Trimethoprim-sulfamethoxazole and nystatin prophylaxis in children with acute lymphoblastic leukemia

Abstract

In 1980 we began a prospective randomized trial of trimethoprim-sulfamethoxazole (TMPSMZ)gnd nystatin prophylaxis in children with acute lymphoblastic leukemia during remission induction therapy. The study was planned (1) to determine the efficacy of this combination in reducing the incidence of fever, infection, and hospitalization, (2) to assess the hematologic toxicity in these patients, and (3) to evaluate the method of randomization proposed by Zelen to expedite clinical studies. Sixty-seven patients were eligible; 30 were randomized by birth date to receive prophylaxis; 37 were randomized to receive no prophylayis (control group). Sixty patients received appropriate therapy and were fully evaluabll-. Eight episodes of infection and seven fevers of unknown origin occurred in the control group; two documented infections and three fevers of unknown origin occurred in the group receiving prophylaxis (p = 0.06). The control group spent 139 of 642 neutropenic days in the hospital; the prophylaxis group spent 39 of 462 neutropenic days in the hospital (p < 0.001). Significant differences in mean neutrophil counts (846 vs. 514) were found in the two groups only at day 14. The method of randomization did expedite enrollment in the study but introduced a bias by allowing for refusals in the prophylaxis group only. We conclude that TMP-SMZ and nystatin prophylaxis benefits this group of patients because of the reduction in hospitalization, but because of potential myelosuppression and changing chemotherapeutic induction regimens, studies with larger numbers of homogeneous groups of patients are needed before routine prophylaxis can be recommended for all patients with this disease.