Using Model-Informed Drug Development to Support OS Endpoints

In a recent video interview with Applied Clinical Trials, Ananth Kadambi, VP of real-world evidence and modeling solutions at Certara, outlined the operational hurdles sponsors face as the FDA increasingly requires overall survival (OS) as the primary endpoint in oncology studies. He emphasized the importance of pre-specified safety monitoring, analytical subgroup planning, and strategies to minimize patient dropouts. This regulatory shift places equal weight on demonstrating efficacy and safety, with implications for post-approval monitoring and patient access. Kadambi noted that model-informed drug development and real-world evidence can support more efficient trial designs. For clinical operations teams, the focus will be on strengthening patient tracking and engaging data monitoring committees earlier in the process to balance trial efficiency with regulatory and payer demands in complex oncology indications.

ACT: Requiring OS data could significantly extend trial timelines. How can model-informed drug development (MIDD) help sponsors design more efficient trials while still aligning with FDA expectations?

Kadambi: Model informed drug development is already recognized by FDA as a viable tool to bring to the table for discussions with sponsors. I don't think it's anything too far off the agency's radar screen as well, but specifically in the context of our overall survival, there are various tools within the MIDD toolkit that allow for the prediction of not just efficacy endpoints, but also safety endpoints. We can use tools like USP to model out some of the potential safety endpoints in oncology studies and perhaps get an early indication of what the expectation would be. Those can be packaged as part of the data that are submitted via sponsors, particularly in the cases where they have a short-term study in indications that have a short natural history, those data will be an integral part of the story to satisfy the FDA and essentially mitigate the risk associated with the lack of having longer term data.