Fluconazole versus amphotericin B as empirical antifungal therapy of unexplained fever in granulocytopenic cancer patients: a pragmatic, multicentre, prospective and randomised clinical trial.

Amphotericin B, despite its intrinsic servere toxicity, is the most commonly used empirical antifungal therapy in cancer patients with unexplained fever not responding to empirical antibacterial therapy. The aim of this study was to show whether fluconazole was as effective as, and less toxic than, amphotericin, with no effort made to compare the antifungal activity of the two drugs. A group of 112 persistently febrile (> 38 degrees C) and granulocytopenic (< 1000 cells/mm3) cancer patients, not receiving any absorbable antifungal antibiotic for prophylaxis, with a mean age of 27 years (range 1-73 years), undergoing chemotherapy for a variety of malignancies and with a diagnosis of unexplained fever after at least 96 h of empirical antibacterial therapy, were randomised to receive either fluconazole (6 mg/kg/day up to 400 mg/day) or amphotericin B (0.8 mg/kg/day) as empirical antifungal treatment. Patients were required to have normal chest X-rays at randomisation, no previous history of aspergillosis and negative surveillance cultures for Aspergillus. The intention-to-treat analysis showed defervescence and survival without treatment modification in 42 of 56 patients (75%) in the fluconazole group and in 37 of 56 (66%) in the amphotericin B group (P = 0.4). Duration of therapy was 6 days (95% CI = 4-8 days) in both groups. Death occurred in 3 patients (5%) in the fluconazole and in 2 (4%) in the amphotericin B group. No fungal death was documented in either group. Adverse events developed in 18 of 56 patients (32%) in the fluconazole group and in 46 of 56 (82%) in the amphotericin B group (P < 0.001). In the amphotericin B group, 5 patients had treatment discontinued because of toxicity, versus none in the fluconazole group, a difference which approached statistical significance (P = 0.06). This study shows that fluconazole is by far less toxic than amphotericin B and suggests that it might be as effective as amphotericin B, in pragmatical terms and for this specific indication. However, numbers are too small to allow definitive conclusions about efficacy, and the use of fluconazole for this indication remains experimental. Future studies should try to identify patients more at risk of fungal infections, with the aim of individualising antifungal approaches.