Pharmacometric Model‐Based Sample Size Allocation for a Region of Interest in a Multi‐Regional Phase 2 Trial: A Case Study of an Anti‐Psoriatic Drug

Phase 2 trials have historically focused on characterizing the dose‐exposure‐response relationship in relatively homogeneous patient populations before proceeding to confirmatory trials. However, with the rise of multi‐regional Phase 2 trials, it is important to strike a balance between this goal and the requirement to make sure that the optimal doses are chosen for patients from various geographic areas. This study uses a dose‐ranging trial for an anti‐psoriatic drug, featuring a typical design with a total sample size of N = 175, to highlight key considerations regarding sample size in multi‐regional exploratory studies. The allocation of sample size to a region of interest (Region X) was evaluated using both a conventional statistical approach and a pharmacometric model‐based (PMx) approach, predicated on the assumption of a minimum treatment improvement in Region X. Further evaluation was performed to assess the probability of reaching reliable conclusions regarding clinically relevant inter‐regional differences in treatment response. The statistical approach, relying solely on end‐of‐trial observations from a single dose group, exhibited a maximum power of less than 40% in detecting treatment differences across regions when Region X accounts for 50% of the total sample size. In contrast, the PMx approach, employing data from multiple dose groups across trial duration, demonstrated that 26% of the total sample size yielded over 80% power to identify the inter‐regional difference. The PMx approach has also been shown to offer a more efficient characterization of the clinical relevance of inter‐regional differences, and has potential to improve decision‐making in development progression by integrating prior knowledge.