Utility-based interim decision rule planning in adaptive population selection designs with survival endpoints

The use of adaptive population selection designs has become widespread in response to the emergence of numerous targeted therapies. Such designs provide an opportunity to stop the recruitment of patients in a population in midcourse when this population does not benefit from the treatment being tested. Several types of designs have been developed in different settings; however, the procedures for setting thresholds for an interim decision have not yet been established. Therefore, in this article, we propose a novel, simple utility-based approach to guide the construction of an interim decision rule for an adaptive population selection design using survival endpoints. This approach is applicable when there is an available biomarker prior to designing a clinical trial. Our simulation studies, which assumed that only a prespecified population will benefit from a targeted therapy, indicated that the proposed approach assists in setting optimal thresholds for selecting an appropriate population at the interim analysis.