PhD Student in Epidemiology UNC- Chapel Hill NC Chapel Hill, United States
Background: Evidentiary standards for drug approvals have traditionally relied on a well-controlled study and confirmatory evidence. Yet, single-arm clinical trials (SAT) have become increasingly common for rare or life-threatening conditions. External control arms (ECAs) may improve interpretation of these trials but pose unique methodological and regulatory challenges.
Objectives: Through literature reviews and interviews across disciplines in drug development, we developed a framework for considerations in SAT strategies and ECAs that may affect likelihood of regulatory success. As a preliminary application, we categorized recent FDA approvals of SATs with ECAs to determine alignment with this framework.
Methods: We examined FDA approvals cited in two recently published reviews. Submissions after 2015 which utilized ECA data from RWD or published literature were included. Those using baseline controls or past trial data were excluded. We documented whether ECA data was included in product labeling and screened FDA review documentation for key framework considerations.
Results: The SAT framework identified key considerations that shift regulatory acceptance to more likely. These considerations include rare disease and oncology indications; gene therapies; unmet medical needs; and labeling expansions. For ECAs, factors were equipoise to trial population, appropriate index date selection, and patient-level data availability. In the framework’s initial application, 19 approvals were examined, the majority (N=16) of which were a first indication. All were for oncology, rare disease, or gene therapy products. FDA reviewers frequently cited inadequate comparability between trial and ECA (N=12); specifically: trial and ECA insufficiently balanced on baseline characteristics (N=6), ECA and trial data not generated among geographically similar populations or practice settings (N=3), outcome measurement inconsistencies between arms (N=2), and ECA data not contemporaneous to trial (N=1). ECA data were most often included in product labeling when in equipoise with the trial population (N=6). This aligns with the proposed framework and recent FDA 2023 guidance on ECA’s. However, some labels (N=4) included ECA data despite methodological concerns raised by the statistical reviewer. Further examination of submissions of SATs with ECAs is planned for EMA and Canada, which are expected to follow similar patterns.
Conclusions: Recent FDA approvals demonstrate the likelihood of regulatory success for SATs with ECAs depends on many design, analytic, and data quality considerations. Our framework is useful in early drug development when considering a single-arm trial strategies for evidence generation.
