Management of Patients With Fever and Neutropenia Through the Arc of Time

Abstract

Key Summary Points Fever and neutropenia due to cytotoxic cancer chemotherapy can result in risk for life-threatening infections. Although neutropenia is a quintessential risk factor, patients with cancer also have a panoply of disease- and treatment-related alterations of their innate and acquired immune defenses, rendering them vulnerable to infection with bacteria, viruses, fungi, and protozoa. Beginning broad-spectrum empirical antibiotic therapy at the first sign of fever in a profoundly neutropenic patient can be life-saving and has been a standard of therapy for nearly 5 decades. Low-risk neutropenic patients can have shorter durations of treatment, including oral regimens, whereas high-risk patients (>7 d of neutropenia) often require additions and modifications of the initial regimen, as well as more prolonged treatment courses. The advent of hematopoietic cytokines (granulocyte colony-stimulating factor) and the selected use of prophylactic antimicrobial regimens have altered risk for infectious complications in high-risk patients. Since Bodey and colleagues seminal study in 1966 (1), the medical community has known that fever in a patient with neutropenia (neutrophil count <0.50109 cells/L) could signal potentially life-threatening infection and should prompt immediate empirical therapy with broad-spectrum antibiotics (13). Although neutropenia is still the quintessential risk factor for infection after cytotoxic chemotherapy, many neutropenic patients also have other disease- or treatment-related perturbations of their host defense matrixfrom breaches of physical and mucosal barriers to alterations in the innate, cellular, and humoral immune systemthat make them vulnerable to bacterial, viral, fungal, and parasitic infections. In 2019, we stand at a time of remarkable progress in cancer therapy. Over past decades, treatment advances have increased survival rates for childhood cancer from less than 10% to nearly 90% (4). Improvements in the treatment of many types of adult cancer have also been notable (5). The 1998 discovery that a tyrosine kinase inhibitor provided a selective and effective treatment of chronic myelogenous leukemia led to the first targeted therapy that might avoid the complications of cytotoxic chemotherapy (6, 7). Many other small molecules have been developed for an array of defined molecular targets and are being integrated into regimens for leukemia, lymphoma, and solid tumors. Targeted therapies also include monoclonal antibodies and an expanding repertoire of immunotherapeutic agents (checkpoint inhibitors and chimeric antigen receptor T cells), some of which also alter the host s microbiome and risk for infection (8). Despite the effect of novel immune-based therapies, cytotoxic chemotherapy (with its associated risk for fever and infection) is still the basis of most cancer treatment regimens. Fortunately, the introduction of hematopoietic cytokines, including granulocyte colony-stimulating factor, can shorten the duration of neutropenia from 6 days to 1 day and can reduce the incidence of infection and its complications, but not mortality, by 50% in high-risk patients (911). Granulocyte colony-stimulating factor is currently recommended as primary prophylaxis against neutropenia in patients whose risk is higher than 20%, as well as in patients who are older than 65 years or are receiving intensive chemotherapy regimens. With these advances, risk stratification and empirical antimicrobial treatment of patients receiving cytotoxic chemotherapy has evolved. In 1990, the International Immunocompromised Host Society published guidelines for the design, analysis, and reporting of clinical trials on the antibiotic management of neutropenic patients (12). This report developed standards for the conduct of such trials, and many of its recommendations remain relevant today. Of note, the report coincided with the first guidelines for antimicrobial therapy in febrile neutropenic patients (13), which were updated in 2002 (14) and 2011 (15). In 2018, the Infectious Diseases Society of America and American Society of Clinical Oncology released recommendations for the outpatient management of fever and neutropenia in lower-risk adults having cancer treatment (generally defined as those with neutropenia <7 days in duration) (16). Table 1 summarizes the evolution of recommendations from 1993 to 2018. Table 1. Evolving Principles for the Management of Fever and Neutropenia Over Time In concert with changes in cancer therapy, advances in the antimicrobial armamentarium and the development of host-modifying agents have led to changes in the management of neutropenic patients, especially those at lower risk (<7 days of neutropenia). Once empirical antibiotic therapy is initiated, when and how it should be modified when patients remain neutropenic for periods longer than 1 week remains an important issue. Table 2 summarizes recommendations for modifications to antimicrobial therapy in 2018 and compares them to 1993 recommendations. Table 2. Common Modifications or Additions to Initial Empirical Antibiotic Therapy in Patients With Neutropenia and Fever* Observations beginning as early as the 1970s and 1980s showed that nearly 80% of the microorganisms associated with infection in the febrile neutropenic patient arose from endogenous microbial flora (18, 19). These observations forecast the future understanding of the balance between aerobic and anaerobic organisms, the role of selective decontamination, and the role of the microbiome in risk for infection and modulation of host defenses, including risk for graft-versus-host disease. The importance of colonization and alterations in the microbiome was shown in 2017 in a multicenter study of 1118 recipients of allogeneic stem cell transplants and 1625 of autologous hematopoietic stem cell transplants, in whom colonization with resistant gram-negative bacteria was associated with infection by the same bacteria after transplant (18, 20, 21). In the 1970s, researchers observed a shift in the oral and gastrointestinal microflora from the normal gram-positive, anaerobic environment when a patient is first diagnosed with cancer to one dominated by gram-negative organisms after onset of illness and associated therapies that alter the balance of aerobes and anaerobes. Van der Waaij (19) showed that anaerobes are necessary to help balance the proliferation of aerobes on the basis of studies that found significant differences in the log-rank number of aerobes needed to colonize a mouse whose anaerobes were preserved compared with mice whose anaerobes were decreased by antibiotics. The gut is a complex microenvironment where anaerobes inhibit colonization by new aerobes by altering metabolism and nutrient availability and producing inhibitory toxins and fatty acids. The gut microbiome can influence the response to chemotherapy, including stem cell therapy, and modulate the immune system (22, 23). Several studies have shown that the diversity and composition of the gut microbiota is correlated with risk for fever, neutropenia, and infection. For example, Hakim and colleagues (24) showed in children with acute lymphocytic leukemia that an abundance of Proteobacteria before chemotherapy was associated with development of fever and neutropenia. Whether the gastrointestinal microbiota is dominated by Enterobacteriaceae or Streptococcaceae during chemotherapy predicts risk for subsequent infection (24). Infection in immunocompromised neutropenic patients can be caused by fungi, viruses, and parasites, as well as by bacteria. Important fungi include Candida, Aspergillus, Mucor, Trichosporon, Fusarium, Scedosporium, and dematiaceous molds (2530). Viruses are also important causes of infection in immunocompromised patients and include adenovirus (31, 32); the herpesviruses (herpes simplex virus, cytomegalovirus, varicella-zoster virus, EpsteinBarr virus, and human herpesvirus 6); and respiratory viruses, such as influenza, parainfluenza, respiratory syncytial virus, coronavirus, human metapneumovirus, and rhinovirus. When and Where to Initiate Antibiotics and Which Ones Should be Administered In the early days of cytotoxic chemotherapyinduced neutropenia, clinicians learned that neutropenic patients must be promptly evaluated and, ideally, start empirical antibiotic therapy within an hour of fever onset. Over decades, researchers have sought strategies to differentiate patients with life-threatening infection from those whose fever might not be of infectious cause and ways to determine which patients require longer versus shorter antibiotic courses, but these questions remain unresolved. Despite advances in culture-independent technologies (for example, inflammatory markers, molecular sequencing techniques and polymerase chain reaction, immunodiagnostics, and imaging [33, 34]), their usefulness in identifying serious infections, including invasive mycoses, remains limited. Although many assays have been developed to diagnose invasive mycoses, sensitivity is limited (galactomannan, 44% to 90%; [1,3]--D-glucan, 30% to 100%; and polymerase chain reaction, 84%) (17, 3437). Further, they can be unreliable when a single sample is measured, although serial sampling has better performance (3136, 3841). Of note, many of the assessment tools, such as physical examination and blood cultures, that were pillars of diagnosis 40 to 50 years ago still serve as standard diagnostic tools. Selected imaging studies have proven valuable for diagnosing invasive fungal infections of the lung, sinuses, and brain in patients with prolonged neutropenia (31, 34). When principles for initial empirical antibiotics for neutropenic fever were first being formulated, no single antibiotic could provide coverage against the broad array of potential gram-positive and gram-negative aerobes and anaerobes. Thus, combination antibiotic therapy was the rule, generally with a first-generation cephalospor