Analysis of Supportive Evidence for US Food and Drug Administration Approvals of Novel Drugs in 2020

Abstract

Importance: In recent years, drug approvals have been based on fewer, smaller, and less rigorous pivotal trials. Less robust preapproval testing raises questions about the efficacy and clinical value of these drugs. Objective: To assess the regulatory context, pivotal design characteristics, and postmarket requirements (PMRs) and postmarket commitments (PMCs) of novel 2020 drug approvals to characterize the state of evidence at the time of approval. Design, Setting, and Participants: This cohort study identified novel drugs approved by the US Food and Drug Administration's (FDA) Center for Drug Evaluation and Research in 2020. The Drugs@FDA database was used to extract key characteristics of each drug's pivotal trials. Drug approval packages provided regulatory information. The prevalence of key trial design features was compared between oncology and nononcology drugs. Exposures: Drug names, date of approval, indication on labeling, and clinical and regulatory details. Main Outcomes and Measures: Number of pivotal trials, pivotal trial design (randomization, masking, groups), trial comparator, trial hypothesis, trial end points, results, number and type of expedited pathway designations, and number and type of PMRs and PMCs. Results: The 49 novel therapeutics approved in 2020 were supported by 75 pivotal trials. More than half of drugs (28 [57.1%]) were supported by a single pivotal trial. Trial sizes ranged from 19 to 2230 participants. More than three-fourths of trials (57 [76.0%]) had a randomization component, and nearly two-thirds (46 [61.3%]) were double-masked. Most used a superiority approach. Roughly half (39 [52.0%]) compared the novel therapeutic with a placebo or vehicle control; 13 (17.3%), an active control; 2 (2.7%), both a placebo and active control; and 21 (28.0%), a historical, external, or other control. Nearly half of pivotal trials (34 [45.3%]) used a surrogate measure as a primary end point. Pivotal trials supporting oncology approvals were much more likely to have historical controls than nononcology approvals (13 of 18 [72.2%] vs 8 of 57 [14.0%]; P