题目：Randomized Phase Ib Decision Making in a Bayesian Framework
摘要：Confronted with the increasing complexities in drug development and driven by the continuous desire to increase research and development productivity, there is a general trend of pharmaceutical companies moving forward to more innovative drug development approaches of futility-rule based and accelerated approval (AA) filing strategy integrated seamless early and late development, to achieve the goal of quick-win and fast-fail. In this background, I will introduce a phase 1b/3 adaptive framework with evaluating the efficacy and safety of a New Molecular Entity (NME) in oncology setting. The framework incorporates an adaptive two-stage design with an regimen selection and futility-rule based efficacy gating analysis performed in Stage 1 (Phase 1b), and includes in Stage 2 (Phase 3) a randomized comparison of the investigated regimen selected in Stage 1 versus standard of care treatment. The regimen selection and efficacy gating thresholds of stage 1 were defined upon Bayesian posterior probabilities simulated through a Dirichlet-Multinomial Bayesian model in order to bridge historical information and adapt to co-primary endpoints of response rates. In this talk, I will describe the adaptive design philosophy as well as key statistical challenges in such drug development model.
报告人简介：Dr. Fan Xia currently works in BeiGene and supports the company PD-1 molecule Tislelizumab global and China development in lung and bladder cancer, including both early phase and registrational studies. Before joining BeiGene, Dr. Xia worked in Roche and supported global MRCTs, China/Asia Pacific regional studies, as well as clinical development plans of multiple molecules. Dr. Xia received her Ph.D. from statistics from the University of Hong Kong and Bachelor degree in statistics from Renmin University of China.