(C-202) A Comparative Analysis of Premarket and Postmarket Safety Data Between Non-Oncology Externally Controlled and Internally Controlled Orphan Drug Approvals

Background: Evaluating safety in rare diseases is challenging due to limited premarketing data, study subject heterogeneity, and uncertainties regarding differentiating adverse reactions from natural history in most rare diseases.

Objectives: This study primarily investigates whether there are significant differences in postmarketing safety outcomes between non-oncology orphan drug approvals that were based on externally controlled trials (ECTs) and those that were based primarily on randomized, internally controlled trials (ICTs).

Methods: Non-oncology orphan new molecular entities (NMEs) approved by CDER between 2010 and 2022 were identified in the Data Analysis Search Host database and classified as subset 1 (SS1), approvals based solely on ECTs, or subset 2 (SS2), approvals based on at least one ICT. Postmarket variables were collected through 4/30/2024. Variables extracted from manual labeling review and internal FDA documents included number of subjects treated with study drug in pivotal studies, number of new labeled adverse reactions, number of supplement approval letters, time to first new labeled adverse reaction, number of 505(o)(3) postmarketing requirements (PMRs) at approval, sources of safety data triggering supplement approval letters, and affected labeling sections. Two authors independently extracted these data and validated their findings with LabelComp, an artificial intelligence tool.

Results: There were 18 SS1 approvals and 108 SS2 approvals between 2010 and 2022. The only statistically significant difference between the subsets was number of subjects treated with study drug in pivotal studies; 78.9 (95% CI: 0, 163.3) for SS1 and 185.2 (150.1, 220.3) for SS2 (p = 0.023). Number of new labeled adverse reactions for SS1 and SS2 were 2.2 (0.75, 3.7) and 2.6 (2.0, 3.2), respectively (p = 0.68). Number of supplement approval letters for SS1 and SS2 were 1.3 (0.53, 2.1) and 1.6 (1.3, 1.9), respectively (p = 0.54). Time to first new labeled adverse reaction for SS1 and SS2 were 804 (459, 1148) days and 716 (576, 857) days, respectively (p = 0.64). Number of 505(o)(3) PMRs issued at approval for SS1 and SS2 were 3.2 (1.8, 4.6) and 2.6 (2.0, 3.2), respectively (p = 0.46). The sources of safety data triggering supplement approval letters and the sections of the label changed exhibited comparable distributions between SS1 and SS2.

Conclusions: Our findings suggest there may not be a significant impact of type of premarket approval or number of subjects exposed to drug in pivotal trials on number of, and time to, new labeled adverse reactions in FDA approved non-oncology orphan drugs. Given the small ECT cohort and the evolving regulatory landscape in orphan drug development, further retrospective and prospective studies are warranted to validate these findings.