OBJECTIVES: To examine FDA approvals with Real-time Oncology Drug Review (RTOR), characterize their supporting evidence, and determine, when applicable, whether postmarketing studies were required to confirm clinical efficacy.

METHODS: We identified oncology approvals that used RTOR between 2018-2023 from Drugs@FDA database and determined approval type and pathway, first-in-class status, and study design, efficacy endpoints, and sample size of pivotal trials supporting approval. For indications approved based on trials using only surrogate marker endpoints, we determined whether postmarketing studies were required to confirm clinical efficacy.

RESULTS: FDA approved 363 new oncology indications between 2018-2023, of which 76 (21%), including 22 original and 54 supplemental approvals, underwent RTOR based on 84 pivotal trials. Nine (41%) RTOR original approvals were first-in-class.

Among RTOR approvals, 15 (20%) received accelerated and 61 (80%) traditional approval. All accelerated approvals were based on pivotal trials using surrogate markers as primary endpoints and had required postmarketing studies confirming efficacy. Among the 61 traditional approvals, 21 (34%) were approved based on pivotal trials using clinical outcomes for at least one primary endpoint, 7 (11%) for at least one secondary endpoint, while 33 (54%) used only surrogate markers as primary and secondary endpoints. Among those approved using surrogate endpoints, 1 (3%) approval had required postmarketing studies confirming efficacy.

Of the 84 pivotal trials, 16 (19%) supported accelerated and 68 (81%) supported traditional approvals. Pivotal trials supporting accelerated approvals were less likely to include placebo or active comparators (4 [25%] vs 58 [85%], P<0.001) and enrolled fewer patients (median [IQR], 108 [88-131] vs 612 [419-759], P<0.001) than those supporting traditional approvals.

CONCLUSIONS: One-fifth of new FDA oncology indication approvals were reviewed under the RTOR program. These approvals were based on pivotal trials with less robust designs and were not accompanied by requirements to confirm clinical benefit, potentially contributing to clinical uncertainty among patients and clinicians.