Menu 返回 Poster-Presentations-Details

P109: Evaluation of PMR/Cs for Assessing Data in Racial and Ethnic Populations Underrepresented in Premarket Clinical Trials

Poster Presenter

Grace Collins

Regulatory Policy Analyst
Friends of Cancer Research
United States

Objectives

Assess regulatory review documents to identify trends in approval and trial characteristics for drugs with and without PMR/Cs specifying the need for representative data based on race and ethnicity (R/E PMR/C).

Method

Compared approval and trial characteristics for novel oncology drugs approved 2012-2023 with and without a R/E PMR/C (descriptions containing the terms “representation” and/or “representative”) using data in regulatory documents available on Drugs@FDA and CBER’s web page of novel biologic approvals.

Results

Regulatory documents for the 144 novel oncology drugs approved 2012-2023 were reviewed and a total of 759 PMR/C descriptions were compiled. Over half of drugs approved 2021-2023 had a R/E PMR/C (21/40, 53%). Before 2021, only 1/104 drugs had a R/E PMR/C. The remainder of this analysis focuses on the 40 drugs approved 2021-2023. 58% (18/31) of approvals supported by a single-arm pivotal trial had a R/E PMR/C compared to 36% (4/11) of approvals supported by a randomized controlled trial (RCT). 76% (16/21) of accelerated approvals (AA) had a R/E PMR/C compared to 26% (5/19) of traditional approvals. Notably, all AA drugs were supported by single-arm trials. FDA issued R/E PMR/Cs to approvals for various cancer indications, including gynecologic cancers, non-small cell lung cancer (NSCLC), non-Hodgkin lymphoma (NHL), multiple myeloma, gastrointestinal cancers, breast cancer, and nasopharyngeal cancer, with the percentage of drugs receiving a R/E PMR/C for these indications ranging from 33% to 100%. Pivotal trial demographics for indications that had both drugs with (n=15) and without (n=8) a R/E PMR/C were evaluated. On average, White patients were the most represented in clinical trials: 75.3% (36.8-96%) of patients for drugs with a R/E PMR/C and 74.4% (34-95%) for drugs without a R/E PMR/C. For ethnicity, trials included mostly non-Hispanic patients who are non-Hispanic: 81.8% (15-99%) for drugs with R/E PMR/C and 88% (61-97%) for drugs without a R/E PMR/C. On average, pivotal trials for drugs with a R/E PMR/C, enrolled fewer U.S. patients (41% [12-73%])] than pivotal trials for drugs without a R/E PMR/C (66% [14.9-100%]). Safety subgroup analyses based on race and ethnicity for drugs with a R/E PMR/C indicated a potential safety signal more frequently than subgroup analyses for drugs without a R/E PMR/C; though, in several instances, there were safety signals observed for drugs without a R/E PMR/C.

Conclusion

FDA issued PMR/Cs for most novel oncology drugs approved in the last decade. In 2020, the FDA started to incorporate language in certain PMR/Cs specifying the need for improved representation based on race and ethnicity. Our assessment of pivotal trial demographics is consistent with previous analyses that have shown clinical trial populations are under representative of diverse races and ethnicities. Despite this, we found that not all approvals supported by trials lacking representation of diverse races and ethnicities were issued a PMR/C specifying the need for greater representation based on race and ethnicity. Single-arm trial design, accelerated approval, and a higher percentage of patients enrolled outside of the U.S. may factor into the FDA’s decisions to issue a R/E PMR/C. In addition, our review of subgroup analyses indicates that potential safety signals between racial and ethnic subgroups may also factor into the decision to issue an R/E PMR/C. While stakeholders throughout the healthcare system work to implement strategies to support an equitable healthcare system and clinical trial ecosystem, the FDA is leveraging PMR/Cs to further characterize the optimal use of drugs in patient populations underrepresented in clinical trials. Identifying factors that may influence whether a R/E PMR/C will be issued can inform more optimal trial designs and help to ensure clinical trials provide information to characterize use in a diverse and representative patient population. In addition, it may be necessary to ensure clinical trials are sufficiently powered to evaluate differences in safety and efficacy, especially when known disparities are present.