Design of Clinical Trials of Empiric Antifungal Therapy in Patients with Persistent Febrile Neutropenia: Considerations and Critiques

Abstract

Febrile neutropenic patients who receive antibiotics are at risk for fungal infections. This risk increases greatly with the length and severity of neutropenia. Because diagnostic tests for fungal infections lack sensitivity and specificity and because established fungal infections are associated with poor outcomes, empiric antifungal therapy is frequently given to patients with fever that persists despite antibacterial therapy. Early trials of empiric amphotericin B showed reductions in the number of invasive fungal infections and in related morbidity and mortality. However, as a result of infusion-related and renal adverse effects of amphotericin B, newer agents, such as lipid formulations of amphotericin B, extended-spectrum azoles, and echinocandins, have been developed. Although these alternatives have been associated with decreased toxicity, improved efficacy has not been clearly demonstrated. Although empiric antifungal therapy can prevent undetected breakthrough infections and morbidity associated with many fungal infections, its shortcomings include overtreatment, toxicity, and increased costs of unnecessary treatment. Recent studies have highlighted several questions in trial design and data interpretation. For example, what is the appropriate study design? Who should be enrolled in studies of empiric antifungal therapy? How should successful therapy be defined? These issues are reviewed to determine whether new antifungal agents should be evaluated for empiric use in patients with fever and neutropenia.