Crossover in a randomized trial can skew the interpretation of the efficacy of a cancer drug. In this Comment, I use examples from clinical trials presented at the 2023 ASCO annual meeting to highlight why ‘allowing’ crossover in randomized trials testing cancer drugs is problematic, and propose that crossovers should either be mandated or prohibited depending on the context.
Crossover in a randomized cancer drug trial allows patients in the control arm to receive the agent administered in the experimental arm at the time of disease progression. An ongoing debate exists in the oncology community regarding the importance of such crossovers in trials. The incorporation of crossover into the design of clinical trials is inherently neither desirable nor undesirable and depends mostly on the context. At this year’s ASCO Annual Meeting, data were presented from several trials with crossover designs that were problematic. In this Comment, I provide