Direct evidence gap on fixed versus adjusted-dose benznidazole for adults with chronic Chagas disease without cardiomyopathy: Systematic review and individual patient data meta-analysis.

ABSTRACT

OBJECTIVES: To determine the comparative efficacy and safety of a fixed dose of benznidazole (BZN) with an adjusted-dose for Trypanosoma cruzi-seropositive adults without cardiomyopathy. METHODS: We conducted a systematic review and individual participant data (IPD) meta-analysis following Cochrane methods, and the PRISMA-IPD statement for reporting. Randomised controlled trials (RCTs) allocating participants to fixed or adjusted doses of BZN for T. cruzi-seropositive adults without cardiomyopathy were included. We searched (December 2021) Cochrane, MEDLINE, EMBASE, LILACS and trial registries and contacted Chagas experts. Selection, data extraction, risk of bias assessment using the Cochrane tool, and a GRADE summary of finding tables were performed independently by pairs of reviewers. We conducted a random-effects IPD meta-analysis using the one-stage strategy, or, if that was impossible, the two-stage strategy. RESULTS: Five RCTs (1198 patients) were included, none directly comparing fixed with adjusted doses of BZN. Compared to placebo, BZN therapy was strongly associated with negative qPCR and sustainable parasitological clearance regardless of the type of dose and subgroup analysed. For negative qPCR, the fixed/adjusted rate of odds ratios (RORF/A ) was 8.83 (95% CI 1.02-76.48); for sustained parasitological clearance, it was 4.60 (95% CI 0.40-52.51), probably indicating at least non-inferior effect of fixed doses, with no statistically significant interactions by scheme for global and most subgroup estimations. The RORF/A for treatment interruption due to adverse events was 0.44 (95% CI 0.14-1.38), probably indicating no worse tolerance of fixed doses. CONCLUSIONS: We found no direct comparison between fixed and adjusted doses of BZN. However, fixed doses versus placebo are probably not inferior to weight-adjusted doses of BZN versus placebo in terms of parasitological efficacy and safety. Network IPD meta-analysis, through indirect comparisons, may well provide the best possible answers in the near future. REGISTRATION The study protocol was registered in PROSPERO (CRD42019120905).