but no history of an aspiration event, fever, respiratory distress, or cough. This radiograph was obtained after an acute respiratory event during evaluation for constipation. Note complete opacification of the left hemithorax with severe shift of the heart and trachea to the right. Three liters of putrid pus was drained, revealing a left lower lobe abscess. Gram stain and culture revealed polymicrobial anaerobic and facultative oropharyngeal flora. (Courtesy of E.N. Faerber and S.S. Long, St. Christopher’s Hospital for Children, Philadelphia, PA.)

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Persistent and Recurrent Pneumonia

diagnostic initially. Lung abscesses are most commonly found in the right upper, right lower, and left lower lobes.164 CT is often useful to define the extent of disease, underlying anomalies, and the presence or absence of a foreign body (Figure 36-5). Bronchoscopy is diagnostic, and therapeutic on many occasions to facilitate the removal of a foreign body or to promote the drainage of purulent fluid if this has not occurred spontaneously.155 Ultrasound or CT-guided transthoracic aspiration of lung abscess successfully identifies the etiologic agent in > 90% of cases.165 It is only required in complex cases or when the etiology cannot reasonably be ascertained from the clinical circumstances. Specimens for culture, other than those obtained by bronchoscopy or direct aspiration of the lung, are of limited value. Quantitative culture of bronchoalveolar lavage fluid improves the accuracy of identification of aerobic and anaerobic bacteria as causes of lung abscess.155,166

Management Prolonged antibiotic therapy is the mainstay of treatment for necrotizing pneumonia and lung abscess. Duration of therapy is based on clinical response and is usually 4 weeks or at least 2 weeks after the patient is afebrile and shows clinical improvement. Parenteral therapy is usually initiated. Two randomized clinical trials that involved 72 adults found clindamycin to be superior to penicillin for the treatment of anaerobic lung abscess.167,168 A clinical trial in children found no difference between these two drugs.169 Parenteral clindamycin is an appropriate empiric therapy for children with suspected S. Aureus (including MRSA) or anaerobic lung infection. Combination therapy with ticarcillin or piperacillin and a b-lactamase inhibitor, with or without an aminoglycoside, is considered when necrotizing pneumonia follows aspiration in a hospitalized child or in a child for whom an Enterobacteriaceae (e.g., Escherichia coli, Klebsiella, etc.) or Pseudomonas aeruginosa infection is suspected on clinical grounds (as in cystic fibrosis) or has been identified as an isolate from a percutaneous lung aspirate. Necrotizing pneumonia or abscess is frequently complicated by parapneumonic effusion, which benefits from percutaneous drainage or other invasive procedures (as mentioned earlier). Percutaneous abscess drainage157 carries the hazard of bronchopleural fistula, with prolonged morbidity or the necessity for surgical repair.157,170 Nevertheless, it should be considered in those patients with continued systemic illness, 5 to 7 days after initiation of antibiotic therapy,

Figure 36-5. Lung windows of computed tomography study showing right upper lobe abscess.

CHAPTER

37

257

especially if lesions are peripheral or if bronchoscopy fails to drain a more central lesion. Drainage may also be necessary if an abscess is > 4 cm in diameter, causes mediastinal shift, or results in ventilator dependency.171 Surgical wedge resection or lobectomy is rarely required, and is reserved for cases in which medical management and drainage fail or bronchiectasis has occurred.

Prognosis and Complications Necrotizing pneumonia in otherwise healthy children usually resolves with antibiotic treatment alone.153 Similarly, 80% to 90% of lung abscesses resolve with antibiotic therapy alone (provided that the bronchial obstruction is removed). Fever has an average duration of 4 to 8 days. The most common complication of lung abscess is intracavitary hemorrhage with hemoptysis or spillage of abscess contents and spread of infection to other parts of the lung.172 Other complications are empyema, bronchopleural fistula, septicemia, cerebral abscess, and inappropriate secretion of antidiuretic hormone.172

ACKNOWLEDGMENT The authors acknowledge significant contributions of K. McIntosh and M. Harper in the second edition of Principles and Practices of Pediatric Infectious Diseases.

CHAPTER

37

Persistent and Recurrent Pneumonia Dennis L. Murray and Chitra S. Mani In young adults with community-acquired pneumonia, clinical resolution occurs in more than 85% of cases, and radiographic resolution in approximately 75%, by 4 weeks after onset.1 Persistent pneumonia has been defined as continuation of symptoms and radiographic findings beyond this period.2 This concept provides a useful framework in which to begin to consider a variety of causes for persistence; in children, however, particularly when symptoms are resolving, what appears to be persistence represents merely a lag in radiographic resolution. Up to a third of adults with uncomplicated pneumococcal pneumonia have radiographic abnormalities for 6 to 8 weeks.1,3 Although pneumonia due to respiratory syncytial virus or parainfluenza virus typically clears within 2 to 3 weeks,4 pneumonia due to adenovirus can be necrotizing and can cause persistent abnormalities for up to 12 months.5 Recurrent pneumonia has been defined as occurrence of two or more episodes of pneumonia in a 1-year period or more than three episodes in any period, with radiographic resolution between episodes.2 Using this definition, approximately 8% of children requiring hospitalization for pneumonia would be identified as having recurrent pneumonia.6 In children with underlying conditions such as cystic fibrosis (CF), complete resolution does not occur between exacerbations. Radiographic documentation of episodes is essential for categorization, because precise clinical distinctions are made infrequently and “pneumonia” is the diagnosis sometimes conveyed to the parent (or the parent perceives) for conditions such as bronchiolitis, bronchitis, asthma, or persistent cough. Although a chest radiograph is not necessarily indicated to confirm the diagnosis of acute pneumonia in previously healthy outpatients, nor indicated routinely at the end of treatment of the first episode of acute pneumonia requiring hospitalization, a history of prior episodes and persistence or recurrence of symptoms are indications for initial and follow-up radiographic evaluations.7 In a child with a history of “recurrent pneumonia,” documenting a normal radiographic appear-

258

SECTION

D Lower Respiratory Tract Infections

ance 2 months after an acute episode (the time at which radiographic findings are expected to be normal in more than 90% of cases) is the most useful step in shaping a differential diagnosis. Before investigations are initiated for persistent or recurrent pulmonary infection, radiographs should be obtained and reviewed with a radiologist experienced in disorders of children to define the abnormalities precisely and to consider both infectious and noninfectious processes as well as underlying conditions.2 The differential diagnosis for and clinical approach to recurrent pneumonia in children are distinct from those for persistent pneumonia and also depend on: (1) whether the site of parenchymal disease is the same or different with each episode; and (2) whether the infiltrate is, on the one hand, dense, focal, and consolidated, or, on the other, atelectatic, patchy, diffuse, nodular, or interstitial. The specific approach to children whose pneumonia is associated with hospitalization, human immunodeficiency virus (HIV) infection, or some other form of immunologic deficiency is addressed in other chapters in this book.

PERSISTENT OR PROGRESSIVE PNEUMONIA AT A SINGLE SITE Pathogen-Related Causes Unresolved or untreated acute infection is usually responsible for persistent pneumonia (Table 37-1). In a patient receiving empiric antibiotic therapy, the cause is: (1) infection by an organism not eliminated by the chosen antibiotic, either because of antimicrobial resistance found in common bacterial pathogens (b-lactamaseproducing Haemophilus influenzae, Prevotella melaninogenica, penicillin-resistant Streptococcus pneumoniae, methicillin-resistant Staphylococcus aureus (MRSA), resistant gram-negative bacilli); or (2) more commonly, infection with organisms for which the chosen antimicrobial therapy may be ineffective (most often Mycobacterium tuberculosis, but may also include Mycoplasma or Chlamydiaceae spp., Francisella tularensis, viruses, fungi, protozoa, or parasites).8 Complications of appropriately treated bacterial pneumonia, such as necrotizing pneumonia, pleural effusion, and progression to lung abscess or empyema, or bronchiectasis must also be considered. The clinical, radiographic, and other imaging findings as well as the context of the patient’s illness help prioritize possible causes (Figure 37-1). Mycobacterium tuberculosis infection should always be considered in children with persistent pneumonia or subacute presentation even if exposure is not obvious. Immigration from a region of the world with a high prevalence of tuberculosis (Asia, Africa, Latin America, or Eastern Europe) or contact with individuals at high risk for tuberculosis (immigrants from regions in which tuberculosis is endemic; Native Americans; homeless persons; individuals with history of drug abuse or imprisonment; or persons with HIV infection) heightens the likelihood of tuberculosis.9 A history of weight loss, cough, and/or fever may not always be present. The radiographic abnormality is generally more impressive than the limited clinical findings, although occasional patients may manifest hectic fever, respiratory distress, and toxicity simulating acute pyogenic pneumonia. The presence of enlarged thoracic lymph nodes is a suggestive, but inconsistent, finding. Coxiella burnetii (Q fever) and Chlamydophila psittaci infections are considered in patients with persistent fever, malaise, and myalgia, hacking cough, persistent parenchymal infiltrate, and exposure to farm animals or psittacine birds. Systemic illness usually overshadows pulmonary symptoms in young children with Q fever; psittacosis is rare in young children. The likelihood of fungal pneumonia in a previously healthy child depends on place of residence or unusual environmental exposure. Nodular or miliary densities are more common than lobar consolidation, but not consistently. Histoplasma capsulatum is endemic to the Ohio and Tennessee river valleys, as well as eastern and central Canada. Subacute presentation with fatigue and a persistent cough is usual, and hilar lymphadenopathy is a clue. Blastomyces dermatitidis is a rare cause of persistent pneumonia in children. Endemic areas

TABLE 37-1. Diagnostic Considerations for Pneumonia at a Single Site Persistent or Progressive

Persistent or Recurrent

Untreated common acute infection Unresolved common acute infection Complication of acute infection Tuberculosis Uncommon infection

Atelectasis Segmental bronchiectasis Intraluminal obstructing lesions Foreign body Granuloma (infective or foreign body) Right middle-lobe syndrome Bronchial tumor (adenoma, papilloma, lipoma) Extrinsic obstructing lesions Lymph nodes (infective or malignant) Tumor Enlarged heart, pulmonary arteries Vascular ring, sling Congenital abnormalities Bronchial anomalies (bronchomalacia, bronchial stenosis or web, tracheal bronchus) Tracheobronchial cysts (cyst adenomatoid malformation, lobar emphysema, bronchogenic cyst) Pulmonary sequestration

Figure 37-1. Computed tomography of a 9-year-old boy with severe cerebral palsy, seizures, aspiration, and recurrent right lower lobe pneumonia showing severe bronchiectasis. Surgical resection was necessary. (Courtesy of J.H. Brien.)

partially overlap those of histoplasmosis, including parts of the Mississippi, Ohio, Missouri, and St. Lawrence waterways. Older children with Blastomyces pneumonia may produce purulent sputum and have often failed at least one course of treatment with an antibacterial agent prior to the correct etiology being identified. Coccidioides immitis is endemic to the southwestern United States, including west Texas, Arizona, New Mexico, California, Utah, and Nevada, as well as northern Mexico and parts of Central and South America. Erythematous rashes and erythema multiforme are frequent early in the course and erythema nodosum can accompany pneumonia (also seen in Blastomyces or Histoplasma infection). Since environmental exposure is often a cause of fungal pneumonia, clusters of cases within the same family or in other persons simultaneously exposed may occur with Histoplasma, Blastomyces, and Coccidioides infections. Rarely, a previously healthy child with or without undue exposure to dust, pigeon excreta, model terrarium, construction, or

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Persistent and Recurrent Pneumonia

home renovation develops pneumonia due to Cryptococcus or Aspergillus spp. that leads to a protracted infection. In adolescents at risk for HIV infection, Pneumocystis jirovecii (P. carinii) should be considered, especially when persistent pneumonia is bilateral, interstitial, and associated with hypoxemia disproportionate to the severity of clinical illness or physical findings. Risk factors for HIV infection (injection drug use, multiple sexual partners, bisexual partner) need to be evaluated. In the infant with HIV infection, P. jirovecii pneumonia, especially when not properly treated, is rapidly progressive and has a very high mortality. Viruses such as Hantavirus and influenza do not cause persistent infections but can cause rapidly progressive infections not responsive to antibacterial therapy. Adenovirus can also cause a persistent lobar pneumonia nonresponsive to antibiotics.

CHAPTER

37

259

The right middle lobe is predisposed to persistent atelectasis and subsequent infection or delayed resolution of pneumonia because of the acute angle and length of its bronchus, the proximity of its bronchus to hilar nodes, and poor to no collateral ventilation compared with other regions of the lung. This lobe is also a site of aspiration in the upright position. Recurrent right middle lobe pneumonia and atelectasis make up the so-called right middle lobe syndrome. Asthma is the most common noninfectious cause and tuberculosis is the most common infectious cause of this rather unique entity. In children with recurrent abnormalities of the right middle lobe, the workup should include bronchoscopy with direct visualization of the airway and obtaining material for cultures and cytology.14

Extrinsic Obstructing Lesions Host-Related Causes Multiple congenital or acquired anatomic abnormalities are also considered in children whose pneumonia: (1) fails to resolve; (2) responds symptomatically to antimicrobial therapy but does not resolve radiographically; or (3) resolves and then recurs at the same site (see Table 37-1). Atelectasis (parenchymal volume loss) should be differentiated from persistent infiltrate or consolidation (without volume loss). Atelectasis is commonly associated with respiratory syncytial virus bronchiolitis, airway hyperreactivity in infants and toddlers, asthma, and complete bronchial obstruction from intrinsic or extrinsic causes. Except with fixed obstructing lesions, atelectasis is expected to be transient or migratory; if atelectasis persists, the affected lung segment can become secondarily infected. Bronchiectasis, or dilatation of the bronchi, arises most commonly from damage to bronchial walls by infection and/or chronic inflammation. The bronchial wall is then susceptible to further dilatation and distortion during breathing. Severe acute viral, bacterial, or fungal infection or recurrent infection related to CF, immunodeficiency, or local obstruction are known causes of bronchiectasis.10 Adenovirus and measles virus infections, retained foreign body, and tuberculosis are most commonly associated with segmental bronchiectasis in the healthy host. In children with HIV infection or congenital immunodeficiency, lymphocytic interstitial pneumonia or recurrent bronchitis or pneumonia may progress to bronchiectasis.11 Data also now suggest that lower respiratory tract Mycoplasma pneumoniae infections can also be a cause.12 Once bronchiectasis has developed, impaired ciliary function and bronchial mechanics predispose to recurrent pneumonia. Children with extensive bronchiectasis are often easily fatigued, may have slower growth, and commonly have digital clubbing.

Intraluminal Obstructing Lesions Aspirated foreign body (or granulation tissue resulting from presence of a foreign body) is the most common cause of incompletely resolved or recurrent pneumonia at the same site among otherwise healthy children.2 In one-third of cases, an aspirated foreign body is not detected in the week after the event.13 The diagnosis is apparent if history of an event is elicited or a foreign body is visualized radiographically. Neither finding is present in most cases, but the event is frequently remembered by older children once the objects (e.g., timothy grass or other barbed twig) are retrieved at bronchoscopy. Spontaneous hemorrhage from the lower respiratory tract suggests foreign body (or pulmonary sequestration) and can be life-threatening. Pneumonia at a single anatomic site can be related to particular positional vulnerability in children who are relatively immobile and who aspirate oropharyngeal material because of impaired neuromuscular function or coordination. A dependent segment or segments of lung is/are usually involved. Other intraluminal bronchial obstructions can cause associated single-site pneumonia, including bronchial adenoma, lipoma, papilloma, foreign-body granuloma (e.g., peanut, other vegetable matter), granuloma of M. tuberculosis or atypical Mycobacterium spp. or segmental bronchomalacia or bronchial stenosis.

Extrinsic airway compression is most commonly due to lymph node enlargement. Tuberculosis is most common, causing hilar, carinal, and other superior mediastinal sites of compression, leading to secondary bacterial pneumonia. Pulmonary histoplasmosis, blastomycosis, and coccidioidomycosis are also important causes of hilar adenopathy.2 Tumors can cause compression of the airway directly or through lymph node involvement. Congenital or acquired heart disease associated with an enlarged heart can cause compression of left lower lobe bronchus especially, leading to pneumonia. Shunting procedures that cause excessive pulmonary blood flow can lead to airway compression or segmental congestion and impaired drainage, predisposing to localized infection.

Congenital Abnormalities of the Respiratory Tract Congenital abnormalities of airways or pulmonary parenchyma or the diaphragm can cause localized persistent or recurrent pneumonia (Figure 37-2). Tracheal bronchus, an abnormal bronchus arising from the trachea, leads to impaired drainage of the right upper lobe and persistent collapse. Congenital cystic anomalies of the tracheobronchial tree include cystic adenomatoid malformation, congenital lobar emphysema, and bronchogenic cysts15 (Figure 37-3). Congenital cystadenomatoid malformation (CCAM) and congenital lobar overinflation usually cause respiratory distress in infancy (and cystic abnormality on radiographs), whereas bronchogenic cysts and pulmonary sequestrations can manifest as chronic or recurrent pneumonia later in childhood. More than three-fourths of bronchogenic cysts become infected.16 Bronchogenic cysts usually appear radiographically as round or oval lesions with air–fluid levels in the perihilar area.17,18 Pulmonary sequestrations are masses of ectopic pulmonary tissue with a vascular connection but aberrant or no communication with the airways. Usually they are located in the lower lobes, more commonly on the left side, with blood supply arising from the aorta.15,19 Although frequently asymptomatic, pulmonary sequestrations may manifest as recurrent infection because of poor drainage of secretions.16 Congenital pulmonary cysts are located in the periphery of the lung. They arise during alveolar development and, unlike bronchogenic cysts, usually communicate with airways.16,17 Pulmonary cysts can cause respiratory distress in the neonatal period or manifest later in life as recurrent infections.

Approach to Diagnosis The history of the illness, associated symptoms, and environmental or contact exposures, a judgment about adherence to prescribed therapy, the possibility of partial improvement, and a re-evaluation of the appropriateness of prescribed therapy usually identify children with incompletely or inadequately treated acute uncomplicated pneumonia. Continued therapy or change in therapy with follow-up to document resolution is appropriate in many situations. An intradermal Mantoux skin test (5 TU) should be performed in all patients (with anergy assessment in selected individuals).

260

SECTION

D Lower Respiratory Tract Infections

B Figure 37-2. A 9-year-old girl with microcephaly had multiple episodes over multiple years of left lower lobe pneumonia (A) when barium study revealed congenital diaphragmatic hernia as the cause of persistent “consolidation” (B). (Courtesy of J.H. Brien.)

A

A Figure 37-3. Congenital cystadenomatoid malformation (CCAM) in a term neonate who was evaluated because of respiratory distress. Plain radiograph (A) and lung window of axial computed tomography (B) show typical cluster of cysts, varying in size. Note superiority of computed tomography in demonstrating sites and extent of pathology. (Courtesy of E.N. Faerber.)

Caregivers should keep in mind that symptoms of tuberculosis are frequently mild and that partial response to antibiotics is common in children whose tuberculosis causes intrinsic or extrinsic compression of the airway resulting in a secondary bacterial pneumonia. In infants, serial first-morning gastric aspirates are obtained for testing; they may be superior to specimens obtained by bronchoscopy (see Chapter 134, Mycobacterium tuberculosis). Sputum is obtained from older children and adolescents. HIV testing is indicated for all individuals in whom tuberculosis is diagnosed.

B

Serologic tests for Q fever, psittacosis, mycoplasmal infection, histoplasmosis, coccidioidomycosis, or blastomycosis are useful when the setting and clinical findings are compatible. Cross-reactivity may cause false-positive or false-negative test results. Serologic tests do not include all pathogens, are insensitive, and may not be easily or immediately available. Fungal pneumonia due to Candida, Cryptococcus, or Aspergillus can be difficult to diagnose. Testing for cryptococcal antigen and galactomannan antigen in the blood may be helpful diagnostically in those patients for whom other diagnostic

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Persistent and Recurrent Pneumonia

tests are nonrevealing. Bronchoalveolar lavage is usually the next diagnostic modality for persistent pneumonia. Bronchoalveolar lavage has been reported to confirm the cause of persistent pneumonia (including Blastomyces, Histoplasma, viruses, bacteria, and obstructing lesions) in 30% of immunocompetent children,20 in 27% of children with cancer,21 and in 70% of children with acquired immunodeficiency syndrome (AIDS).22 If P. jirovecii pneumonia is suspected in an adolescent, diagnosis can be attempted through identification of the organism with methenamine silver staining of an induced sputum specimen. If the result is negative or a specimen cannot be obtained, bronchoalveolar lavage is performed. Occurrence of Pneumocystis pneumonia requires pursuit of an underlying immunodeficiency (such as HIV infection, a congenital or therapy-induced immune defect). Patients with documented recurrent pneumonia, segmental bronchiectasis, suspected anatomic or obstructing lesions, or pneumonia that persists beyond 8 weeks of appropriate therapy require further evaluation. Poor weight gain, weight loss, digital clubbing, polycythemia, or anemia validates a history of chronicity and the need to proceed aggressively. Bronchoscopy is performed to: (1) exclude, or to detect and remove, a foreign body; (2) detect extrinsic compression or intraluminal anomaly; (3) obtain a biopsy specimens of a mass; or (4) obtain a specimen for microscopy and culture. Computed tomography (CT) is useful to evaluate more distal airways (including for bronchiectasis), mass lesions, tracheal bronchus, and congenital cystic anomalies, and to define precise relationships of cysts with surrounding structures before surgical excision.16,23,24 Additionally, CT is better than chest radiography to detect lung parenchymal complications (cavitary necrosis, abscess, bronchopleural fistula) and loculation of pleural fluid, and can better define pleural versus parenchymal disease.25 CT, magnetic resonance imaging, or Doppler ultrasonography can confirm the specific blood supply to sequestrated pulmonary tissue, obviating invasive studies such as angiography before lobectomy.16,19 Bronchiectasis can be diagnosed with CT as well; compared with bronchography, CT has a sensitivity of 98% and a specificity of 99%.26

CHAPTER

37

PERSISTENT OR RECURRENT PNEUMONIA NOT CONFINED TO A SINGLE SITE Causes of Dense Focal or Multifocal Infiltrates Primary pulmonary as well as a variety of nonpulmonary disorders can lead to recurrent and chronic pneumonia affecting multiple areas of the lungs (Table 37-2). Primary pulmonary disorders include inflammatory diseases (such as asthma), congenital and acquired causes of epithelial dysfunction (ciliary dyskinesia, CF, viral infection), and congenital and acquired structural abnormalities (laryngeal and tracheal anomalies, chronic lung disease in infancy). Nonpulmonary disorders that cause pneumonia are myriad, comprising: (1) conditions that impair or overcome normal pulmonary clearance mechanisms, such as impaired cough or gag reflex, neuromuscular disorders, and gastroesophageal reflux; and (2) conditions that increase the risk of infections in the lung, such as congenital and acquired immunodeficiency states and sicklecell hemoglobinopathies.

Respiratory Tract Disorders and Aspiration Asthma causes most cases of recurrent pulmonary atelectasis or infiltrates in children by inducing diffuse inflammation or mucus plugging of airways (Figure 37-4)27; common manifestations include recurrent wheezing and cough.27,28 Environmental stimuli (e.g., allergens, passive smoke) or infectious agents (e.g., viruses, Mycoplasma or Chlamydiaceae spp.) enhance inflammation and bronchial hyperreactivity. The majority of asthma exacerbations are associated with occurrence of viral infections.29 Nocturnal cough, protracted coughing after upper respiratory illnesses, and, even if not acutely ill, exercise-induced cough are important historical clues to asthma, especially in older chidren. Wandering atelectasis, segmental overaeration, and nonconsolidated, perihilar, peribronchial, and interstitial infiltrates are helpful radiographic clues that asthma is the underlying cause.

TABLE 37-2. Diagnostic Considerations for Recurrent or Chronic Pneumonia not Confined to a Single Site Dense Focal or Multifocal Infiltratesa

Diffuse Interstitial Infiltratesb

RESPIRATORY TRACT DISORDERS AND ASPIRATION

MISCELLANEOUS CONDITIONS

Asthma Cystic fibrosis Ciliary dyskinesia Bronchiolitis obliterans Recurrent aspiration (drugs, seizures, cricopharyngeal incoordination, neuromuscular disorders) Gastroesophageal reflux Laryngotracheal anomalies (laryngeal or submucosal cleft) Esophageal obstruction or dysmotility (webs, stricture, achalasia) Tracheoesophageal fistula Congenital abnormalities of heart and vessels

Bronchopulmonary dysplasia Pulmonary lymphangiectasia Hypersensitivity pneumonitis Allergic bronchopulmonary aspergillosis Vasculitic syndromes Desquamative interstitial pneumonitis Lymphocytic interstitial pneumonitis Histiocytosis Metastatic malignancies (neuroblastoma, Kaposi sarcoma) Alveolar proteinosis Idiopathic pulmonary fibrosis Drug, chemotherapy, radiation, or physical agent (smoke inhalation, kerosene, etc.) injury

IMMUNOLOGIC ABNORMALITIESC

Agammaglobulinemia Common variable immunodeficiency Immunoglobulin G subclass deficiency Cellular immunodeficiency Complement deficiency Phagocytic defects Immunodeficiency secondary to disease or drug OTHER DISEASES

Sickle-cell disease Pulmonary hemosiderosis a

Most pneumonitis is infectious. Most pneumonitis is noninfectious. c Infiltrates can also be diffuse interstitial or nodular, depending on pathogen. b

261

262

SECTION

D Lower Respiratory Tract Infections

B A Figure 37-4. A 9-year-old girl with asthma was hospitalized for treatment of “pneumonia.” Chest radiograph on admission shows dense right lung consolidation (A). Note decreased volume of right side of chest with deviation of trachea and heart to the right, suggesting atelectasis. Chest radiograph 2 days later, following chest physical therapy and treatment of asthma (B). Course is most compatible with mucus plug. (Courtesy of J.H. Brien.)

measure of ciliary function; delayed tasting of saccharin placed in the inferior nasal turbinate confirms dyskinesia.33

Cystic Fibrosis CF is the most frequently inherited disorder, affecting approximately 1 in 2000 to 2600 white children.30 More than 1000 mutations of the CF gene on chromosome 7 have been identified. CF is characterized by mucoviscid respiratory tract secretions that impair ciliary clearance, malabsorption due to failure of the exocrine pancreas, and increased salt loss in sweat. Chronic cough, poor weight gain, digital clubbing, sinusitis, and nasal polyposis at a preadolescent age are distinctive characteristics. Early colonization with Staphylococcus aureus, Haemophilus influenzae, and Streptococcus pneumoniae is followed by chronic infection of the lower respiratory tract due to Pseudomonas aeruginosa. Chronic infection with Burkholderia cepacia, Stenotrophomonas maltophilia, and MRSA can also occur. Chest radiograph may be abnormal in infancy and is never normal again, with multiple persistent and new areas of parenchymal consolidation, overinflation, and eventual bronchiectasis. Measurement of chloride concentration in sweat obtained by iontophoresis after pilocarpine stimulation is the diagnostic method of choice; most affected patients have a value > 60 mEq/L.31 Ability to detect CF mutations by direct DNA analysis for CF transmembrane conductance regulator (CFTR) on chromosome 7 has advanced the diagnosis and detection of heterozygotes. Although 70% of cases are due to F508 mutations, hundreds of mutations account for the remainder, none of which is responsible for > 2% of cases.

Primary Ciliary Dyskinesia Normal ciliary motility is necessary for adequate clearance of fluid and foreign materials from both the upper and lower respiratory tracts. Originally described as the Kartagener triad, consisting of dextrocardia, sinusitis, and bronchiectasis, primary ciliary dyskinesia always leads to recurrent sinopulmonary infections. Affected patients have a chronic productive cough, sinusitis, and otitis media; only half have situs inversus. Affected males have reduced fertility. Ultrastructural abnormalities of tubules forming the cilia prevent normal ciliary beating. The most common structural abnormality is absence of one or both dynein arms. Diagnosis is confirmed by histologic examination of specimens of respiratory tract columnar cells obtained by nasal brush biopsy or mucosal scraping.32 Samples should be obtained when the patient is not acutely infected. Morphologic abnormalities are accompanied by reduced frequency of ciliary beat. This can be tested with use of the saccharin transit time as a direct

Epithelial Damage Damage of epithelium due to infection by viruses, Bordetella pertussis, and Mycoplasma pneumoniae can lead to impairment of defense mechanisms and recurrent pneumonia.34 Additionally, exposure to cigarette smoke and other environmental pollutants impairs ciliary function.

Bronchiolitis Obliterans Bronchiolitis obliterans is a chronic lung disease where, in children, the majority of cases have followed lower respiratory tract infection, usually caused by adenovirus. The incidence is highest when children are infected between 6 months and 2 years of age. Certain populations of Native Americans and Maoris (New Zealand) have a higher incidence of chronic lung disease following pulmonary adenovirus infection in childhood. Bronchiolitis obliterans also occurs after inhalation of various acids and as a late complication following lung transplantation. Lesions, resembling those of bronchiolitis obliterans, have also been identified in bone marrow transplant patients associated with graft-versus-host disease. Bronchiolitis obliterans organizing pneumonia arises from masses of granulation tissue in alveolar ducts that obliterate airspaces. Dyspnea and cough, combined obstructive and restrictive functional abnormalities, radiographic visualization of hyperaeration, and airspace consolidation are typical. In occasional cases, the obliterative process is confined to one lung, which may then result in unilateral hyperlucent lung or Swyer–James syndrome. Patients with bilateral, diffuse obliterative disease often have pulmonary edema. While the clinical course, chest radiograph, and CT are suggestive, the diagnosis is confirmed by lung biopsy and pathologic examination.35 Corticosteroid therapy appears beneficial in some forms and stages of bronchiolitis obliterans.36

Neurologic Dysfunction Neurologic dysfunction is the most common nonpulmonary cause of recurrent pneumonia. Dysfunction predisposes to pneumonia as a result of pharyngeal incoordination or muscle weakness, which leads to failure of the gag, cough, or swallow reflex to protect the airway. Lesions of upper motor neurons, seizures, and decreased level of

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Persistent and Recurrent Pneumonia

consciousness commonly result in motor incoordination of the pharynx, predisposing to aspiration of oropharyngeal secretions. Children who aspirate frequently lose their cough reflex.14 Location of pneumonia depends on the person’s position at the time of aspiration. The middle lobe or lung bases tend to be involved when aspiration occurs in the upright position, whereas upper lobes are involved with aspiration in the supine position.37 Fluoroscopic examination during a barium meal can be performed to examine upper pharyngopalatal coordination. Although gastric fundoplication and placement of a gastrostomy tube are frequently performed in such patients for convenience of feeding, aspiration of oropharyngeal secretions continues. A sensitive and noninvasive method to detect chronic aspiration of oral contents is a radionuclide salivagram, using technicium-sulfur colloid.38

Gastroesophageal Reflux Gastroesophageal reflux can cause pneumonia in infants by spillover of gastric acid, particulate matter, or both into the trachea. A history of postprandial emesis, irritability accompanied by arching of the back (Sandifer syndrome), and stridor are recognized manifestations, but recurrent wheezing or pneumonia can dominate the clinical picture.37,39–41 In children, recurrent pneumonia from aspiration in association with gastroesophageal reflux most commonly occurs in those younger than 2 years of age.6,14 The most sensitive method for diagnosis is 24-hour intraesophageal pH monitoring,37 but false-positive test results occur. Barium swallow, esophageal manometry, and esophagoscopy are alternative diagnostic modalities, but they are less sensitive than pH probe. A barium swallow study has a sensitivity of approximately 50% for diagnosis of gastroesophageal reflux.42 Correlation of findings with causation of clinical pulmonary disease and the relative merits of medical or surgical therapies are controversial.43–45

Congenital Anomalies H-type tracheoesophageal fistula is a rare developmental abnormality that can escape recognition in the neonatal period and manifest later as recurrent pneumonia. Respiratory distress with feeding is typical. Radiographic demonstration of caudad-to-cephalad fistula to the trachea during retrograde filling of the esophagus with barium is diagnostic. Recurrent aspiration can persist after surgical repair because of esophageal dysmotility and pooling of secretions above the surgical anastomotic site.46 Clues to other congenital anomalies of the trachea, great vessels, or esophagus that cause recurrent pneumonia are usually present in associated symptoms or signs (positional respiratory distress, cough, stridor, feeding disorders).

Ventilator-Associated Pneumonia Ventilator-associated pneumonia (VAP) is defined as nosocomial (healthcare-associated) pneumonia in ventilated patients that develops > 48 hours after initiation of mechanical ventilation.47 In adults, VAP is responsible for both significant morbidity and mortality and prolonged hospital stay.48 Mortality associated with and the overall cost of VAP are not as well known in children. In one pediatric series, only 1% of patients admitted to a pediatric intensive care unit (PICU) acquired a nosocomial bacterial pneumonia, but the mortality rate for those patients was 8%. Risk factors for infection included immunodeficiency, immunosuppression, and neuromuscular blockade.49 VAP is considered to be the second most common nosocomial infection in the PICU.50 Bacteria that colonize the oropharynx can more easily evade host defenses in the presence of endotracheal intubation and neuromuscular blockade. VAP has also been studied in extremely premature infants in the neonatal intensive care unit (NICU) setting. In one study, VAP was a significant risk factor for infants born at < 28 weeks who stayed > 30 days in the NICU.51 Pseudomonas aeruginosa, Staphylococcus aureus, and enteric gram-negative bacilli are the most frequently associated organisms recovered from endotracheal aspirates in both premature infants and older children with VAP.51,52

CHAPTER

37

263

The diagnosis of VAP is made using criteria developed by the National Nosocomial Infection Surveillance System (NNIS).50 Clinical criteria differ by age; separate criteria exist for children < 1 year, > 1 and ≤12 years, and ≥13 years. Although no laboratory standard exists for VAP, cultures obtained through bronchoalveolar lavage or blind protected specimen brush may improve specificity.53 Prevention of VAP has been studied extensively in adults, whereas few comparable pediatric studies exist.50 The role of selective digestive decontamination or sucralfate in prevention of VAP in adults remains controversial.54 Intermittent enteral feeding compared with continuous feeding, while intubated, is associated with lower gastric pH and a lower risk of VAP in some studies.50 Aspiration is important in the pathogenesis of VAP in children. Elevation of the head of the bed 30° to 45°, subglottic suctioning, and use of noninvasive positive pressure ventilation may be effective in preventing VAP in children, as they have been in adults.50

Immunologic Abnormalities Immunologic abnormalities constitute an uncommon cause of recurrent pneumonia. The characteristic pattern of pulmonary involvement is recurrent, dense focal infiltrates at different sites, but diffuse interstitial or alveolar and interstitial infiltrates can occur, depending on the pathogen and defect. Qualitative or quantitative defects in phagocytic function, immunoglobulin synthesis, cellular immune function, or complement activity can lead to recurrent pneumonia.37 The etiologic agents of pneumonia and involvement of the sinopulmonary system exclusively, or in concert with other organ involvement, predict the presence and type of immune defect.

Immunoglobulin and Complement Deficiency Recurrent pulmonary infections caused by encapsulated organisms such as Streptococcus pneumoniae and H. influenzae suggest a quantitative or qualitative abnormality in immunoglobulins. Nontypable H. influenzae, Staphylococcus aureus, and Pseudomonas aeruginosa are also pulmonary pathogens associated with X-linked agammaglobulinemia and other antibody deficiency syndromes. Dysfunctional hypergammaglobulinemia of HIV infection is associated with sinopulmonary and invasive infection due to Streptococcus pneumoniae and other encapsulated organisms. Complement deficiency, although rare, can predispose to recurrent pulmonary pyogenic infection.

Phagocytic Cellular Defects Abnormalities in the number or function of phagocytic cells predispose to recurrent pyogenic infection, including pneumonia. Recurrent skin and soft-tissue infections, and other invasive infections, also occur commonly. The most common neutrophil disorder resulting in recurrent infections is chronic granulomatous disease (CGD), a microbicidal neutrophil defect preventing production of peroxidase. In the national registry of CGD patients, 80% had at least one episode of pneumonia.55 Infections are caused exclusively by catalaseproducing organisms and can be a clue to the defect. In the national registry, Aspergillus, Staphylococcus aureus, and Burkholderia cepacia were the most common causes of pneumonia in CGD patients. Other organisms, such as Serratia marcescens, Nocardia spp., and nontuberculous mycobacteria, can occasionally be involved.

Cellular Immune Defects Abnormalities of T lymphocytes, such as those due to congenital combined immunodeficiency, DiGeorge anomalad, HIV infection, or immunodeficiency secondary to disease or therapeutic agents, predispose to recurrent and progressive persistent infections, frequently pneumonia. Progressive disseminated viral, fungal, or mycobacterial infection and P. jirovecii pneumonia are suggestive of cellular immune defects.

264

SECTION

D Lower Respiratory Tract Infections

Other Conditions Acute thoracic, bony infarction or pulmonary infarction in a patient with sickle-cell hemoglobinopathy may lead to the “acute chest syndrome,” which can be difficult to distinguish from pneumonia and is a condition commonly precipitated or complicated by pneumonia (see Chapter 108, Infectious Complications in Special Hosts). Pulmonary hemorrhage with hemosiderosis can masquerade as recurrent pneumonia with alveolar, anatomically confined, “lobar” densities. Occurring predominantly in infants younger than 1 year and postulated to be related to cow’s milk protein allergy in some cases and exposure to molds in others, pulmonary hemorrhage with hemosiderosis is signaled by a characteristic abrupt onset of respiratory distress and panting with concurrent pallor and drop in hematocrit.56–58 Additional clues are rapid resolution of pulmonary infiltrate over 3 to 5 days (uncharacteristic of pneumonia related to infection) and the presence of hemosiderin-laden alveolar macrophages in bronchoalveolar lavage specimens. In older children, hemosiderosis and diffuse pulmonary disease are associated with collagen vascular diseases.

Causes of Diffuse Interstitial Infiltrates A variety of noninfectious entities cause chronic or recurrent lower respiratory tract symptoms and diffuse interstitial or alveolar infiltrates on chest radiographs that can be difficult to distinguish from pneumonia (see Table 37-2).59 Many of the chronic interstitial or diffuse lung diseases of children are rare, idiopathic disorders characterized by diffuse infiltrates, restrictive functional defect, and disordered gas exchange.60 Cardiac abnormalities include congestive heart failure, anomalous pulmonary venous return, and congenital or acquired inequality of pulmonary artery blood flow. Congenital lymphangiectasia or pulmonary veno-occlusive disease can be misdiagnosed as recurrent pneumonia or pulmonary edema.61 Idiopathic or autoimmune disorders that cause infiltrates include hypersensitivity pneumonitis, allergic aspergillosis, fibrosing alveolitis, Hamman–Rich syndrome, desquamative and lymphocytic interstitial pneumonitis, sarcoidosis, vasculitic syndromes, Wegener granulomatosis, and Goodpasture syndrome. Reticulonodular infiltrate suggests a differential diagnosis that includes Kaposi sarcoma, histiocytosis, metastatic neuroblastoma, and lymphoma.62 Children with HIV infections and chronic lung disease most commonly have lymphoid interstitial pneumonitis (see Chapter 111, Diagnosis and Clinical Manifestations of HIV Infection), but infiltrates can also be nonspecific, perhaps related to HIV itself.63 Many children with chronic interstitial lung disease have symptoms that may go unrecognized for years. Typical symptoms include cough, dyspnea, chronic tachypnea, exercise intolerance, and recurring respiratory infections. Diagnostic evaluation includes pulmonary function tests, high-resolution CT, and bronchoalveolar lavage. Lung biopsy is often required to determine the primary cause.

Approach to Diagnosis Clues from history, physical examination, growth and development, radiographic pattern, and confirmed causes of pneumonia or other infections should guide investigation and permit a staged evaluation. A precise history of the frequency, severity, duration, and sites of pulmonary infection should be obtained, and radiographs reviewed. Poor growth or the presence of digital clubbing suggests severity and chronicity. The history of associated respiratory tract signs or symptoms focuses the differential diagnosis. For example, stridor suggests tracheal anomalies; chronic cough productive of purulent sputum suggests bronchiectasis possibly due to disorders of antibody or phagocytic function, CF, or ciliary dyskinesia; severe and complicated sinusitis or otitis media suggests antibody deficiencies or ciliary dyskinesia; wheezing commonly indicates asthma but can be present in a variety of respiratory tract disorders (e.g., foreign body, CF, bronchiolitis obliterans). A history of recurrent problems confined

exclusively to the respiratory tract makes cellular immune defects and phagocytic defects unlikely. Infections involving other organ systems, such as recurrent boils or soft-tissue abscesses, raises the possibility of phagocytic defects. Involvement of the central nervous system, gastrointestinal tract and, in some conditions, development of arthritis, in addition to recurrent pneumonia, suggests the possibility of various forms of antibody deficiency. Confirmation of specific infectious agents heightens suspicion of certain disorders. For example, identification of catalase-producing agents may lead to a diagnosis of CGD; of Pseudomonas aeruginosa, to CF; of Burkholderia cepacia, to CF or CGD; and of Pneumocystis jirovecii, to cellular defects or HIV infection. A complete blood count is used to assess anemia or polycythemia; abnormal numbers, appearance, or differential makeup of leukocytes; and thrombocytopenia (associated with immunological disorders such as Wiskott–Aldrich syndrome, primary hematologic cytopenias, or suppression by disseminated virus infection) or thrombocytosis (associated with an acute inflammatory response). A Mantoux tuberculin skin test (5 TU), with anergy assessment, should be performed routinely. An underlying medical condition is identified in more than 90% of cases of recurrent pneumonia requiring hospital admission.6 In the otherwise healthy child, asthma is the most common diagnosis identified as a result of recurrent pneumonia.6 The following pattern makes the diagnosis of asthma more likely: recurrent wheezing, coughing exaggerated at night or dawn, positive family history, prolonged expiratory phase of respiration, radiograph showing airtrapping, and scattered peribronchial infiltrates or atelectasis. Findings such as poor linear growth and digital clubbing should not be ascribed to asthma. Spirometric measurement of reduced airway caliber, beneficial response to b2-agonist drugs, or 20% or greater reduction in forced expiratory volume in 1 second (FEV1) with methacholine or histamine challenge documents small-airway hyperreactivity typical of asthma. Assessment of clinical response to therapeutic interventions and follow-up examination including spirometry are appropriate first steps. In a 10-year retrospective review involving 238 children with recurrent pneumonia,6 the most common causes for recurrence of pneumonia were: (1) oropharyngeal muscular incoordination and the resulting difficulty in handling respiratory tract secretions; (2) seizure disorders; and (3) neurologic abnormalities (Table 37-3). A pattern of pneumonia in dependent pulmonary segments suggests aspiration. For other patients, certainly those with impaired growth or digital clubbing, and those with radiographs showing lobar or segmental

TABLE 37-3. Underlying Causes of Recurrent Pneumonia Cause Aspiration syndrome Immune disorder Malignant neoplasm Dysgammaglobulinemia Human immunodeficiency virus infection Autoimmune pancytopenia Congenital heart disease Asthma Airway or lung anomaly Tracheoesophageal fistula Cyst adenomatoid anomaly Vocal cord paralysis Subglottic stenosis Tracheomalacia Other Gastroesophageal reflux Sickle-cell anemia No identified predisposing cause

Percent of Cases 48 10 5 2 2 0.12 to ≤ 0.5 mg/mL)

Penicillin G or Cetriaxone plus gentamicin or Vancomycin (if unable to tolerate b-lactam agent)

4 weeks

6 weeks

4 weeks (gentamicin 2 weeks) 4 weeks

6 weeks 6 weeks

Resistant to penicillin (MIC > 0.5 mg/mL) (some streptococci and Abiotrophia, Granulicatella, Gemella spp.)

Penicilllin G (or ampicillin) plus gentamicin or Vancomycin plus gentamicin

4–6 weeks

Ampicillin (or penicillin G) plus gentamicin Vancomycin (if unable to tolerate b-lactam agent) plus gentamicin

4–6 weeks

Vancomycin plus gentamicin

6 weeks

Oxacillin (or nafcillin) plus

(optional gentamcin) Cefazolin plus (optional gentamicin) Vancomycin plus (optional gentamicin)

2 weeks if right-sided in IVDU ≥ 6 weeks plus and no complications gentamcin 2 weeks plus 4 weeks non-IVDU and no rifampin ≥ 6 weeks complications 6 weeks if complications (3–5 days) Penicillin-allergic nonanaphlactoid (3–5 days) Penicillin-allergic anaphylactoid (3–5 days)

Vancomycin

6 weeks

6 weeks

ENTEROCOCCI

Susceptible to penicillin, vancomycin, gentamicn

Resistant to penicillin, susceptible to vancomycin

6 weeks

STAPHYLOCOCCI

(Staphylococcus aureus or coagulase-negative staphylococci) Susceptible to oxacillin

Resistant to oxacillin

≥ 6 weeks plus gentamcin 2 weeks plus rifampin ≥ 6 weeks

MIC, minimal inhibitory concentration; IVDU, intravenous drug user. ModiÀed from reference 1.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Endocarditis and Intravascular Infections

together to maximize potential activity.1 If blood cultures are sterile but the patient’s symptoms respond to therapy, initially selected antibiotics are continued while other methods of diagnosis are pursued. If blood cultures are positive, deÀnitive antibiotic treatment is based on susceptibility test results.

Definitive Antibiotic Therapy Therapeutic recommendations for treatment of endocarditis depend on the pathogen, the susceptibility pattern, the presence of a foreign body in the heart (e.g., prosthetic valve), the need for surgery either to resect the vegetation or to replace an infected valve, and/or embolic complications.1 Regimens for common pathogens are outlined in Table 39-8; these regimens have been developed based on clinical experience primarily in adults. When clinical experience is limited, generally due to small numbers of cases of endocarditis with multidrug-resistant organisms, some recommendations are based on case reports or animal model data. Traditional therapy with outstanding cure rates for native valve endocarditis caused by penicillin-susceptible strains of viridans streptococci (minimal inhibitory concentration (MIC) ≤ 0.12 —g/mL) is penicillin G given intravenously for 4 weeks. Two weeks of therapy using a combination of penicillin plus an aminoglycoside (usually gentamicin) or 4 weeks of therapy with ceftriaxone1,74 has been effective in adults with endocardits. Endocarditis due to viridans streptococci relatively-resistant penicillin (MIC > 0.12 to 0.5 —g/mL) should be treated with a combination of penicillin G or ceftriaxone with an aminoglycoside.75 Endocarditis due to ampicillin-susceptible strains of enterococci is treated with a combination of ampicillin plus an aminoglycoside for 4 (if native valve) to 6 weeks (if prosthetic valve).1,76,77 Synergistic activity of the aminoglycoside is predictable if the isolate is susceptible to gentamicin (< 500 —g/mL) or streptomycin (< 1000 —g/mL).1 Some enterococcal spp., as well as Abiotrophia defectiva, Gemella spp., and Granulicatella spp., are inherently less susceptible to penicillin (MIC > 0.5 —g/mL). These latter three species were formerly known as nutritionally variant streptococci. Clinical studies

CHAPTER

Daily Dose a

Antimicrobial Agent

Children

Adults

Ampicillin

300 mg/kg

12 g

4–6 hours

Cefazolin

100 mg/kg

6g

8 hours

Ceftriaxone if native valve and PCN-susceptible (MIC ≤ 0.12 —g/ml)

100 mg/kgb

2 gc

24 hours

Penicillin G If native valve and PCN-susceptible (MIC ≤ 0.12 —g/ml)

200 000 U/kg

12–18 million U

4–6 hoursd

300 000 U/kg

24 million U

Dose Interval

If prosthetic material

300 000 U/kg

24 million U

If enterococci PCN-susceptible

300 000 U/kg

18–30 million U

Gentamicinb

5–6 mg/kg

3 mg/kg

8 hours or 24 hoursc

Oxacillin, nafcillin

200 mg/kg

12 g

4–6 hours

40 mg/kg

30 mg/kg

Children 8 hours Adults 12 hours

b

Vancomycin

275

have shown that use of a b-lactam agent alone frequently results in treatment failure or relapse, so the addition of gentamicin for 2 weeks is recommended.1,48,76 If the enterococcal isolate is resistant to ampicillin or the patient is allergic to penicillins, a combination of vancomycin and an aminoglycoside should be used for 6 weeks. During the past several years, vancomycin-resistant enterococci, particularly Enterococcus faecium, have emerged. Previously, management of vancomycin-resistant enterococcal endocarditis required multiple-drug, experimental regimens, which had been successful in some cases,78,79 but more recently there have been reports of successful treatment of endocarditis with linezolid,80,81 and with quinupristin/dalfopristin.51 Both Staphylococcus aureus and coagulase-negative staphylococci, primarily S. epidermidis, cause endocarditis and both can infect native valves, although S. aureus does so far more commonly.1 While most community-acquired strains of S. aureus in the United States remain susceptible to penicillinase-resistant penicillins, i.e., nafcillin or oxacillin, increasingly, community-acquired staphylococci are resistant to these agents.82,83 Patients with S. aureus or S. epidermidis resistant to oxacillin should be treated with vancomycin. To date, vancomycin-resistant S. aureus have been rarely described, but obviously are of great concern as treatment of endocarditis caused by such strains would prove very challenging. The addition of a second agent for treatment of staphylococcal endocarditis (generally gentamicin for 3 to 5 days) is optional, with the exception of infection of prosthetic valves or patients failing to respond to antimicrobial therapy (see Table 39-8).1 The duration of therapy varies from 2 weeks for uncomplicated right-sided native valve endocarditis caused by a oxacillin-susceptible strain to 4 weeks for uncomplicated left-sided native valve endocarditis to 6 or more weeks for endocarditis of a prosthetic valve or with a complicated course. Compared with streptococcal endocarditis, the treatment response of staphylococcal endocarditis is more protracted, with a mean 7 days to defervescence, 2 days to sterile blood culture, and fever for more than 14 days in 17% of cases in one pediatric study.11 Endocarditis caused by enteric gram-negative bacilli is uncommon in children. Two bactericidal agents are chosen on the basis of in vitro

TABLE 39-8. Doses of Commonly Used Antibiomicrobial Agents Recommended to Treat Infective Endocarditis

If native valve and PCN-resistant (MIC > 0.12 —g/ml)

39

MIC, minimal inhibitory concentration; PCN, penicillin. a Calculate maximum dose for large children and adolescents. b Dosage adjustment is necessary when renal insufÀciency exists. c Single daily dose not recommended for infective endocarditis caused by staphylococci if prosthetic material is present.

276

SECTION

E Cardiac and Vascular Infections

susceptibility; generally, a combination of an expanded-spectrum blactam agent and an aminoglycoside is used. Surgical intervention is often required.1 The HACEK group of organisms can be resistant to penicillin due to the production of a b-lactamase. As these organisms are very slow-growing, susceptibility testing is difficult to standardize. Thus, experts recommend therapy with a third-generation cephalosporin such as cefotaxime, ceftriaxone, or ampicillin-sulbactam for 4 weeks (if native valve) or 6 weeks (if prosthetic valve).1 Management of fungal endocarditis has traditionally included both medical and surgical approaches to optimize cure. Rare cases of fungal endocarditis have been managed successfully with fungicidal therapy alone.84,85 Traditionally, amphotericin or lipid amphotericin products have been used,86 but more recently in numerous case reports caspofungin87,88 and voriconazole89 have been advocated as more efficacious and less toxic therapeutic alternatives. Unfortunately, pharmacokinetic studies with these newer agents in neonates have not been performed. Notably, fluconazole prophylaxis has been advocated to prevent relapses or as suppression in patients not deemed to be surgical candidates.1

TABLE 39-9. Potential Indications for Surgical Intervention for Endocarditis Indication

Examples

Microbiologic

Inability to sterilize the blood > 7 days of optimal antibiotic therapy Fungal endocarditis Enterococci for which there are no synergistic combinations Left-sided endocarditis with Enterobacteriaceae

Vegetations

One or more serious embolic events occurring within the first 2 weeks of treatment Anterior mitral valve leaflet vegetation > 10 mm Persistent vegetation after embolization Increase in vegetation size despite adequate therapy

Valvular dysfunction

Cardiac failure that cannot be managed medically due to: • Rupture of a valve leaflet or chordae or valve perforation • Mitral or aortic valve insufficiency

Dosing and Duration Endocarditis is difficult to eradicate because the responsible microorganisms are enmeshed in avascular fibrinous vegetations. Because antibiotics must diffuse through this barrier, high concentrations of antibiotics with bactericidal activity must be maintained over an extended period. Thus, longer courses and higher dosages of antibiotics are needed to treat endocarditis than are commonly used for other childhood infections (see Table 39-8). Antibiotics are administered intravenously. Although there are reports of successful oral treatment of endocarditis, including with linezolide80,90 and voriconazole,89 this route has not been systematically studied and is not generally recommended, although there are reports of successful management of endocarditis in intravenous drug abusers who are not candidates for parenteral therapy.1 Duration of treatment is based on the pathogen and whether prosthetic cardiac material is present. In children with native valve endocarditis due to viridans streptococci, 4 weeks of penicillin G alone is effective; this regimen may be preferred for children because of limited experience with the 2-week regimen. A longer treatment course, 6 weeks, is recommended in “high-risk” situations, including infection due to relatively resistant streptococci or to enterococci, staphylococci, or gram-negative bacilli. Individuals with intravascular prosthetic material are generally treated for 6 weeks. More prolonged courses are recommended for complicated endocarditis (e.g., intracardiac abscess) and in patients who have a slow response to therapy. Regardless of surgical intervention, fungal endocarditis should be treated with prolonged courses of antifungal agents. There is a role for outpatient management for endocarditis. Although there are not controlled trials documenting safety of such an approach, careful selection of clinically stable patients at low risk for complications, specifically heart failure and/or embolic phenomena, is warranted, providing close follow-up is ensured.91

Surgery Timely surgical intervention may be critical to manage endocarditis effectively.92 The indications for surgical intervention have recently been modified by the American Heart Association expert advisory panel.1 These guidelines emphasize the need to individualize management of endocarditis, but provide indications for surgery, as shown in Table 39-9.93–96 Surgery should be strongly considered early in patients with congestive heart failure, at high risk for embolization and with prosthetic valve endocarditis; the risk of mortality may be substantially reduced by surgery. Generally, removal of prosthetic material (e.g., a patch, or conduit) in children is not necessary, with the exception of pacemaker wires.

Intracardiac extension Large abscess or extension despite treatment Valve dehiscence Fistula formation New heart block

Complications Complications of endocarditis include direct damage to the heart and heart valves. Complications can be caused by sterile or infected emboli and immunopathologic organ damage. Examples of direct damage to the heart are disruption of leaflet function by vegetations or chordae rupture, abscesses of the valve ring, and myocardial abscesses. A review of published cases of endocardial abscesses in children showed that a preponderance of cases involved native aortic valves and Staphylococcus aureus.97 Clues to a possible abscess were persistent fever and bacteremia; conduction abnormalities were less frequent than in adults. Many of the signs associated with endocarditis are due to emboli from cardiac vegetations. Right-sided endocarditis usually causes embolization and infection in the lung parenchyma. Necrotizing peripheral pneumonia and prolonged febrile course can occur; outcome with prolonged therapy is generally good. Left-sided endocarditis is associated with systemic emboli to brain, visceral organs, limbs, and coronary arteries. The greatest risk of emboli is associated with S. aureus, Candida, the HACEK organisms, and Abiotrophia.1 Septic emboli can result in hemorrhage, infarction, or abscess of involved viscera. Emboli to the central nervous system can result in stroke and/or mycotic aneurysms. Surgical resection of an enlarging mycotic aneurysm in the cerebral vasculature may be necessary if neurologic symptoms or bleeding ensues. Delayed splenic rupture due to emboli is a potentially fatal complication of endocarditis. Large emboli to major blood vessels can lead to cold, painful, and pulseless extremities, requiring immediate surgical intervention. Most embolic events occur within the first 2 weeks of therapy.

Prognosis Cure rates in endocarditis depend on the infecting organism, the site of endocarditis, embolic phenomena, and the underlying clinical state of the host. More than 90% of children with native valve endocarditis due to viridans streptococci have been cured.13,29,54,55,98 Enterococcal endocarditis, if treated with synergistic antibiotics, has a cure rate of 75% to 90%. The cure rate for S. aureus endocarditis is 60% to 75%,

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Endocarditis and Intravascular Infections

and the rate does not appear lower for methicillin-resistant strains. For gram-negative bacilli, the cure rate is lower. Fungal endocarditis and mold endocarditis have the poorest prognosis, with cure rates < 50% and < 20%, respectively. Features associated with a poorer prognosis include presence of prosthetic cardiac material, especially artiÀcial heart valves; extremes of age (infants and the elderly); and the formation of giant vegetations.99 Large (> 1 cm) left-sided vegetations, particularly of the mitral valve, appear to be associated with a higher risk of embolic events, especially during the Àrst 2 weeks of treatment and valve replacement.29,66,67,99,100 Persistence of fever for 2 weeks or longer has been shown to be associated with the presence of myocardial abscesses, requirement for urgent cardiac surgery, nosocomial complications, and death.101

Prevention It is estimated that between 4% and 19% of patients who have endocarditis have recently undergone a procedure associated with a high risk of bacteremia.44 In its 2007 guidelines, the AHA recommended that antibiotic prophylaxis for endocarditis should not be based soley on an increased lifetime risk of acquisition of endocarditis, but rather on high risk of adverse outcome from endocarditis. High risk conditons for which prophylaxis for dental procedures (and certain respiratory tract procedures such as tonsillectomy and adenoidectomy)4 as well as procedures on infected soft tissue is recommended limited to presence of: 1) prosthetic cardiac valve; 2) previous endocarditis; 3) congenital heart disease (CHD) limited to unrepaired CHD (including palliative shunts and conduits), completed repaired CHD with residual defect at or adjacent to prosthetic material; 4) cardiac valvulopathy following cardiac transplantation.4 Antibiotic prophylaxis to prevent endocarditis is no longer recommended for gastrointestinal or genitourinary tract procedures.4 (see Table 39-2 and Chapter 8, Chemoprophylaxis). A recent model has proposed that thousands of courses of antibiotics are prescribed to prevent potentially a single case of endocarditis.102

OTHER “ENDOCARDITIS-LIKE” INTRAVASCULAR INFECTIONS In addition to endocarditis, there are other types of infection that can present as persistent bacteremia or fungemia. These include, most commonly, catheter-related bloodstream infections as well as thrombus-related infections following catheterization of either central veins or, less commonly, peripheral veins, so-called septic (or suppurative) thrombophlebitis.103 Fortunately, septic thrombophlebitis is relatively rare in pediatrics; in one study, the incidence was 0.12% of all admissions, but the rate is likely lower today.104 Risk factors for these types of infection are the prolonged (> 96 hours) use of an intravascular catheter, surgical manipulation of a vessel, infection of adjacent nonvascular structures that spreads to involve the vasculature, or severe burns. Intravascular catheter placement can be complicated by formation of sterile thrombus or a sterile Àbrin clot, generally at the distal tip of the catheter.105 Phlebitis can occur from trauma or irritation of the vessel wall due to infused particles such as can occur among intravenous drug abusers. Similarly, thrombus may develop at the site of a surgical anastomosis or at the juncture of a vascular prosthesis. The clot or inflamed vessel can become infected during bacteremia or infection from a distal site or in the case of superÀcial thrombophlebitis from the skin or contaminated parenteral fluid.106 Suppurative thrombophlebitis occurs if the thrombus or the vessel wall becomes infected during this process. Notably, vessel wall infection occurs rarely during catheter-related bloodstream infection. Progression of infection and inflammation can occlude the vessel, leading to an enlarged, thickened, and tortuous vessel.

CHAPTER

39

277

Septic thrombophlebitis can also result from adjacent nonvascular infections. Examples include Lemierre syndrome, i.e., internal jugular vein septic thrombophlebitis arising from pharyngitis or peritonsillar abscess;107 pylephlebitis, i.e., portal vein septic thrombophlebitis following appendicitis or other intra-abdominal surgery;103 septic thrombophlebitis of the dural sinuses, or pelvic vein thrombosis following septic abortion, pelvic surgery, or pelvic abscess. Although septic thrombophlebitis of the portal vein, pylephlebitis, is rare in children in the antibiotic era, hepatic abscesses may occur as infected friable thrombi break off and enter the liver. SuperÀcial thrombophlebitis may lead to subperiosteal abscesses of the long bones presenting with bone tenderness and signs of inflammation.108 Similarly, infected deep-vein thrombosis has been associated with pulmonary embolus and osteomyelitis.109 The superior vena cava syndrome can occur due to septic thrombophlebitis of the thoracic central veins following insertion of central venous catheters. Symptoms of superÀcial septic thrombophlebitis include fever, chills, induration, warmth, erythema, and tenderness of the infected vessel. Notably, only about half of patients with central septic thrombophlebitis have inflammation and symptoms may not occur until several days after the catheter has been removed.110 As these infections are endovascular in nature, high-grade bacteremia, hypotension, sepsis, and septic emboli are also complications. Metastatic foci often provide the clue to the diagnosis. For example, Lemierre’s syndrome can present with cavitating pulmonary lesions. Symptoms of septic thrombophlebitis of the portal vein include fever, abdominal pain usually in the right upper quadrant, anorexia, weight loss, and malaise. The diagnosis of septic thrombophlebitis is often delayed, particularly as symptoms and signs of intravascular infections are sometimes not referable to the site of the lesion. Computed tomography scans, magnetic resonance imaging, and/or ultrasound may demonstrate thrombophlebitis. Radionuclide scans rarely demonstrate such infections and ultrasonography is only diagnostic if there is a large abscess or vegetation at an extracardiac site, although occlusion or obstruction of a vessel after catheter removal may be detected. Causative organisms are generally recovered from blood, but can also be cultured from pus at the exit site, aspiration of the thrombosed vein, or culture of the surgically resected vessel. S. aureus is the most common etiology of suppurative thrombophlebitis, but Enterobacteriaceae or fungi may also be causative agents. Vaginal flora, including Bacteroides spp., streptococcal spp. and gram-negative bacilli cause pelvic thrombophlebitis and gram-negative pathogens and Enterococci cause intra-abdominal thrombophlebitis. Lemierre syndrome is usually caused by Fusobacterium necrophorum. Septic thrombophlebitis can be life-threatening, particularly if the diagnosis and treatment are delayed. Treatment of suppurative thrombophlebitis requires removal of the catheter, ligation of the infected vessel, and total removal of the infected segment plus the use of antibiotics directed against the organisms isolated from the blood or purulent material or thrombus within the infected segment.103–106 If bacteremia continues, surgical resection of the involved vascular segment or removal of an infected vascular prosthesis may be necessary. Like cardiac vegetations, these lesions are not vascularized, and penetration of antibiotics into such lesions may be poor. Thus, principles of therapy are similar to those for valvular endocarditis. In the absence of thrombophlebitis, bacteremia with an indwelling central venous catheter such as a Broviac catheter can frequently be treated with antibiotics without removing the catheter (see Chapter 102, Clinical Syndromes of Device-Associated Infections). Unfavorable bacteriology (e.g., enterococci, fungi, resistant gram-negative bacilli), as well as cardiovascular instability and complicating underlying conditions, usually mandates immediate catheter removal (see Chapter 102, Clinical Syndromes of Device-Associated Infections). If there is evidence of a tunnel infection (erythema and pain along the catheter tract) or if antibiotics fail to resolve the bacteremia, the device must be removed. Exit site infection, that is, local purulence at the site of insertion, can frequently be treated with a combination of local care and systemic antibiotics.

278

SECTION

CHAPTER

E Cardiac and Vascular Infections

40

Myocarditis Joseph A. Hilinski

Myocarditis can manifest in a variety of ways, ranging from nonspecific systemic symptoms such as fever and myalgia, to fulminant heart failure or sudden death. The condition is defined as “inflammation of the heart muscle,” and generally refers to diseases not associated solely with valvular abnormalities. There is continued debate regarding appropriate diagnosis, classification, and management of myocarditis.1,2

ETIOLOGY Myocarditis can be a manifestation of almost every infectious agent. For most cases in routine clinical practice, a specific etiology is not found. Viruses remain the most common cause in North America and Europe, and are often presumed to be causative in cases without a proven etiology. Initial studies using serologic assays implicated enteroviruses as common causes of myocarditis, particularly coxsackievirus B serotypes 1 through 5.3,4 Newer studies using molecular techniques such as polymerase chain reaction (PCR) assays of endomyocardial biopsies confirm the importance of enteroviruses as causative agents of acute myocarditis (25% to 35%), as well as of idiopathic dilated cardiomyopathy (10% to 30%).5–7 In addition, using PCR techniques, other viruses such as adenovirus7 and parvovirus B195,8 have emerged as important and frequent causes of myocarditis. Influenza, cytomegalovirus, and Epstein–Barr virus have also been implicated, and myocarditis is recognized as a complication during outbreaks of mumps, measles, influenza, and polio.7,9,10 Human immunodeficiency virus and hepatitis C virus have been implicated as important causes of myocarditis; however, the exact role each virus plays in causing disease is unclear.2 Bacteria less frequently cause myocarditis than viruses. Invasion of the bloodstream by any bacterial pathogen can result in myocardial seeding and microabscesses, such as in complications of Staphylococcus aureus, Neisseria meningitidis, Salmonella, and other bloodstream infections.11–13 Myocarditis can be a toxin-mediated complication of tetanus or diphtheria, or can be caused by other bacteria such as Borrelia burgdorferi (occurring in up to 10% of patients with Lyme disease), Rickettsia spp. (especially scrub typhus), and Mycoplasma pneumoniae.14–16 Parasites are a major cause of myocarditis worldwide, with Trypanosoma cruzi (the causative agent of Chagas disease) being the principal cause in South America, with as many as 20% of infected patients developing chronic heart failure.17 Many additional agents have been reported to cause myocarditis in immunocompromised hosts, most importantly cytomegalovirus, Toxoplasma gondii, T. cruzi, Cryptococcus, Candida, and Aspergillus species.

EPIDEMIOLOGY The incidence of myocarditis can coincide with the occurrence of epidemic enterovirus infections. Enteroviruses cause approximately 5 to 10 million symptomatic infections annually in the United States, and occur in all human populations. The rates of infection vary by geographic location, season of the year, and age, although infections can occur across all age groups. In the United States enterovirus activity peaks in the summer and fall months, although in other parts of the world transmission is year round.18 Young children are most susceptible, probably because they lack cross-reacting immunity from few or no prior infections.19 Serologic studies show higher rates of

enteroviral illness among people with myocarditis and controls with similar exposures than among people without common exposures.4 Chagas disease is endemic in South and Central America, and is a leading cause of myocarditis in those regions. Chronic heart failure in these patients is probably precipitated by immune activation.1 Autopsy series are helpful in demonstrating the incidence of myocarditis as a cause of mortality. A 25-year review of sudden death in military recruits implicated myocarditis in 20% of cases.20 Other studies using PCR techniques highlight the possible role of myocarditis in sudden infant death syndrome (SIDS).21,22 Enteroviruses, parvovirus B19, and adenoviruses were among the most commonly detected viruses in cases associated with SIDS.

PATHOPHYSIOLOGY Pathogenesis The pathogenesis of myocarditis has been gleaned in large part from animal models. Both direct myocardial invasion by viruses and host autoimmune responses are important in the pathogenesis of disease. Direct myocyte invasion by infectious agents can initiate the process, which in later stages results in the development of both CD4+ helper and CD8+ cytotoxic T-lymphocyte responses. Additionally, circulating antiheart antibodies directed against contractile, structural, and mitochondrial proteins have been described.2 Mouse models show that exercise,23 cold exposure, malnutrition, pregnancy, and immune suppression can worsen clinical disease in disseminated enteroviral infections, which may also be true in humans.24 A common coxsackievirus-adenovirus receptor (CAR) on cardiac myocytes has been described, which may offer a partial explanation as to why these two groups of viruses are prone to cause myocarditis. Reports have indicated increased expression of CAR in mouse models of myocarditis, as well as in humans with endstage dilated cardiomyopathy, although the overall significance of CAR is not yet clear.25

Pathologic Findings Several standardized criteria have been developed to classify myocarditis using histopathology. The Dallas criteria divide biopsy findings into those of active myocarditis (inflammatory cellular infiltrate with evidence of myocyte necrosis), borderline myocarditis (inflammatory cellular infiltrate without evidence of myocyte injury), and no myocarditis. Inflammatory infiltrates are then described as lymphocytic, eosinophilic, or granulomatous; with mild, moderate, or severe inflammation; and focal, confluent, or diffuse distribution.26 In 1991 a second classification system was described by Lieberman to incorporate clinical features into diagnostic consideration, but this system is seldom used.26a Regardless of which classification system is used, pathologic guidelines likely underestimate cases of myocarditis for several reasons, most notably biopsy sampling site error, and interobserver variability. Myocarditis from bacterial, parasitic, and fungal causes can show specific findings. Bacterial and fungal infection result in polymorphonuclear cell infiltrates, which can be focal or organized into microabscesses. Trypanosomes can often be visualized in the biopsy specimen in chronic Chagas disease. Giant-cell myocarditis is a rare disorder in which multinucleated giant cells in the absence of granulomas are found on biopsy. A specific link to an infectious agent has not been identified, although an autoimmune process is suspected.2

CLINICAL MANIFESTATIONS The presentation of myocarditis can be highly variable, but usually occurs in the setting of a systemic infection. Cardiac manifestations can range from an asymptomatic state with electrocardiogram (ECG) abnormalities to cardiogenic shock. A viral prodrome is often reported by patients; this consists of influenza-like complaints such as fevers, myalgia, gastroenteritis, chest pain, dyspnea, and tachypnea.27 This can be followed abruptly by hemodynamic collapse. Sudden death can

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Myocarditis

be a presenting sign of myocarditis in all age groups, particularly in infants and children.28 Some cases of SIDS are due to myocarditis.21,22 Young infants and children frequently have nonspecific findings such as feeding difficulties, irritability, respiratory distress, and newonset murmur or cardiac findings. In a report of 13 children with myocarditis, 10 had a preceding “viral illness,” 11 presented with tachypnea, 2 had chest pain, 2 presented with Stokes–Adams attacks (syncope related to arrhythmia), and 1 had a seizure. Congestive heart failure was present in all, as were ST-T-wave changes, cardiomegaly, and pulmonary edema on chest radiograph.29 Typical findings of heart failure can also be present, such as pulmonary rales, distant heart sounds, gallop rhythm, jugular venous distention, hepatomegaly, cardiomegaly, cyanosis, poor perfusion, delayed capillary refill, and shock.

DIFFERENTIAL DIAGNOSIS Many conditions mimic infectious myocarditis. Most involve systemic inflammatory conditions, such as autoimmune diseases, drug reactions, envenomation, endocrinopathies, radiation, and transplant rejection (Box 40-1). Several cardiac diseases can present similar to or in conjunction with infectious myocarditis. Endocarditis should be distinguishable by primary valvular disease in the presence of positive blood cultures and echocardiographic evidence of vegetations. Pericarditis can be distinguished by the findings of precordial chest pain, signs of pericardial fluid, and absence of arrhythmia. Myocardial infarction can have findings similar to myocarditis, whether infarction

BOX 40-1. Noninfectious Causes of Myocarditis in Children • INFLAMMATORY DISEASES Inflammatory bowel disease Systemic lupus erythematosus Polymyositis/dermatomyositis Mixed connective tissue disease Kawasaki disease Rheumatic fever Rheumatoid arthritis Sarcoidosis Thyrotoxicosis Transplant rejection Giant-cell myocarditis • DRUG REACTIONS Adriamycin Alcohol Amitriptyline Amphetamines and methamphetamines Arsenic Catecholamines Cefaclor Cocaine Cyclophosphamide Daunorubicin Furosemide Isoniazid Lead Lithium Methyldopa Penicillins Sulfonamides Tetracyclines • BITES AND STINGS Bee Scorpion Snake Spider Wasp • IDIOPATHIC CAUSES (LYMPHOCYTIC) • PERIPARTUM PERIOD • PHEOCHROMOCYTOMA • RADIATION THERAPY

CHAPTER

40

279

is due to coronary artery atherosclerosis, or is secondary to drug ingestions such as cocaine and methamphetamine. The term idiopathic (or lymphocytic if characteristic lymphocytic infiltrates are seen) myocarditis is used to describe cases in which no causative agent is found.

LABORATORY FINDINGS AND DIAGNOSIS A high index of suspicion is required to make the diagnosis of myocarditis accurately in children. If the diagnosis is considered, useful initial tests include chest radiography, ECG, and echocardiography. Other nonspecific laboratory findings are leukocytosis, elevated sedimentation rate, eosinophilia, or an elevation in the cardiac fraction of creatine kinase (CK-MB). Measurements of troponin I and T may be useful in the diagnosis of myocarditis because of higher sensitivity compared with other enzymes tested.30 ECG findings are highly variable in myocarditis; however, suggestive changes include ST-segment elevations in two contiguous leads (54%), T-wave inversions (27%), widespread ST-segment depressions (18%), and pathological Q waves (18% to 27%).2 Lowvoltage complexes can be commonly observed in standard and precordial leads. Other related findings can include tachycardia out of proportion to fever, arrhythmias, and conduction disturbances. Occasionally ST-segment and T-wave abnormalities can be associated with events other than myocarditis, including fever, hypoxia, electrolyte disturbances, and minor childhood viral infections.31 Many ECG abnormalities resolve in 1 to 2 months. Echocardiography is useful to confirm and quantify impaired systolic cardiac function. In addition, exclusion of other causes of cardiac dysfunction such as valvular vegetations or pericardial effusion is important diagnostically and therapeutically. Common echocardiographic findings are segmental or global abnormalities of motion of the heart wall due to inflammation and edema of the myocardium.32 Repeated echocardiograms are useful to follow evolution of disease, with persistent abnormalities suggesting development of dilated cardiomyopathy. Other imaging techniques useful for diagnosis of myocarditis include nuclear imaging with gallium67-labeled or indium111-labeled antimyosin antibodies, and magnetic resonance imaging (MRI). Nuclear imaging techniques with gallium67 or indium111 detect the extent of myocardial inflammation and myocyte necrosis, respectively, and may be highly sensitive for detecting myocarditis.2 MRI is a promising diagnostic modality. Small observational studies show its utility in demonstrating focal myocardial enhancement of diseased areas, which may be useful for pinpointing sites for endomyocardial biopsy.33 Endomyocardial biopsy remains the gold standard for establishing the diagnosis of myocarditis, despite its limited sensitivity and specificity. In large series of patients with cardiomyopathies using the Dallas criteria, positive biopsy results are reported in only approximately 10% of patients.34,35 In general, biopsies performed within weeks of symptom onset have a higher yield than biopsies done later. Numerous studies have demonstrated the usefulness of biopsy to identify possible causative agents,5,7,8,36 to establish the diagnosis, and to follow patients to determine whether myocarditis is active, healing, or healed. Specific diagnosis of infectious agents causing myocarditis can be made in several ways. Agents can be isolated directly from myocardial tissue, visualized using specific stains, or amplified with molecular techniques such as PCR. Indirect associations can be made by isolation or demonstration of likely infectious agents from other tissue or body fluids, such as finding positive nasopharyngeal, throat, or stool cultures for viral pathogens; blood cultures for agents such as Neisseria meningitidis; or rapid antigen assays for respiratory tract viruses. In the acute setting, blood can be obtained for PCR analysis for common viruses such as enteroviruses, adenovirus, cytomegalovirus, Epstein–Barr virus, and others. Serologic assays showing a fourfold rise of titer in paired sera, or a single positive immunoglobulin M assay can implicate infectious causes, but have limited usefulness in the immediate workup and management of myocarditis. Epidemiologic clues such as time of year and geographic location may be

280

SECTION

E Cardiac and Vascular Infections

helpful in evaluating for etiologies, such as in the midst of an influenza or enteroviral epidemic, or to suspect Chagas disease in a patient from South America.

MANAGEMENT Supportive therapy is the most important initial management of myocarditis. Many patients have mild myocarditis as a complication of systemic infection and can be managed with minimal intervention. Patients with symptoms of congestive heart failure or fulminant disease require careful management in an intensive-care setting. Diuretic and vasodilator agents may be necessary to lower ventricular filling pressures. If congestive heart failure is present, an angiotensinconverting enzyme inhibitor agent should be used to decrease vascular resistance, and a beta-blocking agent may be indicated once the patient is stable clinically.1 Digoxin should be used with caution and only at low doses, because a mouse model of myocarditis indicated increased mortality with use of digoxin.37 Arrhythmias may require pharmacologic therapy, or placement of a defibrillator. Bedrest should be recommended for most patients, given the potential for exacerbation of disease with exercise. In extreme cases implantation of a ventricular assist device or extracorporeal membrane oxygenation may be required. Specific infectious agents can be treated with antimicrobial agents. Myocarditis due to bacterial, fungal, or parasitic causes should be treated with appropriate antimicrobial therapy. Diphtheric myocarditis should be treated with a combination of antitoxin and antimicrobial therapy. Clinical trials have not been done to evaluate efficacy of specific antiviral agents for presumed or proven viral myocarditis. However, if a treatable cause of myocarditis can be established, therapy with an antiviral agent may be indicated. Such agents include: neuraminidase inhibitor agents for influenza, ganciclovir for cytomegalovirus, acyclovir for varicella-zoster virus or herpes simplex virus, cidofovir for adenovirus, intravenous ribavirin for parainfluenza, and respiratory syncytial viruses. Pleconaril, an investigational agent active against enteroviruses, is not currently available for compassionate use. Many studies have been done evaluating immune-modulating agents for the treatment of myocarditis with controversial results, using agents such as systemic corticosteroids, immune globulin intravenous (IGIV), interferons, and other immunosuppressive drugs. The high incidence of spontaneous recovery after myocarditis necessitates studies utilizing a control group for accurate analysis for efficacy. In one of the largest trials of use of immunosuppressant agents, the Myocarditis Treatment Trial, 111 patients with biopsyproven myocarditis were randomized to receive either placebo or an immunosuppressive regimen consisting of prednisone and either cyclosporine or azathioprine. No difference in mortality was found between the two groups, and improvement in left ventricular function was identical at the 28-week follow-up.34 Another controlled trial using prednisone 60 mg daily as immunosuppression for unexplained dilated cardiomyopathy showed an initial improvement in left ventricular ejection fraction at follow-up at 3 months in the treated group, which was not sustained at 6 or 9 months.38 A meta-analysis was performed to review the use of immunosuppressive regimens for acute myocarditis in children. The analysis concluded that data suggest that immunosuppressive therapy does not significantly improve outcomes in children.39 For these reasons, and findings in other similar studies, immunosuppression should not be used routinely in the treatment of myocarditis. Use of IGIV as therapy for myocarditis has also been controversial. One pediatric trial suggested benefit of high-dose IGIV (2 g/kg) in 21 children compared with historical controls, but other larger, randomized controlled studies in adults do not support a consistent clinical benefit.40,41 Therefore, the long-term benefit of this treatment remains unknown. Single-center trials of interferon-alpha and interferon-beta have been performed showing benefit in patients with dilated cardiomyopathy and myocarditis. Data need to be confirmed in a large-scale, multicenter trial.42,43

COMPLICATIONS AND PROGNOSIS Complications of myocarditis include congestive heart failure, arrhythmias, cardiac rupture, sudden death, and the progression to dilated cardiomyopathy. One study evaluating risk factors for fulminant progression in acute myocarditis showed higher likelihood of shock occurring in patients with elevated serum C-reactive protein level, elevated serum creatine kinase concentration, decreased left ventricular ejection fraction, and intraventricular conduction disturbance(s) at the time of admission.44 Follow-up data are available from a series of 41 children with suspected myocarditis, of whom 27 (66%) made a complete recovery, 4 (10%) had incomplete recovery, 5 (12%) died, and 5 (12%) underwent cardiac transplantation. Medical management of this cohort varied significantly, with some patients receiving corticosteroid therapy, some receiving IGIV, and some receiving neither therapy.45

PREVENTION Prevention of acute infectious myocarditis relies on prevention of the underlying microbial causes. Since most cases are caused by viruses such as enteroviruses and adenoviruses, targeting prevention of these agents is most likely to be helpful. Routine control measures such as frequent hand hygiene after contact with infectious body fluids, covering the mouth while coughing, and food and water sanitation should be effective.

CHAPTER

41

Pericarditis Joseph A. Hilinski

Pericarditis is inflammation of the pericardium and the proximal part of the great vessels. Pericarditis can be acute, subacute, or chronic in presentation. Pericarditis can have an infectious or noninfectious cause; in either case, pericarditis may be the sole manifestation of a disease or part of a multisystem disorder. Pericarditis can manifest as cardiac tamponade with a fulminant, life-threatening process, constrictive pericarditis from chronic disease, or as an incidental finding of pericardial fluid in an asymptomatic person. Extensive guidelines have been published by the European Society of Cardiology on the diagnosis and management of pericardial diseases.1

ETIOLOGY Pericarditis caused by infection can be classified as purulent, “benign,” or granulomatous. Purulent pericarditis is caused by bacterial infection; benign pericarditis is caused by viral infection or occurs in hypersensitivity, postinfectious, or postpericardiotomy syndromes; and granulomatous pericarditis is usually caused by Mycobacterium tuberculosis and, occasionally, by fungal infection. Up to 90% of acute cases are “idiopathic” or presumed to be of viral origin. Noninfectious causes of acute pericarditis include cardiac injury, uremia, radiation, neoplasia, collagen vascular disease, sarcoid, inflammatory bowel disease, and drug-induced. Viral pericarditis is more common in adults than children. This condition can be clinically subtle, detected because of an enlarged heart on a routine chest radiograph, or fulminant, with severe hemodynamic manifestations. A viral cause is assumed if pericardial fluid is not purulent and spontaneous resolution occurs. Enteroviruses are the most common causes of viral pericarditis, can be associated

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Pericarditis

with seasonal epidemics of virus circulation, or can occur in clusters.2 Other viral causes (often associated with myocarditis) have also been implicated, such as adenovirus, influenza, herpes simplex, cytomegalovirus, Epstein–Barr virus, mumps, lymphocytic choriomeningitis, varicella-zoster, and human immunodeficiency virus (HIV). Cytomegalovirus is an important cause of pericarditis in immunocompromised and HIV-infected people.3 Bacteria invade the pericardium as a result of bacteremia or by contiguous spread from the lungs or pleural space. Staphylococcus aureus and Haemophilus influenzae type b (Hib) historically have been the most frequent causes of purulent pericarditis in children (Table 41-1).4,5 Both pathogens usually cause pericarditis in children under 5 years of age, although occasionally older children and adults are infected.6 The frequency of pericarditis caused by Hib has decreased as a result of universal childhood immunization. Purulent pericarditis can rarely be caused by anaerobic bacteria, either by contiguous spread (especially pulmonary actinomycosis)7 or bacteremic spread (especially Bacteroides fragilis).8 HIV infection is an important predisposing factor worldwide for pericarditis, in particular for that due to bacterial and mycobacterial causes.9 Pericarditis caused by M. tuberculosis is uncommon, occurring in about 1% of cases of tuberculosis, either by direct extension from a pulmonary focus or spread from lymphatic or bacteremic origin. Although an uncommon cause of pericarditis in developed countries, M. tuberculosis can cause up to 70% of cases of pericarditis in areas where tuberculosis remains a major public health problem.10 Pericarditis is occasionally caused by fungal infection after surgery, instrumentation, immunosuppression, or neutropenia.11 Fungi associated with pericarditis include Candida species, Aspergillus, Blastomyces, Histoplasma, Cryptococcus, and Coccidioides species. Less common causes of pericarditis include Rickettsia spp., Mycoplasma pneumoniae,12 M. hominis, Ureaplasma urealyticum, Entamoeba histolytica, Toxoplasma gondii, and Toxocara canis. Inoculation at surgery, invasion associated with instrumentation of the genitourinary tract, and immunosuppression can also result in pericarditis.

EPIDEMIOLOGY Pericarditis is not a reportable disease, and population-based studies are not available. Pericarditis commonly occurs in the late summer and fall because the most common viral agents, the enteroviruses, are epidemic during this time. Pericarditis occurs in people of all ages, although viral pericarditis is more common in adults than in children. In contrast, purulent pericarditis appears to be more common in children; one-half of all cases reportedly occur in children under 13 years of age, with equal incidence in males and females. Tuberculous pericarditis is more common in poor urban populations and among immigrants from countries where tuberculosis is endemic, and is the most common cause of pericarditis in high-risk regions.10

TABLE 41-1. Causes of 163 Cases of Purulent Pericarditis, 1950–1977 Causative Organism Staphylococcus aureus Haemophilus influenzae type b Neisseria meningitidis Streptococcus pneumoniae Salmonella spp. Escherichia coli Othera Unknown a

Number of Isolates (%) 72 (44) 35 (22) 14 (9) 9 (6) 4 (3) 3 (2) 7 (5) 17 (10)

Includes isolates of Klebsiella spp., Streptococcus pyogenes, “Paracolon spp.,” Pseudomonas aeruginosa, Staphylococcus epidermidis, Bacteroides spp., anaerobic streptococci, singly or in combination. Data from Feldman WE. Bacterial etiology and mortality of purulent pericarditis in pediatric patients: a review of 162 cases. Am J Dis Child 1979;133:641.

CHAPTER

41

281

PATHOGENESIS The pericardium consists of two layers, the inner layer, or visceral pericardium, which is continuous with the outer tissue of the myocardium, and the parietal pericardium, which lines the surrounding mediastinal structures. The two layers are 1 to 2 mm thick, with a space between them that contains 10 to 15 mL of clear fluid in healthy children and 15 to 35 mL in adults. The pericardial tissue is normally thin and semitransparent. There is independent blood supply from the internal mammary arteries, and innervation from the phrenic nerve, explaining the severe pain and vagally mediated reflexes that sometimes occur with inflammation. With inflammation, there is influx of fibrin, polymorphonuclear and mononuclear cells, and exudation of fluid into the pericardial space. Pericardial inflammation results in proliferation of fibrous tissue, neovascularization, and scarring, with consequent loss of elasticity and restriction of cardiac filling. Purulent pericarditis frequently results from contiguous extension of pneumonia, empyema, suppurative mediastinal lymphadenitis, liver abscess, or cardiac conditions, such as myocarditis, myocardial abscess, and infectious endocarditis. Bacterial pericarditis can result from inoculation during bacteremia, as commonly it is the pathogenesis in pericarditis due to Staphylococcus aureus infection. Pericarditis can complicate placement of pacer devices or during sternal osteomyelitis/mediastinitis postoperatively. S. aureus, S. epidermidis and nonenteric and enteric gram-negative bacilli are usually responsible. Enteroviruses are transmitted primarily by the fecal–oral route. After initial proliferation in the lymphoid tissues of the intestine, viremia can lead to disseminated or focal infection in skin, brain, meninges, heart, and pericardium. Pericarditis can be the sole manifestation of enteroviral infection or part of multiorgan infection. During an enteroviral epidemic there is a wide range of expression of disease; pericarditis affects only a small proportion of infected children. Myocarditis is often present in people with pericarditis due to enteroviruses. Regardless of cause, acute pericarditis results in collection of fluid in the space between the visceral and parietal pericardium. Small increases in fluid production are clinically insignificant because they are reabsorbed. Large increases in fluid secretion that exceed resorptive capacity can result in significant cardiac dysfunction. As a result of the ability of the heart to dilate in response to long-standing stress, slowly accumulating collections of 1 liter or more of fluid in adults may not interfere with cardiac function. However, 200 to 300 mL of fluid accumulation, if rapid, can exceed maximal distensibility of the pericardial sac. When this occurs, small incremental fluid accumulation causes a sharp increase in intracardiac pressure, which interferes with cardiac filling, leading to decreased stroke volume (cardiac tamponade); death from low cardiac output occurs if fluid is not removed urgently. Tuberculous pericarditis results from lymphatic spread from a focus in the lung or lymph nodes, or by hematogenous spread from a distant site. Granulomas of the pericardium containing M. tuberculosis develop initially and are followed by a serous or serosanguineous effusion containing lymphocytes and monocytes. Healing of tuberculous pericarditis results in deposition of fibrin and collagen, often leading to constrictive pericarditis.

CLINICAL MANIFESTATIONS Precordial chest pain is the most common symptom of pericarditis. This pain is often unrecognized in young children, in whom manifestations are irritability and a grunting expiratory sound as they splint the thoracic cage. Exercise intolerance and fever are common but nonspecific manifestations of pericardial infection. Enteroviral infection is often characterized by fever, malaise, and rash. Pericarditis frequently follows an upper respiratory tract infection, thus fever, weakness, and chest pain after nonspecific febrile or upper respiratory

282

SECTION

E Cardiac and Vascular Infections

tract illnesses should suggest consideration of the diagnosis of pericarditis. Pain is felt over the entire precordium, to the left side over the trapezius ridge, and over the scapula; pain sometimes radiates down the arm and can be aggravated by movement. Pain is generally worse when supine, and can be relieved by sitting. Pain can also be referred, because pericardial pain fibers are located in the diaphragmatic reflection of the pericardium. Pain is more common in acute infectious pericarditis than in more indolent forms. Examination of the heart can reveal muffled heart sounds caused by the surrounding effusion and increasing tachycardia as the effusion impinges on the volume of chambers. A pericardial friction rub may be audible, especially when the effusion is small. The rub is best heard during deep inspiration and with the patient kneeling or in the knee–chest position, leaning forward. The rub is typically a to-and-fro, high-pitched loud rasping or creaking sound, heard throughout the cardiac cycle, although it can be limited to systole. Pleural rubs can be differentiated because they occur in timing with the respiratory cycle. Clinical manifestations of tamponade include tachycardia, peripheral vasoconstriction, reduced arterial pulse pressure, and pulsus paradoxus. Pulsus paradoxus represents a drop of > 10 mmHg in systolic blood pressure during inspiration due to decreased venous return to the heart. Pulsus paradoxus is not pathognomonic for pericardial tamponade; it can occur in the presence of chronic lung disease and in pericarditis without tamponade. Other findings described in pericarditis include the Kussmaul sign (a rise or failure to fall of jugular venous pressure with inspiration), and Beck triad (hypotension, muffled heart sounds, and raised jugular venous pressure).

DIFFERENTIAL DIAGNOSIS Chest pain and fever occur with pneumonia, empyema, myocarditis, or pleurodynia. Myocardial infarction, a common cause of chest pain among adults, is rare in children. Dressler syndrome is a form of pericarditis associated with infarction that can occur weeks to months later. However, myocardial ischemia can follow use of crack cocaine and crystal methamphetamine (“ice”) in adolescents and adults.13 Noninfectious causes of pericardial effusion include blunt trauma, malignant tumors, and radiation. Any direct injury can cause pericarditis. Postpericardiotomy syndrome is considered in the 6 weeks after cardiac surgery. Systemic inflammatory disorders, such as collagen vascular diseases, Kawasaki disease, rheumatic fever, and polyserositis, following infection with organisms such as Neisseria meningitidis, can cause pericarditis; these rarely lead to tamponade. Metabolic diseases, such as uremia and myxedema, and exposure to certain drugs, including hydralazine, procainamide, and phenytoin, can also cause pericarditis.

LABORATORY FINDINGS AND DIAGNOSIS The diagnosis of pericarditis is made on the basis of history, physical examination, and imaging studies. All patients with suspected pericarditis should undergo electrocardiogram (ECG), chest radiography, and echocardiogram.1 The specific cause of pericarditis is best determined by examination of pericardial fluid. Complete cell count and morphology, glucose level, lactate dehydrogenase, and protein concentrations are useful to support likely causes. Serosanguineous or hemorrhagic fluid is suggestive of tumor, trauma, tuberculosis, toxoplasmosis, and streptococcal infection. Systemic leukocytosis and elevated acute-phase reactants are common. Troponin I can be raised in some patients, and unlike in acute coronary artery syndromes, may not be a negative prognostic marker.14 An increase in the size of the cardiac shadow on chest radiograph, especially in the absence of pulmonary congestion, suggests presence of pericardial fluid. The epicardial fat pad can be visualized within the left borders of the cardiopericardial silhouette in about 15% of patients

with pericardial effusion. Concomitant findings of pneumonia or findings suspicious for tuberculosis can be helpful etiologic clues. Typical features of pericarditis on ECG are generalized ST segment elevations without reciprocal ST depression, except in leads V1 and aVR. A few days into illness, elevated ST segments return to baseline. At this time there is flattening or inversion of the T waves. Low-voltage QRS complexes may be evident without the pathologic Q waves of myocardial infarction; these findings result from damping of electrical activity by effusion. With resolution of disease the electrocardiographic results normalize, although T-wave abnormalities may persist. Echocardiographic examination of the heart and its surrounding structures is the most valuable tool for diagnosis and assessment of the severity of pericarditis, and may differentiate between pericarditis, myocarditis, and endocarditis. When pericardial fluid is present, both M-mode and two-dimensional echocardiography demonstrate a sonolucent space between the two layers of pericardium. M-mode is the more sensitive procedure and can reliably estimate the volume of effusion. Two-dimensional echocardiography can be used to direct catheter placement for drainage. Transesophageal echocardiogram may be necessary in certain cases, including suboptimal transthoracic views, evidence of complex diastolic dysfunction, or suspected constrictive pericarditis. Further imaging with computed tomography (CT) or magnetic resonance imaging (MRI) has an important role in the assessment of complications of pericarditis. CT is useful to delineate extracardiac masses and other causes of an enlarged cardiac silhouette: combined studies with flow imaging by MRI are useful to delineate intracardiac masses.15 Microbiologic evaluation of pericardial fluid obtained by pericardiocentesis is the best method for determining the specific cause of pericarditis. In one report on patients with large pericardial effusions undergoing subxiphoid pericardial biopsy, a diagnosis was established in 93% of cases, whether or not the cause was infectious.16 In another study of patients undergoing pericardiocentesis without suspicion of purulent pericarditis, the diagnostic yield was under 10%.17 Evaluation of fluid should include Gram, acid-fast, and fungal stains and culture for bacteria, viruses, mycobacteria, and fungi. Special attention to rapid transport under anaerobic conditions is necessary to optimize recovery of anaerobic bacteria. Blood cultures should be obtained and are positive in up to 70% of patients with purulent pericarditis.18 If pericarditis results from extension from a contiguous focus, cultures of sputum, pleural fluid, and mediastinal fluid or tissue may identify the causative organism. Viral causes of pericarditis are determined by culture, rapid antigen tests, serologic tests, and, in certain cases, molecular genetic techniques. Tissue culture should be inoculated for recovery of viruses; suckling mouse inoculation enhances recovery of certain enteroviruses. Consideration can be given to obtaining polymerase chain reaction (PCR) assays for common viral agents, such as enteroviruses, directly from pericardial fluid, or indirectly from blood.19 However, even with these procedures, recovery of virus from pericardial fluid is unusual. Paired sera can be tested for antibody to prevalent enterovirus serotypes. A fourfold rise between acute and convalescent serum titers for enteroviruses or the presence of specific immunoglobulin M antibody (depending on technique and laboratory) is considered diagnostic. Because of the large number of serotypes, nontargeted testing is impractical. Virus isolated from a site other than the pericardial fluid, such as the stool or throat tissue, can be considered to be the likely cause of concomitant pericarditis. Rapid antigen tests from respiratory tract specimens can be helpful in implicating some viral agents. A rise in serum antibody to an isolate confirms active infection. Serology also offers the best opportunity for diagnosis of Rickettsia and Mycoplasma species infection. Recovery of mycobacteria and fungi is uncommon, but appropriate cultures for these agents should be performed in cases of granulomatous pericarditis. When these agents are suspected, biopsy of the pericardium for histologic examination and culture has a higher yield than direct stain or culture of pericardial fluid alone. If tuberculous pericarditis is suspected, a tuberculin skin test should be

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Pericarditis

performed. PCR for M. tuberculosis directly from pericardial fluid can offer the advantage of rapid diagnosis, but the sensitivity and speciÀcity of PCR need to be determined.20 Postcardiotomy syndrome occurs in up to one-third of children and one-Àfth of adults following open-heart surgery.21 This syndrome is characterized by fever, chest pain, and pericardial friction rub. An autoimmune process, triggered by viral infection or surgical trauma, is the postulated mechanism of disease.

41

283

TABLE 41-2. Antimicrobial Therapy of Major Causes of Pericarditis Antimicrobial Agent

Dosage (mg/kg per day)

Enterovirus and other viruses

Nonea

–

Empiric therapy for purulent pericarditis

Vancomycin or Nafcillinb or Oxacillinb plus Cefotaxime or Ceftriaxone

60 200 200 200 80 (qd) or 100 (divided bid)

Oxacillin or Nafcillin

200 200

Vancomycinc

60

Ampicillin (if susceptible) or Cefotaxime or Ceftriaxone

200–300

Causative Agent

MANAGEMENT Patients who have a small pericardial effusion of apparent viral cause can be managed with bedrest, symptomatic relief of pain, and close clinical monitoring. The presence of a large effusion with tamponade or impending tamponade requires immediate removal of fluid by pericardiocentesis or open drainage, possibly with resection of the anterior pericardium (pericardiectomy). Patients with suspected bacterial or fungal pericarditis should undergo pericardiocentesis promptly. Echocardiographically guided pericardiocentesis provides immediate relief of acute clinical deterioration. Pericardiectomy, the deÀnitive procedure, is indicated if the fluid is too thick to withdraw through a small-bore tube, if fluid persists after pericardiocentesis, or if the process is chronic and constrictive. Whether pericardiectomy can be limited to producing a small subxyphoid “window” or resection of the whole anterior pericardium should be performed is controversial. For cases of purulent pericarditis, irrigating the pericardial cavity is mandatory. A small case series and case reports indicate that irrigation with streptokinase may help liquefy the exudate.9,22 Therapy with broad-spectrum antimicrobial agents is initiated empirically if purulent pericarditis is suspected (Table 41-2). Bacterial pericarditis is usually treated for 3 to 4 weeks with parenteral antimicrobial agents. Given the recent increased rates of communityassociated methicillin-resistant Staphylococcus aureus (MRSA), empiric therapy for presumed purulent pericarditis should include vancomycin.23 Candidal pericarditis almost always requires combined medical and surgical treatment.11 Fungal pericarditis caused by disseminated Aspergillus spp. or dimorphic fungi (Histoplasma, Coccidioides, Blastomyces) is usually treated for several months with amphotericin B preparations, although triazoles such as itraconazole, fluconazole, and voriconazole can be used in certain settings (see chapters on speciÀc pathogens). Empiric antituberculous therapy using three or four drugs should be initiated if tuberculous pericarditis is suspected; corticosteroids may also be indicated. Drainage of purulent pericardial fluid is mandatory to decrease the immediate risk of death. Use of appropriate antimicrobial agents and drainage has improved outcomes for patients with pericarditis. Of 162 children with purulent pericarditis reported in 1979,4 death occurred in 82% of children who were given only antimicrobial therapy treatment and 22% of children who had both antimicrobial therapy and drainage. Young age was an independent risk factor for poor outcome. Children < 1 year of age who were infected with either Staphylococcus aureus or Haemophilus influenzae had a mortality rate of 63%, which is substantially higher than the mortality rate (26%) for children > 1 year of age.4 In a 1994 report,18 42 of 43 patients with purulent pericarditis treated with both antimicorbial agents and drainage recovered completely. The major long-term complication of pericarditis is thickening and stiffening of the pericardial membrane leading to development of

CHAPTER

Therapy when agent is conÀrmed Staphylococcus aureus or Staphylococcus epidermidis (susceptible to methicillin) Staphylococcus aureus or Staphylococcus epidermidis (resistant to methicillin) Haemophilus influenzae

Streptococcus pneumoniae (susceptible to penicillin) or Streptococcus pyogenes Streptococcus pneumoniae (intermediate to penicillin) Streptococcus pneumoniae (highly resistant to penicillin) Mycobacterium tuberculosisd

Penicillin G

200 80–100 (divided bid) 200 000–300 000 units

Cefotaxime

300

Vancomycin

60

Isoniazid + Rifampin + Pyrazinamide

10–15 mg for 6 months 10–20 mg for 6 months 20–40 mg for 2 months 15–25 mg until susceptibility studies are available

± Ethambutol

a

Cytomegalovirus may respond to ganciclovir. Nafcillin or oxacillin should only be used if community rates of methicillinresistant Staphylococcus aureus (MRSA) are signiÀcantly low. c Linezolid would be an alternative; some community-associated MRSA are susceptible to clindamycin. d Addition of corticosteroids to antibiotic regimen is recommended (e.g., prednisone 1 mg/kg per day for 6–8 weeks); in communities where isoniazid-resistant Mycobacterium tuberculosis has been encountered, a four-drug initial treatment is recommended. b

constrictive pericarditis, which usually necessitates surgical stripping of the pericardium. In one series, surgical pericardiectomy was required in 8 of 24 (33%) patients with tuberculous, purulent, or neoplastic pericarditis, versus only 1 of 203 (< 1%) with idiopathic, viral, or Toxoplasma pericarditis.24 Pericarditis can recur in 15% to 30% of people in whom the disease is idiopathic. Fever and pericardial fluid collections in patients with the postpericardiotomy syndrome usually resolve with aspirin, nonsteroidal anti-inflammatory agents, or corticosteroid therapy.

SECTION

F

Central Nervous System Infections

Acute Bacterial Meningitis Beyond the Neonatal Period Xavier Sáez-Llorens and George H. McCracken, Jr.

ETIOLOGIC AGENTS AND EPIDEMIOLOGY

284

PATHOGENESIS AND PATHOLOGY Development of bacterial meningitis progresses through the following five related steps: (1) bacterial colonization of the nasopharynx; (2)

2.0 1.5 1.0 0.5 0

Persons aged ≥ 5 years

States reporting Children aged 90 pneumococcal serotypes (4, 6B, 9, 14, 18F, 19F, and 23F) account for > 80% of invasive disease in children from developed countries; serotypes 5 and 1 are also prevalent in developing countries.3 The incidence of invasive disease, including meningitis, caused by S. pneumoniae has been reduced by > 90% after the implementation of universal heptavalent conjugate vaccination in United States infants.4 Serogroups B and C are the most common serogroups of N. meningitidis causing invasive infections in the American continent, although group Y strains can account for up to 20% to 25% of cases. A conjugated vaccine against meningococcal serogroup C was introduced in 1999 in the British routine immunization schedule, showing an overall 81% reduction of serogroup C disease within 2 years of implementation.5 Children with a basilar skull or cribriform fracture and a cerebrospinal fluid (CSF) leak have greater risk for pneumococcal meningitis, as do children with asplenia (anatomic or functional) or human immunodeficiency virus (HIV) infection. Deficiencies in terminal components of complement lead to greater risk for meningococcal infection. Common causes of meningitis after penetrating head trauma or neurosurgery are staphylococcal species, streptococci, and gram-negative enteric bacilli, especially Escherichia coli, Klebsiella spp., and Pseudomonas aeruginosa. These organisms are also associated with meningitis related to a dermal sinus or

Rate

42

embryopathy of the neurenteric canal. Nontypable H. influenzae meningitis is associated with immunoglobulin deficiencies, and Listeria meningitis with cellular immune defects; Listeria meningitis occasionally occurs in immunocompetent infants and children as well. Salmonella spp. rarely cause meningitis in immunocompetent children beyond the neonatal period, and some cases have been linked to reptile pets. The presence of a ventriculoperitoneal shunt is a risk factor for meningitis; ventriculitis related to shunt contamination and skin organisms are usually causative. Isolation of an organism other than pneumococcus, meningococcus, or Hib from the CSF of a child older than 2 months always requires an explanation or evaluation for unusual host susceptibility. Children with recurrent pneumococcal or meningococcal infections should also undergo thorough investigation, including neuroimaging studies. Patients with cochlear implants have more than a 30-fold increased incidence of pneumococcal meningitis.6

Rate

CHAPTER

1997

Year * Per 100 000 children aged 1000 cells/mm3, few or no WBCs may be present in CSF in the early phase (usually

of rapidly progressive infection). Most WBCs are usually polymorphonuclear (PMNs); presence of immature PMNs is suggestive of bacterial infection. Normally, CSF in children older than 6 months contains < 6 WBCs/mm3 and no PMNs.15 Only 2 of 122 children with proven bacterial meningitis in one study had CSF with < 6 WBCs/mm3.16 The protein concentration is elevated (mean, 100 to 200 mg/dL), and the glucose concentration is usually depressed (CSF-toserum ratio less than 0.6), but the relative severity of abnormalities depends on the offending organism as well as the cadence and duration of infection. Depending on the observer, the CSF Gram stain smear is positive in 80% to 90% of patients with untreated bacterial meningitis. Table 42-2 offers a differential etiologic diagnosis based on findings usually encountered in CSF for various meningeal pathogens. Detection of polysaccharide antigen in CSF by latex agglutination is most reliable for Hib (85% to 95%), followed by S. pneumoniae (50% to 75%) and N. meningitidis (33% to 50%).13 Antigen is readily detected only in urine of children with Hib meningitis but can also be detected after immunization with Hib conjugate vaccines. There is no reliable method to detect the polysaccharide of N. meningitidis serotype B. Thus, in the Hib vaccine era, rapid antigen tests have limited value in untreated patients, are infrequently more discriminating than Gram stain, and cannot be used to limit therapy. Many children with bacterial meningitis receive oral or parenteral antibiotics before diagnosis. In general, prior oral therapy at standard doses does not alter CSF findings to a point that prevents a final diagnosis of bacterial meningitis. The most comprehensive information about the impact of oral antibiotics on CSF abnormalities, however, relates to Hib meningitis.17 Similar data for large numbers of patients with pneumococcal or meningococcal meningitis are not available; the impact could be greater for exquisitely susceptible pathogens and when a parenteral dose of a cephalosporin is administered. Positive results of antigen detection tests are most helpful in patients in whom CSF is abnormal but results of Gram stain and culture are negative. Although measurement of serum C-reactive protein concentration can be useful to distinguish bacterial from viral meningitis in many patients, the main value of this component appears to be related to the detection of complications or treatment failures after serial measurements. A CSF leukocyte aggregation test has been suggested to help distinguish between bacterial and aseptic causes. One study

TABLE 42-2. Usual Cerebrospinal Fluid (CSF) Findings in Children with Meningitis Caused by Various Microbial Etiologies CSF finding Leukocytes/mm

Viral 3

Bacterial a

Partially Treated Bacterial

Lyme

Fungal

TB

< 1000

> 1000

> 1000

< 500

< 500

< 300

Polymorphonuclear cells

20–40%b

> 85–90%

> 80%

< 10%

< 10–20%

< 10–20%b

Protein (mg/dL)

N or < 100

> 100–150

60–> 100

< 100

> 100–200

> 200–300

c

Glucose (mg/dL)

N

UD to < 40

< 40

N

< 40

< 40

Blood-to-glucose ratio

N

< 0.4

< 0.4

N

< 0.4

< 0.4

Positive smeard

–

> 85%e

≥ 80%

–

< 40%

< 30% f

Positive culture

Rare

> 95%

< 90%

–

> 30%

PCR or other methods

Enterovirus, herpesvirus

16S RNA, bacterial DNA

16S RNA, bacterial DNA

Borrelia burdgorferi antibodies

Histoplasma and Cryptococcus antigen, India ink for Cryptococcus

CSF, cerebrospinal fluid; PCR, polymerase chain reaction; TB, tuberculosis; UD, undetectable. a Fewer than 500 leukocytes can be seen on severe pneumococcal meningitis. b Polymorphonuclear nucleocycte cell predominance can be observed in early stages of meningitis. c Low glucose concentrations can be detected in meningitis caused by mumps and herpesviruses. d Gram or acid-fast bacilli staining for bacteria or Mycobacterium, respectively. e Fewer positive smears in Listeria meningitis due to lower bacterial inoculum. f Better culture isolation rates for Candida than for Histoplasma or Cryptococcus organisms.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

< 30% Mycobacterium tuberculosis

Acute Bacterial Meningitis Beyond the Neonatal Period

has indicated that this test might be of value as a sensitive adjunctive screening tool for the timely diagnosis of bacterial meningitis,18 but the test has the limitation of low speciÀcity. Serum procalcitonin level of 0.5 ng/mL in one study of 167 children with meningitis (21, bacterial meningitis) in the postconjugate vaccines era had a sensitivity and speciÀcity of 89% for identifying bacterial meningitis.19 Polymerase chain reaction (PCR) of CSF has been employed to detect microbial DNA in patients with bacterial meningitis. Primers are available for the simultaneous detection of N. meningitidis, S. pneumoniae, and Hib. One study detected species-speciÀc amplicons in 87 of 98 CSF samples (sensitivity of 89%) from patients with meningitis caused by any of these pathogens, with no falsepositive results.20 Another group used a PCR-based assay developed to amplify a conserved region of the pneumococcal autolysin gene.21 The ampliÀed product was tagged and detected with a biotin-labeled probe in an enzyme immunoassay (EIA). This test was used in CSF samples from 11 patients with culture-negative meningitis, 5 of which yielded positive results. The researchers suggested that the PCR-EIA test might be useful when results of CSF culture, Gram stain, and latex agglutination are negative because of prior antibiotic treatment. Realtime PCR is a promising sensitive and speciÀc technique that likely will have great future usefulness in the clinical setting.22 Lumbar puncture for collection of CSF is contraindicated in certain situations. For children with hypotension, respiratory distress, or cardiac disorder, the positioning for lumbar puncture can further compromise ventilation and blood flow, so this procedure is deferred until the child’s condition is stable. In a child with profound thrombocytopenia or a clotting disorder, the lumbar puncture should be delayed until the condition is corrected. An older child or any child with an underlying predisposing condition (cyanotic heart disease or chronic sinusitis) who manifests fever and evidence of increased intracranial pressure or focal neurologic deÀcit should undergo an imaging study before lumbar puncture, to exclude a brain abscess, other mass lesion, or other cause of increased intracranial pressure. Occasionally, the skin overlying the lumbar vertebrae is infected or inflamed by cellulitis or an abscess associated with an infected dermal sinus, precluding a lumbar puncture. Empiric therapy is begun without delay in all of these situations. Blood culture results are positive in most children with bacterial meningitis, especially that caused by Hib or S. pneumoniae. The diagnostic yield of blood culture can be increased in meningococcemia by inoculation of broth without sodium polyanethol sulfonate (see Chapter 286, Laboratory Diagnosis of Infection due to Bacteria, Fungi, Parasites, and Rickettsiae). Aspiration of an inflamed joint, cellulitis (particularly of the cheek), purpuric lesion, or purulent middle-ear fluid maximizes identiÀcation of the infecting organism and provides a pathogen for susceptibility testing. In children with meningitis after surgery or trauma, culture of a specimen protected from infected wounds is useful. Culture specimens obtained from puncture of sinus, middle-ear cavity, or mastoid bone are useful when meningitis complicates these infections.

MANAGEMENT Empiric treatment of children with possible or proven bacterial meningitis is guided by knowledge of likely pathogens and their most current antimicrobial susceptibility. In addition, antibiotics and dosages are selected that produce CSF drug concentrations likely to be at least 10-fold greater than those required to inhibit and kill meningeal pathogens in vitro (minimal inhibitory concentration, or MIC), a level in animal models that predicts successful treatment.23 For b-lactam antibiotics and vancomycin, the time during which the drug concentration exceeds the MIC seems to determine drug effectiveness. Concentrations in CSF should surpass the MIC for at least 50% to 60% of the dosing interval (concentration-independent activity). For aminoglycosides and fluoroquinolones, effectiveness is determined by the ratio between the peak concentration or area under

CHAPTER

42

287

the curve of the antibiotic and the MIC of the pathogen (concentrationdependent activity).24

Antimicrobial Resistance National surveillance conducted by the CDC in 1986 showed that 32% of Hib isolates recovered from children with invasive infections were resistant to ampicillin because of b-lactamase production;25 by the mid-1990s, the rate of resistant organisms had increased to 40%.26 Occasional additional isolates are ampicillin-resistant, as a result of alterations in penicillin-binding proteins, or chloramphenicolresistant, because of production of chloramphenicol acetyltransferase enzyme.25 Chloramphenicol resistance is commonly encountered in other areas of the world. Cefotaxime and ceftriaxone have excellent activity against all Hib strains and are comparable in efÀcacy but superior in CSF sterilization rapidity to chloramphenicol for treatment of ampicillin-resistant Hib meningitis.27–29 The penicillin resistance of S. pneumoniae is a growing problem worldwide.30 Before the mid-1980s, there were only scattered reports of bacterial meningitis due to resistant S. pneumoniae throughout the world, including the United States. In some areas of the United States, however, as many as 40% of pneumococcal isolates are resistant,31 whereas in other countries, rates of resistant organisms as high as 60% have been noted. S. pneumoniae with intermediate resistance (MIC = 0.1 to 1.0 —g/mL) or high resistance (MIC ≥ 2.0 —g/mL) to penicillin have alterations in penicillin-binding proteins and also have reduced susceptibility to other b-lactam antibiotics. Pneumococcal serotypes 6B, 23F, 14, and 19 are the most common serotypes associated with penicillin resistance and also are the serotypes most frequently isolated from children with systemic infection. Furthermore, pneumococcal isolates can be resistant to multiple classes of antibiotics (penicillins and cephalosporins, trimethoprim-sulfamethoxazole, chloramphenicol, and macrolides). In children’s hospitals in the United States, > 20% of pneumococci recovered from children with systemic infections can be expected to show penicillin resistance. Penicillin or ampicillin is inappropriate therapy for meningitis due to pneumococci that are relatively or highly resistant to penicillin; treatment failures have been documented in such cases. Pneumococcal isolates that are tolerant to b-lactam antibiotics and vancomycin have also been documented.32 These strains have a defective control pathway for triggering autolysis, preventing lysis (i.e., killing) of the organism in the presence of b-lactam antibiotics. As a result, organisms can persist in CSF, eventuating in relapse after treatment is stopped. Because of defective autolysis, these bacteria produce fewer cell wall-derived meningeal inflammatory substances, so the CSF abnormalities they produce can be mild.33 Cefotaxime and ceftriaxone remain active against many penicillinresistant pneumococci; however, treatment failures have occurred in some cases of meningitis.34 The MIC for cefotaxime or ceftriaxone in these instances is usually ≥ 1.0 —g/mL. Guidelines for interpreting MIC values of antibiotics for S. pneumoniae are shown in Table 42-3.35 Guidelines undergo frequent revision as additional information accrues, correlating clinical outcome with in vitro activity. Interpretive standards differ in patients with or without meningitis for cefotaxime, ceftriaxone, and cefepime. Vancomycin and rifampin are active against cefotaxime- or ceftriaxone-resistant S. pneumoniae. Microbiology laboratories should routinely screen all pneumococcal isolates from normally sterile sites for penicillin susceptibility. Screening can be accomplished with several methods, including oxacillin disk, E-test, and commercial microtiter methods, with the appropriate use of broths for growing S. pneumoniae and proper testing standards.36 Aqueous penicillin G remains the agent of choice for meningococcal meningitis, although ampicillin and third-generation cephalosporins are also effective and are easier to administer. Clinical isolates of N. meningitidis with relative resistance to penicillin (MIC ≥ 0.125 —g/mL) but not third-generation cephalosporins have been reported.37 Resistance is associated not with b-lactamase

288

SECTION

F

Central Nervous System Infections

TABLE 42-3. National Committee for Clinical Laboratory Standards Guidelines for Interpretation of Susceptibility Testing of Streptococcus pneumoniae Minimal Inhibitory Concentration (μg/mL) Interpretation

Penicillin

Cefotaxime/Ceftriaxone

Susceptible

≤ 0.06

≤ 0.5

Intermediate Resistant

0.1–1.0 ≥ 2.0

a

Imipenem–Cilastatin

Vancomycin

Rifampin

Chloramphenicol

≤ 0.12

≤1

≤1

≤4

1.0

0.25–0.5

–

2

–

≥ 2.0

≥1

–

≥4

≥8

a

Interpretive standard for meningitis.

TABLE 42-4. Recommendations for Pathogen-Specific Antimicrobial Therapy of Children with Bacterial Meningitis Bacteria

Antibiotic of Choice

Other Useful Antibiotics

Neisseria meningitidis Haemophilus influenzae Streptococcus pneumoniaeb 1. Penicillin-susceptible (MIC < 0.1 —g/mL) 2. Penicillin-intermediate (MIC = 0.1–1.0 —g/mL)

Penicillin G or ampicillin Cefotaxime or ceftriaxone

Cefotaxime or ceftriaxone Ampicillin or chloramphenicola

Penicillin G or ampicillin Cefotaxime or ceftriaxone with/without vancomycin Cefotaxime or ceftriaxonec plus vancomycin Cefotaxime or ceftriaxonec plus vancomycin

Cefotaxime or ceftriaxone Cefepime or meropenem

3. Penicillin-resistant (MIC ≥ 1.0 —g/mL) 4. Cephalosporin-nonsusceptible (MIC > 0.5 —g/mL)

Cefepime or meropenem Add rifampin to antibiotics of choice Meropenem + vancomycin (see text) New fluoroquinolonesd Trimethoprim-sulfamethoxazole

Listeria monocytogenes

Ampicillin ± gentamicin

Streptococcus agalactiae

Penicillin G ± gentamicin

Ampicillin ± gentamicin

Enterobacteriaceae

Cefotaxime or ceftriaxone with/without aminoglycoside

Cefepime or meropenem

Pseudomonas aeruginosa

Ceftazidime + amikacin

Cefepime or meropenem

MIC, minimal inhibitory concentration. a In areas with economic constraints: ampicillin for susceptible strains and chloramphenicol for ampicillin-resistant Haemophilus influenzae type b isolates. b Higher dosages might be helpful. c These drugs are currently under clinical investigation.

production but with alterations in penicillin-binding proteins. The clinical signiÀcance of this type of resistance in the treatment of meningococcal meningitis is unclear.

Antimicrobial Therapy Initial empiric therapy using cefotaxime or ceftriaxone plus vancomycin (60 mg/kg per day in divided doses every 6 hours) is prudent for children and infants 1 month and older.38,39 Limited clinical pediatric data preclude reliable recommendations regarding rifampin combined with a cephalosporin. In the Hib vaccine era, ampicillin plus chloramphenicol is not an optimal combination for initial therapy because of its probable inadequacy against resistant pneumococcal strains. Once an organism has been identiÀed and the antimicrobial susceptibility pattern is known, antibiotic therapy can be simpliÀed or modiÀed (Table 42-4). Penicillin G can be used to complete therapy for pneumococcal or meningococcal meningitis due to susceptible organisms. Cefotaxime or ceftriaxone is continued for penicillinresistant pneumococci that are susceptible to these agents (MIC ≤ 0.5 —g/mL).40 Although many patients whose isolates have an MIC of ≥ 1.0 —g/mL for third-generation cephalosporins have been treated successfully, this should not be relied on. It is prudent to repeat the CSF examination after 24 to 48 hours of therapy in patients with infection caused by beta-lactam-resistant pneumococci to document a sterile culture and negative Gram stain. If results of the second CSF culture or Gram stain is positive, or a pneumococcal isolate has an MIC of 1.0 —g/mL or greater for extended-spectrum cephalosporins,

vancomycin (if not previously begun) with or without rifampin should be added.41 Rifampin should be added if the patient is already receiving vancomycin. Vancomycin penetration into CSF during meningitis is unpredictable, and thus, peak serum levels should be in the upper therapeutic range (35 to 40 —g/mL). Chloramphenicol is an option for treatment of bacterial meningitis caused by ampicillin-resistant Hib or in the child with a history of anaphylaxis or respiratory distress associated with penicillin or other b-lactam antibiotics. Because the pharmacokinetics of chloramphenicol are variable, serum concentrations should be monitored to ensure that safe and effective levels are achieved.42 Ideal peak values are 15 to 30 —g/mL at 60 to 120 minutes after the infusion is complete. Levels exceeding 30 —g/mL are associated with bone marrow suppression; levels greater than 50 to 80 —g/mL have been related to “gray-baby” syndrome (which is not conÀned to infants) due to impaired myocardial contractility. Chloramphenicol is avoided in children with septic shock. Additionally, several drugs affect the metabolism of chloramphenicol: Concurrent administration of phenobarbital and rifampin increases the chloramphenicol metabolism, whereas administration of phenytoin generally lowers metabolism. Cefepime and meropenem are useful antimicrobial agents for the treatment of bacterial meningitis. Both have been shown to be equivalent to third-generation cephalosporins in treatment of children infected with common meningeal pathogens.43–46 The efÀcacy of these antibiotics against b-lactam-resistant pneumococci has not been deÀned clearly, although they appear not to be superior to cefotaxime or ceftriaxone. Imipenem-cilastatin is not recommended in children with CNS infection because of its potential epileptogenic activity in

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Acute Bacterial Meningitis Beyond the Neonatal Period

children. Because of the increasing rate of isolation of multiple-drugresistant pneumococci worldwide, there is an urgent need to develop antibiotics with different mechanisms of action against these strains. Accordingly, fluoroquinolones and other novel antibiotics need to be evaluated in clinical trials of children with bacterial meningitis.

Monitoring During Therapy Performance of complete blood counts during therapy is helpful for early detection of neutropenia, which is associated with b-lactam antibiotics and chloramphenicol; the neutropenia is readily reversible when antibiotics are discontinued. Anemia is commonly encountered in children with bacterial meningitis and can also be exacerbated by chloramphenicol. Ceftriaxone has been associated with immunemediated, rapidly fatal hemolysis in patients with sickle-cell disease, HIV infection, and leukemia.47,48 Thrombocytosis (platelet count > 750 000/mm3) occurs frequently, has no apparent adverse clinical effect, and need not be monitored. Diarrhea is the most common adverse effect of ampicillin and the extended-spectrum cephalosporins. Rashes, eosinophilia, and mild elevation in serum hepatic transaminase values can occur with any of these agents as well. Ceftriaxone has been associated with abdominal discomfort and pseudolithiasis (“sludge”) in the gallbladder documented by ultrasonography;49 this process can be reversed with discontinuation of the drug. Routine performance of computed tomography or magnetic resonance imaging of the head during antimicrobial therapy has no therapeutic benefit; imaging is useful if the clinical course is complicated or the pathogen is unusual.50

CHAPTER

42

289

through an Ommaya reservoir. Systemic treatment is continued for a minimum of 21 days or at least 2 weeks beyond the first documented sterile CSF culture, whichever is longer. Surgical drainage of wound infections after trauma or neurosurgery is also required, as is removal of contaminated devices. Imaging studies are performed in patients with gram-negative bacillary meningitis to evaluate predisposing factors or the nature of cerebral involvement and the need for surgical intervention. Meningitis due to multiple-drug-resistant gram-negative bacilli may require the use of other agents, such as meropenem or cefepime, that have greater in vitro activity against gram-negative rods, including cephalosporinresistant strains. Meningitis caused by Staphylococcus aureus is treated with nafcillin or oxacillin when the organism is susceptible. Vancomycin is the alternative when the patient is allergic to penicillin or has a methicillin-resistant isolate. Adding rifampin to the antistaphylococcal regimen may be necessary if CSF culture results remain positive despite the use of agents effective in vitro. Ampicillin plus an aminoglycoside is recommended for patients with meningitis due to susceptible strains of Enterococcus spp. Vancomycin (with aminoglycoside) is used when Enterococcus spp. resistant to ampicillin are encountered (see Chapter 120, Enterococcus Species). Ampicillin is also the drug of choice for Listeria monocytogenes infection; data support the beneficial role of concurrent aminoglycosides (see Chapter 132, Listeria monocytogenes). If staphylococcal infection occurs in a patient with a ventriculoperitoneal shunt, bacterial eradication is best accomplished by removal of the device coupled with antimicrobial therapy. Case reports of successful therapy with linezolid in adults with meningitis caused by resistant staphylococci and enterococci have recently been published.54,55

Duration of Therapy The duration of antibiotic therapy varies with the organism and the clinical response.51 For otherwise uncomplicated cases, meningitis due to S. pneumoniae is treated for 10 to 14 days; that due to N. meningitidis, for 4 to 7 days; and that due to Hib, for 7 to 10 days. When possible, therapy should be completed in the hospital to permit careful assessment of the response to treatment and to prevent complications of disease or therapy. Therapy could be completed at home for carefully selected patients (e.g., intramuscular ceftriaxone) if meningitis is due to an exquisitely susceptible organism and the patient has resumed virtually normal activity. Ceftriaxone is approved for a dosing schedule every 24 hours, making it more convenient and perhaps less expensive than other agents. It is essential, however, to calculate and administer each dose correctly so that the patient is not treated inadequately for 24 hours because an error or delayed intravenous infiltration occurs.

MENINGITIS DUE TO UNUSUAL ORGANISMS Escherichia coli and Klebsiella spp. are the most common gramnegative enteric organisms that cause bacterial meningitis in children other than neonates.52 A combination of an extended-spectrum cephalosporin or ampicillin plus an aminoglycoside administered intravenously is reasonable empirical therapy for suspected gramnegative meningitis, with the choice affected by nosocomial patterns in hospital-associated cases.53 Modifications are made once antimicrobial susceptibility information is available. Ceftazidime (cefepime, meropenem, ticarcillin, or piperacillin) is generally given with an aminoglycoside for Pseudomonas meningitis. Aminoglycoside peak serum levels should be at the upper safe limit in the treatment of meningitis. A CSF specimen is drawn for culture at 24 to 48 hours after therapy begins and every 48 hours thereafter until sterilization of CSF is documented. Occasionally, CSF does not become sterile unless an aminoglycoside is directly instilled into the ventricles (almost exclusively in postoperative or posttraumatic cases). This is best done

SUPPORTIVE CARE Unless the patient is only mildly affected, initial supportive care of the child with bacterial meningitis is best provided in an intensive care setting, where the patient can be observed and monitored continuously.56 Typically, most life-threatening complications of bacterial meningitis (septic shock, herniation, infarctions, seizures, and inadequate ventilation) occur early in the course of treatment and require urgent interventions for optimal outcome. Maintenance of blood pressure within the normal range for age may require infusion of a vasoactive agent, such as dopamine or dobutamine. Initially, the patient is not fed orally. Fluid restriction is only advised in patients without clinical evidence of dehydration but with hyponatremia (i.e., suspicion of inappropriate antidiuretic hormone secretion). If the patient is hypovolemic or in shock, additional intravenous fluids must be given accordingly. There is no evidence that fluid restriction reduces cerebral edema in children with bacterial meningitis. A recent Cochrane meta-analysis found that fluid restriction in bacterial meningitis can be associated with poorer neurologic outcome compared with administration of maintenance intravenous fluids.57 Increased intracranial pressure is a major component of the pathophysiologic alterations of meningitis. In addition to elevation of the patient’s head, some clinicians administer mannitol (0.5 to 2 g/kg) if severely high intracranial pressure is detected (apnea, bradycardia, sluggish pupils, or pupillary dilation). Immediate intubation and hyperventilation can be lifesaving if cerebral herniation develops. Seizures are controlled with standard anticonvulsants, such as phenobarbital and phenytoin. Most seizures occur early in the illness and are generalized, and they are easily controlled. These seizures have no prognostic significance. However, seizures that are focal, occur 48 hours or more after admission, or are difficult to control imply an underlying vascular disturbance, such as venous thrombosis or infarction, and are associated with development of epilepsy and other neurologic sequelae.58 Once adequate therapy begins, the duration of fever is typically 4 to 6 days. Children with meningitis due to Hib tend to have

290

SECTION

F

Central Nervous System Infections

a relatively longer febrile period. Recurrence of fever or persistence of fever beyond 8 days can be caused by several conditions (Box 42-1). Fever patterns should be carefully evaluated, although a specific explanation is often not found. It is unlikely that uninfected subdural effusions cause prolonged fever.

ADJUNCTIVE MEASURES Modulating the host response to infection may be beneficial in decreasing some sequelae of meningitis. Prospective studies have shown that dexamethasone therapy initiated just before or concurrently with the first dose of intravenous antibiotics significantly diminishes the incidence of neurologic and audiologic deficits due to Hib meningitis.26,59–62 Although the benefit of dexamethasone has not been evaluated as extensively as in Hib disease, two meta-analyses of published data indicated that early administration of this agent also improves outcome in pneumococcal disease, in both children and adults.63,64 For pneumococcal meningitis, it seems crucial to administer dexamethasone before or concomitantly with parenteral antibiotics, particularly in emergency departments of both developed and underdeveloped countries.65,66 Because of the better prognosis of meningococcal meningitis, it is likely that thousands of patients would have to be studied to assess the role of corticosteroid therapy reliably. Dexamethasone can decrease the penetration of antibiotics into the CSF and, theoretically, could jeopardize bacterial eradication of b-lactam-resistant pneumococcal strains. Although some clinical reports have indicated that dexamethasone does not affect the bactericidal activity against highly resistant pneumococci when a third-generation cephalosporin is combined with vancomycin,67 careful clinical assessment and second CSF evaluation are warranted in this situation. Dexamethasone administration is associated with a rapid resolution of fever (i.e., 1 to 2 days) in treated patients, but a second, immediate transient rise in body temperature is observed in up to 40% of cases when the drug is discontinued. Its administration has rarely been associated with severe gastrointestinal bleeding. To minimize potential, albeit infrequent, adverse events while maintaining therapeutic anti-inflammatory efficacy, the currently recommended dosing regimen is 0.6 to 0.8 mg/kg daily in 2 or 3 divided doses for 2 days. For optimal results, the first dose of dexamethasone should be administered before or concomitant with the first dose of antibiotic parenterally.

PROGNOSIS AND SEQUELAE With modern management, the mortality rate for bacterial meningitis in children caused by the three common pathogens is less than 5% to 10% in most studies. Case-fatality rate and incidence of neurologic sequelae are greatest with pneumococcal meningitis. The neurologic sequelae of meningitis are listed in Box 42-2. Sensorineural hearing

loss is the most common readily identifiable sequela. Hearing loss occurs in approximately 20% to 30% of patients after meningitis due to Streptococcus pneumoniae and in 5% to 10% of patients after meningitis due to Hib or N. meningitidis. Balance disturbances are common in these children because the vestibular portion of the inner ear is also affected. Hearing loss is more likely if the admission CSF glucose concentration is < 20 mg/dL. Hearing should be tested before discharge or within 1 month of discharge in all children with bacterial meningitis, so that any hearing loss can be detected as early as possible. Reversible deafness has been documented in some children at follow-up testing 4 to 6 months after discharge. Acute hydrocephalus as well as many sequelae related to vascular compromise can improve with time. Hemiparesis can resolve several months to years after the event. Imaging studies demonstrate evidence of infarction in such patients, although these findings do not usually affect or change management. Behavioral and academic problems are more subtle consequences of bacterial meningitis that may not be apparent for several years after infection. Although formal testing is not generally necessary, careful assessment over time is essential; any concerns regarding school performance warrant further investigation.

PREVENTION The greatest advance regarding meningitis is the introduction of conjugated vaccines to prevent meningitis due to the most common microorganisms. The formidable impact of the conjugate Hib vaccines has been extensively discussed. A heptavalent pneumococcal conjugate vaccine has 97% efficacy against invasive infections caused by the pneumococcal serotypes contained in the vaccine.68 A marked reduction of pneumococcal meningitis in vaccinated United States children has been documented in recently published surveillance studies.69,70 A trial of a 9-valent pneumococcal conjugate vaccine conducted in African children, with or without HIV infection, reduced the incidence of a first episode of invasive pneumococcal disease due to vaccine-contained serotypes by 65% and 83%, respectively.71 Ongoing studies are further evaluating 9-, 11-, and 13-valent vaccines that cover most important serotypes circulating in many parts of the world. Children older than 2 years of age who are at risk of developing invasive pneumococcal disease, such as patients with sickle-cell disease and nephrotic syndrome, or meningitis, such as children with cochlear implants, should receive the 23-valent polysaccharide vaccine. A serogroup-specific quadrivalent meningococcal polysaccharide vaccine against A, C, Y, and W-135 strains is recommended for highrisk children older than 2 years of age, such as children with asplenia or terminal complement deficiencies, and suggested for freshmen students living in college dormitories. Current work is dedicated to improving the immunogenicity of meningococcal vaccines and to generating potentially effective candidate vaccines against group B meningococci. Meningococcal serogroup C conjugate vaccine is routinely being administered in the United Kingdom and Canada

BOX 42-1. Causes of Prolonged or Recurrent Fever in Children with Bacterial Meningitis

BOX 42-2. Neurologic Sequelae of Bacterial Meningitis

• • • • • • • • • • • •

• • • • • • • • • • • •

Inadequate treatment Nosocomial infection Discontinuation of dexamethasone Phlebitis Suppurative complication Pericarditis Pneumonia Pyogenic arthritis Subdural empyema Immune-mediated arthritis Drug fever Unknown

Sensorineural hearing loss Ataxia Vascular insults Hemiparesis Quadriparesis Epilepsy Spinal cord infarction Cortical blindness Diabetes insipidus Hydrocephalus Behavior disorder Intellectual deficits

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Chronic Meningitis

with preventive success. Serogroup A, C, Y, W-135 polysaccharideconjugate vaccine currently is given universally before entry to high school in the United States. Close contacts of patients with meningococcal disease should receive chemoprophylaxis with rifampin at doses of 600 mg for adults, 10 mg/kg for children older than 1 month, and 5 mg/kg for younger infants twice daily for 2 days, started ideally within 24 hours of the exposure. A single large oral dose of ciprofloxacin or azithromycin or a parenteral dose of ceftriaxone has been shown to be a suitable prophylactic alternative; the last is preferred for pregnant women. Rifampin prophylaxis is also recommended for all household contacts of an index case with Hib disease when at least one household contact is younger than 4 years and is unimmunized or incompletely immunized.

CHAPTER

43

Chronic Meningitis Ram Yogev

Chronic meningitis is defined arbitrarily as symptoms and signs of meningeal irritation that persist for at least 4 weeks without improvement. The 4-week timeframe is intended to avoid extensive evaluation for individuals with self-limiting processes (e.g., meningoencephalitis, resolving meningitis). In most cases of prolonged meningitis, diagnosis and treatment occur before clinical symptoms have continued for 4 weeks; thus, chronic meningitis is relatively rare.

ETIOLOGY

CHAPTER

43

291

TABLE 43-1. Infectious Causes of Chronic Meningitis Agent

Reference No.

BACTERIA

Mycobacterium tuberculosis Treponema pallidum Brucella Borrelia burgdorferi Nocardia Actinomyces (or Arachnia) Leptospira Mycoplasma or Ureaplasma

1 2 3 4 5 6 7 8

VIRUSES

Human immunodeficiency virus Mumps virus Lymphocytic choriomeningitis virus Enterovirusa Cytomegalovirusb Herpes simplex virusb

9

FUNGI

11 12 13 14 15 16 17

Blastomyces Histoplasma Coccidioides Cryptococcus Candida Aspergillus Pseudallescheria boydii (asexual form, Scedosporium apiospermum) Zygomycetes Cladosporium Sporothrix

10

18 19 20 21

OTHERS

Toxoplasma Taenia (cysticercosis) Acanthamoeba Balamuthia Angiostrongylus Coenurus cerebralis Baylisascaris

22 23 24 25 26 27 28

a

In patients with agammaglobulinemia. In patients infected with the human immunodeficiency virus.

b

In many cases of chronic meningitis, the history, physical examination findings, and laboratory results are not helpful in identifying the cause. Most infectious causes (Table 43-11–28) are associated with predominantly lymphocytic cerebrospinal fluid (CSF) pleocytosis. The absolute number of cells in the CSF may be helpful for differential diagnosis. An infectious agent is rarely responsible when CSF has fewer than 50 white blood cells (WBC) per cubic millimeter; in such cases, no infectious causes should be considered (Box 43-129–39). Fungi are probably the most frequent infectious cause of chronic meningitis. Histoplasma capsulatum, Coccidioides immitis, and Blastomyces dermatitidis should be considered in the immunocompetent host; Candida spp. and Cryptococcus neoformans are the most common agents in the immunocompromised patient. Other fungi, such as Aspergillus, Blastomycoses, Sporothrix, Cladosporium, and Allescheria spp., and agents of mucormycosis and pheohyphomycosis are rare causes of chronic meningitis. Although Mycobacterium tuberculosis can cause chronic meningitis, most patients with untreated tuberculous meningitis die within 3 weeks of the first symptoms. In patients with the human immunodeficiency virus (HIV), tuberculous meningitis is at least five times more common than in patients who have tuberculosis without HIV, but the clinical presentation and mortality rate are similar. Borrelia burgdorferi should be considered in endemic areas. Leptospirosis is a rare cause of chronic meningitis because the clinical signs and symptoms of meningitis generally disappear within 7 to 21 days.7 Other rare bacterial causes include Brucella melitensis,3 Treponema pallidum,2 Nocardia,5 Actinomyces, and Arachnia spp.6 Both Mycoplasma hominis and Ureoplasma urealyticum can cause chronic meningitis in preterm infants.8 It is commonly associated with

BOX 43-1. Noninfectious Causes of Chronic Meningitis Sarcoidosis29 Neoplasm (e.g., non-Hodgkin lymphoma) Systemic lupus erythematosus Polyarteritis nodosa Rheumatoid arthritis Granulomatous angiitis30 Other forms of vasculitis31 Behçet syndrome32 Sjögren syndrome33 Neonatal-onset multisystem inflammatory disease (NOMID)34 Uveomeningoencephalitis syndrome35 Chronic benign lymphocytic meningitis36 Subarachnoid hemorrhage Subdural hematoma Drug-induced (e.g., ibuprofen, cyclooxygenase-2 inhibitor, trimethoprim)37,38 Wegener granulomatosis39

development of hydrocephalus. Protozoa and parasites can also cause chronic meningitis, including Toxoplasma,22 Taenia,23 and rarely Acanthamoeba,24 Balamuthia,25Angiostrongylus,26 and Coenurus cerebralis.27 HIV itself is probably the most common viral cause,9 but patients with the acquired immunodeficiency syndrome (AIDS) can also develop chronic cytomegalovirus (CMV) meningitis. In patients

292

SECTION

F

Central Nervous System Infections

with agammaglobulinemia, echoviruses can cause persistent meningitis.10 Other rare viral causes include mumps virus and lymphocytic choriomeningitis virus. Several parameningeal infections can manifest as chronic meningitis. The most common entities are shown in Box 43-2.

EPIDEMIOLOGY The incidence of chronic meningitis is unknown and is different for each etiologic agent. Approximately 0.5% of children with untreated tuberculosis develop meningitis. The recent increase in tuberculosis in the United States has also increased the number of cases of tuberculous meningitis, with disproportionate representation in African Americans and immigrants from countries where tuberculosis is endemic. The disease is more common in children younger than 6 years but rarely occurs in infants younger than 6 months. Chronic syphilitic meningitis is rare and occurs in fewer than 0.5% of patients with syphilis; the incidence of meningitis is greatest in the first 2 years of the syphilis infection. In recent years, the incidence of syphilis has declined in the United States, but continued attention to populations with increased risk (e.g., HIV-infected individuals) and early treatment prevents its severe consequences. Although neurosyphilis rarely manifests as chronic meningitis, partially treated neurosyphilis can imitate chronic meningitis. Meningitis occurs in fewer than 5% of patients with brucellosis and is the first manifestation of the disease in about 1%. Although brucellosis is relatively rare in the United States, it remains a common disease worldwide. Thus, a history of travel to an endemic area or history of consumption of unpasteurized dairy products is important. Neurologic manifestations commonly occur in patients with Lyme disease, but meningitis develops in only 10% to 15%. Chronic meningitis with serious neurologic sequelae occurs rarely.40 Visits to an endemic area, contact with deer, or history of erythema migrans are clues to the diagnosis. Histoplasma capsulatum is a common cause of infection in the United States, but disseminated disease is rare and usually occurs in immunocompromised individuals. Neurologic involvement is reported in 8% to 18% of cases of disseminated disease,13 but an incidence of 55% has been found in AIDS patients with disseminated disease. In most cases with neurologic manifestations, the duration of symptoms was more than 4 weeks. Blastomyces meningitis, the most common central nervous system (CNS) manifestation of blastomycosis, rarely occurs in the absence of systemic dissemination. Exposure to an endemic area (i.e., the Great Lakes or southeast regions of the United States for Blastomyces species disease and California, New Mexico, or Texas for Coccidioides species, especially if a sandstorm has occurred) provides a clue to the diagnosis. Although respiratory infection with Coccidioides is common, disseminated disease develops in only 0.5% of patients and meningitis in fewer than 0.2%; nonwhites, especially Filipinos, are at greater risk for development of meningitis.14 Infections caused by unusual fungi (e.g., Candida, Aspergillus, or Cryptococcus spp.) have increased as the number of immunocompromised children, performance of organ transplantation, survival

BOX 43-2. Parameningeal Infections That can Manifest as Chronic Meningitis • Encephalitis • Viral • Postinfectious • Brain abscess • Subdural empyema • Cranial osteomyelitis • Mastoiditis • Sinusitis

of premature infants, use of indwelling catheters, and prolonged therapy with antibiotics or corticosteroids have increased. Chronic meningitis is an uncommon manifestation of disseminated candidiasis; an underlying risk factor is found in about 75% of patients. In contrast, disseminated aspergillosis frequently involves the CNS, particularly in patients with prolonged neutropenia (present in more than 60% of cases11). Direct CNS invasion after head trauma or lumbar puncture or from the paranasal sinuses or the orbit can also occur. Cryptococcus neoformans is probably the most common cause of fungal meningitis; the disease can occur in the absence of any risk factor. However, impairment of cellular immunity resulting from Hodgkin disease, high-dose corticosteroid treatment, or other impairment of the reticuloendothelial system increases susceptibility to cryptococcal infection. Cryptococcal meningitis develops in up to 10% of adults with AIDS; it is their most common life-threatening fungal infection and, in 40% of those affected, it is the first AIDSdefining opportunistic infection. The disease occurs more frequently in nonwhites and in the southeastern United States. C. neoformans variant gattii was found to be strongly associated with CNS disease in immunocompetent patients, suggesting that genotypic or phenotypic differences of the fungus, or both, may explain its virulence in normal hosts.41 Cysticercosis is the most common cause of CNS parasitic disease worldwide. The prevalence of cerebral cysticercosis in hospitalized patients in Latin America (an endemic area) ranges from 0.02% in Honduras to 2.4% in Brazil and accounts for 3% of patients in the neurologic wards in Mexico. Other endemic areas include Southeast Asia, Africa, and Eastern Europe. In the United States, most cases are reported from southwestern states, usually among recent immigrants from Latin America. Chronic meningitis is a rare manifestation of CNS cysticercosis and usually appears several years after a visit to an endemic area. The most common manifestation is an afebrile focal seizure or increased intracranial pressure. In HIV-infected patients, Toxoplasma gondii is the most common parasite causing CNS infections. Although CNS toxoplasmosis is common in adult AIDS patients with pre-existing antibodies to T. gondii (with a prevalence of almost 30%), the disease is rare in HIVinfected children. In addition, chronic meningitis is not a common manifestation of CNS toxoplasmosis. The usual course of Candida albicans meningitis is acute. Chronic meningitis is uncommon but increasingly is recognized in recent years.42 Most reported cases were in neonates, children with congenital immunodeficiencies (e.g., chronic granulomatous disease, severe combined immunodeficiency), and patients with indwelling catheters or prolonged neutropenia (or both) or had infection as a postoperative complication of neurosurgery. Although candidiasis is common in AIDS patients, dissemination to the CNS is rare.

DIFFERENTIAL DIAGNOSIS Before testing for the myriad infectious causes of chronic meningitis (see Table 43-1), noninfectious and parameningeal conditions should be considered (see Boxes 43-1 and 43-2). In addition, cases of recurrent meningitis (i.e., repeated episodes of acute meningitis separated by symptomfree periods) should be distinguished from chronic meningitis (see Chapter 44, Recurrent Meningitis). Figure 43-1 presents an algorithm to guide differential diagnosis.

History Although the exact cause of chronic meningitis may not be found in up to one-third of cases,43 close attention to details may provide important clues to the diagnosis. A history of environmental exposure or exposure to persons infected with pathogens spread by person-toperson contact (e.g., tuberculosis, syphilis) should be sought (Table 43-2). Although in most cases, signs and symptoms of illness develop PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Chronic Meningitis

Positive exposure history

Known infected person

Travel or residence in endemic area

Tuberculosis Syphilis

Table 43-2

Negative exposure history

Predisposing condition

Clinical manifestations and physical examination

Table 43-3

Positive findings

Negative findings

Table 43-4

Lumbar puncture features Table 43-5

Confirmatory diagnosis

CHAPTER

43

293

Figure 43-1. Algorithm for evaluation of patients with signs and symptoms of chronic meningitis.

30% undiagnosed

TABLE 43-2. Agent Causing Chronic Meningitis and Likely Exposure History in Patient Agent

Exposure History

Brucella

Ingestion of unpasteurized dairy products, especially in the Mediterranean, India, or Latin America

Leptospira

Occupational (farmer) or recreational (camper, swimmer) exposure to urine of wild and domestic mammals during summer and fall

Borrelia

T/Ra in northeast United States, Wisconsin, Minnesota, California, and Oregon, especially during late summer and early fall, or in Europe, China, Japan, and Australia, with consequent tick exposure

Blastomyces

T/R in Ohio and Mississippi river valleys or North Carolina

Histoplasma

T/R as for Blastomyces, plus Appalachian mountains and Virginia

Coccidioides

T/R in southeastern United States, Mexico, Central America, Venezuela, Argentina, or Paraguay

Paracoccidioides

T/R in Central and South America, and, rarely, Mexico

Sporothrix

T/R in tropical and subtropical Americas, or a prick from a rose thorn

Cysticercus

T/R in Latin America, Southeast Asia, Africa, Eastern Europe, and southwestern United States

Angiostrongylus

T/R in Hawaii, Australia, Southeast Asia, or Philippines

Baylisascaris

Exposure to raccoon feces

a

T/R, travel or residence.

within a short time after the exposure (such as in brucellosis, Lyme disease, and blastomycosis), a long incubation period (such as in cysticercosis) or reactivation of a dormant focus several years after exposure (such as in tuberculosis, histoplasmosis, and toxoplasmosis) can occur. A careful history can also reveal predisposing conditions that increase the risk of opportunistic infections (Table 43-3). For example, prematurity, prolonged usage of corticosteroid therapy, cancer, HIV infection, intravenous drug abuse, or use of indwelling catheters facilitates invasion and dissemination of Candida spp. Direct extension of Aspergillus spp. infection from an adjacent sinus, ear, or orbit or after head trauma or surgery can lead to infection of the CNS. In HIV-infected patients, multiple organisms can cause chronic meningitis: HIV itself, fungi that cause disease in immunocompetent patients (e.g., Histoplasma, Cryptococcus, Coccidioides, Cladosporium, and Blastomyces spp.) or in immunocompromised patients (e.g., Candida and Aspergillus spp.), and M. tuberculosis and Toxoplasma spp. In addition, herpes simplex virus (HSV) and CMV should be considered in HIV-infected patients, especially in those with end-stage disease.

Clinical Manifestations A thorough and careful physical examination is important because, occasionally, the physical findings suggest the diagnosis (Table 43-4).

Skin The skin examination may reveal a typical lesion or provide a source for culture or biopsy to identify etiology. Erythema nodosum (tender, erythematous nodules located on the anterior aspect of the lower leg, which typically do not suppurate) are seen in a limited number of diseases (see Table 43-4); nonerythematous nodules can also be found in patients with systemic candidiasis, blastomycosis, and nocardiosis. The erythema migrans lesion of Lyme disease, the rarely seen umbilicated nodules of cryptococcal infection in HIV-infected patients, and the characteristic palmar lesions of secondary syphilis can guide investigations. Cysticercosis can be associated with subcutaneous nontender nodules, usually on the extremities or the trunk.44 Several other pathogens can cause skin lesions (see Table 43-4); a skin biopsy with appropriate staining and culture can prove the diagnosis.

294

SECTION

F

Central Nervous System Infections

TABLE 43-3. Causes of Chronic Meningitis with Predisposing Conditions Predisposing Conditions Causative Organism

Indwelling Catheter

Trauma or Surgery

Illicit IV Drug Use

Long-term Corticosteroid

Cancer

Transplant

HIV Infection

Candida

•

•

•

•

•

•

•

•

•

•

•

•

•

•

•

•

•

•

Aspergillus Cryptococcus

a

Cladosporiuma

•

Zygomycetes

•

Histoplasma

• •

a

Coccidioides

a

•

•

•

•

•

•

•

Blastomycesa

•

•

•

Toxoplasma

•

•

•

Nocardia

•

•

•

Mycobacterium tuberculosisa

•

•

•

•

a

Also occurs in immunocompetent host. HIV, human immunodeficiency virus; IV, intravenous.

TABLE 43-4. Etiology of Chronic Meningitis Relative to Clinical Manifestations Skin Lesions

Erythema Nodosum

Lung Involvement

Chorioretinitis

Endophthalmitis

Cranial Neuropathy

Hepatomegaly

Blastomyces Candida Coccidioides Cryptococcus Zygomycetes (e.g., mucormycosis) Sporothrix Treponema pallidum Borrelia burgdorferi Nocardia Cysticercus

Mycobacterium tuberculosis Histoplasma Coccidioides Behçet disease Sarcoidosis Systemic lupus erythematosus

Blastomyces Coccidioides Cryptococcus Histoplasma Paracoccidioides Mycobacterium tuberculosis Sarcoidosis

Candida Histoplasma Mycobacterium tuberculosis Treponema pallidum Toxoplasma CMV Sarcoidosis

Candida Cryptococcus Coccidioides Sporothrix Pseudallescheria Treponema pallidum Actinomyces Mycobacterium tuberculosis

Cranial Mycobacterium tuberculosisa Treponema pallidumb Borrelia Histoplasma HIV Cancer Sarcoidosis Peripheral Borrelia Brucella Sarcoidosis Systemic lupus erythematosus CMVc

Brucella Leptospira Histoplasma Sarcoidosis

AIDS, acquired immunodeficiency syndrome; CMV, cytomegalovirus; HIV, human immunodeficiency virus a Especially cranial nerve VI. b Especially cranial nerves VII and VIII. c In AIDS patients.

Lungs The lungs are the portal of entry for many of the pathogens that cause chronic meningitis; therefore, careful examination and laboratory evaluation (chest radiography, microscopy, culture of sputum, and, occasionally, lung biopsy) can be useful. Blastomycosis of the CNS is almost always associated with other organ involvement, including the lungs. Cryptococcal meningitis is also frequently associated with disseminated disease, along with obvious clinical symptoms of respiratory tract involvement or abnormal results on chest radiograph. Almost two-thirds of patients with coccidioidal meningitis have lung

lesions. Immunocompromised patients with Histoplasma infection usually have multiple organs involved, including the lungs, liver, spleen, lymph nodes, or skin. Seventy-five percent of children with tuberculous meningitis have active primary pulmonary complex or miliary tuberculosis concurrently, and in some patients, pleural effusion is evident.

Eye Careful examination of the eye is important in patients with chronic meningitis. Candidal infection, the most frequent cause of fungal

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Chronic Meningitis

endophthalmitis, manifests as fluffy yellow-white retinal or chorioretinal lesions that can progress to vitritis. If vitritis is found, CMV and toxoplasmosis are excluded from the differential diagnosis. If the eye is involved in patients with cryptococcal meningitis, yellow choroidal or chorioretinal lesions can be seen, with signs of uveitis and, rarely, optic atrophy. In the rare case of coccidioidal endophthalmitis, the chorioretinal yellow lesions often have pigmented borders. Ophthalmologic manifestations of disseminated sporotrichosis are anterior uveitis or conjunctival nodules. Histoplasmosis can cause a distinctive form of choroiditis, including lesions in the peripapillary and sometimes the macular areas. M. tuberculosis infection rarely affects the eye; when it does, small elevated yellow-white choroidal nodules without welldemarcated borders are seen. Tuberculous retinitis is also reported. In syphilis, iridocyclitis is the most commonly reported ocular finding, but choroiditis (sometimes referred to as “salt-and-pepper” choroiditis) and chorioretinitis are also reported. Nocardia or Actinomyces infection rarely involves the eye; when it does, it usually causes necrotizing chorioretinitis. CMV infection is a well-known cause of chorioretinitis, manifesting as intraretinal bleeding and yellow-white exudates. Toxoplasma gondii can also cause chorioretinitis, with yellow-white lesions and an overlying hazy vitreous. Sarcoid granulomas or uveitis can also be identified. If papilledema is found, computed tomography (CT) of the head, to rule out increased intracranial pressure, should precede the lumbar puncture.

Neurologic Examination The neurologic examination may reveal cranial or peripheral nerve abnormalities, which are more common in infections that cause extensive inflammation at the base of the brain. For example, syphilitic meningitis causes palsies of the seventh and eighth cranial nerves; tuberculous meningitis causes palsies in the sixth and sometimes the third and fourth cranial nerves in more than one-third of cases. The eighth cranial nerve is most commonly involved in Brucella meningitis, but second, third, sixth, and seventh cranial nerve neuropathy is also reported, as is peripheral neuropathy, manifesting as spastic paraparesis or neuritis. Seizure is a rare manifestation of neurobrucellosis.45

CHAPTER

43

295

Deficits of cranial nerve function or other facial neurologic deficits are rarely reported in patients with chronic Histoplasma meningitis. Paralysis (especially of the seventh cranial nerve), with peripheral motor and painful sensory radiculoneuropathy, can help distinguish neuroborreliosis from other causes of chronic meningitis, except sarcoidosis. Hyperesthesia is relatively common in cases of Angiostrongylus meningitis and radiculoneuropathy in HIV-infected patients with CMV CNS infection.

Liver Hepatomegaly, not a common finding in patients with chronic meningitis, should direct the diagnosis toward a relatively few pathogens that cause this complication (see Table 43-4).

Cerebrospinal Fluid Analysis CSF is obtained after performance of an imaging study of the brain. Most infectious causes of chronic meningitis elicit similar CSF abnormalities (i.e., a mildly elevated protein level, fewer than 500 WBC/mm3 with lymphocyte predominance along with a normal glucose level). In some cases, however, results of CSF analysis are sufficiently distinctive to aid diagnosis (Table 43-5). In most cases of tuberculous meningitis, the CSF shows mild lymphocytic pleocytosis (50 to 500 cells/mm3), with increased protein levels and a low glucose level.46 Unfortunately, these findings are not specific and can occur in patients with viral, fungal, or bacterial meningitis. Polymorphonuclear cell predominance (usually early in the course of the infection) or normal glucose level (> 40 mg/dL) occurs in almost one-third of patients with tuberculous meningitis. However, repeated lumbar punctures to document a progressive decrease in CSF glucose levels may be helpful in distinguishing tuberculous from viral, fungal, or treponemal meningitis. Careful cytologic examination of the CSF may show Langhans multinucleated giant cells. Only 50% of patients with cryptococcal meningitis have a decreased glucose level with mild lymphocytic pleocytosis (50 to 500 cells/mm3 in CSF). In patients with AIDS, CSF abnormalities can be minimal.15 Most (70%) patients with Histoplasma meningitis have a CSF WBC count of less than 100/mm3 and low glucose levels,13 whereas only 50% of patients with Candida

TABLE 43-5. Predominant Leukocyte Found in CSF for Various Causes of Chronic Meningitis Predominant Leukocytes in CSF Causative Agent

Lymphocytes

Neutrophils

Eosinophils

Bacteria

Mycobacterium tuberculosis Treponema pallidum Borrelia Brucella

Nocardia Actinomyces Brucella Leptospira

Mycobacterium tuberculosis Treponema pallidum

Fungi

All

All

Coccidioides

Parasite

Taenia Toxoplasma

Entamoeba histolytica

Angiostrongylus Taenia solium Baylisascaris

Viruses

Lymphocytic choriomeningitis CMV

Mumps CMVc

Other

Parameningeal focusa Sarcoidosisb Chronic idiopathic meningitis Malignant process Behçet disease

Suppurative parameningeal focusa Systemic lupus erythematosus Chemical meningitis Drug hypersensitivity NOMID/CINCA

Hodgkin disease Chemical meningitis Drug hypersensitivity

AIDS, acquired immunodeficiency syndrome; CMV, cytomegalovirus; NOMID/CINCA, neonatal-onset multisystem inflammatory disease/chronic infantile neurologic cutaneous articular syndrome. a See Box 43–2. b See Box 43–1. c In AIDS patients.

296

SECTION

F

Central Nervous System Infections

meningitis have a low glucose level and lymphocytic predominance in CSF.16 In up to 25% of patients with cysticercosis, CSF analysis reveals lymphocytic pleocytosis and low glucose levels, although in almost 40%, analysis of the CSF is normal. In patients with meningitis caused by Borrelia, CSF analysis typically reveals lymphocytic pleocytosis of 100 to 200 cells/mm3 with mildly to moderately elevated protein and normal glucose levels. Some patients have oligoclonal bands, which represent speciÀc antiBorrelia antibodies. Similar CSF Àndings are common in Brucella meningitis. Sarcoid meningitis is usually associated with mild CSF lymphocytosis (60% to 70% of patients) and, in 20% of patients, the glucose level is low. In Behçet disease, CSF analysis reveals minimal pleocytosis (with both lymphocytes and polymorphonuclear cells), normal glucose, and mild elevation in protein level. CSF pleocytosis with predominance of neutrophils is rarely reported in chronic meningitis and may help in the differential diagnosis.47 Neutrophilic pleocytosis can be present early in the course of patients who have tuberculous meningitis, but conversion to lymphocytic pleocytosis occurs within 10 days; neutrophilic pleocytosis rarely persists for longer than 2 weeks. In patients with chronic meningitis, neutrophilic pleocytosis eliminates the diagnosis of M. tuberculosis infection. Leptospira infection, a rare cause of chronic meningitis, produces persistent neutrophilic pleocytosis. Chronic meningitis due to Actinomyces or Nocardia spp. is usually a consequence of spread from a parameningeal focus; thus, CSF analysis reveals lymphocytic pleocytosis, with a normal or low glucose level and negative results on culture. In the few cases of direct involvement of the meninges or rupture of a parameningeal abscess, neutrophils predominate in CSF. Certain fungi can cause persistent neutrophilic pleocytosis. Aspergillus meningitis, which is usually associated with pleocytosis of lymphocytes, can also cause a persistent neutrophilic response.48 CNS zygomycosis typically produces a few cells in CSF, with an equal number of neutrophils and lymphocytes, but predominance of neutrophils is also reported.19 CSF Àndings in candidal meningitis can be similar to those of pyogenic bacterial meningitis, with predominance of neutrophils (in 30% to 50% of patients) and low glucose levels (in 40% to 60% of patients), both of which can persist.16,47 Neutrophilic pleocytosis in other fungal infections (e.g., Histoplasma, Blastomyces, and Coccidioides) is reported in association with a more acute onset of disease, but, rarely, a persistent neutrophilic response of several weeks’ duration is reported. Entamoeba histolytica is the only parasite that causes chronic meningitis with polymorphonuclear predominance in the CSF. Chronic viral meningitis usually produces a lymphocytic response in CSF, yet mumps virus and lymphocytic choriomeningitis virus can induce a predominant neutrophilic response that can continue for a long time. A suppurative parameningeal focus, such as brain abscess, subdural empyema, cranial osteomyelitis, mastoiditis, or sinusitis, can occasionally leak or extend into the CSF, causing a neurophilic pleocytosis, with or without hypoglycorrhachia. It is important to exclude such a process, especially in patients who are receiving antibiotic therapy that can sterilize CSF but fails to treat the parameningeal process. Several noninfectious causes of neutrophilic pleocytosis in CSF also exist. Chronic meningitis is a rare complication of systemic lupus erythematosus, and in these patients, the CSF response resembles that of bacterial meningitis (i.e., predominance of neutrophils and a low glucose level). Meningitis caused by exogenous chemicals (e.g., radiographic contrast material, chemotherapeutic agents) usually causes a neutrophilic pleocytosis, which persists long after the offending chemical has been given. Drug hypersensitivity (to isoniazid, ibuprofen, or sulfa drugs) can also rarely cause a chronic meningitis-like picture with polymorphonuclear predominance. Most patients with neonatal-onset multisystem inflammatory disease (NOMID) and CNS involvement have predominance of neutrophils in the CSF due to activation of interleukin-6 and tumor necrosis factor.34 High eosinophil counts in the CSF during chronic meningitis is a rare occurrence. If eosinophilia exceeds 10% of WBCs, the likely

causes include Angiostrongylus cantonensis, Taenia solium, and Baylisascaris procyonis (ascarid of raccoons). A history of exposure is very important. Other parasites, such as Toxocara canis, Trichinella spiralis, and Echinococcus spp. can rarely cause CSF eosinophilia, but they are not associated with chronic meningitis. Neurosyphilis, tuberculous meningitis, and coccidioidal meningitis can rarely cause mild CSF eosinophilia (less than 10% of WBCs). Noninfectious processes, such as Hodgkin disease, chemical meningitis, and drug hypersensitivity, should also be considered in the differential diagnosis of CSF eosinophilia.

DIAGNOSIS Analysis of CSF (including cell count and measurement of protein and glucose concentration) can provide direction to the diagnostic investigation (see Table 43-5); stains for bacteria, fungi, and acid-fast organisms should be performed. Special strains (using monoclonal antibodies) of lymphocytes in CSF should be considered if leukemia or lymphoma is suspected. Culture of a CSF sample is important and should include appropriate techniques for isolation of aerobic, facultative anaerobic, and fastidious bacteria (e.g., increased CO2 to enhance growth of Actinomyces, Nocardia, and Brucella spp.); fungi; and mycobacteria. Cultures should be incubated for an extended time because growth of some bacteria and fungi is slow. In addition, Leptospira spp., which usually grow within 7 to 10 days, can require up to 6 weeks before growth is detected. Testing for CSF antibodies to agents, such as Treponema pallidum, Borrelia burgdorferi, Histoplasma, Coccidioides, Sporothrix, Toxoplasma, and Taenia spp., is performed with concurrent serum specimen testing. Measurement of antigen in CSF or blood and polymerase chain reaction (PCR) testing are available for a few pathogens (Table 43-6). Because many patients with chronic meningitis also have disseminated disease, cultures of other body fluids (see later for speciÀc sites) and serologic tests for antigen antibodies (see Table 43-6) can be diagnostic. Skin testing has limited usefulness because many infected patients have negative results (e.g., 35% of patients with tuberculous meningitis) or are positive because of previous exposure during residence in an endemic area. CT and magnetic resonance imaging (MRI) should be done before lumbar puncture to exclude space-occupying lesions, hydrocephalus, or increased intracranial pressure. These radiologic techniques may also help identify a parameningeal infectious site, meningeal inflammation, granulomatous lesions, or cerebral infarction. Cortical brain biopsy usually adds little to the diagnostic process, especially if MRI does not reveal meningeal enhancement.49

Viruses Diagnosis based on viral isolation from the CSF or demonstration of speciÀc antibody responses (in blood or CSF) is slow and insensitive. PCR assays offer rapid detection (within hours if necessary) for many viruses (e.g., enterovirus, CMV, HSV) with excellent sensitivity and speciÀcity. In chronic meningitis where viral replication may be low this method is especially valuable. Recently, reverse transcription (RT) multiplex PCR assay evaluating both enteroviruses and neurotropic herpesviruses was shown to be valuable in identifying the etiologic agents in patients with neurologic infections, including chronic meningitis.50

Mycobacterium tuberculosis Acid-fast bacilli are seen infrequently (in fewer than 25% of cases) on microscopy of a CSF smear from patients with tuberculous meningitis. The yield can be increased by withdrawing a large volume of CSF (5 to 10 mL), which is centrifuged; the sediment is then carefully examined. CSF culture shows positive results in 35% to 85% of

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Chronic Meningitis

CHAPTER

43

297

TABLE 43-6. Diagnostic Tests Available for Causative Agents of Chronic Meningitisa CSF Agent Mycobacterium tuberculosis Treponema pallidum Brucella Borrelia burgdorferi Nocardia Actinomyces Leptospira Mycoplasma Ureaplasma Blastomyces spp. Histoplasma Coccidioides Cryptococcus Candida Aspergillus Sporothrix Toxoplasma Angiostrongylus Taenia Viruses

Stain Acid-fast Dark-Àeld • • •b •b

• • •b India ink • • • Wright-Giemsa Giemsa

Culture • • •b • • •b • • •b •b • • • • • •b

Antigen

Blood Antibody

c

•

VDRL, FTA-ABS, IgM • EIA, IFA, WB

PCR • • • • • •

•

• CFd CF, anti-33kd •

Ag

FTA-ABS, VDRL EIA EIA,d WB, IFA • EIA EIA •d •d •d

LA, EIA • EIA •

Ab

•

• EIA EIA, IFA

•

• •

• •

•

WB •

• •

Ab, antibody; Ag, antigen; CF, complement Àxation; CSF, cerebrospinal fluid; EIA, enzyme immunoassay; FTA-ABS, fluorescent antibody absorption; IFA, indirect fluorescent antibody; IgM, immunoglobulin M; LA, latex agglutination; PCR, polymerase chain reaction; VDRL, Venereal Disease Research Laboratory; WB, Western blot. a Bullets indicate test available. b Rarely positive. c Tuberculosteric acid. d Lacks speciÀcity.

patients; cultures of sputum which are positive in 50% of cases and gastric aspirate may increase diagnostic yield. Few rapid diagnostic tests are available. Detection of tuberculostearic acid, a component of Mycobacterium spp. in CSF, is speciÀc and sensitive (almost 90% for each) for M. tuberculosis.51 Nocardia can also express this chemical. Elevated levels of adenosine deaminase (released by T lymphocytes) have been found in patients with tuberculous meningitis but not in those with viral or bacterial meningitis.52,53 Although speciÀcity is generally high, increased levels of adenosine deaminase were also found in patients with CNS sarcoidosis, brucellosis, and tumors. PCR of a CSF sample is the best method for rapid diagnosis of tuberculous meningitis. PCR is more sensitive than is the combination of acid-fast stain and culture; the sensitivity ranges from 30% to 82% (probably related to the density of mycobacteria in the CSF, methods used for preparation of the sample, and the type of primers used), with a detection limit as low as two mycobacteria per reaction.54 The speciÀcity of the test is 95% to 100%. The sensitivity of the PCR method is greater in HIV-infected patients than in immunocompetent patients. The PCR test can remain positive for several weeks after treatment is initiated.55

Treponema pallidum The diagnosis of syphilitic meningitis can be difÀcult. A serum fluorescent treponemal antibody absorption (FTA-ABS) test is positive in more than 95% of patients. If the serum FTA-ABS test result is negative, the probability of syphilitic meningitis is very low (except in patients with AIDS, who may fail to produce antibodies). Useful tests for CSF samples include speciÀc treponemal immunoglobulin (Ig) M antibodies,56 Venereal Disease Research Laboratory (VDRL), and FTA-ABS. The speciÀcity of the VDRL test on CSF is high, but its sensitivity is low (30% to 70%). A nonreactive result does not exclude the diagnosis. False-positive reaction may be due to blood contamination or high CSF protein. In contrast, a negative result of FTA-ABS test rules out neurosyphilis, but a positive test (i.e., speciÀcity) does not conÀrm the diagnosis because false-

positive results can occur as a result of CSF contamination with blood or small amounts of antibodies from the serum. The PCR technique has been shown to be sensitive for diagnosis of acute neurosyphilis but to be less useful in chronic cases.57 The method is also less effective in diagnosis of disease in HIV-infected patients (less than 30% sensitivity).

Brucella Species In Brucella meningitis, CSF culture results are positive in fewer than 50% of patients; cultures of blood and, especially, bone marrow increase the diagnostic yield. The diagnosis of meningitis can be substantiated by detection of speciÀc Brucella agglutination titers in the CSF or serum; enzyme immunoassay (EIA) of CSF sample is both sensitive and speciÀc.58

Borrelia burgdorferi These bacteria can be cultured only rarely from CSF, by using a speciÀc culture medium (Barbour–Stoenner–Kelly (BSK)). The presence of speciÀc antibody (i.e., IgA, IgM, or IgG) in CSF is the most useful evidence for neuroborreliosis.59 The serologic tests available (e.g., EIA, indirect fluorescent antibody (IFA), immunoblotting) should be used with caution. They are more useful in the chronic stage of the disease than in earlier stages. Positive or equivocal results should be veriÀed by Western blot assay.60 Antibody tests on serum do not distinguish active and inactive infection, especially in residents of an endemic area. A DNA PCR method for detecting B. burgdorferi in a CSF sample has been shown to be helpful for diagnosis.61 The PCR technique is better than culture, with a yield of 40% to 50% compared with a yield of < 5% for culture. The PCR yield was lower in patients with chronic neuroborreliosis; the relatively low yield can be explained by the small number of spirochetes in CSF samples and their preference to reside in perivascular locations.

298

SECTION

F

Central Nervous System Infections

Fastidious Bacteria Gram stain of CSF samples is too insensitive for diagnosis of meningitis caused by Nocardia or Actinomyces; centrifugation of a large volume of CSF before staining increases the yield. Cultures should be held for at least 2 weeks. Serologic methods, such as complement fixation (CF) and hemagglutination, lack specificity because of cross-reactivity with Mycobacterium and Streptomyces spp. A 55-kd protein isolated from N. asteroides has been used in a more specific serologic test.62 Leptospira spp. can be isolated from the CSF by using special culture media, but only during the first few weeks of infection. Culture results are routinely negative in cases of chronic leptospiral meningitis. PCR technique may increase the yield identification.63 Serologic tests include an agglutination test, the indirect hemagglutination test, and the highly sensitive and specific EIA.

Mycoplasmataceae Both Mycoplasma hominis and Ureaplasma urealyticum should be considered in neonates (especially if premature) with evidence of chronic meningitis and negative CSF Gram stain and routine culture. The organisms grow easily in Mycoplasma broth medium and on agar enriched with yeast extract and animal serum. They can also be detected by PCR techniques.

Blastomyces Species Although direct examination of the CSF and culture is recommended for diagnosis of meningitis associated with blastomycosis, only 25% of the cases yield positive results on culture. The serum immunodiffusion test is specific and is positive in almost 90% of patients with chronic disseminated disease; the CF test lacks both sensitivity and specificity. A sensitive, commercially available EIA is helpful in initial screening. No diagnostic skin test is currently available.

Histoplasma Species CSF culture results for Histoplasma are positive in only 30% to 65% of cases, and CSF antibody tests (i.e., CF or radioimmunoassay (RIA)) lack specificity because of cross-reactivity with other fungi. Patients with positive serologic test results should be tested for cryptococcal antigen and Coccidioides in endemic settings. Serum antibody tests are also not helpful because up to 15% of residents in an endemic area without active disease have positive results. In contrast, the RIA for antigen is highly specific. In AIDS patients with disseminated histoplasmosis, antigen is present in 50% and 90% of blood and urine samples, respectively; antigen is also detectable in CSF of patients with meningitis,64 but false-negative test results have been reported. The precise sensitivity of the antigen test is not yet determined. The skin test, which is valuable for epidemiologic surveys, is useless for diagnosis, especially in endemic areas.

Coccidioides Species Visualization of Coccidioides immitis in CSF samples is rare, and although the organism is not fastidious in growth requirements, only about one-third of patients with meningitis have a positive culture result. Removal of large volumes of CSF followed by sterile filtration and culture of the filter increases the yield. CF antibodies are present in the CSF of 60% to 95% of patients with chronic coccidioidal meningitis; titers can be used to monitor the course of the disease. Detection of antibodies against a 33-kd antigen from C. immitis by EIA was reported to be more sensitive than the CF antibody test.65 Skin tests using the coccidioidal antigen (spherulin) yield positive

results in almost 50% of patients with chronic meningitis but should be interpreted with caution because there is some degree of crossreactivity with other fungi.

Cryptococcus Species The India ink stain or the fluorescent stain (the commercially available Calco-Fluor white stain) is important in screening CSF samples of patients with suspected cryptococcal meningitis. The test is positive in 50% of patients, but false-positive results can occur; culture of a CSF sample is required to confirm the diagnosis. Large volumes of CSF (5 to 10 mL) should be removed for culture on at least three occasions to increase the yield of isolation. Detection of cryptococcal antigen in CSF is the most clinically useful test. The commercially available latex agglutination test detects antigen in more than 90% of patients with cryptococcal meningitis. When a CSF specimen is tested, pronase treatment is not needed (but such treatment does improve results when serum is tested). Both blood and CSF should be tested for cryptococcal antigen. False-positive results occur, but other tests, such as EIA, are more sensitive and specific.66 The possibility of capsule-deficient cryptococcal meningitis should be considered in the rare event that the diagnosis is strongly suspected but the staining and antigen detection tests are negative. In such cases, the diagnosis can be established by indirect immunofluorescence staining or reformation of the capsule by inoculation into mice.67 The serum antibody test is not useful because healthy persons in endemic areas frequently have positive test results, and many patients with cryptococcal meningitis have negative test results. Cultures of blood, urine, and sputum (if available) should be obtained, even in patients without clinical manifestations of specific organ involvement. Skin tests are not helpful.

Other Fungi The diagnosis of meningitis caused by Candida or Aspergillus can be made by microscopic examination or culture of CSF. A single positive test result on CSF culture for Candida should be interpreted cautiously, unless accompanied by CSF indices suggestive of meningitis and other risk factors for infection (e.g., prematurity, prolonged antibiotic therapy, CNS shunt, immunocompromised host, intravenous catheterization).68,69 Multiple positive culture results for Candida in CSF, or from CSF and other sites (e.g., blood, urine), suggest CNS infection. Recovery of Aspergillus from CSF, even on a single culture result, is highly suggestive of the diagnosis, especially when the expected clinical features (e.g., other CSF parameters, neutropenia, leukemia, HIV infection) are present. Serologic tests are neither sensitive nor specific enough to be useful in the diagnosis. Some studies suggest that antigen detection (quantitative galactomannan EIA) and PCR can help to establish the diagnosis of Aspergillus infection before culture results are available or when cultures are negative.70,71 The diagnosis of meningitis caused by Sporothrix is best made by CSF culture. Serologic tests may be useful in the diagnosis, but the presence of antibodies in individuals without evidence of sporotrichosis suggests that more experience with such tests is needed.21

Toxoplasma Species Toxoplasma gondii can be detected by a simple Wright-Giemsa stain of sedimented CSF. Other stains, such as fluorescent antibody and immunoperoxidase, have been used with variable success. Although isolation of Toxoplasma from mice after peritoneal inoculation of the CSF specimen establishes the diagnosis, most strains isolated from humans are avirulent in the mouse and induce only an antibody response, which delays the diagnosis. Tissue cell culture is a less

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Recurrent Meningitis

sensitive method, but results are available more rapidly (3 to 6 days). An EIA to detect Toxoplasma antigen in serum usually produces positive results in the acute phase of the disease but negative results in chronically infected individuals; its value in patients with chronic meningitis is unknown. The PCR technique has been used successfully to detect the parasite in CSF samples from patients with AIDS.72 Although intrathecal production of antibodies against Toxoplasma has been reported in patients with encephalitis, it may be absent in patients with AIDS.72 The most useful serologic tests are the Sabin-Feldman dye test (now available in only a few reference laboratories), EIA (IgG, double-sandwich IgM and IgA), IFA test (IgM), and the modiÀed agglutination test.73

Angiostrongylus cantonensis

CHAPTER

Taenia Species

THERAPY When the initial evaluation of the patient identiÀes the cause of chronic meningitis, speciÀc therapy should be instituted. When choosing the antimicrobial agent or agents, the drug’s ability to penetrate the blood–brain or CSF barrier in order to achieve consistent and adequate drug levels is of paramount importance. For example, isoniazid and pyrazinamide penetrate the blood–CSF barrier well during meningeal inflammation, achieving high concentrations in the CSF; they are thus Àrst-line drugs for treatment of tuberculous meningitis. Rifampin penetrates the CSF less well; a higher dose (15 to 20 mg/kg per day) is necessary to achieve adequate CSF levels. In contrast, the CSF penetration of streptomycin is limited; thus, it should be used rarely (i.e., in patients with multidrug-resistant M. tuberculosis infection). The use of corticosteroids to reduce the inflammatory response appears to be beneÀcial, but their role in preventing cerebral infarction is questionable.74 Chapters on speciÀc pathogens contain speciÀc recommendations. After the initial evaluation, the cause of chronic meningitis remains unknown in many patients. In these situations, decisions regarding use of empiric treatment and preferred antimicrobial agents are controversial. Because tuberculous meningitis is probably the most common cause of chronic meningitis, an empiric trial with antituberculous drugs is favored by many clinicians. Such an approach was not found to be beneÀcial in one study.36 If empiric antituberculous therapy is initiated, further studies should be done in patients who are seriously ill or who deteriorate rapidly. Corticosteroid therapy should only be added if the patient continues to deteriorate on antituberculous therapy or if fungal meningitis is reasonably excluded. Several patients respond favorably to empiric corticosteroid therapy

44

299

alone,36 but such treatment should be considered cautiously because of the potential catastrophic outcome when given to patients with unrecognized tuberculous or fungal meningitis. Corticosteroids were shown to be beneÀtial in shortening the duration of headaches and the need for repeated lumbar puncture in patients with A. cantonensis meningitis.75 If fungal meningitis is a possibility, antifungal treatment is considered. Choice of agent should be made on the basis of the suspected fungus and the degree of CSF penetration of the agent. In previously healthy patients who have had symptoms of meningitis for several weeks and who are not seriously ill, it is reasonable to investigate the cause of meningitis without initiating treatment. If the cause cannot be found and the patient is not improving, empiric therapy should be considered. If antituberculous medications are initiated and the patient does not improve within a few weeks, antifungal therapy should be considered. In the 30% of chronic meningitis cases in whom no diagnosis is found, the course is benign.

A Giemsa stain of the CSF is the most important test for prompt diagnosis, espically in travelers returning from endemic areas. Western blot analysis can be helpful in the diagnosis when speciÀc antigen bands (29, 31, 55, 85 to 99, 200 to 204 kd) are seen. CT is usually non-revealing, but MRI may show enhancing subcortical lesions.

CT or MRI is suggested for diagnosis of cysticercosis; Àndings usually show intracranial calciÀcation or cystic lesions with enhancing ring following administration of contrast material. Detection of subcutaneous cysts and skeletal muscle calciÀcations on plain radiography can be helpful in the diagnosis, although the yield is low (10% to 20%). Serologic tests of CSF samples are helpful in identifying more than 90% of patients with chronic meningitis, but negative results on serum serologic tests do not rule out the diagnosis because up to 50% of patients with neurocysticercosis have negative results.

CHAPTER

44

Recurrent Meningitis Ram Yogev Recurrent meningitis is rare and is difÀcult to diagnose. For the purpose of this chapter, recurrent meningitis is deÀned as two or more episodes of meningitis separated by a period of complete resolution of signs, symptoms, and laboratory Àndings. Many conditions can cause recurrent meningitis. In most cases, a bacterial pathogen can be identiÀed that is helpful in directing the diagnostic evaluation and the therapeutic efforts. In some cases, a nonbacterial condition, with sterile cultures of cerebrospinal fluid (CSF), can be identiÀed as the cause of the recurrent episodes.

ETIOLOGIC AGENTS Streptococcus pneumoniae is the infectious agent identiÀed in more than 50% of patients with recurrent bacterial meningitis. In contrast to single-episode S. pneumoniae meningitis, most cases of which are caused by 8 to 10 serotypes, no particular serotypes predominate in cases of recurrent meningitis. Neisseria meningitidis is the second most common pathogen, accounting for 10% to 20% of cases of bacterial meningitis, and is more commonly associated with immunologic abnormalities. Haemophilus influenzae has been reported less commonly as a cause since the introduction of the conjugate vaccine, with nonserotypable organisms more commonly isolated than serotypable organisms (i.e., type b). Staphylococcus aureus, gram-negative bacilli such as Escherichia coli, and Enterobacter spp. are rarely associated with recurrent meningitis; the isolation of any of these organisms suggests a communication between the subarachnoid space and the skin (such as a dermal sinus) or intestine (such as posterior neurenteric Àstula or anterior meningomyelocele). Other bacteria, such as oral streptococci, Citrobacter, Proteus spp., Ralstonia mannitolilytica,1 and group B streptococcus, have been sporadically reported to cause recurrent meningitis. If such a bacterium is isolated, nosocomial infection of intraventricular devices (e.g., shunt catheter, Omaya reservior) should be suspected. In cases in which bacteria cannot be found, the differential diagnosis of recurrent aseptic meningitis should be considered (Box 44-1).2–21

300

SECTION

F

Central Nervous System Infections

BOX 44-1. Conditions Associated with Recurrent Aseptic Meningitis 2

Mollaret syndrome Familial Mediterranean fever (FMF)3 Intracranial or intraspinal cyst4 Dermoid or epidermoid tumor5 Behçet syndrome6 Sarcoidosis7 Systemic lupus erythematosus8 Lyme disease9 Intracranial tumors10 Aneurysm11 Drug-induced12,13 Herpes simplex type 1 or 2 infection14,15 Enterovirus16 Hydatid cyst17 Hemosiderosis of the central nervous system17 Migraine syndrome18 After neurosurgery in posterior fossa19 Ferreol–Besnier disease20 Lipoid meningitis21

BOX 44-2. Congenital Anomalies Associated with Recurrent Bacterial Meningitis ANOMALIES OF THE ANTERIOR FOSSA Encephalocele Meningocele Defects of the cribiform plate Enlarged subarachnoid space Intracranial cyst ANOMALIES OF THE TEMPORAL BONE Mondini dysplasia Stapedial anomalies Klippel–Feil syndrome Pendred syndrome Petromastoid fistula Widened cochlear aqueduct Hyrtl fissure SPINAL DEFECTS Dermoid or epidermoid cyst Neuroenteric fistula

EPIDEMIOLOGY The incidence of recurrent meningitis is low and is different for each of the various predisposing conditions. A 1999 review of 463 children with bacterial meningitis identified 6 patients with confirmed recurrent episodes, representing an incidence of 1.3%.22 The most common predisposing condition is a communication between the subarachnoid space and the base of the skull resulting from head trauma or a congenital defect; immunologic defects are less common but important. CSF rhinorrhea or otorrhea occurs after 2% to 3% of all head injuries, 11% of skull base fractures,23 and 25% of fractures involving the paranasal sinuses.24 Most such “CSF leaks” resolve spontaneously. However, 12% to 30% of patients with CSF leak experience recurrent bacterial meningitis.25 The risk of development of meningitis at a particular time after the injury is variable. About a fifth of cases of posttraumatic meningitis occur within a week or two of the head trauma,26 a third within the first month, and half within 6 months. Almost 50% of patients have the first episode of meningitis 7 months or longer after the traumatic event. Delay of as long as 30 years between occurrence of a CSF leak and development of an intracranial infection has been reported.27 Surgical procedures such as transsphenoidal pituitary surgery, craniotomy, nasal and paranasal sinus surgery, mastoidectomy, and stapedectomy can also predispose to recurrent meningitis.24 On average, postneurosurgical meningitis develops within 10 days from the operation with an incidence of 2.5% to 8.9%.28 Gram-negative bacteria are the most common etiologic agents. A variety of congenital defects in the bony structure of the skull contribute to the development of CSF fistulas and thus predispose to recurrent meningitis; Box 44-2 lists some of them. The incidence of recurrent meningitis in patients with congenital defect is unknown but is probably low, although the defects themselves are not uncommon. For example, in children with unexplained (idiopathic) sensorineural hearing loss, computed tomography (CT) examination of the temporal bone revealed that 21% had an inner-ear malformation.29 Although Mondini dysplasia (a developmental arrest characterized by hypoplasia of the cochlear labyrinth, which thereby has less than two and one-half turns) is commonly cited as contributing to recurrent meningitis, a more severe form of cochlear dysplasia (i.e., a single turn) is probably required to lead to spontaneous CSF fistula to the inner ear (Figure 44-1), which may lead to meningitis.30 Additionally, trauma can precipitate a fistulous connection in milder defects and lead to meningitis. Although deafness is frequently associated with Mondini dysplasia, it is unilateral and was previously frequently unrecognized. Epidermoid and dermoid cysts with dermal sinus tract are well-known causes of recurrent meningitis or meningitis caused by unusual bacteria, but the incidence of the meningitis is unknown.31

Figure 44-1. Computed tomography (with contrast enhancement) of the right ear. The cochlea is seen as a single cavity (black arrow). Note contrast material in the common cavity malformation of the cochlea and in the mastoid (open arrows), which suggests a communication between the subarachnoid space and the middle ear. (Courtesy of N.M. Young, MD, Children’s Memorial Hospital, Chicago.)

A rare cause for CSF leak and recurrent meningitis is lymphangiomatosis of the skull.32 The incidence of recurrent bacterial meningitis after a parameningeal focus of infection (Box 44-3) or related to more distant foci (e.g., endocarditis) is also unknown, but these entities should be considered in the differential diagnosis.

Immunologic Defects Several disorders of the immune system can predispose to recurrent meningitis. They include hypogammaglobulinemia, acquired immunodeficiency syndrome (AIDS), leukemia, lymphoma, splenic dysfunction (i.e., surgical splenectomy, congenital asplenia, sickle-cell hemoglobinopathies), and complement deficiencies (see Chapter 105, Infectious Complications of Complement Deficiencies, and Chapter 108, Infectious Complications in Special Hosts). The frequency of a complement deficiency in the general population is estimated to be

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Recurrent Meningitis

BOX 44-3. Parameningeal Foci of Infection Associated with Recurrent Meningitis Sinusitis Mastoiditis Brain abscess Subdural empyema Central nervous system shunt infection Infected dermoid sinus Infected porencephalic cyst Cranial or vertebral osteomyelitis Neurenteric fistula Anterior meningomyelocele

0.09%.33 Deficiency of properdin (the molecule that stabilizes the C3bBb complex) is rare. Meningitis occurs in 39% of patients with complement deficiencies and 6% of those with properdin deficiency.34 Affected individuals, however, can experience fulminant, usually fatal, meningococcal infection, unlike the milder meningococcal infections usually seen in patients with terminal component deficiencies.35 C4 deficiency is relatively common and can lead to recurrent meningitis.36 Deficiency of factor I, which plays an important part in the regulation of both the classical and the alternative C pathways (by cleaving the a chains of C4b andC3b) is rare and may lead to recurrent bacterial meningitis (as well as increased incidence of systemic lupus erythematosus).37 Recurrent bacterial meningitis (especially due to N. meningitidis) also occurs in patients with deficiency of one of the terminal complement components (i.e., C5 to C9). The incidence of these complement deficiencies in patients with systemic infections due to Neisseria spp. has been estimated to be 10% to 15%.33,38 It is noteworthy that reported complement deficiencies in African Americans almost exclusively involve terminal components. Almost 50% of patients with terminal complement deficiencies have N. meningitidis infections; almost a quarter have no adverse effect; and 10% to 15% experience autoimmune manifestations. H. influenzae, gram-negative enteric bacilli, and mycobacteria also cause infections in this population, but so rarely that the incidences may not exceed those in the general population. In general, C5 deficiency is associated with a higher incidence of bacterial meningitis than deficiencies in more terminal components, probably because C5 deficiency is associated with a defect in chemotaxis that limits the ability to localize infection. Almost 50% of individuals with C6 deficiency, and 40% with C7 deficiency, experience bacterial meningitis. Specific defects in C8 are predictable according to race; deficiency of C8 subunit (the portion responsible for the membranolytic function) is seen in African American and Hispanic individuals, whereas deficiency of the C8 portion responsible for attachment of C8 to C5b67 is seen in whites. Most patients with C9 deficiency are healthy and rarely have meningococcal meningitis. The incidence of meningococcal infection is 42% in persons with a complement deficiency, compared with 0.0072% in the general population.31 The median age for the first episode of meningococcal infection in the general population is 3 years (peak age, 3 to 8 months), compared with 17 years in individuals with complement deficiency. In addition, recurrent meningococcal infection rarely occurs in the general population, and relapse (i.e., reappearance of infection during convalescence despite appropriate antimicrobial therapy) is uncommon, with a rate of about 0.6%. In contrast, the recurrence rate in patients with complement deficiencies is high, approximately 45%, and the relapse rate is 10 times higher than in the general population, about 6%. The mortality rate for meningitis due to N. meningitidis is estimated to be 14% in the general population but only 3% in complement-deficient patients. In addition to the well-known association between complement abnormalities and recurrent meningitis, a deficiency in some of the immunoglobulin (Ig) subclasses (such as IgG2 and IgG3 subclasses) is also rarely associated with recurrent meningitis, as well as mannose-

CHAPTER

44

301

binding lectin mutations with susceptibility to invasive meningococcal disease.39,40,40a

PATHOGENESIS Trauma and Congenital Defects Recurrent meningitis after head trauma only occurs if, in addition to fracture, there is a tear of the dura. Because the dura is more tightly bound to the base of the skull in children than in adults, a higher incidence of CSF leak occurs after trauma in children. CSF leak serves as a portal of entry to the central nervous system (CNS) for bacteria of the normal flora of the nasopharynx. Fracture through the paranasal sinuses can manifest as an apparent CSF leak only transiently or not at all, the only indication of the abnormal communication being recurrent bacterial meningitis. Fracture of the temporal bone is the most common cause of CSF leak into the middle ear.41 After fracture of the petrous bone, CSF can occasionally find its way to the nasopharynx, but the expected otorrhea does not occur. It is important to note that recurrent meningitis can occur years or even decades after the skull fracture occurred. Thus, a careful history-taking for previous head trauma is very important.42 Encephalocele is a rare cause of CSF leak which predisposes the patient to recurrent meningitis. The probable etiology is a small congenital defect in the skull bone which allows cerebral tissue to protrude, but trauma, intracranial infections, tumor, or a surgical procedure can be the cause. In patients with congenital abnormalities of the inner ear, several mechanisms contribute to the communication between the CSF and the middle-ear space.43 First, the cochlear aqueduct (i.e., the perilymphatic duct), that normally connects the subarachnoid space with the inner ear, can be abnormally wide, allowing CSF to flow freely to the inner ear. Second, Hyrtl fissure (an embryonic cleft connecting the posterior fossa and the middle ear), which is usually obliterated during ossification of the petrous bone, can remain patent, causing leakage of CSF into the middle ear. Third, there may be an abnormal communication between the internal acoustic canal (traversed by the eighth nerve) and the perilymph of the vestibule. CSF leakage from the vestibule to the middle ear occurs most commonly through the oval window. A hole in the stapes footplate and CSF leakage through the oral window due to increased vestibular perilymph pressure that displaced the stapes footplate were also described.44, 45 In Pendred syndrome (congenital perceptive hearing loss with enlargement of the thyroid and abnormal perchlorate test result), the pathologic abnormality predisposing to recurrent meningitis is a cochlear defect that resembles Mondini dysplasia.46 In Mondini dysplasia, the CSF leaks into the middle ear through a deficient foramen ovale.47 In Klippel–Feil syndrome (low hairline with short neck and limitation of neck movement), several anomalies can predispose to recurrent meningitis.48 The most common are inner-ear abnormalities, including fistulas through the oval window, the stapes footplate, or the round window, or along the facial nerve. Anomalies of vertebral bodies with neuroenteric cyst or fistulas or a dermoid cyst, with or without dermal sinus tract, have also been described in children with Klippel–Feil syndrome and can serve as portals of entry for bacteria. Occult CNS abnormalities should be suspected in patients without obvious predisposing CNS abnormalities or underlying immunologic defects who have recurrent meningitis. A dermal sinus tract ending intracranially (Figure 44-2) or intraspinally (Figure 44-3) in a dermoid cyst allows communication between the skin flora and the CSF.49 Thus, meningitis caused by bacteria that usually inhabit the skin (i.e., Staphylococcus spp., gram-negative bacilli) should prompt a careful search for such a lesion. Other abnormalities, such as an enteric cyst contained within a meningocele,50 continuous sequestration of bacteria on a CNS shunt or in paraventricular sites (i.e., brain abscess51), or neuroenteric fistula, can serve as the entry points. Rare cases of recurrent meningitis secondary to migration of ventriculoperitoneal shunt to the gastrointestinal tract or after spinal arthrodeses for scoliosis have also been reported.52,53

302

SECTION

F

Central Nervous System Infections

process explains recurrent infections, it fails to explain adequately the unique susceptibility to N. meningitidis but not to other gram-negative bacteria, such as H. influenzae, in such persons. Additional factors that influence the activation of the complement system (e.g., the presence of sialic acid on the bacterial surface) may also be important in defining the unique susceptibility of these patients. They could explain the excessive occurrence of N. meningitidis group Y infection in patients with complement deficiency compared with meningococcal infections in the general population (44% versus 11%, respectively).31 The lower rate of mortality may be the result of lower virulence of infecting organisms or more effective recruitment of other host defense mechanisms, such as the C3b complement pathway.

Drug-Induced

Figure 44-2. Magnetic resonance image showing dermal cyst (long arrow) with sinus tract (short arrow) extending to the skin surface of the nose. (Courtesy of T.T. Tomita, MD, Children’s Memorial Hospital, Chicago.)

Figure 44-3. A lumbosacral dermal sinus tract (broad black arrow) leading into a dermal cyst (thin black arrow). Note the white material extruding from the cyst, which may cause chemical meningitis (open arrow). (Courtesy of J. Grant, MD, Children’s Memorial Hospital, Chicago.)

Drug-induced recurrent meningitis has been reported in connection with several antibiotics (e.g., trimethoprim-sulfamethoxazole, isoniazid, penicillin, metronidazole, and ciprofloxacin), nonsteroidal antiinflammatory drugs (e.g., ibuprofen, tolmetin, naproxen), cytotoxic drugs, (e.g., azathioprine, cytosine arabinoside), OKT3 monoclonal antibody, and intravenous immune globulin (IGIV).12,54,55 The pathogenesis of meningitis in these situations is either a direct chemical irritation by the drug or a hypersensitivity reaction. In the case of ibuprofen-induced meningitis, immune complexes have been found in the CSF, suggesting an allergic reaction.56 The CSF findings in many cases of drug-induced meningitis resemble those of bacterial or aseptic meningitis (low glucose concentration and brisk polymorphonuclear pleocytosis, with variable height of protein elevation).

Mollaret Meningitis Mollaret meningitis is characterized by recurrent episodes of aseptic meningitis that last for 3 to 5 days.2 Symptoms usually disappear as rapidly as they appear, and corticosteroid therapy may abort or prevent attacks. Analysis of CSF reveals pleocytosis (up to several thousand cells per mL), predominantly polymorphonuclear cells. Large “endothelial” cells, with an irregular outline of nuclear and cytoplasmic membranes and tendency to lyse rapidly, are often seen early in the attack. Monoclonal antibody techniques suggest that these cells are monocytes–macrophages.57 The protein level is moderately increased, and glucose concentration is normal or slightly decreased. Several causes have been postulated for Mollaret meningitis. First, eosinophilia in a few patients has suggested a possible allergic mechanism. Second, the occurrence of uveitis, transitory facial paralysis, and positive Babinski reflex in a few patients suggests that Behçet syndrome and sarcoidosis may be related etiologically to Mollaret meningitis. Third, resemblance of Mollaret meningitis to familial Mediterranean fever, specifically, the pattern of recurrent attacks and the response to colchicine, suggested the possibility that the conditions were a single disease.57 Fourth, an epidermoid cyst could produce chemical meningitis upon release of its contents into the subarachnoid space. However, recent studies using a herpes simplex virus (HSV) type-specific polymerase chain reaction assay have shown that the CSF of most patients with benign recurrent lymphocytic meningitis, including Mollaret meningitis, contained HSV DNA.14,15 In most reported cases, HSV-2 DNA was detected, suggesting that recurrent lymphocytic meningitis may be a unique presentation of HSV-2 infection of the CNS, in contrast to the “classic” HSV-1 nonrecurrent encephalitis.58

CLINICAL MANFESTATIONS

Immunologic Defects The pathogenesis of bacterial infections in patients with complement deficiencies is not fully understood. In individuals with complement deficiency, as in the general population, specific antibodies against pathogenic bacteria develop after exposure early in life. Complement defect, however, precludes bactericidal activity, and the affected persons experience a lifelong susceptibility to these pathogens. Although this

The signs and symptoms of recurrent meningitis are determined by etiology. Most patients have symptoms suggestive of acute bacterial meningitis but the intensity may be milder, resembling those of aseptic meningitis. The presence of rhinorrhea or otorrhea should prompt a detailed examination for skull fracture or congenital bony malformations. Anosmia, deafness, fluid-filled middle ear, and hemotympanum are important clues to the presence of congenital or

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Recurrent Meningitis

acquired CSF fistula or head trauma. However, the physical findings are usually completely normal. CSF rhinorrhea or otorrhea can be intermittent (and may cease with pressure occlusion of fistulas during acute episodes of meningitis). Certain maneuvers, such as coughing, sneezing, Valsalva maneuver, and bending the head forward, may initiate the rhinorrhea. Some patients complain of a headache while they are sitting up that disappears when they lie down. Special attention should be given during the physical examination to evaluation for a possible dimple or bulge, tract, tuft of hair, nevus, or hemangioma along the craniospinal axis (e.g., occipital, lumbosacral, and midline face). Some patients note recurring fluid discharges from sites of a lesion (Figures 44-4 to Figure 44-6).

DIAGNOSIS After documentation of recurrent episodes of meningitis, the medical history should be carefully reviewed. Specific attention is given to any history of head injuries or cranial surgery, multiple serious non-CNS bacterial infections, family history of recurrent infections, splenectomy, fluid leakage from ears or nostrils, and medication use. In addition, any child with known deafness or a family history of anomalies of the ear who experiences meningitis should be evaluated

A

CHAPTER

44

303

for anatomic abnormalities. So also should individuals whose deafness is discovered during meningitis after trauma. As many cases of recurrent meningitis have an otorhinologic etiology, an otolaryngologist should be consulted, especially if congenital anomalies of the temporal bone or anterior skull are suspected. Drummond et al. suggested an algorithem for diagnostic tests in pediatric patients with recurrent meningitis of unknown origin.22 The results of tests performed on CSF from previous episodes of meningitis help differentiate bacterial causes from aseptic causes (see Boxes 44-1 and 44-2). The type of bacteria isolated can lead to preliminary identification of the site of the predisposing condition. For example, recurrent Streptococcus pneumoniae or H. influenzae meningitis (or meningitis due to oral streptococci or anaerobic bacteria of oropharyngeal flora) is more common in patients with head trauma or congenital CSF communication to the middle ear, nose, or sinuses, whereas isolation of N. meningitidis suggests a complement deficiency. Recurrent pneumococcal infection, including meningitis, rarely occurs in individuals with specific inability to respond to bacterial polysaccharides.59 If Staphylococcus aureus, enteric or environmental gram-negative bacteria, anaerobic bacteria, or a combination of such organisms is isolated, a dermal sinus tract or sequestration of bacteria in the CNS should be suspected. Recurrent meningitis can follow second esophageal or rectal dilatation

B

Figure 44-4. Dermoid cyst. A 6-year-old boy had a 3-week history of severe headache and then seizure, hemiparesis, and obtundation. Enterobacter agglomerans was isolated from a frontal lobe brain abscess and subarachnoid space. Postoperatively a midline “comedone” was noted (A), which on closer inspection was a pit with a tuft of hair (B) overlying a dermoid cyst and tract to the subarachnoid space. The family recalled that the nose lesion had periodically discharged fluid over the child’s lifetime. (Courtesy of E.N. Faerber and S.S. Long, St. Christopher’s Hospital for Children, Philadelphia, PA.)

304

SECTION

F

Central Nervous System Infections

Figure 44-5. Dermoid cyst. “Pit and pucker” on the midline of the back of a boy with a dermoid cyst connecting to the subarachnoid space. (Courtesy of J. Bass through J.H. Brien.)

procedures to correct strictures. Organisms reflect the flora of the mucosal site. If CSF rhinorrhea or otorrhea is suspected, the fluid should be tested for glucose. The test is useful if the result is positive, but is insensitive. Analysis of the fluid for b2-transferrin, a protein found in the CSF but not in blood or ear and nasal secretions, has been shown to be helpful in detecting CSF in patients with suspected CSF leak.60 Several imaging methods are available for the evaluation of suspected CSF fistulas and anomalies. Plain radiography or polytomography can detect large defects or abnormalities. CT and magnetic resonance imaging (MRI) are more precise for localizing smaller lesions; multiple coronal views must be obtained, because small defects can be missed in axial views.61 The fast inversion recovery for myelin suppression (FIRMS) MRI has been shown to be more sensitive than the T1,T2 MRI in detecting encephaloceles.62 Use of dyes or contrast material (Figure 44-7) is helpful in visualizing smaller bony defects.63 Radionuclide cisternography alone or in combination with pledgets placed in the nostrils or ears is considered by some to be the method of choice for localizing occult CSF fistulas in the skull.23 T2-weighted MRI cisternography is better than CT cisternography in identifying bony defects without active CSF leak.64 Using an infusion pump that instills the radiopharmaceutical indium 113 diethylenetriamine pentaacetic acid (113In DTPA) with a constant pressure (i.e., 40 to 50 cm3) and strategic placement of nasal and otic pledgets (pressure-infusion radionuclide cisternography (PIRC)) enhances the ability to identify and localize CSF fistulas undetectable by other methods (JJ Conway, personal communication, 2000). For diagnosis of complement deficiencies, the initial test is the total hemolytic complement (CH50) assay, a test of the hemolyzing activity of complement that requires the presence of all components for a normal result. A low CH50 assay value suggests that one or more of the classic or terminal components is missing. Further testing of individual complement components establishes the specific defect (see Chapter 105, Infectious Complications of Complement Deficiencies). If the complement cascade is found to be intact, analysis of quantitative serum immunoglobulins and subclasses should be performed. In addition, imaging of the abdomen (e.g., ultrasound) to rule out hyposplenism should be done. If Mollaret meningitis is suspected, polymerase chain reaction of the CSF may be helpful in the diagnosis. In addition, serum-to-CSF ratio of antibodies for HSV should be done.15,65

TREATMENT For recurrent bacterial meningitis due to head trauma, congenital malformations, or complement deficiency, the choice of empiric

A

B Figure 44-6. Arachnoid cyst with sinus tract. A 5-year-old boy came to medical attention for draining sacral osteomyelitis. His parents had throughout his life repeatedly “popped” a pustule over his sacrum. Sagittal T2-weighted magnetic resonance image (A) and operative field (B) show a large arachnoid cyst occupying the entire bony central canal from L2 to S1. Note the tip of the spinal cord conus at L1–L2 level (large arrow) and enhancing soft tissue, bone, and bony canal in the sacral region (small arrow). (Courtesy of J.H. Brien.)

antibiotic therapy should be similar to that for single-episode meningitis. For patients with a known dermal sinus or history of meningitis due to staphylococcal or gram-negative bacteria, an antibiotic with a broader spectrum of activity is required. In rare situations when the bacteria is multidrug-resistant and do not respond to intravenous treatment, intraventricular administration of the antibiotic may be of help.66 The length of therapy is similar to that for

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Aseptic and Viral Meningitis

Figure 44-7. T1-weighted sagittal computed tomographic scan with nonionic contrast material. Scan in coronal plane reveals the defect in the bone (black arrow) and collection of contrast material with fluid level (white arrow) in the sphenoid sinus. (Courtesy of S.E. Byrd, MD, Children’s Memorial Hospital, Chicago.)

sporadic cases; there is no advantage to longer courses of therapy. Time to sterilization may be prolonged and should be documented in cases of gram-negative bacillary infections associated with dermoid/epidermoid cysts as squamous collections can act like foreign bodies. Meticulous examination and diagnostic evaluation (occasionally including exploratory surgery) to identify and perform primary repair of the site of CSF Àstula are most important. In the case of otorrhea, packing of the middle ear with fat is inadequate because it fails to close the Àstula permanently. Even packing with muscle or fascia alone seems to be insufÀcient; additional grafting of the vestibule is needed. For CSF rhinorrhea, extracranial surgery may be the preferred technique, because it produces less morbidity than an intracranial approach; the few disadvantages of this procedure are cranial nerve paralysis and postoperative sinusitis. In patients with Mollaret meningitis, anti-HSV therapy and suppression (e.g., acyclovir) may be helpful. Fresh frozen plasma has been given as replacement therapy to patients with complement deÀciency. Although this approach may be reasonable in the few patients who suffer from life-threatening disease, routine administration of plasma can stimulate the production of antibodies against the missing component. Additionally, the short half-life of complement proteins makes this approach impractical. Monthly immunoglobulin infusions for patients with immunoglobulin subclass deÀciency should be individualized according to the severity of the disease (i.e., recurrent bacterial versus aseptic meningitis).

PROPHYLAXIS CT of the temporal bone should be considered in children with idiopathic senorinural hearing loss to identify those who have innerear anomaly. Parents of children with proven inner-ear anomaly should be educated about the risk of recurrent meningitis from middleear infection, contact sports, and activities that may increase the inneror middle-ear pressure (e.g., diving, prolonged Valsalva maneuver). Suspected middle-ear infections should be treated promptly and aggressively and in patients with common cavity abnormalities, exploratory tympanotomy should be considered. If cochlear implantation is considered in patients with inner-ear malformation, the type of implant and the risk for meningitis should be considered.67 All

CHAPTER

45

305

efforts should be made to identify the source of the CSF leak and to seal it because if the leakage continues it increases the risk for postoperative meningitis.68 Although prophylactic antibiotics are often given to patients with recurrent meningitis, their efÀcacy in preventing further episodes is questionable. One review of the efÀcacy of prophylactic antibiotics in preventing meningitis after basilar skull fracture in children suggested that the practice had no beneÀt and might possibly do harm by increasing the risk of infection due to antibiotic-resistant organisms.69 Prophylactic antibiotics have also been recommended for patients with complement or signiÀcant immunoglobulin subclass deÀciency. In some cases, such treatment markedly reduced the incidence of infection; yet the clinical failure of penicillin prophylaxis against meningococcal disease suggests that long-term prophylaxis may be of limited usefulness in this population.35 The advisability of antibiotic prophylaxis in patients with recurrent meningitis should be made on a case-by-case basis; Àrm guidelines are not available. Limited data suggest that chronic oral acyclovir (for year or more) may prevent recurrency of Mollaret meningitis.70 The quadrivalent A, C, Y, W-135 meningococcal vaccine is recommended for patients who have terminal complement component deÀciencies and for those who have anatomic or functional asplenia (i.e., groups at high risk for recurrent meningitis), but the vaccine’s clinical efÀcacy has not been documented in such patients.71 Individuals with deÀciencies of the late complement components can make speciÀc antibodies against meningococci, and vaccination may increase protection in seronegative individuals. Unfortunately, recurrent infection has been reported in vaccinated patients despite the presence of speciÀc antibodies.35,72 Meningococcal vaccination may be advantageous in individuals 2 years or older with previous meningococcal infections who do not have deÀciency of the terminal complement pathway or in those with properdin deÀciency. Although the heptavalent pneumococcal conjugate vaccine is available, it is not sufÀcient for older children and adults in high-risk groups (e.g., patients with CSF leaks), because the proportion of invasive pneumococcal isolates covered by the heptavalent vaccine (50% to 60%) is much lower than that covered by the 23-valent polysaccharide vaccine (PPV23) (80% to 90%).73 Immunization with PPV23 is recommended for older children and adults with recurrent pneumococcal meningitis. Children younger than 2 years should receive the heptavalent vaccine, followed by revaccination with PPV23 when they are 2 years old, as recommended for individuals at high risk of pneumococcal disease.74,75 Failure of such vaccination to prevent recurrent episodes of meningitis has been reported.76 It is possible that: (1) the vaccine failed because it lacked the responsible serotypes; or (2) systemic antibodies are ineffective because the organisms invade the meninges directly from the nasopharynx. Children should complete age-appropriate pneumococcal vaccines before implantation of a cochlear device.74,75

CHAPTER

45

Aseptic and Viral Meningitis Mark H. Sawyer and Harley A. Rotbart

The term aseptic meningitis refers to a clinical syndrome of meningeal inflammation in which common bacterial or fungal agents cannot be identiÀed in the cerebrospinal fluid (CSF). Implicit in the term aseptic meningitis is a benign clinical course and an absence of signs of parenchymal brain involvement (encephalitis) or spinal cord inflammation (myelitis). Certain pathogens more commonly cause “pure”

SECTION

F

Central Nervous System Infections

meningitis or “pure” encephalitis, whereas other pathogens are more likely to result in less discrete manifestations of central nervous system (CNS) infection and are described as meningoencephalitis or encephalomyelitis. Some aseptic meningitis is not due to infection, being caused by drugs, lead intoxication, or systemic inflammatory conditions. This chapter focuses on infectious agents that most often cause aseptic meningitis syndrome.

ETIOLOGY AND EPIDEMIOLOGY Aseptic meningitis is not a reportable disease in the United States, and the number of annual cases is unknown but is estimated to be at least 75 000. The vast majority of cases are due to viral infections; however, nonviral pathogens, including certain bacteria (which are not readily seen after staining or do not grow in standard culture systems, such as Borrelia, Treponema, and Rickettsia), mycoplasma, and fungi, can cause identical clinical illnesses (Box 45-1). The Àrst viruses recognized to cause meningitis were mumps, lymphocytic choriomeningitis virus (LCMV), and poliovirus; these are uncommon causes in most developed countries today. Early case series of aseptic meningitis reported identiÀcation of a speciÀc cause in only 25% of cases.1,2 With the discovery of the group B coxsackieviruses in 19483 and the advent of tissue culture in 1949,4 enteroviruses (EVs) quickly emerged as the leading recognizable cause of aseptic meningitis. EVs comprise more than 60 distinct serotypes within the family Picornaviridae (pico, “small”; rna, ribonucleic acid). EV subgroups include the polioviruses, coxsackieviruses A and B, echoviruses, and the newer numbered EVs. Usually only a few serotypes cause disease in any one community during one season and shifts in serotype predominance can affect the overall incidence of

BOX 45-1. Major Causes of Aseptic Meningitis Syndrome COMMON Enteroviruses (echoviruses, coxsackieviruses, polioviruses) Arboviruses (eastern equine encephalitis, western equine encephalitis, St. Louis encephalitis, Colorado tick fever, California encephalitis, West Nile) Herpes simplex 2 Borrelia burgdorferi (Lyme disease) Partially treated bacterial meningitis UNCOMMON Mumps virus Human immunodeÀciency virus Mycobacterium tuberculosis Parameningeal bacterial infection Fungi (Cryptococcus, Coccidioides, Histoplasma, Blastomyces) RARE Respiratory viruses (adenovirus, influenza, parainfluenza) Lymphocytic choriomeningitis virus Other herpesviruses (herpes simplex 1, human herpesvirus 6, Epstein–Barr, varicella-zoster, cytomegalovirus) Measles virus Miscellaneous viruses (parvovirus B19, rotavirus) Bartonella sp. (cat-scratch disease) Spirochetes Leptospira sp. Brucella sp. Parasites (e.g., Taenia solium (cysticercosis), Toxoplasma gondii, Trichinella spiralis (trichinosis)) Mycoplasma pneumonia Rickettsia NONINFECTIOUS CAUSES Drugs (e.g., nonsteroidal anti-inflammatory agents, agents instilled into cerebrospinal fluid) Biologic products (e.g., immune globulin) Systemically immunologically mediated diseases (e.g., rheumatologic diseases, Behçet disease, Mollaret meningitis) Neoplastic diseases

viral meningitis from year to year.5 Other than this impact on epidemiology the speciÀc serotype causing meningitis is generally of little importance clinically. Once tissue culture techniques were established, case series of aseptic meningitis reported identiÀcation of a speciÀc viral pathogen in as many as 55% to 70% of cases.6–10 Application of the polymerase chain reaction (PCR) test to the diagnosis of aseptic meningitis during the last 15 years has resulted in increased identiÀcation of EVs. In studies of aseptic meningitis in children since the advent of PCR testing, more than 70% of cases for which an etiology is identiÀed are attributable to EV.11–16 For adults the proportion of aseptic meningitis cases due to EVs is somewhat lower, in part due to the increased prevalence of herpes simplex virus type 2 (HSV-2) aseptic meningitis.17 The overall seasonal epidemiology of aseptic meningitis mirrors that of the EVs (Figure 45-1). A summer-to-fall peak in incidence of EV meningitis is seen annually. Arbovirus is a taxonomically defunct term that still has useful practical meaning, encompassing viral pathogens transmitted by arthropod vectors. Hundreds of such viruses have been identiÀed worldwide, each with distinct seasonal and geographic characteristics determined by the biologic patterns of the particular vector and the animal reservoirs. Until recently in the United States, St. Louis encephalitis virus was the most common arbovirus causing CNS infections. Although encephalitis is the most clinically signiÀcant and commonly recognized neurologic manifestation of infection,18 certain viruses also cause aseptic meningitis and meningoencephalitis as part of their disease spectrum. Seasonality and geographic distribution are determined by the life cycle and habitat of the vectors; hence, most infections in this country occur during summer and fall – peak mosquito and tick seasons. Except during large epidemics, far fewer cases of summer and fall meningitis are caused by arboviruses than by EVs. Beginning in 1999 in the United States, West Nile virus (WNV) has caused epidemic viral CNS disease, including meningitis, and has become the most frequently identiÀed arbovirus infection in recent years.19 Fortunately, < 1% of individuals infected with WNV develop CNS disease and the majority are asymptomatic. Neurologic disease is more frequent in adults than in children. Encephalititis and acute flaccid paralysis are more common manifestations of CNS WNV infection than is aseptic meningitis but isolated meningitis does occurs and is difÀcult to differentiate clinically from other viral meningitis illnesses. The mumps virus is a member of the family Paramyxoviridae; aseptic meningitis is the most common neurologic complication of mumps. CSF pleocytosis occurs in more than 50% of patients with Reported cases per 100,000 population

306

1.2 1.1 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 1986

1987

1988

1989

1990

1991

1992

1993 1994

Year (month) Figure 45–1. Seasonal incidence of aseptic meningitis in the United States from 1986 to 1993, as reported to the Centers for Disease Control and Prevention. The summer predominance reflects the role of enteroviruses as the leading cause of the aseptic meningitis syndrome. (Redrawn from Centers for Disease Control and Prevention. Summary of notifiable diseases, United States, 1993, MMWR 1994;42:69.)

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Aseptic and Viral Meningitis

mumps parotitis,20 but most do not have symptoms of meningitis. CNS symptoms are reported in up to 30% of all cases of mumps parotitis within 4 to 10 days of illness but may precede parotitis by as many as 7 days; one-half or more cases of mumps meningitis are not associated with parotitis at all.21,22 The mumps outbreak during 2006 in the Midwest United States led to more than 2500 reported mumps cases, primarily in adolescents and young adults.23 This outbreak included 11 cases of mumps meningitis. LCMV was one of the earliest and most common viruses associated with aseptic meningitis in humans. The virus is transmitted by rodents and historically affected pet owners and laboratory workers exposed to rodents.24 Due to undefined epidemiologic factors or perhaps because it is infrequently tested for, LCMV is now rarely identified in human CNS infections.25 Neurologic complications are well known in infections caused by HSV-1 and 2, varicella-zoster virus, Epstein–Barr virus, cytomegalovirus, and, most recently, human herpesvirus 6 (HHV-6).26 Although aseptic meningitis has been reported in conjunction with all of these pathogens, only meningitis associated with HSV, especially HSV-2, appears to be of numerical significance.27 The clinical course and outcome of aseptic meningitis (versus other neurologic disease) resulting from the herpes family viruses are uniformly good, and patients with this disease are indistinguishable clinically from those with aseptic meningitis from other causes. Encephalitis and other neurologic sequelae of herpesvirus infections may be much more severe. A variety of other infectious agents have been identified in patients with an aseptic meningitis syndrome. Measles virus and adenovirus are well known as causes of encephalitis and meningoencephalitis, particularly in immunocompromised patients. Occasional cases of aseptic meningitis due to adenovirus infection in normal patients have been noted,9 and measles infection may be associated with pleocytosis in as many as 30% of uncomplicated cases, usually without signs or symptoms of meningitis.28 Respiratory viruses are rarely described in association with aseptic meningitis. Several cases of parainfluenza 3 meningitis have been reported29,30 and a variety of neurologic syndromes, including aseptic meningitis, can accompany influenza A and B infections.31 A case of aseptic meningitis concomitant with rotavirus gastroenteritis has been noted.32 Mycoplasma pneumoniae, Chlamydia pneumoniae, M. hominis, and Ureaplasma urealyticum infections have been associated with aseptic meningitis.33–37 In individual cases, Borrelia burdorferi (Lyme) meningitis and viral meningitis are not clinically distinguishable. Endemicity, time of year, and a history of tick exposure or erythema migrans are helpful. Children with Lyme versus viral meningitis are more likely to be afebrile, and have subacute presentation and cranial neuropathy.38 CSF findings are similar to viral meningitis, with mild mononuclear pleocytosis (see Chapter 186, Other Borrelia Species and Spirilum minus, for diagnosis and therapy). Using PCR, cases of parvovirus B19-associated aseptic meningitis have been detected.39,40 Meningitis has been reported in HIV-infected patients, without other HIV-associated CNS manifestations or evidence of secondary pathogens.41,42

PATHOGENESIS AND PATHOLOGY The pathogenesis of viral meningitis, regardless of the virus, is similar. It begins with introduction of the pathogen into the host, followed by viremia, with dissemination to the CNS and other organs. EVs are acquired by fecal–oral contamination and, less commonly, by respiratory droplet.43,44 While some replication occurs in the nasopharynx and spreads to the upper respiratory tract lymphatic system, most of the viral inoculum is swallowed and traverses the stomach en route to the site of primary infection in the lower gastrointestinal tract. Arbovirus infections begin with subcutaneous inoculation by a mosquito or tick vector, followed by local tissue and lymph node replication. Mumps and other respiratory viruses are acquired by the respiratory droplet route, with primary replication in the upper respiratory epithelium. Local viral invasion of the parotid gland via

CHAPTER

45

307

the duct epithelium accounts for parotitis, the most prominent manifestation of infection. Parotitis is not, however, required for – and may be incidental to – dissemination of the virus.45 LCMV is transmitted by rodents (e.g., hamsters, rats, mice). Laboratory workers, pet owners, and individuals living under impoverished and unhygienic circumstances are at greatest risk. Ingestion of food contaminated with animal urine or exposure of open wounds to dirt are presumed routes of transmission. Data from pathology reports of aseptic and viral meningitis are scarce. A report of a child who died of coxsackievirus B5 myocarditis with concomitant meningitis describes inflammation of the choroid plexus of the lateral and fourth ventricles, fibrosis of the vascular walls with focal destruction of the ependymal lining, and fibrosis of basal leptomeninges.46 Parenchymal findings were limited to moderate, symmetric dilatation of the ventricles and an increase in number and size of subependymal astrocytes. The inflammatory reaction at the choroid plexus supports the concept of viremic spread to the CNS. A second patient, an adolescent for whom there was a similar constellation of findings, died of systemic coxsackievirus B3 infection.47 The dura was grossly distended, with swelling of the pia, arachnoid, and brain parenchyma. On microscopy, round cell infiltrates were noted in the meninges overlying the cerebellum; the brain parenchyma was congested, with increased numbers of oligodendrocytes. Lymphocytic infiltration was most prominent around blood vessels in the cerebral white matter and in the basal ganglia, again suggesting viremic access to the CNS; focal areas of necrosis and hemorrhage were also seen. The occasional fatalities from mumps meningitis demonstrate pathologic findings of demyelination near blood vessels, which suggests an autoimmune process, and evidence of acute parenchymal involvement.48

CLINICAL MANIFESTATIONS The clinical disease observed in patients with EV meningitis varies with the host’s age and immune status. Neonates are at risk for severe systemic illness, commonly including meningitis or meningoencephalitis.49 Aseptic meningitis was detected in 48 of 77 (62%) of infants less than 3 months of age who were hospitalized with group B coxsackievirus infection.50 Echoviruses identified in infants less than 2 weeks of age were associated with meningitis or meningoencephalitis in 27% of cases.51 In a prospective study of neonates (2 weeks of age) with proven EV infection, 75% had clinical or laboratory evidence of meningitis.52 The infected neonate appears to be at greatest risk for severe morbidity and mortality when signs and symptoms develop in the first days of life, suggesting possible transplacental acquisition.50–52 Even in the youngest patients, fever is common and is frequently accompanied by nonspecific signs, including vomiting, anorexia, rash, and upper respiratory tract findings.53 Signs of meningeal inflammation, including nuchal rigidity and a bulging anterior fontanel, are variably present. As the neonatal disease progresses, major systemic manifestations, such as hepatic necrosis, myocarditis, and necrotizing enterocolitis, can develop. Disseminated intravascular coagulation and other findings of sepsis result in illness that can be indistinguishable from overwhelming bacterial infection. The CNS disease can progress to a more encephalitic picture with seizures and focal neurologic findings suggestive of HSV infection. The incidence of morbidity and mortality due to perinatal EV infections is not precisely known, but may be as high as 74% and 10% respectively. When death occurs, it is seldom the result of CNS involvement, but rather due to hepatic failure (echoviruses) or myocarditis (coxsackieviruses). EV meningitis outside the neonatal period rarely has a poor outcome, but the short-term morbidity may be substantial and the duration of disease prolonged.50,51 Onset is usually sudden, with fever in 76% to 100% of patients.54–56 Fever can be biphasic, appearing first with nonspecific constitutional symptoms, followed by resolution and reappearance with the onset of meningeal signs. Nuchal rigidity is present in more than one-half of patients, particularly in those > 1 year

308

SECTION

F

Central Nervous System Infections

of age.55,56 Headache is almost always present in adults and children old enough to report it,57 and photophobia is common. Nonspecific and constitutional signs and symptoms of viral infection include vomiting, anorexia, rash, diarrhea, cough, and upper respiratory tract abnormalities (particularly pharyngitis), diarrhea, and myalgia.55,56 Aseptic meningitis characterized by certain EV serotypes produces particular clinical stigmata, e.g., hand, foot, and mouth syndrome, which frequently occurs with EV 71 meningitis,58 and nonspecific rashes with echovirus 9 meningitis,59 although both findings occur with numerous other serotypes. Neurologic abnormalities are rare. Typical febrile seizures can complicate aseptic meningitis in children, without implicating parenchymal brain involvement. The duration of EV meningitis illness is often > 1 week, particularly in older children.54,60 Adults often have symptoms that persist for 2 weeks or longer.57 As a result, missed school and work contribute to the substantial economic impact of EV meningitis.54,57,60 The short-term prognosis for young children with EV meningitis appears to be good; however, there is controversy regarding possible sequelae. In one large and well-controlled study, no differences between patients and controls were reported in any of the neurodevelopmental skills studied.61 Unlike other viruses, which are largely contained by cellular immune mechanisms, the EVs are eliminated by antibody-mediated mechanisms. Individuals with agammaglobulinemia or hypogammaglobulinemia who are infected with EV can develop chronic meningitis or meningoencephalitis that persists and progresses over many years, often with a fatal outcome.62 The clinical manifestations of mumps aseptic meningitis are nonspecific and differ little from those of EV meningitis. However, because the average age of the patients with mumps is older than that of patients with EV meningitis, certain signs and symptoms, such as headache, nausea, vomiting, and meningismus, are more often reported. Fever typically persists for at least 3 days, but occasionally persists for > 1 week.21 Bradycardia, drowsiness, lethargy, and anemia are reported. Arboviral meningitis is clinically indistinct from EV disease, except when an encephalitic component is present. Colorado tick fever (CTF) is caused by an orbivirus introduced by a tick bite. A biphasic illness is characteristic, but actually observed in only half of patients,63,64 and consists of initial sudden onset of high fever and headache with flulike constitutional symptoms. Hepatosplenomegaly and gastrointestinal symptoms may occur. Stiff neck and other meningeal signs are reported in up to 18% cases of confirmed CTF.65 WNV infections are asymptomatic in 80% of cases and clinically evident neuroinvasive disease occurs in < 1%.19 However, among patients with West Nile fever, a presentation lacking clear CNS involvement, neck pain or stiffness is common, suggesting possible mild meningitis.66 Most recognized cases of neuroinvasive WNV manifest as encephalitis or acute flaccid paralysis rather than aseptic meningitis. The most common syndrome of HSV-associated aseptic meningitis occurs concomitantly with or shortly after primary HSV-2 genital infection.67 More than 30% of women and 11% of men in one study of primary HSV-2 infections developed an aseptic meningitis syndrome, including CSF pleocytosis.68 Primary HSV-1 genital infection is less often associated with aseptic meningitis, and nonprimary genital infection with either HSV serotype rarely results in meningitis. Meningitis has also been reported with HSV-1 and HSV-2 in the absence of recent genital lesions27,69; the high frequency of asymptomatic shedding of genital HSV may explain the apparent de novo HSV meningitis.

DIFFERENTIAL DIAGNOSIS Aseptic meningitis syndrome is defined by the absence of a readily identifiable cause. Nevertheless, certain bacterial or fungal infections can occur occultly with classic features of aseptic meningitis. These include spirochetes (leptospirosis, Lyme disease, syphilis, and

borreliosis), mycobacteria, and Brucella. Partially treated bacterial meningitis and parameningeal infection also manifest similarly to viral meningitis and must always be considered, because specific therapy is required. Fungal meningitis is recognized increasingly among immunocompromised hosts; however fungi can also infect the CNS of normal individuals and must be considered in any indolent, atypical aseptic meningitis syndrome. A variety of neurologic manifestations have been reported in association with Mycoplasma pneumoniae infections,33,34 among which aseptic meningitis and encephalitis are the most common. Ureaplasma urealyticum and Mycoplasma hominis have been associated with cases of neonatal meningitis, usually in preterm infants.35,37 Patients who have immune-mediated diseases or who are undergoing immunomodulating drug therapy can develop an aseptic meningitis syndrome, often characterized by recurrent episodes. Meningitis has been described as the initial manifestation of systemic lupus erythematosus (SLE) in several patients but is more often part of the ongoing evolution of disease. Anti-inflammatory agents, particularly nonsteroidal anti-inflammatory agents (NSAIDs), can cause aseptic meningitis in patients with autoimmune or autoinflammatory conditions such as SLE, Sjögren syndrome, and juvenile idiopathic arthritis, and occasionally in patients without underlying disease.70 The acute meningeal syndrome can occur after a single dose of prolonged therapy and frequently recurs with re-exposure. In a review of 73 cases in adults, median CSF white cell count was 280 cells/mm3 (range 9 to 5000), neutrophils predominantly in 72% of cases, median protein level was 132 mg/dL (range 32 to 857), and glucose level was depressed in approximately 10%.70 Other immune-modulating drugs have also been associated with aseptic meningitis in patients with underlying diseases other than connective tissue diseases. Cytosine arabinoside, immune globulin intravenous (IGIV), and the murine monoclonal antibody OKT3, used to reduce graft rejection, are examples. Antibiotics, particularly sulfacontaining compounds, have also been associated with aseptic meningitis, often, but not always, in patients with underlying connective tissue disease.71 Carbamazepine has caused aseptic meningitis. Selfresolving aseptic meningitis has been described in as many as 26% of patients with Kawasaki disease.72 Mollaret meningitis is a rare disease characterized by recurrent, benign episodes of aseptic meningitis with symptomfree intervals between episodes. Using PCR, HSV-2 has been shown to cause typical cases of Mollaret meningitis without signs or symptoms of genital infection.73,74 Occasional cases of Mollaret meningitis associated with HSV-1 and Epstein–Barr virus have been reported.75,76 Each episode is typical of viral meningitis, except that in certain patients, cells resembling endothelial cells are present in the CSF. In patients with proven HSV-associated Mollaret meningitis, HSV antiviral therapy may be useful prophylactically or therapeutically, but this has not been formally studied.

LABORATORY FINDINGS AND DIAGNOSIS Clues to the diagnosis of EV meningitis can be obtained from CSF laboratory findings. Pleocytosis is almost always present, although EVs have been isolated from acellular CSF of patients, usually young infants, who were evaluated for fever and clinical suspicion of meningitis.77 White blood cell count of CSF can reach several thousand78 but a count of 100 to 1000 cells is typical.79,80 Polymorphonuclear cells can predominate early in meningeal infection, usually changing to lymphocyte predominance over the first 8 to 48 hours.81,82 Hypoglycorrhachia and elevated levels of CSF protein, if they occur, are usually mild,55,56 but extreme degrees of both have been reported.83 Wide variations in CSF findings can occur even during an epidemic of a single serotype.84,85 Specific virologic diagnosis of EV meningitis depends upon the detection of virus from CSF, either in tissue culture or by PCR.86

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Aseptic and Viral Meningitis

Although a nonpolio EV isolated from the throat or rectum of a patient with aseptic meningitis is suggestive of the diagnosis, the mean shedding period following infection from these sites is 1 week and several weeks, respectively.44 Hence, shedding from past infection cannot be excluded unless the virus is detected in nonpermissive sites, specifically the CSF or blood.87 The sensitivity of tissue culture for EVs is only 65% to 75%, because of a combination of factors, including the inability to grow many coxsackievirus A serotypes and the neutralization or inhibition by CSF of serotypes that can usually be cultured.86 The use of multiple cell lines improves the recovery of EVs88,89 but may not be cost-effective or efficient enough for most clinical laboratories. Furthermore, the titer of EVs in the CSF of patients with aseptic meningitis can be as low as 10 to 1000 infectious particles per mL of CSF,90 resulting in slower growth than observed with specimens of throat or rectal origin. Investigators report 3.7 to 8.2 days as the mean time for EV isolation from CSF.86 The absence of a widely shared antigen has hampered the development of immunoassays for the EVs.91 PCR is the most widely used alternative to viral culture for EV infection diagnosis.92,93 This technique is typically directed at genomic RNA in the highly conserved 5„ noncoding region designed for reverse transcription PCR (RT-PCR). EV RT-PCR has been tested in clinical settings by numerous investigators and is consistently more sensitive than culture and is almost 100% specific.92–96 In many laboratories, an accurate diagnosis of EV infection is available in < 1 day, offering the potential for positively affecting the quality and cost of patient management. Diagnosis of all arboviruses is theoretically possible using a single RT-PCR protocol and multiple primer pairs97; however, primers have not yet been designed for many of the pathogenic arboviruses. Each of the arbovirus groups has been detected successfully by an immunoglobulin M (IgM) antibody capture enzyme immunoassay (EIA). Paired serology is also an effective retrospective diagnostic method. For CTF, laboratory diagnosis is best made by inoculation of blood (clot or unclotted cellular fraction) into suckling mice.98 Virus may also be detected in peripheral blood smears by indirect immunofluorescence.99 For WNV virus infection, serum or CSF-specific IgM is usually detectable by the time the patient manifests with neurologic symptoms.19 WNV can also be detected by molecular amplification methods and virus isolation performed in reference laboratories. Before tissue culture was available, a reliable serologic test for mumps virus infection accounted for the pre-eminent status of mumps virus as a cause of cases of meningitis and meningoencephalitis.1,2 Complement fixation and hemagglutination inhibition,100,101 hemolysis-in-gel assays,102 paired acute and convalescent serologic studies, specific IgM and IgG assays on a single specimen, and comparative CSF and serum antibody measurements all have roles in the diagnosis of mumps meningitis. Mumps virus from CSF samples can be grown in tissue culture for at least a week after onset of disease, but the sensitivity of the technique is highly variable. Molecular amplification techniques, particularly PCR, have improved our ability to make a specific diagnosis in cases of aseptic meningitis of many types.103 In the case of EV meningitis, the use of PCR testing is associated with reduced hospital stays and reduced use of alternative diagnostic testing, thus demonstrating the value of obtaining a rapid diagnosis using molecular methods.104 Continued standardization of these tests and expanded availability should increase their use in the future.

TREATMENT As with other viral pathogens, there are several steps in the replication cycle of the EVs that are potential targets for antiviral therapy. Cell susceptibility and viral attachment, uncoating, RNA replication,

CHAPTER

45

309

and protein synthesis have been studied as targets for anti-EV compounds.105 Unfortunately, although compounds targeting each of these steps have been developed and studied in clinical trials, there are no established effective therapies. Immune serum globulin (ISG) has been used prophylactically and therapeutically against the EVs in two clinical settings: the neonate and the immunocompromised host. Anecdotal reports of clinical success with maternal serum or plasma, or commercial ISG, against a variety of EV serotypes that cause severe neonatal EV disease have been reported. However, other reports describe progressive disease and death despite such therapy.49 A small, blinded, randomized, controlled study did not demonstrate clinical benefit but showed reduction in viral titer in neonates receiving IGIV: this specific antibody preparation was subsequently shown to contain high antibody titers to the infecting serotype.106 Before the availability of IGIV, mixed results were reported with intramuscular or intrathecal administration of ISG in patients with antibody deficiency. Some patients appeared to benefit from supplemental immunoglobulin; others died despite therapy.62 Since patients with known antibody deficiency have begun receiving maintenance supplementation with IGIV, the incidence of chronic, progressive EV meningoencephalitis has decreased and the clinical profile of patients who have such infections has been modified.107 Pleconaril is an antiviral compound that inhibits EV replication by inhibiting viral uncoating and blocking viral attachment to host cell receptors. Although this compound has promising in vitro properties and has undergone clinical trials for EV meningitis in children and adults, there are no established indications for its use, and currently pleconaril for systemic use is not available.108 There are no specific therapies available for other viral causes of aseptic meningitis. Although numerous specific antiherpes therapies are available, none have been studied in patients with HSV meningitis without concomitant encephalitis. Supportive care for the patient with viral or aseptic meningitis includes pain management, often requiring narcotics.57 Attention to fluid balance is necessary to provide hydration in dehydrated patients but to avoid brain edema if inappropriate secretion of antidiuretic hormone occurs. Serum electrolyte concentrations and, on occasion, urine and serum osmolality may require monitoring. Seizures may result from fever alone, or may reflect direct viral or indirect inflammatory damage to brain parenchyma (in which case encephalitis is the more apt term). Anticonvulsants may be required to manage this complication but a need for prolonged use is uncommon.

PREVENTION The use of poliovirus vaccines does not influence the incidence or severity of infections caused by the nonpolio EVs, for which no vaccines exist. Similarly, there are no vaccines currently available to prevent arbovirus, LCMV, or herpesvirus meningitis. The widespread use of the mumps attenuated live-virus vaccine in the United States has resulted in a drop in incidence of this once common infection. The 2006 outbreak of mumps in the United States illustrated the potential for widespread disease despite a highly immunized population. This has led to recent changes in mumps immunization recommendations that include a second dose of measles, mumps, and rubella vaccine for healthcare workers and students in post secondary education.109 Cases of vaccine-associated mumps meningitis have been reported but are now rare in the United States because of improvements in the strains of mumps virus used in United States vaccine. When it occurs, vaccine-associated mumps meningitis occurs 15 to 35 days after vaccination110,111 and has no sequelae.112

310

SECTION

CHAPTER

F

Central Nervous System Infections

TABLE 46-1. Differential Diagnosis of Encephalitis in the Neonate

46

Infection/Disorder

Frequency

ACUTE INFECTION

Encephalitis, Meningoencephalitis, Acute Disseminated Encephalomyelitis, and Acute Necrotizing Encephalopathy Rodney E. Willoughby, Jr. and Sarah S. Long

Inflammation of the central nervous system (CNS) can involve the brain (encephalitis), spinal cord (myelitis), spinal roots (radiculitis), or a combination of sites (meningoencephalitis, encephalomyelitis, or myeloradiculitis).1 The diagnosis of encephalitis is usually inferred from clinical findings rather than proved by histology. Encephalopathy (clinical findings without inflammation) mimics encephalitis. The term encephalitis is used in this chapter to denote predominant inflammation of the brain, with or without involvement of meninges or spinal cord. Myelitis and radiculitis are addressed in Chapter 47, Cerebellar Ataxia, Transverse Myelitis and Myelopathy, Guillain–Barré Syndrome, Neuritis, and Neuropathy.

Herpes simplex virus Enterovirusa Adenovirus Group B streptococcusb Listeria monocytogenesb Citrobacter spp.b

++ +++ + + + +

CONGENITAL INFECTIONS

Cytomegalovirus Lymphocytic choriomeningitis virus Toxoplasmosis Rubella virus Syphilis

++ ++ + + +

METABOLIC DISORDERS

+c

Propionic acidemia Methylmalonic acidemia Urea cycle defects Maple syrup urine disease PRIMARY CENTRAL NERVOUS SYSTEM DISORDER

Nonconvulsive status epilepticus Hemorrhage Neonatal encephalopathy

+ + ++

+++, most frequent; ++, frequent; +, occasional. a Includes aseptic meningitis. b Cerebritis. c Combined incidence.

ENCEPHALITIS/MENINGOENCEPHALITIS Etiologic Agents Infectious causes of acute encephalitis are myriad but a specific etiology is identified in fewer than one-third of cases,2 even when extensive labratory testing is performed.2,2a The cause is usually inferred from isolation of pathogens at anatomic sites other than the CNS, through the interpretation of antibody titers in acute and convalescent serum samples, detection of specific immunoglobulin M (IgM) antibody in cerebrospinal fluid (CSF) or serum or polymerase chain reaction (PCR) testing of CSF. A single positive antibody titer for a pathogen is acceptable evidence of causation if infection is rare or highly fatal, such as due to eastern equine encephalitis virus (EEEV) or rabies virus. Encephalitis in the neonate is often part of a systemic, multiorgan disease, sometimes difficult to distinguish from manifestations of an inborn error of metabolism. This difficulty may explain, in part, the lower rate of confirmed cause for encephalitis in the first year of life (Table 46-1).3 Encephalitis in later infancy and childhood is caused by a variety of infectious agents and mechanisms. Viruses cause most proven encephalitis. Enteroviruses are the most common cause of CNS infection in children; predominant encephalitis, without meningitis, is an unusual manifestation. Common agents are listed in Table 46-2. Rare causes in the United States include human herpesvirus 6 (HHV-6), mumps virus, respiratory viruses such as parainfluenza virus and respiratory syncytial virus, hepatitis viruses A and B, human immunodeficiency virus-1 (HIV-1: acute seroconversion syndrome), and rabies virus. Few pyogenic bacteria cause encephalitis without overt meningitis. Syphilis, leptospirosis, brucellosis, tuberculosis, and listeriosis are rarely associated with encephalitis. Occasionally, encephalitis is a presenting manifestation of cryptococcosis, histoplasmosis, blastomycosis, or coccidioidomycosis. Inborn errors of metabolism and systemic diseases can be confused with infectious encephalitis in the older child (Box 46-1).4,5

Epidemiology Encephalitis is a rare disease, occurring in 0.3 to 0.5 per 100,000 individuals in the United States. Disease occurs predominantly in

TABLE 46-2. Causes of Infectious Encephalitis Beyond the Neonatal Period Agent

Frequency

VIRUS

Enterovirus Arthropod-borne virusesa Herpes simplex virus Epstein–Barr virus Adenovirus Human immunodeficiency virus 1

+++ + ++ + + +

BACTERIA/OTHER

Borrelia burgdorferi Bartonella henselae Rickettsia rickettsii Mycoplasma pneumoniae

+ + + +

+++, most frequent; ++ frequent; +, occasional. a Especially equine encephalitis viruses and West Nile virus.

children, elderly, immunocompromised hosts, or in individuals with occupational or residential exposure to vectors of arthropod-borne viruses. Incidence of encephalitis is highest in the first year of life (17/100,000 births) and declines with age.3 Postinfectious encephalitis rarely occurs in children younger than 1 year. Rates of epidemic encephalitis fell over several decades as a result of both improvements in living conditions and vector control and the advent of vaccines against many of the childhood exanthematous diseases. Mortality from encephalitis has not declined proportionately, however, because the agents that cause most of the severe cases of encephalitis, such as herpes simplex virus (HSV), remain endemic. HSV is the leading cause of severe encephalitis at all ages. Neonatal HSV encephalitis occurs once in 2600 live births, a frequency approximating that of group B streptococcal meningitis. Neonatal encephalitis due to enterovirus or adenovirus usually occurs as part of disseminated infection; isolated encephalitis is rare. Human

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Encephalitis, Meningoencephalitis, Acute Disseminated Encephalomyelitis, and Acute Necrotizing Encephalopathy

CHAPTER

46

311

parechaviruses (HPeVs) were previously known as echo22 and echo23. HPeV3 is especially associated with sepsis syndrome in young infants, with neurologic manifestations including lethargy, convulsions and paralysis.5a Congenital infection with members of the herpesvirus family (cytomegalovirus, varicella-zoster virus (VZV), HSV, and HHV-6), rubella virus, lymphocytic choriomeningitis virus,6 or Toxoplasma gondii can cause CNS infection with structural brain damage and neurologic symptoms present at birth. Organic acidemia or defects in the urea cycle can cause global CNS dysfunction in the neonatal period. Individually rare, such disorders, when combined, are as common as infectious encephalitis in the young infant.2 Encephalitis in the older infant or child is usually mild and selflimited and occurs as an extension of aseptic meningitis (see Chapter 45, Aseptic and Viral Meningitis) or as a rare complication of other common bacterial or viral infections. Enteroviruses and many arthropod-borne viruses (arboviruses), including West Nile virus (WNV), are the most common causes of mild encephalitis and have expanding geographic ranges (see Table 46-2).7,8 EEEV causes severe encephalitis, with fatality rates estimated at 35% to 75%, and frequent sequelae in survivors.9 Adenoviruses and Epstein–Barr virus (EBV) can cause encephalitis sporadically, complicating EBV mononucleosis in approximately 1 per 1000 cases. Encephalopathy, and occasionally encephalitis, associated with Bartonella henselae (cat-scratch) disease is likely more common than is reported. Encephalitis is an epidemic disease of summer whereas postinfectious encephalitis clusters in the wintertime in countries with low rates of immunization.9,10 Arbovirus encephalitis varies geographically and seasonally because of the life cycle of enzootic and bridge vectors and amplifying nonhuman hosts (frequently wild birds). In North America, arbovirus encephalitis occurs in the late summer and fall and is frequently heralded by encephalitis in horses, crows, and jays. Except for WNV, California serogroup viruses have been most prevalent in the last decade (Figure 46-1).11 St. Louis encephalitis viruses are distributed through most of the United States and can cause major epidemics that peak later than other arboviruses.

BOX 46-1. Conditions Mimicking Encephalitis TOXIC ENCEPHALOPATHY Bordetella pertussis Shigella spp. Campylobacter jejuni Salmonella spp. Bartonella henselae Influenza-associated acute necrotizing encephalopathy Reye syndrome Acute toxic ingestion Lead intoxication Hyperpyrexic shock INBORN ERRORS OF METABOLISM Ornithine transcarbamylase deficiency, heterozygote Glutaric acidemia type 1 MCAD deficiency MELAS syndrome Leber optic neuropathy Acute intermittent porphyria Adrenoleukodystrophy CNS VASCULITIS Systemic lupus erythematosus Periarteritis nodosa Lymphogranulomatous angiitis TUMOR Brainstem glioma OTHER CNS CONDITIONS Intracranial hemorrhage Intracranial thrombosis Pseudotumor cerebri Acute confusional migraine CNS, central nervous system; MCAD, medium-chain acyl coenzyme A dehydrogenase; MELAS, mitochondrial encephalopathy with lactic acidosis and stroke.

200 180 160

California serogroup Eastern equine St. Louis

Number

140 120 100 80 60 40 20 0 1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

Year Figure 46-1. Arboviral encephalitis/meningitis. Number of reported cases by year: United States, 1995 to 2004. Data from the National Center for Infectious Diseases (ArboNet Surveillance). Arboviral diseases are seasonal, occurring during the summer and fall, with incidence peaking in the later summer. The most common arboviruses affecting humans in the United States are West Nile virus (WNV), La Crosse virus (LACV), Eastern equine encephalitis virus (EEEV), and St. Louis encephalitis virus (SLEV). California serogroup viruses (mainly LACV in the eastern United States) causes encephalitis, especially in children. In 2004, cases were reported in 13 states (Florida, Georgia, Illinois, Indiana, Iowa, Louisiana, Minnesota, North Carolina, Ohio, Tennessee, Virginia, West Virginia, and Wisconsin). During 1964 to 2004, a median of 68 (range: 29 to 167) cases per year were reported in the United States. EEEV disease in humans is associated with high mortality rates (> 20%) and severe neurologic sequelae. In 2004, cases were reported in three states (Massachusetts, North Carolina, and South Carolina). During 1964 to 2004, a median of four (range: 0 to 15) cases per year were reported in the United States. Before the introduction of WNV in the United States, SLEV was the nation’s leading cause of epidemic viral encephalitis. In 2004, cases were reported in five states (Arizona, Kansas, Michigan, Oklahoma, and Texas). During 1964 to 2004, a median of 26 (range: 2 to 1967) cases per year were reported in the United States.

312

SECTION

F

Central Nervous System Infections

Most infections are asymptomatic, and < 1% cause neurologic disease. Western equine encephalitis is the usual cause of arbovirus encephalitis in the western United States; disease is concentrated in infants and the elderly. EEEV and California encephalitis viruses occur in the central and eastern United States. EEEV causes a fulminant disease of low incidence. In August to September, 2005, 11 cases of EEEV encephalitis were reported in southeastern New Hampshire and Massachusetts, which was extraordinary for both states; 3 of 11 cases and 1 of 4 deaths were in children < 5 years of age.9 California encephalitis viruses cause mild disease, primarily in children. WNV encephalitis was Àrst diagnosed in the United States in 1999 and is now the most common arbovirus encephalitis. Although cases sometimes occur in children, more severe disease is associated with increasing age.12 Despite the potential for epidemics, most cases of arbovirus encephalitis occur sporadically. Seasonality is also a feature of rickettsial encephalitis; Rocky Mountain spotted fever occurs during the warm months, when ticks are active. Tickborne diseases, such as Rocky Mountain spotted fever and Lyme disease, are often geographically concentrated in hyperendemic areas. Both occur primarily in the eastern United States, although reservoirs of Borrelia burgdorferi, the organism responsible for Lyme disease, include the upper Midwest states and the PaciÀc northwest and are expanding. Neuroborreliosis follows erythema migrans by 3 to 10 weeks and thus can occur in cooler months. The HIV epidemic has resulted in neurologic diseases in a large proportion of infected individuals.13 Encephalitis can occur during seroconversion; degenerative encephalopathy complicates late disease in up to 90% of infected children (see Chapter 111, Diagnosis and Clinical Manifestations of HIV Infection). Rabies is rare in the United States, averaging 1 human case each year. Fifty percent of rabies infections occur in children; bats are the most common source of infection (see Chapter 228, Rabies Virus). Hyperpyretic (hemorrhagic) shock with encephalopathy affects infants younger than 1 year; peak incidence is at 3 to 4 months. A history of an infectious prodrome is elicited from 60% of patients; there is no clustering of cases.14 Reye syndrome is associated with antecedent viral infections, especially chickenpox and influenza.

Pathogenesis Encephalitis can be caused by several mechanisms, depending on the virulence and tropism of the infectious agent and the exuberance of the host response. Infectious encephalitis is the result of direct virus invasion of any cell type in the brain or spinal cord that causes transient cellular dysfunction, cytolysis, or inflammation. Virus can enter the brain by the hematogenous route, resulting in diffuse encephalitis; measles and neonatal HSV infection are examples. Virus and select bacteria can enter the brain by neuronal tracts, causing focal encephalitis.15 Frontal and temporal localization of HSV encephalitis is characteristic and is explained by retrograde spread of virus from a site of latency in the trigeminal ganglion. Virus-induced cytolysis results in focal or generalized loss of neurons. Neuronal loss and demyelination also proceed by apoptosis mediated by infected astrocytes. Perivascular and parenchymal inflammation of the cortical gray matter, adjacent gray–white junction, basal ganglia, or brainstem with neuronophagia and glial nodule formation is characteristic. Certain viruses, including HSV, cytomegalovirus, adenoviruses, rabies virus, JC virus, and Nipah virus produce characteristic inclusions in a small number of infected cells. Inflammation can be severe enough to cause localized vasculitis that leads to hemorrhage and necrosis. Some agents (VZV, Nipah virus, Rickettsia, Neisseria, and treponemes) cause direct endothelial damage to arteries, arterioles, and capillaries, resulting in vasculitis, hemorrhage, and thrombosis. Systemic vasculitis (such as that due to systemic lupus erythematosus and rheumatoid arthritis) can affect the brain, resulting in similar lesions. Infections can precipitate decompensation in children who have metabolic disorders, particularly influenza-associated encephalopathies, Reye syndrome, or hyperpyretic shock with encephalopathy.

BOX 46-2. Encephalitis Associated with Immunodeficiency • HUMORAL IMMUNODEFICIENCY Chronic enteroviral meningoencephalitis • CELL-MEDIATED IMMUNODEFICIENCYa Progressive multifocal leukoencephalopathy (JC virus) Subacute herpes encephalitis Subacute measles encephalitis Progressive rubella panencephalitis Cytomegalovirus Varicella-zoster virus Toxoplasma gondii • REACTION TO INFUSION OF MUROMONAB-CD3 (OKT3), ETANERCEPT, INFLIXIMAB • ACQUIRED IMMUNODEFICIENCY SYNDROME Toxoplasma gondii Cytomegalovirus Cryptococcus neoformans Human immunodeÀciency virus-1 Progressive multifocal leukoencephalopathy (JC virus) a

Includes immunodeÀciency following organ transplantation.

In immunodeÀcient patients, the natural history of infectious encephalitis is frequently subacute or chronic (Box 46-2).16 Pathologic Àndings include cerebral atrophy, neuronal loss, and demyelination with variable inflammatory response (see Chapter 50, Spongiform Encephalopathies: Slow Infections of the Nervous System).

Clinical Manifestations and Differential Diagnosis The manifestations of encephalitis reflect perturbations in brain function; symptoms and signs are protean. Encephalitis can manifest subtly as psychiatric symptoms, emotional lability, or altered sensorium or as obvious signs such as ataxia, movement disorders, focal neurologic deÀcit, paresis, stupor, and coma. The presence of fever is helpful in distinguishing encephalitis from encephalopathy due to toxins or inborn errors of metabolism. Infectious encephalitis commonly begins with a prodrome of fever, headache, weakness, fatigue, personality change, or irritability lasting for hours to days. Lethargy follows and is the extent of progression in most cases. Seizures are common in children. In severe encephalitis, lethargy rapidly progresses to coma and, in some cases, death. Some infectious agents cause both infectious and postinfectious encephalitis. Mumps encephalitis can precede parotitis by several days; infectious or postinfectious encephalitis can occur as parotitis wanes. Virus infection can reactivate, causing symptoms distantly from Àrst occurrence, such as in HSV encephalitis which is associated with virologic relapse in 5% of children and adults after antiviral therapy is discontinued.17,18 Relapse of encephalitis, chronic movement disorders, and progressive neurologic deterioration after HSV encephalitis may be immunologically mediated or sometimes associated with isolation of the virus or detection of viral genome from the CSF. Diagnostic criteria and pathogenesis of pediatric autoimmune neuropsychiatric disorders associated with streptococcus (PANDAS) is controversial (see Chapter 118, Streptococcus pyogenes (Group A Streptococcus)). Apparent response to antibiotic therapy has been reported.19 Listeria monocytogenes, enterovirus 71 and possibly HSV are especially associated with acute brainstem encephalitis.15,20 Parkinsonian symptoms, movement disorders, neuropsychiatric disorders, especially obsessive-compulsive disorders, are associated with WNV and Japanese B encephalitis and poststreptococcal disorders.21 Hyperpyretic shock with encephalopathy is an idiopathic syndrome characterized by cerebral edema and encephalopathy, profound shock, coagulopathy, and diarrhea (often bloody).14 This syndrome has substantial overlap with heat stroke. Reye syndrome is characterized by cerebral edema and encephalopathy.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Encephalitis, Meningoencephalitis, Acute Disseminated Encephalomyelitis, and Acute Necrotizing Encephalopathy

Clinical Approach Clinical approach to the patient with encephalitis is predicated on the following four principles: (1) the incidence of encephalitis is similar to that of meningitis, so the diagnosis must be considered initially; (2) most causes of encephalitis are difficult to diagnose acutely; (3) some encephalitis and diseases mimicking encephalitis are treatable with specific therapy; and (4) prognosis may depend on prompt institution of therapy. With the emergence of new diagnostic tests and new antiinfective or anti-inflammatory therapies, the diagnosis and management of encephalitis should be as interventional as those of meningitis (Box 46-3).21–25 In the neonate, the repertoire of signs and symptoms of illness is limited. Encephalitis should be considered in any infant with fever and signs of poor feeding, irritability, lethargy, or sepsis. A history of maternal fever in the peripartum period predicts encephalopathy26 and may presage enterovirus infection in the neonate. A history of maternal genital herpes infection is sought, with the realization that it is present in only about one-quarter of infants who have neonatal HSV encephalitis.27 Any evaluation for sepsis should include CSF testing for virus culture and PCR assays as well as bacterial cultures. Abnormal serum hepatic enzyme values and consumptive coagulopathy are clues to disseminated disease. Acyclovir should be given empirically to ill infants with signs compatible with HSV encephalitis and no other likely diagnosis. Management of a less ill infant is individualized. An electroencephalogram and magnetic resonance imaging (MRI) should be performed promptly, and acyclovir therapy should be initiated for infants who do not show improvement with empiric antibacterial therapy. Prolonged vomiting, hypoglycemia, or severe acidosis or hyperammonemia in a young infant should lead to prompt evaluation and treatment for metabolic disorders.

BOX 46-3. Initial Diagnostic Evaluation for Patients with Encephalitis • PERFORM LUMBAR PUNCTURE Measure opening pressure Order routine studies and cytology Culture for virus and bacteria Perform PCR panel (and repeat as appropriate)a Consider immunoglobulin analysis Consider specific antibody tests Consider myelin basic protein analysis • COLLECT SPECIMENS FOR OTHER TESTS FOR INFECTIOUS ETIOLOGY Culture of blood for bacteria Culture/antigen tests/PCR as appropriate of nasopharyngeal, stool specimens for virus Consider cultures of blood, buffy coat, and urine for virus Consider nucleic-acid-based viral load tests of serum Consider specific serum antibody tests • PERFORM METABOLIC SCREENING Crisis blood spot on Guthrie card for PCR or tandem mass spectroscopy Serum electrolytes Blood glucose Blood pHb Plasma ammoniab Toxicology screen, urine, serum • PERFORM MAGNETIC RESONANCE IMAGING • COLLECT ACUTE SERUM SPECIMEN Consider Mycoplasma serology Consider ASO, antiDNase B serologies • CONTACT LOCAL HEALTH EPIDEMIOLOGIST ASO, antistreptolysin O; PCR, polymerase chain reaction. a Repeat cerebrospinal fluid (CSF) PCRs are indicated for treatable infections given the poor concordance of CSF with tissue levels of pathogens. b Plasma for amino acids and urine for organic acids are collected and frozen immediately, then sent for assay if blood pH or ammonia values are abnormal.

CHAPTER

46

313

In the older child it is important to obtain a careful history. The prodrome and pace of illness should be established. An antecedent respiratory infection, exanthem, or vaccination raises the possibility of postinfectious encephalitis. New avian influenza strains may be entero- and neurotropic.28 History of protracted fatigue or pharyngitis may indicate encephalitis due to EBV; exposure to kittens suggests Bartonella henselae encephalopathy or encephalitis. The occurrence of encephalitis in warm months is compatible with Rocky Mountain spotted fever or infection caused by enteroviruses, or arboviruses. A history of travel or contact with ticks or mosquitoes should be sought. Local health departments may have knowledge of respiratory virus activity or arbovirus activity in sentinel animals. A careful neurologic examination includes a global assessment (such as the Glasgow coma scale) as well as examination of sensory, motor, and cerebellar functions. Documentation of specific performances should replace generalities such as “lethargy” and “weakness.” The presence of a rash may suggest Rocky Mountain spotted fever or infection with an enterovirus. Parotitis, pharyngitis, or lymphadenopathy may be present with specific viral infections. Respiratory signs are present in < 50% of patients purported to have Mycoplasma encephalitis.29

Laboratory Findings CSF abnormalities seldom correlate with the clinical or histologic severity of encephalitis. CSF cell count and protein value are frequently normal or slightly elevated (< 200 cells/mm3 and 50 to 200 mg/dL, respectively); the glucose level is often normal; pleocytosis consists predominantly of mononuclear cells. Vasculitis or tissue necrosis (common in HSV, VZV, EEEV, and amoebic encephalitis) causes extravasation of red blood cells into CSF and elicits CSF leukocytosis with a higher proportion of polymorphonuclear cells. A high CSF protein level with polymorphonuclear pleocytosis in this setting suggests contiguous brain necrosis. Because encephalitis is a parenchymal disease, isolation of virus from CSF occurs in only 15% to 50% of cases. Rapid diagnostic tests performed on CSF augment the results of viral cultures. PCR assays are, on average, 75% sensitive in the early stages of encephalitis and highly specific when carefully performed. PCR assays of CSF can detect more than one pathogen.30 PCR of CSF for VZV is often positive in patients with and without skin lesions.31,31a HHV-6 is so ubiquitous that its identification in CSF beyond 2 years of life is not generally informative.32,33 Presence of antibody (globulin) in CSF is indicative of local immune response; CSF antibody levels are assessed in relation to the globulin-to-albumin ratio of serum. Measurements of total CSF immunoglobulin and of antibodies to specific pathogens (e.g., WNV) have been studied as diagnostic tests. Minimal time to appearance is 4 to 5 days for acute encephalitis, limiting the usefulness of the measurement during the first few days of the illness. In chronic viral infections, such as subacute sclerosing panencephalitis and subacute rubella encephalitis, specific CSF antibodies (and usually serum antibodies) are extremely elevated. MRI is the most sensitive imaging modality for infectious encephalitis, showing brain edema and inflammation in the cerebral cortex, gray–white-matter junction, or basal ganglia (Figure 46-2). In contrast, postinfectious encephalitis is associated with foci of demyelination in the semilunar white matter, basal ganglia, or spinal cord. Use of gadolinium contrast improves the sensitivity of MRI in detecting vasculitis and cerebral abscesses. Diffusion-weighted MRI may optimally detect pyogenic infections and predict neurologic sequelae better than standard MRI.34 MRI may be insensitive in detecting encephalitis at the onset of symptoms and signs, especially in neonates with higher brain water content; a second study performed 24 to 48 hours later is usually abnormal. Electroencephalography is equally sensitive (80%) and is a useful complementary test for detecting HSV encephalitis. Computed tomography is an inferior test for acute encephalitis but is superior to other methods for detecting intracranial calcifications caused by congenital brain infections and some metabolic diseases.

314

SECTION

F

Central Nervous System Infections

B

A

Figure 46-2. Herpes simplex type 2 encephalitis diagnosed in a 3-week-old infant with seizures, who had received antibiotics in the immediate perinatal period, and then had repeated episodes of fever thought to be due to partially treated bacterial meningitis. The mother did not have a history of genital herpes. The infant had no skin lesions or history of scalp electrode. Axial noncontrast-enhanced T1-weighted (A) and T2-weighted (B) magnetic resonance imaging scans show areas of hypointensity and hyperintensity, respectively, in the temporoparietal lobes bilaterally (arrows) (representing areas of brain necrosis). (Courtesy of E.N. Faerber and S.S. Long, St. Christopher’s Hospital for Children, Philadelphia, PA.)

Brain biopsy is the definitive test for diagnosis of encephalitis; its sensitivity is 97% for HSV encephalitis. The clinical syndrome consistent with HSV encephalitis can also be caused by infection due to enteroviruses or HHV-6,35 tuberculosis, mitochondrial encephalopathy with lactic acidosis and stroke (MELAS) syndrome, acute hemorrhagic leukoencephalitis, and systemic disorders. Because of improved diagnostic methods and the low rate of adverse effects of acyclovir, therapy with this agent is frequently given empirically in the neonate, with biopsy reserved for patients with negative PCR on CSF, atypical features, or progressive disease despite empiric therapy. Biopsy should be considered early in the course in older infants and children, unless evidence of the cause is compelling or the patient’s condition improves rapidly. This should be done routinely at the time intracranial monitoring is implemented. The specific cause of some cases of encephalitis may be determined by analysis of CSF antibody titers early in the course of disease or retrospectively through measurement of acute and convalescent antibody titers in serum. Interpretation is confounded by lack of specificity of results of IgM assays, cross-reactive antibodies, and polyclonal production of intrathecal antibodies during CNS infection. CSF antibodies can be elevated for months to years after an infection. Mycoplasma antigens frequently cross-react with brain tissue, thus confounding interpretation.29 Paired antistreptolysin O and antiDNase B titers may assist with streptococcus-associated movement disorders.19,21,36

Management Early treatment of encephalitis can reduce the rate of mortality and sequelae. Airway protection, cerebral perfusion, seizure control, and general support of the obtunded or comatose patient are paramount.

Metabolic and toxic disorders causing encephalopathy and seizures should be excluded. Encephalitic symptoms caused by bacterial cerebritis (especially Listeria) or cerebral abscess should be considered, and therapy instituted as appropriate. Therapy for specific pathogens is delineated in Part III of this book. Therapy with acyclovir is initiated empirically in many patients with encephalitis, because HSV causes 10% of cases of encephalitis and 50% of the sequelae. There are insufficient data to recommend use of antimicrobial therapy for Mycoplasma encephalitis.37 The decision to treat varicella-associated encephalitis with acyclovir is usually based on the presence and timing of systemic manifestations suggesting active infection. Many forms of encephalitis are mild and self-limited, even after coma. EBV, B. henselae, and Mycoplasma CNS disease may not require therapy.

Complications and Prognosis The overall risks of death and morbidity from encephalitis are 3% to 4% and 7% to 10%, respectively. The rate of complications is inversely correlated with patient age at onset; children younger than 1 year had fatality rates of 40% to 50% in several retrospective series. Not surprisingly, neonates with disseminated viral infection have poor prognosis, with fatality rates of 50% to 80%.38 Cerebral edema and inappropriate antidiuretic hormone secretion are common complications that may result in rapid decline or death. Neurogenic pulmonary edema has been frequent and fatal in outbreaks of enterovirus 71 infection.39 In benign, self-limited cases, lethargy or coma can last from a few days to more than a week, with gradual improvement over days to weeks. Focal deficits can resolve over an extended period.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Encephalitis, Meningoencephalitis, Acute Disseminated Encephalomyelitis, and Acute Necrotizing Encephalopathy

Severe neurologic residua include changes in personality, behavior disorders, mental retardation, blindness, movement disorders, paretic syndromes, spasticity, and persistent ataxia. Clinical signs at presentation predictive of significant sequelae of infectious encephalitis include lethargy or coma, convulsions (for arbovirus infections), and focal motor deficits. Neuropsychological outcome may be worse with onset under 5 years of age.40 Treated HSV encephalitis in neonates is associated with fatality rates of about 14% and more than 50% of survivors have major neurologic sequelae. Outcome may be worse for infants with HSV-2 infection.41 Data support treatment of neonates with disseminated or CNS infection with acyclovir at high dosage (60 mg/kg per day) for 21 days.18 The mortality rate for treated HSV encephalitis in older children is 28%. After a 10-day course of acyclovir therapy, 5% to 8% of patients with HSV encephalitis experience relapse within 1 month. Most cases of virologic relapse are cured with a second course of therapy. Twenty percent of infants with neonatal herpes disease have cutaneous relapses within 1 month after acyclovir therapy is stopped, and 45% within 6 months.17 Association of multiple cutaneous relapses of HSV type 2 infection and neurologic deterioration has been noted; the benefit of oral suppression therapy is not established.42 Complications of arbovirus encephalitis vary with the etiologic agent. Infections with the California virus group have a fatality rate of less than 1%, with seizure disorders in 10% of survivors. EEEV results in death in 50% to 75% of cases, with neurologic residua in most survivors. St. Louis encephalitis and Western equine encephalitis have fatality rates of 2% to 20%. The rate of residual morbidity is 10% to 25% for St. Louis encephalitis and 13% to 56% for Western equine encephalitis. Coma or convulsions in the acute phase are associated with a worse prognosis. Case-fatality rates for WNV are 4% to 14% but encephalitis most often affects the elderly; there are no estimates of morbidity from WNV-associated acute flaccid paralysis in children. Infectious encephalitis may be followed by acute disseminated encephalomyelitis (ADEM)/postinfectious encephalitis within a few days to several weeks. The clinician is challenged to distinguish between virologic relapse, postinfectious encephalitis, and another episode of a noninfectious, chronic neurologic, or metabolic disorder.

Prevention The rate of infectious and postinfectious encephalitis in the United States has dramatically declined since the introduction of vaccines against measles, mumps, and rubella. Widespread use of the varicella vaccine may lower this rate further. Passive immunotherapy is available to prevent varicella and measles in the immunocompromised host. Draining stagnant water, controlling the mosquito population, and avoiding exposure to mosquito, arthropod, and tick vectors is an effective means of preventing vector-borne encephalitis.

POSTINFECTIOUS ENCEPHALOMYELITIS/ACUTE DISSEMINATED ENCEPHALOMYELITIS Etiologic Agents and Epidemiology The term postinfectious encephalomyelitis (PIEM) is frequently used interchangeably with ADEM, although classically PIEM was diagnosed when neurologic signs and symptoms followed known infection. ADEM commonly follows a nonspecific illness and diagnosis is based on clinical findings and abnormalities detectable by MRI. Clearly, neurologic deficits can occur without (or with delayed or minor) corresponding MRI findings, and vice versa.43–47 Beginning in the first to second years of life, PIEM/ADEM accounts for a substantial proportion of clinical encephalitis. PIEM/ADEM is an acute, monophasic, multifocal inflammatory, demyelinating disorder of the CNS causing development over days of multiple neurologic signs. PIEM often follows respiratory tract infections, especially those caused by influenza virus and Mycoplasma pneumoniae. Encephalitis that occurs after measles infection is the paradigm of PIEM, occurring

CHAPTER

46

315

in 1 per 1000 cases.10 The incidence of PIEM declined precipitously in countries after implementation of vaccination against measles, mumps, and rubella.11 In North America, varicella has also declined as a common cause of postinfectious encephalitis, coincident with the introduction of universal vaccination against varicella. Immunizations rarely cause postinfectious encephalitis; for example, the United States measles vaccine is estimated to cause < 1 case of encephalitis per million doses administered – far lower than the attack rates following natural infection. A history of illness within 2 to 3 weeks or vaccination within 2 to 4 weeks of onset of ADEM is present in 50% to 80% of cases46,47; myriad infectious agents and vaccines have been temporally associated.47 The infection is usually nonspecific and minor. Of 360 children with ADEM reviewed by Davis & Booss,46 the median age was 6.7 years and 60% were boys. There has been no apparent seasonality or geographic clustering of cases since the universal use of MMR and VZV vaccines.

Pathogenesis PIEM (such as that after measles or previously used rabies vaccines prepared in neural tissue) is an autoimmune process characterized by a perivenulitis and contiguous demyelinization. It is uncommon in children younger than 1 year. T-lymphocyte response to myelin basic protein has been demonstrated in these cases; it is similar to the mechanism elucidated in the animal model of experimental allergic encephalitis. Dysregulation of immune modulation, which commonly occurs in measles infection, may be partially responsible for the autoimmune process. Some vaccine-associated PIEM cases were directly attributable to contamination of a specific vaccine (i.e., Semple rabies or Japanese B encephalitis vaccines) with animal brain tissue in which it was propagated. The molecular mimicry concept of PIEM/ADEM proposes a partial structural homology between the infectious agent and myelin proteins of the host. In this model, antigen-specific lymphocytes that enter the CNS or antigen-presenting microglial cells encounter the homologue myelin protein, causing a mononuclear cell perivenulitis and myelin protein loss (demyelination). The inciting pathogen, its RNA or DNA, is absent from the CNS.

Clinical Manifestations and Differential Diagnosis Table 46-3 shows clinical characteristics of PIEM/ADEM, which develop over a mean of 5.1 days. The typical presentation is that of multiple neurologic disturbances accompanied by change in mental status.45–48 Multiple levels of the CNS can lead sometimes to predominance of cranial nerve, brainstem, cerebellar, or spinal cord signs. The diagnosis of ADEM is made on a clinical basis, with MRI findings of demyelination. PIEM/ADEM cannot be distinguished by neurologic findings from acute encephalitis; seizures are less frequent in PIEM/ADEM. It may be difficult to distinguish neuroborreliosis. Multiple sclerosis (MS) is also a demyelinating disease with similar clinical and MRI findings. Diagnosis of MS cannot be made without so-called “dissemination in time” (recurrence of demyelination at the same or a different site) and “dissemination in space” (lesions involving more than one site).45,47,49,50 The highest risk that first episode ADEM is the onset of MS occurs in children > 12 years old who present with hemiparesis, optic neuritis, or spinal cord signs but no distinct prodrome or altered mental status.46 CSF was abnormal in 61% of 360 children reviewed by Davis & Booss,46 with mild pleocytosis (mean, 49 cells/mm3; highest, 270 cells/mm3; mean protein, 58 mg/dL). Oligoclonal bands in CSF (but not serum) were present in 6% and CSF myelin basic protein was elevated in 11%. Cellular response is generally mononuclear, but neutrophils can be present early in the course. Oligoclonal banding in childhood cases reviewed by Menge et al.47 was present in 0% to 29% (lower than in adult cases of ADEM), in sharp contrast to high prevalence in MS and other neuroinflammatory diseases.47 A positive

316

SECTION

F

Central Nervous System Infections

TABLE 46-3. Clinical Features and Neuroimaging Findings in 360 Cases of Childhood Acute Disseminated Encephalomyelitis Clinical Features (244)

%

SYSTEMIC

Fever Headache Nausea/vomiting Neck stiffness

%

SPECIFIC

48 36 32 25

NEUROLOGIC

Altered consciousness Motor signs Ataxia Cranial neuropathy Seizures Visual loss Language impairment Sensory abnormality Movement disorder Peripheral nerve signs

Initial MRI Findings (255)

63 62 38 34 29 21 20 6 5 0

Bilateral lesions CNS white-matter lesions Brainstem or cerebellum lesions Thalamus lesions Basal ganglion lesions Spinal cord lesions (56)

96 93 43 27 23 36

GENERAL

Gadolinium enhancement of lesion (98) Mass effect from lesions Hemorrhage

26 16 100 deaths occurred annually in Japanese children from influenza-associated ANE.52 Emergence in Japan occurred in the year after cessation in 1994 of a 1960s policy of universal annual immunization of children against influenza. Influenza A and B were associated with ANE in direct proportion to clinical respiratory tract illness in the population. In the United States, physicians were alerted to ANE in late 2003 and were encouraged to report cases to the Centers for Disease Control and Prevention. More than 100 possible cases have been reported from broad geographic areas, and without Asian racial predilection. Only some met the screening case definition of: (1) proven influenza; and (2) > 24 hours of altered mental status. The median age of cases was 4.5 years. Influenza was the precipitating event; clinically similar cases during varicella and rotavirus illnesses have been reported.53,54 The pathogenesis of ANE is unknown. Speculation includes mitochondrial respiratory or metabolic derangement (such as in Reye syndrome or Leigh encephalopathy), thermolabile phenotype of carnitine palmitoyltransferase II variations,55 venous thrombosis, and “cytokine storm” in the CNS. Virus is present in the respiratory tract but there is no virus detected by culture or RNA amplification in CSF or brain tissue in affected patients. CSF is acellular, and has elevated concentrations of several proinflammatory cytokines and cytokine receptors – interleukin (IL)-6, IL-1b, tumor necrosis factor (TNF)-a, soluble TNF receptor-1 (sTNF-R1)56–58 – in proportion to severity of CNS disease. Upregulation of IL-6, IL-10, and TNF-a genes has been demonstrated in patients with ANE.59 Brain microscopy shows edema, apoptosis, and necrosis of neurons – without inflammation. Systemic inflammatory response syndrome (SIRS) clinically follows CNS malfunction in some patients.60 It is speculated that the occurrence of ANE, and its emergence in Japan, may be associated with the use of nonsteroidal anti-inflammatory agents (NSAIDS; especially diclofenate sodium and mephenamate) in febrile children, which could cause dysregulation of innate immune responses or an aberrant host–virus interaction.

Clinical Manifestations ANE is a diagnosis made clinically, with typical MRI findings.61,62 It is suspected when, during the height of influenza or another febrile

CHAPTER

46

317

BOX 46-4. Clinical Characteristics of Acute Necrotizing Encephalopathy • CLINICAL AT ONSET Typical age < 5 years, rarely > 10 years Onset during the peak of febrile illness Seizures or altered mental status or both Propensity for word-finding difficulty/mutism Normal complete blood count, serum chemical and hepatic enzyme tests Normal CSF (+ increased pressure, or mildly elevated protein level Negative PCR testing for influenza in CSF • CRANIAL MRI Diffuse, bilateral, symmetric high-intensity signal on T2-weighted images Unremarkable T1-weighted images Involvement of periventricular and deep white matter (thalamus, brainstem tegmentum, cerebellum, medulla) No enhancement with gadolinium • SECONDARY CLINICAL Disseminated intravascular coagulopathy/SIRS that follows CNS signs Multiorgan failure that follows CNS signs • AUTOPSY Edema, apoptosis/necrosis of neurons No vasculitis No inflammation CSF, cerebrospinal fluid; MRI, magnetic resonance imaging; PCR, polymerase chain reaction; SIRS, systemic inflammatory response syndrome.

viral illness, the young child has abrupt onset of altered mental status (frequently mutism)53,61,62 or seizures, or both.63,64 The course of the precipitating illness (unusually) is not severe or complicated. Clinical and neuroimaging findings are shown in Box 46-4.61 SIRS, with thrombocytopenia and serum aspartate aminotransferase level > 1000 IU/L, was associated with poor prognosis in Japanese children.51 Brain MRI of T2-weighted and fluid-attenuated inversion recovery (FLAIR) images show high-intensity signal diffusely in the periventricular and deep white matter bilaterally and concentrically, characteristically affecting the thalamus, brainstem, and cerebellum (Figure 46-4).

Management, Outcome, and Prevention Supportive care is given, anticipating SIRS. Antiviral therapy is given if influenza is identified; the effect on the clinical course of ANE is unknown. Preliminary data suggest that up to one-quarter die and onequarter survive with substantial neurologic sequelae. Immunization against influenza and varicella protect against associated neurologic disease. Pediatricians in Japan caution against the use of NSAIDs during influenza.

318

SECTION

F

Central Nervous System Infections

A

B

Figure 46-4. Acute necrotizing encephalopathy in a 3-year-old with influenza B Sichuan group/Shanghai-like. On the third day of fever and influenzal symptoms he had acute onset of diminished arousal and recognition, mutism, ataxia, and left-sided increase in tone and reflexes. Axial fluidattenuated inversion recovery (FLAIR) magnetic resonance imaging shows bilateral symmetrical, extensive hyperintense lesions in the centrum semiovale with periventricular involvement (A) and in the white matter anterior to the frontal horns, in the corpus callosum, and right globus pallidus (B). (Courtesy of E.N. Faerber and S.S. Long, St. Christopher’s Hospital for Children, Philadelphia, PA.)

CHAPTER

47

Cerebellar Ataxia, Transverse Myelitis and Myelopathy, Guillain–Barré Syndrome, Neuritis, and Neuropathy Rodney E. Willoughby, Jr For anatomic reasons, infection-induced dysfunction of the brainstem, spinal cord, and peripheral nervous system often occurs with predictable complexes of symptoms and signs. Inflammatory processes can affect the cord (myelitis), nerve roots (radiculitis), or peripheral nerves (neuritis). Neurologic dysfunction without inflammation is myelopathy or neuropathy. Histologic characterization is usually inferred rather than proved by biopsy or autopsy. Intrinsic damage to the spinal cord or nerves is the subject of this chapter. Extrinsic damage is covered in Chapter 48, Focal Suppurative Infections of the Central Nervous System. Diseases with a substantial encephalitic component are covered in Chapter 46, Encephalitis, Meningoencephalitis, and Postinfectious Encephalomyelitis. The inciting cause of most episodes of acute cerebellitis or cerebellar ataxia, transverse myelitis, Guillain–Barré syndrome (GBS), and cranial neuropathy is not known (Box 47-1). Diagnosis by acute and convalescent serum specimens is confounded because of the

postinfectious pathogenesis of some myelopathies and neuropathies. Rheumatoid factor can cause false-positive reactions in immunoglobulin M (IgM) assays, and polyclonal stimulation of IgM antibodies reactive against different viruses can occur.1,2 Antibodies against Mycoplasma spp. cross-react with neural elements; Lyme antibody tests lack specificity.3

ACUTE CEREBELLAR ATAXIA Etiologic Agents Encephalitis with ataxia is discussed in Chapter 46, Encephalitis, Meningoencephalitis, and Postinfectious Encephalomyelitis. Frequent or important causes of acute cerebellar ataxia are listed in Box 47-2. This clinically benign but uncomfortable syndrome is usually associated with varicella-zoster virus (VZV) infection, mononucleosis syndromes, or malaria. Measles, mumps, and rubella were common precipitants of cerebellar ataxia before the availability of vaccines. Mycoplasma pneumoniae, pertussis, typhoid fever, scarlet fever, diphtheria, enteroviruses, Borrelia burgdorferi, and vaccines have been implicated in small numbers of patients. Acute cerebellar ataxia can be the occasional presentation of bacterial meningitis, brain abscess, or Listeria monocytogenes rhombencephalitis.

Epidemiology The incidence of isolated acute cerebellar ataxia is unknown. Schoolage children are most commonly affected, a pattern that paralleled the epidemiology of VZV before vaccine introduction. The epidemiology will likely shift to adult age groups.4

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Cerebellar Ataxia, Transverse Myelitis and Myelopathy, Guillain–Barré Syndrome, Neuritis, and Neuropathy

BOX 47-1. Causes of Acute Motor Weakness GUILLAIN–BARRÉ SYNDROME (GBS: ACUTE INFLAMMATORY DEMYELINATING POLYNEUROPATHY: AIDP) • Campylobacter jejuni • Cytomegalovirus • Epstein–Barr virus • Mycoplasma pneumoniae • Vaccines • Surgery/trauma MOTOR NEURON OR AXONAL DISEASE, INCLUDING AXONAL GBS • Spinal muscular atrophy (infantile and late-onset) • Poliomyelitis and other enteroviruses • Flaviviruses (Japanese B encephalitis, West Nile virus) • Campylobacter-associated acute paresis syndrome • Paralytic variant rabies MOTOR END-PLATE DYSFUNCTION • Infant botulism • Myasthenia gravis • Tick paralysis HYSTERICAL PARALYSIS

BOX 47-2. Etiologic Considerations for Isolated Acute Cerebellar Ataxia INFECTIOUS CAUSES • Bacterial meningitis • Listeria rhombencephalitis POSTINFECTIOUS CAUSES • Varicella-zoster virus • Epstein–Barr virus • Malaria (Plasmodium falciparum) • Multiple other agents INTOXICATION • Antiepileptic drugs • Alcohols • Heavy metals TUMOR • Neuroblastoma • Posterior fossa tumors

Pathogenesis and Pathologic Findings Because acute cerebellar ataxia is not a fatal disease, pathologic findings are limited. Purkinje cells of the cerebellum are particularly vulnerable to a variety of noxious or toxic states. Direct viral damage is reported in patients with infectious panencephalitis (see Chapter 46, Encephalitis, Meningoencephalitis, and Postinfectious Encephalomyelitis). Postinfectious demyelination, characterized by venulitis, adjacent focal demyelination, and presence of lipid-laden microglial cells, is believed to be the mechanism for most episodes of acute ataxia. It is unclear why viruses in the herpes family frequently elicit cerebellar encephalitis.

Clinical Manifestations Acute cerebellitis or cerebellar ataxia is characterized by vomiting, incoordination, truncal ataxia, and dysarthria.5 Onset frequently follows a febrile illness; the child is generally afebrile when ataxia begins. Sigmoid sinus thrombosis following acute otitis media/ mastoiditis can manifest with ataxia. Gait disturbance (staggering) is a common presenting symptom. Many children refuse to speak when dysarthric. History of antecedent respiratory infection, exanthem, vaccination, medication, and possible intoxication (especially alcohol, phenytoin, or phenobarbital) may be elicited. Neurologic examination

CHAPTER

47

319

generally confirms truncal ataxia, intention tremor, and dysmetria. Asking an older child to walk a line on a square-tiled floor helps quantify gait disturbance. Nystagmus is a less common finding. Abnormality of sensory findings suggests a different diagnosis. Cerebellar ataxia rarely occurs alone, frequently including other signs of postinfectious encephalitis (see Chapter 46, Encephalitis, Meningoencephalitis, and Postinfectious Encephalomyelitis). Papilledema (papillitis) can be associated with optic neuritis; bilateral papilledema suggests intracranial hypertension. Ataxia can be a feature of central nervous system malignancy, either through a focal effect or as a manifestation of increased intracranial pressure. Cerebellar astrocytoma generally has a subacute onset of symptoms.

Laboratory Findings Acute cerebellar ataxia is a clinical diagnosis. Meningitis and rhombencephalitis must be excluded. Postinfectious encephalitis is characterized by normal to mildly elevated cerebrospinal fluid (CSF), white blood cell count (< 200 mononuclear cells/mm3) and protein concentration (50 to 200 mg/dL). Magnetic resonance imaging (MRI) is superior to computed tomography (CT) for delineating posterior fossa disease; acute hydrocephalus from midbrain edema, cerebellar abscess, and posterior fossa tumor, otitic hydrocephalus must be excluded as causes of ataxia. Testing of serum and urine for toxic ingestion or urine for catecholamines (for occult neuroblastoma) should be considered. Serologic analysis of acute and convalescent serum specimens and polymerase chain reaction (PCR) of CSF may be useful in defining a precipitating infection.

Management and Outcome Treatable causes of acute ataxia include meningitis, mastoiditis and abscesses, tumors, and hydrocephalus. Acute hydrocephalus can occur as a result of local cerebellar edema, occasionally requiring urgent placement of a ventricular shunt to prevent herniation. Optic neuritis can occur without specific complaints in the younger child; MRI occasionally identifies multifocal, clinically silent areas of demyelination characteristic of postinfectious central nervous system disease. The therapy for postinfectious cerebellitis is supportive, consisting of physical therapy and emotional support. Recovery from cerebellar ataxia is usually complete and occurs over days to weeks. A few patients have permanent gait or speech disturbances. Role (if any) of thrombolytic/anticoagulant therapy in central nervous system venous thrombosis is unknown.

ACUTE TRANSVERSE MYELITIS Etiology and Epidemiology Acute transverse myelitis can have an infectious, postinfectious, or vascular pathogenesis. No specific agent is strongly associated with transverse myelitis. Most cases are believed to have a postinfectious cause; prodromal respiratory infection occurs in 25% to 37% of patients (Box 47-3). Measles, mumps, and rubella infections and vaccination against Japanese B encephalitis, smallpox, and rabies were associated with postinfectious transverse myelitis in prior decades. Minor infections, herpesvirus, and M. pneumoniae infections and cat-scratch disease are now more frequently associated with the disorder. Dermatomal reactivation of VZV rarely causes infectious transverse myelitis, whereas VZV antibody may be detected in CSF in the absence of a rash.6 Arterial insufficiency and vasculitis, especially meningovascular syphilis, rarely cause anterior spinal artery syndrome in children. The incidence of transverse myelitis is estimated to be 0.1 per 100 000 population per year. Acute transverse myelitis occurs at any age but is most common in early adolescence and young adulthood.

320

SECTION

F

Central Nervous System Infections

BOX 47-3. Etiologic Considerations for Acute Transverse Myelitis POSTINFECTIOUS CAUSES • Respiratory tract infections • Gastrointestinal tract infections EMERGENCIES OF SPINAL COMPRESSION • Epidural abscess • Tumor • Hematoma • Spinal subluxation

Pathogenesis Pathologic information on postinfectious transverse myelitis exists only for severe cases, in which patchy destruction of the spinal cord with demyelination and microglial proliferation occurring along several spinal segments is seen. Direct viral cytolysis with viral inclusions can be seen after cutaneous zoster.

Clinical Manifestations

BOX 47-4. Initial Laboratory Evaluation for Peripheral Motor Neuropathy TRANSVERSE MYELITIS • Urgent computed tomography or magnetic resonance imaging • Lumbar puncturea • Serology and PCR of CSF ACUTE PARALYSIS • Lumbar puncture • Serology and PCR of CSF • Nerve conduction study • Spirometry • If febrile: • Culture of cerebrospinal fluid for virus ± bacteria • Culture of nasopharynx and stool for virus • If no fever: • Consider stool culture for Campylobacter jejuni • Consider stool culture for Clostridium botulinum toxin assay FACIAL NERVE PALSY • Lumbar puncture • Consider cerebrospinal fluid Lyme antibody test • Serology and PCR of CSF CSF, cerebrospinal fluid; PCR, polymerase chain reaction. a Only after imaging study excludes epidural/subdural collection or compression.

Acute transverse myelitis is a catastrophic syndrome mimicking complete transection of the cord.7 Maximal deficit is reached in less than 1 hour in 15% of patients and by 24 hours in 50%. Subacute progression also occurs. Explicit diagnostic criteria for acute transverse myelitis have been proposed.8 Muscle weakness and bladder dysfunction are universal complaints, whereas constipation and fecal incontinence are less common. Sensory symptoms, including paresthesia, numbness, and coldness, are present in 80% of verbal patients. Back pain is associated with localization of disease to the thoracic cord. Muscular weakness is severe in two-thirds of patients. Deep tendon reflexes are decreased or absent during the acute stages. Loss of sensation of pain and temperature below a well-demarcated level is universal; losses of perception of touch and proprioception are very common. Sparing of proprioception and vibration sense suggests possible anterior spinal artery disease. Although transverse myelitis clinically mimics segmental transection of the spinal cord, a significant anatomic length of the cord is affected by myelitis. Symmetric loss of sensory and motor functions and a prominence of autonomic (bladder) dysfunction are essential to the diagnosis. During the evolution to full deficit, some asymmetry can occur. Substantial asymmetry (such as in the Brown–Séquard hemicord syndrome) suggests another diagnosis. Epidural abscess or spinal cord hematoma, tumor, or compression are neurosurgical emergencies requiring urgent consideration. The back must be inspected for cutaneous stigmata of underlying spinal malformations. Serial neurologic examinations are necessary, because rostral progression of deficit can occur.

Laboratory Findings Exclusion of other causes of cord compression is paramount (Box 47-4). Spinal MRI has replaced myelography in the evaluation for acute transverse myelitis. CSF is collected for PCR analysis for viruses, and acute and convalescent serum samples are obtained.

Management and Outcome Autonomic disturbances, including hypotension, hypertension, and cardiac arrhythmias, can be severe and changeable and require immediate intervention. Arrhythmia can be fatal. Ventilatory failure and aspiration pneumonia can complicate cervical cord involvement. Urinary tract infection is the most common complication. Intermittent catheterization results in less infection than the use of an indwelling catheter. Pneumonia, septicemia from urinary tract infections or

decubitus ulcer, and renal failure are major causes of death. The psychological strain caused by the sudden neurologic deficit is overwhelming; emotional support from experienced professionals is necessary even during early medical management. Contractures and articular subluxations quickly develop in the absence of physical therapy. For optimal outcome, a multidisciplinary approach to the rehabilitation of affected children is required. Some children recover fully from acute transverse myelitis within several months. The prognosis is better in children than in adults; 45% of children with transverse myelitis have substantial residual deficits, compared with 75% of older individuals. Many children regain ambulation, although with spastic gait.

GUILLAIN–BARRÉ SYNDROME AND OTHER ACUTE PARALYTIC DISORDERS Etiologic Agents Neuromuscular weakness can result from damage in the central nervous system, along the descending pyramidal tracts, or in the peripheral nervous system. Common causes of encephalitis with paresis and acute transverse myelitis are listed in Box 47-3 and Chapter 46, Encephalitis, Meningoencephalitis, and Postinfectious Encephalomyelitis. Before the introduction of vaccines, poliomyelitis was the principal cause of acute peripheral nervous system paralysis. Now GBS, a group of postinfectious neuropathies, is most commonly associated with acute peripheral motor weakness. Acute inflammatory demyelinating polyneuropathy (AIDP) and Miller Fisher syndrome are demyelinating disorders; acute motor axonal neuropathy (AMAN) and AMSAN (with a sensory component) are axonopathies.9 Campylobacter jejuni is strongly associated with both AIDP and AMAN.10,11 Cytomegalovirus (CMV) is associated with AIDP12 and CMV DNA has frequently been detected in CSF samples from CMV-seropositive patients with GBS.13 Vaccines are implicated circumstantially, especially the 1976 “swine flu” vaccine. Temporal association of GBS in 17- to 18-year-old individuals following receipt of meningococcal quadrivalent conjugate vaccine (Menactra) after universal recommendation in 2005 was reported but rate does not exceed background rate.14 Hepatitis B virus, influenza virus, and enteroviruses are uncommon causes. Patients with systemic lupus erythematosus,

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Cerebellar Ataxia, Transverse Myelitis and Myelopathy, Guillain–Barré Syndrome, Neuritis, and Neuropathy

CHAPTER

47

321

TABLE 47-1. Syndromes of Acute Motor Paralysis Feature Fever Sensory abnormalities Motor abnormalities CSF WBC count (cell/mm3) CSF protein concentration (mg/dL) Electrophysiologic findings MRI findings

Transverse Myelitis

AIDP

AMAN

Poliomyelitis

Botulism

Spinal Muscular Atrophy

No Yes, prominent Fixed Symmetric Normal ˜ 200 Normal ˜ 200

No Occasional, minor Ascending Symmetric Normal > 200

No No Symmetric

Yes No Asymmetric

Normal Normal–100

10–200 100–200

No No Global Symmetric Normal 0

No No Global Symmetric Normal 0

Denervation

Demyelination

Denervation

Denervation

Denervation

Spinal cord inflammation

Cauda equina radiculus

Not defined

Anterior column inflammation

Incremental response Normal

Anterior column atrophy

AIDP, acute inflammatory demyelinating polyneuropathy; AMAN, acute motor axonal neuropathy; CSF, cerebrospinal fluid; WBC, white blood cell count; MRI, magnetic resonance imaging; WBC, white blood cell.

Hodgkin lymphoma, and early human immunodeficiency virus (HIV) infection may be at greater risk of GBS. Surgery or other trauma is a factor in 5% of cases. Other causes of acute motor paralysis are poliomyelitis syndrome, toxins, and genetic causes (Table 47-1; also see Box 47-1). Fever and acute motor paralysis are usually caused by viral infection, especially polioviruses, coxsackievirus A7, enteroviruses 70 and 71, West Nile and Japanese B encephalitis viruses. Tickborne paralysis, botulism (which causes descending paralysis), diphtheria, rabies, and inorganic toxins are uncommon causes of neuromuscular weakness. Genetic syndromes such as spinal motor atrophy and McArdle syndrome can manifest acutely.

Epidemiology GBS affects children at any age, but peak occurrence is in late adult life. The annual incidence increases from 0.8 per 100 000 individuals younger than 18 years to 3.25 per 100 000 in those older than 60 years. Most pediatric cases occur in young adolescents. The highest attack rates for the AMAN form of GBS are in children during the summer. AMAN accounts for a substantial fraction of GBS in developing nations.15 Other causes of paralytic disease are rare. With the universal use of inactivated poliovirus vaccine in the United States, cases of paralytic poliomyelitis should cease (see Chapter 236, Polioviruses). Botulism occurs at a rate of 0.03 to 0.05 per 100 000 population. Infant botulism occurs primarily in infants 2 to 4 months old, especially in California, Hawaii, Ohio, and Pennsylvania.

Pathogenesis The motor unit of the peripheral nervous system is composed of the anterior horn cell, axon, myelin, neuromuscular junction, and muscle fibers. The pathophysiology of GBS is incompletely understood, but AIDP is believed to involve deposition of antibody and complement along the myelin sheath. Inflammatory infiltrate along nerves is perivenular, with associated demyelination. Axons are spared. Lesions are focal and are dispersed throughout the peripheral nervous system, including the anterior and posterior roots, sensory and autonomic ganglia, and proximal and distal nerve segments. The efficacy of plasmapheresis suggests that humoral factors may be involved in the pathogenesis of GBS, but an animal model of experimental allergic neuritis produces identical lesions through a cell-mediated process. Poliomyelitis syndrome is characterized by virus infection and cytolysis of anterior horn cells in a patchy distribution. There is diffuse inflammation of ventral horns and the bases of dorsal horns of the spinal cord, with neuronal degeneration and neuronophagia. Lymphocytic inflammation of the meninges is slight and patchy.

BOX 47-5. Diagnosis of Guillain–Barré Syndrome REQUIRED • Progressive weakness of more than one limb • Areflexia STRONGLY SUPPORTIVE • Relative symmetry • Mild or no sensory symptoms • Cranial nerve involvement • Autonomic dysfunction • Absence of fever • Disease progression halts by 4 weeks • Recovery CEREBROSPINAL FLUID FEATURES • Elevated protein level after first week • < 10 cells/mm3; mononuclear ELECTRODIAGNOSTIC FEATURE • Slowing of nerve conduction

Botulism is caused by a noninflammatory, toxin-mediated inhibition of the release of acetylcholine from synaptic vesicles of terminal nerves.

Clinical Manifestations GBS is a motor polyradiculoneuropathy, characterized by muscle pain and symmetric, ascending paresis with minor sensory abnormality. History of antecedent upper respiratory tract or gastrointestinal illness, recent vaccination, trauma, or surgery may be elicited. AIDP and AMAN are indistinguishable; AMSAN has a more prominent sensory loss. Onset is subtle, with ascending progression of paralysis over several days to weeks. Weakness begins with footdrop and proceeds proximally to involve the legs and trunk. The child may complain of paresthesia of the feet followed by leg, buttock, or back pain. Paresthesias of the fingers and upper-extremity weakness may begin when lower-extremity weakness reaches the level of the shins. The patient is usually afebrile. Loss of deep tendon reflexes precedes and is generally more severe than the muscle weakness. GBS can progress to quadriparesis and involve muscles of respiration or bulbar function. Cranial nerve involvement includes the Miller Fisher syndrome, which consists of ophthalmologic, ataxia, and reflexes. Bulbar nerve paresis appears to be more common in children younger than 5 years. Autonomic instability can occur. GBS is a clinical diagnosis that has considerable overlap with other neurologic diseases. Criteria for the diagnosis have been established and refined (Box 47-5).16 Prominent sensory deficit, bladder or bowel dysfunction, marked persistent asymmetry of the motor deficits, a

322

SECTION

F

Central Nervous System Infections

high CSF white blood cell count, or a predominance of granulocytes suggest an alternate diagnosis. A variant of classic GBS occurs in HIV-infected patients; CSF findings in this subgroup include higher mononuclear cell counts, averaging 23 cells/mm3.17 Poliomyelitis syndrome, botulism, and other motor unit diseases should be considered in the young patient with GBS (see Table 47-1). Paralytic poliomyelitis in the Americas can be caused by nonpoliovirus enterovirus strains or West Nile virus. Poliomyelitis is characterized by fever, intense muscle pain, and muscle weakness. The onset of paralysis is sudden and is preceded by the loss of deep tendon reflexes in the affected area. Asymmetric motor weakness is typical except in severe cases. Autonomic dysfunction occurs in up to 20% of cases; sensory deficit does not occur. Botulism affects autonomic and motor function; most infants have a history of constipation, and physical examination may reveal mydriasis. Spinal muscular atrophy can become apparent suddenly after an environmental stress.

intravenous (IGIV) (400 mg/kg per day for 5 days) may be of some benefit if therapy is instituted early in the clinical course.18,19 Plasmapheresis may not be as effective in C. jejuni-associated GBS. Corticosteroids are not effective.

Laboratory Findings

POLYRADICULITIS AND NEURITIS WITH SENSORY SYMPTOMS

There is no confirmatory laboratory test for GBS. Acellular CSF with elevated protein content is suggestive but may not be found in the first week of illness (see Box 47-4). HIV infection is associated with a variant of GBS with greater CSF pleocytosis. Electrophysiologic testing in AIDP shows reduction in nerve conduction velocity, abnormal temporal dispersion, prolonged distal latencies, and absence of F waves, findings consistent with demyelination. Study findings are normal in up to 14% of patients at first testing, so testing may need to be repeated.17 Proximal sites may yield a higher percentage of abnormal findings.16 AMAN is associated with marked reductions in the compound muscle action potential. MRI of the spine is not recommended but, in rare cases, can show inflammation of motor and sensory nerve roots. Nerve biopsy provides definitive diagnosis but is rarely necessary. Antiganglioside antibodies are frequently present in GBS and are associated imperfectly with inciting infections, GBS types, and longterm outcomes. The role of antiganglioside antibodies in pathogenesis is unclear. Antibodies to ganglioside GM1 are found in all GBS types, but especially in cases with distal paralysis, in pure motor forms, in cases occurring after Campylobacter jejuni infection, and in cases with worse outcome. Antibodies to ganglioside GD1a are specifically associated with cases of AMAN (present in half). Antibodies to GM2 are associated with CMV infection, and antibodies to Gal-ceramide with M. pneumoniae infection. Antibodies to ganglioside GQ1b are associated with the Miller Fisher type and Bickerstaff brainstem encephalitis. Poliomyelitis syndromes are characterized by CSF pleocytosis and elevated protein content consistent with aseptic meningitis (see Table 47-1). Enterovirus can be isolated from CSF, blood, pharynx, urine, or stool specimens; conjunctival cultures can be positive in enterovirus 70 infection. Virus is shed for weeks after onset of the illness. Specific CSF IgM is diagnostic for West Nile virus infection. Nerve conduction studies show anterior horn cell dysfunction or axonopathy; electromyogram indicates muscle denervation. Botulism is associated with normal nerve conduction and amplitude, with electromyographic evidence of denervation. In botulism, high-frequency stimulation of motor nerves elicits the characteristic progressive increase in muscle action potentials as remaining functional motor terminals are recruited. Diagnosis is confirmed by the identification of botulinum toxin in the stool of affected infants, or serum in cases of foodborne botulism.

Management Management of mild GBS or other paralytic disease is supportive. The course of paresis is closely monitored; progression can continue over weeks or can accelerate rapidly. Serial measurements of maximal respiratory effort is an important way of predicting impending ventilatory failure. Forced vital capacity < 10 to 15 mL/kg requires prophylactic intubation. Control of the airway may also be indicated for severe bulbar paresis. Plasmapheresis or immune globulin

Complications and Outcome GBS is complicated by respiratory compromise in 12% to 20% of cases, and fatal outcome in 5%. Pneumonia occurs in 25% of patients, and urinary tract infection in 30%. Dysautonomia can be severe. Relapses occur after natural resolution, immune globulin therapy, and plasmapheresis. Children are believed to have milder illness and faster recovery than adults. In children who require ventilatory assistance, intubation averages 3 weeks, and hospitalization more than 80 days.

Etiologic Agents There are few infectious causes of primary sensory neuropathy. Lyme disease (Borrelia burgdorferi) can be associated with mixed motor and sensory polyradiculoneuropathy with frequent cranial nerve involvement. Advanced HIV infection is associated with many forms of neuropathy, including distal symmetric polyneuropathy, CMVor VZV-associated caudal polyradiculopathy, chronic inflammatory demyelinating polyneuropathy, and mononeuritis multiplex.20 Antiretroviral (nucleoside) therapy causes neuropathy in some patients. Prolonged use of metronidazole and exposure to heavy metals or Vacor (rat poison) can precipitate toxic sensory neuropathy. Syphilis, Hansen disease (leprosy), and diphtheria are rarely encountered as a mononeural cause of anesthesia in the United States.

Epidemiology Most literature describing neurologic manifestations of Lyme disease in children is from Europe, because the disease there characteristically has more neurologic and fewer articular manifestations than the disease in North America.21 Cranial nerve palsy is the most commonly recognized manifestation of neuroborreliosis in children. Bannwarth syndrome (meningitis, radiculitis, and cranial neuropathy) represents only 4% of cases. Distal symmetric polyneuropathy accounts for 90% of all HIV neuropathy, which in turn constitutes 5% to 20% of all neurologic complications of HIV infection. Hansen disease occurs at an annual rate of 0.07 per 100,000 population in the United States, with 20% to 30% of cases in children (see Chapter 135, Nontuberculous Mycobacterium Species).

Pathogenesis Lyme polyradiculoneuropathy is a vasculitis affecting both myelinated and unmyelinated nerve fibers. The sensory polyradiculopathy of late HIV infection is characterized by axonal loss and endoneural vasculitis.

Clinical Manifestations Lyme disease causes a motor and sensory radiculitis characterized by isolated asymmetric weakness, shooting pain, and dysesthesia. Erythema migrans rash, which lasts several weeks, is often waning when neuritis appears; rash occurs in only 30% of cases of neuroborreliosis. Patients with the distal symmetric polyneuropathy of acquired immunodeficiency syndrome (AIDS) complain of burning acral dysesthesia. There is minimal motor involvement; ankle reflexes are diminished. The cauda equina syndrome (lumbosacral radiculitis) in patients with AIDS is caused by CMV or VZV, with motor, sensory, bladder, and bowel dysfunction. Symptoms can spread rostrally.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Cerebellar Ataxia, Transverse Myelitis and Myelopathy, Guillain–Barré Syndrome, Neuritis, and Neuropathy

Laboratory Findings Serologic tests should be positive in children with neuroborreliosis; CSF is typical of aseptic meningitis. The sensory polyradiculitis of HIV infection can be associated with abnormal nerve conduction studies and CSF albuminocytologic dissociation. The cauda equina syndrome seen in AIDS is usually caused by active CMV infection. CSF contains granulocytes and elevated protein concentration; CSF culture for CMV is usually positive. Rapid diagnosis by shell vial culture of CSF or by PCR assay is often possible. The VZV-associated syndrome can be diagnosed serologically and using CSF PCR assay.6

Management and Outcome Lyme polyradiculoneuropathy usually responds to antibiotic therapy (delineated in Chapter 185, Borrelia burgdorferi (Lyme Disease)). Lyme polyradiculitis can be complicated by cranial neuritis or encephalitis. The CMV-associated cauda equina syndrome is treated with ganciclovir or foscarnet; the benefit of antiviral therapy of VZVassociated disease is not clear. There is no well-established therapy for the distal neuropathy of AIDS; benefits of IGIV or plasmapheresis are short-lived. Psychological support and pain control are mainstays of therapy. Prognosis for neuropathy in patients with HIV is affected by the late stage of HIV disease. The distal sensory polyradiculitis of late HIV infection is often associated with progressive HIV encephalopathy.

ACUTE CRANIAL NEUROPATHY Etiologic Agents Acute cranial nerve dysfunction reflects local brainstem disease (e.g., brainstem abscess or hydrocephalus), inflammation or compression of the nerve (e.g., basilar meningitis), intrinsic neuritis, neuropathy or ganglionitis (e.g., GBS, Lyme disease, herpes simplex virus (HSV)), or dysfunction of motor endplate (e.g., botulism) or muscle (e.g., myopathy) (Table 47-2). Most acute cranial neuropathies occur without a definite cause and are presumed to be postinfectious. Pathogens associated with acute postinfectious neuropathies include Epstein–Barr virus, HSV, VZV, CMV, M. pneumoniae, enteroviruses,

CHAPTER

Epidemiology The incidence of optic neuritis in children peaks in late adolescence because of its high frequency in patients with multiple sclerosis. It is estimated that 30% to 50% of patients with optic neuritis eventually have multiple sclerosis. The etiology of optic neuritis is usually autoimmune (e.g., postinfectious, paraneoplastic) or toxic (e.g., ethambutol, lead) but can also be caused by Borrelia and Bartonella infections. Oculomotor paresis is frequently encountered with benign intracranial hypertension (pseudotumor cerebri), hydrocephalus, encephalitis, and infant botulism. Trigeminal nerve lesions are infrequent. Acoustic nerve damage is unusual except as a complication of pyogenic meningitis and Kawasaki disease;22 congenital infections, especially with CMV, should also be considered. Isolated peripheral facial nerve palsy (Bell palsy) is the most common cranial neuropathy. The cause of Bell palsy remains enigmatic but is hypothesized to be postinfectious.2 In areas of Lyme endemicity, 30% of all facial palsies and most bilateral palsies are attributed to neuroborreliosis.21 Bulbar involvement can be seen as part of other diseases, including encephalitis (see Chapter 46, Encephalitis, Meningoencephalitis, and Postinfectious Encephalomyelitis), poliomyelitis, GBS, botulism, and diphtheria. Malignancy, especially neuroblastoma, should be considered. Bulbar involvement in poliomyelitis and GBS appears to be more common in infants than in older patients.

Clinical Manifestations Visual loss from optic neuritis is commonly sudden. Neuritis may be difficult to detect in the young child. A careful funduscopic examination is mandatory in children with other postinfectious neurologic diseases. A history of preceding illnesses and exposures is sought. Isolated acute oculomotor paresis causes diplopia, often with associated retro-orbital pain (Tolosa–Hunt syndrome). The young

Anatomic Site of Disease Brainstem

Optic neuritis

323

adenovirus, and influenza and other respiratory viruses. B. burgdorferi is associated with infectious neuropathy, especially optic neuritis and peripheral facial palsy. Malignancy, hydrocephalus, or brainstem abscesses can cause nerve dysfunction. Multiple sclerosis and sarcoid are considered but rarely have onset in childhood.

TABLE 47-2. Causes of Cranial Neuropathy According to Site of Disease

Cranial Nerves Affected

47

Nerve

Extrinsic to Nerve

Previous infection Lyme disease

Astrocytoma Basilar meningitis Parasellar mass Hydrocephalus Pseudotumor cerebri

Endplate/Muscle

Oculomotor paresis (III, IV, VI)

Encephalitis Glioma

Previous infection

Myasthenia gravis Infant botulism Chronic sinusitis Trichinosis

Trigeminal palsy

Encephalitis Glioma Abscess

Previous infection Varicella-zoster virus Herpes simplex virus

Facial palsy

Encephalitis Glioma Abscess

Previous infection Lyme disease Kawasaki disease

Mastoiditis Temporal bone tumor

Myasthenia gravis Infant botulism

Bulbar palsy (lower cranial nerves)

Encephalitis Poliomyelitis Medulloblastoma Abscess Rabies

Previous infection Diphtheria

Neuroblastoma

Infant botulism

Infant botulism

324

SECTION

F

Central Nervous System Infections

child closes one eye to avoid seeing double. Orbital cellulitis or adjacent sinusitis is excluded through an examination for signs of inflammation, proptosis, sinus tenderness, and purulent drainage from the sinus ostia. Chronic meningitis, especially tuberculous meningitis, as well as causes of increased intracranial pressure or focal compression of the third, fourth, and sixth cranial nerves should be considered. Ipsilateral ear pain is frequently associated with Bell palsy. Antecedent respiratory or gastrointestinal illness suggests a postinfectious cause. History of tick exposure and focal rash are important but insensitive indicators of neuroborreliosis. Otitis media and mastoiditis are occasional causes of facial palsy. Isolated peripheral facial nerve palsy as well as hearing loss can be associated with Kawasaki disease.2 Pooling of oral secretions, a weak cry or voice, nasal twang, and nasal regurgitation are signs of the bulbar dysfunction (associated with GBS, poliomyelitis, diphtheria, or malignancy.)

CHAPTER

48

Focal Suppurative Infections of the Central Nervous System Ram Yogev

Focal suppurative infections of the central nervous system (CNS) reviewed in this chapter include brain abscess, subdural empyema, epidural abscess, spinal subdural and epidural abscess, and septic venous thrombosis.

ETIOLOGY Laboratory Findings Diagnostic evaluation of isolated cranial neuropathy generally consists of careful delineation of the anatomy and extent of deficit and lumbar puncture to exclude chronic meningitis (see Box 47-4). MRI evaluation of optic neuritis is insensitive but commonly highlights other, silent areas of central nervous system demyelination. CT can be helpful in assessing sinus and inner-ear disease.

Management and Complications Patients with Lyme disease or chronic infectious meningitis are treated with appropriate antimicrobial agents (see Chapter 185, Borrelia burgdorferi (Lyme Disease)). The use of corticosteroids, IGIV, and other treatments remains controversial in patients with optic neuritis, facial palsy, and other cranial neuropathy. Bulbar nerve paresis results in the inability to swallow secretions and food or to protect the airway. Aspiration leading to laryngospasm or pneumonia is common. Four percent of children with bulbar poliomyelitis die. Subclinical foci of presumed demyelination may be found incidentally in patients with cranial neuropathy who undergo MRI evaluation. Foci of presumed postinfectious demyelination can persist for months. Diagnosis of multiple sclerosis is not made unless there is supporting evidence and the patient has recurrent episodes or new lesions (so-called dissemination in time and space). Use of visual evoked response tests may aid in earlier diagnosis.23,24

Most microorganisms (i.e., bacteria, fungi, or parasites) can cause focal suppurative infection of the CNS. The Streptococcus milleri group, which consists of three species, S. constellatus, S. intermedius, and S. anginosus, is frequently isolated (50% to 70% of cases). Staphylococci (10% to 30% of cases), and enteric bacteria (e.g., Escherichia coli, Klebsiella, Proteus spp.) (10% to 25% of cases) are also commonly isolated from patients with brain abscess (Table 48-1).1,2 Mixed infections are common in brain abscess (up to 30% of cases), and isolation of anaerobic bacteria (e.g., Bacteroides, Prevotella spp.) or nutritionally variant streptococci (e.g., Abiotrophia species3) is increasing as proper culture techniques are used.4 In neonates, gram-negative bacteria are more likely to cause brain abscess. Cases of Citrobacter species,5,6 Salmonella,7 Serratia,8 Proteus,9 Enterobacter sakazakii,10 and Bacteroides fragilis11 meningitis are especially likely to lead to brain abscess or subdural empyema. A rare case of Mycoplasma hominis and Ureoplasma species causing brain abscess in a neonate was also reported.12 Table 48-2 is a partial list of additional agents that occasionally causes brain abscess at any age. A variety of gram-negative bacteria can constitute a single etiologic agent or can be part of mixed infection. In individuals with chronic middle-ear infections and those with neutrophil defects (most commonly chemotherapy-induced), the incidence of Pseudomonas aeruginosa infection is increased.13 In patients with impaired cell-mediated immune response, particularly macrophage- and Tlymphocyte-related defects, or in patients with phagocytic defects, Listeria monocytogenes14 or Nocardia spp.15 are more commonly isolated from brain abscesses. When Eikenella invades the CNS, it has a propensity to cause an abscess, and rarely Brucella,16 Moraxella,

Prognosis and Sequelae Approximately 30% to 50% of all patients with optic neuritis have multiple sclerosis. The prognosis in children is not clear. Recovery of vision is expected but may require weeks. Some degree of afferent pupillary defect persists. Patients with postinfectious oculomotor palsy recover well, whereas those with trigeminal nerve involvement can have permanent pain or numbness. The rate of recovery of facial nerve function in patients with Lyme disease after specific therapy is 80% to 90% (a rate identical to that for all facial palsy). Bulbar palsy resolves incompletely in 10% to 20% of cases caused by GBS. Residua from bulbar poliomyelitis depend on the degree of paresis; recovery may continue for up to 2 years.

TABLE 48-1. Common Bacterial Causes of Focal Suppurative Infections of the Nervous Systema Epidural Abscess

Bacteria

Brain Abscess

Subdural Empyema

Cranial

Spinal

Streptococcib Staphylococci Enterobacteriaceae Anaerobic bacteria Pathogens of meningitisd Mixed Sterilef

50–70 10–30 10–25 20–30c 5–10 20–30 10–25

40–60 10–20 5–15 10–20c 60–70e 5–10 20–30

40–60 10–20 5–15 10–20 5–10 5–10 20–30

5–10 60–80 5–20 5–10 20% of cases; less common, 5–15% of cases; occasional, < 5%. Data from references 1, 71, and 72.

Hematogenous Spread Hematogenous spread from distant sites is probably the most common cause of brain abscess in children. CCHD is the most common predisposing factor in this category, with tetralogy of Fallot and transposition of the great vessels causing the majority of the cases. The current approach to surgical repair at an early age diminishes the incidence of this complication. Any cardiac lesion with a right-to-left shunt, such as ventricular septal defect or patent foramen ovale,81 allows contaminated venous blood to bypass reticuloendothelial function of the lung. As many as 6% of persons with unrepaired CCHD will develop a brain abscess; infarctions of the brain predispose to infection, with infectious risk increasing with age in unrepaired individuals. Patients with endocarditis can also develop brain abscesses, especially those with left-sided and acute endocarditis, which has an increased likelihood of septic embolization. The intensity and the duration of bacteremia are important factors in establishing CNS infection. Septic embolization from phlebitis of deep neck veins associated with parapharyngeal infection (Lemierre disease) can lead to brain abscess,82 as can chronic pyogenic lung disease, such as abscess, empyema, bronchiectasis, or immunoglobulin A deficiency. As therapeutic improvements have resulted in individuals with cystic fibrosis having increased life expectancy, brain abscess complicating bronchiectasis is also increasing.73 Brain abscess is reported in association with pulmonary arteriovenous malformations78 and the hepatopulmonary syndrome (a complication of cirrhosis, extrahepatic portal hypertension, or acute hepatitis).79 Brain abscess also can occur as a metastatic focus of bacteremia from pyogenic infections at distant sites (e.g., bone, teeth, skin, bronchopulmonary,83 abdomen). Endoscopyassociated cases probably result from penetration during the procedure of mucosal bacteria to Batson plexus of veins, which communicate with the cranial sinuses.77 It was also reported as a result of septic abortion and in situ intrauterine device.80

Contiguous Spread PATHOGENESIS Brain Abscess Box 48-1 summarizes conditions that predispose to the development of brain abscesses through hematogenous or contiguous spread in children.

Spread from a contiguous site of infection (e.g., middle ear, sinus, orbit, face, or scalp) is the second most common pathogenesis of brain abscess. The most common contiguous sites reported are the middle ear in young children and the paranasal sinuses in adolescents.84–86 Development of large air sinuses and rapid growth of the frontal sinuses in the second decade of life plus the ability for more vigorous

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Focal Suppurative Infections of the Central Nervous System

nose-blowing were suggested as causes for the increase in CNS suppurative infections seen in adolescents with sinusitis. In most cases, the spread of the infection is through pre-existing anatomic openings, but extension by means of thrombophlebitis of diploic or emissary veins, osteomyelitis or hematogenous dissemination also occurs. Relatively few patients with bacterial meningitis develop brain abscess. A more severe tissue injury occurring during meningitis appears to be necessary to initiate the formation of a brain abscess. Factors such as thrombophlebitis, venous stasis, ischemia, and infarction probably also contribute to the pathogenesis. Pyogenic infection of the cavum pellucidi and/or vergae (the structures between the lateral ventricles) rarely occurs due to anatomical communication with the ventricular system or the same mechanisms that cause a brain abscess.87 Brain abscess following neurosurgery or head trauma is relatively rare, but penetrating injury to the skull, such as from a dog bite,88 pencil puncture,89 lawn dart,90 or open skull fracture, increases the risk. Rarely, the penetrating trauma is from oral injuries (such as chopsticks).91 CNS shunt-related brain abscess is a rare complication of either gut-associated ascending infection along a ventriculoperitoneal shunt or untreated shunt infection.74 Halo device (used to immobilize the cervical spine) pin penetration can cause cerebrospinal fluid (CSF) leak in 1% of patients and rarely intracranial abscesses.75 Other foreign bodies (e.g., sublaminar or interspinous wires used to stabilize or fuse the cervical spine) can migrate intracranially and cause an abscess.92 Underlying brain pathology, such as intracerebral hematoma, necrosis, or neoplasm, can rarely serve as the nidus of infection.

Subdural Empyema Subdural empyema usually results from the direct spread of infection from the meninges in young children and from infection at contiguous extracranial sites (e.g., middle ear, sinus, or calvarium).33 Bone defects following mastoidectomy or inflammatory process of the vascular channels (i.e., emissary veins) or hematogenous dissemination are some of the possible mechanisms.85 Other causes include infection related to prior craniotomy,35 skull trauma,35 ventriculoperitoneal shunt,74,93 pre-existing hematoma,94 halo-pin traction,95 hematogenous spread such as from the lung, or endoscopic procedures.96

Epidural Abscess Cranial Sites Cranial epidural abscess is almost always the result of contiguous extension of infection from sinuses,97 middle ear, or orbit. Rarely, a penetrating injury to the head can lead to such infections. Epidural abscess has also been reported as a complication of fetal monitoring98 and following a wrestling injury.99 Because the dura adheres tightly to the skull, the abscess usually develops slowly, which explains the insidious clinical presentation of this disease. Often, infection of the epidural abscess spreads to other areas (e.g., subdural space, brain) and its potential coexistence with other intracranial infections should be investigated.

Spinal Sites Hematogenous spread from other sites of infection (e.g., skin, soft tissue, bone, respiratory or urinary tract) is the major source of spinal epidural abscess. Direct extension from local osteomyelitis or retropharyngeal, retroperitoneal, or abdominal abscess can also occur.70a Spinal fracture,100 penetrating injury, spinal surgery, and local invasive procedure (e.g., lumbar puncture, steroid injection,101 epidural analgesia102) have been identified as possible causes. Other mechanisms include: complication of meningitis, fistula in Crohn

CHAPTER

48

327

disease,103 midline neuroectodermal defects (e.g., dermal sinus), and intradural tumors (e.g., lipoma). In almost one-third of reported cases, the pathogenesis is unknown.

Septic Venous Thrombosis The major risk factor for intracranial septic thrombosis is meningitis (especially of the superior sagittal venous sinus). Spread of infection along the emissary veins from contiguous infections (e.g., sinuses, ear, face, or oropharynx) is another common risk factor.104 Most infections of the lateral venous sinus are of otogenic source, whereas infection of the cavernous sinus usually originates from the paranasal and sphenoid sinuses, teeth, or face.105 Septic venous sinus thrombosis can also occur in association with an epidural abscess or by hematogenous spread. Patients with increased risk of thrombosis, such as those with sickle-cell disease, dehydration, or certain malignancies, are at higher risk for developing this complication.

CLINICAL MANIFESTATIONS Brain Abscess Multiple factors contribute to the development of clinical symptoms. For example, the location of an abscess affects the signs and symptoms that are present (Box 48-2). Frequently, a frontal lobe abscess remains asymptomatic for long periods of time, and symptoms and signs only develop when the abscess creates an expanding mass effect and increased intracranial pressure. Parietal lobe abscess also remains clinically silent for extended periods of time until it extends to the sensorimotor cortex. The virulence of the pathogen and the host’s immune status also affect the acuteness of clinical presentation. In children, the mean duration of symptoms and signs before the diagnosis of bacterial brain abscess is 2 weeks,1 but in some cases it can be as long as 4 months. The most common initial signs and symptoms of brain abscess are headache, fever, and vomiting, occurring in more than one-half of cases (Table 48-4). A lateralized headache in the pediatric age group is uncommon and should raise the suspicion of intracranial pathology. Mental status changes are observed in fewer than 50% of cases of brain abscess; coma occurs in 15% to 20% of cases. Generalized seizures are seen in fewer than 50% of the children, whereas focal neurologic abnormalities occur in 35% to 40%, and papilledema and signs of meningeal irritation are seen in about one-third of affected children. It has been suggested that the combination of fever, headache, and focal neurologic deficits is a strong indicator of brain abscess, but this triad of symptoms is found in fewer than 30% of cases. Thus, the clinical presentation of brain abscess is relatively nonspecific in infants and children and must be distinguished from several other processes (Table 48-5). Sudden deterioration of the patient’s clinical status, usually with signs and symptoms of meningitis, suggests rupture of the brain abscess into the ventricles or the subarachnoid space.106

Subdural Empyema Clues that a subdural empyema has developed as a complication of bacterial meningitis include persistent bulging of a fontanel and fever, with new onset of neurologic findings (e.g., depressed responsiveness, seizures) during the course of appropriate antimicrobial therapy. The triad of severe headache, nuchal rigidity, and fever (common in many patients) should raise the suspicion of subdural empyema.35,107 Signs and symptoms of increased intracranial pressure (e.g., vomiting, depressed responsiveness, enlarging head circumference, or papilledema) should also lead to an investigation for subdural empyema.108 In older children, the clinical manifestations of subdural empyema mimic those of brain abscess. Hemiparesis or hemiplegia occurs

328

SECTION

F

Central Nervous System Infections

BOX 48-2. Signs and Symptoms in Relation to Site of Brain Abscess FRONTAL LOBE Headache Behavioral changes Motor speech disorder Forced grasping and sucking Depressed consciousness Hemiparesis Papilledema TEMPORAL LOBE Dyspraxiaa Ipsilateral headache Ipsilateral third cranial nerve palsy Aphasiaa Upper homonymous hemianopia Motor dysfunction of face and arm PARIETAL LOBE Visual field defects in inferior quadrant Homonymous hemianopia Dysphasiaa Dyspraxiab Contralateral spatial neglectb INTRASELLAR Headache Visual field defects Endocrine imbalance CEREBELLAR Dizziness Vomiting Ipsilateral ataxia Ipsilateral tremor Nystagmus (beating toward lesion) Sixth cranial nerve palsy Papilledema BRAINSTEM Headache Dysphasia Vomiting Facial and multiple cranial nerve palsies Hemiparesis

more commonly than in patients with brain abscess. However, other symptoms and signs (e.g., severe headache, fever, meningeal signs, vomiting) are not helpful in differentiating empyema from other intracranial processes. Subdural empyema that develops following surgery or trauma can lead to only localized signs, such as overlying wound infection or focal neurologic signs. Fever or vomiting, or both, can occur, but headache is often absent.

Epidural Abscess Symptoms of cranial epidural abscess can be nonspecific for weeks, followed by fever, headache (sometimes localized pain with tenderness), and changes in mental status. Other signs and symptoms common in other intracranial suppurative processes, such as vomiting, seizures, focal neurologic deficits, papilledema, or meningeal signs, are relatively rare and may develop only when the abscess has grown sufficiently large to cause a mass effect or increased intracranial pressure. In a few situations, the symptoms suggest the location of the abscess. For example, unilateral facial pain with weakness of the

TABLE 48-5. Differential Diagnosis of Brain Abscess Noninfectious Etiologies Infectious Etiologies

Vascular

Other

Meningitis Encephalitisa Meningoencephalitis Subdural empyema Mycotic aneurysmb Epidural abscess Cranial osteomyelitis Suppurative thrombosis Cysticercosis Tuberculomac Cryptococcosis

Hemorrhage Intracerebral Subarachnoid Subdural (chronic) Venous sinus thrombosis Cerebral infarction Migraine Central nervous system vasculitis

Primary tumor Metastatic tumor Multiple sclerosis

a

Especially when the etiologic agent is herpes simplex virus. As a result of septic embolus (e.g., endocarditis). Or tuberculous abscess.

a

b

If abscess is in dominant hemisphere. b If abscess is in nondominant hemisphere.

c

TABLE 48-4. Clinical Manifestations of Focal Suppurative Infections of the Central Nervous System Septic Venous Thrombosis Signs or Symptoms

Brain Abscess (%)

Subdural Empyema (%)

Cranial Epidural Abscess (%)

Superior (%)

Lateral (%)

Cavernous (%)

Headache

60–70

60–80

100

70–90

80–90

75–90

Fever

40–60

80–90

80–90

60–70

75–90

70–85

Vomiting

50–60

60–70

20–30

50–80

60–80

30–50

Seizures

25–40

50–60

20–30

50–60

15–25

15–25

Mental status change

30–40

60–80

50–60

55–75

20–45

15–30

Coma

15–20

20–30

40 mg/dL) in 70% to 85% of cases. The yield of CSF cultures is also low (< 10%) except in patients with coexisting meningitis or with rupture of the brain abscess into the subarachnoid space. Recently, polymerase chain reaction amplification and sequencing of the 16S ribosomal DNA (which is highly conserved among bacteria) from brain abscess material provided rapid identification of the etiologic agent(s) which allowed a better choice of antibiotics.116 An electroencephalogram is modestly helpful in diagnosing brain abscess; low-frequency delta waves are the expected finding.1,72 The CSF findings in spinal epidural abscess or subdural empyema are also nonspecific, ranging from normal to moderate elevation of neutrophils and protein. Caution should be taken when lumbar puncture is done on a patient with soft-tissue symptomatology at the lumbar area, to prevent spread of the infection from the lesion into the CSF. In the presence of partial or complete spinal cord block, the CSF protein level is high (> 1000 mg/dL) with a few hundred WBCs (mostly lymphocytes) and normal glucose.

Imaging Studies The most important diagnostic test in patients suspected of having an intracranial suppurative process is CT or MRI. Since these tests have been used, the mortality rate from brain abscess has been reduced by 90% because lesions are diagnosed earlier and localized more

A

B

CHAPTER

48

331

accurately and surgical drainage is facilitated. CT manifestations of a brain abscess117 typically include a hypodense center, surrounded by ring enhancement, and another area of hypodensity that represents edema of the brain tissue adjacent to the abscess (Figure 48-2). Other lesions, such as tumor, granuloma, resolving hematoma, or an infarct, can have a similar appearance; cerebritis also shares common features. CT findings can lag clinical manifestations by a few days; initial normal CT does not exclude brain abscess. In addition, in patients receiving corticosteroids, the characteristic ring enhancement and edema can be decreased.118 High-resolution ultrasonography may be helpful in the early diagnosis of subdural empyema in infants and in differentiating this condition from subdural effusion.119 CT is also valuable in diagnosing subdural empyema. Although the noncontrast-enhanced CT can demonstrate the hypodense collection of pus, contrast-enhanced CT accentuates this finding by delineating the inflammatory response of the meninges or cerebral cortex (Figure 48-3). But MRI is superior to CT in demonstrating extra-axial fluid and rim enhancement.120 It is important to image the sinuses, middle ears, and orbits to identify the potential source of intracranial infection. Neither CT nor routine MRI can definitely distinquish subdural empyema for reactive subdural effusion (which is a more common complication of meningitis). Diffusion-weighted imaging has been shown to provide more definitive and specific information about the character of subdural fluid.121 CT of cranial epidural abscess shows a focal hypodense collection of pus between the dura and the calvarium. Contrast-enhanced CT shows an enhanced rim around the abscess, with displacement of the dura and early complications, such as the development of a subdural abscess (Figure 48-4). Ultrasonography has been successful in detecting superior sagittal sinus thrombosis in infants.122 CT with contrast enhancement is more helpful in diagnosing septic venous thrombosis by manifesting a filling defect in the thrombosed sinus. CT gives a better definition of the bony structures around the sinus but is inferior to MRI in detecting diminished venous flow. In the case of superior sagittal sinus thrombosis, demonstration of a hypodense triangular area surrounded by a hyperintense ring (the delta sign) by CT is specific for septic thrombus.59

C

Figure 48-2. Computed tomography in a child with headache for 6 weeks and brain abscess of the temporal lobe. (A) Computed tomography with contrast enhancement at diagnosis. Note the intense ring enhancement and edema. (B) Computed tomography scan without contrast enhancement following drainage of abscess in the patient from (A). Without contrast enhancement, the abscess cavity is much smaller (solid arrow) following surgical drainage. Marked edema is still noted (open arrow). (C) With contrast enhancement, note the abscess cavity with air, enhancement of the abscess wall, and marked edema. (Courtesy of M.A. Radkowski, M.D., Children’s Memorial Hospital, Chicago.)

332

SECTION

F

Central Nervous System Infections

A

B

Figure 48-3. Computed tomography in an 8-month-old infant with meningitis and seizures during the course of therapy. (A) There are bilateral extraaxial fluid collections in the frontal region (solid arrow), suggestive of an early subdural empyema. (B) One week later, there has been an increase in the extra-axial fluid collection with extension to the anterior interhemispheric fissure (open arrow) and enhancement of the membrane consistent with subdural empyema. (Courtesy of M.A. Radkowski, M.D., Children’s Memorial Hospital, Chicago.)

Figure 48-4. Computed tomography with contrast enhancement of a 14-year-old adolescent with sinusitis, orbital cellulitis, and continued fever during therapy. There is an extra-axial collection consistent with epidural abscess (arrow). Note ipsilateral proptosis and soft-tissue edema (orbital cellulitis, arrowhead). (Courtesy of M.A. Radkowski, M.D., Children’s Memorial Hospital, Chicago.)

MRI is considered by many to be the diagnostic test of choice for evaluating intracranial suppurative infections because it demonstrates soft-tissue details with a sensitivity and specificity superior to those of CT.123 Lesions can be detected earlier, and multiple smaller lesions undetectable by CT can be made apparent. Either can alter the therapeutic approach. If contrast-enhanced CT is done, both axial and coronal planes that extend through the frontal bone and the sellar region should be done so that the frontal, ethmoid, and sphenoid sinuses are seen. These cuts will help to avoid missing a suppurative process (e.g., epidural or interhemispheric abscess) that may not be seen with maxillofacial or orbital CT alone. The contrast-enhanced MRI may differentiate abscess fluid and CSF, allowing a more accurate evaluation of possible rupture of an abscess.124 If the conventional and diffusion MRIs are not sufficient to differentiate an abscess from a cystic tumor, they can be differentiated by using the perfusionweighted images which show the lesion’s vascularity. Because the abscess capsule is hyposvascular while the tumor’s capsule has a higher vascularization, the relative cerebral blood volume of these lesions is significantly different and helps in differentiation.125 Although MRI is better than CT in demonstrating the location of the venous thrombosis and the surrounding inflammation, the diagnosis is made by magnetic resonance with angiography (MRA), which more effectively demonstrates the absence of flow in the occluded vein. Analysis of amino acid resonance detected by 1H MR spectroscopy suggests that this test may aid in distinguishing bacterial brain abscess from necrotic brain tumor.126 MRI is also superior to CT for the diagnosis of spinal epidural or subdural abscess. The lesions are seen as isointense on T1-weighted images and hyperintense on T2-weighted images. The presence of concomitant diskitis, osteomyelitis, or paraspinal abscess is easily identified. In addition, MRI allows differentiation between an active inflammatory process and chronic granulation tissue, which may help the clinician make the diagnosis. MRI is also superior in differentiating the more common noninfected subdural effusions and

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Focal Suppurative Infections of the Central Nervous System

empyema, and in delineating between subdural and epidural abscess. This modality also offers advantages for management. Although a successfully treated abscess can still show enhancement on CT for up to 1 year, MRI changes resolve within 2 months. The introduction of CT- or MRI-guided stereotactic techniques revolutionized the diagnosis and management of intracranial suppurative infections.127 These techniques are minimally invasive and can be performed promptly with the patient under local anesthesia; the risks of hemorrhage or other complications are minimal. Aspiration of the lesion facilitates the diagnosis and identification of the etiologic agent, which results in a more specific antimicrobial therapy. Gram stain and culture for aerobic and anaerobic bacteria, fungi, and mycobacteria should be performed on aspirated specimens. Acid-fast, India-ink, and specialized immunohistochemical staining should also be performed in many settings.

MANAGEMENT Surgical Intervention Versus Medical Therapy Alone Brain Abscess Several studies have documented the success of antibiotic therapy alone in carefully selected patients with brain abscess.128 The patient must: (1) be alert, clinically stable, and without signs of increased intracranial pressure; (2) have an abscess 2 to 3 cm or less in diameter; and (3) have had symptoms for less than 2 weeks (probably representing the time sequence for development of a well-encapsulated abscess). It is important to note that CT can show ring enhancement even though cerebritis rather than an encapsulated, walled abscess is present. Thus, in patients with a short duration of symptoms, the high rate of success with nonsurgical management may be due to resolution of cerebritis, rather than of an encapsulated abscess. Identification of the specific pathogen (from blood, CSF culture, or CT-guided aspiration of the abscess) and knowing its antibiotic susceptibility are other important factors that underpin successful outcome of a nonsurgical approach. In certain situations, antibiotic therapy alone may be the only alternative. These situations could include: (1) patients with concomitant bacterial meningitis with a known etiology; (2) patients with multiple abscesses (aspiration of one abscess by CT-guided technique is important in identifying the pathogen in optimizing antibiotic treatment); (3) patients with deep-seated abscess or an abscess in a critical area that is not amenable to a safe surgical approach; and (4) patients who are poor surgical candidates (e.g., poorly controlled bleeding). Even among ideal candidates for nonsurgical therapy, failures occur, and patients must be followed carefully with frequent imaging studies (at least once a week during the acute stage). In patients who respond favorably to nonsurgical management, improvement seen by imaging study (i.e., reduced ring enhancement, less edema and mass effect) is generally noted within 1 to 2 weeks of initiating therapy. Reduction in abscess size should occur within 2 to 4 weeks. If the patient improves both clinically and by scan, intravenous antibiotics should be continued to complete a 4- to 6-week course of therapy. A shorter course of 3 to 4 weeks was found to be effective in patients who had surgical excision of the brain abscess in addition to antibiotic therapy.129 Decisions about the duration of therapy should be made on a case-by-case basis. Although up to 6 months of antibiotics given orally following intravenous therapy is sometimes advocated, such a long regimen is unnecessary. Follow-up imaging studies for 1 year (once every 3 to 4 months) is recommended; however, CT can show contrast enhancement for up to 1 year, even with successful resolution of the abscess. The role of corticosteroids in the treatment of brain abscess is controversial. There are two potential disadvantages. First, reduced ability of antibiotics to penetrate the blood–brain barrier can negatively affect the rate of bacterial eradication and the outcome. However, one

CHAPTER

48

333

study suggested that administration of corticosteroids did not increase mortality.71 In addition, no significant difference in patient outcome was found related to corticosteroid usage.130 Second, decreased ring enhancement on CT can be misinterpreted as improvement, or increased ring enhancement when corticosteroid is discontinued can be misinterpreted as a worsening condition. Thus, corticosteroids should only be used in patients with increased intracranial pressure and/or neurologic deterioration, when reduction of edema may be lifesaving. A short course (3 to 6 days) of high-dose dexamethasone (1 to 2 mg/kg per day divided q6 hours) is generally safe. Unless patients fulfill the previously discussed strict criteria for medical treatment alone, surgery is required to optimize outcome. Two surgical approaches should be considered: stereotactic aspiration or excision. Stereotactic aspiration of an abscess is generally safe, accurate, rapid, and associated with low morbidity and mortality, and is frequently successful. The procedure allows drainage of the abscess and a better choice of antimicrobial therapy, avoiding the potential hazards of empirical antibiotic therapy. The disadvantages of the procedure are the need (in some cases) for repeated aspirations, which can cause more tissue damage and bleeding and relative ineffectiveness of aspiration in posterior fossa lesions. Excision should be reserved for: (1) patients with fungal or helminthic abscesses who do not respond to medical therapy; (2) those with posterior fossa abscesses, especially comatose patients with signs of brainstem compression131; (3) patients with multiloculated abscesses in which aspiration has failed; and (4) those with reaccumulation of fluid following repeat aspirations of the abscess.

Extracerebral Abscesses Antibiotic therapy alone may be sufficient for patients with a relatively small subdural empyema (< 3 cm in diameter) or patients who are neurologically stable and rapidly respond to therapy.35,132 Appropriate antibiotic therapy and aspiration of subdural empyema related to meningitis are usually sufficient for identifying the pathogen and affecting positively the outcome of the disease. Unfortunately, empyema secondary to other causes (e.g., contiguous site infection, after penetrating trauma) is not easily drained with simple aspiration, and burrholes or craniotomy is needed. The location and extent of the empyema dictate the type of drainage procedure to be considered. Although craniotomy allows a better evacuation of pus from a larger empyema, the use of burrholes alone and irrigation can be effective.133 As little as 2 weeks of antibiotics intravenously was reported to be sufficient in patients with subdural empyema that was drained by craniotomy or burr holes and showed improved CT findings.134 Irrigation of the empyema with antibiotics is not recommended or supported by the data assessing this procedure. Drainage and debridement of a cranial epidural abscess, with appropriate antibiotic therapy, are usually recommended. In selected patients whose isolated epidural abscess complicates sinusitis and who have only a minimal mass effect, adequate sinus drainage combined with appropriate antibiotics is sufficient without the need for a neurosurgical procedure.135 Successful decompression of the abscess with CT-guided percutaneous aspiration and intravenous antibiotics has also been reported.136 Repeated MRI studies should be done to show that the abscess is resolving (reduced in size), and extension (subdural empyema) has not occurred. Two weeks of treatment may be necessary before a decrease in the abscess size is noted. Rarely, repair of a communication between the abscess and the primary source is needed. In the absence of ostemyelitis, 3 to 4 weeks of antibiotic therapy intravenously should be adequate; if osteomyelitis is present therapy should be continued for at least 6 weeks. Appropriate antibiotic therapy with surgical drainage of the source of infection is needed in most cases of septic venous thrombosis.105 The use of anticoagulant therapy is controversial because of the risk of hemorrhagic infarction.137 A recent review of the literature suggested that the risk of hemorrhage is low, and early treatment with anticoagulant agents is beneficial in patients without hemorrhagic complications by CT or MRI.138 Ligation of the internal jugular vein is not

334

SECTION

F

Central Nervous System Infections

necessary, but frequent MRI/CT monitoring to detect progression of thrombosis is important. Urgent drainage and appropriate antibiotic therapy are critical to halting progressive permanent neurologic deficit in spinal epidural or subdural abscess. Endoscopic surgical techniques should be used instead of open laminectomy.139 Antibiotic therapy alone was used successfully in patients with localized pain who did not progress to have root pain or spinal cord compression or in patients with stable condition (e.g., radiculopathy or signs of partial cord compression) for more than 3 days without deterioration.140 Frequent MRI studies should be done to document regression of the abscess. Antibiotic therapy should be given intravenously for 3 to 4 weeks, and extended to 6 to 8 weeks if osteomyelitis is present.

Antimicrobial Therapy Table 48-6 summarizes the recommended choices of initial empiric antibiotic therapy for focal suppurative infections of the CNS. Recommendations are based on the probability of the bacterial pathogens according to predisposing conditions, expected susceptibility patterns, and the ability of the antibiotic to achieve effective concentrations at the affected area. The penetration of the antibiotic into the suppurative area is affected by its molecular weight, ionization, protein binding, and lipid solubility. A summary of penetration of some antibiotics was published recently.141 Recommended antibiotic regimens are based on clinical experience and the medical literature. There are no controlled trials evaluating the relative efficacy of regimens; adults make up the subjects in most substantial case series. In many reports, surgical intervention is part of treatment, making assessment of the efficacy of a specific antibiotic regimen more difficult. Ceftriaxone (or cefotaxime) plus metronidazole is an appropriate empiric combination for brain abscess and subdural empyema associated with otitis media, mastoiditis, sinusitis, or CCHD. Ampicillin-sulbactam alone was reported to be effective in treating intracranial abscesses,71,142 but there was no comparison with other therapeutic combinations. Small numbers of patients were successfully treated with other antibiotics, such as imipenem,143 meropenem,144 and ciprofloxacin.145 With the increasing role of methicillin-resistant Staphylococcus aureus and S. epidermidis in focal CNS infections (as in cases associated with penetrating head trauma, ventriculoperitoneal shunts, or prosthetic valve endocarditis), vancomycin should be given empirically, plus a third-generation cephalosporin (and possibly metronidazole)

to cover other potential pathogens. Some experts use ceftazidime as the third-generation cephalosporin because Pseudomonas spp. are sometimes pathogens in chronic otitis media or sinusitis or following neurosurgery. In immunocompromised patients, a broad-spectrum antibiotic combination – for example, vancomycin plus ceftazidime plus metronidazole – is recommended for empiric therapy. If the initiating condition is meningitis, the recent increase in Streptococcus pneumoniae strains resistant to penicillin dictates treatment with vancomycin plus ceftriaxone. In neonates, ceftriaxone plus ampicillin is preferred because S. pneumoniae is rare, whereas Listeria monocytogenes is a potential pathogen. For patients with endocarditis that has developed on a natural valve, it seems logical to use high dosages of the same combination that is effective for the endocarditis itself (e.g., penicillin plus aminoglycoside), although the relative ineffectiveness of aminoglycosides in purulent collections should be considered. Therapy with a third-generation cephalosporin appears to be a good alternative. Nocardia infection is treated with the combination of trimethoprim-sulfamethoxazole (15 to 20 mg of trimethoprim/kg per day divided every 6 or 8 hours), although failures occur rarely. Regimens containing imipenem (e.g., plus cefotaxime or amikacin) or ceftriaxone (e.g., plus amikacin or minocycline) were reported to be effective in a few cases. Linezolid was recently reported to be effective against multiple Nocardia brain abscesses.146 The suggested treatment for fungal infections is shown in Table 48-7.

COMPLICATIONS AND PROGNOSIS Currently, fewer than 5% to 15% of children with brain abscess die.71,147 In one study, the mortality rate fell from more than 30% in the preimaging era to 13% after CT was introduced.71 A higher mortality occurs in patients with rapid onset (< 4 days), severe mental changes at diagnosis, or rapidly progressing neurologic impairment.130 In addition, mortality is higher in immunocompromised patients and those with intraventricular rupture of the abscess. The mortality rate of patients with intraventricular rupture who are aggressively managed was still > 38%.148 Patients with better outcome were younger than 21 years of age, with fewer complications from the rupture, and underwent aspiration of the brain abscess with ventricular drainage. Higher mortality rates are also associated with the presence of multiple abscesses and coma at diagnosis.1 About two-thirds of patients with brain abscess recover without sequelae. Seizures develop

TABLE 48-6. Empirical Antibiotic Therapy for Brain Abscess According to Predisposing Factors Predisposing Factor

Antibiotic (Dose)

MENINGITIS

Infants and children Neonates Cyanotic congenital heart disease Dental pathology Sinusitis and/or otitis mediad Penetrating trauma or postsurgery

Ceftriaxonea (100 mg/kg per day) + vancomycinb (60 mg/kg per day) Cefotaxime (200 mg/kg per day) + ampicillin (300–400 mg/kg per day) Ceftriaxone (as above) + metronidazolec (30 mg/kg per day) or Ampicillin-sulbactam (≥ 300 mg/kg per day of ampicillin) Penicillin (300,000 U/kg per day) + metronidazolec (as above) or Ampicillin-sulbactam (as above) Ceftriaxonee (as above) + metronidazolec (as above) or Ampicillin-sulbactam (as above) Vancomycin (as above) + third-generation cephalosporin

ENDOCARDITIS

Natural valve Prosthetic valve Unknown or immunocompromisedg

Ampicillinf (as above) + aminoglycoside (maximum doses) or Ceftriaxone (as above) + vancomycin (as above) Vancomycin (as above) + gentamicin Vancomycin (as above) + ceftazidime (see below) + metronidazole

a

Cefotaxime (200 mg/kg per day) can be substituted. Rifampin (20 mg/kg per day) can be substituted if local resistance of Streptococcus pneumoniae to this drug is low. c If anaerobic bacteria are resistant to metronidazole, change to meropenem (100–120 mg/kg per day) or chloramphenicol (80–100 mg/kg per day). d When staphylococci or penicillin-resistant Streptococcus pneumoniae is suspected, add vancomycin. e With chronic otitis media or sinusitis, change to ceftazidime (150–200 mg/kg per day). f Penicillin (300 000–400 000 U/kg per day) can be substituted. g If no response within 1 week and organism is unknown, consider amphotericin B. b

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Focal Suppurative Infections of the Central Nervous System

in 10% to 30% of survivors. (Seizures sometimes have onset several years after cure.)71,149 Patients with frontal or temporal lobe abscess, or a large abscess, have a higher incidence of seizures. Other neurologic sequelae include hemiparesis (10% to 15%), cranial nerve palsy (5% to 10%), hydrocephalus (5% to 10%), and behavioral and intellectual disorders (more serious in children younger than 5 years of age150). Less common sequelae include spasticity, ataxia, optic atrophy, and visual deficits. TABLE 48-7. Antifungal Therapy for Central Nervous System Focal Infections Pathogen

Therapy

Alternative Therapy

Candida spp.

Amphotericin B + flucytosine

Liposomal amphotericin or fluconazole

Aspergillus spp.

Voriconazole

Amphotericin B + flucytosine or liposomal amphotericin or caspofungina

Agents of mucormycosis

Amphotericin Bb Liposomal amphotericin + itraconazole or amphotericin + rifampina

Coccidioides immitis

Amphotericin B

Fluconazole or liposomal amphotericin or caspofungina

Cryptococcus neoformans

Amphotericin B + flucytosine

Liposomal amphotericin or fluconazole

Blastomyces dermatitidis

Amphotericin B

Liposomal amphotericin or fluconazole

Histoplasma capsulatum

Amphotericin B

Liposomal amphotericin or fluconazole

Pseudallescheria spp. Voriconazolea a

Itraconazolea or caspofungina

Limited experience. Hyperbaric oxygen may be used as adjunct therapy. Doses: Amphotericin B, 1.0–1.5 mg/kg per day; liposomal amphotericin, 5 mg/kg per day starting dose; fluconazole, 10–12 mg/kg per day; itraconazole, 10–12 mg/kg per day; flucytosine, 100–150 mg/kg per day; rifampin, 20 mg/kg per day; voriconazole, 4 mg/kg per day; caspofungin, 70 mg/m2 per day (maximum 70 mg/day).

b

A

CHAPTER

48

335

The morbidity and mortality from subdural empyema also improved significantly since the availability of CT, MRI, and improved surgical techniques.35 Only 1 of 56 children reported in five recent manuscripts had died.151 Mortality is lower if the patient is alert at presentation (< 10%); in comatose patients the prognosis is still poor (> 50% mortality). In patients (5% to 10%) who develop venous sinus thrombosis as a complication of subdural empyema, morbidity and mortality are increased (Figure 48-5). Despite improved survival, many patients (15% to 40%) have persistent neurologic problems, such as seizures or hemiparesis.35 The prognosis of uncomplicated (e.g., no extension to subdural empyema or brain abscess) intracranial epidural abscess is excellent. However, a delay in diagnosis can result in long-term neurologic sequelae (similar to those expected in patients with subdural empyema or brain abscess). The mortality from septic sinus thrombosis remains high (25% to 35%) despite improved diagnostic tools and antibiotic therapy.59 Mortality is even higher when the superior sagittal sinus is completely occluded; if only the anterior segment is involved, the potential for complete recovery is greater.59 Nearly 40% to 60% of patients with cavernous or lateral sinus thrombosis recover without sequelae.105 About one-fifth of patients with lateral sinus thrombosis have hearing loss, decreased visual activity, brain abscess, and, rarely, meningitis. One-third of patients with cavernous sinus thrombosis have blindness, oculomotor paralysis, hemiparesis, or pituitary insufficiency.152 The prognosis of patients with spinal epidural or subdural abscess is excellent if therapy begins before radicular symptoms appear. The overall mortality is less than 10%.153 The long-term outcome depends on the following: (1) the duration of paralysis before surgery is performed; (2) the degree of compression of the thecal sac; and (3) the age of the patient.154 The outcome is more favorable in children than in adults.70 Common neurologic sequelae include sphincter disturbance, spasticity, and paraparesis. Recently, the individual serum dynamic (time course) of the glial-derived protein S-100b was correlated with outcome.155 Patients whose serum S-100b levels normalized within 3 days postoperatively had a significantly better outcome (e.g., restoration of walking capacity) than those who had either a further increase or slower decrease.

B

Figure 48-5. A, T1-weighted magnetic resonance image following contrast enhancement shows thrombosis of the transverse (lateral) sinus. The grayish area (arrow) represents the thrombus itself. B, Three-dimensional, time-of-flight magnetic resonance angiography in the same patient shows complete blockage of the sinus, with extension of the thrombus into the superior sagittal sinus (arrows). (Courtesy of C. Darling, M.D., Children’s Memorial Hospital, Chicago.)

336

SECTION

CHAPTER

F

Central Nervous System Infections

49

Eosinophilic Meningitis Marian G. Michaels

The finding of even a few eosinophils in human cerebrospinal fluid (CSF) raises suspicion of certain pathologic states. Helminthic infestation of the central nervous system (CNS), particularly with the rat lungworm Angiostrongylus cantonensis, is the most common cause of eosinophilic meningitis worldwide.1–10 The differential diagnosis of CSF eosinophilia, however, especially in the United States, is broad and includes infestation by other parasites, especially the raccoon ascarid Baylisascaris procyonis,10–19 reaction to placement or infection of a ventriculoperitoneal shunt,20–23 medications,24,25 malignancies,26,27 hypereosinophilic syndrome,28 and an unusual manifestation of a more common fungal, bacterial, or viral infection of the CNS.29–32 Table 49-1 lists the various causes of eosinophilic meningitis. Eosinophilic meningitis is defined as: (1) at least 10 eosinophils/ mm3 in CSF; or (2) eosinophils making up at least 10% of the white blood cells in CSF.10 Diagnostic examination of CSF should always include Giemsa, Wright, or other stain of a cytocentrifuged specimen to delineate the exact composition of pleocytosis, as well as Gram stain, quantitative cell count, and biochemical tests.7,10

INFECTIOUS CAUSES Helminths

Angiostrongylus cantonensis The rat lungworm, A. cantonensis, is the most common cause of eosinophilic meningitis worldwide. Although no cause is predominant in the United States, it is important that clinicians consider parasitic infections and warn travelers to endemic areas about the risks of dietary indiscretions. This particular infestation has been the subject of several reviews.2–5,9 CNS migration is characteristic of the lungworm’s normal life cycle in the rat as well as the accidental human host. Adult worms live in the rat pulmonary arteries and lay eggs in the lung. After hatching, larvae make their way through the alveolar spaces and up the trachea, from which they are swallowed and then excreted in rat feces. Many species of slugs and snails serve as intermediate hosts, in which the larvae develop into infectious third-stage larvae. When third-stage larvae are ingested by rats, the definitive host, larvae migrate across the intestinal wall and are carried by the circulatory system to the brain. In the CNS, they undergo two more moltings and mature into adult worms; the worms return to the pulmonary arteries to renew the cycle. Human infection occurs after consumption of raw snails, shrimp, or fish that have fed on infected snails. Likewise, infectious larvae can be ingested accidentally on raw vegetables, raw vegetable juice, or fomites contaminated with snail slime.4,5,9,33,34 CNS tropism is retained in human infection; however, the life cycle is disrupted, and generally, the disease is self-limited, with larvae dying in the CNS. Human infection by A. cantonensis has been found in Southeast Asia, notably in Thailand and Malaysia; the South Pacific, including Taiwan, Hawaii, American Samoa, New Guinea, Indonesia, and Australia; India, Egypt, and the Caribbean.1–10,33–37 Worms can migrate on ships within their natural hosts (the rat) to distant countries, including mainland United States. The first reported, nonimported case in the United States was a child from New Orleans who reported eating a raw snail “on a dare.”38 The child had meningitis that abated with supportive care.

In adults, symptoms classically begin 2 to 35 days after infection, with acute onset of headaches.2–4 Other common complaints are weakness, nausea, and vomiting, paresthesias or hyperesthesias, somnolence, and cranial nerve palsies.2–4,39 Fever is usually low-grade if present. Hearing loss has been reported, as has retinal detachment or hemorrhage.40,41 Most patients recover completely, with symptoms beginning to abate within several weeks of the initial neurologic manifestations. Paresthesias may be more persistent.4 Infection in children may not cause headaches. Clinical manifestations in younger patients can be more insidious, with upper respiratory symptoms, cough, and fever preceding mental status changes, seizures, or focal neurologic abnormalities.3,8,42–44 One series of 16 young adults consuming raw infected great African Achatina fulica snails in American Samoa was notable for the absence of headache as a major complaint.45 All of the patients experienced severe radiculomyeloencephalitis; one died. Unlike other intermediate hosts or contaminated vegetables, A. fulica snails can contain thousands of parasites, possibly accounting for the more fulminant course in patients who acquire parasites from consuming these snails. Other reported fatalities or sequelae are primarily observed in young children, who are at risk for a relatively higher infectious load.3,43,44 CSF examination reveals a pleocytosis value typically between 150 and 2000 white blood cells per mm3.2,3,46 Peak eosinophilia occurs between 2 and 4 weeks of illness, with CSF eosinophils representing a median of 49% of the total white blood cells (range, 15% to 97%).2–8,39,43,44 The CSF protein value is often elevated, and the glucose value is normal or mildly decreased.2,3 Peripheral eosinophilia is common but does not correlate with the extent of CNS eosinophilia.2,3,5 Coincident infections with other parasites may contribute to peripheral eosinophilia. A diagnosis of eosinophilic meningitis is usually made on the basis of clinical presentation in patients who are from or are traveling from an area enzootic for A. cantonensis and have a consistent dietary history. Consumption of Caesar salad was strongly associated in a Jamaican outbreak.39 The history of eating raw seafood can be absent in young children who have a propensity for pica.8,42,43 On occasion, worms are found in the CSF, especially if a large-bore cannula is used for removal of CSF or if CSF is aspirated rather than allowed to flow by gravity.3 Enzyme immunosorbent assays to detect antibody in serum or CSF have been developed and are available in enzootic areas and can be facilitated by referral to the Centers for Disease Control and Prevention Disease Surveillance Division of Parasitic Infections (770-488-7775).4,5

Baylisascaris procyonis The ascarid parasite of the raccoon, Baylisascaris procyonis, is found in 20% to 90% of both rural and urban raccoons in the United States.12,47–49 Evidence of infected raccoon latrines near human habitation48 increases the concern for this emerging infection, which has been highlighted in several recent reviews.50,51 Like A. cantonensis, this ascarid migrates through the CNS during its normal life cycle and is neurotropic in humans.11,12,50,51 The ascarid causes little disease in the raccoon, but ingestion of eggs by aberrant hosts, such as foxes, rabbits, and birds (and humans), can result in migration of larvae, which cause severe CNS damage. Experimental disease in nonhuman primates consists of eosinophilic meningitis, neurologic deterioration, coma, and death (discussed in 12,50,51). Despite the high prevalence of this parasite in raccoons, B. procyonis has rarely been documented in humans. In the United States, just over a dozen clinical cases have been reported, most commonly in young children, resulting in eosinophilic meningoencephalitis with dramatic eosinophilic CSF pleocytosis, severe neurologic devastation, and, in 6 patients, death.11–13,50–54 Ocular larva migrans and visceral larva migrans can also occur.50,55 Exposure to raccoon droppings near nesting sites is the expected mode of acquisition. Because clinician awareness of this parasite is limited, and diagnosis is not straightforward, it is possible that the disease is more widespread and that less severe symptoms remain undiagnosed. Asymptomatic infection has been suspected by

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Eosinophilic Meningitis

CHAPTER

49

337

TABLE 49-1. Causes of Eosinophilic Meningitis (EM) Cause

Association with EM

NEMATODES (ROUNDWORMS)

Most common causes

Comments

Angiostrongylus cantonensis

Rat lungworm Snail transmission Most common cause of EM worldwide Not found in mainland United States Neurotropic Usually self-limited

Baylisascaris procyonis

Raccoon roundworm Found in United States Exposure to raccoon feces Neurotropic Prolonged, profound encephalitis

Gnathostoma spinigerum

Dog and cat roundworm Common in Southeast Asia Consumption of raw fish/poultry Peripheral eosinophilia common Visceral and CNS migration

Ascaris lumbricoides

Human roundworm Rarely associated with EM

Trichinella spiralis

Larvae usually found in skeletal muscle Rarely associated with EM

CESTODES (TAPEWORMS)

Occasional

Taenia solium

Pork tapeworm Ingestion of eggs from asymptomatic excretors Neurocysticercosis is a rare cause of EM

Echinococcus granulosus

Hydatid disease Rarely infects CNS and associated with EM

OTHER PARASITES

Rare

Toxoplasma gondii

Anecdotal report with congenital toxoplasmosis

Schistosoma japonicum

Acquired from contact with infected water

Paragonimus westermani

Lung fluke but can invade CNS and other sites

Fasciola hepatica

Found in Southeast Asia Rarely infects CNS

FUNGI

Occasional

Coccidioides immitis

Low level of CSF eosinophils common True EM also reported Diabetes is a risk factor

Cryptococcus neoformans BACTERIA AND VIRUSES

Treponema pallidum Mycobacterium tuberculosis Rickettsia rickettsii Lymphocytic choriomeningitis virus Ventriculoperitoneal shunt complications Bacterial shunt infections Reaction to shunt material NONINFECTIOUS ETIOLOGIES

Case reports

Occasional

Rarely associated with neurosyphilis Questionable association with CNS disease or treatment Rare with Rocky Mountain spotted fever Association made through serologic studies Can occur instead of neutrophilic pleocytosis Pathogens same as in shunt infections without eosinophils Implied by absence of infection and improvement with removal of shunt

Case reports

Systemic drugs CNS neoplasms Hypereosinophilic syndrome Immunologic/hypersensitivity reaction

Reported with quinolones and nonsteroidal anti-inflammatory drugs and intravenous street drugs Reported with Hodgkin lymphoma, acute lymphoblastic leukemia, and primary CNS tumors Reported in patients without other cause Reported with sarcoidosis or rabies vaccine

CNS, central nervous system; CSF, cerebrospinal fluid.

finding a low level of seropositivity in otherwise healthy children (reviewed in 50,51). As improved serodiagnostic assays become more readily available, the true incidence and clinical array will become more apparent. Currently serologic testing of serum and CSF is only available from the Department for Veterinary Pathobiology at Purdue University West Lafayette, IN (765-494-7558)

Gnathostoma spinigerum The nematode of dogs and cats, Gnathostoma spinigerum is commonly found in Thailand and other parts of Southeast Asia. It can cause visceral larva migrans when it infects humans who consume raw fish or poultry containing third-stage larvae.10,22 Gnathostomiasis is

338

SECTION

F

Central Nervous System Infections

most often characterized by migratory subcutaneous nodules. However, the larvae can wander to organs and cause local symptoms56 or enter the CNS via the nerve root, causing a myeloencephalitis.18,22,46,56 Patients can manifest disease over a period of a year with intermittent intense nerve root pain followed by paralysis or sudden deterioration in mental status.2,56 The CSF shows a striking eosinophilia but can often also have many red blood cells or xanthochromia.2 Frequently, G. spinigerum infection is associated with concurrent peripheral eosinophilia, which is pronounced.10,18,56 Infection can be severe and results in death more often than infection with A. cantonensis.2,7

Other Infectious Causes Other parasites can invade the CNS, including Taenia solium, Toxocara canis, Toxoplasma gondii, Paragonimus westermani, Fasciola hepatica, Trichinella spiralis, Ascaris lumbricoides, Echinococcus granulosus, Schistosoma japonicum, and Onchocerca volvulus. Associated CSF eosinophilic pleocytosis is characteristically mild.4,7,10,15,16,19 Bacteria have occasionally been associated with mild degrees of eosinophilic pleocytosis. The first descriptions of eosinophilic meningitis were in patients with neurosyphilis.7 Other implicated agents are Mycobacterium tuberculosis, group B streptococcus, and Rickettsia rickettsii.32,33,35,57 A retrospective review of 27 adults in the southwestern United States with Coccidioides immitis meningitis noted eosinophils in the CSF of 70% of patients; 30% of instances met the definition of eosinophilic meningitis.29 Other fungal infections of the CNS, such as those due to Histoplasma capsulatum or Cryptococcus neoformans, and Aspergillus species are occasionally associated with CSF eosinophilia.30,31 Rarely, viral infections of the CNS (lymphocytic choriomeningitis virus, coxsackieviruses B3 and B4, measles virus, echovirus 6, and herpes zoster virus) have been associated with eosinophilic pleocytosis of CSF.34–36 In some of these cases, diagnosis was suggested by results of serologic testing or by isolation of the virus from sites other than the CNS.34,36

Noninfectious Causes Review of series of children with complications of ventriculoperitoneal shunts have shown varying rates of eosinophilic pleocytosis associated with bacterial infection,22,35 intraventricular administration of antibiotics,25 antibiotic-impregnated ventriculostomy catheters,23 or without any other risk factor other than the shunt itself.20,21,58 Tung and colleagues21 found more than 8% eosinophils in the CSF of their patients after 36 of 558 (6.5%) shunt insertions. Affected children characteristically had experienced excessive numbers of shunt revisions and shunt infections. A study from Bulgaria likewise found that 6.4% of 404 children with ventriculoperitoneal shunts had transient CSF eosinophilia.20 However, in most of these children, the eosinophilia value was no more than 3% or eosinophils were present prior to placement of the shunt. Vinchon and associates22 reported finding eosinophilic meningitis associated with 27 of 81 presumed bacterial shunt infections; CSF culture results were positive in 63% of cases, and the microbes found did not differ from the types of organisms found in patients with noneosinophilic shunt infections. Persistent eosinophilic pleocytosis, failure to isolate infectious agents in many cases, and an association with eventual extrusion of subcutaneous tubing in some populations suggest that hypersensitivity to shunt material occasionally occurs. Replacement of the shunt with one of different material is sometimes necessary and sometimes curative. A recently reported case noted an association between the use of a minocycline and rifampin-impregnated ventricular drain in a youngster with a failed shunt with the development of eosinophilic meningitis.23 The eosinophilia resolved coincident with steroid administration and removal of the tubing. The authors speculate that the minocycline was

the likely cause of eosinophilic meningitis as it has been associated with eosinophilia in other organs. Several other noninfectious possibilities are included in the differential diagnosis of eosinophilic meningitis. Systemic drug administration has been associated on occasion with aseptic meningitis, and, more rarely, quinolones and nonsteroidal anti-inflammatory agents have been associated with eosinophilic meningitis.10,24 In addition, CNS involvement with Hodgkin primary lymphoma26 or acute lymphoblastic leukemia27 has caused eosinophilic meningitis. Other primary CNS tumors (as well as aseptic meningitis after resection of tumors) can be associated with mild eosinophilic pleocytosis.35 Patients with sarcoidosis and hypereosinophilia syndrome have been reported to have accompanying eosinophilic meningitis.28,35 Eosinophilic meningitis has also been described in two intravenous drug users who were negative for human immunodeficiency virus or other identifiable infections.59 The authors postulate that this might represent an eosinophilic response to systemically injected drug adulterants that were used as cut substances for street drugs.

DIAGNOSIS AND TREATMENT Careful attention to the clinical setting (i.e., occurrence, history of exposure or travel) and findings on physical examination almost invariably narrow possible causes of eosinophilic meningitis to one or two categories. Careful examination of large amounts of CSF for parasites and malignant cells and culture of CSF for fungus can yield a diagnosis. Examination of stool for ova and parasites is often unrewarding, because most human neurotropic parasites do not complete the life cycle in humans. However: (1) parasitic infestations often occur concurrently; and (2) on occasion, Ascaris lumbricoides, Schistosoma japonicum, Taenia solium, and Fasciola hepatica have been associated with eosinophilic meningitis. Therefore, stool evaluation is justified. Serologic tests of the CSF and serum for Cryptococcus neoformans, Angiostrongylus cantonensis, or Baylisascaris proycyonis, when epidemiologically indicated, are appropriate. Neuroimaging can be of benefit and can guide biopsy of the meninges or brain if other tests do not confirm a diagnosis. The treatment and prognosis of parasitic and infectious causes of eosinophilic meningitis or meningoencephalitis depend on the cause, underlying conditions, and extent of involvement at time of diagnosis. Recommendations for specific therapies can be found elsewhere in this book, in chapters addressing individual pathogens. No systematic study of treatment with antiparasitic agents has been undertaken for A. cantonensis, and treatment is primarily supportive. A multicenter study prospectively evaluated patients between 2 and 65 years of age with eosinophilic meningitis; 19% of patients were younger than 20 years. Treatment was not dictated by the study; however, no difference was noted in course or outcome among patients who received 5 days of corticosteroid therapy, antibiotics, or analgesics alone.2 Nonplacebo-controlled studies have reported the use of corticosteroids to be beneficial.37,60 Observations that clinical improvement occurs after lumbar puncture or is temporally related to use of steroid and mannitol suggest that some symptoms may be a consequence of the inflammatory response, raised intracranial pressure, or both.2,3,28 Children with severe eosinophilic meningitis due to Baylisascaris have not benefited from treatment; however, experimentally infected mice were protected from neurologic disease if treated with albendazole or diethylcarbazine prior to larvae entering the brain (reviewed in50). Because of the potential for severe, even fatal, outcome, particularly in young children, some experts recommend treatment of children exposed to raccoon excrement with albendazole to prevent disease.48,50,51 Educating the public about potential dangers from exposures to raccoons and their feces is the most important component of prevention. Raccoons have readily adapted to human habitats and are attracted to areas with food and water sources. Therefore their

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Transmissible Spongiform Encephalopathies: Slow Infections of the Nervous System

presence should be discouraged by cleaning up areas and ensuring that food is not left out unwittingly. Parents, of toddlers in particular, should be mindful of potentially contaminated play areas, encourage good handwashing after playing outside, and discourage pica.

CHAPTER

50

339

are scrapie of sheep, BSE, and a chronic wasting disease of North American deer, elk, and moose. All TSEs have a similar etiology and share some clinical manifestations and histopathology (Figure 50-1).

ETIOLOGY CHAPTER

50

Transmissible Spongiform Encephalopathies: Slow Infections of the Nervous System David M. Asher

Infectious agents cause several neurologic diseases, once considered degenerative, that have asymptomatic incubation periods of months to years and protracted clinical courses. Such diseases have been called slow infections.1,2 Some slow infections of the human nervous system are caused by viruses with conventional physical properties: encephalopathies caused by human immunodeficiency virus (acquired immunodeficiency syndrome (AIDS) encephalopathy), JC papovavirus (progressive multifocal leukoencephalopathy), measles virus (subacute sclerosing panencephalitis), and rubella virus (progressive rubella panencephalitis) are described elsewhere in this text. Other slow infections have unique etiologic agents with an array of physical properties so unlike those of conventional viruses that some authorities have concluded that they lack nucleic acid genomes; the diseases caused by these agents are now termed the transmissible spongiform encephalopathies (TSEs)3,4 or prion diseases.5,6 The TSEs (Table 50-1) include at least four diseases of humans: (1) kuru7; (2) Creutzfeldt–Jakob disease (CJD),8 which can be sporadic, iatrogenic, or familial6; (3) the Gerstmann–Sträussler–Scheinker (GSS) syndrome,9 which may be considered as a form of familial CJD; and (4) a new variant of CJD (vCJD)10,11 resulting from human infection with the agent of bovine spongiform encephalopathy (BSE).12 The fatal familial insomnia (FFI)13,14 and sporadic fatal insomnia syndromes15 – extremely rare syndromes differing from the other human diseases in several respects – are included in the same group. There are several TSEs of animals,3 the best known of which

The TSEs are slow infections transmissible to susceptible animals by inoculation of tissues from affected subjects. Kuru, all forms of CJD, and some cases of GSS have been transmitted from humans to a variety of primates16 and, less consistently, to rodents. Attempts to transmit FFI to primates from brain tissues of patients were un0successful, although transmission of disease from thalamic tissues to conventional mice has been reported.17 Transgenic mice engineered to express prion proteins (PrPs) of other species have been susceptible to infection with agents of human TSEs and BSE, providing convenient assays of infectivity.6 Although the infectious agents of the TSEs replicate in some cell cultures,18–20 they achieve much lower titers of infectivity than in brain tissue and elicit no recognizable cytopathic effects. Studies of the properties of the etiologic agents have generally employed in vivo assays of infectivity. Inoculation of susceptible animals with very small amounts of infectivity results in the accumulation in tissues of large amounts of an agent having the same physical and biologic properties as that in the inoculum. The self-replicating pathogens transmitting the infections display a spectrum of extreme resistance to inactivation by a variety of chemical and physical treatments that is unknown among conventional viruses. This characteristic stimulated the widely accepted “prion” hypothesis, that the TSE agents are composed of protein21,22 and devoid of nucleic acid.5,6 However, the prion hypothesis is not universally accepted, and some authors continue to suspect that the TSE agents may contain small amounts of pathogenic nucleic acid, perhaps surrounded and protected by hostderived PrPs.3,23–25 Studies purporting to demonstrate in vitro synthesis of abnormal infectious PrP generated in bacteria26,27 or in highly diluted infected rodent tissue after repeated enrichment with normal tissue extracts,28,29 although appearing to support the prion hypothesis, require confirmation.30 Abnormal proteins, first recognized by electron microscopy as scrapie-associated fibrils31 and then by polyacrylamide gel electrophoresis as PrP 27–30 (PrP27–30),32 are found in protease-treated detergent extracts from brains of patients and animals with TSEs. PrP27–30 is derived from a larger protein referred to as “scrapie-type” PrP(PrPSc), protease-resistant PrP (PrPres), or PrPTSE.4 PrPTSE

TABLE 50-1. Transmissible Spongiform Encephalopathies (TSEs) of Humans and Animals Disease Creutzfeldt–Jakob disease Sporadic Iatrogenic Familial Variant Gerstmann–Sträussler –Scheinker syndrome Kuru Fatal familial insomnia Sporadic fatal insomnia Bovine spongiform encephalopathy (“mad-cow” disease) Chronic wasting disease Scrapie Transmissible mink encephalopathy

Naturally infected hosts Humans

Humans Humans Humans Humans Cattle, other ungulates, zoo felines, domestic cats, humans American deer, elk, moose Sheep, goats Mink

Figure 50-1. Severe vacuolation (status spongiosus) in the cerebral cortex of a patient with familial Creutzfeldt–Jakob disease (hematoxylin and eosin stain). From Asher DM, Gibbs CJ Jr, Gajdusek DC. Subacute spongiform encephalopathies: slow infections of the nervous system. Clin Microbiol Newslett 1985;7:129.

340

SECTION

F

Central Nervous System Infections

molecules originate from a ubiquitous protease-sensitive form of the protein, called “cellular” PrP (PrPC). It is not yet clear whether PrPTSE constitutes the complete infectious particle of TSE agent (prion), is one component of a more complex particle, or is an important pathologic host protein not separable from the actual infectious agent by currently used techniques. Although PrPTSE is host-encoded and has the same primary amino acid sequence as the precursor PrPC widely expressed in normal tissues, many authorities are convinced that it is the etiologic agent of spongiform encephalopathies.6,33–35 However, PrPTSE has not generally been proven to be infectious,36,37 and purified PrP preparations that were infectious contain some residual nucleic acids.38 Besides the prion and virino theories, other hypothetical structures for the TSE agents have also been proposed,39 but none has gained acceptance. PrPTSE is a glycoprotein with the physical properties of an amyloid.6 PrP in humans and several species of animals, though not identical, is similar in amino acid sequence and antigenicity. The amino acid sequence of PrP is not influenced by the source of the infectious TSE agent provoking its formation; each host continues to express PrP with its original primary sequence. A portion of the PrPC of infected animals undergoes a posttranslational change in tertiary structure (folding) as disease progresses. Protease-resistant PrPTSE appears to be mainly folded into a b-sheet structure, while proteasesensitive forms of PrP are mainly a-helical.6 The normal function of PrPC is unknown; it binds copper40 and may play a role in synaptic transmission,41 but it is not required for essentially normal cerebral function, because mice lacking PrP live and behave normally.42,43 Whatever the relationship of PrPTSE to the actual infectious particles, PrP must be central to the pathogenesis of TSEs, susceptibility to infection, and replication of the agents. Transgenic mice in which the PrP-encoding gene was interrupted so that PrPC was not expressed resisted infection with scrapie, neither becoming ill nor supporting replication of the agent as assayed in ordinary mice.42–44

EPIDEMIOLOGY AND MECHANISMS OF TRANSMISSION Kuru once commonly affected many children and adolescents, as well as adults in an area of Papua New Guinea7; it now occurs in the elderly. In the years after it was first described, kuru was frequently recognized in male and female adolescents and children older than 4 years, while many more adult women than men had the disease. This pattern of occurrence probably reflects greater frequency and intensity of exposure to brain tissues during cannibalism by women and

children. Every patient with kuru has had a history of attending a cannibal ritual,45 in some instances more than 50 years before onset of clinical illness.46 Although many children were born to mothers with kuru, only children exposed to infected tissues by cannibalism contracted the disease.45 Taken together, the epidemiologic findings suggest that the practice of cannibalism, which ended before 1960, was the only mechanism by which kuru was spread. No evidence suggests either transplacental or milkborne transmission of kuru. Sporadic CJD, the most common human TSE, affects mainly older adults, with rates of 0.25 to 2 cases/million population per year in most surveys. The Centers for Disease Control and Prevention (CDC) have estimated the lifetime risk of CJD in the United States to be a little more than 1 in 10 000 persons (L. Schonberger, CDC, personal communication). Foci of considerably higher incidence of CJD – attributed to familial disease – occur among Libyan Jews in Israel, in isolated villages of Slovakia, and in other limited areas. CJD affects men and women in equal numbers; the mean age at onset in sporadic cases of CJD is about 60 years.16 Sporadic CJD has been recognized in several adolescents and young adults, but never in children less than 14 years of age.47 GSS and FFI have only been described in adults. Several hypothetical mechanisms for the source of sporadic CJD have been proposed, including contamination of meat products with scrapie agent or some other animal TSE agent, iatrogenic infection from tissues of patients, and spontaneous appearance of PrPTSE, or failure of the host to eliminate small amounts of PrPTSE that might appear normally. In March 1996, authorities in the United Kingdom first reported the occurrence of a new variant of CJD that differed in several respects from sporadic CJD (Table 50-2).10,11 That report realized a fear that the outbreak of BSE among cattle, zoo animals, and domestic cats – all probably infected by eating contaminated food48 – had also spread to people. As of August 2006 there have been 161 cases of definite or probable vCJD recognized in residents of the United Kingdom (where annual deaths attributed to vCJD peaked in 2002), as well as 31 cases in other countries (France, 17; Ireland, 4; Netherlands, 2; United States, 2; Canada, Italy, Japan, Portugal, Saudi Arabia, and Spain, 1 each). Some cases of vCJD, including all 3 in North America, affected persons who had either visited or lived in the United Kingdom and were probably infected while there; other cases – including 15 in France – were in persons who had never been to the United Kingdom and must have been exposed outside that area.4 The new variant of CJD has now affected persons as young as 12 and as old as 74 years; but the mean age at onset of vCJD (about 29 years) is substantially younger than for other forms of human TSE (about 61 years in the UK).10,11

TABLE 50-2. Comparison of Sporadic Creutzfeldt–Jakob Disease (CJD) and Variant CJD10,11 Sporadic CJD

Variant CJD

Mean age at onset

~61 years

29 years

Mean duration

4 months

14 months

Presenting sign

Confusion, sometimes ataxia

Abnormal behavior, dysesthesia, ataxia

Electroencephalogram

Slowing, usually periodic suppression burst pattern at some time

Slowing, usually no periodic complexes

Magnetic resonance imaging

Increased density in anterior basal ganglia

Increased density in pulvinar

PRNP genotype at codon 129

met/met ~80% (versus ~50% in general UK population)

met/met 100% (of clinical cases to date)

Amyloid (PrPTSE) plaques

~15% of patients (rarely, if ever, florid)

100% (florid, surrounded by vacuoles)

PrP core size and relative abundance of glycosylation formsa

Not BSE type

BSE type

Biologic properties in mice

Not BSE type

BSE type

TSE

BSE, bovine spongiform encephalopathy. a The size of the nonglycosylated PrPTSE molecule and relative abundance of nonglcosylated, monoglycosylated, and diglycosylated PrPTSE forms have been proposed as useful features to classify TSEs105, 106; not all authorities are convinced that each type of TSE is associated with a single unique PrPTSE “glycoform.”107,108

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Transmissible Spongiform Encephalopathies: Slow Infections of the Nervous System

Variant CJD provides the first documented example of an animal TSE infecting humans, reinforcing the recommendation to avoid human exposure to animal products likely contaminated with TSE agents. Twenty-three countries have recognized BSE in native cattle, including 2 cows in the United States49 and 7 in Canada50; agricultural and public health authorities in both countries have instituted precautions to reduce the risk of spread of BSE/vCJD to other animals and to humans. The CDC has estimated that, even during the peak of the BSE epidemic in the United Kingdom, where more than 180 000 cattle have been diagnosed with the disease, risk to travelers consuming beef products was small.51 In 1996 the United Kingdom completed the implementation of a number of steps to protect the food supply, and other European countries later instituted measures attempting to do so. Those steps, in addition to efforts to eliminate BSE from national cattle herds, have reduced the risk of human exposure to the BSE agent in food. Current information on the risk of BSE to travelers in various countries is available through the website of the CDC’s National Center for Infectious Diseases52 with links to other useful information about TSEs. Iatrogenic transmissions of CJD from infected human tissues have resulted from contaminated neurosurgical instruments or operating facilities, a contaminated cortical electrode used during epilepsy surgery, transplantation with contaminated corneas, grafts of cadaver dura mater, and injections of human pituitary growth hormone and human pituitary gonadotropin.11 Pharmaceutical products and grafts derived from or contaminated with human neural tissues, especially tissues obtained from unselected donors and from large pools of donors, pose special risks. Before 2003, human blood had never been convincingly incriminated in transmission of human spongiform encephalopathies, even among recipients of blood from a long-time donor who developed CJD.53 However, a few unconfirmed reports claimed to have transmitted infection to rodents injected with blood from humans with sporadic CJD54,55 and blood of rodents experimentally infected with rodent-adapted strains of CJD,56 GSS,57,58 scrapie,59 BSE,60 and vCJD61 was repeatedly found to contain small amounts of transmissible agent. Furthermore, reports of transmission of infection from the blood of sheep with experimental BSE and natural scrapie62,63 were troubling. In 1987 the Food and Drug Administration recommended deferring blood donors identified as at increased risk for CJD.64 Criteria for deferral were subsequently extended to persons residing in the United Kingdom and some other European countries.65,66 In 2003, vCJD was diagnosed in a recipient of nonleuko-reduced red blood cells from an apparently healthy donor from the United Kingdom who became ill with the disease 3 years after the donation.67 Two additional vCJD infections, attributed to transfusions with nonleuko-reduced red blood cells from two asymptomatic donors from the United Kingdom, have subsequently been reported.68,69 One of these cases was of particular concern, because it was only diagnosed by testing for PrPTSE in lymphoid tissues of a person who died without a clinical diagnosis of vCJD. The patient had a genotype heterozygous for methionine and valine at codon 129 of the PRNP gene, a genotype not previously associated with vCJD.4 The incubation periods of CJD for two of the blood recipients were 6 and 8 years. Spouses and household contacts of patients with CJD seem to be at low risk of contracting disease, although conjugal CJD has been reported.1,16 Medical personnel exposed to brains of patients with CJD may be at increased risk: a neurosurgeon, a pathologist, two histopathology technicians, and at least 16 other healthcare workers have been recognized with the disease.1 CJD has not occurred in children born to affected mothers.

PATHOLOGY Typical histopathologic changes of spongiform encephalopathy include vacuolation and loss of neurons (see Figure 50-1), with hypertrophy and proliferation of glial cells. Changes are most pronounced in the cerebral cortex in patients with CJD and in the cerebellum in those with kuru. The lesions are typically most severe in

CHAPTER

50

341

gray matter. Loss of myelin occurs secondary to degeneration of neurons. There is no inflammation. Astrocytes increase in number and size. Amyloid plaques are found in the brains of all patients with GSS and vCJD, in at least 70% of patients with kuru, and less often in those with other forms of CJD. Plaques are most common in the cerebellum but occur elsewhere in the brain as well. The plaques react with antisera prepared against PrP, and in areas of the parenchyma without plaques, extracellular PrP can be detected by immunostaining. A unique diagnostic finding in vCJD has been the consistent presence of unusual “florid” (flower-like) amyloid plaques surrounded by vacuoles. Although accumulations of PrPTSE70,71 and infectivity16 are detectable in tissues outside the central nervous system, no histopathologic changes have been recognized. The fatal insomnia syndromes13–15 are characterized by severe selective atrophy of anterior ventral and mediodorsal thalamic nuclei, usually without spongiform changes.

PATHOGENESIS The portal of entry for the kuru agent is thought to be the integument, probably through lesions rather than intact skin and mucosa. It is not known whether humans can be infected with the agent of sporadic CJD through the intact intestinal tract. The probable portal of entry for most cases of vCJD is the gastrointestinal tract, although direct intravascular inoculation by blood transfusion is a possible source. The first site of replication of the agents may be in tissues of the reticuloendothelial system. In experimentally infected mice the scrapie agent also spreads to the central nervous system by ascending peripheral nerves. The only portal of exit of the kuru agent from the body, at least in quantities sufficient to infect others, seems to have been through tissues exposed during cannibalism.45 In iatrogenically transmitted CJD, the brains and eyes of patients with CJD were sources of contamination. In sporadic cases of CJD, the sources of infection and portals of entry are unknown. Kidney, liver, lung, lymph node, spleen, and cerebrospinal fluid (CSF) have sometimes contained the infectious agent in sporadic CJD16 but their role in pathogenesis and person-to-person spread of infection is unknown. In vCJD, the BSE agent seems most likely to have been carried by beef products – especially by mechanically recovered meat contaminated with bovine brain, spinal cord, or ganglia, which contain the bulk of infectivity.72 vCJD infection has also spread by transfusions of blood from donors incubating disease. At no time during the course of any spongiform encephalopathy have antibodies or cell-mediated immunity to the infectious agents been convincingly demonstrated, in either patients or animals.

CLINICAL MANIFESTATIONS Kuru is a progressive degenerative disease of the cerebellum and brainstem with less obvious involvement of the cerebral cortex.7 The first sign of kuru is usually cerebellar ataxia followed by progressive incoordination and coarse shivering tremors. Variable abnormalities in cranial nerve function appear, with frequent impairment in conjugate gaze and swallowing. Patients die from inanition and pneumonia or accidents, usually less than a year after onset. Although mental changes are common, there is no frank dementia or progression to coma as in CJD. Signs of acute encephalitis – fever, headaches, and convulsions – are absent in kuru. Patients with CJD initially have either sensory disturbances or confusion and inappropriate behavior, progressing over weeks or months to frank dementia and coma. Some patients have cerebellar ataxia early in disease, and most eventually develop myoclonic jerking movements. Mean survival of patients with CJD is less than 1 year from the earliest signs of illness, although about 10% live for more than 2 years. Patients with vCJD often come to medical attention with subtle mental changes, first confused with psychiatric disorders, and have often complained of painful sensations in the extremities – a finding not typical of other forms of CJD. Patients with vCJD typically survive for more than a year after onset of illness.10,11

342

SECTION

F

Central Nervous System Infections

GSS is a familial encephalopathy resembling familial CJD but with more prominent cerebellar ataxia and amyloid plaques.9 Dementia may appear only late in the course. Patients with GSS can survive longer than those with typical sporadic CJD. Patients with the two fatal insomnia syndromes, in addition to sleep disorder, have progressive autonomic insufficiency and motor dysfunction13–15; some also have ataxia and myoclonus.

LABORATORY FINDINGS Patients with CJD usually have abnormal electroencephalogram (EEG) findings as disease progresses; the background becomes slow and irregular with diminished amplitude. A variety of paroxysmal discharges may also appear: slow waves, sharp waves, spike-and-wave complexes. These can be unilateral or focal as well as bilaterally synchronous. Paroxysmal discharges may be precipitated by loud noise. Many patients with sporadic CJD have typical periodic suppression burst complexes of high-voltage slow activity on EEG at some time during the illness; those complexes have rarely been observed in vCJD.73 Magnetic resonance imaging (MRI) may show cortical atrophy with large ventricles late in the course of CJD. The distribution of abnormally increased density on MRI differs between CJD and vCJD. Anterior basal ganglia are most affected in sporadic CJD, whereas the pulvinar is more affected in vCJD.74 In all forms of CJD there may be modest elevation of CSF protein concentration late in disease. Detection of several normal brain proteins, especially the 14-3-3 protein,75 although not unique or specific for CJD, sometimes helps to discriminate CJD from Alzheimer disease. The predictive value of CSF analysis is improved when other normal neuronal proteins (s100, neuron-specific enolase, and tau proteins) are also detected.76 PrPTSE has not been detected in CSF of patients with TSE.77

DIAGNOSIS AND MANAGEMENT Diagnosis of spongiform encephalopathies is most often made on clinical grounds after excluding other diseases. In older adults, the most difficult differential diagnosis is between CJD and the more common Alzheimer disease. Brain biopsy, while often diagnostic of CJD, is only recommended if some other potentially treatable disease cannot otherwise be excluded; biopsy of the palatine tonsil, looking for accumulations of PrPTSE, has been useful to diagnose vCJD but routine use has been questioned.78 Definitive diagnosis of TSE usually requires examination of brain tissue obtained at autopsy. The demonstration of PrPTSE in brain extracts confirms the histopathologic diagnosis. Transmission of disease to susceptible animals by inoculation of brain suspension has been the ultimate diagnostic test for spongiform encephalopathy, although this is reserved for cases of special research interest. CDC’s National Center for Infectious Diseases79 provides useful advice to state public health authorities and healthcare providers. The National Prion Disease Pathology Center, Division of Neuropathology, Case Western Reserve University offers assistance with laboratory diagnosis.80 No treatment has been demonstrated to be effective for any TSE. Experimental therapeutic approaches are currently under investigation18,20,81,82 but most published animal studies83 and clinical trials84,85 have not yielded encouraging results. Appropriate supportive care should be provided. The prognosis for spongiform encephalopathies is uniformly poor.

PREVENTION, CONTAINMENT, AND DISINFECTION Medical and family histories of potential tissue donors should be carefully reviewed to exclude those with dementia suggestive of a spongiform encephalopathy. Universal precautions should be used to handle blood and body fluids of all patients. Materials and surfaces contaminated with tissues or CSF from patients suspected of having CJD must be treated with great care. Whenever possible, contaminated instruments should be discarded by careful packaging and transported to sites of incineration. No practical method can be guaranteed to

remove all infectivity from contaminated surfaces.86–88 Sodium hydroxide, chlorine bleach, and steam autoclaving (with certain limitations) have been recommended by World Health Organization consultants89 and CDC90; concentrated formic acid, a proprietary phenolic disinfectant,86,91 a mild acidified detergent,92 and other treatments have also been reported to reduce TSE infectivity substantially. To enhance the effects of disinfectants, all potentially contaminated surfaces should first be carefully cleaned.93,94 Contaminated tissues and most biologic products probably cannot be completely freed of infectivity without destroying their structural integrity or functional activity.

GENETICS AND COUNSELING Spongiform encephalopathies occur in some families in a pattern consistent with an autosomal-dominant mode of inheritance. In patients with a family history of CJD the clinical and histopathologic findings are the same as those seen in sporadic cases. In the United States about 10% of CJD cases are familial. GSS and FFI are always familial. In some families affected with CJD, about 50% of siblings and children of an affected individual eventually develop the disease95; in other families the penetrance is less. The PrP-encoding gene is closely linked to that controlling the incubation periods of scrapie in sheep and both scrapie and CJD in mice.6 The gene encoding PrP in humans, currently designated the PRNP gene, is located on the short arm of chromosome 20. It has an open reading frame of about 250 codons. Twenty-four point mutations were identified in the PRNP gene as of the year 2000, and many of those, as well as a variety of inserted sequences encoding extra tandem repeated octapeptides and a deletion, have been linked to spongiform encephalopathies in families. The most common mutations encode a proline-to-leucine change at PRNP codon 102, associated with GSS, or a glutamine-to-lysine change at PRNP codon 200, associated with familial CJD. The same nucleotide substitution at codon 178 of the PRNP gene associated with FFI in some families is linked to CJD in other families; however, the mutations at PRNP codon 178 in the two syndromes are linked to different amino acids at codon 129 – a normal polymorphic codon in the general unaffected population. Subjects with iatrogenic CJD, sporadic CJD, and kuru are more often homozygous for alleles in PRNP codon 129, especially methionine, than are unaffected subjects.96 However, heterozygosity at PRNP codon 129 is not protective against TSE. Thus far, all genotyped patients with clinical vCJD have been homozygous for PRNP-129 methionine. However, one of the three transfusion-transmitted infections with vCJD was detected in a person heterozygous for methionine and valine at that codon. A United Kingdom survey of anonymous tissue samples retained after surgery revealed that two of three appendices containing PrPTSE, derived from more than 12 000 samples tested, were from persons homozygous for PRNP-129 valine.97 Thus it appears that all persons must be considered potentially susceptible to infection with vCJD and that a substantially larger number of people may have been infected in the United Kingdom than was previously suspected.98,99 In affected families with an autosomal-dominant pattern of CJD, GSS, or FFI, subjects with a mutation in the PRNP gene often have a high probability of developing spongiform encephalopathy. One TSEassociated mutation was successfully detected in utero.13 The implications of finding mutations in PRNP genes of subjects from families with no history of spongiform encephalopathy are unknown.

OTHER DEGENERATIVE DISEASES OF THE CENTRAL NERVOUS SYSTEM POSSIBLY CAUSED BY UNCONVENTIONAL AGENTS It was reported that two other human diseases might be caused by infections with agents similar to those causing the spongiform encephalopathies – familial Alzheimer disease of adults100 and Alpers disease, a convulsive disorder associated with hemiatrophy and status spongiosus of the cerebral gray matter presenting in young children.101 Attempts to confirm these reports have failed.102–104

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

SECTION

G

Genitourinary Tract Infections

CHAPTER

51

Urinary Tract Infections Jacob A. Lohr, Stephen M. Downs, and Theresa A. Schlager

Urinary tract infections (UTIs) are one of the most common bacterial infections in children, especially young girls, and are the most common nephrologic problem seen by pediatricians.

ETIOLOGIC AGENTS Enterobacteriaceae are the most common causes of uncomplicated UTI, Escherichia coli being responsible for 70% to 90% of overall and > 90% of first UTIs in outpatients.1,2 Although E. coli is the most common cause of infection in children with underlying urinary abnormalities, other microorganisms, including Klebsiella spp., Proteus spp., Enterococcus spp., Pseudomonas spp., and Enterobacter spp. may be prevalent.3 Staphylococcus saprophyticus, a coagulasenegative staphylococcus, accounts for 15% or more of UTIs in female adolescents.4–6 Group B streptococci are unusual urinary tract pathogens but are occasionally isolated from the infected urine of neonates and adolescents.7 S. aureus, including methicillin-resistant S. aureus (MRSA), rarely cause pyelonephritis or cystitis in outpatient practice. Recovery of S. aureus from the urine suggests an additional site of infection, such as renal abscess, osteomyelitis, bacterial endocarditis, or another important bloodstream infection (BSI).8 Gastrointestinal pathogens such as Salmonella, Shigella, and Campylobacter occasionally infect the urine. Haemophilus influenzae type b and anaerobes are rare urinary tract pathogens.9,10 Lactobacillus spp., Corynebacterium spp., and a-hemolytic streptococci are common periurethral flora; under most circumstances, their isolation from urine represents contamination. Mycobacterium tuberculosis infection of the urinary tract is uncommon but should be considered when sterile pyuria is present. Protozoa, significant in some areas of the world, are rare uropathogens in the United States. Enterobiasis can be associated with entry of pinworms into the urethra, which can lead to symptoms (dysuria, frequency) and pyuria suggestive of a UTI. Acute hemorrhagic cystitis, most commonly caused by Escherichia coli and adenovirus types 11 and 21. Adenovirus cystitis is a benign self-limited disease of childhood suggested by the presence of gross hematuria associated with dysuria, urgency, and increased frequency of urination. Nosocomial UTI, largely due to gram-negative organisms and yeasts, remains a significant cause of morbidity in hospitalized pediatric patients. Urinary catheterization of > 3 days is an important risk factor for infection.11 MRSA outbreaks have been reported in intensive care units with isolation of MRSA from blood, nasopharynx, and urine of infected neonates.12

EPIDEMIOLOGY Reported prevalence of UTI is influenced by patient selection, method of urine collection, and laboratory tests used for diagnosis. Lower

prevalence rates are observed if only symptomatic children are evaluated. Falsely high rates of infection are reported if urine samples are collected by application of a bag to the periurethral skin, a method known to have a substantial rate of bacterial contamination. The most important variables influencing prevalence of infection are age and sex. In neonates, the rate for premature infants (2.9%) and very-lowbirthweight infants (4% to 25%)13 exceeds that for full-term infants (0.7%).14 Young boys are 5 to 8 times more likely to be infected than girls.15,16 Male preponderance persists for the first 3 months of life, after which the prevalence rate among females exceeds that in males.17 The prevalence rate reported in girls 1 to 5 years of age is 1% to 3%, whereas few infections occur in boys of those ages.18–20 This is the age range in which children are most likely to experience a first symptomatic infection. Symptomatic infections occur 10 to 20 times more commonly in preschool-aged girls than in preschool-aged boys.21 The prevalence of bacteriuria in school-aged girls (0.7% to 2.3%) exceeds that among boys (0.0% to 0.2%).22–26 In one study that included 4910 girls and 7731 boys in grades 1 to 12, a positive urine culture, without symptoms of infection, was 30-fold more common among girls.22 Prevalence of UTI is lower in school-aged African American girls than in white girls.22 At least 5% of girls have one or more episodes of bacteriuria by age 18.27 Febrile infants and children commonly have UTI, with rates of infection inversely proportionate to age. In four separate studies, specific rates of UTI were 7.5% in 442 febrile episodes in infants younger than 8 weeks28; 5.3% in 945 febrile episodes in those younger than 1 year of age29; 4.1% of 501 episodes in children younger than 2 years30; and 1.7% of 664 episodes in children younger than 5 years.31 The recurrence rate for UTI in girls is substantial regardless of the presence or absence of a urinary tract abnormality. The greatest risk of recurrence is during the first few months after an infection.32 In the United States, approximately 75% to 80% of white and 50% of African American school-aged girls have recurrence of UTI within 3 years of the first infection.32,33 Recurrences are less frequent in males, affecting approximately one-third of those with a UTI.32

PATHOGENESIS Colonization of periurethral mucosa with gastrointestinal bacteria is the initial event in UTI.34–38 Periurethral bacteria can then ascend into the bladder, ureters, and kidneys by undefined mechanisms, establishing a risk for bladder urine or renal parenchymal infection. Presence of pathogens on the periurethral mucosa does not necessarily result in infection.39 Despite the heavy colonization of periurethral mucosa of infants younger than 1 year,40,41 most of them do not experience UTI.

Bacterial Factors Our evolving knowledge of UTI suggests a combination of microbe– host interactions. E. coli isolates from urine specimens in healthy children with UTIs are more likely to derive from E. coli phylogenetic groups B2 and D,42 exhibit a more limited number of O, K, and H antigens43–46 and express specific bacterial properties or “virulence factors” than isolates from stool. Virulence factors are thought to enable bacteria to evade host defense mechanisms and to induce an inflammatory response leading to disease of the urinary tract. E. coli 343

344

SECTION

G Urinary Tract Infections

clones encoding combinations of virulence factors have been associated with enhanced ability to infect the urinary tract. Other uropathogens known to express virulence factors include Proteus spp., Klebsiella spp., Serratia spp., and Staphyloccus saprophyticus.47 Virulence factors, however, may not be necessary for the development of UTI in compromised hosts.48,49

Host Factors The ability to empty the bladder completely and regularly is the most important host defense against infection.50,51 Obstruction to normal flow of urine with urostasis is one of the most important factors predisposing to UTI. Children with anatomic abnormalities, neurogenic bladder, detrusor-sphincter dyssynergia,52 obstruction from calculi,53 instrumentation of the urinary tract, or extrinsic compression from tumors or constipation54,55 are all at increased risk for UTI. Circumcision reduces the frequency of UTI in boys by almost 10fold.56 Host factors that may increase the risk of UTI in adolescent patients include diabetes, frequent sexual intercourse, pregnancy, and human immunodeficiency virus with CD4 lymphocyte count < 200/mm3.57,58 Genetically controlled blood group antigens on the surface of uroepithelial cells may affect bacterial adherence by acting as receptors or, conversely, by blocking bacterial attachment. Patients prone to UTI show a higher density of epithelial cell receptors and individuals of blood group P1 have an increased risk of developing recurrent pyelonephritis.59 Receptor expression may also control which fimbrial type can cause infection. Individuals of blood group A1P1 express the globo-A receptor and become infected with bacteria recognizing this receptor.44 Interleukin-8 chemokines and chemokine receptors have been shown to be necessary for bacterial clearance and tissue integrity in a murine UTI model. Children with at least one episode of pyelonephritis had low chemokine receptor expression compared with controls in one study.59 Urine is an excellent culture medium for most Enterobacteriaceae. In contrast, urine does not usually support the growth of normal periurethral flora (Lactobacillus and Corynebacterium spp.)60 or anaerobic organisms.10 The rate and amount of microbial growth

depend on the pH, tonicity, concentration of urea, and presence of dietary-derived organic acids in the urine.61

Factors Affecting Upper Tract Infection Infected bladder urine can ascend to the upper urinary tract and, possibly, the renal parenchyma. Vesicoureteral reflux (VUR), an important risk factor for the development of pyelonephritis, is found in 30% to 50% of children with UTI.62 VUR is usually congenital, resulting from deficiency of the longitudinal muscle of the submucosal ureter, with shortening of the intramural portion of the ureter as it traverses the bladder wall. Closure of the ureter with bladder filling and micturition is restricted, facilitating retrograde flow of urine from the bladder into the ureter and pelvicaliceal system (Figure 51-1). Urodynamic dysfunction, probably related to inadequate relaxation of the external urethral sphincter, is an increasing accepted cause of VUR, especially in infant males.63 Grades I, II, and III VUR usually resolve spontaneously by 5 years of age with elongation of the intramural portion of the ureter. Grades IV and V VUR resolve spontaneously in only about 40% of affected children.64 A population-based cohort study of 1221 children followed prospectively after their first UTI found that boys had primary congenital, VUR-associated renal scarring and girls had acquired scarring related to recurrent febrile UTI.65 Fetally diagnosed uropathy secondary to VUR is often bilateral high-grade reflux, predominantly in males.66,67

CLINICAL MANIFESTATIONS Clinical manifestations of UTI are highly variable, ranging from asymptomatic to fulminant systemic infection. Fever or temperature instability, decreased activity, poor feeding, and jaundice are common nonspecific findings among infected neonates.68 Infants often have failure to thrive, “feeding problems,” vomiting, and fever.69 Symptoms and signs of respiratory or gastrointestinal infection can predominate31,70; nearly 5% of respiratory syncytial virus-positive infants may also have a UTI, possibly related to decreased flow due to

Figure 51-1. Diagrams illustrating variations within grades I to V of vesicoureteral reflux (VUR). (A) Grade I: VUR does not reach the renal pelvis; (a, b, c) different degrees of ureteral dilation. (B) Grade II: VUR extends up to the renal pelvis without dilation; (a) and (b) incomplete filling of the ureter and calyces. (c) Complete filling of the ureter and calyces. (C) Grade III: VUR extends up to the kidney. (a) Mild dilation of the ureter, renal pelvis, and calyces with no blunting of the calyceal fornices; (b) moderate dilation of the ureter and renal pelvis, and mild ureteral tortuosity with no blunting of the calyceal fornices; (c) mild dilation of the ureter, moderate dilation of the renal pelvis, and slight blunting of the calyceal fornices. (D) Grade IV: Moderate dilation of the ureter with complete obliteration of the sharp angles of the calyceal fornices, but the papillary impressions are visible. (a) Moderate dilation of the renal pelvis and calyces with complete obliteration of the sharp angles in majority of fornices; (b) moderate tortuosity of the ureter and moderate dilation of the renal pelvis; complete obliteration of the sharp angles of all fornices; (c) moderate tortuosity of the ureter and extensive dilation of the renal pelvis, although papillary impressions are visible in the majority of calyces. (E) Grade V VUR. (a) Moderate dilation of the tortuous ureter, moderate dilation of the renal pelvis, and a papillary impression visible in only one of the calyces; (b) gross dilation of a tortuous ureter, renal pelvis, and calyces, and no papillary impressions visible; (c) extreme dilation of the whole upper urinary tract. From Johnston JH (ed) Management of Vesicoureteral Reflux. Baltimore, Williams & Wilkins, 1984, pp 97–109. PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Urinary Tract Infections

dehydration.71 Under 2 years fever occurs with most UTIs in boys and girls. Above age 2 years most girls with UTI are afebrile, and UTI in boys is unusual.72 Abdominal pain may be common in children 2 to 5 years of age.69 Bedwetting in a child who has previously exhibited nighttime bladder control is an important symptom, particularly in school-aged girls.73 Foul-smelling urine can also be a sign of UTI. The rates of dysuria, urgency, and costovertebral angle tenderness increase in frequency in patients older than 5 years with UTI.69 It is not possible to differentiate cystitis (lower tract infection) reliably from pyelonephritis (upper tract infection) on the basis of clinical presentation in children; as many as 25% of children without symptoms of pyelonephritis are found by ureteral catheterization or bladder washout test to have renal bacteriuria.63 The presence of high fever is an imprecise indicator of upper tract involvement.74–77 Some 1% to 2% of older infants and children have bacteriuria that is transient and of no apparent clinical significance.23–26

Differential Diagnosis In addition to UTI, disorders to consider in children with bacteriuria include urethritis, vaginitis, cervicitis, prostatitis, foreign body, nephrolithiasis, renal (intrarenal or perinephric) abscess, vaginovesical fistula, and enterovesical fistula.

Clinical Approach History In addition to identification of symptoms consistent with UTI, the history should be reviewed for previous UTI, family members with a history of UTI,78 undiagnosed febrile episodes, foreign bodies, trauma, and sexual activity.

Physical Examination An accurate blood pressure measurement should be obtained. The urethral meatus is examined, although meatal abnormalities (including obstruction by phimosis) that lead to UTI are rare. The presence of suprapubic tenderness and costovertebral angle tenderness should be assessed, but these symptoms are only moderately sensitive and specific for UTI.75,76

Urine Collection Improper collection of specimens leads to misdiagnosis, overtreatment, and inappropriate radiographic evaluation. Bag specimen collection is not appropriate for diagnosis of UTI at any age because of the substantial rate of contamination.14 However, a urine specimen collected by the bag technique that yields no growth rules out a UTI, unless the patient is receiving antibiotics. Suprapubic aspiration is the most diagnostically accurate technique for obtaining a urine specimen for culture.79,80 The procedure has a low rate of complications but may not successfully obtain a specimen.81 Use of a urethral catheter also provides a reliable specimen for culture.82 The technique is simple,80 but the catheter must be carefully inserted past the curved portion of the posterior urethra in boys to avoid mucosal trauma. These two methods of urine collection have low contamination rates, thus avoiding the risks for overdiagnosis, overtreatment, and unnecessary evaluation associated with bag collection. Spontaneously voided urine may be used to evaluate older infants and children with suspected recurrent infections if they are known to have structurally normal urinary tracts. Cleansing of the urinary meatus and collection of a midstream urine specimen, in both girls and circumcised or uncircumcised boys, do not appear to reduce the frequency of contamination substantially.83–85

CHAPTER

51

345

LABORATORY FINDINGS AND DIAGNOSIS Immediate Tests Urine specimens should be processed immediately; if immediate processing is not possible, specimens should be refrigerated at 4°C and processed within 24 hours. Incubation of urine at room temperature for as little as 1 hour results in loss of sensitivity of the leukocyte esterase test and loss of specificity of the nitrite tests. Results of urinalysis can suggest the presumptive diagnosis of UTI, but results should not be used to confirm the diagnosis of UTI.86 The leukocyte esterase test, nitrite reaction, and microscopic examination for white cells and bacteria in both unstained and Gram-stained urine specimens are often performed. The sensitivities of the leukocyte esterase test (76% to 85%) and nitrite reaction (29% to 70%) are too low to allow these tests to be useful in screening for UTI.86 However, when the nitrite reaction is performed with a freshly voided specimen, a positive result is highly predictive of UTI.87 Testing fresh urine for the presence of leukocyte esterase and microscopic examination for white blood cells are highly correlated; therefore, it is not necessary to perform both tests. Negative microscopy for bacteria on unstained or Gram-stained urine specimens (performed by certified technologists with processing initiated within 10 minutes of collection), when combined with a negative “dipstick” leukocyte esterase test result, has high negative predictive value.87 This combination of negative tests performed under careful conditions can virtually rule out UTI in children beyond the neonatal period. Similarly, the use of a counting chamber to demonstrate fewer than 10 white blood cells per mm3 and a Gram stain showing no bacteria effectively rule out UTI.88,89 Urine microscopy requires additional time and a higher level of technical skill and experience than the dipstick test.

Culture Quantitative culture of properly collected urine specimens establishes the diagnosis of UTI and provides an isolate for susceptibility testing. Certain biochemical tests (e.g., filter paper method) and simple culture techniques (e.g., dipslide) are popular screening tests. Quantitative cultures are performed on both nonselective (e.g., sheep blood agar) and selective (e.g., MacConkey agar) media inoculated with a calibrated loop that delivers an inoculum of 0.01 mL (for a catheter/suprapubic specimen) or 0.001 mL (for a voided specimen). The urine is streaked over the entire agar surface to permit quantification of colony-forming units (CFU). The CFU count is multiplied by 100 (when a 0.01-mL loop is used) or by 1000 (when a 0.001-mL loop is used) to yield CFU/mL. Interpretation of the significance of growth on the agar plate or plates depends on the following variables: CFU/mL of each species isolated, number of species isolated, method of urine collection, sex of the patient, recent antibiotic therapy, timing of urine specimen collection-to-inoculation or refrigeration, time of day of specimen collection, dilution of urine by fluid intake, and level of clinical suspicion of UTI. The presence of more than 105 CFU/mL urine is a widely held standard for diagnosis of UTI. This standard is based on Kass’s observations,90,91 made nearly 50 years ago, that the urinary bacterial count in adult women with clinical diagnosis of pyelonephritis was > 105 CFU/mL. Lower counts, found in asymptomatic patients, were considered to represent contaminants. In women with the acute urethral (“dysuria–pyuria”) syndrome, Stamm and associates92 documented bladder infection in culture of suprapubic aspirate or urethral catheter specimens in patients for whom concurrent midstream cultures had colony counts as low as 102 CFU/mL. It has not been established that counts this low indicate infection in children. Hoberman and colleagues93 have proposed that colony counts of 50 000 CFU/mL or greater from specimens of catheterized urine obtained from children younger than 2 years with fever were most characteristic of infection.

346

SECTION

G Urinary Tract Infections

Most children with normal urinary tracts are infected with a single organism. Therefore, the presence of more than one type of organism suggests specimen contamination and another specimen should be obtained. The method of urine collection, Gram stain character of the isolate, sex of the patient, and quantiÀcation of growth were considered in the criteria compiled by Hellerstein94 for diagnosis of UTI in children (Table 51-1). In a later publication, Hellerstein emphasized that the criteria for diagnosis are operational rather than absolute.95 Timing of the specimen collection is important; Àrst-voided morning specimens optimize sensitivity because pathogens have had the opportunity to incubate overnight in the bladder. The amount of fluids taken before urine collection influences the CFU/mL value for the cultured specimen because of dilution effect. Recent receipt of an antibiotic can decrease the yield of culture.

Other Tests The usefulness of multiple tests (e.g., C-reactive protein and erythrocyte sedimentation rate) for determining the site of the UTI have been studied; however, no noninvasive test has demonstrated a high degree of accuracy. Invasive tests such as bladder washout, renal biopsy, and ureteral catheterization are not appropriate in the usual clinical setting of UTI. Renal cortical scintigraphy offers promise for identifying the patient with parenchymal involvement but has limitations (see below).

MANAGEMENT Management of UTI consists of antibiotic therapy, imaging, and careful follow-up. Hospitalization may be indicated for children with suspected UTI if they are very young infants or have a toxic appearance, severe dehydration, vomiting, or intolerance of oral medication.

TABLE 51-1. Criteria for Diagnosis of Urinary Tract Infections by Culture Method of Collection

Colony Counta

Probability of Infection (%)

SUPRAPUBIC ASPIRATION

Any gram-negative bacilli; > 103 grampositive cocci

> 99

CATHETERIZATION

> 105 104 to 105 103 to 104

95 Infection likely Suspicious; repeat culture needed Infection unlikely

3

< 10 CLEAN-VOIDED

Boy Girl

> 104 > 105 in 3 specimens > 105 in 2 specimens > 105 in 1 specimen 5 × 104 to 105 104 to 5 × 104 104 to 5 × 104 < 104

a

Infection likely 95 90 80 Suspicious; repeat culture needed If symptomatic, suspicious; repeat If asymptomatic, infection unlikely Infection unlikely

Colony-forming units/mL of single isolate. From Hellerstein S. Recurrent urinary tract infections in children. Pediatr Infect Dis J 1982;1:271.

Antimicrobial Therapy The choice of antimicrobial agent is based on the most likely pathogen and is adjusted after results of culture and susceptibility tests are known. Initial parenteral therapy for hospitalized infants and children is ampicillin (100 to 200 mg/kg per day intravenously, given every 6 hours) and an aminoglycoside (e.g., gentamicin, 6 mg/kg per day intravenously, given every 8 hours or, as an option, given once daily96). Patients with underlying urinary tract abnormality or complicated history (frequent recurrences, unusual pathogens, catheter use, or immunodeÀciency) may require a more individualized antibiotic regimen. Duration of parenteral therapy depends on clinical response and is based on expert opinion; parenteral therapy is generally continued for at least 3 days until blood culture result is deÀnitively negative or for at least 24 hours after fever and other symptoms have resolved, whichever is longer. Approximately 10% of infants are not afebrile at 48 hours of therapy; fever > 48 hours does not correlate well with underlying conditions or complications.97 A total of 10 days of therapy is usually appropriate unless the course was complicated (e.g., BSI or central nervous system infection, positive urine culture after 48 hours of therapy, circumstances noted later). Hoberman and colleagues98 have demonstrated the equivalence of oral therapy and parenteral therapy in children 2 to 24 months of age with febrile UTI who did not otherwise require hospitalization. Effective oral antibiotics are available for outpatient treatment. Selected cephalosporins (e.g., ceÀxime, 8 mg/kg orally given once daily) are appropriate agents; their ineffectiveness for enterococcal infection must be kept in mind, as should cephalosporin resistance in patients receiving prophylactic antimicrobial therapy.99 Nitrofurantoin (5 to 7 mg/kg per day orally given every 6 hours) remains an effective agent for treatment of uncomplicated infections due to susceptible agents. The established resistance of Escherichia coli to amoxicillin (> 40%) and trimethoprim-sulfamethoxazole (> 20%) limits the use of these drugs. Patients with underlying urinary tract abnormalities are more likely to be infected with organisms more broadly resistant to antimicrobial agents. The duration of oral therapy is controversial. Many studies in adults with lower UTI and older children have evaluated the effectiveness of short-course therapy (single-dose, 1-day, 3-day, or 4day). Recurrence rates after single-dose or 1-day therapy are higher than those after conventional therapy.100,101 The outcome of 3- or 4-day therapy is variable.100–102 At least 10 days of therapy is recommended for children with presumed pyelonephritis (infants, toxic appearance, high fever, costovertebral angle tenderness, urinary white blood cell casts, or a combination of these Àndings), VUR, or other urinary tract abnormalities and for those who have not yet been evaluated by imaging studies.

Imaging Studies The role of diagnostic imaging in the evaluation of young children with UTIs has been debated. Imaging studies can detect congenital or acquired abnormalities of the urinary tract, including posterior urethral valves in boys, VUR, obstruction, dysplasia, hydronephrosis, other congenital anomalies, renal scarring, and renal stones. Most authorities agree that imaging studies are indicated after a Àrst documented UTI in any boy or girl younger than 5 to 6 years.103–107 Abdominal ultrasonography is the most useful modality for evaluating kidneys, ureters, and bladder. Ultrasonography can deÀne renal structure (size, shape, position) and detect dilatation of the collecting system; it does not provide functional information or reliably demonstrate renal scarring. Cystography (voiding cystourethrogram (VCUG) or radionuclide cystogram (RCG)) deÀnes the presence and grade of VUR; a VCUG with fluoroscopy is used to detect posterior urethral valves in boys. VUR is associated with recurrent UTI108 and with progressive renal

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Renal (Intrarenal and Perinephric) Abscess

scarring, particularly in the presence of recurrent UTI.109–112 However, progressive scarring appears to be unusual after 5 years of age.113 Renal cortical scintigraphy using technetium 99-labeled dimercaptosuccinic acid (DMSA) can suggest the presence of: (1) acute pyelonephritis, which is associated with focal or diffuse areas of decreased cortical uptake of tracer without any loss of volume; or (2) renal scarring, which is associated with areas of decreased uptake accompanied by loss of volume. A positive DMSA scan is associated with increased risk of subsequent scarring, particularly if VUR is present.114–116 However, a clear role for DMSA scan in the evaluation and management of UTI is not established. For the child with presumed renal involvement (high fever, toxic appearance, costovertebral angle tenderness, urine white cell casts, or a combination of these findings) in whom urinary tract imaging studies have not been performed in the past, abdominal ultrasonography is performed early in the course to identify pyonephrosis or abscess. Cystography is performed after the urine is sterile. For the child with a less complicated presentation and an early response to therapy, the performance of ultrasonography and cystography can be delayed until the urine is sterile. For initial cystography, a VCUG with fluoroscopy is performed in boys to rule out posterior urethral valves and to detect VUR, which is frequently due to dysfunctional voiding rather than anatomic abnormality. Urethral obstruction is rare in girls, so visualization of the urethra is not necessary. An RCG can be used for evaluation of VUR in girls and for follow-up evaluation of boys without posterior urethral valves. If findings of either the initial cystogram or abdominal ultrasonography are abnormal, consultation with a pediatric radiologist or pediatric urologist helps determine the necessity for additional studies, including renal cortical scintigraphy.

Follow-Up Careful follow-up of the child who has had a documented UTI is essential. A urine culture of a specimen obtained 48 hours after initiation of therapy confirms antibiotic effectiveness but is unnecessary when the patient shows good clinical response and the infecting organism is susceptible in vitro to the antibiotic chosen.117 A urine culture of a specimen obtained 48 hours after completion of therapy can identify relapses and early reinfections; this culture is of little use if antibiotic prophylaxis has been initiated. The utility of routine follow-up cultures in asymptomatic children is doubtful, unless uropathy is present. Patients with frequent recurrences (> 3 UTIs per year) should receive prophylactic antibiotics for at least 6 months, especially in the presence of uropathy or if recurrences are symptomatic.118,119 The drugs of choice are nitrofurantoin (1 to 2 mg/kg orally qhs) and trimethoprim-sulfamethoxazole (2 mg trimethoprim and 10 mg sulfamethoxazole per kg orally qhs).120 If no infection occurs during the period of prophylaxis, antibiotics can be discontinued. If infection then recurs, prophylaxis should be reinstituted. Long-term prophylaxis in children with VUR is associated with higher rates of resolution of VUR than is individual treatment of each infection.121 Children with severe VUR (grades IV and V) should be managed in conjunction with pediatric subspecialists in pediatric urology, nephrology, or both. This condition can be managed medically or with surgery. To date, there has been no demonstrated benefit of one approach versus the other.122 The choice of treatment is determined by the age of the child, expectation of adherence to the management plan, and the availability of follow-up care. Most VUR resolves over a decade.123

COMPLICATIONS In infants with UTI, the frequency of associated BSI is inversely related to age, occurring in 18% of those 1 to 3 months of age and in 6% of those 4 to 8 months of age.17 Meningitis can result from BSI,

CHAPTER

52

347

especially in infants younger than 3 months. BSI is unusual in children older than 1 year. Unusual complications of UTI include the development of a renal abscess or stones. Early eradication of infection appears to prevent renal scarring and progressive renal dysfunction.124,125 Renal scarring can lead to renal failure with consequent growth failure, hypertension, and renal complications during pregnancy. The risk for these complications after recurrent UTI is unknown, but is heightened in children with obstructive uropathy and high-grade VUR.

CHAPTER

52

Renal (Intrarenal and Perinephric) Abscess Jacob A. Lohr and Sara M. O’Hara

Renal abscess is rare in children, and clinical manifestations are variable. Thus, precise characterization of the typical patient is difficult. Risk factors for acquiring renal abscess and proposed pathogenic mechanisms have evolved over time. Morbidity and mortality are substantial. Sonography and computed tomography (CT) have led to earlier diagnosis and a more precise anatomic delineation as well as allowing a less invasive therapeutic approach. The terminology describing abscess of the kidney or the surrounding tissues has been inconsistent and ambiguous.1

TERMINOLOGY In this chapter, the term acute pyelonephritis describes acute bacterial infection of the kidney, without suppuration; intrarenal abscess describes a collection of purulent material within the kidney (Figure 52-1); perinephric (perirenal) abscess describes an abscess outside the kidney but within the renal (Gerota) fascia; and renal abscess includes both intrarenal and perinephric abscess.

Liver

Pyonephrosis

Multifocal bacterial nephritis (acute form of renal corticomedullary abscess) Pancreas

Renal (Gerota) fascia Renal Perinephric Paranephric capsule abscess abscess Renal cortical abscess (renal carbuncle) Figure 52-1. Anatomy and sites of focal infections related to the kidneys.

348

SECTION

G Urinary Tract Infections

Familiarity with certain terms used most commonly in the past is pertinent as well. Renal cortical abscess (renal carbuncle) describes an infection that results from a coalescence of multiple abscesses originating in the renal cortex (see Figure 52-1). Focal pyelonephritis, also known as focal bacterial nephritis and lobar nephronia, is an acute corticomedullary phlegmon characterized by a leukocytic inÀltrate with focal areas of tissue necrosis but without liquefaction, conÀned to a single renal lobe. Acute multifocal bacterial nephritis is a frank abscess. Xanthogranulomatous pyelonephritis, a chronic corticomedullary abscess, is a granulomatous process in which the affected parenchyma is replaced by lipid-laden macrophages, resulting in yellow discoloration of renal tissue. Stones are usually present, and there is renal dysfunction. This entity is rarely included in reports of renal abscess. Pyonephrosis (suppurative hydronephrosis) occurs when stagnant urine in a dilated urinary tract becomes infected and purulent. A paranephric abscess resides in the region of the kidney but, unlike perinephric abscess, is located outside Gerota fascia.

ETIOLOGIC AGENTS The etiologic agents for renal abscess are listed in Table 52-1.2–18 In the preantibiotic era, 80% of intrarenal and perinephric abscesses resulted from hematogenous seeding by staphylococci.19 Staphylococcus aureus continues to be an important cause of both intrarenal and perinephric abscesses during childhood.4,13 Enterobacteriaceae also constitute an important cause, especially if the abscess is a complication of urinary tract infection. Other gram-negative and gram-positive organisms play lesser etiologic roles, as do fungi.18 Brook17 reported isolation of anaerobic bacteria of oral or gastrointestinal origin from renal abscesses in 9 of 10 children; polymicrobial infections were described, but a single organism was recovered from most of the abscesses.17,20 Mycobacterium tuberculosis is a rare but important cause. Occasionally, cultures obtained from renal abscesses are sterile.2–18

EPIDEMIOLOGY AND PATHOGENESIS The exact incidence of renal abscess in children is unknown; approximately 250 cases of perinephric abscess13 and fewer than 200 cases of intrarenal abscess4–8 have been reported. All age groups are affected, and there is no gender predominance.2–21 Renal abscess occurs most often in otherwise healthy children, but certain identiÀable risk factors are enumerated in Box 52-1. Although diabetes mellitus is frequently noted as a pre-existing condition for renal abscess in adults, it is rarely an associated condition in children.

TABLE 52-1. Causative Agents of Renal Abscess Pathogen

Frequency

Staphylococcus aureus Enterobacteriaceae (especially Escherichia coli and including Salmonella spp.) Pseudomonas spp. Enterococcus spp. Coagulase-negative staphylococci Streptococcus spp. Actinomyces spp. Anaerobic organisms Fungi Mycobacterium tuberculosis

++ ++ + + + + + + +a +

++, frequent cause; +, infrequent cause. a In patients with catheter-related, disseminated, or bladder candidiasis, focal renal infection is relatively common. Data collated from references 2–18; based on 90 patients, 1 month to 19 years of age.

Renal abscess can occur in immunocompetent and immunocompromised hosts. Intrarenal abscess can result from hematogenous spread or as a complication of ascending infection from the lower urinary tract. The affected kidney was previously normal or abnormal (e.g., dysplastic or hydronephrotic). Hematogenous infection is usually caused by invasion of the bloodstream by S. aureus from the skin, another site of infection, or as a spontaneous infection. Experimental evidence suggests that microabscesses form Àrst, with subsequent coalescence leading to a larger focus of infection.22 Intrarenal abscess that follows urinary tract infection is usually caused by a gram-negative bacillus. Vesicoureteral reflux is the most frequently associated urinary tract abnormality. Pyonephrosis can be followed by development of an abscess, especially if a urinary tract obstruction due to stones or congenital abnormality is present. Focal bacterial nephritis can progress to renal abscess.23 Intrarenal abscess can remain conÀned to the renal parenchyma, even with substantial enlargement, or can extend into the perinephric space or the renal pelvis. Most perinephric abscesses are caused by gram-negative bacilli. Organisms gain access to the perinephric space by direct extension from an intrarenal abscess, or infection occurs in association with vesicoureteral reflux or urinary tract obstruction following urinary tract or abdominal surgery. Perinephric abscesses can also result from hematogenous seeding by S. aureus from a distant site. Unusually, perinephric abscess follows an intraperitoneal infection such as a ruptured retrocecal appendix. Rarely, an adjacent infected hematoma is found. A perinephric abscess in a renal transplant patient whose regimen included sirolimus has been reported.24

CLINICAL MANIFESTATIONS AND DIFFERENTIAL DIAGNOSIS The average duration of symptoms before diagnosis ranges from less than 1 week to approximately 3 weeks.4,6,13,15–18 NonspeciÀc symptoms, such as malaise, lethargy, decreased appetite, weight loss, nausea, and vomiting, are typically associated with fever and pain. Among 61 patients 1 month to 18 months of age with renal abscesses reported in the literature, fever occurred in 89%; pain in the flank or abdomen or tenderness in the costovertebral angle was present in 85%.2–13,16–18 Because most abscesses are unilateral, pain or tenderness is usually

BOX 52-1. Conditions that Increase Risk for Renal Abscesses URINARY TRACT CONDITIONS Infection Anomalies (reflux, obstruction, duplications) Neurogenic bladder Urinary tract stones Tumor Polycystic disease Peritoneal dialysis PRIMARY INFECTION ELSEWHERE WITH BACTEREMIA Skin and soft tissue Dental Respiratory tract Gastrointestinal tract Intra-abdominal Cardiac Genital Intravascular catheter-related Intravenous illicit drug-related SURGERY Urinary tract, including transplantation Intra-abdominal OTHER ImmunodeÀciency states Trauma in area of kidney Diabetes mellitus

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Renal (Intrarenal and Perinephric) Abscess

unilateral. However, pain can be referred to the back, periumbilical area, or hip. When a renal abscess is preceded by urinary tract infection, dysuria or frequency is common. A palpable mass occurs in about 5% of cases and is more likely to be found in an infant. Other findings are: (1) scoliosis with splinting of the affected side; (2) pain on bending to the contralateral side; and (3) chest abnormalities, such as decreased respiratory excursion, tenderness over the lower ribs, and pulmonary dullness, decreased breath sounds, and rales on the affected side. The most likely diagnosis in a febrile patient with symptoms or signs referable to the urinary tract is acute pyelonephritis. The differential diagnosis of a unilateral renal mass in a neonate includes hydronephrosis or another obstructive anomaly, multicystic dysplastic kidney, mesoblastic nephroma, and renal vein thrombosis. In an older child, Wilms tumor, other tumors, hematoma, hydronephrosis, and other anomalies are considered. The diagnosis of renal abscess should be considered in the following instances: (1) failure of response of presumed pyelonephritis to therapy; (2) fever without an identifiable source after urinary tract or abdominal surgery or dialysis; (3) fever and urinary tract obstruction; (4) fever after trauma to the area of the kidney; (5) fever and pain in the flank or abdomen, or tenderness at the costovertebral angle; (6) unilateral renal mass; and (7) fever of undetermined origin.

LABORATORY AND IMAGING STUDIES

52

349

acute pyelonephritis who have atypical findings or are unresponsive to antibiotic treatment (Figure 52-2). However, ultrasonography may not permit distinction between abscess, phlegmon (nephronia), and uncomplicated pyelonephritis. Doppler ultrasonography is useful to demonstrate absence of blood flow in a liquefied abscess or a centrally necrotic tumor. Abscesses classically show hyperemia in well-defined walls on Doppler ultrasonography. Contrast-enhanced magnetic resonance imaging (MRI) and CT show similar devascularized nonexcreting areas replacing normal renal parenchyma.27,28 Renal cortical scintigraphy, gallium-67 citrate scanning, and technetium 99-labeled leukocyte scanning occasionally reveal focal masses in the kidney, although additional cross-sectional imaging is typically necessary to characterize the inflammatory process further. Additional case examples of renal abscess can be accessed at www.PediatricRadiology.com and other linked sites.

MANAGEMENT Management of the patient with an intrarenal or perinephric abscess must be individualized. Medical management alone with a trial of intravenous antibiotic therapy is suggested by some.29 Others believe that medical therapy alone cannot be justified in children.30 If medical management is selected and is not successful, percutaneous drainage, open surgical drainage, or nephrectomy, which can now be performed laparoscopically, may be necessary.

A peripheral white blood cell count, erythrocyte sedimentation rate or C-reactive protein, urinalysis, and culture of urine and blood (multiple specimens) are obtained; cultures of skin lesions, wounds, respiratory tract, and other sites can be useful in selected patients. Once a renal abscess is identified, specimen for Gram stain and culture should be obtained by aspiration or at the time of surgery. Culture and stain for aerobic and anaerobic bacteria, fungi, and mycobacteria should be performed as indicated. Causative agents from 90 patients with renal abscess are shown in Table 52-1. Results of other tests in 91 patients with renal abscess are shown in Table 52-2. Imaging studies are useful for detection, localization, characterization, and guidance in aspiration and for follow-up of focal infections of the kidney.25,26 Ultrasonography is the usual initial examination. It should be performed in individuals presumed to have

A

CHAPTER

TABLE 52-2. Laboratory Findings Associated with Renal Abscess in 91 Patients 1 Month to 19 Years of Age Test Result (No. of Patients Tested)

No. of Patients with Positive Test Result (%)

White blood cells > 12 000/mm3 (68) Elevated erythrocyte sedimentation rate (26) Microscopic pyuria (55) Positive urine culture (69) Positive blood culture (35) Positive abscess culture (83)

59 (87) 24 (92) 25 (46) 36 (52) 12 (34) 78 (94)

Data collated from references 3–6, 8–18, and 21.

B

Figure 52-2. (A) Diagnostic ultrasonogram of renal abscess in the left kidney of a 5-year-old, ill-appearing girl with a 5-week history of fever and abdominal pain. She originally had a urinary infection with Escherichia coli and was treated with multiple oral antibiotics. The upper medial portion of the kidney (arrows) is more bulbous and anechoic compared with the rest of the kidney, which is normal in appearance. (B) Enhanced computed tomogram (CT) obtained concurrently. A large renal abscess on the left is seen as a large black oval area (arrow) representing nonenhancement of the liquified center of the abscess. There is a smaller abscess on the right (arrow).

350

SECTION

G Urinary Tract Infections

Initial antibiotic therapy should include agents active against Enterobacteriaceae and S. aureus. A penicillinase-resistant penicillin such as nafcillin (or vancomycin in the setting in which methicillinresistant S. aureus is prevalent) plus an aminoglycoside is a suitable combination. The addition or substitution of an antibiotic with anaerobic activity, such as metronidazole, clindamycin, or ticarcillinclavulanic acid, should be considered, especially in the patient with underlying chronic obstructive disease. Adjustment of therapy is based on the results of culture of abscess specimens. Therapy for the spectrum of potential pathogens should not be narrowed on the basis of blood or urine culture results alone, because they do not always correlate with isolates from abscess specimens.17,20 Many patients have a negative urine culture.2–18A clinical response should be expected in 48 to 72 hours. Ten to 14 days of parenteral therapy followed by 2 to 4 weeks of oral therapy is usually sufficient.31 Percutaneous aspiration for diagnosis, culture, and cytologic studies is usually performed under ultrasonographic guidance (Figure 52-3).32,33 Therapeutic drainage, more often indicated with perinephric abscess or when obstruction or severe reflux exists, can be performed at the same time without significant morbidity, averting the need for open drainage and general anesthesia. Early aggressive drainage is recommended in immunocompromised patients.34 The drainage catheter can be removed when drainage ceases and resolution is evident on ultrasonography. Open surgical drainage is used when antibiotic therapy and percutaneous drainage do not result in a favorable clinical response. Previously, open surgical drainage was used when an abscess ruptured into an adjacent space (intrarenal abscess into the perinephric space or the renal pelvis or a perinephric abscess into the paranephric area); currently, a percutaneous approach can usually provide adequate drainage for the expanded abscess. Nephrectomy is reserved for the patient with massive abscess in whom function of the involved kidney is unlikely to be preserved. Patients recovering from a renal abscess should undergo renal ultrasonography serially to document progress, but it must be recognized that resolution of the ultrasonographic abnormality lags behind clinical and laboratory signs of improvement. In patients with acceptable courses and outcomes, renal masses often take months to resolve (see Figure 52-2).

COMPLICATIONS AND PROGNOSIS A variety of complications, including loss of renal function, can occur. An abscess can extend within the kidney or perinephric space, causing additional tissue destruction and organ dysfunction, or can rupture into an adjacent space (abdominal, pulmonary). Bacteremia from abscess leakage, occurring spontaneously or as a surgical complication, can result in hematogenous spread of infection to other organs. Prognosis depends on the presence of underlying conditions, the timeliness of diagnosis, and the appropriateness of treatment. Morbidity and mortality can be substantial. The development of sequelae is directly related to the loss of renal function and to dysfunction of organs involved either as primary sites of infection or as a result of bacteremia.

PREVENTION Prevention of renal abscesses, when possible, depends on the appropriate management of the conditions (skin, dental, respiratory, cardiac, abdominal, and other infections, vesicoureteral reflux, urinary tract obstruction, and others) that predispose to their development.

CHAPTER

53

Sexually Transmitted Infection Syndromes Margaret J. Blythe and J. Dennis Fortenberry

Healthcare providers should feel comfortable initiating discussion about sexually transmitted infections (STIs) with their adolescent patients, recognize clinical syndromes associated with STIs, and provide confidential screening and appropriate treatment.1 Lack of access to confidential care limits teens’ reported use of healthcare services.2,3 Young people are concerned about possible exposure to STIs and expect inquiry from physicians about potential risk for STIs, as many teens become sexually active during middle and late adolescence.4,5 Additionally, rates of reported sexually acquired infections are high in the 15- to 19-year-old and 20- to 24-year-old age groups of young men and women.6 STIs diagnosed in the prepubertal patient require assessment for sexual abuse; infections acquired during infancy can be the result of sexual abuse or perinatal acquisition (see Chapter 58, Infectious Diseases of Child Abuse).

ASYMPTOMATIC INFECTION

Figure 52-3. Axial image from contrast-enhanced computed tomography scan in a teenager with autosomal-dominant polycystic kidney disease and diabetes, showing low-density fluid collections within the kidney and in the anterior perirenal space (left arrow). The deep, noncommunicating intrarenal perinephric abscesses were drained percutaneously and Staphylococcus aureus was isolated from fluids. Position of the pigtail catheters is shown (right arrow). Air is iatrogenic. Note the multiple low-density cysts in both kidneys in this patient with polycystic kidney disease.

Most of the common STIs are asymptomatic. Thus, regular screening for infections in sexually active teens is required in order to improve early detection and treatment of infections, and minimize risk of adverse medical consequences. In particular, screening of young women and young men once or twice each year for Chlamydia trachomatis infections may be important. Results of a 2004 study from a nationally representative sample of teen and young adult men and women indicate an overall prevalence of 4% for chlamydia and 0.4% for gonorrhea. More than 95% of people infected reported no symptoms in the 24 hours prior to testing.7 Less than two-thirds of people infected thought they were at risk for infection and less than one-third had been tested for STIs in the prior 12 months.8

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Sexually Transmitted Infection Syndromes

SYMPTOMATIC INFECTION Six clinical syndromes make up the majority of symptomatic presentations of adolescents and young adults with an STI: (1) discharge/dysuria syndrome; (2) anal discharge/proctitis syndrome; (3) genital ulcer/lymphadenopathy syndrome; (4) pelvic or scrotal pain syndrome; (5) pharyngeal infection; and (6) dermatologic syndromes (Table 53-1). Table 53-2 shows office and laboratory tests used to determine the etiology of STI syndromes in adolescents and young adults.

Discharge/Dysuria Syndrome The genital discharge/dysuria syndrome among sexually active young women is largely a result of infections caused by one or more of the following: Chlamydia trachomatis, Neisseria gonorrhoeae, and Trichomonas vaginalis. Among young men, genital discharge/dysuria syndrome is most commonly associated with C. trachomatis, N. gonorrhoeae, and T. vaginalis.7,9 Coinfections with chlamydia and gonorrhea are common. Studies of various populations of teens indicate that 30% to 70% diagnosed with gonorrhea also have chlamydia.7,10,11 Other less commonly identified organisms associated with discharge/dysuria syndrome are Ureaplasma urealyticum, Mycoplasma hominis, M. genitalium, and viruses such as herpes simplex virus (HSV). Use of nucleic acid amplification techniques indicates that the prevalence of U. urealyticum or M. genitalium or both is higher in males with symptoms of urethritis compared with males with no symptoms of urethritis.12–15 A high rate of infection has been found among sexual partners of people infected with M. genitalium, suggesting that the infection is transmitted sexually.14,15 Other conditions associated with discharge, primarily among women, include bacterial vaginosis and vaginal candidiasis (Table 53-1). Clinical features are not reliable predictors of the microbiologic etiology of genital discharge/dysuria among women or men. Color,

CHAPTER

53

smell, presence of pruritus, and quantity of discharge do not reliably indicate the etiology of vaginal discharge. Young women often present with dysuria referred to as “sterile pyuria” when white blood cells (WBCs) are found in urine with negative urine culture. Sterile pyuria may accompany cervicovaginal infection due to meatal irritation or urethral infection. Other symptoms that may accompany vaginal discharge/dysuria are abnormal vaginal bleeding, dyspareunia, vulvovaginal itching, and mild pelvic pain. Among men, characteristics of the discharge are useful but are not reliable to distinguish chlamydia from gonococcal infections. In general, urethral discharge due to C. trachomatis is characterized as less purulent and profuse in contrast to discharge due to N. gonorrheoae. A majority of symptomatic men with gonococcal infection complain of both discharge and dysuria, but only about 2% complain of dysuria alone. Urethritis due to Chlamydia or Mycoplasma may more often be associated with dysuria alone, sometimes only with first void of urine. Other findings/symptoms sometimes associated with discharge/dysuria in men are hematuria, meatal itching, and edema of the glans and head of the penis.

Anal Discharge/Proctitis Syndrome Anorectal discharge and pain associated with tenesmus and even anorectal bleeding is a second STI syndrome that occurs in both young men and young women. Asymptomatic infections are common, although the exact prevalence is not known since population-based screening studies have not been done. Receptive anal intercourse is associated with proctitis syndrome. Infections without a history of anal sex can occur in females but not males. In the United States, anogenital infections acquired sexually are primarily caused by Chlamydia trachomatis (including lymphogranuloma venereum (LGV) serovars), N. gonorrhoeae, and HSV.16,17 Human papillomaviruses (HPV) can infect anogenital tissues but remain asymptomatic unless causing anal warts. Other possible organisms

TABLE 53-1. Sexually Transmitted Infection (STI) Syndromes in Adolescents and Young Adults STI Syndrome

Primary Organisms

Other Causal Organisms

Discharge/dysuria

Chlamydia trachomatis Neisseria gonorrhoeae Trichomonas vaginalis Females: Candida albicans, anaerobes

Ureaplasma urealyticum Mycoplasma hominis, Mycoplasma genitalium Herpes simplex 1, 2

Discharge/proctitis

Chlamydia trachomatis Neisseria gonorrhoeae Treponema pallidum Herpes simplex 1, 2

Shigella species Campylobacter species Salmonella species Giardia lamblia Entamoeba histolytica

Genital ulcer/lymphadenopathy

Herpes simplex 1, 2 Trepomena pallidum

Haemophilus ducreyi Lymphogranuloma venereum Calymmatobacterium granulomatis

Pelvic pain (pelvic inflammatory disease)

Chlamydia trachomatis Neisseria gonorrhoeae

Mycoplasma hominis, Mycoplasma genitalium Mixed aerobic/anaerobic bacteri

Scrotal pain (epididymitis)

Chlamydia trachomatis Neisseria gonorrhoeae

Ureaplasma urealyticum Mycoplasma genitalium

PHARYNGITIS

Neisseria gonorrhoeae Herpes simplex 1, 2

Treponema pallidum Human papillomaviruses

GENITO-URINARY SYNDROMES

DERMATOLOGIC SYNDROMES

Genital warts Molluscum contagiosum Rash, alopecia Arthritis/dermatitis syndrome Jaundice/hepatitis Scabies Pubic lice

351

Human papillomaviruses Molluscum contagiosum virus Treponema pallidum Neisseria gonorrhoeae Chlamydia trachomatis Hepatitis A, B, C Sarcoptes scabiei Phthirus pubis

352

SECTION

G Urinary Tract Infections

include Shigella species, Campylobacter species, Salmonella species, Entamoeba histolytica, and Giardia lamblia. These organisms often cause symptoms of diarrhea, bloating, and abdominal pain in addition to symptoms related to proctitis syndrome.

Genital Ulcer/Lymphadenopathy Syndrome Genital Herpes In the United States, genital herpes due to HSV is the most common cause of genital ulcers. In the past, genital herpes typically was associated with HSV-2, with a small proportion of incident cases associated with HSV-1. Recently, the proportion of cases of genital herpes due to HSV-1 has risen and is estimated to be 30% to 50% of new cases.18 Data from national studies indicate that the prevalence of positive serology for HSV-2 in people 12 to 19 years of age is 5.6% and increases to 17.2% for people 20 to 29 years of age. Overall, 1 of every 4 females and 1 of every 5 males 12 years of age or older is seropositive for HSV-2. Up to 90% of people with positive HSV-2 serology indicate no clinical history of anogenital outbreaks.19 Over two-thirds of the United States population has antibody to HSV-1, specifically 44.4% of people 12 to 19 years of age and 56.4% of people 20 to 39 years of age.20 Clinical presentation of lesions due to genital herpes depends somewhat on whether the outbreak is initial (no antibodies to HSV-1 or -2) or recurrent (antibodies to HSV-1 and/or -2) and on the stage of presentation (vesicular, pustular, ulcerated, or crusted). First clinical episode of infections is more likely to be associated with systemic symptoms such as headache, myalgia, photophobia, fever, and malaise, particularly in the absence of HSV-1 and HSV-2 antibodies. Approximately 10% of first clinical infections are recurrent infections.21 Initial symptoms may be only pain, burning, or itching in the area where lesions eventually appear. Development of single or small groups of vesicles typically follows within a few hours. Vesicles are often quite painful, and progress to painful ulcers over a period of a few days. Tender, usually bilateral inguinal lymphadenopathy can appear at this time. New lesions can appear, leading to finding of lesions in various stages of development. The length of viral shedding and time required to heal depend on whether the infection is initial or recurrent; for initial infections, viral shedding continues for 10 to 12 days and healing requires up to 3 weeks; for recurrent infections, shedding occurs for up to 4 days, with healing in 5 to 7 days. Dysuria and urethral or vaginal discharge are also common.

Syphilis Primary syphilis is an important cause of genital ulcer/lymphadenopathy syndrome, although it is relatively uncommon among adolescents. The classic lesion of primary syphilis is a single, deep, indurated, painless ulcer associated with unilateral, nonfluctuant inguinal adenopathy on the same side. Multiple chancres occur in approximately 25% of cases. These lesions occur about 3 weeks (10 to 90 days) after sexual contact with an infected person and heal spontaneously without treatment in 1 to 6 weeks. Lesions can occur almost anywhere in the genital/rectal area, with perianal and rectal chancres usually the result of rectal intercourse. Clinical presentation of genital lesions due to HSV and to primary Treponema pallidum can be similar. Classic findings for genital herpes (multiple, shallow, tender ulcers) and for primary syphilis (deep, indurated, painless ulcers) are about 35% sensitive for a confirmed diagnosis.22

Other Causes Other causes of genital ulcer diseases are relatively rare in the United States, with sporadic outbreaks of LGV primarily recorded in human immunodeficiency virus-positive men who have sex with men in both Europe and the United States.17 LGV is caused by Chlamydia

trachomatis serovars L1, L2, L3, with an incubation time of 3 to 12 days or longer. Most infections begin as an asymptomatic ulcer or papule or erosion at the site of infection and can be associated with nonspecific urethritis. Lesions then progress clinically to lymphadenopathy. Another rare cause of lymphadenopathy/genital syndrome in the United States is Haemophilus ducreyi (chancroid), a gram-negative anaerobic bacillus.23 Incubation is usually between 4 to 7 days with no prodromal symptoms; lesions are tender papules surrounded by erythema. Lesions become pustular, erode, and then ulcerate over 14 to 48 hours. Granuloma inguinale (donovanosis), caused by Klebsiella (Calymmatobacterium) granulomatis, can be considered the etiology of genital ulcerative disease if the infection was acquired in a developing country or by sexual contact with someone traveling in or from a developing country. Lesions are painless, “beefy red,” and progress to ulcerations without regional lymphadenopathy.

Pelvic or Scrotal Pain Syndrome Pelvic (in young women) or scrotal (in young men) pain is often associated with an STI. Pelvic pain is common in patients with pelvic inflammatory disease (PID). Although N. gonorrhoeae and C. trachomatis are most often linked to PID, these pathogens are identified in less than a third of the cases of PID. Other organisms, including anaerobes representing endogenous vaginal and perineal microflora, have also been isolated from the endometrium or fallopian tubes or both.24,25 Some studies have also linked M. genitalium with PID.26–29 Subclinical disease, i.e., “silent” PID, is thought to be common. Symptoms of PID can be mild, consisting of discomfort and abdominal tenderness, or can be fever and generalized abdominal and pelvic pain. In general, the diagnosis is based on the presence of uterine or adnexal tenderness, or cervical motion tenderness. Criteria for making a clinical diagnosis of PID are imprecise, thus making it important to rely on the history as well as other clinical indicators. The history may include dyspareunia, irregular bleeding, dysuria, and even vague gastrointestinal tract symptoms. Other clinical findings supportive of diagnosis of PID include fever (oral temperature greater than 38.3°C), abnormal cervical discharge (mucopus), and/or palpable pelvic mass. A wet mount (saline microscopy) of vaginal secretions should be examined. An abundant number of WBCs or the finding of Trichomonas or clue cells supports the diagnosis.16 Laboratory findings suggestive of PID include elevated erythrocyte sedimentation rate (ESR) or C-reactive protein, or laboratory documentation of infection due to N. gonorrhoeae or C. trachomatis or presence of sterile pyuria. In a publication re-evaluating the Swedish experience using laparoscopy to diagnose PID, the presence of three findings, i.e., fever, adnexal tenderness, and elevated ESR, most often “correctly classified” the finding of PID.30 An incidental urinary tract infection can accompany PID. A urine pregnancy test is performed since ectopic pregnancy can cause pelvic pain. Scrotal pain in the context of STIs most often represents epididymitis. Young men have insidious onset of unilateral scrotal pain, tenderness, and swelling in the affected area. Torsion of the testicle must be ruled out. History of dysuria and urethral discharge is often, but not invariably, present. Fever and bilateral involvement are infrequent. Alleviation of pain with gentle manual elevation of the affected side (Prehn sign) is supportive of the diagnosis. Presence of WBCs and gram-negative intracellular diplococci (N. gonorrhoeae) on a stain are helpful, as is the presence of positive leukocyte esterase on a urine dipstick or WBCs in a spun urinalysis. Among young men, C. trachomatis and N. gonorrhoeae are the most common microbiologic etiologies of epididymitis and can occur concurrently. U. urealyticum and M. genitalium likely cause some cases but their exact role as agents of epididymitis is poorly defined.31 A few young men have epididymitis due to Escherichia coli and other gramnegative enteric organisms, although this presentation is most common among men older than 35 years of age.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Sexually Transmitted Infection Syndromes

CHAPTER

53

353

TABLE 53-2. Office and Laboratory Tests to Determine Etiology of Sexually Transmitted Infection (STI) Syndromes in Adolescents and Young Adults STI Syndrome

Office Tests

Laboratory Tests

Females: wet prep of vaginal secretions; LET, pH paper, rapid tests for TV, test cards for BV Males: Gram stain urethral discharge, LET

Cultures for GC, CT, HSV-1, -2 or NAATs for GC, CT, TV and/or yeast (recurrent yeast infections)

GENITO-URINARY SYNDROMES

Discharge/dysuria

Discharge/proctitis

Cultures for GC, CT, HSV-1, -2; Cultures for Gram negative pathogens, ameba

Genital ulcer/lymphadenopathy

Females, males: Darkfield examination for syphilis

Cultures for HSV-1, -2; serology for HSV-2; RPR or VDRL and confirmatory tests for syphilis if positive; serology for LGV; Gram stain and culture for chancroid; staining for Donovan bodies on tissue biopsy

Pelvic pain (pelvic inflammatory disease)

Wet prep of vaginal secretions; pH paper, rapid tests for TV, test cards for BV; Gram stain urethral discharge, LET

Cultures for GC, CT or NAATs for GC, CT, TV

Scrotal pain (epididymitis)

PHARYNGITIS

Cultures for GC, CT or NAATs for GC, CT, TV Rapid strep test

Cultures for GC, HSV-1, -2; RPR or VDRL and confirmatory tests if positive

DERMATOLOGIC SYNDROMES

Genital warts Molluscum contagiosum

Females: Papanicolaou smear Characteristic lesions

Wright or Giemsa staining for intracytoplasmic inclusions Males/females: RPR or VDRL and confirmatory tests if positive

Rash/alopecia Arthritis/dermatitis syndrome

Males: Gram stain urethral discharge, LET Females: wet prep of vaginal secretions; pH paper; LET

Jaundice/hepatitis

Culture for GC from rectum, pharynx, genital; or NAATs of genital secretions or urine Appropriate laboratory tests for hepatitis

Scabies

Microscopic examination of skin and hair

Pubic lice

Identification of eggs, nymphs and lice with naked eye or microscopy

BV, bacterial vaginosis; CT, Chlamydia trachomatis; GC, Neisseria gonorrhoeae; HSV-1, -2, herpes simplex-1,-2; LET, leukocyte esterase test of urine; LGV, lymphogranuloma venereum; NAATs, nucleic acid amplification tests; RPR, rapid plama reagin; TV, Trichomonas vaginalis; VDRL, Venereal Disease Research Laboratory.

Pharyngeal Infection Gonococcal infections are limited to mucosal surfaces that are lined with columnar epithelium, including the cervix, urethra, rectum, pharynx, and conjunctiva. Gonococcal infection can manifest as an asymptomatic infection of the pharynx or as a mild pharyngitis with sore throat and erythema. Gonococcal infection of pharynx can lead to disseminated gonococcal infection. HSV-1 and -2 can infect the pharynx, with clinical manifestations of sore throat, fever, malaise, myalgia, headache, and tender anterior cervical lymphadenopathy. Mild ulceration and diffuse erythema can also be the initial manifestation. Oral syphilis can cause chancres on the lips, buccal mucosa, tonsils, and pharynx. Lesions on the tonsils and pharynx may be painful. Oral and laryngeal papillomas due to HPV may appear after oral genital contact.16

Dermatologic Syndromes Dermatologic syndromes that occur in adolescents and young adults include genital warts, various dermatologic manifestations of secondary syphilis, the arthritis/dermatitis syndrome associated with disseminated gonococcal infection, jaundice/hepatitis associated with several sexually transmitted viruses, and the sexually transmitted parasites scabies and pubic lice.

Genital warts are visible manifestations of nearly ubiquitous subtypes of human HPV infection. HPV subtypes most commonly associated with exophytic genital warts (condyloma acuminata) are types 6 and 11 or other low-risk types, including 42 and 43. Genital warts can occur at almost any squamous epithelial site in the genital tract. Less frequently, HPV-associated warts are found on the lips or in the mouth and larynx. In young men, common sites of warts are the glans, prepuce, urethral meatus, and shaft of the penis, as well as the perianal area. However, warts can often occur at almost any site on the genital skin. In young women, warts are found in almost any area of the vulva, perianal area, vagina, and cervix. Molluscum contagiosum can be an STI, and can be confused with genital warts. Lesions are typically 1 to 2 mm, smooth, skin-colored papules. Infections can be transmitted by any skin-to-skin contact; lesions on nongenital skin are common. Lesions are usually asymptomatic. Lesions of secondary syphilis are seen 2 to 6 weeks after infection. Often, the primary lesion has not been noticed, or has been ignored. Secondary syphilis has a number of potential dermatologic presentations. Diffuse maculopapular and papulosquamous rashes of the trunk, arms and legs are most common. A papulosquamous rash involving the palms and soles also occurs. Wart-like growths, typically in the perianal or posterior vulvar area, are manifestations of secondary syphilis, referred to as condyloma lata, and initially can be confused with genital warts.

354

SECTION

G Urinary Tract Infections

The arthritis/dermatitis syndrome of disseminated gonococcal infection classically presents with dermatitis, tenosynovitis, and migratory polyarthritis. Knee, ankle, wrist, and metacarpalphalangeal joints are most commonly involved. Dermatitis is usually a painless, maculopapular rash typically occurring on the trunk and limbs. Tenosynovitis usually involves the hands and fingers, and may be asymmetric. Dissemination of infection can occur from any site of genital, anal, or pharyngeal infections by N. gonorrhoeae. Disseminated infections are more common among young women than young men. For many adolescents, jaundice associated with acute viral hepatitis is an STI, including infection due to hepatitis A virus, hepatitis B virus, cytomegalovirus, and Epstein–Barr virus. Scabies is caused by the mite, Sarcoptes scabiei. Sexual transmission is common, although close personal contact such as sharing a bed can lead to transmission in the absence of sexual contact. Genital infestations are common, although concentration of mites in the hands and fingers is more common. However, the rash associated with scabies is caused in part by sensitization and can occur in body parts other than those directly infested. Pubic lice (Phthirus pubis) are also associated with close interpersonal contact, not necessarily sexual contact. Symptoms are related to itching and irritation: lice and/or nits are often visible.

CHAPTER

54

Skin and Mucous Membrane Infections and Inguinal Lymphadenopathy Catherine McLean and Kimberly Workowski

40% to 90% of GUD in crack cocaine users.6,7 Following an all-time low incidence in reported syphilis cases in the 1990s, several large syphilis outbreaks in men who have sex with men (MSM) were identified in cities in North America and Europe.8,9 LGV and donovanosis are also more common in Africa and Asia and typically have represented imported infections when diagnosed in industrial countries. However, in the mid 2000s, clusters of proctitis due to LGV have been reported in MSM in the United States and Europe. Most patients with LGV proctitis in these clusters have been HIV-coinfected and have participated in high-risk sexual behaviors.10,11 Prior to the HIV epidemic in Africa, Southeast Asia, and India, GUD was more commonly due to chancroid than HSV1,12,13 In countries with high HIV prevalence, an increasing proportion of GUD appears to be due to HSV, in family-planning clinic attendees. In Botswana, the proportion of GUD due to HSV increased from 35% in 1998 to 61% in 2002, whereas the proportion due to T. pallidum (52% to 5%) and H. ducreyi (33% to < 1%) declined substantially.14 This pattern is believed to be similar to other sub-Saharan countries where HIV prevalence in the general population is 15% to 40%.10,13,15 GUD due to HSV and the interaction with HIV has been the topic of considerable research.1,16–18 HSV infections are chronic; many persons are not aware they are infected and may have unrecognized or only mild symptoms. The majority of genital HSV infections are transmitted by people unaware that they have the infection. People with HSV-2 infections shed virus intermittently in the genital tract, at which time they can transmit HSV-2 to sexual partners.17 Whereas the majority of recurrent GUD is due to HSV-2, GUD due to HSV-1 in the United States may be increasing. The course of symptoms can be modified by therapy with acyclovir (or its analogues) if treatment is started during the prodrome or within 24 hours of symptom onset.17 GUD increases the risk of transmission of HIV.18,19 Further, in the absence of ulceration, HSV-2 infection increases the risk of HIV transmission to a sex partner at least two-fold and may increase infectiousness among HIV-infected people.16 Treatment of HSV-2 infection (in the absence of GUD) may reduce HIV infectiousness and may reduce HIV transmission when used by HIV-infected people, a hypothesis being investigated in randomized clinical trials. HIV infection itself may increase susceptibility to several GUD pathogens.11,15,19

Clinical Manifestations and Differential Diagnosis GENITAL ULCER WITH LYMPHADENOPATHY Infections that cause lymphadenopathy with and without genital ulcer disease (GUD) are described in this chapter. Infections causing GUD are frequently transmitted sexually and can manifest as unilateral or bilateral inguinal lymphadenopathy. Genital ulcerations due to sexually transmitted infections (STI) pose complex clinical problems because of the multiple possible etiologies, diverse clinical presentations, and challenges to making a definitive diagnosis.1–3

Etiologic Agents and Epidemiology The most common causes of genital ulceration are herpes simplex viruses (HSV-2 and HSV-1), Treponema pallidum (syphilis), Chlamydia trachomatis serovars L1, L2, and L3 (lymphogranuloma venereum (LGV)), Haemophilus ducreyi (chancroid), and Calymmatobacterium granulomatis (donovanosis or granuloma inguinale).1–3 Infrequently, scabies and pubic lice can result in GUD, as can pyoderma, Entamoeba histolytica, and Capnocytophaga spp. infections. In general, GUD is more common in the developing world than in North America or western Europe. Overall, an estimated 1% to 5% of patients seen in STI clinics in North America and Europe have a genital ulcer compared with 20% to 70% of STI clinic attendees in Africa and Southeast Asia. The etiology of GUD is geographically distinct. Genital HSV is a common cause of GUD in patients in North America, whereas syphilis and chancroid are less common.1–4 There is marked variation in prevalence of syphilis and chancroid by city, region, and risk group in North America.5 In the late 1980s and early 1990s, syphilis and chancroid were found to be the cause of

Identification of the etiology of GUD presents several challenges. Experienced STI clinicians are unable to distinguish between GUD etiologies based upon clinical presentation only.8,9 Obtaining a specific etiologic diagnosis frequently requires specialized testing. The specimen collection materials, testing equipment, and trained personnel required to conduct and interpret some of tests are typically not available in many laboratories, especially in the developing world. The classic clinical features of the major causes of GUD are described in Table 54-1. In clinical practice, distinctions are not necessarily apparent, because lesions of different etiologies may coexist, GUD clinical presentations are less distinctive for each etiology than previously believed, and immunocompromising conditions, such as HIV infection, can modify clinical manifestations. Careful examination of the entire genital region, including the perineum and anus, facilitated by a good light source and a handheld magnifying lens, may improve the clinical assessment and guide selection of appropriate diagnostic tests. Additionally, due to the asymptomatic nature of several of these infections, anorectal manifestations may be overlooked by the patient as well as the clinician. Performing an anoscopic examination can assist in the identification of lesions and improve GUD diagnosis and treatment.

Laboratory Diagnosis and Management Confirmation of etiology is usually possible with appropriate specimen collection and testing (Table 54-2). A discussion with local laboratory personnel before collection and submission of a specimen can facilitate the testing process.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Skin and Mucous Membrane Infections and Inguinal Lymphadenopathy

CHAPTER

54

355

TABLE 54-1. Classic Clinical Featuresa of Sexually Transmitted Genital Ulcer Disease Syndromes

Feature Incubation period estimates

Herpes simplex (Herpes)

Treponema pallidum (Syphilis)

Haemophilus ducreyi (Chancroid)

Chlamydia trachomatis (Lymphogranuloma Venereum)

Calymmatobacterium granulomatis (Granuloma Inguinale)

2–7 days

2–4 weeks

1–14 days

3–42 days

8–80 days

Glans/prepuce, penis, anus/rectum Cervix, vulva, perineum, buttocks/legs, anus/rectum

At site of inoculation

At site of inoculation

90% involve genitalia

At site of inoculation

At site of inoculation

At site of inoculation (urethal/rectal) At site of inoculation (urethal/rectal)

SITE

Male Female

90% involve genitalia

Typical primary lesion presentation

Vesicle (variable, depending upon lesion duration, host immune status, etc.)

Papule (ulcerates)

Papule or pustule

Papule, vesicle, vesiculopustular lesion

Subcutaneous nodule that erodes; occasionally verrucous

Number of lesions

Usually multiple; can coalesce, especially in immunocompromised host

Usually single, can be multiple

Often multiple; can coalesce

Usually single

Single or multiple

Ulcer appearance

Small, superficial, smooth; with erythematous edge, circular None

Deep, small to large; undermined, with ragged edge, irregular shape Soft

Variable depth, small to medium size; elevated edge, round/oval

Induration

Superficial, medium size, well demarcated; with elevated edge, circular/oval Firm

Occasionally firm

Small to large lesions; with elevated edge and beefy base, irregular shape Firm

Pain

Exquisitely

Typical painless

Variable

Variable

Not typical

Lymphadenopathy characteristic

Firm, tender, often bilateral

Firm, nontender, bilateral

Tender, can suppurate; unilateral; superinfection

Large, tender, unilateral; can suppurate

Pseudobuboes; regional lymphadenopathy with superinfection

Treatment

Acyclovir, famciclovir or other acyclovir analogues

Penicillin (dose and duration depend upon clinical stage)

Azithromycin, ceftriaxone, doxycycline, or erythromycin

Doxycycline (alternative: erythromycin)

Doxycycline (alternative: erythromycin)

a

Typical but not universal clinical presentation.

TABLE 54-2. Specimen Selection and Laboratory Tests for Diagnosis of Genital Ulcer with Lymphadenopathy Syndrome Condition

Specimen Type

Test

Herpes

Scraping from ulcer base Serology

Antigen detection, tissue culture HSV-1 and HSV-2 antibody detection (IgM for early infection, IgG for previous infection)

Syphilis

Exudate from ulcer Serology

Darkfield microscopy examination for spirochete Nontreponemal (with titer quantification); if positive, confirm with treponemal test

Chancroid

Swab from ulcer base or aspirate from bubo

Semiselective media

Lymphogranuloma venereum

Aspirate from bubo Lesion or swab from ulcer base approved

Tissue culture No FDA-sites of infection, commercially available test, although Chlamydia trachomatis testing will assist with diagnosis

Donovanosis

Crush preparation from lesion

Giemsa or Wright stain for Donovan bodies

FDA, Food and Drug Administration; HSV, herpes simplex virus; Ig, immunoglobulin.

HSV-1 and HSV-2 GUD have different prognoses and counseling messages, and therefore, the clinical diagnosis should be confirmed by laboratory testing (Chapter 287. Laboratory Diagnosis of Infection due to Viruses, Chlamydia, and Mycoplasma).17 Recurrences and subclinical shedding are less frequent for genital HSV-1 than genital HSV-2. Both virologic and type-specific serologic tests for HSV should be available in clinical settings that provide care for patients with STIs or those at risk for STIs. Isolation of HSV in cell culture is the preferred virologic test for patients who seek medical treatment for GUD or other mucocutaneous lesions. However, the sensitivity of culture is low, especially for recurrent lesions, and declines as lesions

begin to heal. Therefore, lack of HSV detection does not indicate a lack of HSV infection, because viral shedding is intermittent. The use of cytologic detection of cellular changes due to HSV infection is insensitive and nonspecific for genital lesions and for cervical Papanicolaou smears and should not be used. Accurate, type-specific HSV serologic assays are based on the HSV-specific glycoproteins G2 (HSV-2) and G1 (HSV-1). These assays first became commercially available in 1999, but older assays that do not accurately distinguish HSV-1 from HSV-2 antibody remain on the market. Therefore, the serologic type-specific glycoprotein G (gG)-based assays should be requested when serology is performed.17 Antibody negativity does not

356

SECTION

G Urinary Tract Infections

exclude early primary infection; antibody positivity does not necessarily indicate current infection.17 Optimal management of GUD depends on speciÀc identiÀcation of the cause (see Table 54-1). If no clinical improvement is evident following treatment, the clinician must consider if the diagnosis is correct, the patient is coinfected with another STI, the patient is infected with HIV, the treatment was not used as instructed, or the microbiologic strain causing the infection is resistant to the prescribed antimicrobial angent.17 Because concurrent infections may be present in people diagnosed with GUD, testing for other concurrent STIs is recommended. A thorough risk assessment and counseling and testing for HIV infection are important to conduct on all patients diagnosed with an STI and with people at ongoing risk of STI infection, including STI and family-planning clinic attendees. Partner counseling and treatment (if indicated), and STI and HIV testing are also recommended as part of comprehensive STI care. Diagnosis of STI in a child or adolescent should trigger a thorough evaluation for sexual abuse (see Chapter 58, Infectious Diseases of Child Abuse).17

INGUINAL ADENOPATHY WITHOUT GENITAL ULCERS While infections associated with GUD often cause inguinal adenopathy (see Table 54-1), some pathogens cause inguinal adenopathy without ulcers (Table 54-3). In these infections, inguinal lymphadenopathy is typically bilateral, although chancroid bubo is usually unilateral. Inguinal lymphadenitis associated with pyogenic infections of the lower extremities, lower abdominal wall, or perineum is usually unilateral and frequently is of acute onset; progression to local abscess formation can occur (see Table 54-3), typically due to Staphylococcus aureus or group A streptococcus. Retrocecal appendicitis is of particular note. In patients with this condition, unexplained limp with right-sided inguinal lymphadenopathy can be the initial symptom, followed by onset of fever. With abscess formation, back and hip pain become increasingly prominent, and extension of the upper leg produces pain. Cat-scratch disease (see Chapter 160, Bartonella Species) and mycobacterial infections can occasionally manifest with inguinal adenopathy (see Chapter 134, Mycobacterium tuberculosis, and Chapter 135, Mycobacterium Species Non-tuberculosis). In bubonic plague, if the fleabite initiating the infection occurs on the leg, inguinal or femoral buboes can occur (see Chapter 148, Yersinia Species). Tularemia can mimic plague but is more likely to cause ulceroglandular syndrome than glandular symptoms alone (see Chapter 171, Francisella tularensis).

Management of isolated inguinal lymphadenopathy depends on the suspected etiology (see Table 54-2). Aspiration of lymph nodes with smear examination and culture is useful in pyogenic infections. Further diagnostic investigations, such as ultrasonography or computed tomography, are helpful if osteomyelitis or retrocecal appendicitis is suspected. SpeciÀc antimicrobial therapy is directed against the inciting pathogen.

GENITAL DERMATOSIS Yeast Infections Many skin infections affect the genitals and genitocrural folds, especially infections of the diaper region and perineum caused by Candida albicans or C. glabrata (see Chapter 244, Candida Species). In diapered infants, eruption often starts in the perianal area and spreads to involve the perineum and upper thighs. Onset can be acute, with scaly macules or papules that coalesce to form well-demarcated, erythematous lesions with irregular borders and frequently with satellite lesions. Treatment consists of keeping the area as dry as possible and applying a topical antifungal cream. In older children and adolescents, genital candidal infection can involve the vagina or the penis. Symptoms of vulvovaginitis pruritus, include pain, and a whitish milky discharge (see Chapter 55, Urethritis, Vulvovaginitis, and Cervicitis). Balanoposthitis can be associated with pruritus, and erythematous lesions with whitish patches can be seen on the glans, prepuce, and shaft of the penis. Topical antifungal therapy is effective, but relapses occur.

Dermatophyte Infections Excessive perspiration and friction in intertriginous areas such as the groin predispose to superÀcial infections with dermatophytes including Trichophyton rubrum, T. mentagrophytes, or Epidermophyton floccosum (see Chapter 255, Dermatophyte and Other SuperÀcial Fungi). These infections are more common in hot weather and occur more in males than in females. The infection, although often symmetric, can be asymmetric with extension from the crural areas to the upper thigh. Lesions may be erythematous scaly patches with papular or vesicular margins. The lesions are typically pruritic; scratching can lead to licheniÀcation. Treatment consists of use of a topical antifungal agent. Systemic therapy (e.g., fluconazole) is occasionally used in widespread or recalcitrant cases, a common manifestation of candidiasis in immunocompromised patients, until the infection is resolved. Some Candida infections have developed

TABLE 54-3. Differential Diagnosis of Inguinal Lymphadenopathy Condition

Organism

Pattern of Lymphadenopathy

Associations

Pyogenic bacterial disease

Group A streptococcus, Staphylococcus aureus

Suppurative, unilateral, and tender; can progress to abscess

Tuberculosis Cat-scratch disease Plague

Mycobacterium tuberculosis Bartonella henselae Yersinia pestis

Caseating, can be unilateral, nontender Unilateral, tender, slowly progressive Large cluster, very tender nodes

Retrocecal appendicitis; osteomyelitis, pyogenic arthritis; infection of skin or abdominal wall Miliary tuberculosis, osteomyelitis Scratch lesion on lower limb Exposure to fleas, rodents, or rabbits

Bilateral nodes (see text for details)

Sexually transmitted

Large, very tender lymph nodes with ulcerated skin lesions

Exposure to rabbit, squirrel, coyote

WITHOUT ULCER(S)

WITH ULCERS

Herpes Syphilis Chancroid Lymphogranuloma venereum Granuloma inguinale (donovanosis) Tularemia

Herpes simplex virus Treponema pallidum Haemophilus ducreyi Chlamydia trachomatis (L1, L2, L3) Calymmatobacterium granulomatis Francisella tularensis

}

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Urethritis, Vulvovaginitis, and Cervicitis

CHAPTER

55

357

topical antifungal agent. Systemic therapy (e.g., fluconazole) is occasionally used in widespread or recalcitrant cases, a common manifestation of candidiasis in immunocompromised patients, until the infection is resolved. Some Candida infections have developed resistance to antifungal agents; therefore, it is important to be aware of any information about selective candidiasis resistance patterns in the local community. People with persistent or recurrent infection should be evaluated for possible immunodeficiency, including HIV infection testing.17

through induction of local inflammation, as with application of podophyllin. In adolescents with multiple facial molluscum lesions, and in children with multiple lesions, concurrent HIV infection should be considered. Lesions associated with HIV infection can be atypical and extensive. Treatment of extensive infections is challenging, and patients with this condition should be referred to a dermatologic specialist for treatment.

Genital Warts (Condyloma Acuminatum)

Scabies and pubic lice also can be transmitted sexually and can cause skin lesions in the genital area.

Genital warts are caused by infection of the epidermis with a human papillomavirus (HPV), especially types 6 and 11, 16, and 18 (see Chapter 211, Human Papillomaviruses) (Figure 54-1).20 Among adolescents and adults, anogenital warts are most often transmitted sexually among adolescents and adults. In children, the mode of acquisition is more difficult to assess (see Chapter 58, Infectious Diseases of Child Abuse).17

Molluscum Contagiosum Infection Molluscum contagiosum is caused by parapoxvirus, and can be transmitted sexually and nonsexually (see Chapter 202, Poxviridae (Molloscum Contagiosum)). The mean incubation period is 2 to 3 months. Diagnosis is usually based on clinical appearance. Unlike genital warts, which occur predominantly on external genitalia and the perianal area, lesions of molluscum contagiosum most often occur on thighs, inguinal region, buttocks, and lower abdominal wall. Children more typically have lesions on the face, trunk, and extremities, but 10% to 50% can have genital lesions. Lesions begin as small papules and can grow to 10 to 15 mm. They become smooth, firm, and domeshaped with central umbilication, from which caseous material can be expressed. Lesions can be removed mechanically with curettage or

OTHER CONDITIONS

CHAPTER

55

Urethritis, Vulvovaginitis, and Cervicitis Paula K. Braverman

URETHRITIS Urethritis represents response of the urethra to an inflammatory process. The clinical presentation includes dysuria, urinary frequency, and urethral discharge or itching. Polymorphonuclear leukocytes usually are found in urethral secretions.1 According to the Centers for Disease Control and Prevention (CDC), urethritis can be documented by one of the following criteria: a visible abnormal urethral discharge; or a positive leukocyte esterase (LE) test in a male younger than 60 years of age without other urinary tract disease that could cause pyuria; or at least 5 white blood cells per high-powered field (WBCs/hpf) on urethral smear Gram stain; or a first-void urine with ≥10 WBCs/hpf.2,3 However, studies have demonstrated that symptoms of urethritis can occur without microscopic evidence of elevated WBCs on Gram stain of urethral swab specimens or in firstvoid urine samples.4,5 It is somewhat easier to establish the diagnosis in men than in women because some pathogens simultaneously infect multiple genital areas, making it difficult for women to localize symptoms.6 In the absence of a documented urinary tract infection, dysuria in a female can represent vulvar inflammation from vaginitis or vulvar dermatoses. Urethral syndrome is a term used for females who have dysuria and pyuria in the presence of sterile urine cultures for bacterial cystitis.6,7

Etiology Infectious Causes

Figure 54-1. Large cauliflower-like genital warts due to papillomavirus with satellite lesions in a 3-year-old girl.

Organisms associated with sexually transmitted infections (STIs) are the most significant etiologic agents in urethritis. With the advent of newer, more sensitive molecular diagnostic testing modalities for STIs, it has become evident that further research is needed to understand better the proportion of cases of urethritis caused by specific pathogens. Males. Chlamydia trachomatis and Neisseria gonorrhoeae are common causes of urethritis in men.1 N. gonorrhoeae has been estimated to cause approximately one-third of the cases of acute urethritis and is differentiated from other causes, which are referred to as nongonococcal urethritis (NGU).8 NGU is the most common clinical STI syndrome in men. Studies have found different rates of specific pathogens depending on geographic location, socioeconomic

358

SECTION

G Urinary Tract Infections

factors, age, race, and sexual orientation or sexual practices.5,8–12 Since the mid-1960s, NGU has increased in frequency more rapidly than has gonococcal urethritis. Although studies have indicated socioeconomic differences between men with gonococcal and NGU, none of these characteristics is useful in distinguishing infectious etiology on an individual basis.1 Co-infection with multiple pathogens can occur and, in 20% to 50% of cases, no pathogen is identified.1,5 Approximately 30% to 50% of NGU in men is caused by C. trachomatis5 and 18% to 46% of men with NGU who are Chlamydianegative have Mycoplasma genitalium.5,10–17 Despite previous literature which seemed to indicate that Ureaplasma urealyticum was the most probable cause of nongonococcal, nonchlamydial infection in men,1 more recent studies of U. urealyticum have yielded conflicting results.5,10,13–15,18–23 Upon further analysis, it appears that Ureaplasma biovar 2 is the biotype most likely associated with urethritis.14,18,19 Trichomonas vaginalis traditionally was considered a less frequent cause of urethritis in men. However, nucleic acid amplification techniques (polymerase chain reaction, PCR) have demonstrated that T. vaginalis is associated more commonly with urethritis than previously expected. Utilizing a variety of detection modalities, T. vaginalis has been identified in 1% to 68% of men with NGU. Many studies report rates in the range of 10% to 20%.9,24,25 T. vaginalis is commonly associated with other STIs and has been demonstrated in the absence of both N. Gonorrhoeae and C. trachomatis.9,26–28 The importance of detecting and treating T. vaginalis has been recognized because of the association of T. vaginalis with enhanced human immunodeficiency virus (HIV) transmission and inceased excretion of the HIV virus in semen of HIV-seropositive men.26 A less frequent cause of urethritis in men includes herpes simplex virus (HSV).5,10,20 Thirty percent of men with primary HSV infection develop urethritis.1 Studies reported in 2006 found that HSV-1 was responsible for more cases of NGU than HSV-2 and HSV-1 was more likely to be associated with men engaging in oral–genital sex as well as men having male partners.5,10 Infrequent causes of urethritis in men include adenoviruses, Haemophilus species, Neisseria meningitidis, and yeast.1,10,29–32 The presence of some of these pathogens suggests that infection with oropharyngeal flora, which are normal nonpathogenic organisms in monogamous partners, may be possible.10 Females. Urethritis in females can be caused by N. gonorrhoeae, C. trachomatis, HSV, and M. genitalium. T. vaginalis typically causes vaginitis in women but is known to infect the urethra and is associated with pyuria.6,9,16,27,33,34

Noninfectious Causes In both males and females, urethritis can accompany noninfectious systemic diseases, such as Stevens–Johnson syndrome, or can result from chemical irritation.

Epidemiology Population based data are available from the year 2001 regarding prevalence rates of N. gonorrhoeae, C. trachomatis, and T. vaginalis. Data were derived from urine-based nucleic acid amplification testing (NAAT) among 18- to 26-year-old subjects who participated in the Add Health study. This study is a population-based, nationally representative sample of over 12 000 youth who were originally in 7th to 12th grade in 1994 and have been followed longitudinally. The rate of C. trachomatis was 3.7% in males and 4.7% in females whereas the rate of N. gonorrhoeae was 0.4% in both males and females.35 T. vaginalis was found in 1.7% of males and 2.8% of females. Prevalence rates of T. vaginalis infection increased with age and were highest among black women who had a prevalence of 10.5% whereas black men had a prevalence of 3.3%. The lowest rates were found among whites, with 1.1% of women and 1.3% of men having T. vaginalis.36 Similar ethnic/racial disparities have been found for C. trachomatis and N. gonorrhoeae. Data from the CDC show that rates of infection tend to be higher among minorities and that the highest rates occur among blacks while the lowest are among whites.3

Population-based studies reflect prevalence in the general population. However, data from the CDC 2004 STI Surveillance Reports illustrate that the STI rates vary and may be higher in specific populations. This information should be considered in the evaluation of patients. For example, infection rates of C. trachomatis among 16- to 24-year-olds in the National Job Training Program are 4.4% to 17.3% for females and 0.8% to 13% for males while rates of N. gonorrhoeae are 0 to 6.4% for females and 1% to 5.5% for males. Juveniles in correction facilities have even higher rates, with a range of 2.4% to 26.5% for C. trachomatis among females and 1.0% to 27.5% for males. Rates for N. gonorrhoeae in this group are 0 to 16.6% for females and 0 to 18.2% for males.3 One study of adolescent women 14 to 18 years of age in clinics and school-based health showed an infection rate of T. vaginalis of 13% among African American females35 whereas another study of 12- to 18-year-old highrisk females at a juvenile detention center found that 48% were positive for T. vaginalis by wet mount examination.37 A study of highrisk adolescent males showed an infection rate of T. vaginalis of 50%.24 Sexual practices and behavior may influence the epidemiology of urethritis. Urethritis due to C. trachomatis, N. gonorrhoeae, or HSV among adolescent women is correlated with having new sex partners.6 Studies of NGU in males have revealed that sexual practices may play a role. One study found that adenovirus and HSV-1 were associated with oral–genital contact and having a male partner, whereas M. genitalium and C. trachomatis were associated with vaginal sex.5 In another study, N. gonorrhoeae and U. urealyticum urethritis were found in heterosexual men, C. trachomatis urethritis was associated with gay and bisexual men, and T. vaginalis was more common in men over the age of 30.21 Urethritis due to STI pathogens can also occur in prepubertal boys and, less frequently, in prepubertal girls. In this age group, transmission commonly results from sexual abuse with genital-togenital contact (see Chapter 58, Infectious Diseases of Child Abuse).

Clinical Manifestations and Differential Diagnosis Symptomatic urethritis in adolescent males is characterized by dysuria, urethral discharge, and/or urethral pruritus. Discharge can be mucoid, mucopurulent, or purulent. Gonococcal urethritis compared with NGU usually has a shorter incubation period, more acute onset, and more profuse discharge (Table 55-1).1,38 Discharge in patients with NGU can be so scant that it is only noted in the morning or is apparent as crusting on the meatus or as stains in underwear.1 Urethral infection with N. gonorrhoeae and the various organisms causing NGU can also be asymptomatic. Among male contacts of women infected with either N. gonorrhoeae or C. trachomatis, only 50% of those men who become infected develop symptoms.1 In another study of male–female dyads in an STI clinic, 20% had unsuspected T. vaginalis infection and among the partners of subjects who were T. vaginalis-positive but N. gonorrhoeae- and C. trachomatisnegative, 11% had C. trachomatis, 5% had N. gonorrhoeae, and 37% had T. vaginalis.39 Urethritis must be differentiated from urinary tract infection, particularly in adolescent boys with dysuria but no discharge. In contrast with urinary tract infection, frequency, hematuria, and TABLE 55-1. Clinical Manifestations of Nongonococcal and Gonococcal Urethritis Nongonococcal Urethritis

Gonococcal Urethritis

Incubation period

2–3 weeks

2–6 days

Onset

Insidious

Abrupt

Dysuria

+; may wax and wane

++; continuous

Discharge

Scant to moderate; may be absent

Profuse; absent in < 10%

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Urethritis, Vulvovaginitis, and Cervicitis

CHAPTER

55

359

TABLE 55-2. Distinguishing Features of Urethritis, Acute Bacterial Cystitis, and Vulvovaginitis in Adolescent Females Urethritis

Acute Bacterial Cystitis

Vulvovaginitis

Symptoms

Internal dysuria

Internal dysuria, frequency, urgency, hematuria

External dysuria, vaginal discharge, vulvar burning, itching

Duration of symptoms

Often > 7 days

Usually < 4 days

May be > 7 days; varies with cause

Signs

Mucopurulent cervicitis Vulvar lesions

Suprapubic tenderness

Vulvar lesions and inflammation; vaginal discharge

Epidemiologic associations

New sex partner Previous STI Sex contact with STI

Previous cystitis Onset of symptoms within 24 hours of intercourse Use of diaphragm Use of a spermicide

History of genital herpes Sex partner with genital herpes Antibiotic use Previous vulvovaginitis Candidiasis

STI, sexually transmitted infection. Adapted from Holmes KK, Stamm WE. Lower genital tract infection syndromes in women. In: Holmes KK, Sparling PF, Mardh P-A, et al. (eds) Sexually Transmitted Diseases, 3rd ed. New York, McGraw-Hill, 1999, pp 761–781.

urgency are uncommon in urethritis. However, if the adolescent male is sexually active, pyuria is more likely to be due to urethritis than to urinary tract infection, since the latter is uncommon in this age group. A focused STI history (see Chapter 53, Sexually Transmitted Disease Syndromes) and past medical history can help establish whether a low or high risk of urethritis or of urinary tract infection exists. In adolescent girls, dysuria is usually the cardinal feature of urethritis. The urethral syndrome must be differentiated from acute bacterial cystitis and vulvovaginitis (Table 55-2). The literature describes differentiation between “internal” and “external” dysuria. Internal dysuria is pain that is felt internally during voiding. External dysuria is discomfort that is felt as urine passes over the labia.7 Internal dysuria, urinary frequency, and isolation of ≥102 uropathogens per milliliter of voided urine suggest acute bacterial cystitis; isolation of < 102 uropathogens per milliliter suggests acute urethritis due to STI pathogens.6,40 Pain that is felt internally only at the end of urination is consistent with bacterial cystitis.7 External dysuria can occur with vulvovaginitis. However, relying on the distinction between internal and external dysuria could be problematic. Adolescent females can have vaginitis alone or a concurrent urinary tract infection, and they may not be able to adequately distinguish between internal and external dysuria.7 Any female suspected of having urethritis requires an STI-directed history and physical examination to determine whether other STI or STI syndromes (e.g., pelvic inflammatory disease (PID)) are present (see Chapter 53, Sexually Transmitted Disease Syndromes). In prepubertal boys and girls, urethritis due to STI pathogens can manifest with dysuria and urethral or vaginal discharge. There may be vague lower abdominal pain, unwillingness to void, and, in boys, irritation in the distal urethra or meatus. Dysuria in a prepubertal child is much more likely to be due to urinary tract infection than urethritis associated with STI. Urethritis is more probable in the presence of a discharge or a history of sexual abuse, especially if genital-to-genital contact has occurred (see Chapter 58, Infectious Diseases of Child Abuse).

Diagnosis Males In males, specimens are obtained both to document urethritis and to detect common causes, N. gonorrhoeae and C. trachomatis. Definitive diagnosis is enhanced if the patient has not voided recently; examination in the morning without voiding is ideal.1 When discharge is present, a meatal swab specimen can be taken for Gram stain. A smear showing gram-negative intracellular diplococci of the typical kidney-bean morphology pattern (Figure 55-1A), or 5 or more neutrophils per oil immersion field (μ1000) is diagnostic of urethritis.1,41 Gram-stain smear is sensitive and specific in diagnosing

gonococcal urethritis if intracellular diplococci are detected. If Gram stain is equivocal or negative, culture of the specimen or NAAT for N. gonorrhoeae is indicated. In all patients, regardless of whether N. gonorrhoeae is suspected by Gram stain, an intraurethral specimen should be obtained for detection of C. trachomatis by culture, antigen detection, or nucleic acid amplication method, or a first-voided urine specimen should be tested by nucleic acid amplication (see Chapter 167, Chlamydia trachomatis). Although culture had previously been considered the “gold standard” for diagnosing C. trachomatis infection, NAAT (e.g., PCR, strand displacement amplification, or transcription-mediated amplification) are more sensitive than culture and the preferred diagnositic method.42 The urinary LE dipstick technique for detecting pyuria has been used as a screening tool for identifying asymptomatic urethritis in adolescent males without risk factors according to the patient history.43,44 If this test was positive, testing for N. gonorrhoeae and C. trachomatis was performed. However, in many studies, LE testing is not sensitive enough to be considered reliable.45 The availability of molecular amplification techniques now presents the possibility of noninvasive testing on urine using assays that have excellent sensitivity and specificity. LE testing may be useful in situations in which the new tests are cost-prohibitive.45 Detection of asymptomatic infection is important because asymptomatically infected males are an important source of transmission of N. gonorrhoeae and C. trachomatis. The diagnosis of T. vaginalis in men has proven more challenging than in women. Studies have shown that the use of PCR to diagnose T. vaginalis in males is superior to culture or wet mount.25–27,46 An advantage of PCR is that it can be performed on a urine specimen. Several studies have demonstrated that PCR testing of urine is superior in sensitivity to testing of urethral specimens.25–27 The sensitivity of PCR on urine specimens has ranged from 93% to 100%.27 At this time, PCR remains a research tool and is not commercially available. Diagnosis of T. vaginalis must rely on wet mount of a urethral smear or culture. Wet mount only detects 30% of T. vaginalis infections in men and is not utilized commonly.26 Culture can be performed on either urine or urethral specimens and appears to yield better results if both are tested.27 The InPouch culture system (BioMed Diagnostics, San Jose, CA) has proven equivalent to the traditional gold-standard Diamond’s media and can be used on urethral or urine sediment specimens.24 A study comparing InPouch T. vaginalis culture and of first-voided urine sediment in males found that culture detected 28% of cases whereas PCR detected 94% of cases.46 Another study found that PCR detected three times as many cases of T. vaginalis (17% prevalence) as compared to 5% by InPouch culture in males.25 Cultures for U. urealyticum are not readily available and M. genitalium is difficult to culture.1,16 Both of these organisms are diagnosed and evaluated utlizing PCR techniques in research studies. At this time PCR is not available commercially for either organism

360

SECTION

G Urinary Tract Infections

A

B Figure 55-1. (A) Gram stain of urethral discharge from an adolescent male with urethritis showing multiple neutrophils and intracellular diplococci with kidney-bean morphology typical of Neisseria gonorrhoeae (magnification ¥1000). (B) Gram stain of vaginal fluid from an adolescent with bacterial vaginosis showing clue cells and squamous vaginal epithelial cells covered with coccobacilli, which gives them a stippled or granular appearance. Note the absence of rods with blunt ends (lactobacilli) (magnification ¥2000). (C) Wet mount of vaginal secretions in an adolescent female with bacterial vaginosis showing clue cells. Note stippled epithelial cells with ragged (bacteria-covered), ill-defined borders (magnification ¥200).

C

and treatment relies on the clinical presentation and ruling out other definable organisms.

Females Endocervical and urethral specimens should be obtained for culture or NAAT for both N. gonorrhoeae and C. trachomatis in adolescent girls with urethritis, because concurrent infection is common.6 Urinalysis and urine culture also are indicated, because simultaneous urinary tract infection and STI diagnosis can occur in sexually active females.33 Studies have indicated that urine testing using NAAT for C. trachomatis and N. gonorrhoeae may not be as sensitive as endocervical specimens in females.42,47 Although studies using NAAT testing vaginal specimens have shown good correlation with cervical specimens, not all NAATs are approved by the Food and Drug Administration for use on vaginal specimens. One advantage of vaginal specimens is that swabs can be self-collected by the patient and patient-collected vaginal samples have similar sensitivities to vaginal swabs collected by a clinician.42,47 NAAT is not recommended for testing of specimens obtained by rectal or pharyngeal swabs. T. vaginalis can be diagnosed by a variety of methods, including wet mount, culture, DNA probe, or antigen-based rapid testing.24 As in males, wet mount has poor sensitivity compared to other methods. In prepubertal children, some experts recommend meatal, rather than intraurethral, specimens because the former are less painful to obtain.48

Management Initial treatment in males is based on Gram stain results, if performed (Table 55-3).2,41 Patients with Gram-stain evidence of N. gonorrhoeae

should be treated with single-dose therapy, orally or intramuscularly. All patients should be tested for C. trachomatis and treated if positive, but if C. trachomatis testing is not possible, empiric treatment is indicated because coinfection is common. For NGU, single-dose azithromycin therapy may be preferred over a week of therapy with doxycycline in adolescents because of compliance issues.2 Alternative treatments include erythromycin or a fluoroquinolone. In addition, confirmed cases of N. gonorrhoeae or C. trachomatis must be reported to the local health authorities, and sexual partners should be contacted for assessment and treatment (see Chapter 53, Transmitted Disease Syndromes). Follow-up and repeat testing for N. gonorrhoeae or C. trachomatis urethritis in adolescents are not routinely recommended if appropriate treatment is completed, if symptoms and signs disappear, and if no re-exposure to an untreated partner occurs.2,41 However, if symptoms persist or recur, the patient should be instructed to return for evaluation. If adherence to therapy was poor or re-exposure was probable, patients are retreated with the initial regimen. If symptoms persist despite good adherence and no re-exposure, culture of an intraurethral specimen and first-void urine for T. vaginalis should be performed because of the high prevalence of T. vaginalis infection in men with nongonococcal, nonchlamydial urethrits as well as coinfection with N. gonorrhoeae and C. trachomatis. Treatment for T. vaginalis as part of the initial treatment for male urethritis has been suggested,28 but that recommendation is not part of current STI treatment guidelines. If T. vaginalis is diagnosed, metronidazole therapy is prescribed (see Chapter 274, Trichomonas vaginalis). However, T. vaginalis culture may miss a significant portion of the cases and current CDC guidelines recommend treatment with metronidazole or tinidazole and azithromycin if not used for the initial episode in all cases of persistent urethritis.1,2,11

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Urethritis, Vulvovaginitis, and Cervicitis

CHAPTER

55

361

TABLE 55-3. Treatment of Urethritis Gram Stain

Suggested Treatment

RESULTS AVAILABLE

Increased Neutrophils Present

Gram-Negative Intracellular Diplococci Present

Present

Absent

Absent

Absent

Treat for gonococcal and chlamydial urethritis. For Neisseria gonorrhoeae: for children and adolescents, ceftriaxone, 125 mg IM once, or for adolescents, cefixime, 400 mg PO once.a For Chlamydia trachomatis: for adolescents, azithromycin, 1 g PO once, or doxycycline, 100 mg bid PO for 7 days. For children < 8 years of age, erythromycin base, 50 mg/kg per day in four divided doses for 10 to 14 days (maximum 2 g/day); for children < 8 years but ≥ 45 kg, azithromycin, 1 g PO once Treat for nongonococcal urethritis. Azithromycin once, or doxycycline for 7 days, or ofloxacin 300 mg PO twice a day for 7 days or levofloxacin 500 mg PO once a day for 7 days or erythromycin base 500 mg PO four times a day for 7 days; for children 45 kg; and azithromycin or doxycycline for older females; an antifungal agent such as clotrimazole for Candida or oral fluconazole if topical therapy is not effective; and penicillin for group A streptococcus. Other antimicrobial agents are chosen on the basis of vaginal bacterial culture results.2,33 If no etiology is determined and a precipitating cause of nonspecific vulvovaginitis is not suggested by history or physical examination, treatment outlined for nonspecific vulvovaginitis is recommended.

Follow-Up and Sequelae The follow-up of children with vulvovaginitis depends on the etiology. Fortunately, gonococcal and chlamydial vulvovaginitis are rarely associated with upper tract disease, especially if treated early.33 Thus, impairment of fertility is unlikely if adequate treatment is given. Sequelae due to other causes of vulvovaginitis depend on the specific agent.

Vaginitis in Pubertal Females Etiology and Epidemiology With puberty, under the influence of estrogen, the vaginal epithelium shifts from cuboidal to a glycogen-containing stratified squamous state.6,33 There is an associated increased growth of lactobacilli, which produce lactic acid from glycogen. This results in a decrease in vaginal fluid pH from a prepubertal level of about 7.0 to 4.0 to 4.5. The lower pH, associated with hydrogen peroxide production by lactobacilli, is important in the regulation of vaginal flora.33 Changes in the type of

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Urethritis, Vulvovaginitis, and Cervicitis

epithelium and colonizing flora render the vaginal environment relatively resistant to infection caused by C. trachomatis and N. gonorrhoeae. Thus, in adolescents, these two organisms cause cervicitis, rather than vulvovaginitis. Leukorrhea, the normal white mucous vaginal discharge that represents the effect of estrogen on the vaginal mucosa in adolescents, must be distinguished from pathologic discharge. Saline wet-mount examination of vaginal secretions would reveal sheets of epithelial cells without inflammatory cells, yeast, clue cells, or trichomonads. Leukorrhea is sometimes considered excessive by patients, and they need reassurance.33 The major causes of vaginitis in adolescents are bacterial vaginosis (BV), candidiasis, and T. vaginalis infection. Bacterial vaginosis has replaced the term nonspecific vaginitis, because this condition is not generally inflammatory but arises from a change in the vaginal flora.33,58 Less common causes of vaginitis in adolescents include ulceration and infection associated with tampons, cervical caps, and other foreign bodies; chemical agents such as those found in douches and spermicides; and toxin-producing Staphylococcus aureus.6,33 Bacterial Vaginosis. BV is the most common cause of vaginitis in postpubertal women.59 It represents a disruption in the normal vaginal flora, with a decrease in lactobacilli and overgrowth of a variety of organisms that can include G. vaginalis, genital mycoplasmas (M. hominis), and anaerobic bacteria, particularly Bacteroides and Mobiluncus species.1,24,60,61 Although many patients with BV have moderate-to-heavy concentrations of vaginal Gardnerella species, detection of this organism is not diagnostic because vaginal colonization with small numbers is common in both sexually active (30% to 60%) and nonactive (20% to 30%) adolescents without BV.58,62 New technologies utilizing nucleic acid amplification of ribosomal DNA are being utilized to characterize bacterial species. One study found newly recognized bacteria in the Clostridiales order that were prevalent in subjects with BV but uncommon in healthy controls.63 In another study, an anaerobic bacteria, Atopobium vaginae (which is resistant to metronidazole) was identified in association with BV.24 Whereas BV is more common in adolescents and young adults who are sexually active and have multiple sexual partners, designating BV as an STI is controversial.58,62,64–67 The prevalence of BV among women attending STI clinics is higher (24% to 37%) than among college women attending student health clinics for routine annual examination (4% to 15%).33,58 However, studies have indicated that BV occurs in women who have not been sexually active.68 One study of women entering the military found that 19% of subjects denying a history of vaginal intercourse met criteria for BV compared to 28% who had been active sexually.65 Although sexual activity and multiple sexual partners are associated with BV,66,67 treatment of sex partners does not appear to prevent recurrence.60 In addition, douching is associated with BV.66,67 Candidiasis. Vulvovaginal candidiasis has an estimated lifetime incidence of 75%59 (see Chapter 244, Candida Species). Candidal colonization of the vagina usually originates from the gastrointestinal tract, and sexual transmission is not an important mode of acquisition.62 As many as 20% of women are colonized with Candida albicans, which is the organism found in most uncomplicated cases. Risk factors include sexual activity, receptive oral sex, oral contraceptive pills, and spermicide use. Antibiotic use is frequently mentioned, but it is not a major cause of infection in most women.59,69 A small group of adolescents and women have recurrent or chronic candidal infection. Approximately 5% of women have complicated vulvovaginal candidiasis, with more than four episodes per year, severe infection, or infection by Candida non-albicans species and Saccharomyces cerevisiae.59 The most common C. non-albicans species reported is C. glabrata.69 The term recurrent vulvovaginal candidiasis (RVVC) is used to describe the clinical presentation of four or more symptomatic episodes of Candida vaginitis over the course of a year.69 Although most of these women do not have diabetes mellitus or immunosuppression, these conditions increase the incidence of RVVC.69 Complicated infections are characterized by a transient and incomplete response to standard therapy rather than reinfection.59 Women with RVVC are more likely to be colonized by Candida

CHAPTER

55

363

species from the vagina or gastrointestinal tract. Further, some patients with RVVC have demonstrated an immediate hypersensitivity reaction to Candida species, which has been hypothesized to represent an organ-specific, antigen-specific abnormality of the lower genital tract that results in vulvovaginal symptoms.69 Trichomonas vaginalis. T. vaginalis is found in 5% to 10% of the general population and 18% to 50% of women with vaginal complaints. This STI has a particular predilection for the squamous epithelium of the genitourinary tract.9 As in other STIs, asymptomatic infection occurs.

Clinical Presentation and Diagnosis Clinical presentation and laboratory features that help distinguish BV from candidal vulvovaginitis and trichomonal vaginitis are shown in Table 55-4. In the sexually active adolescent, cervicitis must also be excluded because cervicitis can occur as a coinfection or as the sole cause of the vaginal discharge.70 Symptomatic BV presents with a thin, white homogeneous vaginal discharge. Women with symptomatic candidal infection commonly complain of vaginal pruritus and burning, dysuria, and dyspareunia. Discharge can be thick and white and is not usually malodorous. Approximately 50% of patients do not complain of discharge.59 Patients with trichomoniasis often have pruritus and a malodorous, frothy discharge, but dysuria, dyspareunia, vulvar erythema and edema, and bloody vaginal discharge can also occur. Bacterial Vaginosis. The diagnosis of BV is established by the presence of at least three of the following four criteria (Amsel criteria): (1) thin, homogeneous vaginal discharge; (2) vaginal pH greater than 4.5; (3) characteristic fishy “amine” odor released when alkali (10% KOH weight/volume) is added to the vaginal fluid specimen (positivewhiff test); and (4) at least 20% of epithelial cells having the appearance of “clue cells.” Clue cells are stippled epithelial cells whose borders are obscured by adherent bacteria, which include G. vaginalis and other organisms (Figure 55-1B and C).71 Accuracy of diagnosis increases with use of the Amsel criteria.24 Alternative tests include use of Gram stain to group bacteria into morphologic types (Nugent’s scoring). Amsel criteria and Nugent scores show good correlation.72 Microscopy and pH paper are frequently not available in many ambulatory office settings. Alternative testing includes point-of-care testing for BV that does not require a microscope. One point of care test is BVBlue (Genzyme Diagnostics, Cambridge, MA), which detects sialidase – an enzyme produced by some of the anaerobic bacteria found in BV. This test was sensitive and specific compared with Amsel and Nugent criteria.73 A DNA probe for G. vaginalis ribosomal RNA (Affirm VP III DNA probe, Becton Dickinson, Franklin Lakes, NJ) is available but not useful if rapid results are needed. This test is most helpful as a supplemental marker to support the diagnosis of BV by detecting the presence of G. vaginalis.24,72 Usually, BV does not produce an inflammatory response. The presence of elevated WBCs on vaginal smear indicates concurrent vaginitis or cervicitis associated with yeast, T. vaginalis, C. trachomatis, or N. gonorrhoeae.72,74 Routine aerobic and anaerobic vaginal cultures are not helpful. Candida Species. Candida vaginitis is diagnosed by demonstrating hyphae and blastospores on microscopic examination of a saline or 10% KOH preparation.59 This technique has a sensitivity of about 50%. C. glabrata and S. cerevisiae only produce blastospores and are more easily missed.59 If the KOH preparation is negative, culture may confirm the diagnosis. Culture of vaginal, rather than cervical, specimens is more sensitive and may be particularly helpful in patients with ongoing nonspecific symptoms in whom BV and trichomoniasis have been excluded.75 In cases of complicated vulvovaginal candidiasis, culture is important in identifying the species of the organism because a longer course of therapy or alternative medications may be needed.59 It is also useful to measure vaginal pH since, unlike BV or trichomoniasis, the vaginal pH remains low in vaginal candidiasis.69 Trichomonas vaginalis. Diagnosis of trichomoniasis is made by visualizing ovoid, motile organisms, slightly larger than neutrophils,

364

SECTION

G Urinary Tract Infections

TABLE 55-4. Characteristics and Recommendations for Treatment of Vaginitis in Adolescents Bacterial Vaginosis

Candidal Vaginitis

Trichomonal Vaginitis

Malodorous discharge

Vaginal discharge Usually not malodorous Vulvar itch or discomfort Dysuria Dyspareunia Often exacerbated symptoms just prior to menses

Vaginal discharge May be malodorous Vulvar itch Dysuria Dyspareunia

Discharge

Thin, homogeneous, white, clings to vaginal wall; ± frothy

Thick, curdlike

Heavy, gray or yellow-green; frothy

Other signs

None

Vulvar and vaginal erythema, vulvar edema

Vulvar and vaginal erythema

pH

> 4.5

< 4.5

> 4.5

KOH preparation

“Fishy,” amine odor when mixed with 10% KOH (positive-whiff test)

Hyphae or pseudohyphae

Occasionally positive-whiff test

Saline preparation

“Clue cells,” few neutrophils

Neutrophils and epithelial cells in equal numbers

Motile trichomonads, neutrophils

Gram stain

Few gram-positive bacilli; abundant mixed flora

Hyphae or pseudohyphae

Trichomonads visualized rarely

Culture

Not useful

Can be useful if KOH-negative

Culture more sensitive than wet mount

TREATMENT

Oral Metronidazole, 500 mg bid for 7 days, or Clindamycin, 300 mg bid for 7 days

Topical intravaginala Butoconazole cream Clotrimazole cream or vaginal tablet Miconazole cream or vaginal suppository Terconazole cream or vaginal suppository Tioconazole ointment Oral Fluconazole, 150 mg once

Oral Metronidazole, 2 g for 1 dose, or 500 mg bid for 7 days or Tinidazole 2 g orally in a single dose

SYMPTOMS

SIGNS

LABORATORY FINDINGS

Topical intravaginal Clindamycin cream 2%, one full applicator amount per vagina at bedtime for 7 days Metronidazole, 0.75% gel, one full applicator amount per vagina once a day for 5 days KOH, potassium hydroxide. a Intravaginal therapies are available in 1- to 7-day regimens.

on the wet mount.76 Wet mount misses the diagnosis 30% to 50% of the time. Culture using Diamond media is more sensitive but is not widely available. The InPouch T. vaginalis culture system (BioMed Diagnostics, San Jose, CA) may provide an alternative to standard culture because it is equivalent to Diamond media and is easier to store and transport.24,77 Culture can be performed on vaginal specimens, including patient self-collected specimens. Conventional Papanicolaou (Pap) smear identification has a high false-positive rate.78,79 However, liquid-based Pap smear appears to be more accurate and in one study had a positive predictive value of 96% and a negative predictive value of 91% compared with culture.80 A DNA probe test (Affirm VP, Becton Dickinson, Franklin Lanes, NJ) is sensitive and specific but is generally best performed in the laboratory rather than the office setting because the test is moderately complex and requires about 45 minutes to complete.24,81 A newer point-of-care test, utilizing an immunochromatographic capillary flow assay with monoclonal antibodies (OSOM Trichomonas Rapid Test, Genzyme Diagnostics, Cambridge, MA) was shown in a multicenter study to have a positive predictive value of 96%, negative predictive value of 95%, and 84% sensitivity compared with culture.82 PCR has excellent sensitivity and specificity but is not available commercially. Studies in women have shown that vaginal PCR specimens are superior to urine, and patient self-collected vaginal specimens in adolescents as well as adults can be utilized.9,76,83,84 Vaginal specimens collected without the use of a speculum for microscopic evaluation to detect trichomoniasis, BV, and vulvovaginal

candidiasis are comparable in sensitivity with specimens obtained during speculum examination. Because NAAT for N. gonorrhoeae and C. trachomatis can be performed on urine and vaginal swabs, it may be possible to avoid the more invasive speculum examination in determining an etiology for vaginitis in adolescents.85

Treatment Orally administered metronidazole (1 g/day in two divided doses for 7 days) is the recommended therapy for BV.2 Single-dose metronidazole therapy is no longer recommended because it is not as effective as treatment administered for 1 week, and recurrent infection is common regardless of therapeutic choice.2,24,60 Alternative regimens, including use of vaginal clindamycin or metronidazole preparations, are outlined in Table 55-4. Topical treatments have fewer gastrointestinal tract side effects but increase the risk for vaginal candidiasis24 and are not recommended in pregnancy. Use of 2% clindamycin cream intravaginally was associated with the onset of premature labor. Clindamycin cream should not be used when condoms are used because the oil base weakens the latex.2,24 Some women experience recurrent episodes of BV. One randomized trial of nonpregnant women treated with either intravaginal metronidazole or clindamycin found high baseline (17%) and posttreatment (53%) resistance to clindamycin in vaginal anaerobic bacteria that persisted for 3 months after treatment. This finding compares with < 1% resistance to metronidazole at baseline

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Urethritis, Vulvovaginitis, and Cervicitis

and no increase after treatment.86 There are no current recommendations for treatment of recurrent infection. However, a large multicenter study demonstrated 70% efficacy in preventing recurrence with twice-weekly metronidazole vaginal gel while on maintenance therapy. For many subjects, the suppression of BV only lasted while they were on this regimen.24 It is not clear that treatment of asymptomatic BV in nonpregnant women is useful. One randomized, placebo-controlled, double-blind study comparing metronidazole gel and placebo found no statistically significant differences in symptom resolution.87 For candidal vulvovaginitis, topical therapy with azoles, such as clotrimazole, terconazole, miconazole, butoconazole, and tioconazole, is effective. The efficacy of the use of single-dose fluconazole (150 mg) is comparable with that of topical therapy.2,69 Sexual partners usually do not require therapy unless candidal balanitis is present. Recurrent or chronic candidal vulvovaginitis (see Chapter 244, Candida Species) merits investigation for predisposing conditions and may require oral therapy with ketoconazole, fluconazole, or itraconazole if a repeated course of topical therapy fails to lead to resolution of signs and symptoms.69 Topical boric acid in the form of suppositories and topical flucytosine have been useful in patients with Candida non-albicans species and azole-resistant C. albicans. Resistance to azole agents is uncommon among C. albicans and more likely in C. non-albicans species. Alternative therapies are useful, especially in cases of C. glabrata and C. krusei.59,69,88 Patients with RVVC can be treated initially with 10- to 14-day courses of conventional topical therapy or a two-dose course of 150 mg fluconazole 72 hours apart.88–90 In women who continue to have recurrent episodes, longer-term maintenance therapy has been successful, utilizing weekly fluconazole,91 daily ketoconazole, or topical therapy with azoles or nystatin.69 Suppressive therapy is usually continued for 6 months, but recurrence is common after therapy is discontinued because organisms are suppressed rather than eradicated. In some cases, suppressive therapy has been continued for several years.69 Systemic therapy with oral metronidazole or tinidazole is indicated for treatment of trichomoniasis.2 Metronidazole has a 95% cure rate with either a 7-day course (500 mg twice daily) or a large single dose (2 g orally). Although uncommon, metronidazole resistance has been reported in approximately 1% to 5% of cases.24,78 Treatment with higher doses of metronidazole and longer courses as well as use of tinidazole (tinidazole, 2 g orally in a single dose) is reported.2,24,59,92 One study found a 92% cure rate utilizing a combination of oral and vaginal tinidazole in patients unresponsive to metronidazole.92 Tinidazole appears to be better tolerated than metronidazole and has fewer gastrointestinal tract and central nervous system side effects. Topical therapy with metronidazole gel is not effective because the gel does not adequately penetrate the urethral and perivaginal glands.2 Sexual partners should be evaluated for STIs and treated for trichomoniasis.

Complications and Outcome BV has been associated with chorioamnionitis, postpartum endometritis, posthysterectomy vaginal cuff cellulitis, postabortion PID, premature rupture of membranes, preterm labor and delivery, and spontaneous abortion.24,60,68,93 The risk of preterm delivery is restricted to a small subset of women. These differences may be related to host response to inflammation and cytokine production and the response may be genetically determined.24,68 There have been conflicting data on whether treatment of pregnant women for BV prevents these complications.24,68,94 Some studies have indicated that treatment of asymptomatic pregnant women with previous preterm delivery reduces this complication. Data on women without a history of preterm delivery are less clear and screening of all pregnant women is not currently recommended.2,68 Current CDC guidelines recommend testing and treatment of all symptomatic pregnant women.2 Studies have also shown an association between BV and PID but a causal relationship remains unproven. Organisms associated with BV

CHAPTER

55

365

have been found in the upper genital tracts of women with PID, women with BV and more likely to have PID, and women with PID are more likely to have BV.24,68 Treatment of BV has been associated with decreased PID in women undergoing abortion and decreased postoperative cuff cellulitis.68 However, the relationship between BV and PID for women not undergoing abortion or uterine instrumentation is not clear.24,68,95 There is no current recommendation for treatment of nonpregnant asymptomatic women as a standard clinical practice.2,60 Trichomoniasis has been associated with premature rupture of membranes and preterm delivery.24 T. vaginalis may also play a role in the development of PID by acting as a vehicle transporting other pathogens to the upper genital tract.24,33 Controversy exists about the treatment of T. vaginalis during pregnancy, since treatment does not appear to reduce the complications. Furthermore, although treatment will eliminate the organism, one study showed an increased risk of preterm birth in treated women, which raises concern for the treatment of all pregnant women, including women who are symptomatic.26,27,94,96 Current CDC recommendations do not recommend screening and treatment for asymptomatic T. vaginalis infection in pregnancy.2 Both BV and T. vaginalis enhance acquisition of HIV and transmission to a partner.24,27,68 Treatment of T. vaginalis reduces HIV viral shedding in vaginal secretions. However, there are no studies to date that have demonstrated similar effects on viral shedding in subjects with BV.24

Vulvitis in Adolescents Inflammation of the vulva in adolescents is most commonly due to HSV and yeasts (see Table 55-4). HSV often causes painful genital ulcers, along with vulvar inflammation and inguinal lymphadenopathy (see Chapter 54, Skin and Mucous Membrane Infections and Inguinal Lymphadenopathy, and Chapter 204, Herpes Simplex Virus).97 Occasionally, inflammation can be associated with T. vaginalis infection.

CERVICITIS Cervicitis is inflammation of the endocervix or ectocervix, or both. Both are common problems among adolescents, but neither is common in prepubertal girls. Under the influence of estrogens following puberty, the vaginal epithelium and ectocervix become cornified and thus relatively resistant to infection with a number of pathogens, including N. gonorrhoeae and C. trachomatis.7 By contrast, the endocervix continues to be lined with columnar epithelium and remains susceptible to infection with these organisms. Therefore, in adolescent and adult women, these organisms usually cause endocervicitis in the absence of vaginitis. A normal developmental finding in adolescents is the presence of the ectropion, which appears as an erythematous area surrounding the os. This represents the area of the squamocolumnar junction, with the erythematous area corresponding to the columnar epithelium and the surrounding pink area corresponding to the stratified squamous epithelium. During adolescence, the ectropion recedes as the result of squamous metaplasia. Although some adolescents with a large ectropion may have significant vaginal discharge, this is not a pathologic or infectious process. The ectropion is not usually friable; the presence of edema or friability suggests infection.6

Ectocervicitis Ectocervicitis represents infection of the stratified squamous epithelium of the ectocervix. Ectocervicitis can occur in conjunction with candidal vulvovaginitis, trichomonal vaginitis, and HSV infection. HSV causes both ectocervicitis and endocervicitis6,97–100 (see Chapter 29, Pharyngitis).

366

SECTION

G Urinary Tract Infections

Endocervicitis Etiology and Epidemiology Endocervicitis represents infection of the endocervical columnar epithelium and can produce mucopurulent cervicitis. Common pathogens that cause endocervicitis are N. gonorrhoeae and C. trachomatis.99 Other agents include HSV, which often causes concurrent exocervicitis, and actinomycetes, which are frequently associated with an intrauterine device.6,98–100 Studies utilizing PCR technology demonstrate the association between Mycoplasma genitalium and mucopurulent cervicitis.16,34,99,101,102 There is also a possible association with BV; cervicitis was more likely to resolve when patients were also treated for BV.24,99 Possible theories for this relationship include the presence of proinflammatory vaginal cytokines in patients with BV and the presence of glucosidases and proteinases produced by BV-associated organisms that may degrade cervicovaginal mucus.99 Endocervicitis is often overlooked and underdiagnosed because symptoms and signs can be mild or absent. PID is one consequence of untreated mucopurulent cervicitis; thus, every effort must be made to detect and treat this infection.6 Sexually active adolescents with vaginal discharge; lower abdominal pain of recent onset; intermenstrual, postcoital, or prolonged abnormal vaginal bleeding; or deep dyspareunia should be evaluated promptly for endocervicitis.103,104 Evaluation is also indicated if a sexual partner has an STI. Infectious STI agents are not solely responsible for clinically apparent cervicitis. Other entities that should be considered include noninfectious agents such as infectious systemic inflammatory processes (e.g., sarcoidosis, tuberculosis, Behçet disease) and agents that cause local insults (e.g., chemical douches, spermicides, foreign bodies).99,100

Clinical Manifestations Cervical abnormalities associated with endocervicitis range from subtle changes to the presence of intensively hyperemic, erythema-

tous, raised, irregular, friable lesions.103 There is no consensus deÀnition for mucopurulent cervicitis, which makes evaluation of the research literature difÀcult.99 Current deÀnitions include inflammation of the endocervix with possible edema; yellow-green endocervical discharge; increased numbers of neutrophils on microscopic examination of cervical secretions; and inducible endocervical bleeding. Mucopus is characterized by a yellow or green color on a cottontipped applicator obtained from the endocervix. The number of neutrophils considered signiÀcant varies in different studies from ≥ 30 cells per 400μ magniÀed microscopic Àeld to > 10 cells per 1000μ microscopic Àeld.2,6,98–100

Diagnosis SpeciÀc microbiologic diagnosis leads to appropriate treatment. Patients should be tested for N. gonorrhoeae, C. trachomatis, and T. vaginalis. If HSV is suspected, a culture should be obtained. Typing HSV strains differentiates between HSV-1 and HSV-2 isolates. Saline wet mount of vaginal secretions can be utilized both to detect T. vaginalis and to help establish the diagnosis of BV.2,6,33,98,99 Because of the poor sensitivity of wet mount, further testing for T. vaginalis by culture or antigen-based tests should be conducted if the wet mount is negative. PCR testing for M. genitalium is not available commercially. Gram-negative intracellular diplococci are seen on Gram stain in about one-half of cases of gonococcal endocervicitis.6 Given the poor sensitivity of the Gram stain and the possibility of infection in the absence of any abnormality, evaluation for gonococcal endocervicitis must include an appropriate culture or NAAT. Inflammatory changes may be even less remarkable with endocervicitis caused by C. trachomatis. Therefore, NAAT, chlamydial culture, or other chlamydial detection tests are recommended in all cases of suspected infection2,6,33,99 (see Chapter 167, Chlamydia trachomatis). NAAT is the most sensitive test available for diagnosing C. trachomatis. Chlamydial endocervicitis can stimulate epithelial and inflammatory changes, which can be seen on Pap smear.6 However, Pap smear should not be relied on to assess for chlamydial infection because its sensitivity, speciÀcity, and positive predictive values are poor.99

TABLE 55-5. Treatment of Cervicitis in Adolescents Gram Stain

Suggested Treatment

RESULTS AVAILABLE

Mucopurulent endocervical discharge and smear with many neutrophils on Gram stain

Treat for Neisseria gonorrhoeae and Chlamydia trachomatis (ceftriaxone, 125 mg IM in a single dose or ceÀxime, 400 mg PO as a single dose, or ciprofloxacin, 500 mg PO as a single dose, or ofloxacin 400 mg PO as a single dose or levofloxacin 250 mg PO as a single dose, plus azithromycin, 1 g in a single dose, or doxycycline, 100 mg PO bid for 7 days)

Gram-negative intracellular diplococci on smear regardless of presence or absence of endocervical discharge or neutrophils on smear

Treat for Neisseria gonorrhoeae and Chlamydia trachomatis

Mucopurulent endocervical discharge and smear without neutrophils

Defer therapy until further microbial results are available or If patient is high-risk by history (e.g., known contact), consider treatment for Neisseria gonorrhoeae and Chlamydia trachomatis unless follow-up can be ensured

Clinical presentation suggestive of herpes simplex virus infection

Consider oral acyclovir or famciclovir or valacyclovir

RESULTS NOT AVAILABLE

Mucopurulent discharge and at least one of the following: Friable area on cervix Edema or erythema in an area of ectopy Patient in high-risk group (see Chapter 53, Sexually Transmitted Infection Syndromes)

Treat for Neisseria gonorrhoeae and Chlamydia trachomatis

No endocervical discharge

Defer therapy until further microbiologic results available Or If patient is high-risk by history (e.g., known contact), consider treatment for Neisseria gonorrhoeae and Chlamydia trachomatis unless follow-up can be ensured Consider oral acyclovir or famcicovir or valacyclovir

Clinical presentation compatible with herpes simplex virus infection

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Pelvic Inflammatory Disease

CHAPTER

56

367

Management The initial management of endocervicitis varies, depending on the clinical findings and the risk of the adolescent for certain STIs (Table 55-5). A positive laboratory test for N. gonorrhoeae, C. trachomatis, T. vaginalis, or HSV defines treatment. However, empiric treatment can be considered if there is a high suspicion of N. gonorrhoeae or C. trachomatis, based on high prevalence of either microbe in the community or a concern that the patient may be at risk for loss to follow-up.2 Treatment of BV can also be considered since higher rates of resolution of endocervicitis occurred with treatment of BV in some studies.99 More data are needed regarding treatment of M. genitalium cervicitis. Although theoretically susceptible to macrolides, fluoroquinolones, and tetracyclines, a 5-day course of azithromycin may be most effective in eradicating the M. genitalium.105 Follow-up after completion of therapy is recommended for adolescents with persistent symptoms.2 The management of mucopurulent cervicitis also requires evaluation and treatment of all sexual partners for STIs, and provides an opportunity to reinforce STI prevention measures (see Chapter 53, Sexually Transmitted Disease Syndromes).

Sequelae Sequelae from untreated mucopurulent cervicitis include PID as well as possible long-term sequelae of ectopic pregnancy and infertility The relationship between N. gonorrhoeae and C. trachomatis and PID is well established and there is now increasing evidence that M. genitalium is also associated with PID.105–107 (see Chapter 56, Pelvic Inflammatory Disease).

CHAPTER

56

Pelvic Inflammatory Disease

A

Figure 56-1. Examination of an adolescent girl shows an area of cervical ectopy (A), where the columnar cells that line the inner area of the cervix extend to the outer area.

TABLE 56-1. Frequency of Microorganisms Isolated from the Endometrium of Women with Clinically Diagnosed Pelvic Inflammatory Disease Microorganism Bacterial vaginosis (Gram stain) Gardnerella vaginalis Any anaerobic gram-negative bacillus Any anaerobic gram-positive coccus Neisseria gonorrhoeae Viridans streptococcus Diphtheroids Chlamydia trachomatis

Percentage of Women 54 31 22 16 13 13 11 10

Adapted from Haggerty CL, Hillier SL, Bass DC, et al. Bacterial vaginosis and aerobic bacteria are associated with endometritis. Clin Infect Dis 2004;39:990–995.

Madeline Y. Sutton Pelvic inflammatory disease (PID) is an upper genital tract inflammation often caused by sexually transmitted infections (STIs) that ascend from the lower genital tract. PID is the most common gynecologic reason for hospitalization and emergency department visits in the United States, especially among adolescent women.1,2 Sexually active adolescent girls are at high risk for PID, mainly because of cervical ectopy (Figure 56-1) that increases the ability of certain organisms to bind to cervical tissue during sexual exposures and then ascend to the upper genital tract. Ascending infection is uncommon in prepubescent girls; the immature cervical tissue environment is a contributing factor.2

ETIOLOGIC AGENTS PID often has a polymicrobial etiology2–4 (Table 56-1). Isolation of both STI pathogens and endogenous pathogens from the upper reproductive tract of women with PID is common. The STI pathogens Neisseria gonorrhoeae and Chlamydia trachomatis are more commonly associated with acute infection. In a study of 84 patients with laparoscopically confirmed acute salpingitis, 77% had N. gonorrhoeae or C. trachomatis isolated.4 However, among women with acute PID, 20% to 50% of N. gonorrhoeae or C. trachomatis infections can be missed by cervical culture; PID, therefore, is often diagnosed without laboratory documentation of N. gonorrhoeae or C. trachomatis infection.2,3 Newer, more sensitive DNA amplification tests are being used in many clinical settings and are expected to have

greater sensitivity for detection of cervical disease, if present, in women with clinical PID. Anaerobic bacteria (including Bacteroides spp., Peptostreptococcus, Clostridium, and Actinomyces spp.) and facultative bacteria (e.g., Escherichia coli, Haemophilus influenzae, and Streptococcus spp.), which constitute the endogenous flora of the lower genital tract, more commonly are associated with chronic, recurrent, or complicated infections, such as tuboovarian abscesses. Anaerobic and facultative bacteria are identified in 25% to 60% of cases of PID.2,4 Bacterial vaginosis (BV) is associated with anaerobic and facultative bacteria when there is an imbalance of the vaginal microflora. BV has a prevalence of 27%, with 28% in sexually experienced and 18% in nonsexually experienced women, and is commonly associated with upper genital tract inflammation and PID.5,6 Some studies have shown an increased risk of chorioamnionitis, intra-amniotic fluid infection, postpartum endometritis, and postabortion PID in women with BV.7–10 However, while BV microorganisms commonly are found in patients with PID, a recent study showed no increased risk of developing incident PID among women with BV.11 Genital tract mycoplasmas, such as Mycoplasma hominis and Ureaplasma urealyticum, frequently are isolated from the endometrium of women with clinically diagnosed PID but their causative role is unclear.8 M. hominis also is prevalent on the cervical mucosa of healthy adolescents and women.2,5 Viral agents, such as herpes simplex virus and cytomegalovirus, and the protozoan Trichomonas vaginalis have been recovered from the upper genital tract but do not appear to cause PID.2

368

SECTION

G Urinary Tract Infections

EPIDEMIOLOGY Risk factors for PID are multiple and are often interrelated.12,13 Adolescent and young adult women 15 to 24 years of age, and living in the southern region of the country, are diagnosed most frequently with PID in hospital settings compared with older women (Table 56-2).1,14,15 Suggested reasons for increased risk among adolescents include increased number of STIs, failure to use barrier contraceptives, estrogen-dominated cervical mucus that easily is penetrated by pathogens, presence of columnar epithelium on the ectocervix for which both C. trachomatis and N. gonorrhoeae have a predilection, and lower levels of local and systemic immunity to pathogens due to lack of previous exposure.2,15,16 Adolescents seek healthcare later in the course of illness, mainly due to financial restraints and limited options for access to healthcare, which may increase the severity of the condition at presentation as well as subsequent sequelae.16 As many as one-third of affected women suffer repeat episodes of PID.13 When properly used, mechanical and chemical barriers, such as a condom, a diaphragm, and vaginal spermicide, appear to decrease the risk.3,8,13 Intrauterine devices (IUDs) may increase risk of PID during the first 20 days after insertion.13,17 Screening and treating for STIs prior to IUD insertion decrease the risk of postinsertion PID.17,18 Oral contraceptive agents have been associated with an apparent increased risk of C. trachomatis infection of the cervix,19 but a decreased risk of clinically overt PID, which may be due to changes in cervical mucus and decreased menstrual flow.20–22 PID that occurs closer to the time of menses is associated more closely with N. gonorrhoeae or C. trachomatis infection23; increased risk may be due to loss of the protective cervical mucus plug, the presence of blood that serves as a culture medium, and reflux of blood into the fallopian tubes.2 Uterine manipulative procedures, such as dilatation and curettage and hysteroscopy, can be associated with subsequent development of PID; antimicrobial prophylaxis decreases the occurrence.13 Vaginal douching has been shown to increase the risk for PID in some studies, but a direct relationship remains controversial.24–26

PATHOGENESIS AND PATHOLOGIC FINDINGS The ascent of gonococci and chlamydiae from the cervix through the endometrium to the fallopian tubes is the most common precipitant of TABLE 56-2. Average Annual Cases and Rates of Hospital Discharges and Ambulatory First Visits for Acute Pelvic Inflammatory Disease, 1995 to 2001, United States: Women 15 to 44 Years of Agea Hospital Dischargesb

Characteristic

a

BOX 56-1. Criteria for Clinical Diagnosis of Pelvic Inflammatory Disease (PID) • MINIMAL CRITERIA Cervical motion tenderness or uterine tenderness or adnexal tenderness • ADDITIONAL CRITERIA Oral temperature > 38.3°C (> 101°F) Abnormal cervical or vaginal mucopurulent discharge Presence of abundant numbers of white blood cells (WBCs) on saline microscopy of vaginal secretions Elevated erythrocyte sedimentation rate Elevated C-reactive protein Laboratory documentation of cervical infection with Neisseria gonorrhoeae or Chlamydia trachomatis • THE MOST SPECIFIC CRITERIA FOR DIAGNOSING PID INCLUDE THE FOLLOWING: Endometrial biopsy with histopathologic evidence of endometritis Transvaginal sonography or magnetic resonance imaging techniques showing thickened, fluid-filled tubes with or without free pelvic fluid or tuboovarian complex, or Doppler studies suggesting pelvic infection (e.g., tubal hyperemia) Laparoscopic abnormalities consistent with PID Adapted from Centers for Disease Control and Prevention. 2006 Guidelines for treatment of sexually transmitted infections. In press.

GYNECOLOGIC CAUSES Pregnancy (ectopic, and spontaneous abortion) Ovarian cyst (rupture, bleeding, and torsion) Ovarian or adnexal torsion Ovarian mass (hemorrhage and torsion) Dysmenorrhea Endometriosis Mittelschmerz (pain associated with ovulation) GASTROINTESTINAL CAUSES Appendicitis Inflammatory bowel disease Constipation Mesenteric adenitis Gastroenteritis Irritable bowel syndrome Intestinal obstruction Meckel diverticulum URINARY TRACT CAUSES Urinary tract infection Pyelonephritis Urinary calculus

Average Number of Cases

Average Rate per 1000 Women/Year (95% CI)

Average Rate Average per 1000 Number Women/Year of Visits (95% CI)

13,933 14,089 10,330 11,920 8651 6611

1.4 (1.2, 1.7) 1.6 (1.3, 1.8) 1.1 (0.9, 1.3) 1.1 (0.9, 1.3) 0.8 (0.6, 0.9) 0.6 (0.5, 0.7)

134,140 14.0 (9.4, 18.6) 166,586 18.4 (12.1, 24.8) 180,966 19.0 (12.0, 26.0) 78,564 7.6 (5.0, 10.1) 88,790 7.8 (4.9, 10.7) 55,280 5.0 (2.3, 7.7)

12,337 14,760 28,910 9527

1.1 (0.9, 1.3) 1.1 (0.9, 1.2) 1.3 (1.1, 1.5) 0.7 (0.6, 0.8)

97,696 8.5 (5.0, 11.9) 158,114 11.3 (7.8, 14.8) 299,222 13.8 (9.3, 18.3) 149,294 10.9 (6.2, 15.7)

Cases from private physicians, outpatient departments, and emergency departments. Data source: National Hospital Discharge Survey, 1995–2001. c Data sources: National Ambulatory Medical Care Survey, 1995–2001, National Hospital Ambulatory Medical Care Survey–Outpatient Department, 1995–2001, and National Hospital Ambulatory Medical Care Survey–Emergency Department, 1995–2001. b

Clinical diagnosis and differential diagnosis of PID remain difficult and subjective despite diagnostic guidelines (Boxes 56-1 and 56-2).

BOX 56-2. Differential Diagnosis of Pelvic Inflammatory Disease

REGION

Northeast Midwest South West

CLINICAL MANIFESTATIONS, DIFFERENTIAL DIAGNOSIS, AND CLINICAL APPROACH

Ambulatory First Visitsc

AGE GROUP (YEARS)

15–19 20–24 25–29 30–34 35–39 40–44

PID. Introduction of organisms into the upper genital tract can occur by direct migration, by transport via sperm or motile organisms such as T. vaginalis, by direct inoculation via instrumentation, or by carriage of refluxed menstrual blood.27,28 In the fallopian tube, gonococci cause inflammation by invading submucosal tissue and producing endotoxin. Chlamydiae cause a lymphocytic infiltrate with inflammation that can be severe.29 The presence of other bacteria may be explained by priming of tissue by the STI organisms (monomicrobial phase) for opportunistic invasion by facultative and anaerobic bacteria from the lower genital tract (polymicrobial phase).2 This “priming” could result from either alteration of the local mucosal defenses, or anatomic interference with mechanical host defenses that drain secretions from the upper genital tract mucosa.

Adapted from Braverman PK. Sexually transmitted diseases in adolescents. Med Clin North Am 2000;84:869–889.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Pelvic Inflammatory Disease

Symptoms and signs vary greatly among women, particularly among adolescents.2,14 Clinical diagnosis remains only 65% sensitive compared with the gold standard of laparoscopy.30 The frequencies of symptoms and signs reported in women with laparoscopy-proven PID are shown in Table 56-3.30 The most common symptoms and signs are lower abdominal pain, increased vaginal discharge, and adnexal tenderness. Because many of the symptoms associated with acute PID are nonspecific, it is important to consider other reproductive tract illnesses and diseases of both the urinary and the gastrointestinal tracts during an evaluation of a sexually active female with lower abdominal pain (Table 56-4 and Boxes 56-1 and 56-2). Pregnancy must also be excluded. No single or combined set of symptoms, signs, or laboratory test results is 100% sensitive and specific for establishing a diagnosis of PID.31 Physicians will often diagnose and treat based on clinical findings and the desire to prevent possible adverse sequelae. Patients

TABLE 56-3. Symptoms and Signs in 623 Patients with Laparoscopically Proven Pelvic Inflammatory Disease Symptom or Sign

56

369

who have minimal criteria for PID should be treated for PID if the evaluation is negative for other diagnoses, particularly appendicitis or ectopic pregnancy. Data from the PID Evaluation and Clinical Health (PEACH) Study and other trials have helped to shape recommendations for clinical diagnosis of PID (see Box 56-2).12,32 These minimal criteria detect approximately 60% of laparoscopically proven cases.2 Whereas additional routine criteria, such as fever, cervicitis, elevated erythrocyte sedimentation rate, and laboratory documentation of cervical infection with N. gonorrhoeae or C. trachomatis, increase the possibility of a correct diagnosis, a significant proportion of women who do not have PID also have some of these additional signs and symptoms.2 In patients with signs or symptoms beyond the minimal criteria, transvaginal sonography is a useful diagnostic tool for helping distinguish pyosalpingitis, hydrosalpinx, and tuboovarian abscess.33 Transvaginal ultrasonography has greater sensitivity than transabdominal examination. A study34 of magnetic resonance imaging (MRI) of the pelvis showed 95% sensitivity compared with laparoscopy. MRI was also shown to be more accurate than transvaginal ultrasonography.

LABORATORY FINDINGS AND DIAGNOSIS

Percentage (%)

Lower abdominal pain of recent onset Marked adnexal tenderness Increased vaginal discharge Irregular menstrual bleeding Fever > 38°C Urinary tract symptoms (cystitis, urethritis) Vomiting Proctitis symptoms

CHAPTER

95 90 55 35 35 20 10 5

Adapted from Jacobson L, Westrom L. Objectivized diagnosis of acute PID. Am J Obstet Gynecol 1969;105:1088–1098.

Finding leukocytes on examination of cervical or vaginal secretions is suggestive of an STI. A cervical culture for N. gonorrhoeae and C. trachomatis or other tests, such as nucleic acid amplification tests (ligase chain reaction, polymerase chain reaction), should be obtained in all patients with suspected PID.34,35 In many settings, the amplification tests are preferred because of increased sensitivity to detect more disease compared with culture. Laboratory confirmation of an STI is helpful because it provides diagnostic corroboration and guidance in therapy, indicates the most appropriate referral plan and

TABLE 56-4. Comparative Clinical Characteristics of Pelvic Inflammatory Disease, Ectopic Pregnancy, and Acute Appendicitis Pelvic Inflammatory Disease

Ectopic Pregnancy

Acute Appendicitis

Age Onset of symptoms

Any after puberty 75% within 7 days of menses

Any Any

Abdominal pain

Dull, crampy, localized to lower abdomen; right upper quadrant pain (perihepatitis) can be present

Vaginal discharge Menstrual pattern Fever

55% Metrorrhagia; missed menstruation not associated > 38°C in up to 35%

Risk increases with age At 4 weeks’ gestation or later; rupture at 6–10 weeks’ gestation Before rupture: localized or diffuse dull abdominal pain After rupture: severe, poorly localized pain, rectal pressure Not associated Intermittent bleeding or spotting; often missed or late menstruation > 38°C in up to 20%

Nausea and vomiting

Uncommon

Uncommon

Cervicitis Cervical motion tenderness

Purulent exudate in 80% > 95%

Not associated Can be present; not as pronounced as in salpingitis

Adnexal mass

5–50%

35–50%

Anemia

Not associated

Leukocyte count Pregnancy test

Normal or elevated Usually negative

Significant (hematocrit < 25%) after rupture Usually elevated Positive 1–2 weeks after conception

Ultrasonography

Common findings: cul-de-sac fluid, adnexal enlargement, complex adnexal mass Inflammation of fallopian tubes

HISTORY

Poorly localized periumbilical or epigastric pain with shift to right lower quadrant after 4–6 hours; increasing severity Not associated No change in usual pattern Usually < 38°C; > 38°C after perforation 50–60%

SIGNS

Not associated Usually none; can be present with perforation if appendix adjacent to adnexa Not associated

LABORATORY EVALUATION

Laparoscopy

Not associated

Intrauterine gestational sac absent

Mildly elevated Usually negative, but pregnant patients can have appendicitis as well Can detect inflamed appendix

Fallopian tube pregnancy with or without rupture

Normal fallopian tubes; inflamed appendix

Adapted from Paradise JE, Grant L. Pelvic inflammatory disease in adolescents. Pediatr Rev 1992;13:216–223.

370

SECTION

G Urinary Tract Infections

BOX 56-3. Principles of Management of Pelvic Inflammatory Disease • • • • • • • •

Exclude pregnancy and appendicitis Use minimal criteria to guide diagnosisa Err on the side of overdiagnosis Screen for and treat lower genital tract infections All women diagnosed with acute PID should be offered HIV testing Treat early and with broad-spectrum antibiotics Reassess patient 48 to 72 hours after initiating therapy Criteria for hospitalization: • Surgical emergency (e.g., appendicitis or ectopic pregnancy) cannot be excluded • Pregnancy • Lack of response clinically to oral antimicrobial therapy • Inability to follow or tolerate an outpatient oral regimen • Severe illness, nausea and vomiting, or high fever • Tuboovarian abscess • Identify, evaluate, and treat sex partners. In settings where only women are treated, male sex partners should be referred for appropriate treatment • Education Educate patient about STI prevention (see Chapter 53, Sexually Transmitted Disease Syndromes), including abstinence, encouraging the use of barrier methods of protection, and regular assessment for STIs HIV, human immunodeÀciency virus; PID, pelvic inflammatory disease; STI, sexually transmitted infection. a See Box 56-1, Criteria for Clinical Diagnosis of Pelvic Inflammatory Disease (PID).

treatment regimen for sex partners, and serves as a baseline for subsequent testing of the patient for microbiologic cure. However, laboratory conÀrmation is not necessary to justify initiation of therapy for PID. A serologic test for human immunodeÀciency virus (HIV) is also recommended.32,36 A pregnancy test should always be performed, to exclude ectopic pregnancy and because PID can occur concurrently with pregnancy.2

MANAGEMENT Prevention of adverse sequelae of infection is a primary goal of managing acute PID. The principles of management and criteria for hospitalization are outlined in Box 56-3. The antimicrobial regimens which have proven to be most effective in achieving clinical cure of PID were reviewed by an evidence-based panel and are summarized in Box 56-4.32,37,38 Clinical cure rates for the recommended therapeutic regimens are excellent, but mainly focus on eradication of microbes from the cervical mucosa. Reliable data are lacking on microbiologic cure rates of infection in the fallopian tubes.2 Parenteral therapy is initially recommended for patients with tuboovarian abscess.28 Abscesses 8 cm in diameter as well as bilateral abscesses may require drainage in addition to medical therapy.2 A woman treated as an outpatient should be reassessed 48 to 72 hours after initiation of therapy to evaluate the clinical response, reinforce the importance of completing the full course of treatment, and to help refer sex partners for treatment. Oral regimens should be as simple as possible to aid adherence. Antimicrobial regimens are designed to treat polymicrobial agents of PID, but they may not always be optimal against anaerobic bacteria. Because anaerobic bacteria can often exist in the genital tract during PID, recommended treatment guidelines now include the option to include an antimicrobial for anaerobic coverage.32,38,39 Antibiotic resistance of certain pathogens has emerged in some geographic areas;40,41 close follow-up of patients is essential to ensure eradication of pathogens.

COMPLICATIONS AND SEQUELAE At least one-fourth of women with PID experience one or more serious long-term sequelae, including ectopic pregnancy (6 to 10 times increased rate), infertility (approximately 10% after Àrst episode,

BOX 56-4. Suggested Treatment Regimens for Acute Pelvic Inflammatory Disease PARENTERAL REGIMENS Regimen A Cefotetan 2 g IV every 12 hours Or Cefoxitin 2 g IV every 6 hours plus Doxycycline 100 mg orally or IV every 12 hours Regimen B Clindamycin 900 mg IV every 8 hours plus Gentamicin loading dose IV or IM (2 mg/kg of body weight) followed by a maintenance dose (1.5 mg/kg) every 8 hours. Single daily dosing may be substituted ORAL REGIMENS Ceftriaxone 250 mg IM in a single dose plus Doxycycline 100 mg orally twice a day for 14 days with or without Metronidazole 500 mg orally twice a day for 14 days Or Ceftriaxone 2 g IM in a single dose plus Probenecid, 1 g orally administered concurrently in a single dose plus Doxycycle 100 mg orally twice a day for 14 days with or without Metronidazole 500 mg orally twice a day for 14 days Or Other parenteral third-generation cephalosporin (e.g., ceftizoxime or cefotaxime) plus Doxycycline 100 mg orally twice a day for 14 days with or without Metronidazole 500 mg orally twice a day for 14 days REGIMEN AFTER DISCHARGE Doxycycline 100 mg orally twice daily to complete at least 14 days of treatment plus Clindamycin 450 mg orally four times a day to complete at least 14 days of treatment. Clindamycin may be preferred if a tuboovarian abscess is present IM, intramuscular; IV, intravenous. From the Centers for Disease Control and Prevention. 2006 Guidelines for treatment of sexually transmitted infections. In press.

approximately 20% after second episode), chronic abdominal pain (3 times increased rate), and recurrent infection (2 to 3 times increased rate).3,22,42 Women infected with HIV may be at higher risk for tuboovarian abscesses.2 Mortality from PID is less than 1% and is usually secondary to rupture of a tuboovarian abscess or to ectopic pregnancy.2 Perihepatitis (inflammation of the liver capsule) occurs in 5% to 15% of patients as a complication of PID. Serum hepatic enzymes are usually normal because the hepatic parenchyma is not involved. Adhesions described as “violin strings” can occur between the liver and the anterior abdominal wall (Fitz–Hugh–Curtis syndrome; gonococcal periphepatitis in women), and patients can come to medical attention because of right upper quadrant pain that must be differentiated from other clinical entities.2 Recent annual estimates of the direct costs of care for acute PID and its sequelae are 2 billion dollars.43 When indirect costs of PID, such as productivity losses and losses from premature death, were considered, total annual cost estimates are as high as 10 billion dollars using year-2000 dollars.43,44 The personal suffering resulting from involuntary infertility due to PID is also great, with STI-related tubal infertility being responsible for an estimated 20% of cases of female infertility.45 A cost analysis revealed that most of the medical care costs related to PID occur during treatment for the acute episode.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Epididymitis, Orchitis, and Prostatitis

CHAPTER

57

Therefore, strategies for preventing PID, such as screening for chlamydial infection, may be immediately cost-effective.46

TABLE 57-1. Microbial Etiology and Predisposing Factors in Acute

PREVENTION

Predisposing Factors

PID is a common diagnosis among adolescents. Since STIs play a major role in PID, screening and early treatment of infected adolescents and their sex partners can help to minimize the risk of acquisition and continued transmission of STIs and subsequent adverse sequelae (see Chapter 53, Sexually Transmitted Disease Syndromes). The use of nucleic acid amplification testing, which can be performed on urine as well as urethral and cervical specimens, may be a cost-effective and a more patient-acceptable screening method for prevention of PID.47–50 Although abstinence from all types of sexual intercourse is the only absolute way to prevent acquisition and transmission of STIs that can lead to PID, barrier contraceptives should be encouraged for those adolescents who choose to be sexually active. Barrier contraceptives have been shown to decrease the risk of transmitting/acquiring HIV, N. gonorrhoeae, C. trachomatis and herpes simplex virus in men and women, and the risk of human papillomavirus and PID among sexually active adolescents and adults.51,52 A guide to STI resources on the internet is available.53

PREPUBERTAL CHILDREN

CHAPTER

57

Epididymitis, Orchitis, and Prostatitis Noni E. MacDonald

EPIDIDYMITIS Epididymitis is an inflammatory reaction or infection of the epididymis, the convoluted tubular structure attached to the upper posterior part of each testicle that collects and stores sperm.

Etiology, Epidemiology, and Pathogenesis For young adult males, epididymitis is common, accounting for more days lost from service in the military than any other disease.1 Although previously thought to be uncommon in prepubertal boys,2 the incidence in a prospective population-based study in 2000 was noted to be 1.2/1000 boys yearly, with peaks for hospital admissions in summer and winter.3 Acute epididymitis in boys may be more common than testicular torsion.4,5 Epididymitis can occur as an inflammatory postinfectious reaction to a number of bacterial and viral pathogens3 or as a complication of urethral infections caused by sexually transmitted pathogens,1,6 by genitourinary pathogens (especially if predisposing obstructive anatomic7,8 or neurologic genitourinary abnormalities9,10 or anorectal malformations11 exist), or, more rarely, through hematogenous spread to the epididymis from a primary focus of infection (such as with Haemophilus influenzae b, Salmonella spp., Streptococcus pneumoniae, or Mycocbaterium tuberculosis). The microbial etiology and predisposing factors (Table 57-1) in acute bacterial epididymitis in children and adolescents vary with age. Occasionally, epididymitis is not caused by inflammation or infection but is related to trauma, systemic diseases such as Henoch–Schönlein purpura or Kawasaki disease, or to medication such as amiodarone.2,3,11,12

Clinical Manifestations and Differential Diagnosis Epididymitis, whether inflammatory or infectious in etiology, typically has an acute onset of unilateral scrotal pain and swelling that

371

Epididymitis in Children and Adolescents

Underlying structural or neurologic abnormalities of the genitourinary tract

Etiology

Incidence

Enterobacteriaceae Pseudomonas aeruginosa

Uncommon

Hematogenous spread Haemophilus influenzae b Uncommon from primary focus Streptococcus pneumoniae Neisseria meningitidis Salmonella spp. Other ADOLESCENTS

Urethritis

Chlamydia trachomatis Neisseria gonorrhoeae Enterobacteriaceae Pseudomonas aeruginosa

Related to frequency of sexual activity

Underlying genitourinary tract pathology

Enterobacteriaceae Pseudomonas aeruginosa

Uncommon

Hematogenous spread from primary focus

Streptococcus pneumoniae Rare Neisseria meningitidis Mycobacterium tuberculosis Other

increases over 1 or 2 days. Common associated symptoms include dysuria and other lower urinary tract symptoms. Sexually transmitted epididymitis usually is accompanied by urethritis, which frequently is asymptomatic. Differentiation of epididymitis from testicular torsion, which requires immediate surgical intervention, is critical.13,14 Bacterial orchitis, an extension of epididymitis, is another important consideration in the differential diagnosis, especially in prepubertal children. Table 57-2 compares the clinical and laboratory differences among disorders. The presence of Prehn sign, relief of pain with testicular elevation, supports the diagnosis of epididymitis but is not definitive.15 Presence of urethral discharge is suggestive but not diagnostic of epididymitis. The cremasteric reflex is usually present in epididymitis but absent in testicular torsion. Ultrasonography, radionuclide scanning, and, occasionally, magnetic resonance imaging11,14–18 are helpful in differentiating among entities (see Table 57-2).

Management and Complications Organisms responsible for infectious epididymitis usually can be isolated from urine and urethral specimens. When a sexually transmitted infection (STI) is suspected, a Gram stain of urethral exudate or intraurethral swab specimen and a culture of intraurethral exudate or a nucleic acid amplification test (NAAT) (either on intraurethral swab or first-void urine) for Chlamydia trachomatis or Neisseria gonorrhoeae is indicated6 (see Chapter 55, Urethritis, Vulvovaginitis, and Cervicitis, Chapter 126, Neisseria gonorrhoeae, and Chapter 167, Chlamydia trachomatis). Empiric therapy, based on microscopic examination of urine sediment and Gram stain of a urethral swab specimen, is indicated before culture or NAAT results are available as appropriate early treatment of STI epididymitis cures the infection, improves the symptoms and signs, and prevents transmission of these STI microbes to others.6 For epididymitis most likely caused by gonococcal or chlamydial infection, the recommended treatment regimen is a single 250 mg dose of ceftriaxone intramuscularly plus a 10-day course of oral doxycycline (100 mg twice a day) orally while awaiting laboratory test results.6 In addition to antimicrobial therapy, bedrest, scrotal elevation, and administration of analgesics are recommended until fever and local inflammation are resolved.6 Failure of symptoms to improve

372

SECTION

G Urinary Tract Infections

TABLE 57-2. Comparison of Clinical Manifestations and Laboratory Findings in Acute Epididymitis, Acute Orchitis, and Testicular Torsion in Children and Adolescents Acute Orchitis Acute Epididymitis

Bacterial

Viral

Torsion Testes

Adolescence Postpuberty: urethritis (sexually transmitted infection) Prepuberty: structural abnormality

Rare in childhood Epididymitis

Adolescence None

Peripubertal None

Gradual Mild–severe Moderate

Gradual Moderate–severe Moderate–high

Gradual/acute Mild–severe Low–high

Acute Severe Uncommon

May occur with severe infection, uncommon Common

Common

Can occur

Uncommon

Common

Can occur

Uncommon

Vertical (normal) ++; may be unilateral Pain ΠUsually present

Vertical ++ Pain ΠUsually present

Vertical ++ No change in pain Usually present

Possibly horizontal + Pain unchanged or ΠUsually absent

+/–; epididymis tender ++ Common ++ Possible

++ ++ Common + Possible

+ (can be unilateral) ++ – – –

++ – – +/– (~30%) –

++ ++ Ø Ø

++ +/– Ø Ø

+/– Negative usually Ø Ø

+/– Usually negative Œ Œ

CLINICAL FEATURES

Most common age Predisposing factors

PAIN

Onset Severity Fever SYSTEMIC SYMPTOMS

Anorexia, vomiting, malaise Dysuria SCROTAL FINDINGS

Testicular lie Swelling Prehn sign Cremasteric reflex Tenderness Testes Spermatic cord Presence of hydrocele Skin inflammation Urethral discharge LABORATORY FINDINGS

Peripheral blood leukocytosis Pyuria (> 10 WBC/hpf) Doppler blood flow Testicular scan activity

+, mild to moderate; ++, moderate to severe; +/–, variably present; –, absent; Ø, increased; Œ, decreased.

within 3 days requires re-evaluation of the diagnosis and therapy, and may require hospitalization. In patients with infectious epididymitis in whom STI pathogens are not suspected, initial antimicrobial therapy is directed against uropathogens, such as Escherichia coli and Pseudomonas.5,6 These patients also need investigation to rule out underlying anatomic or neurologic abnormalities that may be causing urinary tract obstruction. When epididymitis is due to a systemic infection (i.e., hematogenous spread), antimicrobial therapy is directed against the suspected or isolated pathogen. In patients with inflammatory postinfectious reactive epididymitis, treatment includes analgesics and supportive care but antimicrobial therapy is not indicated.3 Complications of epididymitis include testicular abscess, chronic epididymitis, testicular infarction, and infertility. Drainage of scrotal abscesses or orchiectomy is seldom needed if the condition is diagnosed and treated promptly and close follow-up is performed.

ORCHITIS Orchitis, inflammation of the testis, rarely occurs in prepubertal patients. Epididymo-orchitis involves infection spreading from the epididymis to include the testicle. Both viral and bacterial pathogens can cause orchitis. Mumps is the most common viral pathogen (see Chapter 224, Mumps Virus).19,20 Less frequently, enterovirus (see Chapter 237, Enteroviruses)21,22 or, rarely, adenoviruses (see Chapter 210, Adenoviruses),23 varicella virus (see Chapter 205, VaricellaZoster Virus),24 or West Nile virus25(see Chapter 217, Togaviridae, and Chapter 218, Flaviviridae) are responsible.

When associated with mumps, orchitis usually follows parotitis by 4 to 8 days, but it has been reported to develop up to 6 weeks after parotid gland involvement and in the absence of parotitis.20,26 The onset of viral orchitis can be gradual or abrupt and usually is heralded by fever, chills, nausea, and lower abdominal pain (see Table 57-2). When the right testis is involved, appendicitis can be suspected erroneously if the scrotum has not been examined carefully. The affected testis is swollen and tender, and the adjacent skin is edematous and red. In orchitis (unlike epididymitis), a hydrocele is not common.27 Patients with mumps orchitis appear to have higher levels of C-reactive protein than do patients with mumps meningitis.28 The mumps virus can be detected in the semen for 14 days and mumps RNA can be detected for up to 40 days.29 Mumps orchitis is associated with a transient but signiÀcant reduction in sperm count and abnormal sperm morphology as well as development of antisperm antibodies, which may account for the long-term adverse effect on fertility in some patients.29 Bacterial epididymo-orchitis usually occurs as a consequence of contiguous spread from bacterial epididymitis and is due to organisms such as Escherichia coli, Pseudomonas aeruginosa, and Klebsiella or from hematogenous seeding from another source. Signs and symptoms are similar to those accompanying epididymitis (see Table 57-2). Toxoplasma gondii is a rare cause of orchitis, usually in the presence of widely disseminated disease or in a patient with immunodeÀciency.30 Management of viral orchitis is supportive and symptomatic, consisting of bedrest and analgesia. A small randomized controlled trial suggests that early treatment of mumps orchitis with interferona2B may decrease the risk of testicular atrophy.31 Bacterial epididymoorchitis is treated with antimicrobial therapy (see Epididymitis,

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Infectious Diseases of Child Abuse

above). Surgery may be necessary if testicular abscess or pyocele of the scrotum occurs. Infertility can result from viral or bacterial orchitis, but it is uncommon even after bilateral disease due to mumps.19,20

PROSTATITIS Prostatitis is inflammation of the prostate. Prostatitis encompasses a group of poorly defined clinical constellations with complaints referable to the lower genital tract and perineum.32,33 There are four categories of prostatitis syndromes according to the National Institutes of Health consensus classification.34 Prostatitis is an unusual condition in adolescents and young adults and is not reported in prepubertal boys.

Etiology, Pathogenesis, and Epidemiology In an adolescent or young adult, prostatitis is an acute bacterial infection.33,35 Organisms can reach the prostate by reflux of infected urine, hematogenous spread, or lymphatic spread. Although wellcontrolled studies are lacking, risk factors for acute prostatitis include trauma (bicycle riding, horseback riding) and dehydration.35 The most common pathogens associated with acute bacterial prostatitis are facultative gram-negative bacilli (uropathogens), such as Escherichia coli and Proteus spp. Enterococcus faecalis and Staphylococcus saprophyticus sometimes cause prostatitis, but other gram-positive constituents of urethral flora (e.g., S. epidermidis, micrococci, and streptococci) are seldom implicated. Acute prostatitis in men older than 35 years of age can be due to a STI. STIs do not cause prostatitis among adolescents, except when prostatitis accompanies Reiter syndrome precipitated by Chlamydia trachomatis. Reflux of sterile urine, which incites an inflammatory reaction, can be a factor contributing to noninfectious prostatitis. Rarely, mechanical obstruction such as that due to urethral valve dysfunction or prostatic stones causes noninfectious prostatitis among adolescents.36

CHAPTER

58

373

prostatitis is suspected, prostatic fluid is sampled for culture and examined for leukocytes. Examination of expressed prostatic secretions has been the definitive test for differentiating the prostatitis syndromes since Meares & Stamey37 introduced the prostatic localization four-cup urine collection test in 1968. This method, however, is cumbersome, time-consuming, and expensive and may not be reliable; it is not often used in practice.32,33 Preliminary study of a simpler screening test using urine collected before and after prostatic massage reports accuracy similar to the traditional four-cup test.32 The specimens are sent for quantitative bacterial culture and Gram stain of the urinary sediment for examination for leukocytes.

Management If bacterial prostatitis is suspected, antimicrobial therapy active against gram-negative organisms is recommended. Treatment can be difficult because of the limited penetration and activity of antibiotics in acidic prostatic fluid.33 Combination therapy with an aminoglycoside and a b-lactam agent is effective in controlling acute bacterial prostatitis in severely ill hospitalized adult patients. Therapy for less acute cases includes trimethoprim-sulfamethoxazole, a macrolide, doxycycline, or a fluoroquinolone. If mechanical obstruction is contributing to the problem, correction is required. There are generally no accepted principles for treating nonbacterial prostatitis and chronic pelvic pain syndrome, although anticholinergic agents can decrease urinary urgency and frequency, and antiinflammatory agents can reduce pain; little evidence supports the use of antimicrobial therapy.38 Suggested additional modes of therapy are muscle relaxants, sitz baths, biofeedback, analgesics, and psychotherapy.33

Complications Abscess of the prostate and stone formation are rare complications of acute bacterial prostatitis.39

Clinical Manifestations and Differential Diagnosis Acute bacterial prostatitis manifests as sudden onset of systemic and intense focal symptoms. In adolescents, the sudden onset of chills, fever, and malaise with voiding difficulties is much more likely to be due to urinary tract infection (see Chapter 51, Urinary Tract Infections), epididymitis, or urethritis (see Chapter 55, Urethritis, Vulvovaginitis, and Cervicitis) than to acute prostatitis. Rectal examination is usually painful; prostatic massage is not recommended, because it can precipitate bacteremia. The prostate is usually warm, tense, and swollen. Chronic bacterial prostatitis occurs in adults who usually have a history of recurrent urinary tract infection; symptoms include recurring episodes of pain or discomfort in the perineum, groin, lower back, or scrotum, and voiding dysfunction. Prostatic examination is not usually helpful because findings are variable. Nonbacterial (inflammatory) prostatitis and chronic pelvic pain syndrome also seen in adults manifest similarly to chronic bacterial prostatitis, except the patient has a history of preceding urinary tract infection. On examination, the prostate is either boggy with nodular areas caused by inflammation or is fibrous and difficult to massage. Typically, there is a complaint of abnormal urinary flow with pain or discomfort in the perineum, groin, testicles, penis, urethra, or a combination of these areas.

Laboratory Findings and Diagnosis Demonstration of inflammatory cells within the excretory ducts of the prostate is diagnostic of prostatitis.33 Diagnosis of acute bacterial prostatitis is usually based on recovery of bacteria from the urine in the patient in whom digital examination reveals a hot, swollen, tender prostate.37 In patients in whom chronic prostatitis or nonbacterial

CHAPTER

58

Infectious Diseases of Child Abuse David L. Ingram Infectious agents that are spread primarily by sexual contact in adults can cause infections in children after sexual abuse. Detection of an infection possibly due to sexual contact has important medical, legal, and social implications. Appropriate evaluation for infection and interpretation of positive test results for likelihood of the mode of acquisition are the subjects of this chapter. They are critical to the optimal care of children, their families, and the potential perpetrators of abuse. Treatment of infections is addressed in other chapters.

EVALUATION OF SEXUALLY ABUSED CHILDREN The proper evaluation of children for sexually transmitted diseases (STDs) has been concisely outlined in publications of the American Academy of Pediatrics (AAP) and the Centers for Disease Control and Prevention (CDC).1–3 The recommendations of these organizations, which differ in some respects, are summarized in Table 58-1. Both organizations recommend an individualized approach to specific STD testing in each child being evaluated for sexual abuse. STD evaluation might include some combination of specific tests for Neisseria gonorrhoeae and Chlamydia trachomatis and serologic

374

SECTION

G Urinary Tract Infections

TABLE 58-1. Guidelines for Evaluating Sexually Abused Children for Sexually Transmitted Disease (STD): Recommendations of the American Academy of Pediatrics (AAP) and the Centers for Disease Control and Prevention (CDC) AAP Guidelines1,2 (2003 and 2005)

CDC Guidelines3 (2002)

The decision is made on an individual basis; situations involving a high risk for STDs and a strong indication for testing include the following: • A suspected offender is known to have an STD or to be at high risk for STDs (e.g., multiple sex partners or a past history of STDs) • Patient or family requests testing • A postpubertal patient • Child with signs or symptoms of an STD or an infection that can be transmitted sexually • Prevalence of STDs in the community is high • STDs in siblings, other children, or adults in the household or immediate environment On an individual basis, examine for Chlamydia trachomatis, Neisseria gonorrhoeae, syphilis, human immunodeficiency virus (HIV), bacterial vaginosis, Trichomonas vaginalis, and hepatitis B

Same as the AAP guidelines

Obtain specimens at initial evaluation and in 2 weeks if recent exposure • For syphilis: serologic testing at time of abuse and 6, 12, and 24 weeks later and test abuser if possible • For HIV: test at time of abuse and 6, 12, and 24 weeks later

Obtain specimens at initial evaluation. If abuse took place < 2 weeks before, repeat evaluation at approximately 2 weeks after sexual contact; store initial serum specimen for testing if a later specimen is positive. At 12 weeks after sexual contact, test for Treponema pallidum, HIV, or hepatitis B on a case-by-case basis. If abuse occurred for an extended period of time, a single evaluation may be sufficient.

Rectal, throat, vaginal (prepubertal), and endocervical (postpubertal) specimens for culture; urethral specimen in boys

Pharyngeal, anal, vaginal specimens for standard culture, urethral specimen in boys if there is no discharge (meatal specimen if there is discharge).

Chlamydia trachomatis

Rectal, urethral (boys) or vulvovaginal (prepubertal), endocervical (postpubertal) specimens for culture (the gold standard) Nucleic acid tests may be more sensitive but data regarding use in prepubertal children are limited. Confirmatory testing with an alternative test may be important, especially in legal settings

Treponema pallidum

Dark-field microscopy examination of chancre fluid if present; serologic tests

Anal and vaginal specimens for standard culture in girls, and anal and urethral specimen in boys (if there is a discharge). Don’t use nonculture tests unless culture systems are unavailable, and if it can be confirmed by a second Food and Drug. Administration-approved nucleic acid test that targets a different sequence than the first test. Serologic tests

HIV

Serologic test of abuser (if possible); serologic test of child

Serologic tests

Hepatitis B virus

Serologic tests of abuser and child

Serologic tests

Herpes simplex virus

Culture of lesion distinguishing between types I and II

Culture all vesicular or ulcerative genital or perianal lesions

Bacterial vaginosis

Wet mount, pH and KOH vaginal discharge or gram stain Wet mount and culture of vaginal swab specimen

Human papillomavirus

Biopsy of lesion, or clinically diagnosed

—

Trichomonas vaginalis

Wet mount and culture of vaginal discharge

Wet mount and culture of vaginal swab specimen

Pediculosis capita

Examine for eggs, nymphs, and lice

1. Which children should be evaluated for which STDs?

2. What is the timing of evaluation?

3. Which specimens and tests are appropriate? Neisseria gonorrhoeae

tests for syphilis. If evaluation for STDs is performed, the CDC and AAP recommend testing on an individual basis for N. gonorrhoeae, C. trachomatis, Trichomonas vaginalis, bacterial vaginosis (BV), human immunodeficiency virus (HIV), herpes simplex virus (HSV), hepatitis B virus (HBV), human papillomavirus (HPV), and Treponema pallidum.

Some experts recommend that all children suspected of being sexually abused should have cultures for N. gonorrhoeae and C. trachomatis and a serologic test for syphilis performed. Both the AAP and CDC recommend that all specimens be collected at the initial evaluation. If sexual contact occurred more than

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Infectious Diseases of Child Abuse

CHAPTER

58

375

TABLE 58-2. Implications of Commonly Encountered Sexually Transmitted Diseases (STDs) for Diagnosis and Reporting of Suspected or Diagnosed Sexual Abuse of Infants and Prepubertal Children STD Confirmed

Sexual Abuse

Suggested Actiona

Gonorrhea

Diagnosticb,c

Report

Syphilis

b

Diagnostic

Report

Human immunodeficiency virus infection

Diagnosticd

Report

b,c

Chlamydia trachomatis infection

Diagnostic

Trichomonas vaginalis infection

Highly suspicious b

Report Report

Condylomata acuminata infection (anogenital warts)

Suspicious

Report

Herpes simplex (genital location)

Suspicious

Reporte

Bacterial vaginosis

Inconclusive

Medical follow-up

a

Reports should be made to the agency mandated in the community to receive reports of suspected sexual abuse. If not likely to be acquired perinatally. Only culture using standard confirmation methods should be used; DNA probes should not be used as a diagnostic method. d If not likely to be acquired perinatally or by transfusion. e Unless there is a clear history of autoinoculation. Adapted from American Academy of Pediatrics Committee on Child Abuse and Neglect. Guidelines for the evaluation of sexual abuse of children: subject review. Pediatrics 1999;103:186–191. Published correction appears in Pediatrics 1999;103:1049.26 b c

2 weeks previously, a single testing episode should be adequate. If the sexual contact occurred less than 2 weeks previously, evaluation should be repeated 2 weeks after sexual contact. This will allow sufficient time for incubation of possible pathogens or the appearance of a detectable antibody response. The AAP recommends that the selected child undergo serologic tests for syphilis and HIV at the initial visit and at follow-up 6, 12, and 24 weeks later, and that the perpetrator undergo a serologic test for HBV. The CDC recommends that serologic tests for these agents be performed selectively 12 weeks after sexual contact. Specimens from the throat, anus or rectum, and vagina are collected for isolation of N. gonorrhoeae. The AAP recommends that a specimen from the urethra of boys should also be cultured whether or not a discharge is present. The CDC recommends culturing a swab specimen from the urethra if no discharge is present or from the meatus if a discharge is evident. An endocervical culture is only obtained in postpubertal females. A culture is the only appropriate test; a rapid antigen detection test can yield false-positive (as well as false-negative) results, and its use in this setting is not appropriate. Specimens from the anus or rectum and vagina should be collected for isolation of C. trachomatis. The AAP recommends a culture of a specimen from the urethra of boys regardless of the presence or absence of discharge, whereas the CDC recommends a urethral culture only if discharge is evident. The CDC recommends only using nonculture tests if culture systems are unavailable, and if it can be confirmed by a second Food and Drug Administration-approved nucleic acid test that targets a different sequence from the first test.3The AAP warns that new tests such as nucleic amplification tests may be more sensitive in detecting C. trachomatis, but data regarding use in prepubertal children are limited. Because the prevalence of STDs in children is low, the positive predictive value of these tests is lower than in adults. Therefore, confirmatory testing, with an alternative test, is important, especially for legal use.2 One expert provides a thoughtful perspective as to why nonculture tests should be avoided.4 Differences in guidelines reflect assessments of likelihood of a positive test result with a specimen collected from a single site as well as cost issues. Physicians evaluating children for sexual abuse can develop their own protocols, choosing between guidelines when there are differences. Implications of commonly encountered STDs for diagnosis and reporting of suspected or diagnosed sexual abuse of infants and prepubertal children are outlined in Table 58-2.1,3

EPIDEMIOLOGY OF SPECIFIC AGENTS Identification of an infection in a child by an agent usually transmitted by sexual contact between adults requires explanation and investigation. Interpretation is enhanced by an understanding of pathogen-specific modes of transmission, periods of incubation, and possibility of reactivation after vertical transmission. Tables 58-3 and 58-4 summarize the current knowledge about common STD pathogens.5

Bacterial Vaginosis Limited data exist regarding BV and sexual abuse.6 In one study in which “definite disease” was equated with the presence in genital secretions of both clue cells and positive result of the whiff test (see Chapter 55, Urethritis, Vulvovaginitis, and Cervicitis) and “possible disease” with the presence of either finding, 4 of 31 (13%) girls aged 3.5 to 12 years evaluated for possible sexual abuse had definite BV and 4 others had possible BV. All had a history of penile penetration. Of 23 nonabused girls, only 1 had possible disease. It is not known whether BV can be acquired perinatally.

Chlamydia trachomatis All 15 verbal children with vaginal chlamydial infection in one study gave a history of sexual contact,7 as did 3 of 3 children aged 7 to 10 years with the disease in another study.8 The number of children with anal or pharyngeal infections due to C. trachomatis is too small to determine accurately what percentage of these chlamydial infections in children are due to sexual contact. Perinatal infection of the conjunctiva, rectum, vagina, and lungs is common.4 Untreated neonatal vaginal and rectal infections can persist for at least 12 months.9

Herpes Simplex Virus A small number of genital infections due to HSV in children have been reported in children evaluated for sexual abuse. In one report, 4 of 5 girls with genital infection caused by HSV type 2 (HSV-2) had a history or physical findings consistent with sexual contact.10 The other girl and 1 boy had oral and genital infection caused by HSV type 1 (HSV-1) and no history or physical findings suggestive of sexual contact. Self-inoculation of HSV-1 from oral to genital sites is possible.

376

SECTION

G Urinary Tract Infections

TABLE 58-3. Modes of Acquisition in Children of Agents of Sexually Transmitted Disease in Adults Agent/Illness

References

Acquisition by Sexual Contact

Acquisition by Nonsexual Contact

Perinatal Acquisition

Bacterial vaginosis

5

++

++

?

Chlamydia trachomatis (vaginal; > 3 years old)

4, 7, 8

++++

–

Yes

Herpes simplex, 1 and 2 (anogenital)

7, 10, 11, 27

++

++

Yes

Human immunodeficiency virus

3, 13

++++

+

Yes

Human papillomavirus (condylomata acuminata; > 3 years old)

3, 14

+++

+

Yes

Mycoplasma hominis

15, 16

++

++

Yes

Neisseria gonorrhoeae (pharyngeal, anogenital)

3, 18

++++

–

Yes

Treponema pallidum

3

++++

–

Yes

Trichomonas vaginalis

3, 7, 25

+++

+

Yes

Ureaplasma urealyticum

15, 16

++

++

Yes

++++, virtually always the mode; +++, the most likely mode; ++, possible mode; +, unlikely mode; –, negligible mode; ( ), site of culture, age of patient or name of infection.

TABLE 58-4. Epidemiology in Children of Agents of Sexually Transmitted Diseases in Adults

Agent/Illness

Incubation Period

Maximal Duration of Asymptomatic Infectiona

Bacterial vaginosis

7 days

?

Chlamydia trachomatis (vaginal)

≥ 7 days

≥ 3years

Herpes simplex (anogenital)

2–14 days

Life

Human immunodeficiency virus

Weeks– > 12 years

> 12years

Human papillomavirus (condyloma acuminata)

3 months–years

Life

Mycoplasma hominis

?

?

Neisseria gonorrhoeae

2–7 days

> 28 weeks

Treponema pallidum

10–90 days

Life

Trichomonas vaginalis

4–28 days

Up to 1 year

Ureaplasma urealyticum

10–20 days

?

a

Interpretation of positive test result in an asymptomatic child after possible sexual abuse must be made with considerations of this knowledge. ( ), site of culture or name of infection. Data collated from references 3, 6, 8, 27, 28.

Additional reports include a 7-year-old and a 4-year-old girl, both with vaginal HSV-2, 1 with and 1 without a history of sexual contact.7 In 6 cases of genital HSV infection in another series, 4 occurred in sexually abused girls (3 with HSV-1 infection, and 1 with HSV-2), and 2 cases in 2-year-old boys were not associated with sexual contact (both with HSV-1 infection).11

Human Immunodeficiency Virus Children have acquired HIV through sexual abuse. Reports of children with HIV infection have implicated sexual abuse as a source in 0.3% to 10.4% of cases.12,13 The likelihood that HIV infection contracted in childhood is due to sexual contact rather than other routes (such as breastfeeding, infected needles, or blood products) depends on whether there is a history of these other exposures. Vertical

transmission during birth is the usual route of infection, even in children as old as 12 years at the time of clinical presentation.

Human Papillomavirus HPV diagnosed on the basis of condylomata acuminata can be due to sexual transmission in children. However, controversy exists about the percentage caused by sexual contact.12 Of 19 verbal children with condyloma acuminata in one study, 12 gave a history of sexual contact.7 Of the 7 children who did not give a history of sexual contact, only 2 had mothers with known HPV infection at the time of delivery. Gutman and colleagues, in reviewing the literature on HPV infection in children, concluded that most children with condylomata acuminata that first appears after the first few months of life acquired their infection through sexual contact.14 Perinatal exposure and infection are relatively common; the longest subclinical interval before presentation of condylomata acuminata is not known but may be 2 years or more. For children 3 years or younger, vertical and horizontal transmissions should be considered.4,12 Condyloma acuminatum can disappear and recur after years of latency (Table 58-5). The prolonged incubation period and the ability to recur after lengthy asymptomatic periods make it difficult to determine the timing of HPV acquisition. Recurrent respiratory papillomatosis in young children is diagnosed in 75% by age 5 years and is thought to be acquired during delivery (see Chapter 211, Human Papillomaviruses).

Mycoplasma hominis Studies are conflicting as to whether infections due to large-colony mycoplasmas (LCMs), which include Mycoplasma hominis, are more commonly found in sexually abused children. In a study of 83 girls aged 2.5 to 14.5 years, LCMs were isolated from vaginal cultures in 34% of 47 girls with sexual contact, versus 17% of 36 controls; anorectal cultures were positive in 23% of girls with sexual contact, versus 8% of controls. The differences between colonization rates at both sites in those with and without sexual contact were not statistically significant (P > 0.05 < 0.1).15 Reports of the results of vaginal and anorectal cultures for LCM in 416 girls evaluated for sexual abuse revealed no significant association of LCM with sexual contact.16 Overall, anorectal carriage rates were associated with age, being 0 to 2% at 1 to 3 years and rising to 17.5% at 7 to 9 years and to 10% at 10 to 12 years. The vaginal colonization rate was 2% to 3% at 1 to 3 years and 11% at 10 to 12 years. Female and (rarely) male neonates can contract anogenital infection from the mother’s infected birth canal, but these infections rarely persist.17

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Approach to the Diagnosis and Management of Gastrointestinal Tract Infections

Neisseria gonorrhoeae Branch & Paxton found that 160 of 161 children aged 1 to 14 years who were infected with N. gonorrhoeae had a history of sexual contact.18 Occasional cases of isolated gonococcal conjunctivitis, especially in children younger than 2 years, can occur in the absence

CHAPTER

59

377

of sexual contact.19 Failure to elicit a history of sexual contact in children does not preclude abuse, as evidenced by cases of infection due to N. gonorrhoeae in two 8-year-old girls who acknowledged a history of abuse only 12 years later, for fear of retaliation by the perpetrator.20

Treponema pallidum TABLE 58-5. Recurrence of Symptomatic Infection in Children After Latent or Asymptomatic Infection with a Sexually Transmitted Diseasea

Agent/Illness

Recurrence of Symptoms After Perinatally Acquired Infection

Recurrence of Symptoms After Postnatally Acquired Infection

Bacterial vaginosis

?

?

Chlamydia trachomatis (vaginal)

?

?

Herpes simplex (anogenital)

Yes

Yes

Human immunodeficiency virus

Yes

Yes

Human papillomavirus (condyloma acuminata)

Yes

Yes

Mycoplasma hominis

?

?

?

Yes

Treponema pallidum

?

?

Trichomonas vaginalis

?

?

Ureaplasma urealyticum28

?

Yes

Neisseria gonorrhoeae

27,29,30

Virtually all children who acquire syphilis after the neonatal period contract the infection through sexual contact.4 Case reports of transmission to children by a kiss from an infected mother, through breastfeeding in which there was contact with a breast chancre, and from blowing a trumpet just previously used by an infected person should be regarded with skepticism.21–23

Trichomonas vaginalis A small number of cases of trichomoniasis have been reported in sexually abused girls.7,24 Too few cases have been thoroughly evaluated and reported to determine the likelihood that infection is due to sexual contact. Carriage may persist for up to a year.25

Ureaplasma urealyticum In a study of 47 abused girls and 36 controls, Ureaplasma urealyticum was isolated from vaginal secretions of 30% and 8%, respectively, and from anorectal cultures in 19% and 3%, respectively (differences for both vaginal and anorectal recovery rates significantly different (P < 0.025).15 In another study, no greater prevalence of vaginal or rectal Ureaplasma infection was noted in 416 children evaluated for sexual abuse when the findings were controlled for age and race.16 U. urealyticum was found in vaginal or rectal specimens in up to 3% of 1- to 3-year-old girls and 12% to 15% of 7- to 9-year-old girls, regardless of history of sexual contact, in this study.

a

Superscript numbers indicate references.

SECTION

H

Gastrointestinal Tract Infections and Intoxications CHAPTER

59

Approach to the Diagnosis and Management of Gastrointestinal Tract Infections Larry K. Pickering Most infections of the gastrointestinal tract manifest as diarrhea, which is a clinical syndrome of diverse etiology associated with frequent loose or watery stools often accompanied by emesis, fever, and abdominal bloating or pain and occasionally by extraintestinal manifestations. Infectious diarrhea can be caused by bacterial, viral,

and parasitic enteropathogens (Table 59-1), most of which are associated with specific epidemiologic factors or clinical manifestations.1 The diagnosis of a specific cause of an episode of diarrhea is sometimes difficult to establish because of the wide array and complexity of potential etiologic agents. Therapy is directed at fluid and electrolyte replacement and maintenance, and dietary considerations in all cases, and at specific antimicrobial therapy in some.2 Commercial vaccines for prevention of diarrhea are available in the United States only for Salmonella typhi and rotavirus.3–5

EPIDEMIOLOGY Enteropathogens are acquired through the fecal–oral route from person-to-person transfer or via contaminated food or water. Enteric pathogens acquired via person-to-person transmission are generally organisms that require a low inoculum dose. People with certain host defects may be more susceptible to infection with various enteric pathogens and to suffer greater mortality or severe morbidity. Table 59-2 shows the inocula of various enteric pathogens necessary to

378

SECTION

H Gastrointestinal Tract Infections and Intoxications

TABLE 59-1. Causative Agents of Gastroenteritis Bacteria

Parasites

Viruses

Aeromonas species Bacillus cereus Campylobacter jejuni Clostridium difficile Clostridium perfringens Escherichia coli Listeria monocytogenes Plesiomonas shigelloides Salmonella species Shigella species Staphylococcus aureus Vibrio cholerae Vibrio parahaemolyticus Vibrio vulnificus Yersinia enterocolitica

Cryptosporidium parvum Cyclospora cayetanensis Entamoeba histolytica Giardia lamblia Isospora belli Microsporidia (including Enterocytozoon bieneusi and Encephalitozoon intestinalis)

Astrovirus Enteric adenovirus Noroviruses Rotavirus

Childcare centers

Foodborne or waterborne

Antimicrobial agents

Travel

Immunocompromised

Hospitalization Diarrhea

Clinical manifestations • Acute/chronic • Inflammatory/noninflammatory • Extraintestinal

TABLE 59-2. Inoculum Required to Cause Diarrhea in Adult Volunteers and Association of Agents with Outbreaks of Diarrhea Due to Person-to-Person Transmission in Childcare Centers Organism

Inoculum

Outbreaks Reported in Childcare Centers

Shigella Escherichia coli O157:H7 Cryptosporidium Rotavirus Enteric adenovirus Astrovirus Giardia lamblia Entamoeba histolytica Campylobacter jejuni Salmonella Vibrio cholerae Escherichia coli

101–2 Unknown 132 oocysts Unknown Unknown Unknown 101–2 cysts 101–2 cysts 102–6 106 108 108

Yes Yes Yes Yes Yes Yes Yes No Rare No No Yesa

a

Limited to enteropathogenic and O157:H7 strains.

cause diarrhea in adult volunteers. This table also lists outbreaks of diarrhea due to person-to-person transmission reported in childcare centers (see Chapter 3, Infections Associated with Group Childcare). Because volunteer studies are not performed in children, outbreaks of diarrhea in childcare centers must be used to determine which organisms are associated with low-inocula disease in this population.6 In the United States every year, there are 20 to 35 million episodes of diarrhea, resulting in 2.1 to 3.7 million physician visits, 220 000 hospitalizations accounting for almost 1 million hospital days, and 300 to 400 deaths,3,7,8 with the majority of episodes in children less than 5 years of age due to rotavirus.9 Approximately 9% of all hospitalizations of children younger than 5 years of age are due to diarrhea.7 Children at higher risk of death due to diarrhea include young infants who were born prematurely, children of teenage mothers who have had little or no prenatal care, and children with underlying immune deficiencies. In the United States, the incidence of diarrhea in children younger than 3 years of age is estimated to be 1 to 3 episodes per child per year, with rates being higher in children attending childcare.6,10 Worldwide, diarrheal diseases are a leading cause of pediatric morbidity and mortality, with 1.5 billion episodes and 1.5 to 2.5 million deaths estimated to occur each year among children less than 5 years of age.11–13 Repeated early-childhood enteric infections result in long-term disability.14

GENERAL CONSIDERATIONS Evidence-based recommendations, reviews, and meta-analyses that deal with people with diarrhea are available and include information

Diagnosis • Epidemiologic factors • Clinical evaluation • Laboratory studies

Therapy • Fluid and electrolytes • Diet • Antidiarrheal compounds • Antimicrobial therapy Figure 59-1. Steps in assessment of likely causative agents and management of diarrhea in children.

about the following: (1) administration of fluid and electrolyte solutions; (2) clinical and epidemiologic evaluation; (3) ordering of selective studies on stool specimens; (4) avoidance of antimotility agents in people with bloody diarrhea; (5) institution of selective antimicrobial therapy; and (6) prevention of disease by immunization.1–5,15–18 Considering diarrhea by category assists in evaluation. The major categories of diarrhea are those associated with foodborne or waterborne outbreaks, with travel, with use of antimicrobial agents, with immunosuppression, with childcare or extended-care facilities, and with hospitalization (Figure 59-1). There are overlaps among the various categories with regard to organisms associated with the cause of illness, but in each category, specific agents are responsible for most episodes.

Foodborne Diarrhea Foodborne diseases, including foodborne infections and intoxications, are acquired by consumption of contaminated food. A diverse group of chemical contaminants (heavy metals and organic compounds), bacterial toxins (enterotoxins of Bacillus cereus, Staphylococcus aureus, Clostridium perfringens, Escherichia coli, Vibrio cholerae), products accumulated in the food chain of fish and shellfish (scombroid, ciguatera, neurotoxic and paralytic shellfish poisoning, tetrodotoxin, and domoic acid), and disease due to other bacteria (Salmonella, Shigella, E. coli, Brucella, Yersinia, Campylobacter, Vibrio, and Listeria), viruses (norovirus, rotavirus, and hepatitis A virus), and parasites (Giardia, Cyclospora, Cryptosporidium, Taenia, Toxoplasma, and Trichinella) constitute the major causes of foodborne and waterborne diseases (see Chapter 63, Foodborne and Waterborne Disease).15,16 Viruses are considered the most common cause of foodborne illness.15 An outbreak of foodborne or waterborne disease is defined as an incident in which two or more people experience a similar illness,

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Approach to the Diagnosis and Management of Gastrointestinal Tract Infections

usually involving the gastrointestinal tract, after ingesting a common food or water intended for eating or drinking, respectively.15,16 A single case of botulism or chemical poisoning constitutes an outbreak if laboratory studies indicate contamination of the implicated food or water by Clostridium botulinum or the chemical. The incubation period, duration of the resultant illness, clinical manifestations, and population involved in the outbreak are helpful in establishing a diagnosis, but prompt and thorough laboratory evaluation of involved people and implicated food or water is critical for definitive diagnosis. Individual cases are difficult to identify unless a distinct clinical syndrome exists, such as occurs with botulism. Most people with food poisoning respond to supportive care, because most of these illnesses are self-limited. Exceptions include botulism (which causes constipation rather than diarrhea), paralytic shellfish poisoning, long-acting mushroom poisoning, and enterohemorrhagic E. coli infections, all of which can produce significant morbidity and mortality in previously healthy people. Prevention and control of these diseases, regardless of the specific cause, are based on avoidance of food contamination, destruction or denaturation of the contaminants, and prevention of further spread or multiplication of contaminants. Suspected cases of a foodborne illness should be reported to public health officials, because any patient with a foodborne illness may represent the sentinel case of a widespread outbreak. State reporting requirements and information can be obtained online at http://www.cste.org/nndss/reportingrequirements.htm.

Travelers’ Diarrhea Diarrhea is the most common illness encountered among international travelers to developing countries, affecting approximately 40% of people in the first 2 weeks of travel.19–21 Most studies of diarrhea in travelers have been performed in adults, and only limited data are available for the pediatric population.22 Diverse enteric pathogens are implicated in travelers’ diarrhea; bacteria are implicated most frequently, especially enterotoxigenic E. coli, Campylobacter jejuni, Salmonella species, and Shigella species; viruses and parasites account for a smaller percentage of cases. In 10% to 50% of episodes no pathogen is isolated.21 Risk factors for travelers’ diarrhea are young age, season, eating in restaurants, and short duration of stay in the developing country in which the travel is occurring.23–25 Etiologic diagnosis is complex because of the diversity of potential pathogens, requirements of specialized testing for some agents (bacterial and fish toxins), and a general lack of available laboratory facilities in areas where organisms are acquired. Clinical illness is variable, reflecting the diversity of causative agents.19–21,26 Because most diarrheal episodes are bacterial in origin, travelers to developing countries should be given advice on how to diagnose and treat a diarrheal illness empirically with an antibiotic specific for the area of travel, if they cannot obtain medical care. Use of antimicrobial agents as a preventive measure is not recommended routinely.27 Travelers should be informed about: (1) the ubiquitous exposure to pathogens during travel; (2) the need to follow appropriate rules of hygiene, including avoidance of potentially contaminated foods and water, proper preparation of foods, and use of hand hygiene; and (3) the importance of fluid replacement if diarrhea occurs. Children who travel may have a higher risk of disease and a greater severity of diarrhea if disease occurs (see Chapter 9, Protection of Travelers).

59

379

76 children, C. difficile toxins were identified in posttherapy stools; none had positive results before therapy. A previously uncommon strain of C. difficile with variations in toxin genes has become more resistant to fluoroquinolones and has emerged as a cause of outbreaks of C. difficile-associated diarrhea.29,30 The problem of defining the association of C. difficile and diarrhea in infants is illustrated by studies showing that 25% to 65% of children younger than 1 year of age without diarrhea are colonized with toxinproducing C. difficile31,32 and that no association exists in this age group between the presence of toxigenic C. difficile and acute community-acquired or nosocomial diarrhea.33,34 Colonization rates of toxigenic C. difficile have been shown to be 0 to 10% in older children. Because of the lower rate of carriage of toxigenic C. difficile in older children and adults, diagnosis of antimicrobial-associated diarrhea and colitis in this population is more easily established.

Diarrhea in Immunosuppressed Children Diarrhea in children who have immunodeficiencies or are immunosuppressed is caused by enteropathogens that have been shown to produce disease in all hosts as well as by enteropathogens unique to this population.35–38 Gastrointestinal tract disease in children with acquired immunodeficiency syndrome (AIDS) can be due to enteric infection, malignancy, chronic disease, and invasion by the human immunodeficiency virus (HIV). Clinical manifestations and ease of eradication depend on location of involvement in the gastrointestinal tract, immune status of the host, severity of injury or disease that the intestinal tract has sustained, and availability and effectiveness of therapy. Organisms and diseases of the gastrointestinal tract that fulfill the surveillance case definition of the Centers for Disease Control and

TABLE 59-3. Organisms Associated with Gastrointestinal Tract Infections in Immunocompromised Hosts Including People with AIDS Site

Organisms

Esophagus

Candida albicansa Herpes simplex virusa CMVa

Hepatobiliary tract

CMVa Cryptosporidiuma Hepatotropic viruses MACa

Stomach

CMVa MACa

Small intestine

CMVa Campylobacter species Cryptosporidium speciesa Enterocytozoon bieneusi Giardia lamblia Isospora bellia MACa Salmonella species Strongyloides stercoralis

Large intestine

Campylobacter jejuni Clostridium difficile CMVa Entamoeba histolytica Herpes simplex virus Salmonella speciesa Shigella species

Blood

Salmonella speciesa

Antimicrobial-Associated Diarrhea Diarrhea commonly occurs during therapy with antimicrobial agents. Antimicrobial-associated diarrhea can result from changes in smallbowel peristalsis or from alterations in the intestinal microflora, including overgrowth by Clostridium difficile. In a prospectively conducted study, 29% of 76 children who received amoxicillinclavulanate for therapy of otitis media experienced diarrhea, which led to discontinuation of therapy in 5 children (23%).28 In 10 (13%) of the

CHAPTER

AIDS, acquired immunodeficiency syndrome; CMV, cytomegalovirus; MAC, Mycobacterium avium complex. a Indicator diseases of the gastrointestinal tract included in the Centers for Disease Control and Prevention surveillance case definition of AIDS.

380

SECTION

H Gastrointestinal Tract Infections and Intoxications

Prevention (CDC) for AIDS are listed in Table 59-3.36 Guidelines for prevention of opportunistic infections, including infections involving the gastrointestinal tract, in people with HIV infection and among recipients of hematopoietic stem cell transplantation have been published.37,38

Diarrhea in Childcare Centers The greater use of out-of-home childcare has had a significant impact on the epidemiology of diarrheal disease in the United States (see Chapter 3, Infections Associated with Group Childcare).6,10 After respiratory tract illness, diarrhea is the most common disease among children in childcare facilities, occurring at a rate of approximately 3 cases per year among children younger than 3 years of age.6 Episodes of acute infectious diarrhea in childcare settings have been associated with several bacterial, viral, and parasitic enteropathogens. Rotavirus, enteric adenoviruses, astroviruses, caliciviruses, Shigella, Cryptosporidium, E. coli O157:H7, and Giardia lamblia are the most frequently identified organisms associated with outbreaks of diarrhea in childcare centers. The major route of transmission is person-toperson (see Table 59-2). Risk factors include gathering of children, specifically children younger than 3 years of age, new enrollment in a center, centers that house large numbers of children, centers without an exclusion policy for diarrheal disease, and inadequate hand hygiene.6,10 The single most important procedure for minimizing fecal or oral transmission of enteric pathogens is frequent hand hygiene combined with education, training, and monitoring of staff regarding infection control procedures.39

Diarrhea in Hospitals Diarrhea can develop in children as a result of infections acquired before hospital admission or during hospitalization. Each child infected with an enteropathogen can then become a potential source of spread within the hospital population. The CDC considers an episode of diarrhea to be nosocomial if the onset of disease occurs during hospitalization or shortly after discharge and if the infection is not present or is not incubating at the time of the patient’s admission to the hospital.40,41 Reports by the CDC as part of the National Nosocomial Infections Surveillance (NNIS) system from 1985 to 1991 demonstrated that nosocomial diarrhea occurred in general pediatric care units at a rate of 11 cases per 10 000 discharges and in newborn nurseries at a rate of 3 cases per 10 000 discharges. The rate is higher in high-risk nurseries, averaging 20 per 10 000 discharges. The NNIS system also reported that gastrointestinal tract infections cause 8% of all nosocomial infections in high-risk nurseries.42 Studies of pediatric populations have identified viral agents as the most common nosocomial enteropathogens, with rotavirus identified most frequently.43 Guidelines and recommendations for prevention of healthcareassociated infections can be found at www.cdc.gov/ncidod/dhqp/ guidelines.html. Several factors are important in transmission of organisms, including: (1) patient-to-patient transmission via hands of hospital personnel, generally after contact with a child who has diarrhea; (2) asymptomatic carriers; (3) contaminated food, medications, or medical instruments; and (4) hospital crowding. Host risk factors are immunocompromised states, prolonged hospitalization, and young age. Prevention of nosocomial gastrointestinal tract infection is best accomplished by surveillance of methods to improve hospital infection control procedures. Contact isolation is recommended for children with diarrhea to prevent transmission among children and between staff and children.40,41,44

PATHOGENESIS The gastrointestinal tract is barraged constantly by foreign material, including bacteria, viruses, parasites, and toxins. Numerous host

factors counteracting the intrusion of invading enteropathogens are gastric acidity, intestinal motility, enteric microflora, glycoconjugates, and specific immune components (cells and humoral compounds). Host species, genotype, age, personal hygiene, intestinal receptors, past exposure history of the host, and food intake (including human milk) influence these protective factors and are therefore major determinants of colonization and disease due to various enteropathogens. The effect of microbial factors is influenced by the size of the inoculum and specific virulence traits of the enteropathogen. One or more virulence traits possessed by enteric bacteria determine the pathogenic mechanisms associated with diarrhea. The range of these traits is illustrated by the various types of E. coli that produce disease in the gastrointestinal tract (see Chapter 137, Escherichia coli). Depending on the presence of transmissible plasmids or phages, which encode for virulence traits, E. coli can produce secretory heat-stable or heat-labile enterotoxins and cytotoxins, can be invasive, can manifest one of several types of adherence, or can be enteroaggregative.45,46 In addition, some bacteria, such as Clostridium botulinum, Staphylococcus aureus, and Bacillus cereus, produce neurotoxins while others have variations in their toxin genes and become more virulent.15,29,30 All bacteria that are associated with disease in the gastrointestinal tract generally possess one or more of these virulence traits.45 Enteric viruses can cause diarrhea through selective destruction of absorptive cells (villus tip cells) in the mucosa, leaving secretory cells (crypt cells) intact (see Chapter 60, Viral Gastroenteritis). Infection with rotavirus, noroviruses, astroviruses, and enteric adenoviruses alters absorptive fluid balance and also reduces the brush-border digestive enzymes. This imbalance and disruption of the specialized absorptive surface also may be involved in other small-bowel infections associated with villus tip flattening or microvillus destruction, as occurs with E. coli pathotypes and many parasitic infections.45–48

CLINICAL MANIFESTATIONS Clinical manifestations of gastrointestinal tract infections are varied but generally consist of primary gastrointestinal tract and systemic manifestations as well as intestinal or extraintestinal complications. Clinical manifestations localized to the gastrointestinal tract can be categorized for diagnostic and therapeutic considerations into several groups: ● ● ● ● ●

Watery diarrhea due to enteric viruses, enterotoxin-producing bacteria, and protozoa that infect the small intestine Purging, watery diarrhea that occurs in some infections due to Vibrio cholerae and enterotoxigenic E. coli Dysentery with scant stools that contain blood and mucus associated with organisms that invade the large intestine Persistent diarrhea that lasts for 14 days or longer Vomiting with minimal to no diarrhea, typical of chemical contaminants and toxins associated with foodborne and waterborne outbreaks and with viral enteropathogens

Extraintestinal manifestations associated with gastrointestinal tract infections result from infection and immune-mediated mechanisms. Extraintestinal infections related to bacterial enteric pathogens can include local spread that causes vulvovaginitis and urinary tract infection. Remote spread can result in endocarditis, arteritis, osteomyelitis, arthritis, meningitis, pneumonia, hepatitis, peritonitis, chorioamnionitis, soft-tissue infection, keratoconjunctivitis, and septic thrombophlebitis.49 The immune-mediated extraintestinal manifestations of enteric pathogens are shown in Table 59-4. Signs and symptoms usually occur after diarrhea has resolved.

DIAGNOSIS Laboratory studies can be performed to identify the cause of diarrhea and to guide therapy but often are not required because most episodes of diarrhea are self-limited. The main objectives in the approach to a child with acute diarrhea are to: (1) assess the degree of dehydration and provide fluid and electrolyte replacement; (2) prevent spread of

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Approach to the Diagnosis and Management of Gastrointestinal Tract Infections

TABLE 59-4. Immune-Mediated Extraintestinal Manifestations of Enteric Pathogens Manifestation

Related Enteric Pathogen(s)

Erythema nodosum Glomerulonephritis Guillain–Barré syndrome Hemolytic anemia Hemolytic–uremic syndrome Immunoglobulin A nephropathy Reactive arthritis

Yersinia, Campylobacter, Salmonella Shigella, Campylobacter, Yersinia Campylobacter Campylobacter, Yersinia Enterohemorrhagic Escherichia coli Campylobacter Salmonella, Shigella, Yersinia, Campylobacter, Cryptosporidium Shigella, Salmonella, Campylobacter, Yersinia

Reiter syndrome

the enteropathogen; and (3) in select episodes, determine the etiologic agent and provide specific therapy if indicated. In some children, additional laboratory tests on stool may be indicated. Additional evaluation includes observation of stool, microscopy, rapid diagnostic tests, culture, and use of specialized laboratory tests.

History To assist in determination of the etiologic agent, the following history should be obtained: childcare center attendance; recent travel to a diarrhea-endemic area; recent hospitalization; visiting a farm or petting zoo or having contact with reptiles or pets; recent or current use of antimicrobial agents; exposure to contacts with similar symptoms; intake of seafood, unwashed vegetables, unpasteurized milk, contaminated water, or uncooked meats; and status of the immune system (see Figure 59-1). Information also should be sought regarding how and when the illness began, duration and severity of diarrhea, stool frequency and consistency, presence of mucus and blood, and other associated symptoms, such as fever, vomiting, abdominal pain, headache, and seizures. The occurrence of fever suggests an inflammatory process, usually involving the large intestine (see Chapter 61, Inflammatory Enteritis, and Chapter 62, Necrotizing Enterocolitis), but fever also can result from dehydration and systemic spread of an organism. Nausea and emesis are nonspecific symptoms, although vomiting suggests upper intestinal tract infection, as occurs with enteric viruses, enterotoxinproducing bacteria, Giardia, and other protozoa. As a general rule, in people with inflammatory diarrhea, fever is common, abdominal pain is more severe, and tenesmus can occur in the lower abdomen and rectum. In noninflammatory diarrhea, emesis is common; fever is usually absent or low-grade; pain is crampy, periumbilical, and not severe; and diarrhea is watery, indicating upper intestinal tract involvement (see Chapter 60, Viral Gastroenteritis). Because immunocompromised patients require special consideration, information about an underlying primary or secondary immunodeficiency or chronic disease is important. Persistent diarrhea, defined as diarrhea lasting more than 14 days, occurs as a result of multiple consecutive infections, the postgastroenteritis syndrome, or persistent infection.

Examination of Stool Stool specimens should be examined for mucus, blood, and leukocytes, the presence of which indicates colitis. Fecal leukocytes are produced in response to bacteria that diffusely invade the colonic mucosa. A positive fecal leukocyte examination or stool lactoferrin assay indicates presence of an invasive or cytotoxin-producing organism such as Shigella, Salmonella, Campylobacter jejuni, invasive E. coli, enterohemorrhagic E. coli, Clostridium difficile, Yersinia enterocolitica, Vibrio parahaemolyticus, and, possibly, an Aeromonas species or Plesiomonas shigelloides.50 However, not all patients with colitis have leukocytes in stool or a positive lactoferrin assay result; ingestion of human milk can interfere with the lactoferrin assay.

CHAPTER

59

381

Stool cultures should not be performed in all children with acute diarrhea, especially children with mild, self-limited illness.1,51,52 A culture specimen should be obtained as early in the course of disease as possible from patients in whom the diagnosis of hemolytic–uremic syndrome is suspected, who have bloody diarrhea or whose stools contain fecal leukocytes, during outbreaks of diarrhea, in people with diarrhea who are immunosuppressed,1,53 and for public health reasons. Fecal specimens that cannot be inoculated immediately on to solid media can be transported to the laboratory in a nonnutrient-holding medium such as Cary–Blair transport medium to prevent drying or overgrowth of specific organisms. Certain bacterial agents, such as Y. enterocolitica, diarrhea causing E. coli organisms, Vibrio cholerae, V. parahaemolyticus, Aeromonas spp., C. difficile, and Campylobacter spp., require modified laboratory procedures for identification and are often overlooked in routine stool cultures. Laboratory personnel should be notified when one of these organisms is suspected to be the causative agent. For further characterization of E. coli, serotype and toxin assays are available in reference and research laboratories (see Chapter 137, Escherichia coli). Information about specific diagnostic techniques used to detect other enteropathogens is available in the individual chapters. Proctosigmoidoscopy may be helpful in establishing a diagnosis in patients in whom symptoms of colitis are severe or the etiology of an inflammatory enteritis syndrome remains obscure after initial laboratory evaluation. Newer methods that do not require culture of organisms, such as enzyme immunoassay and DNA probes, are becoming increasingly available and are highly sensitive.

TREATMENT All patients with diarrhea require some level of fluid and electrolyte therapy and attention to diet, a few need other nonspecific support, and some may benefit from specific antimicrobial therapy.

Fluid and Electrolyte Therapy Oral rehydration therapy (ORT) is the preferred treatment for fluid and electrolyte losses due to diarrhea in children with mild to moderate dehydration.2,17 Surveys show that healthcare providers do not always follow such recommended procedures.54 Stool losses of water, sodium, potassium, chloride, and base must be restored in children with dehydration to ensure effective rehydration.55,56 In the mid-1960s, the discovery of coupled transport of sodium and glucose (or other small organic molecules) provided scientific justification for oral rehydration as an alternative to intravenous therapy.57 In the initial studies, conducted in patients with cholera in Bangladesh and India in the late 1960s, oral glucose-electrolyte solutions compared favorably with standard intravenous therapy.58,59 The solutions used were similar to the oral rehydration salt solution recommended by the World Health Organization and United Nations Children’s (Emergency) Fund, which has been used successfully throughout the world for more than 20 years. Since the early 1980s, a series of studies from developed countries have proved the effectiveness of ORT compared with intravenous therapy in children with diarrhea from causes other than cholera,60–64 showing a reduction in subsequent unscheduled follow-up visits.60 These studies evaluated glucose-electrolyte ORT solutions with sodium concentrations ranging from 50 to 90 mmol/L, in comparison with rapidly administered intravenous therapy. These ORT solutions were successful in rehydration of more than 90% of dehydrated children and had complication rates lower than rates for intravenous therapy.65 Oral rehydration is less expensive than intravenous rehydration, hospitalization often can be prevented, and ORT can be administered in many settings.62 The frequency and volume of stools, duration of diarrhea, and rate of weight gain are similar with the two therapies.60–64 In 2002 the World Health Organization and the United Nations Children’s Fund recommended that formulation of ORT for treatment of patients with diarrhea be changed to one with reduced osmolarity.2,66 This reduced-osmolarity ORT is not associated with an increased risk for hyponatremia.

382

SECTION

H Gastrointestinal Tract Infections and Intoxications

Most oral solutions available in the United States (Table 59-5) have sodium concentrations ranging from 45 to 50 mmol/L, which is at or just below the lower concentration of the solutions studied. Although these products are best suited for use as maintenance solutions, they can be used to rehydrate otherwise healthy children who are mildly or moderately dehydrated.61,67,68 Glucose-electrolyte solutions such as these, which are formulated on physiologic principles, must be distinguished from other popular but nonphysiologic liquids that have been used inappropriately to treat children with diarrhea, which include cola, apple juice, sports beverages, and chicken broth. These beverages have inappropriately low electrolyte concentrations and are hypertonic because of their carbohydrate content.69

Dietary Intake Early feeding of age-appropriate foods to children with diarrhea after rehydration should be an integral component of the therapeutic regimen. When used with glucose-electrolyte ORT, early feeding can reduce stool output as much as can cereal-based ORT.65,70 Studies of early-feeding regimens including human milk,71–73 dilute or fullstrength animal milk or animal milk formulas,71–73 dilute or fullstrength lactose-free formulas,60,71,74 and staple-food diets with milk73,75–77 have shown that unrestricted diets do not worsen the course of symptoms of mild diarrhea72,73 and can decrease stool output,76,77 in comparison with rehydration therapy alone. Meta-analysis of literature from developed countries on early refeeding72,78,79 shows that the duration of diarrhea may be reduced by 0.43 days (95% confidence interval, –0.74 to –0.12).2 Although these beneficial effects are modest, the added benefit of improved nutrition with early feeding has major importance.56,74 One meta-analysis showed that > 80% of children with acute diarrhea can safely tolerate full-strength milk.80 Although reduction in intestinal brush-border lactase levels is often associated with diarrhea,81 most infants with decreased lactase levels do not have clinical signs or symptoms of malabsorption.81 Infants fed human milk can be safely nursed during an episode of diarrhea.71 If children are monitored so that the few who demonstrate signs of malabsorption may be identified, a regular age-appropriate diet, including fullstrength milk, can be safely used. During the refeeding period, certain foods, including complex carbohydrates (rice, wheat, potatoes, bread, and cereals), lean meats, yogurt, fruits, and vegetables, are better tolerated than other foods.65,70,76,77 Fatty foods and foods high in simple sugars (including juices and soft drinks) should be avoided.2

adsorption of toxins or fluid; and (4) alteration of intestinal microflora. Many of these compounds, especially agents that alter motility, have systemic toxic effects that are augmented in infants and children or in people with diarrheal disease; most are not approved for children younger than 2 or 3 years of age. Few published data are available to support the use of most antidiarrheal agents to treat acute diarrhea, especially in children.2,82 Table 59-6 lists generic and brand names of the drugs commonly used to treat people with diarrhea.

Antimicrobial Therapy Most children with acute infectious diarrhea would not benefit from antimicrobial therapy. Antimicrobial agents are of no benefit for children infected with enteric viruses, including rotavirus, enteric adenovirus, astrovirus, and norovirus. In addition, problems can occur with development of resistance, side effects of treatment, and superinfections when normal flora are eradicated. Patients with diarrhea associated with certain bacterial and protozoal agents may benefit from therapy (Table 59-7). Limited benefit can be achieved with therapy of people infected with Aeromonas species (trimethoprimsulfamethoxazole), Blastocystis hominis (metronidazole or iodoquinol), TABLE 59-6. Medications Used to Relieve Symptoms in Patients with Acute Diarrhea Mechanism

Generic Name

Alteration of intestinal Loperamide motility Difenoxin and atropine Diphenoxylate and atropine Tincture of opium Alteration of secretion Bismuth subsalicylate Adsorption of toxins and water

Attapulgite

Alteration of intestinal Lactobacillus microflora

Trade Name Imodium Imodium A-D Maalox Antidiarrheal Pepto Diarrhea Control Motofena Lomotila Paregorica Pepto-Bismol Kaopectate Diasorb Donnagel Rheaban Pro-Bionate Superdophilus

a

Requires prescription.

TABLE 59-7. Potential Benefit of Antimicrobial Therapy for Enteropathogens

Nonspecific Therapy A variety of pharmacologic agents have been used as nonspecific therapy for people with acute infectious diarrhea. These compounds can be classified according to the following mechanisms of action: (1) alteration of intestinal motility; (2) alteration of secretion; (3)

Potential Benefit

Enteropathogen or Disease

No therapy available

Enteric viruses

Established benefit

Clostridium difficile colitis Cryptosporidium parvum Cyclospora cayetanensis Entamoeba histolytica Enterotoxigenic Escherichia coli Giardia lamblia Isospora belli Shigella species Strongyloides stercoralis Vibrio cholorae

Absolute benefit

Any bacterium that produces bacteremia (e.g., Salmonella typhi)

Limited or unknown benefit

Aeromonas species Blastocystis hominis Campylobacter jejuni Enterohemorrhagic Escherichia coli Intestinal salmonellosis Microsporidia Yersinia enterocolitica

TABLE 59-5. Composition of Representative Glucose-Electrolyte Solutions Concentration (mmol/L) Solutions

CHOa (g/L)

Na

CHO:Na K

Base

Osmolality (mosm/L)

CeraLyte-50 Pedialyte Enfalyte Rehydralyte WHO (1975) WHO (2002)

40 25 30 25 20 13.5

50 45 50 75 90 75

3.1 3.1 1.4 1.9 1.2 1.2

30 30 30 30 30 30

220 250 200 310 310 245

20 20 25 20 20 20

CHO, carbohydrate; WHO, World Health Organization. a All commercial oral rehydration salts solutions except Ceralyte contain glucose. Ceralyte contains a complex mixture of rice CHOs and proteins.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Viral Gastroenteritis

Campylobacter jejuni (erythromycin or azithromycin, if given early in the course of therapy), Cryptosporidium (nitazoxanide), and intestinal microsporidia (albendazole).83 Therapy of patients with intestinal salmonellosis and Y. enterocolitica is of no known benefit. A metaanalysis of antibiotic therapy of children with E. coli O157:H7 did not show a higher risk of hemolytic–uremic syndrome associated with antibiotic administration.84

PREVENTION The most important aspect in control of diarrheal disease is hygiene, both public and personal. Public issues deal with clean water, clean food, and appropriate sanitation facilities. Despite the high-quality water and food supplies available in the United States, outbreaks of foodborne and waterborne disease continue to occur15,16 (see Chapter 63, Foodborne and Waterborne Disease), as illustrated by the large outbreak of cryptosporidiosis in 1993 involving > 400 000 people in Milwaukee due to contamination of the public water supply.85 Personal measures include careful personal hygiene, especially hand hygiene,39 and limited use of antacids, antimotility drugs, and antimicrobial agents. Appropriate diaper-changing facilities and techniques should be available and implemented in childcare centers.10 Written infection control policies must be in place and in use in hospitals, extended-care facilities, and childcare facilities. Guidelines for preventing opportunistic infections in people with HIV and among hematopoietic stem cell transplant recipients have been published.37,38 Currently, vaccines against Salmonella typhi5 and rotavirus3,4 are the only vaccines against enteric disease commercially available in the United States. Hepatitis A vaccines are part of the recommended childhood and adolescent immunization schedule.86 There are no effective vaccines against parasitic enteric infection. Vaccines against other enteric pathogens and improved vaccines against pathogens for which immunizations are available are under study. These vaccines are directed against the organisms themselves, adherence factors, cytotoxins, or enterotoxins.87 Other nonspecific agents that may interfere with microbial adherence or with the virulence mechanisms of toxins are being developed and evaluated, as are compounds that serve as competitors for binding of organisms or toxins to receptors in the intestine. The areas of glycoconjugates and probiotics and their role in disease prevention are under intense investigation.88,89 Breastfeeding provides young infants with significant protection against morbidity and mortality due to diarrheal disease.90 Breastfeeding protects against diarrhea in part through decreased exposure of breastfed infants to organisms present on or in contaminated bottles, food, or water. In addition, immunologic components in human milk protect infants against disease after exposure to an infectious agent.

CHAPTER

60

Viral Gastroenteritis Joseph Bresee Viral gastroenteritis is the most common disease associated with acute vomiting and diarrhea among children, and remains a leading cause of pediatric morbidity and mortality worldwide. With discovery of both Norwalk virus1 and rotavirus2 in the early 1970s, and subsequent development of improved diagnostic strategies for these and other enteric viruses,3,4 the importance of viral agents as causes of diarrheal disease increasingly has been appreciated. The most important viral agents associated with gastroenteritis in children include rotaviruses, caliciviruses, adenoviruses, and astroviruses (Table 60-1). Many other

CHAPTER

60

383

TABLE 60-1. Relative Distribution of Viral Pathogens as Causes of Acute Gastroenteritis among Children Agent Rotaviruses Noroviruses Sapoviruses Astroviruses Adenoviruses 40/41

Hospitalizationsa (%) 25–50 5–30 90% of children have antibody to human astroviruses by 6 to 9 years of age.75 Disease in adults is uncommon, but can occur in outbreak settings.76 Astroviruses have usually been detected in < 10% of young children treated for gastroenteritis in outpatient clinics or in hospitals, but are the most common virus detected in a few studies.4,51,77 While astroviruses primarily cause sporadic disease, outbreaks have been reported in closed settings such as schools,76 childcare centers,31,47 hospitals,78 nursing homes,79 and households.80 Astroviruses have been reported to be responsible for 5% to 16% of nosocomial gastroenteritis in children’s hospitals, second only to rotavirus.81 Most adenovirus infections occur in children < 2 years of age, and they appear to be less important causes of gastroenteritis among adults.82,83 Enteric adenoviruses account for 5% to 10% of hospitalizations for acute gastroenteritis in children and may be a common cause of healthcare-associated diarrhea.82–85 Enteric adenoviruses are generally detected in 1% to 4% of children with communityassociated diarrhea.57,86 In economically developing countries, enteric

CHAPTER

60

385

adenoviruses, compared with other viral agents, appear to account for a smaller proportion of diarrheal disease than in developed countries, although they are occasionally detected in rates similar to rotavirus.58 All age groups are infected by noroviruses, but serosurveys document that antibody is acquired at an early age, indicating that first exposure to these viruses occurs early in life.87 Noroviruses are the most common cause of foodborne illness in the United States.33 They are estimated to cause 23 million illnesses a year in the United States and 93% of all nonbacterial outbreaks reported by the Centers for Disease Control and Prevention between 1997 and 2000. Common foods associated with outbreaks include uncooked foods contaminated by ill foodhandlers who are shedding virus, and shellfish harvested from contaminated water.88 Approximately half of norovirusassociated outbreaks occur through person-to-person spread in closed populations, such as nursing homes, childcare centers, hospitals, and cruise ships. Norovirus outbreaks in institutions can require closing of patient-care units or suspension of cruises. In addition, noroviruses are increasingly appreciated to cause sporadic community-associated gastroenteritis among children.72 In a Finnish cohort of children between 2 months and 2 years of age, noroviruses were detected in 20% of stool specimens from episodes of gastroenteritis and 13% of hospitalized cases, second only to that of rotaviruses.71,73

CLINICAL MANIFESTATIONS After a short incubation period, infections with any of the viruses lead to an acute onset of gastroenteritis (Table 60-2). The clinical characteristics of illnesses caused by the different viruses are generally indistinguishable.85,89–91 Vomiting is often an early sign, and particularly pronounced in norovirus infections. Diarrhea is frequent, watery, and without blood or visible mucus. Fever occurs in approximately half of children and is often an early sign. Vomiting and fever often cease within 1 to 3 days, whereas diarrhea can persist for additional days. Other symptoms include abdominal cramps and malaise. Stools generally do not contain blood or fecal leukocytes. The most important and common complication of viral gastroenteritis is dehydration, often with electrolyte abnormalities. Malabsorption can occur during the illness and persist for weeks following infection. Simultaneous respiratory tract symptoms may occur but are likely due to concurrent wintertime respiratory tract viral infections. Extraintestinal complications are rare, but encephalitis, acute myositis, hemophagocytic lymphohistiocytosis, polio-like paralysis, and sudden infant death syndrome have been described rarely in children with rotavirus infections.5 Their relationship to rotavirus infection remains unclear. Long-term complications are not associated with viral gastroenteritis. Prolonged diarrhea associated with each agent has been reported among children with malnutrition and among immunocompromised patients. While viral etiologies of cases of gastroenteritis are not distinguishable by clinical signs and symptoms, clinical characteristics of cases in outbreak settings have been helpful in predicting the presence of noroviruses. Kaplan and colleagues found that outbreaks that met simple epidemiologic and clinical criteria were likely to have been caused by noroviruses.92 These criteria included: (1) failure to detect a bacterial or parasitic pathogen in stool specimens; (2) the occurrence of vomiting in > 50% of patients; (3) mean duration of illness of 12 to 60 hours; and (4) mean incubation period of 24 to 48 hours. The “Kaplan criteria” have been widely used by local health departments for the diagnosis of outbreaks in the absence of laboratory testing.

DIAGNOSIS Laboratory diagnosis of viral gastroenteritis is best made by detection of viral antigen or nucleic acid in fresh, whole stool samples obtained during the acute illness. Commercially available assays to detect rotavirus antigen in stools offer an easy and inexpensive method to diagnose infection in children. These tests are available as either

386

SECTION

H Gastrointestinal Tract Infections and Intoxications

TABLE 60-2. Epidemiologic Features of Viral Agents of Gastroenteritis Feature

Rotavirus

Noroviruses

Sapoviruses

Astroviruses

Adenoviruses

Age of illness Mode of transmission

< 5 years Person-to- person via fecal–oral route, fomites

< 5 years Person-to- person via fecal–oral route

< 2 years Person-to- person via fecal–oral route

< 2 years Person-to- person via fecal–oral route

Incubation period

1–3 days

All ages Person-to- person via fecal–oral route, fomites, food/water 12–48 hours

12–48 hours

1–4 days

3–10 days

Watery; milder than rotavirus Less common than rotavirus Less common, usually mild 1–4 days

Watery; milder than rotavirus Less common than rotavirus Less common, usually mild 1–5 days

Watery; milder than rotavirus; can be prolonged Less common than rotavirus Less common, usually mild

RT-PCR

Stool EIA (not available Stool EIA in United States)

SYMPTOMS

Diarrhea Vomiting

Explosive, watery (5–10 episodes/day) 80–90%

Fever

Frequent

Illness duration

2–8 days

Watery with acute onset > 50%; often dominant symptom Less common, usually mild 1–5 days

PRINCIPAL METHODS

Stool EIA or LPA

RT-PCR

OF CLINICAL DIAGNOSIS

3–10 days

EIA, enzyme immunoassay; EM, electron microscopy; IEM, immune electron microscopy; LPA, latex particle agglutination; RT-PCR, reverse transcriptase-polymerase chain reaction. Modified from Peck AJ, Bresee JS. Viral gastroenteritis. McMillan JA, In: Feigin RD, De Angelis CD, Jones MD (eds) Oski’s Pediatrics, 4th ed. Philadelphia, PA, Lippincott, Williams and Wilkins, 2006, pp 1288–1294.

enzyme immunoassay (EIA) or latex particle agglutination test for group A rotaviruses, designed to detect the VP6 protein.5 Antigen detection tests generally have a high (90% to 95%) sensitivity and specificity.93 Other methods for rotavirus detection, including electron microscopy, viral isolation, polyacrylamide gel electrophoresis (PAGE) of RNA extracted directly from stool, and reverse transcription-polymerase chain reaction (RT-PCR), are available in research settings, but are rarely used in clinical practice.93 Serologic testing for rotavirus infection is possible but impractical, thus is not widely available in clinical care settings. Immunohistochemical stains have been developed that identify rotavirus antigen in pathologic tissues, and are available in some research and public health settings. Commercial antigen detection kits are available for caliciviruses, but are not available in the United States. Commercial EIA tests for caliciviruses have poor sensitivity but may be useful in outbreak investigations.94 RT-PCR has become the standard diagnostic assay used for caliciviruses, but is seldom used clinically. RT-PCR has become widely available in public health laboratories for outbreak investigations, where sequencing of the PCR product from clinical samples may allow for linking cases to each other and to a common source.88 Caliciviruses have not been reproducibly propagated in cell cultures. Commercial EIAs for detection of astrovirus viral antigen in stool are available in Europe, but not in the United States.4 Similarly, RT-PCR is a sensitive and specific method for detection of astroviruses. RT-PCR, serologic assays, and electron microscopy are primarily used in research settings. Similarly, EIA and latex particle agglutination kits are available commercially and provide highly sensitive and specific antigen detection of enteric adenoviruses.95 All viral gastroenteritis agents are detectable by electron microscopy and immune electron microscopy, but these tests are seldom used because of relatively low sensitivity and specificity, expense, and required expertise.

TREATMENT No specific therapies are available for viral gastroenteritis. Case management depends on accurate and rapid assessment, correction of fluid loss and electrolyte disturbances, and maintenance of adequate hydration and nutrition.96 Oral rehydration therapy with appropriate glucose-electrolyte solutions is sufficient for most patients (Table 60-3). Intravenous rehydration may be required for children with severe dehydration with shock or intractable vomiting. Breastfed infants should continue to nurse on demand. Infants receiving formula should continue their usual formula upon rehydration. Children taking

solid foods should continue to receive their usual diet during episodes of diarrhea, although substantial amounts of foods high in simple sugars should be avoided because the osmotic content might worsen diarrhea. Because viral agents account for the large majority of infectious gastroenteritis in children, appropriate use of antimicrobial agents in patients with acute gastroenteritis should be stressed. Some evidence exists to support the use of oral probiotics, such as Lactobacillus species, that reduce the duration of diarrhea caused by rotavirus.97 Human or bovine colostrums and human serum immunoglobulin that contain antibodies to rotavirus may be beneficial in decreasing or preventing rotavirus diarrhea, but are not used in routine practice.98–101

PREVENTION Except for rotavirus, prevention of viral gastroenteritis is limited to nonspecific strategies. Breastfeeding confers some protection against rotavirus infection, and probably astrovirus infections, in young infants; it is likely mediated through rotavirus antibodies and other nonimmunologic factors in the milk. Good hygiene, including hand hygiene practices, is an effective prevention strategy and should be encouraged, particularly in institutional settings, such as childcare centers and hospitals.102 Noroviruses are relatively resistant to environmental disinfection, but cleaning contaminated surfaces and food preparation areas with household chlorine bleach-based cleaners can decrease spread of infection with these viral agents and is likely effective in settings where rotavirus and astrovirus outbreaks occur.103 However, significantly reducing transmission of viral agents of gastroenteritis is difficult because the disease generally requires a low infectious dose, high quantity of viruses are excreted in stool (and often vomitus) from infected persons, and the agents are quite stable in the environment. The best option for preventing rotavirus morbidity and mortality is the use of live, oral rotavirus vaccines in routine immunization programs. Rotavirus vaccines are attenuated strains given in multiple doses designed to replace a child’s first exposure to wild-type rotavirus with strains that will not cause disease but will generate an adequate immune response to confer protection.104,105 Two rotavirus vaccines are licensed in the world. One was licensed in the United States in 2006 and recommended for use universally.106,107 Additional vaccines are in late stages of FDA submission or development and are expected to be available within the next several years.105 While most vaccines in clinical development are live, orally administered vaccines, parenterally administered vaccines are also being investigated.104

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Inflammatory Enteritis

CHAPTER

61

387

TABLE 60-3. Composition of Commercial Oral Rehydration Solutions and Commonly Consumed Beverages Solution

Basea (mmol/L)

CHO (g/L)

Na (mmol/L)

K (mmol/L)

Cl (mmol/L)

Osmolarity (mosmol/kgH2O)

WHO-ORS (2002)

13.5

75

20

65

10

245

WHO-ORS (1975)

20

90

20

80

10

311

ORAL REHYDRATION SOLUTIONS

ESPGHAN ORS

16

60

20

60

10

240

Enfalyteb

30

50

25

45

34

200

25

45

20

35

30

250

25

75

20

65

30

305

40

50–90

20

40–80

30

220

Apple juicee

120

0.4

44

45

–

730

Coca-Colaf

112

1.6

–

–

13.4

650

46

23.5

2.5

17

3

330

0.5

260

–

450

–

–

6

Pedialytec c

Rehydralyte d

Ceralyte

COMMONLY USED BEVERAGES

g

Gatorade

Chicken broth

e

8

e

Tea

–

260 6

–

WHO-ORS, World Health Organization oral rehydration solution. Reprinted from CDC. Managing acute gastroenteritis among children. MMWR 2003;52:1–16. a Actual or potential bicarbonate, such as lactate, citrate, or acetate. b Mead-Johnson Laboratories, Princeton, NJ. c Ross Laboratories, Columbus, OH (data for Flavored and Freezer Pop Pedialyte are identical). d www.ceralyte.com/index.htm, accessed April 25, 2003. e United States Department of Agriculture. f Coca-Cola Corporation, Atlanta, GA (figures do not include electrolytes, which may be present in local water used for bottling; base = phosphate). g The Gatorade Company, Chicago, IL.

Neither of the licensed vaccines, nor any vaccine in development, has yet been tested in countries with high mortality rates. Because past rotavirus vaccines have shown little efficacy in these settings,68,108–112 trials of rotavirus vaccines are under way; decisions about use of these vaccines in countries where mortality is high will await the results of these trials.68 While experimental vaccines against noroviruses are in early stages of development, proof that these vaccines could be protective remains to be established.113 No vaccines against other caliciviruses, astroviruses, or enteric adenoviruses are yet in human trials.

CHAPTER

61

Inflammatory Enteritis Ina Stephens and James P. Nataro

Inflammatory enteritis is a pathologic diagnosis characterized by ileal and/or colonic inflammation, which can range from small, superficial patches of leukocyte infiltration amid relatively normal mucosa to deep, exudative ulcerations with involvement of the entire intestinal wall.1 The large number of infectious and noninfectious causes of inflammation results in a wide range of manifestations (Table 61-1). Inflammatory enteritis should be considered in the differential diagnosis of patients with abdominal pain, dysentery, or watery diarrhea, particularly when patients present with fever and blood or mucus in stool.

TABLE 61-1. Differential Diagnoses of Inflammatory Enteritis Acute Inflammatory Enteritis

Chronic Inflammatory Enteritis (≥ 2 Weeks)

BACTERIAL

BACTERIAL

Aeromonas hydrophila Campylobacter jejuni/coli Enteroaggregative Escherichia coli Enterohemorrhagic Escherichia coli Enteroinvasive Escherichia coli Plesiomonas shigelloides Salmonella spp. Shigella spp. Vibrio parahaemolyticus Yersinia enterocolitica

Campylobacter jejuni/coli Enteroaggregative Escherichia coli Mycobacterium tuberculosis Salmonella spp. Shigella spp. FUNGAL

Candida spp. Histoplasma capsulatum Paracoccidioides brasiliensis Phycomycosis agentsa

PARASITIC

Balantidium coli Cryptosporidium parvuma Entamoeba histolytica Schistosoma species Strongyloides stercoralis Trichinella spiralis VIRAL

Cytomegalovirusa Enterovirusesa a

Primarily in immunocompromised patients.

Proctitis Chlamydia trachomatis Entamoeba histolytica Herpes simplex virus Neisseria gonorrhoeae Shigella spp. Treponema pallidum

388

SECTION

H Gastrointestinal Tract Infections and Intoxications

CLASSIFICATION Inflammatory enteritides can be classified into several distinct clinical entities, the recognition of which facilitates diagnosis and management. However, many patients with inflammatory enteritis do not easily fit into any of these categories, thereby broadening their differential diagnosis. Acute dysentery is defined by the presence of fecal blood and mucus, associated with frequent, small loose bowel movements. Fever is often present. Dysentery is the result of microbial invasion of the colonic mucosa, with mucosal and submucosal inflammation and destruction. Shigellosis is a classic cause of acute dysentery. Chronic inflammatory enteritis is typically an indolent and slowly progressive illness. The patient can manifest fever, abdominal pain, and weight loss over several weeks. Recurring, relapsing symptoms commonly occur. Mycobacterium tuberculosis is a classic cause of chronic inflammatory enteritis; children in the United States with chronic inflammatory enteritis are more likely to have inflammatory bowel disease. Ulcerative proctitis is characterized by severe anorectal pain, purulent discharge, tenesmus, hematochezia, and fever. Erythematous, friable mucosa is visualized in the rectal vault.2,3 Severe disease can lead to rectal abscesses or fistulas. This condition, which usually occurs in men who have sex with men and in immunocompromised patients, is caused by a variety of agents. Necrotizing enterocolitis (NEC) of the newborn is characterized by varying degrees of mucosal necrosis.4–6 NEC occurs almost exclusively in premature infants, who typically demonstrate a sepsislike illness with abdominal distention and, commonly, grossly bloody stools. Although many factors contribute to development of intestinal necrosis, a common pathway features ischemia, followed by disruption of the mucosal barrier, bacterial invasion, proliferation, and gas formation within the bowel wall.4,5 Epidemics of NEC have been associated with a variety of organisms, including Escherichia coli, Klebsiella, Pseudomonas, and Salmonella spp., and rotavirus. However, prospective studies have failed to implicate any specific infectious cause in most cases (see Chapter 62, Necrotizing Enterocolitis).6 Antimicrobial-associated colitis (AAC) is characterized by the presence of multiple yellow plaquelike ulcerative lesions (1 to 5 mm

in size) overlying an erythematous, friable, colonic mucosa. In severe cases, lesions become confluent, resulting in sloughing of necrotic tissue and formation of the characteristic pseudomembrane.7–9 Clinically, the presentation of AAC varies from mild watery diarrhea (with or without crampy abdominal pain) to fulminant colitis that can progress to toxic megacolon, colonic perforation, shock, and death. Overgrowth of toxigenic Clostridium difficile with release of potent cytotoxins is responsible for almost all cases.7,8 C. difficile colitis in the absence of antibiotic therapy is unusual; however, use of both proton pump inhibitor (PPI) and H2-receptor agonist drugs are associated with increased risk of community-acquired C. difficileassociated disease.9 A potential mechanism for this effect is loss of acidity with increased spore survival. Data suggest that the rate and severity of C. difficile-associated disease in the United States are increasing,9,10 and this increase may be associated with the emergence of newer strains with increased virulence, antimicrobial resistance or both. These strains, which carry variations in toxin genes and have also become more resistant to the fluoroquinolones, have emerged as a cause of outbreaks of C. difficile disease.10 Of particular concern is the frequency of C. difficile colitis in patients with HIV, where C. difficile can account for up to 54% of cases of bacterial diarrhea.11 Noninfectious causes of inflammatory enteritis include ulcerative colitis, Crohn disease, Henoch–Schönlein purpura, eosinophilic gastroenteritis, enterocolitis complicating Hirschsprung disease, Behçet disease, and allergic colitis.12 Intussusception can cause acute onset of bloody, diarrheal stools, but inflammatory exudate is not expected.

PRINCIPAL ETIOLOGIC AGENTS A variety of infectious agents classically are considered as causes of acute inflammatory enteritis, in which patients come to medical attention with fever, abdominal pain, blood or mucus in stool, or a combination of these findings. However, many of these pathogens can cause clinical illnesses without overt characteristics of inflammation. Absence of clinical signs of inflammation does not therefore rule out these agents (Table 61-2). In the United States, most cases of inflammatory enteritis in children are caused by bacterial pathogens, whereas parasitic agents are common among travelers to or indigenous

TABLE 61-2. Pathogenesis and Complications of Principal Infectious Causes of Inflammatory Enteritis Agent

Pathogenesis

Intestinal Complications

Extraintestinal Complications

Shigella spp. and enteroinvasive Escherichia coli

Invasion of colonic enterocytes with lateral spread through mucosa and submucosa77; enterotoxins possibly involved36

Colonic perforation, protein-losing enteropathy

Seizures Septicemia HUS Immune-mediated disease (reactive arthritis, Reiter syndrome)

Salmonella spp.

Invasion of distal ileal mucosa; can Colonic perforation (enteric fever), spread to reticuloendothelial system78 chronic carriage

Septicemia Meningitis, disseminated foci

Campylobacter jejuni/coli

Invasion of colonic enterocytes79

Dissemination Immune-mediated disease (Guillain–Barré syndrome, reactive arthritis)

Yersinia enterocolitica

Invasion of ileal mucosa with Mesenteric adenitis inflammatory response; enterotoxins possibly involved19,37

Chronic disease

Enterohemorrhagic Escherichia coli Attaching and effacing lesions of colonic mucosa; production of Shiga-like toxins linked to HUS29

Toxic megacolon, colonic perforation

Clostridium difficile

Cytotoxin production

Toxic megacolon, colonic perforation

Entamoeba histolytica

Contact-dependent and contactChronic disease, toxic megacolon, independent cytotoxicity; proteases; colonic perforation secretory factors possibly involved26

Dissemination with septicemia, suppurative foci HUS Seizures Septicemia Ameboma, liver abscess, disseminated foci

HUS, hemolytic–uremic syndrome.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Inflammatory Enteritis

populations in developing areas. Generally, differentiation among bacterial enteropathogens on clinical grounds alone is not possible; pathogenesis and potential complications associated with these agents are listed in Table 61-2.

Shigella Species All four Shigella species can elicit prototypic acute bacillary dysentery, but S. sonnei most often elicits an uncomplicated watery diarrhea and is most prevalent in developed countries.13,14 Shigella spp. have no known animal reservoir. Organisms are highly contagious and have a low infectious dose (as low as 102 colony-forming units), a feature hypothesized to be related to the organism’s acid resistance.15 Personto-person and foodborne transmissions are most common in childhood infections14; environmental contamination also can be a source of transmission.

Salmonella Species Nontyphoidal Salmonella serotypes cause a spectrum of illness that includes mild diarrhea, dysentery-like inflammatory enteritis, and disseminated disease, including septicemia, enteric fever, and focal seeding (especially in immunocompromised hosts).16 Most infections due to Salmonella in the United States are foodborne.17 Transmission most commonly occurs via contaminated animal products such as unpasteurized milk, eggs, and improperly cooked meat.

Escherichia coli Diarrheagenic E. coli are divided into five major pathogenic categories on the basis of mechanism of pathogenesis (see Chapter 137, Escherichia coli).18 Categories are: (1) enterotoxigenic (watery diarrhea in infants and travelers); (2) enteroinvasive (dysentery identical to that of S. sonnei); (3) enteropathogenic (acute and chronic watery diarrhea in infants and children); (4) enterohemorrhagic (hemorrhagic colitis associated with the hemolytic–uremic syndrome); and (5) enteroaggregative (acute or persistent watery diarrhea, prevalent in developing countries, and in the immunocompromised host). Each type of E. coli except enterotoxigenic strains can cause inflammatory enteritis. Both enterohemorrhagic and enteroaggregative E. coli pathotypes occur commonly in the United States.

Yersinia enterocolitica Yersinia enterocolitica typically causes acute diarrhea, fever, and abdominal pain.19 Other manifestations of Yersinia infection are septicemia in infants and immunocompromised hosts, mesenteric lymphadenitis (mimicking appendicitis) in older children and young adults, dysentery, and postinfectious syndromes, including reactive arthritis, Reiter syndrome, glomerulonephritis, and erythema nodosum.20 In the United States, most Y. enterocolitica transmissions are foodborne, with outbreaks often due to unpasteurized milk and contaminated pork or chitterlings.21

Campylobacter jejuni Campylobacter jejuni is one of the most commonly documented bacterial causes of diarrhea among people of all ages in the United States and Canada.22 Inflammatory enteritis and acute dysentery with severe abdominal pain and fever are common manifestations.22 Bacteremia and dissemination are rare. The reservoir for C. jejuni is the gastrointestinal tract of wild and domestic birds and, to a lesser extent, animals, allowing for foodborne outbreaks via ingestion of contaminated water, raw milk, or undercooked meat and poultry.

CHAPTER

61

389

Person-to-person spread has been documented, although such transmission is less likely than in shigellosis, as the infectious dose is considerably higher at 104–6 organisms.22

Aeromonas hydrophila Aeromonas is a genus of gram-negative bacilli found in soil and in fresh and brackish waters worldwide. A. hydrophila has been associated with acute gastroenteritis, soft-tissue infection, and bacteremia, especially in immunocompromised hosts. The pathogenetic mechanism of intestinal disease is unknown. Diarrhea due to A. hydrophila is associated with the consumption of seafood.23

Vibrio parahaemolyticus Vibrio spp. other than V. cholerae can elicit acute inflammatory enteritis with explosive, watery diarrhea. Illness usually is caused by V. parahaemolyticus acquired through consumption of undercooked seafood.24 Some non-O1 V. cholerae have also been implicated in inflammatory enteritis.25

Entamoeba histolytica The classic clinical manifestation of Entamoeba histolytica enteritis is amebic dysentery, which can be acute and fulminant (especially in children), or mild and insidious with only abdominal pain, tenesmus, and small, frequent bowel movements.26,27 Illness often follows an intermittent or waxing and waning course over weeks to months. Long incubation periods (1 week to 1 month) allow onset of symptoms long after return home from travel to a developing country where infection was acquired. Dissemination, most frequently to the liver, can occur after resolution of untreated amebic dysentery. Examination of the colonic mucosa of individuals with dysentery often reveals shallow, flasklike ulcers.26 Trophozoites are characteristically found in stools of patients with dysentery or diarrhea; cysts are more commonly found in asymptomatic people.

EPIDEMIOLOGY The epidemiology of infectious inflammatory enteritis is determined by the specific causative agent and susceptibility of the host (see pathogen-specific chapters). Predictive epidemiologic associations should be explored in the patient’s history, including foreign travel, childcare or institutional exposure, swimming and other exposure to water sources, receipt of antibiotics, diet, animal exposure, and underlying conditions. (Risk factors and associated agents are shown in Table 61-3.) Infectious inflammatory enteritis often occurs in outbreaks related to person-to-person, foodborne, or waterborne transmission. Etiology of outbreaks of enteric disease can often be predicted from the incubation period.28 Cases with onset between 1 and 16 hours of source event are not usually inflammatory in nature and are toxinmediated; those with onset between 16 and 72 hours can be caused by infection, such as due to Salmonella or Shigella spp., C. jejuni, enteroinvasive Escherichia coli, V. parahaemolyticus, and Y. enterocolitica; an incubation period > 72 hours suggests enterohemorrhagic E. coli O157:H7.28,29 The implicated vehicle of transmission can also be helpful in predicting the cause. Cysts of parasites, including Cryptosporidium parvum, Balantidium coli, and Entamoeba histolytica, resist chlorination and thus contaminate municipal water supplies. Pathogens associated with specific foods are Salmonella spp. (undercooked poultry and eggs), Campylobacter jejuni (undercooked poultry), V. parahaemolyticus (raw and undercooked seafood), and Y. enterocolitica (pork, chitterlings, milk). Neisseria gonorrhoeae, herpes simplex virus, Chlamydia trachomatis, and Treponema

390

SECTION

H Gastrointestinal Tract Infections and Intoxications

Risk Factor

Infectious Agents

between host and parasite, the result of which favors the survival of both species. In addition, invasive pathogens can secrete enterotoxins.36,37 A summary of pathogenetic mechanisms for principal pathogens is listed in Table 61-2.

Travel

Enterotoxigenic and enteroaggregative Escherichia coli, Shigella, Salmonella, Campylobacter spp., Entamoeba histolytica

DIAGNOSTIC EVALUATION

TABLE 61-3. Epidemiologic Risk Factors Associated with Infectious Agents of Inflammatory Enteritis

Beef consumption

Enterohemorrhagic Escherichia coli, Salmonella

Poultry consumption

Salmonella, Campylobacter

Pork consumption

Yersinia enterocolitica, Salmonella spp.

Seafood consumption

Vibrio parahaemolyticus, Aeromonas hydrophilia

Water contamination

Shigella, Salmonella, Campylobacter, Cryptosporidium spp.

Institutional exposure/ childcare

Shigella, Cryptosporidium, Giardia lamblia, enterohemorrhagic E. coli

Sexual transmission

Treponema pallidum, herpes simplex virus, Neisseria gonorrhoeae, Chlamydia trachomatis, Shigella, Entamoeba histolytica

Unpasteurized milk consumption

Salmonella, Campylobacter, Yersinia enterocolitica, Mycobacterium spp.

Antimicrobial therapy

Clostridium difficile

pallidum are sexually transmitted agents of ulcerative proctitis in men who have sex with men; this population is also at increased risk for colitis caused by routine pathogens.

PATHOGENESIS All inflammatory enteritides share common features of mucosal damage and accompanying release of inflammatory mediators. The mechanism of such damage elicited by infectious agents follows one of two general paradigms, cytotoxin elaboration or mucosal invasion, although some organisms are able to execute both mechanisms. The cytotoxic mechanism is exemplified by Clostridium difficile colitis (see Chapter 190, Clostridium difficile). Two C. difficile exotoxins (toxin A, a 308-kd enterotoxin, and toxin B, a 250- to 270-kd cytotoxin) damage enterocytes, leading to mucosal necrosis, and also may be responsible for secretion of fluid into the intestinal lumen.30,31 Both toxins act by catalyzing glycosylation of Rho proteins (guanosine triphosphate-binding proteins) in target eukaryotic cells, thereby inducing derangement of the cytoskeleton and altering multiple cellular processes. In addition, C. difficile toxins induce release of cytokines from epithelial cells, monocytes, macrophages, and neuronal cells of the lamina propria; these cytokines contribute to the toxin-mediated inflammation and damage of colonic mucosa.8,32 The invasive mechanism of bacterial enteritis is exemplified by shigellosis, yersiniosis, and salmonellosis, the pathogens of which penetrate mucosa of the ileum and colon. The role of inflammation in these infections is not entirely clear, but is apparently complex and may benefit both host and pathogen.33 For example, experimental evidence suggests that migration of polymorphonuclear cells into the intestinal lumen can exacerbate both Shigella and Salmonella diarrhea.34,35 Moreover, mice that are deficient in the apoptotic (and proinflammatory) enzyme caspase 1 are resistant to Salmonella infection. However, mice deficient in certain cytokine responses are also unable to limit Salmonella infection to the intestine and are more likely to succumb to systemic disease. The clinical manifestations of inflammatory enteritis may be the result of a finely tuned co-evolution

Before embarking on diagnostic evaluation of a patient with enteric disease, one should consider whether determining the causative agent will affect management. Most gastrointestinal tract infections in the United States are caused by viruses, for which there is no specific therapy. Most viral infections do not elicit an inflammatory syndrome, and therefore, detecting the presence of acute inflammatory enteritis suggests a bacterial or parasitic etiology. One should also note that some bacterial agents of inflammatory enteritis can be treated to benefit the patient, potential contacts, or both, whereas other agents, like nontyphoid Salmonella confined to the gut, are not responsive to antimicrobial therapy. A practical, selective diagnostic approach to the patient with enteritis should be cost-effective and likely to yield information helpful to management. By far the most common infectious causes of acute inflammatory enteritis in United States children are the bacterial enteropathogens. Relevant epidemiologic factors (see Table 61-3), disease patterns in the community, and patient characteristics are useful in assessing the likelihood of a bacterial etiology.38,39 Suggestive signs and symptoms include fever, abrupt onset of diarrhea, more than four stools per day, absence of vomiting before onset of diarrhea, and presence of blood or mucus in stool. The patient with none of these features is unlikely to have a bacterial pathogen. A scoring system for predicting bacterial enteritis has been proposed, in which risk is weighted for each feature.39 However, such a quantitative approach is unlikely to determine the precise risk of inflammatory enteritis in all locales. The leukocyte assay of stool mucus is a practical, inexpensive screening test that could be used to predict probable inflammatory enteritis, helping to limit performance of stool culture.38,40,41 Sensitivity and specificity of this test in predicting bacterial enteritis have been reported to be > 80% (for 5 or more white blood cells/hpf);38 however, the reported values vary widely among studies, in part owing to differences in the definition of a positive result. Moreover, the predictive value of the assay depends on the likelihood of bacterial inflammatory enteritis as indicated by historical factors and clinical signs. Most leukocytes detected in stool are polymorphonuclear cells (mononuclear cells have also been associated with Salmonella infection).41 Presence of leukocytes is not always indicative of bacterial inflammatory enteritis (Table 61-4), as leukocytes can be present in inflammatory bowel diseases such as ulcerative colitis and Crohn disease, and fecal polymorphonuclear cells occasionally occur in some cases of viral enteritis.42 Fecal lactoferrin, detected with a commercially available rapid agglutination test, correlates closely with the presence of fecal leukocytes.43 In one study, 94% (16 of 17) of patients for whom smears of

TABLE 61-4. Association of Fecal Leukocytes with Intestinal Pathogens Present

Variable

Absent

Campylobacter Enteroinvasive Escherichia coli Shigella

Clostridium difficile Salmonella Schistosoma Vibrio parahaemolyticus Yersinia enterocolitica

Bacillus cereus Clostridium perfringens Entamoeba histolytica Enteric viruses Enteropathogenic Escherichia coli Enterotoxigenic Escherichia coli Giardia lamblia Vibrio cholerae

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Inflammatory Enteritis

fecal mucus had > 1 polymorphonuclear leukocyte/hpf also had a positive lactoferrin latex agglutination test result, whereas 3 of 19 patients with diarrhea and no polymorphonuclear leukocytes in smears and none of 7 normal controls had positive agglutination results. In another study, the fecal lactoferrin test had a better positive predictive value than fecal leukocyte smear or occult fecal blood detection for the detection of bacterial pathogens.44 Lactoferrin is stable in fecal specimens and can be detected even when leukocytes disintegrate during transport or storage. Fecal lactoferrin tests can have falsepositive results in infants who are breastfed.

Selection of Tests to Confirm Etiology When performance of stool culture is indicated on the basis of history, physical examination, or fecal leukocyte assay, specimens should be inoculated into culture medium adequate for recovery of Shigella, Salmonella, Y. enterocolitica, and Campylobacter jejuni. In the United States, all bloody stool specimens should be inoculated on to sorbitolMacConkey agar for detection of enterohemorrhagic Escherichia coli O157:H7. In addition, all stool specimens from patients with diarrhea who have a known human contact or who have exposure to a contaminated vehicle should be cultured for E. coli O157:H7. If certain epidemiologic risk factors are present, such as a history of seafood ingestion or exposure to brackish water, specialized media should be requested for isolation of Vibrio, Aeromonas, and Plesiomonas spp. Other specialized tests are sometimes indicated in evaluation of patients with diarrhea; preferred diagnostic methods for etiologic agents of inflammatory enteritis are shown in Table 61-5. Direct microscopic examination of stool specimens can reveal protozoal trophozoites, cysts, or oocysts; fungal spores; or helminthic larvae. Although parasitic diseases are not as common in the United States as elsewhere and rarely cause inflammatory enteritis, their presence should be suspected, and direct examination for them performed, in the following high-risk situations: (1) recent travel to a developing country; (2) persistent diarrhea (> 14 days); (3) immunocompromised status; (4) contact with a person with intestinal parasitic disease; and (5) occurrence of diarrhea in the presence of an outbreak of undetermined cause. Eosinophilia in a patient with inflammatory enteritis may indicate strongyloidiasis.45 If amebic dysentery is suspected on the basis of risk factors, at least three fresh stool specimens should be examined using the iodine-trichrome method. Yield can be increased by examining stool concentrates stained with iron-hematoxylin.26 Diagnosis is made visually by detection of hematophagous trophozoites or characteristic cysts in stools. Serum antibody to Entamoeba histolytica is present in 85% to 95% of patients with invasive disease but is likely to be absent in asymptomatic cyst excreters and persons with mild diarrhea.26 Intestinal biopsy (for histology, special staining for microscopic identification of pathogen, and culture) should be considered in cases of persistent inflammatory enteritis or in an immunocompromised host when diagnosis cannot be made by noninvasive methods. Specimens obtained at biopsy generally have better yield for parasites such as E. histolytica. Certain entities, such as enteropathogenic Escherichia coli, exhibit characteristic histopathology on electron microscopy (see Chapter 137, Escherichia coli). Diagnosis of Clostridium difficile colitis is established by detection of C. difficile in stool by culture tissue culture assay for cytotoxin A and B, or detection of antigens in stool by rapid enzyme immunoassays.7,46,47 However, problems with these tests include relatively slow turnaround times for stool culture and cell cytotoxin assay, and lack of sensitivity for the enzyme immunoassay.46 Stool culture results can be misleading (especially in infants), because they do not distinguish toxigenic from nontoxigenic strains of C. difficile.46,47 The best approach may be a combination of diagnostic tests used only in patients in whom there is clinical suspicion of C. difficile enteritis. Sigmoidoscopic findings of pseudomembranous colitis are beneficial in confirming diagnosis of antibiotic-associated colitis due to C. difficile.

CHAPTER

61

391

TABLE 61-5. Diagnostic Studies for Enteropathogens Organism

Diagnostic Studies

ACUTE DISEASE

Balantidium coli

Stool examination, microscopic examination of mucosal scrapings

Campylobacter

Stool culture

Cryptosporidium

Modified acid-fast/auramine or fluoresceinlabeled monoclonal antibody stain of stool, duodenal aspirate, or small-bowel biopsy specimen

Entamoeba histolytica

Stool examination; serology for invasive disease; proctosigmoidoscopy with scrapings or biopsy

Escherichia coli

Stool culture followed by gene probes, adherence assays, or serotyping

Salmonella

Stool, blood, bone marrow cultures

Schistosoma

Stool examination

Shigella

Stool culture

Yersinia enterocolitica

Stool culture, blood culture

Cytomegalovirus

Endoscopic mucosal biopsy for microscopy (viral inclusions), immunochemical staining, culture, pp65 antigen

CHRONIC DISEASE

Candida

Potassium hydroxide stain, culture

Enteroaggregative Escherichia coli

Stool culture, DNA probe, adherence to human laryngeal cell line (HEp-2)

Mycobacterium

Acid-fast examination and stool culture, blood culture; microscopic examination of biopsy specimens

PROCTITIS

Chlamydia trachomatis

Culture, gene probe

Herpes simplex

Culture, enzyme immunoassay or direct fluorescent antibody test

Neisseria gonorrhoeae

Culture

Treponema pallidum

Darkfield microscopy examination of mucosal scrapings, serology

Entamoeba histolytica

Microscopic examination of stool or intestinal biopsy tissue; serology

OTHER

Antimicrobialassociated colitis

Toxin assays, sigmoidoscopy with rectal biopsy

Necrotizing enterocolitis Abdominal radiograph

INFLAMMATORY ENTERITIS IN THE IMMUNOCOMPROMISED HOST Clinical Approach Diarrhea is a symptom in up to 50% to 60% of patients with the acquired immunodeficiency syndrome (AIDS) in the United States.11,48 Although hospital admissions attributable to diarrhea have declined, diarrhea is still a debilitating symptom in patients infected with human immunodeficiency virus (HIV), even in the era of highly active antiretroviral therapy (HAART).11,49 Digestive tract disease in children with AIDS and other immunodeficiencies can be produced by a variety of infectious agents, including HIV.11,48,50–53 Depending on the specific cause and the immunologic status of the patient, clinical

392

SECTION

H Gastrointestinal Tract Infections and Intoxications

manifestations can be acute or chronic. Patients with CD4 T-lymphocyte counts < 50 cells/mm3 and low serum albumin levels are more likely to have an infectious etiology.54 Signs and symptoms include diarrhea, vomiting, anorexia, failure to thrive, weight loss, and malabsorption. The histopathologic changes in immunocompromised patients are usually similar to changes in healthy patients, although often more severe. In addition to agents that cause inflammatory enteritis in otherwise healthy individuals, opportunistic agents of acute and chronic inflammatory enteritis have been described in immunocompromised patients, and disease due to the following agents fulfill AIDS-defining criteria of the Centers for Disease Control and Prevention: Candida, Cryptosporidium, Histoplasma, Isospora spp., Mycobacterium avium complex, Salmonella infection with septicemia, cytomegalovirus (CMV), and herpes simplex virus.11,48,51 In early studies, an infectious etiology was not found in a large proportion of patients with AIDS who had enteric disease. But subsequent studies demonstrate that aggressive workup, including endoscopy with biopsy, substantially improve diagnostic yield.52,53 These data, coupled with the broad range of pathogens seen in this compromised population (even in the era of HAART), support an intensive diagnostic approach when reasonable empiric therapy fails.55 NEC of the cecum and surrounding tissues (typhlitis) is a lifethreatening condition that occurs in profoundly neutropenic patients, usually people undergoing aggressive chemotherapy (with or without irradiation) for malignancy.56 Typhlitis has also been described in patients with AIDS.51 Typhlitis manifests as fever, abdominal pain, tenderness (especially right-sided, lower-quadrant tenderness mimicking that of acute appendicitis), and diarrhea. Evidence of right-sided colonic inflammation can be obtained using computed tomography, ultrasonography, and plain radiography.56 An approach to evaluation of the immunocompromised host with inflammatory enteritis is summarized in Box 61-1.

overlying exudate and necrosis; the most common finding is mild patchy colitis. If CMV is suspected, biopsy should be performed even if the mucosa appears normal grossly. Pp65 antigen detection may aid diagnosis.59 Complications of CMV colitis include intestinal perforation, severe hemorrhage, peritonitis, small-bowel obstruction, and toxic megacolon. Ganciclovir therapy has been shown to be beneficial for CMV colitis when initiated early in the disease.57,59 Ganciclovir induction therapy should be administered for 3 to 6 weeks, depending on the degree of involvement and immunologic state of the patient; maintenance therapy may be required thereafter.60 Cryptosporidial infection is one of the most common causes of enteric disease in patients with AIDS, occurring in 10% to 20% of adults with AIDS in the United States.61,62 It may be less common in children. Typical AIDS-related diarrheal syndromes consist of profuse, watery, and often bloody diarrhea associated with anorexia and weight loss; death is common. Mycobacterium avium complex infections are common in patients with AIDS and low CD4 T-lymphocyte counts.63 Signs and symptoms are usually nonspecific and include fever, night sweats, malaise, and diarrhea. The small intestine appears to be more commonly involved than the colon. Mucosal changes include erythema, edema, friability, and, occasionally, small erosions and fine white nodules. Local or systemic fungal infections can cause chronic inflammatory enteritis.64,65 Fungal pathogens include Candida spp., which can cause chronic, bloody, or nonbloody diarrhea; Paracoccidioides brasiliensis (South American blastomycosis), which causes a granulomatous or ulcerative lesion in the gastrointestinal tract; Histoplasma capsulatum, which can manifest with ulceration, bleeding, or obstruction; and the phycomycoses, which can cause abdominal pain, diarrhea, gastrointestinal tract bleeding, and peritonitis.

MANAGEMENT Specific Agents Gastrointestinal tract disease due to CMV occurs almost exclusively in the setting of immunodeficiency, including AIDS, organ transplantation, cancer chemotherapy, and corticosteroid therapy.57–59 Colitis, the most common manifestation of enteric CMV disease, is characterized by fever, diarrhea, abdominal pain, and hematochezia. Abdominal radiographs can reveal pneumatosis intestinalis or free peritoneal air.50 Endoscopic findings in patients with CMV-associated inflammatory colitis range from localized hyperemia to deep ulceration with

BOX 61-1. Diagnostic Evaluation of Immunocompromised Patients with Inflammatory Enteritis l. Routine stool culture for Shigella, Salmonella, and Campylobacter jejuni/coli 2. Specialized stool culture for Mycobacterium avium complex, other Campylobacter species, Yersinia enterocolitica, Escherichia coli 3. Assay for Clostridium difficile toxin if patient is receiving antibiotics 4. Microscopic stool examination: a. Fecal leukocytes/fecal lactoferrin test for assessment of invasive bacterial etiology b. Ova, cysts, and parasites such as Giardia lamblia, Strongyloides, Entamoeba histolytica (see Table 61-1) c. Acid-fast stain for Mycobacterium, Cryptosporidium, Cyclospora, Isospora 5. Blood culture for bacteria, mycobacteria, cytomegalovirus 6. If symptoms persist, if results of preceding workup are negative, and/or if patient is severely ill: a. Obtain abdominal radiographs to detect pneumatosis intestinalis (present in severe cytomegalovirus colitis) b. Consider esophagogastroduodenoscopy or colonoscopy with intestinal biopsy for histology, culture, special stains, molecular diagnosis

Fluid loss is the most important consequence of acute gastroenteritis. The primary aims of management are aggressive rehydration and replacement of electrolytes by oral or intravenous routes, maintenance of blood volume and acid–base balance, and replacement of ongoing diarrheal losses. Nutritional support is also important, especially in immunocompromised and malnourished patients, who are prone to persistent illness. Early feeding minimizes caloric deficit and stimulates repair of the intestinal brush border.66 Breastfeeding of infants should be continued, especially in developing countries, adding the benefit of passively acquired protective factors. In severe persistent cases of inflammatory enteritis in which absorption is compromised, parenteral nutrition is considered. A minority of patients with diarrhea can be treated successfully with antimicrobial agents28; therapy is reserved for the small numbers of patients with suspected or proven bacterial causes of diarrhea in whom benefit is likely. Empiric antibiotic therapy is appropriate in the following patients with inflammatory enteritis: (1) the severely ill patient in whom there is a strong suspicion of bacterial dysentery (such as the patient with high fever and blood and mucus in stools); (2) immunocompromised persons, including patients with AIDS and recipients of immunosuppressive drugs; (3) infants younger than 3 months with Salmonella infection (who are at high risk of bloodstream infection (BSI)); (4) patients with hemoglobinopathy or asplenia; and (5) people with a high likelihood of a treatable cause according to history, physical findings, or epidemiologic factors. Antibiotics are contraindicated in patients in whom Shiga toxin-producing Escherichia coli infection is suspected (typically presenting with afebrile bloody diarrhea), because such therapy may increase the risk of hemolytic– uremic syndrome,67 although this is controversial.68 Antibiotics hasten clinical improvement in persons with shigellosis and severe, early Campylobacter enteritis if given early in the disease.68 Therapy also shortens duration of fecal shedding of these organisms and of Y. enterocolitica. In developed countries, trimethoprinsulfamethoxazole has long been effective empiric treatment for suspected shigellosis in children. However, increasing antibiotic resistance

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Necrotizing Enterocolitis

renders this agent ineffective in many cases. Antibiotic susceptibility testing should be performed on patient isolates, and clinicians should be aware of susceptibility patterns in their area. Cefixime has in vitro activity against most strains of Shigella that are resistant to ampicillin or trimethoprim-sulfamethoxazole and is safe for use in children.69 One dose of ciprofloxacin is usually adequate for treatment of all Shigella species except Shigella dysenteriae 1; however, fluoroquinolones are approved by the United States Food and Drug Administration for use only in people 18 years of age and older. A Scientific Working Group of the World Health Organization has raised the possibility that the efficacy of fluoroquinolones for shigellosis may outweigh the small risk of adverse effects in selected children.70 The role of antibiotics in treating persons with specific causes of bacterial gastroenteritis is summarized in Table 61-6. Nonspecific therapies other than antimicrobial agents are available for symptomatic treatment of patients with diarrhea. Bismuth subsalicylate has some effect in decreasing stool volume and illness in travelers’ diarrhea.71 One study showed a decrease in frequency and volume of diarrhea in children.72 Antiperistaltic agents (loperamide and diphenoxylate hydrochloride) and antiemetic agents can be effective in the management of diarrhea in adults; however, their use is contraindicated in small children and infants as safety and efficacy data are lacking.73 Antiperistaltic agents have been demonstrated to worsen the clinical course of disease due to Shigella spp., Clostridium difficile, and Escherichia coli O157:H7.

CHAPTER

62

393

The value of antimicrobial therapy for Salmonella gastroenteritis without BSI has not been established. There is no evidence of clinical benefit of antimicrobial therapy in otherwise healthy children and adults with nonsevere Salmonella diarrhea.

CHAPTER

62

Necrotizing Enterocolitis C. Michael Cotten and Daniel K. Benjamin, Jr

Necrotizing enterocolitis (NEC) is the most common neonatal gastrointestinal emergency. NEC results in substantial morbidity and mortality, particularly in premature infants. Less commonly, NEC is observed in older children. Bacterial translocation and bacterial overgrowth occur in settings of underlying gastrointestinal pathology.

NECROTIZING ENTEROCOLITIS IN NEONATES Infants with Inflammatory Enteritis

Epidemiology

An infant younger than 3 months of age with inflammatory enteritis should be managed cautiously, because of the propensity of Salmonella strains to cause BSI with disseminated suppurative foci in this age group.74,75 Although the precise frequency of BSI is not known, a prospective study of children with gastroenteritis due to Salmonella spp. reported BSI in 6 of 91 (2 of the 6 infants were 3 months of age or younger).74 Infants who appear toxic with diarrhea as the prominent feature of illness should be hospitalized, cultures of blood and cerebrospinal fluid should be obtained, and intravenous antibiotics with a spectrum of activity against enteric pathogens, as well as common invasive pathogens, should be given. In the infant 3 months of age or younger who does not appear toxic, a consensus panel has recommended obtaining blood and stool cultures if the stool contains blood or leukocytes.75 Ampicillin, cefotaxime, and ceftriaxone are effective for treatment of invasive focal Salmonella infections (see Chapter 146, Salmonella Species).76 If stools contain blood or leukocytes, antimicrobial therapy may be initiated pending culture results.

NEC affects predominantly premature infants in neonatal intensive care units (NICUs). The overall incidence of NEC is 1 to 3 cases per 1000 live births, and the incidence among very-low-birthweight (VLBW, < 1500 grams at birth) infants overall is between 5 and 10%. However, the incidence is variable by site; different tertiary centers report a range from 0.3% to 22%. Besides association with center, the incidence of NEC increases with lower gestational age and birthweight. NEC does not have seasonal, geographic, or sex predilection. Among VLBW infants, an increased incidence has been reported in black male infants relative to nonblack males.1 The age of onset of NEC is inversely related to birthweight and gestational age.1,2 In extremely preterm infants, NEC has been reported to occur as late as 10 weeks of age. In VLBW infants, risk factors are not different between those affected early and those affected later. The overall mortality of VLBW premature infants has remained steady, and is approximately 50% for infants with NEC requiring surgery.1,3,4 NEC is also associated with prolonged hospital stay3,5 and increased risk of neurodevelopmental impairment in survivors.6 Occasionally NEC affects full-term infants. Full-term infants are often affected in their first week of life, and associated factors (Table 62-1) suggest that mucosal injury and ischemia are important.7–9 Isolated gastrointestinal perforation can also occur, especially in the first few postnatal days of extremely-low-birthweight (ELBW, < 1000 grams at birth) infants; the usual site is the distal ileum. It has been associated with the combination of treatment with corticosteroids and indomethacin, and likely is a separate clinical entity from NEC.10,11

TABLE 61-6. Use of Antimicrobial Therapy in Inflammatory Enteritis Antimicrobial Therapy

Enteropathogen

Effective clinically

Shigella, enteroinvasive Escherichia coli

Decreased excretion

Shigella, enteroinvasive Escherichia coli, Yersinia enterocolitica, Campylobacter jejuni

Indicated in certain hostsa Salmonella, Campylobacter, enteropathogenic Escherichia coli, enterotoxigenic Escherichia coli, Clostridium difficile, Yersinia enterocolitica Unclear effect Controversial a

b

Enteroaggregative Escherichia coli, Aeromonas, noncholera Vibrio spp. Enterohemorrhagic Escherichia coli

Antimicrobial agents warranted for patients with severe colitis or with increased risk of disseminated disease: infants < 3 months of age, persons with immunodeficiency disorders (including acquired immunodeficiency syndrome, hemoglobinopathy, and asplenia) as well as recipients of immunosuppressive drugs. b See text.

Pathology and Pathogenesis The most common sites of NEC in preterm infants are the ileum and proximal colon, although lesions have been noted throughout the intestine.12 The earliest pathologic lesions are superficial mucosal ulceration and submucosal hemorrhage and edema. The trivial infiltration of acute inflammatory cells seen initially speaks against a primary infectious process. It may suggest suboptimal barrier function of the immature intestinal mucosal barrier that allows passage of bacteria and undigested food antigens.13 In severe cases, the mucosal

394

SECTION

H Gastrointestinal Tract Infections and Intoxications

TABLE 62-1. Factors and Conditions Associated with Necrotizing Enterocolitis Premature

Full-Term

Lower gestational age Feeding

Cyanotic congenital heart disease Polycythemia Exchange transfusions Perinatal asphyxia Small for gestational age/intrauterine growth restriction (IUGR) Umbilical catheters Maternal pre-eclampsia Antenatal cocaine Premature rupture of membranes Milk allergy Gestational diabetes

inflammatory process progresses to coagulative transmural necrosis, with subsequent perforation. Hepatobiliary gas or intramural gas (pneumatosis intestinalis) is pathognomonic for the disease (Figure 62-1). It is seen as bubbles or gaseous strips in the submucosa and subserosa. If the disease does not progress to complete transmural necrosis and perforation, healing can occur by epithelialization and proliferation of fibroblasts, and granulation tissue is laid down. Strictures can follow.14,15 The pathogenesis of the disease is multifactorial, involving three factors: substrate (usually formula feeding into the bowel lumen), mucosal injury, and presence of bacteria.16 Theoretical, experimental, and clinical data support the importance of all three factors in precipitating disease. The intestine of the premature infant intestine is especially sensitive to all three (Box 62-1);12 however, no compelling evidence suggests a single cause of NEC.12,13,17 NEC is more likely the “final common pathway” of multiple combinations and interactions of immaturity, injury, bacteria, and substrate. Experimental models of NEC incorporate ischemia, microbial toxins, metabolic byproducts, milk components, abnormal bowel motility, abnormal vascular autoregulation, oxidative injury induced by ischemia/reperfusion, blocking of nitric oxide synthesis, and the central action of platelet-activating factor (PAF). Epidemiologic observations support an important but probably not a singular role for microbes, and clustered, epidemic cases have been observed in NICUs.18,19 No single pathogen has been consistently associated with NEC in infants, although gram-negative enteric species, and occasionally Pseudomonas species, predominate.1,20 Bacterial toxins (e.g., hemolysins of coagulase-negative staphylococci and Staphylococcus aureus, enterotoxins of Escherichia coli, and toxins of clostridia) have been associated with NEC. Most bacteria identified from the stool, blood, or peritoneal fluid of infants with NEC also colonize the gastrointestinal tract of asymptomatic infants in NICUs.21 Infectious agents isolated from infants affected with endemic NEC are similar to those associated with epidemic NEC. Viruses (e.g., rotavirus) have been identified in ill infants during outbreaks of NEC, but case-control studies have demonstrated similar rates of identification in well infants. Outbreaks of NEC have been associated with illness in caregivers, reinforcing the need for careful infection control practices.

Clinical Manifestations and Laboratory Findings The clinical course of NEC can vary from a slow indolent process to a fulminant one with progression to death in a few hours. The classic symptoms include abdominal distension and blood in the stool, and many patients have less specific signs indicative of generalized septicemia (Box 62-2). Hematologic abnormalities, including thrombocytopenia and low or high white blood cell counts, are common. Falling leukocyte and platelet counts combined with worsening metabolic acidosis suggest worsening disease.12

Figure 62-1. Abdominal radiograph of an infant with necrotizing enterocolitis and pneumatosis intestinalis. (From Willoughby RE Jr, Pickering LK. Necrotizing enterocolitis and infection. Clin Perinatol 1994;21:307–315.)

BOX 62-1. Limitations of the Preterm Infant’s Intestine that may Contribute to Risk of Necrotizing Enterocolitis IMMUNOLOGIC FACTORS Decreased secretory immunoglobulin A Decreased intestinal lymphocytes Poor antibody response LUMINAL FACTORS Lower H+ output in stomach Lower proteolytic enzyme activity IMMATURE INTESTINAL BARRIER Less protective mucin barrier Less protective microvillous membrane Higher permeability LOWER/LESS ORGANIZED MOTILITY

BOX 62-2. Initial Signs of Necrotizing Enterocolitis Feeding intolerance (decreased oral intake, vomiting, increased gastric residual fluid) Lethargy Temperature instability Apnea Metabolic acidosis Shock Disseminated intravascular coagulation Ileus Clinitest-positive stools Heme-positive stools or heme-positive vomitus

The differential diagnosis includes medical conditions associated with decreased intestinal motility, such as bacterial or viral infection, metabolic disorders including hypothyroidism, and congenital heart disease. The diagnosis of NEC is suspected when the gastrointestinal signs and symptoms predominate. Infectious enterocolitis, smallbowel volvulus, isolated gastrointestinal perforation, colonic aganglionosis with enterocolitis, and other surgical diseases of the intestinal tract should be considered. Confirmation of the diagnosis of NEC depends on radiographic findings (see Figure 62-1). Pneumatosis intestinalis can be subtle, but it is pathognomonic of NEC. Intestinal distension with multiple thickened and dilated loops of small bowel is commonly seen, but, unlike pneumatosis, this finding is nonspecific. Pneumoperitoneum (best assessed on a cross-table or left lateral decubitus position) or presence

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Necrotizing Enterocolitis

of intra-abdominal fluid is a serious sign indicating a need for immediate surgical intervention. Although increased exhaled breath hydrogen, serum levels of interleukin-6, endotoxin, a1-antitrypsin, stool-reducing substances, and a specific short-chain fatty acid profile have been suggested to be early laboratory findings associated with NEC, none has been introduced into clinical practice. Diagnostic criteria for NEC were originally developed by Bell and Conorkers. Modified in 1986, criteria provide uniformity for therapeutic decisionmaking22 (Table 62-2). Cultures of blood and peritoneal fluid are positive in approximately 30% of preterm infants with advanced NEC. Isolates usually reflect the intestinal bacterial flora – Enterococcus, Enterobacter, and Klebsiella species are common. Coagulase-negative staphylococci or Candida species are isolated from peritoneal cultures in < 20% of cases of NEC.1,23 In a study of 96 cases of neonatal culture-positive peritonitis, Enterobacteriaceae were recovered from 75% of cases associated with NEC, whereas Candida species and coagulasenegative staphylococci were found in 44% and 50% of cases associated with focal intestinal perforation, respectively.23

Management Management is largely supportive and directed at preventing progression of damage, restoring homeostasis, and minimizing complications. Oral feedings should be withheld, and a large-bore nasogastric tube placed; intermittent suction is then applied. Intravenous access is required to provide fluids, electrolytes, nutrition, and antimicrobial therapy. The duration of restriction of enteral feeding depends on radiographic findings of NEC, clinical status, and local practice. Many patients whose symptoms resolve with these measures and whose NEC does not progress beyond stage I disease (Table 62-2) are fed after 48 to 72 hours. If the period of enteral restriction required is longer (e.g., > 7 days, which is common if the patient reaches stage II), consideration should be given to providing total parenteral nutrition. Reinstitution of feeding after the resolution of NEC should be undertaken cautiously. It is common to feed low volumes of elemental formula or mother’s milk initially to allow optimal absorption of all nutrients, to avoid osmolar challenge and

CHAPTER

62

395

further potential injury to the intestinal mucosa, and to enhance repair. However, few randomized clinical trials have been performed to guide substance or timing of initiation of feeding.24 Because microorganisms have been implicated in the pathogenesis of NEC, can invade because of NEC, or because the clinical picture is frequently one of septicemia, cultures of blood, stool, urine, and cerebrospinal fluid are obtained, and broad-spectrum antibiotics are given parenterally. Recommendations for initial antimicrobial therapy vary among experts, from initial use of ampicillin and gentamicin, with addition of an agent (e.g., clindamycin or metronidazole) with activity against Bacteroides species when perforation is diagnosed, and vancomycin empirically when resistant Staphylococcus species are suspected;12,25 to recommendations for immediate empiric use of vancomycin plus broad-spectrum gram-negative bacillary coverage.26 Empiric use of an agent with activity against anaerobic bacteria is advocated by some.27 Routine inclusion of antimicrobial agents active against anaerobic bacteria does not seem to be indicated in NEC without perforation, but there are few well-controlled studies; clindamycin use was associated with increased strictures following NEC in one small randomized study.28 The empiric use of vancomycin for possible coagulase-negative staphylococci must be balanced against the risk of selecting vancomycin-resistant enterococci. Reinforcement of strict handwashing is essential; institution of infection control measures such as barrier isolation and cohorting of affected infants and personnel may abort epidemics. Abdominal radiography is used to assess the progression of NEC and to identify perforation or other indications for surgery. Radiographic studies are typically repeated every 6 hours during escalation of signs. Progressive neutropenia or thrombocytopenia may be a useful sign. Fluid and electrolyte status is strictly monitored because large third-space losses can occur with inflammation and edema of the intestine. Severe metabolic acidosis can result from poor perfusion and is often difficult to control. Surgery is indicated for intestinal perforation or when response to medical management is not observed. Indications for surgery include pneumoperitoneum or cellulitis of the anterior abdominal wall – signs of perforation or a gangrenous bowel. The classic operative strategy has been relatively conservative, with resection of grossly gangrenous bowel and creation of a proximal enterostomy. The less invasive approach of drain placement rather

TABLE 62-2. Modified Bell Staging for Necrotizing Enterocolitis Stage

Classification

Systemic Signs

Intestinal Signs

Radiograph

IA

Suspected NEC

Temperature instability, apnea, bradycardia, lethargy

Increased pregavage residuals, midabdominal distension, emesis, heme-positive stool

Normal or intestinal dilatation, ileus

IB

Suspected NEC

Same as above

Bright-red blood from rectum

Same as above

IIA

Proven NEC – mildly ill

Same as above

Same as above, plus absent bowel sounds, with or without abdominal wall tenderness

Intestinal dilation, ileus, pneumatosis intestinalis

IIB

Proven NEC – moderately ill

Same as above, plus mild metabolic acidosis and mild thrombocytopenia

Same as above plus absent bowel sounds, definite tenderness, with or without abdominal cellulitis or right lower-quadrant mass

Same as stage IIA plus definite ascites

IIIA

Advanced NEC – severely ill, bowel intact

Same as stage IIB, plus hypotension, bradycardia, severe apnea, combined respiratory and metabolic acidosis, disseminated intravascular coagulation, and neutropenia

Same as above plus signs of generalized Same as stage IIA plus definite peritonitis, marked tenderness, and ascites distension of the abdomen

IIIB

Advanced NEC – severely ill, bowel perforated

Same as stage IIIA

Same as stage IIIA

NEC, necrotizing enterocolitis.

Same as stage IIB plus pneumoperitoneum

396

SECTION

H Gastrointestinal Tract Infections and Intoxications

than performing extensive laparascopic surgery is also used. A prospective observational trial of drains versus laparotomy, and a randomized trial comparing the two strategies have been completed.4,28a Neither strategy appears to provide clear benefit over the other. We await future comparative trials that will include long term neurodevelopmental follow up in the outcome analysis to help define best practice.

studied in animals and humans. These lipids seem to have antiinflammatory effects, with a reduction in the levels of PAF and leukotrienes in the intestinal mucosa and a significant decrease in the incidence of stage III NEC in infants born at less than 32 weeks of gestation. A meta-analysis of controlled studies has suggested a reduction in the incidence of NEC when antenatal corticosteroids were used,51 and evidence in animal models has indicated that corticosteroid administration is associated with maturation of the intestinal barrier, limitation of bacterial colonization, and reduced inflammation.12

Outcome Approximately 50% of neonates with NEC requiring surgery die.1,3 Survival among preterm infants with NEC not requiring surgery is not markedly different from those without NEC. Morbidity related to the long-term effects of the disease on the gastrointestinal tract is seen in 10% to 30% of affected children, and is especially frequent among those with NEC requiring surgery. Acute and chronic complications related to the gastrointestinal tract are listed in Box 62-3.

Prevention Strategies to reduce the incidence of NEC have targeted prevention of mucosal injury, and modulation of bacterial colonization. The authors of one small single-center study suggest that L-arginine supplementation may reduce the risk of NEC by increasing local intestinal nitric oxide synthesis.29 However, additional appropriately powered studies must be conducted to affirm this conclusion. Recombinant PAF acetylhydrolase, the enzyme found in human milk that degrades PAF, has been proposed as another preventive agent based upon preliminary animal studies.30 In a double-blind placebo-controlled study in septic neonates, the immunomodulating agent pentoxifylline was associated with reduced incidence of NEC.31 In a large multicenter randomized trial of diet-supplemented glutamine (an important micronutrient thought to mitigate stress in the intestine), no incidence of infection or death (the primary outcome variables) or NEC was found.32 Multiple strategies of alteration of intestinal flora have been studied. Prophylactic use of antibiotics has been evaluated in several small studies, with conflicting results.33 Supplementing formula with an immunoglobulin A (IgA)-IgG preparation in low-birthweight infants was shown to prevent NEC in one study,2 but this result has not been replicated.34,35 Lowering intestinal pH may also reduce the risk of NEC.36,37 The use of probiotics to prevent NEC has been evaluated. Reduced incidence of NEC has been observed in pilot studies using probiotic combinations, especially those with Lactobacillus and Bifidobacterium species.38–41 Commensal bacteria appear to stimulate healthy gastrointestinal epithelial cell turnover, provide baseline immune stimulation, and may prevent overgrowth of potentially pathogenic bacteria.42–45 Data from cohort studies suggest that human milk may reduce the risk of NEC.46–48 There are also small randomized trials comparing mother’s milk, donor milk, and preterm formula,49 but the studies suggesting benefit of human milk have not been definitive.50 The biologic basis for protective effect of human milk is strong, and includes the presence of substances that modify bacterial colonization; immune and growth factors, including lactoferrin; PAF acetyl-hydrolase and the coincident intestinal colonization with commensal bacteria.47 Omega-3 fatty acid- and egg phospholipid-containing diets have been

BOX 62-3. Complications of Necrotizing Enterocolitis Intestinal or colonic strictures Enterocolonic fistulas Fluid and electrolyte imbalance Malabsorption Cholestasis Anastomotic leaks and stenosis after surgery Short-gut syndrome after surgery

NECROTIZING ENTEROCOLITIS IN OLDER CHILDREN Pneumatosis intestinalis, a cardinal radiographic feature of NEC, rarely occurs in older children. Reports of NEC in older children are case reports or small single-center case series. NEC can complicate the course of infants after open heart surgery.52 Children with embolic, thrombotic, or vasculitic diseases can have manifestations secondary to mesenteric ischemia. Unrecognized Hirschsprung disease can manifest as enterocolitis in infancy or early childhood, with profound illness. Gangrene of the bowel is usually the result of intestinal volvulus or surgical adhesions, and bacterial invasion with gas production in the bowel wall or portal vein occurs when the gangrene is extensive. Typhlitis is a special form of enterocolitis that occurs in children with neutropenia.53 Typhlitis is characterized by ulceration and bacterial invasion of the cecum; monocytes are found in the inflammatory infiltrate. Pneumatosis intestinalis, typically involving the colon, also occurs in children with acquired immunodeficiency syndrome or those treated with high doses of corticosteroids.54,55 Its pathogenesis is unknown. Air can dissect into the bowel wall or peritoneum, without enterocolitis, during endoscopy or surgery or in patients with cystic fibrosis and other obstructive lung diseases. In developing countries, severe protein-calorie malnutrition can predispose a child to bowel wall pneumatosis after an episode of gastroenteritis.56

Clinical Manifestations and Management NEC in an older child often causes vomiting, abdominal pain, distension, and diarrhea. The diarrhea can become bloody or subside, and ileus is a prominent late feature. Elevations in serum bilirubin and hepatic enzymes suggest increasing bacterial translocation from necrotic bowel. Pneumatosis intestinalis and portal vein air are late findings. Children with neutropenic colitis have right lower-quadrant findings that simulate appendicitis, often with diarrhea. By contrast, pneumatosis intestinalis associated with corticosteroid use may be detected incidentally, or patients can have abdominal pain and diarrhea; ileus is uncommon. Pneumatosis intestinalis with ileus in older children is ominous, with a mortality rate of greater than 20%, with transplant and congenital heart disease patients having the worst prognosis.57 Blood and peritoneal cultures may be positive for Streptococcus species and Enterobacteriaceae. Empiric antimicrobial therapy should include agents active against these bacteria and anaerobic organisms. Children with typhlitis frequently respond to medical management; operative intervention can be complicated by dehiscence of anastomoses or wounds. Corticosteroid-induced pneumatosis resolves when the drug is discontinued.

BACTERIAL TRANSLOCATION Patients sustaining severe trauma or burns most commonly die of septicemia and multiple organ failure. Studies in hypotensive patients after trauma have demonstrated that 56% have positive blood cultures within 3 hours of injury.58 Bacterial translocation has been documented in animals and proposed as the mechanism of infection.59 It has not been proved to occur in patients after trauma, but it has been

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Foodborne and Waterborne Disease

CHAPTER

63

397

demonstrated (by mesenteric lymph node biopsy) in patients with Crohn disease, intestinal malignancies, or intestinal obstruction. Bacterial translocation may also explain the increased incidence of catheter infection in children with short-gut syndrome (usually caused by intestinal bacteria) relative to other groups.60 Increased bacterial translocation also occurs with malnutrition, failure to use the enteral route for feeding (especially in burned patients), bacterial overgrowth syndrome, and in patients receiving peritoneal dialysis.61 When translocation occurs in healthy individuals, bacteria are ingested by enterocytes, translocated into the submucosa, and transported by lymphatics to the mesenteric lymph nodes. Escherichia coli, Klebsiella pneumoniae, and Enterococcus species are commonly isolated, although intracellular organisms such as Salmonella or Listeria monocytogenes translocate and survive most readily. The anaerobic flora protects against translocation by providing “colonization resistance” in healthy individuals.62 Translocation imprecisely describes the passage of bacteria across damaged intestinal epithelium. Bacteria replicating in compromised tissues may overwhelm local host defenses and disseminate beyond the mesenteric lymph nodes, with resultant bacteremia, hepatic abscesses, seeding of implanted or indwelling devices or distant organ sites. Early institution of enteral feeding in burned patients and aggressive management of other predisposing conditions are the only current modes of reduction of translocation.

to voluminous stool loss. Micronutrient deficiencies include vitamin B12, and fat-soluble vitamins A, D, E, and K. Conditions that mimic bacterial overgrowth include gluten-sensitive enteropathy, postinfectious enteropathy, and giardiasis.

SHORT-GUT SYNDROME AND BACTERIAL OVERGROWTH

ACKNOWLEDGMENT

Short-gut syndrome, often complicated by bacterial overgrowth, is seen in surgically treated infants with NEC when large segments of intestine have been resected. To avoid this complication, a conservative approach to surgery is attempted in infants with a large amount of the small bowel affected. A proximal enterostomy is placed without resection acutely, with re-exploration days later to assess damage if clinical conditions do not improve, or weeks to months later for reanastomosis and to assess damage. Serious nutritional consequences usually result if more than 70% of the intestine is removed. Preservation of the terminal ileum and the ileocecal valve has primary importance for long-term nutrition and function, but the predilection of NEC to affect this area sometimes makes preservation difficult. The use of total parenteral nutrition until adaptation of the intestine occurs is a prime therapeutic maneuver for short-gut syndrome. Bacterial overgrowth is a common consequence of short-gut syndrome. Gastric hypersecretion is characteristic in short-gut syndrome. Treatment with antacids and histamine antagonists leads to increased colonization by oral flora in the stomach and upper intestine. Bacterial overgrowth may be exacerbated when resection of the ileocecal valve eliminates the distal barrier to colonic bacteria, which then reflux into the small intestine. Overgrowth of bacteria can result in decreased production of mucosal hydrolases, causing carbohydrate malabsorption and osmotic diarrhea. Deconjugation of bile salts in the small intestine by bacteria leads to malabsorption of fat and fat-soluble vitamins. Bacterial overgrowth can also decrease intestinal transit time.

The authors of this chapter and the editors acknowledge the contributions of Esther Jacobowitz Israel to the previous edition of this chapter, portions of which remain unchanged, particularly those on NEC in older children.

More than 200 known diseases are transmitted through food.1 These diseases result from ingestion of food contaminated with viruses, bacteria, parasites, microbial or chemical toxins, metals, and prions. Symptoms of foodborne illness range from mild gastroenteritis to lifethreatening sepsis, neurologic, hepatic, and renal syndromes. Despite improvements in sanitation, foodborne disease remains a major public health problem in all economically developed as well as developing countries.

Clinical Manifestations

EPIDEMIOLOGY

Short-gut syndrome is primarily characterized by diarrhea. The diarrhea can be due to inappropriate food intake, hypersecretion of gastric acid, rapid intestinal motility, malabsorption, intercurrent infection, or bacterial overgrowth. The role of bacterial overgrowth in the pathogenesis of persistent diarrhea and malnutrition in the developing world has been questioned.63 Even in the developed world, bacterial overgrowth has been implicated as a factor in chronic diarrhea. Clinical manifestations associated with overgrowth include diarrhea, abdominal pain with bloating, malnutrition, and micronutrient deficiencies. The diarrhea is frequent and may be loose and watery, and steatorrhea can contribute

The epidemiology of foodborne disease is changing rapidly. Largescale industrialized production of minimally processed food products and rapid national and international distribution of food have been introduced without proper hygienic precautions. These processes have led to the emergence of new vehicles for recognized pathogens.2 Antimicrobial resistance has become increasingly prevalent among strains of Salmonella, Campylobacter, and Shigella due in part to widespread use of antimicrobial agents administered to feed animals. Postinfectious syndromes have been recognized as important consequences of foodborne infections, including hemolytic–uremic syndrome (HUS) after infections with Escherichia coli O157:H7,3

Management The diagnosis of bacterial overgrowth requires the isolation of > 105 colony-forming units/mL of bacteria from small intestinal fluid specimen; isolation of strictly anaerobic bacteria, such as the Bacteroides fragilis group, also establishes the diagnosis. Bacterial overgrowth can occur in the distal end of the small intestine, which is poorly accessible to intubation or endoscopy. The hydrogen breath test after a lactulose challenge is the best noninvasive marker of bacterial overgrowth. A fasting breath hydrogen screening test can also be helpful; values of hydrogen above 40 ppm are indicative of bacterial overgrowth. Empiric trial of antibiotic therapy is sometimes given to patients with chronic diarrhea or those with abdominal pain and bloating. No single regimen has proved effective for bacterial overgrowth. Metronidazole and oral gentamicin, however, have been used. Although some clinicians may perceive a potential benefit of use of probiotics after lengthy course of antibiotics in post-NEC infants with short gut, bacteremia with probiotic agents used in this situation has been reported.64

CHAPTER

63

Foodborne and Waterborne Disease Wences Arvelo, Andrew Pavia, and Christopher Braden

398

SECTION

H Gastrointestinal Tract Infections and Intoxications

reactive arthritis after salmonellosis,4 and Guillain–Barré syndrome (GBS) after campylobacteriosis.5 From 1998 to 2002, 6647 outbreaks affecting 127 734 people in the United States were reported to the Centers for Disease Control and Prevention (CDC).6 The annual number of foodborne disease in the United States was estimated to be approximately 76 million cases, with 325 000 hospitalizations and 5000 deaths in 1999.7 Data on age-specific incidence of foodborne diseases are not available, but the highest reported rates for several important foodborne enteric pathogens, including Salmonella spp., Campylobacter spp., E. coli O157:H7, and Shigella spp., occur in children younger than 5 years of age, suggesting that children are at higher risk for foodborne diseases than adults.8,9

PATHOGENESIS AND CLINICAL SYNDROMES Foodborne diseases are often classified according to the mechanism of pathogenesis: ingestion of various classes of chemicals or preformed toxins, in vivo production of bacterial toxins, direct pathogen attachment and local invasion, or disseminated infection. Consideration of foodborne disease according to its associated clinical syndrome is more useful for diagnosis. The cause of a foodborne illness is suggested by the clinical symptoms, the incubation period, and epidemiologic clues. Laboratory testing is needed to confirm the identity of specific pathogens. A foodborne outbreak should be suspected when two or more people who have shared a common food develop the same acute illness, most often characterized by nausea, vomiting, diarrhea, or neurologic symptoms.

Foodborne Disease Caused by Microbial Agents or their Toxins Nausea, Vomiting, or Diarrhea within 1 to 16 Hours Toxins caused by Staphylococcus aureus, Bacillus cereus, and Clostridium perfringens are the major causes of rapid-onset foodborne illness. A clinical syndrome with an incubation period from 1 to 6 hours and predominant vomiting is usually caused by S. aureus toxin or short-incubation B. cereus toxin. Illness after a slightly longer incubation period with predominant cramps and diarrhea suggests C. perfringens toxin or the long-incubation B. cereus toxin. The duration of symptoms is generally short. Comparative features are shown in Table 63-1. Staphylococcal food poisoning is characterized by vomiting (82% of cases) and diarrhea (68%); fever is relatively uncommon (16%).10 Illness due to S. aureus is caused by ingestion of preformed enterotoxins A, B, C, D, or E. Staphylococcal enterotoxins are produced when toxin-producing strains of S. aureus contaminate foods that support growth and are held at temperatures that permit bacterial growth and toxin production. Enterotoxin A is most commonly implicated in food poisoning in the United States and England.11 Staphylococcal enterotoxin F is associated with toxic shock syndrome but not with food poisoning. Toxins produce vomiting through stimulation of receptors in the abdominal viscera; the stimulus is carried to the emetic center in the brain via autonomic pathways. Diarrhea is less prominent but may occur as a result of fluid accumulation in the intestinal tract.

TABLE 63-1. Characteristics of Foodborne Illnesses Caused by Microbes and their Toxins Symptoms Organism

Median Hours Incubation (Range)

Emesis

Diarrhea

Fever

Staphylococcus aureus toxin

3 (1–6)

+++

++

0

Bacillus cereus toxin Emetic syndrome Diarrheal syndrome

2 (1–6) 9 (6–16)

+++ +

+/++ +++

0 0

C

2–12 hours 12–24 hours

Clostridium perfringens toxin

12 (6–24)

+

+++

0

C

12–24 hours

Vibrio cholerae non-O1

11 (5–96)

+

+++

+++

C, BD (25%)

1–12 days

Vibrio parahaemolyticus

15 (4–96)

++

+++

++

C, HA, BD (rare)

1–7 days

Norovirus

24 (12–48)

+++

+++

++

HA, My

1–2 days

Shigella spp.

24 (7–168)

+

+++

+++

C, BD

2–10 days

Listeria monocytogenes

31 (11–70)

+

+

+++

Meningitis, bacteremia

Unknown

Vibrio cholerae O1

48 (6–120)

++

+++

+

Dehydration

1–7 days

Salmonella spp.

36 (12–72)

+

+++

++

C, HA, My

3–10 days

Campylobacter jejuni

48 (24–168)

+

+++

+++

C, BD, MA

2–14 days

Shiga-toxin producing Escherichia coli (STEC) 96 (48–120)

++

+++

+

C, BD, HA, HUS

2–12 days

Enterotoxigenic Escherichia coli (ETEC)

(10–72)

+

+++

+

C

1–5 days

Enteropathogenic Escherichia coli (EPEC)

(9–12)

+

+++

+

C

Prolonged

Enteroinvasive Escherichia coli (EIEC)

(20–48)

+

+++

++

C, BD

Unknown

Enteroaggregative Escherichia coli (EAEC)

(20–48)

+

+++

+

C

Unknown

Yersinia enterocolitica

96 (48–240)

+

+++

+++

C, HA, pharyngitis, MA

3–20 days

Cyclospora cayetanensis

156 (24–300)

+

+++

+

Anorexia, weight loss, fatigue

1–5 weeks

Other

Duration 12–24 hours

BD, bloody diarrhea; C, abdominal cramps; HA, headache; HUS, hemolytic–uremic syndrome; MA, mesenteric adenitis; My, myalgias. 0, rare ( 105 cfu/g in food; demonstration of enterotoxin

Bacillus cereus

Isolation from stool (patients + controls)

Isolation of > 105 cfu/g in food

Clostridium perfringens

Isolation of > 106 cfu/g stool

Isolation of > 105 cfu/g in food

Vibrio parahaemolyticus, Vibrio cholerae O1

Stool culture (TCBS agar); serology

Culture (TCBS agar)

Shigella species

Stool culture, serotyping; stool culture of food handlers

Culture

Salmonella species

Stool culture, serotyping

Culture

Campylobacter jejuni

Stool culture by selective media (e.g., Campy-BAP, Skirrow, Butzler)

Culture (selective media)

Shiga toxin-producing Escherichia coli

Stool culture (SMAC agar), serotyping; toxin-testing by EIA; serology

Culture (selective media), toxin testing

Yersinia enterocolitica

Stool culture (CIN agar); serotyping

Culture (cold enrichment)

Listeria monocytogenes

Blood culture; stool culture (cold enrichment)

Culture (cold enrichment)

Cyclospora cayatanensis

Stool examination (acid-fast stain, autofluorescence)

Not established

Norovirus and other enteric viruses

Reverse transcriptase PCR of stool

Reverse-transcriptase PCR

Histamine fish poisoning

Demonstration of 100 mg histamine per 100 g of fish

Ciguatera

Demonstration of ciguatoxin by bioassay or EIA

PSP and NSP

Demonstration of toxin by bioassay

BAP, Brucella agar-amphotericin B-polymyxin; cfu, colony-forming unit; CIN, cefsulidin-Irgasin-novobiocin; EIA, enzyme immunoassay; NSP, neurotoxic shellfish poisoning; PCR, polymerase chain reaction; PSP, paralytic shellfish poisoning; SMAC, sorbitol MacConkey; TCBS, thiosulfate-citrate-bile salt-sucrose.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Foodborne and Waterborne Disease

subtyping at the state public health laboratories has become a critical component of disease surveillance, and identifies outbreaks that would otherwise not be detected. Campylobacter spp., Y. enterocolitica, Vibrio spp., and E. coli O157:H7 require use of selective media for optimal recovery (see Table 63-7). E. coli O157:H7 can be identified presumptively by screening sorbitol-negative colonies on sorbitol MacConkey (SMAC) media, an inexpensive screening procedure that should be implemented in all laboratories. Detection of other STEC requires identification of Shiga toxins or detection of toxin genes; a commercial enzyme immunoassay is widely available, and polymerase chain reaction (PCR) test is available in some reference and state or federal public health laboratories. Enterotoxigenic E. coli infections are confirmed by identifying the production of heat-stable (ST) or heat-labile (LT) toxin by E. coli isolated from stool using a commercial latex agglutination assay or by PCR. All Salmonella spp., Shigella spp., STEC, V. parahemolyticus, and Listeria monocytogenes isolates should be forwarded to the state public health laboratory for full characterization and analysis for surveillance and outbreak detection purposes. Botulism outbreaks can be confirmed by demonstration of botulinum toxin in serum or stool of ill people or in incriminated food by use of the mouse neutralization test.127 Identification of norovirus has been enhanced by availability of PCR-based assays available in many public health laboratories. Stool samples to be tested for virus detection should be kept refrigerated but not frozen. Acid-fast or specific monoclonal antibody stains of stool for Cryptosporidium should also be obtained in cases of suspected waterborne disease. Outbreaks caused by heavy metals, chemicals, histamine fish poisoning, ciguatera and shellfish poisoning may be documented by demonstration of the offending toxin in the incriminated food. If chemical food poisoning is suspected, a urine specimen may yield evidence of the toxin. Additional information on collection of specimens for diagnosis and investigation of foodborne outbreaks is available at http://www.cdc.gov.foodborneoutbreaks/guide_sc.htm.

MANAGEMENT Supportive care is central to the management of acute foodborne and waterborne disease. Depending on the syndrome, some specific measures may also be appropriate. Careful replacement of fluid and electrolyte losses is crucial, particularly in infants. In the absence of severe vomiting, oral replacement with appropriate electrolyte solutions is the preferred treatment.128 Antiemetics may be useful for patients with severe or prolonged vomiting. Antiperistaltic agents may provide some symptomatic relief, but should be avoided in any patient with signs of invasive disease such as high fever, bloody diarrhea, and fecal leukocytes and generally are not recommended for children < 2 years of age.128 Specific antimicrobial therapy is indicated for certain confirmed infections, notably shigellosis, enterotoxigenic and enteroinvasive E. coli, giardiasis, typhoid fever, cyclosporiasis, and cholera. Specifics of therapy are discussed in the pathogen-specific chapters.129–131 Recent outbreaks of salmonellosis and shigellosis have shown the emergence of multidrug-resistant bacterial strains. Antimicrobial therapy for these patients should be tailored to the specific susceptibility profile of the pathogen. The role of antimicrobial therapy is less clear for treatment of diarrheal illness due to Y. enterocolitica, Campylobacter, V. parahaemolyticus, Aeromonas, and Plesiomonas, but therapy may have a role in severe or prolonged gastrointestinal tract illness caused by these pathogens. Antimicrobial agents are not usually indicated for intestinal salmonellosis, except in very young children, persons with an immunodeficiency, or in cases of invasive salmonellosis. The association between antimicrobial agents and increased risk of HUS is not clear.132–134 Therefore, antimicrobial agents are not recommended for children with E. coli O157:H7 and other STEC infections. Vigorous hydration may reduce the risk of HUS. Medical care providers who suspect botulism should immediately contact their state health department’s 24-hour emergency telephone

CHAPTER

63

405

line. The state health department will contact the CDC to arrange for clinical consultation by telephone and, if indicated, release of botulinum antitoxin. The CDC’s 24-hour telephone number for state health departments to report possible botulism cases, obtain emergency consultation, and request botulinum antitoxin is 770-488-7100. Patients with ciguatera, PSP, or botulism must be closely observed for evidence of respiratory compromise. Preliminary data suggest that intravenously administered mannitol may ameliorate the acute neurologic symptoms of severe ciguatera, and tocainide may improve the dysesthesias.135,136 Symptoms of histamine fish poisoning may respond to antihistamines and H2-receptor antagonists. In cases with bronchospasm, epinephrine may be required. For most patients with mushroom poisoning, supportive care is adequate.137,138 Removal of unabsorbed mushroom should be attempted by induced emesis and use of cathartics or enemas. Uneaten mushrooms should be saved, and mycologists or experts in poison control should be consulted for species identification and possible use of specific antidotes. Pyridine hydrochloride and methylene blue may be useful in cases of Gyromitra ingestion. Thioctic acid is an experimental antidote for Amanita poisoning; a regional or nationwide (1-800-222-1222) poison control center should be contacted for information on availability. Liver transplantation has been successfully used in severe Amanita poisoning. Therapy for acute heavy-metal poisoning is supportive. Antiemetics are contraindicated, and emesis should be induced if it does not occur spontaneously. Very severe cases of toxicity may require use of specific antidotes, but this is rarely necessary.

COMPLICATIONS The most common complications of foodborne illnesses are dehydration, electrolyte abnormalities, and hypoglycemia. Children and elderly people are more susceptible to these complications. Severe vomiting can result in subconjunctival hemorrhage, syncope, or Mallory–Weiss esophageal tears. Enteric infection with Salmonella spp., Y. enterocolitica, V. cholerae non-O1, and Campylobacter spp. can be complicated by bacteremia or focal extraintestinal infections, such as osteomyelitis, meningitis, endocarditis, or endarteritis. Infants are also at increased risk of bacteremia and metastatic infections of common bacterial pathogens, particularly Salmonella spp. People with defects of cellular immunity (e.g., HIV infection, leukemia, lymphoma), reticuloendothelial function (e.g., sickle-cell disease, malaria), and iron overload also have increased risk of Salmonella and Campylobacter bacteremia. Other complications include HUS and colonic strictures following enterohemorrhagic E. coli infection, reactive arthritis following Salmonella and Shigella infections, and GBS associated with C. jejuni infection.25,26,139

PREVENTION The risk of foodborne disease can be minimized by careful attention to selection of foods, cleaning of cooking surfaces used for preparation of raw foods, personal hygiene of food handlers, thorough cooking immediately before serving, and proper storage at temperatures too low (less than 4°C) or too hot (more than 60°C) to support bacterial growth. Raw foods of animal origin require particular attention, including poultry, beef, pork, unpasteurized (raw) milk, uncooked eggs, and uncooked shellfish. People at greatest risk of severe disease (i.e., young infants and people with chronic liver disease, decreased gastric acidity, and acquired or congenital immunodeficiency) as well as people who seek to minimize their risk of illness should avoid foods such as uncooked shellfish, raw sprouts, raw milk, or incompletely cooked eggs. Cross-contamination of cooked foods from raw foods via contaminated surfaces and food preparation equipment is a common but avoidable error. Children may be inadvertently infected when food handlers touch their bottles or toys without washing their hands after handling contaminated foods. Attention to the source of fish and

406

SECTION

I

Intra-Abdominal Infections

shellfish may help avoid acquiring an infection or developing food poisoning after consuming these products. Widespread use of hepatitis A vaccine will reduce foodborne transmission of hepatitis A virus. Many foodborne illnesses go undetected. Foodborne illnesses are often misinterpreted as illnesses caused by person-to-person spread (such as “stomach flu”). Stool examinations may not be performed because they are not considered to be cost-effective for an individual patient or may not lead to specific therapy, despite providing a public health benefit. Reporting may be incomplete or delayed, or public health departments may not have adequate resources to conduct an investigation. However, prompt recognition and diagnosis of foodborne illness, along with timely reporting, public health investigation, and use of rapid subtyping methods, can identify causes of outbreaks, recognize emerging pathogens, identify new vehicles and modes of transmission, and prevent additional illness. Recognition of an outbreak may lead to removal of contaminated food items from the marketplace, and enhance industry and regulatory control measures to prevent contamination in the future. Much of foodborne and waterborne illness prevention lies in reducing contamination of food and water before it reaches the

SECTION

consumer. Major revision of meat inspection procedures and revised requirements for drinking water treatment began in the 1990s in the United States. For food and water that may still be contaminated with dangerous pathogens, a systematic disinfection or pathogen reduction is required, such as pasteurization of milk and chlorination of water. In the future treatment of meat and poultry with electron beam irradiation may reduce infection rates even further. Regulatory and industry efforts have been successful in decreasing the incidence of several foodborne pathogens. For example, from 1996 to 2005 the incidence of Listeria decreased by 32%, Campylobacter by 30%, and E. coli O157:H7 by 29%.140 However, foodborne illnesses remain an important public health concern. Clinicians have the important role of providing education and appropriate counseling to high-risk patients, including parents of infants or persons with immunocompromising conditions about the health hazards of foodborne pathogens and their vehicles of transmission. Several websites provide current and educational information about foodborne diseases (cdc.gov/foodnet/, cdc.gov/foodsafety/, nal.usda.gov/foodborne/, and fightbac.org).

I

Intra-Abdominal Infections

CHAPTER

64

Acute Hepatitis John D. Snyder

Acute hepatic inflammation in children can be caused by a large number of infectious and noninfectious causes (Table 64-1). Because the liver has limited mechanisms by which to manifest acute injury, diverse disease states initially can demonstrate similar patterns of hepatic injury. This chapter presents a systematic approach to the evaluation and diagnosis of acute liver injury in immunocompetent children. The pathogenesis of many infectious causes is addressed more fully in other sections of this book.

APPROACH TO EVALUATION Signs and symptoms associated with acute hepatic injury usually include jaundice, vomiting, poor feeding, lethargy, hepatomegaly, and right upper quadrant pain. Assessment of the multiple possible etiologies begins with the patient’s age and a detailed history and physical examination with special emphasis on potential exposures, evolution of symptoms, concomitant health problems, and family history. The physical examination must include a careful evaluation for extrahepatic manifestations of disease as well as a thorough assessment of the abdomen.

Elevated serum hepatic enzyme levels and (often) bilirubin levels are present in persons with acute hepatitis, but the pattern of elevations is rarely diagnostic.1,2 The initial diagnostic tests required to evaluate acute liver injury by age are shown in Box 64-1 and Table 64-2. In most cases, a core group of tests, common for all age groups, is ordered initially. Evaluation for infectious causes is always a central component but requires tailoring for age. The core set of tests given in Box 64-1 is usually ordered simultaneously to provide a complete initial picture of the disease process. Depending on the age of the child, several additional tests are considered (see Table 64-2). The severity of the child’s illness can influence the pace and extent of testing. For example, in a mildly affected child, results of tests for infectious diseases are often evaluated before metabolic disorders are pursued.

INFECTIOUS CAUSES A variety of infectious agents have been implicated in hepatic inflammation in neonates, including bacterial, parasitic, and especially viral pathogens.2–5 Hepatitis in neonates caused by specific agents is usually distinguished from the category of neonatal hepatitis, which has been used to designate hepatic inflammation of no known cause.

Disseminated Infections Disseminated systemic and extrahepatic bacterial and viral infections must always be considered when jaundice is present, especially in the newborn infant.4 Gram-negative bacterial infections are important causes in newborn infants requiring immediate, appropriate therapy.3,4

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Acute Hepatitis

TABLE 64-1. Age of Onset of Infectious and Noninfectious Causes of Acute Hepatitis Age of Occurrence or Recognition Etiology

Neonates and Infants Children Adolescents

INFECTION

Primarily hepatotropic Hepatitis A Hepatitis B Hepatitis C Hepatitis D Hepatitis E Hepatitis G Generalized infection Adenovirus Arbovirus Coxsackievirus Cytomegalovirus Enterovirus Epstein–Barr virus Herpes simplex virus Human immunodeÀciency virus Rubella Varicella Anatomic Biliary atresia Choledochal cyst Congenital hepatic Àbrosis Autoimmune Autoimmune hepatitis Sclerosing cholangitis

– + + + – +

+ + + + + +

+ + + + + +

+ + + + + + + + + +

– – – + – + – + – +

– – – + – + – + – +

+ + + – – –

– + + + + +

– + – + + +

a1-antitrypsin deÀciency + Cystic Àbrosis + Disorders of carbohydrate metabolism Galactosemia + Glycogen storage diseases + Hereditary fructose intolerance + Disorders of protein metabolism Urea cycle deÀciencies + Organic acidemias + Tyrosinemia + Disorders of metal metabolism Neonatal hemochromatosis + Indian childhood cirhhosis – Wilson disease – Lipid storage diseases Gaucher disease + Niemann–Pick disease + Wolman disease + Errors of bile acid metabolism + Inherited Biliary Transport Disorders Byler disease +

+ +

+ +

– + –

– – –

– – –

– – –

– + +

– – +

+ + – –

+ + – –

–

–

METABOLIC DISORDERS

TOXINS/DRUGS

Acetaminophen, alpha-methyldopa, alcohol, amiodarone, chlorpromazine, dilantin, oral contraceptives, halothane, isoniazid, total parenteral nutrition, and amanita toxin

+

+

+

TUMORS IDIOPATHIC

+

+

+

Neonatal hepatitis Reye syndrome

+ +

– +

– +

+, Recognized or usual age of occurrence; –, unexpected or not an age of occurrence.

The pathogenesis of hepatic dysfunction in sepsis is not completely understood but the cholestatic effects of endotoxins and endotoxininduced mediators appear to be important.3,4 Jaundice can also occur in the absence of severe illness, as in gram-negative bacillary urinary

CHAPTER

64

407

BOX 64–1. Initial Diagnostic Evaluation for Suspected Hepatitis in All Age Groups BLOOD Tests of hepatic cell injury and function Bilirubin, total and direct Alanine aminotransferase Aspartate aminotransferase Alkaline phosphatase or gamma-glutamyltranspeptidase Albumin Prothrombin time Ammonia Fasting glucose Tests for infectious causes Serology for hepatotrohic viruses (see Table 64-2) Serology for Epstein–Barr virus Serology, antigen or molecular tests for cytomegalovirus Serology or molecular tests for human immunodeÀciency virus Screening for metabolic disorders Alpha-1-antitrypsin level and protease inhibitor type URINE Shell vial culture for cytomegalovirus RADIOLOGY Ultrasound of liver

tract infection.6 Disseminated infection caused by gram-positive organisms and viruses can also be associated with cholestasis. The diagnosis is usually considered and then established because infected children appear severely ill. A clue to diagnosis is an elevation of conjugated bilirubin greatly out of proportion to elevation of aminotransferases or alkaline phosphatase.3 In the absence of hemolysis associated with mild liver disease, this pattern suggests septicemia. Many viruses in addition to the primary hepatotrophic viruses (hepatitis A, B, C, D, E, and G) must be considered when hepatitis occurs in children. Viruses that can cause hepatic injury as part of a disseminated, multisystem illness include Epstein–Barr virus (EBV), cytomegalovirus (CMV), enteroviruses, adenoviruses, rubella, herpes simplex virus, and human immunodeÀciency virus (HIV)2,4 Congenital rubella is an additional important consideration in neonates. Congenital viral infections are often associated with prematurity, growth retardation, and congenital malformations.2 Other causes of disseminated infection and associated hepatitis in neonates include congenital syphilis, disseminated candidiasis, and toxoplasmosis.2,4 Infection caused by these agents can involve other organ systems, including the skin, central nervous, cardiorespiratory, and musculoskeletal systems.2,4 A wide spectrum of hepatic dysfunction, ranging from mild to fulminant disease, can occur with any of the infectious agents listed above.2–4 Fulminant hepatitis is characterized by rapid progression to very high hepatic enzyme levels, decreased production of coagulation proteins, elevated ammonia, hypoglycemia from loss of glycogen reserves, shock, coma, or death. In 1999, 24 pediatric sites in the United States, Canada, and the UK established an acute liver failure (ALF) registry, with diagnostic and evaluative criteria. The cause of ALF in 348 children is shown in Table 64-3. Only 3 patients had acute hepatitis A, 1 had hepatitis C, and no one had hepatitis B. Infection accounts for only 6% of cases.7

Hepatitis Viruses The six hepatotropic viruses, hepatitis A, B, C, D, E, and G, that cause hepatitis as the primary disease manifestation play a limited role in symptomatic hepatitis in neonates.4,6

Hepatitis A Virus (HAV) HAV infection continues to be the most common cause of acute hepatitis reported in the United States.8 The infection is primarily

408

SECTION

I

Intra-Abdominal Infections

TABLE 64-2. Additional Age-Specific Evaluation for Patients with Suspected Hepatitis Age Group Neonates

Children

Adolescents

Infectious causes

Anti-HBc (IgM), HBsAg Anti-HCV, HCV PCR Toxoplasmosis Ab: mother and child

Anti-HBc (IgM), HBsAg Anti-HCV, HCV PCR Anti-HAV (IgM) Anti-HDV Anti-HEV (if travel history)

Anti-HBc (IgM), HBsAg Anti-HCV, HCV PCR Anti-HAV (IgM) Anti-HDV Anti-HEV (if travel history)

Metabolic disorders

Cystic fibrosis genetic screen or sweat test Serum amino acids

Same as for neonates, plus serum ceruloplasmin

Same as for neonates, plus ceruloplasmin

ESR Quantification of serum immunoglobulins Antinuclear antibody LKM and smooth-muscle antibodies

Same as for children

Same as for neonates 24-hour cooper excretion

24-hour cooper excretion

BLOOD TESTS

Autoimmune diseases

URINE TESTS

Metabolic disorders

Reducing substances Organic acid screening

ESR, erythrocyte sedimentation rate; HBc, hepatitis B core antigen; HA(B, C, D, E)V, hepatitis A (B, C, D, E) virus; HBsAg, hepatitis B surface antigen; IgM, immunoglobulin M; LKM, liver–kidney–microsomal; PCR, polymerase chain reaction.

transmitted through direct human fecal–oral contamination, but food and water have also been implicated.9 HAV is common in early childhood, especially in developing countries with poor conditions of sanitation and hygiene. Almost all children in these countries become seropositive before 5 years of age.8,9 In the United States, transmission often occurs in group childcare facilities.5,9 Approximately 10% of people less than 20 years of age in the United States have serologic evidence for infection but rates are much higher in groups such as American Indians, Alaska Natives, and Hispanics.8,9 The illness is usually mild or unrecognized in young children and often manifests with symptoms of an influenza-like illness.5,8,9 Outbreaks in childcare facilities are usually recognized by illnesses with jaundice in staff or parents of attendees.9 The advent of effective vaccines makes HAV a potentially vaccine-preventable disease.9

TABLE 64-3. Causes of Acute Liver Failure in Children (n = 348) Cause of Acute Liver Failure

Percent

Acute acetaminophen toxicity

14

Metabolic disease

10

Autoimmune disease

6

Nonacetaminophen drug toxicitya

5

b

Infection

Other diagnosed conditions

6 c

Indeterminate cause

10 49

a

Hepatitis B Virus (HBV) HBV is the only hepatotrophic virus that is not directly cytopathic; it causes disease through an immune response against virus-infected hepatocytes. The severity of the infection is inversely related to the effectiveness of the immune system’s ability to diminish viral replication.10 Two main patterns of transmission occur. In endemic areas like China, Southeast Asia, and sub-Saharan Africa, transmission is usually caused at birth or through horizontal transmission among children less than 5 years of age.10,11 In the United States, the most common routes of transmission are injection drug use, sexual contact, and noscomial infection; no risk factor is identified in about 30% of cases.10–12 Infections in neonates are much more common than for HAV, but the incidence is declining in countries with neonatal immunization programs.10–13 Most neonatal infections are not associated with clinically evident disease but chronic infection develops in people in whom infection is not prevented by administration immediately after birth of hepatitis B immune globulin and vaccine.11,12 Children and adolescents are more likely than neonates to develop clinical illness if infected with HBV, which is usually acquired by close contact with an infected adult, sexual contact, or through use of intravenous drugs.10–12 Viral genotypes are important predictors of clinical outcome, antiviral drug response, and mutations.10 The implementation of universal vaccination programs has greatly reduced the disease burden but only about one-third of

17 cases include: trimethoprim-sulfamethazole (1), isoniazid (2), minocycline (1). b 20 cases include: Epstein–Barr virus (6), herpes simplex virus (6), hepatitis A (3), hepatitis B (0), hepatitis C (1), adenovirus (2), cytomegalovirus (1), enterovirus (1). c 34 cases include: shock (16), hemophagocytic syndrome (4). Data from Squires RH, Shneider BL, Bucuvalas J, et al. Acute liver failure in children: the first 348 patients in the pediatric acute liver failure study group. J Pediatr 2006;148:652–658.

the global birth cohort is now vaccintated.12 HBV is potentially a vaccine-preventable disease but only if all countries establish effective vaccination programs for several generations and if at-risk adults are also vaccinated.14

Hepatitis C Virus (HCV) HCV infection in young children is primarily acquired by perinatal transmission in the United States but use of unsafe injections and medical procedures is an important cause in poorer countries.15 In older people most cases are caused by injection drug use (68%) or sexual contact with an infected partner (18%).15 Because of improvements in serologic diagnosis, HCV infection is now rarely caused by blood transfusion or organ transplantation. In contrast to HAV and HBV, no effective HCV vaccine is available so the number of mother-to-infant cases may increase.11 The efficiency of perinatal transmission appears to be low in the general population but increases substantially if the mother is coinfected with HIV or has a high titer

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Acute Hepatitis

of HCV RNA.15,16 HCV infections are usually mild or asymptomatic but a high proportion of infected people go on to develop fibrosis, cirrhosis, and hepatocellular cancer.11,15,16

Hepatitis D Virus (HDV Delta Virus) HDV infections, which can only occur in conjunction with HBV infection, have rarely been described in neonates and are infrequent in all age groups in the United States.17,18 The importance of perinatal transmission appears to be minimal.17 In older infants and children, the disease is uncommon and usually occurs in children with chronic HBV infection.17,18

Hepatitis E Virus (HEV) HEV has only been identified in people living in endemic areas outside the United States and in travelers to those areas, especially the Middle East and Asia; HEV has not been reported as a cause of neonatal hepatitis.19 The disease is similar to hepatitis A except that it primarily affects older children and adults.20

Hepatitis G Virus (HGV) HGV, the most recently identified hepatitis virus, appears to be the mildest of the hepatotrophic viruses.21 Although most infected persons have evidence of persistent viremia, histologic evidence of HGV infection is rare and serum aminotransferase values are usually normal. Currently, there is no conclusive evidence that HGV causes fulminant or chronic disease and it appears that HGV may not be a pathogen.21

Summary of Approach Hepatitis virus infection should be suspected in patients with predominant or severe hepatocellular dysfunction and in fulminant hepatitis. The evaluation for infectious causes of hepatitis in children relies on serologic tests to identify antibodies (usually immunoglobulin M) or antigens or the use of molecular diagnostic techiques, especially polymerase chain reaction (PCR), to diagnose infection.1,2,6 The series of tests ordered, beyond the core group, should be individualized and based on the child’s age and exposures (see Table 64-2). For example, neonates are not evaluated for hepatitis A, D, E, or G except in unusual circumstances. Rubella testing is rarely required since maternal testing is included in routine prepregnancy or prenatal care. The “TORCH” screen is inappropriate because certain agents are unlikely to cause hepatitis as a cardinal feature (toxoplasmosis and rubella) and others (CMV, rubella) are more efficiently diagnosed by culture of urine.2 Liver biopsy is rarely required to make the diagnosis of infectious hepatitis.

ANATOMIC CAUSES Biliary Atresia Biliary atresia is the most common cause of neonatal cholestasis and is usually manifest by 1 month of life.21 The disease was previously named extrahepatic biliary atresia but since fibrotic injury occurs throughout the biliary tract, the term “extrahepatic” is no longer used.21 Although no clinical or laboratory findings are diagnostic of biliary atresia, infants are typically healthy and well grown at birth and are asymptomatic for the first several weeks of life.2,21 By contrast, infants with neonatal infections are more likely to be prematurely born, small for gestational age, and ill-appearing at birth.2 The finding of situs inversus, polysplenia, and congenital cardiac defects associated with cholestasis should suggest the possibility of extrahepatic biliary atresia.21

Intrahepatic Biliary Atresia Intrahepatic biliary atresia, which affects only the intrahepatic biliary tree, can be grouped into syndromic or nonsyndromic varieties; both

CHAPTER

64

409

are usually evident in infancy with jaundice and hepatomegaly. Children with the syndromic type (Alagille syndrome) have a variety of associated anomalies, including peculiar facies (broad forehead, hypertelorism, small chin) and cardiac (most commonly peripheral pulmonic stenosis), ocular (posterior embryotoxon), and vertebral arch (butterfly vertebrae) abnormalities.22 The nonsyndromic form of biliary atresia does not include such anomalies. Byler syndrome is an inherited form of nonsyndromic paucity of bile ducts. Signs and symptoms begin in infancy and progress over a few years to cirrhosis and death.

Choledochal Cysts Choledochal cysts are often recognized in infancy but can become symptomatic in any age group.23 The spectrum of disease is wide, ranging from solitary lesions involving the extrahepatic biliary tree to diffuse intrahepatic involvement (Caroli disease).23,24

Congenital Hepatic Fibrosis Congenital hepatic fibrosis, a syndrome that includes hepatomegaly, cholestasis, cystic disease of the kidneys, and portal hypertension, varies in clinical manifestation depending on the age of the child.25 The renal form of disease (autosomal-recessive polycystic disease) usually predominates in infancy, whereas the hepatic-related form is more common in older children and adults.25,26

Summary of Approach The diagnosis of most anatomic lesions is made by ultrasound (e.g., choledochal cyst) or liver biopsy (e.g., intrahepatic cholestasis, congenital hepatic fibrosis, and sclerosing cholangitis).1,2 Endoscopic retrograde cholangiopancreatography is often helpful in diagnosing biliary tract lesions, especially in older children.1,2

Autoimmune Causes Autoimmune Hepatitis Autoimmune hepatitis usually causes insidious onset of malaise, anorexia, and fatigue in adolescents, with a striking female predominance (see Chapter 65, Chronic Hepatitis). However, the disease can occur in younger children, in both sexes, and with acute manifestations precipitated by another event, such as an intercurrent viral illness.27 The diagnosis of autoimmune hepatitis is made using the clinical and serologic criteria included in a scoring system developed by a group of experts.28 The constellation of findings in autoimmune hepatitis includes hypergammaglobulinemia, autoantibodies, and evidence for other disorders known to be associated with disturbances in immunoregulation.27,28

Sclerosing Cholangitis Sclerosing cholangitis is rare in young children, usually manifesting in adolescence or adult life.29,30 Jaundice and right upper quadrant pain are the most common manifestations.29 The disease is often associated with inflammatory bowel disease, especially ulcerative colitis or Crohn colitis.29,30

Metabolic Causes Metabolic conditions include a large group of disorders that must be considered in every infant with hepatitis, especially children in whom usual infectious causes have been excluded.2,31 Initial manifestations include jaundice, lethargy, vomiting, hepatomegaly, and failure to thrive, mimicking many other causes of hepatic dysfunction. Developmental delay can be an important clue to metabolic disease.

410

SECTION

I

Intra-Abdominal Infections

Diagnostic tests indicated for the initial evaluation for metabolic disease are included in Box 64-1 and Table 64-2. Many of the metabolic diseases occur in early childhood so that extensive testing for these disorders is not usually required when previously healthy children and adolescents come to medical attention with hepatic dysfunction.

children younger than 4 years of age.37,38 By contrast, Indian childhood cirrhosis typically occurs in children younger than 3 years of age and hepatic insufficiency develops in the first week of life in neonatal hemochromatosis.2

a1-Antitrypsin Deficiency

The lipid storage diseases are usually associated with hepatosplenomegaly and progressive central nervous system deterioration. Signs and symptoms can be apparent early in the first year and can quickly progress to death (Wolman disease), or onset can occur throughout childhood and even into adulthood (Niemann–Pick and Gaucher diseases).31

a1-Antitrypsin deficiency can cause a hepatitis-like illness in children of any age group, including neonates.32 Approximately 10% to 20% of a1-antitrypsin-deficient individuals develop signs and symptoms of liver dysfunction at some time.33 Neonatal liver disease is almost always limited to infants with homozygous protease inhibitor phenotype zz.32,33 Clinical manifestations are not distinctive; disease should be considered in any infant, child, or adolescent with jaundice or abnormal liver function tests.32 The condition is frequently unmasked during an intercurrent infection or hepatic insult.

Cystic Fibrosis Hepatobiliary abnormalities have been reported in about one-half of children with cystic fibrosis.34 In the neonatal period, severe cholestatic disease can occur in the absence of pulmonary involvement. Hepatic disease is more frequently found in infants with cystic fibrosis who have associated small-bowel atresia or meconium ileus. The “gold standard” test for diagnosis remains the sweat test.34 However, serologic genetic screening can be helpful in suspected patients of European descent because they often have the most common mutations, including delta*-F508.35 If genetic screening is negative, a sweat test is still required to rule out the diagnosis. Genetic screening can be especially valuable in neonates, in whom a sweat test is often unreliable due to technical difficulties in obtaining an adequate sample.

Lipid Storage Diseases

Disorders of Mitochondrial Fatty Acid Oxidation A growing number of disorders of fatty acid oxidation are being discovered. Children with these disorders can come to medical attention in infancy but often do not develop symptoms until later in life when prolonged fasting, often in conjunction with an acute infection, causes a Reye-like syndrome.39 Cardinal features include hypoglycemia, acidosis, and hepatic injury, often with little elevation in serum bilirubin level.

Disorders of Bile Acid Metabolism An increasing number of disorders that cause errors in metabolism of bile acids are being discovered; these disorders routinely cause cholestasis and hepatitis in neonates.40 These rare disorders were previously placed in the category of neonatal, or idiopathic, hepatitis. Clinical manifestations are not usually diagnostic, but infants often have low alkaline phosphatase levels, a finding that is otherwise rare in conditions of hepatic dysfunction.40

Disorders of Carbohydrate Metabolism

Other Causes

Several disorders of carbohydrate metabolism can cause hepatic dysfunction in infants and children.36 A family history of liver disease should always be sought since these disorders are inherited and the pattern of symptom onset can help to guide the diagnosis. For example, the onset of diarrhea and liver dysfunction occurring after ingestion of fructose (e.g., in fruit juices) raises the likelihood of fructosemia. Galactosemia must be ruled out by testing urine for reducing substances in any neonate with liver dysfunction who receives lactose. In addition, fasting hypoglycemia, in the absence of liver failure or endstage hepatic disease, is an important clue to the possibility of disorders of carbohydrate metabolism, including six of the eight forms of glycogen storage disease.36 Splenomegaly is not common in these disorders and usually occurs only in amylopectinosis disease (type IV glycogen storage disease).36

Toxins and Medications A variety of hepatotoxins, including medications and chemicals, can be associated with a hepatitis-like picture.41 The clinician should obtain a history about exposure to medications, including acetaminophen, valproic acid, tegretol, isoniazid, rifampin, sulfonamides, phenytoin, carbamizine, and phenobarbital. Drug-related hepatitis in neonates is rare but exposure to these compounds must be considered in older infants and children, since the diagnosis is usually made on a clinical basis.41 Children receiving total parenteral nutrition, especially prematurely born infants, have substantial risk for developing hepatic injury.42 Since there is no diagnostic test for this disorder, it must be considered as a diagnosis of exclusion following rigorous evaluation. A rare but important toxic cause of severe hepatic injury in children is ingestion of Amanita mushrooms.

Disorders of Protein Metabolism Disorders of protein metabolism often manifest in infancy, but some diseases, such as the chronic form of tyrosinemia, can occur later in life. Developmental delay and seizures are often associated with these disorders but may not be manifest initially; laboratory evaluation or biopsy may be required to establish the diagnosis.2 Hyperammonemia can be an important clue to disorders involving protein metabolism, including disorders of the urea cycle and organic acidemias.2 Many of these disorders are apparent at birth or shortly thereafter.

Disorders of Metal Metabolism Consideration of disorders of metal metabolism is influenced by the age of the child. For example, Wilson disease, which can present as acute, chronic, or fulminant liver dysfunction, rarely presents in

Tumors Hepatic tumors can cause hepatomegaly and abnormalities of hepatic biochemical tests but are usually identified by initial imaging studies. The liver often has a hard, rock-like feel on palpation.

Idiopathic Causes Age helps in differentiating disorders in the idiopathic category. Neonatal hepatitis (also called giant-cell hepatitis) is defined as a group of disorders of unknown etiology associated with cholestasis in the neonate and young infant. Findings in neonatal hepatitis can be similar to those of biliary atresia but are usually distinct from neonatal infectious hepatitis, which characteristically is part of an illness affecting multiple organ systems.2

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Chronic Hepatitis

CHAPTER

65

Chronic Hepatitis Walter E.B. Sipe and John D. Snyder

Chronic hepatitis, which is defined as hepatic injury that persists for at least 6 months, is a clinical and pathologic syndrome associated with a wide variety of diseases and conditions1,2 (Table 65-1). Most, but not all, cases of chronic hepatitis are judged by a combination of histologic, clinical, and serologic abnormalities.1,2 Although continuous activity for 6 months provides definite proof of the persistent nature of the hepatitis, a combination of clinical, laboratory, and histologic findings can establish the diagnosis more expeditiously1 and, in the case of autoimmune hepatitis (AIH), lead to earlier treatment.3 The goal of this chapter is to provide a practical approach to the evaluation of a child with chronic hepatitis.

APPROACH TO EVALUATION The evaluation of people with chronic hepatitis begins with a careful history and physical examination. The evaluation can be challenging since onset is often insidious and many patients are asymptomatic.2 Even when signs and symptoms are present, they can take a number of

CHAPTER

65

411

forms. Often initial presentation resembles that of acute hepatitis, including fatigue, malaise, abdominal pain, anorexia, weight loss, dark-colored urine, clay-colored stools, and fever.2 Other cases may only be diagnosed after months or years of nonspecific constitutional symptoms that can include relapsing jaundice. Alternatively, variceal bleeding or organomegaly can be the presenting sign. Several additional historical clues should always be sought. For example, history of exposure to blood products, use of intravenous drugs, or maternal infection are important risk factors for hepatitis B, C, and D.4 The presence of thyroiditis, Sjögren syndrome, or idiopathic colitis in a female raises the possibility of autoimmune liver disease. History of exposure to drugs and toxins must also be sought. The pattern of biochemical abnormalities is not usually diagnostic for a specific etiology. Serum levels of aminotransferases are often elevated but can be intermittently normal in hepatitis C infection.4 In addition, levels of aminotransferases do not reliably reflect the severity of disease by liver biopsy.4 Patients with chronic hepatitis can progress to cirrhosis with normal serum levels of aminotransferases.1,4,5 Bilirubin elevations are variable, and levels of other serum enzymes such as alkaline phosphatase and gamma-glutamyltranspeptidase can be normal or minimally increased. In contrast to acute hepatitis, liver biopsy is often an essential tool in the diagnosis and management of patients with chronic hepatitis. In addition to diagnosis, biopsy is used to assess prognosis by grading the severity of disease and staging its progression. The terms chronic active, chronic persistent, and chronic lobular hepatitis have been replaced by new terminology that grades the severity of liver inflammation and fibrosis from minimal to severe.1

BOX 65-1. Laboratory Evaluation of Children with Chronic Hepatitis

TABLE 65-1. Causes of Chronic Hepatitis by Age of Onset Age of Onset Causative Factor

Neonates, Infants

Children Adolescents

HEPATOTROPIC INFECTION

Hepatitis B Hepatitis C Hepatitis D

+ + +

+ + +

+ + +

+ + + + +

+ + + – +

+ + + – +

Biliary atresia Congenital hepatic fibrosis Sclerosing cholangitis Primary biliary cirrhosis

+ + – –

– + + –

– – + +

AUTOIMMUNE DISORDERS METABOLIC DISEASES

–

+

+

+ + +

+ + +

+ + +

– – +

+ + –

+ + –

TOXINS AND DRUGS IDIOPATHIC

+

+

+

Cryptogenic Neonatal hepatitis

+ +

+ –

+ –

GENERALIZED INFECTION

Cytomegalovirus Epstein–Barr virus Human immunodeficiency virus Rubella virus Varicella virus ANATOMIC ABNORMALITIES

a1-Antitrypsin deficiency Cystic fibrosis Carbohydrate, protein, and lipid disorders Obesity Wilson disease Errors of bile acid metabolism

+, recognized or usual age of occurrence; –, unexpected age or never an age of occurence

INITIAL TESTS Blood Tests of hepatic injury and function Bilirubin, total and direct Alanine aminotransferase Aspartate aminotransferase Alkaline phosphatase or gamma glutamyl transpeptidase Albumin Prothrombin time Ammonia Fasting glucose Tests for infectious agents Hepatitis B surface antigen Antihepatitis B core antigen Antihepatitis C virus Antihepatitis D virus Epstein–Barr serology Cytomegalovirus Human immunodeficiency virus Urine Shell vial culture for cytomegalovirus for neonates Imaging Ultrasonography of the liver FURTHER EVALUATIONS IF TEST RESULTS FOR INFECTIOUS CAUSES ARE NEGATIVE Blood Erythrocyte sedimentation rate Quantitative serum immunoglobulin levels Antitissue transglutaminase and serum immunoglobulin A level Autoantibody tests • Antinuclear antibody • Liver–kidney–microsomal antibody • Smooth-muscle antibody Metabolic tests • a1-antitrypsin level with protease inhibitor type • Ceruloplasmin level Urine Quantification of copper in 24-hour specimen Liver Biopsy

412

SECTION

I

Intra-Abdominal Infections

The list of diagnostic possibilities included in Table 65-1 should be considered for each patient with chronic hepatitis. The approach to the initial evaluation of these patients is shown in Box 65-1. Since infectious causes are the most common, they are sought first.4,5 If this initial evaluation, which should take only 1 or 2 days, is negative, the second stage of testing is undertaken.

CAUSES Infectious Causes Evaluation for infectious agents begin with testing for hepatitis B, C, and D, the three hepatitis viruses that cause chronic disease. Hepatitis B virus (HBV) infections are the most common cause of chronic hepatitis in children.4,5 The risk of developing chronic hepatitis is directly related to the age of acquisition of infection. Untreated perinatal infection is associated with a 90% to 95% risk of chronic hepatitis in infants born to mothers who are hepatitis B e-antigen-positive, whereas infection in adults results in less than 10% incidence of chronic hepatitis.4,5 Exposure to blood products, intravenous drug use, and institutionalized settings all increase the risk of HBV and subsequent chronic hepatitis.6,7 Hepatitis D virus (HDV) is an incomplete virus that only occurs in conjunction with HBV infections. Perinatal transmission is rare, and most cases of HDV infection occur as superinfections in patients who are long-term carriers of HBV.8 Superimposed infection with HDV appears to convert an often mild illness into severe hepatitis that can become fulminant or progress to cirrhosis.9 The diagnosis is made serologically and should only be considered in patients with HBV infection.8,9 The signs and symptoms of chronic HBV infections are variable; many children have a mild or asymptomatic course.6,7 Symptomatic cases of chronic HBV infection are difficult to distinguish from other causes of acute and chronic hepatitis. The diagnosis is usually made on serologic criteria; liver biopsy is primarily performed for prognosis or for evaluation of experimental therapies.10 Hepatitis C virus (HCV) has emerged as one of the most common causes of chronic hepatitis in the United States,11 but only a small proportion of cases occur in the pediatric population.12 Prior to 1990 and the advent of effective screening techniques, children who had received blood products were at greatest risk for infection.11,12 Most new pediatric cases are caused by vertical transmission, with a risk of infection in infants born to seropositive mothers of approximately 5%.12 Diagnosis is usually made by detection of antibodies to core and nonstructural antigens but detection of active infection is best established by molecular diagnostic techniques, using polymerase chain reaction. Diagnosis of hepatitis C in infants should be confirmed after the first year of life, because transient viremia has been observed in neonates.11 The laboratory findings of chronic HCV are notable for a fluctuating pattern of aminotransferase values with swings from normal to mild or moderate elevations.12 Liver biopsy is primarily performed for prognosis or to help with evaluation of experimental therapies.10,11,12 Over one-half of children and adults infected with HCV develop chronic hepatitis.7,11,12 Disease can progress to cirrhosis, hepatocellular carcinoma, and death, but most adults have milder disease than that associated with chronic HBV infection.11 Although few data are available for children, the progression of disease also appears to be slow in children.12 Cytomegalovirus and Epstein–Barr virus infections can cause chronic hepatitis, although progression of these infections to cirrhosis and liver failure is rare.13

Nonalcoholic Fatty Liver Disease Nonalcoholic fatty liver disease (NAFLD) refers to a spectrum of conditions associated with fatty infiltration of the liver.14 Findings range from simple fatty infiltration to nonalcoholic steatohepatitis (NASH), a condition in which inflammation or fibrosis is also present and which can advance to cirrhosis and advanced liver disease.14

NAFLD is typically associated with obesity, insulin resistance, and hyperlipidemia, and an estimated one-third of the adult population may have steatosis.15 Because of the dramatic increase in prevalence of overweight children and adolescents, NAFLD may now be the most common cause of chronic liver disease in pediatrics.16 Patients are often asymptomatic, and come to attention when elevated aminotransferase levels are noted incidentally. The physical examination is usually only remarkable for excess weight (body mass index > 85% for age) and hepatomegaly.14 Acanthosis nigricans may also be present as a sign of associated insulin resistance. Ultrasound demonstrates a homogeneous pattern of fat in the liver. A liver biopsy is almost always required because other causes of liver disease must be excluded before the diagnosis of NAFLD can be made with certainty.15 Hepatic dysfunction usually normalizes if body mass index can be improved.

Autoimmune Causes Autoimmune liver disease is a heterogeneous group of diseases consisting of AIH, primary sclerosing cholangitis (PSC), and primary biliary cirrhosis (PBC).17–19 AIH causes a diffuse pattern of injury to liver whereas PSC and PBC primarily cause injury to the biliary tract, with secondary cholestasis and hepatocyte injury. Although AIH is a relatively rare cause of chronic hepatitis, it should always be considered in patients who have the distinctive clinical and serologic findings of female sex, coexistent autoimmune disease, low ratio of albumin to total protein or elevated immunoglobulins, and a relative elevation of serum transaminases compared with alkaline phosphatase.17 In addition, elevated titers of nonorganspecific autoantibodies are characteristic of AIH, and are used to identify subgroups of disease. Type 1 AIH usually has antinuclear (ANA) or antismooth-muscle (SMA) antibodies, whereas type 2 has antiliver–kidney–microsomal (LKM) antibodies in the absence of ANA and AMA.3,17 Although clinical, laboratory, and histologic findings can be helpful in suggesting the diagnosis, there is no single diagnostic test for AIH, and careful exclusion of infectious, toxic, and metabolic causes is critical. A widely applied scoring system has been developed to improve diagnostic consistency among clinicians and researchers.20 Chronic liver disease is rarely associated with sclerosing cholangitis in children and adolescents, but if found, is most commonly seen in people with evidence of colitis caused by inflammatory bowel disease.16,18 Jaundice and right upper quadrant pain are the most common sign and symptom.18 Diagnosis often requires endoscopic retrograde cholangiopancreatography (ERCP). The continuing improvements in magnetic resonance cholangiopancreatography (MRCP) may allow this technique to supplant use of the more invasive ERCP in the future.21 PBC is primarily a disease of middle-aged females but it has been identified in adolescents.19 The presence of antimitochondrial antibodies is a characteristic serologic feature.22

Metabolic Causes Metabolic disorders are uncommon causes of chronic hepatitis but should be considered, especially when initial testing for infectious causes does not result in a diagnosis.2 The clinical presentation of these disorders is not usually distinctive enough to permit diagnosis without laboratory testing, which may include a liver biopsy. Although a1-antitrypsin deficiency most commonly affects lungs, it can cause chronic liver disease in all age groups, including neonates.23 Rarely, patients manifest both lung and liver disease. Liver involvement in some infants progresses to cirrhosis and early death, but many individuals lead active lives through childhood before developing signs of chronic liver failure in late adolescence.23,24 When the disease develops later in childhood or adolescence, a wide spectrum of illness can occur, ranging from mild portal fibrosis to cirrhosis or hepatoma.24 The progression of disease is usually slow in older patients. Diagnosis is made by measuring the serum a1-

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Granulomatous Hepatitis

antitrypsin level and the protease inhibitor (PI) type of the affected person.23 Usually the most severe disease develops in people who are homozygous (PI type ZZ), but disease also occurs in heterozygotes having MZ, SZ, and Z null phenotypes.23,24 Chronic hepatic dysfunction occurs in as many as one-half of children with cystic fibrosis.25 Except in the first year of life, liver disease is rarely the initial manifestation of cystic fibrosis, so diagnostic difficulties are not usually encountered. The diagnosis of cystic fibrosis is made by sweat test or by genetic screening for the most common mutations.25 Disorders of carbohydrate (e.g., glycogen storage disease), protein (e.g., chronic form of tyrosinemia), and lipid (e.g., Niemann–Pick and Gaucher disease) metabolism rarely manifest after the neonatal period as chronic hepatitis.13 These diseases often require liver biopsy as well as laboratory evaluation for diagnosis. They are not usually tested for as part of the initial evaluation unless the patient has signs and symptoms such as hypoglycemia, seizures, or developmental delay. Wilson disease can manifest as acute, chronic, or fulminant hepatitis, and since it is potentially treatable, it must be part of the differential diagnosis of all children and adolescents evaluated for liver disease.26 The disease rarely has not been reported in children younger than 4 years of age. Wilson disease should also be considered when liver disease is associated with hemolytic anemia, osteoporosis, renal disease, or neurologic deterioration.26 Kayser–Fleischer ring on ophthalmologic examination is the most distinctive finding on physical examination. The diagnosis is often suggested by low serum ceruloplasmin level and elevated urinary copper level (especially after penicillamine treatment) and is confirmed by elevated levels of copper in the liver biopsy specimen.26 Liver disease, though seldom severe, has been observed in association with celiac disease (gluten-sensitive enteropathy). The diagnosis should be considered in all cases of hepatitis when no other cause can be found.27

Toxins/Drugs Patients should always be questioned for a history of exposures to toxins and intake of drugs, since several agents can cause chronic hepatic injury.28 The most important agents include a-methyldopa, nitrofurantoin, isoniazid, dantrolene, propylthiouracil, valproic acid, and sulfonamides. Affected persons often have abnormalities of autoimmune testing, including positive lupus erythematosus slide preparation test and ANA.28 Discontinuation of the drug usually results in improvement in the biochemical and histologic abnormalities. Timing of toxic drug reactions is unpredictable; reactions can occur after years of regular use of a medication. Many drugs, in addition to the more common ones listed above, have caused such toxicity.29 The toxic effect of long-term use of total parenteral nutrition can also be an important cause of chronic hepatitis in children.3

Idiopathic and Anatomic Causes Cryptogenic hepatitis (hepatitis of no known cause) is a diagnosis of exclusion when all the above etiologies have been ruled out. Depending on the location and age of the population studied, cryptogenic hepatitis accounts for about 25% of the cases of chronic hepatitis.13,30 Analysis of risk factors in adults suggests that silently progressive fatty liver may be an underrecognized cause of cryptogenic cirrhosis.30 Neonatal hepatitis, defined as a group of disorders of unknown etiology that cause neonatal cholestasis, is also a diagnosis of exclusion. This diagnosis is usually made early in the neonatal period; rarely do patients go unrecognized and unevaluated until later infancy. The number of cases of neonatal hepatitis is shrinking as more sophisticated methods of diagnosis identify new diseases, especially errors in bile salt metabolism.31 The anatomic causes of hepatitis listed in Chapter 64 (Acute Hepatitis) can be associated with chronic hepatitis, but diagnosis is almost always made before hepatitis becomes chronic.

CHAPTER

CHAPTER

66

413

66

Granulomatous Hepatitis Nada Yazigi and Beverly Connelly

The presence of histologically evident granulomas in the liver is referred to as granulomatous hepatitis. Patients may or may not be symptomatic and hepatic function is usually normal or only minimally deranged.1 Granulomas have been reported as an incidental finding in liver tissue specimens obtained from adults at the time of screening for living-related liver donation2 and in 2% to 10% of liver biopsy specimens overall in adults.3–6 In one review of 521 liver biopsy specimens from pediatric patients, 4% showed granulomas.7 It is estimated that granulomatous hepatitis in children accounts for 5% to 7% of hepatitis cases overall.8 Increased utilization of computed tomography (CT) and magnetic resonance imaging (MRI) in patient evaluations has led to an increasing recognition of granulomatous hepatitis and increased search for the cause(s).

PATHOGENESIS AND PATHOLOGIC FINDINGS The epithelioid cell is the hallmark of hepatic granulomas. Granuloma formation is initiated when monocyte-macrophages migrate into an area of inflammation. Various stimuli can cause the macrophage to transform into an epithelioid cell. It is well established in some disorders that immunomodulatory molecules, such as proinflammatory cytokines, chemokines, and other cytokines, regulate Tlymphocyte function and lead to the formation and maintenance of granuloma.9,10 In most situations more than one mechanism is involved. The triggers are varied, and include intracellular microbial antigens, foreign-body reactions, and hypersensitivity responses. Hepatic granulomas vary in size (50 to 300 mm in diameter) and morphology (from clusters of epithelioid cells to well-developed granulomas rimmed by lymphocytes).1 Epithelioid cells can merge to form multinucleated giant cells. Central caseation or abscess formation may occur. Distribution of epithelioid cells is often patchy, and the small granulomas can only be identified with serial tissue section analysis. The quantity and distribution of granulomas are best established with the periodic acid–Schiff stain.11 Some histologic features orient the pathologist towards a particular diagnosis.11 In tuberculosis, the granulomas display central caseation and the epithelioid cells are in a radial array at the periphery; Langhan giant cells may be seen. In sarcoidosis, the granulomas are large and loose, the epithelioid cells show no pattern and there is no central caseation, but there may be central eosinophilic necrosis and multinucleated giant cells.12 In chronic granulomatous disease of childhood, pigmented macrophages are found in architecturally normal liver, and necrotizing granulomas are found in areas of active inflammation.13 Bartonella henselae causes granulomas typically with stellate microabscess. Toxocara gives palisading granulomas with numerous eosinophils,8 and caseation is rare.14 Eosinophilic infiltrates distinguish hepatic granulomas caused by visceral larva migrans. A portion of the larva can be seen within the granulomatous inflammation.8 Eosinophilic granules are also often seen with histoplasmosis. Special histologic stains and techniques should be used when an infectious cause is suspected. Acid-fast bacilli (AFB) are demonstrated in fewer than 10% of cases of granulomatous hepatitis caused by Mycobacterium tuberculosis, whereas large numbers of AFB are usually present in patients with the acquired immunodeficiency syndrome (AIDS) who are infected with M. aviumintracellulare.15 Immunohistochemistry techniques can aid in

414

SECTION

I

Intra-Abdominal Infections

identifying viruses, especially cytomegalovirus (CMV) and Epstein– Barr virus (EBV). Nucleic acid amplification techniques may be useful to identify bacteria, viruses, fungi, and rickettsiae in tissues.

CLINICAL MANIFESTATIONS AND DIFFERENTIAL DIAGNOSIS Hepatic granulomas generally reflect a systemic disease.1 In a comprehensive review of 6000 liver biopsies, only 4% reflected a disease limited to the liver.16 When granulomas are limited to the liver, primary biliary cirrhosis is an important diagnosis.5 Clinical manifestations vary with the underlying disease. Fever and chills are often primary symptoms. In 44% of patients with granulomatous hepatitis in one series, diagnosis was pursued because of fever of unknown origin.17 Mild right upper quadrant abdominal pain and tenderness are common. Serum hepatic enzyme levels are generally normal or mildly abnormal.1 Granulomatous inflammation of the liver is induced by numerous infectious agents. Drugs and toxins, inert materials and chemicals, as well as noninfectious conditions of the host can also lead to granulomatous hepatitis. The differential diagnosis is broad and, despite aggressive evaluation, no etiology can be established in many cases.18,19

Infectious Causes The leading causes of granulomatous hepatitis in children are infections,7 the diversity of which are listed in Table 66-1. Geographic variability and host immunocompetence influence the prevalence of the various causes. Mycobacteria are the most frequent bacterial cause of granulomatous hepatitis.19,20 In one review of 63 patients with granulomatous hepatitis, 9 of 11 patients younger than 20 years had tuberculosis.17 In a large biopsy series in India, 55% of granulomatous hepatitis was related to tuberculosis.6 The liver is probably seeded

TABLE 66-1. Agents and Conditions Associated with Granulomatous Hepatitis Infectious Agents

Noninfectious Agents/Conditions

Bacteria Bartonella henselae Brucella spp. Coxiella burnetii Francisella tularensis Listeria monocytogenes Mycobacterium tuberculosis Nontuberculous mycobacteria Nocardia spp. Pasteurella multocida Treponema pallidum Yersinia enterocolitica

Immune Dysregulation Autoimmune hepatitis Biliary cirrhosis, primary Chronic granulomatous disease Common variable immunodeficiency Inflammatory bowel disease Juvenile idiopathic arthritis Sarcoidosis Systemic lupus erythematosus Wegener granulomatosis

Viruses Cytomegalovirus Epstein–Barr virus Hepatitis C virus Human immunodeficiency virus Fungi Candida spp. Coccidioides immitis Histoplasma capsulatum Parasites and Protozoa Amoebas Schistosoma spp. Strongyloides spp. Toxocara spp. Toxoplasma gondii

Drugs/Chemicals Allopurinol Barium Carbamazepine Mebendazole Methyldopa Norfloxacin Penicillins Phenytoin Pyrazinamide Sulfasalazine Talc Quinine Neoplasms Hodgkin disease Lymphoma Idiopathic

during the initial lymphohematogenous dissemination of M. tuberculosis; involvement can occur at any stage of the infection.21 In patients with AIDS M. avium-intracellulare and other nontuberculous mycobacteria can be associated with granulomatous hepatitis. Disseminated bacillus Calmette-Guérin (BCG) infection has been noted to cause granulomatous hepatitis in the selected population of adult patients who receive BCG instillation into the bladder for therapy of bladder carcinoma.22,23 Among other bacteria, B. henselae infection is undoubtedly a major and underdiagnosed cause of granulomatous hepatitis in normal and immunosuppressed children, often manifesting with abdominal pain and fever.24 Infection with B. henselae leads to granulomatous hepatitis in 11% of infected patients presenting with atypical disease.25 Histologically, necrotizing granulomas, with or without periportal and retroperitoneal lymphadenopathy and splenomegaly, are evident.26,27 Progress in diagnostic testing in the 1990s has led to the recognition that chronic hepatitis C may be a leading cause of viralinduced granulomatous hepatitis.4,28 In a recent review in the United Kingdom, hepatitis C accounted for 10% of all cases, surpassing tuberculosis more than twofold.28 Characteristically, in chronic hepatitis C, granulomas are noted in the portal tracks.4 Chronic infection with hepatitis A, hepatitis B, CMV, and EBV have been implicated in granulomatous hepatitis as well. Several tickborne infections, including ehrlichiosis, tularemia, Q fever, and Lyme disease, have rarely been associated with granulomatous hepatitis.29 Among parasites and protozoa, schistosomiasis is a leading cause of hepatobiliary disease worldwide, particularly in tropical and subtropical regions of Asia, the Caribbean, South America, and Africa, where the pathogens are prevalent.30 Histopathologic examination of biopsy tissue reveals granulomatous inflammation in the portal vessels in response to schistosomal ova. The inflammatory reaction ultimately leads to fibrosis, portal hypertension, and massive splenomegaly.31 Migrating larvae of Toxocara canis and T. cati lead to hepatic or splenic granulomatous disease associated with visceral larva migrans.32 Peripheral eosinophilia is often significant and organ enlargement can be severe. Histoplasmosis is the most common fungal infection reported in association with granulomatous hepatitis in immunocompetent as well as immunocompromised patients.6,7,33 Hepatic and splenic granulomas and abscesses due Candida species are well-recognized complications in the neutropenic and otherwise immunocompromised population, and occasionally in children with bloodstream infection related to a central venous catheter.

Noninfectious Causes Noninfectious causes of hepatic granulomas are noted in Table 66-1. Systemic conditions with abnormal immune regulation are most associated.34 Sarcoidosis is by far the most frequently encountered entity with multisystem manifestations that causes granulomatous hepatitis.6,17,18,35,36 In chronic granulomatous disease of childhood, the liver is one of a number of organs that are chronically infected, containing abscesses or noncaseating granulomas.37 Hepatic granulomas can occur in a variety of lymphoreticular disorders, such as Hodgkin disease.3,28 Primary biliary cirrhosis, affecting mostly adult women, is the most common cause of granulomas isolated to the liver.5,28 When present in primary sclerosing cholangitis, hepatic granulomas have been associated with more severe systemic immune dysregulation and linked to worse outcome.39 Numerous drugs and toxins have been associated with noncaseating granulomatous hepatitis.39,40 Notable on the list of agents are antibiotic, anticonvulsant, and anti-inflammatory agents. It has been speculated that many of the “idiopathic” cases of granulomatous hepatitis may be drug-related.40 Hepatic granulomas can be seen in the context of inflammatory bowel disease (Crohn) and in chronic inflammatory systemic disorders

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Granulomatous Hepatitis

such as juvenile idiopathic arthritis1,8 and in individuals with abnormal immune function such as common variable immunodeficiency.34 In all of these individuals with immune dysfunction, any of the infectious causes can occur as well.

DIAGNOSIS Granulomatous hepatitis should be suspected in any patient with fever of unknown origin or with a systemic disorder in which right upper quadrant abdominal pain or tenderness, hepatomegaly, or a mild elevation of serum concentration of hepatic enzymes is seen. Granulomas are also sometimes seen on imaging studies as part of the evaluation of such symptoms. Ultrasonography, but mostly CT scan and MRI, are sensitive screening diagnostic modalities; MRI may allow differentiation between caseating and noncaseating granulomas.41 As depicted in Figure 66-1, testing, guided by a careful history to capture epidemiologic clues, and the physical exam, paying attention to clues for underlying systemic disease, can often lead to the diagnosis, and guide the treatment without having to resort to a liver biopsy. Because of the prevalence of viral infections such as EBV, CMV, and hepatitis C, testing for these in the absence of tissue examination may not be specific. In immunocompromised individuals and those with progressive symptoms without a forthcoming

FUO or other systemic disease RUQ pain or tenderness Hepatomegaly

Imaging studies as part of evaluation for systemic disease (abdominal ultrasound, CT, or MRI)

Suspected Granulomatous hepatitis

Diagnosis forthcoming and/or patient’s condition stable?

Yes Treatment options in symptomatic patients: • Withdraw offending medication • Treat underlying disease/condition • Treat infection

CHAPTER

66

415

diagnosis, a tissue diagnosis should be sought. Tissue from a lymph node or skin lesion in a patient with systemic disease may provide the diagnosis, avoiding the more invasive liver biopsy. Tissue cultures may yield an offending pathogen. Histopathologic clues from routine and special stains can guide immunohistochemistry and nucleic acid amplification testing (PCR and in situ hybridization) of tissue specimens and substantially increase detection of most infectious agents.42 Further serologic studies and serum PCR studies may then be indicated. Thorough investigation for all possible causative diseases should be performed before the condition is labeled idiopathic.

MANAGEMENT AND OUTCOME Granulomatous hepatitis of infectious origin regresses with appropriate targeted therapy. Treatment of any underlying disease, however, is the cornerstone of management. Empiric therapy with corticosteroids has been advocated in idiopathic cases in adults, although illness may resolve without therapy in these patients.43 There are no collective data regarding the outcome of children with idiopathic granulomatous hepatitis. Thus, empiric therapy with corticosteroids is rarely indicated in children in view of the predominance of infectious causes of pediatric cases.

Carefully review: Medical history • Underlying diseases, conditions • Exposure history, including tuberculosis, animals, ticks, chemicals, blood products • Travel history • Drug history (therapeutic and illicit) Physical examination • Clues to systemic illness

Targeted evaluation for infectious causes • PPD • Selective cultures • Selective serologic tests • Selective antigen or PCR studies Evaluation for noninfectious causes • Serum angiotensin-converting enzyme level • Serum antinuclear, antimitochondrial antibody level No Biopsy evaluation: • Tissue cultures • Histopathology, special stains • Immunohisochemistry • In situ hybridization • Tissue PCRs

Figure 66-1. An approach to the diagnosis of granulomatous hepatitis. CT, computed tomography; FUO, fever of unknown origin; MRI, magnetic resonance imaging; PCR, polymerase chain reaction; PPD, purified protein derivative; RUQ, right upper quadrant.

416

SECTION

CHAPTER

I

Intra-Abdominal Infections

67

Acute Pancreatitis Nada Yazigi and Beverly Connelly Acute pancreatitis is a rare disorder in children. In its mild form, pancreatitis is likely underdiagnosed in pediatrics, and resolves spontaneously with no sequelae. The etiologic factors leading to pancreatitis cover a wide variety of disorders that vary in incidence by age. In adults, alcohol and biliary tract disease are the major causes of acute pancreatitis.1 In children, pancreatitis is most often due to trauma (22% to 25% of cases); structural anomalies, systemic disease (e.g., cystic fibrosis, metabolic disorders), drugs and toxins each account for 10% to 15% of cases; the remainder of cases are attributed to infections (10% to 15% of cases), and increasingly recognized hereditary forms.2–7

sensitivity of 95% to 100% if tested within the first 24 hours after onset of symptoms; serum amylase level tends to fall within 48 hours of onset. Specificity (70%) and positive predictive value (15% to 72%) of elevated serum amylase level are low.13 Serum lipase is more specific for pancreatic inflammation and is more sensitive in later stages of the disease.14 When used together, elevated serum amylase and serum lipase values yield a specificity of 90% to 95%.15 Persistence of elevated amylase beyond 48 hours after the onset of symptoms raises the possibility of the development of pancreatic pseudocyst. Abdominal ultrasonography is the best initial imaging study when pancreatitis is suspect. Ultrasonography can show swelling of the pancreas, but most importantly helps exclude surgically treatable causes of pancreatitis such as gallstones, and extrapancreatic masses. A plain film of the abdomen is useful to exclude other causes of abdominal pain; pancreatic calcification can occur in patients with acute recurrent or chronic pancreatitis. Intravenous contrast-enhanced abdominal computed tomography is the most useful modality for diagnosis of significant abdominal trauma, necrotizing pancreatitis, and pseudocyst.16 It should be done in all cases of severe pancreatitis to guide therapy. Endoscopic retrograde cholangiopancreatography or magnetic resonance cholangiopancreatography can be useful in selected cases, particularly where therapeutic intervention is needed.

PATHOGENESIS Inflammation of the pancreas in response to a variety of insults is due to autodigestion of the organ by pancreatic enzymes. The precise circumstances and chronology of events leading to enzyme activation vary by etiology.8 The end result, however, is that enzymes are activated within the pancreatic parenchyma, leading to local inflammation and damage. The inflammatory cascade that follows pancreatic injury results in the systemic inflammatory response syndrome.9 Interleukin 1 (IL-1) and tumor necrosis factor (TNF) mediate this response. These agents, in turn, increase production of platelet-activating factor (PAF), nitric oxide, IL-6, and IL-8. Inflammatory mediators are responsible for the clinical findings of severe acute pancreatitis: adult respiratory distress syndrome, vascular leakage, renal dysfunction, hypovolemia, and shock. Other factors, such as activated complement (C) 5 and IL-10, may act to dampen the inflammatory response.10

CLINICAL MANIFESTATIONS Abdominal pain is the predominant symptom in most patients, and can be isolated. The pain is typically in the epigastric area and is continuous and dull in nature. It begins abruptly, increases in severity, and peaks in a few hours. Sometimes the pain is mild, is localized elsewhere in the abdomen, or radiates to the middle of the back. Nausea and vomiting occur in 70% of patients. The child’s position of comfort is usually one with the knees flexed on the abdomen. Epigastric or midabdominal tenderness with guarding is often found on physical examination. Bowel sounds disappear as ileus develops. An abdominal mass can develop as a manifestation of a pancreatic pseudocyst. Rigid abdomen or cutaneous signs of hemorrhagic pancreatitis, such as bluish flank (Grey Turner sign) and periumbilical discoloration (Cullen sign), are rare and represent signs of severity.11 Care should be taken to look for complications such as shock. In pediatrics, the vast majority of pancreatitis remains mild, and has spontaneous resolution.

DIAGNOSIS The diagnosis of acute pancreatitis relies on clinical symptoms and should be considered in any child manifesting upper or diffuse abdominal pain or shock.12 No single definitive diagnostic test is available. A comprehensive history highlighting potential causes should be elicited. The diagnosis is usually confirmed by the finding of elevated serum pancreatic enzymes. Elevated total serum amylase has a

DIFFERENTIAL DIAGNOSIS The differential diagnosis of acute pancreatitis varies with the severity of the pancreatitis attack. In its mild forms (the most frequent presentation in children), it can mimic acute gastroenteritis. Mild acute pancreatitis can also occur during acute viral gastroenteritis. Biliary tract disease (cholelithiasis, choledocholithiasis, choledochal cyst), peptic ulcer disease (penetrating or perforated peptic ulcer), intestinal obstruction, and factitious pancreatitis should be excluded. Renal failure and diabetic ketoacidosis can be responsible for falsely elevated serum amylase levels and should be excluded as well.

Infectious Causes Although many infectious agents have been implicated in acute pancreatitis (Box 67-1), evidence supporting their causality is usually indirect, based on results of serologic assays or epidemiologic association. Undiagnosed but self-limited viral illnesses likely account for many “idiopathic” cases of acute pancreatitis. Viral agents, particularly mumps, are the most common infectious pathogens.2 Mild pancreatitis occurs in up to 15% of patients with mumps. Parotitis usually precedes abdominal pain and vomiting by 4 to 5 days; pancreatitis rarely occurs alone. Enteroviruses, especially group B coxsackieviruses, are also implicated frequently.17,18 Epstein–Barr virus (EBV) infection has been associated with pancreatitis in multiple case reports.19,20 Hepatitis B surface and core

BOX 67-1. Common Infectious Causes of Acute Pancreatitis in Children VIRUSES Mumps virus Group B coxsackieviruses Other enteroviruses Hepatitis A Hepatitis B Epstein–Barr virus Cytomegalovirus Human immunodeficiency virus Influenza A virus Mycoplasma pneumoniae

PARASITES Ascaris lumbricoides Clonorchis sinensis Cryptosporidium parvum Echinococcus granulosus

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Cholecystitis and Cholangitis

antigens (HBsAg and HBcAg) have been detected in pancreatic acinar cells, a finding that may explain the pancreatitis associated with hepatitis B infection.21 A report from Brazil documented 4 cases of pancreatitis associated with measles infection; 3 of the 4 patients were immunosuppressed.22 Pancreatitis in patients with human immunodeficiency virus (HIV) infection is attributed to direct HIV infection, superinfection with cytomegalovirus or another opportunistic agent (e.g., Toxoplasma gondii), tumor, or drug therapy (e.g., pentamidine or dideoxyinosine).23–26 Pancreatitis unrelated to antiretroviral therapy is a poor prognostic indicator in HIV-infected children. Ascaris lumbricoides, Clonorchis sinensis, and Cryptosporidium parvum cause pancreatitis via physical obstruction of the pancreatic duct as they migrate from the intestinal lumen into the biliary or pancreatic ducts or both.27–29 Bacteria are unusual causes of acute pancreatitis, but enteric organisms complicate necrotizing pancreatitis and worsen the prognosis. Fungal pathogens have not been reported as causative agents in pancreatitis, but can infect a necrotic pancreas.

Noninfectious Causes The many noninfectious causes of pancreatitis are shown in Box 67-2. Traumatic pancreatitis occurs in association with blunt abdominal trauma, child abuse, or penetrating wounds or after surgery. Systemic inflammatory disorders like Kawasaki disease,30 juvenile idiopathic arthritis, systemic lupus erythematosus, and Crohn disease have all been associated with acute pancreatitis. Prescription medications as well as the accidental ingestion of the drugs listed in Box 67-2 can be causal. Corticosteroids are the most commonly implicated drugs causing pancreatitis in children. Pancreatitis following bone marrow transplantation can occur during graft-versus-host disease,31 possibly as a consequence of corticosteroid treatment. Acute pancreatitis following liver transplantation is thought to be multifactorial.32 Hereditary pancreatitis, which manifests first in infancy or adolescence, is an autosomal-dominant disorder described in more than 40 kindreds, and linked to chromosome 7q35 in a kindred from the United States.33 Juvenile tropical pancreatitis, which occurs in Africa and Asia, manifests as recurrent abdominal pain in childhood. Both entities characteristically progress to chronic pancreatitis.

Therapy for acute pancreatitis is primarily supportive. Gut rest is only indicated if pain continues or a complication arises. Withdrawing an offending agent is important whenever possible and is particularly relevant for drugs causing cholelithiasis. Pancreatitis may signal the need for treatment of an underlying systemic inflammatory disorder.

BOX 67-2. Common Noninfectious Causes of Acute Pancreatitis in Children • TRAUMA • ANATOMIC Congenital duct anomalies Biliary duct stones Tumor (blocking pancreatic duct) • DRUGS Corticosteroids L-Asparaginase Thiazides Furosemide Dideoxyinosine Azathioprine Valproic acid Others

• INFLAMMATORY DISEASE Kawasaki syndrome Other vasculitic disorders Juvenile idiopathic arthritis Crohn disease Systemic lupus erythematosus • METABOLIC DISORDERS Hyperlipidemia (type I, IV, V) Type I diabetes Glycogen storage disease Cystic fibrosis Mitochondrial disorder • HEREDITARY • IDIOPATHIC

68

417

Supportive care is dictated by the severity of the multiorgan dysfunction that can complicate pancreatitis. Meperidine hydrochloride is usually selected for pain control because it has less of a constrictive effect on the sphincter of Oddi than does morphine. Nasogastric suction is only used in patients with ileus or severe vomiting, for symptomatic relief.34 If fasting is required for more than 3 days and no improvement is seen, parenteral nutrition or jejunal feeding is indicated. Clinical trials suggest a possible role for early enteral antimicrobial therapy (given by oral and rectal routes) to prevent systemic infection in severe pancreatitis.35–37 The administration of broad-spectrum antibiotics to patients with severe, necrotizing pancreatitis results in improved mortality and morbidity.38 The use of agents to inhibit pancreatic secretion, such as cimetidine, somatostatin, calcitonin, glucagon, and fluorouracil, has proven to be ineffective in improving outcomes. Testing of protease inhibitors, PAF antagonist, and cholecystokinin A receptor antagonist have yielded mixed results.39–41 Surgery may be required to exclude peritonitis if the initial diagnosis is uncertain, to correct structural abnormalities, or to treat acute complications (i.e., to remove necrotic tissue, or provide drainage) or delayed complications (e.g., to remove or drain cysts or abscess).

CLINICAL COURSE Most children recover completely from acute pancreatitis; mortality rates in children are not defined clearly. In one series in children, 21% died, but all had serious underlying multisystem disorders.6 Other series report 2% to 20% mortality.3,12 Death or complications are more likely in those with pancreatic or peripancreatic hemorrhage or necrosis, which presages a complicated acute course that can include hemodynamic instability, serious biochemical abnormalities, and multiorgan failure. Local complications include formation of abscess or pseudocyst. Development of a secondary fever indicates a possible complication; abdominal ultrasonography or computed tomography is indicated, and drainage of abscess or guided-needle aspiration of necrotic tissue is required.

CHAPTER

TREATMENT

CHAPTER

68

Cholecystitis and Cholangitis Nada Yazigi and Beverly Connelly

Cholecystitis and cholangitis refer to inflammation of the gallbladder and extrahepatic bile ducts, respectively. They are much more common in adults than in children but occur often enough to be important to clinicians caring for children with acute abdominal illnesses.

PATHOGENESIS Most cases of cholecystitis and cholangitis are initiated by obstruction of normal bile flow through the cystic duct and common bile duct, respectively, by bilary stones, anatomic abnormalities, or devices such as stents. Biliary stones can be composed of pure pigment produced by hemolysis as a complication of hemoglobinopathies (e.g., spherocytosis, thalassemia, and sickle-cell anemia), or in infants after extracorporeal membrane oxygenation support.1 Biliary stones are more often cholesterol-based; these stones occur in association with sudden weight loss, overweight problems, prolonged total parenteral nutrition and with no enteral feeding, chronic liver disorders (affecting the integrity of bile acids secretion) and prolonged use of medications that

418

SECTION

I

Intra-Abdominal Infections

can result in biliary sludge (i.e., some diuretics and antibiotics). Some patients have a familial predisposition to stone formation, particularly as young adults; in these instances the prevalence is higher in women. Portoenterostomy (Kasai procedure), the classic surgical procedure performed for biliary atresia, predisposes children to recurrent cholangitis. In one study, 46 of 101 infants who underwent a portoenterostomy had one to eight episodes of postoperative cholangitis, despite restoration of bile flow.2 Most of the 105 episodes of cholangitis occurred during the first 3 postoperative months; only a few children had cholangitis after 1 year of age.2 Late cases, occurring more than 5 years after successful surgical repair, are also reported.3 Use of the intestinal conduit to restore bile flow from the liver to the small intestine is thought to lead to the ascent of intestinal flora and cholangitis.4,5 Because of this well-recognized risk, patients are usually given antibiotic prophylaxis for 6 to 12 months. Similarly, cholangitis occurs, although less frequently, in children who have undergone orthotopic liver transplantation using a Roux-en-Y procedure for a biliary conduit. Other rare causes of impaired biliary drainage predisposing to cholangitis include choledochal cyst, Caroli disease, primary sclerosing cholangitis, biliary stricture after abdominal trauma, prior duct surgery, cystic fibrosis, and tumors of the extrahepatic ducts (e.g., rhabdomyosarcoma, neuroblastoma). Infection of the biliary tree can also occur during instrumentation for percutaneous or endoscopic retrograde cholangiography.6 In Asia, biliary obstruction can result from migration into the bile ducts by the Chinese liver fluke Clonorchis sinensis or the roundworm Ascaris lumbricoides. Rarely, bile ducts are partially obstructed by daughter cysts from hepatic hydatid disease. Echinococcus granulosus, the dog tapeworm, is prevalent in sheep-raising communities in southern Europe, Australia, and New Zealand. The ductal obstruction results in increased intraductal pressure and/or distention of the gallbladder. Superinfection of the stagnant bile with gut flora organisms follows, along with edema and congestion, further compromising blood supply and lymphatic drainage. Tissue necrosis follows, favoring further bacterial proliferation. In cholangitis, the risk of bacteremia and septicemia is very high,7 accounting for the high morbidity and mortality of this condition, and making cholangitis a medical and surgical emergency. Although obstruction to biliary drainage remains the most frequent cause for cholecystis, occasionally this entity can be seen without it. This condition, known as acalculous cholecystitis, occurs in critically ill patients after trauma, burns, and major surgery, notably spinal instrumentation and fusion for scoliosis.8 It has also been reported in children with infections caused by Mycoplasma pneumoniae, Salmonella typhi, Shigella spp., and Giardia lamblia.9 Acute noncalculous distention (hydrops) of the gallbladder can occur acutely in infants and children. It is associated with a variety of infections and inflammatory conditions (notably Kawasaki disease, streptococcal pharyngitis, septicemia, toxic shock syndrome, and Henoch– Schönlein purpura), and during prolonged fasting and total parenteral nutrition. The affected patient can have pain or a palpable mass or both; ultrasonography shows markedly distended echofree gallbladder without dilatation of the biliary tract. The course is usually selfresolving.

ETIOLOGY Bile is normally sterile. However, in patients with obstructed biliary drainage, bacteria are commonly found in the bile and in the gallbladder wall. The likelihood of bactibilia is greater when obstruction results from gallstones or benign stricture than when it results from malignancy.10 The explanation of this phenomenon is not clear, although the likelihood of a complete obstruction with malignancy is higher, which may prevent the passage of bacteria from the intestinal tract into the biliary system. A few clinical settings and symptoms predict bactibilia: acute cholecystitis, fever, rigors, history of biliary tract intervention, jaundice, and diabetes mellitus.10

Organisms isolated from bile are usually part of the intestinal flora, namely enteric gram-negative bacilli, enterococci, and anaerobes. Common gram-negative bacilli include Escherichia coli, Klebsiella pneumoniae, Enterobacter spp., and Proteus spp. Common anaerobes include Bacteroides spp., Clostridium spp., and Fusobacterium spp. When present, anaerobes are usually part of polymicrobic infections. Bacterial growth was observed in 85% of 145 biliary tract specimens processed for both aerobic and anaerobic bacteria at two hospitals over a 4-year period.11 About 80% of the specimens were from adults with cholecystitis. Aerobic and facultative bacteria were present in 48% of the culture-positive specimens, aerobic bacteria alone were present in 3%, and mixed growths of anaerobic and aerobic or facultative bacteria were present in 49%. The six most common bacterial isolates were Escherichia coli (71 isolates), Enterococcus spp. (42 isolates), Klebsiella spp. (29 isolates), Bacteroides spp. (28 isolates), Clostridium spp. (27 isolates), and Enterobacter spp. (16 isolates).11 Anaerobic infections are more common in patients who have had biliary tract surgery, biliary–intestinal anastomosis, or chronic biliary tract infections. Bacteria isolated from the blood of patients with cholangitis are the same as those isolated from bile, except that bloodstream infection (BSI) caused by enterococci or anaerobic organisms is uncommon. BSI is reported in 21% to 70% of patients with acute cholangitis.12 Bacteria have also been isolated from the liver in patients with acute cholangitis; isolates correlate with those found in the bile.13

CLINICAL AND LABORATORY MANIFESTATIONS Patients with cholecystitis or cholangitis usually have an antecedent history compatible with biliary tract disease, surgery with or without endoprosthesis, or predisposing factors such as a hemoglobinopathy, prolonged total parenteral nutrition, chronic liver disease, cystic fibrosis, or a family history of gallstone formation. Clinical manifestations of cholecystitis are typically milder than those of cholangitis, but there can be substantial overlap. The clinical triad of fever, right upper quadrant pain, and jaundice classically alerts for the presence of biliary tract disease. Cholecystitis is usually heralded by mild epigastric pain with nausea and vomiting, followed by a shifting of the pain to the right upper quadrant, sometimes with radiation to the right shoulder or scapula. Pain is usually aggravated by movement or deep breathing. The gallbladder is palpable in less than 50% of cases. Moderate temperature elevation and mild icterus can be evident, but high spiking fevers, chills, prominent jaundice, circulatory collapse, and other findings of gram-negative BSI are more suggestive of cholangitis. There can also be signs of localized peritonitis, including guarding, rigidity, and rebound tenderness. The white blood cell count is normal or slightly elevated in the presence of cholecystitis and, typically, is substantially increased with a “shift” to immature forms with cholangitis. Slightly elevated serum bilirubin and aspartate aminotransferase (AST) concentrations occur in about 50% of patients with cholecystitis and elevated alkaline phosphatase levels in about 25% of patients. By contrast, most patients with cholangitis have elevated bilirubin, AST, and alkaline phosphatase levels; the degree of abnormality is greater in those with cholangitis than in those with cholecystitis. Serum amylase concentrations can be elevated in both disorders. Disseminated intravascular coagulopathy can occur with cholangitis. Blood cultures are rarely positive in children with cholecystitis, whereas about 50% of children with cholangitis have bacteremia. Culture of liver biopsy specimens can also be positive.13

DIAGNOSIS The differential diagnosis of cholecystitis and cholangitis includes appendicitis, pancreatitis, perforating ulcer, acute pyelonephritis, biliary colic without cholecystitis, hepatitis, perihepatitis (Fitz–Hugh–Curtis syndrome), and right lower lobe pneumonia.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Cholecystitis and Cholangitis

Ultrasonography is the most useful diagnostic test in cholecystitis and cholangitis and should be obtained in suspected cases. It can establish the diagnosis of cholecystitis by demonstrating thickening and edema of the gallbladder wall and surrounding tissue and can support the diagnosis of cholangitis by demonstrating dilatation of bile ducts. Ultrasonography can also uncover the underlying cause of biliary tract disease by demonstrating stones, a choledochal cyst, or other obstruction to bile flow. Radionuclide hepatobiliary scanning can demonstrate obstruction of the common bile duct or cystic duct (excluding the gallbladder). Percutaneous transhepatic cholangiography and endoscopic retrograde cholangiopancreatography (ERCP) are also valuable in evaluating and treating obstructions of the biliary tract. Magnetic resonance cholangiopancreatography (MRCP) is emerging as a noninvasive diagnostic modality but does not have the advantage of being therapeutic. Since the advent of the above testing modalities, use of cholecystography has declined in the diagnostic evaluation of suspected cholecystitis or cholangitis. Oral cholecystography is time-consuming and cannot be used in patients with jaundice or vomiting. Intravenous cholecystography is more rapid and is not affected by vomiting, but is not sensitive if the serum bilirubin concentration is > 4 mg/dL, a level often present in patients with cholangitis. Furthermore, cholecystography is less sensitive than ultrasonography or nuclear medicine scans. Culture of a liver biopsy specimen is the most definitive diagnostic procedure for cholangitis. Nonetheless, even in the presence of cholangitis, biopsy specimen culture results can be falsely negative. Blood cultures can be diagnostic and appear to be especially sensitive when performed immediately after liver biopsy (such as in cases in children after liver transplantation). Bile culture can sometimes be useful to guide therapy and is diagnostic when the specimen is obtained at the time of a percutaneous or endoscopic biliary drainage procedure.

TREATMENT The initial treatment of cholecystitis and cholangitis in children consists of supportive therapy and use of broad-spectrum antibiotic therapy effective against enteric gram-negative bacilli and anaerobic bacteria. Drainage of the obstructed biliary tract is the mainstay of treatment for patients with cholangitis who are severely ill. Percutaneous or endoscopic drainage has replaced surgical intervention as the method of choice for emergent re-establishment of bile flow. Antibiotics are complementary to biliary drainage and may circumvent surgery in patients who are not critically ill. Appropriate initial antimicrobial therapy employs an aminoglycoside such as gentamicin or a cephalosporin in combination with an agent active against anaerobic bacteria, such as clindamycin or metronidazole, or an agent combined with a b-lactamase inhibitor. Cefazolin, cefoperazone, ceftriaxone, trimethoprim-sulfamethoxazole (TMP-SMX), mezlocillin, or piperacillin is sometimes selected because of each agent’s high biliary concentration.10,12,14,15 Ampicillin is often included because of the possibility of enterococcal infection. Patients who receive TMP-SMX prophylactically are likely to have resistant enteric bacilli. Antibiotic prophylaxis with piperacillin, cefazolin, cefuroxime, cefotaxime, or ciprofloxacin before therapeutic ERCP has been shown to reduce the risk of bacteremia and cholangitis.16,17 If organisms are isolated from cultures, the antibiotics are adjusted according to sensitivity profiles and then continued for 7 to 10 days. In most cases of treated cholecystitis and cholangitis, symptoms and signs resolve promptly with medical treatment. If the symptoms do not respond to initial medical management, and if gallstones or obstructing lesions are demonstrated on ultrasonography, biliary drainage is indicated. In the past, biliary drainage was accomplished by open cholecystectomy with exploration and drainage of the common bile duct or by percutaneous transhepatic drainage. Generally, current practice is immediate establishment of biliary drainage via endoscopic cholangiography, followed by nonemergent laparoscopic cholecystectomy once the active inflammatory process has subsided.18,19

CHAPTER

68

419

COMPLICATIONS Perforation occurs in 10% to 15% of cases of cholecystitis; sometimes peritonitis results, but more often, the perforation is spontaneously localized by surrounding omentum and serosa of contiguous viscera. When localized, perforation results in persistent fever, a palpable mass, and, occasionally, a friction rub over the liver. Pancreatitis and cholangitis are other complications of cholecystitis. Complications of cholangitis include perforation, macroscopic hepatic abscesses, pancreatitis, and septicemia. This entity is a high medical–surgical emergency.

CHOLECYSTITIS AND CHOLANGITIS IN SPECIAL CLINICAL CONDITIONS Biliary Atresia Recurrent episodes of bacterial cholangitis occur in about 50% of patients with biliary atresia after a successful Kasai procedure.2,4,5 Repeated bouts of cholangitis result in progression of the liver disease to cirrhosis. Therefore, reducing the frequency of cholangitis may prolong the survival after a Kasai procedure. The causative organisms are usually susceptible to TMP-SMX or third-generation cephalosporin agents during the initial episode but are likely to be resistant after repeated episodes of cholangitis and courses of antibiotic therapy.2 A number of reports have suggested that chronic antibiotic prophylaxis can reduce the frequency of recurrent attacks of cholangitis; however, no definitive, placebo-controlled evaluation has been conducted. Although the Roux-en-Y loop (which retains adherence properties of intestinal mucosa) is likely to become colonized with enteric flora despite long-term antibiotic therapy, there may be sufficient suppression of bacterial growth to reduce the concentration of bacteria below the threshold that results in symptoms of recurrent cholangitis.10 TMP-SMX is the antibiotic commonly recommended for the prevention of recurrent episodes of cholangitis.12,20 TMP-SMX is active against most common biliary pathogens, has low toxicity even after prolonged use, is relatively inexpensive, has excellent bioavailability after oral administration, and is excreted in substantial concentrations into bile (especially after repeated dosages).21 Breakthrough infections and superinfections due to resistant Enterobacteriaceae, Pseudomonas, or Candida organisms can occur.17

Sickle-Cell Disease The incidence of gallstones in individuals with sickle-cell disease (SCD) is reported to be between 17% and 33% in those younger than 18 years and > 50% in the adult population in a number of studies. A report of gallstone disease in sickle-cell patients in Jamaica who were studied prospectively from birth to age 17 to 24 years revealed that approximately 30% of homozygous individuals experienced gallstones, whereas only 11% of patients with sickle-cell hemoglobinopathy experienced gallstones.22 Of the 99 patients with SCD anemia and gallstones, only 7 became symptomatic, requiring cholecystectomy, but no patient with sickle-cell disease had symptoms. The risk of gallstones in patients with SCD correlated with higher reticulocyte counts and higher mean corpuscular volume; the risk correlated minimally with elevated unconjugated bilirubin levels and low fetal hemoglobin levels. When gallbladder sludge was noted, gallstones developed within 1 to 2 years in most patients, none of whom had symptoms. One case report of a patient with SCD in whom hepatic duct strictures developed raises the possibility that hypoxic injury to the biliary system results in gallstone formation in patients with chronic anemia.23 Patients with sickle-cell disease who have gallstones and experience repeated bouts of abdominal pain have generally been treated with elective cholecystectomy, and reports have

420

SECTION

I

Intra-Abdominal Infections

indicated a resultant decrease in frequency of hospitalization for abdominal pain.24 Management of acute cholecystitis or cholangitis in the patient with SCD is the same as that for other patients.

Orthotopic Liver Transplantation In children, the biliary anastomosis post liver transplantation is often done via an intestinal Roux-en-Y. This predisposes patients to ascending cholangitis. A chronic subclinical form of cholangitis can occur and manifests as chronic graft dysfunction. This entity, felt to be due to poor drainage of the Roux-en-Y limb, is difficult to diagnose. With improved surgical technique, this chronic situation is currently thought to be rare. It would be unlikely to occur in patients with pre-existing Roux-en-Y (for biliary atresia) who did not have recurrent cholangitis prior to transplantation. Chronic cholangitis should only be considered as the cause of hepatic graft dysfunction when all other causes have been ruled out, and the liver biopsy still suggests acute biliary tract inflammation. A hepatobiliary nuclear scan could be helpful to support the diagnosis. No prospective studies exploring this entity, its diagnosis, or treatment are available. Acute cholangitis is more commonly seen with, and is a hallmark of, liver graft bile duct stenosis. Bloodstream infection (BSI) due to biliary sepsis can be the initial presentation of bile duct stenosis or the condition can be unmasked by a liver biopsy for graft dysfunction; in the latter situation, the diagnostic liver biopsy typically is followed within 12 hours by symptomatic BSI. Right upper quadrant pain is notably absent in these patients due to lack of innervation of the liver graft. The absence of dilated intrahepatic bile ducts by ultrasonography does not exclude the diagnosis; computed tomography is often more sensitive to demonstrate dilated biliary ducts in liver grafts. As the viability of liver graft bile ducts depends on the integrity of the hepatic arterial flow, special attention in these situations should be directed to blood flow in the hepatic artery. Another common cause of biliary stenosis after liver transplantation is chronic rejection. After transplantation, patients have higher risk of BSI due to their immunosuppression. Clinical presentation can initially be subtle, followed by a quick and severe deterioration if not treated promptly. Primary bilary BSI or BSI secondary to hepatic artery thrombosis occur most frequently in the early posttransplantation time. They can also occur years posttransplantation, raising the additional possibility of chronic rejection. Therefore special attention and high index of suspicion for biliary sepsis should always be exercised in this special population, and addressed as a medical emergency.

CHAPTER

69

Peritonitis Shawn J. Rangel and R. Lawrence Moss

Peritonitis is an inflammatory process involving the peritoneum, a specialized lining of the abdominal cavity that functions to lubricate abdominal organs and clear the cavity of infectious particles and other debris. Infectious peritonitis has classically been described as being either primary or secondary in etiology, depending upon on the continuity of gastrointestinal (GI) tract. Primary, or spontaneous, bacterial peritonitis (SBP) is a relatively rare infection that develops in the pre-

sence of an anatomically normal peritoneal cavity. Secondary bacterial peritonitis, which occurs much more frequently, arises from inoculation of the peritoneal cavity with bacteria and other inflammatory debris following intestinal perforation or postoperative anastomotic leak. Catheter-related infections (e.g., ventriculoperitoneal (VP) shunts) comprise a third category of peritoneal infections, which arise following direct or indirect contamination of the indwelling foreign body. Differentiating between primary and secondary causes of peritonitis is crucial as surgical intervention is not indicated in the treatment of SBP and catheter-related infections (except sometimes to remove the catheter).

EPIDEMIOLOGY The overall incidence of infectious peritonitis in children is difficult to estimate due to the heterogeneity in diagnostic criteria and the multitude of underlying disease processes associated with the infection. Primary and catheter-related bacterial peritonitis is relatively uncommon compared with secondary infections and accounts for only 3% to 10% of all cases of peritonitis in the pediatric age group.1–4 In the past three decades, the incidence of catheter-related peritonitis has been steadily decreasing, in large part due to the improvements in catheter and connecting hardware technology used for continuous abdominal peritoneal dialysis (CAPD).5–7 The incidence of SBP in children peaks between the ages of 5 and 9 years, and affects males and females equally. Although SBP has been reported in previously healthy children, this is exceedingly rare.8 In a 22-year review of documented peritonitis in children without pre-existing comorbidities, Freij et al. reported only 7 cases of SBP compared with 1840 cases of secondary peritonitis from acute appendicitis alone.9 In general, immunocompromised children are at greater risk for developing primary and catheter-related peritonitis than are healthy children. The most common predisposing condition is nephrotic syndrome (NS), but children with cirrhosis and portal hypertensive ascites also have increased risk. The first episode of SBP typically occurs during the first 2 years following diagnosis of NS, and the overall incidence of SBP in these children may be as high as 5%.10–15 Children with NS who experience one episode of peritonitits have increased risk for developing subsequent infection.10,13 Children with ascites due to portal hypertension are at high risk for developing SBP, with incidence rates exceeding 20%.16,17 Catheter-related peritonitis remains one of the major complications in children receiving CAPD and is the primary reason for catheter discontinuation outside renal transplantation.18,19 Up to two-thirds of all children receiving CAPD experience at least one episode of peritonitis in their lifetime. Epidemiologic studies have documented a decrease in CAPD-associated peritonitis over the past two decades from one episode per 3 to 6 patient months of dialysis to one episode in 14 to 24 months, although there remains substantial variability in rates of infection.20–24 The overall decrease in incidence is accounted for by a selective decrease in the rate of peritonitis due to Staphylococcus epidermidis.25 Similar decreases in CAPD-associated peritonitis have been reported in Europe, and a remarkably low incidence of peritonitis has been observed in Japanese children (1 episode in every 29 months of dialysis).20,21,26 These results have been attributed to an intensive CAPD training program and the rigorous attention to hygiene practiced in the Japanese culture. Despite the overall decrease in CAPD-associated peritonitis, the incidence of this infection in the pediatric population has remained persistently higher than that observed in adults.27 Most of this difference is attributable to the relatively high incidence of CAPDassociated peritonitis in children under 6 years of age.23,28 The proximity of the catheter exit site to gastrostomy tubes and diapers containing urine and feces has been proposed as a possible explanation. Technical considerations may also play a role, particularly given the frequent use of single-cuffed catheters in children and their anatomically shorter tunnel lengths.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Peritonitis

PATHOPHYSIOLOGY Peritonitis can be caused by a variety of insults, infectious and noninfectious, that produce an inflammatory response involving the visceral and parietal peritoneum. The sequence of pathophysiologic events leading to peritonitis depends upon the complex interplay of many factors, including the nature of the offending agent(s), the ability of the peritoneum to sequester and ultimately clear the infection, and the status of the host immune response. Infectious peritonitis can be exacerbated by a chemical component when there is concomitant intra-abdominal hemorrhage, urine extravasation, or, in the case of perforation of a hollow viscus, leakage of gastric juice and bile. Chemical peritonitis can exacerbate the severity of an adynamic ileus, leading to bacterial stasis and secondary infection through translocation of intestinal flora. The initial inflammatory response of the peritoneum to bacterial infection is characterized by vasodilation, tissue edema, transudation of fluid, and the influx of inflammatory leukocytes. Tissue macrophages residing in the peritoneal space provide the initial phagocytic response to invading organisms, followed by neutrophils recruited from the systemic circulation. Lymphatic channels located in the undersurface of the diaphragm facilitate the clearance of bacteria, endotoxin, and other infectious particles from the peritoneal cavity. This drainage provides an important adjunctive defense mechanism to the local cellular immune response. Impairment of this process by fibrin and inflammatory debris can result in accumulation of peritoneal fluid and dilution of immunoglobulins and opsonins. This may be particularly relevant in the pathophysiology of peritonitis in children with pre-existing ascites, where serum and peritoneal concentrations of these immunoreactive compounds are lower than that seen in healthy children.29,30. Selective IgG subgroup deficiencies have also been characterized in infants undergoing CAPD, which may further predispose this cohort to a more severe course of peritonitis.31,32 Disintegration of gram-negative bacteria following phagocytosis results in the release of lipopolysaccharide, and disintegration of gram-positive bacteria results in the release of peptidoglycans and other inflammatory mediators. Macrophages, neutrophils, and other leukocytes release a host of secondary inflammatory mediators in response to these antigens, including complement, cytokines, interleukins, and other arachidonic acid derivatives. These compounds are absorbed by the peritoneal lymphatics into the systemic circulation where they exert major hemodynamic effects. The large surface area of the peritoneum and its ability to allow bidirectional diffusion of water and electrolytes has great physiologic

CHAPTER

69

421

implications in cases of severe peritonitis. Inflammatory changes involving such a large permeable surface can result in physiologic derangements similar to those seen with large full-thickness thermal burns. Capillary leakage of protein and fluid into the peritoneal space can be substantial and lead to massive fluid shifts away from the intravascular space. This can markedly exacerbate the hemodynamic effects of the systemic inflammatory response, and, in severe cases, lead to rapid and potentially fatal circulatory collapse.

ETIOLOGY Primary Bacterial Peritonitis The precise mechanism leading to primary bacterial peritonitis is unknown, although seeding from a hematogenous source is thought to be the most likely route of infection. Other proposed mechanisms include lymphatic seeding, retrograde inoculation from a genitourinary source, and translocation of intestinal flora (Table 69-1). The bacteriologic profile of peritoneal infections largely depends upon the source of infection and underlying co-morbidities (Table 69-2). Primary peritoneal infection in healthy children is often monomicrobial; Streptococcus pneumoniae is the most common etiology. This pathogen causes over two-thirds of culture-positive cases of peritonitis in children with NS.10,13 In the past few decades, an increasing incidence of peritoneal infections caused by gram-negative organisms (especially Escherichia coli) has been observed.33,34 Immunocompromised patients are also at increased risk for gram-negative peritonitis. Gram-positive organisms (particularly group B streptococcus) predominate as the cause of primary peritoneal infections in the neonate. These infections, also known as “neonatal idiopathic primary peritonitis,” are usually caused by hematogenous seeding, although infection has also been associated with oomphalitis.35–38 Mycobacterium tuberculosis (TB) is a rare cause of peritonitis but should be considered if the history suggests exposure to a known infected individual. Abdominal TB accounts for nearly 12% of extrapulmonary disease and approximately 1% to 3% of all TB-associated infections.39–41 Such peritoneal infections rarely result from the ingestion of infectious particles from an active pulmonary infection. Reactivation of latent disease and subsequent hematogenous spread are more likely mechanisms of infection, although direct extension from caseating mesenteric lymph nodes has also been proposed. Peritoneal infections caused by M. bovis can occur following the ingestion of unpasteurized milk.42

TABLE 69–1. Proposed Mechanisms and Associated Conditions Leading to the Development of Peritonitis in Children Primary Bacterial Peritonitis

Catheter-Associated Peritonitis

Secondary Bacterial Peritonitis

Hematogenous seeding (sepsis) Direct extension of localized process Ascending urinary tract infection Pelvic inflammatory disease Mesenteric adenitis (TB) Omphalitis Lymphogenous seeding Translocation of intestinal flora

Poor sterile technique during dialysis Migration of skin flora along catheter Direct extension of local infection Exit-site infection (with CAPD) CNS infection (with VPS) Hematogenous seeding of dialysate/CSF

Neonatal period Necrotizing enterocolitis Spontaneous gastric perforation Feeding tube perforation (iatrogenic) Intestinal volvulus Hirschsprung disease Postneonatal period and childhood Acute appendicits Intestinal volvulus Intussusception Incarcerated hernia

CAPD, continuous abdominal peritoneal dialysis; CNS, central nervous system; CSF, cerebrospinal fluid; TB, tuberculosis; VPS, ventriculoperitoneal shunt.

422

SECTION

I

Intra-Abdominal Infections

TABLE 69–2. Organisms Commonly Associated with Peritonitis in Children Primary Bacterial Peritonitis

Catheter-Associated Peritonitis

Secondary Bacterial Peritonitis

Previously healthy children Streptococcus pneumoniae Group A streptococcus Staphylococcus aureus Gram-negative bacilli

CAPD-associated Staphylococcus epidermidis Staphylococcus aureus Escherichia coli Other enteric gram-negative bacilli Fungi

Proximal GI perforations Facultative organisms Escherichia coli Klebsiella sp. Enterobacter sp. Pseudomonas sp. Gram-positive cocci

Neonatal idiopathic primary peritonitis Group B streptococcus Streptococcus pneumoniae Staphylococcus aureus Enteric gram-negative bacilli

VPS-Associated Staphylococcus epidermidis Staphylococcus aureus Escherichia coli (late) Other gram-negative bacilli (late)

Anaerobic organisms Bacteroides sp. Clostridium sp. Peptostreptococcus sp.

Nephrotic syndrome Staphylococcus epidermidis Streptococcus species Staphylococcus aureus Enteric gram-negative bacilli

Colonic Perforations Bacteroides fragilis Gram-negative bacilli (Escherichia coli) Other anaerobic organisms Enterobacter sp. (NEC)

Cirrhosis Escherichia coli Klebsiella sp. Streptococcus pneumoniae Organisms are listed in the order of relative incidence by category, although there may by substantial variation in microbiologic data over time and between published series. CAPD, continuous abdominal peritoneal dialysis; GI, gastrointestinal; NEC, necrotizing enterocolitis; VPS, ventriculoperitoneal shunt.

Other infectious causes of primary peritonitis are rare but must be kept in mind under the appropriate clinical circumstances. Peritoneal infection caused Haemophilus influenzae, Neisseria gonorrhoeae, and other encapsulated organisms should be considered in previously healthy children who have had a splenectomy. Gonococcal peritonitis can occur in healthy adolescent girls due to ascending pelvic inflammatory infection. This infection can progress to involve the liver and perihepatic tissues, resulting in gonococcal perihepatitis (Fitz–Hugh–Curtis syndrome).

Catheter-Associated Peritonitis Catheter-associated peritonitis remains one of the major complications in children treated with chronic abdominal peritoneal dialysis. Peritoneal infections may arise from migration of bacteria along the catheter tunnel or from poor sterile technique leading to contamination of connecting tubing or dialysate fluid. Other than renal transplantation, infection-related complications are the most common cause for discontinuation of peritoneal dialysis catheters.18,19 This is a particularly serious issue for children with end-stage renal disease as CAPD is the preferred long-term dialysis modality. Peritoneal cultures are positive in approximately 80% to 85% of children with clinically apparent peritonitis.43 Between one-half and two-thirds of all episodes are caused by gram-positive organisms, and most of the remainder are caused by gram-negative bacteria.23,28,44–46 Coagulase-negative staphylococci are responsible for the greatest number of peritoneal infections. Improvement in hygienic practices and connector tubing technology has resulted in a reduction of S. epidermidis infections. Staphylococcus aureus is the causative organism in about 30% of culture-positive cases of peritonitis in patients receiving CAPD. In approximately half of these cases S. aureus is found at the catheter exit site, suggesting a mechanism for direct migration of bacteria along the catheter tunnel.47 Other grampositive organisms such as streptococci and enterococci are relatively rare causes of peritonitis in this population. Gram-negative pathogens include Pseudomonas, Enterobacter, Escherichia coli, and Klebsiella species.21,23,46,47 Pseudomonas species are isolated from approximately

10% to 20% of catheter exit site infections, further suggesting a mechanism for bacterial migration along the catheter tunnel. Although gram-negative peritonitis in patients receiving CAPD can be caused by a secondary source (e.g., ruptured appendicitis), isolation of a single organism on culture strongly suggests a primary etiology. Fungal peritonitis is a rare but serious complication of CAPD and is associated with substantial morbidity and mortality.48–52 Although fungal pathogens account for only 3% to 6% of all CAPD-associated infections in children, the incidence of fungal peritonitis has been increasing over the last decade.49–52 Many patients who develop fungal peritonitis have had a recent catheter-related infection treated with antibiotics. Fungal peritonitis can be particularly difficult to treat, resulting in a greater likelihood for catheter loss and conversion to hemodialysis compared with bacterial infections.48,50–52 Candida species are the most common cause of fungal peritonitis. Infectious peritonitis is a rare complication in children with indwelling VP shunts, affecting less than 1% of all patients.53 Peritonitis can arise from contamination of intra-abdominal cerebrospinal fluid (CSF) by hematogenous seeding, bacterial translocation, or from contamination of the peritoneal cavity by an infection originating in the central nervous system. Bacterial peritonitis in the presence of a VP shunt must be carefully distinguished from a CSF pseudocyst, which is a much more common etiology for abdominal pain in children with a VP shunt. Early VP shunt-associated peritoneal infections are usually caused by gram-positive cocci (S. epidermidis and S. aureus), whereas late infections are more likely caused by gram-negative organisms.53 Secondary peritonitis following erosion of the catheter into the colon has also been reported in children with indwelling VP shunts.

Secondary Bacterial Peritonitis Secondary peritonitis can arise from any condition that results in the loss of GI tract continuity (see Table 69-1). In the neonatal period, secondary peritonitis most often results from intestinal ischemia and perforation, such as during necrotizing enterocolitis, gastric perforation, meconium ileus, intestinal atresia and Hirschsprung

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Peritonitis

disease.35 Acute appendicitis is the most commonly associated condition leading to secondary peritonitis in older children.54 Other conditions that can lead to perforation and secondary peritonitis in the postneonatal period include volvulus, intussusception, and incarcerated hernia, among others.35 The bacteriologic causes of secondary infections depend on the location of GI perforation. Aerobic gram-negative enteric organisms, including E. coli, Klebsiella, Pseudomonas species, and others, are frequently isolated from perforations of the proximal GI tract. The colon contains predominantly anaerobic organisms, and Bacteroides species is the most commonly isolated organism following colonic perforation. In neonates, culture results from distal intestinal perforations are much more likely to yield enteric gram-negative rods than anaerobes, although Bacteroides and Clostridium species are often isolated.55 Enterobacter species were the most common cause of peritonitis in a series of neonates with perforated necrotizing enterocolitis; anaerobes were isolated from only 6% of cases.55 In contrast to primary bacterial peritonitis, secondary infections tend to be polymicrobial, often involving multiple strains of mixed aerobic and anaerobic organisms.

Clinical Presentation The clinical presentation of a child with peritonitis depends on many factors, including age, underlying cause of infection, associated comorbidities, and immune status. Peritonitis in the neonatal period may occur during the first few days of life with obstructive pathology, or within the first few weeks of life in the case of necrotizing enterocolitis. In either case, the clinical presentation of the neonate can be relatively nonspecific due to a paucity of localizing signs. The neonate generally appears ill, often with marked abdominal distention, hypothermia, emesis, and respiratory distress.35,36,38 The abdominal wall may appear erythematous or edematous when there is severe peritoneal inflammation. Fever is not a sensitive sign of peritonitis in the neonate, being present in fewer than 20% of documented cases.35,36,38 The presentation of older children with peritonitis can include abdominal pain and tenderness, distention, vomiting, hypoactive bowel sounds, and varying degrees of systemic toxicity. Sepsis may be present and evidenced by leukocytosis, fever, and hemodynamic lability. Abdominal pain can be generalized, or in the case of secondary peritonitis, localized to the site of pathology, such as right lowerquadrant tenderness with acute appendicitis. The child may also experience tenderness on rectal or vaginal exam. The clinical presentation in neutropenic children or those receiving high doses of corticosteroids (e.g., NS) may be benign. These children may present without fever and with minimal abdominal tenderness, often leading to a misdiagnosis of gastroenteritis or other less serious ailment. The presence of fever and abdominal pain in any child undergoing CAPD should alert the clinician to the possibility of catheterassociated peritonitis, especially in the context of cloudy dialysis effluent. The clinical presentation of VP shunt-associated peritonitis can be subtle, with the only symptoms being mild abdominal pain and low-grade fever. Peritonitis associated with tuberculosis is often insidious in onset, characterized by weight loss, chronic weakness, fever, night sweats, and anorexia.41 The abdominal pain may be chronic, and children with tuberculous peritonitis generally have less discomfort than those with pyogenic infections. Approximately 40% of patients with tuberculous peritonitis have evidence of active pulmonary tuberculosis.

DIAGNOSIS Laboratory Evaluation Laboratory findings in a child with peritonitis are usually nonspecifically abnormal, often suggesting the presence of a systemic inflammatory process rather than a specific diagnosis. Leukocytosis is

CHAPTER

69

423

variably present and the white blood cell (WBC) count may be depressed in children who are immunocompromised. With untreated or advanced peritonitis, signs of systemic inflammatory response syndrome may become evident with leukopenia, thrombocytopenia, metabolic acidosis, and coagulation disorders. Bacteremia occurs in up to 75% of children with peritonitis.43 Intravascular depletion from thirdspacing of fluid into the bowel wall and lumen as well as the peritoneal cavity can lead to profound dehydration and electrolyte abnormalities. Analysis of peritoneal fluid aspirates should be done when possible and may aid in differentiating primary from secondary peritonitis. The presence of stool, amylase, and bile is indicative of intestinal perforation, while the presence of blood does not differentiate primary from secondary peritonitis. Elevated protein content (> 1 g/dL), decreased glucose concentration (< 50 mg/dL), and elevated lactate dehydrogenase concentration (> 25 mg/dL) are more suggestive of secondary peritonitis, although these are less specific than the presence of stool or bile. In cases of severe secondary peritonitis, WBC counts in excess of 5000 cells/mm3 may be observed. Peritoneal fluid with a WBC > 250 cells/mm3 and a left shift (> 50% polymorphonuclear leokocytes) is suggestive of primary peritonitis. In children undergoing CAPD, the presence of cloudy effluent with WBC > 100 cells/mm3 and PMNs > 50% is adequate for the diagnosis of peritonitis and an indication to begin empiric antibiotics.56,57 Peritoneal fluid from infections caused by M. tuberculosis is characterized by a lymphocytic predominance (> 50%), high protein content (> 2.5 g/dL), and decreased glucose levels (< 30 mg/dL).58,59 Peritoneal fluid aspirates should be sent for Gram stain and culture in all cases where the diagnosis of peritonitis is suspected. Peritoneal cultures are negative in up to 20% of documented cases and this finding should not exclude the diagnosis.43 Aspirates should be routinely cultured for aerobic and anaerobic organisms, and additional cultures should be sent for acid-fast bacteria and fungal pathogens in the appropriate clinical situation. Culture results that reveal only gram-positive cocci are suggestive of primary peritonitis (especially in patients on CAPD), whereas polymicrobial infections with anaerobic organisms are more suggestive of secondary infections. Gram-negative pathogens (especially E. coli) can be seen in both primary and secondary peritonitis. If tuberculous peritonitis is suspected but acid-fast stains and culture are unrevealing, image-guided peritoneal biopsies to identify M. tuberculosis and granulomas may be helpful to rule out the diagnosis.58,60 Tuberculin skin testing should also be done in patients suspected of mycobacateria exposure.

Imaging Studies The primary utility of imaging in patients with suspected peritonitis is to help determine the etiology (i.e., primary versus secondary causes) and the necessity of laparotomy. Plain films of the abdomen can demonstrate a number of nonspecific findings, including abnormal bowel gas pattern, adynamic ileus, and presence of ascites. Other findings may be more specific, such as loss of peritoneal fat lines and psoas shadow with acute appendicitis. The presence of free air is strongly suggestive of a perforation of a hollow viscus and mandates urgent exploratory laparotomy. Ultrasonograpy provides a useful adjunct to plain films. Abdominal sonography can accurately identify the presence of ascites and the presence of abdominal abscesses and other masses, and may be useful in directing peritoneal aspirations for the purpose of diagnostic culture. Computed tomography (CT) is the most useful imaging modality in the evaluation of the child with an acute abdomen. When oral and intravenous contrast agents are used appropriately, CT scanning provides the most sensitive means for differentiating primary from secondary infections and the need for laparotomy. CT can also delineate the source of secondary infections in many cases, particularly where there is a localized inflammatory process (e.g., acute appendicitis). The finding of complex ascites, with or without visualization of the appendix, is strongly suggestive of perforated appendicitis in

424

SECTION

I

Intra-Abdominal Infections

older children. In the case of primary or catheter-related peritonitis, typical imaging findings include simple ascites, high-attenuation peritoneal fluid, and nodular irregularity of the peritoneal lining.61–63 Nonspecific inflammatory changes involving the mesentery, omentum, and serosal surfaces also occur with primary and catheter-related infections. CT findings in patients with abdominal tuberculosis include lymphadenopathy, splenomegaly, and focal mesenteric calcifications.59

Cloudy dialysis effluent

Send effluent for analysis: Gram stain and culture Cell count and differential

Begin empiric antibiotic therapy

MANAGEMENT Successful management of bacterial peritonitis involves early recognition, aggressive fluid resuscitation and correction of electrolyte abnormalities, and initiation of appropriate antimicrobial therapy. Differentiating between primary or secondary sources of peritonitis is critical as prompt surgical intervention is necessary to control the infection with perforation of a hollow viscus. Surgical consultation should be sought early to determine the need for exploratory laparotomy (or laparoscopy) if the diagnosis of secondary peritonitis is suspected. The choice of empiric antibiotics is based on the most likely source of contamination (primary, secondary, or catheter-associated peritonitis). Management of VP shunt infections related to a primary central nervous system source is addressed in Chapter 102, Clinical Syndromes of Device-Associated Infections.

Primary Peritonitis In a child with an anatomically normal peritoneal cavity who develops primary bacterial peritonitis, antibiotic treatment should be directed against gram-positive cocci and gram-negative rods (see Table 69-2). Both gram-positive and gram-negative organisms have been implicated as causative organisms in children with NS, whereas enteric gram-negative infections are much more common in children with cirrhosis.17,33,34 Empiric coverage should include a third- or fourth-generation cephalosporin or a broad-spectrum penicillin or ampicillin plus an aminoglycoside. Antibiotic coverage can subsequently be narrowed based on culture results from peritoneal aspirates. Aggressive removal of infected peritoneal fluid has not been shown to improve outcomes and is therefore not recommended.

Catheter-Associated Infections Empiric antibiotics should be initiated as soon as the diagnosis of bacterial peritonitis is suspected (Figure 69-1). The wide variability of organisms associated with peritonitis in children treated with CAPD warrants the empiric use of broad-spectrum agents. In patients presenting with abdominal discomfort and a cloudy dialysis effluent but without evidence of severe infection (fever and severe abdominal pain), the combined intraperitoneal administration of ceftazidime plus a first-generation cephalosporin is recommended by some experts.20,56,64,65 In patients presenting with frank peritonitis or systemic signs of infection, vancomycin should be substituted for the first-generation cephalosporin and concomitant intravenously administered antibiotics considered. A glycopeptide is also recommended for children with a recent history of methicillin-resistant Staphylococcus aureus infection, recent catheter exit site infection, or known nasal colonization with S. aureus. Further tailoring of antibiotics should be done once Gram stain and culture results are available (Figure 69-2). If Gram stain and culture reveal only gram-positive organisms, the third-generation cephalosporin (ceftazidime) can be discontinued. A first-generation cephalosporin should be used for methicillin-susceptible S. aureus, and clindamycin, vancomycin, or linezolid could be used for resistant strains. Ampicillin is sufficient to treat most strains of enteorocci and streprococci. Antibiotic treatment for gram-positive organisms should continue for 2 to 3 weeks.

No risk factors for complicated infection

Risk factors for complicated infection: Recent MRSA infection Nasal carrier of MRSA Peritoneal signs Fever Age less than two years

1st generation cephalosporin + ceftazidime

Vancomycin + ceftazidime

Figure 69-1. Initial approach in the management of a child with suspected catheter-associated peritonitis. MRSA, methicillin-resistant Staphylococcus aureus

If culture and sensitivity results identify only gram-negative organisms susceptible to ceftazidime, gram-positive coverage (firstgeneration cephalosporin) may be discontinued. If cultures identify anaerobic organisms or multiple strains of gram-negative rods, anaerobic coverage should be added and an intra-abdominal source for secondary peritonitis sought. Antibiotic treatment should be continued for approximately 2 weeks in the case of nonpseudomonal infections, and about 3 weeks when Pseudomonas species are cultured. If the patient is improving on empiric therapy and the initial cultures remain unrevealing, the initial combination of antibiotics should be continued for a total of 2 weeks. Fungal peritonitis could be treated with either intravenous amphotericin B or a combination of an azole agent (e.g., fluconazole) and flucytosine.48 Although it may be possible to treat the fungal infection without immediate removal of the dialysis catheter, the catheter should be removed if no improvement is seen after the first 72 hours of therapy. If clinical improvement is observed and the dialysis catheter is left in place, antifungal treatment should be continued for at least 4 to 6 weeks following complete resolution of clinical symptoms and sterilization of fluid. If no improvement is seen, the catheter should be removed and treatment continued for at least 2 weeks following resolution of clinical symptoms.48,66 Improvement in the clinical manifestations of infectious peritonitis should be observed within 72 hours of initiating antibiotic therapy. This may be evidenced by reduction in fever and abdominal pain, and antibiotic-decreased cloudiness of the peritoneal effluent. Subsequent aspirates of dialyate fluid in a patient responding to antibiotics should reveal a decrease in WBC counts by at least 50%. If no clinical improvement is observed within 72 hours, modification of antibiotic coverage and removal of the dialysis catheter should be considered.56,66 With the exception of cases involving Pseudomonas species, removal and replacement of the infected catheter can be performed simultaneously once the patient has clinically responded to antibiotics and effluent leukocyte counts are less than 100/mm3. Antibiotics should be continued for 3 weeks following catheter replacement.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Peritonitis

CHAPTER

69

425

Organism identified on culture

Gram-positive organism

Gram-negative organism

Discontinue empiric gram-negative coverage

Discontinue empiric gram-positive coverage

Enterococcus Streptococcus

MRSA

Gram-positive non-MRSA

Pseudomonas species

Non-pseudomonal Gram-negative bacillus

Anaerobes or multiple strains of gramnegative bacilli

Substitute ampicillin for empiric therapy (treat for 2 weeks)

Substitute clindamycin or vancomycin for empiric therapy (treat for 3 weeks)

No further modifications (treat for 2 weeks)

Add second agent active against Pseudomonas (treat for 3 weeks)

No further modifications (treat for 2 weeks)

Add metronidazole and search for sources of secondary peritonitis

Figure 69-2. Algorithm for the modification of empiric antibiotic therapy for continuous abdominal peritoneal dialysis-associated peritonitis based on culture results. MRSA, methicillin-resistant Staphylococcus aureus.

Relapsing peritonitis is defined as the recurrence of peritonitis with the same organism within 4 weeks of completing antibiotic treatment. Relapsing peritonitis occurs in up to 20% of episodes.67 Coagulasenegative staphylococci can survive antibacterial treatment within a biofilm secreted on the catheter surface, and P. aeruginosa can prove difficult to treat due to the formation of microabscesses within the catheter tunnel. Empiric treatment for a patient presenting with recurrent symptoms is the same as for the initial infection. The use of fibrinolytic agents and high-dose intraperitoneal antibiotics for catheter decontamination may be useful in cases where relapsing infection is attributed to coagulase-negative staphylococci.68,69 Catheter removal is not necessary after the first episode of recurrent peritonitis, although removal should be strongly considered with the second relapsing infection and in all cases due to Pseudomonas species. Catheter removal is also suggested when the source of reinfection is thought to originate from within the soft-tissue tunnel. Relapsing peritonitis should be treated with intraperitoneal antibiotics for at least 3 weeks.

Secondary Peritonitis Patients with secondary peritonitis require urgent laparotomy to identify and control the source of peritoneal contamination. All infectious debris should be removed and necrotic tissue debrided. Indwelling foreign bodies should be removed. Intraoperative peritoneal irrigation with saline has been shown to decrease the bacterial burden and reduce postoperative infectious complications; irrigation with antibiotic solutions does not confer any additional benefit and is avoided. Inadequate control of the source of infection at initial laparotomy leads to recurrent peritonitis and is associated with increased mortality and morbidity. This most commonly occurs when there has been massive contamination (fecal peritonitis) or when intestinal viability is questionable at abdominal closure (necrotizing enterocolitis). Mandatory re-exploration 24 to 48 hours after the initial laparotomy (“second-look” operation) is practiced by many and has been shown to decrease morbidity and mortality compared with selective re-exploration.70 Empiric antibiotic treatment for secondary peritonitis should be directed against enteric pathogens, including anaerobic bacteria and gram-negative bacilli. Treatment of secondary peritonitis with the

three-drug combination of ampicillin, gentamicin, and metronidazole has long been the gold standard, but does not confer any additional benefit over the broad-spectrum monotherapeutic agents (pipercillintazobactam). Clindamycin plus a second-generation cephalosporin or plus ampicillin and gentamicin is used less commonly because of increasing resistance of Bacteroides fragilis in many areas. There are no consensus guidelines for the length of antibiotic treatment once the source of infection has been controlled with operative management, although 7 to 10 days of therapy following complete resolution of clinical and laboratory indicators of infection is typical.

COMPLICATIONS AND PROGNOSIS Infectious peritonitis is associated with many life-threatening complications, including mesenteric vein thrombosis, adult respiratory distress syndrome, progressive multiorgan failure, and death. Severe complications are most often associated with secondary peritonitis. Other complications include prolonged ileus, surgical wound infection, intra-abdominal abscess, enteric fistula, and the development of inflammatory adhesions. Inflammatory thickening of peritoneal surfaces and compartmentalization of the peritoneal cavity by infectious debris can limit the effectiveness of peritoneal dialysis. Short-term hemodialysis may be required in these patients until the infection is controlled. The prognosis of children with peritonitis depends on the etiology of the infection, associated comorbidities, and the host response to the inflammatory process. Mortality rates for nonsecondary cases of peritonitis in children are close to 10%, but vary depending on etiology and associated comorbidities.71 Contemporary estimates of casefatality rates for patients receiving chronic peritoneal dialysis are approximately 6%, with peritoneal infections comprising over 15% of all-cause mortality and 69% of infection-related mortality.71 Fungal peritonitis carries a particularly poor prognosis in children, with some series reporting case-mortality rates exceeding 25%.71 Estimates of case-fatality rates for SBP in children with NS are around 9%.10 Spontaneous peritonitis in a child with cirrhosis and portal hypertension carries a much worse prognosis, with mortality rates exceeding 30% to 40% at 1 month following diagnosis.72 The observation that overall survival at 6 months following the first episode of

426

SECTION

I

Intra-Abdominal Infections

SBP is less than 30% suggests that the diagnosis may be a marker of rapidly progressing hepatic dysfunction. This has led some to consider the first episode of SBP in these children as a relative indication for liver transplantation. Mortality rates associated with primary and catheter-related peritonitis are generally lower than that seen with secondary peritonitis. Prior to the era of broad-spectrum antibiotic use and aggressive surgical management, secondary peritonitis from all causes carried a mortality rate in excess of 60% to 70%. Recent estimates of mortality vary depending on the etiology of the perforation, age of the child, and associated comorbidities. Mortality in healthy children who develop perforated appendicitis is rare, whereas mortality associated with necrotizing enterocolitis in low-birthweight neonates exceeds 30% to 50%.73–75

PREVENTION Strategies to prevent bacterial peritonitis in children have largely focused on patients undergoing peritoneal dialysis. Observational studies conducted from registry data and a few controlled trials identified decreased rates of peritonitis and exit-site infections with specific technical factors during catheter placement.76 These include the use of a double-cuffed dacron catheter (versus a single-cuffed model) and directing the peritoneal opening inferiorly (versus cephalad).77–79 However, a systematic review of these data by the Cochrane Collaboration failed to demonstrate a significant benefit with any specific catheter design or insertion technique with respect to the incidence of peritonitis, tunnel site infections, or need for catheter removal.80 The use of strict sterile technique during the dialysis process is crucial to prevent contamination-associated peritonitis in children treated with CAPD. The competency of CAPD training appears to be critical in this regard, as a study in 50 dialysis units suggested a strong correlation between the duration of training and incidence of peritonitis.81 The lowest incidence of CAPD-associated peritonitis has been observed in Japan, where parents and technical personnel typically receive 6 to 7 weeks of intensive training on sterile technique and CAPD protocol.82 Prophylactic antibiotic therapy with first-generation cephalosporin significantly decreases the incidence of postoperative peritonitis; a single dose should be given immediately prior to catheter insertion.83 Children with a recent history of methicillin-resistant S. aureus or infection colonization should receive vancomycin. Nasal carriers of S. aureus (methicillin-susceptible) have been identified in up to 45% of families of children treated with CAPD.84 The use of mupirocin applied to the nares or catheter exit site in patients who are known carriers and their family members has been shown to decrease the incidence of exit site infections and peritonitis in controlled clinical trials.84 One to 2 days of a first-generation cephalosporin is also recommended when there is suspected contamination of the dialysate or connecting hardware, and prior to procedures involving the GI or genitourinary tract.85 Long-term prophylactic systemic or intraperitoneal antibiotics is not indicated.

CHAPTER

70

in the United States.1 Although appendicitis is no longer responsible for significant mortality, morbidity associated with appendicitis, particularly complicated appendicitis, can be substantial. Recent efforts to decrease morbidity have focused on improving methods of diagnosis, optimizing surgical management, and improving postoperative care.

EPIDEMIOLOGY Approximately 80 000 children < 18 years of age suffer from appendicitis each year. Appendicitis can affect children of all ages, including infants < 1 year of age; however, the peak incidence is among 12- to 18-year-old adolescents. There is a slight male predominance with an 8.7% lifetime risk for boys compared with 6.7% risk for girls.2 The incidence increases from 1 to 2 cases per 10 000 in children < 4 years of age to 25 cases per 10 000 in older children.2 There may be a small seasonal predominance during summer months. Epidemiologic factors that may increase the risk for appendicitis include a low-fiber diet3 and possibly a family history of appendicitis.3 About one-third of patients with proven appendicitis have a positive family history compared with 14% of children with right lower quadrant pain due to other causes.4 The rate of perforation at the time of operation ranges from 20% to 76%.5 The highest incidence of perforation is in infancy, occurring in 70% to 95% of cases.6 The incidence of perforation decreases with increasing age, from 70% to 90% in 1- to 4-year-old children to 10% to 20% in adolescents.3,6,7 Perforation rates have also been shown to be increased among minorities and economically disadvantaged children.8–10 Perforated appendicitis is associated with a more severe clinical illness, high morbidity, and a longer hospital course than nonperforated disease. Surgery performed more than 36 hours after the onset of symptoms is associated with an increased risk of perforation compared with surgery performed earlier.3,11,12 This delay usually results from a delay in coming to medical attention or from difficulties in establishing a diagnosis.13,14

PATHOGENESIS Histologic examination of tissue removed has facilitated an understanding of the pathogenesis of appendicitis.15,16 An initial event leads to occlusion of the lumen of the appendix between the cecal base and the tip. This can be caused by true occlusion of the lumen by a foreign body such as an appendicolith, occlusion by inflamed mucosa due to an infectious process, or by hyperplasia of intramural lymphoid tissue. Luminal occlusion is followed by increased intraluminal pressure leading to swelling, congestion, and distention of the mucosa and appendiceal wall. Translocation of normal bowel flora may then occur, leading to an acute inflammatory infiltrate in the wall of the appendix. Progression of the infection leads to increased bowel wall cellulitis and intraluminal pressure. When full-thickness inflammation occurs, localized peritonitis ensues. If untreated necrosis and frank gangrene of the appendiceal wall occur, rupture leads to spread of peritonitis and the development of either an inflammatory phlegmon or an intraabdominal abscess. If this process is not recognized and treated, septicemia may ensue.

ETIOLOGY

Appendicitis Marion C.W. Henry and R. Lawrence Moss

Appendicitis is the most common reason for emergent abdominal surgery in children, accounting for more than one-third of hospitalizations per year for abdominal pain in people under the age of 18

The organisms responsible for appendicitis and subsequent intraabdominal infections are organisms that colonize the bowel. Historically, cultures of peritoneal fluid were routinely obtained at the time of operation. However, with improved antibiotic therapeutic options several studies have suggested that obtaining cultures may not be cost-effective as results infrequently lead to changes in management.17,18 In most series, the most commonly isolated organisms are anaerobic bacteria, including Bacteroides fragilis, Clostridium and

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Appendicitis

Peptostreptococcus species and gram-negative bacilli, including Escherischia coli, Pseudomonas, Enterobacter, or Klebsiella species.

CLINICAL MANIFESTATIONS Although the classically described symptoms of appendicitis are periumbilical pain that migrates to the right lower quadrant, followed by nausea, occasional vomiting, and low-grade fever, the clinical presentation can be highly variable depending on the location of the appendix, the age of the patient, and the host response to infection (Table 70-1). In children < 2 years of age, the most common symptoms of appendicitis are vomiting, abdominal pain, fever, abdominal distention, diarrhea, irritability, right hip pain, and limp or refusal to bear weight.14 In children 2 to 5 years of age, abdominal pain precedes vomiting and is usually associated with fever and anorexia. Diarrhea is also common in infants and toddlers with appendicitis and may lead to misdiagnosis.6 School-aged children describe abdominal pain that is constant and worse with movement or coughing.14 Nausea, vomiting, anorexia, diarrhea, and dysuria are also reported in this age group.14 However, older children sometimes report that they are hungry.1 Two retrospective studies focused on children19,20 and one prospective study that included children and adults21 have evaluated the sensitivity and speciÀcity of vomiting, anorexia, pain with cough and movement, and pain migration in patients with appendicitis. Not surprisingly, none of these symptoms reliably predicts the presence or absence of appendicitis; neither sensitivity nor speciÀcity exceeds 80%.22 A scoring system for appendicitis was developed from a large cohort of mainly adult patients. This system is called the Alvarado scoring system and is given the acronym MANTRELS based on its components (Table 70-2). Its use has been tested in children, and though some authors found the scores to be accurate,23 most studies show only moderate sensitivity (76% to 90%) and speciÀcity (50% to 81%) when the score is 7 or higher. In older children (> 16 years of age), a score of 7 yields high sensitivity (100%) and speciÀcity (93%).22,24 Overall, the Alvarado scoring system seems more useful for risk-stratifying children than for guiding clinical management. Physical Àndings also vary by the patient’s age. In children < 2 years of age, nonspeciÀc signs such as fever and diffuse tenderness are the most common Àndings. Preschool children 2 to 5 years of age, however, demonstrate right lower quadrant tenderness, fever and involuntary guarding. School-aged children are more likely to have localized right lower quadrant tenderness, or diffuse guarding and rebound tenderness in cases of perforated appendicitis.22 Two studies25,26 have evaluated the reliability of rebound tenderness in children, observing low sensitivity (50%) and speciÀcity (60%). Common signs equated with appendicitis, such as Rovsing sign, obturator sign, and psoas sign, have not been critically evaluated.22

CHAPTER

DIAGNOSIS Appendicitis can mimic many other diseases. Other inflammatory diseases, most notably gastroenteritis, can produce signs and symptoms similar to appendicitis (Box 70-1).1,27 In girls, ovarian pathology such as ruptured ovarian cyst and ovarian torsion can cause right lower quadrant pain and peritoneal irritation. In neutropenic patients, differentiating appendicitis from typhlitis or neutropenic enterocolitis can be difÀcult and imaging studies are often necessary.1 Mesenteric lymphadenitis is probably a frequent cause of abdominal pain in children who are not pursued for appendicitis, or may be found at the time of surgery or by imaging studies when appendicitis cannot be excluded (see Chapter 21, Abdominal and Retroperitoneal Lymphadenopathy).

LABORATORY STUDIES Laboratory tests are often used in conjunction with the history and physical examination in order to establish a diagnosis in children seeking medical attention because of abdominal pain. The most commonly ordered test is the white blood cell (WBC) count. In patients with appendicitis, the WBC count is typically in the range of 1,100 to 16,000/mm3. Very high WBC count suggests perforation or a different diagnosis. In a review of studies examining the utility of the WBC count variable sensitivity (19% to 88%) and speciÀcity (53% to 100%) were noted.22 Serum C-reactive protein level has been examined as a potentially useful test in diagnosing appendicitis. However, studies have found that sensitivity ranges from 48% to 75% and speciÀcity from 57% to 82%.22 While in adults the combination of a normal WBC count and normal C-reactive protein level makes the diagnosis of appendicitis unlikely, this is not true in children.28 Serum procalcitonin level also has been studied as a marker for appendicitis. In a small study of 212 children, the sensitivity of a procalcitonin level > 0.5 ng/mL for gangrene or perforation was 73%,

TABLE 70-2. MANTRELS Score for Acute Appendicitis Symptoms

Signs Investigation Total possible score

Migration of abdominal pain from the epigastrium to the right lower quadrant Anorexia Nausea/vomiting Right lower quadrant tenderness Rebound tenderness Elevated temperature Leukocytosis Left shift of neutrophils

1 1 2 1 1 2 1

Appendicitis unlikely Appendicitis possible Appendicitis likely

6

TABLE 70-1. Clinical Signs or Symptoms in Appendicitis by Age of Patient86

BOX 70-1. Differential Diagnosis of Acute Abdominal Pain Age of Children with Finding (%)

Clinical Sign or Symptom

≥2 years

2–5 years

6–12 years

Abdominal pain Right lower quadrant pain Diffuse tenderness Vomiting Fever Anorexia

35–77 < 50 55–97 85–90 40–60 NR

89–100 58–85 19–28 66–100 80–87 53–60

100 > 90 15a, 83b 68–95 64 47–75

NR, not reported. a Without perforation. b With perforation.

427

70

Infectious gastroenteritis Pneumonia Urinary tract infection Mesenteric lymphadenitis Intussusception Inflammatory bowel disease Meckel diverticulum Hernia Primary peritonitis Orchitis Testicular torsion Blunt abdominal trauma Ovarian cyst

1

428

SECTION

I

Intra-Abdominal Infections

with a specificity of 95%.29 However, in the majority of children with simple acute appendicitis, the procalcitonin level was < 0.5 ng/mL.29 Pyuria and hematuria have been reported in up to 30% of children with appendicitis.30–32 Abnormal urinalysis is most commonly observed when the inflamed appendix is adjacent to the ureter or bladder.

of CT in the evaluation of children with suspected appendicitis ranged from 18% to 89% in a study conducted across 30 children’s hospitals. In one study, an approach based on selective imaging for equivocal cases, starting with US first, followed by CT for inconclusive studies, has been shown to be accurate and cost-effective,47,48 with sensitivity and specificity of 99% and 92%, respectively, and accuracy of 97%.49

IMAGING STUDIES Radiologic studies can be a valuable adjunct to history and physical examination in the diagnosis of appendicitis. In the past, plain abdominal radiograph was the most common imaging study performed in patients with suspected appendicitis. However, this study has proven to be of little value. Plain abdominal radiograph has little merit. In fact, one study found the radiograph to be normal or misleading in 82% of children with appendicitis.14 Plain film may demonstrate a fecalith in about 5% of cases of perforated appendicitis.1 The presence of a fecalith on plain film in a patient with abdominal pain is highly suggestive of appendicitis; however, absence has no meaning.33 Ultrasonography (US) has been studied extensively as a tool for diagnosing appendicitis. The diagnostic findings observable on ultrasound include appendiceal diameter > 6 mm, a “target sign” with five concentric layers, distention or obstruction of the appendiceal lumen, high echogenicity surrounding the appendix, an appendicalith, fluid surrounding the appendix, enlarged and thickened bowel wall, and a lack of peristalsis.22 Sensitivity of US ranges from 71% to 92% and specificity from 96% to 98%.34–37 However, the test is both operator- and subject-dependent. Obesity and gaseous distention of bowel loops limit utility of US. In up to 30% of cases the appendix is not visualized, and thus the study is inconclusive.22 In general, ultrasound can be used to confirm but not to exclude the diagnosis of appendicitis.22 Many centers recommend US as the first-line study for children with suspected appendicitis due to its low cost, freedom from ionizing radiation, and superiority to clinical judgment in equivocal cases.38–41 Computed tomography (CT) is widely used in the diagnosis of many abdominal conditions and has been proposed as a diagnostic tool for appendicitis. A CT is considered positive for appendicitis if there is an enlarged appendiceal diameter, appendiceal wall thickening, or periappendiceal inflammatory changes, including fat streaks, edematous phlegmon, fluid collection, and/or extraluminal gas.22 In adults, CT has been shown to have higher specificity and sensitivity than ultrasound.42 Studies in children have found CT to be both sensitive (range 94% to 99%) and specific (range 87% to 99%) for the diagnosis of appendicitis.36,42,43 Several of these studies were prospective, but were not blinded. In many studies, the population included large numbers of patients in whom the diagnosis of appendicitis was clinically apparent and an operation was already planned. Despite the reported high sensitivity and specificity, CT has several limitations: expense, exposure to radiation and contrast material, and requirement for sedation in some children. Some studies have found that CT offers no increased accuracy over simple history, physical examination, and laboratory analysis.44,45 Diagnostic accuracy of clinical examination alone can also be improved by repeated examinations and laboratory testing in 12 to 24 hours. Observation with serial examinations can be a cost-effective method of diagnosing appendicitis versus other causes of abdominal pain; fewer than 2% of patients perforate during observation.46 Given the accuracy of imaging studies, some practitioners obtain studies in all children with abdominal pain to reduce negative laparotomy rates. However, the accuracy of physical examination and clinical judgment varies, just as the accuracy of imaging studies varies. Imaging studies result in additional costs and delayed treatment, and can be both falsely positive and falsely negative. Indeed, the falsenegative rate of a study is increased if the study is used in a child with a high clinical suspicion of disease, and the false-positive rate is increased if the study is used in a child with a low clinical suspicion of disease.47 Therefore, imaging studies are most useful when used selectively in children with clinically equivocal presentations. The use

MANAGEMENT The management of appendicitis has gradually improved as new technologies and medications have been introduced. Although some aspects of care have become standardized, great variability remains in the care of children with appendicitis.5,50 Even within a single state, there may be major regional variations in the care and outcomes of children with appendicitis.51 In cases of acute, noncomplicated, appendicitis, prompt appendectomy continues to be the mainstay of treatment. Delaying appendectomy until antibiotic treatment has been administered for 12 hours does not appear to increase perforation rates or clinical morbidity in comparison to immediate appendectomy.52,53 Laparoscopic and open techniques for appendectomy are both widely utilized, with no clear outcome advantage for either. Some studies have suggested that laparoscopic procedures lead to a shorter hospital stay, less pain, and better cosmetic result. A randomized trial in Europe compared the costs of laparoscopic and open appendectomy and found only a marginal difference in the cost of the two different procedures but a quicker return to activities in children who underwent laparoscopic appendectomy.54 Another randomized trial found that children undergoing laparoscopy had less pain and shorter duration of hospitalization compared to those undergoing open appendectomy. Although operating times and costs were slightly increased with laparoscopy, there was an overall reduction in costs because of the decreased length of hospital stay.55 A review of 45 published trials in adults and children of laparoscopic versus open appendectomy showed that laparoscopy decreased the wound infection rate by about 50% but increased the rate of intra-abdominal abscesses fourfold.56 For cases of perforated appendicitis, some studies have suggested an increase rate of intra-abdominal abscess development after laparoscopy.57–61 A large study of four centers treating children with perforated appendicitis, however, found no difference in the rate of abscess development between those treated by laparoscopic or open appendectomy.62 Preoperative antibiotics have long been considered key to the prevention of postoperative infections. While several studies have challenged the use of routine preoperative antibiotics, these studies were limited by their small numbers of patients.63–66 One prospective, randomized study found no difference in complication rates of those patients with simple appendicitis who received antibiotics and those who did not. However, there was a higher rate of postoperative infection observed among those with gangrenous appendicitis at operation if they did not receive preoperative antibiotics.67 In simple appendicitis, postoperative antibiotics are often used for 24 hours or not used at all. Many institutions have adopted clinical algorithms for the management of appendicitis. The majority of the data to support these algorithms comes from retrospective reviews of pediatric populations. However, the studies examining the benefit of these guidelines have generally noted a reduction in cost and length of hospitalization for patients treated by guidelines.68 It is difficult to know, however, what effect just having a guideline and standardization of care may have in decreasing complication rates in these patients.69

COMPLICATED APPENDICITIS Most of the controversies and the majority of the variability of care of children with appendicitis involves issues related to the management of patients with complicated appendicitis. The controversy begins with the very definition of complicated appendicitis as some surgeons

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Intra-abdominal, Visceral, and Retroperitoneal Abscesses

argue that gangrenous and perforated appendicitis cannot be differentiated. This distinction is important, as gangrenous appendicitis is associated with outcomes and morbidity rates that are consistent with simple appendicitis, whereas those of perforated appendicitis are much higher. The use of nonoperative management for children with perforated appendicitis is controversial and still under investigation. If peritonitis is present, many surgeons advocate immediate appendectomy. However, some have advocated the treatment of these patients with intravenous antibiotics followed by interval appendectomy.70–72 Surgeons in favor of this approach cite the high rate of complications when operating during a period of intense inflammation and peritonitis. However, high failure rates have been reported. If a child does not respond in 24 to 72 hours, then an appendectomy should be performed. One study has shown an 84% failure rate in those patients with > 15% bands.72 The patients who do not respond to nonoperative treatment and undergo delayed appendectomy often have high complication rates and prolonged hospitalizations.71,73 Some proponents of nonoperative management have started to question the necessity of performing an interval appendectomy. These surgeons note an insignificant rate of recurrence74 and argue that, after perforation, the lumen of the appendix is obliterated. A histopathologic analysis of a small number of interval appendectomy specimens, however, found that all of the specimens had patent lumens and the presence of their tips.16 A survey of members of the American Pediatric Surgical Association found that 86% of surgeons perform an interval appendectomy.50 In a survey of pediatric surgeons, postoperative antibiotics were routinely used.50 However there was substantial variation in the type of antibiotics prescribed and the duration of treatment. Antibiotic regimens commonly include ampicillin, gentamicin, and clindamycin or metronidazole. Monotherapy with broad-spectrum antibiotics such as piperacillin-tazobactam may be equally effective.75 In one randomized trial, there was no difference in infectious complications between those patients who were treated with a set course of a minimum of 5 days of antibiotics and those treated with a course based on clinical factors with no set minimum.76 In a large review of the subject, there was no difference in infectious complication rates in those treated with only 3 days of antibiotics compared to those treated longer than 3 days.77 Treating with antibiotics until a child has been afebrile for 24 hours and the WBC count has returned to normal has been shown to be effective in a prospective study.78 Studies have demonstrated that it is safe and cost-effective to treat children with complicated appendicitis with outpatient parenteral antibiotics79 or oral antibiotics after a course of intravenous antibiotics.80,81 Traditionally it has been thought that the natural history of appendiceal rupture was within the control of the physician and that a high rupture rate reflected a failure of care. In order to decrease the rupture rate, early surgical intervention has been the gold standard and high rates of negative exploration have been acceptable, in order to decrease the likelihood of rupture. However, despite efforts to decrease the rates of complicated appendicitis, rates of perforation remain high. ranging from 30% to 74%.82 These high rupture rates may not actually be related to the medical care provided but are due to delay in diagnosis and treatment because of inadequate access to medical care. Several studies have linked race with an increased risk of perforation.82,83 In a review of a national pediatric discharge database, the likelihood of perforation differed by race, even while controlling for age and health insurance status.83 In another review of a large pediatric health systems database, the rate of rupture in schoolaged children was associated with race and insurance status and not with negative appendectomy rate.82,84,85 Examination of a statewide database also revealed that children with Medicaid or no insurance were more likely to develop appendiceal perforation than those children with private insurance.84 These findings suggest that efforts on focusing improved access to healthcare would be more beneficial in reducing rates of complicated appendicitis than altering hospital management.

CHAPTER

CHAPTER

71

429

71

Intra-abdominal, Visceral, and Retroperitoneal Abscesses Karen A. Diefenbach and R. Lawrence Moss

INTRODUCTION Abscesses in children are not infrequent occurrences. When they occur, they are sometimes difficult to diagnose and even more difficult to treat. Even when managed appropriately, morbidity can be high. If not managed appropriately, the results can be devastating. Abdominal abscesses are categorized into three types, with the first being the most common: intra-abdominal (intraperitoneal), visceral, and retroperitoneal abscesses. Bacteria from the original site of colonization or infection can travel hematogenously, lymphatically, or directly to form an abscess. Sample from the abscess may or may not show bacteria on Gram stain or culture. In intra-abdominal abscesses, there is frequently a mixed flora of facultative and anaerobic bacteria, but in visceral and retroperitoneal abscesses, single organisms are more common. The types of bacteria isolated can indicate the origin of the infection. The most common facultative bacteria isolated in intra-abdominal abscesses are Escherichia coli, Staphylococcus aureus, and Enterococcus spp.1,2 Anaerobic bacteria include Bacteroides fragilis group, Peptostreptococcus spp., Clostridium spp., and Fusobacterium spp.1,2 These enteric organisms might also cause visceral abscesses, but this type of abscess can also be caused by the gram-positive bacteria of a systemic infection or by fungi in immune-compromised hosts. Retroperitoneal abscesses contain organisms specific to the site of the primary infection.3 For example, a perinephric abscess is usually caused by organisms that cause pyelonephritis, whereas an iliopsoas abscess (IPA) is usually caused by bacteremic S. aureus.3–5 The tenets of management of an abscess typically include drainage, appropriate antibiotic therapy, and correction, if possible, of any underlying pathology that can cause recurrence. Drainage can be accomplished in a variety of ways, including open surgical exploration, laparoscopic surgical exploration, and percutaneous, imageguided drainage. Empiric antibiotic therapy should be initiated immediately upon diagnosis considering likely abscesses by site (Table 71-1), and changed, if necessary, to definitive therapy when culture and susceptibility test results become available. Coordinated management by the surgical, interventional radiology, and infectious

TABLE 71-1. Summary of Common Organisms in Abscesses by Site1–3,7,24–29,33,36 Site

Organisms

Intra-abdominal

Escherichia coli, Enterococcus spp., Bacteroides spp., Peptostreptococcus spp., Clostridium spp., Fusobacterium spp., Staphylococcus aureus, Klebsiella spp., and Pseudomonas spp.

Visceral (liver, spleen, and pancreas)

Staphylococcus aureus, Escherichia coli, Enterococcus faecalis, Klebsiella spp., Enterobacter spp., Pseudomonas spp., Streptococcus spp., Salmonella, Bartonella, and Candida spp.

Retroperitoneal Staphylococcus aureus, Streptococcus spp., (iliopsoas, other) Escherichia coli, Klebsiella spp., and Candida spp.

430

SECTION

I

Intra-Abdominal Infections

disease teams, in consultation with the child’s primary care physician, is needed to deliver care in a timely and efficient manner.

INTRA-ABDOMINAL ABSCESSES Etiology and Clinical Manifestations Complicated appendicitis is by far the most common cause of intraabdominal abscesses in children. Patients can come to medical attention with an abscess after perforation of the appendix or can develop an abscess postoperatively. Several other potential causes of intra-abdominal abscesses originate from the gastrointestinal tract, but rates are much lower. Any operation on the GI tract, whether elective or emergent, has the potential for causing a postoperative abscess. Elective operation on the bowel carries the risk of contamination of the peritoneum at the time of surgery or later from a complication such as an anastomotic leak. Emergent surgery for perforation is performed in the setting of contamination and has an even higher risk. Other emergent causes of a perforated viscus and resultant abscess include necrotizing enterocolitis, inflammatory bowel disease, peptic ulcer, Meckel diverticulum, and trauma.1 Location of an abscess within the abdominal cavity, such as subphrenic and pelvic sites, can predict an underlying disease and vice versa, as can the isolation of certain pathogens. The bacteria most frequently isolated in abscesses after bowel disease or surgery include Escherichia coli, Bacteroides species, and Enterococcus species, but any enteric organism can be present.2 Brook studied the microbiology of intra-abdominal abscesses in children and found that 8% of cultures grew facultative bacteria only, 17% grew anaerobic bacteria only, and 75% grew mixed flora.1,6 A postoperative patient with persistent, spiking fevers and abdominal pain must be evaluated for the possibility of an abscess. Symptoms from compression of the abscess on adjacent structures – early satiety for abscesses near the stomach, urinary frequency or urgency for those near the bladder, or diarrhea or tenesmus for abscesses near the colon or rectum – and anorexia with or without weight loss can also be seen in patients with an abscess. Although not common, the incidence of biliary tract disease in children is increasing. Of note, children with hemolytic disorders such as sickle-cell anemia and hereditary spherocytosis are at increased risk of cholelithiasis and biliary tract disease. Although most cases of cholelithiasis in children are managed electively, some patients present with complicated disease and associated cholecystitis. These patients are at risk for postoperative abscess, and this complication should be suspected in a patient who has undergone a recent cholecystectomy and has spiking fevers and right upper-quadrant pain. Additional symptoms can include nausea, vomiting, and back or shoulder pain. Elevated white blood cell count and abnormal hepatic enzymes can also be present. Bilirubin levels are usually normal, but if elevated should prompt an evaluation for ascending cholangitis. This is usually the result of obstruction of the common bile duct by a retained common duct stone or occurs in children with biliary atresia post Kasai procedure or liver transplantation. If identified, urgent decompression of the biliary tree is required. Bilirubin levels can also be elevated in the presence of a postoperative bile leak. This usually occurs at the raw surface of the liver bed in the gallbladder fossa. The resulting collection of bile can become infected and require drainage. Finally, another source of intra-abdominal abscess in a patient with a history of cholecystectomy is a retained gallstone in the peritoneal cavity.7 The operative report can indicate this etiology of the abscess. The most common pathogens in cholecystitis include E. coli, Klebsiella, Pseudomonas, and Bacteroides species.7 Antibiotic therapy in a patient with a history of cholecystitis presenting with a postoperative abscess should be targeted against these organisms. When an intra-abdominal abscess from any cause is suspected, imaging is indicated. Whereas plain radiographs can show indirect evidence of abscess such as a pleural effusion or ileus, the diagnosis is confirmed by cross-sectional imaging as in ultrasound or computed

Figure 71-1. A postoperative abscess visualized on computed tomography scan. This rim-enhancing fluid collection has the typical appearance of an abscess.

tomography (CT) (Figure 71-1). In addition, these modalities can facilitate treatment by providing guidance for aspiration of material for culture and placement of drainage catheters. In general, one must maintain a high index of suspicion for abscess in patients with known risk factors. These would include patients who have undergone recent surgery or trauma, immunocompromised patients, and patients who have a chronic illness, such as Crohn disease, that places them at risk for abscess formation.

Management Broad-spectrum antibiotics that cover facultative and anaerobic bacteria should be used empirically.1,6 Therapy should initially be based on location and presumed etiology. Continued therapy should be based on culture and susceptibility tests. Depending on the location and source of the abscess, antibiotic therapy parenterally may be required for an extended period of time. The manner in which an intra-abdominal abscess is managed depends on the circumstances of the individual patient. Antibiotics alone can be effective with fluid collections < 2 cm in diameter. Patients with a large inflammatory mass (comprised of mostly solid viscera and fibrin) with small fluid pockets are considered to have a phlegmon and are usually best treated with antibiotics alone. When imaging reveals a collection of fluid > 2 cm or when a smaller collection has failed to respond to antibiotics, drainage combined with antibiotics is essential. Percutaneous drainage with image guidance has become a mainstay of therapy for pediatric patients with abscesses (Figure 712). Different modalities have been used to facilitate drainage, such as ultrasound and CT guidance for placement of the drain with fluoroscopy and CT to re-evaluate and confirm adequate drainage of the abscess cavity.2,8 Percutaneous drainage is advantageous because it is well tolerated, less invasive than laparotomy for the child, and provides accurate confirmation of response to therapy2,8–10 (Figure 71-3). Although it does frequently require general anesthesia, percutaneous drainage decreases the risk of bowel manipulation, visceral injury, and probably of adhesive small-bowel obstruction. Clinical setting, location, size, organization, and number of abscesses all play a role in determining if the patient is a candidate for percutaneous drainage;8,9 these factors have been confirmed to be predictive of success of

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Intra-abdominal, Visceral, and Retroperitoneal Abscesses

CHAPTER

71

431

A Figure 71-3. Resolution of the fluid collection, catheter still visualized next to small bowel.

better than the other.11 Proponents for laparoscopic drainage cite better visualization of the entire abdominal contents and ability to irrigate the area adequately. In other cases, the need for a concomitant surgical procedure, such as to form stoma or revise an anastomosis, is an indication for an open procedure. Several case series have suggested laparoscopy is an adequate method to drain abscesses.11–14

VISCERAL ABSCESSES Etiology and Clinical Manifestations

B Figure 71-2. (A and B) Computed tomography-guided placement of a drainage catheter into the abscess.

management.8,9 Factors associated with a lower likelihood of successful percutaneous drainage include presence of complex abscesses, loculated or poorly organized abscesses, or those with multiple or extensive collections.1,8 In one study the presence of air–fluid levels and location of air bubbles within the abscess were assessed as possible indicators of success.9 Although the sample size was small, the study results suggest that abscesses with less viscous contents, as indicated by the presence of air–fluid levels or peripheral air bubbles, are likely to be drained successfully by percutaneous methods in 95% of cases. Thicker fluid collections, as evidenced by deep air bubbles, had only a 66% success rate of percutaneous drainage.9 Some intra-abdominal abscesses are best managed by operative or reoperative laparotomy. Indications for surgical intervention are not limited to, but include, uncontrolled septicemia, a cause of the abscess that requires surgical correction (i.e., an anastomotic leak), abscess(es) not amenable to percutaneous drainage, and unsuccessful percutaneous drainage. Surgical drainage can be performed by open or laparoscopic technique. There is no clear evidence that one method is

Prior to the development of antibiotics, abscesses in the liver were not uncommon in otherwise healthy children. However, with the development of appropriate antimicrobial therapies, most liver abscesses that occur currently are seen in underdeveloped countries where antibiotics are less available or in immunocompromised patients such as those with cancer receiving chemotherapy, or those with chronic granulomatous disease or human immunodeficiency virus (HIV)/acquired immunodeficiency syndrome (AIDS).1,15 Immunologically normal children who develop pyogenic liver abscesses typically have had a recent infection such as appendicitis, cholecystitis, and, less frequently, skin or respiratory tract infections.16 Pyogenic liver abscesses in children can be caused by aerobic, facultative or anaerobic bacteria, or a mixture of bacteria.1,3,17,18 The bacteria associated with these infections include Staphylococcus aureus, Escherichia coli, Enterococcus faecalis, or Klebsiella, Enterobacter, Pseudomonas, or Salmonella species. Bartonellosis, catscratch disease, can also cause multiple microabscesses of the liver and spleen in otherwise healthy children and in immunocompromised children with cancer or HIV.1,19 Abscesses in the liver can be single or multiple, simple or complex. They can occur as a result of direct extension of local infection, as in severe cholecystitis or cholangitis. They can also result from hematogenous spread through the hepatic artery or enter through the portal vein from sources within the abdominal cavity. Pyelephlebitis, for example, occurs when enteric organisms access the liver through the portal system. Amebiasis is a frequently discussed cause of liver abscess, although some would argue that it is not a true abscess.1 This parasite reaches the liver through the portal venous system, and usually causes a single large abscess. Other types of parasitic infections, including ascariasis, have been associated with hepatic abscesses. The

432

SECTION

I

Intra-Abdominal Infections

granulomas induced by these infections have been suspected as a predisposing factor for pyogenic liver abscesses in children in endemic areas.16 Traumatic injury to the liver has been associated with liver abscess in about 1% of cases.20 It is more common in patients managed operatively and in those with concurrent injury to a hollow viscus, but it can occur in others as well. Symptoms of hepatic abscesses include fever; abdominal pain, which can be localized in the right upper quadrant or diffuse; nausea; vomiting; and anorexia.21 Jaundice is less common, but can be present if the abscess is compressing the biliary tree. It is more common in some of the parasitic infections, such as ascariasis, when parasites present within the biliary tract cause obstruction of the lumen of the bile ducts leading to ascending cholangitis. The white blood cell count is usually elevated but serum hepatic enzymes may or may not be elevated. A detailed history is important, including kitten exposure or travel through endemic areas of parasitic infestation and history of recent trauma, infection, or surgery. Imaging studies in the form of ultrasonography or CT are confirmatory. Specific antibody tests can be helpful if a parasitic infection is suspected, such as the hemagglutination test which is almost always positive with amebic abscesses, or serum antibodies for Bartonella henselae if cat-scratch disease is suspected. The incidence of splenic abscesses appears to be increasing, although it is unclear whether this is a true increase in number of cases or a higher index of suspicion coupled with better diagnostic tools.22–24 Some reports indicate that the increasing number of immunecompromised patients is contributing significantly to the changing pattern of splenic abscesses.24 Aggressive cancer chemotherapy, increasing numbers of patients with HIV/AIDS, and immunosuppression for organ transplantation may be responsible.24 Other populations at risk are patients with hemolytic disorders such as sickle-cell anemia and those with a history of splenic trauma. Splenic abscesses can be large or small, solitary or multiple. Infections can be due to single organisms or can be polymicrobial. Organisms frequently identified include Staphylococcus aureus, E. coli, and Salmonella, Streptococcus, Candida, and Klebsiella species.24–27 The triad of fever, leukocytosis, and left upper-quadrant pain is common. Nausea and vomiting can also be present as well as splenomegaly and tenderness on physical examination. The diagnosis can be suspected on chest radiograph, when the left hemidiaphragm is elevated or a left pleural effusion is present. Ultrasonography or CT can confirm the diagnosis. Renal and perinephric abscesses are to be mentioned for completeness as they are visceral abscesses found in the retroperitoneal space. They are the result of urogenital pathology infections such as pyelonephritis, urinary obstruction from ureteral stone, or seeding from systemic infections causing bacteremia or fungemia. They are discussed in detail in the section on genitourinary tract infections (see Chapter 52, Renal (Intrarenal and Perinephric) Abscesses). Pancreatic abscesses in children are rare, and can be the result of acute pancreatitis or trauma. The former is discussed in Chapter 67 (Acute Pancreatitis). Injuries of the pancreas due to blunt or penetrating trauma can precipitate an abscess in the presence of either devitalized, necrotic tissue or a pancreatic duct leak. If not diagnosed preoperatively or identified at the time of exploration, the resulting fluid collection can be secondarily infected. Mortality associated with pancreatic abscesses has been reported to be 15% to 50%, with morbidity rates as high as 75% to 80%.28,29 The most common organisms include E. coli and Klebsiella, Enterococcus, Staphylococcus, Streptococcus, and Pseudomonas species, as well as fungi.28,29 Symptoms include fever, epigastric or left upper-quadrant pain, and hemodynamic instability. Laboratory tests show an elevated white blood cell count with variable elevation of serum amylase and lipase. CT with rapid contrast injection is the most sensitive test for pancreatic abnormalities and can show a nonenhancing portion of the pancreas or a fluid collection in or near the pancreas.28 The presence of gas within the fluid collection is an ominous sign.

Management Antibiotic therapy should be broad-spectrum, effective against the most common gram-positive and gram-negative organisms associated with the type of abscess present. In immunocompromised patients, antifungal therapy is considered. Decisions regarding drainage are made based on the clinical presentation, condition of the patient, and need for microbiologic confirmation. In liver abscesses, there have been reports of treating multiple microabscesses successfully with antibiotics alone, especially in cases of cat-scratch disease.19,21 Abscesses < 4 cm in diameter can be successfully treated by aspiration, with antibiotic therapy continued until the patient’s clinical status improves and CT shows resolution of the abscess. However, in patients with abscesses > 4 cm or recurrence after aspiration, continuous drainage using a CT- or ultrasound-guided percutaneous catheter is indicated in addition to antibiotic therapy.21 Surgical drainage is recommended in patients with liver abscesses that do not resolve with percutaneous drainage, those that present with rupture of the abscess cavity into the peritoneal cavity, and in patients with chronic granulomatous disease who present with persistent or recurrent liver abscesses.15,21 The management of splenic abscesses is usually similar to that of liver abscesses. However, splenectomy is curative. Although salvage of the spleen is preferable and usually possible, the primary goal is eradication of infection. Splenectomy should be considered in anyone who does not respond to drainage and antibiotics. It should also be considered in patients who are immunocompromised and have ongoing septicemia, in patients with sickle-cell disease or other hemoglobinopathies, in those who present with rupture of the abscess into the peritoneal cavity, and in those who have an abscess following a traumatic splenic injury.24–26,30 This can be approached by either an open or laparoscopic technique. Pancreatic abscesses should be managed with drainage. In cases of pancreatic fluid collections that may not be infected or in cases of pancreatic phlegmon, CT-guided aspiration of pancreatic or peripancreatic fluid can provide specimens for culture to guide antibiotic therapy. Percutaneous drainage catheters can be left in place if the fluid appears purulent or Gram stain is positive for organisms. In patients who have ongoing septicemia or whose fluid collections are not amenable to percutaneous drainage, open operative exploration is required.28,29,31 Percutaneous drainage has had success rates reported in the range of 69% to 86%.28,29 In those patients who ultimately require surgery, percutaneous drainage is often a good temporizing measure in patients who are too unstable to undergo surgery due to severe sepsis.

RETROPERITONEAL ABSCESSES Etiology and Clinical Manifestations Retroperitoneal abscesses include visceral abscesses of the kidney and pancreas, which have already been discussed. Other causes of abscesses in the retroperitoneum include traumatic duodenal injury and IPAs. The former may be a result of penetrating trauma or blunt trauma associated with “handlebar” injuries by bicycles or motorcycles. Several cases have been reported of penetrating trauma to the duodenum presenting as a retroperitoneal abscess after the ingestion of a sharp foreign body such as a needle, toothpick, or fish bone.32 Abscesses associated with duodenal injuries or perforations are extremely difficult to manage and have a high rate of morbidity and mortality.33–35 Evacuation of the abscess cavity is difficult, and the underlying problem, a disruption of the duodenal wall, must be addressed to prevent further complication, such as duodenal fistula. Populations at risk include those with a history of trauma who manifest right upper-quadrant or epigastric pain, fever, nausea and vomiting, an elevated white blood cell count, and, in many cases, an elevated serum amylase level. Findings on physical examination that

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Intra-abdominal, Visceral, and Retroperitoneal Abscesses

should cause one to suspect duodenal injury include a contusion of the upper abdomen with signs of pertionitis. Diagnostic evaluation includes plain radiographs of the abdomen, which may not be diagnostic unless free air is present. An upper intestinal contrast study or CT is the next step and is usually diagnostic.33,34 Organisms reported in these abscesses include Staphylococcus and Streptococcus species, Escherichia coli, Klebsiella, and Candida species.3,33,36 IPAs can be described as either primary or secondary.37 Primary IPAs have no detectable source, and are the most common type in children. The most common organism in primary IPA is Staphylococcus aureus, indicating that an unidentified, cutaneous source leading to bacteremia is most likely.14,38,39 Secondary IPA is associated with a known source of infection. Because the psoas muscle is intimately related to the ureter, renal pelvis, spine, appendix, terminal ileum, pancreas, jejunum, and sigmoid colon, an infectious process deriving from any of these structures can cause a secondary IPA.14,38,40 Examples include pyelonephritis, appendicitis, osteomyelitis, pancreatitis, and inflammatory bowel disease. Systemic infections can also be responsible. The most common organism is Staphylococcus aureus. Other organisms reported include Streptococcus pneumoniae (the most common causative organism in patients with histories of recent respiratory infections) and gram-negative organisms, such as E. coli (in patients with a history of appendicitis or other gastrointestinal infection or surgery). Underlying illness such as diabetes, HIV infection, or connective tissue disease may be present and increase susceptibility, but this is more commonly the case in adults. A history of trauma can also increase risk. There have been several reports of IPA in neonates, and there is commonly a history of femoral venous catheter use or omphalitis in these patients.5,38,39 The most common organism is Staphylococcus aureus, with a higher rate of methicillin-resistant S. aureus isolated in this circumstance. Presenting symptoms of IPA include fever, leg pain or swelling, and decreased range of motion in the ipsilateral hip.37 In patients old enough to walk, the degree of dysfunction of the hip joint is widely variable, from no apparent dysfunction to complete pseudoparalysis. A positive psoas sign may be found on physical examination, as might inguinal swelling or lymphadenopathy.

CHAPTER

71

433

The diagnosis of septic arthritis of the hip must be distinguished from IPA. Laboratory testing commonly demonstrates an elevated WBC count and an elevated erythrocyte sedimentation rate in both conditions. Plain radiographs can show obscuring of the sacroiliac joint or an effusion of the hip, evidenced by joint space widening. An effusion is not necessarily indicative of pyogenic arthritis of the hip as it can occur as a sympathetic effusion due to the surrounding inflammatory process. However, because a delay in treatment of pyogenic arthritis must be avoided, an aspiration of the joint is the safest approach to exclude this diagnosis.37 Ultrasonography or CT is the best diagnostic tool to characterize an IPA and can assist in obtaining samples for culture, placement of drains, and follow-up for resolution.

Management The management of retroperitoneal abscesses resulting from duodenal pathology is extremely difficult. Several factors should be considered when deciding the best approach. These include time since injury, clinical presentation and hemodynamic stability of the patient, as well as concurrent injuries or illnesses. Any approach should accomplish drainage of the retroperitoneal abscess cavity and address the primary pathology of the duodenum.33–35 The possible surgical approaches include primary repair and drainage with or without pyloric exclusion versus a more aggressive approach of a pancreaticoduodenectomy. There have also been reports of successful drainage of the abscess percutaneously and stabilization of the patient prior to surgery.34 Nutritional support is extremely important in this patient population as oral feeding will necessarily be avoided after surgery.33,35 IPAs are managed by drainage and antibiotic therapy. Pyogenic arthritis of the hip should be excluded by joint aspiration. Except in the case of a known prior infection, empiric antibiotic therapy should be selected to cover Staphylococcus aureus, since this is the most common organism. Recently, drainage has been accomplished more frequently by ultrasound-guided aspiration or drainage using a percutaneous catheter, with good success rates.41,42 A retroperitoneoscopic approach as well as an open operative approach have been used when percutaneous drainage is not successful.27

SECTION

J

Skin and Soft-Tissue Infections

CHAPTER

72

Cellulitis and Superficial Skin Infections John Browning and Moise Levy

An infectious agent tends to localize in one (or more) of several skin layers, making recognition of the type of lesion produced, as well as its depth, essential in establishing a correct diagnosis.1,2 The stratum corneum, the outermost layer of the epidermis, as well as the deeper layers of the epidermis constitute the principal barrier against infection. The dermis is composed of three primary connective tissues: collagen, elastic tissue, and reticular fibers. The subcutis or subcutaneous tissue, which lies below the dermis, is principally composed of connective tissue and fat cells. Cutaneous vessels pass through both the subcutis and the dermis. Several types of primary lesions can form in the skin as the result of a primary infectious process, including macules (flat lesions < 1 cm), patches (flat lesions > 1 cm), papules (elevated lesions < 1 cm), plaques (elevated broad flat lesions > 1 cm), nodules (domed or rounded lesions > 1 cm arising from the dermis or subcutis), pustules (pus-filled lesions), vesicles (elevated lesions < 1 cm, filled with serous fluid), and bullae (elevated lesions > 1 cm, filled with serous fluid). An agent usually produces a characteristic type of primary lesion, which can evolve from one type to another and has a characteristic pattern of spread, secondary changes, and resolution.

Cutaneous manifestations of bacterial disease can occur through a variety of pathogenic mechanisms, including: (1) primary infection with local replication, leading to a local inflammatory response; (2) circulating exotoxins (e.g., in staphylococcal scalded-skin syndrome); (3) immunologic mechanisms (as in vasculitis associated with streptococcal infections); (4) involvement of the skin as part of a systemic infectious disease; and (5) as a manifestation of disseminated intravascular coagulopathy. Primary infection of the skin involves the epidermis, dermis, or subcutaneous tissue, whereas soft-tissue infection involves the deeper fascia or muscle or both. A superficial skin infection is primarily limited to the epidermis, the dermis, or both, although secondary inflammation can extend to the subcutis. Superficial lesions are generally small; tenderness, if present, is localized; tissue necrosis, gangrene, or abscess formation is minimal to absent; and few or no systemic manifestations develop. Primary, superficial infections of the skin are the focus of this chapter (Table 72-1).

MECHANISM OF INFECTION Normal skin microflora can be categorized into two broad groups: (1) resident flora, which are attached to the skin, and are present in relatively stable numbers; and (2) transient flora, which are introduced from the environment and only attach if integrity of the skin is disrupted. The most important transient bacteria are group A streptococcus (GAS: Streptococcus pyogenes) and Staphylococcus aureus. Whereas coagulase-negative staphylococci are most numerous among resident flora, there is a high degree of natural resistance to colonization of the skin with Staphylococcus aureus. Persistent nasal carriage is detected in 20% to 40% of immunocompetent adults, however, and up to 20% can be colonized on the perineum3,4; other

TABLE 72-1. Primary Superficial Cutaneous Bacterial Infections Disease Entity

Skin Lesions

Infectious Agent(s)

Anthrax Blistering distal dactylitis Cellulitis Diphtheria Ecthyma Erysipelas Erysipeloid Folliculitis Sycosis barbae Gram-negative folliculitis

P, V, B V Pl P, Pu, U Pu, Pl, U PI, V, B Pa P, Pu P, Pu P, Pu P, N P, Pu, N Pa N N V, B, Pu, Pl V, Pu, Pl P Pa, Pl E

Bacillus anthracis Streptococcus pyogenes, group B streptococcus, Staphylococcus aureus Streptococcus pyogenes, Staphylococcus aureus Corynebacterium diphtheriae Streptococcus pyogenes Streptococcus pyogenes, groups B, C, G streptococci Erysipelothrix rhusiopathiae Staphylococcus aureus, coagulase-negative staphylococci, Candida albicans, Malassezia furfur Staphylococcus aureus Klebsiella spp., Enterobacter spp., Escherichia coli, Pseudomonas aeruginosa Proteus spp. Pseudomonas aeruginosa Corynebacterium minutissimum Staphylococcus aureus Staphylococcus aureus, Streptococcus anginosus group, Escherichia coli, anaerobic streptococci Staphylococcus aureus Streptococcus pyogenes Mixed aerobic and anaerobic infection (see text) Streptococcus pyogenes Coryneform bacteria

Hot-tub folliculitis Erythrasma Furuncles, carbuncles Hidradenitis suppurativa Impetigo Paronychia Perianal dermatitis Pitted keratolysis

B, bulla; E, erosion; N, nodule; P, papule; Pa, patch; Pl, plaque; Pu, pustule; V, vesicle.

434

Cellulitis and Superficial Skin Infections

body sites are only transiently contaminated. By contrast, in individuals with atopic dermatitis, S. aureus is recovered from 90% or more of skin lesions and 70% of cultures from unaffected skin.5,6 Nasal carriage of S. aureus in neonates or in patients with atopic dermatitis has been associated with a higher number of days of hospitalization.7 Coryneform organisms are lipophilic, pleomorphic gram-positive bacilli that are normal residents of moist, intertriginous areas of skin. Brevibacterium preferentially grows in toe webs, especially in association with tinea pedis, and is implicated in foot odor.8 Propionibacterium spp. are anaerobic gram-positive bacilli that grow in hair follicles and sebaceous glands; P. acnes is the predominant organism at sites of high sebum production, such as the scalp, forehead, and back. Whereas micrococci and coagulase-negative staphylococci usually colonize the surface of the stratum corneum and upper parts of hair follicles, Corynebacterium and Propionibacterium spp. are found deep in follicular canals. Candida spp., most commonly C. albicans, colonize the oral mucous membranes in up to 40% of individuals; normal skin is seldom colonized except in individuals in whom overgrowth occurs because of antibiotic therapy or immunosuppression. Balance between host defense mechanisms and virulence of organisms is the major determinant of skin infection.9–12 Factors that modify the resident flora facilitate transient colonization by pathogens such as Streptococcus pyogenes and Staphylococcus aureus. These factors include increased temperature, humidity, presence of cutaneous or systemic disease, young age, and antibiotic treatment.9,13,14 Decreased temperature favors the nonpathogenic, coagulase-negative staphylococci over S. aureus and decreases the virulence of infection with the latter organisms.13 Treatment with antibiotics (systemic or topical) can reduce the density of coryneforms, with a corresponding increase in density of micrococci and gram-negative bacilli.14 Topical application of corticosteroid has little effect on microflora, except in individuals with atopic dermatitis, in whom the density of S. aureus can diminish, perhaps through healing of the skin and the resultant enhancement of epidermal barrier function.15 The process of epithelial colonization is thought to involve irreversible adherence to a specific receptor on the host cell via an adhesin, typically an antigen on a filamentous projection from the bacterial cell wall.16 Alternatively, colonization might involve relatively weak, nonspecific bonds (e.g., hydrophobic), whereas a stronger bond involving a specific interaction between cell surface receptors and bacterial adhesins may be necessary for initiation of infection. The hyaluronic acid capsule of Streptococcus pyogenes impedes the interaction of bacterial adhesins with keratinocyte receptors. The molecular basis for attachment of skin strains of S. pyogenes to keratinocytes remains unidentified,9 but protein F may mediate adherence to epidermal Langerhans cells.17,18 The inability of streptococci and staphylococci to colonize intact epithelium may be due to the absence of receptors on normal skin surface; attachment can occur when breaks in the skin reveal the receptors. Competitive blocking of pathogens by topical application of a purified bacterial adhesin or of cell surface receptors has been demonstrated in the prevention and treatment of bacterial skin infections (e.g., mupirocin).19–21 Competitive binding to cell surface receptors by resident flora or transient flora of low virulence can also prevent pathogenic bacteria from colonizing skin. Purposeful nasal colonization with a nonvirulent strain of Staphyococcus aureus has been used in neonates during nursery epidemics of staphylococcal disease and in adults to interrupt recurrence of staphylococcal abscesses within families.22,23 After successful colonization, several other host defense mechanisms must be breached before infection develops. Intact, overlapping cells of the stratum corneum and its dry, acidic, inhospitable environment provide the first mechanism of defense. Breakdown products of the stratum corneum, including free fatty acids, polar lipids, and glycosphingolipids, have antistaphylococcal and antistreptococcal activity.24 Many of the resident flora, particularly the lipophilic corynebacteria, release lipases that liberate

CHAPTER

72

435

fatty acids from triglycerides of sebum, creating an acid mantle.25 This mantle favors growth of propionibacteria, which in turn produce propionic acid, a compound that has relatively more antimicrobial activity against transient organisms than against resident flora. Cationic antimicrobial peptides (e.g., human b-defensin-2, LL-37, secretory leukocyte protease inhibitor, skin-derived antileukoprotease) are produced by keratinocytes and have broad-spectrum antimicrobial activity, providing innate immunity. Their role in human cutaneous defense remains to be defined.26 Antigen presentation by epidermal Langerhans cells and cytokine secretion by keratinocytes also play a key role in defense against infection.11,27 Bacteria produce an array of compounds that make the environment noxious to competitors. Gram-negative bacteria tend to produce substances with a wide spectrum of activity, whereas the substances secreted by gram-positive organisms tend to affect only the same or closely related species. These narrow-spectrum substances are called bacteriocins.28 Bacteriocin-producing organisms, including Staphylococcus aureus, tend to increase in number and to become the predominant species when skin is damaged.29

IMPETIGO Nonbullous Impetigo Nonbullous impetigo accounts for more than 70% of cases of impetigo.30–34 Lesions of nonbullous impetigo form on skin at sites of trauma, most commonly at varicella lesions, insect bites, abrasions, lacerations, and burns.35 Lesions have an erythematous base with “honey-crusted” exudates. They are associated with minimal pain and surrounding erythema; constitutional symptoms are generally absent. Regional adenopathy occurs in up to 90% of cases, and leukocytosis is present in approximately 50% of persons. Without treatment, most cases resolve spontaneously and without scarring in approximately 2 weeks.36 The differential diagnosis of nonbullous impetigo includes contact dermatitis, viral (herpes simplex, varicella-zoster), fungal (dermatophyte), and parasitic (scabies, pediculosis) infections, all of which can become infected secondarily. S. aureus is the predominant cause of nonbullous impetigo in the United States; Streptococcus pyogenes is less frequently implicated.34,37–40 Staphylococci generally spread from the nose to normal skin.36 By contrast, S. pyogenes colonizes the skin an average of 10 days before development of impetigo and persists in the nasopharynx an average of 2 to 3 weeks after the appearance of lesions.36,41,42 Lesions of nonbullous impetigo that yield growth of staphylococci or S. pyogenes cannot be distinguished clinically. Whereas Staphylococcus aureus causes impetigo in children of all ages, Streptococcus pyogenes is most common in children of preschool age and is unusual before 2 years of age, except in highly endemic areas. Staphylococcal types causing nonbullous impetigo are variable but are not generally of phage group 2, the group associated with toxin production and bullous impetigo.32 Several serotypes of S. pyogenes, termed impetigo strains, are found most frequently in lesions of nonbullous impetigo and are generally different from those that cause pharyngitis.42 The anti-DNAase B titer is the best serologic test for detecting preceding streptococcal impetigo.43

Bullous Impetigo Impetigo with flaccid bullae occurs mainly in infants and young children (Figure 72-1). Bullae may be single or clustered and often lack underlying erythema but can become purulent. The term “staph pustulosis” refers to multiple small pustules on the abdomen and diaper area. It is caused by Staphylococcus aureus, most often phage group 2.36,44 Lesions of bullous impetigo are a manifestation of localized toxin production (exfoliatin or epidermolytic toxins A and B) and develop on intact skin.45,46 Culture of fluid swabbed from an intact blister or from beneath the lifted edge of a crusted lesion is

436

SECTION

J

Skin and Soft-Tissue Infections

Figure 72-1. Bullae and an erosion on the left medial upper thigh of an infant with bullous impetigo caused by Staphylococcus aureus.

usually positive for S. aureus. The differential diagnosis of bullous impetigo in the neonate includes epidermolysis bullosa, bullous mastocytosis, herpetic infection, early scalded-skin syndrome, and group B streptococcal infection.47,48 In older children, insect bites, contact dermatitis, burns, erythema multiforme, chronic bullous dermatosis of childhood, pemphigus, and bullous pemphigoid must be considered, particularly if the lesions do not respond to therapy.

Complications Complications of either nonbullous or bullous impetigo are rare. Cellulitis occurs in approximately 10% of patients with nonbullous impetigo but less frequently with the bullous form.32 Lymphangitis, suppurative lymphadenitis, guttate psoriasis, and scarlet fever can follow streptococcal disease. There is no correlation between number of lesions of streptococcal impetigo and development of lymphangitis or cellulitis. Infection with nephritogenic strains of Streptococcus pyogenes can result in acute poststreptococcal glomerulonephritis. The most commonly affected age group is children 3 to 7 years old. The clinical character of preceding impetiginous lesions is not unusual. The latent period from onset of impetigo to development of poststreptococcal glomerulonephritis averages 18 to 21 days.42,49 Poststreptococcal glomerulonephritis occurs epidemically after either skin infection or pharyngitis. After skin infection, it is usually caused by strains M2, M49, M53, M55, M56, M57, and M60.42,49 Strains of S. pyogenes associated with endemic impetigo in the United States have little or no nephritogenic potential.50 Acute rheumatic fever does not occur as a result of impetigo.

Treatment Topical mupirocin (pseudomonic acid A) or systemic antibiotic treatment is superior to placebo or cleansing with 3% hexachlorophene soap for treatment of impetigo.51 Mupirocin exerts its bactericidal effect through inhibition of bacterial isoleucyl-tRNA synthetase. Applied 3 times daily for 7 to 10 days, its effectiveness is comparable with oral erythromycin ethylsuccinate, 30 to 50 mg/kg per day for 7 to 10 days.31,33,34,38,39,52 Bacterial resistance to mupirocin is reported, most often in patients treated sporadically or prophylactically for prolonged periods (> 2 weeks).53 Systemic therapy should be used in patients with widespread lesions; lesions near the mouth (where topical medication may be licked off); in those with evidence of deeper involvement, such as cellulitis, furunculosis, abscess formation, or suppurative lymph-

adenitis; or in those with constitutional symptoms.54 Erythromycin was frequently used in the past, but resistance to all macrolides occurs frequently with Staphylococcus aureus and sporadically among Streptococcus pyogenes.55 Other antibiotics that are effective for impetigo include dicloxacillin56; amoxicillin plus clavulanic acid57–59; clindamycin60; azithromycin61; clarithromycin62; or a cephalosporin such as cephalexin,30 cefaclor,58,60,63,64 cefadroxil,62 cefprozil,63 or cefpodoxime.65 With the emergence of methicillin-resistant Staphylococcus aureus (MRSA), trimethoprim-sulfamethozalole and clindamycin are options for outpatient oral antibiotic therapy. Intravenous vancomycin can be used to treat hospitalized patients with more severe infections and oral doxycycline can be used to treat children older than age 8 years. Linezolid is an option for patients who fail other agents.66 Seven days of therapy are usually adequate. A patient with recurrent impetigo should be evaluated for carriage of S. aureus in the nares, although approximately 25% of carriers harbor S. aureus exclusively at sites other than the nares, such as the perineum and axillae.67 Nasal carriage of methicillin-susceptible S. aureus (MSSA) (and, less frequently, MRSA) can usually be eliminated with 2 to 4 days of topical application of mupirocin,68,69 thereby reducing rates of staphylococcal infection in patients who have atopic dermatitis or are undergoing hemodialysis.70 Use of mupirocin to eliminate nasal carriage should be reserved for outbreaks, for patients with recurrent staphylococcal impetigo, and for furunculosis.69,71 Skin antisepsis is also useful for patients with repeated bacterial skin infections of these types.

ECTHYMA Ecthyma resembles nonbullous impetigo in onset and appearance but gradually evolves into a deeper, more chronic infection. The initial lesion is a vesicle or vesicopustule with an erythematous base that erodes through the epidermis into the dermis to form a crusted ulcer with elevated margins up to 4 cm in diameter. Lesions occur most commonly on the legs, particularly in the setting of poor hygiene and malnutrition, at sites of pruritus such as insect bites, scabies, or pediculosis, which are subject to frequent scratching. Complications include lymphangitis, cellulitis, and (rarely) poststreptococcal glomerulonephritis. The causative agent is usually Streptococcus pyogenes; Staphylococcus aureus is also isolated from culture of most lesions but is probably a secondary pathogen. Systemic antibiotic therapy, as for impetigo, is indicated.54 Ecthyma gangrenosa is a necrotic ulcer covered with a gray-black eschar (Figure 72-2). It is generally a sign of disseminated Pseudomonas aeruginosa infection in neutropenic patients with leukemia or aplastic anemia, but it has been described in previously healthy children, possibly in relation to excessive exposure to water from bathing, whirlpools, or bath sponges.72 Ecthyma gangrenosa can also occur as a primary cutaneous infection after inoculation. Similar lesions can also develop as a result of infection with other agents, such as S. aureus, Aeromonas hydrophila, Enterobacter spp., Escherichia coli, Proteus spp., Burkholderia cepacia, Serratia marcescens, Stenotrophomonas maltophilia, Aspergillius spp., Mucor organisms, and Candida albicans.73

CELLULITIS Cellulitis is characterized by edema, warmth, erythema, and tenderness. The lateral margins tend to be indistinct, because the process is deep in the skin, involving the subcutaneous tissues in addition to the dermis (Figure 72-3). Regional lymphadenopathy and the constitutional symptoms fever, chills, and malaise are common. A break in the skin due to previous trauma or an underlying skin lesion predisposes to cellulitis. The most common etiologic agents are Streptococcus pyogenes and Staphylococcus aureus. Occasionally, groups G, C, and (in neonates) B streptococci are the causal organisms. Complications of streptococcal cellulitis include

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Cellulitis and Superficial Skin Infections

CHAPTER

72

437

A

B

C

lymphangitis, osteomyelitis, pyogenic arthritis, thrombophlebitis, bacteremia, and necrotizing fasciitis. In patients who are immunocompromised or who have diabetes mellitus, a number of other bacterial or fungal agents can be implicated, notably Pseudomonas, Enterobacteriaceae, and Cryptococcus neoformans. Prior to universal immunization in infancy, bacteremic Haemophilus influenzae type b was an important cause of facial cellulitis with bluish discoloration.74 Facial cellulitis associated with a portal of entry such as tooth abscess or cutaneous trauma is generally due to oral anaerobic bacteria, Streptococcus pyogenes, or Staphylococcus aureus.75 Streptococcus pneumoniae and Salmonella spp. can also cause cellulitis resembling that due to Haemophilus influenzae type b.76,77 Reflex sympathetic dystrophy, a syndrome of unclear etiology, is underrecognized in children and can mimic cellulitis when pain, edema, and flushing are predominant features. Absence of fever and leukocytosis and presence of hyperesthesia, vasomotor instability, tache cérébrale, and diminished tracer uptake on technetium scanning can be helpful clues to diagnosis.78

Figure 72-2. Pseudomonas aeruginosa septicemia and ecthyma gangrenosa in a young male with severe neutropenia (A). Microscopic appearance of echythma gangrenosa in fatal P. aeruginosa septicemia in an infant with leukemia (B). Haematoxylin-eosin stain shows superficial necrosis and elevation as well as bland ischemic necrosis beneath. Gram stain of fluid from bullous lesion (B) shows dense gram-negative bacilli with rare inflammatory cells (C). (Courtesy of J.H. Brien.)

Culture of aspirate from the site of inflammation, skin biopsy, and blood cultures collectively allow identification of the causal organism in approximately 25% of cases.79 Yield is higher (by approximately one-third) when pre-existing abrasion or ulcer is present. An aspirate taken from the point of maximum inflammation yields the causal organism more often than does a leading-edge aspirate.80 Empiric therapy directed against Streptococcus pyogenes and Staphylococcus aureus is indicated for cellulitis in an immunocompetent host.54 In most cases of cellulitis on an extremity, bloodstream infection (BSI) is rare. If fever and lymphadenopathy are absent, outpatient treatment using a penicillinase-resistant penicillin or a first-generation cephalosporin is appropriate. The need for an agent for coverage of community-associated (CA) MRSA should be guided by local prevalence and antibiotic susceptibilities. Parenteral therapy is instituted if fever, rapid progression, lymphangitis, or lymphadenitis is present. When erythema, warmth, edema, and fever have decreased substantially in uncomplicated cases, a 10-day course of treatment can be completed with oral therapy.

438

SECTION

J

Skin and Soft-Tissue Infections

are only helpful in retrospect.84 Skin biopsy shows edema and vascular dilation in the dermis and upper subcutis; occasionally, organisms can be seen within lymphatic vessels, particularly early in the course of infection. Prompt institution of a penicillin parenterally is the mainstay of treatment. Change to oral therapy can be instituted in most cases in immunocompetent hosts within 3 days to complete a 10-day course of therapy. In more indolent cases, oral therapy alone can be sufficient. Local wound care and attention to predisposing factors are also important. If the presence of staphylococci is a concern, a blactamase-resistant antibiotic should be used. In areas with a high prevalence of CA-MRSA, an appropriate antibiotic, such as clindamycin, should be considered. Immunocompromised hosts may require prolonged therapy, and patients with recurrent disease may benefit from prophylactic therapy with penicillin, erythromycin, or an alternative antistreptococcal drug.81

ERYSIPELOID

Figure 72-3. Young infant with lymphadenitis and cellulitis due to Staphylococcus aureus. (Courtesy of J.H. Brien.)

ERYSIPELAS Erysipelas is a superficial form of cellulitis with lymphatic involvement. It occurs sporadically and usually begins at a break in the skin. Common portals of entry include ulcers on lower extremities, sites of local trauma, dermatoses, intertriginous or pedal fungal infections, insect bites, and heel fissures.81 In neonates, infection can originate at the umbilical stump and spread to the abdominal wall or can originate at the circumcision site. Predisposing host factors include venous or lymphatic obstruction, nephrotic syndrome, and diabetes mellitus. Approximately one-third of affected patients have a preceding upper respiratory tract infection, presumably viral in origin. The most common site of erysipelas is the lower extremities; in the past, the face was most common.81,82 Onset of erysipelas is abrupt, with fever, chills, and malaise, followed within 1 or 2 days by cutaneous signs. A small area of burning and redness develops into a warm, shiny, bright red, confluent, indurated tender plaque with a brawny, peau d’orange appearance and elevated, sharply demarcated margins. Vesicles, hemorrhagic bullae, and ecchymoses can develop in the plaque, and regional lymphadenitis can be present. Fine desquamation and, sometimes, residual pigmentation accompany resolution of the plaque. Complications include bloodstream infection, abscess formation, gangrene, thrombophlebitis, and glomerulonephritis. Pyrogenic exotoxin-producing strains of Streptococcus pyogenes are associated with local invasiveness of lesions, septicemia, and toxic shock. Patients with seemingly localized skin infections must be questioned and carefully examined for signs of these complications (see Chapter 13, Toxic Shock Syndrome, Chapter 15, Mucocutaneous Symptom Complexes, and Chapter 118, Streptococcus pyogenes (Group A Streptococcus)). In most cases, S. pyogenes is the cause of erysipelas. Group G as well as groups B, C, and D streptococci are occasionally responsible, particularly in hosts compromised by surgery or other illness. Staphylococcus aureus, Streptococcus pneumoniae, Klebsiella pneumoniae, and Yersinia enterocolitica are rarely implicated. Diagnosis of erysipelas is made primarily on clinical grounds. Skin culture from the portal of entry can be helpful in identifying the causal organism.82 Pharyngeal culture results are infrequently positive. Tests for streptococcal antigens in skin biopsy specimens may be more sensitive than skin culture, but their utility must be verified.83 Culture of saline aspiration of the lesion is usually futile because of the low number of organisms present. Elevated titers of antistreptolysin-O and anti-DNAase B antibodies occur in approximately 40% of cases but

Erysipeloid is a rare cutaneous infection due to inoculation of Erysipelothrix rhusiopathiae from contaminated animals, birds, fish, or their products. There are three forms of infection. The localized cutaneous form is most common and is characterized by welldemarcated, diamond-shaped, erythematous to violaceous patches at sites of inoculation. Local symptoms are generally not severe, constitutional symptoms are rare, and the lesions resolve spontaneously after weeks but can recur at the same site or develop elsewhere weeks to months later. The diffuse cutaneous form manifests as lesions at several areas of the body in addition to the site of inoculation and is also self-limited. The systemic form, due to hematogenous spread, is accompanied by constitutional symptoms and can include endocarditis, pyogenic arthritis, cerebral infarction with abscess formation, meningitis, and pulmonary effusion. Diagnosis is confirmed by skin biopsy, demonstrating gram-positive bacilli, and culture. The treatment of choice is parenteral penicillin or erythromycin.

BLISTERING DISTAL DACTYLITIS Blistering distal dactylitis is a superficial blistering infection of the distal volar fat pad of the phalanges.85 More than one digit can be involved, as can the volar surfaces of the proximal phalanges, the toes, and the palm. Blisters are filled with a purulent fluid that contains polymorphonuclear leukocytes and infecting organisms. There is usually no preceding history of trauma, and systemic symptoms are generally absent. The infection is most commonly caused by Streptococcus pyogenes; group B streptococcus and Staphylococcus aureus are less frequent etiologies.86,87 If left untreated, blisters can enlarge and extend to the paronychial area. The infection responds to incision and drainage and a 10-day course of systemic therapy with a b-lactamase-resistant penicillin or cephalosporin, such as dicloxacillin, cloxacillin, or cephalexin. Penicillin-allergic patients can be treated with erythromycin or clindamycin.

PARONYCHIA Acute paronychia usually results from local injury to the nail fold. It occurs most commonly in children who suck their fingers, bite their nails or cuticles, or have poor hygiene. The lateral nail fold becomes warm, erythematous, edematous, and painful. In most cases, the infection is caused by mixed oropharyngeal flora. The most common aerobic organisms are Streptococcus pyogenes, Staphylococcus aureus, and Eikenella corrodens; anaerobic pathogens include Prevotella spp., Fusobacterium nucleatum, and gram-positive cocci, particularly Peptostreptococcus spp.88,89 Chronic infection can be caused by Candida albicans. Warm compresses are generally curative

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Cellulitis and Superficial Skin Infections

for superficial lesions. Antibiotic therapy, in addition to incision and drainage, is needed for treatment of deeper lesions. Both aerobic and anaerobic cultures of purulent material are recommended. Because of the emergence of penicillin-resistant strains of anaerobic organisms, therapy with clindamycin or amoxicillin plus clavulanic acid is indicated.90 Herpes simplex infection of the fingers can also occur after sucking or nail-trimming by biting; herpetic whitlow can resemble staphylococcal infection.91 Multiple vesicles and dusky appearance are typical of whitlow. These should not be incised or debrided; instead, acyclovir therapy should be given.

PERIANAL DERMATITIS Perianal dermatitis due to Streptococcus pyogenes occurs most commonly in boys (70% of cases) between the ages of 6 months and 10 years and manifests as perianal erythema (90% of cases) and pruritus (80% of cases).92,93 Approximately half of patients have rectal pain, most commonly described as burning inside the anus during defecation, and a third have blood-streaked stools. Refusal to defecate is common. The rash is superficial, erythematous, well-marginated, nonindurated, and confluent from the anus outward. Initially, the rash tends to be bright red, moist, and tender. With time, painful fissures, a dried mucoid discharge, or psoriasiform plaques with yellow, peripheral crusts become more prominent, and the erythema subsides. In girls, the perianal rash can be associated with vulvovaginitis, vaginal discharge, and vulvar redness.94 Patients with guttate psoriasis can present with perianal lesions, emphasizing the importance of examining the anus in all cases of guttate psoriasis.95 Although local induration or edema can occur, subcutaneous involvement is absent. Familial spread of perianal dermatitis is common, particularly when family members use the same bath water. The differential diagnosis of perianal dermatitis includes psoriasis, seborrheic dermatitis, candidiasis, pinworm infestation, sexual abuse, and inflammatory bowel disease. Isolation of S. pyogenes from culture of perianal swab specimens confirms the diagnosis; antigen detection test on a swab specimen is less sensitive. Characteristically, antistreptolysin-O and anti-DNAase B titers do not rise. A 10-day course of oral penicillin is curative in most patients; however, recurrence rates of 40% to 50% have been reported, emphasizing the need for close follow-up. Clindamycin or macrolide therapy may be useful in individuals who fail penicillin therapy. Mupirocin has been used in conjunction with oral antibiotics to treat recurrence but has not been evaluated as single-drug therapy.

FOLLICULITIS A superficial infection of the hair follicle, folliculitis, manifests as a discrete, dome-shaped pustule with an erythematous base, located at the ostium of the pilosebaceous canal, usually on the scalp, buttocks, or extremities. A moist environment, maceration, poor hygiene, and drainage from adjacent wounds and abscesses can be provocative factors. In patients infected with human immunodeficiency virus, confluent erythematous patches with satellite pustules in intertriginous areas and violaceous plaques composed of superficial follicular pustules can develop on the scalp, axillae, or groin.96 The causative organism of folliculitis can be identified from Gram stain and culture of purulent material from the follicular orifice. Staphylococcus aureus is the predominant cause, although coagulase-negative staphylococci are occasionally involved. Candida spp. characteristically cause satellite follicular papules surrounding patches of candidal intertrigo, particularly in patients receiving corticosteroid or antibiotic therapy. Malassezia furfur can also cause pruritic, 2- to 3-mm, erythematous, perifollicular papules and papulopustules on the back, chest, and extremities, particularly in patients who have diabetes mellitus or are receiving systemic corticosteroids or antibiotics. Diagnosis is made by examination of a potassium hydroxide-treated scraping from a lesion. A skin biopsy is

CHAPTER

72

439

often necessary to identify grapelike clusters of yeast as well as the short, septate branching, “spaghetti-and-meatballs” hyphae of M. furfur in widened follicular ostia, mixed with keratinous debris; isolation in culture requires use of olive oil overlaid on agar. Topical antibiotic cleansers, such as chlorhexidine and hexachlorophene, are usually effective for mild cases of folliculitis; more severe cases may require a penicillinase-resistant systemic antibiotic, such as cephalexin, dicloxacillin, or cloxacillin. In chronic recurrent folliculitis, daily application of a benzoyl peroxide lotion or gel can facilitate resolution. Folliculitis due to M. furfur can be treated with topical application of selenium sulfide 2.5% lotion, applied daily for 5 to 10 minutes for 14 days. Topical miconazole, clotrimazole, ketoconazole, or tolnaftate is also effective. Treatment of M. furfur folliculitis in patients infected with human immunodeficiency virus requires the use of an oral imidazole (e.g., ketoconazole, 200 mg/day for 10 to 14 days).97 Sycosis barbae is a deeper, more severe, recurrent, inflammatory form of folliculitis due to Staphylococcus aureus and involving the entire depth of the follicle. Erythematous, follicular papules and pustules develop on the chin, upper lip, and angle of the jaw, primarily in young black males. Papules can coalesce into plaques, and healing may occur with scarring. Affected individuals are frequently found to be nasal carriers of S. aureus. Treatment with warm saline compresses and topical antibiotics such as mupirocin is generally effective. More extensive, recalcitrant cases may require therapy with a systemic b-lactamase-resistant antibiotic as well as elimination of S. aureus carriage.54 Antibiotics effective against CA-MRSA may be necessary in communities with a high rate of CA-MRSA infections. Folliculitis due to gram-negative organisms occurs primarily in patients with acne vulgaris who have received long-term therapy with broad-spectrum systemic antibiotics. A superficial pustular form, due to Klebsiella spp., Enterobacter spp., Escherichia coli, or Pseudomonas aeruginosa, occurs around the nose and spreads to the cheeks and chin. A deeper, nodular form on the face and trunk is caused by Proteus spp. Culture of infected follicles is necessary to establish the diagnosis. Treatment consists of incision and drainage of the deeper, larger cysts; topical antibiotics (e.g., bacitracin, mupirocin); or selection of an oral antibiotic based on the susceptibility test results of the pathogenic organism. For severe, recalcitrant cases, 13-cis-retinoic acid (Accutane), 1 mg/kg per day, is helpful; however, because of the potential for severe side effects, use of 13-cis-retinoic acid should be limited and only prescribed by experienced providers. Hot-tub folliculitis is caused by Pseudomonas aeruginosa, predominantly serotype O mX11. Lesions are either pruritic papules and pustules or deeply erythematous or violaceous nodules that develop 8 to 48 hours after exposure. Lesions are most dense in areas of exposure to water and abrasive garments (e.g., bathing suits). Fever, malaise, and lymphadenopathy develop occasionally. The organism is isolated from purulent lesions. The eruption usually resolves spontaneously within 1 to 2 weeks, often leaving postinflammatory hyperpigmentation. Sometimes, topical agents with antipseudomonal activity, such as potassium permanganate and gentamicin cream, are necessary. Consideration should be given to the use of a systemic antibiotic such as ciprofloxacin in adolescent patients with constitutional symptoms. Immunocompromised patients, especially those with neutropenia or neutrophil dysfunction, are susceptible to complications of Pseudomonas folliculitis, such as cellulitis, and are advised to avoid hot-tub bathing.98 Occasionally, immunocompetent children have severe local or systemic complications of Pseudomonas folliculitis, including BSI, multiple metastatic foci, neutropenia, and shock.72,99

FURUNCLES AND CARBUNCLES A furuncle can originate from preceding folliculitis or can arise initially as a deep-seated, tender, erythematous, perifollicular nodule. Central necrosis and suppuration lead to rupture and discharge of a

440

SECTION

J

Skin and Soft-Tissue Infections

central core of necrotic tissue, destruction of the follicle, and scarring (Figure 72-4). Sites of predilection are hair-bearing areas on the face, neck, axillae, buttocks, and groin. Pain can be intense if the lesion is situated in an area where the skin is relatively fixed, such as in the external auditory canal or over nasal cartilage. Individuals with furuncles usually have no constitutional symptoms, although BSI can occur, particularly in association with malnutrition or manipulation of the lesion.100 Rarely, severe lesions on the upper lip or cheek can lead to cavernous sinus thrombosis.101 A carbuncle is an infection of a group of contiguous follicles, with multiple drainage points, and inflammatory changes in surrounding connective tissue. Carbuncles can be accompanied by fever, leukocytosis, and BSI. The causative agent is almost always Staphylococcus aureus, which has a predilection for binding to abraded perifollicular skin. Conditions that predispose to furuncle formation include obesity, hyperhidrosis, maceration, friction, and pre-existing dermatitis. Furunculosis is common in individuals with low serum iron concentration, diabetes mellitus, malnutrition, human immunodeficiency virus infection, or other immunodeficiency states. Recurrent furunculosis is frequently associated with carriage of S. aureus in the nares, axillae, or perineum or with sustained close contact with someone who is a carrier. Other bacteria or fungi occasionally cause furuncles or carbuncles; Gram stain and culture of purulent exudate are indicated. Treatment consists of frequent application of a hot, moist compress to facilitate drainage. Large lesions may require surgical drainage. Carbuncles as well as large or multiple furuncles, particularly those located on the central face, should be treated with an oral penicillinase-resistant antibiotic, or clindamycin or trimethoprimsulfamethoxazole in areas where CA-MRSA is prevalent. Penicillinallergic patients can be treated with clindamycin or erythromycin, although macrolide resistance is high among MRSA and MSSA in many geographic areas. Data on decolonization are extremely limited for MRSA (see Chapter 115, Staphylococcus aureus). S. aureus carriage can be temporarily reduced in some instances by application of mupirocin cream three times daily to the anterior nares for 5 days. Attention to personal hygiene and use of chlorhexidene wash may be beneficial. Patients with frequent recurrences may benefit from prophylactic or pre-emptive therapy.

HIDRADENITIS SUPPURATIVA Hidradenitis suppurativa is a chronic, inflammatory, suppurative disorder of the apocrine glands in the axillae and anogenital region, and occasionally on the scalp, the posterior aspect of the ears, breasts, and around the umbilicus. Onset usually occurs during puberty or early adulthood. Solitary or multiple painful, erythematous nodules, deep abscesses, and contracted scars are confined to areas of skin containing apocrine glands. When the disease is severe and chronic, sinus tracts, ulcers, and thick, linear fibrotic bands develop. Hidradenitis suppurativa tends to persist for many years, punctuated by relapses and partial remissions. Complications include cellulitis, ulceration, and abscesses that burrow into adjacent structures in the anogenital region, forming fistulae to the urethra, bladder, rectum, or peritoneum. Episodic inflammatory arthritis develops in some patients. A minority of patients have the follicular occlusion tetrad consisting of hidradenitis suppurativa, acne conglobata, dissecting cellulitis of the scalp, and pilonidal sinus inflammation. The disease is probably initiated by plugging of apocrine gland ducts with keratinous debris. Bacterial infection, particularly with Staphylococcus aureus, Streptococcus anginosus group, Escherichia coli, and possibly anaerobic streptococci, appears to be important in the progressive dilatation below the obstruction. The underlying mechanism of hidradenitis suppurativa is controversial, but it appears to be an androgen-dependent condition.102 Early lesions are often mistaken for infected epidermal cysts, furuncles, scrofuloderma, actinomycosis, cat-scratch disease, granuloma inguinale, or lymphogranuloma venereum. Sharp localization to areas of the body with apocrine glands, however, should suggest hidradenitis. When involvement is limited to the anogenital region, the condition can be difficult to distinguish from, and can coexist with, Crohn disease. Patients should be counseled to avoid tight-fitting clothes, because occlusion can exacerbate the condition. The effectiveness of topical antibiotics is limited. Systemic antibiotics, chosen on the basis of results of culture and susceptibility tests, should be administered in the acute phase. Empiric therapy can be initiated with tetracycline, doxycycline, or minocycline (for patients older than 8 years); clindamycin or cephalosporins are also effective. Some patients require long-term treatment. Intralesional triamcinolone acetonide (5 to 10 mg/mL) is often helpful in early disease. The addition of prednisone (40 to 60 mg/day for 7 to 10 days in adolescents, tapering gradually as inflammation subsides) to the regimen of patients whose disease responds poorly to antibiotics may decrease fibrosis and scarring. Systemic retinoids may be helpful. Ultimately, surgical measures may be required for control or cure. Tumor necrosis factora inhibitors have recently been found to be of use in some cases.103,104

ANTHRAX

Figure 72-4. Furuncle due to Staphylococcus aureus. (Courtesy of J.H. Brien.)

Anthrax infection in the United States has been almost entirely limited to cutaneous infection of individuals who work with contaminated animal products, such as meat, carcasses, animal hair, and wool, although cases resulting from acts of terror have been reported.105 Primary infection develops approximately 2 to 5 days after inoculation of Bacillus anthracis (see Chapter 129, Bacillus Species (Anthrax)) organisms into the skin. The classic lesion is a “malignant pustule” that develops on exposed sites. A painless papule evolves into a serosanguineous vesicle or bulla, which is surrounded by brawny, nonpitting edema. The site becomes progressively discolored, hemorrhagic, and necrotic, with central eschar formation. Satellite vesicles and regional adenitis can be present, but lymphangitis is absent (Figure 72-5). The lesion should not be incised and drained, because this may precipitate bacteremia. Parenteral penicillin G should be given until edema subsides and then continued for 3 to 5 days after the lesion becomes sterile. Alternative antibiotics involve erythromycin, tetracycline, and ciprofloxacin.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Cellulitis and Superficial Skin Infections

A

CHAPTER

72

441

B

Figure 72-5. Cutaneous anthrax. Note ulcer with vesicular ring, induration, and erythema (A). As eschar forms, induration lessens, surrounding desquamation occurs, but erythema persists (B). Source: http://www.bt.cdc.gov/agent/anthrax/anthrax-images.

DIPHTHERIA Cutaneous infection with Corynebacterium diphtheriae (see Chapter 130, Corynebacterium diphtheriae (Diphtheria)) occurs primarily in malnourished, unimmunized children in crowded, unsanitary conditions. Outbreaks of cutaneous infection have occurred in the United States, primarily in Seattle in the 1970s and 1980s among Native Americans and indigent alcoholics. Secondary infection of skin affected by pre-existing trauma, dermatitis, or pyoderma with Streptococcus pyogenes is the most common form of cutaneous diphtheria in the United States. The infected site is erythematous, edematous, and purulent and is usually partially covered with a membrane. Primary cutaneous diphtheria, a disease mainly of tropical environments, begins as a tender papulopustule that becomes a punched-out ulcer with a membranous base and edematous, heapedup, bluish margins.106,107 Cutaneous diphtheria can serve as an important source for person-to-person transmission of pathogenic organisms; greater shedding into the environment appears to occur from a cutaneous compared with a respiratory source. The incidence of neurologic symptoms due to elaboration of toxin from C. diphtheriae in ulcerated cutaneous lesions is low (3% to 5%), less common than that after symptomatic pharyngeal infection. Additional complications are secondary bacterial skin infection, myocarditis, nephritis, gastritis, and hepatitis. Antitoxin is probably of no value for the treatment of cutaneous diphtheria, but some authorities recommend use of 20 000 to 40 000 units of intravenous antitoxin.108 If this agent is administered, it must be given promptly, because the antibodies neutralize toxin only prior to its entry into cells. Thorough cleansing, together with procaine penicillin (25 000 to 50 000 U/kg per day in two divided doses), or erythromycin (40 to 50 mg/kg per day) for 14 days, is the treatment of choice. Erythromycin resistance has been reported among some isolates of C. diphtheriae.105 A second culture of a specimen collected 2 weeks after completion of therapy is necessary to document eradication of the organism.

PITTED KERATOLYSIS Pitted keratolysis occurs most frequently in people whose feet are moist for prolonged periods, because of conditions such as hyperhidrosis, or after prolonged wearing of boots or immersion in

water. Lesions consist of 1- to 7-mm, irregularly shaped, superficial erosions of the horny layer on the soles, particularly at weight-bearing sites. Brownish discoloration of involved areas may be apparent. The condition is usually asymptomatic but frequently is malodorous. A rare, painful variant occurring predominantly in soldiers manifests as thinned, erythematous to violaceous plaques in addition to the typical pitted lesions. Scaly, ringed collarettes of keratolysis can form on the palms. The most likely etiologic agent of pitted keratolysis is a species of Corynebacterium that is able to invade soft, macerated skin. A potassium hydroxide-treated preparation of scrapings from the lesion demonstrates filamentous organisms. Actinomyces spp., Dermatophilus spp., and Micrococcus spp. have also been isolated from lesions and implicated as pathogens. Avoidance of moisture and maceration leads to slow, spontaneous resolution of the infection.1 Effective therapeutic regimens include topical application of 2% buffered glutaraldehyde, aluminum chloride 25% in alcohol solution (Drysol), and topical erythromycin, clindamycin, or clotrimazole.

ERYTHRASMA Erythrasma is a benign, chronic, superficial infection caused by the filamentous diphtheroid Corynebacterium minutissimum. Predisposing factors include heat, humidity, obesity, skin maceration, and poor hygiene. Approximately 20% of healthy individuals with erythrasma have involvement of the toe webs. Other commonly affected sites are moist intertriginous areas such as the groin and axillae; the inframammary and perianal regions are occasionally involved. Sharply demarcated, irregularly bordered, brownish red, slightly scaly, pruritic patches are characteristic. C. minutissimum is a complex of related organisms that produce porphyrins, which fluoresce a brilliant coral-red color under ultraviolet light. The diagnosis is readily made and is differentiated from both dermatophyte infection and tinea versicolor by Wood lamp examination. Bathing within 20 hours before the Wood lamp examination can remove the water-soluble porphyrins. Staining of skin scrapings with methylene blue or Gram stain reveals the pleomorphic, filamentous coccobacilli. Most cases represent colonization, are asymptomatic, and require no therapy. Effective treatment consists of topical erythromycin,

442

SECTION

J

Skin and Soft-Tissue Infections

clindamycin, miconazole, or Whitfield ointment, or a 10- to 14-day course of oral erythromycin. Recurrence can be minimized by the use of an antibacterial soap or astringent such as 10% to 20% aluminum chloride in anhydrous ethyl alcohol.

CHAPTER

73

Erythematous Macules and Papules John Browning and Moise Levy

Erythematous macules and papules are the most common primary lesions seen during acute febrile illness in children. Lesions can result during infection from myriad agents, including viruses, bacteria, treponemes, rickettsia, protozoa, helminths, and fungi. West Nile virus infection is one of the more recent examples of viral pathogens that can cause a macular and popular cutaneous eruption.1 Almost all viral and many bacterial exanthems that occur in the first years of life are of this type, and many conditions that ultimately manifest purpuric, vesicular, urticarial, or ulcerative cutaneous lesions first appear as erythematous macules or papules. Macules are circumscribed, flat, discolored lesions that are not palpable. If an individual lesion is > 1 cm in diameter, the term patch is used. When accompanying an infectious condition, both types of lesions are usually erythematous or purpuric. Erythematous macules, caused by local dilation of dermal blood vessels, can be identified by their blanching on pressure; petechiae or purpuric macules, caused by extravasation of red blood cells, do not blanch (see Chapter 75, Purpura). Papules are circumscribed, solid, elevated lesions, less than 1 cm in diameter. The term plaque refers to an elevated (flat-topped) lesion that is greater than 1 cm in diameter. Papules accompanying an infectious process usually form within a macule after inflammation, localized edema, or cellular infiltration that causes elevation of the skin surface. Papules can be identified clinically by touch or by the shadow they cast when illuminated by a light source placed tangentially on the skin. Papules that are covered with fine scale, such as in pityriasis rosea or guttate psoriasis, are referred to as papulosquamous lesions. Major noninfectious causes of erythematous macules and papules are shown in Table 73-1. Some authorities consider the term maculopapular to be contradictory, because macules are flat and papules are raised, and prefer the terms morbilliform (resembling morbilli, or measles) when lesions coalesce2; rubelliform (resembling rubella) when lesions remain discrete; and scarlatiniform (resembling scarlet fever) when lesions have the characteristic sandpaper appearance and are exaggerated in skinfolds. Despite objections, the term maculopapular is in widespread use and is generally understood to refer to a cutaneous eruption characterized by the simultaneous presence of erythematous lesions that vary in morphology from flat to slightly elevated.

ETIOLOGY Among the many infectious causes of erythematous macules and papules in children, the most common are the enteroviruses.3,4 At least 36 enteroviruses cause human disease characterized by rash. Most of the exanthems are macular, papular, or morbilliform, occurring in association with mild, febrile upper respiratory tract or gastrointestinal tract illness. Enteroviral exanthems are protean, however, and include vesicular, vesiculobullous, scarlatiniform, urticarial, and purpuric eruptions.

TABLE 73-1. Major Noninfectious Causes of Erythematous Macules and Papules in Infants and Children Disease

Skin Lesions

Acne vulgaris Actinic prurigo Acute graft-versus-host reaction Acute sunburn reaction Arthropod bite Atopic dermatitis Chemotherapy-associated acral erythema Contact dermatitis Dermatomyositis Eosinophilic pustular folliculitis Granuloma annulare Hemangiomatosis, disseminated Drug reaction Juvenile rheumatoid arthritis Kawasaki syndrome Livedo reticularis Miliaria rubra Nevus flammeus Panniculitis, cold Papular urticaria Phototoxic or photoallergic dermatitis Pityriasis lichenoides Pityriasis rosea Pityriasis rubra pilaris Polymorphous light reaction Psoriasis Subcutaneous fat necrosis Systemic lupus erythematosus Telangiectasias

P P M, Pa, P Pa P Pa, P, Pl M, Pa M, Pa, P, Pl M, Pa, P, Pl P M, P, Pl P M, Pa, P, Pl M, Pa, P M, Pa, P M, Pa P M, Pa Pl P M, Pa, P, Pl P P, Pl M, Pa, P, Pl P, Pl Pa, P, Pl Pl M, Pa, P, Pl M

M, macule; P, papule; Pa, patch; Pl, plaque.

Because maculopapular rashes are nonspecific, a review of epidemiologic and physical findings is most helpful in establishing a diagnosis. Pertinent questions include: (1) child’s age; (2) season; (3) history of exposure to medications or toxins; (4) history of exposure to a person with an illness, particularly one with a rash; (5) geographic location and recent travel history; (6) immunization history; (7) history of previous illnesses; (8) exposure to domestic or wild animals; (9) pattern and evolution of the rash; and (10) associated signs and symptoms.5 Drug reaction is the most common condition that mimics an eruption of erythematous macules and papules caused by an infectious agent. Such reactions have been described in up to 5% of individuals receiving medication, particularly antibiotic agents (e.g., amoxicillin, cephalosporins, and trimethoprim-sulfamethoxazole).6 It is likely, however, that many children thought to have an allergic reaction to an antibiotic agent are exhibiting a viral exanthem incorrectly attributed to medication. In fact, viral exanthems are more common in children, whereas drug eruptions are more commonly seen in adults. In some cases, exanthems that occur in the setting of concurrent viral infection and antibiotic administration are related to both, but do not recur on administration of the antibiotic once the viral infection has resolved or become latent. In this setting, drug metabolites may disturb the balance between cytotoxic and regulatory mechanisms of T lymphocytes, favoring cytotoxicity over viral-infected cells in the skin.7 There are few maculopapular rashes with sufficient typical features for a diagnosis to be established with certainty.8–10 The first phase of the rash of erythema infectiosum, with a “slapped-cheek” appearance, although strongly suggestive of parvovirus B19 infection, has been described with enteroviral infection (echovirus 9) and can be confused with the flushed face and circumoral pallor of early scarlet fever or erythroderma of toxic shock syndrome. The sudden appearance of a rash (exanthem subitum) following defervescence of a sustained high fever in an infant, though classically associated with human

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Vesicles and Bullae

herpesvirus 6 infection (roseola), has also been noted in outbreaks of echovirus 16 infection (“Boston exanthem”), as well as in infections caused by other echoviruses, coxsackieviruses A and B, adenoviruses, parainfluenza virus type 1, and human herpesvirus 7. A rash similar to that of scarlet fever caused by Streptococcus pyogenes can be seen in conjunction with infection caused by Arcanobacterium haemolyticum, Staphylococcus aureus, parvovirus B19, and echovirus 14, as well as Kawasaki disease. The cutaneous eruption caused by drug hypersensitivity reactions can mimic all of these and other infectious exanthems.6,11 However, the hallmark of drug hypersensitivity reactions is facial edema. Conjunctivitis, pharyngitis, respiratory tract symptoms, abdominal pain, and myalgia accompany a variety of infectious conditions; their presence and specific characteristics help to narrow the spectrum of possible diagnoses (see Chapter 15, Mucocutaneous Symptom Complexes). Pharyngoconjunctival fever suggests infection resulting from adenovirus; wheezing suggests respiratory syncytial virus; pharyngeal vesicles suggest echovirus or coxsackievirus; severe exudative pharyngitis suggests infection with Streptococcus pyogenes or Epstein–Barr virus; myalgia suggests influenza or an enterovirus. Few, if any, findings are diagnostic; even lesions similar to Koplik spots, generally considered pathognomonic of rubeola, have been described in patients with echovirus 9 and coxsackievirus A16 infection. Most macular and papular infections associated with eruptions in childhood are brief, self-limited, and uncomplicated. When it is important to establish a cause, cultures of affected mucous membranes, antigen detection tests, serologic studies, and ancillary laboratory tests (e.g., polymerase chain reaction) and radiographic studies can be useful.

PATHOGENESIS Maculopapular eruptions can be caused by: (1) direct infection of cells in the epidermis, dermis, or vascular endothelium; (2) a host immunologic reaction directed against the infecting organism; (3) circulating toxins; or (4) a combination of these mechanisms. In most cases, infection of dermal blood vessels accompanied or followed by a host immune response produces the rash. The predominant pathogenic factor is often unclear, such as in meningococcal infection, in which endothelial infection, direct action of endotoxin, and an allergic response (Shwartzman reaction) have all been implicated.12–14 Direct infection and replication of rubeola virus in the corneal layer, the keratinocytes of the epidermis, and the superficial dermis appear to initiate the rash in measles.15 Subsequent inflammatory infiltrates of mononuclear cells, producing epidermal edema or spongiosis, contribute further to the eruption. A similar process produces the eruption in rubella, except that viral replication and mononuclear cell infiltration occur throughout the dermis and extend into the subcutis, sparing the epidermis. Invasion of endothelium in dermal capillaries has been confirmed in patients with rubeola and cytomegalovirus infection by electron microscopy and immunofluorescent staining. Acute rickettsial disease (such as typhus, Rocky Mountain spotted fever) is also characterized by infection of small blood vessels; organisms can be demonstrated within vascular tissues with secondary vasodilation, edema, and perivascular cellular infiltrates. A delay between exposure to an organism and the appearance of a cutaneous reaction, coincident with a rise in antibody titers, suggests a role for host immune response in the genesis of many infectious exanthems.16 The observed variation in clinical course among children with altered immune responses is also instructive. An immunocompetent child with measles exhibits a morbilliform rash on the face that spreads rapidly downward to the neck, trunk, and extremities, remaining most concentrated and becoming confluent on the face and trunk. Immunosuppressed children, lacking both adequate cellular and antibody immunity, do not usually have a rash, despite progressive systemic infection that often leads to fatal interstitial (“giant-cell”) pneumonitis.

CHAPTER

74

443

Evidence for the role of immunologic factors in the genesis of some maculopapular exanthems is illustrated with parvovirus B19 infection. Erythema infectiosum rash is seen in immunocompetent children, is not seen in children with chronic viremia, but develops in the latter patients within hours after infusion of intravenous immune globulin. Immune complexes can be identified in the skin of patients with a rash accompanying hepatitis B or chronic meningococcal infection. In the latter condition, for example, unlike the rash of acute meningococcemia, bacteria are absent from skin lesions, endothelial cells are intact, and mononuclear rather than polymorphonuclear cells predominate in the perivascular infiltrate.12 The role of circulating toxins in the development of the erythroderma of S. pyogenes and Staphylococcus aureus infection is well described, although the factors directly responsible for the vasodilation and erythema are unknown.17 The rash of scarlet fever has been attributed to the vascular effects of streptococcal pyrogenic exotoxins A, B, and C. A virtually identical eruption, considered to be a mild variant of staphylococcal scalded-skin syndrome, is caused by staphylococcal epidermolytic exotoxin A or B. In these conditions, dermal vessels show diffuse vasodilation with surrounding edema and mononuclear cell infiltrates, but no evidence of vasculitis. Pyrogenic toxins of Streptococcus pyogenes and Staphylococcus aureus can also act as superantigens to stimulate exuberant responses of T lymphocytes and accessory cells, such as macrophages, Langerhans cells, and keratinocytes. Cytokine release by these stimulated cells is thought to mediate the clinical effects of superantigens, perhaps including the exanthem of Kawasaki disease.18–20

CHAPTER

74

Vesicles and Bullae John Browning and Moise Levy Vesicles and bullae result from a disturbance of cohesion of epidermal cells or components of the basement membrane zone associated with influx of fluid into or beneath the site of disturbance. Although easily recognized in an intact state, primary lesions (vesicles or bullae) often rapidly evolve into erosions, ulcers, or crusts. Diagnosis of vesiculobullous eruptions must be made promptly because, although some are benign, others are rapidly progressive and life-threatening. The skin consists of stratified epidermis, composed primarily of keratinocytes, and an underlying dermis of connective tissue. The epidermis is constantly renewed by mitotic division of the lowermost keratinocyte layer (the basal layer) and by the orderly shedding of fully differentiated keratinocytes from the outermost layer (the stratum corneum) (Figure 74-1). Stratum corneum cells are held together tightly relative to the keratinocytes of the spinous layer, which rely largely on glycoprotein-rich intercellular cement substance and desmosome–tonofilament complexes for cohesion. The basement membrane zone anchors the epidermis to the dermis. Cohesion between layers of the skin is weakest at the stratum corneum–spinous cell transition zone and particularly within the least electron-dense region of the basement membrane zone, called the lamina lucida. A plane of cleavage, or blister, is therefore most apt to develop at these levels (see Figure 74-1). Application of a lateral, sliding force to nonblistered skin (usually at a site adjacent to a bullous lesion) that produces a plane of cleavage is the Nikolsky sign. This maneuver can be useful in the clinical recognition of a blistering disorder but cannot be used to judge the depth of the blister within the skin. When the depth of the plane of cleavage is in doubt, histopathologic examination of a frozen section of denuded skin can provide diagnostic information rapidly.

444

SECTION

J

Skin and Soft-Tissue Infections

Subcorneal Candidiasis Impetigo Staphylococcal scalded-skin syndrome Granular cell layer Spinous layer Dermatophyte infection Scabies Viral infections

Basal cell layer Erythema multiforme Figure 74-1. Sites of vesiculobullous diseases in the skin.

Vesicles and bullae, commonly called blisters, are circumscribed, elevated lesions filled with clear fluid. Depending on the mechanisms responsible for their formation, blisters can contain a combination of edematous or lymphatic fluid, serum proteins, antigen–antibody complexes, and soluble inflammatory mediators. Cellular elements are also often present, including inflammatory cells, erythrocytes, detached epidermal cells, and infectious agents. Vesicles measure < 1 cm in diameter and bullae measure > 1 cm. Vesicles and bullae associated with infection can be any of the following: (1) solitary, such as the lesion of streptococcal blistering dactylitis; (2) localized, as in staphylococcal bullous impetigo; (3) grouped or clustered, as in herpes simplex virus (HSV) infection; (4) arranged linearly, as in shingles, caused by varicella-zoster virus (VZV); or (5) generalized, as in varicella. A vesicle or bulla that is located in the epidermis tends to be flaccid and to rupture easily; when located subepidermally, it is more often tense, has greater structural integrity, and ruptures less easily. As a vesicle or bulla matures, an influx of leukocytes and accumulation of cellular debris can occur, leading to development of a pustule. Depending on the cause of the inflammatory response, a pustule can be infected or sterile. Rupture and detachment of the roof of a subcorneal or intraepidermal blister can form a moist, slightly depressed erosion. Erosion does not extend below the epidermal– dermal junction and heals without scarring. Postinflammatory pigmentary changes, however, can be present for weeks to months. When an unroofed blister extends into the dermis or subcutaneous tissue, it forms an ulcer. Scarring or postinflammatory pigmentary changes can follow healing of a wound that involves the dermis. Erosion or ulceration is accelerated in areas of friction or maceration, such as in the axillae or perineum, and on the mucous membranes of the oropharynx and vagina. Crusts, or scabs, are the dried remnants of serum, blood, and cellular debris; they form quickly over denuded areas. It is not unusual for several types of lesions to be present at the same time in an individual. For example, the eruptions caused by varicella can comprise a variety of lesions in various stages of evolution, including macules, papules, vesicles, and pustules, mixed with erosions, ulcers, crusts, and self-induced linear excoriations. Bullae can form if staphylococcal superinfection occurs (Figure 74-2).

ETIOLOGY Although the inventory of infectious agents capable of causing vesiculobullous rashes is lengthy, the number of common agents is limited to enteroviruses, VZV, HSV, Staphylococcus aureus, and Streptococcus pyogenes. The list of noninfectious conditions that closely mimic the eruption caused by these organisms is also lengthy,

Figure 74-2. Chickenpox lesions, infected secondarily with Staphylococcus aureus, forming crusted plaques and bullae on the face of an infant.

often leading to difficult or delayed diagnosis.1–3 The cause can frequently be established by consideration of season, patient age, history of recent exposure to infectious agents or medications, previous disease, and concurrent symptoms, as well as the morphology, distribution, and evolution of the eruption. For example, although both varicella and enteroviruses cause aphthous erosions in the mouth and exanthems that appear identical in the early papulovesicular stage, distinction is almost always possible as the lesions mature. Additionally, enterovirus eruptions rarely itch, ulcerate, or crust. Distribution of lesions also provides clues to cause of the lesion. For example, the characteristic appearance of some coxsackievirus A and B and enterovirus 71 lesions in a hand, foot, and mouth (and, often, buttocks) pattern is a distinguishing feature of these agents and contrasts with the more centralized truncal rash of VZV. The dermatomal rash of zoster is also distinctive and easily recognized, but can occur with HSV (zosteriform simplex) (Figure 74-3) and echovirus 6 infections. When the cause of vesicular lesions remains in doubt, culture, biopsy for histopathologic evaluation, immunofluorescence staining, and, occasionally, electron microscopy are helpful, particularly in excluding noninfectious bullous disorders.1 Bullae are less common. Bullous impetigo caused by Staphylococcus aureus, occurring independently or as a superinfection of varicella lesions, is probably the most common cause of bullae. Other bullous eruptions, such as staphylococcal scalded-skin syndrome (SSSS); erythema multiforme (with HSV infection); the hemorrhagic bullae that can accompany septicemia caused by gram-negative organisms, particularly Pseudomonas aeruginosa (ecthyma gangrenosum) and Neisseria meningitidis; and necrotizing soft-tissue infection due to Streptococcus pyogenes must be considered. The differential diagnosis of a vesiculobullous eruption is extensive. There are few experienced physicians who have not misdiagnosed varicella as insect bites – and vice versa – or confused

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Vesicles and Bullae

CHAPTER

74

445

A

Figure 74-3. A young girl hospitalized for febrile pneumonia had reactivation herpes simplex virus 1 in a dermatomal distribution on her face. (Courtesy of J.H. Brien.)

shingles (Figure 74-4A) with a linear vesicular rash caused by contact dermatitis (Figure 74-4B). Dyshidrotic eczema and dermatophytid reactions on the palms and soles can be mistaken for enteroviral disease in a child with fever resulting from another cause. The appearance of bullae caused by thermal injury or a hypersensitivity response to insect bite is identical to that seen with bullous impetigo. Distinction between the skin lesions of drug-induced toxic epidermal necrolysis (TEN) and SSSS can require histopathologic examination of a skin biopsy specimen; TEN results in full-thickness epidermal necrosis, whereas the cleavage plane in SSSS is subcorneal. Distinguishing between these conditions is particularly important, because mortality rates as high as 30% have been reported with TEN and avoidance of the offending drug is crucial in preventing recurrence.4

PATHOGENESIS Three basic patterns of intraepidermal vesicle formation have been described: spongiosis, acantholysis, and cytolysis.5 Spongiosis, or intercellular epidermal edema, is caused by the influx of fluid between cells, stretching the intercellular desmosomal attachments of keratinocytes and giving the tissue a spongelike appearance due to formation of microvesicles. Marked spongiosis, aided by hydrostatic pressure, can result in rupture of intercellular bridges, formation of intraepidermal clefts, and confluence into a macroscopic vesicle. Spongiosis is characteristic of many viral and superficial fungal (e.g., dermatophytic) infections and also of cutaneous autosensitization reactions caused by infectious agents (e.g., dermatophytid reaction). Acantholysis is the loss of intercellular cohesion resulting from direct damage to desmosomal attachments between keratinocytes. As cells separate, fluid accumulates between cells, and clefts develop and coalesce to form a blister. In SSSS and bullous impetigo (a localized form of SSSS), disruption of cellular adhesion appears to be caused by staphylococcal epidermolytic toxin that disrupts desmosomes by proteolysis.6 Secondary acantholysis occurs when intercellular spongiotic fluid exerts sufficient pressure to rupture intercellular attachments. Cytolytic blisters, usually associated with viral infection, form as a result of epidermal cell necrosis and death. As damaged cells degenerate, they leave spaces into which fluid accumulates and forms a vesicle. In the case of most viral infections, the cytolytic process is initiated by intracellular viral replication and appears to be aided by the host immune response to viral antigens present in infected cells. Characteristic eosinophilic inclusion bodies are found in the

B Figure 74-4. (A) Bullae on the dorsal hand and forearm of an infant resulting from reactivation of varicella-zoster virus, without superinfection. (B) Vesicles and bullae on the hand and wrist of a girl with contact dermatitis caused by poison oak.

cytoplasm of keratinocytes infected with coxsackievirus, orf virus, and paravaccinia virus (milker’s nodules), whereas eosinophilic inclusion bodies are in the nuclei of keratinocytes infected with HSV and VZV. Examination of cells scraped from the base of a virus-induced blister can reveal swollen keratinocytes with cytoplasmic vacuolization (ballooning degeneration), which causes marked swelling of the infected cells and often leads to acantholysis. Reticular degeneration, a process in which epidermal cells become greatly distended by intracellular edema, can lead to rupture of epidermal cells and usually develops in association with spongiosis and ballooning degeneration.7 Reticular degeneration is nonspecific and is seen in several viral infections, such as HSV and VZV, hand, foot, and mouth disease, orf virus and milker’s nodules, and several noninfectious diseases, such as erythema multiforme, fixed drug eruption, and irritant contact dermatitis.8 A more specific finding is formation of acantholytic balloon cells with one or several nuclei (multinucleated giant cells) at the base of vesicles; in addition to herpesvirus infection, multinucleated giant cells can be seen in measles and hand, foot, and mouth disease.9 Occasionally, cutaneous infection can lead to a severe inflammatory infiltrate in the dermis or epidermis with destruction of tissue in the upper dermis, forming a dermolytic blister. Epidermal and dermal necrosis is responsible for formation of bullae in bullous erysipelas, whereas extensive edema of the upper dermis can lead to subepidermal blister formation in dermatophytosis. The processes of spongiosis, acantholysis, and cytolysis seldom occur independently. Frequently, the histopathology of cutaneous viral infection includes a combination of intracellular and intercellular edema, secondary acanthosis, and cytolysis. For example, vesicles

446

SECTION

J

Skin and Soft-Tissue Infections

caused by infection with HSV, VZV, or hand, foot, and mouth disease develop initially from ballooning of keratinocytes, and, to a lesser extent, from spongiosis; ballooning degeneration then leads to secondary acantholysis. As intraepidermal spongiosis intensifies and reticular degeneration of keratinocytes occurs, intraepidermal clefts coalesce into blisters. Blisters can rupture, and become subepidermal.9

CHAPTER

75

Purpura

A

John Browning and Moise Levy

Purpuric lesions are of great concern because of their association with life-threatening infections from multiple causes.1 Lesions result from vascular injury or disorders of hemostasis, particularly those associated with platelet depletion or dysfunction, which leads to extravasation of red blood cells into the skin or mucous membranes.2 Purpura can be subdivided into three forms on the basis of size, depth, and extent of hemorrhage, as follows: petechiae, ecchymoses, and palpable purpura.

TYPES OF LESIONS Petechiae Petechiae are purpuric macules up to 2 mm in diameter, caused by extravasation of blood from capillaries (Figure 75-1). They can be found in the skin, mucous membranes, conjunctivae, or retinae. When extravasation of red blood cells occurs in the nailbed, it usually appears in linear streaks oriented along the axis of the nail in the form of “splinter hemorrhages” (which are especially meaningful if located in the proximal nail). Petechiae often appear over a short time in “crops” or “showers”; isolated petechiae are a common finding in early bloodstream infection (BSI). Initially scarlet or brick-red, the breakdown of heme pigments in the skin causes petechiae to darken briefly to purple and then fade to green and yellow-brown, disappearing in most cases over 2 to 3 days. Depending on the cause of petechiae, individual lesions can evolve into ecchymoses/purpura, palpable purpura, vesicles, pustules, or necrotic ulcers. Petechiae can be distinguished clinically from erythematous macules caused by vasodilation and from most telangiectasias (“spider hemangiomas”) by pressing a glass slide or lens firmly over the lesion. Lesions that blanch under pressure do not represent extravascular blood; those that persist are purpuric. Occasionally, small angiomas and angiokeratomas can be confused with petechiae, although these lesions are blanchable.

Ecchymoses Ecchymoses, or bruises, are areas of bleeding into the skin that differ from petechiae only by their larger size. When petechiae occur in an individual with a particularly severe bleeding tendency, they can evolve rapidly into ecchymoses, sometimes over a period of several minutes. Ecchymoses go through the same color changes as described for petechiae; however, depending on their size, the interval from onset to disappearance is usually 1 to 3 weeks. Purpura fulminans is an acute, severe, and often rapidly fatal form of hemorrhage, occurring most commonly in children infected with Neisseria meningitidis or varicella-zoster virus.2 It is characterized by large confluent purpuric patches, symmetrical in distribution, that

B Figure 75-1. An 8-year old boy has rapidly progressive meningococcemia. Upon first examination, he had predominant petechiae on his arm (A) with exaggeration distal to the site of a prior tourniquet, and petechiae and ecchymoses/purpura on his back (B). (Courtesy of J.H. Brien.)

subsequently undergo necrosis and eschar formation. Symmetrical peripheral gangrene, an ischemic necrosis involving the distal portion of two or more extremities in the absence of large-vessel obstruction, is considered to be a variant of purpura fulminans.3 Similar to purpura fulminans, it occurs most frequently in infants and children.

Palpable Purpura Palpable purpura refers to elevated, firm, hemorrhagic papules or plaques up to several centimeters in diameter, most commonly located on dependent surfaces, such as the lower legs (Figure 75-2), buttocks, or the back in a recumbent patient. Lesions can develop a nodular, vesicular, pustular, necrotic, or ulcerative center. They are usually asymptomatic but can become tender, particularly when nodular or ulcerative. Whereas nonpalpable purpura usually suggests hemorrhage from a platelet or coagulation disorder, palpable purpura is the classic and most typical cutaneous lesion in patients with inflammatory vascular injury (leukocytoclastic vasculitis).4 Palpable purpura can also develop as a result of infectious microemboli (as can especially occur in severe Staphylococcus aureus BSI); these lesions are not vasculitic on histopathologic examination. Osler nodes and Janeway lesions seen in conjunction with bacterial endocarditis are examples of embolic lesions that can be both purpuric and palpable. Osler nodes are small, tender, nodular lesions that occur most commonly in the palms, soles, and pads of the fingers and toes. Janeway lesions are small, painless, hemorrhagic macules, papules, or small nodules, also located primarily on the palms and soles. In many cases, lesions first appear as erythematous macules, petechiae, or ecchymoses before adopting a characteristic form; both can progress to form pustules. Lesions similar to Osler nodes have also been described in typhoid fever, polyarteritis nodosa, systemic lupus erythematosus, and Wegener granulomatosis.3

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Purpura

CHAPTER

75

447

newborn, acquired thrombocytopenia from infection or maternal antiplatelet antibodies must be excluded. Blood dyscrasias should be considered in older children, especially if fever is absent. A platelet count should be obtained from patients with fever and a generalized petechial eruption; however, meningococcal vasculitis may not cause thrombocytopenia early in the course, and viral infections often cause modest thrombocytopenia.

Ecchymoses

Figure 75-2. Purpuric plaque on the lateral distal leg of a girl with Neisseria meningitidis septicemia and purpura fulminans.

ETIOLOGY Petechiae In a retrospective review of causes of petechiae accompanied by fever in hospitalized children, almost 60% had presumed or proven viral illness; approximately 20% had invasive, potentially life-threatening bacterial disease, most commonly septicemia, with or without meningitis caused by N. meningitidis; 20% had a variety of other infectious and noninfectious conditions.5 The relative prominence of viruses as causes of petechial lesions results from their overall prevalence and may also occur because viruses generally cause vasculitis, involving small vessels, whereas bacteria tend to invade vessels of a variety of sizes, more often resulting in palpable lesions.1 Two prospective studies also attributed 60% to 90% of episodes of petechiae accompanied by fever to viral illness.6 Only 2% and 8% of children in these two series had invasive bacterial disease, which was caused by N. meningitidis. An ill appearance was 100% sensitive for BSI in one study of febrile children with petechiae.7 Approximately 10% had an upper respiratory tract infection caused by Streptococcus pyogenes, an association emphasized in the past.8 Although petechiae on the palate are also characteristic of streptococcal pharyngitis,9 they have been described in a number of other conditions, including Epstein–Barr virus infection, rubella, roseola, viral hemorrhagic fevers, thrombocytopenia, and palatal trauma. Parvovirus B19 can cause papular purpuric, “gloves-and-socks” syndrome characterized by confluent erythema and pinpoint erythematous papules, evolving to petechiae and purpuric macules and patches on the palms and soles and, occasionally, the dorsum of the hands and feet in conjunction with fever and leukopenia.10 Repeated severe retching or coughing and rapid and sustained elevations in intrathoracic pressure can cause punctate capillary bleeding in the vascular bed drained by the superior vena cava. Although these eruptions appear above the nipple line, not all petechiae in this distribution can be dismissed as benign; in one study, 25% of children described with fever and petechiae confined to the upper torso had bacterial disease.5 Petechiae confined to an extremity distal to tourniquet placement have no clinical significance. Trapping of the presenting part in the birth canal during delivery is a common mechanical cause of petechiae among newborns; in these instances, lesions are usually localized. When petechiae are widespread in the

Ecchymoses resulting from infection in children are uncommon. When they occur, they most frequently result from BSI caused by N. meningitidis and are often accompanied by coagulopathy.2 Other gram-negative bacteria, as well as gram-positive organisms, including Staphylococcus aureus, Streptococcus pyogenes, and S. pneumoniae, can produce a similar clinical picture.2,11–13 Rocky Mountain spotted fever, dengue fever, and other viral hemorrhagic fevers can cause ecchymoses in children. Trauma is the most common cause of ecchymoses in children. When it is unexplained or widespread, physical abuse must be considered. Purpuric rashes caused by vasculitis (e.g., Henoch– Schönlein purpura) or defects in clotting factors (e.g., hemophilia) are relatively less common. When coagulation disorders present as large ecchymoses, they are often accompanied by bleeding into joints or muscles. Ecchymoses that are characteristic of abnormalities in platelet number or function are often superimposed on a petechial eruption. Purpuric eruptions caused by hypersensitivity reactions, vasculitis caused by drugs, infectious agents, or immune deregulation, and capillary fragility caused by prolonged corticosteroid therapy must also be considered.14 Despite the well-known association with septicemia, purpura fulminans and symmetrical peripheral gangrene can also occur in infants and young children in association with relatively benign provocative infections, such as varicella or S. pyogenes.2,15 Measles and the bacterial and rickettsial agents that cause ecchymoses are other causes of purpura fulminans. Congenital or acquired protein C or S or antithrombin III deficiency can cause a similar clinical picture.

Palpable Purpura Palpable purpura most often results from drug-induced, necrotizing, small-vessel vasculitis.1,2 Among the numerous possible infectious causes, meningococcal, staphylococcal, and gonococcal are most common (Boxes 75-1 and 75-2).

PATHOGENESIS Purpuric skin lesions attributable to infection can result from one of several pathogenic processes, including the following, alone or in combination: (1) dissemination of microorganisms through the bloodstream, leading to direct invasion of dermal capillaries and, in some cases, formation of microabscesses in the adjacent papillary dermis; (2) immune complex formation and vasculitis generated by host response to infection; (3) release of toxins by infectious agents; (4) coagulopathy due to thrombocytopenia or disseminated intravascular coagulation (DIC). Examples of capillary invasion, with demonstration of microorganisms in the capillary lumen or endothelium, include bacteremia caused by N. meningitidis, N. gonorrhoeae, Pseudomonas aeruginosa, and organisms contained in microemboli released during severe BSI or during acute and subacute bacterial endocarditis (particularly Staphylococcus aureus and Streptococcus spp.) or intravascular catheter-related infections. Diffuse vasculitis with multiplication of rickettsiae within endothelial cells of capillaries, arterioles, and venules is characteristic of Rocky Mountain spotted fever and epidemic typhus. In patients with a purpuric eruption associated with

448

SECTION

J

Skin and Soft-Tissue Infections

BOX 75-1. Infectious Causes of Purpura in Children and Neonates

BOX 75-2. Major Noninfectious Causes of Purpura

CHILDREN Bacteria Bacteroides fragilis, Bartonella hensleae, Borrelia spp., Campylobacter jejuni, Capnocytophaga canimorsus,a Enterococcus spp., Escherichia coli, Francisella tularensis, Haemophilus influenzae type b, Klebsiella spp., Leptospira spp., Listeria monocytogenes, Moraxella catarrhalis, Mycobacterium leprae, Mycobacterium tuberculosis, Neisseria meningitidis,a Neisseria gonorrhoeae,a Pseudomonas aeruginosa,a Salmonella typhimurium, Staphylococcus aureus, Streptobacillus moniliformis, Streptococcus pneumoniae, Streptococcus pyogenes,a viridans streptococci, Yersinia pestis, Yersinia enterocolitica Fungus Candida, Aspergillus, Fusarium species and agents of mucormycosis Helminths Ascaris spp., Strongyloides stercoralis (hyperinfection syndrome), Trichinella spiralis Mycoplasma Mycoplasma pneumoniae Protozoa Plasmodium falciparum Rickettsiae Ehrlichia canis, Ehrlichia chafeensis, Ehrlichia equi, Anaplasma phagocytophilum, Rickettsia akari, Rickettsia prowazekii (epidemic typhus),a Rickettsia rickettsii (Rocky Mountain spotted fever),a Rickettsia typhi Viruses Adenoviruses; Colorado tick fever virus; coxsackievirus A4, A9, B2–B5; cytomegalovirus,a echovirus 3, 4, 7, 9, 18; dengue fever virus; Epstein–Barr virus; hantavirus; hepatitis virus A, B,a Ca; human immunodeficiency virus; Lassa virus; Marburg virus; parvovirus B19a; respiratory syncytial virus; rotavirus; rubella virus; rubeola virus (typical and atypical measles); varicella-zoster virus; yellow fever and other hemorrhagic fever viruses (e.g., Ebola virus) NEONATES Bacteria Gram-positive and gram-negative bacteria associated with septicemiaa Treponema pallidum Protozoa Toxoplasma gondii Viruses Enteroviruses; cytomegalovirus; herpes simplex virus types 1, 2; rubella virus

PLATELET DISORDERSa Thrombocytopenia Decreased production of platelets: drugs, marrow infiltration or aplasia, toxins Decreased survival of platelets: disseminated intravascular coagulation, drugs, hemolytic–uremic syndrome, idiopathic and autoimmune thrombocytopenic purpura, prosthetic heart valve Hypersplenism Sequestration Platelet dysfunction (inherited or acquired) Thrombocytosis VASCULAR DISORDERS Noninflammatory purpura Amyloidosis, angiokeratoma corporis diffusum, antiphospholipid syndrome, atrophie blanche, corticosteroid therapy, Cushing disease, Ehlers–Danlos syndrome, fat embolism, hereditary hemorrhagic telangiectasia, increased capillary pressure (choking, coughing, forced restraint, tourniquet, vomiting), Langerhans cell histiocytosis, progressive pigmentary dermatosis (pigmented purpura), scurvy, trauma Vasculitis and inflammatory purpura Behçet disease, Churg–Strauss syndrome, cystic fibrosis, dermatomyositis, drug- or toxin-induced hypersensitivity vasculitis (nonsteroidal antiinflammatory drugs, penicillin, quinidine, propylthiouracil, sulfonamides, allopurinol, diphenylhydantoin, phenothiazines, thiazides), dysproteinemias (Waldenström Adom hyperglobulinemic purpura, Waldenström Adom macroglobulinemia, essential mixed cryoglobulinemia, paraproteinemia), Henoch–Schönlein purpura, inflammatory bowel disease, Kawasaki syndrome, leukemia, lymphoma, pityriasis lichenoides, polyarteritis nodosa, pyoderma gangrenosum, rheumatoid arthritis sarcoidosis, scleroderma, primary and acquired immunodeficiencies (graft-versus-host disease, human immunodeficiency virus infection), second component of complement deficiency, serum sickness, Sjögren syndrome, Sweet syndrome (acute febrile neutrophilic dermatosis), systemic lupus erythematosus, urticarial vasculitis, Wegener granulomatosis COAGULATION DISORDERSb Clotting factor deficiencies Depletion of clotting factors through increased use or proteolysis, inadequate production, presence of inhibitor of coagulation, synthesis of abnormal form of clotting factor Thrombotic disorders Antiphospholipid syndrome, antithrombin III deficiency, plasminogen deficiency, protein C deficiency, protein S deficiency

a

Most common.

a

Primarily petechiae. Primarily ecchymoses.

b

cytomegalovirus infection and measles, light and electron microscopic examination of a skin biopsy specimen demonstrates viral particles and characteristic histologic changes within endothelial cells. Several enteroviral infections are also characterized by this process. Depending on the invasive properties of the microorganisms involved, infection of vascular tissue leads to varying degrees of endothelial cell degeneration, thrombosis of involved vessels, and areas of microinfarction with subsequent extravasation of erythrocytes. A local inflammatory response, with development of edema and leukocytic infiltration, is responsible for the subsequent papular lesions that often evolve at sites of dermal hemorrhage. Direct vascular injury involving larger vessels, such as arterioles and venules, can lead to extensive thromboses with widespread areas of tissue hemorrhage, infarction, and skin necrosis. Immune-mediated vasculitis is responsible for rashes that are characteristic of infections caused by hepatitis B, atypical measles, chronic meningococcemia, and endocarditis. The proposed mechanism of this type of rash is the deposition of circulating immune complexes in cutaneous blood vessels. The resulting inflammation, vascular damage, and necrosis result in an eruption most commonly manifested as palpable purpura (but occasionally as petechiae alone). Based on the presumed pathogenesis and histopathologic features, this condition is described as hypersensitivity vasculitis, cutaneous necrotizing venulitis, or leukocytoclastic vasculitis. Henoch–

Schönlein purpura is the prototypic leukocytoclastic vasculitis of childhood; it frequently develops after an upper respiratory tract infection, most commonly Streptococcus pyogenes, and involves IgA immune complex deposition in postcapillary venules. Toxin-mediated rashes, so-called toxic erythemas, are typified by the scarlatiniform petechial eruptions of S. pyogenes and Staphylococcus aureus infection. In streptococcal scarlet fever, the pathogenesis is believed to be vasodilation caused by erythrogenic toxin acting in concert with a delayed-type hypersensitivity reaction to one or more bacterial antigens. A large number of infections are accompanied by thrombocytopenia with the consequent development of a petechial or ecchymotic rash. Although infection-associated purpura fulminans and symmetrical peripheral gangrene are generally considered to be consequences of DIC, not only can these lesions occur in the absence of DIC, but no correlation exists between the severity of the purpura and the presence of DIC.3 Purpura in septicemia may result from thrombosis in damaged vessels. A Shwartzman-like reaction, with local vascular inflammation, deposition of antigen–antibody complexes with complement activation, and increased permeability, may be the primary cause of purpura fulminans and symmetrical peripheral gangrene.3

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Urticaria and Erythema Multiforme

CHAPTER

76

Urticaria and Erythema Multiforme John Browning and Moise Levy

Urticaria and erythema multiforme are important examples of immune-mediated diseases in children and can be initiated by a variety of infectious agents.

URTICARIA Urticaria is a common problem, developing in 10% or more of individuals by adolescence and in 15% to 20% of individuals during a lifetime.1 Commonly referred to as “hives” or “welts” by the general population, urticaria consists of the sudden onset of circumscribed, erythematous, edematous papules or plaques, often showing central clearing. Lesions vary from a few millimeters to centimeters in diameter, and they are generally annular or circular but can assume bizarre, irregular shapes. Cholinergic urticaria, a particularly common form in adolescents, manifests as numerous, 2- to 3-mm, erythematous, closely set, well-defined papules. Urticarial lesions can occur anywhere on the body, can change shape from minute to minute, and can persist for minutes to hours; by definition, an individual urticarial lesion resolves within 24 hours and usually resolves without a trace within 4 hours. Rarely, urticarial vasculitis occurs and is characterized by persistence of urticaria for > 24 hours; purpura is usually faint. Approximately one-third of patients with urticarial vasculitis have decreased complement levels.2 Pruritus, a burning sensation, or both are generally present, but focal excoriation is generally absent. Occasionally, ecchymoses develop as a result of trauma from rubbing and scratching. When urticaria is associated with anaphylaxis, systemic symptoms develop, because of release of vasoactive mediators and associated swelling of viscera. Systemic symptoms and signs include hoarseness, respiratory distress, emesis, diarrhea, abdominal pain, arthralgia, flushing, syncope, hypotension, and cardiovascular collapse. Nearly all cases of urticaria in infants and most cases in children are acute, and the problem resolves in less than 6 weeks.3,4 Acute urticaria occurs in a higher proportion of atopic individuals compared with the general population.5,6 Chronic urticaria is present when lesions recur repeatedly over a period of more than 6 weeks.7 Nearly half of those with chronic urticaria have daily episodes.8 Some reports suggest that chronic urticaria is also more common in atopic individuals, but this issue is controversial.3,6,9 Approximately half of children with chronic urticaria continue to experience lesions for more than 1 year, with a median duration of 16 months.5,10 Angioedema is a circumscribed, nodular swelling of deep cutaneous and subcutaneous or submucosal tissues, most commonly on the face, genitals, and mucous membranes. A burning or stinging sensation can be present, but pruritus is not usually prominent. Angioedema occurs with urticaria in 5% to 10% of cases; some series report an association in approximately half of children with urticaria.3 Angioedema tends to be most serious in association with anaphylaxis or serum sickness and in the hereditary form with complement (C)1 esterase deficiency. If systemic symptoms have not developed within the first couple of hours of a given episode, they are unlikely to appear later. Papular urticaria is a reaction to an arthropod bite, most commonly fleas, mites, bedbugs, gnats, mosquitoes, chiggers, or animal lice. The cutaneous reaction varies according to the extent of previous exposure to the same or related species of arthropod. After repeated bites, sensitivity develops, producing a pruritic papule within

CHAPTER

76

449

approximately 24 hours; this is the most common reaction seen in young children. With prolonged, repeated exposure, a wheal develops within minutes after a bite, followed 24 hours later by papule formation; this combination of reactions is commonly seen in older children. By adolescence or adulthood, characteristically only a wheal forms, unaccompanied by the delayed papular reaction. Ultimately, as the individual becomes insensitive to the bites, there is no reaction. Individuals who experience papular urticaria are in the transitional phase between development of a delayed, papular reaction and an immediate urticarial reaction.

Etiology The cause of both acute and chronic urticaria can be difficult to determine (Boxes 76-1 and 76-2).5,9,11 Although causes of urticaria are similar for children and adults, age-dependent differences in etiology exist. In infants younger than 6 months, urticaria is largely due to cow’s milk allergy.3 Between age 6 months and 14 years, up to half of all cases of urticaria, particularly acute urticaria, are associated with infections.3,5,10,12 Streptococcus pyogenes is the most commonly associated infectious agent and has been associated with angioedema as well.13 Chronic urticaria is most often triggered by physical

BOX 76-1. Major Noninfectious Causes of Urticaria in Infants and Children ARTHROPOD BITES Ants (Solenopsis saevissima), bedbugs (Cimex lectularius), bees,a body lice (Pediculus humanus),a caterpillars, fleas (Pulex irritans),a chiggers (Trombicula irritans), flies, gypsy moths, kissing bugs (Triatoma sanguisuga), mosquitoes, scabies mites (Sarcoptes scabiei),a scorpions, spiders, wasps COMPLEMENT ACTIVATION Blood transfusion reactions, cryofibrinogenemia, cryoglobulinemia, hereditary angioedema, hypocomplementemia, serum sickness, urticarial vasculitis CONTACTANTS Animal danders, caterpillars, chemicals, cosmetics, epoxy resins, fish, foods, medications, moths, nickel, parabens, saliva, water (i.e., aquagenic), wood dust DRUGS Acetylsalicylic acid,a allopurinol, amoxicillin,a barbiturates, cephalosporin antibiotics, tetracycline, codeine, curare, meperidine, morphine,a nonsteroidal anti-inflammatory agents (e.g., indomethacin),a penicillin,a phenytoin, polymyxin B, procainamide, quinidine, radiocontrast materials, sulfa-derived antibiotics, sulfonylureas, thiamine, thiazides, vancomycin, zidovudine FOOD ADDITIVES, PRESERVATIVES, AND DYES Azo dyes (e.g., sunset yellow, tartrazinea), butylhydroxyanisole, butylhydroxytoluene, 4-hydroxybenzoic acid,a sodium benzoate,a sodium metabisulfite FOODS Chocolate, egg,a fish,a fresh berries, milk,a nuts,a peanuts,a shellfish,a tomatoes GENETIC CONDITIONS C3b inactivator deficiency, erythropoietic protoporphyria, hereditary angioedema (C1 esterase inhibitor deficiency), Muckle–Wells syndrome INHALANT ALLERGENS Animal danders, mold spores, pollens PHYSICAL FACTORS Aquagenic, cholinergic stimuli, including emotional stress (psychogenic), exercise, heat; colda; dermatographisma; heat; pressure; sunlight; sweating; vibration SYSTEMIC DISEASE Autoimmune thyroid disease, bullous pemphigoid, carcinomas, dermatomyositis, inflammatory bowel disease, juvenile rheumatoid arthritis, Kawasaki disease, leukemia, lupus erythematosus, lymphoma, polymyositis, rheumatic fever, Sjögren disease a

Most common.

450

SECTION

J

Skin and Soft-Tissue Infections

BOX 76-2. Infectious Agents Associated with Urticaria BACTERIA Borrelia burgdorferi, Escherichia coli, Neisseria gonorrhoeae, Neisseria meningitidis, Pseudomonas aeruginosa, Shigella sonnei, Streptococcus pyogenes,a Yersinia enterocolitica FUNGI Candida albicans, Cladosporium spp., Coccidioides immitis, Histoplasma capsulatum, Candida glabrata, Trichophyton spp. HELMINTHS Ancylostoma duodenale, Ascaris lumbricoides, Echinococcus spp., Enterobius vermicularis, Fasciola hepatica, Necator americanus, Onchocerca volvulus, Schistosoma spp., Strongyloides stercoralis, Toxocara canis, Trichinella spiralis, Trichobilharzia spp. (avian blood flukes), Wuchereria bancrofti MYCOPLASMA Mycoplasma pneumoniae PROTOZOA Entamoeba histolytica, Giardia lamblia, Plasmodium spp., Trichomonas vaginalis RICKETTSIA Coxiella burnetii TREPONEME Treponema pallidum VIRUSES Adenovirusa; coxsackieviruses A9, A16, B4, B5a; echovirus 11a; Epstein–Barr virusa; hepatitis viruses A, B, Ca; influenza B virusa; human immunodeÀciency virus; measles virus, attenuated; mumps virus; respiratory syncytial virusa a

Most common.

factors,5,14,15 but up to 45% of cases of chronic urticaria are also associated with infections.5,10 Viral upper respiratory tract and gastrointestinal tract infections are the primary infectious triggers of acute urticaria in children.5,12,16,17 Urticaria can also occur in association with infection with mycoplasmal, treponemal, rickettsial, fungal, or parasitic organisms, in Kawasaki disease, or in association with the infestation or bites of arthropods.14,18,19 Occasionally, urticaria is due to occult infection, such as dental caries, sinusitis, or urinary tract infection. Treatment of associated infections results in an improvement in urticaria in only a minority of patients. Furthermore, it is commonly unclear whether the infectious agent, the therapeutic drug, or the combination was responsible for triggering the episode of urticaria. Aspirin, in particular, is known to exacerbate urticaria due to a variety of other causes, including infections.6,20 The medication that most commonly causes urticaria is penicillin and its derivatives; up to 25% of patients with chronic urticaria have antibodies to penicillin.21

Pathogenesis Urticaria is due to an immediate hypersensitivity response, causing release of mediators from cutaneous mast cells and basophils as well as transudation of fluid from cutaneous blood vessels. Histamine is a primary mediator of urticaria and acts via both H1 and H2 receptors to produce vasodilation and altered vascular permeability. Other mediators of urticaria are kinins, eicosanoids, and neuropeptides.7 Mediator release can be triggered by a number of mechanisms and agents. Physical agents or drugs can produce urticaria by nonimmunologic means. The most common immunologic mechanism for release of mediators is interaction of an allergenic antigen (e.g., food or drug) with immunoglobulin E (IgE) bound to a basophil or mast cell, producing a type I hypersensitivity response.22 IgG autoantibodies can be responsible for urticaria in some patients by interacting with and causing cross-linkage of adjacent IgE receptors, suggesting that urticaria may be a manifestation of autoimmune

mast cell disease.7 Complement activation with formation of the anaphylotoxins C3a, C4a, and C5a is an alternative immunologic mechanism for generation of urticaria. The anaphylotoxins interact directly with the cell surface of basophils or mast cells to trigger mediator release. Immune complex reactions can stimulate mediator release via the complement system; this process is particularly important in urticaria associated with hereditary angioedema, serum sickness, blood transfusion reactions, cryoglobulinemia, collagen vascular disease, urticarial vasculitis, and the urticaria-like lesions of Henoch–Schönlein purpura. The mechanism of urticaria due to infectious agents is postulated to involve either the formation of immune complexes with activation of complement and release of anaphylotoxins or the development of IgE antibodies to microbial antigens.5,12

Diagnosis Diagnosis is established from the history and clinical characteristics. Nonurticarial conditions that can have an urticarial phase include erythema toxicum neonatorum, erythema multiforme, erythema nodosum, anaphylactoid purpura, Kawasaki disease, diffuse cutaneous mastocytosis, and bullous pemphigoid. Biopsy can be helpful in equivocal cases or when urticarial vasculitis is suspected. Biopsy of an urticarial lesion is expected to show dermal edema, dilatation of blood and lymphatic vessels, and a sparse perivascular mononuclear inÀltrate with variable numbers of eosinophils. The histopathology of urticaria can often look like “normal skin.” Urticarial vasculitis, however, is characterized by swelling of endothelial cells and Àbrinoid necrosis of postcapillary venules, red blood cell extravasation, leukocytoclasis (i.e., fragmentation of neutrophil nuclei), and perivascular and vascular inflammation with neutrophils. Diagnostic tests in a child with acute urticaria are performed when Àndings other than urticaria suggest a speciÀc etiology (e.g., streptococcal pharyngitis, bacterial enteritis). In an attempt to identify the cause of chronic urticaria (> 6 weeks in duration), diagnostic tests should be selected on the basis of history and physical examination. Generally, laboratory evaluation, even if it is thorough, is unrevealing unless a speciÀc allergen (food, drug, contactant, inhalant), infection, infestation, or underlying systemic disease is suspected.16,23 If a physical cause for urticaria is identiÀed, additional testing is unnecessary. If angioedema is a prominent, consistent feature, the patient should be tested for hereditary deÀciency of C1 esterase inhibitor by measuring plasma C4 complement concentration. Useful tests for identifying a cause of urticaria are: (1) a complete blood cell count with differential count (particularly the eosinophil count); (2) throat swab collection for Streptococcus pyogenes culture or serum for antistreptolysin O titer; (3) Epstein–Barr virus serologic test; (4) serum hepatic enzyme measurements; (5) serologic tests for hepatitis viruses; (6) urinalysis and urine culture; (7) vaginal smear for Candida and Trichomonas; (8) stool examination for ova and parasites; and (9) radiographs of the sinuses and teeth.13,17 Of course these tests should only be ordered when clinically indicated.

Management Infectious agents of chronic urticaria should be identiÀed and treated, when possible, and stimuli of physical urticaria, such as heat, cold, exercise, and emotional stress, should be minimized. The most successful symptomatic approach to treatment of urticaria with medication is blockage of the effect of histamine at its receptor on cutaneous blood vessels. Regimens include therapies with H1 antihistamines, combination H1 and H2 receptor blockade, and tricyclic antidepressants. Antihistamines that block H1 receptors are the mainstay of treatment for urticaria. Hydroxyzine hydrochloride is the most effective of the classic H1 antihistamines for suppression of the wheal-and-flare response, pruritus, dermatographism, and cholin-

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Urticaria and Erythema Multiforme

CHAPTER

76

451

ergic urticaria,24–26 whereas cyproheptadine is the drug of choice for cold-induced urticaria.27 For patients who experience excessive sedation or mucosal dryness, the newer, nonsedating H1 antihistamines, such as fexofenadine, loratadine, and cetirizine, are effective alternatives.28 These agents have been shown to be comparable in efficacy with hydroxyzine and one another for the treatment of urticaria. The therapeutic response to the H1 antihistamines can be improved with the addition of an H2 receptor-blocking agent, such as cimetidine or ranitidine.29 H2-blocking agents should not be used alone, because their extent of H1 blockade is insufficient. The dosage of antihistamine is often more important than the particular agent. Antihistamine therapy should be initiated at the upper end of the recommended dosage range and then gradually increased as needed until symptoms are relieved or side effects, particularly sedation, become prohibitive. In general, the dosage should not exceed approximately twice that recommended by the manufacturer. The effectiveness of antihistamines in children, particularly in those with dermatographism, appears to be produced, at least in part, by their sedative action. Bothersome drowsiness can be minimized by administering most of the daily dose just before bedtime. Topical corticosteroid agents are ineffective in urticaria, and systemic corticosteroids are rarely indicated in children, except in severe unremitting cases, especially those associated with asthma, laryngeal edema, or circulatory instability, or when angioedema is particularly severe on the face.

ERYTHEMA MULTIFORME

A

In the past, erythema multiforme has been divided into two subtypes: erythema multiforme minor and erythema multiforme major, which was compatible with Stevens–Johnson syndrome (SJS). Due to confusion over terminology, erythema multiforme minor is now called erythema multiforme and erythema multiforme major is referred to as SJS. As the name implies, erythema multiforme has numerous manifestations in the skin. Morphology of skin lesions varies from erythematous macules, papules, vesicles, bullae, or urticarial plaques to patches of confluent erythema. The eruption appears most commonly in patients between the ages of 10 and 30 years and is usually asymptomatic, although a burning sensation or pruritus can be present. Diagnosis is established by the finding of doughnut-shaped target (iris or bull’s-eye) lesions with an erythematous outer border, an inner pale ring, and a purple center.

Definition Erythema multiforme is characterized by a symmetric cutaneous eruption, most commonly on the dorsal hands and feet and extensor surfaces of the arms and legs (particularly the elbows and knees), often involving the palms, soles, and face, with relative sparing of the trunk and mucous membranes. Prodromal symptoms are generally absent, and the eruption often appears initially as red macules or urticarial plaques that expand centrifugally to form lesions up to 2 cm in diameter. SJS is characterized by lesions that develop predominantly on the extremities (Figure 76-1) and are accompanied by involvement of two or more mucosal surfaces – the eyes; the oral cavity, upper airway, or esophagus; the gastrointestinal tract; and anogenital mucosa (Figure 76-2). A burning sensation, edema, and erythema of the lips and buccal mucosa are often the initial manifestations, followed by development of bullae, ulceration, and hemorrhagic crusting. Pain from mucosal ulceration is often severe, but skin tenderness may be mild. In at least half of the cases of SJS, a history of a prodromal illness of fever and flu-like symptoms is present.30 Corneal ulceration, anterior uveitis, panophthalmitis, pancreatitis, bronchitis, bronchiolitis obliterans, pneumonitis, myocarditis, hepatitis, enterocolitis, polyarthritis, hematuria, and acute tubular necrosis leading to renal failure can also

B Figure 76-1. (A, B) Hands of two children with drug-related Stevens–Johnson syndrome showing typical erythema multiforme lesions. (Courtesy of S.S. Long.)

occur. Disseminated cutaneous bullae and erosions can result in significant blood loss, increased insensible fluid loss, and a high risk of bacterial superinfection and septicemia. New lesions occur in crops, and complete healing can take 4 to 6 weeks; ocular scarring with visual impairment as well as stricture of the esophagus, bronchi, vagina, urethra, or anus can result. Nonspecific laboratory abnormalities in SJS can include leukocytosis, increased levels of serum amylase, lipase and hepatic transaminases, elevated erythrocyte sedimentation rate, and decreased serum albumin concentration. The incidence is increased up to three times in human immunodeficiency virus-infected people.31 It should be noted that in SJS the lesions are nonscarring because they represent sites of epidermal necrosis and detachment rather than full-thickness dermal necrosis. Severe ocular complications can occur that threaten sight, including corneal ulceration and synechiae. Toxic epidermal necrolysis (TEN) is part of the same spectrum as SJS but represents a more severe form of the disease, involving

452

SECTION

J

Skin and Soft-Tissue Infections

BOX 76-3. Infectious Agents Associated with Erythema Multiforme

Figure 76-2. Severe erosion of the labial mucosa as well as vesicles and crusted plaques on the face of a girl with Stevens–Johnson syndrome. Culture of vesicular lesions for herpes simplex virus was negative.

considerable constitutional toxicity.32,33 SJS is said to involve less than 10% of the body surface area (BSA) and TEN more than 30% of the BSA. Between 10% and 30% BSA involvement is considered to be SJS/TEN overlap. TEN appears most commonly in adults but can occur in children.34 Although erythema multiforme can manifest initially as urticarial lesions, it can be distinguished from urticaria by failure of a given lesion to fade within 24 hours. Unlike urticaria, lesions have dusky centers and are often distributed symmetrically. Serum sickness-like reaction (SSLR) to cefaclor has also been described as causing erythema multiforme-like lesions.35,36 Although the lesions can develop a dusky to purple center, the eruption of cefaclor-induced SSLR is in most cases pruritic, transient, and migratory, probably representing urticaria rather than the fixed tissue lesion of erythema multiforme. Anticonvulsant hypersensitivity syndrome is a multisystem reaction that appears approximately 4 weeks to 3 months after initiation of phenytoin, carbamazepine, phenobarbitone, or primidone therapy. The mucocutaneous eruption can be identical to that seen in erythema multiforme, SJS, and TEN, but the reaction typically also includes lymphadenopathy as well as fever, hepatitis, eosinophilia, and leukocytosis.37

Etiology Numerous factors and inciting infections have been associated with the development of erythema multiforme; infection with herpes simplex virus (HSV) is the most common (Box 76-3).38 HSV labialis or genitalis has been implicated in at least 60% of episodes of erythema multiforme and is believed to trigger nearly all episodes of recurrent disease; however, HSV is an uncommon cause of SJS and TEN.38–41 HSV antigens and DNA are present in skin lesions from patients with erythema multiforme but are absent in unaffected skin of such patients.42–45 Presence of the human leukocyte antigens B62, B35, and DR53 is associated with a higher risk of HSV-induced disease, particularly the recurrent form.46,47 Mycoplasma pneumoniae is the most convincingly demonstrated infectious cause of SJS and TEN48; the organism has also been detected in lesional skin.49 Approximately 50% of cases of SJS and TEN occur with no identifiable cause. Drugs, particularly sulfonamide-containing products, nonsteroidal anti-inflammatory agents (butazones, pyrazolones, ibuprofen, piroxicam, and salicylates), and anticonvulsants (phenytoin, lamotrigine), are the most common precipitating agents for SJS and TEN.50 Recurrent erythema multiforme has been reported most commonly after infection with HSV.39,40,50 Lesions of HSV-induced recurrent

BACTERIA Bartonella henselae, Corynebacterium diphtheriae, Francisella tularensis,a Legionella pneumophila, Mycobacterium leprae, Mycobacterium tuberculosis,a Neisseria gonorrhoeae, Proteus mirabilis,a Pseudomonas aeruginosa, Salmonella spp.,a Staphylococcus aureus,a Streptococcus pneumoniae, Streptococcus pyogenes,a Vibrio parahaemolyticus,a Yersinia spp.a CHLAMYDIA Chlamydophila psittaci, Chlamydia trachomatis (lymphogranuloma venereum) FUNGI Coccidioides immitis, Histoplasma capsulatuma IMMUNIZATIONS Calmette-Guérin bacillus, diphtheria and tetanus toxoid, hepatitis B, measles, mumps, rubella, poliomyelitis MYCOPLASMA Mycoplasma pneumoniaea PARASITES Trichomonas vaginalis TREPONEME Treponema pallidum VIRUSES Adenovirusa 7; coxsackieviruses A10, A16, B5; echovirus 6; Epstein–Barr virusa; hepatitis viruses A, B; herpes simplex viruses 1, 2a; human immunodeficiency virus; influenza A virus; measles virus; mumps virus; orf virus; paravaccinia virusa; parvovirus B19; varicella-zoster virus a

Most clearly established.

disease typically develop 10 to 14 days after onset of recurrent HSV eruptions (or asymptomatic reactivation), have a similar appearance (frequently at the same site, e.g., the palms) from episode to episode, but can vary considerably in frequency and duration in a given patient.51 Not all episodes of recurrent HSV elicit erythema multiforme in susceptible patients.40 SJS and TEN can follow re-exposure to drugs.52

Pathogenesis and Diagnosis The pathogenesis is incompletely understood, but evidence increasingly implicates a host-specific cell-mediated immune response to an antigenic stimulus, resulting in damage to keratinocytes.53 Cytokines released by activated mononuclear cells and keratinocytes may contribute to epidermal cell death and constitutional symptoms. Although circulating and tissue-bound immune complexes can be demonstrated, they do not appear to play a central role in the pathogenesis; histopathologic examination does not detect vasculitis. Microscopic findings of erythema multiforme and SJS/TEN (as well as the gross appearance of the cutaneous eruption) are variable but are diagnostically significant. Early lesions typically show slight intercellular edema, rare dyskeratotic keratinocytes, basal vacuolization in the epidermis, and perivascular lymphohistiocytic infiltrate and edema in the upper dermis. More mature lesions demonstrate accentuation of these characteristics, with development of lymphocytic exocytosis and an intense perivascular and interstitial mononuclear infiltrate (without significant numbers of eosinophils or neutrophils) in the upper third of the dermis. The entire epidermis becomes necrotic in severe cases.

Management In most cases, supportive treatment is given while a search is made for the inciting agent. Debridement of necrotic skin is not recommended before disease activity has subsided. For treatment of erythema multiforme, topical emollients as well as systemic antihistamines, salicylates, and nonsteroidal anti-inflammatory agents do not alter

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Papules, Nodules, and Ulcers

the course of the disease but can provide symptomatic relief. No controlled, prospective studies support the use of corticosteroids.54,55 Corticosteroid therapy may decrease the healing time but not the frequency of recurrence of oral lesions,52 and systemic corticosteroids have sometimes been used to treat early, severe cases of SJS.56 Most authorities discourage their use, however, because of reports of the association of higher morbidity and mortality.33 Intravenous immunoglobulin has been used with anecdotal success to treat SJS/TEN; no prospective controlled trial has been performed.57,58 Transplantation of human limbal epithelium cultivated on amniotic membrane has been used for treatment of severe oculoulcerative disease.59 Oral acyclovir has been shown to be effective in controlling recurrent episodes of HSV-associated erythema multiforme when given prophylactically or early in the course of the eruption.41,60

Disease Entity

John Browning and Moise Levy

Papules, nodules, and ulcers are primary lesions of the skin that can be caused by a variety of infectious and noninfectious agents (Tables 77-1 and 77-2). A papule is a raised superficial lesion that is < 1 cm in size; a nodule is a solid palpable lesion that is > 1 cm in size; an ulcer is a loss of skin to the level of the dermis or deeper. A single type of cutaneous lesion can be associated with a specific organism (e.g., papule of molluscum contagiosum), or multiple morphologic lesions can occur within the natural history of infection due to a single organism. For example, in tuberculosis or leishmaniasis, an initial papule can enlarge to form a nodule and then break down to an ulcer. The composition of a papular or nodular lesion can consist of a proportionately large volume of the infectious agent (e.g., molluscum contagiosum), almost exclusively of inflammatory cells (e.g., wellcontrolled primary cutaneous tuberculosis), or, most often, both. Papules, nodules, or ulcers can also result from a disordered immune response after infection; at the time of development of these lesions, infectious organisms are not generally recoverable (Table 77-3 and Box 77-1). Drug reactions in children can also manifest as papules or nodules, often in the setting of a concurrent viral infection. Nodular lymphangitis is a distinctive, underrecognized syndrome resulting from cutaneous inoculation of a limited number of agents (e.g., Sporothrix schenckii, Nocardia brasiliensis, Mycobacterium marinum, and Francisella).1 This chapter contains a brief discussion of disease entities that serves to illustrate the principles of primary skin infection and immune reactivity to the presence of an infectious agent that manifests as papules, nodules, or ulcers. Infections in which skin lesions do not develop primarily but are manifestations of systemic disease (e.g., bartonellosis, tularemia, and trypanosomiasis) are discussed elsewhere.

MOLLUSCUM CONTAGIOSUM Molluscum contagiosum produces a papule in the skin caused largely by the presence of virions.2,3 This disease is unlike warts, in which the relative abundance of human papillomavirus particles in the verruca

Skin Lesionsa

Infectious Agent

P, N P P

Dermatobia hominis Sarcoptes scabiei Tunga penetrans

P, U P, U P, U U P, U P P P P, N P P N

Bacillus anthracis Haemophilus ducreyi Corynebacterium diphtheriae Streptococcus pyogenes Pseudomonas aeruginosa Enterobacter spp. Escherichia coli Klebsiella spp. Proteus spp. Pseudomonas aeruginosa Staphylococcus aureus Staphylococcus aureus

P, N, U

Calymmatobacterium granulomatis Mixed skin flora Staphylococcus aureus Streptococcus pyogenes Chlamydia trachomatis

ARTHROPODS

Cutaneous myiasis Scabies Tungiasis BACTERIA

Anthrax Chancroid Diphtheria Ecthyma Ecthyma gangrenosum Folliculitis

Furunculosis, carbunculosis Granuloma inguinale

Papules, Nodules, and Ulcers

453

Ulcers

The authors acknowledge significant use of the work of G.L. Darmstadt from the second edition.

77

77

TABLE 77-1. Primary Infectious Causes of Papules, Nodules, and

ACKNOWLEDGMENTS

CHAPTER

CHAPTER

Hidradenitis suppurativa N Impetigo P, U P, U Lymphogranuloma P, U venereum Melioidosis N, U Nocardiosis P, N P, N Rhinoscleroma N, U Sycosis barbae P

Pseudomonas pseudomallei Nocardia brasiliensis Nocardia asteroides Klebsiella rhinoscleromatis Staphylococcus aureus

FUNGI

Blastomycosis Candidosis Folliculitis Mycetoma Sporotrichosis Tinea barbae, tinea capitis Tinea corporis

P, N, U P P P P P, N, U P, N, U P,N

Blastomyces dermatitidis Candida albicans Candida tropicalis Candida albicans Malassezia furfur Multiple organisms Sporothrix schenckii Trichophyton spp.

P,N P P

Microsporum spp. Trichophyton spp. Microsporum canis

HELMINTHS

Dracunculosis U (guinea worm) Larva currens P Cutaneous larva migrans P P Ground itch P Onchocerciasis Cercarial dermatitis Schistosomiasis

Dracunculus medinensis

P, N P P

Strongyloides stercoralis Ancylostoma braziliense Ancylostoma caninum Necatur americanus Ancylostoma duodenale Onchocerca volvulus Trichobilharzia spp. Schistosoma spp.

P, N, U

Mycobacterium marinum

N, U N ,U N, U P, N, U P, N, U

Mycobacterium kansasii Mycobacterium scrofulaceum Mycobacterium ulcerans Mycobacterium avium complex Mycobacterium fortuitum, Mycobacterium chelonae Mycobacterium leprae Mycobacterium tuberculosis Mycobacterium bovis Calmette-Guérin bacillus

MYCOBACTERIA

Nontuberculous mycobacteriosis

Cutaneous tuberculosis

P, N, U P, N, U P, N, U P, N, U

454

SECTION

J

Skin and Soft-Tissue Infections

TABLE 77-1. Primary Infectious Causes of Papules, Nodules, and

TABLE 77-2. Major Noninfectious Causes of Papules, Nodules, and

Ulcers—cont’d

Ulcers

Disease Entity

Skin Lesions

a

Infectious Agent

PROTOZOA

Leishmaniasis

P, N, U

Leishmania spp.

P P, N, U P, N, U

Treponema carateum Treponema pallidum Treponema pertenue

P

Human papillomaviruses

P, N P, N

Paravaccinia virus Molluscum contagiosum virus

P, N P, N

Orf virus Human papillomaviruses

TREPONEMES

Pinta Syphilis Yaws VIRUSES

Epidermodysplasia verruciformis Milker’s nodule Molluscum contagiosum Orf Warts a

N, nodule; P, papule; U, ulcer.

varies with the type of human papillomavirus and the clinical setting, but the papular lesions develop largely from hyperproliferation of basal cells and retention of upper epidermal keratinocytes. The poxvirus that causes molluscum contagiosum is a doublestranded DNA virus that replicates in the cytoplasm of host epithelial cells (see Chapter 202, Poxviridae). The disease is acquired through direct contact with an infected person or from fomites and is spread by autoinoculation. School-aged children who are otherwise well and individuals with impaired cellular immune function are most commonly affected.4 The incubation period is estimated to be 2 to 8 weeks but may be longer. Boys are affected approximately three times as frequently as girls. One mechanism the molluscum contagiosum virus uses to evade host mechanisms for killing virally infected cells is its viral flice-like inhibitory protein (FLIP). Viral FLIP inhibits the death effector domain of caspase 8 (the initiator for the cell surface death receptors in the tumor necrosis factor family).5 The lesions of molluscum contagiosum are discrete, pearly, skincolored, dome-shaped, smooth papules varying in size from 1 to 5 mm (Figure 77-1). Typically, they have a central umbilication from which a plug of cheesy material can be expressed. Papules can occur anywhere on the body, but there is predilection for the face, eyelids, neck, axillae, and thighs. When the lesions are found in clusters on the genitalia or in the groin of a sexually active adolescent, other sexually transmitted infections should be sought. Lesions commonly involve the genital area in children and frequently are not acquired by sexual transmission; a search should be undertaken, however, for other signs of sexual abuse. Mild surrounding erythema or an eczematous dermatitis can accompany the papules. Lesions in children with the acquired immunodeficiency syndrome (AIDS) tend to be large and numerous, particularly on the face, and can have an atypical appearance, with coalescent crusted lesions and large separate lesions on an erythematous base; exuberant lesions also occur in children with leukemia and other immunodeficiencies.6 Histopathologically, the epidermis is hyperplastic and hypertrophied, extending into the underlying dermis and projecting above the skin surface. The papule of molluscum contagiosum consists of a lobulated adhesive mass of virus-infected epidermal cells. Eosinophilic viral inclusion bodies (Henderson–Patterson or molluscum bodies) become more prominent as infected keratinocytes move upward from the basal layer to the stratum corneum. The central plug of material, which is composed of virus-laden cells, can be shelled out from a lesion and examined under the microscope with 10% potassium hydroxide or with Wright or Giemsa stain. The rounded, cup-shaped mass of homogeneous cells, often with identifiable lobules, is diagnostic.

Disease Entity

Skin Lesionsa

Amyloidosis Arthropod bite hypersensitivity reaction Dermatofibroma Elastosis perforans serpiginosa Epidermal cyst Eruptive vellus hair cyst Erythema induratum Erythema nodosum leprosum Factitial panniculitis Foreign-body reaction Fox–Fordyce disease Granuloma annulare Juvenile xanthogranuloma Kawasaki disease Keratosis follicularis (Darier disease) Keratosis pilaris Langerhans cell histiocytosis Leukemia Lichen nitidus Lichen planus Lipoma Lupus erythematosus Lupus panniculitis Lymphoma Lymphomatoid granulomatosis Miliaria rubra Milium Necrobiosis lipodica diabeticorum Neuroblastoma Pilar cyst (trichilemmal cyst) Pilomatricoma Pityriasis rubra pilaris Polyarteritis nodosa Polymorphous light eruption Pseudoxanthoma elasticum Pyoderma gangrenosum Pyogenic granuloma Rheumatoid nodule Sarcoidosis Spitz nevus Steatocystoma multiplex Subcutaneous fat necrosis Superficial thrombophlebitis Trichoepithelioma Tuberous sclerosis Urticaria Urticaria pigmentosa

P, N P, N P, N P P, N P N, U N N P, N P P, N P, N P P P P, N, U P, N P P, U N P, U N P, N N, U P P N, U N P, N P, N P N, U P P N, U P N P, U P, N P, N N N P, N P P P, N

a

N, nodule; P, papule; U, ulcer.

The differential diagnosis of molluscum contagiosum includes trichoepithelioma, basal cell carcinoma, ectopic sebaceous glands, syringoma, hidrocystoma, keratoacanthoma, and warty dyskeratoma. In individuals with AIDS, disseminated cryptococcosis can be indistinguishable clinically from molluscum contagiosum. Molluscum contagiosum is a self-limited, epidermal disease and should not be overtreated to the point that scarring results.7 The average duration lasts 6 to 9 months, although lesions can persist for years, may spread to distant sites, and can be transmitted to others. Affected persons should be advised to avoid sharing baths and towels until the infection has cleared. Infection can spread rapidly and produce hundreds of lesions in children with atopic dermatitis or immunodeficiency.5 A brief, 6- to 9-second topical application of liquid nitrogen is effective and, in many instances, is the treatment of choice. Particularly in younger children, in whom liquid nitrogen therapy is not well tolerated, cantharidin 0.9% can be applied to each lesion without occlusion; this agent frequently causes enough

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Papules, Nodules, and Ulcers

CHAPTER

77

455

TABLE 77-3. Conditions with Sterile Papular, Nodular, or Ulcerative Skin Lesions After Infection Disease Entity

Skin Lesionsa Infectious Agent

Autosensitization P, N Erythema elevatum diutinum P, N Erythema multiforme P Erythema nodosum Gianotti–Crosti syndrome Guttate psoriasis Henoch–Schönlein purpura Polyarteritis nodosa Postscabitic nodule Reiter disease Circinate balanitis Keratoderma blenorrhagicum Rheumatic fever

N P P P, N P, N, U P, N

See Box 77-3 Streptococcus pyogenes See Chapter 76 (Urticaria and Erythema Multiforme) See Box 77-1 See Box 77-2 Streptococcus pyogenes Streptococcus pyogenes Streptococcus pyogenes Sarcoptes scabiei

U P, N N

–b –b Streptococcus pyogenes

a

N, nodule; P, papule; U, ulcer. Sexually transmitted form follows infection with Chlamydia trachomatis. Postdysenteric form follows infection due to Salmonella and Shigella species, Yersinia enterocolitica, Yersinia pseudotuberculosis, Campylobacter fetus, and Clostridium difficile. b

Figure 77-1. Umbilicated papules on the upper arm and chest of a child with molluscum contagiosum.

BOX 77-1. Infectious Agents Associated with Erythema Nodosum BACTERIA Bartonella henselae,a Brucella spp., Campylobacter jejuni, Corynebacterium diphtheriae, Francisella tularensis, Haemophilus ducreyi, Leptospira interrogans, Neisseria meningitidis, Salmonella spp., Streptococcus pyogenes,a Yersinia enterocolitica,a Yersinia pseudotuberculosis CHLAMYDIA Chlamydophila psittaci, Chlamydia trachomatis (lymphogranuloma venereum) FUNGI Blastomyces dermatitidis, Coccidioides immitis,a dermatophytoses (tinea capitis), Histoplasma capsulatum, Sporothrix schenckii HELMINTHS Ancylostoma duodenale, Necator americanus MYCOBACTERIA Mycobacterium leprae, Mycobacterium marinum, Mycobacterium tuberculosisa MYCOPLASMA Mycoplasma pneumoniae PROTOZOA Toxoplasma gondii TREPONEME Treponema pallidum VIRUSES Epstein–Barr virus; hepatitis viruses B, C; herpes simplex virus; mumps virus; paravaccinia virus a

Most common.

inflammation to facilitate disappearance of the lesion. Cantharidin should be used cautiously because severe inflammatory reactions can occur. The papules can also be destroyed by expressing the plug with a needle, a sharp curette, or a comedo extractor. Once-daily topical application of retinoic acid can also incite an inflammatory response, leading to resolution of the lesions. The role of topical application of imiquimod is currently under investigation.

MYCOBACTERIAL INFECTIONS OF THE SKIN The type of cutaneous lesion that develops in tuberculous and nontuberculous mycobacterial infection depends on virulence properties

of the mycobacterial organism, the general health and immune responsiveness of the host, and the mode of introduction of the mycobacteria into the skin. Cutaneous tuberculosis is caused by Mycobacterium tuberculosis, M. bovis, and, occasionally, the Calmette-Guérin bacillus (an attenuated form of M. bovis). Manifestations caused by the various Mycobacterium spp. are indistinguishable from one another. After invasion of the skin, mycobacteria either multiply intracellularly within macrophages, leading to progressive disease, or are controlled by the host immune reaction, which depends largely on macrophages activated by lymphokines released from lymphocytes. Delayed-type hypersensitivity is also important, producing caseous necrosis through cytotoxic T-lymphocyte-mediated destruction of bacilli-laden nonactivated macrophages and nearby tissue.8 A primary lesion, called a tuberculous chancre (when resulting from M. tuberculosis or M. bovis), occurs at the skin or mucous membrane site where organisms gain access through trauma. Sites of predilection are the face, lower extremities, and genitalia. The initial lesion develops 2 to 4 weeks after inoculation. A red-brown papule gradually enlarges to form a nodule and then a shallow, firm, sharply demarcated ulcer; satellite abscesses can be present. Some lesions acquire a crust resembling that in impetigo, and others become heaped-up and verrucous at the margins. Mycobacteria can be isolated from skin lesions and local lymph nodes, but acid-fast staining of histologic sections often fails to reveal the organism, particularly in infection that is well controlled by the host. Use of a polymerase chain reaction assay to identify organisms in biopsy material is helpful. Clinically, the differential diagnosis is broad, including multiple nontuberculous mycobacteria9–11 and deep fungal infections, cat-scratch disease, sporotrichosis, nocardiosis, leishmaniasis, syphilis, leprosy, tularemia, papular acne rosacea, reaction to foreign substances (e.g., zirconium, beryllium, silk or nylon sutures, talc, starch), and lupus miliaris disseminatus faciei. Treatment hastens healing and prevents dissemination (see Chapter 134, Mycobacterium tuberculosis, and Chapter 135, Mycobacterium Species Non-tuberculosis). Untreated, spontaneous healing with scarring generally occurs within 12 months. At that time, skin lesions and infected nodes can become calcified. Alternatively, the infection can reactivate, and M. tuberculosis can cause scrofuloderma or lupus vulgaris, or rarely, can progress to the acute miliary form.

456

SECTION

J

Skin and Soft-Tissue Infections

ERYTHEMA NODOSUM Erythema nodosum consists of the sudden appearance of exquisitely tender, erythematous, 1- to 10-cm nodules usually located symmetrically on the extensor surfaces of the legs12 (see Figure 160-1I and J). Lesions can also develop elsewhere, including the calves, thighs, trunk, upper extremities, head, and neck. Erythema nodosum is rare in children younger than 2 years and is most common in the pediatric population during adolescence. Nodules typically enlarge over a few days, remain stable for 1 to 2 weeks, and then gradually resolve over 3 to 6 weeks with color changes typical of a bruise. Ulceration, suppuration, and scarring do not occur, but residual hyperpigmentation can persist for weeks to months. Fever, chills, malaise, leukocytosis, and elevated erythrocyte sedimentation rate are often present. Erythema nodosum in children in the United States is most commonly associated with group A streptococcal pharyngitis (see Box 77-1). Infection precedes onset of erythema nodosum by approximately 3 weeks. A variety of other infectious agents have also been associated with the development of erythema nodosum, most notably tuberculosis (worldwide), coccidioidomycosis (in the southwestern United States), Yersinia enterocolitica (in Europe), cat-scratch disease (Bartonella henselae), and tularemia. Erythema nodosum can also be associated with: (1) use of drugs, such as sulfonamides and oral contraceptive agents; (2) sarcoidosis and, occasionally, malignancies such as leukemia and lymphoma; and (3) systemic diseases, including Behçet disease, Reiter disease, systemic lupus erythematosus, and inflammatory bowel disease. Differential diagnosis includes insect bite hypersensitivity reaction, lupus panniculitis, factitial panniculitis, superficial thrombophlebitis, polyarteritis nodosa, necrobiosis lipoidica diabeticorum, erythema induratum, erythema nodosum leprosum, and sarcoidosis. Sweet syndrome (painful, indurated cutaneous plaques accompanied by fever and leukocytosis) should also be considered; it can be associated with hematologic malignancies or chronic recurrent multifocal osteomyelitis.13,14 The pathogenesis of erythema nodosum is largely unknown. It is classified as a septal panniculitis without vasculitis and is regarded as an immunologic reaction to a variety of antigenic stimuli, most commonly group A streptococcal antigens in children. Diagnosis is usually made clinically. Agreement as to the typical histologic appearance is not complete, but in the early phase there is usually an acute inflammatory reaction in the connective tissue of the lower dermis and subcutaneous tissue, with swelling and degeneration of collagen bundles. Venous walls are edematous, endothelial cells can proliferate, and mixed infiltrates can be present; however, the changes of leukocytoclastic and lymphocytic angiitis are not present. Foreign body-type giant cells are typical. Erythema nodosum is self-limited and usually resolves within 3 to 6 weeks. A recurrent form exists; in children, it is often due to repeated streptococcal infections. Attention should be directed to identifying and treating potential causes of the reaction. Nonsteroidal anti-inflammatory agents may provide symptomatic relief. Leg elevation and bedrest may minimize lower-extremity edema and, thus, decrease discomfort. Potassium iodide treatment may hasten resolution of the nodules.15

BOX 77-2. Infectious Agents Associated with Gianotti–Crosti Syndrome IMMUNIZATIONS Calmette-Guérin bacillus, diphtheria, tetanus, pertussis VIRUSES Adenovirusesa; coxsackievirus A16, B4, B5a; cytomegalovirus; echovirus 7, 9a; Epstein–Barr virusa; hepatitis virus A, Ba; human immunodeficiency virus; parainfluenza virusesa; poliovirus; respiratory syncytial virusa a

Most common.

BOX 77-3. Infectious Agents Associated with a Papular or Nodular Autosensitization (“Id”) Reaction FUNGI Blastomyces dermatitidis, Candida albicans, Coccidioides immitis, Epidermophyton floccosum, Histoplasma capsulatum, Sporothrix schenckii, Trichophyton mentagrophytes,a Trichophyton rubrum,a Trichophyton tonsuransa MYCOBACTERIA Mycobacterium bovis, Mycobacterium leprae, Mycobacterium tuberculosis PROTOZOA Leishmania spp. TREPONEMES Treponema carateum a

Most common.

Gianotti–Crosti syndrome is now recommended, irrespective of hepatitis B surface antigenemia. In the United States, Epstein–Barr virus infection is the most common cause of Gianotti–Crosti syndrome. The disorder consists of the sudden eruption of symmetric, flattopped, discrete, skin-colored to erythematous, 2- to 10-mm papules on the malar face, upper and lower extremities, and buttocks. The trunk and antecubital and popliteal fossae are usually spared. Pruritus and the Koebner phenomenon (accentuation of lesions at sites of trauma) can be present. Peak incidence is between the ages of 1 and 4 years. Differential diagnosis includes lichen planus, lichen nitidus, lichenoid dermatitis or drug eruption, and Langerhans cell histiocytosis. The pathogenesis of the syndrome has been suggested to involve an immune-reactive process, perhaps mediated by immune complexes or due to a delayed hypersensitivity response.17 In the form associated with hepatitis B infection, however, lesions have not yielded evidence of immune complex vasculitis, and the surface antigen has not been demonstrated in the skin lesions. The syndrome is generally self-limited, resolving within 3 to 6 weeks. Lesions can occasionally persist for a few months, however, and relapses, although rare, have been reported. No specific treatment is helpful, although antihistamines can provide symptomatic relief.

GIANOTTI–CROSTI SYNDROME

AUTOSENSITIZATION

Gianotti–Crosti syndrome, also known as papular acrodermatitis of childhood, was initially described in association with generalized lymphadenopathy and anicteric hepatitis B surface antigenemia. The disorder has since been reported as occurring after a variety of viral infections, most commonly infections of the upper respiratory tract (Box 77-2). In the United States, hepatitis B infection is not commonly associated with the eruption, and the association has become rare after universal hepatitis B immunization during infancy. In the past the form of the disorder not associated with hepatitis B was referred to as papulovesicular acro-located syndrome, although it was indistinguishable clinically from papular acrodermatitis of childhood.16 The term

Dermatophytid (“id”) reactions are inflammatory autosensitivity reactions of the skin thought to occur in association with dermatophytosis but at sites distant from infection (Box 77-3).18,19 Culture and potassium hydroxide examination of id lesions do not reveal the organism, but Trichophyton skin reactivity is positive. The most common dermatophyte infection triggering a dermatophytid reaction is tinea pedis, although id lesions occur in the setting of tinea capitis and tinea corporis as well. The autosensitization reaction to dermatophytes has been postulated to involve local reactions to fungal antigens absorbed systemically from the distant site of infection. Epidermal cytokines released from sites of skin infection have been

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Subcutaneous Tissue Infections and Abscesses

implicated as causing the heightened sensitivity of skin, at sites distant from the site of infection, to a variety of stimuli that would usually be innocuous.20 Tuberculid reactions are another example of autosensitization; these reactions, however, appear to differ in genesis from dermatophytid reactions. Skin lesions are rare, usually appearing in a host who: (1) has moderate to strong tuberculin reactivity; (2) has a history of previous tuberculous infection of other organs; and (3) usually, but not always, has shown a therapeutic response to antituberculous therapy. Lesions exhibit tuberculoid features histologically but do not contain detectable mycobacteria. Tuberculid reactions are postulated to result from hematogenous dissemination of bacilli during a transient waning of immunity; an Arthus-like reaction followed by a delayed hypersensitivity reaction to the mycobacteria may lead to the clinical lesions. Rapid local destruction may account for the absence of bacilli within lesions. The most commonly observed tuberculid is the papulonecrotic (vesicular eczema) form. Recurrent crops of symmetrically distributed, asymptomatic, firm, sterile, dusky red papules appear on the extensor aspects of the limbs, the dorsum of the hands and feet, and the buttocks. Papules can undergo central ulceration and eventually heal, leaving sharply delineated, circular, depressed scars. Lesions are characterized histopathologically by a wedge-shaped area of upper dermal necrosis, with the broad base toward the epidermis, surrounded by a tuberculoid inflammatory infiltrate. Obliterative vasculitis is noted in the deep dermis. The duration of the eruption is variable, but the lesions usually disappear promptly after treatment of the primary infection. Lichen scrofulosorum, another form of tuberculid, is characterized by asymptomatic, grouped, pinhead-sized, often follicular, pink or red papules that form discoid plaques, mainly on the trunk. Healing occurs without scarring. On histopathologic examination, granulomas are seen in the superficial dermis, often surrounding follicles and ducts of sweat glands. The inflammatory infiltrate contains epithelioid cells, giant cells, and a rim of lymphocytes. Caseation is generally absent. These lesions can be difficult to distinguish both clinically and histopathologically from lichenoid sarcoidosis. Of note, id reactions can also occur in response to noninfectious stimuli. One example is an id reaction occurring with severe contact dermatitis. In this particular example, an eczematous dermatitis can occur at sites distant from the affected area of contact dermatitis. ACKNOWLEDGMENT

The authors acknowledge significant use of the work of G.L. Darmstadt from the second edition.

CHAPTER

78

Subcutaneous Tissue Infections and Abscesses John Browning and Moise Levy

DEFINITION Infections of the skin and subcutaneous tissues can be classified as primary, secondary, or tertiary (Box 78-1). Primary infections originate directly in skin that appears to be clinically normal, although minor breaks in the integrity of the barrier function may be required (see Chapter 72, Cellulitis and Superficial Skin Infections).

CHAPTER

78

457

BOX 78-1. Subcutaneous Tissue Infections and Abscesses CUTANEOUS/SUBCUTANEOUS ABSCESSES Breast abscess Carbunclea Folliculitisa Furunclea Paronychiaa Periporitis (sweat gland abscess) Perirectal abscess SCALP ABSCESS Suppurative sialadenitisb NON-NECROTIZING SUBCUTANEOUS TISSUE INFECTIONS Cellulitis Erysipelas Secondarily infected ulcers: decubitus, diabetic NECROTIZING SUPRAFASCIAL INFECTIONS Anaerobic cellulitis • Clostridial • Nonclostridial Bacterial synergistic gangrene • Meleney ulcer Gangrenous cellulitis Necrotizing fasciitis • Type I: mixed flora • Type II: Streptococcus pyogenes • Other monomicrobial infections (see text) Fournier gangrene • Nomab • Tropical ulcer NECROTIZING DEEP FASCIAL AND MUSCLE INFECTIONS Synergistic necrotizing cellulitis Myositis Pyomyositis a

See Chapter 72, Cellulitis and Superficial Skin Infections. See Chapter 27, Infections of the Oral Cavity. See Chapter 79, Myositis, Pyomyositis, and Necrotizing Fasciitis.

b c

Some infections, such as folliculitis, furunculosis, carbunculosis, and paronychia, evolve into abscesses and occasionally extend from the epidermis or dermis to involve deeper subcutaneous tissues. Secondary infections, which occur in previously diseased or wounded skin, include infection of cysts (e.g., epidermal inclusion or pilar cyst), ulcers (e.g., decubitus ulcer), wounds (e.g., surgical wound, burn, scalp electrode site, arthropod, animal, or human bite, or burrow due to scabies, flies, or fleas), and dermatitis (e.g., atopic dermatitis, psoriasis). These infections can also spread contiguously to deeper subcutaneous tissue. Tertiary infection develops when pathogens are spread hematogenously or via lymphatics to soft tissues from a distant focus (e.g., Clostridium septicum necrotizing fasciitis or myonecrosis, Staphylococcus aureus abscess). These infections can involve any of the deeper soft tissues, but they spare direct involvement of the epidermis because of its lack of vasculature. Compromise of nutrient vessel blood flow or extension of an infectious and inflammatory nidus to the epidermis and dermis, however, can result in cutaneous disease. Infection of soft tissues can involve the skin, subcutaneous tissues and fascia, skeletal muscle, or a combination of these structures. The subcutaneous compartment is continuous over the entire body and consists of loose connective tissue containing blood and lymphatic vessels and fat. The fascia is subdivided into superficial and deep components. The superficial fascia, between the dermis and the deep fascia, is further subdivided into two layers. The outer layer, of variable thickness, contains loose collagenous tissue and fat. The inner layer of the superficial fascia is a thin membrane that has relatively little fat but is rich in elastic tissue. Superficial arteries, veins, nerves, and lymphatics lie within the superficial fascia. The deep fascia is a membranous sheet surrounding muscles and separating them into functioning units and forming the deepest boundary of the subcutaneous tissue compartment.

458

SECTION

J

Skin and Soft-Tissue Infections

This chapter describes infections that lead to abscess formation and tissue necrosis involving subcutaneous tissues as deep as the inner layer of the superficial fascia. Infections that commonly extend into the deep fascia and muscle are covered elsewhere (e.g., see Chapter 79, Myositis, Pyomyositis, and Necrotizing Fasciitis).

Abscess An abscess is a localized collection of pus in a cavity formed by disintegration or necrosis of tissue. It is recognized clinically by the presence of a firm, tender, erythematous nodule that becomes fluctuant. Histopathologic examination reveals that a cutaneous abscess has a normal epidermis, but that the dermis contains a dense aggregate of acute inflammatory cells surrounded by a fibrinoid wall. Subcutaneous abscess most commonly evolves by local extension of a primary infectious process in the epidermis or dermis, such as a cutaneous abscess originating from a skin appendage (e.g., furuncle, carbuncle, infundibular cyst, periporitis) or secondarily from a site of skin disease or injury. A subcutaneous abscess can also arise by direct traumatic implantation or invasion of pathogens into subcutaneous tissue or, occasionally, by hematogenous spread. In the patient with an abscess, constitutional symptoms are generally absent, unless the process has extended into deeper tissues or the bloodstream. Often, no predisposing factor can be identified, although a number of processes that disrupt the integrity of the barrier function of the skin or the integrity of local immunologic processes, particularly neutrophil function, are associated with abscess formation (Box 78-2). Despite the extensive list of immunodeficiency diseases

BOX 78-2. Predisposing Factors for Subcutaneous Tissue Infections and Abscesses Alteration of normal skin flora Burn wound Chronic dermatoses Corticosteroid therapy Foreign body Immunodeficiency diseases Chédiak–Higashi syndromea Chronic granulomatous diseasea Congenital neutropeniaa Cyclic neutropeniaa Griscelli syndrome Hyperimmunoglobulin E syndromea Leukocyte adhesion deficiencya Leukocyte alkaline phosphatase deficiencya Neutrophil-specific granule deficiencya Transient hypogammaglobulinemia of infancya Wiskott–Aldrich syndromea X-linked hypogammaglobulinemiaa Intravenous drug abuse Malnutrition Peripheral vascular disease Obstruction of drainage Ischemia Skin trauma Surgical wound Circumcision Umbilical cord stump Systemic disease Cachexia Cirrhosis Diabetes mellitus Neoplasia Leukemia, lymphoma Solid tumor Organ transplantation Renal tubular acidosis Renal failure a

Associated with abscess formation.

associated with abscess formation, most individuals with recurrent furunculosis lack evidence of immunodeficiency.

Necrotizing Cellulitis Subcutaneous Tissue Infection and Fasciitis Necrotizing soft-tissue infections are potentially life-threatening conditions characterized by rapidly advancing, local tissue destruction and systemic toxicity. Tissue necrosis distinguishes them from cellulitis (see Chapter 72, Cellulitis and Superficial Skin Infections), in which an inflammatory infectious process involves but does not destroy subcutaneous tissues. Unlike an abscess, necrotizing softtissue infections involve diffuse tissue necrosis and lack localized purulence, although a subcutaneous abscess can occasionally form. Such infections can occur anywhere on the body, the most common locations being the extremities, abdomen, and perineal region. Incidence is highest in hosts with immunocompromise, either systemic or of local tissue, particularly in those who have diabetes mellitus, neoplasia, peripheral vascular disease, have undergone recent surgery, or are receiving immunosuppressive treatment, especially with corticosteroids. Necrotizing soft-tissue infections can also occur in healthy individuals at enterostomy sites or after: (1) minor puncture wounds, abrasions, or lacerations; (2) vesicular viral infections; (3) blunt trauma; (4) surgical procedures, particularly of the abdomen, gastrointestinal, or genitourinary tracts, or perineum; or (5) hypodermic needle injection. Since the mid-1980s, there has been a resurgence of fulminant necrotizing soft-tissue infections due to Streptococcus pyogenes and most recently due to Panton–Valentine leukocidin-positive Staphylococcus aureus, often in healthy individuals with little (e.g., varicella-zoster virus infection) or no apparent compromise of immunologic or skin integrity. This observation stresses the importance of bacterial virulence factors as well as absence of specific host resistance factors in the pathogenesis.1 Necrotizing soft-tissue infections form a continuum, some developing primarily in the more superficial layers of the subcutaneous tissues and others typically extending to the deep fascia and muscle. Although the rapidity and extent of tissue destruction as well as the causative agent vary, patients characteristically manifest a paucity of early cutaneous signs relative to the rapidity and extent of destruction of the subcutaneous tissues. Early clinical findings include ill-defined cutaneous erythema and edema that extends beyond the area of erythema. Also, in distinction from cellulitis, which may have a more distinct border, the pain, tenderness, and constitutional signs are often out of proportion to the cutaneous findings in necrotizing soft-tissue infection. This is particularly true with involvement of the deeper tissues, such as deep fascia and muscle. In general, patients with involvement of the superficial or deep fascia or muscle tend to be more acutely and systemically ill and to have more rapidly advancing disease than those with infection and tissue destruction largely confined to subcutaneous tissues above the fascia. Other cutaneous signs, such as vesiculation, bulla formation, ecchymoses, crepitus, anesthesia, and necrosis, are ominous and indicative of advanced disease.2 Classification of necrotizing soft-tissue infections has undergone multiple revisions over the years, resulting in a confusing body of literature on the subject, often with more than one name for the same condition. Failure to recognize differences among these infections can lead to suboptimal therapy, because the need for aggressive surgical intervention and optimal antibiotic therapy varies. The most important question is whether the soft-tissue infection is necrotizing or nonnecrotizing; the former requires prompt surgical removal of all devitalized tissue in addition to antimicrobial therapy, whereas the latter responds to antibiotic therapy alone. If the distinction is in doubt, magnetic resonance imaging or incisional biopsy can be helpful.3,4 These procedures, however, should not delay surgical exploration and intervention in the course of destructive or potentially fulminant

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Subcutaneous Tissue Infections and Abscesses

infection. Ultimately, precise determination of the planes of tissue involved in the infection cannot be made on clinical examination, but rather must be made definitively in the operating room. Current classification schemes for necrotizing soft-tissue infections focus on the depth of soft-tissue destruction and the etiologic agent.5 Disease entities that primarily involve subcutaneous tissues (sparing fascia and muscle) include clostridial and nonclostridial anaerobic cellulitis, bacterial synergistic gangrene, and Meleney ulcer. Other conditions that can evolve secondarily into subcutaneous tissue infection are infected decubitus or diabetic ulcers and tropical ulcers. When infection involves the deep layer of the superficial fascia but largely spares the adjacent skin, deep fascia, and muscle, it is termed necrotizing fasciitis. Necrotizing fasciitis encompasses a variety of clinical presentations, ranging from subacute to fulminant, and can be due to a number of pathogens, sometimes apparently acting alone and sometimes synergistically. Synergistic necrotizing cellulitis is indistinguishable clinically from necrotizing fasciitis in most cases, but it frequently extends to involve the deep fascia and muscle.6

ETIOLOGY Identification of the pathogen or pathogens causing a particular subcutaneous soft-tissue infection or, in some cases, exclusion of a variety of noninfectious diseases in the differential diagnosis of subcutaneous tissue infections (Box 78-3), requires proper use of diagnostic tests and correct interpretation of their results. If the surface of a wound or site of infection is sampled with a swab and cultured, a number of organisms that are colonizing the area but are not contributing to the disease process may be identified. If an organism is identified on both Gram stain (or special stains in the case of fungi) and culture, the likelihood of its pathogenicity is increased. Chances of identifying the true pathogen are further improved if culture specimens are obtained by swabbing of the exudate from a site of suppuration, by fine-needle aspiration, or by excisional biopsy.7 Cultures should be obtained for both routine

CHAPTER

78

459

and fastidious, aerobic and anaerobic organisms as well as molds if appropriate.

Abscess The principal pathogens of subcutaneous abscesses in children vary with body site. The single most common pathogen, Staphylococcus aureus,8 is prevalent, especially in abscesses of the neck, trunk, and extremities, although it can be found originating at all skin surfaces. For abscesses from which only one organism is isolated, the agent is S. aureus in most cases.8–10 Most abscesses of the perineal (inguinal, buttock, perirectal, vulvovaginal) region contain multiple species of facultative and anaerobic fecal organisms, particularly Bacteroides spp. Occasionally, anaerobic bacteria appear to act alone to cause abscess. Encapsulated anaerobic bacteria generally appeared to be more important than a variety of aerobic bacteria.11 This finding contrasts markedly with observations in superficial skin infections in children, in which anaerobic bacteria play little pathogenic role.12 Likewise, anaerobic bacteria do not appear to be capable of acting alone to cause necrotizing soft-tissue infections.13 In general, cultures from perineal or perioral abscesses and ulcers in both children and adults contain organisms that reside normally on adjacent mucous membranes rather than skin, whereas lesions remote from the rectum or mouth primarily contain organisms that reside normally on skin at that site.8–10,14–18 Overall, the most common aerobic and facultative bacteria isolated from cutaneous and subcutaneous abscesses and decubitus ulcers in children are S. aureus, Streptococcus pyogenes, a-hemolytic and nonhemolytic streptococci, Enterococcus spp., Enterobacter spp., Escherichia coli, and Pseudomonas aeruginosa (Box 78-4).8,17 The predominant anaerobic isolates are Bacteroides, Peptostreptococcus, Peptococcus, Fusobacterium, Prevotella, and Porphyromonas species. Immunocompromised hosts are at risk of abscess formation because of a broader spectrum of pathogenic microbes, particularly fungi.19

Necrotizing Subcutaneous Tissue Infections BOX 78-3. Noninfectious Diseases in the Differential Diagnosis of Subcutaneous Tissue Infections in Infants and Children Acne conglobata Acne fulminans Acute hemorrhagic edema of infancy Burn Calciphylaxis Crohn disease Dissecting cellulitis of the scalp Eosinophilic fasciitis Envenomation Jellyfish, scorpion, snake, spider (brown recluse) bites Factitial disease Hemangioma (Deep) Hemangioendothelioma/Kasabach–Merritt Syndrome Hidradenitis suppurativa Langerhans cell histiocytosis Leukemia Lymphoma Myiasis Necrobiosis lipoidica diabeticorum Nodular fasciitis Panniculitis Polyarteritis nodosa Purpura fulminans Pyoderma gangrenosum Subcutaneous gas: esophageal perforation, peribronchial dissection, traumatic cutaneous installation of air, wound irrigation with H2O2 Superficial thrombophlebitis Sweet syndrome (acute febrile neutrophilic dermatosis) Tungiasis Wegener granulomatosis

Relatively few organisms possess sufficient virulence to cause necrotizing soft-tissue infections when acting alone (Box 78-5). Fulminating infections are caused by S. pyogenes, although a clinically indistinguishable infection of necrotizing fasciitis can occasionally be caused by Staphylococcus aureus (including methicillin-resistant S. aureus (MRSA)), Clostridium perfringens, C. septicum,, Pseudomonas aeruginosa, Vibrio spp., particularly Vibrio vulnificus, and fungi of the

BOX 78-4. Agents of Subcutaneous Abscesses in Infants and Children BACTERIA Actinomyces spp., Bacteroides spp.,a Brucella spp., Clostridium spp.,a Eikenella corrodens, Enterobacter spp.,a Enterococcus, Escherichia coli,a Fusobacterium spp.,a Klebsiella spp.,a Mycobacterium spp., Neisseria gonorrhoeae, Nocardia spp., Peptococcus spp.,a Peptostreptococcus spp.,a Porphyromonas spp.,a Prevotella spp.,a Propionibacterium acnes, Proteus spp.,a Pseudomonas aeruginosa, Pseudomonas mallei, Pseudomonas pseudomallei, Serratia marcescens, Staphylococcus aureus,a streptococci (a-hemolytic, nonhemolytic),a group B streptococcus, Streptococcus pyogenesa FUNGI Aspergillus spp., Candida albicans, Histoplasma capsulatum, hyalohyphomycoses, Penicillium marnefei, pheohyphomycoses, Pseudallescheria boydii, Rhizopus spp. MYCOPLASMA Mycoplasma hominis PROTOZOA Entamoeba histolytica a

Most common.

460

SECTION

J

Skin and Soft-Tissue Infections

BOX 78-5. Agents of Necrotizing Soft-Tissue Infections in Infants and Children MONOMICROBIAL Bacteria Clostridium perfringens, Clostridium septicum, Haemophilus influenzae type b, Mycobacterium chelonae, Mycobacterium kansasii, Pseudomonas aeruginosa, Staphylococcus aureus, Streptococcus agalactiae, streptococci (groups C, G, and F), Streptococcus pneumoniae, Streptococcus pyogenes, Vibrio alginolyticus, Vibrio cholerae (nonserogroup 01), Vibrio parahaemolyticus, Vibrio vulnificus Fungi Absidia spp., Aspergillus spp., Fusarium spp., Mucor spp., Rhizopus spp., Saksenaea spp. Protozoa Entamoeba histolytica POLYMICROBIAL Anaerobic bacteria Bacteroides spp., Clostridium spp., Fusobacterium spp., Peptococcus spp., Peptostreptococcus spp., Porphyromonas spp., Prevotella spp. Aerobic or facultative bacteria Aeromonas spp., Enterobacter spp., Escherichia coli, Klebsiella spp., Proteus spp., Pseudomonas spp., Staphylococcus aureus, streptococci (hemolytic, nonhemolytic nongroup A).

order Mucorales, particularly Rhizopus spp., Mucor spp., and Absidia spp.1,20–22,23 The characteristic feature of necrotizing fasciitis due to the Mucorales is occurrence in an immunocompromised host as well as the presence of angioinvasive, broad, nonseptate hyphae, 10 to 20 μm in diameter, with right-angle branching. On rare occasions, nongroup A streptococci, such as group B, C, G, or F streptococci, Streptococcus pneumoniae, and Haemophilus influenzae type b, have been reported to cause necrotizing fasciitis; the adequacy of culture methods for anaerobic bacteria to furnish evidence of a synergistic infection is not clear in all these cases, however. Necrotizing soft-tissue infections are often polymicrobial. In most cases, a mixture of anaerobic, aerobic, and facultative anaerobic bacteria appears to act together to cause tissue necrosis.13 Anaerobic isolates are similar to those found in subcutaneous abscesses, consisting of Bacteroides, Peptostreptococcus, Peptococcus, Prevotella, Porphyromonas, Clostridium, and Fusobacterium species.13,24 The most common aerobic or facultative bacteria are Escherichia coli, Enterobacter spp., and a variety of other Enterobacteriaceae, Pseudomonas spp., several species of hemolytic or nonhemolytic, nongroup A streptococci, and Staphylococcus aureus. Bacterial synergistic gangrene is due to microaerophilic streptococci and either S. aureus or a gram-negative bacillus, particularly Proteus spp., acting together to cause tissue damage.25,26 Entamoeba histolytica and nontuberculous mycobacteria in steroid-dependent patients, particularly Mycobacterium kansasii and M. chelonae, can cause an identical clinical syndrome.25,27 The most common association of organisms in tissue of necrotizing fasciitis is Bacteroides spp. with either nongroup A streptococci or Enterobacteriaceae, particularly Escherichia coli.13,24 No particular combination of organisms, however, is diagnostic for any given clinical entity that involves necrotizing soft-tissue infection. Gas in tissue in association with infection is a hallmark of clostridial infection, but most soft-tissue infections with gas production are due to facultative gram-negative bacilli, such as E. coli, Klebsiella spp., Proteus spp., and Aeromonas spp.28 Crepitant anaerobic cellulitis due to Clostridium spp. is also more common than life-threatening clostridial myonecrosis (i.e., gas gangrene) (see Chapter 79, Myositis, Pyomyositis, and Necrotizing Fasciitis). Tertiary or hematogenous infections of the subcutaneous tissues occur most commonly in immunocompromised hosts and can lead to abscess or necrotizing soft-tissue infection. Although a wide range of organisms, including bacteria and fungi, can be involved, the most common are Staphylococcus aureus and Pseudomonas aeruginosa.

PATHOGENESIS Some of the same organisms that cause necrotizing soft-tissue infections sometimes cause cellulitis with involvement of fascia or tissue necrosis. Responsible factors are poorly understood, involving a complex interaction among predisposing tissue factors such as local trauma, anaerobic wound environment, systemic and local host defense, and bacterial virulence factors and synergism (see Box 78-2). Compromise of immune defense is a primary predisposing factor for development of subcutaneous infections. This process can occur locally, through trauma or surgery that compromises the barrier function of the skin, or can involve systemic immunodeficiency, especially in individuals with defective neutrophil function. Systemic immunodeficiency is especially important for the development of necrotizing infections due to P. aeruginosa or fungi. Innate cutaneous defenses, including the physical barrier afforded by the lipids and the cross-linked cornified envelope in the stratum corneum, appear to be sufficient in most cases to prevent infection. Through trauma to skin or mucosal surfaces, however, a pathogen may be able to adhere to a previously unexposed receptor or a receptor that is newly synthesized in the injured skin. In some infections, such as those due to Clostridium spp., the depth of the initial injury is the most important determinant of depth of infection.28 Other infections, such as necrotizing fasciitis due to Streptococcus pyogenes, can spread rapidly from superficial layers of the skin to deeper subcutaneous tissues, aided by as yet poorly understood virulence factors. In infections that follow varicella, access to tissue beneath the stratum corneum at sites of chickenpox lesions may allow initiation of local infection.29–33 In diabetes, both compromised blood flow to the skin due to small-vessel vasculopathy and impaired function of neutrophils may predispose to necrotizing infection. The relatively high incidence of subcutaneous infections in those with peripheral vascular disease due to other causes supports this concept of pathogenesis, as does the fact that thrombosis of vessels is a major histologic feature of many necrotizing infections.34,35 Furthermore, a higher degree of vascular thrombosis has been correlated with more acute presentations of necrotizing fasciitis.34 In necrotizing fasciitis, the primary disease is in the superficial fascia, with thrombosis of nutrient vessels leading secondarily to ischemia and necrosis of subcutaneous tissue. The cause of vascular thrombosis is not completely understood, but probably involves direct cytolytic or thrombogenic factors released from bacterial pathogens, immune-mediated vascular damage due to the inflammatory infiltrate surrounding the blood vessels, or noninflammatory intravascular coagulation.35 Disease in the dermis and epidermis occurs secondarily, after thrombosis of deeper vessels. Synergism of microorganisms does not appear to be necessary for abscess formation, but it does appear to be operative in many necrotizing soft-tissue infections, because anaerobic or facultative organisms are rarely found alone in these infections.13 The concept of bacterial synergism in the pathogenesis of soft-tissue infection, as advanced in the 1920s by Brewer & Meleney,36 involved microaerophilic (probably anaerobic) streptococci and Staphylococcus aureus, causing a progressive, chronic subcutaneous tissue infection called “bacterial synergistic gangrene.” Since then, numerous combinations of organisms have been found to act synergistically in subcutaneous infections.37 Even streptococcal necrotizing fasciitis and clostridial anaerobic cellulitis, in some cases, are caused by more than one agent acting synergistically, S. aureus being isolated from the infected tissue of some patients who have streptococcal necrotizing fasciitis.13,24,34 Synergistic pathogenesis of anaerobic, aerobic, and facultative bacteria is supported by clinical data and animal models, including studies in tropical ulcer showing synergistic associations of Fusobacterium spp., other anaerobic bacteria, and spirochetes.38,39 In an animal model of necrotizing fasciitis, S. aureus or crude staphylococcal a-lysin potentiated the pathogenesis of Streptococcus pyogenes.40 The mechanism of synergy appears to vary according to the organisms present, but it may result from mutual protection

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Subcutaneous Tissue Infections and Abscesses

from phagocytosis and intracellular killing, promotion of bacterial capsule formation, production of essential growth factors or energy sources, and utilization of oxygen by facultative bacteria, thus lowering tissue oxidation reduction potential to facilitate growth of anaerobic organisms.37,41,42 Extracellular bacterial toxins and enzymes are important factors in the destruction of subcutaneous tissue.43 In necrotizing soft-tissue infections due to Clostridium spp. and S. pyogenes in children, few acute inflammatory cells can be identified in necrotic fascia and subcutaneous tissue. Necrotizing tissue infections appear to be related to a production of a variety of potent endotoxins and enzymes.44 Initiation of necrotizing streptococcal infection appears to require a break in the skin or mucous membrane, although the injury is often blunt, trivial, or inapparent.2 Adherence of S. pyogenes to respiratory epithelium has been postulated to involve a two-step mechanism: a relatively weak, nonspecific hydrophobic bond followed by highaffinity, specific attachment.45 The mechanism of adherence of the bacterium to the skin is not well understood. Keratinocyte differentiation promotes adherence of S. pyogenes,46 whereas the primary epidermal cytokines tumor necrosis factor-a (TNF-a) and interleukin-1a (IL-1a) decrease adherence.47 The hyaluronic acid capsule of S. pyogenes impedes adherence, and modulation of capsule expression may be important in the pathogenesis of skin infections.48 The molecular interactions (i.e., bacterial adhesin and keratinocyte receptor) involved in initial attachment to the skin, however, remain unknown. After attachment, invasion of S. pyogenes is related to numerous extracellular enzymes and toxins, and membrane-bound proteins that function as virulence factors.49,50 Patients with streptococcal necrotizing fasciitis can also experience toxic shock-like syndrome, associated with the presence of the streptococcal pyrogenic exotoxins A or B or both.51 These toxins, as well as certain M-protein fragments, appear to have the ability to interact simultaneously with the major histocompatibility complex (MHC) class II antigen on antigen-presenting cells as well as specific Vß regions of T-lymphocyte receptors, inducing massive synthesis and release of monokines, including TNF-a, IL-1b, IL-6, and the lymphokines TNF-b, IL-2, and interferon-g.33 These cytokines, particularly TNF, may mediate, at least in part, the rapid, massive tissue destruction seen in necrotizing fasciitis.2 Protease activity and production of Panton–Valentine leukocidin are postulated to correlate better with the ability of Staphylococcus aureus to cause invasive disease and may be a mechanism for invasion of pyogenic bacteria from a cutaneous site.20,21,52,53 A central feature of the disease in necrotizing fasciitis, as in many other necrotizing soft-tissue infections, is vascular injury and thrombosis of the arteries and veins passing through the fascia.3,34 A possible mechanism for the vascular injury leading to tissue ischemia and necrosis in streptococcal necrotizing fasciitis is greater adherence of neutrophils to vascular endothelium as a result of streptolysin O-induced upregulation of receptors.54

CLINICAL CONDITIONS Abscess Breast Abscess Breast abscess is an uncommon infection of neonates usually due to S. aureus and occasionally caused by group B streptococcus, E. coli, Salmonella spp., Proteus mirabilis, or Pseudomonas aeruginosa. Although anaerobic organisms can be isolated from up to 40% of infections, their pathogenic role in neonates is questionable, and therapy directed specifically against them is unnecessary.55,56 Breast abscess develops in full-term neonates during the first 1 to 6 weeks of life, most commonly during the second to third week.55–59 Incidence of breast abscess is twofold higher in girls overall, but is approximately equal in boys and girls during the first 2 weeks of life.57 Physiologic breast enlargement, which is more common in

CHAPTER

78

461

infant girls than in boys after, but not before, 2 weeks of age,60 is probably a factor in the pathogenesis, as also evidenced by the absence of both breast enlargement and breast abscess in premature infants.57 Other factors are involved, however, because a relatively large retrospective review of neonates with breast abscess found that only about half had breast enlargement before development of abscess.57 It is likely that organisms from the nasopharynx or umbilicus colonize the skin of the nipple and move retrogradely via glandular ducts of physiologically enlarged breasts.61 Breast enlargement, accompanied by varying degrees of erythema, induration, and tenderness, is present initially and can progress to fluctuation, depending in part on how promptly antibiotic therapy is initiated. Bilateral infection is extremely rare. Fever or constitutional symptoms, such as irritability and toxicity, are present in approximately a third of cases, although leukocytosis (> 15 000 cells/mm3) is found in approximately one-half to two-thirds.55 Breast abscess due to Staphylococcus aureus is accompanied by cutaneous pustules or bullae on the trunk, particularly in the periumbilical or perineal region, in 25% to 50% of patients.57 The symptoms, age at presentation, and clinical findings for infections due to gramnegative bacilli or anaerobes are similar to those for staphylococcal abscesses, except that infants infected with Salmonella spp. also have gastrointestinal symptoms.59 The most common complication of breast abscess is cellulitis, which develops in approximately 5% to 10% of affected infants.57 Cellulitis is generally localized but can rarely extend rapidly over the shoulder or abdomen.58 Other complications, such as bacteremia, pneumonia, osteomyelitis, and septicemia, are unusual. Scar formation leading to decreased breast size after puberty can be a late sequela.57 Results of Gram stain and culture of material expressed from the nipple or obtained by needle aspiration or incision and drainage help guide antibiotic therapy. If fluctuation is present, the abscess must be drained; antibiotic therapy is adjunctive. If fluctuation is absent, antibiotic therapy alone may be curative. Warm compresses may also be beneficial. A b-lactamase-resistant antistaphylococcal antibiotic is given parenterally or in areas with a high incidence of MRSA infections, vancomycin or clindamycin is given. If gramnegative bacilli or no organisms are seen on Gram stain or if the infant appears ill, initial therapy should also include an aminoglycoside agent or cefotaxime. In most instances, a 5- to 7-day course of therapy is sufficient, although therapy is sometimes continued for 10 to 14 days; duration of parenteral therapy depends on the isolate, the presence of bacteremia, and the clinical course.

Sweat Gland Abscess (Periporitis) Sweat gland abscesses develop rarely in neonates, most often in association with malnutrition or debilitation.62 The infection has also been termed periporitis staphylogenes because of the almost uniform presence of S. aureus in the lesions. It appears that lesions of miliaria become infected secondarily, followed by extension of the infection into the sweat gland apparatus and, occasionally, into the adjacent subcutaneous tissue. Miliaria-like lesions, however, are not a constant feature. The 1- to 2-cm, round to oval nodular abscesses occur most commonly on the neck, occiput, back, and buttocks, and unlike furuncles and carbuncles of follicular origin, they are nontender, nonpointing, and cold. Constitutional symptoms can accompany numerous large abscesses, and lymphangitis or cellulitis occurs rarely. Therapy consists of control of factors such as skin occlusion or fever that predispose to miliaria, correction of malnutrition, local care of abscesses, and use of antistaphylococcal antibiotics. Healing occurs over several weeks, generally without scarring.

Perirectal Abscess Perirectal abscess tends to occur in healthy neonates and infants, more often in boys than girls. There may be no apparent predisposing

462

SECTION

J

Skin and Soft-Tissue Infections

factor, or the abscess can develop after minor abrasions or fissures, particularly in association with diarrhea or constipation. Older children who experience perirectal abscess more frequently have a predisposing condition, such as: (1) neutropenia in association with neoplastic disease, autoimmunity, or chemotherapy; (2) neutrophil dysfunction due to immunodeficiency disease, such as chronic granulomatous disease; (3) acquired immunodeficiency syndrome; (4) diabetes mellitus; (5) corticosteroid therapy; (6) ulcerative colitis; (7) Crohn disease; (8) hidradenitis suppurativa; or (9) prior rectal surgery. In those with granulocytopenia, the risk of development of an abscess or perirectal cellulitis increases with the presence of perirectal mucositis, hemorrhoids, rectal fissure, or manipulation. A break in the mucosal barrier or occlusion of anal crypts usually initiates infection. The predominant organisms in perirectal abscesses are mixed anaerobic and aerobic flora of the intestine and skin of the anal verge, particularly Bacteroides, Peptococcus, Peptostreptococcus, Porphyromonas, Fusobacterium, Clostridium species, E. coli, Staphylococcus aureus, Streptococcus pyogenes, Pseudomonas aeruginosa, Klebsiella spp., and Proteus spp.63–67 Perirectal abscess is rarely due to Entamoeba histolytica, Mycobacterium spp., Nocardia spp., and Actinomyces spp. Buttock abscesses are most commonly due to Staphylococcs aureus, even in the neonate.63 Superficial abscess in the infant usually manifests as pain on defecation, sitting, or walking and the presence of redness, swelling, and tenderness in the perianal region. A superficial abscess in the epithelium can potentially extend in any of the following directions: inferiorly along the anal sphincter to exit next to the anus on the buttock (fistula in ano), laterally through the external sphincter to the ischiorectal fossa to form a deep abscess, or superiorly to the deep space between the internal sphincter and the levator ani muscles.68–70 An abscess in deeper tissues can be accompanied by poorly localized, deep pain and constitutional symptoms. An anorectal abscess may not be apparent externally, but pain is generally elicited upon rectal examination. In immunocompetent infants, a superficial perianal abscess can drain spontaneously (and can be self-limited) or should be drained promptly. Surgical exploration of the abscess for fistula depends on clinical circumstances and isolation of organisms other than S. aureus. Empiric therapy with antibiotics to cover anaerobic and aerobic gram-negative bacilli and S. aureus may prevent regional spread of the infection and decrease the incidence of complications.64 A 7- to 10-day course of therapy is recommended. Children with granulocytopenia may have delayed development of erythema, induration, and fluctuation. In the absence of fluctuation, extensive soft-tissue disease, or septicemia, a trial of parenteral antimicrobial therapy may be initiated alone. If progression of disease or fluctuation becomes apparent, surgery should be undertaken. Complications, which occur more commonly in children with underlying disease, include anorectal fistula, recurrence of abscess, bloodstream infection (BSI), and necrotizing fasciitis. A chronic fistula may require a fistulotomy.69

Scalp Abscess Scalp abscess appears as a localized, erythematous area of induration 0.5 to 2 cm in diameter usually at the insertion site of a fetal scalp monitoring electrode. The site can become fluctuant, pustular, or suppurative. Presentation occurs most commonly on the third or fourth day of life (range, 1 day to 3 weeks).71 Regional lymphadenopathy can be present, but other, more serious complications, such as cranial osteomyelitis, subgaleal abscess, necrotizing fasciitis of the scalp, and BSI, are uncommon. Death associated with a complication of fetal scalp electrode placement has been described in a premature infant who had E. coli scalp abscess and BSI.72 The incidence of scalp abscess after placement of a spiral fetal scalp electrode (the type used since the early 1970s) ranged from 0.1% to 1.0% in retrospective studies of approximately 18 000 neonates71; prospective studies have reported the incidence to be

4.5%73 and 0.56%.74 Reported presence of predisposing factors is conflicting. Okada and associates73 found significant association with longer duration of placental membrane rupture and of monitoring; monitoring for high-risk indications, particularly prematurity; and nulliparous birth, possibly because of a higher risk of infection of an edematous, hypoxic caput succedaneum. Scalp trauma and compression per se, however, are questionable factors in the pathogenesis of scalp abscess, because most reports have not noted abscess formation at sites of scalp trauma and abrasion due to forceps or vacuum extraction, at sites of fetal blood sampling, or in association with hematomas.71,75 There are exceptions, because cephalohematomas can become infected, and similar rates of scalp abscess have been reported among infants who were monitored with scalp electrodes and those who were not monitored but were delivered by forceps or vacuum extraction.76 A threefold greater but not statistically significant rate of scalp abscess has been noted among infants delivered by vacuum extraction compared with those born by spontaneous vaginal delivery.73 Plavidal & Werch77 also found an association between prolonged rupture of membranes and development of scalp abscess, but Wagener and colleagues74 did not. Risk factors in this latter study included number of vaginal examinations, concurrent monitoring with an intrauterine pressure catheter, use of more than one spiral electrode, fetal scalp blood sampling, maternal diabetes, and endomyometritis.74 Scalp abscess is typically a polymicrobial infection. Cultures from approximately one-third of abscesses reveal aerobic or facultative organisms alone, 10% to 25% show anaerobic isolates alone, and 40% to 60% yield a mixture of aerobic or facultative and anaerobic bacteria.71,73,78 The most common aerobic isolates are Staphylococcus aureus, group A, B, or D streptococci, S. epidermidis, and, occasionally, Haemophilus influenzae, E. coli, Klebsiella pneumoniae, Enterobacter spp., Pseudomonas aeruginosa, and Neisseria gonorrhoeae. The role of S. epidermidis as a pathogen is questionable. Common anaerobic isolates are Peptococcus, Peptostreptococcus, and Bacteroides species, Propionibacterium acnes, and Clostridium spp. The anaerobic flora in the abscesses reflects that found in the normal cervix during labor. Neonatal necrotizing fasciitis of the scalp has been reported in association with intrapartum fetal scalp electrode monitoring.79 The primary differential diagnostic concern in fetal scalp abscess is herpes simplex virus (HSV) infection.71 The time of appearance of herpetic lesions (peak incidence, 4 to 10 days) overlaps with that for scalp abscess, and the lesions can be indistinguishable clinically. Dissemination of HSV can also occur.80 If suspicion of HSV exists, therapy with acyclovir should be initiated while awaiting results of diagnostic tests. Infants who are subjected to scalp electrode monitoring during birth should be followed closely during the first weeks of life for evidence of infection. Parents should be instructed in surveillance. If weeping, vesiculopustular lesions or abscess is noted, specimen should be obtained by needle aspiration or swabbing of the exudate from the puncture site for direct testing and culture for both aerobic and anaerobic organisms and HSV. Many bacterial infections of the fetal scalp resolve spontaneously, but if fluctuation develops without spontaneous suppuration, incision and drainage are necessary; extensive debridement should not be performed, however. If surrounding cellulitis is present, a 5- to 7-day course of parenteral antibiotic therapy is usually sufficient, with culture results guiding antibiotic choice. HSV scalp lesions can also heal spontaneously, but reactivation or dissemination or both can occur, with serious consequences; parenteral acyclovir therapy should be administered.

Necrotizing Subcutaneous Suprafascial Infections Anaerobic Cellulitis Anaerobic cellulitis can be caused by Clostridium spp. or by nonclostridial pathogens.81 Clostridium spp. are associated with a spectrum of clinical infections, ranging from wound contamination

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Subcutaneous Tissue Infections and Abscesses

to anaerobic cellulitis to myonecrosis (gas gangrene). Contamination occurs when clostridia grow in devitalized tissue but do not invade surrounding healthy tissue. Contamination can progress to cellulitis, which spares the fascia and muscle. Cellulitis is usually due to C. perfringens or sometimes other Clostridium spp. The incidence of anaerobic cellulitis due to Clostridium is severalfold greater than that of clostridial myonecrosis.82 Nonclostridial anaerobic cellulitis also spares the fascia and underlying muscle. Nonclostridial anaerobic cellulitis is generally due to Bacteroides spp., Peptostreptococcus spp., or Peptococcus spp. acting synergistically with facultative streptococci and staphylococci, or with facultative or aerobic gram-negative bacilli such as E. coli, K. pneumoniae, and Aeromonas spp. Anaerobic cellulitis develops in devitalized subcutaneous tissues after spread of a local primary infection or introduction of the pathogen into subcutaneous tissues by trauma or during surgery. It is most prevalent on areas of the body subject to fecal soiling, such as the perineum, abdominal wall, buttocks, and lower extremities. In one series, 5 cases of penetrating wound trauma that did not involve any aquatic environment were complicated by rapidly progressive infections.81 Rarely, anaerobic cellulitis occurs as a tertiary infection after BSI with C. septicum in an immunocompromised patient who has a solid tumor or hematologic malignancy and neutropenia.82 In this setting, progression to myonecrosis is usually rapid. Onset of nonclostridial anaerobic cellulitis is relatively gradual, after an incubation period of several days, with development of localized, ill-defined erythema, edema, and tenderness. Once it is established, destruction of subcutaneous tissue can occur rapidly, but the usual course is indolent compared with that of clostridial myonecrosis, with little tissue edema, necrosis, or pain, and few constitutional symptoms. Generally, there is minimal cutaneous discoloration, even late in the infection. By contrast, as clostridial myonecrosis advances, the skin becomes bronze-colored, with dark bullae and patches of necrosis. Gas formation with tissue crepitus is usually remarkable in anaerobic cellulitis, out of proportion to the clinical appearance of the skin, and beyond that seen with myonecrosis. Gas in tissue is usually visible on radiographs. A thin, dark, foul-smelling exudate may drain from the wound. Gram-stained smears of tissue or exudate may help distinguish clostridial from nonclostridial infection. In addition, the inflammatory infiltrate is often more substantial with nonclostridial anaerobic cellulitis, particularly early in the course of the disease, because tissue damage with Clostridium infection appears to be due in large part to histotoxic exotoxins. Culture of aspirates or tissue specimens obtained during surgery is definitive. Identification of Clostridium species in a patient without acute, severe illness is reassurance that myonecrosis has not occurred. Surgical exploration is necessary, however, to distinguish this entity definitively from clostridial myonecrosis. At the time of surgery, all areas of necrotic tissue must be exposed and debrided. Recommended initial antibiotic therapy consists of intravenous penicillin or ampicillin along with clindamycin or metronidazole. An aminoglycoside or third-generation cephalosporin should also be given if gram-stained specimens suggest the presence of gramnegative bacilli. Final antibiotic selection can be tailored according to results of cultures and antibiotic susceptibility testing.83

Bacterial Synergistic Gangrene Bacterial synergistic gangrene is a rare, chronic gangrenous infection of the skin and subcutaneous tissue that occurs almost uniformly on the trunk after abdominal surgery. The most common sites are at the exit of a fistulous tract, particularly after appendectomy or drainage of empyema, and in association with an ileostomy or colostomy.25 Occasionally, bacterial synergistic gangrene develops in close proximity to chronic ulceration on an extremity. A related lesion, chronic undermining ulcer of Meleney, occurs most commonly after lymph node surgery in the neck, axilla, or groin or occasionally after colonic or gynecologic surgery.

CHAPTER

78

463

Bacterial synergistic gangrene is due to microaerophilic or anaerobic streptococci in combination with Staphylococcus aureus or, occasionally, Proteus spp. or other gram-negative bacilli.26,84 Identical lesions, from both clinical and histopathologic points of view, can be caused by cutaneous amebiasis due to Entamoeba histolytica.25 Bacterial synergistic gangrene is characterized by severe pain and slow but inexorable progression of gangrenous ulceration. It begins with localized tenderness, erythema, and edema, which progresses to ulceration. The ulcer characteristically has a central floor of red granulation tissue with gray to yellow exudate, and a sharply demarcated gangrenous serpiginous border. A raised, dusky red to purple margin surrounds the ulcer, with a peripheral ring of erythema and edema. Meleney ulcer is characterized by burrowing necrotic sinus tracts that emerge at distant sites to form additional ulcers. Systemic signs are minimal in both conditions. Without treatment, these lesions have minimal tendency to heal, ultimately destroying large areas of skin and subcutaneous tissue. Although antimicrobial therapy alone has been curative in some cases, nutritional support, local surgical debridement, and antibiotic therapy (e.g., with clindamycin plus gentamicin, or another agent for facultative and aerobic bacteria, plus metronidazole) are recommended.26 Broad-spectrum antibiotic therapy should be initiated and then adjusted according to results of culture and susceptibility testing of facultative and aerobic organisms recovered from surgical material.

Necrotizing Fasciitis Causes of necrotizing fasciitis cannot be distinguished clinically, although development of crepitation signals the presence of Clostridium spp. or gram-negative bacilli such as Escherichia coli, Klebsiella, Proteus, and Aeromonas species. Necrotizing fasciitis with toxic shock-like syndrome due to Streptococcus pyogenes is perhaps the most fulminant infection known85; bloodstream invasion of Panton–Valentine leukocidin-producing Staphylococcus aureus can lead to severe consequences of septic thrombosis.20,21 Necrotizing fasciitis typically begins at a site of blunt or cutaneous trauma. Common predisposing conditions in neonates are omphalitis and balanitis after circumcision.86,87 Gross compromise of the integument is a predisposing factor, but in many instances, the trauma is trivial to inapparent. Cases reported in children in the mid-1990s highlighted the occurrence of necrotizing fasciitis due to Streptococcus pyogenes after superinfection of varicella lesions, sometimes in association with the use of nonsteroidal antiinflammatory agents.29–32 However, with the advent of the varicella vaccine, this incidence has decreased.88 Underlying systemic disease, such as diabetes, peripheral vascular disease, or immunosuppression, is often present, particularly in adults, but the invasive infections due to S. pyogenes reported since the 1990s have characteristically affected young, previously healthy individuals. Necrotizing fasciitis begins with acute onset of local swelling, erythema, tenderness, and heat. Fever is usually present, and pain is out of proportion to cutaneous signs.29–31 Lymphangitis and lymphadenitis are usually absent. The infection advances along the superficial fascial plane, and initially, there are few cutaneous signs to herald the serious nature and extent of subcutaneous tissue necrosis that is occurring. Skin changes may occur over 24 to 48 hours as thrombosis of nutrient vessels and resultant cutaneous ischemia develops. Cutaneous signs include formation of bullae filled initially with straw-colored fluid and later with bluish to hemorrhagic fluid, and darkening of affected tissues from red to purple to blue. Anesthesia of skin and, finally, frank tissue gangrene and slough develop as a result of the ischemia and necrosis. Children with varicella and invasive S. pyogenes may initially show no cutaneous signs of superinfection, such as erythema or swelling.29 Systemic toxicity is significant and can advance rapidly to shock, organ failure, and death within hours. In most cases of toxic shock-like syndrome due to S. pyogenes, necrotizing fasciitis has

464

SECTION

J

Skin and Soft-Tissue Infections

been present.50 The combined mortality among children and adults with necrotizing fasciitis due to S. pyogenes has been approximately 30% to 60%2; both patients without shock and children have a better prognosis.29 Fournier gangrene is a form of necrotizing fasciitis that occurs in the male genital area, sometimes confined to the scrotum, but involving the perineum, penis, or abdominal wall in other cases. It is caused by a synergistic infection or, occasionally, by spread of streptococcal balanitis. Histopathologic examination shows that necrotizing fasciitis consists of necrosis and suppuration of the superficial fascia; edema and an acute inflammatory infiltrate in the deep dermis, subcutaneous fat, and fascia; microorganisms present within destroyed tissue; and thrombosis of arteries and veins at all levels of tissue.34 Some thrombosed vessels are surrounded by an inflammatory infiltrate or microorganisms or both, others may show vasculitis, and still others may demonstrate noninflammatory intravascular coagulation. Early in infections due to S. pyogenes or Clostridium spp., inflammatory cells can be lacking at sites of tissue damage.28 Definitive diagnosis is made from surgical exploration, which must be undertaken as soon as the diagnosis is suspected. Necrotic fascia and subcutaneous tissue are gray and offer little resistance to blunt probing. Although magnetic resonance imaging can aid delineation of the extent and the tissue planes of involvement, this procedure should not delay surgical intervention.4 Frozen-section biopsy specimens taken early in the course of the infection can aid management by decreasing the time to diagnosis and helping establish margins of involvement. Gram stain of tissue usually shows a mixture of organisms, but it can indicate solo infection with Staphylococcus aureus or Streptococcus pyogenes. Devitalized tissue must be removed; a second exploration is generally performed within 24 to 36 hours to confirm that no necrotic tissue remains; surgical exploration may be required on several occasions, until devitalized tissue has ceased to form. Daily, meticulous wound care is also paramount. The testes can generally be saved in Fournier gangrene, because their blood supply is separate from that of the adjacent fascia and skin. Parenteral antibiotic therapy must be initiated as soon as possible, with coverage of all potential pathogens. Empiric therapy usually consists of a beta-lactam agent, plus clindamycin, and an aminoglycoside for coverage of S. pyogenes and potential anaerobic and gram-negative pathogens.28,33 Vancomycin is included where MRSA is prevalent. Because of the presence of large numbers of organisms (“inoculum effect”) in S. pyogenes infections and their rapid entry into a slower or stationary phase of growth, S. pyogenes can become less susceptible to the b-lactam antibiotics, such as penicillins and cephalosporins, which act by interrupting cell wall synthesis.89,90 The “inoculum effect” may be due to alterations in penicillin-binding proteins (decreasing the target site for the b-lactam antibiotics).91 Clindamycin, an inhibitor of protein synthesis, is not adversely altered by the inoculum effect; clinical data suggest that this agent has a relative benefit over cell wall-active antibiotics for the treatment of deep-seated, serious soft-tissue infections due to S. pyogenes.90,92 Additional theoretic reasons for the superiority of clindamycin are its suppression of bacterial toxin synthesis and monocyte synthesis of TNF-a and its facilitation of phagocytosis through inhibition of M-protein synthesis.93–95

ACKNOWLEDGMENT The authors acknowledge significant use of the work of G.L. Darmstadt, who wrote this chapter in the second edition.

CHAPTER

79

Myositis, Pyomyositis, and Necrotizing Fasciitis Anna Norrby-Teglund and Donald Edward Low

Myositis, pyomyositis, and necrotizing fasciitis are rare conditions in children. Myositis is an inflammation of the skeletal muscles, often caused by infection or autoimmune disease. Pyomyositis is an infection of the skeletal muscle that is usually caused by Staphylococcus aureus. Necrotizing fasciitis is a rapidly spreading soft-tissue infection affecting the subcutaneous tissue and underlying fascial layers. The list of possible etiologic agents is long, and an approach to diagnosis and management can be facilitated by categorizing the diseases as acute myositis, chronic myositis, pyomyositis, and necrotizing fasciitis (Table 79-1).

TRANSIENT ACUTE MYOSITIS Clinical Manifestations The classic form of transient acute myositis is often termed benign acute childhood myositis.1 Abrupt onset of bilateral calf pain usually develops in a school-aged child recovering from an upper respiratory tract illness. The gastrocnemius or soleus muscles are typically tender, and the child may toe-walk or walk with a wide-based, stiff-legged gait. Swelling, overlying erythema, or warmth is rarely noted. Fever and systemic signs are minimal or absent. Occasionally, muscles of the thigh, neck, or upper extremity are also involved. Usually, serum creatine kinase values are markedly elevated but subside along with the muscle pain within 3 to 5 days. No muscular sequela occur.2 In several case reports, transient acute myositis has been associated with more extensive rhabdomyolysis and consequent myoglobinuria.3,4 Affected patients tend to have generalized distribution of muscle involvement. Urine is dark and yields a positive result on dipstick for blood, but microscopic examination reveals no red blood cells. Signs and symptoms of renal insufficiency can ensue. Epidemic pleurodynia is a unique clinical syndrome frequently termed Bornholm disease, after an epidemic that occurred in 1930 on the Baltic island of Bornholm.5,6 Peak incidence is in children of school age, and typically follows mild upper respiratory or gastrointestinal tract illness. Sudden onset of severe, stabbing pain in the chest wall and upper abdominal musculature is the outstanding feature. The pain comes in spasms (“devil’s grip”) lasting a few minutes to several hours and is aggravated by deep inspiration or coughing. Fever is usually present and peaks with exacerbations of pain. Tenderness to palpation without muscle swelling is characteristic. A pleural friction rub may be appreciated, but chest radiography is typically normal. Symptoms resolve gradually over 3 to 5 days.

Etiologic Agents and Pathogenesis Transient acute myositis has been associated with a large number of viral agents, in particular influenza B virus. Other reports have described the syndrome with infection due to influenza A viruses, parainfluenza viruses, various enteroviruses, herpes family viruses, rotavirus,7 and Mycoplasma pneumoniae.8 Microscopic examination

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Myositis, Pyomyositis, and Necrotizing Fasciitis

CHAPTER

79

465

TABLE 79-1. Infections of Skeletal Muscle in Children Disease Category

Syndrome

Infectious Agent(s)

Clinical Features

Influenza A or B viruses, parainfluenza viruses, enteroviruses, herpes group viruses, rotavirus, Mycoplasma pneumoniae

URI or diarrheal prodrome with abrupt-onset bilateral calf pain Overlying inflammation minimal Elevated serum creatine kinase level Self-limited, rare cases of myoglobinuria

Epidemic pleurodynia

Coxsackie B viruses

Fever, URI prodrome Paroxysms of pleuritic chest pain (“devil’s grip”) Tender intercostal and upper abdominal muscles Resolves spontaneously

Trichinellosis

Trichinella spiralis

Ingestion of undercooked meat from infected animal Fever, myalgia, leukocytosis with eosinophilia

Cysticercosis

Taenia solium

Ingestion of eggs from pork tapeworm Typically asymptomatic unless extraocular muscles involved

Toxoplasma myositis

Toxoplasma gondii

Rare cases of progressive polymyositis syndrome in immunosuppressed patients

Lyme myositis

Borrelia burgdorferi

Usually second stage of disease Localized near areas of arthritis, neuropathy, or skin lesions

AIDS-related myositis

HIV

Chronic myopathy with myalgia, weakness, and atrophy

Juvenile dermatomyositis

Questionable links to viral and bacterial infections, listed in Box 79-1

Chronic inflammatory inÀltrate, symptomatic weakness Heliotrope rash and Gottron papules

Abscess formation

Staphylococcus aureus (95%), also streptococci, Serratia spp., Klebsiella spp., Yersinia spp., Pasteurella spp.

Endemic in tropics (“tropical pyomyositis”), sporadic elsewhere Large muscles of limbs or trunk Fever, pain, swelling, muscle spasm

Group A streptococcus, rarely Staphylococcus aureus, anaerobic streptococci or a polymicrobial synergistic infection

Varicella a predisposing factor Severe pain, fever, confusion Fulminant infection with associated toxic shock syndrome

Clostridium perfringens, less commonly Clostridium septicum

Complication of contaminated wound or intestinal tract disease Rapid muscle degeneration, crepitus secondary to gas formation High fatality rate

MYOSITIS

Acute

Chronic

PYOMYOSITIS

NECROTIZING FASCIITIS

Clostridial myonecrosis

AIDS, acquired immunodeÀciency syndrome; HIV, human immunodeÀciency virus; URI, upper respiratory tract infection.

of biopsy samples usually reveals scattered areas of muscle necrosis with sparse mononuclear inÀltrate. Severe rhabdomyolysis and myoglobinuria are rare and are seen predominantly following influenza virus infection. The exact pathogenesis of transient acute myositis is unclear. The appearance of muscular pain as other symptomatology wanes, together with the diversity of associated agents, favors a postinfectious autoimmune mechanism. However, in some case reports, myxovirus-like particles were visualized in,9 or influenza A virus was isolated from, muscle biopsy specimens.10 Epidemic pleurodynia is a manifestation of acute coxsackievirus B infection and is occasionally seen in association with myopericarditis or aseptic meningitis. Most epidemics of pleurodynia occur in the late summer months, and coxsackievirus is often isolated from the stool. An outbreak among members of a high-school football team was traced to contamination of a common water container contaminated by an infected player.11 Although coxsackieviruses are myotrophic in

humans, data demonstrating direct viral invasion in pleurodynia are lacking. Genetic factors are likely to play a role in pathogenesis. In a large coxsackievirus B3 epidemic in South Africa, white children predominantly developed pleurodynia, whereas myocarditis was the major clinical manifestation in the black population.

Treatment No speciÀc treatment is required for most forms of transient acute myositis, because the disease is benign and self-limited. Rest and analgesic therapy beneÀt most patients. When myoglobinuria is present, renal function should be monitored. Rarely, patients experience acute tubular necrosis requiring dialysis. The use of antiviral therapy in severe influenza-associated myositis has not been evaluated.

466

SECTION

J

Skin and Soft-Tissue Infections

CHRONIC INFLAMMATORY MYOSITIS A number of parasitic and other agents can produce a chronic inflammatory reaction in skeletal muscle. In these disorders, persistence of the infectious agent appears necessary to cause chronic myopathy, either by direct tissue damage or as a consequence of the normal immunologic response directed toward the infected tissue. More controversial is whether certain acute infections, through molecular mimicry or elicitation of an anti-idiotypic response, may be responsible for chronic autoimmune myopathies such as juvenile dermatomyositis (JDM) and polymyositis.12

worm develops in the gastrointestinal mucosa and releases larvae that circulate via lymphatics and the bloodstream to striated muscle, where they become encysted and survive for several years. The vast majority of infections are subclinical; autopsies reveal Trichinella cysts in > 4% of randomly selected diaphragms.32 Typical manifestations of symptomatic disease include fever, extreme malaise, muscle pain, weakness, and periorbital edema.33 Serum muscle enzymes are elevated, and eosinophilia is present. Serodiagnostic testing is available and may preclude the need for diagnostic biopsy. Most mild cases respond to analgesic therapy. In severe cases, anthelmintic therapy with thiabendazole or mebendazole is effective.34

Juvenile Dermatomyositis JDM is the most common form of pediatric idiopathic inflammatory myopathy, with incidence ranging from 1.9 to 3.2 per million children.13,14 The two major clinical features of JDM are a characteristic rash and symmetric proximal muscle weakness.15 The gastrointestinal tract and lung are also often involved.15 JDM is clinically distinct from the adult form of disease, the former demonstrating more vasculitis features, more frequent skin ulceration and calcinosis.16 Patients with JDM often experience a flu-like illness approximately 3 months prior to the onset of the disease. Coxsackievirus B and other enteroviruses are the most frequent infectious agents temporally associated with the clinical onset of JDM (Box 79-1).17–20 Coxsackievirus B1 can induce chronic myositis of the proximal hindlimbs in a murine model.21 Serologic responses consistent with acute coxsackievirus B infection were found in some studies17 but not others.22 Although picornavirus-like particles have been observed on electronmicroscopic examination of muscle biopsy samples from some children with JDM, immunofluorescence, polymerase chain reaction,23 in situ nucleotide hybridization,24 and tissue culture19 have failed to identify virus. If an infectious agent precipitates JDM, it is hypothesized that autoimmune injury, rather than direct infection, is the mechanism of injury. This pathogenic mechanism is supported by gene expression profile analyses of untreated JDM muscle biopsies, demonstrating an intense interferon a/b-induced response typical of that seen during an immune response to a viral antigen.25 Several reports have demonstrated an association between the human leukocyte antigen (HLA)-DQA1*0501 and JDM.25–27 Studies indicate that this association was linked to an enhanced cytokine response partly contributed to by the tumor necrosis factor-alpha (TNF-a)-308A allele,25,28 which is known to promote an enhanced TNF-a synthesis. An enhanced TNF-a production was demonstrable in both circulation28 and in muscle biopsies29 from untreated JDM patients positive for the TNF-a-308A allele. Furthermore, this allele has been shown to be associated with pathologic calcifications and disease chronicity of JDM.28,30 In the case of group A streptococcal infection preceding the onset of JDM, the pathogenesis of JDM has been suggested to involve cytotoxic and cytokine responses elicited by specific epitopes of the streptococcal M-protein homologues to the human skeletal myosin.31

Trichinellosis Infection with Trichinella spiralis occurs after ingestion of larval cysts in undercooked pork or the meat of certain wild carnivores. The adult BOX 79-1. Infectious Agents Proposed to be Associated with the Clinical Onset of Dermatomyositis Coxsackievirus B17,18 Enterovirus19,20 Parvovirus B19166 Hepatitis B and C167,168 Toxoplasma169 Borrelia burgdorferi170 Group A streptococcus171 Staphylococcus aureus172

Cysticercosis After ingestion of eggs of the pork tapeworm Taenia solium, larval spread leads to cyst formation within skeletal muscle with the formation of calcific densities and granulomatous nodules.35 Although fever, myalgia, and eosinophilia have been reported, most cases of muscular cysticercosis are asymptomatic. An exception arises when extraocular muscles are involved, and treatment with albendazole plus prednisolone or surgical removal of the encysted parasites may be necessary to preserve vision.36–38 Patients discovered to have musculoskeletal cysticercosis should be evaluated for central nervous system involvement.

Toxoplasmosis During acute Toxoplasma gondii infection, widespread dissemination of tachyzoites can occur, and skeletal muscle is frequently involved. The vast majority of cases of acquired toxoplasmosis are asymptomatic, although some patients experience a mononucleosis-like syndrome with nonspecific myalgia. Chronic inflammatory myositis associated with T. gondii infection has been reported in human immunodeficiency virus (HIV-1)-infected and other immunocompromised hosts.39 Weakness, muscle wasting, and high serum levels of creatine kinase are characteristic features. The acute phase may be responsive to pyrimethamine plus sulfadiazine, whereas the chronic phase is associated with an altered immune response requiring corticosteroid therapy.40

Lyme Disease Chronic myositis, an uncommon manifestation of Lyme borreliosis, occurs as a localized process in proximity to involved joints or regional neuropathy.41,42 Spirochetes are apparent on biopsy, along with mononuclear cell infiltrates and limited fiber necrosis.43 Muscle inflammation typically abates after oral or intravenous antibiotic therapy.

HIV-Related Myositis A variety of inflammatory myopathies resembling polymyositis can be observed in the course of HIV-1 infection. Generalized myalgia, proximal muscle weakness, and elevated serum creatine kinase are frequent features.44 Muscle biopsy shows lymphocytic infiltration and necrosis of muscle fibers, allowing distinction from the mitochondrial myopathy related to extended zidovudine therapy.45 Although one study has shown that proviral DNA can be detected in myocytes and muscle macrophages by polymerase chain reaction,46 other researchers have concluded that the disease probably represents a T-lymphocyte-mediated autoimmune process.47 Corticosteroid therapy may be beneficial in HIV-associated polymyositis.48

PYOMYOSITIS Pyomyositis is a bacterial infection of skeletal muscle. It has a predilection for large-muscle groups and often results in localized abscess formation. It is related to tropical myositis or pyomyositis PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Myositis, Pyomyositis, and Necrotizing Fasciitis

CHAPTER

79

467

tropicans because most cases have been reported in patients living in tropical areas, where pyomyositis accounts for 3% to 5% of hospital admissions.49–51 Levin et al. in 1971 reported the first case from a temperate region.52 Since then many cases have been reported from various parts of the world.53–55 This increase in incidence is attributed to heightened awareness, a greater number of immunocompromised patients, and improvement in diagnostic techniques. Crum56 summarized all reported cases of bacterial myositis in the United States from 1981 to 2002: common underlying disorders included diabetes mellitus, malignancy, a rheumatologic condition, and HIV infection. Within North America, the highest incidence of pyomyositis is in the southernmost regions.57 In children, the peak incidence is between 5 and 9 years of age.58

Clinical Manifestations and Etiologic Agents Pyomyositis occurs most often in children and young adults,59 exhibits a 2:1 male preponderance, and has been reported in the neonatal period.60 Methicillin-susceptible Staphylococcus aureus (MSSA) is isolated from the purulent material in approximately 90% of cases in tropical areas and 75% in cases from temperate countries.53 Group A streptococcus accounts for 1% to 5% of cases. Other bacterial species isolated from pyogenic muscle abscesses are streptococci (group B, C, and G), Escherichia coli, Citrobacter freundii, Serratia marcescens, Yersinia enterocolitica, Klebsiella spp., and Salmonella spp. Individuals infected with HIV-1 are at increased risk for development of bacterial pyomyositis, sometimes with multifocal involvement.61 This condition should be considered in patients with muscle pain, fever, leukocytosis, elevated erythrocyte sedimentation rate (ESR) and C-reactive protein. Magnetic resonance imaging (MRI) is the imaging technique of choice during the first stage of disease. Staphylococcal (“tropical”) pyomyositis most frequently affects the quadriceps, hamstring, or gluteal muscles but can also affect the paraspinous, shoulder girdle, psoas, and other muscles.59 Symptoms generally appear insidiously, with low-grade fever, muscle aches, and cramping evolving over several days. Multiple abscesses are present in about 25% of patients. In the early stages, examination may reveal only a hard, rubbery firmness to the muscle belly, with no other signs of inflammation. Within 1 to 3 weeks, boggy swelling, erythema, tenderness, and warmth appear, and the lesion becomes fluctuant. Leukocytosis with a left shift and elevated ESR are found. Although substantial muscle destruction can develop with delayed treatment, serum levels of muscle enzymes are generally normal. Tropical pyomyositis is occasionally complicated by metastatic disease such as empyema, pericarditis, or lung abscess. In rare cases, fulminant septicemia or toxic shock syndrome can occur.62 Pyogenic abscess in the psoas muscle produces a distinct clinical syndrome with lower abdominal or back pain radiating to the hip.63 The febrile child may limp or hold the hip in fixed flexion because of muscle spasm. Confusion with pyogenic arthritis is common. Pain on hyperextension or abduction of the hip is elicited on examination. Staphylococcus aureus remains the most common cause, but psoas abscess can occasionally develop as an extension of an abdominal process such as a ruptured appendix.63 In such cases, a mixed infection with anaerobic and facultative bowel flora is likely. With the resurgence in invasive disease due to group A streptococci since the late 1980s, these organisms have been recognized as an increasingly important cause of pyomyositis.64,65 Group A streptococcal infection of skeletal muscle can present as a localized phlegmon, an abscess, or more fulminant necrotizing myositis.66,67 Necrotizing group A streptococcal myositis is rapidly progressive, evolving over several hours (Figure 79-1). Intense pain is the most common presenting symptom, often out of proportion to clinical signs of inflammation. The child may refuse to bear weight or to move an extremity. Ultimately, high fever, localized swelling, and overlying erythema are seen. Tachycardia, hypotension, oliguria, confusion, lethargy, and scarlatiniform eruptions are early signs of associated streptococcal toxic shock syndrome. Mild to moderate leukocytosis with bandemia (often more than 50%) and

Figure 79-1. Necrotizing pyomyositis due to Streptococcus pyogenes in a 6-month-old infant with acute onset of fever, erythema, swelling, and pain of her right shoulder, and decreased use of her arm. There was no break in the skin. Coronal fluid-attenuated inversion recovery (FLAIR) magnetic resonance imaging shows diffuse hyperintensity of the deltoid muscle with associated subcutaneous inflammatory stranding and edema. Osseous and fascial structures are normal. (Courtesy of E.N. Faerber and S.S. Long, St. Christopher’s Hospital for Children, Philadelphia, PA.)

elevations of blood urea nitrogen, creatinine, and creatine kinase are common. Blood culture is positive in most cases. In addition to the extensive muscle destruction, renal failure, myocardial dysfunction, and adult respiratory distress syndrome may develop. Mortality in group A streptococcal necrotizing myositis has been as high as 80% in some series but is probably diminishing with the greater awareness and earlier institution of therapy.65,68 Although a portal of entry is not always evident, the exanthem of primary varicella and various forms of minor skin trauma are important predisposing factors to the development of group A streptococcal deep-tissue infection in children. A prospective population-based active surveillance for pediatric invasive group A streptococcal disease in Ontario revealed that varicella-zoster virus infection is associated with a 58-fold increased risk of acquiring invasive group A streptococcal disease in children.64 Although the attack rate of invasive group A streptococcal infections was relatively low (5.2/100 000), 15% of all pediatric invasive group A streptococcal infections, including 50% of necrotizing fasciitis cases, followed varicella infection. Gas gangrene is a fulminant, often fatal form of myonecrosis caused by clostridial infection. Classically, this disorder arises in muscle tissue damaged by trauma or surgical wounds that become contaminated with spores of Clostridium perfringens from the soil or feces. The symptoms typically appear 2 to 3 days after contamination and are heralded by onset of severe and persistent pain at the site of the wound. Inflammation is initially absent but infection progresses

468

SECTION

J

Skin and Soft-Tissue Infections

within hours to produce tenderness, pallor, and tense edema. Subcutaneous emphysema and crepitation, secondary to gas production by the organism, is common. Darkening of the skin with development of hemorrhagic bullae and cutaneous necrosis ensues, and the lesion may release a serosanguineous discharge in which an abundance of boxcar-shaped gram-positive rods and a striking absence of neutrophils are found. Muscle disintegration progresses rapidly over hours, along with shock, coagulopathy, and acidosis. Mortality is high if intervention is delayed. Gas gangrene can also develop in the absence of an external wound. In such cases, Clostridium septicum is the most common isolate, and an intestinal tract abnormality (e.g., necrotizing enterocolitis) or immunodeficiency (e.g., granulocytopenia) is usually present.69,70 Less frequent causes of necrotic pyomyositis include anaerobic streptococci and Aeromonas hydrophila.

Pathogenesis Skeletal muscle tissue is intrinsically resistant to bacterial infections, likely due to sequestration by myoglobin of iron that is required for proliferating bacteria. Staphylococcal muscle abscesses appear to be a complication of transient bacteremia and typically develop without penetrating injury or other clear portal of entry. Blood cultures are usually negative.71 Muscular trauma, strain, or vigorous exercise may be predisposing factors.57,72 The high prevalence of the disease in the tropics has led to speculation that in patients with migrating parasitic infections, such as toxocariasis, microscopic foci of necrotic muscle develop that are predisposed to bacterial seeding.73 Alternatively, a viral infection may be the precipitant; ultrastructural studies of nonsuppurating lesions in some patients with tropical pyomyositis reveal intracellular particles and a lymphocytic infiltrate.74 Group A streptococcal infection of skeletal muscle can take the form of a localized phlegmon, an abscess, or more fulminant necrotizing myositis,66,67 associated with septicemia and toxic shock syndrome. The association of invasive group A streptococcal disease with primary varicella infection may simply implicate the full-thickness skin lesion of chickenpox, serving as a portal of entry for the organism (see Figure 79-1).75–77 Varicella also produces a transient immunologic derangement, predisposing to secondary bacterial infection. Both Staphylococcus aureus and group A streptococci express virulence factors, including adhesins, toxins, superantigens, and immunomodulatory proteins, which contribute to the pathogenesis of infection (Tables 79-2 and 79-3). Although certain virulence factors are implicated in certain disease manifestations, such as superantigens as mediators of systemic toxicity and tissue injury,78–80 no single virulence factor is sufficient to provoke a severe staphylococcal or streptococcal infection. Rather a coordinated action of these virulence factors is required in order for the bacteria to colonize successfully and spread within the host.81–84 The virulence factors can be grouped on the basis of their mechanisms of action, and, as can be seen in Tables 79-2 and 79-3, many of the virulence factors have multiple functions. Generally, surface-associated molecules are important for evasion of phagocytosis and bacterial adherence to host cells and tissue, whereas many of the secreted proteins such as peptidases, streptolysins, superantigens, and proteases are important for spread and growth of the bacteria and result in induction of inflammatory responses.82–85 Tissue injury and systemic toxicity are largely due to excessive inflammatory responses. A direct relation between the magnitude of the inflammatory response and severity of invasive group A streptococcal infections has been demonstrated.80,86,87 Although streptococci and staphylococci have many molecules with proinflammatory activities, including the gram-positive cell wall components peptidoglycan and lipoteichoic acid, the superantigens are the most prominent inducers of inflammatory responses.79,82,88 Superantigens are microbial toxins that induce powerful immune responses by circumventing the normal antigen processing and presentation, resulting in excessive activation and release of inflammatory mediators.88 The best-characterized superantigens are those produced by Staphylococcus aureus and group A streptococci.79,89 The classical

TABLE 79-2. Virulence Factors of Staphylococcus aureus and their Mechanisms of Action Mechanism of Action

Virulence Factor

Reference

Adherence/colonization

Fibronectin-binding protein 173 A and B Fibrinogen-binding protein (Efb) Clumping factor A and B Collagen-binding protein Elastin-binding protien Extracellular adherence protein (EAP)

Inhibition of neutrophil chemotaxis

Chemotaxis inhibitory protein (CHIPS) EAP

174 175

Resistance to phagocytosis Capsule Protein A Clumping factor A

176

Inactivation of complement

Efb Staphylococcal complement inhibitor Staphylokinase

176

Resistance to antimicrobial peptides

DltABCD MprF proteins Aureolysin

176

Killing of host leukocytes

a-haemolysin g-hemolysin Panton–Valentine leukocidin Leukocidin E/D Leukocidin M/F-PV-like

177 176

Proinflammatory activities

Superantigens: staphylococcal enterotoxins, TSST-1 Toll-like receptor ligands: Peptidoglycan Lipoteichoic acid CpG DNA TNFR1 ligand: protein A

89 81

178

TSST-1, toxic shock syndrome toxin-1; TNFR1, tumor necrosis factor-alpha receptor 1.

superantigens include the staphylococcal enterotoxins, toxic shock syndrome toxin-1, and the streptococcal pyrogenic exotoxins. Streptococcal and staphylococcal isolates do not harbor the genes for all superantigens, but rather the strains have a certain repertoire of superantigens. Another mechanism that likely contributes to the systemic toxicity and development of shock is the ability of M-protein to form complexes with fibrinogen, which activate neutrophils to release massive amounts of heparin-binding protein and consequently induce increased vascular leakage.90 Clostridium perfringens elaborates at least two exotoxins (a-toxin and perfringolysin O) that are cytolytic to host tissues and are lethal when purified and injected into animals.91 C. septicum also expresses a pore-forming cytolysin, a-toxin, which triggers fulminant myonecrosis as well as inhibition of leukocyte influx into the lesion.92 Exotoxin-induced microvascular dysfunction is an important factor producing the anaerobic environment that favors C. perfringens replication and ischemic necrosis.22 By contrast, C. septicum is relatively aerotolerant, a feature that may partially explain its ability to spread through the bloodstream and establish infection in otherwise healthy muscle.69

Diagnosis Early in the course and depending on location, pyomyositis or pyogenic muscle abscess can be difficult to distinguish from a muscle

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Myositis, Pyomyositis, and Necrotizing Fasciitis

TABLE 79-3. Virulence Factors of Group A Streptococcus and their Mechanisms of Action Mechanism of Action

Virulence Factor

Reference

Adherence/colonization

Capsule M protein Protein F/Sfb1 Fibronectin-binding protein Glyceraldehyde-3-phosphate dehydrogenase Fibronectin-binding protein Vitronectin-binding protein Collagen-binding proteins

82

Inhibition of neutrophil chemotaxis

Cell envelope proteinase (SpyCEP)

179,180

Resistance to phagocytosis

Capsule M protein M-like proteins

82

Inactivation of complement

M protein M-like proteins Streptococcal inhibitor of complement (SIC) Streptococcal cystein protease (SpeB)

181–184

Resistance to antimicrobial peptides

SIC GRAB/SpeB

185,186

Dissemination of infection

Spreading factors (DNAses, hyaluronidase) Plasminogen-binding proteins Streptokinase

82,184,187

Killing of host leukocytes

Streptolysins O and S

188

Inhibition of proteolysis

GRAB a2-macroglobulin-binding proteins

189,190

Induction of vascular leakage

M-protein

90

Proinflammatory activities

Superantigens: Spe, SmeZ, SSA Toll-like receptor ligands: Peptidoglycan Lipoteichoic acid CpG DNA

78,79,191

SmeZ, streptococcal mitogenic exotoxin Z; Spe, streptococcal pyrogenic exotoxins; SSA, streptococcal superantigen; GRAB, protein G-related alpha 2M-binding protein.

hematoma or tumor, pyogenic arthritis, osteomyelitis, or appendicitis. Plain radiographs are usually normal, but other imaging techniques can be useful in defining the extent of muscle disease. Computed tomography (CT) or ultrasonography may delineate a low-density (or hypoechoic) fluid collection, thereby facilitating diagnostic aspiration or percutaneous drainage.93,94 Diagnostic MRI findings include hyperintense muscle mass with edema in pyomyositis, and a hyperintense rim on unenhanced T1-weighted images and peripheral enhancement after intravenous infusion of gadolinium if abscess has formed95 (see Figure 79-1). Gram stain of purulent fluid from muscle abscesses and subsequent culture identification are crucial to guide the nature and duration of antibiotic therapy. Computed tomography or MRI can also detect marked skeletal muscle abnormalities in group A streptococcal necrotizing myositis (Figure 79-2).94,96 However, group A streptococcal myositis can rapidly progress to a severe, life-threatening systemic illness, and definitive diagnosis of the nature and extent of such necrotizing myositis or fasciitis is only made with surgical exploration.97 Infrared thermography may help reveal the extent of tissue viability in clostridial myonecrosis.98

CHAPTER

79

469

Treatment Effective surgical drainage of all pyogenic muscle abscesses is paramount. This can often be accomplished percutaneously with ultrasound or CT guidance,99 but an open surgical procedure may be required. Initial antibiotic therapy should target Staphylococcus aureus; a b-lactamase-resistant penicillin or clindamycin is an appropriate agent. Definitive antibiotic therapy will be guided by results of Gram stain, culture, and susceptibility testing of material obtained during surgery. Response to therapy is assessed from serial examinations, resolution of fever, normalization of ESR, and follow-up imaging studies. Persistent fever and inflammatory signs suggest undrained pyogenic foci of infection. In necrotizing group A streptococcal pyomyositis, surgical debridement will be necessary once the patient has been stabilized medically.100 Fasciotomies are performed if there is evidence of increased compartment pressures (> 20 mmHg is abnormal; > 45 mmHg indicates substantial compromise of perfusion). Group A streptococci are uniformly susceptible to penicillin and other betalactam antibiotics, and penicillin remains an appropriate antibiotic treatment of group A streptococcal infections. However, in patients with aggressive group A streptococcal infections, such as myositis and necrotizing fasciitis, a beta-lactam antibiotic should be combined with clindamycin (40 mg/kg per day). In a murine model of group A streptococcal myositis, Stevens et al.101 demonstrated a lower mortality with clindamycin compared with penicillin. One explanation for this finding is that, as group A streptococci reach a high density, penicillin loses effectiveness, probably due to the “Eagle effect.”102 The theoretic basis of this phenomenon is that group A streptococci initially proliferate rapidly until a steady state of growth is reached, after which the bacteria are relatively metabolically inative. b-lactam antibiotics are less effective during this steady-state phase, because they inhibit cell wall synthesis of actively replicating organisms. Clindamycin, however, is not affected by the decreased growth rate, because it works at the ribosomal level. Clindamycin offers the added benefit of inhibiting toxin synthesis, including superantigens and M-protein, by the organism.101,103,104 Epidemiological studies of patients with group A streptococcal deep-tissue infections have observed that patients are more likely to have a favorable outcome if the initial antibiotic regimen includes a combination of a protein synthesis-inhibiting antibiotic and a cell wall-inhibiting antibiotic compared with only a cell wall-inhibiting agent.105,106 Therapy of clostridial gas gangrene consists of prompt and radical debridement of involved muscles; amputation may be required. Combination antibiotic therapy with penicillin G and clindamycin is employed most commonly. Hyperbaric oxygen therapy may serve an adjunctive role by retarding growth of C. perfringens, inhibiting a-toxin production, and increasing oxidative killing by host neutrophils.107

COMMUNITY-ACQUIRED METHICILLIN-RESISTANT STAPHYLOCOCCUS AUREUS SKIN AND SOFT-TISSUE INFECTIONS Methicillin resistance first appeared among nosocomial isolates of S. aureus in 1961.108 It is now one of the most common causes of bacterial nosocomial infections, accounting for 40% to 70% of the S. aureus infections in intensive care units.109,110 In the past, acquisition of methicillin-resistant S. aureus (MRSA) colonization and infection was generally considered to be restricted to the nosocomial setting. However, in the 1980s, MRSA infections were reported in persons who lacked traditional MRSA risk factors.111 These infections appeared to be acquired in the community and are now known as community-associated (CA) MRSA infections. Outbreaks have occurred in many settings and among different populations worldwide and especially in children.112–119

470

SECTION

J

Skin and Soft-Tissue Infections

A

B

Figure 79-2. A 5-year-old girl experienced severe left posterior calf pain, tenderness, and swelling 1 week after onset of primary varicella. (A) With magnetic resonance imaging technique in which fat signal intensity is suppressed, marked signal enhancement is seen diffusely in the calf musculature, indicative of a widespread inflammatory process. (B) Operative exploration of the calf revealed liquefaction necrosis of the soleus muscle and the lateral head of the gastrocnemius muscle, which were radically debrided. Group A streptococcus was isolated from culture of tissue and blood. (Courtesy of J.H.T. Waldhausen, M.D.)

Clinical Features CA-MRSA are most frequently associated with skin and soft-tissue infections, especially furunculosis, both in isolated cases and in epidemics.120 Sattler et al. found 11% of CA-MRSA infections were deep-seated as compared with 30% of CA-MSSA infections.121 Frank et al. found that 70% of CA-MRSA infections involved skin, wounds, or abscess formation, with only 16% deemed serious invasive infections.122 In a retrospective study of all age groups at several sites in Minnesota, 84% of CA-MRSA infections involved the skin.123 CA-MSSA infections were initially reported to be more severe than CA-MRSA infections,121 but subsequent reports have observed the opposite.124,125 Of 10 reported cases of CA-MRSA infections in children and adults with pneumonia, 7 had evidence of necrotizing pneumonia, and 6 patients died.126–131 In a retrospective review of hospitalized children with S. aureus infections, Ochoa et al.132 found patients with CA-MRSA infections were significantly younger and more likely to be African American than patients with CA-MSSA infections. Patients with CA-MRSA tended to have longer duration of bacteremia and required more surgical interventions (incision, aspiration, drainage, or debridement). In a retrospective study of outcome of children with musculoskeletal infections admitted to hospital, Martinez-Aguilar et al.133 found that duration of fever and hospitalization were great in children with CA-MRSA compared with CA-MSSA infection.

Genetics of CA-MRSA Methicillin resistance results from the production of an altered penicillin-binding protein known as PBP2a, which has decreased affinity for most b-lactam antibiotics. PBP2a is encoded by the gene mecA, which is carried on a mobile genetic element staphylococcal cassette chromosome type IV in CA-MRSA strains.134 CA-MRSA have a high frequency of virulence factor expression, particularly of Panton–Valentine leukocidin (PVL).135–137 PVL is cytotoxic to monocytes, macrophages, and polymorphonuclear leukocytes. PVL appears to be one of the key virulence determinants of CA-MRSA strains. Lina and colleagues135 found that 50% to 93% of the S. aureus strains causing primary skin infections produce PVL. Furthermore, a

study by Gillet et al.128 demonstrated a strong association between PVL and necrotizing pneumonia in healthy children and young adults. Said-Salim et al.136 analyzed the distribution and expression of PVL among different CA-MRSA strains. They evaluated the lysis of human polymorphonuclear leukocytes during phagocytic interaction with PVL-positive and PVL-negative CA-MRSA strains. Unexpectedly, there was no correlation between PVL expression and polymorphonuclear leukocyte lysis, suggesting that additional virulence factors underlie leukotoxicity and, thus, the pathogenesis of CA-MRSA.

Epidemiology The evidence to date suggests that the acquisition of the SCCmec IV element by MSSA strains in the community has given rise to most of the emerging CA-MRSA strains.138,139 In the United States the CAMRSA strains belong to just two of the eight pulsed-field types of MRSA-USA300 and USA400.140 The USA300 clone was not seen before 2000. It has caused numerous community outbreaks of MRSA skin infections in several distinct populations, including inmates in Los Angeles and San Francisco jails, men who have sex with men in California, professional football players in St. Louis, and prison inmates in Mississippi, Georgia, and Texas.141 Moran et al. monitored the prevalence of CA-MRSA among a group of emergency department patients with skin and soft-tissue infections.142 The proportion caused by MRSA increased from 29% in 2001 to 2002 to 64% in 2003 to 2004. No clinical or historical features reliably predicted MRSA etiology. Ochoa et al.132 performed a retrospective chart review of hospitalized pediatric patients with S. aureus infections during a 3-year interval. From 2000 to 2003, CA-MRSA accounted for 67% (159/239) of community-associated S. aureus infections in hospitalized pediatric patients (56% in 2000 to 2001, 57% in 2002, and 78% in 2003, P < 0.01 for trend).

Treatment Incision and drainage is the most important therapy of skin and softtissue abscesses caused by CA-MRSA. The role of antibiotic treatment is less clear. The majority of 69 normal children with skin or

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Myositis, Pyomyositis, and Necrotizing Fasciitis

soft-tissue abscess disease caused by CA-MRSA were successfully treated with drainage, despite the fact that only 7% received an antibiotic effective against their infecting organism.143 The 4% of children who required hospitalization for additional therapy were primarily those who had a lesion of more than 5 cm at the time of presentation. There are a number of antibiotics available for the treatment of more severe infections caused by CA-MRSA (Table 79-4). In contrast to healthcare-associated MRSA, CA-MRSA strains are often susceptible to trimethoprim-sulfamethoxazole, clindamycin, doxycycline or minocycline, and fluoroquinolones, although susceptibility to these agents may vary by geographic area.144 Because the majority of reported CA-MRSA infections are skin and soft-tissue infections, clindamycin represents an attractive option. One of the major concerns with regard to the use of clindamycin for CA-MRSA infection is the possible presence of inducible resistance to clindamycin.145 Reports of cases of clindamycin failure in infections caused by isolates that have inducible clindamycin resistance primarily include patients with deep-seated infections such as osteomyelitis,146 endocarditis,147 and pneumonia with pneumatocele.148 Conversely, case reports of successful clindamycin therapy for cellulitis and wound infection, caused by S. aureus exhibiting inducible clindaymcin resistance, indicate that clindamycin therapy can be effective in such circumstances.146,149

TABLE 79-4. Antimicrobial Agents for Treatment of CommunityAcquired Methicillin-Resistant Staphylococcus aureus Infections in Children Route

Antimicrobial

Daily Dosage

Intravenous

Daptomycin1

4–6 mg/kg per day q24 hours

Linezolid

30 mg/kg per day divided q8 hours

Vancomycin

40 mg/kg per day divided q6 hours

Clindamycin

10–30 mg/kg per day divided q6–8 hours

Oral

Trimethoprim/ 20 mg/kg per day divided q6 hours sulfamethoxazole Linezolid

30 mg/kg per day divided q8 hours

Daptomycin should not be used to treat pneumonia; not approved by the Food and Drug Administration for the treatment of children.

CHAPTER

79

471

NECROTIZING FASCIITIS Necrotizing fasciitis (also known as hospital gangrene or hemolytic streptococcal gangrene) was described as early as the fifth century BC by Hippocrates.150 Necrotizing fasciitis is a rapidly progressive, deepseated bacterial infection of the subcutaneous soft tissue that may involve any area of the body. It often follows a fulminant course and has a high mortality rate, ranging from 25% to 75%.151,152 Many terms have been used to describe necrotizing soft-tissue infections. A simplified classification is provided in Table 79-5. Although more than 500 cases of necrotizing fasciitis have been reported in North America, it is an uncommon disease and the true incidence is not known.68,153 Males are affected slightly more commonly than females.151,152,154 An increased frequency is reported in persons with diabetes, intravenous drug users, alcoholics, immunosuppressed patients, and patients with peripheral vascular disease.68,152,154 Necrotizing fasciitis also occurs in young, previously healthy patients, including children. Mortality rates in children and previously healthy individuals tend to be much lower than among the elderly and those with underlying disease.68

Clinical Features Patients may report a history of recent surgery, trauma, omphalitis, or varicella infection.155,156 In children, necrotizing fasciitis often presents 1 to 4 days after trauma, with soft-tissue swelling and pain over the affected area. Patients may appear well at initial presentation. When associated with varicella, the findings typically begin 3 to 4 days after onset of the exanthem.156 Infants and toddlers may be fussy or irritable. Toddlers and young children may limp or refuse to bear weight. Initially, pain with manipulation of an affected extremity tends to be out of proportion to the cutaneous signs of infection. Induration and edema are generally apparent within the first 24 hours and are rapidly followed by blistering and bleb formation.65,157 Infection spreads in the plane between the subcutaneous tissue and the superficial muscle fascia, which results in the progressive destruction of fascia and fat. The skin takes on a dusky appearance, and a thick, foul-smelling fluid is produced (Figure 79-3). Pain and tenderness in the subcutaneous space are exquisite and often seem out of proportion to the cutaneous appearance, but destruction of the nerves that innervate the skin may eventually lead to anesthesia of the overlying skin. High fevers are common. The rapidly progressing infection can lead to toxic shock syndrome and severe systemic toxicity, including

TABLE 79-5. Necrotizing Infections of the Soft Tissues Type

Usual Etiologic Agent

Predisposing Causes

Clinical Manifestations

Meleney synergistic gangrene

Staphylococcus aureus, microaerophilic streptococci

Surgery

Slowly expanding ulceration confined to superficial fascia

Clostridial cellulites

Clostridium perfringens

Local trauma or surgery

Gas in skin, fascial sparing, little systemic toxicity

Nonclostridial anaerobic cellulitis

Mixed aerobes and anaerobes

Diabetes mellitus

Gas in tissues

Gas gangrene

Clostridial species (Clostridium perfringens, Clostridium histolyticum, or Clostridium septicum)

Trauma, crush injuries, epinephrine Myonecrosis, gas in tissues, systemic injections; spontaneous cases related toxicity, shock to cancer, neutropenia, cancer chemotherapy

Necrotizing fasciitis type 1

Mixed anaerobes, gram-negative aerobic bacilli, enterococci

Surgery, diabetes mellitus, peripheral vascular disease

Destruction of fat and fascia; skin may be spared; involvement of perineal area in Fournier gangrene

Necrotizing fasciitis type 2

Group A streptococcus

Penetrating injuries, surgical procedures, varicella, burns, minor cuts, trauma

Systemic toxicity, severe local pain, rapidly extending necrosis of subcutaneous tissues and skin; gangrene, shock, multiorgan failure

Adapted from Bisno & Stevens192 with permission from Massachusetts Medical Society.

472

SECTION

J

Skin and Soft-Tissue Infections

C A

B

Figure 79-3. Necrotizing fasciitis of the abdominal wall due to Streptococcus pyogenes as a complication of chickenpox in a 7-year-old girl. Note duskiness, ecchymosis, purpura, and edema of the abdominal wall (A). There was full-thickness necrosis of skin, subcutaneous tissue, and fascia (B). After multiple surgical debridements, the patient is ready for grafting (C). (Courtesy of J.H. Brien.)

A B Figure 79-4. Pyomyositis and necrotizing fasciitis due to methicillin-resistant Staphylococcus aureus in a 17-year-old male with a 1-day history of fever, poor appetite, and left hip pain, precluding ambulation. Physical examination showed fullness and exquisite tenderness of the buttock and posterior thigh, without a break in the skin or erythema. Axial T2-weighted magnetic resonance imaging (A, B) shows hyperintensity of the fascia surrounding the tensor fascia lata and biceps femoris muscles from midthigh to the pelvic girdle. Surgery revealed necrotizing fasciitis. (Courtesy of E.N. Faerber and S.S. Long, St. Christopher’s Hospital for Children, Philadelphia, PA.)

renal and hepatic failure, acute respiratory distress syndrome, and decreased myocardial contractility.

Etiologic Agents and Pathogenesis Necrotizing fasciitis is often polymicrobial in origin151,152,158 and involves gram-negative bacilli, enterococci, streptococci, S. aureus and anaerobes such as Bacteroides spp., Peptostreptococcus spp., and Clostridium spp. (see Table 79-5). This polymicrobial form of the disease is described as necrotizing fasciitis type 1 and is often seen postoperatively or in patients with diabetes mellitus or peripheral vascular disease. Necrotizing fasciitis type 2 is an infection due to group A streptococcus that can occur postoperatively or as a result of penetrating trauma, varicella infection, burns, or minor cuts. It is characterized by rapidly extending necrosis and severe systemic toxicity. Type 2 disease is the most common form of necrotizing fasciitis in children.68,154,159 Necrotizing fasciitis type 3 is rare and is caused by marine Vibrio spp., which enter through skin lesions that have been exposed to seawater or marine animals. Extension of the infection along fascial planes leads to necrosis of the superÀcial fascia and the deeper layers of the dermis. Destruction and thrombosis of the small blood vessels in the area lead to necrosis of the surrounding tissues. The extensive tissue damage often leads to systemic symptoms, including multiorgan failure and shock. Predisposing factors include trauma, surgery, burns, and eczema.151 In neonates, necrotizing fasciitis can complicate omphalitis or cir-

cumcision. Less commonly, precipitating factors include insect bites, perirectal abscess, incarcerated hernia, and subcutaneous insulin injection. Necrotizing fasciitis is commonly reported as a complication of varicella infection.68,75,155 Necrotizing fasciitis can also occur with a preceding group A streptococcal pharyngitis or without any previous evidence of trauma or infection. An association between the use of nonsteroidal anti-inflammatory drugs and necrotizing fasciitis has been reported.155,160,161 It is difÀcult to separate confounding by indication for more severe primary disease or early streptococcal infection versus a true cause-and-effect relationship.161

Diagnosis Although white blood cell counts can be normal or elevated, there is often a pronounced bandemia. Thrombocytopenia and evidence of a coagulopathy can also be apparent. Attempts to identify causative organisms should be made through collection of anaerobic and aerobic blood cultures.75,152 Cultures of the wound and surgically debrided tissue should also be obtained. Frozen section biopsies can be helpful in making a timely diagnosis. Imaging studies may support the diagnosis, but they should not delay surgical intervention. Plain Àlm can show gas or soft-tissue edema but otherwise is nonspeciÀcally abnormal. Although CT may be useful in deÀning the extent of soft-tissue involvement, MRI is the preferred modality. MRI can reveal extension of inflammation along the fascial plains and distinguish compartments and structures (bone, muscle, fascia, fat) involved (Figure 79-4).

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Myositis, Pyomyositis, and Necrotizing Fasciitis

Treatment Emergent, wide surgical debridement is critical and may need to be repeated immediately and frequently.151,152,158 Delays in surgery are associated with increased mortality, and antibiotic therapy in the absence of surgical debridement is ineffective.151,158 Gram stain of surgical or aspirated material from the infected site may be helpful in guiding antibiotic selection. In patients with suspected group A streptococcal infection, the antibiotic regimen should include intravenous penicillin (150 000 U/kg per day, divided every 4 to 6 hours) and clindamycin (40 mg/kg per day). In patients with polymicrobial infections, a b-lactam/b-lactamase inhibitor combination, such as ampicillin/sulbactam or piperacillin/tazobactam, should be considered. Supportive therapy includes careful fluid management, pain control, and management of multisystemic organ failure. The role of hyperbaric oxygen therapy is controversial. Patients who survive may require amputation, skin grafting, or reconstructive surgery. In severe group A streptococcal infections, the use of high-dose intravenous polyspecific immunoglobulin (IGIV) has been proposed as adjunctive therapy. The mechanism of action of IGIV in this setting is believed to include inhibition of the superantigen activity through neutralizing antibodies, opsonization through M-specific antibodies, and a general anti-inflammatory effect.83 Large controlled studies of IGIV therapy in patients with severe invasive group A streptococcal infections have not been conducted, but case reports and small controlled studies have reported use of IGIV in streptococcal toxic

CHAPTER

No. of Patientsa

NF and myonecrosis, case series identified through active surveillance

165

Case-Fatality Rate (%)

IGIV: 10 no IGIV: 4

19 25

NF, case series identified through active surveillance164

IGIV: 10 No IGIV: 67

10 37

Severe soft-tissue infections, observational cohort study100

7

0

STSS + NM, case report STSS + NF, case report194 STSS + NF, case report195 STSS + NF, case report196 STSS, case report197 STSS, case report198 STSS, case report199 STSS, case report200 STSS, case series201

1 1 1 1 1 1 1 1 5

0 0 0 0 0 0 0 0 20

STSS, multicenter placebo RCTb,163

IGIV: 10 Placebo: 11

10 36

STSS, case-control study162

IGIV: 21 No IGIV: 32

33 66

193

473

shock syndrome, necrotizing fasciitis, and necrotizing myositis (Table 79-6). The clinical efficacy of IGIV has been best studied in the management of streptococcal toxic shock syndrome. A favorable outcome was reported for invasive group A streptococcal infections in an observational cohort study of Canadian patients identified through active surveillance162 and in one European multicenter placebocontrolled trial.163 The data on the efficacy of IGIV for necrotizing fasciitis are limited (see Table 79-6).100,164,165 In one study, mortality was 10% among IGIV-treated patients, compared with 37% in nontreated control subjects.164 In an observational case study, the use of an aggressive medical regimen including high-dose IGIV appeared to mitigate the need for aggressive surgical intervention.160 Seven patients with severe soft-tissue infection caused by group A streptococci were treated with antibiotics parenterally and high-dose IGIV. Surgery was either not performed or limited to exploration. Six of the patients had toxic shock syndrome. All patients survived. The study suggested that the use of a medical regimen including IGIV in patients with severe group A streptococcus soft-tissue infections may allow an initial nonoperative or minimally invasive approach, limiting the need to perform immediate wide debridements and amputations in unstable patients. This study, although limited in numbers, suggests that an initial conservative surgical approach combined with the use of immune modulators, such as IGIV, may reduce the morbidity associated with extensive surgical exploration in hemodynamically unstable patients.

TABLE 79-6. Intravenous Polyspecific Immunoglobulin (IGIV) as Adjunctive Therapy in Severe Invasive Group A Streptococcal Infections Study Design

79

NF, necrotizing fasciitis; NM, necrotizing myositis; STSS, streptococcal toxic shock syndrome; RCT, randomized control trial. a All received IGIV unless otherwise specified. b Trial prematurely terminated due to slow patient recruitment.163

SECTION

K

Bone and Joint Infections

CHAPTER

80

Osteomyelitis Kathleen M. Gutierrez Osteomyelitis is inflammation of bone. Bacteria are the usual etiologic agents, but fungal osteomyelitis occasionally occurs. Osteomyelitis in children is primarily of hematogenous origin, occurring less commonly as a result of trauma, surgery, or infected contiguous soft tissue. Osteomyelitis due to vascular insufÀciency is rare in children.

ACUTE HEMATOGENOUS OSTEOMYELITIS Pathogenesis Acute hematogenous osteomyelitis (AHO) is primarily a disease of young children, presumably because of the rich vascular supply of their rapidly growing bones.1–3 Infecting organisms enter the bone through the nutrient artery and then travel to the metaphyseal capillary loops, where they are deposited, replicate, and initiate an inflammatory response (Figure 80-1). Metaphyseal localization results from

sluggish blood flow, the presence of endothelial gaps in the tips of growing metaphyseal vessels, and lack of phagocytic cells lining the capillaries.3–6 Bacteria proliferate, spread through vascular tunnels, and are anchored to areas of exposed cartilaginous matrix. Large colonies of bacteria surrounded by glycocalyx obstruct capillary lumens, impairing phagocytosis and antibiotic penetration.7 Age-related differences in the anatomy of the bone and its blood supply influence the clinical manifestations of osteomyelitis.2,3 Transphyseal vessels are present in most children younger than 18 months, providing a vascular connection between the metaphysis and the epiphysis.8 As a result, in infants, infection originating in the metaphysis can spread to the epiphysis and joint space. The risk of ischemic damage to the growth plate is greater in the young infant with osteomyelitis.9 Before puberty, the periosteum is not Àrmly anchored to underlying bone. Infection in the metaphysis of a bone can spread to perforate the bony cortex, causing subperiosteal elevation and extension into surrounding soft tissue. Bony destruction may spread to the diaphysis. The cartilaginous growth plate usually prevents extension of infection to the epiphysis and into the joint space. When the metaphysis of the proximal femur or humerus is involved, however, infection can extend into the hip or shoulder joint, because at these sites, the metaphysis is intracapsular. Histologic features of acute osteomyelitis include localized suppuration and abscess formation, with subsequent infarction and necrosis of bone. Segments of bone that lose blood supply and become separated from viable bone are called sequestra. An involucrum is a layer of living bone surrounding dead bone. A Brodie abscess is a subacute, well-demarcated focal infection, usually in the metaphysis but sometimes in the diaphysis of bone.

Epidemiology Approximately half of all cases of AHO occur in the Àrst 5 years of life.3,10 Boys are affected twice as frequently as girls, except in the Àrst year of life.10,11 One-third of patients have minor trauma to the affected extremity before infection, but the signiÀcance of this observation is unclear, because virtually all young children experience frequent mild trauma.12 The relative risk of developing osteomyelitis appears to be higher in some populations. Polynesian and Maori children are more likely to develop complicated osteomyelitis with Staphylococcus aureus than other children in one community in New Zealand.13 It is unclear whether genetic or socioeconomic factors account for this.

Microbiology

Figure 80-1. Gross specimen showing osteomyelitis of the proximal humerus in a 6-week old infant. Note metaphyseal location and bony destruction (arrows). (Courtesy of S.S. Long.)

474

S. aureus is the most common cause of AHO.1,12,14–20 Kingella kingae, Streptococcus pneumoniae,21 and S. pyogenes22 are the organisms isolated in most other cases of AHO in children. K. kingae and S. pneumoniae infections are most common in children less than 3 years of age. S. pneumoniae accounts for a relatively small proportion of infections; this number is decreasing further subsequent to widespread immunization with the heptavalent pneumococcal vaccine.23,23a K. Kingae is associated with small outbreaks of bone and joint infections in childcare centers.19,23 Coagulase-negative staphylococci (almost exclusively as a complication of medical

Osteomyelitis

intervention), enteric gram-negative bacilli, and anaerobic bacteria are uncommon causes of AHO. Bone disease caused by Bartonella henselae is reported.24,25,25a Actinomyces spp. cause facial and cervical osteomyelitis.26 Infection with Serratia spp. and Aspergillus spp. should be considered in children with chronic granulomatous disease.27 Before widespread use of Haemophilus influenzae type b (Hib) conjugate vaccines, 10% to 15% of cases of osteomyelitis in children younger than 3 years were caused by this organism.14,17 Invasive disease with Hib is rare in children who have been immunized during infancy.28 Community-acquired methicillin-resistant Staphylococcus aureus (CA-MRSA) osteomyelitis is being reported with increasing frequency. Initial reports of CA-MRSA infections described predominantly skin and soft-tissue infection, but recent case series have reported an increasing number of infections involving the bones or joints.29–32 Some CA-MRSA isolates causing osteoarticular infection carry the genes for Panton–Valentine leucokidin (PVL), a virulence factor associated with complicated osteomyelitis.29

Laboratory Diagnosis Bacteriologic diagnosis can be conÀrmed in 50% to 80% of cases of AHO; the yield is highest when multiple specimens, including blood, bone, and joint fluid, are cultured.3,11,12,16,17,33 Diagnosis of K. kingae infection is enhanced with intraoperative inoculation of culture material directly into liquid media or on to agar plates or when polymerase chain reaction (PCR) analysis is performed.19,20,36 Cultures should be held for at least 7 to 10 days. The erythrocyte sedimentation rate (ESR) is elevated in up to 90% of cases of osteomyelitis, and the C-reactive protein (CRP) level in 98%.16,17,32a,37 The mean ESR is 40 to 60 mm/h, but levels > 100 mm/h can occur. ESR generally peaks 3 to 5 days after initiation of therapy and returns to normal in approximately 3 weeks; the CRP level peaks by the second day (mean, 8.3 mg/dL) and returns to normal (< 2.0 mg/dL) after approximately 1 week of therapy.37 Patients infected with PVL-positive S. aureus are more likely to have positive blood cultures and higher ESR and CRP levels at presentation.32a Higher levels of CRP at diagnosis may predict greater risk of sequelae.38 The peripheral white blood cell count can be elevated or

80

475

Humerus 12% Ulna 3%

Hands and feet 13%

Clinical Characteristics and Differential Diagnosis Most patients with AHO have symptoms for < 2 weeks before they are brought to medical attention, although a small proportion have lowgrade fever and intermittent bone pain for several weeks.12,16 The most common manifestations are fever, pain at the site of infection, and reluctance to use an affected extremity.17,33 Less common complaints are anorexia, malaise, and vomiting. Physical Àndings consist of focal swelling, tenderness, warmth, and erythema (usually over the metaphysis of a long bone). Rarely, a draining Àstulous tract is seen over the affected bone.12 Tenderness out of proportion to soft-tissue Àndings suggests osteomyelitis rather than soft-tissue infection or cellulitis. Exaggerated immobility of the joint and lack of point tenderness over the metaphysis suggest pyogenic arthritis. Other causes of bone pain are fracture, bone infarction secondary to hemoglobinopathy, leukemia, and bony neoplasms such as metastatic neuroblastoma and Ewing sarcoma. Osteomyelitis most frequently occurs in the long bones (Figure 80-2), although in some series, 10% to 25% of cases involve short or nontubular bones, including the pelvis, clavicle, calcaneus, skull, ribs, and scapula.16,17,34 Multiple bones are involved in about 5% of cases. Patients with infection caused by CA-MRSA have prolonged fever and duration of hospitalization compared to those with communityacquired methicillin-sensitive S. aureus (CA-MSSA). They are also more likely to have multiple foci of infection, pyomyositis, and subperiosteal and intraosseous abscesses.32,32a,35,35a Patients with PVL + CA-MRSA are at increased risk for deep-vein thrombosis or septic emboli to the lungs.29,31,32 Infection with PVL + S. aureus may also be more likely to result in chronic osteomyelitis.29

CHAPTER

Radius 4%

Pelvis 9%

Femur 27%

Tibia 22%

Fibula 5%

Figure 80-2. Sites of acute osteomyelitis in 657 children in whom a single bone was involved. Shaded areas constitute sites of approximately 75% of cases. Miscellaneous sites accounting for 5% are not shown. (Data collated from references 11, 16, 17, and 33.)

normal; thrombocytosis may be noted, especially if symptoms have been present for more than 1 week.

Radiologic Diagnosis Plain Radiographs Radiographic abnormalities of osteomyelitis reflect inflammation, destruction, and new formation of bone.3,39 The earliest abnormalities, seen within the Àrst 3 days of onset of symptoms, are deep soft-tissue swelling and loss of the normally visible tissue planes around the affected bone. Osteopenia or osteolytic lesions from destruction of bone are not usually visible until approximately 50% of bone has been demineralized. Lytic lesions, periosteal elevation due to subcortical purulence, and periosteal new bone formation appear approximately 10 to 20 days after onset of symptoms. Sclerosis of bone is seen when infection has been present for longer than a month. If deep soft-tissue swelling is noted on plain radiograph in a patient with a short history of symptoms and with point tenderness over the affected metaphysis, no further imaging studies are necessary to support the diagnosis of osteomyelitis.

Radionuclide Scanning Radionuclide scans are useful in the early diagnosis of osteomyelitis, even when plain radiographs are normal. Technetium-labeled methylene diphosphonate isotope is most frequently used because its uptake by infected bone is enhanced when osteoblastic activity is increased. The reported sensitivity of technetium-99 bone scanning is between 80% and 100% (Table 80-1),40–44 (Figures 80-3 and 80-4). The bone scan can be normal in 5% to 20% of children with osteomyelitis in the Àrst few days of illness.43,45–47 Bone scans have the

476

SECTION

K Bone and Joint Infections

TABLE 80-1. Technetium Bone Scans for Osteomyelitis in Children Studya

Year

Patient Age

Tuson et al.40 Hamdan et al.41 Howie et al.42 Sullivan et al.43 Bressler et al.44

1994 1987 1983 1980 1984

6 months–13 years 6 months–14 years 6 weeks–13 years 3 weeks–15 years < 6 weeks

Sensitivity (%) 92 89 89 81b 100

a

Superscript numbers indicate references. The 2 patients younger than 6 weeks of age had normal scans.

b

advantage of being less expensive compared with MRI, and sedation is not usually required. Bone scans are particularly useful when multifocal bone involvement is suspected.48 A variety of disorders, including malignancy, deep soft-tissue infection, cellulitis, pyogenic arthritis, trauma, fracture, and bone infarction, can result in positive scan results and contribute to decreased speciÀcity of radionuclide scan for the diagnosis of AHO. Metaphyseal site of maximal uptake and lack of uptake in bone on both sides of a joint support the diagnosis of AHO, rather than pyogenic arthritis. Diaphyseal uptake suggests tumor, trauma, or infarction. In some cases of osteomyelitis, bone scans show areas of decreased technetium uptake (“cold scans”), probably reflecting compromised vascular supply to the bone from ischemia or thrombosis40,49; such Àndings make differentiation of osteomyelitis from infarction associated with sickle-cell disease difÀcult. Patients with osteomyelitis who present with decreased uptake on bone scan may have more aggressive infection.49

Nuclear scanning using gallium-67 citrate or indium-111-labeled leukocytes has been evaluated for the diagnosis of osteomyelitis. Gallium scan results are positive in diseases characterized by increased bone turnover and, thus, have limited speciÀcity for osteomyelitis.50 In contrast, indium scanning, which reflects migration of white blood cells into areas of inflammation, is useful in the diagnosis of osteomyelitis associated with trauma, surgery, chronic ulcers, or prosthetic devices.51 A limitation of indium scanning is poor localization of infection (i.e., bone versus soft tissue). This problem can be overcome in some cases by simultaneous performance of a technetium bone scan.52 Other limitations of indium scans are decreased sensitivity in diagnosing infection in the central skeleton and relatively high radiation exposure.

Magnetic Resonance Imaging Magnetic resonance imaging (MRI) is an effective means of imaging bone.53,54 Its reported sensitivity for detection of osteomyelitis ranges from 92% to 100%. Normal red and fatty portions of the bone marrow have a characteristic appearance on MRI. Fatty marrow produces a bright signal on T1-weighted images.55 Changes in marrow caused by infection and inflammation produce an area of low signal intensity within the bright fatty marrow. Areas of low signal intensity in infected marrow seen on a T1-weighted image change to bright signal intensity on a T2-weighted image. These changes are not speciÀc for osteomyelitis and can be seen with malignancy, fracture, and bone infarction. MRI can detect signal alterations in soft tissue and is particularly useful in differentiating cellulitis from osteomyelitis. MRI may also be able to differentiate acute from chronic osteomyelitis.56 Gadolinium-enhanced MRI may be particularly useful in the diagnosis of soft-tissue, muscle, or bone abscesses associated with

B A

C

Figure 80-3. Typical technetium-99 bone scan findings of acute osteomyelitis in a 4-year-old girl who had a 2-day history of fever, ankle pain, and swelling, and refusal to bear weight. Plain radiograph was unremarkable. Triple-phase anterior images of bone scan show increased tracer activity in the region of the right ankle in the angiographic (immediate) phase (A), increased tracer activity in the soft tissues in the region of the ankle in the blood pool (15-minute) phase (B), and localization of tracer in the distal tibial metaphysis (arrow) without periarticular distribution in the delayed (2.5-hour) phase (C). Diagnosis was confirmed at surgery with finding of subperiosteal pus; Streptococcus pyogenes was isolated. (Courtesy of E. Geller and S.S. Long, St. Christopher’s Hospital for Children, Philadelphia, PA.) PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Osteomyelitis

CHAPTER

80

477

D

A B

C

Figure 80-4. Plain film and technetium-99 bone scan of acute osteomyelitis and pyogenic arthritis in a 7-month-old male who had a 3-day history of high fever, fussiness and redness, and swelling of the lateral right leg from the thigh to the lower leg, with limitation of motion of the knee. Aspiration of the knee revealed 169 000 white blood cells/mm2 and gram-positive cocci. Plain film shows obscuration of right lateral subcutaneous fat–muscle plain, from the thigh to the lower leg (A). Triple-phase bone scan shows increased tracer activity in the region of the right knee in the angiographic phase (B) and in the periarticular soft tissues in the blood pool phase (C), as well as localization of tracer in the distal femur (arrow) but not in the proximal tibia in the delayed phase (D). Diagnosis of osteomyelitis of the femur was confirmed at surgery. Methicillin-susceptible Staphylococcus aureus was isolated from blood, joint and bone specimens. (Courtesy of E. Geller and S.S. Long, St. Christopher’s Hospital for Children, Philadelphia, PA.)

MRSA infection.35 MRI has been used to identify bone marrow infection in cases of Bartonella henselae infection, particularly when plain films and computed tomography (CT) of the affected bones appear normal.25

TREATMENT Antibiotic Choice Appropriate antibiotic selection is critical to successful treatment of osteomyelitis. Considerations in choosing specific agents are the age of the child, underlying medical conditions, suspected pathogens and their susceptibility pattern, antibiotic pharmacodynamics, and the safety and efficacy of the antibiotic. Most b-lactam antibiotics achieve therapeutic concentrations in bone, both in experimental animal models and in vivo.57,58 Clindamycin has particularly good bone penetration, attaining a high bone-to-serum ratio.57,59–61 Vancomycin has excellent penetration into the bones of experimental animals.57 Aminoglycoside agents theoretically have poor bactericidal activity in bone because of local tissue hypoxia and acidosis. Although ciprofloxacin penetrates well into bone, it is not routinely recommended for use in young children.62 Most cases of AHO in any age group are caused by Staphylococcus aureus. Empiric therapy should include coverage for this organism. Parenterally administered nafcillin, clindamycin, or a first-generation cephalosporin have historically been usual choices for empiric therapy, however, the increase in CA-MRSA infections has made the choice of initial antibiotic therapy more challenging. Direct sampling of bone for culture is increasingly important. Clindamycin remains a good choice for empiric therapy in communities where resistance (both constitutive and inducible) occurs in fewer than 10% of S. aureus isolates.35 In communities where CA-MRSA resistance to clindamycin and methicillin is greater than 10% to 15%, vancomycin is the drug of choice for empiric therapy. Clindamycin and vancomycin are active against most isolates of Streptococcus pyogenes and S. pneumoniae. Neither offers coverage for K. kingae infection, which is susceptible to most b-lactam antibiotics. Ampicillin-sulbactam

should be considered in addition to clindamycin or vancomycin, especially in young children. Before the widespread use of Hib conjugate vaccines, empiric therapy for young children included a third-generation cephalosporin or chloramphenicol in addition to nafcillin. A second-generation cephalosporin such as cefuroxime alone was a reasonable alternative.63 With the dramatic reduction of cases of Hib disease in the United States, it is reasonable to use only an antistaphylococcal–antistreptococcal antibiotic in the fully immunized and immunocompetent child. Neonates with osteomyelitis should be treated with antibiotics active against Staphylococcus aureus, group B streptococcus (GBS), and gram-negative enteric organisms. Suggested initial empiric parenteral antibiotic therapy for neonates and children with AHO is outlined in Table 80-2. Antibiotic therapy should be modified according to results of culture and susceptibility testing. Nafcillin, oxacillin, a firstgeneration cephalosporin, or clindamycin are the drugs of choice for infections caused by MSSA. S. aureus that is resistant to erythromycin but susceptible to clindamycin should be tested for the presence of inducible macrolide-lincosamide-streptogramin B resistance by means of the D test. Clindamycin should not be used if inducible resistance is detected because of reports of treatment failure under this circumstances.64,65 Alternative antibiotic choices for CA-MRSA infection in children are limited. Limited data support efficacy of linezolid.66,67 In one group of pediatric patients, linezolid was as effective as vancomycin in treating infections caused by MRSA, although efficacy in treating bone or joint infections was not specifically evaluated.67 Linezolid has excellent oral bioavailability and offers an alternative to prolonged intravenous therapy. However, it is expensive and many children object to the taste of the oral suspension. Long-term use has been associated with neutropenia, thrombocytopenia, and elevated transaminases.68 There are rare reports of lactic acidosis and optic neuropathy associated with linezolid therapy.69,70 Most CA-MRSA isolates are susceptible to trimethoprimsulfamethoxazole, tretrcyclines, and rifampin. Clinical experience with these drugs for osteoarticular infections is limited.71,72 Rifampin should never be used as a single agent because of rapid development of resistance. Although daptomycin and tigecycline have activity

478

SECTION

K Bone and Joint Infections

TABLE 80-2. Antibiotic Selection for Initial Treatment of Osteomyelitis Dosage Patient Age

Likely Pathogen

Antibiotic Selection

> 3 years

Staphylococcus aureus

Nafcillin or Clindamycin or Vancomycin

150

Nafcillin or Clindamycin or Vancomycin plus Cefotaxime or Ceftriaxone or Cefuroxime or Ampicillin-sulbactam

300

4

Nafcillin or Vancomyin plus Gentamicin OR Nafcillin or Vancomycin plus Cefotaxime

100

4

30

2

< 3 years

Neonate

Staphylococcus aureus Haemophilus influenzae type ba Kingella kingaeb

Staphylococcus aureus Group B streptococcus Enteric gram-negative bacteria

mg/kg per day

Doses/Day 4

30

3–4

45

3

150

4

30

3–4

45

3

100–150

4

100

2

75–150

5–7.5

3

3

100

4

30

2

150

3

a

If patient is fully immunized against Haemophilus influenzae type b, consider using only antistaphylococcal coverage. b If patient is treated with either clindamycin or vancomycin, consider adding an ampicillin or cephalosporin agent for Kingella kingae.

against MRSA, neither is approved for use in children. Daptomycin is also not indicated for treatment of pulmonary infections, which may accompany osteomyelitis caused by CA-MRSA. If no organism is isolated, and the symptoms are resolving, initial empiric therapy should be continued. In one recent case series, most cases of culture-negative osteomyelitis responded to therapy with antistaphylococcal antibiotics.73 Management of children with osteomyelitis should include concurrent care by an orthopedic surgeon. Indications for surgery include prolonged fever, erythema, pain, and swelling; persistent bacteremia despite adequate antibiotic therapy; soft-tissue or periosteal abscess; formation of a sinus tract; and presence of necrotic, nonviable bone.15,74–76

Sequential parenteral–oral antibiotic regimens can be successful.77–82 The change to oral antibiotics is generally made when fever, pain, and signs of local inflammation have resolved and laboratory values, especially CRP, are normalizing. The willingness of the child to take oral medication and the likelihood of adherence to the regimen must also be assessed. The oral antibiotic should have the same spectrum of coverage as the parenteral drug. If a b-lactam agent is used, dosage required is generally two to three times the usual oral dose (Table 80-3). Once the change to oral therapy is made, the child should be monitored to ensure continued clinical improvement. The CRP level is expected to return to normal 7 to 10 days after initiation of appropriate therapy, and the ESR normalizes within 3 to 4 weeks.37,81

Duration of Therapy

SPECIAL CLINICAL SITUATIONS

Duration of antibiotic therapy depends on the cause and extent of infection as well as the clinical course. The usual duration ranges from 3 to 6 weeks and should be individualized on the basis of severity of illness and clinical response. Historical evidence suggests that < 3 weeks of treatment is associated with higher rates of relapse or recurrence than longer duration of therapy.12 Plain radiographs obtained at the end of therapy may be useful as a marker of maximal anticipated destruction, a baseline for further studies, and a guide to possible complications. Central venous catheters (CVCs) are often placed to continue intravenous antibiotics at home. Use of CVCs for > 2 weeks in children with osteomy can be associated with an increase in CVC related complications, including catheter malfunction, thrombosis and catheter associated bloodstream infections.76a

Neonatal Osteomyelitis Osteomyelitis is uncommon in the neonatal period. The incidence is unknown but is estimated to be approximately 1 to 3 cases for every 1000 intensive-care nursery admissions.83 Associated risk factors include prematurity, low birthweight, preceding infection, bacteremia, exchange transfusion, and the presence of an intravenous or umbilical catheter.84–87 Osteomyelitis of the skull secondary to contiguous spread of infection has occurred as a complication of fetal scalp electrode monitoring88–90and in association with infected cephalohematoma.91,92 Osteomyelitis of the calcaneus has complicated heel lancet puncture.85,93 The diagnosis of osteomyelitis in neonates is often delayed because of nonspeciÀc symptoms.94 Signs and symptoms include

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Osteomyelitis

CHAPTER

80

479

TABLE 80-3. Dosage of Antibiotics Commonly Used in the Oral Phase of Treatment of Osteomyelitisa Dosage Antibiotic Dicloxacillin Cephalexin Clindamycin

mg/kg per day 75–100 100–150 30

Doses/Day 4 4 3–4

a In general, the dose of b-lactam antibiotics used for osteomyelitis is 2–3 times the usual dose.

fever, irritability, swelling or decreased movement of a limb (pseudoparalysis), erythema, and tenderness over the affected bone.83–85,95 Preterm infants are more likely than term infants to manifest symptoms of septicemia.96 Approximately 20% to 50% of neonates with osteomyelitis have infection of multiple bones,83,84,97 and about 75% have suppurative arthritis of contiguous joints84 (Figure 80-5). Staphylococcus aureus (including CA-MRSA98), GBS, and enteric gram-negative bacilli are the most common causes (Table 80-4); fungi,99 Ureaplasma urealyticum,100 S. epidermidis,101 Neisseria gonorrhoeae,95 and anaerobic88 bacteria are unusual causes. Infants with GBS bone infection have usually had an uncomplicated neonatal course and have infection of a single bone. There is as a predilection for involvement of long bones on the right, particularly the right proximal humerus.101,102 This predilection may be secondary to trauma to the upper arm as it passes below the symphysis pubis during vaginal delivery.101 Misdiagnosis of bone infection as trauma in these mildly ill infants is common. Since the release of the 1996 consensus guidelines for prevention of GBS disease, the incidence of early-onset perinatal GBS disease has declined in the United States.103 The white blood cell count is commonly normal; ESR and CRP are often elevated.104 In most infants, an osteolytic lesion is visible on plain radiograph 10 to 12 days after onset of symptoms (frequently at the time of diagnosis).76,84,85,95,105,106 Radionuclide bone scans may be positive44,83,94 but sometimes are less sensitive than plain radiographs.85 Neonatal osteomyelitis can lead to permanent joint abnormalities or disturbance in skeletal growth secondary to damage to the cartilaginous growth plate, including arthritis, decreased range of motion, limb length discrepancy, and gait abnormalities.107 The reported incidence of permanent sequelae varies from 6% to 50%.84,85,96,101

Vertebral Osteomyelitis Vertebral osteomyelitis accounts for approximately 1% to 3% of cases of osteomyelitis in children.12,17,108,109 Boys are affected twice as frequently as girls.110 Infection usually occurs as a result of bacterial seeding of the vertebral bodies by hematogenous arterial or venous spread. It can also result from extension of soft-tissue infection or complication of a surgical procedure.15,111 Clinical manifestations can be indolent and nonspecific, leading to delayed diagnosis. Young infants can have nonspecific signs of septicemia.112 Symptoms in older children include back, chest, abdominal, or leg pain as well as loss of normal curvatures.108,113 Rarely, children manifest dysphagia secondary to a paravertebral or retropharyngeal abscess or acute spinal cord paralysis due to paraspinal compression.108,114 Fever is common, and tenderness over the involved vertebrae is expected. Neurologic deficits are found in 15% to 20% of cases.108,110 S. aureus is isolated in most cases.15,108,110,115 In one review, Salmonella spp. caused 12% of cases of childhood vertebral osteomyelitis.108 Gram-negative bacilli such as Escherichia coli cause vertebral osteomyelitis in adults, particularly those with a history of recent urinary tract infection or instrumentation. Vertebral

Figure 80-5. Plain radiograph of an infant at 2 months of age who had Staphylococcus aureus bloodstream infection during a stay in the neonatal intensive care unit in the first month of life. He had multiple sites of infection, including the proximal and distal femur and hip joint. Note widespread destruction of femur, acetabulum, and joint. (Courtesy of S.S. Long, St. Christopher’s Hospital for Children, Philadelphia, PA.)

TABLE 80-4. Frequency of Organisms Causing Neonatal Osteomyelitis (n = 128) Single Organism

%

Staphylococcus aureus b-Hemolytic streptococcus Group B Group A Enteric gram-negative bacilli Staphylococcus epidermidis Nonhemolytic streptococcus Streptococcus pneumoniae Neisseria gonorrhoeae Multiple organisms

60.6 16.5 3.1 8.7 2.4 2.4 0.8 0.8 4.7

osteomyelitis in intravenous drug users is commonly caused by Pseudomonas aeruginosa and less commonly by Staphylococcus aureus, Serratia spp., Klebsiella spp., Enterobacter spp., or Candida spp.116,117 Tuberculosis and brucellosis should be considered if symptoms and radiographs suggest chronic infection.118,118a Characteristic findings on plain radiograph consist of narrowing of the involved disk space, lucency of the adjacent vertebral bodies, and, eventually, reactive sclerosis of the bone with fusion of vertebral bodies.15,108,115 Vertebral osteomyelitis is differentiated from diskitis radiographically by the minimal vertebral endplate involvement associated with diskitis (see Chapter 82, Diskitis). MRI is reported to be highly sensitive (96%) and specific (92%) for diagnosis of vertebral osteomyelitis.119,120 Blood culture results are positive in only about 30% of acute cases. When blood culture results are negative, strong consideration should be given to obtaining a biopsy specimen from the vertebral body for culture and histologic examination.120 Children with uncomplicated vertebral osteomyelitis and no evidence of abscess formation should be treated with at least 4 weeks of parenterally administered antibiotics.108 Treatment of shorter duration is associated with therapeutic failure.110 Surgical decompression or debridement or both are indicated in the presence of spinal epidural abscess, signs of spinal cord compression, or extensive bony destruction. Complications of vertebral osteomyelitis include neurologic deficits secondary to epidural abscess,121 paravertebral abscess,122 and infected aneurysms of the aorta.123

480

SECTION

K Bone and Joint Infections

Pelvic Osteomyelitis Approximately 6% to 9% of all cases of hematogenous osteomyelitis involve the bones of the pelvis.34,124,124a The ilium and ischium are most commonly involved124; infection of the sacrum, acetabulum, or pubic symphysis is rare.124–126 Pathogenesis is not clearly defined, but risk factors include a history of pelvic trauma, intravenous drug use, and genitourinary procedures.124,126 Most patients with pelvic osteomyelitis have fever, gait abnormalities, and pain that is often localized to the hip, groin, or buttock.124,126–129 Pain with hip movement and point tenderness over the affected bone is often observed. Clinical features can mimic those of pyogenic arthritis of the hip124; however, pelvic osteomyelitis is more likely to be associated with near-normal range of motion of the hip, absence of referred pain to the knee, specific point tenderness over the affected bone (or pain on rocking of the pelvic girdle), and abnormal rectal findings.129 Responsible pathogens are similar to those that cause osteomyelitis of long bone. Plain radiographs of the pelvis are often normal. Technetium scanning,124,124a,127–129 MRI, or CT can suggest the correct diagnosis and may be useful for differentiating osteomyelitis from bacterial infection of the muscles of the pelvic girdle. Patients should be treated for at least 4 weeks with antibiotics.124–126,129 Surgical drainage or debridement should be considered in cases of extraosseous abscess formation or in patients whose symptoms do not respond rapidly to intravenous antibiotic therapy.

Children with Sickle Hemoglobinopathies Children with sickle-cell disease have increased susceptibility to bacterial infections, including osteomyelitis.130 Although the pathogenesis of osteomyelitis is poorly understood, microscopic areas of infarction in the intestinal mucosa and bone probably develop during sickling, resulting in bacteremia and focal bone infection. Splenic hypofunction, impaired opsonization, impaired macrophage function, and microembolism as well as tissue infarction are likely contributing factors in osteomyelitis.131,132 Salmonella spp. plus other gram-negative enteric bacilli cause > 70% of cases of osteomyelitis in children with hemoglobinopathies.133–136 Staphylococcus aureus is also an important cause of osteomyelitis in this population. Other organisms causing osteomyelitis in children with sickle hemoglobinopathies are listed in Box 80-1. Manifestations of osteomyelitis in children with sickle-cell disease are difficult to differentiate from those of acute vaso-occlusive crisis. Fever, bone pain, and leukocytosis are common to both conditions. Temperature > 39°C, toxic appearance, and an absolute band count > 500 cells/mm3 are more consistent with infection; however, there is considerable clinical and laboratory overlap.133 Distinctive features of osteomyelitis in children with sickle-cell disease are frequent involvement of the diaphyses of long bones, flat bones, and small bones of the hands and feet as well as multifocal, symmetrical bone involvement.131,137 Plain radiograph, technetium scanning, and MRI cannot differentiate infarction from infection.137,138 Therefore, if fever and

BOX 80-1. Bacterial Causes of Osteomyelitis in Children with Sickle-Cell Disease COMMON Salmonella spp. Staphylococcus aureus LESS COMMON Escherichia coli Haemophilus influenzae type b Shigella spp. Streptococcus pneumoniae

bone pain have not improved after supportive care has been given for vaso-occlusive crisis, needle aspiration of the affected area of bone for Gram stain and culture should be performed. Prolonged courses of parenteral antibiotic therapy (6 to 8 weeks) may be necessary for treatment of osteomyelitis in patients with sickle hemoglobinopathy; oral therapy can be substituted for parenteral treatment once a pathogen has been confirmed and there is clinical improvement.131,133

OSTEOMYELITIS DUE TO UNUSUAL ORGANISMS Fungal Osteomyelitis Fungal osteomyelitis is unusual in healthy children, occurring occasionally in neonates, immunocompromised patients, and intravenous drug users.99,139–142 Osteomyelitis caused by Candida spp. is reported in intravenous drug users and in premature or ill neonates.99,139–143 Aspergillus spp. cause osteomyelitis in children with chronic granulomatous disease, usually resulting from contiguous spread of pulmonary infection.144 Blastomyces dermatitidis, Coccidioides immitis, Histoplasma capsulatum, and Cryptococcus neoformans cause osteomyelitis in indigenous geographic regions and in immunosuppressed hosts.145–148

Tuberculous Osteomyelitis Skeletal lesions occur in approximately 1% of children with tuberculosis.118,149 Bones and joints are infected through hematogenous or lymphatic dissemination of Mycobacterium tuberculosis. Infection can smolder for years before clinical signs are apparent. The most commonly involved bones are the spine (tuberculous spondylitis), femur, long bones around knees and ankles, and small bones of the hands and feet.150 Other sites less frequently infected are the ribs, mandible, sternum, clavicle, and other long bones. Multifocal osteomyelitis is reported in 10% to 15% of cases.151,152 Clinical signs and symptoms of skeletal tuberculosis include lowgrade fever, weight loss, pain, and soft-tissue swelling at the site of infection. Vertebral involvement begins in the anterior vertebral body, eventually causing disk space collapse and anterior wedging of vertebral bodies, and sometimes gibbus deformity. The lower thoracic spine is the usual site of involvement (Pott disease), followed by the lumbar spine. The Mantoux tuberculin skin reaction is usually positive. Plain radiographic findings include periarticular osteopenia, lytic lesions in the body of the vertebra, joint space narrowing, and soft-tissue swelling.153 The chest radiograph is often normal. CT is useful for the evaluation of bone destruction, adjacent soft-tissue abscess formation, and calcification, and in guiding percutaneous biopsy.154,155 MRI is helpful in determining extent of bone and soft-tissue disease.156 Biopsy specimens should be obtained in an attempt to demonstrate the organism with stains and culture. Antituberculous therapy given for 2 months with four drugs, followed by 7 to 10 months of isoniazid and rifampin daily or twice weekly, is recommended (see Chapter 134, Mycobacterium tuberculosis).157 When infection with multidrug-resistant M. tuberculosis is proved or strongly suspected, modification of the treatment regimen may be necessary. Surgical intervention is indicated in cases of spinal instability and neurologic impairment secondary to paravertebral abscess formation and for drainage of soft-tissue abscesses. Atypical Mycobacterium spp. infrequently cause osteomyelitis in immunocompromised individuals.

Anaerobic Bacterial Osteomyelitis Anaerobic bacteria are commonly associated with chronic and nonhematogenously acquired osteomyelitis.158–160 Risk factors include PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Osteomyelitis

surgery, trauma, diabetes, human bites, chronic otitis media or sinusitis, dental infection, fibrous dysplasia of bone, presence of a prosthesis, and decubitus ulcers (Figure 80-6). Children are more likely than adults to experience anaerobic osteomyelitis of the skull and facial bones.158 Osteomyelitis of ribs follows contiguous spread from aspiration lung infection; Actinomyces spp. are the primary pathogens. Soft-tissue swelling or abscess can be the presenting abnormality. Similarly, Actinomyces spp. can cause osteomyelitis of the maxilla or mandible, frequently without dental pathology. Infection is usually polymicrobial, often due to both anaerobic and aerobic bacteria. Gram-positive cocci, Bacteroides spp., Prevotella spp., and Fusobacterium spp. are the most common anaerobes; S. aureus is the most commonly associated aerobic isolate.159 Therapy consists of treatment of underlying conditions, surgical debridement of necrotic bone, and appropriate antibiotic therapy. Examples of effective antibiotics are clindamycin, metronidazole, imipenem, and amoxicillin-clavulanate. Many anaerobic isolates are susceptible to penicillin. The choice of antibiotic depends on the specific organisms isolated and their potential for b-lactamase production. Therapy is protracted, frequently exceeding 1 year of oral penicillin or amoxicillin plus probenecid for actinomycosis.

CHAPTER

80

481

affected bone is typical. Plain radiograph shows bony destruction. Peripheral white blood cell count is often normal, and ESR and CRP may be normal. Nonhematogenous osteomyelitis is often caused by S. aureus, although coinfection with gram-negative and anaerobic organisms occurs.159 Examination of specimens from an associated sinus tract is unreliable in defining the etiology of osteomyelitis.169 Because some cases of nonhematogenous osteomyelitis are secondary to multiple or unusual organisms, biopsy with culture of the affected bone is the best method of determining appropriate antibiotic therapy. Therapy should be prolonged if infection is chronic, sometimes with parenteral therapy followed by oral therapy for a total of 4 to 6 months. The rate of recurrence in nonhematogenous osteomyelitis is as high as 40%, even with prolonged courses of antibiotic therapy.168 Aggressive surgical debridement or other interventions are required in addition to antibiotic therapy. Implanted devices must commonly be removed to effect a cure. A muscle flap procedure to re-establish the blood supply in decubitus ulcers with osteomyelitis is sometimes successful.169a

PUNCTURE WOUND OSTEOCHONDRITIS OF THE FOOT NONHEMATOGENOUS OSTEOMYELITIS Contiguous Infection Factors associated with the development of nonhematogenous osteomyelitis include open fractures requiring surgical reduction,161,162 implanted orthopedic devices, decubitus ulcers,158 and neuropathic ulcers.163,164 Facial osteomyelitis is usually secondary to untreated mastoiditis, sinusitis, or periodontal abscess.158,162 Osteomyelitis has been reported after local soft-tissue infection or direct involvement of bone from human and animal bites.165,166 Puncture wounds can lead to osteomyelitis of the foot (e.g., stepping on a nail or toothpick167) or the patella (from kneeling on a needle). Indolent presentation is common among children with nonhematogenous osteomyelitis. Fever is present in less than half of patients.168 Persistent drainage or ulceration of the soft tissue over the

CT

Bacterial osteochondritis is a complication of puncture wounds of the foot, occurring in approximately 1.5% of such injuries.167,170–172 Symptoms of osteochondritis, which appear several days to weeks after the initial injury, include increasing tenderness, erythema, and swelling at the site of the puncture. Infection is usually caused by Pseudomonas aeruginosa,173 often as a result of inoculation from the colonized moist soles of tennis shoes.174 S. aureus is the usual pathogen if symptoms began within 3 to 5 days of injury. Obtaining specimens from bone for microbiologic diagnosis is desirable. Empirical antibiotic therapy should include coverage for P. aeruginosa and S. aureus. Ticarcillin-clavulanate, piperacillintazobactam, ceftazidime or cefepime, with or without an aminoglycoside, would be appropriate empiric therapy. With appropriate surgical debridement, short-duration (7 to 10 days) parenteral antibiotic therapy has been effective,173,175 as has 3 to 4 weeks of antibiotic therapy without debridement for acute Pseudomonas osteomyelitis of the foot.174 Treatment with oral ciprofloxacin in conjunction with surgical debridement is also successful176 (see Chapter 292, Antimicrobial Agents).

CHRONIC RECURRENT MULTIFOCAL OSTEOMYELITIS

Figure 80-6. Actinomycosis of the mandible in a 9-year-old girl with history of painless expansion of the jaw over several months. Reconstructed spiral computed tomography shows expansion of bone with complex lytic and sclerotic mass (arrow). Surgical biopsy confirmed Actinomyces osteomyelitis and fibrous dysplasia of bone. (Courtesy of L. Kaban and M. Pasternack, Massachusetts General Hospital; and S.S. Long, St. Christopher’s Hospital for Children, Philadelphia, PA.)

Chronic recurrent multifocal osteomyelitis (CRMO) is an inflammatory disease of children and young adults characterized by recurring episodes of low-grade fever, swelling, and pain over affected bones and by radiologic abnormalities suggestive of osteomyelitis.177,178 Females are more frequently affected than males.179 The median age of onset of illness is 10 years. CRMO is sometimes associated with palmoplantar pustulosis,180 psoriasis, arthritis, sacroiliitis, inflammatory bowel disease,181 and Sweet syndrome. Radiographic abnormalities occur most commonly in the metaphysis of long bones and are characterized by radiolucent bone lesions with reactive sclerosis and soft-tissue swelling.179,182,183 The sternal end of the clavicle, the vertebral bodies, and the smaller bones of the hands and feet are often involved. Radiographic changes are similar to those seen in acute osteomyelitis, but multiple, often symmetrical lesions are present in CRMO. Bone scanning and MRI are useful in determining the extent and evolution of disease.183–185 An infectious cause of CRMO is not usually determined; cultures of bone biopsy and blood specimens are commonly sterile. The course of CRMO consists of prolonged bone pain with remissions and relapses over several years; the mean duration of disease is 6 years.186 In a long-term follow-up study of 23 patients with CRMO, 26% had active disease at a median of 13 years after

482

SECTION

K Bone and Joint Infections

diagnosis.181 Treatment with a variety of antibiotics has no apparent effect on the course or outcome. Some experts have advocated the use of corticosteroids or nonsteroidal anti-inflammatory agents for relief of symptoms.179,182 Interferon-g therapy was curative in one case report.187 Because multifocal bone lesions in childhood are seen with neuroblastoma, histiocytosis X, leukemia, and staphylococcal osteomyelitis, histologic examination and culture of bone specimens should be performed. Histologic findings in CRMO are nonspecific acute and chronic inflammatory changes; in the chronic phase of the disease, granulomatous changes can be seen.188

CHRONIC OSTEOMYELITIS Chronic osteomyelitis develops in fewer than 5% of cases of AHO11,189; it more often follows nonhematogenous osteomyelitis.168 Chronic osteomyelitis is characterized by alternating periods of quiescence and recurrent pain, swelling, and sinus tract drainage, persisting for years despite prolonged antibiotic therapy. Infections are often polymicrobial, and the original metaphyseal infection, with skeletal growth, moves to become a lytic lesion in the diaphysis.160,169a Surgical debridement of necrotic bone is often necessary. Alternative therapeutic approaches include the use of antibioticimpregnated polymethyl methacrylate beads190,191; local antibiotic delivery via implantable pumps192; and suction vacuum devices or bone grafts, skin grafts, and muscle flaps to eliminate dead space and improve vascularity.193–195

CHAPTER

81

Infectious and Inflammatory Arthritis Kathleen M. Gutierrez Infectious arthritis in children can be caused by bacteria, viruses, fungi, or mycoplasma. Pyogenic arthritis is characterized by a purulent inflammatory response, usually caused by bacterial infection. Reactive (inflammatory) arthritis is inflammation of one or more joints that can result from a response to infection elsewhere in the body or from a systemic inflammatory or autoimmune disorder.

INFECTIOUS ARTHRITIS Pyogenic (Bacterial) Arthritis

Animal models of H. influenzae type b bacterial arthritis illustrate possible mechanisms of articular damage.12 Bacterial endotoxin within the joint space stimulates release of tumor necrosis factor and interleukin-1.13,14 These cytokines stimulate production of proteinases by synovial cells and chondrocytes, enhancing leukocyte migration. Neutrophil elastases augment destruction of cartilage matrix within the joint.15,16 Bacteria also can spread to joints from contiguous osteomyelitis.17 The presence of transphyseal blood vessels in the child younger than 18 months facilitates spread of infection from the metaphysis across the growth plate to the epiphysis and adjacent joint space.18 In addition, the joint capsule of the hip and shoulder overlies the bony metaphysis of the femur and humerus, allowing direct extension of bone infection into these joint spaces. Primary pyogenic arthritis rarely extends into the bone to cause a secondary osteomyelitis. Joints also can be infected from penetrating wounds, intra-articular injections of medications, arthroscopy, and prosthetic joint surgery.19

Etiology Age is the most important predictor of etiology of pyogenic arthritis.1,5,20–23 S. aureus, enteric gram-negative organisms, and group B streptococcus (GBS) are the most frequent causes of pyogenic arthritis among neonates. S. aureus (methicillin susceptible [MSSA] and methicillin resistant [MRSA]), Kingella kingae, Streptococcus pyogenes and Streptococcus pneumoniae cause pyogenic arthritis in children younger than 5 years of age. In one series, K. kingae was the most common cause of pyogenic arthritis in children younger than 36 months.22 It is being reported with increasing frequency in the United States.24–27 H. influenzae type b infection is now rare in immunized immunocompetent children.28–30 Approximately one third of bone and joint infections caused by S. pneumoniae are caused by strains with decreased susceptibility to penicillin.31 Although S. pneumoniae is reported to cause approximately 6%32 of cases of pyogenic arthritis, the percentage of infections caused by vaccine serotypes likely will decrease in children who have the received the heptavalent pneumococcal vaccine.33,33a S. aureus and S. pyogenes are the most common causes of pyogenic arthritis in children older than 5 years. Cases of bone and joint infections caused by community acquired MRSA (CA-MRSA) are being reported with increasing frequency.34 CA-MRSA infections are more aggressive than CA-MSSA infections, involving multiple bones and joints35 and sometimes are associated with venous thrombosis and pulmonary disease.34,36 Other organisms reported to cause pyogenic arthritis in children include K. kingae,7 Neisseria meningitidis,37,37a Salmonella spp.,38,38a beta-hemolytic streptococci other than serogroups A or B,39 and rarely, anaerobic bacteria.40,41 Joint infections caused by Pseudomonas aeruginosa and Candida spp. are reported in intravenous drug abusers.9 Brucella spp. infection should be considered if a history of travel to endemic areas, contact with livestock or ingestion of unpasteurized dairy products is elicted.42,42a Arthritis related to Bartonella henselae infection has been reported.43

Epidemiology Although pyogenic arthritis occurs in all age groups, the peak incidence of disease is in children under 3 years of age.1–3 A history of trauma temporally related to the onset of arthritis caused by Staphylococcus aureus is common.4,5 Upper respiratory tract infection frequently precedes pyogenic arthritis caused by Haemophilus influenzae type b and Kingella kingae.5–7 Although most children have no underlying disorder, risk factors for pyogenic arthritis include immunodeficiency, hemoglobinopathy, diabetes, intravenous drug abuse, and rheumatoid arthritis.8–10

Pathophysiology Most cases of pyogenic arthritis in childhood follow hematogenous spread of organisms to the vascular synovium of the joint space.11

Clinical Manifestations Systemic symptoms commonly associated with pyogenic arthritis include fever, malaise, poor appetite, and irritability. Pain in the affected joint usually occurs early in the course of the illness. As infection progresses, the joint becomes swollen and red. Limp or refusal to walk occurs with infection of a lower extremity. If the affected joint is in the upper extremity, “pseudoparalysis” or refusal to use the affected joint is seen; manipulation causes pain. The infected joint is swollen, red, warm, and tender on palpation. Range of joint motion is decreased. The joints of the lower extremities are the most common sites of pyogenic arthritis.1 The knee is most often infected, the hip and ankle6 are less often affected (Table 81-1). More than 90% of cases of pyogenic arthritis are monoarticular.1,20 However, multiple joints can

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Infectious and Inflammatory Arthritis

be involved, particularly with infections caused by Neisseria gonorrhoeae, N. meningitidis, Salmonella spp., and occasionally, S. aureus. The diagnosis of pyogenic arthritis of the hip can be difficult because often there is no obvious joint swelling, and signs and symptoms are nonspecific, especially in infants and young children. Infants with pyogenic arthritis of the hip are irritable when the hip is moved (e.g., during diaper changes); soft-tissue swelling around the hip joint occasionally is noted and can extend to involve the entire leg.44 The affected hip often is held in a flexed, externally rotated and abducted position.45 Older children with pyogenic arthritis of the hip limp or refuse to walk and complain of localized pain. Range of motion of the hip joint is markedly decreased.

The erythrocyte sedimentation rate (ESR) is more than 20 mm/hour (mean, 44 to 65 mm/hour) in most patients with pyogenic arthritis.1,46,47 Similarly, the level of C-reactive protein (CRP) is often increased (mean, 8.5 mg/dL).47 A normal CRP is a good negative predictor for pyogenic arthritis. In one study, the probability that the patient did not have pyogenic arthritis was 87% if the CRP was < 1.0 mg/dL.48 Analysis of joint fluid is helpful in differentiating bacterial and other causes of arthritis (Table 81-2). Joint fluid in bacterial arthritis typically has a cloudy appearance. A leukocyte count of more than 50 000 cells/mm3, with a predominance of polymorphonuclear cells, is strongly suggestive of bacterial infection, even if culture of the joint fluid is negative.49,50 However, synovial fluid white blood cell (WBC) counts of less than 50 000/mm3 may occur in bacterial arthritis,20,45 and counts of more than 50 000/mm3 may occur in children with juvenile rheumatoid arthritis or Lyme disease.51,52 Synovial fluid glucose

Children with suspected pyogenic arthritis should have plain radiographic studies to exclude osteomyelitis or other osseous abnormalities. Soft-tissue swelling and widening of the joint space, as well as occasional associated osteomyelitis, can be observed radiographically in children with pyogenic arthritis. Erosion of subchondral bone can manifest only 2 to 4 weeks after onset of infection.55 Swelling of the hip capsule and lateral displacement or obliteration of the gluteal fat planes are early radiographic findings in pyogenic arthritis of the hip.55 With continued swelling of the hip capsule, the femoral head is displaced upward and outward, and lateral subluxation can occur. Concomitant osteomyelitis of the femur may be present.45 These findings are particularly common in infants, although in this age group, radiographic findings are difficult to interpret because of minimal ossification of the proximal femur. Plain radiograph is sometimes normal in children with proven pyogenic arthritis of the hip.56 Ultrasonography should be performed in suspected pyogenic arthritis of the hip. If fluid is present in the joint, a diagnostic aspiration should be performed under ultrasonographic guidance.57,58 False-negative ultrasound results are reported in children later diagnosed with pyogenic arthritis; these are the result of either inadequate imaging or imaging performed very early (< 24 hours) after onset of symptoms.59 Although technetium phosphate radionuclide scans generally are not used in the diagnosis of pyogenic arthritis, they are valuable in evaluating involvement of deep joints, such as the hip or sacroiliac joint. A characteristic finding is increased activity in the early (blood pool phase) and increased bony uptake on both sides of the joint (which would be uncharacteristic in osteomyelitis). Similarly, computed tomographic (CT) scan may be helpful in the diagnosis of

Pyogenic Arthritis No. 467 287 143 116 53 70

41 25 13 10 5 6

1,136

100

Totalb

%

a

Includes sacroiliac joint, joints of hands and feet, sternoclavicular joint. Some children had more than one joint affected. Data from references 1, 2, 4, 6, 81, 83.

b

TABLE 81-2. Characteristic Synovial Fluid Findings Diagnosis

WBC/mm3 (Typical)

WBC/mm3 (Range)

% PMNs (Typical)

Normal

< 150

—

< 25

Bacterial arthritis

> 50,000

2000–300,000

> 90

Tuberculous arthritis

10,000–20,000

40–136,000

> 50 (10–99)

Lyme arthritis

40,000–80,000

180–140,000

> 75

Candidal arthritis

–

7500–150,000

> 90

Viral arthritis

15,000

3000–50,000

< 50 (variable)

Reiter syndrome

15,000

10,000–22,000

> 70 (37–98)

Rheumatoid arthritis

–

2000–50,000

> 70

Rheumatic fever

25,000

2000–50,000

> 70

PMNs, polymorphonuclear cells; WBC, white blood cells. Data from references 50, 51, 52, 130, 134, 161.

483

Imaging Studies

TABLE 81-1. Frequency of Joint Involvement in 1050 Children with

Knee Hip Ankle Elbow Shoulder Othera

81

and protein levels do not differentiate reliably among most infectious and inflammatory processes and, therefore, have limited value.53 Blood culture should be obtained and synovial fluid sent for Gram stain, culture, and WBC count. Isolation of K. kingae is enhanced when synovial fluid is inoculated directly into fluid blood culture medium.27 Use of 16S ribosomal DNA PCR increased the identification of K. kingae in one group of children with osteoarticular infections.54 In the adolescent, specimens also should be obtained from the cervix or urethra, throat, skin lesions, and rectum for potential isolation of N. gonorrhoeae. When appropriate cultures are obtained, the bacterial cause is confirmed in 60% to 70% of cases of pyogenic arthritis.1,20 Blood cultures are positive in 40%2,19 of cases and joint fluid culture is positive in 50% to 60%.1,2,6,20

Diagnosis

Site

CHAPTER

484

SECTION

K Bone and Joint Infections

arthritis in areas of complex anatomy, such as the shoulder, hip, and sacroiliac joint. Magnetic resonance imaging (MRI) is highly sensitive for the early detection of infected joint fluid.60 Abnormal MRI Àndings in pyogenic arthritis include high signal periarticular changes and periarticular abscesses in some cases. MRI delineates abnormalities of adjacent bone, soft tissue, and extent of cartilage destruction. Alterations in signal intensity of bone marrow are seen more frequently in patients with pyogenic arthritis compared with those with toxic synovitis.60a

Treatment Children with pyogenic arthritis should be managed in conjunction with an orthopedic surgeon who is experienced in treating children. Goals of therapy include decompression, sterilization of the joint space, and removal of inflammatory debris. All children with pyogenic arthritis of the hip require prompt surgical drainage and irrigation of the joint space.21,61,62 Delay in drainage increases the likelihood of permanent damage because increased intra-articular pressure can compromise blood supply, resulting in avascular necrosis of the femoral head; muscle spasms may occur, predisposing the patient to dislocation of the hip. Open surgical drainage of joints other than the hip usually is not required. However, aspiration must be performed promptly to decompress the joint and obtain synovial fluid for analysis. Repeated aspirations are often necessary when fluid reaccumulates. Concurrent osteomyelitis may be associated with the need for repeat debridement of the joint.63 Debridement by arthroscopy has been undertaken in some cases of pyogenic arthritis of the knee and hip.64,64a

The initial choice of antibiotics is based on age, clinical history, and physical examination. Adequate penetration into the joint is essential. Penicillin, ampicillin, nafcillin, methicillin, dicloxacillin, some Àrst- and third-generation cephalosporins, clindamycin, vancomycin, and aminoglycosides attain acceptable concentrations in joints after intravenous or intramuscular administration, and agents that are well absorbed from the gastrointestinal tract do after oral administration.65–70 Because most antibiotic agents achieve high synovial fluid-to-serum ratios, there is no role for intra-articular instillation of antibiotics, which can produce chemical irritation and inflammation. Parenterally administered therapy is used initially (Table 81-3). Antistaphylococcal therapy should be given for a child of any age. Infants younger than 3 months of age should be treated with antibiotics active against S. aureus, gram-negative enteric organisms, and GBS. Children 3 months to 5 years of age should receive empiric therapy for S. aureus, K. kingae, S. pneumoniae, and S. pyogenes. Infection with H. influenzae type b is uncommon in immunized children, although other serotypes of Haemophilus occasionally cause pyogenic arthritis.71 Children older than 5 years are most likely to be infected with S. aureus or streptococci; initial therapy with an antistaphylococcal antibiotic is recommended. Empiric therapy for N. gonorrhoeae is indicated for the sexually active adolescent.72 In communities where more than 10% of S. aureus isolates are resistant to methicillin, empiric therapy with either vancomycin or clindamycin is suggested.73 Increasing rates of resistance to clindamycin may preclude its use as empiric therapy in some communities. Most S. pyogenes and S. pneumoniae isolates are susceptible to vancomycin and clindamycin. Neither of these drugs is effective in

TABLE 81-3. Empiric Antibiotic Therapy for Pyogenic Arthritis in Children Dosage Age

Likely Pathogen

Antibiotic

mg/kg per day

Doses/day

Neonate (doses are for infants > 2000 g and > 7 days old with normal serum creatinine)

Staphylococcus aureusa Group B streptococcus Gram-negative bacilli

Nafcillin or Vancomycin or Clindamycin plus Cefotaxime or Gentamicin

100

4

30

2

20–30

3

100–150

3

5–7.5

3

150

4

45

3

30

3

100–150

3–4

150

3

300 Unasyn

6

150

4

45

3

30 50

3 1

Child, ≤ 5 years

Child, > 5 years

Adolescent (sexually active)

a

Staphylococcus aureus Haemophilus influenzaeb Kingella kingaec Streptococcus pyogenes Streptococcus pneumoniae

Staphylococcus aureusa Streptococcus pyogenes

Neisseria gonorrhoeae (consider)

Nafcillin or Vancomycin or Clindamycin plus Cefotaxime or Cefuroxime or Ampicillin-sulbactamd Nafcillin or Vancomycin or Clindamycin Ceftriaxonee

a

If more than 10% of community-acquired isolates are methicillin-resistant Staphylococcus aureus, consider empiric therapy with either vancomycin or clindamycin until culture and susceptibility results are available. Children who have been completely immunized are less likely to have Haemophilus influenzae type b infection. c If empiric therapy with vancomycin or clindamycin is used, consider adding a second- or third-generation cephalosporin for Kingella kingae coverage in patients < 36 months of age. d The dose is for the drug Unasyn (300 mg Unasyn = 200 mg ampicillin plus 100 mg sulbactam). Sulbactam dose in an adult should not exceed 4 g. Unasyn is not approved for infant > 1 year of age, or for this indication. e The dose of ceftriaxone for children ≥45 kg is 1 g/day, given in a single dose. b

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Infectious and Inflammatory Arthritis

treating infection caused by K. kingae. Most beta lactam antibiotics, including ampicillin and second- and third-generation cephalosporins have activity against K. kingae.23 Nafcillin, oxacillin, or a first generation cephalosporin remain the drugs of choice if MSSA is isolated. Choices of antibiotics for MRSA infections in children are limited. Vancomycin is effective, but not available in a bioavailable oral formulation. If the isolate is susceptible to clindamycin but resistant to erythromycin, a D test should be performed to evaluate for the presence of inducible macrolidelincosamide-streptogramin B resistance; if negative result, clindamycin is an excellent oral agent. Linezolid, trimethoprimsulfamethoxazole and doxycycline have been used in some patients with serious MRSA infection (see Chapter 80, Osteomyelitis). Empiric therapy is given until an organism is isolated. Specific therapy based on culture results and susceptibility testing is continued parenterally until the child is afebrile; joint pain, swelling, and erythema have decreased; and joint mobility has increased. In addition, WBC counts of peripheral blood and synovial fluid and erythrocyte sedimentation rate (ESR), or CRP levels should be normalizing; elevated ESR persists longer than CRP during adequate treatment and longer in pyogenic arthritis compared with osteomyelitis. Open drainage of any joint (with lysis and irrigation of loculated collections) should be undertaken when aspirations yield samples that are persistently positive on culture. Orally administered antibiotic therapy can be substituted for parenteral treatment after adequate control of infection and inflammation has been achieved, if an oral antibiotic with appropriate coverage is available and if compliance and careful monitoring are ensured74–77 (Table 81-4). Use of clinical practice guidelines have been successful in decreasing the number of days parenteral antibiotics are given and duration of hospitalization, without increasing complications or sequelae.78 For children in whom oral therapy is not feasible, outpatient parenteral antibiotic therapy administered through central venous lines or a peripherally inserted central catheter, has been successful. Catheter related mechanical and infectious complications occur with prolonged intravenous administration of antibiotics and the risk versus benefit of prolonged central venous access should be considered carefully.79 Duration of therapy is determined by the specific pathogen, clinical and laboratory response, and whether adjacent osteomyelitis is present. Joint infections caused by S. aureus and gram-negative enteric organisms generally are treated for at least 3 to 4 weeks; a longer course of therapy may be necessary for pyogenic arthritis of the hip.45 Arthritis caused by H. influenzae, S. pneumoniae, S. pyogenes and K. kingae are treated for 2 to 3 weeks, depending on clinical response.

Prognosis Sequelae of pyogenic arthritis in children include abnormalities of bone growth, limitation of joint mobility, unstable articulation, and chronic dislocation of the joint. Joint dysfunction may not become apparent for months to years after infection.80 An estimated 10% to 25% of children with pyogenic arthritis have residual dysfunction.1,80

TABLE 81-4. Dosage of Antibiotics Commonly Used in Oral Treatment of Pyogenic Arthritisa

CHAPTER

81

485

A number of risk factors for development of sequelae have been identified and include the following: (1) age younger than 6 months1–6; (2) infection of the adjacent bone, which is evident in 10% to 16% of children with pyogenic arthritis and increases the likelihood of sequelae to approximately 50%1,21,36,81,82; (3) infection of the hip or shoulder1,80,83; (4) a delay of 4 days or more before decompression and antibiotic therapy10,45,84,85,85a; and (5) prolonged time to sterilization of synovial fluid.86 Staphylococcal and gram-negative bacillary infections carry a high risk of sequelae, whereas meningococcal and gonococcal infections carry a low risk.

SPECIAL SITUATIONS AND PATHOGENS Neonatal Arthritis Risk factors for pyogenic arthritis in the newborn include umbilical vessel catheterization, presence of a central venous catheter,87 femoral vessel blood sampling,44,88,89 and possibly fetal breech presentation.90 Pyogenic arthritis often is a complication of osteomyelitis, and the onset is insidious (see Chapter 94, Bacterial Infections in the Neonate). The hip and knee are the most frequently involved joints.89 S. aureus, N. gonorrhoeae, and Candida spp. frequently cause polyarticular infection. If infection is contracted in the hospital, CA-MRSA and MSSA, enteric gram-negative organisms, and Candida spp. are common. GBS, S. aureus, and N. gonorrhoeae are the pathogens most commonly isolated from neonates who develop joint infections after hospital discharge.91

Neisseria gonorrhoeae Arthritis caused by N. gonorrhoeae must be considered in sexually active adolescents.72 The incidence of disseminated gonococcal infection in individuals with urethritis or cervicitis is approximately 1%.92 Disseminated gonococcal infection is characterized by mild fever, polyarthralgia, rash, tenosynovitis, and suppurative arthritis. It is more likely to occur in girls, often during menstruation.10 Suppurative arthritis most often involves the knee; the hand, wrist, ankle, elbow, and foot are involved less often. Infection of the shoulder or hip is uncommon.93 Skin lesions occur in approximately 40% of patients. Lesions typically are few in number and represent vasculitis. Lesions that occur on extremities or over affected joints are papular with a hemorrhagic component and evolve into vesiculopustular lesions on an erythematous base37; other skin lesions, including bullae and purpura, have been described.10 Culture of joint fluid is positive in only 25% to 35% of cases. Cultures of blood, cervix, urethra, rectum, vagina, skin lesions, or throat specimens may be positive when joint fluid culture is negative. Cultures obtained from normally sterile sites should be inoculated onto chocolate agar. Cultures from nonsterile sites should be inoculated immediately onto Thayer-Martin agar and incubated in carbon dioxide. N. gonorrhoeae also can be detected by polymerase chain reaction (PCR) or other DNA amplification procedures on firstvoided urine specimens and urethral and cervicovaginal swab samples. Because of the increasing prevalence of penicillin-resistant N. gonorrhoeae, 7 days of treatment with a parenterally administered third-generation cephalosporin, such as ceftriaxone or cefotaxime, is recommended.93–95 Marked improvement in fever and joint pain usually occurs 1 to 2 days after beginning therapy. Sequelae are rare.

Dosage Agent

mg/kg per day

Doses/day

Dicloxacillin Cephalexin Clindamycin

75–100 100–150 30

4 4 3–4

a Doses can be modified depending on results of serum bactericidal levels. In general, the oral dose of b-lactam antibiotics used for osteoarticular infections is two to three times the usual dose.

Polyarthritis, Fever, and Rash Bacterial causes of the clinical syndrome of fever, polyarthritis, and rash include infection with N. meningitidis and N. gonorrhoeae, rat bite fever (Streptobacillus moniliformis or Spirillum minus), bacterial endocarditis, and rheumatic fever96; multiple viruses also are considered (see below). Noninfectious causes include Kawaski disease, serum sickness, erythema multiforme and other autoinflammatory and autoimmune diseases.

486

SECTION

K Bone and Joint Infections

Lyme Arthritis

Rubella Virus

Lyme disease is caused by Borrelia burgdorferi, a tickborne spirochete.97 Arthralgia occurs early in the course of disease and is recurrent in 18% of individuals. Arthritis occurs in approximately one half of patients with Lyme disease,98 typically within 1 to 2 months after development of erythema migrans.99 Lyme arthritis is characterized by the sudden onset of monoarticular or oligoarticular joint pain. Joints involved, in descending order of frequency, include the knee, shoulder, elbow, temporomandibular joint, ankle, and wrist. Involvement of the hip or small joints is unusual, but reported.52 Patients characteristically are not ill, although about half will be febrile on presentation. Affected joints are warm and swollen, have large effusions but motion typically is not severely limited. The peripheral white blood cell count is often normal, but the ESR and CRP are usually elevated. Synovial fluid leukocyte counts range from 180 to 140,000/mm3; polymorphonuclear cells predominate.52,99 B. burgdorferi DNA has been detected by PCR of synovial fluid of patients with Lyme arthritis.100 Patients with Lyme arthritis are more likely to have MRI findings of myositis, lymphadenopathy and lack of subcutaneous edema compared with children with pyogenic arthritis (P < 0.01).101 Lyme arthritis is treated with amoxicillin or doxycycline, depending on age (see Chapter 186, Borrelia burgdorferi [Lyme Disease]). Duration of therapy usually is 4 weeks. Children with multiple recurrences or persistent arthritis sometimes require intravenous or intramuscular ceftriaxone for 14 to 28 days, or intravenous penicillin for 14 to 28 days. Prognosis in children is excellent.102 If untreated, recurrences of arthritis are common. Recurrences usually are separated by months to years. Frequency and duration of attacks decrease over time. Chronic synovitis develops in approximately 11% of untreated individuals.

Arthritis following rubella infection is uncommon in childhood but is reported in 30% of women and 15% of men. Arthritis usually develops 1 to 2 days after onset of rash, although it has been reported to precede the rash in a few cases. Symmetrical involvement of small joints of the hands is most common. Wrists and knees are sometimes affected. Analysis of joint fluid shows a predominance of mononuclear cells. Rubella virus has been isolated from synovial fluid. Symptoms resolve after several days, and long-term sequelae do not occur. Arthralgia and arthritis occur in approximately 25% of postpubertal females who receive live attenuated rubella vaccine. Joint symptoms begin 7 to 21 days after vaccination and are usually mild and self-limited.103,104

Viral Arthritis Arthritis as a result of viral infection can occur by direct viral invasion of the synovium or through immune complex deposition (Box 81-1). The viruses most commonly associated with the development of arthritis include rubella, parvovirus B19, certain arboviruses, and hepatitis B.

Parvovirus Symptoms of arthritis or arthralgia were reported in 80% of adults and 8% of children during an outbreak of erythema infectiosum.105 Arthritis caused by parvovirus B19 infection can occur in the absence of rash. Infection is often symmetric and polyarticular, and the joints of the hands, wrists, and knees are most commonly involved. Children are more likely to have asymmetric involvement of a few joints. Levels of total hemolytic complement are low in some individuals with parvovirus B19 arthritis, suggesting an immune complex-mediated pathogenesis. Arthritis associated with parvovirus B19 infection usually is self-limited; resolution of symptoms occurs within 1 to 2 months.

Hepatitis Viruses Arthralgia may occur as a prodromal symptom of infection with hepatitis A or B viruses, but arthritis occurs only with hepatitis B infection.106,107 Joint symptoms precede the onset of icterus by 1 or 2 weeks. Multiple small joints of the hands usually are involved. In addition, sometimes symmetric involvement of knees, elbows, ankles, and shoulders is seen. An urticarial or maculopapular rash, usually involving the lower extremities, appears simultaneously with the joint findings in 30% to 40% of patients with arthritis. Hepatitis C viral infection has been associated with development of polyarthralgia and polyarthritis in a few patients.108

Arboviruses BOX 81-1. Viruses that Cause Arthritis TOGAVIRIDAE Rubella virus Ross River virus Chikungunya virus O’nyong-nyong virus Mayaro virus Sindbis virus Barmah Forest virus PARVOVIRIDAE Parvovirus B19 PARAMYXOVIRIDAE Mumps virus PICORNAVIRIDAE Echovirus Coxsackie B virus RETROVIRIDAE Human immunodeficiency virus type 1 Human T-lymphotropic virus type 1 HERPESVIRIDAE Herpes simplex virus 1 Varicella-zoster virus Cytomegalovirus Epstein–Barr virus FLAVIVIRIDAE Hepatitis C virus HEPADNAVIRIDAE Hepatitis B virus

Several arboviruses found in Australia, Africa, Asia, and South America cause systemic illness in which arthritis is a predominant manifestation. They belong to the family Togaviridae and are members of the genus Alphavirus. All are transmitted by bites of mosquitoes or ticks. Epidemic polyarthritis caused by Ross River virus occurs most frequently in Australia.109 Clinical manifestations include fever, papular, petechial or morbilliform skin rash, adenopathy, and polyarthritis. Small joints of the hands and feet are affected most commonly. Most patients spontaneously recover within 2 weeks. Barmah Forest virus, also endemic to Australia, causes fever, polyarthritis, and rash.110 Chikungunya viral infection occurs in Africa, India, and Southeast Asia. Infection results in a biphasic illness, heralded by abrupt onset of fever, nausea, vomiting, and intense pain in one or more joints. The first phase of illness lasts 1 to 6 days, the patient becomes afebrile for 3 days, and then fever recurs. The second phase of illness is characterized by pharyngitis, rash, lymphadenopathy, and persistent arthritis. Children sometimes have febrile seizures and severe hemorrhagic manifestations. O’nyong-nyong virus (East Africa), Sindbis virus (Africa, Australia, Asia, Europe and the Middle East), and Mayaro virus (Central and South America) cause febrile illnesses that are characterized by rash, adenopathy, arthralgia, and arthritis.

Other Viruses Other viruses less commonly associated with arthralgia or arthritis include Epstein–Barr virus,111 enteroviruses (echovirus and human

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Infectious and Inflammatory Arthritis

coxsackie virus B),112 mumps virus,113 varicella zoster virus,114–117 cytomegalovirus, and herpes simplex virus.118 Persistent or intermittent arthralgia involving the knee or shoulder has been reported in 35% to 45% of adults with human immunodeficiency virus infection119,120 but in only 15% of children with this infection.121

Mycoplasma Species Mycoplasma pneumoniae, M. hominis, M. salivarium, and Ureaplasma urealyticum have been isolated from the joint fluid of patients with arthritis.122–127 Most patients are immunocompromised or have suffered trauma to the joint. Characteristically, onset is insidious, with minimal systemic signs and a mildly affected, boggy joint with relative preservation of movement.

CHAPTER

81

487

Amphotericin B followed by prolonged (total duration of therapy 6–12 months) treatment with fluconazole has been successful in certain circumstances.136,137 Candida spp. should be tested for susceptibility to fluconazole if this drug is used. Adequate debridement of the joint is necessary for successful therapy.

Sporothrix schenckii Sporothrix schenckii is a dimorphic fungus found worldwide in soil and decayed plant material. Individuals at risk for infection include those whose occupations place them in frequent contact with plant debris and moist soil. Infection in children is rare. S. schenkii causes both cutaneous and extracutaneous infection. Osteoarticular sporotrichosis is the most common manifestation of extracutaneous infection.138 Joints most frequently involved include the knee, ankle, wrist, and elbows. Itraconazole is the treatment of choice, with amphotericin b recommended as alternative therapy.139

Mycobacterium Species Skeletal tuberculosis occurs in 1% to 6% of all cases of tuberculosis. Isolated tuberculosis of the joint is uncommon.128 Articular infection can represent reactivated or primary infection. The knees and hips are affected most commonly, but infection of other joints may occur. Chronic swelling or pain of the affected joint without systemic symptoms is common. The Mantoux purified protein derivative test result is usually positive.129 Synovial fluid WBC count typically is between 10 000 and 20 000 cells/mm3, with polymorphonuclear leukocytes predominating. Synovial fluid cultures are positive in 79% of cases; synovial biopsy is diagnostic in more than 90%.130 Magnetic resonance imaging of articular tuberculosis shows joint effusion with high signal intensity on T2-weighted images and post contrast enhancement on T1-weighted images. Hypointense internal debris, synovial thickening and cartilage destruction may also be noted.131 Nontuberculous mycobacteria may cause osteoarticular infections in immunocompromised hosts.132,133

Aspergillus Species

Fungi

Cryptococcus neoformans

Fungal arthritis is unusual in healthy children, except in areas endemic for specific fungi. Chronic monoarticular arthritis is typical of most fungal joint infections. The diagnosis of arthritis usually requires microscopic evaluation of synovial biopsy specimens and culture of synovial tissue and fluid.

Cryptococcus neoformans is a yeastlike fungus found in soil contaminated by bird droppings. Infection typically involves the lungs, skin, or the central nervous system. Although bone lesions are found in 5% to 10% of cases, joint involvement is rare and usually secondary to infection in adjacent bone.142

Candida Species

Histoplasma capsulatum

Arthritis caused by Candida spp. occurs by hematogenous spread or, rarely, by direct inoculation of the organism into the joint space.134 Risk factors in neonates include prematurity, use of broad-spectrum antibiotics, intravenous alimentation, and presence of an intravascular catheter.87,91 Risk factors in older children include immunosuppression and intravenous drug use. Clinical manifestations of candidal arthritis vary. Children with disseminated disease have an acute onset of fever, systemic illness, and joint symptoms. In other cases, systemic symptoms are mild or absent. Joint symptoms may persist for months to years before a diagnosis is established. Neonates often have polyarticular involvement, but monoarticular infection is typical in older children. The knee is the most frequently affected joint. Arthritis caused by Candida spp. in intravenous drug users often occurs in fibrocartilaginous joints, such as the sacroiliac joint, costochondral joints, and intervertebral disks.135 Synovial fluid WBC counts range from 7500 to 150,000 leukocytes/mm3, with polymorphonuclear cells predominating. Diagnosis of fungal arthritis is confirmed by culture of synovial fluid or tissue. Culture of blood, urine, or cerebrospinal fluid may be positive in cases of systemic disease and especially in neonates.134

Histoplasma capsulatum is endemic to the central and southeastern United States, where large quantities of fungus are found in soil contaminated by bat or bird droppings. Infection is usually asymptomatic. Symptomatic infection is characterized by fever, chills, headache, cough, and chest pain. Approximately 10% of symptomatic patients have arthritis or severe arthralgia accompanied by erythema nodosum. Arthritis and arthralgia can be prolonged. Antifungal therapy is not always indicated in the immunocompetent host (see Chapter 251, Histoplasma Capsulatum [Histoplasmosis]).143,144 Nonsteroidal anti-inflammatory drugs are recommended for relief of joint pain.144

Aspergillus infection of the joint is uncommon and usually occurs secondary to extension of infection from adjacent bone. Children at risk include those with chronic granulomatous disease, chronic neutropenia, underlying cancer, and prolonged immunosuppression.140

Coccidioides immitis Coccidioides immitis is a fungus found in soil in the southwestern United States and northern Mexico. Infection usually is asymptomatic or associated with localized pulmonary disease. Extrapulmonary manifestations include cutaneous lesions, lymphadenopathy, central nervous system infection, and osteoarticular infection. Joint involvement is usually unifocal and often adjacent to sites of osteomyelitis.141

Blastomyces dermatitidis Blastomyces dermatitidis is commonly found east of the Mississippi River in warm moist soil containing decayed vegetation. Blastomycosis typically involves the lungs, skin, and genitourinary system. Skeletal disease occurs in 10% to 15% of cases.145 Arthritis usually results from extension of osteomyelitis from adjacent bone and is

488

SECTION

K Bone and Joint Infections

usually monoarticular but may be oligoarticular.146 Fungal organisms are identifiable on a wet preparation of synovial fluid.147 Itraconazole or other imidazoles have been used to treat blastomycoses.148 Other fungal organisms, such as Scedosporium spp., are rare causes of arthritis.149

isolated from synovial fluid but, in other cases, infection at another site is associated with reactive arthritis. In addition to being a response to a pathogen, reactive arthritis can occur in association with a more generalized inflammatory or immunologic disorder, such as Crohn disease, ulcerative colitis, rheumatoid and rheumatic disorders, Kawasaki disease, serum sickness, or Henoch–Schönlein purpura.

REACTIVE ARTHRITIS Reactive arthritis is defined as inflammation in one or more joints related to an infection at a site distant from the joint.150 Infections of the gastrointestinal, genitourinary, and respiratory tract are associated with reactive arthritis and an increasing number of pathogens are implicated.151–155 Children are less likely than adults to develop reactive arthritis after enteric infection. Immune complex associated arthritis may is seen in 2% to 16% of cases of meninogococcal disease.156 Organisms most commonly associated with reactive arthritis are listed in Box 81-2.150–152,154,155,157,158 A genetic susceptibility exists for development of reactive arthritis due to distant infection; individuals who are HLA-B27 antigenpositive have an increased incidence of disease.159 Reiter syndrome consists of arthritis, urethritis, and bilateral conjunctivitis. In children, symptoms usually follow a diarrheal illness.160 In adults, Reiter syndrome also can follow an episode of nongonococcal urethritis. Although Reiter syndrome and reactive arthritis are sometimes used interchangeably, reactive arthritis is diagnosed in many children who do not have the triad of symptoms. Reactive arthritis is usually polyarticular and involves the large joints of the lower extremities. Small joints, wrists, and elbows are involved less frequently. Sacroiliitis is more common among adults than children. Urethritis, if present, manifests with dysuria and pyuria. Mucous membrane ulcers (in the mouth, rectum, or vagina or on the glans penis) are sometimes present. Abnormalities of the eye include keratitis, uveitis, and corneal ulcerations. In children, joint symptoms persist for 1 to 12 months and recurrences are rare. The long-term prognosis of the disease is unknown. WBC count and ESR usually are elevated. Synovial fluid WBC count is less than 50 000 cells/mm3, with a predominance of polymorphonuclear cells.161 ESR can range from 20 mm/hour to more than 100 mm/hour. Responsible pathogens are sometimes cultured from stool or urethral specimens. Nonsteroidal anti-inflammatory agents are useful in controlling symptoms. Antibiotic treatment of the predisposing bacterial organism may be appropriate when cultures are positive at the time of onset of joint symptoms.158,162 Some bacteria cause both direct infection of the joint and reactive arthritis. For example, S. pyogenes causes infective pyogenic arthritis and is also associated with postinfectious reactive arthritis and rheumatic fever.163 Poststreptococcal reactive arthritis typically occurs 3 to 14 days after streptococcal infection, and is differentiated from the arthritis of acute rheumatic fever in that it is generally symmetric, may involve both large and small joints and is nonmigratory.164 Similarly, N. meningitidis, N. gonorrhoeae, and Salmonella spp. sometimes are

BOX 81–2. Bacteria Associated with Reactive Arthritis GASTROINTESTINAL PATHOGENS Shigella spp. Salmonella spp. Yersinia enterocolitica Campylobacter spp. Clostridium difficile SEXUALLY TRANSMITTED PATHOGEN Chlamydia trachomatis PYOGENIC AND REACTIVE ARTHRITIS-CAUSING ORGANISMS Streptococcus pyogenes Neisseria gonorrhoeae Neisseria meningitidis

CHAPTER

82

Diskitis Kathleen M. Gutierrez

Diskitis is an inflammatory process involving the intervertebral disks and the endplates of the vertebral bodies, and associated with characteristic clinical and radiologic findings. Our understanding of diskitis is primarily derived from retrospective studies and case reports of small numbers of patients. Diskitis has been reported under a variety of other names, including spondylodiskitis, pyogenic infectious spondylitis, spondylarthritis, acute osteitis of the spine, intervertebral disk space infection, and benign osteomyelitis of the spine.1–6 The definition of diskitis varies among studies, and it is often difficult to distinguish disk space inflammation alone from vertebral osteomyelitis. In fact, disk space inflammation is likely part of a spectrum of disease that includes vertebral osteomyelitis; however, diskitis appears to have a more benign clinical course.7,8 There has been no consistent approach to the diagnosis, treatment, and long-term follow-up of patients with diskitis. Thus, knowledge of incidence, causes, and optimal treatment of this infection is limited.

EPIDEMIOLOGY AND ETIOLOGY Although the incidence of diskitis is unknown, it appears to be uncommon. In one center, diskitis was reported to occur at a rate of approximately 1 to 2 per 30,000 clinic visits per year.1 Most cases occur in children age 6 years and younger.1,9–13 Cases have been reported in older children and adolescents. Disk space infection in adults most commonly occurs in the postoperative setting. Spontaneous diskitis is uncommon in adults,14,15 but when reported it has been associated with older age, diabetes mellitus, and systemic infection.16 Diskitis is probably the result of low-grade bacterial infection of the disk space.2–4,17,18 Some investigators believe that intervertebral disk inflammation is noninfectious and the result of antecedent trauma to the spine.19 Most blood cultures obtained from patients with diskitis are negative. If positive, Staphylococcus aureus is the most common isolate (Table 82-1). Other organisms isolated from either blood culture or disk aspirate from affected children have included S. epidermidis,20 Kingella kingae,6,14,21–23 anaerobes,24 gram-negative enteric organisms,20,25,26 and Streptococcus pneumoniae.2 In most cases, cultures of intervertebral disk specimens are sterile (Table 82-2). Viruses have not been isolated from disk space cultures. Mycobacterium species and Candida species have been isolated by disk space aspiration from older patients.27

PATHOGENESIS AND PATHOLOGIC FINDINGS The lumbar or lower thoracic spine is involved in most cases.28 There are rare reports of diskitis involving the cervical spine.3,25 Usually only one disk space is involved, although patients with involvement of two intervertebral disk spaces have been reported.13

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Diskitis

TABLE 82-1. Results of Blood Cultures in Children with Diskitis

CHAPTER

82

489

TABLE 82-3. Frequency of Presenting Symptoms in Patients with Diskitis (n = 165)

Blood Cultures No. of Patients

No.

20 41 29 36 18 11

24 22 9 28 5 11

0 10a 0 2b 0 0

Author

No. of Positive Results

13

Smith & Taylor (1967) Wenger et al. (1978)3 Scoles & Quinn (1982)25 Crawford et al. (1991)41 Ryoppy et al. (1993)12 Brown et al. (2001)42

Symptoms Limp, leg pain, or refusal to walk Back pain Abdominal pain

% a

41 41 15

a Children whose presenting symptom was refusal to bear weight or walk were younger (≥ 5 years) at time of presentation. Data from references 2, 3, 20, 25, 36, 41.

a

Staphylococcus aureus (9); diphtheroids (1). Staphylococcus aureus (2).

b

some children with the characteristic clinical and radiologic findings of diskitis, histologic results are normal.2,12 TABLE 82-2. Results of Disk Space Aspirates in Children with Diskitis Author

No. of Patients

No. of Biopsies/ Aspirates

No. of Positive Culture Results

Smith & Taylor (1967)13 Spiegel et al. (1972)2 Wenger et al. (1978)3 Scoles & Quinn (1982)25 duLac et al. (1990)9 Crawford et al. (1991)41 Ryoppy et al. (1993)12 Brown et al. (2001)42 Garron et al. (2002)6

20 48 41 29 12 36 18 11 42

3 15 9 6 8 3 17 3 35

1/3a 5/15b 6/9c 2/6d 6/8e 0/3 0/17 0/3 22/35f

a

Staphylococcus aureus (1). Moraxella spp., Staphylococcus aureus, Streptococcus pneumoniae, diphtheroids, and micrococcus (1 each). c Staphylococcus aureus (6; 1 also a-hemolytic streptococcus). d Klebsiella spp., Staphylococcus aureus (1 each). e Moraxella spp. (1), Staphylococcus aureus (5). f Staphylococcus aureus (12), Kingella kingae (6), Staphylococcus epidermidis (1), Clostridium clostridiiforme (1), Streptococcus sp. (1), Coxiella burnetii (1). b

The difference in blood supply to the vertebral bodies and disk spaces in children compared with adults may explain the age-related incidence of diskitis.29–32 In the young child, there are widespread anastomoses between intraosseous arteries supplying the vertebrae. These vessels begin to involute at about 8 months of age and are few in number by age 7. By the third decade of life, the anastomoses have fully atrophied and peripheral periosteal arteries develop. As a result of this rich blood supply to the vertebral endplates, a septic embolus in a child leads only to small areas of vertebral endplate infarction or infection, with disproportionate involvement of the intervertebral disk. By contrast, a septic embolus occurring in an adult results in a much larger area of infarction and subsequent infection of the vertebral body, which leads to vertebral osteomyelitis.1,30 The intervertebral disk is composed of the cartilaginous plate, the annulus fibrosus, and the nucleus pulposus.29 Examinations of cadavers have demonstrated that blood vessels are present in the cartilaginous endplates until age 7 and in the annulus fibrosus until age 20.31 This may offer another explanation for the finding of disk space inflammation or infection with relative sparing of the vertebral body that is seen in children. Magnetic resonance imaging (MRI) findings in children with diskitis suggest a pathophysiologic sequence in which infection or inflammation begins in the metaphyseal bone near the vertebral endplates, with anterior spread to the disk region and the adjacent vertebral endplate.5,33 An ovine animal model of diskitis has been developed. This model demonstrates that infection of the disk impedes disk development but does not seem to affect vertebral body growth.34 Histologic examination of disk biopsy specimens from children with diskitis reveals subacute or chronic nonspecific inflammation. In

CLINICAL MANIFESTATIONS The clinical manifestations of diskitis vary, depending on the age of the child.2,3, 9,11,13,25 Onset is gradual, with symptoms often present for several days to weeks before coming to medical attention. Younger children are more likely to have a history of irritability and reluctance to walk or bear weight. Older children complain of back pain, hip pain, abdominal pain, or pain with walking. Some patients have abdominal complaints, such as anorexia, vomiting, abdominal pain, and constipation (Table 82-3). On physical examination, the child frequently has a low-grade fever and appears well. Comfortable when lying still, the child refuses to bear weight or walks with a limp. Irritability with sitting or with flexion of hips and pain with palpation over the lower back are common. Spasm of paraspinous muscles may occur. Alterations in the normal curvature of the spine are sometimes noted, with attenuation or loss of the normal lumbar lordosis.10,20,35,36 Gower sign (use of hand “push-off ” rather than pelvic girdle to rise from sitting) may be present early in children with diskitis and disappears following treatment.37,38 Abnormal neurologic findings are not commonly described with diskitis; however, one retrospective case series described decreased muscle tone, muscle weakness, and decreased tendon reflexes in 7 of 17 children diagnosed with diskitis.39 Therefore, neurologic abnormalities should not exclude a diagnosis of diskitis; however, when present, imaging studies are needed to exclude intraspinal involvement or an alternative diagnosis. Failure to consider a diagnosis of diskitis, to examine the lower back carefully, and to order the appropriate radiologic studies often results in delay in diagnosis and inappropriate use of tests.

DIFFERENTIAL DIAGNOSIS The differential diagnosis of diskitis includes vertebral osteomyelitis or osteomyelitis of the pelvis, septic arthritis of the hip or sacroiliac joint, abscess of psoas muscle or pelvic structures, spinal epidural abscess, meningitis, appendicitis, malignant processes, pyelonephritis, and tuberculosis of the spine. If a child appears to be ill and has a high fever, marked leukocytosis, neurologic abnormalities, or extensive involvement of the vertebral body, diskitis is not the likely diagnosis.40 An immediate workup to rule out a more serious cause of symptoms should be initiated.

DIAGNOSIS The erythrocyte sedimentation rate (ESR) is almost always elevated (rarely more than 60 mm/h) in children with diskitis.10,11,25 Peripheral white blood cell (WBC) count is normal or slightly elevated.11,41 Creactive protein (CRP) may be elevated.42 Blood culture is obtained, but the result is often negative. Placement of a purified protein derivative (PPD) skin test and performance of chest radiograph are

490

SECTION

K Bone and Joint Infections

Figure 82-1. Diskitis involving the L4 to L5 space in an 18-monthold girl. Note the narrowed disk space and irregularities of the adjacent vertebral endplates. (Courtesy of Bruce Parker, M.D.)

useful in circumstances where tuberculosis infection is possible. When culture results are positive, therapy can be specifically targeted to the causative pathogen. Early in the course of diskitis, findings of plain film of the spine may be normal. Two to 4 weeks after onset of symptoms, the involved disk space is narrow and the margins of the vertebral endplates exhibit demineralization and irregularity (Figure 82-1). Two to 3 months after onset of symptoms, the disk space remains narrow and remineralization of vertebral endplates has occurred. Long-term follow-up in most patients shows persistence of a narrow disk space and sclerotic changes at the vertebral endplates (Figure 82-2). In some cases, fusion of involved vertebral bodies or changes in their shape have been noted.10,11,25,43,44 Other imaging modalities are useful if plain film results are normal. Bone scan has been used to make the diagnosis, because it frequently shows increased uptake of technetium 99m (99mTc) at the level of disk space involvement.35,41 Computed tomography (CT) can demonstrate narrowing of disk space and vertebral body involvement early in the course of disease.20 MRI is a sensitive technique for confirming the diagnosis of diskitis.5,9,33,36 MRI may be particularly useful in detecting widespread paravertebral abscess, epidural abscess, severe protrusion of the disk, or significant vertebral body involvement.42,45 Controversy exists over whether a child with the clinical and radiologic findings of diskitis should undergo aspiration of the disk space for culture. Some investigators have reported a good yield from culture material obtained from aspirate.9,17,18 Identification of an organism, especially an unusual one, such as K. kingae, may affect choice of therapy. In many studies, most aspirates from a disk space are sterile (even when antibiotics have not been administered).2,13,25,41 Therefore, aspiration of disk space may be limited to cases in which there is no response to immobilization or empirical antibiotic therapy, cases with extensive involvement of the vertebral body, or cases in which other features are atypical for diskitis.4,13,26,27,41 If disk space aspiration or biopsy of the disk is performed, care must be taken to send the specimen for appropriate culture for bacteria, mycobacteria, and fungus.

Figure 82-2. Follow-up films of the same patient as in Figure 82-1, 6 months after onset of symptoms. Some narrowing of disk space persists. Note sclerotic changes of the adjacent vertebral endplates. (Courtesy of Bruce Parker, M.D.)

MANAGEMENT Most children with diskitis respond promptly to bedrest, with decreased pain within 48 hours. Immobilization of the spine is sometimes required. Failure to respond to immobilization suggests that the diagnosis is incorrect. Optimal duration of immobilization is unknown. There has been no study of correlation between long-term outcome and duration of immobilization of the spine. Some investigators suggest the use of antibiotic therapy only if culture results are positive or the child appears to be systemically ill or has not responded promptly to immobilization.2,12,23,25,35,41 Although antibiotic therapy for diskitis does not appear to alter the long-term prognosis, one retrospective analysis of 47 patients showed that use of a short course of parenterally administered antibiotic agents, followed by oral therapy, hastened resolution of symptoms.5 Because a number of cases have been associated with positive culture results, particularly for Staphylococcus aureus, antibiotic treatment for an arbitrary length of time seems prudent.1,3–5 In the absence of a positive culture result, empirical use of an antibiotic agent with antistaphylococcal activity is appropriate. Antibiotic agents can be administered intravenously for several days, until fever and pain have resolved and the ESR is decreasing, and then given orally. Optimal duration of therapy is unknown, but it should be continued until the patient is asymptomatic and the ESR or CRP is normalizing.

OUTCOME Most children with diskitis have persistently abnormal radiographic findings long after clinical symptoms have resolved.12,33,43 The involved disk space may never regain its original height, and sclerotic changes in the vertebral endplates persist. There are reported cases of recurrence of symptoms.2 Although reports of follow-up of patients

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Transient Synovitis

with diskitis have included small numbers of patients and relatively short periods of time, most patients are asymptomatic with no limitation of activity.2,3,12,46 Younger children appear to have a lower rate of bony ankylosis compared to older children, presumably as a result of the presence of vascular channels which may promote more rapid healing.46 Some patients have mild, chronic back pain.12,35 A study of 35 patients who were followed for an average of 17 years found that 42% complained of persistent backache. Extension of the spine was restricted in 30 patients. Most patients had been treated with bedrest or with a lengthy course of antibiotic agents and immobilization.44

CHAPTER

83

Transient Synovitis Kathleen M. Gutierrez Transient synovitis is a self-limited, unilateral inflammation of the synovium, usually of the hip joint. It is a common cause of limp in childhood. Synonyms for this condition include toxic synovitis, “observation” hip, irritable hip, and “benign aseptic” arthritis.1–3

ETIOLOGY AND EPIDEMIOLOGY The cause of transient synovitis is unknown. In the few cases for which biopsy specimens of synovial membrane have been obtained, histologic examination shows a nonspecific inflammation.4,5 Considering that many children with transient synovitis have had a recent or concurrent upper respiratory or gastrointestinal viral infection and that elevated serum interferon levels have been demonstrated in some patients, transient synovitis is likely to represent either a selflimited infection involving the synovial membrane or a postinfectious inflammatory response.6–8 Some investigators have noted positive results on throat culture for group A streptococcus or elevated antistreptolysin O (ASO) titer, or both, in a small number of children with toxic synovitis.7,9 However, no single organism has been implicated consistently as the cause of the syndrome. In one series, an increase in antibody concentrations against rubella, enterovirus, Epstein–Barr virus, or mycoplasma was demonstrated in 67 of 80 children with transient synovitis, but there was no unaffected control group; results of viral culture for synovial fluid were negative; and culture samples were not obtained from other sites.10 Other reports, including one prospective study in which a control group was included and one study looking for serologic evidence of acute parvovirus B19 or human herpesvirus 6 infection, showed no evidence for viral or streptococcal causes of transient synovitis.4,11–14 Parvovirus B19 DNA has been detected by PCR in serum of a small number of children with transient synovitis.14a Transient synovitis occurs predominantly in children from 18 months to 12 years of age, with the mean age of occurrence being 5.6 to 5.9 years.3,11,13,15 Boys are affected approximately twice as often as girls.3,7,11,13–17 Although the annual incidence of transient synovitis in the United States is unknown, it is frequently referred to as the most common cause of hip pain in children.18 The average annual incidence in one study of Swedish children was 0.2%.17

CLINICAL MANIFESTATIONS The usual manifestation of transient synovitis is acute onset of leg pain or limp in an otherwise healthy child. Symptoms are unilateral, with

CHAPTER

83

491

right and left hips affected equally. There have been rare reports of children with involvement of both hips.3,11 Leg pain may be localized to the hip, thigh, or knee. The intensity of pain ranges from mild to severe enough to awaken the child at night.2,17 Most children with transient synovitis are afebrile or have minimal temperature elevation.19,20 The child appears minimally ill. There may be pain with movement of the knee, thigh, or hip on the affected side. Internal rotation and adduction of the affected hip are usually limited.3,7 Duration of symptoms ranges from 1 day to 3 weeks. Most children become asymptomatic approximately 1 week after onset of symptoms.12,13,21 Prolonged hip pain suggests a diagnosis other than transient synovitis.

DIAGNOSIS Laboratory and radiologic studies are helpful only to distinguish transient synovitis from more serious causes of hip pain. The white blood cell (WBC) count is usually normal, and the erythrocyte sedimentation rate (ESR) is normal or mildly elevated (no greater than 30 mm/h). The mean C-reactive protein (CRP) level in one study of 64 patients with transient synovitis was 1.0 mg/dL.22 Nonetheless, there may be considerable overlap in the characteristic WBC count, ESR, and CRP level between pyogenic (bacterial) arthritis and transient synovitis.20,22 Blood and joint fluid cultures are helpful because a positive culture rules out transient synovitis. Plain radiographs of the pelvis and hip of children with transient synovitis are often normal;16,17,23 films should be obtained to rule out osteomyelitis, cancer, or Legg–Calvé–Perthes (LCP) disease.9,23,24 Ultrasound is a useful noninvasive method for delineating the presence of joint effusion,21,25–28 which is found in approximately 70% of children with transient synovitis.11 Ultrasonography is helpful in guiding joint aspiration, particularly in children < 8 years, because it shows cartilaginous structures well; in older children fluoroscopic guidance is preferred.24 Bone scan may be normal or show increased uptake, especially in the blood pool phase.29–31 Magnetic resonance imaging (MRI) results are normal or reveal abnormal thickening of the articular surfaces and joint effusion.32,33 Signal intensity alterations detected by MRI in bone marrow may differentiate pyogenic arthritis and transient synovitis. Low-signal-intensity alterations on fatsuppressed T1-weighted images and high-signal-intensity alterations on fat-suppressed T2-weighted images are seen in bone marrow of the affected joint in many of patients with septic arthritis and not seen in bone marrow of patients with transient synovitis.34,34a

DIFFERENTIAL DIAGNOSIS Transient synovitis is a diagnosis of exclusion.8,17 The differential diagnosis includes bacterial arthritis of the hip; osteomyelitis of the proximal femur, acetabulum, or sacroiliac joint; pyomyositis, psoas muscle or intra-abdominal abscess; trauma; pauciarticular arthritis; rheumatic fever; malignant tumor; slipped capital femoral epiphysis; LCP disease; fracture, or reactive arthritis. Early differentiation between transient synovitis and bacterial arthritis of the hip is crucial, because delay in diagnosis and surgical decompression for pyogenic arthritis results in increased morbidity. In contrast to those with transient synovitis, children with pyogenic arthritis usually have a history of high fever as well as elevated ESR and WBC count.19,20 An evidence-based clinical prediction algorithm using 4 independent multivariate predictors, including fever, non-weight-bearing, an ESR of > 40 mm/h, and a peripheral WBC count of > 12,000 cells/mm3, was proposed by one group of investigators to estimate the probability of pyogenic arthritis. In a retrospective analysis of immunocompetent children, the predicted probability of pyogenic arthritis was reported as less than 0.2% if no predictors were present, 3.0% for 1 predictor, 40% for 2 predictors, 93.1% for 3 predictors, and

492

SECTION

L

Eye Infections

99.6% for 4 predictors.35 When the same 4 independent predictors of septic arthritis were used prospectively by the same investigators on a new cohort of children, the performance was somewhat diminished compared to the previously published prediction rule.36 In the new cohort, the area under the receiver operating characteristic curve was 0.86 compared to 0.96 in the original study. Another group of investigators identified 5 independent clinical and laboratory markers that were useful in differentiating pyogenic arthritis from transient synovitis. These included a temperature > 37°C, ESR > 20 mm/h, CRP > 1.0 mg/dL, WBC > 11,000 cells/ mm3, and a difference of joint space distance > 2 mm when one hip was compared with the other on plain radiographs. Patients with 4 or 5 predictors had a high predictive probability (between 90.1 and 99.1%) of pyogenic arthritis and those with 0 or 1 predictor had a low predictive probability of pyogenic arthritis (between 0.1 and 1.7%).14a,37 A recent prospective study confirms utility of including CRP in predictive alorithms to differentiate transient synovitis from pyogenic arthritis.37a. Other investigators have concluded that the validity of clinical prediction algorithms may vary substantially from institution to institution.38 Aspiration of fluid from the joint space is frequently attempted in patients with suspected synovitis, because this is the most reliable method of excluding the diagnosis of pyogenic arthritis. The procedure also relieves pain. With transient synovitis, fluid is expected to be clear or slightly turbid, usually less than 5 mL in volume, and to contain less than 1000 WBC/mm3. LCP disease frequently manifests as a limp in an otherwise healthy 5- to 9-year-old afebrile child who has a normal peripheral WBC count and ESR. Boys are affected four times as often as girls. Initially, the clinical features, laboratory test results, and ultrasonographic find-

SECTION

ings in a child with LCP disease may be indistinguishable from findings in transient synovitis. Plain films, bone scan, and MRI are all useful in differentiating LCP disease from transient synovitis; LCP disease shows ischemic necrosis of the head of the femur.17,24,30,32,39, In addition, in children with LCP disease, bone age is often delayed from 6 months to 3 years.2

TREATMENT AND PROGNOSIS The treatment of transient synovitis of the hip is bedrest. The usual duration of symptoms ranges from 5 to 7 days. One small double-blind placebo-controlled trial of ibuprofen versus placebo demonstrated that ibuprofen shortened the duration of symptoms from 4.5 to 2 days in children with a clinical diagnosis of transient synovitis of the hip.40 Close follow-up is necessary to detect development of signs and symptoms of pyogenic arthritis or osteomyelitis, which is especially challenging if patients have received anti-inflammatory drugs. Children who are older than 6 years or who have protracted symptoms require careful follow-up to distinguish LCP disease from other disorders. Transient synovitis is a self-limited condition that resolves with supportive therapy. Symptoms recur in a small number of children.4,7,8 In some series, up to 30% of children who were followed for 1 to 30 years after an initial episode of transient synovitis developed asymptomatic overgrowth (coxa magna) of the femoral head.1,15 The long-term significance of this radiologic finding is unknown. Some investigators suggest that children with transient synovitis are at a higher risk for developing LCP disease,6,8,16 whereas others have found no relationship between the two entities other than the similarity in initial clinical presentation.13

L

Eye Infections

CHAPTER

84

Conjunctivitis in the Neonatal Period (Ophthalmia Neonatorum) Avery H. Weiss

Ophthalmia neonatorum is defined as conjunctivitis occurring within the first month of life. It is the most common eye disease of neonates, with an incidence ranging from 1.6% to 12.0%.1–3 High incidence is directly related to the prevalence of sexually transmitted diseases in adults. One hundred years ago, Neisseria gonorrhoeae was the most

common pathogen and a major cause of blindness among children. With the advent of Credé prophylaxis and changing trends in sexually transmitted diseases, Chlamydia trachomatis has become the most common pathogen (Table 84-1).4–14 In general, offspring of women who receive prenatal care have a lower incidence of infectious conjunctivitis.12–14 In the absence of maternal sexually transmitted disease, neonatal conjunctivitis is acquired postnatally and is caused by Staphylococcus aureus, Streptococcus pneumoniae, Haemophilus spp., and viridans streptococci. Viruses, with the exception of herpes simplex virus (HSV), are not important pathogens in this age group. Hospital-associated infection can occur in neonatal intensive care units due to Staphylococcus aureus, gram-negative bacilli (especially Pseudomonas aeruginosa), and adenovirus. Significant predictors of conjunctivitis include low birthweight, use of ventilator or nasal cannula, and continuous positive airway pressue.15 Ophthalmologic examination has been associated with nosocomial bacterial conjunctivitis,15 and most notably adenovirus conjunctivitis (sometimes with dissemination and fatal outcome).16

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Conjunctivitis in the Neonatal Period (Ophthalmia Neonatorum)

CHAPTER

84

493

TABLE 84-1. Prevalence of Chlamydia trachomatis, Neisseria gonorrhoeae, and Other Bacterial Pathogens as Causes of Neonatal Conjunctivitis Country of Study

Year

No. of Infants

Chlamydia trachomatis (%)

Neisseria gonorrhoeae (%)

Other Bacterial Pathogensa,b (%)

Reference

United States United States United Kingdom United States United States United States Kenya United Kingdom Sweden Belgium Kenya United States

1977 1979 1982 1984 1985 1986 1986 1986 1987 1987 1988 1993

302 100 42 61 90 100 149 73 107 42 169 109

28 13 1 8.2 44 46 13 51 12.2 10 11 2

15 0 0 0 0 0 43 1 0 0 5 0

22 39 33 31 NS 50 9 10 66 80 NS 44

4 5 2 6 7 8 9 10 11 12 13 14

NS, not stated. a Percentages represent approximations, because most studies show total number of isolates, and multiple isolates were recovered from individual patients. b Excludes Staphylococcus epidermidis, Corynebacterium spp., Propionibacterium acnes, and diphtheroids.

PATHOGENESIS Microbial pathogens can be transmitted to the eye by a variety of routes. Premature rupture of placental membranes allows for the retrograde spread of organisms to the fetal conjunctiva and cornea. Hematogenous spread can occur transplacentally as well. During vaginal delivery, the neonate’s eyes can become infected by contact with infected maternal genital secretions. After birth, caregivers can transmit pathogens to the neonate’s eyes through direct contact or aerosolization. In addition to the immaturity of the neonate’s immune system, reduced tear secretion and lysozyme activity enhance susceptibility to ocular infections. The secretory rate of tears under basal conditions in premature infants is about 20% of that in full-term infants,17 and the blink frequency is decreased. The concentration of lysozyme, the enzyme in tears that catalyzes the breakdown of the bacterial cell wall, is also diminished in premature infants compared with full-term infants and adults.18 Furthermore, a neonate’s tears lack immunoglobulin (Ig) A.19

ETIOLOGIC AGENTS

Chlamydia trachomatis Epidemiology C. trachomatis is the most common cause of infective neonatal conjunctivitis (see Table 84-1). Frequency of conjunctivitis reflects a high prevalence of maternal genital infections, ranging from 18% to 23%.20,21 The likelihood of transmission from untreated infected mothers to infants ranges from 18% to 61% (Table 84-2).22–28

Clinical Manifestations Chlamydial conjunctivitis is usually clinically evident within 5 to 14 days of birth, although it may appear as early as 3 days, especially after premature rupture of membranes, or as late as 60 days.22,29 Typical findings are eyelid swelling, erythema, and unilateral or bilateral mucopurulent conjunctivitis; the cornea is not usually involved. On occasion, pseudomembranes and (rarely) true membranes develop. Because neonates lack lymphoid tissue, the expected follicular response of the conjunctiva does not appear unless the infection persists beyond 6 weeks of age. In treated cases, healing usually occurs without sequelae. Infection in untreated or inadequately treated cases can persist for 2 to 12 months, and some cases remain clinically apparent for years. Persistent infection can lead to conjunctival scar formation and corneal micropannus.30,31

TABLE 84-2. Prevalence of Chlamydial Conjunctivitis Among Infants Born to Mothers with Chlamydial Cervicitis Country of Study

No. of Infants

Neonates Infected (%)

Reference

United States United States Sweden United States United States United States United States Kenya

18 18 23 60 95 131 120 201

44 61 22 20 28 18 25 31

22 23 24 25 20 27 28 13

The major nonocular complication of chlamydial conjunctivitis is pneumonia. Its reported incidence ranges from 11% to 20% of infected infants and it is typically manifested between 1 and 3 months of age.29,32,33 Infants with chlamydial pneumonia are often afebrile but have nasal congestion, prolonged cough, tachypnea, and rales. Hyperinflation with interstitial or alveolar infiltrates is evident radiographically. Total serum IgG and IgM antibody values are elevated, and eosinophilia sometimes occurs. Antibodies to C. trachomatis are detectable in tears and serum of infants with chlamydial pneumonia. Although the infection is usually self-limited, treatment is recommended because it shortens the duration of illness and because systemic treatment prevents pneumonia.33–35

Diagnosis In the past, C. trachomatis conjunctivitis was diagnosed by Giemsa staining of conjunctival scrapings; the presence of blue-stained intracytoplasmic inclusions within epithelial cells is diagnostic. The sensitivity of Giemsa staining ranges widely, from 22% to 95%, reflecting varied technical and examiner skill.5,6,36 Cultivation of organisms in yolk sac or tissue culture is a sensitive means of confirming the diagnosis, but these techniques are laborious and slow. Detection of C. trachomatis antigen by direct fluorescent staining of elementary bodies with monoclonal antibodies, enzyme immunoassay, or DNA probes is sensitive and rapid but these tests are difficult to standardize.8,37–41 Nucleic acid amplification tests are more sensitive than these antigen tests and have become the new “gold standard.” 42 To be adequate for these techniques, the specimen must contain conjunctival cells, not exudate alone.

Treatment Systemic therapy for conjunctivitis due to C. trachomatis is recommended. Erythromycin estolate or ethylsuccinate, 50 mg/kg per

494

SECTION

L

Eye Infections

day in four divided doses for 14 days, is curative in most cases.34 Preliminary studies suggest that a short course of azithromycin (20 mg/kg), given as a single dose or daily for 3 days, is also effective, and may be less associated with idiopathic hypertrophic pyloric stenosis than erythromycin.25 Topical erythromycin may be clinically effective against chlamydial conjunctivitis, but this therapy does not eradicate nasopharyngeal colonization or prevent subsequent pneumonitis.26 Topical corticosteroid therapy is contraindicated because it can prolong the infection and possibly lead to conjunctival scarring.

Neisseria gonorrhoeae Epidemiology In the past, N. gonorrhoeae was the most serious cause of ophthalmia neonatorum because of this organism’s capacity to damage the eye; this infection accounted for 24% of children enrolled in schools for the blind between 1906 and 1911.43 After the widespread use of perinatal ocular prophylaxis, the incidence of gonococcal ophthalmia neonatorum decreased, and the proportion of admissions to schools for the blind as a result of this infection dropped to 0.6% by 1959. In the United States, the emergence of penicillin-resistant gonococci and the increase in frequency of asymptomatic genital gonococcal infection resulted in a higher prevalence of infections.44,45 In developing countries, the estimated incidence of gonococcal ophthalmia neonatorum is much higher than in the United States, ranging from 15% to 34%. In Kenya, where ocular prophylaxis was discontinued in the mid-1970s and the prevalence of penicillinase-producing N. gonorrhoeae has risen to 60%, gonococcal ophthalmia has reached epidemic proportions.9

Clinical Manifestations The clinical manifestations of gonococcal ophthalmia neonatorum typically appear 2 to 7 days after birth, although initial manifestations in the second week are not uncommon (Table 84-3). Later onset suggests postnatal exposure to the organism. Most neonates have edema of the eyelid and purulent conjunctivitis; however, some have a mild or catarrhal response.46 The inflammatory response of the external ocular surface in infants tends to be less severe than that in the older child or adult. Even with treatment, exudative conjunctivitis can continue for 7 to 14 days before subsiding, contributing to the common complication of conjunctival scarring. Corneal involvement is the most serious complication, because ulceration with resultant scarring is the major cause of visual impairment.47 Initially, superficial keratitis gives the corneal surface a lackluster appearance. Subsequently, marginal and, less frequently, central infiltrates appear, which then ulcerate, sometimes forming a ring abscess. Corneal perforation with loss of the anterior chamber allows abnormal apposition of the iris with the posterior cornea, which can persist as an adherent leukoma. The ulcerated or perforated cornea can also be the portal for invasion by other organisms, resulting in endophthalmitis. With widespread use of antibiotic therapy to treat gonococcal ophthalmia neonatorum, such ocular complications are rare.

TABLE 84-3. Differential Diagnosis of Neonatal Conjunctivitis Cause of Conjunctivitis

Time of Onset

Discharge

Chemical Neisseria gonorrhoeae Chlamydia trachomatis Herpes simplex virus

24 hours 2–7 days 5–14 days 6–14 days

Serous Mucopurulent Mucopurulent Serous

Corneal Involvement No Infiltrate No Dendrite, epithelial ulcer

Extraocular gonococcal infection is more common than previously thought. Fransen and colleagues9 reported that N. gonorrhoeae was isolated from pharyngeal specimens in 15% of infants with gonococcal ophthalmia. Although gonococcal pneumonia is not well documented, pharyngeal colonization can lead to bacteremic spread to distant sites, especially joints.48,49

Diagnosis Gram stain of ocular exudate typically shows predominance of neutrophils with intracellular gram-negative diplococci. Moraxella catarrhalis and Neisseria meningitidis can be confused with N. gonorrhoeae on Gram stain.50–53 Diagnosis of gonococcal ophthalmia is confirmed by culture on chocolate agar or Thayer–Martin media incubated at 37°C in humidified carbon dioxide. Appropriate culture specimens should also be obtained from the mother. Blood and cerebrospinal fluid cultures (in an infant who is febrile or otherwise not behaving normally) are also obtained, as are tests for concurrent C. trachomatis infection.

Treatment Although ocular prophylaxis is effective in preventing gonococcal conjunctivitis, topical therapy is ineffective in the treatment of established infection and does not prevent disseminated disease. Neonates with gonococcal ophthalmia should be hospitalized and treated with systemic therapy. Because of the increased prevalence of penicillin resistance in N. gonorrhoeae, penicillin cannot be used as the initial therapy. The treatment of choice is ceftriaxone (25 to 50 mg/kg, not to exceed 125 mg, given once intravenously or intramuscularly).54 Ceftriaxone should not be given to neonates with hyperbilirubinemia; a single 100 mg dose of cefotaxime given intravenously or intramuscularly is an alternative.54 In addition, the eye should be irrigated with saline solution at first hourly and then every 2 to 3 hours until discharge is eliminated. Instillation of topical antibiotic should be avoided because it is unnecessary and can be sensitizing.

Herpes Simplex Virus Epidemiology HSV can infect and seriously damage the neonatal eye. Before antiviral therapy, many infants with isolated herpetic conjunctivitis or skin or other mucous membrane infection progressed to more serious infection.55,56 In 70% to 80% of cases, infants contract the infection when their abraded skin or mucosal surfaces come in contact with infected maternal genital secretions at delivery. Most neonatal infections are caused by HSV type 2, for which there is an estimated prevalence of nearly 25% among adults in the United States.57

Clinical Manifestations Ophthalmia neonatorum due to HSV typically becomes evident 6 to 14 days after birth, although infants born after prolonged rupture of placental membranes can have manifestations at birth.58 Conjunctivitis, which can be unilateral or bilateral, is associated with ipsilateral eyelid edema and serous discharge. In the absence of herpetic lesions of the cornea or skin, herpetic conjunctivitis is indistinguishable from ophthalmia from other causes. Depending on the duration and severity of infection, examination of the cornea can reveal superficial keratitis, epithelial dendrite, geographic ulcer, or disciform stromal keratitis.59 Although keratoconjunctivitis can be the sole manifestation of herpes infection in the neonate, it more commonly appears in association with other manifestations of systemic infections.55,56 Most corneal infections in the neonate are detected with the use of a speculum to hold the eyelids apart and a cobalt blue light to illuminate the corneal surface, stained with a fluorescent dye.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Conjunctivitis Beyond the Neonatal Period

CHAPTER

85

495

Diagnosis

PROPHYLAXIS

The diagnosis of HSV infection is confirmed by culture of specimens obtained from skin vesicles or conjunctival or corneal lesions.

In 1881, Credé introduced ocular prophylaxis, consisting of cleansing of the eyelids immediately after birth followed by instillation of 2% silver nitrate drops into the conjunctival sac. This simple treatment successfully prevented most cases of ophthalmia neonatorum due to N. gonorrhoeae. The combined results of 24 studies conducted in North America and Europe between 1930 and 1979 indicated that only 0.06% of infants receiving topical silver nitrate experienced gonococcal ophthalmia.67 In 1989, Hammerschlag and associates68 reported the same incidence among 12 431 infants born between 1986 and 1988. However, silver nitrate was limited by its ocular toxicity and failure to prevent all cases of gonococcal ophthalmia.69–71 Chemical conjunctivitis lasting up to 72 hours can occur after instillation of silver nitrate and corneal scarring after inadvertent use of a high concentration of silver nitrate has been reported. The introduction of single-use wax ampoules of silver nitrate, designed to prevent evaporative loss and to maintain the concentration of silver nitrate at 1%, has reduced the incidence and severity of these complications. Changes in the epidemiology of ophthalmia to a predominance of C. trachomatis in the early 1970s prompted consideration of alternate prophylactic strategies.72 Topical erythromycin and tetracycline were proposed because of their in vitro effects against C. trachomatis.24 Two United States studies have shown that, although silver nitrate, erythromycin, and tetracycline preparations are each effective prophylaxis against gonococcal ophthalmia, none prevents chlamydial ophthalmia.13,68 The American Academy of Pediatrics and the Centers for Disease Control and Prevention recommend the use of 1% silver nitrate solution, 1% tetracycline ointment, or 0.5% erythromycin ointment for prophylaxis against N. gonorrhoeae.54 Although this treatment is highly effective when administered properly and within 1 hour of birth, infants born to mothers who have gonococcal infection at parturition should also be treated with a single dose of ceftriaxone given intravenously or intramuscularly (25 to 50 mg/kg; maximum dose 125 mg).54 In developing countries a single application of a 2.5% solution of povidone-iodine may be considered for ocular prophylaxis because it is inexpensive and has been shown to be more effective and less toxic than erythromycin and silver nitrate.73 A second application of 2.5% povidone-iodine was found to be no more efficacious than a single application.74 Eradication of sexually transmitted diseases and treatment of genital infection in mothers before parturition are the best means of preventing infection in their offspring.75

Treatment All infections caused by HSV in the neonatal period are treated with systemically administered acyclovir (60 mg/kg per day in three divided doses intravenously for 14 days; duration may be longer if other sites are involved; see Chapter 204, Herpes Simplex Virus). The presence of keratoconjunctivitis requires the addition of topical therapy. Trifluridine (Viroptic) 1% drops are given every 2 to 3 hours for 1 week, and then the dosage is tapered. Alternative medications are 3% vidarabine ointment and idoxuridine, 0.5% ointment or 0.1% solution, each given five times per day. If treatment failure occurs with one medication, another topical regimen can be tried. Topical cycloplegics are indicated for relief of ciliary spasm if keratitis is present. Topical corticosteroid agents should be avoided, because they may augment corneal damage.

Other Infectious Causes The predominant bacterial pathogens recovered from young infants with nonchlamydial, nongonococcal conjunctivitis include Haemophilus influenzae, Streptococcus pneumoniae, and S. mitis (see Table 84-1).2–6,8–12,14 These organisms are frequently isolated from the nasopharyngeal passages of caregivers. Less commonly, various gramnegative organisms, such as Escherichia coli, Klebsiella pneumoniae, and Pseudomonas aeruginosa (especially in nosocomial settings), and Neisseria cinerea, can cause conjunctivitis. Collectively, these bacteria account for a sizable proportion of infants with ophthalmia. Isolation of these bacteria from conjunctival specimens is not a priori evidence of causation. Demonstration by Giemsa stain of neutrophils and predominance of single bacteria (especially with some organisms present intracellularly) supports a pathogenic role. Staphylococcal organisms have been implicated in ophthalmia neonatorum, but in most cases it is unclear whether the organisms resided on the conjunctiva or the eyelid (where they may be part of normal flora).60,61 Staphylococcus aureus can be acquired postnatally from the skin or nares of hospital personnel or from family members. Infection usually manifests as a mild catarrhal conjunctivitis with mucoid or mucopurulent discharge. Potential ocular complications include corneal infiltrates and ulcers that can perforate, resulting in endophthalmitis. Systemic complications include infection at contiguous or distant sites (sinusitis, pneumonia, osteomyelitis, septicemia) or toxin-mediated disease (scalded-skin syndrome).62 Localized staphylococcal conjunctivitis can be treated with topical erythromycin, bacitracin ointment, or gentamicin drops. Susceptibility tests should be performed as community-associated and hospitalassociated S. aureus can be multiply resistant. Pseudomonas aeruginosa is a rare but important cause of neonatal conjunctivitis in hospitalized infants. Infants predisposed to this infection tend to be premature, or neonates receiving ventilatory support and systemic antibiotic therapy. Conjunctivitis can be selflimited and associated with moderate discharge or can progress rapidly, resulting in corneal abscess and endophthalmitis followed by septicemia and death.63–65 The diagnosis should be suspected when Gram stain of exudate reveals slender gram-negative bacilli. Treatment is topical gentamicin or amikacin drops, initially instilled hourly. Some experts advocate the addition of subconjunctival amikacin because of its high intraocular penetration.66 Lack of response to therapy or progression to endophthalmitis or periocular infection requires systemic or intraocular antibiotic therapy or both (see Chapter 88, Endophthalmitis).

CHAPTER

85

Conjunctivitis Beyond the Neonatal Period Avery H. Weiss Conjunctivitis is the most common infectious disease of the eye in childhood. Clinically, it is useful to separate conjunctivitis into acute infection (abrupt onset, lasting less than 10 to 14 days) and chronic infection (insidious onset, often persisting for several weeks, months, or even years). The vast majority of acute conjunctivitis is self-limited and can be managed by generalist physicians. Evaluation and successful management of chronic conjunctivitis, however, usually require consultation with an ophthalmologist for specialized diagnostic techniques and to prevent eventual vision-threatening damage.

496

SECTION

L

Eye Infections

Organisms can be transmitted to the ocular surface in a number of ways, but direct contact with contaminated fingers is the most common. Bacterial pathogens are usually found in nasopharyngeal secretions as well. Viral pathogens have specific tissue tropism, some rarely affecting conjunctival mucosa (e.g., influenza, respiratory syncytial virus), with others having a proclivity for conjunctival mucosa (e.g., adenovirus, herpes simplex). Organisms are also inoculated into the conjunctiva by airborne droplets produced by coughing and sneezing and through contact with common-use medical instruments. Vector-borne ocular infections occur in some developing countries. For example, the filarial parasite Onchocerca volvulus is transmitted to the eye by the bite of an infested black fly. Conjunctivitis is a clinical diagnosis based on the presence of conjunctival hyperemia and ocular discharge. Typically, the palpebral conjunctiva is more inflamed than the bulbar conjunctiva, and the area surrounding the cornea (limbus) is spared. Thus, pulling down the lower lid and noting the predominant area of inflammation enables one to distinguish conjunctivitis from keratitis, uveitis, and other causes of red eye.1 Conjunctival hyperemia associated with Kawasaki disease or bacterial toxin-mediated syndrome is distinguished by predominant involvement of bulbar conjunctivae and lack of exudate. The concurrent presence of a serous or purulent discharge confirms the diagnosis of conjunctivitis. Scant amounts of conjunctival discharge are best confirmed by asking the patient about the presence of eyelid crusting upon awakening, because dried exudate accumulates along the lid margins during sleep.

infections of the conjunctivae.4,5 An outbreak of 698 cases of acute conjunctivitis on a college campus in 2002 was caused by an atypical, unencapsulated strain of S. pneumoniae that was identical to strains that had caused outbreaks two decades previously.6

Staphylococci Staphylococcus aureus and S. epidermidis are often recovered from “eye” cultures but are infrequently a cause of acute conjunctivitis.3 Studies in which the lids and conjunctivae are sampled separately show that staphylococci can be recovered in relatively large numbers from lids, whereas few organisms are recovered from conjunctivae. Although staphylococci commonly colonize the lids without causing disease, they can cause blepharoconjunctivitis, a chronic infection of the lid margins.

Gram-Negative Bacilli Gram-negative bacilli other than H. influenzae occasionally cause conjunctivitis. Special clinical circumstances are usually pertinent, such as: (1) exposure to broad-spectrum antibiotics that promote the emergence of gram-negative nasopharyngeal flora; (2) circumstances of poor hygiene in which children rub their eyes with fingers contaminated by feces or urine; and (3) prolonged hospitalization, especially in intensive care settings, where immobilization, exposure keratitis, and dragging of a tracheal suction catheter across the face raise the risk of direct inoculation.7

ACUTE CONJUNCTIVITIS

Neisseria gonorrhoeae

Bacteria and viruses cause acute conjunctivitis. An epidemiologic study of 99 children with conjunctivitis published in 1981 observed that 65% of cases of acute conjunctivitis were caused by bacteria and 20% of cases were caused by a virus.2 A similar study of 95 children published in 1993 confirmed that the majority of infections confirmed that the majority (78%) of infections were caused by bacteria.3 Differentiating features of bacterial conjunctivitis and viral conjunctivitis are shown in Table 85-1.

Neisseria gonorrhoeae rarely causes conjunctivitis but is important because it can cause corneal ulceration and blindness.8,9 The organism binds avidly to surface receptors on the conjunctivae and cornea, triggering the release of bacterial toxins and inflammatory celldegradative enzymes that damage the corneal epithelium and underlying collagenous stroma. Clinical hallmarks of infection are the onset of purulent conjunctivitis after an incubation period of less than 7 days and the presence of corneal opacification. Unusual cases of infection after incubation periods of up to 19 days and infection associated with minimal symptoms have been reported.10,11 Beyond the first year of life, gonococcal conjunctivitis is a result of sexual activity or abuse.12 N. gonorrhea can be isolated from the pharynx, rectum, or genital mucosa as well as conjunctivae.

Bacterial Causes

Haemophilus influenzae and Streptococcus pneumoniae Nontypable Haemophilus influenzae is the predominant organism isolated from infected conjunctivae. Streptococcus pneumoniae, other streptococcal species (particularly S. mitis), and Moraxella catarrhalis are the next most common bacterial isolates.2,3 Collectively these organisms are still responsible for 55% to 72% of acute bacterial TABLE 85-1. Clinical Findings in Acute Bacterial and Viral Conjunctivitis Clinical Finding

Bacterial Disease

Viral Disease

Bilateral disease at onset Conjunctival responsea Conjunctival discharge Conjunctival membrane Preauricular adenopathy Concurrent otitis media

50–74% Papillaryb or nonspecific Mucopurulent Late onset Nod 20–73%

35% Follicularc Watery Early onset Yes 10%

a

Conjunctival response refers to conjunctival appearance on slit-lamp examination. b Papillary response denotes focal area of inflamed conjunctiva centered on a blood vessel. c Follicular response represents focal accumulation of lymphocytes encircled by blood vessels. d Except granulomatous bacterial infection. Data from Gigliotti F, Williams WT, Hayden FG, et al. Etiology of acute conjunctivitis in children. J Pediatr 1981;98:531–536; Weiss A, Brinser JH, Nazar-Stewart V. Acute conjunctivitis in childhood. J Pediatr 1993;122:10–14.

Neisseria meningitidis Neisseria meningitidis, a rare cause of conjunctivitis, is important because it can be complicated by meningococcemia and meningitis.13–16 This relationship suggests that the conjunctivae or nasopharynx can be a portal of entry for potentially invasive organisms. Bilateral hyperacute conjunctivitis is usual. Conjunctival scrapings typically show the presence of gram-negative intracellular diplococci. In a report of 21 cases and a review of 63 previously reported cases, Barquet and colleagues15 noted that 18% of patients with conjunctivitis due to N. meningitidis experienced meningococcemia. Systemic infection was more common among those treated with topical antibiotics alone.

Haemophilus aegyptius Haemophilus aegyptius deserves special mention because it can cause a meningococcal-like illness (see Chapter 173, Other Haemophilus Species). In South America, this organism is the cause of Brazilian purpuric fever.17,18 Especially prominent in children younger than 10 years, this catastrophic illness typically begins as hyperacute conjunctivitis, which is followed by fever within 3 to 5 days. Disseminated purpura, hypotensive shock, and death ensue within 48 hours. Molecular studies show that all isolates causing Brazilian purpura fever are genetically related.19 Initially, all reported cases

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Conjunctivitis Beyond the Neonatal Period

came from São Paulo, Brazil, and the neighboring state of Parana, but new strains emerged in other regions, raising concern about the disease’s potential to spread worldwide.20 The genome of pathogenic H. aegyptius is larger than that of nonpathogenic strains. Frequent gene exchange between bacterial species has been shown to underlie this genetic addition.21,22

Viral Causes Viral conjunctivitis has an acute onset, spreading from one eye to the other within a week, and the inflammation lasts 4 days to 2 weeks, depending on severity. Conjunctival discharge is watery, and slit-lamp examination shows follicular hyperplasia of conjunctivae. Inflammatory membranes over the conjunctival surface can develop. Lid swelling can be minimal or marked. Invasion of the corneal epithelium (punctate keratitis) is associated with pain and photophobia. Superficial keratitis is usually transient but can evolve into an immunemediated stromal keratitis that reduces vision. Ipsilateral preauricular adenopathy is common. Acute follicular conjunctivitis can be caused by a number of viruses, with associated ocular and nonocular clinical manifestations (Table 85-2).

Adenovirus Pharyngoconjunctival fever is characterized by the concurrent presence of fever, pharyngitis, and conjunctivitis.23,24 It is caused by adenovirus (serotypes 3 and 7) and usually affects children younger than 10 years. Although direct contact with airborne droplets is the usual mode of transmission, prolonged fecal excretion of the virus may be responsible for epidemics associated with swimming pools.24 Epidemic keratoconjunctivitis is the most common ocular infection due to adenovirus (usually serotypes 8, 19, and 37) in older children.25,26 Unassociated with fever or pharyngitis, this conjunctivitis is often associated with corneal inflammation. The epithelium of the cornea and conjunctiva share a membrane cofactor protein (CD 46) that attaches to and promotes entry of adenovirus type 37.25 Diffuse punctate lesions of epithelial keratitis evolve over 7 to 10 days into circumscribed subepithelial opacities that can impair vision and result in local discomfort for weeks to months. Resolution occurs without scarring. In young children, adenovirus conjunctivitis often causes ocular adnexal inflammation, eyelid edema, and erythema simulating periorbital cellulitis; an inflammatory pseudomembranous or palpebral conjunctivitis is distinctive.27 Direct contact with infected individuals is the usual mode of transmission,28,29 but indirect spread by common-use instruments, particularly those in ophthalmologists’ offices, can also occur.30,31 Adenoviral conjunctivitis can be associated with an acute respiratory infection that can be mild or severe. Faden et

CHAPTER

85

497

al. reported an outbreak of adenovirus 30 disease in a neonatal nursery in which 6 infants with pre-existing respiratory disease expired and ophthalmologic procedures were the infectious source.32,33

Herpes Simplex Virus Both primary and recurrent herpes simplex virus (HSV) infections can result in unilateral follicular conjunctivitis or blepharoconjunctivitis.26 Single or multiple clusters of vesicles characteristically appear on the eyelid and progress through pustular and crusted stages before healing (Figure 85-1). When present, accompanying corneal ulcers manifest as marginal infiltrates, punctate epithelial defects, or classic dendrites.

Picornaviruses Acute hemorrhagic conjunctivitis is a highly contagious illness usually caused by the picornaviruses enterovirus 70 and coxsackievirus A24. It is characterized by the sudden onset of bilateral conjunctivitis associated with profuse watery discharge, lid edema, and fine, punctate epithelial keratitis.34–36 A prominent distinguishing feature is the presence of subconjunctival hemorrhage, which can be pinpoint or confluent. The disease lasts for 3 to 5 days and resolves without adverse ocular sequelae. However, this pandemic infection can be accompanied by neurologic involvement, in particular radiculomyelitis with extremity weakness, unilateral facial nerve palsy, or other cranial neuropathies.37,38

Figure 85-1. Herpetic skin lesions of the upper and lower eyelids associated with ipsilateral conjunctivitis.

TABLE 85-2. Diagnostic Features of Acute Follicular Conjunctivitis Clinical Syndrome

Etiologic Agent

Eyelid Lesions

Corneal Lesions

Nonocular Findings

Pharyngoconjunctival fever Epidemic keratoconjunctivitis

Adenoviruses 3 and 7 Adenoviruses 8, 19, and 37

None Lid swelling

Fever, pharyngitis None

Herpetic keratoconjunctivitis

Herpes simplex virus

Vesicles

Acute hemorrhagic conjunctivitis

Enterovirus 70, coxsackievirus A24

None

Punctate epithelial keratitis Early: epithelial keratitis Late: subepithelial opacities Punctate epithelial keratitis Dendritic keratitis Punctate epithelial keratitis

New Castle disease

New Castle disease virus

None

Punctate epithelial keratitis

Rubella, rubeola

Rubella and rubeola viruses

Skin exanthem

Punctate epithelial keratitis

None Neurologic sequelae Facial palsy Radiculomyelitis Usually occurs in poultry workers or veterinarians Fever, diffuse exanthem, cough, rhinorrhea; occipital, postauricular adenopathy (rubella); Koplik spots (rubeola)

Modified from Dawson CR, Sheppard JD. Follicular conjunctivitis. In: Tasman W, Jaeger EA (eds) Duane’s Clinical Ophthalmology, vol 4. Philadelphia, Lippincott-Raven, 1991, pp 1–26, with permission.

498

SECTION

L

Eye Infections

CONJUNCTIVITIS–OTITIS SYNDROME The frequent association of conjunctivitis with otitis media has received increased attention. Because 20% to 73% of children with conjunctivitis have otitis media, children presenting with conjunctivitis should have otoscopic examinations. Nontypable H. influenzae is the major pathogen implicated.39–42 In one study from France, 16% of children with acute otitis media had concurrent conjunctivitis; conjunctival exudates yielded growth of nontypable H. influenzae in 89% of 419 cases.42

Diagnosis Although conjunctivitis is a clinical diagnosis, the determination of its microbial cause requires laboratory support. Bacterial cultures are more useful than viral cultures, because their diagnostic yield is higher and their costs are lower. The combination of sterile culture of conjunctiva and low density of organisms from lid cultures is strong indirect evidence of a viral infection.3 Molecular detection of adenovirus with real-time, automated polymerase chain reaction assay is rapid and an acceptable alternative to viral cultures.43,44 However, when HSV is a diagnostic consideration, viral cultures should be performed. Conjunctival scraping can be helpful when performed properly3,45; specimens should include conjunctival epithelial cells to allow detection of intracellular pathogens. The conjunctival surface is gently scraped with a spatula after the instillation of topical anesthetic drops; swabbing of surface exudate with a cotton-tipped applicator is more likely to yield an amorphous mixture of fibrin and cellular debris than intact cells. Gram stain of the scraping provides preliminary information about the presence and relative quantity of bacterial pathogens. Giemsa stain helps distinguish bacterial infections from viral infections by revealing predominantly neutrophils in the former and lymphocytes in the latter.

Treatment Bacterial Disease Acute conjunctivitis is usually self-limited. Topical antibiotic treatment is recommended because it speeds eradication of the offending pathogen and shortens the duration of symptoms.46 In general, treatment is empiric because cultures and cytologic examinations are not performed routinely. Sodium sulfacetamide is inexpensive and is available as drops and ointment, but its bacteriostatic mechanism of action as well as low concentration in tears, and adverse reactions upon instillation (stinging, burning, hypersensitivity reactions), limit its use. Tear levels show minimal or limited activity at 30 minutes with the 15% solution, and at 2 hours with the 30% solution. Two antibiotic ointments, erythromycin and bacitracin-polymyxin, are commonly prescribed for the treatment of acute conjunctivitis. Erythromycin is inexpensive, well tolerated, and active against many pathogens. However, Haemophilus species and M. catarrhalis are resistant, as are increasing numbers of staphylococci and streptococci.4,6 The combination of bacitracin and polymyxin (Polysporin) provides a broader spectrum of activity and excellent cure rates.5,46 Disadvantages of both ointments are difficulty in administration to young children and objection to associated blurred vision in older children. A combination containing polymyxin B and trimethoprim (Polytrim) is another option. It is active against most organisms, as effective as other topical antibiotics, available in eyedrop form, and causes only local irritation with no serious side effects.47–49 Topical aminoglycoside agents alone (tobramycin, gentamicin) or in combination as neomycin, polymyxin B, and gramicidin (Neosporin) are widely used to treat conjunctivitis. Of these

antibiotics, neomycin has the narrowest bacterial spectrum and the highest allergic risk. Tobramycin has the broadest spectrum and is less toxic than gentamicin to the corneal and conjunctival epithelium. One major disadvantage of aminoglycosides is that streptococci are resistant. In view of the potential epithelial toxicity, suboptimal streptococcal activity, and the fear that emerging resistance will limit their systemic usefulness, aminoglycosides should be reserved for the treatment of infections caused by gram-negative enteric organisms. Topical fluoroquinolone agents have a broad spectrum of activity against gram-positive and gram-negative organisms. Ciprofloxacin, ofloxacin, and norfloxacin were first introduced and shown to be highly effective in the treatment of bacterial conjunctivitis in children.50–53 All three antibiotics have similar gram-negative spectra but ofloxacin has superior gram-positive coverage.54 Levofloxacin, a third-generation fluoroquinolone with expanded activity against grampositive organisms, was compared with ofloxacin and found to have the same clinical cure rate but higher microbial eradication rate.55 Moxifloxacin and gatifloxacin are fourth-generation fluoroquinolones with higher solubility, increased tissue penetration, lower toxicity, and broader gram-positive activity compared with other topical fluoroquinolones.56–57 Because their widespread use has led to the emergence of resistant organisms, however, topical fluoroquinolones should be restricted to the treatment of serious infections or infections due to resistant organisms.56,58,59 Systemic antibiotics are effective in the treatment of bacterial conjunctivitis. Intravenous antibiotics attain levels in the conjunctiva that are sufficient to eradicate potential pathogens. Efficacy of oral therapy is less clear, because the conjunctival levels of drugs are considerably lower. In one study, 46 of 48 (96%) children with acute conjunctivitis treated with oral antibiotics alone had negative bacterial cultures after 3 to 5 days of therapy.60 Oral therapy may have the advantage of preventing the development of otitis media. Harrison and associates40 reported that 11 of 42 (26%) patients treated with topical antibiotics alone, but only 2 of 41 (5%) patients treated with topical and systemic antibiotics, had evidence of otitis media at a 2-week follow-up visit. In comparison, Wald et al. reported that oral cefixime was no more effective than topical polymyxin-bacitracin in either the eradication of conjunctival colonization or prevention of acute otitis media.4 A controlled study by Isenberg et al.61 indicates that topical solution of 1.25% povidone-iodine is as effective as neomycinpolymyxin-gramicidin in the treatment of acute conjunctivitis. Low cost and lack of microbial resistance make this an attractive option, especially in developing countries.

Herpes Simplex Virus Treatment of HSV eye infections should usually be undertaken in collaboration with an ophthalmologist. Topical antiviral therapy is essential. A variety of DNA inhibitors (1% to 2% trifluridine, 3% vidarabine, 1% iododeoxyuridine) are useful when applied topically for superficial keratitis. Oral acyclovir is sometimes beneficial in more severe or recurrent episodes. Topical corticosteroid agents are contraindicated in suspected conjunctivitis and superficial keratitis due to HSV but are used by ophthalmologists on occasion, in conjunction with antiviral agents, for immune-mediated disease of the corneal stroma.

Adenovirus and Other Viruses Treatment is supportive with cool compresses, artificial tears, topical vasoconstrictors, and ointment alone or in combination. Topical ketorolac 0.5%, a nonsteroidal anti-inflammatory agent, used four times daily is no better than artificial tears.62 Topical cidofovir, a broad-spectrum antiviral agent, is not effective at 0.2% concentration, and the 1% concentration is limited by local toxicity.63,64 Corticosteroid drops are contraindicated because they may predispose to corneal involvement and prolong virus shedding in an animal model.65

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Conjunctivitis Beyond the Neonatal Period

CHRONIC CONJUNCTIVITIS Nasolacrimal Duct Obstruction Nasolacrimal duct obstruction is the most common cause of chronic or recurrent eye infection of infants.66 Persistence of an imperforate membrane along the nasolacrimal duct blocks tear drainage and predisposes to secondary infection of retained tears. Consequently, tears pool in the conjunctival cul-de-sac and then spill over the lid margins, flowing on to the cheek (epiphora). At night or during sleep, spillage of infected material leads to accumulation of crusted exudate along the lid margins. Digital compression of the lacrimal sac region usually elicits reflux of mucopurulent discharge. When reflux cannot be elicited, delayed clearance of 5% fluorescein dye instilled on to the eye confirms the diagnosis of a blocked tear duct. The conjunctiva appears normal or only minimally inflamed. Topical antibiotics and digital massage of the lacrimal sac are the initial treatments of choice. Because 95% of obstructions resolve by 6 months of age, lacrimal duct probing is reserved for patients in whom obstruction persists beyond this age.

CHAPTER

85

499

Treatment is continued for 1 month after all signs of inflammation abate. Topical corticosteroids are sometimes required to suppress the inflammatory component. Systemic erythromycin in children under the age of 8 years and systemic tetracycline in older children is an effective treatment option in severe cases of blepharoconjunctivitis.68,69 Erythromycin and tetracycline appear to accumulate in the meibomian glands where they inhibit bacterial protein synthesis and lipase breakdown of sebum.

Ulcerative and Angular Blepharitis Herpes simplex infections can cause ulcerative blepharitis. The presence of grouped vesicles is suggestive of herpetic infections. Diagnosis is confirmed by viral culture. Angular blepharitis refers to an infection localized to the lateral canthal region and, less frequently, to the medial canthal area. The involved area is erythematous, with eczematous changes, maceration, and fissuring of affected skin. It is most common among adolescents living in warm climates. S. aureus and S. pneumoniae are the most commonly implicated bacteria. Treatment is topical antibiotics.

Staphylococcal Blepharitis

Chlamydial Infections

Blepharitis is a primary infection of the eyelid with secondary inflammation of the conjunctiva. It is the most common cause of chronic conjunctivitis in older children.66,67 Staphylococcus aureus or S. epidermidis are the usual pathogens. Affected children come to medical attention because of red eyes, ocular irritation, chronic eye rubbing, and pulling of the eyelashes. Photophobia is a sign of corneal involvement. Clinically, it is useful to separate blepharitis into anterior and posterior types. Anterior blepharitis is accompanied by crusting along the lid margin because of the buildup of epithelial debris and fibrin centered on the eyelashes (cilial collarettes). Close inspection reveals superficial excoriations of the lid margin and telangiectatic vessels. Chronic rubbing of the eyelid and eyelash pulling lead to breakage and partial loss of eyelashes. In posterior blepharitis, the meibomian gland orifices, located along the posterior lid margin, are blocked, causing retention of sebum and predisposing to meibomian cysts, chalazion, and secondary infections. The incidence of meibomitis peaks in young children68 and during adolescence, when the sebaceous glands undergo hormonal stimulation. Examination reveals inspissation of meibomian glands, from which exudative material can be expressed by gentle compression of the lids with cotton applicators. Although blepharitis is a clinical diagnosis, the recovery of relatively more staphylococci from lid cultures than conjunctival cultures, as well as cytologic evidence of neutrophils on Giemsa stains of conjunctival scrapings, helps to confirm the diagnosis. The most common complication of blepharitis is chalazion formation. This results from obstruction of the meibomian gland orifices by debris and secondary infection. Blepharitis should be suspected in any child who demonstrates recurrent or multiple chalazia. Toxic epithelial changes of the cornea (punctate keratitis), immune-mediated subepithelial infiltrates, and marginal ulcerations can result from corneal contamination with exotoxins of staphylococci. Chronic staphylococcal keratitis can lead to corneal neovascularization and visual loss. Blepharitis cannot be cured, but it can be controlled symptomatically. The mainstay of treatment is lid hygiene and topical antibiotics. Lid hygiene entails applying a washcloth presoaked in warm water at least 2 to 4 times a day and scrubbing the lid margin with a cotton-tipped applicator presoaked in a 50:50 mixture of a nonirritating shampoo (or soap) and water. Attention to personal hygiene and avoidance of chronic eyelid rubbing are helpful in preventing contamination of the lids. Topical antibiotics should be used to treat secondary infection. Bacitracin or erythromycin ointments are appropriate choices. They should be applied one to four times daily for 1 to 2 weeks and then once daily for 4 to 8 weeks.

Trachoma Trachoma is the most important ocular infection worldwide because it is a major, preventable cause of blindness.70 The disease is caused by Chlamydia trachomatis, usually serotype A, B, or C. It has an insidious onset in infants and young children, in whom the organism attaches to and invades the conjunctival and corneal epithelium, eliciting the formation of lymphoid follicles. Chronic inflammation causes a loss of conjunctival goblet cells with resulting mucin deficiency, and subconjunctival fibrosis (particularly of the upper lid). When lymphoid follicles of the limbus regress, they leave semilunar areas of thinned cornea known as Herbert pits, which are pathognomonic of the disease. Progressive cicatrization of the conjunctiva can pull the lid margin inward, causing the eyelashes to rub on the cornea (trichiasis). The combined effects of chronic inflammation, mucin deficiency, and trichiasis can lead to corneal scarring, neovascularization, and blindness. Trachoma is treated with topical erythromycin or tetracycline, or sulfacetamide ointment applied either twice daily for 2 months or twice daily for the first 5 days of the month for 6 months. Oral erythromycin or tetracycline for 40 days is given if the infection is severe. Single-dose azithromycin (20 mg/kg) is also effective and can be beneficial, especially in hyperendemic areas.71–75 In one large study, the prevalence of infection fell from 9.5% before introduction of single-dose azithromycin therapy to 0.8% 24 months after introduction.75 On the basis of a chlamydial recurrence rate of 11% at 6 months, one study proposed that repeat single-dose treatment biannually could potentially eliminate ocular Chlamydia trachomatis.76

Inclusion Conjunctivitis of the Adolescent Chlamydia trachomatis (serotypes D through K) is the most common sexually transmitted disease with disproportionate occurrence in sexually active adolescents.77,78 Recovery of C. trachomatis from younger children should prompt concerns about sexual abuse. The clinical hallmark of eye involvement is a unilateral follicular conjunctivitis with mucopurulent discharge, eyelid swelling, and ipsilateral preauricular adenopathy. Photophobia and severe irritation are symptoms consistent with corneal involvement. The organism invades the corneal epithelium, resulting in a superficial keratitis that can progress to subepithelial infiltrates and micropannus formation. The disease can persist for months to years if left untreated. Diagnosis is based on positive cultures, detection of chlamydial antigens, or molecular techniques. Treatment consists of systemic tetracycline, doxycycline, or erythromycin for 7 days or longer.

500

SECTION

L

Eye Infections

come to medical attention because of a painful red eye. Distinguishing infectious keratitis from other causes of red eye accurately and promptly is extremely important.

PATHOGENESIS

Parinaud oculoglandular conjunctivitis is a syndrome characterized by the combined presence of granulomatous conjunctivitis and ipsilateral preauricular or submandibular adenopathy (Figure 85-2). Usually it is unilateral, and there may be one or more granulomas of the upper or lower palpebral conjunctiva. Fever, malaise, and other systemic signs can be present. Bartonella henselae is the most common cause,79 but differential diagnosis includes tularemia, sporotrichosis, tuberculosis, syphilis, and infectious mononucleosis.

The cornea is susceptible to microbial invasion because of its exposed position, avascularity, and limited inflammatory repertoire. Apart from epithelial Langerhans cells, the normal cornea lacks white cells, which must be recruited from the tear film, limbal lymphoid aggregates, and perilimbal circulation.1,2 The epithelium and its basement membrane serve as a relative barrier to most infectious agents, but, once breached, the hypocellular stroma is vulnerable to infection, often with organisms of low virulence. Release of proteases, collagenases, and oxygen-derived free radicals from invasive organisms and infiltrating neutrophils degrades collagen and proteoglycans, the major constituents of the corneal stroma. Trauma is the most common factor predisposing to bacterial and fungal keratitis (Box 86-1).3–6 Trauma from contact lenses can cause epithelial damage and corneal hypoxia. The risk of bacterial keratitis is 10- to 15-fold greater in subjects who wear contact lenses overnight compared with users of daily-wear lenses.7 Orthokeratology, a procedure for reshaping the cornea by wearing nocturnal contact lenses, is popular among teenagers and predisposes to corneal infections.8 Dry eyes related to a deficient tear film or abnormality of the ocular surface, corneal anesthesia, and prior treatment with topical corticosteroid agents are other factors increasing the risk of bacterial keratitis. In developing countries, malnutrition, vitamin A deficiency, poor hygiene, and trachoma also predispose to corneal infections.9 Inadequate eyelid closure is another important risk factor for infective keratitis. Corneal exposure in the severely ill or sedated child predisposes to breakdown of the epithelium and secondary infection.3 Children with eyelid defects, facial nerve weakness, and proptotic globes are at increased risk of corneal infections. Premature infants and older children on mechanical ventilators are particularly susceptible to conjunctival colonization with Pseudomonas aeruginosa and to the development of pseudomonal keratitis.10–12

Viruses

CLINICAL PRESENTATION

Viruses rarely cause chronic conjunctivitis, with the exception of molluscum contagiosum, a poxvirus infection. One or more umbilicated papules of varying size on the eyelid are the cardinal features. Rupture of lesions on the eyelid margin releases molluscum bodies on to the conjunctiva, inciting a follicular response.80 Molluscum contagiosum of the eyelids can be a manifestation of human immunodeficiency virus infection.81 Diagnosis is based on the presence of typical lesions on the lid and elsewhere on the skin. Treatment consists of excision, curettage, or cryopexy of individual lesions.82

Severe pain is the hallmark of infective keratitis. Reflex tearing, redness of the eye, photophobia, and decreased vision are also prominent symptoms. Keratitis is usually distinguished by the presence of grayish corneal opacification. Loss of the epithelium over the corneal

Figure 85-2. Eight-year-old girl with cat-scratch disease and Parinaud syndrome has granulomatous palpebral conjunctivitis and ipsilateral preauricular lymphadenopathy (at tip of examiner’s index finger). (Courtesy of M.C. Fisher, M.D., St. Christopher’s Hospital for Children, Philadelphia.)

Parinaud Oculoglandular Syndrome

CHAPTER

86

Infective Keratitis Avery H. Weiss

Infective keratitis is an uncommon but important infection because it has the potential to cause blindness and can be produced by a wide variety of microorganisms. Most children with corneal infections

BOX 86-1. Risk Factors for Bacterial Keratitis in Children TRAUMA Corneal foreign body Corneal abrasion/laceration Contact lens wear Trichiasis/dystichiasis Prior ocular or eyelid surgery CORNEAL EXPOSURE Congenital and acquired disorders of the eyelids Globe proptosis Facial palsy Moribund or sedated state ABNORMALITIES OF THE OCULAR SURFACE Dry-eye syndrome Mucin deficiency from loss of goblet cells Malnutrition Corneal anesthesia IMMUNODEFICIENCY STATES Topical corticosteroid therapy Immunosuppressive therapy Immune deficiency syndrome

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Infective Keratitis

CHAPTER

86

501

Figure 86-2. Slit-lamp appearance of recurrent dendritic keratitis due to herpes simplex virus stained with 1% fluorescein. Figure 86-1. Classic appearance of primary herpes simplex virus keratitis stained with 1% fluorescein, showing multiple epithelial dendrites.

infiltrate dulls the corneal light reflex and permits topically applied fluorescein dye to stain the area. Progressive destruction can lead to corneal thinning and eventual perforation. The anterior chamber may contain dispersed inflammatory cells or visible aggregates of neutrophils layering inferiorly (hypopyon). In an uncooperative child examination of the eye is facilitated by the use of a topical anesthetic and a lid speculum.

SPECIFIC CAUSES Herpes Simplex Virus Keratoconjunctivitis caused by herpes simplex virus (HSV) is usually mild and indistinguishable from other causes of viral conjunctivitis, except by the presence of skin or corneal lesions. The most common manifestations are unilateral follicular conjunctivitis, watery ocular discharge, and preauricular adenopathy.13 Inspection of the swollen lids often reveals the presence of vesicles or lid margin ulcerations. Initially, the cornea is spared, but within a few days to 2 weeks mild punctate keratitis or dendritic keratitis develops in about one-half of infected children (Figure 86-1). Primary herpetic keratitis is selflimited in most children beyond the neonatal period. Most episodes of herpetic keratitis represent recurrent disease after primary infection has led to latent infection of the cornea or trigeminal ganglion. Only 20% of recurrences appear during childhood; their number, frequency, and type are highly variable.13,14 The spectrum of recurrent HSV keratitis includes epithelial dendrites (Figure 86-2), larger geographic defects, immune-mediated disciform edema (Figure 86-3), necrotizing keratitis complicated by corneal thinning or perforation, and chronic neurotrophic ulcerations. There are few serious sequelae if infection remains confined to the corneal epithelium, but if stromal disease ensues, progressive corneal scarring can result in permanent visual loss.

Varicella-Zoster Virus Chickenpox (varicella-zoster virus; VZV) is sometimes associated with small vesicular or papular eruptions at the limbus. “Pox” usually resolve without sequelae, but the affected conjunctiva is often red and painful. Less frequent corneal manifestations include superficial punctate epithelial defects, linear dendrites, and disciform or necrotizing kerstitis with ulceration. Recurrent epithelial or stromal keratitis can also occur.13,15

Figure 86-3. Herpes simplex virus disciform keratitis, consisting of a disk-like area of stromal edema (asterisk) believed to be a cell-mediated immune response to residual herpes simplex virus antigen.

Herpes zoster ophthalmicus refers to reactivation of VZV along the sensory distribution of the ophthalmic division of the trigeminal nerve. Although uncommon in childhood, it may accompany immunosuppression or inadequate immunity when contracted in first year of life.16 A vesicular eruption is confined to the branches of the ophthalmic nerve. Uveitis and keratitis are the most common ocular manifestations. In the acute infectious stage, examination of the cornea can reveal punctate, dendritic, or geographic epithelial inflammation, which can progress to necrotizing stromal keratitis.13,17 An immunemediated response to zoster-related antigens appears as a delayedonset disciform or nummular (coin-shaped) keratitis. A chronic and recurrent epithelial or stromal keratitis can occur.18 Although herpes zoster infections in previously vaccinated children tend to be mild, individual children can have severe sclerokeratitis and anterior uveitis.19

Other Viruses Measles, mumps, rubella, adenovirus, coxsackievirus A24, and enterovirus 70 are commonly associated with a self-limited punctate epithelial keratitis; however, there are important exceptions. Epithelial

502

SECTION

L

Eye Infections

lesions associated with adenovirus infections sometimes progress over 10 to 14 days to subepithelial opacities that can persist for months.20 Measles keratitis is usually mild but can cause severe corneal disease when contracted by malnourished, immunocompromised, or vitamin A-deficient children.21,22 Localized edema of the corneal stroma can, on occasion, be a delayed complication of mumps.23 Epstein–Barr virus can invade all layers of the cornea. Superficially there can be multiple epithelial dendrites. Stromal disease is more common and is characterized by multiple coin-shaped lesions in the anterior and mid-stroma or recurrent infiltrate of the deeper layers associated with vascularization.24–26 Treatment is supportive.

Bacteria Bacterial infection of the cornea is considered a medical emergency because it can progress rapidly and lead to severe visual loss. The presence of a dense grayish infiltrate and surface ulceration in an actively inflamed eye should be considered bacterial infection until proven otherwise (Figure 86-4). Infection may be caused by a wide range of gram-positive and gram-negative organisms.3–6 Staphylococcus aureus and streptococcal species are among the most common gram-positive isolates, especially in cooler climates. S. epidermidis is frequently isolated from corneal cultures, but this often represents contamination from the eyelids. Pseudomonas aeruginosa is the most common gram-negative isolate, especially among contact lens wearers. Ulcerative keratitis can also complicate bacterial conjunctivitis due to Haemophilus influenzae and Moraxella catarrhalis. Shigella and other enteric pathogens can cause keratitis following transfer to the eyes by fingers contaminated from extraocular sites of infection.27

Fungi Fungal keratitis is rare in childhood and is usually the consequence of ocular trauma, especially with vegetable matter. The prior use of topical corticosteroid agents, systemic immunosuppression, preexisting corneal disease, and tropical environment also increase the risk of fungal infection. Fusarium species were isolated in a recent outbreak of corneal ulcers associated with the use of a specific contact lens solution, ReNu with MoistureLoc.27a Typically, the ulcer has a subacute onset and progresses insidiously. At slit-lamp examination the yellow-white infiltrate has feathery edges, a dry, raised surface, and satellite lesions.28,29 A wide range of fungi can be corneal

Figure 86-4. Perforated keratitis due to Streptococcus pneumoniae in a 3-year-old child after an eye injury.

pathogens. Aspergillus and Fusarium species are the most common pathogens worldwide.3–6,29,30 Within the United States, F. solani is the most common pathogen in southern states, whereas Aspergillus species and Candida albicans are most common elsewhere.31,32 Dematiaceous (pigmented) fungi, particularly Curvularia species, are emerging as important corneal pathogens.29

Acanthamoeba Acanthamoeba causes a recalcitrant keratitis that frequently leads to visual loss. It usually occurs in contact lens wearers or in persons exposed to contaminated water.33,34 Severe pain, out of proportion to the severity of the keratitis, is common. This seems to be related to a propensity of the organism to infiltrate corneal nerve endings.35 Initially, corneal epithelium is involved, and later a stromal infiltrate develops. As the infection progresses the infiltrate becomes densest at the periphery, giving rise to a very characteristic ring-shaped lesion.36–38

Interstitial and Marginal Keratitis Interstitial keratitis is usually a delayed manifestation of a systemic infection with a variety of bacterial, viral, and parasitic pathogens. Congenital syphilis was formerly the predominant cause of interstitial keratitis.39,40 More recently, Mycobacterium tuberculosis, nontuberculous mycobacteria, Borrelia burgdorferi, herpesviruses, and onchocerciasis have become more common causes. Interstitial keratitis represents an immune-mediated reaction to retained microbial antigens rather than active infection. Pain, photophobia, and reflex tearing are the major symptoms. Examination is characterized by an intact corneal epithelium with underlying cellular infiltration of the superficial or deep stroma. Limbal blood vessels often proliferate and extend into the opacities. Once the inflammation subsides, the opacities and blood vessels regress, leaving stromal scars and ghost vessels. Depending on the severity of scarring, final visual acuity ranges from 20/20 to 20/200.41 Staphylococcal lid disease is associated with a marginal keratitis in which infiltrates form in the peripheral cornea, often where the lid margins cross the limbus (at 2, 4, 8, and 10 o’clock). Cultures of the infiltrates are sterile, suggesting that they result from hypersensitivity reactions to staphylococcal antigens or exotoxins deposited on the ocular surface (Figure 86-5).

Figure 86-5. Extensive marginal keratitis in a child secondary to staphylococcal lid disease. Note the characteristic lucid interval between arcuate infiltrate and corneal limbus.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Infective Keratitis

DIAGNOSTIC PROCEDURES Evaluation of infectious keratitis should include corneal scrapings for smears and cultures. Corneal specimens are obtained by scraping the leading edge and base of the ulcer with a sterilized Kimura spatula or Calgi swab.42,43 The lids and conjunctiva should be cultured separately. Ideally, specimens should be inoculated directly on to the culture media since corneal scrapings contain small numbers of fastidious organisms. Amies transport media without charcoal is a useful alternative in the clinical setting.44 Media should include blood and chocolate agar, anaerobic blood agar, Sabouraud dextrose agar (without cycloheximide), and enriched thioglycolate broth. Tissue culture medium is inoculated for isolation of viruses. Acanthamoeba species can be isolated using nonnutrient agar plates overlaid with a dried broth culture of Escherichia coli. Staining of corneal scrapings is important for the early identification of bacteria, fungi, and Acanthamoeba and may provide the sole etiologic evidence in culture-negative cases.42,43 Multiple specimens should be taken and set aside for Gram, Giemsa, and other specialized stains. Methenamine silver, acridine orange, and Calcofluor white stains are useful for the detection of fungi and Acanthamoeba.45,46 Microbial antigens can be detected using immunodiagnostic methods. Enzyme-linked immunoassays and fluorescein-labeled monoclonal antibodies are especially useful in the diagnosis of chlamydial, herpetic, and other viral infections. Recombinant DNA methods have been applied to the detection of corneal fungi, herpes, and other viruses, and Acanthamoeba.47–49 Fungi and Acanthamoeba can be directly visualized in the cornea with confocal slit-lamp microscopy.50

CHAPTER

86

503

including ciprofloxacin, ofloxacin, and norfloxacin. Several studies have shown ciprofloxacin to be an effective single agent in the treatment of bacterial keratitis.53,54 Increasing resistance of grampositive cocci, Actinomycetes, and Pseudomonas species to ciprofloxacin prompted the concurrent use of a cephalosporin and treatment of bacterial keratitis with a newer generation of fluoroquinolones.55–59 Levofloxacin effectively penetrates ocular tissues and has superior gram-positive coverage, but increasing resistance to this agent is being reported.59,60 Fourth-generation moxifloxacin demonstrates lower minimum inhibitory concentrations for gram-positive organisms and gatifloxacin for gram-negative organisms and are active against methicillin-susceptible Staphylococcus aureus isolates.61 The newer fluoroquinolones are reported to have better in vitro activities, increased ocular penetration, and potentially lower rates of development of resistance than other fluoroquinolones.62 In summary, mild and moderate cases of bacterial keratitis may respond to a single fluoroquinolone but severe cases should be treated with a cephalosporin combined with a fluoroquinolone or aminoglycoside. Choices for methicillin-resistant Staphylococcus aureus are limited to gentamicin and vancomycin. Because of their rapid clearance, topical antibiotics should be administered frequently. Initially these should be given as single drops for 5 consecutive minutes and then every 15 minutes for 4 consecutive doses.63 After this initial regimen, drops are administered every 30 to 60 minutes for at least 48 hours. Subsequent therapy is modified according to culture results and clinical course, but treatment is usually continued for 7 to 14 days. A favorable therapeutic response is indicated by diminished pain, healing of the epithelium, decrease in size and density of the corneal infiltrate, and decrease in corneal edema and inflammation in the anterior chamber.

TREATMENT The mainstay of treatment for corneal infections is the intensive use of topical anti-infectives. Corneal ulcers are medical emergencies. Although some childhood keratitis can be successfully managed in an outpatient setting,51 many children will require inpatient care. Mydriatic drops should be used to avoid pupillary synechiae.

Viruses The treatment of viral infections, excluding herpetic keratitis, is largely symptomatic because they are self-limited and no effective therapy is currently available. Epithelial HSV infections can be debrided or treated with topical antiviral drugs, either vidarabine ointment (Vira-A) or trifluorothymidine (Viroptic) drops. Drops are administered while the patient is awake every 2 hours during the first week and every 6 hours during the second week, and then rapidly tapered since these drugs are toxic to the corneal epithelium.13 Oral acyclovir is effective in the treatment of herpetic keratitis and represents an alternative to topical therapy in noncompliant children.52 Stromal HSV disease may require the addition of topical corticosteroid agents, but concomitant use with topical antiviral agents is necessary to prevent the reactivation of epithelial infection. Herpes zoster keratouveitis is unresponsive to available topical antiviral agents and is best managed with frequent application of corticosteroid topically and acyclovir (80 mg/kg per day) orally.

Fungi Fungal keratitis is treated with topical natamycin, flucytosine, amphotericin B, miconazole, or flucytosine.28,29 Frequent initial instillation (hourly) is slowly reduced over several weeks. Adequate treatment requires 6 to 12 weeks owing to poor corneal penetration and the slow growth of fungi. Lack of a therapeutic response is not uncommon and should prompt the addition of parenteral therapy or consideration of excisional keratoplasty.64 Deep fungal keratitis requires parenteral therapy from the outset because of the risk of fungal endophthalmitis. Subconjunctival fluconazole can be helpful in severe fungal keratitis unresponsive to topical and systemic therapy.65

Acanthamoeba Treatment of Acanthamoeba keratitis is often made difficult by delay in diagnosis. Early and aggressive therapy with cationic disinfectants (polyhexamethylene biguanide or chlorhexidine) combined with propamidine isethiocyanate (Brolene) and neomycin can be curative.66–71 Both cationic compounds show rapid killing (2 to 12 hours) of the trophozoites and cysts at concentrations that are not toxic to corneal cells.72 Clearance of Acanthamoeba trophozoites and cysts requires months of treatment.71 Corneal transplantation is only recommended when the infection continues.

COMPLICATIONS Bacteria The mainstay of treatment for bacterial keratitis has been a combination of a cephalosporin (50 mg/mL) drops and a fortified aminoglycoside, either tobramycin (15 mg/mL) or gentamicin (14 mg/mL).43 Cefazolin is selected for gram-positive coverage and ceftazidime when there is concern about Pseudomonas aeruginosa. Toxicity, drug instability, and variability of concentration and pH of fortified antibiotics led to the use of topical fluoroquinolone antibiotics,

Long-term complications are usually related to the loss of corneal transparency or refractive changes, or both. Central corneal scars (leukomas) that obstruct the visual axis can cause serious visual loss. Although the severity of scarring tends to diminish over time, even short periods of visual deprivation in children younger than 8 years of age can result in the development of amblyopia. Although corneal transplants can restore vision,73 results of corneal grafting in children are often disappointing due to the increased likelihood of graft rejection.

504

SECTION

L

CHAPTER

Eye Infections

scars between the iris and lens surface (synechiae). Dystrophic calcification of the epithelial basement membrane (band keratopathy), iris nodules, lens opacities, and phthisis bulbi can occur.

87

Uveitis, Retinitis, and Chorioretinitis

Specific Causes Viruses

Avery H. Weiss

Uveitis refers to any intraocular inflammation. Anterior uveitis is characterized by inflammation of the iris and ciliary body (iridocyclitis). Intermediate uveitis includes inflammation of the pars plana and vitreous (pars planitis, vitritis). Posterior uveitis features inflammation of the retina and choroid (retinitis, choroiditis). Uveitis is an uncommon disorder in childhood; among affected children, the anterior form is more common than the posterior form. At least half of the children with an identifiable cause of anterior uveitis have juvenile arthritis, either juvenile inflammatory arthritis or human leukocyte antigen (HLA)-B27-related spondyloarthropathy.1,2 In the United States, infections are much more likely to be associated with posterior uveitis.

Many viruses are associated with an acute nongranulomatous uveitis that is mild and self-limited.3,4 Varicella-zoster virus (VZV) demonstrates the spectrum of disease involvement. Anterior uveitis resulting in mild inflammation occurs in up to 25% of patients with varicella.5,6 It tends to be self-limited, resolving within 10 to 14 days. By contrast, uveitis associated with herpes zoster ophthalmicus is more severe and almost always requires aggressive treatment with a topical corticosteroid and systemic acyclovir. Devastating visual complications can occur when there is concurrent involvement of structures of the posterior segment. Uveitis associated with herpes simplex virus (HSV) infection is usually a complication of a deep keratitis but can occur without antecedent keratitis.7

Mycobacterium tuberculosis Chronic granulomatous uveitis, sometimes in association with iris nodules, was once commonly caused by Mycobacterium tuberculosis.8 However, it is now uncommon, except in developing countries and in patients with acquired immunodeficiency syndrome (AIDS).

ANTERIOR UVEITIS Table 87-1 summarizes the important pathogens and disorders responsible for anterior uveitis, with distinguishing features of each.

Clinical Manifestations Patients with anterior uveitis come to the physician’s attention because of pain, photophobia, and blurred vision. The whole eye looks red, but closer inspection reveals that the predominant area of hyperemia is at the corneal limbus owing to the vascular engorgement of the underlying ciliary body. Slit-lamp examination allows visualization of: (1) cells floating in the anterior chamber or adherent to the corneal endothelium (keratic precipitates); and (2) turbidity of the normally optically clear aqueous humor resulting from the spillover of protein (flare). Distortion of the pupil is related to formation of inflammatory

Treponema pallidum The incidence of congenital syphilis parallels the incidence of syphilis among heterosexuals in the United States. Congenital syphilis also remains a problem in underdeveloped countries. Chorioretinitis is a classic sign of early congenital syphilis, “salt-and-pepper” pigmentary changes of the peripheral retina being the most common finding.9,10 In severe cases, the pigmentary changes can be extensive, mimicking the fundus findings of retinitis pigmentosa. In comparison, the increased incidence of acquired cases reflects the emergence of human immunodeficiency virus (HIV) infections. Although anterior uveitis is a common presentation, some patients predominantly have an intense posterior uveitis with necrotizing retinitis, retinal vasculitis, or optic neuritis.11,12

TABLE 87-1. Common Causes of Anterior Uveitis Organism or Condition

Duration

Associated Ocular Findings

Associated Systemic Findings

Varicella

Acute

Iris atrophy (rare)

Vesicular skin rash

Herpes simplex

Acute

Keratitis

None

Mumps

Acute

NS

Fever, headache, parotitis

Influenza

Acute

NS

Febrile respiratory illness

Infectious mononucleosis

Acute

NS

Fever, pharyngitis, lymphadenopathy

Measles

Acute

NS

Fever, coryza, Koplik spots

Kawasaki disease

Acute

NS

Fever, mucocutaneous findings, peripheral edema, unilateral lymphadenopathy

Herpes zoster

Chronic

Iris atrophy (common), KP

Fever, vesicular skin rash

Syphilis

Chronic

Iris nodules, chorioretinitis, optic neuritis

Maculopapular skin rash (palms and soles)

Tuberculosis

Chronic

Iris nodules, granulomatous KP, choroidal granulomas

Pulmonary lesions

Lyme disease

Chronic

Keratitis, optic neuritis

Erythema migrans, arthritis

Leprosy

Chronic

“Iris pearls” of iris atropy

Peripheral neuropathy, skin lesions

KP, keratitic precipitates that represent aggregates of inflammatory cells on the corneal endothelium; NS, nonspecific.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Uveitis, Retinitis, and Chorioretinitis

Borrelia burgdorferi Borrelia burgdorferi, the organism responsible for Lyme borreliosis, can invade the cornea (keratitis) and uveal tract (iridocyclitis), and infection of the vitreous retina and choroid is being documented with growing frequency.13,14 Lyme uveitis can be a late manifestation of Lyme disease.15,16

Kawasaki Disease Anterior uveitis is common in Kawasaki disease. In one prospective study, 83% of 41 patients with Kawasaki disease had or experienced such findings in the first week of the disease.17 Iridocyclitis and periorbital vasculitis have also been described. Conjunctival hyperemia is present in > 90% of cases.

POSTERIOR UVEITIS (RETINITIS)

CHAPTER

87

505

toxoplasmosis are tractional distortion of the macula, optic atrophy, cataract, and retinal detachment. Recurrences usually appear at the edge of a quiescent scar, where encysted organisms lie dormant until they are reactivated.22,23 Toxoplasmosis acquired postnatally often appears with isolated ocular involvement.21,26,27 Solid organ transplantation is a source of acquired toxoplasmosis with delayed onset due to the protective benefit of bactrim prophylaxis.28 Although necrotizing retinochoroiditis is the major sequela, early manifestations can be limited to vitritis, anterior uveitis, or retinal vasculitis.28 The presence of multiple active lesions or extensive retinitis is common in patients with AIDS.29 The clinical diagnosis of ocular toxoplasmosis is confirmed in 50% to 80% of cases by demonstration of Toxoplasma-specific immunoglobulin (Ig) G, IgM, and IgA in the serum.30,31 Detection of specific IgG and IgA in serum or aqueous humor by immunoblotting is the most sensitive method, providing confirmatory evidence in 70% of cases.31 Identification of toxoplasmic-specific gene sequences by polymerase chain reaction (PCR) testing of aqueous humor has

Table 87-2 summarizes the predominant ocular manifestations and associated findings for the infectious agents most commonly implicated in retinitis.

Specific Causes

Toxoplasma gondii Toxoplasmosis is the most common cause of posterior uveitis in immunocompetent individuals. Infection may be contracted prenatally or postnatally18–21; the proportion of prenatal versus postnatally acquired infections varies geographically. Congenital infections can be identified at birth as active retinal lesions, inactive chorioretinal scars, or may not become clinically evident until months or years later.22,23 Acquired infections can have an acute or delayed onset of ocular disease. Peak occurrence of acquired infection is between the second and fourth decades. Active lesions appear as fluffy, white areas of focal necrotizing retinitis attributable to proliferation of live parasites and reactive inflammatory cells in the overlying vitreous.23 In some cases, the vitreal reaction is so severe that it obscures the underlying retina, giving the so-called headlight-in-a-fog appearance (Figure 87-1). Healing of these lesions leaves one or multiple atrophic chorioretinal scars, which can be located in the macula or the peripheral retina (Figure 87-2). In addition to focal retinitis there are atypical presentations. Punctate outer retinal toxoplasmosis is characterized by the presence of gray-white lesions of the outer retina with little or no overlying vitritis.24 Healed lesions appear as granular white opacities. Neuroretinitis features optic disk swelling with macular exudates and visual loss, and multifocal retinal infiltrates.25 Optic neuritis can occur with infections of the nerve or surrounding retina. Trapping of the organism within the terminal branches of the perifoveal capillaries probably explains the high incidence of macular involvement and visual loss. Additional causes of visual loss in

Figure 87-1. Fundus photograph shows active toxoplasmic lesion situated at the edge of a pigmented chorioretinal scar; the view is hazy because there are inflammatory cells in the overlying vitreous.

TABLE 87-2. Common Infectious Causes of Retinitis Organism

Predominant Manifestation

Toxoplasma gondii Rubella virus

Necrotizing retinitis Pigmentary retinopathy

Cytomegalovirus Hemorrhagic retinitis Herpes simplex virus Necrotizing retinitis Varicella-zoster virus Necrotizing retinitis Toxocara canis

Retinal granuloma

Associated Findings Vitritis, optic neuritis Cataract, microphthalmia Cottonwool spots Vitritis, vasculitis, optic neuritis Occlusive vasculitis, vitritis Vitritis, endophthalmitis

Figure 87-2. Fundus photograph shows portion of toxoplasmic chorioretinal scar centered on the macula with scattered clumping of retinal pigmentation.

506

SECTION

L

Eye Infections

Figure 87-3. Fundus photograph shows “salt-and-pepper” retinopathy typical of congenital rubella.

Figure 87-4. Fundus photograph shows geographic chorioretinal scar related to congenital cytomegalovirus infection.

Cytomegalovirus reported specificities of 83% to 100% and sensitivities ranging from 28% to 53%.32,33 Invasive procedures to obtain aqueous humor or ocular tissue may be justified in severe cases. Although active infections are usually treated with antiparasitic drugs, toxoplasmosis is a self-limited disease in immunocompetent individuals. Therefore the long-term benefits of treatment for acute and recurrent toxoplasmic chorioretinitis are uncertain. Current regimens include the following: (1) pyrimethamine, sulfadiazine, and folinic acid; (2) trimethoprim and sulfamethoxazole; and (3) adjunctive clindamycin for coverage against the encysted form.34–39 In murine models the combination of atovaquone and sulfadiazine is a promising alternative with few side effects.40 Systemic corticosteroid therapy is indicated when there is visual loss due to inflammation involving the optic nerve or macula.23 Local destruction of solitary lesions with photocoagulation, cryotherapy, or surgical vitrectomy to remove preretinal membranes or relieve retinal traction may be beneficial in selected cases.23,41

Rubella The most common ocular manifestations of congenital rubella are retinitis, cataract, microphthalmos, and glaucoma.42–44 Retinitis is characterized by pigmentary mottling of the retinal pigment epithelium (RPE) due to viral invasion, which gives the fundus a “saltand-pepper” appearance (Figure 87-3). At birth, the fundi can be normal or can show pigmentary changes. Persistent infection of the RPE causes the pigmentary disturbances to develop or to progress during the first few years of life, after which they tend to remain stable.43,45 Visual function is normal in patients with involvement of this epithelium alone. Choroidal neovascularization, macular detachment, and visual loss can rarely be late complications of rubella retinopathy.46,47 Despite newer technology, cataract surgery is frequently associated with poor visual outcomes owing to postoperative inflammatory complications, glaucoma, and cortical visual impairment.43,44 Virus can be isolated from lens material for several years after intrauterine infection.48 Rubella virus infection can contribute to the development of Fuch heterochromic cyclitis, a chronic, unilateral uveitis with mild inflammation that leads to iris atrophy. Topical corticosteroids do not suppress the inflammation unlike other types of anterior uveitis. Assay of aqueous humor revealed IgG rubella antibodies in 100% of 52 patients, and the presence of rubella-specific gene sequence in 18% of patients.49

Chorioretinitis is the major ocular sequela in infants congenitally infected with cytomegalovirus (CMV).50–52 Postinfectious scars appear as geographic areas of chorioretinal atrophy with surrounding hyperpigmentation (Figure 87-4). Lesions vary in size from 0.5 to 5.0 mm (1/2 to 5 disk diameters) and are located in the macula or peripheral retina. Chorioretinal scars are present in 10% to 21% of symptomatic infants but are uncommon in asymptomatic infants. Typically, chorioretinitis in congenital CMV is inactive, but individual cases with late onset and reactivation of chorioretinitis have been reported.53 Visual loss is associated with macular scarring but can be due to optic atrophy or cortical visual impairment. Although neuropathologic correlation is lacking, CMV invasion of neural tissues during development could account for the observed optic nerve and brain damage. CMV retinitis can also be an acquired infection in immunosuppressed children, such as transplant recipients, those receiving chemotherapy for malignancy, and those with congenital or acquired immunodeficiencies, such as HIV/AIDS.54–58 In the era of highly active antiretroviral therapy (HAART) for HIV, the incidence of CMV retinitis has decreased and the rates of visual loss, retinal detachment, and mortality have decreased.59–62 The clinical characteristics of CMV retinitis are similar for patients with and without HIV infection.63 Early retinitis appears as small white lesions that can be mistaken for cottonwool spots. The white lesions correspond histologically to areas of intracellular and extracellular edema and necrosis of the retina. These inflammatory changes are frequently associated with intraretinal hemorrhage and vasculitis of nearby retinal vessels, but the overlying vitreous is clear. As infection spreads, the size of the lesions increases along the course of the vessels, and new lesions develop (Figure 87-5). When lesions heal, the necrotic retina is replaced by glial scar tissue with associated pigmentary clumping.64–66 Visual loss is correlated with the location of retinitis. Involvement in the region of the optic nerve and macula is associated with decreased visual acuity and a central visual field defect. Peripheral retinitis causes loss of peripheral visual field but acuity can be normal. Both neonates and immunosuppressed children with CMV-induced chorioretinitis should be treated, because active CMV infection can progress relentlessly to blindness.66–68 Ganciclovir, foscarnet, and cidofovir are parenteral antivirals approved for initial induction and maintenance therapy of CMV retinitis.69–79 Oral ganciclovir may be an alternative to intravenous medications for maintenance, with fewer side effects.80 For patients with progressive retinitis or drug-related

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Uveitis, Retinitis, and Chorioretinitis

CHAPTER

87

507

Varicella-Zoster Retinochoroiditis and the Acute Retinal Necrosis Syndrome

Figure 87-5. Fundus photograph shows localized area of cytomegalovirus retinitis in the peripheral retina; necrotizing lesion has spread along the course of a retinal vessel and is surrounded by hemorrhage.

toxicity (neutropenia, renal disease), local therapy given as an intravitreal injection or surgically implanted “pellet” can be considered.81 Local treatment is combined with a systemic antiviral to prevent systemic disease and to protect the second eye. Current drugs are virostatic and must be given on a long-term basis to prevent progression or recurrence of active disease. Maintenance anti-CMV therapy can be discontinued in patients with immune reconstitution.82 Human monoclonal antibody directed against CMV appears to enhance the benefit of standard antiviral therapy.83,84 A subset of patients who become immune-reconstituted with CD4+ T-cell counts greater than 100 cells/mm3 after initiation of HAART develop floaters due to vitritis. The incidence of this immune recovery uveitis varies between centers but can be complicated by papillitis, macular abnormalities, and visual loss.85,86

Herpes Simplex Virus Keratoconjunctivitis may be a manifestation of localized neonatal HSV infection, acquired during passage through the birth canal. HSV retinitis in neonates generally indicates disseminated or CNS disease.87,88 The presence of skin lesions, chorioretinitis, and brain abnormalities in the neonate is suggestive of intrauterine infection.89 The lesions consist of yellowish areas of necrotizing retinitis, which can be localized or diffuse. Marked vitritis, retinal vasculitis, and persistent fetal vasculature can occur; full-thickness necrosis can lead to optic atrophy and blindness.90–92 Opsoclonus can be an early sign of encephalitis.93 Acyclovir is the treatment of choice.

Chickenpox or herpes zoster ophthalmicus can be associated with optic disk swelling, macular exudates, multifocal retinitis, and retinal vascular occlusion.98–100 Acute retinal necrosis (ARN) is an uncommon disease of normal and immunocompromised individuals usually caused by VZV followed by herpes simplex (HSV1, HSV2), CMV and, rarely, Epstein–Barr virus.101–106 HSV-2 is the most frequently identified cause of ARN in childhood.107 Therefore the occurrence of ARN in childhood is often attributed to reactivation of congenital or neonatal HSV infection.104,107,108 Characterstic features include multifocal necrotizing retinitis, occlusive arteritis, panuveitis, and rapidly deteriorating vision. Thinning and atrophy of the retina in necrotic areas combined with vitreous organization and traction predispose to retinal detachment. Inflammatory optic neuropathy and panuveitis with visual loss can precede the retinopathy.109 Diagnosis is based on detection of virus-specific gene sequence in aqueous or vitreous samples using the PCR assay. Disease progression is so rapid that visual loss is often irreversible despite aggressive antiviral treatment. However, early treatment with antivirals may prevent further visual loss and the development of ARN in the second eye.110 In patients with AIDS, and, less frequently, in immunocompetent individuals, the retinopathy can be atypical, featuring a rapidly progressive necrotizing retinitis and severe bilateral visual loss.111–114 Although progressive outer retinal necrosis has been emphasized, there is histopathologic evidence of panretinal necrosis. Because patients are immunologically suppressed, the immune-mediated vasculitis and vitritis are less severe.

West Nile Virus West Nile virus (WNV) has become an important cause of arboviral infections of the central nervous system (CNS) in the United States.115 The combined presence of acute encephalitis with multifocal chorioretinitis and vitritis should prompt concern about WNV and other arboviruses. Detection of IgM antibody to the virus confirms the diagnosis.116,117

Toxocara canis Systemic migration of the nematode Toxocara canis usually occurs in 1- to 4-year-old children and results in the clinical syndrome of visceral larval migrans.118,119 Ocular involvement is not concurrent with the visceral larva migrans stage and is usually detected between 4 and 8 years of age. Ocular toxocariasis takes one of the three following forms: (1) inflammatory granuloma of the posterior pole; (2) diffuse endophthalmitis; and (3) solitary granuloma of the peripheral retina.120,121 Common clinical findings include decreased vision, white pupil (leukocoria), and strabismus. The diagnosis is confirmed by the enzyme-linked immunosorbent assay test, which has high sensitivity and specificity. Treatment with periocular or systemic corticosteroids is directed at the inflammatory response incited by the death of the worm.122 Surgical removal of vitreal membranes and reattachment of the retina may improve vision in individual cases.123

Diffuse Unilateral Subacute Neuroretinitis Congenital Varicella During the first 20 weeks of pregnancy, approximately 2% of fetuses exposed to maternal chickenpox develop an embryopathy.94,95 Damage is mediated by direct neuropathic effects of the virus or by secondary vasculitis. Clinical features include skin scarring (76%), eye defects in 51% (chorioretinitis, cataracts, and microphthalmia), skeletal anomalies (49%), and microcephaly and other neurologic deficits (60%).96,97 Chorioretinitis associated with VZV is indistinguishable ophthalmoscopically from that due to other intrauterine infections. Reduced visual acuity is common because lesions frequently involve the macula.

Diffuse unilateral subacute neuroretinitis is an infection of the eye caused by nematodes endemic to the southeastern and midwestern regions of the United States. Although the organisms have not been isolated, migratory worms have been visualized directly in the subretinal space. One possible worm is Ancylostoma caninum, a dog hookworm, which is a common cause of cutaneous larval migrans. Initially, acute visual loss in one eye is noted. Examination of the fundus reveals gray-white lesions in the outer retina that fade over a few days and then reappear elsewhere. Invariably, cells are present in the vitreous, and the optic disk is swollen. Progressive involvement leads to further visual loss, diffuse pigmentary changes, and optic

508

SECTION

L

Eye Infections

atrophy. Diminution of electroretinographic responses confirms retinal damage, and visual loss has probably occurred.124,125 Baylisascaris procyonis, a common intestinal raccoon roundworm, is another cause of neural larva migrans and ocular larva migrans. Although the neurological deterioration predominates, ophthalmologic findings include a diffuse unilateral neuroretinitis and choroidal infiltrates.126,127 Treatment consists of either oral thiabendazole or ivermectin in early stages in patients with active vitritis or direct photocoagulation of the worm.128–130

Multifocal choroiditis occurring in the context of AIDS deserves special mention because of its association with life-threatening systemic infections. Etiologic agents include Pneumocystis carinii, Cryptococcus neoformans, M. tuberculosis, Candida, Mycobacterium avium-complex, and H. capsulatum.140–142

CHAPTER

Histoplasma capsulatum Histoplasmosis of the choroid is believed to be due to invasion by Histoplasma capsulatum. Supporting evidence is largely epidemiologic, because the organism has never been isolated from choroidal lesions.131,132 Recent evidence for persistence of H. capsulatum DNA sequences in the lesions suggests these products serve as immunogens and incite chronic inflammation.133 Viable organisms have only been isolated from immunocompromised patients with endophthalmitis who have disseminated histoplasmosis.134,135 This syndrome is characterized by the presence of multiple areas of atrophy in the peripheral retina, referred to as histo spots. Curvilinear areas of clumped pigment and hypopigmentation surround the disk.136 Focal scars do not contain viable organisms, but there is a recurrent lymphocytic infiltration that predisposes to the development of neovascular membranes and hemorrhagic macular detachment. Treatment options include laser photocoagulation, photodynamic therapy with verteporfin, and intravitreal corticosteroids, particularly when the macula is threatened.137–139 Antifungal therapy has no role in treatment of this disease because there are no actively replicating organisms.

Focal and Multifocal Choroiditis Infectious organisms that spread hematogenously are more likely to become trapped in the choroid than in other parts of the eye because of high choroidal blood flow and large number of fenestrated capillaries. There may be one or multiple foci of choroidal infiltrates (Figure 87-6). Tuberculosis can present as a focal choroidal mass, multiple choroidal masses, diffuse or multifocal choroiditis, or retinal vasculitis.140 Additional infections associated with choroiditis include coccidioidomycosis, cryptococcosis, and nocardiosis. Such infections tend to occur in regions where Mycobacterium tuberculosis is endemic or in debilitated and immunocompromised patients.

Figure 87-6. Fundus photograph shows multifocal choroidal infiltrates deep to the retina; there are no inflammatory cells in the overlying vitreous.

88

Endophthalmitis Mark J. Greenwald

The term endophthalmitis is applied to bacterial or fungal infection involving intraocular tissues (retina, uveal tract, or lens) or fluids (vitreous or aqueous). Two broad categories of infectious endophthalmitis are distinguished. Exogenous infection results from introduction of organisms into the eye through a surgical or traumatic penetrating wound; endogenous (or metastatic) infection is caused by organisms that enter the eye via the bloodstream. Both categories of infection are extremely serious, threatening blindness and even loss of the globe.1–3

ETIOLOGIC AGENTS A wide variety of microorganisms can cause endophthalmitis. Grampositive cocci (both staphylococci and streptococci) are leading pathogens for both exogenous and endogenous endophthalmitis. In recent years gram-positive isolates have increasingly tended to show resistance to antibiotics, including early-generation fluoroquinolones, but fortunately resistance to vancomycin and fourth-generation fluoroquinolones remains uncommon.4,5 Staphylococcus epidermidis is the organism that is most often identified in the postoperative setting and is a frequent agent of posttraumatic endophthalmitis as well.1,2 Often, S. epidermidis infection has a subacute onset 1 week to 1 month after surgery. Neonatal group B streptococcal septicemia can be associated with endogenous endophthalmitis.3 The gram-positive anaerobic bacillus Propionibacterium acnes is one of the most common causes of endophthalmitis after cataract surgery. Characteristically (although not invariably), P. acnes endophthalmitis develops 1 to several months postoperatively and follows a chronic smoldering course.6 Another bacillus, Bacillus cereus, is, in contrast, the most virulent organism inside the eye, capable of destroying the entire retina within hours of introduction by trauma or hematogenous seeding.7 Neisseria meningitidis was the most common cause of endogenous endophthalmitis before the advent of antibiotics and must still be recognized as having a predilection for intraocular localization.3 Nontypable Haemophilus influenzae can infect eyes after accidental trauma and surgery.8 Over the past decade, Klebsiella pneumoniae has emerged as the predominant cause of endogenous endophthalmitis in east Asia.7,9,10 Numerous other gram-negative bacilli (notably Pseudomonas aeruginosa and Escherichia coli) have been linked to both exogenous and endogenous endophthalmitis.1,3,11 Candida albicans bloodstream infection (BSI) remains the most common fungal precursor of endogenous endophthalmitis, although this complication appears to be decreasing in frequency with earlier initiation of systemic treatment for candidemia and invasive candidiasis.1,12–14 Meta-analysis of neonatal Candida BSI showed 3% with endophthalmitis.13 Non-albicans Candida spp. and a variety of other fungi (notably Aspergillus and Fusarium, Alternaria, and Scedosporium spp.) have been implicated in both endogenous and exogenous endophthalmitis, especially in posttraumatic cases or

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Endophthalmitis

immunocompromised hosts.15,16 Fungal endophthalmitis tends to follow a more indolent course than that of most bacterial infections.

EPIDEMIOLOGY AND HOST FACTORS Postoperative endophthalmitis most often follows cataract extraction but can be associated with any form of intraocular procedure or extraocular operations, such as strabismus repair (in which organisms are presumably introduced into the globe by inadvertent needle perforation of the sclera).17,18 The rate of endophthalmitis after pediatric intraocular surgery is similar to that found in adults – approximately 1 in 1000 cases.19–21 Extracapsular cataract surgery with intraocular lens implantation (increasingly being performed in childhood) predisposes particularly to chronic endophthalmitis caused by Propionibacterium acnes.6 A recent trend in cataract surgery toward use of unsutured “self-sealing” incisions through temporal clear cornea (as opposed to the superior limbus) has been associated with a small but significant increase in the frequency of postoperative endophthalmitis.22 Filtering operations for glaucoma (which produce a fistulous connection for aqueous flow through the corneoscleral limbus between the anterior chamber and a conjunctival bleb) lower the physical resistance of the globe to invasion by microorganisms, creating a higher risk of intraocular infection that can persist for decades if the drainage tract remains patent. The frequency of late bleb-related endophthalmitis appears to have grown considerably in children as well as adults because of increasing use of intraoperatively or postoperatively administered periocular antifibrotic agents (Mitomycin C or 5-fluorouracil). These drugs improve the rate of successful lowering of intraocular pressure after filtering surgery but tend to result in blebs that are large and thin-walled and, thus, particularly vulnerable to bacterial invasion.8,23 Endophthalmitis is a major concern after any penetrating trauma to the globe, especially in childhood, when contamination of the causative instrument with soil, saliva, or fecal material is relatively common.15,24,25 Occult penetration of the globe by a needle, thorn, or similarly shaped object can be unsuspected in a nonverbal child until days later when infection leads to obvious inflammatory signs. In a child with posttraumatic endophthalmitis, the possibility of an intraocular foreign body (which must be removed surgically) must be considered, and should be excluded, if necessary, with ultrasonographic or radiographic imaging. Endogenous endophthalmitis usually occurs in a host already known to have BSI or who is immunocompromised, but occasionally, ocular involvement is the first indication of the underlying problem.3 Endocarditis and meningitis are the most important localized infections associated with endophthalmitis. Diabetic patients have the propensity for retinal spread (often bilaterally) of K. pneumoniae from hepatic abscesses and of E. coli from an infected urinary tracts.10,11 Intravenous drug abusers have increased risk for intraocular fungal infection and also for devastating B. cereus endophthalmitis.3,7 Although otherwise healthy persons are occasionally affected, most cases of endogenous candidal endophthalmitis are nosocomial, occurring in association with indwelling intravascular catheters, use of broad-spectrum antibiotics, or immunosuppressive or major surgical treatment. Infants with intraocular candidiasis usually have a history of premature birth and pulmonary disease.12

CHAPTER

88

509

or choroiditis. Diffuse involvement of the anterior ocular segment can develop rapidly after invasion of the iris, but posterior extension usually remains limited in such cases. Spillover of organisms and inflammatory cells from infected tissues of the posterior ocular segment leads to involvement of vitreous, which can remain concentrated locally or become diffuse and can extend ultimately to the anterior segment, sclera, and orbit (panophthalmitis). With initial occlusion of the central retinal artery by a relatively large embolus, initial ischemia followed by massive dissemination of organisms directly into the retina can lead to irreversible blindness within a short time.

CLINICAL MANIFESTATIONS The first external sign of endophthalmitis is usually injection of conjunctival vessels, which can be striking and associated with marked conjunctival and eyelid edema or so mild that it scarcely attracts attention, particularly in postoperative or posttraumatic settings, in which some degree of inflammation is expected. Endophthalmitis must therefore be scrupulously included in the differential diagnosis of every “red eye.” Careful inspection of the anterior segment with a hand light usually permits identification of a whitish hypopyon, which results from layering of abundant leukocytes from aqueous fluid on to the surface of the iris or in the most dependent portion of the anterior chamber (Figure 88-1). Sometimes, however, the only readily identifiable sign in the anterior segment is blue-gray haziness or loss of clarity in visualization of the iris and the pupil (which typically shows absence of or diminished reactivity to light as well). Vision is usually reduced by obvious degrees. Low-grade endophthalmitis localized to the posterior segment (including most cases of intraocular candidiasis) may produce no external signs at all. Typically there is decreased clarity of ophthalmoscopic visualization and dimming of the fundus red reflex (resulting from vitritis), and visual acuity is significantly reduced. In some cases, however, the problem only becomes evident when detailed examination of the fundus through a dilated pupil reveals the presence of one or more small whitish chorioretinal or vitreal infiltrates with indistinct borders.

DIAGNOSTIC EVALUATION Suspicion of endophthalmitis mandates consultation with an ophthalmologist, who can confirm the diagnosis, quantify the vision loss,

PATHOPHYSIOLOGY Microbial proliferation begins and remains concentrated in the aqueous or vitreous fluid in most cases of exogenous endophthalmitis. Damage to the retina and other intraocular structures occurs secondarily from exposure to toxins elaborated by the organisms and inflammatory cells. By contrast, endogenous endophthalmitis typically originates in tissue within which a septic embolus has lodged.3 When this process occurs in a small terminal branch of a vessel, ocular infection begins as either a microabscess in the iris or a focal lesion of retinitis

Figure 88-1. Right eye of a 15-year-old girl with acute meningococcal meningitis and bilateral endophthalmitis. Note the small hypopyon (arrow), layered temporally in the anterior chamber because she had been lying on her right side.

510

SECTION

L

Eye Infections

and determine the extent of posterior-segment involvement (employing B-scan ultrasonography if the fundus cannot be visualized). In the context of well-defined systemic infection with an identifiable organism, it is not generally necessary to perform additional microbiologic studies when making the diagnosis of endophthalmitis.2 Otherwise, aspiration of ocular fluid for culture is required. If the patient is cooperative or easily immobilized, aqueous fluid (about 0.1 mL) can be obtained at the bedside with the use of a 27- or 30-gauge needle on a 1-mL syringe. Aspiration of the vitreous is a surgical procedure best performed in the operating room for pediatric patients (Figure 88-2), despite increasing performance in the outpatient setting in adults. Vitreous culture has a substantially higher yield than aqueous culture in exogenous and posteriorly localized endogenous infection.3,26,27 When fungal infection is suspected, polymerase chain reaction testing of vitreous aspirate can provide rapid confirmation of the diagnosis.28 In one-quarter to one-half of cases of presumed infectious endophthalmitis, eye fluid cultures are sterile. For maximum yield, aspirates should be inoculated into appropriate media promptly and properly, including those for isolation of anaerobic and fungal organisms. Inoculation of fluid specimens into blood culture bottles has also been used successfully.1 With endophthalmitis of unknown origin, multiple blood cultures should be performed, followed by an exhaustive search for remote infections (particularly meningitis, endocarditis, hepatic abscess, and urinary tract infection) and predisposing conditions (occult ocular trauma, diabetes, acquired immunodeficiency syndrome, and other immunodeficiency syndromes).

MANAGEMENT Antimicrobial agents for endophthalmitis can be delivered by a variety of routes. Intravenous antibiotic administration is the mainstay of treatment for endogenous bacterial infection.3,9 The choice of drugs should be based on known or presumed sensitivities of the documented or suspected pathogens, with dosage and duration at levels appropriate for meningitis (or greater if required for associated infection at another site). Endogenous fungal endophthalmitis usually responds well to treatment with intravenous amphotericin B; however, newer agents such as voriconazole (which can achieve therapeutic introacular concentration after oral as well as intravenous administration) are increasingly viewed as appropriate and possibly preferable alternative therapies.29 In the past, many cases of exogenous endophthalmitis were cured with combined intravenous, subconjunctival, and topical antibiotics, but intravitreal administration is now regarded as the most important

route of drug delivery for this condition (see Figure 88-2).1,17,18 Initially, only a single injection is given (immediately after aspiration of vitreous for culture), followed by one or more additional injections if response to the first is incomplete. Gentamicin and cefazolin were formerly the intraocular antibiotics most often used, but growing concerns about aminoglycoside retinal toxicity and drug-resistant organisms (mainly among Staphylococcus epidermidis, and streptococcal isolates) have led to a shift toward other agents, particularly ceftazidime and vancomycin (Table 88-1, and see Table 294-4).30 Controversy persists regarding the need for concurrent use of multiple routes of therapy.31 Outcomes reported after combined intravitreal and systemic administration of antibiotics are similar to those seen with intravenous treatment alone in endogenous infections and with intravitreal treatment alone in most exogenous infections.3,9,32,33 Many experienced clinicians continue to employ both modalities routinely. Fourth-generation fluoroquinolones such as moxifloxacin can achieve therapeutic intraocular levels after oral administration (initial loading dose in an adult of 400 mg twice daily, followed by 400 mg once daily for 1 to 2 weeks) and have a favorable spectrum of coverage, making them an attractive and increasingly used adjunct to intravitreal injection.34 Minimum inhibitory concentrations for methicillin-resistant S. aureus are at least 10-fold those for methicillinsusceptible S. aureus (MIC90 2 to 16 μg/mL). Subconjunctival antibiotics (given as once- or twice-daily injections) are also sometimes still used, although data to indicate incremental benefit are lacking. Many clinicians also include corticosteroid agents (intraocular, subconjunctival, topical, systemic, or a combination) in the regimen in attempt to reduce ocular damage from the inflammatory process; no adverse effect on control of infection from this practice has been documented when appropriate antimicrobial therapy is given.1,2,17 The role of vitrectomy in the management of exogenous endophthalmitis was clarified by the results of a major multicenter, randomized clinical trial. The investigators compared outcomes from treatment of postoperative endophthalmitis after cataract surgery using intravitreal antibiotics (amikacin and vancomycin) with and without immediate surgical removal of most of the infected vitreous. Vitrectomy provided significant benefit only when initial visual acuity was severely reduced to light perception. With better initial vision, outcomes were the same with and without vitrectomy.33 Vitrectomy with removal of infected lens material appears to be required for cure of most cases of chronic endophthalmitis caused by P. acnes.6 Most authorities continue to view vitrectomy as warranted for endophthalmitis associated with penetrating trauma and glaucoma filtering blebs.2,17

TABLE 88-1. Intravitreal Injection for Endophthalmitis Drug

Figure 88-2. Diagrammatic illustration of needle passing through the pars plana of the ciliary body for diagnostic aspiration of vitreous or intravitreal injection of antibiotic. (Adapted from Bohigian GM. Endophthalmitis. In: Krupin T, Kolker AE, Rosenberg LF (eds) Complications in Ophthalmic Surgery, 2nd ed. St. Louis, Mosby, 1999, pp 19–36.)

Dosea (mg)

Comment

Vancomycin

1

Unless gram-positive bacteria can be excluded

Amikacin

0.4

Now seldom used because of potential retinal toxicity

Ceftazidime

2.25

Preferred agent for gram-negative coverage

Amphotericin B

0.005

If fungus suspected

Voriconazole

0.05

Increasingly used alternative to amphotericin

Dexamethasone

0.4

Recommended for anti-inflammatory effect by some authorities

a Each drug should be diluted to provide the indicated dose in a volume of 0.1 mL and should be injected using a separate 1-mL syringe. (Recommendations from references 1 and 17, and Hariprasad SM, Mieler WF, Flynn HW, et al. Advances in Endophthalmitis Management. Instruction course syllabus, American Academy of Ophthalmology Annual Meeting, Las Vegas, Nevada 2006.)

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Periorbital and Orbital Infections

CHAPTER

89

Occasionally, even after intensive treatment, an eye is left blind and intractably inflamed by endophthalmitis. In such cases, it may be necessary to remove the entire ocular contents surgically (evisceration) or to remove the complete globe (enucleation).

CHAPTER

OUTCOME

Periorbital and Orbital Infections

With current optimal therapy, the prognosis for infectious endophthalmitis is often good. Recovery is the rule in: (1) postoperative endophthalmitis due to S. epidermidis and P. acnes; (2) endogenous Candida endophthalmitis; and (3) endogenous bacterial endophthalmitis that shows either primarily anterior-segment or only focal posterior-segment involvement at the time of diagnosis, regardless of organism.1,3,35 Mild to moderate permanent visual loss is common even in these favorable cases, however. In situations other than those mentioned, endophthalmitis remains a grave disorder. Cases of bacterial infection due to trauma or to a hematogenous spread from other sites of infection that show diffuse posterior-segment involvement have a particularly poor prognosis, typically leaving the patient with a nonseeing eye or no eye at all.3,7,15 Overall, with current therapeutic methods, useful vision (fingercounting or better) is retained in about 60% of eyes treated for acute or subacute exogenous bacterial endophthalmitis and in 40% of eyes treated for endogenous bacterial endophthalmitis. These figures have changed little over several decades.3,36

PREVENTION Preoperative application of a drop of 5% povidone-iodine (halfstrength dilution of the common commercially available solution) to the ocular surface has been shown to reduce the conjunctival bacterial population significantly and appears to lower the incidence of postoperative endophthalmitis.20,37 It also remains common practice among ophthalmologists to administer a single subconjunctival injection of antibiotic (usually an aminoglycoside or a cephalosporin) at the conclusion of intraocular surgery and to prescribe application of a topical solution several times a day for a week or longer. Use of moxifloxacin topically before as well as after surgery has been increasing. A minority of surgeons add antibiotic to the infusion fluid that circulates through the eye during cataract or vitrectomy surgery. Recent studies support the injection of antibiotic into the anterior chamber at the conclusion of surgery, but this approach has not been widely adopted in North America.38 More intensive prophylaxis against endophthalmitis is employed by most ophthalmologists after penetrating trauma. Intravenously administered vancomycin plus ceftazidime has largely replaced the traditional use of a first-generation cephalosporin, with or without an aminoglycoside.2 Oral administration of a fluoroquinolone (moxifloxacin, currently the preferred agent) can achieve therapeutic drug levels in the vitreous and reasonably good coverage for common infecting organisms.39 Some authorities recommend intravitreal injection as a preventive measure in cases of traumatic endophthalmitis that are likely to involve heavy contamination.2,21,24 Culture of foreign material removed from the eye, aqueous or vitreous, exposed uveal tissue, or the conjunctiva (in roughly descending order of usefulness) at the time of surgical wound repair can provide information of value in optimizing treatment if endophthalmitis develops later. Because of the risk of endophthalmitis, conjunctivitis that develops in an eye with a glaucoma filtering bleb should be treated promptly and aggressively with topical and, possibly, systemic antibiotics.8 Ophthalmologic evaluation to rule out ocular involvement in patients with Candida albicans fungemia or deep tissue infection appears to be justified, given the paucity of external signs when it develops in such cases, but the yield is likely to be low if systemic treatment has been started.14 Ophthalmologic screening has a very low yield in patients who have fever of unknown origin or in whom Candida is isolated only from a superficial site or a catheter, and so is no longer generally recommended.

511

89

Ellen R. Wald

The practitioner frequently has the opportunity to manage the child for whom the chief complaint is a “swollen eye” (Figure 89-1). Some children have trivial or self-limited disorders, but others can have sight- or life-threatening problems.

DIFFERENTIAL DIAGNOSIS The noninfectious causes of swelling of or around the eye include: (1) blunt trauma (leading to the proverbial “black” eye); (2) tumor; (3) local edema; and (4) allergy. In cases of blunt trauma, history provides the key to the diagnosis. Eyelid swelling continues to increase for 48 hours and then resolves over several days. Tumors that characteristically involve the eye include hemangioma of the lid, ocular tumors such as retinoblastoma and choroidal melanoma, and orbital neoplasms such as neuroblastoma and rhabdomyosarcoma.1 Tumors usually cause gradual onset of proptosis in the absence of inflammation. Orbital pseudotumor, an autoimmune inflammation of the orbital tissues, manifests as eyelid swelling, red eye, pain, and decreased ocular motility.2 Hypoproteinemia and congestive heart failure cause eyelid swelling due to local edema. Characteristic findings are bilateral, boggy, nontender, nondiscolored soft-tissue swelling. Allergic inflammation includes angioneurotic edema or contact hypersensitivity.3 Superficially, these problems can resemble the findings in acute infection. However, the presence of pruritus and the absence of tenderness are helpful distinguishing characteristics of allergic inflammation.

PATHOGENESIS The anatomy of the eye is important for an understanding of its susceptibility to spread of infection from contiguous structures. Veins that drain the orbit, the ethmoid and maxillary sinuses, and the skin of the eye and periorbital tissues (Figure 89-2) constitute an anastomosing and valveless network.1 This venous system provides opportunities for spread of infection from one anatomic site to another and predisposes to involvement of the cavernous sinus. Figure 89-3 demonstrates the relationship between the eye and the paranasal sinuses. The roof of the orbit is the floor of the frontal sinus, and the floor of the orbit is the roof of the maxillary sinus. The medial

Figure 89-1. A 10-year-old boy with the complaint of swollen eye.

512

SECTION

L

Eye Infections Periorbita Periosteum

Frontal v. Superior ophthalmic v. Angular v.

Cavernous sinus

Anterior facial v.

Maxillary sinus

Pterygoid plexus Inferior ophthalmic v. Figure 89-2. The valveless venous system of the orbit and its many anastomoses. (Redrawn from Harris GJ. Subperiosteal abscess of the orbit. Arch Ophthalmol 1983;101:753–754.)

Lamina papyracea Ethmoid sinuses

Frontal sinus

Orbit

Nasal septum

Maxillary sinus

Inferior turbinate

Oval cavity

Figure 89-3. The relationship between the eye and the paranasal sinuses is shown schematically. The roof of the orbit, the medial wall, and the floor are shared by the frontal, ethmoid, and maxillary sinuses, respectively. (Redrawn from Shapiro ED, Wald ER, Brozanski BA. Periorbital cellulitis and paranasal sinusitis: a reappraisal. Pediatr Infect Dis J 1982;1:91–94.)

wall of the orbit is formed by the frontal maxillary process, the lacrimal bone, the lamina papyracea of the ethmoid bone, and a small part of the sphenoid bone.4 Infection originating in the mucosa of the paranasal sinuses can spread to involve the bone (osteitis with or without subperiosteal abscess) and the intraorbital contents. Orbital infection can occur through natural bony dehiscences in the lamina papyracea of the ethmoid or frontal bones or via foramina through which the ethmoidal arteries pass.3 Figure 89-4 shows the position of the orbital septum. This structure is a connective tissue extension of the periosteum (or periorbita) that is reflected into the upper and lower eyelids. Infection of tissues anterior to the orbital septum are described as periorbital or preseptal.5

Orbital septum

Figure 89-4. The orbital septum is a connective tissue extension of the periosteum that is reflected into the upper and lower lid. (Redrawn from Shapiro ED, Wald ER, Brozanski BA. Periorbital cellulitis and paranasal sinusitis: a reappraisal. Pediatr Infect Dis J 1982;1:91–94.)

The septum provides a nearly impervious barrier to spread of infection to the orbit. Although preseptal cellulitis, or periorbital cellulitis (the terms may be used interchangeably), is often considered a “diagnosis,” the term is an inadequate diagnostic label unless accompanied by a modifier that indicates likely pathogenesis. Infectious causes of preseptal cellulitis occur in the following three settings: (1) secondary to a localized infection or inflammation of the conjunctiva, eyelids, or adjacent structures (e.g., conjunctivitis, hordeolum, acute chalazion, dacryocystitis, dacryoadenitis, impetigo, traumatic bacterial cellulitis); (2) secondary to hematogenous dissemination of nasopharyngeal pathogens to the periorbital tissue; and (3) as a manifestation of inflammatory edema in patients with acute sinusitis (Box 89-1).5 Infections behind the septum that cause eye swelling include subperiosteal abscess, orbital abscess, orbital cellulitis, cavernous sinus thrombosis, panophthalmitis, and endophthalmitis. Although all of these entities can be labeled “orbital cellulitis,” a systematic approach allows a more specific diagnosis, thereby directing management. Infections intrinsic to the eye (i.e., conjunctivitis, keratitis, endophthalmitis, and panophthalmitis) are discussed in Chapters 84 through 88.

PRESEPTAL INFECTIONS Conjunctivitis Conjunctivitis is the most common disorder of the eye for which children are brought for medical care. In most cases, the lids are crusted and thickened with hyperemic conjunctiva. The usual causes of conjunctivitis in children older than neonates but less than 6 years old are Haemophilus influenzae (nontypable) and Streptococcus pneumoniae. In approximately 20% to 25% of children with conjunctivitis due to H. influenzae, acute otitis media is a complicating feature. In this case, systemic antibiotics are preferable to topical ophthalmic preparations. Adenovirus is the most common cause of viral conjunctivitis in children older than 6 years.6 Occasionally young children with adenovirus infection have diffuse swelling of the lids that can be mistaken for a more serious problem7 (see Chapter 84, Conjunctivitis in the Neonatal Period (Ophthalmia Neonatorum)).

Hordeolum and Chalazion An external hordeolum, or stye, is a bacterial infection of the glands of Zeis or Moll (sebaceous gland or sweat gland, respectively)

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Periorbital and Orbital Infections

CHAPTER

89

513

BOX 89-1. Infectious Causes of Preseptal and Orbital Cellulitis PRESEPTAL CELLULITIS Localized infection of the eyelid or adjacent structure Conjunctivitis Hordeolum Dacryoadenitis Dacryocystitis Bacterial cellulitis (trauma) Hematogenous dissemination Bacteremic periorbital cellulitis Acute sinusitis Inflammatory edema ORBITAL CELLULITIS Acute sinusitis Subperiosteal abscess Orbital abscess Orbital cellulitis Cavernous sinus thrombosis Hematogenous dissemination Endophthalmitis Traumatic inoculation Endophthalmitis

associated with a hair follicle on the eyelid. In most cases, infection is localized and points to the lid margin as a pustule or inflammatory papule. The lid can be slightly swollen and erythematous around the area of involvement. An external hordeolum usually lasts a few days to a week and resolves spontaneously. An internal hordeolum is a bacterial infection of a meibomian gland, a long sebaceous gland whose orifice is at the lid margin.8 The infection usually causes inflammation and edema of the neck of the gland, which can result in obstruction. If there is no obstruction, infection points to the lid margin. If obstruction is present, infection points to the conjunctival surface of the eye.8 Sometimes the swelling caused by an acute internal hordeolum is diffuse rather than localized, and a pustule is not obvious on the lid margin. To clarify the cause, it is necessary to evert the eyelid and examine the tarsal conjunctiva. A tiny, delicate pustule is diagnostic of an internal hordeolum. The usual cause of acute internal or external hordeola is Staphylococcus aureus. An antibiotic ophthalmic ointment containing bacitracin can be applied to the site of infection. The main purpose of the topical therapy is to prevent spread of infection to adjacent hair follicles. Warm compresses may facilitate spontaneous drainage. In contrast to the internal hordeolum, a chalazion manifests as a persistent (more than 2 weeks in duration), nontender, localized bulge or nodule (3 to 10 mm) in the lid; the overlying skin is completely normal. It is a sterile lipogranulomatous reaction. When a chalazion is large and causes local irritation, incision may be required.

Dacryoadenitis Dacryoadenitis is an infection of the lacrimal gland. Sudden onset of soft-tissue swelling that is maximal over the outer portion of the upper lid margin is typical. Occasionally, the eyeball is erythematous and the eyelid swollen, and the patient can have remarkable constitutional symptoms. The location of the swelling is a distinguishing characteristic (Figure 89-5). When dacryoadenitis is caused by viral infection (mumps virus, Epstein–Barr virus,9 cytomegalovirus, coxsackievirus, echoviruses, and varicella-zoster virus), the area is only modestly tender.10 By contrast, when the infection is caused by bacterial agents, discomfort is prominent. In addition to S. aureus, which is the most common cause of bacterial dacryoadenitis, etiologic agents include streptococci, Chlamydia trachomatis, Brucella melitensis, and, occasionally, Neisseria gonorrhoeae.11,12 Fungal and rare parasitic infections of the lacrimal gland have been reported, including those with Cysticercus cellulosae and Schistosoma

Figure 89-5. A patient with dacryoadenitis due to an unspecified viral infection. The nontender swelling over the lateral portion of the left upper lid evolved while an antibiotic was being administered for acute otitis media. Swelling resolved in several days without any change in medical treatment.

haematobium.10 If parenteral therapy is required for suspected bacterial dacryoadenitis due to S. aureus, nafcillin, at 150 mg/kg per day divided into doses every 6 hours, is appropriate. If methicillinresistant S. aureus is suscepted, vancomycin (40 mg/kg divided into doses every 6 hours) should be initiated. Oral treatment of acute dacryoadenitis is undertaken with a semisynthetic penicillin such as dicloxacillin (100 mg/kg per day divided into doses every 6 hours), cephalexin or cefadroxil (100 or 50 mg/kg per day, respectively, divided into doses every 6 or 12 hours, respectively), or sulfamethoxazoletrimethoprim (based on 40 mg/kg per day of trimethoprim divided into doses every 12 hours) or clindamycin (40 mg/kg per day divided into doses every 6 hours). Treatment is continued until all signs and symptoms have disappeared. The differential diagnosis of swelling of the upper outer aspect of the eyelid includes inflammatory noninfectious problems such as Sjögren syndrome and sarcoidosis as well as benign and malignant tumors.10

Dacryocystitis Dacryocystitis is a bacterial infection of the lacrimal sac. Although it is uncommon, it can occur at any age as a bacterial complication of a viral upper respiratory tract infection (URI). Because of the course traversed by the lacrimal duct, which drains to the inferior meatus within the nose, it is surprising that the duct and sac are not infected more often. Delayed opening, inspissated secretions, and anatomic abnormalities lead to disproportionate representation of infants younger than 3 months among children with dacryocystitis. Patients with dacryocystitis have usually had a viral URI for several days. They then experience fever and impressive erythema and swelling in addition to exquisite tenderness, which is most prominent in the triangular area just below the medial canthus (Figure 89-6). Pressure over the lacrimal sac causes considerable discomfort but can result in expression of purulent material from the lacrimal puncta. Common causative organisms are gram-positive cocci. Streptococcus pneumoniae is most common in neonates, although S. aureus, H. influenzae, and Streptococcus agalactiae have also been reported.11,13 S. aureus and S. epidermidis are most commonly implicated in acquired dacryocystitis in the older patient.14 It is important to obtain material from the punctum because other organisms (including enteric gram-negative bacilli, anaerobic bacteria, and yeast) have occasionally been observed. Unusual pathogens, such as Pasteurella multocida and Aeromonas hydrophilia, have been reported rarely.15 Most patients with dacryocystitis require admission to hospital. Often they appear ill or toxic. Because of the potential for any case of bacterial facial cellulitis to result in cavernous sinus thrombosis, therapy with parenteral antibiotics is indicated until the infection

514

SECTION

L

Eye Infections

Figure 89-6. A 16-year-old boy with dacryocystitis. The area beneath the medial canthus is erythematous, indurated, and exquisitely tender.

begins to subside. Nafcillin (at a dose of 150 mg/kg per day divided into doses every 6 hours) or cefazolin (at a dose of 100 mg/kg per day divided into doses every 8 hours) is appropriate. In penicillin-allergic patients, vancomycin or clindamycin (40 mg/kg per day divided into doses every 6 hours) suffices. After substantial improvement is observed in local findings, an oral agent can be substituted to complete a 10- to 14-day course of therapy. The role of nonmedical management of dacryocystitis is controversial. Although surgical manipulation of the lacrimal duct is not necessary for most patients, both probing of the duct and incision and drainage have been reported to be successful in neonates.13 Incision and drainage and direct application of antibiotics inside the sac have been promoted by some practitioners who care for adults.16

Preseptal Cellulitis After Trauma Occasionally, preseptal cellulitis results from secondary bacterial infection of sites of local skin trauma (including insect bites) or with spread of infection from a focus of impetigo. The traumatic injury may be extremely modest or completely inapparent. Loosely bound periorbital soft tissues permit impressive swelling to accompany minor infection. The overlying skin can be bright red with subtle textural changes, or intense swelling can lead to shininess (Figure 89-7). Some patients have fever, but many are afebrile despite dramatic local findings. The peripheral white blood cell count is variable. In these cases, cellulitis, similar to that on any other cutaneous area, is caused by S. aureus or group A streptococcus.17 Several less common causes of lid cellulitis have been reported. Periocular cellulitis and abscess formation have resulted from infection with Pasteurella multocida in a healthy child who sustained a cat bite and cat scratch to the eyelid.18 Ringworm (caused by Trichophyton species) has also been recognized as a cause of lid infection (leading to preseptal cellulitis) characterized by redness, swelling, ulceration, and vesicle formation.19,20 Palpebral myiasis involving the eyelid of a 6-year-old child was reported from the Massachusetts Eye and Ear Infirmary.21 A small draining fistula through which the larvae was extracted was noted at the site of the erythematous and edematous lid. Several cases of cellulitis of the eyelid due to Bacillus anthracis have been reported from Turkey.22 The diagnosis was suspected when the erythematous and swollen lid developed an eschar. Scrapings showed the presence of gram-positive bacilli which were confirmed by culture. A primary case of lymphocutaneous Nocardia brasiliensis of the eyelid has been reported in an adult hunting in England 2 weeks before presentation following a small abrasion on this lower eyelid.23 In countries where Mycobacterium tuberculosis is endemic, this etiology should also be considered in patients who present with a swollen lid. Raina et al. reported 7 children with tuberculous lesions of their eyelids. In most cases the presentation was relatively indolent (2 days to 2 months) and fistulas occurred during the course of conventional antibiotic treatment for more typical bacterial disease.24 Diagnosis was confirmed by a positive tuberculin skin test, the

Figure 89-7. A 3-year-old boy with rapid onset of left-eyelid swelling and erythema after he incurred a small laceration at the lateral margin of the left eye. He had had an upper respiratory tract infection for 10 days. Group A streptococcus was recovered from the wound.

identification of a primary focus of tuberculosis in lung or bone and response to antituberculous therapy. Patients with bacterial cellulitis of traumatized areas rarely have bacteremia. Precise bacteriologic diagnosis is made through culture of exudate from the wound. If there is no drainage, a careful attempt at tissue aspiration is undertaken if it can be done safely (i.e., at a distance far enough from the orbit that there can be no potential damage to the eye). A tuberculin syringe with a 25-gauge needle can be used for aspiration of “tissue juice.” Usually, only a minuscule amount of infected material can be aspirated. A small volume of nonbacteriostatic saline (0.2 mL) is drawn into the syringe before the procedure. The saline is not injected into the skin; instead, it is used to expel the small volume of tissue fluid on to chocolate agar for culture.25 Parenteral treatment similar to that advised for dacryocystitis is recommended in patients with bacterial cellulitis, to hasten resolution and avoid spread of infection to the cavernous sinus.

Bacteremic Periorbital Cellulitis The child with bacteremic periorbital cellulitis, which is most often seen in infants younger than 18 months, has had a viral URI for several days. There is a sudden increase in temperature (to > 39°C) accompanied by the acute onset and rapid progression of eyelid swelling. Swelling usually begins in the inner canthus of the upper and lower eyelid and can obscure the eyeball within 12 hours. Periorbital tissues are markedly discolored and usually erythematous, although if the swelling has been rapidly progressive, the area may have a violaceous discoloration.26,27 The child’s resistance to examination commonly leads to the erroneous impression of tenderness. Retraction or separation of the lids reveals that the globe is normally placed and extraocular eye movements are intact. If retraction of the lids is not possible, orbital computed tomography scan may be necessary.28 The young age, high fever, and rapid progression of findings differentiate bacteremic preseptal cellulitis from other causes of swelling around the eye. In the era before universal H. influenzae type b (Hib) immunization, this organism was the most common cause of bacteremic periorbital cellulitis in approximately 80% of cases. S. pneumoniae accounted for the remaining 20%. The substantial decline that has been observed in the total number of cases of bacteremic periorbital cellulitis is attributable to the widespread use of the Hib vaccine since 1991 and the introduction of pneumococcal conjugate vaccine in 2000.29 A precise bacteriologic diagnosis is made by recovery of the organism from blood culture. If a careful tissue aspiration is performed, culture of the specimen may have a positive result. The pathogenesis of most of these infections, which usually occur during the course of a viral URI, is hematogenous dissemination from a portal of entry in the nasopharynx. This process is akin to the mechanism of most infections caused by Hib and some infections caused by S. pneumoniae.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Periorbital and Orbital Infections

In patients with bacteremic periorbital cellulitis, radiographs of the paranasal sinuses are often abnormal. However, the abnormalities almost certainly reflect the viral respiratory syndrome that precedes and probably predisposes to the bacteremic event, rather than a clinically significant sinusitis.5 Bacteremic cellulitis rarely arises from the paranasal sinus cavities, as evidenced by the finding that typable H. influenzae organisms are almost never recovered from maxillary sinus aspirates and likewise are rarely recovered from abscess material in patients who have serious local complications of paranasal sinus disease, such as subperiosteal abscess. Although S. pneumoniae can cause subperiosteal abscess in patients with acute sinusitis, such patients are not usually bacteremic. Treatment for suspected bacteremic periorbital cellulitis requires parenteral therapy. S. pneumoniae is the most likely cause in a child who has received both the Hib and pneumococcal conjugate vaccine series. Because this infection is usually bacteremic in the age group in whom the meninges are susceptible to inoculation, it may be prudent to use an advanced-generation cephalosporin such as cefotaxime or ceftriaxone (150 or 100 mg/kg per day, respectively, divided into doses every 8 or 12 hours, respectively). Lumbar puncture should be performed unless the clinical picture precludes meningitis. Addition of vancomycin (60 mg/kg per day divided into doses every 6 hours) or rifampin (20 mg/kg once daily, not to exceed 600 mg/day) is appropriate if cerebrospinal fluid pleocytosis is present. When evidence of local infection has resolved and there is no meningitis, oral antimicrobial therapy is prescribed to complete a 10-day course.

Preseptal (Periorbital) Cellulitis Caused by Inflammatory Edema of Sinusitis Several complications of paranasal sinusitis can result in the development of swelling around the eye. The most common and least serious complication is often referred to as inflammatory edema or a sympathetic effusion.4 This is a form of preseptal cellulitis, although infection is confined to the sinuses. Typically, a child at least 2 years old has had a viral URI for several days when swelling is noted. Often, there is a history of intermittent early-morning periorbital swelling that resolves after a few hours. On the day of presentation, the eyelid swelling does not resolve typically but progresses gradually (see Figure 89-1). Surprisingly, striking degrees of erythema can also be present. Eye pain and tenderness are variable. Eyelids can be very swollen and difficult to evert, requiring the assistance of an ophthalmologist. However, there is no displacement of the globe or impairment of extraocular eye movements. Fever, if present, is usually low-grade. The peripheral white blood cell count is unremarkable. Blood culture results are always negative. If a tissue aspiration is performed, culture of the specimen has a negative result. Sinus radiographs show ipsilateral ethmoiditis or pansinusitis. The age of the child, gradual evolution of lid swelling, and modest temperature elevation differentiate inflammatory edema from bacteremic periorbital cellulitis. The pathogenesis of sympathetic effusion or inflammatory edema is attributable to the venous drainage of the eyelid and surrounding structures. The inferior and superior ophthalmic veins, which drain the lower lid and upper lid, respectively, pass through or just next to the ethmoid sinus. When the ethmoid sinuses are completely congested, physical impedance of venous drainage occurs, resulting in soft-tissue swelling of the eyelids, maximal at the medial aspect of the lids. In this instance, infection is confined within the paranasal sinuses. The globe is not displaced, and there is no impairment of the extraocular muscle movements. However, inflammatory edema is part of a continuum with more serious complications resulting from the spread of infection outside the paranasal sinuses into the orbit.30 Rarely, infection progresses despite initial optimal management of sympathetic effusions. The infecting organisms in cases of inflammatory edema are the same as those that cause uncomplicated acute sinusitis (i.e., S. pneumoniae, nontypable H. influenzae, and Moraxella catarrhalis). Antibiotic therapy can be given orally if, at the time of the first

CHAPTER

89

515

examination, the eyelid swelling is modest, the child does not appear toxic, and the parents will adhere to management. Otherwise, admission to the hospital and parenteral treatment should be undertaken. The only source of bacteriologic information is that obtainable by maxillary sinus aspiration, which is not usually performed. Appropriate agents for outpatient therapy have activity against ß-lactamaseproducing organisms (e.g., amoxicillin-potassium clavulanate, cefuroxime axetil, and cefpodoxime proxetil). Parenteral agents include cefuroxime (150 mg/kg per day divided into doses every 8 hours) and ampicillin-sulbactam (200 mg/kg per day divided into doses every 6 hours). The latter combination, although not approved for children younger than 12 years, is an attractive choice. Although the use of topically applied intranasal decongestants such as oxymetazoline has not been systematically evaluated, such agents may be helpful during the first 48 hours. After several days, once the affected eye has returned to near normal, an oral antimicrobial agent is substituted to complete a 14-day course of therapy.

ORBITAL INFECTION The child or adolescent with true orbital disease secondary to sinusitis usually has sudden onset of erythema and swelling about the eye after several days of a viral URI (Figure 89-8). Eye pain can precede swelling and is often dramatic. The presence of fever, systemic signs, and toxicity is variable. Orbital infection is suggested by proptosis (with the globe usually displaced anteriorly and downward), impairment of extraocular eye movements (most often upward gaze), or loss of visual acuity or chemosis (edema of the bulbar conjunctiva). Fortunately, orbital infection is the least common cause of the “swollen eye.” Most orbital infections involve the formation of a subperiosteal abscess. In young children, such an abscess results from ethmoiditis and ethmoid osteitis. In the adolescent, subperiosteal abscess can be a complication of frontal sinusitis and osteitis. Rarely, orbital cellulitis evolves, without formation of subperiosteal abscess, by direct spread from the ethmoid sinus to the orbit via natural bony dehiscences in the bones that form the medial wall of the orbit. Imaging studies are usually performed if orbital disease is suspected. They help determine whether subperiosteal abscess, orbital abscess, or orbital cellulitis is the cause of the clinical findings (Figure 89-9). In the presence of a large, well-defined abscess, complete ophthalmoplegia, or impairment of vision, prompt operative drainage of the paranasal sinuses and the abscess is commonly performed.31–33 Several studies have reported on the successful drainage of a subperiosteal abscess via endoscopy. This method, performed through an intranasal approach, has been successful, and has avoided an external incision.34,35 In many cases, a well-defined abscess is not seen. Instead, inflammatory tissue is observed interposed between the lateral border of the ethmoid sinus and the swollen medial rectus muscle. Usually,

Figure 89-8. A 12-year-old boy with orbital cellulitis. He had a 5-day history of eye pain and progressive swelling of the eyelids, which were markedly erythematous. When his eyelids were retracted, anterior and lateral displacement of the globe and impairment of upward gaze were noted.

516

SECTION

M Infections Related to Trauma Figure 89-9. Axial (A) and coronal (B) computed tomography scans show a subperiosteal abscess extending from the left ethmoid sinus.

patients with these findings are managed successfully with antimicrobial therapy alone.33,36–38 On occasion, the computed tomography scan can be misleading, suggesting abscess when inflammatory edema is present;30,39 accordingly, the clinical course is the ultimate guide to management. Empiric antimicrobial therapy should be chosen to provide activity against S. aureus, Streptococcus pyogenes, and anaerobic bacteria of the upper respiratory tract (anaerobic cocci, Bacteroides spp., Prevotella spp., Fusobacterium spp., and Veillonella spp.) in addition to the usual pathogens associated with acute sinusitis (i.e., S. pneumoniae, H. influenzae, and M. catarrhalis).40–42 Appropriate selections include cefuroxime (150 mg/kg per day divided into doses

SECTION

every 8 hours) or ampicillin-sulbactam (200 mg/kg per day divided into doses every 6 hours). Clindamycin (40 mg/kg per day divided into doses every 6 hours) or metronidazole (30 to 35 mg/kg per day divided into doses every 8 to 12 hours) can be added if cefuroxime is used and anaerobic infection is likely. If surgery is performed, Gram stain of material drained from the sinuses or the abscess guides consideration of additional drugs or an altered regimen. When final results of culture are available, antibiotic therapy may be changed, if appropriate. Intravenous therapy is maintained until the eye appears nearly normal. At that time, oral antibiotic therapy can be substituted to complete a 3-week course of treatment.

M

Infections Related to Trauma

CHAPTER

90

Infection Following Trauma Felice C. Adler-Shohet and Jay M. Lieberman

Trauma is a major cause of morbidity in children and the leading cause of death in children older than 1 year. About 90% of childhood injuries

are secondary to blunt trauma, often caused by motor vehicle accidents, falls, or abuse.1 In patients who survive their initial injuries, infection is the leading cause of death.2 Infection that follows trauma can be due to the injury itself or, more commonly, can be a consequence of procedures and prolonged hospitalization. In one study, 10% of children admitted to a pediatric intensive care unit after trauma experienced infection.3 The literature estimates the incidence of infection following trauma in adults to be as high as 25%; wound infections or other hospital-associated infections appear to be equally likely.3 Hospital-associated infections are addressed in detail in Chapter 101 (Healthcare-Associated Infections). Trauma-associated infections, which can cause considerable morbidity, are the focus of this chapter.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Infection Following Trauma

PATHOGENESIS Severe injuries increase a child’s susceptibility to infection for several reasons. Breaks in the skin and mucosal barriers allow pathogens to gain entry, and accumulation of blood provides a favorable environment for bacterial growth. Devitalized or necrotic tissue at the injury site can harbor pathogens that can invade the host’s defense mechanisms. Foreign bodies introduced by the injury itself or as a consequence of hospitalization (e.g., catheters) allow entry and persistence of pathogens. Risk factors for infection in patients hospitalized for trauma include the severity of the injury, the presence of shock, the number of organs injured, and the amount of blood lost.4,5 Other potential risk factors are underlying host factors and the number and virulence of organisms introduced as a result of injury. Trauma can adversely affect host defense mechanisms. Monocyte activation occurs rapidly after severe injury and hemorrhagic shock. After monocyte activation, inflammatory mediators are released in a dysregulated pattern that can lead to the systemic inflammatory response syndrome6 (see Chapter 12, The Systemic Inflammatory Response Syndrome (SIRS), Sepsis, and Septic Shock). Complement is activated by injured tissue, thus decreasing complement levels, and antibody production also diminishes after blunt trauma.7 Suboptimal nutrition can further impair immune responsiveness.

GENERAL PRINCIPLES OF MANAGEMENT Patients who sustain trauma can exhibit signs and symptoms consistent with infection in the absence of infection. Retroperitoneal blood can cause fever, as can atelectasis. Pulmonary contusions can mimic pneumonia, and the signs and symptoms of hemorrhagic shock can be indistinguishable from those of septic shock. Thorough evaluation may be hampered by a patient’s immobility or diminished neurologic status and by dressings that can obscure foci of infection. The decision to begin antibiotic therapy should be made carefully, after acquisition of appropriate microbiologic specimens. Appropriate use of antibiotics may benefit the patient, but their inappropriate use can lead to adverse effects and promote the development of resistant organisms. Measures that prevent infection include rapid surgical intervention when necessary, removal of unnecessary catheters, strict attention to infection control procedures and hand hygiene, prophylactic or presumptive antibiotic therapy in some circumstances, and review of tetanus immunity. Immunity to tetanus should be assessed in any patient with an open wound. Tetanus vaccine should be given to any child in whom the primary vaccination series has been delayed or who has not received tetanus vaccine in more than 5 years. Tetanus immune globulin should be used in conjunction with the vaccine in high-risk wounds (e.g., puncture wounds, wounds contaminated with soil, feces, or saliva) if the child has not received adequate prior immunization. Children 11 to 18 years who were vaccinated against tetanus 5 years earlier and require a tetanus toxoid-containing vaccine as part of wound management should receive a tetanus toxoid, reduced diphtheria toxoid, acellular pertussis vaccine (Tdap) if they have not received Tdap previously.8 The use of prophylactic or presumptive antibiotic therapy in select situations can reduce the incidence of infection (Table 90-1); however, duration of antibiotic therapy longer than 24 hours has not demonstrated additional benefit, as discussed in detail below. Nutritional status can also affect infection risk. Patients with trauma who are given enteral nutrition have a lower risk of infectious complications than those given intravenous nutrition.9 There is conflicting evidence that injured patients fed an enteral diet rich in glutamine, arginine, and omega-3 fatty acids have a lower risk of infection than those fed a standard enteral diet or those who receive no early enteral nutrition.10,11

CHAPTER

90

517

TABLE 90-1. Common Pathogens and Recommended Prophylaxis Following Trauma Type of Trauma

Common or Important Pathogens

Recommended Prophylaxis

All trauma

Clostridium tetani

Ensure tetanus immunity or give vaccine ± tetanus immune globulin

Penetrating abdominal injuries

Enteric gram-negative bacilli and anaerobic bacteria

Ceftriaxone plus metronidazole, or cefoxitin, or gentamicin plus metronidazole

Splenectomy

Encapsulated organisms, e.g., Streptococcus pneumoniae, Haemophilus influenzae, and Neisseria meningitidis

Ensure vaccination against all three organisms. Give penicillin prophylaxis

Basilar skull fracture

Meningitis due to respiratory tract organisms

No prophylaxis recommended

Open fractures Staphylococcus aureus, (including skull) gram-negative bacilli

Nafcillin, or cefazolin, or clindamycin;a add gram-negative coverage for grade III fractures

Thoracic trauma requiring chest tube

Nafcillin, or cefazolin, or clindamycina

Staphylococcus aureus and Streptococcus species

a Vancomycin may be appropriate if MRSA is prevalent in the community or if the patient is known to be colonized with MRSA.

SKIN AND SOFT-TISSUE TRAUMA Factors associated with a greater risk of infection after injury to the skin and soft tissues are the severity of the wound, shock, and blood loss.4,7 Because wounds sustained in trauma are inherently contaminated, thorough cleansing and debridement of devitalized tissue have pre-eminent importance to prevent infection. Cellulitis can follow minor or major trauma to the skin. Rapidly spreading cellulitis that occurs within 2 days of injury is more likely due to Streptococcus pyogenes (group A streptococcus (GAS)), whereas staphylococcal cellulitis may not manifest for several days. Although GAS and Staphylococcus aureus are the most common causes of cellulitis, other pathogens should be considered in certain settings. Animal or human bites can lead to infection with Pasteurella multocida or Eikenella corrodens, respectively, as well as with anaerobes, S. aureus and GAS (see Chapter 118, Streptococcus pyogenes (Group A Streptococcus)). Body piercings and tattoos, which are increasingly common in adolescents, can also lead to bacterial infection. Although skin flora is usually the cause, Pseudomonas aeruginosa is a common cause of infection following piercing of the auricular cartilage12 and there have been outbreaks of methicillinresistant Staphylococcus aureus (MRSA) infections in tattoo recipients.13 Hepatitis B, hepatitis C, and human immunodeficiency virus (HIV) infection can also be transmitted through piercing and tattooing. Aeromonas hydrophila infection is associated with lacerations obtained while swimming in fresh water, and Vibrio vulnificus can cause skin or soft-tissue wound infections after contact with salt water or drippings from raw seafood. Mycobacterium marinum can cause cutaneous infection after freshwater or salt-water injuries from fish spines or marine shellfish, or from contamination during exposure to fishtanks.14

518

SECTION

M Infections Related to Trauma

Blood culture and needle aspirate have low yield in typical cases of cellulitis, but can be revealing in cases of cellulitis associated with animal bites or other unusual injuries. For mild to moderate cases of typical wound cellulitis, empiric therapy with a first-generation cephalosporin provides coverage for the most common pathogens, including S. aureus and GAS. The emergence of community-associated MRSA (CA-MRSA) makes empiric treatment of posttraumatic infections problematic. Skin and soft-tissue infections due to CA-MRSA are increasingly common in adults and children.15,16 Rates of MRSA colonization are increasing in children17 and colonized children would have greater risk for developing a CAMRSA wound infection following trauma. Therefore, for more severe cases of cellulitis, for wound infections requiring hospitalization, or for an infection not responding to initial b-lactam antibiotic therapy, a diagnostic specimen should be obtained and treatment should include coverage of MRSA. While many CA-MRSA are susceptible to clindamycin,18 some have inducible or constitutive clindamycin resistance.19 Most strains are also susceptible to trimethoprimsulfamethoxazole, tetracycline (which should be avoided in children < 9 years), and rifampin. For life-threatening infections possibly due to MRSA, appropriate therapy would include vancomycin or linezolid.20 Knowledge of local antibiotic resistance patterns should guide empiric therapy. Pyomyositis can occur after blunt or penetrating trauma and is usually caused by S. aureus or GAS. Treatment involves both surgical drainage and debridement and antibiotic therapy. Necrotizing fasciitis, due to GAS or S. aureus, and gas gangrene, usually due to Clostridium spp., can complicate traumatic wounds and are surgical emergencies. The diagnosis should be suspected in a patient with intense pain, often, but not always, with discoloration of the overlying skin, swelling, and crepitus. Patients with necrotizing fasciitis can also have concomitant toxic shock syndrome. Computed tomography (CT) or magnetic resonance imaging may show muscle swelling, fluid collections, or, in the case of gangrene, gas in the muscle. Leukocytosis, thrombocytopenia, a rising creatine kinase level, and hypocalcemia can be seen in these patients. Therapy for necrotizing fasciitis involves immediate surgical fasciotomy and debridement. Penicillin or nafcillin plus clindamycin should be given for streptococcal and clostridial infections.21,22 Although penicillin G is the drug of choice for invasive GAS infections, clindamycin should also be used for deep infections such as necrotizing fasciitis. In a mouse model of streptococcal myositis, clindamycin was found to be superior to penicillin.23,24 The likely explanation is that the high density of GAS leads to reduced replication and decreased expression of penicillin-binding proteins.24 Clindamycin, which inhibits protein synthesis, maintains its activity against these slowly replicating bacteria and also inhibits the production of bacterial toxins. Clindamycin should not be used alone until results of susceptibility testing are available because some GAS are resistant.

ABDOMINAL TRAUMA Trauma to the abdomen can be blunt or penetrating. Penetrating trauma is less common but carries a higher risk for intestinal injury. Either penetrating or blunt trauma can cause intestinal tears that lead to spillage of bowel contents into the peritoneum. Injuries to the colon are associated with a higher risk of infection because of the higher density of organisms in the more distal bowel.25 Blunt trauma related to physical abuse can cause hematoma of the duodenum and pancreatitis. Requirement for a large volume of transfusion is also a risk factor for infection after abdominal trauma.5 Infection can manifest acutely as peritonitis, with or without septicemia, or later, as an abscess or surgical site infection. Diagnosis is aided by CT or, in a patient who cannot be moved, ultrasonography. Needle aspiration of peritoneal fluid is performed in patients with peritonitis to confirm the diagnosis and establish the etiology. Treatment of an established infection requires empiric therapy with antibiotics effective against gram-negative and anaerobic bacteria, and

surgical or CT-guided drainage if an abscess is present. Metronidazole plus an aminoglycoside or third-generation cephalosporin would be appropriate. It is not clear that empiric therapy with antienterococcal agents improves outcome in these patients; however, antienterococcal therapy should be considered for children who have persistent symptoms when Enterococcus is isolated from peritoneal fluid cultures. Surgical exploration is part of the routine management of penetrating abdominal injuries and antibiotic therapy should be started as soon as possible. Fullen and associates26 showed that patients with penetrating abdominal injuries who received antibiotics before surgery had a lower incidence of infection than patients who received the first dose of antibiotics intraoperatively or postoperatively. Because patients with penetrating abdominal injuries are presumed to have bacterial contamination of the peritoneum at the time of injury, antibiotic therapy should be considered presumptive therapy. The empiric antibiotic regimen should include agents effective against gram-negative enteric and anaerobic bacteria, which play a critical role in intra-abdominal infections after injury.27 Single-drug therapy with a second-generation cephalosporin, such as cefoxitin, which is active against Bacteroides fragilis, appears to be as effective as traditional three-drug antibiotic regimens in preventing infection in adults with penetrating abdominal injuries.28,29 Other appropriate therapies include a third-generation cephalosporin plus metronidazole or an aminoglycoside plus metronidazole. Patients with trauma have a higher volume of drug distribution, so higher doses of aminoglycosides may be needed.30,31 Several studies have evaluated the appropriate length of presumptive therapy for patients with penetrating abdominal injuries. No significant difference in infection rate was found between patients treated for 24 hours and those treated for 5 days.25,29,32–34 Therefore, current evidence does not support the use of antibiotics beyond 24 hours in patients with penetrating abdominal injuries, even if they have colonic injuries. Longer duration of antibiotic therapy promotes bacterial resistance and fungal overgrowth and may lead to additional side effects without further reducing the likelihood of infection.35 Splenic rupture after trauma deserves special consideration (see Chapter 108, Infectious Complications in Special Hosts). If the rupture is managed nonoperatively, multiple blood transfusions may be required, placing the patient at risk for bloodborne pathogens. If splenectomy is performed, the child is at risk for overwhelming postsplenectomy septicemia, especially due to encapsulated organisms. The risk of septicemia in a child who has undergone traumatic splenectomy is 50 times higher than that in a healthy child, with the greatest risk in the first 2 years after splenectomy; therefore, observation or splenic repair is preferred if the child is clinically stable.36,37 If splenectomy is performed, immunizations should be reviewed to ensure that the child has been appropriately immunized with the Haemophilus influenzae type b and pneumococcal conjugate vaccines. Meningococcal conjugate vaccine should be administered if vaccine is licensed for the patient’s age group (currently 11 years and older). In children 2 to 10 years of age, meningococcal polysaccharide vaccine should be used until there is a licensed meningococcal conjugate vaccine for this age group. Children 2 years and older who have been adequately immunized with the pneumococcal conjugate vaccine should also be given a dose of the 23-valent polysaccharide vaccine. They should receive a second dose 3 to 5 years after the first. Partially immunized children older than 2 years should receive one dose of the pneumococcal conjugate vaccine as well as a dose of the polysaccharide vaccine 8 weeks and 3 to 5 years later. For children 5 years of age and older, some experts recommend the same sequential vaccine schedule, although either vaccine alone is also acceptable. In addition, penicillin prophylaxis should be considered for all children younger than 5 years, and for at least 1 year after splenectomy.36

HEAD TRAUMA Head trauma is common in children, and central nervous system injury increases mortality among injured children.1 As many as one-half of patients with head trauma experience an infectious complication,

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Infection Following Trauma

most often a hospital-associated infection.38 Trauma-related infectious complications include meningitis, ventriculitis, and brain abscess; common pathogens are Streptococcus pneumoniae and other upper respiratory tract organisms such as H. influenzae, Neisseria meningitidis, and GAS. Staphylococcal and gram-negative bacillary meningitis can occur after penetrating trauma, with open wounds, or after prolonged hospitalization.39 It is essential to obtain cerebrospinal fluid (CSF) for culture if the diagnosis of meningitis is being considered after head trauma; CSF cell count as well as glucose and protein level can be difficult to interpret in the presence of subarachnoid hemorrhage. Treatment includes antibiotic therapy as well as possible surgical intervention in patients with persistent CSF leakage. Basilar skull fracture accounts for up to 20% of skull fractures and is associated with substantial risk of bacterial meningitis. Meningitis is a consequence of the communication created between the subarachnoid space and the colonized paranasal sinuses, nasopharynx, and middle ear. After basilar skull fracture, the incidence of meningitis can be as high as 17% and approaches 50% in patients who also have a CSF leak. However, a 1998 meta-analysis of 12 studies involving 1241 patients with basilar skull fractures suggested that antibiotic prophylaxis did not prevent meningitis.40 Even when children and patients with CSF leakage were analyzed separately, there was no evidence that antibiotic prophylaxis decreased the risk of meningitis. In addition, the rate of meningitis in children after basilar skull fracture was only 3%, much lower than rates reported in adults. A 2006 Cochrane Database review found that antibiotic prophylaxis had no significant effect on reducing the frequency of meningitis, allcause mortality, or meningitis-related mortality in patients with basilar skull fractures.41 Available data therefore do not support the use of prophylactic antibiotics in children with basilar skull fractures. Prophylactic antibiotic is usually given following an open skull fracture, although there are few data to guide recommendation. Antibiotics chosen should have good CSF penetration and cover grampositive organisms. Coverage for gram-negative organisms may be needed if the wound is heavily contaminated or there is extensive softtissue damage.

FRACTURES Fractures are common childhood injuries and can be either closed or open. Open fractures can be graded on a scale of I to III on the basis of wound size and amount of soft-tissue damage. Grade III fractures are further subdivided (IIIA, IIIB, and IIIC) according to vascular injuries and soft-tissue defects. Dellinger and colleagues,42 evaluating risk factors for infection after open-extremity fractures, found that the severity of fracture (grade IIIB or IIIC), the placement of an internal or external fixation device, and involvement of the lower leg were independent risk factors. Organisms most commonly implicated in fracture-associated infections were Staphylococcus spp., Enterobacter spp., and Pseudomonas spp.42–44 Therapy consists of surgical removal of infected tissue and antibiotic therapy. After open fractures, infection rates range from as high as 9% for grade I fractures to 50% for grade III fractures. Infection after open fracture can lead to delayed bone healing, prolonged hospitalization, and permanent disability. Open fractures have a higher risk of infection than closed fractures, because bone and soft-tissue contamination may have occurred at the time of injury. Several studies have demonstrated a reduced risk of infection after open fractures in patients given prophylactic therapy with antibiotics effective against S. aureus.45–47 In addition to S. aureus, gram-negative bacilli such as Pseudomonas and Enterobacter can cause open fracture-associated infections, especially with grade III fractures. The Eastern Association for the Surgery of Trauma (EAST) guidelines from 2000 recommend prophylactic antibiotics with gram-positive coverage as soon as possible after injury for all trauma patients with open fractures. For grade III fractures, additional coverage for gramnegative organisms should be given. For grade I and II fractures, antibiotics should be discontinued 24 hours after wound closure. For

CHAPTER

90

519

grade III fractures, antibiotics should be continued for only 72 hours after the time of injury or not more than 24 hours after soft-tissue coverage of the wound, whichever comes first. A Cochrane Database review from 2000 also found evidence for antibiotic prophylaxis for surgery for closed long-bone fractures.48 The use of antibiotic-impregnated beads, cement, and polymers has shown some promise in reducing open-fracture-associated infections.44,46,49 Prospective randomized trials are needed to determine effectiveness.

PUNCTURE WOUNDS The management of infection after a puncture wound to the foot deserves special consideration. Up to 18% of these wounds are complicated by cellulitis or a soft-tissue abscess, and up to 2% by osteochondritis. Pseudomonas aeruginosa is responsible for as many as 90% of cases of osteochondritis after a puncture wound to the foot.29 The liner of sneakers has been found to contain Pseudomonas, making it a likely source; however, Pseudomonas can cause infection after a puncture wound through other shoes and even bare feet, and at other sites.50 Other gram-negative bacilli as well as S. aureus and GAS can also cause osteochondritis following a puncture wound. Eikenella corrodens or oral anaerobic bacteria can cause infection following a toothpick puncture injury to the foot.51,52 Rarely, nontuberculous mycobacteria can be the primary pathogen.53,54 Surgical debridement is the mainstay of management, both to obtain specimens to determine infectious etiology and to remove necrotic cartilage and any retained foreign body. After debridement, antipseudomonal therapy, if necessary, should only be continued for a week. Osteochondritis due to other organisms should be treated with a conventional course of therapy.

THORACIC TRAUMA Thoracic injuries are a major cause of death in childhood trauma. Such injuries are usually blunt and are often due to a motor vehicle accident.1 Significant chest trauma can lead to pulmonary contusion, hemothorax, and pneumothorax. Although any major trauma increases the risk of hospital-associated pneumonia due to prolonged intubation, poor pulmonary toilet, and the risk of aspiration, pulmonary contusions also lead to decreased pulmonary function, thereby further increasing the risk of pneumonia. The diagnosis of pneumonia can be challenging because fever and a pulmonary density can be due to the contusion itself. New fever, leukocytosis, purulent secretions, and the need for increased ventilator settings are clues to the diagnosis. The antibiotic regimen for suspected pneumonia in this setting should consist of agents effective against nasopharyngeal organisms, including anaerobic and gram-negative bacteria. Clindamycin plus either a third-generation cephalosporin or an aminoglycoside is appropriate, as are b-lactam plus b-lactam-inhibiting agents. However, antibiotic choices should be based on knowledge of the pathogens commonly identified in the local intensive care unit as well as results of culture of deep respiratory tract specimens, such as those collected by bronchoalveolar lavage when possible. Hemothorax due to blunt or penetrating thoracic trauma increases the risk of empyema, as does a thoracostomy tube placed to drain a hemothorax or pneumothorax. The risk increases with incomplete drainage of the pleural space or prolonged chest tube placement.55 Staphylococcus aureus and gram-negative bacilli predominate in these infections.56 Prophylactic antibiotic therapy has been studied in the setting of tube thoracostomy after chest trauma. A meta-analysis of nine prospective trials and two earlier meta-analyses found that prophylactic antibiotic therapy reduced the risk of subsequent pneumonia after tube thoracostomy.57 A brief (24-hour) course of an antibiotic with good antistaphylococcal and antistreptococcal activity is appropriate after chest tube placement. Antibiotic prophylaxis for < 24 hours is also recommended when open thoracotomy is performed for penetrating chest injury.58

520

SECTION

CHAPTER

M Infections Related to Trauma

91

Infection Following Burns Jane M. Gould and Gail L. Rodgers

In 2002, unintentional injury was the leading cause of death in children 1 to 18 years of age in the United States. Of these deaths, fire/burns were the fourth leading cause in children less than 1 year of age and the third leading cause in children 1 to 18 years of age.1 Children aged 4 years and under are at the greatest risk, with an injury death rate more than twice that of children aged 5 to 14 years.2 Many more children suffer serious disability. The vast majority of these injuries are preventable. Most young children suffer scald injuries from hot liquids or flame burns from house fires. Older children are more likely to sustain flame burns from accidents with flammable liquids and from house fires. On occasion, burn injuries can result from medical therapies, such as therapeutic application of heat, ignition of flammable medications (rubbing alcohol and hot oils), or burns from hot-air vaporizers.3 The survival of children with burns depends on the following factors: (1) age; (2) the percentage of total body surface area (TBSA) burned; (3) the depth of the burn injury; (4) the type of burn; and (5) management. Young children, particularly those younger than 2 years, have a higher incidence and a lower survival for the same amount of body surface area burned than older children and adults. Advances in burn care (improvements in resuscitation, intensive care, and care of the burn wound) have narrowed the gap in survival for small children.4 Currently, the extent of burn associated with a 50% survival in patients older than 1 year is approximately 80%.5 Data suggest that the most common causes of death in burned children are smoke inhalation and hypoxic–ischemic brain injury. A trend of decreased fatal septicemia has been seen in centers, mainly as a result of aggressive early burn excision, prompt surgical treatment of infected wounds, and judicious use of systemic antibiotics.6 The most reliable predictor of outcome of burns is the percentage of TBSA burned, followed by depth. TBSA can be estimated either from the Lund and Browder chart or with use of the size of the patient’s palm, which is roughly 1% of the TBSA at any age, as a measure.7 Depth of burn injury is determined by the extent of damage to tissue and is classified by degrees. A first-degree burn involves the epidermis only, is painful, red, and dry, and resembles a sunburn. A second-degree burn involves the dermis, is severely painful, is usually erythematous, moist, and weepy, and may have blisters and bullae. A third-degree burn involves the subcutaneous tissue and is usually white or waxy-appearing, dry, avascular, and painless. Depth is alternatively classified as either partial-thickness, involving the epidermis or superficial dermis, or full-thickness, involving the deep dermis and subcutaneous tissue. The larger the percentage of TBSA involved and the deeper the burn, the higher the mortality, both immediately after injury (primarily from shock or occasionally from other associated injuries) and after successful resuscitation (from infectious complications). The type of burn is important in survival, because scald burn, the most common childhood burn, is less commonly fatal than flame burn, especially if the latter is associated with pulmonary injury. The enormous progress in burn therapy is attributable to the following factors: care of patients in centers specializing in the treatment of burns; knowledge of the pathophysiology of shock and aggressive treatment of patients with fluid resuscitation and other adjunctive therapies; recognition of the importance of the caloric needs of the burned patient and its role in wound healing; and advances in the care of the burn wound itself, especially early debridement and excision of the wound, use of topical antimicrobial agents, and improved grafting techniques.

The incidence of serious infections in burned patients increases in proportion to the percentage of TBSA burned. Children with burns > 30% of TBSA, flame and inhalation injuries, and full-thickness burns are at highest risk of infectious complications.8 Burn victims are susceptible to a wide variety of infections associated with relative immunosuppression (which occurs with burns of 30% TBSA or more) and complications of intensive care. Virtually any organ can become the target of an infection in such patients. The most common infections in burned children are those related to the burn wound and catheter-associated septicemia.8 Infections related to intravascular and urinary catheters and endotracheal tubes are discussed in Chapter 102 (Clinical Syndromes of Device-Associated Infections). Infection of the burn wound occurs with greatest frequency in children.9 Burn wound septicemia is associated with at least an 80% mortality in children. Diagnosis of infectious complications in a burn victim is challenging. Although fever and elevated peripheral white blood cell count with a left shift are usual indicators of infection, their positive predictive value for diagnosis of infectious complications in burn victims is very low because they are commonly seen in uninfected burned children whose wounds are uncovered.10 Neither severity of fever (frequently > 39°C) nor response to antipyretic therapy is a reliable indicator of infection.10 Peak fever in burned children without infection usually occurs on the second day after the burn, with a second peak around the sixth and seventh days.11 Fever is probably the result of an increase in metabolic rate and an alteration of hypothalamic temperature regulation. Fever usually subsides without specific therapy, coincident with re-epithelialization of the burn wound or successful grafting of all open areas. Thus, fever alone in the burned child is not a reliable indicator of infection or of the need to investigate for infection or prescribe antibiotic therapy. Hypothermia is usually a more reliable indicator of infection, although children with burns affecting a high percentage of TBSA who are left uncovered for prolonged periods during dressing changes may demonstrate hypothermia that requires external warming. Hypotension due to fluid shifts is common early in the postburn period and can usually be differentiated from infectious causes because it can be rapidly corrected with fluid resuscitation. Several inflammatory mediators have been evaluated in attempts to distinguish infection from the normal response to thermal injury. They include tumor necrosis factor-a,12–14 interleukins (IL) 1b,12,14 6,12–14 and 8,13,15 procalcitonin,16,17 and C-reactive protein.18 Levels of mediators in burned patients were higher than those in controls, probably secondary to nonspecific inflammation associated with thermal injury. Although several studies have shown significant differences in levels of these inflammatory mediators between burned patients with septicemia and those with uncomplicated thermal injury,12,13,15–17 none is highly discriminatory. However, changes in serial C-reactive protein levels may permit detection of septicemia an average of 2.3 days sooner than using a decrease in platelet count or clinical manifestations and is more useful than procalcitonin level.18 Thus, diagnosis of infectious complications is made through evaluation of all clinical signs, examination of the burn wound and all catheter sites, and consideration of supporting laboratory evidence, and results of culture of blood, urine, or quantitative burn wound biopsy specimens.

BURN WOUND INFECTION Burn wound infection is defined as the invasion of microorganisms into viable tissue under the wound. Local infection can result in prolonged wound healing or sloughing of graft, toxin production leading to distant organ damage, and septicemia and infection at distant sites.19 The diagnosis of burn wound infection is made from local signs of wound infection, with or without systemic signs indicating septicemia or toxemia (Box 91-1), in conjunction with histologic and microbiologic evidence of infection in biopsy specimens of burn wounds (Box 91-2). The technique for quantitative biopsy requires a 1 gram

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Infection Following Burns

CHAPTER

BOX 91-1. Clinical Characteristics of Burn Wound Infection

TABLE 91-1. Immunologic Dysfunction in Burned Children

LOCAL SIGNS Focal areas of discoloration or necrosis Edema, erythema, discoloration of wound margin Conversion of partial- to full-thickness burn Unexpectedly rapid eschar separation Hemorrhagic discoloration of subeschar tissue Purulent exudate on burn wound SYSTEMIC SIGNS Hyperthermia Hypothermia Hypotension Altered mentation Glucose instability Organ dysfunction

Type of Dysfunction

Adapted from Pruitt BA Jr, Yurt RW. Treating burn and soft tissue infections. Infect Surg 1983;2:623–650.

91

521

References

HUMORAL IMMUNE FUNCTION

Decreased numbers of B lymphocytes Decreased total immunoglobulins Decreased fibronectin Aberrant production of immunomodulators

25–27 27–35 26, 27, 36–38 24, 26, 28, 39, 40

CELLULAR IMMUNE FUNCTION

Anergy Decreased mitogen and antigen responses Increased suppressor T lymphocytes Decreased helper T lymphocytes

26, 27, 41 26, 29, 42 27, 28, 42–44 28, 42, 43

NEUTROPHIL FUNCTION

Decreased adherence Decreased phagocytosis Decreased killing Decreased chemotaxis

36 29, 45–47 27, 29, 45, 46, 48, 49 26, 27, 29, 45, 46, 50, 51

OTHER

BOX 91-2. Biopsy Findings of Burn Wound Infection HISTOLOGY Characteristics of tissue underlying burn Presence of microorganisms Thrombosis or hemorrhage Necrosis Intense inflammatory response Presence of intracellular viral inclusions MICROBIOLOGY Positive quantitative Gram stain reaction Isolation of single or multiple organisms, each > 105 colony-forming units per gram of tissue Adapted from Pruitt BA Jr, McManus AT. Opportunistic infections in severely burned patients. Am J Med 1984;76:146–154.

specimen of eschar, which is homogenized and cultured.20 This procedure consists of cleaning the open wound with alcohol and obtaining the specimen with a scalpel or a dermal punch. Processing of the biopsy consists of aseptic weighing, alcohol dip and flaming to remove surface contamination, dilution in fixed-volume thioglycolate broth or saline, homogenization, and inoculation on to nutrient agar.21,22 Although selection of the biopsy site, retained activity of topical antimicrobial agents, and multistep processing can be sources of error, quantitative biopsy culture yielding growth of a single or multiple organisms, each with a density of > 105 colony-forming units per gram of tissue, correlates with infection. Histologic diagnosis of infection is supported by the presence of bacteria invading viable tissue and can be obtained with frozen or permanent sections, the latter of which is thought to be more accurate.20 Blood culture results are positive in only 40% to 50% of patients with burn wound septicemia. Factors that influence burn wound infection pertain to the wound, the host, and the causative organisms.

Wound Factors Skin is a primary, critical local defense mechanism against infection, providing a mechanical barrier to penetration of organisms that normally reside on the surface. Skin saprophytes are thought to inhibit colonization by more pathogenic bacteria. Skin also produces antibacterial substances, such as unsaturated free fatty acids, that inhibit a number of microorganisms, particularly group A streptococcus (GAS).23 Thermal injury rapidly disrupts normal functions and produces an ideal culture medium. After burn injury, the wound site rapidly becomes colonized with normal skin flora and gram-positive pathogens, and then by endo-

Decreased macrophage and monocyte function Complement-induced increase in release of immunosuppressive mediators

52 27

genous or environmental gram-negative bacilli and fungi if the hospital course is protracted. Isolation of microbes from a burn wound is not a priori evidence of infection, because invasion into underlying viable tissue must occur for infection to develop. Adequacy of blood supply to the wound is critical. Avascularity of the burned tissue results from coagulation of vessels and tissue that make a protein-rich eschar (an ideal site for microbial proliferation) and renders the site inaccessible to systemic antibiotics as well as humoral and cellular defenses. Infection in the burn wound can extend into the blood vessels, causing thrombosis that can compromise the blood supply further, converting a partial-thickness burn into a fullthickness burn. Wound factors such as the acidic, anaerobic, moist environment of the avascular burn tissue favor growth of certain pathogens such as fungi and impair activity of aminoglycoside antibiotics. The cooler temperature of the burn wound may influence infection, because lower temperatures restrict blood flow, possibly causing further tissue necrosis and impairing phagocytic cell metabolism.24 However, cooler temperatures may also limit microbial multiplication. In addition, the location of the wound can contribute to the overall risk of infection, as is seen with exposure of the globe, bones, cartilage, and joints.20 Likewise, the presence of foreign bodies in the wound can also promote infection.

Host Factors Several host factors have been identified that influence the likelihood of burn wound infection. Children with underlying medical conditions, such as diabetes, neurologic disorder, and immunodeficiency, are more likely to experience burn wound septicemia, resulting in very high mortality. Dysfunction of the immune system has been well described in patients who sustain burn injuries of > 30% of body surface area, the degree of which correlates directly with TBSA burned. Abnormalities have been documented in all aspects of humoral and cellular immune function (Table 91-1). The mechanism is unknown. It has been suggested that immunosuppression after injury evolves to protect against autoimmunity that might otherwise result from antigenic bombardment after intense tissue injury.42,53,54 Immunosuppression of burned patients greatly increases their susceptibility to infection. Polymorphisms in the genes that encode Toll-like receptor 4 and tumor necrosis factor-a are significantly associated with an increased risk for severe septicemia following burn trauma.55

522

SECTION

M Infections Related to Trauma

Causative Organisms For a patient with a disrupted mechanical barrier and suppressed immune function, any organism is a potential pathogen. Organisms causing burn wound infection have changed over the past century. In the 1930s and 1940s, GAS was the predominant pathogen, followed by Staphylococcus aureus. In the 1950s and early 1960s, the predominant pathogen in such infections was S. aureus. Gram-negative bacilli, especially Pseudomonas aeruginosa, became the predominant pathogen in the 1960s and early 1970s. Since then, S. aureus has regained prominence (becoming the most common pathogen in many centers8,56), including methicillin-resistant S. aureus (MRSA), the spectrum of gram-negative bacilli has broadened, and fungal and viral pathogens have become increasingly important. These changes are direct reflections of complications of aggressive therapy and survival of more severely affected patients. The number of organisms present on the burn wound is an important factor in infectivity. It is theorized that high temperature initially sterilizes the burn wound. However, rapid colonization by normal skin flora and existent pathogens follows. At the time of hospitalization in our pediatric burn unit, routine cultures reveal that 9% and 54% of patients are colonized with GAS and S. aureus, respectively. Colonization of the wound surface is not equivalent to infection. Usually, a bacterial density of 105 organisms per gram of tissue is required before invasion of underlying viable tissue (burn wound infection) occurs. Infection can occur with lower bacterial densities, but rarely. The most common organisms to cause wound infections are S. aureus and P. aeruginosa.20 An organism’s unique virulence, invasiveness, and motility also affect pathogenicity. It has been demonstrated that nonmotile strains of Pseudomonas spp. rarely cause infection.57

patients who have not had boosters within 60 months. Tetanus immunoglobulin is given in addition to toxoid to the patient with heavily contaminated wounds whose immunization status is unknown or who has received fewer than three doses of adsorbed tetanus toxoid.61

Later Infections During the second week after injury, gram-negative bacilli continue to be important pathogens, but the burn wound also becomes colonized with fungi. Risk factors associated with fungal wound infection are preceding antibiotic therapy, presence of indwelling central venous catheters, and infusion of parenteral nutrition, especially of solutions containing lipids. Colonization with Candida spp. is common in burned patients, but burn wound septicemia is infrequent. When invasive infection occurs, mortality exceeds 90% despite aggressive medical and surgical therapy.62 Fungi such as Aspergillus spp. and Fusarium spp. and members of the Mucoraceae family are rare causes of infection; infection with such organisms is invasive, however, and usually fatal. These infections frequently follow successful treatment of gram-negative bacillary infection with broad-spectrum antibiotics, and they occur in patients with acidosis. Fungi invade tissue rapidly and cause thrombosis, leading to tissue infarction and systemic dissemination. Extremely aggressive surgical measures are indicated urgently, but mortality remains high. Other fungi isolated from burn wounds are Geotrichum, Rhodotorula, Cephalosporium, Penicillium, Trichosporon, Trichophyton, and Fonsecaea species; they do not usually invade tissue.63

Special Sites of Infection

Early Infections Specific colonization of burn wounds is somewhat predictable over time. Initially, gram-positive organisms are present; infection that occurs in the first 48 hours after the burn is usually secondary to GAS. Because streptococcal infection can be rapidly invasive and fatal, it was once common practice to administer penicillin prophylactically to all burned patients at the time of admission. The incidence of GAS infections in burned patients has decreased, probably secondary to immediate use of topical antimicrobial therapy. Routine administration of penicillin prophylaxis is not recommended because it may hasten colonization and potential infection with more resistant organisms. S. aureus also causes early septicemia. If there is concomitant inhalation injury, organisms that colonize the respiratory tract can cause bacteremia and invade the burn wound. In the latter part of the first week after injury, environmental and endogenous gram-negative bacilli colonize the wound. P. aeruginosa and Enterobacteriaceae are usual. Colonization of the gut is probably the primary event, followed by bacterial translocation and wound colonization. Bacterial translocation is frequent in burned patients because of the disruption of the mucosal barrier from nutritional factors, immunosuppression, disturbance of the normal barrier of bowel flora (from selective pressure of antibiotic administration), and overgrowth of hospital-associated pathogens. Early institution of enteral nutrition satisfies caloric requirements, improves immune function and tissue healing, and preserves the mucosal barrier, thereby decreasing the incidence of septicemia in burn victims.58 Another proposed mechanism for acquisition of environmental organisms and subsequent infection is immersion of the patient in water or rolling of the patient in soil in an attempt to control the burning at the time of injury.59 Complication of burn injuries by tetanus is uncommon in the United States. In the nonimmune patient who has a wound contaminated with Clostridium tetani, neurologic manifestations of tetanus toxin and minor local infection can occur in the first week after injury.60 Tetanus immunization is routinely administered to burned

Anaerobic infections are rare in burned pediatric patients. Anaerobic bacteria have been found to colonize burn wounds around the mouth and anus and can have a synergistic role in burn wound infection, however.64 Infections of burns involving the ear cartilage can result in suppurative auricular chondritis, which presents with pain, fever, and rapidly progressive edema of the auricle followed by liquefaction of cartilage. Avascularity of this tissue makes treatment difficult. Iontophoresis, a technique that uses direct current to drive charged compounds (antibiotics) into local tissues, has been used in conjunction with topical antimicrobial therapy and grafting, with encouraging results.65,66 Corneal infections secondary to direct thermal injury, chemical burn, or ectropion and desiccation can cause permanent scarring requiring corneal transplantation. In addition, infected corneal ulcers can perforate and cause herniation of the lens and loss of the eye. Globe exposure secondary to progressive contracture of burned eyelids and facial skin can require acute eyelid release.67 Pyomyositis can occur secondary to vascular compromise in deep burns that leads to muscle necrosis or unsuspected deep muscle injury in electrical burns. Intracompartmental infection in which there may be associated pyomyositis can result from delayed escharotomy or extravasation of infused fluids or can occur as a complication of splinting and positioning.68 Treatment is surgical. Osteomyelitis and septic arthritis can occur when bone and joints are exposed.

OTHER INFECTIONS Respiratory Tract Infections Pneumonia is the most common infectious cause of death in the burned patient. In patients with inhalation injuries, pneumonia occurs most frequently in the first week after injury. Thermal damage produced by inhalation injury predisposes the airway to infection by:

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Infection Following Burns

(1) producing structural damage to the respiratory tract epithelium; (2) impairing surfactant production, mucociliary transport, and macrophage function; and (3) producing atelectasis.69 Pneumonia or tracheobronchitis occurs in up to 35% of those with inhalation injury.20 Less commonly, pneumonia can result from hematogenous spread of infection from other areas, most notably the infected burn wound; this typically occurs late in the postburn period. Chest tubes that must be placed through a wound can result in empyema; they should be removed as soon as is practical.20 Sinusitis and otitis media are infectious complications of the child requiring nasotracheal intubation for inhalation injury.

CHAPTER

91

523

Central Nervous System Infections Infectious complications in the central nervous system can occur in burned patients. They are usually associated with bacteremia or endocarditis in the patient with a burn > 30% of TBSA, correlating with a higher level of immunosuppression and an inability of the host to localize infection. In a postmortem review of burned victims, 53% had central nervous system complications, 16% of which were infectious. Microabscess, septic infarction, and meningitis occurred, and most were caused by S. aureus, Candida spp., or P. aeruginosa.77

Viral Infections Bacteremia Bacteremia is not uncommon in the burned patient. Risk factors include wound manipulation and the presence of an intravascular catheter. Suppurative thrombophlebitis or infected intravascular thrombus can cause persistent bacteremia. Endocarditis must be considered in any patient with prolonged bacteremia. Daily dressing changes and surgical wound debridement have been associated with bacteremia in 7.7% to 65% of episodes.70–73 The greater the percentage of TBSA burned (and thus manipulated), the higher the risk of bacteremia. Use of prophylactic systemic antibiotics before burn wound manipulation, particularly debridement, has been shown to reduce the incidence of bacteremia but has not had a beneficial effect on subsequent clinical course or incidence of burn wound infection.70,72,73 Bacteremia can occur in the absence of wound manipulation or other identifiable risk factors, presumably from translocation of gut organisms. Bacteremia secondary to burn wound manipulation or gut translocation is usually transient and does not result in infection at distant sites or interfere with graft adherence. It is possible, however, that the higher incidence of endocarditis demonstrated in burned patients is attributable to the increased number of episodes of transient bacteremia. Pyogenic arthritis and brain abscesses secondary to S. aureus bacteremia have been described.74,75

Catheter-Related Infections Burned patients have a high risk of acquiring intravascular catheterrelated infections. Catheters are used extensively for volume resuscitation, medications, blood products, and alimentation. In one study, the rate of catheter infection correlated inversely with the distance of the catheter insertion site from the burn wound; infection occurred more commonly when the insertion site was < 30 cm from the burn wound.76 Suppurative thrombophlebitis, both peripheral and central, can also occur in burned patients and is difficult to diagnose if the affected vein underlies the burn. These conditions may require prolonged antibiotic therapy, systemic anticoagulation, or surgery.20 Urinary tract infections are common in burned patients, being associated with the urinary catheters commonly used to care for perineal burns or to monitor fluid status. Catheters placed for fluid status monitoring should be removed as soon as the patient’s clinical status is stable, and their use should never be prolonged for the convenience of caregivers.

Intraabdominal Infections Acute cholecystitis can occur in older children and adolescents with burns if enteral nutrition is not instituted. Findings include fever without localizing signs and serum hepatic enzyme abnormalities consistent with cholestasis (mimicking findings in septicemia). Acute pancreatitis as well as pancreatic abscess can also be seen in severely burned patients. Peritonitis, as a result of splanchnic ischemia, increased gastrointestinal permeability, and bacterial translocation can occur.20 Early enteral feeding decreases the likelihood of these complications.

Viral infections complicate the hospital course of a burned patient and must be considered in any burn victim with unexplained fever whose burn wound is healing adequately.78–81 These infections can occur at any time but are most common in the second week after injury. They can be due to reactivation or acquisition. Herpes simplex virus (HSV) and cytomegalovirus (CMV) have been isolated from up to 30% of pediatric burned patients.78 Their presence correlates directly with the surface area involved. CMV infection is often asymptomatic, being diagnosed solely on the basis of a fourfold increase in antibody titers. It is a purported cause of unexplained fever in burned patients as well as of hepatitis, neutropenia, and thrombocytopenia. Other serious manifestations of CMV infections, such as pneumonia, are infrequent. CMV infection can result from reactivation of latent infection or from primary infection. The cell-mediated immune dysfunction experienced by burned patients enhances reactivation of CMV and exacerbates symptomatic infection. Primary infection can be acquired from blood transfusions or cadaveric skin allografts from seropositive donors. CMV infection is associated with greater risk of bacterial infection in patients with burns. Orolabial reactivation of HSV is common, and the virus can disseminate locally or cause infection of the burn wound, impeding wound healing. Herpetic infections have been treated successfully with intravenous acyclovir. Disseminated infection with HSV has also been described in burned patients.82,83 Other viral infections associated with excessive morbidity, delayed wound healing, or unexplained fevers are varicella-zoster virus, Epstein–Barr virus, and adenovirus. One study, performed before testing of the blood supply for hepatitis C virus (HCV), demonstrated that 18% of burned patients acquired HCV. Chronic hepatitis was observed in 83% of burned patients, which is a high rate in comparison with that in other populations.84 Nosocomially acquired respiratory syncytial virus, adenovirus, and rotavirus add considerable morbidity during the protracted hospitalization of burned patients.

TREATMENT Specialized care in burn centers, the surgical trend toward aggressive early and repeated debridement and wound closure, and the routine use of topical antimicrobial agents contribute immensely to the prevention and control of infections after burns. Topical antimicrobial therapy is a mainstay of burn care; such therapy has had an enormous impact on the rates of wound infection and septicemia. The general objectives of topical therapy are to decrease water vapor loss, prevent desiccation of exposed viable tissues, contribute to pain control, and inhibit bacterial and fungal growth.20 Although topical agents do not sterilize the wound, numbers of colonizing organisms decrease, thus reducing the risk of bacterial invasion of underlying tissue. The topical agents commonly used in burned patients differ in their antimicrobial spectrum and side effects (Table 91-2). Other topical agents used in burn care are discussed in Chapter 294 (Topical Antimicrobial Agents). There is no ideal topical antibiotic. An agent is selected on the basis of known or expected colonizing organisms, timing of injury, knowledge of organisms

524

SECTION

M Infections Related to Trauma

TABLE 91-2. Topical Antimicrobial Agents Commonly Used in the Treatment of Burns Antimicrobial Activity Against Agent

Gram-Positive Organisms

Gram-Negative Organisms

Anaerobes

Fungi

Adverse Effects

Silver sulfadiazine (Silvadene)

++

+++

++

++

Allergic reaction in 5% Leukopenia

Mafenide acetate (Sulfamylon)

+

+++

++

+

Allergic reaction in 10% Pain Metabolic acidosis Pulmonary edema

Nitrofurazone (Furacin)

+++

+++

++

–

Dermatitis Fungal overgrowth

+++, excellent activity; ++, good activity; +, limited activity; –, no activity.

indigenous to the burn unit, and possible adverse reactions. Initially, silver sulfadiazine is usually used, and subsequent choices are based on results of isolation of organisms from the wound. Silver sulfadiazine has fair to poor eschar penetration, unlike mafenide acetate, a carbonic anhydrase inhibitor, which is capable of eschar penetration.20 Debridement of the wound is essential. Both enzymatic and chemical means of debridement have been used in the care of burns as alternatives to surgical excision. Proteolytic and mucolytic enzymes rapidly debride nonviable tissue, but they have been associated with a higher risk of septicemia.85 Temporary membranes are also available to use on superficial wounds or donor sites to decrease infection and facilitate comfort.86 Fresh or reconstituted porcine xenograft, synthetic bilaminates, hydrofibers, semipermeable membranes, hydrocolloid dressings, and human allograft are examples, some of which are impregnated with silver to reduce bacterial and fungal growth. Patients with heavily contaminated wounds or septicemia or both are best managed with allografts initially and later with autografts.20 Systemic antimicrobial therapy is only used when systemic infection is strongly suspected, but its role is adjunctive and it never replaces aggressive surgical debridement and the use of topical agents. Parenteral agents should be of the narrowest antimicrobial spectrum possible, being directed specifically at known pathogens, to avoid colonization with resistant bacteria or fungi. In burned patients, the pharmacokinetics of antibiotics are altered as a result of the multiple pathophysiologic changes that occur after the burn injury.87,88 These changes must be considered in selection of agents and optimal dosage. Aminoglycosides have a short elimination half-life, thus requiring increases in both dose and frequency of administration.89–91 Oncedaily aminoglycoside dosing (8 mg/kg) in burned patients (TBSA burned, 18% to 81%) has been evaluated and shown to cause elevation of creatinine > 0.5 mg/dL in 35% of patients and ototoxicity in 1 of 13 patients.92 Thus, without further study, once-daily aminoglycoside dosing cannot be recommended in burned patients. Vancomycin clearance is increased in burned patients, possibly because of increased renal tubular secretion; higher dosages may be necessary.93 Because burned patients may have renal dysfunction, serum levels of aminoglycosides or vancomycin are monitored to allow individual adjustment of dosages. The pharmacokinetics of ceftazidime, ticarcillin-clavulanate, imipenem-cilastatin, and aztreonam have been evaluated in burned patients.88,94–97 Alterations were found for all except imipenemcilastatin (i.e., increased total clearance and volume of distribution for ceftazidime and ticarcillin/clavulanate, and increased volume of distribution for aztreonam). Creatinine clearance does not adequately predict dosage requirements for these drugs; thus, specific dosage recommendations were not made in any study. It seems prudent to use the highest recommended dose in a burned patient who does not have evidence of renal or hepatic impairment.

MANAGEMENT OF OUTPATIENTS Partial-thickness burns covering a small TBSA (< 10% to 15%) can be treated on an outpatient basis in the following circumstances: (1) the burns are not circumferential or do not involve the face, hands, feet, or perineum; (2) no other injuries are present; (3) child abuse is not suspected; and (4) the caregivers are capable of adhering to management and follow-up. Appropriate treatment of smaller burns is imperative, because inadequate care can lead to wound infection, septicemia, poor wound healing, and a poor functional or cosmetic result. Outpatient management of burns consists of cleaning the wound thoroughly and removing devitalized tissue. Small blisters may be left intact. Controversy exists as to the best management of large blisters. Because large blisters can interfere with wound healing, some physicians advocate debridement, whereas others believe that it raises the risk of infection. After cleaning and debridement, gauze dressing containing a broad-spectrum topical antibiotic such as silver sulfadiazine is applied. The dressing is changed with new application of antibiotic twice daily, and the parent or caregiver is educated about the expected changes in appearance of the wound. The wound is inspected 24 to 48 hours later for expected healing and signs of local infection. At follow-up, the physician ensures that parents have an adequate understanding of wound care.

PREVENTION Prevention of infection in a burn is paramount to prevent mortality and morbidity. Urgent and adequate care of all burns and use of dressings impregnated with topical antimicrobial agents, followed by meticulous care and assessment, are usually successful. Early closure of the burn wound is the most protective treatment against infection. Pulmonary toilet in the patient with inhalation injury and rapid removal of intravenous and urinary catheters are essential in prevention of nosocomial infections. Strict attention to infection control practices is essential. Routine use of prophylactic antibiotics is not beneficial in prevention of infections and can predispose to colonization with resistant organisms.98 Replacement therapy with immune globulin intravenous (IGIV) has been studied, because most burned patients show a decrease in serum immunoglobulin (Ig) G concentration immediately after injury (mean values, 300 mg/dL), with a gradual return to normal after 14 to 60 days.34,35 In one study, patients younger than 4 years had the most severe and prolonged decreases of IgG.30 Patients with persistent serum IgG concentrations of < 500 mg/dL may have a higher risk of septic complications.31 Although IgG levels normalized with administration of IGIV in the patients studied (from mean of 400mg/ dL before infusion to > 1200 mg/dL after), no beneficial effect on

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Infection Following Bites

development of septicemia was evident.99 Administration of hyperimmune globulin G against P. aeruginosa and S. aureus has been used as adjunctive treatment for septicemia caused by these organisms in burned patients, with evidence of a beneficial effect.100–102 Interferon-g, because of its ability to enhance macrophage activity and its many other actions that stimulate the host immune system, has been evaluated as prophylaxis for infectious complications in burned patients. In a European multicenter trial of 216 patients with severe burn injuries, administration of interferon-g did not give protection from infectious complications or affect the mortality from infections.103 A vaccine directed against P. aeruginosa has also been developed and shown to be efficacious.104–108 In one study of 322 burned patients, administration of maximal-dose vaccine reduced the rate of invasive disease. A lower mortality was only seen, however, in patients who received parenteral antibiotic therapy and adjunctive hyperimmune globulin. Vaccination did not prevent colonization with P. aeruginosa, and 90% of the patients had an adverse reaction to vaccination.105 Additional controlled studies are required to assess the benefit of immunotherapy, in view of changes in predominant pathogens, improved supportive measures, and the trend toward early excision and wound closure.

CHAPTER

92

Infection Following Bites Marvin B. Harper

In the United States, it is estimated that 4.7 million dog bites, 400,000 cat bites, and 250,000 human bites occur every year.1 Dog bites alone result in approximately 330,000 emergency department visits (42% among children < 14 years of age), 13,000 hospitalizations, and 20 deaths annually (mostly children).2 Dogs, cats, and humans account for > 90% of noninsect-related bite injuries, with rabbits and rodents responsible for most of the remainder.1,3 The trauma caused by these bites can be quite serious, and infection is the most common late complication.

ETIOLOGIC AGENTS The organisms causing bite wound infections generally derive from the microbial flora of the biting animal’s mouth rather than the victim’s skin. Therefore, the infecting organisms vary by species. Bite wound infections should be considered polymicrobial infections; in a large, prospective study of infected dog and cat bites, cultures done at a central reference laboratory yielded a median of five bacterial isolates per culture.4 Mixed aerobic and anaerobic infection occurred in more than 50% of cases. Specimens sent concomitantly to local microbiology laboratories identified significantly fewer organisms, emphasizing the need for careful microbiologic analysis. Table 92-1 shows the aerobic and anaerobic bacteria isolated from the culture of bite wound infections from 50 patients with dog, 57 with cat, and 50 with humanbite wounds.4,5 Pasteurella species were the most common isolates from both types of bites, with P. canis the predominant organism isolated from dog bites and P. multocida the most common from cat bites. Other species isolated from these patients included streptococci, staphylococci, and Moraxella and Neisseria species. Common anaerobic isolates included Fusobacterium, Bacteroides, Porphyromonas, Prevotella, Proprionobacterium, and Peptostreptococcus species. A number of other

CHAPTER

92

525

bacterial isolates had not been previously recognized as bite wound pathogens. Other studies confirm that wound infections resulting from human and animal bites are often polymicrobial and represent mixed aerobic and anaerobic infections.6,7 Pasteurella multocida is only seen in animal bites, more commonly from cats,8 whereas Eikenella corrodens and Streptococcus pyogenes are more closely associated with human bites, although they are sometimes seen with animal bites as well. Capnocytophaga canimorsus,9,10 C. cynodegmi,9 Neisseria weaveri (formerly M-5),11,12 Bergeyella zoohelcum,13 Neisseria canis,14 Staphylococcus intermedius,15 NO-1,16 and EO-217 are all uncommon organisms to recover from general clinical specimens but are isolated from bite wound infections, particularly from dog bites. The cause of infections resulting from bites of species other than humans, dogs, and cats is less well described. Simian bites appear to be quite similar to human bites in microbiology.18 Polymicrobial infections with Staphylococcus aureus, Streptococcus pyogenes, Pseudomonas aeruginosa, and Bacillus species have been reported with camel bites.19 Larger cat species such as leopards and tigers have also been associated with organisms typical of domestic cat bites, with isolation of Pasteurella multocida and Neisseria weaveri.20 Snake bites do not routinely require antibiotic management unless there is necrosis, in which case aerobic gram-positive cocci and gramnegative bacilli are thought to predominate.21 A P. caballi infection was seen after a horse bite,22 P. aerogenes and a Chryseobacteriumlike organism have been isolated from infected pig bites,23,24 Actinobacillus species have been reported from horse and sheep bite wounds,25 and Halomonas venusta has been isolated from a fish bite.26 In bites occurring in marine settings, organisms associated with water, such as Vibrio species, Aeromonas hydrophila, Plesiomonas shigelloides, and Pseudomonas species, have caused infections from bites of catfish, eel, crocodile, and swan.27–31 Additionally, systemic diseases, such as tularemia from cats,32 rat bite (Haverhill) fever, and sodoku from rats (see Chapter 183, Other Treponema Species), herpesvirus B infection from monkeys, hepatitis B from humans, and leptospirosis from dogs and rodents (see Chapter 184, Leptospira Species (Leptospirosis)) have all been transmitted via bite wounds. A rare infectious entity acquired from contact with seals, known as seal finger, is likely caused by a marine mycoplasma.33 Human immunodeficiency virus (HIV) appears to be difficult to transmit by human bite.34–36

EPIDEMIOLOGY Dog bites account for > 80% of bite wounds that come to medical attention. The annual incidence of dog bites of children has been estimated at 1 to 3 per 1000 children per year37–39 in developed countries, to rates as high as 26 per 1000 per year in developing countries.40 Dog bite injuries alone account for 0.3% to 0.4% of all emergency department visits.2,37 The incidence and body part involved with dog bites vary by age. Children are more likely than adults to sustain a dog bite, with the highest risk in the second year of life and steadily decreasing each year thereafter (based on data from Austria) or with peak incidence in the 5- to 9-year age group (based on United States data).37,38 Injuries to the face, head, and neck are most common, accounting for two-thirds of dog bite injuries in preschool children (extremities account for 27%). Injuries to the extremities (upper more than lower) become more common with increasing age, accounting for 55% beyond 14 years of age (whereas only 9% involve the head or neck).38 Between 7% and 25% of children sustaining dog bites require hospital admission, depending on use of general anesthesia for primary wound management and the rate of infectious complications.37,41 Fatalities from bites occur primarily from massive blood loss after mauling by large dogs, particularly Rottweiler and pitbull-type dogs or as the result of intracranial injury, particularly in smaller children.42 Children less than 5 years of age are particularly vulnerable to attack and are more likely to sustain injury from smaller dogs. Overall the relative risk of attack is significantly higher from German shepherd,

526

SECTION

M Infections Related to Trauma

TABLE 92–1. Frequency of Isolation of Aerobic and Anaerobic Bacteria Following 50 Dog, 57 Cat, and 50 Human Bite Wounds at Presentation for Management of Infection Dog

Cat

Human

+ +

+ + + + + +

AEROBIC AND FACULTATIVE

Acinetobacter Actinobacillus Aerococcus viridans Aeromonas Agrobacterium radiobacter Alcaligenes Bacillus Brevibacterium Candida Capnocytophaga CDC group EF-4a CDC group EF-4b CDC group NO-2 Citrobacter Corynebacterium species Corynebacterium jeikeium Corynebacterium pseudodiphtheriticum Dermabacter hominis Eikenella corrodens Enterobacter cloacae Enterococcus Erysipelothrix Flavimonas Flavobacterium Gemella Haemophilus Kingella Klebsiella Kocuria Lactobacillus Leclercia Micrococcus Moraxella Neisseria Oerskovia Pantoea endophytica Pasteurella Pasteurella canis Pasteurella multocida ssp. multocida Pasteurella multocida ssp. septica

+ + + + + ++ + ++ + + + + ++ + + +

+ ++ + + ++ +++ +

+ + ++ +

+ + ++ + +

++++ + +

+

++ +++ + + + + + + + +

+

+

+

+

+ ++ ++ +

++++ ++

+++++ +++ ++ ++

+ +

+ +++++ + +++++ +++

Dog Pediococcus Proteus Pseudomonas aeruginosa Pseudomonas non-aeruginosa Reimerella anatipestifer Rhodococcus Rothia Staphylococcus aureus Staphylococcus coagulase-negative Stenotrophomonas maltophilia Stomatococcus mucilaginosus Streptococcus species Streptococcus pyogenes (group A) Streptococcus agalactiae (group B) Streptococcus group C/G Streptococcus group F Streptococcus viridans group Streptococcus milleri group (Streptococcus intermedius, Streptococcus anginosus, Streptococcus constellatus) Streptomyces Weeksella

Cat

+ + + +

+

+ + + + +++ + +++++ +++++ + + +++++ +++++ ++ + + + + +++++ +++++ ++ +

Human

+ + +

+ ++++ ++++ + +++++ ++ + +++++ +++++

ANAEROBIC

+

Doberman, pitbull, and Rottweiler breeds than from Labrador/ retriever or cross-bred dogs, but because cross-bred dogs are so much more common, they represent the majority of bite injuries. The majority of dog bites are from dogs owned by the family or friends of the family. Each year in the United States there are approximately 20 deaths as a result of dog bite attacks. Infectious deaths due to dog bites are uncommon but may occur as the result of septicemia or intracranial infection. In developing countries, rabies remains a signiÀcant late cause of death from dog bites. The estimated risk of infection in the otherwise healthy child depends on the location and extent of the dog bite injury, and the initial management, but is estimated to be from 5% to 15%. Cat bites are spread more evenly across age groups and cause much less overall initial injury because cats have smaller mouths and less biting force, which results in less tearing of the tissues. However, their thin, sharp teeth produce small deep wounds that are difÀcult to cleanse and are more likely to result in clinical infection, estimated to be from 10% to 50%. Human bites may be sustained somewhat differently than bites of other animals and are generally deÀned as any disruption in the skin as a result of contact with the mouth. These can be self-inflicted (e.g.,

Actinomyces Arcanobacterium bernardiae Bacteroides Campylobacter Clostridium Collinsella (Eubacterium) aerofaciens Dialister pneumosintes Eubacterium Filifactor villosus Fusobacterium species Fusobacterium necrophorum Fusobacterium nucleatum Lactobacillus Peptostreptococcus Porphyromonas Prevotella Propionibacterium Veillonella

++++

+++

+ + + ++

+ + ++++ ++ + ++ +++ +++ +++

+ + ++++ +++ + ++++ ++ ++ +

+ + ++ ++++ + ++++ + +++ + ++++ + +++

thumb-sucking resulting in skin breakdown, or self-injury as with autism), the result of a punch to the mouth of another person that results in a laceration to the hand, or may be the result of a typical bite but resulting from sexual activity or abuse that might be concealed from the medical provider. One study of human bite wounds in children seen at a single institution over a 6-year period identiÀed 322 patients, representing 0.2% of all Emergency Department visits. Infection was seen in 9%, and 2%, overall, were hospitalized for treatment of infection.43 Table 92-2 displays the site of bite injuries from any source summarized from two studies.43,44 Each study enrolled subjects at the time of initial presentation to the outpatient setting, allowing a more accurate estimation of the subsequent rate of infection such as cellulitis or abscess. Bites to the hand are signiÀcantly more likely to become infected compared with bites at other sites. Wounds to the face are the least likely to become infected, probably because of its extensive vascularity. In one of these studies,44 12% of the wounds became infected, and the risk of infection differed on the basis of the biting species. Only 3 (4%) of 80 dog bites became infected, but 6 (16%) of 37 human bites and 11 (50%) of 22 cat bites became infected. The most common organisms causing infection from dog and

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Infection Following Bites

CHAPTER

92

527

TABLE 92–2. Rates of Infection of Bite Wounds by Body Site Location of Bite

Rate (%)

Face/head/neck Arm/leg Trunk Hand

6 10 10 28

# Infected / # Bites 3/49 7/68 1/10 21/76

Total

16

32/203

Data from Baker MD, Moore SE. Human bites in children. Pediatrics 1991;88:55; and Aghabian RV, Conte JE Jr. Mammalian bite wounds. Ann Emerg Med 1980;9:79.

cat bites are Pasteurella multocida, Staphylococcus aureus, streptococci, Capnocytophaga canimorsus, and oral anaerobes. Cat bites have also been associated with transmission of cat-scratch disease (Bartonella henselae) and sporotrichosis.45,46 Human bite wound infections additionally can be caused by Eikenella corrodens and are not associated with P. multocida infections. Fusobacterium, Prevotella, Peptostreptococcus, and Candida are also seen.5 The paucity of reports regarding infection after rodent or rabbit bites makes the rate difficult to estimate but appears to be small, perhaps related to smaller and less penetrating wounds. Rat bites are associated with Streptobacillus moniliformis infections, hamster bites have caused tularemia, and Pasteurella aeroginese infections47,48 and guinea pigs have caused Haemophilus influenzae infections.49 Other types of bite wounds can also be important causes of infection. These include bites due to other mammals such as bats where rabies is the predominant concern. Infections transmitted by biting insects are beyond the scope of this chapter.

PATHOGENESIS The primary mechanism by which bite wounds become infected is via direct inoculation of bacteria into subcutaneous tissues. Modifying factors include the type of inoculation, the depth of inoculation, the amount of crush injury and devitalized tissue, the involvement of infection-prone structures, the ability to cleanse and debride the wound, and the pathogenicity of the infecting bacteria. Most series show a higher infection rate in bite wounds to the hand. Cat bites have the highest infection rates, probably because they are usually puncture wounds that inoculate the bacteria deeply, making wound care more difficult, and because of the high prevalence of Pasteurella multocida (Figure 92-1). Dog bites are more likely to cause lacerations or avulsion injuries that are easier to clean and debride. Most infections manifest as local cellulitis or simple abscess. Infection can extend to adjacent areas, particularly if the teeth penetrate structures such as bones or joints in bites to an extremity or the cranium in injuries to the head and face. Regional lymphadenopathy, lymphangitis, fasciitis, toxic shock syndrome, septicemia, and shock can also develop. Septicemia occurs primarily in immunocompromised hosts, particularly asplenic individuals, who are prone to fulminant septicemia and shock with infection caused by Capnocytophaga canimorsus.50–54 or Eubacterium plautii55 after dog bites.

CLINICAL MANIFESTATIONS, DIFFERENTIAL DIAGNOSIS, AND CLINICAL APPROACH The history and physical examination readily lead to the likely organisms of most bite wound infections (Box 92-1). Signs and symptoms, such as erythema, pain, tenderness, and swelling, almost always become apparent within 24 to 48 hours after the bite. Infections caused by Pasteurella multocida tend to have a more rapid onset, often within 12 to 18 hours. There may be purulent or

Figure 92-1. Cat-bite abscess of the wrist with presentation for care delayed 1 month after injury. Pasteurella multocida was isolated in pure culture.

BOX 92-1. Key Elements of History and Physical Examination in Evaluating Bite Wounds with Regard to Possible Infection HISTORY The animal Record type of animal, health of animal, whether attack was provoked or unprovoked, and consider the availability for testing or quarantine for signs of rabies, if applicable. Were local authorities notified (e.g., animal control or police officers)? The patient Consider special risks such as immunosuppression, asplenia, diabetes mellitus, antibiotic allergies, and history of tetanus and hepatitis B immunizations PHYSICAL EXAMINATION Systemic Note fever, tachycardia, tachypnea, hypotension, or widened pulse pressure Local Note type of wound (e.g., puncture, laceration, avulsion), depth of penetration, involvement of underlying structures (e.g., joint, tendon, bone, cranial contents), extent of edema, erythema, tenderness (with measurements), range of motion, type of drainage (e.g., purulent, serosanguineous, malodorous), neurovascular function, lymphangitic streaking, and regional lymphadenopathy. Consider possibility of retained foreign body (e.g., tooth)

serosanguineous discharge. Most patients are afebrile,4 although fever can occur in patients with severe cellulitis or in the rare patient with bacteremia. The presence of eschariform lesions at bite sites in individuals who appear ill may indicate C. canimorsus infection.53 The extent of the infection must be carefully determined. Even apparently trivial bites require careful evaluation for penetration of the skull or other body cavities or possible damage to underlying structures, such as joint spaces, bone, tendons, nerves, and blood vessels. For bite wounds the physician should consider the size of the mouth of the animal involved and if there are puncture wounds, examine carefully for punctures due to the apposing teeth. For instance, with cosmetically important bite wounds to the face in young children there may be more life-threatening puncture wounds found in the scalp or neck that could easily be missed unless the child is carefully evaluated.56 The need to assess involvement of deep structures is particularly important for scalp injuries that can penetrate the skull, and for clenched-fist injuries where joint or tendon injury may be missed.57 Lacerations sustained when punching and striking teeth most commonly occur on the dorsal aspect of the third metacarpophalangeal joint. Because the injury occurs with a closed fist the injury may not be recognized when examined in the relaxed

528

SECTION

M Infections Related to Trauma

position; in addition, the initially linear path of injury becomes obstructed, hindering drainage of deep structures. Paronychial infections, in young children, may result from thumbor finger-sucking, and oral flora should be considered as a possible cause, especially when a history of injury is lacking. Herpetic whitlow, which can be confused with bacterial infection, should be suspected when erythematous or dusky-appearing vesicular or vesiculobullous lesions are present, particularly in satellite arrangement. Seal finger, which occurs in association with either seal bites or scratches or other contact with seals, has a slightly different clinical presentation. The incubation period is longer (4 to 8 days), edema and severe pain are prominent, and lymphangitis, lymphadenitis, and arthritis occur more frequently than in other bite wound infections.58 Water organisms, such as Pseudomonas aeruginosa and Aeromonas species, should be considered when bites from water animals occur.

LABORATORY FINDINGS AND DIAGNOSIS Microbiologic confirmation of infection can be highly beneficial in managing bite wound infections regardless of the species causing the bite. Laboratory personnel should be notified that the specimen is from a bite wound, thus allowing consideration of multiple isolates as well as organisms that can be difficult to identify, such as Pasteurella multocida or Eikenella corrodens. Specimens should be sent for both aerobic and anaerobic bacterial culture, taking special care that proper anaerobic transport medium is used. For wounds contaminated by soil or vegetative debris, culture for mycobacteria and fungi should be considered. However, Gram stain and culture of fresh, uninfected bite wounds are not recommended, because they do not predict the risk of subsequent infection or the pathogens that cause them. Blood specimens for aerobic and anaerobic culture should be obtained from febrile children, particularly if they are immunocompromised or asplenic. Radiographs should be obtained after penetrating injuries overlying bones or joints, when a fracture is suspected, or when a foreign body is suspected.

Devitalized tissues should be cautiously debrided, and operative exploration and debridement should be considered if there is extensive tissue damage, involvement of a joint space, or cranial injury. There is no role for culture of the wound unless there is clinical evidence of infection.

3. Consider wound closure The role of suturing, in general, is controversial for bite wounds. One study evaluated the practice of primary closure for mammalian bites in a series of 145 consecutive bites (88 dog, 45 cat, 12 human) treated with primary closure. Patients presented a mean of 2 hours after injury. Six percent (95% confidence interval 2% to 9%) experienced wound infections, suggesting that when cosmesis is a primary concern, this rate of infection may be acceptable.60 Increasing time to closure has been associated with increasing risk for infection but more likely the time to closure is simply a proxy for the time to adequate wound cleansing (often done together with the repair). This has led to specific recommendations against closure for high-risk wounds if more than 8 to 12 hours have passed from the time of injury. This has not been specifically studied in children and there are no prospective data validating this approach with bite wounds. Nonetheless, in the absence of further data it is recommended that primary closure should not be performed when it is more than 8 hours since trauma or it is difficult to irrigate a deep puncture wound (unless the wound(s) must be repaired for important cosmetic or functional reasons).1,61 When closure is appropriate, the placement of a drain, microdrains, or only loose approximation of the skin is the optimal approach when feasible. The use of subcutaneous sutures has been associated with increased risk of infection in retrospective studies and their use should be avoided. An underutilized practice is delayed closure (after 48 to 72 hours) of this type of wound. Delayed closure has been recommended for human bites to the head and neck, especially those with exposed cartilage.62 Elevation and immobilization of the wound are common practice.

4. Consider postexposure prophylaxis MANAGEMENT AND PRESUMPTIVE THERAPY Immediate Postexposure Management 1. Evaluate the extent of the injury and contamination New, uninfected bites should be carefully examined for foreign bodies, and visible dirt and debris should be sponged or irrigated away.

2. Debride and cleanse the wound The wound should be copiously irrigated with sterile normal saline by high-pressure syringe irrigation, taking care not to inject into the tissue or inflict additional trauma. a. Consider use of a surgical-type scrub sponge for highly contaminated wounds, especially if particulate matter is present. (Hibiclens or Betadine should not be used, unless diluted to 1%, as these can cause tissue toxicity. Agents such as Pluronic-68 or ShurClens are suggested.) b. Irrigate with > 250 mL or more of saline. There is no demonstrated benefit to irrigation with other solutions such as 1% Betadine solution or Pluronic F-68.59 Animal models suggest high-pressure irrigation reduces bacterial wound counts better and reduces wound infection rates. Irrigation pressures of 5 to 8 pounds per square inch (psi) have been recommended and correspond to gentle use of commercially available splash guard shields (e.g., Zerowet) or an 18G catheter/needle attached to a 30 or 60 mL syringe. Normal saline solutions work just as well as other irrigants. Antibiotic irrigation has not been proven to have additional benefit; rarely it may be used in special, high-risk circumstances.

a. Antibiotic prophylaxis. The role of prophylactic antibiotimicrobial treatment is discussed in Chapter 8 (Chemoprophylaxis). Limited data upon which to base treatment decisions are available from clinical trials. A Cochrane analysis concluded that there is evidence that antibiotic administration significantly reduces infection rates with human bites but not dog or cat bites.63 One meta-analysis suggested that antibiotic treatment also lowered the risk of infection after dog bites.64 One small study randomized patients with human bite wounds to placebo or cephalexin/penicillin combination and identified a low rate of infection in both groups (1 of 62 receiving placebo and 0 of 63 individuals receiving antibiotic).65 Despite limited data regarding efficacy, antibiotic administration for postexposure wound prophylaxis is commonly recommended for infection-prone wounds. In the absence of better data, antibiotic prophylaxis is generally recommended following moderate to severe wounds (especially if edema or crush injury is present), wounds that are difficult to clean or debride adequately, penetrating cat bite(s), and human bite wounds. Less severe wounds or more easily cleansed dog bite wounds to areas other than the face, hands, feet, or genital area do not require antibiotic prophylaxis routinely. The ability of the host (e.g., immunocompromised patients) to deal with infection, should it occur, should also be taken into consideration (Box 92-2).1,57 The drug of choice is amoxicillin-clavulanic acid, which is given for 2 to 3 days. All patients with bite wounds should be re-evaluated within 24 to 48 hours for signs of infection (Box 92-3). b. Tetanus prophylaxis. Bites are not generally tetanus-prone injuries unless there is additional contamination with soil. They are, however, an opportunity to assess the patient’s vaccination status. For persons with tetanus-prone injuries who have not completed their primary series, tetanus immune globulin and the vaccine should be administered (see Chapter 188, Clostridium tetani (Tetanus)).

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Infection Following Bites

BOX 92-2. Indications for Infection Prophylaxis ANTIBIOTIC RECOMMENDED Characteristics of the Wound Bite wounds of the face, hands, feet, or genital area Wounds that cannot be reliably cleansed or completely debrided, e.g., deep punctures Bite wounds involving tendon, bone, or joint Wounds with moderate or severe edema, crush injury, or devitalized tissue Characteristics of the Host Immunocompromised or asplenic host OTHER POSTEXPOSURE PROPHYLAXIS RECOMMENDED Hepatitis B Begin vaccine if a human bite and the patient is not previously immunized. The use of hepatitis B immune globulin is limited to cases in which the biter is known to be hepatitis B surface antigen-positive Human Immunodeficiency Virus (HIV) Prophylaxis is limited to high-risk human bite exposures from known HIV-positive biter (the need for HIV postexposure prophylaxis of bite wounds is uncommon) Rabies Administer rabies immune globulin and begin vaccine series for mammalian bites, indicating possible exposure Administer rabies immune globulin and begin vaccine series Tetanus Give DT to those < 7 years if primary vaccine series is not up to schedule; give Td to those 7 years to 10 years of age or Tdap (if not previously received) to these 10 or 11 years of age or older (depending on product licensure) if series is not up to schedule or if 5 years since last booster. Give tetanus immune globulin to those not known to have received at least three previous doses of tetanus toxoid vaccine or those unlikely to have immune response to the vaccine

BOX 92-3. Recommended Antibiotic Therapy WOUND PROPHYLAXIS IF INDICATED Amoxicillin-clavulanate; if penicillin-allergic, trimethoprimsulfamethoxazole plus clindamycin TREATMENT OF WOUND INFECTION Oral therapy as above unless febrile, rapidly spreading, or high-risk: then use parenteral therapy Parenteral therapy if indicated: • Ampicillin-sulbactam (consider aminoglycoside if reptile or waterrelated species) • If penicillin-allergic without anaphylaxis, consider using extendedspectrum cephalosporin or carbapenem (e.g., ceftriaxone, cefipime, or meropenem) or if severe allergy, use trimethoprim-sulfamethoxazole plus clindamycin and consider addition of aminoglycoside if reptile or water-associated • Vancomycin should be considered in severe infections if methicillinresistant Staphylococcus aureus is a possibility • Bite wound infection in the immunocompromised host can include many other pathogens; treatment individualized

c. Rabies prophylaxis. The possibility of rabies must be considered after animal bites (see Chapter 228, Rabies Virus). Local public health authorities can assist in determining the need for postexposure rabies prophylaxis. d. Consideration of infectious agents related to human bites, such as hepatitis B, hepatitis C, and HIV. Hepatitis B prophylaxis should be considered for human bites. Human bites do not efficiently transmit HIV: postexposure prophylaxis for HIV is not routinely indicated, although it may need to be considered in some special circumstances.34,35 Most jurisdictions require bite injuries to be reported to the local health department. In cases with multiple or severe bites there is a high risk of posttraumatic stress disorder and referral for counseling should be considered.66

CHAPTER

92

529

EMPIRIC MANAGEMENT WHEN INFECTION OCCURS When a patient comes to medical attention with an infected bite wound, basic principles of management should be followed. Sutured wounds should be opened, purulent collections should be drained, and necrotic tissue, foreign bodies, and debris removed. Specimens should be obtained for aerobic and anaerobic culture. Infected wounds should not be sutured. If bone or joint penetration has occurred, exploration and debridement of these areas may be necessary. Empiric antimicrobial therapy for infected wounds should be based on pathogens that are most likely to be associated with the biting species and the clinical setting. Oral antibiotic therapy is appropriate for mild to moderate infections when adherence to the treatment plan is assured. Amoxicillin-clavulanic acid is an excellent choice because it provides activity against all common bite wound pathogens, including anaerobic bacteria, methicillin-susceptible Staphylococcus aureus, group A streptococcus, Eikenella corrodens, and Pasteurella multocida. Comparable parenteral agents, such as ampicillinsulbactam or ticarcillin-clavulanic acid, can be used for patients who require hospitalization because of rapidly spreading cellulitis or lymphangitis, large abscesses, bone or joint infection, or suspected sepsis. Appropriate alternative regimens include an extendedspectrum cephalosporin (cefotaxime or ceftriaxone) plus clindamycin. Trimethoprim-sulfamethoxazole has activity against Pasteurella and Eikenella species, and a combination of trimethoprim-sulfamethoxazole and clindamycin is acceptable for individuals with a history of anaphylactic reactions to penicillins or cephalosporins. Antibiotics commonly used for other skin and soft-tissue infections, such as the antistaphylococcal penicillins, first-generation cephalosporins, vancomycin, clindamycin, and erythromycin, are less active against Pasteurella and Eikenella species and should not be used as single agents in bite wound infections.4,67 Azithromycin and fluoroquinolone agents have in vitro activity against the organisms that commonly cause bite wound infections,68,69 but data on their clinical effectiveness are limited. Tetracycline is the treatment of choice for seal finger and is also active against most aerobic bacteria that cause bite wound infections, including Pasteurella;68 however, it is not generally recommended for children under age 8, because it may stain the teeth. A 7- to 14-day course of antibiotic treatment is usually sufficient for infections limited to the soft tissues. For bone or joint infections, > 3 weeks of treatment is generally required. In all cases, the duration and route of antibiotic therapy should be individualized, based on the infected site, culture results, antimicrobial susceptibility testing, and response to treatment.

COMPLICATIONS The most common complications of bite wound infections are related to tissue destruction. Soft-tissue necrosis can result from infection or from crush injury during the bite. Pyogenic arthritis or osteomyelitis, particularly in cases in which diagnosis is delayed, can result in permanent injury to the joint or bone. Depending on the site of the bite, other complications, such as brain abscess, can develop.70 Septicemia and meningitis can occur, especially in immunocompromised individuals.

PROGNOSIS AND SEQUELAE Long-term follow-up studies of children sustaining bite wound infections are not available. In general, soft-tissue infections that are treated promptly and appropriately are expected to heal completely, unless the injury itself has resulted in extensive tissue devitalization. Infections involving underlying bone, tendon, or joint structures have a more guarded prognosis for return to normal function. For example, Goldstein57 reports residual sequelae in 25% to 50% of clenched-fist injury patients, and other series report smaller numbers of individuals with permanent disabilities of the involved extremity, often related to delayed or inappropriate therapy.71,72 A delay in therapy, particularly that caused by a failure to recognize deep (bone, joint, or tendon)

530

SECTION

N Infections of the Fetus and Newborn

involvement, may be the most important factor leading to a poor outcome.

PREVENTION The most important factor for dealing with bite wounds is the prevention of bites before they occur. This can be accomplished, to some degree, through education of the public to limit exposure of

SECTION

preschool-age children to dogs, selection of specific dog breeds or spayed or neutered dogs that are less likely to bite, and through education of children on proper behavior around animals and the proper respect for dogs. Proper cleansing and debridement of wounds remain the cornerstone for the prevention of wound infections, although antimicrobial prophylaxis may also play an important role in preventing infection in persons who sustain bite wounds.

N

Infections of the Fetus and Newborn

CHAPTER

93

Clinical Approach to the Infected Neonate

first trimester often result in stillbirth or severe anomalies. With organisms such as CMV, transmission of infection is dependent on whether the maternal infection is a primary infection or a recurrence. Nearly 40% of neonates born to mothers experiencing a primary CMV infection during the pregnancy are affected whereas only 1% of neonates are affected in pregnancies of seropositive mothers.2 Perinatal infections are acquired just prior to birth (often after rupture of membranes) or as the neonate passes through the birth

P. Brian Smith and Daniel K. Benjamin, Jr TABLE 93-1. Periods of Transmission of Neonatal Pathogens6,28–31 Neonates have unique susceptibilities to infection. The immaturity of the neonatal immune system and the neonatal intensive care unit environment contribute to the variety of organisms affecting this population. These immunologic deficiencies and the nature of the pathogens are detailed in Chapter 10, Immunologic Development and Susceptibility to Infection. This chapter outlines the clinical approach to the neonate with signs or symptoms consistent with infection. The approach should focus first on supportive measures: respiratory support, correction of metabolic and hematologic derangements, fluid management, and use of inotropic drugs when indicated. The clinician should then obtain diagnostic studies and institute antimicrobial therapy considering the timing of infection and the neonate’s previous exposure to maternal and environmental factors.

Pathogen

Prenatal

Perinatal

Postnatal

Cytomegalovirus

μ

μ

μ

Enterovirus

Rare

μ

μ

Hepatitis B virus

Rare

μ

Herpes simplex virus

Rare

μ

Human immunodeficiency virus

μ

μ

μ

Parvovirus B19

μ

Rubella virus

μ

Varicella-zoster virus

μ

μ

μ

VIRUSES

BACTERIA

EPIDEMIOLOGY Timing is one of the most important factors in determining the cause of neonatal infections. Infections can be acquired prenatally, perinatally, or postnatally (Table 93-1). Detailed discussions of each of these individual infections may be found in Chapter 94, Bacterial Infections in the Neonate; Chapter 95, Viral Infections in the Fetus and Neonate, and Chapter 96, Nosocomial Infections in the Neonate. Infections acquired in utero result from organisms such as Toxoplasma gondii, rubella virus, cytomegalovirus (CMV), herpes simplex virus (HSV), parvovirus B19, and Treponema pallidum. With the exception of HSV, these infections usually result from transplacental transmission. Risk of transmission varies depending on trimester of pregnancy during which the maternal infection occurs. For example, Toxoplasma transmission rates range from < 10% during the first weeks of pregnancy to 60% in the third trimester.1 Severity of infection also depends on the stage of pregnancy at which the maternal infection occurred. Congenital toxoplasmosis and rubella acquired during the

Chlamydia trachomatis

μ

Group B streptococcus

μ

μ

Enterococcus species

μ

μ

Enterobacteriaceae

μ

μ

μ

μ

Listeria monocytogenes

μ

μ

Neisseria gonorrhoeae

μ

Staphylococcus species Treponema pallidum

μ μ

Ureaplasma urealyticum FUNGI

μ

Candida species PROTOZOA

Toxoplasma gondii

μ

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

μ

Clinical Approach to the Infected Neonate

CHAPTER

93

531

and often manifests as focal infection: meningitis, urinary tract infection, septic arthritis, osteomyelitis, or pneumonia. Over 90% of neonates with early-onset sepsis present in the first 24 hours, with most of the remaining neonates presenting by 48 hours of life.8 Hence the acutely ill neonate requires empiric antimicrobial therapy in addition to multisystem supportive care. Signs of perinatally and nosocomially acquired septicemia are often nonspecific and subtle, but septicemia is uncommon in asymptomatic neonates.8,9 Signs of neonatal septicemia include: temperature instability, lethargy, irritability, apnea, respiratory distress, hypotension, bradycardia, tachycardia, cyanosis, abdominal distension, hyperglycemia, hypoglycemia, jaundice, and feeding intolerance.10 These signs overlap with a myriad of other disease processes presenting in the neonatal period, including: anemia, congenital heart disease, respiratory distress syndrome, and metabolic disorders.

canal. The neonate is initially colonized after exposure to the variety of microflora present, including nonpathogenic organisms such as Lactobacillus, Peptostreptococcus, and Saccharomyces. However, the neonate is also exposed to potential pathogens such as group B Streptococcus (GBS), Escherichia coli, HSV, human immunodeficiency virus (HIV), and Candida species. Perinatally acquired infections can manifest any time from immediately after birth to weeks or months later (e.g., HIV). Early-onset sepsis (sepsis presenting before 7 days of life) is almost always caused by perinatally acquired infections. Risk factors for development of sepsis, such as preterm delivery, prolonged rupture of membranes (> 18 hours), maternal fever, and chorioamnionitis, are often present in neonates developing early-onset sepsis.3 Although GBS is still the most common cause of early-onset sepsis, the institution of intrapartum chemoprophylaxis has decreased the incidence of this organism as a cause of early-onset disease by 70%,4 with a corresponding trend, at least in very-lowbirthweight infants (VLBW, < 1500 grams birthweight), of an increase in early-onset sepsis caused by gram-negative organisms.5 Late-onset sepsis (sepsis presenting 7 to 30 days of life) can be caused by perinatally or postnatally acquired infections. Term neonates may still experience late-onset sepsis from GBS as intrapartum prophylaxis has shown little effect on the rates of lateonset infection.4 Late-onset and late-late onset (> 30 days of life) infections are increasingly found in VLBW neonates.6 In this population, the organisms are likely to be commensal organisms such as coagulase-negative staphylococci, Staphylococcus aureus, Candida species, or gram-negative bacilli.6,7

LABORATORY EVALUATION If bacterial or fungal infection is suspected, prompt investigation should be accompanied by administration of empiric antimicrobial therapy. Screening tests such as white blood cell counts and acutephase reactants such as C-reactive protein (CRP) have poor positive predictive values in septic neonates (40% in symptomatic neonates and as low as 1% to 2% in asymptomatic neonates at risk for GBS sepsis).8,11 For bacterial and fungal infections, culture of normally sterile body fluids remains the gold standard for diagnostic purposes. Positive blood cultures for pathogenic bacteria are evident by 48 hours of incubation in the vast majority of cases.12 Blood cultures were found to be positive for Candida species in 97% of cases by 72 hours, but sensitivity of blood cultures for detection of invasive candidiasis is < 50%.13 There is no consensus as to the recommended number of blood cultures or volume of blood to culture. Obtaining multiple-site blood cultures increases the ability to distinguish between cultures contaminated with skin flora and those representing true infection.14 However, the sensitivity increases by only a few percentage points with the addition of a second blood culture. Although a blood culture inoculum of 0.5 mL has demonstrated good sensitivity in neonates, several other studies have shown that 0.5 mL of blood may not detect low-level bacteremia.15,16 Sensitivity of blood cultures may be further

CLINICAL MANIFESTATIONS Although a variety of organisms can cause congenital infections, some aspects of the clinical presentation are common to many (Table 93-2). Findings such as intrauterine growth retardation (IUGR), jaundice, hepatosplenomegaly, rash, intracranial calcifications, microcephaly, chorioretinitis, and thrombocytopenia can occur with several types of infections. Although many of the sequelae of congenitally acquired infections are present at birth, others (hearing loss, developmental delay) may not manifest for months or years. Early-onset bacterial septicemia is often nonfocal and fulminant in onset, in contrast to late-onset disease which can progress more slowly TABLE 93-2. Characteristic Manifestations of Congenital Infections32–36 Toxoplasma

Rubella

CMV

Anemia Bony abnormalities

μ

μ

Treponema

Parvovirus B19

μ

μ

μ

VZV

Herpes

μ

μ

μ

Cardiac anomalies

μ

Cataracts

μ

μ

Chorioretinitis

μ

μ

Hearing impairment

μ

μ

μ

Hepatosplenomegaly

μ

μ

μ

Hydrocephalus

μ

μ

μ

μ μ

μ

μ

μ

μ

μ

μ

μ

μ

μ

μ

μ

μ μ

μ μ μ

Hydrops fetalis μ

IUGR

μ

Microcephaly

μ

Rash

μ

μ

μ

μ

μ

Thrombocytopenia

μ

μ

μ

μ

μ

CMV, cytomegalovirus; IUGR, intrauterine growth retardation; VZV, varicella-zoster virus.

μ

μ

Intracranial calcifications

μ

Enterovirus

μ

532

SECTION

N Infections of the Fetus and Newborn

compromised by maternal intrapartum antibiotic administration or current antibiotic exposure in the neonate. The incidence of bacterial meningitis is higher in the first month of life than at any other time and complicates up to one-third of the cases of septicemia in this population.17 Diagnosis of meningitis and identification of the offending organisms require examination of cerebrospinal fluid (CSF). Unless neurologic signs are present at the time of the sepsis evaluation, some clinicians defer the lumbar puncture until the blood cultures are positive for a pathogenic organism.18 However, blood cultures can be negative in neonates with bacterial meningitis. In one series, blood cultures were negative in 28% (12/43) of neonates with meningitis diagnosed in the first 72 hours of life.19 In another study, Stoll et al. found that blood cultures were negative in 34% (45/134) of VLBW neonates with culture-proven meningitis.20 These data suggest that clinicians should continue to include the lumbar puncture as part of the evaluation of neonatal sepsis in symptomatic neonates provided that the infant is stable enough to tolerate the procedure. Urine culture should be obtained as part of the sepsis evaluation in neonates after day of life 3, but they are of low yield before this age.21 Urine culture should be obtained by suprapubic aspiration or catheterization. Bag specimens are often difficult to evaluate due to contamination and may lead to unnecessary antibiotic administration and radiologic studies. Disseminated HSV infection can manifest as sepsis. Evaluation of neonates suspected of having HSV infection should include surface cultures obtained from swabs of the conjunctivae, mouth, skin, and anus. CSF should also be obtained for routine studies and DNA polymerase chain reaction. Characteristics of the clinical presentation and a review of the maternal history may provide additional clues as to which laboratory tests should be obtained. Table 93-3 provides an outline for current diagnostic methods for congenital infections. Evaluation of perinatally and nosocomially acquired bacterial infections is discussed further in Chapter 94, Bacterial Infections in the Neonate and Chapter 96, Nosocomial Infections in the Neonate.

considered standard empiric therapy. Ampicillin provides coverage for gram-positive infections (GBS and Listeria monocytogenes). Gentamicin is active against many gram-negative pathogens (E. coli and other Enterobactericeae) and also provides synergy with ampicillin against GBS and Listeria. Cefotaxime, a third-generation cephalosporin with superior CSF penetration compared with gentamicin, may be considered in cases of suspected or proven gramnegative bacillary meningitis. Vancomycin or nafcillin is sometimes substituted for ampicillin for suspected nosocomial infections, especially when the neonate has an indwelling central venous catheter. Recommendations regarding prevention of catheter-related infections are included in a comprehensive review.22 Duration of antibiotic therapy is often 10 days for confirmed uncomplicated bloodstream infection. Duration of therapy for meningitis is pathogen-specific. For meningitis-caused GBS, therapy is given for 14 to 21 days (depending on clinical course and CSF culture results after 24 hours of therapy), for Listeria is 14 days, and for enteric bacilli is 14 days after sterilization of CSF or 21 days (whichever is longer). Therapy is often extended beyond the standard duration if complicated infection is indicated by course of imaging studies. Although uncommon in term infants, the cumulative incidence of candidemia in extremely-low-birthweight neonates (ELBW, < 1000 grams birthweight) is 7%.7 Empiric therapy for Candida should be considered in symptomatic neonates with risk factors for fungal infections, including extreme prematurity, thrombocytopenia, and exposure to broad-spectrum antibiotics.23 Because amphotericin B deoxycholate has a wide spectrum of activity and superior tolerability in neonates compared with adults, it should be considered first-line treatment of Candida infections in this population.24,25 Fluconazole is an alternative when treating an isolate known to be sensitive; i.e., C. albicans, C. parapsilosis, and C. tropicalis. Although not part of the standard empiric therapy for neonates, acyclovir should be administered in cases of suspected herpes infection;26 neonates who should be considered for empirical acyclovir therapy include those with elevated hepatic transaminases, those with skin vesicular lesions, and those who do not improve following antibiotic therapy and have negative blood cultures at 48 to 72 hours.27

TREATMENT A high index of suspicion and timely antimicrobial therapy are paramount in the management of septic neonates. For newborns with suspected bacterial infections, ampicillin and an aminoglycoside are TABLE 93-3. Diagnostic Tests for Congenital Infection28,31,34,37,38 Pathogen

Method

Cytomegalovirus

Shell vial assay, culture of virus from urine, DNA PCR, detection of IgM antibody

Human immunodeficiency virus

DNA PCR

Parvovirus B19

DNA PCR, IgM antibody

Rubella virus

IgM antibody, culture from nasal specimens

Varicella-zoster virus

IgM antibody, DNA PCR

Herpes simplex virus

DNA PCR of CSF, cell culture of skin lesions, mouth, nasopharynx, eyes

Treponema

Quantitative treponemal test of serum, CSF VDRL, dark-field examination for spirochetes, direct fluorescent antibody test of lesions

Toxoplasma

Detection of IgM or IgA antibody, blood or urine culture, DNA PCR

CSF, cerebrospinal fluid; IgM, immunoglobulin M; PCR, polymerase chain reaction; VDRL, Venereal Diseases Research Laboratory.

CHAPTER

94

Bacterial Infections in the Neonate Morven S. Edwards and Carol J. Baker

Systemic bacterial infections affect 1 to 5 of 1000 liveborn neonates. Neonatal sepsis is a clinical syndrome characterized by systemic signs of infection and bacteremia occurring in the first month of life. However, the survival of very-low-birthweight (VLBW) (1500 g or less) and especially extremely low-birthweight (ELBW) (1000 g or less) infants has necessitated extension of this definition to include infants requiring prolonged hospitalization for complications of prematurity. The past two decades witnessed changes in the epidemiology of neonatal bacterial infections. The mortality rate declined from 30% to 40% to approximately 5% to 10%, and the incidence of meningitis as a complication of early-onset sepsis declined from 25% to 30% to approximately 3% to 10%. Enhanced awareness of maternal and infant factors increasing risk for septicemia led to earlier treatment, which provides a partial explanation for the declines in mortality and the rate of meningitis. The greater survival of VLBW and ELBW infants has contributed to a shift in late-onset bacterial and fungal infections due to commensal species as well as to an overall higher

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Bacterial Infections in the Neonate

incidence in neonatal septicemia.1 Also, outbreaks of sepsis in the neonatal intensive care unit (NICU) have become frequent.2 Molecular techniques can provide proof that causative bacteria from several infants have a common chromosomal DNA, suggesting that transmission from infant to infant via the hands of caregivers occurs with some frequency.3

EPIDEMIOLOGY Neonatal bacterial infections are best understood when characterized by age at onset. Since the 1970s, the onset has been termed early or late, reflecting the age at which signs of infection begin (Table 94-1). Neonates with early-onset infection have illness before 7 days of age, but the majority are ill within 24 hours of birth. Maternal complications of labor or delivery are common, and the source of pathogens is the maternal genital tract. The typical clinical presentation is that of respiratory distress or nonspecific signs, often indistinguishable from those of noninfectious disorders, without evidence of focal infection. For infants with late-onset infection, maternal obstetric complications other than preterm delivery are less common. The source of the causative agent can be the maternal genital tract, the hospital if neonatal care in this setting lasts more than a few days, or the community. The clinical presentation is often focal; softtissue infection, pneumonia, and meningitis are examples. The survival of VLBW infants or ELBW neonates has prompted the use of a third category of neonatal sepsis – late, late or very late onset. Although such infants are, strictly speaking, no longer neonates, their median gestational age (< 28 weeks) and their continuing hospitalization for postnatal complications accord them an extended interval as “newborns” on the basis of postconceptual age. These infants almost always have intravascular access catheters, and infection is often caused by commensal species.1 The associated mortality rate is 5% to 60% and dependent on infecting agent.

ETIOLOGIC AGENTS Over the past three decades there has been a decline in the incidence of early-onset sepsis and an increase in cases of late-onset and late, late-onset sepsis.4 The incidence of early-onset group B streptococcal (GBS) infection has declined by ~65% to 0.3 cases per 1000 live births since the institution of maternal intrapartum antibiotic prophylaxis.5 However, the predominant pathogens causing early-onset bacterial

TABLE 94-1. Features Distinguishing Early-Onset from Later-Onset Bacterial Infection in Neonates Feature

Early-Onset

Late-Onset

Time of onset 30 Common

Median birthweight < 1000 g

Source of organism

Maternal genital Maternal genital Nosocomial; tract tract; nosocomial; community community

Usual clinical presentation

Nonspecific or respiratory distress

Mortality rate (%) 5–15

Focal

Focal

5–60

94

533

infections, GBS and Escherichia coli, have not changed (Table 94-2). Taken together, these organisms account for approximately two-thirds of infections.4–6 The remaining early-onset infections are caused by other streptococci, including Enterococcus faecalis, E. faecium, ahemolytic streptococci, group A Streptococcus and Streptococcus pneumoniae; Listeria monocytogenes; Haemophilus influenzae; or other maternal genital flora. Among VLBW infants there has been a marked reduction in GBS as a cause of early-onset sepsis and an increase in sepsis caused by Escherichia coli.7 More than one-half of the early-onset infections in the 2002 to 2003 cohort of VLBW infants at centers of the National Institute for Child Health and Human Development (NICHD) Neonatal Research Network were caused by gram-negative organisms, especially E. coli.8 Each of the microorganisms causing early-onset bacterial infection can also cause late-onset disease. Septicemia due to E. coli or GBS remains common. However, the predominant organisms are coagulasenegative staphylococci (CONS), which virtually always occur in association with a medical device.4 Gram-negative enteric bacilli that are frequently more resistant to antimicrobial agents, such as Enterobacter and Citrobacter spp., as well as nosocomial gramnegative pathogens, such as Pseudomonas aeruginosa and Serratia marcescens, are also encountered. Streptococcus pneumoniae is an uncommon causal organism. Infants with late-onset pneumococcal sepsis are usually full-term and present in the third week of life.9

TABLE 94-2. Bacteria Causing Neonatal Septicemia Importance of Pathogen Early-Onset

Later-Onseta

Group B Streptococcus

+++

+

Viridans streptococci

+

+

Enterococcus spp.

+

++

Coagulase-negative staphylococci

–

+++

Staphylococcus aureus

–

+++

Streptococcus pneumoniae

+

+

Listeria monocytogenes

+

+

Escherichia coli

+++

++

Klebsiella spp.

+

++

Enterobacter spp.

+

++

Citrobacter spp.

–

+

Serratia marcescens

–

+

Pseudomonas spp.

–

+

Salmonella spp.

–

+

Haemophilus influenzae

+

–

Neisseria meningitidis

–

+

Other nonenteric gram-negative bacilli

–

+

Other enteric gram-negative bacilli

+

+

Bacteroides spp.

+

+

Clostridium spp.

–

+

Others

–

+

Bacteria GRAM-POSITIVE BACTERIA

GRAM-NEGATIVE BACTERIA

ANAEROBIC BACTERIA

a

2–10

CHAPTER

Includes late- and late, late-onset. +++, commonly associated; ++, frequently associated; +, occasionally associated; –, rarely associated.

534

SECTION

N Infections of the Fetus and Newborn

The increased survival rates for VLBW infants have contributed to the shift to commensal species causing late-onset infection. According to the definition from the Centers for Disease Control and Prevention, septicemia due to commensal species requires isolation of such a species (normal skin flora or resident flora of the intestine) from a blood culture specimen and one of four signs of illness (apnea, bradycardia, hypothermia, or hyperthermia), and either a second positive blood culture result or placement of an intravascular access device before illness develops, followed by antibiotic treatment for > 96 hours. CONS are the most common cause of septicemia due to commensal species, accounting for 39% of all late-onset infections and 25% of late, late-onset infections in a contemporary series.4 Among VLBW infants, a majority of late-onset infections were caused by gram-positive organisms (70%) and CONS accounted for nearly one-half of all infections in the experience of the NICHD Neonatal Research Network.10 CONS, more often than Staphylococcus aureus infections, are associated with VLBW, use of intravascular catheters, and total parenteral nutrition.11 S. aureus is also a common cause of healthcare-associated infection, accounting for 8% of late-onset infections in the NICHD Neonatal Research Network.10 Community-associated methicillin-resistant S. aureus (CA-MRSA) strains have emerged as a significant cause of late-onset sepsis in neonates hospitalized in the NICU since birth.12 MRSA strains from 75% of infants with late-onset bacteremia due to S. aureus in one large NICU carried the staphylococcal cassette chromosome mec (SCCmec) genes that are characteristic of CAMRSA. Outbreaks of neonatal pustulosis (especially on the lower abdomen) due to CA-MRSA have been reported in multiple states, associated with term well-baby nurseries.13 Enterococci and Candida must be considered in infections that arise from the intestine as well as those that occur in association with use of intravascular lines. In a recent national point prevalence survey, enterococci and fungi were the next most common bloodstream isolates after staphylococci in the NICU setting.14 Anaerobic bacterial infections can occasionally result from acquisition of organisms from maternal genital flora. Clinical manifestations are similar to those of aerobic infections, except in the setting of late-onset polymicrobial septicemia, in which the intestine is the source of infection (e.g., necrotizing enterocolitis). In addition, a number of unusual bacteria have caused infections in neonates.15 A bacterial isolate from blood or a normally sterile body site in a clinically ill neonate should be considered a true pathogen unless there is sufficient cause to conclude that it is a contaminant.

TABLE 94-3. Maternal Peripartum Risk Factors for Early-Onset Bacterial Infection Risk Factor

Comment

Preterm delivery

Attack rate inversely related to gestation < 37 weeks

Premature rupture of membranes

Rupture of membranes > 1 hour before onset of labor at any gestation

Chorioamnionitis

Risk of neonatal septicemia is 5–15%

Urinary tract infection Higher neonatal risk even when mother asymptomatic Multiple pregnancy

Only noted for group B streptococcal septicemia

Prolonged rupture of membranes

Attack rate directly proportional to duration of rupture of membranes > 12 hours

Early postpartum febrile morbidity

Maternal fever (> 38°C) during the first 24 hours postpartum

No prenatal care

Higher neonatal risk

Fetal hypoxia

Apgar score < 6 associated with higher risk

units (cfu) per mL) of GBS, for example, are at a greater risk of septicemia than those exposed to a maternal carrier who has lowdensity colonization. Premature neonates are at increased risk of septicemia because of (1) acquisition of lower levels of maternally derived total immunoglobulin (Ig) G and specific antibodies to bacterial pathogen species than term neonates; (2) immature function of neutrophils and decreased neutrophil storage pools; and (3) immature immune responses to pulmonary invasion and bacteremia. Metabolic factors such as hypoxia, acidosis, and hyperbilirubinemia further compromise host response. Interruption of mucosal or skin barriers by endotracheal or nasogastric tubes, intravascular access devices, blood sampling, and monitoring equipment promote bacterial invasion, particularly for late, late-onset infection. Prior therapy with broad-spectrum antimicrobial agents increases the likelihood that bacteria more resistant to routinely employed antimicrobial regimens will cause late-onset infection.

CLINICAL MANIFESTATIONS PATHOGENESIS

Septicemia and Meningitis

Infants who experience early-onset sepsis, especially preterm infants, usually have a history of one or more risk factors associated with their mother’s labor and delivery (Table 94-3). Although most term infants with early-onset sepsis have associated maternal pregnancy and delivery complications, 30% to 50% of those with early-onset GBS sepsis have no identifiable maternal risk factor. Maternal factors that should suggest risk for early-onset septicemia include preterm delivery, premature rupture of placental membranes (rupture of membranes before onset of labor at any point in gestation), prolonged rupture of membranes, chorioamnionitis, meconium-stained amniotic fluid, GBS bacteriuria during pregnancy, multiple pregnancy, early postpartum febrile morbidity (including maternal bacteremia, endometritis, and wound infection), no prenatal care, and complications of delivery causing or associated with fetal hypoxia. An inverse relationship between the duration of rupture of membranes and the attack rate of GBS septicemia16 is likely applicable to other pathogens causing early-onset sepsis because this increases the infant’s exposure time to potential pathogens. The pathogenesis of neonatal septicemia in neonates with one or more maternal risk factors is multifactorial and encompasses microbial, host, and metabolic components. Infants with exposure to a high-density maternal inoculum (more than 1 μ 105 colony-forming

The risk of bacterial infection in healthy-appearing neonates is low.17 However, clinical signs of bacterial infection are usually subtle, and even minimal deviation from usual activity should be regarded as a possible indication of invasive infection (Table 94-4). In one study of 647 infants, hypoglycemia and hypothermia were the most common findings in both early-onset and late-onset sepsis.4 In another report, hyperthermia was the most common sign of septicemia among 455 neonates.15 Even so, only half of the infants had fever as a sign of septicemia. Temperature elevation without infection in full-term infants is uncommon (except for fever in conjunction with prostaglandin therapy). Only 1% of infants have a febrile episode, defined as an axillary temperature in excess of 37.5°C to 37.8°C. Environmentally induced abnormality (e.g., “isolette fever”) is rarely persistent. A temperature elevation sustained for > 1 hour is frequently associated with systemic infection and cannot be casually dismissed. Infants with a temperature abnormality should be examined closely for other accompanying signs. One-third to one-half of infants with bacteremia have signs of respiratory distress, including tachypnea, grunting, or retractions. Among full-term infants returning for evaluation after the first week of life, jaundice can be a presenting feature. Findings such as lethargy, irritability, abdominal distention, and diarrhea are less common clues, occurring in less than one-fourth of infants.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Bacterial Infections in the Neonate

TABLE 94-4. Clinical Signs of Bacterial Infections in the Newborn

CHAPTER

94

TABLE 94-5. Tympanic Membrane Signs in Healthy Term Neonates

Frequency of Sign

Age (days)

Clinical Sign

Septicemia

Meningitis

Sign

1–3

21–41

Hyperthermia

+++

+++

Usually gray (~80%)

++

++

Tympanic membrane color

Often pink (33%)

Hypothermia Respiratory distress

++

++

Apnea

+

+

Jaundice

++

++

Lethargy

++

+++

Anorexia or vomiting

++

++

Irritability

+

++

Convulsions

–

++

Bulging or full fontanel

–

++

Diarrhea or abdominal distension

+

+

Hypotension

++

+

+++, encountered in ≥ 50%; ++, frequently associated (25–49%); +, occasionally observed (15–24%); –, rarely associated (< 15%). Adapted from Palazzi DL, Klein JO, Baker CJ. Bacterial sepsis and meningitis. In: Remington JS, Klein JO, Wilson CB, et al. (eds) Infectious Diseases of the Fetus and Newborn Infant, 6th ed. Philadelphia, Elsiever Saunders, 2006, p 267.

Neonates with meningitis manifest the same general early signs as those with septicemia (see Table 94-4). In addition, seizures occur in 40%, a bulging or full fontanel in 28%, and nuchal rigidity in 15%.18 These signs, if present, suggest meningitis, but their absence does not exclude central nervous system infection. Infants should be examined for other signs suggesting foci of bacterial infection, including otitis media, conjunctivitis, pneumonia, cellulitis, and abdominal signs such as distention and diminished bowel sounds (see Chapter 62, Necrotizing Enterocolitis). Extremities should be examined for limitation of motion, swelling, erythema, and pain with motion, because the signs of osteomyelitis and septic arthritis are also subtle. Skin lesions, including the pustular lesions consistent with listeriosis or staphylococcal infection, should be sought.

Otitis Media Healthy term infants have otoscopic findings in the immediate neonatal period (such as diminution or absence of mobility or a pink appearance) that would suggest middle-ear inflammation in an older infant (Table 94-5).19 These features are progressively less common by age 1 month. In most infants younger than 1 month diagnosed as having otitis media, established presumptively with pneumatic otoscopy, middle-ear effusion is confirmed by tympanic aspiration.20 These infants are not generally ill, uncommonly have fever, and often have findings consistent with upper respiratory tract infection. Despite their young age, middle-ear pathogens are usually those expected in older infants, such as Streptococcus pneumoniae, Moraxella catarrhalis, and Haemophilus influenzae; gram-negative enteric organisms and Staphylococcus aureus constitute < 10%. Young infants with isolated pneumococcal otitis media are usually full term and often have bilateral disease.9 Approximately 40% of middle-ear aspirates are sterile or contain species considered nonpathogenic. In contrast to neonates evaluated as outpatients, premature infants remaining in the NICU often have systemic signs associated with bacterial otitis media. Fever, abdominal distention, emesis, diarrhea, irritability, and poor feeding are frequent findings, whereas nasal congestion is encountered less frequently.21 Nasotracheal intubation

535

Light reflex adequacy Usually decreased

Normal in 33%

Landmarks

Normal

Normal

Mobility

Usually none or limited Normal in 50%; usually limited in remainder

Data from Cavanaugh RM Jr. Pneumatic otoscopy in healthy full-term infants. Pediatrics 1987;79:520.19

for > 7 days correlates significantly with the impaired tympanic membrane mobility suggestive of otitis media. The pathogens commonly encountered are similar to those expected for late, late-onset sepsis and include S. aureus, CONS, and enteric bacilli.

Conjunctivitis The most common cause of conjunctivitis in the neonate in the United States is chemical irritation resulting from topical administration of silver nitrate.22 Infectious conjunctivitis in neonates is caused by Chlamydia trachomatis, various gram-positive and gram-negative bacteria, including S. aureus, Streptococcus pneumoniae, enterococci, and Haemophilus spp.23 Neisseria gonorrhoeae, herpes simplex virus, and adenovirus are infrequent but important pathogens. Regardless of the pathogen, eyelid edema, hyperemia of the palpebral conjunctivae, and purulent discharge are common findings. Findings are frequently bilateral. In hospitalized infants, especially those born prematurely, Pseudomonas aeruginosa conjunctivitis can be associated with systemic complications such as bacteremia and meningitis.24

Osteomyelitis and Pyogenic Arthritis Two distinct clinical syndromes have been associated with bone or joint infection in neonates (Table 94-6). The first, termed the “mild” or “benign” form, is probably the consequence of low-grade and transient bacteremia. At presentation, there is little to no evidence for systemic infection, and signs can persist for 2 to 4 weeks before evaluation. These signs include edema and swelling of the extremity or joint, usually without warmth or erythema. There is decreased spontaneous movement of the involved extremity, resulting either from pain (“pseudoparalysis”) or weakness due to neuropathy, which in turn is caused by stretching or edema of the nerve plexus. Infants with GBS osteomyelitis usually have this “benign” presentation. Although this finding is typically evident at 3 to 4 weeks of age, parents may describe the infant as having diminished movement of an extremity dating from the time of hospital discharge or pain with lifting or (when the femur is involved) during diaper changes. History of minor trauma at birth, such as from shoulder dystocia or breech carriage, may be present. Erb palsy can be the erroneous diagnosis for osteomyelitis involving the proximal humerus. The severe presentation is a consequence of prolonged or intense bacteremia. Staphylococci, including MRSA, and, to a lesser extent, E. coli, are the usual pathogens. Severe systemic signs overshadow early localizing signs of hematogenous seeding of bones. Bone or joint foci can be noted concurrently with bacteremia or within the days or weeks after initiation of therapy, emphasizing the need for repeated thorough physical examinations. Inflammatory changes, such as swelling, tenderness, erythema, and warmth, are prominent. Because of spread across transphyseal vessels, inflammatory changes

536

SECTION

N Infections of the Fetus and Newborn

TABLE 94-6. Clinical Presentation of Bone and Joint Infections in

TABLE 94-7. Screening Tests for Septicemia: Uses and Limitations

Newborn Infants Manifestation

“Mild” Form

Severe Form

Preceding bacteremia Low-grade or transient Prolonged or intense Duration of signs

Can be several weeks

Simultaneously with bacteremia or in days after initiation of therapy

Physical findings

Subtle

Prominent

Multiple bone involvement

Uncommon

Common

characteristically are poorly localized. Involvement of multiple bones is common.25 The presenting signs of osteomyelitis in 121 neonates reported between 1965 and 1990 consisted of swelling in 64%, pseudoparalysis in 55%, tenderness in 32%, erythema in 30%, fever in 45%, and irritability or lethargy in 36% of infants.26 As with older children, tubular bones are predominantly involved, with infection beginning in the metaphysis. The femur is the most frequently involved site, with the humerus and tibia next most affected. Cuboidal bones, primarily bones of the hands or feet, and the flat bones of the ribs, skull, sternum, and scapula, account for one-fourth of sites. Infants with osteomyelitis due to Candida spp. have a pathogenesis similar to that of the aforementioned bacterial species. However, signs are more subtle. Predominant findings are: (1) swelling of the involved area without erythema; and (2) pain with motion. Multiple sites can be involved.

Test Finding Total white blood cell count (cells/mm3)

LABORATORY FINDINGS AND DIAGNOSIS Septicemia Screening Tests The difficulties inherent in determining the diagnosis of bacterial infection in neonates have prompted the development of a number of screening tests. These are variably useful adjuncts to clinical assessment, but no one test or combination of tests has proved sufficiently sensitive and specific to obviate evaluation of infants with clinical signs of illness.28 The final judgment as to which infants undergo full evaluation and pre-emptive treatment is achieved through clinical assessment weighted by evaluation of risk factors. Taken in the context of their limitations, the total white blood cell (WBC) count with differential count and other screening tests listed in Table 94-7 can be useful. Standards from the 1970s29 provide a framework within which WBC abnormalities can be detected. These were not intended as a septicemia screening tool but have been widely used for this purpose. In one study 79% of 61 bacteremic neonates younger than 3 days had an abnormal WBC count according to these criteria.30 In those with

< 5000 or > 20 000

Comment(s) < 50% of those with finding have proved infection

Total neutrophil < 4000 (polymorphonuclear: PMN) count (cells/mm3)

Particularly useful in first hours of life

Total immature PMN count (cells/mm3)

Relatively insensitive; finding unusual in uninfected infants

> 1100 (cord blood) > 1500 (12 hours) > 600 (> 60 hours)

Ratio of immature > 0.2 PMNs to total PMNs

Sensitivity 30–90%; good negative predictive value

Platelet count (cells/mm3)

< 100 000

Insensitive, nonspecific, and late finding

C-reactive protein (mg/dL)

> 1.0

Sensitivity 50–90% at onset

Interleukin-6 (pg/mL)

> 15

Sensitivity > 80%; cutoff points vary; serial determinations may be required

Procalcitonin (ng/mL)

> 0.5

Promising for early- and late-onset infection

Erythrocyte sedimentation rate (mm/h)

> 5 (1st 24 hours) > Infant’s age in days + 3 (through age 14 days) > 20 (> 2 weeks of age)

Individual laboratories must establish normal values; normal value varies inversely with hematocrit

Fibronectin (mg/mL)

< 120–145

Sensitivity 30–70%

Haptoglobin (mg/dL)

> 10 (cord blood) > 50 (after delivery)

Unreliable due to poor sensitivity

Granulocyte colonystimulating factor (pg/mL)

> 200

Good sensitivity but low specificity

Skin and Soft-Tissue Infections Bacteremia can occur in association with skin or soft-tissue foci of infection. Pustular lesions can be a presenting feature of staphylococcal bacteremia, more often with Staphylococcus aureus than with CONS. Other skin manifestations of infection in the neonate include cellulitis, abscess, impetigo, omphalitis, and necrotizing fasciitis.9,12,27 These are more often a consequence of gram-positive infection, especially S. aureus, than of gram-negative infection, but the latter should be considered until culture confirmation. On occasion, infection due to Candida spp. can present with soft-tissue foci, especially abscesses.

Finding Supporting Possible Infection

“falsely normal” results, there was a delay in collection of blood culture from time of the screening test (15 hours) compared with the group with abnormal WBC count (3 hours). A normal WBC should not preclude evaluation for septicemia in an infant with maternal risk factors. Rather, risk factors for septicemia, postnatal clinical signs, and screening tests are used together. When abnormality is detected, prompt initiation of empiric antimicrobial therapy is essential. Interpretation of WBC and absolute neutrophil counts in VLBW neonates is difficult, especially during the first 72 hours of life, despite published reference ranges for these infants.31 C-reactive protein (CRP) is an acute-phase reactant that is in common use to evaluate for and monitor response to therapy in neonates with bacterial infections. It has high sensitivity for diagnosis of infection but can also be elevated in many noninfectious conditions (e.g., respiratory distress syndrome, hypoxia, or intraventricular hemorrhage) that are associated with tissue injury or inflammation.32,33 Interleukin-6 is a proinflammatory cytokine that has been well studied and may assist in the diagnosis of early- or lateonset infection; serial determinations may be needed in the first days of life to avoid overdiagnosis.34 Procalcitonin, a prohormone of calcitonin, undergoes proteolysis into calcitonin in the thyroid gland. Procalcitonin secretion in neonatal sepsis is thought to be induced by bacterial toxins. Elevated procalcitonin levels show promise as a screening tool for neonatal sepsis.32,33

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Bacterial Infections in the Neonate

Microbiologic Techniques Blood culture is the “gold standard” for the diagnosis of neonatal septicemia. The minimum volume for optimal sensitivity is 0.75 to 1.0 mL. Multiple blood cultures increase sensitivity but are not recommended. Documentation of septicemia will be “missed” by a single blood culture in approximately 10% to 15% of instances, but in these infants, sepsis can be presumed from the clinical course despite a single sterile blood culture. It is appropriate to initiate antimicrobial therapy after an evaluation for septicemia has been performed. With computer-assisted automated blood culture systems, virtually all cultures containing clinically significant gram-positive and gram-negative organisms are positive by 24 to 36 hours of incubation, and cultures containing coagulasenegative staphylococci or yeast are positive within 48 hours.35 With current technology, antibiotic therapy can be discontinued in an infant with a benign clinical course if culture results remain negative at 36 to 48 hours. Generally, a lumbar puncture is performed when the blood culture specimen is obtained. If this procedure is deferred because of the infant’s unstable condition, a single negative blood culture result cannot reliably exclude meningitis; the cerebrospinal fluid (CSF) should be examined for evidence of meningitis as soon as the infant’s clinical condition permits. In one review of 39 infants with meningitis confirmed by recovery of a pathogen from CSF culture, 6 (16%) of 39 patients had negative blood culture results with meningitis caused by GBS, gram-negative enteric bacilli, or Staphylococcus aureus.36 The yield of the lumbar puncture is low among healthy-appearing term neonates who are evaluated for bacteremia on the basis of maternal risk factors and in premature infants when respiratory distress is the major sign of possible sepsis.37,38 Some experts propose omission of the lumbar puncture in evaluating these neonates. However, Wiswell and associates39 observed that, in up to one-third of neonates younger than 7 days who had bacterial meningitis, the diagnosis would have been missed or delayed if the CSF had not been assessed. Weighing the benefits of early diagnosis and initiation of appropriate therapy against the minimal risk of the procedure, we believe that CSF should be examined in every neonate evaluated for suspected septicemia. Urine culture need not be performed routinely in the evaluation of infants with early-onset infection. Bladder tap or catheter-obtained urine should be taken for culture when infants are evaluated for

CHAPTER

94

537

septicemia beyond the first few days of life. Cultures from foci of apparent infection are obtained for Gram stain and culture. Examples are middle-ear fluid obtained by tympanocentesis, purulent drainage from the eye, joint fluid obtained by arthrocentesis, metaphyseal aspirate for suspicion of osteomyelitis, peritoneal fluid after rupture of an abdominal viscus, and purulent material from pustules or softtissue abscesses. Cultures from surface sites or mucous membranes are not helpful. Isolates from sites such as the ear canal, nasopharynx, axilla, umbilicus, groin, rectum, stomach, and endotracheal tube are uncommonly the same as those from blood, CSF, or other sterile sites and have a positive predictive value of < 10%.40,41 Cultures from the placenta or gastric aspirate indicate only exposure to a potential pathogen. Positive results of surface, mucosal site, or placental cultures do not dictate further evaluation if an infant has no maternal risk factors predisposing to, or signs of, septicemia.

MANAGEMENT AND PRESUMPTIVE THERAPY The choice of empiric antimicrobial therapy is influenced by: (1) the likely etiologic agents by age and endemicity in a NICU; (2) susceptibility patterns of isolates from infants in a specific NICU; (3) the antimicrobial agent’s central nervous system penetration; (4) toxicity of drugs in neonates (especially ELBW infants); and (5) the infant’s hepatic and renal function. New agents that have not been specifically evaluated for safety and efficacy in the neonate should be avoided. Information concerning dosage schedules by age and birthweight should be available for the agents selected. For example, the half-life of aminoglycosides is prolonged in VLBW neonates and serum levels are monitored to ensure efficacy and to avoid toxicity. Furthermore, a thorough knowledge of an antimicrobial agent’s elimination avoids adverse effects arising from physiologic immaturity of liver and kidney. Empiric therapy for early-onset sepsis, unchanged after three decades of use, consists of ampicillin in combination with gentamicin. Once meningitis is excluded, ampicillin and gentamicin at the doses shown in Table 94-8 are employed.42 Gentamicin serum levels are not required unless therapy is given for > 72 hours, renal function is abnormal or unstable, or birthweight is < 1500 g. Combination therapy provides bactericidal activity against GBS and other streptococci,

TABLE 94-8. Empirical Antimicrobial Therapy for Neonatal Bacterial Infections Clinical Presentation

Antibiotic(s) (dose/kg, IV)

Frequency (h)

Expected Duration (days)

Early-onset (term infant)

Ampicillin (50 mg) plus gentamicin (2.5 mg)

q 8–12a

10

Late-onset term infant readmitted

Ampicillin (50 mg) plus gentamicin (2.5 mg)

q 8–12a

7–10

SEPTICEMIA

Inpatient

a–c

Vancomycin plus gentamicin (2.5 mg) or amikacin (10 mg)

q 12

10–14

Early-onset

Ampicillin (100 mg) plus gentamicin (2.5 mg) plus cefotaxime (50 mg)

q 8a q 12

14–21

Late-onset

Ampicillin (75 mg) plus gentamicin (2.5 mg) or amikacin (15 mg) plus cefotaxime (50 mg)

q6 q 8a,c

14–21

BONE OR JOINT INFECTION

Vancomycinb plus gentamicin (2.5 mg)

q 8a

3–6 weeks

SUSPECTED GASTROINTESTINAL INFECTION

Include clindamycin (10 mg) or piperacillin (50 mg) with aminoglycoside

q 6–8 q 8a,c

10–14

MENINGITIS

a For postconceptual (PC) age < 30 weeks, unit dosage of gentamicin is 3 mg/kg and frequency is determined by serum levels. For PC age 30–37 weeks, unit dosage is 3 mg/kg (≥ 7 days of age) and 2.5 mg/kg thereafter. For PC age > 37 weeks, unit dosage is 2.5 mg/kg. Serum levels should be monitored to achieve a peak of 5–12 mg/mL and a trough of < 2.0 mg/mL. b For PC age < 30 weeks, unit dosage of vancomycin is 20 mg/kg and frequency is determined by serum levels. For PC age 30–37 weeks, unit dosage is 20 mg/kg (≥ 7 days of age) and 15 mg/kg thereafter. For PC age > 37 weeks, unit dosage is 15 mg/kg. Serum levels should be monitored to achieve a peak of 25–40 mg/mL and a trough of 5–15 mg/mL. c Serum levels should be monitored to achieve an amikacin peak of 20–35 mg/mL and a trough of < 10 mg/mL. Adapted from Baker CJ. Neonatal sepsis. In: Kaplan SL (ed) Current Therapy in Pediatric Infectious Disease, 3rd ed. St. Louis, BC Decker, 1993, p 278.42

538

SECTION

N Infections of the Fetus and Newborn

Enterococcus spp., Listeria monocytogenes, many Escherichia coli, and some other enteric bacilli. Use of maternal intrapartum antibiotic prophylaxis (IAP) for prevention of early-onset GBS infection has raised concern that ampicillin resistance will increase among E. coli isolates. Active surveillance from the Centers for Disease Control and Prevention’s Active Bacterial Core indicates that the proportion of E. coli infections that are resistant to ampicillin has increased significantly among preterm infants but not in full-term infants.43 The finding that ampicillin-resistant E. coli are the most common cause of serious bacterial infection in febrile infants less than 90 days in Utah led Byington and colleagues44 to suggest adding a third-generation cephalosporin to the initial treatment of presumed meningitis in young infants. Ongoing surveillance by geographic area will be required to monitor trends in resistance. Although some experts advocate substitution of cefotaxime or another third-generation cephalosporin for gentamicin for empiric treatment of presumed nonmeningeal early-onset infection in neonates, this change is not recommended for the following reasons: (1) superior efficacy to ampicillin and gentamicin has not been demonstrated; (2) third-generation cephalosporins are not active against Listeria or Enterococcus spp.; (3) the routine use of these agents for empiric therapy of sepsis in NICUs is associated with “outbreaks” of sepsis due to multiple-drug-resistant enteric organisms, typically Enterobacter and Serratia spp.;45 and (4) for NICU patients, the concurrent use of ampicillin and cefotaxime (or something for which cefotaxime is a surrogate measure) within the first 3 days of life is associated with an increased risk for death, compared with use of ampicillin and gentamicin.46 In addition, antibiotic-induced derepression of b-lactamase leads to resistance to all third-generation cephalosporins. One crossover study found an 18-fold higher rate of colonization with strains resistant to the empiric regimen for earlyonset sepsis in infants receiving amoxicillin and cefotaxime than in those being treated with a regimen of penicillin and tobramycin.47 Thus, cephalosporins must be prescribed selectively, not routinely, for suspected neonatal infections. For the term infant up to 8 weeks of age who is readmitted to the hospital for possible septicemia without an apparent focus of infection, ampicillin and gentamicin are also appropriate for initial empiric therapy unless Staphylococcus aureus infection is suspected. Additional or alternative agents are given for suspected or apparent focal infections. For meningitis, cefotaxime is given either in addition to ampicillin and gentamicin or instead of gentamicin. For skin, softtissue, bone, or joint infection, vancomycin is substituted for ampicillin. For septicemia of suspected gastrointestinal origin, therapy for anaerobic pathogens should be improved by adding clindamycin or another suitable agent. For the patient previously treated with gentamicin, amikacin could replace gentamicin, but gentamicinresistant gram-negative organisms are uncommon outside NICU outbreak settings. For suspected late, late-onset sepsis occurring within the nursery for which commensal species and MRSA are possible pathogens, vancomycin and an aminoglycoside are empiric agents of choice. Staphylococci rarely cause central nervous system infection, except in association with intraventricular hemorrhage, instrumentation, or neurosurgery. Limited experience indicates that linezolid is well tolerated and is as effective as vancomycin in the treatment of resistant gram-positive infections in neonates.48 If nosocomial meningitis is suspected or proved, more appropriate initial therapy would consist of vancomycin rather than ampicillin, a b-lactam agent such as cefotaxime or perhaps ceftazidime if Pseudomonas is suspected, plus gentamicin. Surveillance of aminoglycoside activity against gramnegative bacilli in NICU infants may lead to the need to cross-over empiric choice to amikacin periodically. Prompt drainage of suppurative infections and removal of intravascular catheters for persistent bacteremia, severe clinical illness, or unfavorable microbiology are paramount (see Chapter 96, Nosocomial Infections in the Neonate).

MENINGITIS The neonate with suspected or proven bacterial meningitis deserves special comment. First, only examination of the CSF excludes this diagnosis: no clinical or indirect laboratory test provides a definitive answer. If the CSF is not obtained initially, therapy for meningitis should be given until the CSF is examined. Second, infants with meningitis often have progression of clinical illness after antimicrobial therapy is initiated. Observation under intensive care is required during the first 24 hours. Third, bactericidal rather than bacteriostatic agents should be chosen, and for gram-negative meningitis, and concurrent use of two active agents is ideal. Fourth, doses used should be high enough to achieve bactericidal concentrations in the CSF but to avoid associated toxicity (see Table 94-8). The dose of cefotaxime for suspected meningitis is 50 mg/kg per dose at a frequency of every 12 hours in the first week of life and every 8 hours in weeks 1 through 4 of life. Once the pathogen has been identified, antimicrobial susceptibility determined, and the CSF proven to be sterile, therapy is modified to the most active and least toxic agent(s). For GBS, aqueous penicillin G alone at a dose of 450 000 to 500 000 mg/kg per day, and for L. monocytogenes, ampicillin at 300 mg/kg per day are drugs of choice. The high dose of penicillin G for GBS meningitis is suggested because it is often a highinoculum infection (median, 1 μ 106.5 cfu/mL of CSF) in which the concentration of ampicillin required for bactericidal activity is 100 times that for penicillin-susceptible pneumococci or meningococci.42 For enteric pathogens, combination therapy is suggested for the first 7 to 14 days of treatment, consisting of a b-lactam with an aminoglycoside agent, usually gentamicin. Some exceptions occur, such as ampicillin-susceptible Escherichia coli or Proteus mirabilis meningitis, in which ampicillin alone has been demonstrated to be efficacious. Failure of antibiotic therapy to sterilize the CSF within 24 to 36 hours suggests a complication of meningitis, such as cerebritis, ventriculitis (often with obstruction), subdural empyema, or early abscess formation. The duration of therapy is longer than for older infants and children and is pathogen-specific. For meningitis caused by GBS, therapy is administered for a total of 14 to 21 days; for L. monocytogenes, 14 days is recommended; and for enteric bacilli, 14 days after sterilization of CSF or 21 total days (whichever is longer) is indicated.42 For the optimal management of neonatal meningitis, it is necessary to document sterilization of CSF 24 to 48 hours after optimal therapy has been initiated. Finally, the diagnosis of meningitis in the neonate also mandates supportive care not routinely given to the infant with septicemia, such as careful observation for and control of respiratory failure and seizures, fluid restriction to prevent or treat the syndrome of inappropriate antidiuretic hormone secretion, and care that minimizes elevation of intracranial pressure.

RECENT ADVANCES Ongoing morbidity and mortality from neonatal sepsis despite the use of potent antimicrobial agents and advances in critical care medicine have prompted interest in adjunctive therapies. Most are aimed at downregulating the putatively harmful host inflammatory response or improving host defense mechanisms. High levels of interleukin-1b in the CSF have been shown to correlate directly with mortality and permanent neurologic impairment in neonates with meningitis. In animal models of meningitis, dexamethasone treatment decreases these levels after experimental challenge; thus, corticosteroids have been considered as adjunctive therapy for meningitis. Until a controlled clinical trial documents the efficacy and safety of corticosteroids, however, their use is not recommended. Several clinical trials have evaluated the efficacy of immune globulin intravenous (IGIV) in reducing the mortality of neonatal septicemia.49–51 These indicated a reduction in mortality in IGIV-

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Viral Infections in the Fetus and Neonate

treated patients, but none of the studies was placebo-controlled or randomized, and none had > 50 patients in each study group. In addition, some of the investigators used clinical rather than microbiologic criteria to document septicemia. A meta-analysis showed a significant decrease in mortality rate for neonates who received IGIV treatment.52 It appears that IGIV in this setting is safe when doses of 1 g/kg or less are employed, but until multicenter, appropriately controlled clinical trials are performed, the use of IGIV as adjunctive therapy in neonatal septicemia is experimental, and its routine use is inappropriate. Improvement of the opsonic capacity of neonatal serum through infusion of commercially available IGIV has been demonstrated for GBS, but most investigators believe that the relatively low concentrations of antibodies to the capsular antigens of these organisms contained in such products mandate the development of high-titered pathogen-specific IGIV preparations or monoclonal antibodies. Both have been developed experimentally for GBS and have improved survival in animal models of sepsis; neither has been prepared commercially or undergone clinical trial. Compared with those of older infants, polymorphonuclear leukocytes of neonates, and especially preterm neonates, show deficient mobilization of receptors for immunoglobulin or complement, chemotaxis, and deficient and rapid exhaustion of storage pools during “stress,” especially septicemia. No large, wellcontrolled study of polymorphonuclear leukocyte transfusion has been performed, but most early studies suggested improvement in mortality among transfused infants. Concerns arise regarding the safety and practicality of this mode of therapy, given the need for irradiated cells collected 24 hours a day from donors who are seronegative for human immunodeficiency virus, cytomegalovirus, and hepatitis B. This mode of therapy, as well as IGIV, should be considered experimental. Other potential adjunctive therapies are the use of recombinant granulocyte, monocyte, and granulocyte–macrophage colonystimulating factors. Of these, recombinant human granulocyte colonystimulating factor offers the greatest theoretic promise. It has been shown to increase the neutrophil count significantly and to reduce the incidence of neonatal sepsis in critically ill ventilated neonates with prolonged pre-eclampsia-associated neutropenia compared with conventional therapy.53 The future is bright for the development of adjunctive agents, but their use by clinicians should await results from multicenter, controlled clinical trials documenting their safety as well as efficacy.

PREVENTION Three general approaches for the prevention of neonatal septicemia have been suggested, and the first two have been efficacious: (1) improvement in prenatal care resulting in delivery of infants at term gestation and without maternal risk factors for septicemia; (2) maternal IAP for prevention of early-onset GBS septicemia;54 and (3) maternal immunoprophylaxis providing IgG-mediated passive immunity for the infant to prevent early- and late-onset GBS septicemia and, potentially, septicemia caused by other etiologic agents. Although the first approach, provision of prenatal care, especially for women younger than 20 years, is self-evident, its achievement continues to elude urban centers. A decline in the rate of preterm deliveries has been shown to reduce the incidence of neonatal septicemia and would eliminate most very late-onset septicemia and nosocomial infections in NICUs. The second approach, IAP, targets women who are antenatally identified as carriers of GBS to prevent early-onset neonatal septicemia caused by this organism. Sepsis caused by other organisms is more often a disease of prematurity.55 Implementation of IAP has been associated with a 65% decrease in the incidence of early-onset GBS infections.56 A discussion of this and other approaches can be found in Chapter 119, Streptococcus agalactiae (Group B Streptococcus).

CHAPTER

95

539

Whereas IAP is a desirable interim approach, prevention of all GBS infections, irrespective of age of onset or presence of maternal risk factors, awaits the development of suitable vaccines. Such capsular polysaccharide-protein conjugate vaccines are immunogenic in experimental animals, and phase 1 and 2 clinical trials have been completed. If such vaccines are effective in phase 3 trials, women of childbearing age could be immunized to protect their neonates against GBS and possibly other neonatal pathogens. This approach has been successful in the prevention of tetanus.

CHAPTER

95

Viral Infections in the Fetus and Neonate Robert F. Pass

Infection of the fetus and newborn accounts for much of the public health significance of rubella virus, parvovirus B19, human immunodeficiency virus (HIV), cytomegalovirus (CMV), herpes simplex virus (HSV), and hepatitis B virus (HBV). Less commonly encountered in the perinatal setting are varicella-zoster virus (VZV), enteroviruses, measles virus, mumps virus, and human T-cell lymphotropic virus type I (HTLV-I), among others. Viral infections of the fetus and newborn can be considered in terms of source, chronicity of infection in both mother and neonate, and clinical manifestations. Although neonates are subject to community-acquired infections with viruses such as respiratory syncytial virus (RSV), rotavirus, and enteroviruses, infection, primarily from a maternal source, accounts for the distinctive features of perinatal virus infections. Virus can be transmitted from mother to infant transplacentally, during birth, and through breastfeeding. Viruses that establish chronic infection in the mother with persistence of infectious virus (herpesviruses, HIV, HTLV-I, hepatitis B and C) in blood, mucosa, or milk are often transmitted from mother to fetus or neonate. Whether viruses that produce acute, self-limited infections in the mother, such as rubella, VZV, enteroviruses, and parvovirus B19, are transmitted to the fetus or newborn and produce disease depends on the timing of maternal infection in relation to gestation. Viral infections in the newborn may be symptomatic or clinically silent; however, their impact may not be fully evident for years. This chapter provides a framework for the diagnosis and management of viral infection in the fetus and neonate. Detailed discussions of diagnosis, treatment, and prevention of specific infections are presented in the chapters focused on individual viruses.

PATHOGENESIS Many viral infections produce disease that is much more severe in the fetus or neonate than in adults, children, or infants. The pathogenesis of most postnatal viral infections involves: (1) inoculation of virus on to mucosal surfaces; (2) local replication; (3) primary viremia; (4) further replication at the site of inoculation and in regional and distant reticuloendothelial tissue; and (5) secondary viremia with dissemination of virus to multiple tissues, possibly resulting in clinical illness or organ dysfunction.1 Time from inoculation to secondary viremia corresponds to the clinical incubation period. During this interval, local and systemic host immune responses develop. Viral infection of the fetus probably follows maternal secondary viremia or viral

540

SECTION

N Infections of the Fetus and Newborn

replication in the placenta. Developmental immaturity of fetal cellular and humoral immune function is undoubtedly important (see review by Lewis & Wilson2). Viral infections that reach the fetus early in gestation are usually more virulent.3–5 When the fetus is infected before the third trimester of gestation, fetal viremia occurs in the absence of substantial concentrations of maternal antibodies. Infections early in gestation encounter an immature immune system and developing fetal organs. Tissue damage, organ dysfunction, and teratogenicity are likely consequences of these infections. Maternal antibody acquired in utero and antibody and immunocompetent cells present in colostrum and mother’s milk are important components of the neonate’s defense against viral infection. A neonate infected by a virus to which the mother lacks immunity is prone to severe infection; neonatal infections caused by herpesviruses are illustrative. Transfusion-acquired CMV infections are rarely evident clinically in infants of seropositive mothers, but neonatal CMV infections result in severe illness in small, premature infants born to seronegative women. HSV and VZV are more likely to cause severe disease in the neonate with absent or low concentrations of maternal antibodies. Vaginal birth and breastfeeding provide opportunities for virus transmission that are unique to the newborn and infant. Labor and delivery prolong the contact between the neonate’s mucosal surfaces and maternal secretions and blood, facilitating transfer of viruses. Although the quantity of virus present in human milk is usually low, the nursing infant is exposed to this potential source of infection multiple times per day for months. This likely accounts for the consistent transmission of CMV, HIV, and HTLV-I by this route.

not have sequelae. For example, most cases of congenital CMV infection are not associated with illness; about 15% to 20% cause neonatal disease or long-term neurological, auditory, or visual sequelae. In contrast, most cases of congenital varicella, neonatal herpes, congenital rubella, and perinatal HIV are associated with substantial morbidity. Congenital parvovirus B19 infection is usually asymptomatic in infancy, without known sequelae; only a small proportion of affected fetuses experience severe anemia in utero and are stillborn. Perinatal hepatitis B and C infections are important because a high proportion of infected newborns develop chronic infection and chronic liver disease. For viruses that cause acute, self-limited infections, such as rubella, enteroviruses, RSV, influenza, and parvovirus B19, the incidence of maternal infection is determined by epidemic activity in the community, which can vary substantially by season, from year to year, and among subgroups of the population. For example, outbreaks of rubella infection have occurred in unimmunized immigrants; these outbreaks were accompanied by cases of congenital rubella.7 Congenital infection due to lymphocytic choriomeningitis virus is likely underdiagnosed. Maternal infection, especially in the first trimester, can lead to fetal infection with subsequent chorioretinitis, micro- or macrocephaly, and intracranial calcifications. Maternal exposure to rodents may be the key epidemiologic clue.15 The incidence of fetal and neonatal viral infections varies substantially among different populations according to prevalence and incidence of maternal infection during pregnancy. Immunization practices have had a significant effect on maternal and perinatal infections due to VZV, rubella virus, and HBV.

The Importance of Chronic Viremia EPIDEMIOLOGY Leading Causes of Perinatal Viral Infection Estimated rates of neonatal infection from transplacental or intrapartum transmission of selected viruses in the United States are shown in Table 95-1. The numbers in this table are rough estimates for comparison purposes; they are derived from a variety of sources, including cohort studies, expert opinions, and cases reported to public health authorities.6–12 Congenital varicella infection is probably more common than congenital rubella, but it is not included as recent data on rates of gestational or congenital infection are not available. Enterovirus infections appear to be common in neonates due to either sporadic cases or nursery epidemics associated with community outbreaks.13,14 However, it is difficult to determine the role of vertical transmission versus postnatal acquisition of enteroviruses in neonatal infections. Clinical impact on the newborn is not directly apparent from the estimates of numbers of infections listed in Table 95-1 because a variable proportion of infected newborns are normal at birth and do

TABLE 95-1. Estimated Annual United States Cases of Selected Fetal and Neonatal Viral Infectiona Virus 6

Cytomegalovirus (congenital) Herpes simplex9 Hepatitis C virus8 Human immunodeficiency virusc12 Parvovirus B1910 Hepatitis B virusd11 Rubella (1994–2003)7 a

Routeb

Cases/Year

T I I I/T T I T

40,000 1500–2200 2400 436 11,000 9500 1–9

Estimates are for cases transmitted from mother to offspring, assuming 4 million births a year. Principal route of acquisition: T, transplacental; I, intrapartum. c Provisional data, cases in children ≤ 13 years, 2004. d Estimate based on neither prenatal screening nor universal newborn hepatitis B virus immunization. b

For viral infections characterized by chronic viremia, the prevalence of maternal infection determines the rate of congenital or perinatal infection in a population. From 14% to 42% of infants born to HIVpositive mothers are infected.16 Around 50% of infants born to chronic carriers of HBV are infected and most become chronic carriers. Around 3% of infants born to hepatitis C virus (HCV)-positive mothers acquire the virus.17–19 Transmission rates of hepatitis C are higher if the mother has concurrent HIV infection. Intrapartum transmission of hepatitis C is suspected because of the strong association between maternal viremia (RNAemia) at term, the fact that in most infants HCV RNA appears in plasma initially between 1 and 4 months of age, and the observation that greater exposure to maternal blood during delivery seems to increase risk of transmission.19,20 However, elective cesarean section delivery does not appear to decrease the risk for mother-to-infant transmission of HCV.21 GB virus C, or hepatitis G virus (HGV), is a flavivirus presumed to be hepatotropic because it was initially detected in sera from patients with hepatitis.22 Serologic evidence of HGV infection was detected in 9% of blood donors, and viremia was detected by polymerase chain reaction testing in up to 2%.23,24 Infection rates are higher in intravenous drug users, recipients of multiple blood transfusions, and people with other bloodborne infections, such as HIV-1, HBV, and HCV. Studies of infected mothers and their offspring show that 60% to 80% of infants born to mothers with HGV RNA in their blood are infected.25,26 Sequencing and phylogenetic analysis of the NS3 region of the HGV RNA from maternal and neonatal viral strains have supported vertical transmission.27 Although mildly and transiently elevated serum alanine aminotransferase concentrations have been noted, there has been no associated illness in infants with perinatal HGV infection.25,28 TT virus, a nonenveloped, single-stranded DNA virus, was discovered by Japanese investigators in a patient with posttransfusion hepatitis and was named from the patient’s initials.29 Epidemiologic studies using polymerase chain reaction testing suggest that TT virus is common in healthy persons of all ages, though its role as a cause of liver disease or other clinical problems is uncertain. Some experts have suggested that detection of TT virus DNA in cord blood from neonates born to TTV-positive mothers and the finding of high rates of infection

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Viral Infections in the Fetus and Neonate

in infants are evidence of vertical transmission.30–32 Others, however, find little evidence of transplacental or parenteral transmission early in infancy, but report rising rates of infection after the first postnatal months, consistent with breast milk transmission or horizontal infection from family or community sources early in life.33–35

Intrapartum and Breast Milk Transmission Viruses present in the maternal genital tract and those that persist in blood can be transmitted to the neonate at the time of birth. Intrapartum transmission of CMV is common; about 10% of seropositive mothers shed CMV in the cervix or vagina at term, and approximately 50% of offspring exposed to the virus during birth are infected.36 Approximately 85% of neonatal HSV infections are acquired by exposure to the virus during birth. Evidence suggests that most vertically acquired HBV infections result from exposure to maternal blood at delivery.37 Transmission of HIV can occur transplacentally but infection at delivery is more common.38 Vertical transmission of viruses through ingestion of human milk is important in the epidemiology of CMV, HTLV-I, HIV, and possibly HGV and TT virus (Table 95-2). Breastfeeding is the major route for transmission of CMV from mother to offspring. In populations in which mothers nurse their infants routinely and rates of maternal seropositivity are high, most infants acquire CMV during the first year of life.39 Transmission of CMV through mother’s milk rarely causes acute illness or the types of sequelae that follow congenital infection; a possible exception may be breast milk-acquired CMV infection in very-low-birthweight premature newborns (see Chapter 206, Cytomegalovirus). Both HIV-1 and HTLV-I can be transmitted through human milk, although the onset of infection is usually after the neonatal period.40–42 Because of the consistent association of higher infant mortality with formula feeding, the potential transmission of maternal viral infection should rarely be a reason to interdict breastfeeding in developing countries, with the possible exception of HIV infection. The World Health Organization has acknowledged that use of formula may be preferable in some settings to reduce transmission of HIV if a safe, nutritionally adequate substitute for mother’s milk can be provided. In developed countries with low rates of infant mortality due to contaminated formula, mothers infected with HIV or HTLV-I should not breastfeed their infants.42,43 HBV can be detected in milk from most carrier mothers; however, breastfeeding does not appear to affect the rate of vertical transmission. Breast milk samples from 73 chronically HCV-infected women had no detectable virus, despite viremia in 60% of women. Therefore, hepatitis C is not a contraindication to breastfeeding.44 HSV and rubella virus have been detected in human milk in unusual circumstances, but milk is not a substantial factor in the vertical transmission of either agent.

Horizontal and Nosocomial Routes of Transmission Horizontal transmission of viruses to neonates from caregivers or family members occurs primarily through infected droplets or TABLE 95-2. Viruses Transmitted Through Milk from Mother to Offspring

CHAPTER

95

541

contaminated hands. Neonates are more vulnerable than older hosts because they are immunologically naive and their care requires repeated handling and close contact. In addition, hospitalized neonates are exposed to a continual influx of hospital personnel and new patients, creating multiple opportunities for the introduction and spread of viruses prevalent in the community. Nosocomial outbreaks of enterovirus, RSV, HSV, VZV, and rotavirus have been described. Neonates are infected by the mother, other family members, or hospital personnel, or infection can be endemic to nurseries.45–49 Blood products are a potential source of nosocomial infection due to HIV, HTLV-I, HBV, HCV, and CMV.

CLINICAL MANIFESTATIONS Prematurity and Low Birthweight Maternal viral infections during pregnancy can affect fetal growth or lead to premature birth, although their importance in these morbidities is ill defined.50–53 A virologic and serologic study of small-forgestational-age neonates in Sweden did not find evidence that viral infection was causal.51 Clinically apparent congenital infections are associated with premature birth and low birthweight, although there is considerable variability among causative agents. For example, congenital infections due to rubella and CMV are associated with prematurity and low birthweight, but those due to parvovirus B19 and HIV infections are not.

Spontaneous Abortion and Stillbirth Associations between spontaneous abortion or stillbirth and a number of viruses, including poliovirus, measles, rubella, mumps, influenza, parvovirus B19, HSV, CMV, and nonpolio enteroviral infection, have been reported.54–62 Most agents have been isolated from products of conception, aborted fetuses, or stillborn infants. A study of outcome of pregnancy in HIV-infected intravenous drug users suggests that, in the absence of acquired immunodeficiency syndrome (AIDS) in the mother, HIV infection does not increase the risk of spontaneous abortion or stillbirth.63 A review of spontaneous abortions and fetal deaths found evidence of viral infection in 16 of 21 cases and in none of 26 controls; enterovirus/coxsackievirus accounted for 10 of the cases.64 The same virus was detected in placenta and fetal tissue. Many unexplained spontaneous abortions and fetal deaths could be due to unrecognized viral infection.

Syndrome of Congenital Infection The presence of hepatomegaly, splenomegaly, microcephaly, petechiae, jaundice, dermal manifestations of erythropoiesis, chorioretinitis, intracranial calcifications, deafness, thrombocytopenia, direct hyperbilirubinemia, or hepatitis in the neonate suggests prenatal infection. Other findings occasionally associated with congenital infection are cardiac defects, hydrocephalus, prematurity, hemolytic anemia, and low birthweight for gestational age. Clinical findings suggestive of infection by a specific viral agent are listed in Table 95-3. However, clinical findings are rarely diagnostic of a specific

Virus

Epidemiologically Significanta

Associated Disease

TABLE 95-3. Clinical Findings in the Newborn Suggesting a Specific

CMV HIV-1 HTLV-I Hepatitis B Rubella vaccine virus

Yes Yes Yes No No

None AIDS Adult T-cell leukemia Carrier, chronic liver disease None or mild rash

Congenital Viral Infection

AIDS, acquired immunodeficiency syndrome; CMV, cytomegalovirus; HIV-1, human immunodeficiency virus type 1; HTLV-I, human T-cell lymphotropic virus type I. a Shown to influence the incidence of infection in the infant population.

Agent

Features

Cytomegalovirus

Microcephaly, hearing loss, petechiae, cholestatic jaundice Cataracts, heart disease, deafness Scarring of skin, limb hypoplasia, ocular abnormalities Hydranencephaly, ocular disease, skin scars

Rubella Varicella-zoster virus Herpes simplex virus

542

SECTION

N Infections of the Fetus and Newborn

cause, so laboratory evaluation is essential in patients with suspected congenital infection.

Central Nervous System Infection Viral encephalitis in the neonate can result from prenatal or postnatal infection. Abnormalities suggestive of viral central nervous system disease are often subtle or nonspecific; they include lethargy, hypotonia, irritability, poor feeding, apnea, fever and seizures. Prenatal viral infections can affect brain growth, leading to microcephaly. Although encephalopathy or encephalitis in the neonate has been noted with a number of viral infections, HSV and enteroviruses are the most common causes.

Sepsis Syndrome Clinical manifestations suggestive of septicemia are sometimes associated with neonatal enterovirus, HSV, or RSV infections. Neonates with echovirus or coxsackievirus infection can manifest pallor, lethargy, hypotension, apnea, acidosis, and respiratory impairment.46 Rapid progression of disseminated HSV infection can manifest as shock, coagulopathy, fulminant hepatitis, and diffuse lung disease either before or in the absence of vesicular skin lesions.65 Neonatal RSV infection can also cause nonspecific signs that mimic those of septicemia, such as apnea, lethargy, irritability, and poor feeding.45

Cardiac Insufficiency Viral infections can produce congestive heart failure in the fetus or neonate by causing anemia or by directly damaging the myocardium. Fetal infection with human parvovirus B19 characteristically leads to profound anemia. Although many fetuses appear to recover in utero, hydrops fetalis and fetal death can result.66 Rare cases of nonimmune hydrops fetalis have also been attributed to intrauterine CMV or adenovirus infection.67,68 Viral myocarditis in the neonate is usually due to coxsackie B viruses or echoviruses.46 Enterovirus myocarditis can present as shock or cardiac failure and has a poor prognosis.69

TABLE 95-4. Newborn Ocular Abnormalities Associated with Congenital and Neonatal Viral Infection Abnormality

Agents

Cataracts Chorioretinitis Optic atrophy Microphthalmia Coloboma Keratoconjunctivitis Pigment retinopathy Glaucoma Iritis Anophthalmia Peter anomalya Horner syndromeb

Rubella, HSV, VZV HSV, VZV, rubella, CMV HSV, VZV, rubella, CMV CMV, rubella CMV HSV Rubella Rubella HSV, rubella CMV CMV VZV

CMV, cytomegalovirus; HSV, herpes simplex virus; VZV, varicella-zoster virus. a Central corneal/anterior-chamber synechiae, cataract. b Ptosis, meiosis, and ipsilateral absence of facial sweating.

are important in congenital or neonatal viral infection because they may be predictive of central nervous system involvement.

Deafness Deafness is commonly associated with congenital infection by rubella virus and CMV. Newborns with diminished hearing of unknown etiology should be evaluated for congenital infection. Congenital CMV infection accounts for approximately a third of all cases of sensorineural hearing loss in children in the United States.80 Because congenital CMV infection can cause progressive hearing loss, serial hearing evaluations throughout infancy and early childhood are recommended.

APPROACH TO THE NEONATE WITH SUSPECTED VIRAL INFECTION Differential Diagnosis

Pneumonia Lower respiratory tract disease is an unusual manifestation of fetal viral infection. When it occurs, lower respiratory tract disease is usually part of multisystem infection. Viruses associated with pneumonia in the neonate include RSV, parainfluenza virus, influenza A, adenovirus, enteroviruses, CMV, rubella virus, VZV, and HSV.45,65,70–74 Nonspecific signs are often more prominent than lower respiratory disease in neonates infected by common respiratory viruses. A surveillance study in preterm infants on mechanical ventilation found that 29.5% had a viral infection; RSV and enteroviruses accounted for the majority of infections.75

Ocular Abnormalities Table 95-4 lists ocular abnormalities found in patients with congenital and neonatal viral infections.76–79 Careful examination with the use of indirect ophthalmoscopy is an important part of the evaluation of neonates with suspected congenital infection. Abnormalities of the cornea and iris, chorioretinitis, vitritis, optic atrophy, and pigment retinopathy are the abnormalities most often detected in neonates with congenital infection. Chorioretinitis can be evident as scarring or as active lesions, sometimes accompanied by vitritis. Ocular abnormalities are usually associated with evidence of infection of other organs. In addition to association with visual impairment, ocular signs

The presence, singly but especially in combination, of hepatomegaly, splenomegaly, petechiae, purpura, jaundice, microcephaly, encephalopathy, ocular abnormalities, anemia, thrombocytopenia, conjugated hyperbilirubinemia, or elevated serum hepatic transaminases should prompt the consideration of congenital viral infection. Nonspecific signs, such as fever, lethargy, anorexia, respiratory symptoms, and a sepsis-like syndrome, also suggest the possibility of perinatal viral infection as well as bacterial or fungal infection. Prenatal viral and nonviral infections, especially syphilis and toxoplasmosis, can have clinically indistinguishable manifestations. Miliary tuberculosis is rare and should be easily distinguished from viral infection, although central nervous system manifestations, hepatomegaly, and splenomegaly might initially suggest congenital infection. Differential diagnosis includes noninfectious causes. Inborn errors of metabolism that manifest in the neonatal period can cause encephalopathy, elevation of serum hepatic enzymes, thrombocytopenia, anemia, enlarged liver and spleen, jaundice, retinal pigment defects, and cataracts. Hypoglycemia, acidosis or alkalosis, hyperammonemia, crystalluria, urinary reducing substances, and a positive urine ferric chloride test result are clues to the presence of metabolic disease. Liver disease associated with neonatal giant-cell hepatitis, biliary atresia, choledochal cyst, or intestinal obstruction can lead to hepatomegaly, splenomegaly, serum transaminase elevation, and cholestatic jaundice. Hematologic abnormalities accompanying rhesus or ABO isoimmunization, red blood cell biochemical defects, red blood cell structural defects leading to anemia and hyperbilirubinemia,

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Nosocomial Infections in the Neonate

and immunologically mediated thrombocytopenia, can be confused with congenital infection. The fetus can be adversely affected by maternal medication or illicit drug use. Fetal exposure to alcohol, anticonvulsants, or cocaine can impair brain growth, leading to microcephaly and neonatal encephalopathy similar to those observed in congenital infection. The prolonged use of intravenous vitamin E in premature infants has been associated with thrombocytopenia, encephalopathy, and cholestatic jaundice that could be confused with signs of congenital or neonatal viral infection. Congenital leukemia and neuroblastoma can manifest as anemia, thrombocytopenia, and organomegaly. Finally, some infants with chromosomal defects have microcephaly, encephalopathy, jaundice, anemia, and thrombocytopenia.

Laboratory Diagnosis Laboratory testing should be targeted and should focus on identification of virus by culture or detection of viral nucleic acid or proteins in the appropriate specimen. The most likely etiologies should be selected on the basis of clinical findings. The specimens required and the approach used depend on the specific viral infection being considered, and are discussed in the chapters dealing with each virus. Measurement of maternal or neonatal antibody responses is of limited value but can contribute useful information in certain circumstances. Negative immunoglobulin (Ig) G antibody results for specific agents indicate that maternal infection is not present or is very recent, virtually eliminating the possibility of congenital infection. Newborn IgM antibody response has been used to diagnose congenital infections due to CMV, rubella, VZV, and parvovirus B19 as well as nonviral infections, such as syphilis and toxoplasmosis. However, the accuracy of IgM antibody testing of neonatal serum for viral diagnosis is highly variable, depending on the agent, assay used, and laboratory. Most commercial assays are not reliable. When clinical evidence suggests congenital infection, virus culture or polymerase chain reaction testing should be used to confirm positive or negative IgM antibody test results. For viruses that produce chronic viremic infection, such as HTLV-I, HIV, HBV, and HCV, it is useful to know whether the mother has been infected in order to gauge the risk to the neonate. Tests for IgG antibody can provide this information. Results of test panels for IgG or IgM antibody to multiple possible causes of infection (“TORCH titers”) usually fail to establish an etiologic diagnosis or are not relevant; they should not be used as the sole laboratory diagnostic assay.81

Treatment Antiviral treatment is playing an increasingly important role in the management of congenital and neonatal viral infections. Acyclovir treatment of neonates with HSV infection can be life-saving and may improve the quality of life for survivors.82 When HSV infection is suspected, initiation of empiric antiviral treatment is usually indicated. Acyclovir is also used for perinatal VZV infection.9 Antiviral treatment for severe symptomatic, congenital CMV infection with 6 weeks of intravenous ganciclovir decreases the risk for progressive or late-onset hearing loss.83 Pleconaril was studied for the treatment of severe enterovirus infection in newborns; results under compassionate use were encouraging but production of drug by Vira Pharma was halted; Schering-Plough has taken up development.84 Perinatal HIV infection can be diagnosed within the first month of life, and antiretroviral treatment should be initiated as soon as the diagnosis is made (and Pneumocystis carinii prophylaxis is initiated when HIV exposure is confirmed). Among other bloodborne, vertically transmitted agents, HBV and HCV may require treatment during childhood if chronic liver disease occurs. Management of infants with congenital or neonatal viral infection involves provision of supportive care and anticipation of complications, such as hearing loss, mental retardation, cerebral palsy, and chronic liver disease. Anticipatory follow-up

CHAPTER

96

543

should be planned, with focused examinations for possible sequelae and intervention as needed.

CHAPTER

96

Nosocomial Infections in the Neonate M. Gary Karlowicz and E. Stephen Buescher

As progressively smaller premature infants survive beyond the first few days of life, nosocomial infections have emerged as a major cause of morbidity and late mortality in the neonatal intensive care unit (NICU). Effective prevention and treatment of nosocomial infections in the NICU require understanding of the distribution of pathogens that cause these infections, the various patient-related risk factors for these infections, and the roles of medications and invasive procedures in predisposing to their occurrence.

EPIDEMIOLOGY AND ANATOMIC SITES OF INFECTION Bloodstream infections (BSIs) are the most common nosocomial infections in the NICU, and they can occur in isolation or in association with organ infections, including endocarditis, osteomyelitis, and septic arthritis. Less commonly, nosocomial infections in neonates can involve the lungs, meninges, urinary tract, peritoneum, bowel, conjunctivae, or skin (Table 96-1).

Late-Onset Sepsis Late-onset sepsis is usually defined as BSIs occurring on or after 7 days of age in neonates. It accounts for the majority of nosocomial infections in all birthweight groups in NICUs. Late-onset sepsis is especially important in very-low-birthweight (VLBW, birth weight < 1500 g) infants, in whom its occurrence increases hospital length of stay by 19 days and causes 45% of deaths beyond 2 weeks of age.1 Stoll et al.1 reported that late-onset sepsis occurred in 21% of VLBW infants who survived beyond 3 days of age in the National Institute for Child Health and Human Development (NICHD) Neonatal Research Network, and similar rates have been reported for the Neonatal Networks in Canada2 (24%) and Israel3 (30%). At the institutional level, the prevalence of late-onset sepsis in VLBW infants is more variable: 11% to 32% in the NICUs of the NICHD Neonatal Research Network,1 and 7% to 74% in the NICUs participating in the Canadian Neonatal Network.2 The rate of lateonset sepsis was strongly and inversely associated with birthweight and gestational age,1 decreasing from 43% for infants with birthweights 401 to 750 g, down to 7% for those 1251 to 1500 g; and decreasing from 46% for neonates with gestational age < 25 weeks, down to 10% for 29 to 32 weeks. Consequently, institutions caring for more extremely premature infants have higher apparent rates, and management practices, particularly those concerning utilization of central venous catheters (CVCs) or peripherally inserted central catheters (PICCs), can further impact these figures. Most cases of late-onset sepsis in neonates are associated with CVCs or PICCs,4 and are referred to as CVC-BSIs. The Centers for Disease Control and Prevention (CDC) definition5 for a CVC-BSI includes: (1) isolation of a pathogen from one blood culture or of a skin commensal from two blood cultures; (2) one or more clinical signs of infection (e.g., apnea, bradycardia, or temperature instability); and (3) presence of a CVC at the time the blood culture is obtained. A

544

SECTION

N Infections of the Fetus and Newborn

TABLE 96-1. Common Sites and Causes of Nosocomial Infections in the Neonatal Intensive Care Unit Anticipated Causal Organisms Site of Infection

CONS

S. aureus

Enterococci

GNR

Candida

Viruses

BSI CVC-BSI Osteomyelitis/septic arthritis Endocarditis Meningitis VAP Peritonitis UTI Conjunctivitis Skin or subcutaneous tissue

+++ +++ – + +++ – + – + +

++ ++ +++ +++ + + – – + +++

++ + – + + – + + – –

++ ++ + + ++ +++ +++ +++ + +

+ ++ + + ++ + + ++ – +

– – – – + +a – – – +

BSI, bloodstream infection; CONS, coagulase-negative staphylococci; S. aureus, Staphylococcus aureus; CVC-BSI, central venous catheter-related bloodstream infection; GNR, gram-negative rods; UTI, urinary tract infection; VAP, ventilator-associated pneumonia. +++, most common isolate; ++, frequently; +, occasionally; –, rarely or not. a Includes respiratory syncytial virus, influenza virus, parainfluenza viruses, and enterovirus.

nosocomial infection rate that is linked with device utilization, like the CVC-BSI, helps control for variability in management practices from institution to institution, and the preferred unit of measure is infections per 1000 catheter-days. The National Nosocomial Infection Surveillance System (NNIS) currently recommends that CVC-BSI be the major focus of surveillance and prevention efforts in NICUs, and to that end, provide summary data on CVC-BSI dates for different birthweight groups in the NICU. In 2004, these data on CVC-BSI rates from 104 NICUs reported a median of 8.5 BSIs per 1000 catheter-days for infants with birthweights ≤ 1000 g (extremely-low-birthweight or ELBW) and a median of 4.0 BSI per 1000 catheter-days for those weighing 1001 to 1500 g.6 These values represent a 30% to 38% decrease in birthweight-specific CVC-BSI in NICUs since 1998, when the NNISreported median CVC-BSI rates for ELBW infants and infants weighing 1001 to 1500 g were 12.1/1000 days and 6.4/1000 days, respectively.7 The NNIS data help individual NICUs determine where their CVC-BSI rate is relative to other NICUs. Values at the extremes of the NNIS data may indicate problems with effective infection control (outlying above 90 percentile) or underreporting of CVC-BSI events (outlying below the 10 percentile).6 Individual NICUs are encouraged to monitor and compare their CVC-BSI rates with NNIS data, which are updated annually and usually published in December.6 Coexistence of endocarditis, osteomyelitis, or pyogenic arthritis should be considered whenever BSIs persist in neonates. Although Staphylococcus aureus is the most common cause of both endocarditis8–10 and osteomyelitis11 in neonates, these complications are uncommon, with one series reporting a prevalence of 0.07% for bacterial endocarditis in the NICU.10

Ventilator-Associated Pneumonia According to the NNIS, ventilator-associated pneumonia (VAP) is the second most common nosocomial infection in neonates. Unfortunately, even the 2004 NNIS report considered its VAP data as provisional6 due to the vagaries in the definition of VAP when applied to the neonatal population. Diagnosis of VAP in neonates is more difficult because noninfectious conditions such as respiratory distress syndrome and bronchopulmonary dysplasia are common and frequently cause radiologic abnormalities. Consequently, the data concerning the incidence, risk factors, microbiology, and outcomes of VAP in critically ill newborns are limited. A few investigators have attempted to establish reproducible criteria for VAP specific to the neonatal population. Cordero et al.12 showed that isolated purulent tracheal aspirates with positive tracheal cultures in mechanically ventilated neonates, in the absence of

worsening clinical or radiologic findings, are more consistent with clinically insignificant tracheal colonization than with VAP. Apisarnthanarak et al.13 performed a prospective cohort study addressing the risk factors, the microbiology, and outcomes of VAP in neonates. Their definition of VAP required new and persistent radiologic evidence of focal infiltrates > 48 hours after initiating mechanical ventilation and that the neonate received antibiotics for > 7 days to treat VAP. By this definition, the prevalence of VAP was 28% (19 of 67) in mechanically ventilated VLBW infants and the VAP rate was 6.5 per 1000 ventilator-days,13 which places their results at the 75th to 90th percentile of the NNIS VAP rates. Gram-negative bacteria were isolated from tracheal aspirates in 94% of VAP episodes and most cases were polymicrobial. VAP developed in neonates on a median of day 30 and the risk of VAP increased by 11% for every additional week an infant was mechanically ventilated. VAP was strongly associated with mortality in neonates who required NICU care > 30 days.13 A large multicenter study is needed to validate this definition of VAP and their findings.

Late-Onset Meningitis Until recently, there were few surveillance data on the incidence of late-onset meningitis in the NICU. Consequently, considerable variability has existed in clinical practice concerning lumbar puncture in neonates with suspected late-onset sepsis. Stoll et al.14 prospectively studied late-onset meningitis in 9641 VLBW infants who survived > 3 days, finding that it occurred in 134 infants. This represented 1.4% of all infants and 5% of those who had a lumbar puncture performed. Compared with nonseptic infants, VLBW infants with meningitis were more likely to have seizures (25% versus 2%), and were more likely to die (23% versus 2%).14 One-third (45 of 134) of the infants with meningitis had simultaneous negative blood cultures. Because many neonatologists do not perform lumbar punctures when late-onset sepsis is suspected, it is likely that late-onset meningitis has been underdiagnosed in VLBW infants. Because meningitis can alter longterm prognosis and duration of antibiotic therapy, all VLBW infants with suspected late-onset sepsis should have a lumbar puncture as part of the initial diagnostic evaluation, unless they are too critically ill to tolerate the procedure. In the latter case, it should be performed when clinical stabilization is achieved.

Urinary Tract Infection Urinary tract infection (UTI) is the most common nosocomial infection in adults.6 The high rate of UTI in hospitalized adults is associated with frequent use of indwelling urinary catheters, which

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Nosocomial Infections in the Neonate

are seldom used in VLBW infants. But, as with lumbar puncture, there is considerable practice variability in performing urine culture and analysis by either suprapubic bladder aspiration or urethral catheterization when late-onset sepsis is suspected.15 Urine specimens obtained by bag collection from infants have notoriously high rates of contamination – up to 63%16 – and are not recommended. Clinicians have tended to avoid suprapubic bladder aspiration in neonates because of the risk of serious, albeit rare, complications like bowel perforation.17 Fortunately, sterile urethral catheterization can easily be performed by experienced nurses, even in ELBW infants, and has a significantly higher rate of success in obtaining urine than suprapubic bladder aspiration – 100% versus 46% in one report.18 The prevalence of late-onset UTI in NICUs is uncertain: there has not been a prospective study of UTI, as there has been for late-onset sepsis1 and meningitis.14 The reported prevalence of UTI in premature infants ranges from 4% to 25%, but these reports are from the 1960s, and consequently not based on the typical population of infants in today’s NICUs. A recent retrospective study has reported an 8% prevalence of late-onset UTI in 762 VLBW infants in one NICU over an 11-year period.19 UTI was more common in ELBW infants (12%) than in infants with birthweight 1001 to 1500 g (6%). These findings suggest that UTI may be the second most common cause of nosocomial infections in the NICU, but a large multicenter prospective study is needed. Tamim et al.15 examined paired blood and urine cultures in a group of 189 VLBW infants suspected of having late-onset sepsis. UTIs were detected in 25%. Among the VLBW infants with UTIs, 62% (30 of 48) had negative blood cultures. Phillips & Karlowicz20 reported a case series of 60 UTIs in an NICU, primarily documented through specimens obtained by urethral catheterization when late-onset sepsis was suspected. Simultaneous BSIs with the same pathogen were present in 52% of cases of Candida UTI and 8% of cases of bacterial UTI. Since many neonatologists do not obtain a urine culture when late-onset sepsis is suspected, it is likely that UTIs are underdiagnosed. As most VLBW infants with UTI do not have BSIs, we recommend obtaining urine for culture, either by sterile urethral catheterization or by suprapubic aspiration, whenever late-onset sepsis is suspected in neonates.

CHAPTER

96

545

and coagulase-negative staphylococci (CONS) were present in 50% of FIP cases versus 14% of NEC cases. Most importantly, results from peritoneal fluid cultures resulted in changes in antimicrobial therapy in 40% (46 of 116) of cases.21 These findings suggest that a peritoneal fluid culture should be obtained in all neonates with intestinal perforation, regardless of cause, since it helps direct the choice of the most effective antimicrobial treatment.

Other Infections Although conjunctivitis is common in neonates, there are few studies addressing its occurrence in the NICU. Diagnosis can be complicated because conjunctival colonization, especially with CONS, is common among infants in the NICU.22 Haas et al.23 reported the results of a prospective study of conjunctivitis in the NICU and found a prevalence of 5%. Most neonatal skin infections are caused by Staphylococcus aureus. Clinical manifestations include impetigo, cellulitis, soft-tissue abscesses, and toxin-mediated diseases such as staphylococcal scaledskin syndrome and toxic shock syndrome.24 Pseudomonas aeruginosa can cause ecthyma gangrenosum lesions in the premature infant,25 and Enterobacteriaceae can cause purpura fulminans.26 Zygomycetes can cause progressive necrotizing skin lesions in neonates.27 Vesicular lesions in neonates are usually associated with herpes simplex and enteroviral infections.

DISTRIBUTION OF PATHOGENS CAUSING LATE-ONSET SEPSIS AND CASE-FATALITY RATES Multiple organisms cause late-onset sepsis in the NICU. Grampositive organisms are predominant (57% to 70% of cases), but gramnegative organisms (19% to 25% of cases) and fungi (12% to 18% of cases) also cause disease.1,28 It is noteworthy that many studies have the same organisms causing most of the episodes: CONS, Candida species, S. aureus, and Enterobacteriaceae (Table 96-2).

Usual Pathogens Intestinal Perforation and Peritonitis Peritonitis in NICUs is associated with intestinal perforation. Coates et al.21 reported striking differences in the distribution of pathogens associated with peritonitis in 36 infants with focal intestinal perforation (FIP) compared with 80 infants with necrotizing enterocolitis (NEC). Enterobacteriaceae were present in 75% of NEC cases compared with 25% of FIP cases. In contrast, Candida species were found in 44% of FIP cases compared with 15% of NEC cases,

Frequency of pathogens causing late-onset sepsis in a NICU is a consideration when empiric antibiotics are selected; it should not be the only consideration. All pathogens do not have equal likelihood of causing severe complications and death; clinicians should be concerned especially about late-onset sepsis in which infants die in less than 48 hours of onset of illness (fulminant sepsis), often before pathogens are identified or their antibiotic susceptibilities known. Karlowicz et al.28 reported that, although gram-negative organisms caused only 25% of

TABLE 96-2. Pathogens Commonly Causing Late-Onset Sepsis in the Neonatal Intensive Care Unit (NICU) Pathogen

Relative Frequency of Isolation

Comment

CONS

+++

Most common cause of CVC-BSI

Staphylococcus aureus

++

Highest rate of focal complications; MRSA is a problem in some NICUs

Fungi

++

Candida albicans and Candida parapsilosis are the most common species

GNR

++

GNR are most common cause of fulminant sepsis; Klebsiella species is the most common GNR

Pseudomonas aeruginosa

+

GNR with highest case-fatality rate

Enterococcus species

+

Increased in importance as a nosocomial pathogen since the 1990s

Group B streptococci

+

Rate of late-onset cases unchanged, in contrast to dramatic decrease in early-onset cases with intrapartum antibiotics

CONS, coagulase-negative staphylococci; CVC-BSI, central venous catheter-related bloodstream infection; GNR, gram-negative rods; MRSA, methicillin-resistant Staphylococcus aureus. +++, most frequent; ++, common; +, occasional

546

SECTION

N Infections of the Fetus and Newborn

sepsis cases in their series, these agents caused 69% of fulminant lateonset sepsis. Of gram-negative bacilli, Pseudomonas aeruginosa was the most prominent pathogen (42% of fulminant sepsis cases) and overall had a case-fatality rate of 56% – in contrast to a case-fatality rate of < 1% for CONS.28 Similar findings have been reported by others.1,29 VLBW infants infected with gram-negative organisms or fungi had the greatest risk of death, infants infected with gram-positive organisms were not more likely to die than infants who were not infected, and infants infected with gram-negative organisms were more likely to succumb to acute mortality, with Pseudomonas aeruginosa having the highest fulminant case-fatality rate. CONS are the most common pathogens causing late-onset sepsis in neonates, with a prevalence between 35%28 and 48%1 of cases. Identification of definite CONS sepsis is problematic because pseudobacteremia is common.30 Some of the reported variation in the prevalence of CONS infection depends upon how it was defined. Karlowicz et al.28 required at least two positive blood cultures. Stoll et al.1 distinguished between: (1) definite CONS infection, defined as either two positive blood cultures drawn within 2 days of each other or one positive blood culture and an elevated C-reactive protein within 2 days of blood culture; (2) possible infection, defined as one positive blood culture in a patient treated with vancomycin or a semisynthetic antistaphylococcal antibiotic for > 5 days; and (3) probable contaminant, defined as one positive blood culture without an elevated C-reactive protein or antibiotic therapy as described above. Both definite CONS infections and possible CONS infections were combined by Stoll et al.1 to make CONS responsible for 48% of cases of late-onset sepsis. If the 353 possible CONS cases are excluded, then the prevalence of CONS sepsis in their series was 29% (276/960),1 similar to the rate of 35% reported by Karlowicz et al.,28 which still makes CONS the single most common pathogen causing late-onset sepsis in neonates.

Emerging Pathogens Frequent antibiotic use in the NICU results in heavy antibiotic pressure on the organisms within this environment, and often, emergence of antibiotic-resistant strains. The placement of NICUs within larger healthcare facilities provides an additional opportunity for the introduction of antibiotic-resistant organisms. Antibiotic use increases the rates of infant colonization with antibiotic-resistant organisms.31 Gram-positive bacteria, including methicillin-resistant Staphylococcus aureus – both hospital- and community-acquired32–34 – and vancomycin-resistant Enterococcus faecium35 have been clearly identified as serious problems in some NICUs. Both gram-negative enteric organisms (extended-spectrum b-lactamase-carrying Escherichia coli and Klebsiella spp.,36 AmpC b-lactamase-carrying Enterobacter spp.,37 multidrug-resistant Serratia marcescens38) and nonenteric organisms (Pseudomonas aeruginosa,39 Burkholderia cepacia40) have emerged in NICU environments. In many instances, reservoirs containing the organism are present within the healthcare environment; patients are exposed either through the use of contaminated medical equipment or via the hands of their caretakers. The former often results from breakdowns in the cleaning procedures used in the NICU or hospital environment35,37,38,40,41 and the latter from ineffective use of hand hygiene by healthcare workers.42 Control of NICU outbreaks of antibiotic-resistant organisms frequently requires vigorous application of infection control procedures (surveillance cultures, patient and staff cohorting) and active education about the factors that predispose to infection. Molecular fingerprinting of organisms has been useful for characterizing and controlling some outbreaks.31,39–41

Nosocomial Viral Infections Nosocomial infections caused by viruses are infrequent in the NICU, with an incidence < 1%,43 but because of their propensity to spread

from patient to patient, they can cause significant clinical problems. Respiratory syncytial virus,44,45 influenza virus,46 enteroviruses,47,48 rotavirus,49 adenovirus,50 and coronavirus51 have been described in NICU outbreaks, sometimes concurrently.52 Attack rates can be as high as 33%.47,49,51 Patients can be asymptomatic or have disease that is lethal,50 and the attributable costs can be high.45 Respiratory syncytial virus infections can present as cough, congestion, apnea, increasing oxygen requirement, or respiratory failure.44,45 Adenovirus can have the same clinical presentation, in addition to causing epidemic conjunctivitis.50 Of note, ophthalmologic procedures can contribute to adenovirus spread.50 Coronavirus infection can be associated with respiratory decompensation or abdominal distention and fever.51 Enteroviruses can be associated with a clinical picture suggestive of NEC,48 overwhelming septicemia, rash, or aseptic meningitis.47 Rotavirus infection is associated with diarrhea that is frequent and watery in term infants, whereas in preterm infants it is more frequently bloody and associated with abdominal distention and intestinal dilatation.49

CLINICAL MANIFESTATIONS The clinical features of nosocomial sepsis in neonates are nonspecific. There are no clinical features that are characteristic exclusively of infections caused by nosocomial pathogens. Fanaroff et al.53 reviewed data from the Neonatal Research Network Intravenous Immunoglobulin trial in order to identify the predominant clinical features of late-onset sepsis in neonates. The most common clinical features were increased apnea/bradycardia (55%), increased gastrointestinal problems (46%) (such as feeding intolerance, abdominal distention, or bloody stools), increased respiratory support (29%), and lethargy/ hypotonia (23%). The predominant laboratory indicators were an abnormal white blood cell count (46%), (e.g., leukocytosis, increased immature white blood cells, or neutropenia), unexplained metabolic acidosis (11%), and hyperglycemia (10%). Unfortunately, the predictive value was low for any of these clinical or laboratory features, with the best positive predictive value only 31%, for hypotension.53 None of the clinical features was pathogen-specific. In addition, sepsis caused by CONS or Candida species can be more indolent in clinical presentation.54 Abnormal heart rate characteristics (reduced variability and transient decelerations) occur early in the course of neonatal sepsis.55 Technology has been developed to monitor changes in heart rate characteristics continuously and noninvasively in an attempt to provide a clinical tool with the potential for alerting medical personnel in advance of overt clinical illness from late-onset sepsis.56 The most common signs of CVC-BSI in neonates are fever (49%) and pulmonary dysfunction (30%).57 Erythema or purulent discharge at the insertion site was present in only 20% of cases of CVC-BSI in neonates.

LABORATORY DIAGNOSIS Whenever a nosocomial infection is suspected enough to begin antibiotics, the pretreatment diagnostic evaluation should include blood cultures, cerebrospinal fluid (CSF) culture, and urine culture. Late-onset meningitis and UTIs commonly occur in neonates with negative blood cultures14,15 and will be missed if CSF or urine cultures are not obtained. The Pediatric Prevention Network recommends routinely obtaining two blood cultures from neonates with suspected sepsis because isolation of CONS from a single blood culture, especially without an elevated C-reactive protein within 2 days,1 should generally be interpreted as a contaminant in order to reduce vancomycin use in neonates.58 A definitive diagnosis of nosocomial infection due to bacterial or fungal species requires isolation of the organism from blood or another normally sterile body site or fluid. Exceptions are fungi such as Aspergillus and those that cause zygomycosis, as they can cause

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Nosocomial Infections in the Neonate

disseminated multiorgan infection but are rarely isolated from blood.27 Skin biopsy is the sole means of establishing the diagnosis of zygomycosis in premature infants with progressive necrotizing skin lesions.27 When viral infections are suspected, clinicians can make a presumptive diagnosis by rapid diagnostic testing (e.g., a positive direct fluorescent antibody test or enzyme immunoassay (EIA) for influenza A, respiratory syncytial virus, adenovirus, or EIA for rotavirus) and a definitive diagnosis by identification of a viral pathogen from nasal wash, tracheal secretions, bronchoalveolar lavage fluid, or stool, as appropriate. Other laboratory studies provide clinicians with an index of the severity of infection, help guide therapy, and monitor response to treatment. As with infections in older children, laboratory evidence of organ dysfunction(s), particularly renal failure, consumption coagulopathy, neutropenia, thrombocytopenia, or hypoperfusion, usually indicate severe illness and a more guarded prognosis. Attempts to identify dependable serum markers for diagnosis, severity, or prognosis have been variably successful. The utilities of C-reactive protein,59 various proinflammatory cytokines,60 and/or procalcitonin61 levels as useful markers for diagnosis and severity of neonatal sepsis continue to be debated.

TREATMENT Empiric Therapy Empiric antimicrobial therapy for suspected nosocomial infections without a clinical focus in neonates should be guided by knowledge of the distribution and case-fatality rates of pathogens and the susceptibility patterns of likely pathogens in a particular NICU. Empiric antibiotic regimen should effectively treat gram-negative pathogens, particularly Pseudomonas aeruginosa, because these organisms account for the majority of cases of fulminant late-onset sepsis in neonates.28 An aminoglycoside should be used for empiric treatment of possible gram-negative sepsis, the choice of which is determined by the antimicrobial susceptibility patterns of isolates from the NICU. During an outbreak of gentamicin-resistant gramnegative septicemia, amikacin may be the preferred aminoglycoside. Third-generation cephalosporins are not recommended for routine empiric therapy in neonates (unless knowledge of the patient’s flora or the NICU pattern of infections specifically dictates) because: (1) they do not effectively treat most Pseudomonas aeruginosa and some Enterobacteriaceae; (2) routine use in NICUs has been associated with emergence of cephalosporin-resistant gram-negative bacilli62,63; and (3) they have been associated with increased risk of candidemia in VLBW neonates.54 Ampicillin may be considered for empiric treatment of possible gram-positive septicemia, especially if Enterococcus and Streptococcus agalactiae are common pathogens causing late-onset sepsis in the NICU. In the past, a semisynthetic penicillin, such as oxacillin, was considered for empiric treatment of possible gram-positive late-onset septicemia in neonates, because methicillin-susceptible Staphylococcus aureus (MSSA) was prevalent.1,28 The recent emergence of community-acquired methicillin-resistant S. aureus (CA-MRSA) as a neonatal pathogen may require modification of this approach in the future.33,34 Broad empiric usage of vancomycin, as recommended by some,64 because CONS sepsis is common and MRSA sepsis is possible in VLBW infants, creates additional problems. Stoll et al.1 found it alarming that 44% of all VLBW infants in the NICHD Neonatal Research Network were treated with vancomycin whether or not they had CONS sepsis. The Hospital Infection Control and Practices Advisory Committee of the CDC recommends avoiding empiric vancomycin therapy in patients with suspected sepsis to prevent the emergence and spread of vancomycin-resistant enterococci.65 Karlowicz et al.28 showed that avoidance of vancomycin as an empiric antibiotic had no impact on the very low rate of fulminant sepsis for

CHAPTER

96

547

CONS sepsis in neonates and that the practice of starting vancomycin only after CONS were identified in blood cultures did not prolong the duration of sepsis caused by CONS. Sanchez et al.66 reported their experience with restricted use of vancomycin in neonates, even though CONS was the predominant pathogen, and found no change in the number of deaths as a result of CONS sepsis despite a 53% reduction in vancomycin use. Consequently, the Pediatric Prevention Network has recommended restricting the use of vancomycin for empiric therapy of neonates with suspected late-onset sepsis.58 Because, in the past, MRSA has been rare in most NICUs,58 it is recommended that empiric use of vancomycin be reserved for neonates in NICUs with persisting MRSA colonization and infections that do not resolve with cohorting and standard infection control measures. How CA-MRSA will modify this recommendation remains to be seen.33,34 Some investigators have suggested empiric antifungal therapy for VLBW infants at high risk of candidemia,67 hoping to reduce morbidity and mortality in an approach analogous to the algorithm for persistent fever and neutropenia in patients with cancer.68 A predictive model for empiric antifungal therapy in neonates has been developed,67 but it has not been validated in prospective trials. Consequently, empiric antifungal therapy cannot be recommended in neonates with suspected late-onset sepsis. The suggested duration of therapy for nosocomial infections by anatomic site is summarized in Table 96-3. The duration of treatment for individual patients should be determined by virulence of the pathogen, time it takes for follow-up cultures to become negative, rapidity of clinical response, removal or retention of CVC, and adequate drainage of purulent foci, if present.

Adjunctive Therapy Several adjunctive therapies have been investigated in late-onset sepsis, including immunoglobulin intravenous (IGIV), hematopoietic growth factors (granulocyte colony-stimulating factor (G-CSF) and granulocyte–macrophage colony-stimulating factor (GM-CSF)), and granulocyte transfusions. IGIV,69 G-CSF and GM-CSF70 have been evaluated by the Cochrane Database of Systematic Reviews, with the conclusion that there is currently insufficient evidence to support routine use in the treatment of neonates with sepsis, even though mortality was decreased. Zipursky71 advised against the use of G-CSF or GM-CSF in newborns because of considerable potential for harm due to formation of antibodies against these factors.72 A review of the data from studies of granulocyte transfusions in septic neonates demonstrated improved outcome in the situation of neutropenia depletion of the marrow storage pool, but associated morbidities, including fluid overload, worsening hypoxia and respiratory distress from leukocyte sequestration in the lung, graft-versus-host disease,

TABLE 96-3. Suggested Duration of Therapy for Selected Nosocomial Infections Site or Manifestation of Infection

Duration of Therapy (days)

BSI Meningitis CVC-BSI without removal of CVC Osteomyelitis/septic arthritis VAP UTI Endocarditis Candidemia, catheter removed, rapidly resolving Fungemia, disseminated Skin or subcutaneous lesion

10–14 14–21 14a 4–6 weeks 10–14 10–14 4–6 weeks 10–14 ~4 weeks 7–10

BSI, bloodstream infection; CVC-BSI, central venous catheter-related bloodstream infection; CVC, central venous catheter; UTI, urinary tract infection; VAP, ventilator-associated pneumonia. a After first negative blood culture.

548

SECTION

N Infections of the Fetus and Newborn

and risk of transmission of viral infections.73 Careful assessment of the risks versus the beneÀts of leukocyte administration is required.

MANAGEMENT OF CENTRAL VENOUS CATHETER-RELATED BLOODSTREAM INFECTIONS Catheters are intravascular foreign bodies; removal is the optimal management when a BSI occurs. Nevertheless, the vital importance of CVCs in critically ill neonates must be acknowledged, especially since successful in situ treatment of CVC-BSI has become more common.74,75 It is clear that BSI can occur with or without CVC involvement. Differentiation of these two conditions is difÀcult, however. No data derived from large randomized trials are available as guidelines for the management of CVC-BSI. However, several large observational studies have compared outcomes of late-onset sepsis in neonates with CVCs treated with and without CVC removal. Data suggest that management strategies can be different, depending on the pathogen and clinical condition of the infant. If treatment with the CVC in situ is attempted, antimicrobial agents should be administered through the infected catheter. The algorithm shown in Figure 96-1 may help clinicians manage CVC-BSI in neonates until better evidence becomes available from randomized trials.

Candida Species Until recently, the attitude towards removal of CVCs at time of diagnosis of Candida sepsis varied among pediatric specialists. A 1998 survey reported that only 35% of neonatologists and 53% of infectious disease specialists would immediately remove a CVC from a neonate with the Àrst positive blood culture for Candida species.76 A single-center retrospective study of 104 cases reported that failure to remove CVCs as soon as Candida sepsis was detected in neonates was associated with signiÀcantly increased mortality in C. albicans sepsis (case-fatality risk increase of 39%, number needed to harm of 2.6)

and signiÀcantly prolonged duration of Candida sepsis regardless of Candida species (median of 6 days versus 3 days).77 These Àndings were conÀrmed in a retrospective multicenter study of ELBW infants with candidiasis.78 Consequently, the Infectious Diseases Society of America guidelines for treatment of candidiasis79 strongly recommend that CVCs be removed as soon as Candida sepsis is detected in neonates, if feasible. Unfortunately, in some neonates, CVCs cannot be removed because of severe generalized skin breakdown or unstable critical condition, but these conditions should be present in less than 15% of cases.80

Coagulase-Negative Staphylococci In contrast to cases of Candida sepsis, CVC-BSI in neonates associated with bacterial pathogens have a reasonable chance of success with in situ treatment. CONS bacteremia is the most common cause of CVC-BSI in neonates.4 It has been difÀcult to interpret clinical studies of CONS CVC-BSI in neonates because many studies required only a single positive blood culture to diagnose CONS bacteremia,81–83 allowing inclusion of many cases of pseudobacteremia.30 A series of 119 cases84 in which two positive blood cultures yielding CONS within 3 days of each other were required for entry consideration (deÀnition of CONS bacteremia consistent with CDC guidelines)65 concluded that in situ treatment could be successful in many neonates with CONS CVC-BSI, but observed it was unclear how long clinicians should wait before abandoning sterilizing attempts and removing the CVC.85 Karlowicz et al. reported that in situ treatment with vancomycin was successful in 46% of cases with CONS CVC-BSI.84 In their report, none of 19 patients with CONS bacteremia > 4 days’ duration after institution of antibiotic therapy showed resolution of bacteremia until CVC were removed. In contrast, 79% of cases with CONS bacteremia for 1 or 2 days were successfully treated without CVC removal, but successful treatment decreased to 44% when bacteremia persisted for 3 to 4 days.84 Therefore, when CONS bacteremia persists in neonates who have CVCs that are vital

Figure 96-1. Suggested management of central venous catheter (CVC)-related bloodstream infections in neonates. CONS, coagulase-negative staphylococci; MSSA, methicillin-susceptible Staphylococcus aureus. aAfter commencement of appropriate antibiotic therapy.

Central venous catheter-related bloodstream infection

Ventricular reservoir or ventriculo-peritoneal shunt? Yes

No

• Remove CVC • Treat with systemic antibiotics

No

Isolate identification

Clinically stable? Yes

Candida spp.

CONS

Enterobacteriaceae

Polymicrobial

S. aureus

• Remove CVC • Treat with amphotericin B

• Treat with vancomycin • Repeat blood culture daily • Consider CVC removal if blood culture (+) × 3–4 days • Remove CVC if blood culture (+) >4 days

• Treat with systemic antibiotics • Repeat blood culture daily • Remove CVC if platelet count 1 day

• Remove CVC • Treat with systemic antibiotics

• Treat with systemic anti-staphylococcal antibiotics • Remove CVC if blood culture (+) >1 day and vigorously search for metastatic disease

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Nosocomial Infections in the Neonate

to clinical care, we recommend that clinicians give antibiotic treatment for 2 days, perhaps as long as 3 to 4 days in special circumstances, but never beyond 4 days of persistent bacteremia, before removing CVCs.

Enterobacteriaceae Enterobacteriaceae are a common cause of late-onset sepsis in neonates,86 but, until recently, data were limited concerning Enterobacteriaceae CVC-BSI. In a report of 53 cases of Enterobacteriaceae CVC-BSI in neonates, Enterobacteriaceae bacteremia resolved with use of gentamicin or tobramycin without removal of CVCs in 45% of cases in which it was attempted.87 In contrast to successful in situ treatment despite several days of CONS bacteremia, successful treatment of Enterobacteriaceae bacteremia of greater than 1 day’s duration was unlikely without removal of CVCs, decreasing from 85% to 24% successful in situ treatment. Attempting to treat Enterobacteriaceae bacteremia with CVC in situ was not associated with increased mortality, increased morbidity, or increased recurrence. In addition, Enterobacteriaceae cases with severe thrombocytopenia (platelet count < 50,000/mm3) on the first day of bacteremia did not resolve until CVCs were removed in 82% of cases, compared with 32% of cases without severe thrombocytopenia.87 Therefore, we recommend that CVCs be removed in cases associated with severe thrombocytopenia or if Enterobacteriaceae bacteremia persists > 1 day after commencing appropriate antibiotic treatment.

Polymicrobial Agents Polymicrobial sepsis in neonates now accounts for about 14% of cases of late-onset sepsis.88 In a preliminary report of 70 cases of polymicrobial CVC-BSI in neonates,89 successful in situ treatment of polymicrobial sepsis was uncommon, occurring in only 20% of cases, even when one of the pathogens was CONS. The risk of case fatality was significantly increased by 20% when in situ treatment was attempted. On the basis of this experience, our recommendation is for removal of CVCs, as soon as possible, in cases of polymicrobial sepsis.

Staphylococcus aureus Finally, S. aureus is one of the most common and most serious causes of late-onset sepsis in the NICU.90 In adults, removal of CVC is advised in cases of S. aureus bacteremia, unless there is a compelling reason to conserve the catheter.74 There are few published reports concerning S. aureus CVC-BSI in neonates. In a case series of 11 infants, many developed end-organ damage, and only 1 case was successfully treated with CVC in situ.83 Another case series concerning invasive staphylococcal disease in neonates included 25 cases of S. aureus bacteremia, and found no difference in occurrence of complications or outcome in cases treated with or without CVC removal.64 Recently, there was a preliminary report that MSSA CVCBSIs were successfully treated with CVCs in situ in 13 (62%) of 21 cases.91 Most cases that were successfully treated with CVC in situ showed resolution of the S. aureus bacteremia within 24 hours of starting a penicillinase-resistant penicillin. Focal complications occurred in 34% of 47 cases of MSSA bacteremia, but, unlike a previous report,83 at least 75% of focal complications were already present when the first blood culture was obtained.91 The duration of bacteremia was significantly longer in cases with focal complications compared with those without complications – a median of 6 days versus 1 day, respectively. Focal complications, like soft-tissue abscesses, endocarditis, and osteomyelitis, may be more important risk factors for persistent S. aureus bacteremia than retention of CVCs. Similar to that for Enterobacteriaceae CVC-BSI, and we advise a cautious approach and recommend that CVCs be removed immediately if MSSA CVC-BSI persists > 1 day with appropriate antibiotic treatment.

CHAPTER

96

549

Other Considerations Another clinical feature to consider when deciding whether to treat CVC-BSI without CVC removal is the presence of other foreign bodies. We reported an infant with a ventricular reservoir who developed CONS meningitis on day 7 of 9 days of persistent CONS bacteremia despite vancomycin therapy with the CVC in situ.84 The long-term consequences of meningitis and shunt revision are so potentially devastating that we consider the presence of a ventricular reservoir or ventriculoperitoneal shunt to be a contraindication to attempting to treat CVC-BSI with CVC in situ.

MANAGEMENT OF PERSISTENT BLOODSTREAM INFECTIONS The likelihood of adverse outcomes, such as focal complications, increases when BSIs persist in neonates. Although it is uncertain whether focal complications are the cause or the consequence of persistent BSIs, it is imperative that clinicians obtain serial blood cultures to document resolution of BSIs and perform thorough diagnostic evaluations searching for focal complications if BSIs persist. In addition, when BSIs persist, clinicians must make management decisions concerning timing of CVC removal and changes in antimicrobial therapy. Removal of CVC in cases of persistent bacterial BSI, defined as > 1 day of appropriate antibiotic therapy for Enterobacteriaceae and MSSA and > 2 to 4 days for CONS, should be done promptly. Of note, two case series reported success treating persistent staphylococcal bacteremia in neonates with CVC in situ, without adverse consequences, by adding rifampin to standard antistaphylococcal antimicrobial therapy.92,93 Some pathogens, especially Candida species, may continue to grow from blood cultures even when CVCs are removed promptly, and the neonate is given antibiotic therapy to which the organism is susceptible. In one report of 96 cases, candidiasis lasted > 7 days in 30% of cases, even when CVCs were removed on the day the first positive culture report was received, and the infants were promptly given systemically therapeutic doses antifungal therapy.94 The risk of focal complications was significantly increased in cases with persistent invasive candidiasis compared with nonpersistent cases (48% versus 13%). The most common focal complications in neonates with persistent candidiasis are “fungus ball” uropathy (29%), renal infiltration (20%), abscess (19%), and endocarditis (9%).94 Since more than half of neonates with persistent candidiasis do not have focal complications, Chapman & Faix94 suggested that aggressive imaging for focal complications be reserved for cases in which blood cultures remain positive after several days of antifungal therapy, or if there are clinical signs suggesting focal complication. On the other hand, Noyola et al.95 documented focal complications in 23% of 86 neonates with candidemia, including some with only one positive blood culture, and they recommended renal, cardiac, and ophthalmologic diagnostic evaluations in all neonates with candidemia because the presence of focal complications may affect the duration of therapy and outcome. The prevalence of persistent bacteremia, defined as recovery of the same pathogen > 24 hours after initiation of antibiotic therapy to which the organism is susceptible, was reported to be 22% in a series of 335 cases of bacteremia in one NICU.93 In this case series, the frequent decision to treat bacterial BSI with CVC in situ contributed to the high prevalence of persistent cases. The prevalence of focal suppurative complications (osteomyelitis, septic arthritis, abscess, infected thrombus, or endocarditis) was significantly increased in infants with persistent non-CONS BSI compared with persistent CONS BSI (28% versus 3%).93 S. aureus caused 50% of persistent non-CONS BSIs and 67% of the cases with focal complications. The duration of persistence of bacteremia significantly correlated with the presence of focal complications in non-CONS cases. Consequently, Chapman & Faix93 recommended that all neonates with persistent bacteremia undergo extensive evaluation for focal complications,

550

SECTION

N Infections of the Fetus and Newborn

especially looking for endocarditis, osteomyelitis, and soft-tissue abscesses. This evaluation is especially important in cases of persistent BSI caused by S. aureus or Enterobacteriaceae, because the bacteremia will not resolve until the soft-tissue abscesses or bone or joint infections are surgically drained, or the intravascular clot dissolves.

PREVENTION OF NOSOCOMIAL INFECTIONS The largest proportion of healthcare-associated infections in the NICU result from patient requirements for technological support and handson care by healthcare workers. The most commonly utilized devices in the NICU are mechanical ventilation and intravascular access. Risk factors for CVC-BSI have been extensively examined and include blood and blood component transfusion,96 blood drawing and manipulation (including disinfection) of the catheter hub,97 catheter hub colonization, exit site colonization, and duration of parenteral nutrition.98 Thus, approaches that minimize catheter manipulation, blood drawing, and blood product administration via the catheter prevent colonization of the skin entry site and prevent infections.99 Hand decontamination by healthcare workers is the most effective means of preventing healthcare-associated infection in the NICU, but is often overlooked or performed poorly in the NICU environment.100 A remote video camera study on random nursing shifts reported handwashing only 24% of the time before infant contact by healthcare workers in a regional referral NICU.101 Activities such as skin contact, respiratory care, and diaper changes are independently associated with increased hand contamination.42 Recently, the CDC Guideline for Hand Hygiene in Health Care Settings recommended that healthcare workers use alcohol-based hand rubs over antimicrobial soaps.102 Alcohol-based hand rubs have excellent antimicrobial spectrum against bacteria, fungi, and viruses. In addition, alcohol-based hand rubs have fast speed of action and are the least likely to cause dermatitis on the hands of healthcare workers.103 Use of alcohol-based hand rubs in a NICU has been associated with significantly improved hand hygiene by healthcare workers compared with use of antimicrobial soap.104 Unfortunately, compliance was still unacceptably low (23%). Support of respiratory and enteral function figure prominently in the care of the premature infant. Systemic corticosteroid and H2blocking agents have been used to prevent chronic lung disease and enhance gastrointestinal function, respectively. Dexamethasone therapy in VLBW infants is associated with increased risk of lateonset sepsis.105 Use of H2-blocking agents in VLBW infants is associated with higher rates of NEC,106 BSIs,105 and candidemia.107 Avoiding the use of dexamethasone and H2-blocking agents should

reduce rates of late-onset sepsis. Establishing full enteral feeds with human milk is associated with lower risks of late-onset sepsis in ELBW infants.108 Prophylactic administration of IGIV has been associated with decrease in infection rates by 3% to 4% in premature infants without improving mortality rates109; it is not recommended. Several clinical trials have examined prophylactic administration of fluconazole for prevention of Candida colonization and BSI in ELBW infants, observing that daily or twice-weekly administration can decrease Candida BSI and attendant mortality.110,111 However, it is unclear whether prophylaxis significantly alters morbidity or how great is the risk of emergence of fluconazole resistance. Use of fluconazole prophylaxis to prevent Candida BSI in ELBW infants remains controversial, especially because the rates of Candida BSIs vary greatly among NICUs.112 In their editorial review of fluconazole prophylaxis studies in NICUs, Long & Stevenson concluded that fluconazole prophylaxis for ELBW infants should not be implemented in NICUs.113 They identified the need for a multicenter study to answer the questions of benefit and risk in ELBW infants of fluconazole prophylaxis which is adequately powered to investigate all BSIs, all-cause mortality, and weighing potential unintended consequences (such as prolongation of use of catheter, antibiotics, corticosteroids) and the long-term expectation of emergence of resistant pathogens. Rates of infection among regional NICUs vary by as much as fivefold.113 Lower rates are attributable to more restrictive use of CVCs and parenteral nutrition, but other, undefined practices also influence rates. Systematic investigation of clinical practices in NICUs with low rates of nosocomial infection may lead to strategies for prevention of infection that can be applied to other NICUs.114 Schelonka et al.115 reported that comprehensive infection control practices resulted in a sustained reduction in both bacterial and fungal nosocomial infection rates in a single regional NICU. The comprehensive infection control intervention115 included: (1) improving hand hygiene and reducing microbial contamination of patient care space; (2) limiting duration of umbilical catheters and developing a highly trained core of nurses to insert and maintain peripherally inserted CVCs; (3) increasing rates of breastfeeding and shortening the interval to full enteral feeding; (4) reducing the duration of empiric antibiotic therapy with negative culture results; and (5) educating and empowering the entire NICU staff to encourage maximal adherence to comprehensive infection control practices. A similar program using evidence-based infection control practices virtually eliminated CVCBSIs in an adult surgical intensive care unit.116 The best hope for sustained prevention of nosocomial infections in NICUs appears to be changing the NICU culture to one of collaborative continuous quality improvement and more pervasive use of human milk feeding.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

SECTION

O

Infections and Transplantation

CHAPTER

97

Infections in Solid-Organ Transplant Recipients Michael Green and Marian G. Michaels

Solid-organ transplantation is now accepted therapy for end-stage disease of the kidneys, liver, heart, and lungs. Intestinal and multivisceral transplantation are increasingly performed. Accordingly, an expanding number of immunosuppressed children are at risk for developing infection after transplantation.

PREDISPOSING FACTORS Factors predisposing to infection after organ transplantation can be divided into those that exist before transplant and those secondary to intraoperative and posttransplant events (Box 97-1).

Pretransplant Factors The organ undergoing transplantation is the most important determinant of the location of infection, especially during the first 3 postoperative months.1 For example, the chest, abdomen, and urinary tract are the most common sites of infection after thoracic, liver, and kidney transplantation, respectively. Explanations for these sitespecific infections include local ischemic injury and bleeding, as well as potential contamination.2

BOX 97-1. Predisposing Factors for Infections After Organ Transplantation in Children PRETRANSPLANT FACTORS Underlying disease(s), malnutrition Specific organ to be transplanted Age of patient Previous exposures to infectious agents Previous immunizations INTRAOPERATIVE FACTORS Duration of transplant surgery Exposure to blood products Technical problems Organisms transmitted in donor tissue POSTTRANSPLANT FACTORS Immunosuppression • Immediate posttransplant immunosuppression • Maintenance immunosuppression • Augmented therapy for rejection episodes Indwelling cannulas Nosocomial exposures Community exposures

The underlying illnesses that lead to organ failure may also be associated with an increased risk for developing infection after organ transplantation. For example, a history of cystic fibrosis predisposes to pseudomonal and fungal infections after lung transplantation. A history of palliative surgery before the transplant increases the technical difficulty of the transplant procedure, enhancing the risk of developing a posttransplant infection.3 The severity of disease at the time of transplantation is proportional to the risk of postoperative morbidity and mortality.4 Similarly, long-standing malnutrition predisposes children to infections before and after transplantation. Attempts to correct nutritional deficits with parenteral alimentation carry attendant risks of catheter-associated infection. Finally, mechanical ventilation while awaiting transplantation increases the risk of colonization and infection with nosocomial pathogens, many of which are resistant to multiple antimicrobial agents. Age is an important determinant of both susceptibility to certain pathogens and severity of infection after transplantation. Young children undergoing these procedures can experience moderate to severe disease with certain viral pathogens, such as respiratory syncytial virus (RSV) or parainfluenza virus, or bacterial pathogens, such as coagulase-negative staphylococci, compared with milder symptoms observed in older children or adult recipients. By contrast, other pathogens, such as Cryptococcus neoformans, are rarely found before young adulthood. Age is also associated with severity of infection with cytomegalovirus (CMV) and Epstein–Barr virus (EBV), because primary infections (which are more likely to occur in young children) in transplant recipients are more severe than reactivation infections.5,6 Younger age at the time of transplantation is also associated with a substantially increased rate of infection during the first few years after transplantation.7 Finally, young children who are not fully immunized remain susceptible to vaccine-preventable infections or receive vaccination after transplantation, at a time when their ability to mount an immune response may be hampered.8 Transplant recipients are at risk for acquiring pathogens from their donors who have active or latent infections at the time of organ harvesting. Examples include CMV,9–11 EBV, Toxoplasma and Histoplasma spp., West Nile virus (WNV), hepatitis B and C viruses, and human immunodeficiency virus (HIV). Screening to preclude use of a donor with WNV, hepatitis B surface antigenemia, or HIV infection is standard practice. The use of a hepatitis C-positive donor is controversial. Another donor-related concern is the presence of bacteria or fungi colonizing the respiratory tract of a lung donor; such organisms can cause infection in the postoperative period.12 Similarly, acute, unrecognized bacteremia or viremia at the time of organ harvest is an additional risk to the recipient.

Intraoperative Factors Operative factors unique to each solid-organ transplant procedure can predispose to infectious complications. For example, the type of biliary reconstruction used in liver transplantation influences the likelihood of developing an infectious complication.13 Surgical events during the operation also alter the risk of infection. Injury to the phrenic, vagal, or recurrent laryngeal nerves during surgery affect pulmonary toilet and predispose the patient to pneumonia after lung transplantation.14 Ischemic injury to the allograft during the transplant procedure reduces its viability and increases the risk of infection. Additional factors, 551

552

SECTION

O Infections and Transplantation

including prolonged operative time, contamination of the operative Àeld, and bleeding at or near surgical sites, also increase the risk of postoperative infections.

Posttransplant Factors

occur within the Àrst 180 days after transplantation. Risk periods can be divided into three intervals: (1) early (0 to 30 days after transplantation); (2) intermediate (30 to 180 days); and (3) late (more than 180 days). In addition, some infections can occur throughout the posttransplant course. These divisions provide a useful framework for the approach to a patient with fever after transplantation and serve as a guide to differential diagnosis (Table 97-1).

Immunosuppression is the major risk factor for infection following transplantation. Immunosuppressive regimens are continually evolving in an attempt to achieve more speciÀc control of rejection with the least impairment of immune function. However, all immunosuppressive regimens interfere with host defenses. Treatment of episodes of rejection exacerbates this risk. The use of antilymphocyte preparations, especially OKT3, is associated with an enhanced risk of infection.11,13 Technical problems at the time of transplant surgery are major risk factors for infectious complications. These include thrombosis of the hepatic artery predisposing to hepatic abscesses and bloodstream infection (BSI) after liver transplantation15; the presence of vesicoureteral reflux predisposing to graft pyelonephritis in renal transplant recipients14,16; and mediastinal bleeding requiring re-exploration predisposing to mediastinitis and BSI in thoracic transplant recipients.9 The prolonged use of indwelling cannulas is a major risk factor for infection after transplantation. The presence of a central venous catheter is a risk for BSI; the use of a urethral catheter predisposes to urinary tract infection; the use of a cannula in an obstructed biliary tract predisposes to cholangitis; and prolonged endotracheal intubation is associated with pneumonia. Nosocomial exposures constitute the Ànal group of posttransplant risk factors. All transplant recipients are at risk for developing infection with transfusion-associated pathogens. Children undergoing transplantation during the winter months are often exposed nosocomially to common viruses (e.g., RSV, rotavirus). The presence in the hospital of areas of heavy contamination with pathogenic fungi, such as Aspergillus spp., increases the risk of invasive fungal disease in these patients. Finally, nosocomial transmission of multiply resistant bacteria predisposes to infection with these pathogens.

The intermediate period (31 to 180 days after transplant) is the typical time of onset of infections transmitted from the donor (via either organ or blood product) or those associated with the recipient’s own reactivated viruses, as well as opportunistic infections. During this period, CMV infection peaks9–11 and EBV-associated posttransplant lymphoproliferative disorders (PTLD),6,18,19 Pneumocystis jiroveci (formerly P. carinii) pneumonia (PCP),20–22 and toxoplasmosis manifest.23 A review of autopsies found infections to be the most common cause of death during this period after pediatric lung or heart–lung transplantation; disseminated adenovirus and Aspergillus infection predominated, followed by CMV and EBV disease.24

TIMING OF INFECTIONS

Late Infections

The timing of speciÀc infections is generally predictable regardless of the type of organ transplanted. Most clinically important infections

Only limited data on late infectious complications (more than 180 days after transplant) have been published. In general, rates and severity of

Early Infections Early infections (0 to 30 days after transplant) are usually associated with the presence of pre-existing conditions or complications of surgery. Bacteria or yeast are the most frequent pathogens recovered during this period.9,17 Fifty percent or more of all bacterial infections that develop after transplantation occur during the early posttransplant period.9,17 SuperÀcial or deep surgical wound sites are the most common infections during this period. Technical difÀculties, particularly those resulting in anastomotic stenoses, are important risk factors for the development of invasive infection in the Àrst month after most types of organ transplantation.

Intermediate Period

TABLE 97-1. Timing of Infectious Complications Following Transplantationa Early Period (0–1 months)

Middle Period (1–6 months)

Late Period (> 6 months)

BACTERIAL INFECTIONS

VIRAL INFECTIONS

VIRAL INFECTIONS

Gram-negative enteric bacilli Small bowel, liver, neonatal heart Pseudomonas/Burkholderia spp. Cystic Àbrosis: lung Gram-positiveorganisms All transplant types

Cytomegalovirus All transplant types Seronegative recipient of seropositive donor Epstein–Barr virus All transplant types Seronegative recipient Small-bowel highest-risk group Varicella-zoster virus All transplant types

Epstein–Barr virus All transplant types, but less than middle period Varicella-zoster virus All transplant types Community-acquired viral infections All transplant types

FUNGAL INFECTIONS

All transplant types VIRAL INFECTIONS

OPPORTUNISTIC INFECTIONS

Herpes simplex virus All transplant types Nosocomial respiratory viruses All transplant types

Pneumocystis jirovecii All transplant types Toxoplasma gondii Seronegative recipient of a heart from a seropositive donor

BACTERIAL INFECTIONS

Pseudomonas/Burkholderia spp. Cystic Àbrosis: lung Lung recipients with chronic rejection Gram-negative bacillary bacteremia Small bowel FUNGAL INFECTIONS

Aspergillus spp. Lung transplants with chronic rejection

BACTERIAL INFECTIONS

Pseudomonas/Burkholderia spp. pneumonia Cystic Àbrosis: lung Gram-negative enteric bacilli Small bowel a

Listed in decreasing order of relative importance.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Infections in Solid-Organ Transplant Recipients

infection in children more than 6 months after transplantation are similar to those observed in otherwise healthy children.7 This is most likely explained by the fact that the majority of pediatric transplant recipients are maintained on only low levels of immunosuppressant agents. However, chronic or recurrent infections do occur in a subset of transplant recipients who have an uncorrected anatomic or functional abnormality (e.g., vesicoureteral reflux, biliary stricture). An important exception to this rule is the frequent finding of disease caused by Pseudomonas, Stenotrophomonas, and Aspergillus in lung transplant recipients with bronchiolitis obliterans.12,24 Finally, lymphoproliferative disease continues to manifest in the late period.25

Infections Occurring Throughout the Postoperative Course While iatrogenic factors are associated with the development of bacterial or fungal infections at any time after transplantation, the highest risk occurs in the early posttransplant period. Central venous lines, urethral catheters, biliary tract, peritoneal or pleural catheters, and nasotracheal or endotracheal tubes are maintained for a variable period; increased risk of infection persists for the entire time of cannulation. The risk of nosocomial acquisition of community viruses, such as RSV, rotavirus, influenza, and parainfluenza virus, is seasonal. Diagnostic studies should be modified according to these considerations.

CHAPTER

97

553

account for more than one-half of episodes. Bacterial infections involving the abdomen or surgical wound are common in most series. Infectious complications of the transplanted liver also occur. The most important complication is hepatic abscess, associated with hepatic artery or portal vein thrombosis, often accompanied by persistent BSI. However, the introduction of frequent surveillance Doppler studies early after transplantation to monitor development of thrombosis, coupled with the use of operative thrombectomy and thrombolysis, has essentially eliminated the development of hepatic abscesses in this population. Ascending cholangitis is common after liver transplantation and is usually associated with biliary tract abnormalities. This diagnosis is typically made on clinical grounds in a patient with fever and biochemical evidence of biliary inflammation (see Chapter 68, Cholecystitis and Cholangitis). Enteric gram-negative bacteria and enterococcal species predominate. However, this clinical picture can be identical to that of acute graft rejection; liver biopsy should be performed to differentiate these processes. A cholangiogram is performed to assess the status of the biliary tract for patients with proven cholangitis. As many as 40% of children undergoing liver transplantation develop a fungal infection during the first year following this procedure.35 Candida spp. are the most common fungal pathogens, and infection is usually associated with an intra-abdominal focus or indwelling catheter. Episodes of invasive aspergillosis are uncommon but can be fatal.37 Children undergoing liver transplantation for cystic fibrosis especially have risk of developing infection due to Aspergillus spp.35

BACTERIAL AND FUNGAL INFECTION With the exception of infections related to the use of indwelling catheters, sites of bacterial infection tend to occur at or near the transplanted organ; recovery of bacteria with resistance to multiple antibiotics is common. Knowledge of surveillance cultures and local antimicrobial resistance patterns helps guide empiric antibiotic therapy in these patients.

Renal Transplantation Septicemia originating from the urinary tract, the lower respiratory tract, or the transplant wound accounts for most life-threatening infections in the first month after renal transplantation.14,26 Urinary tract infection, especially pyelonephritis, is the most common infectious complication, accounting for up to 50% of episodes.27 Gram-negative organisms predominate.26,27 Reportedly, one-third of pediatric renal transplant recipients experience recurrent urinary tract infection.28 The incidence may be decreased with the use of uretoneocystostomy and prophylactic use of trimethoprim-sulfamethoxazole (TMP-SMX).16,29 Infection of the lower respiratory tract is seen in 10% to 25% of adult renal transplant recipients and can occur more than a year after transplantation.26,28 Episodes of pneumonia, caused by both gramnegative and gram-positive bacteria, can be severe.30 Data from the North American Pediatric Renal Transplant Cooperative Study (NAPRTCS) suggest that the rates for hospitalizations between 6 and 12 months after pediatric kidney transplantation have increased due to bacterial and viral infections.31 Fungal infections are uncommon after renal transplantation. When present, Candida spp. predominate, and the urinary tract is the most common site.14 However, patients are also at risk of infection with opportunistic fungi, such as Aspergillus spp.28,32 Those with fungal infections appear to be at higher risk for graft loss.31

Liver Transplantation Bacterial and fungal infections are a common early problem after liver transplantation.33–36 BSI often occurs in association with intraabdominal infection or with use of a central venous catheter, but it can occur without an obvious source. Enteric gram-negative organisms

Intestinal Transplantation A relatively small number of children have received intestinal transplants. Many have undergone combined transplantation of liver and intestine or multivisceral transplantation. Bacterial infection occurs frequently in these patients.17,38,39 BSI, which can be explained in part by disruption of the mucosal barrier associated with harvest injury or rejection, is a common finding.38,39 Coagulase-negative staphylococci, enterococci, and gram-negative enteric bacilli account for most episodes. Candidemia can also occur in these settings. While the majority of episodes of BSI occur in the first 3 to 6 months following transplantation, later episodes also occur. Intra-abdominal and wound infections are also frequent in this population, occurring in more than one-third of patients, and are typically detected during the first month after transplantation.

Heart Transplantation Infections account for approximately 15% of early deaths and 10% of late deaths after cardiac transplantation.40 Infection of the lower respiratory tract (including both pneumonia and lung abscess) is the most common site of infection reported in most, but not all, series of pediatric heart transplantations.9,41–43 Mediastinitis is an important infection after thoracic transplantation, particularly if re-exploration of the chest is required. Pathogens associated with mediastinitis include Staphylococcus aureus and gram-negative enteric bacilli. Children who undergo heart transplantation at a young age are at increased risk for invasive infection due to Streptococcus pneumoniae disease.44 Fungal infections, although less common than bacterial infections, can be severe after heart transplantation. Candida spp. predominate, but serious Aspergillus infection can also occur.41 Cryptococcus spp., which rarely cause disease in young children, were identified in only 1 of 192 patients under the age of 19 years who underwent heart transplantation at the Children’s Hospital of Pittsburgh; the patient was an older adolescent. Infants undergoing heart transplantation represent a unique population and have increased risk of serious bacterial or fungal disease. Backer and colleagues45 reported that over 20% of infants who underwent heart transplantation had serious bacterial infection. Similarly

554

SECTION

O Infections and Transplantation

high rates of bacterial or fungal infection have been noted in infants younger than 6 months of age who underwent heart transplantation in Pittsburgh.

Heart–Lung and Lung Transplantation Infection accounts for approximately 50% of all deaths in the first year after pediatric lung transplantation.40 Infection was either the primary or contributing cause of death in the majority of lung and heart–lung recipients who died after the perioperative period in one series.24 Recipients of lung transplantation are at high risk for developing bacterial infection of the respiratory tract. Pneumonia, the most important infectious complication, is difficult to diagnose definitively because differentiation between chronic colonization and lower respiratory tract infection can be problematic. Gram-negative pathogens and Staphylococcus aureus, often with resistance to multiple antimicrobial agents, can be recovered in the presence or absence of disease. Radiographic abnormalities are present almost universally in patients with pneumonia or graft rejection.46–48 Therefore, bronchoalveolar lavage or transbronchial biopsy is often required to help distinguish between causes. Children undergoing lung transplantation because of cystic fibrosis have a high rate of infectious complications. They are usually colonized with Pseudomonas or Aspergillus spp. Empyema necessitans developed in 2 of 3 patients with cystic fibrosis in Pittsburgh who were colonized with Burkholderia cepacia before transplant.49 Further studies recognized that colonization with B. cenocepacia (formerly B. cepacia genomovar III) was associated with excessive mortality.50–52 BSI due to organisms present before transplantation is common after lung transplantation in patients with cystic fibrosis.49,53 Because of the importance and difficulty in treating these bacterial complications, protocols at most transplant centers include thorough evaluation of the microbial flora of candidates prior to transplantation. This should include antibiotic synergy testing and evaluation of isolates of B. cepacia complex species at reference laboratories.

VIRAL INFECTIONS Viral pathogens, especially herpesviruses, are a major source of morbidity and mortality after solid-organ transplantation. Patterns of disease associated with individual viral pathogens are generally similar among all transplant recipients. However, frequency, mode of presentation, and relative severity can differ according to the type of organ transplanted and the pretransplant serologic status of the recipient.

Cytomegalovirus CMV remains one of the most common and important causes of viral infection after solid-organ transplantation.54 Infection can be asymptomatic or symptomatic and can be due to primary infection, reactivation of latent infection, or superinfection with a different strain in a previously seropositive individual. Before the use of prophylaxis against CMV, the reported incidence of disease in children after transplantation was 22% for kidney,50 40% for liver,11 and 26% for thoracic9 organs. Preventive strategies and the use of ganciclovir treatment have resulted in decreased rates and severity of CMV disease. Primary CMV infection, typically acquired from an organ donor, is associated with the highest morbidity and mortality rates. Reactivation of latent infection or superinfection with a new strain of CMV tends to result in milder illness.5 Patients treated with unusually high doses of immunosuppressive agents, especially antilymphocyte products, have increased rates of CMV disease, regardless of previous immunity.9,13,50 Symptomatic CMV disease typically manifests during the intermediate period after transplantation. However, this time of onset

can be later if antiviral prophylaxis is administered.55 A characteristic constellation of fever (which can be high-grade, prolonged, and hectic) and hematologic abnormalities (including leukopenia, atypical lymphocytosis, and thrombocytopenia) is frequently seen. Disseminated CMV disease is manifest by visceral organ involvement; common sites include the gastrointestinal tract, liver, and lungs. The site of involvement can vary according to the type of transplant. Unlike in patients with the acquired immunodeficiency syndrome (AIDS), CMV chorioretinitis is uncommon in organ transplant recipients. Before the availability of ganciclovir, fatal, disseminated CMV disease occurred in 4%, 14%, and 19% of infected children after kidney,50 heart or heart–lung,9 and liver transplantation,11 respectively. Although mortality due to CMV disease has become rare in the ganciclovir era, CMV viremia is still associated with an increased rate of death and retransplantation in pediatric lung transplant recipients.56 Diagnosis of CMV disease is confirmed in a patient with a compatible clinical syndrome by means of histopathology, culture, or a molecular-based quantitative CMV assay such as a pp65 antigen assay, or quantitative detection of CMV DNA in blood by polymerase chain reaction (PCR).57 However, results of viral cultures of the urine and respiratory secretions (including bronchoalveolar lavage specimens) can be difficult to interpret because patients frequently shed CMV asymptomatically in these secretions. Quantitative tests have been most valuable in predicting which patients are at risk for disease, thus allowing pre-emptive initiation of ganciclovir. Histologic examination of involved organs to confirm the presence of CMV remains the gold standard when invasive CMV disease is suspected. A variety of CMV-preventive strategies have been used in pediatric transplant recipients.54,58 Approaches include the use of antiviral agents as prophylaxis, monitoring for virus detection to inform pre-emptive administration of antiviral agents, administration of immunoglobulin products, or a combination of approaches. In addition, most centers use CMV-seronegative or leukocyte-reduced blood products. Antiviral agents with activity against CMV (e.g., ganciclovir and foscarnet) have dramatically improved the survival of transplant recipients with CMV disease. For clinical CMV disease (fever with cytopenia or visceral disease), ganciclovir therapy is given in conjunction with reduction of the immunosuppressant regimen, unless there is evidence of rejection concurrently. Clinical response usually occurs 5 to 7 days after treatment is begun. Duration of therapy is based on demonstrating clearance of CMV viral load, as demonstrated by performance of serial quantitative assays.54 The role of CMV hyperimmune globulin in combination with ganciclovir in the treatment of CMV disease is controversial, although some evidence for improved outcome with this therapy has been reported in the treatment of CMV pneumonia in adult transplant recipients.59 Studies of oral valganciclovir in adults have shown it to be comparable with intravenous ganciclovir for clearance of viremia.60 However, data are still being gathered on appropriate dosing of valganciclovir in the pediatric population. Use of foscarnet or cidofovir should be restricted to patients with apparent or proven resistance to ganciclovir.

Epstein–Barr Virus Recognition of mortality and morbidity caused by EBV after solidorgan transplantation is evolving.10,18,19,61 A wide spectrum of EBV disease is recognized, including nonspecific viral illness, mononucleosis, and PTLD, including lymphoma. Histologic evaluation is important in differentiating among these categories; manifestations can evolve in individual patients. Asymptomatic seroconversion also occurs. Variation in severity and extent of disease is related to the degree of immunosuppression and adequacy of the host immune response. Symptomatic EBV infection in general, and PTLD in particular, is more common after primary EBV infection, thus affecting children disproportionately.6 In one study, 4% of children undergoing solidorgan transplantation and 10% of children with primary EBV infection developed PTLD between 1 month and 5 years after transplantation6; 75% of cases occurred during the first postoperative year in patients

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Infections in Solid-Organ Transplant Recipients

undergoing cyclosporine-based immunosuppression. Cumulative occurrence can be 12% to 20% by 7 to 12 years after liver transplantation.25,62 Onset of viral syndrome, mononucleosis, and PTLD occurs primarily within the first year, whereas lymphoma tends to occur later. Immunosuppressive regimens based on the use of tacrolimus appear to have affected a later timing of PTLD; only rare cases occur longer than 18 months after transplant.63,64 The impact of newer immunosuppressive agents and regimens on EBV disease remains to be determined. The diagnosis of EBV-associated PTLD is made on the basis of clinical, laboratory, and histopathologic examination and should be suspected in patients with protracted fever, exudative tonsillitis, lymphadenopathy, organomegaly, leukopenia, or atypical lymphocytosis (see Chapter 207, Epstein–Barr Virus Infections (Mononucleosis and Lymphoproliferative Disorders)).18,19 Gastrointestinal involvement should be suspected in patients with persistent fever and diarrhea. Serologic diagnosis is often confounded by the presence of passive antibody acquired at the time of transplantation or during subsequent blood product transfusions. The detection of increased EBV viral load identified by EBV PCR is gaining wide acceptance as an assay to predict risk for, or presence of, EBV or PTLD.19,65–67 Although extremely sensitive, these assays are limited by their lack of specificity; viral load is often elevated in asymptomatic patients.68 Accordingly, every effort should be made to confirm the diagnosis of EBV or PTLD histologically. Occult sites of PTLD are assessed by performance of computed tomography of chest and abdomen. Palpable nodes or lesions (or both) identified by surveillance imaging should be biopsied. Endoscopic evaluation should be considered in patients with an elevated viral load and diarrheal illnesses. Histologic evaluation for typical features can be augmented through the use of the Epstein–Barr encoded RNA (EBER) probe.69 Management of patients with PTLD is controversial.18,19,63 Reduction of immunosuppression is widely recommended. Antiviral agents are typically used,65,70 although their role has not been studied formally; the potential impact of monoclonal antibodies,71 interferon,72 and chemotherapy73 awaits formal clinical trials. Resection of tumor may also be of value for patients with lymphoma.

Adenovirus Adenovirus is the third most important viral infection after liver transplantation, occurring in 10% of 484 pediatric liver transplant recipients in one series.74 Symptomatic disease (ranging from self-limited fever, gastroenteritis, or cystitis, to devastating illness with necrotizing hepatitis or pneumonia) occurred in over 60% of infected patients. Infections occurred within the first 3 months after transplantation. The frequency of invasive adenovirus infections after pediatric liver transplantation appears to have decreased markedly with the use of tacrolimus-based immunosuppression. This may be due to the ability to decrease the overall amount of immunosuppressive agents.75,76 Adenovirus infection in other pediatric organ recipients is less well characterized but can be particularly severe after lung transplantation and often leads to fatal disease.24,77,78 The presence of adenovirus DNA in cardiac biopsies after pediatric heart transplantation was significantly associated with poor graft survival in one series.79 Adenovirus has also been associated with hemorrhagic cystitis and graft dysfunction in adult renal transplant recipients.80 It has also been found in high rates after pediatric intestinal transplantation.76 As adenovirus, like CMV, can be latent and can reactivate asymptomatically, ascribing a causative role in the pathologic process can sometimes be difficult.

Varicella-Zoster Virus Many children undergo solid-organ transplantation before development of immunity against varicella-zoster virus (VZV). Nineteen of 160 pediatric renal transplant recipients developed varicella after transplantation in one series; 8 had severe disease, including 1 death.81 Similarly, 14 of 47 susceptible pediatric liver transplant recipients

CHAPTER

97

555

developed varicella in another series.82 Despite the use of intravenous acyclovir in 13 of these children, 2 patients died. Disease developed in approximately 50% of patients who received postexposure varicellazoster immune globulin (VZIG), including a spectrum of severity and 1 fatal case. However, a recent report of 22 children who developed varicella following liver transplantation was more encouraging, with none developing any complications of VZV.83 All of these children were treated with intravenous acyclovir and none of them had been treated with high-dose corticosteroids for rejection before the onset of VZV infection. An aggressive response to varicella exposure and disease is still warranted. A VZV immunoglobulin product should be administered within 72 hours of a varicella exposure in nonimmune patients. Since production of VZIG has been halted, various recommendations have been made, including the use of intravenous immunoglobulin and investigational use of a Canadian VZV immunoglobulin product (Varizig), or, if prompt acquisition is not feasible, oral acyclovir starting at day 7 postexposure as off-label use. If varicella lesions develop, it is prudent to hospitalize patients and administer acyclovir intravenously until fever abates, no new lesions erupt, and existent lesions begin to crust. Limited data in stable patients suggest that oral acyclovir may also be acceptable if close follow-up is assured. The use of VZV vaccine posttransplantation of solid organs is an area of active research. Limited experience in pediatric renal84 and liver transplant recipients85 is encouraging; immunogenicity and safety appear to be acceptable in children who are more than 1 year out from transplant who have stable and relatively low levels of immunosuppression. Additional experience is needed before recommending a standard protocol for VZV vaccination following transplantation.

Common Community-Acquired Viruses Although the course of illness has been poorly documented, most children who receive solid-organ transplants experience the usual childhood respiratory and gastrointestinal tract illnesses without significant problems. This is especially true when they occur long after transplantation and are not associated with an episode of rejection.7 Bailey and associates42 noted that most of 43 pediatric heart transplant recipients experienced typical childhood illnesses without resultant fatality. However, infections caused by RSV, influenza virus, or parainfluenza virus have led to more severe disease in young children, especially if they occur soon after transplant and during periods of maximal immunosuppression.86,87 Shirali and colleagues found that the presence of DNA from community-associated viruses such as enterovirus, adenovirus, and parvovirus in heart biopsies of children after cardiac transplantation was associated with increased adverse events but the timing of the biopsies relative to the time of transplantation was not evaluated.79

OPPORTUNISTIC INFECTIONS

Pneumocystis jirovecii (P.carinii) Pneumonia PneumoCystis Pneumonia (PCP) is a well-documented complication of solid-organ transplantation. Before the widespread use of prophylaxis, the incidence of PCP was 4% to 35%.20–22 However, use of TMP-SMX prophylaxis has essentially eliminated this problem. PCP typically occurs after the first month following transplantation, reflecting indolence of this pathogen. Most cases occur within the first year, but PCP can occur later and should remain in the differential diagnosis for patients with fever and lower respiratory tract symptoms, particularly if they are not receiving PCP prophylaxis.

Toxoplasmosis Toxoplasma gondii causes a significant infection in immunocompromised hosts.88 A rare cause of disease in renal and liver transplant recipients,14,89 it is more often an infectious complication after cardiac

556

SECTION

O Infections and Transplantation

transplantation.23 The risk in cardiac recipients may be explained by tropism of the organism for cardiac muscle and subsequent donor transmission. Reactivation of cysts within the graft occurs in the immunosuppressed recipient without previous immunity. This is in contrast to adult patients with AIDS who develop reactivation disease despite the presence of antibody. Four Pittsburgh pediatric cardiac transplant patients developed disease after primary infection.90 Two had disseminated disease with myocardial parasites documented on tissue biopsy; 1 also developed focal neurologic impairment and severe chorioretinitis. Both patients died despite treatment. Two other patients remain asymptomatic several years after transplantation. Clinical manifestations usually occur 2 to 24 weeks after transplantation and include fever, pulmonary disease, chorioretinitis, myocarditis, and neurologic disorders. Use of pyrimethamine for prophylaxis in adult patients has shown promising results,73 and several studies have shown efficacy of TMP-SMX prophylaxis.91,92

Tuberculosis Tuberculosis is a special concern in immunosuppressed hosts, including recipients of solid organs. Although the development of tuberculosis is rare in the Pittsburgh experience, an incidence of 2.4% was reported from a pediatric liver transplant center in the United Kingdom.93 While most cases in this series were felt to represent primary infection acquired after transplantation, transplant recipients with a reactive Mantoux purified protein derivative (PPD) test or who come from areas where tuberculosis is endemic are also at increased risk for symptomatic reactivation after transplantation.94–96 Although the risk of reactivation appears greatest in patients who received inadequate therapy for tuberculosis, disease can also occur in patients who received appropriate therapy prior to transplantation. Accordingly, a careful history of exposure, a Mantoux PPD, and a chest radiograph are evaluated prior to transplantation. Patients with a positive tuberculosis history or a positive PPD test result should receive isoniazid for at least 9 months after transplant.93,97 Treatment prior to transplantation is recommended in children. Attempts at a more definitive diagnosis are indicated in a patient from an endemic area who has a negative PPD test result but a suspicious radiograph. Evidence of side effects, particularly hepatotoxicity, should be carefully monitored in all pediatric transplant recipients receiving chemotherapy for tuberculosis.

Other Opportunistic Infections Additional opportunistic infections include cryptococcosis, coccidioidomycosis, and histoplasmosis. Previous infection with these pathogens is associated with exposure to geographic areas where pathogens are endemic. Because patients often travel to transplant centers distant from their homes, physicians caring for candidates or recipients of solid-organ transplants must be cognizant of the environmental risk for each patient. Experience with coccidioidomycosis in transplant recipients suggests that a minimum of 4 months of antifungal therapy, such as fluconazole, should be given to transplant recipients with this history.98 Similarities between coccidioidomycosis and other fungal infections suggest that such strategies may be necessary for patients with a history of fungal infection with pathogens prone to recurrence after resolution of primary infection.

MANAGEMENT AND PREVENTIVE MEASURES Pretransplant Evaluation Pretransplant evaluation permits preventive intervention and anticipation of posttransplantation complications. History and physical examination should be performed, with particular attention to previous

infections, immunizations, and drug allergies. Children with cystic fibrosis or those with a prolonged intensive care unit stay just before transplantation may be colonized with resistant organisms. Pretransplant surveillance cultures are useful in guiding subsequent selection of antimicrobial agents in these patients. Evaluation for tuberculosis is performed. Serologic studies for CMV, EBV, hepatitis B and C viruses, syphilis, HIV, and VZV should be obtained in all candidates. In addition, other serology, such as for Toxoplasma gondii, particularly for heart transplant candidates and hepatitis A serology for liver transplant candidates, should be performed by protocol at individual centers. Similar donor serology is performed (Tables 97-2 and 97-3). WNV has been transmitted by organ donation from undiagnosed cases, resulting in severe disease in the recipient. Currently, blood donor screening but not organ donor screening by WNV nucleic acid amplification test is routinely performed.99 Inactivated and purified antigen vaccines are given to maximize protection before immunosuppression. Children expected to have a prolonged wait before transplantation should be given live virus vaccines as scheduled (and often in an accelerated fashion). In many centers, immunologic response is assessed by obtaining serology for measles, mumps, and/or rubella. If seroconversion is not documented, revaccination can be attempted if time permits.

Preventive Strategies Prophylactic regimens for solid-organ transplantation vary by center and type of transplant. Current protocols used at the Children’s

TABLE 97-2. Screening Tests for Transplant Candidatesa Test and Pathogen

Comment

SEROLOGIC TESTb

HIV-1 and -2 HTLV-1 and -2 Hepatitis A virus Hepatitis B virus

Hepatitis C virus Hepatitis D virus CMV EBV Herpes simplex virus Varicella-zoster virus Toxoplasma gondii Measles virus Mumps virus Rubella virus

IgG and IgM screening test Hepatitis B surface antigen and anti-core antibody indicate active disease; anti-hepatitis B surface antigen indicates serologic conversion after immunization If Hepatitis B-positive IgG test; urine culture if positive Anti-viral capsid antigen IgG and EBNA Obtain for heart and heart–lung transplant candidates in particular; some centers screen all candidates If serology is negative, consider immunization if ≥ 3 months anticipated before transplantation If serology is negative, consider immunization if ≥ 3 months anticipated before transplantation If serology is negative, consider immunization if ≥ 3 months anticipated before transplantation

OTHER TESTS

Mycobacterium Mantoux intermediate PPD skin test with tuberculosis anergy panel Respiratory tract pathogens Sputum culture for patients with cystic fibrosis and other lung transplant candidates CMV, cytomegalovirus; EBNA, Epstein–Barr virus nuclear antigen; EBV, Epstein–Barr virus; HIV, human immunodeficiency virus; HTLV, human T-lymphotropic virus; Ig, immunoglobulin; PPD, purified protein derivative (tuberculin). a All tests performed on all candidates except where noted. b IgG antibody measured except where noted.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Infections in Solid-Organ Transplant Recipients

Hospital of Pittsburgh are shown in Table 97-4. Perioperative antibiotics are used for the first 48 to 72 hours to provide prophylaxis against intraoperative soilage, septicemia, and wound infection. The choice of antimicrobial agents is dictated by the organ being transplanted, patient characteristics, expected flora, and knowledge of the antimicrobial susceptibilities of local pathogens. Surveillance cultures of the donor bronchi or trachea are also useful in heart–lung or lung transplantation. The frequency and severity of CMV infection in transplant recipients prompt consideration of prophylactic strategies; optimizing the target population and timing of interventions require further study.100 Potential roles exist for intravenous and oral ganciclovir101,102 and oral valganciclovir.103,104 Currently we recommend the use of intravenous ganciclovir alone (for varying durations) for liver, heart,

TABLE 97-3. Serologic Screening of Organ Donora Pathogen

Comment

HIV-1 and -2 HTLV-1 and -2

Positive result contraindicates organ use Positive result is relative contraindication for organ use Positive result IgM test contraindicates organ use Obtain complete serologic panel; result suggesting current infection contraindicates organ use Some centers use positive donor only for positive candidate Obtain IgG test; obtain urine culture if positive neonatal donor Obtain complete serologic panel Obtain on heart and heart–lung donors in particular Obtain reagin test (specific test if positive); positive result contraindicates organ use

Hepatitis A virus Hepatitis B virus Hepatitis C virus CMV EBV Toxoplasma gondii Treponema pallidum

CMV, cytomegalovirus; EBV, Epstein–Barr virus; HIV, human immunodeficiency virus; HTLV, human T-lymphotropic virus; Ig, immunoglobulin. a IgG antibody measured except where noted.

CHAPTER

97

557

and lung transplant recipients and ganciclovir plus intravenously administered immune globulin containing a high titer of antibody against CMV for high-risk (donor CMV-positive/recipient CMVnegative) intestinal transplant recipients if available. At the present time, the use of oral valganciclovir is an alternative to prophylaxis with intravenous ganciclovir for adolescents. Pharmacokinetic studies for dosing oral valganciclovir suspension in children after transplantation are ongoing. Upon completion of these studies, use of valganciclovir suspension will likely be of merit as an alternative to intravenous treatment for younger children. Serial monitoring of the blood CMV viral load by using either the pp65 antigen assay or quantitative CMV PCR as an indicator for the use of pre-emptive antiviral therapy has been proposed as an alternative to these chemoprophylactic and immunoprophylactic strategies.105 In this approach, only patients demonstrating increased risk because of increased viral load are treated with intravenous or oral ganciclovir. Although this strategy has gained acceptance at some centers, experience in pediatric transplant recipients remains limited. The use of viral load monitoring after completion of chemoprophylaxis is gaining acceptance. A number of strategies are currently being explored (e.g., immunoprophylaxis, monitoring, and pre-emptive therapy) as an attempt to prevent EBV disease PTLD106,107; the efficacy of these approaches have not been established. The use of viral load monitoring to inform pre-emptive reductions in immunosuppression appears to be the most promising of these strategies.108 Nystatin suspension can be used in pediatric transplant recipients for the first 3 months after transplantation to prevent oropharyngeal candidiasis. TMP-SMX is used to prevent PCP. This strategy has also been shown to decrease the incidence of posttransplant urinary tract infections in renal transplant recipients.27 In these patients, TMP-SMX is used twice a day during the initial hospitalization; renal transplant recipients are usually given a single daily dose in the early posthospitalization period. The duration of prophylaxis for PCP is controversial. Most cases occur during the first year after transplantation. However, because late cases occur, TMP-SMX is given indefinitely at some centers.

TABLE 97-4. Antimicrobial Prophylaxis After Solid-Organ Transplantation at the Children’s Hospital of Pittsburgh Prophylaxis Type

Organ Type Kidney

Liver

Heart a

Lung

Intestine

Cefazolin (24 hours after procedure)

Piperacillin/tazobactam (2 days)

Cefazolin (2–3 days)

Clindamycin plus ceftazidime (2–3 days)b

Piperacillin/ tazobactam (2–3 days)

Cytomegalovirus

Ganciclovire

Ganciclovird,g

Ganciclovird,g,i

Ganciclovirg,h,i

Ganciclovird,f Anti-CMV IGIVc

Antifungal

Nystatin (2 months)

Nystatin (2 months)

Nystatin (2 months)

Nystatin (2 months) Amphotericin lipsomal formulation or voriconazolej

Nystatin (2 months)

TMP-SMXl

TMP-SMXl

PERIOPERATIVE AGENT-SPECIFIC

Pneumocystis jirovecii TMP-SMXk

TMP-SMXl

TMP-SMXl

CMV, cytomegalovirus; IGIV, immune globulin intravenous; TMP-SMX, trimethoprim-sulfamethoxazole. a For infants, use ampicillin and cefotaxime for 2–3 days. b For patients with cystic fibrosis, perioperative antimicrobial therapy is individualized according to pretransplant culture and perioperative cultures and susceptibility data. Duration is 2 weeks. c High-risk patients (CMV-positive donor/CMV-negative recipient) only. d Dose of 5 mg/kg q12 hours μ 14 days. e Oral ganciclovir or valganciclovir μ 6 months for all patients. f All transplant recipients. g All high-risk and CMV-positive recipients. h Dose of 5 mg/kg q12 hours μ 14 days; then 5 mg/kg qd μ additional 14 days. i High-risk patients get a combination of intravenous ganciclovir and valganciclovir until 3 months after transplantation. j Amphotericin liposomal formulation or voriconazole is given intravenously until intraoperative cultures from the donor and the recipient are negative; a more prolonged course is given for cystic fibrosis patients with a history of aspergillis colonization. k TMP-SMX (5 mg/kg TMP) in single daily dose μ 4–6 months; then same single dose 3 times a week indefinitely. l Given as 5 mg/kg TMP in single dose 3 times a week indefinitely (maximum daily dose, 80 mg TMP).

558

SECTION

CHAPTER

O Infections and Transplantation agents, the incidence of associated disease has decreased. However, the incidence of infections due to adenovirus and Epstein–Barr virus (EBV) is increasing with the use of mismatched HSCT and Tlymphocyte-depleted HSCT.8,9 A third period associated with deÀcits in humoral immune responses, cellular immune responses, and reticuloendothelial function begins at 100 days after HSCT. Varicellazoster virus (VZV) and encapsulated bacteria, particularly Streptococcus pneumoniae and Haemophilus influenzae type b (Hib), are major pathogens during this period.

98

Infections in Hematopoietic Stem Cell Transplant Recipients Jorge Luján-Zilbermann

EPIDEMIOLOGY Hematopoietic stem cell transplantation (HSCT) has broad indications in pediatrics, including patients with cancer, primary immunodeÀciency syndromes, bone marrow failure syndromes, hemoglobinopathies, and an assortment of genetic conditions, including inborn errors of metabolism.1,2 Patients undergoing HSCT have an increased risk for infectious complications that are somewhat predictable based on the acquired immune deÀciencies that occur after HSCT.3–6

ETIOLOGIC AGENTS Patients who have had HSCT have immune deÀciencies in the phagocytic, humoral, and cellular arms of the immune system.3 These immune defects cause disease to occur from infectious agents in three predictable time periods (Figure 98-1).3 After the conditioning regimen, a period of neutropenia occurs for 3 to 4 weeks. During this period, bacteria and fungi cause most infections; in addition, herpes simplex virus (HSV) and seasonal respiratory viral infections can occur.3,6,7 The second phase occurs after granulocyte recovery and continues until approximately 100 days after HSCT. Infectious complications during this period are associated with profound impairment of humoral and cellular immunity. Although bacterial and fungal infections can still occur, they are much less frequent than during the neutropenic period. Cytomegalovirus (CMV) is a major infecting agent, but Pneumocystis jirovecii (P.carinii) disease can also occur. With the advent of effective prophylaxis against both of these

HSV RSV Parainfluenza virus Adenovirus

Viral

CMV EBV VZV Gram-negative Gram-positive

Bacterial

Encapsulated bacteria

Most infecting agents in HSCT patients are derived from the patient’s microbial flora or by reactivation of a latent infection. Multiple factors account for high risk of HSCT recipients for an infectious complication (Box 98-1). Immune deÀciency associated with the underlying disease and its status are determinants of the degree of immune defects. Allogeneic HSCT recipients (e.g., matched unrelated donor or unrelated cord blood transplantation) are at high risk for graft-versus-host disease (GvHD), which enhances the infection rate by delaying return of normal immune function and by ulceration of the gastrointestinal tract. Moreover, the risk of infection is directly related to the degree of donor–recipient mismatch. To abrogate GvHD, cyclosporine and methotrexate are administered as prophylaxis. Both of these agents increase the risk of infection by depressing the cellmediated immune response and, in the case of methotrexate, by disrupting mucosal barriers. The conditioning regimen with or without concomitant irradiation compromises the immune system and can also disrupt mucosal barriers. In addition, purging of the bone marrow in autologous transplants to reduce the load of malignant cells, and T-lymphocyte depletion used in allogeneic transplants to reduce the incidence of GvHD, predispose the host to infection. The serologic status of the donor and the recipient is important because many infections in patients with transplants are due to reactivation (Table 98-1). This is most notable with the herpesvirus group. CMV is a major cause of pneumonitis in allogeneic transplant recipients. Additionally, other herpesviruses, especially HSV, VZV, and EBV, as well as Toxoplasma gondii and adenovirus, are prone to reactivate after the transplant. All HSCT recipients have a central venous catheter (CVC) placed before transplant, providing a potential site for infection. In addition, patients who have other indwelling medical devices (e.g., cerebrospinal fluid shunts) have attendant risk for infection. Patients with a history of invasive aspergillosis who are undergoing HSCT require adequate antifungal prophylaxis and treatment to prevent relapse of aspergillosis.10 Knowledge of the epidemiology of pathogens associated with the hospital and the transplant unit allows an assessment of risk for environmental organisms, such as Aspergillus and Legionella species. Rates of infection can be substantially reduced by preventive mechanisms that inhibit aerosolization of organisms, such as the use of laminar flow rooms or high-efÀciency particulate air (HEPA)Àltered rooms.5

Candida and Aspergillus Fungal

BOX 98-1. Predisposing Factors for an Infectious Complication in Hematopoietic Stem Cell Transplant Recipients

Pneumocystis jirovecii Toxoplasma

Protozoal 0

30

60

90

120

150

180

365

Days following transplantation Figure 98-1. Temporal association of infectious agents and hematopoietic stem cell transplantation. Day 0 is the time of stem cell infusion. Boldness of line denotes increasing frequency of infection with offending agent.

• • • • • • •

Underlying illness Type of transplant Conditioning regimen Infectious disease history Presence of an indwelling medical device Occurrence and severity of graft-versus-host disease Immunosuppressive regimen administered to prevent graft-versus-host disease • Epidemiology of infection in hospital or transplant care unit

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Infections in Hematopoietic Stem Cell Transplant Recipients

CHAPTER

98

559

TABLE 98-1. Pretransplant Evaluation for Patients Undergoing Hematopoietic Stem Cell Transplantation Study Type

Organism or Test

Serum antibody

CMV EBV Hepatitis B (HBsAg, HBsAb, HBcAb) Hepatitis C RPR Human immunodeficiency virus Serum hepatic enzymes (ALT, AST, bilirubin) Renal function tests (BUN, creatinine) Complete blood cell count with differential Stool for ova and parasites Mantoux purified protein derivative Chest, posteroanterior and lateral Sinus series, if clinically indicated

Other laboratory

Skin Radiographic

ALT, alanine aminotransferase; AST, aspartate aminotransferase; BUN, blood urea nitrogen; CMV, cytomegalovirus; EBV, Epstein–Barr virus; RPR, rapid plasma reagin.

Figure 98-2. Invasive aspergillosis in a patient after hematopoietic stem cell transplantation. Computed tomography of the chest shows bilateral pulmonary involvement, including aspergilloma in the right middle lobe (arrow).

CLINICAL SYNDROMES, DIFFERENTIAL DIAGNOSIS, AND CLINICAL APPROACH The approach to a patient who has had HSCT is based on understanding the infections that can occur during each of the risk periods (see Figure 98-1). This provides the framework for matching possible causative agents with the clinical syndrome. Because some of these clinical syndromes are highlighted in Chapters 99 (Fever and Granulocytopenia) and 100 (Infection in Children with Cancer), the following discussion focuses on a clinical approach to the HSCT recipient.

Early Period (Before Engraftment) Herpes Simplex Virus Gingivostomatitis HSV infection occurs primarily as a result of reactivation in seropositive patients undergoing HSCT. The diagnosis is difficult because lip lesions are rare and mucosal ulcerations are similar to those that occur as a result of the conditioning regimen.

Central Venous Catheter-Related Bloodstream Infections During the neutropenic period, HSCT recipients have high risk for bacterial infection, comparable with the risk for patients with cancer who develop chemotherapy-induced neutropenia. Catheter-related bloodstream infections (BSIs) are common because of the uniform use of indwelling CVCs in HSCT patients for administration of medications, hyperalimentation, blood products, and for blood sampling. Staphylococcus epidermidis is the most frequent BSI during the three phases posttransplant.11–13 Other gram-positive organisms associated with BSI in HSCT recipients include viridans streptococci and S. aureus.11–13 Viridans streptococci infections have been associated with chemotherapy-induced mucositis and poor dental hygiene.14 Most gram-positive catheter-related infections can be treated successfully without removal of the catheter. Gram-negative bacillary infections occur after mucosal damage with bacterial translocation from the intestinal mucosa into the bloodstream and are the second most frequent cause of BSI. The predominant organisms in this class include Escherichia coli, Klebsiella spp., and Pseudomonas aeruginosa, although other gramnegative organisms (frequently indigenous to the facility) are common.11,13,15 Antibiotic resistance among these organisms is common. Catheter-related infections with some gram-negative bacilli, Candida spp. and Bacillus cereus can be problematic and require catheter removal.

Fungal Infections The major causes of fungal infection include Candida spp., Aspergillus spp., and agents of mucormycosis (e.g., Mucor, Absidia, and Rhizopus spp.).16–19 Additionally, other fungi recognized as pathogens include Trichosporon spp., Fusarium spp., Curvularia spp., and Alternaria spp.20,21 Infection with these organisms usually occurs after a period of antibiotic therapy and is correlated with the degree and duration of neutropenia. Although Candida albicans is the most frequent Candida spp. causing BSI, C. tropicalis may cause more severe disease. Other Candida spp., including C. glabrata, C. parapsilosis, and C. krusei, have also emerged owing to resistance to fluconazole, which is used as prophylaxis.16,17 The portal of entry for Aspergillus and agents of mucormycosis is the respiratory tract, as opposed to that of Candida spp., which is the gastrointestinal tract. Aspergillus is associated with sinopulmonary disease and dissemination (Figure 98-2), but is rarely recovered from blood cultures. Diagnosis usually depends on tissue histology and culture of material obtained from bronchoscopy, lung aspiration, or open-lung biopsy. An assay for Aspergillus galactomannan is undergoing validation tests to establish value in early diagnosis of invasive aspergillosis in high-risk patients.22

Hemorrhagic Cystitis Hemorrhagic cystitis is associated with a variety of infectious and noninfectious causes (Box 98-2).23 The onset can occur at any time during the transplantation period; chemotherapy-induced cystitis occurs soon after commencing the conditioning regimen. The most common infectious causes are polyoma viruses (BK virus and JC virus) and adenovirus.23,24 Bacterial and fungal pathogens must also be considered.

Enteric Infections Diarrhea after transplantation is more commonly caused by mucositis and GvHD. Enteric infections can occur throughout the transplantation period. Antibiotic-associated diarrhea, including that due to Clostridium difficile, usually occurs during the neutropenic period, when antimicrobial agents are frequently administered. Other infecting agents, including enteric adenovirus, rotavirus, and coxsackievirus, should also be considered.9,25 The enteric viruses are generally seasonal in occurrence.

560

SECTION

O Infections and Transplantation

BOX 98-2. Differential Diagnosis of Hemorrhagic Cystitis in Hematopoietic Stem Cell Transplant Recipients

BOX 98-3. Causes of Pneumonitis in Hematopoietic Stem Cell Transplant Recipients

INFECTIOUS CAUSES Virus Adenovirus Cytomegalovirus Polyoma viruses, especially BK virus and JC virus Herpes simplex virus Bacteria Urinary tract infection, predominantly gram-negative bacilli Fungus Urinary tract infection Fungus ball NONINFECTIOUS CAUSES Chemotherapy (e.g., cyclophosphamide) Graft-versus-host disease Mechanical trauma from Foley catheter

INFECTIOUS Bacteria Enterobacteriaceae Staphylococcus aureus Legionella pneumophila Fungi Aspergillus species Mucormycosis Pneumocystis jirovecii Candida species Virus Parainfluenza types 1–4 Adenovirus Respiratory syncytial virus Human metapneumovirus Cytomegalovirus Influenza Human herpesvirus 6 Coxsackievirus and echoviruses Rhinovirus NONINFECTIOUS Pulmonary damage by radiation Pulmonary damage by chemotherapeutic agents (e.g., bleomycin) Underlying cancer Pulmonary edema Alveolar hemorrhage Idiopathic interstitial pneumonia Pulmonary vascular disease Pneumomediastinum

Middle Posttransplant Period (Early Engraftment) The middle period (from days 30 through 100 after transplantation) was once dominated by CMV infection, but the incidence of this infection has diminished with the use of ganciclovir as pre-emptive therapy. Bacterial infections are less problematic during this period, except for those associated with indwelling catheters. Fungal infections are still evident in patients with GvHD.

Pneumonitis Pulmonary infiltrates can have an infectious or noninfectious cause (Box 98-3). Although signs and symptoms can occur throughout the transplantation period, viral infections are more common during the early engraftment period. CMV manifests at a median time of 40 to 50 days after the transplantation period (see Figure 98-1). CMV most commonly occurs because of reactivation of latent virus in seropositive individuals but can also occur in seronegative patients who receive a transplant from a seropositive donor. Risk factors for CMV pneumonia include seropositivity of donor, type of transplant (allogeneic more than autologous), human leukocyte antigen mismatch transplant, older age of the patient (> 10 years of age), and development of acute GvHD.26 CMV infection occurs in 30% to 50% of patients undergoing HSCT; pneumonia occurs in 10% to 15% of these patients and has a mortality rate of 85%. CMV infection and disease also occur in autologous HSCT recipients, but at a much lower incidence. The clinical manifestations of CMV disease vary from asymptomatic infection, to the constellation of fever, hepatitis, and leukopenia, to life-threatening diseases, such as interstitial pneumonitis, esophagitis, and encephalitis.27 Other viral causes of respiratory tract infections include adenovirus, respiratory syncytial virus (RSV), the human metapneumovirus, parainfluenza virus, influenza virus, human herpesvirus 6, coxsackievirus, rhinovirus, and echoviruses.28–34 RSV, influenza virus, and parainfluenza virus can cause sinusitis and life-threatening pneumonia.28,31 Pneumocystis jirovecii also causes pneumonia after engraftment.35 Clinical manifestations are similar to that in other immunocompromised hosts. The incidence of PneumoCystis pneumonia (PCP) has been curtailed by routine prophylaxis with trimethoprimsulfamethoxazole (TMP-SMX).

Adenovirus Infection Adenovirus infection occurs in approximately 30% of pediatric HSCT recipients and can become latent in lymphoid tissue and kidneys.36 The most common clinical manifestations of infection include diarrhea, febrile illness, hemorrhagic cystitis, and pneumonia, but can also include hepatitis and encephalitis.36

Toxoplasmosis Toxoplasmosis is a rare but almost always fatal infection after HSCT. It usually occurs 2 to 6 months after HSCT, and the central nervous system is most often affected. Symptoms include focal neurologic signs, fever, seizures, and altered mental status. Imaging of the brain typically shows multiple lesions in both hemispheres and basal ganglia.37

Encephalopathy Encephalopathy is a poorly characterized complication of HSCT that has a very poor prognosis in pediatric patients, with a mortality rate of 65% in one study. Several infectious and noninfectious etiologies have been described (Box 98-4).30,38,39 HHV-6 has been associated with encephalitis and delayed platelet engraftment.30,39

Late Posttransplant Period At 100 days after transplantation, the late period begins.

Bloodstream Infection Bacterial infections occur less commonly during this period, but patients continue to be immunosuppressed and susceptible to bacterial infection and BSI. The presence of chronic GVHD augments this immunosuppression. Encapsulated bacteria, especially S. pneumoniae and Hib, are the most common agents responsible for bacterial infections that are not related to presence of an indwelling catheter.3

Varicella-Zoster Virus Infection VZV infections occur in 25% to 40% of patients following HSCT.40 Reactivation occurs more frequently in association with chronic GVHD but also has been described in patients following autologous

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Infections in Hematopoietic Stem Cell Transplant Recipients

CHAPTER

98

561

BOX 98-4. Differential Diagnosis of Meningitis, Encephalitis, and Encephalopathy in Hematopoietic Stem Cell Transplantation Recipients

BOX 98-5. Differential Diagnosis of Hepatitis After Hematopoietic Stem Cell Transplantation

INFECTIOUS CAUSES Virus Adenovirus Human herpesvirus 6 and 7 Polyoma viruses, especially BK virus and JC virus (progressive multifocal leukoencephalopathy) Herpes simplex virus Postviral (acute disseminated encephalomyelitis) Bacteria Streptococcus pneumoniae Fungus Aspergillus species Protozoa Toxoplasma gondii NONINFECTIOUS CAUSES Medications (e.g., cyclosporine, amphotericin B) Nonconvulsive seizures Thrombotic thrombocytopenic purpura Multiorgan system failure Stroke

INFECTIOUS CAUSES Hepatitis viruses A, B, C, and D Herpes simplex virus Cytomegalovirus Adenovirus Echovirus Epstein–Barr virus Varicella-zoster virus Human herpesvirus 6 NONINFECTIOUS CAUSES Veno-occlusive disease Graft-versus-host disease Chemotherapy-induced hepatitis Hepatopathy of total parenteral nutrition Cholestatic liver injury secondary to septicemia Nonchemotherapeutic drugs, including acetaminophen and antibiotics

HSCT. A prodrome of burning or pain over the involved dermatome can occur. Groups of vesicles appear in the distribution of one to three sensory dermatomes. If appropriate therapy with acyclovir is not instituted, dissemination can occur in 36% of infected patients, with a mortality rate of 10%.

Hepatitis Acute or chronic hepatitis can follow HSCT and has infectious and noninfectious causes (Box 98-5). Hepatitis C virus can infect recipients of HSCT, with activation at the time of discontinuation of immunosuppressive therapy.41

LABORATORY FINDINGS AND DIAGNOSIS Except for microbiologic evaluation, laboratory tests are of limited value. Microbiologic tests include blood cultures for bacteria, fungi, and detection of viruses; specific serologic tests; and special stains. The Bactec Myco/F lytic culture medium or other lysis centrifugation systems enhances fungal isolation from blood. A buffy coat culture for viruses, such as CMV, may improve isolation. Additionally, shell vial techniques (based on a low-speed centrifugation to enhance attachment and monoclonal antibody detect virus replication) can rapidly identify viruses, including CMV, herpesviruses, and adenovirus, particularly those involving the respiratory tract. Polymerase chain reaction (PCR) testing is extremely helpful for both screening and diagnosis of viral infections. PCR is available for all the herpesviruses, adenovirus, and polyoma viruses.30,36 Real-time PCR of respiratory samples allows the diagnosis of influenza virus, RSV, parainfluenza virus, metapneumonovirus, and adenovirus.29 Other laboratory tests are of limited value. Complete blood cell counts can assess engraftment status; absolute neutrophil plus count < 500/mm3 is associated with higher incidence of bacterial infections. Serum hepatic enzyme and renal function tests can provide evidence of pathology from an infectious agent or insults from chemotherapy or GvHD (see Boxes 98-2 and 98-5).

ceftazidime.42 This combination provides effective therapy for viridans streptococci infections, which are common in patients who are given high-dose cytosine arabinoside in the conditioning regimen. In addition, ceftazidime provides adequate therapy for Pseudomonas aeruginosa and many other gram-negative bacteria. This combination avoids the use of aminoglycosides in patients who have received nephrotoxic drugs. Use of ceftazidime can increase incidence of extended-spectrum beta-lactamase production by many gram-negative bacilli, and vancomycin can increase incidence of vancomycinresistant gram-positive cocci. Consideration of antibiotic regimens should take into account isolates indicating the patient’s colonization status, the hospital environment, and the antimicrobial resistance pattern within the community and the hospital. Empiric amphotericin B is begun in the patient who is persistently febrile after 5 to 7 days of antibiotic therapy and who does not have an identified bacterial cause.42 Lipid or liposomal formulations of amphotericin B can be used as substitutes if nephrotoxicity occurs during use of amphotericin B.43 Azole agents, such as itraconazole and voriconazole, are effective for the treatment of invasive aspergillosis in immunocompromised hosts.43,44 The echinocandins, caspofungin and micafungin, alone or in combination therapy with an azole or amphotericin B, are effective therapies for Candida spp. infections and invasive aspergillosis.45,46 Viral infections are treated after a diagnosis is made, if therapy is available. Acyclovir is the treatment of choice for both HSV and VZV infections. Valacyclovir and famciclovir can also be used to treat HSV infections. CMV pneumonitis is treated with both ganciclovir and intravenous immune globulin (IGIV), to inhibit proliferation of CMV and possibly to abrogate the immune response contributing to the pneumonitis. Foscarnet can be used in cases of ganciclovir-resistant CMV or acyclovir-resistant HSV infections. Cidofovir is a safe and effective therapy for adenoviral infections in pediatric patients, with no dose-limiting nephrotoxicity.36 Adoptive immunotherapy, by transferring virus-specific cytotoxic T lymphocytes to patients after allogeneic HSCT, has been used to treat infections caused by EBV and CMV and selectively to reconstitute immune function against these viruses.47,48 Aerosolized ribavirin can be used for the treatment of parainfluenza virus and RSV infections; for the latter, RSVIG is also recommended.49,50

PREVENTION MANAGEMENT AND PRESUMPTIVE THERAPY Initial management of patients with fever and neutropenia is similar to management of patients with chemotherapy-induced fever and neutropenia (see Chapter 99, Fever and Granulocytopenia). One empiric antibiotic therapy regimen may be vancomycin plus

Prevention of infectious complications is a high priority for recipients of HSCT. Prophylaxis with TMP-SMX against PCP is started after engraftment and is given orally three times a week until 6 months after the transplant. The dose is 5 to 10 mg/kg per day divided into 2 doses, or 150 mg/m2 per day divided into 2 doses, for 3 consecutive days per week.

562

SECTION

O Infections and Transplantation

Antifungal prophylaxis includes the use of oral fluconazole or lowdose amphotericin B to reduce the incidence of systemic and superficial fungal infections. Each agent is effective in preventing fungal infections, but fluconazole is better tolerated.51 However, the use of fluconazole has been complicated by the emergence of resistant pathogens, such as Candida krusei and some strains of C. albicans.17 Micafungin is also useful for prophylaxis against invasive fungal infections in HSCT recipients.52 Primary CMV infection can be decreased in the CMV-seronegative HSCT recipient by using leukocyte-filtered blood products for transfusion.53Since the introduction of prophylaxis with ganciclovir for CMV infection, the incidence of CMV disease has been substantially reduced.27 Pre-emptive therapy with ganciclovir or foscarnet is used in seropositive recipients of HSCT or in patients receiving a transplant from a seropositive donor. The use of ganciclovir as prophylaxis can cause neutropenia with an increased risk of a bacterial infection; however, the risk of CMV pneumonia justifies its use. Pre-emptive ganciclovir prophylaxis administered early in the course of HSCT delays the median time of onset of CMV infection from 1 to 2 months to 4 to 6 months.54

Active Immunizations Reimmunization is important because most allogeneic and a large proportion of autologous recipients of HSCT lose their immunity to vaccine-preventable diseases. When reimmunizing transplant recipients, recent administration of immunoglobulin preparations (except monoclonal antibody to RSV) must be considered because they interfere with the response to live virus vaccine (e.g., measles, mumps, rubella and varicella). Immunization can be started 12 months after HSCT. One schema is shown in Table 98-2. Transplant recipients and their household contacts should only be given inactivated poliovirus vaccine (IPV). Inactivated influenza vaccine should be administered annually in early autumn to recipients of HSCT beginning 6 months after transplantation. The live attenuated influenza vaccine should not be administered to HSCT recipients. Varicella vaccine is contraindicated until 24 months after HSCT and is restricted to research protocols 24 months or longer after HSCT in patients who are presumed to be immunocompetent.5,55–57 Vaccination of donors prior to transplant and to recipients in the peritransplant period is being investigated in clinical trials.58 There is one study on which to base standard recommendations currently for the use of meningococcal conjugate vaccine,59 and no study on acellular pertussis vaccine for adults and adolescents. These vaccines would be expected to be safe and effective as other protein or protein conjugate vaccines in these patients (see Table 98-2). Rotavirus vaccine is a live-virus vaccine indicated in healthy children with dose 1 given by 12 weeks

TABLE 98-2. Immunization Schedule for Hematopoietic Stem Cell Transplant Recipients Months after Transplant

Vaccines

12

Td,a IPV, Hib, pneumococcal,b meningococcalc and hepatitis B Td,a IPV, Hib, and hepatitis B MMR,d varicella,e Td,a IPV, Hib, pneumococcal, and hepatitis B

14 24

DTaP, diphtheria and tetanus toxoids and acellular pertussis; GvHD, graft-versushost disease; Hib, Haemophilus influenzae type B; IPV, poliovirus vaccine inactivated; MMR, measles, mumps, rubella; Td, diphtheria and tetanus toxoids. a DTaP or DT if patient < 7 years of age; Tdap, if ≥ 11 years of age and indicated. b 23-valent pneumococcal vaccine; use of 7-valent conjugate vaccine is considered depending on age and indication (see Chapters 7, Active Immunization, and 123, Streptococcus pneumoniae). c Meningococcal conjugate vaccine if indicated by age. d Do not use live virus vaccines in patients with chronic GvHD or in patients receiving corticosteroid therapy. A second dose of MMR should be given 4 weeks or more after the first dose if there is no serologic response to measles after the first dose. e Restricted to research protocols.

of age and the last dose by 32 weeks of age. Rotavirus vaccine thus could not be given post-HSCT. Immunogenicity of a roughly similar vaccine schedule has been shown in a UK study; in that study immunizations are begun at 12 months post identical sibling, syngeneic and autologous HCSTS and 18 months after allogeneic HSCT, with MMR given 6 months after first series.59

Passive Immunization and Immunoglobulin Utilization IGIV, because of its antimicrobial and immunomodulatory effect, has been administered to HSCT recipients with mixed results.60,61 Differences in dosage, schedule, duration, and preparation used in various studies may contribute to ambiguity of results. IGIV has been reported to decrease the rate of infection and incidence of GvHD in allogeneic transplant recipients.62 The optimal dose is not known. However, prolonged administration of IGIV can lead to delayed immune reconstitution and an increased incidence of infections after discontinuation. The indications for passive immunization with specific immune globulin preparations (hepatitis B, tetanus, and rabies) in patients post-HSCT are similar to those in otherwise healthy individuals. Passive immunization is recommended for susceptible people with known exposure to varicella55(see Chapter 205, VaricellaZoster Virus.)

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

SECTION

P

Infections and Cancer

CHAPTER

99

Fever and Granulocytopenia Andrew Y. Koh and Philip A. Pizzo

The most common cause of immune compromise in children is related to cancer and its treatment. Infection remains a major cause of morbidity in children with cancer, although mortality from infectious complications has diminished significantly in the past 35 years. The risk factors, etiologic organisms, pathogenesis, diagnosis, and management of these infections have been the subjects of investigation since the advent of cytotoxic therapy and the identification of chemotherapy-induced neutropenia as a primary predisposing factor. The use of empiric antimicrobial regimens in this patient population has evolved from the observation that febrile neutropenic patients with cancer who had potentially fatal infections could not be distinguished from those who had less serious illnesses.

ETIOLOGIC AGENTS Bacteria Infections in immunocompromised children can result from bacteria, fungi, viruses, or protozoa, but most significant infections arise from the host’s endogenous bacterial flora. Studies have documented that, with hospitalization and antibiotic therapy, there is a shift in normal flora to include potentially pathogenic gram-negative bacteria.1 Hospitalization can result in the acquisition of antibiotic-resistant organisms, especially gram-negative bacteria, but not all patients who become colonized with these organisms experience clinical infections. In fact, the relative incidence of gram-negative infections has decreased since the early 1980s. During this same period, the incidence of gram-positive infections in this population has been rising,2,3 with coagulase-negative staphylococci and streptococci becoming more important.4 Although oral antibiotic regimens for prophylaxis against gram-negative bacillary septicemia (especially fluoroquinolone agents) and the growing use of indwelling venous catheters are risk factors for these infections, some changes antedated these evolving medical practices.5 Currently, the gram-positive organisms (particularly coagulase-negative staphylococci, Staphylococcus aureus, a-hemolytic streptococci, enterococci, and Corynebacterium spp.) account for more than half of documented infections in patients with cancer, and Escherichia coli, Pseudomonas, and the Enterobacter–Klebsiella–Serratia group are identified in a smaller but substantial proportion of patients.2,3,6,7 Although bacteremia with the gram-positive organisms is generally associated with lower mortality than that associated with bacteria with gram-negative organisms, the syndrome of a-hemolytic streptococcal septicemia deserves special attention. The a-hemolytic streptococci are normal inhabitants of the oral cavity and have been noted to cause infection in patients receiving cytosine arabinoside or other agents that predispose to the development of severe mucositis in the mouth and gastrointestinal tract.8 Streptococcus mitis and S. sanguis are the most

common a-hemolytic streptococci associated with this syndrome, which manifests as septic shock, adult respiratory distress syndrome, and rapid progression to death.9 Rare reports of secondary myositis have also been reported with a-hemolytic streptococcal bacteremia.10 Of particular concern are reports of penicillin resistance among these previously susceptible organisms.7, 11 Increasing resistance of enterococci, particularly Enterococcus faecium, to ampicillin and vancomycin raises the specter of even greater morbidity and mortality due to the gram-positive bacteria in the future. Traditionally, methicillinresistant Staphylococcus aureus (MRSA) has been considered a healthcare-associated pathogen; however, recently MRSA has emerged in adult and pediatric patients without established risk factors: community-acquired MRSA (CA-MRSA).12–18 CA-MRSA tends to cause localized skin and soft-tissue infections, although more invasive disease does occur (i.e., sepsis,16 necrotizing fasciitis,19 and pneumonia20). CA-MRSA isolates are often likely to be susceptible to nonbeta-lactam antibiotics such as clindamycin, trimethoprimsulfamethaxazole, and rifampin. However, inducible macrolidelincosamide-streptogramin resistance in a subset of CA-MRSA isolates14 increasing resistance to clindaymcin17 potentially limits the effectiveness of this agent. Increasing antibiotic resistance has also been observed among gram-negative organisms. Resistance to multiple agents, including extended-spectrum penicillins and cephalosporins, carbapenems, aminoglycosides, and quinolones, is reported. Of particular concern are Enterobacter and Serratia species, which are prone to rapid development of resistance due to inducible b-lactamases. The pattern of antibiotic resistance as well as the relative distribution of predominant organisms varies by medical center and should be considered in the empiric choice of antimicrobial therapy for the newly febrile neutropenic patient. Anaerobic organisms are less commonly associated with bacteremia (< 5%) in febrile, neutropenic patients despite predominance in the normal flora. The most commonly isolated anaerobic organisms are Bacteroides spp. and Clostridium spp. These organisms have been associated with specific syndromes, such as peritonitis, abdominal or pelvic abscesses, and perirectal cellulitis. They can also contribute to infections of the oral cavity, especially necrotizing gingivitis. Clostridium septicum can cause a devastating infection characterized by septic shock and rapidly progressive necrotizing fasciitis with myonecrosis. Infection arises from a traumatic or surgical wound or spontaneously from necrotic bowel. Pseudomembranous colitis caused by toxins of C. difficile can result in a wide spectrum of clinical manifestations, ranging from mild diarrhea and cramping to toxic megacolon and intestinal perforation.

Fungi With improved treatment and better survival from bacterial infections in febrile neutropenic patients, prevention and treatment of fungal infections have assumed greater importance, especially in patients with prolonged neutropenia (i.e., neutropenia lasting > 10 days). These infections tend to occur as secondary infections, rarely being identified at the onset of a febrile episode. Most fungal infections in children with cancer are caused by Candida species, with C. albicans accounting for most isolates.21 In the last 20 years, other Candida species (e.g., C. tropicalis, C. parapsilosis, C. krusei, and C. (Torulopsis) glabrata) have become more 563

564

SECTION

P Infections and Cancer

common. This changing pattern of infections may be due, in part, to the use of oral antifungal agents such as fluconazole.22 Aspergillus spp. remain the second most common cause of invasive fungal infections in neutropenic patients in most hospitals: A. fumigatus and A. flavus are the most common species isolated. These organisms are commonly environmental contaminants, and clusters of cases have been linked to environmental exposure (e.g., via air conditioning units or during construction). Other fungi that cause serious infection in neutropenic children are Mucor spp., Fusarium, Trichosporon, dematiaceous molds, and the pheohyphomycoses (including Curvularia, Bipolaris, Alternaria, and Exserohilum). None of these organisms is part of the normal flora of the respiratory or gastrointestinal tract, but after colonization, they can become invasive when chemotherapy renders the host susceptible. Yeasts such as Histoplasma, Cryptococcus, and Coccidioides are more commonly associated with defects of cell-mediated immunity, but they can occur in neutropenic patients as primary infection or as reactivation disease. Another major fungal pathogen in children with cancer is Pneumocystis jirovecii (P. carinii).23 In patients with cancer, P. jirovecii infection is generally due to reactivation of latent organisms and appears to be associated with a more rapidly progressive pneumonia than that seen in adults and older children infected with human immunodeficiency virus (HIV).24 The risk of PneumoCystis Pneumonia (PCP) is related to the use of corticosteroids, a feature that explains its occurrence in children with leukemia, lymphomas, and brain tumors. The temporal occurrence of PCP is strongly associated with the tapering and discontinuation of steroids.25,26

week. The rate approaches 100% when newer, more dose-intensive chemotherapeutic regimens result in periods of neutropenia for 2 weeks or longer.30 “Tumor fever,” resulting from cellular necrosis, is uncommon in children with leukemia but can occur in up to 25% of those with solid tumors. Although fever can sometimes be a manifestation of the cancer itself, it is far more likely to reflect underlying, often cryptogenic infection. Studies completed in the late 1970s suggested that more than half of children with fever and neutropenia had a clinically or microbiologically proven infection and that a causative agent could be documented in almost two-thirds of infectious episodes.31 In more recent studies, the incidence of documented infection has been lower. Currently, fever without apparent cause accounts for approximately two-thirds of febrile episodes, perhaps because rapid administration of broad-spectrum antibiotic therapy empirically masks documentation of infection. Concurrently, the mortality among febrile neutropenic patients has dramatically diminished over the last three decades. Several large series have noted fatal outcomes in only 1% to 5% of patients with fever and neutropenia.2,32–35 An attempt has been made to stratify patients into groups according to relative risk of infection and to identify other parameters that could predict the development of infectious complications. The need to identify children at low risk of serious infection has assumed greater importance as changes in supportive care and medical economics have shifted the emphasis to outpatient management and the rates of documented infections in febrile neutropenic patients have decreased.

Viruses

PATHOGENESIS AND ETIOLOGY

Early studies of febrile, neutropenic patients probably underestimated the pathogenic role of viruses, because techniques for culture and identification of these organisms were relatively primitive. Development of rapid and reliable methods of isolation over the last decade has facilitated a better understanding of their impact on children with cancer. Herpes simplex virus and varicella-zoster virus have consistently been identified as the most common viral pathogens in children with leukemia. Influenza, parainfluenza, cytomegalovirus, human herpesviruses 6 and 8, adenovirus, measles, respiratory syncytial virus, and enteroviruses can also be associated with substantial morbidity. The incidence of viral infections is higher in children undergoing induction or during relapse than during remission.27 Although some viral illnesses (e.g., dermatomal zoster or herpetic stomatitis) cause clinically recognizable lesions that are accessible to culture, others manifest as nonspecific symptoms or localize to areas that are less accessible to direct diagnostic studies (e.g., cytomegalovirus hepatitis). Common respiratory tract viruses, especially respiratory syncytial virus, parainfluenza virus, and influenza virus, cause lower respiratory tract infection and substantial morbidity in immunocompromised patients. As new antiviral agents become available and rapid diagnostic methods expand, identification of these organisms offers an important opportunity for specific intervention.

The interactions of complex factors in cancer and its treatment provide the basis for the higher risk for serious infection in patients undergoing cancer treatment. Sometimes, specific alterations in host defense mechanisms contribute to an increase in risk for certain viral, bacterial, fungal, and parasitic infections. In most cases, however, children receiving cytotoxic chemotherapy have multiple defects in immune responsiveness that contribute to the infectious complications (Tables 99-1 to 99-4).

Parasites Fortunately, parasites such as Toxoplasma gondii and Cryptosporidium are not common causes of infection in children with cancer, because resulting illness can be severe. T. gondii can cause fulminant disseminated disease but most often is localized to the central nervous system (CNS); stem cell transplant patients are at the highest risk of developing CNS toxoplasmosis.28 Cryptosporidium should always be considered in any pediatric cancer patient with severe or prolonged diarrhea.29

EPIDEMIOLOGY The syndrome of fever and neutropenia in a child undergoing cancer chemotherapy is common, occurring in about one-third of patients whose neutrophil counts fall below 500 cells/mm3 for more than 1

Anatomic Disruptions and Devices Alterations of skin or mucosal integrity or obstruction of an organ or body cavity predispose to infection (see Table 99-1). Cytotoxic chemotherapy damages the gastrointestinal tract mucosa, facilitating local flora access to the bloodstream. Mucositis has been implicated as a risk factor for development of infections, with the severity of mucosal disruption perhaps more important than its duration. Dose-intensive chemotherapeutic regimens that cause severe mucositis are associated with almost universal occurrence of febrile episodes and a high incidence of systemic bacterial infections, despite relatively short periods of neutropenia. Similarly, invasion of mucosal surfaces or obstruction of a viscus (e.g., by tumor mass, lymph nodes, or surgical scarring) can lead to stasis and proliferation of bacteria, with translocation of intestinal bacteria across the bowel wall, or to intestinal perforation. The recombinant human keratinocyte growth factor, palifermin, has been shown to reduce the duration and severity of oral mucositis after intensive chemotherapy and radiotherapy, resulting in a lower incidence of bloodborne infections than that in placebo recipients.36 Devices that violate skin integrity, such as central venous catheters (CVCs) and peripherally inserted central catheters (PICCs), also predispose to infection. One report estimated a four-fold increase in incidence of bacteremia in neutropenic patients who had catheters compared with those who did not, and a 40-fold increase in nonneutropenic patients compared with matched patients who did not.37 Depending on patient population, techniques of catheter insertion and care, treatment regimens, and the definition of CVCrelated infections, such complications are identified in 9% to 80% of patients.30

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Fever and Granulocytopenia

CHAPTER

99

565

TABLE 99-1. Predominant Pathogens Infecting Children with Anatomic Disruptions Site

Bacteria

Fungi

Other

Oral cavity

a-Hemolytic streptococci Oral anaerobes: Peptococcus, Peptostreptococcus

Candida spp.

Herpes simplex virus

Esophagus

Staphylococci Other colonizing organisms

Candida spp.

Herpes simplex virus Cytomegalovirus

Lower gastrointestinal tract

Gram-positive: group D streptococci Gram-negative: enteric organisms Anaerobes: Bacteroides fragilis, Clostridium perfringens

Candida spp.

Strongyloides stercoralis

Skin (intravenous catheter)

Gram-positive: staphylococci, streptococci, corynebacteria, Bacillus spp. Gram-negative: Pseudomonas aeruginosa, enteric organisms

Candida spp. Aspergillus spp. Malassezia furfur

Urinary tract

Gram-positive: group D streptococci Gram-negative: enteric organisms, Pseudomonas aeruginosa

Candida spp.

TABLE 99-2. Predominant Pathogens Infecting Children with Neutropenia Category

Organisms

Bacteria

Gram-negative enteric organisms Escherichia coli, Klebsiella pneumoniae, Enterobacter spp., Citrobacter spp., Pseudomonas aeruginosa, Bacteroides spp. Gram-positive Staphylococci: coagulase-negative, coagulase-positive Streptococci: group D, a-hemolytic, anaerobic Clostridia

Fungi

Candida spp. (Candida albicans, Candida tropicalis, other species) Aspergillus spp. (Aspergillus fumigatus, Aspergillus flavus)

TABLE 99-3. Predominant Pathogens Infecting Children with Defects in Cell-Mediated Immunity Bacteria

Fungi

Viruses

Other

Legionella Nocardia asteroides Salmonella spp. Mycobacteria Mycobacterium tuberculosis and nontuberculous mycobacteria Disseminated bacille Calmette-Guérin

Cryptococcus neoformans Histoplasma capsulatum Coccidioides immitis Candida spp. Pneumocystis jirovecii

Varicella-zoster virus Herpes simplex virus Cytomegalovirus Epstein–Barr virus Hepatitis B

Toxoplasma gondii Cryptosporidium Strongyloides stercoralis Disseminated infection from live virus vaccines (vaccinia, measles, rubella, mumps, yellow fever, or poliovirus)

Because CVCs facilitate delivery of chemotherapy and supportive therapy, their benefits to the patient generally outweigh the infectious risks. Use of subcutaneously implanted venous access devices has become common in the management of children with cancer. Although some studies have suggested that the incidence of infection is lower for patients with totally implanted devices (e.g., Port-A-Cath) compared with external, tunneled CVCs (Hickman-Broviac-type catheters),38–40 the only prospective, randomized study failed to document a difference in rates of infection.41 The use of central catheters for delivery of parenteral nutrition further raises the risk of infection, independent of the type of catheter used, by about 2.4-fold according to a multivariate analysis performed at one pediatric oncology center.42 Foreign devices other than CVCs have also been implicated in the risk of focal and disseminated infection in immunocompromised

patients. Among the most common of these devices are Ommaya (intraventricular) reservoirs, used to deliver chemotherapeutic agents directly into the ventricular space in patients with malignancies of the CNS. A review of infections associated with these devices in 61 patients at the National Cancer Institute (NCI) revealed that 75% never experienced a device-related infectious complication. Unlike previously published series of ventricular shunt infections in children with hydrocephalus, this study reported no mortality associated with Ommaya reservoir infections, and many of the infections were successfully treated without removal of the device.43 Limb-sparing procedures for patients with osteosarcoma utilize prosthetic bone–joint devices that can also be associated with infections. Such infections can follow an episode of bacteremia, or the prosthesis can be the primary site of infection after minor trauma. In some institutions, children with such prosthetic devices are not

566

SECTION

P Infections and Cancer

TABLE 99-4. Predominant Pathogens Infecting Children with Defects in Immunoglobulins, Complement, and Splenic Functiona Defect

Organisms

Immunoglobulin abnormalities Gram-positive Streptococcus pneumoniae Staphylococcus aureus Gram-negative Haemophilus influenzae Neisseria spp. Enteric organisms Viruses Enteroviruses (including polioviruses) Protozoa Giardia lamblia Complement abnormalities C3–C5

C5–C9

Splenectomy

Gram-positive Streptococcus pneumoniae Staphylococcus spp. Gram-negative Haemophilus influenzae Neisseria spp. Enteric organisms Neisseria spp. Neisseria gonorrhoeae Neisseria meningitides Gram-positive Streptococcus pneumoniae DF2 bacillus Gram-negative Haemophilus influenzae Salmonella spp. Babesia

a

See also Chapter 104, Infectious Complications of Antibody Deficiency; Chapter 105, Infectious Complications of Complement Deficiencies; Chapter 108, Infectious Complications in Special Hosts.

considered candidates for permanent CVCs because of the potential higher risk of bacteremia and seeding of the prosthesis. One retrospective study showed that, in children and adolescents with bone malignancies who underwent limb-sparing surgery, focal bacterial infections developed in 67% and bacteremia in 21% of patients.44 The relative risk of infection in children with limb prostheses is not known, and optimal management has not been defined.

Neutropenia The most important risk factor in the development of infections in children with cancer is chemotherapy-induced granulocytopenia. The relationship between neutrophil numbers and the risk of infectious complications in patients with leukemia was first described in 1966 by Bodey and colleagues45 at the NCI. Following a cohort of children and adults with leukemia throughout chemotherapy, recording blood counts, and identifying infectious complications, the investigators concluded that: (1) the risk of infection was directly related to the absolute neutrophil count (ANC), severe infections being more prevalent when the ANC fell below 100 cells/mm3; (2) relapse of leukemia was associated with higher rates of infection than remission at all levels of neutrophil count; and (3) duration of neutropenia was the single most important factor in predicting risk of infection. Severe neutropenia that lasted longer than 3 weeks was associated with 100% risk of infection and the highest mortality rates.45 These observations were confirmed in numerous studies and led to the current approach to management of patients with fever and neutropenia. Organisms causing infections in children with neutropenia are listed in Table 99-2.

Other Defects The observation that some children undergoing cancer therapy experienced infections such as PCP, typically associated with defects in cell-mediated immunity, led investigators to evaluate multiple aspects of chemotherapy-induced immune suppression. Repeated cycles of cytotoxic therapy not only decrease circulating neutrophils but also deplete lymphocytes. Although neutrophils, monocytes, and platelets recover to nearly normal numbers between cycles of chemotherapy, CD4+ and CD8+ lymphocyte populations progressively decrease and remain deficient for several months after completion of chemotherapy. In one series, reduced lymphocyte subset populations were linked to the occurrence of opportunistic infections in a group of patients receiving dose-intensive chemotherapeutic regimens.46 Organisms causing infections in children related to defects in cell-mediated immunity are listed in Table 99-3, and those related to immunoglobulin and complement abnormalities as well as splenic compromise are listed in Table 99-4. The pathogenesis of specific infectious clinical syndromes is discussed in Chapter 100, Infections in Children with Cancer.

DIFFERENTIAL DIAGNOSIS AND CLINICAL APPROACH TO DIAGNOSIS Myriad infectious or noninfectious entities can be responsible for fever in a child with chemotherapy-induced neutropenia. Because of the potential for life-threatening infection, each patient must be evaluated expeditiously and repeatedly and treated promptly with empirical broad-spectrum antimicrobial therapy. Fever can be a sign of localized or disseminated infection, a manifestation of cancer, or a side effect of a chemotherapeutic agent. Many oncology centers have adopted a standardized approach to the evaluation of febrile, neutropenic patients to ensure that all potentially dangerous causes of fever are considered (Box 99-1). For example, it is reasonable to define neutropenia as an existing ANC of fewer than 500 neutrophils/mm3 or an expected count at that level within 24 hours. Fever is defined as a single oral-equivalent temperature of 38.5°C (101.3°F) or higher, or a series of three temperatures recorded above 38.0°C (100.4°F) within a 24-hour period. Documentation of fever by the patient or a family member should always be accepted, even if the child is afebrile at the time of medical evaluation. It is also important to remember that some lifethreatening infections can occur in the absence of fever in profoundly neutropenic patients with localizing signs or symptoms (e.g., right lower quadrant abdominal pain). A careful physical examination, including evaluation of the anal cavity and perirectal area, must be performed at the time of presentation. Unfortunately, initial examinations often do not identify the site of infection. Repeating the physical exam at least daily as long as the patient remains febrile and neutropenic can help define occult sites of infection. However, the traditional markers of inflammation can be muted or even absent in children who lack neutrophils, and thus the usual manifestations of serious infection, such as chills, rigors, and “toxic appearance,” may not help distinguish patients with bacteremia or other serious infection from those with an unexplained fever. Specimens for culture should be obtained from any suspicious sites, urine, and blood. At least two blood cultures should be obtained from all patients before institution of antibiotic therapy. If the patient has a PICC or CVC, blood culture specimens should be obtained from all accessible lumens of the device as well as from a peripheral vein. The site of a catheter-related infection can be localized to one lumen of a multiluminal device. During seasonal outbreaks of viral infection of the respiratory tract, a throat or nasopharyngeal specimen should be obtained for antigen detection or culture. A chest radiograph is frequently obtained at the time of initial evaluation, but the utility of this practice has been challenged.47–49 One prospective pediatric study showed that the only children with abnormal chest radiographs were those who also had abnormal

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Fever and Granulocytopenia

BOX 99-1. Initial Diagnostic Evaluation of Children with Fever and Neutropenia • Careful medical history • Physical examination with attention to the skin, perirectal area, and other mucosal sites (which should be repeated at least daily in febrile neutropenic patients) • Specimens for culture • Blood: peripheral venipuncture and specimens from every lumen or access port of intravascular catheters • Urine • Respiratory secretions: for bacteria and viruses if symptoms are present • Any site with clinical signs of infection, including Clostridium difficile toxin assay if diarrhea is present • Chest radiograph • Other imaging studies or diagnostic procedures as clinically indicated • Sinus radiograph or computed tomography • Abdominal ultrasonography or computed tomography

respiratory findings.50 Chest radiograph at time of initial evaluation, however, can establish a baseline for future comparison in patients anticipated to have prolonged neutropenia (e.g., > 7 to 10 days). Pulmonary infiltrates not present initially may become apparent with bone marrow recovery, because neutrophils are recruited to a site of previously “silent” infection. Multiple studies have investigated the value of surveillance cultures in patients undergoing chemotherapy. Although colonization generally precedes infection, knowledge of colonization has little clinical benefit. In a survey of 652 episodes of fever and neutropenia at the NCI in which surveillance cultures of nose, throat, urine, and stool were performed, 62% of patients who became bacteremic were found to be colonized with the infecting organism. However, management of these patients was not influenced by this information because multiple potential pathogens were frequently isolated from the same site, no single site was predictive of bacteremia, and the blood isolate was frequently identified before results of the surveillance cultures were available.51 The cost of routine surveillance cultures is not justifiable. Surveillance cultures may be useful in subsets of patients, such as those who have had procedures that increase risk of infection (e.g., nephrostomy, placement of urinary conduit) and for patients at centers experiencing a high rate of infections with a resistant or highly virulent organism (such as resistant Enterococcus, Pseudomonas, or Aspergillus).

MANAGEMENT AND PRESUMPTIVE THERAPY One-third of patients with fever and neutropenia have an identifiable source of infection at the time they seek medical attention. Empiric antibiotic therapy for febrile neutropenic episodes has been the standard of care since the 1960s, when it became clear that unsuspected bacteremia in such patients could be rapidly fatal, even before the causative organism was identified. An empiric regimen should: (1) provide a broad spectrum of activity against a variety of pathogenic organisms, including Pseudomonas; (2) be bactericidal in the absence of neutrophils; and (3) have low potential for adverse effects or emergence of resistant organisms.

CHAPTER

99

567

antimicrobial agents in neutropenic adults with unexplained fever that outline several empiric regimens.3 For many institutions, the “standard” antibiotic regimen for a child admitted with fever and neutropenia continues to include an aminoglycoside and an antipseudomonal b-lactam antimicrobial agent (either an extended-spectrum penicillin or a third-generation cephalosporin). These regimens were developed in the 1970s, when Pseudomonas and other gram-negative organisms were the predominant isolates from febrile patients with cancer. They have proved to be effective and generally well tolerated, although the use of aminoglycosides necessitates careful monitoring of renal function and serum drug concentrations. The potential for nephrotoxicity and ototoxicity has led many investigators to search for regimens with better safety profiles, because several antineoplastic agents (such as cis-platinum) also cause renal toxicity. A recent meta-analysis of randomized controlled trials that compared the ciprofloxacin/b-lactam combination versus an aminoglycoside/b-lactam combination for the empiric treatment of febrile neutropenia showed comparable or better outcomes with the ciprofloxacin/b-lactam combination. The authors of this analysis emphasized that this combination of quinolone/ b-lactam should only be considered for patients who have not received a quinolone for prevention of infections and in settings in which quinolone resistance is not common.52 Double b-lactam regimens (usually consisting of an extendedspectrum penicillin and a third-generation cephalosporin) have the advantage of lower toxicity. The extended-spectrum penicillins with beta-lactamase inhibitor have good activity against many enterococci and anaerobic bacteria. The major drawback of these regimens is the potential for emergence of b-lactam-resistant gram-negative bacteria. Monotherapy (using the broad-spectrum, third-generation cephalosporin ceftazidime) was pioneered by the NCI in the interest of simplifying antibiotic regimens and reducing toxicity in children who are often receiving complex chemotherapeutic protocols. Monotherapy had equivalent efficacy (i.e., survival of patients with fever and neutropenia) compared with combination regimens regardless of whether a site of infection could be documented. However, many patients required modifications of the initial regimen if they had a documented site of infection or more than 7 days of neutropenia. Additional studies evaluating other cephalosporins (e.g., cefepime53–55), imipenem-cilastatin,32,56 and meropenem,57–59 as single agents for fever and neutropenia show roughly comparable efficacy and provide options for management. Neither the third-generation cephalosporins nor the carbapenems are effective against MRSA and other selected gram-positive bacteria. The increasing frequency of antibiotic-resistant a-hemolytic streptococci,7,60,61 S. aureus, coagulase-negative staphylococci, Corynebacterium jeikeium, and enterococci as causes of bacteremia in neutropenic patients with cancer has led to the inclusion of vancomycin in empiric regimens at some centers. At the NCI, and in randomized trials conducted by the European Organization for Research and Treatment of Cancer before the advent of CA-MRSA, delayed administration of vancomycin (i.e., at the time of isolation of the pathogen) had no adverse effect on outcome.62,63 Empiric use of vancomycin, especially in closed units, heightens the risk of resistance. The decision to administer vancomycin as part of the initial empiric management should be based on regional resistance patterns, the presence of localized infection likely to be caused by grampositive bacteria, and increased risk for potentially resistant a-hemolytic streptococcal infections (e.g., in patients who have received continuous or high-dose cytosine arabinoside).

Empiric Therapeutic Regimens A number of empiric regimens have been evaluated. Most large comparative clinical efficacy trials have included both adults and children. There is little evidence that children respond differently to the regimens that have been investigated. Many of the regimens studied are equivalent in their efficacy, although study designs and definitions of success have not been uniform. The Infectious Diseases Society of America has published general guidelines for use of

Duration and Modification of Therapy Even when a bacterial pathogen is isolated from a febrile, neutropenic child, broad-spectrum antibiotic therapy should be continued.64 Specific clinical events that may require modification of the initial treatment regimen are described in Chapter 100, Infections in Children with Cancer; recommendations are summarized in Table 99-5.

568

SECTION

P Infections and Cancer

TABLE 99-5. Modifications of Antimicrobial Therapy During Fever and Neutropenia Clinical Event

Possible Modifications in Therapy

Breakthrough bacteremia

If gram-positive isolate, add vancomycin If gram-negative isolate (presumably resistant), change regimen

Catheter-associated soft-tissue infection

Add vancomycin (and gram-negative coverage, if not already being given)

Severe oral mucositis or necrotizing gingivitis

Add agent active against b-lactamase-producing anaerobic bacteria (clindamycin, metronidazole); consider acyclovir

Esophagitis

Clotrimazole, ketoconazole, fluconazole, voriconazole, caspofungin, or amphotericin B; consider acyclovir

Diffuse or interstitial pneumonitis

Trimethoprim-sulfamethoxazole and erythromycin

New infiltrate in neutropenic patient on antibiotics

If neutrophil count rising, can observe If neutrophil count not recovering, pursue biopsy; if biopsy not possible, add amphotericin B

Perianal cellulitis

Add agent active against b-lactamase-producing anaerobic bacteria to broad-spectrum therapy

Prolonged fever and neutropenia

After 1 week of antibiotics, add amphotericin B

The duration of antibiotic therapy depends on a number of factors, including isolation of a pathogen, clinical identification of a presumed infectious process, duration of both fever and neutropenia, and the patient’s schedule for chemotherapy3 (Figure 99-1). For patients in whom infection has been documented (either microbiologically or clinically) and resolution of fever and neutropenia has been prompt, the course of antibiotic treatment should be determined by the specific type of infection identified. For patients without a defined site of infection, antibiotics can generally be discontinued, within 2 to 5 days of attaining an ANC of approximately 500/mm3. For patients who remain neutropenic and without an obvious site of infection, recommended duration of antibiotic therapy depends upon the course of fever. If fever resolves, antibiotics may be discontinued 5 to 7 days later, even in the presence of continued neutropenia.3 However, some experts would continue antibiotics for a longer duration, in the hope of preventing recrudescence of fever due to persistent or recurrent infection. If fever persists in the presence of continued neutropenia, antibiotics should be continued for at least 2 more weeks and discontinued if no site of disease is found.3

Fever and neutropenia

Empiric antibiotic Rx

Infection documented

Yes

No

Rx

Febrile at 3 days

No

Yes

WBC >500

WBC >500

Empiric Antifungal Therapy Fungi have emerged as an important cause of superinfections in patients with prolonged neutropenia and may affect 9% to 31% of this population.65,66 Traditionally, amphotericin B was the only available systemic antifungal agent. Its use in empirical regimens for prolonged or recurrent fever in neutropenic patients reduced the incidence of documented fungal infections and attributable mortality.65 It has become standard practice to administer amphotericin B empirically in patients who remain neutropenic and have persistent or recurrent fever after 4 to 7 days of antibiotic therapy. One-quarter of all modifications to empirical antibiotic regimens in a large multicenter trial involved the addition of systemic antifungal agents; fungi were responsible for 7 bloodstream and 21 lung infections (causing 12 deaths) among 784 febrile episodes.33 A dose of 0.5 mg/kg per day of amphotericin B is used routinely for empirical antifungal therapy at the NCI and many other centers. However, doses of 1.0 to 1.5 mg/kg per day are recommended if there is a substantial risk of infection with Aspergillus spp., such as in patients with protracted neutropenia.3,67 The search for alternative antifungal agents has been driven by the substantial toxicity of amphotericin B and the increasing prevalence of drug resistance. The azoles represent a less toxic class of antifungal agents. Ketoconazole, fluconazole, itraconazole, and voriconazole have all been effective in treatment of mucosal candidiasis. Although fluconazole has been reported to be as effective as amphotericin in

No

Yes

No

Yes

Stop antibiotics when afebrile for ≥48 hrs

Continue antibiotics for at least 7 days

Stop antibiotics 4–5 days after WBC >500

Continue antibiotics at least 2 weeks, adding empiric antifungal after 4–7 days

Figure 99-1. Decision tree for management of antimicrobial therapy in children with cancer, fever, and neutropenia. WBC, white blood cell count/mm3.

the treatment of candidemia in patients without neutropenia or other major immunodeficiency condition,68 efficacy is less well established in febrile and neutropenic patients with cancer. Two prospective studies suggest that fluconazole is an equally effective but less toxic alternative to amphotericin B when given as empirical antifungal

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Fever and Granulocytopenia

therapy in patients with cancer who have prolonged fever and neutropenia.69,70 The azoles, however, may be less active than amphotericin B against some species. Fluconazole has no activity against Aspergillus and has less activity than amphotericin B against Candida tropicalis, C. krusei, C. lusitaniae, and C. glabrata. Itraconazole and voriconazole have activity against Candida spp., Aspergillus spp., and some of the less common fungi. Voriconazole compared favorably with liposomal amphotericin B when used empirically in adult cancer patients with persistent fever and neutropenia.71 None of the azoles has significant activity against the etiologic agents of mucormycosis. There is considerable interest in preparations of liposomal or lipidassociated amphotericin because of lower toxicity. Open-label phase I and II studies in patients with cancer and neutropenia suggested that liposomal amphotericin B (AmBisome) was well tolerated and had minimal nephrotoxicity.72,73 In the only randomized, double-blind trial comparing liposomal amphotericin with conventional amphotericin B as empirical antifungal therapy, the outcomes were similar with respect to survival, resolution of fever, and discontinuation of study drug because of toxic effects or lack of efficacy.74 Liposomal amphotericin B was also associated with fewer breakthrough fungal infections, less infusion-related toxicity, and less nephrotoxicity.74 In a follow-up pharmacoeconomic analysis of liposomal versus conventional amphotericin therapy, the acquisition costs of liposomal amphotericin B were substantially higher than those of conventional amphotericin B. However, when the costs of study drugs are excluded from the analysis, hospital costs were lower for use of liposomal amphotericin B, most likely because of the additional costs associated with management of nephrotoxicity of the conventional agent.75 Finally, the higher therapeutic index of liposomal amphotericin B allows the use of substantially higher doses than of conventional amphotericin – perhaps most relevant to the management of invasive aspergillosis. The only prospective randomized clinical trial comparing the efficacy of two dosages of liposomal amphotericin (1 mg/kg per day versus 4 mg/kg per day) for treatment of invasive aspergillosis showed that the higher dose was not more efficacious.76 The newest class of antifungal agents are the echinocandins, capsofungin and micafungin. These large lipopeptide molecules are inhibitors of beta-(1,3)-glucan synthesis, which is essential for fungal cell wall synthesis. The echinocandins are rapidly fungicidal against most Candida spp. and fungistatic against Aspergillus spp., but they are not active against Zygomycetes, Cryptococcus neoformans, or Fusarium spp.77 Since the drug target is not present in mammalian cells, adverse events are generally mild, including local phlebitis, fever, abnormal liver function tests, and mild hemolysis. Oral bioavailability is suboptimal, thereby limiting use to the intravenous route. In a prospective randomized, double-blind trial comparing the efficacy and safety of caspofungin with that of liposomal amphotericin B, caspofungin was found to be as effective as and generally better tolerated than liposomal amphotericin B when given as empirical antifungal therapy in patients with persistent fever and neutropenia.78 The optimal caspofungin dosing for children is not yet known. One small prospective study investigated the pharmacokinetics and safety of caspofungin in pediatric patients and demonstrated that a caspofungin dose of 50 mg/m2 per day provided comparable exposure to that of adult patients treated with 50 mg/day without developing any serious drug-related adverse events or toxicity.79

PROGNOSIS AND SEQUELAE The mortality associated with bacteremia and other infections during neutropenia has decreased substantially during the past 30 years. However, substantial morbidity, often attributed to secondary infections, or antimicrobial toxicity persists. Rates of microbiologically documented superinfections after initial therapy for fever and neutropenia range from 5% to 9% in several investigations of different empirical antibiotic regimens.32,33,35 The type and frequency of drug toxicities depend upon the specific agents. Regimens containing

CHAPTER

99

569

aminoglycosides carry the risk of nephrotoxicity (3% to 9%) or ototoxicity (1% to 8%);33,35 imipenem causes nausea and vomiting in up to 20% of patients;32 and rash, other allergic manifestations, and elevated serum hepatic transaminase levels have each been reported in about 5% of patients receiving a number of different regimens.

RECENT ADVANCES Strategies to simplify further the approach to the management of febrile, neutropenic patients continue to evolve. It has been hoped that stratifying patients by risk of infection might facilitate less aggressive management of those at low risk. Although there is no definitive consensus about the criteria used to distinguish high-risk from lowrisk patients, the following key factors that may raise the risk of infectious complications can be surmised from studies that have been conducted: (1) anticipated duration of neutropenia;80 (2) significant medical comorbidity;80,81 (3) cancer status and cancer type; (4) documented infection on presentation; (5) evidence of bone marrow recovery;82,83 and (6) magnitude of fever.2,82 One small study found that discharging well-appearing children in whom blood culture results were not positive and rapid defervescence occurred after short hospitalizations was both safe and cost-effective.84 In contrast, another study raised concerns about a less aggressive approach to febrile and neutropenic cancer patients considered to be at low risk for infection. Four of 30 patients who were discharged early to continue intravenous antibiotics at home had medical complications and another 5 required readmission for recurrent or prolonged fever.85 Investigations of oral antibiotic therapy have used many antibiotic regimens (pefloxacin,86 ofloxacin,87–91 ciprofloxacin,92–94 cefixime,95 moxifloxacin96) for empirical coverage of low-risk febrile and neutropenic patients with cancer. Although the results of these studies are encouraging, many of the clinical trials were statistically underpowered and limited by other methodologic issues. Two large, prospective randomized studies of low-risk patients have evaluated the relative efficacy of oral ciprofloxacin plus oral amoxicillin-clavulanate compared with parenteral antibiotic therapies in an inpatient setting.97, 98 Although the efficacies of oral and intravenous regimens were comparable in both of these clinical trials, limited experience precludes this practice as an established standard for treating low-risk patients. Low-risk criteria are not reliable in identifying uninfected patients. Although 58% of 775 episodes of fever and neutropenia occurring in cancer patients cared for at M.D. Anderson were unexplained, 21% of the episodes were associated with a clinically documented infection and an additional 21% were associated with microbiologically documented infections.99 Among microbiologically documented infections, gram-positive (49%), gram-negative (36%), and polymicrobial (15%) infections occurred. Thus, although these patients were considered at low risk for developing infectious complications, many developed serious infections necessitating broadspectrum antibiotic therapy. The use of hematopoietic growth factors, such as granulocyte– macrophage colony-stimulating factor (GM-CSF) and granulocyte colony-stimulating factor (G-CSF), has been widely implemented in the management of neutropenia in patients with cancer and in bone marrow transplant recipients. Both G-CSF and GM-CSF have been evaluated as adjuncts to chemotherapy to assist bone marrow reconstitution, minimize neutropenia, and reduce infectious complications. Primary use of G-CSF has been shown to lower the incidence of febrile neutropenic episodes by about 50% in three randomized studies in which the incidence of fever and neutropenia in the control group was > 40%. Primary administration of CSFs should be reserved for patients who are expected to experience rates of fever and neutropenia (≥ 40%) that are comparable to or greater than those seen in the control patients in these randomized trials. Use of GM-CSF has been less consistently helpful and associated with more side effects than G-CSF. No study to date has demonstrated an advantage in rates of tumor response, fatal infections, or overall survival. Secondary use of G-CSF or GM-CSF in subsequent cycles of

570

SECTION

P Infections and Cancer

chemotherapy has not demonstrated diseasefree or overall survival benefits when the dose of chemotherapy was maintained. Therefore, reduction in chemotherapy dose should be considered the primary therapeutic option in a patient who experiences neutropenic fever or severe or prolonged neutropenia after the previous cycle of treatment. If G-CSF is administered within the first few days of chemotherapy for the initial induction or first postremission course for acute lymphoblastic leukemia, the duration of neutropenia of < 1000/mm3 can be shortened by approximately 1 week. These studies and guidelines for the use of hematopoietic growth factors are summarized in a consensus paper published by the American Society of Clinical Oncology.100 Clinical trials have reported conflicting results when attempting to evaluate whether the addition of colony-stimulating factors (CSFs) to antibiotics in the treatment of fever and neutropenia improves outcomes. A meta-analysis of 13 studies revealed that the use of CSFs in patients with established fever and neutropenia reduces the amount of time spent in the hospital and the time to neutrophil recovery.101 Overall mortality was not influenced by use of CSFs, but a marginally significant decrease in infection-related mortality was noted.101

PREVENTION Preventive strategies have been evaluated, including use of hematopoietic growth factors, isolation measures, active and passive immunization, and prophylactic antimicrobial regimens. Strict isolation using laminar-airflow rooms and sterilization of everything entering the environment may offer some protection to profoundly neutropenic patients, including those who are undergoing bone marrow transplantation or have severe combined immunodeficiency syndrome. But the cost of this kind of program and the hardship it imposes on the patient make this approach less attractive for most oncology patients, for whom the benefit is less appreciable.

Active Immunization The American Academy of Pediatrics provides guidelines for use of immunizations in immunocompromised children.102 In general, livevirus vaccines are not recommended for use in immunosuppressed children, with the exception of the live-attenuated varicella vaccine; use of this vaccine under protocol can be recommended for children with acute lymphocytic leukemia in remission.103 Killed virus or subunit vaccines are considered safe for even the most compromised patients, but the efficacy in these individuals may be reduced. Children recovering from allogeneic bone marrow transplantation may acquire the immunity of their donor but generally should be regarded as unimmunized; they are candidates for revaccination with inactivated vaccines beginning at 12 to 18 months depending on transplant type.104 Live-virus vaccines (with the exception of the varicella vaccine) can be administered 2 years after successful transplantation if there is no evidence of graft-versus-host disease.

Passive Immunization Passive immunization with immunoglobulins has also been evaluated extensively. In certain circumstances, infection-specific or hyperimmune globulins have proven to be particularly helpful, as in the case of postexposure use of high-titer varicella-zoster immune globulin in susceptible individuals. Passive immunization reduces the incidence of primary varicella and its complications in nonimmune individuals if given within 72 hours of exposure. Because protection only lasts 3 to 4 weeks after administration, repeated doses may be necessary in the setting of community outbreaks of varicella and multiple episodes

of exposure. The use of immune globulin intravenous may also be beneficial in specific circumstances, such as for prophylaxis for measles in susceptible patients during epidemics.

Prophylactic Antimicrobial Therapy Many clinical trials have focused on the use of prophylactic antibiotics to prevent infections in neutropenic patients. Because most infections in neutropenic patients arise from the patient’s flora, attempts have been made to “decontaminate” the gastrointestinal tract with oral nonabsorbable antibiotics; these regimens have frequently been poorly tolerated. The oral fluoroquinolone antibiotics have been investigated because of their good bioavailability and broad activity against aerobic bacteria. A meta-analysis of published randomized controlled trials of quinolone prophylaxis (18 trials with 1408 patients) found that quinolone prophylaxis significantly reduced the incidence of gramnegative bacterial infections, microbiologically documented infections, total infections, and fevers but did not alter the incidence of gram-positive infections or infection-related deaths.105 A large randomized prospective trial evaluating the use of oral levofloxacin (500 mg daily) in adult patients with cancer (prior to an anticipated chemotherapy-induced neutropenia) showed that the levofloxacin group had a lower rate of microbiologically documented infections, bacteremia, and single-agent gram-negative bacteremia than did the placebo group. Mortality was similar in the two groups, and the effects of prophylaxis were similar between patients with acute leukemia and those with solid tumors or lymphomas.106 Two distinct studies evaluating the use of prophylactic oral levofloxacin (500 mg daily) in patients receiving chemotherapy for either solid tumors or lymphoma107 or hematologic malignancies108 both showed a reduction in documented infections. Finally, a meta-analysis of antibiotic prophylaxis in neutropenic cancer patients (95 trials performed between 1973 and 2004) concluded that antibiotic prophylaxis (various antibiotic regimens) significantly decreased the risk for death compared with placebo or no treatment.109 When trials that utilized quinolone prophylaxis (52 trials) were separately analyzed, there was a significant reduction in the risk for all-cause mortality, as well as infection-related mortality, fever, clinically documented infections, and microbiologically documented infections. Although these results are encouraging, many of these studies reported a rapid increase in the rate of antimicrobial resistance to quinolone agents.106,108,109 Two large trials using fluconazole as antifungal prophylaxis have been conducted in adult patients with leukemia and bone marrow transplant recipients.110,111 Although both studies confirmed a decrease in fungal colonization and superficial infections, a reduction in systemic fungal infections and associated mortality was only identified in the patients undergoing bone marrow transplantation. However, in several of the studies investigating fluconazole, infections with fungi known to be resistant to fluconazole (i.e., Aspergillus or other molds and non-albicans Candida spp.) or Candida species developing resistance have been a major cause of morbidity and mortality.69,112,113 The incidence of PCP in children with cancer can be significantly decreased by the use of trimethoprim-sulfamethoxazole.114,115 The incidence of recurrent herpes simplex disease in patients receiving intensive chemotherapy or bone marrow transplantation has been reduced with the prophylactic use of acyclovir.116 Similarly, use of ganciclovir has been shown to be effective in preventing cytomegalovirus disease after bone marrow transplantation.117 The goal of both basic science and clinical research in this area is to reduce the duration and intensity of immune dysfunction during the course of chemotherapy and thus avoid the risk of infectious morbidity and mortality. Use of cytokines and hematopoietic growth factors may ultimately allow reconstitution of the immune system and limit the need for medical intervention.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Infections in Children with Cancer

CHAPTER

10 0

Infections in Children with Cancer Andrew Y. Koh and Philip A. Pizzo

Although children receiving cytotoxic chemotherapy for treatment of cancer frequently develop febrile episodes during periods of neutropenia, only about one-third have a clinically or microbiologically proven site of infection. When a clinical or microbiologic site of infection is identified, addition to or modification of the empirical regimen may be necessary. Table 100-1 summarizes documented infections identified in three large clinical trials of empiric antibiotic therapy administered to patients admitted to the National Cancer Institute for fever during episodes of neutropenia.1–3

DIAGNOSIS AND MANAGEMENT OF SPECIFIC CLINICAL SYNDROMES OF INFECTION Catheter-Associated Bacteremia and Soft-Tissue Infections Several studies evaluating the use of central venous catheters in patients with cancer have confirmed that these devices increase the incidence of bacteremia, regardless of the level of bone marrow suppression.4 Although gram-positive organisms, particularly coagulasenegative staphylococci, have emerged as the predominant pathogens, almost any bacteria can be responsible for catheter-associated bacteremia. Treatment protocols that cause severe or prolonged mucositis can lead to bloodstream invasion by agent(s) from flora at that site. Fungi, most often Candida spp., can also cause catheter-related septicemia and are difficult to eradicate without removal of the contaminated device.5,6 Bacteremia with multiple organisms is simultaneously much less common and generally indicative of a significant break in sterility during line entry and routine care or of serious gut-associated infection. Mixed-flora bacteremia should prompt an investigation to identify a causative event (e.g., swimming with the catheter unprotected, dropping of catheter tubing into bathwater, chewing on catheter tubing by young children, disconnection of catheter caps or tubing) so that appropriate education can be provided. Because of the increased risk of catheter-related bacteremia, antibiotics may be warranted in nonneutropenic patients with central venous catheters who have fever with no localizing findings. In eva-

CHAPTER

100

luating children for presumed catheter-associated bacteremia, cultures should be obtained from all lumina of a multilumen device and administration of antibiotics should be rotated through all lumens because the pathogen is sometimes present in only one portion of the catheter. Most episodes of catheter-associated septicemia resulting from bacterial pathogens can be treated adequately with the catheter in place, even when multiple organisms are identified.4,7 The choice of agent for empiric antibiotic therapy in children with cancer should take into consideration those organisms most likely to cause catheterrelated infections and should provide broad-spectrum activity until an organism is identified.8 Severe clinical illness due to Staphylococcus aureus or Pseudomonas aeruginosa infection often requires immediate removal. Specific organisms that are infrequently eradicated and usually require removal of the catheter include a number of Bacillus spp.,9 the rapidly growing mycobacteria (Mycobacterium chelonae and M. fortuitum),10 Candida spp.,5,6 and vancomycinresistant enterococci. Polymicrobial infections also usually require removal of the catheter. Other instances that may necessitate removal of the catheter include severe illness associated with any pathogen or persistently positive results on blood cultures from an infected catheter after 48 hours of appropriate antimicrobial therapy.11 In one prospective study at the time of removal of implantable ports, 50% of children had deep venous thrombosis at the site, often without symptoms. Thrombosis did not appear to affect risk or outcome of bacteremia in these children.12 Bacteremia can occur in conjunction with a catheter-related softtissue infection or without localizing findings. These infections may be confined to the skin immediately surrounding the exit site of a catheter or can involve deeper soft tissue around the subcutaneously tunneled portion of the catheter or access port. Exit site infections without evidence of purulent discharge or deep subcutaneous involvement can sometimes be managed with oral antibiotics in children who do not have neutropenia. If Pseudomonas species is isolated from the site, catheter removal is usually required. Erythema, tenderness, or purulent drainage from the site necessitates intravenous antibiotics. Any localizing findings in a neutropenic child should be considered evidence of infection, mandating hospitalization and administration of broad-spectrum antibiotics. Induration, erythema, tenderness, or fluctuance along the subcutaneous tunnel tract of the catheter generally requires removal of the catheter (in addition to antibiotic therapy) and debridement of infected tissue. Inclusion of an empiric antistaphylococcal agent is recommended in the presence of a catheter-related soft-tissue infection.

Skin Infections Cutaneous infections are common in immunocompromised patients, accounting for 22% to 33% of infections in one older series13 and 16% of infections present at the time of hospitalization for fever and

TABLE 100-1. Documented Sites of Infection in Patients with Cancer, Fever, and Neutropenia Site or Type of Infection

Ref. 1: No. (%)

Ref. 2: No. (%)

Ref. 3: No. (%)

Total No. (%)

Bloodstream

81 (43)

109 (27)

20 (17)

210 (29)

Pulmonary

28 (15)

88 (21)

16 (13)

132 (18)

Cutaneous

42 (22)

43 (10)

19 (16)

104 (14)

Head, eyes, ears, nose, throat

11 (6)

69 (17)

28a (23)

108 (15)

4 (2)

35 (9)

30 (25)

69 (10)

22 (11)

29 (7)

NR

51 (7)

2 (1)

38 (9)

Gastrointestinal Urinary tract Other NR, not reported separately. a Includes cases of severe mucositis.

571

7 (6)

47 (7)

572

SECTION

P Infections and Cancer

neutropenia in another study.3 Infections of the skin can be caused by viruses, fungi, or bacteria. Bacterial skin infections are usually caused by staphylococci or streptococci, often beginning at sites of skin breakdown, such as surgical or biopsy sites or areas of previous radiation therapy. In immunocompromised patients, both gram-positive and gram-negative bacteria, including enteric organisms and Pseudomonas species, have been isolated from blood cultures and material obtained by fine-needle aspiration.14 Pseudomonas aeruginosa infections are particularly associated with the severe necrotic skin lesions of pyoderma gangrenosum (Figure 100-1). Treatment for presumed bacterial cellulitis, therefore, should provide broad-spectrum coverage and every effort should be made to establish a specific microbiologic diagnosis. Fungal pathogens, including Candida, Aspergillus, Fusarium spp., and Mucor, can cause cutaneous lesions either as isolated findings or as manifestations of disseminated infection. Black, rapidly progressing, necrotic eschars should prompt immediate evaluation for fungal infection. Tender, erythematous skin nodules or pustules are most consistent with Candida infection. Superficial cultures of skin lesions may not provide adequate material for diagnosis of fungal infections; biopsy is more useful. Cutaneous fungal infections resulting from Aspergillus species or other molds require surgical debridement in addition to prolonged courses of amphotericin B, since they are often a manifestation of hematogenous spread or local angioinvasive infection. Cutaneous infections secondary to viruses are common in immunocompromised patients. Both herpes simplex virus (HSV) and varicella-zoster virus (VZV) can cause painful or pruritic vesicular lesions. Both can become secondarily infected with bacteria. Diagnosis is typically made on the basis of clinical findings of typical skin lesions and can be confirmed by unroofing a fresh vesicle, scraping the base of the lesion, and sending the samples to be stained with direct fluorescent antibody specific for each virus. HSV and VZV can be isolated from fluid in fresh lesions, and therefore diagnostic samples should always be sent for viral culture. Differentiation of the two viruses is important because, although both respond to acyclovir therapy, VZV requires doses of 500 mg/m2 given intravenously every 8 hours, whereas HSV is treated with half the dose. Patients with cancer who develop primary varicella, especially those actively receiving chemotherapy, are at increased risk of serious disseminated disease, including pneumonia, encephalitis, hepatitis, and purpura fulminans. Dissemination and subsequent mortality (esti-

Figure 100-1. Pyoderma gangrenosum in an 8-year-old boy with newly diagnosed acute lymphocytic leukemia, fever, and bacteremia due to Pseudomonas aeruginosa.

mated as 7% to 20% in untreated patients15) have been reduced by rapid initiation of therapy with intravenous acyclovir.16 Although recurrent disease in the form of zoster is rarely associated with severe complications when it remains localized to the skin, dissemination occurs in up to 25% of patients with immunocompromising conditions.17 Because the inhibitory concentration of acyclovir for VZV is at the upper limit of levels that can be achieved with oral dosing, the National Cancer Institute policy has been to hospitalize and treat intravenously all patients with cancer who are being actively treated in whom either primary or recrudescent VZV infection develops. A severe variant of scabies (Sarcoptes scabiei), called Norwegian or crusted scabies, occurs in immunodeficient patients and is characterized by the presence of 103 to 106 viable mites, resulting in widespread, hyperkeratotic, crusted lesions. This form is highly contagious and can also be recalcitrant to topical scabicidal therapy. Ivermectin, an oral antiparasitic agent, has been shown to be effective in curing both routine scabies and Norwegian scabies after a single dose.18,19

Pulmonary Infections The lungs are the most common site of localized infection in patients with neutropenia, and pulmonary infection in this population produces a wide variety of symptoms, signs, and radiographic appearances. The differential diagnosis of respiratory tract symptoms in patients with cancer is influenced by the radiographic manifestations (i.e., the type of pulmonary infiltrate seen) and the degree and type of immunosuppression. Table 100-2 summarizes causative agents based on specific radiographic findings.

Localized Infiltrates Acute illness with localized infiltrates radiographically at the onset of fever in a patient with cancer, with or without neutropenia, is most often due to bacterial pneumonia. Because both community-acquired and nosocomial pathogens can be responsible for pneumonia in neutropenic patients, initial antibiotic therapy should provide coverage for organisms such as Streptococcus pneumoniae and Haemophilus influenzae, as well as Pseudomonas spp. and other gram-negative bacteria. Community-associated Panton-Valentin leukocidin-positive Staphylococcus aureus can cause rapidly progressive necrotizing pneumonia. Legionella infection should also be considered in immunocompromised patients with patchy infiltrates; antigen test on urine is sensitive and specific for L. pneumophila. Patients with pulmonary metastases can also have increased risk of postobstructive pneumonia with a variety of pathogens, including anaerobic bacteria. The antibiotic regimens used for empirical therapy of fever without localizing findings in a patient with neutropenia (e.g., a thirdgeneration cephalosporin, an extended-spectrum penicillin in combination with an aminoglycoside, or carbapenem) are appropriate for initial management of pneumonia. In some cases the addition of a macrolide (i.e., erythromycin, clarithromycin, or azithromycin) for possible Mycoplasma or Legionella infections is warranted. Table 100-3 lists recommendations for empiric antibiotic therapy for most common pathogens in patients with extensive pulmonary infiltrates (see Chapter 38, Pneumonia in the Immunocompromised Host). Determining the cause of localized infiltrates that develop during antibiotic therapy is aided by knowledge of the patient’s state of immunosuppression. In one series, infiltrates developing at the time of bone marrow recovery from neutropenia were more likely to have no identifiable causative agent, and more than 90% of patients recovered clinically. This may represent localization of neutrophils and the recovering inflammatory response to a previously unrecognized area of treated bacterial infection. In contrast, patients who develop a new infiltrate during prolonged neutropenia are highly likely to have a fungal pathogen; they have a survival rate of only 32%.20

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Infections in Children with Cancer

CHAPTER

100

573

TABLE 100-2. Causes of Pulmonary Processes in Patients with Cancer Based on Radiographic Abnormality Radiographic Manifestation

Infectious Cause

Noninfectious Process

Focal consolidation (lobar or segmental)

Bacteria (routine and nosocomial pathogens) Legionella Oral flora (aspiration and postobstructive) Mycobacterium tuberculosis Cryptococcus, Histoplasma, Coccidioides

Pulmonary hemorrhage Pulmonary infarction Atelectasis Radiation pneumonitis Drug-related pneumonitis Tumor

Diffuse interstitial infiltrate

Viruses Pneumocystis jirovecii Miliary tuberculosis Disseminated fungi (Cryptococcus, Histoplasma, Coccidioides) Mycoplasma Chlamydia

Pulmonary edema Adult respiratory distress syndrome Drug-related pneumonitis Radiation pneumonitis Lymphangitic metastasis Lymphocytic interstitial pneumonitis (HIV)

Nodular infiltrate (with or without cavitation)

Aspergillus, Mucor, Fusarium Nocardia Bacteria (especially Staphylococcus aureus, Pseudomonas, Klebsiella, anaerobic bacteria) Mycobacterium tuberculosis

Tumor

HIV, human immunodeficiency virus.

TABLE 100-3. Approach to Empiric Antibiotic Therapy in Immunocompromised Patients with Extensive Pulmonary Infiltrates Patient Characteristic

Empiric Regimen

Pathogens Likely to Be Treated

PATIENT WITH DEFICIENT CELL-MEDIATED IMMUNITY (NOT HIV)

Trimethoprim-sulfamethoxazole plus Erythromycin with or without Nafcillin or vancomycin plus Aminoglycoside or third-generation cephalosporin

Pneumocystis, Nocardia Legionella, Mycoplasma, Chlamydia Aerobic gram-positive cocci Facultative gram-negative bacilli

PATIENT WITH NEUTROPENIA

Trimethoprim-sulfamethoxazole plus Erythromycin plus Nafcillin or vancomycin plus Aminoglycoside

Pneumocystis, Nocardia Legionella, Mycoplasma, Chlamydia Aerobic gram-positive cocci Facultative gram-negative bacilli, Pseudomonas aeruginosa

plus Ceftazadime or extended-spectrum penicillin

Facultative gram-negative bacilli, Pseudomonas

PATIENT WITH HIV

Trimethoprim-sulfamethoxazole plus Erythromycin

Pneumocystis Legionella, Mycoplasma

HIV, human immunodeficiency virus.

Viruses, such as influenza and parainfluenza viruses, respiratory syncytial virus, and adenovirus, can also cause localized infiltrates in immunocompromised children, although a diffuse process is more common. In one series, respiratory tract viruses were documented in about 25% of episodes of pneumonia in febrile, neutropenic patients, but the role of these organisms could not be predicted on the basis of presenting symptoms, radiographic findings, or degree or duration of neutropenia.9 Consequently, consideration of a viral process does not obviate the need for broad-spectrum antibiotic therapy in a patient with neutropenia, fever, and a pulmonary infiltrate. Localized infiltrates in patients who fail to respond to broadspectrum antibiotics can also represent infection caused by fungi, Nocardia, mycobacteria, or antibiotic-resistant, hospital-acquired bacteria. Tuberculosis should always be considered. Endemic fungi,

Histoplasma, Cryptococcus, and Coccidioides can also cause localized pneumonitis in selected geographic regions and, in compromised hosts, can be associated with extrapulmonary infection as well. The presence of Candida in respiratory tract secretions, even in patients with pulmonary infiltrates, correlates poorly with causation because of the high frequency of colonization of the oral cavity and tracheobronchial tree. Blood cultures positive for Candida or typical lesions of endophthalmitis are associated with disseminated infection, which can include pneumonia. In the absence of these findings, definitive diagnosis usually requires histopathologic confirmation. The fungal pathogens of most concern in the patient with neutropenia include Aspergillus, Fusarium, Mucor, Pseudallescheria boydii, and Trichosporon, since these organisms cause rapidly progressive, extensively destructive infection. Unlikely to be identified

574

SECTION

P Infections and Cancer

at the onset of neutropenia and fever, the finding of a progressive or new infiltrate, accompanied by fever, nonproductive cough or hemoptysis, and pleuritic chest pain, in a persistently neutropenic patient during broad-spectrum antibiotic therapy suggests the diagnosis of invasive fungal pneumonia (usually with Aspergillus). Computed tomography can reveal multiple pulmonary nodules, sometimes with cavitation, that are not readily apparent on routine chest radiograph.21 The classic histopathologic finding in these infections is invasion of the blood vessels with thrombosis and resulting infarction and hemorrhage. Whereas recovery of Aspergillus from the respiratory tract of patients without neutropenia can represent colonization, isolation of the organism in the setting of prolonged neutropenia and pulmonary infiltrate is highly predictive of invasive disease22 and is considered sufficient evidence for initiation of antifungal therapy. Neither bronchoalveolar lavage (BAL), which has a recovery rate of about 50% in biopsy-proven Aspergillus pneumonia, nor transbronchial biopsy, which has a yield of about 20%, is highly sensitive for the diagnosis; open-lung biopsy may be required.23 Historically, the treatment of choice for Aspergillus pneumonia has been amphotericin B at 1.0 to 1.5 mg/kg per day. Invasive pulmonary aspergillosis can develop in patients who are already receiving empirical amphotericin B for persistent fever and prolonged neutropenia. Higher dosages of amphotericin B (1.5 mg/kg per day) or liposomal amphotericin B (3 to 5 mg/kg per day) are often used in this setting, although increased efficacy has not been proved.24 With the advent of new antifungal agents, however, the choices for treatment of invasive aspergillosis are no longer limited to amphotericin B. In a prospective randomized, unblinded trial comparing the efficacy and safety of voriconazole with that of amphotericin B for the treatment of invasive aspergillosis, initial therapy with voriconazole led to better responses, improved survival (at 12 weeks, 70.8% for the voriconazole group versus 57.9% in the amphotericin B group), and fewer adverse effects.25 Caspofungin, a member of the echinocandins, was initially approved for use in cases of refractory aspergillosis. A prospective evaluation of this drug enrolled 90 patients with invasive aspergillosis who were refractory to (86%) or intolerant of amphotericin B, liposomal amphotericin B, or triazoles (14%) to be treated with caspofungin. A favorable response to caspofungin therapy was observed in 45% of patients (50% with pulmonary aspergillosis and 23% with disseminated infection). Only 2 patients discontinued caspofungin therapy because of adverse effects.26 While successful use of caspofungin as first-line monotherapy for invasive aspergillosis has been reported,27,28 no randomized controlled trials comparing efficacy to other antifungal agents have been conducted. The combination of voriconazole and caspofungin for invasive aspergillosis has been reported to be successful in isolated cases,29–31 but randomized trials are needed in order to determine whether this combination is advantageous (from both an efficacy and safety perspective) over monotherapy or other combination therapy regimens. Other fungal pathogens, such as Histoplasma capsulatum, Coccidiodes immitis, and Cryptococcus neoformans, can also cause focal infiltrates in patients with neutropenia who are receiving corticosteroid therapy.32

diagnosis of PCP can be established by finding cysts or trophozoites on smears of respiratory tract secretions obtained by sputum induction, BAL, or open-lung biopsy. In nonneutropenic patients with diffuse infiltrates who are unable to undergo BAL or biopsy, an empiric course of trimethoprim-sulfamethoxazole (TMP-SMX) plus a macrolide is recommended.34 The treatment of choice for patients with proven or suspected PCP is TMP-SMX at a total daily dose of 15 to 20 mg/kg, divided into four doses given every 6 hours, for 14 days. Prednisone is usually given to patients with moderate or severe PCP (i.e., room air PaO2 70 mmHg or PAO2 ≥35 mmHg). Other therapies found to be effective for PCP in patients with human immunodeficiency virus (HIV) infection include pentamidine, trimetrexate, atovaquone, and clindamycin plus primaquine. The incidence of PCP in patients with cancer has been substantially reduced with the routine use of prophylactic TMP-SMX. Recurrences of PCP or breakthrough infections are unusual in children with cancer, although they have been reported in children with HIV infection.35 Viral infections can also cause pneumonia with diffuse or interstitial infiltrates. Whereas the common respiratory tract viruses cause more severe disease in immunocompromised patients, the most serious cases are caused by cytomegalovirus (CMV). CMV pneumonitis occurs most often in allogeneic bone marrow transplant recipients. Clinical presentation, usually within the first 3 months of transplantation, includes fever and rapidly progressive diffuse pulmonary infiltrates, often causing pulmonary dysfunction requiring ventilatory assistance. Risk factors identified for development of CMV pneumonitis include presence of pretransplant seropositivity for CMV in either recipient or donor, total body irradiation as part of the pretransplant regimen, and development of graft-versus-host disease.36 Historically, CMV pneumonitis had a mortality rate of over 80%. Treatment with ganciclovir and high-dose intravenously administered immune globulin has improved survival substantially, but, even with this regimen, 50% of patients die.37,38 Although definitive diagnosis of this process requires identification of CMV in lung tissue obtained by biopsy, standard practice in many transplant centers now includes preemptive therapy with intravenous ganciclovir for patients who have CMV isolated from any site during the early posttransplant period.39

Infections of the Ears and Sinuses In the neutropenic child, otitis media and otitis externa can be accompanied by pain, with minimal erythema of the tympanic

Diffuse or Interstitial Infiltrates Diffuse or interstitial infiltrates most likely represent a nonbacterial process, although both gram-positive and gram-negative bacteria can cause interstitial pneumonitis. In children not receiving prophylaxis, Pneumocystis jirovecii (P. carinii) is the organism most often identified in the setting of diffuse, interstitial infiltrates. P. jirovecii should be considered in children receiving corticosteroid therapy or in children with deficient cell-mediated immunity who have fever, nonproductive cough, tachypnea, and hypoxia. In one review of oncology patients with confirmed PnemoCystis Pneumonia (PCP), 70% of cases became symptomatic while corticosteroid dosages were being tapered.33 The chest radiograph typically reveals bilateral perihilar interstitial (or alveolar) infiltrates that spread to involve all lobes (Figure 100-2), but virtually any radiographic pattern has been described with PCP. The

Figure 100-2. Chest radiograph showing bilateral interstitial and alveolar infiltrates in a child with acute lymphocytic leukemia and Pneumocystis pneumonia.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Infections in Children with Cancer

membrane and canal. In addition to the usual bacterial pathogens responsible for otitis media (Streptococcus pneumoniae, Haemophilus influenzae, Moraxella catarrhalis), nosocomial pathogens (such as gram-negative organisms) must be considered. Malignant otitis externa resulting from invasive infection of the auditory canal with Pseudomonas aeruginosa (and occasionally invasive fungi) occurs more commonly in patients with diabetes mellitis or those receiving corticosteroids, but it may occur in patients receiving chemotherapy. Aggressive intravenous antimicrobial therapy and debridement are required for effective therapy. Although mastoiditis is an uncommon infection in pediatric cancer patients, there is a subset of patients who are at increased risk, which includes those with anatomic abnormalities of the middle ear and individuals with prolonged neutropenia. Mastoiditis in neutropenic patients may be caused by typical bacterial pathogens or by fungi (particularly Aspergillus), which often require surgical debridement for cure.40 The paranasal sinuses can also become infected with typical pathogens or the altered bacterial flora present in hospitalized patients. Sinusitis was found more commonly in children with hematologic malignancies than in those with solid tumors in one series, and 41% of 91 children with acute lymphoblastic leukemia had abnormal sinus radiographs at the time of induction chemotherapy.41 Children with tumors involving the sinuses are at particular risk of recurrent or chronic sinusitis. Distinguishing between progressive or necrotizing tumor and infection is challenging. Sinus radiograph and computed tomography are helpful tools in diagnosis. Broad-spectrum antibiotics, including an agent with activity against anaerobic organisms, are necessary in neutropenic patients who demonstrate signs and symptoms of sinusitis. Symptoms not responding to empirical antibiotic therapy within 48 to 72 hours should be further evaluated with aspiration or biopsy of the involved sinuses. Fungal sinusitis is most common in the presence of prolonged neutropenia. In one series of fungal sinusitis in children, facial pain or headache, fever, facial swelling (which developed in 50% of patients), and abnormal Àndings on sinus radiographs were the most common clinical Àndings.41 A 10-year retrospective analysis of invasive fungal sinusitis revealed that the majority of cases were caused by Aspergillus flavus, and the most common presenting symptom was periorbital swelling (41% of patients).42 Other fungal causes of sinusitis include Mucor, Fusarium, Pseudallescheria boydii, and Rhizopus. Infection can begin as a small area of blackened eschar within the nose or sinus but can become rapidly progressive with signiÀcant tissue invasion and necrosis secondary to vascular thrombosis, and ultimate extension into the orbits or brain. Nasal cultures for surveillance to diagnose invasive Aspergillus infection have been of variable utility.22,41 Biopsy material for histopathologic examination and culture is necessary to establish the diagnosis deÀnitively. Therapy for the rhinocerebral syndrome of fungal infection requires aggressive surgical debridement and long-term therapy with amphotericin B, usually at doses of 1.0 to 1.5 mg/kg per day. Liposomal amphotericin is an alternative.

Gastrointestinal Tract Infections Infectious complications of the gastrointestinal tract are common in children with cancer. Both the underlying diseases and the cytotoxic and radiation regimens used in therapy predispose to alterations of the mucosal surface, leading to local infections and serving as a portal of entry for the microorganisms residing in the gastrointestinal tract.

Mucositis and Esophagitis Mucositis as a side effect of cytotoxic chemotherapy can range in severity from isolated small oral ulcers to extensive mucosal sloughing of the oral cavity and more distal gastrointestinal tract. A recognizable complication of this process is necrotizing or marginal gingivitis, characterized by a periapical line of erythema and tenderness. Presumably resulting from anaerobic infection, this Ànding indicates a need to include an agent effective against anaerobic bacteria in the

CHAPTER

100

575

antibiotic regimen (e.g., clindamycin, metronidazole, a beta-lactamase inhibitor combination drug imipenem, or meropenem). In addition, children who have been hospitalized for a prolonged period or who have received broad-spectrum antibiotics may have oral colonization with gram-negative organisms or fungi. Chemotherapy-induced mucositis can also become superinfected with Candida, Aspergillus,43 or HSV. A Gram stain of lesions may identify yeast or hyphae; Candida infection usually responds to topical treatment with clotrimazole troches, or, in more severe cases, fluconazole, voriconazole,44 caspofungin,45–47 or amphotericin B. Locally invasive or disseminated fungal infection can follow stomatitis. HSV infection can be documented by culture from a swab of the involved area; infection usually responds to oral or intravenous acyclovir. Because patients with a past history of HSV infection have a 70% to 80% incidence of reactivation during induction therapy for leukemia or after transplantation, it has been recommended that they receive prophylactic acyclovir during the high-risk period of leukopenia.48 Alphahemolytic streptococcal septicemia with high rates of shock and central nervous system (CNS) infection can also complicate mucositis; up to 75% of isolates of a-hemolytic streptococci are not sensitive to penicillin.49 The processes predisposing to mucositis in the oral cavity, both infectious and noninfectious, can also produce esophagitis. In addition, mediastinal irradiation can produce a syndrome of esophagitis that is clinically similar to infectious esophagitis. Determining whether substernal burning chest pain and odynophagia result from an infectious or noninfectious cause can be difÀcult. Neither barium swallow nor endoscopic visualization reliably distinguishes esophagitis caused by Candida, HSV, and noninfectious causes. One study showed that 21 of 22 cancer patients with a clinical and microbiologic diagnosis of oral candidiasis also had endoscopic and microbiologic Àndings diagnostic for candidal esophagitis, suggesting that oropharyngeal candidiasis may represent a reliable marker for esophageal candidiasis in patients with cancer.50 DeÀnitive diagnosis, however, requires endoscopically performed biopsy and culture. When biopsy is considered too risky (e.g., a patient with a platelet count of < 50,000/mm3), an alternative approach is to begin empirical therapy for Candida esophagitis with one of the antifungal azoles, and if symptoms fail to respond within 48 hours, institute a trial of amphotericin B or caspofungin. Failure to improve within 48 hours after initiation of amphotericin B or caspofungin makes Candida a less likely cause of esophagitis. Endoscopic biopsy should be reconsidered or an empirical trial of acyclovir for presumptive HSV infection should be considered. In the patient without neutropenia, infectious causes of esophagitis are uncommon; symptomatic treatment with antacids or histamine-blocking agents may be appropriate.

Intra-Abdominal Infections The immunosuppressed patient with cancer is at increased risk of intra-abdominal infections because of invasion or obstruction of the bowel by tumor, extension of mucosal ulcerations, or sloughing secondary to chemotherapy. Typhlitis, or neutropenic cecitis, is a necrotizing process involving the cecum and terminal ileum, usually occurring in patients who are experiencing profound neutropenia. Typhlitis manifests with acute right lower quadrant abdominal pain (mimicking appendicitis), which may become generalized and is usually accompanied by fever and systemic symptoms. Because of neutropenia, many patients lack the classic signs of peritonitis. Histopathologic examination of the bowel reveals inÀltration of the bowel wall with bacteria, usually gram-negative bacilli (especially Pseudomonas), with little or no surrounding inflammation, and progression in some areas to necrosis. In a review of 24 cases of typhlitis over 30 years, bacteremia was documented in 8 of 24 children, and thickening of the bowel wall was seen on computed tomographic scan in 17 of 20 patients studied.50 Surgical resection of necrotic bowel may be necessary when perforation, abscess formation, uncontrolled bleeding, or septic shock occurs. However, most cases of typhlitis respond to medical management, including use of broad-

576

SECTION

P Infections and Cancer

spectrum antibiotic therapy, and recovery from effects of cancer therapies.51 Although uncommon, a specific and devastating variation of typhlitis is one that culminates in spontaneous peritonitis and septicemia due to Clostridium septicum. Peritonitis involving this species classically manifests with a fulminant picture, including rapidly progressive abdominal wall myonecrosis with crepitance, hemolysis, and shock. Importantly, this devastating syndrome can also occur in the absence of fever. C. septicum, which is more aerotolerant than C. perfringens and can grow in viable, nonnecrotic tissue, may be part of the natural flora of the gastrointestinal tract. Over 80% of patients with this syndrome have an underlying malignant process.52,53 The presumed pathophysiology of the process involves breaks in mucosal integrity secondary to malignancy or chemotherapy, microbial invasion of the bowel wall, and rapid proliferation in the setting of immunosuppression or granulocytopenia.54 Additionally, C. septicum produces an a-toxin that may play some role in the severity of infection. Pseudomembranous or antibiotic-associated colitis has been reported after the use of many of the antibiotics that are part of empirical therapy in febrile neutropenic patients. This process is associated with production of toxins by C. difficile; isolation of the organism from stool is not in itself diagnostic. Many patients become colonized during hospitalization. Children with C. difficile-associated diarrhea can have symptoms ranging from mild abdominal pain to severe bloody diarrhea. Stool samples should be tested for C. difficile toxin whenever diarrhea or abdominal pain develops in a patient with neutropenia. C. difficile-associated colitis usually responds to treatment with oral metronidazole or vancomycin; relapses, which may occur in 10% to 20% of patients, can usually be treated with a second course of the same agent (see Chapter 190, Clostridium difficile). Hepatosplenic candidiasis (also referred to as chronic disseminated candidiasis) is a problematic intra-abdominal infection that afflicts patients with cancer, especially those with prolonged episodes of neutropenia. The usual clinical setting is that of a child with persistent fever that is unresponsive to broad-spectrum antibiotics during neutropenia, sometimes accompanied by right upper quadrant abdominal pain or tenderness and an elevated alkaline phosphatase level. Most often, multiple cultures from blood, urine, are negative. Typical hepatic “bull’s-eye” lesions on ultrasonography or hypodense lesions on computed tomographic scan may be seen as the neutrophil count begins to recover. Magnetic resonance imaging may be the most sensitive study. Before recovery of the inflammatory response, hepatic lesions can be missed by all imaging techniques available. Blood culture or detection of Candida subcellular components do not diagnose all invasive candidiasis.55,56 Confirmation of the diagnosis may require liver biopsy (especially in children with risk factors for hepatic metastasis). Treatment of hepatic candidiasis is difficult, usually requiring a long course of amphotericin B. Combination therapy with flucytosine may be advantageous, particularly if the kidney or CNS is involved. Amphotericin B lipid complex has been shown to be effective in the treatment of hepatosplenic candidiasis.57 Fluconazole has been shown to have efficacy in the treatment of hepatosplenic candidiasis in patients in whom amphotericin B failed or who have experienced serious amphotericin B-related toxicities.21,32,58,59 Treatment failures with fluconazole in this setting, however, have been reported. The major therapeutic dilemma remains the appropriate length of therapy, because the endpoint often depends on resolution or calcification of the lesions that were visualized.

Perianal Cellulitis In general, children have fewer chemotherapy-related perirectal lesions than adults. However, children can develop mucositis involving the rectal mucosa, providing focus for local cellulitis or abscess. Multiple organisms are identified in most cases in which surgical drainage or needle aspiration is performed; anaerobic organisms are most commonly isolated.60 Most cases can be treated effectively with antibiotics alone. Meropenem, an extended-spectrum penicillin, or

third-generation cephalosporin in combination with metronidazole or clindamycin is appropriate; in some cases, an aminoglycoside is also used.

Central Nervous System Infections Infections of the CNS occur infrequently in children undergoing cancer chemotherapy. Alpha-streptococcal bloodstream is associated excessively with CNS infection.49 CNS infections must be considered in children who have undergone a neurosurgical procedure. Placement of an intraventricular shunt or Ommaya reservoir poses an additional risk of infection; 25% of children experience a related CNS infection, but death is rare.61 Indolent infections are most common with organisms of low pathogenicity, such as Staphylococcus epidermidis and Propionibacterium acnes. Because of the vague nature of associated symptoms, any child with a CNS device who has fever and no localizing findings should have fluid sampled for cell count, chemical evaluation, and culture. Although meningitis in cancer patients is uncommon (one study noted that fewer than 0.9% of febrile neutropenic episodes in pediatric cancer patients were caused by CNS infections62), it is associated with significant mortality and morbidity. Patients with fever and a change in mental status should be evaluated promptly. Some studies have reported meningeal signs in a minority of neutropenic patients or those who have had neurosurgical manipulation who develop meningitis.63–65 Thus the absence of meningeal signs does not exclude the diagnosis of meningitis. In a retrospective review of 40 cases of bacterial or fungal meningitis in pediatric cancer patients, most patients (65%) had recent neurosurgery, a CNS device placed, or a cerebrospinal fluid leak.66 Fever and altered mental status were the most consistent signs at presentation, with meningismus notably less frequent in neutropenic patients. Staphylococcus aureus and Streptococcus pneumoniae, and, more recently, oral streptococci are the most common pathogens. Of the 5 patients with fatal outcomes, all were neutropenic at presentation. Although the incidence of CNS aspergillosis is quite low, the mortality approaches 100%. A retrospective review of 81 patients with definite or probably CNS aspergillosis who were treated with voriconazole showed that a complete or partial response was recorded in approximately one-third of patients and that those who underwent neurosurgical therapeutic procedures had improved survival.67

Cystitis Hemorrhagic cystitis, manifesting as bladder pain and gross hematuria, has been reported in bone marrow transplant recipients. Adenovirus (particularly type 11) and polyomavirus (BK virus) are likely etiologies. Polymerase chain reaction detection of adenovirus is possible.68 Case reports documenting successful treatment of adenoviral and polyomaviral hemorrhagic cystitis with intravenous ribavirin, vidarabine, and ganciclovir have been reported, but casecontrolled randomized studies have not been performed.69–72

Osteoarticular Infections Bone and joint infections occurring in patients with cancer are uncommon. Children with soft-tissue or bone tumors are at risk of infectious complications in areas of surgery or of destruction by the underlying tumor. It may be difficult to distinguish osteomyelitis from tumor or the effects of local radiation. Any organism that can lead to bacteremia, including gram-negative and gram-positive bacteria and fungi (e.g., Aspergillus73), can cause osteomyelitis or arthritis. Diagnosis and treatment can be particularly complex in children who have had limb-sparing procedures involving bone and joint prostheses and in whom long-term stability of the prosthesis is of concern. Serial imaging studies may suggest infection, but biopsy of the involved bone is necessary for definitive diagnosis. Long-term antimicrobial therapy is mandatory when infection occurs, and revision of the prosthesis is often required for cure.74

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

SECTION

Q

Infections Associated with Hospitalization and Medical Devices CHAPTER

10 1

Healthcare-Associated Infections Susan E. Coffin and Theoklis E. Zaoutis

A healthcare-associated (HA) infection is typically defined as any infection not present or incubating at the time of the patient’s initial contact with a healthcare setting. Originally called nosocomial, hospital-acquired, or hospital-onset infections, the scope of HA infections has expanded as patients now receive medical care in a wide variety of healthcare settings. Many HA infections are preventable and thus should be a target for aggressive infection control programs (see Chapter 2, Pediatric Infection Control). In 1985, the Study on the Efficacy of Nosocomial Infection Control (SENIC) project found that 5.7% of 169 526 patients hospitalized in 338 randomly selected United States hospitals experienced an infection.1 In addition, this study estimated that 32% of these hospital-onset infections could have been prevented by the application of known principles and practices of infection control (e.g., the presence of one infection control practitioner for every 250 hospital beds, the presence of a hospital epidemiologist, a comprehensive surveillance program to detect infections, and active programs that implement specific interventions to prevent nosocomial infections). Although this study specifically excluded children’s hospitals, many of its findings have been generalized to facilities that provide pediatric care. As in adults, HA infections are common in pediatric patients. However, the rates of specific types of infection differ among pediatric and adult facilities (Table 101-1). A variety of factors help explain why the microbiology of pediatric and adult HA infections differs (Table 101-2). Children are less likely than adults to have comorbid conditions, such as diabetes mellitus, which increase the risk of specific HA infections. In contrast, the reservoir of viral infections, such as respiratory syncytial virus (RSV) or rotavirus, is greater in sites where children seek medical care. The only source of national data on hospital infections in the United States is the National Healthcare Safety Network (NHSN), originally named the National Nosocomial Infections Surveillance (NNIS) system. This program was organized in 1970 and is coordinated by the Centers for Disease Control and Prevention (CDC). This system currently monitors HA infections and uses uniform definitions of infection and standard surveillance protocols.2 These surveillance definitions are not intended for the diagnosis of clinical disease. They are designed to standardize the detection and reporting of HA infections so that consistent data are provided from all reporting hospitals. Through the use of standardized definitions and appropriate denominators that express the population at risk, healthcare epidemiologists can track the rates of selected infections in their institutions as well as compare their rates with peer insitutions.3,4 NHSN definitions for HA infections recognize that the clinical manifestations of some illnesses differ in infants from those in older children and adults. Thus, two sets of criteria are provided for the definitions of pneumonia and bloodstream infection (BSI), one for infants < 12 months and one for patients > 12 months.

Rates of HA infections among pediatric patients vary according to birthweight, age, underlying diseases, and intensity of medical care. Rates of infection are highest in children < 1 year of age and in children who require intensive care, particularly infants in neonatal intensive care units (NICUs).4–7 These vulnerable patients often have many of the risk factors that predispose patients of all ages to infections, such as severe underlying illness, loss of skin integrity, or the presence of multiple medical devices (e.g., an endotracheal tube bypassing the mucociliary elevator) that breech normal defense mechanisms. In addition, pediatric HA infections differ from those seen in adults by site and pathogen. For example, in children BSIs are the most common HA infection whereas in adults, catheter-related urinary tract infections (UTIs) predominate.8–10

BLOODSTREAM INFECTIONS Device-Associated Infections BSIs are the most common HA infections experienced by pediatric patients (see Table 101-1). Among hospitalized and chronically ill patients, the majority of HA BSIs are related to the use of intravascular catheters and are discussed in depth in Chapter 102 (Clinical Syndromes of Device Associated Infections). In addition, pediatric patients are at risk for BSIs as a result of contaminated intravenous fluids or medications, blood or blood products, or dissemination from a distant site of infection.

TABLE 101-1. Distribution of Healthcare-Associated Infections by Site in Pediatric Patients at Children’s Hospitals and Nonchildren’s Hospitals, from the National Nosocomial Infections Surveillance (NNIS) System, 1990–1999a

Site of Infection

Children’s Hospitals

Nonchildren’s Hospitals

Total

n = 8081 (%)

n = 11,400 (%) n = 19,481(1%)

Bloodstream

31

22

26

Pneumonia

12

22

18

Urinary tract

13

16

15

Surgical site

9

14

12

Lower respiratory tract

14

6

9

Ear, eye, nose, throat

6

7

7

Gastrointestinal tract

8

5

6

Skin/soft tissue

4

4

4

Cardiovascular system

2

2

2

Central nervous system

1

–

–

Reproductive system

–

2

5 was predictive of VAP.73 In 1991, Pugin and colleagues proposed the Clinical Pulmonary Infection Score (CPIS) as a tool to identify adult ICU patients with pulmonary infection and demonstrated good correlation between a threshold score of 6 and quantitative bacteriology of BAL samples.81 Subsequently, several attempts have been made to validate the CPIS in adult patients, with variable results. This score was subsequently modified and evaluated as a tool to limit unnecessary antibiotic use in

Management and Outcome Ventilator-Associated Pneumonia Empiric management of suspected bacterial pneumonia depends on both patient-specific and hospital-related considerations. All of the following should be considered: the patient’s underlying disease, previous antimicrobial therapy, mental status (and other conditions that increase the risk of aspiration), length of hospitalization, and results of Gram stain examination of respiratory tract specimens, and detailed knowledge of the local antimicrobial susceptibility patterns of common bacterial pathogens derived from microbiology surveillance data. Empirical treatment for VAP is generally given when visualization of a new infiltrate on chest radiograph is associated with fever, an increase in quantity or purulence of respiratory tract secretions, rising neutrophil count, or an unexplained decrease in oxygen saturation. Gram stain and culture of respiratory tract secretions are commonly used to support the diagnosis and direct therapy. The American Thoracic Society and the Infectious Diseases Society of America have authored guidelines for the treatment of HA pneumonia and VAP in adults,83 but many of the principles apply equally well to children. Because delays in the initiation of appropriate therapy have been associated with increased morbidity and mortality,84,85 initial therapy should be broad. Negative lower respiratory tract cultures, in the absence of a recent change in antibiotic therapy, can be used to stop empiric therapy. Short courses of therapy (5 to 7 days) may be adequate in patients with uncomplicated VAP who have a good initial clinical response to treatment.86

Viral Respiratory Tract Infections Although ribavirin has in vitro activity against RSV, it is not generally recommended.87 Studies have demonstrated that the clinical improvements achieved during ribavirin administration are limited or sometimes insignificant. Although several studies have reported that clinical improvement has been seen with subsequent reductions in length of mechanical ventilator support, these findings have been inconsistent, and treatment may not be cost-effective in children

582

SECTION

Q Infections Associated with Hospitalization and Medical Devices

hospitalized with RSV.88,89 Treatment may be appropriate in selected patients, such as those who have hemodynamically significant congenital heart disease or are profoundly immunosuppressed. Immunotherapy with RSV monoclonal antibody has also been evaluated for use in therapy. Although therapy is associated with reduced viral loads, there is no significant impact on the clinical course or duration of hospitalization.90–92 Children with HA influenza infection should be treated with age-appropriate antiviral therapy. The outcome of HA lower respiratory tract infections varies greatly and is determined by characteristics of both the host and the pathogen. VAP is associated with significant morbidity and mortality, particularly in neonates; the case-fatality rate for neonates with VAP is reported to be 10%, with extremely premature infants having a rate of 27%. Among pediatric ICU patients, the reported rate is approximately 6%. VAP is associated with prolonged ICU stay and prolonged hospital stay in both NICU and pediatric ICU (PICU) patients.47

Prevention The Hospital Infection Control Practices Advisory Committee of the CDC has published a comprehensive collection of recommendations for the prevention of HA pneumonia.93 This “Guideline for the prevention of nosocomial pneumonia” has specific recommendations for the prevention of bacterial pneumonia, Legionnaires’ disease, pulmonary aspergillosis, influenza, and RSV. A summary of the recommendations to prevent bacterial pneumonia is given in Box 101-1. Major areas for preventive interventions include technique for intubation, proper positioning, use of aseptic technique during suctioning, appropriate maintenance of respiratory therapy equipment, and careful handwashing before and after contact with an intubated patient or patient’s secretions. Taken together, these interventions, sometimes called “bundles,” substantially reduce the incidence of VAP. Additional study is needed to define the optimal methods to prevent HA viral respiratory tract infection in children. Macartney and colleagues demonstrated that a targeted, multicomponent program designed to reduce nosocomial transmission of RSV was effective and cost-efficient.72 Components included: (1) prompt laboratory confirmation of infection; (2) cohorting of patients and nursing staff; and (3) use of contact precautions (gloves and gowns). However, little is known about the efficacy and cost-efficiency of similar programs targeted at other pathogens.

GASTROINTESTINAL TRACT INFECTIONS Epidemiology and Pathogenesis HA gastroenteritis has been estimated to occur in a minimum of 10.5 patients in 10 000 discharges.94 Infants and the elderly are at greatest risk. Estimated rates of HA gastroenteritis for neonatal, pediatric, and high-risk nursery patients are 3.0, 11.3, and 20.3 per 10 000 discharges, respectively. Among cases reported to NHSN, an etiologic agent was identified for 97% of episodes; Clostridium difficile and rotavirus were the most common agents, accounting for 91% and 5% of reported infections, respectively. However, because many hospitals do not have diagnostic virology laboratories, the rates of HA viral gastroenteritis are probably higher than reported. Both intrinsic and extrinsic factors raise the risk of gastrointestinal infections in hospitalized patients. Intrinsic factors include impairment of immunity, alterations in gastric acidity, and changes in gastric motility and flora. Several studies have shown that bone marrow transplant recipients and patients with severe immunodeficiencies are at risk of repeated and prolonged episodes of HA viral gastroenteritis.95,96 Immunocompromised patients also excrete large numbers of virus particles for prolonged periods and thus are an important reservoir of infection within the healthcare setting.95 Extrinsic factors influencing the risk of HA gastroenteritis include the adherence to hand hygiene97 and use of nasogastric tubes or

antacids. The most common mode of transmission of gastrointestinal tract pathogens is via the fecal–oral route, frequently carried from patient to patient by the hands of healthcare workers. Indirect contact via fomites or environmental contamination can also play a role in the transmission of C. difficile. Contaminated food or other common vehicles have been associated with the transmission of Salmonella spp., Shigella spp., Yersinia enterocolitica, Escherichia coli, noroviruses, and Cryptosporidium species.5,7,98–103

Viruses The majority of episodes of gastroenteritis among children are caused by viruses and rotavirus is typically the most common and serious cause of viral gastroenteritis among hospitalized children.104,105 Rotavirus was first recognized as an important cause of HA infection on pediatric wards.106–108 The incidence of HA rotavirus infections parallels that of community-acquired infections; both sporadic and epidemic infections (outbreaks) occur.104,109 HA rotavirus infections have been reported to occur in virtually all settings that provide medical care to children.110–112 Risk factors associated with transmission include the use of shared toys and poor attention to hand hygiene.113 Several reports have highlighted the challenges of recognizing and controlling outbreaks of HA rotavirus infection in neonates.104 In some studies, up to 75% of rotavirus-infected neonates are asymptomatic.114–116 In addition, rotavirus can be shed for prolonged periods after a primary infection.116 In 1982, Rodriguez and colleagues117 reported two outbreaks of rotavirus infections in neonatal nurseries. In the first outbreak, 10 infants in three of six nursery rooms were infected. The involved rooms were closed to further admissions after the infected patients were isolated, and infants and staff were cohorted (i.e., infected infants had a separate nursing staff). No further cases were found, and no source or index case was identified. In the second outbreak, 12 infants developed clinically apparent illness after the admission of patients with failure to thrive who were later found to be with positive rotavirus antigen test. During the two outbreak periods, rotavirus was detected in approximately 35% of stools tested. Of note, not all infants excreting rotavirus were symptomatic; rotavirus was found in 15 (25%) of 61 stools designated as loose, mucoid, or watery, and in 24 (11%) of 214 normal stools. Data suggest that effective control measures after detection of a rotavirus nursery outbreak include screening all patients for rotavirus and creating cohorts of infected and noninfected. Infected but asymptomatic staff have been identified during some outbreak investigations.118 In addition, nosocomial outbreaks are often difficult to control when rotavirus is circulating in a community because of the ongoing admission of infected children (i.e., new reservoirs of disease).119–121 Adenoviruses, noroviruses, and astroviruses have also been reported as causes of HA gastroenteritis in children. Adenovirus can cause severe gastroenteritis in immunocompromised patients.122,123 Several clusters of HA adenovirus gastroenteritis have been described. One study specifically surveyed enteric adenoviruses in hospitalized infants with gastroenteritis.124–126 Noroviruses, members of the Caliciviridae family, have been associated with outbreaks of gastroenteritis and hospital transmission.127–129 Noroviruses can be transmitted through direct and indirect contact, including through food or water sources.101,127,130,131 Environmental contamination has frequently been implicated as a cause of protracted outbreaks.132 Immediate cohorting of ill patients, furloughing of ill healthcare workers, and implementation of stringent environmental control measures are necessary to stop continued transmission. Clostridium difficile. In both children and adults, Clostridium difficile is the most important bacterial pathogen associated with HA gastrointestinal disease. C. difficile was first identified in the stool of neonates in 1935133 but initially was not thought to be an important human pathogen. In 1977, however, a C. difficile toxin was discovered in the stools of patients with pseudomembranous colitis.134,135 The epidemiology of C. difficile is complex and somewhat confusing in the pediatric population (see also Chapter 190, Clostridium

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Healthcare-Associated Infections

difficile). Up to 50% of healthy neonates may be colonized with toxigenic forms of C. difficile.136,137 After the age of 2 years, colonization rates decrease to adult levels of < 5%, although colonization is much more prevalent in hospitalized patients with or without exposure to antimicrobial agents.138,139 Some studies have questioned the frequency of HA transmission of C. difficile in situations other than outbreaks.140,141 Molecular analysis of isolates has shown that C. difficile isolates from asymptomatic or mildly symptomatic hospitalized patients are often polyclonal and likely represent endogenous strains.140,141 Clinically apparent disease occurs when the microbial ecology of the gut is upset, such as when a patient is undergoing antimicrobial therapy. With the decline of normal intestinal flora, C. difficile proliferates and elaborates potent toxins A and B, which stimulate secretory diarrhea and inflammatory colitis.142–144 Although not all cases of antibiotic-associated diarrhea (AAD) or pseudomembranous colitis are caused by C. difficile,145 it is the most commonly identified pathogen in these clinical syndromes. The major risk factor for pseudomembranous colitis is exposure to antimicrobial agents. Pseudomembranous colitis in children has been associated with a wide variety of antimicrobial agents, most commonly cephalosporins and clindamycin.146,147 Other risk factors are underlying gastrointestinal disease, bowel stasis, and anatomic obstruction.144,148,149 Recent reports suggest that the epidemiology of C. difficile infections is changing. In 2002, some adult hospitals noted an increase in the frequency of severe C. difficile infections.150 Subsequent analysis demonstrated the existence of novel strains that hypersecrete toxins, leading to increased severity of clinical disease.151 C. difficile-associated diarrhea has recently been reported to be increasing in patients without apparent risk factors for disease, such as healthy individuals without contact with a healthcare setting or exposure to antibiotics.152

Clinical Manifestations and Laboratory Diagnosis Rotavirus Rotavirus typically causes a secretory diarrhea, often associated with fever and vomiting, and is more likely than other causes of viral gastroenteritis to lead to dehydration.153,154 Children between the ages of 6 months and 3 years are most susceptible. However, repeated infections can occur at any age. Case-series of children with HA rotavirus have demonstrated that hospitalized children of all ages are vulnerable.106,107 Rotavirus infection is typically diagnosed by detection of rotavirus antigen in the stool of a symptomatic or exposed patient. Alternative methods of diagnosis have included documented seroconversion, detection of viral particles by electron microscopy, or PCR.

Clostridium difficile Symptoms of uncomplicated AAD are usually mild and resolve with discontinuation of the antibiotics. Symptom onset is usually 1 to 21 days (mean, 4 days) after initiation of antimicrobial therapy. However, occasionally symptoms begin after the agent is discontinued.144 Associated signs and symptoms include fever, cramping abdominal pain, distention, nausea, and vomiting. Typically, hematochezia does not occur. However, symptoms can progress to pseudomembranous colitis, which usually manifests as the acute onset of profuse watery diarrhea, typically within 1 week of initiation of antimicrobial therapy. The spectrum of disease in pseudomembranous colitis is broad; complications include dehydration, protein-losing enteropathy, ascites, peripheral edema, toxic megacolon, peritonitis, septicemia, and intestinal perforation.142–145,148 Diagnostic tests for C. difficile include culture of stool using selective media, testing of stool or stool filtrates for toxin using either tissue culture or enzyme immunoassay tests, genetic fingerprinting

CHAPTER

101

583

with serotyping, and PCR.155 Because the presence of the bacteria or toxin does not necessarily confirm causality of symptoms, endoscopy is sometimes required for definitive diagnosis.144,149 On direct examination, lesions in the colon or small bowel are red, edematous, and friable, with multiple raised, yellowish plaques (pseudomembranes) that consist of mucus, fibrin, necrotic cells, and polymorphonuclear cells.

Management and Outcome Rotavirus Rotavirus is typically a self-limited infection. Supportive care with careful attention to hydration status remains the cornerstone of management of infected children.156 Rotavirus has been shown to contaminate the fomites and persist on inanimate objects for long periods of time. Thus, appropriate cohorting of patients and staff, as well as environmental cleaning, is critical to prevent the transmission of rotavirus in hospitals.157,158

Clostridium difficile AAD is treated by discontinuation of the antibiotic and provision of supportive care; invasive studies are not necessary. Patients with pseudomembranous colitis that persists after discontinuation of antibiotics require specific therapy. Oral metronidazole, bacitracin, or vancomycin have all been documented to ameliorate symptoms and eradicate C. difficile.143,147,148,159,160 In an era of vancomycin-resistant enterococci and growing antimicrobial resistance among other HA pathogens, oral metronidazole is the preferred treatment; it is also less expensive than the other agents.161,162 Vancomycin should be reserved for cases that do not respond to metronidazole. When vancomycin therapy is necessary, oral administration is preferable to intravenous therapy, because the latter results in low antimicrobial concentrations in the colon. In patients who are unable to take medications orally or by nasogastric tube, and in patients with adynamic ileus, intravenous metronidazole is preferable, or sometimes vancomycin by colonic tube is required. Response to treatment generally occurs within 24 to 48 hours; diarrhea and other symptoms usually resolve within 5 days. Relapse after therapy has been reported in up to 35% of patients and reflects: (1) reinfection; (2) relapse secondary to failure to eradicate the initial organism; or (3) germination of persistent spores.163,164 Whether recurrences are more often caused by reinfection or relapse remains controversial, but host factors such as immunologic impairment, vascular disease, and repeated exposure to antimicrobial agents as well as the severity of the initial episode may affect the frequency.163 Some studies of the use of probiotic agents, such as Lactobacillus and Saccharomyces boulardii, as adjuncts to standard antimicrobial therapy have demonstrated reductions in severity of symptoms as well as in recurrence rates,165,166 although this strategy remains controversial.167 Antiperistaltic agents are contraindicated because they decrease toxin clearance. Barium enema is contraindicated in the early phase of illness because it can induce toxic megacolon or perforation. Strict infection control practices and isolation precautions are critical to prevent nosocomial transmission or recurrences of C. difficile gastroenteritis. Contact isolation precautions should be instituted for any patient with symptomatic infection. Because C. difficile commonly contaminates the inanimate environment, care should be taken to wash hands after all patient contact or after any contact with objects in the patient’s room.168,169 Transmission of C. difficile most commonly results from person-to-person transmission after inadequate handwashing by healthcare workers. Soap and water must be used to perform hand hygiene when caring for a patient with C. difficile, because alcohol-based hand rubs are not sporocidal.170 Careful terminal environmental cleaning of patient’s rooms is essential.

584

SECTION

Q Infections Associated with Hospitalization and Medical Devices

SURGICAL SITE INFECTIONS Epidemiology and Pathogenesis Surgical site infections (SSIs) are the third most commonly reported infection in the NHSN system, representing 14% to 16% of all infections.171 In 1991, a SSI risk index was introduced in the NHSN system that included: (1) site of surgery; (2) wound classification; (3) procedure-based cutoffs for duration of surgical time; and (4) the presence of comorbid conditions.172 On the basis of data from almost 85 000 operations, the NHSN surgical infection index predicted the occurrence of SSIs better than wound classification alone.172 For most common operative procedures, the infection rate rose as the risk index increased; neurosurgical procedures were exceptions. SSI rates for commonly performed procedures in pediatric patients, calculated with data from January 1992 through April 2000, are presented in Table 101-4.173 The NHSN index has not been validated for children undergoing surgery, and significant differences between adults and children may affect the usefulness of the index for pediatric patients. For instance, the 75th percentile cutoff of 5 hours applied to cardiac procedures other than coronary artery bypass grafting is mostly based on procedures performed on adults and may not apply to the pediatric population. In addition, because a large number of pediatric surgical procedures consist of low-risk operations, such as elective hernior-

rhaphy, the NHSN risk index may not accurately identify the small proportion of children who experience SSIs. To date, most studies of SSIs among children have been based on a single hospital’s experience, and few have attempted to adjust infection rates for the presence of known risk factors.174–177 The requisite elements for the development of an SSI include a reservoir of microorganisms, a mode of transmission, and a suitable wound. However, whether an individual patient experiences a postoperative infection is also determined by the patient’s nutritional status, underlying health, and immune status; the size of the wound and magnitude of tissue destruction; and microbial virulence and concentration. Well-accepted risk factors for postoperative infections include the type and duration of the procedure,172,178 surgical technique,179–181 obesity,179,182,183 extremes of age,179,184 and remote infection.179,185 Surgical infection usually results from intraoperative seeding of the surgical site by endogenous bacteria or by transmission of exogenous bacteria. Almost all microorganisms causing SSIs (Table 101-5) are thought to be acquired during the operation; rarely, small clusters of SSIs may be caused by pre- or postoperative exposures.186,187 Nasal carriage of S. aureus has been associated with an increased risk of SSI after adult cardiothoracic surgery, although no standard recommendations are in practice to eradicate nasal carriage (see later discussion of prevention).188–190

TABLE 101-4. Selected Surgical Site Infection Rates by Operative Procedure and National Nosocomial Infections Surveillance (NNIS) Risk Index Score, NNIS System, January 1992 to April 2000 Risk Indexa

Procedure Cardiac surgery

b

Procedures (n)

Pooled Mean Infection Rate (%)

1 2, 3

23,731 7243

2 3

Appendectomyc

0–Yes 0–No 1 2 3

1342 5343 6808 2569 295

1 1 3 5 9

Colon surgeryd

M 0 1 2 3

384 10,751 18,856 8165 1126

1 4 6 9 11

Laparotomy

0 1 2 3

4884 5678 2999 501

2 3 5 9

Herniorrhaphy

0 1 2 3

8806 5120 1141 36

1 2 4 11

Craniotomy

0 1, 2, 3

3065 11,665

1 2

Ventricular shunt

0 1, 2, 3

2346 5562

4 5

Spinal fusion

0 1 2, 3

22,437 12,112 3134

1 3 7

Open fracture reduction

0 1 2 3

11,045 17,525 3476 394

1 1 2 5

a A point is given for each of the following risk factors: an American Society of Anesthesiologists preoperative assessment score of ≥ 3; an operation classified as contaminated or dirty; and duration of surgery > 75th percentile for the particular procedure. b Does not include coronary artery bypass graft procedures. c If the procedure was performed laparoscopically and the risk index was equal to 0, it is denoted as “0–Yes”; otherwise, it is denoted as “0–No.” d The risk index for these procedures is modified such that if the procedure was performed laparoscopically, 1 point is deducted from the standard total. When the risk index was equal to 1 and the procedure was laparoscopic, the score is given a rating of –1 or “M.”

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Healthcare-Associated Infections TABLE 101-5. Microbiology of Postoperative Infections in Surgical Patients Nature of Operation “CLEAN” SURGERY

Likely Pathogens

a

CARDIACb

Prosthetic valve, coronary artery bypass, other open -heart surgery, pacemaker implant

Staphylococcus epidermidis, Staphylococcus aureus, Corynebacterium spp., enteric gramnegative bacilli

VASCULAR

Arterial surgery involving the abdominal aorta, a prosthesis, or a groin incision

Staphylococcus aureus, Staphylococcus epidermidis, enteric gram-negative bacilli

Lower-extremity amputation for ischemia

Staphylococcus aureus, Staphylococcus epidermidis, enteric gram-negative bacilli, clostridia

ORTHOPEDIC

Total joint replacement, internal fixation of fractures

Staphylococcus aureus, Staphylococcus epidermidis

OPHTHALMIC

Staphylococcus epidermidis, Staphylococcus aureus, streptococci, enteric gram-negative bacilli, Pseudomonas HEAD AND NECK

Incisions through oral cavity or pharynx

Anaerobic bacteria, enteric gramnegative bacilli, Staphylococcus aureus

Craniotomy

Staphylococcus aureus, Staphylococcus epidermidis

ABDOMINAL

Gastroduodenalc

Enteric gram-negative bacilli, gram-positive cocci

Biliary tractd

Enteric gram-negative bacilli, enterococci, clostridia

Colorectal

Enteric gram-negative bacilli, anaerobic bacteria, enterococci

Appendectomy, nonperforated

Enteric gram-negative bacilli, anaerobes, enterococci

“DIRTY” SURGERYe

Ruptured viscus

Enteric gram-negative bacilli, anaerobes, enterococci

Traumatic woundf

Staphylococcus aureus, group A streptococci, clostridia

a Parenteral prophylactic antimicrobial agents can be given as a single intravenous dose just before the operation. For prolonged operations, additional intraoperative doses should be given every 4–8 hours for the duration of the procedure. b An additional dose can be given when patients are removed from bypass during open-heart surgery. c Morbid obesity, decreased gastric acidity, or decreased gastrointestinal motility. d Acute cholecystitis, nonfunctioning gallbladder, obstructive jaundice, or common duct stones. e For “dirty” surgery, therapy should usually be continued for 5–10 days. f For bite wounds, likely pathogens can also include oral anaerobic bacteria, Eikenella corrodens (human), and Pasteurella multocida (dog and cat). Adapted from Antimicrobial prophylaxis in surgery. Med Lett 1999;41:75–80

Clinical Manifestations and Laboratory Diagnosis A surgical wound is not infected if it heals primarily without discharge. It is infected if purulent discharge develops, even if microorganisms are not recovered from the discharge. Surgical sites that are inflamed without discharge or that drain culture-positive serous fluid

CHAPTER

101

585

are possibly infected. Traditionally, a surgeon has been given latitude to diagnose clinically as infected a wound that does not meet the these criteria.191 For surveillance purposes, NHSN classifies SSIs into incisional or organ or space (i.e., involving organs or spaces other than the incision) infections. Incisional infections are subdivided into superficial (involving only the skin and subcutaneous tissue) and deep (involving fascial and muscle layers) categories.190,192 By definition, a superficial incisional SSI must occur within 30 days of the operative procedure and meet at least one of the following criteria: (1) purulent fluid is draining from the incision; (2) an organism is isolated from culture of a specimen that was obtained aseptically from the incision; (3) there is pain, tenderness, localized swelling, redness, or heat, and the incision is opened by the surgeon; or (4) diagnosis of infection is made by the surgeon. A deep incisional SSI must occur within 30 days after the procedure in the absence of an implanted prosthesis, or within 1 year if prosthetic material was placed. In addition, it must meet at least one of the following criteria: (1) purulent fluid is draining from the deep incision but not from the organ-space component of the site; (2) the deep incision becomes dehiscent or is opened by the surgeon when the patient has fever (> 38°C), localized pain, or tenderness, unless the incision is culture-negative; (3) an abscess is found on direct examination, during operation, or by histopathologic or radiographic examination; or (4) diagnosis of infection is made by the surgeon. An organ or space infection involves deeper spaces entered or manipulated during the operative procedure; examples are subdiaphragmatic abscess after an appendectomy, or osteomyelitis after laminectomy (see Chapter 80, Osteomyelitis). An organ or space SSI must occur within 30 days after the procedure in the absence of an implanted prosthesis, or within 1 year if prosthetic material was placed. In addition, it must meet at least one of the following criteria: (1) purulent fluid is draining from a drain placed through a stab wound; (2) an organism is isolated from culture of a specimen that was obtained aseptically from the organ or space; (3) an abscess is found on direct examination, during operation, or by histopathologic or radiographic examination, or (4) diagnosis of infection is made by the surgeon.

Management and Outcome Antimicrobial agents are only one aspect of management of SSIs, along with surgical intervention and nutritional support. In pyogenic abscesses, clostridial myonecrosis, streptococcal gangrene, or other infections with extensive tissue necrosis, surgical drainage or debridement or both is the primary treatment, with antimicrobial therapy assuming a secondary role. However, in most postoperative infections in which the infected tissues are well vascularized, antimicrobial therapy is paramount. Pending culture results, empiric antimicrobial therapy for an infection from an operation in which the abdominal or genitourinary tract was entered should include agent(s) with polymicrobial activity, including activity against anaerobic bacteria. For other SSIs, S. aureus is the most likely pathogen; empiric therapy must take into account the antimicrobial susceptibility patterns of S. aureus at the facility and in the community.

Prevention Several measures to prevent postoperative SSIs were introduced into surgical practice in the last century, and they have been reviewed comprehensively.193 They include preoperative measures such as removal of hair and application of antiseptic agents to the patient’s skin, performance of surgery in a properly designed and ventilated operating room, and handwashing, gowning, and gloving by the operating room staff. Finally, the appropriate selection and timing of administration of perioperative antimicrobial prophylaxis has repeatedly been shown to reduce the risk of SSIs. Many reviews of

586

SECTION

Q Infections Associated with Hospitalization and Medical Devices

antimicrobial prophylaxis have been published, and most have similar recommendations for the procedures in which prophylaxis is desirable.194–197 The selection of specific antimicrobial agents should take into consideration the most likely pathogens and the antimicrobial susceptibility data of individual facilities. The efficacy of perioperative antimicrobial prophylaxis in decreasing the incidence of SSIs is well established for a variety of surgical procedures.196–198 Most recommendations for children are extrapolated from studies of adult surgical patients. Such extrapolation may be appropriate for older children, but among neonates and young infants, for whom SSI rates appear to be higher, specific data regarding the optimal agent for, frequency of, and duration of prophylaxis are needed.199 Although difficulties exist in establishing evidence-based recommendations for antimicrobial prophylaxis in pediatric patients, there remains a core group of well-accepted recommendations for adult patients that can reasonably be applied to children.194 In general, antimicrobial prophylaxis is only indicated for procedures for which the postoperative infection rate is high or in which prosthetic material is implanted. Such procedures include all operative procedures during which the gastrointestinal tract is entered, with the sole exception of elective surgery for duodenal ulcer disease, which has a very low rate of SSI. For procedures involving the biliary tract, prophylaxis is only indicated for acute cholecystitis, obstructive jaundice, or when there are stones in the common bile duct. Preoperative antimicrobial agents decrease the incidence of infection after colorectal surgery. Other widely accepted indications for antimicrobial prophylaxis are most cardiothoracic procedures, prosthetic joint replacement, internal fixation of a fracture, craniotomy, urinary tract procedures in the setting of bacteriuria or obstructive uropathy, and head and neck procedures with an incision through the oral or pharyngeal mucosa.190,194,196,197,200 A wide variety of agents have been recommended for perioperative prophylaxis on the basis of a vast number of studies. Selection of a particular agent should be determined by the nature of the procedure and potential pathogens, knowledge of the antimicrobial susceptibility patterns, and the patient’s age, hepatic, and renal function. For most procedures involving the gastrointestinal tract, the agent should have activity against Enterobacteriaceae and the Bacteroides fragilis group; cefoxitin (not approved for use in patients younger than 3 months) is the most widely recommended agent in adults. For almost all other operations, the most commonly recommended agent is cefazolin.196–198 Cefazolin has activity against the common grampositive pathogens recovered from postoperative infections at these sites (especially S. aureus) and has a low incidence of side effects as well as a long serum half-life. Regional changes in the prevalence of methicillin resistance among S. aureus isolates recovered from previously healthy children might require alteration of prophylactic agents in the future. Without age-specific guidelines, many centers have developed their own practices based on local consensus and available literature, but these protocols often vary greatly and reflect the controversy that continues regarding some protocols. A review of 43 academic centers performing pediatric cardiothoracic surgery showed that physicians continued prophylactic antimicrobial agents at variable rates after surgery while certain interventions remained in place, as follows: thoracostomy tube (67%), mediastinal tube (72%), transthoracic vascular catheter (51%), central venous catheter (30%), or pacing wire (14%).201 The general consensus was that first- or second-generation cephalosporins were used as single prophylactic agents and were continued until the removal of transthoracic devices. However, a prospective comparative study in a PICU showed that discontinuation of antimicrobial agents before these devices were removed did not result in higher infection rates.177 In fact, this program, which involved restriction of prophylactic agents after cardiothoracic surgery to 48 hours or 1 day after sternal closure in combination with aggressive removal of intravascular and urinary catheters, resulted in lower infection rates and better patient outcomes. It remains a challenge to determine the balance between the protective benefits of appropriate

antimicrobial prophylaxis and the potential harmful effects of injudicious use that can lead to resistance and other complications.

POSTOPERATIVE MEDIASTINITIS AND STERNAL OSTEOMYELITIS Mediastinitis is inflammation or infection of the mediastinal area (defined as the extrapulmonary area of the thoracic cavity between the lungs). This area contains the thymus, trachea and bronchi, esophagus, aorta and aortic arch, heart, pericardium, lymph nodes, and nerve tissue. Historically, infections of the mediastinum have been divided into acute (abrupt onset, with or without toxic appearance) and chronic (indolent) processes.202–204 Although acute mediastinitis is uncommon, it can be serious and life-threatening.

Epidemiology and Pathogenesis Mediastinitis after cardiac surgery is an infrequent yet serious complication of median sternal incision, with an estimated incidence ranging from 0.15% to > 5% of all cardiothoracic operations.205 In the few studies restricted to children, mediastinitis rates after cardiac surgery varied between 0.1% and 5%, with the highest rates found in neonates.206–208 Because most pediatric cardiac operations involve correction of congenital anomalies, infection rates and risk factors for infection may differ from those identified. No data are available on whether rates of mediastinitis in children vary according to type of reconstructive surgery or congenital anomaly. The pathogenesis of infection is multifactorial but generally requires intraoperative introduction of pathogenic bacteria into the operative site, making the quality and consistency of both the perioperative aseptic technique and surgical technique critical aspects of patient management.209,210 Risk factors for mediastinitis in adults include diabetes mellitus, chronic obstructive pulmonary disease, prolonged duration of perfusion (cardiopulmonary bypass) or aortic cross-clamping time, higher body mass index or obesity, preceding infections (e.g., pneumonia, tracheal infections), and receipt of corticosteroids.202,203,205,209–213 Outbreaks of mediastinitis have been associated with preoperative colonization, operating room personnel (e.g., anesthesiologist, intraoperative nurses, or surgeons), other intraoperative factors (e.g., contaminated cardioplegia solution or inadequate sterile surgical technique), and postoperative exposures.186,187,214,215 Staphylococci are the predominant pathogens causing postoperative mediastinitis.211,216,217 In children, S. aureus accounts for 38% to 96%, and coagulase-negative staphylococci for 10% to 52%, of episodes of mediastinitis in some series.207,218 Gram-negative organisms and Candida spp. are less common pathogens in children.219,220 A wide variety of pathogens have been reported to cause mediastinitis after cardiac surgery in adults, including staphylococci, Enterobacter cloacae, Escherichia coli, Klebsiella spp., Pseudomonas spp., Proteus spp., enterococci, Bacteroides fragilis, Corynebacterium xerosis, Mycoplasma spp., nontuberculous mycobacteria, Aspergillus spp., Haemophilus spp., Nocardia asteroides, N. farcinica, and Rhodococcus bronchialis.221,222

Clinical Manifestations and Laboratory Diagnosis Common signs and symptoms of postoperative mediastinitis include local tenderness, wound dehiscence, increased erythema of the wound with or without purulence, and an unstable sternum (movable, appositional edges). Infants can be merely fussy and display expiratory grunting. Fever occurs on average 5 days after surgery, and local signs occur a mean of 9 days after surgery.207,218,219 In older children, it may take longer for signs and symptoms to become apparent; in one study, infections were diagnosed at a mean of 15 days postoperatively.207 Infections due to Nocardia, Rhodococcus, Candida

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Clinical Syndromes of Device-Associated Infections

spp., or nontuberculous mycobacteria can have an indolent onset, and incubation periods > 30 days are possible. Infections associated with these organisms may have minimal local signs, only serosanguineous drainage, and little or no fever. Laboratory tests usually reveal a moderate leukocytosis with an increased frequency of polymorphonuclear cells or an elevation in the erythrocyte sedimentation rate or C-reactive protein value. The diagnosis of acute suppurative mediastinitis is suggested by radiographic evidence of widened mediastinum, mediastinal emphysema, and pleural effusions. The presence of gas in the soft tissues is highly suggestive of perforation. Computed tomography (CT) has been used to differentiate mediastinitis from mediastinal abscess223,224; however, in the absence of mediastinal gas, CT may not differentiate mediastinitis from benign postoperative changes. Among heart transplant recipients, gallium scintigraphy has been used to confirm mediastinitis when CT scan is not diagnostic.225 Gadolinium-enhanced magnetic resonance imaging (MRI) can also help delineate the extent of infection and differentiate among mediastinal tissues, masses, and inflammatory tissue.226,227 A definitive diagnosis can be made by Gram and acid-fast staining and bacterial, mycobacterial, and fungal cultures performed on specimens obtained via CT-guided aspiration. If esophageal perforation is suspected, fluoroscopy with water-soluble contrast media may aid in the diagnosis of mediastinitis and allow localization of the level of perforation.

Management and Outcome Aggressive surgical drainage and debridement are generally required to cure mediastinitis after cardiac surgery. For superficial mediastinal infections, incision, drainage, packing of the wound, and antimicrobial therapy may be effective. For deep infections, debridement with removal of infected and devitalized tissue, mediastinal irrigation, and antimicrobial therapy may be necessary. Surgical debridement (with closed-tube irrigation) and systemic antimicrobial therapy are usually sufficient. In severe infections, it may be necessary to leave the wound open, with subsequent secondary closure by transposition of omentum or muscle.228,229 For postoperative mediastinitis, selection of empiric agents should be based on the prevalent pathogens associated with cardiac SSIs in the institution and the patient’s endogenous flora if known; therapy effective for coagulase-positive and coagulase-negative staphylococci must be provided. The combination of vancomycin and an aminoglycoside is a common empiric regimen. No studies have been conducted to evaluate the optimal regimen or duration of antimicrobial therapy for mediastinitis; however, 3 to 8 weeks is generally recommended, depending on the severity of the infection and the extent of bone involvement.230 Sternal osteomyelitis can accompany severe, deep mediastinal infections. These infections most commonly follow surgery involving a median sternotomy. The pathogens responsible for sternal osteomyelitis are similar to those causing mediastinitis, and coagulasepositive and coagulase-negative staphylococci are the predominant organisms. Treatment of sternal osteomyelitis requires debridement of infected bone and a minimum of 4 to 6 weeks of antimicrobial therapy; transposition of pectoralis major or rectus abdominis muscles, or omentum, is increasingly performed to close the defect and improve blood supply to the affected site.208,231,232

Prevention Antimicrobial prophylaxis has not been shown in placebo-controlled trials to decrease the risk of mediastinitis; however, because this infection can be catastrophic, perioperative prophylaxis is frequently used because of possible benefit. It should not be extended beyond 24 hours unless an infection is suspected and appropriate diagnostic tests have been performed. Studies from centers with lower rates of mediastinitis demonstrate that rates can be reduced by: (1) strict

CHAPTER

102

587

perioperative adherence to careful aseptic technique; (2) surgical measures, including hemostasis and precise sternal closure; and (3) interventions targeted to identified risk factors.177,210,233 Some authorities have proposed the use of intraoperative ultraviolet irradiation, preoperative decolonization, or preoperative chlorhexidine washes of patients who are carriers of S. aureus to help prevent infections.188,189,233

CHAPTER

10 2

Clinical Syndromes of Device-Associated Infections Theoklis E. Zaoutis and Susan E. Coffin

Many healthcare-associated infections are related to the use of medical devices.1 This chapter addresses device-associated infections in children beyond the neonatal period. For device-related infections in the neonate, see Chapter 96, Nosocomial Infections in the Neonate. Some general principles apply to infections involving medical devices. Factors that influence the risk of developing a deviceassociated infection include the properties of the device, of potential pathogens, and of the host (Box 102-1).2 In addition, colonization and subsequent infection associated with medical devices are facilitated by biofilm formation. Biofilms are complex structures in which colonizing organisms are ensconced in a matrix of extracellular proteins.3,4 Immediately after implantation biofilms begin to form with the deposition of host proteins on the surface of the device. Microorganisms adhere, multiply, and contribute their own extracellular proteins to this coating. In addition, properties inherent to biofilms can inhibit host defenses5,6 and thus help organisms to evade host factors and antimicrobial agents.7–9 The microbiology of device-related infections varies with the type of device (Box 102-2). Organisms found as part of the normal skin flora are the most common etiologic agents in foreign-body infections in infants and children.6,10 Certain microorganisms are particularly well suited to colonizing devices. Staphylococcus aureus adheres specifically to fibrinogen, a component of biofilms.11,12 Coagulase-negative staphylococci (CONS), Pseudomonas aeruginosa, and Corynebacterium spp. produce proteinaceous slime, capsular polysaccharides, and adhesins that enhance adherence and protect the organism from host defenses and the effects of antimicrobial agents.10,13–15 Common findings in foreign-body infection include malfunction of the device (e.g., occlusion of catheters and shunts or loosening of prosthetic joints),16,17 pain at the site of the device, and failure of

BOX 102-1. Determinants of Infection of Medical Devices PROPERTIES OF THE DEVICE Material: inert, plastic, rubber Design: hollow, solid Location: implanted, percutaneous, vascular/nonvascular site PROPERTIES OF POTENTIAL PATHOGENS Source: endogenous, environment, contaminated disinfectant/device/solution Susceptibility: in vitro and in biofilm PROPERTIES OF THE HOST Host defenses: skin, mucous membranes, immune system Medication: anti-inflammatory agents, immunosuppressive agents Altered flora due to diet, medications (antacids, antimicrobial agents)

588

SECTION

Q Infections Associated with Hospitalization and Medical Devices

BOX 102-2. Pathogens Recovered from Infected Medical Devices

BOX 102-3. Routes of Infection of Medical Device

MOST COMMON Coagulase-negative staphylococci Staphylococcus aureus Other skin flora COMMON Gram-negative enteric bacilli Environmental organisms Candida spp., especially Candida albicans and Candida glabrata OCCASIONAL Other fungi Nontuberculous mycobacteria

INOCULATION AT THE TIME OF INSERTION OF THE DEVICE Endogenous flora of the host Environmental flora INOCULATION DURING MANIPULATION OF THE DEVICE Breaks in aseptic technique HEMATOGENOUS SEEDING Transient bacteremia Nonintact mucous membranes or gastrointestinal tract EXTENSION OF LOCAL INFECTION

wound healing after implantation. Fever, leukocytosis, and other systemic signs may be absent. A high index of suspicion is therefore required for timely diagnosis. Cultures are essential to establish the diagnosis and guide definitive therapy. Antimicrobial therapy alone is sometimes curative, but frequently the foreign body must be removed. The risk of device-related infections can be reduced by consistent application of several principles and practices; these practices together are sometimes referred to as “bundles.” First, devices should only be used in circumstances in which they are essential. Strict aseptic technique must be observed during insertion or manipulation of a device. In addition, limiting the frequency of device manipulation reduces the likelihood of contamination. Finally, these temporary foreign bodies should be removed as soon as they are no longer needed. Although the use of antimicrobial prophylaxis at the time of insertion is commonplace, the efficacy of such measures for many devices has not been established.18–20 Recent advances in materials used for catheters and prosthetic devices have decreased rates of infection. The use of central venous catheters (CVCs) impregnated with either minocycline and rifampin or chlorhexidine and silver sulfadiazine have been shown to be effective in reducing the incidence of both catheter colonization and catheter-related bloodstream infection (BSI) in adult patients.21–23 However, few efficacy studies of these devices have been performed in children.24–27

INTRAVASCULAR CATHETER-RELATED INFECTIONS

BOX 102-4. Pathogens Recovered in Vascular Catheter-Associated Infections MOST COMMON Coagulase-negative staphylococci COMMON Enterobacter spp. Escherichia coli Klebsiella spp. Pseudomonas aeruginosa Staphylococcus aureus Enterococcus spp. Candida spp. OCCASIONAL Other gram-negative bacilli, including Acinetobacter spp. and Citrobacter spp. Nontuberculous mycobacteria Corynebacterium spp. Bacillus spp.

United States.36,37 Other relatively common organisms are Corynebacterium, Propionibacterium, and Bacillus spp., Serratia marcescens, Acinetobacter spp, Citrobacter spp., and nontuberculous mycobacteria.30 Children with CVCs managed predominantly in home healthcare settings are at higher risk for nonendogenous (e.g., from water and other environmental sources) gram-negative pathogens, such as Pseudomonas, Acinetobacter, and Agrobacterium spp., particularly during summer months.38,39

Epidemiology and Pathogenesis Commonly used in hospitals for more than a decade, intravascular catheters are now also widely used in outpatient and homecare settings. Catheter-related infections represent the most important healthcare-associated infection experienced by children. Critically ill pediatric patients who develop catheter-related BSIs have longer duration of intensive unit and hospital stay. In addition, the crude mortality has been estimated to be 19%.28 The attributable cost of a primary nosocomial BSI for critically ill pediatric patients was estimated to be $40 000.29 Several different types of catheter-related infections can be identified. Infection can occur at the exit site, in the subcutaneous tunnel or pocket, or in the bloodstream.30 Pathogens can be introduced during insertion, manipulation, calibration, and flushing of monitoring devices, or via contaminated parenteral fluids, flush solutions, or skin antiseptics (Box 102-3).31–34 The organisms most commonly isolated in intravascular catheterassociated infections are skin organisms, particularly CONS, which account for approximately 40% of pathogens isolated from patients with CVC BSI. Gram-negative aerobic bacilli account for approximately 25% of nosocomial BSI, followed by Staphylococcus aureus and enterococci.28,35 The most commonly encountered gramnegative bacilli include Enterobacter spp., Pseudomonas aeruginosa, Klebsiella pneumoniae, and Escherichia coli. (Box 102-4). Over the last two decades, Candida spp. have become increasingly prevalent and are now the fourth most common cause of nosocomial BSI in the

Catheter-Related Factors Infection risk varies according to the type of catheter used. Rates are higher for CVCs of all types than for peripheral venous catheters or arterial catheters.30 The focus of this section is restricted to CVCs. Although some studies have shown similar infection risks for tunneled and nontunneled catheters,40,41 later randomized studies noted that tunneling of femoral catheters42 and internal jugular catheters43 is associated with lower infection risk (Box 102-5). In recent years, peripherally inserted central catheters (PICCs) are used more commonly. In the outpatient setting, PICCs have been shown to have a relatively low risk of infection. However, one analysis of critically ill patients revealed that the risk of catheter infection was similar among patients with a PICC and those with a centrally inserted percutaneous catheter.44 Additional study is needed to address the relative risk of infection associated with various modes of central venous access in critically ill patient populations. A correlation between increasing number of lumens and incidence of infection has been noted in some studies, but not others.45,46 In addition, the presence of multiple venous catheters was found to be an independent risk factor associated with nosocomial BSIs.28 Adult studies have demonstrated that the site of catheter placement also influences the risk of infection. Organisms present at the site of catheter insertion can move from the entry site along the dermal tunnel to the catheter tip within hours.47 The site of placement determines the types and burden of organisms in proximity

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Clinical Syndromes of Device-Associated Infections

BOX 102-5. Risk Factors for Vascular Catheter-Associated Infections CATHETER Site of insertion Manipulations: entry into the system Duration of catheterization Thrombosis Lumens, stopcocks, monitoring devices Antiseptic/antibiotic-coated or not Tunneled or not; implanted or not INFUSANT Parenteral nutrition Lipids Blood HOST Skin integrity Skin flora Immune competence

to the catheter entry site. Quantitative cultures of potential catheter sites in adults and neonates show that the upper extremity and chest wall have the lowest burden of bacteria, whereas the groin and jugular areas are more heavily colonized.48 In adults, catheters placed in the groin are at greater risk of infection than those placed in the neck.49 The impact of site selection on infection risk remains unclear among children.50–52 Catheters that are coated or impregnated with antimicrobial or antiseptic agents can decrease the risk for infection and have been shown to decrease hospital costs, despite the additional expense of these coated catheters.21 However, the anti-infective effect of these catheters is not permanent and may be limited to relatively short times after insertion (i.e., 1 to 2 weeks). Data are lacking on the efficacy of these catheters for longer insertion times.53 To date, the efficacy of these catheters has not been demonstrated in children. Nonetheless, these catheters are approved by the Food and Drug Administration for use in patients weighing > 3 kg and are in use in some pediatric settings. The duration of catheter placement is directly related to the risk of infection. Biofilm formation, which increases with dwell time, clearly increases the opportunity for catheter colonization. In addition, each manipulation of a catheter, stopcock, or needleless device provides an opportunity for the introduction of organisms.30,54,55 Outbreaks of bacteremia have resulted from contamination of pressure transducers.56–59 Replacing the catheter over a guidewire does not decrease the incidence of infection, and may even increase it.60 Finally, current studies show similar rates of infection when a transparent or a gauze dressing is used.61,62 Certain infusates are more likely than others to be associated with BSI. Use of parenteral nutrition fluids increases the risk of catheterrelated BSIs, particularly those caused by Candida spp.63 Infusion of lipid emulsion was associated with a dose-dependent increase in the risk of CONS bacteremia in one study of infants.64 Intrinsically or extrinsically contaminated infusates can cause catheter-related infections. Intrinsic contamination of intravenous fluids is rarely reported in the United States; episodes occur more commonly in international settings. In contrast, episodes of extrinsic contamination of intravenous fluids and medications continue to occur in the United States and throughout the world.

Impact of Host Factors Host factors can influence the risk of primary and secondary catheter infections. Skin flora varies according to a person’s hygiene and the medications being used. Topical antimicrobial agents may alter the nature and density of local flora. Systemic antimicrobial agents also alter stool and skin flora and increase the risk of infection by Candida spp.63 Factors that lead to a break in the integrity of skin or mucous

CHAPTER

102

589

membranes can increase the risk of catheter infections. Thus, patients with immature skin (premature infants), impaired integrity of the gut epithelium (patients with short-gut syndrome),65,66 or burn patients are at high risk of bacteremia and subsequent colonization of an indwelling catheter and secondary BSI. Similarly, patients with a distant site of infection (such as an abscess) or who develop a spontaneous bacteremia are also at risk of developing a secondary catheter-related infection. Other host factors influence the risk of infection. Children undergoing cancer chemotherapy are at risk of infection because of neutropenia; infection rates in these patients are increased two- to fourfold by the use of indwelling central catheters.67 In addition, mucositis raises the risk of bacterial gastrointestinal wall translocation and bacteremia. These factors likely account for differences in the microbiology of BSIs that occur in neutropenic, compared with nonneutropenic, patients. In a prospective study of over 24 000 episodes of nosocomial BSIs, researchers noted that neutropenic patients were at greater risk of infection due to Candida spp., enterococci, and viridans group streptococci.68 Patients with human immunodeficiency virus (HIV) infection are at greater risk for a variety of infections; use of CVCs for nutritional support and for medications increases the risk of bacteremia in HIV-infected and other patients.69,70 Patients with uremia are at higher risk of catheter-related infection, particularly those with hemodialysis catheters.71–73 In patients undergoing hemodialysis, the type of bloodstream access used influences the risk of infection; the risk is lowest with native arteriovenous fistulae, intermediate with artificial arteriovenous grafts, and highest with CVCs.74,75 Clinical factors that are likely related to the severity of illness have also been found to influence the risk of infection. In a cohort of critically ill pediatric patients, the need for transport out of the unit and for a procedure performed within an intensive care unit were both independently associated with a three- to fourfold increase in the risk of developing a nosocomial BSI.28

Clinical Manifestations and Laboratory Diagnosis Fever is often, but not universally, present in patients with catheterrelated infections. Patients with localized infection of the exit site, pocket, or tunnel frequently have infection in the absence of fever. When a catheter-related infection is suspected the exit site should be examined for the presence of local suppuration or cellulitis; however, normal appearance of the exit site does not exclude a local or systemic infection involving the catheter. Any catheter malfunction (including decrease in flow or unidirectional flow) should provoke an evaluation for potential infection. Findings suggestive of disseminated infection include the presence of emboli in the retina, skin, bone, and viscera (lungs, kidneys, liver, and spleen) and organ dysfunction due to immune complex deposition (e.g., nephritis). Infections attributable to a CVC include exit, tunnel, and pocket infections as well as CVC BSI (Table 102-1).76 A CVC BSI is defined as bacteremia or fungemia in a patient with a centrally placed intravascular catheter in which the catheter is the presumed source of infection. Establishing that the catheter is the source of infection is not always straightforward. For example, a BSI in a patient with an indwelling catheter can originate from an undocumented source of infection (e.g., a postoperative wound infection or a urinary tract infection) rather than from the catheter. In adult patients, only 15% to 20% of CVCs removed in the context of a BSI are ultimately implicated as the primary source of infection. Several methods have been used to diagnose CVC BSI, including simultaneous quantitative cultures of blood obtained through the catheter and a peripheral vein, quantitative and semiquantitative cultures of a catheter segment, and differential time to blood culture positivity (see Table 102-1). A CVC BSI can be diagnosed when there is an incremental increase in the quantity of bacteria obtained from simultaneous blood cultures from the catheter and peripheral blood (see Table 102-1). Because many institutions have recently adopted

590

SECTION

Q Infections Associated with Hospitalization and Medical Devices

TABLE 102-1. Types of Catheter-Related Infections Infection

Clinical Diagnosis

Exit site infection

Erythema or induration within 2 cm of catheter exit site

Tunnel infection

Tenderness, erythema, or induration along the subcutaneous tract of a tunneled catheter and more than 2 cm from catheter exit site

Pocket infection

Purulent fluid in the subcutaneous pocket of a totally implanted venous access device. May be accompanied by overlying tenderness, erythema, induration, visible drainage, and skin necrosis

Catheter-associated bloodstream infection

Positive simultaneous blood cultures from the central venous catheter and peripheral vein yielding the same organism in the presence of at least one of the following: • Simultaneous quantitative blood cultures in which the number of cfus isolated from blood drawn through the central catheter is at least fivefold greater than the number isolated from blood drawn peripherally • Positive semiquantitative (≥ 15 cfu/catheter segment) or quantitative (≥ 100 cfu/catheter segment) catheter tip cultures • Simultaneous blood cultures in which the central blood culture has growth in an automated system ≥ 2 hours earlier than the peripheral blood culture

cfu, colony-forming unit.

automated blood culture systems, this technique is now less commonly used. Until recently, the most commonly accepted methods of diagnosing a CVC BSI have involved either quantitative or semiquantitative cultures of the catheter tip. This strategy relies upon the prompt removal of the catheter and is thus not appropriate in many clinical settings. A positive catheter tip culture result is defined as growth of at least 15 colonies (cfu) on an agar surface over which the catheter was rolled.77 An alternative method involves sonicating the catheter tip in fluid to dislodge the intraluminal biofilm; this method is associated with a higher rate of positive cultures than the roll technique.78 A catheter tip culture that yields at least 15 colonies is a valid indication of catheter colonization, which in turn is associated with a higher risk of bacteremia. However, a relatively small number of patients with colonized catheters have catheter-related septicemia.79 Thus, the positive predictive value (PPV) of a positive catheter tip culture result is low, ranging in most studies from 5% to 30%. Also, this method has not been evaluated for antimicrobial agent- or antiseptic-impregnated catheters. Electron microscopy of catheter tips reveals universal colonization with bacteria.78,80 Differential time to positivity of paired blood cultures is an alternative method to evaluate catheter-related BSI. This technique is the simplest of the three methods and does not require either specialized laboratory culture methods (other than a continuousmonitoring blood culture system) or catheter removal. This method is only accurate if the peripheral and central cultures are drawn simultaneously and are of equal volume. If the same organism is isolated from both cultures and the time to positivity of the catheterobtained specimen is > 2 hours shorter than the peripherally obtained culture, catheter colonization and CVC BSI are likely.81 Compared with quantitative and semiquantitative methods, differential time to positivity of > 2 hours had a sensitivity of 93% and a specificity of 75% for catheters in place for > 30 days and a sensitivity of 81% and a specificity of 92% for catheters in place for < 30 days.82 Several smaller studies have demonstrated similar results.83–85 The diagnosis of a CVC BSI can be more difficult in children than adults because current National Nosocomial Infection Surveillance (NNIS) definitions for catheter-related BSI require that data be obtained from cultures of a catheter tip or the peripheral blood in addition to blood cultures obtained through the catheter. These cultures are frequently difficult to obtain in pediatrics for a variety of reasons. Attempts are being made to develop and validate alternative methods for diagnosis of CVC BSI in children. Investigators recently found that analysis of semiquantitative cultures obtained from different lumens of multilumen catheters yielded a high PPV for CVC BSI. In pediatric oncology patients with a double-lumen catheter in place and no peripheral blood culture available, the PPV of greater

than fivefold difference in cfu mL of isolates from samples from the two lumens was 92.2% predictive of a CVC BSI.86 Few studies have evaluated the optimal timing of blood cultures. In clinical conditions with continuous bacteremia, such as endocarditis or septic thrombophlebitis, this issue is less relevant. In CVC BSI, bacteremia can be intermittent when bacteria residing within the catheter lumen are not continuously exposed to the intravascular space. Obtaining multiple samples over a 24-hour period appears to increase the ability to detect intermittent bacteremia compared with obtaining multiple specimens at the same time.85,87 In critically ill patients who are hemodynamically unstable, two sets of blood cultures should be drawn promptly prior to initiation of empiric antibiotic treatment. In less urgent cases, blood should be drawn at least twice within a 24-hour period prior to beginning empiric therapy. In patients already receiving antibiotics, samples obtained close to the time that antibiotic concentrations have reached trough levels (i.e., just before next dose) could theoretically improve recovery of organisms in blood cultures;88 however, this issue has not been studied and may not be practical clinically. The magnitude of bacteremia affects blood culture yield, especially when small blood volumes are used to inoculate blood culture bottles. The likelihood of growth is lower and the time to detection is delayed when small volumes (< 0.5 mL) of blood are used to inoculate blood culture bottles.89 Although multiple cultures enhance the recovery of pathogens, the volume of blood cultured is more important than the total number of blood cultures obtained. One study found that the pathogen recovery rate at 24 hours was 72% for a large-volume (6 mL) single culture compared with a 47% combined yield of two smaller (2 mL) samples inoculated into separate culture bottles.90 Similar results were found in studies of adult patients in which standard “adult-volume” cultures (mean, 8.7 mL) had a higher detection rate (92%) than “low”-volume cultures (mean, 2.7 mL) where the detection rate was only 69%.91 These investigators estimated that the yield of adult blood cultures increased approximately 3% per mL of blood cultured. Too much blood inoculum (rarely an issue in pediatrics) as well as too little blood can influence culture yield. Diluting the blood into the blood culture broth enhances recovery of pathogens, perhaps by diluting antimicrobial agents (if applicable) and blood components such as phagocytes, antibodies, and complement factors that are known to have bactericidal activity.92 The ideal blood-to-broth ratio depends on the blood culture system used but a ratio between 1:5 and 1:10 is generally considered optimal.93 Certain bacteria as well as fungi and mycobacteria require special handling of blood cultures. For example, filamentous fungi require solid media for growth, Malassezia furfur requires lipid supplementation, and certain gram-

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Clinical Syndromes of Device-Associated Infections

negative species and some mycobacteria need prolonged incubation (see Chapter 286, Laboratory Diagnosis of Infection Due to Bacteria, Fungi, Parasites, and Rickettsiae). Failure of culture to identify a pathogen in the patient with continued clinical evidence of infection should prompt consideration of less common, more fastidious organisms and discussion with microbiology laboratory personnel about additional collection and culture techniques. Distinguishing between cultures that represent “contamination” from those that represent “true” BSI can be challenging, especially when a skin organism (e.g., CONS, Bacillus species, micrococci) is isolated. Obtaining multiple cultures can clarify most situations. A study using a mathematical model of blood cultures positive for CONS in patients with a CVC found that the PPV of a single positive culture (if only one culture was obtained) was 55%. However, if only 1 of 2 cultures obtained was positive, the PPV was 20%; if only 1 of 3 cultures was positive, the PPV was only 5%. Investigators developed a similar model for blood cultures obtained from a peripheral vein. If 2 of 2 cultures were positive, the PPV was 98% if both samples were obtained from a peripheral vein, 96% if 1 sample was obtained through a catheter and the other was obtained by through the vein, and only 50% if both samples were obtained through a catheter.94 Furthermore, the distinction between pathogen and contaminant is affected by age and underlying condition(s). CONS as a true pathogen in neonates has been well described but the issue remains controversial.95 Because the isolation of CONS from a blood culture often results in a clinical intervention and administration of vancomycin, it is particularly important to obtain multiple blood cultures. Overuse of vancomycin has implications for the continued increase of vancomycin resistance among gram-positive organisms.96–98 Repeatedly positive blood culture results suggest the presence of an intravascular focus of infection such as the catheter itself or complicating suppurative thrombophlebitis or endocarditis (especially if the catheter has been removed).99 Among patients with intravascular catheters, the risk of endocarditis is highest with Swan-Ganz catheters and lowest with peripheral catheters. In an autopsy series, 53% of patients who had undergone pulmonary artery catheterization had right-sided endocardial lesions; 7% had infective endocarditis.100 Centrally placed catheters often traverse and damage the tricuspid valve, increasing the risk of right-sided endocarditis. Thrombi within the heart can become infected or can obstruct outflow. Rapid diagnosis of bacteremia can be achieved by direct staining of blood withdrawn from the catheter; this technique is highly specific but not sensitive.101 Sensitivity is increased by lysis of blood with hypotonic saline followed by centrifugation and staining of the sediment with acridine orange. If the acridine orange stain reaction is

CHAPTER

102

591

positive, a Gram stain is performed to characterize the pathogen. In one study, this technique was 87% sensitive and 94% specific in detecting bacteremia in patients with central catheters.102 Gram stain of the catheter tip has also been shown to be sensitive and specific for the detection of catheter colonization.79 Ultrasonography is useful in detecting thrombosis of vessels or formation of vegetations. The value of a negative transcutaneous ultrasonography result is debated; in adults, the use of transesophageal echocardiography increases the likelihood of detecting intracardiac vegetations.103 Unfortunately, children often require sedation or general anesthesia before the transesophageal echocardiography can be performed. Gadolinium-enhanced magnetic resonance venography can be useful in detecting central venous thromboses.104

Management and Outcome Exit site infections can be treated by removal of the catheter and local care. If the catheter remains in place, local topical therapy can be successful in controlling this type of infection. However, systemic antimicrobial therapy is typically required if an exit site infection extends to involve the soft tissues that surround the catheter (i.e., the subcutaneous tunnel or pocket) or is associated with a catheter-related BSI. Limited data guide the management of CVC BSI in children. Even in adults, there are no randomized or controlled studies to address optimal management of CVC BSI.76 Empiric therapy in children with suspected CVC BSI should include an antimicrobial agent with activity against gram-positive bacteria, such as nafcillin, oxacillin, or vancomycin, and an agent effective against gram-negative bacteria, including Pseudomonas species, such ceftazidime or cefipime with or without an aminoglycoside. The empiric use of both an antipseudomonal beta-lactam and an aminoglycoside may be appropriate in severely ill patients or when infection with a resistant gram-negative organism is suspected. In institutions in which methicillin-resistant isolates of Staphylococcus aureus are prevalent, the use of vancomycin is appropriate. Fluoroquinolones are commonly used in adults but have been approved for only limited indications in children. Limited data guide clinical decisions regarding the need for catheter removal (Table 102-2). In adults with CVC BSI, it is recommended that most nontunneled CVCs should be removed.76 In children, removal of a catheter may not always be feasible because of the potential for complications associated with reinsertion and limited vascular access sites. Therefore, treatment of CVC BSI without removal of the catheter is often attempted. In patients with CVC BSI associated with a tunneled catheter or implantable device such as a

TABLE 102-2. Management of the Catheter in Patients with a Central Venous Catheter (CVC)-Related Infection Type of Infection

Catheter Management

Exit site infection

Remove CVC if: • No longer required • Alternate site exists • Patient critically ill (e.g., hypotension) • Infection due to Pseudomonas aeruginosa or fungi

Tunnel infection

Remove CVC

Pocket infection

Remove CVC

Catheter-related blood stream infection

Remove CVC if: • No longer required • Infection caused by Staphylococcus aureus, Candida species, or mycobacteria • Patient critically ill • Failure to clear bacteremia in 48–72 hours • Persistent symptoms of bloodstream infection beyond 48–72 hours • Noninfectious valvular heart disease (increased risk of endocarditis) • Endocarditis • Metastatic infection • Septic thrombophlebitis

592

SECTION

Q Infections Associated with Hospitalization and Medical Devices

port, the decision to remove the catheter is more complicated. However, it is strongly recommended based on good evidence that in patients with evidence of a tunnel infection or pocket infection (the subcutaneous pocket of an implanted device), the catheter should be removed.76 When culture data are available, treatment decisions can be tailored to the specific organism isolated. Several studies have reported successful treatment of CVC BSI without catheter removal depending on the pathogen identified.105–108 Few data exist regarding the duration of antibiotic therapy for CVC BSI. The duration of therapy depends in part on the pathogen, whether the catheter is removed, and whether infection is complicated by septic thrombosis, endocarditis, osteomyelitis, or other metastatic foci of bacteria. For complicated infections, the duration of therapy is based on the length of therapy needed to treat the suppurative complication. There are no data to determine the optimal duration of intravenous versus oral antibiotics for the treatment of CVC BSI. Certain antibiotics with excellent oral bioavailability may be considered an alternative to parenteral therapy once a patient has cleared the bacteremia and has shown clinical improvement. Obviously, compliance with oral therapy is of the utmost importance. Some pathogen-specific recommendations are provided below.

Coagulase-negative staphylococci CONS is considered less virulent than other pathogens that cause CVC BSI and CONS infections usually present with fever alone or with inflammation of the catheter exit site. CONS CVC BSI can resolve with catheter removal alone. Some experts recommend a short course of antibiotic therapy, 3 to 5 days, even after removal of the catheter.109 If the catheter is retained, the recommended duration of treatment is 10 to 14 days after a negative culture from blood taken from the CVC has been obtained. In neonates with CONS bacteremia, treatment without removal of the catheter can be attempted. However, if a neonate has 3 positive blood cultures despite appropriate antimicrobial therapy, the catheter should be removed because of the increased risk for end-organ damage.110 The relapse rate in adult patients with CONS CVC BSI is 20% if the catheter is not removed, compared with 3% if the catheter is removed.111

CVC BSI caused by gram-negative bacilli have been successfully treated without catheter removal.76 However, catheter removal has been shown to be beneficial in the treatment of infections with specific gram-negative bacilli such as Pseudomonas spp., Burkholderia cepacia, Acinetobacter baumanni, and Stenotrophomonas spp.34,76 In a recent study of adult patients with CVC BSI caused by gramnegative bacilli, catheter removal was associated with a reduced rate of relapse.116 In general, antimicrobial therapy should be administered for 10 to 14 days after blood cultures become negative.76

Fungi Treatment of fungemia without removal of the catheter has been associated with poor outcomes in children and adults.117–119 However, these studies have not accounted for confounding effect of the severity of illness.120 Failure to remove the catheter promptly can lead to prolonged candidemia, which in turn has been associated with higher rates of disseminated infection. Among 153 children with candidemia at the Children’s Hospital of Philadelphia, the overall rate of disseminated candidiasis (lung, liver, spleen, eye, brain, heart) was 17%; the crude mortality was 26%.121 The consensus opinion is that catheters should be removed in patients with candidemia whenever clinically feasible.122 Patients with candidemia should be treated with amphotericin B or fluconazole for at least 2 weeks. Newer azole and echinocandin class antifungal agents are currently under study in children and will likely have a role in the future. Patients with candidemia should have an ophthalmologic examination to evaluate for candidal endopthalmitis, preferably after the infection is controlled and further disseminated disease is unlikely.122

Nontuberculous mycobacteria Infections due to nontuberculous mycobacteria should be managed by removal of the catheter and antimicrobial therapy. The choice of agent is individualized. The combination of cefoxitin and amikacin,123 or other drugs, such as quinolones and macrolides,124 should be considered depending on the species of pathogen and its antimicrobial susceptibility.

Staphylococcus aureus Serious complications, including endocarditis and other deep-tissue infections, have been reported in association with S. aureus CVC BSI.34 Adult patients with S. aureus bacteremia (with medical devices) are at significant risk for endocarditis and often have echocardiography performed routinely as part of their management.112,113 In contrast, the frequency of infective endocarditis is low in children with structurally normal hearts who have S. aureus bacteremia; therefore, echocardiography is not routinely recommended.114 Echocardiography should be considered in children with prolonged bacteremia prior to treatment or persistent bacteremia while receiving appropriate antimicrobial therapy or in whom a new murmur is identified on physical examination. In a prospective study of 51 children with S. aureus bacteremia, definite or possible endocarditis was diagnosed in 52% of patients with congenital heart disease but in only 3% of those with structurally normal hearts.114 Neonates may be more vulnerable than older infants to complications of S. aureus BSI.110 Some investigators recommend catheter removal for neonates for a single positive blood culture for either S. aureus or a gram-negative bacillus, as this significantly improves outcome. Two weeks of appropriate antimicrobial therapy, chosen based on susceptibility test results, is recommended for uncomplicated S. aureus CVC BSI.76 Longer durations of therapy may be necessary for patients with prolonged bacteremia (> 3 days), persistent fever, or complicated infection.115

Gram-negative bacilli There are a paucity of data addressing need for catheter removal in patients with CVC BSI caused by gram-negative bacilli. Children with

Complications Treatment is complicated in patients with catheter-related thrombosis. Phlebitis of peripheral veins is commonly delayed in onset, often becoming apparent 24 hours or longer after removal of the catheter.125,126 Suppurative thrombophlebitis sometimes resolves after catheter removal. However, fluctuance at a subcutaneous venous site, persistent erythema, swelling, and tenderness, or persistent bacteremia despite antimicrobial therapy are clues that surgical drainage or resection of the vein may be necessary for cure.127 CVC-related thrombosis is surprisingly common. In one study of children who had CVCs for home parenteral nutrition, venography revealed blockage of central vessels in 66%.128 In a study of femoral venous catheters, evidence of thrombosis was found in 69% of patients. Positive blood culture results were obtained in 38% of patients who had thrombosis versus 3% of patients who did not. The incidence of thrombosis was lower in patients who had heparinbonded catheters.129 If a patient with a thrombosed vessel develops bacteremia, the clot can become infected and lead to thrombophlebitis and persistent bacteremia.130 The optimal treatment for septic thrombophlebitis of a CVC is unclear. Removal of the catheter and prompt therapy with appropriate antimicrobial agents are essential; surgical ligation may also be necessary. Although urokinase has been demonstrated to be effective in restoring catheter patency in up to 75% of patients with occluded catheters,131 the role of anticoagulation or thrombolytic agents in the face of a catheter infection is uncertain. Complications of catheter-related infection include sepsis syndrome, endocarditis, septic thrombophlebitis, and metastatic infection with seeding of bone and organs (lung, kidneys, liver, spleen,

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Clinical Syndromes of Device-Associated Infections

brain, skin). The prognosis for patients with a catheter-related infection is influenced by factors related to the pathogen, host, and therapy. If the infection is recognized and appropriate antimicrobial therapy is promptly instituted, there can be a high rate of cure, and frequently salvage of the catheter. Delayed recognition or therapy of an infection is associated with increased morbidity and mortality. One study demonstrated that only the presence of underlying malignancy or immunodeficiency was independently associated with an increased risk of mortality due to a CVC BSI.132 Overall, fatality rates for CVC BSI are 10% to 25%, with rates as high as 50% in critically ill patients and neonates.133,134 Relapse of catheter-associated infection is unusual, but reinfection with a new organism is more common in catheters that have previously been infected.115,123 Frequent recurrences should raise concerns regarding the care of the catheter, particularly in children who are outpatients. Munchausen syndrome by proxy has been implicated as a cause of recurrent central catheter septicemia.135

Prevention The simplest way to reduce the risk of intravascular catheterassociated infection is to avoid catheterization (Box 102-6). However, if a catheter is essential, many practical measures can be taken to lower the risk of intravascular catheter-associated infection.136,137 Strict asepsis and the use of maximal barrier precautions (the wearing of sterile gloves, long-sleeved sterile gown, mask, and cap, and the use of a large sterile sheet drape) during insertion of central catheters help reduce infection risk substantially.138 These measures have been shown to be cost-effective over a wide variety of clinical conditions.139 Careful skin antisepsis before insertion is an important preventive measure. Although povidone-iodine is widely used, three randomized trials have demonstrated that 2% chlorhexidine is superior for preventing CVC BSI.140–142 Routine changing of percutaneously inserted lines over guidewires is associated with a higher risk of infection and should be avoided.60,136

BOX 102-6. Prevention of Vascular Catheter-Associated infections GENERAL • Avoid unnecessary catheter insertion • Remove unneeded catheters as soon as possible CATHETER INSERTION • Apply skin site antisepsis: chlorhexidine or tincture of iodine • Use strict aseptic technique • Use sterile technique and maximal barrier precautions when inserting central venous or arterial catheters CATHETER CARE • Minimize manipulation of catheters • Wash hands before and after palpating, inserting, replacing, or dressing any catheter • Inspect hub design; avoid preslit cover • Prepare hub before injection or aspiration through the hub • Replace end caps frequently CATHETER SELECTION • Select a catheter with the fewest lumens needed for management of the patient • Consider use of antiseptic- or antimicrobial-coated central catheters and silver-impregnated, collagen-cuffed catheters when infection rates remain high despite other measures (e.g., maximal barrier precautions) • For long-term (> 30 days) access in children older than 4 years, consider peripherally inserted central catheters or tunneled or totally implantable devices; for children younger than 4 years, consider totally implantable devices STOPCOCKS • Cover all openings • Minimize use MONITORING DEVICES (TRANSDUCERS, ETC.) • Sterilize if being reused; avoid contamination during use

CHAPTER

102

593

Attention to catheter care is essential to preventing infection. Infection rates may be lower when catheter care is provided by a specially trained team.143 Catheter site dressings should not be changed at fixed intervals but rather when they are loose, soiled, or damaged; the type of dressing (transparent or gauze) is not crucial in determining infection risk.61,62 Consistent and careful disinfection of catheter hubs is also essential to prevent catheter infections. Catheter hubs have repeatedly been demonstrated to be common sites of catheter colonization.144 Thus the hub of the catheter should be cleansed each and every time it is accessed. Both 70% ethanol and 1% to 2% tincture of iodine are equally effective;145 70% ethanol was found to be superior to 1% chlorhexidine in one study.146 Intravenous delivery systems need to be changed no more frequently than at 72-hour intervals,147–149 unless blood- or lipid-containing solutions are administered, in which case tubing should be changed within 24 hours after the infusion is initiated.136,150 Several studies suggest that the use of needleless connection systems reduces the risk of needlestick injuries to healthcare workers, but if appropriate infection control practices are not fully implemented, such systems may be associated with a higher risk of BSI.151–155 Practices to reduce such infections include replacing the end cap every 48 to 72 hours, using continuous rather than intermittent infusions,151 and educating caregivers (including family members) about aseptic technique and avoiding exposing the catheter to tap water.153 If needleless systems are used, hubs should be carefully disinfected before each access.156 If stopcocks are essential for patient care, the ports must be covered at all times, and sterile technique must be used when the bloodstream is accessed via the port. Catheters that are no longer needed should be removed promptly. The presence of an idle catheter is very common, accounting for 20% of catheter days in one study, and poses an unacceptable risk to the patient.157 Antimicrobial prophylaxis for CVC-associated BSIs is controversial. Short-course, perioperatively administered (during catheter insertion) intravenous vancomycin has not been demonstrated to reduce the risk of BSI and is not recommended.137,158 Continuous infusion of vancomycin through the CVC was effective in reducing CONS BSI in neonates in one study but did not reduce either length of stay in the intensive care unit or mortality.159 Moreover, routine use of systemic vancomycin for prophylaxis carries the risk of inducing vancomycin resistance and is not recommended.160 A proposed alternative to systemic antibiotics is the antibiotic lock technique: a minute quantity of an antimicrobial agent is confined to the catheter lumen and theoretically should pose less risk of inducing resistance. Studies of tunneled catheters161,162 and nontunneled catheters163 demonstrated that a vancomycin lock solution was associated with longer time to first episode of BSI by vancomycin-susceptible microorganisms. A large, double-blind randomized trial in 1513 children with tunneled cuffed CVCs or ports showed an 80% reduction in BSIs with routine use of a lock solution containing vancomycin, ciprofloxacin, and heparin.164 Vancomycin was undetectable in the serum of these patients, and there was no infection or colonization with vancomycin-resistant organisms. In some studies, infection rates for totally implanted ports have not been lower than those for Hickman or Broviac catheters, whereas in other studies, patients with implanted CVCs had significantly lower rates of infection.165–167 Catheters impregnated with antimicrobial or antiseptic agents, such as chlorhexidine-silver sulfadiazine27 or minocycline-rifampin,168 have been shown to be associated with lower likelihood of infection. In general, these studies have been conducted in high-risk populations (such as patients in intensive care units and patients with cancer) and in units with higher rates of BSI. Some clinicians have recommended the use of these catheters in clinical situations in which the incidence of BSI is > 3.3 per 1000 catheter days despite implementation of other preventive measures.136 A chlorhexidine-impregnated disk affixed directly to the skin surrounding the catheter was shown in a randomized trial performed in adult patients to be protective against CVC BSIs.169 In a randomized study performed in children with cardiac disease, use of a

594

SECTION

Q Infections Associated with Hospitalization and Medical Devices

BOX 102-7. Routes of Infection in Prosthetic Valve Endocarditis

BOX 102-8. Diagnosis of Prosthetic Valve Infection

EARLY INFECTION Inoculation during insertion Seeding from catheter-associated bacteremia Spread from infected wound LATE INFECTION Hematogenous spread: bacteremia due to tooth/gum disease, surgical procedures, infection elsewhere

PHYSICAL FINDINGS New murmur Congestive heart failure Arrhythmia Skin lesions (Osler nodes, Janeway lesions) Petechiae; splinter and subconjunctival hemorrhages LABORATORY FINDINGS Positive blood culture Elevated C-reactive protein/erythrocyte sedimentation rate Elevated white blood cell count Microscopic hematuria ECHOCARDIOGRAM Cardiac failure/valve dysfunction Vegetation

chlorhexidine-impregnated dressing reduced the rate of catheter colonization but not of catheter-related BSI.170 Luminal contamination may be reduced through use of closed catheter connection systems171 or a self-disinfecting CVC hub (containing iodinated alcohol).172 There are surprisingly few guidelines for outpatient management of vascular catheters. Issues that require further study include the safety of bathing or swimming, the optimal frequency of dressing, cap, and tubing changes, and optimal care of the insertion site. One study has suggested that children with subcutaneously implanted catheters who swim (while the catheter is not being accessed) are not at higher risk of infection.173 Nevertheless, when patients are discharged with indwelling venous catheters, it is important that they and their parents or caregivers understand the need to protect the catheter and catheter site and avoid their contamination with tap water.

INFECTIONS OF PROSTHETIC VALVES, PATCHES, AND VASCULAR GRAFTS Epidemiology and Pathogenesis Prosthetic patch and valve infections are classified as early (occurring ⭐ 60 days after implantation) or late (occurring > 60 days after implantation). Agents causing early infection are presumed to have been introduced at the time of the surgery or soon thereafter (Box 102-7). Early infections are most commonly caused by CONS or Staphylococcus aureus.10,174–176 Other pathogens are Enterococcus spp., Corynebacterium spp., fungi, and gram-negative enteric bacilli.177,178 Organisms rarely responsible for early infection of prosthetic valves, patches, or vascular grafts include Mycoplasma hominis,179 Brucella melitensis,180 Legionella spp.,181 Pasteurella multocida,182 and Kingella kingae.183 Late infection is typically the result of bacteremic seeding of the prosthetic material and is usually caused by oral streptococci. Bacteremia in patients with prosthetic valves is often related to medical devices, particularly vascular catheters. In a prospective study of patients with prosthetic valves who had nosocomial bacteremia, 43% experienced endocarditis.184 The rate of infection of prosthetic valves is 1% to 4% for early infection and about 1% annually for late infections (after the prosthesis has endothelialized).175,176,185 Reported rates of infection of vascular grafts range from 1% to 6%. Grafts in the lower extremities, especially the groin, are more likely to become infected. Hemodialysis grafts and fistulae are most often infected with S. aureus. Infection with gram-negative bacilli or nontuberculous mycobacteria is often associated with improper sterilization of hemodialyzers.186–190

Clinical Manifestations and Laboratory Diagnosis Patients with infected prosthetic valves or patches typically present with fever. Hemodynamic changes can also be present in patients infected with virulent organisms (Box 102-8).175,176 Valve ring abscesses may lead to perivalvular leakage. The leakage may be appreciated as a new murmur and may be complicated by congestive heart failure and rhythm disturbances (particularly heart block). Late infections generally give rise to the usual signs and symptoms of subacute bacterial endocarditis.

Blood cultures are essential to the identification of the causative organism. Bacteremia is generally continuous; this feature may be demonstrated by obtaining at least three culture specimens drawn at half-hour intervals. However, culture results can be negative because of prior exposure to antimicrobial therapy or the presence of a fastidious organism. Echocardiography is another key component of the evaluation of patients with suspected endocarditis.191 In children, good-quality imaging can often be achieved by transthoracic echocardiography. However, the sensitivity of transesophageal echocardiography is greater than transthoracic studies.192 Finally, a negative echocardiograph does not exclude the possibility of endocarditis. Negative studies can occur if vegetations are small or have already embolized, or if prosthetic material obscures or distorts the images. Graft infection can manifest as local wound purulence, pseudoaneurysms at the graft–vessel anastomosis, hemorrhage, or thrombosis. Physical examination often identifies infection in superficially located grafts. Ultrasound examination is useful to detect pseudoaneurysms and thromboses. Other studies, such as angiography, computed tomography, and magnetic resonance imaging, are occasionally needed to determine the extent of infection and the presence of complications.186

Management and Outcome Antimicrobial therapy and surgery are the main therapeutic modalities for infections of prosthetic valves and patches. For both early and late infections, combination therapy using synergistic antimicrobials is usually employed.175 Because strains of CONS that often cause prosthetic valve endocarditis are almost always resistant to b-lactam agents,174 vancomycin and an aminoglycoside plus rifampin are used empirically until culture results are known. Surgical consultation is advisable, particularly for patients with early and acute infection. Indications for surgery in early infection include hemodynamic instability, repeated major embolizations, inability to control infection, and relapse of infection. Indications for surgery in late infection include new regurgitant murmur, moderate to severe heart failure, or a myocardial abscess unresponsive to medical therapy.174–176 When surgery has been undertaken during acute infection, studies have demonstrated that infection of valves implanted at this time is relatively uncommon.174–176,193 A retrospective analysis of S. aureus prosthetic valve endocarditis revealed significantly lower mortality when valves were replaced during antimicrobial therapy.194 Therapy for infected grafts often requires surgical resection or revision for cure.186 Antimicrobial agents have limited penetration in areas with poor blood flow (i.e., a thrombosed graft). The role of anticoagulation in infection of prosthetic valves or grafts is still debated. Most patients with prosthetic valves are treated with anticoagulants; discontinuation of this therapy is associated with a higher incidence of thrombosis and embolic phenomena.175,176 However, the risk of death from neurologic events is higher if anticoagulation is continued.195

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Clinical Syndromes of Device-Associated Infections

Complications of intravascular infections include thrombosis, hemorrhage, embolic infarction and infection, graft dysfunction, heart failure, overwhelming infection, and death. Prognosis of prosthetic valve endocarditis depends on the organism, the condition of the patient at the time of diagnosis, the rapidity and appropriateness of antimicrobial therapy, and the timely use of surgical intervention if appropriate.185,193,194

Prevention Proper timing of surgical procedures (e.g., avoiding elective surgery in a patient with an intravascular device and an active distant infection), appropriate administration of perioperative prophylactic antibiotics, meticulous attention to surgical asepsis, and prompt removal of catheters and other potential infectious foci are important measures to prevent infection of prosthetic valves and grafts (Box 102-9). The common practice of continuing perioperative antimicrobial prophylaxis until catheters or drains are removed is inappropriate and may prevent the timely identiÀcation of pathogens in early prosthetic valve endocarditis. Patients who have prosthetic valves are considered to be at high risk for the development of endocarditis for the rest of their lives; perioperative prophylactic antimicrobial agents are recommended for those who are to undergo certain dental, respiratory, gastrointestinal, and genitourinary procedures, according to American Heart Association guidelines196 (see Chapter 8, Chemoprophylaxis). Prevention of graft infection starts with meticulous attention to sterile technique at the time of implantation. Hemodialysis grafts are at risk of contamination and subsequent infection with each dialysis procedure, so good hygiene and care of the graft site with each access are essential.185 Because S. aureus is by far the most common pathogen associated with dialysis grafts, attempts have been made to decrease patient colonization through the use of intranasal mupirocin or systemic rifampin and cloxacillin; these attempts have met with some success.197–200 If dialyzers are reused, strict adherence to sterilization and reuse protocols is crucial. The Association for the Advancement of Medical Instrumentation and the Centers for Disease Control and Prevention recommend that hemodialysis fluids be monitored at least monthly with quantitative culture and endotoxin assay. Water used to prepare dialysate must have ≤ 200 cfu/mL in culture, and water used to reprocess hemodialyzers should have ≤ 200 cfu/mL in culture and < 1 ng/mL of endotoxin on assay. Dialysate colony counts should be ≤ 2000 cfu/mL.201

INFECTIONS ASSOCIATED WITH PACEMAKERS AND LEFT VENTRICULAR ASSIST DEVICES Epidemiology and Pathogenesis Temporary pacemaker wires, which are generally inserted percutaneously into the bloodstream, are associated with infection risks similar to those for vascular catheters. In addition, pacemaker

CHAPTER

102

595

wires are typically embedded in the myocardium, posing risk of myocardial abscess. Permanent pacemakers consist of a battery, which is inserted into a soft-tissue pocket, and wires that pass across tissue planes to reach the myocardium. During initial placement either the battery or wires can become contaminated.202 Soft-tissue infection of the battery pocket and, less commonly, of the wire track occurs in 1% to 7% of patients.1,203,204 Left ventricular assist devices (LVAD) are an effective treatment option for patients with end-stage heart failure awaiting transplantation. However, LVAD-related infection is a common complication, occurring in 25% to 70% of patients.205,206 The LVAD consists of a pump with an inflow conduit from the left ventricular apex and an outflow conduit to the ascending aorta. Within the device are two porcine valves that maintain unidirectional flow. The pump is placed into a pocket formed in the lateral rectus sheath. A driveline, tunneled from the pump, connects to an external power source via an exit site on the lower abdominal wall. Two types of LVAD infection occur: local infections, which involve the interaction of von Willebrand factor with Staphylococcus aureus and bacteremia.207

Clinical Manifestations and Laboratory Diagnosis Signs of pacemaker infection include tenderness and erythema around the battery pocket, fever, and, occasionally, bacteremia.203,208 The most common organisms are S. aureus and CONS. Gram-negative enteric bacilli, Pseudomonas, and yeast are occasionally isolated. Signs of LVAD driveline infection are localized purulent drainage from the abdominal exit site. Device pocket infection is characterized by tenderness and erythema around the pump pocket and purulence within the subcutaneous space. Bacteremia is characterized by fever, chills, and other constitutional symptoms. The distribution of pathogens identiÀed in LVAD infection is similar to those encountered in pacemaker infections.205,206 Diagnosis is made by culturing relevant purulent drainage and obtaining blood cultures.

Management and Outcome Treatment of pacemaker infections depends on the extent of the infection.202 Localized cellulitis can often be managed by systemic antimicrobial therapy without removal of the pacemaker battery. However, wire track infection necessitates removal and replacement of the wire. Endocarditis with associated myocardial abscess requires prolonged therapy and usually removal and replacement of the wire.208 The outcome depends on the site of infection, the pathogen, and the condition of the patient. Local infections usually resolve without sequelae; deeper infections are associated with greater morbidity and mortality. Treatment for LVAD-associated infection consists of antimicrobial therapy. Although the optimal duration of therapy for LVAD-associated infection is unknown, continuous antimicrobial therapy before, during, and after transplantation appears to be an effective strategy. The results of a large single-center study suggested that continuous antimicrobial therapy is associated with fewer relapses than a limited course of therapy.206 However, the potential beneÀts of prolonged antimicrobial therapy must be balanced against the risk of developing antimicrobial resistance.

BOX 102-9. Prevention of Prosthetic Valve Infection • Use prophylactic antimicrobial agents at time of placement and all subsequent dental/surgical procedures • Minimize intraoperative contamination • Remove catheters as soon as possible • Use meticulous wound care: handwashing, removal of necrotic tissue, avoidance of contamination • Minimize antimicrobial agents to maintain normal flora • Avoid antacids to diminish colonization of stomach and subsequent spread to lungs

Prevention Evidence-based data on the efÀcacy of measures to prevent infections in patients with pacemakers and LVADs are lacking. Antibiotic prophylaxis recommendations for pacemakers and LVADs are modeled after those used to prevent surgical site infection. For LVAD drivelines, exit site care may consist of daily sterile cleansing with povidone-iodine and isopropyl alcohol, dilute hydrogen peroxide, or chlorhexidine and placement of an occlusive dressing.206

596

SECTION

Q Infections Associated with Hospitalization and Medical Devices

INFECTIONS OF CENTRAL NERVOUS SYSTEM DEVICES

peritoneal shunt in whom the diagnosis of meningitis is suspected if a sample of ventricular shunt fluid sample is normal.

Epidemiology and Pathogenesis Central nervous system devices include cerebrospinal fluid (CSF) shunts, subarachnoid screws and bolts, and subdural catheters.209 Infections involving these devices are usually caused by skin flora introduced during surgical insertion. Ascending infection can occur in patients with ventriculoperitoneal shunts. These infections develop when the distal end of a ventriculoperitoneal shunt penetrates the bowel and are typically polymicrobial. Rarely, the ventriculoperitoneal catheter tip perforates the genitourinary system and becomes contaminated with local flora. Ventriculostomy catheters exit percutaneously to an external monitoring system and can become contaminated during placement or manipulation.210,211 Moreover, skin flora can ascend from the exit site along the catheter and into the ventricle. Subcutaneous reservoirs can be contaminated during placement or subsequent access. Infection rates for ventriculoperitoneal shunts range from 3% to 35%.1,212–217 In general, rates are higher for second surgical procedures (especially for replacement of an infected shunt, although the causative organisms may be dissimilar),209 and for shunts placed for hydrocephalus due to intraventricular hemorrhage fluid shunt infections in children.217 One study found that rates of infection were highest in July, correlating with the arrival of new neurosurgical fellows in hospital programs.214 Most infections occur within 2 months of shunt placement.218 As with other indwelling catheters, the rate of infection depends on the duration of catheterization.219 Infection rates of subcutaneous reservoirs have been reported to be approximately 20% to 25%.219 The rate of infection of percutaneous ventriculostomies has been estimated to be 10%.

Clinical Manifestations and Laboratory Diagnosis Catheter malfunction is the most common manifestation of infection. For shunts, malfunction produces signs of increased intracranial pressure (i.e., headache, vomiting, irritability, and mental status changes). Fever is not always present.209,210 Abdominal symptoms can predominate when the gastrointestinal tract is the source of shunt infection or if primary ventricular infection causes clinical peritonitis. Seizure is occasionally a presenting sign. Shunt malfunction due to occlusion or disconnection is the major differential diagnosis. A malfunctioning shunt should be considered infected until proven otherwise, even in the absence of systemic manifestations and CSF pleocytosis.210 Up to 30% of malfunctioning shunts are proved to be infected.214 Most infections are caused by CONS and other skin flora; gramnegative bacteria account for about 20%.210–214,220 Skin flora, including Corynebacterium spp., and Propionobacterium acnes, can cause particularly indolent infection. Fungal infections are most commonly caused by Candida spp. and are more common in prematurely born infants.221 The incidence of meningitis due to Haemophilus influenzae type b in patients with ventriculoperitoneal shunts has decreased since the advent of H. influenzae vaccine.222 Diagnosis requires examination and culture of ventricular fluid; examination of lumbar CSF is not sufficient.223 Pleocytosis is usual but can be absent.217,223 Ventricular fluid leukocyte counts can range from zero to thousands of cells per mm3,212 but are usually less than 150 cells/mm3. Elevated ventricular fluid protein is more common than abnormal glucose concentration and may be the only abnormality. Clinical and laboratory findings are often subtle in infections due to CONS. Blood culture results are expected to be positive with ventriculoatrial shunt infection; likewise, pleural fluid culture can be positive in ventriculopleural shunt infections, and peritoneal fluid in ventriculoperitoneal shunt infections. Abdominal complications of ventriculoperitoneal shunts, including formation of pseudocyst (sometimes called “CSF-oma”), are best diagnosed with ultrasonography or computed tomography.224,225 A lumbar puncture for CSF examination should be performed for a child with a ventriculo-

Management and Outcome Treatment includes antimicrobial agents and surgery.209,210,214 Cure with systemic antimicrobial therapy alone is unusual. A common approach to treatment involves: (1) removing the ventriculostomy or shunt or externalizing the peritoneal end of the shunt; (2) beginning systemic antimicrobial therapy; and (3) sampling ventricular fluid daily. Empiric therapy consists of an antistaphylococcal agent (nafcillin, oxacillin, or vancomycin); if abdominal signs are present or if a gram-negative infection is suspected, a broad-spectrum cephalosporin, such as cefotaxime, ceftriaxone, or ceftazidime, plus an aminoglycoside is added. Definitive therapy is based on susceptibility results of isolate(s) and ability of the antimicrobial agent to penetrate the CSF. When given intravenously, vancomycin penetrates into CSF poorly. Rifampin is useful in combination therapy for patients with infection due to CONS. If the infection does not resolve promptly, antimicrobial agents can also be administered intraventricularly via the catheter.226,227 Clearance of agents depends on the rate of fluid production, absorption, and drainage; therefore consideration should be given to direct measurement of drug concentrations to determine dosing. Intraventricular administration of antimicrobial agents can cause inflammation, which mimics CSF findings in infection. Cultures should be repeated frequently to ensure that the infection is controlled and to use time of sterilization as a basis for duration of therapy. The system is replaced after the infection is fully controlled. Antimicrobial therapy should not be discontinued for “proof of cure” before the shunt replacement. There are no data regarding duration of antimicrobial therapy (5 to 10 days after documentation of sterilization of CSF is usual) or optimal timing for shunt replacement.210,211,228 Timing of shunt replacement must balance the risks of superinfection associated with prolonged externalization against the risk of premature internalization with contamination of a new system. The major complication of ventriculitis is brain injury. Studies of children with myelomeningocele reveal that the number of shunt infections is a major determinant of low intelligence.229 The prognosis for ventriculitis depends on the virulence of the pathogen and the underlying host. It has been suggested that infections caused by slimeproducing strains of CONS are more difficult to cure.230 Infection in the immunocompromised host is more likely to result in morbidity or mortality. Complications of peritonitis include adhesions and bowel obstruction. Respiratory distress due to pleural effusion or infection can complicate infections of ventriculopleural shunts (Box 102-10). Ventriculoatrial infection can lead to endocarditis and immune complex-related glomerulonephritis.

Prevention Prevention of infection depends on the appropriateness of antimicrobial prophylaxis as well as meticulous sterile technique during

BOX 102-10. Complications of Central Nervous System Devices ALL SHUNTS Ventriculitis: loss of neurons VENTRICULOPERITONEAL SHUNTS Peritonitis: bowel obstruction Abdominal pseudocyst (“CSF-oma”) VENTRICULOPLEURAL SHUNTS Pleural effusion, empyema VENTRICULOATRIAL SHUNTS Bacteremia, endocarditis, nephritis CSF, cerebrospinal fluid.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Clinical Syndromes of Device-Associated Infections

device placement and with each sampling of CSF. Shaving the scalp over the shunt button before CSF collection damages the skin and raises the risk of infection at subsequent samplings. Although most individual studies of antimicrobial prophylaxis have failed to show a significant decrease in infection rate associated with its use, metaanalysis suggests that perioperative prophylaxis reduces the incidence of shunt infection. The optimal antimicrobial regimen remains unclear.213,231,232Administration of antimicrobial agents for the duration of continuous intracranial pressure monitoring is not appropriate.233 Prophylaxis for central nervous system devices should be limited to the perioperative period to avoid replacement of pathogens by resistant CONS. Catheters and shunts impregnated with antimicrobial agents, such as rifampin and clindamycin, have been shown in vitro and in animal models to inhibit bacterial colonization.234,235 Two small clinical trials in humans are promising, and this technique warrants further investigation.236,237

INFECTIONS ASSOCIATED WITH PERITONEAL CATHETERS Epidemiology and Pathogenesis Peritoneal catheters are most often inserted for peritoneal dialysis. Peritoneal dialysis removes complement and immunoglobulins and thus impairs host defenses.238,239 In addition, the low pH of dialysate inhibits neutrophil function.239 Peritoneal catheter-associated infections include catheter exit site and tunnel infections as well as peritonitis. The incidence of peritonitis is approximately 1 to 8 episodes per patient-year.1,10,240–242 Longer duration of dialysis therapy and decreased serum concentrations of immunoglobulin G are associated with a higher incidence of peritonitis.242 Organisms can enter the peritoneum during placement and subsequent manipulation of the catheter, with intrinsically or extrinsically contaminated peritoneal dialysis fluid, and after perforation of the bowel by the catheter tip.240 Hematogenous seeding, ascending infection via the fallopian tubes, and transmural migration of bacteria across the bowel wall are less common routes of infection.238 The most common organisms are those that colonize the skin around the catheter insertion site and that may enter the peritoneum via the catheter tract.238 Other pathogens are gram-negative enteric bacilli, such as Escherichia coli and Klebsiella spp., enterococci, Pseudomonas aeruginosa, yeast, nontuberculous mycobacteria, molds, and Nocardia.243,244 The most common fungal pathogen is Candida albicans.244

Clinical Manifestations and Laboratory Diagnosis Clinical findings in peritonitis are fever, abdominal tenderness, and cloudy dialysate. Presentations range from subtle pain to sepsis syndrome.240 The major differential diagnosis is chemical peritonitis or allergic reaction to some component of the dialysate. Examination of the dialysate fluid usually reveals pleocytosis. Cell counts vary, depending on fluid dwell time and the causal pathogen. However, leukocyte counts > 50 cells/mm3 suggest infection.238,245 There is no consensus as to the optimal way to culture fluid; methods include inoculation of a small sample or filtration of a large volume with inoculation of the filter disk on to solid media238,245,246 Ultrasonography of the catheter tunnel can be useful in diagnosing tunnel site infection (particularly when exit site infection is apparent or in the setting of peritonitis) and after response to therapy.247–249

Management and Outcome Exit site infections often respond to local therapy and oral antimicrobial therapy. Infections of the tunnel often require catheter removal.238 Peritonitis is usually managed with intraperitoneal antimicrobial therapy and continuous dialysis.240 Most patients with catheter-associated peritonitis can be treated in the ambulatory setting. If the patient is systemically ill or if attempts to eradicate the bacteria

CHAPTER

102

597

fail within a short time, hospital admission and systemic antimicrobial therapy are warranted, and catheter removal should be considered. Catheter removal is essential for cure of fungal infection244and is often necessary for enterococcal and Pseudomonas infections.243 Several reports indicate that simultaneous removal and replacement of the catheter are successful in selected instances,250–252 particularly when infection is caused by staphylococci, the tunnel is involved, or in cases of recurrent peritonitis that clears intermittently with treatment. This approach is less likely to be effective in infections due to gramnegative organisms or fungi, or if there is ongoing inflammation between episodes of recurrent peritonitis.252 There are no controlled studies on optimal duration of therapy; the usual course is 10 to 14 days.240 Complications include intra-abdominal abscess formation, adhesions with subsequent bowel obstruction, and impairment of the peritoneal surface available for dialysis.

Prevention Staphylococcal carriage predisposes to peritonitis. Eradication of nasal or skin carriage with antistaphylococcal agents decreases the incidence of peritonitis due to S. aureus.253–255 Systemic administration of rifampin (20 mg/kg once daily) combined with an antistaphylococcal agent for 7 days appears to be most effective.253 Topical application of mupirocin to the nose two to four times daily for 5 to 10 days often eliminates nasal carriage; elimination of nasal carriage was accompanied by elimination of hand carriage in one study.254 Aseptic technique during catheter insertion and manipulation is essential.

INFECTIONS OF ORTHOPEDIC DEVICES Epidemiology and Pathogenesis Prosthetic joints are used less commonly in children than in adults, but pins, medullary nails, and rods are commonly placed in children for stabilization of fractures and during corrective surgery. Infection is a major complication of orthopedic devices;256,257 the incidence varies with type of procedure and patient population.1 Prosthetic joint infections are the best studied; rates vary from 1% to 9%, depending on the joint involved.256–258 Risk of infection is higher: (1) in the presence of underlying malignancy; (2) in a surgical site infection not involving the prosthesis; (3) with an NNIS system surgical patient risk index of ≥1;259 and (4) after surgery related to a previously infected site.260,261 Pin site infections are common, because the pin track creates a direct pathway for skin flora and contaminating organisms to reach soft tissues and bone. Infectious risk depends on the mode and site of injury, amount of associated soft-tissue injury, and the time before pin removal. Solid medullary nails are associated with lower infection risk than hollow nails.262 Most infections are caused by CONS and S. aureus. Other grampositive cocci, such as group A streptococci, oral streptococci, and enterococci, as well as gram-negative enteric bacilli, are occasionally involved. Anaerobic bacteria account for < 10% of infections. Organisms gain access to the joint and bone at the time of implantation, and later via spread from contiguous infections, or (less commonly) by hematogenous seeding.256,257

Clinical Manifestations and Laboratory Diagnosis Clinical manifestations of implanted orthopedic devices can be subtle. Findings include pain, wound dehiscence, or loosening or malfunction of prosthetic joints. It is difficult to distinguish aseptic loosening from loosening due to infection. With pin-related infections, there is usually purulence at the skin exit site, but the depth of involvement is difficult to determine; infection can extend to the intramedullary cavity.263 Toxic shock has been reported as a complication of pin site infection,264 as has brain abscess secondary to a halo orthosis pin site infection.265

598

SECTION

Q Infections Associated with Hospitalization and Medical Devices

The diagnostic utility of most radiographic techniques in pin, nail, and joint infection is limited, because the appearance of infection mimics that of healing bone. For prosthetic joints, radiographs are sometimes helpful; periosteal reaction occurs with infection but can also occur with normal healing. An area of radiolucency surrounded by a sclerotic edge is suggestive of infection.266 Radionuclide scans (using technetium 99 compounds and indium 111-labeled leukocytes) are sensitive but not specific and do not reliably differentiate infection from fracture and healing bone.267,268 Computed tomography and magnetic resonance imaging (if not contraindicated) are rarely helpful because the device often causes artifacts, obscuring the area of interest.256 Culture of superficial drainage from pin sites or draining sinuses may yield colonizing or contaminating organisms rather than the true etiologic agent. Direct sampling of the bone or joint for histologic examination and culture is usually required for definitive diagnosis.257 Intraoperative specimens evaluated by frozen-section examination are useful in excluding active infection during surgery to revise loosened prostheses.269

hours afterward) to minimize excessive use of such agents. The clinical efficacy of dipping implants in antimicrobial solutions, antimicrobial coating of implants, or use of antimicrobial carriers (e.g., cements, beads) has not been established.278,279 Debate exists about antimicrobial prophylaxis in persons with prostheses who undergo dental or surgical procedures. Prophylaxis is generally not recommended unless the patient underwent joint replacement < 2 years earlier or also has: (1) an inflammatory arthropathy (rheumatoid arthritis or systemic lupus erythematosus); (2) drug-, disease-, or radiation-induced immunosuppression; (3) type 1 diabetes mellitus; (4) hemophilia; (5) malnutrition; or (6) a history of prior prosthetic joint infection.280 If the gums are infected or if the procedure is to be performed in an infected area, antimicrobial prophylaxis is appropriate to control the spread of the infection and to prevent hematogenous seeding.257

INFECTION OF URINARY CATHETERS Epidemiology and Pathogenesis

Management and Outcome Therapy for superficial pin track infections involves pin removal and antimicrobial therapy. Although oral antistaphylococcal agents are often used, a longer course of intravenous therapy is needed for contiguous deep soft-tissue infection or osteomyelitis. For infected rods and joints, removing the device offers the best chance of cure but is complicated by prolonged disability and deformity. Often, a two-step approach is used, in which the infection is first brought under control with parenteral antimicrobial therapy and the device and all devitalized tissue and cement are removed. At a later date, the device is replaced.270,271 There is no consensus regarding the ideal interval before replacement. If removal of the device is not feasible, a combination of antimicrobial therapy and debridement can be attempted. This approach has a failure rate of 70% to 80%, however, and is best attempted only when the duration of symptoms before treatment is brief.272,273 One study of adults with infected hip and knee prostheses demonstrated successful treatment with a combination of ciprofloxacin and rifampin for 3 to 6 months and retention of the prosthesis after initial debridement.274 However, the patients in this study had a short duration of symptoms. Antimicrobial therapy for infected orthopedic devices is typically administered for prolonged periods. Ideally, agents documented to be bactericidal against the causative organism are used. For infections due to staphylococci, the combination of a b-lactam or quinolone plus rifampin should be considered if the organism is susceptible and the device cannot be removed.275 In most cases in which the device is left in place, suppressive therapy may have to be continued indefinitely. If there are large tissue defects over the involved area, skin and muscle flaps may help by providing a vascular source to enhance host defenses and improve antimicrobial delivery.276

Infection is the most common complication of urinary catheterization.281 Isolated cystitis occurs most commonly, but pyelonephritis and secondary BSI can also develop. Contamination may occur during catheter insertion, organisms can enter the bladder by ascending from the perineum on the outside of the catheter, or contamination can be introduced from the collection bag.282,283 Infection rates are approximately 1% for each catheterization episode.284 The rate of infection for an indwelling catheter is 5% to 10% a day if a closed system is used.1,283,284 There is rapid infection of the bladder when an open drainage system is used. Infection rates are higher with diarrhea (presumably due to contamination of the urethral meatal area),87,285 diminished urine flow, or urinary stasis (as occurs with ureteral reflux, bladder diverticula, and stenosis of the urethra). The most common organisms that cause catheter-related urinary tract infections are members of the perineal flora. Escherichia coli is by far the most common isolate; other enteric gram-negative bacilli, Enterococcus, and Candida spp. are also frequently isolated. Patients receiving antimicrobial agents are at high risk of infection due to resistant bacteria and fungi.283,286

Clinical Manifestations and Laboratory Diagnosis Catheter-associated urinary tract infections are often asymptomatic.287 Signs and symptoms of such infections include urgency, frequency, enuresis, dysuria, and cloudy or foul-smelling urine. Urinalysis and culture are useful to confirm infection. Urinalysis can reveal pyuria, but pyuria is often absent in bacteriuria associated with indwelling catheters. Diagnosis of catheter-associated urinary tract infection is established by isolation of organisms with colony counts of > 102 cfu/mL from catheter urine.

Prevention

Management and Outcome

Many methods designed to decrease contamination of the operative field during implantation of orthopedic devices have been studied, including increasing the number of operating room air exchanges as well as use of high-efficiency particulate air filters, laminar flow systems, and special barrier suits for surgeons.277 A large prospective study involving > 7000 patients showed no difference in rates of infection between the use of conventional operating suites and that of laminar flow rooms.257 Perioperative prophylaxis with intravenous antistaphylococcal agents is beneficial and considered standard practice. Administration of prophylactic agents should be confined to the immediate perioperative period (i.e., from just before the procedure to 24 to 48

Therapy for bacteriuria and infection should include removal of the catheter if possible. If the catheter remains in place, the infection often persists despite appropriate antimicrobial treatment or recurs immediately after cessation of therapy. There is debate regarding the need for antimicrobial therapy for bacteriuria or cystitis, because these conditions often resolve spontaneously with removal of the catheter.283 For complicating pyelonephritis and BSI, systemic therapy is necessary; the optimal duration is unclear. The major complications of urinary tract infection are pyelonephritis and renal damage.284 There is also significant risk of bacteremia or BSI in patients with urinary tract infections.284,288 Thus, it is important to exclude urinary tract infection before major surgical

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Evaluation of the Child with Suspected Immunodeficiency

interventions in catheterized patients. Prognosis is related to extent of the infection, the pathogen involved, and underlying host factors.

Prevention Important strategies to prevent urinary catheter infections are: (1) limiting catheterization to those for whom it is medically necessary; (2) attention to strict aseptic technique during catheter insertion; and (3) removal of the catheter as soon as possible. Catheters are commonly left in place for too long and “forgotten”; one study of adult medical inpatients revealed that clinicians were unaware that their patients had a catheter 28% of the time.289 Prophylactic antimicrobial therapy has been demonstrated to provide brief protection against bacteriuria.290 However, few guidelines recommend this strategy because the risk of progression from asymptomatic bacteriuria to clinically significant urinary tract infection remains low for most patients. In addition, prolonged anti-

CHAPTER

103

599

microbial therapy is not protective and is often harmful because of the increased risk of infection with resistant organisms.282 Catheter care includes attention to handwashing, maintaining a closed system, minimizing entry to the system, and positioning the collecting system below the level of the bladder to ensure complete drainage of the bladder and avoid urine stasis or reflux.20 In general, applying creams or attempting to wash the perineum raises the risk of urinary tract infection, probably because of movement of the catheter and an increase in introduction of perineal flora along the catheter and into the bladder.291 Clean intermittent catheterization is preferred to indwelling catheterization in individuals with neurogenic bladder. The abdominal Credé method does not empty the bladder and may increase risk of infection.17,282,292 Antiseptic-impregnated catheters have been shown to delay the onset of bacteriuria but few studies have examined the cost-efficacy and the associated risk of emerging resistant organisms associated with the use of these catheters.293

SECTION

R

Infections in Patients with Deficient Defenses CHAPTER

10 3

Evaluation of the Child with Suspected Immunodeficiency E. Stephen Buescher

Although primary immunodeficiency disorders are typically rare, referral for evaluation of these conditions is common in children with frequent infections. Epidemiologic studies show that normal children younger than 2 years have on average five to six acute respiratory tract illnesses per year (with a range up to 11 or 12 per year).1,2 Infections such as otitis media and gastroenteritis occur with similar frequencies in children younger than 2 years of age, with up to 14 episodes a year at the far end of the normal spectrum.3 Attendance at group childcare and exposure to secondhand smoke further increase frequency of these infections.4 A carefully obtained history, a thorough physical examination, and selected laboratory tests are required to differentiate the uncommon, immunologically abnormal child who requires more extensive evaluation from the common, “normal but unlucky” child.

mean number of infections is “normal but unlucky.”5 These “normal but unlucky” children can experience large numbers of infections, but this experience is not due to an immunodeficiency. Criteria other than the number of illnesses, as shown in Box 103-1, must be used to conclude that a child is “normal but unlucky.” Further investigation is limited to those who lack these characteristics, and attempts to categorize patients into groups of likely underlying conditions.

Anatomic and Physiologic Abnormalities A variety of anatomic abnormalities can alter natural host defenses and predispose a child to recurrent infections (Table 103-1). These infections often localize to or near the site of the abnormality; thus, infections recur at the same site. In instances of congenital malformation, infections usually begin during infancy. Compared with incidence of primary immunodeficiency disorders, congenital malformation as a cause of recurrent infections is common.

Underlying Conditions The presence of underlying conditions, either natural or iatrogenic, can alter host defenses and predispose to recurrent infections or may be associated with an immunodeficiency (Box 103-2). In addition

BOX 103-1. History of a “Normal but Unlucky” Child

IDENTIFICATION The concept of the “normal but unlucky” child presumes that, in a group of normal children, the half that experiences more than the

• • • •

Lack of documented, deep infections at multiple sites Normal growth and development Normal morphology and physiology between episodes of infection Lack of a family history of immunodeficiency

600

SECTION

R Infections in Patients with Deficient Defenses

TABLE 103-1. Anatomic and Physiologic Abnormalities that Predispose to Recurrent Infections Type of Infection

Predisposing Abnormality

Bloodstream infection

Asplenia Cardiac valve abnormality Intravascular cannula or thrombus Neutropenia

Bone infection

Foreign body Orthopedic device

Meningitis

Cochlear implant Dura mater (meningeal) defect Midline dermal sinus Mondini defect of inner ear Neurenteric fistula Occult skull fracture Ventricular cannula

Pneumonia

Abnormal cough reflex Atelectasis Bronchiectasis Endotracheal intubation Extrinsic airway compression Foreign body Gastroesophageal reflux Polyps Pulmonary cyst, fistula Pulmonary sequestration Tracheal web Tracheoesophageal fistula Tracheomalacia Tracheostomy Vascular ring

Soft-tissue infection

Diminished sensation Foreign body Lymphedema Thermal injury

Urinary tract infection

Genitourinary tract duplication, cyst, fistula, or obstruction Nephrostomy Urinary catheter Vesicostomy

to the 45 primary immunodeficiency syndromes, immunodeficiency occurs in 105 other syndromes. Of these syndromes, 45 are associated with growth deficiency, 39 with specific organ system dysfunctions, 17 with inborn errors of metabolism, 17 with miscellaneous anomalies, and 17 with chromosomal anomalies.6 Some conditions only become serious considerations when specific data are obtained during a detailed history. Other conditions are suspected when constellations of noninfectious signs and symptoms are revealed by history and physical examination. Recurrent infections associated with underlying conditions can be either localized or disseminated and may or may not respond to appropriate therapy in the expected manner. Compared with the incidence of primary immunodeficiency disorders, this category of causes of recurrent infections is also common.

Primary Immunodeficiency Disorders Recurrent infections due to primary immunodeficiency disease are rare in the general population (Table 103-2) and are relatively rare compared with other causes of recurrent infection. Common characteristics in children with primary immunodeficiency disorders are shown in Box 103-3. Relative frequency of primary immunodeficiency disorders is shown in Table 103-3.7,8 The component due to other innate immune system disorders, e.g., those affecting pattern recognition receptors and/or their associated signal transduction systems, is not known, but

BOX 103-2. Underlying Conditions that can Predispose to Recurrent Infections Asthma Collagen vascular diseases Cystic fibrosis Diabetes mellitus Genetic/metabolic conditionsa Growth deficiency/immunodeficiency syndrome Neurologic syndrome with immunodeficiency Dermatologic syndrome with immunodeficiency Gastrointestinal tract syndrome with immunodeficiency Down syndrome Ichthyosis Myotonic dystrophy Werdnig–Hoffmann disease Galactosemia a-Mannosidosis Mucolipidosis II Galactosemia Orotic aciduria Methylmalonic aciduria Propionic acidemia Isovaleric acidemia Glycogen storage disease type IB Hematopoietic/immunologic conditions Hemoglobinopathy Immunosuppression Corticosteroid therapy Chemotherapy Radiation therapy Bone marrow and solid-organ transplantation Lymphohematopoietic malignancy Drug-induced cytopenia Asplenia syndrome Myelokathexis (WHIM syndrome) Newborn state Congenital malformation Prematurity Nutritional conditions Malnutrition Protein-losing enteropathy Sarcoidosis Renal conditions Renal failure Nephrotic syndrome Tumor necrosis factor antagonist therapy WHIM, warts, hypogammaglobulinemia, infections, myelokathexis. a See reference 8.

is probably low.8 The list of specific immunodeficiency conditions is long (Tables 103-4 to 103-6), and their degrees of characterization vary. Approximately 100 genes are associated with these conditions.9 Some conditions are well characterized, their pathologic mechanisms are understood, and numerous affected patients have been described (chronic granulomatous disease,10 X-linked severe combined immunodeficiency syndrome,11 X-linked agammaglobulinemia,12 leukocyte adhesive deficiency type I,13 and adenosine deaminase deficiency14). Other conditions remain incompletely characterized, poorly understood, or have been observed in so few patients that they have not been extensively studied (lazy leukocyte syndrome,15 common variable immunodeficiency,16 reticular dysgenesis,17 Toll-like receptor dysfunction8).

EVALUATION OF A CHILD WITH RECURRENT INFECTIONS History A detailed history, the initial step in patient evaluation, should gather information from all available sources, including medical records,

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Evaluation of the Child with Suspected Immunodeficiency

CHAPTER

103

TABLE 103-2. Estimated Frequencies of Selected Chronic Underlying

TABLE 103-3. Relative Frequency of Primary Immunodeficiency

Illnesses, Primary Immunodeficiency Disorders, and HIV Infection

Disorders

a

Condition 26

Asthma IgA deficiency Diabetes mellitus26 HIV infection Sickle-cell disease26 Cystic fibrosis25 Acute lymphocytic leukemia26 Phenylketonuria26 Agammaglobulinemia26 Severe combined immunodeficiency Chronic granulomatous disease27

Frequency

Factors

Percentage

1 in 26 1 in 500 to 1 in 700 1 in 556 1 in 1000 1 in 2200 1 in 2500 1 in 9000 1 in 10,000 1 in 50,000 to 1 in 100,000 1 in 100,000 to 1 in 500,000 1 in 255,000

B lymphocytes

50–70

HIV, human immunodeficiency virus; IgA, immunoglobulin A. a Superscript numbers indicate references.

BOX 103-3. Characteristics of Children with Primary Immunodeficiency Disorders • Infectious symptoms often begin in first days to weeks of life • Therapeutic response is slow despite identification of a pathogen and administration of appropriate antimicrobial therapy • Infection is suppressed rather than eradicated by appropriate therapy • Common organisms cause severe manifestations or recurrent infection • Unusual (sometimes sentinel) or “nonpathogenic” organisms cause infections • Growth and development are commonly delayed • Multiple infections occur simultaneously • Infection with common organisms leads to unexpected complications

parent or patient interviews, radiographs, and laboratory test results. Information is assembled into a chronology of data and events to determine whether episodes have been well enough characterized to attest that the patient has truly had recurrent infections, whether treatment of infectious episodes was appropriate, whether response to treatment was as expected, and whether other explanations for “infections” exist (e.g., fever from inflammatory rather than infectious disease).

Medical History A description of each infectious episode is obtained, specifically: (1) date, duration, and site of infection; (2) how the diagnosis was established (i.e., specific cultures and diagnostic tests performed and their results); (3) severity of the episode (shock, tissue destruction, ventilator support, skin grafting); (4) what treatment was used and responses to specific treatments; (5) need for surgical intervention and quality of wound healing; and (6) temporal relationships to previous episodes. This level of detail often necessitates a prolonged interview, but is crucial for establishing whether an immunodeficiency is likely. Additional helpful aspects of the patient’s history are immunizations administered and any associated clinical events, pattern of growth and development, medications given (including long-term antibiotic use) and their effect on course of disease, compliance with treatments, presence of conditions that may predispose to or masquerade as infections, unusual blood phenotypes (e.g., McLeod phenotype in chronic granulomatous disease, Bombay phenotype in leukocyte adhesion deficiency type II), and healing responses after injuries or surgery. If infections have been localized around an anatomic site, details should be obtained to focus on foreign-body aspiration, injuries or surgery at the site, medical problems involving the site, or timing of onset of infections. Historical details that decrease concern for a primary immunodeficiency are rapid responses to appropriate therapy, rapid resolution (versus suppression) of infected foci, and inability to

T lymphocytes

20–30

T and B lymphocytes

10–15

Phagocytic cells

15–20

Complement Other innate immunity factors

601

2–5 1 IgG subclass low; IgA normal or low; total IgG and IgM normal

Normal

Unknown Cg

Most normal; recurrent infections; autoimmune disorders

Selective IgA deficiency

IgA low, IgG and IgM normal; some low IgG2

Normal

Unknown

Most normal; recurrent infections; autoimmune disorders

Specific antibody deficiency

IgG, IgM, IgA normal; IgG subclasses normal; impaired response to PS antigens

Normal

Unknown

Recurrent or invasive bacterial infection

Transient hypogammaglobulinemia of infancy

IgG decreased, IgA and IgM variable

Usually normal

Unknown

Recurrent infection; rarely, invasive bacterial infections

AID, activation-induced cytidine deaminase; BLNK, cytoplasmic adapter B-cell linker protein; Btk, Bruton agammaglobulinemic tyrosine kinase; CD, cluster designation (of antigens); CD179B, l5/14.1, surrogate light-chain complex; Cg, IgG heavy-chain constant region gene; Ck, k light-chain constant region gene; Cm, IgM heavy-chain constant region gene; CVID, common variable immunodeficiency; HIM, hyperimmunoglobulinemia M syndrome; ICOS, inducible costimulator; Ig, immunoglobulin; Iga, immunoglobulinassociated signal tranducing chain; IgCH, immunoglobulin heavy-chain constant region gene; LRRC8, leucine-rich repeat containing 8; LCK, lymphocyte-specific protein tyrosine kinase; PS, polysaccharide; SAP, SLAM-associated protein; TACI, transmembrane activator and calcium-modulator and cyclophilin ligand interactor; UNG, uracil-DNA glycosylase; XLA, X-linked agammaglobulinemia.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Infectious Complications of Antibody Deficiency M

AID/UNG Defn.

IgM

Ag

SAD IGGSD

XLA, Autosomal agammaglobulinemia

G1,2,3,4 M

μ

μ

M

D

IgG1 IgG2 IgG3 IgG4

μ

Pre-B cell

Immature B cell

CHAPTER

104

609

Figure 104-1. Normal antibody synthesis and primary deficiencies. m, heavy chain of immunoglobulin (Ig) M; M, IgM; G, IgG; A, IgA; D, IgD; Ag, antigen; AID, activation-induced cytidine deaminase deficiency; CVID, common variable immunodeficiency; IGGSD, IgG subclass deficiency; SAD, specific antibody deficiency; SIGAD, selective IgA deficiency; T, T cell; THI, transient hypogammaglobulinemia of infancy; UNG, activation-induced cytidine deaminase; XLA, X-linked agammaglobulinemia.

Mature B cell CVID

THI SIGAD

T

A1,2 IgA1 IgA2

Stimulated B cell

lymphocyte activation pathway necessary for response to polysaccharide antigens; and (4) isotype restriction of the response. Specific immune responses to polysaccharide antigens typically do not begin to develop until the second year of life and there are substantial age- and antigen-dependent variations in the magnitude of these responses throughout early and middle childhood. Fortunately, this delay can often be circumvented by the use of vaccines in which polysaccharide antigens are conjugated to protein carriers (see Chapter 123, Streptococcus pneumoniae; and Chapter 172, Haemophilus influenzae).

X-LINKED AGAMMAGLOBULINEMIA The prototype of antibody deficiency syndromes is X-linked agammaglobulinemia (XLA), which has an incidence of approximately 1 in 200 000 persons (Table 104-2).1–3 In the first few months of life, passively transferred maternal antibody protects patients with XLA from infection. Thereafter, serum IgG, IgA, and IgM levels are very low, and patients have very poor or no specific antibody responses following infection or immunization. Circulating CD19+ B lymphocytes are absent or present in very low numbers. This feature may distinguish patients with XLA from those with transient hypogammaglobulinemia. Plasma cells are absent from lymph nodes and bone marrow, and lymph nodes and tonsils are absent or small. T lymphocytes are normal in both number and function. Neutropenia may occur in association with acute infections.4 Children with XLA have recurrent bacterial infections in infancy or early childhood, including otitis media, sinusitis, and pneumonia. Other invasive bacterial infections, such as bacteremia, meningitis, and osteoarticular infections, are also common. Most patients become symptomatic in the first 18 months of life, but onset is later in 10% to 20%.2 Most infections are caused by encapsulated bacteria, particularly H. influenzae type b (Hib), S. pneumoniae, Staphylococcus aureus, and Pseudomonas spp. Up to 50% of adults and a smaller proportion of children with XLA also experience chronic diarrhea, steatorrhea, or malabsorption. In some cases, these symptoms have been associated with chronic rotavirus and Giardia

Plasma, memory cells

lamblia infections.5–7 Enteritis due to Salmonella spp and Campylobacter jejuni is more common in patients with XLA than in healthy persons.2 Persistent Mycoplasma and Ureaplasma infections of the respiratory tract, joints, and urogenital tract also occur. An interesting exception to the general rule that antibody deficiencies lead primarily to greater susceptibility to bacterial, but not viral, infection is the common occurrence of severe and chronic enteroviral infections in patients with XLA. Chronic enteroviral meningoencephalitis often has an insidious onset with ataxia, loss of cognitive skills, and paresthesias.8 Occasionally, children have a more acute meningoencephalitis, with fever, headache, and seizures. Cerebrospinal fluid (CSF) pleocytosis with a lymphocytic predominance, increased protein levels and, frequently, decreased glucose levels is typical, although some children have normal or mildly abnormal findings. CSF abnormalities tend to worsen with clinical exacerbations. Enteroviral polymerase chain reaction analysis of the CSF is more sensitive than culture for the diagnosis of encephalomyelitis.9 Some individuals with chronic enteroviral infections experience a dermatomyositis-like syndrome, characterized by muscle weakness, edema, and woody induration of the skin, and a violaceous rash over the extensor surfaces of the joints. The syndrome is often accompanied by hepatitis. An important clinical observation is that enteroviral meningoencephalitis can be the initial manifestation of XLA and can occur despite intravenous IgG (IVIG) therapy.8 Patients with XLA are at risk of vaccine-associated paralytic disease after receiving oral poliovirus vaccine. Disease has been reported both in individuals who received the vaccine directly and in contacts of recent recipients of the vaccine.10 XLA is caused by mutations in Bruton agammaglobulinemia tyrosine kinase (BTK) gene, Btk, which is required for the normal differentiation of pro-B cells to pre-B cells and mature B lymphocytes.11,12 Monocytes and platelets also fail to express BTK. A family history of male relatives with recurrent infections is common; however, about half of patients represent new mutations. Over 100 distinct mutations have been described, some of which are associated with relatively mild clinical manifestations.13 The severity of infectious complications can also differ among family members with the same mutation.

610

SECTION

R Infections in Patients with Deficient Defenses

TABLE 104-2. X-Linked Agammaglobulinemia

TABLE 104-3. Autosomally Inherited Agammaglobulinemia

Onset

Infancy, early childhood

Onset

Infancy, early childhood

Clinical presentation

Recurrent respiratory and invasive infections, especially Haemophilus influenzae type b, Streptococcus pneumoniae Chronic diarrhea, malabsorption, chronic Giardia and rotavirus Chronic enteroviral meningoencephalitis, poliomyelitis, dermatomyositis-like syndrome Neutropenia with infection Lymph nodes and tonsils absent or small

Clinical presentation

Same as X-linked agammaglobulinemia

Diagnosis

IgG usually < 2 g/L, IgM/IgA < 0.2 g/L Peripheral blood CD19+ B cells < 2% Poor specific antibody responses, isohemagglutinins Mutation in Cm, CD179B, Iga, BLNK, LRRC8, autosomal inheritance

Treatment

IGIV

Diagnosis

IgG usually < 2 g/L, IgM/IgA < 0.2 g/L Peripheral blood CD19+ B cells < 2% Poor specific antibody responses, isohemagglutinins Mutation in Btk, X-linked

Ig, immunoglobulin; IGIV, intravenous immunoglobulin.

Treatment

IGIV

Ig, immunoglobulin; IGIV, intravenous immunoglobulin.

Patients with XLA are treated with IGIV to maintain normal serum IgG concentrations. Chronic enteroviral meningoencephalitis is treated with higher doses of IGIV (to maintain trough levels of > 1000 mg/dL) or with plasma or immune serum known to have high antibody titers to the enterovirus causing the infection.8 Intrathecal IGIV has also been used. Aggressive and prolonged treatment of bacterial infections is critical. The prognosis of XLA has greatly improved with IGIV therapy, particularly when it is initiated at an early age. The most common causes of death remain chronic enteroviral and pulmonary infections. Prenatal diagnosis and carrier detection are possible if the mutation is known.

AUTOSOMALLY INHERITED AGAMMAGLOBULINEMIA Approximately 10% to 15% of patients presenting with a clinical history and laboratory findings consistent with XLA are females or males who do not have mutations in Btk.1,13 As in XLA, these patients have low or absent serum Ig, poor specific antibody responses, and low or absent numbers of peripheral blood CD19+ B lymphocytes (Table 104-3). Mutations in the IgM heavy-chain constant region gene, Cm, the surrogate light-chain complex, CD179B or l5/14.1, the Ig-associated signal-transducing chain Iga, CD79A, and the cytoplasmic adapter B-lymphocyte linker protein (BLNK) cause autosomal-recessive agammaglobulinemia.14–17 Autosomal-dominant agammaglobulinemia in a female patient who also had minor facial anomalies has been attributed to a mutation in the gene encoding leucine-rich repeat-containing 8 (LRRC8).18 Treatment of autosomally inherited agammaglobulinemia consists of Ig replacement as for XLA.

IMMUNOGLOBULIN HEAVY-CHAIN DELETION In rare persons, large deletions in the heavy-chain locus affect the expression of one or more heavy-chain constant regions.19 Expression of antibody of the classes encoded by genes proximal or distal to these deletions is not affected. Six deletion haplotypes have been described. Decreased serum concentrations of IgG1 or IgG2 and IgG4, with or without IgA1, IgA2 or IgE deficiency, are most common. Most affected persons do not have significant problems with recurrent or severe infections and have relatively normal immune responses to vaccine antigens.20 Some subjects have had elevated IgG1 or IgG3 levels, suggesting that this may compensate for the absence of other Ig isotypes.20

IMMUNOGLOBULIN KAPPA-CHAIN DELETIONS Rare persons with partial or complete deficiencies of serum antibody containing kappa light chain have been described.21,22 The clinical manifestations of the disorder are variable. In one case, light-chain deficiency appeared to result from a point mutation in the kappa lightchain constant region gene (Ck).22

ACTIVATION-INDUCED CYTIDINE DEAMINASE AND URACIL-DNA GLYCOSYLASE DEFICIENCY Children with the hyperimmunoglobulinemia M (HIM) syndrome have recurrent bacterial infections in the first and second years of life, normal or increased levels of IgM, and absent or low levels of IgG, IgA, and IgE (see Chapter 107, Infectious Complications of CellMediated Immunity – Primary Immunodeficiencies). Approximately 70% of cases are inherited in an X-linked recessive fashion, and are due to mutations in CD40L (HIM type 1). HIM type 3, due to mutations of CD40, is a rare autosomal-recessive disorder. In each of these conditions, B-lymphocyte signaling through CD40, which is required for isotype switching and the affinity maturation of antibody responses, is impaired. HIM type 2 is an autosomal-recessive disorder caused by mutations of the activation-induced cytidine deaminase gene, AID (Table 104-4). A similar phenotype is caused by mutations in activation-induced cytidine deaminase, UNG. The protein products encoded by these genes are required for isotype switching from IgM to IgG and IgA. These disorders, therefore, are the result of a primary B-lymphocyte abnormality, rather than the defective interaction between B and T lymphocytes that is characteristic of HIM type 1 and HIM type 3. Serum IgM levels of patients with AID and UNG deficiency are usually greatly increased and IgG and IgA levels are very low or undetectable at the time of diagnosis. Circulating T- and Blymphocyte numbers are within the normal range and neutropenia is rare. Most patients have normal isohemagglutinins, which are of the IgM isotype, but poor or absent IgG antibody responses to vaccine antigens. The median age of presentation of patients with AID deficiency is about 2 years, several years younger than that of patients with UNG deficiency.23,24 Most patients experience recurrent or severe infections, particularly affecting the respiratory and gastrointestinal tracts. In contrast to HIM type 1, opportunistic infections are rare in these patients. Lymphoid hyperplasia, especially of peripheral lymph nodes, is common in untreated patients. Autoimmune and inflammatory complications, including diabetes, arthritis, autoimmune hepatitis, inflammatory bowel disease, uveitis, and Sjögren syndrome, have been reported in approximately 20% of patients, including young children. Unlike HIM type 1, there is no apparent increased risk of malignancy; however, only small numbers of patients with these disorders have been identified and followed for long periods of time.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Infectious Complications of Antibody Deficiency TABLE 104-4. Activation-Induced Cytidine Deaminase and Uracil-DNA Glycosylase Deficiency Onset

Infancy, early childhood

Clinical presentation

Recurrent and severe sinopulmonary infections Gastrointestinal infections, Giardia Lymphoid hyperplasia Autoimmune and inflammatory disorders (diabetes, arthritis, inflammatory bowel disease, uveitis)

Diagnosis

IgM usually elevated IgG usually < 2 g/L, IgA 40 kg

Comments81

Trimethoprimsulfamethoxazole

10 mg/kg

160 mg

Penetrates phagocytes

Dicloxacillin

12–25 mg/kg

500 mg

Remains extracellular

Clindamycin

10 mg/kg

300 mg

Penetrates phagocytes

a

Divided into two equal doses.

TABLE 106-6. Antibiotics and Dosages for Surgical Prophylaxis Figure 106-5. Abdominal computed tomography (CT) study of a 2-year-old male with chronic granulomatous disease and persistent vomiting. Contrast material in the stomach forms an “apple-core” shadow as it passes through the pylorus (arrow) which is thickened by granulomatous inflammation. The posterior antral wall is also severely thickened (lower arrow). Vomiting resolved within 24 hours of starting oral prednisone treatment.

Drug

Dose

Schedule

Alternative Druga

Oxacillin

Child: 200 mg/kg IV

20–30 min before and 8 hours and 16 hours after procedure

Vancomycin

Adult: 2 g IV

Child: 10–20 mg/kg IV over 1 hour Adult: 0.5–1.0 g IV over 1 hour

PROGNOSIS AND SEQUELAE The prognosis for patients with phagocyte functional disorders has steadily improved over time. With aggressive management of infectious episodes and careful follow-up, 10-year survival in conditions such as CGD has improved from 15% to 25%20 to 80% or greater.7,21 Survival data for other conditions are not readily available. Rates of infection in patients with CGD can be reduced by cotrimoxazole22,23 and itraconazole prophylaxis24,25 and by interferong prophylaxis.26 Patients with CGD and fungal pneumonia or disseminated fungal infection have mortality rates as high as 45%.8 Bone marrow and stem cell transplantation have been performed in a limited number of patients, including some with CGD27 and Chédiak–Higashi syndrome.28 Gene insertion therapies have been successful in animal models of CGD,29 but remain unavailable for humans.

RECENT ADVANCES Use of recombinant human hematopoietic growth factors has been examined in many phagocytic cell disorders and, in general, has not proven helpful, except in situations in which neutropenia accompanies abnormal function, as in glycogen storage disease Ib.30 Use of cytokines in nonmalignant syndromes of cytopenia (congenital neutropenia, myelokathexis) also shows some promise.31–33 Hematopoietic stem cell transplantation and gene therapy for CGD offer promise, but are still investigative.34

PREVENTION Prevention of infections in people with phagocyte function disorders is based on prophylactic antibiotic administration, prompt attention and care of minor injuries, and avoidance of potentially harmful environments. Chronic administration of oral antibiotics daily has not been examined definitively in any of these conditions. Daily cotrimoxazole (Table 106-5) appears to prolong infectionfree periods in CGD22,23 but has less clear effects in other conditions. Prophylactic antibiotics should be administered before elective surgical and dental

plus Gentamicin

Child: 1–2 mg/kg IM or IV

30–60 min before and 8 hours and 16 hours after procedure

Adult: 1.5 mg/kg IM or IV

Cefotaxime

Child: 50 mg/kg IV Adult: 1 g IV

IM, intramuscularly; IV, intravenously. a Given on same schedule as drug of choice.

procedures (Table 106-6). Continuous prophylactic use of itraconazole to prevent fungal infections in patients with CGD results in significantly fewer invasive fungal infections.25 Administration of ascorbic acid may be effective in reversing functional defects of phagocytes in Chédiak–Higashi syndrome28,35 and CGD,36 although beneficial effects have not been observed in all patients.37 In several conditions, administration of recombinant interferon-g appears to prolong infectionfree periods. In a large randomized and placebocontrolled study, administration of subcutaneous interferon-g three times a week (50 μg/m2) significantly prolonged periods between serious infections in patients with CGD26 and long-term treatment is safe.38 In hyperimmunoglobulinemia E syndrome, in vitro interferong exposure alters polymorphonuclear leukocyte locomotive responses, but clinical trials have not been performed.39 Minor wounds should be cleaned promptly and thoroughly and treated topically with an antiseptic or antibiotic agent (e.g., 1.5% to 3% solution of H2O2 or Neosporin ointment). To minimize gingivitis and periodontal disease, oral hygiene should include twice-daily brushing of teeth with 3% H2O2 and baking soda. Activities or environments that might predispose to infection should be avoided, including active and passive smoking; environments containing decaying plants, vegetables or wood (e.g., hay, sawdust, compost, garden mulch), where Burkholderia species (see Figure 106-2), spores of Aspergillus or other fungi might be aerosolized;40 and situations in which medical care is inaccessible.

624

SECTION

R Infections in Patients with Deficient Defenses

SPECIFIC CONDITIONS Chronic Granulomatous Disease CGD is characterized by a defective or absent phagocyte respiratory burst in both polymorphonuclear and mononuclear phagocytes.41 Absence of the respiratory burst results in failure to produce reactive oxygen metabolites and H2O2. Microbicidal activity against catalasepositive bacteria and fungi is defective; recurrent, life-threatening infections with these organisms result. CGD occurs in 1:250 000 births.7 For initial screening, the nitroblue tetrazolium test is rapid, relatively simple, and accurate. For diagnosis, a quantitative test of the respiratory burst (superoxide assay or H2O2 assay) is necessary.42 Both infectious and inflammatory syndromes are associated with CGD (see Table 106-2). Hepatic abscesses commonly cause fever, anorexia, and variable degrees of abdominal pain, often without abnormalities in serum hepatic enzyme levels. Abscesses can be multiple and variable in size (1 cm to 20 cm or more in diameter) (Figure 106-6). Surgical excision is the preferred management.43,44 Suppurative adenitis can involve a single node, multiple contiguous nodes, or multiple geographically separate nodes. Surgical excision is preferred, followed by prolonged antibiotic therapy. Pneumonia, caused by either bacterial or fungal agents, requires aggressive diagnostic efforts (e.g., bronchoalveolar lavage, open-lung biopsy) early in management. Fungal pneumonia may be asymptomatic and is difficult to eradicate; excisional biopsy should be seriously considered if pneumonia is localized discretely to one lung area. Osteomyelitis can cause fever with or without initial bone pain. Hematogenous spread or local extension from a nearby nonosseous infection can occur. Bone scan is usually diagnostic and bone biopsy for identification of organism should be performed. Vertebral osteomyelitis caused by Aspergillus species is difficult to treat and is associated with an ominous prognosis.45 Inflammatory syndromes include granulomatous changes in viscera that produce mass effects (see Figure 106-5), poorly controlled inflammatory responses (Figure 106-7), poor scar formation, and slow wound healing.

A

B

Myeloperoxidase Deficiency Myeloperoxidase (MPO) deficiency is the most common syndrome of phagocyte defect, occurring in approximately 1 in 2100 persons.46–50 MPO deficiency affects both polymorphonuclear and mononuclear phagocytes, resulting in deficient conversion of chloride and H2O2 to hypochlorous acid and its microbicidal byproducts. The diagnosis of MPO deficiency relies on biochemical or histochemical determination of absence of peroxidase activity in blood polymorphonuclear leukocytes. In vitro, MPO-deficient cells have only minor (if any) microbicidal abnormalities against bacterial pathogens. However, in vitro killing of Candida spp. by MPO-deficient granulocytes is dramatically deficient46,49 and may explain the severe candidal infections reported in some affected individuals.46,49–51 Appropriate surgical debridement with antifungal treatment appears to be effective for these infections. MPO deficiency should be considered in any patient with unexplained (not caused by an indwelling device or cytopenia) severe candidal infection.

Hyperimmunoglobulinemia E/Recurrent Infection Syndrome Originally termed Job syndrome,52,53 hyperimmunoglobulin E/recurrent infection syndrome is a multisystem autosomal-dominant disorder characterized by extreme elevations of serum immunoglobulin E (at least 10 times the upper limit of normal levels),54,55 recurrent skin and sinopulmonary infections, and eczematoid rashes. Clinically, patients often have coarse facial features, retention of primary teeth,53 and candidiasis of the mouth, nails, or vagina. Recurrent sinopulmonary and cutaneous infections often begin in infancy; pneumonia is frequently severe and often associated with

C Figure 106-6. Intraoperative photograph (A) and abdominal computed tomography (CT) section (B) through the liver of a 3-monthold patient with chronic granulomatous disease, illustrating multiple liver abscesses, the number and location of which preclude resection en bloc. Abscesses were successfully drained using intraoperative, ultrasound-guided aspirations, followed by prolonged antibiotic therapy. (C) Single complex liver abscess (white rectangle) in a 12-year-old patient with chronic granulomatous disease. Abscess is in the medial left lobe (box), in contact with the portal vein. Drainage was achieved by intraoperative ultrasound-guided needle aspiration, followed by prolonged antibiotic therapy.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Infectious Complications of Cell-Mediated Immunity Other than AIDS: Primary Immunodeficiencies

CHAPTER

107

625

pulmonary cyst or pneumatocele formation.56,57 Bronchitis and sinusitis are common. Cutaneous infections are typically manifested as either abscesses (often with little associated evidence of inflammation) or cellulitis, but other types of cutaneous infections can occur.58 When osteomyelitis occurs, it is often in proximity to a softtissue focus.55 Expected infectious agents include Staphylococcus aureus (predominant), Haemophilus influenzae, Candida albicans, Streptococcus pneumoniae, and S. pyogenes,54,55 but other organisms are also described.57,59 Noninfectious manifestations include arthritis, arthralgia, multiple bone fractures,53 hyperextensibility of joints, scoliosis, keratoconjunctivitis,55 pulmonary cysts or pneumatocele,56 and lymphoid cancer.55,60 No single laboratory study establishes the diagnosis. A number of laboratory abnormalities can be present that aid in diagnosis: 1. Variable chemotactic defect in granulocytes54,55 2. Excess production of a chemotaxis inhibitor61 3. Absent Staphylococcus aureus-specific immunoglobulin A (IgA)62 4. Elevated S. aureus-specific IgE62–64 5. Abnormal IgE production by B lymphocytes65 Although originally proposed as a syndrome of primary phagocyte defect, it is now clear that the hyperimmunoglobulinemia E/recurrent infection syndrome is a multisystem disease involving dentition, bones, connective tissues, and defense systems and that aberrant lymphocyte function may cause the associated abnormalities of phagocyte function (i.e., chemotaxis).66

nystagmus, and giant intracellular granules, is a rare, autosomalrecessive disorder of humans74 and several types of animals.75 Abnormal packaging of lysosomes is characteristic; all blood leukocytes show the diagnostic characteristic of abnormal, giant intracellular granules. In polymorphonuclear cells, these granules contain both primary and secondary granule markers76 and are deficient in elastase and cathepsin G.77 In vitro, microbial killing is delayed, degranulation is abnormal, and chemotactic responsiveness is impaired.34 Microtubular dysfunction may underlie the functional abnormalities in Chédiak–Higashi syndrome phagocytes.78 Recurrent infections involving the respiratory tract and skin are most common, with Staphylococcus aureus being the most common causative agent.79 Progressive neuropathy is common, and a lymphoma-like “accelerated phase,” accompanied by hepatosplenomegaly, lymphadenopathy, lymphohistiocytic proliferation or infiltration, anemia, neutropenia, and thrombopenia, develops in most patients during their first decade of life.79 Limited studies addressing the use of prophylaxis with antibiotic agents in these patients have shown little effect.34 In some patients, ascorbic acid treatment is reported to improve phagocytic cell function,28,35 but this effect does not occur in all patients.37 Ascorbic acid treatment does not alter progression to the accelerated phase, prompting investigation of bone marrow transplantation to correct this syndrome.28

Leukocyte Adhesion Deficiency Types I, II, and III

Secondary granule deficiency is a rare condition characterized by severe, recurrent infections and deficiency of the secondary granule marker lactoferrin.80,81 These patients’ phagocytes also lack all defensins, a class of phagocytic antimicrobial proteins.77 Phagocytes from these patients have multiple functional abnormalities.82,83 The diagnosis is made by demonstrating absence of antigenic lactoferrin in polymorphonuclear leukocytes. Clinically, patients manifest delayed localization of acute inflammatory cells at sites of injury, which leads to recurrent bacterial infections (both cutaneous and deep) and poor wound healing.82

To the present, three separate conditions involving abnormalities of leukocyte adhesion are identified as leukocyte adhesion deficiency type I, type II, or type III. The first descriptions of the condition now called leukocyte adhesion deficiency type I reported patients with recurrent infections, defective neutrophil mobility, and delayed separation of the umbilical cord.67,68 Subsequently, the additional characteristics of severe periodontal disease, poor wound healing, and multiple adhesion-related abnormalities of phagocyte function, including poor adherence, spreading, chemotaxis, phagocytosis, and antibody-dependent cellular cytotoxicity responses, were described.69 Partial (“moderate phenotype”) or complete (“severe phenotype”) absence of b2 integrins (“leukocyte integrins,” CD11/CD18) on mononuclear and polymorphonuclear phagocytic cells causes the condition.69,70 The result is slow mobilization and poor localization of phagocytic cells at inflammatory foci, resulting in severe, systemic, life-threatening infections, often resulting in death during infancy.69 Infections involving cutaneous ulceration, delayed wound healing, pneumonia, and peritonitis are common.69 Periodontal disease is common and can result in premature tooth loss. Leukocytosis without infection is also common, with marked leukocytosis often accompanying episodes of infection. In vitro natural killer cell responses are diminished,69 but whether this results in abnormal cellmediated immunity is controversial. Leukocyte adhesion deficiency type II is an unrelated disorder in which absence of fucosylated ligands (sialyl-Lewis X antigen), which are the ligands for endothelial cell selectins on the surface of phagocytes, results in ineffective phagocyte–endothelial cell interactions.71 Associated findings are severe mental retardation, short stature, and the Bombay erythrocyte phenotype. Dietary supplementation with oral fucose reverses the adhesive defect.72 Leukocyte adhesion deficiency type III is the most recently recognized of these conditions, involving defective activation of integrins with resultant defects in leukocyte adhesion and platelet aggregation. Although it is at present unclear whether the type III condition is one or several distinct entities, all cases recognized to date have the common clinical characteristics of recurrent infection and severe bleeding tendencies.73

Chédiak–Higashi Syndrome Chédiak–Higashi syndrome, consisting of recurrent pyogenic infections, partial oculocutaneous albinism with photophobia and

Secondary Granule Deficiency

CHAPTER

10 7

Infectious Complications of Cell-Mediated Immunity Other than AIDS: Primary Immunodeficiencies David B. Lewis

This chapter focuses on the infectious complications of primary immunodeficiencies in which lymphocyte-mediated immunity is compromised. Other portions of this textbook address the infectious complications of decreased cell-mediated immunity (CMI) that occur in the context of human immunodeficiency virus (HIV) infection (Chapter 112, Infectious Complications of HIV Infection), bone marrow and solid-organ transplantation (Chapter 97, Infections in Solid-Organ Transplant Recipients, and Chapter 98, Infections in Hematopoietic Stem Cell Transplant Recipients), cancer chemotherapy (Chapter 100, Infections in Children with Cancer), and during the neonatal period (Chapter 95, Viral Infections in the Fetus and Neonate). Genetic disorders of CMI that are mainly associated with hemophagocytic syndrome are discussed in Chapter 14 (Hemophagocytic Lymphohistiocytosis and Macrophage Activation Syndrome). Inherited immunodeficiencies that result in an autoimmune diathesis rather than

626

SECTION

R Infections in Patients with Deficient Defenses

compromised host defense, such as FoxP3 deficiency, are not included in this discussion.

OVERVIEW OF CELL-MEDIATED IMMUNITY AND ITS ROLE IN CONTROL OF INFECTIONS Disorders of CMI that predispose to infection can be due to quantitative or qualitative deficiencies of thymus-derived lymphocytes (T lymphocytes) or natural killer lymphocytes (NK cells), as well as abnormalities in antigen-processing or presentation to T lymphocytes by antigen-presenting cells (APCs).1 Limitations in cytokine production or cytokine responsiveness, e.g., deficiencies in cytokine receptors or downstream signaling molecules, are also important causes of genetic cell-mediated immunodeficiency.2 T lymphocytes play a critical role in the initiation and maintenance of antigen-specific immunity, and express heterodimeric surface receptors for antigen, known as T-lymphocyte receptors (TCRs). T cells express TCRs that consist of either a and b chains (a/b T lymphocytes) or g and d chains (g/d T lymphocytes). a/b T lymphocytes recognize antigens in the form of short peptides bound to histocompatibility leukocyte antigen (HLA) molecules on APCs. The CD4 subset of a/b T lymphocytes recognize peptide antigens presented by major histocompatibility complex (MHC) class II molecules, which consists of HLA-DR, -DP, and -DQ proteins in humans. CD4+ T lymphocytes regulate the adaptive immune response by producing soluble cytokines, such as interleukin-2 (IL-2), interferon-gamma (IFN-g), and tumor necrosis factor-alpha (TNF-a), and by expressing surface molecules, such as CD40-ligand (CD154), which interact with cognate ligands on other cells. This CD4+ T-lymphocyte regulation includes promoting B-lymphocyte responses to protein antigens, helping maintain CD8+ effector T lymphocytes that are involved in the control of persistent viral infections, and augmenting the microbicidal activity of mononuclear phagocytes. CD4+ T lymphocytes are particularly important for the control of pathogens that infect cells for at least part of their life cycle. These include fungi (e.g., Pneumocystis, Candida, and Aspergillus), protozoa (e.g., Toxoplasma and Leishmania), some bacteria (e.g., Listeria, Mycobacteria, and Salmonella), and viruses (e.g., herpesviruses). The human CD8 a/b T-lymphocyte subset recognizes peptide antigens presented by class I MHC molecules, which consists of HLA-A, -B, and -C molecules. CD8+ T lymphocytes are particularly important in killing infected host cells, such as those harboring viruses, by inducing them to undergo apoptosis (cell-mediated cytotoxicity). CD8+ T lymphocytes also produce cytokines, such as IFN-g and TNF-a, which contribute to antimicrobial resistance. APCs are not only involved in antigen presentation but also produce cytokines, such as IL-12 and IL-18, which play an important role in directing the adaptive immune response. Myeloid dendritic cells are particularly important in initiating the immune response of naive CD4+ and CD8+ T lymphocytes to pathogens. Plasmacytoid dendritic cells are an APC population that appears to be particularly important in the early response to viruses, and this cell type is capable of producing very high levels of type I interferon. Since the antigens that g/d T lymphocytes recognize and the role of these cells in host defense in humans are not well defined, they will not be discussed further. NK cells are distinct from T lymphocytes in that they lack TCRs, and have the innate ability to lyse host cells (natural cytotoxicity) that are infected with intracellular pathogens, particularly herpesviruses. NK cells also secrete cytokines, such as IFN-g, which may contribute to the early cell-mediated immune response. Most genetically defined primary lymphocyte-mediated immunodeficiency disorders are inherited as either autosomal-recessive (AR) or X-linked disorders, and compromise immunity from birth, e.g., severe combined immunodeficiency (SCID). Some inherited disorders require an infectious trigger for immunodeficiency to become manifest, e.g., Epstein–Barr virus (EBV) infection in the X-linked proliferative (XLP) syndrome. A few syndromes, such as selective

absence of NK cells or common variable immunodeficiency (CVID), can be classified as idiopathic, since their inheritance or genetic defect has not yet been documented in most cases. Although primary lymphocyte-mediated immunodeficiencies collectively are rarer than acquired immunodeficiency disorders, they often present as unusual, severe, or recurrent infections, and frequently require more aggressive approaches for specific microbial diagnosis and therapy. It is particularly important for the clinician to be aware of the frequent initial presentation of these immunodeficiencies so that a specific diagnosis can be made rapidly and potentially life-saving therapies, such as immune reconstitution, can be considered.

SEVERE COMBINED IMMUNODEFICIENCY SYNDROME Definition SCID is an inherited severe immunodeficiency of T- and B-lymphocyte function that occurs in about 1/50 000 live births. There is variable loss of NK-cell function, depending on the specific defect (Table 107-1).3 The “combined” term in SCID reflects the fact that severe T-lymphocyte deficiency invariably compromises B-lymphocyte function, even if B lymphocytes are present in normal numbers. This is because CD4+ T-lymphocyte help for B-lymphocyte antibody production in the form of surface CD40 ligand and secreted cytokines is required for most protein antigens4 and probably also for capsular polysaccharide antigens in the context of intact bacteria.5 In addition, in some forms of SCID the B lymphocytes that are present may have intrinsic functional defects.6

Specific Gene Defects and their Inheritance Pattern SCID can be due to single X-linked or autosomal gene defects3 or, much less commonly, to chromosomal abnormalities, e.g., complete DiGeorge syndrome (DGS) secondary to interstitial deletions involving chromosomes 22 or 10. Peripheral T-lymphocyte dysfunction in all forms of SCID is due to perturbed intrathymic maturation of ab-T lymphocytes. X-linked SCID comprises about 50% of cases in most series, and is due to deficiency of the common gamma chain (gc) gene, which is encoded in the Xq13.1 region. The gc protein (CD132) is a component of the receptor for the cytokine IL-7, and a lack of functional IL-7 receptors (IL-7R) on developing thymocytes results in their arrested development.3 Deficiency of the gc protein also results in the loss of functional surface receptors for IL-2, IL-4, IL-9, IL-15, and IL-21, and compromises the ability of gc-deficient B lymphocytes in this disorder to receive IL-21-mediated help for immunoglobulin production.7 Other components of the IL-7 receptor, such as IL-7R-a chain or the Janus kinase-3, which is associated with the gc, result in a form of SCID similar to gc deficiency except that these have an AR rather than an X-linked inheritance pattern.3 SCID with an AR inheritance pattern can also be due to defects in the ability of TCRs to transmit survival signals to the thymocyte (e.g., deficiencies of CD3-d, CD3-zeta-associated protein (ZAP-70) kinase, CD45, or Orai1, a molecule involved in calcium signaling), defective thymocyte rearrangement of TCR genes (deficiencies of recombination activating gene encoding RAG-1 and RAG-2 proteins), Artemis, DNA ligase IV, increased thymocyte death from toxic effects of accumulated purine metabolites (deficiency of adenosine deaminase (ADA) or purine nucleoside phosphorylase (PNP)), a failure of the development of bone marrow precursors that normally colonize the thymus and give rise to thymocytes (reticular dysgenesis), or deficiencies in the expression of class II MHC molecules, which are required for the intrathymic development of CD4+ T lymphocytes from CD4+CD8+ thymocytes as well as antigen presentation to mature peripheral CD4 T cells.3,8–11 DGS results in SCID in about 5% of cases, a form of the disorder that is referred to as complete DGS. Complete DGS may be more common, with interstitial deletion of the chromosome 10p14-13 region rather than the chromosome 22q11.2 region,12 which is the

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Infectious Complications of Cell-Mediated Immunity Other than AIDS: Primary Immunodeficiencies

CHAPTER

107

627

TABLE 107-1. Clinical Characteristics of Some Well-Defined Forms of Severe Combined Immunodeficiency (SCID) Syndrome/Gene Defect(s) and Mode of Inheritance

Gene Product Function

Effect on lymphocyte numbers Characteristic NoninfectionMechanism of Immunodeficiency and function Related Features

Common gamma chain (gc); Component of IL-2, IL-4, X-linked IL-7, IL-9, IL-15, and IL-21 receptors

Lack of IL-7 signaling leads to ŒŒ T and NK; nl B but ŒŒ thymocyte development; nonfunctional lack of IL-15 signaling arrests NK-cell development

None

JAK-3 kinase; AR

Signaling of cytokine receptors that use gc

Same as for gc chain deficiency

None

IL-7 receptor; AR

Specific component of IL-7 Same as for gc, except that intact ŒŒ T; nl NK and B receptor IL-15 function allows NK-cell development

RAG-1 or RAG-2; AR

Enzymes required for TCR and Ig gene rearrangement

ŒŒ T and B precursors

ŒŒ T and B, nl NK

None

Artemis, AR

Required for DNA repair process involved in TCR and Ig gene rearrangement

ŒŒ T and B precursors

ŒŒ T and B, nl NK

Radiation sensitivity

Class II MHC deficiency Transcription of class II (defects in CTIIA, RFX5, MHC (HLA-DR, -DP, RFXAP, or RFX-AP), AR and DQ) genes

Œ Intrathymic maturation of CD4+ T lymphocytes and ŒŒ antigen presentation to peripheral CD4+ T lymphocytes

Œ or ŒŒ CD4+ T, nl or Ø CD8+ T; nl NK and B

None

ZAP-70 kinase; AR

TCR signaling of thymocytes and T lymphocytes

Œ Intrathymic development of ŒŒ CD8+ T, nl or Ø CD4+ None CD8+ with sparing of CD4+ T lymphocytes, nl B and NK lineage; peripheral CD4+ T NK lymphocytes have ŒŒ function

CD45 tyrosine phosphatase; AR

Intracellular signaling of T and B cells

Œ Intrathymic development of ŒŒ CD4+ and Œ CD8+ T CD4+ and CD8+ T lymphocytes; lymphocytes; ØØ B B lymphocytes develop but lymphocytes but low have signaling abnormalities immunoglobulin levels

Cartilage hair hypoplasia (CHH) syndrome, AR

Gene for RMRP RNA

RMLP RNA/protein complex may be required for thymocyte growth

Œ to ŒŒ T lymphocytes – SCID only occurs with severe ŒŒ

Absence of scalp and eyebrow hair, short-limb dwarfism

Adenosine deaminase, AR

Enzyme in purine salvage pathway

Thymocytes and immature B lymphocytes die from toxic effects of accumulated purine metabolites

ŒŒ T and B in infantile-onset cases

Rachitic flaring of costochondral junctions (50% of infantile-onset cases); renal mesangial sclerosis

Purine nucleoside phosphorylase, AR

Enzyme in purine salvage pathway

Purine metabolites that accumulate are less toxic to B lymphocytes than T lymphocytes

ŒŒ T; variable Œ in B with poor function

Central nervous system disorders; autoimmune/ allergic disorders

Reticular dysgenesis, AR(?)

Unknown

Unknown, ? of a defect in stem ŒŒ T, B, and NK cells required for lymphocytes ŒŒ PMNs and granulocytes

Omenn syndrome – AR, partial RAG deficiency, some cases of deficiency of IL-7 receptor or CHH syndrome

See RAG-1, RAG-2, IL-7 receptor deficiencies, CHH syndrome

Œ T- and ŒŒ B-lymphocyte development, Ø peripheral T lymphocytes with ŒŒ TCR repertoire

Same as for gc chain deficiency

None

None

Bilateral sensorineural deafness

Nl or Ø T lymphocytes with Congenital/neonatal Th2 cytokine profile and ŒŒ erythroderma, B lymphocytes lymphadenopathy, hepatosplenomegaly, ØØ eosinophils and IgE

AR, autosomal recessive; HLA, human leukocyte antigen; Ig, immunoglobulin; IL, interleukin; MHC, major histocompatibility complex; NK, natural killer; nl, normal; PMN, polymorphonuclear leukocyte.

more common cause of the disorder. AR SCID can also occur in some cases of the cartilage hair hypoplasia syndrome,13 which is due to mutations in the RMRP RNA molecule.14 The RMRP RNA is a component of a ribonuclease complex, but the molecular basis for thymic hypoplasia in this disorder remains unclear. ADA deficiency and RAG protein deficiencies have been estimated to comprise about 35%, and 5% of cases of AR SCID.3,11

Characteristic Infections SCID disorders almost always become manifest in early infancy with infection.13,15–19 This is due to the marked susceptibility of the infant to opportunistic infections that are not observed in the immunocompetent host, and to severe and/or protracted courses of common pathogens, particularly viruses. The infectious complications of SCID

628

SECTION

R Infections in Patients with Deficient Defenses

are similar regardless of the particular genetic defect. Infections with adenovirus, enterovirus, parainfluenza and influenza viruses, respiratory syncytial virus (RSV), and herpesviruses, such as cytomegalovirus (CMV), herpes simplex virus (HSV), or varicella-zoster virus (VZV), can be severe, e.g., with marked pulmonary, liver or central nervous system involvement.13,15–19 This is due to a failure of adaptive immune mechanisms, such as the development of antigen-specific cytotoxic CD8+ T lymphocytes, to control and ultimately eliminate intracellular viruses in these and other tissues. Failure to thrive is a prominent manifestation in most patients, and is usually due to persistent gastroenteritis from common pathogens, such as rotavirus, adenovirus, or enteroviruses. SCID patients also have increased risk for complications from live vaccines, such as paralytic disease from oral poliovirus vaccine. Fungal infections, which are suggestive of severe CD4+ T-lymphocyte immune deficiency, include persistent and severe mucocutaneous candidiasis, and, especially, Pneumocystis carinii (P. jiroveci) pneumonia (PCP). Pneumocystis infection should always prompt a comprehensive search for SCID or another primary immunodeficiency involving CD4+ T lymphocytes in HIV-negative patients. Severe bacterial infections with pathogens that have a predominant intracellular presence, such as Legionella, Listeria, Mycobacterium (e.g., after vaccination with bacille Calmette-Guérin (BCG)), and Salmonella can occur and illustrate the requirement for CD4+ Tlymphocyte-mediated immunity in control of these pathogens. Other gram-negative bacteria, such as Pseudomonas, Serratia, Klebsiella, and Escherichia coli, have also been reported as causes of infection even prior to hematopoietic cell transplantation.15,16,19 Some cases of SCID can be complicated pretransplant by hemophagocytic syndrome,20 which can include neutropenia that increases the risk of invasive bacterial disease. As maternally derived immunoglobulin G (IgG) level decreases over the first several months of life, patients develop marked and persistent hypogammaglobulinemia, and have increased risk for recurrent sinopulmonary infections from encapsulated bacteria, such as Streptococcus pneumoniae and Hemophilus influenzae. In addition to infectious complications, some infants with SCID can develop a skin rash and other organ dysfunction, such as hepatitis and gastrointestinal inflammation, from graft-versus-host-disease (GvHD). GvHD is mediated by T lymphocytes acquired transplacentally from the mother or from unirradiated blood transfusions. Severe GvHD should also be distinguished from Omenn syndrome (OS), a form of SCID that classically is due to partial deficiency of one of the RAG proteins, but can also result from deficiency of Artemis, IL-7R alpha chain, or the cartilage hair hypoplasia syndrome.21,22 In OS, there is often extensive erythroderma and failure to thrive. OS is typically accompanied by a marked eosinophilia, highly elevated serum levels of IgE, increased levels of CD4+ T lymphocytes, and markedly reduced CD8+ T lymphocytes and B lymphocytes.

Diagnosis A history of other family members who experienced severe or recurrent infections in infancy or who died of unknown cause is suggestive of the diagnosis. However, a substantial number of cases of SCID result from new mutations, in which case the family history is not informative, even in cases of X-linked disease. Since SCID is usually due to severe thymic hypoplasia, there is typically a reduced or absent thymic shadow on imaging studies. In cases of early-onset ADA deficiency, lateral chest radiograph can also reveal the rachitic-like flaring of the costochondral junctions. On physical examination, peripheral lymph node tissue is typically reduced, although there are exceptions, such as in OS, in which lymphadenopathy and hepatosplenomegaly are common. In a minority of cases, findings point to a specific disorder, e.g., the findings of short-limbed dwarfism and the absence of eyebrows and hair in the cartilage hair hypoplasia syndrome, or the characteristic facies and congenital heart disease typical of complete DGS or velocardial facial syndrome (VCS: see Table 107-1).

In most cases of SCID, a complete blood count reveals a reduced absolute lymphocyte count (ALC) for age. This finding in early infancy should never be ignored, even in a well-appearing infant, as it can lead to an early diagnosis of immunodeficiency.18 A severely reduced ALC is typical of forms of SCID that result in an arrest of both T- and B-lymphocyte development, such as complete RAG deficiency or ADA deficiency. A less severely decreased ALC is characteristic of forms of SCID due to defects in IL-7R signaling, since this spares human B-lymphocyte and NK cell development (see Table 107-1). However, the ALC in some forms of SCID can be decreased to only moderately normal (e.g., some cases of X-linked SCID, class II MHC deficiency, and ZAP-70 deficiency) or even increased (e.g., some cases of ZAP-70 deficiency and most cases of OS). Normal or increased counts may reflect a compensatory increase in one lymphocyte population that masks the loss of another (e.g., the absolute increase in circulating numbers of CD4+ T lymphocytes masking the absence of CD8+ T lymphocytes in ZAP-70 deficiency) or a pathologic expansion (e.g., the marked expansion of a small population of activated CD4+ T lymphocytes with a limited TCR repertoire in OS). In some cases maternal engraftment of T lymphocytes can partially offset the lymphopenia. Cases are described of X-linked SCID in which the numbers of CD4+ and CD8+ T lymphocytes are normal.23 Therefore, regardless of the result of the ALC or T-lymphocyte subset analysis, if SCID is clinically suspected it is critical to enlist the help of a clinical immunologist so that appropriate tests evaluating lymphocyte subpopulations (e.g., flow cytometry), and T-lymphocyte function (e.g., mitogen-, alloantigen-, and antigen-induced proliferation, TCR repertoire, and gene sequencing) can be performed. Although delayed-type hypersensitivity skin test reactions to antigens, such as tetanus toxoid, can be useful as a screening test for intact CMI in older children, they are not reliably elicited during early infancy.24

Treatment Once SCID is diagnosed and appropriate blood samples for total and antigen-specific immunoglobulin levels have been obtained, intravenous immunoglobulin (IGIV) is usually administered pending a definitive treatment plan. All live vaccines, including oral poliovirus vaccine, rotavirus vaccine, measles, mumps and rubella, varicella vaccine, vaccinia, BCG, and attenuated influenza vaccine and oral Salmonella typhi vaccine, are contraindicated. IGIV therapy obviates the benefit of vaccination with “killed” vaccines (e.g., inactivated poliovirus vaccine, inactivated influenza vaccine (TIV)) and antigen vaccines by providing passive antibody. Therefore, these vaccines, although safe in SCID and other immunodeficiencies, are not typically administered once IGIV therapy is begun. For ADA deficiency, specific enzyme replacement with polyethylene glycol-associated ADA (Pegademase) is an option for immune reconstitution. For most SCID patients, the transplantation of hematopoietic stem cells contained in bone marrow or peripheral blood, ideally from an HLAmatched relative, is currently the standard treatment option for immune reconstitution. Significant T-lymphocyte and NK-cell immune reconstitution can also be achieved in X-linked SCID or ADA deficiency by reinfusion of patient-derived hematopoietic stem cells that are genetically engineered ex vivo by retroviral transduction to express intact gc or ADA proteins, respectively.3 Clinical trials of this form of gene therapy were halted after the development of Tlymphocyte leukemia in several treated X-linked SCID patients, but have recently been cautiously restarted.

Pneumonia in SCID Pneumonia is often a presenting feature of SCID, as well as other serious disorders of T-lymphocyte immunity. Because of myriad organisms that can cause pneumonia in these hosts, it is imperative to establish a specific etiologic diagnosis, particularly if the patient shows signs of severe disease or deterioration. Although some etiologies of pneumonia in SCID patients can be established by

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Infectious Complications of Cell-Mediated Immunity Other than AIDS: Primary Immunodeficiencies

bronchoalveolar lavage (BAL), lung biopsy should be strongly considered if there is not a prompt response to initial therapy.25 Results of BAL tests include a higher risk of both false-positive and falsenegative results compared with lung biopsy.26 Open-lung biopsy poses a risk of morbidity and mortality, especially in patients with serious respiratory compromise. Therefore, depending on the location of the pneumonia, consideration should be given to using less invasive biopsy methods, such as thoracoscopically guided sampling.26

SELECTED NON-SCID DISORDERS OF T LYMPHOCYTES AND NK CELLS Numerous non-SCID primary immunodeficiencies of T lymphocytes and/or NK cells can manifest with serious infections.1–3 The genetic basis, clinical features, including infection predilection, diagnosis, and treatment of some of the better characterized disorders are discussed below and summarized in Table 107-2.

ADA and PNP Deficiency (late-onset) ADA or PNP deficiency should be considered in HIV-negative patients of any age, including adults, in whom there is unexplained lymphopenia and recurrent infections. Notable infections include recurrent sinopulmonary disease, pneumonia, bacteremia, severe local papilloma infections, and recurrent herpes zoster.27 Late-onset infections associated with allergic or autoimmune hematologic disorders suggest partial enzyme defects.28

Ataxia-telangiectasia Ataxia-telangiectasia (AT) is an AR progressive disorder with cerebellar degeneration and a high risk of cancer. About two-thirds of patients also develop immunologic dysfunction, including reduced antibody responses to polysaccharide antigens, absent IgA in serum, and progressive loss of T lymphocytes. Patients frequently have severe and recurrent sinopulmonary infections,29 most likely a reflection of markedly decreased antibody-mediated immunity. Infections typical of compromised T-lymphocyte immunity are not common, although reduced T-lymphocyte immunosurveillance may contribute to the high risk of EBV-related lymphoma. The AT gene is a cytoplasmic molecule that helps arrest the progression of the cell cycle following radiation, allowing DNA repair to be completed prior to resumption of cycling. This accounts for the unusual sensitivity of AT patients to radiation-induced chromosomal breakage and the development of neoplasms. How AT gene deficiency also leads to neurologic and immunologic abnormalities remains unclear. The serum concentration of a-fetoprotein is usually elevated and is a useful screening test for this disorder. Treatment of these patients is problematic, as the immunodeficiency and neurologic disease are progressive and the risk of developing cancer is high.

Autoimmune Polyendocrinopathy–Candidiasis– Ectodermal Dysplasia Chronic mucocutaneous candidiasis is an invariable component of autoimmune polyendocrinopathy–candidiasis–ectodermal dysplasia (APECED), also known as autoimmune polyglandular syndrome type I, an AR disorder. APECED can include autoimmune-mediated hypoadrenalism, hypothyroidism, growth hormone abnormalities, hypogonadism, hepatitis, and vitiligo, as well as corneal and nail abnormalities. Chronic, intractable Candida infection of the oral mucosa, fingers, toes, and face is frequent,30 and occasional patients have disseminated fungal infection. Prolonged antifungal therapy and prophylaxis is the mainstay of treatment. Patients with chronic mucocutaneous candidiasis and APECED do not develop normal

CHAPTER

107

629

Candida-specific CD4+ T-lymphocyte proliferative responses or Candida delayed-type hypersensitivity skin test response despite chronic infection, suggesting a Candida-specific T-lymphocyte immunodeficiency. The rest of immunologic function is usually normal. APECED is due to a defect in the autoimmune regulator (AIRE) gene, and specific diagnosis is based on DNA sequence analysis. The AIRE gene encodes a putative transcriptional regulatory protein that is expressed at high levels in the thymic medulla.31 The autoimmunity in APECED might be due to a disturbance of negative selection, a process in which medullary thymocytes with autoreactivity are eliminated by apoptosis. This may account for the frequent detection of T-lymphocyte-dependent autoantibodies against endocrine organ targets,31 but it is unclear how this increases susceptibility to candidal infection. Some patients have chronic mucocutaneous candidiasis and decreased Candida-specific CD4+ T-lymphocyte responses but lack other manifestations of APECED. Whether these patients have a variant of APECED or some other disease is not clear.

DiGeorge Syndrome (Anomalad) and Velocardial Facial Syndrome Due to Interstitial Deletions of Chromosome 22 (22q11.2 Deletion Syndrome) Patients who are hemizygous for an interstitial deletion that includes the chromosomal 22q11.2 region may have features of either classic DGS or VCS. These include T-lymphocyte immunodeficiency due to thymic hypoplasia, transient neonatal hypocalcemia secondary to hypoparathyroidism, and truncoconal cardiovascular anomalies, especially interrupted aortic arch, truncus arteriosus, and tetralogy of Fallot, and characteristic facial anomalies.32 DGS and VCS are the result of a developmental failure of the third and fourth pharyngeal pouches. Haploinsufficiency of the TBX1 gene in the 22q11.2 region is likely responsible for the thymic hypoplasia and other characteristic features of DGS and VCS.33 It is important to note that the DGS can also result from other chromosomal abnormalities, such as deletion within the chromosome 10p14-13 region,12 and this should be sought if screening for a 22q11.2 deletion is negative. T lymphocytes that are produced by the thymus in 22q11.2 deletion syndrome appear to function normally in most respects. This probably accounts for the observation that patients with peripheral Tlymphocyte counts in the range of 500 to 1500/mm3 do not usually show signs of immunodeficiency.17 In most moderately lymphopenic infants with this syndrome, the peripheral CD4+ and CD8+ T-lymphocyte counts normalize during the first year of life.32,34 The assay of antigen-specific T-lymphocyte responses is particularly important, as peripheral T lymphocytes can undergo homeostatic expansion in a lymphopenic environment, giving the misleading impression that thymic function is relatively intact when it is actually substantially reduced.32 If such responses are normal, and the patient has a CD8+ count > 400/mm3, live vaccines appear to be well tolerated. Moderately lymphopenic patients may have risk for autoimmune hematologic disease in later childhood, suggesting that the thymic defect includes limitations in the induction of Tlymphocyte tolerance or its maintenance, e.g., reduced production regulatory T lymphocytes. In a small fraction of cases (< 5%), interstitial deletions of the chromosome 22q11.2 or, more commonly, chromosome 10p14-13, can result in severe thymic hypoplasia or aplasia, so that peripheral naive T-lymphocyte counts are persistently < 100/mm3. These patients typically have SCID-like manifestations with severe infections (see Table 107-1). In this case, immune reconstitution, such as via thymic transplant,12 may be effective.

Hyperimmunoglobulin M (Hyper-IgM) Syndrome The hyper-IgM syndrome is so named because of the frequent finding of an elevated serum level of IgM, and low levels of other immunoglobulin isotypes, particularly IgG.35 The classic X-linked

630

SECTION

R Infections in Patients with Deficient Defenses

TABLE 107-2. Genetic and Clinical Features of Selected T-Cell and Natural Killer (NK)-Cell Immunodeficiencies Syndrome/Gene Defect/ Mode of Inheritance

Gene Product Function and Mechanism Characteristic Noninfectious of Immunodeficiency and Immune Features

Characteristic Infections

Ataxia telangiectasia/AT gene/AR

Protects from radiation-induced chromosomal damage by cell cycle arrest, allowing DNA repair; mechanism of immunodeficiency?

Severe sinopulmonary infections and bronchiectasis, usually no opportunistic infections due to retention of T-lymphocyte immunity

Progressive cerebellar ataxia, bulbar and cutaneous telangiectasia, cellular hypersensitivity to radiation; ΠIgA,IgM

Autoimmune polyendocrinopathy– Elimination of autoreactive thymocytes Autoimmune endocrinopathies, candidiasis–ectodermal dysplasia in thymic medulla; relationship with hepatitis, vitiligo; keratopathy, Œ (APECED) syndrome/AIRE candidiasis and ŒCandida-specific Candida-specific T lymphocytes gene/AR T lymphocytes?

Chronic mucocutaneous candidiasis with decreased Candida-specific CD4+ T-lymphocyte proliferation and DTH skin tests

DiGeorge syndrome/unknown/ interstitial deletion of chromosome 22q11.2

TBX1 gene defect likely an important contributor to congenital thymic hypoplasia

Usually asymptomatic if circulating T lymphocytes > 500/mL; complete DiGeorge syndrome with severe T lymphopenia may present as SCID

Hyperimmunoglobulin M syndrome/CD40-ligand (CD154) gene/X-linked

Binds to CD40 and activates B Neutropenia and stomatitis common; lymphocytes and APCs; CD40 normal or elevated IgM in 50% ligand/CD40 interaction required for with ŒŒ serum IgG and IgA; poor generation of memory T and B specific antibody formation to lymphocytes, and isotype switching protein antigens

Sinopulmonary infections and chronic parvovirus (ΠB-lymphocyte immunity), Pneumocystis, Cryptococcus, Cryptosporidium, Toxoplasma, CMV, PML (ΠT-lymphocyte immunity)

Hyperimmunoglobulin M syndrome/CD40 gene/AR

Binds to CD154 and activates B lymphocytes and APCs; interaction with CD40 ligand required for memory T and B lymphocytes, isotype switching

Neutropenia, stomatitis common; elevated IgM; ŒŒ IgG and IgA; poor specific antibody formation to protein antigens

Sinopulmonary infections and T-lymphocyte immunodeficiency infections similar to X-linked hyperimmunoglobulin M syndrome.

Interferon-gamma (IFN-g) receptor deficiency/IFN-gR1 or IFN-gR2 genes/AR or, rarely, autosomal dominant

Specific cell surface receptor for IFN-g, a cytokine with pleiotropic effects on immunity

Poor granuloma formation in response Disseminated BCG and to mycobacterial infections; DTH nontuberculous Mycobacteria, response may be intact Salmonella, Listeria; recurrent oral and respiratory viral infections?

IL-12 and IL-12 receptor deficiency/IL-12 and IL-12Rb1 genes/AR

IL-12 induces NK-cell and Tlymphocyte IFN-g production and CD4 T-cell Th1 immunity

Granuloma formation in response to mycobacterial infection is normal

Truncoconal congenital heart disease, characteristic facies, hypocalcemia, ΠT lymphocytes (CD8 often Πmore than CD4)

Nontuberculous mycobacteria, Salmonella, Listeria

NEMO mutation with immunoImpaired activation of the NF-kB Hypohidrosis; conical/peg teeth, Disseminated nontuberculous deficiency/NEMO gene/X-linked transcription factor impairs innate oligodontia; delayed tooth eruption; Mycobacteria, gram-positive or (e.g., NK cell function) and adaptive Œ specific T-lymphocyte and ŒŒ gram-negative bacteremia/sepsis, immunity by T and B lymphocytes specific antibody responses and sinopulmonary infection, severe NK-cell-mediated cytotoxicity herpesviral infections, Pneumocystis NK-cell deficiency/unknown/ unknown

Unknown; lack of NK results in ? Myelodysplasia; ? malignancy; lack initially severe herpesvirus infections of NK cells by CD16 or CD56 until T-lymphocyte immunity can staining develop

Severe primary varicella, HSV, and CMV infection, but no Ø in recurrent infection with these viruses

TAP deficiency/TAP1 or TAP2 gene/AR

Transports peptides from cytoplasm to endoplasmic reticulum MHC class I binding

Skin ulcers (? noninfectious); decreased numbers of CD8+ T lymphocytes with ab-TCRs

Bronchiectasis from recurrent bacterial infections

Wiskott–Aldrich syndrome/WASP gene/X-linked

Regulates leukocyte cytoskeletal function; required for normal function of T lymphocytes, B lymphocytes, and APCs, including dendritic cells

Thrombocytopenia with decreased mean platelet volume, eczema, IgA-mediated autoimmune disease; ØIgA and IgE, ŒIgM and antigenspecific B cell responses

Recurrent sinopulmonary infections, herpesvirus infections, EBV lymphoproliferative disease, Pneumocystis, Aspergillus, mucocutaneous candidiasis

X-linked lymphoproliferative syndrome/SH2D1A gene/ X-linked

Lymphocyte signal transduction molecule involved in T-lymphocyte and NK-cell function

Lymphoid neoplasm can occur in the absence of EBV infection (rare)

Primary EBV infection with fulminant hepatitis, hemophagocytic syndrome, neoplasm, or later hypogammaglobulinemia

APC, antigen-presenting cell; BCG, bacille Calmette-Guérin; CMV, cytomegalovirus; DTH, delayed-type hypersensitivity skin tests; EBV, Epstein–Barr virus; HSV, herpes simplex virus; IFN, interferon; Ig, immunoglobulin; IL, interleukin; MHC, major histocompatibility complex; PML, polymorphonuclear leukocyte; TCR, T-lymphocyte receptor.

form of the disease is due to genetic deficiency in CD40 ligand (CD154), a surface protein that is expressed at high levels by activated CD4+ T lymphocytes. The interaction between CD40 ligand on the CD4+ T lymphocytes and the CD40 molecule on APCs, mononuclear phagocytes, and B lymphocytes, is essential for many adaptive immune responses, but is not required for normal T- and Blymphocyte development. These males have severe qualitative defects

in B-lymphocyte function and a partial but important qualitative defect in T-cell function. Markedly reduced B-lymphocyte function frequently results in recurrent sinopulmonary infections with encapsulated bacteria and, rarely, in chronic parvovirus-induced anemia or enteroviral meningoencephalitis. These patients are extremely susceptible during infancy to PCP, an indication of the importance of CD40 ligand (CD154) produced by CD4+ T

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Infectious Complications of Cell-Mediated Immunity Other than AIDS: Primary Immunodeficiencies

lymphocytes in the normal control of this fungus. A variety of other infections indicative of T-lymphocyte immunodeficiency have also been reported,36,37 including hepatobiliary infection with Cryptosporidium, cryptococcal meningitis, disseminated histoplasmosis, oral candidiasis, disseminated toxoplasmosis, severe CMV disease, and progressive multifocal leukoencephalopathy due to JC virus. Since the number of T and B lymphocytes and the ALC is normal, diagnosis requires assaying for functional CD40 ligand surface expression on activated CD4+ T lymphocytes using flow cytometry. Another frequent abnormality is neutropenia, the mechanism for which remains unclear. Oral ulcers often accompany neutropenia, but can occur independently. Patients should be treated with IGIV to prevent the complications of B-lymphocyte immunodeficiency, and receive PCP prophylaxis. Some cases of severe neutropenia may respond to treatment with granulocyte colony-stimulating factor. Prognosis is guarded even with these therapies; approximately 50% die by the fourth decade of life. Transplantation of hematopoietic stem cells from bone marrow, cord blood, or peripheral blood is curative of all features of the disease. The hyper-IgM syndrome can also affect males and females in an AR inheritance pattern.21 CD40 deficiency is a rare form of AR hyperIgM syndrome and, other than occurring in girls, is clinically indistinguishable from its X-linked counterpart.38 The identical immunologic phenotype and clinical consequences for CD40 ligand and CD40 deficiency is consistent with CD40 being the only known receptor for CD40 ligand. Other forms of AR hyper-IgM syndrome do not result in T-lymphocyte immunodeficiency but are associated with selective B-lymphocyte intrinsic defects in isotype switching and somatic hypermutation of immunoglobulin genes. These include defects in two enzymes, activation-induced cytidine deaminase (AID) and uracil N-glycolyase (UNG).35 Therefore, patients with AID or UNG deficiency are predisposed to recurrent sinopulmonary and gastrointestinal infections associated with severe antibody deficiency, but are not susceptible to severe infections with intracellular pathogens, such as PCP.

Interferon-g Receptor Deficiency The IFN-g receptor consists of two surface proteins, IFN-gR1, which bind IFN-g and IFN-gR2, resulting in intracellular signaling. Complete genetic deficiency of either of these receptor components results in an AR disorder characterized by marked susceptibility to intracellular bacterial infections, particularly with Mycobacterium, Salmonella, and occasionally, Listeria. Disseminated infection with BCG vaccine and environmental nontuberculous (“atypical”) Mycobacterium, including species that do not cause disease in immunocompetent humans, such as M. smegmatis, is characteristic.2 These mycobacterial infections typically occur before 3 years of age and have a high mortality rate, with survivors requiring continuous multiple-drug antimycobacterial therapy.2 Tissue that is infected with Mycobacterium characteristically lacks well-defined granulomas, demonstrating the importance of IFN-g in the induction of this form of tissue reaction. Despite poor granuloma formation, these patients may have delayed-type hypersensitivity responses to mycobacteria antigens administered intradermally.39 Severe oral or respiratory tract viral infections, caused by CMV, HSV, VZV, parainfluenza virus, and RSV, may occur in patients with prior mycobacterial infection.40 It remains unclear if IFN-g receptor deficiency directly predisposes to such viral infections or if these viral infections occur mainly as a secondary complication of severe lung damage. Partial IFN-gR1 deficiency can occur, due to homozygous or compound heterozygous mutations that are hypomorphic, in which there is recovery from childhood infection with disseminated BCG, M. tuberculosis, and Salmonella, without specific chronic therapy.2 Autosomal-dominant forms of IFN-gR1 deficiency have also been described. These forms tend to be less severe clinically than complete deficiency but more severe than partial deficiency. Diagnosis of IFN-g receptor deficiency requires specialized tests to evaluate IFN-g receptor expression and function, e.g., the production

CHAPTER

107

631

of TNF-a by monocytes after priming by exogenous IFN-g, and confirmation of mutations using DNA-based methods. Treatment is problematic since patients with complete deficiency do not benefit from recombinant IFN-g. Immune reconstitution using hematopoietic cell transplantation has been successful, but is a high-risk procedure in patients who are chronically infected with mycobacteria or who have sustained lung damage.

IL-12 and IL-12 Receptor Deficiency IL-12 is a heterodimeric cytokine, consisting of p35 and p40 chains, which stimulates IFN-g production by NK cells and T lymphocytes. IL-12 also promotes the differentiation of CD4+ T lymphocytes into a Th1-effector cell population with a high capacity for IFN-g production. Human genetic deficiencies of the IL-12 p40 chain and of one of the chains of the IL-12 receptor, the IL-12Rb1 chain, have recently been identified.2 Both result in a predisposition to infections that is very similar to that of IFN-g receptor deficiencies, i.e., disseminated infections with nontuberculous Mycobacterium,Salmonella, and, in one case, Listeria.41,42 However, granuloma formation in response to mycobacteria may occur normally, in contrast to patients with complete IFN-g receptor deficiencies, and extraintestinal infections with Salmonella are more prominent than in IFN-g receptor deficiency. Treatment with recombinant IL-12 or IFN-g, in the case of IL-12 deficiency, or with IFN-g, in the case of IL-12 receptor deficiency, bypasses the blockade of the Th1 response, and should be considered in cases unresponsive to conventional antimicrobial therapy.

Natural Killer Cell Deficiency A few patients of both sexes have been identified with an apparent selective and complete deficiency of NK lymphocytes. These patients have presented after infancy with initially severe primary herpesvirus infections, including VZV (D. Lewis, unpublished observations), HSV, and CMV, requiring intravenous antiviral drug therapy.43 After recovery from a particular herpesvirus infection, the patients do not experience an increased frequency of recurrent infection, indicating the development of durable viral-specific T-lymphocyte immunity. There is a complete absence of circulating NK cells, based on flow cytometry using the CD16+ and CD56+ markers, and absent in vitro NK cell activity in peripheral blood samples.43 T-lymphocyte numbers and function appear to be normal. It is not clearly established that this is an inherited disorder, since families with more than one affected individual have not been described. IGIV has been used for therapy, since this may theoretically increase virucidal function of cells other than NK cells, such as neutrophils and monocytes.

NEMO Mutation with Immunodeficiency NF-kB essential modulator (NEMO) protein is encoded on the X chromosome and plays a critical role in the activation of the NF-kB protein by facilitating its release from the inhibitor of NF-kB (IkB) protein complex.44,45 NEMO immunodeficiency occurs in boys with hypomorphic mutations of the NEMO gene; null mutations are lethal for the embryo. Ectodermal dysplasia, including hypohidrosis, sparse hair, conical or peg-shaped teeth, oligodontia, or delayed tooth eruption, occur in > 90% of cases, and reflect impaired function of the ectodysplasin A receptor, which utilizes NF-kB for signaling. NF-kB activation also plays a central role in both innate and adaptive immune responses. The immunologic phenotype and predilection to infectious disease complications are complex and also vary depending on the particular mutation. Decreased lipopolysaccharide (LPS)-induced production of TNF-a by NEMO-deficient myeloid cells is observed, consistent with a central role for NEMO in the activation of Toll-like receptor (TLR) NF-kB and NF-kB-dependent cytokine production

632

SECTION

R Infections in Patients with Deficient Defenses

(in this case, TLR-4).44 This block likely compromises the activation of mononuclear phagocytes and dendritic cells by pathogens and pathogen-derived products in vivo, which accounts for patients developing sepsis from a variety of bacteria, e.g., Listeria, Streptococcus pneumoniae, Klebsiella, Haemophilus influenzae, and Pseudomonas. Severe and disseminated infections with nontuberculous mycobacteria are a hallmark of the disorder, and may reflect both limitations in innate immunity as well as antigen-specific T-lymphocyte immune defects. PCP, CMV colitis, and severe herpesviral infections have also been observed, further suggesting impairment of T-lymphocyte immunity. NEMO-deficient NK cells have markedly impaired natural cytotoxic activity, which may contribute to the severity of herpesvirus infections,45 particularly at early stages. Finally, antibody response to both T-dependent and T-independent antigens is markedly decreased, and likely reflects the impact of impaired help to B cells from dendritic cells and CD4+ T lymphocytes, as well as intrinsic B-lymphocyte signaling defects, e.g., via CD40 engagement. Consistent with impaired CD40 signaling by B lymphocytes, some patients have a phenotype reminiscent of deficiency of CD40 ligand or CD40, i.e., increased levels of IgM and low levels of IgG and IgA.44 For unclear reasons, other patients may have markedly elevated levels of IgA.45 Poor specific antibody responses likely contribute to the risk of bacteremia/invasive infection from encapsulated organisms, such as Streptococcus pneumoniae44,45 and recurrent sinopulmonary infections. Patients can be screened for this disorder using assays for TLR function, such as LPS-induced production of TNF-a, and NK-cellmediated cytotoxicity, followed by confirmatory genetic sequencing. IGIV replacement therapy is indicated because of the severity of humoral immunodeficiency. Hematopoietic cell transplantation is curative for the immunodeficiency component of this disorder.

TAP Deficiency (Type I Bare Lymphocyte Syndrome) The transporter associated with antigen processing (TAP) is required for the proper loading and assembly of antigenic peptides on to the human class I MHC molecules, HLA-A, HLA-B, and HLA-C. TAP is a heterodimer consisting of the TAP-1 and TAP-2 proteins, and genetic deficiency of either protein results in TAP deficiency and in an antigen presentation defect characterized by reduced MHC class I surface expression. As MHC class I is also involved in selecting for CD8+ T lymphocytes that utilize ab-TCR from intrathymic precursors, some patients may have decreased numbers of these cells in the peripheral blood. Mice in which either of the TAP genes has been functionally inactivated by selective gene disruption are highly susceptible to severe and persistent viral infections, presumably because of a reduced ability of CD8+ T lymphocytes to recognize virally infected cell targets. An unexpected finding has been that deficiency of TAP-1 or TAP-2 in humans is strongly associated with chronic bacterial lung infections, especially bronchiectasis.46 Skin ulcers with chronic granulomatous inflammation are also frequent. This tendency for bronchiectasis does not appear to be sequelae of severe antecedent viral respiratory infections. TAP deficiency should be considered in cases of unexplained bronchiectasis, and can be screened for by assessing the expression of class I MHC molecules on peripheral blood leukocytes.

Wiskott–Aldrich Syndrome (WAS) Wiskott–Aldrich syndrome is an X-linked disorder characterized by thrombocytopenia with small platelets, eczema of variable severity, and recurrent pyogenic infections, especially otitis media. Immune abnormalities include severely depressed antibody response to unconjugated polysaccharide antigens, moderately depressed response to protein antigens, low serum IgM levels, and elevated serum IgA and IgE levels.47 IgE elevation is presumably associated with the

atopic diathesis of these patients. The defective gene encodes Wiskott–Aldrich syndrome protein (WASP), a cytoplasmic molecule that is widely expressed by hematopoietic cells and involved in the regulation of the cytoskeleton during cell signaling. In vitro studies indicate that WASP deficiency may compromise the intrinsic function of T and B lymphocytes, and antigen-processing cells, such as dendritic cells, but it remains unclear how these defects are precisely related to the abnormalities of immune function in vivo. In addition to pyogenic infections, Wiskott–Aldrich syndrome patients are also at increased risk for severe infection with certain viruses, especially VZV and HSV.48,49 Wiskott–Aldrich syndrome patients, particularly those with mutations that result in undetectable levels of WASP, may also develop infections suggestive of severely compromised T-lymphocyte-mediated immunity, such as Pneumocystis and Aspergillus infections, CMV encephalitis, and severe mucocutaneous candidiasis.50 A substantial fraction of Wiskott–Aldrich syndrome patients eventually develop autoimmune vasculitis associated with deposition of IgA-containing immune complexes, having many similarities to Henoch–Schönlein purpura. Clinical autoimmune manifestations can include intussusception, arthritis, and nephritis. In cases of Wiskott–Aldrich syndrome with profound thrombocytopenia, bleeding may be severe or life-threatening. Splenectomy may be helpful in raising platelet counts, but this also increases the risk of life-threatening bacteremia. Wiskott–Aldrich syndrome patients are at high risk for developing malignancy, especially lymphomas, starting after the second decade of life. The diagnosis is usually straightforward given the distinct features of the syndrome, and can be confirmed genetically by molecular analysis of the WASP gene. A determination of the precise gene defect in the WASP gene may not only allow screening of other potentially affected family members, but may also have prognostic value for the patient. Immune reconstitution by hematopoietic cell transplantation is curative of all disease manifestations, and appears to have the best outcome if performed early in life.

X-Linked Lymphoproliferative Disease XLP disease is an X-linked immunodeficiency of males that affects T and B lymphocytes and is due to a mutation in the SH2D1A (SH2 domain protein 1A) gene (also known as SAP and DSHP).51 SH2D1A is a cytoplasmic protein that is involved in T-lymphocyte and NK-cell signal transduction. The immunodeficiency is usually clinically silent until primary infection with EBV results in a fatal infectious mononucleosis syndrome in about 50% to 60% of cases, and in lymphoproliferative disorders, including malignant lymphoma in about 25 to 30% of cases. About 30% of patients manifest a persistent dysgammaglobulinemia characterized by decreased circulating levels of IgG and, in some cases, increased levels of IgM. Rarely, genetically affected individuals develop either lymphoma or dysgammaglobulinemia without prior EBV infection, indicating that EBV is not required for all manifestations of the immunodeficiency. The mean age of development of IM in this syndrome is about 5 years. Infectious mononucleosis is usually accompanied by severe hepatitis with periportal lymphocytic infiltration and bone marrow dysfunction with accompanying viral-associated hemophagocytic syndrome. Hemophagocytosis, in which there is the appearance of highly activated mononuclear phagocytes, many of which have internalized red blood cells, is also found in the bone marrow, lymph nodes, liver, and spleen (see Chapter 14, Hemophagocytic Lymphohistiocytosis and Macrophage Activation Syndrome). Meningoencephalitis is also common. Although there are no simple screening tests for diagnosis of this disorder, the diagnosis of XLP can be definitively established by analysis of SH2D1A protein expression or its gene sequence. Patients with post-EBV hypogammaglobulinemia are treated with IGIV. The prognosis without definitive treatment, such as hematopoietic cell transplantation, is poor, with most patients dying before 40 years of age.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Infectious Complications in Special Hosts

CHAPTER

10 8

Infectious Complications in Special Hosts Jane L. Burns, Janet Englund and Alice S. Prince

SICKLE HEMOGLOBINOPATHY The decrease in splenic function in young children with hemoglobin SS (SS) disease results in increased susceptibility to fulminant bacterial infection, especially in early childhood. After recognition of excessive rates of septicemia and meningitis due to Streptococcus pneumoniae in patients with sickle-cell disease in the early 1970s,1 mortality and morbidity due to pneumococcal disease have substantially decreased.1,2 Retrospective studies reported rates of invasive infection in children 0 to 10 years of age with hemoglobin SS disease prior to licensure of pneumococcal conjugate vaccine (PCV) to be 63.4 cases per 1000 person-years, a rate > 10 times that in the general population. Mortality rates due to pneumococcal infections in United States children with SS disease have been reported to be as high as 2.8 per 1000 person-years, a rate 100 times that in the general population.3 Patients with less severe hemoglobinopathies (sickle-cell hemoglobin C (SC) disease, sickle-cell thalassemia disease (S-thalassemia)) appear to have lower but increased risk for severe pneumococcal disease.4,5 Bone and joint infections are relatively more common in patients with hemoglobinopathies, and the Salmonella spp. are isolated with greater frequency.6,7

Etiologic Agents Encapsulated organisms, S. pneumoniae, Haemophilus influenzae type b, Neisseria meningitidis, and Salmonella spp. are the most common pathogens in patients with sickle-cell disease.8,9 In African countries, infection with Staphylococcus aureus, Escherichia coli, Salmonella, and Klebsiella predominates, with S. pneumoniae responsible for fewer cases of infection than previously reported.10 The incidence of H. influenzae type b infections has dramatically decreased with the use of conjugate vaccines in infancy. Other microbes with special significance for patients with SS disease are Edwardsiella tarda, Yersinia enterocolitica, Mycoplasma spp.,8 Chlamydophila spp.,11,12 and parvovirus B19.13,14

Epidemiology Several studies of the natural history of SS hemoglobinopathy in populations in the United States, Saudi Arabia, and Jamaica have demonstrated increased mortality in young children: rates of pneumococcal infection have been 20-fold to 100-fold higher than in normal children in the first 5 years of life.3,15 Persistence of fetal hemoglobin is associated with fewer episodes of infection. Nasopharyngeal carriage rates and serotypes of pneumococci infecting children with SS disease are the same as those infecting normal hosts.2 In a 1996 study, 27% of 226 patients with SS disease surveyed had at least one positive nasopharyngeal culture for S. pneumoniae.16 Because of the frequent use of antimicrobial agents, close contact with many children in group childcare, and penicillin prophylaxis, 50% to 60% of the pneumococci colonizing children with SS disease are not susceptible to penicillin.17,18 Although infection was the major cause of death in a cohort of patients between the ages of 1 and 3 years with SS disease who were

CHAPTER

108

633

monitored from 1979 to 1989, cerebrovascular accidents and trauma were more common causes in patients older than 10 years.19 Other life expectancy studies suggest that at least 50% of patients with SS disease currently followed survive beyond the fifth decade and that mortality is most often related to renal failure, chest syndrome, or stroke rather than infection.20,21No data are available for median survival of patients with SS disease in the African continent, but it is estimated to be less than 5 years of age.10 The incidence of bacterial infection in SC disease, although greater than that in healthy children, is less than that in SS disease. Functional asplenia has been documented in adults with SC disease, by radionuclide liver–spleen scans, or by quantification of pitted erythrocytes. In one series, 4 of 51 children with SC disease observed for 370 person-years were found to have 7 serious but nonfatal bacterial infections.4 A second report describes 7 fatal episodes of pneumococcal septicemia in patients with SC disease aged 1 to 15 years.5

Pathogenesis and Pathology The increased incidence and morbidity of infections due to encapsulated microorganisms in patients with sickle hemoglobinopathies are primarily attributable to splenic dysfunction. The spleen is important as a reticuloendothelial filter and is involved in processing bacterial antigens and subsequent helper T-lymphocyte and Blymphocyte responses. Encapsulated organisms cannot be phagocytosed efficiently without opsonization; thus, presence of type-specific antibody is critical to clearance of organisms. The spleen and the liver (to a lesser extent) are important in clearing pneumococci from the blood. Additionally, the activation of complement by the alternative pathway, which is critical for phagocytosis in the absence of specific antibody, appears to be deficient in patients with SS disease, although the specific defect has not been defined. Pneumococcus itself elicits a profound inflammatory response. Pneumococcal cell wall fragments trigger the expression of interleukin-1 and tumor necrosis factor, cytokines that in turn mediate systemic reactions associated with the clinical syndrome of septic shock. Thus, inability of the functionally asplenic, nonimmune child with SS disease to phagocytose and kill pneumococci efficiently results in bacteremia, multiple metastatic foci of infection, and unremitting upregulation of inflammatory mediators.

Clinical Manifestations The clinical manifestations of infection (i.e., fever, pain, erythema, swelling) are no different from those in normal hosts. However, differentiation of symptoms due to vaso-occlusive ischemia from those due to infection is problematic. Frequently processes occur together. Children with SS disease who appear “ill” should be treated aggressively for presumed septicemia while a specific diagnosis, including meningitis, is sought.

Pulmonary Symptoms Bacterial pneumonia caused by S. pneumoniae is suspected in a patient with fever, cough, chest pain, sputum production (older patients), and abnormal chest film (lobar infiltrate or other parenchymal or pleural abnormalities). Patients often appear toxic. Laboratory findings include leukocytosis, frequently with an increase in immature forms. Causative agents include encapsulated bacteria, Mycoplasma pneumoniae22 and Chlamydophila pneumoniae,11 all of which have been shown to cause excessive morbidity in patients with sickle hemoglobinopathy. Acute chest syndrome, described as fever and new pulmonary symptoms in a patient with SS or SC disease, is commonly precipitated by fat embolism and infection, usually community-acquired pneumonia.23 Specific causes may be difficult to differentiate.24,25

634

SECTION

R Infections in Patients with Deficient Defenses

Patients have chest pain, rales on physical examination, and infiltrates caused by focal vaso-occlusive necrosis or frank pulmonary infarction, although the latter may be less frequent in children. Radiographic abnormalities may not be detected for several days after the onset of symptoms. Precipitating infections are bacteria, Chlamydophila,11 viruses, and especially influenza. Conversely, primary vaso-occlusive crisis involving ribs can cause splinting, diminished pulmonary toilet, and secondary bacterial pneumonia. Laboratory data helpful in differentiating bacterial infection from vaso-occlusive processes are increased band count9 and either extremely high or low white blood cell count (> 30 000 or < 5000 cells/mm3). Blood gas determinations from patients with acute pneumonia show oxygen desaturation, but patients with acute chest syndrome due to vaso-occlusion can have dramatic abnormalities as a result of ventilation–perfusion mismatch. Patients with significant vaso-occlusive pulmonary disease require prompt treatment with transfusion or exchange transfusion.

symptomatology over the metaphysis and shaft of long bones suggest infarction. Laboratory findings may be helpful. Elevated erythrocyte sedimentation rate (which is usually depressed in SS disease) suggests bacterial infection. Bony destruction (“crumbling-bone disease”) and periosteal new bone formation on radiograph can occur with either condition. Imaging studies can be useful in establishing an etiology, but differentiation of infarction from infection is difficult (Figure 108-1). Radionuclide bone scanning is difficult to interpret. Magnetic resonance imaging may be useful for differentiating muscle and bone involvement. Bone aspiration or biopsy is frequently necessary to

Central Nervous System Symptoms In 1971, Barrett-Connor1 calculated that children with SS disease had a 300-fold higher risk of development of pneumococcal meningitis than normal children. Children with SS disease have a 6% to 8% chance of experiencing bacterial meningitis. Of all children with SS disease infected with pneumococcus, meningitis has been seen to develop in 20%, and 15% of these died.3 Factors associated with death in children with SS and SC disease include age > 4 years, serotype 19F S. pneumoniae, and not being followed by a hematologist.3 The rate of deafness after meningitis is increased in patients with SS disease, with rates as high as 40%. Several epidemiologic studies performed during the 1980s demonstrated a significant decline in the rate of pneumococcal meningitis in these patients, and this rate has been further reduced in recent years by the prophylactic use of antimicrobial agents in young children with SS disease, pneumococcal immunization, and prompt use of antimicrobial agents for febrile illnesses in these children.2,8,9,19 Lumbar puncture should be performed in all febrile children with SS disease who appear ill. This practice is particularly important in infants whose mental status may be difficult to evaluate or who have received oral antibiotics that can mask signs of infection. Central nervous system (CNS) vaso-occlusive disease (stroke) in young children with SS disease is relatively uncommon, and typically, neurologic abnormalities are found in the absence of signs of infection. The management of pneumococcal meningitis includes consideration of prompt use of corticosteroids to decrease the CNS inflammatory response; thus, it is important to establish this diagnosis immediately.

A

Bone and Joint Infections Bone and joint infection can manifest as dactylitis (in infants), pyogenic arthritis, or osteomyelitis with erythema, swelling, and pain.6,7 An infectious cause may be suggested by the clinical presentation (e.g., as a complication of septicemia due to S. pneumoniae as a subacute presentation with less systemic illness due to Salmonella as a cause). Limp or refusal to bear weight may be the only clinical complaint. Bone and joint infections are associated with S. aureus in patients with SS disease, as in normal children. Salmonella spp., particularly the serotypes associated with bone and joint infections (S. choleraesuis, S. heidelberg), are also important because of splenic dysfunction and can manifest either as systemic infection or with localized signs and symptoms.26 Specimens of blood and stool as well as aspirates from infected sites should be obtained for culture before empirical antibiotic therapy is started. It is important to determine specific etiology and antimicrobial susceptibility, because patients generally require prolonged therapy. The differentiation between osteomyelitis and vaso-occlusive ischemia in the bone is difficult. Pain and tenderness out of proportion to physical findings, bilateral, symmetric involvement, and diffuse

B Figure 108-1. (A) Radiograph of elbow of 30-month-old patient with hemoglobin S (SS) disease and osteomyelitis due to Salmonella spp. There is joint effusion, marked destruction of bone, involucrum, and periosteal new bone formation. (B) Magnetic resonance imaging of elbow showing fat saturation fast T2-weighted image in the axial plane. There is joint effusion, edema of the supinator muscle, and an active process within the radius more consistent with osteomyelitis than with infarction.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Infectious Complications in Special Hosts

establish a bacterial cause. Prompt surgical decompression of infected joints is also critical for good outcome and it is often diagnostic.

Management and Presumptive Therapy Meningitis, Septicemia, and Pneumonia Infants and young children with SS disease and fever should be evaluated and treated for presumed bacterial infection. Initial evaluation usually includes a complete blood cell count, blood culture, and chest radiograph, particularly in young children in whom the physical examination may not be optimal. Patients who are at high risk for infection (i.e., who look ill and who have a body temperature > 40°C or a white blood cell count > 30 000 or < 5000 cells/mm3) are likely to have pneumococcal infection and should be hospitalized for aggressive therapy.9 Children who have had previous episodes of pneumococcal sepsis are at increased risk for repeated infection and warrant special consideration.27 Patients who are febrile but do not appear ill have been managed successfully in a variety of ways, including: (1) observation in a shortstay area of the hospital for several hours after institution of antimicrobial therapy; or (2) parenteral administration of ceftriaxone and close observation on an outpatient basis.9 Enthusiasm for the latter approach is tempered by case reports of severe and rapidly fatal immune-mediated hemolysis due to ceftriaxone in patients with SS disease and other patients who have received multiple courses of the drug.28,29 All febrile children with SS disease should be given an antibiotic. In the case of disease highly likely to be a vaso-occlusive crisis, therapy could be given orally with close follow-up. S. pneumoniae that is not susceptible to penicillin or cephalosporins poses a particular challenge,18,30,31 especially for children younger than 3 years who are receiving penicillin prophylaxis. Ampicillinsulbactam plus gentamicin or a third-generation cephalosporin, plus vancomycin, are considered for presumptive therapy for septicemia. For meningitis, particularly if corticosteroids are used, rifampin (20 mg/kg per day) plus vancomycin (60 mg/kg per day) plus cefotaxime or ceftriaxone are considered for empiric use.

Bone and Joint Infections Empiric therapy for bone or joint infection includes agents effective against S. aureus (e.g., oxacillin, nafcillin, vancomycin, or clindamycin) and Salmonella (e.g., third-generation cephalosporin or quinolone). Extensive disease, particularly that due to Salmonella spp., usually requires debridement and prolonged antimicrobial therapy (e.g., 6 months or more) as for chronic osteomyelitis.

Prevention Substantial reduction in morbidity and mortality due to S. pneumoniae has been demonstrated in children younger than 3 years with SS disease through the use of prophylactic penicillin therapy.2 The major placebo-controlled study (performed prior to the development of PCV) demonstrated a decrease in the rate of pneumococcal infection from 9.8 to 1.5 in 100 patient-years.8 Prophylactic penicillin is begun at 3 months of age. There is no consensus regarding the age at which prophylaxis should be stopped; studies do not support continuation beyond 5 years of age.32 Adherence to an oral regimen may be erratic. Monthly injections of benzathine penicillin G are efficacious. A PCV series is recommended for all children, beginning at 2 months of age (see Chapter 123, Streptococcus pneumoniae), including those with SS disease; PCV given in infancy decreases nasal colonization with S. pneumoniae as well as invasive disease.33 For children with SS disease or other causes of functional asplenia, 23-valent pneumococcal polysaccharide vaccine is also recommended at 24 months of age (after at least two doses of conjugate vaccine). A

CHAPTER

108

635

second immunization with the polysaccharide vaccine is suggested at 3 to 5 years of age. Opsonophagocytic activity has been documented in patients with SS disease who have undergone immunization with the PCV followed by polysaccharide vaccine. Influenza vaccine should be given annually to children with SS disease because they have significant risk of complications due to influenza, including pneumococcal superinfection. This vaccine is recommended for children 6 months or older and is given annually in the autumn, before the influenza season.

Special Considerations Parvovirus B19 Infection Transient aplastic crisis in SS disease (worsening anemia and profound reticulocytopenia) has been associated with parvovirus B19 infection (see Chapter 214, Human Parvoviruses).13 The cellular receptor for this virus is the blood group P antigen, explaining the tropism of the virus for erythroid progenitor cells; viral infection appears to trigger apoptosis.14 Serologic studies of patients with SS disease who have transient aplastic crisis demonstrate acute parvovirus B19 infection in 70%.13 Patients have a high burden of virus, are highly contagious, and do not demonstrate the characteristic rash of erythema infectiosum. Neither chronic nor recurrent disease is reported in this patient group, and anti-B19 immunoglobulin (Ig) G antibodies remain detectable for several years following infection, suggesting protection.

ASPLENIA AND POLYSPLENIA Asplenic patients have increased risk for overwhelming, lifethreatening infections, most commonly due to S. pneumoniae.34–36 The level of risk appears to correlate inversely with the amount of time a patient has had a functioning spleen. Adults who undergo splenectomy after trauma have a lower risk of serious infection than infants with congenital asplenia syndrome and children who undergo splenectomy after trauma. Children who undergo splenectomy as part of treatment for malignancy are at great risk. Overall, approximately 5% of children whose spleens are removed before the age of 4 years have significant infections, with a mortality rate of 30% to 60%. Risk of infection is greatest in the year after splenectomy (regardless of age) and continues to be significant for the next 7 to 10 years, after which time risk is low but never as low as for people with normal splenic function. Children with congenital asplenia and polysplenia syndromes (e.g., Ivemark syndrome) are at increased risk of overwhelming septicemia.37 Splenic function should be evaluated in children with congenital abnormalities of embryonic lateralization, such as conotruncal abnormalities, transposition of the great vessels, anomalous pulmonary venous return, endocardial cushion defects, and situs inversus. Asplenia and polysplenia are often associated with other types of congenital abnormalities, including gastrointestinal malformations such as biliary atresia, neural tube anomalies, genitourinary defects, skeletal defects, and bronchopulmonary defects, including Kartagener syndrome, as well as bilobed and mirror image lungs. Because of the success of liver transplantation from living related donors in children with polysplenia syndrome, clinicians may be caring for more such patients who have significant risks for infection. The polysplenia syndrome has also been reported as an incidental finding on computed tomography in adults, suggesting that some individuals may have immunologic impairment that is not as significant as that of asplenic patients. Functional asplenia is confirmed by the presence of Howell–Jolly bodies in a peripheral blood smear (obtained after the first week of life), absence of splenic uptake on a technetium99 sulfur colloid scan, or increased percentage (more than 3%) of pitted or pocked erythrocytes in peripheral blood. The last test is a useful means of

636

SECTION

R Infections in Patients with Deficient Defenses

monitoring splenic function after trauma and partial splenectomy.38 Patients with severe or repeated pneumococcal infection should undergo evaluation of splenic function as well as tests for immunoglobulin deficiency and human immunodeficiency virus infection.

Etiologic Agents Encapsulated organisms represent the greatest risk for asplenic patients. S. pneumoniae has been associated with 50% to 90% of the overwhelming infections occurring after splenectomy. Other pathogens are H. influenzae b and N. meningitidis. Fatal cases are frequently associated with meningitis.35 It is not well documented that N. meningitidis infection is more fulminant in asplenic patients. Other streptococci, such as Streptococcus agalactiae (group B streptococcus) and Enterococcus spp., can cause fatal infection in asplenic hosts. Infections due to Salmonella spp. have been reported in asplenic patients, although the risk appears to be lower than in patients with sickle hemoglobinopathy. Fulminant septicemia due to Capnocytophaga canimorsus (DF-2 bacillus), part of the mouth flora of dogs, occurs in asplenic patients. The increased rates of gram-negative bacillary infections reported in asplenic hosts could be attributed to underlying malignancy, immunosuppression, and chemotherapy or to the loss of splenic function. Asplenic patients also have substantially higher risk for infection due to Babesia microti, an intraerythrocytic parasite that is endemic on islands off the coast of the eastern United States. Whether asplenic patients are at greater risk of severe infection from Plasmodium species is not well established.

Prevention Routinely recommended immunizations, especially PCV, H. influenzae type b vaccine, and varicella vaccine, are critically important for all asplenic children, with additional use of the polysaccharide pneumococcal vaccine after age 2 years. The response to this vaccine is variable, even in children older than 2 years.41 Meningococcal conjugate vaccine should be administered at the youngest age of recommendation; current licensure for adolescents may extend to younger children in the future. When splenectomy is elective, children should be immunized at least 2 weeks beforehand if at all possible.25,42 Children have a more rapid decline in antibody titers after pneumococcal vaccination and should be revaccinated at 3- to 5-year intervals with the 23-valent polysaccharide vaccine.43 Minimal side effects – chiefly induration and erythema at the injection site – are associated with revaccination: these side effects are thought to be due to an Arthus-type reaction. It is less clear whether children with asplenia benefit from prophylactic antibiotics to the same extent as children with SS disease. Most experts recommend penicillin prophylaxis for children who: (1) have congenital asplenia; (2) undergo splenectomy for hemolytic anemia, malignancy, or liver transplantation at any age; (3) undergo splenectomy before age 5 years; and (4) in the first years after splenectomy, for those who undergo the procedure at any age. There is no consensus as to when penicillin prophylaxis should be discontinued. It is important to recognize that patients can experience fulminant infection while receiving antimicrobial prophylaxis and, thus, require the same careful and urgent evaluation for febrile illness. To date, invasive disease due to penicillin-resistant S. pneumoniae has not occurred at high rates in asplenic patients receiving prophylaxis; it is, however, an important consideration for therapy.

Pathogenesis The central roles of the spleen include: (1) mechanical clearance of antigen and foreign material; (2) synthesis of factors such as tuftsin, that enhance phagocytosis; and (3) coordination of interactions of the T-lymphocyte and B-lymphocyte responses to organisms. The spleen is important in the initial response to a pathogen, perhaps as a result of its role in IgM production; the asplenic host is at a disadvantage when encountering a polysaccharide antigen for the first time. Asplenic patients are unduly affected by encapsulated organisms that require antibody and complement for opsonization and clearance. In the absence of preformed antibody, the spleen is critical in clearing the bloodstream. This may explain why adults who have developed a broad immunologic repertoire and older children who have been given pneumococcal vaccines have a lower risk than asplenic children for overwhelming infection after splenectomy. Infants with congenital asplenia have profoundly impaired reticuloendothelial clearance mechanisms. Additionally, they have diminished T-lymphocyte responsiveness to a variety of antigens compared with age-matched controls.39

Management Seemingly trivial febrile illnesses can herald life-threatening pneumococcal septicemia in the asplenic patient. In a retrospective study, clinical presentation of 26 episodes of bloodstream infection in children with asplenia included fever in 22 children, shock in 7 children, petechiae or purpura, respiratory distress in 5 children, and disseminated intravascular coagulopathy in 5 children.35 Febrile episodes require careful evaluation; empiric antibiotic therapy is begun urgently. Many physicians instruct patients to begin taking oral antibiotics at the earliest signs of infection, particularly if medical evaluation is not immediately available. For patients requiring parenteral antibiotics, empiric therapy is the same as for patients with SS disease. Vancomycin and third-generation cephalosporin are usually given.40

RENAL DISEASE Children with renal disease have increased risk for infection for a variety of reasons, including underlying disease and therapeutic modalities, such as immunosuppressants, corticosteroid therapy, and indwelling intravascular and peritoneal dialysis catheters. Children with nephrotic syndrome and accompanying hypogammaglobulinemia have increased risk for infection by encapsulated organisms, particularly S. pneumoniae, which require preformed IgG for efficient opsonization. A 1999 review of 452 admissions of 231 children with nephrotic syndrome described 10 episodes of septicemia (4 due to S. pneumoniae, 2 of which were fatal) and 8 episodes of peritonitis in 18 patients.44 Gram-negative organisms, including N. meningitidis and Salmonella spp. as well as gut flora (Escherichia coli, Klebsiella, and Enterobacter spp.), accounted for 50% of the infections. Such patients require careful evaluation and empiric therapy effective against bowel flora as well as the encapsulated pathogens associated with septicemia in patients with immunoglobulin deficiency. Patients with nephrotic syndrome also have an apparently higher frequency of urinary tract infections. Children with nephrotic syndrome as well as other patients with chronic renal disease should be immunized with the PCV followed by the polysaccharide vaccine.45 Antibody levels in these patients are lower than in healthy children, but geometric mean antibody titers are generally within the protective range following immunization.46 Duration of antibody protection may be reduced.47 The benefit of prophylactic penicillin in children with nephrotic syndrome is unclear because the risk of selection of multiple-drug-resistant organisms is of concern. Some clinicians use prophylactic penicillin in children with nephrotic syndrome younger than 2 or 3 years who have had an episode of S. pneumoniae septicemia, as for children with sickle-cell disease. Patients with end-stage renal disease who are undergoing ambulatory peritoneal dialysis or hemodialysis have additional risk of infection. Patients receiving chronic ambulatory peritoneal dialysis

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Infectious Complications in Special Hosts

commonly experience low-grade peritonitis, which is often due to catheter-related coagulase-negative staphylococci. Gram-negative organisms or fungi from the bowel can contaminate peritoneal fluid and cause frank peritonitis. The repeated use of vancomycin in these patients may predispose to multiple-drug-resistant pathogens, such as vancomycin-resistant enterococci (VRE). Linezolid is currently considered the drug of choice for VRE infections in adults and there is increasing use of this drug in children.48 The use of indwelling atriovenous shunts can also lead to infection, usually due to coagulasenegative staphylococci but also to Staphylococcus aureus or other nosocomial pathogens. Children with nephrotic syndrome who are being treated with corticosteroids (> 2 mg/kg or > 20 mg/day of prednisone) are at significant risk for severe varicella infection. A study of varicella vaccination in 20 children with corticosteroid-sensitive nephrotic syndrome found that antivaricella antibodies remained high 2 years postvaccination, with only 3 subjects becoming mildly infected.49 The use of varicella-zoster immune globulin (VZIG) after known exposures is recommended, and some clinicians also treat exposed patients with acyclovir, although the oral dosages used are unlikely to achieve sufficient levels to be effective.

IRON OVERLOAD STATES Patients with increased availability of free iron are at greater risk of serious infection due to Yersinia enterocolitica50 and, less commonly, Listeria monocytogenes, Vibrio vulnificus, and Klebsiella spp.51–54 Children at risk include: (1) those with b-thalassemia who require chronic transfusions and iron-aluminum chelation therapy with desferrioxamine (an iron chelator that competes with ferritin for free iron)55; (2) those with hemolysis due to glucose-6-phosphate dehydrogenase deficiency and other conditions; (3) those with idiopathic hemochromatosis; and (4) those with chronic renal failure that is managed with chronic transfusions and chelation.56 The use of erythropoietin in chronic anemias should decrease the incidence of these infections. Many in vitro studies demonstrate the importance of ironscavenging systems in the virulence of bacteria and parasites as well as of free iron-depleting defense mechanisms of the host during acute infection.51 Most patients with iron overload syndromes have other deficiencies in immune function due to reticuloendothelial blockade and functional asplenia. However, increased free iron itself predisposes to Yersinia spp. infections and has been reported as a complication of acute iron ingestion.57 In addition, a high rate of gramnegative septicemia has been associated with the intramuscular injection of iron-dextran in neonates.58 Acidosis, particularly diabetic ketoacidosis, leads to a greater availability of iron through a number of mechanisms and has been associated with infection due to Zygomycetes in patients with acidosis.53

Etiologic Agents and Pathogenesis Y. enterocolitica has been associated with septicemia, mesenteric lymphadenitis, liver and splenic abscesses, and an acute appendicitislike syndrome in patients with thalassemia and hemochromatosis.55,59 Organisms can be found in blood, stool, and lymph nodes. V. vulnificus can cause severe and often fatal bloodstream infection in these patients, and has also been reported in necrotizing fasciitis in a child with congenital spherocytosis.60 Infections caused by L. monocytogenes, Salmonella spp., Klebsiella spp., and Zygomycetes are reported in patients who are treated with desferrioxamine for iron overload associated with chronic transfusion.54 Yersinia spp., like other bacteria and fungi, require iron for growth. Unlike most other organisms, however, Yersinia spp. and the Zygomycetes are able to use iron complexed with the siderophore

CHAPTER

108

637

desferrioxamine and thrive under conditions of increased iron.57 Several studies have demonstrated diminished immune function (decreased phagocytic function, impaired T-lymphocyte responsiveness) in the presence of elevated serum iron levels. Thus, both host and bacterial factors contribute to the greater susceptibility to infection.58

Clinical Manifestations and Management The most common presentation of Y. enterocolitica infection in patients with thalassemia consists of fever, chills, and gastrointestinal complaints ranging from bloody diarrhea to frank peritonitis. Findings suggestive of acute appendicitis are associated with recovery of Y. enterocolitica from mesenteric lymph nodes. Normal hosts generally limit the organism to the intestinal mucosa, resulting in self-limited enteritis, appendicitis-like syndromes, and occasional extraintestinal infection. Early recognition and specific treatment are important. Iron chelation therapy with desferrioxamine should be temporarily discontinued. Some Y. enterocolitica are susceptible to trimethoprimsulfamethoxazole and aminoglycosides; most are susceptible to fluoroquinolones and third-generation cephalosporins. Parenteral therapy with a third-generation cephalosporin (cefotaxime) is appropriate for febrile, ill patients without focal signs of infection.

CILIARY DYSFUNCTION Primary ciliary dyskinesia is an autosomal-recessive disease with an incidence of approximately 1 in 15,000 to 30,000 births. At least 11 distinct ultrastructural abnormalities of cilia have been associated with the immotile cilia syndrome.61 Approximately half of patients with immotile cilia syndrome have situs inversus, which may be associated with Kartagener syndrome, dextrocardia, asplenia, or polysplenia. Lack of normal mucociliary clearance in the respiratory tract also leads to a higher incidence of local infection. Patients with ciliary dysfunction have respiratory symptoms in the first month of life and are often colonized with H. influenzae.61 Recurrent mucopurulent rhinitis, otitis media, and bronchitis lead to bronchiectasis in as many as 85% of patients. Risk of invasive infection is not increased (unless there is an associated splenic abnormality), and pulmonary function is usually preserved until late in the course of disease. Longitudinal studies suggest that patients diagnosed early in life and aggressively treated for infection have significantly better pulmonary function than those who are not treated for infection.62 Bronchiectasis is the most significant complication, but young infants can have frequent otitis media, leading to hearing impairment.63 The most common infecting organisms associated with bronchiectasis in primary ciliary dyskinesia are nontypable H. influenzae (47%), S. pneumoniae (32%), and Pseudomonas aeruginosa (16%).64 In upper-airway infections, typical agents of otitis media and sinusitis are common. Bronchodilator therapy may be useful in some patients. Antimicrobial therapy is aimed at treating existent infection and decreasing inflammation. Physical therapy aids in clearance of secretions and prevention of atelectasis. Unlike cystic fibrosis, ciliary dysfunction does not reduce life expectancy significantly.

ASCITES Patients with ascites due to either hepatic or renal dysfunction have increased risk for infection, particularly primary peritonitis. Historically, these infections were often due to group A streptococci or pneumococci. Although pneumococcal peritonitis remains a major problem in children with nephrotic syndrome, children with ascites due to hepatic disease are also at increased risk for peritonitis due to gram-negative bacteria. The close association of primary peritonitis

638

SECTION

R Infections in Patients with Deficient Defenses

and ascites due to hepatic disease suggests that the compromised reticuloendothelial function of the liver attributed to cirrhosis and portal hypertension allows organisms that would normally be cleared to contaminate hepatic lymph and pass into the peritoneal fluid. Because the liver is an important site for the clearance of bacteria from the bloodstream, significant compromise of this function leads to peritoneal infection, bloodstream infection, or both.65 In studies of adults with peritonitis, up to 75% of patients have bacteremia. Experimental data suggest that bacteria can also traverse the intestinal wall to seed the peritoneal cavity. Thus, gram-negative bacilli can cause peritonitis. Although peritoneal fluid cultures are positive in the majority of patients with ascites and peritonitis, culture techniques inadequate for recovery of anaerobic organisms may explain negative culture results in some patients.66 The composition of the ascitic fluid contributes significantly to the development of clinically significant peritonitis. Low levels of complement and immunoglobulins allow bacterial proliferation; there is a direct correlation between the amount of protein in ascitic fluid and susceptibility to infection. The diagnosis and treatment of primary peritonitis in patients with ascites hinge on examination of peritoneal fluid and exclusion of other sources of infection, particularly intra-abdominal infection. Despite the presence of polymorphonuclear leukocytes, culture results can be negative; empiric treatment is often required. Choice of therapy is directed by Gram stain reaction and likely sources of infection. A third-generation cephalosporin such as cefotaxime may be useful for activity against possible aerobic and facultative bowel flora and S. pneumoniae. In children repeatedly exposed to nosocomial pathogens who have received multiple courses of antibiotics, the use of a carbapenem plus metronidazole, for activity against P. aeruginosa, multiple-drug-resistant Enterobacteriaceae (Enterobacter, Citrobacter, Serratia, Klebsiella spp.), and anaerobic bacteria, may be warranted until a specific cause is identified. In adults with uncomplicated spontaneous bacterial peritonitis, a therapeutic course of a fluoroquinolone orally is often used.67 In adults with cirrhosis, prophylactic use of a fluoroquinolone decreases the number of episodes of primary peritonitis, although it also raises the risk of selecting multiple-drug-resistant pathogens.68

CYSTIC FIBROSIS Cystic fibrosis is the most common lethal genetic disease of white persons. It is caused by mutations in the cystic fibrosis transmembrane regulator (CFTR), an adenosine triphosphate-dependent chloride channel that is primarily expressed in the lung and exocrine glands.69 Patients have chronic pulmonary infection with airway inflammation, and poor growth secondary to pancreatic exocrine insufficiency and malabsorption. Sinusitis secondary to abnormal secretions and nasal polyps as well as pancreatic endocrine insufficiency can cause significant morbidity. Aggressive medical management of the pulmonary and gastrointestinal manifestations of cystic fibrosis has more than doubled the life expectancy of affected patients from a median of 14 years in 1969 to a median of 37 years in 2005.70,71

Etiologic Agents The spectrum of bacterial pathogens colonizing and infecting the lungs of patients with cystic fibrosis is relatively limited (Table 108-1).72 S. aureus, H. influenzae (mostly nontypable), and P. aeruginosa are the most commonly isolated bacterial pathogens, and several studies suggest that infections due to agents can be seen within the first 3 years of life.73,74 P. aeruginosa is the predominant pathogen in cystic fibrosis and is seen in between 45 and 62% of pediatric patients.75 Although infection with both nonmucoid and mucoid phenotypes of P. aeruginosa is associated with a more rapid decline in lung function, mucoid isolates signal chronic infection and acceleration in decline of lung function, and are very difficult to eradicate.76,77 Later in the course of disease, multiple-drug-resistant

TABLE 108-1. Age-Related Appearance of Pathogens in Cystic Fibrosis Patient Age

Pathogens

Early infancy

Staphylococcus aureus

Early childhood

Haemophilus influenzae

Adolescence

Pseudomonas aeruginosa

Late in disease

Burkholderia cepacia complex Stenotrophomonas maltophilia Achromobacter xylosoxidans Aspergillus fumigatus

gram-negative organisms such as Burkholderia cepacia complex, Stenotrophomonas maltophilia, and Achromobacter xylosoxidans are isolated. Although B. cepacia complex organisms, particularly B. cenocepacia (genomovar III) and B. multivorans (genomovar II), have been associated with both fulminant pulmonary infection and chronic colonization, the clinical significance of S. maltophilia and A. xylosoxidans is less well characterized.78 Fungal species such as Aspergillus also cause disease in patients with cystic fibrosis. A. fumigatus is important in patients who have allergic bronchopulmonary aspergillosis79 or have undergone lung transplantation.80 In a multicenter study, nontuberculous mycobacteria were isolated from sputum in up to 13% of patients with cystic fibrosis, more commonly in older patients, those with a higher forced expiratory volume in 1 second, those with lower body mass, and those who had S. aureus, but not P. aeruginosa, isolated from sputum. However, whether mycobacterial infection affects pulmonary decline remains unclear.81 The role of infection with respiratory viruses and Mycoplasma spp. in patients with cystic fibrosis has been examined in several studies. There appears to be no difference in the numbers of upper respiratory tract viral infections between children with cystic fibrosis and normal controls. However, children with cystic fibrosis have significantly more episodes of lower respiratory tract disease.82,83 Studies support a causative role for influenza in acute pulmonary exacerbations84; high morbidity and occasional fatality during acute influenza infection are recognized. During infancy, respiratory syncytial virus may be an important cause of early pulmonary morbidity, requiring hospitalization.85

Epidemiology The epidemiology of the acquisition of P. aeruginosa in cystic fibrosis is increasingly being examined. The majority of children with cystic fibrosis appear to have unique strains based on genotyping86,87 and these early isolates share many phenotypic characteristics with environmental isolates.88 Studies in siblings and at summer camps identify common strains in patients with close contact; it is uncertain whether the finding results from person-to-person spread or a common environmental source. However, outbreaks of multiple-drug-resistant P. aeruginosa in cystic fibrosis centers in Europe and Australia have identified epidemic strains that appear capable of person-to-person transmission.89,90 Epidemiologic studies also support the conclusion that B. cepacia complex can be spread from person to person. Because molecular typing of B. cepacia complex has identified unique strains in each cystic fibrosis center, studies of the epidemiology in summer camps attended by patients from different geographic areas are able to demonstrate cross-infection.91 Social contact outside the hospital has also been implicated in the spread of B. cepacia complex.92 Of the nine named species in the complex, B. cenocepacia has specific clones, especially ET12, that have most clearly been associated with transmissibility.93 A significant problem in understanding the epidemiology of B. cepacia complex is the demonstration that up to 2

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Infectious Complications in Special Hosts

years may elapse from the time of acquisition of the organism until it is detected by sputum culture.94

Pathogenesis The molecular events responsible for the close association between P. aeruginosa and the lungs of patients with cystic fibrosis have not been fully defined. Inspissated secretions and dehydrated mucus decrease normal mucociliary clearance of organisms and facilitate chronic infection. The advantage afforded P. aeruginosa and other opportunistic pathogens by these abnormalities is uncertain. Expression of increased numbers of asialylated glycolipid receptors for Pseudomonas pili on the epithelium of the lung in such patients may predispose them to colonization.95 Some studies suggest that naturally occurring airway antimicrobial peptides are inhibited by the milieu in the lungs of patients with cystic fibrosis.96 After colonization, P. aeruginosa grows in microcolonies, forming biofilms within the airway.97,98 The environment within the lung in cystic fibrosis favors the proliferation of mucoid mutants of P. aeruginosa that produce large amounts of alginate exopolysaccharide, interfering with effective phagocytosis. Once a chronic infection is established, organisms are rarely, if ever, eradicated. Infection is endobronchial in location, with bacteria loosely enmeshed in a viscous layer of mucus that contains large amounts of cellular debris, neutrophil DNA, and actin. Histopathologic analysis shows that areas of focal infection are surrounded by polymorphonuclear leukocytes. Eventually, chronic infection, fibrosis, and loss of the normal pulmonary parenchyma result. Much of the observed pathology is due to the inflammatory response to P. aeruginosa.

CHAPTER

108

639

radiograph may be helpful; however, most pulmonary exacerbations, especially in patients with advanced disease, are unassociated with significant radiographic changes, probably because of the marked abnormalities present at baseline (Figure 108-2). Sputum culture can accurately identify the bacterial pathogens colonizing lower-airway secretions102; oropharyngeal culture (a suggested alternative in the nonexpectorating patient) is a relatively insensitive measure of lower-airway pathogens.103 A technique combining the use of selective media and quantitative culture methods to identify infecting organisms in sputum is also used to evaluate therapeutic response.104 This method circumvents the problem of rapidly growing mucoid strains of P. aeruginosa, which obscure the growth of more fastidious organisms such as H. influenzae.

Clinical Manifestations Because of their inability to eradicate bacterial pathogens from the lower airways, patients with cystic fibrosis have chronic pulmonary symptoms, including chronic cough and expectoration of sputum, and progressive deterioration in pulmonary function. Acute episodes of clinical exacerbation are superimposed on chronic pulmonary symptoms. Because of wide variation in the severity of underlying manifestations, it is often difficult to define a pulmonary exacerbation on clinical criteria alone. Rather than new symptomatology, an exacerbation is more often heralded by a quantitative change in ongoing symptoms, for example, an increase in frequency and intensity of cough or in the quantity and purulence of sputum. The acute and chronic symptoms are the result of a vigorous inflammatory response to endobronchial infection. Increased airway disease manifests as tachypnea and increased work of breathing (retractions, use of accessory respiratory muscles, and wheezing). Systemic manifestations of pulmonary exacerbation include malaise, myalgia, anorexia, weight loss, and (rarely) low-grade fever. Nonpulmonary symptoms of cystic fibrosis that can mimic infection include immune complex-mediated manifestations, such as vasculitic rashes99 and arthritis.100 In addition to pancreatic insufficiency, gastrointestinal manifestations include severe constipation, intestinal obstruction, and unrecognized or unusual presentations of acute appendicitis. Sinusitis secondary to nasal polyps is common.101 Etiologic agents include the usual sinus pathogens plus organisms colonizing the lower airways in this population. In addition to causing acute and chronic disease, sinuses can serve as a reservoir for antibiotic-resistant organisms in patients undergoing lung transplantation.

Laboratory Findings Decline in results of pulmonary function studies is the most accurate and objective indicator of pulmonary exacerbation. Unfortunately, most patients younger than 5 years cannot perform pulmonary function tests reproducibly, making the diagnosis of an exacerbation more difficult. The appearance of a new pulmonary infiltrate on a chest

A

B Figure 108-2. (A) Chest radiograph of a 14-year-old boy with cystic fibrosis demonstrating typical changes of hyperinflation, bronchial wall thickening, and areas of bronchiectasis and consolidation. (B) Computed tomography demonstrates the magnitude of changes and diffuse nature of the pulmonary involvement.

640

SECTION

R Infections in Patients with Deficient Defenses

Other laboratory studies helpful in defining a pulmonary exacerbation are white blood cell count with differential and measurements of acute-phase reactants, such as erythrocyte sedimentation rate and C-reactive protein level.105 These tests are of limited value without baseline measurements for comparison. Bloodstream infection is rarely seen in patients with cystic fibrosis,106 except in a subpopulation of patients infected with B. cepacia complex,107 in patients with an indwelling central venous catheter, and in patients who are immunosuppressed as a result of lung transplantation. It is hypothesized that the low incidence of pulmonaryassociated bloodstream infection is due to the relatively low virulence of the infecting organisms, intense antibody response, and endobronchial rather than parenchymal location of the infectious process. Routine blood cultures are not indicated in patients with cystic fibrosis unless they have an indwelling central venous catheter, or are colonized with B. cepacia complex, or have undergone lung transplantation.

Management The goal of treatment is to prevent pulmonary damage from chronic infection and inflammation in the lung. Mainstays of therapy include nutritional support with pancreatic enzyme supplementation, chest physiotherapy to improve the drainage of pulmonary secretions, bronchodilator therapy to treat reactive airway disease, and antibiotic treatment to decrease bacterial density in the lungs. Antibiotic therapy in cystic fibrosis can take three forms: (1) early treatment of colonization (pre-emptive therapy); (2) suppression of chronic infection; and (3) treatment of acute pulmonary exacerbations. Aggressive pre-emptive therapy for cystic fibrosis patients with their first P. aeruginosa-positive airway culture is the standard of care in several Scandinavian countries and has been studied and is being implemented in other populations.108–112 Since 1989, studies from Denmark in intermittently colonized cystic fibrosis patients have examined the clinical and microbiological efficacy of treatment with oral ciprofloxacin in combination with inhaled colistin and reported a decrease in acquisition of chronic P. aeruginosa infections as well as improvement in lung function.112 Several other studies have demonstrated eradication of intermittent P. aeruginosa colonization with use of inhaled tobramycin.108,110,111 In these studies, when P. aeruginosa reappeared, it was most frequently a new strain.109,111 The role of suppressive therapy in individuals with chronic P. aeruginosa infections has also been demonstrated, using antibiotics that demonstrate both in vitro and in vivo antibacterial activity and those that demonstrate neither. Tobramycin has been used parenterally in pulmonary exacerbations for many years; most isolates of P. aeruginosa in patients with cystic fibrosis are susceptible. Use of high-dose inhaled tobramycin is associated with improvement in pulmonary function, decrease in the risk of hospitalization, and reduction in P. aeruginosa density in sputum without detectable ototoxicity or nephrotoxicity, and without a substantially higher rate of tobramycin resistance.113,114 Several randomized controlled trials of treatment with oral azithromycin, a macrolide agent with antiinflammatory properties, to which P. aeruginosa is routinely resistant in standard in vitro susceptibility testing, have also demonstrated clinical improvement in lung function, weight, number of exacerbation, and quality of life, but without a decrease in sputum bacterial density.115–117 Optimal antibiotic management of pulmonary exacerbations depends on knowledge of each patient’s recent sputum bacteriology and susceptibility test results. Two drugs with antipseudomonal activity are normally given – usually a b-lactam agent or fluoroquinolone plus an aminoglycoside. Because of the potential for the induction of expression of the Pseudomonas chromosomal blactamase, combinations containing two b-lactam agents are avoided. Late in the course of disease, multiple-drug-resistant strains are frequently isolated, and synergy studies may be useful to identify combinations of antibiotics with potential efficacy.118 The clinical

utility of the results of synergy testing have not been examined in cystic fibrosis. However, in a controlled clinical trial in subjects with clinical exacerbations, antibiotic therapy directed by combination susceptibility testing of P. aeruginosa, B. cepacia complex, Stenotrophomonas maltophilia, and A. xylosoxidans did not result in better clinical or bacteriologic outcomes compared with therapy directed by standard testing.119 Several problems are inherent to the selection and use of antibiotics in patients with cystic fibrosis. Chronic inflammation in the lung and localization of infection in the lumen of the airway make achievement of adequate levels of bioactive drug at the site of the infection difficult. The lack of a microbiologic endpoint for treating infections (organisms are never eradicated) makes definition of antibiotic resistance problematic. The differences in antibiotic pharmacokinetics identified in individuals with cystic fibrosis require the modification of drug dosing because the clearance of many drugs is higher compared with healthy controls.120–122 Bilateral lung transplantation is a potential life-extending therapy for selected patients with endstage pulmonary disease. However, infectious complications caused by multiple-drug-resistant organisms are a serious postoperative problem because of the immunosuppression required for transplantation.123 Although infected lungs are removed, colonization of upper respiratory and gastrointestinal tracts persists. Bronchiolitis obliterans is also a potentially fatal complication of transplantation. Recombinant human deoxyribonuclease has been demonstrated to improve pulmonary function and decrease the frequency of pulmonary exacerbations in patients with cystic fibrosis.124 Improved mucociliary clearance results from lysis of the high concentrations of leukocyte DNA in the sputum of such patients.125 Nebulized hypertonic saline, which has been shown to improve mucociliary clearance, has also been demonstrated in several short-term trials to improve mucociliary clearance, but it is not routinely recommended in the management of individuals with cystic fibrosis.126

Recent Advances Investigational efforts in cystic fibrosis have been divided between anti-infective–anti-inflammatory strategies and potentially curative therapies. Aerosol administration of other antimicrobial agents in addition to tobramycin is potentially advantageous, because high concentrations of drug can be delivered directly to the site of infection and potential toxicity can be avoided. Inhaled aztreonam has demonstrable safety and potential efficacy in the management of infections in cystic fibrosis.127 The potential role of anti-inflammatory medications, including corticosteroids and ibuprofen, has been demonstrated in well-controlled clinical trials, although treatment with systemic corticosteroids has unacceptable side effects.128,129 Other anti-inflammatory strategies currently being examined include inhibitors of proinflammatory mediators, anti-inflammatory cytokines, modulators of proinflammatory signaling cascades, antioxidants, and protease inhibitors.130 The effect of azithromycin in cystic fibrosis may be due in part to its anti-inflammatory properties.117 Potentially curative therapies include gene therapy and pharmacologic approaches to correcting the phenotype resulting from mutant CFTR. Gene therapy has received widespread attention as a potential cure for cystic fibrosis. Although the genes responsible for normal physiology and many cystic fibrosis-associated mutations have been identified, the development of effective vectors to deliver the gene to the appropriate target cells and maintain its expression is a formidable, not a trivial, challenge.128 Administration of CFTR cDNA to humans using adenovirus, adeno-associated and cationic lipid vectors has demonstrated that CFTR can be expressed in appropriate tissues, that the expression can complement the defect in vivo, and that gene transfer can be safe.131 Pharmacologic agents that address each of the classes of CFTR mutations are also under development, and some have reached phase I/II studies.132

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Epidemiology and Prevention of HIV Infection in Children and Adolescents

Prevention Approaches to the prevention of P. aeruginosa colonization in patients with cystic fibrosis have included the aggressive use of prophylactic antibiotics described above and the development of antipseudomonal vaccines for use in noncolonized patients. Preliminary studies demonstrated that a polyvalent P. aeruginosa conjugate vaccine was safe and immunogenic.133 Long-term follow-up of 26 young children from that clinical trial without prior history of P. aeruginosa infection demonstrated a longer time to infection and fewer chronic infections in the study group compared with matched controls.134

CHAPTER

109

641

Patients with cystic fibrosis should receive routine childhood immunizations (including heptavalent PCV, varicella-zoster vaccine, and H. influenzae type b vaccine).135 In addition, the patients and their families should receive influenza vaccine annually because of the high rate of morbidity associated with infection in the patients.84 There is no specific recommendation for routine administration of pneumococcal vaccine in patients with cystic fibrosis beyond childhood.

SECTION

S

Human Immunodeficiency Virus and the Acquired Immunodeficiency Syndrome CHAPTER

10 9

Epidemiology and Prevention of HIV Infection in Children and Adolescents Avinash K. Shetty and Yvonne A. Maldonado

The four recognized routes of human immunodeficiency virus (HIV) transmission are sexual contact with an HIV-infected individual, vertical transmission from an HIV-infected mother, receipt of HIVinfected blood or blood products, and parenteral exposure to HIVcontaminated equipment. Perinatal transmission is the most common source of HIV infection among infants and children in the United States.1 Since the initial description of acquired immune deficiency syndrome (AIDS) cases in infants and children more than 20 years ago,2 the epidemiology of pediatric HIV epidemic in the United States has changed substantially because of implementation of strategies to prevent vertical transmission as well as improved survival of HIVinfected children into adolescence and adulthood due to the availability of highly active antiretroviral therapy (HAART).3–6 Remarkable progress has been made in the prevention of perinatal transmission of HIV during the last decade in the United States and Europe.1,7–9 During the first 15 years of the epidemic in the United States, an estimated 15 000 infants acquired HIV infection, primarily perinatally (after routine screening of the blood supply was instituted). Approximately 3000 children died of AIDS during the same period.10 Preventing perinatal HIV transmission became a reality in 1994 when the Pediatric AIDS Clinical Trials Group (PACTG) 076 data showed that a long course of zidovudine (ZDV) prophylaxis given to HIV-1-infected mothers during early gestation and labor and then postnatally to the baby reduced perinatal HIV-1 transmission by almost two-thirds.3 In 1995, the United States Public Health Service (USPHS) issued guidelines recommending universal counseling and

testing for pregnant women and use of ZDV to reduce perinatal transmission.11 Since then, rates of perinatal HIV-1 transmission in the United States and Europe have decreased to 2% or less due to widespread implementation of universal antenatal HIV testing, combination antiretroviral treatment during pregnancy, elective cesarean section, and avoidance of breastfeeding.1,12,13 Currently, fewer than 400 infants acquire HIV from their mothers in the United States annually, primarily due to missed prevention opportunities.4,12,14 In contrast, prevention of HIV infection in children is a major public health challenge in many resource-limited nations.15 As of December 2006, the World Health Organization (WHO) estimated that 2 to 3 million children younger than 15 years are living with HIV/AIDS.16 More than 90% of these affected children reside in subSaharan Africa.16 In 2006 alone, between 410,000 and 660,000 infants became newly infected with HIV globally. An estimated half a million children die annually from HIV infection.16 Although several effective, simple, and less expensive prophylactic antiretroviral regimens are available to prevent perinatal HIV transmission, these interventions have not been widely implemented in the developing world.15

EPIDEMIOLOGY OF HIV/AIDS IN WOMEN OF CHILDBEARING AGE Characteristics of the HIV epidemic among women of reproductive age affect the pediatric HIV epidemic.17 The rate of HIV infection among women of childbearing age has continued to rise. Of reported AIDS cases in adults in the United States, women accounted for 7% in 1985, 13% in 1993, and 23% in 1999.4 In 2004, females accounted for 27% of 44 615 reported AIDS cases among adults and adolescents.18 African American and Hispanic women are disproportionately affected by the HIV epidemic. More than 75% of women with AIDS are in the reproductive age group at the time of diagnosis. Despite improved survival due to effective HAART regimens, transmission is ongoing, and young women are at highest risk.17 The average HIV seroprevalence among women in the United States has been estimated at 1.5 to 1.7 in 1000 women of childbearing age.17 Regional HIV seroprevalence rates vary, with the highest rates found among women residing in the northeast and in the south,

642

SECTION

S Human Immunodeficiency Virus and the Acquired Immunodeficiency Syndrome

especially in New York, Florida, Texas, California, and New Jersey.17 Based on these rates, it is estimated that approximately 6000 to 7000 HIV-infected women give birth in the United States each year; with a projected perinatal HIV transmission rate of 25%, an estimated 1000 to 2000 HIV-infected infants would be born each year in the United States, without preventive therapy for HIV-infected women.4,14 Of the estimated 160,963 female adults and adolescents living with HIV/AIDS at the end of 2005, 72% contracted infection through heterosexual contact, and 26% had been exposed through injection drug use.18 Whereas the association of intravenous drug use (IVDU) as a risk factor for HIV infection among women has been relatively constant, HIV-infected women with heterosexual contact as the only risk factor for HIV infection increased from 14% in 1982 to 40% of all HIV-infected women in 2000.19 Heterosexual transmission of HIV to women of childbearing age will likely continue to account for most perinatal HIV infection in the United States.

CHARACTERISTICS OF HIV/AIDS IN CHILDREN Cases of AIDS in children have accounted for 1% of all reported cases in the United States.7 Through June 2003, a total of 9419 cases of AIDS in individuals younger than 13 years of age had been reported in the United States.18 The prevalence of AIDS among children was estimated at 3.2 per 100,000 at the end of 2004.18 Perinatal transmission accounts for approximately 93% of AIDS cases in children younger than 13 years of age. The remaining 7% acquired infection through receipt of contaminated blood or blood products, sexual abuse or assault, and, very rarely, household, school; childcare, hospital, or clinic exposure.7,17,20 Fewer than 1% of cases have been reported to have no identifiable risk factor. The racial, ethnic, and geographic distribution of AIDS cases in children parallels that of women with AIDS. Perinatally acquired HIV infection occurs more frequently among black, not Hispanic (61%), and Hispanic (23%) children. In contrast, infection related to blood or blood product transfusion is more proportional to the racial and ethnic distribution of the general population.17 More than 80% of children with vertically acquired AIDS are diagnosed before 5 years of age. Approximately 95% of children with hemophilia who contract AIDS and more than 70% of those infected by transfusion are diagnosed at 5 years of age or older.17

TRANSMISSION Perinatal Transmission Mother-to-child transmission of HIV can take place in utero via transplacental infection, intrapartum by exposure to maternal blood at the time of labor and delivery, and postnatally through breastfeeding.1,8,9,21 In the absence of specific interventions, perinatal HIV transmission rates vary from about 15% to 30% among nonbreastfeeding HIVinfected women in the United States and Europe to 25% to 45% among breastfeeding populations in sub-Saharan Africa.1,9,22–24 Variability in estimated rates likely reflects differences in breastfeeding patterns, maternal and obstetric risk factors, and viral factors, as well as methodological differences between studies. Knowledge about the precise timing of transmission is crucial for the design of preventive strategies. In the nonbreastfed infant, about one-third of transmissions occur during gestation and the remaining two-thirds during delivery.8 The absolute risk for intrauterine transmission is approximately 5% and for intrapartum transmission is approximately 13% to 18%. In the breastfed infant, however, as much as one-third to one-half of overall transmission may occur after delivery during lactation.21 In 1992, Bryson et al.25 proposed a working definition of timing of vertical transmission in nonbreastfeeding infants: infants with a positive HIV culture or DNA polymerase chain reaction test in the first 48 hours of life are usually considered to have had infection in utero, whereas those with a

negative virologic test in the first week of life who subsequently become positive before 90 days of life are probably infected during the intrapartum period. Transmission in utero has been documented as early as 8 weeks of gestation by isolation of HIV from the tissue of aborted or miscarried fetuses.26–28 Intrapartum transmission can occur in a variety of ways, including direct exposure of the fetus/infant with infected maternal secretions during birth, ascending infection after rupture of membranes, or maternal–fetal microtransfusions during uterine contractions.29 Intrapartum transmission is supported by studies failing to detect HIV in infants born to HIV-infected women in the first month of life but subsequent detection of virus after 1 to 3 months of life.30–33 Retrospective studies of twins born to HIV-infected women found a higher transmission rate among those born by vaginal delivery compared with those born by cesarean delivery, and among first-born compared with second-born twins. These data support exposure to maternal virus during delivery as a likely route of transmission.34–36 The role of cesarean delivery in reducing the risk of perinatal transmission may be beneficial in certain situations, such as for women who are not receiving antiretroviral therapy or have high levels of HIV-1 in their blood at the time of labor and delivery.35,36 Although numerous maternal-, obstetric-, infant-, and virus-related factors may modify perinatal HIV transmission risk,37 the strongest predictor of both intrauterine and intrapartum transmission is the maternal serum level of HIV RNA.38–45 However, transmission can rarely occur among pregnant women with low or undetectable serum levels of HIV around the time of labor and delivery.46 Other maternal risk factors associated with higher rates of perinatal HIV infection include women with progressive symptoms of AIDS, acute HIV infection during pregnancy, and low CD4 counts.47–51 HIV viral burden in cervicovaginal secretions is an independent risk factor for perinatal HIV transmission.52 Obstetric risk factors associated with increased risk of transmission include vaginal delivery, rupture of membranes for more than 4 hours, chorioamnionitis, and invasive obstetric procedures.51,53,54 Premature infants born to HIV-infected women have a higher rate of perinatal HIV infection than full-term infants.24,40,51,55,56 Studies evaluating the relative risk of HIV transmission by route of delivery have demonstrated decreased transmission with abdominal compared with vaginal delivery.35,36 Increased transmission of HIV strains that are fetotropic is reported; isolation of HIV strains with highly conserved gene sequences from HIV-infected infants has been demonstrated, despite the large number of genetically diverse strains isolated from their mothers.57 Maternal–fetal human leukocyte antigen (HLA) concordance increases the risk of perinatal transmission,58 whereas CCR5 haplotype may be permissive or protective, depending on the specific mutation.59 In resource-poor settings where breastfeeding is the cultural norm, postnatal transmission of HIV through human milk remains a serious problem.21,60,61 HIV has been isolated from cellular and cell-free fractions of milk from HIV-infected women.62,63 Available data among breastfeeding African populations suggest that 33% to 50% of transmission may occur through breastfeeding.21,64–67 A meta-analysis estimated the overall additional risk of breastmilk transmission as 14% (95% confidence interval (CI), 7% to 22%) for established maternal infection to 29% (95% CI, 16% to 42%) for primary infection.64 Most breastmilk HIV-1 transmission occurs during the first few months of life, with a lower but continued risk thereafter.67,68 In a randomized controlled trial of breastfeeding versus formula feeding on HIV-1 transmission in Kenya, investigators found that formula feeding reduced transmission by 44% at age 2 years and that 75% of postnatal infections were acquired during the first 6 months of life.67 In a prospective breastfed cohort study in Malawi, Miotti et al.68 reported cumulative transmission rates of 3.5%, 7.0%, 8.9%, and 10.3% after 5 months, 11 months, 17 months, and 23 months of breastfeeding, respectively. Several studies have evaluated the risks of late postnatal breastmilk transmission.69–71 In an international multicenter pooled meta-analysis of over 900 mother–infant pairs, the risk of late postnatal transmission (after age 4 months) was 3.2 cases (CI 3.1 to 3.8) per year per 100 breastfed infants.71

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Epidemiology and Prevention of HIV Infection in Children and Adolescents

Risk factors for breastmilk HIV-1 transmission include women seroconverting during lactation, detectable HIV-1 RNA level in milk, bleeding or cracked nipples, subclinical and clinical mastitis, and breast abscesses.21,61

Transfusion and Hemophilia-Associated Transmission Approximately 7% of children now living with HIV infection in the United States were infected through receipt of contaminated blood or blood products.17 When given by transfusion, HIV-contaminated blood is highly infectious, with transmission rates in excess of 90%. The highest incidence of transfusion-associated AIDS occurred during the late 1970s, before routine testing of blood donors for antibody to HIV began in March 1985. As recipients of pooled coagulation factors collected from hundreds to thousands of donors, individuals with hemophilia were at highest risk of HIV infection by this route. Statistical backcalculation methods have estimated that more than 8000 hemophiliacs in the United States were infected with HIV through contaminated coagulation factors before March 1985.72 This represents 40% to 83% of hemophiliacs who received coagulation factors during that period.73,74Among transfusion recipients who did not receive pooled coagulation factors, the attack rate of HIV infection was substantially lower. By the end of 1993, more than 5000 cases of AIDS among persons with hemophilia or other coagulation disorders and 9000 cases among recipients of blood transfusions, blood components, or tissue had been reported in the United States.19 Because of the effectiveness of current HIV antibody screening of blood and blood products and heat treatment of coagulation factors, hemophiliacs and other recipients of blood or blood products represent a small proportion of the population with HIV infection in the United States. There is a small risk of transfusion-associated HIV infection caused by “silent” infections among blood donors who have recently been infected with HIV and who still test negative for HIV antibody. As a result of p24 antigen testing in 1996, the risk of transfusionassociated HIV infection was estimated at 1 in 200 000 to 1 in 2 000 000 per unit of blood transfused.75 In 1999, nucleic acid testing (NAT) of blood and plasma donations for HIV was initiated, which reduced the window period of potential silent HIV infection to approximately 13 to 15 days.7 With the implementation of NAT testing, in 2001 the Red Cross estimated the current risk of transfusion-related HIV infection in the United States to be 1 in 2 135 000.76 HIV-contaminated transfusions continue to be a problem in parts of the world where blood screening for HIV is not available.

Transmission Among Adolescents Both within the United States and globally, adolescents represent a growing population of those acquiring HIV and other sexually transmitted diseases (STDs).19 Youths between the ages of 15 and 24 account for 50% of all new HIV infections and 25% of new STDs reported annually in the United States.19 In recent years, the proportion of young people with a diagnosis of AIDS has increased. In 1999, 3.9% of all persons with a diagnosis of AIDS were aged 13 to 24, whereas in 2003, 4.7% were aged 13 to 24.77 An estimated 3897 young people received a diagnosis of HIV/AIDS in 2003, representing about 12% of all persons given this diagnosis during that year.77 Males account for 73% of all HIV/AIDS cases among adults and adolescents.18 Internationally, the HIV epidemic in adolescents is an even more major problem. It is estimated that more than 10 million people aged 15 to 24 are living with HIV/AIDS worldwide, with approximately 7000 new infections occurring each day.16 Routes of acquisition of HIV for adolescents with AIDS in the United States include the following: (1) men who have sex with men (MSM) in 44%; (2) heterosexual contact with an HIV-infected person in 19%; (3) injecting drug use (IDU) in 16%; (4) homosexual and IDU exposure in 7%; (5) hemophilia or coagulation disorders in 3%; (6) blood or blood product transfusion in 1%; and (7) other or no

CHAPTER

109

643

identified risk in 10%. Because the incubation period from the time of HIV infection by sexual contact and the development of AIDS is estimated at 10 to 12 years, it can be assumed that a large proportion of adults with AIDS in the United States were infected with HIV during adolescence. Studies among urban and inner-city adolescents have demonstrated high rates of behaviors that increase risk of HIV infection, including early initiation of sexual activity, unprotected anal and vaginal sexual activity, IVDU, and tattooing.78–86 Racial and ethnic minorities account for a disproportionate number of adolescent AIDS cases in the United States, reflecting the increase in high-risk behavior among these youth.87 African Americans represent the largest group of young people affected by HIV, accounting for 56% of all HIV infections ever reported among those aged 13 to 24.77 Young MSM, especially those of minority races or ethnicities, have high risk for HIV infection. In the seven cities that participated in the CDC’s Young Men’s Survey during 1994 to 1998, 14% of African American MSM and 7% of Hispanic MSM aged 15 to 22 were infected with HIV.88 Cross-sectional studies of HIV-infected adolescents compared with non-HIV-infected adolescents from the same community have demonstrated that HIV-infected individuals are more likely to have a history of sexual abuse; to engage in unprotected sexual activity, especially under the influence of drugs; to have multiple sexual partners; and to have multiple STDs. Efforts to develop effective preventive behavioral strategies for adolescents have largely been unsuccessful because of the lack of adequate medical and educational access for adolescents and because of the difficulty in changing adolescent high-risk behavior.

Transmission of HIV in the Healthcare Setting Accidental exposure of healthcare personnel to HIV, such as from percutaneous, mucous membrane, and cutaneous exposure, can result in HIV infection.89,90 Blood is the vehicle for most nosocomial spread of HIV. Saliva, tears, and semen have been shown to contain viable HIV particles, but they have not been implicated in nosocomial spread. Since the risk of transfusion-borne HIV has decreased in developed countries, the relative importance of needlestick transmissions has increased. The overall transmission risk from a single percutaneous needlestick accident involving HIV-contaminated blood is approximately 0.3%.91,92 The following factors appear to affect the potential for infection: use of hollow-bore needles (which carry more blood than solid needles and are thus of greater concern), depth of penetration, level of viremia of the source person’s blood, age of the contaminating material (i.e., time since it was drawn from the infected person), and whether the needle penetrated through gloves before entering the recipient’s skin. Cutaneous and mucous membrane exposures are also documented routes of transmission, although fortunately the rate is extremely low – approximately 0.09%.93 Cutaneous transmissions have almost always involved recipients with nonintact skin (because of eczema, chapped hands, abrasions, other rashes).94 Cases have been reported from broad geographic regions in the United States, but true incidence rates cannot be calculated in the absence of an active surveillance system.94 After cutaneous exposure, there is a relative lack of prospectively detected seroconversion, but several cases of seroconversion have occurred after percutaneous exposures during surveillance.94 The risk of transmission after either form of exposure is small but sufficiently appreciable to warrant medical staff concern and appropriate precautions in situations in which blood exposure is likely. Virus has been detected in saliva, but no known transmissions have occurred from salivary contamination.95,96 It is thought that neutralizing antibodies and other salivary inhibitors probably limit contagiousness.97 HIV does not appear to be found in urine98 or sweat.99 It has been isolated from tears,100 and retroviral nucleic acids have been found in feces,101 although isolation of HIV from fecal material is difficult because of bacterial and fungal contamination. Virus has been

644

SECTION

S Human Immunodeficiency Virus and the Acquired Immunodeficiency Syndrome

detected in cool aerosols produced by surgical instruments, although the risk of these vapors is unknown and almost certainly extremely low.102

pathogenic mechanisms of perinatal HIV acquisition is currently being investigated. One hypothesis suggests that short-term survivors may represent infants with HIV infection acquired in utero and longterm survivors may reflect intrapartum HIV transmission.25

NATURAL HISTORY AND PROGNOSIS

Prognostic Factors

Perinatal Infection

Considerable progress has been made in understanding the natural history of HIV infection in children. Infants with perinatally acquired HIV-1 infection have widely variable clinical courses and durations of survival. Early reports suggest a bimodal disease expression, with 20% to 25% of untreated HIV-1-infected infants rapidly progressing to AIDS or death in the first year of life, whereas the remaining 75–80% have a better prognosis, some now surviving into young adulthood.109–112 A number of clinical and immunologic factors in perinatal HIV infection have been associated with a poor prognosis. The development of AIDS-defining conditions such as PCP, HIV encephalopathy, and severe wasting is associated with a poor prognosis.108–110,113 Children with onset of symptoms at 6 to 12 months of life, especially with an initial diagnosis of PCP or other AIDS-defining conditions, die earlier than those whose symptoms begin later.114 In one cohort studied, PCP occurred at a median age of 5 months, with a median survival of 1 month, and children with other AIDS-associated diseases had a median survival of more than 12 months.24 Advanced maternal HIV-1 disease during pregnancy and high maternal viral load late in pregnancy or shortly after delivery are independently associated with disease progression in HIV-1-infected infants.115,116 Late onset of clinical symptoms, occurrence of lymphoid interstitial pneumonitis (LIP), and slow loss of CD4+ T-lymphocyte count are associated with improved survival. Longer median survival times ranging from 54 to 72 months have been reported for children who develop LIP.110,117 One study reported that progression by 6 months of age was predicted by depressed maternal and infant CD4 T-lymphocyte count are determinants of disease progression by 6 months of age; and progression by 18 months of age was associated with elevated infant viral load.118 Early diagnosis and treatment appear to improve the prognosis of perinatal HIV infection.5,6,119,120 Laboratory markers associated with progression of HIV symptoms in children with perinatal HIV infection include p24 antigenemia,121,122 decreased CD4 count,118,123 and increasing levels of b2-microglobulin,124 neopterin,125 and viral burden.120,126 In general, the presence of high viral load (> 300,000 copies/mL in infants < 12 months of age; > 100,000 copies/mL in young children) are associated with higher long-term mortality, especially if the CD4+ T-cell percentage is less than 15%.127 Syncytium-inducing HIV strains, which are more cytopathic in vitro than other strains, have been associated with rapidly progressive disease in HIV-infected adults,128–130 but not in HIV-infected infants or children to date.131 By contrast, rapidly growing HIV strains have been associated with progressive disease compared with slow- or intermediate-growing strains isolated from HIV-infected children.132

Time to AIDS The first classification system for pediatric HIV infection was published in 1987 and revised in 1994 and was developed to stage the severity of symptoms (see Chapter 111, Diagnosis and Clinical Manifestations of HIV Infection).103 The onset of clinical symptoms occurs more rapidly following perinatal exposure to virus than to transfusion-related expsoure.104 The observed period to development of AIDS is estimated to be 12 months for perinatal HIV infection compared with 41 months for children with transfusion-acquired HIV infection.105 In addition, the type of AIDS-defining condition at diagnosis varies markedly by age (Figure 109-1). Most infants diagnosed with AIDS in the first 6 months of life present with Pneumocystis pneumonia (PCP), whereas, in older children, other AIDS-defining conditions are more prevalent. Statistical modeling of the observed AIDS “incubation period” among children with perinatal HIV infection suggests that there may be two distinct patterns of disease progression, with the first sharp peak at about 4 months of age and a second median onset at about 6 years of age.106 This model projects a 20% incidence of AIDS in the first year of life, followed by an 8% annual incidence thereafter, and a median age for AIDS diagnosis of 4.8 years. Another mathematical model of the period from perinatal infection to AIDS predicted that 14% of HIV-infected infants manifest AIDS in the first year of life and 11% to 12% a year later.107 This model projects the overall rate of AIDS diagnosis of 80% to 86% by 84 months of age, which is higher than rates of AIDS found in prospective clinical studies. A major drawback of both models is that they were derived from pediatric AIDS cases reported to a county public health department and not from longitudinal cohort studies. Prospective clinical data also support the hypothesis that perinatal HIV infection manifests in a bimodal fashion, with early, severe disease in the group with short-term survival and mild to moderate symptoms and longer survival in the second group.108,109 Whether these short- and long-term survivors correlate with different 350 300

Pneumocystis jirovecii pneumonia

AIDS cases

250 200 150

Other AIDS-defining conditions

100 50 0 0 1 2 3 4 5 6 7 8 9 10 11121314 1516171819 20 212223 Age in months Figure 109-1. Acquired immune deficiency syndrome (AIDS)-defining condition for perinatally acquired cases by age at diagnosis, United States, through December 1993. Redrawn from Centers for Disease Control and Prevention. 1994 revised classification system for human immunodeficiency virus in children less than 13 years of age; official authorized agenda: human immunodeficiency virus infection codes and official guidelines for coding and reporting ICD-9 CM. MMWR 1994;43:1–17.

Survival and Antiretroviral Therapy Prognosis and survival of perinatally HIV-infected children has improved since the introduction of combination antiretroviral therapy.5,6,116,119,120 There are few population-based5,113 or longitudinal studies116,119,120,133 examining temporal trends in clinical outcomes, including disease progression and mortality. Survival estimates among children with perinatal HIV infection demonstrate a median life expectancy of 96 months.134 In a prospective study of 124 children with perinatal HIV infection, the European Collaborative Study found that 23% of the children had AIDS in the first year of life and 39% had AIDS by 4 years of age. Mortality was 10% by 1 year of age and 28% by 5 years of age. Forty-eight percent of children were still alive 2 years after the diagnosis of AIDS.135 The short-term effects of antiretroviral therapy on progression of HIV disease have been reported in small clinical studies, such as a

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Epidemiology and Prevention of HIV Infection in Children and Adolescents

documented decrease in p24 antigenemia and viral burden and an increase in CD4 counts following ZDV and dideoxyinosine therapy.136, 137 Few data are available regarding the effect of antiretroviral therapy on the long-term outcome for HIV-infected children. Data from the Italian Register for HIV Infection in Children demonstrated improved survival among children with perinatal HIV infection from 1996 through 1998 as a result of availability of combination antiretroviral therapy for these children.120 A population-based study published in 2005 attributed decreased early HIV progression and improved survival at age 3 years to more advanced antiretroviral therapy.5 A true picture of the evolution and natural history of perinatal HIV infection will only be possible when sufficient numbers of children, enrolled prospectively, have been followed over long periods of time. Current data suggest that the incidence of AIDS is highest early in life but tapers thereafter to a low but constant rate. Survival is inversely linked to development of AIDS, and prognosis is poorest among children who manifest AIDS in the first year of life. Data are still insufficient to detect definitive differences in survival based on year of birth, the temporal differences in availability of therapeutic interventions, and the effect of therapy and prophylaxis on the temporal incidence of AIDS and survival.

Transfusion and Hemophilia-Associated Infection Between 1981 and 1989, 212 cases of transfusion-associated AIDS in children younger than 13 years were reported in the United States.138 Most cases occurred in children who received transfusions in the first year of life. The median age at diagnosis of AIDS was 4 years, with a steady distribution of AIDS diagnoses of about 10% a year from 1 to 8 years after transfusion. The estimated incubation period from transfusion to development of AIDS was 3.5 years. After the diagnosis of AIDS, the median survival time was 13.7 months, not statistically different from the median survival time of 14.3 months after the diagnosis of AIDS following perinatally acquired infection. The most common AIDS-related illnesses among children with transfusionacquired infection were PCP (34%), other opportunistic infections (43%), recurrent bacterial infections (27%), LIP (21%), HIV encephalopathy (13%), and wasting syndrome (13%). A review of 72 children who acquired HIV infection through neonatal blood transfusions showed a median symptomfree period from birth to symptomatic infection of 17.8 months. This is significantly longer than the median symptomfree survival of 6.4 months that was documented among 166 children with vertically acquired perinatal HIV infection. Overall survival in the transfusion group was significantly longer (71 months) compared with the vertically infected children (44 months).139 The natural history of HIV infection associated with transfusion of coagulation factors has been described among populations in the United States and Europe. Some studies suggest that HIV infection may progress to AIDS more rapidly among transfusion-acquired cases than among coagulation factor-acquired cases, such as those in persons with hemophilia. However, transfusion recipients are usually older than persons with hemophilia and may have other medical conditions that influence outcome.140,141 Persons with hemophilia who contract HIV infection after 17 years of age have higher rates of early progression to AIDS than those who contract infection earlier.73,142,143 Predictors of progression to AIDS include a decrease in the CD4 count, increases in p24 antigenemia and level of HIV viremia, and elevated neopterin and b2-microglobulin levels.144–148

Adolescents Limited data are available regarding the long-term outcome of nonhemophiliac adolescents with HIV infection or AIDS. One study in New York city reviewed the natural history of HIV infection among the first 50 patients followed in a comprehensive evaluation and treatment program.149 Most of these patients were referred because of high-risk behavior or were self-referred to the program. The mean age

CHAPTER

109

645

of HIV testing in this population was 18.2 years; 66% of the HIVinfected adolescents were male and 80% belonged to racial or ethnic minorities. Although specific risk factors for HIV infection were not known for all of the HIV-infected adolescents, when compared with age-matched HIV-negative adolescents there were no differences in age at first sexual intercourse, prevalence of drug use, history of STDs, or behavior such as exchange of sex for money or drugs. However, HIV-infected male adolescents were more likely than age-matched uninfected males to have engaged in anal intercourse and to report a history of sexual abuse. Among the female adolescents with HIV infection, 82% acquired HIV from heterosexual contact. The evolution of HIV infection in adolescents may be influenced by the ability of primary care physicians to provide effective interventions for persons with high-risk behaviors in this population as well as adequate HIV screening for high-risk youth and early treatment for HIV-infected adolescents.81,150,151

PREVENTION OF HIV INFECTION Efforts to prevent infection with HIV have been pursued since the discovery of the virus in 1983.152 Prevention of pediatric HIV infection is primarily the prevention of perinatal HIV transmission and prevention of transmission in adolescents.153 Strategies to prevent perinatal HIV infection are different from those used to prevent adolescent and adulthood infection. Development of a safe and effective vaccine and the use of antiviral therapies as postexposure prophylaxis (PEP) are additional strategies being investigated for control of vertical and horizontal transmission of HIV.

Interventions to Reduce Perinatal HIV Transmission Major advances have been made in the prevention of perinatal transmission of HIV during the last decade in the United States and Europe.1,7–9 Strategies to prevent perinatal HIV infection must take into account the three potential points at which transmission to the fetus or infant can occur: in utero, intrapartum, and postpartum via breast milk.

Antiretroviral Prophylaxis When Administered During Pregnancy Zidovudine Prophylaxis (Long-Course Regimen) The single greatest success in the campaign against HIV infection and disease has been the > 10-fold decrease in the rate of vertical HIV transmission through the use of antiretroviral therapy in mother and child. The critical proof-of-concept study was PACTG protocol 076.3 In this study, published in 1994, monotherapy with ZDV (formerly azidothymidine (AZT)) was effective in the prevention of perinatal HIV transmission when compared with placebo. Asymptomatic, HIVinfected pregnant women with CD4 counts > 200/mm3 were enrolled between 14 and 34 weeks of gestation and were treated with ZDV orally (100 mg 5 times a day), followed by ZDV intravenously at the onset of labor (2 mg/kg initially, then continuous infusion of 1 mg/kg per hour until delivery). Infants were formula-fed and were treated with ZDV orally (2 mg/kg per dose every 6 hours for the first 6 weeks of life). The rate of perinatal HIV transmission was reduced to 8.3%, compared with 25.5% in recipients of placebo; this rate represented a 67.5% reduction in transmission among ZDV recipients (Table 109-1). Minimal adverse effects were noted in this study, the primary effect being transient anemia in the ZDV-treated neonates that did not require transfusion. The basis for the effectiveness of ZDV in reducing perinatal infection is not fully understood. In the PACTG 076 study, ZDV only modestly reduced maternal HIV-1 RNA and change in maternal HIV-1 RNA levels accounted for only 17% of the reported efficacy

646

SECTION

S Human Immunodeficiency Virus and the Acquired Immunodeficiency Syndrome

TABLE 109-1. Antiretroviral (ARV) Prophylaxis Regimens to Reduce Perinatal Human Immunodeficiency Virus-1 (HIV-1) Transmission in Nonbreastfeeding Infants Study/Sample Size/Country

Schedule of ARV Prophylaxis

Transmission Rate and Relative Efficacy

ZDV/placebo PACTG 076 n = 477 United States, France

AP + IP + PP AP = oral ZDV 100 mg 5 times/day starting at 14 weeks’ gestation IP = 2 mg/kg, then 1 mg/kg per hour IV PP (mother) = none PP (infant) = oral ZDV 2 m/kg q 6 hours for 6 weeks

At 18 months: 8.3% ZDV versus 25.5% placebo Efficacy 68%

ZDV/placebo155 n = 392 Thailand/CDC

AP + IP AP = oral ZDV 300 mg q 12 hours starting at 36 weeks’ gestation IP = oral ZDV 300 mg q 3 hours PP = none

At 6 months: 9.4% ZDV versus 18.9% placebo Efficacy 50%

ZDV/comparative156 n = 1437 Thailand PHPT/ Harvard

AP + IP + PP ZDV (long-long, LL) arm AP = oral ZDV 300mg q 12 hours starting at 28 weeks’ gestation IP = oral ZDV 300 mg q 3 hours PP (mother) = none PP (infant) = oral ZDV 2 mg/kg q 6 h for 6 weeks

At 6 months final analysis: 6.5% (LL) versus 4.7% (LS) versus 8.6% SL

3

ZDV (long-short, LS) arm AP = oral ZDV 300mg q 12 hours starting at 28 weeks’ gestation IP = oral ZDV 300 mg q 3 h PP (mother) = none PP (infant) = oral ZDV 2 mg/kg q 6 hours for 3 days

In utero transmission: 1.6% (LL + LS) versus 5.1% (SL +SS) At 6-month interim analysis: 4.1% (LL) versus 10.5% (SS) (SS arm stopped)

ZDV (short-long, SL) arm: AP = oral ZDV 300 mg q 12 hours starting at 36 weeks’ gestation IP = oral ZDV 300 mg q 3 hours PP (mother): none PP (infant) = oral ZDV 2 mg/kg q 6 hours for 6 weeks ZDV (short-short, SS) arm: AP = oral ZDV 300 mg q 12 hours starting at 36 weeks’ gestation IP = oral ZDV 300 mg q 3 hours PP (mother): none PP (infant) = oral ZDV 2 mg/kg q 6 hours for 3 days NVP/placebo164 PACTG 316 n = 1248 United States, Europe, Brazil, and Bahamas

AP + IP + PP NVP arm: AP = standard ART starting from 14 weeks’ gestation (77% combination, 23% ZDV alone) IP = ZDV 2 mg/kg, then 1 mg/kg per hour IV plus NVP 200 mg ¥ 1 PP (mother) = ART if needed PP (infant) = ZDV 2 mg/kg for 6 weeks plus NVP 2 mg/kg ¥ 1 at birth

At 6 months: 1.4% NVP versus 1.6% NVP placebo

Placebo arm: AP = standard ART starting from 14 weeks’ gestation (77% combination, 23% ZDV alone) IP = ZDV 2 mg/kg, then 1 mg/kg per hour IV plus NVP placebo PP (mother) = ART if needed PP (infant) = ZDV 2 mg/kg q 6 hours for 6 weeks plus NVP placebo at birth AP, antepartum; ART, antiretroviral therapy; BF, breastfeeding; CDC, Centers for Disease Control and Prevention; IP, intrapartum; IV, intravenous; NVP, nevirapine; PACTG, Pediatric AIDS Clinical Trials Group; PHPT, Perinatal HIV Prevention Trial PP, postpartum; 3TC, lamivudine; ZDV, zidovudine.

of ZDV.46 In addition, ZDV reduced transmission at all levels of maternal HIV-1 RNA levels. The continued efficacy of ZDV in reducing transmission even in women with low viral loads suggest that pre- and PEP of the infant during labor and delivery may have been a substantial component of protection.1

ZDV Prophylaxis (Short-Course Regimen) Although the PACTG 076 regimen is very effective, it is not feasible to implement this intervention in resource-limited countries because of cost. In addition, most transmission occurs late in pregnancy or at

the time of delivery.154 Therefore shorter, less expensive regimens that are more applicable to resource-limited countries were studied in subsequent trials. Initial studies focused on regimens that were modifications of ZDV monotherapy. The efficacy of short-course prophylaxis regimens of ZDV in reducing perinatal HIV transmission was studied among nonbreastfeeding populations in Thailand155,156 (see Table 109-1). In 1998, a placebo-controlled, randomized trial in Thailand demonstrated that a short course of twice-daily oral ZDV starting at 36 weeks’ gestation and during labor reduced mother-to-child transmission of HIV-1 by 50% at age 6 months.155

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Epidemiology and Prevention of HIV Infection in Children and Adolescents

A recent trial of shortened ZDV regimen in Thailand showed that longer (starting at 28 weeks) antepartum prophylaxis was more effective than shorter (starting at 36 weeks) antepartum prophylaxis, suggesting that a substantial proportion of in utero infection occurs between 28 and 36 weeks of gestation (see Table 109-1). Also, when women received the longer three-part ZDV prophylaxis during pregnancy, prolonged treatment of the infant did not provide additional benefit. However, when the antenatal regimen is shortened, longer treatment of the infant was found to be beneficial.156 The identical short-course ZDV antepartum/intrapartum regimen used in Thailand,155 when evaluated in a placebo-controlled trial in Abidjan, Côte d’Ivoire, involving breastfeeding populations, showed an efficacy of 37% reduction in transmission compared to placebo at age 3 months (Table 109-2).157 Another placebo-controlled trial in West Africa added a 1-week postpartum course of ZDV to shortcourse antepartum/intrapartum maternal ZDV prophylaxis. Efficacy of 38% was observed at 6 months of age among predominantly breastfed infants (see Table 109-2).158 Thus, an additional week of postpartum maternal ZDV therapy did not seem to confer any additional benefit over the antepartum/intrapartum ZDV-alone prophylaxis regimen. Long-term pooled analysis showed an efficacy of 26% by 24 months of age compared with 37% to 38% efficacy noted in infants at 3 and 6 months of age in this population with long-term breastfeeding.159 Thus, the overall efficacy of short-course ZDV is less in breastfeeding populations in sub-Saharan Africa than in formulafed populations157,158 and the early efficacy seems to diminish with prolonged periods of breastfeeding.159

Zidovudine plus Lamivudine (3TC) Prophylaxis Once the efficacy of short-course ZDV was established, studies explored whether short-course combination regimens might improve efficacy. Investigators from the United States, France, and Thailand evaluated whether combining a second antiretroviral agent such as 3TC would further enhance the efficacy of short-course ZDV in reducing transmission in nonbreastfeeding populations.160,161 The French open-label, nonrandomized study assessed the safety of perinatal 3TC-ZDV combination prophylaxis in infants, and its effects on perinatal transmission of HIV-1 in a nonbreastfeeding population.160 A total of 445 HIV-1-infected pregnant women were enrolled and received 3TC at 32 weeks’ gestation through delivery in addition to the standard PACTG 076 study ZDV prophylaxis regimen. Infants received 3TC for 6 weeks in addition to the standard 6-week course of ZDV. The transmission rate in the study group was 1.6% compared to 6.8% in a historical control group of HIV-infected mother–infant pairs in France who had received only ZDV.160 The Thailand open-label, nonrandomized trial studied the efficacy of 3TC added to short-course ZDV prophylaxis. A total of 106 HIV-1-infected pregnant women were enrolled and 3TC and ZDV were begun at 34 weeks’ gestation and given orally during labor. Infants received a 4-week course of ZDV alone. The transmission rate in the study group was 2.8% compared to 11.7% in a historical control group of HIV-infected women in Thailand who had received only short-course ZDV.161

CHAPTER

109

647

and elective cesarean section was made available (see Table 109-1). All neonates in this study received the standard 6-week ZDV course, and the overall risk of perinatal HIV transmission was very low (1.5%).164 In this study, NVP resistance developed in 15% of the women who received single-dose intrapartum NVP,164,165 and therefore, the addition of intrapartum/newborn NVP dose is not recommended in HIV-infected women who have received HAART during pregnancy.13

Antiretroviral Prophylaxis When Administered During Labor When the woman has not received any therapy during pregnancy, several efficacious intrapartum/postpartum regimens are available based on data from several international clinical trials (see Table 109-1). However, these regimens are not as effective as regimens that include antenatal as well as intrapartum and postpartum prophylaxis. In a study in Uganda, a single dose of NVP given to the mother at onset of labor and a single 2 mg/kg oral dose given to the infant at 48 to 72 hours of life was safe and reduced perinatal HIV transmission by 47% at 14 to 16 weeks of life and by 41% at 18 months in breastfeeding infants (see Table 109-2).166,167 NVP is a very potent nonnucleoside analogue with a long half-life, and excellent penetration across the placenta.168 A multicenter, placebo-controlled trial conducted among breastfeeding populations in Africa showed an efficacy of 63% at 6 weeks for ZDV/3TC given from 36 weeks’ gestation, intrapartum, and for 1 week postpartum to mothers and infants; 42% efficacy for intrapartum-postpartum ZDV/3TC; and no efficacy for intrapartum treatment only (see Table 109-2).169 Another trial conducted in South Africa comparing the efficacy of NVP (intrapartum and a single dose postpartum to mothers and to infants) versus ZDV/3TC (intrapartum and for 1 week postpartum to mothers and infants) showed that the risk of infant HIV-1 infection at 8 weeks of age was similar in the two groups (see Table 109-2).170

Postnatal Antiretroviral Prophylaxis When the mother has not received any therapy during pregnancy, or during labor and the delivery period due to detection of HIV in the mother or infant after delivery, observational data suggest that administration of oral ZDV for 6 weeks to the infant, when started within 24 hours after birth, may provide some benefit against transmission.171 A trial conducted in breastfeeding infants in Malawi comparing the efficacy of single-dose infant NVP versus single-dose infant NVP plus 1 week infant ZDV showed that the combined regimen had a superior efficacy (14.4% transmission) compared with single-dose infant NVP alone (21.9% transmission) at 6 weeks of age.172

Maternal HAART During Pregnancy Single-Dose Nevirapine (NVP) Plus Short-Course ZDV Prophylaxis Two trials conducted in Thailand and West Africa have shown that the addition of a single maternal intrapartum/neonatal NVP dose to short-course maternal ZDV (with oral ZDV during labor and either no infant prophylaxis or 1 week of infant ZDV prophylaxis) may provide increased efficacy for reducing perinatal HIV-1 transmission compared with short-course maternal ZDV prophylaxis alone.162,163 In contrast, an international, blinded, placebo-controlled phase III trial conducted in nonbreastfeeding populations in the United States, Europe, Brazil, and the Bahamas (PACTG 316) found no additional benefit from two-dose intrapartum/newborn NVP dose when women received prenatal care and standard antenatal antiretroviral therapy,

Since maternal plasma viral load is a critical predictor of perinatal HIV-1 transmission,15 the effect of maternal HAART and combination antiretroviral treatment on transmission has been studied in open-label and epidemiologic studies. Investigators from France and Thailand have reported that ZDV/3TC prophylaxis is more effective than ZDV alone in reducing perinatal HIV-1 transmission.160,161 In the United States a large multicenter prospective cohort study of 1542 HIV-1infected pregnant women and their infants (The Women and Infant Transmission Study) found that HIV-1 transmission was 20.0% (95% CI, 16.1% to 23.9%) for 396 women with no prenatal antiretroviral therapy, 10.4% (95% CI, 8.2% to 12.6%) for 710 receiving ZDV monotherapy, 3.8% (95% CI, 1.1% to 6.5%) for 186 receiving dual therapy without protease inhibitors, and 1.2% (95% CI, 0 to 2.5%) for 250 receiving combination antiretroviral therapy with protease

648

SECTION

S Human Immunodeficiency Virus and the Acquired Immunodeficiency Syndrome

TABLE 109-2. Antiretroviral (ARV) Prophylaxis Regimens to Reduce Perinatal Human Immunodeficiency Virus-1 (HIV-1) Transmission in Breastfeeding Infants Study

Schedule of ARV Prophylaxis

Transmission Rate and Relative Efficacy

157

AP + IP AP = oral ZDV 300 mg q 12 hours starting at 36 weeks’ gestation IP = oral ZDV 300 mg q 3 hours PP (mother) = none PP (infant) = none

At 3 months: 16.5% ZDV versus 26.1% placebo Efficacy 37% at 3 months

ZDV/placebo158 n = 400 Ivory Coast

AP + IP + PP AP = oral ZDV 300 mg q 12 hours starting at 36 weeks’ gestation IP = oral ZDV 600 mg ¥ 1 PP (mother) = 300 mg q 12 hours for 1 week PP (infant) = none

At 6 months: 18% ZDV versus 27.5% placebo Efficacy 38% at 6 months

ZDV-3TC/placebo (PETRA)169 n = 1797 South Africa, Uganda, and Tanzania

AP + IP + PP IP + PP IP only

At 6 weeks: Arm 1, 5.7% Arm 2, 8.9% Arm 3, 14.2% Placebo, 15.3%

ZDV/placebo n = 280 Ivory Coast

ZDV/3TC arm 1: AP = oral ZDV 300 mg q 12 hours plus 3TC 150 mg q 12 hours starting at 36 weeks’ gestation IP = oral ZDV 300 mg q 3 hours plus 3TC 150 mg q 12 hours PP (mother) = oral ZDV 300 mg q 12 hours plus 3TC 150 mg q 12 hours for 1 week PP (infant) = oral ZDV 4 mg/kg q 12 hours plus 3TC 2 mg/kg q 12 hours for 1 week ZDV/3TC arm 2: AP = none IP = oral ZDV 300 mg q 3 hours plus 3TC 150 mg q 12 hours PP (mother): oral ZDV 300 mg q 12 hours plus 3TC 150 mg q 12 hours for 1 week PP (infant) = oral ZDV 4 mg/kg q 12 hours plus 3TC 2 mg/kg q 12 hours for 1 week

Efficacy: at 6 weeks Arm 1, 63% Arm 2, 42% Arm 3, not significant At 18 months: Arm 1, 14.9% Arm 2, 18.1% Arm 3, 20.0% Placebo, 22.2% Efficacy: at 18 months Arm 1, 33% Arm 2, not significant Arm 3, not significant

ZDV/3TC arm 3: IP = oral ZDV 300 mg q 3 hours plus 3TC 150 mg q 12 hours PP (mother) = none PP (infant) = none NVP/ZDV (HIVNET 012)166,167 n = 626 Uganda

NVP/ZDV-3TC170 n = 1331 South Africa

IP + PP NVP arm: AP = none IP = NVP 200 mg μ 1 PP (mother) = none PP (infant) = NVP 2 mg/kg μ 1 at birth

At 14–16 weeks: 13.1% NVP versus 25.1% ZDV Efficacy 47% at 14–16 weeks

ZDV arm: AP = none IP = ZDV 600 mg, then 300 mg q 3 hours PP (mother) = none PP (infant) = ZDV 4 mg/kg q 12 hours for 1 week

At 18 months: 15.7% NVP versus 25.8% ZDV Efficacy 41% at 14–16 weeks

IP + PP NVP arm: AP = none IP = NVP 200 mg μ 1 PP (mother) = NVP 200 mg μ 1 PP (infant) = NVP 2 mg/kg μ 1 at birth

At 8 weeks: 12.3% NVP versus 9.3% ZDV/3TC Not statistically significant (P = 0.11)

ZDV/3TC arm: AP = none IP = ZDV 300 mg q 3 hours plus 3TC 150 mg q 12 hours PP (mother) = ZDV 300 mg q 12 hours plus 3TC 150 mg q 12 hours for 1 week PP (infant) = ZDV 4 mg/kg q 12 hours plus 3TC 2 mg/kg q 12 hours for 1 week AP, antepartum; BF, breastfeeding; FF, formula feeding; IP, intrapartum; NVP, nevirapine; PACTG, Pediatric AIDS Clinical Trials Group; HIVNET, HIV Network for Prevention Trials PP, postpartum; 3TC, lamivudine; ZDV, zidovudine.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Epidemiology and Prevention of HIV Infection in Children and Adolescents

inhibitors.13 Transmission also varied by maternal HIV-1 RNA level at delivery: 1.0% for < 400; 5.3% for 400 to 3499; 9.3% for 3500 to 9999; 14.7% for 10 000 to 29 999; and 23.4% for > 30 000 copies/mL. The odds of transmission increased 2.4-fold (95% CI, 1.7 to 3.5) for every log10 increase in viral load at delivery. In multivariate analyses adjusting for maternal viral load, duration of therapy, and other factors, the odds ratio for transmission for women receiving combination therapy with or without protease inhibitors compared with those receiving ZDV monotherapy was approximately 0.30.13 Thus, levels of maternal viral load at delivery and antenatal antiretroviral therapy were independently associated with transmission. The protective effect of therapy increased with the complexity and duration of the regimen, and maternal HAART was associated with the lowest rates of transmission.13

Role of Elective Cesarean Delivery The role of elective cesarean section delivery in reducing perinatal HIV-1 transmission was recognized before the advent of combination antiretroviral therapy during pregnancy.35,36 Data from a large international meta-analysis of 15 prospective cohort studies and a randomized controlled trial from Europe have shown that cesarean section performed before labor and rupture of membranes reduces perinatal transmission of HIV-1 by 50% to 87% independent of the use of antiretroviral therapy or ZDV prophylaxis.35,36 These studies were performed prior to the advent of HAART during pregnancy, and there was no information on maternal serum HIV-1 RNA level. Since level of maternal serum HIV-1 RNA level is an important predictor of perinatal HIV-1 transmission,38,39 it is unclear if elective cesarean section would offer any additional benefit in women successfully treated with HAART who have low or undetectable viral load.1

Avoidance of Breastfeeding Because of the documented risk of postnatal HIV-1 transmission through breastfeeding,21,61 HIV-infected mothers should be advised not to breastfeed their infants in resource-rich countries where infant formulas are safe and readily available.12,173 In resource-poor settings, exclusive breastfeeding is recommended since safe, affordable, and feasible alternatives to breastfeeding are not available. Therefore, there is an urgent need to make breastfeeding by HIV-1-infected women safer in order to prevent postnatal transmission of the virus.60 Innovative strategies have been proposed for prevention of breastmilk HIV-1 transmission in resource-poor settings, including maternal and/or infant postpartum antiretroviral prophylaxis during breastfeeding, infant vaccines, and passive immunization.21,174

Perinatal HIV Prevention Guidelines Perinatal transmission rates as low as 1% have been reported in the United States with the advent of maternal HAART, widespread implementation of the PACTG 076 regimen, access to elective cesarean section, and avoidance of breastfeeding.1,12,13,41,173 Since publication of the positive results of PACTG protocol 076, guidelines regarding prenatal HIV screening and AZT treatment of HIV-infected pregnant women have been formulated and regularly updated by the USPHS.175 Several case scenarios from those guidelines are presented in Box 109-1. In the United States, combination antiretroviral therapy with three or more drugs during pregnancy is recommended if maternal viral load is > 1000 copies/mL. In addition, elective cesarean delivery is recommended if maternal viral load is > 1000 copies/mL near delivery. Because of the proven benefit of antiretroviral prophylaxis in preventing perinatal HIV-1 transmission in women, including those with viral load < 1000 copies/mL,41 all HIV-infected women should receive prophylaxis using the PACTG ZDV regimen alone or in

CHAPTER

109

649

combination with other antiretrovirals. ZDV monotherapy administered to HIV-infected women with viral load < 1000 copies/mL has been shown to reduce perinatal HIV transmission to 1%.41 In addition, no long-term effects on women’s health have been noted among United States women enrolled in the PACTG 076 trial in terms of disease progression, mortality, viral load, or ZDV resistance.176 When the woman has not received any therapy during pregnancy, or during labor and the delivery period, ZDV should be prescribed to the neonate for 6 weeks. In such instances, other antiretroviral agents could be added to the postnatal ZDV regimen.1

Safety and Toxicity of Antiretroviral Prophylaxis With widespread use of antiretroviral prophylaxis to prevent perinatal HIV transmission in resource-poor countries, and the availability of combination antiretroviral therapy for HIV-infected mothers during pregnancy, increasing number of infants will be exposed to antiretroviral agents in utero and during the postnatal period.4,177 Animal data have shown that nucleoside analogues may be carcinogenic and can cause mitochondrial dysfunction.175 However, an extensive review of short- and medium-term follow-up data from several studies indicates that antiretroviral therapy during pregnancy has been well tolerated by mothers and infants.4,177 Except for mild, transient anemia,1 no short-term maternal or infant adverse effects have been noted with prophylactic ZDV regimens.178,179 An association between low birthweight or preterm delivery with the use of combination antiretroviral agents during pregnancy has been reported in a European study.180 However, data from a large meta-analysis of seven studies performed in the United States found no association between increased rates of low birthweight, preterm delivery, low Apgar scores, or stillbirths and the use of combination antiretroviral therapy.181 Most studies have indicated that infants exposed to commonly used antiretroviral agents such as ZDV, 3TC, stavudine, NVP, and nelfinavir during early pregnancy are no more likely to have a congenital anomaly than those in the general population.180,182 However, the French perinatal cohort study group reported possible mitochondrial abnormalities resulting in fatal outcomes in a large cohort of uninfected infants exposed to ZDV alone or ZDV-3TC during pregnancy or in the neonatal period.183 Another study from France suggested a possible association of early febrile seizure with perinatal exposure to nucleoside analogues.184 In contrast to the French studies, a retrospective review of 16 000 uninfected United States children born to HIV-infected mothers, with and without antiretroviral exposure, failed to identify any deaths related to mitochondrial dysfunction.185 Also short- to medium-term follow-up data from the European Collaborative Study involving 2414 uninfected children born to HIV-infected mothers and exposed to antiretroviral agents in utero or early life did not show any serious adverse events, including febrile seizures and clinical manifestations suggestive of mitochondrial abnormalities.186 However, the long-term outcomes of infants exposed to combination antiretroviral therapy in utero is unknown. Thus, long-term follow-up of all infants born to mothers exposed to antiretroviral therapy is recommended.175

Antiretroviral Resistance Emergence of viral resistance is a concern with global use of antiretroviral prophylaxis to prevent perinatal HIV-1 transmission.187 NVP resistance has been a topic of major interest because the WHO recommends NVP-based treatment regimens as first-ine therapy in resource-limited nations.188 A single gene mutation in HIV-1 reverse transcriptase can confer rapid resistance to commonly used antiretroviral agents such as 3TC and NVP.187 In an open-label study from France, the addition of 3TC to ZDV after 32 weeks’ gestation was associated with 3TC resistance (M184V mutation) in 39% of

650

SECTION

S Human Immunodeficiency Virus and the Acquired Immunodeficiency Syndrome

BOX 109-1. Clinical Scenarios and Recommendations for the use of Antiretroviral Drugs to Reduce Perinatal Transmission of Human Immunodeficiency Virus (HIV)a CENARIO 1: HIV-INFECTED PREGNANT WOMAN WHO HAS NOT RECEIVED PRIOR ANTIRETROVIRAL THERAPY HIV-infected pregnant women must receive standard clinical, immunologic, and virologic evaluation. Recommendations for initiation and choice of antiretroviral therapy should be the same as those used for women who are not pregnant, although the known and unknown risks and benefits of such therapy during pregnancy must be considered and discussed. The three-part zidovudine (ZDV) chemoprophylaxis regimen, initiated after the first trimester, is recommended for all HIV-infected pregnant women, regardless of prenatal HIV RNA level, to reduce the risk of perinatal transmission. The combination of ZDV chemoprophylaxis with additional antiretroviral drugs for treatment of HIV infection is recommended for infected women whose clinical, immunologic, or virologic status requires treatment and should be strongly considered for any infected woman with an HIV RNA rate > 1000 copies/mL regardless of clinical or immunologic status, and can be considered for women with HIV RNA rate < 1000 copies/mL. Women who are in the first trimester of pregnancy may consider delaying initiation of therapy until after 10–12 weeks of gestation. Elective cesarean delivery is recommended if maternal HIV RNA level remains > 1000 copies/mL SCENARIO 2: HIV-INFECTED WOMAN WHO IS RECEIVING ANTIRETROVIRAL THERAPY DURING A CURRENT PREGNANCY HIV-1-infected women receiving antiretroviral therapy in whom pregnancy is identified after the first trimester should continue therapy. ZDV should be a component of the antenatal antiretroviral treatment regimen after the first trimester whenever possible, although this measure may not always be feasible. For women receiving antiretroviral therapy in whom pregnancy is recognized during the first trimester, the woman should be counseled regarding the benefits and potential risks of antiretroviral administration during this period, and continuation of therapy should be considered. If therapy is discontinued during the first trimester, all drugs should be stopped and should be reintroduced simultaneously to avoid the development of drug resistance.

Regardless of the antepartum antiretroviral regimen, administration of ZDV is recommended during the intrapartum period and for the neonate. Recommendations for resistance testing in HIV-infected pregnant women are the same as for nonpregnant patients, that is, acute HIV infection and after virologic failure or suboptimal viral suppression after initiation of antiretroviral therapy SCENARIO 3: HIV-INFECTED WOMAN IN LABOR WHO HAS UNDERGONE NO PRIOR THERAPY Several effective regimens are available. They are as follows: (1) intrapartum intravenous ZDV followed by 6 weeks of ZDV for the neonate; (2) oral ZDV and lamivudine (3TC) during labor, followed by 1 week of oral ZDV and 3TC for the neonate; (3) single dose of nevirapine for the mother at the onset of labor followed by a single dose of nevirapine for the neonate at age 48 hours; and (4) the two-dose nevirapine regimen combined with intrapartum intravenous ZDV and 6 weeks of ZDV for the neonate. In the immediate postpartum period, the woman should have appropriate assessments (e.g., CD4+ count and HIV-1 RNA copy rate) to determine whether antiretroviral therapy is recommended for her own health SCENARIO 4: INFANT BORN TO A MOTHER WHO HAS RECEIVED NO ANTIRETROVIRAL THERAPY DURING PREGNANCY OR DELIVERY The 6-week neonatal ZDV component of the ZDV chemoprophylactic regimen should be discussed with the mother and offered for the neonate. ZDV should be initiated as soon as possible after delivery – preferably within 6–12 hours of birth. Some clinicians may choose to use ZDV in combination with other antiretroviral drugs, particularly if the mother is known or suspected to have ZDV-resistant virus. However, the efficacy of this approach for prevention of transmission is unknown, and appropriate dosing regimens for neonates are incompletely defined. In the immediate postpartum period, the woman should undergo appropriate assessments (e.g., CD4+ count and HIV-1 RNA copy rate) to determine whether antiretroviral therapy is required for her own health. The infant should undergo early diagnostic testing so that if the infant is HIV-infected, treatment can be initiated as soon as possible.

a

Discussion of treatment options and recommendations should be noncoercive, and the final decision regarding the use of antiretroviral drugs is the responsibility of the woman. A woman’s decision not to accept treatment with ZDV or other drugs should not result in punitive action or denial of care. Use of ZDV should not be denied to a woman who wishes to minimize exposure of the fetus to other antiretroviral drugs and who therefore chooses to receive only ZDV during pregnancy to reduce the risk for perinatal transmission. Adapted from Centers for Disease Control and Prevention: Public Health Service Task Force recommendations for use of antiretroviral drugs in pregnant HIV-1-infected women for maternal health and interventions to reduce perinatal HIV-1 transmission in the United States. MMWR Recomm Rep 2002;51:1–40 (for updates, please refer to http://AIDSInfo.nih.gov).

women postpartum.160 In the HIV Network for Prevention Trials (HIVNET) 012 study, 19% of women exposed to the single intrapartum NVP dose acquired NVP-resistant mutations (predominantly the K103N mutation); however, resistance was transient and no longer detectable 12 to 24 months after delivery.189 In the same study, NVP-induced genotypic resistance was detected in 46% of NVP-exposed infants who subsequently became infected, but the mutations were no longer present by 12 months of age.189 Furthermore, the mutations were different in both mothers and infants and no resistant virus was transmitted from mother to baby. Thus, in the absence of continued drug exposure, these resistance mutations become undetectable in women and infants over time and transmission of NVP-resistant virus appears less likely.187 In another study, NVP resistance mutations were detected 10 days after delivery in 32% of women who had received intrapartum NVP. Furthermore, women who received intrapartum NVP were less likely to have virologic suppression after 6 months of postpartum treatment with an NVPbased regimen.190 However, the long-term clinical consequences of NVP-induced genotypic resistance on future treatment options is unknown.191 Emergence of NVP-resistant HIV-1 in this setting is more common among women who have high viral loads, low CD4 counts, and subtype D viral infection.192 This subset of women should be considered for HAART to improve their own health and reduce the

risk of perinatal and possibly postnatal HIV-1 transmission.187,191 In contrast, NVP resistance is less likely to emerge in healthy women who do not require HAART during pregnancy, and could benefit from single-dose NVP prophylaxis to prevent perinatal HIV-1 transmission.191

Prevention of Bloodborne Transmission The risk of HIV infection from blood and blood products has dropped dramatically in the United States with the implementation of HIV antibody detection tests and NAT assays.75,76 Nevertheless, continued scrutiny, including thorough evaluation and screening of blood, organ, and tissue donors, is warranted. Cases of blood, blood product, and organ donation from HIV-infected, antibody-negative individuals have been reported.193–195 Prevention of HIV transmission in such circumstances depends on self-deferral of individuals with recent high-risk behaviors. Coagulation factors are cell-free and therefore potentially contaminated only with free virus. The use of viral inactivation procedures such as heat and solvent/detergent treatments and purification with monoclonal antibody has led to complete inactivation of HIV in pooled plasma products. Bloodborne HIV transmission is still a problem in resource-limited countries where screening tests are not available or have inadequate sensitivity.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Epidemiology and Prevention of HIV Infection in Children and Adolescents

Prevention of HIV Transmission in Adolescents Behavioral Interventions Adolescents who indulge in high-risk behaviors (sexual promiscuity and injection drug use) represent a major challenge to healthcare providers.87 A comprehensive HIV/AIDS prevention approach among youths must take into consideration their psychosocial as well as medical needs. Innovative programs that educate teens about sexual decision-making are urgently needed.196,197 Reduction of HIV infection could be accomplished through strict adherence to “safe sex” practices and avoidance of HIV-contaminated needles and syringes during illicit drug use. Safe sex practices involve monogamous relationships between HIV-negative individuals as well as consistent use of condoms and avoidance of the sharing of bodily secretions during sexual activity for all other sexual partners.196 Abstinence-only messages or delay of sexual activity are validating for many adolescents, but ineffective for adolescents who are already sexually active.87 Education regarding HIV and prevention programs has been developed and evaluated for efficacy since the beginning of the HIV epidemic. Most programs have shown little or short-lived effect in changing high-risk behaviors, despite participants’ knowledge of modes of transmission.79,82–86 In some studies, having a friend, relative, or close acquaintance with HIV infection was the strongest predictor of behavior modification. Strategies for education about and prevention of high-risk behaviors have focused primarily on informational campaigns, in which the information was delivered in a familiar and understandable format and repeated at intervals over several weeks or months.81,150,151 Targeted HIV prevention programs, which focus intense preventive approaches on high-risk, high-seroprevalence populations and areas (i.e., crack cocaine-using adolescents in large urban areas) may be effective in decreasing HIV infection rates.198 The AIDS Risk Reduction Model (ARRM) is a useful strategy for primary HIV prevention as well as secondary prevention for youths infected with HIV.199 Based on the ARRM model, for behavior change to occur, one must first label that behavior as risky, then make a commitment to change or reduce the behavior and finally, take action to perform the desired change. Community outreach education, particularly through the use of peer counselors, is critical to programs targeting HIV prevention in adolescents. Social marketing campaigns and targeted community-based HIV testing can make a major impact.200 Recent data from Uganda indicate that male circumcision may prevent HIV infection.201

CHAPTER

109

651

blood. Transmission can be minimized by strict adherence to infection control measures. Standard precautions should be followed by all HCWs. Safe work practices and techniques, personal protective equipment, and training are critical to reduce the risk of transmission.206 In particular, gloves and sharps containers are mandatory. There appears to be almost no risk from casual contact, so the HIV-infected patient can be examined in the same manner as an uninfected child. The clinician should be careful to emphasize the use of appropriate precautions and to limit more extensive precautionary measures (masks, goggles, gowns) to situations where their use is appropriate, such as during bronchoscopy. Guidelines for the management of occupational HIV exposures are summarized in Box 109-2.207 The first step after a needlestick accident is to clean the wound site immediately with a disinfectant such as alcohol.91,208 If the source person’s blood can be tested for HIV antibody in < 4 hours (preferably < 1 hour) and the result is negative, prophylaxis is not given. If rapid source testing cannot be accomplished or the source person has HIV antibody, chemoprophylaxis is given. Although failures of ZDV have occurred,91,209 PEP with ZDV is associated with a decrease of approximately 79% in the risk for HIV seroconversion after percutaneous exposure of HCWs to HIVinfected blood.210 USPHS recommendations for chemoprophylaxis and counseling after occupational exposure to HIV according to type of exposure and source material have been published.207 If the source person is HIVinfected, and the risk of transmission due to percutaneous injuries is low (i.e., solid needle or superficial injury), a basic two-drug PEP is recommended. If the source person is HIV-infected, and the risk of transmission due to percutaneous injuries is high (i.e., large-bore needle, deep puncture, visible blood on device, needle used in patient’s artery or vein), an expanded three-drug PEP regimen is recommended; addition of a fourth drug may be needed for high-risk exposures, especially if the HIV-infected source person is symptomatic, or has acute retroviral illness, end-stage AIDS or high viral load. The basic two-drug PEP regimen could include combinations of two nucleoside reverse transcriptase inhibitors, such as ZDV and 3TC, stavudine and 3TC. The expanded PEP should be protease inhibitorbased. The protease inhibitor preferred for expanded PEP regimen is lopinavir/ritonavir, although other protease inhibitors, including atazanavir, ritonavir-boosted indinavir, or nevfinavir, may be used. Prophylaxis should be initiated promptly, preferably within 1 to 2 hours of exposure, and is continued for 4 weeks, if tolerated. A pediatric infectious disease specialist with expertise in HIV should be consulted. Close follow-up is indicated to improve adherence to PEP, and monitor for adverse events, including seroconversion.207

Postexposure Prophylaxis for Possible Sexual or Other Nonoccupational Exposure to HIV The success of antiretroviral drugs in perinatal prophylaxis and needlestick exposure has led to the hope that antiretroviral PEP might also be effective after sexual intercourse between HIV-infected and uninfected partners or other nonoccupational exposure to HIV. Formal guidelines for administration of nonoccupational PEP have been published.202 These guidelines recommend the earliest possible initiation of therapy (within 72 hours of exposure) with a three-drug regimen for 28 days.202 To date, the effectiveness of PEP following nonoccupational exposure is unknown; feasibility studies have generally shown successful application of the concept.203,204 In addition to the question of efficacy,205 PEP is expensive and associated with side effects. In general, antiretroviral agents should not be used if the transmission risk is low (e.g., needlestick injury from unknown nonoccupational exposure) or if the patient presents more than 72 hours after reported exposure.7

Prevention of HIV Transmission in the Healthcare Setting Healthcare workers (HCWs) are at risk for occupational acquisition of HIV infection, primarily due to percutaneous exposure to infected

BOX 109-2. Appropriate Steps After Exposure to Human Immunodeficiency Virus (HIV) Through Needlestick Injury 1. Clean wound with disinfectant 2. Ascertain the source’s HIV and hepatitis B status 3. If the source is HIV- and hepatitis B-negative, no further steps are necessary 4. If the source is HIV-positive and the needlestick injury occurred within 6 hours, consider antiretroviral prophylaxis for the needlestick recipient (see text) 5. If the source is HIV-positive, ascertain the needlestick recipient’s HIV status. If the recipient’s baseline HIV status is negative, perform HIV enzyme immunoassay (and Western blot) at 4–6 weeks, 12 weeks, and 6 months after exposure to determine whether transmission has occurred 6. If the source has hepatitis B surface antigen, determine the needlestick recipient’s hepatitis B status. If the recipient has no prior hepatitis B infection or immunity, administer hepatitis B immunoglobulin (HBIG) and begin hepatitis B vaccination. If the source’s and recipient’s hepatitis B status cannot be ascertained within 48 hours, administer HBIG to the needlestick recipient and ascertain need for hepatitis B vaccination

652

SECTION

S Human Immunodeficiency Virus and the Acquired Immunodeficiency Syndrome

VACCINES There are many obstacles to successful development of an effective HIV vaccine.211 The single biggest obstacle has been the limited understanding of the correlates of protection.212 For example, it is not known: (1) whether humoral or cellular immunity is more important in blocking the passage of HIV from one person to another; or (2) whether free virus particles or cell-associated viruses play a larger role in transmission. The other major challenges surrounding vaccine design include the ability of HIV to establish lifelong infection, antigenic diversity, and hypervariability, and transmission of disease by mucosal route. Because HIV infection produces disease only in humans and because compromises are involved in the use of most animal models, the ultimate means to test HIV vaccines for their efficacy will be in humans. One consequence of this limitation is an ethical dilemma, particularly in regard to testing attenuated vaccines. With all attenuated vaccines, there is the concern that the vaccine will produce disease if the attenuation process is inadequate.213 Nevertheless, only by testing in humans can the proper degree of attenuation be determined. It is inevitable that the development of HIV vaccines will be a slow and controversial process,214 even though the goal of an effective vaccine is critical to international public health. The first generation of HIV vaccines focused on eliciting neutralizing antibodies to selected gp120 epitopes, because those antibodies are able to inhibit HIV replication in vitro and to prevent attachment of cell-free virus to CD4+ lymphocytes.215–218 Success with these vaccines was limited, owing both to the variability of the gp120 in wild-type HIV and to the fact that HIV transmission can occur with cell-associated virus that may not come into contact with serum-neutralizing antibodies. Attention has been focused on the development of HIV-specific cytotoxic responses, such as antibodydependent cellular cytotoxicity and major histocompatibility complexspecific cytotoxic T-lymphocyte responses, based on the assumption that these mechanisms could prevent infection by recognizing and destroying HIV-infected CD4+ cells.219–223 In addition, it has become clear that an ideal vaccine should elicit immunity at the mucosal sites of primary infection and should provide long-lasting, HIV-primed memory T and B lymphocytes. There may be, for example, an avenue to protecting the host at the point where HIV attaches to dendritic cells, probably the first step in HIV infection through mucosal surfaces. Several innovative vaccine types targeted to both cell- and antibody-mediated immunity have been developed. Whole HIV inactivated and live-attenuated viruses, recombinant synthetic peptides, DNA vaccines, and viral or recombinant subunit vaccines have been tested in animals or in humans.224–229 Another approach in vaccine trials includes generating a mixed T- and B-lymphocyte response. This is being accomplished through a sequence of vaccines that consists of replicating viral vaccines (like a recombinant canarypox vector with HIV proteins inserted) followed by subunit glycoprotein vaccines. This scheme is generically referred to as “prime boost,” with the primer being the live vaccine and the boost being the inactive glycoprotein.230 The recombinant envelope proteins (rgp120) and poxvirus vector constructs have been studied extensively in humans. Two phase III trials (one in the United States and the other in Thailand) of rgp 120 vaccines developed by VaxGen were not effective in preventing HIV infection.231,232 Concerns about the safety and potential efficacy of HIV vaccines include the possibilities that: (1) humoral immunity in response to certain HIV vaccines may produce anti-idiotypic antibodies that could interfere with aspects of normal cellular function233; (2) augmenting the humoral response to HIV might stimulate further replication of HIV via major histocompatibility complex II, the pathway traditionally invoked during development of humoral immunity to other immunogens; and (3) broad genetic diversity of HIV strains with high potential for rapid and extensive mutation could compromise the effectiveness of HIV vaccines.234 Despite these concerns, human vaccine trials have shown promise in inducing antibodies and cytotoxic T lymphocytes while also demonstrating short-term safety.

CHAPTER

110

Immunopathogenesis of HIV-1 Infection Katherine Luzuriaga and John L. Sullivan

Improved understanding of the role of virus replication in pediatric HIV-1 pathogenesis, along with the availability of potent antiretroviral therapies, has led to improved therapeutic strategies. This chapter will describe how HIV-1 interacts with the developing immune system of infants and children and how antiretroviral therapy allows preservation or reconstitution of immune function.

HIV-1 AND THE DEVELOPING IMMUNE SYSTEM HIV-1 infection is characterized by high levels of virus replication (see Chapter 234: Human ImmunodeÀciency Viruses). In infancy and early childhood, a large and renewable pool of host cells permissive to HIV-1 infection (CD4+ T lymphocytes) may contribute to persistently high plasma HIV-1 RNA levels.1 Additionally, thymic mass is high relative to body size and thymopoiesis is particularly active in the fetus and young infants. Finally, infection at a time of reduced ability to generate virus-speciÀc immune responses also may affect infection outcome. In human infants, delayed production of several antiviral effector mechanisms have been reported.2,3 A diminished capability of neonatal natural killer (NK) cells to mediate antibody-dependent, cellmedicated cytotoxicity (ADCC) of HIV-1 infected target cells has been documented. After the decline of passively-acquired maternal ADCC antibodies, delayed active generation of HIV-1 speciÀc ADCC antibodies by infected infants has been reported.4 HIV-1 speciÀc cytotoxic T-lymphocyte (CTL) responses in vertically-infected infants are less vigorous and appear later in primary infection than in adults.5 These relative deÀciencies in HIV-1-speciÀc immunity may preclude the effective containment of virus replication in early infection and thus contribute to the relatively rapid disease progression often observed following vertical infection.

IMMUNE CONSEQUENCES OF UNTREATED HIV-1 INFECTION HIV-1 infection results in a wide spectrum of immune defects (Box 110-1). While the bases for several defects are not well-understood, many are thought to result from either the direct infection or the activation of antigen-presenting cells (dendritic cells and macrophages) and T lymphocytes. The activation and direct infection of CD4+ T lymphocytes are thought to be the major bases for high rates of CD4+ T-lymphocyte turnover and depletion that occur throughout HIV-1 infection. Reduced production of bone marrow precursors, thymic dysfunction, and immune-mediated destruction of infected CD4+ T-lymphocyte by immune mechanisms also may contribute to CD4+ lymphocyte depletion. In addition to a reduction in CD4+ T-lymphocyte numbers, alterations in the CD4+ T-lymphocyte repertoire and functions have been documented. These alterations in CD4+ T-lymphocyte numbers and function result in reduced CD4+ T lymphocytes help for several immune effector functions, including antibody formation and antiviral CD8+ T-lymphcoyte responses. Most studies of HIV pathogenesis have relied on the examination of numbers and function of CD4+ T lymphocytes in peripheral blood. However, several investigators demonstrated that regardless of the route of infection, the gastrointestinal tract appears to be a major site

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Immunopathogenesis of HIV-1 Infection

BOX 110-1. Immunopathologic Consequences of Human Immunodeficiency Virus (HIV) Infection T LYMPHOCYTES Clinical manifestations Impaired delayed-type hypersensitivity (new and recall) Opportunistic infections Chronic active viral infections (varicella-zoster virus, cytomegalovirus, Epstein–Barr virus) ? Neoplasms Laboratory Manifestations Lymphopenia Selective deÀciency of CD4+ T lymphocytes Impaired mitogen–antigen responses Impaired alloantigen reactivity Impaired production of cytokines (IL-2, IFN-g, others) Impaired T-lymphocyte cytotoxicity B LYMPHOCYTES Clinical manifestations Hypergammaglobulinemia Impaired antibody responses (primary and secondary responses; T-lymphocyte-dependent and -independent antigens) Laboratory Manifestations Elevated numbers of circulating B lymphocytes spontaneously secreting immunoglobulins Circulating immune complexes Mononuclear Phagocytes Clinical manifestations Impaired delayed-type hypersensitivity Impaired granuloma formation Opportunistic infections (Mycobacterium tuberculosis, Mycobacterium avium complex) ? Pneumococcal bacteremia Laboratory manifestations Impaired splenic clearance (antibody-coated red blood cells) Increased levels of circulating “pocked” red blood cells Elevated levels of tumor necrosis factor NATURAL KILLER LYMPHOCYTES Clinical manifestations Chronic herpesvirus infection Neoplasms Laboratory manifestations Impaired cytotoxicity of large granular lymphocytes NEUTROPHILS Clinical manifestations Candidiasis Laboratory manifestations Neutropenia Impaired chemotaxis IFN, interferon; IL, interleukin.

of viral replication and CD4+ T-lymphocyte depletion from early through chronic infection.6,7 Intestinal lamina propria memory CD4+ T lymphocytes, in particular, appear to be depleted early in HIV-1 infection and depletion of intestinal CD4+ T lymphocytes often precedes decreases in peripheral blood CD4+ T lymphocytes. This work supports the concept that antiretroviral therapy should be initiated in early infection in order to avoid immune depletion.

EARLY EFFECT OF ANTIRETROVIRAL THERAPY ON IMMUNE FUNCTION Several studies have evaluated the virologic and immunologic consequences of early therapy in HIV-1 positive infants. Longterm control of plasma viremia to levels below the detection limits of currently available assays has been observed in the majority of infants who began combination antiretroviral therapy prior to 3 months of

CHAPTER

110

653

age.8,9 While total CD4+ lymphocyte counts and general immune function appeared intact in children who received early therapy, persistent HIV-1 speciÀc immune responses (antibodies, CD4+ lymphoproliferative responses, and CD8+ T-lymphocyte responses) were not detected in the majority of infants. The lack of persistent HIV-1 speciÀc immune responses contrasted with the development of antibody and lymphoproliferative responses to routine infant vaccines (e.g., tetanus) and CD8+ T-lymphocyte responses to other viral infections, suggesting that the defect was HIV-1 speciÀc. The lack of persistent HIV-1 speciÀc immune responses in infants contrasts with the persistent HIV-1 speciÀc immune responses reported in adults who receive combination antiviral therapy within 3 to 4 months of acquisition of infection. These observations have led to a proposed trial, which will provide an HIV-1 speciÀc vaccine following early combination therapy, in an attempt to boost HIV-1 speciÀc immune responses.

IMMUNE RECONSTITUTION FOLLOWING ANTIRETROVIRAL THERAPY OF ESTABLISHED HIV-1 INFECTION Gastrointestinal lymphoid tissues are a major site of viral replication and CD4+ T-lymphocyte depletion. Intestinal CD4+ T lymphocytes are less effectively restored than peripheral blood CD4+ T lymphocytes following intensive antiretroviral therapy. However, because peripheral blood CD4+ T lymphocytes are more easily obtained, peripheral blood CD4+ T-lymphocyte numbers have been more commonly used to measure responses to antiretroviral therapy. Increases in peripheral blood CD4+ T-lymphocyte numbers have been observed following antiretroviral therapy of adults and children, even in individuals with signiÀcant CD4+ T-lymphocyte depletion prior to therapy.10,11 An early increase in CD4+ T-lymphocyte numbers following antiretroviral therapy is likely due to the expansion of existing peripheral CD4+ T lymphocytes. This can occur even in the absence of complete suppression of viral replication, but sustained increases in CD4+ T lymphocyte numbers and return of CD4+ Tlymphocyte function appears to occur only following durable and profound suppression of viral replication. The potential for peripheral blood CD4+ T-lymphocyte reconstitution appears to be greater in children than in adults12 and probably reflects greater thymic activity in early life.13 The restoration of pathogen-speciÀc CD4+ T-lymphocyte function is thought to account for the decreased incidence of several opportunistic infections observed following the widespread use of potent antiretroviral regimens. Improvements in peripheral blood CD4+ T-lymphocyte counts following antiretroviral therapy permit the discontinuation of prophylaxis for speciÀc opportunistic infections.14 An acute inflammatory condition known as “immune reconstitution inflammatory syndrome (IRIS)” has been described in up to 10% to 15% of adults and children who initiate antiretroviral therapy (reviewed in references 15,16). IRIS most commonly occurs within the Àrst 90 days of treatment and appears to be more common in individuals who initiate antiretroviral therapy with advanced disease or marked peripheral blood CD4+ T-lymphocyte depletion. As the immune system improves following therapy, IRIS is thought to represent an over exuberant inflammatory response to an opportunistic infection, with the worsening of clinical symptoms associated with a previously treated opportunistic infection or the unmasking of a subclinical opportunistic infection. Herpes simplex virus, herpes zoster virus, cytomegalovirus, atypical mycobacteria, tuberculosis, cryptococcosis, and hepatitis C virus are the opportunistic infections most commonly implicated in IRIS. Many experts thus recommend evaluation and prophylaxis or treatment for opportunistic infections prior to the initiation of antiretroviral therapy; this appears to be particularly important for atypical mycobacteria, tuberculosis, and cryptococcal infections.17 Most individuals who develop IRIS on therapy can be managed by continuing antiretroviral therapy while treating the opportunistic infection and administering antiinflammatory agents.

654

SECTION

CHAPTER

S Human Immunodeficiency Virus and the Acquired Immunodeficiency Syndrome

111

Diagnosis and Clinical Manifestations of HIV Infection Heidi Schwarzwald and Mark W. Kline

The diagnosis of human immunodeficiency virus (HIV) infection in children less than 18 months of age requires specialized laboratory testing. The clinical manifestations of HIV infection in children are varied and often nonspecific. Prompt recognition is a prerequisite to early provision of potentially life-prolonging prophylactic and therapeutic medications as well as appropriate management of intercurrent medical illnesses. This chapter focuses on clinical manifestations directly related to the pathologic effects of HIV on various organ systems. Infectious complications of HIV infection are discussed in Chapter 112.

DIAGNOSIS The detection of HIV infection in a child older than 18 months (or an adult) is straightforward because antibody detection by enzyme immunoassay (EIA), confirmed by Western blot, is both sensitive and specific. In younger children the presence of transplacentally acquired maternal antibodies complicates evaluation.1 Antibody assays can determine whether an infant was exposed to the risk of HIV infection in utero but cannot separate the at-risk infant from the infected one. The critical determination of infection status is performed using one of several strategies: detection of the virus in culture, detection of viral-specific proteins in serum (generally the major capsid protein, p24), or finding virus-specific genetic sequences through the use of polymerase chain reaction (PCR) assays or equivalent tests. These techniques vary in cost, sensitivity, and time required to obtain a result (Table 111-1). A schema for diagnostic approach is age-dependent (Table 111-2).

normal during the neonatal period. The most common acquired immunodeficiency syndrome (AIDS)-defining illness in infants is pneumonia caused by Pneumocystis jirovecii (P. carinii). Common signs of HIV infection in infancy include failure to thrive, oral candidiasis, chronic diarrhea, and hepatosplenomegaly. All EIA and rapid-test kits are based on the detection of antibodies to HIV. Because maternal immunoglobulin G (IgG) can cross the placenta and can remain detectable in an infant’s serum up to 18 months of age, antibody-based EIA, and Western blot assays are not reliable diagnostic tests for HIV-exposed infants. Several other diagnostic tests are currently in use and under study for the purpose of early diagnosis of HIV in infants. According to the Centers for Disease Control and Prevention (CDC), an infant less than 18 months of age is considered HIVinfected if he or she has positive results on two separate determinations (excluding cord blood) from one or more of the following tests: HIV culture, HIV PCR, HIV p24 antigen; or meets criteria for AIDS diagnosis based on the 1987 AIDS surveillance case definition.2 Cord blood is excluded because it has a high likelihood of being contaminated with maternal blood. Two tests are required because none of the current testing methods is 100% sensitive and specific, particularly during the first 2 weeks of life. For infants who are not breastfed, HIV infection can be excluded if HIV DNA PCR assay results obtained at birth, 4 to 6 weeks of age, and 8 to 16 weeks of age are negative.3 If an infant is breastfed, HIV infection cannot be definitively excluded until the infant has a negative diagnostic test 6 months after breastfeeding has ceased. It is common practice to confirm any diagnosis or exclusion of HIV infection with a negative EIA, rapid test, and/or Western blot at age 18 months. However, this approach is probably neither necessary nor cost-effective.4

Nonsubtype B Infection Because many of the diagnostic tests for HIV detect antigens and/or genetic material, test sensitivity and specificity may be compromised in populations where HIV subtype B does not predominate.5 A study in Thailand, however, found that in a population where 92% of mothers were infected with subtype E, DNA and RNA PCR specifically designed to be sensitive to subtype E were 100% sensitive at 2 months of age.6

Diagnosis in Children

Diagnosis in Infants An infant can acquire HIV from the mother during pregnancy, labor, or breastfeeding. Infants with vertically acquired HIV can be clinically

Children > 18 months of age can be diagnosed using a combination of clinical signs and symptoms and laboratory tests. Children often manifest recurrent bacterial infections, failure to thrive or wasting,

TABLE 111-1. Assays to Detect Human Immunodeficiency Virus Infection Sensitivity (+ to +++)

Cost ($ to $$$)

Time

Advantages

Disadvantages

Antibody detection (enzyme immunoassay/Western blot)

+++

$

Hours

Sensitive, suitable for screening

Transplacentally acquired maternal antibodies produce false-positive results in infants

p24 antigen detection

+

$

Hours

Inexpensive, quantitative

Poor sensitivity; false-positive results in the first week of life

Immune complexdissociated p24 detection

++

$

Hours

More sensitive than p24; quantitative

Moderately sensitive; reproducibility difficult; false-positive results in the first week of life

Culture

+++

$$$

Weeks

Sensitive, highly specific; virus available for genotypic and phenotypic analyses (e.g., antiviral sensitivity)

Requires weeks to complete, isolation facility, trained personnel; limited availability (research); quantification, although possible, is difficult and time-consuming

Polymerase chain reaction (PCR)

+++

$$

Days

Sensitivity equal to culture, but faster and cheaper; quantitative

Risk of false-positive results; limited data about virus phenotype/antiviral sensitivity

Assay

+, lowest sensitivity; +++, highest sensitivity; $, lowest cost; $$$, highest cost.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Diagnosis and Clinical Manifestations of HIV Infection

CHAPTER

111

655

TABLE 111-2. Schema for Evaluating Children at Risk for Human Immunodeficiency Virus (HIV) Infection Age at Initial Presentation < 2 weeks

2 weeks–6 months

6–15 months

> 15 months

Stage of Evaluation

Test

Schedule

Comments

Maternal HIV status unknown

EIA/Western blot Immediate (on mother or infant)

If positive, warrants further evaluation

Mother HIV-infected

PCR (or culture)

Birth, 4–6 weeks, and 8–12 weeks

PCR and culture have equal sensitivity. Negative PCRs at 4–6 weeks and 8–12 weeks are evidence that the child is not infected. Start PCP prophylaxis at 6 weeks of age pending results.

Maternal HIV status unknown

EIA/Western blot on mother or infant

Immediate

If positive, proceed to next step to determine whether the child is also infected.

Mother HIV-infected

PCR (or culture)

Initial visit, 4–6 months of age

Consider two PCR assays 4 weeks apart, then one at ≥ 4 months old.

Maternal HIV status unknown

EIA/Western blot on mother and infant

Immediate

If EIA/Western blot test for either person is positive, proceed to next step.

Mother HIV-infected

EIA/Western blot, and PCR (or culture)

EIA/Western blot and PCR/culture once immediately; repeat EIA every 3 months until negative

EIA in uninfected children seroreverts to negative at a median age of 10 months. Perform PCR (or culture) initially to determine need for PCP prophylaxis or antiretroviral treatment. If EIA/Western blot does not serorevert to negative by 15–18 months old, child is infected.

All at-risk children

EIA/Western blot

Once as a screen, twice EIA/Western blot suffice at > 15 months of age; if confirmation needed positive assay indicates infection. To exclude lab errors, repeat assay in a high-risk child may be worthwhile, regardless of result of first assay.

EIA, enzyme immunoassay; PCP, Pneumocystis carinii pneumonia; PCR, polymerase chain reaction.

persistent lymphadenopathy, developmental delay, or oral and pharyngeal thrush. Like adults, children > 18 months of age can be diagnosed using an EIA or rapid serum test plus a confirmatory test.

DIAGNOSTIC TESTS Enzyme Immunoassay The EIA test uses purified HIV-specific protein as antigen and is often produced using recombinant technology. EIA tests are highly sensitive (> 99%), but have some limitation. Because maternal antibodies are present in an infant’s blood for up to 18 months after birth, the test is only accurate in patients over 18 months of age. Additionally, there is a “window period” (usually 6 to 12 weeks after infection) during which an infected patient can have a negative EIA antibody result. Newer EIAs utilize improved technology that can narrow the window. EIAs are designed for screening large numbers of patients, making them suitable for centralized laboratories, but they may not be costeffective in other circumstances. Because the EIAs are screening tests, all positive results must be confirmed by a secondary test, usually the Western blot test. However, in some settings with a high prevalence of HIV, a second EIA or use of a rapid diagnostic test can be an acceptable confirmatory test.7,8

Western Blot A Western blot is a polyacrylamide gel electrophoresis methodology that detects bands of proteins specific to HIV antibodies. If no bands are seen, the Western blot is negative. If most or all of the specified protein bands are seen, the Western blot is positive. The Western blot test can be inconclusive or indeterminate. A Western blot assay is more specific for HIV than an EIA but is generally less sensitive and is more difficult technically. A few individuals have persistently

indeterminate Western blot results, with the presence of one or more bands that do not meet the diagnostic criteria. If at low risk, these patients are almost certainly not infected.9 If the patient is at high risk for HIV infection, an indeterminate Western blot can represent a transition to positive status. The combination of EIA and Western blot has proved highly accurate, with sensitivity and specificity sufficiently adequate to be used for screening. Illnesses that are known to cause false-positive HIV EIA with an indeterminate Western blot include autoimmune diseases, certain viral infections, syphilis, and hematologic malignancies. Pregnancy can also cause a false-positive EIA.10 A survey of a low-prevalence population of United States Army recruits found a false-positive rate of 1 in 134 187 individuals.11

Rapid Tests Several rapid tests detect HIV antibodies, many as accurately as EIA. These tests are also only accurate in those > 18 months of age, have the same window period of 6 to 12 weeks, and all positive rapid HIV tests must be confirmed by another test. Most rapid tests can be performed on blood obtained from a fingerprick method, which performance requires little training. Many commercial rapid tests are available, and most are 99% to 100% sensitive and specific.12,13 Although some rapid tests can be performed on saliva or urine, their quality and reliability have not been adequately validated. One study, conducted by the World Health Organization (WHO) and Joint United Nations Programme on HIV/AIDS (UNAIDS), did not find the urine and saliva test kits they investigated to be sufficiently sensitive or specific.14 Other saliva tests have been found to be highly reliable. In a 2000 study in a high-prevalence area in South Africa the rapid tests performed well for diagnosing. In addition, when two different rapid test kits (testing for two different HIV-associated antibodies) were used, two positive test results were almost 100% sensitive and specific for HIV infection.7 Because the results were available promptly, the effectiveness of posttest counseling was increased.8

656

SECTION

S Human Immunodeficiency Virus and the Acquired Immunodeficiency Syndrome are expensive and labor-intensive, and they require several weeks to obtain results.

DNA PCR DNA PCR technology has allowed for the rapid, accurate diagnosis of HIV-infected infants. In developed countries, DNA PCR has replaced HIV cultures as the primary methodology used for the diagnosis in children < 18 months of age. Sensitivity is 38% at 48 hours of life and 93% at 2 weeks of life.15 These tests have a sensitivity of 90% to 100% and a specificity of 95% to 100% in infants > 1 month of age.16–18 For infants born to HIV-infected mothers, current recommendations include performance of DNA PCR at delivery, 4 to 6 weeks, and 8 to 16 weeks.3 Cord blood should not be used.

CLASSIFICATION OF INFECTION IN CHILDREN The CDC employs two separate definitions for HIV infection in children, reflecting the difficulty of establishing the diagnosis in infants with vertical acquisition of HIV and those with transplacentally derived maternal HIV antibody (Table 111-3 and Box 111-1).2 An infant younger than 18 months is considered HIV-infected if born to an HIV-infected mother and testing positive for HIV or antigen on at least two different blood specimens. Cord blood should not be used. An infant is also considered HIV-infected if the CDC surveillance case definition for AIDS is met. The AIDS case definitions for children and adults are similar, with noteworthy exceptions.23 Lymphoid interstitial pneumonitis (LIP) and multiple or recurrent serious bacterial infections are only AIDSdefining conditions for children. Some infections, including certain cytomegalovirus and herpes simplex virus infections, and central nervous system (CNS) toxoplasmosis, are only AIDS-defining conditions for children older than 1 month and adults. The expanded definition for AIDS in adolescents and adults, which became effective in 1993, does not apply to individuals younger than 13 years.24 The CDC has developed a separate classification system to describe the spectrum of HIV disease in children, including HIVexposed infants with undetermined infection status. This system was most recently updated in 1994 (see Table 111-3).2 The revised system employs two axes to indicate severity of clinical signs and symptoms and degree of immunosuppression. Once classified as to HIV status, an infant or child is not subsequently reclassified into a less severe category, even if improvement in clinical or immunologic status occurs in response to antiretroviral therapy or other factors. The classification code for an infant with HIV vertical exposure and indeterminate (unconfirmed) infection status has an “E” (for vertically exposed) as a prefix (e.g., EN1 for an infant with no signs or symptoms of disease, no evidence of immunosuppression, and indeterminate infection status).

RNA PCR The sensitivity of HIV RNA measurements has been compared with DNA PCR for the diagnosis of neonatal HIV infection. In one study, specimens obtained from 49 HIV-infected infants and 8 uninfected infants were tested by both assays.19 Quantitative plasma RNA testing was equivalent to DNA PCR specificity, and equivalent or better in sensitivity.20 Only 2 of the 49 HIV-infected infants had received antiretroviral therapy before plasma specimens were collected. This is a potentially important issue, because the sensitivity of HIV RNA measurements for the diagnosis of neonatal HIV infection may be decreased if specimens are collected while the infant is receiving antiretroviral therapy. Since most infants have high plasma HIV-1 RNA measurements, HIV RNA testing should be sensitive despite antiretroviral therapy. Plasma HIV-1 RNA, therefore, may be a useful test for the early diagnosis of perinatal HIV-1 infection.20

p24 Antigen p24 is a major core protein of HIV. Several EIAs that detect p24 antigen have been developed and are simple to perform and inexpensive. However, they are not appropriately sensitive in the first 6 months of life (e.g., 6% to 53% in one large Ugandan study).21 p24 antigen tests do not detect p24 antigen that is bound to antibody; heating plasma or serum or incubating in acid breaks antibody– antigen complexes and increases the sensitivity of the p24 antigen test.22 Because of its lack of sensitivity, p24 antigen testing is not recommended for diagnosis of HIV in infancy.

CLINICAL MANIFESTATIONS Early Signs and Symptoms of Vertical Infection Infants with vertically acquired HIV infection are usually asymptomatic and have normal physical findings during the neonatal period. A purported congenital HIV syndrome, encompassing microcephaly, a prominent boxlike forehead, flattened nasal bridge, short nose with

HIV Peripheral Blood Lymphocyte Coculture Before PCR technology became available for the diagnosis of HIV, culture was used to diagnose infection in infants. However, cultures

TABLE 111-3. Summary of 1994 Revised Classification of Human Immunodeficiency Virus (HIV) Infection in Children Younger Than 13 Yearsa Immunologic Categoriesb

Summary of Categories

Immunologic Category 1. No evidence of suppression 2. Evidence of moderate suppression 3. Severe suppression

N No Signs/ Symptoms

A Mild Signs/ Symptoms

Bc Moderate Signs/ Symptoms

Cc Severe Signs/ Symptoms

< 12 Months Old per mm3 %

1–5 Years Old per mm3 %

6–12 Years Old per mm3 %

N1

A1

B1

C1

≥ 1500

≥ 25

≥ 1000

≥ 25

≥ 500

≥ 25

N2

A2

B2

C2

750–1499

15–24

500–999

15–24

200–499

15–24

N3

A3

B3

C3

< 750

< 15

< 500

< 15

< 200

< 15

a

Children whose HIV infection status is not confirmed are classified by using the table above with a letter E (for perinatally exposed) placed before the appropriate classification code (e.g., EN2). Immunologic categories based on age-specific CD4+ T-lymphocyte counts and percentage of total lymphocytes. c Both category C and lymphoid interstitial pneumonitis in category B are reportable to state and local health departments as acquired immunodeficiency syndrome (AIDS). From Centers for Disease Control and Prevention. 1994 revised classification system for human immunodeficiency virus infection in children less than 13 years of age. MMWR 1994;43:1. b

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Diagnosis and Clinical Manifestations of HIV Infection

CHAPTER

111

657

BOX 111-1. Clinical Categories for Children with Human Immunodeficiency Virus (HIV)a CATEGORY N: NOT SYMPTOMATIC Children who have no signs or symptoms considered to be the result of HIV infection or who have only one of the conditions listed in category A CATEGORY A: MILDLY SYMPTOMATIC Children with two or more of the conditions listed below but none of the conditions listed in categories B and C: • Lymphadenopathy (≥ 0.5 cm at more than two sites; bilateral, one site) • Hepatomegaly • Splenomegaly • Dermatitis • Parotitis • Recurrent or persistent upper respiratory infection, sinusitis, or otitis media CATEGORY B: MODERATELY SYMPTOMATIC Children who have symptomatic conditions other than those listed for categories A or C that are attributed to HIV infection; examples of conditions in clinical category B include but are not limited to: • Anemia (< 8 g/dL), neutropenia (< 1000/mm3), or thrombocytopenia (< 100,000/mm3) persisting ≥ 30 days • Bacterial meningitis, pneumonia, or sepsis (single episode) • Oropharyngeal candidiasis, persisting > 2 months in children > 6 months of age • Cardiomyopathy • Cytomegalovirus infection, with onset before 1 month of age • Diarrhea, recurrent or chronic • Hepatitis • HSV stomatitis, recurrent (> 2 episodes within 1 year) • HSV bronchitis, pneumonitis, or esophagitis with onset before 1 month of age • Herpes zoster (shingles) involving ≥ 2 distinct episodes or > 1 dermatome • Leiomyosarcoma • LIP or pulmonary lymphoid hyperplasia complex • Nephropathy • Nocardiosis • Fever persisting > 1 month • Toxoplasmosis, onset before 1 month of age • Varicella, disseminated (complicated chickenpox) CATEGORY C: SEVERELY SYMPTOMATIC Children who have any condition listed in the 1987 surveillance case definition for AIDs, with the exception of LIP • Serious bacterial infections, multiple or recurrent (i.e., any combination of ≥ 2 culture-confirmed infections within a 2-year period), of the following types: septicemia, pneumonia, meningitis, bone or joint infection, or abscess of an internal organ or body cavity (excluding otitis media, superficial skin or mucosal abscesses, and indwelling catheter-related infections) • Esophageal or pulmonary (bronchi, trachea, lungs) candidiasis • Coccidioidomycosis, disseminated (at site other than or in addition to lungs or cervical or hilar lymph nodes) • Extrapulmonary cryptococcosis • Cryptosporidiosis isosporiasis with diarrhea persisting > 1 month • Cytomegalovirus disease with onset of symptoms at age > 1 month (at a site other than liver, spleen, or lymph nodes) • Encephalopathy (at least one of the following progressive findings present for at least 2 months in the absence of a concurrent illness other than HIV infection that could explain the findings): (1) failure to attain or loss of developmental milestones or loss of intellectual ability, verified by standard developmental scale or neuropsychological tests; (2) impaired brain growth or acquired microcephaly, demonstrated by head circumference measurements, or brain atrophy, demonstrated by CT or MRI (serial imaging is required for children < 2 years of age); (3) acquired symmetric motor deficit manifested ≥ 2 of the following: paresis, pathologic reflexes, ataxia, or gait disturbance • HSV infection causing a mucocutaneous ulcer that persists for > 1 month; or bronchitis, pneumonitis, or esophagitis for any duration affecting a child > 1 month of age • Histoplasmosis, disseminated (at a site other than or in addition to lungs or cervical or hilar lymph nodes) • Kaposi sarcoma • Lymphoma, primary, in brain • Lymphoma, small, noncleaved cell (Burkitt), or immunoblastic or large-cell lymphoma or B-cell or unknown immunologic phenotype • Mycobacterium tuberculosis, disseminated or extrapulmonary • Mycobacterium avium complex or Mycobacterium kansasii, disseminated (at site other than or in addition to lungs, skin, or cervical or hilar lymph nodes) • Pneumocystis jirovecii pneumonia • Progressive multifocal leukoencephalopathy • Salmonella (nontyphoid) septicemia, recurrent • Toxoplasmosis of the brain with onset at > 1 month of age • Wasting syndrome in the absence of a concurrent illness other than HIV infection that could explain the following findings: (1) persistent weight loss > 10% of baseline or (2) downward crossing of ≥ 2 of the following percentile lines on the weight-for-age chart (e.g., 95th, 75th, 50th, 25th, 5th) in a child ≥ 1 year of age or (3) < 5th percentile on weight-for-height chart on two consecutive measurements ≥ 30 days apart, plus: (1) chronic diarrhea (i.e., at least two loose stools a day for ≥ 30 days) or (2) documented fever (for ≥ 30 days, intermittent or constant) AIDS, acquired immunodeficiency syndrome; CT, computed tomography; HSV, herpes simplex virus; LIP, lymphoid interstitial pneumonia; MRI, magnetic resonance imaging. a Centers for Disease Control. Classification system for human immunodeficiency virus in children under 13 years of age. MMWR 1987;36:225. Adapted from Centers for Disease Control and Prevention. 1994 revised classification system for human immunodeficiency virus in children less than 13 years of age. MMWR 1994;43:1.

flattened columella, well-formed triangular philtrum, and patulous lips with prominent vermilion border,25,26 has been described but lacks specificity for HIV infection.27 Early manifestations of maternally derived HIV infection are frequently nonspecific.28 Lymphadenopathy, often associated with

hepatosplenomegaly, can be an early sign of infection. Oral candidiasis, parotitis, failure to thrive, dermatitis, and developmental delay are other common presenting features. P. jirovecii pneumonia accounts for about half of all AIDS-defining conditions diagnosed during the first year of life.29

658

SECTION

S Human Immunodeficiency Virus and the Acquired Immunodeficiency Syndrome

Hyperimmunoglobulinemia G, A, E, or M is present in up to 90% of HIV-infected infants by 6 months of age.30,31 By 1 or 2 months of age, HIV-infected infants have CD4+ lymphocyte counts that are significantly lower than those in HIV-exposed infants who are uninfected.32

Neurologic Depending on the definition employed and the rigor with which diagnostic studies are performed, CNS abnormalities can be found in most HIV-infected children.33–35 Progressive HIV encephalopathy accounts for about 15% of all pediatric AIDS-defining conditions reported to the CDC.36 Clinical features of HIV-associated progressive encephalopathy are shown in Table 111-4. Characteristic computed tomographic findings are cerebral atrophy (in about 85% of cases) and bilateral symmetric calcification of the basal ganglia (in about 15% of cases)33 (Figure 111-1). Results of routine cerebrospinal fluid (CSF) studies are usually normal; mild pleocytosis and elevated protein concentration are observed in some cases. A study from France suggests that the immature infant brain may be particularly sensitive to the effects of HIV.37 This study found the rate of encephalopathy in HIV-infected infants younger than 1 year of age to be almost 30-fold that of adults. The rate decreased to 15-fold the adult rate in the second year of life, and thereafter the rates of development of encephalopathy were similar in children and adults. HIV-exposed infants are not different from unexposed controls with respect to neurologic function.38 Encephalopathy is likely the result of direct HIV infection of the CNS rather than from complicating infection or malignancy. Evidence includes virus isolation from CSF,39 intrathecal synthesis of HIV antibody,39 and identification of HIV nucleotide sequences in brain tissue at autopsy.40,41 Neuropathologic features noted in the brains of HIV-infected children include atrophy, subcortical inflammatory lesions, multinucleated giant cells, and vascular calcification.41,42 Blood-derived macrophages, resident microglia, and their derivatives (including multinucleated giant cells) are the only cells that have consistently been shown to harbor HIV in the CNS.34,42 It is hypothesized that activation of these cells by HIV results in overproduction of certain cytokines, arachidonic acid metabolites, nitric oxide, and quinolinic acid, which in turn may be responsible for some of the neuropathologic changes observed.43–46 Elevated serum concentration of tumor necrosis factor (TNF) has been associated with progressive encephalopathy in HIV-infected children,44–46 and TNF can produce white-matter destruction in vitro similar to that observed clinically and pathologically in children with encephalopathy.47 Platelet-activating factor and products of arachidonic acid metabolism may also mediate CNS injury, possibly through upregulation of TNF or other cytokines.45 The rate of progression and severity of HIV encephalopathy are variable. Theoretically, the extent of penetration of antiretroviral agents into the CNS will affect outcome. Zidovudine attains therapeutic concentrations in CSF, and therapy with this agent results in stabilization or reversal of encephalopathy in some cases.48,49 CSF concentrations of stavudine are appreciable50; CSF concentrations of

didanosine are variable. However, treatment with all three nonnucleoside reverse transcriptase inhibitor agents (NRTIs) has been associated with improvement in neurodevelopmental testing.51 As a class, protease inhibitors are highly protein-bound and do not pass well into CSF. However, potentially effective concentrations of indinavir were found in CSF specimens in children in one study.52 Frequent evaluation of developmental milestones and neuropsychological testing are critical components in the care of HIV-infected children.

PULMONARY Lymphoid Interstitial Pneumonia or Pulmonary Lymphoid Hyperplasia LIP and the clinically similar pulmonary lymphoid hyperplasia (PLH) are second only to P. jirovecii pneumonia among pediatric AIDSdefining conditions, accounting for about 20% of cases.36 Although these conditions are assigned to clinical category B because of their more benign course than other AIDS-defining conditions, one retrospective study showed more than twice the rate of hospitalization due to serious respiratory infections in patients with LIP-PLH than in controls with HIV but no LIP-PLH.53 Although the radiographic features of LIP-PLH and P. jirovecii pneumonia can be similar, clinical features distinguish the two conditions (Table 111-5). The onset of LIP-PLH is usually insidious. Cough and tachypnea are often present. Examination of the chest generally reveals few auscultatory abnormalities. Marked generalized lymphadenopathy, hepatosplenomegaly, and salivary gland enlargement may be noted. Digital clubbing is observed in advanced cases. A chest radiograph typically reveals symmetric bilateral reticulonodular

TABLE 111-4. Features of Progressive Encephalopathy Associated with Human Immunodeficiency Virus Clinical Features

Neuroimaging Features

Impaired brain growth Developmental delay or regression Spastic weakness Pathologic reflexes Dystonia Gait disturbance Expressive language impairment

Cerebral atrophy Symmetric calcifications in basal ganglia Figure 111-1. Computed tomography scan of the brain showing bilateral symmetric calcification of the basal ganglia in a 2-year-old boy with vertically acquired human immunodeficiency virus infection.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Diagnosis and Clinical Manifestations of HIV Infection

CHAPTER

111

659

TABLE 111-5. Clinical Differentiation of LIP-PLH from PCP Pneumonia Finding

LIP-PLH

PCP

No No Late No Late Yes Yes

Yes Yes Early Yes Early No No

Yes No

No Yes

CLINICAL

Acute onset Fever Retractions Rales Hypoxia Clubbing Salivary enlargement BAL FLUID

Lymphocyte predominance Neutrophil predominance

BAL, bronchoalveolar lavage; LIP-PLH, lymphoid interstitial pneumonitis or pulmonary lymphoid hyperplasia; PCP, Pneumocystis pneumonia.

and interstitial infiltrates, sometimes in association with hilar adenopathy (Figure 111-2). Differentiation from miliary tuberculosis is sometimes difficult, although patients with LIP-PLH are more likely to be afebrile and relatively well-appearing; the chest radiograph shows less distinct (sometimes larger) lesions and more interstitial abnormality than in tuberculosis. There is no typical laboratory abnormality, but a marked increase in serum immunoglobulin concentrations is often present in patients with LIP-PLH. Presumptive diagnosis of LIP-PLH can be made on the basis of characteristic radiographic features persisting for 2 months or longer. Definitive diagnosis is made by open-lung biopsy. Histopathologic and immunocytochemical analyses reveal a mononuclear interstitial infiltrate composed of immunoblasts, plasma cells, and CD8+ lymphocytes. The pathogenesis of LIP-PLH is poorly understood, although Epstein–Barr virus has been implicated as a cofactor.54 The clinical course of LIP-PLH is variable. Spontaneous clinical remission is sometimes observed. Exacerbation of clinical signs and symptoms can occur in association with intercurrent viral respiratory illnesses. In severe cases, there is progressive hypoxia and respiratory failure. The management of children with LIP-PLH is largely supportive. Some patients require supplemental oxygen. Anecdotal reports suggest that some cases with progressive hypoxemia may respond to corticosteroid therapy.55 Children taking highly active antiretroviral treatment (HAART) exhibit marked clinical improvement in their disease.56

MALIGNANT Malignancies are uncommon among HIV-infected children, accounting for only about 2% of AIDS-defining illnesses.36 The most common cancer is non-Hodgkin lymphoma (Burkitt or immunoblastic types).57 Patients often manifest fever, weight loss, and evidence of “extranodal” disease (e.g., hepatomegaly and jaundice, abdominal distention, myelophthisis, or neurologic abnormalities). Children with CNS lymphoma can have delay or loss of developmental milestones, cranial nerve palsies, seizures, or hemiparesis.58 Contrast-enhanced computed tomography of the brain reveals hyperdense periventricular mass lesions similar to those seen in cerebral toxoplasmosis. However, CNS lymphoma more commonly causes an isolated brain mass in children than does toxoplasmosis or other infectious agensts.59 In addition to elevated protein concentration and hypoglycorrhachia, examination of CSF can reveal malignant cells. Brain biopsy for definitive diagnosis is indicated for those children who have rapidly progressive disease or clinical or radiographic progression despite empiric therapy for toxoplasmosis. The pathogenesis of AIDS-associated non-Hodgkin lymphoma is poorly defined; both Epstein–Barr virus60 and mutations or rearrangements in the c-myc oncogene61 have been implicated causally. Effective chemotherapeutic regimens are available. Leiomyosarcoma, leiomyoma, and Kaposi sarcoma also occur with

Figure 111-2. Chest radiograph demonstrating bilateral reticulonodular interstitial infiltrates of lymphoid interstitial pneumonitis/pulmonary lymphoid hyperplasia in a 4-year-old boy with vertically acquired human immunodeficiency virus infection.

greater frequency in children with HIV infection than in the general pediatric population.57 A strong association exists between Epstein– Barr virus and leiomyosarcoma or leiomyoma. Endobronchial leiomyosarcoma or leiomyoma can manifest with fever or cyanosis.62 Multiple pulmonary parenchymal nodules can be visualized radiographically. Clinical manifestations of pulmonary Kaposi sarcoma include fever, cough, dyspnea, and hemoptysis. The radiographic features are diffuse reticulonodular infiltrates resembling those seen in LIP-PLH or P. jirovecii pneumonia. Kaposi sarcoma can also present with involvement of lymph nodes, the oral cavity (hard palate or tonsils), or skin. Gastrointestinal lesions can occur with any of these malignancies, resulting in gastrointestinal bleeding, abdominal pain, or bowel obstruction.

GASTROINTESTINAL Hepatitis Hepatomegaly, moderate increases in serum concentrations of hepatic enzymes (e.g., 5- to 10-fold increases in aspartate transaminase (AST) and/or alanine transaminase (ALT)), are common in HIV-infected children. Laboratory evidence of hepatitis is often unaccompanied by clinical manifestations of liver disease. The differential diagnosis of hepatitis is broad and includes a variety of infectious causes, neoplasms, and effects of medications (Box 111-2).63 Coinfection with HIV and either hepatitis B or C can hasten the course of both diseases, increasing the risk for liver failure.64 Detection of HIV RNA and gp41 antigen in hepatocytes, and the absence of evidence for other causes, suggests that HIV itself is the cause of some cases of hepatitis.63 The clinical course of HIV-associated hepatitis is variable. Histopathologic studies of liver tissue obtained from HIV-infected children have identified giant-cell transformation of hepatocytes, fatty degeneration, and lymphoplasmocytic infiltration as prominent features.65,66 Elevation of transaminase values has been reported in association with the use of all antiretroviral medications, but increases are especially common with NRTIs and protease inhibitor agents. These elevations are often clinically insignificant in children.

Wasting Syndrome HIV infection was originally referred to as “slim disease” in African culture. Weight loss, or failure to gain appropriate weight, is a

660

SECTION

S Human Immunodeficiency Virus and the Acquired Immunodeficiency Syndrome

BOX 111-2. Differential Diagnosis of Hepatitis in Children Infected with Human Immunodeficiency Virus (HIV) Infection • Bacteria (e.g., Mycobacterium avium-intracellulare complex, Mycobacterium tuberculosis, or Bartonella spp.-induced bacillary peliosis hepatis) • Virus (e.g., hepatitis A, B, or C; cytomegalovirus; Epstein–Barr virus; herpes simplex virus; adenovirus; or HIV) • Fungus (e.g., Candida spp., Cryptococcus neoformans, Histoplasma capsulatum, Pneumocystis jirovecii) • Protozoa (e.g., Toxoplasma gondii) • Malignancy Lymphoma Sarcoma • Drug-induced Sulfonamides Antiretroviral agents Antimycobacterial medications Ketoconazole Pentamidine

common presenting symptom in HIV-infected infants and children. Several factors can contribute to wasting. HIV infection itself increases a child’s nutritional requirements. In addition, opportunistic infections, especially those associated with fever, contribute to increased metabolic requirements. Oral lesions can inhibit the intake of food. Chronic diarrhea and malabsorptive syndromes common in HIV can inhibit caloric intake. Treatment for wasting includes HAART and nutritional support.

CARDIOVASCULAR Cardiac abnormalities are recognized with growing frequency in HIVinfected children. Electrocardiographic abnormalities include conduction defects, voltage abnormalities compatible with chamber enlargement, and dysrhythmias; echocardiographic findings include ventricular dysfunction and pericardial effusion.67 Many abnormalities are not clinically apparent; however, in one retrospective study of 81 HIV-infected children, serious dysrhythmia, congestive heart failure, and unexpected cardiac arrest occurred in 28 (35%), 8 (10%), and 7 (9%) patients, respectively.67 The pathogenesis of cardiac disease in HIV-infected children is poorly understood, but epidemiologic associations between coinfection with Epstein–Barr virus and either bradycardia or congestive heart failure have been reported.68 Adenovirus and cytomegalovirus sequences have been identified in a postmortem study of cardiac tissue from HIV-infected children with myocarditis.69 HIV-infected children who had been treated with zidovudine have been reported to have an 8.4-fold higher risk of cardiomyopathy.70 The relative importance of HIV-related and drugrelated cardiac dysfunction, as well as value and timing of echocardiographic screening, is evolving with changing antiretroviral therapies.71,72 Many HIV-infected children with congestive heart failure demonstrate good response to medical management.

RENAL Nephropathy has been described predominantly among children with advanced HIV disease.73–75 Persistent proteinuria is a common finding. In one study, progressive renal disease with nephrotic syndrome and renal failure developed in 5 of 12 children with AIDS and proteinuria.73 In a larger series of 556 HIV-infected children, 72 (12.9%) met the definition of HIV nephropathy. Nineteen percent of those with nephropathy progressed to chronic renal insufficiency.76 Histopathologic examination of the kidney has revealed a variety of lesions, including focal glomerulosclerosis and mesangial hypercellularity. Immune complex deposition may be involved in the

pathogenesis of HIV-associated nephropathy.74,75 Most cases respond poorly to corticosteroid therapy,72–74 but cyclosporine-induced remission has been reported.75,76

HEMATOLOGIC Anemia, leukopenia, neutropenia, and thrombocytopenia are common manifestations of HIV infection in children. Many cases of normocytic or microcytic anemia are attributable to chronic disease; macrocytic anemia is most often secondary to zidovudine therapy. Other causes of anemia are iron deficiency, hemoglobinopathies, and red blood cell enzyme defects. Chronic human parvovirus infection is an important cause of anemia in HIV-infected children.77 The treatment of anemia in HIV depends on the etiology. Therapy with recombinant human erythropoietin can allow continued use of zidovudine in patients with otherwise dose-limiting anemia.78 Some patients with chronic parvovirus infection respond to high-dose immune globulin intravenous (IGIV) therapy.77 Leukopenia and neutropenia often result from medication-induced bone marrow depression. Zidovudine, trimethoprim-sulfamethoxazole (TMP-SMX), and ganciclovir commonly are implicated agents. Neutropenia also is observed during the course of various opportunistic infections. Granulocyte colony-stimulating factor therapy can be helpful in reversing drug- or disease-induced neutropenia and preventing infectious complications.79 Thrombocytopenia in HIV-infected children can occur as a result of either underproduction or shortened survival of platelets. Immunemediated destruction is a common cause of thrombocytopenia, and antiplatelet antibodies are detectable in most patients. Treatment of HIV-associated thrombocytopenia is difficult. Zidovudine produces a temporary rise in platelet counts in most cases. Some children exhibit an initial response to corticosteroid or immune globulin intravenous therapy, but improvement is rarely sustained. Splenectomy may be indicated for the HIV-infected child with severe thrombocytopenia that is refractory to other measures.80 Rituximab has also been used in chronic, recurrent HIV-associated immune thrombocytopenia.81

DERMATOLOGIC A wide variety of bacterial, fungal, viral, and parasitic infections of the skin occur in children with HIV infection. However, HIV-infected children also have a higher than expected incidence of noninfectious inflammatory, eczematoid, psoriatic, neoplastic, and drug hypersensitivity skin conditions.82 Seborrheic dermatitis manifests as erythema and scaling of the face and scalp in HIV-infected infants or of the nasolabial folds, eyebrows, and postauricular areas in older children. Pruritic papular eruption (PPE) is a chronic eruption of papular lesions on the skin: its etiology is unclear. It may represent an overexuberant reaction to insect bites and has been associated with increased IgE.83 PPE can cause significant discomfort for affected children. Cutaneous eruptions occur with a variety of oral and parenteral medications, trimethoprim-sulfamethoxazole being the most commonly implicated agent. Rash usually begins 7 to 10 days after initiation of the drug. Characteristically, it is maculopapular or morbilliform in appearance and disappears promptly after discontinuation of the drug. Cutaneous Kaposi sarcoma occurs in HIVinfected children internationally, but is rarely seen in the United States.

PROGNOSIS In general, children with vertically acquired HIV infection have shorter clinical latent periods and more rapid disease progression than other individuals with HIV infection.28,84 However, there is bimodal disease expression, with late onset of symptoms and long-term survival in many cases.84,85 Both age at diagnosis of an

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Infectious Complications of HIV Infection

AIDS-defining condition and clinical presentation are important determinants of prognosis. In one study before the advent of highly effective therapy, the overall median survival time for children with vertically acquired HIV infection was 30 months, but children who demonstrated AIDS in the first year of life had a median survival time of only about 7 months.28 Presentation during infancy with opportunistic infection (especially P. jirovecii pneumonia) or HIV encephalopathy28,33,84 portends a particularly poor prognosis, whereas slow decline of CD4+ lymphocyte count, late onset of signs and symptoms of HIV infection, and occurrence of LIP-PLH are associated with a longer survival.85 The importance of weight growth velocity as a prognostic indicator for HIV-infected children has been highlighted in several studies.86,87 In one study of HIV-infected children receiving zidovudine therapy, 9 of 28 (32%) children who gained weight at a rate lower than the 10th percentile during the first 6 months of therapy died within 24 months, whereas only 10 of 75 (13%) children with more normal growth velocity died during the same period of follow-up.86 The outlook for children with HIV continues to improve. Combination HAART has become the standard of care. The use of HAART therapy can often stop or even reverse many of the clinical manifestations of HIV. Although adherence to medication regimen and side effects of medications remain problems, the number of pediatric deaths due to HIV/AIDS in the United States has declined dramatically over the past several years. Hospitalizations have also decreased as a result of effective prophylaxis against opportunistic infections and potent antiretroviral medications.

CHAPTER

112

Infectious Complications of HIV Infection Mark W. Kline

Opportunistic infections are central to the definition of the acquired immunodeficiency syndrome (AIDS) and are major causes of AIDSassociated morbidity and mortality. An unusual clustering of cases of Pneumocystis jirovecii (also known as PneumoCystis) pneumonia (PCP) led in part to recognition of AIDS in the early 1980s. Other AIDSassociated opportunistic infections, including disseminated mycobacterial disease, cryptococcal meningitis, cerebral toxoplasmosis, and cytomegalovirus (CMV) retinitis were recognized soon thereafter. In the mid-1990s, improved treatment strategies and the introduction of highly active antiretroviral therapy (HAART) dramatically reduced the incidence of opportunistic infections. Yet, despite these advances, opportunistic infections continue to occur, and physicians must be familiar with the epidemiology and features of these infections to effect early treatment and prophylaxis for those at risk.

PATHOGENESIS The naivety of the developing immune system is invoked as a possible explanation for susceptibility of the infant or child infected with the human immunodeficiency virus (HIV) to rapid disease progression and complicating infections. The normal neonate is at risk of serious infection because of immaturity of several components of the immune system, including B and T lymphocytes, phagocytes, and complement. As a consequence, defenses against primary HIV infection and other opportunistic and nonopportunistic infections, as well as immune control of progression of HIV disease, are probably impaired.

CHAPTER

112

661

HIV infection of the fetus or neonate can have profound effects on cellular immunity. Destruction of the thymus gland is observed in spontaneous abortuses from HIV-infected mothers. As early as 1 or 2 months of age, HIV-infected infants have CD4+ lymphocyte counts significantly lower than those of HIV-exposed infants who escape infection.1 Suppression of cell-mediated immunity is responsible in large part for the susceptibility of HIV-infected individuals to PCP and opportunistic mycobacterial, fungal, and viral infections. Despite the presence of hypergammaglobulinemia, humoral immune dysfunction can be demonstrated in most children with symptomatic HIV infection. Humoral deficiency results either as a primary effect of HIV or as a consequence of dysregulation of T-lymphocyte-mediated responses and polyclonal stimulation of B lymphocytes by poorly controlled infectious agents.2–6 In vitro lymphoproliferative responses to B-lymphocyte mitogens and specific antigens are often poor. In vivo, there is impaired specific antibody production after immunization with either T-lymphocyte-independent antigens (e.g., capsular polysaccharide of Streptococcus pneumoniae), or T-lymphocyte-dependent antigens (e.g., bacteriophage OX174). Both primary humoral immune responses and recall responses (e.g., amplification and immunoglobulin class switch) are defective in children with HIV infection. Phagocytic function can also be altered by HIV infection. Neutropenia and defects in neutrophil chemotaxis and bactericidal activity have been described.7,8 Neutrophil superoxide production can be depressed in children with advanced HIV disease.9 This combination of phagocyte abnormalities compromises the ability of an HIVinfected individual to kill bacterial and fungal pathogens. Although the precise pathogenesis of these abnormalities is unknown, studies suggest that neutropenia and neutrophil dysfunction are at least partly mediated by abnormal regulation by cytokines (e.g., granulocyte colony-stimulating factor (G-CSF)).10 Recombinant G-CSF administration in HIV-infected adults has been shown to reverse neutropenia and correct neutrophil-killing defects.11,12

EPIDEMIOLOGY AND ETIOLOGIC AGENTS In children, as in adults, untreated HIV infection and AIDS are characterized by a wide variety and increased frequency of serious opportunistic and nonopportunistic infections. Also apparent is an increased frequency of common infections of a less serious nature, which contribute to the overall morbidity of the disease. The frequency of the various opportunistic infections among HIV-infected children in the years preceding HAART varied by age, pathogen, and immunologic status.13 The most common opportunistic infections among United States children included serious bacterial infections (SBIs, e.g., pneumonia and bacteremia), PCP, nontuberculous mycobacterial infections, and CMV disease (Table 112-1).14 Substantially reduced rates of mortality and morbidity, including opportunistic infections, have been observed among HIV-infected children receiving HAART.15 Despite this decrease in the rate of opportunistic infections, the types of infections observed have not changed.16 Before the advent of HAART and widespread use of P. jirovecii prophylaxis in early infancy, PCP was the most common AIDSassociated serious opportunistic infection reported to the Centers for Disease Control and Prevention (CDC) (Table 112-2).17 More than 50% of reported cases occurred between 3 and 6 months of age.18 The disease accounted for about 60% of all AIDS-defining illness occurring during the first year of life but only about 19% occurring subsequently. Between 7% and 20% of all HIV-infected infants manifested PCP. A low age-adjusted CD4+ lymphocyte count or percentage was the major determinant of risk. Although PCP is still observed in HIV-infected infants and children, the incidence has dramatically decreased in the United States. Esophageal and pulmonary candidiasis account for approximately 10% and 2%, respectively, of all AIDS-defining illnesses in children.19 Whereas oral candidiasis is common at all stages of HIV disease, esophageal and pulmonary candidiasis are far less common, and their occurrence is typically restricted to patients with advanced HIV disease.

662

SECTION

S Human Immunodeficiency Virus and the Acquired Immunodeficiency Syndrome

TABLE 112-1. Incidence of Infections in Human Immunodeficiency Virus (HIV)-Infected Children Aged 12 Years or Younger Before the Advent of Highly Active Antiretroviral Therapy (HAART) Condition

Cases/100 Patient-Years

OPPORTUNISTIC INFECTIONS

Pneumocystis jirovecii pneumonia Nontuberculous mycobacterial infection Cytomegalovirus disease Cryptosporidiosis

6.4 2.6 2.4 1.1

OTHER INFECTIONS

Otitis media Upper respiratory infection Sinusitis Bacterial pneumonia Bacteremia Herpes zoster Varicella Urinary tract infection Meningitis Tuberculosis

57.0 47.0 15.0 11.0 8.2 5.0 4.7 3.7 1.1 1.0

Data from Van Dyke RB. Opportunistic infections in pediatric HIV disease. Ann NY Acad Sci 1993;February:1.

TABLE 112-2. Common Acquired Immunodeficiency Syndrome (AIDS)-Defining Infections in Children AIDS-Defining Infection

AIDS-Defining Events (%)

Pneumocystis jirovecii pneumonia Recurrent bacterial infections Esophageal or pulmonary candidiasis Cytomegalovirus disease Mycobacterium avium complex infection Herpes simplex disease Cryptosporidiosis

25 18 12 8 7 3 2

Data from Centers for Disease Control and Prevention. HIV-AIDS surveillance report. Atlanta, Centers for Disease Control and Prevention, 1997;9(No. 2):1.

Recognition of the frequently serious and repetitive nature of bacterial infections in HIV-infected children led to the designation of certain of these as indicator diseases for AIDS in children. A child younger than 13 years with laboratory evidence of HIV infection is defined as having AIDS if any combination of the following bacterial infections occurs twice within 2 years: septicemia, pneumonia, meningitis, osteomyelitis or pyogenic arthritis, or abscess of an internal organ or body cavity. Recurrent bacterial infections accounted for 18% of all pediatric AIDS-defining illnesses reported to the CDC in 1997.19 In a review of the medical histories of 46 children with symptomatic HIV infection, 26 (57%) had at least one episode of SBI, 80% of which included at least one episode of bloodstream infection (BSI).20 In another study, 27 (38%) of 71 HIV-infected children were reported to have experienced 125 episodes of SBI over 3.5 years of observation.21 In 35 of these 125 episodes, BSI and focal infection occurred concomitantly, and in another 17, bacteremia occurred without an apparent focus of infection. A variety of SBIs was observed in these studies, including pneumonia, urinary tract infection, meningitis, cellulitis, and lymphadenitis. In a multicenter placebo-controlled trial of intravenous immunoglobulin (IVIG), 376 HIV-infected children underwent 545 patientyears of follow-up.22 Laboratory-proven (i.e., confirmed by bacterial culture or antigen assay) or clinically diagnosed (i.e., without a defined bacterial cause) SBI occurred in 141 (38%) of the children. Clinically diagnosed infections occurred twice as often as laboratoryproven infections, and minor infections (e.g., otitis media, urinary tract infection, or infections of skin or soft tissue) were three times

more common than serious infections (e.g., bacteremia, certain types of pneumonia or sinusitis, meningitis, osteomyelitis, septic arthritis, mastoiditis, or abscess of an internal organ). The most frequent laboratory-proven serious infections were BSI and pneumonia; pneumonia and sinusitis were the most common clinically diagnosed serious infections. The overall incidence of SBIs was 44 per 100 patient-years of follow-up. Risk factors for SBI in HIV-infected children have not been defined precisely, but studies have demonstrated that the incidence is highest in vertically infected children less than 1 year old who have low CD4+ lymphocyte counts.23 The Pediatric Spectrum of Disease Project evaluated 21167 vertically infected children for SBI; 570 children had 1063 infections. Sixty-four percent were severely immunocompromised (category 3) when the first SBI occurred. Twenty-five percent of all children with SBIs had their first infection in the first 6 months of life, and 37% of infections occurred by 12 months of life. BSI, pneumonia, and urinary tract infection were the most common BSI.24 Streptococcus pneumoniae is the single most common cause of SBI in HIV-infected children.20–22,24,25 In a study of HIV-infected children younger than 36 months of age, pneumococcal BSI occurred with an annualized rate of 11.3 episodes per 100 patient-years26; this rate is more than three times greater than that observed in children with sickle-cell anemia27 and 12 times that reported in one study of adults with AIDS.28 Other AIDS-defining infections are recognized substantially less frequently than PCP, esophageal or pulmonary candidiasis, or recurrent bacterial infections. CMV infection is common, but disseminated disease, chorioretinitis, and colitis caused by CMV are uncommon.29 The role of CMV as a cause of pneumonia in HIVinfected children is controversial; it is usually found in association with another pathogen (often P. jirovecii). Herpes simplex virus commonly causes infection of the oral mucous membranes30; esophageal or pulmonary disease occurs rarely. CMV disease, cryptosporidiosis,31 and disseminated Mycobacterium aviumintracellulare complex (MAC) disease32,33 all occur predominantly among children with advanced HIV disease and severely depressed CD4+ lymphocyte counts. Up to 20% of HIV-infected children with CD4+ lymphocyte counts of < 50 cells/mm3 have disseminated infection with MAC; nearly all children with this infection have CD4+ lymphocyte counts of < 100 cells/mm3.

APPROACH TO THE HIV-INFECTED CHILD WITH SUSPECTED SYSTEMIC OR FOCAL INFECTION Fever Clinical manifestations and diagnostic considerations for HIV-infected children with fever are diverse. Febrile episodes can be acute or prolonged. Some acutely febrile HIV-infected children have fever alone, or fever without localizing signs, whereas others have focal infections. Similarly, prolonged fever (arbitrarily defined as more than 7 days) may be associated with a discernible cause, or a cause may not be evident after careful physical examination and initial laboratory testing (i.e., fever of unknown origin). Fever is a common reason for unscheduled outpatient clinic visits and hospital admission of HIV-infected children. At Texas Children’s Hospital, 26 (11%) of 231 hospitalizations of HIV-infected children were because of fever without localizing signs, either acutely or with a duration of more than 7 days. Fifty (31%) of 161 HIV-infected children had a total of 77 episodes of unexplained fever lasting 1 month or longer. Most HIV-infected children with acute febrile illnesses have mild, self-limited illnesses that appear to be caused by viral infection. However, some HIV-infected children with acute onset of fever have bacteremia or other serious complicating medical illnesses. Differentiation can be difficult (Box 112-1). Febrile HIV-infected children should be initially evaluated by means of a thorough medical

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Infectious Complications of HIV Infection

BOX 112-1. Causes of Fever of Unknown Origin in Human Immunodeficiency Virus (HIV)-Infected Children Focal bacterial infection (e.g., sinusitis, pneumonia, or internal abscess) Salmonellosis Tuberculous or nontuberculous mycobacterial infection Fungal infection (e.g., candidal esophagitis, cryptococcal meningitis, or pneumonia) Pneumocystis jirovecii infection Toxoplasmosis Cytomegalovirus infection Epstein–Barr virus infection Herpes simplex virus infection Hepatitis Lymphoma and other types of malignancy Drug fever

history and physical examination. The character and duration of symptoms, HIV disease status, and history of recent exposures to illness in others are particularly relevant. A careful search for evidence of focal infection or inflammation and a general assessment of severity of illness or “toxicity” should be performed. Clinical features of focal infection in HIV-infected children are similar to features observed in immunologically normal children. Fever and local signs of inflammation are often present. BSI should be suspected when certain serious focal infections (e.g., pneumonia, cellulitis, or osteomyelitis) are present or in the child who appears to be ill. Bacterial cultures of blood should be obtained from all such children. The approach to the acutely febrile HIV-infected child without localizing signs is more problematic. Not all patients require diagnostic testing or antibiotic therapy, although both should be considered, especially in those with high-grade fever (higher than 39.4°C) or advanced HIV disease. The white blood cell count, often used as a screening test for SBI in immunologically normal children, must be interpreted in the context of the patient’s baseline values. Several other diagnostic studies are considered individually and include blood culture, chest radiograph, lumbar puncture, urinalysis, and urine culture. Cultures for mycobacteria and fungi seldom have positive results (or are indicated) in the setting of acute fever without localizing signs. An individual approach to the therapy of acutely febrile, HIVinfected children with localizing signs is appropriate. The selection of antibiotic agents for expectant therapy is influenced by several factors, including the likely source of infection (e.g., community-acquired versus hospital-acquired infection) and the child’s HIV disease status. Expectant antibiotic therapy, when used, should be directed against a limited number of likely pathogens. Unless convincing clinical or laboratory evidence exists for serious opportunistic fungal or viral infection, therapy is usually only directed against potential bacterial pathogens. In general, antibiotic therapy for asymptomatic or mildly symptomatic HIV-infected children with fever should be similar to that used in the treatment of immunologically normal children with comparable clinical manifestations. Oral antibiotic therapy is appropriate in most cases. Children with more advanced HIV disease may require agents with broader activity, administered parenterally. Children with granulocytopenia resulting from HIV infection or from various medications should be given antibiotics similar to those given to children with granulocytopenia caused by other conditions (e.g., acute leukemia or chemotherapy). Evaluation of fever of unknown origin in children infected with HIV often requires extensive diagnostic testing because of the myriad potential infectious and noninfectious causes (see Box 112-1). Such studies may include radiographs of the chest and sinuses; culture of blood for routine bacteria, mycobacteria, and fungi; culture of the throat, nasopharynx, urine, and blood for viruses (especially CMV); cryptococcal and other fungal antigen tests; Epstein–Barr virus serologic tests; hepatic enzyme measurements; and ophthalmologic

CHAPTER

112

663

examination for chorioretinitis. Cerebrospinal fluid (CSF) should be obtained for routine cell count; protein and glucose concentration; bacterial, fungal, and viral polymerase chain reaction (PCR) assays; cryptococcal antigen test; and cytologic tests. Bone marrow aspiration and biopsy are occasionally useful, particularly if hematologic abnormalities exist or if routine cultures and serologic test results are negative. Specimens should be obtained for Gram stain and bacterial culture, fungal stains and culture, stain for acid-fast bacteria, culture for mycobacteria and virus, histopathology, and cytology. The diagnostic evaluation for fever of unknown origin is determined by the pace of the illness and the patient’s HIV disease status. In the case of a child with less advanced HIV disease who appears to be mildly ill, diagnostic studies are conducted in a stepwise manner with common causes of disease excluded first. A patient with more advanced HIV disease or fulminant illness may require immediate and extensive testing to address the full range of diagnostic possibilities. Studies attempting to detect opportunistic conditions, such as disseminated infection with MAC or Cryptococcus neoformans or CMV retinitis, are more likely to have positive results if the child has advanced HIV disease with a markedly depressed CD4+ lymphocyte count. As in immunologically normal children, expectant antibiotic therapy is rarely indicated in HIV-infected children with fever of unknown origin.

Pneumonia More than 50% of all infants with vertically acquired HIV infection initially manifest signs and symptoms of a pulmonary disorder.34 Many infectious and noninfectious pulmonary complications of HIV infection are recognized, but PCP, lymphocytic interstitial pneumonitis/pulmonary lymphoid hyperplasia (LIP/PLH), and bacterial pneumonia are the most common. Predominant bacterial causes of pneumonia include Streptococcus pneumoniae, Haemophilus influenzae type b, Klebsiella pneumoniae, Pseudomonas aeruginosa, and Mycobacterium tuberculosis. Other infectious causes of pneumonia in HIV-infected children include common viruses (e.g., respiratory syncytial virus, parainfluenza virus, influenza virus, and adenovirus), CMV, herpes simplex virus, and fungi (e.g., Candida, Cryptococcus neoformans, and Histoplasma capsulatum). Diagnostic evaluation of the HIV-infected child with pneumonia is individualized on the basis of HIV disease status and clinical presentation. The likelihood of an unusual or opportunistic infectious agent increases as the patient’s CD4+ lymphocyte count decreases. The usual manifestations of PCP include acute onset of fever, cough, tachypnea, and respiratory distress. Decreased breath sounds or rales may be present. The onset of LIP/PLH is usually insidious and the course is slowly progressive. Cough and wheezing may be present. Associated findings include generalized lymphadenopathy, hepatosplenomegaly, parotid enlargement, and digital clubbing. The clinical presentation of bacterial pneumonia in children with HIV infection is similar to that seen in immunologically normal children. Radiographic features of pneumonia in HIV-infected children are generally nonspecific. The typical radiographic feature of PCP is diffuse interstitial and alveolar infiltrate, most prominent in the perihilar areas (Figure 112-1). Air bronchograms, focal infiltrates, pneumatoceles, or pneumothorax are seen in some cases. The chest radiograph can be entirely normal at the time of presentation. The typical radiographic features of LIP/PLH include diffuse reticulonodular infiltration and hilar adenopathy (Figure 112-2). Bacterial pneumonia may be associated with focal or diffuse infiltration. A Mantoux tuberculin skin test (purified protein derivative, PPD) should be done in every case of pneumonia in a child with HIV infection. The use of anergy skin testing (e.g., using mumps, candidal antigens) is no longer recommended for screening for Mycobacterium tuberculosis infection among HIV-infected individuals; a negative PPD does not exclude M. tuberculosis infection regardless of results of anergy testing.35 A complete blood cell count and routine blood cultures are obtained in most cases. Measurement of the serum lactate dehydrogenase concentration can be helpful when PCP is suspected.

664

SECTION

S Human Immunodeficiency Virus and the Acquired Immunodeficiency Syndrome

Figure 112-1. Chest radiograph showing Pneumocystis jirovecii pneumonia in a 2-year-old boy with vertically acquired human immunodeficiency virus infection. Note bilateral interstitial lung disease.

Figure 112-2. Typical lymphocytic interstitial pneumonitis/pulmonary lymphoid hyperplasia (LIP/PLH) as the presenting manifestation of vertically acquired HIV infection in a 23-month-old infant with slowly worsening “bronchiolitis,” and hypoxia. Fungal, bacterial, and other viral causes were excluded. Plain radiograph shows diffuse reticulonodular infiltration and perihilar adenopathy. There was complete resolution 4 months after commencement of highly active antiretroviral therapy. (Courtesy of E.N. Faerber, J. Chen, and J. Foster, St. Christopher’s Hospital for Children, Philadelphia, PA.)

A nasopharyngeal specimen for rapid diagnostic tests (e.g., fluorescent antibody test or enzyme immunoassay) and culture for viruses can also be helpful. Bronchoalveolar lavage is indicated for confirmation of suspected PCP and exclusion of other opportunistic infections in children with advanced HIV disease. If a high index of suspicion exists for tuberculosis in a child with a pulmonary infiltrate and negative results on PPD skin test, gastric or bronchoalveolar lavage is warranted. Atypical features of pneumonia may necessitate open-lung biopsy to evaluate the wide array of diagnostic possibilities. Most children with normal or unchanged results on chest radiograph and normal arterial oxygen saturation have self-limited (presumably viral) illness. P. jirovecii is the most likely etiology in a

child with acute onset of a diffuse reticulonodular infiltrate associated with hypoxemia. Marked elevation of the serum lactate dehydrogenase concentration (> 1000 IU/L) also suggests PCP. Because cutaneous anergy can be present in children with advanced HIV disease, a negative PPD skin test result (defined in the HIV-infected child as an induration of 5 mm or less) does not exclude tuberculosis. A presumptive diagnosis of LIP/PLH can be made on the basis of characteristic clinical features and radiographic changes persisting for longer than 2 months in the absence of another explanation. Atypical features may necessitate open-lung biopsy for definitive diagnosis. Positive results of blood culture or evidence of a segmental or lobar infiltrate suggests bacterial pneumonia. Empiric antibiotic therapy is indicated for HIV-infected children with characteristic features of bacterial pneumonia. A second- or third-generation cephalosporin is appropriate for children with relatively well-preserved CD4+ lymphocyte counts, whereas combination therapy with a broad-spectrum b-lactam antibiotic plus an aminoglycoside is indicated for children with advanced disease and profoundly depressed CD4+ lymphocyte counts. Caution is warranted regarding repeated courses of ceftriaxone because rapidly fatal immune-mediated hemolysis has been described in HIV-infected children.36

Central Nervous System Infections Except for children with advanced HIV disease, the causative agents, clinical manifestations, and CSF findings of meningitis and encephalitis are similar to those observed in immunocompetent hosts. Cryptococcus neoformans and JC virus are two unusual pathogens that can cause central nervous system (CNS) infection in HIV-infected children. C. neoformans is a common cause of meningitis in patients with advanced HIV disease and severely depressed CD4+ lymphocyte counts. Cryptococcal meningitis is often characterized by an indolent course of illness. Complaints of fever, intermittent headache, and vomiting are common. Signs of meningeal inflammation can be absent. CSF complete cell count and glucose and protein levels can be normal. Important CSF diagnostic studies include India ink stain, cryptococcal antigen testing, and culture. JC virus is a papovavirus that causes progressive multifocal leukoencephalopathy (PML). PML is a progressive degenerative disease of the CNS, characterized by focal neurologic abnormalities (e.g., visual deficits and motor weakness), personality changes, and confusion. Neuroimaging reveals white-matter disease in the absence of mass effect (Figure 112-3). The incidence of PML in HIV-infected children is unknown; 4% or 5% of adults with AIDS may be affected. Most cases of PML occur in individuals with advanced HIV disease. The usual clinical course is inexorable deterioration of CNS function, leading to death within a few months of diagnosis; however, in some cases remission has occurred spontaneously or in temporal association with treatment with cytosine arabinoside, cidofovir, or HAART. CNS mass lesions in children with HIV infection pose special diagnostic considerations. Definitive diagnosis usually requires examination of tissue. Infectious causes include bacterial brain abscess, toxoplasmosis, cryptococcoma, and tuberculoma; noninfectious causes are also possible. Cerebral toxoplasmosis occurs relatively frequently in patients with advanced HIV disease and severely depressed CD4+ Tlymphocyte counts. Presenting features include fever, headache, seizures, focal neurologic abnormalities, and altered consciousness. Computed tomographic scan or magnetic resonance imaging of the brain typically reveals multiple ring-enhancing lesions with surrounding edema. Toxoplasmosis almost invariably represents reactivated infection in HIV-infected individuals. Therefore, the usefulness of serologic testing for Toxoplasma is restricted to patients who are found to be seronegative and therefore are unlikely to have cerebral toxoplasmosis. Brain biopsy, the definitive diagnostic procedure, is usually reserved for those patients who fail to respond clinically and radiographically to empirical therapy for toxoplasmosis.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Infectious Complications of HIV Infection

A

CHAPTER

112

665

C

B

Figure 112-3. A 12-year-old boy, recently immigrated from Zambia, and newly diagnosed with acquired immunodeficiency syndrome (AIDS) had a 2- to 3-year history of severe headache and diarrhea. Magnetic resonance imaging showed predominantly right-sided white-matter lesions. Antimicrobial and highly active antiretroviral therapy (HAART) were begun; 9 weeks after starting HAART, human immunodeficiency virus (HIV) viral load was low but he was subdued, tearful, and had decreased strength and reflexes on the left. Imaging showed progression of previous abnormalities. Cerebrospinal fluid (CSF) was normal; CSF polymerase chain reaction was positive for JC virus. Cidofovir intravenously plus probenecid orally were begun; 2 months later his strength was 4/5 and he was back to school and walking independently (but with poor balance). He has residual left-sided pronator drift, ankle clonus, increased reflexes, and upgoing toe. Imaging was unchanged from commencement of therapy. Note fluid-attenuated inversion recovery (FLAIR) (A, B) and T2-weighted (C) magnetic resonance images showing increased signal within the periventricular white-matter tracts extending outward to the subcortical areas – involving parietal lobes, both hemispheres, the corpus callosum and brainstem – with no mass effect. (Courtesy of A.R. Feingold and D. Meislich, Cooper University Hospital, Camden, NJ.)

In addition to a variety of bacterial and fungal processes, the differential diagnosis of CNS toxoplasmosis includes CNS lymphoma.

Diarrhea Many episodes of diarrhea in HIV-infected children are acute and selflimited, with no apparent long-term effects. By contrast, persistent or recurrent diarrhea can adversely affect quality of life and place the child at risk of malnutrition and further immunologic impairment. There are two common syndromes of infectious diarrhea. Enteritis is caused by pathogens that affect the small bowel. Some of these organisms produce toxins that stimulate secretion of fluid and electrolytes; others selectively damage the absorptive cells of the small bowel. High-volume watery diarrhea is characteristically observed in patients with enteritis. By contrast, ileal and colonic pathogens are typically invasive, producing ileocolitis, which results in a dysenterylike illness characterized by the presence of fever, abdominal cramping, and frequent stools of small volume. Mucus and blood are often present in stool. Enteritis can be produced by a variety of common pathogens, including rotavirus, enteric adenovirus, enterotoxigenic Escherichia coli, Vibrio cholerae, and Giardia lamblia. Opportunistic pathogens, such as CMV, MAC, Cryptosporidium, and Isospora spp., are frequently implicated in children with advanced HIV disease. The causative agents of ileocolitis include virulent bacterial pathogens, such as Salmonella, Shigella, Campylobacter, Clostridium difficile, and Entamoeba histolytica, and opportunistic agents such as CMV. Data on the frequency of various diarrheal pathogens in HIVinfected children are not available. Common enteric pathogens (e.g., rotavirus or Salmonella species) can produce disease at any stage of HIV infection, whereas opportunistic disease (e.g., CMV ileocolitis) is generally only observed in patients with advanced HIV infection with severely depressed CD4+ lymphocyte counts. Therefore, to some extent, the diagnostic evaluation should be guided by the patient’s HIV

disease status, as well as the clinical syndrome (i.e., enteritis versus ileocolitis). Many episodes of acute enteritis do not require diagnostic evaluation. However, if gastrointestinal symptoms are severe or persistent, a number of investigations should be considered. An antigen test for the presence of rotavirus may be useful. Examination of the stool for ova and parasites by routine methods as well as by an acid-fast staining procedure (for detection of Cryptosporidium and Isospora spp.) should be considered. Stool cultures for bacterial enteric pathogens are obtained if the child has fever or stools containing blood or mucus. Blood culture is also obtained to exclude Salmonella or Shigella bacteremia. Studies for Clostridium difficile should be obtained if there is a history of recent antibiotic use. Stool cultures for CMV and MAC are not generally useful. Upper endoscopy is reserved for the HIV-infected child with persistent diarrhea of unknown cause. Colonoscopy should be considered in patients with dysentery who do not harbor an identified pathogen. Biopsies are obtained for evaluation of bowel histology and detection of Giardia, CMV, Cryptosporidium, and MAC. Management of the HIV-infected child with diarrhea must include close attention to nutritional support. Empiric antibiotic therapy for Salmonella spp. should be considered for the child with dysentery and signs of toxicity. Other therapeutic interventions are guided by the results of diagnostic studies.

Esophagitis Esophagitis is common among HIV-infected individuals. Affected children can have irritability, pain, difficulty swallowing, hiccups, or only fever. Esophagitis can be due to a number of infectious and noninfectious causes. Candida species are the most common cause. Children with esophageal candidiasis usually have concomitant oral candidiasis. Viral causes of esophagitis include CMV, herpes simplex virus, and perhaps HIV per se.37 Esophageal disease caused by

666

SECTION

S Human Immunodeficiency Virus and the Acquired Immunodeficiency Syndrome

bacteria (including mycobacteria) is uncommon. Other causes of esophageal disease in HIV-infected children include lymphoma, acid reflux, and aphthous ulcers. Empiric antifungal therapy is a reasonable initial step in the management of the HIV-infected child with oral candidiasis and typical clinical features of esophagitis. Barium esophagography is considered if presentation is atypical or oral thrush is absent or if the patient fails to improve within 5 to 7 days. Radiographic differentiation of various types of esophageal ulcerative disease (e.g., candidiasis versus CMV disease) and recognition of dual infection are impossible. Therefore, if ulcerative or nodular lesions are identified radiographically, esophagoscopy and biopsy should be performed.

(HIV culture or PCR). Both tests must be performed at or after 1 month of age, and at least one must be performed at or after 4 months of age. After the first year of life, prophylaxis is indicated in the presence of severe immunosuppression (immunologic category 3). Children who have had an episode of PCP should receive lifelong prophylaxis to prevent recurrence. Trimethoprim-sulfamethoxazole (150 mg trimethoprim per square meter of skin per day) given orally in two divided doses on 3 consecutive days per week is the drug of choice for prophylaxis against infection caused by P. jirovecii.37 Alternative regimens for children who are intolerant of trimethoprim-sulfamethoxazole include oral dapsone and intravenous or aerosolized pentamidine.

Oral Infections

Pneumococcal Infection

Candidiasis is the most common oral infection of HIV-infected children.30 Several clinical variants are recognized, including pseudomembranous, hyperplastic, and erythematous candidiasis and angular cheilitis. Most cases of oral candidiasis respond to oral nystatin or clotrimazole therapy. Cases refractory to these first-line therapies may respond to oral administration of an azole agent (e.g., fluconazole). Oral ulcers are also common in HIV-infected children. CMV and herpes simplex virus are common infectious causes. Aphthous ulcers of undetermined cause also occur frequently. Aphthous ulcers are of three basic types: aphthous major (large ulcers persisting for longer than 3 weeks), aphthous minor (smaller ulcers that heal spontaneously within 5 to 10 days), and aphthous herpetiformis (multiple, discrete ulcers resembling those caused by herpes simplex virus infection). Aphthous major is the most common form observed in HIV-infected individuals; concomitant pharyngeal and esophageal ulcers are present in some cases. Diagnostic evaluation and management of HIV-infected children with oral ulcers are based on clinical presentation. A surface culture for herpes simplex virus and empiric therapy with acyclovir may suffice for patients with lesions typical of herpetic stomatitis. Biopsy and cultures for viruses, mycobacteria, and fungi should be considered if initial diagnostic studies are negative, the ulcers are large or deep, or the ulcers fail to heal in an appropriate length of time, with or without empiric therapy.

Heptavalent pneumococcal conjugate vaccine is recommended for universal use in children 23 months and younger and for HIV-infected children who are 24 to 59 months old.40 Heptavalent vaccine should be administered at 2, 4, 6, and between 12 and 15 months, concurrently with other childhood immunizations. HIV-infected children older than 23 months should receive additional pneumococcal prophylaxis based on the following schedule. 1. For HIV-infected children who have received 4 doses of heptavalent vaccine, a dose of 23-valent pneumococcal polysaccharide (23PS) vaccine is recommended at 24 months of age, to be given at least 6 to 8 weeks after the last dose of heptavalent vaccine. 2. For HIV-infected children who have received 1 to 3 doses of heptavalent vaccine before 24 months of age, a single additional dose of heptavalent vaccine should be given at least 6 to 8 weeks after the last dose of heptavalent vaccine. This should be followed by a dose of 23PS vaccine at least 6 to 8 weeks later. An additional dose of 23PS vaccine should be given no earlier than 3 to 5 years after the initial dose. 3. For HIV-infected children 24 to 59 months old who have received only a single previous dose of 23PS vaccine, there are minimal data regarding the safety of subsequent doses of pneumococcal conjugate vaccines. However, 2 doses of heptavalent vaccine are recommended, to be given at an interval of 6 to 8 weeks. Commencement of the heptavalent vaccine series should begin no earlier than 6 to 8 weeks after the last dose of 23PS vaccine. An additional dose of 23PS vaccine is recommended 3 to 5 years after the first dose. 4. For HIV-infected children 24 to 59 months old who have received no previous doses of either 23PS vaccine or heptavalent vaccine, 2 doses of heptavalent vaccine are recommended, to be given at an interval of 6 to 8 weeks, followed by a single dose of 23PS vaccine no less than 6 to 8 weeks after the last dose of heptavalent vaccine. An additional dose of 23PS vaccine is recommended 3 to 5 years after the last dose.

PROPHYLAXIS AGAINST INFECTION Guidelines for the prevention of opportunistic infections in individuals infected with HIV have been published by the United States Public Health Service and the Infectious Diseases Society of America38 and are updated periodically at http://aidsinfo.nih.gov/.

Pneumocystis jirovecii Pneumonia Identification of children who would benefit from PCP prophylaxis is complicated by two factors. First, some HIV-infected infants at greatest risk of PCP are either not known to be at risk for HIV infection (i.e., the mother is not known to be HIV-infected) or are known to be at risk but do not have confirmed infection (i.e., their HIV infection status is indeterminate). Second, because normal CD4+ lymphocyte counts are age-specific, a fixed CD4+ lymphocyte threshold cannot be used for defining the risk of PCP in young children. Guidelines for prophylaxis against P. jirovecii pneumonia published by the CDC take these issues into consideration.39 Prophylaxis is recommended for all HIV-exposed infants, beginning at 4 to 6 weeks of age. The guidelines recommend that all HIV-infected infants and those with indeterminate status of infection continue prophylactic therapy until they are at least 12 months of age. Prophylaxis can be discontinued if HIV infection has been reasonably excluded on the basis of two or more negative viral diagnostic tests

Intravenous Immunoglobulin Several investigators have reported clinical and immunologic improvement in HIV-infected children receiving IGIV. Early reports included relatively small numbers of children, lacked uniformity in patient selection, and used a variety of treatment regimens. Evidence for beneficial effects of IGIV emerged from two National Institutes of Health-sponsored multicenter trials of IGIV for children with HIV infection.22,41 The results of these trials formed the basis for Food and Drug Administration approval of IGIV for certain HIV-infected children. Improved immunologic health of HIV-infected children resulting from HAART and widespread use of trimethoprim-sulfamethoxazole for PCP prophylaxis has limited administration of IGIV. Current guidelines38 recommend its use in the following situations: 1. For HIV-infected children who have hypogammaglobulinemia (immunoglobulin G < 400 mg/dL).

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Management of HIV Infection

2. For HIV-infected children with a demonstrated propensity for recurrent bacterial infections.

Mycobacterium avium Complex Infection Clinical trials have demonstrated the effectiveness of clarithromycin and azithromycin as prophylaxis against MAC infection.42,43 Although rifabutin is also effective for the prevention of MAC infection, it should be used as an alternative agent for patients who are unable to tolerate macrolide therapy. Current guidelines from the United States Public Health Service and the Infectious Diseases Society of America recommend that HIV-infected adults and adolescents receive clarithromycin or azithromycin if the CD4+ lymphocyte count is less than 50 cells/mm3. Recommendations for pediatric MAC prophylaxis are based on the child’s age and CD4+ lymphocyte count. Clarithromycin or azithromycin should be administered to the following groups of children: 6 years of age, < 50 cells/mm3; 2 to 6 years of age, < 75 cells/mm3; 1 to 2 years of age, < 500 cells/mm3; and < 12 months, < 750 cells/mm3.

Fungal Infections Oral candidiasis occurs in approximately 15% to 40% of children with HIV infection and is more common among children with low CD4+ lymphocyte counts or symptomatic HIV disease than among those with normal counts or no symptoms.44 Long-term survivors (> 5 years) of vertical HIV infection are significantly less likely to have histories of oral candidiasis than are children who die during infancy or early childhood.45 Several trials of prophylaxis in adults using daily or weekly fluconazole indicate that clinical relapses can be prevented. Emergence of resistance to fluconazole, particularly among species of Candida other than C. albicans, is of potential concern.46,47 Primary prophylaxis of candidiasis in HIV-infected infants and children is not recommended.38 For children with infrequent recurrences, it is reasonable to treat each individual clinical episode. Children with frequent recurrences or a history of esophageal candidiasis are candidates for prophylaxis with oral nystatin, clotrimazole, or fluconazole. Specific circumstances dictate the use of antifungal agents for prophylaxis of systemic noncandidal fungal infections in HIV-infected patients.38 Prospective controlled trials indicate that fluconazole and itraconazole are effective in the prevention of cryptococcosis in adults with advanced HIV disease (CD4+ lymphocyte count < 50 cells/mm3).38

Viral Infections Prevention of most viral infections in HIV-infected children is not possible. Two doses of live attenuated measles, mumps, rubella and varicella vaccine should be administered to HIV-infected children who are asymptomatic and not immunosuppressed (i.e., CDC immunologic category 1) beginning at 12 to 15 months of age.48 Measles, mumps, and rubella and varicella vaccines should not be administered to other HIV-infected children because of the potential for disseminated viral infection. Influenza immunization is recommended on an annual basis. Varicella-zoster immunoglobulin (see Chapter 205, VaricellaZoster Virus) should be administered as early as possible after exposure of a susceptible HIV-infected child to an individual with chickenpox or zoster.48

PROGNOSIS Certain opportunistic and nonopportunistic infections can be important clinical indicators of HIV disease progression, immunologic deterioration, and early demise. PCP, especially when it occurs

CHAPTER

113

667

during the first year of life, appears to portend a poor prognosis for long-term survival.49,50 In one study, children with recurrent bacterial infections or LIP/PLH had a survival time at least four times longer than those with PCP, disseminated MAC disease, CMV infection, tumors, or PML.49 Pediatric HAART is associated with a marked reduction in rates of progression to AIDS, opportunistic infections, and hospitalizations, as well as improved survival.51

CHAPTER

113

Management of HIV Infection Elizabeth D. Lowenthal, Juan Carlos Millon and Mark W. Kline

At the end of 2004, it was estimated that 2.2 million children worldwide were infected with human immunodeficiency virus (HIV). About 640 000 of those were infected during 2004 alone.1 The majority of pediatric infections are acquired through mother-to-child transmission. Yet worldwide, only a small minority of pregnant women have access to services for the prevention of mother-to-child transmission (PMTCT). Without antiretroviral (ARV) treatments, a large percentage of perinatally HIV-infected children die during the first few years of life. With appropriate care and medical treatments, however, HIVinfected children often thrive. The disease requires the long-term administration of multiple medications according to complicated schedules. Effective ARV therapy must be combined with a holistic approach that considers the functioning and quality of life of the child and family. A multidisciplinary team is essential to optimize medical management, minimize psychosocial problems, improve nutrition, and monitor developmental progress.

ANTIRETROVIRAL THERAPY Therapeutic Agents No ideal ARV drug has yet been developed. An ideal drug would be curative and without side effects. Currently licensed drugs can decrease the viral burden profoundly, but have significant side effects and demand strict adherence to dosing schedules. Regimens of highly active ARV therapy (HAART) involve combinations of ARVs that attack multiple sites in the virus life cycle (Figure 113-1).

Nucleoside and Nucleotide Reverse Transcriptase Inhibitors Reverse transcriptase (RT) is the viral enzyme responsible for transcription of viral RNA into a DNA provirus molecule. Nucleoside and nucleotide analogues take advantage of the relatively high affinity of virus RT for certain nucleotide modifications. Because human polymerases are able to recognize the modified nucleotides as “defective,” the nucleoside and nucleotide analogues selectively inhibit completion of viral replication. As an example, zidovudine is an adapted thymidine molecule that is incorporated by RT into viral DNA in preference to normal thymidine, although the modification is such that most host polymerases reject it. Because zidovudine has been altered in such a way as to preclude linkage to another nucleotide, when a polymerase incorporates zidovudine into a DNA molecule, the growing chain is terminated before completion. There are now seven licensed nucleoside RT inhibitors (NRTIs) in the United States: zidovudine (Retrovir), didanosine (Videx),

668

SECTION

sCD4 CD4-IgG

Penetration

S Human Immunodeficiency Virus and the Acquired Immunodeficiency Syndrome Reverse transcriptase inhibitors Zidovudine Didanosine Zalcitabine Stavudine Lamivudine Nevirapine

Uncoating

Reverse transcription

Figure 113-1. Human immunodeficiency virus (HIV) life cycle and location of action of various antiretroviral medications.

Protease inhibitors Saquinavir Indinavir Tat Ritonavir antagonists Nelfinavir

Trans- TransIntegration cription lation Assembly

Release HIV

HIV Host Double-stranded chromosome unintegrated DNA Proviral cDNA DNA Genomic RNA CD4 and chemokine receptors

Viral mRNA Nucleus

Cytoplasm

zalcitabine (Hivid), stavudine (Zerit), lamivudine (Epivir), emtricitabine (Emtriva), abacavir (Ziagen), and one nucleotide reverse transcription inhibitor (NtRTI): tenofovir (Viread) (Table 113-1). Most of the drugs have substantial toxicities, and, unfortunately, all have a limited duration of activity as monotherapy or in two-drug combinations. Toxicities common with one NRTI/NtRTI usually also occur with other drugs in the class. Those that are most commonly attributed to certain drugs in the class are specified below. The only currently recommended role for ARV monotherapy in children is the use of zidovudine for PMTCT. After maternal PMTCT measures have been instituted, most babies born to HIV-positive mothers receive 4 to 6 weeks of zidovudine monotherapy to decrease the chance of viral transmission further. The most common toxicities of zidovudine observed in children are anemia, thrombocytopenia, and neutropenia, all of which usually resolve with cessation of drug or dose modification.2 Other side effects commonly reported with zidovudine use include nausea, headaches, and myopathy. Stavudine, like zidovudine, is a thymidine analogue.3 Because these two drugs are activated by the same cellular kinases, they compete metabolically with each other and should not be used together. Typically, one of these thymidine analogues is combined with one or more nonthymidine NRTIs as the nucleoside backbone for a HAART regimen. Stavudine and zidovudine have similar resistance patterns, so that the second tier of therapy (with the alternate drug) may not be as effective as the first. Peripheral neuropathy, lipodystrophy, and lactic acidosis sometimes occur in patients receiving stavudine. These side effects are less common among children than adults. Didanosine and zalcitabine also produce peripheral neuropathy as a primary toxicity. This most frequently manifests as painful paresthesias in the hands and feet. Hyporeflexia, loss of proprioception, and diminished sensation of light touch can also occur. To varying degrees, didanosine, zalcitabine, stavudine, and lamivudine can also cause pancreatitis, which usually manifests as abdominal pain associated with elevated serum amylase and lipase concentrations.4,5 Because serum amylase can derive from other body sources, particularly salivary glands, it must be fractionated into pancreatic and salivary bands, or a serum lipase level must be obtained, to confirm a diagnosis of pancreatitis. Lipase can also be elevated by nonpancreatic sources, but in general, the combination of elevated serum amylase and lipase concentrations in a child with abdominal pain can be considered diagnostic. Didanosine and stavudine are sometimes used together in HAART regimens. Because of their overlapping toxicities, some practitioners avoid this combination.

Generally, lamivudine is used in combination with zidovudine or stavudine.6 When used as monotherapy or as the sole NRTI in a combination regimen, resistance to lamivudine develops quickly. However, the mechanism by which HIV most commonly becomes resistant to lamivudine (development of the M184V mutation) may make the virus more sensitive to zidovudine and stavudine, making those two drugs more effective than they would otherwise be. The M184V mutation also decreases viral fitness. Thus, lamivudine is used by some specialists in salvage regimens to help make the virus less aggressive. Lamivudine is generally well tolerated compared with other drugs in the class. Lamivudine also has activity against hepatitis B virus, making it a particularly good therapeutic option for HIV/hepatitis B virus coinfected patients. Like lamivudine, zalcitabine and emtricitabine are cytosine analogues. Zalcitabine is rarely used because of its less favorable sideeffect profile and lack of clear benefit compared with other drugs in the class. Lamivudine and emtricitabine should not be used together because a single RT mutation can render both inactive. Abacavir is also a potent NRTI when given as a part of HAART.7 The primary shortcoming of abacavir is the development in some patients of a hypersensitivity reaction consisting of fever, nausea, headache, and frequently a rash.8 The mechanism of the reaction is unclear. If a patient has a reaction of this type, especially in the first 2 to 3 weeks of abacavir therapy, the drug should be stopped and the patient should not be rechallenged. Some patients on rechallenge have had a lethal hypotensive reaction. Tenofovir is the only nucleotide RT inhibitor currently marketed in the United States. It currently has Food and Drug Administration (FDA) approval for use in adolescents > 16 years and adults. Unlike the NRTIs, tenofovir is monophosphorylated. It therefore requires less intracellular processing before being utilized by the RT enzyme and terminating the viral DNA chain. Tenofovir should be used with caution in patients with renal impairment because it has been shown to affect renal function in some patients. The use of tenofovir in children is currently being studied. Preliminary evidence suggests that tenofovir may play an important role in optimizing salvage regimens for treatment-experienced children.9 Some of the most severe side effects reported with NRTI/NtRTI drugs can be caused by any drug in the class. The most worrisome side effect of the NRTIs in general is a syndrome of hepatic failure and lactic acidosis. This acute condition is often lethal, probably reflects mitochondrial toxicity, and can occur when patients are otherwise stable. Fortunately, it is rare. Lipodystrophy is also a common longterm side effect observed among patients receiving NRTIs and

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Management of HIV Infection

CHAPTER

113

669

TABLE 113-1. Medications and Common Side Effects Associated Toxicities Agent

Mode of Action

Hematologic

Neuropathic

Comments

Zidovudine (ZDV, AZT, Retrovir)

Nucleoside RT inhibitor

+++ (anemia, + neutropenia, thrombocytopenia, macrocytosis)

Pancreatic

+

Headache, fatigue, myopathy, myositis, perhaps hyperactivity

Stavudine (D4T, Zerit)

Nucleoside RT inhibitor

+

+

+++

Peripheral neuropathy may be less common in children than in adults; keep liquid refrigerated

Abacavir (ABC, Ziagen)

Nucleoside RT inhibitor

0

+

0

Hypersensitivity reaction (fever, rash, nausea, vomiting, abdominal pain) should lead to permanent discontinuation of the drug

Didanosine (ddI, Videx)

Nucleoside RT inhibitor

0

++

+++

Must be given with antacid or as entericcoated capsules; keep liquid refrigerated; best given 1–2 hours away from meals. Can produce GI disturbances related to antacid

Zalcitabine (ddC, Hivid)

Nucleoside RT inhibitor

+

++

+++

Stomatitis and aphthous esophageal ulcers

Lamivudine (3TC, Epivir)

Nucleoside RT inhibitor

+

++

+

Headache, fatigue (uncommon)

Emtricitabine (FTC, Emtriva)

Nucleoside RT inhibitor

0

0

0

Hyperpigmentation of palms and soles more common in children

Tenofovir (TDF, Viread)

Nucleotide RT inhibitor

0

0

0

Nevirapine (NVP, Viramune)

NNRTI

0

Efavirenz (EFZ, Sustiva)

NNRTI

0

+

0

Central nervous system effects: sleeping problems, abnormal thinking, depression; rash.

Amprenavir (APV, Agenerase)

PI

0

0

0

Nausea, vomiting; perioral paresthesias; may cause less body habitus change than other PIs

Indinavir (IDV, Crixivan)

PI

0

0

0

Hyperbilirubinemia; kidney stones due to drug precipitation in the renal collecting system; nephritis; upset stomach; hepatitis

Lopinavir/R (LPV/R, Kaletra)

PI

0

+

0

Nausea, vomiting (especially with liquid preparation); hepatitis; headache; fatigue

Nelfinavir (NLV, Viracept)

PI

+

0

0

Diarrhea; nausea; flatulence; hepatitis; asthenia

Ritonavir (RTV, Norvir)

PI

0

0

0

Nausea, vomiting (especially with the liquid preparation); perioral paresthesias; hepatitis; headache, fatigue

Saquinavir (SQV, Fortovase)

PI

+

0

0

GI upset

GI, gastrointestinal; NNRTI, non-nucleoside reverse transcriptase inhibitor; PI agent, protease inhibitors; RT, reverse transcriptase. 0, not expected; +, rare; ++, occasional; +++, frequent. Note: All nucleoside RT inhibitor agents are associated with lactic acidosis and fatty liver. All PIs have the potential to cause fat redistribution, with thinning of the face and extremities, and truncal and breast deposition of fat. All PIs have interactions with the cytochrome P-450 enzyme system that affect clearance of other hepatically metabolized drugs. Most PIs can cause hypercholesterolemia and hypertriglyceridemia.

protease inhibitors. Notwithstanding these side effects, NRTI drugs are generally well tolerated. Several fixed-dose combinations of NRTIs are available to help simplify dosing. Combivir is a fixed-ratio combination of zidovudine and lamivudine that, for older children and adolescents, allows twicedaily dosing of both drugs with a single pill. Emtricitabine and tenofovir are also formulated as a fixed-dose combination tablet called Truvada, which can be given as a once-daily dose. Trizivir, a triplenucleoside formulation, is also available. It is a fixed-dose tablet of abacavir, zidovudine, and lamivudine that can be given as a single tablet twice a day. When first available, it was hoped that Trizivir would be as effective as regimens containing three ARVs from at least two drug classes. Studies have shown, however, that Trizivir is less efficacious than regimens containing two NRTI drugs plus efavirenz

in treatment-naive patients.10 Because of this, Trizivir is rarely recommended alone for first-line treatment of HIV-infected patients.

Non-Nucleoside Reverse Transcriptase Inhibitors There are two licensed non-NRTIs used in children, nevirapine (Viramune) and efavirenz (Sustiva). They produce noncompetitive inhibition of the RT enzyme to block initiation of the reverse transcription process. The two drugs have long half-lives, relatively complex pharmacokinetic properties, excellent activity, and almost complete cross-resistance. In a patient in whom one of these two drugs has failed virologically, there is seldom reason to try the other. However, they do not overlap entirely in terms of hypersensitivity, so a patient who has an adverse reaction to one may still be able

670

SECTION

S Human Immunodeficiency Virus and the Acquired Immunodeficiency Syndrome

to tolerate the other. Although generally very effective as part of treatment for HIV-1 infections, NNRTIs have little activity against HIV-2.

Nevirapine Nevirapine is available in tablet and liquid formulations. Its pharmacokinetic properties and dosing have been established for all ages, including neonates. One unusual aspect of nevirapine dosing is that, early in therapy, the drug induces its own metabolism, such that the half-life becomes shorter with use. Thus, it is given at a lower dose for 14 days (once daily) before being advanced to the standard dose (twice daily). Nevirapine is associated with two frequent and potentially significant side effects. First, it can cause serious hepatotoxicity, particularly during the first 6 weeks of treatment. Serum hepatic enzymes must be closely monitored and the medication held or dosage modified if toxicity becomes evident. The other serious reaction is a skin rash, which can progress from red macules to Stevens–Johnson syndrome. If a patient develops a moderate or severe rash on nevirapine treatment, it is prudent to stop the drug.

Efavirenz Efavirenz is available in capsule and tablet forms. Its pharmacokinetic properties vary with the age of the patient, and dosing is not yet established in children under 3 years of age. A liquid formulation of efavirenz has been studied and found to be less bioavailable than the capsule form.11 The capsules can be opened and the contents sprinkled on small amounts of food. Like nevirapine, efavirenz can produce skin rashes, including Stevens–Johnson syndrome. The most common significant adverse effect of efavirenz is central nervous system (CNS) toxicity. The drug can cause sleep disturbances, lethargy, difficulty concentrating, hallucinations, and personality changes. In children, manifestations of efavirenz-related CNS toxicity can be subtle. Young children receiving efavirenz should be monitored for changes in behavior and school performance in order to facilitate detection of CNS toxicities. Efavirenz should be avoided in women of childbearing potential. It is classified as FDA class D, contraindicated during early pregnancy due to the possibility of potentiating neural tube defects.

Protease Inhibitors The HIV protease inhibitors act specifically on the viral rather than the host’s enzyme. Most of the current protease inhibitors are peptidomimetic; they emulate the substrate of the enzyme but bind in the protease, preventing cleavage. Unfortunately, the protease inhibitors are difficult compounds to administer. They are not readily made into suspensions or water-based solutions, are usually unpalatable, and interact with hepatic cytochrome P-450 cytochromes in complex ways. Some protease inhibitors induce certain cytochrome P-450 isoforms and inhibit others. Ritonavir, as an example, is a relatively broad inhibitor of the cytochrome P-450 enzyme system that induces its own metabolism and also complicates the use of many other drugs. Ritonavir’s potent inhibition of cytochrome P-450 3A4 isoenzyme allows it to be used to therapeutic advantage as a “booster” of other protease inhibitors when given at low doses with the boosted drug. There are currently eight protease inhibitors licensed for use in the United States: nelfinavir (Viracept), ritonavir (Norvir), saquinavir (Invirase), indinavir (Crixivan), amprenavir (Agenerase), fosamprenavir (Lexiva), tipranavir (Aptivus), atazanavir (Reyataz), and lopinavir/ritonavir (LPV/r, Kaletra). Only amprenavir, ritonavir, nelfinavir, and lopinavir/ritonavir have pediatric liquid or powder formulations.

Ritonavir Ritonavir was the first protease inhibitor approved by the FDA for use in children and is available in both a liquid form and pills.

Unfortunately, the liquid has a bad taste and contains 43% ethanol. Ritonavir is highly potent but, because of significant side effects and drug interactions, must be used cautiously. Resistance mutations of HIV to ritonavir are almost identical with those to indinavir, so substitution of one drug for the other because of HIV resistance is unlikely to be of benefit (although simultaneous use may have some effect through pharmacologic boosting). Some viruses with high-level resistance to nelfinavir are also refractory to ritonavir, depending on the presence of specific mutations. Ritonavir is well absorbed orally, regardless of food intake. Its halflife is 3 to 4 hours. Ritonavir interacts metabolically with many other medications and care must be taken to adjust doses as needed for other drugs metabolized through the cytochrome P-450 system. Ritonavir is metabolized in the liver by the cytochrome P-450 system and acts as a potent inhibitor of cytochrome P-450 3-A4 isoenzyme. Pharmacologic properties of ritonavir have been used to extend the half-life of other medications, particularly other protease inhibitors. For example, lopinavir/ritonavir (Kaletra) is a combination of lopinavir with a low dose of ritonavir. Ritonavir extends the otherwise short half-life of lopinavir, improving trough levels and enhancing convenience of dosing. Ritonavir is associated with many side effects. Nausea and vomiting are produced both by the ritonavir itself and by the ethanol solvent. In some children, nausea becomes chronic and does not decrease over time. Ritonavir can also cause diarrhea, anorexia, headaches, circumoral paresthesias, and elevated serum hepatic transaminases. In rare circumstances, ritonavir can cause insulin resistance and diabetes. Hypercholesterolemia and hypertriglyceridemia are common with prolonged use.

Nelfinavir Nelfinavir has fewer adverse effects than ritonavir, but has a higher pill burden. The medication is available in tablets and as a granular powder. The powder is somewhat bulky and bad-tasting, although most children can take it when it is mixed with nonacidic liquids, puddings, or icecream. In general, tablets are preferable to the powder formulation. Nelfinavir levels cannot be boosted by administration with low-dose ritonavir as reliably as can other protease inhibitors.12 Nelfinavir has proven to be less effective than ritonavir-boosted protease inhibitors.13 Nelfinavir is similar to other unboosted protease inhibitors in its antiviral effect. It induces a different set of resistance mutations than ritonavir or indinavir. The most important mutation is of D30N, which was found in 56% of patients who received nelfinavir monotherapy for 12 to 16 weeks. There is marked cross-resistance between nelfinavir and the other protease inhibitors. Patients in whom nelfinavir fails and who have low-level viral resistance may respond to other protease inhibitors or combinations of protease inhibitors. However, 65% to 80% of isolates with more than 10-fold resistance to nelfinavir also had more than a fourfold increase in resistance to other protease inhibitors.14 Nelfinavir has a more tolerable adverse-effect profile compared with other drugs in this class. Diarrhea is most common, but is generally manageable with symptomatic measures. In some children, however, diarrhea is significant enough to preclude use of nelfinavir. Less common problems include fatigue, abdominal pain, and rashes. Diabetes can occur rarely during treatment with all protease inhibitors.

Lopinavir Lopinavir is only available commercially as a fixed-ratio combination along with ritonavir (Kaletra). The drug is available in gel cap, tablet, and liquid forms, although the liquid contains a high percentage of alcohol and tastes poorly. Lopinavir has resistance mutations that appear to be unique among the protease inhibitors. Side effects of lopinavir/ritonavir are similar to those of ritonavir. Lopinavir/ritonavir is generally better tolerated than ritonavir given alone in therapeutic doses.

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Management of HIV Infection

Saquinavir There is less information about pediatric use of saquinavir compared with some other protease inhibitors. Saquinavir must be used in combination with ritonavir. This combination is rational both because saquinavir has a resistance pattern distinctly different from that of ritonavir and because ritonavir increases serum saquinavir levels. Coadministration of saquinavir with ritonavir allows for twice-daily dosing. Concomitant nelfinavir, indinavir, or ritonavir therapy increases saquinavir levels by inhibiting hepatic metabolism (approximately 4-, 6-, and 20-fold, respectively). Drug interactions associated with saquinavir are similar to the other protease inhibitors, although they are less significant than those of ritonavir.

CHAPTER

113

671

mutations. Both of these protease inhibitors are best used in combination with ritonavir. They have been approved for adult use, but not yet for use in children < 18 years of age. The newest class of FDA-approved ARV drugs is the fusion inhibitor (FI) class. These drugs target the entry complex between the human cell and the HIV virion in order to prevent entry of the viral capsid into the cell. FIs bind to the gp41 protein on the viral surface and block the conformational changes necessary for membrane fusion to occur. The only commercially available drug in this class is fuzeon (Enfuvirtide, T-20). It is available as a subcutaneous injection that must be administered twice daily. Enfuvirtide’s main side effect is a local inflammatory reaction at injection sites, which occurs in approximately 95% of patients. Enfuvirtide is mainly used as a part of a salvage regimen in ARV-experienced patients.

Amprenavir and Fosamprenavir Amprenavir has a potent antiviral effect, similar to that of other protease inhibitors. Interactions with indinavir and nelfinavir appear to be favorable.15 Fosamprenavir (Lexiva) is a prodrug of amprenavir that is rapidly converted to amprenavir by cellular phosphatases in vivo. Amprenavir is associated with a relatively unique resistance pattern, which makes it a candidate to use in combination with other protease inhibitors. Amprenavir is cleared by the cytochrome P-450 system, which it also inhibits. However, amprenavir does not appear to induce or inhibit its own metabolism. The plasma half-life is somewhat variable but averages approximately 7 hours, which makes it suitable for twice-daily dosing. The most common adverse effects of amprenavir are gastrointestinal disturbances. Some patients have headache, malaise, or fatigue. In 2005, production of the adult formulation of amprenavir was stopped as fosamprenavir lowered the pill burden and became preferred for use in adults. Fosamprenavir dosing for pediatric patients has not yet been established. Pediatric formulations of amprenavir are still being produced and can be used in children over 4 years of age.

Indinavir Indinavir is widely used among adult patients, but experience in children is limited because of the lack of a pediatric formulation. Indinavir is potent, has a similar resistance pattern to ritonavir, and produces several adverse effects. Although pediatric studies are not complete, it appears that indinavir will be administered every 8 hours in children and must be given on an empty stomach. Like the other protease inhibitors, indinavir has interactions with numerous other medications but to a lesser degree than ritonavir, making drug interactions somewhat easier to manage. Because of its solubility characteristics, indinavir can form crystals in the kidney after filtration through the glomeruli. These crystals can form small kidney stones. To minimize the formation of indinavir stones, patients should drink large amounts of water or other fluids after taking the drug. The proportion of children who will have stones is not yet known, although some studies have suggested that a higher peak serum concentration of indinavir is achieved in children after oral dosing, which might lead to more stone formation. Other side effects of indinavir include hyperbilirubinemia (occurring in 5% to 10% of patients), nausea, abdominal pain, headache, and rarely, diabetes. As is often observed with other protease inhibitors, a pattern of fat redistribution occurs in adults who use indinavir chronically.

Recently Approved Drugs Atazanavir (Reyataz) and Tipranavir (Aptivus) are some of the most recently FDA-approved protease inhibitors on the market. Atazanavir has a less pronounced effect than other protease inhibitors in elevating lipid levels. Tipranavir is the newest protease inhibitor and has been marketed for use in patients who are heavily experienced with protease inhibitors and whose HIV has multiple protease inhibitor

Practical Use of Antiretroviral Therapy Because of the rapid development of resistance when ARV drugs are administered as monotherapy, they are almost always given in combination. Maximal durable response to ARV therapy occurs when at least three medications are given which attack the virus at two or more different points in the life cycle. Combinations of ARV drugs that meet these criteria are referred to as HAART. The goals of HAART therapy are to suppress the virus maximally, to restore and maintain CD4+ T-lymphocyte counts, to improve the body’s immune response, and to promote normal growth and development of the child. Through these successes a good quality of life may be achieved for a long period of time. Successful administration of ARV therapy relies on the ability to overcome difficulties related to palatability, high pill burdens, storage requirements, and medication-related food requirements. Didanosine, for example, although a useful member of the NRTI class, can be difficult to administer to children. Didanosine is acid-labile and must be given on an empty stomach. Older formulations of didanosine that included an antacid were often rejected by children.16 For older children, didanosine is now available in an enteric-coated formulation that does not release the didanosine until it has passed through the stomach. Ritonavir and lopinavir/ritonavir tolerance are strongly affected by their taste. Indinavir, like didanosine, needs to be given on an empty stomach. However, the antacid given with didanosine decreases the absorption of indinavir, so the two should not be given together on the same schedule. Physical habitus may change as a result of long-term drug administration of many NRTI and protease inhibitor medications. Any of the drugs, if taken at less than 95% of the prescribed dosing schedule, have significant potential to select for resistant HIV strains. Once resistance to a drug is established, it remains present for life. Thus, relatively brief periods of carelessness with medication can have lifelong repercussions. Because of this potential, great care should be taken in deciding when a patient and family are ready for therapy. Controversy exists about when to start ARV drugs.17,18 Most controlled data are from studies conducted in adults and extrapolated to children.19 There are several important differences in pediatric and adult HIV disease that make translation tenuous, including the following: 1. Perinatally transmitted infection appears to be more aggressive than infection in adults. 2. The virus has different effects on a growing immune system than on a mature one. 3. Children are more likely to have CNS involvement. Children who are infected perinatally do not limit viral replication as effectively as do adults during primary infection,20 and more children have a rapidly progressive course, including acquired immunodeficiency syndrome (AIDS) or death, within 2 years of the time of infection.21,22 Data in adults demonstrate that treatment during “primary infection,” the time immediately after HIV inoculation, appears to

672

SECTION

S Human Immunodeficiency Virus and the Acquired Immunodeficiency Syndrome

have therapeutic benefits.23,24 However, adults are rarely aware of infection in time for early therapy. By contrast, HIV-infected infants are identified within months of infection (either at birth or in utero) and can be treated at a time when virus dissemination is less well established. Consequently, many specialists believe that all infected children, particularly infants, should receive antiviral therapy at the time of identification. This hypothesis has not yet been verified systematically, but data from animal studies are supportive.25 Timing of the initiation of therapy in HIV-infected infants and children has been an area of controversy in the field of pediatric HIV. Many experts recommend starting ARV therapy in all children under 12 months of age regardless of immune category and viral load. This is a rational approach because infants and children under 12 months of age are at high risk of disease progression and to date no reliable predictors of disease progression in this age group have been identified. Some experts, however, recommend delay of initiation of therapy for children older than 6 months of age until there has been some decay in immunologic function or clinical condition. Delaying therapy is justified on the basis of concerns for side effects and the risk of formation of early, lifelong resistance when adherence or absorption is suboptimal.26

CHEMOPROPHYLAXIS OF OPPORTUNISTIC INFECTIONS Probably the single most effective therapeutic intervention for HIVinfected children other than the use of ARVs has been the introduction of prophylaxis for Pneumocystic jirovecii (formerly P. carinii) pneumonia (PCP). Because PCP is most common in children younger than 8 months of age, prophylaxis should be initiated in early infancy. The Centers for Disease Control and Prevention (CDC) published consensus guidelines regarding the timing and method of PCP prophylaxis in 1995.27 All HIV-seropositive infants should be started on trimethoprimsulfamethoxazole (TMP-SMX, Bactrim, Septra, co-trimoxazole) at 4 to 6 weeks of age. The child should continue prophylaxis until HIV infection is excluded. Specific criteria considered to be sufficiently reliable for cessation of TMP-SMX include two or more negative polymerase chain reaction (PCR) assays (one of which > 4 weeks of age and the other at > 4 months of age) and antibody seroreversion (two negative enzyme immunoassays at > 6 months of age). The original CDC PCP prevention guidelines suggested withholding prophylaxis until the CD4+ T-lymphocyte count in an infant was < 1500/mm3; however, PCP can occur at CD4+ T-lymphocyte counts above that level, making the broader use of PCP prophylaxis desirable. TMP-SMX is inexpensive and easy to administer and has a low rate of toxicity at the doses used for prophylaxis. Many dosing schemes have been recommended for TMP-SMX prophylaxis, including dosing 150 mg/m2 of TMP daily or 3 days each week, either as single or divided doses. In many households with HIVinfected children, adherence to an intermittent schedule is likely to be poor. Another frequently used option is 0.5 mL/kg of TMP-SMX suspension (4 mg/kg TMP) as a daily dose. HIV-infected children between 1 and 5 years of age with a CD4+ T-lymphocyte count of < 500/mm3 or a CD4+ T-lymphocyte percentage of < 15% should receive PCP prophylaxis, as should children 6 years of age with a CD4+ count of < 200/mm3 or < 15% CD4+ T lymphocytes. Many experts believe that any child who has had PCP, regardless of CD4 count, should remain on lifelong PCP prophylaxis. Recent data, however, suggest that both primary and secondary PCP prophylaxis can be safely discontinued once immune reconstitution has occurred.28 If a child does not tolerate TMP-SMX, alternative prophylactic options can be utilized. Dapsone (1 to 2 mg/kg once daily, not to exceed 100 mg/day) has been shown to be almost as effective as TMP-SMX for prophylaxis against PCP,29 but is sometimes not tolerated in patients with TMP-SMX-related toxicities. An alternative is pentamidine, which can be given as an aerosol (300 mg via

nebulizer) to those children who are old enough to breathe deeply on command, or as an intravenous infusion (4 mg/kg once every 3 to 4 weeks). Atovaquone has also been proven to be effective for the prevention of PCP in HIV-infected adults.30,31 Children with HIV infection have an increased rate of both serious and minor bacterial infections. There may be a role for antimicrobial prophylaxis to limit the number and severity of infections. Decreases in mortality, diarrhea, malaria, and CD4+ T-lymphocyte declines have been reported among HIV-infected patients without access to specific ARV treatments who took TMP-SMX daily.32 Immune globulin for intravenous use (IGIV) was once widely used to prevent infections among HIV-infected children. IGIV produces no further reduction in the rate of bacterial infections in children who receive TMP-SMX concurrently, however.33 Another bacterial infection that may warrant antimicrobial prophylaxis is Mycobacterium avium-intracellulare (MAI) disease. This infection only occurs in the late stage of HIV infection, causing fever, fatigue, weight loss, anemia, and neutropenia. To prevent MAI infection in end-stage HIV disease, azithromycin can be administered weekly. Daily dosing of clarithromycin can also be used for prophylaxis. Studies in adults have also demonstrated that rifabutin can be used to decrease the rate of symptomatic disease by 50%.34 Rifabutin prophylaxis in children is controversial; the decision is best made locally, considering regional incidence of mycobacterial infection and individual patients.

NUTRITION HIV-infected children commonly do not grow at normal rates. Height and weight are usually comparably affected, preserving a normal ratio. Children become wasted (i.e., develop a low ratio of weight for height) only in the late stages of HIV disease or when suffering concomitant illnesses or malnutrition (Figure 113-2). Throughout disease, balanced nutrition should be emphasized. High-protein and high-calorie formulations are often recommended. Multivitamin supplementation has also been shown to delay disease progression in low-resource settings.35 There is, however, no proof of efficacy among children with consistently good nutritional access and ARV therapies. Children in late stages of disease are often given nutritional supplements with high calorie and high protein contents. These can provide a balanced diet, but their main advantage is the ease with which they can be swallowed. Some children are fed via a nasogastric or gastrostomy tube, but this step has a cost, particularly in quality of life. Intravenous nutrition is most useful to support a child with an acute gastrointestinal problem. Chronic total parenteral nutrition is expensive and requires placement of a central venous access line, making it an unsatisfactory therapy generally. Appetite stimulants, such as Megace and Periactin, may increase oral intake and produce some growth. Unfortunately, in late HIV disease almost all interventions produce more increase in fat than in lean body mass, rendering their utility marginal. Effective ARV therapy and aggressive treatment of opportunistic infections are probably the most important determinants of optimal growth.

COMMON MANIFESTATIONS OF HIV INFECTION Clinical manifestations of HIV-infection are discussed in greater detail in Chapter 111 (Diagnosis and Clinical Manifestations of HIV Infection). In the sections below, some common manifestations are highlighted with a focus on therapy-related manifestations and side effects.

Hematologic Anemia, thrombocytopenia, and neutropenia are commonly observed in children infected with HIV and can be related to HIV infection

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Management of HIV Infection

CHAPTER

113

673

Cardiovascular SigniÀcant cardiovascular effects of HIV infection and ARV therapies have been recognized in children. Left ventricular dysfunction and increased left ventricular mass have been shown to occur frequently and to be associated with earlier mortality.37 HIV-infected children also develop early vascular disease and lipid abnormalities that may be worsened by the use of certain ARVs, particularly protease inhibitors.38 HIV-infected children treated with HAART have been found to have elevations in cholesterol similar to those seen in patients heterozygous for familial hypercholesterolemia. Because of this, they are thought to have similar risks of premature atherosclerotic disease.39 Cardiomyopathy is a frequent complication of HIV in children: it often manifests as a subacute or acute deterioration in cardiac function. Whereas some children ultimately recover normal function, most have some residual cardiac dysfunction. No single therapeutic strategy can be used, but most patients respond to a combination of digoxin, an afterload-reducing medication such as enalapril, and diuretic agents.

Metabolic Both HIV infection itself and the ARV therapies used for treatment have been implicated in the development of several metabolic abnormalities. In addition to lipid abnormalities, children with HIV frequently suffer from lipodystrophy, insulin resistance, hyperlactatemia, and decreased bone mineral density.40 Often, these conditions can be improved through alterations in the ARV regimens. Lipodystrophy syndrome consists of changes in fat distribution, typically manifesting as peripheral lipoatrophy with or without central adiposity. NRTI drugs are most commonly implicated in the development of lipodystrophy, although the syndrome is sometimes thought to be related to the use of protease inhibitors (Figure 113-3).

Dermatologic

Figure 113-2. Human immunodeficiency virus-associated wasting syndrome.

itself, toxicity of commonly used drugs, or as a consequence of superinfections, such as with parvovirus B19 or M. avium-intracellulare. Anemia directly caused by HIV infection is often improved with successful ARV therapy. The treatment of zidovudine-related anemia may require adjustments in zidovudine dose (typically through downward increments of 30%). The choice of PCP prophylaxis (to a drug other than TMP-SMX or dapsone) and treatment of superinfections (e.g., parvovirus with IVIG, M. avium-intracellulare infection with a combination of antimycobacterial drugs) should also be considered. Chronic anemia can be treated with transfusions, but this form of therapy is expensive and has attendant risks. Chronic anemia can also be treated with recombinant erythropoietin.36 The cost for this therapy is high – 25% to 50% of patients still require periodic transfusions – and the agent must be given subcutaneously several days each week. It is, however, a reasonable option for patients with chronic anemia. Neutropenia is also a common problem in HIV-infected children. Granulocyte colony-stimulating factor can be effective treatment for this condition. The usual dose is 1 to 20 —g/kg per day as a single subcutaneous injection. Dosage is begun at the low end of the range and advanced as needed. The cost is high, however, and attempts should be made to adjust dosing of zidovudine or TMP-SMX to assess their role in neutropenia.

Numerous skin and mucous membrane abnormalities occur in HIVinfected patients. Some common dermatologic problems are observed with increased frequency and severity among HIV-infected children (i.e., molluscum contagiosum, human papillomavirus lesions, eczema). Other skin abnormalities, such as Kaposi sarcoma lesions, are rarely seen among children without HIV infection. HIV-related skin abnormalities can be divided into four main groups: (1) infectious (i.e., impetigo, herpes zoster); (2) malignant (i.e., Kaposi sarcoma); (3) abnormal immune responses (i.e., eczema, papular pruritic eruption); and (4) drug-induced. Drug-induced skin changes commonly occur among patients receiving ARV therapies. Drug eruptions of varied severity and frequency have been reported among patients taking most approved ARV medications. One of the more common side effects of both nevirapine and efavirenz is the development of a generalized macular, pruritic rash. In the most severe cases the rash can progress to Stevens–Johnson syndrome, requiring discontinuation of therapy (Figure 113-4). Of the protease inhibitors, indinavir has been implicated in the largest number of cutaneous manifestations, including alopecia, paronychia with lateral nailfold pyogenic granuloma-like lesions, and drug eruptions. Within the NRTI class, zidovudine is known to increase the pigmentation of skin and nails in the hands and feet. Hypertrichosis and leukocytoclastic vasculitis have also been reported with zidovudine use.41 Increased pigmentation of palms and soles of patients is commonly seen among those taking emitricitabine.

Pulmonary Pulmonary infections are frequent causes of morbidity and mortality among HIV-infected patients. Bacterial PCP and tuberculosis are

674

SECTION

S Human Immunodeficiency Virus and the Acquired Immunodeficiency Syndrome Figure 113-3. Lipodystrophy.

commonly implicated in childhood deaths. In addition to these infections, structural changes in the lungs are commonly seen. Lymphocytic inÀltrations of the lung are common and may progress to bronchiectatic and cystic changes. Focal lymphocytic hyperplasia within the lung parenchyma is thought to progress to more diffuse inÀltration of the alveolar septa and interstitial tissues. The latter histology is referred to as lymphoid interstitial pneumonitis (LIP). Chest radiographs of patients with LIP characteristically demonstrate diffuse reticulonodular inÀltrates, with or without hilar adenopathy (see Figure 113-2). When bronchiectasis and cystic lung changes occur as a result of advanced recurrent infections and lymphocytic inÀltration, a state of chronic respiratory insufÀciency may result. HAART is the only known effective therapy for LIP.

Neurologic HIV infection commonly affects both the central and peripheral nervous systems. CNS involvement can manifest as focal Àndings, and delay or regression of development in children. Careful monitoring of developmental progress for early detection of CNS problems is an important component of care of the HIV-infected child. Focal cerebral mass lesions, myelopathies, myopathies, progressive encephalopathy, seizures, cerebral vascular accidents, and peripheral neuropathies are all more commonly observed among HIV-infected compared with HIV-uninfected children. Peripheral neuropathy may occur in HIV-infected children who are not being treated with ARV therapies. Many of the nucleoside analogues (didanosine, stavudine, zalcitabine) can be neurotoxic and can exacerbate or trigger peripheral neuropathy. Patients with HIV infection can manifest neuropsychiatric symptoms. Efavirenz is commonly implicated in causing impaired concentration, sleep disturbances, anxiety, vivid dreams, nightmares, and even psychosis. Psychiatric side effects of other ARVs are rare. Secondary causes of psychiatric symptoms (such as hypothyroidism and vitamin B12 deÀciency in a depressed patient) should be excluded before psychiatric symptoms are deemed to be related to HIV or therapeutic interventions.

Renal Renal complications can affect the outcome of HIV infection.42 HIVassociated renal parenchymal disease has been classiÀed into four groups: (1) acute tubular dysfunction with electrolyte fluid abnormalities and/or renal failure caused by infection or nephrotoxic drugs; (2) immune-mediated HIV glomerulopathies (immunoglobulin A nephropathy, lupus-like syndrome, and HIV-associated immune complex renal disease); (3) HIV-associated thrombotic microangiopathies, including atypical forms of hemolytic–uremic syndrome; and (4) HIV-1-associated nephropathy (large edematous kidneys with focal segmental glomerulosclerosis and tubulointerstitial lesions). Crystallization of indinavir, particularly in patients with inadequate hydration, is commonly implicated in causing renal calculi. Following resolution of symptoms in patients with renal colic, indinavir can usually be safely restarted. Several of the nucleoside/nucleotide RT inhibitors are eliminated through the kidneys, requiring dose adjustments in patients with renal failure.

THE FUTURE OF HIV The introduction of highly active ARV therapies has led to remarkable improvements in morbidity and mortality in HIV-infected children and adults. For people with access to HAART, HIV is now a chronic disease rather than an inevitably terminal condition. Survival trends have been most impressive among HIV-infected children in the United States and other areas where early initiation of HAART has become the standard of care for most HIV-infected children.43 Unfortunately, despite recent ambitious international initiatives, most HIV-infected children worldwide do not have access to HAART. With dedication and resources, ARV treatment programs for children can be successful even in low-resource settings.44,45 The expansion of HIV programs to include all children in need of treatment in heavily affected regions and resource-limited regions worldwide will depend upon numerous factors. These include improvements in healthcare infrastructure, stafÀng levels, and Ànancial inputs. SimpliÀcation of ARV therapies, more widespread availability of pediatric ARV formulations, and increasing knowledge of pediatric-speciÀc

PART II Clinical Syndromes & Cardinal Features of ID: Approach to Diagnosis & Initial Management

Management of HIV Infection

treatment issues in low-resource settings are also likely to have significant impacts on the survival of children in these areas. Despite the promise of well-implemented treatment strategies, the greatest hope for long-term successful management of the HIV pandemic is the scale-up of prevention programs. For children, universal implementation of PMTCT programs is most likely to have a significant impact on reducing the incidence of HIV. Currently, it is estimated that only 3% of HIV-infected pregnant women worldwide have access to effective PMTCT interventions. For older children, adolescents, and adults, there is hope that one day vaccinations will stem the spread of HIV. A great deal of research effort is aimed at the development of therapeutic vaccines for the treatment of HIV infection. Early studies of candidate vaccines suggest that immune-modulating vaccines may some day play an important role in helping to improve virologic control in HIV-infected patients.46 Ultimately, control of HIV depends on the development of an effective prevention vaccine. Numerous barriers to the development of an effective HIV prevention vaccine exist. Our understanding of specific immune

CHAPTER

113

675

responses that are protective against HIV infection is still rudimentary. The inherent and increasing genetic variance of the HIV virion and the structural complexity of potential target sites increase the challenge of initial vaccine design plans. The lack of an ideal animal model for HIV infection increases the difficulty of testing early candidate vaccines. Because HIV can be transmitted as either a free viral particle or as cell-associated virus, the ideal vaccine must stimulate both immunoglobulin and cell-mediated immune responses at the mucosal and systemic levels. The most promising vaccination strategies currently being tested are those that combine different vaccination approaches to administer the same antigens in order to generate a more complete immune response. The three vaccines of this type, however, that have been studied in phase III clinical trials have all failed to show a significant protective effect.47 As we move through the third decade of HIV-related research and treatment, we have a great deal to offer our pediatric patients. Current therapies promise that many of our HIV-infected children will live to have families of their own. With available and developing preventive strategies, these children may live to see an HIV-free generation.

PA RT

III

Etiologic Agents of Infectious Diseases

SECTION

A

Bacteria

CHAPTER

114

Classification of Bacteria Joseph W. St. Geme III

The ability to discriminate between distinct groups of organisms and to communicate with a common language about organisms in the context of disease is essential for clinical microbiologists and for physicians caring for patients. The official taxonomic ranks for naming bacterial organisms are kingdom, division, class, order, family, genus, and species. A bacterial species is defined as a distinct group of organisms that share a constellation of properties. This definition is somewhat subjective, and as new data become available, reclassification is occasionally necessary.

IDENTIFICATION BY PHENOTYPIC CHARACTERISTICS Historically, bacterial classification has been based on phenotypic characteristics. A useful first approach to classification was developed in 1884 by Christian Gram, who observed that some bacteria retain crystal violet dye after decoloration with ethanol and others do not. Organisms that retain crystal violet dye appear blue and are called gram-positive. Bacteria that are decolorized appear red when counterstained with safranin and are referred to as gram-negative. The staining characteristics of gram-positive and gram-negative bacteria reflect differences in cell wall structure. As shown in Figure 114-1, gram-positive cells have a relatively simple cell wall composed of an inner membrane and a surrounding peptidoglycan layer between 30 and 200 molecules thick. The gram-positive cell wall also contains teichoic acids and lipoteichoic acids, which are water-soluble polymers of polyol phosphates. Gram-negative organisms have a more complex cell wall, characterized by the presence of an inner membrane, a surrounding peptidoglycan layer one to two molecules thick, and an outer membrane (see Figure 114-1). The outer membrane contains lipopolysaccharide (endotoxin), which is unique

to gram-negative organisms. The compartment between the inner and outer membranes is called the periplasm and contains a number of degradative enzymes and transport-related proteins. Whereas most bacteria are gram-positive or gram-negative, some stain poorly or fail to stain at all with Gram reagents. Examples are mycobacteria, some actinomycetes, treponemes, rickettsiae, anaplasmae, chlamydiae, and mycoplasmas. Mycobacteria possess a complex cell wall that is rich in lipids and have been referred to as acid-fast because of their resistance to decolorization with acid solutions. These organisms are best visualized microscopically with Ziehl–Neelsen or Kinyoun acid-fast stain. Most actinomycetes are gram-positive, but Nocardia and Rhodococcus characteristically take up Gram stain irregularly. They have a cell wall structure similar to that of mycobacteria and, like the mycobacteria, are acid-fast. The family Treponemataceae contains five genera, namely Treponema, Borrelia, Leptospira, Brachyspira, and Spirillum. The treponemes generally do not stain with Gram reagents or other standard laboratory stains and are usually best seen with darkfield microscopy. Members of the family Rickettsiaceae and the family Anaplasmataceae (Ehrlichia, Anaplasma, Wolbachia, and Neorickettsia) possess a cell wall typical of gram-negative bacilli but react weakly with Gram stain and are seen optimally with Giemsa or Gimenez stain. Members of the family Chlamydiaceae (Chlamydia trachomatis, Chlamydophila psittaci, and C. pneumoniae) are obligate intracellular organisms that possess inner and outer membranes similar to those of gram-negative bacteria but lack a peptidoglycan layer and do not take up Gram stain. The two important members of the family Mycoplasmataceae are Mycoplasma and Ureaplasma. Mycoplasmas do not have a cell wall; their cytoplasmic contents are enclosed only by a well-developed plasma membrane. Cellular morphology also plays an important role in classification. Bacteria can be separated into five major groups on the basis of morphology as viewed through the light microscope. Cells that are spherical or oval in appearance are described as cocci. Those that are rod-like or cylindrical are referred to as bacilli. Bacteria referred to as vibrios have a comma-like or curved-rod appearance. Bacteria with a helical or more undulating appearance are called spirochetes if they are flexible and spirilla if they are rigid. After cell division, the cell wall between daughter cells in some species may not separate completely, giving rise to cell arrangements. These cell arrangements are typically distinctive and may be helpful in identifying related organisms. Common arrangements are cocci in 677

678

SECTION

A Bacteria Teichoic acid Lipoteichoic acid

Peptidoglycan

Cytoplasmic membrane Gram positive

Lipopolysaccharide

Outer membrane Peptidoglycan Periplasm

Cytoplasmic membrane Gram-negative Figure 114-1. Schematic diagram of the cell wall structure of grampositive and gram-negative bacteria.

pairs, cocci in chains, cocci in irregular clusters, cocci in packets of four or eight (called sarcina), bacilli in pairs, bacilli in chains, unusually short bacilli (called coccobacilli), and bacilli with tapered ends (called fusiform bacilli). A wide range of other phenotypic characteristics is also commonly employed for the purpose of bacterial classification. For example, aerobic and facultative anaerobic organisms are distinguished from strictly anaerobic organisms, and spore-forming bacteria are distinguished from nonspore-forming species. Additional useful properties are carbohydrate fermentation abilities, susceptibilities to antibiotics and inorganic compounds, and reactivity with well-defined serologic reagents.

IDENTIFICATION BY MOLECULAR TECHNIQUES Beyond phenotypic characteristics, molecular analysis also plays a major role in determining phylogenetic relationships among organisms and in classifying bacteria. The historical standard in this respect is DNA-DNA hybridization, which allows the total DNA from one organism to be compared with that of any other organism.1 This technique involves incubating single-stranded DNA from one strain with single-stranded DNA from a second strain and assessing formation of a double-stranded DNA molecule. DNA reassociation is a specific, temperature-dependent reaction. The optimal temperature for DNA reassociation is 25°C to 30°C below the temperature at which native double-stranded DNA is denatured into single strands. Experience with several hundred species has led taxonomists to conclude that organisms with 70% or greater DNA-DNA relatedness belong to the same species. Occasionally, the 70% species-relatedness rule has been ignored when the existing nomenclature is both deeply ingrained and useful. One such example is Escherichia coli and the four species of Shigella. These organisms are all 70% or more related and should therefore be grouped into a single species instead of the present five species in two genera. However, to avoid confusion among members of the medical community, this change has not been made.

Mole percent guanine plus cytosine (G + C) content and genome size are two other measurements of DNA that are often helpful in classifying organisms.2 The G + C content in bacterial DNA ranges from about 25% to about 75%. The G + C percentage is specific for a given species but is not unique for that species. Therefore, two strains with similar G + C contents may or may not belong to the same species. On the other hand, if the G + C contents are very different, the strains cannot be members of the same species. Genome size, or the molecular mass of bacterial DNA, ranges between 1 ¥ 109 and 8 ¥ 109 da. In certain circumstances, genome size determinations can distinguish between groups. Such determinations were used to distinguish Legionella pneumophila from Bartonella (formerly Rochalimaea) quintana. L. pneumophila has a genome size of 3 ¥ 109 da, whereas the B. quintana genome is roughly 1 ¥ 109 da. In a number of cases, multilocus enzyme electrophoresis (MEE) has been applied for the purpose of classifying bacteria.3 This method detects differences in the electrophoretic mobilities of individual soluble metabolic enzymes. The cellular proteins of an organism are separated by starch gel electrophoresis, and the individual enzymes are detected with the use of specific substrates. Variations in electrophoretic mobility of a given enzyme typically reflect amino acid substitutions that alter protein charge; these variations thus represent different alleles of the gene encoding the enzyme. Through the use of a relatively large number of enzymes, a positive correlation between estimates of relatedness obtained by MEE and by DNA hybridization of whole chromosomal DNA has been demonstrated. It is therefore possible to use MEE to determine the level of relatedness of two strains or a group of strains. This method has been particularly useful in yielding quantitative data about the relationships between organisms within a given species. In addition, it has provided insights into bacterial evolution. Two additional methods that are useful for determining genetic relatedness between strains and for characterizing the population structure of a given species are pulsed-field gel electrophoresis (PFGE) and amplified fragment length polymorphism (AFLP) analysis.4,5 PFGE has been the accepted gold standard for many species and involves extraction of chromosomal DNA and then digestion of the DNA with a rare-cutting restriction enzyme, giving rise to a number of large (80 to 100 kb) fragments, which can be separated by applying an electrical field alternatively in two directions. The critical experimental variable is the pulse time, defined as the time a field is applied in one direction before it is abruptly switched to another direction. Each strain gives rise to a reproducible pattern of bands (a fingerprint), and the number of bands shared between strains provides a measure of relatedness. AFLP analysis involves extraction of genomic DNA and then digestion of this DNA with two restriction enzymes. Subsequently, two different double-stranded oligonucleotide adapters are added to the digested DNA, one compatible with the first restriction enzyme and the other compatible with the second enzyme. After ligation, an aliquot of the DNA sample is subjected to PCR amplification under highly stringent conditions with adapter-specific primers that have an extension of one to three bases at their 3„ ends, running into the unknown chromosomal fragment. An extension of one base (e.g., adenine) in one primer results in amplification of one-fourth of the ligated fragments (e.g., those with a thymine, but not those with adenine, guanine, or cytidine at the complementary position), whereas an extension of two or three bases results in amplification of a smaller fraction of fragments. After polyacrylamide gel electrophoresis, a highly informative pattern of 40 to 200 bands is obtained. The molecular technique that has taken on greatest importance for classifying bacteria is 16S (small subunit) ribosomal RNA (rRNA) gene sequencing. The 16S rRNA folds in a precise fashion with the large subunit rRNA to form ribosomes. Ribosomes are highly conserved structures found in all living cells that perform the crucial task of protein synthesis. Because the genes for 16S rRNA contain some sequences that have been highly conserved through the course of evolution and others that are highly variable, 16S rRNA gene sequence can be used to determine evolutionary and genetic relationships between organisms.6 A further attribute of 16S rRNA gene sequencing

PART III Etiologic Agents of Infectious Diseases

Staphylococcus aureus Green sulfur bacteria BacteroidesDeinococci Spirochetes flavobacteria and relatives

Green nonsulfur bacteria

115

679

115

Staphylococcus aureus Planctomyces and relatives Chlamydiae

Thermotoga

CHAPTER

CHAPTER

Gram-positive bacteria Cyanobacteria Proteobacteria

Figure 114-2. Phylogenic tree of the domain bacteria based on comparison of 16S ribosomal RNA sequences. Divisions that contain human pathogens are denoted by boldface branches. (Redrawn from Relman DA. The identification of uncultured microbial pathogens. J Infect Dis 1993;168.1.)

is that cultivation of the organism is not required; for example, through the use of PCR and oligonucleotide primers that correspond to conserved regions of the bacterial 16S rRNA gene, a stretch of bacterial DNA can be amplified directly from diseased tissue or any other relevant source.7 The variable regions of the resultant DNA sequence form the basis for specific phylogenetic analysis. In general, consensus primers can be designed for any discrete group of organisms with common ancestry. On the basis of 16S rRNA sequence comparisons, the evolutionary relationships among all known extant bacterial species can be represented in a phylogenetic tree (Figure 114-2).7 Although 16S rRNA gene sequencing has clear strengths related to defining genetic relationships and classifying bacteria, using this technique alone has limitations. For example, isolates that have divergent 16S rRNA gene sequences are consistently unrelated, but isolates that have nearly identical 16S rRNA gene sequences may or may not belong to the same species, resulting in the need for another method to explore further whether isolates are sufficiently similar to be assigned to the same species. With this information in mind, multilocus sequence typing (MLST) is becoming more common, circumventing the potential effects of simple stochastic variation or recombination at a single locus. MLST uses profiles of alleles at multiple housekeeping genes (usually seven or so) to group isolates and determine relatedness between strains.8 Over the past decade there has been a steady increase in the number of completely sequenced bacterial genomes, reflecting advances in sequencing technology. The availability of this information has led to the development of DNA microarrays and has allowed comparison of the whole genomes of individual isolates. In the coming years, it is likely that whole genome information will be increasingly used to assist with bacterial classification in general terms and in clinical settings.9,10 A summary of current nomenclature of clinically relevant microbes is published.11,12 Additional genetic information continues to spur reclassifications.13

Gina S. Lowell and Robert S. Daum Staphylococcus aureus is the most common pathogen isolated in pediatric patients in North America, and is a major cause of morbidity and mortality. It is the most virulent species of the genus Staphylococcus. Its pathogenicity reflects its ability to acquire and integrate accessory genetic elements that confer virulence. S. aureus is responsible for community- and healthcare-associated infections and toxin-mediated diseases. The species is particularly adept at evolving strategies to elude antimicrobial therapy, constantly modifying and limiting therapeutic options.

MICROBIOLOGY AND PATHOGENESIS Staphylococci are aerobic or facultatively anaerobic gram-positive bacteria that can persist in distressed environments such as acidic conditions, high sodium concentrations, and wide temperature variations. Staphylococci can survive on fomites, in dust, or on clothing for at least several days. The defining characteristics of S. aureus are the production of the extracellular enzyme coagulase and protein A (Figure 115-1). Clinical manifestations of S. aureus infection can result from direct local invasion with inflammation, hematogenous dissemination, or toxin release inciting inflammatory cascades and tissue necrosis.

Capsule and Cell Wall Most clinical S. aureus isolates have a polysaccharide capsule; 11 capsular serotypes have been described, but most experts believe that probably only four exist.1 The high prevalence of two of the capsular polysaccharide types (5 and 8) in almost all collections of clinically important human and veterinary isolates suggests an important role for these polysaccharides in pathogenesis, the nature of which is uncertain. Bloodstream infection (BSI) has also been caused by unencapsulated organisms, for example the so-called type 336 isolates.2 Immunization with the type 5 and 8 polysaccharide conjugated to Pseudomonas exotoxoid A is was basis for two recent immunization trials, neither of which met its primary endpoints.3–5 The cell wall of S. aureus is composed of peptidoglycan, capsular polysaccharide when present, ribitol teichoic acid, lipoteichoic acid, and many surface proteins, including protein A, which binds to the Fc region of the immunoglobulin G (IgG) molecule.6 IgG antibody binding to the staphylococcal cell surface in this nonphysiologic manner decreases the efficiency by which S. aureus are opsonized and phagocytosed.7–9

Surface Proteins Many proteins found on the surface of S. aureus isolates have been implicated in pathogenesis. Adherence of S. aureus to mammalian extracellular matrix components is mediated by a family of adhesins, the microbial surface components recognizing adhesive matrix

680

SECTION

A Bacteria

IgG Protein A Microcapsule

Fatty acid modifying enzyme (FAME) Lipase

Panton–Valentine leukocidin (PVL) V8 protease

Leukocidin R

Enterotoxins A, B, C1-3, D, E, H Toxic shock syndrome toxin-1 Exfoliative toxins A, B

Cell wall (peptidoglycan) INVASINS a-Toxin b-Hemolysin g-Hemolysin d-Hemolysin Phospholipase C Metalloprotease (elastase) Hyaluronidase, hyaluronate lyase

Capsular polysaccharide types 1, 2, 5, and 8 + ADHESINS

+ Clumping factor + Fibrinogen-binding protein + Fibronectin-binding protein A + Fibronectin-binding protein B + Collagen-binding protein A + Coagulase + Polysaccharide/adhesin + Polysaccharide intracellular adhesin + 220kd adhesin

Lipoteichoic acid Penicillinbinding Damaged tissue proteins b-Lactamase Extracellular matrix (ECM) Ribitol Foreign material teichoic acid Figure 115-1. Virulence factors and relevant surface adhesins of Staphylococcus aureus. Redrawn from Daum RS. Staphylococcus aureus vaccine. In: Plotkin SA, Orenstein WA (eds) Vaccines, 5th ed. Philadelphia, PA, Elsevier, 2007, in press.

molecules (MSCRAMMs).10 Coagulase is found on the bacterial cell surface as well as in its environment. It binds to host prothrombin and forms staphylothrombin, an enzyme that catalyzes the formation of fibrin from fibrinogen.11 The A and B clumping factors are cell surface proteins that bind to fibrinogen, producing the typical clusters of staphylococci when mixed with plasma.12 Coagulase and the clumping factors breech host defenses by causing localized clotting; the clumping factors may also aid in adherence to traumatized skin, endothelial structures, and foreign surfaces. Recognition of this role for the S. aureus clumping factors has prompted investigation into their potential use as vaccine antigens. Iron is an essential component of cytochromes and other redox proteins. A cell wall-bound system of iron acquisition and importation has been identified in S. aureus and is under investigation to define an apparently essential role in pathogenesis.13

Toxins The virulence of S. aureus is due to a combination of variously elaborated virulence proteins that include extracellular products, such as a, b, g, and d hemolysins (also called toxins), leukocidins, proteases, lipase, deoxyribonuclease, a fatty acid-modifying enzyme, and hyaluronidase.14–20 a-Hemolysin is the best-studied of the exotoxins. It causes hemolysis of erythrocytes, necrosis of skin, and release of cytokines and eicosanoids that may produce shock. It is lethal when injected into animals, and S. aureus mutants lacking a-toxin are less virulent.21 However, pathogenic S. aureus isolates that do not produce a-hemolysin have been identified. b-Hemolysin is a sphingomyelinase that can injure membranes that are rich in this lipid; most human isolates do not express b-toxin. The high prevalence of g-leukocidin genes in isolates causing necrotizing skin infections suggests the importance of this toxin in the production of dermonecrosis. S. aureus possesses several leukocidins: synergohymenotropic toxins that damage membranes of certain cells as a result of the concerted action of two elaborated, secreted proteins. The best studied is the Panton–Valentine leukocidin (PVL), a pore-forming cytotoxin,22 the toxic effect of which is said to result from the synergistic action of the proteins lukS-PV and lukF-PV.23 PVL has been associated with severe inflammatory lesions, presumably through activation of granulocytes.24 PVL-producing strains have been associated with

severe skin infections and lung infections,25 specifically necrotizing pneumonia.26 Despite these compelling associations the role of PVL in the pathogenisis of S. aureus infections remains controversial.26a,26b S. aureus causes a variety of so-called “toxin-mediated” diseases, including staphylococcal scalded-skin syndrome (SSSS), toxic shock syndrome (TSS), and staphylococcal food poisoning, which are caused by the action of exotoxins and enterotoxins. Exfoliative toxins A and B (ETA and ETB) cleave the glycoprotein desmoglein 1, promoting spread of S. aureus under the stratum corneum,27,28 resulting in blistering of the superficial epidermis that is characteristic of SSSS and bullous impetigo. TSS toxin 1 (TSST-1) and staphylococcal enterotoxins B and C SEB and SEC have been implicated in most cases of TSS. These toxins have superantigen activity, i.e., stimulate T lymphocytes nonspecifically, resulting in cytokine release and clinical toxic shock. At a skin or mucosal port of entry, TSST-1 can interfere with release of inflammatory mediators locally, which may be responsible for a surprisingly benign appearance at the local infective site. An expanding family of enterotoxins has been implicated in staphylococcal food poisoning; the most frequently implicated molecule is enterotoxin A.

Genetic Basis and Regulation of Pathogenicity Factors and Antimicrobial Resistance S. aureus is the most sequenced of all bacterial genomes. It has a circular chromosome of about 2,800,000 basepairs.29 Genes for many housekeeping functions are highly conserved. The genome also carries mobile genetic elements such as plasmids, transposons, prophages, and pathogenicity islands, many of which encode virulence factors or determinants of antibiotic resistance. These factors represent a subset of a large class of accessory gene products (including surface proteins, exotoxins, and other enzymes) that, although not required for growth and cell division, are advantageous in particular environments. S. aureus has evolved a remarkable network of regulatory mechanisms that control subsets of genes regulated under certain growth and environmental conditions. The best studied is the agr locus with its two-component signal transduction system and its effector RNAIII molecule. agr upregulates genes for capsular polysaccharides, a-d toxins, two-component synergohymenotropic toxins, enterotoxins,

PART III Etiologic Agents of Infectious Diseases

Staphylococcus aureus

exfoliatins, and proteases, and downregulates others such as MSCRAMMS, other adhesins, and protein A. Other two-component signal transduction systems include saeRS, srrAB, arlSR, and lytRS. DNA-binding protein systems such as sarA and its regulatory homologs also regulate virulence factors. Additionally the sigma factor sB can combine with RNA polymerase core enzyme to form a holoenzyme that can recognize promoter elements for at least 36 S. aureus genes involved in bacterial stress responses. Taken together, membrane sensing systems such as agr as well as DNA-binding regulatory systems afford S. aureus a versatile environmental response system and confer the capacity to respond to a myriad of environmental stimuli such as high or low NaCl concentrations, subinhibitory concentrations of protein synthesis inhibitors, low pH, low PO2 or limitation of essential nutrients such as amino acids. S. aureus can overcome environmental antibiotic pressure through the acquisition and transmission of resistance genes from other, usually less pathogenic, species and between isolates of the same species. Resistance to b-lactam antibiotics such as penicillins, cephalosporins, and carbepenems is one example. b-Lactam antibiotics bind to S. aureus penicillin-binding proteins (PBP), thereby inhibiting cell wall synthesis. Resistance to penicillin was documented almost immediately after its introduction into clinical practice in the 1940s and is mediated by the elaboration of a b-lactamase, encoded by a transposon borne on a plasmid. Almost all clinical isolates of S. aureus can elaborate this enzyme, rendering clinically ineffective antibiotics that are susceptible to b-lactamase hydrolysis. The development of b-lactamase-resistant semisynthetic penicillins temporarily overcame the clinical problem created by the wide prevalence of b-lactamase-producing S. aureus. However some strains became resistant to semisynthetic compounds by acquiring mecA, a gene that elaborates PBP2a, a peptidoglycan-synthesizing enzyme that has decreased affinity for b-lactam antibiotics.30,31 These resistant strains are termed methicillin-resistant Staphylococcus aureus (MRSA) and are responsible for nosocomial outbreaks and, more recently, a community-based MRSA epidemic. MecA is located on a mobile genetic element called the staphylococcal chromosome cassette mec (SCCmec),32 which is present in all MRSA isolates with a single exception. SCCmec contains a mec complex, comprised of mecA and the variably present mecI and mecRI regulatory genes, and a ccr complex, comprised of genes that mediate insertion and excision of SCCmec from the genome. Insertion of SCCmec into the S. aureus genome is the genetic event that converts a methicillin-susceptible strain into a methicillin-resistant one. There are currently six such SCCmec elements that have been sequenced or partially characterized; SCCmec types I to III are generally found in hospital- (or healthcare-) associated MRSA (HA-MRSA) isolates, whereas SCCmec types IV and V are generally found in communityassociated MRSA (CA-MRSA) isolates, although these distinctions are blurring.32a Although the mechanism for the movement of SCCmec elements from strain to strain is unknown, the large size of types I to III is believed to limit easy transfer of the elements. Types IV to VI, however, are smaller and probably, therefore, are more mobile. Types IV and V have been found in multiple S. aureus genetic backgrounds, supporting the hypothesis that they are readily transferred from strain to strain.33–35

EPIDEMIOLOGY Colonization Humans and other mammals are the natural reservoir for S. aureus. Asymptomatic colonization is frequent in humans and is most commonly detected in the anterior nares. Other areas that can be colonized include the skin, nails, pharynx, axillae, perineum, and vagina (which may be more frequent for CA-MRSA than MSSA). Colonization rates range from 25% to 50%, with higher rates observed in people with dermatologic conditions (e.g., eczema), frequent needle use (e.g., intravenous drug abusers), indwelling intravascular devices (e.g., dialysis patients), and people who are healthcare workers.

CHAPTER

115

681

Children have a higher colonization rate, possibly due to their frequent exposure to respiratory tract secretions. Three patterns of S. aureus colonization have been observed: about 20% of the population is persistently colonized, 60% is intermittently colonized, and 20% is almost never colonized.36 Nasal S. aureus carriage is a risk factor for developing infection,36–39 although most colonized individuals do not do so.

Hospital-Associated Infections S. aureus has demonstrated the ability to develop resistance to all classes of antimicrobial agents. Soon after the introduction of penicillin in the 1940s the acquisition by S. aureus of b-lactamase genes and their rapid dissemination resulted in widespread resistance among clinical isolates. Less than a year after the introduction of semisynthetic penicillins in the early 1960s that were active against so-called methicillin-susceptible S. aureus (MSSA), the first MRSA strain was reported. These multidrug-resistant, HA-MRSA strains spread worldwide within hospital settings. The National Nosocomial Infections Surveillance System (NNIS) of the Centers for Disease Control and Prevention (CDC) reports S. aureus as a major pathogen among nosocomial infections, with MRSA accounting for 64% of nosocomial S. aureus isolates in intensive care units in United States hospitals in 2003, an increase from 36% in 1992.40 In pediatric intensive care units in the United States, S. aureus is the major nosocomial pathogen in a variety of clinical situations, accounting for 9% of HA blood stream infections, 17% of cases of HA pneumonia, and 20% of surgical site infections.41 Current patterns of HA-MRSA isolates reveal decreasing resistance to non-b-lactam antibiotics, suggesting a shift in nosocomial MRSA epidemiology and probably reflecting the movement of “biologically fit” CA-MRSA isolates into the hospital.32a

Community-Associated Infections The first report implying prevalent MRSA infection occurring outside of the healthcare setting was in 1982 in Detroit, MI, among intravenous drug users.42 This and subsequent reports of CA-MRSA were associated with risk factors for infection similar to those known for HA-MRSA, including intravenous drug use, recent hospitalization or surgery, indwelling catheters or devices, dialysis, or residence in a long-term care facility.43–46 In the late 1990s, CA-MRSA infections occurring mostly in children with no identifiable predisposing MRSA risk factors began to emerge in case reports, retrospective reviews, and surveillance studies.47–51 Infections caused by these strains most commonly resulted in skin and soft-tissue infections, although some manifested as serious infections requiring hospitalization or resulting in death.52 Recognition of CA-MRSA causing infections ranging from skin and soft-tissue infections (SSTIs) to invasive disease has grown substantially in number and geographic distribution. Substantial evidence supports de novo rise of MRSA from MSSA in the community rather than movement of HA-MRSA into the community to explain most CA-MRSA disease although both CA-MRSA and HA-MRSA can be found to circulate in the community.53,54 Infection with CA-MRSA has been defined as isolation of the organism in an outpatient or within 72 hours of admission to the hospital in a patient without the following factors used to define risk for HA-MRSA: recent hospitalization or surgery, prolonged antibiotic therapy, underlying chronic disease, indwelling catheter or other device, healthcare contact, or residence in a long-term facility.55 Outbreaks of CA-MRSA infections have been reported in group childcare centers,46,56 sports teams,57,58 correctional facilities,59,60 and military units,61,62 suggesting that close contact and suboptimal hygiene practices play a role in spread. CA-MRSA isolates are distinguished from HA-MRSA by their lack of multidrug resistance. Most CA-MRSA isolates are susceptible to clindamycin, trimethoprim-sulfamethoxazole (TMP-SMX), and doxycycline, whereas HA-MRSA isolates are more often resistant to these agents. Genetic investigations such as multilocus sequence

682