Disparities in Clinical Trial Participation Among Medicare Beneficiaries with Hematologic Malignancies from 2006 to 2019: A SEER-Medicare Analysis

Introduction

Representative participation in clinical trials (CTs) has been identified as an important dimension of health equity, particularly for older adults. Prior studies have suggested a lack of representativeness amongst CTs for hematologic malignancies (HM), but these findings have relied on synthesized participation data from published trials, which lack comprehensive reporting of participant race and ethnicity. Further, the lack of individual data has limited the multi-level evaluation of patient demographic, geographical, and disease-related factors contributing to CTs for older adults with HM. The purpose of this study is to systematically examine CT disparities among older adults with HM who are Medicare beneficiaries using national population-based data.

Methods

This retrospective cohort study was conducted using the linked Surveillance, Epidemiology, and End Results (SEER)-Medicare database. We identified all patients aged ≥65 years with a new diagnosis of lymphoma (Hodgkin lymphoma [HL], non-Hodgkin lymphoma [NHL], chronic lymphocytic leukemia [CLL]), acute leukemias, chronic myeloid neoplasms, and multiple myeloma (MM) between 2006 and 2018 (end of follow-up December 2019). The main outcome was CT participation, defined as the presence of a Medicare claim for the delivery of CT services (ICD9/10 code V70.7/Z00.6, or either of HCPCS modifiers “Q0” and “Q1”). Cumulative incidence was used to estimate the incidence of CT participation while accounting for the competing risk of death using Fine-Gray subdistribution hazard model, reported as adjusted hazard ratios (aHRs) with their 95% confidence intervals (CIs). Covariates included patient factors (age, sex, income, education, race, ethnicity, distance from NCI center), treatment status, and comorbidities. Participation was assessed overall and by HM subtype.

Results

The study cohort (N=54,121) was 50% female with a median age of 78 years (y) (IQR 72-84). The cumulative incidence of unadjusted CT participation at 1, 2, and 5 y was 2.7% (95% CI 2.5-2.8%), 3.5% (95% CI 3.4-3.7%), and 4.3 (95% CI 4.1-4.4%), respectively. Overall, reduced odds of participation were observed in patients who were older (compared to 66-69 yrs: age 70-74, aHR 0.79, 95% CI 0.71-0.88; age 75-79, aHR 0.63, 95% CI 0.56-0.70; age 80-84, aHR 0.41, 95% CI 0.36-0.46; age 85+, aHR 0.21, 95% CI 0.17-0.24; p<0.001), female (aHR 0.79, 95% CI 0.74-0.86, p<0.001), Black (compared to White, aHR 0.73, 95% CI 0.59-0.90, p<0.001), had greater comorbidities (cardiovascular aHR 0.85, 95% CI 0.72-0.98, p=0.025; pulmonary aHR 0.76, 95% CI 0.68-0.85, p<0.001; renal aHR 0.67, 95% CI 0.59-0.76, p<0.001), and lived ≥ 12.5 km from the nearest NCI center (compared to < 12.5km: 12.5-49 km, aHR 0.89, 95% CI 0.81-0.98, p=0.02; 50-249 km, aHR 0.84, 95% CI 0.74-0.95, p=0.004; 250+ km, aHR 0.64, 95% CI 0.48-0.86, p=0.003) . Treatment received within 1-year of diagnosis was associated with CT participation (aHR 2.2, 95% CI 1.96-2.42, p<0.001).

Across other HM subtypes, consistent patterns of disparities were observed for age, sex, and comorbidities, but not for race and ethnicity or geography. Compared to white participants, Black individuals showed lower participation in indolent NHL (aHR 0.22, 95% CI 0.07-0.70, p=0.010), acute leukemias (aHR 0.57, 95% CI 0.33-0.99, p=0.044) and MM (aHR 0.54, 95% CI 0.36-0.81, p=0.003) trials, but not other subtypes. Distance to NCI centers was associated with lower odds of participation among patients with lymphoid malignancies (compared to < 12.5km: indolent NHL 50-249 km aHR 0.72, 95% CI 0.53-0.99, p=0.040; aggressive NHL 50-249 km aHR 0.74, 95% CI 0.61-0.91, p=0.005; 12.5 – 49 km aHR 0.79, 95% CI 0.62-0.99, p=0.047), but not others.

Conclusions

Amongst older adults with HM malignancies, significant sociodemographic underrepresentation was observed amongst CT participants compared to the population affected, and these disparities persisted across the duration of illness. These disparities compromise the generalizability of trial results and reflect reduced access to trials for patients from vulnerable populations. These data identify several targets for future research and intervention to improve equitable access to innovative therapies across diverse patient demographics. Differences in the patterns of underrepresentation HM subtypes suggest that a disease-specific approach to addressing these disparities may be needed.