TITLE: Adjust Overall Survival in Randomized Clinical Trials with Treatment Switching
SPEAKER: Songzi Li, Hong Tian and Kaifeng Lu, BeiGene USA Inc.
MODERATOR: Jingjing Ye
Abstract:
⦁ Motivations
Treatment switching is commonly allowed in randomized clinical trials on novel interventions driven by ethical considerations. When control group patients switch to experimental arm and benefit from the experimental treatment, statistical inference on overall survival based on data according to the arms to which patients were randomized will be biased. The question of clinical interest “what is the overall survival benefit of treatment” cannot be addressed adequately without proper adjustment.
Rank preserving structural failure time model, inverse probability of censoring weights and two-stages were most widely used model to adjust overall survival for treatment switching. However, there is no available validated SAS macro/R package for those statistical models. In this course, we will introduce the traditional treatment adjustment models (RPFSTM, IPCW, and Two-stage) and walk through the SAS macro/R package with audience. Furthermore, we pick a few typical industry case studies which represent difference scenarios of K-M curve.
Course Structure
⦁ Overview existed statistical model and result to adjust overall survival (RPFSTM, IPCW, and Two-stage) (60 mins)
⦁ SAS/R program example (30 mins)
⦁ Break (15 mins)
⦁ Industry Case Study (60 mins)
⦁ K-M curves diverges then parallel in the tail
⦁ K-M curves diverges then cross in the tail
⦁ K-M curves overlaps across time
⦁ Q&A (15 mins)
Instructors’ Biography:
Dr. Songzi Li has over 7 years of experience in design and analysis of clinical trials. Currently he is an Associate Director of biostatistics at BeiGene, serving as a statistical lead for multiple clinical studies. Dr. Li’s research interest includes treatment switching, go-no-go, sequential parallel comparison design (non-oncology), and clustering analysis. He co-developed SAS macros and R functions for RPSFTM, IPCW and two-stage method.
Prior to joining BeiGene, Dr. Li spent 4 years at PPD, working at the studies in multiple therapeutic areas including oncology, inflammatory bowel disease, and neurology, which support submission and successful approvals. He received a doctorate degree in Statistics from Bowling Green State University.