Value of the Run-In Period to Evaluate the Safety of Conventional Trypanocidal Treatment: A Subanalysis of a Colombian Randomized Clinical Trial

When testing poorly tolerated or long-term treatments, including a run-in phase enhances clinical trial efficiency and internal validity by selecting more adherent participants. This report describes the adherence and tolerance of Colombian participants in EQUITY (a randomized, concealed, parallel-group, placebo-controlled trial testing nifurtimox and benznidazole among Trypanosoma cruzi–seropositive adults without cardiomyopathy). Our design included a 10-day, single-blind, placebo run-in phase. On completion, willing participants reporting good adherence ($80%) and tolerance were randomized to any five 120-day, blinded treatments: four with either active medication, each given as 120-day half dose or 60-day full dose (followed/preceded by a randomly allocated 60-day placebo treatment), or a 120-day placebo. Side effects were compared after the run-in (day 0) between excluded and enrolled participants, and between those randomized to active medications or placebo 20 days after starting each treatment period (days 20/80). Those excluded (44/351, 12.5%) more often reported gastrointestinal (15.9/4.6%), nonspecific (13.7/4.2%), and musculoskeletal symptoms (9.1/1.6%) than those randomized. Participants given nifurtimox (n 5 84) or benznidazole (n 5 86) versus placebo (n 5 126) on day 20 reported more nonspecific (15.5/10.5/4.8%, respectively) or cutaneous side effects (6.0/12.8/3.2%). When starting active treatments (transition OFF–ON, n 5 171 and n 5 61 in first and second 60-day treatment periods), more participants reported emerging-worsening than receding-ending side effects (49/13 and 15/5, respectively). Despite inducing a nocebo effect, the run-in phase highlighted more closely related side effects, strengthening causal inference and informing adherence and tolerance to conventional trypanocides.