Optimal Bayesian Adaptive Clinical Trial Design in Cancer Personalized Medicine

Clinical biomarkers play an important role in personalized medicine in cancer clinical trials. An adaptive trial design enables researchers to use treatment results observed from early patients to aid in treatment decisions of later patients. Combining personalized medicine and adaptive trial design will make big contribution to new drug development. However, there has not been previous results on Bayesian optimal adaptive trial designs in cancer clinical trials. We describe a biomarker-incorporated Bayesian adaptive trial design that maximizes the patient responses in the trial and in a total patient horizon . The trial design applies the idea of dynamic programming, and we derive closed-form solution at each step of the optimization process. We study the the effects of the biomarker and marker group proportions on the total utility and compare the optimal strategy with other adaptive trial designs.