Ceftazidime with or without vancomycin vs. cephalothin, carbenicillin and gentamicin as the initial therapy of the febrile neutropenic pediatric cancer patient

Abstract

In a 28-month randomized trial we compared ceftazidime (CAZ), an extended spectrum cephalosporin, with cephalothin, carbenicillin and gentamicin (KCG) as empiric therapy for febrile neutropenic pediatric cancer patients. Because of the occurrence of ceftazidime-resistant Gram-positive primary infections, vancomycin was added to CAZ after the first year of study. Of 206 evaluable episodes 76 (37%) were documented infections including 20 bacteremias; 130 (63%) episodes were caused by fever of unknown origin. The number of complete responses to initial therapy in patients with documented infections did not differ among regimens: 26 of 43 (61%) for KCG, 9 of 16 (56%) for ceftazidime and 8 of 16 (50%) for CAZ + vancomycin (not significant). In patients with fever of unknown origin, response without modification of the initial regimen was 52 of 62 (84%) in the KCG arm, 32 of 40 (80%) on CAZ and 23 of 29 (80%) in patients treated with CAZ + vancomycin (not significant). Modifications of the regimen were similar among all three groups and were due primarily to the use of empiric antifungal or antiviral therapy and to empiric treatment of interstitial pneumonia. Hypokalemia occurred in 25 of 105 patients treated with KCG and in 4 of 101 treated with CAZ or CAZ + vancomycin (P less than 0.001). No differences between the efficacy of KCG, CAZ and CAZ + vancomycin as initial empiric therapy were demonstrated.