Use of real-world evidence in the U.S. Food and Drug Administration and the European Medicines Agency approvals of oncology products supported by single-arm trials (2022-2025).

INTRODUCTION: Regulatory interest in real-world evidence (RWE) to supplement single-arm trials (SATs) in oncology has increased, yet how the Food and Drug Administration (FDA) and European Medicines Agency (EMA) evaluate such evidence remains incompletely understood. OBJECTIVE: To comprehensively evaluate recent oncology approvals in which SATs were supported by RWE, characterizing regulatory feedback, methodological approaches, role of RWE in decision-making. METHODS: We reviewed FDA New Drug Applications and Biologics License Applications (NDAs/BLAs), and EMA Marketing Authorizations (MAAs) (January 2022-August 2025). Approvals were identified through publicly available FDA and EMA databases. Regulatory documents were screened using predefined RWE- and SAT-related keywords, with full-text review for data extraction. RESULTS: Among 883 original approvals (553 FDA, 330 EMA), 200 were oncology products (134 FDA, 66 EMA). Eighteen applications (7 FDA, 11 EMA) incorporated SATs supplemented by RWE, representing 15 unique products, predominantly for hematologic malignancies (53.3%). All 15 products included at least 1 RWE approach, and 8 used multiple approaches including historical clinical trial data and literature. Among 25 distinct RWE or observational approaches, 16 (64.0%) were retrospective cohort designs drawing on patient-level data from electronic health records (24.0%), disease (20.0%) or product registries (4.0%), or chart abstractions (28.0%). RWE primarily served as contextual/supportive evidence (only 2 use cases viewed as providing "substantial evidence"). Methodological limitations commonly included small sample sizes, population heterogeneity, missing or unmeasured covariates, non-aligned outcome assessments, post-hoc design, and non-contemporaneous treatments. CONCLUSIONS: RWE is increasingly used to support oncology SATs and holds promise as supportive or contextual evidence in regulatory decision-making.