Background: The aim of the study was to compare the efficacy of empirical antifungals in combination with broad spectrum antibiotics with that of antibiotics alone in high risk febrile neutropenic cancer patients not responding to initial antibacterial therapy.
Patients and methods: A prospective, randomized controlled trial was conducted at 22 cancer centers in Germany. Patients with fever of unknown origin were randomized to either piperacillin (Pip) plus an aminoglycoside (AMG) (arm A) or a third generation cephalosporin (Ceph) plus AMG (arm B). Patients not responding after 4-6 days were randomized to either imipenem (Imi) plus glycopeptide (GLP) (arm C), or Imi/GLP plus amphotericin B deoxycholate (AmB) plus 5-flucytosine (5-FC) (arm D), or Imi/GLP plus fluconazole (Fluco) (arm E). A successful outcome was defined as resolution of fever.
Results: In arm A, 192 of 373 patients (51.5%) responded as compared to 176 of 344 patients (51.2%) in arm B. The response rates of 155 patients randomized for further empirical treatment were 55.6%, 77.8% and 62.5% in arm C, D and E, respectively. The difference between arm C and D was of borderline statistical significance (p = 0.06) after correction for multiple testing.
Conclusion: In neutropenic cancer patients with persistent fever the combination of antibiotics with AmB/5-FC is superior to salvage antibacterial therapy alone. There is no difference in efficacy between Pip and third generation Ceph given as initial empirical therapy in combination with an AMG.