The primary goal of an exploratory oncology clinical trial is to identify an effective drug for further development. To expedite the drug development process and increase the chance of finding active tumor indications where the treatment works, multiple tumor cohorts are often investigated simultaneously in a basket trial. In this article, we extend the optimal basket trial design in Zhou et al. (2019) and propose a generalized framework of an optimal basket trial design in the exploratory setting where tumor indications can be homogeneous (objective response rates of all indications are the same under both the null and alternative hypotheses) or heterogeneous (objective response rates are not the same under at least one of the null or alternative hypotheses). The proposed design prunes the inactive tumor indication in stage 1 and pools the remaining tumor indications at the end of stage 2 to evaluate the overall effectiveness of whether the treatment works in at least one tumor indication. The design parameters are optimized to minimize the expected sample size while explicitly controlling the global type I and type II error rates. In addition, we consider reallocating the planned stage 2 sample size of pruned indications to achieve a higher power when the total planned sample size is fixed. Simulation studies are conducted to show the favorable operating characteristics of the proposed design under certain scenarios.

The primary goal of an exploratory oncology clinical trial is to identify an effective drug for further development. To expedite the drug development process and increase the chance of finding active tumor indications where the treatment works, multiple tumor cohorts are often investigated simultaneously in a basket trial. In this article, we propose a generalized framework of an optimal basket trial design in the exploratory setting where tumor indications can be homogeneous (objective response rates of all indications are the same under both the null and alternative hypotheses) or heterogeneous (objective response rates are not the same under at least one of the null or alternative hypotheses). The proposed design prunes the inactive tumor indication in Stage 1 and pools the remaining tumor indications at the end of Stage 2 to evaluate the overall effectiveness of whether the treatment works in at least one tumor indication. The design parameters are optimized to minimize the expected sample size while explicitly controlling the global Type I and Type II error rates. In addition, we consider reallocating the planned Stage 2 sample size of pruned indications to achieve a higher power when the total planned sample size is fixed. Simulation studies are conducted to show the favorable operating characteristics of the proposed design under certain scenarios.

The primary goal of an exploratory oncology clinical trial is to identify an effective drug for further development. To expedite the drug development process and increase the chance of finding active tumor indications where the treatment works, multiple tumor cohorts are often investigated simultaneously in a basket trial. In this article, we propose a generalized framework of an optimal basket trial design in the exploratory setting where tumor indications can be homogeneous (objective response rates of all indications are the same under both the null and alternative hypotheses) or heterogeneous (objective response rates are not the same under at least one of the null or alternative hypotheses). The proposed design prunes the inactive tumor indication in Stage 1 and pools the remaining tumor indications at the end of Stage 2 to evaluate the overall effectiveness of whether the treatment works in at least one tumor indication. The design parameters are optimized to minimize the expected sample size while explicitly controlling the global Type I and Type II error rates. In addition, we consider reallocating the planned Stage 2 sample size of pruned indications to achieve a higher power when the total planned sample size is fixed. Simulation studies are conducted to show the favorable operating characteristics of the proposed design under certain scenarios.

Supplemental material, sj-zip-2-smm-10.1177_0962280220921553 for Dynamic prediction of competing risk events using landmark sub-distribution hazard model with multiple longitudinal biomarkers by Cai Wu, Liang Li and Ruosha Li in Statistical Methods in Medical Research

Supplemental material, sj-zip-2-smm-10.1177_0962280220921553 for Dynamic prediction of competing risk events using landmark sub-distribution hazard model with multiple longitudinal biomarkers by Cai Wu, Liang Li and Ruosha Li in Statistical Methods in Medical Research