时 间：2023年5月26日（周五）10：00-11：00
地 点： 腾讯会议 651-711-605
题 目：DEMO: Bayesian Adaptive Dose Exploration-Monitoring-Optimization Design Based on Short, Intermediate, and Long-term Outcomes
报告人：Ruitao Lin 教授 The University of Texas MD Anderson Cancer Center
主持人：徐进 教授
主 办：统计学院，统计与数据科学前沿理论及应用教育部重点实验室
摘 要：
The advancements in targeted therapy or immunotherapy have challenged the conventional ``more-is-better'' paradigm in oncology. Due to the deadly nature of cancer, an optimal dose recommended from early-phase trials should ultimately possess a promising long-term survival profile. Conventional early-phase trial designs that rely on short-term toxicity or efficacy data may identify a dose with suboptimal long-term benefits. Moreover, many existing designs fail to account for short-term pharmacodynamics when making decisions. Following the guideline of Project Optimus, a generalized dose optimization procedure consisting of three seamlessly connected stages is proposed. Throughout the three stages, the proposed design employs various endpoints, ranging from short-term to long-term, to make informed decisions. In the first dose-exploration stage, short-term toxicity and pharmacodynamics data are utilized to timely screen out doses that are overly toxic or biologically inactive. In the subsequent stages, patients are randomized to admissible doses, which are further monitored based on intermediate outcomes such as toxicity and tumor responses. At the end of the trial, an optimal dose is determined through maximizing the restricted mean survival time. Results from simulation studies indicate that the proposed design can identify a desired dose that is favorable in terms of toxicity, biological activity, tumor response, and survival benefits. Sensitivity analyses indicate that the design is robust to changes in prior specifications and model misspecifications.
报告人简介：
Ruitao Lin, Ph.D., is a tenure-track Assistant Professor in the Department of Biostatistics at The University of Texas MD Anderson Cancer Center. He also holds an adjunct faculty position in the Department of Statistics at Rice University. Motivated by the unmet need for the development of precision medicine, Dr. Lin has developed many innovative statistical designs to increase trial efficiency, optimize healthcare decisions, and expedite drug development. He is the leading author of several generalized BOIN or other model-assisted designs for drug-combination trials, trials with late-onset outcomes, and dose optimization trials. His innovative designs have been widely applied in trials conducted in the US and overseas. Dr. Lin has published over 50 papers in top statistical and medical journals. He is currently an Associate Editor of Biometrical Journal, Pharmaceutical Statistics, and Contemporary Clinical Trials. He is the co-PI (with Dr. Peter Thall) of the recently NCI-funded R01 project titled Bayesian Methods for Complex Precision Biotherapy Trials in Oncology.