Racial disparity in cancer clinical trials: An analysis of clinicaltrials.gov protocols for race-related language and description.

Abstract

e18545 Background: Despite high cancer mortality rates in Black and Hispanic patients, they remain underrepresented in clinical trials. We analyze clinicaltrials.gov trial protocols to understand how sponsors discuss race, ethnicity, or diversity-related keywords, and map the distribution of cancer trials in the USA in relation to cancer incidence rates among Black and Hispanic Americans. Our study highlights the potential of enhanced sponsor information and data-driven site selection to reduce racial disparities in clinical trials. Methods: The US oncology trials from NIH’s clinicaltrials.gov were searched for racial disparity-related keywords using Kognitic© NLP algorithms. Keywords such as "racial," "ethnicity," "minority," "African American,”, "Hispanic” etc. were included. Only Hispanic and African American races were considered. Data were verified, classified based on keywords found. Zip codes of recruiting clinical trial sites were mapped to high-cancer-incidence counties among Hispanics and Black Americans (sources: statecancerprofiles.cancer.gov and unitedstateszipcodes.org). Results: We found 8338 (21% of all US cancer trials) clinical trials addressing racial disparity in cancer, with 4.3% being industry-sponsored, 8.3% academic-industry collaborations, and 87.3% non-industry-sponsored. 50% of the trials were interventional and the other 50% were observational, with biopharma as the primary industry represented. In industry-sponsored trials, only 10.3% directly explore race-based outcome differences, with the remainder collecting demographic data or having patients self-report their race. After comparing the locations of cancer clinical trials currently underway in the United States to counties with the highest cancer incidence rates among Black and Hispanic Americans, we discovered that only a limited number of clinical trial sites are in those high-incidence counties. Only four out of the top ten counties with the highest cancer incidence rates for Blacks and one out of the top ten for Hispanics have clinical trials actively recruiting patients, with most trials in both populations being conducted by non-industry sponsors. Conclusions: Our findings indicate that a limited number of cancer clinical trials mention race and ethnicity in publicly available trial protocols, and even fewer industry trials include race-related information. Our site analysis shows that a major gap in access to trials in areas with high cancer incidence rates, particularly for populations with low socioeconomic status. More analysis is needed to understand the relationship between trials and demographic conditions in America. Providing details on plans to address racial disparities in recruitment, trial design, and site selection will promote inclusive trial design that are accurate assessment of how drugs work in various types of patients.