Amphotericin B vs high-dose ketoconazole for empirical antifungal therapy among febrile, granulocytopenic cancer patients. A prospective, randomized study.

We compared high-dose ketoconazole (800 mg/kg per day, orally) with amphotericin B (0.5 mg/kg per day, intravenously) for empirical antifungal therapy in a prospective, randomized study of persistently or recurrently febrile granulocytopenic cancer patients. Among 97 patients eligible for empirical antifungal therapy, 20 (21%) of these patients were ineligible for randomization to ketoconazole treatment because of their inability to tolerate oral medications. Among 72 patients eligible for randomization, 64 were assessable (32 in each arm of the study). Five of six patients with proved fungal infections who were randomized to receive ketoconazole treatment required crossover to amphotericin B treatment because of progressive infection. The conditions of three of these five patients improved after receiving amphotericin B. The frequency of transaminase elevation was higher in those receiving ketoconazole, while the frequency of azotemia was higher in those receiving amphotericin B. Bioavailability of ketoconazole was unpredictable. Amphotericin B remains the drug of choice for empirical antifungal therapy in granulocytopenic patients; whereas, lack of a parenteral formulation, ineffectiveness against proved mycoses, and unreliable bioavailability preclude high-dose ketoconazole from being an appropriate compound for this purpose.