Bayesian Adaptive Phase I/II Clinical Trial Design with Competing Risk Model in Personalized Medicine (306315)*Wenjing Hu, University of Notre Dame
Ick Hoon Jin, University of Notre Dame
Yong Zang, Indiana University
Keywords: Personalized Medicine, Competing Risk, Survival Model
We propose a Bayesian adaptive phase I/II clinical trial with competing risk endpoints. The proposed design is motivated by an ongoing lung cancer clinical trial, which treats patients at different levels of radiotherapy(RT). Patients are classified into the radiation-resistant(RE) subgroup and radiation-sensitive(SE) subgroup. The patients are randomized to receive the high-dose, standard or low-dose RT in these two subgroups. The RT on tumor cell can prevent the tumor progression but the RT on the normal cell can induce the normal tissue complication. The purpose of the trial is to study whether the RE patients perform better in the high-dose RT arm and whether the SE patients perform better in the low-dose RT arm, both compared with the standard RT.