PS31: Case Studies of Statistical Borrowing in Clinical Trials: From Design to Regulatory Approval

Incorporating external data into clinical trial designs has become an increasingly important strategy for improving efficiency and feasibility of modern clinical research. Leveraging information from historical trials or real-world evidence may reduce the number of participants required for new studies, improve precision of treatment effect estimates, and address recruitment challenges that make research in rare diseases impractical. However, to meet the regulatory requirements for approval and maintain the rigorous standards that make RCTs the "gold standard" of clinical research, it is vital that trial designers utilize statistical approaches that mitigate the potential biases that arise due to differences between external and current datasets. With the increased utilization of these designs, we now have examples of trials that have been designed, reviewed by regulatory agencies, conducted and completed, and resulted in regulatory approval. We propose a session to review two case studies of completed trials leading to approval, with an assessment of the benefits of the statistical borrowing and lessons learned for future trials. This session will feature three presentations focusing on applications of statistical borrowing in clinical trials that led to successful regulatory approval. The first two talks will showcase two real trials with a focus on the statistical borrowing approaches utilized, the interactions with FDA, and the impact of borrowing on the trial outcomes. The session will conclude with a presentation by an FDA representative who will provide a regulatory perspective on the two highlighted trials.