Reduction of Fever and Streptococcal Bacteremia in Granulocytopenic Patients With Cancer

Abstract

Objective. —To determine the effect of oral penicillin V combined with a fluoroquinolone (pefloxacin) on the occurrence of fever and streptococcal and other gram-positive coccal bacteremic infections in granulocytopenic patients with cancer. Design. —Prospective randomized double-blinded placebo-controlled prophylactic trial. Setting. —Inpatient setting in multiple cooperating cancer centers. Subjects. —Convenience sample with a total of 551 granulocytopenic patients, 95% of whom had leukemia or underwent bone marrow transplantation. Interventions. —Penicillin V (500 mg twice a day) vs placebo given in combination with oral pefloxacin (400 mg twice a day). Main Outcome Measures. —Occurrence of fever and/or infection. Results. —Fever or infection (without fever) developed in 190 (71%) of 268 evaluable patients in the penicillin arm compared with 213 (80%) of 268 evaluable patients in the placebo arm ( P =.03; 95% confidence interval [CI] for the difference, -16% to -1%). Bacteremia occurred in 58 (22%) of 268 placebo-treated patients and in 38 (14%) of 268 penicillin-treated patients ( P =.03; 95% CI for the difference, -14% to -1%), primarily due to a reduction in streptococcal bacteremic episodes that occurred in 14 penicillin-treated patients (5%) and in 27 placebo-treated patients (10%) ( P =.05; 95% CI for the difference, -9% to -0.3%). Gram-negative rod bacteremias occurred in only two patients (1%) and in five patients (2%), respectively. Logistic regression analysis also supported the treatment effect on the development of bacteremia. Conclusions. —These results demonstrate that the addition of penicillin V to fluoroquinolone prophylaxis in granulocytopenic patients with cancer effectively reduces febrile episodes and the incidence of bacteremia, especially that due to streptococcal species. ( JAMA . 1994;272:1183-1189)