|
Abstract:
|
Multi-Regional clinical trials (MRCTs) play a crucial role in global new drug development. They include patients from different regions within a single trial under a single protocol, which greatly facilitate the new drug registration approval in several countries simultaneously. However, despite of the benefits, MRCTs bring up new challenge especially when evaluating the consistency of treatment effect across regions. There are no generally acceptable approaches for consistency evaluation in existing regulatory guidance. In this simulation study, we first focused on evaluating a hazard ratio (HR) categories approach on the level of consistency between regional and the global results. Similar patterns were found for the four HR categories under varying hazard rate of control group, power of the study, and hazard ratio. We then compared a Bayesian approach with the Japanese approach recommended by PMDA. In general, Bayesian approach with mean posterior probability produced a higher rate of consistency results than Japanese approach in different scenarios. It was difficult to show consistency between regional and global results when the regional sample size was very small (e.g < 5%).
|
