Randomised trial of efficacy of benznidazole in treatment of early Trypanosoma cruzi infection

Benznidazole, a nitroimidazole derivative, has been recommended for the treatment of acute and congenital Trypanosoma cruzi infection (Chagas’ disease). We have examined the safety and efficacy of this drug in the treatment of the early chronic phase of T cruzi infection. Between 1991 and 1995, we carried out a randomised, double-blind, placebo-controlled trial in a rural area of Brazil with endemic Chagas’ disease. 82% of 2434 schoolchildren (aged 7–12 years) identified in a census were screened for antibodies to T cruzi by indirect immunofluorescence, indirect haemagglutination, and ELISA. 130 were positive in all tests and were randomly assigned benznidazole (7·5 mg/kg daily for 60 days by mouth) or placebo. The primary endpoint for efficacy was the disappearance of specific antibodies (negative seroconversion) by the end of 3-year follow-up. The secondary endpoint was the reduction of antibody titres on repeated serological tests. One child moved away from the area just after randomisation and was excluded from the analyses. Insecticidal measures were taken throughout the trial to reduce the risk of reinfection. Minor side-effects requiring no specific medication were recorded in a small proportion of individuals. On a chemiluminescent ELISA with purified trypomastigote glycoconjugate, serum from all participants was positive at the beginning of the trial. At the end of follow-up, 37 (58%) of the 64 benznidazole-treated participants and 3 (5%) of those who received placebo were negative for T cruzi antibodies. The efficacy of benznidazole treatment estimated by intention to treat was 55·8% (95% Cl 40·8–67·0). At the end of follow-up, children who received benznidazole had five-fold lower geometric mean titres by indirect immunofluorescence than placebo-treated children (196 [147–256] vs 1068 [809–1408], p