|
Abstract:
|
Creative yet complex oncology clinical trial designs have emerged in response to the need to rapidly evaluate novel targeted agents in multiple contexts. One such class of designs has been termed "basket trials", whereby treatment allocation is biomarker-driven rather than disease-driven. In these trials, investigators are essentially screening for specific subpopulations with a given somatic mutation that respond to treatment. Depending on previous regulatory approval in other disease indications, investigators may be inclined to expect broad efficacy across all baskets at the onset of a trial. Bayesian modeling is an appealing approach for a basket trial design to capitalize on the expected correlated efficacies between baskets and potentially improve power and efficiency, as compared to independent designs for each basket. Both designs using Bayesian hierarchical and mixture modeling have been proposed. In preliminary work, we have found in simple settings there is little performance improvement by introducing such modeling complexity of multiple mixtures. We present our findings from the investigation of potential gains of such complexities and when they are needed.
|
