Randomized comparison between two ceftazidime-containing regimens and cephalothin-gentamicin-carbenicillin in febrile granulocytopenic cancer patients

Because the results of our published trial [R. Ramphal, B. S. Kramer, K. H. Rand, R. S. Weiner, and J. W. Shands, Jr., J. Antimicrob. Chemother. 12(Suppl. A):81-88, 1983] of ceftazidime versus cephalothin, gentamicin, and carbenicillin (KGC) revealed a preponderance of gram-positive superinfections, including those caused by clostridia, in patients treated with ceftazidime, we added vancomycin to the ceftazidime regimen at study entry 49 and continued with a 2:1 randomized comparison of ceftazidime-vancomycin (CV) versus KGC. Criteria for study entry were fever (temperature, greater than or equal to 38.5 degrees C on one occasion or greater than or equal to 38 degrees C on two occasions 6 h apart) and granulocytopenia (less than 500/mm3 or a falling count anticipated to be less than 500/mm3). Ninety-five entries (79 patients) were evaluable. The numbers of initial clinical responses for ceftazidime-, KGC-, and CV-treated patients were 9 of 21 (43%), 21 of 37 (57%), and 21 of 37 (57%), respectively; differences were not significant. The death rate was lower with CV (2 of 37 patients) than with KGC (10 of 37 patients) (P less than 0.05 by two-tailed analysis) or with ceftazidime alone (7 of 21 patients) (P less than 0.025). Death from presumed infections occurred in 9 of 37 KGC-treated patients versus 1 of 37 CV-treated patients (P less than 0.025). Superinfections occurred in five ceftazidime-treated patients (24%) versus 7 KGC-treated patients (19%) but not in CV-treated patients (CV versus KGC, P less than 0.05; CV versus ceftazidime, P less than 0.01). CV appears to be superior to KGC or ceftazidime alone in the management of febrile granulocytopenic cancer patients.