Real-World Comparison of Healthcare Costs and Resource Utilization Among Patients with Relapsed-Refractory Large B-Cell Lymphoma Treated with CAR T-Cell Therapy Versus Historical Standard-of-Care: A Cost-Consequence Analysis in Ontario, Canada

Background Chimeric Antigen Receptor T-cell therapy (CAR-T) has transformed the management of relapsed-refractory large B-cell lymphoma (RR LBCL) by offering the potential for long-term survival to patients who have exhausted other curative intent treatment. In 2019, the Canadian Agency for Drugs and Technologies in Health recommended funding CAR-T for patients with RR LBCL progressing after two or more lines of systemic therapy. However, CAR-T is resource-intensive to deliver and is associated with significant toxicity and health resource utilization. Real world data on healthcare spending associated with CAR-T is required to assist decision-makers and ensure its efficient and equitable delivery. We compared healthcare costs among a cohort of patients receiving CAR-T with a cohort of historical patients treated prior to CAR-T approval at Princess Margaret (PM) Cancer Centre in Toronto, Canada. Methods Using linked, institutional and population-based clinical and administrative databases in Ontario, Canada, patients with RR LBCL consecutively treated at PM with CAR-T (2020-2022) were compared to a historical PM control cohort of RR LBCL patients treated prior to CAR-T approval (2012-2017). Patients were followed from the date of progression following 2 lines of chemotherapy (2L) in the historical controls and date of progression after last therapy (2L or higher) prior to receiving CAR-T for up to 3-years, with maximum follow-up to March 31, 2023. Stabilized inverse probability of treatment weighting (sIPTW) was used to balance baseline covariates between the cohorts (age, sex, lactate dehydrogenase, and comorbidities). sIPTW-weighted Kaplan-Meier curves and Cox proportional hazard regression were used to estimate OS probability and hazard ratios (HR) for CAR-T patients versus historical controls. Generalized linear regression modelling was used to estimate total and resource-specific healthcare costs (2024 Canadian dollars), including mean 3-year incremental costs between the two arms (excluding CAR-T drug cost to focus on healthcare resource expenses [Axi-cel $485,021; Tisa-cel $450,000]). Costs were adjusted for censoring using inverse probability of censoring weighting (IPCW). Results Cohorts of 86 CAR-T-treated patients and 150 historical control patients were evaluated. After applying sIPTW, baseline variables were balanced between the two groups: mean age was 56 years and males comprised 61%. The 3-year OS probability was 57% (95% CI 39-71%) in the CAR-T group and 10% (95% CI 5-16%) in the historical control group, with an IPTW-adjusted HR of 0.22 (95% CI 0.15-0.33) for all-cause death. IPCW-adjusted 3-year mean [standard deviation (SD)] total healthcare cost per CAR-T patient was $141,870 [$125,251] vs $55,388 [$40,833] in historical controls (p<0.001), resulting in a mean 3-year incremental total healthcare cost of $86,482 (95% CI $64,422-108,542) in the CAR-T cohort. The majority of healthcare spending in both cohorts was incurred through inpatient days (42% of CAR-T total expenditure vs 44% in historical controls). CAR-T patients incurred a mean 3-year incremental inpatient cost of $34,720 [95% CI $22,625-46,816], which represented 40% of the incremental total healthcare cost for CAR-T patients. The CAR-T cohort also had significantly higher mean 3-year incremental healthcare resource costs than historical control patients in: ambulatory cancer clinics ($14,663 [95% CI $9,609-19,717]; 17% of incremental total healthcare cost), outpatient clinic visits ($11,782 [95% CI $9,122-14,442]; 14%), physician costs ($8718 [95% CI $5,913-11,523]; 10%), oral drug costs ($7,676 [95% CI $2,605-12,747]; 9%), and systemic chemotherapy drug costs ($4,729 [95% CI $1,435-8,022]; 5%). Conclusion In this real-world analysis, CAR-T therapy was associated with improved survival as well as higher healthcare costs compared to patients treated with historical standard-of-care therapies. Greater inpatient care needs amongst the CAR-T cohort were a significant contributor to the higher overall spending observed, followed by clinic, physician, and additional drug costs. These data provide important considerations for funders and decision-makers in determining the value of CAR-T therapy in Ontario. These cost parameters can also inform future economic modelling of CAR-T therapy.