Management of Infectious Complications in Cancer Patients

Management of Infectious Complications in Cancer Patients

Cancer Treatment and Research Steven T. Rosen, MD, Series Editor

Nathanson L (ed): Malignant Melanoma: Genetics, Growth Factors, Metastases, and Antigens. 1991. ISBN 0-7923-0895-6. Sugarbaker P H (ed): Management of Gastric Cancer. 1991. ISBN 0-7923-1102-7. Pinedo HM, Verweij J, Suit H D (eds): Soft Tissue Sarcomas: New Developments in the Multidisciplinary Approach to Treatment. 1991. ISBN 0-7923-1139-6. Ozols R F (ed): Molecular and Clinical Advances in Anticancer Drug Resistance. 1991. ISBN 0-7923-1212-0. Muggia FM (ed): New Drugs, Concepts and Results in Cancer Chemotherapy 1991. ISBN 0-7923-1253-8. Dickson RB, Lippman ME (eds): Genes, Oncogenes and Hormones: Advances in Cellular and Molecular Biology of Breast Cancer. 1992. ISBN 0-7923-1748-3. Humphrey G Bennett, Schraffordt Koops H, Molenaar WM, Postma A (eds): Osteosarcoma in Adolescents and Young Adults: New Developments and Controversies. 1993. ISBN 0-7923-1905-2. Benz CC, Liu ET (eds): Oncogenes and Tumor Suppressor Genes in Human Malignancies. 1993. ISBN 0-7923-1960-5. Freireich EJ, Kantarjian H (eds): Leukemia: Advances in Research and Treatment. 1993. ISBN 0-7923-1967-2. Dana BW (ed): Malignant Lymphomas, Including Hodgkin's Disease: Diagnosis, Management, and Special Problems. 1993. ISBN 0-7923-2171-5. Nathanson L (ed): Current Research and Clinical Management of Melanoma. 1993. ISBN 0-7923-2152-9. Verweij J, Pinedo HM, Suit H D (eds): Multidisciplinary Treatment of Soft Tissue Sarcomas. 1993. I SBN 0-7923-2183-9. Rosen ST, Kuzel TM (eds): Immunoconjugate Therapy of Hematologic Malignancies. 1993. ISBN 0-7923-2270-3. Sugarbaker PH (ed): Hepatobiliary Cancer. 1994. ISBN 0-7923-2501-X. Rothenberg ML (ed): Gynecologic Oncology: Controversies and New Developments. 1994. ISBN 0-7923-2634-2. Dickson RB, Lippman ME (eds.): Mammary Tumorigenesis and Malignant Progression. 1994. ISBN 0-7923-2647-4. Hansen H H (ed): Lung Cancer. Advances in Basic and Clinical Research. 1994. ISBN 0-7923-2835-3. Goldstein LJ, Ozols R F (eds.): Anticancer Drug Resistance. Advances in Molecular and Clinical Research. 1994. ISBN 0-7923-2836-1. Hong WK, Weber RS (eds.): Head and Neck Cancer. Basic and Clinical Aspects. 1994. ISBN 0-7923-3015-3. Thall P F (ed): Recent Advances in Clinical Trial Design and Analysis. 1995. ISBN 0-7923-3235-0. Buckner C D (ed): Technical and Biological Components of Marrow Transplantation, 1995. ISBN 0-7923-3394-2. Winter JN (ed.): Blood Stem Cell Transplantation. 1997. ISBN 0-7923-4260-7. Muggia FM (ed): Concepts, Mechanisms, and New Targets for Chemotherapy. 1995. ISBN 0-7923-3525-2. Klastersky J (ed): Infectious Complications of Cancer. 1995. ISBN 0-7923-3598-8. Kurzrock R, Talpaz M (eds): Cytokines: Interleukins and Their Receptors. 1995. ISBN 0-7923-3636-4. Sugarbaker P (ed): Peritoneal Carcinomatosis: Drugs and Diseases. 1995. ISBN 0-7923-3736-3. Sugarbaker P (ed): Peritoneal Carcinomatosis: Principles of Management. 1995. ISBN 0-7923-3727-1. Dickson RE, Lippman ME (eds.): Mammary Tumor Cell Cycle, Differentiation and Metastasis. 1995. ISBN 0-7923-3905-3. Freireich EJ, Kantarjian H (eds.): Molecular Genetics and Therapy of Leukemia. 1995. ISBN 0-7923-3912-6. Cabanillas F, Rodriguez MA (eds.): Advances in Lymphoma Research. 1996. ISBN 0-7923-3929-0. Miller AB (ed.): Advances in Cancer Screening. 1996. ISBN 0-7923-4019-1. Hait WN (ed.): Drug Resistance. 1996. ISBN 0-7023-4022-1. Pienta KJ (ed.): Diagnosis and Treatment of Genitourinary Malignancies. 1996. ISBN 0-7923-4164-3. Arnold AJ (ed.): Endocrine Neoplasms. 1997. ISBN 0-7923-4354-9. Pollock R E (ed.): Surgical Oncology. 1997. ISBN 0-7923-9900-5. Verweij J, Pinedo HM, Suit HD (eds.): Soft Tissue Sarcomas: Present Achievements and Future Prospects. 1997. ISBN 0-7923-9913-7. Walterhouse DO, Cohn SL (eds.): Diagnostic and Therapeutic Advances in Pediatric Oncology. 1997. ISBN 0-7923-9978-1. Mittal BB, Purdy JA, Ang KK (eds.): Radiation Therapy. 1998. ISBN 0-7923-9981-1. Foon KA, Muss H B (eds.): Biological and Hormonal Therapies of Cancer. 1998. ISBN 0-7923-9997-8. Ozols RF (ed.): Gynecologic Oncology. 1998. ISBN 0-7923-8070-3.

Management of Infectious Complication in Cancer Patients edited by

GARY A. NOSKIN Associate Professor of Medicine Northwestern University Medical School Medical Director, Infection Control Healthcare Epidemiology Northwestern Memorial Hospital Chicago, Illinois

KLUWER ACADEMIC PUBLISHERS BOSTON/DORDRECHT/LONDON

'w Ld

Distributors for North, Central and South America: Kluwer Academic Publishers 101 Philip Drive Assinippi Park Norwell, Massachusetts 02061 USA Distributors for all other countries: Kluwer Academic Publishers Group Distribution Centre Post Office Box 322 3300 AH Dordrecht, THE NETHERLANDS Library of Congress Cataloging-in-Publication Data Management of infectious complications in cancer patientsledited by Gary A. Noskin. cm. - (Cancer treatment and research; v. 96) p. Includes bibliographical references and index. ISBN 0-7923-8150-5 (alk. paper) 1. Infection - Treatment. 2. Cancer - Complications - Treatment. 3. Immunosuppression. I. Noskin, Gary A. 11. Series. RC112.M35 1998 616.99'4 - dc21 98-16254 CIP Copyright O 1998 by Kluwer Academic Publishers 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, mechanical, photocopying, recording, or otherwise, without the prior written permission of the publisher, Kluwer Academic Publishers, 101 Philip Drive, Assinippi Park, Norwell, Massachusetts 02061

Printed o n acid-,free paper. PRINTED IN THE UNITED STATES OF AMERICA

Contents

Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii

......................................................

xi

1. Host impairments in patients with neoplastic diseases . . . . . . . . . . . BEN E. DE PAUW, J. PETER DONNELLY, and BART-JAN KULLBERG

1

2. Epidemiology of infectious complications in cancer patients. . . . . . TERESA ZEMBOWER

33

3. Approach to fever in the neutropenic host . . . . . . . . . . . . . . . . . . . . ATHENA STOUPIS and STEPHEN H. ZINNER

77

Preface

4. Infections associated with solid tumors . . . . . . . . . . . . . . . . . . . . . . . . 105 SARAH H. SUTTON and JOHN P. FLAHERTY

5. Role of the clinical microbiology laboratory in the diagnosis of infections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143 RICHARD B. THOMSON, JR. and LANCE R. PETERSON 6. Recent advances in the management of fungal infections. . . . . . . . . 167 JASON SANCHEZ and GARY A. NOSKIN 7. Recent advances in the management of viral infections. . . . . . . . . . 183 JOHN R. WINGARD 8. Cytokines and biological response modifiers in the treatment ofinfection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201 BRIGITTA U. MUELLER and PHILLIP A. PIZZO 9. Prevention of infection in immunocomprised hosts. GARY A. NOSKIN

. . . . . . . . . . . . . 223

10. Pharmacologic considerations with antimicrobials used inoncology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247 MICHAEL POSTELNICK and SARA R. HALBUR 11. Economic impact of infections in patients with cancer.. . . . . . . . . . 283 DAVID J. SHULKIN and LAWRENCE J. ANASTASI Index

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291

Contributors

Lawrence J. Anastasi, DO University of Pennsylvania Medical Center 3400 Spruce Street Philadelphia, Pennsylvania 19104-4283 Ben E. De Pauw, MD, PhD Departments of Haematology and Blood Transfusion University Hospital Nijmegen Geert Grooteplein 8 PO Box 9101 6500 HB Nijmegen The Netherlands J. P. Donnelly, MD Department of Haematology University Hospital Nijmegen Geert Grooteplein 8 PO Box 9101 6500 HB Nijmegen The Netherlands John P. Flaherty, MD University of Chicago Hospitals 5841 S. Maryland Avenue Mail Code 5065 Chicago, Illinois 60637 Sara R. Halbur, Pharm D Department of Pharmacy Northwestern Memorial Hospital 250 E. Superior Street Chicago, Illinois 60611

B. J. Kullberg, MD Departments of Internal Medicine and Infectious Diseases University Hospital Nijmegen Geert Grooteplein 8 PO Box 9101 6500 HB Nijmegen The Netherlands Brigitta U. Mueller, MD Pediatric Branch National Cancer Institute Building 10, Room 13N240 Bethesda, Maryland 20892 Gary A. Noskin, MD Division of Infectious Diseases Northwestern University Medical School 710 N. Fairbanks Court-216 Chicago, Illinois 60611 Lance R. Peterson, MD Director, Clinical Microbiology Northwestern University Medical School 250 East Superior Street, Wesley 565 Chicago, Illinois 60611 Philip A. Pizzo, MD Pediatric Branch National Cancer Institute Building 10, Room 13N240 Bethesda, Maryland 20892 Michael Postelnick, R. Ph Department of Pharmacy Northwestern Memorial Hospital 250 E. Superior Street Chicago, Illinois 60611 Jason Sanchez, MD Department of Medicine Northwestern University Medical School 250 E. Superior Street Chicago, Illinois 60611

Athena Stoupis, MD Brown University School of Medicine Rhode Island Hospital Providence, Rhode Island 02903 David J. Shulkin, MD Chief Medical Officer University of Pennsylvania Medical Center 3400 Spruce Street Philadelphia, Pennsylvania 19104-4283 Sarah Sutton, MD MacNeal Hospital 3231 S. Euclid Avenue Suite 405 Berwyn, Illinois 60402 Richard B. Thompson, Jr., PhD Department of Pathology Northwestern University Medical School Evanston Hospital 2650 Ridge Road Evanston, Illinois John R. Wingard, MD Division of HematologylOncology University of Florida College of Medicine P.O. Box 100277 Gainesville, Florida 32610-0277 Teresa Zembower, MD Division of Infectious Diseases Northwestern University Medical School 303 E. Superior Street, 8E Chicago, Illinois 60611 Stephen Zinner, MD Brown University School of Medicine Rhode Island Hospital Roger Williams Medical Center 825 Chalkstone Avenue Providence, Rhode Island 02908-4728

Preface

Infection is a major cause of morbidity and mortality in patients with neoplastic disease because of compromised host defenses. These defects result in an increased risk of infection and its complications. The nature of the underlying malignancy, the immunodeficiencies associated with it, and the treatments directed against it are all important determinants of infection. In recent years the introduction of more intensive chemotherapeutic regiments and the widespread use of bone marrow and peripheral stem cell transplantation have changed the pattern of infection in many patients. Furthermore, the increasing use of central venous access devices and antimicrobial prophylaxis has changed the epidemiology of infection in these patients as well. To complicate matters, the microbiology of infections is continuously evolving. Whereas in the past the majority of serious infections were caused by gram-negative bacilli, this has changed to favor more gram-positive pathogens and fungi. This has resulted in the emergence of vancomycinresistant enterococci, glycopetide-insensitive Staphylococcus aureus, and Candida krusei. The objective of this volume in the Cancer Treatment and Research series is to emphasize that whereas the management of infection in cancer patients is common, it is constantly changing. With the increasing complexity of these patients, optimal management requires a multidisciplinary approach. Ultimately, it is hoped that this book will assist clinicians in the diagnosis, management, and prevention of infection in order to optimize care for patients with cancer.

1. Host impairments in patients with neoplastic diseases Ben E. De Pauw, J. Peter Donnelly, and Bart-Jan Kullberg

1. Introduction In the course of evolution nature has provided the normal human individual with an impressive and effective defense system against microbial enemies that eclipses even Star Wars, arguably the most advanced and ingenious defense program ever designed by human beings. On its own, the normal defense system recognizes foreign invaders, alerts the relevant protective mechanisms, launches counterattacks, ceases hostilities as soon as the job is done, and clears up the battlefield, causing only negligible collateral damage. An intact system offers protection against most microbial aggressors through a complex interrelationship of protecting surfaces, cells, and soluble factors. Optimal nutritional status and normal organ function form the basis of resistance to potentially dangerous microorganisms; therefore, it is somewhat artificial to further delineate the separate lines of defense [I] because all components are more or less dependent upon each other in attaining maximum efficacy. For instance, the skin and mucosal membranes are ranked amongst the first line of defense but they can only exert optimal activity in conjunction with the immunoglobulin (Ig) A and other secretory substances. Moreover, the surfaces of the human body exhibit a clear propensity to interact with colonizing microorganisms. The so-called commensal resident flora are normally avirulent, do not cause infection, and protect against more aggressive pathogens by competing for binding sites on the surfaces and for the available nutrients. Therefore, white blood cells (granulocytes, macrophages, and lymphocytes), platelets, soluble factors of the immunoglobulins, complement, lymphokines, and other cytokines, as well as the physical barriers, have to be considered as integral and virtually indispensable components of a unitary defense system (Figure 1). Given its complexity, it is not surprising that such a finely tuned system is subject to profound perturbation by therapeutic manipulation and hematologic malignancies. Any qualitative or quantitative defect in one of the components of the human defense system may predispose to infection, which remains a major Gary A . Noski11 (en), M A N A G E M E N T O F INFECTIOUS C O M P L I C A T I O N S IN C A N C E R PATIENTS. 0 1998. Kl~iwerAcadenllc Ptrblrshr~rs,Boston. All rights reserved.

THROMBOCYTES

MUCOUS MEMBRANES

MACROPHAGE

COMMENSAL FLORA

CELLULAR IMMUNITY

RGAN FUNCTION Figure I. Normal defense systems.

cause of morbidity and mortality among patients undergoing treatment for malignancy. However, isolated deficiencies are rarely encountered because malfunction of one part of the system exerts an impact on several other parts. Moreover, therapeutic interventions and the underlying disease conspire to afflict a range of defense mechanisms. The effects of the various different noxious events that occur while treating malignancy differ in severity as well as in primary targets (Figure 2). To complicate things further, hazardous events do not remain static but rather exert their impact dynamically as the degree of disturbance varies with time during or after a course of treatment (Figure 3). The human defense system is capable of coping with a tremendous number of insults before it finally begins to show the first signs of surrender. It should therefore be obvious to physicians that their activities put the entire defense system of patients at considerable risk. At the very least the patient should receive proper instruction on the prevention of infection and optimal hygiene should be maintained at all times. Physicians also have to select the most appropriate treatment regimen, thereby avoiding potentially dangerous medications, and to limit invasive procedures to those that are absolutely warranted. This complex interaction between host defenses and therapeutic modalities has a profound effect on patient outcome.

I SKIN PENETRATION:1

Figure 2. Factors influencing the human defense systems.

2. Basic clinical condition and organ function

2.1 Nutritional status Weight loss correlates inversely with survival in patients with cancer. This occurs whether or not intensive treatment is given because the integrity of host defenses can be endangered by the catabolic state induced by cachexia and malnutrition, resulting in a quantitatively deficient intake of calories and protein, with insufficient vitamin levels and trace metal concentrations [2,3]. Cachexia will be exacerbated by anorexia, chemotherapy-induced nausea and vomiting, gastrointestinal obstructions, and metabolic derangements. The final extent of the damage to the defense system depends upon the degree of cachexia and malnutrition. This may result in delayed wound healing, mucosal atrophy with a decrease in the secretions of lysozyme and secretory IgA, as well as impairment of both the classical and alternative complement pathways. A deficiency of vitamin A may also have a detrimental effect on the cellular immune system [4]. Furthermore, deficiencies of trace elements may undermine host defense in compromised patients. Zinc deficiency, as has been observed during total parenteral nutrition, generates a disturbed function of phagocytes and T cells,

CELLULAR IMMUNITY

HUMORAL IMMUNITY Figzlr-e 3. Evolution of impairment of defense systems after treatment for malignancy.

which can be neutralized by the addition of this mineral [5]. The in vitro microbicidal capacity of neutrophils and T-lymphocyte function are reduced in patients with iron deficiency, but it is uncertain whether this has any clinical significance. On the hand, it is well known that iron overload, a conceivable consequence of multiple blood transfusions, may lead to an increased susceptibility to infection, which is possibly related to a direct interaction between the iron available and the fungus Mucor. A phosphate deficit, which may occur during episodes of starvation and insufficient parenteral nutrition, is associated with a decrease in the chemotactic, phagocytic, and microbicidal functions of granulocytes in vitro, and clinically with bacterial and fungal infections [61. In elderly patients, the atrophy and dryness of the skin and mucosal membranes that occurs with age may lead to an increased susceptibility to infections. In addition, the primary and secondary humoral responses, as well as the oxidative metabolism of neutrophils and T-cell functions, decline with age, but the exact role of these regularly found abnormalities with regard to susceptibility to infection is unclear [7]. Concomitant chronic illnesses enhance the risk of infection in many patients. Even mild graft-versus-host disease is deleterious to the integument [8], and patients with a pre-existing immune disturbance, such as HIV infection or

a congenital immunodeficiency syndrome, are placed in double jeopardy. Much more common, however, is the detrimental effects of smoking, particularly in patients with primary lung tumors, due to colonization of their airways with virulent microorganisms and impaired clearance of secretions

PI.

Patients with poorly controlled diabetes mellitus are more likely to develop wound infections after all kinds of skin penetrations, and they frequently suffer from concurrent vascular disease and neuropathy. High concentrations of glucose in the urine and oral secretions promotes colonization by Candida species and other pathogens [lo]. Diabetes mellitus has been associated with notorious infections, such as rhinocerebral mucormycosis and malignant external otitis [Ill, which is not difficult to explain in view of several other aberrations that are associated with diabetes, such as impaired opsonization and decreased chemotactic activity of granulocytes and monocytes. Reduction of phagocytic adherence and defective phagocytosis, as well as bactericidal function of granulocytes, have been shown during episodes with high glucose concentrations and a low pH, putatively due to an impaired glucose metabolism of the phagocytes. A remarkable observation in this context is the relation between myeloperoxidase deficiency and serious fungal infections in patients with diabetes [12]. 2.2 Physiological and psychological status Psychological stress is thought to suppress host defense mechanisms. This general assumption has been corroborated by the observations that psychological stress has a negative influence on the function of T cells and NK cells. Indeed, stress appears to be connected with an increased risk of acute viral respiratory illness, a risk that is related to the amount of stress. This is most likely mediated by endogenous opioids, hormones from the hypothalamicpituitary-adrenal axis, catecholamines, and cytokines [13]. Tumors themselves may also predispose to infection by local organ dysfunction. In patients with solid tumors, obstruction of natural passages can lead to inadequate drainage of secretory or excretory fluids from nasal sinuses, bronchi, and bile ducts. Furthermore, tissue invasion may create connections between normally sterile spaces and the environment through disruption of epithelial surfaces. Examples include perforation of the esophagus by mediastinal tumors, invasive gynecological malignancies with local pelvic abscesses caused by gram-negative rods and anaerobes, skin ulcerations with cellulitis and deep soft-tissue infections, and invasion of the bowel wall by tumors of the lower gastrointestinal tract, resulting in bacteremia. Localizations in the central nervous system, spinal cord compression, and paraneoplastic neuropathy are associated with an increased risk of infection due to lethargy and, for instance, a diminished ability to cough and swallow, and incomplete emptying of the bladder [9]. Of course, hematologic malignancies are notorious for infectious com-

plications because the neoplasm resides within the immune system itself and interferes directly and indirectly with its function. In patients undergoing splenectomy, the risk that they will develop overwhelming sepsis at some time during their life is approximately 5 % . Encapsulated bacteria such as Streptococcus pneumoniae and Haemophilus in.uenzne are the prevalent pathogens, but Neisseria meningitidis and staphylococci are occasionally encountered [14]. Several factors might explain this well-established increased susceptibility to microbial infection. Encapsulated bacteria are able to elude phagocytosis because specific opsonizing antibodies are necessary for efficient phagocytosis. Furthermore, a reduced level of the complement factor properdin, which may lead to suboptimal opsonization, and a decrease in functional tuftsin have both been demonstrated after splenectomy [I].The spleen is the principal organ for eliminating particles that are not opsonized, and it is left to the macrophages that occupy strategic positions within the organ to remove them. The primary immunoglobulin response also takes places in the spleen, and low levels of circulating IgM have been observed after splenectomy in children. Because of the risk of pneumococcal infection after splenectomy, immunization with a polyvalent pneumococcal vaccine is recommended, preferably prior to splenectomy to ensure a better immune response during later life. However, the protection from vaccination is probably limited to several years, and, although vaccination has been shown to be effective, infection still may occur [15]. Therefore, for children it is recommended that vaccination be supplemented by antibiotic prophylaxis. In splenectomized adults suffering from a hematologic malignancy, patient-initiated treatment with oral amoxicillin at the onset of fever should be considered because the response to vaccines is usually suboptimal in patients with pre-existing immune deficiencies.

3. Integument and commensal microflora The integument comprises the skin, respiratory tract - including the nasal cavity, ears, and conjunctiva - the alimentary tract, and the genitourinary tract (Figure 4) and provides the first line of defense against microbial invasion. In physical terms the only difference between the skin and the other parts of the integument is that it is dry, whereas the others are bathed in mucins and therefore continually moist. Thus, while both surfaces are normally colonized with a variety of microorganisms, including many different genera of bacteria and yeasts, the range and number of species and the biomass associated with mucosal surfaces is much greater than those of the skin. However, the resident microbial flora of each surface play an integral role in helping maintain the function and integrity of these first lines of defense. Moreover, when intact and healthy, both the mucosa and skin are capable of resisting colonization with foreign or allochthonous organisms found in the immediate environment and

NASAL SECRETIONS 10%organlsmdrn DENTAL PLAQUE 10'' organism9 SALIVAlO' organlsrnstrnL

SKIN (axilla 8 groin) 10" organ~smskm~ SKIN (other sites) lo3 organ~srnslcrn; ONJUNCTIVA
LUNGS stcnlc PANCREAS stenre STOMACH to3 organlsmdrnL

KIDNEYS stenle

LIVER stenle BLADDER
GINAL SECRETIONS 109arganismdg

Figure 4. Body surfaces and their resident microbial flora. The integument comprises the skin, respiratory tract - including the nasal cavity, ears, and conjunctiva - the alimentary tract, and the genitourinary tract. and provides the first line of defense against microbial invasion. These body surfaces are normally colonized with a variety of microorganisms, including many different genera of bacteria and yeasts, but the range. number of species, and microbial biomass associated with the mucosal surfaces of the alimentary tract far exceed those of the skin.

maintaining an ecological balance within the indigenous microbial flora as a result of intimate host-parasite interactions. 3.1 Skin

The skin of an adult has an estimated surface area of 1.5-2.3m2and possesses features that are inimical to microbial invasion, provided it remains healthy and intact. The cells are composed of keratin and resemble loose paving stones. They are joined together by desmosomes and are continually sloughed off during desquamation so that adherent bacteria are also lost. This rapid cell turnover occurs every 2 weeks and helps to limit the opportunities for transient organisms to establish residence. Resident flora also have to continually re-establish themselves and do so because they are able to attach quickly. In ecological terms, the skin presents an arid milieu even though its humidity exceeds 90% because there is, in biological terms, very little water available for microorganisms due to the production of sebum, which is composed of lipids including triglycerides, long chain fatty acids, wax, and cholesterol esters, squalene, and other lipophilic substances. This oily, parched environment is particularly hostile to the establishment of microbial settlements involving gram-negative bacteria, which are vulnerable to desiccation and require an aqueous environment for survival. Moreover, only those microorganisms that elaborate lipases are capable of acquiring carbon from these

Table I . Microbial residents of the normal skin -

-

--

Major group

Genus

Opportunistic pathogens

Gram-positive cocci

Staphylococcus spp. Micrococclls spp. Corynebncterium spp. Brevibacterizlnz spp. Propionibncreriz~nrspp. Acinetobacter spp. Pity rosporum spp. Cnndidn spp.

S. epidermidis

Gram-positive bacilli

Yeasts

C. jeikeium

A . baumnnii C. parapsilosis

lipids. The skin also forms an acid mantle, having a pH of 5.0-6.0, and its surface temperature is, on average, about 5°C lower than that of the core body temperature. Sweat and transepidermal water loss introduce water to the surface of the skin, which contains a range of potential carbon and nitrogen sources, including lactate and pyruvate, glucose, amino acids, creatinine, urea, and urocanic acid. However, the resident flora modulate the microecology of the skin by releasing fatty acids, such as oleic, stearic, and palmitic acids, from the sebaceous secretions, as well as short-chain lactic and propionic acids, resulting in a hydrophobic, acid milieu, which counterbalances the nutrient potential of sweat. IgA is also secreted in sweat. Thus the range of organisms that are able to reside on the skin is strictly limited to a few, mainly gram-positive bacteria, such as various members of the coagulase-negative staphylococci, particularly Staphylococcus epidermidis, Corynebacterium jeikeium and other coryneforms, Propionibacterium spp., and certain yeasts (e.g ., Pity rosporum spp.) that can withstand these hostile conditions and compete successfully for binding sites and nutrients to establish a permanent and intimate attachment to the epidermis [16]. Many of the resident bacteria also elaborate toxins that inhibit other closely related microorganisms, allowing individual species to retain their foothold and consolidate their territory. Resident species also grow as biofilms, which consist of microcolonies enmeshed in a glycocalyx, rather than the planktonic growth found in laboratory cultures. Thus each microbial consortium possesses a boundary and exists as a distinct unit separate from its neighbors, rather like the various plants found in a garden. 3.2 Erosion of the skin integument The effectiveness of the skin as a defense barrier can be eroded in a variety of ways. Topical antibiotics and those secreted in sweat will disturb the balance within the resident commensal flora, leaving the surface vulnerable to colonization by exogenous potential pathogens such as the gram-negative bacteria. Antibiotics will also exert selective pressure on the resident flora, causing

resistance to emerge, as has been observed during treatment with ciprofloxacin because the drug is secreted with sweat [17]. Chemotherapy and irradiation can bring about radical changes in the normal skin by interrupting normal cell replacement, resulting in hair loss, dryness, and loss of sweat production. In addition, steroids also can exert a profound effect on sebum secretions. When the skin is broken, the release of fibronectin is thought to assist colonization with Staphylococcus aureus, and other changes facilitate colonization with gram-negative bacilli such as Acinetobacter baumanii and enterobacteria. Cutaneous infection results from the loss of integrity and reduced local immunity of the skin as well as disturbances within the resident flora. Abraded skin and the associated exudates and minor breaches in the integument can lead to local infection as well as provide a reservoir that assists further spread to other body surfaces, including the oral cavity. When the balance is lost between the host defenses and resident commensal flora around the hair follicles, they can become inflamed and necrotic, forming a potential nidus of infection.

3.3 Impact of intravenous catheters o n the integrity of the skin Cutaneous infections in the immunocompromised patient can also develop from needle punctures, but the insertion of catheters provides the single most effective means of breaching the natural protective barrier of the skin and creating access for microorganisms to both the stratum corneum and the bloodstream. Intravenous catheters are often essential for the successful management of immunocompromised patients because they provide ready and safe access to the bloodstream with minimal trauma and discomfort. With good technique, the complications resulting from inserting indwelling catheters, such as the Hickman device, are minimal. However, the devices frequently become colonized with organisms that have a predilection for hydrophobic surfaces and a proclivity to form biofilms. Indeed, intraluminal colonization by coagulase-negative staphylococci may be unavoidable because these organisms can be recovered from almost all devices [18,19]. These staphylococci are commonly resistant to tobramycin, trimethoprim, and methicillin, and may also be resistant to ciprofloxacin [20]. Unless the catheter ends in an implanted port, skin commensals have open access directly into the bloodstream depending upon how frequently it is used [21]. This is thought to be the common route of infection by other, normally minor, residents of the skin flora such as Corynebacteriurn urealyticurn that are given a selective advantage by antimicrobial agents [22]. Unusual saprophytic bacteria, such as Comonzonas acidovorans [23], Ochrobactrum anthropi, and Agrobacteriurn spp. [24], as well as the more common and familiar species of gram-negative bacilli, including Pseudomonas aeruginosa, that are found in aqueous environments, also gain access this way and have the ability to attach to the silicone used to make catheters and to grow in biofilms, producing copious amounts of slime. Moreover, when such organisms colonize the

device, antibiotic treatment rarely achieves a complete cure, necessitating removal of the catheter. Infections related to the external surface of the catheter, particularly exit-site infections and tunnel infections, occur much less frequently than does intraluminal colonization and tend to involve other resident grampositive bacteria, including Corynebacterium jeikeium and Stomatococcus mucilaginosus, and occasionally some gram-negative bacilli (Acinetobacter spp. and Stenotrophornonas (Xanthomonas) maltophilia), which have established colonization beforehand. The catheter is also a portal of entry for fungi, including Candida parapsilosis [25] and other Candida [26,27] and for molds such as Aspergillus and Mucor. Infections associated with intravenous catheters more often represent colonization of the lumen, which gives rise to bacteremia. While many, if not most, cases of intraluminal colonization do not represent a threat to the patient, metastatic infection can occur when S. aureus, Candida spp., and gramnegative bacilli are involved. Persistent colonization with coagulase-negative staphylococci or Corynebacterium jeikeium often leads to repeated episodes of bacteremia, which may only be noticed when the patient experiences shaking chills, tachycardia, hypotension, and peripheral cyanosis after the line is manipulated, the classic manifestations of intraluminal colonization. Colonization of the external surface of a catheter may lead to infection of the catheter-tissue interface at the exit site, which occasionally extends along the track, causing cellulitis when a tunneled device such as the Hickman catheter is involved or phlebitis. 4. Upper respiratory, alimentary, and genitourinary tracts The surface area of the upper respiratory, alimentary, and genitourinary tracts available for microbial colonization is greater than that afforded by the skin because of the folds, crypts, and villi. The surfaces of each anatomical region are also very different, ranging from the hard enamel of the teeth to the microvilli of the bowel. Extreme changes in the local environment also occur, ranging from the neutrality of the mouth to the acidity of the stomach. Moreover, both extremes of the alimentary tract are anaerobic and together contain approximately 10"microbial cells, representing some 700 different species. The relationship between these commensals and the host are both complex and poorly understood. Nevertheless, some generalizations are possible and useful in understanding how the mucosal surfaces play their part as a first line of defense. Two principal physical host factors influence the microbial ecology of the mucosal surfaces. Dilution of the inoculum is achieved by sneezing and coughing of microbes trapped in mucus, flushing of the mouth and esophagus by saliva, micturition, and peristalsis of the intestines. Acidity plays a crucial role both in disinfecting the stomach and in regulating the microbial milieu of the

vagina. The upper respiratory, alimentary, and genitourinary tracts are essentially composed of epithelial cells interspersed by cells that produce the glycoproteins known as mucins. These hydrophilic substances perform various functions, including lubrication, waterproofing, and sudden changes in osmotic pressure [28]. They also contain inhibitory substances, such as lactoferrin, lysozyme, and peroxidase, as well as secretory IgA. Mucins also appear to interfere with adherence of foreign bacteria to epithelial cells and prevent access of antigens to antibodies while allowing the biofilm or glycocalyx formed by resident bacteria to blend or fuse so that the bacteria can form a more intimate contact with the epithelial cells. The resident flora probably play a crucial role in maintaining the integrity of this part of the integument. They compete with one another for sites of attachment and nutrients as they continually modulate the microecology. Their activity is sometimes beneficial to the human host, such as the synthesis of vitamin B,,, but can also be harmful, as in the production of carcinogens from nitrates. On the whole, however, the microflora are commensals exhibiting stable symbiosis. The human host is probably immunologically tolerant to all members of resident flora because fewer than 50h of the genera have ever been implicated as opportunistic pathogens, even in the most profoundly immunosuppressed individuals. For example, even when translocation into the bloodstream occurs, the resident bacteria are poorly adapted to the environment within the body proper and only rarely establish an intracorporeal infective process. 4.1 Resident microorganisms of the upper respiratory tract and oral cavity

Although there are over 40 different bacterial species that reside on the epithelia of the upper respiratory tract and oral cavity, very few are capable of successful translocation into the bloodstream and fewer still of establishing disseminated infection (Table 2). In fact, during the course of a normal day, the acts of chewing and brushing the teeth may lead to transient bacteremia due to viridans streptococci; hence, their involvement in bacterial endocarditis. Most of the species capable of causing infection do so only to a limited extent, and primarily in individuals with poor oral hygiene. 4.2 Effect of chemotherapy and irradiation on the oral cavity

Cytotoxic chemotherapy and irradiation interrupt cell division, leading to breakdown in the integrity of the oral mucosa. The production of saliva may also be impaired, leading to a dry mouth and, if mucin is produced, may be extremely viscous and difficult to either swallow or expectorate. Periodontal disease may be exacerbated and minor oral cuts and abrasions may become inflamed or ulcerated. The nonkeratinized surfaces of the mouth, including the dorsal surface of the tongue, the roof of the mouth, and the buccal mucosa, may become erythematous, inflamed, and edematous, limiting the intake of

Table 2. Resident flora of the upper respiratory tract and oral cavity Major group

Genus

Gram-positive cocci

Micrococcus spp. Staphylococcus spp. Stomatococczts spp. Streptococc~lsspp, nonhemolytic group Streptococc~isspp. viridans group Actinomyces spp. Arachnia spp. Bacillus spp. Bacterionema spp. Bifidobacterilrnz spp. Closrridiiim spp. Corynebacterium spp. Ezlbacterium spp. Lactobacillus spp. Propionibacteritolz spp. Rothia spp. Moraxella spp. Neisseria spp. Veillonella spp. Actinobacill~lsspp. Capnocyrophaga spp. Eikonella spp. Fzisobacteriurn spp. Haernophilus spp. Leptotrichia spp. Prevotella spp. Selenomonas spp. Wolinella spp. Treponenza spp. Mycoplasma spp. Cnndidn spp.

Gram-positive bacilli

Gram-negative cocci

Gram-negative bacilli

Spirochetes Mycoplasma Yeasts

Opportunistic pathogens S. epidermidis S. nz ~~cilaginosus S. milleri S. oralis, S. nzitis A. israelii

C. sporogenes

A, actinornyceremcon~itans C. ochracea E. corrodens F. nucleatum H. parainfl~lenzae L. buccalis P. melanogeniczls

M. salivarilim C. albicans

both solids and liquids [29]. This phenomenon is now generally referred to as mucositis, although some prefer the older term, stomatitis. Thus, when mucositis is present, the mouth loses its normal ability to dilute foreign bacteria. Mucositis also occurs at the same time as other manifestations of toxicity, particularly bone marrow depletion and gut toxicity, manifested by nausea, vomiting, and diarrhea. Moreover, mucosal changes normally progress to a peak severity and coincide with the nadir of bone marrow aplasia and then begin to recover as hematopoiesis returns (Figure 5) [29-321. Exposing oral commensal flora to the antimicrobial agents used for prophylaxis and local antisepsis will inevitably select for more resistant species. Very susceptible bacteria, such as the oral Neisseria spp., will be suppressed by a wide range of antimicrobials, whereas others that are marginally susceptible to agents frequently used, such as co-trimoxazole, penicillin, and fiuoroquinolones, will thrive. This partly explains why the viridans strepto-

Days after starting chemotherapy

Figure 5. Mucositis. Mucositis and bone marrow aplasia, leading to profound neutropenia, are both manifestations of toxicity frequently occurring together with gut toxicity manifested by nausea, vomiting, and diarrhea. As with neutropenia, mucosal changes normally progress to a peak severity, which coincides with the nadir of bone marrow aplasia and then begins to recover as hematopoiesis returns.

cocci have become one of the most frequent causes of bacteremia in neutropenic patients who have undergone cytostatic chemotherapy for leukemia or who have received a bone marrow transplant [33], although the chemotherapeutic agents may be a more important factor, especially when it induces severe mucosal damage [34]. One particular viridans streptococcus species, Streptococc~ismitis, many of which are actually S. oralis (formerly S.sanguis 11) [35], is causing concern because its appearance in the bloodstream following treatment with high-dose cytarabine is associated with the sepsis syndrome and the adult respiratory distress syndrome (ARDS). Bacteremia due to other unusual oral commensals, such as Stomatococcus mucilaginosus, Capnocytophaga spp,, and Leptotrichia buccalis, are likely to be selected for by quinolones because they are also only marginally susceptible. In addition, gingivitis as the source of bacteremia due to S. epidermidis has been reported [36]. Similar risk factors are associated with the development of bacteremia following chemotherapy due to members of the

Streptococcus milleri group [37]. The chlorhexidine mouthwashes used to minimize infective complications arising from the oral toxicity induced by chemotherapy also influence the microflora [30,38,39]. The oral flora may also change as a direct result of chemotherapy [40], and it is likely that more intensive conditioning regimens will aggravate mucositis, leading to a commensurate increase in the number of unusual bacteria. Use of the growth factors, granulocyte-colony stimulating factor (G-CSF) and granulocyte macrophage-colony stimulating factor (GM-CSF) [41,42] does not appear to have any influence on mucositis [43]. Therefore, neutropenic patients will continue to experience varying degrees of mucosal damage depending upon the nature of their therapy because some chemotherapeutic agents (e.g., methotrexate, high doses of cytarabine and melphalan) produce extensive damage, often with the production of thick mucus [44,45]. Mucositis can also be particularly severe when anthracyclines are combined with total body irradiation and cyclophosphamide to condition patients for an allogeneic transplant [46]. Fortunately, the morbidity associated with gram-positive infections is usually mild and the attributable mortality is negligible [47-521. The lung appears to be particularly vulnerable to damage by cytotoxic chemotherapy and irradiation, and is exquisitely susceptible to infection. Immunopathological reactions mediated by the pulmonary macrophages that survive chemotherapy can lead to various other syndromes, including respiratory distress. Pulmonary hemorrhage as a result of profound thrombocytopenia further imperils the lung, increasing the risk of infection. However, the risk of invasion and dissemination is high when the integrity of the mucosa is impaired and the ecology of resident flora is disturbed, and an exogenous microorganism such as a gram-negative bacillus or other potential pathogen establishes colonization. Resident flora such as Candida spp. can result in superficial infection, often as a consequence of reactivation of herpes simplex virus [53,54]. Clinically, the presence of pseudomembranes over the ulcerated tissue can initiate local invasion and progressive spread to the esophagus and gastrointestinal tract, resulting in disseminated candidiasis. Aspiration and inhalation of spores and hyphal elements of Aspergillus spp. and other molds permits colonization of the sinuses and bronchial tree, which may extend into the alveolar spaces, resulting in invasive disease that is often fatal. 5. Microflora of the intestinal tract

The alimentary tract is the major reservoir of gram-negative bacilli, which are either endogenous (e.g., Escherichia coli) or have been acquired by ingestion (e.g., Klebsiella pneumoniae and Pseudomonas aeruginosa) [55-571. Normally, the alimentary tract flora contains in excess of 1014microorganisms, amounting to several grams, but only very few species are capable of establishing infec-

Table 3. Resident flora of the lower alimentary tract Major group

Genus

Opportunistic pathogens

Gram-negative anaerobic bacilli

Bacteroides spp. Desr-llfomonas spp. Leptotrichia spp. Fusobacterium spp. Butyrvibrio spp. Sucinimonas spp. Vibrio spp. Escherichia spp. Citrobacter spp. Klebsiella spp. Enterobncter spp. Morgnnella spp. Proteus spp. Lnctobacillus spp.

B. fragilis

Gram-negative facultatively anaerobic bacilli

Gram-positive facultatively anaerobic bacilli Gram-positive anaerobic bacilli

Gram-positive facultatively anaerobic cocci

Gram-positive anaerobic cocci

Gram-positive anaerobic cocci

Yeasts

BGdobacteriurn spp. Clostridiurn spp. ELL bacterium spp. Lachnospira spp. Propionibacterium spp. Enterococcus spp. Staphylococcus spp. Streptococcus spp. Peptococcus spp. Peptostreptococcus spp. Acidaminococcus spp. Megasphaera spp. Rurninococcus spp. Sarcina spp. Veillonella spp. Coprococcus spp. Gemella spp. Candicln spp.

L. biiccalis F, nucleatum

E. coli C. freundii K. pneumoniae E, cloacae M. rnorgnnii P. rnirabilis

C, tertiurn, C. dificile, C. sporogenes

P. acne E. faecalis, E. faecium S. epidermidis S.rnilleri, S.rnitis, S. oralis, S, hovis

C. nlbicnns, C. glabrntn, C. krusei, C. lusitnnia

tion, even in the most profoundly immunosuppressed patient. Most of the microbial flora is densely distributed around the surfaces of the oral cavity and the large bowel, where scores of different microorganisms, including spirochetes, spore formers, bacilli, and cocci, compete for the available surfaces and nutrients. Anaerobes predominate and play a crucial role in maintaining a healthy commensal flora, preventing the establishment of exogenous or allochthonous organisms, which is known as colonization resistance [58,59]. The integrity of the mucosa, the production of saliva and mucus, peristalsis, gastric pH, bile acids, digestive enzymes, and the levels of secretory IgA also play an important role in maintaining colonization resistance [60].

5.1 Impact o f antimicrobial agents o n colonization resistance o f the alimentary tract Exposure to antimicrobial agents is one of the most effective means for destroying colonization resistance, as is manifest by fungal overgrowth, and increases in the enterococcal populations [61-631. The most likely contributors to colonization resistance, the gram-positive nonsporing, lactic acid-producing bacilli, particularly bifidobacteria, are particularly susceptible to antibiotics known to impair colonization resistance, including the penicillins, rifamycin, clindamycin, erythromycin, bacitracin, and vancomycin [60,64-681. Some cephalosporins are also detrimental to colonization resistance, whereas other (3-lactams (e.g., meropenem) and the quinolones have been declared "friendly" [61,68-73). Some drugs such as aztreonam and imipenem only appear "friendly" because they are inactivated by feces [63,74], whereas under the circumstance of diarrhea, parenteral feeding, and gut toxicity, normal stool is no longer produced so these agents may remain sufficiently active to destroy what remains of the colonization resistance. Initially co-trimoxazole was thought to be neutral [60,68,75-791, but recent evidence suggests otherwise [80]. Individual antibiotics that appear to spare colonization resistance, such as ceftazidime and piperacillin, might have a marked impact when given in combination, leading to an increase in both Clostridium difficile as well as yeasts [81].These bacteria can lead to enterocolitis, which responds to treatment with metronidazole or oral vancomycin, but the latter may select for resistant bacteria such as Enterococcus faecium and Lactobacill~tsrhamnosz~s[82]. The widespread use of fluoroquinolones for prophylaxis has led to the emergence of resistance among the Escherichia coli, which are indigenous to the bowel [83-861.

5.2 Effect of chemotherapy and irradiation o n the intestinal tract One of the most important consequence of the loss of colonization resistance is that cell surfaces become vacant, allowing some exogenous bacteria such as P. aeruginosn to establish residence, leading to chronic colonization, with the attendant risks of invasion and systemic dissemination. The ecology of the bowel flora is also altered markedly by diarrhea induced by treatment with certain chemotherapy [87], graft-versus-host disease [88], and total body irradiation [89]. When severe chemotherapy-induced mucositis extends to the cecum, typhlitis or neutropenic enterocolitis, can occur and the recovery of Clostridium septicum from the blood confirms the diagnosis 1901. Gut permeability also increases following conditioning therapy for bone marrow transplant [91]. Agents used either for the treatment of neoplasms or supportive care may even exert an influence on gut and oral flora, either alone or in combination. Some chemotherapeutic agents have been shown to have antibacterial activity and even to enhance the effects of antimicrobial agents [92-971. The

antifungal, miconazole, is also inhibitory to gram-positive bacteria [98]. Gut motility is reduced during parenteral nutrition due to the low amounts of fiber and reduced microbial biomass, which result in dilute feces. When the gut fails to function normally, the protective "anaerobic wallpaper" may still be intact but will be unusually fragile to the effect of antimicrobial agents. Thus, unless placed in a degree of isolation and supplied with Iow-microbial content diets, patients will be vulnerable to acquiring other gram-negative bacilli from the environment [55-571.

6. Platelets The protective role of platelets [99] in normal individuals is often underestimated but becomes obvious during treatment for a malignant disease (see Figure 2). Thrombocytopenia is an almost inevitable repercussion of intensive chemotherapy and irradiation, but a decreased function of thrombocytes is a similar matter of concern. Such a thrombocytopathy is either disease related or caused by concurrent medication (Table 4). The consequences for both an increased susceptibility to infection and a decreased capacity to repair damaged tissues can be considerable and may have an impact on the eventual outcome of a treatment episode. Thrombocytopenia also appears to be an independent risk factor for bacteremia [loo], and the incidence of major hemorrhages at autopsy of patients who die with or from an infection is striking.

7. Granulocytopenia Under normal circumstances the proliferation of neutrophil precursors is regulated by hematopoietic growth factors such as interleukin-3, GM-CSF, and G-CSF. Starting from a pluripotent stem cell, it takes approximately 6 days to

Table 4. Causes and sequelae of thrombocytopenia and thrombocytopathy Causes of thrombocytopenia Disease related Treatment related Causes of thrombocytopathy Disease related Treatment related Hazardous sequelae

Leukemia and lymphoma, bone marrow metastasis Chemotherapy, radiotherapy Leukemia and myeloma, renal insufficiency Chemotherapy, (3-lactam antibiotics, antiinflammatory drugs, antihistamines, heparin Hemorrhagic lesions facilitate growth of microorganisms and interfere with organ function Decrease of platelet-derived growth factor, epidermal cell growth factor, endothelial cell growth factor, fibronectin (diminished adhesion), P-selectin (diminished transmigration)

form metamyelocytes by sequential divisions and another 6 days to mature into polymorphonuclear granulocytes [loll. Approximately 90% of the total population of neutrophils resides in the bone marrow, only to be released into the circulation upon an inflammatory stimulus. Neutrophils that enter the bloodstream are distributed over two compartments of equal size in dynamic equilibrium: a free circulating pool of neutrophils and the marginating pool, consisting of neutrophils that adhere loosely to the vascular endothelium. The size of these respective pools is under the influence of several factors. Adherence of neutrophils to endothelial cells is mediated by a number of adhesion molecules on neutrophils, which are induced by factors such as complement factor C5a, which acts as a ligand. Likewise, there is a whole series of adhesion molecules on the endothelial cells themselves, with cytokines such as interleukin-1 and tumor necrosis factor-a being important inducers of these molecules [I]. Other inflammatory impulses and glucocorticosteroids are also potent inhibitors of margination. Circulating neutrophils disappear after approximately 6 hours in blood, whereas they survive 1-3 days in tissues. During an acute inflammatory reaction, an increase in neutrophils, sometimes accompanied by eosinophils and followed by macrophages, can be seen at the site of inflammation. The formation of this inflammatory exudate is the result of activation of several humoral factors, such as cytokines, prostaglandins, and complement, which enhance the blood flow and increase vascular permeability. This occurs in conjunction with chemotactic activity, which results from other soluble factors, especially C5a, leukotriene B, interleukin-8, and bacterial products. In the peripheral blood, granulocytosis evolves as a consequence of the release of the marrow reserve and increased granulocytopoiesis on stimulation by factors such as interleukin-1. However, the mere presence of granulocytes at the site of an infection is meaningless if they are not able to execute their normal functions. Phagocytosis, a Fc- and C3b receptor-mediated process with IgG1, IgG3, and C3b as ligands or opsonins, results in the uptake of particles larger than 1pm via pseudopods until they enclose in a vacuole (phagosome). The rate of ingestion by neutrophils is impressive in comparison with that of other phagocytes. As soon as the particles, with or without opsonins, make contact with the cell membrane of a granulocyte, oxidases in the membrane are triggered to activate oxygen-dependent microbicidal mechanisms, and superoxide, hydrogen peroxide, and hydroxyl radicals are formed. During and after ingestion, the lysosomes, which are microscopically visible as azurophilic granules, fuse with the phagosome and pour their digestive enzymes into the vacuole, a process known as degranulation. One of these lysosomal enzymes, myeloperoxidase, triggers the reaction of H,O, with chloride, which results in the formation of hypochlorite, a potent microbicidal product. Usually this operation of phagocytosis and intracellular killing of microorganisms is a suicidal act for the neutrophils, leaving the remainder for consumption and enzymatic digestion by the more powerful macrophages. However, even

macrophages may require cooperation with products from activated T lymphocytes for the optimal killing of some microorganisms. The proliferation and maturation of eosinophilic precursors is under the control of interleukin-3, GM-CSF, and interleukin-5, and has a time span similar to that of neutrophils [102], whereas survival in the tissues appears to be considerably longer. Eosinophils are able to kill several parasites, largely by means of an extracellular process mediated by IgE and, probably, complement. 7.1 Impairment of granulocyte function Virtually all cytotoxic drugs used in the treatment of malignant diseases have a dose-dependent deleterious effect on the proliferation of normal hematopoietic progenitor cells, including those of the myeloid series. After destruction of the mitotic pool by one or more cytotoxic compounds and depletion of the marrow pool reserve, granulocytopenia with a duration of days or weeks will ensue, particularly in the treatment of hematologic malignancies and following bone marrow transplant conditioning regimens. Likewise, therapeutic radiation may induce a clinically significant granulocytopenia, depending on dose rate, total dose given, irradiated area of the body, and field size. Total body irradiation, as used in bone marrow transplant procedures, is the most illustrative of the potential deleterious effects of irradiation. However, both chemotherapeutic drugs and irradiation do not only inhibit the proliferating cell pool, they also interfere with nonproliferating cells and their function (see Figure 2). In granulocytes this may result in decreased chemotaxis, diminished phagocytotic capacity, and defective intracellular killing. Glucocorticosteroids seem to enhance granulocytopoiesis and mobilize the marginal as well as marrow pool reserve, but these supposedly positive effects on the granulocytes are counterbalanced by numerous disadvantages. Indeed, these drugs restrain the accumulation of neutrophils at the site of inflammation through impaired migration - probably due to reduced adherent capacity of the granulocytes - and diminished chemotactic activity. Furthermore, they negatively influence phagocytosis and intracellular killing by neutrophils in a dose-dependent fashion, and are associated with a reduction in the number of eosinophils in the blood. Finally, many other drugs, including antibiotics, that are regularly used in cancer patients are known to interfere with the production and function of granulocytes, which also may lead to an increased susceptibility to infection. Although they usually occur simultaneously, any substantial reduction of the number of granulocytes or qualitative defect in the phagocytic process can, in fact, make the patient prone to recurrent bacterial and fungal infections. Granulocytopenia is probably the paramount factor responsible for the increased frequency of infection in cancer patients because only one fifth of the febrile episodes occur when they are not granulocytopenic. It has been shown that an inverse correlation exists between the number of circulating

neutrophils and lymphocytes, and the frequency of infection. Depending on the duration of neutropenia, the risk of a febrile episode varies between 30% and 80%. In a study by Bodey and coworkers [103], all patients with a neutrophil count of less than 100 per yL for more than 3 weeks developed an infectious complication, and the risk for secondary infections increases proportionally with the duration of granulocytopenia. Moreover, infectionrelated mortality increased with the duration of hospitalization and the number of days of granulocytopenia.

7.2 Diagnostic consequences of granulocytopenia It may be difficult to establish an unequivocal diagnosis of infection because the inflammatory response in patients without properly functioning granulocytes is muted, thereby obscuring the classic signs and symptoms of infection [104]. Of the episodes of fever associated with granulocytopenia, a definite infectious etiology can be established in about a quarter of cases on the basis of microbiological confirmation. Local infections, if detected at all, are frequently complicated by bacteremia, which accounts for more than 90% of culture-documented infections in cancer patients [105,106]. Microorganisms that cause a local infection or that colonize damaged mucosa or skin can easily gain access to the bloodstream in the absence of granulocytes. Between 40% and 70% of patients who become febrile while granulocytopenic have unexplained fevers, but nevertheless improve after treatment with broad-spectrum antibacterials, suggesting an occult bacterial source as the cause of fever. A small inoculum, sufficient to cause symptoms of infection in a patient with a defective defense mechanism, might be below the detection limit of standard blood culture techniques. Infectious complications usually arise insidiously in granulocytopenic patients with little or no inflammatory signs and without the formation of pus. Fever is often the only hallmark of a possible infection, but without prompt and appropriate treatment such infections may run a fulminant course in these patients (see Chapter 3). After bacteria, fungi are the next most common pathogens, especially in immunosuppressed patients who have prolonged and profound granulocytopenia. Autopsy evidence of significant fungal infections can be found in one half of these patients. Most of these infections are not diagnosed or treated antemortern, but they account for 20-30% of fatal infections in patients with acute leukemia [107-1091. Besides granulocytopenia, the use of pharmacological doses of corticosteroids and indwelling catheters may also foster the development of systemic fungal infection [I101 (see Chapter 6).

8. Cellular immunity and cytokines Whereas humoral immunity is primarily responsible for clearing extracellular bacteria, the cellular immune system serves to eliminate intracellular patho-

gens and virus-infected cells. Both antigen-specific and nonspecific cells contribute to the development of cellular immunity. Specific cells including T-helper (TH) and cytotoxic (T,) T cells and nonspecific cells include macrophages, neutrophils, and natural killer (NK) cells. Normal macrophages have a limited capacity for killing ingested microorganisms, and various organisms (e.g., Toxoplasma gondii, Pneumocystis carinii, Cryptococcus spp., Listevia spp., Salmonella spp., and Legionella spp.) are capable of surviving and replicating within the cell, unless the macrophage becomes activated. The activation process is complex and is primarily under the control of cytokines. Interferon (1FN)-y, produced by NK cells and TH cells, is the most important player in the field, but the presence of tumor necrosis factor (TNF)-a is also required. The activated macrophage is characterized by morphological changes, expression of certain antigens, increased oxygen consumption, and a marked upregulation of both oxygen-dependent and nitric oxide-based microbicidal mechanisms, enabling the cell to kill intracellular pathogens. The antigen-specific branch of cell-mediated immunity can be divided into two major categories. One category involves cytotoxic effector cells, which are able to lyse virus-infected or foreign lymphocytes and macrophages. The second category involves various subpopulations of TI, cells that mediate delayed-type hypersensitivity reactions. These T cells can be activated following antigen recognition only when the antigen is displayed together with major histocompatibility complex (MHC) on the surface of specialized cells called antigen-presenting cells (APC). These antigen-presenting cells, mainly rnacrophages and dendritic cells, phagocytose the microorganism and then express a part of the antigen on their membrane, enabling T,, cells to recognize the antigen (Figure 6). TNF-a and IL-12 are induced, and synergistically activate NK cells to produce 1FN-y. IFN-y serves to activate macrophage microbicidal systems, through induction of nitrate synthase and reactive oxygen metabolites. These new insights place greater stress on the pivotal role of the NK cell in the earliest phase of the immune response and imply that activation of rnacrophages is not necessarily T-cell mediated. Rather, T cells become involved at a later stage in the process after an average of 24 hours following contact with the antigen. At the level of the so-called naive, uncommitted THOlymphocyte, IL-12 induces the maturation of THO cells to the T , , phenotype, which produces IL-2 and IFN-y. Under other circumstances, such as allergic reactions or infection with certain parasites, the naive THOcells differentiate into a TI,, subset, which produces an entirely different repertoire of cytokines specifically tailored to respond to the character of the insult. The TH,derived IL-2 functions in an autocrine manner to amplify the population of cytokine-producing T cells (see Figure 6). Moreover, IL-2 potentiates the effect of IL-12 and TNFa in stimulating NK cells to produce IFN-y. Other cytokines that are produced by the activated TH, cell include IL-3 and GM-CSF, which stimulate the granulocyte-monocyte lineage. IFN-y and macrophage-derived IL-1 induce a

APC

IL-2

3

Figure 6. Development of cell-mediated immunity. (1) Microorganisms are phagocytosed by macrophages and antigen is presented. (2) Macrophages produce TNF-a and IL-12 to activate NK cells to produce IFN-g. (3) IFN-g activates microbicidal functions of macrophages. (4) IL-12 also induces naive T-helper (THO)cells to differentiate into T,, cells, secreting IL-2 and IFN-g. (5) In an autocrine fashion, IL-2 stimulates T cells and also activates NK cells to produce IFN-g. (6) Macrophages begin to produce IL-10, which inhibits TNF-a, IL-12, and IFN-g production in order to terminate the process of the cellular immune response after several days.

number of changes in nearby endothelial cells, facilitating extravasation of monocytes and other nonspecific inflammatory cells. Also, the expression of cellular adhesion molecules, including ICAM, VCAM, and ELAM, is increased, and IL-8 is secreted. Together, these reactions lead to an influx of neutrophils and monocytes to the site of infection. Somewhat later, with a peak occurring 24-48 hours after the infection, macrophages begin to produce IL-10, which acts as a negative feedback by inhibiting IFN-y production. This effect of IL-10 is accomplished by two mechanisms: inhibition of macrophage TNF-a and IL-12 production, and direct suppression of IFN-y production by NK cells. This downmodulation of the response occurs at the time that specific T cells begin to further modulate the cellular response to infection. The importance of cell-mediated immunity in protecting the host against various intracellular pathogens is evident from the opportunistic infections occurring in various groups of cancer patients. Patients treated with prednisone or other immunosuppressive agents that specifically affect cellmediated responses may become unable to cope with intracellular pathogens such as Listeria rnonocytogenes, Pnet~mocystiscarinii, or ToxopEasma gondii. The classic concept of Hodgkin's lymphoma is that of a disease associated with

impaired cellular immunity, although delayed hypersensitivity responses are intact in the majority of untreated patients [lll].No anergy has been found in patients with stage I disease, and the incidence of cellular immune defects increases with disease stage to affect only approximately 25% of patients with stage IV disease. The defect in T-cell-mediated immunity in patients with Hodgkin's disease is probably due to an excess of T suppressor cells, which leads to increased susceptibility to varicella-zoster virus (VZV) infections, tuberculosis, cryptococcosis and, in the United States, endemic mycoses such as histoplasmosis and coccidioidomycosis. Cytotoxic treatment of patients with lymphoma that leads to severe neutropenia may render these patients at risk for severe gram-negative and other fungal infections, thereby obscuring the classic association of both Hodgkin's and non-Hodgkin's lymphoma with defects in cellular immunity. Allogeneic bone marrow transplantation (BMT) brings about a long-lasting dysfunction of T and B cells, and the opportunistic infection due to these defects may become manifest long after transplantation and recovery from neutropenia. The most prominent example is VZV infection, which occurs in up to 50% of BMT recipients in some series. The median time of onset is 5 months after transplant. The spectrum of the disease ranges from primary infection (varicella) to localized or often disseminated herpes zoster. Prophylaxis with acyclovir is advocated in many centers for the first 9 months after transplant. Chronic graft-versus-host disease and the immunosuppressive agents used for its treatment further render these patients at risk for infection with VZV, Aspergillus spp. and Pneumocystis cnrinii.

9. Humoral immunity The humoral branch of the immune system involves interaction of B cells with antigen and their subsequent proliferation and differentiation into antibodysecreting plasma cells. An important difference in antigen recognition by T cells and B cells is that the latter can recognize an antigen, whereas T cells can only do so once the antigen has been phagocytosed and is presented on the surface of an antigen-presenting cell. In this way, the immune system is able to cope with invaders under a variety of different circumstances. The humoral system recognizes a plethora of bacterial or viral microorganisms as well as the soluble proteins they release. The cell-mediated system is suited to recognizing altered cells belonging to the "self", that is, infected phagocytes as well as cancer cells. Upon challenge with an antigen, immunoglobulins are produced by the humoral branch of the system and bind to the antigen. IgM is secreted early and during differentiation. Plasma cells then become committed to produce the other classes of immunoglobulin, such as IgG, IgA, IgE, and IgD [I]. The specific functions of IgG and IgM include not only neutralization of the antigen, but also complement activation and opsonization. Secretory IgA, which is found on mucosal surfaces, is not an opsonin but it inhibits the

motility of bacteria, neutralizes their toxins, and prevents their adherence to epithelial cells. Circulating IgA probably plays only a minor role in host defense. Tuftsin is a small peptide that binds to the Fc portion of IgG and becomes involved in activating phagocytes after being released from the immunoglobulin. Two enzymes are required for this process: leukokinase, which is bound to the membranes of neutrophils and macrophages, and tuftsin endocarboxypeptidase, which is produced in the spleen. The decreased availability of functional tuftsin is one of the mechanisms that contribute to the increased risk of severe infections after a staging splenectomy in patients with malignancies. Other mechanisms include the clearance of particles that have not been opsonized by complement. The spleen plays an important role in the humoral immune response as the primary immunoglobulin response takes place there, as shown by the low concentrations of IgM found after splenectomy. Reduced concentrations of the complement factor properdin have also been found, leading to suboptimal opsonization. Functional asplenia develops in a large proportion of patients after allogeneic BMT and is also associated with increased risk for bacterial infections. Humoral immunity is impaired in patients with malignancies, leading to decreased production of immunoglobulins, such as in chronic lymphocytic leukemia (CLL), multiple myeloma, and other lymphoproliferative disorders. Humoral immunity is generally well preserved in patients with acute lymphocytic or myelogenous leukemia. However, with intensive chemotherapy and/or progression of the disease, the capacity to produce immunoglobulins decreases. This may lead to defective opsonization of bacteria and subsequent impairment of phagocytosis by neutrophils and macrophages, adding to the quantitative effect of chemotherapy-induced neutropenia. Although the humoral response in patients with malignant lymphomas is unimpaired, subsequent radiotherapy and chemotherapy, particularly if both treatment modalities are combined, lead to reduced antibody titers and increased susceptibility to infections with pneumococci and Haernophilus influenzne. Splenectomy potentiates the reduction of immunoglobulins by chemotherapy in these patients. Therefore, combined therapy may increase the risk of postsplenectomy bacteremia in patients with lymphoma, and, even after curing Hodgkin's disease, patients are left with a potentially lifethreatening humoral immunodeficiency, due to the effects of treatment rather than to the underlying disease itself. Thus, the advent of more aggressive chemotherapy has changed the classic concept of specific defects of host defense mechanisms in the various types of leukemia and lymphoma. The effects of chemotherapy and radiation are now the primary factor determining the nature and depth of the defect in host defense. Likewise, the increased susceptibility to pneumococci and Haemophilus influenzae in patients with CLL or multiple myeloma may be replaced by a defect in cellular immunity and neutrophil function when these patients are being treated with glucocorticosteroids or other agents. Whether

patients with hypogammaglobulinemia due to CLL should routinely receive intravenous immunoglobulins has been a matter of considerable debate. A cost-effectiveness analysis has suggested that indiscriminate replacement may not improve quality or length of life in this patient group, and that it is extraordinarily expensive [112]. However, such a decision analysis model cannot be applied to the individual patient who actually has suffered from recurrent bacterial infections. Therefore, it seems reasonable to institute immunoglobulin replacement in those patients who have had a documented infection with pneumococcus or Hnemophilus injluenzae and have decreased serum IgG concentrations.

10. Sequence of infective events The sequence of risk factors (Figure 7) determines to a large extent the order of infectious events in granulocytopenic cancer patients and the types of infection. The first few days after a treatment course are dominated by profound granulocytopenia and mucosal damage, placing the patient in double jeopardy. After the first week the number of positive blood cultures gradually decreases and remains relatively constant after the fourth week. From day 10 onwards, infections related to central venous catheters occur with increasing risk, depending on the length of time that the catheter is left in place [19]. Few patients develop invasive pulmonary aspergillosis early in the course of granulocytopenia [110]. The initial period of risk of bacterial and fungal infection resolves with recovery of the granulocyte, whereafter the infectious complications are determined by the pace of reconstitution of other components of the immune system, as is illustrated in bone marrow recipients [8]. The major factors that influence immunologic reconstitution after allogeneic bone marrow transplant are graft-versus-host disease and its treatment. Cytomegalovirus, adenovirus, fungi, and protozoa all constitute major pathogens in these patients. In bone marrow transplant recipients, a third major risk period begins approximately 3 months after the procedure, at the time that chronic graft-versus-host disease develops. Sinopulmonary and cutaneous infections, probably related to the IgA deficiency and sicca syndrome, are common. Varicella zoster is the most important cutaneous infection, and pulmonary infections caused by cytomegalovirus and Pneumocystis cnrinii may also be encountered. Months, if not years, after successful engraftment or recovery from other very aggressive treatment, encapsulated organisms can cause rapidly fatal bacteremias and severe respiratory infections due to the lack of opsonizing antibodies.

11. Conclusions It is clear from the foregoing that patients with neoplastic diseases seldom suffer impairment of a single defense mechanism. Rather, the risk of infection

Cytoreductive chemotherapy

Fever

Potential pathogens Mucosa

8 =J

6

-/

\\

Bacteremia

Catheter-related infection Pulmonary infiltrate

1 Neutropenia

u

0 0

10

Days 20

30

40

Figure 7. The sequence of events during neutropenia. Profound granulocytopenia and inucosal damage usually develop about a week after the start of cytoreductive chemotherapy. Thereafter, infectious and other complications tend to coincide with one another, placing the patient at most risk. Fever develops around a week later and, if there is bacteremia, it occurs at this time. The risk of infections related to the central venous catheter increase with the length of time that the catheter is left in place, but signs and symptoms usually manifest themselves during the first few days of fever, that is, during the third week after starting chemotherapy. Infectious complications related to the lung tend to occur a few days later, often being recognized only after 5-6 days of fever. The period of risk of bacterial and fungal infection diminishes with recovery of the granulocytes, when the clinical manifestations of tissue infections may be temporarily exacerbated before finally resolving.

is the product of an interplay between the many lines of defense, all of which can be breached simultaneously. Moreover, any attempt to confine the damage inflicted upon the host defenses by protecting only one specific line of defense (e.g., use of growth factors to stimulate hematopoiesis), is likely to be frustrated and to offer only limited benefit. What is required is a two-pronged approach involving more selective treatment, on the one hand, to avoid damaging healthy tissue and, on the other hand, strategies that prevent, or at least ameliorate, any toxicity that is unavoidable. This requires a holistic approach involving both the laboratory and the clinician in continuing to refine therapeutic regimens that are effective and in designing others to cope with the

morbidity associated with impaired host defenses. Both are essential to successfully achieve remission of neoplastic disease and to maintain the best quality of life for the patient.

References 1. Van der Meer JWM. Defects in host defense mechanisms. In: Rubin RH, Young LS, eds. Current Approach to Infection in the Compromised Host. New York: Plenum Medical, 1994. pp. 33-66. 2. Louria DB. Introduction and epidemiology. Am J Med 1984;76:414-420. 3. Chandra RK. Nutrition, immunity and infection: Present knowledge and future directions. Lancet 1983;1:688-691. 4. Salimonu LS, Ojo-Amaize E, Williams AIO, et al. Depressed natural killer cell activity in children with protein calorie malnutrition. Clinl Immunol Immunopathol 1982;24:1-7. 5. Sugarman B. Zinc and infection. Rev Infect Dis 1983;5:137-147. 6. Craddock PR, Yawata Y, van Santen L. Acquired phagoctye dysfunction: A complication of the hypophosphatemia of parenteral hyperalimentation. N Engl J Med 1974;290:1403-1407. 7. Saltzman RL. Peterson PK. Immunodeficiency in the elderly. Rev Infect Dis 1987;9:11271139. 8. Meyers JD. Infection in bone marrow transplant recipients. Am J Med 1986;81(Suppl. 1A):27-38. 9. McGeer A. Feld R. Epidemiology of infection in immunocompromised oncological patients. Baillikre's Clin Infect Dis 1994;1:415-438. 10. Allen JC. The diabetic as a compromised host. In: Allen JC, ed. Infection and the Compromised Host: Clinical Correlations and Therapeutic Approaches, 2nd ed. Baltimore: Williams & Wilkins, 1981, pp. 229-270. 11. Rubin J, Yu VL. Malignant external otitis: Insights into pathogenesis, clinical manifestations, diagnosis and therapy. Am J Med 1988;139:557-563. 12. Cech P, Stalder H, Widmann JJ, et al. Leucocyte myeloperoxidase deficiency and diabetes mellitus associated with Candidn nlbicnns liver abscess. Am J Med 1979;66:149-153. 13. Schwartz MN. Stress and the common cold. N Engl J Med 1991;325:654-656. 14. Weitzman S, Aisenberg AC. Fulminant sepsis after the successful treatment of Hodgkin's disease. Am J Med 1977;62:47-50. 15. Broome CV, Facklam RR, Fraser DW. Pneumococcal disease after pneumococcal vaccination; an alternative method to estimate the efficacy of pneumococcal vaccine. N Engl J Med 1980;303:549-552. 16. Roth RR, James WD. Microbial ecology of the skin. Ann Rev Microbiol 1988;42:441464. 17. Kotilainen P, Nikoskelainen J. Huovinen P. Emergence of ciprofloxacin-resistant coagulasenegative staphylococcal skin flora in immunocompromised patients receiving ciprofloxacin. J Infect Dis 1990;161:41-44. 18. Weightman NC, Simpson EM, Speller DCE, Mott MG. Oakhill A. Bacteraemia related to indwelling central venous catheters: Prevention, diagnosis and treatment. Eur J Clin Microbiol Infect Dis 1988;7:125-129. 19. Raad I.I., Bodey GP. Infectious complications of indwelling vascular catheters. Clin Infect Dis 1992;15:197-210. 20. Hedin G, Hambraeus A. Multiply antibiotic-resistant Staphylococczls epider~?zidis in patients, staff and environment - a one-week survey in a bone marrow transplant unit, J Hosp Infect 1991;17:95-106. 21. Groeger JS, Lucas AB, Thaler HT, Friedlanderklar H, Brown AE, Kiehn TE, et al. Infectious morbidity associated with long-term use of venous access devices in patients with cancer. Ann Intern Med 1993:119:1168-1174.

22. Wood CA, Pepe R. Bacteremia in a patient with non-urinary tract infection due to Corynebacterizrnz zlrealyticzrin. Clin Infect Dis 1994;19:367-368. 23. Castagnola E, Tasso L, Conte M, Nantron M, Barretta A, Giacchino R. Central venous catheter-related infection due to Conzanzonns acidovorans in a child with non-Hodgkin's lymphoma. Clin Infect Dis 1994;19:559-560. 24. Alnor D, Frimodt-Moller N, Espersen F, Frederiksen W. Infections with the unusual human pathogens Agrobacteriiirrz species and Oclzrobnctrnrn anthropi. Clin Infect Dis 1994;18:914920. 25. Weems JJ. Candida yc~rpsilosis:Epidemiology, pathogenicity, clinical manifestations and antibiotic susceptibility. Clin Infect Dis 1992;14:756-766. 26. Lecciones JA, Lee JW, Navarro EE, Witesby FG. Marshall D, Steinberg SM, et al. Vascular catheter-associated fungenlia in patients with cancer: Analysis of 155 episodes. Clin Infect Dis 1992;14:875-883. 27. Morrison VA, Haake RJ, Weisdorf DJ. The spectrum of non-candida fungal infections following bone marrow transplantation. Medicine 1993:72:78-89. 28. McNab PC, Tomasi TB. Host defense mechanisms at mucosal surfaces. Ann Rev Microbiol 1981;35:477-496. 29. Kolbinson DA, Schubert MM. Fluornoy N, Truelove EL. Early oral changes following bone marrow transplantation. Oral Surg, Oral Med Oral Path01 1988;66:130-138. 30. Weisdorf DJ. Bostrom B, Raelher D, Mattingly M, Walker P, Pihlstrom B, et al. Oropharyngeal mucositis complicating bone marrow transplantation: Prognostic factors and the effect of chlorhexidine mouth rinse. Bone Marrow Transplant 1989;4:89-95. 31. Donnelly JP, Muus P, Schattenberg A, Dewitte T, Horrevorts A. De Pauw BE. A scheme for daily monitoring of oral mucositis in allogeneic BMT recipients. Bone Marrow Transplant 1992;9:409-413. 32. Rocke LK, Loprinzi CL, Lee JK, Kunselman SJ, Iverson RK, Finck G, et al. A randomized clinical trial of two different durations of oral cryotherapy for prevention of 5-fluorouracil related stomatitis. Cancer 1993;72:2234-2238. 33. Bochud PY, Calandra T, Francioli P. Bacteremia due to viridans streptococci in neutropenic patients: A review. Am J Med 1994;97:256-264. 34. Donnelly JP. Dompeling EC, Meis JF, De Pauw BE. Bacleremia due to oral viridans streptococci in neutropenic patients with cancer: Cytostalics are a more important risk factor than antibacterial prophylaxis. Clin Infect Dis 1995;20:469-470. 35. Beighton D, Carr AD, Oppenheim BA. Identification of viridans streptococci associated with bacteraemia in neutropenic cancer patients. J Med Microbiol 1994;40:202-204. 36. Lina B, Forey F, Troncy J, Greenland T. Fleurette J, Etienne J. Oral source of Sr((phy1ococcus r~~idernziclissepticemia in a neutropenic patient. Eur J Clin Microbiol Infect Dis 1994~13~773-774. 37. Jacobs JA, Pietersen HG, Stobberingh EE, Soeters PB. Bacteremia involving the "Streptococcus milleri" group: Analysis of 19 cases. Clin Infect Dis 1994;19:704-713. 38. Ferretti GA, Ash RC, Brown AT, Largent BM, Kaplan A, Lillich TT. Chlorhexidine for in patients receiving bone prophylaxis against oral infections and associated complicatio~~s marrow transplants. JADA 1987;114:461467. 39. Meurman JH. Laine P, Murtomaa H. Lindqvist C. Torkko H, Teerenhovi L. et al. Effect of antiseptic mouthwashes on some clinical and microbiological findings in the mouths of lymphoma patients receiving cytostalic drugs. J Clin Periodont 1991;18:587591. 40. Bergmann OJ. Alterations in oral microflora and pathogenesis of acute oral infections during remission-induction therapy in patients with acute myeloid leukaemia. Scand J Infect Dis 1991:23:355-366. 41. Schuster MW. Granulocyte-macrophage colony-stimulating factor (GM-CSF) - what role in bone marrow transplantation. Infection 1992;20(Suppl. 2):S95-S99. 42. Bronchud M. Can hematopoietic growth factors be used to improve the success of cytotoxic chemotherapy. Anti-Cancer Drugs 1993;4:127-139.

43. D e Witte T, Van Der Lely N, Muus P, Donnelly JP, Schattenberg T. Recombinant human granulocyte macrophage colony stimulating factor (rhGM-CSF) accelerates bone marrow recovery after allogeneic T-cell depleted bone marrow transplantation. L'Ospedale Maggiore 1993:87:42-46. 44. McGuire DB, Altomonte V, Peterson DE, Wingard JR. Jones RJ, Grochow LB. Patterns of mucositis and pain in patients receiving preparative chemotherapy and bone marrow transplantation. Oncol Nursing Forum 1993:20:1493-1502. 45. Sable CA, Donowitz GR. Infections in bone marrow transplant recipients. Clin Infect Dis 1994:18:273-284. 46. Raemaekers J, De Witte T, Schattenberg A , Van Der Lely N. Prevention of leukaemic relapse after transplantation with lymphocyte-depleted marrow by intensification of the conditioning regimen with a 6-day continuous infusion of anthracyclines. Bone Marrow Transplant 1989;4:167-171. 47. Rubin M, Hathorn JW, Marshall D, Gress J. Steinberg SM, Pizzo PA. Gram-positive infections and the use of vancomycin in 550 episodes of fever and neutropenia. Ann Intern Med 1988:108:30-35. 48. The EORTC International Antimicrobial Therapy Cooperative Group. Gram-positive bacteraemia in granulocytopenic cancer patients. Eur J Cancer 1990;26:569-574. 49. The EORTC International Antimicrobial Therapy Cooperative Group and National Cancer Institute of Canada. Vanco~nycinadded to empirical combination antibiotic therapy for fever in granulocytopenic cancer patients. J Infect Dis 1991;163:951-958. 50. Awada A , Van Der Auwera P, Meunier F, Daneau D. Klastersky J. Streptococcal and enterococcal bacteremia in patients with cancer. Clin Infect Dis 1992;15:3348. 51. Devaux Y, Archimbaud E, Guyotat D, Plottoll C, Maupas J, Fleurette J, et al. Streptococcal bacteremia in neutropenic adult patients. Nouv Rev Fr Hematol 1992;34:191-195. 52. Menichetti F. Gram-positive infections in neutropenic patients - glycopeptide antibiotic choice. J Antimicrob Chemother 1992;29:461-462. 53. Beattie G, Whelan J, Cassidy J, Milne L, Burns S. Leonard R. Herpes si~nplexvirus, Cnndidn ~lbicrrlzs and mouth ulcers in neutropenic patients with non-haematological malignancy. Cancer Chemother Pharmacol 1989:25:75-76. 54. Bergmann OJ. Oral infections in haematological patients - pathogenesis and clinical significance. Danish Med Bull 1992:39: 15-29. 55. Van der Waaij D. The colonization resistance of the digestive tract of man and animals. In: Clinical and Experimental Gnotobiotics. New York: Gustav Fischer Verlag. 1979. 56. Schimpff SC. Infection prevention during profound granulocytopenia: New approaches to alimentary canal microbial suppression. Ann Intern Med 1980:93:358-361. 57. Young LS. Antimicrobial prophylaxis against infection in neutropenic patients. J Infect Dis 1983;147:611-614. 58. Van der Waaij D. The ecology of the human intestine and its consequences for overgrowth by pathogens such as Clostridizrrn dificile.Ann Rev Microbiol 1989:43:69-87. 59. Vollaard EJ. Clasener HAL. Colonization resistance. Antimicrob Agents Chemother 1994:38:409414. 60. Van der Waaij D . Effect of antibiotics on colonization resistance. In: Medical Microbiology. London: Academic Press, 1984, pp. 227-237. 61. Louie TJ. Chubb H, Bow EJ, Conlp JM. Harding GKM. Rayner E. et al. Preservation of colonization resistance parameters during empiric therapy with aztreonam in febrile neutropenic patient. Rev Infect Dis 1985;7:S747-S761. 62. Meijer-Severs GJ, Van Santen E. Short-chain fatty acids and succinate in feces of healthy human volunteers and their correlation with anaerobic cultural counts. J Gastroenterol 2987;22:672-676. 63. Welling GW. Groen G. Inactivation of aztreonam by faecal supernatants of healthy volunteers as determined by HPLC. J Antimicrob Chemother 1989;24:805-810. 64. Dietrich M, Rasche H, Rommel K, Hochapfel G. Antimicrobial therapy as a part of the decontamination procedures for patients with acute leukemia. Eur J Cancer 1973:9:443-447.

65. Bender JF, Schimpff SC, Young VM, Fortner CL, Brouillet MD, Love LJ, et al. Role of vanconlycin as a component of oral nonabsorbable antibiotics for microbial suppression in leukemic patient. Antimicrob Agents Chemother 1979;15:455-460. 66. Pizzo PA, Robichaud KJ, Edwards BK, Schumaker C, Kramer BS, Johnson AJ. Oral antibiotic prophylaxis in patients with cancer: A double-blind randomized placebocontrolled trial. J Pediatr 1983;102:125-133. 67. Walsh TJ, Schimpff SC. Prevention of infection among patients with cancer. Eur J Cancer Clinical Oncology 1983;19:1333-1344. 68. Nord CE, Kager L, Heirndahl A. Impact of antimicrobial agents on the gastrointestinal microflora and the risk of infections. Am J Med 1984;15:99-106. 69. Jones PG, Bodey GP, Swabb EA, Rosenbaum B. Effect of aztreonam on throat and stool flora of cancer patients. Antimicrob Agents Chemother 1984;26:941-943. 70. Rozenberg-Arska M, Dekker AW, Verhoef J. Ciprofloxacin for selective decontamination of the alimentary tract in patients with acute leukemia during remission induction treatment: The effect on fecal flora. J Infect Dis 1985;152:104-107. 71. Jones RN, Barry AL, Thornsberry C. In-vitro studies of meropenenl. J Antimicrob Chemother 1989;24(Suppl. A):9-29. 72. Vollaard EJ, Clasener HAL, Janssen AJHM. The contribution of Escherichia coli to microbial colonization resistance. J Antimicrob Chemother 1990;26:411-418. 73. Vollaard EJ, Clasener HAL, Janssen AJHM. Decontamination of the bowel by intravenous administration of pefloxacin. J Antimicrob Chemother 1990;26:847-852. 74. Welling GW, Slootmakervandermeulen C, Jansen GJ. Inactivation of imipenem by faecaI fractions from human volunteers and the effect of clavulanate and cilastatin. J Antimicrob Chemother 1993;31:617-619. 75. Van der Waaij D, Hofstra H , Wiegersma N. Effect of B-lactam antibiotics on the resistance of the digestive tract of mice to colonization. J Infect Dis 1982;146:417-422. 76. Wiegersma N, Jansen G. Van Der Waaij D. Effect of twelve antimicrobial drugs on the colonization resistance of the digestive tract of mice and on endogenous potentially pathogenic bacteria. J Hygiene (Cambridge) 1982;88:221-230. 77. Clasener HA, Vollaard EJ, van Saene HK. Long-term prophylaxis of infection by selective decontamination in leukopenia and in mechanical ventilation. Rev Infect Dis 1987;9:295328. 78. Vollaard EJ, Clasener HAL, Van Griethuysen AJA, Janssen AJHM, Sanders-Reimers AHJ, Muller NF, et al. Influence of cefaclor, phenethicillin, co-trimoxazole and doxycycline on colonization resistance in healthy volunteers. J Antimicrob Chemother 1988;22:747-758. 79. Van der Leur JJJPM, Thunnissen PLM, Clasener HAL, Muller NF, Dofferhoff ASM. Effects of imipenem, cefotaxime and cotrimoxazole on aerobic microbial colonization of the digestive tract. Scand J Infect Dis 1993:25:473-478. 80. Vollaard EJ, Clasener HAL, Janssen AJHM. Co-trimoxazole impairs colonization resistance in healthy volunteers. J Antimicrob Chemother 1992;30:685-691. 81. Meijer-Severs GJ, Joshi JH. The effect of new broad-spectrum antibiotics on faecal flora of cancer patients. J Antimicrob Chemother 1989;24:605-613. 82. Chomarat M, Espinouse D. Lnctnhacill~~s rlznrnnoszrs septicemia in patients with prolonged aplasia receiving ceftazidime-vancomycin [letter]. Eur J Clin Microbiol Infect Dis 1991;10:44. 83. Cometta A, Calandra T, Bille J, Glauser MP. Escherichict coli resistant to fluoroquinolones in patients with cancer and neutropenia. Engl J Med 1994;330:1240-1241. 84. Kern WV, Andriof E, Oethinger M, Kern P, Hacker J, Marre R. Emergence of fluoroquinolone-resistant Escherichin coli at a cancer center. Antimicrob Agents Chemother 1994;38:681-687. 85. Kern WV, Markus A, Andriof E. Bacterernia due to fluoroquinolone resistant Escherichia coli in two immunocompromised patients. Eur J Clin Microbiol Infect Dis 1994;13:161-165. 86. Zinner SH, Calandra T, Meunier F, Gaya H, Viscoli C, Klastersky J, et al. Reduction of fever and streptococcal bacteremia in granulocytopenic patients with cancer - a trial of oral penicillin V or placebo combined with pefloxacin. JAMA 1994;272:1183-1189.

87. Peters WG, Willemze R, Colly LP, Guiot HFL. Side effects of intermediate- and high-dose cytosine arabinoside in the treatment of refractory or relapsed acute leukaemia and onHodgkins lymphoma. Neth J Med 1987;30:64-74. 88. Guiot HFL, Biemond J, Klasen E, Gratama JW, Kramps JA, Zwaan FE. Protein loss during acute graft-versus-host disease: Diagnostics and clinical significane. Eur J Haematol 1987;38:187-196. 89. Callum JL, Brandwein JM, Sutcliffe SB, Scott JG, Keating A. Influence of total body irradiation on infections after autologous bone marrow transplantation. Bone Marrow Transplant 1991;8:245-25 1. 90. Johnson S, Driks MR, Tweten RK, Ballard J, Stevens DL, Anderson DJ, et al. Clinical courses of seven survivors of Clostridir~m septicum infection and their immunologic responses to a-toxin. Clin Infect Dis 1994;19:761-764. 91. Fegan C, Poynton JA, Whittaker JA. The gut mucosal barrier in bone marrow transplantation. Bone Marrow Transplant 19905373-377. 92. Moody MR, Morris MJ, Young VM, Moye LA 111, Schimpff SC, Wiernik PH. Effect of two cancer chemotherapeutic agents on the antibacterial activity of three antimicrobial agents. Antimicrob Agents Chemother 1978;14:737-742. 93. Jacobs JY, Michel J, Sacks T. Bactericidal effect of combinations of antimicrobial drugs and antineoplastic antibiotics against Sfrtphylococcus azirells. Antimicrob Agents Chemother 1979;15:580-586. 94. Michel J, Jacobs JY, Sacks T. Bactericidal effect of combinations of antimicrobial drugs and antineoplastic antibiotics against gram-negative bacilli. Antimicrob Agents Chemother 1979;16:761-766. 95. Bodet CA 111, Jorgensen JH, Drutz DJ. Antibacterial activities of antineoplastic agents. Antimicrob Agents Chemother 1985;28:437-439. 96. Neuman M. The antimicrobial activity of non-antibiotics - interactions with antibiotics. APMIS 1992;lOO(Supp1. 30):15-23. 97. Bergstrom P, Grankvist K, Henriksson R. Interaction between antibiotics and antineoplastic drugs on antibacterial activity in vitro: Estramustine phosphate sensitizes pneumococci to amikacin. Inte J Oncol 1994;4:43-439. 98. Van Cutsem JM, Thienpont D. Miconazole, a broad-spectrum antimycotic agent with antibacterial activity. Chemotherapy 1972;f 7:392-404. 99. Ware AJ, Coller BS. Platelet morphology, biochemistry, and function. In: Beutler E, Lichtman MA, Coller BS, Kipps TJ eds. Williams Hematology, (5th ed.) New York: McGraw-Hill, 1995, pp. 1161-1201. 100. Viscoli C, Bruzzi P, Castagnola E, Boni L, Calandra T. Gaya H, et al. Factors associated with bacteraemia in febrile, granulocytopenic cancer patients. Eur J Cancer 1994;30A:430-437. 101. Cronkite EP, Fliedner TM. Granulopoiesis. N Engl J Med 1964:270:1347-1352. 102. Weller PF. The immunobiology of eosinophils. N Engl J Med 1991;324:1110-2114. 103. Bodey GP, Buckley M, Sathe YS, Freireich EJ. Quantitative relationships between circulating leukocytes and infection in patients with acute leukemia. Ann Intern Med 1966:64:328340. 104. Sickles EA, Greene WH, Wiernik PH. Clinical presentation of infection in granulocytopenic patients. Arch Intern Med 1975:135:715-719. 105. EORTC International Antimicrobial Therapy Cooperative Group. Three antibiotic regimens in the treatment of infection in patients with cancer. J Infect Dis 1978;137:14-29. 106. Pizzo PA. Robichaud KJ, Wesley R, Commers JR. Fever in the pediatric and young adult patient with cancer. A prospective study of 1001 episodes. Medicine 1989;61:153-165. 107. DeGregorio MW, Lee WMF, Linker CA, Jacobs RA, Ries CA. Fungal infections in patients with acute leukemia. Am J Med 1982;73:543-548. 108. Anaissie E, Bodey GP, Kantarjian H, et al. New spectrum of fungal infections in patients with cancer. Rev Infect Dis 1989:11:369-378. 109. Bodey GP, Bueltman B, Duguid W. et al. Fungal infections in cancer patients: An international autopsy survey. Eur J Clin Microbial Infect Dis 1992;11:99-109.

110. Gerson SL, Talbot GH, Hurwitz S, et al. Prolonged granulocytopenia: The major risk factor for invasive pulmonary aspergillosis in patients with acute leukemia. Ann Intern Med 1984:100:345-351. 111. Young RC, Corder MP, Haynes HA, et al. Delayed hypersensitivity in Hodgkin's disease. Am J Med 1972;56:63-72. 112. Weeks JC, Tierney MR, Weinstein MC. Cost effectiveness of prophylactic intravenous immune globulin in chronic lymphatic leukemia. N Engl J Med 1991;325:81-86.

2. Epidemiology of infectious complications in cancer patients Teresa Zembower

1. Introduction Infectious complications are a serious cause of morbidity and mortality in cancer patients, especially those with underlying hematologic malignancies. Several large studies have estimated that 70-75% of deaths in patients with acute leukemia are caused by infection, with or without hemorrhage [I-31. Fewer data exist on infectious mortality in patients with solid organ tumors; however, approximately 50% of these patients are estimated to have an infection as either the primary or an associated cause of death [4,5]. Because patients with underlying malignancies are a heterogeneous group, an epidemiologic review of infections in these patients must take into account the diversity of the patient population. The nature of the underlying malignancy, the immunodeficiencies associated with it, and the treatments directed against it are all important determinants of infection. For ease of understanding, the factors that predispose to infection are divided into those that are host associated and those that are treatment associated. Host-associated factors include deficiencies of cellular and humoral immunity, disrupted anatomic barriers, central nervous system dysfunction, prior splenectomy, and alterations in the microbial flora. Treatmentassociated factors include surgery, radiation, and chemotherapy, antibiotic use, and diagnostic or invasive procedures, including blood transfusions and bone marrow infusions (Table 1). Clearly, more than one predisposing factor may exist simultaneously in a given patient. To a large extent, however, these risk factors are associated with specific infectious pathogens, and an understanding of each individual risk factor can help direct diagnosis and empiric therapy (Table 2).

2. Host-associated factors 2.1 Granulocytopenia Granulocytopenia (neutropenia) is defined as an absolute neutrophil count (ANC) lower than 1000 or 500 c e l l s l m m ~ n dis the single most important risk Gary A. Noskin (erl), M A N A G E M E N T OF INFECTIOUS COMPLICATIONS IN C A N C E R PATIENTS. O 1998. K h w e r Academic. hiblishers, Boston. All rights reserved.

Table I . Factors predisposing to infection in cancer patients Factor Host associated Granulocytopenia Cell-mediated immunodeficiency Humoral immunodeficiency Disrupted anatomic barriers Splenectomy Central nervous system dysfunction Changes in microbial flora Treatment associated Surgery Radiation Chemotherapy Antibiotics Diagnostic and invasive procedures Central venous catheters Blood transfusions and bone marrow infusions

Comments Absolute neutrophil count less than 1000 or 500 cells/mm"; most important risk factor for infection in cancer patients Defective monocyte/macrophage or T-cell function Defective B-cell function, leading to loss of opsonizing an tibodies Caused by the invading malignancy or the treatments directed against it Loss of monocytes, macrophages, B lymphocytes, and properidin (a component of the alternate complement pathway) Loss of gag reflex, impaired micturition, and loss of mobility Occur secondary to the severity of the underlying illness, invasive procedures, and antimicrobial usage Especially pelvic, gastrointestinal, and maxillofacial; largest interventions carry greatest infectious risk Preoperative irradiation leads to fistula formation and impaired wound healing Damages anatomic barriers and leads to cytopenias in a dose-related fashion Rapidly and radically alter microbial flora Bacterial colonization occurs within days of insertion; 15-20% develop overt infection Infection occurs most commonly through collection of specimens from infected donors

factor for the development of bacterial infection in cancer patients. Patients with acute leukemia, those who have received intensive myelosuppressive therapies for their underlying malignancies or as part of their bone marrow transplantation, or those with aplastic anemia are most likely to develop or present with granulocytopenia [6]. Patients with lymphoma may also have marrow involvement severe enough to result in neutropenia [7]. Although less common, solid organ tumors, such as metastatic carcinoma of the breast, prostate, lung, adrenal, thyroid, and kidney, can all infiltrate the bone marrow and result in granulocytopenia [8]. The incidence of infection increases when the neutrophil count falls below 500 cells/mm3. Patients with severe neutropenia, having an ANC below 100 cells/mmi, represent a unique subset of patients who are at highest risk of infection [9]. This level of severe neutropenia decreases the mobilization of white blood cells to the site of inflammation. Therefore, when bacterial pathogens are encountered, the usual acute pyogenic response is muted or absent, making it difficult for these patients to control the spread of an infection [lo].

Table 2. Predisposing factors and their associated infections Factor

Infectious pathogen

Neutropenia

Bacteria Staphylococcus aureus Staphylococcus epidermidis Alpha-hemolytic streptococci Escherichia coli Klebsiella spp. Pseudomonas neruginosa Viruses Herpes simplex Varicella-zoster Cytomegalovirus Yeastlfungi Candida spp. Aspergillus spp. Bacteria Listeria nzonocytogenes Salmonella spp. Nocardia asteroides Mycobacteria Legionella pneltmophila Viruses Herpes simplex VariceIIa-zoster Cytomegalovirus Epstein-Barr Yeast Cryptococcus neoformans Histoplasma capsulatum Coccidioides immitis Protozoa Pneumocystis carinii Toxoplasrna gondii Cryptosporidium Helminth Strongyloides stercoralis Streptococcus pneumoniae Haemophilus influenzae Skin Staphylococcus aureus Staphylococcus epidermidis Streptococcus pyogenes Oral cavitylnasopharynx Anaerobes Streptococci Haemop hilus influenzae Gastrointestinal tract Enterobacterianceae Fungi Female genital tract Enterobacteriaceae Anaerobic GNB Enterococci Clostridia spp. Streptococcus pneumoniae Haemophilus influenzae Neisseria nzeningitidis

Cell-mediated immunodeficiency

Humoral immunodeficiency Disruption of anatomic barriers

Splenectomy

Table 2. (continued)

Factor

Infectious pathogen

Central venous catheters

Staphylococci Streptococci Bacillrls spp. Corynebacterium spp. Cnndidn nlbicans Mycobncterizlm chelonei Mycobacteri~imfortuitum Bacteria P. JEuorescens/putida Yersinia enterocolitica Viruses Hepatitis viruses HIV Epstein-Barr Cytomegalovirus Protozoa Leishmaniasis Trypanosomiasis Chagas' disease Microfilarial diseases Malaria Bnbesiu nzicroti

Blood transfusions

The rapidity in the decline of the neutrophil count is also important in determining the risk of infection. For example, a rapidly falling count is much more likely to be associated with infection than is either a slowly declining count, as observed with syndromes such as cyclic neutropenia, or a stable neutropenia, as is often seen in patients with stable aplastic anemia or benign idiopathic neutropenia [ll]. The duration of neutropenia also correlates with the risk of infection. In fact, once a patient is febrile and neutropenic, the duration of the neutropenia is a stronger predictor of infection than is the absolute neutrophil count [12]. Virtually all patients who remain severely neutropenic (ANC less than 100 cellslmm3) will develop an infection within 3 weeks [3,8]. Patients whose neutrophil counts are falling are also at greater risk than those patients whose counts are recovering. For example, the patient with an ANC of 500 cells/mm3 who has just received cytotoxic chemotherapy and is expected to be severely neutropenic within a few days is at greater risk than is the patient whose neutrophil count is 200 cells/mm3 and rising. A large multicenter study by the European Organization for Research on Treatment of Cancer (EORTC) demonstrated that the change in granulocyte count was the most important factor in determining success or failure of antibiotic therapy for gram-negative bacteremia. Only 22% of patients whose granulocyte count did not rise by at least 100 cellslmm3 during therapy were successfully treated, whereas 88% of those whose count rose by at least 100 cells/mm3 completely improved [13].

Several studies have examined the sites and types of infection that occur in the neutropenic host. The neutropenic cancer patient usually has a variety of other predisposing factors that act together to increase the risk of infection. Cytotoxic chemotherapy damages the mucosal membranes of the alimentary and respiratory tracts, drug-induced vomiting with stomach acid reflux predispose to infection in the distal esophagus, underlying periodontal disease or reactivation of herpes virus can serve as a cofactor for the development of oral infection, frequent bowel movements and the associated high pressure damage the anal mucosa, and vascular catheters and other invasive procedures damage the integument [14]. The infecting organisms in neutropenic patients are usually the organisms that colonize these particular sites (Figure 1). Organisms that colonize the lower gastrointestinal tract, such as Escherichia coli, Klebsiella pneun:wniae, and Pseudomonas aeruginosa, are the most commonly encountered gramnegative pathogens causing infections in these patients. Staphylococcus epidermidis, Staphylococcus aureus, and a-hemolytic streptococci colonize the skin and upper respiratory tract and cause the majority of gram-positive bacterial infections [15]. However, the microbial flora changes with prolonged hospitalization, largely as a result of antibiotic selective pressure. Gramnegative bacilli begin to colonize the upper respiratory tract, resulting in gram-negative bacterial infections of the pharynx, lung, and esophagus [16].

Oropharynx and Upper Respiratory Tract

.- a-hemolytic streptococci fl fl

..-

Stephylococcus sureus Gram-negative bacilli Candida sp. Herpes simplex virus

fl fl

Srephyhnmcusepidermidis Staphyloecrccus aureus

.- Gram-negative bacilli /

S t a p h y l m c u s 8U.feUS

.- Aspergillus sp.

Gastrointestinal Tract v Escherichia coli v KlobsioNa sp. v Other Enlerobacteriaceae v Pseubomon8s m g i n a s a v Anaerobes v Enterococci v Candida sp.

Figure I . Sites and types of infections in the neutropenic host.

r. Acinerobacter ca/atawtkus v Other gram-negative bacllti

v Enterococci

Specific antibiotics, such as the fluoroquinolones, predispose to colonization with streptococci and quinolone-resistant gram-negative bacteria [17]. Other bacteria, such as Serratia marcescens, Proteus spp., Enterobacter spp., Aeromonas spp., and Acinetobacter calcoaceticus, and fungi such as Candida spp. and Aspergillus spp., less frequently colonize these patients and can cause infections in the neutropenic host [8,18-211. 2.2 Cell-mediated immunity Cell-mediated immunity depends upon the interaction of T lymphocytes with the monocytelmacrophage system. Working together, these cells enhance the phagocytosis and killing of intracellular pathogens. T-lymphocyte precursors are released from the bone marrow and migrate to the thymus gland, where maturation into T lymphocytes occurs. The mature T lymphocytes then exit the thymus and are present in the circulation, the lymph nodes, and the spleen. Upon antigenic stimulation, T lymphocytes secrete lymphokines, which act in two important ways. One group, the immunoregulatory lymphokines, modulates the function of both T and B cells. The second group, the inflammatory lymphokines, regulates the function of macrophages. The two most important lymphokines are migration inhibitory factor (MIF) and macrophage activating factor (MAF). MIF slows the rate of migration of macrophages, keeping them at the site of inflammation. MAF greatly enhances the killing capacity of macrophages for intracellular organisms. Mononuclear phagocytes arise in the bone marrow from immature precursors, undergo differentiation in the marrow as promonocytes, and are then released into the circulation as monocytes. Monocytes remain in the circulation for only a brief time (half-life of 22-24 hours), after which they leave the blood to become tissue macrophages. These macrophages exist throughout the body in the brain, lung, bone, liver, kidney, spleen, and skin. Under the direction of lymphokines, macrophages become activated. In this form, the macrophages demonstrate increased phagocytosis and microbial killing. In fact, the activated macrophages are responsible for the killing of many intracellular pathogens that are resistant to killing by monocytes, quiescent macrophages, and neutrophils [22,23]. Hodgkin's disease and HIV infection are the prototypical illnesses associated with cellular immune dysfunction [24,25]. However, impairment of cellmediated immunity can occur with most cancers, including acute and chronic leukemia; solid organ tumors; such as breast, lung, brain, gastrointestinal tract, and urogenital tumors and following bone marrow transplantation [26-331. Alternatively, irradiation and certain medications, such as azathioprine, cyclosporine, and corticosteroids, can result in cellular immunodeficiency [34-361. Several predominantly intracellular pathogens are associated with deficiencies of cell-mediated immunity. These include the bacteria Listeria monocytogenes, Salmonella spp., Nocardia asteroides, mycobacteria (both

M. t~lberculosisand the nontuberculous mycobacteria), and Legionella; the yeast Cryptococcus neoformnns, Histoplasma capsulatum, Coccidioides immitis, and Pneumocystis carinii; the viruses varicella-zoster virus (VZV), cytomegalovirus (CMV), Ebstein-Barr virus (EBV), herpes simplex virus (HSV), and adenovirus; the protozoa Toxoplasma gondii and Cryptosporidium; and one helminth, Strongyloides stercornlis [10,37-471. 2.3 Humoral immunity B lymphocytes are the cells primarily involved in humoral immunity. Like T lymphocytes, they arise from precursor stem cells in the bone marrow. In humans, B lymphocytes mature in gut-associated lymphoid tissue of Peyer's patches in the ileum, in the submucosal lymphoid follicles of the appendix, and in the bone marrow. Following maturation, they are distributed to the spleen and lymph nodes. Under proper antigenic stimulation, they differentiate into immunoglobulin (antibody)-producing plasma cells. These plasma cells produce opsonizing antibodies. Coating or opsonizing certain bacteria, particularly encapsulated bacteria, greatly enhances their phagocytosis. Patients with defects in humoral immunity lack opsonizing antibodies to the common encapsulated pyogenic bacteria and thus are susceptible to infections with organisms such as Streptococcus pneumoniae and Hnemophilus influenzae [22]. Multiple myeloma is the neoplasm classically associated with altered humoral immunity. Multiple myeloma patients are hypogammaglobulinemic, producing normal immunoglobulins at only 10% of the normal rate. Interestingly, some myeloma patients have no infections with S. pneumoniae. These patients produce the specific opsonizing antibodies needed to defend them, whereas other myeloma patients who do not produce the necessary opsonizing antibodies have recurrent pneumococcal infections with the same or different strains [48-501. Waldenstrom's macroglobulinemia and chronic lymphocytic leukemia are also neoplasms affecting B lymphocytes. They too are characterized by defective humoral immunity [51,52]. Likewise, patients who undergo prolonged immunosuppresive therapies, such as those used to prevent graft-versus-host disease (GVHD) following allogeneic bone marrow transplantation, may not regain their ability to synthesize IgG and IgM for up to 1 year, and the synthesis of IgA may remain depressed for years [53,54]. 2.4 Anatomic barriers The anatomic barriers of the skin and soft tissues are essential for protection of the immunocompromised host. These barriers can be damaged by invasion from the malignancy itself or by treatments directed at the malignancy, such as radiation and cytotoxic chemotherapy. Primary or metastatic tumors of the skin can disrupt this important anatomic barrier. These tumors increase the

risk of bacteremia with organisms such as Staphylococcus aureus, Staphylococcus epidermidis, and Streptococcus pyogenes, which colonize the skin [55]. Mucous membranes of the oral cavity and nasopharynx may be affected by invasive carcinomas, resulting in local infection in the mouth, nose, throat, or sinuses. This predisposes to anaerobic necrotizing infections, streptococcal infections, and infections due to H. influenzae. These infections can spread to the meninges, causing meningitis, or can locally invade the sinuses, causing osteomyelitis with or without subsequent cerebral abscess. The mucous membranes of the gastrointestinal tract can be disrupted by invasive carcinomas, causing local abscess formation, bacteremia with gram-negative bacilli, or perforation with resulting peritonitis. The possibility of fungal infection increases if these patients have received broad-spectrum antibiotics. Disruption of barriers in the female genital tract can occur with gynecological malignancies. Pathogens such as anaerobic gram-negative bacilli, enteric gram-negative bacilli, Clostridia spp., and enterococci will occasionally invade the bloodstream, causing systemic infection [8,56]. Mucous membranes can also be damaged by radiation or chemotherapy. Regimens that include cytosine arabinoside, anthracyclines, methotrexate, 6-mercaptopurine, and 5-fluorouracil are most likely to cause gastrointestinal mucositis and stomatitis [51. In addition to damaging mucous membranes, malignancies can cause obstruction of various orifices or passages, leading to stasis of body fluids and subsequent infection. Carcinoma of the prostate, ovary, cervix, and rectum commonly obstruct the urinary tract, whereas central nervous system tumors can impair micturition, all leading to urinary retention and recurrent urinary tract infection. Primary or metastatic lung tumors can obstruct the bronchi, causing postobstructive pneumonia and abscess formation. Obstruction of the biliary tract by lymphoma or pancreatic cancer predisposes to ascending cholangitis. Tumors that obstruct the blood vessels can cause septic thrombophlebitis or ischemia, which predispose to infection [5,8].

2.5 Splenectomy The spleen contains large numbers of monocytes, macrophages, and B lymphocytes, enabling it to perform many important immune functions. First, the spleen filters particles from the bloodstream using strategically positioned tissue macrophages. These macrophages can engulf circulating, opsonized organisms, helping to rid the body of many encapsulated bacteria [57]. These macrophages can also remove non-opsonized bacteria, making the spleen critically important to the individual who encounters infection with a new serotype of pneumococci to which he or she is not immune. The spleen is also important in humoral immunity; it is here that the immunoglobulin response primarily takes place. In addition, splenectomized patients have reduced levels of properidin, an important component of the alternate complement pathway.

Patients who have undergone splenectomy are at increased risk for infections with encapsulated bacteria such as Streptococcus pneumoniae, H. influenzae, and Neisseria meningitidis. In fact, patients who undergo splenectomy for staging or treatment of a hematologic malignancy have approximately a 5% risk of developing overwhelming sepsis at some time during their lifetime [58]. A syndrome of overwhelming pneumococcal sepsis has been described in splenectomized children, and much Iess commonly in splenectomized adults [59]. For this reason, patients who undergo splenectomy should receive the polyvalent pneumococcal vaccine, preferably prior to splenectomy in order to obtain the best immune response [60]. Another group of patients who are at increased risk of invasive pneumococcal infections are those who have undergone allogeneic bone marrow transplantation. The reasons are twofold: (1) these patients demonstrate a decreased ability to synthesize IgG and IgM, and (2) they fail to switch from IgM to IgG production if chronic GVHD develops, producing a state of functional asplenia [53,61].

2.6 Central nervous system dysfunction Primary or metastatic tumors of the central nervous system (CNS) can predispose to a variety of infections. Patients who have either a partial or complete loss of their gag reflex are at greater risk for aspiration pneumonia. Impaired micturition is a common finding in these patients and can lead to urinary retention and recurrent urinary tract infections. Likewise, patients with impaired mobility are predisposed to skin breakdown, which can result in decubitus ulcers and osteomyelitis. Interestingly, meningitis, encephalitis, and brain abscesses are uncommon in patients with CNS tumors unless related to problems of surgery [58,62]. 2.7 Changes in microbin 1 flora Dramatic changes in microbial flora can occur in debilitated patients. The severity of the underlying illness, invasive procedures, and antibiotic usage are all associated with alterations in the normal flora. A study by Johanson and colleagues demonstrated that severity of illness and antimicrobial usage can both change the normal flora. Throat cultures were obtained from normal volunteers and from patients hospitalized on a psychiatric ward, an orthopedic ward, and two medical wards. The patients on both medical wards had severe underlying medical illnesses; on one ward they were receiving antibiotics and on the other they were not. Throat cultures from the normal volunteers and the psychiatric patients revealed normal flora. However, the throat cultures from 16% of the orthopedic patients, 57% of the medical patients without antibiotics, and 80% of the medical patients with antibiotics revealed gramnegative bacilli [16]. This suggests that severity of illness and antibiotics, not hospitalization per se, are associated with changes in endogenous flora.

Invasive procedures, such as placement of urinary or intravascular catheters or tracheostomies, can also alter normal flora. Urinary and intravascular catheters can become colonized with organisms that track along the catheter and colonize these normally sterile body sites. Patients with tracheostomies generally become colonized with gram-negative bacteria within a few days following placement. If pneumonia develops, it is usually due to these same bacterial pathogens with which the patient is colonized. Indeed, the majority of patients are infected with the organisms with which they are colonized; however, 50% of these organisms are acquired after hospitalization [14]. Studies have demonstrated that serial axillary surveillance cultures grow primarily Staphylococcus epidermidis and Corynebacterium spp. on admission. As the illness and the hospitalization progress, the resident flora shifts toward gram-negative bacteria and yeast such as Candida nlbicans [10,63]. Of all the predisposing conditions, antibiotic use is the single most important factor leading to changes in host flora. Although necessary for both prophylaxis and treatment of infections, antimicrobial agents can cause rapid and radical alterations in endogenous flora. One of the most common examples is Clostridium dificile colonization and infection induced by antibiotic therapy [64,65]. However, broad-spectrum antibiotics are more apt to suppress normal, noninvasive flora, particularly anaerobes, and to cause a shift toward gram-negative bacteria and yeast. In one study, surveillance cultures were monitored in 10 patients receiving ampicillin for 3 weeks. Nine of these patients became rapidly colonized with ampicillin-resistant gram-negative rods, and several isolates were multiply drug resistant. Only one patient in the control group acquired a multidrug-resistant organism [lo]. Likewise, the incidence of fungal infections is related to prophylaxis and treatment with broad-spectrum antibiotics. For example, fluconazole prophylaxis has resulted in the development of resistant strains of C. albicans and Torlklop~i~ glabrata [66-691 and to outbreaks of inherently fluconazole-resistant Candida krusei [70,71]. To understand the changing microbial flora, it is important to understand a concept known as colonization resistance. Individuals are colonized with noninvasive flora that, in a sense, can be considered "protective." This normal flora prevents colonization and subsequent infection with more invasive, pathogenic bacteria. Patients who have lost their normal flora, such as those receiving broad-spectrum antibiotics, are at greater risk of colonization and infection with these more invasive organisms. In an animal model of infection, van der Waaij elegantly depicts this phenomenon. In this study, three groups of mice were used: one group was rendered completely germ-free, a second group retained their anaerobic flora but were rendered free of aerobes, and the third group of normal mice served as the control. The mice were given different oral doses of streptomycin-resistant E. coli for ease of detection, and persistent colonization was determined by evaluation of fecal flora. The control group required lo7 E. coli to become persistently colonized, the mice with only anaerobic flora required approximately lo5E. coli, and the germ-free

mice, who had no colonization resistance, required only 10' to lo2 E. coli, confirming the importance of colonization resistance [72].

3. Treatment-associated factors Although essential to patient care, no procedure or treatment is without risk. The following treatment-associated factors have all been shown to predispose patients with underlying malignancies to an increased risk of infection.

3.1 Surgery Extensive surgery, especially in the pelvic, gastrointestinal, or maxillofacial regions, increases the risk of infection in cancer patients [8]. Although the procedures are often necessary, especially for advanced invasive tumors, they remove large areas of otherwise protective tissue and disrupt anatomic barriers that predispose to leakage of material already containing bacterial flora. The infectious complications following surgery vary depending on the site and extent of the operation, and the type of procedure performed; even so, postoperative infections have been shown in one series to be twice as common in cancer versus noncancer patients [73]. The site and specific type of surgery is an important determinant of infection. For example, intraabdominal procedures such as Hartman's operation, which involves sigmoid resection with a diverting colostomy [74], are frequently complicated by infection in patients with underlying malignancies [75]. Likewise, craniotomy in cancer patients who have previously had an arteriovenous shunt placed predisposes to an increased risk of meningitis andlor sepsis [76]. Extensive surgery of the paranasal sinuses has also been shown to predispose to Pseudomonas meningitis in these patients [77]. The extent of the operation plays a major role in determining infection. As expected, the largest interventions are associated with the maximum risk. Other factors such as obesity can also increase the infectious risk in these patients [78]. The surgical management of the patient with neutropenic enterocolitis (typhlitis) is a frequently encountered, although controversial, issue. A review of 438 leukemic patients demonstrated a 13% incidence of major gastrointestinal complications, and neutropenic enterocolitis was one of the most ominous [79]. Along with the increased risk of infection, these neutropenic and usually thrombocytopenic patients have a high risk of operative mortality from the surgery itself. Consequently, the care of these patients should be individualized. Non-operative management with bowel rest, decompression, nutritional support, and broad-spectrum antibiotics is usually recommended initially. Operative intervention is often reserved for patients with bowel perforation or for those whose condition clinically deteriorates despite conservative management [80].

Splenectomy, as previously outlined, increases the risk of infection by depressing cell-mediated and humoral immunity. In fact, except in young children, the splenectomy itself does not pose an increased risk of infection. In splenectomized patients, the specific type and severity of the underlying malignancy is the major determinant of infection [26,81]. 3.2 Radiation In addition to surgery, preoperative irradiation increases the risk of infection. In one series, preoperative irradiation given to patients undergoing surgery for breast cancer was associated with a twofold increase in infectious complications. However, postoperative irradiation was not associated with an increased risk [82]. Infection is also the most common complication in patients who receive preoperative irradiation prior to oncologic surgery of the upper respiratory or gastrointestinal tract. This is predominantly due to fistula formation or impaired wound healing [83]. In addition to causing local tissue damage, radiation can also result in stenosing lesions, leading to obstruction [$I. Radiation of the spleen or lymph nodes can depress cell-mediated immunity and antibody production. Total body irradiation predictably results in substantial depression of cellular immune function for months to years. Finally, radiation can also result in marrow depression and neutropenia [62]. 3.3 Chemotherapy

Chemotherapeutic agents predispose to infection in a variety of ways (Table 3). Many of these agents damage the body's anatomic barriers. Most notably, they can cause ulceration of the gastrointestinal tract, allowing for erosion and invasion by the endogenous microorganisms. Two agents, bleomycin and methotrexate, are associated with skin lesions, which can predispose to bacteremia with staphylococci and other skin organisms. BCNU, ara-C, and daunomycin irritate veins, increasing the risk of phlebitis and subsequent bacteremia. Most chemotherapies cause bone marrow suppression and neutropenia in a dose-related fashion. Some of these drugs can also inhibit neutrophilic migration and chemotaxis. Regimens that include corticosteroids inhibit the bactericidal activity of neutrophils. Humoral immunity is altered by agents such as methotrexate, cyclophosphamide, and 6-mercaptopurine. Other agents, such as interleukin-2 and deferoxamine, are associated with specific types of infections. Interleukin-2 has been linked to increased infection, particularly with staphylococci. This is predominantly due to druginduced defects in neutrophil function. Deferoxamine is associated with increases in bacterial infections and zygomycosis, most likely due to the increased availability of free iron necessary for fungal growth [8]. Currently, deferoxamine is being used experimentally to treat neuroblastomas, and a case of Fusarium in a child receiving deferoxamine for this indication was recently reported [$4].

Table 3. Chemotherapeutic agents that predispose to infection

Ulceration of the gastrointestinal tract Actinomycin D Adriamycin Cyclophosphamide Cytosine arabinoside Daunomycin 5-Fluorouracil Hydroxyurea 6-Thioguanine Skin lesions/ulceration Bleomycin Methotrexate Vascular irritationiphlebitis BCNU Ara-C Daunomycin Dose-related bone marrow suppression Adriamycin Busulfan BCNU Methyl CCNU Cisplatin Chlorambucil Cyclophospharnide Daunomycin Hexamethylmelamine Hydroxyurea Melphalan Methotrexate Nitrogen mustards Procarbazine 6-Thioguanine Thiotepa Vinblastine Vincristine Inhibition of neutrophilic function Corticosteroids Humoral immunodeficiency Methotrexate Cyclophosphamide 6-Mercaptopurine Other Interleukin-2 Deferoxamine

3.4 Antibiotics

Antibiotics used for both prophylaxis and treatment alter a patient's microflora and select for resistant organisms. Examples of this have been proven repeatedly in cancer patients. Trials of nonabsorbable antibiotics were previously used to decrease colonization of the alimentary canal. These trials were halted in part due to the emergence of resistant organisms [85-871.

The use of fluconazole for fungal prophylaxis in bone marrow transplant recipients has been associated with outbreaks of C. krusei infections [71,88]. The use of fluoroquinolones in neutropenic cancer patients is associated with an increase in infection with resistant staphylococci, streptococci, and anaerobes, as well as increased fluoroquinolone resistance in the gram-negative rods [17,89]. The total amount of ceftazidime, the duration of treatment, and the number of days of therapy with this agent have all been implicated in the emergence of vancomycin-resistant Enterococcus faecium bacteremia [90,91]. Although antibiotic use in cancer patients is essential in many situations, the emergence of resistant organisms can be directly linked to antibiotic selective pressure. These antibiotic-resistant organisms pose a major health threat to all patients, especially to those who are immunocompromised.

3.5 Diagnostic and invasive procedures Diagnostic and invasive procedures are well-recognized factors predisposing to infection. Any procedure that breaks the natural protective barrier between the internal and external environment can allow bacteria or fungi to enter the bloodstream. Biopsies, bone marrow aspirations, endoscopy, and indwelling urinary catheters are but a few examples. Strict attention to sterile technique, when applicable, can decrease but cannot completely eliminate the infectious risk associated with these procedures.

3.6 Central venous catheters Semipermanent centrally placed venous access devices, such as those developed by Broviac and Hickma.n, are commonly used in cancer patients to administer chemotherapy, blood products, antibiotics, and parenteral nutrition, and to obtain blood specimens for laboratory analyses. Infection is a common and potentially life-threatening complication of these vascular access devices. The most commonly isolated pathogens are staphylococci and streptococci, although infections with Bacillus spp., Corynebacteriurn spp., and yeast such as C. albicans are also frequently diagnosed [92,93]. The rapidly growing nontuberculous mycobacteria, M. chelonei and M. fortuiturn, have also been associated with exit-site or tunnel infections around these intravascular catheters E94f. Almost all of these long-term intravenous devices become colonized within a few days of insertion; however, only 15-20% develop a clinically significant infection. On average, when correlated with the number of catheter days, the incidence of infection is approximately 1.37 episodes per 1000 patient days of long-term catheter use [95]. Data from Memorial Sloan-Kettering Cancer Center suggest that completely implanted ports (Port-A-Caths) are less prone to device-related infection than are intravenous catheters, probably due to the more frequent manipulation of the Hickman or Broviac-type catheters when not in use [92]. Data on peripherally inserted central catheters (PICCs)

suggest that infection rates are lower than for the centrally placed venous catheters. One study of PICC lines in critical care patients found the incidence of infection to be 0.48% per 1000 catheter days [96]. However, PICC lines are more prone to certain insertion and maintenance problems, such as catheter fracture, phlebitis, and occlusion [97,98]. 3.7 Blood transfusions and bone marrow infusions Nosocomially acquired infections from blood transfusions occur despite modern blood banking techniques designed to prevent this complication. Cancer patients, especially those with hematologic malignancies or those undergoing bone marrow transplantation, may require several transfusions during the course of their illness, and thus are at increased risk of transfusion-related infection. Contamination of blood products can occur during processing and storage, but most commonly occurs through collection of blood from infected donors [99]. For an organism to cause an infection in a transfused patient, it must (1) be present in the donor's blood at the time of collection while producing few or no symptoms; (2) escape detection by current screening methods; (3) remain viable in citrated, refrigerated blood for prolonged periods of time; and (4) be of sufficient virulence and quantity to produce infection in the transfusion recipient [100]. Viruses are the most frequently encountered pathogens associated with blood transfusions. These include hepatitis viruses, human immunodeficiency virus, Epstein-Barr virus, and cytomegalovirus (CMV). Although most of these are detected by present screening procedures, CMV remains a significant risk for cancer patients. Protozoal diseases, such as leischmaniasis, trypanosomiasis, Chagas' disease, and microfilarial infections, are acquired through transfusion in developing countries. An increased incidence of transfusion-related Chagas' disease has also been reported in the United States. Malaria is uncommon in the United States but does occur, especially in people who have returned from travel in endemic areas. Therefore, transfusion-related malaria remains a potential risk in this country [101-1041. Babesia microti, a tick-borne protozoan parasite, has been transmitted through transfusion along coastal regions of the northeastern United States. It can cause a life-threatening infection in immunocompromised, especially asplenic, patients [104,105]. The procedure of storing citrated blood at 4OC for prolonged periods has greatly reduced the risks of transfusion-transmitted bacterial infections. Although up to 6% of stored blood contains some form of bacterial contamination, most of these organisms are normal skin flora such as S. epidermidis and diptheroids, which do not cause significant infections in transfusion recipients. On the other hand, Pseudomonas fluorescens/putida and Yersinia enterocoliticn can survive and multiply in cold storage. These organisms have been associated with life-threatening sepsis following blood transfusions

[106,107]. Platelets are often stored at room temperature to enhance their post-transfusion function. Thus, bacterial infections are more likely to occur following platelet transfusions [108]. With increasing numbers of patients undergoing bone marrow transplantation, the transmission of infectious agents during bone marrow reinfusion needs to be considered. To minimize this risk, strict attention must be paid to the health status of the donor and to the procedures for processing and storing of specimens in the time between donation and transplantation.

4. Infections related to underlying malignancies Infections in cancer patients are largely determined by the underlying malignancy and the treatments directed against it. Understanding the risk factors that predispose to infection and applying this knowledge to a specific clinical situation helps to logically guide diagnosis and therapy (Table 4). 4.1 Acute leukemia and lynzphorna

Patients with acute leukemia and lymphoma who are neutropenic, either due to their underlying disease or to cytotoxic chemotherapy, are at risk for a different set of infections than those who are not neutropenic. Classically in neutropenic patients, gram-negative bacilli such as E. coli, Klebsielln spp., and P. nevuginosa cause the earliest infections. These usually occur within the first 2-3 weeks after the initiation of chemotherapy and are due to the rapid decrease in the neutrophil count. These infections are characterized by acute febrile episodes, which can progress to overwhelming sepsis if not treated promptly [13,15,109-1141. During the 1980s however, investigators noted a relative decrease in the number of gram-negative bacteremias and a significant increase in infections caused by gram-positive aerobic bacteria, namely, staphylococci and streptococci. Several reasons for this observation have been postulated [115,116]. The use of both prophylactic and empiric antibiotic regimens targeting gram-negative bacteria diminishes recovery of gram-negative pathogens while selecting for gram-positive infections [117,118]. One example is the emergence of streptococcal infections in populations of patients receiving fluoroquinolones [119]. The use of intravascular catheters also increases the likelihood of infection with gram-positive bacteria, such as staphylococci, that colonize the skin [1171. Chemotherapeutic regimens that cause oral mucositis predispose to infection with bacteria that ordinarily colonize the oropharynx, namely, alpha-hemolytic streptococci. Although the mortality associated with gram-positive infections is less than that of gram-negative, the morbidity is significant. For example, alpha-hemolytic streptococci have been associated with cases of adult respiratory distress syndrome (ARDS) in patients receiving cytarabine [119]. Furthermore, patients who remain

Table 4. Infections related to the underlying malignancy

Malignancy

Immunodeficiency

Infection

Neutropenia

Bacteria Gram-negative E. coli Klebsiella spp. P. rteruginosa Gram-positive S. rrrdrerts S. epirierrnidis Streptococci Y eastlfungi Canriich spp. Aspergilllrs spp. Viruses HSV

I. Acute leukemia and lymphoma

vzv

CMV Adenovirus Hepatitis A, B, C Cell-mediated immunity (in the nonneutropenic)

Bacteria L. motzocytogeaes Salmonella spp. N. asteroides Mycobacteria L. przez{rnoptlila Viruses HSV

vzv

CMV EBV Adenovirus Measles Yeastifungi C. neofornzans Aspergill~lsspp. Protozoa P. carinii T. gondii Cryptosporidi~4rn Helminth S. stercoralis 2. Chronic lymphocytic leukemia

Hypogammaglobulinemia As the disease progresses: combination chemotherapy, with or without steroids, splenectomy, and occasionally radiation therapy predispose to further opportunistic infections

S. ptteiimonirte H. inffllenzae E. coli

Table 4. (continued) Malignancy 3. Multiple myeloma

Immunodeficiency

Infection

Humoral immunodeficiency; comple~nentdeficiency

S. ynertmoniae H. influenzae N. nzeningiticlis

Neutropenia in late-stage disease

Gram-negative bacilli

Cell-mediated immunodeficiency

Bacteria Salmonella spp. L. nzonocytogenes M. kansasii MAC M. chelonei Yeast Candida spp. C. 17eoforrrzans Viruses HSV CMV

Neutropenia

Gram-negative bacilli

Disruption of anatomic barriers

Skin Staphylococci Streptococci

4. Hairy cell leukemia

5. Solid organ tumors

Oral cavitylnasopharynx Anaerobes Streptococci H. influenzne GI tract Enterobacteriaceae Fungi Female gential tract Enterobacteriaceae Anaerobic GNB Enterococci Clostridiunz spp. Mechanical obstructions

Biliary, urinary, and respiratory tract infection, and vascular obstruction

Loss of gag reflex Impaired micturition Impaired mobility

Aspiration pneumonia Recurrent UTIs Decubitus ulcers with or without osteomyelitis

CNS tumors

CMV = cytomegalovirus; EBV = Ebstein-Barr virus; GNB = gram-negative bacteria; HSV = herpes simplex virus; MAC = Mycobncteri~unaviltm complex; UTIs = urinary tract infections; VZV = varicella zoster virus.

neutropenic for prolonged periods of time are more likely to develop infections with multidrug-resistant bacteria such as Corynebacterium jeikeium, Serratia spp., Enterobacter spp., Acinetobacter spp., Pseudomonas cepacia, and Stenotrophomonas (Xanthomonas) maltophilia. These emerge as a consequence of protracted courses of broad-spectrum antibiotics [111,112,120]. Because neutrophils play a major role in controlling infections due to Candida and Aspergillus, invasive fungal infections are also frequently encountered in neutropenic patients [121-1261. Autopsy series have documented invasive fungal infections in 1 0 4 0 % of patients with underlying hematologic malignancies (see Chapter 6) [27,127-1291. Besides prolonged granulocytopenia, extended hospital stays, previous antibiotics, corticosteroids, central venous catheters, and total parenteral nutrition are risk factors for fungemia. In addition, many other uncommon fungi have been reported to cause infection in this patient population, as is discussed later in this chapter. Viruses commonly infect neutropenic hosts (see Chapter 7). Reactivation of HSV is by far the most common viral infection encountered. VZV, CMV, adenovirus, and the viral hepatitides have also been reported in the neutropenic patient with acute leukemia or lymphoma [130-1371.

4.2 Nonneutropenic patient Leukemia or lymphoma patients who are not neutropenic demonstrate deficiencies in cell-mediated immunity, due either to their underlying disease or to the treatment regimens they receive. This cellular immunodeficiency predisposes them to infections with a variety of intracellular organisms. In addition, many of them have undergone splenectomy, putting them at risk for bacteremia with the encapsulated bacteria. Bacterial infections in this subgroup of patients are most commonly caused by Listeria monocytogenes, Legionella pneumophilin, Salmonella spp., Mycobacterium tuberculosis, the nontuberculous mycobacteria, and Nocardia spp. [lo]. Patients who have undergone splenectomy are at risk for Streptococcus pneumoniae, N. rneningitidis, and H. influenzae [81]. Fungal infections other than Cryptococcus and occasionally Aspergillus are uncommon [138]. Protozoal infections, on the other hand, are much more common in the nonneutropenic patient. Infections with Pneumocystis carinii, Toxoplasma gondii, Strongyloides stercoralis, and Cryptosporidium have all been reported [42-441. Viruses such as HSV, VZV, and CMV are all encountered in these patients [lo], and cases of measles have been reported in nonneutropenic children with acute lymphocytic leukemia [136]. 4.3 Chronic lymphocytic leukemia

Chronic lymphocytic leukemia (CLL) represents a clonal expansion of neoplastic B lymphocytes in inore than 95% of cases [139]. These mature-

appearing B lymphocytes are found in the peripheral blood. They also infiltrate the bone marrow, spleen, and lymph nodes. Much of the gamma globulin produced by patients with CLL is nonfunctional, leading to defects in humoral immunity. This humoral immunodeficiency worsens as the disease progresses and does not revert after chemotherapy. CLL patients are at risk for infection with encapsulated pyogenic bacteria such as S. pneumoniae and H. influenzae, as well as E. coli. As the malignancy progresses, treatment modalities may include corticosteroids, combination chemotherapy, splenectomy, and occasionally radiation therapy for control of localized disease, all of which increase the risk of opportunistic infections [62]. 4.4 Multiple myeloma

Like CLL, patients with multiple myeloma (MM) classically present with defects in humoral immunity. MM patients are hypogammaglobulinemic, producing normal immunoglobulins at only 10% the normal rate. Therefore, they are predisposed to infections with the encapsulated bacteria such as Streptococcus pneurnoniae, H. influenzae, and N. meningitidis [48,49]. As disease progresses, the malignant plasma cells proliferate within the bone marrow to such an extent that the marrow is unable to produce adequate numbers of neutrophils. Therefore, patients with advanced disease may become neutropenic, increasing their risk of gram-negative bacterial infections [50].

4.5 Hairy cell leukemia This chronic B-cell lymphoproliferative disorder presents with cytopenias in the majority of patients. In particular, patients have monocytopenia, granulocytopenia, and defective T-cell function. This results in a cellular immunodeficiency and predisposes to a variety of infections. As in other patients, the neutropenia predisposes to gram-negative bacterial infections. Defects of cell-mediated monocytelmacrophage and T-cell function predispose to other bacterial infections with organisms such as Salmonella and Listeria; fungal infections with Candida and Cryptococcus; viral infections with HSV and CMV; and nontuberculous mycobacterial infections with M. kansasii, M. avium complex, and M. chelonei [140,141]. In one review from the University of Chicago, five of nine hairy cell leukemia patients with nontuberculous mycobacterial infections had disseminated disease at presentation [142]. 4.6 Solid organ tumors

Patients with solid organ tumors do not have the same risk of infection as patients with underlying hematologic malignancies. This is largely because the standard chemotherapeutic regimens used to treat these malignancies do not usually result in either long-term or profound levels of neutropenia. Excep-

tions include patients with small cell carcinoma of the lung, testicular carcinoma, and some sarcomas. Aggressive chemotherapeutic regimens used to treat these malignancies may result in periods of neutropenia for 7-10 days or more [62]. Likewise, malignancies such as metastatic carcinoma of the breast, prostate, lung, adrenal, thyroid, and kidney have a propensity to infiltrate the bone marrow and can result in neutropenia in the advanced stages of disease. Patients with tumors of the central nervous system, either primary or metastatic, are at risk for a unique set of infections based on the associated neurologic deficit. An impaired gag reflex can lead to aspiration pneumonia, impaired micturition to recurrent urinary tract infections, and impaired mobility to decubitus ulcers with or without osteomyelitis. Any solid organ tumor that invades and disrupts anatomic barriers may predispose to infection. These include tumors of the skin, oral cavity, nasopharynx, gastrointestinal, respiratory, and urogenital tracts. These malignancies and their associated pathogens were discussed previously. 4.7 Bone marrow transplantation The patient who has undergone bone marrow transplantation serves as an excellent example of how various host immunodeficiencies predispose to specific infectious pathogens. From the standpoint of infection, the posttransplant course can be divided into three separate time periods (Table 5). The first 20-30 days following transplantation, the pre-engraftment period, is characterized by a precipitous loss of circulating granulocytes. Also during this time, many of the anatomic barriers have been disrupted, such as the oral and gastrointestinal mucosa, resulting in a predominance of bacterial and fungal infections. For the most part, these organisms include staphylococci, streptococci [143], Corynebacterium, Propionibacterium, E. coli, Klebsiella, and Pseudomonas, followed by Candida and Aspergillus as prior antibiotic therapy shifts the local flora [144,145]. Reactivation of HSV is also common during this stage. By 20-30 days post-transplant, most patients have recovered their granulocytes. The second period, early postengraftment, encompasses the time from recovery of circulating granulocytes through day 100. As the neutropenia resolves, the patient's clinical condition improves rapidly; however, neutrophil function, especially chemotaxis, is not entirely normal. The recovery of cellmediated immunity is also delayed. The proportion and sometimes the absolute number of T cells may be abnormal for prolonged periods, occasionally for several years after transplantation. Therefore, bacterial and fungal infections may continue to occur, especially during the early part of the second stage. However, viral and protozoal infections increase in incidence during this stage [146,147]. A diffuse interstitial pneumonia of unknown etiology has also been reported during this period [148,149]. The pathogens to expect 30-100 days post-transplant include CMV, Pneumocystis carinii, toxoplasmosis, and cryptosporidiosis.

Table 5. Time course for infection in bone marrow transplant recipients Day post-transplantation Pre-engraftment 0-30

Early post-engraftment 30-100

Immunodeficiency

Common pathogens

Granulocytopenia and disrupted anatomic barriers

Staphylococci Streptococci Corynebacterium spp. Propionobacteritlm spp. E. coli Klebsiella spp. P. aeruginosn Candidn spp. Aspergillus spp. Reactivation of HSV

Delayed recovery of neutrophil function, esp. chemotaxis

Bacteriallfungal infections may continue to occur as in the pre-engraftment stage CMV, PCP, T. gondii Cryptosporidiosis, diffuse interstitial pneumonitis (unclear etiology)

Cell-mediated immunodeficiency

Late post-engraftment >I00

Hypogammaglobulinemia; functional asplenia with decreased IgG and IgM levels for 6-8 months; decreased IgA levels for 1 year or longer If GVHD develops, immunologic function recovers more slowly and patients are susceptible to opportunistic infections for longer periods of time

VZV; S, pneumoniae; chronic viral hepatitides, such as HCV acquired in the early pre-engraftment period

HCV = hepatitis C virus; GVHD = graft-versus-host disease; CMV herpes simplex virus; PCP = Pnel~mocystiscarinii pneumonia; VZV

= =

cytomegalovirus; HSV varicella zoster virus.

=

The third time period, after day 100, is the late postengraftment period. It is characterized by low levels of circulating immunoglobulins. IgG and IgM levels may remain depressed for 6-8 months or longer, and IgA levels may remain depressed for at least 1 year. Patients who develop chronic GVHD recover their immunologic function more slowly and are susceptible to opportunistic infections for longer periods of time. Characteristic infections occurring after day 100 include VZV and bacteremic pneumococcal infection. Nearly 40% of bone marrow transplant patients develop VZV infection, and the median time of onset is 5 months after transplantation. The incidence of pneumococcal bacteremia relates to chronic GVHD-induced hyposplenism with loss of opsonizing antibodies to the encapsulated gram-positive organisms. After day 100, infection with hepatitis C virus that may have been acquired through transfusions during the first 3 weeks after transplantation can become clinically manifest [SO-1521.

5. Emerging pathogens All clinicians caring for cancer patients must be aware of the emerging or newly recognized pathogens that are increasingly affecting this patient population. Infections with multidrug-resistant organisms are becoming more and more common. Liberal use of broad-spectrum antibiotics, prophylactic antimicrobial therapy, and complacency in prescribing techniques underlies the development of these resistant organisms [89,153,154]. Furthermore, organisms that were previously considered commensal or nonpathogenic have been shown to cause serious infections in these immunocompromised hosts. An awareness of emerging pathogens is essential for the diagnosis and management of patients with underlying malignancies (Table 6).

5.1 Bncteria 5.1.1 Aerobic gram-positive bacteria. Enterococci have emerged as important nosocomial pathogens. Estimates suggest that enterococci play a causal role in 12% of all hospital-acquired infections. Two species of enterococci, Enterococcus faecalis and Enterococcus faecium, account for the majority of clinical infections [90]. Vancomycin resistance among enterococci was first described clinically in 1988 [I551 and has since become a global problem. Indeed, many strains are not only vancomycin resistant but are multidrugresistant, and some multidrug-resistant strains of E. faecium are untreatable. Risk factors for invasive infection in oncology patients include the total amount and duration of antibiotic treatment, and the number of days of ceftazidime use [91]. Montecalvo and colleagues described an outbreak of multidrug-resistant E. faecium bacteremia on an oncology unit. Of the five patients who had multiple positive blood cultures, four died [156]. A report by Noskin and associates suggests that once patients are colonized with vancomycin-resistant enterococci (VRE), this colonization can last indefinitely and can later lead to invasive infection [157]. Although several antibiotics and antimicrobial combinations have been tried, standard therapeutic regimens have not been defined. Therefore, therapy may require unconventional or investigational agents [158]. Streptococci have recently received renewed interest as an important pathogen in the neutropenic host [56,159,160]. Streptococcus mitis, a viridans streptococci, has been associated with sepsis and ARDS in leukemic patients. The emergence of infection with both Streptococcus pneumoniae and the viridans streptococci has been demonstrated in cancer patients receiving Auoroquinolone prophylaxis. Fluoroquinolones, such as ciprofloxacin, have little activity against these organisms, giving them a selective advantage [161]. Streptococcus pneumoniae are becoming increasingly resistant to penicillin, and many strains are multidrug-resistant. Researchers at Northwestern University in Chicago have reported that in 1996, 35% of their strains possessed intermediate resistance to penicillin while 11% were highly penicillin resistant

[162]. Outbreaks of penicillin-resistant pneumococci on oncology wards have recently been reported. Because of this, when life-threatening infections with pneumococci are encountered or suspected, many centers are using either vancomycin or ceftriaxone as first-line therapy. Leuconostoc species, previously regarded as a commensal, have been reported to cause bacteremia in patients with intravascular catheters. These organisms can be confused microbiologically with enterococci, viridans Table 6. Emerging pathogens Pathogen

Comments -

Bacteria Aerobic gram-positive bacteria VRE Streptococci Viridans streptococci S, pneunzoniae Le~lcorlostocspp.

Aerobic gram-negative bacteria 3. ~1zr~lfop1ziLia GNB-containing ESBLs P. cepacia, M. exforqriens, A. radiobacfer, 0. rrnthropi, A. xylosoxidarzs A. plitre,facier.rs

Capnocytophagia Anaerobic bacteria C. septicrlm C. tertizirn F. nuclenninz, L, br~ccalis Mycobacteria M. tcrberculosis M. avirtm complex M. fortuirrrm, M. chelonae M. haemophil~im

No standard therapy; may require unconventional or investigational agents Increased incidence noted in patients receiving fluoroquinolone prophylaxis Associated with sepsis and ARDS in leukemic patients Increasingly penicillin and multidrug-resistant: empiric vancomycin or ceftriaxone used for life-threatening infections Inherently resistant to vancomycin; variably sensitive to penicillin and first-generation cephalosporins Mucositis is a risk factor, treat with penicillin or vancomycin Multiply drug resistant; remains vancomycin susceptible p-lactam antibiotics rarely effective in vitro: empiric therapy should consist of vancomycin or clindamycin with or without gentamicin 61 % survival with antibiotics alone; 75% for antibiotics plus surgical resection Nortoriously multidrug resistant; usually resistant to all aminoglycosides Plasmid-mediated resistance to many penicillinsithirdgeneration cephalosporins Associated with catheter-related bacteremias Fulminant syndrome of overwhelming sepsis and DIC in immunocon~promisedhosts Risk factors are mucositis and neutropenia Necrotizing enterocolitis Perirectal cellulitis Part of the normal oral flora; mucositis and pharyngitis are risk factors Combined medical and surgical approaches may be necessary in multidrug-resistant M. tuberculosis A c c o u ~ ~for t s 27% of nontuberculous mycobacteria in cancer patients Most commonly cause catheter infections Optimal treatment regimens not well defined

Table 6. (contirzued) Pathogen Viruses VZV, HSV CMV HTLV-I HTLV-I1 HIV KSHV (HHV-8) Fungi Candida spp.

Aspergillrrs spp. Fzisarircnz spp., Scoplt lariopsis spp., P. boydii

Comments Treatment with foscarnet in acyclovir-resistant strains Treatment with ganciclovir or foscarnet Associated with T-cell NHL Associated with hairy cell leukemia Associated with B-cell NHL Herpes virus associated with Kaposi's sarcoma

C. albicans still predominates, although non-albicans species are increasing At some institutions, A . ,flavus is now more common than A. filrnigafrrs Severe, often fatal infection in neutropenic hosts; recovery of neutrophil count is required for a successful outcome

Crlrv~rlariaspp.. Bipolaris spp., Exserohilum spp., Alternaria spp., Aspergillus spp.

Fungal sinusitis, which may lead to brain abscess

T, beigelii

Disseminated disease may occur; treatment is difficult and relapse is common Severe, disseminated infection in neutropenic patients Catheter-related sepsis in patients receiving parenteral lipids: sensitive to amphotericin B and the azoles; catheter removal and discontinuation of lipids Wide, deep surgical debridement or cryosurgery in early-stage disease; medical therapy has been disappointing All have been associated with catheter infections

B. capitatus M. firrfrlr

E. jeanselrnei, E. pisciphila, E. spinifera, S. inJIntzrm

R. rubra, H. anorrzalrt, G. candidrm, S. cerevi~eae. Drechslera spp., P. parasitica, Acrenzonirrn? spp., P. filrinosa P. carirzii Protozoa and parasites T. gondii

Other B. quintana. B. henselue

P, wickerhamii

Increased incidence in patients with brain tumors attributed to intensive chemotherapy and high doses of corticosteroids Primary treatment should be followed by maintenance therapy to prevent relapse in chronically immu~~osuppressed patients No reliable palliative or curative therapy exists Very high mortality in immunocompromised hosts; thiabendazole is the only effective therapy Causative agents of bacillary angiomatosis and bacillary peliosis; doxycycline or erythromycin is the treatment of choice Surgical debridement plus amphotericin B have been used successfully

ARDS = adult respiratory distress syndrome; CMV = cytomegalovirus; DIC = disseminated intravascular coagulation: ESBLs = extended-spectrum beta-lactamases; GNB = gram-negative bacteria; HHV-8 = human herpes virus 8; HSV = herpes simplex virus; HTLV = human T-cell lymphotropic virus; KSHV = Kaposi's sarcoma herpes virus; NHL = non-Hodgkin's lymphoma; VRE = vancomycin-resistant enterococci; VZV = varicella zoster virus.

streptococci, or lactobacillus. It is important to differentiate Leuconostoc because this organism is inherently vancomycin-resistant and only variably sensitive to penicillins and the first-generation cephalosporins [163]. Stomatococcus nzucilaginosus is a slime-producing, gram-positive coccus that is found in the normal oral flora of humans [164]. McWhinney and coworkers recovered S. m~ucilaginosus from eight febrile, neutropenic patients. Seven of these patients had bacteremia and one had positive cerebrospinal fluid cultures resulting in a fatal meningitis. Four of these patients had proven infections with this organism, and all eight had mucositis attributable to chemotherapy. All but one of these isolates was sensitive to penicillin, and all were sensitive to vancomycin [165]. Prior to 1976, Corynebacterium spp., bacteria normally abundant on the skin and mucous membranes, rarely caused infections and were susceptible to most antibiotics. However, in 1976, four cases of sepsis at the National Institutes of Health caused by Corynebacterium jeikeium were reported [166]. This organism is highly antibiotic-resistant but is vancomycin susceptible. Since then C. jeikeium has been increasingly recognized to cause intravascular catheter-associated bacteremias in neutropenic patients with underlying hematologic malignancies [167]. It has also caused meningitis and transverse myelitis in one neutropenic patient [168]. Cases of primary cutaneous Bacillus cereus infections have occurred in neutropenic patients with cancer or aplastic anemia. In these patients, the lesions were vesicular or pustular, only occurred on the limbs, and arose in the spring or summer. They all responded to antibiotic therapy [169]. B. cereus can also cause more severe clinical syndromes, such as necrotizing fasciitis, pneumonitis, and meningitis. In vitro data suggest that (3-lactam antibiotics are rarely effective; therefore, empiric therapy for a suspected infection with Bacillus spp. should consist of vancomycin or clindamycin with or without an aminoglycoside [170]. Rhodococcus equi is an uncommon pathogen that has been reported to cause infection in patients with impaired cellular immunity. HIV infection is the most common predisposing risk factor; however, infection in patients with other forms of cellular immunodeficiency have also been described. R. equi is most frequently associated with a cavitary pneumonia, which may mimic a fungal infection or tuberculosis. In a study by Harvey and Sunstrum, the survival rate for patients receiving antibiotics alone was 61% compared with 75% for those receiving both antibiotics and surgical resection [171].

5.1.2 Aerobic gram-negative bacteria. Although the overall incidence of gram-negative infections is decreasing in cancer patients, the gram-negative aerobic bacteria still cause significant morbidity and mortality. The use of broad-spectrum prophylactic and ernpiric antibiotics targeting the gramnegative bacteria has also led to increased antimicrobial resistance among these organisms.

Stenotrophomonas (Xanthomonas) maltophilia is an organism that is frequently isolated from the environment, particularly from water supplies. Both colonization and infection among immunocomprornised patients is increasing, especially in those receiving broad-spectrum antibiotics, particularly imipenam. S. maltophilia causes pneumonia, urinary tract infections, bacteremia, and wound infections in debilitated patients and is notoriously multidrug-resistant, making treatment difficult [I721. Enterobacteriaceae are also becoming more drug resistant. For example, a study by the International Antimicrobial Therapy Cooperative Group (IATCG) of the EORTC between 1983 and 1993 showed that the number of patients receiving fluoroquinolones had increased from 1.4% to 45%. During this same period, E. coli resistance to these drugs increased from 0% between 1983 and 1990 to 27% between 1991 and 1993 [17]. Certain Enterobacteriaceae, namely, E. coli and Klebsiella spp., have developed extended-spectrum beta-lactamases (ESBLs), enzymes that render them resistant to many penicillins and cephalosporins, particularly the third-generation cephalosporins. Although first- and second-generation cephalosporins may appear susceptible in vitro, clinical failures have been observed when these antibiotics are used. Therefore, imipenem, meropenem, or possibly a filactaml(3-lactamase inhibitor combination such as piperacillinltazobactam is now the treatment of choice for patients who develop life-threatening infections with these organisms. This antibiotic resistance is plasmid mediated and thus can be spread from patient to patient, leading to nosocomial outbreaks [173-1751. Therefore, for patients infected or colonized with bacteria expressing ESBLs, contact isolation is appropriate. Several unusual gram-negative pathogens have been associated with catheter-related bacteremia in cancer patients. An outbreak of Pseudomonas cepacia occurred on an oncology ward related to a contaminated heparin solution that was used to flush central venous catheters [176]. Methylobacterium extorquens bacteremia occurred in three acute leukemics, two of whom had undergone bone marrow transplantation and one who was receiving consolidation chemotherapy [177]. Catheter-associated bacteremia has also been reported with Agrobacteriuvrz radiobacter. This organism has also been associated with peritonitis, urinary tract infection, and endocarditis [178]. The first case of Ochrobactrum anthropi infection was reported in a 3year-old girl undergoing chemotherapy for retinoblastoma [179]. The experience with Achromohacter xylosoxidans bacteremia at M . D. Anderson Cancer Center from 1983 to 1988 has been reviewed. During this period, 10 cancer patients had positive blood cultures for this organism. In four, the infection was believed to be catheter related. An additional four had associated gastrointestinal pathology, and in the remaining two the predisposing factor could not be determined. Interestingly, neutropenia did not appear to be a risk factor [180]. Alteromonas (Pseudornonas) putrefaciens is an unusual organism that causes two distinct syndromes of bacteremic infection. One syndrome is

associated with chronic lower extremity infection, is mild, and responds to antimicrobial therapy. The second, more fulminant syndrome is associated with severe underlying illnesses such as liver disease and malignancy. These patients are more likely to experience overwhelming sepsis and disseminated intravascular coagulation [181]. Capnocytophagia is a facultative anaerobe that constitutes part of the normal oral flora and is an unusual cause of systemic infection. Capnocytophagia bacteremia has been reported following autologous bone marrow transplantation for Hodgkin7sdisease. Infection followed pretreatment conditioning that was associated with severe oral mucositis and neutropenia [182].

5.1.3 Anaerobic bacteria. Traditionally, anaerobes have accounted for relatively few primary infections in cancer patients, except as part of mixed infections, such as necrotizing gingivitis, perianal cellulitis, or perirectal abscesses. However, recently, a few anaerobes have been recognized as important pathogens in these immunocompromised hosts. Two species of Clostridium, C. septicurn and C. terti~1n.1,can cause significant infections in cancer patients. C. septicum has been known to cause necrotizing enterocolitis, and C. tertiurn has been isolated in neutropenic patients with perirectal cellulitis or another presumed gastrointestinal tract source [183,184]. Fusobacterium nucleat~tnzwas recovered from a leukemic patient with ulcerative pharyngitis and nodular pulmonary infiltrates suggestive of septic emboli [185,186]. Leptotrichia buccalis, part of the normal oral flora, has caused bacteremia in neutropenic patients whose only predisposing factor was mucositis [1871.

5.1.4 Mycobacteria. Strains of Mycobacterium tuberc~llosis resistant to isoniazid and many other first-line agents have recently emerged. When infected with tuberculosis, the immunocompromised cancer patient is at increased risk for disseminated or miliary disease. The second-line agents that must be used, in many instances, are less effective; thus, combined medical and surgical approaches may be necessary in these patients. Mycobacteriurn avium complex (MAC) is most commonly seen in AIDS patients but has also been reported in patients with underlying malignancies. MAC accounts for 27% of nontuberculous mycobacterial infections in cancer patients, especially patients with lung cancer. The most common clinical presentation is pulmonary disease; however, dissemination can occur [188]. M. fortuiturn and M. chelonae, two rapidly growing strains of mycobacteria, are known to cause catheter-related infections. They cause disseminated disease less commonly than MAC, but often result in cellulitis, skin abscesses, or subcutaneous nodules. Treatment of the catheter infections requires antimicrobial therapy, catheter removal, and surgical excision if tunnel infection is present. For other skin and soft tissue infections, a combination of antimicrobial therapy and surgical debridement may be required [189,190].

Reports of M. haernophil~lminfection are increasing in lymphoma patients, bone marrow and renal transplant patients, and AIDS patients. Infection with this mycobacteria most commonly presents as cutaneous ulcerations, joint effusions, or osteomyelitis. Optimal treatment regimens are not well defined [I911.

5.2 Viruses Varicella-zoster virus, HSV, and CMV are well-known pathogens in cancer patients. However, prophylactic use of acyclovir and ganciclovir has led to both acyclovir and ganciclovir-resistant strains, which must then be treated with foscarnet [192]. Recently, cidofivir became available for therapy of CMV retinitis in AIDS patients [193,194]. Retroviruses have been increasingly recognized as important pathogens in cancer patients over the past decade. Human T-cell leukemia virus-I (HTLV-I) is associated with adult T-cell non-Hodgkin7slymphoma, HTLV-I1 with hairy cell leukemia, and HIV with adult B-cell non-Hodgkin's lymphoma. Recently, a member of the gamma-herpesvirus family, referred to as Kaposi's sarcoma-associated herpesvirus (KSHV) or human herpesvirus 8 (HHV-8), has been associated with four types of Kaposi's sarcoma: classic KS, AIDS-associated KS, post-transplantation KS, and KS that occurs in endemic areas for the disease. It has also been associated with body-cavity-based lymphomas [195,196].

5.3 Fungi Candida spp. and Aspergillus spp. are the most common fungal infections in immunocompromised cancer patients, and recently several important trends have been noted. The incidence of nosocomial candidal bacteremia rose sharply in the late 1980s and early 1990s. At some institutions, Candidn bacteremia now surpasses that of the Enterobacteriaceae, Pseudornonas spp., and Entvrococc~lsspp. [197]. Candida albicans is still the most common species, accounting for more than half of fungal isolates from cancer patients, although the incidence of non-albicans species is increasing. Among these are C. tropicalis, C. parpsilosis, C. krusei, and C. (Torulopsis) glnbmta [198,199]. Central venous catheters, total parenteral nutrition, and the increasing use of azoles for antifungal prophylaxis are some of the presumed mechanisms thought to account for this rising trend. Of the Aspergillus species, Aspergillus fumigatus is the most commonly isolated species to cause invasive disease. However, at some institutions, A. flavus has supplanted A. fumigatus as the most common cause of aspergillosis [200]. Many fungi are ubiquitous in nature and were thought, until recently, to be commensal or nonpathogenic. Several of these have now been proven to cause serious infections in immunocompromised cancer patients. Fusarium spp., Scopulariopsis spp., and Pseudallescheria boydii, members of the

hyalohyphomycoses group, are important pathogens. Fusarium causes severe, often fatal, infections in neutropenic patients, particularly those who have undergone bone marrow transplantation. Fusariurn is highly resistant to conventional antifungal drugs, and rising neutrophil counts are required for a successful response. Subsequent neutropenic episodes are associated with a high incidence of recurrence [201,202].Scopulariopsis spp. and P. boydii have likewise been shown to cause serious disseminated disease in these patients [200]. Among the phaeohyphornycoses, Cttrvt~laria spp., Bipolaris spp., Exserohilum spp., and Alternaria spp. are known pathogens. These darkwalled molds are responsible for allergic fungal sinusitis. Patients may present with an indolent onset of sinus pain or painless proptosis. These infections can extend from the ethmoid or frontal sinuses into the frontal lobe causing brain abscesses. Differentiation of these fungi from Aspergillus is important. Other clinical syndromes, such as subcutaneous abscesses, cutaneous granuloma, disseminated disease, pneumonitis, prosthetic valve endocarditis, osteomyelitis, and septic arthritis, have been reported [203]. Trichosporon beigelii (T. cutaneum) can be part of the normal human flora. Patients with leukemia or bone marrow aplasia may develop disseminated trichosporonosis. In patients with fungemia due to T. beigelli, multiple red papular skin lesions may develop. An infection of hair shafts, known as white piedra, has also been described. Treatment of these infections is difficult and relapse is common [204-2061. Blastoschizornyces capitatus (Trichosporon capitatum) has caused severe, disseminated infection in at least 25 patients, 19 of whom were neutropenic. Six of these patients had papulonecrotic skin lesions similar to those seen in disseminated candidiasis in leukemic patients [207]. Malassezia furfur, a lipophilic yeast, colonizes normal skin and is also the causative agent of tinea versicolor. In cancer patients, it has been associated with catheter-related sepsis in patients receiving parenteral lipids. In vitro, it is susceptible to amphotericin B and irnidazoles. Therapy requires both catheter removal and discontinuation of parenteral lipids [208]. The dematiaceous soil fungi, Exophiala jeanselrnei, E. pisciphila, E. spinifera, and Scedosporiurn inflat~~rn, have caused infection in neutropenic patients [209]. In particular, S. inflaturn has been associated with catheterrelated fungemia, retinal lesions, esophagitis, and hepatosplenic infections [210]. In vitro, these organisms are resistant to amphotericin B and fluconazole. In early stage disease, wide and deep surgical excision or cryosurgery are most effective. Medical therapy has been disappointing. In small series, long-term itraconazole therapy has shown some success; however, relapse after treatment has been reported [211]. Several other fungi have been associated with infections in immunocompromised patients. Some of these include Rhodotorula rubra, Hansen~tla anomala, Geotrichurn candidurn, Saccharornyces cereviseae, Drechslera spp., Phialophora parasiticia, Acrernoniurn spp., and Pichia farinosa. Infections

with these organisms are primarily associated with central venous catheters [212,213]. Recently, Pneumocystis carinii has been reclassified as a fungus based on DNA homology. This organism results in pneumonia and is the most common opportunistic infection in HIV-infected patients. Although infrequent, the incidence is increasing in patients with underlying malignancies. In the 1970s, most cancer patients who developed Pneumocystis carinii pneumonia (PCP) had hematologic malignancies. However, in the 1980s the incidence of PCP increased most among patients with solid organ tumors. Indeed, 31% of patients who developed PCP had primary or metastatic brain tumors. The increase in patients with brain tumors has been attributed to more intensive chemotherapeutic regimens and the use of higher doses of corticosteroids [45,214]. The concomitant decrease in patients with hematologic malignancies and those undergoing bone marrow transplantation reflects the use of PCP prophylaxis in this patient population.

5.4 Protozoa and parasites

Toxoplasma gondii is an intracellular protozoan parasite. Reactivation of 7'. gondii has been reported in cancer patients. In fact, more than one third of the cases of toxoplasmosis in non-HIV patients occurs in patients with Hodgkin's disease or leukemia. Toxoplasmosis is associated with corticosteroid use and can present with pneumonitis, uveitis, or central nervous system disease. The diagnosis of reactivation disease includes an assay for specific toxoplasma IgG antibodies. However, in an immunocompromised patient, the IgG titer is often low and IgM assays may be negative. In this situation, a definitive diagnosis of toxoplasmosis depends on the demonstration of the tachyzoites on tissue histology or on specific T. gondii DNA by the polymerase chain reaction (PCR). Chronic (latent) asymptomatic infection in immunodeficient patients is not treated; however, immunodeficient patients with acute infection should always be treated. The first-line therapy includes pyrimethamine and folinic acid plus either sulfadiazine or clindamycin for 4-6 weeks after resolution of all signs and symptoms of disease (often for 6 months or longer). For chronically immunosuppressed patients, particularly AIDS patients, acute treatment is followed by life-long maintenance therapy to prevent relapse of disease [215]. Cryptosporidiurn is an intracellular protozoan parasite first described in 1907 [216]. The first cases of human diarrhea were described in 1976, and today it is most commonly associated with diarrhea in patients with HIV [217]. It has also been detected with increasing frequency in patients with underlying hematologic malignancies, those who have undergone bone marrow transplantation, and those who are receiving corticosteroid therapy. In one review of 20 patients with hematologic malignancies and cryptosporidiosis, 5 patients had severe diarrhea, 10 had moderate diarrhea, and 5 were asymptomatic carriers. Extraintestinal cryptosporidiosis with pulmonary involvement was

observed in one case. Resolution occurred spontaneously, although relapse was common, occurring in four patients [218]. Diagnosis is typically made by recognizing the characteristic oocysts in the material sampled, usually stool, duodenal aspirates, bile, or respiratory secretions. There is no reliable palliative or curative treatment for cryptosporidiosis. In immunocompetent or temporarily immunocompromised patients, the disease is self-limited. In patients who are irreversibly immunocomprornised, therapy is symptomatic and supportive. Strongyloides stercoralis is a nematode that can cause an overwhelming fatal infection in immunocompromised patients. S. stercoralis is increasingly recognized in patients with underlying hematologic malignancies, in patients treated with corticosteroids, and in patients with HIV. Humans are usually infected through skin contact with soil containing the infective filariform larvae or through autoinfection via the lower GI tract or perianal region by larvae that transform into infective organisms during their passage with feces. After infection and tissue invasion, the larvae enter the bloodstream and travel to the lungs. Here they break into the alveolar spaces and ascend to the glottis, where they are swallowed to reside in the small intestines. Contrary to most other worm infections, the patient's worm burden in strongyloidiasis is dependent not only on the size of the larval inoculum but also on the degree of autoinfection, which may be enhanced in immunocompromised hosts. This "autoinfection" cycle can result in an overwhelming larval invasion seen in strongylides hyperinfection syndrome, and these patients can develop massive larval invasion of the lungs and other tissues. This syndrome of hyperinfection strongyloidiasis is characterized by severe generalized abdominal pain, diffuse pulmonary infiltrates, ileus, shock, and meningitis or sepsis from gram-negative bacilli. Eosinophilia may be absent in the immunocompromised host. Definitive diagnosis is made by demonstrating the larvae in feces or in duodenal fluid. Thiabendazole is the only effective therapy; however, mortality remains very high in immunocomprornised patients. Therefore, patients with a past history of exposure to S. stercoralis should be thoroughly examined and treated prior to receiving any immunosuppresive therapy [219-2241.

5.5 Other pathogens 5.5.1 Bartonella spp. Bartonella (formerly Rickettsia) quintana and Bartonella (formerly Rochalimaea) henselae are slow-growing, motile, curved, gram-negative bacilli that have recently been recognized as human pathogens. Originally described in HIV patients, they have now been documented in cancer patients, including those with neutropenia. Both B. quintana and B. henselae can cause bacteremia or localized tissue infections, such as bacillary angiomatosis or bacillary peliosis. Bacillary angiomatosis was originally described in HIV patients as a neovascular proliferative disorder involving the

skin and regional lymph nodes. Characteristically, the lesions are cutaneous, can number from few to hundreds, bleed copiously when incised, and may resemble Kaposi's sarcoma, from which they need to be differentiated. Bacillary angiomatosis involving the liver, spleen, bone, and brain has also been documented. Bacillary peliosis is a distinct clinical syndrome that is less dramatic than bacillary angiomatosis because its lesions are exclusively visceral and are only associated with nonspecific symptoms. These organisms can be detected in blood cultures; however, they require prolonged periods of incubation. When they are suspected, the laboratory should be instructed to hold the blood cultures for up to 30 days. These organisms can also be detected in tissue biopsy specimens by Warthin-Starry staining or by electron microscopy. The initial treatment of choice is either oral doxycycline or oral erythromycin, often for prolonged periods of time. Tetracycline, minocycline, chloramphenicol, azithromycin, and clarithromycin have also been used successfully. Relapses can occur, and chronic suppressive therapy with doxycycline or erythromycin may need to be considered in these cases [225].

5.5.2 Prototheca wickerhamii. This unicellular, achloric algae is found in a wide range of environmental sites and has been isolated as the cause of infection in at least 61 cases. One was a child with Hodgkin's disease who had a central venous catheter in place and had P. wickerhamii isolated from the bloodstream [226]. The typical clinical presentation is a single lesion of the skin or subcutaneous tissue that gradually enlarges over weeks to months and may ulcerate. Characteristically, these lesions do not heal spontaneously. P. wickerhnnzii is resistant to fluconazole. Surgical debridement and amphotericin B have been used successfully [227-2291.

6. Summary Patients with underlying malignancies are at risk for a wide array of infectious diseases that cause significant morbidity and mortality. To develop a clear etiologic understanding of the infectious agents involved first requires a knowledge of the factors that predispose to infection. Neutropenia is clearly the single most important risk factor for infection in the cancer patient. However, a variety of both host and treatment-associated factors act together to predispose these patients to opportunistic infections. Approaching the individual malignancies with a knowledge of the underlying risk factors helps logically guide diagnosis and therapy. The astute clinician must also be aware of new and emerging infections in this patient population. As new pathogens are discovered and established pathogens become increasingly drug resistant, they will continue to present challenges for physicians caring for these patients in the years ahead.

References 1. Chang H, Rodriguez V, Narboni G, et al. Causes of death in adults with acute leukemia. Medicine 1976;55:259-268. 2. Hughes WT. Fatal infections in childhood leukemia. Am J Dis Child 1971;122:283-287. 3. Hersh EM, Bodey GP, Nies BA, et al. Causes of death in acute leukemia. JAMA 1965;193:99-103. 4. Morrittu L, Earl HM, Souhami RL, et al. Patients at risk of chemotherapy-associated toxicity in small cell lung cancer. Br J Cancer 1989;59:801-804. 5. Viscoli C, Castagnola E. Factors predisposing cancer patients to infection. In: Klastersky J, ed. Infectious Complications of Cancer. Boston: Kluwer Academic, 1995, pp. 1-30. 6. Bodey GP, Buckley M, Sathe YS, et al. Quantitative relationships between circulating leukocytes and infection in patients with acute leukemia. Ann Intern Med 1966;64:328340. 7. Feld R, Bodey GP. Infections in patients with malignant lymphoma treated with combination chemotherapy. Cancer 1977;39:1018-1025. 8. Glauser MP, Zinner SH. Mechanisms of acquisition and development of bacterial infections in cancer patients. In: Klastersky J, ed. Infections in Cancer Patients. New York: Raven Press, 1982, pp. 13-30. 9. De Jongh CS, Joshi JH, Newn~anKA, et al. Antibiotic synergism and response in gramnegative bacteremia in granulocytopenic cancer patients. Am J Med 1986;80:96-100. 10. Wade JC, Schimpff SC. Epidemiology and prevention of infection in the compromised host. In: Rubin RH, Young LS, eds. Clinical Approach to Infection in the Compromised Host. New York: Plenum Medical, 1988, pp. 5-40. 11. Dale DC, Guerry DI, Wewerka JR, et al. Chronic neutropenia. Medicine 1979;58:128-144. 12. Viscoli C, Bruzzi P, Castagnola E, et al. Factors associated with bacteraemia in febrile, granulocytopenic cancer patients. Eur J Cancer 1994;30A:430-437. 13. The EORTC International Antimicrobial Therapy Pro,ject Group. Three antibiotic regimens patients with cancer. J Infect Dis in the treatment of infection in febrile gra~~ulocytopenic 1978;137:14-29. 14. Schimpff SC, Young VM, Greene WH, et al. Origin of infection in acute nonlymphocytic leukemia: Significance of hospital acquisition of potential pathogens. Ann Intern Med 1972;77:707-714. 15. The EORTC International Antimicrobial Therapy Project Group. Ceftazidime combined with a short or long course of arnikacin for empirical therapy of gram-negative bacteremia in cancer patients with granulocytopenia. N Engl J Med 1987;317:1692-1698. 16. Johanson WG, Pierce AK, Sanford JP. Changing pharyngeal bacterial flora of hospitalized patients. Emergence of gram-negative bacilli. N Engl J Med 1969;281:1137-1140. 17. Cometta A, Calandra T, Bille J, et al. Escherichin coli resistant to fluoroquinolones in patients with cancer and neutropenia. N Engl 3 Med 1994;330:1240-1241. 18. Levine SS, Schimpff SC, Graw RG Jr., et al. Hematologic malignancies and other marrow failure states: Progress in the management of complicating infections. Semin Hematol 1974;11:141-202. 19. Chanock S. Evolving risk factors for i~~fectious con~plicationsof cancer therapy. Hematol Oncol Clin North Am 1993;7:771-793. 20. Koll BS, Brown AE. Changing patterns of infection in the immunocompromised patient with cancer. Hematol Oncol Clin North Am 1993;7:753-769. 21. Guiot HFL, Fibbe WE, van't Wout JW. Risk factors for fungal infection in patients with malignant hematologic disorders: Implications for empirical therapy and prophylaxis. Clin Infect Dis 1993;18:525-532. 22. Warren JR, Scarpelli DG, Reddy JK, et al. Chronic inflammation and repair. In: Essentials of General Pathology. New York: Macmillan, 1987, pp. 27-46. 23. Johnston RB, Jr. Monocytes and macrophages. N Engl J Med 1988;318:747-752.

24. Young RC, Corder MP, Haynes HA, et al. Delayed hypersensitivity in Hodgkin's disease. A study of 103 untreated patients. Am J Med 1972;52:63-72. 25. Seligmann M, Chess L, Fahey JL, et al. AIDS - An immunologic reevaluation. N Engl J Med 1984;311:1286-1292. 26. Aubertin J, Lacut JY, Hoerni B, et al. Factors favoring infections in cancer patients. In: Armstrong D, ed. Opportunistic Infections in Cancer Patients. New York: Masson, 1978, pp. 16-28. 27. Krick J, Remington J. Opportunistic fungal infection in patients with leukemia and Iymphoma. Clin Hematol 1976;5:249-310. 28. Lum LG. Immune recovery after bone marrow transplantation. Bone Marrow Transplant Hematol Oncol Clin North Am 1990;4:659-675. 29. Whittaker K, Rees K, Clark CG. Reduced lymphocyte transformation in breast cancer. Lancet 1971;1:892-893. 30. Han T, Takita H. Immunologic impairment in bronchogenic carcinoma: A study of lymphocyte response to phytohemagglutinin. Cancer 1972;30:616420. 31. Brooks WH, Netsky MG, Normansell DE, et al. Depressed cell-mediated immunity in patients with primary intracranial tumors. J Exp Med 1972;136:1631-1647. 32. Orita K, Miwa H, Fukada H, et al. Preoperative cell-mediated immune status of gastric cancer patients. Cancer 1976;38:2343-2348. 33. Schellhammer PF, Bracken RB, Bean MA, et al. Immune evaluation with skin testing. A study of testicular, prostate and bladder neoplasms. Cancer 1976;38:149-156. 34. Mackall CL, Fleisher TA, Brown MR, et al. Lymphocyte depletion during treatment with intensive chemotherapy for cancer. Blood 1994;84:2221-2228. 35. Mackall CL, FIeisher TA, Brown MR, et al. Age, thymopoiesis, and CD4+ T-lymphocyte regeneration after intensive chemotherapy. N Engl J Med 1995;332:143-149. 36. Kim JH, Perfect JR. Infection and cyclosporine. Rev Infect Dis 1989;11:677-690. 37. Bishop JF, Schimpff SC, Diggs CH, et ai. Infections during intensive chemotherapy for nonHodgkin's lymphoma. Ann Intern Med 1981;95:549-555. 38. Bookman MA, Longo DL. Concomitant illness in patients treated for Hodgkin's disease. Cancer Treat Rev 1986;13:77-111. 39. Hardy I, Gershon AA, Steinberg SP, et al. The incidence of zoster after immunization with Iive attenuated varicella vaccine. N Engl J Med 1991;325:1545-1550. 40. Hughes WT, Townsend TR. Nosocomial infections in immunocompromised children. Am J Med 1981;70:412-416. 41. Ampel NM, Wing EJ. Legionellosis in the compromised host. In: Rubin RH, Young LL, eds. Clinical Approach to Infection in the Compromised Host. New York: Plenum Medical, 1988, pp. 305-319. 42. Luft BJ, Remington JS. Toxoplasmosis of the central nervous system. Curr Clin Top Infect Dis 1985;6:315-358. 43. Longworth DL, Weller PF. Hyperinfection syndrome with strongyloidiasis. Curr Clin Top Infect Dis 1986;7:1-26. 44. Varthalitis I, Meunier F. Pneurnocystis carinii pneumonia in cancer patients. Cancer Treat Rev 1993;19:387-413, 45. Sepkowitz KA, Brown AE, Telzak EE, et al. Pneumocystis carinii pneumonia among patients without AIDS at a cancer hospital. JAMA 1992;267:832-837. 46. Sepkowitz KA. Pneumocystis carinii pneumonia in patients without AIDS. Clin Infect Dis 1993;17(Suppl. 2):S416-S422. 47. Hughes WT, Feldman S, Aur RJA, et al. Intensity of immunosuppressive therapy and the incidence of Ptzerlmocystis carinii pneumonitis. Cancer 1975;36:2004-2009. 48. Salmon SE. In~munoglobulinsynthesis and tumor kinetics of multiple myeloma. Semin Hematol 1973;10:135-147. 49. Fahey JL, Scoggins R, Utz JP, et al. Infection, antibody response and gamma globulin components in multiple myeloma and macroglobulinemia. Am J Med 1963;35: 698-707.

50. Jacobson DR, Zolla-Pazner S. Immunosuppression and infection in multiple myeloma. Semin Oncol 1986;13:282-290. 51. Mordasini RC, Keller H. Schlumpf E, et al. Humorale Infektabwherschwache bei paraproteinamischen Erkrankungen. Schewiz Med Wschr 1972:102:625-635. 52. Homberg JC, Cartron J, Ropars C, et al. Anomalies immunologiques observees au cours des leucemies lymphoides chroniques hyperlymphocytaires. Nouv Rev Fr Hematol1971;11:476483. 53. Sheridan JF, Tutschka PJ, Sedmak DD, et al. Immunoglobulin G subclass deficiency and pneumococcal infection after allogeneic bone marrow transplantation. Blood 1990;75:15831586. 54. Atkinson K. Reconstruction of the haemopoietic and immune systems after bone marrow transplantation. Bone Marrow Transpl 1990;5:209-226. 55. Axelrod PI, Lorber B, Vonderheid EC. Infections complicating mycosis fungoides and Sezary syndrome. JAMA 1992;267:1354-1358. 56. Awada A, van der Auwera P, Meunier F. et al. Streptococcal and enterococcal bacteremia in patients with cancer. Clin Infect Dis 1992;15:33-48. 57. Hosea SW, Brown EJ, Hamburger MI, et al. Opsonic requirements for intravascular clearance after splenectomy. N Engl J Med 1981;304:245-250. 58. Van Der Meer JWM. Defects in host-defense mechanisms. In: Rubin RH, Young LS, eds. Clinical Approach to Infection in the Compromised Host. New York: Plenum Medical, 1988, pp. 41-63. 59. Gopal V, Bison AL. Fulminant pneumococcal infections in "normal" asplenic hosts. Arch Intern Med 1977;137:1526-1530. 60. Advisory Committee on Itnmunization Practices. Pneumococcal polysaccharide vaccine. MMWR 1989;38:64-68, 73-76. 61. Kahls P, Kier P, Lechner K. Functional asplenia after bone marrow transplantation. Ann Intern Med 1990;114:805-806. 62. Schimpff SC. Infections in the cancer patient - Diagnosis, prevention, and treatment. In: Mandell GL, Bennett JE, Dolin R, eds. Principles and Practices of Infectious Diseases. New York: Churchill Livingstone, 1995. pp. 2666-2675. 63. Schwartz SN, Dowling JN. Benkovic C, et al. Sources of gram-negative bacilli colonizing the trachea of intubated patients. J Infect Dis 1978:138:227-231. 64. Kelly CP, Pothoulakis C, Lamont JT. Clostridi~~rn dificle colitis. N Engl J Med 1994;330:257262. 65. Anand A, Glatt A. Clostridiurn dificle infection associated with antineoplastic chemotherapy: A review. Clin Infect Dis 1993;17:109-113. 66. Fan-Havard P, Capano D, Smith SM, et al. Development of resistence in Cnndida isolates from patients receiving prolonged antifungal therapy. Antimicrob Agents Chemother 1978:35:2302-2305. 67. Meunier F, Aoun M. Bitar N. Candidemia in immunocompromised patients. Clin Infect Dis 1992;14(Suppl. 1):S120-S125. 68. Boken DJ, Swindells S, Rinaldi MG. Fluconazole-resistant Candidn albicans. Clin Infect Dis 1993;17:1018-1021. 69. Aisner J, Schmipff SC, Sutherland JC, et al. Torz~lopsisglabrata infections in patients with cancer. Am J Med 1976;61:23-28. 70. Merz WG, Karp JE, Schron D, et al. Increased incidence of fungemia caused by Cnndida krusei. J Clin Microbiol 1986;24:581-584. 71. Wingard JR, Merz WG, Rinaldi MG, et al. Increase in Cal~didakrrrsei infection among patients with bone marrow transplantation and neutropenia treated prophylactically with fluconazole. N Engl J Med 1991;325:1274-1277. 72. van der Waaij DD, Berghuis J, Lekkerkerk JEC. Colonization resistance of the digestive tract of mice during systemic antibiotic treatment. J Hyg (Camb) 1972;70:605-610. 73. Wilson BD, Surgalla MJ. Yates JW. Aerobic and anaerobic surgical wound contamination in patients with cancer. Surg Gpnecol Obstet 1974;139:329-332.

74. Rolandelli R, Roslyn JJ. The colon and rectum. Surgical anatomy and operative procedures. In: Sabiston DCJ, Lyerly HK, eds. Textbook of Surgery. The Biological Basis of Modern Surgical Practice. Philadelphia: W.B. Saunders, 1997, pp. 971-975. 75. Gongaware RD, Slanetz CAJ. Hartmann procedure for carcinoma of the sigmoid and rectum. Ann Surg 1973;178:28-30. 76. Naito H, Toya S, Shizawa H, et al. High incidence of acute post operative meningitis and septicemia in patients undergoing craniotomy with vertriculoatrial shunt. Surg Gynecol Obstet 1973;137:810-812. 77. Geelhoed GW, Ketcham AS. Pseudomonas meningitis complicating radical resection for radiorecurrent cancer of paranasal sinuses: Report of two patients successfully treated with intrathecal polymyxin. J Surg Oncol 1973;5:365-374. 78. Harris M, Gumport S, Berman I, et al. Ilioinguinal lymph node dissecton for melanoma. Surg Gynecol Obstet 1973;136:33-39. 79. Villar HV, Warneke JA, Peck MD, et al. Role of surgical treatment in the management of complications of the gastrointestinal tract in patients with leukemia. Surg Gynecol Obstet 1987;165:217-222. 80. Wade DS, Nava HR, Douglass HOJ. Neutropenic enterocolitis. Clinical diagnosis and treatment. Cancer 1992;69:17-23. 81. Schimpff SC, O'Connel MJ, Greene WH, et al. Infections in 92 splenectomized patients with Hodgkin's disease. Am J Med 1975;59:695-701. 82. Say CC, Donegan W. A biostatistical evaluation of complications from mastectomy. Surg Gynecol Obstet 1974;138:370-376. 83. Smith MGM, Golding PL, Eddleston ALWF, et al. Cell-mediated immune responses in chronic liver diseases. Br Med J 1972;1:527-530. 84. Castagnola E. Garaventa A , Conte M, et al. Survival after fungemia due to Frlsariurn monilifornte in a child with neuroblastoma [letter]. Eur J Clin Microbiol Infect Dis 1993;12:308-309. 85. Wade JC, Schimpff SC, Hargadon MT. et al. A comparison of trimethoprimsulfamethoxazole plus nystatin with gentamicin plus nystatin in the prevention of infections in acute leukemia. N Engl J Med 1981:304:1057-1062. 86. Schimpff SC, Greene WH, Young WM, et al. Infection prevention in acute nonlymphocytic leukemia. Ann Intern Med 1975;82:351-358. 87. Hathorn JW. Critical appraisal of antimicrobials for prevention of infections in immunocompromised hosts. Hematol Oncol Clin North A m 1993;7:1051-1099. 88. McQuiHen DP, Zingman BS, Meunier F, et al. Invasive infections due to Candida krusei: Report of ten cases of fungemia that include three cases of endophthalmitis. Clin Infect Dis 1992;14:472-478. 89. Shlaes DM, Binezewski B, Rice BL. Emerging antimicrobial resistance and the immunocornpromised host. Clin Infect Dis 1993;17(Suppl. 2):S527-536. 90. Stosor V , Noskin G A , Peterson LR. The management and prevention of vancomycinresistant enterococci. Infect Med 1996;13:488,493-498. 91. Livornese LL, Dias S, Same1 C, et al. Hospital-acquired infection with vancomycin-resistant Enterococcrls fi~eciur?? transmitted by electronic thermometers. Ann Intern Med 1992;117:112-116. 92. Groeger JS, Lucas AB, Thaler HT, et al. Venous access devices in patients with cancer. Ann Intern Med 19933 19:1168-1174. 93. Tenney J H , Moody MR. Newman KA, et al. Adherent microorganisms on lumenal surfaces of long-term intravenous catheters. Arch Intern Med 1986;146:1949-1954. 94. Pizzo PA. Management of fever in patients with cancer and treatment-induced neutropenia. N Engl J Med 1993;328:1323-1332. 95. Clarke DE, Raffin TA. Infectious complications of indwelling long-term central venous catheters. Chest 1990;97:966-972. 96. Abi-Nader JA. Peripherally inserted central venous catheters in critical care patients. Heart Lung 1993;22:428-434.

97. Lam S, Scannell R, Roessler D, et al. Peripherally inserted central catheters in an acute-care hospital. Arch Intern Med 1994;154:1833-1837. 98. Loughran SC, Borzatta M. Peripherally inserted central catheters: A report of 2506 catheter days. J Parent Ent Nut 1995;19:133-136. 99. Gottesdiener KM. Transplanted infections: Donor-to-host transmission with the allograft. Ann Intern Med 1989;110:1001-1016. 100. Barbara JAJ, Contreras M. Infectious complications of blood transfusion: Bacteria and parasites. Br Med J 1990;300:386-389. 101. Soulier JP. Diseases tramsmissible by blood transfusion. Vox Sang 1984;47:1-6. 102. Alter HJ, Purcell RH, Shih JW, et al. Detection of antibody to hepatitis C virus in prospectively followed transfusion recipients with acute and chronic non-A, non-3 hepatitis. N Engl J Med 1989;321:1494-1500. 103. Sayers MH, Anderson KC, Goodnough LT, et al. Reducing the risk for transfusiontransmitted cytomegalovirus infection. Ann Intern Med 1992;116:55-62. 104. Mayer KH, Opal SM. Unusual nosocomial pathogens. Infect Dis Clin North Am 1989;3:883899. 105. Wittner M, Rowin KS, Tanowitz HB, et al. Successful chemotherapy for transfusion babesiosis. Ann Intern Med 1982;96:601-604. 106. Tabor E, Garety RJ. Five cases of Psetldomonas sepsis transmitted by blood transfusions. Lancet 1984;1:1403. 107. Wright DC, Selss IF, Vinton KJ, et al. Fatal Yersinia enferocolitica sepsis after blood transfusion. Arch Path01 Lab Med 1985;109:1040-1042. 108. Yomtovian R, Lazarus LT, Goodnough NV, et al. A prospective microbiologic surveillance program to detect and prevent the transfusion of bacterially contaminated platelets. Transfusion 1993;33:902-909. 109. Schimpff SC. Therapy of infection in patients with granulocytopenia. Med Clin North Am 1977;61:1101-1118. 110. Bodey GP, Rodriguez V, Chang H, et al. Fever and infection in leukemic patients. A study of 494 consecutive patients. Cancer 1978;41:1610-1622. 111. Sanders JW, Powe NR, Moore RD. Ceftazidime monotherapy for empiric treatment of febrile neutroenic patients: A metaanalysis. J Infect Dis 1991;164:907-916. 112. Sanders CC. New B-lactams: New problems for the internist. Ann Intern Med 1991;115:65& 651. 113. Whimbey E, Kiehn TE, Brannon P, et al. Bacteremia and fungemia in patients with neoplastic disease. Am J Med 1987;82:723-730, 114. Singer C, Kaplan MH, Armstrong D. Bacteremia and fungemia complicating neoplastic disease. Am J Med 1977;62:731-742. 115. The EORTC International Antimicrobial Therapy Cooperative Group. Gram-positive bacteremia in granulocytopenic cancer patients. Eur J Cancer 1990;26:569-574. 116. Viscoli C, Van der Auwera P, Meunier F. Gram-positive infections in granulocytopenic patients: An important issue? J Antimicrob Chemother 1988;21(Suppl. C):149-156. 117. Winston DJ, Dudnick DV, Chapin M, et al. Coagulase-negative staphylococcal bacteremia in patients receiving immunosuppressive therapy. Arch Intern Med 1983;143: 32-36. 118. Karp JE, Dick J. Angelopulos C, et al. Empiric use of vancomycin during prolonged treatment-induced granulocytopenia: Randomized, double-blind, placebo-controlled clinical trial in patients with acute leukemia. Am J Med 1986;81:237-242. 119. Arning M, Gehrt A, Aul C, et al. Septicemia due to Streptococcrls n~itisin neutropenic patients with acute leukemia. Blut 1990;61:364-368. 120. Neu HC. The crisis in antibiotic resistance. Science 1992;257:1064-1073. 121. Anaissie E. Opportunistic mycoses in the immunocompromised host: Experience at a cancer center and review. Clin Infect Dis 1992;14(Suppl. 1):S43-53. 122. Meunier-Carpentier F, Kiehn TE, Armstrong D. Fungemia in the immunocompromised host: Changing patterns, antigenemia, high mortality. Am J Med 1981;71:363-370.

123. Lecciones JA, Lee JW, Navarro EE, et al. Vascular catheter-associated fungemia in patients with cancer: Analysis of 155 episodes. Clin Infect Dis 1992;14:875-883. 124. Wey SB, Mori M. Pfaller MA, et al. Hospital-acquired candidemia: The attributable mortality and excess length of stay. Arch Intern Med 1988;148:2642-2645. 125. Gerson SL, Talbot GH, Hurwitz S, et al. Prolonged granulocytopenia: The major risk factor for invasive pulmonary aspergillosis in patients with acute leukemia. Ann Intern Med 1984:100:345-351. 126. Fischer BD, Armstrong D, Yu B, et al. Invasive aspergillosis: Progress in early diagnosis and treatment. Am J Med 1981;71:571-577. 127. De Gregorio M, Lee W, Linker C, et al. Fungal infections in patients with acute leukemia. Am J Med 1982;73:543-548. 128. Hawkins C, Armstrong D. Fungal infections in the immunocompron~ised host. Clin HaematoI 1984;3:599-630. 129. Bodey G. Bueltmann B, Duguid W, et al. Fungal infections in cancer patients: An international autopsy survey. Eur J Microbiol Infect Dis 1992:11:99-109. 130. Core! L, Spear PG. Infections with Herpes simplex virus. N Engl J Med 1986;314:686-691, 749-757. 131. Rand KH, Kramer B, Johnson AC. Cancer chemotherapy associated symptomatic stomatitis: Role of Herpes simplex virus (HSV). Cancer 1982;50:1262-1265. 132. Buss DH, Scharyj M. Herpesvirus infection of the esophagus and other visceral organs in adults: Incidence and clinical significance. Am J Med 1979;66:457-462. 133. Ramsey PG, Fife KH, Hackman RC, et al. Herpes simplex virus pneumonia: Clinical, virologic. and pathologic features in 20 patients. Ann Intern Med 1982;97:813-820. 134. Straus SE, Ostrove JM, Inchauspe G, et al. VaricelIa-zoster virus infections: Biology, natural history, treatment, and prevention. Ann Intern Med 1988;108:221-237. 135. Meyers JD, Flournoy N, Thomas ED. Risk factors for cytomegalovirus infection after human marrow transplantation. J Infect Dis 1986;153:478-488. 136. Kaplan LJ, Daum RS, Smaron M. et al. Severe measles in immunocompromised patients. JAMA 1992;267:1237-1241. 137. Zahradnik JM, Spencer MJ, Porter DD. Adenovirus infection in the immunoco~npromised patient. Am J Med 1980;68:725-732. 138. Meunier F. Infections in patients with acute leukemia and lymphoma. In: Mandell GL, Bennett JE, Dolin R, eds. Principles and Practices of Infectious Diseases. New York: Churchill Livingstone, 1995, pp. 2675-2686. 139. Cotran RS. Kumar V, Robbins SL. Diseases of white cells, lymph nodes, and spleen. In: Robbins Pathologic Basis of Disease. Philadelphia: W.B. Saunders, 1989, pp. 703-754. 140. Goyal R, Bajpai S, Chopra HK, et al. Hairy cell leukemia - an unusual presentation. Leuk Res 1995;19:485-487. 141. Mackowiak PA, Demian SE, Sutker WL, et al. Infections in hairy cell leukemia: Clinical evidence of a pronounced defect in cell-mediated immunity. Am J Med 1980:68:718-723. 142. Bennett C, Vardinlan J, Golomb H. Disseminated atypical rnycobacterial infection in patients with hairy cell leukemia. Am J Med 1966;80:891-896. 143. Attal M. Schlaifer D, Rubie H. et al. Prevention of gram-positive infections after bone marrow transplantation by systemic vancomycin: A prospective. randomized trial. J Clin Oncol 1991;9:865-870. 144. Goodinan JL, Winston DJ, Greenfield RA, et al. A controlled trial of fluconazole to prevent fungal infections in patients undergoing bone marrow transplantation. N Engl J Med 1992;326:845-851. 145. Goodrich JM, Reed EC, Mori M, et al. Clinical features and analysis of risk factors for invasive candidal infection after marrow transplantation. J Infect Dis 1991;164:731-740. 146. Derouin F, Devergie A, Auber P, et al. Toxoplastnosis in bone marrow-transplant recipients: Report of seven cases and review. Clin Infect Dis 1992;15:267-270. 147. Miller W, Flynn P, McCulough J, et al. Cytomegalovirus infection after bone marrow transplantation: An association with acute graft-v-host disease. Blood 1986;67:1162-1167.

148. Wingard JR, Mellits D, Sostrin MB, et al. Interstitial penumonitis after allogeneic bone marrow transplantation: Nine year experience at a single institution. Medicine 1988;67:175-186. 149. Pannuti CS, Gingrich RD, Pfaller MA, et al. Nosocomial pneumonia in adult patients undergoing bone marrow transplantation: A 9 year study. J Clin Oncol 1991;9:77-84. 150. Meyers JD, Thomas ED. Infection complicating bone marrow transplantation. In: Rubin RH, Young LS, eds. Clinical Approach to Infection in the Compromised Host. New York: Plenum Medical, 1988, pp. 525-556. 151. Hiemenz JW, Greene JN. Special considerations for the patient undergoing allogeneic or autologous bone marrow transplantation. Hematol Oncol Clin North Am 1993;7:961-1002. 152. Rubin RH, Ferraro MJ. Understanding and diagnosing infectious complications in the immunocompromised host. Hematol Oncol Clin North Am 1993;7:795-812. 153. Cohen ML. Epidemiology of drug resistance: Implications for a post-antimicrobial era. Science 1992;257:1050-1055. 154. Kunin CM. Resistance to antimicrobial drugs - A worldwide calamity. Ann Intern Med 1993;118:557-561. 155. Leclercq R, Derlot E, Duval J, et al. Plasmid-mediated resistance to vancomycin and teicoplanin in Enterococc~lsfcrecilirn. N Engl J Med 1988;319:157-161. 156. Montecalvo MA, Horowitz H, Gedris C, et al. Outbreak of vancomycin-, ampicillin-, and aminoglycoside-resistant E~~ferococczls faeciurn bacteremia in an adult oncology unit. Antimicrob Agents Chemother 1994;38:1363-1367. 157. Noskin GA. Cooper I, Peterson L. Vancomycin-resistant Enterococcrrs faecium sepsis following persistent colonization. Arch Intern Med 1995;155:1445-1447. 158. Infectious Diseases Society of America. IDSA disseminates the latest on infectious diseases. Infect Dis Alert 1996;16:38-39. 159. Bouchud PY, Calandra T. Francioli P. Bactere~nia due to viridans streptococci in neutropenic patients: A review. Am J Med 1994;97:256-264. 160. Kern W, Kurrle E, Sclimeiser T. Streptococcal bacteremia in adult patients with leukemia undergoing aggressive chemotherapy. A review of 55 cases. Infection 1990;18:138-145. 161. Elting LS, Bodey GP, Keefe BH. Septicimia and shock syndrome due to viridans streptococci: A case-control study of predispoisng factors. Clin Infect Dis 1992;14:1201-1207. 162. Borek AP, Dressel DC, Hussong J, et al. Changing susceptibility of Streptococclrs pneumoniae to antimicrobial agents in Chicago, Illinois between 1993 and 1996:The need for accurate laboratory identification and susceptibility testing. Diagn Microbiol Infect Dis 1997;29:209-214. 163. Handwerger S, Horowitz H, Coburn K, et al. Infection due to Le~lconostocspecies: Six cases and review. Rev Infect Dis 1990;12:602-610. 164. Ascher DP, Zbick C, White C, et al. Infections due to Stor~zatococcusn~~icilaginosi~s: 10 cases and review. Rev Infect Dis 1991;13:1048-1052. 165. McWhinney PH, Kibbler CC, Gillespie SH, et al. Stonzatococous nzucilnginosus: An emerging pathogen in neutropenic patients. Clin Infect Dis 1992;14:641-646. 166. Young VM, Meyers WF. Moody MR, et al. The emergence of coryneform bacteria as a cause of nosocomial infections in compromised hosts. Am J Med 1981;70:646-650. 167. Riebel W, Frantz N. Adelstein D, et al. Coryizebacte~.irmiJK: A cause of nosocomial devicerelated infection. Rev Infect Dis 1986;8:42-49. 168. Johnson A, Hulse P, Oppenheim BA. Corynebacterir~mjeikeilim meningitis and transverse meylitis in a neutropenic patient. Eur J Clin Microbiol Infect Dis 1992;11:473474. 169. Henrickson K, Shenep J, Flynn P, et al. Primary cutaneous Bncill~lscereus infection in neutropenic children. Lancet 1989;1:601-603. 170. Tuazon CU. Other Bacillus species. In: Mandell GL, Bennett JE, Dolin R, eds. Principles and Practice of Infectious Diseases. New York: Churchill Livingstone, 1995, pp. 1890-1894. 171. Harvey RL, Sunstrum JC. Rhodococc~iseqrii infection in patients with and without human immunodeficiency virus infection. Rev Infect Dis 1991;13:139-145. 172. Khardori N, Elting L, Wong E, et al. Nosocomial infections due to X~mthomonrisn.zaltophilia ( P s e ~ ~ d o m o nmaltophilicl) as in patients with cancer. Rev Infect Dis 1990;12:997-1003.

173. Garau J. Beta-lactamases: Current situation and clinical importance. Intensive Care Med 1994;2O:S5-S9. 174. Sirot D. Extended-spectrum plasmid-mediated B-lactamases. J Antimicrob Chemother 1995;36(Suppl. A):19-34. 175. Meyer KS. Urban C, Eagan JA, et aI. Nosocomial outbreak of Klebsiella infection resistant to late-genertation cephalosporins. Ann Intern Med 1993;119:353-358, 176. Pegues DA, Carson LA, Anderson RL, et al. Outbreak of Pseudonzonas cepacin bacteremia in oncology patients. Clin Infect Dis 1993;16:407-411. 177. Kaye KM, Macone A, Kazanjian PH. Catheter infection caused by Methylobcrcterilcnz in immunocompromised hosts: Report of three cases and review of the literature. Clin Infect Dis 1992;14:1010-1014. 178. Edmond MB. Riddler SA, Baxter CM, et al. Agrobacterium radiobacter: A recently recognized opportunistic pathogen. Clin Infect Dis 1993;16:388-391. 179. Cieslak TJ, Robb ML, Drabick CJ, et al. Catheter-associated sepsis caused by Ochrobactr~im anthropi: Report of a case and review of related nonfermentative bacteria. Clin Infect Dis 1992;14:902-907. 180. Legrand C. Anaissie E. Bacteremia due to Achromobacter xylosoxidans in patients with cancer. Clin Infect Dis 1992;14:479-484. 181. Kim JH, Cooper RA, Welty-Wolf KE, et al. Pseudonzonas prrtrefciciens bacteremia. Rev Infect Dis 1989:11:97-104. 182. Bilgrami S, Bergstrom SK, Peterson DE, et al. Capnocytoyhcigcr bacteremia in a patient with Hodgkin's disease following bone marrow transplantation: Case report and review. Clin Infect Dis 1992;14:1045-1049. 183. Lorber B. Gas gangrene and other Clostridium-associated diseases. In: Mandell GL. Bennett JE. Dolin R. eds. Principles and Practice of Infectious Diseases. New York: Churchill Livingstone. 1995, pp. 2182-21 95. 184. Valtonen M, Sivonen A, Elonen E. A cluster of seven cases of Clostridium tertitim septicemia in neutropenic patients. Eur J Clin Microbiol Infect Dis 1990;9:40-42. 185. Zinner SH. New and unusual infections in neutropenic patients. In: Klastersky J, ed. Infectious Complications of Cancer. Boston: Kluwer Academic, 1995. pp. 173-184. 186. Huyghebaert MF, Dreyius F. Paul G, et al. Septicemie a Frlsobacteri~~rrznucleatz~nz, producteur de beta-lactamase chez un sujet neutropenique. Ann Med Interne (Paris) 1989:140:225-226. 187. Weinberger M, Wu T. Rubin M, et al. Leptotriclzn bllccalis bacteremia in patients with cancer: Report of four cases and review. Rev Infect Dis 1991;13:201-206. 188. Simon HB. Mycobacterial and nocardial infections in the compromised host. In: Rubin RH, Young LS. eds. Clinical Approach to Infection in the Compromised Host. New York: Plenum. 1988, pp. 221-251. 189. McWhinney PH. Yates M, Prentice HG, et al. Infection caused by Mycobacteri~rmchelonae: A diagnostic and therapeutic problem in the neutropenic patient. Clin Infect Dis 1992;14:1208-1212. 190. Raad I.I., Vartivarian S, Khan A, et al. Catheter-related infections caused by the Mycobcrcferirrin forfuitrliu complex: 15 cases and review. Rev Infect Dis 1991;13:1120-1125. 191. Kristjansson M, Bieluch VM, Byeff PD. Mycobacferzrim haernophilunz infection in immunocompromised patients: Case report and review of the literature. Rev Infect Dis 1991;13:906-910. 192. Chatis PA, Crumpacker CS. Resistance of Herpesviruses to antiviral drugs. Antimicrob Agents Chemother 1992;36:1589-1595. 193. Lalezari J, Stagg R, Kupperman B, et al. Intravenous cidofovir for peripheral Cytomegalovirus retinitis in patients with AIDS. Ann Intern Med 1997;126:257-263. 194. Studies of ocular complications of AIDS Research Group and the AIDS Clinical Trials Group. Parenteral cidofovir for Cytomegalovirus retinitis in patients with AIDS: The HPMPC peripheral Cytomegalovirus retinitis trial. Ann Intern Med 1997;126:264274.

195. Foreman KE, Friborg J, Kong W, et al. Propatation of a human Herpesvirus from AIDSassociated Kaposi's sarcoma. N Engl J Med 1997;336:163-171. 196. Strauchen JA, Hauser D, Burstein D, et al. Body cavity-based malignant lymphoma containing Kaposi sarcoma-associated Herpesvirus in an HIV-negative man with previous Kaposi sarcoma. Ann Intern Med 1996;125:822-825. 197. Pfaller M, Wenzel R. Impact of the changing epidemiology of fungal infections in the 1990's. Eur J Clin Microbiol Infect Dis 1992;11:287-291. 198. Walsh TJ, Lee J, Lecciones J, et al. Empiric therapy with amphotericin B in febrile granulocytopenic patients. Rev Infect Dis 1991;13:496-503. 199. Wingard J. Merz W, Sara1 R. Canditin fropicalis: A major pathogen in immunocompromised patients. Ann Intern Med 1979;91:539-543. 200. Anaissie E, Bodey G. Nosocomial fungal infections: Old problems and new challenges. Infect Dis Clin North Am 1989;3:867-882. 201. Gamis AS, Gudnason T, Giebink GS, et al. Disseminated infection with Fzdsnriurn in recipients of bone marow transplants. Rev Infect Dis 1991;13:1077-1088. 202. Nucci M, Spector N, Lucena S. et al. Three cases of infection with F~isnriunzspecies on neutropenic patients. Eur J Clin Microbiol Infect Dis 1992;11:1160-1162. 203. Bennett J. Miscellaneous fungi, and prototheca. In: Mandell G, Bennett J, Dolin R, eds. Principles and Practices of Infectious Diseases. New York: Churchill Livingstone, 1995, pp. 2389-2393. 204. Nahass GT, Rosenberg SP, Leonardi CL, et al. Disseminated infection with Trichosporon beigelii. Arch Dermatol 1993;129:1020-1023. 205. Walsh TJ, Newman KR, Moody M, et al. Trichosporonosis in patients with neoplastic disease. Medicine 1986;65:268-279. 206. Kalter DCA, Tschen JA, Cernoch PL, et al. Genital white piedra: Epidemiology, microbiology and therapy. J Am Acad Dermatol 1986;14:982-993. 207. Martino P, Venditti M, Micozzi A , et al. Blastoschizoirlyces cnpitntzls: An emerging cause of invasive fungal disease in leukemia patients. Rev Infect Dis 1990;12:570-582. 208. Danker W, Spector S, Fierer J. Malnssezia fungemia in neonates and adults: Complication of hyperalimentation. Rev Infect Dis 1987;9:743-837. 209. Sudduth E, AJ CI, WE F. Phaeohyphomycosis due to Exophinla species: Clinical spectrum of disease in humans. Clin Infect Dis 1992;15:639-644. 210. Wood GM, McCormack JG. Muir DB, et al. Clinical features of human infection with Scedospori~inziizflnttinz. Clin Infect Dis 1992;14:1027-1033. 211. Wagner KF. Agents of chro~nomycosis.In: Mandell GL, Bennett JE, Dolin R, eds. Principles and Practice of Infectious Diseases. New York: Churchill Livingstone, 1995, pp. 2324-2327. 212. Anaissie E, Bodey GP, Kantarjian H, et al. New spectrum of fungal infections in patients with cancer. Rev Infect Dis i989;11:369-378. 213. Kiehn TE, Gorey E, Brown AE, et al. Sepsis due to Rhodotorr,la related to use of indwelling central venous catheters. Clin Infect Dis 1992;14:841-846. 214. Henson JW, Jalaj JK, Walker RW, et al. Pnerrnzocystis cnrirlii pneumonia in patients with primary brain tumors. Arch Neurol 1991;48:406-409. 215. Beaman MH, McCabe RE, Wong S, et al. Toxoplasrna gondii. In: Mandell GL, Bennett JE. Dolin R. eds. Principles and Practice of Infectious Diseases. New York: Churchill Livingstone, 1995, pp. 2455-2475. 216. Ledingham JCG, Penfold WJ, Woodcock HM. Recent bacteriological experiences with typhoidal disease and dysentery: With notes on the protozoan parasites in the excreta. Br Med J 1915;2:704-711. 217. Faust EC, Giraldo LE, Caicedo G. Human isosporosis in the western hemisphere. Am J Trop Med Hyg 1961;10:343-349. 218. Gentile G, Venditti M, Micozzi A, et al. Cryptosporidiosis in patients with hematologic malignancies. Rev Infect Dis 1991;13:842-846. 219. Plumelle Y, Edouard A. Strongyloides stercornlis in T-cell leukemia/lymphoma in adults and acquired immunodeficiency syndrome. Rev Med Interne 1996;17:125-129.

220. Nucci M, Portugal R, Pulcheri W, et al. Strongyloidiasis in patients with hematologic malignancies. Clin Infect Dis 1995;21:675-677. 221. Savage D, Foadi M, Haworth C, et al. Marked eosinophilia in an immunosuppressed patient with strongyloidiasis. J Intern Med 1994;236:473-475. 222. Aydin H, Doppl W, Battmann A, et al. Opportunistic Strongyloides stercoralis hyperinfection in lymphoma patients undergoing chemotherapy and/or radiotherapy - report of a case and review of the literature. Acta Oncol 1994;33:78-80. 223. Simpson WG, Gerhardstein DC, Thompson JR. Disseminated Strongyloides stercoralis infection. South Med J 1993;86:821-825. 224. Wilkinson R, Leen CL. Chronic lymphocytic leukaemia and overt presentation of underlying Stronglyloides stercoralis infection. J Infection 1993;27:99-100. 225. Slater LN, Welch DF. Rochalimaea species (recently renamed Bartonella). In: Mandell GL, Bennett JE, Dolin R, eds. Principles and Practice of Infectious Diseases. New York: Churchill Livingstone, 1995, pp. 1741-1747. 226. Heney C, Greeff M, Davis V. Hickman catheter-related Prorothecal algaemia in an immunocompromised child. J Infect Dis 1991;163:930-931. 227. Kwon-Chung KJ, Bennett JE. Medical Mycology. Philadelphia: Lea & Febiger, 1992. 228. Kaminski ZC, Kapila R, Sharer LR, et al. Meningitis due to Prototheca wickerhamii in a patient with AIDS. Clin Infect Dis 1992:15:704-706. 229. Iacoviello VR, DeGirolami PC, Lucarini J, et al. Protothecosis complicating prolonged endotracheal intubation: Case report and literature review. Clin Infect Dis 1992;15:959-967.

3. Approach to fever in the neutropenic host Athena Stoupis and Stephen H. Zinner

1. Introduction The approach to fever in the neutropenic host in large part depends on an understanding of the basic immune defect. White blood cells are critically important in maintaining an infection-free state. Nearly 30 years ago, the risk of serious bacterial infections in patients with a low number of circulating leukocytes was first recognized by Bodey and coworkers [I]. Diminished polymorphonuclear leukocyte numbers or function and impaired humoral immunity result in the development of serious infections, with the attendant risk of morbidity and mortality. The risk of infection progressively increases as the neutrophil count decreases below 1000 cells/mL. The most severe risk for bacteremia occurs when the granulocyte count falls below 100 cellslmL [I]. The depth of neutropenia, as well as the duration, are also considered to be important parameters that influence the development of infection [2]. One study revealed that in patients with neutropenia lasting less than 1 week ("short neutropenia"), fewer than 30% developed fever or other evidence of infection, whereas almost all patients with neutropenia lasting longer than 1week ("long neutropenia") developed fever or evidence of infection [2-51. In patients with malignancy, neutropenia may result from the use of cytotoxic chemotherapy, marrow irradiation, or neoplastic infiltration of the bone marrow. Another coexisting factor associated with malignancy that affects host defense is loss of the integrity of the skin and mucosa, which can occur by direct extension of solid tumors or by erosive mucosal lesions often encountered in patients with leukemia. The placement of intravenous or urinary catheters, and of endotracheal tubes, and the presence of decubitus ulcers often encountered in the hospital setting also violate physical defense barriers and may result in infections caused predominantly by nosocomial pathogens, particularly staphylococci, streptococci, or aerobic gram-negative rods that colonize the skin and gut of these patients. Impairment of humoral immunity with resultant deficits in antibody responses that normally potentiate phagocytosis can be found in patients who have undergone splenectomy for staging purposes or have developed splenic Gory A . Noskin (ed), M A N A G E M E N T O F INFECTIOUS C O M P L I C A T I O N S IN C A N C E R PATIENTS. O 1998. Klrlwer Academic Pnblishers, Bostorz. Ail rights reserved.

infiltration with malignant cells, as in patients with Hodgkin's disease, multiple myeloma, or leukemia [6,7]. These patients are at high risk for the development of overwhelming infection due to encapsulated bacteria such as Streptococcus pneumoniae, Haernophilus influenzae, Neisseria meningitidis, and others. The problem is further complicated by an inability to mount an appropriate antibody response to prophylactic polysaccharide vaccines following splenectomy [7,8]. Cell-mediated immunity also may be impaired in many patients who receive cytotoxic chemotherapy, radiation, and corticosteroids. This results in infection with intracellular organisms such as Listeria monocytogenes, Legionella, and Salmonella spp., and less commonly Mycobacterium species [9,10]. These patients also may be infected with herpes viruses (herpes simplex, varicella zoster, cytomegalovirus) and are subject to fungal infections, including disseminated candidiasis, cryptococcal meningitis, and aspergillosis, among others [7,10]. Other risk factors for infection, such as the patient's clinical instability, advanced malignant disease, and overt organ dysfunction, also contribute to the risk of serious infection in patients with cancer [I I]. 2. Risk factor assessment

In recent years, several retrospective studies have evaluated neutropenic cancer patients with fever in an attempt to identify subsets of patients at high and low risk for the development of bacteremia and other serious infectious complications [12]. Talcott and coworkers [13,14] studied various clinical variables that could be assessed within the first 24 hours of presentation of fever and neutropenia, with the aim of developing a predictive model that could stratify patients according to risk of complications. The majority of patients were projected at high risk on the basis of "comorbidity factors" such as cardiac or renal disease, uncontrolled cancer, or existing clinical status at onset of admission that included hypotension, respiratory failure, altered mental status, dehydration, and inadequate oral intake. Of the 444 patients studied, those with any of these high risks were more likely to have serious medical complications (35%) than patients without any risk factors (5%). These studies have important implications for the selection of patients amenable to oral or outpatient therapy. Pappo and Buchanan [I51 also retrospectively examined multiple clinical and laboratory criteria to identify low-risk features at the onset of fever in pediatric patients. Over a 3-year period, 600 hospital admissions were assessed and 93 episodes of bacteremia were noted and analyzed. The analysis revealed that all but 7 of the 93 positive blood cultures (7.5%) occurred in patients defined as "high risk" by the presence of one or more of these factors: (1) primary disease not in remission; (2) age <1 year; (3) fever developing 10 days or less from the first day of the most recent course of chemotherapy; (4) no

'

evidence of bone marrow recovery, defined as an absolute neutrophil count <100/mm3 and/or a platelet count <75,000/mm3. Patients without these risks were classified at "low risk" for the development of bacteremia [Ill. Six of the seven episodes of bacteremia in the low-risk patient category were caused by gram-positive organisms, including coagulase-negative staphylococci, Bacillz~sspecies, and Propionibacteriurn species, showing the relatively indolent course of these gram-positive organisms, even in neutropenic patients [16,17]. The introduction of more intensive chemotherapy, autologous bone marrow transplantation, or peripheral stem cell infusions and the use of hematopoietic growth factors have had a major impact on the risk of infectious complications of the febrile neutropenic patient [2]. The development of a predictive model for the assessment of specific risk factors to classify patients at high or low risk of infection may prove to be important clinically as well as economically. The spiraling costs of health care in this era of cost containment and managed care make the need for risk stratification of the febrile neutropenic patient even more critical. Future studies may provide additional costeffective modifications of the management of the febrile neutropenic patient, such as early discharge from the hospital or oral and home IV antibiotics as alternatives to prolonged hospitalization [18].

3. Clinical presentation Fever is the most common presentation of infection in the neutropenic host, and often it is the sole sign of infection [19]. Although fever may be related to noninfectious causes, such as the use of medications, transfusion of blood products, or the underlying disease itself, elevated temperature remains strongly associated with infection in the immunocompromised host [19]. Febrile neutropenia (fever defined as ~38.5"C)is associated with microbiologically documented infections in about 40% of patients, with bacteremia accounting for about half of these [20]. Approximately 30% of infections may present as fever of unknown origin (FUO). Even in these cases, the usual common sites of infection should influence therapeutic decisions. The most frequently encountered sites of infection as derived from five consecutive trials of the European Organization for Research and Treatment of Cancer (EORTC) - International Anitmicrobial Therapy Cooperative Group (IATCG) include the mouth and oropharynx (25%); the lower respiratory tract (25%); skin, soft tissue, and intravascular catheters (15%); gastrointestinal tract (15%); gerineal region (10%); the urinary tract (5-10%); and the nose and sinuses (5%) [21,22]. Many of the classic hallmarks of infection seen in healthy individuals, such as fluctuance, calor, rubor, and lymphadenopathy, may be absent in the neutropenic patient and the immunocompromised host [23]. Exudates may be

less common in pharyngeal infections, urinary tract infections may not be accompanied by pyuria, and meningitides may present without classical meningeal irritations, all due to decreased migration of leukocytes to the sites of infection [lo]. A careful history and physical examination are critical to establish a prompt diagnosis, and even minor signs of infection must be pursued vigorously. For example, the development of pneumonia may not be associated with cough, expectoration, purulent sputum, or rales, and an abnormal chest radiograph may be the only clinical aid [23]. Pulmonary infiltrates may be caused by a wide variety of organisms, which include bacterial pathogens such as Enterobacteriaceae and Pse~ldomonasneruginosa, Legionelln, Nocardin spp. (which may form cavities), Mycobncteria, and fungi. Aspergillus is a common pathogen in patients with prolonged, profound neutropenia [2]. Prompt use of diagnostic procedures, such as bronchoscopy and/or needle thoracentesis of pleural effusions with pleural or lung biopsy, should be performed for early diagnosis, although these procedures may be hazardous in the presence of severe thrombocytopenia [24]. Open lung biopsy may be necessary for definitive diagnosis, especially in patients whose underlying disease process has a good prognosis, such as organ transplant recipients or patients with Hodgkin's disease [25,26]. Oral mucositis or stomatotoxicity is frequently associated with chemotherapy and can be quite severe. Drug-induced mucositis can become superinfected with Candidn spp., cytomegalovirus (CMV), or herpes simplex virus (HSV), which can be clinically indistinguishable. Periodontal or gingival involvement may indicate infection by anaerobes that requires antianaerobic antibiotics [2,27]. Other mucosal sites, including the perianal area, must be examined carefully for the presence of painful ulcerations or symptoms of proctalgia. If the patient complains of rectal pain, rectal examination is critical to rule out inflammation and early abscess formation. Other infections of the gastrointestinal tract include intraabdominal infections such as Clostridiunz difficile colitis, which can progress to toxic megacolon and can be fatal [28,29]. Relapse after treatment is not uncommon. In patients with diarrhea, stool should be tested for C. difJicile toxin if the diagnosis is considered. Other pathogens, such as Snlrnonelln spp., Shigelln spp., and Strongyloides stercornlis, are important but less common [lo]. Typhlitis, inflammatory colitis involving the cecum, is another intraabdominal process associated with mortality as high as 30-50% in profoundly neutropenic patients. It presents with fever and right lower quadrant pain, and is usually a manifestation of polymicrobial infection. Although medical therapy is preferred, surgical resection may be necessary if there is no response to bowel rest and broad-spectrum antibiotics [30]. Thorough examination of the retina, skin, sinuses, and lymph nodes also must be performed. Skin pustules or subcutaneous nodules may be indicative of disseminated fungal infection [31], and extensive cutaneous necrosis can be seen with ecthyma gangrenosum, which is usually associated with Pseudomo-

nas neruginosa bacteremia [32]. Iatrogenic foci of infection also should be sought at sites of indwelling venous and urinary catheters or other sites of foreign bodies [33,34]. Despite advances in laboratory diagnostic methods (lysis-centrifugation blood cultures for bacteria and yeasts, shell-vial cultures for viruses, polymerase chain reaction, etc.) and the advent of new imaging studies (indium scan, CT, and MRI), the cause of neutropenic fever remains unknown in 6070% of patients. However, sufficient samples of blood, sputum, urine, and other body fluids should be obtained for culture and can be important diagnostic tools in the approach to an immunocompromised patient, although the yield of surveillance cultures is usually low [35].

4. Infecting pathogens in the neutropenic patient Bacterial infections constitute the single most common group of infections in the neutropenic host and may produce serious complications. The pathogens associated with the highest mortality rates include the Enterobacteraciae (Escherichia coli, Klebsiella spp., Prote~isspp.) and Pseudornonas aeruginosa, However, in the past 5-20 years most centers have shown a decline in the incidence of gram-negative pathogens and an increase in the incidence of gram-positive infections, particularly those caused by coagulase-negative staphylococci and alpha-hemolytic streptococci (Figure 1). This increase in frequency was recently confirmed by the EORTC International Antimicrobial Therapy Cooperative Group Trial [36], which showed in a group of 694 patients that over 65% of single-agent bacteremias were caused by grampositive bacteria. Several factors have been implicated to explain the increase in grampositive bacteremias, including: (1) the increased use of fluoroquinolone antibiotics for prophylaxis in neutropenic patients, which alters the gramnegative gastrointestinal flora, with little effect on gram-positive organisms; (2) the increased use of long-term indwelling intravenous devices; (3) aggressive cytotoxic therapy, particularly cytarabine, with resultant severe mucositis; (4) the use of histamine type 2 (H2) blockers. An increased incidence of viridans streptococcal bacteremia and the associated shock syndrome, characterized by hypotension, rash, palmar desquamation, and acute respiratory distress syndrome, from 1 case per 10,000 admissions to 47 cases per 10,000 admissions (P = 0.0001), was noted between 1972 and 1989 at the University of Texas M. D. Anderson Cancer Center in Houston [37]. Furthermore, the risk of streptococcal bacteremia was noted to increase in patients with profound neutropenia who had received a fluoroquinolone or trimethoprim/sulfamethoxazole for antibacterial prophylaxis. A sevenfold increase in risk (P < 0.002) in viridans streptococcal infection also was noted with the use of H2 antagonists. In summary, it appeared that gastric overgrowth of organisms resistant to fluoroquinolones and

fl Gram-Neg Gram-Pos

Figure 1. Trends in single organism bacterelnias in 1973-1995 according to the EORTC-IATCG trials.

trimethoprimisulfamethoxazole, the HZ antagonist-induced alkaline environment, and chemotherapy-induced gastric ulceration, which provides a port of entry for the infecting organism, were predisposing factors. Other bacterial pathogens have been reported with increasing frequency. Leuconostoc species, which had been regarded as commensals or nonpathogens, have recently been identified as important causes of intravenous catheter-associated bacteremias [38]. Corynebacterium jeikeium, Rhodococcus equi (formerly Corynebacterium equi), Stomatococcus mucilaginosus (formerly known as Neisseria mucosa or Staphylococc~~s salivarius), and Bacillus cereus also cause bacteremia or serious infections in neutropenic hosts. Bacillus cereus has been specifically associated with primary cutaneous infection. Vesicular or pustular lesions usually occur on extremities during warm weather. This organism also has been associated with deep soft-tissue infections and occasionally can result in bacteremia [39]. Other gram-positive bacteria that can cause invasive bloodstream infections include Enterococcus species (especially those resistant to vancomycin and aminoglycosides), Lactobacillus rhamnosus, Corynebacterium striatum, Clostridium septicum, and Clostridium tertium [40]. Although bacteremia due to gram-negative bacilli is decreasing in frequency, these organisms are still responsible for serious infections, with significant morbidity and mortality in the immunocompromised host. Of

interest is the emergence of newly recognized gram-negative organisms, such as Stenotrophomonas maltophilia (formerly Xanthomonas maltophilia, Pseudomonas maltophilia), which exhibits high levels of antibiotic resistance. Alteromonas putrefaciens, Vibrio parahaemolyticus, Capnocytophaga spp., Burkholderia cepacia, Achromobacter xylosoxidans, Ochrobactrum anthropi, Agrobacteruim radiobacter, and others also are emerging, probably as a result of selection pressures of excess antimicrobial use [40] (Table 1). Although bacterial infections are usually encountered first in febrile neutropenic patients, fungal, viral, parasitic, and protozoal infections, as well as mixed infections, also are seen usually later in the course of fever in neutropenic and other immunocompromised patients. Invasive fungal infections are important causes of morbidity and mortality, and have been documented with increased frequency in autopsies of patients with hematologic malignancies. Predisposing factors for fungal infection include the use of broad-spectrum antibiotic regimens, corticosteroids, parenteral nutrition, and indwelling intravenous catheters. The most commonly isolated fungal pathogens are Candida spp. and Aspergillus spp., although other fungi, such as Torulopsis glabrata, C. krusei, Cryptococcus neoformans and Mucor spp., are also seen. Disseminated Fusarium infections also have been reported in patients with hematologic malignancy and after bone marrow transplantation [41]. Fusarium spp. are emerging pathogens in immunocompromised patients. Over the last 10 years approximately 80 new cases of disseminated Fusariurn spp. infections have been reported. Portals of entry for deep-seated infection include the respiratory tract, gastrointestinal tract, skin and skin structures, and central venous catheters. The prognosis for systemic fungal infections is usually poor, and mortality rates are high despite antifungal therapy. The antifungal agent that has been shown to be most effective in vitro appears to be amphotericin B; however, optimal efficacy requires a concomitant increase in the absolute neutrophil count [42,43]. Prophylactic and therapeutic administration of intravenous amphotericin B has failed to prevent disseminated Fusarium infections in these patients [44]. Other deep-seated fungal infections include those caused by Pseudallescheria boydii and Scedosporium prolijicans. Trichosporon beigelii and other species also cause infection in neutropenic patients. The pathogenetic process caused by this organism remains unclear, and amphotericin B remains the mainstay of treatment for Trichosporon spp. despite its limited clinical efficacy [45,46]. Aggressive diagnostic testing for these pathogens is imperative in guiding further management. Viral infections also are found in neutropenic patients and include infections caused by HSV, varicella-zoster, and CMV. Reactivation of HSV frequently results in mucositis with the formation of painful ulcerations. HSV and CMV also are implicated in the development of esophagitis. CMV disease after bone marrow transplant most commonly presents as pneumonitis or gastroenteritis, and may be associated with a mortality rate of 50% or higher

B c

n

Ji.l,iiir

=3t

CQ

-Q 0

S •iJ -Ci 0
o "9

:s.s

t

-3 = t; o

s-'^i. o:t

d

!0

^ s'-a 2 § i o . "a o .a Si r=s ^ -« .• Q ttj

O .^ r- S ^ ^ ^ c ^ ^ s; o a ^ ci,

^

!s -S t^ N

K

^

§1 tij bj t^

u. .3 S -O

o 1^

<

ESS 3-3 K

SSS-:|

E VD

(D C c« 03 C W) O

Co:

N -^

s Sy cq fct,

U d.2 S

lit

1^2 ^a

la

t« !3

•£

^ o

13 i : •^j •J:'

to

' t ::i-

^ c S 3 <3 - ^

C •?

II

c S u -C) „ ^. •S '-^ o 3 <J CQ re;

if

O '^C 0

11

OH S

^ ti)

-. « 3 ^ a; ^

o 84

3 ^

c, g

3 3 3 fill ei

' I I R I I , IHlll a s «

i-J O

r \ U

r^

a^Co

CQ IjQ s j

sa

g 5 s: •g


b s

0 a 0 ><

s s

1441. Disseminated herpesvirus infections and extensive warts caused by the human papillomavirus also have been encountered in these patients. Nonbacterial pathogens such as Pneurnocystis carinii are less commonly associated with pneumonitis in neutropenic patients than in patients with AIDS, but bronchoscopy with lavage andlor needle biopsy are required for diagnosis in these patients. Prior to the HIV epidemic, Pneumocystis carinii was classically encountered in patients with lymphoma. Expectorated, induced, or lavaged sputum specimens frequently suffice to diagnose this infection in patients with AIDS, but not in patients with hematological malagnancies.

5. Treatment strategies The high morbidity and mortality associated with gram-negative bacteremia and the paucity of physical signs and symptoms of infection other than fever prompted the use of empirical therapy with broad-spectrum antibiotics as early as 1971, as reported by Schimpff and colleagues 1471. This practice still remains the standard of care. Although an increase in the number of gram-positive infections has occurred over the past 5 years [37,48,49], gram-negative bacteremia continues to be associated with high morbidity and mortality [48,50,511, even when appropriate therapy is initiated [52]. Therefore, any empiric antibiotic regimen should provide activity against pseudomonas and other gram-negative bacteria. The predominant organisms and their antibiotic susceptibilities in a given hospital environment also should guide the choice of empiric antibiotic therapy in the febrile neutropenic patient. Controversy remains as to the empiric treatment regimen of choice in the febrile neutropenic patient, and there is no single agent or combination drug regimen that is consistently superior to others 1531. 5.1 Beta-lactam compound plus aminoglycoside The traditional choice of antibiotics is the combination of a beta-lactam compound, such as an antipseudomonal penicillin or cephalosporin, plus an aminoglycoside. The rationale for using such antibiotic combinations rests on several considerations including: (1) broad-spectrum antibacterial coverage; (2) high bactericidal activity with synergistic antibacterial effects, especially against P. aertlginosn, which are important factors in patients with profound and prolonged neutropenia 140,541; (3) decreased risk of the emergence of resistance and reduction in superinfections. Various combination drug regimens have been used over the years for the empiric treatment of febrile neutropenic patients (Table 2). From the early trials of the EORTC (IATCG) it can be concluded that triple antibiotic regimens were not more efficacious than double combination antibiotic

.3 S H

(11

u

2 =s 3 >

u.s ^

= :rr P ><

o §.-2 i3 t^ S 1-1 G

B-^ U

U

U

>,

o

N

lO

'£:^ ,^o. o^ ^^:;'^ t---

_g 'o

o a o

C3

O

o

3

^6

s

'a

o, fc;

,^ 4-<

U U

t ) '^^ T3

"

o S s y c S 5 rt

c o^ p <^ M o S

S .S : 5 io^

o

;:^ •^XCO) 3 3 o ca o N « u
3 '"3 "yCJ it

OH

ffl

86

Ui

a

< !S ££ £

CL,

H U O

regimens [55]. One study used up to five antibiotics as empirical therapy, but obtained only a 53% response rate [56]. Early EORTC (IATCG) trials also revealed that an antipseudomonal penicillin plus amikacin was more potent than the combination of gentamicin with cephalothin, adding further evidence that the use of an antipseudomonal beta-lactam was useful [57]. Patients with profound and prolonged granulocytopenia who have gramnegative rod bacteremia appear to do better with two antimicrobial agents that are synergistically active than with a single drug [lo]. The importance of the synergistic effect was reviewed early by Klastersky and Zinner, who reported an 82% clinical success rate versus 53% depending on the presence or absence of a synergistic outcome in vitro [21,54]. Aminoglycosides provide rapid bactericidal activity against most gramnegative rods. Their synergistic activity in combination with a beta-lactam was shown to be beneficial in a trial of ceftazidime plus amikacin; neutropenic patients with gram-negative rod bacteremias had better clinical outcomes with a full course of both antibiotics than when the aminoglycoside was discontinued after 3 days [58]. This was especially important in profoundly, persistently granulocytopenic patients with gram-negative rod bacteremia. In addition, bacteremic patients have improved outcomes if the gram-negative rod is susceptible to both the beta-lactam as well as the aminoglycoside [58]. Taking into account local susceptibility patterns, ceftazidime plus an aminoglycoside was a reasonable initial drug regimen. If there is no evidence of gram-negative rod bacteremia, after 3 or 4 treatment days the aminoglycoside can be discontinued. However, the recent rise in streptococcal bacteremia in neutropenic patients suggests that the use of ceftazidime might no longer be optimal initial therapy. Recent studies have shown that the use of single daily dosing of aminoglycosides is as effective as multiple dosing [36,59]. The EORTCIATCG studied 858 febrile episodes in 677 febrile neutropenic patients who were randomized to receive amikacin (20mglkg IV in a single daily dose) plus ceftriaxone or amikacin (20mglkg IV in three divided doses) plus ceftazidime. Response rates were similar in both groups, as was associated nephrotoxicity (2-3%), which occurred later in patients receiving single daily doses [36]. The once-daily administration of aminoglycosides is also considerably simpler than multiple dosing schemes and may be a more realistic option in outpatient antibiotic therapy. Martino et al. [60] administered once-daily ceftriaxone plus amikacin to 46 patients with fever and neutropenia initially in a short-term hospital ward and then either at home or at daily clinic visits, once they were afebrile and had no clinical signs of infection. This antibiotic regimen without modification was successful in 37 of 49 febrile episodes (76% response rate). Four patients who ultimately responded to therapy developed septic shock requiring hospitalization. This nonrandomized, uncontrolled pilot study reveals that selected patients who have rising granulocyte counts and/or are expected to have a short duration of neutropenia and have improving signs of infection may be candidates for early discharge and completion of antibiotic

therapy as outpatients. Single daily dosing of aminoglycosides, therefore, may play an important role in the future outpatient management of febrile neutropenic patients.

5.2 Double beta-lnctam combination therapy Another approach to the management of the granulocytopenic patient is the use of double beta-lactam combinations, such as a penicillin plus a cephalosporin, which avoid the nephrotoxicity of the aminoglycosides and provide broad-spectrum activity. However, there is little potential for synergism, and, in fact, antagonism 1611might result and resistant organisms might be selected. Some beta-lactam antibiotics, for example, cefotaxime, can induce bacterial beta-lactamase production and further disrupt the therapeutic potential of the double beta-lactam combination [62,63]; therefore, their role in empiric treatment remains equivocal.

5.3 Morzotherapy The new array of recently introduced expanded-spectrum antibiotics offers the potential for empiric monotherapy, perhaps minimizing the nephrotoxicity and ototoxicity associated with the use of aminoglycosides [64]. The newer antibiotics include third- and fourth-generation cephalosporins, carbapenems, extended-spectrum penicillins with beta-lactamase inhibitors, monobactams, and quinolones. Among the third-generation cephalosporins, ceftazidime and cefoperazone provide adequate activity against Pseudomonns aeruginosa as well as other gram-negative rods. In most studies, ceftazidime has been compared with a beta-lactam compound plus an aminoglycoside [65-671. As shown in a metaanalysis, ceftazidime monotherapy was as effective as the standard combination regimens [68]. De Pauw and colleagues compared ceftazidime alone to piperacillin plus tobramycin and reported similar success rates with both regimens. They studied 876 febrile neutropenic episodes in 696 patients with acute leukemia or bone marrow transplantation. A 63% satisfactory response rate was found compared with 61% with the combination of piperacillin and tobramycin. Adverse events were significantly lower in the mono therapy group (8% of episodes treated with ceftazidime compared with 20% of episodes treated with combination therapy) [65]. A recent outbreak of ceftazidime-resistant Klebsielln pne~rrrzoniaeand E. coli raised concern and signals the need for the use of other antibiotic regimens [69]. Also, in hospitals where there is excess infection with Enterobacter cloacae, Serratia spp., or Psezidornonas aeniginosa, a combination of ceftazidime and aminoglycoside might be necessary to minimize the induction of chromosomal betalactamases. In a pilot study by Eggimann and colleagues [70], the efficacy and safety of cefepime, a new cephalosporin with extended-spectrum activity against both

gram-positive and gram-negative bacteria as empirical monotherapy, was evaluated in 108 febrile episodes in 84 granulocytopenic cancer patients. Response rates were 86% for gram-negative infections and 44% for grampositive infections. The survival rate after resolution of granulocytopenia was 96% three patients died of primary bacterial infection and one from secondary disseminated candidiasis. Cefepime monotherapy appeared relatively safe and effective as empiric therapy in febrile neutropenia; however, further studies are needed to document whether it is superior to other advanced generation cephalosporins. In an early study, Pizzo et al. [67] compared ceftazidime monotherapy with a double beta-lactam (cephalothin, carbenicilIin) plus aminoglycoside (gentamicin) combination. There was no significant difference in terms of "success of therapy." and these authors introduced the concept of modification of therapy in the analysis of success of therapy. Patients were evaluated at 72 hours and then further treatment was based on modification of the initial empiric therapeutic regimen. In patients with documented infection who failed initial therapy, the antibiotic regimen was tailored with an ner~iginosnwas isolated or with vancomycin aminoglycoside if Psezi~lor~zotzns for the treatment of identified gram-positive infections, such as Staphylococcr~s spp. or Streptococczrs spp. This approach resulted in initial success rates of 60-70% and overall success rates "with modification" of approximately 90%. Most other groups of investigators (including the EORTC-IATCG) consider trials to be a "failure" if any nonprotocol antibiotics are added. Modifications of initial empiric therapy are discussed later. The carbapenem, imipenern, has been used as empiric therapy in febrile neutropenic patients. Compared with combination regimens, as shown by Leyland and coworkers [71] in 312 episodes of febrile neutropenia, imipenem and piperacillin plus an aminoglycoside achieved success rates of 55% and 53%, respectively. Rolston and associates [72] also compared imipenem with ceftazidime plus amikacin and reported slightly better results with imipenem. On the other hand, other investigators have shown equivocal results with imipenem and ceftazidime considering n~odifications of ~s (MRSA) initial therapy [73]. Methicillin-resistant S t ~ t p h y l o c o c c ~allreus and rnethicillin-resistant Stnylzylococcris epidermidis (MRSE) and Steuzotr.ophonzo~zas nzaltophilirr (formerly Xanthor~zonns mnltophilin) are usually resistant to imipenem, and these organisms may emerge in institutions where it is used extensively. Recently, a direct comparison of ceftazidime and imipenem monotherapy for fever and neutropenia was performed that reported similar overall success rates with each agent ( 9 8 % ) among 204 ceftazidime recipients and 195 imipenem recipients [73].However, antianaerobic agents were added more frequently to ceftazidime, and imipenem was significantly associated with greater toxicity, manifested by nausea and vomiting, requiring discontinuation of imipenem in 10% of recipients and a higher incidence of C. d(ficileassociated diarrhea.

Meropenem, a new carbapenem, also has been used successfully for empiric therapy, as reported by De Pauw and coworkers [74] and by the Meropenem Study Group [75]. Also, the EORTC-IATCG recently reported results of a large study of meropenem alone as monotherapy compared with ceftazidime plus amikacin [76]. A total of 958 patients were evaluated and the results were similar in both arms; a successful outcome was reported in 270 of 483 (56% patients treated with rnonotherapy compared with 245 of 475 (52%) treated with the combination of ceftazidime plus amikacin (P = 0.20). Nephrotoxicity was more frequent in the combination group: six patients in the ceftazidime plus amikacin group and one patient in the meropenem group ( P = 0.07). Gastrointestinal side effects were rarely seen with meropenem in this trial, unlike patients given high-dose imipenem in previous investigations [73].

5.4 Beta-lactamuse inhibitors The poor response of gram-positive infections to ceftazidime plus amikacin in previous trials prompted an investigation by the EORTC-IATCG of the direct comparison of an extended-spectrum penicillin (piperacillin) combiiled with a new beta-lactamase inhibitor (tazobactam) plus amikacin with ceftazidime plus amikacin [49]. Antibiotic treatment was successful in 196 of 364 febrile neutropenic episodes (54%) treated with ceftazidime plus amikacin compared with 210 of 342 (61%) episodes treated with piperacillin-tazobactam plus amikacin (P = 0.05). A significant difference in the response to bacteremic infections between the two groups was also found. Piperacillin/tazobactam plus amikacin was successful in 40 of 80 episodes (50%) compared with 35 of 101 episodes (35%) in the ceftazidime plus amikacin group ( P = 0.05). However, this difference in response rate was not seen when the analysis was limited to gram-positive bacteremic infections. This may have been related to the large number of bacteremias caused by methicillin-resistant coagulase-negative staphylococci (36% of single gram-positive bacteremic episodes). Perhaps the better overall clinical response rate in the giperacillinl tazobactam-treated patients can be attributed to other gram-positive infections. Piperacillin/tazobactam may be particularly useful as initial empiric antimicrobial therapy given the rising incidence of these gram-positive infections. Further investigations addressing the use of this drug as monotherapy are in progress.

5.5 Quinolones The fluoroquinolones have many properties that might benefit the neutropenic host, including [21]: (1) rapid bactericidal activity against gramnegative rods; (2) high peak serum levels (relative to minimum inhibitory concentrations [MICs] of most Enterobacteriaceae) with oral and intravenous administration; (3) relative safety; and (4) no need to monitor serum drug levels.

However, their use is limited by minimal activity against staphylococci, especially MRSA and MRSE, and the quinolones have been implicated in the emergence of streptococcal bacteremia in neutropenic patients [21]. The EORTC-IATCG [77] conducted a trial comparing the use of low-dose intravenous ciprofloxacin with piperacillin plus amikacin for empiric antibiotic therapy of febrile granulocytopenic patients with solid tumors and relatively short duration neutropenia. The study was discontinued prematurely because of the very poor overall success rate of ciprofloxacin [31 of 48 patients (65%) vs. 48 of 53 patients (91%; P = 0.002)J. Patients with gram-positive bacteremia had a particularly poor outcome. This study was performed with lower doses of ciprofloxacin than are acceptable today. Bow et al. [78] conducted an open, nonrandomized study of 53 patients who had 60 neutropenic episodes and reported a reduction in the amount and duration of antibiotic therapy directed against gram-negative organisms as a result of the use of quinolone antibacterial prophylaxis. A gram-negative bacillus was isolated in only 1(1.8%) of 55 febrile neutropenic episodes. These results showed a decrease in gram-negative organism-associated infections; however, carefully randomized controlled studies are needed to further investigate the definitive benefit of prophylactic fluoroquinolones. In summary, there is no antibiotic or combination of antibiotics that is ideal for the empirical treatment of febrile neutropenic patients. However, knowledge of the local patterns of antibiotic susceptibility and resistance at an individual institution and close monitoring of the results of bacteriologic tests and overall responses of patients are helpful to guide initial therapy. Careful daily examination of patients is important to determine early failure of empiric therapy and to identify superinfections or other infectious and noninfectious complications. 5.6 Outpatient management of the febrile ne~ltropenicpatient Traditionally, cancer patients with fever and neutropenia have been treated emergently with broad-spectrum intravenous antibiotics in a hospital setting. Recently, Talcott et al. [14,79] described risk factors predictive of poor outcome in febrile neutropenic patients. Patients at high risk include those with prolonged neutropenia, low platelet counts, high fever, mental status changes, and respiratory distress. These patients require hospitalization and parenteral antimicrobial therapy. Patients in this high-risk category were more likely to have infectious complications (35%) in contrast to the low-risk category ( 5 % ) . No mortality was observed in the low-risk category, but 1.1% of patients in the high-risk category had a fatal outcome. Patients at low risk for complications represent a patient population that may be treated with oral regimens at home or in the hospital. Buchanan and coworkers [11,80] also used various laboratory and clinical criteria to form risk factor assessments in their patients. These studies suggested that both early patient discharge and total outpatient therapy for febrile neutropenia are possible in selected patients (Table 3).

Table 3. Outpatient therapy in low-risk febrile neutropenic patients Therapy IV therapy Ceftazidime or mezlocillin (Talcott et al., 1979)

Hospital discharge criteria

+ gentarnicin

Aztreonam + clindamycin (Rubenstein et al., 1981) Single daily dose of ceftriaxone + amikacin (Martino et al., 1960) Oral therapy acid Ciprofloxacin + arnoxicilli~~/clavula~~ic (Rolston et al., 1982) Ciprofloxacin + clindamycin (Rubenstein et al., 1981) Ofloxacin (Malik et al., 1983)

"Low-risk" patient (absence of comorbidity) < 1 hour to hospital Presence of ho111e co~npanion 2-day observation in hospital "Low-risk" patient (absence of comorbidity, short duration neutropenia) 2-8 hours observation in clinic Short-term hospitalization

"Low-risk" patient (absence of comorbidity. short-duration neutropenia) "Low-risk" patient (absence of co~norbidity. short-duration neutropenia) 2-8 hours observation in clinic "Low-risk" patient (absence of comorbidity. short-duration neutropenia) 2-hour observation in clinic

Varying definitions of "low-risk patient" amongst studies.

Rubenstein et al. [81] performed the first randomized, prospective outpatient trial, which included 83 febrile neutropenic episodes considered to be low risk who were assigned either oral (clindamycin and ciprofloxacin) or IV therapy (clindamycin and aztreonam). Only six patients receiving the oral regimen had to be readmitted to the hospital secondary to persistent fever or associated nephrotoxicity. This trial suggested the feasibility of outpatient therapy for febrile neutropenia in a well-defined subset of patients. Similar results have been observed wtih oral ciprofloxacin plus amoxicillin/clavulanic acid [82]. Malik and coworkers [83] also reported a randomized trial comparing inpatient with outpatient therapy in 182 patients with febrile neutropenia. All patients received oral ofloxacin either in the hospital or as outpatients. The overall outcome was similar in both settings. Twenty-one percent of the patients were readmitted because of persistent fever andlor infection, and mortality was low in both settings. In an attempt to decrease the likelihood of hospital acquired infections for patients receiving intensive chemotherapy, Gillis et al. [84] discharged from the hospital 29 patients treated for acute myelogenous leukemia (AML) who were afebrile and not receiving antibiotics at the completion of chemotherapy (total of 50 courses). Patients were followed closely as outpatients. Of the 50 ambulatory nadir periods, only 3 did not result in hospitalization; 47 of the 50 patients at nadir white blood cell counts required rehospitalization at a mean of 7.2 days after discharge. Four patients had life-threatening complications.

Two patients were admitted with septic shock but rapidly recovered, and two cases resulted in death. After 45 ambulatory nadir periods, patients were discharged a mean of 12.7 days following readmission. It was estimated that approximately 383 hospital days were saved by this policy, representing 16% of total inpatient hospital days. This study further suggests that outpatient monitoring of "low-risk" neutropenic patients following chemotherapy is feasible and may reduce the incidence of nosocomial infections. Future studies may investigate other cost-effective modifications of patient management that include early discharge from the hospital or oral and home IV antibiotics as an alternative to hospitalization [18]. Most recently, De Marie and associates [85] studied patients who received "selective decontamination." Specific antibiotic therapy was given only for demonstrated or proven infections. This approach needs further documentation and may prove to be a cost-effective modification in the future.

5.7 Modificntions of ernpiric nntibiotic therapy After the initial 72-hour period of empiric antibiotic therapy, patients must be reevaluated for their initial clinical response and results of the initial cultures and laboratory studies considered. The expected duration of neutropenia as well as the degree of mucosal damage and other deficits of the immune system must also be recognized, and the treatment regimen may require further modification. In situations where there is a documentated infection, treatment should be tailored to the specific infection encountered. Successful outcome with modification of the empiric antibiotic regimen has previously been demonstrated [67]. Among 156 episodes in which evaluation after 72 hours of empric antibiotics revealed a documented infection, antibiotic therapy modification was required in 22 episodes (34%) of patients receiving combination antibiotics (cephalothin, gentamicin, carbenicillin) and in 45 episodes (40%) of patients receiving monotherapy alone (ceftazidirne). With these modifications, successful outcome was obtained in 98% of patients treated with combination therapy and in 98% of patients treated with monotherapy. Among patients with unexplained fever at the onset of treatment, a total of 394 patients, 78% of those receiving combination therapy and 77% receiving monotherapy, had successful responses without modification of the initial drug regimen. With modification of therapy, successful outcomes were achieved in 199 patients (98%) with combination therapy and in 186 patients (98%) in those receiving monotherapy [2,67]. In patients in whom fever persists after 72 hours of treatment, a new bacterial infection, emergence of bacterial resistance, or superinfection with nonbacterial organisms such as viruses or fungi might be responsible for persistent fever. In these instances, modification of antibiotic treatment is necessary with a change of the antibiotic regimen or the addition of an antifungal, antiviral, or antiprotozoal agent. Reasons for modification of empiric antibiotic therapy include: (1) persistent fever, (2) resistant pathogen, (3)

progression of primary infection, (4) breakthrough bacteremia, ( 5 ) persistence of bacteremia, (6) withdrawal of antibiotic due to toxicity, (7) documented viral or fungal infection, (8) development of shock, and (9) relapse of the primary infection. The risk for secondary infections or superinfections increases the longer granulocytopenia persists, and modification of the initial treatment regimen is needed if they occur. Bacterial infections that are resistant or become resistant during the initial empiric course of antibiotic therapy may be encountered. Also, there may be evidence of infection with a gram-positive organism resistant to the initial beta-lactam and the addition of vancomycin may be necessary, particularly if there is evidence of a catheter-site infection. Breakthrough bacteremia with resistant gram-negative organisms such as Enterobacter, Serrntia, Citrobncter, and Acinetobncter species is of concern, especially when a beta-lactam antibiotic is used alone. The addition of an aminoglycoside or a change in the antibiotic regimen according to bacterial susceptibilities is often necessary. The development of gingivitis or perianal cellulitis, both of which are frequently colonized with gram-negative bacilli and anaerobes, may signal the need for the addition of an antianaerobic agent. Clinical signs suggesting a new site of infection may require the addition of an antifungal or antiviral agent. For example, the development of a pustular skin rash may signify disseminated fungal infection and require the addition of amphotericin B once diagnosis is established with skin biopsy. Retrosternal chest pain or odynophagia may be symptomatic of esophagitis caused by Candida spp., herpes simplex virus, or CMV, The development of new pulmonary infiltrates may be secondary to Pneumocystis carinii, or other fungal or viral pneumonia. Consequently, a decision must be made regarding the addition of trimethoprim/sulfamethoxazole and erythromycin, or whether a diagnostic procedure such as brochoscopy with lavage and/or biopsy or open lung biopsy should be performed. When clinically feasible, a definitive diagnosis should be sought prior to the addition of empiric antimicrobial therpay. Other risk factors, such as the use of corticosteroids or bone marrow transplant, must be also taken into account when determining the most appropriate treatment strategy.

5.8 Addition of vancomycin therapy With the increase in gram-positive infections in the febrile neutropenic patient, controversy has arisen over the use of empiric vancomycin at the onset of treatment. Most suggest that the addition of vancomycin 3 or 4 days into the therapeutic course is safe and effective in patients from whom a gram-positive organism is isolated or who are not responding to initial empirical agents. This approach is reasonable because the reported mortality from gram-positive bacteremia is less than 5% [17]. Of 550 patients studied by Rubin et al., 63% had documented gram-positive cocci infection [17]. Vancomycin was added

subsequent to the initial regimen in 91 % of these patients after gram-positive bacteremia was documented and a 93% success rate was achieved. An EORTC-IATCG study [86] on the use of empiric vancomycin also supports this approach. Ramphal and coworkers [87] compared ceftazidime alone with vancomycin plus ceftazidime as initial therapy. Based on their results they advocate the empiric use of vancornycin only if fever persists for more than 4 days or when fever recurs following an initial response. Specific empirical antistaphylococcal coverage is not routinely recommended for the initial regimen unless the prevalence of MRSA infections in a given institution is very high. 5.9 Addition of antifungal therapy The clinical presentation of invasive fungal infection may only include persistent fever, and fungal infections are often difficult to reliably diagnose early. Because of high mortality rates and evidence of clinical response to ernpiric antifungal therapy [88], treatment can be started early in antibiotic-treated patients with persistent fever. However, some differences of opinion exist

[

IS VANCOMYCllV NECESSARY?

DAYl -

I

I I

I

MONOTHERAPY: CEFTAZIDIhlE, OR CEFEPIME, OR DMIPENEM. OR MEROPENEM

VANCOMYCM

+

CEFTAZJDIME

AMINOGLYCOSIDE

+

ANTIPSEUDOMONAL BETA-LACTAM

DAY 3-4 PERSISTENT FEVER DURING 1ST 3 DAYS NO ETIOLOGY--REASSESS I

I

CONTINUE INITIAL ANTIBIOTICS

ADD AMPHO B WITHIWITHOUT ANTIBIOTIC CHANGE

IF NO CHANGE IN PT. CONSIDER DIC VANCO

IF FEBRILE THROUGH DAY 5 TO 7 AND RESOLUTION O F N+ IS NOT IMMINENT

DAY 3-4 AFEBRILE WITHIN IST 3 DAYS I

I

I NO ETIOLOGY IDENTIFIED] I

I LOW RISK I CHANGE TO ORAL P.O. ANTIBIOTICS: CEFIXIME OR QUJNOLONE-HOME

[ ETIOLOGY

I

IDENTIFIED

1

I

I HIGH RISK I

ADJUST TO MOST APPROPRIATE THERAPY

CONTINUE SAME ANTIBIOTICS

Figure 2. Suggested ernpiric antimicrobial therapy according to recent guidelines of the Infectious Diseases Society of America (88a). DIC = discontinue; GNB = gram-negative bacteria; vanco = vancomycin; N + = neutropenia.

regarding the optimal time to add empiric amphotericin B. The EORTC [88] reported a benefit of empiric amphotericin B added on day 4 of empiric antibacterial therapy. Other investigators 12,891 started amphotericin B following 7 days of persistent fever and neutropenia, and this approach has been associated with defervescence and a reduction in fungal deaths. Present guidelines from several sources suggest starting empiric amphotericin B in the presence of persistent fever >1 week duration, recurrence of fever after 1 week or later in patients with persistiilg neutropenia, and persistent fever at time of recovery from neutropenia (suspicion of hepatosplenic candidiasis, the presence of which can be suggested by the typical appearance of punched out lucencies in these organs on CT) [2,21,90]. The optimal dose of amphotericin B remains uncertain; however, most centers begin with 0.6-1 .O mglkglday. At present, the only antifungal agent of proven benefit for empiric treatment of persistently febrile neutropenic patients is ainphotericin B [2,89]. The toxicity of amphotericin B and the advent of new imidazoles has led to continued search for optimal empiric management of fungal infections in the neutropenic host. Fluconazole may be an alternative empiric treatment regimen for candidal infections in the iminunocompromised host, especially those with hepatosplenic candidiasis [91.92]. De Pauw et al. [93] performed an open study of fluconazole in 24 patients, 9 of whom had acute and 15 chronic disseminated candidiasis. Successful clinical response occurred in 67% of patients with acute disseminated candidiasis and 86% of cases with chronic developed in five patients who disseminated candidiasis. Aspergillus fi~nzignt~ts were persistently neutropenic. Although this clinical trial showed promising results, caution must be taken with some species that may be intrinsically resistant to fluconazole, such as C. kvzisei and C. glnbrnta. Although fluconazole is effective in the treatment of candidal infections, it is not active against Aspergillzrs spp. Itraconazole is effective in patients with Aspergillus spp. and other invasive fungal infections. Van't Wout et al. 1941 compared the efficacy of itraconazole with amphotericin B against systemic fungal infection in neutropenic patients. Of the 32 patients studied, 10 of 16 (63%) had a clinical response with amphotericin B and 9 of 16 (56%) with itraconozole (P > 0.9). Itraconazole seemed to be more effective against Aspergillus spp., whereas amphotericin B appeared more effective against candidal infections; however, the sample size was sinall and the difference was not statistically significant. Additional studies are needed to further compare the role of itraconazole as an alternative to amphotericin B in invasive systemic fungal infections. New preparations of lipid-associated amphotericin B have been formulated in an effort to reduce nephrotoxicity. These include ~mbisome"'(liposomal arnphotericin B), ~ b e l c e t ' (Amphotericin B Lipid Complex, ABLC), and ~mphotec@ (Amphotericin B cholesteryl sulfate complex, ABCD). These compounds differ in structural and pharmacokinetic characteristics; however, clinical data on their efficacy when compared with conventional amphotericin

B are scant and additional randomized comparative trials are needed to further assess their clinical value with respect to efficacy and reduction of toxicity [95-981. 6. Duration of empiric antimicrobial therapy The duration of empiric antimicrobial therapy remains controversial. Pizzo and colleagues showed that patients who remain granulocytopenic and febrile and had antibiotics discontinued after 7 days developed rebound infection and/or shock [99]. Young suggested that antibiotics could be stopped in febrile patients with no bacterial diagnosis after 7 days if the granulocyte count exceeds 500/mmzfor 2 days [loo], but they should be continued if fever and/or neutropenia persists (Figure 3). Failure to respond to initial empiric antibiotic therapy warrants further diagnostic testing for occult infections of the intraabdominal, perirectal, pleural, and sinus spaces, and in these instances antibiotic therapy should be continued despite an increase in the granulocyte count. Prolonged antibiotic therapy and profound granulocytopenia predispose irnmunocompromised patients to fungal superinfections, and empiric amphotericin B for antifungal therapy is recommended in patients who remain febrile after 4-7 days of antibiotic therapy.

7. Adjuvant therapy The adjuvant use of colony stimulating factors (CSFs), immunoglobulins, and granulocyte transfusions from CSF-stimulated donors and the use of other cytokines have been suggested to bolster host defenses in neutropenic

+ -1

DC AND REASSESS

+ PERSISTENT FEVER

AFEBRILE BY DAY 3

1

ANC > 500 STOP IN 4-5 DAYS: REEVALUATE

ANC < 500 CONTINUE FOR 14D

a CONTIhIVE UNTIL

Figure 3. Duration of antibiotic therapy according to recent guidelines from the Infectious Diseases Society of America (88a). ANC = absolute neutrophil count; D = days; DC = discontinue.

patients. The most commonly used agents are the hematopoietic growth factors, granulocyte macrophage-colony stimulating factor (GM-CSF) and granulocyte-colony stimulating factor (G-CSF). These cytokines are glycoproteins that stimulate the proliferation and maturation of bone marrow stem cell lines, resulting in an increase in peripheral granulocyte counts. Several studies [101-1031 have revealed a reduction in the duration and severity of neutropenia in patients receiving chemotherapy. However, in these studies hematopoeitic growth factors were administered prophylactically with the intent of shortening the period of risk and lowering the incidence of infectious complications. More recently, investigations exploring the interventional rather than prophylactic administration of CSFs demonstrated an equivocal effect on neutropenia and no reduction in the frequency of infections. Also, with respect to their use in patients with acute myelocytic leukemia, CSFs have not stimulated regrowth of leukemic cells but also did not result in an increase of disease free survival [104]. Hence, hematopoietic growth factors are currently recommended only for prophylactic use in patients with prolonged and severe neutropenia and at high risk of infectious complications. According to the American Society of Clinical Oncology (ASCO) guidelines [105], CSFs are indicated for primary prophylaxis if the expected incidence of neutropenic fever is greater that 40% and for secondary prophylaxis if chemotherapy doses cannot be reduced to minimize infection. The ASCO guidelines do not advocate the use of these cytokines routinely in all patients receiving chemotherapy. Although CSFs clearly have decreased the duration of neutropenia, beneficial results of the incidence of severe infection and the long-term effects on antileukemia treatment remain to be determined definitively [106]. Other less commonly used adjunctive therapies also have been utilized to improve the host defenses in neutropenic patients. Passive immunization with antiendotoxin antibodies is still undergoing clinical evaluation. In an early clinical trial using antibody to core glycolipid of the Enterobacteriaceae (J5 antiserum), mortality was decreased in patients with gram-negative bacteremia, except in patients who had neutropenia [107]. This early trial ignited further investigational work on the use of core glycolipid for passive immunization. More recent trials using monoclonal core glycolipid antibodies have produced inconclusive results. Clinical studies of anti-lipid A monoclonal antibodies [108,109], specifically E5 and HA-lA, seemed promising because both antibodies appeared to protect subsets of patients with sepsis syndrome. E5 appeared to improve the survival of patients with gram-negative sepsis with refractory shock, but only when they were bacteremic. Questions concerning the clinical efficacy and cost effectiveness of these agents in patients with neutropenia and their ultimate impact on survival mandate the need for further investigation. Antibody to tumor necrosis factor (TNF) and interleukin-1 receptor antagonist have been disappointing. Recently, the efficacy and safety of antiTNF-alpha monoclonal antibody (Mab) was studied in a randomized

double-blind placebo-controlled trial [110]. A total of 971 patients were prospectively stratified into shock or nonshock groups and then randomized to receive a single infusion dose of 15mgJkg of TNF-alpha Mab, 7.5mg/kg of TNF-alpha Mab or placebo. There was no reduction in mortality between placebo and TNF-alpha Mab in all infused patients. Since the cloning of the interleukin-1 receptor antagonist (IE-lra) there has been intensive research on the genetics and potential therapeutic value of this protein; however, the role of IL-lra in normal physiology or in host defense mechanisms remains unclear. Preliminary results of clinical trials in animal models and in humans indicated a possible benefit of IL-lra in sepsis syndrome, rheumatoid arthritis, and graft-versus-host disease, but clinical studies were disappointing [I 11-1131. Intravenous immunoglobulins have been found to reduce the frequency of respiratory infections [I071 in patients with chronic lymphocytic leukemia in whom antibody production is deficient. CMV-antibody-rich IV immunoglobulin to prevent CMV infection in CMV-seronegative blood recipients has also been investigated [114].

References 1. Bodey GP, Buckley M, Sathe YS, Freireich EJ. Quantitative relationships between circulating leukocytes and infection in patients with acute leukemia. Ann Intern Med 1966;64:328340. 2. Lee JW, Pizzo PA. Management of the cancer patient with fever and prolonged neutropenia. Hematol Oncol Clin North Am 1993;7:937-960. 3. Pizzo PA, Commers JR, Cotton D, et al. Approaching the controversies in the antibacterial management of cancer patients. Am J Med 1984;76:436439. 4. Schimpff SC. Empiric antibiotic therapy for granulocytopenic cancer patients. Am J Med 1989;8O(S~ppl.5C):13-20. 5. Pizzo PA, Robichaud KJ, Gill FA, Witebsky PG. Empiric antibiotic and antifungal therapy for cancer patients with prolonged fever and granulocytopenia. Am J Med 1982; 72:101-111. 6. Desser RK. Ultman JE. Risk of severe infection in patients with Hodgkin's disease or lymphoma after diagnostic laparotomy and splenectomy. Ann Intern Med 1972;77:143-146. 7. Mayer KH, Opal SM. Antibacterial therapy in patients with malignancies. Cancer Metastas Rev 1987;5:271-293. 8. Siber GR, Weitzman SA, Aisenberg AC, et al. Impaired antibody response to pneumococcal vaccine after treatment for Hodgkin's disease. N Engl J Med 1978;299:442-448. 9. Lamb D, Pilney F, Kelly WD, Good RA. A comparative study of the incidence of anergy with carcinoma, leukemia, Hodgkin's disease and other lymphomas. J Immunol1962;89:555558. 10. Zinner SH. Treatment and prevention of infection in immunocompromised hosts. In: Gorbach SL, Bartlett JG, Blacklow NR, eds. Infectious Diseases. Philadelphia: W.B. Saunders, 1992, pp. 1057-1064. 11. Buchanan GR. Approach to treatment of the febrile cancer patient with low risk neutropenia. Hematol Oncol Clin North Am 1993;7:919-935. 12. Freifeld AG, Pizzo PA. The outpatient management of febrile neutropenia in cancer patients. Oncology 1996;10:599-612.

13. Talcott JA, Finberg R, Mayer RJ, et al. The medical course of cancer patients with fever and neutropenia. Clinical identification of a low risk subgroup at presentation. Arch Intern Med 1988;148:2561-2568. 14. Talcott JA, Siege1RD, Finberg R, Goldman L. Risk assessment in cancer patients with fever and neutropenia: A prospective, two-center validation of a prediction rule. J Clin Oncol 1992:10:316-322. 15. Pappo AS, Buchanan GR. Predictors of bacteremia in febrile neutropenic children with cancer (abstr.). Proc Annu Meet Am Soc Clin Oncol 1991:10:A1167. 16. Patrick CC. Coagulase-negative staphylococci: Pathogens with increasing clinical significance. J Pediatr 1990;116:497-507. 17. Rubin M, Hathorn JW. Marshall D, et al. Gram-positive infections and the use of vancomycin in 550 episodes of fever and neutropenia. Ann Intern Med 1988;108:30-35. 18. Anaissie EJ. Vadhan-Raj S. Is it time to redefine the management of febrile neutropenia in cancer patients? Am J Med 1995:98:224-231. 19. Klastersky J. Febrile neutropenia. Support Care Cancer 1993:1:233-239. 20. Zinner SH, Gaya H. Glauser M, et al. The International Antimicrobial Therapy Group of the European Organization for Research and Treatment of Cancer. Con~binationof amikacin and carbenicillin with or without cefazolin as empirical treatment of febrile neutropenic patients. J Clin Oncol 1983:1:597-603. 21. Giamarellou H. Empiric therapy for infections in the febrile, neutropenic, compromised host. Med Clin North Am 1995:79:559-580. 22. Meunier F. Infections in patients with acute leukemia and lymphoma. In: Mandell GL, Douglas RG, Bennett JE. eds. Principles and Practice of Infectious Disease, 4th edi. New York: Churchill Livingstone, 1995, pp. 2675-2686. 23. Sickles EA, Greene WH, Wiernick PH. Clinical presentation of infection in granulocytopenic patients. Arch Intern Med 1975:135:715-719. 24. Rubin RH, Green R. Clinical approach to the compromised host with fever and pulmonary infiltrates. In: Rubin RH, Young LS, eds. Clinical Approach to Infection in the Compromised Host. New York: Plenum, 1994. pp. 121-161. 25. Cockerill FR 111, Wilson WR, Carpenter HA, et al. Open lung biopsy in immunocon~promisedpatients. Arch Intern Mcd 1985;145:1398-1404. 26. Cheson BD. Samlowski WE, Tang TT, Spruance SL. Value of open lung biopsy in 87 immunocompromised patients with pulmonary infiltrates. Cancer 1985:55:453-459. 27. Peterson DE, Minah GE, Overholser CD, et al. Microbiology of acute periodontal infection in myelosuppressed cancer patients. J Clin Oncol 1987;5:1461-1468. 28. Gerard M, Defresne N, Daneau D, et al. Incidence and significance of Clostridi~ir?~ dificile in hospitalized cancer patients. Eur J Clin Microbial Infect Dis 1988:7:274-278. 29. Milligan DW, Kelly JK. Pseudomembranous colitis in a leukemia unit: A report of fatal cases. J Clin Path01 397932: 1237-1243. 30. Varki AP, Armitage JO, Feagler JR. Typhlitis in acute leukemia: Successful treatment by early surgical intervention. Cancer 1979;43:695-697. 31. Anaissie E. Opportunistic mycoses in the immunocompromised host: Experience at a cancer center and review. Clin Infect Dis 1992;14(Suppl. 1):S43-S53. 32. Bodey GP. Jadeja L, Elting L. Pseudomonas bacteremia. Retrospective analysis of 410 episodes. Arch Intern Med 1985:145:1621-1629. 33. Hiemenz J. Skelton J, Pizzo PA. Perspective on the management of catheter-related infections in cancer patients. Pediatr Infect Dis 1986;5:6-11. 34. Johnson PR, Decker MD, Edwards KM, et al. Frequency of broviac catheter infections in pediatric oncology patients. J Infect Dis 1986:154:570-578. 35. Kralner BS, Pizzo PA, Robichaud KJ. et al. Role of serial microbiologic surveillance and clinical evaluation in the management of cancer patients with fever and granulocytopenia. Am J Med 1982;72:561-568. 36. The International Antimicrobial Therapy Cooperative of the European Organization for Research and Treatment of Cancer (Writing Committee: Calandra T. Zinner SH, Viscoli C,

37. 38. 39.

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

48. 49.

50.

51. 52. 53. 54. 55.

56. 57.

de Bock R, Gaya H, Meunier F, Klastersky J, Glauser M). Efficacy and toxicity of single daily doses of amikacin and ceftriaxone versus multiple daily doses of amikacin and ceftazidime for infection in patients with cancer and granulocytopenia. Ann Int Med 1993;119:584593. Elting LS, Bodey GP, Keefe BH. Septicemia and shock syndrome due to viridans streptococci: A case-control study of predisposing factors. Clin Infect Dis 1992:14:1201-1207. Handwerger S. Horowitz H, Coburn K, Kolokathis A , Wormser GP. Infection due to Lerlconostoc species: Six cases and review. Rev Infect Dis 1990;12:602-610. Swartz MN. Cellulitis and subcutaneous tissue infections. In: Mandell GL, Douglas RG. Bennett JE, eds. Principles and Practice of Infectious Disease, 4th ed. New York: Churchill Livingstone, 1995. pp. 909-929. Zinner SH. New and unusual infections in neutropenic patients. In: Klastersky J, ed. Infectious Complications of Cancer. Boston: Kluwer Academic. 1995, pp. 173-184. Anaissie E. Kantarjian H. Ro J , et al. The emerging role of Fusarium infections in patients with cancer. Medicine 1988;67:77-83. Martino P, Gastaldi R, Raccah R, Girmenia C. Clinical patterns of Fzunrir~nzinfections in immunocompromised patients. J Infect 1994:28(Suppl. 1):7-15. Raad I, H a c h e ~ nR. Treatment of central venous catheter-related fungemia due to F~rstrriunz oxysponinr. Clin Infect Dis 1995:20:709-711. Walter E A , Bowden R A . Infection in the bone marrow transplant recipient. Infect Dis Clin North Am 1995;9:823-847. Samonis G , Bafaloukos D. Fungal infections in cancer patients: An escalating problem [review]. In Vivo 1992;6:183-193. Gokaslan A, Anaissie E. A novel murine model of disseminated trichosporonosis. Infect Immun 1992;60:3339-3344. Schimpff SC. Satterlee W, Young VM. Serpick A. Empiric therapy with carbenicillin and gentamicin for febrile patients with cancer and granulocytopenia. N Engl J Med 1971; 284: 1061-1 065. Gibson J. Johnson L, Snowdon L. et al. Trends in bacterial infections in febrile neutropenic patients: 1986-1992. Aust N Z 9 Med 1994:24:374-377. Coinetta A, Zinner S, de Bock R, et al. and the International Antimicrobial Therapy Cooperative Group of the European Organi~ationfor Research and Treatment of Cancer. Piperacillin-tazobactam plus alnikacin versus ceftazidime plus arnikacin as empiric therapy for fever in granulocytopenic patients with cancer. Antimicrob Agents Chemother 1995: 39:445-452. Fergie JE, Shema SJ, Lott L. Crawford R. Patrick CC. Pseudonzotlns cieruyiizosti bacteremia in immunocompromised children: Analysis of factors associated with a poor outcome. Clin Infect Dis 1994;18:390-394. Schimpff SC, Greene WH, Young VM. et al. Significance of Pseudon~onasrreniginosn in the patient with leukemia and lymphonla. J Infect Dis 1974;130(Suppl.):S24-S31. Hersch EM. Bodey GP, Nies BE, Freireich EJ. Causes of death in acute leukemia. JAMA 1965:L93:105-109. Rubin M. Hathorn JW, Pizzo PA. Controversies in the management of febrile neutropenic cancer patients. Cancer Invest 1988;6:167-1 84. Klastersky .I.Zinner SH. Synergistic combination of antibiotics In gram-negative bacillary infections. Rev Infect Dis 1982;4:294-298. Klastersky J. Schimpff SC. Gaya H. Glauser MP. Zintler SH. EORTC Intertlational Antimicrobial Therapy Cooperative Group. Prospective randomized comparison of three antibiotic regimens for empirical therapy of suspected bacteren~icinfection in febrile granulocytopenic patients. Antimicrob Agents Chemother 1986:29:263-270. Tattersall MH. Spiers ASD. Darrell JH. Initial therapy with combination of five antibiotics in febrile pat~entswith leukaemia and neutropenia. Lancet 1972:1:162-165. Zinner SH, Gaya H. Glauser M, Klastersky J, Schimpff SC. Tattersall MHN. EORTC International Antimicrobial Therapy Cooperative Group. Combination of amikacin and

58.

59.

60.

61. 62.

63. 64. 65.

66. 67.

68. 69.

70.

71.

72.

73.

74. 75.

76.

carbenicillin with or without cefazolin as empirical treatment of febrile neutropenic patients. J Clin Oncol 1983;1:597-603. Calandra T, Klasterksy J, Gaya H, Glauser MP, Meunier F, Zinner SH. The EORTC International Antimicrobial Therapy Cooperative Group. Ceftazidime combined with a short or long course of amikacin for empirical therapy of gram-negative bacteremia in cancer patients with granulocytopenia. N Engl J Med 1987;317:1692-1698. Gibson J, Johnson L, Snowdon L, et al. Single daily ceftriaxone and tobramycin in the empirical management of febrile neutropenic patients: A randomised trial. Int J Hematol 1993;58:63-72. Martino P, Girmenia C, Raccah R, et al. Single daily dose ceftriaxone plus amikacin treatment of febrile episodes in neutropenic patients attending day hospital for hematologic malignancies. Oncology 1992;49:49-52. Gutmann L, Williamson R,Kitzis M, et al. Synergism and antagonism in double beta-lactam antibiotic combinations. Am J Med 1986;80(Suppl.5C):21-29. Joshi JH, Newman KA, Brown BW, et al. Double beta-lactam regimen compared to an aminoglycoside/beta-lactam regimen as empiric antibiotic therapy for febrile granulocytopenic cancer patients. Support Care Cancer 1993;1:186-194. Young LS. Double beta-lactam therapy in the im~nunocompron~ised host, J Antimicrob Chemother 1985;16:4-6. Pizzo PA, Thaler M, Hathorn J, et al. New beta-lactam antibiotics in granulocytopenic patients: New options and new questions. Am J Med 1985;79(Suppl. 2A):75-82. De Pauw BE, Deresinski SC, Feld R, Lane-Allman EF, Donnelly JP for The Interncontinental Antimicrobial Study Group. Ceftazidime compared with piperacillin and tobramycin for the empiric treatment of fever in neutropenic patients with cancer. A multicenter randomized trial. Ann Intern Med 1994;120:834-844. Bizette GA, Brooks BJ Jr., Alvarez S. Ceftazidime as monotherapy for fever and neutropenia: Experience in a community hospital. J La State Med Soc 1994;146:448-452. Pizzo PA, Hathorn JW, Hiemenz J, et al. A randomized trial comparing ceftazidime alone with combination antibiotic therapy in cancer patients with fever and neutropenia. N Engl J Med 1986;315:552-558. Sanders JW, Powe NR, Moore RD. Ceftazidime monotherapy for empiric treatment of febrile neutropenic patients: A meta-analysis. J Infect Dis 1991;164:907-916. Naumovski L, Quinn JP, Miyashiro D, et al. Outbreak of ceftazidime resistance due to a novel extended-spectrum beta-lactamase in isolates from cancer patients. Antimicrob Agents Chemother 1992;36:1991-1996. Eggimann P, Glauser MP, Aoun M, Meunier F, Calandra T. Cefepime monotherapy for the empirical treatment of fever in granulocytopenic cancer patients. J Antimicrob Chemother 1993;32(Suppl. B):151-163. Leyland MJ, Bayston KF, Cohen J, et al. A comparative study of imipenem versus piperacillin plus gentamicin in the initial management of febrile neutropenic patients with haematological malignancies. J Antimicrob Chemother 1992;30:843-854. Rolston KV. Berkey P, Bodey GP, et al. A comparison of imipenem to ceftazidime with or without amikacin as empiric therapy in febrile neutropenic patients. Arch Intern Med 1992;152:283-291. Freifeld AG, Walsh T, Marshall D, et al. Monotherapy for fever and neutropenia in cancer patients: A randomized comparison of ceftazidime versus imipenem. J Clin Oncol 1995; 13:165-176. Donnelly JP, Horrevorts AM, Sauerwein RW, De Pauw BE. High-dose meropenem in meningitis due to Pseirdon~onasaerrlginosa [letter]. Lancet 1992;339:1117. The Meropenern Study Group of Leuven, London and Nijmegen. Equivalent efficacies of meropenem and ceftazidime as empirical monotherapy of febrile neutropenic patients. J Antimicrob Chemother 1995;36:185-200. Cometta A, Calandra T, Gaya H, et al. and the International Antimicrobial Therapy Cooperative Group of the European Organization for Research and Treatment of Cancer and

77.

78.

79.

80.

81. 82.

83.

84. 85.

86.

87.

88.

88a. 89.

90. 91. 92.

The Gruppo Italiano Malattie Ematologiche Maligne Dell'Adulto Infection Program. Monotherapy with meropenem versus combination therapy with ceftazidime plus amikacin as empiric therapy for fever in granulocytopenic patients with cancer. Antimicrob Agents Chemother 1996;40:1108-1115. Meunier F, Zinner SH, Gaya H, Calandra T, Viscoli C, Klastersky J, Glauser M (EORTCIATCG). Prospective randomized evaluation of ciprofloxacin versus piperacillin plus amikacin for enlpiric antibiotic therapy of febrile granulocytopenic cancer patients with lymphomas and solid tumors. Antimicrob Agents Chemother 1991;35:873-878. Bow EJ, Loewen R, Vaughan D. Reduced requirement for antibiotic therapy targeting gram-negative organisms in febrile, neutropenic patients with cancer who are receiving antibacterial chemoprophylaxis with oral quinolones. Clin Infect Dis 1995;20:907-912. Talcott JA, Whalen A, Clark J, Rieker PP, Finberg R. Home antibiotic therapy for low-risk cancer patients with fever and neutropenia: A pilot study of 30 patients based on a validated prediction rule. J Clin Oncol 1994;12:107-114. Mullen CA, Buchanan G R . Early hospital discharge of children with cancer treated for fever and neutropenia: Identification and management of the low-risk patient. J Clin Oncol 1990;8:1998-2004. Rubenstein EB. Rolston K, Benjamin RS, et al. Outpatient treatment of febrile episodes in low-risk neutropenic patients with cancer. Cancer 1993;71:3640-3646. Kolston K, Rubenstein EB, Elting LS. et al. Ambulatory management of febrile episodes in low-risk neutropenic patients - an acceptable alternative to hospital based therapy. In: Program Proceedings and Abstracts of the 6th International Symposium Supportive Care in Cancer, 1994. Malik IA. Khan WA, Karim M, Aziz Z, Khan MA. Feasibility of outpatient management of fever in cancer patients with low-risk neutropenia: Results of a prospective randomized trial. A m J Med 1995;98:224-231. Gillis S, Dann EL. Rund D . Selective discharge of patients with acute myeloid leukemia during chemotherapy-induced neutropenia. A m J Hemtaol 1996;51:26-31. D e Marie S, Van den Broek PJ. Willemze R, Van Furth R. Strategy for antibiotic therapy in febrile neutropenic patients on selective antibiotic decontamination. Eur J Clin Microbiol Infect Dis 1993;122397-906. E O R T C International Antimicrobial Therapy Cooperative Group and the National Cancer Institute of Canada CIinical Trials Group. (Writing Committee: Calandra T. Zinner SH, Glauser MP, Meunier F, Viscoli C, Feld R, Gaya H, and Klastersky J.) Vancomycin added to empirical combination antibiotic therapy for fever in granulocytopenic cancer patients. J Infect Dis 1991;163:951-958. Ramphal R, Bolger M. Oblon DJ, et al. Vancomycin is not an essential component of the initial e~npirictreatment regimen for febrile neutropenic patients receiving ceftazidime: A randomized propsective study. Antimicrob Agents Chemother 1992;36:1062-1067. E O R T C International Antimicrobial Therapy Cooperative Group. (Writing Committee: Meunier F, Gaya H, Calandra T. Glauser M, Klastersky J, Sylvester R, and Zinner S.) Empiric antifungal therapy in febrile granulocytopenic patients. Am J Med 1989;86:668672. Hughes WT. Armstrong D, Bodey GP, et al. 1997 guidelines for the use of antimicrobial agents in neutropenic patients with unexpected fever. Clin Infect Dis 1997;551-573. Pizzo PA. Empirical therapy and prevention of infection in the immunocompromised host. In: Mandell G, Douglas RG. Bennett JE, eds. Principles and Practice of Infectious Diseases. New York: Churchill Livingstone, 1995, pp. 2686-2696. AnttiIa VJ, Ruutu P, Bondestam S, et al. Hepatosplenic yeast infection in patients with acute leukemia: A diagnostic problem. Clin Infect Dis 1994;18:979-981. Anaissie E , Pinczowski H. Invasive candidiasis during granulocytopenia. Recent Results Cancer Res 1993;132:137-145. Flannery MT, Simmons DB. Saba H. Altus P, Wallach PM, AdeLman HM. Fluconazole in the treatment of hepatosplenic canclidiasis. Arch Int Med 1992;152:406-408.

93. D e Pauw BE. Raemaekers JM, Donnelly JD, Kullberg BJ, Meis JF. An open study on the safety and efficacy of fluconazole in the treatment of disseminated Candida infections in patients treated for hematologic malignancy. Ann Hemato1 1995;70:83-87. 94. Van't Wout JW, Novakova I, Verhagen CA, et al. The efficacy of itraconazole against systemic fungal infections in neutropenic patients: A randomised conlparative study with amphotericin B. J Infect 1991;22:45-52. 95. D e Marie S, Janknegt R, Bakker-Woudenberg IA. Clinical use of liposomal and lipidcomplexed amphotericin B [review]. J Antimicrob Chemother 1994;33:907-916. 96. Janknegt R , D e Marie S, Bakker-Woudenberg IA. Liposornal and lipid formulations of amphotericin B. Clin Pharmacokinet 1992;23:279-291. 97. Adler-Moore J. AmBisoine targeting to fungal infections. Bone Marrow Transplant 1994; 14(Suppl. 5):S3-S7. 98. Lopez-Berestein G , Bodey GP, Fainstein V. et al. Treatment of systemic fungal infections with liposomal amphotericin B. Arch Intern Med 1989;149:2533-2536. 99. Pizzo PA, Robichaud KJ, Gill FA, et al. Duration of empiric antibiotic therapy in granulocytopenic patients with cancer. A I J~Med 1979;67:194-200. 100. Young LS. Neutropenia: Antibiotic combinations for empiric therapy. Eur J Clin Microbiol Infect Dis 1989;8:118-122. 101. Gabrilove JL, Jakubowski A, Scher H, et al. Effect of granulocyte colony-stimulating factor on neutropenia and associated morbidity due to chemotherapy for transitional cell carcinoma of the urothelium. N Engl J Med 1988;318:1414-1422. 102. Dibenedetto SP, Ragusa R, Ippolito AM. et al. Assessment of the value of treatment with granulocyte colony-stimulating factor in children with acute lymphoblastic leukemia: A randomized clinical trial. Eur J Haernatol 1995;55:93-96. 103. Stone RM, Berg DT, George SL, et al.. for The Cancer and Leukemia Group B. Granulocyte-macrophage colony-stimulating factor after initial chemotherapy for elderly patients with primary acute inyelogenous leukemia. N Engl J Med 1995;332:1671-1677. 104. Maher DW, Lieschke GJ. Green M, et al. Filgrastim in patients with chemotherapy-induced febrile neutropenia. A double-blind, placebo-controlled trial. Ann Intern Med 1994;121:492501. 105. American Society of Clinical Oncology. American Society of Clinical Oncology recommendations for the use of hematopoietic colony stimulating factors: Evidence based. clinical practice guidelines [special article]. J Clin Oncol 1994;12:2471-2508. 106. Roilides E , Pizzo PA. Perspectives on the use of cytokines in the management of infectious complications of cancer. Clin Infect Dis 1993:17(Suppl. 2):S385-S389. 107. Pizzo PA. Management of fever in patients with cancer and treatment-induced neutropenia. N Engl J Med 1993;328:1323-1332. 108. Ziegler EJ, Fisher CJ Jr.. Sprung CL. et al. Treatment of gram-negative bacteremia and septic shock with HA-1A human n~onoclonalantibody against endotoxin: A randomized. double-blind, placebo-controlled trial. N Engl J Med 1991:324:429-436. 109. Greenman RI, Schein RMH, Martin MA, et al. and The Xoma Sepsis Study Group. A controlled clinical trial of ES murine monoclonal IgM antibody to endotoxin in the treatment of gram-negative sepsis. JAMA 109 1 :266:1097-1 102. 110. Abraham E , Wunderink R, Silverman H, et al. for The TNF-u MAb Sepsis Study Group. Efficacy and safety of monoclonal antibody to human tumor necrosis factor a in patients with sepsis syndrome. A randomized, controlled, double-blind, multicenter clinical trial. J A M A 1995;273:934-941. 11 1. Dinarello CA. Blocking interleukin-1 receptors. Int J Clin Lab Res 1994:24:61-79. 112. Arend WP. Interleukin-1 receptor antagonist. Adv Iinrnunol 1993;54:167-227. 113. Lennard AC. Interleukin-l receptor antagonist. Crit Rev Irnmunol 1995;15:77-105. 114. Winston DJ, H o WG, Bartani K, Champlin RE. Intravenous immunoglobulin and CMVseronegative blood products for prevention of CMV infection and disease in bone marrow transplant recipients. Bone Marrow Transplant 1993;12:283-288.

4. Infections associated with solid tumors Sarah H. Sutton and John P. Flaherty

1. Introduction According to American Cancer Society estimates, leukemias and lymphomas accounted for only 7% of male and 6% of female cancers diagnosed and 8% of cancer deaths; the remaining cancers were solid tumors [I]. Although solid tumors account for the vast majority of cancer in adults, severe infectious complications in these patients are far less common than in patients with hematologic malignancies. For instance, Elting et al. [2] reported that polymicrobial sepsis was 16 times less common per patient admission in patients with solid tumors than patients with acute leukemia. Similarly, Mayo and Wenzel [3] found that nosocomial bloodstream infections were 15 times less likely in solid tumor patients than leukemia patients. Nevertheless, patients with solid tumors complicated by infection are not unusual. We reviewed the last 100 cancer patients seen by our infectious disease inpatient consultation service and the majority (62) had underlying solid tumors (J. P. Flaherty, unpublished observation). Common scenarios included wound infection, pneumonia, intravascular catheter-related sepsis, and fever and neutropenia following intensive chemotherapy. The function of the immune system is a major factor in determining the spectrum of infections to which cancer patients are vulnerable. Unlike solid tumor patients, those with henlatologic malignancies have a malignant leukocyte clone that does not function correctly, replaces the marrow, or interferes with specific immune functions. The presence of a solid tumor may have indirect deleterious effects upon the immune system, but these are poorly understood. For example, depressed CD4 and CD8 lymphocyte counts have been measured in solid tumor patients [4], but the clinical significance of these abnormalities is uncertain. A subset of solid tumor patients receive intensive chemotherapy that is complicated by neutropenia. Patients with hematologic malignancies undergo prolonged, often uninterrupted, courses of combination chemotherapy, resulting in extended periods of neutropenia. Because the periods of neutropenia that most solid tumor patients experience tend to be brief, this patient population has a lower risk of infection than neutropenic patients with hematologic Grrry A . Noskit1 (en), M A N A G E M E N T O F INFECTIOUS COIMPLICATIONS IN C A N C E R PATIENTS.

0 1998. Kllr~vrrAcndernic Publishers, Boston. All rights reserved.

malignancies. When febrile and neutropenic, the solid tumor patient does have a high rate of infection. Pizzo and coworkers [5] documented a specific infectious source in 52% of febrile, neutropenic children and young adults with solid tumors, which is similar to the 55% of those with leukemia and 46% of those with lymphoma. While the majority of solid tumor patients do not experience severe treatment-related immunosuppression, they cannot be considered normal hosts. First, because solid tumors are more likely to occur in the elderly, many solid tumor patients have changes in the immune system associated with the aging process. Also, because solid tumor patients often have indolent presentations, malnutrition and cancer cachexia may be severe. Several comorbidities, such as chronic obstructive pulmonary disease (COPD) and poor dentition, may increase infection risk at presentation and throughout therapy. Solid tumor patients are at risk of acquiring nosocomial infections because they often undergo invasive diagnostic and therapeutic procedures, intravenous line placement, and hospitalization. As a result of antibiotic therapy, they are at increased risk of acquiring Clostridium difJicile colitis and infection with resistant organisms such as vancomycin-resistant Enterococci (VRE). Infections in of the non-neutropenic solid tumor patient contribute to significant morbidity and mortality. For example, the non-neutropenic counterparts in Pizzo's study mentioned earlier had identifiable sources of infection in 17 of 112 febrile episodes (15%) for solid tumor patients, which compared with 21 O/O for leukemic patients and 17% for lymphoma patients. In a review of bacteremic and fungemic episodes in patients with solid and liquid tumors at Memorial Sloan-Kettering Cancer Center (New York City), investigators reported a 26.6% mortality rate for 192 non-neutropenic episodes of sepsis in solid tumor patients [6]. As treatment regimens rate for these malignancies become more aggressive, the role of infection in these patients is likely to grow.

2. Infectious complications of lung cancer Lung cancer is the leading cause of cancer deaths in the United States, contributing to 141,285 deaths in 1990 [I]. In 1994, lung cancer accounted for 16% of all new cancer cases among males and 13% among females [I]. Infection plays a potentially critical role in the outcome of patients with lung cancer. The lung must continue to function as an organ of gas exchange during and after cancer therapy. Exposure to the external environment, and the attendant risk of exposure to infectious pathogens, must be maintained for gas exchange to occur. In addition, the lung serves as a vast vascular bed, and hematogenous seeding by bacteria and fungi can occur. The lung's defenses against these insults may be impaired by lung cancer and its treatment. The lung cancer patient has altered host defenses in several ways. Lung cancer commonly occurs in older adults and host defenses diminish with age.

Age-dependent changes that predispose to respiratory infections include an increased tendency to aspirate, decreased cough reflex, decreased mucociliary clearance, and increased oropharyngeal colonization by aerobic gramnegative bacilli [7]. Increasing age is also associated with decreased cellular and humoral immunity. In addition, there are changes that are associated with the cancer itself. The most apparent is local bronchial obstruction by a tumor mass, leading to impaired clearance of respiratory secretions. Lung cancer, particularly with advanced disease, has also been associated with diminished delayed cutaneous hypersensitivity reactions [8f.The lung cancer patient may present with serious comorbidities such as COPD or malnutrition, which can prolong postsurgical or postpneumonia ventilatory support or require tracheostomy. Such complications can increase pulmonary and nonpulmonary infection risks. Finally, treatment of lung cancer can increase infection risk. Most patients with stage I and I1 non-small cell carcinoma of the lung undergo surgery as initial therapy. In a study of 103 such patients, the risk of postthoracotomy pneumonia was 22% [9]. Small cell lung cancer patients, and recently some non-small cell lung cancer patients, have disease that can be responsive to chemotherapy and radiation; therapy-related immunosuppression and injury can increase infection risk, particularly in the lung. 2.1 Bacterial infections

Lung cancer patients are predisposed to develop focal lung infections secondary to bronchial obstruction by the tumor. Several older studies documented the simultaneous diagnoses of lung cancer and bacterial pneumonia, lung abscess, or empyema. In a review of 579 hospitalized patients with lung cancer at a Japanese university hospital over 15 years, 139 (24%) developed respiratory infections, most of which were bacterial [lo]. Patients with extensive disease were more likely to develop pulmonary infection than those with cancer at early stages. Twenty-seven percent of the pneumonias were postobstuctive by chest radiograph. Several older studies documented the simultaneous diagnoses of lung cancer and bacterial pneumonia, lung abscess, or empyema. Strang and Simpson [11] reported 70 patients with lung abscesses among 1,930 patients with a lung cancer diagnosis in Great Britain over an 11year period, an incidence of 3.6%. In many cases, partial obstruction by tumor led to postobstructive atelectasis and pneumonia. In a minority of cases, infection occurred secondarily within an area of tumor necrosis. Rarely, when an abscess was found distant to the tumor, aspiration was thought to be its etiology. Patients generally presented with the abrupt onset of cough productive of sputum, fever, and chest pain. A subset presented in a more indolent manner, with weight loss and anorexia as prominent symptoms. Sputum cultures were usually polymicrobial. In contrast to most patients with simple abscesses who improved with penicillin, most patients with cancer and abscess did not show clinical benefit or radiologic improvement after penicillin therapy alone.

Fig~rreI . A 7L-year-old Vietnamese male with squamous cell carcinoma of the soft palate and primary adenocarcinolna of the lung iollowing surgical resection of both tumors, radiation therapy. and chemotherapy with taxol. 5-fluorouracil, and hydroxyurea presented with recurrent fever. chills, and productive cough. The chest x-ray (A) showed a right upper lobe deformity and pleural thickening attributed to postoperative changes following right upper lobectomy. The C T scan (B) showed a large, thick-walled cavitary lesion in the right apex. Thc sputum Gram stain showed many polyn~orphonuclearleukocytes. The sputum culturc grcw moderate Alcoligertes .~ylasoxirlnns.Fever persisted despite ceftazidime and tobramycin. Because of his upbringing in Southeast Asia, the possibility of tuberculosis was considered. The sputum smear was positive for acid-fast bacilli. The sputum culture grew Mycobncreriutn luberc.l~los.is,and he responded to antitubercukous therapy.

Postobstructive pneumonias are thought to develop secondary to partial obstruction of an airway with overgrowth of bacteria distal to the obstruction; however, this may be an overly simplistic. Other factors likely contribute to the risk of pneumonia associated with an endobronchial tumor. The organisms

recovered from lung abscesses secondary to obstructing tunlor are frequently more virulent than those recovered from primary lung abscesses. In a review of 97 lung abscesses, Perlman and associates [12] found that Stpl~ylococcus nzrrezis and gram-negative enteric organisms were recovered from patients with underlying lung cancer more often than those without lung cancer. Cultures from primary abscesses (noncancer patients) were more likely to reflect normal upper respiratory flora, especially alpha-hemolytic streptococci. This shift to more virulent organisms in the Iung abscesses of cancer patients likely results from aspiration of altered oropharyngeal flora. Oropharyngeal colonization changes during illness, probably due to alterations in epithelial cell surface receptors, resulting in increased proliferation of aerobic gram-negative rods. Empyema occasionally complicates postobstructive pneumonia. In a review of 105 cases of empyema, only 7 (6.7%) were associated with postobstructive pneumonia secondary to bronchogenic carcinoma [13]. Kohno and

others [lo] reported only 2 empyemas among 148 episodes of pulmonary infection in hospitalized patients with lung cancer. Empiric therapy for postobstructive pneumonia or abscess should emphasize coverage of anaerobes, Staphylococcus aureus, and aerobic gram-negative bacilli. A variety of antibiotic regimens may be appropriate and should be guided by the results of sputum gram stains (and later cultures), previous antibiotic exposure (especially recent), and knowledge of local (community and institutional) antibiotic susceptibility patterns. Usual lung abscess treatment (e.g., clindamycin) may be adequate if cultures fail to identify aerobic gram-negative bacilli. Prolonged therapy may be required when bronchial obstruction prevents adequate drainage of the infected lung. 2.2 Mycobacterial infection The frequency of mycobacterial disease may be increased in patients with cancer. In a retrospective review at M. D. Anderson in Houston, Texas, the incidence of mycobacterial disease among their cancer patients was 65 cases per 100,000 persons, in comparison with 45 cases per 100,000 among Texans age 45-65 years old [14]. Kaplan et al. [15] reviewed 201 cases of tuberculosis that developed in cancer patients at Sloan-Kettering Cancer Center over 20 years. Lung cancer patients had the highest prevalence, 920 per 100,000, among solid tumor patients, which was second only to Hodgkins' disease overall. Lung cancer and head and neck cancer patients were more likely to present with tuberculosis at the time of cancer diagnosis; patients with the other neoplasms were more likely to develop tuberculosis while receiving cancer therapy [IS]. Historically, two mechanisms of tuberculosis reactivation in lung tumor patients have been invoked: first, tumor can break down granulomas harboring sequestered mycobacteria, or, second, malignancy-associated cachexia may impair cell-mediated immunity, resulting in reactivation [16]. As therapy for lung cancer has become more aggressive, chemotherapy-related immunosuppression may contribute to reactivation (Figure 1). In a population with high baseline rates of tuberculosis, autopsy data revealed that corticosteroids plus antineoplastic agents increased the incidence of mycobacterial infection compared with antineoplastic agents alone [17]. Of 304 lung cancer patients who came to autopsy, four died of tuberculosis. Each of the four patients with fatal tuberculosis had abnormalities on prechemotherapy chest x-rays consistent with old tuberculosis, suggesting that reactivation had occurred following chemotherapy. Three of four cases occurred within 3 weeks of institution of corticosteroid therapy. A high index of suspicion for tuberculosis in a cancer patient is indicated if the patient's history or epidemiological background suggests prior exposure or if unexplained or rapidly progressive pulmonary symptoms, signs, or chest xray findings develop. At diagnosis of solid tumor disease, we recommend that a tuberculin skin test be placed. Regardless of tuberculin skin test status, at

lung cancer diagnosis, sputum samples and lung biopsy samples should be sent for acid-fast bacillus (AFB) stain and culture. If the tuberculin skin test is positive and there is no evidence of active tuberculosis, prophylaxis with daily isoniazid for 6-12 months is indicated. If a patient has been previously adequately treated for a positive tuberculin skin test, a repeat course of isoniazid is not recommended. Whenever acid-fast bacilli are identified on smears or histopathology, or when mycobacteria are identified in respiratory cultures, therapy for presumed active pulmonary tuberculosis is indicated. Some of these smears or cultures will prove to represent contamination of specimens by nonpathogenic mycobacteria (e.g., Mycobacterium gordonae) and therapy can be discontinued. Some others may prove to represent true infection caused by atypical organisms (e.g., Mycobacterium kansasii or Mycobacterium aviumintracellulare) and therapy can be altered appropriately.

2.3 Fungal infection Individual case reports of relatively immunocompetent lung cancer patients with focally invasive Aspergillus infection are scattered throughout the literature. In these cases, necrotic tumor itself serves as the substrate in which Aspergillus germinates, colonizes, andlor invades. Saprophytic colonization nearby or within the tumor appears to be the most common presentation in relatively immunocompetent lung cancer patients. Smith and Bveneck [18] noted that these focal Aspergillus infections rarely cause life-threatening hemorrhage and uncommonly form fungal balls, in contrast to post-tuberculous aspergillomas. Symptoms from a growing tumor may result in an earlier recognition of Aspergill~lsinfection than post-tuberculous cases, before complications associated with more long-standing infection can develop. As a result, aspergillomas forming in the presence of lung cancers have rarely been described. In one case, misdiagnosis contributed to development of a fungal ball over several months. A 61-year-old male presented with hemoptysis and a multiloculated cystic lung lesion; over the next several months, while the patient was treated empirically for tuberculosis, a fungal ball developed within the cystic cavity [19]. At lobectomy, a mass of Aspergillus f~lmigatuswas found within a previously undiagnosed necrotic, cavitating adenocarcinoma. No evidence of tuberculosis was identified. Only rarely is focal fungal disease detected at the site of the tumor prior to any cancer therapy. Most Aspergillus infections in solid tumor patients are much more aggressive, developing in the setting of intensive immunosuppression associated with chemotherapy andlor radiation therapy. Aspergill~ispneumonia has become increasingly common in a subset of solid tumor patients. This increase appears to be associated with intensification of chemotherapy and prolonged neutropenia. At Memorial SloanKettering Cancer Center, a retrospective study noted twice as many Aspergillus infections during 1969-1970 as during 1964-1965 [20]. Of the 93 collected cases of Aspergillus infection in cancer patients, 14 involved solid

Figure 2. A 62-year-old woman with recurrent breast cancer following surgery, radiation, and chemotherapy was admitted to the hospital with several days of malaise, myalgias. and blurred vision. She was febrile to 39.4"C and pus could be expressed from her percutaneously inserted central venous catheter (PICC) site. Conjunctival hemorrhages (A) and purpuric skin lesions (B) were evident. Retinal exam also identified multiple chorioretinal abscesses. PICC line site, catheter tip, and blood cultures were positive for rnethicillin-resistant Sftrphylococclrs aureus. A transesophageal echocardiogra~nshowed no evidence of cardiac valvular vegetations. Nevertheless, a presumptive diagnosis of endocarditis was made, and she responded to 6 weeks of intravenous vancomycin.

tumor patients. Like the affected leukemia and lymphoma patients, the solid tumor patients who developed invasive Aspergillus were more likely to have leukopenia or a history of recent chemotherapy or corticosteroid therapy. A common presentation was the abrupt onset of unremitting fever and pulmonary infiltrates that did not respond to broad-spectrum antibacterial therapy.

The lung was the most common organ involved with Aspergillus infection, and bronchopneumonia and hemorrhagic infarction were the most common manifestations. The role of corticosteroids, whether exogenous or endogenous, in the development of invasive Aspergillus is illustrated by the following cases. Borkin et al. [21] reported a case of a 53-year-old male with history of adenocarcinoma of the left lung, following resection, who presented with brain metastasis. He was place on dexamethasone; 5 weeks later he developed fever and right chest pain while hospitalized for brain irradiation. Chest x-ray showed a dense right lower lobe infiltrate. He developed respiratory distress over 2 days and subsequently died after massive hemoptysis. Sputum cultures among other pathogens. Autopsy revealed grew Aspergillus f~~mignt~ls, necrotizing pneumonia of the right lung; microscopically, vascular invasion by fungal hyphae was seen. No evidence of malignancy was found in the lung. Smith et al. [22] reported a case of a 47-year-old male who presented with a 3-month illness associated with a 19-kg weight loss and a 2-week history of cough. He was found to be grossly cushingoid in appearance. A chest x-ray demonstrated a right hilar mass, a focal infiltrate, and lymphadenopathy. Hemoptysis prompted a transbronchial biopsy, which revealed invasive pulmonary aspergillosis. Small cell carcinoma was found on bone marrow examination. The Cushing's syndrome was attributed to ectopic hormone secretion by tumor. Following his death on the ninth day of hospitalization, autopsy showed widespread fungal abscesses. Animal models demonstrated the impact of corticosteroids on clearance of an aerosol challenge of AspergilIus spores [23]. The macrophages of untreated control mice effectively phagocytized the spores and the animals remained healthy. The macrophages of mice receiving corticosteroids failed to effectively phagocytize spores. The Aspergill~~s spores germinated and produced invasive hyphae; hemorrhagic bronchopneumonia developed and the majority of animals died. Sputum cultures that demonstrate AspergiEEus species in an immunocompetent patient with deteriorating pulmonary status or new infiltrates should prompt a more thorough investigation for evidence of invasive disease. In severely immunosuppressed individuals with deteriorating pulmonary status or new infiltrates, recovery of AspergiEE~.lsfrom the respiratory tract should prompt empiric antifungal therapy. If possible, immunosuppressive therapy should be discontinued. Amphotericin B is the usual therapy for invasive aspergillosis, but itraconazole has also proven effective and may be a reasonable compromise when the suspicion of invasive disease is not high or the risk of amphotericin B-related toxicity is considered substantial. Other fungal infections have been noted rarely in patients with lung cancer, but no clear association has been made. These have included blastomycosis [24], candidal infections, and cryptococcosis. In a survey of 170 Veterans Affairs hospitals over 12 years, 198 cases of blastomycosis were found; only 3 had underlying bronchogenic carcinoma and 2 had metastatic lung

lesions [25]. Isolation of Cnndida species from the lung has proved to be airway colonization in most cases. Thirty-one cases of Candida pneumonia, however, were documented by autopsy over a 20-year period at the M. D. Anderson Cancer Center [26]. Sixteen (52%) of these cases had underlying solid tumors; the remainder had hematologic malignancies. Associated with the development of Candida pneumonia were broad-spectrum antibiotics (28 patients), corticosteroid therapy (15 patients), and neutropenia (9 patients). 2.4 Pneumocystis carinii pneumonia

The risk of acquiring Pneumocystis carinii pneumonia (PCP) in non-AIDS patients appears to be highest in patients with hematologic malignancies and those with severe T-cell depression. PCP is a rare but recognized risk in immunosuppressed solid tumor patients. Yale and Limper [27], in a review of 116 non-AIDS patients with PCP at the Mayo Clinic, reported that 13% had underlying solid tumors, 30% had hematologic malignancies, 25% had received organ transplant~,and 22% had inflammatory disorders. The underlying solid tumors included brain tumors, lung carcinoma, breast carcinoma, colon carcinoma, renal cell carcinoma, and melanoma. Accumulated evidence from human and animal studies documents that chronic corticosteroid administration is a significant risk factor for PCP. In the aforementioned [27], 91% of non-AIDS patients with PCP had recently received several weeks of systemic corticosteroids. The median steroid dose associated with PCP in solid tumor patients was equivalent to 30mg of prednisone per day (range, 6.6-240mg per day); the median duration of corticosteroid use was 12 weeks (range, 4-14 weeks). A review of 142 PCP cases in non-AIDS cancer patients at Memorial Sloan-Kettering Cancer Center reported that 31% had underlying solid tumors and 67% had hematologic malignancies [28]. Eighty-seven percent of these patients had been on corticosteroids within 3 months of PCP diagnosis. A small, retrospective study among patients with Cushing's syndrome and opportunistic infections identified the highest morning cortisol levels in those with PCP [29]. The animal model of PCP is based on the fact that rats routinely develop progressive PCP when treated with corticosteroids [30,31]. In some patients, the development of symptomatic PCP may be associated with tapering of chronic steroids. Henson et al. [32]reported 10 patients with primary brain tumors complicated by PCP. All of these patients were receiving dexamethasone for greater than or equal to 1.5 months at the time of PCP diagnosis. Eight of the 11 cases presented during steroid taper. Similarly, when Poplin and colleagues [33] reported two cases of PCP in solid tumor patients; a patient with metastatic prostate carcinoma developed PCP while tapering his oral steroids. The mechanism by which chronic steroid use predisposes to PCP is unclear. Chronic corticosteroid therapy causes CD4 cell depletion and impaired macrophage function, which may allow the development of P. cnrinii

infection while limiting the host inflammatory response. Withdrawal of steroids may remove the antiinflammatory effects before the immunosuppressive effects have resolved, leading to clinical exacerbation. In general, the presentation of PCP in non-AIDS patients is more acute than in AIDS patients. Kovacs et al. [34] found that prior to presentation, non-AIDS patients had a median duration of pulmonary symptoms of 5 days (range, 1-42 days), whereas AIDS patients had a median duration of 28 days (range, 1-270 days). Non-AIDS patients were more likely to have fever and severe hypoxemia, and showed a wider range of respiratory rates. In the report by Henson et al. [32],brain tumor patients with PCP had pulmonary symptoms a median of 7.4 days before admission (range, 1-30 days). Of 10 patients, 8 had dyspnea and 6 had fever. Chest x-rays upon admission varied from normal (I), to an isolated focal infiltrate (21, to diffuse or bilateral infiltrates (7). Lactate dehydrogenase levels were elevated, ranging from 336 to 1284UlL (median 510UlL). Non-AIDS patients with PCP, and specifically those with solid tumors, have a 10-fold lower organism load [35]. While the low organism load may reduce the diagnostic yield of bronchoscopy, it does not appear to reduce the severity of illness. In the series reported by Yale and Limper [27], 7 of 15 patients with PCP and underlying solid tumors developed respiratory failure and died. In Kovac et al. [34], the survival of AIDS and non-AIDS patients presenting with PCP was not significantly different (57% and 41%, respectively). Solid tumor patients with severe cell-mediated immunosuppression, for example, those who are severely malnourished, bone marrow transplant recipients, and those who receive chronic corticosteroids, are candidates for PCP prophylaxis (trimethoprim-sulfamethoxazole, dapsone, or monthly aerosolized pentamidine). Individuals receiving chronic steroid therapy, especially high-dose therapy, should continue on PCP prophylaxis during steroid tapering [36].

2.5 Viral infection Viral pneumonia is a rare cause of infectious complications in solid tumor patients. Camazine et al. [37] reported three cases of herpes simplex virus (HSV) type 1 pneumonia occurring in patients who had undergone recent thoracotomy for carcinoma involving the lung. A 52-year-old female presented with fever and hypoxemia on postoperative day 2 following thoracotomy for pulmonary metastasis of rectal carcinoma, a 72-year-old male developed fever and hypoxemia on postoperative day 3 after thoracotomy for squamous cell carcinoma of the lung, and a 72-year old male with mesothelioma presented with fever, hypoxemia, and respiratory failure on postoperative day 4. Two of the three patients developed diffuse interstitial infiltrates; one developed a focal, progressive infiltrate. Respiratory cultures were positive for herpes simplex virus (HSV) and, in one case, cytomegalovirus (CMV) as well; no other pathogens were isolated. Bronchial washings

Figzrre3. A 49-year-old woman with recurrent squamous cell cascino~naof the cervix, treated with radiation therapy and multiple wide excisions, presented with persistent pelvic pain. On admission, she was febrile and had a WBC coutlt of 5300/mm3and an ESR ol119. CT scan of the pelvis showed a fluid collection within the pelvis with an enhancing rim of soft tissue. Inlraoperative cultures grew many beta-hemolytic streptococci. group F. Bone scan and MRI showed evidence of osteomyelitis of the left ileum.

revealed intranuclear inclusions consistent with HSV infection. All three patients responded promptly to acyclovir. The authors recommend delaying cardiothoracic surgery in the presence of perioral HSV lesions. Significant immunosuppression, similar to that experienced by organ transplant patients and bone marrow transplant patients, appears to be necessary to develop pneumonia secondary to CMV, respiratory syncytial virus (RSV), parainfluenza virus, and adenovirus. It is likely that solid tumor patients undergoing autologous bone marrow transplantation are at risk for these infections. 2.6 Noninfectio~cscauses of pulr?zonary bzfiltrrrtes in solid tlinzor patients

Radiation pneumonitis typically presents 3-4 months after irradiation with insidious onset of nonproductive cough, fever, and shortness of breath [38]. Symptoms frequently become apparent when corticosteroids are tapered [39]. Physical findings are uncommon, but include pulmonary consolidation, pleural

friction rub, or pleural effusion [39]. When patients become symptomatic, radiologic changes are evident and are almost always precisely limited by the edges of the radiation field. During the period 2-4 months after irradiation, many more individuals develop abnormalities on chest x-ray than develop symptoms. Ground glass opacification or haze is common in the early stages [39], possibly followed by dense infiltrates. Because radiation pneumonitis can respond dramatically to steroids, diagnosis is clinically important. Some chemotherapeutic agents, such as actinomycin D and adriamycin, have been associated with reactivation of radiation pneumonitis [40]. Other noninfectious causes of pulmonary infiltrates in patients with lung cancer include congestive heart failure, pulmonary emboli, pulmonary hemorrhage, adult respiratory distress syndrome, drug toxicity, chemical aspiration, and progression of tumor.

2.7 Evnluntion The urgency with which diagnosis of pulmonary infiltrates in a lung cancer patient must be made depends upon the level of immunosuppression. A lung cancer patient who is severely malnourished, neutropenic, or receiving highdose corticosteroid therapy who develops pulmonary infiltrates should undergo an urgent diagnostic procedure. In addition, empiric broad-spectrum antimicrobial therapy should be considered. If a patient is relatively immunocompetent, on the other hand, a more methodical approach can be undertaken, although invasive procedures are often indicated. The relatively immunocompetent patient with lung cancer is more likely to develop focal infiltrates. These patients may benefit from a CT scan of the chest with intravenous contrast to better define the extent of disease and facilitate obtaining a biopsy. Sputum for gram stain, culture, AFB smear, mycobacterial culture, and fungal culture should be obtained. A tuberculin skin test should be placed. Bronchoscopy with bronchoalveolar lavage (BAL), transbronchial biopsy (TBB), or percutaneous transthoracic needle biopsy (if the lesion is located perphierally) are options for rapid diagnosis. If tissue is obtained, aerobic, anaerobic, fungal, and rnycobacterial studies, as well as cytology, are indicated. Because fungi can colonize the airways of relatively immunocornpetent individuals, biopsy to detect tissue invasion is often indicated. Lung cancer patients who develop diffuse pulmonary infiltrates typically occur in those with severe immunosuppression. Diffuse pulmonary infiltrates are more likely to be secondary to infection when the patient has received chemotherapy, when fever develops with the radiographic changes, and when the radiographic changes occur rapidly [41]. Nevertheless, clinical presentation is not predictive enough to direct therapy with confidence. In the setting of diffuse pulmonary infiltrates and immunosuppression, an invasive diagnostic procedure (BAL, TBB, or open lung biopsy) should be considered early. Meanwhile, specimens of blood and sputum (if accessible) should be sent for

bacterial, fungal, and mycobacterial culture. Because viruses can cause diffuse pulmonary infiltrates in the setting of severe immunosuppression, a nasopharyngeal or oropharyngeal swab may be sent for viral culture and, if available, direct staining for influenza or RSV. Skin should be inspected daily in the immunocompromised patient, and new skin lesions should be biopsied promptly. Bronchoscopy and BAL is often utilized as the first-line diagnostic procedure in the immunosuppressed patient with new diffuse pulmonary infiltrates. Specific studies of solid tumor patients have not been performed, so recommendations must be derived from reports in which leukemia and lymphoma patients predominate 1421. When the patient is severely immunosuppressed, specimens should be sent for gram stain and quantitative culture, acid-fast stain and mycobacterial culture, fungal stain and culture, Pneumocystis stains, Legionella direct fluorescent antibody (DFA) staining and culture, viral culture and DFA or in situ hybridization for HSV and CMV, possibly cultures for Chlamydia pneumoniae and Mycoplasma, and DFA and culture for respiratory viruses (e.g., influenza virus, parainfluenza virus, and RSV). Because BAL generally dilutes lower respiratory tract secretions by a factor of 1:10 to 1:100, the diagnostic threshold for bacterial culture is 1 0 4 C F U / m ~However, . some organisms, including Mycobacterium tuberculosis, Legionella species, and Nocardia species, should be considered pathogens whenever isolated. A sample should also be sent for cell count and differential, and for cytopathology. The cell count and differential may be used to assess specimen adequacy. For example, squamous and bronchial epithelial cells accounting for >1% of the total cells suggests contamination by oropharyngeal flora. Detection of plentiful hemosiderin-laden macrophages by direct microscopy suggests pulmonary hemorrhage. BAL appears to have a good yield in the diagnosis of PCP, tuberculosis, CMV, and aspergillosis [43]. TBB is more invasive than BAL alone, but improves the diagnostic yield in some infectious and noninfectious processes. However, TBB is associated with an increased risk of pneumothorax and hemorrhage. Open lung biopsy (OLB) is associated with greater morbidity than either BAL or TBB, but may be considered when BAL and TBB fail to provide a diagnosis or are contraindicated because of respiratory instability or bleeding risk.

3. Infectious complications of breast cancer Breast cancer accounts for approximately one third of the new cancer cases diagnosed annually in females in the United States [I]. Almost all of these patients undergo some surgical procedure of the involved breast and ipsilatera1 lymph nodes. Most infectious complications in breast cancer patients involve skin and soft tissue, which has been altered by surgery and/or irradia-

tion. These infections occur both immediately postoperatively and very late, up to several years after cancer therapy. Increasingly, patients with a wide range of disease stages are now receiving some combination of chemotherapy, radiation, and surgery. Limited information is available regarding the impact of combined-modality therapy on infection rates, although chemotherapy and radiation can theoretically delay wound healing and impede wound drainage. The subset of patients with advanced breast cancer or metastatic disease who undergo bone marrow transplantation or other intensive chemotherapy protocols are at risk of additional complications, similar to those that leukemia and lymphoma patients experience (Figure 2). 3.1 Postoperative infectio~lscomplications of breast cancer patients Despite the fact that most breast cancer surgery is classified as clean, acute wound infection rates in the range of five times the rate of other clean surgeries has been reported [44]. Disruption of the skin integrity and contamination by skin flora accounts for most wound infections within the first 6 weeks after breast surgery [44]. By far the most common organisms involved in these wound infections are streptococcal species, Staphylococcus nureus, and coagulase-negative staphylococci [44,45]. A recent surveillance study of surgical wound infections revealed that variable wound infection rates appear to be dependent on the type of breast cancer procedure performed [44]. More extensive surgery resulted in higher wound infection rates: simple breast biopsies had a rate of 2.3%, lumpectomy with lymph node dissection had a wound infection rate of 6.6%, and mastectomies had a rate of 19% (P < .05). In addition, variation in surgical technique, such as drain type (e.g., closed suction, Jackson-Pratt) and placement (e.g., new skin incision separate from the cancer incision vs. through the wound), also influenced infection rates [44]. Alternatively, other investigators have reported similar wound infection rates for modified radical mastectomy and lumpectomy. For example, Vinton et al. [46] reviewed 387 modified radical mastectomies and 173 lumpectomies between 1983 and 1989. The wound infection rates were not significantly different for modified radical mastectomy (15%) and lumpectomy (13%). The impact of the shift toward more conservative surgery on breast cancer wound infection rates is unclear.

3.2 Late-onset infectious conzplications of breast cancer

A subset of breast cancer patients experience skin and soft tissue infections months to years after therapy concludes. Historically, the most common late infection in postmastectomy patients was upper extremity cellulitis ipsilateral to the breast cancer surgery and lymph node dissection. A classic presentation is sudden onset of a painful and erythematous rash, which spreads rapidly on the involved upper extremity, with or without fever. Blood cultures, positive in only a minority of cases, may yield skin flora. Many, but not all, of the patients

who experience one or more episodes of cellulitis had obvious pre-existing chronic swelling and lymphedema of the ipsilateral upper extremity. Lymphedema was a common complication after radical and, less frequently, modified radical mastectomy. The incidence of chronic lymphedema following these surgeries has been estimated at 15% [47]. Risk factors for the development of lymphedema included the following: a greater number of lymph nodes removed, and delayed postoperative wound healing, cellulitis, radiodermatitis, hematoma, seroma formation, and skin flap necrosis [47]. The rates of upper extremity cellulitis following lumpectomy, radiation therapy, and chemotherapy are unavailable, but appear to be lower than those following radical or modified radical mastectomy. Clinical lymphedema appears to be a marker for increased risk of cellulitis, but subclinical abnormalities of lymphatic drainage can also predispose to infection. Bertelli et al. [48] reported that 7 of 21 patients with ipsilateral upper extremity cellulitis following breast cancer surgery did not have clinically detectable lymphedema. Breast cancer patients who experience a single episode of cellulitis appear to be at risk of recurrence; in one series, 11 of 15 patients with cellulitis after breast cancer surgery had more than one episode [49]. A breast cancer patient, especially one who experiences difficulties with wound healing postoperatively, should be considered at life-long risk of developing upper extremity cellulitis. Osteomyelitis and septic arthritis of the ipsilateral shoulder can present years after breast cancer therapy; this complication is apparently also linked to preexisting lymphedema. Chaudhuri et al. [50] described five cases in which ipsilateral humeral osteomyelitis and septic arthritis presented 2-12 years after breast cancer surgery. All five patients had received a radical or modified radical mastectomy followed by radiation therapy. A striking characteristic of these cases was indolent presentation - ipsilateral shoulder pain and restricted movement for 4 or more months, without fever. Four patients had an erythrocyte sedimentation rate (ESR) of >100mm/h. Radiologic studies confirmed the diagnosis of osteomyelitis: all five patients had positive bone scans, three of the five patients had findings of osteomyelitis on plain radiographs, and none had findings consistent with radiation necrosis of adjacent bones within the radiation field. It is unclear whether these serious late complications, osteomyelitis and septic arthritis, will occur following lumpectomy. Despite the shift in management of breast cancer to more conservative surgery, skin and soft tissue infections continue to be the major delayed complication of breast cancer. Instead of ipsilateral arm cellulitis, focal cellulitis of the involved breast has been reported following lumpectomy with axillary lymph node dissection. Rescigno and coworkers [51] documented 20 episodes of breast cellulitis in 11 patients. The authors estimated the incidence of this complication among patients after lumpectomy to be 2.5-3.0%. Time from completing radiation to first episode of cellulitis ranged from 9 days to 4 years (median, 4.3 months). Each of these patients presented acutely with rapidly spreading erythema, warmth, and tenderness of the breast. The origi-

Figirre 4. A 52-year-old man with squamous cell carcinoma of the tongue was treated with surgery, radiation, and administration of cisplatin. 5-fluorouracil. and hydroxyurea. H e p1,esented t o the hospital with fever and drainage from the site of a lormer lcft subclavian Port-A-Cath. C T scan of thc chest showed air or gas collections in the left chest wall and anterior mediastinum. and a soft tissue mass in the left superior rncdiastinum. At surgery. infection was identified to have tracked from the port site along the catheter tract into the metliastinuni. Intraoperative cultures grew Klcbsiell~ptzertri?oriioe, alpha-hemolytic streptococci, and Eikenellrr c o r r o r l e ~ ~ . ~ .

nal site of erythema was often removed from the surgicaI scar. In eight cases, erythema spread beyond the breast tissue to the back, shoulder, or arm. Fever or breast swelling was present in some, but not all, patients. Six cases required hospitalization. Two patients experienced repeated episodes of chronic recurrent celIulitis. Episodes of acute and recurrent breast cellulitis responded to antibiotics directed against gram-positive cocci (e.g., oxacillin, cefazolin). Finally, four patients experienced chronic persistent cellulitis, which did not fully resolve despite prolonged antibiotic therapy. This apparent noninfectious erythema may be to radiation-induced inflammation [52].This latter condition is treated with corticosteroids. Late development of breast abscesses may be unique to the breast cancer patients who received lumpectomy and radiation. Keidan et al. [53] reviewed

112 lumpectomies, finding 7 breast abscesses that developed 1.5-8 months following surgery. This 6% abscess incidence was higher than expected for clean surgery. Aspirated fluid grew Staphylococcus aureus in 3 cases and S. epidermidis in 3 cases. Postsurgical manipulation, such as previous seroma aspiration, was a risk factor for abscess development. Notably, each of the seven patients with breast abscess had received local irradiation. Irradiation may result in impaired lymphatic drainage and local ischemia, which may increase the risk of infection. Axillary node dissection likely contributes to poor lymphatic drainage.

3.3 Infectious complications in breast cancer patients undergoing bone marrow transplantation Breast cancer patients who receive high-dose chemotherapy and autologous bone marrow transplant are at risk for complications similar to those of leukemia and lymphoma patients. The risk of infection is related to the duration of severe neutropenia, and clinical manifestations are similar to other patients with chemotherapy-related neutropenia. Mudad et al. [54] reported that 12 of 206 patients with breast cancer following autologous bone marrow transplantation developed CT scan findings consistent with hepatosplenic candidiasis 6-12 weeks after transplantation. Only 4 of the 12 breast cancer patients were febrile when hepatosplenic candidiasis was diagnosed by CT scan. Only three patients had positive blood cultures for Candida. As with other patients, the most common laboratory abnormality was an increased alkaline phosphatase. A subset of patients underwent purging of their bone marrow complicated by prolonged neutropenia (median 24 days vs. 12 days in nonpurged bone marrow patients). Seven (22%) of the 32 patients on the purge protocol developed hepatosplenic candidiasis. In contrast, only 5 (2.8%) of 174 patients receiving nonpurged marrow developed hepatosplenic candidiasis. Outcome may be better for the breast cancer patient than patients with hematologic malignancies. Two patients received no antifungal therapy, yet survived. The remaining patients were treated with fluconazole or amphotericin B with or without Aucytosine. No deaths were attributed to the fungal infections, although five died from cancer progression and one died from drug toxicity.

3.4 Management of infections in breast cancer patients New onset of erythema in the breast tissue or ipsilateral upper extremity in a patient with breast cancer is most likely cellulitis. However, noninfectious causes of skin erythema should also be considered, especially when erythema occurs within the radiation field and when it persists after a trial of antibiotic therapy. Radiation changes alone can cause chronic, sometimes progressive, inflammation. In addition, "radiation recall" of the chest has been reported when patients receive chemotherapy agents long after their radiation course

[55].Differential diagnosis of a focal breast abscess includes fat necrosis, which often presents as a tender hard mass with or without appreciable inflammation [56]. If fluid is not detected, biopsy may be necessary to differentiate tumor recurrence, fat necrosis, and abscess. Stewart-Treves syndrome, a very rare type of angiosarcoma that occurs exclusively in the ipsilateral arm of patients following mastectomy, develops subacutely in the setting of chronic lymphedema; skin biopsy should be considered if swelling and erythema do not resolve with a seemingly appropriate course of antibiotic therapy. The underlying pathophysiology of most infectious complications of breast cancer patients is damaged local host defense secondary to surgery and radiation therapy, namely, an impaired skin barrier, injured microcirculation, and impaired lymphatic drainage. Patients usually present with acute onset of erythema, pain, possibly with fever and leukocytosis. The infections tend to be local, involving skin, underlying dermis, and possibly fat. Because the infections tend to be focal, blood cultures are rarely positive, but should be performed. Tissue samples for culture are also rarely positive; unless there is fluctuance, tissue aspiration is not recommeded, so as to avoid further damage to the skin barrier. When cultures are positive, isolated pathogens are almost always gram positive; thus, in most cases anti-staphylococcal coverage is adequate. The patient should be treated with anti-staphylococcal coverage, such as nafcillin, cefazolin, clindamycin, or vancomycin. Oral antibiotics, such as cephalexin and dicloxacillin, may be used to complete a course once swelling and erythema have been reduced. Any patient who has undergone lymph node dissection should be educated about protecting the ipsilateral arm from trauma, pressure, or damage. Patients are advised to avoid phlebotomy, intravenous catheters, or blood pressure monitoring in the upper extremity ipsilateral to previous breast cancer surgery [57]. The use of compression garments and elevation of the affected arm can reduce lymphederna. Early aggressive therapy of cutaneous fungal infections can minimize local skin breakdown [58]. Those who experience poor wound healing initially, develop clinical lymphedema, or experience an episode of cellulitis should take measures to prevent repeated episodes. Selected patients with recurrent cellulitis may benefit from chronic suppressive antibiotic therapy. 4. Infectious complications of abdominal and pelvic cancer Malignant tumors do not honor mucosal barriers. As a consequence, infectious complications of solid tumors in the abdomen and pelvis appear to occur mainly secondary to mucosal invasion of tumor, with subsequent local abscess formation or dissemination. The gastrointestinal tract serves as the dominant source of bacteria within the abdomen and pelvis. When the gastrointestinal tract is intact, a sterile peritoneum lays within millimeters of colonic luminal contents bearing a bacterial density of 10" per gram of dry weight [59].

Similarly, an intact ileocecal valve and proper peristalsis separate the mainly aerobic small bowel luminal contents, bearing only 10' to lo6 bacteria per milliliter, from colonic contents, most of which is anaerobic [59]. Tumor invasion of gastrointestinal structures causes contamination of previously sterile structures, spaces, or fluids. As a result, the infections that result from cancers in the abdomen and pelvis are often polymicrobial. In a 10-year review of polymicrobial septicemia in cancer patients, Elting and associates [2] found three types of underlying malignancies most frequently: hematologic malignancy in 47%, genitourinary cancers in 16%, and gastrointestinal cancers in 13%. In addition, bacterial flora within the gastrointestinal lumen may shift toward more virulent organisms secondary to chemotherapy, irradiation, or in association with the malignancy itself. 4.1 Streptococcus bovis colonization and endocarditis

The relationship of Streptococcus bovis bacteremia and colon carcinoma was uncovered during the 1970s. Klein et al. [60] reported two patients with adenocarcinoma of the colon who developed S. bovis endocarditis. They showed that fecal carriage of S. bovis in patients with colon carcinoma was significantly greater (56%) than in healthy controls (lo%), patients with nongastrointestinal carcinomas, and patients with other gastrointestinal disorders. The authors recommended evaluation for colon carcinoma in all cases of S. bovis endocarditis. Subsequently, these same investigators prospectively evaluated 29 patients with 30 episodes of S. bovis endocarditis [61]. Twelve patients were found to have colonic neoplasms, 3 of which were malignant; 3 other patients had undiagnosed colonic masses and 10 had diverticulosis. Of note, the majority of the patients with colonic neoplasms and endocarditis did not have neoplasms that invaded the muscularis mucosa: the presence of a friable intraluminal mass or diverticuli, which also bleed easily, may allow bacteria access to the bloodstream. Following the reported association of S. bovis and colon carcinoma, there have been case reports of patients with endocarditis involving other streptococci (e.g., S. snlzguis, S. agnlncfiae) who were found to have colonic neoplasms [62,63] and gastric carcinoma [64]. The diagnosis of Streptococc~isbovis bacteremia and endocarditis is based on recovery from blood cultures. Once a blood culture isolate is identified as S. bovis, it is necessary to establish whether endocarditis is likely by performing serial blood cultures and echocardiography. In addition, the lower gastrointestinal tract should be evaluated for the presence of disease. Colonoscopy or a barium enema may be performed; if a barium enema is negative, however, a colonoscopy is indicated. 4.2 Clostridium septicum myonecrosis and bacterenzia

An association between malignancy and Clostridi~imsepticutn myonecrosis has been noted. Particularly vulnerable to this acute, life-threatening bacterial

infection are those with leukemia or occult colon or rectal carcinoma [65-671. In a literature review covering 42 years, Kornbluth et al. [67] identified 162 cases of C. septicum infection; 34% had colorectal carcinoma and 47% had hematologic malignancies. Occult malignancies were found in 37%. C. septicur?~,a sporulating gram-positive, toxin-producing rod, is an uncommon pathogen in humans. In a review of 114 cases of clostridial infection by Gorbach and Thadepalli [68], only 3 were caused by C. septicurn. In a similar review, Kornbluth and colleagues [67] review four studies of clostridial infection: of a total of 612 clinical isolates of CIostriditirn, only 6 (1.3%) were C. septicurn. The relative risk of C. septicurn infections in cancer patients is unknown, but an increased risk of Clostridiurn septiciirn infections among patients with colorectal cancers seems apparent. Typically, a patient with C. septicurn myonecrosis presents with a history of acute onset of severe focal pain, usually of an extremity, fever, and a toxic appearance. The painful site may initially look normal, but within hours the involved skin becomes discolored and edematous, bullae form, and the discolored area enlarges rapidly [67]. As a late finding, the subcutaneous tissue becomes crepitant [69]. Some patients present with diffuse abdominal pain as the most prominent initial symptom [67]. Patients progress rapidly from a toxic presentation to shock and ultimately death, often within 48 hours. The site of myonecrosis, such as the shoulders, limbs, or chest wall, is usually is distant from the carcinoma; therefore, hematogenous spread of Clostridi~~m presumed. The most common sites of underlying adenocarcinoma in patients with C. septicum are the cecum and distal ileum. Tumor invasion into the mucosa is thought to supply access of the organism to the blood stream. In a review of the published literature, Kaiser et al. [70] described 23 cases of distant clostridial myonecrosis, 12 of which had underlying colon or rectal carcinoma. C. perfringens was isolated in approximately half of these cases, causing a syndrome indistinguishable from that caused by C. septicurn. The 12 patients with underlying colorectal cancer had mucosal breakdown at the ileum, colon, or rectum documented at surgery or autopsy. Seventeen of the 23 patients died, many within a few hours of admission. Of 59 cases with C. septicurn bacteremia reviewed by Koransky et al. [66], 21 had solid tumors, 14 of which were colon cancers. Of the 28 patients autopsied, a colonic lesion was documented in 17. Seven autopsies demonstrated evidence of "fecal peritonitis from bowel perforation or gangrene." Additional evidence of hematogenous spread comes from case reports of colon cancer patients with C. septicum septic arthritis [71,72], bacteremia, and septic shock [73], and with bacteremia and polymicrobial abscess within a hepatic metastasis [74]. Host factors such as diabetes meIlitus [67,75], granulocytopenia [66], and atherosclerosis [66] may increase the risk of developing C. septicurn infection. Nineteen of 100 patients with C. septicurn infection in the review by Kornbluth et al. [67] had diabetes rnellitus. The diabetic patients were significantly more likely to have an occult malignancy than nondiabetic patients. Kudsk [75] reported five cases

of C. septicurn myonecrosis in diabetic patients, each of whom was found to have an occult malignancy. The cellulitis, bullae formation, and myonecrosis are thought to occur secondary to several locally produced toxins. The diagnosis of C. septicurn myonecrosis should be made rapidly, based on clinical presentation in those with and without known malignancies. Blood cultures should be obtained before initiating antibiotics. Gram stain and culture of percutaneous tissue aspirates and bullae aspirates should be performed emergently [70]; identification of short, plump gram-positive rods suggests clostridial infection in the appropriate clinical setting . Because gas on x-ray may be a late finding, performing such studies should not delay debridement. If the patient presents with abdominal symptoms or signs, exploratory laparoscopy may be considered as well. Patients may require multiple surgeries over several days following presentation. High-dose penicillin G is the traditional drug of choice for C. septicum infection. Clindamycin is favored by some investigators because its mechanism of action may inhibit clostridial toxin production and more effectively halt the progression of established disease. Initial antibiotic management may be broadened to include gram-negative, staphylococcal, and anaerobic coverage, until the diagnosis is confirmed by culture results and the extent of intraabdominal disease is known. The patient should be aggressively treated for sepsis and monitored closely for hemodynamic deterioration. It is unclear whether patients benefit from adjunctive hyperbaric oxygen therapy.

4.3 Pyogenic abscesses Pyogenic liver abscess is another extremely rare entity that can complicate gastrointestinal malignancies. There appear to be two typical presentations for pyogenic liver abscesses in patients with gastrointestinal malignancies. First, abscesses may herald the discovery of a previously undiagnosed, usually advanced, luminal or pancreaticobiliary malignancy. Secondly, pyogenic liver abscesses occur in patients with known malignancies, many of whom have undergone recent gastrointestinal procedures. In a review of 20 cases of pyogenic liver abscess, 5 had underlying gastrointestinal carcinomas [76]. Patients frequently presented with fever of unknown origin. Most pyogenic liver abscesses associated with colon carcinomas were polymicrobial and included anaerobes and enteric gram-negative rods. Organisms presumably spread from areas of mucosal breakdown to the liver via the portal circulation. Pancreaticobiliary malignancy may obstruct the biliary tract, resulting in ascending cholangitis, and then in multiple hepatic abscesses. The development of one or more liver abscesses in a known cancer patient is extremely rare. A review of liver abscess in cancer patients at the National Cancer Institute yielded only 37 patients over 35 years [77].The etiology of these abscesses was bacterial in 17 and fungal in 20. Twelve of the 17 patients

with bacterial liver abscesses had solid tumors; the remaining 5 had hematologic malignancies. Most bacterial abscesses were polymicrobial, with gramnegative and anaerobic organisms recovered on culture. Marcus and associates [77] found that recent gastrointestinal instrumentation was a strong risk factor for liver abscess development. Ten of the 17 patients with bacterial abscesses had undergone either a surgical or radiologic procedure on the gastrointestinal system, such as surgical resection of liver metastases and biliary stent placement. Recent neutropenia did not appear to be a risk factor for development of bacterial liver abscesses; only one such patient had been neutropenic within 60 days of presentation. Hepatic and splenic abscesses have occurred after invasive procedures for hepatocellular carcinoma. Okada et al. [78] reported a case of a 56-year-old woman with hepatocellular carcinoma and distant cholecystoduodenostomy who presented with fever and leukocytosis while hospitalized after her third percutaneous ethanol injection procedure. One of her two injected liver lesions had become more hyperechoic, consistent with gas formation. This lesion was drained percutaneously, yielding Klebsiella pneumoniae. Isobe et al. [79] reported a case of probable splenic abscess in a cirrhotic woman with hepatocellular carcinoma who became febrile with new left upper quadrant pain 1 day after percutaneous ethanol injection. By ultrasound, her spleen had multiple hyperechoic lesions. These lesions resolved by 10 days with antibiotics alone. Whereas breakdown of the mucosal barrier of the gut or pancreaticobiliary system is the basis of pyogenic liver abscess in the undiagnosed colon carcinoma patient, in the known cancer patient undergoing therapy, recent gastrointestinal instrumentation appears to be the primary risk factor. Pyogenic hepatic abscesses should be suspected in a patient with fever, malaise, right upper quadrant pain, andlor jaundice or in a patient with persistent unexplained fever. Liver abscesses may be identified by CT scan or by ultrasound. Blood cultures should be obtained upon admission arid when the patient is febrile. If a patient with a pyogenic abscess has no clear pancreaticobiliary obstruction, a search for a lesion within the gastrointestinal tract is indicated. Treatment of pyogenic liver abscesses is controversial and evolving. Many continue to rely on surgical or percutaneous drainage [SO]. Recently, some patients have done well with prolonged antibiotic therapy alone, following percutaneous aspiration and identification of the infecting organisms and their antimicrobial susceptibilities [Sl]. Empiric broadspectrum antibiotics, including gram-negative and anaerobic coverage, is indicated initially; choices such as piperacillinltazobactam, ticarcillinl clavulanate, imipenem-cilistatin, or meropenem are among the many appropriate initial regimens. Use of an aminoglycoside can often be avoided or used only in the initial days of therapy while awaiting culture results. Antibiotic coverage should be tailored to the results of antimicrobial susceptibility testing. Because polymicrobial infections are common in these infections, retaining

broad coverage agent gram-negative and anaerobic bacteria is usually indicated on the assumption that more organisms may be involved than can readily be identified. Periodic imaging studies can be used to monitor the resolution of the abscesses; up to 4 months of antibiotic therapy may be warranted. 4.4 Salmonella infections

Salmonella species infect humans via the gastrointestinal tract. In the healthy host, Salrnorzella infections are associated with ingestion of a large inoculum that manages to evade the acid barrier of the stomach. Individuals who have added risk of salmonellosis include those with cellular immune dysfunction, the elderly, and those with achlorohydria. An increase in Salmonella infections have been associated with malignancies, especially metastatic disease. In a retrospective study of 95 patients with salmonellosis and neoplastic disease at Memorial Cancer Center [82], there was a clear relationship between Salrrzonella spp. infections and leukemias and lymphomas. The hematologic malignancies accounted for 46 cases; the most common organism isolated from In contrast, nontyphimurium Salmonella these patients was S. fyphirnl~ri~mt. were responsible for most infections among the 40 solid tumor patients. Of solid tumors, gastrointestinal malignancies were most commonly associated with SnIrnonella infection. Intraabdominal malignancy appears to be a risk factor for development of serious Salrr.zonella infections. In a 7-year review of Salr?zonella infections, Han et al. [83] found 4 such infections among 1258 admissions with underlying colon cancer (0.3 %); 3 cases among 2891 admissions with bladder, uterine, or ovarian cancers (0.1%); and 9 cases among 1235 admissions with hematologic malignancies (0.7%). The rate of salrnonellosis for patients admitted with all types of malignancy was approximately 13 times the rate for those without malignancy. Manifestations of infection in patients with intraabdominal cancers varied from gastroenteritis to focal infections to sepsis. Other solid tunloss may rarely be associated with salmonellosis. Two case reports of elderly patients with advanced, undiagnosed lung cancer presenting with Salt~zonellapulmonary infections have been reported [84,85]. Host factors that appear to predispose cancer patients to Snlrqaonella infections include increasing age, impaired cellular immunity, altered gastrointestinal function (e.g., postoperative ileus), acl~lorohydria (secondary to atrophic gastritis or H2 blocker or antacid therapy), chemotherapy, and corticosteroid therapy. Salrr~onella infections can be treated with quinolones, trimethoprimsulfan~ethoxazole,ampiciIlir1, cliloramphenicol, or tetracycline. Antibiotic choice should be tailored to the sensitivities obtained because S~rlrr.zonelln species have high resistance rates. Uncomplicated gastroenteritis in a healthy host is generally not treated with antibiotics. Because of the increased risk of dissemination in solid tumor patients, antibiotic therapy should be considered.

4.5 Gynecologic cancers In a retrospective study of infectious morbidity on a university gynecologic oncology service, Brooker and others [86] found 20 (6%) of 494 patients had a serious infection on admission and 54 patients (11%) developed serious infections during hospitalization. The infection rate per admission varied by cancer origin: 8% for cervical cancer, 7% for uterine cancer, 3% for ovarian cancer, and 21 % for vulvar cancer. Bacteremia in gynecologic cancers may be caused by a single organism or be polymicrobial, with the primary tumor the likely portal of entry [Z]. In general, the organisms involved in gynecologic infections are normal flora of the vagina, gastrointestinal tract, and skin (Figure 3). Infectious complications of gynecologic cancers at diagnosis highlight underlying problems that may occur secondary to changes in endogenous flora. When infection complicates stage I cervical cancer, the infection is typically limited to the vagina, covering only the surfaces of the tumor itself [87].The abnormal neoplastic tissue allows bacterial overgrowth of normal flora to take place. Streptococcal species are usually isolated from the purulent debris. Rose and Wilson [88]presented a case of toxic shock syndrome in a patient with previously undiagnosed advanced cervical cancer. The cervical cancer was believed to be the portal of entry for staphylococcal toxins to reach the bloodstream. Obstruction probably contributes to adnexal infections in patients with advanced cervical disease. Barton et al. [89] presented three unusual cases in which patients presented with cervical cancer complicated by tuboovarian abscesses. In two cases, the patients were initially overstaged secondary to inflamed adnexal masses, which were later found to be free of cancer. The third patient developed an acute abdomen secondary to a ruptured tuboovarian abscess shortly after detection of an exophytic cervical mass. Patients with cervical disease that has invaded surrounding tissues by direct extension may be more likely to develop pyometra (pus in the uterus) [87]. Some patients with pyometra present with the classic triad of purulent vaginal drainage, fever, and lower abdominal pain [90]. Pyometra, however, can often be asymptomatic, presenting without fever or pelvic pain [87]. Typical organisms are aerobic and anaerobic streptococci [87]. Not all patients with this infection have a malignancy, and the pathogenesis is not always associated with obstruction at the cervical 0s. An early study [91] demonstrated that a fixed obstruction of the cervix was not necessary to maintain a pyometra; in this collection of 52 removed uteri with pyometra, only 24 had obstructed cervices. Of note, 26 patients developed infection following radiation therapy. Of 52 patients, only 8 had malignancies of the endometrium or cervix and 3 patients had gastrointestinal carcinomas. Peritonitis can occur if a pyo~netraruptures. A collection of 15 cases in the literature of spontaneous perforation of pyometra found that one third of the patients had malignant disease [90]. All 15 patients presented with fever, 53%

with vomiting, and 20% with atypical genital bleeding. The most common organisms in peritoneal fluid were Escherichia coli, Bacteroides spp., and polymicrobial. Douvier et al. [92] reported two cases of perforation of the uterus at the site of endometrial carcinoma, resulting in peritonitis. Very advanced, usually undiagnosed, carcinoma of the cervix has been associated with spontaneous rupture into the retroperitoneum [87]. The severity of infectious complications at cancer diagnosis appears to correlate with the extent of tumor invasion and secondary obstruction; because the vast majority of gynecologic cancers in the United States are now diagnosed when disease is localized, complications such as peritonitis are extremely rare at the time of cancer detection. Surgery, chemotherapy, andlor radiation therapy contribute to infectious complications of gynecologic oncology patients. Under such abnormal conditions, a normally benign vaginal commensal can proliferate, invade the bloodstream, and cause sepsis. For example, Andriessen et al. [93] reported a nonneutropenic patient who presented with sepsis following chemotherapy for metastatic choriocarcinoma; multiple blood cultures grew LactobaciElus acidophilus. A gallium scan showed only diffuse uptake in her uterus. A second patient became septic 2 days following surgery and was found to have a Lactobacillus spp. pelvic abscess and bacteremia [94]. In a retrospective study of infections associated with gynecologic cancers, cervical cancer had the highest surgical infection rate (22%); examples of such infections were peritonitis, pelvic hematomas, groin abscesses, and drainage tube infections [86]. Similar types of infections were documented as complications of uterine and ovarian cancers. Prior radiation therapy and surgery appeared to be risk factors for infection in patients with cervical and uterine cancers. Preoperative subclinical pelvic infections, invasive diagnostic procedures, and invasive devices of supportive care (nasogastric tubes, urinary catheters, central lines) may also contribute to the development of postoperative infections [86]. Graham [95] suggested that two additional factors contributed to gynecologic surgical infections: first, removed organs and tissue create a space that fills with blood and serum, an excellent culture medium, and, second, bowel obstruction or ileus can result in poor nutritional status preoperatively and postoperatively. Ovarian carcinoma is the most common gynecologic cancer to be treated with chemotherapy; these patients are at risk for the infectious complications associated with neutropenia. Complications of pelvic irradiation for cervical cancer, such as fistula formation and small bowel obstruction or perforation, are rare and may be associated with previous pelvic inflammatory disease [96]. Infections associated with gynecologic cancers can be life-threatening. Empiric antibiotics for a febrile patient with suspected advanced gynecologic cancer should cover anaerobes and aerobic gram-negative bacilli. Candidrr spp. may play a role in some infections because they are part of normal and abnormal vaginal flora. Enterococci and group B streptococci can also be involved in gynecologic infections, particularly abscesses. Cultures of

blood, vagina, pyometra, and abscess drainage abscesses can help guide therapy.

4.6 Bacillus Calmette-Guirin dissemination Dissemination of intravesicular bacillus Calmette-Gukrin (BCG) therapy is an unusual outcome of a unique anticancer therapy commonly used in bladder cancer. The antineoplastic mechanism of action of this Mycobacterium bovis strain is thought to be as an immune modulator. Cases of disseminated BCG have been rare. Lamm et al. [97] found granulomatous hepatitis and1 or pneumonitis in 0.7% of bladder cancer patients receiving intravesicular BCG therapy. M. bovis is presumed to spread hematogenously from the bladder. The time from installation to symptomatic presentation appears to be highly variable, ranging from hours to several months after exposure. Proctor et al. [98] presented a case of an elderly male who experienced fever and rigors 5 hours after installation. A blood culture from that day grew M. bovis. The patient developed mild hepatitis, hyperbilirubinemia, and AFB-positive hepatic granulomas. The patient improved on isoniazid and rifampin for 12 months plus 2 months initial ethambutol. In contrast, Hakim and colleagues 199) reported a case in which an elderly bladder cancer patient presented with a M. bovis psoas abscess 9 months after BCG therapy. Katz et al. [loo] reported a case of a man who presented with lumbar vertebral osteomyelitis and psoas abscess approximately 4 months after completing a year of BCG therapy. Cultures from bone and abscess fluid grew M. bovis. Subsequent vertebral surgery revealed necrotic bone with AF'B-positive caseating and noncaseating granulomas. He responded well to abscess drainage and isoniazid and rifampin. Other cases of BCG infection that are consistent with hematogenous spread included: acute prosthetic knee arthritis [loll, septic arthritis of the elbow [102], and pulmonary infections [103]. Pulse field gel electrophoresis has been used to confirm that the instilled organism is identical to the infecting pathogen [103]. Infectious arthritis, which should be treated with antituberculous drugs, should be differentiated from BCG-associated reactive polyarthritis [I041 or Reiter's syndrome [105], which are treated with antiinflammatory agents. One needs to have a high index of suspicion that BCG may be the cause of infection in a febrile patient who is receiving or received BCG therapy for bladder cancer. Attempts should be made to identify the organism because prolonged, potentially toxic treatment is required. When BCG disseminates, therapy requires prolonged antituberculous medication and drainage of any abscess. M. bovis is intrinsically resistant to pyrazinamide. Most cases in the literature reported cure when patients were treated with isoniazid and rifampin for 6-9 months. A patient who suffers invasive BCG infection should not receive any additional BCG therapy.

5. Infectious complications of head and neck cancer Infections occur commonly in patients with head and neck cancer. Because patients typically become symptomatic at late stages of disease, they often present when large tumors obstruct airways or inhibit swallowing. Tumor involvement of the oral mucosa, a reservoir of substantial numbers of bacteria, provide a ready source of pathogens (Figure 4). In the late stages of disease, many patients experience profound weight loss secondary to cachexia and restricted intake. The resulting malnutrition and related immunosuppression, as well as frequent comorbidities of chronic obstructive pulmonary disease (COPD), liver disease, and poor dental hygiene, further diminish the ability of these patients to effectively fight infection. As in other solid tumor patients, surgery, radiation therapy, and chemotherapy further impair local and systemic host defenses. The combination of these factors results in a significant risk of infection. 5.1 In,fections after radiation therapy

Head and neck cancer patients commonly receive treatment with chemotherapy, radiation, and surgery. Radiation can result in delayed healing, posing an increased risk of wound infection. In addition, radiation can dramatically reduce saliva production, causing an alteration in oral flora favoring more virulent organisms [106-1081. Microbial samples from plaque and saliva of patients before and after irradiation revealed significant increases in Streptococcus mutans, Lactobacillus spp., Candida spp., Staphylococc~tsspp., enteric gram-negative bacilli, and anaerobes [108]. A progressive drop in salivary production was identified, beginning within 2 weeks of bilateral parotid gland irradiation and decreasing to 6% of initial flow rates in patients by 3 months; at that time dental caries were noted with increased frequency [109]. Fungal infections cause a great deal of morbidity in the head and neck patient following irradiation [107], including increased pain and difficulty with speech [110]. Irradiation to the mouth or the larynx changes the skin flora and the mouth flora for up to 6 months, resulting in significant overgrowth by a variety of yeast. 5.2 Postoperative woun,d infections Most surgery for head and neck cancer, because it involves the upper respiratory and gastrointestinal tract, is considered to be contaminated or clean contaminated. During surgery and throughout the healing process, most wounds are in intimate contact with the rnucosal surfaces, or secretions, of the oropharynx and respiratory tract. Salivary bacterial counts are in the range of 10~109/m [Ill]. ~ Anaerobes account for 90% of the organisms in the oral cavity; the remainder are gram-positive and gram-negative aerobic organisms. Contamination during and after surgery by oral flora is thought to contribute

greatly to the high rate of wound infections after head and neck surgery. Infection rates of over 80% have been recorded when prophylactic antibiotics were not used. The impact of anaerobic organisms in the pathogenesis of wound infections in head and neck cancer became evident during the 1980s, coinciding with improved techniques of isolating anaerobic bacteria. Sawyer [I121 clarified the need for anaerobic coverage in most head and neck cancer surgeries by demonstrating a significantly reduced infection rate following prophylaxis with metronidazole and cefazolin when compared with recent historical controls using cefazolin alone. A prospective, randomized trial in head and neck cancer patients confirmed improved infection rates with cefazolin and metronidazole (9.5% infection rate vs. 18.6% infection rate for cefazolin alone) [113]. The majority of head and neck postoperative wounds have been polymicrobial. Brook and Hirokawa [I141 cultured 24 postoperative wounds of head and neck cancer patients, finding that 88% of the wounds were mixed anaerobic and aerobic flora. Peptostreptococcus spp., Bacteroides spp. (non-frcrgilis), and Fzwobacterium spp, were the most commonly identified anaerobes. Increased risk of wound infection can be attributed to the extent of disease, duration of surgery, and technical constraints. While tumor removal is of primary importance, surgeons attempt to preserve the airway, cough reflex, diaphragm function, speech, facial muscles and nerves, hormonal function, lymphatic drainage, and saliva production [115]. Seroma and hematoma formation can contribute to the risk of abscess; effective hemostasis and use of drains can minimize bleeding and edema. Patients with advanced disease often require removal of large amounts of tissue, thereby exposing extensive wounds. Technical decisions that influence perfusion of a graft or skin flap can have significant impact on infection risk. Tandon et al. [I161 found that patients who underwent a muscle flap procedure, which reflects extensive disease, had increased risk of infection. Eight of 12 patients who underwent a pectoralis major flap developed wound infections. Robbins et al. [I171 reviewed 400 head and neck patients who underwent surgery and found a wound infection rate of 20%. Presence of advanced disease, duration of surgery greater than 6 hours, placement of a flap, as well as absence of anaerobic coverage perioperatively were found to be significant risk factors for wound infections. Brown and coworkers [1181 recorded an overall 7% wound infection rate among 245 head and neck patients receiving perioperative antibiotics; the subset of patients with stage 1%' disease had a 15% risk of wound infection. Similarly, those patients who underwent a myocutaneous flap procedure had an infection rate of 36%, whereas those receiving simpler procedures had a risk of 6%. These investigators also identified probable errors in surgical decision-making or technique in 10 of the 17 patients who developed wound infections, many of which resulted in flap or skin graft failure from ischemia, bleeding, tension, or trauma. Some risk factors for wound infection in head and neck cancer patients may be difficult to modify,

namely, the extent of disease at presentation and the technical and physical constraints that result.

5.3 Nonwound infections in head and neck patients The lower respiratory tract is the primary site of nonwound infections in head and neck cancer patients. These infections are a major cause of mortality in this patient group. Hussain et al. [I191 reviewed 12 months of admissions to a university head and neck cancer service. Eighty-six infections were documented among 102 febrile episodes in 67 patients. Forty-three percent of the infections were attributed to pneumonia or tracheobronchitis. Eighteen percent of deaths were directly attributed to pneumonia. Papac [I201 reported 78 infectious complications among 191 patients with advanced head and neck cancer hospitalized on a medical oncology service at a Veterans Affairs hospital. Of 111 reported deaths, pneumonia was the most frequent cause of death (26%), twice as common as the next leading cause of death, tumor or metastasis. Pneumonia was the most common infection in a study of perioperative morbidity among patients with head and neck cancer [121]. Twenty-two of 225 patients experienced lower respiratory tract infections. Of the 22 patients, 19 developed postoperative pneumonia and 3 had tracheobronchitis. Duration of surgery greater than 6.2 hours increased the risk of nonwound infection from 4.5% to 15.3%. Having a greater than 70 pack-year history of smoking and receiving a blood transfusion perioperatively also significantly increased the risk of pulmonary infection. Length of stay for those with pneumonia increased from a mean of 14.6 days to 23 days (P < .05). Other sources of nonwound infections were the urinary tract (3), septic phebitis (I), and acute sinusitis (1).

5.4 Management of infections in head and neck cancer At the time of diagnosis of head and neck cancer, some steps can be taken to prevent infectious complications. Because tuberculosis can become reactivated in these patients, it is prudent to administer isoniazid prophylaxis to those who are tuberculin positive, but without disease, despite their advancing age and possible underlying liver disease. Pneumococcal vaccination prior to instituting therapy and annual influenza vaccination may be protective. Anticipating problems involving the oral cavity is key to minimizing infectious complications during and after radiation therapy, and to a lesser extent, during and after surgery. Before receiving radiation therapy or undergoing surgery, a head and neck cancer patient should have a complete dental evaluation, including dental x-rays [115,122]. Carious teeth should be removed or restored. Institution of oral hygiene can reduce the risk of subsequent infectious complications. Patients should use salivary substitutes throughout the day and receive regular fluoride treatments [122]. Some investigators recommend selective decontamination with topical antibiotics during the weeks of

irradiation, but this remains unproven. The patient should be well educated regarding the benefits of rigorous oral hygiene during and after radiation. Postoperative management is critical in the efforts to reduce morbidity in these patients. In the event of a postoperative wound infection, aerobic and anaerobic cultures should be sent. Empiric antibiotic therapy covering anaerobes, gram-negative bacilli, and aerobic gram-positive cocci is appropriate. Infected fluid collections should be drained. Coleman [I151 urged prompt exploration of the surgical wounds if fistulae or necrosis (or other evidence of infection) develop. The index of suspicion of infection should be higher if the patient has received preoperative irradiation. Should a fistula form near or overlying the carotid sheath, an emergency exploration should be performed; infection of the carotid artery in a previously irradiated site can result in septic emboli or carotid artery rupture with exsanguination [115]. Attempts to prevent aspiration in a patient with advanced head and neck cancer are often futile. Pneumonia is relatively common and often responds to empiric broadspectrum antibiotics.

6. Conclusions The medical literature describing infectious complications in patients with solid tumors is limited to isolated case reports and retrospective studies dominated by patients with hematologic malignancies. There are no infections unique to solid tumor patients. Two broad categories of infection in patients with solid tumors can be described. Some infections are directly attributable to the tumor. These infections develop because the neoplastic process causes focal injury, breaks down normally intact barriers, or causes local obstruction. Examples include S. bovis endocarditis or C. septicurn myonecrosis in patients with colon cancer, postobstructive pneumonia complicating lung cancer, and pyometra complicating advanced cervical carcinoma. The second major category of infections may be attributed to the effects of cancer treatment: surgery, chemotherapy, and radiation therapy. Examples include upper extremity cellulititis complicating the lymphedema that results from surgery and radiation therapy for breast cancer, intraabdominal sepsis following attempted resection of a solitary hepatic metastasis in a patients with colon cancer, and catheter-related sepsis in a patient with head and neck cancer. Wound infection deserves particular attention in patients with solid tumors. For all types of surgery, the level of contamination and the length of the procedure are the critical determinants of wound infection risk. Neoplasia appears to have a definite, if indirect, impact on wound infection rates because the location and extent of tumor define the level of contamination and influence the duration of surgery. A recent large study at Memorial SloanKettering Cancer Center identified four significant risk factors for wound infections after cancer surgery: prolonged surgery, a poor preoperative physical assessment score, and contaminated or dirty surgeries and obesity [123].

Preoperative chemotherapy, radiation therapy, and the neoplastic process itself no doubt continue to impair wound healing and the risk of infection. Only a small subset of those with solid tumors sustain the level of chemotherapy-related immunosuppression associated with an increased risk of Pneumocystis cnrinii pneumonia, disseminated aspergillosis, hepatosplenic candidiasis, or neutropenic enterocolitis. Nevertheless, the types of cancers that are treated with intensive chemotherapy regimens and autologous bone marrow transplantation appear to be increasing. Advanced non-small cell lung cancer patients demonstrate improved 5-year survival with the addition of induction chemotherapy to radiation [124]. Ongoing trials of autologous bone marrow transplantation for advanced breast cancer are attempting to establish whether such intensive chemotherapy results in survival advantage. As more solid tumor patients receive intensive chemotherapy, we can anticipate an increase in the incidence of severe infectious complications. Recognizing the subset of solid tumor patients who are severely immunosuppressed is critical when infection is first suspected; outcome in these individuals may depend upon considering an expanded differential diagnosis at the outset, obtaining appropriate microbiologic and pathologic specimens, often via invasive procedures, and employing empiric antibiotics thoughtfully and rapidly. Management of infectious complications of solid tumors is aided by understanding their pathogenesis, minimizing nosocomial risks, and utilizing modern imaging, invasive procedures, and modern culture techniques.

References 1. Boring CC, Squires TS, Tong T. Montgomery S. Cancer Statistics, 1994. CA Cancer J Clin 1994;44:7-26. 2. Elting LS, Bodey GP, Fainstein V. Polymicrobial septicemia in the cancer patient. Medicine 1986;65:218-225. 3. Mayo JW, Wenzel RP. Rates of hospital-acquired bloodstream infections in patients with specific malignancy. Cancer 1982;50:187-190. 4. Mathe G, Meyer P, Brienza S. Gil-Delgado M, Orbach-Arbouys S, Itzhaki M. Retrospective study correlating clinical infectious history and peripheral blood T-cell subpopulations in cancer, G v H and HIV+ patients. Biomecl Pharmacother 1992:46:17-19. 5. Pizzo PA, Robichaud KJ, Wesley R, Con~mersJR. Fever in the pediatric and young adult patient with cancer: A prospective study of ZOO1 episodes. Medicine 1982;61:153-165. 6. Singer C, Kaplan MH, Armstrong D. Bacteremia and fungemia complicating neoplastic disease: A study of 364 cases. Am J Med 1977;62:731-742. 7. Krumpe PE, Knudson RJ, Parsons G, Reiser K. The aging respiratory system. Clin Geriatr Med 1985;1:143-175. 8. Brugarolas A, Takita H. Immunologic status in lung cancer. Chest 1973;64:427-430. 9. Busch E, Verazin G, Antkowiak JG, Driscoll D, Takita H. Pulnlonary complications in patients undergoing thoracotomy for lung carcinoma. Chest 1994;105:760-766. 10. Kohno S. Koga H. Oka M, Kadota J. Kaku M, Soda H, Tomono IS, Hara K. The pattern of respiratory infection in patients with lung cancer. Tohoku J Exp Med 1994;173:405-411. 11. Strang C, Simpson JA. Carcinon~atousabscess of the lung. Thorax 1953;8:11-25.

12. Perlman LV, Lerner E , D'Esopo N. Clinical classification and analysis of 97 cases of lung abscess. A m Rev Respir Dis 1969;99:390-398. 13. Snider GL, Saleh SS. Empyema of the thorax in adults: Review of 105 cases. Dis Chest 1968;54:12-11. 14. Feld R. Bodey GP, Groschel D. Mycobacteriosis in patients with malignant disease. Arch Intern Med 1976:136:67-70. 15. Kaplan MH, Armstrong D, Rosen P. Tuberculosis complicating neoplastic disease: A review of 201 cases. Cancer 1974;33:850-858. 16. Snider GL. Placik B. The relationship between pulmonary tuberculosis and bronchogenic carcinoma: A topographic study. Am Rev Resp Dis 1969:99:229-236. 17. Nagata N, Nikaido Y, Kido M, Ishibashi T, Sueishi K. Terminal pulmonary infections in patients with lung cancer. Chest 1993:103:1739-1742. 18. Smith FB, Bveneck B. Localized aspergillus infestation in primary lung carcinoma: Clinical and pathological contrasts with post-tuberculous intracavitary aspergilloma. Chest 1991; 100:554-556. 19. D e Vuyst P, D e Troyer A , Yernault J, Verhest A, Vanderhoeft P. Aspergilloma in a necrotic bronchial adenocarcinoma. Eur J Respir Dis l980;6 1:213-217. 20. Meyer RD, Young LS. Armstrong D , Yu B. Aspergillosis complicating neoplastic disease. An1 J Med 1973;54:6-15. 21. Borkin MH, Arena FP, Brown A E , Armstrong D. Invasive aspergillosis with massive fatal hemoptysis in patients with neopla5tic disease. Chest 1980;78:835-839. 22. Smith JM, Hampton JR, Webb RW. Disseminated aspergillosis, Cushing's syndrome, and oat cell carcinoma of the lung. Chest 1979:76:112-113. 23. Sidran3ky H. Friedman L. The effect of cortisone and antibiotic agent on experimental pulmonary aspergillosis. A m J Path01 1959:35:169-183. 24. Hasan FM, Jarrah T, Nassar V. The association of adenocarcinotna of the lung and blastomycosis from an unusual geographical location. Br J Dis Chest 1978;72:242-246. 25. Blastomycosis Cooperative Study of the Veterans Administration: A review of 198 collected cases in Velerans Administration hospitals. Ail1 Rev Resp Dis 1964;89:659-672. 26. Haron E , Vartivarian S. Anaissie E, Dekmezian R, Bodey GP. Primary Crindlda pneumonia: Experience at a large cancer center and review of the literature. Medicine 1993;72:137142. 27. Yale SH. Limper AH. Przeut?zoc~~stis corrnii pneumonia in patients without acquired immunodeficiency syndrome: Associated illnesses and prior corticosteroid therapy. Mayo Clinic Proc 1996;7 15-13. 28. Sepkowitz KA, Brown AE, Tolzak E E , Gottlieb S, Armstrong D . Pneunzocysti~carinii pneunlotlia among patients without AIDS at a cancer hospital. JAMA 1992;267:832-837. 29. Graham BS. Tucker WS. Opportunistic infections in endogenous Cushing's syndrome. Ann Intern Med 1984:101:334-338. 30. Frenkel JK. Good JT. Shultz JA. Latent Pnez(nzocystis infection in rats, relapse, and chemotherapy. Lab Invest 1966;15:1559-1577. 31. Walzer PD. LaBine M. Redington TJ, Cushion MT. Lymphocyte changes during chronic administration of and withdrawal from corlicosteroids: Relation to Pnerrinocyst~scarinii pneumonia. J Immunol 1984:133:2502-2508. 32. Henson JW, Jalaj JK, Walker RW, Stover DE. Fels AOS. Pt~eumocystiscar.il~iipneumonia in patients with pri~narybrain tumors. Arch Neurol 1991:48:406-409. 33. Poplin EA, Gordon CJ, Piskorowski TJ. Chandrasekar PH. Pnertnzocystis carinii pneumonia in patlents with solid tumors without acquired imtnune deficiency syndrome. Cancer 199 1:68:193-194. 34. Kovacs J A , Hiemenz JW, Macher AM, Stover D , Murray HW, Shelhamer J, Lane HC, Urmacher C, Honig C. Longo DL, Parker MM. Natanson C, Parrillo J E , Fauci AS. Pizzo PA, Masur H. Ptzerrmocystis cllrirril pneumonia: A comparison between patients with the acquired in~munodeficiencysyndrome and patients with other immunodeficiencies. Ann Intern Med 1984;100:663-671.

35. Limper AH, Offord KP, Smith TF, Martin WJ. Pneumocystis carinii pneumonia: Differences in lung parasite number and inflammation in patients with and without AIDS. Am Rev Respir Dis 1989;140:1204-1209. 36. Sepkowitz KA. Pneumocystis carinii pneumonia without acquired immunodeficiency syndrome: Who should receive prophylaxis? Mayo Clin Proc 1996;71:102-103. 37. Camazine B, Antkowiak JG, Nava MER, Lipman BJ, Takita H. Herpes simplex viral pneumonia in the postthoracotomy patient. Chest 1995;108:876-879. 38. Coggle JE, Lambert BE, Moores SR. Radiation effects in the lung. Environ Health Perspect 1986;70:261-291. 39. Gross NJ. Pulmonary effects of radiation therapy. Ann Intern Med 1977;86:81-92. 40. McInerney DP, Bullimore J. Reactivation of radiation pneumonitis by adriamycin. Br J Radiol 1977;50:224-227. 41. Schwarz MI, Whitcomb ME, Goldman AL. The spectrum of diffuse pulmonary infiltration in malignant disease. Chest 1973;64:88-93. 42. Pisani RJ, Wright AJ. Clinical utility of bronchoalveolar lavage in imnlunocompromised hosts. Mayo Clin Proc 1992;67:221-227. 43. Shure D. Transbronchial biopsy and needle aspiration. Chest 1989;95:1130-1138. 44. Rotstein C, Ferguson R, Cummings KM, Piedmonte MR, Lucey J, Banish A. Determinants of clean surgical wound infections for breast procedures at an oncology center. Infect Control Hosp Epidemiol 1992;13:207-214. 45. Sauter ER, Eisenberg BL, Hoffman JP, Ottery FD, Boraas MC. Goldstein LJ, Solin LJ. Postmastectomy morbidity after combination preoperative irradiation and chemotherapy for locally advanced breast cancer. World J Surg 1993;27:237-242. 46. Vinton AL, Traverso LW, Jolly PC. Wound coinplications after modified radical mastectomy compared with tylectomy with axillary lymph node dissection. Am J Surg 1991;161:584-588. 47. Mozes M, Papa MZ, Karasik A, Reshef A, Adar R. The role of infection in post-nlastectomy lymphedema. Surg Ann 1982;14:73-83. 48. Bertelri G, Venturini M, Forno G, Macchiavello F, Dini D. An analysis of prognostic factors in response to conservative treatment of postrnastectomy lymphedema. Surg Gynecol Obstet 1992;175:455-460. 49. Simon MS, Cody RL. Cellulitis after axillary lymph node dissection for carcinoma of the breast. Am J Med 1992;93:543-548. 50. Chaudhuri K, Lonergan D, Portek I, McGuigan L. Septic arthritis of the shoulder after mastectomy and radiotherapy for breast carcinoma. J Bone Joint Surg Br 1993;75-B:318-321. 51. Rescigno J, McCormick B, Brown AE, Myskowski PL. Breast cellulitis after conservative surgery and radiotherapy. Int J Radiat Oncol Biol Phys 1994;29:163-168. 52. Olivotto IA, Weir L, Kim-Sing C. In response to Rescigno et al. Int J Radiat Oncol Biol Phys 1995;31:203. 53. Keidan RD, Hoffman JP, Weese JL, Hanks GE, Solin LJ, Eisenbe BL, Ottery FD, Borass M. Delayed breast abscesses after lumpectomy and radiation therapy. Am Surg 1990;56:440444. 54. Mudad R, Vredenburgh J. Paulson EK, Ross M, Meisenberg B, Hussein A, Peters WP. A radiologic syndrome after high dose chemotherapy and autologous bone marrow transplantation, with clinical and pathologic features of systemic candidiasis. Cancer 1994;74:13601366. 55. Parry BR. Radiation recall induced by tamoxifen. Lancet 1992;340:49. 56. Rostom, AY, ME El-Sayed. Fat necrosis of the breast: An unusual complication of lumpectomy and radiotherapy in breast cancer. Clin Radiol 1987;38:31. 57. Bertelli G, Dini D, Forno GG, Gozza A. Preventing cellulitis after axillary lymph node dissection. Am J Med 1994;97:202-203. 58. Dixon JM. Breast infeclion. Br Med J 1994;309:946-949. 59. Finegold SM. Anaerobic bacteria: General concepts. In: Mandell GL, Bennett JE, DoIin R, eds. MandelI, Douglas and Bennett's Principles and Practice of Infectious Diseases, 4th ed. New York: Churchill Livingstone, 1995, pp. 2156-2173.

60. Klein RS, Recco RA, Catalano MT, Edberg SC, Casey JI, Steigbigel NH. Association of Streptococcils bovis with carcinoma of the colon. N Engl J Med 1977;297:800-802. 61. Klein RS, Catalano MT, Edberg SC, Casey JI, Steigbigel NH. Streptococcus bovis septicemia and carcinoma of the colon. Ann Intern Med 1979;91:560-562. 62. Fass R, Alim A, Kaunitz JD. Adenocarcinoma of the colon presenting as Streptococcus snngrlis bacteremia. Am J Gastroenterol 1995;90:1343-1345. 63. Wiseman A, Rene P, Crelinsten GL. Streptococcils agalactiae endocarditis: An association with villous adenomas of the large intestine. Ann Intern Med 1985;103:893-894. 64. Gelfand MS, Hughey JR, Sloas DD. Group B Streptococcal bacteremia associated with adenocarcinoma of the stomach. Clin Infect Dis 1994;39:364. 65. Katlic MR. Derkac WM, Coleman WS. Clostridiunz septiciinz infection and malignancy. Ann Surg 1981;193:361-364. 66. Koransky JR, Stargel MD, Dowel1 VR. Clostridirlm septicurn bacteremia: Its clinical significance. Am J Med 1979;66:63-66. 67. Kornbluth AA, Danzig JB, Bernstein LH. Clostridiilnz septiciinz infection and associated malignancy: Report of 2 cases and review of the literature. Medicine 1989;68:30-37. 68. Gorbach SL, Thadepalli H. Isolation of Clostridiilrn in human infections: Evaluation of 114 cases. J Infect 1975;131:S81-S85. 69. Jendrzejewski JW. Jones SR, Newcombe RL, Gilbert DN. Nontraumatic clostridial myonecrosis. Am J Med 1978;65:542-546. 70. Kaiser CW, Milgrom ML, Lynch JA. Distant nontraumatic clostridial myonecrosis and malignancy. Cancer 1986;57:885-889. 71. Hovenden JL, Murdoch GE, Evans AT. Non-traumatic Clostridium septicurn arthritis in a patient with caecal carcinoma. Rr J Rheumatol 1992;31:571-572. 72. Fallon SM, Guzik HJ, Kramer JE. Clostridium septicunz arthritis associated with colonic carcinoma. J Rheumatol 1986;13:662-663. 73. Alpern RJ, Dowel1 VR. Clostricliiinz septic~iminfections and malignancy. JAMA 1969; 209:385-388. 74. Kahn SP, Lindenauer SM, Wojtalik RS, Hildreth D. Clostridia hepatic abscess: An unusual manifestation of metastatic colon carcinoma. Arch Surg 1972;104:209-212. 75. Kudsk KA. Occult gastrointestinal malignancies producing metastatic Clostridiiirn septicirm infections in diabetic patients. Surgery 1992;112:765-772. 76. Cohen JL, Martin M, Rossi RL, Schoetz DJ. Liver abscess: The need for complete gastrointestinal evaluation. Arch Surg 1989:124:561-564. 77. Marcus SG. Walsh TJ, Pizzo PA, Danforth DN. Hepatic abscess in cancer patients: Characterization and management. Arch Surg 1993;128:1358-1364. 78. Okada S, Aoki K, Okazaki N, Nose H, Yoshimori M, Shimada K, Yamalnoto J, Takayama T, Kosuge T, Yamaski S, Takayasu K, Moriyama N. Liver abscess after percutaneous ethanol injection (PEI) therapy for hepatocellular carcinoma. A case report. Hepatogastroenterology 1993;40:496-498. 79. Isobe H, Sakamoto S, Nawata H. Splenic abscess as a complication of percutaneous ethanol injection therapy for hepatocellular carcinoma. J Clin Gastroenterol 1992;15:167169. 80. Teitz S, Guidetti-Sharon A, Manor H, Halevy A. Pyogenic liver abscess: Warning indication of silent colonic cancer: Report of a case and review of the literature. Dis Colon Rectum 1995;38:1220-1223. 81. Lonardo A, Grisendi A, Pulvirenti M, Della Casa G, Melini L, DiGregorio C, Nasi G, Sarti M. Tamborrino E, Lonardo F. Right colon adenocarcinoma presenting as Bactevoides fragilis liver abscesses. J Clin Gastroenterol 1992;14:335-338. 82. Wolfe MS, Armstrong D, Louria DB, Blevins A. Salmonellosis in patients with neoplastic disease: A review of 100 epidsodes at Memorial Cancer Center over a 13-year period. Arch Intern Med 1971;128:546-554. 83. Han T, Sokal JE, Neter E. SalmoneHosis in disseminated malignant diseases: A seven-year review (1959-1965). N Engl J Med 1967;276:1045-1052.

84. Kate P, Osei K, Chiemchanya S, Zatuchni J. E~npyemadue to SnL?zonella typhinzliriun~with underlying alveolar ccll carcinoma. South Med J 1984:77:234-236. 85. Berkeley D , Mangels J. Snlntoncll~~ pneutnonia in a patient with carcinoma of the lung. Am J Clin Path01 1980;74:476-478. 86. Brooker DC, Savage JE, Twiggs LB. Adcock LL, Prem KA. Sanders CC. Infectious morbidity in gynecologic cancer. An1 J Obstet Gynecol 1987;156:513-520. 87. WeIton JA. Significance of infection in carcinonia of the cervix. Obstet Gynccol1965;26:441446. 88. Rose PG, Wilson G. Advanced cervical carcinoma presenting with toxic shock syndrome. Gynecol Oncol 1994:52:264-266. 89. Barton DPJ, Fiorica JV. Hoffman MS. Roberts WS, Finan MA, Cavanagh D. Cervical cancer and tuboovarian abscesses: A report of three cases. J Reprod Med 1993;38:561-564. 90. Imachi M, Tanaka S, Ishikawa S, Matsuo K. Spontaneous perforation of pyometra preTenting as generalized peritonitis in a patient with cervical cancer. Gynecol Oncol 1993;50:384388. 91. DeVoe RW, Randall LM. Pyometra: A clinical and pathologic study. Am J Obstet Gynecol 1949;58:784-789. 92. Douvier S, Nabholtz J-M, Friedman S, Cougard P, Ferry C. Aupecle P. Infectious pneumoperitoneum as an uncommon presentation of endometrial carcinoma: Report of two cases. Gynecol Oncol 1989;33:392-394. 93. Andriessen MPHM. Mulder JG. Sleijfer DT. Lactobacillus septicemia, an unusual complication during the treatment of metastatic choriocarcinoma. Gynecol O~icol1991;40:87-89. 94. Connor JP, Buller RE. Lactobacillus sepsis with pelvic abscess. Gynecol Oncol 1994;54:99100. 95. Graham JE. Infectious morbidity in gynecologic oncology. J Reprod Med 1990;35:348-352. 96. Van Nagell JR. Parker JC, Maruyama Y, Utlep J. Hager WD, Donaldson ES, Wood EG. The effect of pelvic inflammatory disease on enteric complications following radiation therapy for cervical cancer. Am J Obstet Gynecol 1977;128:767-77 1. 97. Lamm DL, van der Meijdcn PM, A Morales A, Brosman SA. Catalona WJ. Herr HW, Soloway MS, Steg A, Debruyuc FM. Incidence and treatment of colnplications of bacillus Calmette-Guirin intravesical therapy in supcrficial bladder cancer. J Urol 1992;147:596-600. 98. Proctor DD, Chopra S, Rubenstein SC. Jokela JA. Uhl L. Mycobacterernia and granulomatous hepatitis following initial intravesical bacillus Calmette-Gukrin installation for bladdcr carcinoma. Am J Gastroenterol 1993;S8:1112-1115. 99. Hakim S, Heaney JA. Heinz T. Zwolak RW. Psoas abscess following intravesical bacillus Calmette-Guerin for bladdcr cancer: A case report. J Urol 1993;I50:188-189. 100. Katz DS, Wogalter H, D'ETposito RF. Cunha BA. Mycohacter.rlir?z bovis vertebral osteornyelitis and psoas absccss after intravcsical BCG thcrapy for bladder carcinoma. Urology 1992;40:63-66. 101. Chazerain P, Desplaces N. Mamoudy P, Leonard P, Ziza J. Prosthetic total knee infection with a bacillus Calmette Gukrin (BCG) strain after BCG therapy for bladder cancer. J Rheumatol 1993:20:2171-2172. 102. Van Thiel RJ. Nossent GD. Tjonpiangi NP. Articular complications of intravesical BCG treatment for bladder carcinoma. Br J Urol 1992:70:446-447. 103. Kristjansson M. Green P. Manning HL, Slutsky AM, Brecher SM. Fordharn von Reyn C. Arbeit RD, Maslow JN. Molecular confirmation of bacillus Calruettc-Guirin as the cause of pulmonary infect~onfollowing urinary tract installation. Clin Infect Dis 1993:17:228230. 104. Jawad ASM. Kahn L. Copland RFP, Henderson DC, Abdul-Ahad AK. Reactive arthritis associated with bacillus Calmette-Gukrin immunothcrapy for carcinoma ot the bladder: A report of two cases. Br J Rheumatol 1993:32:1018-1020. 105. Pancaldi P, Van Linthoudt D, Alboriiio D, Haefliger JM, Ott H. Reiter's syndrome after intravesical bacillus Calmette-Guerin treatment for superficial bladder carcinoma. Br J Rheuniatol 199332:1096-1 098.

106. Ariyan S. Infections following surgery for head and neck cancer. Clin Plast Surg 1979:6:523535. 107. Curtis TA. Griffith MR, Firtell DN. Complete denture prosthodontics for the radiation patient. J Prosthet Dent 1976;36:66-76. 108. Brown LR, D r e i ~ e nS, Handler S. Johnston D A . Effect of radiation-induced xerostomia on hurnan oral microflora. J Dent Res 1975:54:740-750. 109. Dreizen S. Brown LR. Handler S, Levy BM. Radiation-induced xcroslomia in cancer patients: Effect on salivary and serum electrolytes. Cancer 1976;38:273-278. 110. Martin MV, Al-Tikriti U, Bramley PA. Yeast flora of the mouth and skin during and after irradiation for oral and laryngeal cancer. J Med Microbiol 1981:14:457-467. 111. Bartlett JG, Gorbach SL. Anaerobic infections of the head and ncck. Otolarygol Clin North 1976:9:655-678. 112. Sawyer R. Clinical implications of metronidazole antianaerobic prophylaxis in major head and neck surgical procedures. Ear Nose Throat J 1988;67:655-662. 113. Robbins KT. Byers RM, Cole R. Fainstein V. Guillamondegui OM, Schantz SP, Weber RS, Wolf P, Goepfert H. Wound prophylaxis with metronidazole in head and neck surgical oncology. Laryngoscope 1988:98:803-806. 114. Brook I. Hirokawa R. Microbiology of wound infection after head and neck cancer surgery. Ann Otol Rhino1 Laryngol 198?:98:323-325. 1 15. Coleman JJ. Complications in hcad and ncck surgery. Surg Clin North Am 1986;66:149-167. 116. Tandon D A , Bahadur S. Laldirla HC, Sundaram KR. Role of prophylactic antibiotics in surgery for advanced head and neck canccr. J Laryngol Otol 1992;106:621-624. 117. Robbins KT. Favrot S. Hanna D , Cole R. Risk of wound infection in patients with head and neck cancer. Head Neck 1990;12:143-148. 118. Brown BM. john so^^ JT, Wagner RL. Etiologic factors in head and neck wound infections. Laryngoscope 1987;?7:587-590. 119. Hussain M. Kish JA. Crane L. llwayda A. Cummings G, Ensley JF. Tapazoglou E. Al-Sarraf M. The role of infection in the morbidity and mortality of patients with head and neck cancer undergoing multirnodal therapy. Cancer 1991:67:716-721. 120. Papac RJ. Medical aspects of head and neck cancer. Cancer Invest 1985;3:435-444. 121. Weber RS, Hankins P, Kosenbaum B. Rand I. Nonwound infections following head and neck oncologic surgery. Laryngoscope 1993:103:22-27. 122. Jansma J. Vissink A. Spijkervet FKL. Roodenburg JLN. Panders AK. Vermey A, Szabo BG. Johannes's-Gravcnmade E. Protocol for the prevention and treatment of oral sequelae resulting from head and neck radiation therapy. Cancer 1992;70:2171-2180. 123. Barber G R , Miransky J, Brown AE. Coit DG, Lewis FM, Thaler HT, Kiehn TE, Armstrong D. Direct observations of surgical wound infections at a con~prehensivecancer center. Arch Surg 199.5;130:1042-1047. 124. Dillman RO. Herndon J. Seagren SL, Eaton WL, Green MR. Improved survival in stage I11 non-small-cell lung cancer and leukemia group B (CALGB) 8433 trial. J Natl Cancer Inst 1996:88:1210-1215.

5. Role of the clinical microbiology laboratory in the diagnosis of infections Richard B. Thornson, Jr. and Lance R. Peterson

1. Introduction The role of the microbiology laboratory and microbiologist are especially important to the clinician caring for a patient with compromised host defenses. The microbiologist can assist in establishing a differential diagnosis and selection of laboratory tests to make an infectious diagnosis. Complete understanding of microbiology test results not only improves patient management but reduces the cost of medical care.

2. Infections associated with specific immune deficiencies Abnormalities of the immune system that affect T lymphocytes, B lymphocytes, granulocytes, or a combination of these cell functions predispose an individual to specific infectious diseases [I]. In addition, splenectomy, diminished serum complement, organ transplantation, and the use of corticosteroid or cytotoxic therapy further depress immune function and predispose one to infection. Table 1 summarizes the types of immunosuppressive impairments often observed in patients with cancer and the associated microorganisms.

3. Selection, collection, and transport of high-quality specimens, and interpretation of stain and culture results The initial evaluation of a cancer patient suspected of having an infectious disease involves collection of specimens to detect disseminated (e.g., blood) as well as localized (e.g., urine or sputum) disease. It is important to understand why a laboratory may "reject" a specimen, because this risks a delay in establishing a specific diagnosis. Especia.11~in infectious diseases, the likelihood of establishing a diagnosis is directly related to the quality of the specimen submitted for analysis. A clinical specimen that is contaminated with normal flora that overgrow important pathogens, is exposed to prolonged transport, contributing to the Gary A. Noskin (ed), M A N A G E M E N T O F INFECTIOUS COMPLICATIONS IN C A N C E R PATIENTS. 0 1998. Kluwer Acadenzic Publishers, Boston. All rights reserved.

Table 1. Selected immunosuppressive impairments and associated microorganisms

Microorganisms Defect

Bacteria -

Impaired antibody formation

Impaired cellmediated immunity

Reduced granulocyte function

Decreased serum conlplenlent

Fungi

Viruses

Parasites

Enteroviruses

Ginrdin

-

Pneumococcus H. irzflrlenzae Meningococcus Gastrointestinal pathogens Listeria Nocardia Snlnzonell~r Mycobacteria S. ailrezls P. aencgir~osa Other gram-negative bacilli Nocardill Pneumococcus H, iq41~enzae Meningococcus

C. r~eo,forr?~rmsHerpes viruses. H. crrysrdlatr/in esp. CMV B. dernzatifidis C. inznzitis Ptzei~nzocystis Candidn Asyergillzls

Toxoylasrna Strongyloide~

death of fastidious bacteria, or is representative of material overlying but not within a pathologic process generates results that mislead, delay, or prevent proper diagnosis and therapy.

3.1 Blood Blood cultures are one of the most important high-volume microbiology specimens. Collection practices that impact on obtaining reliable culture results include volume of blood per specimen, number of specimens, specimen collection from different sites, and proper skin antisepsis prior to venipuncture (Table 2) [2]. The volume of blood collected to diagnose bacteremia is a critical determinant. Every study addressing the issue of blood volume concludes that more blood is better. The usual bacteremic adult is intermittently bacteremic with fewer than one bacterium per milliliter of blood; in fact, 20% have fewer than 1 microbe per 10mL of blood. A czllture set is optimal when inoculated with at least 20mL of blood. Each culture, referred to as a culture set, represents a set of two or three bottles inoculated with the blood specimen from a single venipuncture. Higher volumes, such as 30mL, improve recovery of pathogenic bacteria but are less commonly used because of perceived difficulties with large-volume collection and the potential for overphlebotomizing patients. However, three sets containing 20mL each are equivalent in volume to two sets of 30mL, and this volume per day appears optimal for detection of

Table 2. Collection of blood for culture Type of culture Adult stat bacterial Adult routine bacterial

Geriatric bacterial Pediatric bacterial

When ordered Acute febrile episode Antimicrobials to be started or changed immediately Non-acute disease Antimicrobials will not be started or changed immediately Altered mental status vomiting, >6% bands Acute, non-acute febrile episode

No. of cultures

Volume of blood

2

20mL each culture

Collected immediately, before antimicrobials started

2-3

20mL each culture

2

20mL each culture

With 8-12 h Intervals no closer than 1 h, drawing at time of fever spike not routinely necessary Same as stat or routine above

2

1-20 mL each culture, depending on weight of patient

Timing

Immediately?

bacteremia. A review of blood volume studies concluded that each additional milliliter of blood collected increased significant positive cultures by 3% [3]. The most common contaminants of blood cultures are the coagulasenegative staphylococci, which reside on the skin and, in spite of adequate antisepsis, contaminate 1-3% of all blood culture phlebotomies. Two separate blood cultures can be used to evaluate the potential significance of an isolate of coagulase-negative staphylococcus [4]. Significant bacteremia with a coagulase-negative staphylococcus usually develops from an intravascular source, resulting in a continuous bacteremia with all cultures positive. Therefore, a single positive blood culture for a coagulase-negative staphylococcus is less likely to be significant than a strain of coagulase-negative staphylococcus isolated from two or three separate cultures. The accepted recommendation for number of blood cultures is two to three per day from separate venipuncture sites. This takes into consideration the relatively large blood volume needed to improve sensitivity and the need for more than one culture to evaluate staphylococcal contamination. More than three cultures in 1 day are rarely needed and should not be ordered unless there is a significant change in the patient's clinical status. Proper skin antisepsis is required to prevent excess blood culture contamination. At a minimum, skin antisepsis should include the use of an iodinecontaining compound that is allowed to dry before blood collection. Tincture of iodine dries more quickly than iodophor preparations and may represent a more useful antiseptic in a busy clinical environment [5]. Transport of blood collected for culture from the patient's bedside to the laboratory should be rapid to ensure immediate incubation and shortened detection times for positive samples. Blood should be collected directly

into culture bottles and held at room temperature until received by the laboratory. Test-of-cure blood cultures, or those collected after diagnosis and treatment to document clearing of the bacteremia, are rarely necessary. In general, monitoring the clinical progress and ordering additional blood cultures only when the patient does not respond properly is preferable. However, the technique of repeat blood cultures on the day following a documented episode of sepsis can be helpful in diagnosing a suspected intravascular infection (endocarditis or catheter-associated bacteremia), because only rarely does bacteremia due to an extravascular source persist beyond 24 hours. In addition, patients who have failed therapy, who have a difficult bacterium to eradicate (e.g., Brrrcella spp.), or who have disease caused by multiply resistant bacteria (e.g., vancomycin and ampicillin-resistant enterococci) may require a test-of-cure culture because data are not available to indicate what treatment regimens are likely to be successful.

3.2 Respiratory specimens Transport of all lower respiratory tract specimens must be rapid. Overgrowth of potential pathogens by contaminating bacteria can occur within 1-2 hours at room temperature and the death of fastidious pathogens, such as S. pneumonine, can occur in even less time. Specimens should be stored at room temperature and must arrive in the laboratory within 2 hours of collection. The three most common respiratory tract specimens received by the microbiology laboratory are expectorated or induced sputum, endotracheal aspirates, and bronchoalveolar lavage (BAL) fluid. Ideally, all three specimens represent bronchial and alveolar secretions that are largely free of normal oral bacterial flora. Most etiologies of community- and hospital-acquired lower respiratory tract disease are oral flora that colonize the patient at the time of aspiration and tracheal inoculation. For this reason, the mere presence of potentially pathogenic bacteria does not definitively establish the cause of disease. Respiratory specimens can provide clinically useful information from staining and culturing if screened for gross oropharyngeal contamination, indicated by elevated counts of squamous epithelial cells, and rejected if contaminated. Purulent, bloody, or tenacious portions of sputum are used to prepare a smear and to inoculate culture media. Comparison of sputum culture results with those of transtracheal aspirates and bronchoscopy specimens confirms that useful information is gained only from those sputums that contain low numbers of squamous epithelial cells. Most laboratories use either <10 or <25 squamous epithelial cells per 10x microscope objective field to indicate a specimen with minimal contamination, whereas >I0 or >25 squamous epithelial cells indicates that culture of the specimen is colonized with oropharyngeal flora and will likely provide no useful information. In some laboratories, acceptable sputum specimens are identified by both low numbers of squamous

epithelial cells and elevated counts of polymorphonuclear leukocytes. The increased complexity of screening based on two cell types may not be justified because studies have shown squamous epithelial cell screens are equivalent to screening procedures utilizing squamous epithelial cells and white blood cells. In addition, severely neutropenic patients may have an infected lower respiratory tract with minimal oropharyngeal contamination that is devoid of inflammatory cells. Gram stains of acceptable sputum specimens should have inflammatory cells and predominating or intracellular bacteria quantitated and reported [6]. Identification and reporting of potential pathogens detected in culture should include those predominating over normal respiratory flora or those with morphotypes noted in the Gram stain as predominant or intracellular . Endotracheal aspirates can be screened and cultured, and results interpreted, in a manner similar to sgutums, with one exception. Any endotracheal aspirate that shows no bacteria in the Gram-stained smear does not require a bacterial culture. Studies have demonstrated that cultures of Gram stain"negative" aspirates do not provide information that is clinically useful, even from neonates [7]. Bronchoalveolar lavage specimens provide useful information when processed using both the Gram-stained smear and a semiquantitative culture. A cytospin smear preparation of lavage fluid should be examined for inflammatory cells and bacteria. If more than 1% of cells present are squamous epithelial cells, excessive contamination with upper respiratory tract material has occurred and relative numbers of bacteria are not helpful. If very few ( < I % ) cells are squamous epithelial cells, the quantity of neutrophils and bacteria should be noted. Intracellular bacteria (within neutrophils) should also be reported. Potential pathogens and obvious contaminating oral flora should be quantitated and reported. Bronchoalveolar lavage smear and culture results should he interpreted according to Table 3 [S]. Using a "protected" catheter to obtain the bronchoalveolar lavage (BAL) specimen can avoid the problem of upper airway contamination and is recommended [9]. Specimens obtained using a protected bronchial brush also may be processed quantitatively. However, they are not as useful as lavage specimens due to the small amount of material obtained with this technique. One exception is in the diagnosis of invasive fungal (i.e., Aspergillus) pneumonia. In this setting a bronchial brushing is likely to be positive even when a lavage specimen is negative. 3.3 Urine specimens Urine is considered an easy specimen to collect but, because it is easily contaminated, interpretation of results may be difficult and misleading. Three common specimens received by the microbiology laboratory include midstream urines, straight (removed) catheter urines, and urine from indwelling catheters. Difficulties encountered when collecting urine by each

Table 3. Guidelines for interpretation of bronchoalveolar lavage fluid smear and culture results

Direct Gram stain

Culture

Interpretation

Elevated number of squamous epithelial cells (SEC), i.e., >1% of all cells are SECs

Any quantity of potential pathogen and respiratory flora

< I % of all cells are SECs;

<10,000 potential pathogeninll in unconcentrated BAL fluid 10.000-100,000 potential pathogenlml in unconcentrated BAL fluid Quantity of respiratory flora less than potentiaI pathogen >100,000 potential pathogens1lnL in unconcentrated BAL fluid Quantity of respiratory flora less than potential pathogen

BAL fluid likely contaminated with orophryngeal flora or aspiration pneumonia; many PMN's with mixed intra and extracellular respiratory flora suggests aspiration Does not suggest disease by routine (cultivatable) bacterium Significance of potential pathogen indeterminate: intracellular bacteria in direct Gram stain suggest potential pathogen causing disease

PMNs may or may not be present < I % of all cells are SECs; PMNs usually present

<1% of all cells are SECs; PMNs usually present

Suggests potential pathogen causing disease, especially if intracellular bacteria in direct Gram stain

Table 4. Guidelines for interpretation of urine culture results Specimen type

Patient

Midstream urine

Female - cystitis Female - pyelonephritis Female - asymptomatic Male All

Straight (removed) catheter urine Catheter (indwelling) urine

All (only culture if symptomatic)

Urine culture significant colony count > I 0' potential pathogenslmL

>10' potential pathogens1mL >lo' potential pathogenslmL >I@ potential pathogens1mL Any number (usually >lo3) potential pathogenslmL >los potential pathogenslnll

method and the pathogenesis of urinary tract infection in a variety of hosts necessitate different interpretative criteria for urine cultures (Table 4) [lo]. Urine collected by any method requiring transit through the urethra is likely to be contaminated. Most bacteria divide rapidly in urine at room temperature. Refrigeration will stabilize colony counts for several days. Boric acid-containing tubes are commercially available for transportation at room temperature when transit and processing times are expected to be greater than 1 hour. Colony counts of bacteria in boric acid are stabilized without altering viability for culture. Urinalysis can be performed on specimens from the boric acid tubes for at least 24 hours after collection. Any urine specimen can be evaluated rapidly using a cytocentrifugation method. The cytospin is an instrument that uses high-speed centrifugation

and capillary action to concentrate cells in a smear approximately l c m in diameter. For all patient groups except outpatient women, sensitivity is great enough that culture is not necessary when a negative Gram stain is reported [Ill. 3.4 Cerebrospinal fluid specimens Detection of microorganisms in cerebrospinal fluid (CSF) is a medical emergency. Rapid interpretation of a stained smear, combined with immediate inoculation of specimen to culture media, is essential. Bacteria that infect the meninges, such as N. meningitidis, S. pneumoniae, and H. influenzae, are fastidious, requiring room temperature transportation and rapid inoculation onto culture media to ensure recovery. The issue of which tube, of the multiple tubes collected during the lumbar puncture, to use for culture is less important than sending a sufficient volume of specimen to ensure detection. It is vitally important to have sufficient fluid for all the microbiology tests ordered, bacterial (1.5mL), fungal (5mL), mycobacterial (5mL), and viral (1 mL). Many of these infections involve a basilar meningitis and unless sufficient volume is submitted for processing, a false-negative result is likely, which may complicate proper diagnosis. Antigen tests have been used for more than 15 years to detect bacteria in body fluids, but, unfortunately, have proven themselves to be of limited value [12], Antigens of H. influenzne, N. meningitidis, S. pneumonine, and group B Streptococcus can be detected with fair sensitivity and high specificity. However, evaluations have shown that performance is similar to that of the Gram stain, and when an antigen test is positive changes are infrequently made in patient management, leading only to increased medical costs without improvement in patient care. At this point, antigen test results offer very little to the diagnosis of infections in patients with cancer and should be used sparingly or not at all.

3.5 Gastrointestinn1 specinzens Fecal specimens and endoscopic bowel biopsies are common and useful specimens for the diagnosis of gastrointestinal infections in cancer patients. Spontaneously passed stool or endoscopically collected aspirates are used to document the bacterial or parasitic etiology of infectious diarrheas, whereas biopsies collected during endoscopy are used to detect viral causes of gastrointestinal disease. Infectious causes of gastrointestinal disease are listed in Table 5. Freshly passed stool should be processed and cultured within 1 hour. Clostridium dificile is nearly the exclusive cause of diarrhea acquired within the hospital, especially in patients with prior antimicrobial therapy. This exclusive association suggests that stool cultures for other bacterial causes of diarrhea are not warranted if diarrhea begins in the hospital [13].

Table 5. Usual infectious causes of gastrointestinal disease Etiologies -

-

Setting

Bacteria

Parasites

Viruses

Travel

E. coli Salmonella Carnpylobacter A eromonas Carnpylobacter Salmonella Shigella E. coli (enterohemorrhagic) C. dificile C. dificile Same as above plus M. avium

Giardia Cryptosporid ium Cyclospora Amoeba Giardia

Unknown

Outpatient

Inpatient Compromised

Same as above plus: Microsporidium Isospora Biastocystis hominis

Rotaviruses (pediatric)

Rotaviruses (pediatric) CMV

Small and large bowel biopsies are used to detect cytomegalovirus (CMV) disease. CMV-positive cultures of washes or feces have poor predictive value for active disease. Histologic confirmation, which includes finding CMV inclusions within epithelial lining cells, is strongly suggestive of significant gastrointestinal disease. Biopsy tissue for viral culture should be transported to the laboratory in sterile viral transport medium (VTM). 3.6 Wound and cutaneous specimens

Wound and cutaneous specimens are very helpful when collected aseptically from a "deep" site. Material swabbed from the skin or wound surface is frequently colonized with contaminating bacterial flora and should be avoided. Wound specimens are best collected by aspiration or biopsy, as a way to bypass the contaminated surface. Irrigating with or injecting nonbacteriostatic saline may be necessary to ensure sufficient sampling volume. If anaerobic culture is needed, transport must occur in an oxygen-free container. Small pieces of tissue and purulent fluid should be added or injected into a transport vial or tube. Swabs are difficult to maintain in an oxygen-free environment, easy to contaminate with normal flora, may actually "trap" or kill the bacteria causing an infection, and sample too little material. Therefore, swabs as collection devices for wound and cutaneous specimens should not be used. 3.7 Intravascular catheter specimens

Skin and soft-tissue infections are confirmed by culturing purulent material or tissue from the cutaneous and intracutaneous area around the outside of the

catheter. Documenting the catheter as the source of a bacteremia is more difficult. When the catheter is removed, the distal 2 inches of the intravascular portion can be cultured by one of several methods to assess the catheter as the likely source of bacteremia. The method of Maki and colleagues is the simplest, requiring that the catheter be rolled on an agar plate [14]. Less than 15 colonies of a single isolate suggests the catheter is not the source of the bacteremia. Greater than 15 colonies, and in most cases greater than 100 colonies, suggests the catheter is the source of the bacteremia, but requires a positive blood culture to confirm the association. Alternatively, sonicating the catheter tip, a method that samples both the external and internal surfaces, improves the detection of infected catheters by approximately 10%. The sonicate is then quantitatively cultured. Less than 1000microorganisms per milliliter of a single isolate suggests the isolate is not causing the bacteremia [15]. Transport of catheter segments to the laboratory should occur rapidly and in a sterile container. As the catheter surface dries, adherent bacteria can die. Bacteria also continue to multiply in the moist slime on the catheter surface. Therefore, in order to ensure accurate colony counts and viability, transport should be <1 hour. Evaluating a catheter in situ is more difficult. Measuring the time to positivity of a blood culture is based on the assumption that a true positive bacteremic isolate would be in relatively high numbers and would result in the blood culture becoming positive quickly, that is, within 1-2 days. On the other hand, contaminants in low numbers would require 3 or more days before detection. While this method is frequently used clinically, it is crude and subject to many exceptions. Evaluation of an existing catheter is accomplished best by performing simultaneous cultures of blood collected from a venipuncture and the catheter port. If only the catheter specimen is positive, a contaminated collection hub or needle insertion site may be the cause. If both blood specimens are positive with the same organism, a true bacteremia is likely, but it is not known whether the catheter is the source or not. If quantitative cultures using the Isolator lysis centrifugation blood culture system are used, the catheter is likely to be the source if counts of bacteria isolated from the catheter are >5 times the counts of the venipuncture specimen [16].

4. Methods for the diagnosis of infections caused by unusual or fastidious etiologies There are many organisms that require special culture media or laboratory preparation to optimize recovery. If clinical microbiology laboratory techniques cannot be optimized, then false-negative cultures may result. 4.1 Legionella Legionella spp., especially L. pneurnophilia, are causes of both communityand hospital-acquired pneumonia. The laboratory diagnosis of pneumonia

,

Table 6. Summary of Legionella diagnostic tests

Test

Sensitivity

Specificity

Turn-around time

Culture Fluorescent antibody stain (DFA) of smear of respiratory specimen Serology

40-100% 25-75%

100% 95-99%

3-5 days Hours to 1 day

25-75%

95-99%

Urinary antigen"

40-75%

95%

Weeks to months (convalescent specimen needed) Hours to days (send out)

"Detects disease caused by L. pnez~rnophilnserogroup 1 only.

caused by the legionellae can be accomplished by detection of serum antibody, fluorescent antibody staining, and culture of respiratory secretions. A test for detection of legionella antigen in urine is also available, with comparative studies suggesting that it performs as well as fluorescent antibody staining or culture. A summary of available legionella diagnostic tests occurs in Table 6. Although a positive culture provides definitive identification of disease, it is not highly sensitive unless the sputum undergoes careful preparation (washing) with plating to several media in the laboratory. In addition, positive cultures require 3-5 days to be recognized. The direct fluorescent antibody (DFA) test is the most rapid test for the diagnosis of legionellosis but lacks both sensitivity and specificity. It should not be done in the absence of culture as an accompanying test. Serology requires testing of acute and convalescent serum specimens, necessitating a delay of several weeks to months before a significant increase in antibody is detected. Detection of urinary antigen is a rapid and very specific test but has not been widely offered because of its use of radioisotopes. Fortunately, an enzyme-based test is now commercially available [17]. DFA testing and culture of specimens obtained during bronchoscopy or by lung biopsy are the most rapid and accurate diagnostic approaches, and are recommended when legionnaires' disease is strongly suspected.

4.2 Bartonella The bartonellae are a large group of organisms, with a limited number recently recognized as causing disseminated infection in compromised hosts, including those not infected with HIV. Reports suggest that B. henselne, and possibly B. quintann and B. elizabethae, can be recovered from blood using automated blood culture methods, acridine orange staining, wet mount testing for motility, and culture on blood agar following lysis centrifugation (Isolator) treatment [18]. Figure 1 shows a culture and detection algorithm that can be used by most laboratories. Here, too, consultation with laboratory personnel is necessary to determine when blood submitted for culture should be processed by this more laborious pathway.

Blood. tissue, or fluid specimen

Automated Blood Culture System Inoculate up to 5 m L of specimen into aerobic broth medium Incubate 14 days with rockinglagitation

1

Positive growth reading

Wet mountlphase microscopy Use oil immersion objective to examine for rachety motility

Stain one drop of concentrate with acridine orange. Bartor~ellaare small pleon~orphic bacilli

Process blood from positive bottle by injecting into IsoIator (Wampole Laboratories) tube

Inoculate concentrate onto chocolate and sheep blood agar plates Incubate taped plated at 36' C 0 2for 1 month or until growth is evident

Figure I . Proposed algorithm for culture and identification of Bartonella spp. The Automated Blood Culture System is a continuous monitoring system, including BacTiALert (Organon Leknika Corp.), Bactec 9000 Series (Becton Dickinson), and ESP (Difco Laboratories).

4.3 Capnocytophngin Capnocytophagia spp. are common inhabitants of the oral cavity, and are also well-documented opportunistic pathogens in patients with impaired host responses. C. sputigena, C. ochracea, and C. gingivalis can cause sepsis in patients with granulocytopenia [19]. Detection is relatively easy using normal laboratory media, including usual blood culture broth formulations. 4.4 Gmm-positive bacilli

Gram-positive bacilli that can act as opportunists should be suspected in cancer patients with undiagnosed infections [20]. These include Listericr rnoaocytogenes, Nocardia asteroides complex, Rhodococc~lsequi, and Bacillus cereus. Although these microbes grow on usual culture media, an infectious diagnosis may be missed if they are ignored as contaminants.

4.5 Mycobacteria Mycobacteria are ubiquitous environmental bacteria and can result in serious disease in patients with cancer. In spite of the severity and continued presence of Mycobacterium tuberculosis disease, species of mycobacteria other than tuberculosis (MOTT) are responsible for more disease in most laboratory settings [21]. Table 7 summarizes the most pathogenic mycobacteria, including the type of diseases reported along with the hosts' immunocompromising conditions. Molecular detection of M. tuberculosis by polymerase chain reaction (PCR) techniques can aid in the rapid detection of infected patients. Although PCR detection is less sensitive than culture and, therefore, should not replace traditional detection methods, it can be used to rapidly confirm the presence or absence of M. tuberculosis in specimens that contain acid-fast bacilli in the direct smear. Settings where the expense of this additional diagnostic test are justified, which can run between $150 and $200 in direct laboratory cost when a single specimen is tested on an urgent basis, must be established for each practice, institution, and laboratory. Table 7. Pathogenic mycobacteria other than Mycobacterium tuberculosis Mycobacterium M. bovis

Disease

Host

Pulmonary, extrapulmonary, Normal to severely bladder, or, rarely, severe compromised complications following intravesical inoculation to treat superficial bladder cancer Leprosy Not endemic in U.S.A. M. leprae Photochromogens (colony pigment produced following exposure to Iight) M. kansasii Pulmonary with dissemination, Normal or compromised skin, bone M. rnarinlinz Skin, soft tissue, musculoskeletal Normal M. sinziae Pulmonary Normal or compromised Scotochromogens (colony pigment produced in light and dark) M. szulgai Pulmonary, musculoskeletal Normal or compomised M. xenopi Pulmonary, disseminated Generally infects patients with underlying respiratory disease M. scroj%laceum Lymphadenitis, lung, disseminated Normal or compromised M. rhernzoresistible Pulmonary, skin Normal or compromised Nonchromogens (no colony pigment) M. avium-intmcellulare Pulmonary, disseminated Normal or compromised (MA11 M. ~ ~ l c e r a n s Skin, soft tissue Not endemic in U.S.A. M. malnzoense Respiratory, skin and disseminated Normal or compromised less common M. haernophilirrm Skin and subcutaneous tissues Normal and compromised M. genavence Multisystem disease AIDS Rapid growers M. fortuitum complex Skin, soft tissue, lung, disseminated Normal or compromised M. fortuitum M. chelonae M. nbscesus

4.6 Fungi Most fungi grow readily on a wide variety of media if incubated for a sufficiently long period of time. It is important to remember that yeasts such as Candida, Torulopsis, and Cryptococcus spp. should be reported when detected from clinical specimens. Other fungi capable of causing invasive disease, such as Histoplasma, Blastor?zyces, Coccidioides, Rhizopus, and Aspergillus, are not routinely detected by bacterial culture. A specific fungal culture request should be used to ensure 3-4 weeks incubation and the use of selective media to prevent bacterial overgrowth for these latter organisms. This includes blood cultures specified for the detection of fungi. Fungi are strict aerobes and grow best on the surface of an agar plate, not in a broth medium away from oxygen. The Isolator lysis centrifugation blood culture system, which uses a blood concentrate inoculated to the surface of agar plates, has been found to be the best all-around method for detecting yeast and fungi in blood [221. However, automated, continuously monitoring bacterial blood culture instruments are reliable for the detection of Candida and Torulopsis species causing a fungemia. 4.7 Viruses The commercial availability of high-quality reagents and cell cultures for the detection of viruses allows many clinical laboratories to rapidly and reliably detect pathogenic viruses. Viral diseases in cancer patients may be thought of as those that reactivate in the immunocompromised host, such as the herpes viruses, and those that are usual community pathogens, such as influenza and adenoviruses. The quickest method for virus detection is direct staining of a clinical specimen using an antibody specific for the virus suspected that is conjugated to fluorescein isothiocyanate (FITC). In the appropriate setting, fluorescent antibody staining to detect influenza virus, respiratory syncytial virus (RSV), adenovirus, herpes simplex virus (HSV), cytomegalovirus (CMV), or varicella-zoster virus (VZV) can quickly establish a diagnosis. Viruses that are not cultivatable in normally available cell cultures can be detected by molecular methods, such as the PCR test. High detection sensitivity for viral nucleic acids by PCR must be balanced against the high cost and apparent detection of latent viral genome in situations where no clinically significant viral disease is present. At the current time it appears that PCR is the test of choice for detection of herpes group viruses (HSV, VZV, CMV) in spinal fluid. Its utility in other settings is less well defined. Isolation of viruses in cell culture is the standard laboratory method for detection [23]. Conventional cell culture, which requires the appearance of a cytopathic effect (CPE) in the infected cells, takes 1 day to 2 weeks before the presence of a virus can be confirmed. Rapid cell culture, using the shell vial method, grows and detects viruses within 1-2 days. Shell vials are more rapid

because fluorescent antibody staining is used to detect virally encoded proteins produced within hours of host cell infection. Conventional culture is the only available method where viable virus is recovered that can then be used for susceptibility testing or epidemiologic investigation. Table 8 summarizes detection methods for common viruses.

4.8 Parasites In addition to the traditional list of pathogenic animal parasites, such as Entamoeba histolytica, Giardia lamblia, and the nematodes, an increasing number of opportunistic gastrointestinal and systemic pathogens are being identified in the immunocompromised host [24]. Table 9 summarizes these opportunists, especially those that have been recognized within the last decade. To ensure that all parasitic forms are adequately preserved, a specific transport medium for parasitic analysis should be used if transportation time will exceed 1 hour. In addition, stool specimens containing barium are unacceptable for examination for 5-10 days after barium is given to the patient. Although most parasites, which are opportunistic gastrointestinal pathogens, are detected by examining a single diarrheal stool specimen, multiple specimens may occasionally be necessary. It is important to notify the laboratory of unusual pathogens that may be suspected or underlying diseases that can predispose the patient to infrequently encountered parasites. 4.9 Pneumocystis carinii Pneumocystis carinii is a pathogen of major importance in many centers treating immunocompromised hosts. While infection is common is patients with HIV, Pne~~rnocystis carinii pneumonia can also occur in patients with leukemia and following bone marrow transplantation. Recent studies have shown the high sensitivity of a DFA stain applied to properly collected, induced sputum of sufficient volume [25]. Under these circumstances, and when disease is highly suspected, the sensitivity of DFA with a single specimen exceeds 90%. A negative stain result should suggest the need for bronchoscopy to directly obtain pulmonary specimens for laboratory analysis.

5. Tests for rapid detection of microorganisms The best practice of medicine requires prudent use of medical resources, including laboratory testing. The typical turn-around times and relative laboratory costs for common microbiology tests are detailed in Table 10. It is important to note that many of the most rapid tests for diagnosing infectious disease still involve direct microscopic inspection of the specimen from a suspected site of infection. Commonly used stains are summarized in Table 11. The Gram-stained slide can provide rapid and useful information within 30-60

^ (^

u


<

3

3 U

3

la W c c c

C

rt

rt

c^

c/::

«™ < <<< tlH U^ tL. t/D

Q

t/3

.o-g

3

<

PLH

u 3

3o

^

W

a: oiai,c^Q£.

u Poi u Cl^ L) u OH

(1H

CL.

N

N

i-

>-

> >,

5 ^ 3

H w;

_o •Q

< <'5 fc [ i , U

C O

C O

C O

3 O


U

(D

(U

(U

v:i

M V2

LO VJ

V5 '^i

V^ v^

i j

V-c

t^

(-

U
CJ O OJ O

C

3

d

C

C

TJ '35

:/3

>

lo

{/3

c/5

CO

c/D

CO CO CO CO CO

^ ^ ^~> ^^ ^~^ JO O O O O u

;-<

;-H

i-<

c3

CO

rt

rt

^U

I-

l-i

_VH

O - O H D_

CL,

"U

(U

O

(U

VH

VH

U

Cd

CO

CO

» ^ »^ 1/3

C3

CO

C3

y:

(/)

CC C3

o o o o ji;

jr

^

ji;

HHHH

itl

•Q-CQ

•5<

U<

Ui

U

3 0b ^

1-

^

k^

CO CO tU 1)

;-<

CO CO CO

CO CO

G

eE

3

C

C

G

C

CO CO

C

a cx cu & D. c/5

y]

0 0V H U X Xi U i> l^ 1_ U tD

c/3

V)

0^ H 0 U 0 1 - , X) X) -C ^ 1-1 0 l-H0 U U 0 (D

i:^.i= (1) -C*

-6

03

^ [^ s

^

0

w u

-a 0 _o

3

(U

S

^ 3 I- F O^ X 1)

XT' cu t^

(D

o p "

rt)

ao s

a OJ

3

0

u u > !>n

0 N

E>

,

"co

:A

^aFi s I > =pE cO

0 CD

'3 03

yi

uuu0u

to T :

1 < Cui CL,

C3

X) X) X) JO

O O 0 O O T3 T3 T3 T3 JO •3 ' 3 ' 3 ' 3 ' 3 q3 C !3! T : C

)-H

Jli X> X l J3 CO

)^

U

<

t/)

U CD 3 3

UH

CJ OJ

Wi

>

.3

c c CO 03

•"> ^

.b

>

o

^ ?

5 > -^ i I

(U CO

p:;>

> 157

Table 9. Opportunistic parasites Parasite

Disease

Specimen

Diagnostic test

Cryptosporidium parvurn Cyclospora cayetanesis Isospora belli

Diarrhea Diarrhea Diarrhea

Stool Stool Stool

Sarcocystis spp. Microsporidia Acantharnoeba spp.

Diarrhea Diarrhea Cutaneous, neurologic, nasal Hyperinfection syndrome

Stool Stool Biopsy (tissue) Stool, sputum

Acid-fast or fluorescent stain Acid-fast stain Wet mount of stool concentrate and acid-fast stain Wet mount of stool concentrate Modified trichrome stain Hematoxylin-eosin stain, culture Microscopic examination

Strongyloides stercoralis

Table 10. Turn-around times (TAT) and relative laboratory costs for common microbiology tests Test Stains Gram, acid-fast, Calcofluor white (fungi) Fluorescent antibody Trichrome (parasites) Antigen detection Latex agglutination Enzyme immunology assay Molecular (nucleic acid) detection DNA probe PCR~LCR' Culture Bacterial Mycobacterial Fungal Viral

TAT

Relative cost"

2-6 h 6h to 2 days 1-4 d 2-4 weeks 1-3 weeks 2 d to 3 weeks

"Approximate cost to laboratory: +, $5-10; + +, $10-20; + + +, $20-30; 'PCR = polymerase chain reaction; LCR = ligase chain reaction.

+ + + +, $30-50.

minutes of specimen receipt by the laboratory. While it appears to be a simple technique, considerable experience is needed to avoid misinterpretation of what is seen. For this reason it is best to view the slide with a well-trained observer in the clinical microbiology laboratory before using the results as the primary basis for initiating specific treatment. Quantitation of inflammatory cells and recognition of bacterial morphology provide a narrowing of etiologic possibilities, and a guideline for selection of empiric antimicrobial therapy. Table 12 summarizes common bacterial microscopic morphologies and their implied etiology. Historically, fungi have been detected using a KOH wet mount examination. In many laboratories a fluorochrome, calcofluor white, is added to KOH solutions to help differentiate fungal structures from specimen debris. Table 13 summarizes common fungal morphologies seen in clinical material.

Table 11. Stains used to demonstrate microorganisms in clinical material Microorganism group

Stain

Comment

Bacteria

Gram Acridine-orange Fluorescent antibody KOH/Calcofluor white Gomori methenamine silver Carbofluchsin based Kinyoun Ziehl-Neelsen Flurochrome based Auramine-rhodamine Iodine wet mount Trichrome Fluorescent antibody Giemsa Fluorescent antibody Giemsa

Gram negative, gram positive Stains nucleic acid Must use organism-specific reagents

Fungi Mycobacteria

Parasites - intestinal

Parasites - blood Viruses

Used to stain fungi in tissue Can be modified to stain Nocardia

Best for ova and larvae Best for protozoa Must use parasite-specific reagents Must use virus-specific reagents Stains viral inclusions

Table 12, Interpretation of Gram stains Observed morphology

Interpretation

Gram-positive cocci clusters Gram-positive cocci chains Gram-positive diplococci Gram-negative diplococci Gram-negative coccobacilli Gram-positive rod diphtheroid Gram-positive rod boxcar Gram-positive rod branching Gram-positive rod Gram-negative rod

Staphylococci Streptococci Pneumococci Nei~seria!Moraxella Haernophil~is Cor~nebacteri~~rnlPropionibacteri~~nz Clostridi~smiBacill~~s Nocardia!Actinomyces Other gram-positive rods Enteric and pseudomonas-like Gram-negative rods and Cry,vtococcus) Yeast (e.g., Cnnclidn, Tor~~lopsis, Candida Mold (e.g., Aspergillus)

Yeast cells Yeast cells with pseudohyphae Hyphae

Acid-fast stains are used to demonstrate the presence of mycobacteria. The main reason for low sensitivity of stains for mycobacteria is that these infections are often present with a low number of organisms, and even with appropriate concentration of the collected material, false-negative results occur in 50% of cases. Most parasitic worms (adults, larvae, and ova) are large enough to be detected without the use of special stains. In contrast, protozoans in blood and most protozoans in stool need staining to be seen. Direct staining of clinical material for viruses makes use of the Giemsa and fluorescent antibody stains. Fluorescence staining, using antibody specific for

T a b k 13. Identification of fungi seen in stained smears of clinical material Fungus

Morphology in clinical material

Candida, Tor~ilopsis, Cryptococcus, and other yeasts Aspergillus, Fusariunz, and other nonpigmented molds Rhizopus and other related Zygomyceies Exophilia, Curvlilaria and other brown-pigmented molds

Budding yeast cells; in addition, Canrlida spp. produce pseudohyphae and cryptococci produce capsules

Histoplasma capsulatum Blastomyces dernlatitidis

Hyaline (clear), septate hyphae Wide, ribbon-like, nonseptate, hyaline hyphae Dematiaceous (brown pigmented) septate hyphae, grains or compact masses of dematiaceous hyphae, or muriform (cells divided by fusion plane) Small budding yeasts (intra- and extra-cellular) Budding yeast with broad neck between parent and daughter cell Large spherules containing smaller, nonbudding endospores

a suspected virus, is the more sensitive method and can be used for rapid, specific diagnosis of adenovirus, influenza virus, RSV, HSV, VZV, and CMV in the appropriate specimen. Close communication between the clinical and laboratory services can help optimize diagnostic results, improving speed, accuracy, and overall cost-efficient testing.

6. Determining microbial susceptibility or resistance

Standard in vitro tests used to measure antimicrobial activity are the disk diffusion and the minimum inhibitory concentration (MIC) tests. Both tests measure inhibition of microbial growth as an endpoint. Although bactericidal testing is occasionally needed, the vast majority of experience comparing clinical outcomes to laboratory test results are with inhibitory tests. The MIC test is performed in antimicrobial-containing broth and is defined as the lowest concentration of antimicrobial (expressed in micrograms per milliliter) that inhibits growth of the test organism. The disk diffusion test can be thought of as an MIC test on the surface of the agar plate. Antimicrobial concentration is greatest close to the disk and least away from the disk, where less drug has diffused. Results of both tests are correlated with the breakpoint of the antimicrobial in question, resulting in a report of susceptible or resistant. The breakpoint can be thought of as the concentration of antimicrobial that can be safely achieved after a normal dose in a patient. Breakpoints for various antimicrobials are different because of dosing and pharmacokinetic variables. If the MIC result is lower than the breakpoint, the antimicrobial is reported as susceptible. If the MIC is greater than the breakpoint, the result is resistant. MIC results that fall near the breakpoint may be reported as intermediate. Susceptible implies that the microorganism tested will be inhibited by normally achievable levels following a standard dose. Resistance implies that

inhibition should not be expected. Because intermediate results imply the MIC is near the breakpoint, they can be interpreted to mean that the microorganism will be susceptible in areas of the body where the antimicrobial in question is concentrated, such as urine, or that the isolate will be susceptible if higher than normal doses can and will be safely administered. A common error when examining a laboratory report listing many antimicrobials is to assume the lowest MIC value represents the most effective drug to use. 6.1 Antibiograms

Empiric therapy is often initiated in cancer patients, especially those with neutropenic fever. Initial therapy while awaiting culture results is dependent upon an antibiogram, which is a periodic (usually annual) compilation of susceptibility results. All common pathogens and antimicrobials tested by the laboratory are reported as percent susceptible, for example, the percentage of S. aureus susceptible to oxacillin. The antibiogram for every medical center will be different, depending on the resident bacterial flora. Knowledge of the local antibiogram is essential for the selection of proper empiric therapy. 6.2 Specinl susceptibility tests In addition to routine bacterial susceptibility testing, tests that are useful in limited situations include: antimicrobial assays, combination agent testing (synergylantagonism), and serum bacteriocidal testing. Antimicrobial susceptibility testing for microorganisms other than rapidly growing bacteria also may be appropriate in limited circumstances. These tests are those for antimycobacterial, antifungal, and antiviral agents. Antimicrobial assays measure the quantity (measured in micrograms per milliliter) of drug in a fluid, usually serum. Most often, assays are used to avoid toxic concentrations and to optimize therapeutic levels of the antibiotic [26]. Vancomycin and aminoglycoside assays comprise the majority of assays performed by hospital laboratories. Specimen collection and transport must ensure that labile drugs do not degrade, resulting in a falsely low level. Ideally, clotted blood should reach the laboratory within 4 hours of collection. If storage or transport is prolonged, the sample should be placed on ice or refrigerated at 2-8OC. Storage for greater than 3 days should be at -20 or -70°C. Information necessary for proper interpretation and dosing calculations that should be included with the assay request includes the exact time of specimen collection; the exact time of last dose; whether the specimen represents a peak, trough, or random level; what other antimicrobial agents the patient is receiving; and which drug should be assayed. Combination testing refers to measuring the inhibitory or bacteriocidal (killing) effects of testing two antimicrobials at the same time. If together they perform better than predicted by their individual effects, they are synergistic. If their combined effect is less than that predicted by their individual effects,

they are antagonistic. Traditionally, combination testing has been used to document synergy for serious infections such as endocarditis or osteomyelitis [27]. Combination testing may also be necessary if rifampin is to be added to a cell wall-active antistaphylococcal drug. In this instance, antagonism should be measured because reports have documented this combination may be synergistic, antagonistic, or indifferent [28]. In the era of emerging resistance to many of our potent antibacterial compounds, use of these tests may become more prevalent and important when new multidrug resistant microbes make our past clinical response data no longer applicable. The need for serum bacteriocidal testing remains controversial [29]. The test involves collecting serum after the antimicrobial regimen has been selected and initiated. The serum is diluted and a standard suspension of the patient's pathogen, which has been cultured previously, is inoculated to each dilution. Ideally, lower dilutions of the serum inhibit and kill the inoculum. Absence of these effects suggests the therapeutic regimen may be inadequate. Historically, testing was reserved for serious infections requiring prolonged therapy with bacteriocidal drugs such as osteomyelitis and endocarditis. More standardized in vitro tests and experience with clinical outcomes has obviated the need for most serum bacteriocidal testing. An expanded role for less conventional antimicrobial susceptibility testing may arise with the emergence of multiply antibimicrobial resistant organisms in patients with cancer. This is most evident with vancomycin-resistant enterococci, multiply antibiotic-resistant Pseudornonas aeruginosa, and azoleresistant Candida species. Because combination and bacteriocidal susceptibility tests are not fully standardized and lack a national consensus on interpretation of the results, it is important for both medical and legal purposes that trained personnel interpret the results of this testing. This can be done through written consultation by the microbiology laboratory director documenting the medical interpretation of results.

7. Summary

The proper use and interpretation of clinical microbiology test results may be complicated but critical to the care of cancer patients. The microbiology laboratory director is often available to offer advice concerning the differential diagnosis, choice of specimens, as well as the optimal stains and cultures to facilitate diagnosis. Additionally, the rapid interpretation of Gram-stained smears provides useful, occasionally lifesaving, information relative to the etiologic diagnosis and empiric antimicrobial therapy. The microbiology laboratory director should also provide further interpretation of culture and antimicrobial testing results that allow the clinical service to focus on the most critical data. Person-to-person or telephone conversations discussing important laboratory information should be followed up by a written summary

MICROBIOLOGY REPORT

Patient's Name (Last-First-Middle) Patient No:

Sex:

Room No:

Specimen Number:

DOBIAGE:

Date Specimen Received:

Attending Physician:

SUMMARY:

Filamentous fungus detected in direct Gram stain and culture. Morphology suggests a dematiaceous fungus resembling Exoohiala species. Dematiaceous fungi are responsible for a number of subcutaneous diseases. The morphology in the Gram stain suggests subcutaneous phaeohyphomycosis.

TEST REQEST:

Bacterial and fungal culture

SPECIMEN TYPE:

Aspirate of leg lesion

COMMENT:

The presence of septate, swollen hyphae in the Gram stain and a heavy growth of mold on culture plates suggests this b g u s is a "significant" isolate, rather than an incidental contaminant or colonizer. Dematiaceous fungi are a family of yeasts and molds that contain brown or black pigment in their cell walls. They cause a range of subcutaneous diseases, including; phaeohyphomycosis, mycetorna, and chromoblastomycosis. The septate hyphae seen in the direct examinatiok with no detectable grains (granules) or muriform bodies, suggests phaeohyphomycosis. The literature suggests surgical excision, especially in immunocompromised hosts, is the treatment of choice. Newer azole antifungal agents, however, have been used successfully. Case discussed with Dr. Berlin.

Figure 2. Special microbiology written reports.

report placed in the patient's chart so all services involved share the same interpretation (Figure 2). The clinical service has an important responsibility to communicate with the laboratory to optimize care of the patient with cancer. The laboratory compiles data collected from groups of patients that is available and useful to physicians. Review and discussion of test utilization is essential for costeffective, quality health care. This may include analysis of blood cultures documenting an acceptable level of contamination, appropriate number collected per day, and sufficient blood volume per culture. In addition, information about changing resistance patterns or nosocomial transmission can be provided to the clinician. As patients with malignancies become more complex and their infections increasingly difficult to treat, regular interaction between the laboratory and clinician is likely to improve patient care. References 1. Rosenow E, Wilson W, Cockerill F. Pulmonary disease in the immunocompromised host. Mayo Clin Proc 1985;60:473-487.

2. Reimer L, Wilson M, Weinstein P. Update on dectection of bacteremia and fungemia. Clin Microbiol Rev 1997;10:444-465. 3. Mermel A, Maki D. Detection of bacteremia in adults: Consequences of culturing an inadequate volume of blood. Ann Intern Med 1993;119:270-272. 4. Kloos W, Bannerman T. Update on clinical significance of coagulase-negative staphylococci. Clin Microbiol Rev 1994:7:117-140. 5. Schifman R, Pindur A. The effect of skin disinfection materials on reducing blood culture contamination. J Clin Microbiol 1993;99:536-538. 6. Murray P, Washington J. Microscopic and bacteriologic analysis of expectorated sputum. Mayo Clinic Proc 1975;50:339-334. 7. Morris A, Tanner D, Reller B. Rejection criteria for endotracheal aspirates from adults. J Clin Microbiol 1993;31:1027-1029. 8. Kahn F, Jones J. Diagnosing bacterial respiratory infection by bronchoalveolar lavage. J Infect Dis 1987;155:862-869. 9. Baselski V, Baselski VS, el-Torky M, Coalson JJ, Griffin JP. The standardization of criteria for processing and interpreting laboratory specimens in patients with suspected ventilatorassociated pneumonia. Chest 1992:102(Suppl. 1):571S-579s. 10. Hooton T, Stamm W. Diagnosis and treatment of uncomplicated urinary tract infection. Infect Dis Clin North Am 2997;11: 551-581. 11. Goswitz JJ, Willard KE, Eastep SJ, et al. Utility of slide centrifuge Gram's stain versus quantitative culture for diagnosis of urinary tract infection. Am J Clin Pathol 1993:99:132136. 12. Kiska DL, Jones MC, Mangum ME, Orkiszewski D, Gilligan PH. Quality assurance study of bacterial antigen testing of cerebrospinal fluid. J Clin Microbiol 1995;33:1141-1144. 13. Fan K, Morris A, Reller 3 . Application of rejection criteria for stool cultures for bacterial enteric pathogens. J Clin Microbiol 1993:31:2233-2235. 14. Maki DG. Weise CE, Sarafin HW. A semiquantitative culture method for identifying intravenous-catheter-related infection. N Engl J Med 1977;296:1305-1309. 15. Kelly M. Wciorka LR, McConico S, Peterson LR. Sonicated vascular catheter tip cultures, quantitative association with catheter-related sepsis and the non-utility of an adjuvant cytocentrifuge Gram stain. Am J Clin Pathol 1996;105:210-215. 16. Yagupsky P, Nolte F. Quantitative aspects of septicemia. Clin Microbiol Rev 1990;3:269-279. 17. Kashuba A, Ballow C. Legionella urinary antigen testing: Potential impact on diagnosis and antibiotic therapy. Diagn Microbiol Infect Dis 1996;24:129-139. 18. Tierno PM Jr, Inglima K, Parisi MT. Detection of Bartonella (Rochalinznen) lzenselae bacteremia using bactlalert blood culture system. Am J Clin Pathol 1995:104:530-536. 19. Warren J, Allen S. Clinical. pathogenetic, and laboratory features of Capnocytophaga infections. Am J Clin Pathol 1986;86:513-518. 20. Berkowitz FE. The Gram-positive bacilli: A review of the microbiology, clinical aspects, and antimicrobial susceptibilities of a heterogeneous group of bacteria. Pediatr Infect Dis J 1994;13:1126-1 138. 21. Woods G, Washington J. 1987. Mycobacteria other than mycobacterium tuberculosis: Review of microbiologic and clinical aspects. Rev Infect Dis 1987;9:275. 22. Wilson M, et al. Controlled comparison of the bactec high-blood-volume fungal medium, BACTEC plus 26 aerobic blood culture bottle, and a 10-milliliter isoIator blood culture system for detection of fungemia and bacteremia. J Clin Microbiol 1993;31:865-871. 23. Thomson RB. Laboratory methods in basic virology. In: Baron EJ, Peterson LR, Finegold SM, eds. Baily & Scotts Diagnostic Microbiology, 9th ed. St. Louis, MO: Mosby, 1994, pp. 634-688. 24. Sun T. Current topics in protozoal diseases. Am J Clin Pathol 1994;102:16-29. 25. Homer KS, Wiley EL, Smith A, et al. Monoclonal antibody to Pneumocystis carinii. Am J Clin Pathol 1992;97:619-624. 26. Catchpole C, Hastings JGM. Measuring pre- and post-dose vancomycin levels - time for a change? J Med Microbiol 1995;45:309-311.

27. Eliopoulos GM, Eliopoulos CT. Therapy of enterococcal infections. Eur J Clin Microbiol Infect Dis 1990;9:118-126. 28. Hackbarth C, Chambers H, Sande M. Serum bactericidal activity of rifampin in combination with other antimicrobial agents against Stnphylococcl~s aurezls. Antimicrob Agents Chemother 1986;29:611-613. 29. Peterson L, Shanholtzer C. Tests for bactericidal effects of antimicrobial agents: Technical performance and clinical relevance. Clin Microbiol Rev 1992;5:420-432.

6. Recent advances in the management of fungal infections Jason Sanchez and Gary A. Noskin

1. Introduction The treatment of cancer has undergone many advances in recent years, leading to significant increases in patient survival. Improvements have come not only in the therapies themselves, but also in our ability to sustain critically ill patients. This, in turn, has allowed more intensive cytotoxic chemotherapies and an increased application of bone marrow transplantation for a broader range of neoplasms. More aggressive chemotherapy regimens have led to more profound immunosuppression and increased risk of infection. Longer survival in more critically ill patients, as well as the relative success of antibacterial therapies, have contributed to a dramatic increase in the incidence of invasive fungal disease among oncology patients. Additional risk factors for fungal infection, common in cancer patients, include the interruption of natural protective barriers by indwelling intravascular catheters, extensive surgical procedures, and chemotherapy-induced disruption of mucosal integrity. Opportunistic fungal infections have emerged as among the leading causes of mortality in cancer patients with severe, prolonged immunocompromised states. This trend has been most evident among patients with prolonged episodes of neutropenia, underscoring the primary importance of polymorphonuclear leukocytes in preventing and containing fungal infections. Candidn and Aspergillus have historically represented the most common fungal pathogens affecting cancer patients and continue to be the most frequent offenders. As the number of severely immunocompromised patients continues to grow, however, a greater variety of fungal opportunists are emerging. The last decade has witnessed a startling increase in invasive disease from more obscure fungi or those that were previously considered harmless commensals. Clinical experience with these organisms as pathogens is limited, necessitating increased awareness of their presentation, diagnosis, and management.

Gary A . Noskin (en), M A N A G E M E N T OF INFECTIOUS COMPLICATIONS IN C A N C E R PATIENTS. 0 1998. Klrlwer Academic Prthlisher~,Bosrotz. All rights re.rervecl.

2. Candida 2.1 Epidemiology Candida species represent the largest and most important group of opportunistic fungi in oncology patients. The increased incidence of invasive candidiasis in this setting has been well documented [1,2] and is associated with a high mortality [3,4]. The attributable mortality of hematogenous candidiasis alone has been estimated to be approximately 38%, and this figure increases as organ involvement is confirmed [5]. Candida albicans is the most common; however, other candidal species are beginning to assume a greater significance, especially in cancer patients [2,6]. The advent of molecular typing techniques has revolutionized the understanding of Candidn as a pathogen in the immunocompromised host. The ability to identify not only different species but unique strains within the same species is invaluable to the epidemiology of Candida. It is now clear that a hospitalized patient's endogenous flora is not the only potential source of infection but that organisms can be spread nosocomially from patient to patient by healthcare workers. Several outbreaks of infection with a particular strain of Candida have been documented. One investigation revealed that Candida was quite prevalent (17%) on the hands of intensive care unit workers and that these same strains were responsible for patient infection [7]. These findings underscore the importance of simple hand washing, as well as other more stringent infection control measures, in specific patient populations. Candidn nlbicans is an obligate human parasite that can often be isolated from the gastrointestinal tract, female genital tract, and the skin [2,8]. Candida albicans has been shown to be one of the most virulent species in animal models and has traditionally accounted for one half to two thirds of invasive candidal infections [2,8,9]. Candida species other than albicans comprise the remainder of invasive candidal infections. This group collectively tends to occur more often among patients with neoplastic diseases than with other immunocompromised states [2,10]. A recent review examined 1591 cases of systemic infections caused by Candida in patients with cancer between 1952 and 1992, and found that 46% were caused by the non-albicans Candida species. In this review, C. tropicnlis accounted for 25 %, C. glnbratn S%, C. parapsilosis 7%, and C. krusei 4% of the species identified [2]. Cnndida tropicalis, which colonizes patients less often than C. nlbicans, has been shown to be as virulent and continues to be a major pathogen, especially in cancer patients 121. Infections resulting from C. parapsilosis are also increasing, and they have been associated with both hyperalimentation and prosthetic devices [S]. Despite its relatively low virulence and previously infrequent role as a pathogen, C. krusei has emerged as an organism of increasing concern in the last decade. The widespread use of fluconazole both as a therapeutic and prophylactic agent has been associated

with an increased incidence of C. krusei infections in some medical centers [ll-151. One retrospective study of 463 bone marrow transplant and leukemic patients revealed a sevenfold increase in the incidence of bloodstream or visceral infections caused by C. krusei in the group that received fluconazole prophylaxis compared with those that did not (8.3 % vs. 1.2%) [12]. Many factors contribute to the prevalence of candidal infections in patients undergoing treatment for cancer. Neutropenia, the breakdown of anatomic barriers, use of broad-spectrum antibacterial agents, and fungal colonization are among the most important, while other predisposing factors have also been implicated (Table 1). Polymorphonuclear cells exert their effect by phagocytizing and killing candidal blastospores, and are also capable of damaging pseudohyphae [16-181. Monocytes and eosinophils are also involved in the killing of Candida by phagocytosis. The risk of hematogenous candidiasis that is incurred with the onset of neutropenia increases substantially as the duration exceeds 1 week and is greatest when the absolute neutrophil count falls below 100cells/mL [I]. Such profound neutropenia, previously seen primarily in the treatment of leukemia, is now more common as chemotherapy regimens for solid tumors have become increasingly aggressive [19]. The survival of neutropenic patients who acquire disseminated candidiasis is primarily dependent on the recovery of their granulocyte count, even if prompt antifungal therapy is instituted. The primary importance of polyrnorphonuclear cells is also evidenced by the observation that AIDS patients, despite frequent, severe mucocutaneous candidiasis, rarely develop disseminated infection [20]. These patients often have multiple risk factors for disseminated candidiasis but are protected by their relatively preserved neutrophil numbers and function. Anatomic barriers provided by intact skin and mucosal surfaces are a vital part of protection against invasive fungal disease. The disruption of these barriers is extremely common in the treatment of cancer and plays an important role in the pathogenesis of invasive candidiasis. Intravascular catheterization has been shown to increase the incidence of hematogenous

Table I. Risk Factors for Systemic Fungal Infections -

-

Underlying host defects Neutropenia Defects in humoral or cell-mediated immunity Jmmunosuppression Diabetes mellitus Cytotoxic chemotherapy High dose corticosteroids Long-term antimicrobial therapy Prosthetic devices Vascular or bladder catheters Prolonged hospitalization Solid organ transplantation Severe burns

candidiasis [19,21]. These catheters may serve as a portal of entry for primary infection or become secondarily infected during candidemia from another source [21]. There is mounting evidence that the gastrointestinal tract is the primary route by which Candidn invades patients with cancer. The process first requires colonization of the GI tract, usually by endogenous Candida, but strains acquired nosocomially are increasingly being recognized [7,22]. One prospective study of 139 neutropenic patients with hematologic malignancies documented invasive candidiasis in 22.2% of patients who had been colonized by Cnndida species at multiple sites, 4.0% of patients colonized at a single site, and none of the patients who had not been colonized 1231. The risk of colonization with Candida increases with the duration of hospitalization [2] and is associated with the administration of broad-spectrum antibiotics, corticosteroids, H2 blockers, and antacids [24]. The stomach and esophagus are reported to be the most commonly colonized sections of the GI tract [24]. Human and animal studies have shown that candidemia will occur in normal hosts given a high enough oral inoculum; [25,26] however, most disseminated disease in cancer patients is thought to involve a combination of mucosal damage and neutropenia or neutrophil dysfunction. Disruption of the GI mucosa occurs most commonly as a direct result of cytotoxic chemotherapy, especially cytarabine, resulting in mucositis and ulceration [24]. Other conditions that lead to mucosal damage include surgery, graft-versushost disease, [25] hypotension, [27] and concomitant infection with HSV, CMV, or bacterial pathogens [28]. Once colonization and mucosal damage coincide, local invasion is facilitated and disseminated infection can rapidly occur in the absence of neutrophils. 2.2 Clinical manifestations

In addition to localized mucocutaneous infections, Candida species are capable of a variety of deep tissue infections, including primary and secondary fungemia, meningitis, peritonitis, pyelonephritis, endocarditis, myocarditis, osteomyelitis, arthritis, endophthalmitis, pneumonia, and diseases of the spleen, liver, and gastrointestinal tract. [29,30]. The clinical presentation of disseminated candidiasis in neutropenic cancer patients may vary widely, presenting as an acute illness with the initial fungemic episode or as a more indolent disease, appearing only after granulocyte recovery. These different presentations have been labeled acute disseminated candidiasis (ADC) and chronic disseminated candidiasis (CDC), respectively. ADC is characterized by the onset of sudden illness, fungemia, and occasionally shock with multiorgan failure [31]. Characteristic macronodular skin lesions occur in up to 10% of neutropenic patients with disseminated candidiasis [32]. CDC, previously referred to as hepntosplenic candidinsis, may present over several months as a progressive debilitating illness. The liver, spleen, kidneys, and lungs are the most common sites for the large candidal abscesses that characterize this condition [31,33]. Chronic disseminated candidiasis is seldom asso-

ciated with documented fungemia or hypotension, but may have more subtle findings of upper abdominal tenderness and mild liver enzyme abnormalities [3,32]. Persistent fever, despite broad-spectrum antibiotic coverage, is a common but unreliable finding, especially in patients receiving corticosteroid therapy. Computed tomography has been useful in demonstrating the abdominal lesions of CDC, which are often visible only after granulocyte recovery [33]. It is important to view ADC and CDC as different ends of a clinical spectrum of disease that may present with elements of both or neither. 2.3 Diagnosis The diagnosis of candidiasis in the immunocompromised host relies primarily on the culture of organisms from normally sterile sites. A heightened awareness of the prevalence of Candida as a pathogen in cancer patients and the recognition that cultures may remain negative despite advanced fungal disease have led to the search for a reliable serodiagnostic test or marker. Despite many novel approaches, no method currently exists that has enough sensitivity and specificity to be clinically useful. The isolation of Candida from sites such as urine, feces, sputum, and skin should be viewed in the context of the overall clinical picture because it may represent either infection of colonization. The presence of a positive blood culture, however, is significant and carries the same prognosis as multiple positive blood cultures [34,35]. Candida is rarely a contaminant [36,37], even if the one positive blood culture is obtained from a central venous catheter. Eecciones and colleagues [35] reported that the incidence of disseminated candidal infection was 75% for patients with positive cultures obtained from both a peripheral vein and a catheter, and 68% when obtained from the catheter alone. Furthermore, mortality was higher in the later group (55% vs. 46%). Awareness of the cumulative risk factors for candidal infection, careful attention to commonly affected organ systems, and frequent blood cultures remain central to a diagnostic evaluation. 2.4 Treatment

The treatment of candidiasis depends on several factors, including the location and severity of infection, the underlying host, and hospital resistance patterns. The prominence of nosocomially acquired candidemia and the excess mortality associated with this infection mandate prompt initiation of antifungal therapy. Delays in the treatment of candidemia of even 1 day may have adverse effects on mortality [35]. Amphotericin B has long been the standard of therapy, and it remains the optimal initial treatment for candidemia in patients who are immunocompromised or critically ill. The addition of 5-flucytosine (5-FC) can provide synergistic activity, but its use is complicated by a narrow therapeutic window that requires monitoring of blood levels as well as its toxic effect on the bone marrow [38]. However, the combination

of amphotericin B and 5-FC is the preferred treatment of CNS and intraocular infections. Amphotericin B should be used in cases of organ infection with Candida and in cases of endocarditis, combined with surgical intervention. The most common toxicity of amphotericin B is dose-dependent nephrotoxicity, which has prompted a search for safe and effective alternatives. The advent of the triazoles in the late 1980s provided drugs with superior safety profiles; however, questions of their efficacy in immunocompromised patients have not been completely answered. A recent study that compared the use of amphotericin B (0.5-0.6mglkglday) with fluconazole (400 mglday) in patients with candidemia in the absence of neutropenia revealed no statistically significant difference in their effectiveness [39]. The use of lipid-based amphotericin B has shown much promise in the treatment of candidal infections, with comparable efficacy and greatly reduced renal toxicity [40]. Routine use of these formulations await large, randomized trials to verify their efficacy, but they appear to be reasonable alternatives in patients with worsening renal function who are taking conventional amphotericin B. Ultimately, the decision regarding which antifungal agent to choose should also take into account the prevalence of resistant Candidn species at a particular institution. However, the role of antifungal sensitivity testing requires additional testing to determine its optimal role. The use of indwelling intravascular catheters poses a definite risk of disseminated fungal infection and is a particular problem in cancer patients who require access devices for the administration chemotherapy and blood products [19]. Once candidemia has been established, it is necessary to remove intravascular catheters whenever possible. Failure to do so has been clearly associated with increases in morbidity and mortality [21,41,42]. Despite removal of any intravascular devices, antifungal therapy must accompany management, because catheter removal alone has also been associated with worse outcome [21].

3. Aspergillus Ubiquitous in the environment, Aspergillus grows particularly well on decaying vegetation [43]. This mold reaches the majority of patients through airborne spores that are inhaled into the lower respiratory tract, but it can also invade the nose and paranasal sinuses [44,45] as well as infect patients through interruption of the skin. Pulmonary disease is the most common clinical manifestation of aspergillosis in patients with cancer, and it usually manifests as a rapidly progressive pneumonia with high fevers followed by dense pulmonary consolidation [46,47]. Aspergilloma, or fungus ball of the lung, is another form of pulmonary aspergillosis that occurs in patients with pre-existing lung disease such as sarcoidosis, tuberculosis, bronchiectasis, or previous cavitary lesions. It occurs

almost exclusively in the upper lobes of the lung [43]. Massive hemoptysis remains both a common and fatal complication of pulmonary aspergillosis and occurs most frequently shortly after recovery from granulocytopenia [48,49]. Once disseminated, Aspergill~isspp. can invade any organ and is characterized pathologically by extension along blood vessels, leading to hemorrhagic infarction and necrosis [46]. Aspergilltls fumigatus is the most common cause of aspergillosis. A. flavus, although not as common, is also an important pathogen that tends to involve the nose and paranasal sinuses of severely immunocompromised patients [50,51]. Individuals who are at the highest risk of aspergillosis include those receiving bone marrow or solid organ transplants and those who are neutropenic following cytotoxic chemotherapy [52-541.

3.1 Diagnosis

A definitive diagnosis of invasive aspergillosis requires the demonstration of tissue invasion on biopsy and is facilitated by culture of this tissue. Blood, cerebrospinal fluid, and bone marrow are rarely positive, even in patients with advanced disease [43]. Aggressive pursuit of new pulmonary infiltrates, especially in patients with neutropenia, may lead to early diagnosis. In patients at high risk, isolation of hyphae from pulmonary secretions obtained during bronchoscopy can be highly suggestive of disease, whereas computed tomographic (CT) scans can often identify suspicious patterns not seen on chest x-ray. Radiographically, the "halo" and "air crescent" signs that correspond to a central fungal nodule surrounded by a rim of coagulative necrosis, are relatively specific for aspergillosis in the appropriate clinical setting [55,56]. Computed tomographic or magnetic resonance imaging (MRI) scans of the paranasal sinuses have also proven to be quite useful in the detection of occult rhinocerebral disease. The use of serodiagnostic tests, as well as enzymatic evaluation of bronchoalveolar lavage fluid, has proven too insensitive and nonspecific for accurate diagnosis. The propensity of Aspergillus to invade blood vessels with subsequent tissue necrosis should prompt a search for this organism in patients with black nasal discharge or eschars.

Invasive aspergillosis responds poorly to medical therapy alone and requires large doses of amphotericin B in the range of 1.0-1.5 mglkgld. Because of the need for such high doses and the risk of nephrotoxicity, the use of lipid-based amphotericin B has been investigated. Currently, there are three lipid-based preparations that are approved for use in the United States: amphotericin B lipid complex (Abelcet), amphotericin B cholesteryl sulfate complex (Amphotec), and liposomal amphotericin B (AmBisome). The major advantage of these products is reduced nephrotoxicity compared with conventional amphotericin B. This is particularly beneficial in patients with pre-existing

renal impairment or receiving other concurrent nephrotoxins. Although there is cautious optimism with the lipid-based preparations, their cost, which often exceeds $500 a day, may limit their widespread use. Itraconazole, an oral triazole, has activity against Aspergill~lsand has been used with some success in more indolent cases [57,58]. For rhinocerebral aspergillosis, surgical debridement in addition to medical therapy is necessary to optimize response. Surgical resection has also been advocated in the treatment of localized pulmonary disease during neutropenia; however, the benefits are not as clear and it remains controversial. 4. Mucormycosis

Mucormycosis is the term used to describe disease caused by fungi of the order Mucorales. Rhizopus and Rhizomucor are the two genera most commonly associated with pathogenesis. These fungi exist as molds and are common throughout the environment, gaining access to the majority of patients through the inhalation of airborne spores. Cutaneous infections can also occur by direct inoculation of skin in areas of breakdown. In industrialized nations, mucormycosis occurs primarily in neutropenic cancer patients, diabetics, and victims of trauma. Mucormycosis often involves the paranasal sinuses, and it can extend into surrounding structures to invade the central nervous system [59]. Rhinocerebral invasion is observed most often in leukemia patients with prolonged neutropenia and diabetics in association with acidosis [60-621. Profoundly neutropenic patients are also at increased risk of pulmonary involvement as well as disseminated disease [63,64]. As with aspergillosis, mucormycosis has a proclivity for vascular invasion, leading to hemorrhage, infarction, and necrosis, and can produce the same findings of black nasal discharge or eschar. Diagnosis depends on demonstration of tissue invasion on biopsy and is facilitated by the same imaging studies described earlier for aspergillosis. Effective therapy involves a combination of surgical debridement and high-dose amphotericin B, which is minimally effective alone. Even with optimal treatment, the response in neutropenic patients is relatively poor. The azoles have no activity against these molds and therefore no role in treating mucormycosis. 5. Fungi of emerging importance

There now exists nearly 300 species of fungi that are known to cause human disease and the list is rapidly expanding [65]. Oncology patients are particularly vulnerable to these new pathogens because host defense mechanisms are increasingly compromised by aggressive chemotherapy and radiation. The majority of these fungi are ubiquitous in the environment and consist of both yeasts and molds.

5.1 Yeasts In addition to Candida, emerging yeasts in neutropenic hosts include Trichosporon spp., Blastoschizornyces capitatus (previously Trichosporon cutaneurn), Malassezia spp., Rhodotorula spp., and Hansenula an.ornala. Fungemia or catheter-associated infection is the primary manifestation of disease in many of these new yeast pathogens; however, most are also capable of deep tissue invasion in severely immunocompromised patients. Risk factors for infection with these yeasts are similar to those associated, with candidiasis with neutropenia, previous antibiotic use, indwelling intravascular catheters, and gastrointestinal disruption being among the most important. As with Candida, nosocomial infection from organisms carried on the hands of healthcare workers has also been reported with several of these emerging yeast pathogens [66,67].

5.1.1 Trichosporon. The term trichosporonosis encompasses infection from either Trichosporon species or Blastoschizornyces capitatus. The major predisposing factors for invasive infection with these organisms are severe neutropenia and corticosteroid therapy [65,68]. As with candidiasis, trichosporonosis in the neutropenic patient may result in a wide range of clinical presentations, from acute sepsis to a more indolent process. Acute disseminated infection may yield positive blood or urine cultures and, in addition to fever, manifest with skin lesions, pulmonary infiltrates, or ocular involvement [69,70]. Alternatively, persistent fever in a patient recovering from myelosuppression may be the only clue to a more chronic infection, often characterized by abdominal organ abscesses. The latter presentation is more commonly seen with infection from B. capitatus [65,71]. Of note, the serum latex agglutination test for Cryptococcus neoformans is frequently positive in patients infected with Trichosporon species but not B. capitatus [72]. Microbiologically, B. capitatus can be distinguished from Trichosporon spp. by the production of anelloconidia rather than arthroconidia, both in vitro and in vivo [71]. Despite variable activity against Trichosporon, high-dose amphotericin B remains the first-line therapy for invasive disease, although the response is generally poor in the absence of immune recovery. There have been some reports that fluconazole may be more efficacious than amphotericin B and that the two agents should be used in combination for serious infections 1681. With the frequent use of fluconazole as a prophylactic agent, however, fluconazole-resistant strains of Trichosporon spp. and B. capitatus are increasingly being identified [66]. Both Trichosporon spp. and B. capitatus are resistant to 5-FC as well [73]. 5.1.2 Malassezia, Rhodotorula, and Hansenula. The majority of infections with these emerging yeasts occur as isolated fungemias in cancer patients with indwelling intravascular catheters. Despite their relatively low virulence com-

pared with Candida, each is capable of a variety of deep tissue infections in the compromised host [65,68,73]. Malassezia furf~uand M. pachydernzatis cause fungemia and other deep tissue infections in immunocompromised or debilitated patients, and have been associated with the administration of intralipid infusions [65,68]. Malassezia can be distinguished from Candida and other yeasts by its ovoid shape and its need for lipid-enriched media for growth. Rhodotorula is a common commensal that can be isolated from human skin, sputum, urine, and feces [74]. Rhodotorula, along with C. parapsilosis, were the two most common yeasts isoIated on the hands of healthcare workers in a study that documented this to be a common finding (greater than 70% of personal tested) [67]. Finally, Hansenula is another yeast being increasingly reported as cause of bloodstream infection in both cancer patients and other immunocompromised individuals [73]. Amphotericin B, with or without 5-FC, is generally more effective in treating Malnssezia, Rhodoforula, and Hansenula compared with Trichosporon; however, immune recovery is still critical in cases of advanced infection. In vitro susceptibility testing has indicated that Rhodotor~rlnand Hansenula are generally resistant to fluconazole; however, it remains a viable alternative in the treatment of Malassezia [68,73]. The issue of catheter removal in hematoginous infections with these low-virulence yeasts remains controversial because insufficient experience exists with these organisms. One study demonstrated equal resolution of Rhodotorula fungemia in 23 patients, half of whom received therapy with amphotericin B alone and half in conjunction with catheter removal [74]. However, a retrospective evaluation of fungemia from unusual yeasts in general (including atypical Cnndidn species) indicated that failure to remove an indwelling intravascular catheter was associated with higher mortality [73]. At present, it is recommended that all fungemias in immunocompromised hosts be treated with appropriate antifungal therapy and that indwelling catheters be removed whenever possible.

5.2 Molds Fungal infection due to uncommon molds has increased in the past few decades, especially among bone marrow transplant and leukemic patients who endure prolonged episodes of neutropenia. Acquisition of these organisms is usually via the respiratory route or through breaks in the skin. hyalohyphornycosis refers to infection from molds whose basic tissue form is that of hyphal elements consisting of branched or unbranched hyphae without pigment in their cell walls. Important molds in this group include F~lsari~~rn, Scedosporium, and Acrenzonium species. The dematiaceous fungi produce disease referred to as phaeohyphornycosis and comprise a group of molds with melanin in their cell wall imparting a darkly pigmented color. Bipolaris, Exserohil~~m, Exop hialn, Curvularia, Alternaria, and Phialophora are the most frequent genera to cause human disease in this group [68].

5.2.1 Hyalohyphomycosis Fusarium spp. F~isariumspp. are common plant pathogens that are ubiquitous in the soil. In humans they had long been associated with superficial infections such as onychomycosis, keratomycosis, and colonization of wounds and burns. Disseminated fusariosis was first documented in the early 1970s in a patient with acute leukemia and has since been recognized as an increasingly common is now pathogen in severely immunocompromised individuals [75]. F~lsari~lm the second most common pathogenic mold (following Aspergill~ls)in patients receiving cytotoxic chemotherapy and BMT for hematologic malignancies [76,77]. Severe neutropenia is the greatest risk factor for disseminated infection, which occurs most often in patients with hematologic malignancies [77,78].In most cases, the portal of entry is unknown; however, the respiratory tract and disruption of the skin have been the most often reported. Disseminated F~isarium infection is clinically characterized by fever, myalgias, and a high frequency of both skin lesions and positive blood cultures. The skin lesions, which occur in 60-80% of individuals, are usually multiple and either resemble ecthyma gangrenosum or form subcutaneous nodules [65,68,75]. F~isariunz,like Aspergillus, is vasotropic; however, the frequency of im (-60%) far exceeds that of positive blood cultures in F ~ ~ s n r i ~infections Aspergillus (<5%) [65,68]. This adventitial differentiation is not unique to F~isarium, occurring in other agents of hyalohyphomycosis as well, and appears to correlate with rates of fungemia [65]. Diagnosis of fusiarosis depends on culture of the mold from either blood or tissue biopsy because histological findings are indistinguishable from that of aspergillosis. Invasive or disseminated fusariosis is associated with a mortality rate exceeding 75%, and successful outcome is dependent on resolution of the myelosuppressed state [65,75]. Early diagnosis has not proven to impact survival because currently available antifungal agents have failed to significantly alter the clinical course of disseminated infection [75]. Amphotericin B appears to have the greatest activity in vitro and may be beneficial in high doses to provide more time for immune recovery [75]. Once diagnosis is established, all efforts should be focused on resolution of neutropenia with the aid of growth cell factors and possibly white blood cell infusions [65,75]. Scedosporium spp. and others. Other molds that are emerging as sporadic causes of hyalohyphomycosis in immunocompromised cancer patients include Scedosporium spp., Acremoniurn spp., Perzicillitlm spp., and Paecilomyces spp. The two most common species in the genus Scedosporium known to cause disease in humans are S. prolificans and S. apiospermum (also known as Pseudallesceria boydii when in the perfect state) [68]. P. boydii and S. prolificnns are capable of a wide range of infections, often involving the sinuses or lower respiratory tract, the eye, and the central nervous system [68]. As with Fusarium, dissemination is associated with a high rate of positive blood

cultures and diagnosis is dependent on culture of the organism because they are histopathologically similar to Aspergillus [65]. Recovery from infection is again highly dependent on recovery of the patients' immune status. The azoles have some activity against P. boydii ,while S. prolificans has proven resistant to all systemic antifungal therapy [68]. Increasingly, molds such as Acremoniurn spp., Penicillium spp., and Paecilomyces spp. are being reported to cause disease in immunocompromised patients, especially those with hematologic malignancies. These molds cause similar clinical syndromes to that described with Fusarium spp. and Scedosporium spp.

5.2.2 Phaeohyphomycosis. The dematiaceous fungi are darkly pigmented as a result of melanin in their cell walls. The genera most often responsible for human disease include Bipolaris, Curvularia, Exserohilum, Exophiala, Alternarin, and Phialopora [65,79]. These molds may result in variable tissue forms, from solitary yeastlike cells and short chains to septate hyphae that are often irregularly swollen to toruloid and branched or unbranched [68,79]. Sinusitis is the most common manifestation of disease in humans, accounting for the majority of fungal sinusitis in some institutions [65]. Sinusitis and local keratitis may occur in many patients who are not obviously immunocompromised; however, these lesions may go on to disseminate in severely immunocompromised patients [68,79]. Bipolaris is the most common cause of sinusitis, whereas Alternaria accounts for the majority of subcutaneous infections, although any of the genera may produce these infections [69,80]. Diagnosis of phaeohyphomycosis is usually made by culture of the organism from biopsy in addition to histopathological examination after Fontana-Masson staining, which demonstrates the melanin in the cell walls [79]. Surgical resection of localized disease, especially involving the paranasal sinuses, is an important part of therapy when feasible. Immunocompromised hosts with disseminated disease have a poor prognosis in the absence of neutrophil recovery. Limited experience with these organisms has revealed that itraconazole, and occasionally flucytosine, appear to have greater activity than amphotericin B, but antifungal therapy alone is rarely effective [65,68,79]. 6. Conclusions Advances in both the treatment of cancer and management of its complications have led to significant improvement in patient survival. Increases in the use of aggressive cytotoxic chemotherapies and radiation, while providing new hope for remission of malignancy, expose patients like to an ever increasing array of potential infectious pathogens. Opportunistic fungal infections have emerged as one of the leading causes of mortality in patients with prolonged

immunocompromised states. Candida and Aspergillus continue to be the most common fungal pathogens associated with the treatment of malignancy. However, new and unusual fungal pathogens consisting of both yeasts and molds are emerging in significant numbers in cancer patients. These new organisms are frequently resistant to currently available antifungal agents, and patient survival is often dependent on recovery of immune status. The emergence of fungal opportunists, both new and old, has provided new challenges to today's clinician. A high index of suspicion, frequent culturing, and aggressive diagnostic techniques are critical to early diagnosis and effective therapy.

References 1. Anaissie E. Opportunistic mycoses in the immunocompromised host: Experience at a cancer center and review. Clin Infect Dis 1992;14(Suppl. 1):S43-S53. 2. Wingard JR. Importance of Candida species other than C. albicans as pathogens in oncology patients. Clin Infect Dis 1995;20:115-125. 3. Swerdloff JN, Filler SG, Edwards JE. Severe candidal infections in neutropenic patients. Clin Infect Dis 1993;17(Suppl.2):S457-S467. 4. Slavin MA, Osborne B, Adams R, Levenstein MJ, Schoch HG, Feldman AR, Meyers JD, Bowden RA. Efficacy and safety of fluconazole prophylaxis for fungal infections after marrow transplantation - A prospective, randomized, double-blind study. J Infect Dis 1995;171:1545-1552. 5. Wey SB, Mori M, Pfaller MA, Soolson RF, Wenzel RP. Hospital-acquired candidemia: The attributable mortality and excess length of stay. Arch Intern Med 1988;148:2642-2645. 6. Merz WG. Candidn lusiraniae: Frequency of recovery, colonization, infection, and amphotericin B resistance. J Clin Microbiol 1984;20:1194-1195. 7. Voss A, Pfaller MA, Hollis RJ, Rhine-Chalberg J , Doebbeling BN. Investigation of Candida albicnns transmission in a surgical intensive care unit cluster with genomic DNA typing methods (abstract no. 1638). In: Program and Abstracts of the 33rd Interscience Conference on Antimicrobial Agents and Chemotherapy (New Orleans). Washington DC: American Society for Microbiology, 1993. 8. Weems JJ. Candida parnpsilosis: Epidemiology, pathogenicity, clinical manifestations, and antimicrobial susceptibility. Clin Infect Dis 1992;14:756-766. 9. Beck-Sague CM, Jarvis WR, Banerjee SN, Culver DH, Gaynes RP. Nosocomial fungal infections in U.S. hospitals, 1980-1990 (abstract no. 1129). In: Program and abstracts of the 30th Interscience Conference on Antimicrobial Agents and Chemotherapy. Washington DC: American Society for Microbiology, 1990. 10. Komshian SV, Uwaydah, Sobel JD, Crane LR. Fungemia caused by Candida species and Torulopsis glabrata in the hospitalized patient. Rev Infect Dis 1989;11:379-390. 11. Merz WG, Karp JE, Schron D, Saral R. Increased incidence of fungemia cuased by Candida krluei. J Clin Microbiol 1986;24:581-584. 12. Wingard JR, Merz WG, Rinaldi MG, Johnson TR, Karp JE, Saral R. Increase in Candida krusei infection among patients with bone marrow transplantation and neutropenia treated prophylactically with fluconazole. N Engl J Med 1991;325:1274-1277. 13. Bignardi GE, Savage MA, Coker R, Davis SG. Fluconazole and Candida krusei infections [letter]. J Hosp Infect 1991;18:326-327. 14. McIlroy MA. Failure of fluconazole to suppress fungemia in a patient with fever, neutropenia and typhlitis [letter]. J Infect Dis 1991;163:420-421.

15. Person DA, Laughlin M, Tanner D, Perfect J, Gockerman JP, Hathorn JW. Fluconazole and Carzdida k r ~ ~ sfungemia ei [letter]. N EngI J Med 1991;325:1315. 16. Diamond RD, Lyman CA, Wysong DR. Disparate effects of interferon and tumor necrosis factor in early neutrophil respiratory burst and fungicidal responses to Candida albicans hyphae in vitro. J Clin Invest 1991;87:711-720. 17. Lyman CA, Simons ER. Melnick DA. et al. Induction of signal transduction in human neutrophils by Candida albicans hyphae: The role of pertussis toxin-sensitive guanosine triphosphate-binding proteins. J Infect Dis 1988;158:1056-1064. 18. Fleischmann J, Lehrer RI. Phagocytic mechanism in host response. In: Howard DM, ed. Fungi Pathogenic for Humans and Animals, Part B2. New York: Marcel Dekker, 1985, pp. 123-149. 19. Brown AE. Overview of fungal infections in cancer patients. Semin Oncol 1990;17(Suppl. 6):2-5. 20. Odds FC, Schmid J, Sol1 DR. Epidemiology of Candida infections in AIDS. In: Vanden Bossche H, MacKenzie DWR, Cauwenbergh G, Van Cutsem J, Drouhet E, Dupont B, eds. Mycoses in AIDS patients. New York: Plenum Press, 1990, pp. 67-74. 21. Lecciones JA, Lee JW, Navarro EE, et al. Vascular catheter-associated fungemia in patients with cancer: Analysis of 155 episodes. Clin Infect Dis 1992;14:875-883. 22. Isenberg HD, Tucci V, Cintron F, Singer C, Weinstein GS, Tyras DH. Single-source outbreak of Candida tropicalis complicating coronary bypass surgery. J Clin Microbial 1989;27:24262428. 23. Martino P, Girmenia C, Venditti M, et al. Candida colonization and systemic infection in neutropenic patients: A retrospective study. Cancer 1989;64:2030-2034. 24. Cole GT, Halawa AA, Anaissie EJ. The role of the gastrointestinal tract in hematogenous candidiasis: From the laboratory to the bedside. Clin Infect Dis 1996;22(Suppl. 2):s73s88. 25. Krause W, Matheis H, Wulf K. Fungemia and funguria after oral administration of Candirln albicans. Lancet 1969;1:598-599. 26. Stone HH, Kolb LD, Currie CA, et al. Candida sepsis: Pathogenesis and principles of treatment. Ann Surg, 1974:697-711. 27. Wade JC. Epidemiology of Candida infection. In: Bodey GP, ed. Candidiasis: Pathogenesis, Diagnosis, and Treatment, 2nd ed. New York: Raven Press, 1993, p. 85. 28. McDonald G, Sharma P, Hackman R, Meyers J, Thomas ED. Infectious esophagitis in immunocompromised patients after marrow transplantation. Gastroenterology 1985,88:11111117. 29. Pfaller MA. Epidemiology and control of fungal infections. Clin Infect Dis 1994;19(Suppl. 1):S8-S13. 30. Edwards JE. Candidn species. In: Mandell GL, Bennett JE, Dolin R, eds. Principles and Practice of Infectious Diseases, 4th ed. New York: Churchill Livingstone, 1995, pp. 2289-2305. 31. Walsh, TJ, Lee JW, Roilides E, Pizzo PA. Recent progress and current problerns in management of invasive fungal infections in patients with neoplastic deseases. Curr Opin Oncol 1992;4:647-655. 32. Bodey GP. Disseminated candidiases in neutropenic patients. Int J Infect Dis 1997;1(Suppl. 1):S2-S6. 33. Flynn PM, Shenep JL, Crawford R, Hughes WT. Use of abdominal computed tomography for identifying disseminated fungal infection in pediatric cancer patients. Clin Infect Dis 1995;20:964-970. 34. Meunier F, Aoun M, Bitar N. Candidemia in immunocompromised patients. Clin Infect Dis 1992;14(Suppl. 1):S120-S125. 35. Lecciones J, Witebsky F, Marshall D, Gress J, Pizzo P, Walsh TJ. Catheter-associated fungemia in cancer patients: Characteristics and clinical outcome (abstract no. 68). In: Program and Abstracts of the 29th Interscience Conference on Antimicrobial Agents and Chemotherapy. Washington DC: American Society for Microbiology, 1989, p. 112. 36. Edwards JE. Invasive Candida infection. N Engl J Med 1991;324:1060-1062.

37. Young RC, Bennett JE, Geelhoed GW, et al. Fungemia with compromised host resistance. A study of 70 cases. Ann Intern Med 1974;80:605-612. 38. Polak A. 5-Fluorocytosine and combinations. Ann Biol Clin (Paris) 1987;45:669-672. 39. Rex JH, Bennett JE, Sugar AM. et al. A randomized trial comparing fluconazole with amphotericin B for the treatment of candidemia in patients without neutropenia. N Engl J Med 1994;33I:1325-1330. 40. Stevens DA. Overview of amphotericin B colloidal dispersion (amphocil). J Infect 1994;28(Suppl. 1):S45-S49. 41. Eppes SC, Troutman JL, Gutman LT. Outcome of treatment of candidemia in children whose central catheters were removed or retained. Pediatr Infect Dis J 1989;8:99-104. 42. Dato VM, DAjani AS. Candidemia in children with central venous catheters: Role of catheter removal and amphotericin B therapy. Pediatr Infect Dis J 1990;9:309-314. 43. Bennett JE. Aspergillzts species. In: Mandell GL, Bennett JE, Dolin R, eds. Principles and Practice of Infectious Diseases, 4th ed. New York: Churchill Livingstone, 1995,pp. 2306-2310. 44. Talbot GH, Huang A, Provencher M. Invasive Aspergillus rhinosinusitis in patients with acute leukemia. Rev Infect Dis 1991:13:219-232. 45. Peterson DE, Schimpff SC. Asl~ergill~rs sinusitis in neutropenic patients with cancer: A review. Biomed Pharmacother 1989;43:307-312. 46. Young RC, Bennett JE, Vogel CL, et al. Aspergillosis. The spectrum of the disease in 98 patients. Medicine 1970;49:147-1 73. 47. Meyer RD, Young LS, Armstrong D. et al. Aspergijlosis complicating neoplastic disease. Am J Med 1973;54:6. 48. Pagano L, Ricci P, Nosari A, Tonso A, Buelli M, Montillo M, Cudillo L, et al. Fatal haemoptysis in pulmonary filamentous mycosis: An underevaluated cause of death in patients with acute leukaemia in haematological con~pleteremission. A retrospective study and review of the literature. Br J Haematol 1995;89:500-505. 49. Wallace TM Jr., Sais GJ, Frank I, Gefter WB, Aronchick JM, Wallace TM. Pulmonary aspergillosis in patients with AIDS. Chest 1994;105:37-44. 50. Young RC, Jennings A. Bennett Je. Species identification of invasive aspergillosis in man. Am J Clin Pathol 1972;58:554-557. 51. Green WR, Font RI. Zimmerman LE. Aspergillosis of the orbit. Report of ten cases and review of the literature. Arch Ophthalmol 1969:82:302-313. 52. Allo MD, Miller J, Townsend T, Tan C. Primary cutaneous aspergillosis associated with Hickman intravenous catheters. N Engl J Med 1987;317:1105-1108. 53. Kusne S, Torre-Cisneros J, Manez R, et al. Factors associated with invasive lung aspergillosis and the significance of positive Aspergillrrs cultures after liver transplantation. J Infect Dis 1992;166:1379-1383. 54. Boon AP, O'Brien D. Adams DH. 10 year review of invasive aspergillosis detected at necropsy. J Clin Pathol 1991;44:452-454. 55. von Eiff M, Roos N, Schulten R, Hesse M. Zuhlsdorf M, van de Loo J. Pulmonary aspergillosis: Early diagnosis improves survival. Respiration 1995;62:341-347. 56. Hruban RH, Meziane MA, Zerhouni EA, Wheeler PS, Dumler JS, Hutchins GM. Radiologicpathologic correlation of the CT halo sign in invasive pulmonary aspergillosis. J Comput Assist Tomogr 1987:11:534-536. 57. Faggian G. Livi U, Bortolotti U, et al. Intraconazole therapy for acute invasive pulmonary aspergillosis in heart transplantation. Transplant Proc 1989:21:2506-2507. 58. Denning DW, Tucker RM, Hanson LH, et al. Treatment of invasive aspergillosis with itraconazole. Am J Med 1989;86:791-800. 59. Berthier M, Palmieri 0, Lylyk P, et al. Rhino-orbital phycomycosis complicated by cerebral abscess. Neuroradiology 1982;22:221-224. 60. Meyers BR, Wormser G, Hirschman SZ, et al. Rhinocerebral mucormycosis. Premortem diagnosis and therapy. Arch Intern Med 1979;139:557-560. 61. Maniglia AJ, Mintz DH, Novak S. Cephalic phycomycosis: A report of eight cases. Laryngoscope 1982;92:755-760.

62. Abedi E, Sismanis A, Choi K, et al. Twenty five years' experience treating cerebrorhino-orbital mucormycosis. Laryngoscope 1984;94:1060-1062. 63. Lehrer RI, Howard DH, Sypherd PS, et al. Mucormycosis. Ann Intern Med 1980;93:93108. 64. Kolbeck PC, Makhoul RG, Bollinger RR, et aI. Widely disseminated Cunninghamella mucormycosis in an adult renal transplant patient: A case report and review of the literature. Am J Clin Path01 1985;83:747-753. 65. Perfect JR, Schell WA. The new fungal opportunists are coming. Clin Infect Dis 1996;22(Suppl. 2):S112-S118. 66. D'Antonio D, Mazzoni A, Iacone A, Violante B, Capuani MA, Schioppa F, Romano F. Emergence of fluconazoIe-resistant strains of Blastoschizomyces capitatus causing nosocomial infections in cancer patients. J Clin Microbiol 1996;34:753-755. 67. Strausbaugh LJ, Sewell DL, Ward TT, Pfaller MA, Heitzman T, Tjoelker R. High frequency of yeast carriage on hands of hospital personnel. J Clin Microbiol 1994;32:2299-2300. 68. Vartivarian SE, Anaissie EJ, Bidey GP. Emerging fungal pathogens in immunocopromised patients: Classification, diagnosis, and management. Clin Infect Dis 1993;17(Suppl. 2):S487S491. 69. Walsh TJ, Melcher GP, Rinaldi MG, et al. Trichosporon beigelii, an emerging pathogen resistant to amphotericin B. J Clin Microbiol 1990;28:1616-1622. 70. Hoy J, Hsu K-C, Rolston K, Hopfer RL, Luna M, Bodey GP. Trichosporon beigelii infection: A review. Rev Infect Dis 1986:8:959-967. 71. Martino P, Venditti M, Micozzi A, Morace G, Polonelli L, Mantovani MP, Petti MC, Burgio VL, Santini C, Serra P, Mandelli F. Blastoschizonzyces capitatus: An emerging cause of invasive fungal disease in leukemia patients. Rev Infect Dis 1990;12:570-582. 72. McManus EJ, Bozdech MJ, Jones JM. Role of the latex agglutination test for cryptococca~ antigen in diagnosing disseminated infections with Trichosporon beigelii. J Infect Dis 1985;1167-1169. 73. Hazen KC. New and emerging yeast pathogens. Clin Microbiol Rev 1995;8:462-478. 74. Kiehn TE, Gorey E, Brown AE, Edwards FF, Armstrong D. Sepsis due to Rhodotorula related to use of indwelling central venous catheters. Clin Infect Dis 1992;14:841-846. 75. Martino P, Gastaldi R, Raccah R, Girmenia C. Clinical patterns of Fusarium infections in immunocompromised patients. J Infect 1994;28(Suppl. 1):S7-S15. 76. Morrison VA, Haake RJ, Weisdorf DJ. Non-Candida fungal infections after bone marrow transplantation: Risk factors and outcome. Am J Med 1994;96:497-503. 77. Merz WG, Karp JE, Hoagland M, Jett-Gogeen M, Junkins JM, Hood AF. Diagnosis and successful treatment of fusariosis in the compromised host. J Infect Dis 1988;158:1046-1055. 78. Anaissie E, Kantarjian H, Ro J, et al. The emerging role of Fusnriurn infections in patients with cancer. Medicine (Baltimore) 1988;67:77-83. 79. Fotergill AW. Identification of dematiaceous fungi and their role in human disease. Clin Infect Dis 1996;22(Suppl. 2):S179-S184.

7. Recent advances in the management of viral infections John R. Wingard

1. Introduction

The immunocompromised cancer patient is vulnerable to a wide spectrum of viral pathogens (Table 1).There has been increasing recognition of viruses as clinically important pathogens in cancer patients during the past decade. In part this is attributable to improved diagnostic techniques to better recognize viral pathogens as causes of illnesses not earlier recognized to be virally caused. In part this is also due to the increasing dose intensity of cytoreductive regimens used to control cancer, the increasing use of bone marrow transplantation (BMT) in the treatment of neoplastic diseases, and improvements in supportive care that permit patients to survive bacterial and fungal infections that in the past might have led to death before viral illness became manifest. Thus, there are greater numbers of highly immunosuppressed patients with severe compromise in cell-mediated immunity, the major host defense against most viral pathogens. Not only are viral infections increasingly recognized today, but a wider array of pathogens have been noted to cause complications of cancer therapy that in the past have been attributable to toxicities. Pneumonitis, cystitis, myelosuppression, mucositis, enteritis, and hepatitis are examples of syndromes that in the past have been attributable to tissue damage from cytoreductive regimens, or in the case of BMT patients, graft-versus-host disease (Table 2); in a number of instances, however, it is clear that viral pathogens are either sole causes of the syndrome or there is an interplay between viral pathogenesis, tissue damage, and disordered immune responses to the virus. The increased recognition of viral pathogenicity has fortunately been accompanied by the introduction of new therapeutics. Several nucleoside analogues, several biologic agents, and new vaccines all offer the clinician tools to prevent or reduce the morbidity associated with these organisms. Thus, prompt diagnosis of these potentially treatable syndromes and an understanding of how to use these new therapeutic modalities is important for optimal management of the cancer patient. Gary A. Noskit1 (ed), M A N A G E M E N T OF INFECTIOUS COMPLICATIONS IN C A N C E R PATIENTS.

01998. Kl~dwerAcadetnic Puh1isl1er.g Ro.rron. AN rights reserved.

Table I. Viral pathogens in immunocompromised cancer patients Herpesviruses Herpes simplex type 1 Herpes simplex type 2 Cytomegalovirus Varicella-zoster virus Epstein-Barr virus Human herpesvirus 6 Hepatitis viruses Hepatitis A Hepatitis B Hepatitis C Non-A. non-B, non-C hepatitis Adenoviruses Intestinal viruses Rotavirus Norwalk virus Adenoviruses Astroviruses Coxsackie Caliciviruses Respiratory viruses Respiratory syncytial virus Influenza Parainfluenza Papovaviruses Jc BK Human papilloma Retroviruses HTLVI HIV

Tnble 2. Syndromes due to viral pathogens often attributed to treatment toxicity p

p

p

p

p

Syndrome

Patient population

Viral pathogen

Oral mucositis Myelosuppression Interstitial pneu~nonia

Lymphoma, leukemia, BMT BMT BMT

Hemorrhagic cystitis Diarrhea

BMT BMT

Fever, unknown etiology Treatment-related lymphoma

BMT BMT

HSV CMV. HHV-6 CMV, HHV-6, adenovirus. RSV BK virus, adenovirus CMV, adenovirus, rotavirus, coxsackie CMV EBV

BMT = bone marrow transplantation; CMV = cytomegalevirus; EBV = Ebstein-Barr virus: HHV-6 = human herpesvirus-6; HSV = herpes simplex virus; RSV = respiratory syncytial virus.

2. Herpesviruses The most frequently recognized viral pathogens in cancer patients have been members of the herpesvirus family. These have long been recognized to be potential causes of serious and life-threatening illness. Patients receiving therapy for lymphoma, leukemia, and those undergoing bone marrow transplantation are especially susceptible. The human herpesviruses that cause clinically recognizable infection are herpes simplex virus type 1 (HSV-I), herpes simplex virus type 2 (HSV-2), varicella zoster virus (VZV), cytomegalovirus (CMV), Epstein-Barr virus (EBV), and human herpes virus type 6 (HHV-6). These DNA viruses are prevalent in the normal population. Initial infection often occurs early in life, is mild, is self-limited, and generally requires no therapy. After resolution of the primary infection, the virus typically establishes a latent infection that can be lifelong. HSV and VZV reside latently in sensory nerve ganglia; leukocytes harbor CMV, EBV, and HHV-6. With compromises in cell-mediated immunity, reactivation can occur and lead to subsequent morbidity. In the non-immunocompromised patient, reactivation can also occur but is generally associated with milder symptomatology than associated with the primary infection. In contrast, in immunocompromised patients reactivation is both more likely and more apt to lead to serious morbidity. The severity of manifestations tends to correlate with the degree of compromised immunity. 2.1 Herpes simplex vivzis The lesions from HSV infection are typically orofacial. Labial lesions may be absent. Intraoral mucosal ulcerations may be the sole manifestation. These lesions can be indistinguishable from the tissue damage that results from chemotherapy or radiotherapy. Thus, a pathogenic role for HSV in stomatitis has been often missed in the past; indeed, the reactivation of HSV and the occurrence of tissue damage from cytoreductive treatment often occur concomitantly, and these can result in severe oral mucositis. HSV-2 infection in cancer patients is less problematic because the virus is less common in the general population. However, reactivation can occur at high rates in patients who harbor latent HSV-2 and severe manifestations can result, especially in patients with hematologic malignancies and BMT recipients. Genital lesions (especially ulcerations) are frequent manifestations, but extragenital vesicles, in the gluteal and anal regions, can also occur. Although oral and genital mucosa are the major sites of HSV lesions, extension to the esophagus, urethra, bladder, and tracheal mucosa may also occur. Endoscopic biopsy may be necessary to distinguish a viral etiology from fungal or other possible causes. In profoundly immunocompromised patients dissemination and involvement of visceral tissues can occur. Culture of material from an infected lesion can confirm the diagnosis. In recent years rapid detection methods using antigen detection or polymerase

chain reaction (PCR) procedures offer quicker and easier alternatives [I-31. Cytologic examination of cells removed from infected lesions using the Tzanck procedure can demonstrate multinucleated cells but do not permit distinction between HSV or VZV [4]. Serologic tests can be helpful in identifying patients harboring latent virus (and thus susceptible for reactivation) but are of no value in documenting acute infection. Acyclovir, a purine analogue, is very active against HSV-1 and HSV-2, and has been shown in numerous clinical trials to be an effective treatment for HSV infection [5-91. Several oral and intravenous regimens have been evaluated and found to be effective and suitable for different clinical situations. Shortening of the time of viral shedding, time to cessation of pain, and time to prompt healing of lesions have been demonstrated in various studies. Acyclovir has also been shown to be efficacious in prophylaxis [lo-141. For patients at high risk for HSV reactivation and who are susceptible for serious morbidity, prophylaxis may be preferable to treatment [15-181. In adult patients undergoing intensive induction therapy for acute leukemia and in patients undergoing BMT who are HSV seropositive, reactivation occurs in approximately 70% and manifestations can be quite severe. Thus, in that setting acyclovir prophylaxis may be warranted. Indeed, the emergence of drug resistance appears to be less common where acyclovir is used prophylactically than when used as treatment of established infections where repetitive courses of acyclovir may be necessary and the frequency of drug resistance increases with each subsequent treatment episode [17-191. The emergence of acyclovir resistance has been increasingly noted in patients with the acquired immunodeficiency syndrome. Resistance is less frequent in patients receiving cancer therapy but appears most frequent in BMT recipients who have received repetitive courses of acyclovir for repeated infection. Acyclovir resistance usually is conferred by mutations in the genes encoding for the viral-specified thymidine kinase [20,21]. This viral-encoded enzyme is necessary for acyclovir phosphorylation, and without it little drug is converted to its active form. Thus, acyclovir and other nucleoside analogues that similarly rely on phosphorylation for their activity are inactive against acyclovir-resistant mutants. For patients with acyclovir-resistant HSV, foscarnet, a pyrophosphate analogue, which directly inhibits viral DNA polymerase and does not require thymidine kinase for its activity, is an alternative [22,23].

2.2Cytomegalovirus CMV, another member of the herpesvirus family, infects a substantial proportion of the general population. Infection is generally inapparent in the non-immunocompromised host, and although reactivation is frequent in immunocompromised patients, it rarely causes serious manifestations, except in highly immunocompromised patients such as BMT recipients, solid organ transplant recipients, and patients with the acquired immunodeficiency syn-

drome. Leukocytes are a reservoir of latent virus; thus, blood component transfusions as well as organ grafts can be sources of viral transmission. CMV can cause fever, hepatitis, pneumonitis, leukopenia, thrombocytopenia, esophagitis, enterocolitis, retinitis, a mononucleosis-like syndrome, and occasionally central nervous system manifestations. In BMT patients, the most common and severe manifestation is interstitial pneumonitis, which if untreated results in death in 80-90% of cases. Enterocolitis is less common but can represent a cause of severe diarrhea in the transplant recipient, and appears to be increasing in frequency. Chorioretinitis, a common clinical manifestation of CMV infection in HIV-infected patients, is uncommon in BMT recipients. Myelosuppression, a frequent accompaniment of cancer therapies, can have a variety of etiologies but CMV is one treatable cause. Ganciclovir, a nucleoside analogue structurally similar to acyclovir, is very active against CMV. It has assumed an important role in the treatment and prevention of CMV infection in transplant recipients. Ganciclovir exerts a potent antiviral effect in BMT patients with CMV pneumonitis, with a marked reduction in viral titers in infected tissue. However, when ganciclovir was used alone, there was no corresponding clinical benefit and most patients succumbed to relentless ventilatory failure [24]. However, several studies have shown that when ganciclovir is used in combination with immune globulin both antiviral and clinical benefits ensue [25-281. Thus, the mortality rate of 80-90% from CMV pneumonitis has been reduced to approximately 50%. Ganciclovir has also been evaluated as prophylaxis in allogeneic BMT patients who are seropositive and thus at high risk for CMV disease [29,30]. This approach has been found to be highly effective in reducing the risk for serious morbidity from CMV. Unfortunately, ganciclovir's side effects, especially myelosuppression, have led to episodes of neutropenia and bacteremia; thus, survival has not been appreciably improved. An alternative strategy, frequently referred to as early "preemptive therapy," has also been explored [31,32]. In this approach, patients undergo surveillance screening for viral reactivation. Those patients found to have active infection are then treated with ganciclovir to prevent clinical manifestations, which generally do not occur until several days to several weeks after reactivation. Screening is generally done on specimens of blood, urine, throat secretions, or bronchoalveolar fluid using the shell-vial methodology [33,34]. In recent years PCR and leukocyte antigen detection assays have been developed and are capable of detecting virus 1-2 weeks earlier than culture [35-431. Several recent reviews have discussed the advantages and disadvantages of prophylaxis versus preemptive therapy [44-461. In general, ganciclovir prophylaxis is more effective in preventing CMV disease, with fewer breakthrough episodes of CMV disease, while early preemptive ganciclovir is associated with fewer episodes of neutropenia and spares a sizable proportion of patients (in which reactivation does not occur) from the cost and toxicity of ganciclovir. With the introduction into clinical use of PCR and antigen detec-

tion assays, it can be expected that there will be fewer failures associated with the preemptive therapy approach. Several centers are evaluating a shorter course of preemptive ganciclovir to ascertain if its benefits can be achieved with less toxicity [47]. Oral ganciclovir, found to be potentially useful as maintenance therapy in HIV-infected patients [48,49], is being evaluated in the BMT setting. Because of poor bioavailability, low serum levels are achieved. Whether these levels will be adequate for prophylaxis in the BMT setting remains uncertain. Resistance to ganciclovir has occasionally been encountered in HIVinfected patients on chronic maintenance dose schedules [SO] but is rare in cancer patients. Resistance occurs by mutations in the UL97 gene region [51]. Foscarnet can be used for ganciclovir-resistant viral mutants [52]. Acyclovir has not been clinically useful in the treatment of CMV disease. However, several studies in both BMT and solid organ transplant recipients have indicated that as prophylaxis it is effective in reducing the risk for developing CMV disease 153,541. The explanation for this is not clear, but it would appear that a low level of acyclovir phosphorylation occurs despite the fact that CMV does not encode for a viral-specific thymidine kinase, the enzyme that most avidly phosphorylates acyclovir to its active metabolites. Thus. low levels of phosphorylated acyclovir may be effective when the viral burden is low, although not efficacious in instances in which the viral burden would be high (the treatment scenario). For patients who are CMV seronegative, infection can occur through acquisition of virus from blood transfusion or organ donation because leukocytes are a reservoir of latent virus. Accordingly, transfusion of only CMVseronegative blood products is an effective strategy in preventing CMV infection and disease [5S-581. Unfortunately, a substantial proportion of the blood donors are CMV-seropositive and harbor potentially transmissible virus. Accordingly, significant costs are incurred in the provision of CMV-negative blood products by blood banks. An alternative approach is the use of leukocyte filters, which are capable of eliminating most leukocytes that are present in erythrocyte and platelet products [59,60]. A controlled trial recently demonstrated that this approach is as effective as CMV screening [60]. This may have the added advantage of also reducing the risk for alloimmunization, another concern for patients who receive multiple blood products. CMV hyperimmune globulin and plasma have also been shown to reduce the risk for CMV disease in the BMT recipient [61-6.51. Because the antiviral potency of CMV immunoglobulin appears modest in studies in which it was used for treatment of CMV disease, speculation has been raised as to the mechanism of its action; it has been suggested that it may acting more as an immunomodulatory agent affecting antigen presentation or immune responses to CMV antigens rather than as an antiviral agent. Indeed, conventional lots of immune globulin not specifically chosen for high antiviral titers against CMV seem to be comparable with high-titer lots of immune globulin in preventing CMV disease. It should be noted that most studies have been

conducted in CMV-seronegative patients. Only one study conducted in seropositive patients has shown a benefit, and the benefit was modest [66]. Although widely used by many transplant centers, its utility is being reconsidered by many centers because of its high cost and the advent of other alternatives for prevention. 2.3 Varicella zoster virus VZV infection is highly prevalent in the general population. Cancer treatment regimens are associated with a risk for reactivation that is, compared with the nonimmunocompron~isedhost, slightly greater in solid tumor patients, substantially greater in patients treated for hematologic malignancies, and greatest in patients undergoing BMT. The most common manifestation is a dermatomal vesicular eruption, which may be preceded by a prodrome of localized pain and pruritus. Postherpetic neuralgia can persist for many months, especially in older individuals. Dissemination only occasionally occurs, but with highly immunocompromised patients such as allogeneic BMT recipients, dissemination can occur in up to 30-40% of individuals. Cutaneous dissemination, the most common form of spread, can be complicated by bacterial superinfection. Visceral dissemination can be life threatening, and VZV pneumonia is the most common lethal manifestation. In allogeneic BMT recipients a unique manifestation of VZV infection is abdominal pain, which can be quite severe and mimic a perforated viscus [67]. This can antedate the appearance of cutaneous vesicles and may be associated with mesenteric adenitis, hepatitis, or pancreatitis. It can be life threatening if not recognized and if treatment is not initiated promptly. Acyclovir is very active against VZV and has become the treatment of choice [68-721. Higher concelztrations of acyclovir are required to control VZV than HSV. Because of acyclovir's poor bioavailability, intravenous administration is the preferred method of treatment in immunocompromised patients. Although high-dose oral acyclovir, valacyclovir, and famciclovir have shown efficacy in the nonimmunocompromised hosts, their role remains undefined in the immunocompromised host. Acyclovir-resistant VZV has only been rarely encountered to date [73]. Foscarnet can be used for resistant pathogens [74]. Immune globulin can be given to susceptible immunocompromised patients if exposure is recognized within 3-4 days [75]. An attenuated vaccine has been found to be safe and protective for susceptible children with acute lyrnphoblastic leukemia [76-791. Safety has not been evaluated in the early convalescent BMT period. 2.4 Epstein-barr virus

EBV, the cause of infectious mononucleosis in the nonimmunocompromised host, only occasionally causes morbidity in the immunocompromised host despite high rates of reactivation. However, in transplant recipients, severe

morbidity can result from a severe mononucleosis-like syndrome or a variety of lymphoproliferative disorders. These can range from polyclonal lymphadenopathy to rapidly progressive monoclonal malignancy. Although these lymphoproliferative diseases are clearly EBV associated, molecular techniques have demonstrated mutations of oncogenes such as C-myc and tumor suppressor genes, which occur in the transition from benign to malignant disease [80,81]. The risk for EBV-associated lymphoproliferative diseases correlates with the degree of immunodeficiency. The use of multiple immunosuppressive agents, especially antithymocyte globulin, the use of T-cell depletion techniques in the BMT setting, and the occurrence of multiple rejection episodes in the solid organ transplant setting or severe graft-versus-host disease in the BMT setting all contribute to the risk for these disorders. The treatment approach that has been most fruitful is reduction in immunosuppressive therapy, which can effect a remission in the benign lymphoproliferative disorders. Although antiviral agents such as acyclovir and ganciclovir are active against EBV, their effectiveness in treating EBVassociated lymphoproliferative diseases has been disappointing in most cases. Once mutations in oncogenes and tumor suppressor genes occur, most treatment approaches have been largely ineffectual. 2.5 H~lnznnherpesviriis type 6 HHV-6 rarely causes clinical illness in the normal population despite being very prevalent. A self-limited eruption, exanthem subitum, has been noted in children. Recently, HHV-6 has been implicated as a potential pathogen causing some cases of interstitial pneumonitis, several CNS syndromes, and sometimes appears to be a cause of myelosuppression in BMT recipients [82-861. Ganciclovir and several other nucleoside analogues are active against HHV-6 in vitro, but to date there are no clinical trials [87].

3. Immune responses to the herpesvirus family Both humoral and cellular immune responses occur in response to infection by all of the herpesviruses. The immune responses felt to be most important in the control of active infection are the cytotoxic response mediated by T lymphocytes or natural killer (NK) cells. This has been most convincingly demonstrated in CMV infection [88-903. In the BMT recipient, resolution of active infection occurs only with the development of cytotoxic T-cell or NK responses. In the absence of the development of these responses, most patients succumb from infection. In BMT recipients with graft-versus-host disease, the orderly development of cytotoxic responses may be severely impaired and patients are at much greater risk for more frequent and more severe CMV infection and illness. Similarly, patients who are the recipients of Tlymphocyte-depleted bone marrow grafts are unable to mount robust T-cell

responses and are similarly more susceptible for more frequent and severe CMV infection and disease. These observations have led to consideration of cloning cytotoxic T cells (CTL) with anti-CMV activity and expanding them ex vivo for use as lymphocyte transfusions to bolster host immunity in an attempt to prevent severe CMV disease [91-941. Clinical trials are currently under way. EBV-specific cytotoxic T-cell precursors are more frequent in the circulation than CMV-specific CTL precursors. Buffy-coat transfusions have been successfully used in the treatment of EBV-associated lymphoproliferative disorders in transplant recipients without the need for ex-vivo clonal expansion [94,95]. These approaches of adoptive transfer of cellular immunity appear quite promising for the future. Bolstering the host immunity through the use of viral vaccines has been hampered by the lack of safe and highly immunogenic vaccines. A live attenuated varicella vaccine is useful in children with acute leukemia (as noted earlier); however, it have been felt to be too risky for use in the bone marrow transplant setting, except in patients 2 or more years after transplant without active GVHD. Attenuated CMV vaccines have been tested in clinical trials in solid organ transplants, but have been similarly felt to be too risky in the BMT setting. Recombinant vaccines are under development.

4. Hepatitis viruses The hepatitis viruses are a heterogeneous group of RNA (hepatitis A and C) and DNA (hepatitis B) pathogens whose portal of entry is generally the enteric route, with transmission by fecal-oral contact, but sexual and blood transmission can also occur. Recognition of the potential of transmission through blood products and the development of screening tests has led to a marked reduction in transmission of hepatitis B and C. However, there still remains the potential for the occurrence of viral hepatitis from transfusion of a putative, as yet unidentified, viral pathogen referred to non-A, non-B, nonC hepatitis. Inactivated hepatitis A and hepatitis B vaccines have been found to be safe and highly immunogenic. Immune globulin can be protective for those who must come in close contact with infected individuals to reduce the risk for infection. After exposure, immune globulin can also be efficacious for hepatitis A and B. Interferon have been shown to suppress hepatitis B and hepatitis C replication in clinical trials. The magnitude and durability of clinical benefit have been debated [96-991. A bone marrow graft from a seropositive individual has the potential for transmitting hepatitis B or C to the recipient. Different reports have suggested different rates of transmission and different degrees of severity of illness in the recipient of such transmission [10110.51. Donors in whom the presence of HCV RNA has been documented by PCR appear to have a greater risk for transmission than donors who are

RNA negative. Further study is needed to define the exact nature of the risks before firm recommendations call be made regarding the suitability of the seropositive marrow donor. Adoptive transfer of immunity to hepatitis B in the BMT setting from an immune donor may be a possible option for some patients [106].

5. Adenovirus Adenovirus is a viral pathogen capable of causing respiratory illness, conjunctivitis, gastroenteritis, interstitial pneumonitis, and hepatitis. Type 11 has been associated with hemorrhagic cystitis. Adenovirus isolation is noted in 5% of all allogeneic BMT recipients. Illness ensues in approximately 20% of infected individuals. Types 1, 5, and 7 appear to be the most common types causing invasive disease, which can be fatal in approximately half of cases. BMT patients who are the recipients of unrelated donor grafts, mismatched grafts, or grafts in which T-cell depletion has been performed appear to be a greater risk [107]. Currently there is no known effective antiviral therapy.

6. Intestinal viruses

Outbreaks of a variety of enteric pathogens occur in the community with seasonal variation. Immunocompromised patients can become infected during these community outbreaks. Common pathogens include Coxsackie virus, rotavirus, the Norwalk agent, caliciviruses, and astroviruses. The allogeneic BMT recipient is especially vulnerable for severe, even life-threatening, diarrheal illness. There are no effective antiviral therapies. Electrolyte and fluid replacement are important adjunctive measures. Immune globulin given orally has been suggested as a treatment for these illnesses, but adequate clinical trials are lacking.

7. Respiratory viruses Respiratory syncytial virus (RSV), influenza, and parainfluenza viruses are frequent causes of upper and lower respiratory tract illness. Transmission is frequent in the community and often is the source of infection in immunocompromised cancer patients. Inactivated influenza vaccine is available and may be potentially protective for immunocompromised patients [108], but severely immunocompromised patients, such as early convalescent allogeneic BMT recipients, unfortunately do not respond adequately. Amantadine and rimantidine are effective in the prevention of influenza in patients at risk for exposure. Rimantidine has less toxicity than amantadine [109]. Ribavirin, a nucleoside analogue, can be clinically useful for RSV

infection [110,111]. Immune globulin may have a synergistic effect in combination with ribavirin [112]. Cautionary measures must be exercised to avoid nosocomial transmission of these airborne organisms during community outbreaks [113,114].

8. Papovaviruses (polyomaviruses)

JC and BK viruses cause asymptomatic infection in children but establish a persistent infection in renal and urogenital epithelial cells. JC virus has been associated with progressive multifocal leukoencephalopathy. BK virus has been associated with hemorrhagic cystitis in allogeneic BMT recipients [1151191. At present there are no effective therapies.

9. Retroviruses Human T-cell lymphotrophic virus, type-1 (HTLV-1) is an endemic retrovirus in some areas of the world. Transmission can occur by breast feeding, sexual contact, or blood transfusion. It has been associated with the development of the adult T-cell 1eukemiaJlymphoma syndrome. Latency between infection and onset of disease is often more than a decade, and the risk for development of disease may be dependent on the age of infection with infection, with early childhood being most risky. HIV (formerly HTLV-3) is a retrovirus that is the causative agent of AIDS. Sexual transmission and transmission via blood transfusion or organ transplant is well established. The institution of routine screening tests for blood products and organ donors has reduced the risk for transmission substantially. Several nucleoside analogues are active inhibitors of reverse transcriptase, and protease inhibitors have recently been found to be useful in the suppression of viral replication, with corresponding clinical benefits. The emergence of antiviral resistance has plagued the development of effective and enduring antiviral strategies, however.

10. Conclusions The increase in viral infections in immunocompromised patients and the increasing numbers of immunocompromised patients have given a sense of urgency to improve our diagnostic techniques and to develop an armamentarium of antiviral agents for use in the control of these prevalent and opportunistic microorganisms. Recognition of the relevant protective immune responses is likely to lead to new biological strategies to supplement pharmacologic measures to control serious morbidity from these pathogens in the future.

References 1. Pruneda RC, Almanza I. Centrifugation-shell vial technique for rapid detection of herpes simplex virus cytopathic effect in vitro cells. J Clin Microbiol 1987;25:423. 2. Schmidt NJ, Dennis J, Devlin V, Callo D. Mills J. Comparison of direct immunofluorescence and direct immunoperoxidase procedures for detection of herpes simplex virus antigen in lesion specimens. J Clin Microbiol 1983;18:445-448. 3. Gonik B, Seibel M, Berkowitz A, Woodin MB, Mills K. Comparison of two enzyme-linked immunosorbent assays for detection of herpes simplex virus antigen. J Clin Microbiol 1991:29:436438. 4. Nahass GT, Goldstein BA, Zhu WY, Serfing NS, Leonardi CL. Comparison of Tzanck smear, viral culture, and DNA diagnosis methods in detection of herpes simplex and varicella-zoster infection. JAMA 1992;268:2541-44. 5. Meyers JD, Wade JC, Mitchell CD, et al. Multicenter collaborative trial of intravenous acyclovir for the treatment of mucocutaneous herpes simplex virus infection in the immunocompromised host. Am J Med 1982;73A:229. 6. Mitchell CD, Gentry SR, Boen JR, et al. Acyclovir therapy for mucocutaneous herpes simplex infections in immunocompromised patients. Lancet 1981;2:1389. 7. Shepp DH, Newton BA, Dandliker PS, et al. Oral acyclovir therapy for mucocutaneous herpes simplex virus infection in in~munocompromisedmarrow transplant recipients. Ann Intern Med 1985;102:783. 8. Wade JC, Newton B, McLaren C, et al. Treatnlent of mucocutaneous herpes simplex virus infection after marrow transplantation with intravenous acyclovir: A double blind trial. Ann Intern Med 1982;96:265. 9. Selby PJ, Jameson B, Watson JG, et al. Parenteral acyclovir therapy for herpes virus infections in man. Lancet 1980;2:1267. 10. Saral R, Burns WH, Laskin OL, et al. Acyclovir prophylaxis of herpes-simplex-virus infections. A randomized. double-blind, controlled trial in bone marrow transplant recipients. N Engl J Med 1981;305:63. 11. Saral R, Ambinder RF, Burns WH, el al. Acyclovir prophylaxis against recrudescent herpes simplex virus infections in leukemia patients: A randomized, double-blind placebo controlled study. Ann Intern Med 1983;99:773. 12. Anderson H, Scarfee JH, Sutton RNP, Hickmott E, Brigden D, Burke C. Oral acyclovir prophylaxis against herpes simplex virus in non-Hodgkin lymphoma and acute lymphoblastic leukaemia patients receiving remission induction chemotherapy. A randomised double blind placebo controlled trial. Br J Cancer 1984;50:45-49. 13. Wade JC, Newton B. Flournoy N, Meyers JD. Oral acyclovir prophylaxis of herpes simplex virus infection after marrow transplant. Ann Intern Med 1984;100:823. 14. Wade JC, Newton B, Flournoy N, et al. Acyclovir for prevention of herpes simplex virus reactivation after marrow transplantation. Ann Intern Med 1984;99:773-776. 15. Saral R. Management of acute viral infections. NCI Monogr 1990;9:107-110. 16. Wingard JR. Viral infections in leukemia and bone marrow transplant patients. Leuk Lymphom 1993;ll :I 15-125. 17. Ambinder RF, Burns WH, Lietman PS, et al. Prophylaxis: A strategy to minimize antiviral resistance. Lancet 1984;1:1154-1155. 18. Angelopulos A, Gore S, Braine H, Humphrey R, Miller C, Rossiter C, Burke PJ. Successful intravenous acyclovir (ACV) prophylaxis against Herpes Simplex Virus (HSV) infection1 reactivation during intensive chemotherapy of leukemia. 35th Interscience Conference on Antimicrobial Agents and Chemotherapy, San Franciso, CA, September 17-20, 1995, abstract # H143. 19. Wade JC, McLaren C, Meyers JD. Frequency and significance of acyclovir-resistant herpes simplex virus isolated from marrow transplant patients receiving multiple courses of treatment with acyclovir. J Infect Dis 1983;148:1077-1082.

20. Burns WH, Santos GW, Saral R, et al. Isolation and characterizations of resistant herpes simplex virus after acyclovir therapy. Lancet 1982;2:421. 21. McLaren C, Chen MS, Ghazzouli I, Saral R, Burns WH. Drug resistance patterns of herpes simplex virus isolates from patients treated with acyclovir. Antimicrob Agent Chemother 1985;28:740-744. 22. Safrin S, Assakeen T, Follansbee S, Mills J. Foscarnet therapy for acyclovir-resistant mucocutaneous herpes simplex virus infection in 26 AIDS patients: Preliminary data. J Infect Dis 1990;161:1078. 23. Safrin S, Crumpacker C, Chatis, et al. A controlled trial comparing foscarnet with vidarabine for acyclovir-resistant mucocutaneous herpes simplex in the acquired immunodeficiency syndrome. N Engl J Med 1991;325:551-555. 24. Shepp DH, Dandliker PS, deMiranda P, et al. Activity of 9-[2-hydroxy-1(hydroxymethyl)ethoxymethyl]guanine in the treatment of cytomegalovirus pneumonia. Ann Intern Med 1985;103:368. 25. Bratanow NC, Ash RC, Turner PA. Successful treatment of serious cytomegalovirus (CMV) disease with 9(1,3-dihydroxy-2-propoxymethy1)guanine(ganciclovir), DHPG) and intravenous immunoglobulin (IVIG) in bone marrow transplant (BMT) patients (abstract no. 254). Exp Hematol 1987;15:541. 26. Emanuel D, Cunninghanl I, Jules-Elysee K, et al. Cytomegalovirus pneumonia after bone marrow transplantation successfully treated with the combination of ganciclovir and highdose intravenous immune globulin. Ann Intern Med 1988;109:777-782. 27. Reed EC, Bowden RA, Dandliker PS, et al. Treatment of cytomegalovirus pneumonia with ganciclovir and intravenous cytomegalovirus immunoglobulin in patients with bone marrow transplants. Ann Intern Med 1988;109:783-788. 28. Schmidt GM, Kovacs A, Zaia JA, et al. Ganciclovirlimmunoglobulin combination therapy for the treatment of human cytomegalovirus-associated interstitial pneumonia in bone marrow allograft recipients. Transplantation 1988;46:905-907. 29. Goodrich JM, Bowden RA, Fisher L, et al. Ganciclovir prophylaxis to prevent cytomegalovirus disease after allogeneic marrow transplant. Ann Intern Med 1993;118:173178. 30. Winston DJ, Ho WG, Bartoni K, et al. Ganciclovir prophylaxis of cytomegalovirus infection and disease in allogeneic bone marrow transplant recipients. Ann Intern Med 1993;118:179184. 31. Schmidt GM, Horak DA, Niland JC, et al. A randomized, controlled trial of prophylactic ganciclovir for cytomegalovirus pulmonary infection in recipients of allogeneic bone marrow transplant patients. N Engl J Med 1991;324:1005-1011. 32. Goodrich JM, Mori M, Gleaves CA, et al. Early treatment with ganciclovir to prevent cytomegalovirus disease after allogeneic bone marrow transplant. N Engl J Med 1991;325:1601-1607. 33. Leland DS, Hansing RL, French ML. Clinical experiences with cytomegalovirus isolation using conventional cell cultures and early antigen detection in centrifugation-enhanced shell vial cultures. J Clin Microbiol 1989;27:1159-1 162. 34. Gleaves CA, Smith TF, Shuster EA, Pearson GR. Comparison of standard tube and shell vial cell culture techniques for the detection of cytomegalovirus in clinical specimens. J Clin Microbiol 1985;21:217. 35. Boeckh M, Bowden RA, Goodrich JM, Pettinger M, Meyers JD. Cytomegalovirus antigen detection in peripheral blood leukocytes after allogeneic marrow transplantation. Blood 1992;80:1358-1364. 36. van der Bij W, Torensma R, van Son WJ, et al. Rapid immunodiagnosis of active cytomegalovirus infection by monoclonal antibody staining of blood leucocytes. J Med Virol 1988;25:179-188. 37. Jiwa NM, van de Rijke FM, Mulder A, et al. An improved immunocytochemical method for detection of human cytomegalovirus antigens in peripheral blood leucocytes. Histochemistry 1989:91:345-349.

38. Spector SA, Merrill R, Wolf D, et al. Detection of human cytomegalovirus in plasma of AIDS patients during acute visceral disease by DNA amplification. J Clin Microbiol 1992;30:2359-2365. 39. Brytting M, Xu W, Wahren B, et al. Cytomegalovirus DNA detection in sera from patients with active cytornegalovirus infections. J Clin Microbiol 1992;30:1937-1941. 40. Einsele H, Steidle M, Vallbracht A, Saal JG, Ehninger G, Muller CA. Early occurrence of human cytomegalovirus infection after bone marrow transplantation as demonstrated by the polymerase chain reaction technique. Blood 1991;77:1104-1110. 41. Nolte FS, Emmens RK, Thurmond C, Mitchell PS. Pascuzzi C, Devine SM, Sara1 R, Wingard JR. Early detection of human cytomegalovirus viremia in bone marrow transplant recipients by DNA amplification. J Clin Microbiol 1995;33:1263-1266. 42. Landry ML, Fergussen D. Comparison of quantitative cytomegalovirus antigenemia assay with culture methods and correlation with clinical disease. J Clin Microbiol 1993;31:28512856. 43. Schmidt CA, Oettle H, Wilborn F, Jessen J, Timm H, Schwerdtfeger R. Oertal J, Siegert W. Demonstration of cytomegalovirus after bone marrow transplantation by polymerase chain reaction, virus culture and antigen detection in buffy coat leukocytes. Bone Marrow Transplant 1994;83:71-75. 44. Goodrich JM, Boeckh M, Bowden R. Strategies for the prevention of cytomegalovirus disease after marrow transplantation. Clin Infect Dis 1994;19:287-98. 45. Forman SJ, Zaia JA. Treatment and prevention of cytomegalovirus pneumonia after bone marrow transplantation: Where do we stand? Blood 1994;83:2392-2398. 46. Devine SM, Wingard JR. Viral infections in severely immunocompromised cancer patients. Support Care Cancer 1994,2:355-368. 47. Singhal S, Mehta J, Powles R, Treleaven J, Horton C, Carrington D, Tryhon Y, Jameson B. Three weeks of ganciclovir for cytomegalovirus after allogeneic bone marrow transplantation. Bone Marrow Transplant 1995;15:777-781. 48. Polis MA, Masur H. Promising new treatments for cytomegalovirus retinitis. JAMA 1995;273:1457-1459. 49. Drew WL, Ives D, Lalezari JP, Crumpacker C, Follansbee SE, Spector SA, Benson CA, Friedberg DN. Hubbard L, Stempien MJ, Shadman A. Buhles W (for the Syntex Cooperative Oral Ganciclovir Study Group). Oral ganciclovir as maintenance treatment for cytomegalovirus retinitis in patients with AIDS. N Engl J Med 1995;333:615-620. 50. Drew WL, Miner RC, Busch DF, et al. Prevalence of resistance in patients receiving ganciclovir for serious cytomegalovirus infection. J Infect Dis 1991;163:716-719. 51. Chou S, Guentzel S, Michels R. Miner RC, Drew WL. Frequency of UL97 phosphotransferase mutations related to ganciclovir resistance in clinical cytomegalovirus isolates. J Infect Dis 1995;172:239-242. 52. Razis E, Cook P, Mittelman A, Ahmed T. Treatment of ganciclovir resistant cytomegalovirus with foscarnet: A report of two cases occurring after bone marrow transplantation. Leuk Lymphom 1994;12:477-480. 53. Meyers JD, Reed EC, Shepp DH, et al. Acyclovir for prevention of cytomegalovirus infection and disease after allogeneic marrow transplantation. N Engl J Med 1988;318:7075. 54. Prentice HG, Gluckman E, Powles RL, et al. Impact of long-term acyclovir on cytomegalovirus infection and survival after allogeneic bone marrow transplantation. Lancet 1994;343:749-753, 55. Bowden RA, Slichter SJ, Sayers MH, Mori M, Cays MJ. Meyers JD. Use of leukocytedepleted platelets and cytomegalovirus-seronegative red blood cells for prevention of primary cytomegalovirus infection after marrow transplant. Blood 1991;78:246-250. 56. Bowden RA, Sayers M, Flournoy N, Newton B, Banaji M, Thomas D, Meyers JD. Cytomegalovirus immune globulin and seronegative blood products to prevent primary cytomegalovirus infection after marrow transplantation. N Engl J Med 1986;314:1006-1010.

57. Mackinnon S, Burnett AK, Crawford RJ, Cameron S, Leask BGS, Sommerville RG. Seronegative blood products prevent primary cytomegalovirus infection after bone marrow transplantation. J Clin Path01 1988;41:948-950. 58. Miller W, McCullough J , Balfour HH, et al. Prevention of CMV infection by blood products: A randomized trial (abstract K128). J Cell Biochem 1988;12(Suppl. C):93. 59. De Witte T, Schattenbereg A, Van Dijk BA, Galama J, Olthuis H, Van Der Meer JWW, Kunst VAJM. Prevention of primary cytomegalovirus infection after allogeneic bone marrow transplantation by using leukocyte-poor random blood products from cytomegalovirusunscreened blood-bank donors. Transplantation 1990;50:964-968. 60. Bowden RA, Slichter SJ, Sayers M, Weisdorf D , Cays M, Schoch G, Banaji M, Haake R , Welk K, Fisher L, McCullough J, Miller W. A comparison of filtered leukocyte-reducted cytomegalovirus (CMV) seronegative blood products for the prevention of transfusionassociated CMV infection after marrow transplant. Blood 1995:85:3598-3603. 61. Winston DJ, H o WG, Lin CH, et al. Intravenous immune globulin for prevention of cytomegalovirus infection and interstitial pneumonia after bone marrow transplantation. Ann Intern Med 1987;106:12-18. 62. Bass EB. Powe NR, Goodman SN, et al. Efficacy of intravenous immune globulin in preventing complications of bone marrow transplantation: A meta analysis. Bone Marrow Transplant 1993;12:273-282. 63. Messori A, Rampazzo R, Scroccaro G, Martini N. Efficacy of hyperimmune anticytomegalovirus immunoglobulins for the prevention of cytomegalovirus infection in recipients of allogeneic bone marrow transplantation: A meta-analysis. Bone Marrow Transplant 1994;13:163-167. 64. Bowden RA, Fisher LD. Rogers K, Cays M, Meyers JD. Cytomegalovirus (CMV)-specific intravenous immunoglobulin for the prevention of primary CMV infection and disease after marrow transplant. J Infect Dis 1991:164:483-487. 65. Guglielmo BJ, Wong-Beringer A, Linkcr CA. Immune globulin therapy in allogeneic bone marrow transplant: A critical review. Bone Marrow Transplant 1994;13:499-510. 66. Sullivan KM. Kopecky KJ. Jocum J. et al. Immunon~odulatoryand antimicrobial efficacy of intravenous imn~unoglobulinin bone marrow transplantation. N Engl J Med 1990;323:705712. 67. Schiller GJ, Nimer SD. Gajewski JL, Golde D. Abdominal presentation of varicella-zoster infection in recipients of allogeneic bone marrow transplantation. Bone Marrow Transplant 1991;7:489-491. 68. Balfour HH Jr, Bean B, Laskin OL, et al. and the Burroughs Wellcome Collaborative Acyclovir Study Group. Acyclovir halts progression of herpes zoster in immunocompromised patients. N Engl J Med 1983;308:1448-1453. 69. Meyers JD, Wade JC, Shepp DH, Newton B. Acyclovir treatment of varicella-zoster virus infection in the compromised host. Transplantation 1984:37:571. 70. Shepp D H , Dandliker PS, Meyers JD. Treatment of varicella-zoster virus infection in severely immunocompromised patients: A randomized comparison of acyclovir and vidarabine. N Engl J Med 1986:314:208-212. 71. McKendrick MW, McGill JI, White JE. Wood MJ. Oral acyclovir in acute herpes zoster. Br Med J 1986:293:1529-1532. 72. Ljungman P, Lonnqvist B, Ringden 0. Skinhoj P. Gahrton G for the Nordic Bone Marrow Transplant Group. A randomized trial of oral versus intravenous acyclovir for treatment of herpes zoster in bone marrow transplant recipients. Bone Marrow Transplant 1989:4:613615. 73. Boivin G. Edleman CK, Pedneault L. Talarico CL, Biron KK, Balfour HH Jr. Phenotypic and genotypic characterization of acyclovir-resistant varicella-zoster viruses isolated from persons with AIDS. J Infect Dis 1994:170:68-75. 74. Safrin S, Berger T G , Gilson I, et al. Foscarnet therapy in five patients with AIDS and acyclovir-resistant varicella-zoster virus infection. Ann Intern Med 1991;115:19-21.

75. Zaia JA, Levin MJ, Preblud SR, et al. Evaluation of varicella-zoster immune globulin: Protection of immunosuppressed children after household exposure to varicella. J Infect Dis 1983;147:737. 76. Gershon AA. Live attenuated varicella vaccine: Perspective. J Infect Dis 1985;152:859. 77. Gershon AA, Steinberg SP, and the Varicella Vaccine Collaborative Study Group of the NIAID. Persistence of immunity of varicella in children with leukemia immunized with live attenuated varicella vaccine. N Engl J Med 1989;320:892. 78. Gershon AA, Steinberg SP, and the National Institute of Allergy and Infectious Diseases Varicella Vaccine Collaborative Study Group. Live attenuated varicella vaccine: Protection in healthy adults compared with leukemic children. J Infect Dis 1990;161:661. 79. Lawrence R, Gershon AA, Holzman R, et al. The risk of zoster after varicella vaccination in children with leukemia. N Engl J Med 1988;318:543. 80. Ballerini P, Gaidano G, Inghirami G, Gong JZ, Saglio G, Knowles DM, Dalla-Favera R. Multiple genetic alterations in AIDS-associated lymphomas. Blood 1991;78:327. 81. Knowles DM, Cesarman E, Chadburn A, et al. Correlative morphologic and molecular genetic analysis demonstrates three distinct categories of posttransplantation lymphoproliferative disorders. Blood 1995;85:552-565. 82, Carrigan DR, Drobyski WR, Russler SK, Tapper MA, Knox KK, Ash RC. Interstitial pneumonitis associated with human herpesvirus six (HHV-6) infection in marrow transplant patients. Lancet 1991;338:147. 83. Cone RW, Hackman RC, Huang MW, et al. Human herpesvirus 6 in lung tissue from patients with pneumonitis after bone marrow transplantation. N Engl J Med 1993;329: 156. 84. Drobyski WR, Knox KK, Majewski D, Carrigan DR. Fatal encephalitis due to variant B human herpesvirus 6 infection in a bone marrow transplant recipient. N Engl J Med 1994;330:1356. 85. Drobyski WR, Dunne WM, Burd EM. Human herpesvirus 6 (HHV-6) infection in allogeneic bone marrow transplant recipients: I. Evidence for a marrow suppressive role for HHV-6 in vivo. J Infect Dis 1993;167:735. 86. Carrigan DR, Knox KK. Human herpesvirus 6 (HHV-6) isolation from bone marrow: HHV6-associated bone marrow suppression in bone marrow transplant patients. Blood 1994;84:3307-33 10. 87. Burns WH, Sandford GR. Susceptibility of human herpesvirus 6 to antivirals in vitro. J Infect Dis 1990;162:634-637. 88. Quinnan GV Jr., Kirmani N, Rook AH, et al. Cytotoxic cells in cytomegalovirus infections. HLA-restricted T-lymphocyte and non-T-lymphocyte cytotoxic responses correlate with recovery from cytomegalovirus infection in bone marrow transplant recipients. N Engl J Med 1982;307:7. 89. Reusser P, Riddel SR, Meyers SD, Greenberg PD. Cytotoxic T-lymphocyte response to cytomegalovirus after human allogeneic bone marrow transplantation: Pattern of recovery and correlation with cytomegalovirus infection and disease. Blood 1991;78:13731380. 90. Quinnan GV Jr, Burns WH, Kirmani N, et al. HLA-restricted cytotoxic T lymphocytes are an early immune response and important defense mechanism in infections. Rev Infect Dis 1984;6:156-163. 91. Riddel SR, Walter BA, Gilbert MJ, Greenberg PD. Selective reconstitution of CD8 cytotoxic T lymphocyte responses in immunodeficient bone marrow transplant recipients by the adoptive transfer of T ceH clones. Bone Marrow Transplant 1994;14:S78-S84. 92. Riddell SR, Watanabe KS, Goodrich JM, Li CR, Agha ME, Greenberg PD. Restoration of viral immunity in immunodeficient humans by the adoptive transfer of T-cell clones. Science 1992;257:238. 93. Riddell SR, Reusser P, Greenberg PD. Cytotoxic T-cells specific for cytomegalovirus: A potential therapy for immunocompromised hosts. Rev Infect Dis 1991;13:966. 94. Walter EA, Greenberg PD, Gilbert MJ, Finch RJ, Watanabe KS, Thomas ED, Riddell SR.

Reconstitution of cellular immunity against cytomegalovirus in recipients of allogeneic bone marrow lransplant by transfer of T-cell clones from the donor. N Engl J Med 1995;333:10381044. 95. Papadopopoulos EB. Ladanyi M, Emanuel D, et al. Infusions of donor leukocytes to treat Epstein-Barr virus-associated lymphoproliferative disorders after allogeneic bone marrow transplantation. N Engl J Med 1994;330:1185-1 191. 96. Davis GL, Balart LA, Schiff EK, et al. Treatment of chronic hepatitis C with recombinant interferon alpha: A multicenter randomized, controlled trial. N Engl J Med 1989:321:15011506. 97. Di Bisceglie AM, Martin P, Kassianides C, et al. Recombinant interferon alpha therapy for chronic hepatitis C. A randomized, double-blind, placebo-controlled trial. N Engl J Med 1989:321:1506-1510. 98. Terrault N, Wright T. Interferon and hepatitis C. N Engl J Med 1995;332:1509-1511. 99. Greenberg HB, Pollard RB, Lutwixk LI, et al. Effect of hurnan leukocyte interferon on hepatitis B virus infection on patients with chronic active hepatitis. N Engl J Med 1976;295:517. 100. Reed EC, Myerson D. Corey L, Meyers JD. Allogeneic marrow transplantation in patients positive for hepatitis B surface antigen. Blood 1991;77:195-200. 101. Locasciulli A. Bacigalupo A, VanLint MT. Cavalletto D, Pontisso P. Testa M. Masera G, Shulman HM, Portmann B, Alberti A . Hepatitis C virus infection and liver failure in patients undergoing allogeneic bone marrow transplantation. Bone Marrow Transplant 1995:16:40741 1. 102. Fujii Y, Kaku K, Tanaka M, et al. Hepatitis C virus infection and liver disease after allogeneic bone marrow transplantation. Bone Marrow Transplant 1994:13:523-526. 103. Shuhart MC, Myerson D, Childs BH. Marrow transplantation from hepatitis C virus seropositive donors: Transmission rate and clinical course. Blood 1994:84:3229-3235. 104. Frickhofen N, Wisneth M, Jainta C, et al. Hepatitis C virus infection is a risk factor for liver failure from veno-occlusive disease after bone marrow transplantation. Blood 1994;83:19982004. 105. Ljungman P, Johansson N, Aschan J, Glaumann H, Liinnqvist I3, Ringden 0. Sparrelid E , Sonnerborg J, Winiarski J, Gahrton G. Long-term effects of hepatitis C virus infection in allogeneic bone marrow transplant recipients. Blood 1995:86: 1614-161 8, 1995. 106. Ilan Y. Nagler A, Adler R , Tur-Kaspa R. Slavin S, Shouval D. Ablation of persistent hepatitis B by bone marrow transplantation from a hepatitis B-immune donor. Gastroenterology 1993;104:1818-1821. 107. Flomenberg P. Babbitt J, Drobyski WR, et al. Increasing incidence of ade~lovirusdisease in bone marrow transplant recipients. J Infect Dis 1994;169:775-781. 108. Engelhard D, Nagler A, Hardan I, ct al. Antibody response to a two-dose regimen of influenza vaccine in allogeneic T cell-depleted and autologous BMT recipients. Bone Marrow Transplant 1993;ll:l-5. 109. Wingfield WL, Pollack D, Grunert RR. Therapeutic efficacy of amantadine HC1 and rimantidine HC1 in naturally occurring influenza A2 respiratory illness in man. N Engl J Med 1969;281:579. 110. Hertz MI, Englund JA, Snover D, et al. Respiratory syncytial virus-induced acute lung injury in adult patients with bone marrow transplants: A clinical approach and review of the literature. Medicine 1989:68:269-281. 111. Smith DW, Frankel LR. Mathers LH, et al. A controlled trial of aerosolized ribavirin in infants receiving mechanical ventilation for severe respiratory syncytial virus infection. N Engl J Med 1991;325:24-29. 112. Whimbey E , Champlin RE. Englund JA, Mirza NQ, Piedra PA, Goodrich JM, Przepiorka D. Luna MA, Morice RC, Neulnann JL. Elting LS, Bodey GP. Combination therapy with aerosolized ribavirin and intravenous imn~unoglobulinfor respiratory syncytial virus disease in adult bone marrow transplant recipients. Bone Marrow Transplant 1995;16:393399.

113. E~lglundJA, Anderson LJ, Rhame FS. Nosocomial transmission of respiratory syncytial virus in immunocompromised adults. J Clin Microbiol 1991;29:115-119. 114. Whimbey E, Elting LS, Cough RB, Lo W, Williams L, Champlin RE, Bodey GP. Influenza A virus infections among hospitalized adult bone marrow transplant recipients. Bone Marrow Transplant 1994:13:437-440. 115. Arthur RR, Shah KV, Baust SJ, Santos GW, Saral R. Association of BK viruria with hemorrhagic cystitis in recipients of bone marrow transplants. N Engl J Med 2986;315:230234. 116. Russell SJ. Vowels MR. Vale T. Haemorrhagic cystitis in paediatric bone marrow transplant patients: An association with infective agents. GVHD and prior cyclophosphamide. Bone Marrow Transplant 1994;13:533-539. 117. Azzi A, Fanci R, Bosi A, et al. Monitoring of polyomavirus BK viruria in bone marrow transplantation patients by DNA hybridization assay and by polymerase chain reaction: An approach to assess the relationship between BK viruria and hemorrhagic cystitis. Bone Marrow Transplant 1994;14:235-240. 118. Schneider EM. Dorries. High frequency of polyomavirus infection in lymphoid cell preparations after allogeneic bone marrow transplantation. Transplant Pro 1983;25:1271-1273. 119. Drummond JE. Shah KV, Saral R, Santos GW, Donnenberg AD. BK virus specific humoral and cell mediated immunity in allogeneic bone marrow transplant (BMT) recipients. J Med Virol 1987;23:331-344.

8. Cytokines and biological response modifiers in the treatment of infection Brigitta U. Mueller and Phillip A. Pizzo

1. Introduction Patients undergoing myelosuppressive cancer therapy (chemotherapy or radiation) are at an increased risk to develop potentially life-threatening infections. The incidence appears to be strongly correlated with the duration and degree of absolute neutropenia [I].A neutrophil count of 4 0 0 cells/mmi, and especially of
2. Cytokines Numerous interleukins (IL), colony-stimulating factors (CSF), as well as interferons (IFN) and tumor necrosis factor (TNF), all with a biologic response modifier action, have been classified as cytokines. The nomenclature has been changed from a descriptive system based on perceived biological action (e.g., TNF) to a more rationale system using sequential numbering (e.g., Gat y A . N o ~ k i l l( e d ) , M A N A G E M E N T O F I N F E C T I O U S C O M P L I C A T I O N S IN C A N C E R P A T I E N T S . O 1998. Kllr~vrrAcatlrnzic Puhlrsilcr.,, Bo,tor~.All figllt\ r e ~ e r v e d

IL-1, IL-2, etc.). In the last few years it has become clear that most cytokines have multiple functions, determined in part by the type of the target cell, the kind of disease triggering a response, and the host itself [lo]. Table 1 lists the cytokines that are currently known to play a role in infectious diseases. The myelopoietic growth factors, including G-CSF, GM-CSF, rnonocyte colony-stimulating factor (M-CSF), and interleukin-3 (IL-3), increase the number of circulating, activated neutrophils, similar to the physiological reaction triggered by an acute infection [11,12]. Several other cytokines have a strong immunomodulatory effect, normalizing an impaired, hyperactive, or deficient immune response, thus improving the host's defenses against infections. These include IL-1, IL-2, IL-4, IL-6, and IL-12, as well as TNF, all of them affecting both T- and B-cell function [13-161. 2.1 Colony-stimulating factors The major source of endogenous G-CSF, GM-CSF, IL-3, and M-CSF are monocytes, T cells, fibroblasts, and endothelial cells (See Table 1) 1121.

2.1.1 Granulocyte colony-stimulating factor (G-CSF). The gene for G-CSF, a glycoprotein produced by T lymphocytes, macrophages, endothelial and epithelial cells, as well as fibroblasts, is located on chromosome 17. The action of G-CSF, resulting in stimulation of growth and differentiation, is directed towards hematopoietic progenitor cells that are already committed to the neutrophil lineage [17,18]. In addition, G-CSF enhances the antimicrobial activity of mature circulating neutrophils by augmenting the superoxide production and increasing phagocytosis and che~notaxis[19-211. Neutrophils produced in response to G-CSF express neutrophil Fcg IgG receptors (FcgRI, CD64 antigen), enabling antibody-dependent cellular cytotoxicity in an FcgRI-dependent way [22]. Numerous studies evaluating G-CSF have been performed both in adults and children with bone marrow suppression secondary to either chemotherapy or disease. The role of G-CSF in decreasing the number of febrile, neutropenic days has been investigated in a double-blind, placebo-controlled study in patients with fever (>38.2"C) and neutropenia (neutrophil count
2 «

CO

73

ID ^

2


S

••
5 K

a 2? oc

O

O

o

JJ C3

2"0

•g g T3 -C u

2 g S - a ^ ^ 5 S'^ §^6

>

CL,

H

<3 J^

C T3 O C

»-i|I •g cj ^ E .2 -^ U &.•- —
Q U O

.E^ 73

K

R2

^

S

^J

U

CB T 3 E ir

C

G,,C

II

•—

^feEigsfo'S^

I S S .§ 1 5 'o d

i: c o

IS "^

5 «-S

o 2

^E^l

^

in

111 c.o

E C

|_

*-

r-.

3 R S S E

S -g •£ ?.

C •^

o

ra c

3

.3 E

O

E

e

'•5

C1.X!

1)

c .2 73 g -j^ c

3

0

00 3 3 £ C 00

p ra

o .ii C

U

1 \-< o C

00

3

00 9 - ?^

S.SE.I^:?. S

G.

1
"o E ,E o.

^

13 i" c C

1)

0)

CO

"o St:

c

OH'-O

"a 3

O C

,2

aj 00

o Jn!3 o O C O

O


CO C3

6Z

a E E

CO

:s 2 ^ a. °o. a. 9 P —

u

O —

J- u

3

u

till. ^

o •£ TJ S

i ^

C

D

U

III

"S

~W

^-1 C X) U

OJ c O C

00

lis

C

E C

H

00

00

[i,

Itilii 3

CB -TT

« S O

a

^o

CO

ililS

-a

lie

x ; >,JS . c/5 D. c 3 g CJ O O C S V

203

c "5 O


i _«

, 'tA

S1-- l

„ .y !U O

o

gJlt

a

iS

ii S i. 1 ^ 1.^^ 2

•£ "3

U

^ "I C

u^ J^ - 5 a

O T3

1 T3

lti.^il

>

> £

C

S3

" e 3

E 2

i:

^

O S TJ .S C

E

2

E -c

O

y -a 3 y -a 3 o 3

E m E 3

ca

= x: 2 -T3cora'a!jsc..r!o

s^l^i

/

3

r-

3

S V

c/5

S U

3 3-5

UH

1^^

..II

u

>N'3

S " E § S S^ o

aj

0 ^

>^ C

i2 E o 3 O -7

O

^

"J

id

> S

tC ao

o —

<

u 1)

J JS o j5 ^

2 2 1 1 1I-' ra '-I3 ^ '

3

„•

>

1)

02.

ao E

204

E a< cfl
1^ 3-11 i |

u O

»

S >:, ii 3 = J^ b CL

'3 j3
- §• ^ i s &

I ^

oB^

3 E tS

J

TJ

c CD o c .3 ):; 3 S 2 'S CJ E

^ J

5 2 tu >.& Z pffl-H 3 -

^ >< ^3

1 1 s 1

03

^

g J : O, *: n

E

•2 O D

ii '3 ^

-

.C% gz 3 b . 5 - 4 9 gysp+

." d

m

C

~

~g g3 ?jk&k h 9 C Z

cl;

-7 ,gE .6 o u G

kIL2 23 L 3 -

'C1

2

c

CI

s s .E

patients receiving intensive induction chemotherapy for inflammatory breast cancer [26]. This study also found a statistically significant decrease in the incidence of microbiologically documented infections and a decreased need for rehospitalization for infectious complications. G-CSF has been studied in patients who are expected to have a prolonged time to bone marrow recovery. A Japanese study in patients undergoing intensive induction chemotherapy for relapsed or refractory leukemia demonstrated an accelerated recovery of neutrophil counts, shortened by about a week, and a significantly decreased incidence of infectious episodes in the patient group treated with G-CSF [27]. In a pilot study in children undergoing allogeneic bone marrow transplantation, G-CSF combined with erythropoietin resulted not only in a reduction of the time to neutrophil engraftment, but this subsequently lead also to a decrease in the number of febrile episodes [28]. G-CSF has also been used as a preventive mode in neutropenic states caused by the underlying disease, such as cyclic neutropenia, infection with the human immunodeficiency virus, congenital neutropenia, and myelodysplastic syndromes [29-361. The depth and duration of neutropenia can clearly be decreased by the use of G-CSF; however, whether this therapy has an impact on the frequency of infectious complications is more difficult to assess. Many of these diseases require prolonged, if not lifelong support with G-CSF to maintain an adequate neutrophil count, and the efficacy is often limited while the risk for side effects is potentially increased. Especially in myelodysplastic diseases, there is concern about a possible increased risk for the development of leukemia, especially acute myelocytic leukemia [37-391. The timing of G-CSF administration has been studied not only because of the potential therapeutic advantage in increasing the sensitivity of leukemic cells to chemotherapy, but also because of the concern that the pool of vulnerable precursor cells that are then exposed to chemotherapeutic agents may be increased. In a trial of oral fluorouracil(5-FU), the concurrent administration of G-CSF resulted in an unusually high incidence of severe neutropenia, which was not seen when G-CSF treatment was delayed until after the administration of 5-FU [40]. Other studies raised similar concerns 141-431. G-CSF administration is therefore commonly stopped 24-48 hours prior to cytotoxic chemotherapy and reinstituted 24-72 hours after the ablative chemotherapy. How long to continue the administration of G-CSF in patients with chemotherapy-induced neutropenia is even less well defined. The package insert for filgrastim recommends to continue G-CSF administration until an absolute neutrophil count of >10,000 c e l l s / m m ~ a sbeen reached; however, costs, patient comfort, and potential benefits have to weighed carefully. Based on one of the initial studies in patients with small cell lung cancer, a subcutaneous dose of 5-20pgIkg daily is commonly used, although higher doses have safely been administered as well [25,44-46]. G-CSF can also be administered as an albumin-containing intravenous infusion. The dosage, tolerance, and activity in children does not appear to be different from adults

[47]. Side effects associated with G-CSF are relatively rare. Medullary bone pain has been described in 15-39% of the patients, compared with 0-21% of patients receiving placebo and appears to be dose related [23,25,45]. Less frequent side effects that have been described include a painful, localized neutrophilic infiltration of the skin (Sweet syndrome) [48,49]; exacerbations of pre-existing inflammatory skin disorders, such as psoriasis, excema, and vasculitis [35,50-531; and rare allergic reactions [54]. A decrease in platelet counts has been observed in several studies which may limit the benefits in higher dose-intensity chemotherapeutic regimens [55-571. With chronic administration, mild alopecia, splenomegaly, and, more worrisome, the potential for a disseminated intravascular coagulopathy have been described [58,59].

2.1.2 Granulocyte-macrophage colony-stimulating factor (GM-CSF). GMCSF has broader activity than G-CSF, and subcutaneous administration for several days results not only in an increase in the number of circulating neutrophils but also in eosinophilia and monocytosis, and an increased number of circulating stem cells [18]. GM-CSF can modulate the function of terminally differentiated neutrophils and enhances the phagogocytic activity against bacteria and yeasts [12,60-641. Recombinant human GM-CSF has been evaluated in several studies, and overall results in a shortened period of neutropenia after myelotoxic chemotherapy and a reduced incidence of infections [65-721. In two randomized studies of patients treated with intensive chemotherapy for non-Hodgkin's lymphoma, the group treated with adjunctive GM-CSF experienced significantly less infections, including severe infections, less antibiotics, and shorter hospitalizations [70,72]. Similar results were found in patients treated with cisplatin and etoposide or vinblastine for solid tumors during the first treatment cycle; however, no significant difference was demonstrated for later cycles [68]. In a study of patients with head and neck cancer treated with cisplatin, 5-FU, and leucovorin, a significant decrease in the incidence of mucositis was observed when GM-CSF was given [73]. A decrease of grade 111-IV mucositis and infections has been described in patients undergoing allogeneic or autologous bone marrow transplantation [74-781. Recently, a double-blind, randomized study in febrile, neutropenic children treated with imipenem with or without GM-CSF demonstrated a significantly decreased number of episodes requiring antimicrobial therapy for longer than 10 days in the GM-CSF group (12% compared with 50% in the placebo group) [79]. Two problems have been associated with GM-CSF that limit its usefulness in the treatment of patients with cancer. Most patients experience mild fever due to the growth factor, often necessitating the empirical use of antibiotics [67,69,80]. Although the period of neutropenia is shortened, thrombocytopenia often becomes more pronounced and prolonged, prohibiting further dose intensification of the chemotherapeutic regimen [67,69]. Other side effects include headaches, bone pain, myalgia, and chills

[8,66,69,81,82]. Dyspnea, probably due to a transient pulmonary hyperleukocytosis and fluid retention, have been observed with the first few doses of GM-CSF. The recommended dose for GM-CSF is 250 yglm21d, given either subcutaneously or as an intravenous infusion over 2 hours. 2.1.3 Macrophage colony-stimulating factor (M-CSF). M-CSF stimulates the growth of monocytelmacrophage colonies in vitro and results in a marked increase in circulating monocytes when given intravenously [83-851. When mice are treated with M-CSF, they are protected from the effects of a subsequent lethal challenge with Cnndida albicnns [86]. Patients who received MCSF show enhanced monocyte function, demonstrated by an oxidative burst, migration, phagocytosis, and killing of C. albicnns [87]. Nemunaitis et al. conducted a phase I dose escalation trial with M-CSF in bone marrow transplant recipients with invasive fungal infections [85,88]. MCSF was well tolerated at doses of 100-2000pg/m21d,and the only side effect was a dose-related, transient thrombocytopenia. Survival of patients receiving M-CSF was greater than that of historical controls (27% compared with 5 % ) , mainly because of a better survival rate of patients with C. albicnns infection. However, no benefit was noted in patients with Aspergillus infections. 2.2 Interleukins

Two members of the interleukin family have, in addition to immunostimulatory activities, a direct effect on hematopoiesis and have potential applications in the treatment of infectious episode in the cancer patient. Interleukin-1 (IL-1) and interleukin-3 (IL-3) appear to have the capability of inducing myeloid recovery and enhancing myeloid function [89-911.

2.2.1 Interleukin-1 (IL-1). The three structurally related polypeptides - ILl a , IL-lb, and an IL-1 receptor antagonist - are synthesized by a variety of cells. IL-1 (a and b) acts as an endogenous pyrogen and can also induce sleep, hypotension, and anorexia [89,92]. In vitro, IL-1 has been shown to induce the production of GM-CSF, G-CSF, M-CSF, IL-3, and IL-6 [93-971. The administration of recombinant 1L-1 (murine or human) to rodents is followed by an increased number of circulating granulocytes, an increase in granulocytic progenitor cells in spleen and bone marrow, as well as an increase in platelet counts and megakaryocytes [98-1001. There is also evidence that IL-3 may have a myeloprotective function in mice and rats treated with chemotherapeutic agents or radiotherapy [98,101,102]. IL-1 enhances the function of neutrophils, including migration, phagocytosis, and the production of 0, and H,O, [103]. The survival rate of neutropenic and non-neutropenic mice challenged with Pseudomonns neruginosa or Klebsielln pneumonine increased if IL-la was administered prior to or very early during the infection [104,105]. When highly purified recombinant IL-la and IL-lb (rHuIL-la and rHuILl b ) became available for human use, several investigators explored the poten-

tial applications of IL-1 in the treatment of adults with cancer. Local inflammatory reactions or fever, rigors, and chills occurred in the majority of patients, and influenza-like symptoms, manifested by generalized fatigue, arthralgia, and headaches, were common [85,89,106]. Hypotension, myocardial infarction, confusion, severe abdominal pain, and renal insufficiency were dose limiting in a phase I trial of IL-la given daily as an intravenous infusion over 15 minutes [loti]. Nemunaitis et al. demonstrated a beneficial effect of IL-lb on myeloid recovery after autologous bone marrow transplantation for acute myelogenous leukemia [85]. The time to reach an absolute neutrophil count of >500 cells/mm3 was 25 days compared with 34 days in a historical control group (P = 0.02), and this correlated with a decrease in the rate of infections (12% vs. 23%; P = 0.049) and increased survival (30% vs. 20%; P = 0.04). In another study, IL-la was administered to patients with advanced solid malignancies, and a significant, dose-related increase in white blood cells, as well as platelets, was observed [106]. Because of the common side effects of IL-1, especially at higher doses, combinations with indomethacin or other antiinflammatory agents are being explored [106]. The potential myeloprotective effect for chemotherapy and radiotherapy warrants further studies. Other possible clinical indications might include treatment of invasive fungal diseases in cancer patients [1071091 or the combination of IL-1 with other cytokines, such as M-CSF, to enhance its bone marrow-stimulating effect [97,110].

2.2.2 Interleukin-3 (IL-3). IL-3, also called multi-CSF, promotes the production and differentiation of several different bone marrow progenitor cell types, and also enhances the function of mature myeloid cells [ I l l ] . Its activity is dependent, to a large degree, on interactions and the presence of other cytokines, especially GM-CSF, IL-1, and IL-6 [112,113]. A small phase I study in patients with small cell lung cancer found IL-3 to be well tolerated at doses up to 10yg/kg/d [90]. A recombinant fusion protein, combining IL-3 and GM-CSF (PIXY 321), is currently being evaluated for its neoplastic and myeloproliferative effects [114,115]. 2.3 Interferons Of the three species of interferons (a, (3, y), only IFN-y has some promise for the treatment of infectious complications, whereas both IFN-a and IFN-fi are used in the treatment of hepatitis. IFN-a has also been used in the treatment of Kaposi's sarcoma, a malignancy that possibly has an infectious origin. IFN-91 enhances macrophage activity against intracellular organisms as well as neutrophil and monocyte function against bacteria and fungi [116,117]. IFN-TIis an approved therapy in patients with chronic granulomatous disease [16,118-1201 and is being explored for its use in the treatment of infections with the Mycobacteriutn nviunz-intracellulare complex in patients with HIV infection [121]. However, in the treatment of patients with cancer, there is

currently no standardized indication for the use of IFN-y, although its antifungal activity may be of potential use.

3. Immunization Active immunization plays a minor role in the management of infectious complications in patients with cancer. These patients have often impaired Band T-cell function and are therefore not able to mount a protective antibody response. Adequate antibody responses to polysaccharide vaccines (pneumococcal, meningococcal, and anti-Hnernophilus vaccines) are difficult to maintain because of the repeated immunosuppression through subsequent cycles of chemotherapy [122,123]. However, pneumococcal vaccine (and potentially influenza vaccine) are recommended in patients before bone marrow transplantation and in patients with functional or anatomical asplenia, for example, in Hodgkin's disease [124]. A live attenuated varicella vaccine has recently been licensed by the U.S. Food and Drug Administration and appears to result in a persistent immunity in children with leukemia. It is currently not recommended for use in immunosuppressed patients because of the possibility of vaccine-related morbidity [125,126]. Passive immunization can be achieved with polyclonal immunoglobulin preparations (e.g., intravenous immunoglobulins [IVIG]), with immunoglobulin preparations containing high titers of specific antibodies (e.g., varicella zoster immunoglobulin [VZIG]), or with monoclonal antibodies. 4. Other biological response modifiers Another group of biological response modifiers include imidazoles, such as levamisole, which appears to augment delayed hypersensitivity to recall agents and may have a role in the treatment of opportunistic infections in patients with congenital or acquired T-cell deficiency states, as well as in some viral infections [127]. Diethyldithiocarbamate (DTC) and N-acetyl-cysteine are thiols that restore both T- and B-cell function and induce T-cell differentiation. DTC appears to have some activity in the treatment of tuberculosis and was shown to reduce infections in surgical patients [128,129]. Agents other than the above-mentioned cytokines that have an effect on the immune system include corticosteroids, nonsteroidal antiinflammatory agents, thymic hormones, as well as specific T-cell immunosuppressants (e.g., cyclosporin A), or even some chemotherapeutic agents (e.g., cyclophosphamide). Neither of these drugs is currently thought to have a role in the treatment or prevention of infectious complications in cancer patients.

5. When are biological response modifiers indicated? The high costs and the potential risk for known as well as yet unknown side effects mandate a judicious use of cytokines. The American Society of Clinical

Oncology (ASCO) based their recommendations on an extensive review of available data and considered both evidence for activity and costs [8]. As new cytokines or other biological response modifiers become available, it is important that the same degree of critical appraisal and cost-benefit analyses are applied.

5.1 Chemotherapy-associated neutropenia There are three potential applications of cytokines, especially hematopoietic growth factors, in cancer patients undergoing ablative therapy. The most commonly used indication is as preventive therapy (primary administration), and less commonly as adjuvant or pathogen-directed therapy in the event of fever and neutropenia, or in the presence of an established infection (secondary administration). Primary administration of CSFs is recommended when the likelihood of febrile neutropenia is >40% because it has been shown that a 50% reduction in incidence can be achieved under these circumstances [25,45,130,131]. Administration of CSFs is also indicated to try to avoid further infections after a documented episode of febrile neutropenia in a prior cycle, if a dose reduction in the chemotherapeutic regimen is not deemed appropriate [132]. In addition, there may be special circumstances warranting the primary use of cytokines. Patients who have certain risk factors, such as pre-existing disease-related neutropenia, extensive prior chemotherapy, or a history of radiotherapy to the pelvis or long bones, may benefit from such supportive care [133,134] For the majority of adult patients with febrile neutropenia, available data do not clearly support the initiation of therapy with CSFs as adjuncts to antibiotic treatment [79,135]. A recent double-blind, placebo-controlled trial in 218 adult cancer patients with febrile neutropenia demonstrated a significant decrease in the number of neutropenic days and the time to resolution of febrile neutropenia, but not in the number of days with fever [23]. However, the most important difference was the fact that the number of days with profound neutropenia (
gal therapy if the granulocyte count remains low for more than 1-2 weeks [140]. In addition to the use of G-CSF or GM-CSF to decrease the depth and duration of neutropenia, there may be an additional benefit because of the capability of these agents to enhance the antifungal activity of neutrophils [20,141]. Trials are ongoing to determine the role of the transfusion of G-CSFinduced donor-derived neutrophils, M-CSF given directly to the patient, or the transfusion of M-CSF activated, elutriated monocytes [85].

5.2 Transplantation The ASCO guidelines support the primary administration of CSFs (especially GM-CSF) in patients receiving high-dose chemotherapy with autologous progenitor-cell transplantation as well as for the mobilization of peripheral blood progenitor cells because studies have demonstrated both an accelerated recovery of myeloid cells as well as a decreased incidence of bacterial infections [75,77]. M-CSF has been used for the treatment of invasive fungal infections in marrow transplant recipients, but the preliminary results are inconclusive [85].

5.3 Myeloid malignancies and myelodysplastic syndrome The incidence of fever and infection is high in patients undergoing induction therapy for acute myeloid leukemia, and fungal infections are a common problem in patients with myelodysplastic syndromes [138]. Theoretically, CSF administration could shorten the duration of severe neutropenia and diminish the risk for infections [32,35,142]. In vitro and in vivo data also appear to demonstrate a "priming effect" of CSFs on leukemic cells if administered before andlor concurrently with chemotherapeutic agents, enhancing drug sensitivity to S-phase-specific agents such as cytarabine [17,66,143-1471. However, these patients have an increased potential for adverse side effects with CSF administration. The major concern is the occurrence of postchemotherapy leukemia induced by CSFs, based on the observation that most leukemic cells have CSF receptors as well [148]. Similar concerns apply to myelodysplastic syndromes, which carry an inherent risk of conversion into a myeloid leukemia [37-39,1491. The guidelines issued by ASCO do not give a strong recommendation for the use of CSFs in myelodysplastic syndromes or the treatment of acute myeloid leukemia but conclude that their use is probably not detrimental and possibly beneficial in special circumstances, such as patients with recurrent bacterial infections [8]. However, the use of CSFs as "priming" agents should be restricted to the controlled research setting. 5.4 Pediatric indications

Most neoplastic diseases occurring in children are being treated as part of a research protocol and commonly include intensive myelosuppressive chemo-

therapy [150]. Although randomized studies on the use of cytokines have been mainly limited to the adult population, the guidelines recommended for adults are generally applicable to the pediatric age group as well [151].

6. Conclusions and future directions Infections remain a major cause for morbidity in the patient with cancer. Colony-stimulating factors are able to shorten the depth and duration of severe neutropenia, especially the time with an absolute neutrophil count below 100 cells/mm3. However, not every chemotherapeutic regimen will render patients neutropenic, and the costs associated with the use of cytokines make their rationale use mandatory. The guidelines issued by ASCO represent a thorough and well-documented outline for the cost-conscious and appropriate use of these potent agents [a]. More research is needed to evaluate the role of cytokines as means to increase the dose intensity of chemotherapeutic regimens and their potential use as immunomodulatory agents that enhance and restore the body's defense system against infections.

References 1. Bodey GP. Buckley M. Sathe YS, Freireich EJ. Quantitative relationships between circulating leukocytes and infections in patients with acute leukemia. Ann Intern Med 1966:64:328340. 2. Pizzo PA. Infectious conlplications in the child with cancer. I. Pathophysiology of the compromised host and the initial evaluation and management of the febrile cancer patient. J Pediatr 1981;98:341-354. 3. Pizzo PA. Considerations for the prevention of infectious complications in patients with cancer. Rev Infcct Dis 1989:11(Suppl. 7):s 1551-S1563. 4. Richet HM, Andreniont A, Tancrede C, Pico JL, Jarvis WR. Risk factors for candidemia in patients with acute lymphocytic leukemia. Rev Infect Dis 1991:13:211-215. 5. Wingard J R , Santos GW, Sara1 R. Differences between first and subsequent fevers during prolonged neutropenia. Cancer 1987;59:844-849, 6. Chanock S. Evolving risk factors for infectious complications of cancer therapy. In: Pizzo PA, ed. HematologyiOncology Clinics of North America. Philadelphia: W.B. Saunders, 1993;7:4:771-794. 7. Frost JD, Sonde1 PM. Immunotherapy for infection and malignancy in children with cancer. Adv Pediatr 1994:41:385413. 8. American Society of Clinical Oncology. American Society of Clinical Oncology recommendations for the use of hematopoietic colony-stimulating factors: Evidence-based, clinical practice guidelines. J Clin Oncol 1994;12:2471-2508. 9. Yim JM, Matuszewski KA. Vermeulen LC Jr., Ratko TA, Burnett D A , Vlasses PH. Surveillance of colony-stimulating factor use in US academic health centers. Ann Pharmacother 1995;29:47548 1. 10. Paul WE, Seder RA. Lymphocyte response and cytokines. Cell 1994;76:241-251. 11. Groopman JE. Molina J-M, Scadden DT. Hematopoietic growth factors. Biology and clinical applications. N Engl J Med 1989;321:1449-1459. 12. Roilides E, Pizzo PA. Biologicals and hematopoietic cytokines in prevention or treatment of

13. 14. 15. 16. 17.

18. 19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

infections in immunocompromised hosts. In: Pizzo PA, ed. Infectious Complications in the Immunocompromised Host. Philadelphia: W.B. Saunders, 1993;7:4:841-864. Bocci V. Interleukins. Clinical pharmacokinetics and practical implications. Clin Pharmacokinet 1991;21:274-284. O'Garra A. Interleukins and the immune system 1. Lancet 1989;1:943-947. O'Garra A. Interleukins and the immune system 2. Lancet 1989;1:1003-1005. Hill HR. Modulation of host defense with interferon-g in pediatrics. J Infect Dis 1993;167(Suppl. 1):S23-S28. Souza LM, Boone TC, Lai PH, Zsebo KM, Murdock DC, Chazin VR, Bruszewski J, Lu H, Chen KK, Barendt J, Platzer E, Moore MAS, Mertelsmann R, Welte K. Recombinant human granulocyte colony-stimulating factor: Effects on normal and leukemic myeloid cells. Science 1986;232:61-65. Lieschke GJ, Burgess AW. Granulocyte colony-stimulating factor and granulocytemacrophage colony-stimulating factor. First part. N Engl J Med 1992;327:28-35. Wang JM, Chen ZG, Colella S, Bonilla MA, Welte K, Bordignon C, Mantovani A. Chemotactic activity of recombinant human granulocyte colony-stimulating factor. Blood 1988;72:1456-1460. Roilides E, Walsh TJ, Pizzo PA, Rubin M. Granulocyte colony-stimulating factor enhances the phagocytic and bactericidal activity of normal and defective human neutrophils. J Infect Dis 1991;163:579-583. Nelson S. Role of granulocyte colony-stimulating factor in the immune response to acute bacterial infection in the nonneutropenic host: An overview. Clin Infect Dis 1994;18(Suppl. 2):S197-S204. Kerst JM, de Haas M, van der Schoot CE, Slaper-Cortenbach ICM, Kleijer M, von dem Borne AEGK, van Oers RHJ. Recombinant granulocyte colony-stimulating factor administration to healthy volunteers: Induction of immunophenotypically and functionally altered neutrophils via an effect on myeloid progenitor cells. Blood 1993;82:3265-3272. Maher DW, Lieschke GJ, Green M, Bishop J, Stuart-Harris R, Wolf M, Sheridan WP, Kefford RF, Cebon J, Olver I, McKendrick J, Toner G, Bradstock K, Lieschke M, Cruiskank S, Tomita DK, Hoffman EW, Fox RM, Morstyn G. Filgrastim in patients with chemotherapy-induced febrile neutropenia. A double-blind, placebo-controlled trial. Ann Intern Med 1994;121:492-501. Bronchud MH, Scarffe JH, Thatcher N, Crowther D, Souza LM, Alton NK, Testa NG, Dexter TM. Phase 1/11 study of recombinant human granulocyte colony-stimulating factor in patients receiving intensive chemotherapy for small cell lung cancer. Br J Cancer 1987;56:809-813. Crawford J, Ozer H, Stoller R, Johnson D, Lyman G, Tabbara I, Kris M, Grous J, Picozzi V, Rausch G, Smith R, Gradishar W, Yahanda A, Vincent M, Stewart M, Glaspy J. Reduction by granulocyte colony-stimulating factor of fever and neutropenia induced by chemotherapy in patients with small-cell lung cancer. N Engl J Med 1991;325:164-170. Chevallier B, Chollet P, Merrouche Y, Roche H, Fumoleau P, Kerbrat P, Genot JY, Fargeot P, Olivier JP, Fizames C, Clavel M, Yver A, Chabernaud VC. Lenograstim prevents morbidity from intensive induction chemotherapy in the treatment of inflammatory breast cancer. J Clin Oncol 1995;13:1564-1571. Ohno R, Tomonaga M, Kobayashi T, Kanamaru A, Shirakawa S, Masaoka T, Omine M, Oh H, Nomura T, Sakai Y, Hirano M, Yokomaku S, Nakayama S, Yoshida Y, Miura AB, Morishima Y. Dohy H, Niho Y, Hamajima N, Takaku F. Effect of granulocyte colonystimulating factor after intensive induction in relapsed or refractory acute leukemia. N Engl J Med 1990;323:871-877, Locatelli F, Zecca M, Ponchio L, Beguin Y, Giorgiani G, Maccario R, Bonetti F, De Stefano P, Cazzola M. Pilot trial of combined administration of erythropoietin and granulocyte colony-stimulating factor to children undergoing allogeneic bone marrow transplantation. Bone Marrow Transplant 1994;14:929-935. Miles SA, Mitsuyasu RT, Lee K, Moreno J, Alton K, Egrie JC, Souza L, Glaspy JA.

30.

31.

32.

33. 34.

35.

36.

37. 38.

39.

40.

41.

42.

43.

44.

45.

Recombinant human granulocyte colony-stimulating factor increases circulating burst forming unit-erythron and red blood cell production in patients with severe human immunodeficiency virus infection. Blood 1990:75:2137-2142. Miles SA. Golde DW, Mitsuyasu RT. The use of hematopoietic hormones in HIV infection and AIDS-related malignancies. In: Mitsuyasu RT, Golde DW, ed. HematoliOncol Clin North Amer. Philadelphia: W.B. Saunders, 1991, pp. 267-280. Weston B, Axtell RA, Todd R F 111, Vincent M, Balazovich KJ, Suchard SJ. Boxer LA. Clinical and biologic effects of granulocyte colony-stimulating factor in the treatment of myelokathexis. J Pediatr 1991;118:229-234. Dale DC, Bonilla MA, Davis MW, Nakanishi AM, Hammond WP, Kurtzberg J, Wang W, Jakubowski A, Winton E , Lalezari P, Robinson W, Glaspy JA, Emerson S, Gabrilove J, Vincent M. Boxer LA. A randomized controlled phase 111 trial of recombinant human granulocyte colony-stimulating factor (Filgrastim) for treatment of severe chronic neutropenia. Blood 1993;81:2496-2502. Hammond WP, Price TH, Souza LM, Dale DC. Treatment of cyclic neutropenia with granulocyte colony-stimulating factor. N Engl J Med 1989;320:1306-1311. Bonilla MA, GilIio AP, Ruggeiro M, Kernan NA, Brochstein J A , Abboud M, Fumagalli L, Vincent M. Gabrilove JL, Welte K, Souza LM, O'Reilly RJ. Effects of recombinant granulocyte colony-stimulating factor on neutropenia in patients with congenital agranulocytosis. N Engl J Med 1989;320:1574-1580. Negrin RS, Haeuber D H , Nagler A, Olds LC, Donlon T, Souza LM, Greenberg PL. Treatment of myelodysplastic syndromes with recombinant human granulocyte colonystimulating factor. A phase 1-11 trial. Ann Intern Med 1989;110:976-984. Kearns CM, Wang WC, Stute N, Ihle JN. Evans WE. Disposition of recombinant human granulocyte colony-stimulating factor in children with severe chronic neutropenia. J Pediatr 1993;123:471-479. Kojima S, Tsuchida M, Matsuyama T. Myelodysplasia and leukemia after treatment of aplastic anemia with G-CSF. N Engl J Med 1992;326:1294-1295. Imashuku S, Hibi S, Nakajima F, Mitsui T, Yokoyama S, Kojima S, Matsuyama T, Nakahata T, Ueda K, Tsukimoto I. et al. A review of 125 cases to determine the risk of myelodysplasia and leukemia in pediatric neutropenic patients after treatment with recombinant human granulocyte colony-stimulating factor. Blood 1994:84:2380-2381. Imashuku S, Hibi S, Kataoka-Morimoto Y, Yoshihara T, Ikushima S, Morioka Y, Todo S. Myelodysplasia and acute myeloid leukaemia in cases of aplastic anaemia and congenital neutropenia following G-CSF administration. Br J Haematol 1995:89:188-190. Meropol NJ, Miller LD. Korn EL, Braitman LE, Macdermott ML, Schuchter LM. Severe myelosuppression resulting from concurrent administration of granulocyte colonystimulating factor and cytotoxic chemotherapy. J Natl Cancer Inst 1992:84:1201-1204. Broxmeyer HE, Benninger L, Patel SR, Benjamin RS, Vadhan-Raj S. Kinetic response of human marrow myeloid progenitor cells to in vivo treatment of patients with granulocyte colony-stimulating factor is different from the response to treatment with granulocytemacrophage colony-stimulating factor. Exp Hematol 1994:22:100-102. Ohno R, Tomonaga M, Ohshima T, Masaoka T, Asou N, Oh H , Nishikawa K, Kanamaru A, Murakami H, Furusawa S. Japan Adult Leukemia Study Group. A randomized controlled study of granulocyte colony stimulating factor after intensive induction and consolidation therapy in patients with acute lymphoblastic leukemia. Int J Hematol 1993:58:7381. Butler RD, Waites TM, Lamar RE, Hainsworth JD, Greco FA, Johnson DH. Timing of GCSF administration during intensive chemotherapy for breast cancer. ASCO, Anaheim, CA, 1992, p. 403. Diihrsen U, Villeval JL, Boyd J , Kannourakis G, Morstyn G, Metcalf D. Effects of recombinant human granulocyte colony-stimulating factor on hematopoietic progenitor cells in cancer patients. Blood 1988;72:2074-2081. Pettengell R, Gurney H, Radford JA, Deakin DP, James R, Wilkinson PM, Kane K, Bentley

46.

47. 48. 49. 50. 51.

52.

53. 54. 55.

56. 57.

58.

59.

60.

61.

62.

63.

64.

J , Crowther D . Granulocyte colony-stimulating factor to prevent dose-limiting neutropenia in non-Hodgkin's lymphoma: A randomized controlled trial. Blood 1992;SO:1430-1 436. Ottmann O G , Hoelzer D , Gracien E, Ganser A , Kelly K. Reutzel R, Lipp T, Busch FW, Schwonzen M, Heil G. et al. Concomitant granulocyte colonjl-stimulating factor and induction chemoradiotherapy in adult acute lyn~phoblasticleukemia: A randomized phase I11 trial. Blood 1995;86:444-450. Liang D-C, Chen S-H, Lean S-F. Role of granulocyte colony-stimulating factor as adjunct therapy for septicemia in children with acute leukemia. A m J Hematol 1995;48:76-81. Park JW, Mehrotra B, Barnett BO, Baron AD, Venook AP. The Sweet syndrome during therapy with granulocyte colony-stimulating factor. Ann Intern Med 1992;116:996-998. Samlaska CP, Noyes DK. Localized cutaneous reactions to granulocyte colony-stimulating factor. Arch Dermatol 1993;129:645-646. Ross HJ, Moy LA, Kaplan R, Figlin RA. Bullous pyoderina gangrenosum after granulocyte colony-stimulating factor treatment. Cancer 1991:68:441-443. Glaspy J A , Baldwin GC, Robertson PA, Souza L, Vincent M, Ambersley J, Golde DW. Therapy for neutropenia in hairy cell leukemia with recombinant human granulocyte colonystimulating factor. Ann Intern Med 1988:109:789-795. Schliesser G , Pralle H, Lohmeyer J. Leukocytoclastic vasculitis complicating granulocyte colony-stimulating factor (G-CSF) induced neutrophil recovery in Tg-lymphocytosis with severe neutropenia. Ann Hematol 199255:151-1 52. Ostlere LS, Harris D , Prentice H G , Rustin MHA. Widespread folliculitis induced by human granulocyte colony-stimulating factor therapy. Br J Dermatol 1992;127:193-194. Jaiyesimi I. Giralt SS. Wood J. Subcutaneous granulocyte colony-stimulating factor and acute anaphylaxis. N Engl J Med 1991;325:587. Lindemann A , Herrmann F, Oster W, Haffner G, Meyenburg W, Souza LM, Mertelsmann R. Hematologic effects of recombinant h u n ~ a ngranulocyte colony-stimulating factor in patienls with malignancy. Blood 1989;74:2644-2651. Wun T. The Felty syndrome and G-CSF-associated thronlbocytopenia and severe anemia. Ann Intern Med 1993:118:318-319. Mueller BU, Jacobsen F, Butlcr KM, Husson RN, Lewis LL, Pizzo PA. Combination treatment with azidothymidine and granulocyte colony-stimulating factor in children with human imn~unodeficiencyvirus infection. J Pediatr 1992:121 :797-802. Litam PP, Friedman H D , Loughran TP Jr. Splenic extramedullary hematopoiesis in a patient receiving intermittently ad~ninisteredgranulocyte coIony-stimulating factor. Ann Intern Med 1993;118:954-955. Mueller BU, Burt R, Gulick L, Jacobsen F, Pizzo PA, Hornc M. Disseminated intravascular coagulation associated with granulocyte colony-stimulating factor therapy in a child with human immunodeficiency virus infection. J Pediatr 1995;126:749-752. Fleischmann J, Golde DW. Weisbart RH, Gasson JC. Granulocyte-n~acroplagecolonystimulating factor enhances phagocytosis of bacteria by human neutrophils. Blood 1986:68:708-7 11. Baldwin GC, Gasson JC, Quan SG, Fleischrnann J , Weisbart R. Oette D. Mitsuyasu RT, Golde DW. Granulocytc-macrophage colony-stimulating factor enhances neutrophil function in acquired inlmunodeficiency syndrome patients. Proc Natl Acad Sci USA 198838512763-2766. Lopez AF, Williamson DJ. Gamble JR, Begley CG, Harlan JM. Klebanoff SJ, Waltersdorph A , Wong G , Clark SC, Vadas MA. Recombinant human granulocyte-macrophage colonystimulating factor stimulates in vitro mature human neutrophil and eosinophil function. surface receptor expression, and survival. J Clin Invest 1986;78:1220-1228. Duhrsen U, Villeval JL, Boyd J, Kannourakis G. Morstyn G, Metcalf D. Effects of recombinant human granulocyte colony-stimulating factor on hematopoietic progenitor cells in cancer patients. Blood 1988:72:2074-2081. Roilides E, Holmes A, Blake C, Venzon D, Pizzo PA, Walsh TJ. Antifungal activity of elutriated human monocyles agains Aspergillrrs fi~r.rzigat~rshyphae: Enhancenlent by

65.

66.

67.

68.

69.

70.

71.

72.

73.

74.

75.

76.

77.

granulocyte-macrophage colony-stimulating factor and interferon-g. J Infect Dis 1094; 17018944899. Antman KS, Griffin JD, Elias A , Socinski MA, Ryan L, Cannistra SA. Oette D. Whitley M. Frei E, Schnipper LE. Ef'fect of reconlbinant human granulocyte-macrophage colonystimulating factor on chemotherapy-induced inyelosuppression. N Engl J Med 1988:319:593508. Bettelhcim P, VaIent P, Andreeff M, Tafuri A. Hairni J, Gorischek C. Muhm M. Sillaber C, Haas 0 , Vieder L, Maurer D, Schulz G, Speiser W, Geissler K, Kier P, Hinterberger W, Lechner K. Reco~nbinanthuman granulocyte-macrophage colony-stimulating factor in combination with standard induction chemotherapy in de novo acute mpeloid leukemia. Blood 1991;77:700-711. Bunn P A Jr., Crowlep J, Kelly K, Hazuka MB, Beasley K. Upchurch C. Livingston R. Chemoradiotherapy with or without granulocyte-macrophage colony-stimulating factor it1 the treatment of limited-stage small-cell lung cancer: A prospective phase 111 randomized study of the Southwest Oncology Group. J Clin Oncol 1995:13:1632-1641. Bajorin DF. Nichols CR, Schmoll HJ, Kantoff PW. Bokemeyer C, Demelri GD, Einhorn LH, Bosl GJ. Recombinant human granulocyte-macrophage colony-stimulating factor as an adjunct to conventional-close ifosfamide-based chemotherapy for patients with advanced or rcIapsec1 germ cell tumors: A randomi7ed tr~al.J Clin Oncol 1995;13:79-86. H a m ~ nJ, Schiller JH. Cui'iie C, Oken M, Fisher RI. Shepherd F, Kaiser G. Dose-ranging study of rccombinant human granulocyte-macrophage colony-stimulating factor in small-ceH lung carcinoma. J Clin Oncol 1994;12:2667-2676. Engelhard M. G e r h a r t ~H. Brittinger G, et al. Cytokine efficiency in the treatment of highResults of a randomized double-blind placebograde malignant non-Hodgkin's lyn~phon~as: controlled study with intensified COP-BLAM +/- rhGM-CSF. Ann Oncol 19945 (Suppl. 2):123-125. Laughlin MJ. Kirkpatrick G, Sabiston N, Petcrs W, Kurtzbcrg J. Hematopoietic recovery following high-dose combined alkylating-agent chemotherapy and autologous bone marrow support in patients in phase-I clinical trials of colony-stimulating factors: G-CSF, GM-CSF. IL- 1 , IL-2, M-CSF. Ann Hematol 1993:67:267-276. Gerhartz HH, Engelhard M, Meusers P, et al. Randomized, double-blind, placebocontrolled. phase I11 study of rccombinant human granulocyte-macrophage colonystirnulaling factor as adjunct to induction treatment of high-grade malignant non-Hodgkin's lymphomas. Blood 1993:82:2329-2339. Chi K-H, Chen C-H. Chan W-K, Chow K-C. Chen S-Y, Yen S-H, Chao JY, Chang C-Y, Chen KY. Effcct of granulocyte-macropl~agecolony-stimulating factor on oral inucositis in head and neck cancer patients after cisplatin, fluorouracil, and leucovorin chen~otherapy. J Clin Oncol 1995:13:2620-2628. Brandt SJ, Peters WP. Atwater SK. Kurtzberg J. Borowitz MJ, Jones RB. Shpall EJ, Bast RC. Gilbert CJ, Oette DH. E f f c c ~of recombinant granulocyte-macropllage colonystimulating factor on hematopoietic reconstitution after high-dose chemotherapy and autologous bone marrow transplant. N Engl J Med 1988;318:869-876. Advani R. Chao NJ. Horning SJ, Blume KG, Ahn DK. Lamborn KR. Fleming NC, Bonnem EM, Greenberg PL. Granulocyte-macropl~agccolony-stimulating factor (GM-CSF) as an adjunct to autologous hen~opoicticstem cell transplantation for lptnphorna. Ann Intern Mcd 1992:116:183-189. Gordon B. Spadinger A, Hodgcs E. Ruby E , Stanley R, Coccia P. Effect of granulocytemacrophage colony-stimulating factor on oral mucositis after hematopoiet~cstem-cell transplantation. J Clin Oncol 1994:12:1917-1922. Nemunaitis J, Rabinowe SN, Singer JW, Bierman PJ, Vose JM, Freedman AS, Onetto N, Gillis S, Oette D, Gold M, Buckner CD, Hansen JA. Ritz J, Appelbaum FR, Armitage JO, Nadler LM. Recombinant granulocyte-macrophage colony-stimulating factor after autologous bonc marrow transplantation for lymphoid cancer. N Engl J Med 1991:324:17731778.

78. Nemunaitis J. Rosenfeld CS, Ash R, et al. Phase I11 randomized, double-blind placebocontrolled trial of rhGM-CSF following allogeneic bone marrow transplantation. Bone Marrow Transplant 1995;15:949-954. 79. Riikonen P, Saarinen UM, Makipernaa A, Hovi L, Komulainen A, Pikhala J, Jalanko H. Recombinant human granulocyte-macrophage colony-stimulating factor in the treatment of febrile neutropenia: A double blind placebo-controlled study in children. Pediatr Infect Dis J 1994;13:197-202. 80. Gluck S, Gagnon A. Neutropenic fever in patients after high-dose chemotherapy followed by autologous haematopoietic progenitor cell transplantation and human recombinant granulocyte-macrophage colony stimulating factor. Bone Marrow Transplant 1994;14:989990. 81. Lieschke GJ, Cebon J, Morstyn G. Characterization of the clinical effects after the first dose of bacterially synthesized recombinant human granulocyte-macrophage colony-stimulating factor. Blood 1989;74:2634-2643. 82. Lieschke GJ, Burgess AW. Granulocyte colony-stimulating factor and granulocytemacrophage colony-stimulating factor. Second part. N Engl J Med 1992;327:99-106. 83. Roilides E, Mertins S, Eddy J, Walsh TJ, Pizzo PA, Rubin M. Impairment of neutrophil chemotactic and bactericidal function in children infected with human immunodeficiency virus type 1 and partial reversal after in vitro exposure to granulocyte-macrophage colonystimulating factor. J Pediatr 1990;117:531-540. 84. Roilides E, Pizzo PA. Modulation of host defense by cytokines: Evolving adjuncts in prevention and treatment of serious infections in immunocompro~nised hosts. Clin Infect Dis 1992:15:508-524. 85. Nemunaitis J, Meyers JD, Buckner CD, Shannon-Dorcy K, Mori M, Shulman H, Bianco JA, Higano CS, Groves E, Storb R, Hansen J, Appelbaum FR, Singer JW. Phase I trial of recombinant macrophage colony-stimulating factor in patients with invasive fungal infections. Blood 1991;78:907-913. 86. Cenci E, Bartocci A, Puccetti P, Mocci S, Stanley ER, Bistoni F. Macrophage colonystimulating factor in murine candidiasis: Serum and tissue levels during infection and protective effect of exogenous administration. Infect Imlnun 1991:59:868-872. 87. Khwaja A, Johnson B, Addison IE, Yong K, Ruthven K, Abramson S, Linch DC. In vivo effects of macrophage colony-stimulating factor on human monocyte function. Br J Haematol 1991;77:25-31. 88. Nemunaitis J, Shannon-Dorcy K, Appelbaum FR, Meyers J, Owens A, Day R, Ando D, O'Neill C, Buckner D, Singer J. Long-term follow-up of patients with invasive fungal disease who received adjunctive therapy with recombinant human macrophage colony-stimulating factor. Blood 1993;82:1422-1427. 89. Starnes HFJ. Biological effects and possible clinical applications of interleukin-1. Sem Hematol 1991;28:34-41. 90. D'Hondt V, Weynants P, Humblet Y, Guillaume T, Canon J-L, Beauduin M, Duprez P, Longueville J, Miill R, Chatelain C, Symann M. Dose-dependent interleukin-3 stimulation of thrombopoiesis and neutropoiesis in patients with small-cell lung carcinoma before and following chemotherapy: A placebo-controlled randomized phase Ib study. J Clin Oncol 1993;11:2063-2071. 91. Geissler K, Forstinger C, Kalhs P, Knob1 P, Kier P, Kyrle P, Lechner K. Effect of interleukin3 on responsiveness to granulocyte colony-stimulating factor in severe aplastic anemia. Ann Intern Med 1992;117:223-225. 92. Dinarello CA, Wolff SM. The role of interleukin-1 in disease. N Engl J Med 1993;328:106114. 93. Broudy VC, Kaushansky K, Harlan JM, Adamson JW. Interleukin 1 stimulates human endothelial cells to produce granulocyte-macrophage colony-stimulating factor and granulocyte colony-stimulating factor. J Immunol 1987;139:464-468. 94. Fibbe WE, Van Damme J, Billiau A , Duinkerken N, Lurvink E, Ralph P, Altrock BW, Kaushansky K, Willemze R, Falkenburg JH. Human fibroblasts produce granulocyte-CSF,

95.

96. 97.

98.

99.

100.

101.

102.

103. 104.

105. 106.

207.

108.

109. 110. 111. 112.

macrophage-CSF, and granulocyte-macrophage-CSF following stimulation by interleukin-1 and poly(rI).poly(rC). Blood 1988;72:860-866. Fibbe WE, Daha MR, Hiemstra PS, Duinkerken N, Lurvink E, Ralph P, Altrock BW, Kaushansky K, Willemze R, Falkenburg JH. Interleukin 1 and poly(rI).poly(rC) induce production of granulocyte CSF, macrophage CSF, and granulocyte-macrophage CSF by human endothelial cells. Exp Hematol 1989;17:229-234. Crown J, Jakubowski A, Gabrilove J. Interleukin-1: Biological effects in human hematopoiesis. Leuk Lymphoma 1993;9:433-440. Kovacs CJ, Harrell JP, Gooya JM, Evans MJ, Johnke RM. Synergy between recombinant human IL-1 alpha (rHuIL-1) and M-CSF (rHuM-CSF) during the recovery of murine hematopoietic activity in myelosuppressed animals: abbreviated versus chronic administration of rHuM-CSF. Exp Hematol 1992;20:582-589. Gallicchio VS, Hulette BC, Messino MJ, Gass C, Bieschke MW, Doukas MA. Effect of various interleukins (IL-1, IL-2. and IL-3) on the in vitro radioprotection of bone marrow progenitors (CFU-GM and CFU-MEG). J Biol Response Modif 1989;8:479-487. Gallicchio VS, Shedlofsky SI, Swim AT, Robinson JM, Hulette BC, Messino MJ, Doukas MA. Modulation of murine hematopoiesis in vivo with recombinant murine interleukin-1. J Biol Response Modif 1989;8:422-439. Johnson CS, Keckler DJ, Topper MI, Braunschweiger PG, Furmanski P. In vivo hematopoietic effects of recombinant interleukin-1 alpha in mice: Stimulation of granulocytic, monocytic, megakaryocytic, and early erythroid progenitors, suppression of late-stage erythropoiesis, and reversal of erythroid suppression with erythropoietin. Blood 1989;73: 678-683. Castelli MP, Black PL, Schneider M, Pennington R, Abe F, Talmadge JE. Protective, restorative, and therapeutic properties of recombinant human IL-1 in rodent models. J Immunol 1988;140:3830-3837. Neta R, Oppenheim JJ, Wang JM, Snapper CM, Moorman MA, Dubois CM. Synergy of IL1 and stem cell factor in radioprotection of mice is associated with IL-1 upregulation of mRNA and protein expression for c-kit on bone marrow cells. J Immunol 1994;153:15361543. Ozaki Y. Ohashi T. Kume S. Potentiation of neutrophil function by recombinant DNSproduced interleukin la. J Leukoc Biol 1987;42:621-627. Ozaki Y, Ohashi T, Minami A, Nakamura S. Enhanced resistance of mice to bacterial infection induced by recombinant human interleukin-la. Infect Immun 1987;55:14361440. Minami A, Fujimoto K, Ozaki Y, Nakamura S. Augmentation of host resistance to microbial infections by recombinant human interleukin-1 alpha. Infect Immun 1988;56:3116-3120. Smith JWd, Urba WJ, Curti BD, et al. The toxic and hematologic effects of interleukin-1 alpha administered in a phase I trial to patients with advanced malignancies. J Clin Oncol 1992;10:1141-1152. Kullberg BJ, van 't Wout JW, PoelI RJ, van Furth R. Combined effect of fluconazole and recombinant human interleukin-1 on systemic candidiasis in neutropenic mice. Antimicrob Agents Chemother 1992;36:1225-1229. van der Meer JW, Vogels MT, Kullberg BJ. Interleukin-1 and related pro-inflammatory cytokines in the treatment of bacterial infections in neutropenic and non-neutropenic animals. Biotherapy 1994;7:161-167. Dinarello CA. The role of interleukin-1 in host responses to infectious diseases. Infect Agents Dis 1992;1:227-236. Marley S, Reilly I, Russell N. Interleukin 1 has positive and negative regulatory effects in human long-term bone marrow culture. Exp Hematol 1992;20:75-79. Sunderland MC. Roodman GD. Interleukin-3. Its biology and potential uses in pediatric hematology oncology. Am J Pediatr Hematol Oncol 1991;13:414-425. Donahue RE, Seehra J, Metzger M, et al. Human IL-3 and GM-CSF act synergistically in stimulating hematopoiesis in primates. Science 1988;241:1820-1823.

113. Leary AG, Ikebuchi K, Hirai Y, Wong GG. Yang YC, Clark SC, Ogawa M. Synergism between interleukin-6 and interleukin-3 in supporting proliferation of human henlatopoietic stem cells: Comparison with interleukin-1 alpha. Blood 1988;71:1759-1763. 114. Bhalla K, Tourkina E, Huang Y, Tang C, Mahoney ME, Ibrado AM. Effect of hemopoietic growth factors G-CSF and PIXY 321 on the activity of high dose ara-C in human myeloid leukemia cells. Leuk Lymphoma 1993;10(Suppl):123-131. 115. McCrady CW, Li F, Pettit GR. Grant S. Modulation of the activity of a human granulocytemacrophage colony-stimulating factorlinterleukin-3 fusion protein (PIXY 321) by the nlacrocyclic lactone protein kinase C activator bryostatin 1. Exp Hematol 19?3;21:893-900. 116. Murray HW. Interferon-g, the activated macrophage, and host defense against microbial challenge. Ann Intern Med 1988;108:595-608. 117. Murray HW. Gamma interferon, cytokine-induced macrophage activation, and antimicrobial host defense. In vitro, in animal models, and in humans. Diagn Microbial Infect Dis 1990:13:411-421. 118. Gallin JI. Farber JM. Holland SM, Nutman TB. Interferon-g in the management of infectious diseases. Ann Intcrn Med 1995:123:216-224. 119. Quie PG. Chronic granulomatous disease of childhood: A saga of discovery and understanding. Pediatr Infcct Dis J 1993;12:395-398. 120. The International Chronic Granulonlatous Disease Cooperative Study Group. A controlled trial of interferon gamma to prevent infection in chronic granulo~natousdisease. N Engl J Med 1991;324:509-516. 121. Squires KE, Murphy WF, Madoff LC. Murray HW. Interferon-g and Mycobrrcteri~!tnnvizunint~~rrcelllilare infection. J Infect Dis 1989;159:5?9-600. 122. Schmid GP, Smith RP, Baltch AL, Hall CA, Schiffman G. Antibody response to pneumococcal vaccine in patients with multiple myeloma. J Infect Dis 1981;143:590-597. 123. Siber GR, Gorham C. Martin P, Corkery JC. Schiffman G. Antibody response to pretreatment immunization and post-treatment boosting with bacterial polpsaccharide vaccines in patients with Hodgkin's disease. Ann Intern Med 1986;104:467-475. 124. Centers for Disease Control. Recommendations of the I~ninunizationPractices Advisory Committee (ACIP): Pncumococcal polysaccharide vaccine. MMWR 1?89:**:64-76. 125. Gershon AA, Steinberg SP. Persistence of immunity to varicella in children with leukemia immunized with live attenuated varicclla vaccine. N Engl J Med 1989:320:892-897. 126. Hardy I, Gershon AA, Steinberg SP, LaRussa P. The incidence of zoster after immunization with live attenuated varicella vaccine. A ~ t u d yin children with leukemia. Varicella Vaccine Collaborative Study Group [see comments]. N Engl J Med 1991:325:1545-1550. 127. Ruiz-Moreno M, Garcia R, Rua MJ, Serrano B, Moraleda G, Feijoo E, BartolomC J. Ortiz F, Castillo I. Carreno V. Levamisole and interferon in children with chronic hepatitis B. Hepatology 1993:18:264-269. 128. Hersh EM. Brewton G, Abrams D. Bartlett J, Galpin J, Gill P, Gorter R. Gottlicb M, Jonikas JJ, Landesnlan S. Levine A. Marcel A. Petersen EA, Whiteside M. Zahradnik J. Negron C, Boutitie F, Caraux J, Dupuy J-M, Saltni R. Ditiocarb sodium (diethyldithiocarbamate) therapy in patients with syrnptomaytic HIV infection and AIDS. JAMA 1991;265:1538-1544. 129. Hers11 EM. Imn~unomodulatory drugs of relevance to the management of microbial infections. Adv Exp Med Biol 1992:319:1-11. 130. Trillet-Lenoir V, Green J, Manegold C, et al. Recombinant granulocyte colony stimulating factor reduces the infectious complications of cytotoxic chenlotherapy. Eur J Cancer 1993;29A:319-324. 131. Maiche AG, Muhonen T. Granulocyte colony-stimulating factor (G-CSF) with or without a quinolone in the prevention of infection in cancer patients. Eur J Cancer 1993;29A:14031405. 132. Talcott JA, Siegel RD. Finberg R, Goldman L. Risk asscssrnent in cancer patients with fever and neutropenia: A prospective, two-center validation of a prediction rule. J Clin Oncol 1992;10:316-322.

133. Gabrilove JL, Jakubowski A , Fain K. Grous J, Scher H , Sternberg C, Yagoda A, Clarkson B, Bonilla MA, Oettgen HF, Alton K, Boone T, Altrock B. Welte K. Souza L. Phase I study of granulocyte colony-stimulating factor in patients with transitional cell carcinoma of the urothelium. J Clin Invest 1988;82:1454-1461. 134. Gabrilove JL, Jakubowski A, Scher H. Sternberg C. Wong G, Grous J, Yagoda A. Fain K, Moore MAS, Clarkson B, Oettgen HF, Alton K, Welte K, Souza L. Effect of granulocyte colony-stimulating factor on neutropenia and associated morb~ditydue to chemotherapy for transitional-cell carcinoma of the urothelium. N Engl J Med 1988:318:1414-1422. 135. Biesnla B, de Vries EG, Willemse PH, Sluiter WJ, Postmus PE. Limburg PC. Stern AC, Vellenga E. Efficacy and tolerability of recombinant huinan granulocyte-~nacrophage colony-stimulating factor in patients with chemotherapy-related leukopenia and fever. Eur J Cancer 1990;26.932-936. 136. Walsh TJ. Invasive pulmonary aspergillosis in patlents with neoplastic diseases. Semin Respir Infcct 1990;S:l 1 1-122. 137. WaIsh TJ, Lee JW. Roilides E. Pizzo PA. Recent progress and current problems in inanagement of invasive fungal infections in patients with neoplastic disezes. Curr Opin Oncol 1992:4:647-655. 138. Weinberger M, Elattar I. Marshall D. Steinberg SM, Redner RL. Young NS, Pizzo PA. Palterns of infections in patients with aplastic anemia and thc emergence of Aspergilllrs as a major cause of death. Medicine 1992;71:24-42. 139. Walsh TJ. Hiemenz J. Pizzo PA. Editorial response: Evolvitlg risk factors for invasive fungal infections - All neutropenic patients are not the same. Clin Infect Dis 1994:18:793-798. 140. Walsh TJ. Management of i~nn~unocompron~ised patients with evidence of invasive mycosis. Hernatolit/Oncol Clini North A m 1993:7:1003-1026. 141. Rollides E, Scin T. Holmes A, Chanock S, Blake C, Pizzo PA, Walsh TJ. Effects of macrophage colony-stimulating factor on antifungal activity of mononuclear phagocytes against Aspergillus fumigatus. J Infect Dis 1995:172:1028-1034. 142. Naparstek E. Granulocyte colony-stimulating factor, congenital neutropenia. and acute myeloid leukernia. N Engl J Med 1995;333:516-51 8. 143. Vellenga E, Young DC, Wagner K. Wiper D, Ostapovicz D, Griffin JD. The effects of GMCSF and G-CSF in promoting growth of clonogenic cells in acute nlyeloblastic leukemia. Blood 1987:69:177 1-1 776. 144. Tafuri A. Andreeff M. Kinctic rationale for cytokinc-induced recruitment of myeloblastic leukemia followed by cycle-specific chemotherapy in vitro. Leukemia 1990;4:826-834. 145. Aglietta M, Piacibello W. Sanavio F. Stacchini A, Apra F, Schena M, Mossetti C, Carnino F. Callgaris-Cappio F, Gavosto F. Kinetics of human hemopoletic cells after in vivo administration of granulocyte-macrophagc colony-stimulating factor. J Clin Invest 1989;83:551557. 146. Hiddemann W, Kiehl M, Zuhlsdorf M, Busemann C. Schleyer E, Wormann B, Buchner T. Granulocyte-macrophage colony-stimulating factor and interleukin-3 enhance the incorporation of cytosine arabinoside into the D N A of leukemic blasts and the cytotoxic effect on clonogenic cells froin patients with acute mpeloid leukemia. Semin Oncol 1992;19:31-37. 147. Smith MA, Singer CR, Pallister CJ, Smith JG. The effect of haemopoietic growth factors on the cell cycle of AML progenitors and their sensitivity to cytosine arabinoside in vitro. Br J Haematol 1995;90:767-773. 148. Motoji T, Watanabe M. Uzumaki H. Kusaka M, Fukainachi H, Shirnosaka A. Oshimi K. Mlzoguchi H. Granulocyte coiony-stimulating factor (G-CSF) receptors on acute myeloblastic lcukaemla cells and their relationship with the proliferative response to G-CSF in clonogenic assay. Br J Haematol 199 1:77:54-59. 149. Dong F, Brynes RK. Tidow N. Welte K. Lowenberg B, Touw IP. Mutations in the gene for the granulocyte colony-stimulating factor receptor in patients with acute n~yeloidleukernia preceded by severe congenital neutropenia. N Engl J Med 1995;333:487-493.

150. Marsoni S, Ungerleider RS, Hurson SB, Simon RM, Hammershaimb LD. Tolerance to antineoplastic agents in children and adults. Cancer Treat Rep 1985;69:1263-1269. 151. Fink FM, Maurer-Dengg K, Fritsch G, Mann G, Zoubek A, Falk M, Gadner H. Recombinant human granulocyte-macrophage colony-stimulating factor in septic neutropenic pediatric cancer patients: Detection of circulating hematopoietic precursor cells correlates with rapid granulocyte recovery. Med Pediatr Oncol 1995;25:365-371.

9. Prevention of infection in immunocompromised hosts Gary A. Noskin

1. Introduction Infections are a significant cause of morbidity and mortality in patients with neoplastic diseases. In addition, for some malignancies an important factor that limits the ability to administer cytotoxic chemotherapy is the risk of infection. This risk varies among patients based on their underlying disease and host defenses. Neutropenia is the most common host factor that predisposes patients to infection. Other important factors that increase the risk of infection include: defects in cell-mediated and humoral immunity, impaired mucosal or skin integrity, obstruction by mass lesions, and intravascular devices. From the time the first bone marrow transplants were performed, the risk of infection was readily identified [I] and strategies were developed in an attempt to prevent them. Almost all cancer centers have designed protocols to prevent infections that range from early empiric therapy instead of antimicrobial prophylaxis to a total protective environment [2]. This variability confirms that prevention of infection is an important, but complicated, endeavor and remains in evolution (Table 1).Immunocompromised individuals may acquire a wide range of pathogens, either exogenously or by reactivation of latent organisms (Table 2). Therefore, strategies to prevent infection require a comprehensive approach that identifies the most likely organisms that can result in disease within an individual. This chapter reviews the antimicrobial and environmental factors that have been used to prevent infection; however, it must be emphasized that the simplest and most effective measure to prevent infection remains proper hand washing [3].

2. Host defense mechanisms and infecting microorganisms An understanding of underlying host defense is necessary to determine the microorganisms that infect patients with neoplastic diseases. Prophylaxis is based on an interplay between immune dysfunction, granulocytopenia, and the microorgainisms that occur in patients with cancer. The most common G a r y A. Noskitz (ed), M A N A G E M E N T OF INFECTIOUS C O M P L I C A T I O N S IN C A N C E R PATIENTS. O 1998. Kluwer Acridemic Publishers, Boston. All rights reserved.

factor that predisposes patients to infection is neutropenia. Granulocytopenia is often associated with acute leukemia and following cytotoxic chemotherapy. In these patients, the risk of infection correlates with the absolute number and function of their neutrophils [4]. The greatest risk of infection occurs when the granulocyte count falls below 5001pL; however, most severe infections occur when the absolute neutrophil count is less than 1001pL. The microorganisms most frequently associated with granulocytopenia are the aerobic gramnegative bacilli and the gram-positive cocci. Anaerobic infections occur less commonly and are generally associated with gastrointestinal disease such as typhlitis. In addition to bacterial pathogens, patients with neutropenia are also at increased risk for systemic fungal infections. The majority of these infections are caused by Candidn and AspergiIl~tsspecies [5].Because both candidiasis and aspergillosis are usually nosocomially acquired in immunocompromised hosts, strategies to prevent these pathogens are important. Defects in cell-mediated immunity are evident in patients with Hodgkin's disease and HIV-1 infection. Cellular immunity is also impaired in patients on certain immunosuppressive medications such as azathioprine, cyclosporine, and prednisone. These T-cell defects predispose patients to a wide range of infections due to intracellular pathogens, including bacterial (Listeria nzonocytogenes, Snlmonelln, Legionelln, and Mycobacterin), fungal (Cryptococcus neoforulznns, Histoplasnzn cnpsulntum, and Coccidiodes immitis), viral (herpes simplex virus, varicella-zoster virus, and cytomegalovirus), and protozoal (Pne~~mocystis cnrinii, Toxoplnsmu gondii, and Cryptosporiditinz). Patients with multiple myeloma or chronic lymphocytic leukemia are often hypogammaglobulinemic and may develop infections due to encapsulated bacteria such as Streptococc~ispneun~onineor Hnemophilus influenzne. This defect in humoral immunity is also present in patients who are splenectomized. Other factors that predispose patients to infection include medical devices or procedures that break the natural protective barrier of the skin, such as intravascular and indwelling bladder catheters. In addition, prior treatment with antin~icrobialagents can alter a patient's endogenous flora to increase the risk of infection with resistant organisms [6]. More importantly, once the host acquires a resistant organism, it may be difficult and, in some cases, impossible to eradicate [7].

3. Prevention of bacterial infections

It is well known that patients with cancer are at increased risk for bacterial infections, and methods to prevent them have been examined for over three decades [8-101. More importantly, however, is that patients should be free of an active infectious process prior to the initiation of chemotherapy [Ill. While there have been many clinical trials evaluating methods to prevent bacterial

infections, the results must be interpreted with caution because they often contain small sample sizes, varying methodologies, and different endpoints. The most common infections that occur in neutropenic patients are those due to the aerobic bacteria that colonize the skin, oral cavity, and gastrointestinal tract. 3.1 Protected environment Initially there was enthusiasm for the use of a total protected environment as an effective method to prevent bacterial infections in cancer patients [ll-141. In a study by Yates and Holland [Ill, patients with acute myelogenous leukemia were randomized to one of four treatment groups: (1) conventional care, (2) reverse isolation with oralltopical antibiotics, (3) barrier isolation with filtered air, and (4) barrier isolation, filtered air, and oralltopical antibiotics. The average duration of neutropenia was slightly longer than 3 weeks in all groups. They found that there was no difference in empiric antimicrobial therapy among the groups, but that there were less infection-related deaths ainong the patients receiving antibiotics. The downside of this regimen was the observation that more patients treated with prophylactic antibiotics died from hemorrhagic complications. In addition, there was no benefit to nonabsorbable antibiotics or a protected environment for the first 21 days of neutropenia. Finally, despite these interventions, there was no difference in survival between the four groups. A similar study was performed by Levine and coworkers [13], which revealed that leukemic patients in a protected environment, treated with prophylactic antibiotics, and on a low microbial diet had a significant reduction in sepsis. These patients also had fewer life-threatening infections and infectionrelated deaths compared with patients treated outside the protected environment or without antibiotics. Interestingly, there was no difference in infection rates or outcome between the patients receiving only prophylactic antibiotics and the untreated controls. Once again, this study failed to demonstrate any difference in rates of duration of remission. Schimpff and colleagues [14] also performed a clinical trial examining the role of a protected environment and nonabsorbable antibiotics on the risk of infection in patients with leukemia. Patients were randomized to (1) laminar airflow and prophylactic oral antibiotics (gentamicin, vancomycin, and nystatin), (2) routine care plus oral nonabsorbable antibiotics, and (3) routine ward care. These investigators found that oral nonabsorbable antibiotics resulted in a significant reduction of infections, but that laminar airflow did not provide any additional benefit in patients who tolerated the antibiotics. In addition, the laminar airflow patients and those receiving antibiotics achieved a higher complete remission rate and longer survival than those on the general ward. There have been other studies performed during the 1970s that have demonstrated a reduced infection rate with nonabsorbable antibiotics, but these

agents are not routinely used in clinical practice today. The most important reason to avoid these is the selection of more resistant organisms. This is especially true for oral vancomycin, which may increase a patient's risk for acquiring vancomycin-resistant enterococci [6]. Furthermore, patient compliance with oral nonabsorbable antibiotics is generally poor due to medication intolerance. 3.2 Selective decontamination The idea of selective decontamination was fostered when studies revealed that virtually the entire intestinal flora could be eliminated in animals treated with certain antibiotics [15,16]. When these animals were given oral gramnegative bacilli, they became recolonized with a smaller inocula. It was then hypothesized that intestinal anaerobes served as a protective barrier against infection with more virulent gram-negative bacilli. This led to the use of trimethoprim-sulfamethoxazole (TMPISMX) for prophylaxis because this antibiotic would eliminate enteric gram-negative bacilli, without altering anaerobic flora. The interest in this agent for the prevention of infection surfaced when Hughes and associates [17] demonstrated that TMPISMX was more effective than placebo in preventing Bneurnocystis carinii pneumonia in children with leukemia. They also observed that children who received TMPISMX for prophylaxis had a reduction in other infections such as bacteremia, pneumonia, otitis media, and cellulitis. Subsequently, other investigators have examined the role of TMP/SMX as well as other antibiotics, most recently the quinolones, in systemic prophylaxis. The first study designed to evaluate TMPISMX for prophylaxis evaluated 30 patients with acute leukemia and examined those who received framycetin, colistin, and nystatin (FRACON) with or without TMPISMX [la]. The TMPISMX group developed fewer infections and received less days of antibiotics; however, there was no untreated control group. A subsequent study comparing TMPISMX with FRACON plus TMPISMX failed to demonstrate any differences between the two groups [19]. Therefore, the authors concluded that TMPISMX was preferable due to less cost and better tolerability. In patients with acute leukemia, several studies have demonstrated that TMPISMX prophylaxis has resulted in a reduction of total infections and infection-related mortality, but an increase in TMPISMX-resistant bacteria [20-221. However, Weiser and colleagues [23] were unable to show a difference in the incidence of febrile episodes, hospitalizations for fever, or documented infections in patients treated with TMPISMX compared with controls. In 1983, Wade and coworkers [24] questioned the benefit of antimicrobial prophylaxis in patients with acute leukemia undergoing induction chemotherapy. They randomized 62 patients to prophylaxis with TMPISMX or nali-

Table I. Concepts of antimicrobial prophylaxis

Prophylaxis should be targeted against frequent infections that are associated with significant morbidity and mortality Prophylaxis should continue for the high-risk period of the specific infection Goals of prophylaxis Improved survival Decrease incidence of infection Avoid selection of resistant organisms Prevent superinfection Characteristics of an "ideal" agent for prophylaxis Low toxicity profile Few or no drug-drug interactions Narrow spectrum of activity Convenient dosing regimen Good bioavailability (oral and intravenous formulations) Well tolerated Inexpensive Reprinted with permission from Momin F, Chandrasekar PH. Antimicrobial prophylaxis in bone marrow transplantation. Ann Intern Med 1995;123:206.

dixic acid in an attempt to prevent infection. Both groups also received nystatin. The TMPISMX-treated patients experienced a significantly longer period of neutropenia and more frequent acquisition of fungal infections, whereas gram-negative infections were more common in the nalidixic acid-treated patients. Five of the 7 deaths in the TMPISMX-treated patients were due to Aspergill~isflavus.They concluded that although both regimens offered some advantages to patients with leukemia, the risk of bacterial andlor fungal superinfection was significant enough to limit widespread use. The largest study examining the use of TMPISMX in the prevention of infection in neutropenic patients was performed by the EORTC [25]. They randomized 542 patients with anticipated prolonged neutropenia to receive TMPISMX or placebo. The overall incidence of infection was significantly lower in the TMPISMX group, but there was no difference in bacteremia. If patients with acute nonlymphocytic leukemia are removed from analysis, the lower incidence of bacteremia in the TMPISMX patients becomes significant. Not surprisingly, there was a higher incidence of TMPISMX-resistant organisms in patients who received this medication for prophylaxis. Many studies have documented that TMPISMX is effective in preventing infections in granulocytopenic patients, but an unanswered question was whether the effect was on gastrointestinal flora or due to therapeutic levels in the blood thereby preventing tissue invasion [26]. In a small study comparing neomycin plus colistin with TMPISMX plus colistin, there was a benefit in the TMPISMX-treated patients. This suggests that tissue levels of antimicrobial agents are important in the prevention of infection. The exact role of TMPISMX in the prevention of infection in the neutropenic host varies among clinicians. Although it is difficult to make definitive recommendations based on published studies due to differences in

methodology, several conclusions can be drawn [27]: (1) Patients with an anticipated duration of profound neutropenia (<1001pL) for greater than 7 days benefit. (2) Prophylaxis is well tolerated. (3) Prophylaxis may prolong the period of neutropenia. (4) There is an increased risk of TMPISMX-resistant bacteria in patients on prophylaxis. Recently many investigators have advocated the use of Auoroquinolones as prophylactic agents in patients with neutropenia or undergoing bone marrow transplantation. The major advantage that the quinolones have over TMPl SMX is a broader spectrum of activity (including Psetldomonas aeruginosa) and a lower risk of bone marrow suppression. These agents are safe and well tolerated, but significantly more costly than TMPISMX. The major disadvantage of the fluoroquinolones is the risk for selecting resistant bacteria P-81. The first prospectively randomized study to evaluate the quinolones compared norfloxacin 400mg twice daily with placebo in patients with acute leukemia [29]. This investigation demonstrated that norfloxacin delayed the onset of the first febrile episode and decreased the number of gram-negative infections. Therapy was well tolerated and there was no increase in the development of quinolone-resistant bacteria. However, there was no difference in survival or the number of febrile episodes between the norfloxacin and the placebo groups. In a study of 56 patients with acute leukemia, Dekker and colleagues [30] compared ciprofloxacin (500mg twice daily) with TMPISMX (1601800mg) and colistin (200mg TID). The quinolone-treated patients had less serious infections than the TMPISMX-colistin patients. Additionally, there were no gramnegative infections in the ciprofloxacin treated group compared with seven in the TMPISMX-colistin group. Both regimens were well tolerated, but the TMPISMX-colistin patients had a higher incidence of resistant bacteria. In a follow-up study [31], the addition of roxithromycin was included with the quinolone regimen because of the high incidence of streptococcal infections in this group and was successful in decreasing the rate of a-hemolytic streptococci. In the largest study to examine the role of quinolones for prophylaxis, 801 patients with hematologic malignancies were randomized to receive either ciprofloxacin (500mg twice daily) or norfloxacin (400mg twice daily) [32]. There were significantly fewer patients requiring antibiotics and a lower rate of documented infections due to gram-negative bacilli in the ciprofloxacin group. Although both regimens were safe and well tolerated, patients without prolonged neutropenia (
tion. There was no difference between the groups with respect to neutropenic fever or overall infection rates. However, there was a significantly higher rate of Clostridiurn difJicile-associated enterocolitis in the TMPISMX-treated patients. While there was a higher incidence of gram-negative infections and prolongation of neutropenia in the TMPISMX-treated patients, these differences were not statistically significant. In 99 women undergoing autologous bone marrow transplantation for metastatic breast cancer, prophylaxis with ciprofloxacin (500 mg every 8 hours) plus rifampin (300 mg twice daily) was effective in preventing bacterial infections [34]. Compared with historical controls, the patients receiving prophylaxis had a lower incidence of neutropenic fever, fewer documented infections, and less bacteremia. There was no increase in the isolation of rifampin-resistant organisms reported in this study. The concern about delaying neutrophil recovery in patients receiving TMPI SMX who undergo bone marrow transplantation has also been examined [35]. Patients treated with ciprofloxacin were compared with a control group of patients receiving TMPISMX. Both groups started prophylactic antibiotics and granulocyte-macrophage colony stimulating factor the day after tr?,nsplantation. The duration of neutropenia was significantly shorter in the ciprofloxacin group compared with the TMPISMX group (16 days vs. 22 days; P = 0.0006). There were no differences between the groups with respect to the time of first febrile episode or the incidence of bacteremia. Ofloxacin has also been evaluated as a prophylactic agent. Winston and coworkers [36] compared the safety and efficacy of ofloxacin (300r:lg twice daily) plus nystatin with vancomycin (500mg orally every 8 hours) plus polymyxin (100mg every 8 hours) in 62 patients with hematologic malignancies. In the ofloxacin-treated patients, there were significantly fewer gram-negative infections and septicemia. Neither regimen was effective in preventing colonization or infection with gram-positive bacteria. Colonization developed with three ofloxacin-resistant gram-negative bacilli, but none caused infection. Ofloxacin-nystatin was associated with improved patient compliance and tolerability compared with vancomycin-polymyxin. Liang and associates [37] siudied the efficacy of ofloxacin (300mg twice daily) versus TMPISMX (1601800mg twice daily) in 102 patients with hematologic malignancies. There were fewer febrile episodes and gram-negative bacteremias in the ofloxacin group. No documented cases of gram-positive bacteremia occurred during this investigation. Surveillance cultures were performed during this investigation and demonstrated that more bacterial isolates were TMPISMX resistant than ofloxacin resistant. Ofloxacin was associated with less adverse reactions, including rash than TMPISMX. Oral quinolones have also been shown to decrease the requirement for antimicrobial therapy in neutropenic patients with hematologic malignancies [38]. In an open, nonrandomized study, patients who received prophylaxis with norfloxacin (400mg twice daily) or ciprofloxacin (500mg twice daily) were followed to determine if they required less antibiotic therapy directed at

gram-negative organisms. Suspected infections were treated with vancomycin and ceftazidime. There was only 1 documented gram-negative infection in the 55 febrile episodes. The antimicrobial regimen was modified, if necessary, to include metronidazole or amphotericin B. This investigation did not include a control group, but permitted a reduction in the amount and duration of ceftazidime in febrile neutropenic patients. Obviously, a randomized, controlled study would be required to confirm the results of this observational strategy. While currently available data suggest that quinolones are safe and effective in the prevention of gram-negative infections in neutropenic patients, the two emerging problems are the increased incidence of gram-positive bacteria and the development of resistance. The issue of gram-positive infections has been addressed with the use of roxithromycin [31], penicillin [39], or vancomycin [40]; however, the issue of antimicrobial resistance is a growing concern. Carratal5 and colleagues [28] observed that 37% of their E. coli from bacteremic cancer patients were quinolone resistant (norfloxacin MICs, 16128yg/mL and ciprofloxacin MICs, 8-64 pglmL). In a case-control study, they found the only factor that was associated with the development of quinoloneresistant E. coli bacteremia was prophylaxis with norfloxacin. Quinoloneresistant E. coli bacteremia in cancer patients is not unique to norfloxacin. Kern and coworkers [41] noted that the incidence of quinolone-resistant E. coli bacteremia in patients with leukemia increased from <0.5% to 4.5% over a 6-year period during which most patients received ofloxacin for prophylaxis. Furthermore, increasing use of these agents throughout a medical center may also select for resistance in Pseudomonas aeruginosa [42]. Although the quinolones appear to be attractive agents for antibacterial prophylaxis in neutropenic patients, they must be used cautiously due to their lack of grampositive activity and the emergence of resistance. In summary, antibacterial prophylaxis with TMPISMX or quinolones has resulted in a decreased frequency of gram-negative infections in neutropenic patients. However, there are no data indicating that this therapy improves outcome, that is, prolongs survival or increases remission rates. If antibacterial prophylaxis is used, because of the relative increase in gram-positive infections, consideration should be given to the addition of an agent with activity against streptococci. Although it may be difficult to conduct a large, controlled clinical trial of antibacterial prophylaxis with well-defined endpoints, this would be necessary prior to definitive recommendations regarding the prevention of bacterial infections in neutropenic patients. 4. Prevention of fungal infections Fungal infections are an important cause of morbidity and mortality in patients with neoplastic diseases [26,27,39]. The major risk factors for systemic fungal infections, such as neutropenia, broad-spectrum antibiotics, immuno-

suppression, corticosteroids, hyperalimentation, and central venous catheters, are frequently present in patients with cancer. It is estimated that nearly one half of bone marrow transplant recipients will develop a fungal infection if neutropenia persists more than 3 weeks, and this is associated with significant mortality [43]. While Candida and Aspergillus are the most frequently encountered fungal species, emerging pathogens include Fusarium, Alternaria, and Trichosporon. Unlike bacteria1 infections, the diagnosis of an invasive fungal infection in patients with malignancies is often challenging and early identification is difficult. Furthermore, outcome in patients with systemic fungal infections is generally poor. Therefore, strategies for the prevention of invasive fungi can be anticipated to improve survival. The earliest studies evaluating antifungal prophylaxis examined either nystatin or oral amphotericin B [44,45]. In a retrospective study of children with leukemia, Carpentieri and associates [45] showed that nystatin was effective in preventing candidiasis. Although there was initial optimism with this agent, subsequent studies have failed to demonstrate a significant benefit because fungemia and disseminated candidiasis may develop in patients even with high-dose (20-30 million units daily) therapy [46,47]. In addition, noncompliance and unpalatabilty have limited its use. Several investigators have also evaluated the role of oral amphotericin B in the prevention of fungal infections in cancer patients. Oral amphotericin B (50mg four times a days) was shown to be superior to placebo in preventing disseminated candidiasis in patients with hematologic malignancies. However, other investigations have failed to demonstrate a benefit with this agent as well. Similar to nystatin, oral amphotericin B is poorly tolerated. The most significant advances with respect to antifungal prophylaxis have occurred with the imidazoles (clotrimazole, ketoconazole, and miconazole) and triazoles (fluconazole and itraconazole). Clotrirnazole has been shown to be better than placebo in preventing oral candidiasis in patients with solid tumors and leukemia [48,49]. However, clotrimazole is only available as an oral preparation and does not have any systemic absorption. Therefore, this imidazole would not be expected to prevent invasive fungal infections in immunocompromised hosts. In a double-blind study, intravenous miconazole (5mg/kg every 8 hours) was compared with placebo in patients with prolonged neutropenia [50].At the onset of fever, therapy was initiated and continued until neutropenia resolved or an infectious endpoint was reached. This trial revealed that fungemia occurred less frequently in the miconazole-treated patients; however, there was no difference in infection-related mortality. Intravenous miconazole was well tolerated in this study, but serious ventricular arrhythmias have been associated with this antifungal agent [51]. Ketoconazole was the first orally administered imidazole evaluated for antifungal prophylaxis. In a prospective, double-blind study, Hansen and colleagues [52] compared ketoconazole (400mg daily) with placebo for the prevention of fungal infections in patients with cancer. Ketoconazole was

effective in preventing oral candidiasis but failed to prevent Candicln esophagitis and vulvovaginitis. However, the prophylactic use of ketoconazole did not alter outcome nor the need for amphotericin B. Ketoconazole has been compared with nystatin for the prevention of fungal infections in patients with neutropenia [47,53,54].The major conclusions that can be drawn from these studies is that ketoconazole is superior to nystatin and is better tolerated. In patients on ketoconazole, however, an increase in more resistant fungi (especially Candida (Ton~lol~sis) glabrntcl) has been noted. The use of ketoconazole is limited by its requirement for gastric acidity for optimal absorption and lack of a parenteral preparation. Therefore, in patients with achlorhydria, on H2 blockers, or with severe mucositis, absorption is erratic, which may result in subtherapeutic serum levels. In addition, ketoconazole has many drug interactions and is contraindicated in patients on nonsedating antihistamines (astemizole, terfenadine, and loratadine). The introduction of the triazole antifungal agents has been a significant advance in the prevention of fungal infections in patients with cancer. Fluconazole is available as both an oral as well as parenteral preparation and achieves high levels in tissue. Itraconazole, like ketoconazole, is only available as an oral preparation but provides activity against Aspergillzts fi~rnignt~ts. Several well-controlled trials have examined the use of fluconazole for antifungal prophylaxis. In a large, multicenter study, Goodman and associates [55] examined the role of fluconazole for the prevention of fungal infections in patients undergoing bone marrow transplantation. Patients were randomized to receive fluconazole (400mg daily) or placebo from the initiation of the conditioning regimen until granulocytopenia resolved, toxicity, or an infectious endpoint. In the Auconazole-treated patients, there was a significant reduction in superficial as well as systemic fungal infections compared with placebo. Therapy was well tolerated, although patients on fluconazole had more frequent liver function test abnormalities. There was no difference in overall mortality between the two groups, but there were less deaths attributed to fungal infections in the fluconazole group. In addition, empiric use of amphotericin B was delayed in the patients receiving fluconazole for prophylaxis. Another study compared the use of fluconazole (400mg daily) with placebo in patients undergoing bone marrow transplantation [56]. In this investigation, therapy continued until 75 days following transplantation or until a toxicity or infectious endpoint was reached. Fluconazole significantly reduced the incidence of systemic fungal infections as well as empiric alnphotericin B use. This study also demonstrated a survival benefit at 110 days following transplantation for the patients randomized to fluconazole for prophylaxis. Unlike the previous study, there was no increase in colonization with non-C. nlbicans candidal organisms. Fluconazole has also been evaluated for the prevention of fungal infections in patients with acute leukemia. In a multicenter study, Winston et al. [57]

randomized 257 patients with acute leukemia to receive fluconazole (400mg orally once daily or 200mg intravenously every 12 hours) or placebo. Similar to previous investigations, fluconazole was effective in preventing both fungal colonization and documented fungal infections, especially by Candida species (except C. krusei). Other fungal infections, such as aspergillosis, were uncommon in this trial, but were no different between the two groups. In this study fluconazole was not clearly effective in preventing invasive fungal infections, reducing amphotericin B usage, or improving outcome. Most recently, Schaffner and Schaffner [58] performed a double-blind, placebo-controlled study to determine if fluconazole (400mg) was effective in preventing fungal infections in patients with hematologic malignancies. They evaluated 96 patients and found that the time to initiation of amphotericin B was shorter in the placebo group than those treated with fluconazole. However, fluconazole did reduce the number of febrile days by 20%, but the incidence of systemic fungal infections and outcome was unaffected. The previous studies have compared fluconazole with placebo, but this triazole has also been compared with other oral antifungal agents. Ellis and coworkers 1.591 studied the efficacy of fluconazole (200mg daily) versus clotrimazole (10mg twice daily) plus mycostatin (500,000IU four times daily) in patients with hematologic malignancies or undergoing bone marrow transplantation. The fluconazole-treated patients had a significant reduction in fungal infections and survival was higher in patients receiving prophylaxis with fluconazole compared with those treated with clotrimazole. Another large clinical trial compared fluconazole (50mg daily) with oral amphotericin B (500mg four times a day) andlor nystatin (lOO,OOOIU four times a day) in 536 patients with cancer undergoing treatment [60]. Although this was an open-label study with a heterogenous patient population, the fluconazole-treated patients had fewer fungal infections, primarily oral candidiasis, than the polyene group. There was no difference in systemic fungal infections, amphotericin B usage, or fungal colonization between the two groups. In another study comparing fluconazole (50mg daily) with oral amphotericin B (200mg four times a day), Rozenberg-Arska and associates [61] found both regimes equally effective in preventing disseminated fungal infections in patients with leukemia. There was no difference in the use of intravenous amphotericin B between the groups, but fluconazole was better tolerated. Itraconazole has been investigated as prophylaxis against fungal infections; however, the trials have been nonrandomized. Thunnissen and colleagues [62] compared 92 patients prophylaxed with itraconazole (200mg twice daily) versus 80 control patients treated with nystatin during the previous 18 months. It is difficult to draw any significant conclusions from this study design; however, itraconazole was shown to be safe and effective in preventing fungal infections compared with historical controls. Another study evaluated the use of

itraconazole (200mg daily) plus amphotericin B nasal snray (10mg daily) for the prevention of aspergillosis in patients with hematologic malignancies [63]. Compared with historical controls, there was a significant reduction in invasive aspergillosis and mortality in patients on itraconazole. Tricot and others [64] compared itraconazole (200mg twice daily) with ketoconazole (200mg twice daily) in two nonrandomized studies for antifungal prophylaxis in neutropenic patients. There was a significantly higher incidence of serious fungal infections and fatal infections due to Aspergillus in the ketoconazole-treated group. Unfortunately, 24% of the itraconazole-treated patients had a proven fungal infection while on therapy, and there was no difference in the requirement for systemic amphotericin B between the two groups. An explanation for the high failure rate in this study was the observation that 21 of 42 itraconazole patients had inadequate serum levels and 11 of these developed proven or suspected fungal infections. In the only double-blind, placebo-controlled study, Vreugdenhil and coworkers [65] compared itraconazole (200mg daily) or placebo in addition to oral amphotericin B in patients with hematologic malignancies. While there was a decreased incidence of C. nlbicans infections in the itraconazole group, prophylaxis did not improve outcome. An advantage of itraconazole over fluconazole is its broader spectrum of activity, especially against Aspergillus. However, itraconazole is only available as an oral agent, which may limit its use in many patients treated for neoplastic diseases who are unable to tolerate medications by mouth. Controlled clinical trials comparing fluconazole with itraconazole are now needed to determine the optimal dose and agent for antifungal prophylaxis. Systemic amphotericin B has been the mainstay of therapy for documented or suspected fungal infections in immunocompromised hosts. Due to its broad spectrum of activity against most fungi, this agent has also been examined for prophylaxis. The downside of amphotericin B is nephrotoxicity, especially in patients with granulocytopenia who may be receiving other concurrent nephrotoxins. To avoid the problem of toxicity, Perfect and coworkers [66] assessed the safety and efficacy of low-dose amphotericin B (0.1mglkg daily) in patients undergoing autologous bone marrow transplantation. In this placebo-controlled study, there was a reduction in fungal colonization, but no difference in systemic fungal infections between the two groups. Although there was a reduction in mortality in the patients receiving prophylactic amphotericin B, this was not due to a prevention of fungal infections. Other than infusion-related adverse reactions, there was no difference in toxicities between the two groups. In a similar study, Riley and associates [67] randomized patients undergoing bone marrow transplantation to low-dose amphotericin B (0.1mglkg daily) or placebo for antifungal prophylaxis. There was a significantly higher incidence of systemic fungal infections and increased requirement for high-dose amphotericin B (l.Omg/kg daily) in patients randomized to placebo. While there was a trend toward decrease length of stay and improved survival in

the low-dose amphotericin B group, these differences were not significant. There were no differences in toxicity between the two groups; however, both groups received acetaminophen and intravenous meperidine before the infusion to diminish amphotericin B-related side effects and to facilitate blinding. In a retrospective analysis of 331 allogeneic bone marrow transplant patients, O'Donnell and others [68] identified prophylaxis with low-dose amphotericin B (5-10mg daily) as an important factor in reducing the incidence of systemic fungal infections. Following the institution of prophylaxis, the incidence of candidemia decreased from 15.8% to 2.7%. In addition, there was a reduction of invasive aspergillosis from 15.8% to 5.6%. There were no serious amphotericin B-related toxicities in this observational study. Of note, there was a significantly greater incidence of graft-versus-host disease in the amphotericin B-treated patients, which corresponded with lower cyclosporine levels. In summary, antifungal prophylaxis remains in evolution. However, the routine use of high-efficiency particulate air (HEPA) filtration has been demonstrated to be an important factor in decreasing the incidence of nosocomial aspergillosis in bone marrow transplant recipients. An important risk factor for systemic fungal infections is the duration (>3 weeks) and depth (
5. Prevention of viral infections The herpesviruses (herpes simplex virus, varicella-zoster virus, cytomegalovirus, Epstein-Barr virus, and human herpes virus-6) are the most common group of viruses that result in clinical infections in patients with cancer and usually represent reactivation of latent disease [26,27,39]. Patients with impaired cell-mediated immunity are at highest risk, especially those with leuke-

mia or undergoing bone marrow transplantation. While herpes simplex virus (HSV) infections are the most common, cytomegalovirus results in the greatest morbidity and mortality. Herpes simplex viruses have a high propensity for reactivation in patients receiving bone marrow transplantation. They generally develop within the first 7-14 days following transplantation and most commonly cause stomatitis and esophagitis. Sara1 and colleagues [69] conducted a double-blind, placebocontrolled study of acyclovir for prophylaxis in HSV-seropositive patients undergoing bone marrow transplantation. They found a significant reduction in documented HSV lesions in the acyclovir-treated patients compared with placebo. Therapy was well tolerated but did not eradicate latent infection. Prophylaxis with acyclovir (125-250 mg/m2every 8 hours) has also been shown to prevent reactivation of HSV in patients receiving chemotherapy for leukemia [70]. Oral acyclovir (400mg five times a day) is equally efficacious and considerably less costly than the parenteral preparation in patients who can tolerate oral therapy [71]. Although acyclovir is clearly effective as prophylaxis against HSV, acyclovir-resistant isolates have been identified. Foscarnet, an inhibitor of DNA polymerase, is effective as therapy for acyclovir-resistant HSV but has not been examined for prophylaxis. Varicella-zoster virus (VZV) is a common cause of mucocutaneous eruptions in immunocompromised hosts, and reactivation frequently occurs following bone marrow transplantation. Primary infection is less common, but can rapidly disseminate and is associated with a high mortality rate in immunocompromised patients. In a retrospective study, 21 patients with acute leukemia who underwent bone marrow transplantation were treated with acyclovir for 6 months to prevent reactivation of VZV [72]. No patient developed VZV while on prophylaxis; however, five had localized zoster within 1 month of discontinuing acyclovir. The data suggest that acyclovir prophylaxis delays but does not prevent reactivation of VZV in patients undergoing transplantation. Therefore, the role of long-term acyclovir in preventing VZV is controversial. In patients without previous infection with VZV, the administration of varicella zoster immune globulin (VZIG) to immunocompromised hosts reduces the incidence and severity of varicella [73]. Treatment should be given within 72 hours if a susceptible host is exposed to varicella. While VZIG is an effective method of prophylaxis against varicella, it has no effect on the prevention of zoster. A live attenuated varicella vaccine is currently avaiIable in the United States. Universal immunization of susceptible children has been advocated due to the safety and efficacy of the vaccine [74]. The vaccine should not be given routinely to immunocompromised individuals, except children with acute lymphocytic leukemia (ALL). In children with ALL in remission for at least 1 year, immunization has been shown to be safe and protective [75].Due to the potential that varicella can be transmitted following immunization, vaccinated individuals who develop a rash should avoid contact with susceptible immunocompromised hosts. If this vaccine is routinely

Table 2. Common pathogens associated with infection in immunocompromised hosts

Neutropenia Bacteria Enterobncter spp. E.~chericl~ia coli Klebsielln pneunzonicre Psel~donzonasaerrlginosa Stc~phylococ-cus arrrerrs Sfnplzylococci (coagulase-negative) Streptococci (a-hcmolytic) Fungi Asyergil/~i,\ spp. Crrndi~lnspp. Rhizoplis spp. Defects in cell-mediated immunity Bactcria Legiotlellcr pne~tnropl?ila Listerin r?zonocytogenes Mycobacteriu spp. Snlnzonellrr spp. Fungi Coccicliodes imrnrtis Cryptococc~lsneoforitlnrr~ Hisroplci.snrcr ccry;rrilatlrtn Viruses Cyton~egalovirus Herpes simplex Varicella-7oster Protozoa Pnerirfnocysti~cnril~ii Toxoplasnla gorzdii Defects in humoral immunity Bacteria Hner~iophilrlsinfllienzae Strrptococc~tspneltnroniae Emerging pathogens in cancer patients Bacteria Acinetobcrcter spp. Clo.rrricliltrn septicrrnt Corynebacteril~~z jeikeirrrn P.serrdor~~ona.s cepc~cirl Stenotrophorr~onns(Xnrzttzornonas) n7nltophilia Vancomycin-resistant enterococci Fungi Cetnclitla krusei Fuscrriirnz spp. Pser~dallesctzeriahoydii Torrilopsi~glabrrrta Triclzospor.or1 heigelli

administered to healthy children and protective immunity persists, varicella should be an uncommon infection in patients with cancer. Famciclovir, a diacetyl ester prodrug of penciclovir, inhibits viral DNA synthesis and has been approved for treatment of zoster in immunocompromised hosts. The

efficacy of this new antiviral agent for treatment or prophylaxis in immunocompromised patients is unknown. Cytomegalovirus (CMV) infection is a common infection in patients undergoing bone marrow transplantation and the most frequent cause of infectionrelated mortality. CMV is an uncommon infection in patients with solid tumors. In persons seropositive for CMV, reactivation occurs in 70-80% of allogeneic bone marrow transplant recipients and 10% of those receiving autologous transplantation. CMV infection arises either from exogenous sources such as blood products or by reactivation. Prevention of infection in the seronegative patient can be accomplished by using CMV-negative blood products or blood filters [76]; therefore, most CMV disease results from reactivation. The earliest attempts to prevent CMV infection in transplant recipients were with acyclovir. Meyers and coworkers [77] administered intravenous acyclovir (500mg/m2 every 8 hours) to allogeneic bone marrow transplant recipients in an attempt to prevent CMV infection and disease. Patients who were seropositive for both HSV and CMV received prophylaxis and those seropositive only for CMV served as controls. Patients who received acyclovir had a significant reduction in CMV infection and disease following transplantation as well as improved survival at 100 days compared with the control group. Despite the use of high-dose acyclovir in this study, invasive CMV disease still developed in 22% of the acyclovir recipients. In a study of 310 patients undergoing allogeneic bone marrow transplantation, Prentice and associates [78] randomized patients to one of three treatment arms: intravenous acyclovir (500mg/m2 three times daily) for 1 month followed by oral acyclovir (800mg four times daily) for an additional 6 months, intravenous therapy alone, or low-dose oral acyclovir (200-400mg four times daily) for the prevention of CMV infection. They observed that patients receiving intravenous acyclovir had a significantly reduced risk of developing CMV infection, but the addition of oral therapy provided no additional benefit. In addition, the patients treated with high-dose acyclovir followed by oral therapy had improved survival. More recently, Boeckh and colleagues [79] retrospectively reviewed their experience with high-dose acyclovir (500mg/m2every 8 hours intravenously) in 266 CMV-seropositive autologous bone marrow transplant recipients. They observed that there was no significant difference in the incidence of CMV pneumonia at day 100 between patients receiving acyclovir and those who did not receive prophylaxis. In addition, there was no difference with respect to mortality due to CMV pneumonia or survival following transplantation. Although acyclovir is effective in preventing CMV infections, the introduction of ganciclovir, a more potent inhibitor of CMV, is the preferred agent. There have been two approaches taken to prevent CMV disease with ganciclovir: preemptive therapy for patients who have a positive culture for CMV or prophylactic treatment prior to infection. Because isolation of CMV in culture is a strong predictor of active disease, Schmidt et al. [80] used

asymptomatic pulmonary infection as the determinant for ganciclovir therapy (5 mg/kg twice daily for 14 days, then daily until 120 days) in patients undergoing allogeneic bone marrow transplantation in an attempt to decrease the incidence of CMV pneumonitis. On day 35 following transplantation, patients with CMV cultured from bronchoalveolar lavage were randomized to ganciclovir or no treatment. The incidence of CMV pneumonitis was significantly reduced from 70% in the patients without treatment to 25% in the ganciclovir-treated group. Of note, no patient who completed therapy with ganciclovir developed CMV pneumonitis. An alternative approach was explored by Goodrich and colleagues [81], who randomized allogeneic transplant recipients with CMV cultured from blood, urine, throat, or bronchoalveolar lavage to ganciclovir (5mglkg twice daily for 1 week then daily for 100 days) or placebo. CMV disease developed in only 1 of 37 ganciclovir-treated patients compared with 15 of 35 patients in the placebo group. Overall survival at both 100 days and 6 months was greater in the ganciclovir group. The most common adverse reaction was bone marrow suppression, which resolved when the ganciclovir was discontinued. An important observation in both of the previous trials was that patients can develop CMV disease without having positive surveillance cultures. Therefore, the administration of ganciclovir prophylaxis to all CMVseropositive patients receiving allogenic transplantation has been evaluated. In a nonrandornized study, Atkinson and coworkers [82] treated patients with ganciclovir (5mglkg twice daily) for 1 week prior to transplantation, then thrice weekly starting on day 21 or when the absolute neutrophil count was >1 x 1 0 " ~until day 84. Compared with historical controls, the ganciclovirtreated patients had a significant reduction in CMV disease and pneumonitis. Although a small uncontrolled study, the results suggest that prophylactic ganciclovir given thrice weekly following engraftment can prevent CMV disease. In a double-blind, placebo-controlled trial, Goodrich and others [83] randomized CMV-seropositive patients to receive ganciclovir (5 mglkg twice daily for 5 days then once daily until day 100) or placebo at the time of engraftment. A significant reduction in both CMV infection and disease was identified in the ganciclovir group. No patient given prophylaxis with ganciclovir developed CMV disease. Despite the marked reduction in CMV, there was no difference in mortality between the two groups. There was a higher incidence of neutropenia in the ganciclovir-treated patients and an increased risk of bacterial infections. In a similar study, Winston and associates [84] randomized patients to ganciclovir (2.5 mglkg three times daily for 1week then 6mglkg daily five times a week until day 100) or placebo. As in the previous study, daily therapy began at the time of engraftment. There was a significant reduction in the incidence of CMV infection but not CMV disease. Once again, there was no difference in mortality between the patients treated with ganciclovir or placebo.

An oral preparation of ganciclovir is currently available in the United States and Europe, but has not been evaluated for the prevention of CMV in bone marrow transplant patients. The role of foscarnet in the prevention of CMV infection in bone marrow transplant recipients requires additional investigation. An advantage of foscarnet over ganciclovir is no myelosuppression; however, this is offset by nephrotoxicity. Another recently approved antiviral agent. with potent activity against CMV as well as other herpes viruses, is cidofovir, but this compound has not been investigated in cancer patients. Improved methods to detect CMV by the polymerase chain reaction may provide additional insights into the patients most likely to benefit from prophylaxis. In summary, for patients undergoing autologous bone marrow transplantation, the risk of CMV disease is relatively low and data evaluating prophylaxis are lacking. Because the incidence of CMV disease is significantly higher in seropositive patients following allogeneic transplantation, two methods of prophylaxis have been evaluated. While the optimal role of ganciclovir prophylaxis remains to be determined, it is clearly effective in preventing CMV disease following engraftment. The risk factors for CMV disease, cost, and toxicity must be considered when deciding whether patients benefit most from preemptive therapy following a positive culture or routine prophylaxis.

6. Prevention of protozoal infections The most commonly identified protozoal infection in patients with cancer is Pne~~rnocystis cnrinii; however, recent evidence suggests that this organism is a fungus. Although P. cnrinii pneumonia (PCP) is a relatively common infection due to the AIDS epidemic, two features of this organism remain controversial - taxonomy and transmission. Examination of ribosomal RNA suggests that P. cnrinii more closely resembles fungi [85], but the life cycle and microbiologic characteristics suggest the organism is a protozoan [86]. Most of the data suggest that P. cnrinii arises from reactivation of latent infection because virtually everyone has been exposed to this organism by adulthood. However, there have been clusters of PCP reported in hospitals [87] and possible person-to-person transmission [88]. The patients at greatest risk for PCP are those with deficiencies in cellmediated immunity. While most infections in patients with cancer develop during periods of neutropenia, PCP can occur at any time T-cell function is impaired. This results most commonly from chemotherapy and/or corticosteroids. A pivotal study by Hughes and coworkers [17] demonstrated that TMP/ SMX was successful in preventing PCP in children with leukemia. Although no controlled studies have examined the role of PCP prophylaxis in patients undergoing bone marrow transplantation, treatment with TMPiSMX is gener-

ally administered. The optimal dose and duration of prophylaxis are unknown. Most centers administer TMPISMX to allogeneic transplant recipients either daily or three times a week from engraftment until 6-12 months following transplantation [39]. Prophylaxis to prevent PCP is generally not necessary for patients undergoing autologous bone marrow transplantation. Recently an increased incidence of PCP has been noted in patients receiving corticosteroids for brain tumors; therefore, prophylaxis should be considered in these patients 1891. PCP prophylaxis is also indicated in patients with ALL. Although TMPISMX is safe and effective,myelosuppression as well as cutaneous reactions can occur. In patients who are allergic or do not tolerate TMPISMX, aerosolized pentamidine or dapsone are effective alternatives. Atovaquone, a recently approved hydroxynapthoquinone, has been used for treatment of PCP, but its role in prophylaxis is unknown. Reactivation of Toxoplnsmn gondii is an unusual infection in patients with cancer. Most cases occur in patients with Hodgkin's disease or within the first 6 months following allogeneic bone marrow transplantation [90]. Similar to P. carinii, this protozoan usually results in a self-limited disease in immunocompetent individuals, but clinical infection in those with defects in cellmediated immunity. Prophylaxis against toxoplasmosis in transplant patients has been examined by Foot and colleagues [91]. Pyrimethamine/sulfadoxine was administered to 69 patients with positive serology against T. gondii. Treatment was initiated at the time of engraftment and continued for 6 months or longer if immunosuppression persisted. Although there was no control group, no patients on prophylaxis developed toxoplasmosis nor pneumocystosis. The nlost common adverse reaction of this medication was myelosuppression. Slavin and associates [92] reviewed their experience with toxoplasmosis in nearly 4000 bone marrow transplant recipients. An estimated 15% of the patients were seropositive and 2% of these patients developed toxoplasmosis. Therefore, infection occurred at a rate of 0.31 cases per 100 allogeneic transplants. Infection generally occurred within the first 6 months of transplant and in patients with severe graft-versus-host disease. There were no cases among individuals receiving an autologous transplant. Routine prophylaxis against toxoplasn~osisis not indicated in nonendemic areas.

7. Summary The prevention of infection in patients with cancer has changed tremendously over the last decade, but remains in evolution. Despite many clinical trials examining the role of antibacterial, antifungal, and antiviral prophylaxis, there is still discussion among physicians about not only which patients require prophylaxis, but also the optimal regimen. Nevertheless, many of these regimens offer the hope to prevent infection in patients with underlying ileoplastic diseases. There is no therapy that is uniformly effective in all settings. This is

generally due to the severity of the defects in host defenses and the virulence of the microorganism. Hopefully, the future will hold many new therapeutic options to help prevent infection in patients with cancer.

References 1. Clift RA, Buckner CD, Fefer A, et al. Infectious complications of marrow transplantation. Transpl Proc 1974;6:389-393. 2. Watson JG. Problems of infection after bone marrow transplantation. J Clin Path01 1983;36:638-692. 3. Garner JS, Favero MS. CDC guidelines for handwashing and hospital environmental control. Infect Control 1986;7:231-235. 4. Bodey GP, Buckley M, Satha YS, Freidrich EJ. Quantitative relationships between circulating leukocytes and infections in patients with acute leukemia. Ann Intern Med 1966;64:328-340. 5. Pfaller MA. Epidemiology and control of fungal infections. Clin Infect Dis 1994;19(Suppl. 1):S8-S13. 6. Morris JG Jr., Shay DK, Hebden JN, et al. Enterococci resistant to multiple antimicrobial agents, including vancomycin. Establishment of endemicity in a university medical center. Ann Intern Med 1995;123:250-259. 7. Noskin GA, Cooper I, Peterson LR. Vancomycin-resistant Enterococc~~s fneciunz sepsis following persistent colonization. Arch Intern Med 1995;155:1445-1447. 8. Nauseef WM, Maki DG. A study of simple protective isolation in patients with granulocytopenia. N Engl J Med 1981;304:448-453. 9. Schimpff SC, Hahn DM, Brouillet MD, Young VM, Fortner CL, Wiernik PH. Comparison of basic prevention techniques, with standard room reverse isolation or with reverse isolation plus added filtration. Leuk Rcs 1978;2:231-240. 10. Remington JS, Schimpff SC. Please don't eat the salads. N Engl J Med 1981;304:433435. 11. Yates JW, Holland JF. A controlled study of isolation and endogenous microbial suppression in acute myelocyte leukemia patients. Cancer 1973;32:1490-1498. 12. Buckner CD, Clift RA, Sanders JE. et al. Protective environment for marrow transplant recipients. A prospective study. Ann Intern Med 1978:89:893-901. 13. Levine AS, Siege1 SE, Schreiber AD, et al. Protected environments and prophylactic antibiotics. A prospective controlled study of their utility in the therapy of acute leukemia. N Engl J Med 1973;288:477-483. 14. Schimpff SC. Green WH, Young VM, et al. Infection prevention in acute nonlymphocytic leukemia. Laminar air flow room reverse isolation with oral, nonabsorbable antibiotic prophylaxis. Ann Intern Med 1975;82:351-358. 15. Van der Waaij D, Berghuis-de Vries JM. Selective elimination of Enterobacteriaceae species from the digestive tract in mice and monkeys. J Hyg (Camb) 1974;72:205-211. 16. Van der Waaij D, Berghuis-de Vries JM. Determination of the colonization resistance of the digestive tract of individual mice. J Hyg (Camb) 1974;72:379-387. 17. Hughes WT, Kuhn S, Chaudhry SC, et al. Successful chemoprophylaxis for Pne~~r71ocysri.s cnritzii pneumonitis. N Engl J Med 1977;297:1419-1426. 18. Enno A, Catovsky D , Darrell J, Goldman JM, Hows J, Galton DAG. Co-trimazole for prevention of infection in acute leukaemia. Lancet 1978;2:395-397. 19. Starke ID, Donnelly P, Catovsky D, et al. Co-trimazole alone for prevention of bacterial infection in patients with acute leukaemia. Lancet 1982;1:5-9. 20. de Jongh CA, Wade JC, Finley RS, et al. Trimethoprim/sulfamethoxazole versus placebo: A double blind comparison of infection prophylaxis in patients with small cell carcinoma of the lung. J Clin Oncol 1983;1:302-307.

21. Dekker AW, Rozenberg-Arska M, Sixma JJ, Verhoef J. Prevention of infection by trimethoprim-sulfamethoxazole plus amphotericin B in patients with acute nonlymphocytic leukemia. Ann Intern Med 1981;95:555-559. 22. Gurwith MJ, Brunton JL, Lank BA, Harding GK, Ronald AR. A prospective controlled investigation of prophylactic trimethoprim/sulfamethoxazole in hospitalized granulocytic patients. Am J Med 1979;66:248-256. 23. Weiser B, Lange M, Fialk MA, Singer C, Szatrowski TH, Armstrong D. Prophylactic trimethoprim-sulfamethoxazole during consolidation therapy for acute leukemia: A controlled trial. Ann Intern Med 1981;95:436-438. 24. Wade JC, de Jongh CA, Newman KA, Crowley J, Wiernik PH, Schimpff SC. Selective antimicrobial modulation as prophylaxis against infection during granulocytopenia: Trimethoprim-sulfamethoxazole vs. nalidixic acid. J Infect Dis 1983;147:624-634. 25. EROTC. Trimethoprim-sulfamethoxazole in the prevention of infection in neutropenic patients. EROTC International Antimicrobial Therapy Project Group. J Infect Dis 1984;150:372-379. 26. Verhoef J. Prevention of infection in the neutropenic patient. Clin Infect Dis 1993;17(Suppl. 2):S359-S367. 27. Hathorn JW. Critical appraisal of antimicrobial for prevention of infection in immunocompromised hosts. Hematol Oncol Clin North Am 1993:7:1051-1099, 28. Carratali J, Fernidez-Sevilla A, Tubau F, Callis M, Gudiol F. Emergence of quinoloneresistant Escherichia coli bacteremia in neutropenic patients with cancer who have received prophylactic norfloxacin. Clin Infect Dis 1995;20:557-560. 29. Karp JE, Merz WG, Hendricksen C, et aI. Oral norfloxacin for prevention of gram-negative bacterial infections in patients with acute leukemia and granulocytopenia: A randomized, double-blind, placebo-controlled trial. Ann Intern Med 1987;106:1-7. 30. Dekker AW, Rozenberg-Arska M, Verhoef J. Infection prophylaxis in acute leukemia: A comparison of ciprofloxacin with trimethoprim-sulfamethoxazole and colistin. Ann Intern Med 1987;106:7-11. 31. Rozenberg-Arska M, Dekker A, Verdonck, Verhoef J. Prevention of bacteremia caused by ahemolytic streptococci by roxithromycin (RU-28,965) in granulocytopenic patients receiving ciprofloxacin, Infection 1989;17:240-244. 32. GIMEMA Infection Program. Prevention of bacterial infections in neutropenic patients with hematologic malignancies. A randomized, multicenter trial comparing norfloxacin with ciprofloxacin. Ann Intern Med 1991:115:7-12. 33. Lew MA, Kehoe K, Ritz J, et al. Ciprofloxacin versus trimethoprim/sulfamethoxazolefor prophylaxis of bacterial infections in bone marrow transplant recipients: A randomized, controlled trial. J Clin Oncol 1995;13:239-250. 34. Gilbert C, Meisenberg B, Vredenburgh J, et al. Sequential prophylactic oral and empiric once-daily parenteral antibiotics for neutropenia and fever after high-dose chemotherapy and autologous bone marrow support. J Clin Oncol 1994:12:1005-1011. 35. Imrie KR, Prince HM, Couture F, Brandwein JM, Keating A. Effect of antimicrobial prophylaxis on hematopoietic recovery following autologous bone marrow transplantation: Ciprofloxacin versus co-trimazole. Bone Marrow Transplant 1995:15:267-270. 36. Winston DJ, Ho WG, Bruckner DA, Gale RP. Champlin RE. Ofloxacin versus vancomycinl polymyxin for prevention of infections in granulocytopenic patients. Am J Med 1990;88: 36-42. 37. Liang RH, Yung RW, Chan TK, et al. Ofloxacin versus co-trimazole for prevention of infection in neutropenic patients following cytotoxic chemotherapy. Antimicrob Agents Che~nother1990;34:215-218. 38. Bow EJ, Loewen R, Vaughan D. Reduced requirement for antibiotic therapy targeting gramnegative organisms in febrile, neutropenic patients with cancer who are receiving antibacterial chemoprophylaxis with oral quinolones. Clin Infect Dis 1995;1995:907-912. 39. Momin F, Chandrasekar PH. Antimicrobial prophylaxis in bone marrow transplantation. Ann Intern Med 1995;123:205-215.

40. Attal M, Schlaifer D, Rubie H. et al. Prevention of gram-positive infections after bone marrow transplantation by systemic vancomycin: A prospective, randomized trial. J Clin Oncol 1991;9:865-870. 41. Kern WV, Andriof E, Oethinger M, Kern P, Hacker J, Marre R. Emergence of fluoroquinolone-resistant Esclzerichia coli at a cancer center. Antilnicrob Agents Chemother 1994;38:681-687. 42. Peterson LR, Postelnick M, Pozdol TL. Reisberg B, Noskin G A . Mailagement of emerging fluoroquinolone resistance in Pse~{do~nonas aerl~ginosa:Outcome of focusing institutional fluoroquinolone use in a referral medical center. Clin Infect Dis, in press. 43. Goodrich JM, Reed EC, Mori M, et al. Cliilical features and analysis of risk factors for invasive candidal infection after marrow transplantation. J Infect Dis 1991;164:731-740. 44. Ezdinli EZ, O'Sullivan D D , Wasser LP. Kim U. Stutzrnan L. Oral amphoterici~lfor candidiasis in patients with hematologic neoplasms. JAMA 1979:242:258-260. 45. Carpenticri U. Haggard ME, Lockhart LH, Gustavson LP. Box QT, West EF. Clinical experience in prevention of candidiasis by nystatin in children with acute lymphocytic leukemia. J Pediatr 1978:92:593-595. 46. DeGregorio MW, Lee WM, Ries CA. Clrirdiria infections in patients with acute leukemia: Ineffectiveness of nystatin prophylaxis and relationship between oropharyngeal and systemic candidiasis. Cancer 1982:50:2780-2784. 47. Jones PG, Kauffman CA, McAuliffe LS, Licpmail MK, Bergman A G . Efficacy of ketoconazole v. nystatin in prevention of fungal infections in neutropenic patients. Arch Intern Med 1984;144:549-551. 48. Owens NJ, Nightingale CH, Schweizer RT, Schauer PK, Dekker PT. Quintilliani R. Prophylaxis of oral candidiasis with clotri~nazoletroches. Arch Intern Med 1984;144:290293. 49. Cuttner J, Troy KM, Funaro L, Brenden R. Bottone EJ. Clotrimazole treatment for prevention of oral candidiasis in patients with acute leukemia undergoing chemotherapy. A m J Med 1986;81:771-774. 50. Wingard JR, Vaughn WP, Braine HG. Merz WG, Sara1 R. Prevention of fungal sepsis in patients with prolonged ncutropenia: A randomized double-blind placebo-controlled trial of intravenous n~iconazole.Am J Med 1987;83:1 103-1 110. 51. Fainstein V. Bodep GP. Cardiorespiratory toxicity due to miconazole. Ann Intern Med 1980;93:432433. 52. Hansen RM, Reinerio N, Sohnle PG, et al. Ketoconazole in the prevention of candidiasis in patients with cancer. Arch Intern Med 1987;147:710-712. 53. Shcpp D H , Klosterman A, Siege1 MS, Meycrs JD. Comparative trial of ketoconazole and nystatiil for prevention of fungal infection in neutropenic patients treated in a protective environment. J Infect Dis 1985:152:1257-1263. 54. Vogler WR, Malcolm LG, Winton EF. A randomized trial comparing ketoconazole and nystatin prophylactic therapy in neutropenic patients. Cancer Invest 1987;5:267-273. 55. Goodman JL, Winston DJ, Greenfield RA, et al. A controlIed trial of flucoilazole lo prevent fungal infectioils in patients undergoing bone marrow transplantation. N Engl J Med 1992;326:845-85 1. 56. Slavin MA, Osborne B. Adams R, ct al. Efficacy and safety of fluconazole prophylaxis for fungal infections after marrow transplantation - a prospective. randomized, double-blind study. J Infect Dis 1995:171:1545-1552. 57. Winston DJ, Cllandrashekar PH. Lazarus HM, et al. Fluconazole prophylaxis of fungal infections in patients with leukemia. Ann Intern Med 1993:118:495-503. 58. Schaffner A , Schaffner M. Effect of prophylactic fluconazole 011 the frequency of fungal infections, amphotericin B use, and health care costs in patients unclcrgoi~lgintensive chemotherapy for hematologic neoplasms. J Infect Dis 1995:172:1035-1041. 59. Ellis ME, Clink H. Ernst P, et al. Controlled study of fluconazole in thc prevention of fungal infections in neutropenic patients with hematological malignancies and bone marrow transplant recipients. Eur J Clin Microbiol Infect Dis 1994;13:3-11.

60. Philpott-Howard JN, Wade JJ, Mufti GJ, Brammer KW, Ehninger G. Randomized comparison of oral fluconazole versus oral polyenes for the prevention of fungal infection in patients at risk of neutropenia. J Antimicrob Chemother 1993;31:973-984. 61. Rozenberg-Arska M, Dekker AW, Branger J, Verhoef J. A randomized study to compare oral fluconazole to amphotericin B in the prevention of fungal infections in patients with acute leukemia. J Antimicrob Chemother 1991;27:369-376. 62. Thunnissen PLM, Sizoo W, Hendricks WDH. Safety and efficacy of itraconazole in prevention of fungal infections in neutropenic patients. Neth J Med 1991;39:84-91. 63. Todeschini G, Murari C. Bonesi A, et al. Oral itraconazole plus nasal amphotericin B for prophylaxis of invasive aspergillosis in patients with hematological malignancies. Eur J Clin Microbiol Infect Dis 1993;12:614-618. 64. Tricot G , Joosten E, Boogaerts MA, Vande Pitte J, Cauwenbergh G. Ketoconazole vs. itraconazole for antifungal prophylaxis in patients with severe granulocytopenia: Preliminary results of two nonrandomized studies. Rev Infect Dis 1987;9(Suppl. 1):S94-S99. 65. Vreugdenhil G, Van Dijke BJ, Donnelly JP, et al. Efficacy of itraconazole in the prevention of fungal infections among neutropenic patients with hematologic malignancies and intensive chemotherapy. A double blind, placebo controlled study. Leuk Lymphoma 1993;11:353-358. 66. Perfect JR, Klotman ME, Gilbert CC, et al. Prophylactic intravenous amphotericin B in neutropenic autologous bone marrow transplant recipients. J Infect Dis 1992;165:891-897. 67. Riley DK. Pavia AT, Beatty PG. et al. The prophylactic use of low-dose amphotericin B in bone marrow transplant patients. Am J Med 1994:97:509-514. 68. O'Donnell MR, Schmidt GM, Tegtmeier BR, et al. Prediction of systemic fungal infection in allogcneic marrow recipients: Impact of amphotericin prophylaxis in high-risk patients. J Clin Oncol 1994;122327-834. 69. Sara1 R, Burns WH. Laskin OL, Santos GW, Lietman PS. Acyclovir prophylaxis of herpessimplex-virus infections. N Engl J Med 1981;305:63-67. 70. Meyers JD, Wade JC, Mitchell CD, et al. Multicenter collaborative trial of intravenous acyclovir for treatment of mucocutaneous herpes simplex virus infection in the immunoconlpromised host. Am J Med 1982;73:229-235. 71. Straus SE, Seidlin M, Takiff H, Jacobs D, Bowden D, Smith HA. Oral acyclovir to suppress recurring herpes simplex virus infections in immunodeficient patients. Ann Intern Med 1984;100522-524. 72. Senlpere A. Sanz GF, Senet L, et al. Long-term acyclovir prophylaxis for the prevention of varicella zoster virus infection after autologous blood stem cell transplantation in patients with acute leukemia. Bone Marrow Transplant 1992;10:495498. 73. Zaia JA. Levin MJ, Preblud SR, et al. Evaluation of varicella-zoster immune globulin: Protection of immunosuppressed children after household exposure to varicella. 3 Infect Dis 1993:147:737-743. 74. Committee on Infectious Diseases. Recommendations for the use of a live attenuated varicella vaccine. Pediatrics 1995:95:791-795. 75. Gershon A A , Steinberg SP. Persistence of immunity of varicella in children with leukemia immunized with live attenuated varicella vaccine. N Engl J Med 1989;320:892-897. 76. Zaia JA. Prevention and treatment of cytomegalovirus pneumonia in transplant recipients. Clin Infect Dis 1993;17(Suppl. 2):S392-399. 77. Meyers JD. Reed EC, Shepp DH. et al. Acyclovir for prevention of cytomegalovirus infection and disease after allogeneic marrow transplantation. N Engl J Med 1988:318:70-75. 78. Prentice HG, Gluckman E, Powles RL, et al. Impact of long-term acyclovir on cytomegalovirus infection and survival after allogeneic bone marrow transplantation. European acyclovir for CMV prophylaxis study group. Lancet 1994;343:749-753. 79. Boeckh M. Gooley TA, Reusser P, Buckner CD, Bowden RA. Failure of high-dose acyclovir to prevent cytomegalovirus disease after autologous marrow transplantation. J Infect Dis 1995;172:939-943. 80. Schmidt GM, Horak DA, Niland JC, et al. A randomized, controlled trial of prophylactic ganciclovir for cytomegalovirus pulmonary infection in recipients of allogeneic bone marrow

81.

82.

83. 84.

85. 86. 87. 88.

89. 90. 91.

92.

transplants: The City of Hope-Stanford-Syntex CMV study group. N Engl J Med 1991;324:1005-1011. Goodrich JM, Mori M, Gleaves CA, et al. Early treatment with ganciclovir to prevent cytomegalovirus disease after allogeneic bone marrow transplantation. N Engl J Med 1991;325:1601-1607. Atkinson K, Downs K, GoIenia M, et al. Prophylactic use of ganciclovir in allogeneic bone marrow transplantation: Absence of clinical cytomegalovirus infection. Br J Haematol 1991;79:57-62. Goodrich JM, Bowden RA, Fisher L, Meyers JD. Ganciclovir prophylaxis to prevent cytomegalovirus disease after allogeneic marrow transplant. Ann Intern Med 1993;118:173-178. Winston DJ, Ho WG, Bartoni K, et al. Ganciclovir prophylaxis of cytomega1ovirus infection and disease in allogeneic bone marrow transplant recipients. Results of a placebo-controIled, double-blind trial. Ann Intern Med 1993;118:179-184. Edman JC, Kovac JA, Masur H, Santi DV, Elwood HJ, Sogin ML. Ribosomal RNA sequence shows Pneumocystis carinii to be a member of the fungi. Nature 1988;334:519-522. Jackson HC, Colthurst D, Hancock V, Marriott MS, Tuite MF. No detection of characteristic fungal protein elongation factor EF-3 in Pneumocystis carinii. J Infect Dis 1991;163:675-677. Singer C, Armstrong D, Rosen PP, Schottenfeld D. Pneumocystis cnrinii pneumonia: A cluster of eleven cases. Ann Intern Med 1975;82:772-777. Chave JP, David S, Wauters JP, van Melle G , Francioli P. Transmission of Pneumocystis carinii from AIDS patients to other immunosuppressed patients: A cluster of Pneumocysfis carinii pneumonia in renal transplant recipients. AIDS 1991;5:927-932. Sepkowitz KA. Pneunzocystis carinii pneumonia among patients with neoplastic diseases. Semin Respir Infect 1992;7:114-121. Israelski DM, Remington JS. Toxoplasmosis in patients with cancer. Clin Infect Dis 1993;17(Suppl. 2):S423-S435. Foot AB, Garin YJ, Ribaud P, Devergie A, Derouin F, Gluckman E. Prophylaxis of toxoplasmosis infection with pyrimethamine/sulfadoxine (Fansidar) in bone marrow transplant recipients. Bone Marrow Transplant 1994;14:241-245. Slavin MA, Meyers JD, Remington JS, Hackman RC. Toxoplnsrna gondii infection in marrow transplant recipients: A 20 year experience. Bone Marrow Transplant 1994;13:549-557.

10. Pharmacologic considerations with antimicrobials used in oncology Michael Postelnick and Sara R. Halbur

1. Introduction Due to the severely immunocompromised state that occurs in a number of oncology patients (with or without chemotherapy), antimicrobial therapy is often indicated. Therapy may be administered in response to a given pathogen at a specific site of infection or oftentimes is administered empirically. Multiple agents are frequently indicated to provide the necessary activity against likely microbial pathogens. This chapter addresses some of the pharmacologic considerations that are involved in the use of these agents. The antimicrobial agents that are commonly used in patients with neoplastic diseases are discussed. The areas of focus are the pharmacokinetic and pharmacodynamic characteristics of these agents as well as the more common adverse events associated with their use in this patient population and the clinically significant drug interactions that may be encountered.

@-lactamantibiotics are the mainstay of antibacterial therapy for oncology patients. These agents have established themselves as the primary antimicrobials used in the treatment of febrile neutropenic patients. In this setting, agents with significant antipseudomonal activity, such as piperacillin or ceftazidime, are usually chosen. Other 6-lactam antimicrobials, such as the carbapenems, imipenem, and meropenen, or the monobactam, aztreonam, have also been studied and have achieved significant usage in this patient population. 2.1 Pharmacokinetics

2.1.1 Ceftazidime. Single-dose studies of ceftazidime indicate that a 2-g dose administered over 20-30 minutes in healthy adults achieves peak serum concentrations of 160-185rnglL with concentrations 6 hours later of approximately 8mgiL. Ceftazidime is widely distributed into body fluids and tissues, Gary A. N o ~ k i n(ell), IMANAGEIMENT O F INFECTIOUS C O M P L I C A T I O N S IN C A N C E R PATIENTS. 0 1998. Klii~verAcademic P~~blishers, Boxtor?.All right? reserved.

and achieves concentrations in the central nervous system adequate to treat meningitis. The average ceftazidime volume of distribution at steady state ranges from 0.18 to 0.31Llkg. Ceftazidime is not metabolized and is excreted unchanged in the urine. Ceftazidime clearance parallels creatinine clearance, with the average ceftazidime clearance ranging from 98 to 122mLlmin in healthy adults. The elimination half-life in patients with normal renal function is approximately 2 hours. The elimination half-life in anephric patients is approximately 35 hours. Ceftazidime is readily removed by both hemodialysis and peritoneal dialysis. However, the elimination half-life of ceftazidime is minimally affected by hepatic dysfunction [I].

2.1.2 Piperacillin. Piperacillin exhibits nonlinear, dose-dependent pharmacokinetics. Peak concentrations and area under the concentration time curve increase more than proportionally with increases in dose. This is likely due to capacity-limited nonrenal and renal elimination of the drug. Following intravenous infusion of a single 4-g dose in healthy adults with normal renal function, peak concentrations range from approximately 150 to 300mglL and trough concentrations range from 6 to 24mglL. Piperacillin attains widespread distribution into body fluids and tissues. While only low concentrations are normally attained in the cerebrospinal fluid (CSF), high doses may achieve concentrations capable of treating more susceptible organisms in patients with inflamed meninges. Due to biliary excretion, piperacillin achieves high concentrations in the bile. In patients with normal renal and hepatic function, the elimination half-life of piperacillin is dose dependent, ranging from approximately 30 minutes for a 1-g dose to a little over an hour for a 6-g dose. The drug appears to be eliminated unchanged both by tubular secretion and glomerular filtration as well as excretion in the bile. The elimination half-life of piperacillin increases in patients with renal dysfunction, approaching 6 hours in anephric patients with normal hepatic function. In patients with both renal and hepatic dysfunction, the piperacillin elimination half-life approaches 32 hours [I]. 2.1.3 Imipenem. In adults with infections who received multiple infusions of 500-1000mg over 30-60 minutes every 6 hours, peak serum concentrations range from 19 to 67 mglL, with trough concentrations of 1-3 mglL. Imipenem achieves widespread distribution throughout the body, with a volume of distribution approximating that of extracellular fluid (approximately 25% of body weight). Imipenem does not achieve concentrations in the CSF adequate to treat meningitis. Imipenem is eliminated primarily by both glomerular filtration and tubular secretion. The serum half-life of imipenem is approximately 2 hours in patients with normal renal function and increases to 7 hours in patients with renal failure. Imipenem is administered with cilastatin to prevent renal metabolism of imipenem by the enzyme DHP 1. In patients with normal renal function the cilastatin half-life is approximately 2.5 hours, which increases to 17 hours in patients with renal failure. Accumulation of cilastatin

in renal failure is thought to contribute to the CNS side effects attributed to this drug combination in patients with renal failure [I].

2.1.4 Aztreonam. After a single 2-g dose of aztreonam in healthy adults, peak concentrations ranged from approximately 200 to 250mg/L, with concentrations 8 hours later of approximately 8 mglL. Aztreonam is widely distributed in body tissues and fluids, including the central nervous system (particularly in patients with inflamed meninges). Aztreonam is primarily excreted unchanged both by glomerular filtration and tubular secretion. A small amount of the administered dose is metabolized and excreted in the bile. The elimination half-life in patients with normal renal and hepatic function is approximately 2 hours, which increases to about 8 hours in anephric patients. The elimination half-life is only slightly prolonged to about 3 hours in patients with alcoholic cirrhosis but normal renal function. Aztreonam is removed by both hemodialysis and peritoneal dialysis [I]. 2.2 Appropriate dosing in this patient pop~llation p-lactam antibiotics, unlike aminoglycosides and fluoroquinolones, do not exhibit concentration-dependent killing, nor do they demonstrate a postantibiotic effect. The most important factor that has been associated with successful therapy with these agents is the time the concentration of drug is maintained above the minimum inhibitory concentration (MIC) of the infecting organism. Optimal efficacy in several animal models of infection has been achieved for Enterobacteriaceae when the cephalosporin concentration has been maintained above the MIC of the organism for 6070% of the dosing interval [2]. Some investigators have used extremely short dosing intervals or continuous-infusion administration of (S-lactams in oncology patients to ensure maintenance of the antibiotic concentration above the MIC of the infecting organism for the entire dosing interval [3]. The pharmacodynamics of continuous-infusion ceftazidime dosed to maintain concentrations above the MIC of the infecting organism was shown to be equivalent to conventionally dosed ceftazidime in both an in vitro model and in seriously ill patients. The dose of ceftazidime required when administered by continuous infusion was one half of the usually administered intermittent bolus dose [3,4]. These data suggest that to achieve optimal therapy in this group of patients, the dose of f3-lactam administered should be high enough to achieve concentrations above the MIC of potential pathogens such as Pseudomonns neruginosa. In addition, dosing intervals should be short enough so that the (3lactam concentration remains above this concentration for at least 70% of the dosing interval. For drugs such as piperacillin, this will require administration of at least 3g every 4 hours to patients with normal renal function. Ceftazidime has routinely been administered every 8 hours, and a 2-g dose administered at this interval is likely to achieve these objectives in the majority of patients. The

administration of this agent at shorter dosing intervals in this patient population has been evaluated and, while pharmacodynamically appealing, has not been shown to be superior to the conventional dosing regimen. 2.3 Pharmacodynarnics The activity of p-lactam antibiotics is due to inhibition of different cytoplasmic membrane-associated enzymes responsible for the synthesis of peptidoglycan. These enzymes are known as penicillin-binding proteins. In gram-positive organisms these sites are relatively accessible; however, in gram-negative bacteria reaching these sites requires penetration of the outer cell membrane and traversing the periplasmic space. This provides gram-negative organisms with a number of potential mechanisms of resistance that are not available to grampositive organisms [S]. Of particular concern is the ability of organisms to produce (3-lactamaseenzymes capable of inactivating the administered antimicrobial. While most of the (3-lactamsused in the treatment of oncology patients are relatively resistant to the effects of these enzymes, exposure of organisms to some of these agents (particularly third-generation cephalosporins and imipenem) induces the production of (3-lactamase. Concentrations of these enzymes in the periplasmic space may reach such high levels that even the activity of these agents that are normally resistant to the effects of (3-lactamase is compromised. 2.4 Synergy with other antimicrobial agents p-lactam antimicrobials have been shown to be reliably synergistic when combined with aminoglycosides in the treatment of gram-negative infections, including Pseudomonas aeruginosa [6]. This is the basis for their use combined with aminoglycosides in the regimens most commonly used for the empiric treatment of febrile neutropenic patients. Due to the potential nephrotoxicity associated with aminoglycosides, a number of investigators have examined the combination of two (3-lactam antimicrobial agents as a replacement for aminoglycoside-containing regimens for the empiric treatment of the febrile neutropenic patient [7]. However, the ability of two 13-lactam antimicrobials to exert a synergistic effect against gramnegative pathogens remains an area of some controversy. There are two potential mechanisms by which (3-lactams can exert a synergistic effect when administered in combination with other (3-lactams. The first potential mechanism is predicated on the administered agents having primary affinity for different penicillin binding proteins. Most studies have been conducted with the amino-penicillin, mecillinam, which has a high affinity for penicillin binding protein 2. In contrast, (3-lactams such as aztreonam and ceftazidime bind preferentially to penicillin binding protein 3 [8]. Synergy with mecillinam and other fl-lactam antibiotics has been demonstrated in a number of in vitro and in vivo models [9-121. These data lack relevance, however,

because this agent is not in clinical use. The second potential mechanism of synergy entails inhibition of (3-lactamase. This mechanism has been exploited in combination (3-lactaml(3-lactamaseinhibitors, such as ampicillinlsulbactam and piperacillinltazobactam. Sulbactam and tazobactam, which are p-lactam antibiotics with weak affinity for penicillin-binding proteins, irreversibly bind to (3-lactamase, allowing the (3-lactam agent with which they are combined unimpeded access to the penicillin-binding proteins [12]. Despite the lack of a theoretical basis for synergy, a number of (3-lactams have been combined in vitro and studied for synergistic interactions. As would be expected, the vast majority of these combinations was determined to be either additive or indifferent. This contrasts with data for aminoglycosides and fluoroquinolones, which suggest that synergy occurs in a majority of situations [13-151. In combinations where synergism was demonstrated, high concentrations of antibiotic were usually necessary to achieve this effect [16-191. Lack of synergy has also been demonstrated in an in vivo model of Pseudomonas aeruginosa infection [20]. Antagonism when a combination of two (3-lactamsis used concurrently may occur as a result of two possible mechanisms. The first mechanism involves penicillin binding proteins. Concomitant inhibition of different penicillin binding proteins may interfere with bactericidal effect or penetration of the more bactericidal agent through the porin. Another involves the production of (3lactamase; agents that induce the hyperproduction of (3-lactamase, such as cefoxitin, will induce inactivation of the concurrently administered (3-lactam antimicrobial [21,22]. The principal mechanism that has been reported results from the induction of hyperproduction of p-lactamase. Cefoxitin has been the agent that has been most often implicated; however, third-generation cephalosporins and imipenem have also been shown to induce (3-lactamase production [22]. In summary, (3-lactam antibiotics exhibit synergistic interactions in the majority of situations when combined therapeutically with aminoglycoside or fluoroquinolone antimicrobiak. The combination of two (3-lactam agents, while at times synergistic, is most often additive or indifferent. When two (3lactam antibiotics are combined therapeutically, there is a potential for an antagonistic interaction, especially if one of the (3-lactams is likely to induce the production of (3-lactamase by the infecting organism.

2.5 Potential for adverse effects The most common potentially serious adverse effects that occur with (3lactam antibiotics are hypersensitivity reactions. Manifestations of allergy to penicillins in order of decreasing frequency are maculopapular rash, urticaria1 rash, fever, bronchospasm, vasculitis, serum sickness, exfoliative dermatitis, Stevens-Johnson syndrome, and anaphylaxis. The incidence of such reactions has been reported to range from 0.7% to 10%. The incidence of anaphylactic or anaphylactoid reactions ranges from 0.004% to 0.04% [6].

Because of the structural similarity between the penicillins and the cephalosporins, cross-allergenicity has been of concern. Early iinmunologic studies indicated a potential crossreactivity of approximately 20% [23]. Clinical data, however, suggest a much lower incidence. Anne and Reisman reviewed published reports and manufacturer's postmarketing data on cephalosporininduced allergic reactions and concluded that the incidence of cephalosporin allergy in penicillin-allergic patients receiving third-generation cephalosporins is no greater that the incidence of cephalosporin allergy in the general population [25].Other authors agree that the risk of administering a thirdgeneration cephalosporin to a penicillin-allergic patient who exhibits a nonurticaria1 reaction to penicillin is low [25].The risk of administering aztreonam to a patient who has exhibited a penicillin allergy has been demonstrated from both an immunological and clinical standpoint to be extremely low [26]. Imipenem appears to exhibit a greater degree of immunologic crossreactivity with penicillin; therefore, the risk may be greater that the risk for thirdgeneration cephalosporins. In summary, hypersensitivity reactions, although relatively uncommon, remain potentially one of the most serious adverse effects associated with (Slactam administration. Patients with anaphylactic or anaphylactoid reactions to penicillins or cephalosporins should avoid drugs of either class. However, there appears to be little risk in administering a third-generation cephalosporin to a patient who has had a non-urticaria1 allergic type reaction to a penicillin. Likewise, there appears to be little to no risk in the administration of aztreonam to patients who have a history of any type of allergic reaction to penicillin, whereas administration of imipeneln may constitute a greater risk.

2.5.1 Hematologic. Penicillins and cephalosporins have rarely been implicated in cases of drug-induced bone marrow suppression resulting in neutropenia and thrombocytopenia [27]. Charak and colleagues [28] have examined the effect of a number of common antimicrobials on bone marrow progenitor cells. These investigators found that aztreonam, ceftazidime, and imipenem caused significant suppression of human erythroid and granulocyte colony forming units in an in vitro model. This may have been reflected clinically by data suggesting that patients who received double (3-lactam therapy for neutropenic fever had a significantly longer duration of neutropenia than patients who received a IS-lactam and aminoglycosidecontaining regimen [29]. However, other investigators were unable to confirm this observation [30]. The exact effect of these agents on the granulocyte recovery rate of neutropenic patients remains to be elucidated. 2.5.2 Renal. (3-lactams antibiotics that are commonly used in oncology patients have not been reported to result in significant nephrotoxicity. Interstitial nephritis has been reported with penicillins; however, methicillin has been the agent that is most frequently implicated. First-generation cephalosporins, such as cephaloridine and cephalothin, have been implicated in renal toxicity and

acute tubular necrosis; however, this has not been the case for the thirdgeneration cephalosporins.

2.5.3 Central nervous system. (3-lactam antibiotics have been reported to be associated with seizures. Seizures are rare with agents such as piperacillin, ceftazidime, and aztreonam [4]. However, seizures are a significant concern when imipenem is administered, especially to patients with impaired renal function. In addition, when imipenem is administered at the high end of the dosing range (approximately 4gsJd) to oncology patients, the incidence of seizures has been reported to be approximately 10%. In contrast, the seizure rate for patients who received 2gsJday of imipenem was 2.2%. This compares with no seizures having occurred in trials for patients who received ceftazidime [311. 2.5.4 Electrolyte disturbances. Antipseudomonal penicillins have been associated with delivering significant sodium loads to patients. The sodium content of piperacillin is relatively low, with each gram of piperacillin containing 1.85mEq of sodium. Nevertlieless, it is suggested that serum electrolytes, particularly serum potassium, be monitored in patients receiving piperacillin who may potentially have low potassium reserves. This particularly may occur in patients receiving cytotoxic chemotherapy. 2.6 Drug interactions

Because (3-lactams are commonly administered concurrently with aminoglycosides in oncology patients and aminoglycoside serum concentrations are routinely monitored, the ability of these agents to inactivate aminoglycosides and cause falsely depressed aminoglycoside levels is of some importance.

3. Quinolones 3.1 Phnvnzncokinetics Fluoroquinolones exhibit a favorable pharmacokinetic profile. Oral administration provides serum concentrations adequate to inhibit a number of serious bacterial pathogens. Bioavailability ranges from 50% to 100% in healthy, fasting adults [I]. However, the oral bioavailability in patients following the administration of cytotoxic chemotherapy has not been systematically studied. Due to decreased bioavailability caused by the dysregulation in motility that occurs in these patients, intravenous therapy in the acute postchemotherapy period is often warranted [32]. It is important to ensure that adequate doses are used when these agents are administered intravenously. Early studies of ciprofloxacin for the treatment of

neutropenic fever yielded less than favorable results, likely due to the low intravenous dose used (200mg every 12 hours) [33]. The peak concentration following the administration of a 400-mg intravenous dose of ciprofloxacin is similar to the peak concentration of a 750-mg oral dose, whereas the area under the curve (AUC) of 400mg every 12 hours intravenously is similar to the AUC of 500mg every 12 hours administered orally. An intravenous dose of 400mg every 8 hours produces an AUC similar to what is achieved with a 750mg every 12 hour oral dose [33-361, In most situations when ciprofloxacin is administered intravenously to febrile neutropenic patients, a dose of 400mg every 8 hours should be used.

Currently available ffuoroquinolones are active in vitro against most gramnegative aerobic bacteria and methicillin-susceptible staphylococci. These agents are relatively inactive against methicillin-resistant staphylococci, streptococci, enterococci, and obligate anaerobic organisms. Although the newer agents, levoff oxacin and sparfloxacin, have improved activity against staphylococci and streptococci, their role in the management of patients with cancer remains to be defined. Quinolones exhibit their effect by inhibition of DNA gyrase. Gould and colleagues [37] examined the pharmacodynamics of ciprofloxacin as they relate to bacterial killing, postantibiotic effect, and adaptive resistance. What these investigators found was that bacterial kill rate was related to the concentration of ciprofloxacin, with no reduction of killing rate at high concentrations. The duration of the postantibiotic effect was concentration dependent, with adaptive resistance (decreased susceptibility of the organism to the antimicrobial in the postantibiotic period) being evident. This is similar to the properties seen with aminoglycoside antibitoics [38]. There is a substantial body of literature examining the likelihood of a synergistic interaction occurring when a quinolone (ciprofloxacin is the quinolone that has been most often studied) is combined with various other antimicrobials against Pseudomonns aeruginosn. The vast majority of these trials indicate that when ciprofloxacin is combined with a p-lactam (uridopenicillin or a third-generation cephalosporin) the likelihood of a synergistic interaction taking place is high. Conversely, the combination of ciprofloxacin with various aminoglycosides has demonstrated a low probability of synergy, with rare instances of antagonism being demonstrated [39-5 I]. There is limited and somewhat conflicting data available examining the likelihood of synergistic activity resulting when quinolones are combined with other antimicrobials against organisms other than Pseudomonas aeruginosa. Guerillot and colleagues [52] examined the combination of ceftibuten with either the aminoglycosides or ciprofloxacin against strains of E. coli and Klebsielln pneumonine. These authors were unable to demonstrate synergy

with the combination of the cephalosporin and ciprofloxacin. Unal and colleagues [53] examined the combination of vancomycin and ciprofloxacin to determine the extent of synergy against enterococcal isolates. These authors found synergy in six Enterococcus faeciurn strains that were resistant to both vancomycin and ciprofloxacin. However, synergy occurred at such high concentrations that the clinical relevance of this interaction was unlikely. Huovinen and colleagues 1.541 found synergism occurring in approximately one third of clinical isolates tested (primarily E. coli, staphylococci, and enterococci) with the combination of trimethoprim and ciprofloxacin. Against anaerobes, ciprofloxacin has demonstrated a varied ability to produce synergy when combined with antimicrobials with significant anaerobic activity, such as metronidazole, clindamycin, cefoxitin, ceftizoxime, and mezlocillin. Synergy in approximately 40% of the isolates tested was the maximum rate that was determined for these isolates [55,56]. As with all antimicrobials, a number of organisms have demonstrated the ability to develop resistance to the fluoroquinolones. This has been primarily evident with organisms that have relatively high minimal inhibitory concentrations (MICs) to these agents, including methicillin-resistant Staphylococcus aureus, Pseudomonas aeruginosa, streptococci, and enterococci [57]

3.3 Potential for adverse effects The fluoroquinolones are extremely well tolerated, and most serious adverse reactions are rare. Initial animal studies suggested that fluoroquinolones had the potential to cause central nervous system (CNS) excitation, which could result in seizures. This risk is variable and dependent on the ability of the fluoroquinolone to penetrate into the CNS. Of the two most commonly administered agents, the reported incidence of seizures with ofloxacin is greater than that with ciprofloxacin [58,59]. Indirectly, the ability of various fluoroquinolones to inhibit the metabolism of methylxanthine compounds, such as theophylline and caffeine, may result in a greater degree of CNS toxicity that the compounds themselves [60]. Gatrointestinal toxicity is usually mild and dose related. This is most commonly manifested by nauesa, vomiting, or asymptomatic elevations in liver function tests. A number of the more recently marketed fluoroquinolones are associated with a significant incidence of photosensitivity reactions. Lomeffoxacin is associated with a 2.4% incidence of photosensitivity reactions [61]. Sparfloxacin is associated with an 8% incidence of skin reactions. Patients need to be cautioned to avoid exposure to sunlight throughout the course of therapy and as long as 5 days after completion of a course of sparfloxacin therapy. Sunscreens are not protective in these situations because the trigger for these reactions is the UVA band not the UVB band [62]. Other potentially serious adverse reactions have been reported with sparfloxacin that can increase the QTc interval. Therefore, its use is contraindicated in patients with a prolonged QTc interval. Sparfloxacin is also

contraindicated in patients receiving medications that predispose to the development of torsade de pointes, such as class Ia antiarrhythmic agents (e.g., quinidine, procainamide), class I11 antiarrhythmic agents (e.g., sotalol), bepridil, terfenadine, azithromycin, erythromycin, cisapride, pentamidine, tricyclic antidepressants, and phenothiazines [63]. 3.4 Drug interactions

The administration of divalent and trivalent cations concurrently with fluoroquinolones significantly decreases their bioavailability and has been associated with clinical failures resulting from this interaction. Sources of these cations include antacids, sucralfate, as well as medications that require buffers to aid in their absorption, such as didanosine, ferrous sulfate, and multivitamins with zinc. A decrease in the quinolone bioavailability of 40-90% has been reported. Milk and yogurt have also been shown to decrease ciprofloxacin absorption by as much as 30%. This interaction can be avoided if the cation-containing medication is administered at least 2 hours before the quinolone [64-691. Although patients with cancer are frequently receiving multiple medications in addition to antibiotics, careful planning of the administration of these medications can avoid this interaction. Fluoroquinolones inhibit the metabolism of ~nethylxanthinesto varying extents. The ability of the quinolone to inhibit cytochrome P450 1A2 appears to predict this interaction. The most potent inhibitor of theophylline metabolism is enoxacin, followed by ciproffoxacin, perfloxacin, sparfloxacin, offoxacin, and norfloxacin [69]. Quinolones also inhibit caffeine metabolism, with enoxacin resulting in approximately 80% inhibition and ciprofloxacin decreasing caffeine clearance in the range of 30-45%. Adverse effects resulting from this combination have been reported with enoxacin 170-721. Although there have been case reports of fluoroquinolones causing alterations in the anticoagulation status of patients stabilized on warfarin, the effect of concurrent fluoroquinolone administration on the elimination of the more pharmacologically active S-isomer of warfarin appears to be minimal. One would expect little problem in administering fluoroquinolones to patients concurrently receiving warfarin [69]. However, because there appears to be a subset of patients who appear sensitive to the ability of the quinolones to inhibit the less active R-isomer of warfarin, close monitoring of patients on concurrent therapy is required. The likelihood of significant drug interactions occurring in oncology patients receiving fluoroquinolone antimicrobials is minimal. When these agents are being administered orally, care must be taken to ensure that if the patients is receiving concurrent divalent or trivalent cations, appropriate separation of the cation dose from the fluoroquinolone occurs. Patients who are maintained on theophylline should have serum concentrations of theophylline monitored when fluoroquinolone therapy is initiated. Due to its narrow therapeutic

window, patients receiving warfarin should have their anticoagulation status monitored closely.

3.5 Summary Quinolone antimicrobials, with their unique pharmacodynamic profile and low toxicity, offer a number of advantages in the treatment of infections in oncology patients. Concerns remain, however, regarding the rapid development of resistance in organisms with relatively low-level susceptibility to these agents, such as staphylococci and Pse~idornonasaeruginosa, as well as the lack of large-scale clinical trials using these agents for treatment of infection in this group of patients.

4. Aminoglycosides

4.1 Pharmaco kinetics The relationship of serum concentrations to efficacy and toxicity with aminoglycosides, particularly in patients with cancer, has been a topic of much discussion. The traditional goal for aminoglycoside serum concentrations has been to maintain peak concentrations from four to eight times the MIC90 of the most likely gram-negative pathogens, with trough concentrations being maintained as low as possible without dropping below the MIC90 for an appreciable length of time. The basis for these recommendations has been thoroughly evaluated 1731. Currently the trend in aminoglycoside dosing has been to administer higher doses at prolonged dosing intervals. This methodology takes advantage of a number of pharmacodynamic properties associated with aminoglycosides. These include concentration-dependent killing, concentration-dependent postantibiotic effect, and minimization of the adaptive resistance that is exhibited by microorganisms after exposure to aminoglycosides [74-761. The majority of studies examining extended-interval or once-daily aminoglycoside dosing have been accomplished in patients with normal immune function. There are data, however, that support this dosing methodology in patients with neutropenia. The efficacy of extended-interval aminoglycoside dosing regimens when combined with effective b-lactam therapy has been demonstrated in animal models of neutropenic infections [77]. The largest trial of once-daily aminoglycoside dosing in febrile neutropenic patients was undertaken by the EORTC. In this trial, febrile neutropenic patients receiving conventionally dosed ceftazadime and amikacin were compared with febrile neutropenic patients receiving ceftriaxone once daily and amikacin 15mgJkg once daily. No differences in outcomes were observed between the groups [78]. At this time, although there

appears to be sufficient evidence to support the use of once-daily dosing of aminoglycosides in this patient population, most clinicians continue to utilize conventional dosing regimens.

Aminoglycoside antimicrobials provide activity against a variety of gramnegative aerobic organisms, including Pseudonzonns aerugin.osa. Activity is also provided against most methicillin-susceptible staphylococcal isolates. Methicillin-resistant staphylococci are frequently resistant to the aminoglycosides. These agents are inactive against anaerobic organisms and streptococci. Aminoglycosides are bacteriocidal and their activity is generated by inhibition of bacterial 30s ribosomes. They are reliably synergistic against gramnegative pathogens when combined in a therapeutic regimen with most (3-lactam antimicrobials. This synergistic activity has been demonstrated to result in improved clinical outcomes in patients with infections caused by serious pathogens such as Pseudomonns aeruginosn [94]. Aminoglycosides also provide synergistic activity when combined with p-lactam or glycopeptide antimicrobials against enterococci, streptococci, and staphylococci [79]. They are often used in combination for the treatment of serious, deep-seated infections such as endocarditis. Synergy with other classes of antimicrobials, such as fluoroquinolones, is less reliable [39-51,80,81]. As with all antimicrobials, bacteria have expressed a number of mechanisms of resistance to the aminoglycosides. Aminoglycoside-inactivating enzymes are the most commonly expressed mechanism of resistance [82]. Amikacin is the least likely aminoglycoside to be inactivated by these enzymes and as a result has maintained activity in a number of situations in which gentamicin or tobramycin may no longer be effective [82].

4.3 Potential for adverse effects Aminoglycosides are directly toxic to renal tubular cells. Nephrotoxicity has been reported to range from 8% to 26% of patients who are administered these agents [84]. The assessment of the true incidence of aminoglycosideinduced nephrotoxicity is complicated by the fact that cancer patients often receive a number of other concurrent nephrotoxic agents with the aminoglycoside. The clinical status of these patients may also predispose them to renal insult. Aminoglycoside nephrotoxicity usually does not occur until 5-7 days of continuous therapy and manifests itself as a non-oliguric renal failure [85]. Both cochlear and vestibular toxicity have been described with aminoglycoside therapy. Development of ototoxicity appears to be related to the magnitude and duration of patient exposure to aminoglycoside therapy. High peak concentrations have been thought to be associated with ototoxicity,

however, transiently high peaks, such as those experienced with the administration of single daily dosing, do not seem to cause an increased incidence of ototoxicity [76]. 4.4 Drug interactions

Aminoglycosides form complexes with p-lactam antimicrobial agents. This effect is most pronounced with antipseudomonal penicillins and the aminoglycosides tobramycin and gentamicin. These complexes inactivate the aminoglycoside. This interaction may result in falsely depressed aminoglycoside serum concentrations when samples are drawn and not assayed for an appreciable period of time. This interaction may also result in a shorter than expected aminoglycoside half-life in renal failure patients who are being treated with a combination of an antipseudomonal penicillin and an aminoglycoside [86]. Aminoglycosides are known to potentiate neuromuscular blocking agents, which may prolong the time to recovery following anesthesia. The mechanism of aminoglycoside-induced potentiation of neuromuscular blockade involves interference with calcium and the release of acetylcholine. It is characterized by respiratory failure and muscle weakness. The process usually resolves upon discontinuation of the aminoglycoside.

5. Vancomycin Gram-positive cocci, including coagulase-negative staphylococci, have long been recognized as common pathogens in the febrile neutropenic patient 1871. This has prompted widespread use of vancomycin in this patient population. Although there has long been good evidence that the routine use of vancomycin in intial empiric antimicrobial therapy of febrile neutropenic patients is unwarranted [88], this practice has persisted until recently. Increasing concern regarding the emergence of vancomycin-resistant enterococci has prompted the Centers for Disease Control and Prevention to issue guidelines for appropriate vancomycin use that discourage the routine use of vancomycin in this patient population [89]. In light of these recommendations, the overall use of vancomycin in this patient population is likely to decline. However, there is still a substantial role for the appropriate use of vancomycin in patients with neoplastic diseases.

5.1 Pharmaco kinetics Except in patients with Clostridiurn difJicile enterocolitis, vancomycin is minimally absorbed after oral administration. Although appreciable concentrations may be obtained in patients with pseudomembranous colitis, these concentrations are usually not high enough to prompt concerns regarding

toxicity. Vancomycin serum concentrations decline in a biphasic manner, with the distribution half-life ranging from 30 to 60 minutes and an elimination halflife in patients with normal renal function of approximately 4-8 hours. Vancomycin distributes extensively into body tissues and fluids. In patients with inflammed meninges, intravenously administered vancomycin may achieve CSF concentrations adequate to treat most gram-positive organisms. However, because adequate concentrations are not achieved in all situations, administration of an intrathecal dose of 10-20mg is often recommended. Vancomycin clearance ranges from 50% to 75% of creatinine clearance. The vancomycin elimination half-life in patients with end-stage renal disease has been reported to range from 120 to 190 hours [90].

Vancomycin inhibits cell wall synthesis by binding to peptides that contain D-alanyl-D-alanine, thus blocking the synthesis of peptidoglycan [91]. Vancomycin pharmacodynamics mimic those of p-lactam antibiotics, with little concentration-dependent killing and a very short postantibiotic effect [92]. As a result, to achieve optimal efficacy vancomycin serum concentrations should be maintained above the minimum inhibitory concentration of the infecting organism throughout the dosing interval,

5.3 Potential for adverse effects The most serious adverse effects associated with vancomycin are ototoxicity and nephrotoxicity. An extensive review of the data regarding the associatioil of vancomycin with these adverse effects has been published [93]. There is little evidence that the current formulation of vancomycin, when administered alone, has significant ototoxic or nephrotoxic potential. Synergistic nephrotoxicty when vancomycin is administered concurrently with aminoglycosides has been demonstrated [94]. The most commonly experienced adverse reaction associated with vancomycin is know as red nzan syndrome, which is characterized by erythema of the face, neck, and upper torso. This reaction generally occurs during the infusion and is mediated by histamine release. The incidence is reported to approach 80% in normal healthy volunteers when a 1-g dose is administered over 1 hour; however, in clinical practice the incidence is much lower [95]. Increasing the infusion rate to 2 hours or pretreatment with an antihistamine such as hydroxyzine decreases the incidence of this reaction [96,97]. Other adverse effects that have been associated with vancomycin therapy include skin rashes (2-6.5% incidence) and neutropenia. Bone marrow suppression appears to be associated with prolonged vancomycin administration and is immunologically mediated.

5.4 Serum concentration monitoring As with aminoglycosides, vancomycin serum concentration monitoring has become somewhat a standard of practice. The data to support this practice, however, are relatively weak. Cantu and colleagues thoroughly reviewed the literature and concluded that routine monitoring of vancomycin levels increases the cost of therapy without improving the safety or efficacy of treatment [93]. In a companion editorial, Moellering concurred with these authors' conclusion and offered suggestions for identifying the few patients for whom vancomycin serum concentration monitoring may be of value [97]. In oncology patients receiving concurrent vancomycin and aminoglycoside therapy who have alterations in renal function, measurement of a trough vancomycin concentration is warranted. Maintaining trough concentrations of vancornycin less than 10mglL in this group may result in a lower incidence of nephrotoxicity [93,97].

6. Antifungal agents

Invasive fungal infections are an important cause of both morbidity and mortality among patients with neoplastic disease. The primary subset of patients at risk include those with prolonged neutropenia as a result of intensive cytotoxic chemotherapy or ablative therapy, patients on corticosteroid therapy, or those who have undergone allogenic bone marrow transplantation [98,99]. Another risk factor for developing systemic infections in this patient population is the use of broad-spectrum antibiotics, which can deplete endogenous bacteria and ultimately lead to fungal infection [loo]. Not only are antifungal agents used for treatment, but they also are administered for prophylaxis and empiric therapy. 6.1 Amphotericin B

6.1.2 Pharmacokinetics. Amphotericin B has negligible absorption from the gastrointestinal tract, which requires the parenteral administration for therapy of systemic fungal infections [101,102]. A 30-mg intravenous dose of amphotericin B deoxycholate infused over several hours produces a peak concentration of about 1pg/mL. When the dose is increased to 50mg. average peak serum concentrations are approximately 2 pg/mL [loll. Amphotericin B cannot be given intramuscularly. Although amphotericin B has been used for decades, little information exists on its distribution, but it is believed to be multicompartmental. Low concentrations are achieved in aqueous humor, pleural, pericardial, peritoneal, and synovial fluids [101,102]. Because CSF concentrations are approximately 3% of those in serum, amphotericin B must be given intrathecally to achieve fungistatic concentrations within the CSF. Amphotericin B is heavily protein bound (90-95%), primarily to lipoproteins.

The elimination half-life in adults with normal renal function is approximately 24 hours, although with long-term administration the elimination half-life can be as long as 15 days. It has been suggested that this increased half-life may be due to the slow release of the drug from peripheral compartments [loll. In patients who are anephric, amphotericin B is poorly hemodialyzable.

6.1.3 Spectrum of activity. Amphotericin B provides activity against a wide variety of fungal infections that are commonly seen in the oncology patient. Amphotericin B has clinical activity against Aspergillus fumigatas, Candida spp., Mucoracae, Cryptococcus neoformans, Histoplasma capsulatum, Cocccidiodes immitis, Penicillium spp., Trichosporon, and Rh.izopus [1031051. 6.1.4 Mechanism of action. Amphotericin B exerts its activity by binding to sterols in the cell membranes of both fungal and human cells. In concentrations achieved clinically, amphotericin B is usually fungistatic, although it may be fungicidal in high concentrations or against very susceptible organisms. The drug is not active against bacteria or other organisms that do not contain sterols in their cell membrane. As a result of this sterol binding, membrane integrity is impaired, which leads to the loss of intracellular potassium and other cellular contents, ultimately resulting in cell death [101,102]. 6.1.5 Development of resistance. Coccidioides immitis and some species of Candida have shown acquired resistance in vitro, although there is no substantial clinical evidence of acquired resistance to amphotericin B [106,107] However, the degree of correlation between the results of sensitivity testing in vitro and clinical outcomes remains controversial. 6.1.6 Adverse effects. Amphotericin B has the propensity to cause frequent and potenitially severe adverse reactions. These may result in dose reductions or premature discontinuation of therapy. In addition, this limits the use of amphotericin B as a prophylactic agent [108]. The most common adverse reactions are infusion related and occur during or shortly after the infusion. These include headache, chills, fever, rigors, hypotension, tachypnea, nausea, and vomiting [loll. Symptoms usually begin 1-3 hours after starting the infusion and may last several hours [109]; however, with subsequent doses the incidence reactions generally decrease. Amphotericin B has been shown to stimulate prostaglandin synthesis, which is thought to account for its infusion-related effects [110]. Several agents have been used prior to administration to prevent or reduce the severity of the reactions. These include acetaminophen, corticosteroids, antiemetics, diphenhydramine, and meperidine [I11-1 131. Many clinicians administer a "test dose" of 1-5mg before beginning the full dose of therapy, although the manufacturer has not made this recommendation. Despite the widespead use of this agent, the optimal infusion time has not been established. In an effort

Table I . Amphotericin B dosing guidelines for febrile neutropenic patients or with documented fungal infection The following "rapid dose" method of administering amphotericin B allows a therapeutic dose to be given within a 24-hour period. This method should be used regardless of renal function. Amphotericin dose = 0.5-1.5 mglkgld Base solution: 5% dextrose in water Concentration: 0.1 mglmL Additives to IV solution: 25 mg hydrocortisone sodium succinate 500 units heparin (for peripheral line ONLY) Administration 1. lOmL (1 mg) over 30min: assess after 30-60min for anaphylactic reaction 2. 50mL (5rng) over 2 hours 3. 100mL (10mg) over 4 hours 4. Remainder of bottle over 6-8 hours One 25-mg dose of meperidine IVP may be given for rigors Continue with same total dose per day, over 4-6 hours

Rate 25 mL/h 25 mLlh Variable

to reduce untoward events, many clinicians infuse amphotericin B over 4-6 hours. A small study comparing 4-hour infusions with 45-minute infusions determined that toxicity was greater in the rapid infusion group [112]. Infusions of 1-2 hours are commonly used in clinical practice without sequelae in patients without renal insufficiency. Table 1 details the amphotericin B infusion protocol at Northwestern Memorial Hospital. Nephrotoxicity is the major dose-limiting toxicity of amphotericin B deoxycholate and may occur in more than 80% of patients [101,109]. It may manifest as renal insufficiency, azotemia, renal tubular acidosis, or frank renal failure. It has been shown in patients that sodium "loading" with sodium chloride prior to the infusion decreases nephrotoxicity, even in the absence of water or salt deprivation [113-11.51. Administration of 1 L of saline IV on the day that amphotericin B is given is recommended for adults who are able to tolerate the sodium load. Renal tubular acidosis and electrolyte loss also occur frequently with amphotericin B administration. Hypokalemia and hypomagnesemia are the most common, although hypochloremia and hypocalcemia can develop as well [101,109]. Most patients who receive amphotericin B develop a normochromic, normocytic anemia, which is thought to be caused by inhibition of erythropoietin synthesis [110]. This condition does not usually require transfusioris and is reversible upon discontinuation of therapy. Amphotericin B has also been reported to suppress the bone marrow, resulting in leukopenia and thrombocytopenia. 6.1.7 Drug interactions. Flucytosine may be synergistic when combined with amphotericin B against Cryptococcus neoformans, Cnndida albicans, and Candida tropicalis. This synergistic activity may allow for a reduction in the

total daily dose of amphotericin B; however, there is the potential for increased bone marrow toxicity from flucytosine with concomitant amphotericin B therapy [117]. Amphotericin B can have additive nephrotoxicity when used concomitantly with many of the common antiinfectives used to treat oncology patients. These drugs include aminoglycosides, vancomycin, pentamidine, cisplatin, and cyclosporine. Intensive monitoring of renal function is required if these drugs are used concurrently and may require dosage reduction [110,118]. Amphotericin B-induced hypokalemia due to renal potassium wasting may be potentiated by concomitant use of loop diuretics, corticosteroids, ACTH, or carbonic anhydrase inhibitors. Because amphotericin B can cause hypokalemia, it may predispose patients receiving cardiac glycosides or nondepolarizing neuromuscular blockers to life-threatening conditions. Serum potassium concentrations should be monitored closely in patients receiving amphotericin B with cardiac glycosides and skeletal muscle relaxants. 6.2 Liposonzal an,d lipid form~ilationsoafarnphotericin B Three new lipid or liposomal forms of amphotericin are currently available in the United States: amphotericin B lipid complex (ABLC, Albecet), amphotericin B cholesteryl sulfate complex (ABCD, Amphotec), and amphotericin B liposome (AmBisome). When amphotericin B is encapsulated into liposomes or bound to other lipid carriers, its toxicity is greatly reduced [I19- 1211. These newer formulations have been designed to deliver more amphotericin B directly to the fungal cells with a higher therapeutic index [120]. The pharmacological properties of infused amphotericin B are drastically altered by the lipid formulations, resulting in reduced nephrotoxicity [121]. The higher clearance achieved with these formulations may be responsible for the lower toxicity of these products. All of the preparations seem to preferentially accun~ulatein organs of the reticuloendothelial system instead of the kidney [121]. Amphotericin B liposome is the only formulation that is FDA-approved for the empiric therapy in febrile, neutropenic patients. Drug interactions, as well as the possibility of long-term adverse effects, require additional investigation, especially in patients who receive prolonged therapy [121]. Lipid formulations of amphotericin B should be considered for the management of patients with invasive fungal infections who have doselimiting renal insufficiency, those patients who are intolerant of amphotericin B, and in patients with specific fungal infections that are progressive despite treatment with the conventional amphotericin B [121]. 6.3 Azoles

The earliest azoles, ketoconazole and clotrimazole, were limited in antifungal activity primarily to Cnndidn species. Clotrimazole is only available as an oral

lozenge. Due to its poor bioavailability, clotrimazole is reserved for local treatment of mild oropharyngeal or esophageal candidiasis in cancer patients. Concentrations persisting in saliva are believed to be due to clotrimazole binding to oral mucosa. Ketoconazole has a broader spectrum of activity than clotrimazole, and offers the advantage of being orally bioavailable. However, its absorption is inhibited by the concomitant use of antacids or H2 blockers and can be erratic in patients receiving chemotherapy. For use as an empiric antifungal for persistent fever during neutropenia, ketoconazole is inferior to aniphotericin B in preventing fungal disease. Thus, ketoconazole use in cancer patients is limited and will not be discussed.

6.3.1 Pharmacokinetics. The pharmacokinetics of fluconazole are similar with both the oral and intravenous administration [loll. Both are useful in the management of immunocompromised patients. Fluconazole is almost completely absorbed from the gastrointestinal tract, with a bioavailability in excess of 90% in fasting adults. Peak serum concentrations can be obtained within 12 hours following oral administration [loll. Peak serum concentrations are achieved within 5-10 days at doses within the range of 50-400mg/day, and within 2 days when a loading dose of twice the usual daily dose is given. Fluconazole is widely distributed into body tissues and fluids following either oral or IV administration. In adults with normal renal function, concentrations approach plasma levels in saliva, sputum, nail, blister, and vaginal secretions; in urine and skin, concentrations are approximately 10 times that of plasma. Following IV administration, high concentrations can also be achieved in the cornea, aqueous humor, and vitreous body [loll. Unlike other azole derivatives, fluconazole distributes well into the CSF, and achieves concentrations ranging from 50% to 94% of plasma concentrations, regardless of the degree of meningeal inflammation. Protein binding is minimal and ranges from 11% to 12%. Fluconazole does not appear to undergo first-pass metabolism. Elimination is principally via renal excretion, with 60430% of a dose excreted in the urine unchanged and 11% as metabolites. The half-life of fluconazole in patients with normal renal function is approximately 30 hours. However, in renal impairment the plasma concentrations of fluconazole are higher and the half-life is prolonged. In addition, the elimination half-life of fluconazole is inversely proportional to the patient's creatinine clearance. Hepatic impairment appears to have no effect on the elimination half-life. Finally, both hemodialysis and peritoneal dialysis effectively remove fluconazole [loll. Itraconazole is only available for oral administration. Unlike fluconazole, itraconazole has unpredictable pharmacokinetics in immunocomprornised patients, resulting in variable levels in the blood [102]. The oral bioavailability of itraconazole from the capsules is approximately 40-55 % if administered on an empty stomach, yet can be as high as 90-100% if administered with a meal. In contrast, the oral bioavailability of the oral solution (elixir) is about 55%

under fed conditions and 70% under fasting conditions. Because of the oral pharmacokinetic differences in bioavailability between the two formulations, the capsules and the solution should not be used interchangeably. Peak plasma concentrations are reached only after several days of dosing, so a loading dose is frequently used for the first 3 days [102]. Itraconazole is hepatically metabolized to its active metabolite, hydroxyitraconazole. Both itraconazole and its metabolite are extensively bound to plasma proteins. Although extensive amounts of drug are also distributed into lipophilic tissues, aqueous tissues contain negligible amounts; no detectable drug appears in CSF, and little or no intact drug appears in urine. Impaired renal function and hemodialysis do not affect plasma levels [102]. Itraconazole capsules require an acidic environment for absorption from the gastrointestinal tract. In addition, many factors impair the bioavailability of itraconazole capsules, such as intrinsic gastric achlorhydria, oral antacids, H2-receptor antagonists, and mucosal disruption due to cytotoxic chemotherapy and radiation therapy. The nausea and vomiting associated with cytotoxic chemotherapy may also impair compliance and absorption, leading to subtherapeutic serum concentrations. Thus, while the oral solution addresses some of these problems, the fact remains that the lack of an intravenous formulation further limits the utility of itraconazole in cancer [98].

6.3.2 Appropriate dosing in cancer patients. For neutropenic patients receiving antifungal prophylaxis, the recommended dosage of fluconazole is 400mg PO once daily. In these patients, therapy should begin several days prior to the onset of neutropenia, and continue for 7 days after the neutropenia resolves. Clinical studies have shown that doses of 400mg PO once daily or 200mg IV every 12 hours are more effective than placebo in preventing fungal infections in this patient population [122,123]. A lower dose of 100mg PO once daily fluconazole has also been investigated in patients undergoing bone marrow transplantation. The recommendations for this dosage regimen are to initiate therapy 7 days before transplantation and continue for 180 days posttransplant, or for the duration of immunosuppression [124]. In patients with AIDS, one study has suggested that fluconazole is an effective alternative to amphotericin B as primary treatment of meningitis caused by Cryptococc~isneoformans, although in patients with altered mental status the use amphotericin B is preferable [125]. For this infection, a dosage of 400mg PO or IV once daily until clinical response occurs, followed by 200400mg PO or IV once daily, is recommended. For the treatment of aspergillosis in immunocompromised patients who respond to amphotericin B therapy or for primary therapy, the recommended dosage of itraconazole is 200-400mg PO once daily; a loading dose of 200mg PO three times daily for the first 3 days in life-threatening cases should be considered. Only the itraconazole oral solution has been proved effective for esophageal and/or oral candidiasis. For the treatment of esophageal candidiasis, the recommended dose is lO0mg (10mL) PO once daily for a

minimum of 3 weeks, although doses up to 200mg PO once daily may be used depending on the patient's response to therapy. The dose for oropharyngeal candidiasis is 200mg (20mL) once daily for 1-2 weeks, with clinical signs and symptoms generally resolving within several days.

6.3.3 Spectrum of activity. Fluconazole provides activity against many fungi, including yeast and dermatophytes. The following organisms commonly acquired by oncology patients are generally considered susceptible to fluconazole in vitro: Candida species, including some strains of C. albicans, C. tropicalis, and Torulposis glabrata (C. glabrata), and some strains of Cryptococcus neoformans, Histoplasma capsulatum, and Coccidioides immitis [101]. Fluconazole does not appear to be effective in the prevention or treatment of aspergillosis. As with other azoles, there is no activity against mucormycosis. Itraconazole has activity against many of the same fungi as fluconazole and ketoconazole, including Candidn species and Histoplasma capsulatum, but has greater activity against Aspergillus [102]. 6.3.4 Mechanism of action. The imidazoles as a class are usually fungistatic agents. Like the other drugs in this class, both fluconazole and itraconazole exert their mechanism of action by altering the fungal cell membrane [126,127]. They accomplish this by inhibiting ergosterol synthesis by interacting with an enzyme that is required to convert lanosterol to ergosterol, which is an essential component of the membrane. This inhibition of ergosterol synthesis ultimately leads to leakage of cellular contents, caused by increased cellular permeability [126,127]. Other mechanisms of action that have been proposed include inhibition of endogenous respiration, interaction with membrane phospholipids, and inhibition of yeast transformation to mycelial forms. 6.3.5 Development of resistance. Emergence of C. albicans resistance to fluconazole has not emerged as a clinically overt problem in patients with neoplastic disease [128]. However, the introduction and nosocomial transmission of azole-resistant strains of C. albicans into an oncology unit may present future problems with breakthrough infections in the neutropenic host. The risk of resistance may increase as more HIV-infected patients are treated with cytotoxic chemotherapy for their neoplasms. 6.3.6 Adverse effects. Mild, transient elevations in aminotransferases, alkaline phosphatase, and bilirubin have been reported in 5-7% of patients taking fluconazole. These hepatic abnormalities usually return to pretreatment levels after completion of therapy. Serious hepatic reactions, which include necrosis, hepatitis, abdominal pain, and anorexia, have rarely been reported in patients receiving fluconazole therapy. Mild, transient increases in serum liver enzyme concentrations have also been reported in a limited number of patients taking

itraconazole, although the association with itraconazole administration is less clear. Both fluconazole and itraconazole should be discontinued if signs or symptoms of liver toxicity develop during therapy. Exfoliative skin disorders, such as Stevens-Johnson syndrome and diffuse rash accompanied by eosinophilia, have been reported in patients taking fluconazole who have concurrent malignancy [loll. However, a definite causative relationship between exfoliative skin disorders and fluconazole has yet to be determined. Alopecia has been reported in patients receiving fluconazole therapy, which resolved after discontinuation or reduction in the dose of fluconazole [129]. Adverse dermatological effects reported during therapy with itraconazole include rash, which occurs more frequently in immunocompromised patients. Hematologic abnormalities, such as thrombocytopenia, anemia, neutropenia, and eosinophilia, have been reported rarely in patients taking fluconazole. Thrombocytopenia and leukopenia have been reported with itraconazole administration [109]. Nausea, vomiting, diarrhea, and abdominal pain have been reported in approximately 10% of patients receiving fluconazole, although these effects are rarely severe enough to require discontinuation of the drug. Similar to fluconazole, the most common adverse effects of itraconazole are dose-related nausealvomiting, diarrhea, and abdominal discomfort [109]. 6.3.7 Drug interactions. Fluconazole has been shown to exert dosedependent effects on warfarin metabolism. Low doses appear to affect warfarin hepatic metabolism only slightly, whereas higher doses may significantly increase the prothrombin time [118]. The prothrombin time should be monitored carefully in patients receiving a concurrent warfarin and fluconazole. Concomitant administration of fluconazole and cyclosporine may result in increased plasma cyclosporine concentrations, especially in renal transplant recipients [118]. In several studies in patients undergoing allogeneic bone marrow tranplantion, administration of 100-200mg of fluconazole once daily resulted in only slight increases in plasma cyclosporine levels and were not considered clinically significant. Itraconazole has also been shown to increase cyclosporine serum concentrations. A 50% reduction in cyclosporine dosage is recommended if itraconazole is administered concurrently [118]. Plasma cyclosporine levels and renal function should be monitored carefully if either fluconazole or itraconazole is administered concomitantly with cyclosporine. Rifampin is a potent enzyme inducer and can both increase the metabolic clearance and decrease the area under the curve (AUC) for fluconazole [118]. Several case reports have reported therapeutic failures when using these two drugs concomitantly, suggesting that the fluconazole dose may have to be increased in patients also receiving rifampin in order to ensure adequate fluconazole plasma levels. Fluconazole may also alter rifabutin pharmacokinetics by significantly increasing rifabutin AUC by at least 80% [130,131]. Uveitis has been reported after low-dose rifabutin administration if

fluconazole was used concomitantly [132]. This is thought to be a direct result of the increased rifabutin plasma concentration, and it is suggested that these patients be monitored closely for this adverse reaction. Rifampin can also decrease plasma concentrations of itraconazole; increased doses of itraconazole may be required in these patients. A potential serious drug interaction exists when the imidazole antifungals and cisapride are used together; this is because cisapride is metabolized by cytochrome P450 3A4 and the imidazole agents are potential inhibitors of this isoenzyme [1181. Cardiac arrhythmias, such as QT prolongation and torsnde de pointes, have been reported with this drug combination. The combination of itraconazole or fluconazole with astemizole should also be avoided due to the association between these drug combination and cardiac arrhythmias. Similar caution should be observed during concomitant administration of the azoles with terfenadine, although fluconazole doses of 400-800mgld appear necessary for the interaction to be clinically significant. Itraconazole has also been shown to greatly inhibit the metabolism of midazolam and triazolam in healthy volunteers [133,134]. The interaction with midazolam occurred after oral administration; the significance of an interaction with itraconazole and IV midazolam is unknown but may be less of an issue due to absence of an effect on presystemic midazolam clearance [133].

7. Antivirals Patients with defects in cell-mediated immunity are at risk for severe infections caused by varicella-zoster virus (VZV), herpes simplex virus (HSV), and cytomegalovirus (CMV). Successful treatment and chemoprophylaxis against viral infections has been achieved with acyclovir and ganciclovir; however, recently there has been an increase in the number of viruses resistant to both acyclovir and ganciclovir. Famciclovir is a new oral antiviral agent. Although it is the prodrug of penciclovir, which is structurally related to ganciclovir, it is pharmacologically and microbiologically related to acyclovir [135,136]. Famciclovir has a similar spectrum of activity to acyclovir but has a longer duration of action; thus, famciclovir can be dosed three times a day, in contrast to oral acyclovir, which requires five doses per day. In spite of the fact that famciclovir has more advantageous dosing, acyclovir possesses a higher affinity for the target enzyme than famciclovir. Penciclovir is similar to acyclovir in its spectrum of activity and potency against HSV and VZV. The efficacy of famciclovir has not been studied in immunocompromised patients. 7.1 Phnrr?zacokinetics 7.1.1 Acyclovir. Following oral administration, acyclovir is poorly absorbed from the gastrointestinal tract; bioavailability is approximately 20%. Food has

little effect on absorption. Peak plasma concentrations average 0.4-0.8 pg/mL after 200-mg doses and 1.6pgImL after 800-mg doses. Peak plasma concentrations usually occur within 1.5-2 hours after administration [41]. Following intravenous dosing, peak plasma concentrations average 9.8 pg1mL after 5 mgl kg per 8 hours, and 20.7 yglmL after a 10mglkg dose per 8 hours [136]. Acyclovir distributes extensively into body tissues and fluids, including vesicular fluid, aqueous humor, saliva, and cerebrospinal fluid. In comparison with plasma, salivary concentrations are low. Acyclovir CSF concentrations are approximately 50% that of plasma [135]. Acyclovir is minimally metabolized; almost 70% of circulating drug is eliminated unchanged in the urine. Renal elimination is via glomerular filtration and tubular secretion. The half-life in patients with normal renal function is about 2.5 hours, with a range of 1.5-6 hours [136]. In patients with impaired renal function, the half-life may be as long as 19 hours. Six hours of hemodialysis will remove approximately 60% of a single 2.5mgIkg dose when given 48 hours prior to dialysis [135]. 7.1.2 Valacyclovir. Valacyclovir is the L-valyl ester of acyclovir and is almost completely metabolized to acyclovir after oral administration. Like acyclovir, absorption is not affected by administration with food. In healthy volunteers a comparison of valacyclovir with acyclovir showed the relative oral bioavailability of valacyclovir to be approximately 3-5 times greater; acyclovir bioavailability is 54% when administered as valacyclovir, but only 15-30% when administered as acyclovir itself [137]. Because of its improved oral bioavailability, valacyclovir requires less frequent dosing than acyclovir, offering an attractive alternative for prophylaxis against viral infections. Within minutes of absorption from the gastrointestinal tract, valacyclovir undergoes enzymatic metabolism to produce acyclovir. The elimination of acyclovir derived from valacyclovir is similar to that of IV or oral acyclovir. Excretion is renal and fecal, with dosage adjustment required in those patients with impaired renal function [135]. 7.2 Dosing in patients with malignancy 7.2.1 Acyclovir. For the secondary treatment and prophylaxis of herpes genitalis in immunocompromised patients, the Centers for Disease Control (CDC) recommend acyclovir 400mg PO twice daily for up to 1year [138]. The recommended dosage of acyclovir for the treatment of mucocutaneous herpes simplex infection in immunocompromised patients is 200-400mg PO five times daily for 10 days or 5 mglkg infused at a constant rate over 1hour, every 8 hours for 7-10 days. For herpes zoster in bone marrow and organ transplant recipients, patients receiving chemotherapy, and other immunocompromised patients, the recommended dose of acyclovir is 10mgIkg IV infused over 1 hour, every 8 hours for 7-10 days [139]. The dosing of acyclovir for postherpetic neuralgia prophylaxis in immunocompromised patients has also been investigated. Seventy-four patients who had acute herpes zoster were

treated with either placebo or acyclovir 800mg PO every 4 hours while awake, 5 times a day for 7 days; these patients were followed for 5 years. Only 7% of the acyclovir-treated patients developed postherpetic neuralgia at the 6-month and 5-year follow-up, compared with 37% of the placebo group at each followup [140].

7.2.2 Valacyclovir. Valacyclovir is not currently approved for use in immunosuppressed patients; however, in a controlled trial in immunocompromised adults greater than 50 years old with herpes zoster, researchers found that valacyclovir I g PO three times a day for 7 days produced more rapid resolution of zoster-associated pain and a shorter duration of postherpetic neuralgia than acyclovir [137]. For treatment of the first episode of herpes genitalis or for recurrence in immunocompetent patients, valacyclovir twice daily is as effective as acyclovir given 5 times a day. 7.3 Spectrum oafactivity

The antiviral spectrum of acyclovir is limited to herpes viruses, including herpes simplex, herpes zoster, Epstein-Barr virus, and cytomegalovirus [139]. Prophylactic administration of oral or parenteral acyclovir has shown a beneficial effect for preventing reactivation of latent HSV infections in bone marrow transplant recipients, patients undergoing radiation therapy, and other immunocompromised patients, such as those with malignancies and receiving chemotherapy. Latent infections usually reactivate following discontinuance of the drug.

7.4 Mechanism of action Acyclovir requires phosphorylation in order to be active. Intracellularly, acyclovir is converted to the monophosphate derivative by a herpesvirus thymidine kinase [136]. It is then phosphorylated to the diphosphate and triphosphate by cellular enzymes. Fully active acyclovir triphosphate then competes with the natural substrate, deoxyguanosine triphosphate, for incorporation into the viral DNA chain [139]. Once incorporated, it terminates DNA synthesis. Uninfected cells convert very little or no drug to the phosphorylated derivatives. It is postulated that mutations in the viral thymidine kinase or polymerase genes lead to acyclovir resistance [141]. Because acyclovir is only active against the actively replicating virus, it cannot eliminate the latent herpes virus genome.

7.5 Development of resistance Resistance to acyclovir in HSV has been linked to either partial production or absence of thymidine kinase, altered thymidine kinase substrate specificity, or altered viral DNA polymerase [136]. Acyclovir resistance by herpes simplex

virus has gradually progressed from in vitro testing and animal models to immunocompromised patients. Acyclovir-resistant isolates of both HSV and VZV have been reported with increasing frequency, especially in patients with AIDS; however, it is rarely a problem in patients undergoing intensive cancer chemotherapy. The development of viral mutants can occur in patients who receive repeated systemic acyclovir or in patients with prolonged neutropenia. It has been suggested that repeated treatment of recurrent viral infections with acyclovir may favor the selection of or development of drug-resistant strains [135]. 7.6 Adverse effects Acyclovir is generally well tolerated with minimal adverse reactions. The most frequent adverse reactions to IV acyclovir are inflammation at the site of infusion or extravasation [135]. These injection site reactions can be severe, and phlebitis or vesicular eruptions can occur. These effects usually occur with solutions containing more than 8mglmL. Acyclovir is potentially nephrotoxic, although this occurs much less frequently than with amphotericin B or aminoglycosides. Approximately 5-10% of patients receiving IV acyclovir will have some transient increases in serum creatinine concentrations and a reduction in creatinine clearance. These transient elevations usually resolve spontaneously or return to pretreatment levels following dosage adjustment [135]. Acyclovir nephrotoxicity appears to be caused by crystallization of the drug within the renal tubule, resulting in renal tubular obstruction. The risk of adverse renal effects during parenteral acyclovir therapy depends on the degree of hydration, urine output, and rate of acyclovir administration [135]. Patients receiving greater than 2g per day of IV acyclovir should be kept well hydrated. Headache can occur in 2% of patients receiving oral acyclovir for chronic, suppressive therapy. Vertigo, dizziness, fatigue, and confusion also have occurred rarely in patients receiving this drug. Encephalopathy has been reported during systemic acyclovir administration, which can manifest as confusion, tremor, hallucinations, seizures, or coma [135]. In previous clinical trials with valacyclovir, cases of sometimes fatal thrombotic thrombocytopenic purpuralhemolytic uremic syndrome (TTP/HUS) occurred in patients with advanced HIV disease, as well as in bone marrow and renal transplant patients [135]. 7.7 Drug interactions

Acyclovir has been reported to enhance the antiretroviral activity of zidovudine in vitro. Although acyclovir and zidovudine have been administered concomitantly in some patients with HIV infection without evidence of toxicity, patients should be monitored closely when receiving both therapies [135, 1421. Administration of probenecid with acyclovir has report-

edly increased the mean plasma half-life and AUC of acyclovir. Probenecid decreases the renal tubular secretion of acyclovir and can increase serum and CSF concentrations of acyclovir, which can potentiate its toxicity [135]. 7.8 Ganciclovir

7.8.1 Pharmacokinetics. Ganciclovir is administered intravenously, orally, and by intravitreal implantation. Following oral administration, ganciclovir is poorly absorbed from the gastrointestinal tract. Absolute bioavailability is approximately 5% under fasting conditions, but can be increased to 6-9% when administered with food [135]. When administered with a high-fat meal c0ntainin.g 46.5% fat, the AUC was increased by about 22% and the peak serum concentrations were delayed from 1.8 hours fasted to 3.0 hours after a high-fat meal. Following IV administration, distribution into body tissues and fluids is extensive, including significant intraocular penetration [143]. Ganciclovir crosses the blood-brain barrier and achieves CSF concentrations of approximately 40% (range, 24-70%) that of plasma concentrations. Protein binding of ganciclovir is roughly 1-2%. 7.8.2 Dosing in patients with malignancy. In immunocompromised patients with CMV retinitis, the initial dosage for induction treatment with ganciclovir is Smglkg IV every 12 hours for 14-21 days. This dose is associated with an improvement or stabilization in about 85% of patients; reduced viral excretion is usually evident by 1 week, and funduscopic improvement is evident at 2 weeks [136]. The maintenance dosage recommended for this population is either Smglkg IV once daily for 5 days a week or 1000mg PO three times a day with food. An alternate regimen is 500mg PO 6 timeslday administered every 3 hours with food. In a study of 159 AIDS patients, ganciclovir 500mg PO 6 timeslday was equivalent to IV ganciclovir for maintenance therapy, although progression of retinitis began earlier in the group receiving oral drug [143]. The effectiveness of prophylactic IV ganciclovir for CMV-positive patients undergoing bone marrow allograft has been confirmed in several studies. In one study ganciclovir was started on the day of engraftment [144], whereas in the other study ganciclovir administration was delayed until surveillance cultures for CMV became positive [145]. The patients in both studies were dosed at 5 mglkg IV twice daily for 7 days, and continued at a dose of 5 n~glkgIV once daily until day 100 post-transplant. In both studies, patients received high-dose acyclovir until the engraftment took place [144,145]. 7.8.3 Spectrum of activity. Ganciclovir has antiviral activity in vitro and in vivo against various Herpesviridae. To date the drug's principal use has been against cytomegalovirus infections, most commonly retinitis, colitis, and esophagitis [136].

7.8.4 Mechanism of action. Although the exact mechanism of action of ganciclovir is unknown, it appears that the drug exerts its antiviral effect on cytomegalovirus and human herpesviruses by interfering with DNA synthesis [135]. The activity is dependent on the intracellular formation of ganciclovir triphosphate. Infected cells initially convert ganciclovir to the monophosphate form, and eventually to the diphosphate and triphosphate forms by cellular kinases [136]. Ganciclovir is incorporated into both viral and cellular DNA; once incorporated, it terminates DNA synthesis. Intracellular ganciclovir triphosphate concentrations are 10-fold higher than those of acyclovir triphosphate and decline much more slowly, with an intracellular half-life exceeding 24 hours; these reasons may explain why ganciclovir provides better anti-CMV activity than acyclovir [136]. 7.8.5 Development of resistance. Resistance to ganciclovir can occur with repeated treatment, eventually leading to the production of drug-resistant strains. Several mechanisms of resistance to ganciclovir appear to exist. Proposed mechanisms include decreased phosphorylation, low or absent concentrations of virus-coded thymidine kinase, or alterations in substrate specificity of the enzyme [136]. Persistent viremia or progressive disease may also be associated with ganciclovir resistance. 7.8.6 Adverse effects. Ganciclovir can cause significant hematological toxicity, including granulocytopenia, neutropenia, and thrombocytopenia. Neutropenia occurs in up to 20-50% of patients receiving ganciclovir, and it the most common dose-limiting adverse effect of the drug [135]. Neutropenia is most commonly observed during the second week of therapy and is usually reversible within 1 week of cessation. Ganciclovir-induced neutropenia and thrombocytopenia appear to result from a dose-dependent myelotoxic effect of the drug. Less frequent adverse hematological effects of ganciclovir include anemia and eosinophilia, occurring in approximately 2% and less than 1% of patients, respectively. Careful monitoring for potential hematologic toxicity is necessary during therapy. Adverse CNS effects may occur in 15-20% of patients receiving ganciclovir, and occur with either IV or oral therapy. Symptoms include headache, dizziness, confusion, mood changes, paresthesias, and peripheral neuropathy. More serious effects, such as seizures and coma, also can occur, although very few cases of ganciclovir-induced seizures have been reported [146]. Ganciclovir is moderately nephrotoxic, resulting in slight to moderate increases in serum creatinine during therapy. Impaired renal function has frequently been reported in patients receiving ganciclovir for prevention of cytomegalovirus disease following transplantation [135]. If renal function worsens during therapy, the ganciclovir dosage should be reduced because ganciclovir is extensively eliminated via glomerular filtration.

7.8.7 Drug interactions. Generalized seizures have occurred during concomit ant therapy with imipenem/cilastatin and ganciclovir [I461. Although it is well known that imipenem can cause seizures, especially in patients with renal impairment, the frequency and mechanism for this effect with this drug cornbination is unknown. Therefore, this drug combination should be avoided whenever possible. The combination of ganciclovir with zidovudine does not alter the pharmacokinetics of either drug; however, it substantially increases the risk of significant hematological toxicity. Clinicians often withhold zidovudine in AIDS patients when ganciclovir is required. Regular monitoring of hematologic function should be performed if both drugs are administered together. 8. Conclusions

There are a great number of antimicrobial agents that are effective in the treatment and prevention of infection in patients with cancer. Although, in general, they are safe and effective, there are some significant toxicities of this therapy. A detailed understanding of the antimicrobial spectrum, pharmacokinetics, adverse reactions, and toxicity is required to provide optimal care for these immunocompromised patients. References 1. AHFS 96 Drug Information. Washington DC: American Society of Hospital Pharmacists. 1996, pp. 515-556. 2. Craig WA. Interrelationship between pharmacokinetics and pharmacodynamics in determining dosage regimens for broad-spectrum cephalosporins. Diagn Microsc Infect Dis 1995;22:89-96. 3. Cappelletty DM, Kang SL, Palmer SM, Rybak MJ. Pharmacodynamics of ceftazidirne administered as continuous infusion or intermittent bolus alone and in combination with single daily-dose amikacin against Pseudomonas aerzlginosa in an in vitro infection model. Antimicrob Agents Chemother 1995;39:1797-801. 4. Rolston KVI, Berkey P, Bodey GP, et al. A comparison of imipenem to ceftazidime with or without amikacin as empiric therapy in febrile neutropenic patients. Arch Intern Med 1992:152:283-291. 5. Benko AS, Cappelletty DM, Kruse JA, Rybak MJ. Continuous infusion versus intermittent administration of ceftazidime in critically ill patients with suspected gram-negative infections. Antimicrob Agents Chemother 1996;40:691-695. 6. Gutmann L, Williamson R, Kitzis M, Acar J. Synergism and antagonism in double betalactam combinations. Am J Med 1986;80:21-29. 7. Giamarellou H. Aminoglycosides plus beta-lactams against gram-negative organisms. Evaluation of in vitro synergy and chemical interactions. A m J Med 1986;80:126-137. 8. Young LS. Double beta-lactam therapy in the immunocompromised host. J Antimicrob Chemother 1985;16:4-6. 9. Kondo M, Tuchiya K. Effect of combination of cefsulodin and mecillinam. J Antibiot 1981;34:727-738.

10. Tybing L, Melchior NH. Mecillinam (FL 1060). a 6 beta amdinopenicillanic acid derivative: Bactericidal action and synergy in vitro. Antimicrob Agents Chemother 1975;8:271-276. 11. Kramer MJ, Mauriz YR, Times MD, Roberton TL. Cleeland R. Morphologic changes produced by amdinocillin alone and in combination with beta lactam antibiotics in vitro and in vivo. Am J Med 1983;75:30-40. 12. Cleeland R, Squires E. Enhanced activity of beta-lactam antibiotics with amdinocillin. Am J Med 1983;75:21-29. 13. Gerceker AA, Gurler B. In-vitro activities of various antibiotics, alone and in combination with amikacin against Pse~ldonzouascreruginosa. J Antimicrob Chernother 1995;36:707711. 14. Baltch AL, Bassey C, Hammer MC, Smith RP, Conroy JV, Michelsen PB. Synergy with cefsulodin or piperacillin and three aminoglycosides or aztreonam against aminoglycoside aerzcginosa. J Antimicrob Chemother 1991;27:801-808. resistant strains of Pse~idomotz~rs 15. Bosso JA, Saxon BA, Matsen JM. In vitro activity of aztreonam combined with tobramycin and gentamicin against clinical isolates of Pseuclornotzns neruginosa and Pse~tdonlonns cepacin from patients with cystic fibrosis. Antimicrob Agents Chemother 1987;31:14031405. 16. Moody JA, Leterson LR, Gerding DN. In vitro activities of ureido-penicillins alone and in combination with amikacin and three cephalosporin antibiotics. Antimicrob Agents Chemother 1984;26:256-259. 17. Wu DH, Baltch AL, Smith RP, Conlly PE. Effect of aztreonam in combination with azlocillin nerugiilosn. Antimicrob Agents Cheinother 1984;26:519or piperacillin on P.seudo~not~n~s 521. 18. Kurtz TO, Winston DJ, Bruckner DA, Martin WJ. Con~parativein vitro synergistic activity of new beta-lactam antimicrobial agents and amikacin against Pse~idorvlonnsaerliginosn and Sermtin ntnrcescens. Antimicrob Agents Chemother 1981;20:239-243. 19. Neu HC. Combination of ceftizoxiine with azlocillin, mezlocillin, piperacillin, and ticarcillin. J Antimicrob Chemother 1982;10:63-68. 20. Peterson, LR, Gerding DN, Moody JA, Fasching CE. Comparison of azlocillin, ceftizoxime, cefoxitin, and amikacin alone and in conlbination against Pseudomoizns neruginosn in a neutropenic site rabbit model. Antimicrob Agents Chemother 1984;25:545-552. 21. Acar JF, Sabath LD, Ruch PA. Antagonism of the antibacterial action of some penicillins by other penicillins and cephalosporins. J Clin Invest 1975;53:446-453. 22. Sanders CC. Sanders WE. Microbial resistance to newer generation beta-lactam antibiotics: Clinical and laboratory implications. J Infect Dis 1985;151:399-406. 23. Mandell GL, Petri WA. Penicillins, cephalosporins, and other beta-lactan1 antibiotics. In: Hardman JG, Limbird LE, eds. Goodman and Gilman's The Pharmacologic Basis of Therapeutics, 9th ed. New York: McGraw-Hill, 1996. 24. Levine BB. Antigenicity and cross reactivity of penicillins and cephalosporins. J Infect Dis 1973;128:S364-S366. 25. Anne S, Reisman RE. Risk of administering cephalosporin antibiotics to patients with histories of penicillin allergy. Ann Allergy Asthma Immunol 1995;74:167-170. 26. Lin RY. A persepctive on penicillin allergy. Arch Intern Med 1992;152:930-937. 27. Adelman DC. New beta-Iactam antibiotics. In: Saxon A, moderator. Immediate Hypersensitivity Reactions to Beta Lactanl Antibiotics. Ann Intern Med 1987;107:204-215. 28. Charak BS, Louie R, Malloy B, Twomey P, Mazumder A. The effect of amphotericin B, aztreonam, inlipenem and cephalosporins on the bone marrow progenitor cell activity. J Antimicrob Chemother 1991;27:95-104. 29. Winston DJ, Barnes RC, Ho WG, Young LS. Champlin RE, Gale RP. Moxalactan~plus piperacillin versus moxalactam plus amikacin in febrile granulocyctopenic patients. Am J Med 1984;77:442-450. 30. Kibbler CC, Prentice HG, Sage RJ, et al. A comparison of double beta lactam combinations with netilmicin/ureidopenicillin regimens in the empirical therapy of febrile neutropenic patients. J Antimicrob Chemother 1989;23:759-771.

31. Winston DJ, Ho WG, Bruckner D A , Champlin RE. Beta-lactam antibiotic therapy in febrile granulocytopenic patients: A randomized trial comparing cefoperazone plus piperacillin, ceftazidime plus piperacillin, and imipenem alone. Ann Intern Med 1991:115:849-859. 32. Johnson EJ. MacGowan AP, Potter MN, et al. Reduced absorption of oral ciprofloxacin after chemotherapy for haematological malignancy. J Antimicrob Chemother 1995:35:837842. 33. Meunier F, Zinner SH, Gaya H, et al. Prospective randomized evaluation of ciprofloxacin versus piperacillin plus amikacin fro empiric antibiotic therapy of febrile granulocytopenic cancer patients with lymphomas and solid tumors. Antimicrob Agents Chemother 1995;35:873-878. 34. Shah A, Lettieri J, Kaiser L, Echols R, Heller AH. Comparative pharmacokinetics and saftey of ciprofloxacin 400mg iv thrice daily versus 750mg po twice daily. J Antimicrobial Chemother 1994;33:795-801. 35. Catchpole C. Andrews JM, Woodcock J, Wise R. The comparative pharmacokinetics and tissue penetration of single-dose ciprofloxacin 400nlg iv and 750mg po. J Antimicrobial Chemother 1994;33:103-110. 36. Echols RM. The selection of appropriate dosages for intravenous ciprofloxaxin. J Antimicrobial Chemother 1993:31:783-787. 37. Gould IM, Milne K. Jason C. Concentration-dependent bacterial killing, adaptive resistance and post-antibiotic effect of ciprofloxacin alone and in combination with gentamicin. Drugs Exp Clin Res 1990;16:621-628. 38. Craig WA, Vogelman B. The postantibiotic effect. Ann Intern Med 1987;106:900-902. 39. Bustamante CI, Wharton RC, Wade JC. In vitro activity of ciprofloxacin in combination with ceftazidime, aztreonam, and azlocillin against multiresistant isolates of Pseridomonns aenlginosu. Antimicrob Agents Chen~other1990;34:1814-1815. 40. Orlando PL, Barriere SL, Hindler JA, Frost RW. Serum bactericidal activity from intravenous ciprofloxacin and azlocillin given alone and in combination to healthy subjects. Diagn Microbiol Infect Dis 1990:13:93-97. 41. Bosso JA, Saxon BA, Matsen JM. In vitro activities of combinations of aztreonam, ciprofloxacin. and ceftazidime against clinical isolates of Pselidon~onns crerrrgirlo~riand Pseudomonas cepacia from patients with cystic fibrosis. Antimicrob Agents Chemother 1990;34:487-488. 42. Eliopoulos GM, Eliopoulos CT. Ciprofloxacin in combination with other antimicrobials. Am J Med 1989:87:17S-22s. 43. Meyer RD. Liu S. In vitro synergy studies with ciprofloxacin and selected p-lactam agents and aminoglpcosjdes against multidrug-resistant Pseridonronas nrrrryinosa. Diagn Microbiol Infect Dis 1988;11:151-157. 44. Stratton CW, Franke JJ, Weeks LS, Manion FA. Comparison of the bactericidal activity of ciprofloxacin alone and in combination with selected antipseudomonal b-lactam agents against clinical isolates of Psez~tlotnonnsuerr~ginosa.Diagn Microbiol Infect Dis 1988;11:4152. 45. Fuursted K, Gerner-Smidt P. Analysis of the interaction between piperacillin and ciprofloxacin or tobramycin against thirteen strains of Psezrdor~zonasaenrginosrr, using killing curves. Acta Path01 Microbiol Immunol Scan B, Microbiol 1987;95:193-197. 46. Giamarellou H, Petrikkos G. Ciprofloxacin interactions with imipenem and amikacin against multiresistant Pse~rc1onzona.smerilginosrr. Antimicrob Agents Chemother 1987;31:959-961. 47. Bustamante CI, Drusano GL, Wharton RC, Wade JC. Synergism of the combinations of imipenem plus ciprof-loxacin and imipenem plus amikacin against Pseticlo~?zonasaerrlgino,c.a and other bacterial pathogens. Antimicrob Agents Chemother 1987:31:632-634. 48. Chin NX, Neu HC. Synergy of imipenem - a novel carbapenem, and rifampin and ciprofloxacin against Pselrdomonn~ aeruginosa, Serrntin nrnrccscens and Enterobncter species. Chemotherapy 1987;33:183-188. 49. Moody JA, Gerding DN, Peterson LR. Evaluation of ciprofloxacin's synergism with other agents by multiple in vitro methods. Am J Med 1987;82(Supyl. 4A):44-54.

50. Gerceker AA, Gurler B. In-vitro activities of various antibiotics, alone and in combination with amikacin against Pseudonzonas aeruginosa. J Antimicrob Chemother 1995;36:707711. 51. Day CA, Marceau-Day ML, Day DF. Increased susceptibility of Pseudornonas aeruginosa to ciprofloxacin in the presence of vancomycin. Antimicrob Agents Chemother 1993;37:25062508. 52. Guerillot F, Carret G, Flandrois JP. A statistical evaluation of the bactericidal effects of ceftibuten in combination with aminoglycosides and ciprofloxacin. J Antimicrob Chemother 1993;32:685-694. 53. Unal S, Flokowitsch J, Mullen DL, Preston DA, Nicas TI. In-vitro synergy and mechanism of interaction between vancomycin and ciprofloxacin against enterococcal isolates. J Antimicrob Chemother 1993;31:711-723. 54. Houvinen P, Wolfson JS, Hooper DC. Synergism of trimethoprim and ciprofloxacin in vitro against clinical bacterial isolates. Eur J Clin Microbiol Infect Dis 1992;11:255-257. 55. Esposito S, Gupta A, Thadepalli H. In vitro synergy of ciprofloxacin and three other antibiotics against Bacteroides fragilis. Drugs Exp Clin Res 1987;13:489-492. 56. Whiting JL. Cheng N, Chow AW. Interactions of ciprofloxacin with clindamycin, metronidazole, cefoxitin, cefotaxime, and mezlocillin against gram-positive and gramnegative anaerobic bacteria. Antimicrob Agents Chemother 1987;31:1379-1382. 57. Coronado VG, Edwards JR, Culver DH, Gaynes RP. Ciprofloxacin resistance among nosocomial Pseudomonas aeruginosn and Staphylococcus allreus in the United States. National Nosocomial Infections Surveillance (NNIS) System. Infect Control Hosp Epidemiol 1995;16:71-75. 58. Schwartz MT, Calvert JF. Potential neurologic toxicity related to ciprofloxacin. DICP 1990;24:138-140. 59. Traeger SM, Bonfiglio MF, Wilson JA, Martin BR, Nackes NA. Seizures associated with ofloxacin therapy. Clin Infect Dis 1995;21:1504-1506. 60. Karki SD, Bentley DW, Raghavan M. Seizure with ciprofloxacin and theophylline combined therapy. DICP 1990;24:595-596. 61. G.D. Searle & Co. Package literature for Maxaquin. February, 1995. 62. Rhone-Poulenc Rorer Pharmaceuticals Inc. Package literature for Zagam. November, 1996. 63. Domagala JM. Relationship of quinolone structure to side effects. J Antimicrob Chemother 1994;33:685-706. 64. Grasela TH, Schentag JJ, Sedman AJ, et al. Inhibition of enoxacin absorption by antacids or ranitidine. Antimicrob Agents Chemother 1989;33:615-617. 65. Nix DE, Watson WA, Lener ME, et al. Effects of aluminum and magnesium antacids and ranitidine on the absorption of ciprofloxacin. Clin Pharmacol Ther 1989;46:700-705. 66. Van Slooten AD, Nix DE, Wilton JH, et al. Combined use of ciprofloxacin and sucralfate. Ann Pharmacother 1991;25:578-582. 67. Sahai J, GallicanoK, Oliveras L, et al. Cations in the didanosine tablet reduce ciprofloxacin bioavailability. Clin Pharmacol Ther 1993;53:292-297. 68. Polk RE, Healy DP, Sahal J, et al. Effect of ferrous sulfate and multivitamins with zinc on absorption of ciprofloxacin in normal volunteers. Antimicrob Agents Chemother 1989;33: 1841-1844. 69. Gillum JG, Israel DS, Polk RE. Pharmacokinetic drug interactions with antimicrobial agents. Clin Phamacokinet 1993;25:450-482. 70. Barnett G, Segura J. De La Torre R, Carbo M. Pharmacokinetic determination of relative potency of quinolone inhibition of caffeine metabolism. Eur J Clin Pharmacol1990;39:63-69. 71. Healy DP, Polk RE, Kanawati L, et al. Interaction between ciprofloxacin and caffeine in normal volunteers. Antimicrob Agents Chemother 1989;33:474-478. 72. Harder S, Staib AH, Beer C, et al. 4-Quinolones inhibit biotransformation of caffeine. Eur J Clin Pharmacol 1988:35:651-656. 73. McCormack JP, Jewesson PJ. A critical reevaluation of the therapeutic range of aminoglycosides. Clin Infect Dis 1992;14:320-339.

74. Daikos GL, Jackson GG, Lolans VT, Livermore DL. Adaptive resistance to aminoglycoside antibiotics from first exposure down-regulation. J Infect Dis 1990;162:414-420. 75. Vogelman BS, Craig WA. Postantibiotic effects. J Antimicrob Chemother 1984;25:433-437. 76. Nicolau DP, Freeman CD, Belliveau PP, Nightingale CH, Ross JW, Quintiliani R. Experience with a once-daily aminoglycoside program administered to 2,184 adult patients. Antimicrob Agents Chemother 1995;39:650-655. 77. Kapusnik JE, Hackbarth CJ, Chambers HF, Carpenter T, Sande MA. Single, large, daily dosing versus intermittent dosing of tobramycin for treating experimental pseudomonas pneumonia. J Infect Dis 1988;158:7-12. 78. Anonymous. Efficacy and toxicity of single daily doses of amikacin and ceftriaxone versus multiple daily doses of amikacin and ceftazidime for infection in patients with cancer and granulocytopenia. The International Antimicrobial Therapy Cooperative Group of the European Organization for Research and Treatment of Cancer. Ann Intern Med 1993;119:584-593. 79. Eliopoulos GM, Eliopoulos CT. Therapy of enterococcal infections. Eur J Clin Microbiol Infect Dis 1990;9:118-126. 80. Orlando PL, Barriere SL, Hindler JA, Frost RW. Serum bactericidal activity from intravenous ciprofloxacin and azlocillin given alone and in combination to healthy subjects. Diagn Microbiol Infect Dis 1990;13:93-97. 81. Eliopoulos GM, Eliopoulos CT. Ciprofloxacin in combination with other antimicrobials. Am J Med 1989;87(Suppl.5A):17S-22s. 82. Gerding DN, Larson TA, Hughes RA, Weiler M, Shanholtzer C, Peterson LR. Aminoglycoside resistance and aminoglycoside usage: Ten years of experience in one hospital. Antimicrob Agents Chemother 1991;35:1284-1290. 83. Ristuccia AM, Cunha BA. An overview of amikacin. Therap Drug Monitor 1985;7:12-25. 84. Lietman PS, Smith CR. Aminoglycoside nephrotoxicity in humans. Rev Infect Dis 1983;5:S284-S292. 85. Ervin FR, Bullock WE, Nuttall CE. Inactivation of gentamicin by penicillins in patients with renal failure. Anitimicrob Agents Chemother 1976;9:1004-1011. 86. Halstenson CE, Wong MO, Herman CS, et al. Effect of concomitant administration of piperacillin on the dispositions of isepamicin and gentamicin in patients with end-stage renal disease. Antimicrob Agents Chemother 1992;36:1832-1836. 87. Wade JC, Schimpff SC, Newman KA, Wiernik PH. Staphylococcus epidermidis: An increasing cause of infection in patients with granulocytopenia. Ann Intern Med 1982; 97503-508. 88. Rubin M, Hathorn JW, Marshall D, Gress J, Steinberg SM, Pizzo PA. Gram-positive infections and the use of vancomycin in 550 episodes of fever and neutropenia, Ann Intern Med 1988;108:30-35. 89. Centers for Disease Control and Prevention. Recommendations for Preventing the Spread of Vancomycin Resistance Recommendations of the Hospital Infection Control Practices Advisory Committee (HICPAC). September 22, 1995, Vol. 44, No. RR-12, p. 4. 90. Matzke GR. Vancomycin. In: Evans WE, Schentag JC, Jusko W, eds. Applied Pharmacokinetics: Principles of Therapeutic Drug Monitoring, 3rd ed. Spokane, WA: Applied Therapeutics, 1992. 91. Watanakunakorn C. The antibacterial action of vancomycin. Rev Infect Dis 1981;3:S21S215. 92. McDonald PJ, Craig WA, Kunin CM. Persistent effect of antibiotics on Staphylococcus arrreus after exposure for limited periods of time. J Infect Dis 1977;135:217-223. 93. Cantu TG, Yamanaka-Yuen A, Lietman P. Serum vancomycin concentrations: Reappraisal of their clinical value. Clin Infect Dis 1994;18:533-543. 94. Cimino MA, Rotstein C, Slaughter RL, Emrich LJ. Relationship of serum antibiotic concentrations to nephrotoxicity in cancer patients receiving concurrent aminoglycoside and vancomycin therapy. Am J Med 1987;83:1091-1097. 95. Healy DP, Sahal JV, Fuller SH, Polk RE. Vancomycin-induced histamine release and "red

man syndrome": Comparison of 1- and 2-hour infusions. Antimicrob Agents Chemother 1990;34:55-5.54. 96. Sahai J, Healy DP, Garris R, Berry A, Polk RE. Influence of antihistamine pretreatment on vancomycin-induced red-man syndrome. J Infect Dis 1989:160:876-881. 97. Moellering RC. Monitoring serum vancomycin levels: Climbing the mountain because it is there? Clin Infect Dis 1994;18:544-546. 98. Walsh TJ, Lee JW. Prevention of invasive fungal infections in patients with neoplastic disease. Clin Infect Dis 1993;17(Suppl. 2):S468-S480. 99. Armstrong D. Infections in patients with neoplastic disease. In: Schlossberg D, ed. Current Therapy of Infectious Disease. St. Louis, MO: Mosby-Year Book, 1996, pp. 294-296. 100. Sable CA. Donowitz GR. Infections in bone marrow transplant patients. Clin Infect Dis 1994;18:273-284. 101. AHFS 97 Drug Information. Washington DC: American Society of Hospital Pharmacists, 1997, pp. 74-101. 102. Bennett JE. Antifungal agents. In: Hardman JG, Limbird LE, eds. Goodman and Gilman's The Pharmacologic Basis of Therapeutics, 9th ed. New York: McGraw-Hill, 1996, pp. 11751190. 103. Armstrong D. History of opportunistic infection in the immunocompromised host. Clin Infect Dis 1993;17(SuppI. 2):S318-S321. 104. Walzer P, Whimbey E. Overview of prevention of infections in the immunocompromised patient. Clin Infect Dis 1993;17(Suppl. 2):S376-S377. 105. Vartivarian SE, Anaissie EJ, Bodey GP. Emerging fungal pathogens in immunocompromised patients: Classification. diagnosis, and management. Clin Infect Dis 1993;17(S~ppl.2):S487-S491. 106. Sterling TR, Gasser RA, Ziegler A. Emergence of resistance to amphotericin B during therapy for Cnndida glcibrclta infection in an i~nmunocomprised host. Clin Infect Dis 1996;23:187-188. 107. Nguyen MH, Peacock JE. Morris AJ, et al. The changing face of candidemia: Emergence of non-Crrnrlida nlhican,c species and antifungal resistance. Am J Med 1996;100:617-623. 108. Schmitt HJ. New methods of delivery of amphotericin B. Clin Infect Dis 1993;17(Suppl. 2):S501-S506. 109. Anonymous. Systemic antifungal drugs. Medical Lett Drugs Ther 1997;39:86-88. 110. Gigliotti F, Shenap JL, Lott L, et al. Induction of prostaglandin synthesis as the mechanism responsible for the chills and fever produced by infusing amphotericin B. J Infect Dis 1987;156:784-789. 111. Burks LC. Aisner J, Fortner CL, et al. Meperidine for the treatment of shaking chills and fever. Arch Intern Med 1980;140:483-484. 112. Ellis ME, Al-Hokail AA, Clink HM, et al. Double-blind randomized study of the effect of the infusion rates on toxicity of amphotericin B. Antimicrob Agents Chemother 1992:36: 172-1 79. 113. Feely J, Heidemann H, Gerkins J, et al. Sodium depletion enhances nephrotoxicity of amphotericin B. Lancet 1981:l (S23.5):1420-1421. 114. Stein RS, Alexander JA. Sodium protects against nephrotoxicity in patients receiving amphotericin B. Am J Med Sci 1989:298:299-304. 115. Heidemann HT, Gerkins JF, Spickard WA, et al. Amphotericin B nephrotoxicity in humans decreased by salt repletion. Am J Mcd Sci 1983;75;476-481. 116. MacGregor RR, Bennett JE, Erslev AJ. Erythropoietin concentration in amphotericin Binduced anemia. Antimicrob Agents Chemother 1978;14:270-273. 117. Bennett JE, Dismukes WE, Duma RJ, et al. A comparison of amphotericin B alone and combined with flucytosine in the treatment of cryptococcal meningitis. N Engl J Med 1979;301:126-131. 118. Gillum JG, Israel DS. Polk RE. Pharmacokinectic drug interactions with antimicrobial agents. Clin Pharmacokinet 1993;25:450-482.

119. Janknegt R, de Marie S, Bakker-Woudenberg IA, et al. Liposomal and lipid formulations of amphotericin B. Clin Pharmacokinet 1992;23:279-291. 120. de Marie S, Janknegt R, Bakker-Woudenberg IA. Clinical use of liposomal and lipidcomplexed amphotericin B. J Antimicrob Chemother 1994;33:907-916. 121. Hiemenz JW, Walsh TJ. Lipid formulations of amphotericin B: Recent progress and future directions. Clin Infect Dis 1996;22(Suppl. 2):S133-S144. 122. Goodman JL, Winston DJ, Greenfield RA, et al. A controlled trial of fluconazole to prevent fungal infections in patients undergoing bone marrow transplantation. N Engl J Med 1992;326:845-85 1. 123. Winston DJ, Chandrasekar PH. Antimicrobial prophylaxis in bone marrow transplantation. Ann Intern Med 1995:123:305-315. 124. Momin F, Chandrasekar PH. Antimicrobial prophylaxis in bone marrow transplantation. Ann Intern Med 1995;123:205-215. 125. Saag MS. Powderly WG, Cloud GA, et al. Comparison of amphotericin B with fluconazole in the treatment of acute AIDS-associated cryptococcal meningitis. N Engl J Med 1992;326:83-89. 126. Como JA, Dismukes WE. Oral azole drugs as systemic antifungal therapy. N Engl J Med 1994;330:263-272. 127. Borgers M. Mechanism of action of antifungal drugs, with special reference to the imidazole derivatives. Rev Infect Dis 1980;2:520-534. 128. Koll BS, Brown AE. The changing epidemiology of infections at cancer hospitals. Clin Infect Dis 1993;17(Suppl.2):S322-S328. 129. Pappas PG, Kauffman CA, Perfect J, et al. Alopecia associated with fluconazole therapy. Ann Intern Med 1995;123:354-357. 130. Trapnell CB, Narang PK, Li R, et al. Increased plasma rifabutin levels with concomitant fluconazole therapy in HIV-infected patients. Ann Intern Med 1996:124:573-576. 131. Narang PK. Trapnell CR, Schoenfelder JR, et al. Fluconazole and enhanced effect of rifabutin prophylaxis. N Engl J Med 1994;330:316-317. 132. Havlir D, Torriani F, Dube M. Uveitis associated with rifabutin prophylaxis. Ann Intern Med 1994;121:510-512. 133. Olkkola KT, et al. Midazolam should be avoided in patients receiving the systemic antimycotics ketoconazole or itraconazole. Clin Pharmacol Ther 1994:55:481. 134. Varhe A, Olkkola KT, Neuvonen PJ. Oral triazolam is potentially hazardous to patients receiving systemic antimycotics ketoconazole or itraconazole. Clin Pharmacol Ther 1994;56:601-607. 135. AHFS 97 Drug Information. Washington DC: American Society of Hospital Pharmacists, 1997, pp. 440-484. 136. Hayden FG. Antiviral agents. In: Hardman JG, Limbird LE, eds. Goodman and Gilman's The Pharmacologic Basis of Therapeutics, 9th ed. New York: McGraw-Hill, 1996, pp. 11911223. 137. Weller S, Blum R, Doucette M, et al. Pharmacokinetics of the acyclovir pro-drug valacyclovir after escalating single- and multiple-dose administration to normal volunteers. Clin Pharmacol Ther 1993;54:595-605. 138. Centers for Disease Control and Prevention. 1993 Sexually transmitted diseases and treatment guidelines. MMWR 1993;42:1-102. 139. Whitley RJ, Gnann JW Jr. Acyclovir: A decade later. N Engl J Med 1992;327:782789. 140. McGill HI, White JE. Acyclovir and post-herpetic neuralgia and ocular involvement. Br Med J 1994;309:1124. 141. Kost RG, Hill EL, Tigges M, et al. Recurrent acyclovir-resistant genital herpes in an immunocompetent patient. N Engl J Med 1993;329:1777-1782. 142. Collier AC, Bozzette S, Coombs RW, et al. A pilot study of low-dose zidovudine in human immunodeficiency virus infection. N Engl J Med 1990;323:1015-1021.

143. The Oral Ganciclovir European and Australian Cooperative Study Group. Intravenous versus oral ganciclovir: EuropeanlAustralian comparative study of efficacy and safety in the prevention of cytomegalovirus retinitis recurrence in patients with AIDS. AIDS 1995;9:471477. 144. Goodrich JM, Bowden RA, Fisher L, et al. Ganciclovir prophylaxis to prevent cytomegalovirus after allogenic marrow transplant. Ann Intern Med 1993;118:173-178. 145. Goodrich JM, Mori M, Gleaves CA, et al. Early treatment with ganciclovir to prevent cytomegalovirus disease after allogenic bone marrow transplantation. N Engl J Med 1991;325:1601-1607. 146. Barton TL, Roush MK, Dever LL. Seizures associated with ganciclovir therapy. Pharmacotherapy 1992;12:413-415.

11. Economic impact of infections in patients with cancer David J. Shulkin and Lawrence J. Anastasi

1. Introduction The cost of caring for patients with cancer has been reported to total over $100 billion dollars per year [I]. Patients who have cancer have high utilization rates of medical care and have more medical interventions than patients with non-cancer-related diagnoses [2]. One study, that examined the costs associated with the treatment of cancer found treatment costs to be $29,708 a year [3]. Infections are a leading cause of mortality, morbidity, and cost in patients with cancer. Despite progress in the management of cancer, concern over the risk of infections in this patient population has remained high. Many advances in oncology, including intensive chemotherapy and radiotherapy regimens, immunotherapies, and bone marrow transplantation, result in immunosuppression and increase susceptibility to infection. Other factors related to the treatment of cancer can also put outpatients at risk for infection, including prolonged use of central venous catheters, administration of parenteral nutrition, use of corticosteroids, and use of broad-spectrum antibiotics. 2. Economic impact of infections in patients with cancer Studies performed in the 1980s suggested that nosocomial infections in the non-oncologic general medicine population prolong hospitalization by 3.1-4.5 days [4]. Another study from this time period suggested that 77% of charges associated with infections were due to additional hospital stays, 21 % to antibiotic treatment, and 2% to laboratory costs [5].Detailed examination of the cost of infections in oncology patients has been limited. A study from Spain examined the economic impact of neutropenia induced by myelotoxic chemotherapy in patients with cancer and found approximately $3000 in extra resources were used, in part due to the risk of infection [6]. Another study of patients with cancer who underwent surgical intervention found infected patients to be more costly than patients without infection, after severity and case Gary A. Noskin (ed), MANAGEMENT OF INFECTIOUS COMPLICATIONS IN CANCER PATIENTS. O 1998. Klirwer Academic Publisizers, Boston. All rights reserved.

mix adjustments 151. The cost of an infection was determined to add $12,542 in additional medical resources. These costs were due to a prolonged length of hospital stay (37% of costs), laboratory testing (22% of costs), radiology tests (9% of costs), pharmaceuticals (7% of costs), and the use of other hospital services (24% of costs). In addition, oncology patients who were febrile, but without documented infection, were found to have $9145 in added costs from prolonged hospitalization and additional diagnostic and therapeutic care [ 5 ] .

3. Economic aspect of the prevention of infections Recent efforts have focused on the use of cytokines, such as colony stimulating factors and myeloid growth factors, in preventing infections in patients with cancer. These biologic response modifiers have been effective in reducing the duration of neutropenia and the frequency and severity of infections following the administration of myelosuppressive chemotherapy or bone marrow transplantation [7]. Intravenous immunoglobulin has also been shown to have potential for prevention of bacterial infections in immunocompromised patients [7]. Data on the use of intravenous immunoglobulin in patients with chronic lymphocytic leukemia have shown a reduction in the number of bacterial infections but no impact on survival [8]. These reduced infections come at an additional cost of $15,470 per year. Another study of intravenous immunoglobulin in patients with chronic lymphocytic leukemia showed that the therapy is extraordinarily expensive. Intravenous immunoglobulin was found to result in an additional 0.8 day of life at an incremental cost of $13,984 per patient, or approximately $6,000,000 per additional year of quality-adjusted life [9]. Hematopoietic growth factors, such as G-CSF or GM-CSF, have also been shown to reduce infections [lo]. Filgrastim has been shown to lower infection rates by 50% after rnyelosuppressive antineoplastic chemotherapy in patients with severe chronic neutropenia [Ill. Patients treated with filgrastim were found to have a significant reduction in hospital days and total costs. Other studies have also found reduced hospital use in patients with neutropenic fever treated with colony stimulating factors [12,13].Mayordemo and associates [13] reported the mean overall treatment cost was reduced by $1300-1400 when using colony stimulating factor in patients with neutropenic fever. In another study, Lawless [14] documented that costs to insurance companies were reduced by more than 50% in patients using colony stimulating factors following autologous bone marrow transplantation. Hematopoietic growth factors have been suggested to be cost justified when the cost and duration of hospitalization increase as the proportional risk of hospitalization and medications decrease [10,15]. Preliminary cost analyses have also been performed on the prophylactic use of white blood cell transfusions [16]. Additional investiga-

tions are necessary to better elucidate the reduction in risk of infection and costs associated with this treatment. Many nosocomial infections can be prevented by using simple measures, such as adhering to proper hand washing procedures and other standard infection control practices [17]. Some other measures to prevent infections can be costly, yet if used in the appropriate clinical setting can be cost effective. For example, the use of negative-pressure airflow and laminar flow isolation rooms in patients undergoing cancer treatments is increasingly common in tertiary care medical centers. This technology is expensive to build and maintain, yet when used in populations that have a high probability of infections, can help avoid preventable infections. Patients undergoing induction chemotherapy housed in vertical airflow rooms using HEPA filters were found to have fewer febrile episodes than patients in standard rooms [18].Other infection prevention measures include the use of sterile techniques through the creation of total protective environments and the strict use of sterile gowns and other devices. Ueda and colleagues [19] reported that patients with leukemia undergoing treatment in laminar airflow rooms with strict infection control practices had reduced infections when compared with patients treated in laminar airflow rooms without strict isolation practices. Prophylactic antibiotics may also be cost effective when they are found to reduce the incidence of infections. Cimino and coworkers [20] suggested there may be a potential benefit from involving clinical pharmacists in the choice of antibiotics in the management of infections in patients with cancer. Other studies have demonstrated that the use of prophylactic antibiotics may be effective in preventing infections in patients with neutropenia who have germ cell tumors and in patients with head and neck cancer [21]. Oral antibiotic regimens may be as effective and less costly in some distinct patient populations than intravenous medications [22]. However, there is contrary evidence that suggests oral antibiotic therapy in patients with neutropenic fever may have a higher incidence of adverse reactions than parenteral regimens. In addition, one study found the use of antifungal prophylaxis with fluconazole in patients with hematologic malignancies increased costs without improved outcome [24]. A question that remains unanswered is the cost associated with treating resistant infections that occur as the result of failed prophylaxis. Further studies are needed to evaluate the cost of using these preventive measures and the benefits of prevention of infections in these high-risk patients. New tools and methods designed to prophylactically treat patients against infections will increasingly become available. Technology related to vaccines and immunostimulants will likely play an important role in future strategies designed to prevent infections in patients with cancer. A critical assessment of the costlbenefit ratio will be necessary to determine the optimal strategy.

4. Economic analyses of infections in oncology Although the number of economic analyses involving infections in oncology are limited, even fewer studies are done with appropriate methodological rigor. Much of the literature uses hospital charges, which often have little relationship to actual medical costs, Other studies lack precise definitions of infections and are not case controlled nor use well-defined outcomes. Several factors must be taken into consideration when performing rigorous economic analyses on epidemiology of infections. These analyses should include measures of direct costs, indirect costs, and intangible costs. Direct medical costs associated with infections should include the cost of diagnostic testing, such as microbiological studies, radiological examinations, and other testing procedures. Costs associated with pharmaceuticals, hospital room stays, and ambulatory care should be included as a direct cost in the economic analyses of infections. Staffing costs associated with drug preparation and administration can often exceed the acquisition cost of the drug. Other direct costs to be included in economic analyses are prolonged hospitalizations or other treatments associated with complications or adverse events that can occur during the care of a patient with infection. For example, the costs associated with renal dysfunction that may be related to the use of aminoglycosides or other potentially nephrotoxic agents should be included in the economic analyses. Complications related to surgical drainage procedures of infections may lead to prolonged hospitalizations or the use of other medical resources, and should also be included in these analyses. Indirect costs should also be accounted for when performing economic analyses. In the treatment of patients with cancer, costs can be included that are associated with a patient's lost wage while undergoing treatment and the expense associated with family members caring for patients who require supportive care. Family out-of-pocket expenses and lost wages have been reported to account for approximately one half of the total diseaserelated costs in children with cancer [3]. Intangible costs that are often associated with cancer and its complications include pain and suffering. These intangible costs can be assigned an economic value and can be reported in economic studies. There are three types of economic studies that are commonly performed. These include: (1)cost identification studies, (2) cost-effectiveness studies, and (3) cost-benefit studies. Cost-identification studies determine the actual cost of a particular course of treatment. These studies are best suited to identify the least costly method of treatment when outcomes of two treatments are considered to be equal. Cost-effectiveness studies define both costs and outcomes of an intervention. Results of cost-effectiveness studies are usually expressed in terms of a cost per improved unit of outcome. Cost-effectiveness studies are most useful in comparing one type of medical treatment with another and are not considered in isolation from other treatment options. Cost-benefit analyses report both costs and outcomes in dollars. The interven-

tion in question is likely to be a useful clinical strategy if the benefits outweigh the costs.

5. Protocols, critical pathways, and disease management Use of treatment protocols and enrollment in clinical trials are common methods of delivering care in oncology. This standardization of care lends itself well to economic evaluations and allows the clinical to make decisions regarding the most cost-efficient strategies for delivering health care. For example, using two standardized protocols in the management of patients with neutropenic fever would allow for comparisons to be made between the two and the most effective management strategy employed. The recognition that inefficient practices are contributing to rising medical costs has facilitated the development of practice guidelines. Professional groups and consortiums of oncology providers are joining together to develop guidelines for the treatment of patients with cancer. Guidelines may be useful in establishing preferred practices and in reducing variation in the delivery of clinical care. Guidelines for the diagnosis and management of patients with infection in cancer may also be useful in reducing healthcare costs. Critical pathways detail the core aspect of a clinical plan of care and may be another helpful tool for defining the most efficient treatment regimen. Early data suggest that critical pathways have been effective in helping to reduce length of stay and improving other clinical and financial outcomes [25].Critical pathways are probably most effective in reducing variations that may result in inefficiencies or suboptimal outcomes. They are also an effective means to provide consistent education to patients with similar neoplasms. Standardized approaches to the prevention and treatment of infections in patients with cancer are particularly well suited to a critical pathway approach. Disease management strategies are broader than critical pathways and take into account the most efficient and effective ways to deliver care across the full continuum of inpatient and ambulatory settings. Disease management in the oncologic populations holds significant potential in improving the cost and maximizing the outcomes of patient care. With new diagnostic and treatment strategies becoming available, it will be important for to have ready access to accurate outcomes and cost data. The development of large databases and on-lines services is increasingly making such information accessible. Systems such as those developed by the National Cancer Institute and the National Library of Medicine are sources of data that can be used for the evaluation of practice and technologies used in oncology.

6. Practice innovation in managing infections in the oncology population Medical practices are continuing to evolve and become more cost efficient, in large part due to economic pressures being imposed by managed care and

other market forces. Standard medical practices are under increased scrutiny and are now being reconsidered. For example, in chemotherapy-induced neutropenia the use of intravenous antibiotics until resolution of fever and recovery of neutrophil counts is being questioned. Recent practice innovations have included early hospital discharge in patients with neutropenia who previously would have remained in the hospital due to concern about infections [23,26,27]. These patients discontinue intravenous antibiotics 48 hours after becoming afebrile and are discharged on oral antibiotic therapy. One major cancer center estimated that outpatient treatment of a subgroup of low-risk patients with neutropenic fever who did not have any significant comorbid illness would have resulted in a cost savings of $4659 per patient [28]. Another study of chiIdren with neutropenic fever found that early discharge from the hospital resulted in a savings of $5058 per patient [29]. Other innovations in practice for patients with chemotherapy-induced neutropenia that may shorten or avoid hospitalizations altogether are being studied. Mor and colleagues [30] found that medical costs are one third lower in oncology patients cared for in a day hospital as opposed to a traditional inpatient hospital care [30]. Use of subacute facilities and home care is also being aggressively explored as alternatives to acute hospital settings for the delivery of antibiotics and other treatments for acute and chronic infections.

7. Conclusions Infection prevention and treatment remain an important part of caring for patients with cancer. Studies examining the fiscal impact of infections in this population have shown that considerable resources are consumed by the management of infections. Economic studies have been useful in determining efficient practices, but further economic studies will be essential for developing cost-effective strategies designed to prevent, diagnose, and treat infections.

References 1. Leake AR. The economic impact of cancer. Nurse Practioner Forum, No. 4 (December), 1995, pp. 207-214. 2. Munoz E, Chalfin D, Rosner F, et al. Hospital costs, cancer patients and medical diagnosisrelated groups. Onclogy 1988;45:401-404. 3. Bloom BS, Knorr, Evans AE. The epidemiology of disease expense. The costs of caring for children with cancer. JAMA 1985;253:2393-2397. 4. Haley RW, Schaberg DR. Extra charges and prolongation of stay attributable to nosocomial infections: A prospective interhospital compatisot~.Am J Med 1981;70:51-57. 5. Shulkin DJ, Kinosian 3 , Glick H, Glen-Puschett C, Daly J, Eisenberg JM. The economic impact of infections. An analysis of hospital costs and charges in surgical patients with cancer. Arch Surg 1993:128:449-452.

6. Montero MC, Valdivia ML, Carvajal E, et al. Economic study of neutropenia induced by myelotoxic chemotherapy. Pharm World Sci 1994;16187-192. 7. DeVita, Hellman, Rosenburg (eds). Cancer: Principles and Practice of Oncology, 4th ed. 1993. 8. Smith TJ, Hillner BE, Desch CE. Efficacy and cost-effectiveness of cancer treatment: Rational allocation of resources based on decision analysis. J Natl Cancer Inst 1993:85:460474. 9. Weeks JC, Tierney MR, Weinstein MC. Cost-effectiveness of prophylactic intravenous immune globulin in chronic lyrnphoctic leukemia. N Engl J Med 1991;325:81-86. 10. Lyman GH. Lyman CG, Sanderson R A , Balducci L. Decision analysis of hematopoietic grwoth factor use in patients receiving cancer chemotherapy. J Natl Cancer Inst 1993;85:488493. 11. Glaspy JA, Bleecker G. Crawford J, et al. The impact of therapy with filgrastim (recombinant granulocyte colony-stumulating factor) on the health care costs associated with cancer chemotherapy. Eur J Cancer 1993;29A(Suppl.):S23-S30. 12. Riikonen P, Saarrian UM. Makipernaa A. et al. Recombinant human granulocytemacrophage colony-stimulating factor in the treatment of febrile neutropenia: A double blind placebo-controlled study in children. J Pediatr Infect Dis 1994:13:197-202. 13. Mayordomo JI, Rivera F, Diaz-Puente MT, et al. Improving treatment of chemotherapyinduced neutropenic fever by administration of colony-stimulating factors. J Natl Cancer Inst 1995;87:803-808. 14. Lawless GD. Health resources utilization in ABMT with and without G-CSF in stage IIIIIV breast cancer patients. Oncology 1995(Suppl.): 107-1 10. 15. Morstyn G , Foote M, Lieschke GJ. Hematopoietic growth factors in cancer chernotherpay. Cancer Chernother Biol Response Modif 1996:16:295-314. 16. Rosenshein MS. Farewell VT, Prire TH, et al. The cost effectiveness therapeutic and prophylactic leukocyte transfusions. N Engl J Med 1980:302:1058-1062. 17. Guiguet M, Rekacewicz C, Leclerq B, et al. Effectiveness of simple measures to control an i ~ s infections in an intensive outbreak of nosocomial methicillin-resistant S t ~ ~ p h y l o c o c cnureus care unit. Infect Control Hosp Epidemiol 1990;11:23-26. IS. Sawamura N, Akagi E, Narakoa I, Masunaga Y, Yokoyama S, Saijo N. Significance of the bioclean room during treatment of lung cancer. Jpn J Cancer Chemother 1984;11:253-259. 19. Ueda T, Shibata H, Nakamura H. et al. Efficacy of laminar air flow room with or without clean nursing for preventing infection in patients with acute leukemia. Jpn J Clin Oncol 1983;13(Suppl. 1):151-157. 20. Cimino MA, Rotstein CM. Moser JE. Assessment of cost-effective antibiotic therapy in the management of infections in cancer patients. Ann Pharmacother 1994;28:105-111. 21. Counsel1 R, Pratt J, Williams MV. Chelnotherapy for germ cell tumors: Prophylactic ciprofloxacin reduces the incidence of neutropenic fever. Clin Oncol 1994:6:232-236. 22. Velasco E. Costa MA, Martins CA, Nucci M. Randomized trial comparing oral ciprofloxacin plus penicillin V with amikacin plus carbenicillin or ceftazidime for empirical treatment of febrile neutropenic cancer patients. Am J Clin Oncol 1995;18:429-435. 23. Rubinstein EB. Rolston K, Benjamin RS, et al. Outpatient treatment of febrile episodes in low-risk neutropenic patients with cancer. Cancer 1993;71:3640-3646. 24. Schaffner A, Schaffner M. Effect of prophylactic fluconazole on the frequency of fungal infections. amphotericin B use, and health care costs in patients undergoing intensive chemotherapy for hematologic neoplasias. J Infect Dis 1995;172:1035-1041. 25. Shulin D. Critical pathways. In: Kelley. ed. Textbook of Internal Medicine. 3rd ed. Philadelphia: Lippincott-Raven, 1997, pp. 237-241. 26. Mullen CA, Buchanan GR. Early hospital discharge of children with cancer treated for fever and neutropenia: Identification and managment of the low-risk patient. J Clin Oncol 1990:8:1998-2004. 27. Tomiak A. Yau J, Huan S, et al. Duration of intravenous antibiotic and hospital stay for patients with febrile neutropenia after chemotherapy: Experience of Ottawa Regional Cancer Center. Proc Ann Mtg Am Soc Clin Oncol 1993;12:A1500.

El-Deiry W, Morris L. High cost of hospital admissions for cancer patients at low risk to develop complications from febrile neutropenia. Proc Ann Mtg Am Soc Clin Oncol 1993;12:A1606. 29. Bash RO, Katz JA, Cash JV, et al. Safety and cost-effectiveness of early hospital discharge of lower risk children with cancer admitted for fever and neutropenia. Cancer 1994;74:189-196. 30. Mor V, Stalker MZ, Gralla R, et al. Day hospital as an alternative to inpatient care for cancer patients: A random assignment trial. J Clin Epidemiol 1988;41:771-785.

Index

Abdominal cancer, 123-131 Abscesses breast, 121-122, 123 lung, 107, 110 pyogenic liver, 126-128 N-Acetyl-cysteine, 210 Achromobacter xylosoxidans, 59 Acquired immunodeficiency syndrome (AIDS). See Human immunodeficiency virus Acremonium, 62, 176, 177 Acute disseminated candidiasis (ADC), 170 Acute leukemia, 20, 34, 38, 105,228,235 infections related to, 48-51 Acute lymphocytic leukemia (ALL), 24, 51,236-238,241 Acute lymphoma, 48-51 Acute myelocytic leukemia, 98 Acute myelogenous leukemia (AML), 24,92-93,209,225 Acute myeloid leukemia, 212 Acute nonlymphocytic leukemia, 227 Acyclovir, 23, 61, 186, 188, 189, 236, 238, 269-273 dosing with, 270-271 pharmacokinetics of, 269-270 Adenovirus, 25, 51, 155, 192 Adverse effects of acyclovir, 272 of aminoglycosides, 258-259 of amphotericin B, 262-263 of azoles, 267-268 of beta-lactams, 251-253 of ganciclovir, 274 of quinolones, 255-256 of vancornycin, 260 Aerobic bacteria gram-negative, 58-60, 254 gram-positive, 55-58

Agrobacterizim radiobacter, 59 Alimentary tract, 10-14 Alpha-hemolytic streptococci, 37, 48, 81, 109 Alternaria, 62, 176, 178,231 Alteromonas (Pseudomonas) putrefaciens, 59-60 Amantadine, 192 Amikacin, 87, 89, 90, 91,257,258 Aminoglycosides, 58, 82,89, 161, 250, 252,253,254,258,260,261,286 for febrile neutropenia, 85-88, 94, 257-258 pharmacologic considerations with, 257-259 Amoxicillin, 6 Amoxicillin/clavulanic acid. 92 Amphotericin B, 62, 83, 94; 113, 171172, 173, 174, 175, 176, 177, 178, 230,231,232,233,234-235,262, 266 liposomal and lipid formulations of, 96-97,173,264 pharmacologic considerations with, 261-264 Ampicillin, 42, 128 Ampicillin/sulbactam, 251 Anaerobic bacteria, 60 Anatomic barriers, 35, 39-40, 169 Anthracyclins, 14, 40 Antibiograms, 161 Antibiotics, 40, 42, 45-46. See also specific types for Candidn, 169 cost-effectiveness of, 285 for febrile neutropenia, 88-90, 93-94, 97 microflora of skin and, 8-9 monotherapy with, 88 protected environment and, 225-226

Antifungal agents. See also specific types for febrile neutropenia, 95-97 pharmacologic considerations with, 261-269 Antigen-presenting cells (APC), 21 Antigen tests, 149 Antimicrobial agents. See also specific types colonization resistance and, 16 pharmacologic considerations with, 247-275 Antimicrobial assays, 161 Antiviral agents, 269-275. See also specific types Aplastic anemia, 34, 36 Ara-C. See Cytarabine Aspergilh~s,14,23, 53, 62, 167, 172-173, 224,231 diagnosis of, 173 itraconazole for, 266, 267 lung cancer and, 111-113 M-CSF for, 208 microbiology laboratory evaluation of, 155 prevention of, 234 treatment of, 173 Aspergillzis flavus, 61, 227 Aspergillzls fi~nzigatus,61, 96, 111, 172173,232,262 Astemizole, 269 Azathioprine, 224 Azithromycin, 65 Azoles. 162,264-269 Aztreonarn, 16, 92,249, 250,252,253 Babesia nzicroti, 47 Bacillru, 46 Bacillus Calmette-Guerin (BCG), 131 Bacillzis ceuezis, 58 Bacitracin, 16 Bacteremia, 13-14, 40, 46. 59, 81, 82. 124-126,144-146 Bacterial irlfections in acute leukemia and lymphoma, 48-51 emerging pathogens in, 55-61 febrile neutropenia and, 81-83 in lung cancer, 107-108 prevention of, 224-230 Btrcteroides, 130, 133 Barionella, 64-65, 152 Bartonella elizabethae, 152 Bartonella henselae, 64, 152 Bnrtonella quintana, 64, 152

BCNU, 44 Beta-lactam antibiotics, 16, 58, 257, 258, 259,260 in double combination therapy, 88 for febrile neutropenia, 85-88, 89, 94, 247 pharmacologic considerations with, 247-253 Beta-lactamase inhibitors, 90, 251 Biological response modifiers, 210-213 Bipolaris, 62, 176, 178 BK virus, 193 Blastomyces, 155 Blastoschizonzyces capitatrrs, 62, 174, 175 Bleomycin, 44 Blood cultures, 144-146 Blood transfusions, 36,47-48 Bone marrow transplantation (BMT), 13, 14, 19, 23,24, 25, 34, 38, 46, 83, 183,210,212 adenovirus and, 192 BK virus and, 193 in breast cancer patients, 122 Carzdida and, 53, 169 cytomegalovirus and, 53, 186, 187, 188, 238-239,240 Epstein-Barr virus and, 190 fungal infections and, 231, 232,234235 G-CSF and, 206 hepatitis viruses and, 191-192 herpes simplex virus and, 186,271 herpesviruses and, 185,190-191,236 infections related to, 47-48, 53-54 interleukin-1 and, 209 Mycobacterium haemophilurn and, 61 Pnellntocystis carinii pneumonia and, 24 1 selective decontamination and, 228229 varicella zoster virus and, 189 Breast abscesses, 121-122, 123 Breast cancer, 118-123,229 Bronchoalveolar lavage (BAL), 117, 118,146-147,148 C~irzdida,5, 14, 53, 167, 168-172, 224, 231,262,267 azoles for, 264 breast cancer and, 122 clinical manifestations of, 170 diagnosis of, 170-171 epidemiology of, 168-170 gynecologic cancers and, 130

head and neck cancer and, 132 itraconazole for, 266-267 lung cancer and, 113-114 microbiology laboratory evaluation of, 155,162 prevention of, 232,233 treatment of, 171-172 Candida albicans, 42, 46, 61, 168, 208, 234,263,267 Candida (Tor~~lopsis) glabrata, 42, 61, 96,168,232,267 Candida krusei, 42,46, 61,96,168-169 Candida parapsilosis, 61, 168, 175 Candida rropicalis, 61, 168, 263, 267 Capnocytophagia, 60,153 Carbenicillin, 89, 93 Catheters, 59 central venous, 36,46-47, 63 indwelling, 20, 81, 83, 174, 175, 224 intravascular. See lntravascz~lar catheters intravenous, 9-10, 81,82,83 Cefazolin, 123, 133 Cefepime, 88-89 Cefoperazone, 88 Cefotaxime, 88 Cefoxitin, 251, 255 Ceftazidime, 16, 46, 87, 88, 89, 90, 93, 95, 230,252,253,257 dosing with, 249-250 pharmacokinetics of, 247-248 Ceftibuten, 254 Ceftizoxime, 255 Ceftriaxone, 87 Cellular immunity, 20-23, 35, 38-39, 190, 224 Cellulitis, 120-121, 123 Central nervous system toxicity, 253 Central nervous system tumors, 41, 50, 53 Central venous catheters, 36, 46-47, 63 Cephalexin, 123 Cephalothin, 87,89,93 Cerebrospinal fluid specimens, 149 Cervical cancer, 129 Chagas' disease, 47 Chemotherapy, 24,37,44 agents that predispose to infection, 45 for breast cancer, 122-123 intestinal tract effects of, 16-17 neutropenia associated with, 211-212, 288

oral cavity effects of, 11-14 skin effects of, 9 for solid organ tumors, 52-53 Children biological response modifiers for, 212213 TMP-SMX for leukemic, 240 vaccines for, 6, 41, 191,210, 236-238 Chloramphenicol, 65, 128 Chronic disseminated candidiasis (CDC), 170 Chronic leukemia, 38 Chronic lymphocytic leukemia (CLL), 24-25,39,49,51-52,224,284 Cidofivir, 61 Cilistatin, 127 Ciprofloxacin, 9, 55, 91, 92, 228-229, 230, 253-255,256 Cisapride, 269 Clarithromycin, 65 Clindamycin, 16, 58, 63, 92, 123, 126, 255 Clostridium dificile, 16, 42, 106, 149, 229,259 Clostridium perfringens, 125 Clostridium septicum, 16, 60, 124-126 Clostridium tertium, 60 Clotrimazole, 231, 233, 264-265 Coagulase-negative staphylococci, 9, 81, 119, 145 Coccidioides, 155 Coccidioides immitis, 262, 267 Coccidioidomycosis, 23 Colistin, 226, 227, 228 Colonization resistance, 15, 16,42-43 Colony-stimulating factors (CSF), 97, 201,202-208,211. See also specific tYPes Combination testing, 161-162 Commensal microflora. See Microflora Corticosteroids, 20, 44, 63, 110, 112, 113, 114-115,169,170 Corynebacterium, 42,46, 53, 58 Corynebacreriurn jeikeium, 58 Cost-benefit studies, 286-287 Cost-effectiveness studies, 286 Cost identification studies, 286 Co-trimoxazole. See Trimethoprim/ s ~Efnmethoxazole t Cryptococcosis, 23 Cryptococcus, 155 Cryptococcus neoformans, 175,262,263, 266,267 Cryptosporidiunz, 63-64 Cultures, 143-151

Curvularia, 62, 176, 178 Gushing's syndrome, 113,114 Cutaneous specimens, 150 Cyclophosphamide, 14,44 Cyclosporine, 224,268 Cytarabine (ara-C; cytosine arabinoside), 13, 14, 40,44,48,81 Cytokines, 20-23,97-98,201-210. See also specific types Cytomegalovirus (CMV), 25,51, 53, 61, 99,185,236,269,271,273 febrile neutropenia and, 83 immune responses to, 190 lung cancer and, 115 microbiology laboratory evaluation of, 150,155,160 prevention of, 238-240 properties and treatment of, 186-189 Cytomegalovirus (CMV) hyperimmune globulin and plasma, 188-189 Cytosine arabinoside. See Cytarabine Dapsone, 115 Daunomycin, 44 Deferoxamine, 44 Degranulation, 18 Diabetes, 125-126 Diagnostic procedures, 46 Dicloxacillin, 123 Diethyldithiocarbamate (DTC), 210 Dosing with antivirals, 270-271 with azoles, 266-267 with beta-lactams, 249-250 with ganciclovir, 273 Doxycycline, 65 Drechslera, 62 Drug interactions with acyclovir, 272-273 with aminoglycosides, 259 with amphotericin B, 263-264 with azoles, 268-269 with beta-lactams, 253 with ganciclovir, 275 with quinolones, 256-257 Drug resistance acyclovir, 61,186,189,236,271-272 aminoglycoside, 82, 161,258 amphotericin B, 262 ampicillin, 42 azole, 162 ciprofloxacin, 9 fluconazole, 42,267 ganciclovir, 61, 188, 274

methicillin. See Methicillin-resistant entries microbiology laboratory evaluation of, 160-162 multi-, 51, 55, 59 penicillin, 55-56, 58 quinolone, 38, 59 streptomycin, 42-43 tobramycin, 9 trimethoprim, 9 vancomycin, 46, 55, 58, 82, 106, 161, 162,226 Economic impact of infections, 283-288 Ecthyma gangrenosum, 80-81 Electrolyte disturbances, 253 Endocarditis, 124 Enoxacin, 256 Enterobacteriaceae, 59 Enterococcus, 61 Enterococcus faecalis, 55 Enterococcus faecium, 46, 55, 106, 255 Epidemiology, 33-65 of Candida, 168-170 emerging pathogens in, 55-65 host impairments in, 33-43 related to underlying malignancy, 48-54 treatment-associated factors in, 43-48 Epstein-Barr virus (EBV), 185, 189-190, 271 Erythromycin, 16, 65, 94 Erythropoietin, 206 Escherichia coli, 14, 16, 37, 42-43, 59, 81, 88,254,255 acute leukemia and lymphoma and, 48 bone marrow transplantation and, 53 gynecologic cancers and, 130 selective decontamination and, 230 European Organization for Research on Treatment of Cancer (EORTC), 36, 59,79, 81, 85,87, 90, 91,95, 96,257 Exophiala, 62, 176, 178 Exserohilum, 62, 176, 178 Famciclovir, 189, 238, 269 Febrile neutropenia, 77-99, 106, 202, 247,257-258 biological response modifiers for, 211212 clinical presentation of, 79-81 defined, 79 infecting pathogens in, 81-85 outpatient management of, 91-93

risk factor assessment in, 78-79 treatment strategies for, 85-97 unusual pathogens in, 84 Fluconazole, 42, 46, 62, 96, 175, 176, 231, 232-233,234,267,268-269,285 adverse effects of, 268 for Candida, 168-169, 171,267 dosing with, 266 pharmacokinetics of, 265 spectrum of activity, 267 5-Flucytosine (5-FC), 171, 176, 178, 263-264 Fluorescence stains, 159-160 Fluoroquinolones. See Q~iinolones 5-Fluorouracil (5-FU), 40,206 Folinic acid, 63 Foscarnet, 61, 186, 188,236,240 Framycetin, colistin, and nystatin (FRACON), 226 Fungal infections, 5,20, 25,40, 42 in acute leukemia and lymphoma, 51 of emerging importance, 174-176 emerging pathogens in, 61-63 febrile neutropenia and, 83, 211-212 in lung cancer, 111-114 microbiology laboratory evaluation of, 155,158,160 prevention of, 230-235 recent advances in management of, 167-178 Fusarium, 44,61-62,83,176-177,231 Fusobacterium, 133 Fuso bacterium nucleatum, 60 Ganciclovir, 61, 187-188, 238-240, 273275,274 Gastrointestinal specimens, 149-150 Genitourinary tract, 10-14 Gentamicin, 87, 89, 93,225, 258, 259 Geotrichum candidum, 62 Giemsa stains, 159 Glucocorticosteroids, 18, 19 Graft-versus-host disease, 4, 16, 23,25, 54,190,235 Gram-negative bacteria, 37-38, 42, 81, 82-83,226,228,229,230,250,258 acute leukemia and lymphoma and, 48 aerobic, 58-60,254 lung cancer and, 109 Gram-positive bacteria, 153, 230 acute leukemia and lymphoma and, 48 aerobic, 55-58 antibiotics and, 16

beta-lactams and, 250 granulocytopenia and, 37 Gram stains, 156-158, 159 Granulocyte colony-stimulating factor (G-CSF), 14,98,201,202-207,208, 212,284 Granulocyte macrophage-colony stimulating factor (GM-CSF), 14, 19, 21, 98, 201, 202,209,212,229, 284 therapeutic indications of, 203,207208 Granulocytes, 19-20 Granulocytopenia, 17-20, 33-38. See also Neutropenia defined, 33 Group B Streptococcus, 149 Gynecologic cancers, 129-131 Haemophilus infizdenzae, 6, 24, 25, 149 Haemophilus vaccines, 210 Hairy cell leukemia, 50, 52, 61 Hansenuln, 175-1 76 Hansenula anomala, 62, 174 Head and neck cancer, 132-135 Hematologic malignancies, 5-6, 83, 229, 231,233 Hematologic toxicity, 252 Hematopoietic growth factors, 98, 284285 Hepatitis A, 191 Hepatitis B, 191, 192 Hepatitis C, 54, 191 Hepatitis viruses, 191-192, 209 Hepatosplenic candidiasis, 170 Herpes simplex virus (HSV), 51, 53, 61, 83,236,238,269,271,272 microbiology laboratory evaluation of, 155,160 properties and treatment of, 185 Herpes simplex virus type 1 (HSV-I), 115-116,185,186 Herpes simplex virus type 2 (HSV-2), 185,186 Herpesviruses, 185-190,271. See also specific types immune responses to, 190-191 prevention of, 235-240 Herpes zoster, 270-273 ~ i ~ h - e f f i c i e nparticulate cy air (HEPA) filtration, 235, 285 Histamine type 2 (H2) antagonists, 8182,169,234,265,266 Histoplnsma, 155

Histoplasma capsulatum, 262,267 Histoplasmosis, 23 Hodgkin's disease, 22-23, 38, 60, 63, 78, 80,110,210,224,241 Host defense mechanisms, 223-224 Host impairments, 1-27. See also Immunocompromised hosts in cellular immunity, 20-23, 38-39 in epidemiology of infectious complications, 33-43 in humoral immunity, 23-25, 39 in integument, 6-10 in intestinal tract, 14-17 nutritional status and, 3-5 physiological and psychological status and, 5-6 in upper respiratory, alimentary, and genitourinary tracts, 10-14 HPMPC, 240 Human herpes virus type 6 (HHV-6), 185,190 Human immunodeficiency virus (HIV), 4, 58, 61, 193, 266, 267, 272, 273,275 Bartonella and, 64-65 Candida and, 169 cellular immunity and, 38 cytomegalovirus and, 186-187, 188 G-CSF and, 206 Mycobacterittm aviuin con~plexand, 60,209 Mycobacterizkm haemophilr~rnand, 61 Pneumocystis carinii pneumonia and, 63,85,114,115,156,240 Strongyloides stercoralis and, 64 Human papillomavirus, 85 Human T-cell leukemia virus-I (HTLV-I), 61, 193 Human T-cell leukemia virus-I1 (HTLV-11), 61 Humoral immunity, 23-25, 35, 39, 190 Hyalohyphomycosis, 176-177

protozoal infections in, 240-241 viral infections in, 184, 235-240 Indwelling catheters, 20, 81, 83, 174, 175, 224 Influenza virus, 155,160 Integument, 6-10 Interferon (IFN), 201,209-210 Interferon-a (IFN-a), 205, 209 Interferon-13 (IFN-f3), 209 Interferon-y (IFN-y), 21-22, 205, 209210 Interleukin (IL), 208-209 Interleukin-1 (IL-I), 18,21-22,202,204, 208-209 Interleukin-2 (IL-2), 21, 44,202,204 Interleukin-3 (IL-3), 19, 21, 202, 204, 208,209 Interleukin-4 (IL-4), 202 Interleukin-5 (IL-5), 19 Interleukin-6 (IL-6), 202, 208, 209 Interleukin-8 (IL-8), 18, 22 Interleukin-10 (IL-lo), 22 Interleukin-12 (IL-12), 21, 22, 202 Interleukin-1 receptor antagonist (IL-lra), 98, 99,208 International Antimicrobial Therapy Cooperative Group (IATCG), 59, 79, 85, 87, 90, 91, 95, 96 Intestinal tract, 14-17 Intestinal viruses, 192 Intravascular catheters, 174, 175, 224 Intravascular catheter specimens, 150151 Intravenous catheters, 9-10, 81, 82, 83 Intravenous immunoglobulin (IVIG), 25, 99,210,284 Invasive procedures, 46 Iron defiency or overload, 4 Isoniazid, 111, 131, 134 Itraconazole, 96, 113, 173, 178, 231, 232, 233-234,265-267,268,269

Imidazoles, 62, 96, 231 Imipenem, 16,59,89,127,248-249,250, 252,253,275 Immune deficiencies, 143. See also I ~ z m i ~ n o c o ~ ~ p r o ~hosts ~zised Immune globulin, 192, 193 Immune responses, 190-191 Immunization, 210 Immunocompromised hosts, 223-242. See also Host impairments bacterial infections in, 224-230 fungal infections in, 230-235

JC virus, 193 Kaposi's sarcoma, 61, 65,209 Kaposi's sarcoma-associated herpesvirus (KSHV), 61 Ketoconazole, 231-232,234,264,267 Klebsiella, 48, 53, 59, 81 Klebsiella yneumoniue, 14, 37, 88, 208, 254 Lrictobacillus, 132 Lactobacill~~s acidophilus, 130

Laminar flow isolation rooms, 285 Legionella pneumophilia, 151-1 52 Leischmaniasis, 47 Leptotrichia buccalis, 60 Leuconostoc, 56-58,82 Leukemia, 13, 55, 63, 105, 169, 185, 210, 226-227,236,240,285 acute. See Acute leukemia biological response modifiers and, 212 chronic. See Chronic leukemia febrile neutropenia and, 78 fungal infections and, 233 G-CSF and, 206 hairy cell, 50, 52, 61 in nonneutropenic patient, 51 Leukotriene B, 18 Levofloxacin, 254 Lisferia rnorzocytogerzes, 22 Lomefloxacin, 255 Lung abscesses, 107, 110 Lung cancer, 5 , 1 4 bacterial infections in, 107-108 evaluation of infection in, 117-118 fungal infections in, 111-114 infectious complications of, 106-118 mycobacterial infections in, 60, 110111 non-small cell, 107 small cell, 53, 107, 113, 202, 206 viral infections in, 115-116 Lymphoma, 34,61,105,185 acute, 48-51 Hodgkin's. See Hodgkin's disease non-Hodgkin's, 23, 61, 207 Macrophage activating factor (MAF), 38 Macrophage (monocyte)-colony stimulating factor (M-CSF), 202, 203,208,209 Malaria, 47 Malassezia, 174, 175-176 Malassezia furfur, 62, 175 Malassezia pacydermatis, 175 Mechanical obstructions, 50 Mecillinam, 250-251 Melphalan, 14 Meningitis, 43 Meningococcal vaccines, 210 6-Mercaptopurine, 40, 44 Meropenem, 16,59,90,127 Methicillin-resistant staphylococci, 9, 90 Methicillin-resistant Staphylococcus aurelcs (MRSA), 89, 91, 255

Methicillin-resistant Staphylococcus epidermidis (MRSE), 89, 91 Methicillin-susceptible staphylococci, 254,258 Methotrexate, 14,40, 44 Methylobacterium extorquens, 59 Methylxanthine, 255 Metronidazole, 16, 133,230, 255 Mezlocillin, 255 Miconazole, 17, 231 Microbial resistance. See Drug resistance Microbial susceptibility, 160-162 Microbiology laboratory, 143-163 blood cultures, 144-146 cerebrospinal fluid specimens, 149 gastrointestinal specimens, 149-150 infections of unusual or fastidious etiologies, 151-156 intravascular catheter specimens, 150151 rapid detection tests, 156-160 respiratory specimens, 146-147 urine specimens. 147-1 49 wound and cutaneous specimens, 150 Microflora, 1 changes in, 41-43 integument and, 6-10 of intestinal tract. 14-17 of upper respiratory tract and oral cavity, 11, 12 Midazolam, 269 Migration inhibitory factor (MIF), 38 Minimum inhibitory concentration (MIC) tests, 160-161 Minocycline, 65 Molds, 176-177 Monoclonal antibodies, 98-99 Mucor, 4 Mucoracae, 262 Mucormycosis, 174 Mucosal membranes, 1,4, 169 Mucositis, 12, 14, 80 Multiple myeloma, 24, 50, 52,78, 224 Mycobacteria, 60-61, 3 10-1 11,154,159 Mycobacteria other than tuberculosis (MOTT), 154 Mycobacterium aviunz-intmcellulare complex (MAC), 60,209 Mycobacteriunz chelonae, 46, 60 Mycobncterium fortuitum, 46, 60 Mycobacterium haemophilum, 61 Mycobacterium tuberculosis, 60, 154

Mycostatin, 233 Myelocytic leukemia, 206 acute, 98 Myelodysplastic syndrome, 212 Myeloid malignancies, 212 Myeloperoxidase deficiency, 5 Myonecrosis, 124-126 NafcilIin, 123 Nalidixic acid, 226-227 Negative-pressure airflow, 285 Neisseria meningitidis, 6, 149 Neomycin, 227 Nephrotoxicity, 250, 252-253, 258, 260, 263,264,272,274 Neutropenia, 105-106, 224, 228, 229, 266, 288. See also Febrile nezrtropenia; Granulocytopenia Candida and, 169 chemotherapy-associated, 211-212, 288 Fusarirrm and, 177 ganciclovir and, 274 G-CSF and, 206 GM-CSF and, 207 interleukin-1 and, 208 ketoconazole and, 232 miconazole and, 231 Non-A, non-B, non-C hepatitis, 191 Non-Hodgkin7s lymphoma, 23, 61,207 Non-small cell lung cancer, 107 Norfloxacin, 228,229-230,256 Nutritional status, 3-5 Nystatin, 225, 226, 229, 231, 232,233 Ochrobactrum anthropi, 59 Ofloxacin, 229, 230,255,256 Oral cavity, 11-14 Organ transplantation, 186, 188. See also Bone marrow transplantation; Renal transplantation Ototoxicity, 258-259, 260 Ovarian cancer, 130 Paecilomyces, 177 Papovaviruses (polyomaviruses), 193 Parasites, 63-64, 156, 158, 159 Pelvic cancer, 123-131 Penicillin-binding proteins, 250, 251 Penicillin G, 126 Penicillins, 16, 55-56, 58, 85, 87, 88, 107, 230,252,253,259 Penicillium, 177, 262 Pentamidine, 115

Peptostreptococcus, 133 Perfloxacin, 256 Phaeohyphornycosis, 176,178 Pharmacod ynamics of aminoglycosides, 258 of beta-lactams, 250 of quinolones, 254-255 of vancomycin, 260 Pharmacokinetics of aminoglycosides, 257-258 of amphotericin B, 261-262 of antivirals, 269-270 of azoles, 265-266 of beta-lactams, 247-249 of ganciclovir, 273 of quinolones, 253-254 of vancomycin, 259-260 Phialophora, 176, 178 Phialophora parasiticia, 62 Pichia farinosa, 62 Piperacillin, 16, 59, 88, 90, 91, 127, 202, 248,249,251,253 PIXY 321,209 Platelets, 17 Pneumococcal vaccines, 41 Pnez~rnocystiscarinii pneumonia (PCP), 22, 23, 25, 53, 85,226 lung cancer and, 114-115 microbiology laboratory evaluation of, 156 prevention of, 240-241 reclassification of organism, 63 Pneumonia, 107,108-109,134,135 Pneumocystis carinii. See Pneumocystis carinii pneumonia radiation, 116-117 Polymyxin, 229 Polyomaviruses (papovaviruses), 193 Polysaccharide vaccines, 210 Prednisone, 22, 114,224 Probenecid, 272-273 Propionibacteriurn, 53 Protected environment, 225-226 Proteiis, 81 Prototheca wickerhamii. 65 Protozoal infections, 25; 51, 63-64, 240241 Pseudallesceria boydii, 177 Psei~dallescheriaboydii, 61-62 Pseudornonas, 43, 53, 61 Pseudomonas aeruginosa, 14, 37, 48, 81, 88,89, 162,230 aminoglycosides for, 258 beta-lactams for, 249,250,251

interleukin-1 for, 208 quinolones for, 254,255,257 Pseudomonas cepacia, 59 Pyogenic liver abscesses, 126-128 Pyrimethamine, 63 Pyrimethamine/sulfadoxine, 241 Quinolones, 16, 38, 46,48,55, 59,228, 230 aminoglycosides and, 258 for febrile neutropenia, 81-82, 90-91 pharmacologic considerations with, 253-257 Salmonella and, 128 Radiation, 19, 24, 38, 44 for breast cancer, 122-123 for head and neck cancer, 132 intestinal tract effects of, 16-17 oral cavity effects of, 11-14 skin effects of, 9 Radiation pneumonitis, 116-1 17 Radiation recall, 122-123 Rapid detection tests, 156-160 Red man syndrome, 260 Renal transplantation, 61, 268 Respiratory specimens, 146-147 Respiratory syncytial virus (RSV), 116, 155,160,192-193 Retroviruses, 193 Rhdotorula rubra, 62 Rhizom~icor,174 Rhizopus, 155,174,262 Rhodococcus equi, 58 Rhodotorula, 174,175-176 Ribavirin, 192-193 Rifampin, 131,162,229,268-269 Rifamycin, 16 Rimantidine, 192 Roxithromycin, 228,230 Saccharovlzyces cereviseae, 62 Salmonella, 128 Scedosporiurn, 176, 177 Scedosporium injlatum, 62 Scopulnriopsis, 61-62 Seizures, 253,275 Selective decontamination, 93, 226-230 Serum concentration monitoring, 261 Sicca syndrome, 25 Skin, 1, 4, 7-10, 3940, 169 Small cell lung cancer, 53, 107, 213,202, 206 Smoking, 5, 134

Soft tissues, 39-40 Solid tumors, 50. See also specific types cellular immunity and, 38 infections related to, 52-53, 105-136 Sparfloxacin, 254,255-256 Splenectomy, 6,24, 35, 40-41, 44, 51, 224 Stains, 143-151 Staphvlococcus, 132 ~ta~h)lococcus aureus, 37,89,91,109, 110,119,122,255 Staphylococcus epidermidis, 13, 37, 42, 89,91,122 Stenotrophomonas (Xanthomonas) malfophilia, 59, 89 Stevens-Johnson syndrome, 268 Stewart-Treves syndrome, 123 Stomatococcus mucilaginosis, 58 Streptococci, 55-56 alpha-hemolytic, 37, 48, 81, 109 viridans, 12-13, 55 Streptococcus bovis, 124 Streptococcus mitis, 13, 55 Streptococcus mutans, 132 ~tre~tococcus pneumoniae, 6, 55-56, 149 Streptomycin, 42-43 Strongyloides stercoralis, 64 Sulfadiazine, 63 Surgery, 43-44 Synergy, 250-251,255,258,260 Tazobactam, 59,90,127,251 Terfenadine, 269 Testicular carcinoma, 53 Tetracycline, 65, 128 Theophylline, 255, 256-257 Thiabendazole, 64 Thrombocytopenia, 17, 207,274 Ticarcillin/clavulanate, 127 Tobramycin, 9, 88, 202,258,259 Torsade de pointes, 256, 269 Torulopsis, 155 Torulopsis glabrata. See Candida glabrata Toxoplasma gondii, 22,63,241 Triazolam, 269 Triazoles, 231, 232 Trichosporon, 174,175,231,262 Trichosporon capitatum. See Blastoschizomyces capitatus Trichosporon cutaneum (Trichosporon beigelii), 62, 174 Trimethoprim, 9,255

Trimethoprim/sulfamethoxazole (TMPSMX; co-trimoxazole), 16, 81-82, 94,115,128,226-229,230,240-241 Trypanosomiasis, 47 Tuberculosis, 23, 110-1 11, 134, 210 Tumor necrosis factor (TNF), 201,202, 205 Tumor necrosis factor-a (TNF-a), 18, 21,98-99 Upper respiratory tract, 10-14 Urine specimens, 147-149 Uterine cancer, 129 Vaccines, 6,41, 191,210,236-238 Valacyclovir, 189,270, 271 Vancomycin, 16, 89, 123,225, 255 for febrile neutropenia, 94-95 pharmacologic considerations with, 259-261 resistance to, 46, 55, 58, 82, 106, 161, 162,226 selective decontamination and, 229, 230 Varicella vaccines, 191, 210,236-238 VaricelIa zoster immune globulin (VZIG), 236 Varicella zoster virus (VZV), 23,25, 51, 54, 61,83, 185, 186,269,272

microbiology laboratory evaluation of, 155,160 prevention of, 236-238 properties and treatment of, 189 Viral infections emerging pathogens in, 61 febrile neutropenia and, 83-85 in immunocompromised hosts, 183 in lung cancer, 115-116 microbiology laboratory evaluation of, 155-156,157 prevention of, 235-240 recent advances in the management of, 183-193 syndromes due to, 184 Viridans streptococci, 12-13, 55 Vitamin A deficiency, 3 Vulvar cancer, 129 Waldenstrom's macroglobulinemia, 39 Warfarin, 256 Wound specimens, 150 Yeasts, 174-175 Zidovudine, 272, 275 Zinc deficiency, 3-4