Disparities in CAR-T Therapy Access and Outcomes: Variation in Referral Patterns for CAR-T Therapy in Patients with Newly Diagnosed Malignancies at an Academic Center in North Carolina

Introduction: Chimeric Antigen Receptor (CAR)- T cell therapy has revolutionized the treatment of hematologic malignancies. However, access to CAR-T therapy remains limited, especially for underserved populations. We hypothesize that disparities in age, gender, race, and insurance type affect access and progression through key stages of CAR-T therapy—referral, evaluation, and infusion of CAR- T cells. These disparities contribute to inequities in the utilization of CAR-T therapy.

Method: This retrospective study analyzed data from July 2019 to June 2022, incorporating records from the North Carolina Central Cancer Registry (NCCCR) and Duke Cancer Institute. The study focused on adults aged 19-79 with lymphoma or multiple myeloma residing in 67 North Carolina counties, which are designated as Duke's primary catchment area. Cohort 1 comprised NCCCR-Eligible patients—those identified in the NCCCR dataset as eligible for CAR-T infusion. Cohort 2 included Duke Referred patients, further divided into: 2A) Referral Only—patients who were referred but did not receive an appointment; 2B) Appointment without infusion—patients who had an appointment but did not receive CAR-T therapy; and 2C) CAR-T —patients who underwent CAR-T infusion. Statistical analyses included chi-square tests, ANOVA or Wilcoxon Signed Rank tests, and logistic regression.

Referrals to other centers within the DCI catchment and out-of-state counties were excluded from the study.

Results:

NCCCR eligible vs. Duke Referred (Cohort 1 vs 2): Comparing the NCCCR cohort with CAR-T candidates referred to Duke, Duke referrals are younger median 64 [22-79] versus 66 years [19-79] in NCCCR, p = 0.004) and with fewer Caucasian/White patients (60% DCI vs. 71% NCCCR) and more Black or African American (30% DCI vs. 24% NCCCR). Private insurance is higher at DCI (50% vs. 31% NCCCR). Both cohorts are predominantly in metropolitan areas (81% DCI vs. 78% NCCCR, p < 0.001) and Medicare was less common than in NCCCR (40%, vs. 53%, p<0.001). DCI has a higher median income ($67,000 vs. $58,646, p < 0.001) and a lower social vulnerability index (0.61 vs. 0.66, p < 0.001).

Referral Only vs Appointment without Infusion (Cohort 2A vs. 2B): Analysis of the referral-only cohort versus those who attended an initial evaluation appointment for CAR-T revealed significant sociodemographic differences. African Americans were more prevalent among the referral-only group (37% vs. 30%, p = 0.004) and had lower odds of attending initial evaluation appointments (OR = 0.55, 95% CI [0.35-0.87], p = 0.010). Univariate analysis indicated that a lower proportion of the referral-only group had private insurance (44% vs. 49%, p = 0.007) and Medicare coverage (40% vs. 44%, p = 0.007). However, these differences did not persist in multivariate analysis (OR = 0.78, 95% CI [0.49-1.23], p = 0.3 for Medicare). Non-private, non-Medicare insurance showed a trend towards lower attendance at initial appointments (OR = 0.35, 95% CI [0.14-1.03], p = 0.040), but this trend was not significant, likely due to the small sample size. The median time to appointment was 20 days (range 1-253). Age, sex, RUCA classification, income, and social vulnerability index did not show significant effects (p > 0.1).

Appointment without Infusion vs CAR-T (Cohort 2B vs. 2C): Comparison of patients evaluated without infusion versus CAR-T recipients revealed higher odds of proceeding to CAR-T for males (OR = 2.68, 95% CI [1.63–4.56], p < 0.001) and Black or African American patients (OR = 1.87, 95% CI [1.06–3.23], p = 0.027). The median time to CAR-T was 141 days (range 41-1,248). No significant differences were observed in age, insurance type, RUCA classification, income, or social vulnerability index.

Conclusion: Barriers to CAR-T therapy access at Duke appear to vary by stage. Initially, African American race and non-private insurance are linked to lower attendance at evaluation appointments post-referral. After evaluation, female gender and African American race are associated with a lower likelihood of proceeding to CAR-T infusion, highlighting significant gender and racial disparities. Further research is needed to determine if these disparities are due to demographic, socioeconomic, or healthcare system factors. Tailored interventions and collaborations with community partners and patient navigators are crucial for addressing these disparities and ensuring equitable access to CAR-T therapy.