Chimeric Antigen Receptor T-Cell Access in Patients with Relapsed/Refractory Large B-Cell Lymphoma: Association of Access with Social Determinants of Health and Travel Time to Treatment Centers

Background Large B-cell lymphoma (LBCL) is the most common type of non-Hodgkin lymphoma. Chimeric antigen receptor T-cell (CAR T) therapy represents a novel treatment with curative potential for relapsed or refractory (R/R) LBCL, but there are access barriers to this innovative therapy that are not well-studied. Objectives (1) Assess the impact of geographic factors and social determinants of health (SDOH) on access to treatment with CAR T in a sample of patients with R/R LBCL and ≥ 2 prior lines of treatment (LOT). (2) Compare and contrast patient characteristics, SDOH, and travel time between patients with R/R LBCL who received CAR T and those who did not. Study Design An observational, nested case-control study of patients with R/R LBCL, ≥ 2 prior LOT, not in a clinical trial, identified using 100% Medicare Fee-For-Service and national multi-payer claims databases. Patients were linked to near-neighborhood SDOH using 9-digit ZIP-code address. Driving distance and time between residence and nearest CAR T treatment center (TC) was calculated. Patients were stratified based on treatments received upon 3rd LOT initiation (Index Date) or later: 1) received CAR T and 2) did not receive CAR T. Multivariable logistic regression was used to evaluate factors associated with CAR T. Results 5,011 patients met inclusion criteria, with 628 (12.5%) in the CAR T group.Regression models found the likelihood of receiving CAR T decreased with patient age (OR=0.96, p<.001), and males were 29% more likely to receive CAR T (OR=1.29, p=.02). Likelihood of CAR T increased with Charlson Comorbidity Index (CCI; OR=1.07, P<.001) indicating patients with more comorbidities were more likely to receive CAR T. Black patients were less than half as likely to receive CAR T than White patients (OR=0.44, p=.01). Asian patients did not significantly differ from White patients (OR=1.43, p=.24), and there was a trend for Hispanic patients to have a slightly lower likelihood of CAR T (OR=0.50, p=0.07). Higher household income was associated with receipt of CAR T, with the lowest income group more than 50% less likely to receive CAR T than the highest (OR=0.44, p=.002), and the second lowest income group more than 30% less likely (OR=0.68, p=.02). Finally, likelihood of CAR T therapy was reduced when the driving time to the nearest TC was 121 to 240 minutes (reference group: ≤ 30 minutes; OR=0.64, p=.04). Travel times between 31 and 121 or greater than 240 minutes were not significantly different from ≤ 30 minutes. Payer type was collinear with age and could not be included in the regression analysis, but patients with commercial insurance were 1.5-3 times more likely to receive CAR T than other payers on an unadjusted basis. Conclusions We identified significant disparities in access to CAR-T related to demographics and SDOH. Patients who were older, female, low income, or Black were less likely to receive CAR-T. The positive association of CCI with CAR T requires further research. Given the promising outcomes of CAR T, there is urgent need to address identified disparities and increase efforts to overcome access barriers.