Data transparency and demographic representation in approval packages of FDA-approved oncology drugs from 2012-2021.

person Jingyi Liu Brigham and Women's Hospital, Boston, MA info_outline Jingyi Liu, Gowri Jayaram, Tanvee Varma, Stephen Sammut, Carmen Guerra, Jennifer E. Miller

Authors person Jingyi Liu Brigham and Women's Hospital, Boston, MA info_outline Jingyi Liu, Gowri Jayaram, Tanvee Varma, Stephen Sammut, Carmen Guerra, Jennifer E. Miller Organizations Brigham and Women's Hospital, Boston, MA, Medidata, New York City, NY, Yale School of Medicine, New Haven, CT, Wharton School of Business, Philadelphia, PA, University of Pennsylvania School of Medicine, Philadelphia, PA Abstract Disclosures Research Funding No funding received None. Background: Over the past decade, the U.S. Food and Drug Administration (FDA) has leveraged regulations and guidance documents to accelerate improvements in demographic reporting and encourage adequate representation of minorities in clinical trials. This cross-sectional study aims to assess the impact of these interventions by characterizing demographic data reporting and representation by sex, age, and racial and ethnic identity in pivotal trials supporting novel cancer therapeutics approved by the FDA from 2012 through 2021. Methods: Study demographic data were abstracted from FDA drug approval packages and US cancer population demographic data was abstracted from US Cancer Statistics. The percentages of trials reporting sex, age, racial and ethnic identity of participants in drug approval packages were calculated and participation to prevalence ratios (PPRs) were calculated by dividing the percentage of participants in each demographic by the percentage of the US cancer population in each group. PPRs were constructed for each pivotal trial by cancer type. Adequate representation was defined as PPR 0.8-1.2 and the proportion of trials which were adequately representative of each demographic was calculated and trended over time. Results: From 2012 through 2021, the FDA approved 118 cancer therapeutics, based on 131 pivotal trials. For drugs approved 2012-2016, the majority of associated pivotal trials reported participation based on sex (100% of trials), age (100%) and White (98%), Black (75%) or Asian (84%) racial identity, but not American Indian or Alaskan Native racial identity (18%) or Hispanic ethnic identity (41%). For drugs approved 2017-2021, there were no significant improvements in demographic data reporting, although there was a trend towards increased reporting on Hispanic ethnic identity from 2012-2016 (41%) to 2017-2021 (65%) (p= 0.06). For drugs approved 2012-2016, most pivotal trials adequately represented female participants (78% of trials) and participants identifying as White (87%), but few pivotal trials adequately represented older adults (7%), participants identifying as Black (11%), American Indian or Alaskan Native (0%), Asian (18%) or Hispanic (5%). Representation in pivotal trials did not significantly improve for any demographic for drugs approved 2017-2021 (p>0.05). Conclusions: This study found that the last decade’s efforts to improve demographic data transparency and representation in clinical trials have not yet borne fruit for novel cancer therapeutics. More consequential FDA regulations may be necessary. Beyond the FDA, demographic data transparency and adequate minority representation in oncology clinical trials is an all-stakeholder responsibility. A more just clinical trial infrastructure requires greater collaboration between the FDA, sponsors, enrollers, contract research organizations and cooperative groups.