OBJECTIVES: To examine how real-world evidence (RWE) is used in cell and gene therapy (C&GT) regulatory and health technology assessment (HTA) submissions/decisions in and across jurisdictions.

METHODS: FDA multi-disciplinary reviews for C&GTs (excluding cord blood treatments) from 2017-2021 were screened to determine if RWE was included or planned, the type of real-world data (RWD) used, and the role of the RWE in the approval package. For two case studies (onasemnogene abeparvovec and idecabtagene vicleucel), approval documentation from additional regulatory (EMA) and HTA agencies (NICE, ICER, and HAS) were also reviewed. The use and impact of RWE in these two submissions were compared.

RESULTS: Ten C&GTs, 9 of which included RWE in the application, were initially approved by the FDA during 2017-2021. Uses ranged from providing substantial (2) or supportive (5) evidence of effectiveness to providing real-world context around the therapy/indication (1) and planned fulfillment of post-marketing requirements (3; categories are not mutually exclusive). In their FDA applications, onasemnogene abeparvovec and idecabtagene vicleucel both utilized RWD-based external control arms. For onasemnogene abeparvovec, RWE influenced the decision for the FDA, EMA, and HAS submissions; however, NICE and ICER noted challenges with the RWE regarding variable natural history across patients and small patient numbers. For idecabtagene vicleucel, FDA, EMA, HAS, and ICER noted uncontrolled confounding, data loss, and inconsistent standard of care over time; despite these limitations, EMA found that the RWE helped to contextualize the trial results.

CONCLUSIONS: RWE plays a strong role in C&GT approvals by the FDA. However, in the two case studies examined, there was variability among the regulators and HTA agencies in the level of support that RWE contributed to their decisions. These use cases thus provide important learnings for sponsors and agencies seeking to navigate the evolving field of using RWD to support regulatory decision-making.