Evaluating factors contributing to low rates of breast cancer clinical trial accrual in a diverse patient population.

Abstract

100 Background: In an ideal world, the populations studied in cancer clinical trials (CCT) would be representative of the patients seen in clinic. Unfortunately, significant disparities exist in trial enrollment. Patients who are white, male, insured, or of high socioeconomic status (SES) are overrepresented in NCI-sponsored CCT. Despite data indicating equal willingness for participation in CCT across all racial groups, lack of access, cultural barriers, and social determinants of health contribute to poor accrual rates among racial and ethnic minority patients. The Dan L. Duncan Comprehensive Cancer Center (DLDCCC) provides equal access to breast CCT at Smith Clinic (SC) within the safety-net Harris Health system and Baylor St. Luke’s Medical Center (BSLMC). The patient populations differ greatly at these two sites, with BSLMC serving > 95% insured, largely Caucasian patients, and SC serving 60% uninsured, mostly low SES patients, with > 80% racial and ethnic minorities. Despite equal access, patients at SC have a significantly higher CCT refusal rate. Methods: We performed a retrospective review of a prospectively maintained database of new patients seen at DLDCCC dating from 5/2015 to 9/2021, which included 3043 patients screened for breast CCT. 366 patients were found to be eligible for CCT. Some patients were eligible for multiple CCT, so there were 431 total offers of CCT. We performed logistic regression to evaluate whether differences in age, clinic, race, trial type, and primary language may be underlying the observed differences in CCT enrollment rates. Results: In the BSLMC cohort, 61% (116/204) of eligible patients enrolled in a CCT, while in the SC cohort only 39% (74/227) of eligible patients elected to enroll in CCT. This difference was significant on univariate but not multivariate analysis. There were significant differences when comparing race and trial type in the overall patient set. On univariate analysis, SC patients, African American (AA) patients, Hispanic/Latino patients, and Spanish speaking patients were significantly more likely to decline CCT participation. However, on multivariate analysis, only the AA patient category was associated with enrollment refusal (odds ratio 0.261, 95% CI 0.116-0.563, p < 0.001). On both univariate and multivariate analyses, patients were significantly more likely to accept biobanking trials (multivariate: odds ratio 12.799, 95% CI 3.777-61.403, p < 0.001). Conclusions: Based on these findings, it is likely an oversimplification to assume that equal access will lead to a complete elimination of CCT disparities. Our AA patients were significantly less likely to agree to participate in clinical trials, challenging the commonly held view that lack of access is a major barrier. We are exploring interventions designed to improve our AA patient population’s views of trial enrollment.