(SLC-039) Rare Findings: Characterizing External Control Arm Studies Assessed by FDA for Drugs and Biologics in Rare Disease

Background: External control arms (ECAs) are increasingly utilized when randomized control trials are not ethical or feasible, particularly for rare diseases. While extensive information about oncology ECA studies is available, the design, methodology, and regulatory acceptance of ECA studies in rare diseases are less well characterized.

Objectives: To characterize ECA studies assessing drugs and biologics for rare diseases, focusing on their objectives, design, methodology, and regulatory feedback.

Methods: We identified drugs and biologics indicated in rare diseases that sponsors submitted ECA study results to the US FDA. Candidate drugs were found by reviewing the literature on ECA studies in non-oncology settings, including PubMed, real-world evidence use cases, and public FDA product review documents. Products with FDA approval from 2015 to 2025 were considered. Data were extracted on the treatment of interest, trial arm, ECA, and parameters related to the ECA study and regulatory review. Results were summarized descriptively.

Results: Twelve rare disease ECA studies were identified, each for a unique product and indication. Most studies (10) were comparative effectiveness research for single-arm trials. The most common ECA data source type was prospectively collected data (i.e., natural history, registries). Standard of care was the most frequent comparator, though details on its definition were often lacking. Six of the studies assessed continuous primary outcomes. Methods for controlling confounding included propensity score matching and weighting. Despite potential biases such as differences in baseline characteristics between the trial and ECA populations, unmeasured confounding, and outcome assessment differences, the FDA generally accepted results as supportive evidence.

Conclusions: Few rare disease ECA studies are described as supporting evidence in public-facing FDA review documents. The predominance of ECAs being derived from prospectively collected data contrasts with oncology, where robust real-world data sources like electronic health records (EHRs) and administrative and medical claims databases are frequently used. This difference may be due to the nature of rare diseases; patients with rare diseases are less represented in EHR and claims data compared to disease-specific sources like registries, and those who are may be difficult to identify due to phenotypic diversity within conditions and necessity of non-routinely collected data elements to ascertain patient eligibility. Continuous primary outcomes were frequently assessed, highlighting challenges in comparing study arms from different data sources due to potential differences in temporality and ascertainment methodology.