Multi-Level Factors Underlying Racial/Ethnic Inequities in Clinical Trial Participation Among Patients with Hematologic Cancers: Lessons for the Development of Diversity Plans

Background: The US Food & Drug Administration announced that drug sponsors will be required to implement plans to improve racial/ethnic diversity in clinical trials starting in 2024. While historically marginalized racial/ethnic populations are less likely to participate in cancer trials, there is limited information regarding inequities in access to hematologic cancer trials. It is also unknown whether social determinants of health (SDOH) explain racial/ethnic inequities in trial participation. Here we examined the extent to which commonly used trial inclusion criteria and SDOH factors explain racial/ethnic inequities in clinical trial participation among patients with six hematologic cancers.

Methods: This retrospective study used the nationwide Flatiron Health electronic health record-derived de-identified database (2011-2022) of adult patients with multiple myeloma, non-Hodgkin's lymphoma (diffuse large B-cell, follicular, or mantle cell lymphoma) or leukemia (chronic lymphocytic or acute myeloid leukemia). Area-level SDOH ( e.g. predominant race/ethnicity - a measure of segregation) characterizing where patients lived were assessed using the American Community Survey census tract data. Patients were followed from initial diagnosis to clinical study drug receipt (a proxy for trial participation), death, or last recorded activity. Associations of race/ethnicity and SDOH factors with trial participation were assessed using cause-specific hazards models adjusted for clinical confounders (diagnosis year, age, sex, and cancer type). To determine which commonly used trial inclusion criteria ( i.e. abnormal creatinine serum [≥ 2] and performance status) and SDOH factors underlie racial/ethnic inequities, mediation analysis was performed using nonlinear multiple additive regression tree models.

Results: This study included 51,382 patients (median age: 67 years, 42.2% women; 57.7% non-Latinx [NL] White, 6.9% NL Black, 6.1% Latinx, 29.4% other/unknown). Black and Latinx patients were more likely to live in marginalized areas ( e.g. segregated neighborhoods color and higher percentages of limited English proficiency and low vehicle ownership) than White patients. Overall, 5.6% of patients participated in trials. Patients who received care from community practices or lived in marginalized neighborhoods were less likely to participate in trials. Black (4.2%; hazard ratio [HR] 0.62, 95% confidence interval [CI]: 0.52- 0.73) and Latinx patients (3.7%; HR 0.53, CI: 0.44-0.65) were less likely to participate in trials than White patients (6.4%). Black-White trial inequities were mediated (77%, CI: 42%-117%) by practice type ( e.g., community vs. academic center, 19%, CI: 9%-28%) and area-level SDOH such as segregation (49%, CI: 21%-77%) and vehicle ownership (12%, CI: 1%-24%). Mediators explained 55% (CI: 23%-87%) of the Latinx-White trial participation inequity; area-level SDOH-including segregation (30%, CI: 3%-58%), limited English proficiency (21%, CI: -3% to 45%), and vehicle ownership (17%, CI: 0%-45%)-were the most important mediators. Clinical factors did not substantially mediate racial/ethnic inequities (< 3% proportion mediated).

Conclusion: Black and Latinx patients diagnosed with hematologic cancers are 40% to 50% less likely to participate in clinical trials than their White counterparts. SDOH factors explained most of these inequities. These findings suggest that efforts to increase racial/ethnic diversity in cancer trials must address the persistent effects of segregation, along with everyday obstacles to clinical care, such as transportation and language barriers. Without concerted efforts to increase trial participation among historically marginalized racial/ethnic groups, trials may perpetuate inequities in investigational drug access and adversely impact the generalizability of pivotal cancer treatment trials.