Abstract PS4-11-25: Evaluating Underrepresentation in Breast Cancer Clinical Trial Enrollment at Yale Cancer Center: A Retrospective Demographic Study

Abstract Background: As therapeutic advances continue to reshape the standard of care, balanced representation in breast cancer clinical trials is essential to ensure that novel treatments benefit all populations. Yet, racial and ethnic minorities, older adults, non-English speakers (NES), and individuals with lower socioeconomic status remain underrepresented in trial cohorts. To address these imbalances, our team developed an automated Clinical Trial Patient Matching (CTPM) system powered by machine learning to identify eligible patients using real-time clinical data. To quantify existing gaps and assess outcomes of CTPM, we conducted a benchmarking analysis comparing the demographic characteristics of breast cancer clinical trial participants at Yale Cancer Center (YCC) over the past 10 years, before, and after CTPM implementation. Methods: Demographic data for breast cancer patients seen at YCC between 2013 and 2024 were obtained from three sources: the Epic electronic health record via the Yale New Haven Hospital (YNHH) Computational Health Platform, the Clinical Trials Management System (OnCore), and the CTPM system, which provides lists of trial-eligible patients identified using machine learning algorithms. We conducted a retrospective benchmarking analysis comparing the general breast cancer population, CTPM-identified eligible patients, and patients who consented to therapeutic clinical trials, as recorded in OnCore. Comparisons were stratified by time period (pre- and post-CTPM implementation: before July 2022 and before July 2024, respectively) and focused on age (18-39 and 70 years), race, ethnicity, primary language, geographic location, and insurance status. Results: Among 45,380 breast cancer patients treated between 2013 and 2024, baseline demographics included 1.89% young adults (18-39 years), 52.9% older adults (70 years), 19.8% racial/ethnic minorities, 4.7% non-English speakers (NES), 3.2% Medicaid-insured, and 8.0% rural residents. Between 2022 and 2024, demographics shifted with a decrease in older adults (32.1%) and an increase among young adults (4.8%), racial/ethnic minorities (25.6%), NES patients (6.6%), Medicaid-insured (4.1%), and rural residents (8.5%). Among patients enrolled via manual chart review, 1.98% were aged 18-39, 36.1% were 70 years, 21.6% were racial/ethnic minorities, 4.6% Medicaid-insured, 5.8% NES, and 9.2% rural residents. Following implementation of the CTPM system across four breast cancer clinical trials, 662 patients were pre-screened and deemed eligible, and 99 patients were consented. Trial participation increased for young adults ages 18-39 (10.7%), racial/ethnic minorities (28.1%), and Medicaid-insured patients (7.2%). Conclusions: Manual chart review, while useful, does not provide a comprehensive analysis of breast cancer patients and obscures recent demographic changes. In contrast, the automated CTPM system identifies more patients eligible for clinical trials and reflects population trends. We successfully identified additional patients who met eligibility criteria from underrepresented groups such as young adults, racial/ethnic minorities, and Medicaid-insured patients. However, most patients who were identified and confirmed as eligible did not consent or enroll. These findings highlight the importance of combining automated patient identification with novel strategies to overcome barriers to consent and improve representative participation in breast cancer clinical trials. Citation Format: G. Gong, J. Liu, M. Taylor, S. Pandya, C. Taborda, J. Xie, J. Parikh, W. Wei, M. Stefanou, P. Kunz, N. Fischbach, T. Battaglia, L. Mendez, J. Gaddy, I. Krop, P. LoRusso, M. Lustberg, A. Silber. Evaluating Underrepresentation in Breast Cancer Clinical Trial Enrollment at Yale Cancer Center: A Retrospective Demographic Study abstract. In: Proceedings of the San Antonio Breast Cancer Symposium 2025; 2025 Dec 9-12; San Antonio, TX. Philadelphia (PA): AACR; Clin Cancer Res 2026;32(4 Suppl):Abstract nr PS4-11-25.