Background: The use of combinations of antibiotics has been the cornerstone of therapy for febrile patients with cancer and severe neutropenia. Each empirical regimen should be selected according to the epidemiology and susceptibility patterns in each center. We describe here the experience with empirical antimicrobial treatments in pediatric patients with cancer, fever and severe neutropenia, and identify the risk factors associated with treatment failure.
Methods: This is a prospective study including 145 patients with cancer, and 171 episodes of neutropenia and fever. Blood cultures were taken before initiating empirical treatment: a) carbenicillin (400 mg/kg/day) plus amikacin (21 mg/kg/day) (Cb/ak), and b) ceftazidime (100 mg/kg/day), plus amikacin at the same dosage (Cz/ak).
Results: The overall response rate was 54.9% and 56.3% for Cb/ak and Cz/ak, respectively. Fifty-seven episodes (33.3%) were microbiologically documented, gram-positive isolated in 38% and gram-negative in 49%. Risk factors associated significantly with treatment failure were acute myelocytic leukemia (AML) (RR 2.59, CI 95% 1.42-4.7, p = 0.003); bacteriological identification (RR = 4.41, CI 95% 2.21-8.8, p < 0.001), and the presence of two or more sites of infection (RR = 2.89, CI 95% 1.03-8.11, p = 0.03).
Conclusions: The rates of response are similar to the combinations used in the hospital (Cb/ak, Cz/ak). The risk factors associated with treatment failure were AML diagnosis, bacteriological identification, and the presence of two or more sites of infection.