Cost and Healthcare Utilization in Relapsed/Refractory Diffuse Large B-Cell Lymphoma: A Real-World Analysis of Medicare Beneficiaries Receiving Chimeric Antigen Receptor T-Cell Vs. Autologous and Allogeneic Hematopoietic Cell Transplants

Introduction: Patients with relapsed or refractory diffuse large B-cell lymphoma (DLBCL) can be treated with 3 resource intense options: autologous hematopoietic cell transplant (Auto-HCT), allogeneic HCT (Allo-HCT), or autologous anti-CD19 chimeric antigen receptor T-Cell (CAR T) therapy after two or more lines of systemic therapy, depending on the clinical scenario. Our aim was a comparative analysis of healthcare resource utilization (HCRU) and associated costs for matched samples of Medicare patients for treatment planners, payers, and policy-makers.

Methods: This study utilized a retrospective, observational cohort design. Data were derived from the Center for Medicare and Medicaid Services (CMS) 100% Medicare Fee-for-Service (FFS) Part A and B claims data. Part D data for the study period were not yet available, so pharmacy claims for oral medications were not evaluated. Patients with DLBCL were included if they received CAR T or Auto-HCT between 10/1/2017 and 3/31/2019 or Allo-HCT between 7/1/2012 and 3/31/2019. Patients with more than 1 CAR T or HCT were excluded. The index date was the date of initiation of the procedure. To allow for evaluation of patient characteristics and treatments pre- and post-procedure, patients must have been continuously enrolled in Medicare FFS for 6 months prior to (PRE) and after (POST) the index date. Patients who died during the POST period were included.

The 3 cohorts CAR T, Auto-HCT, and Allo-HCT were matched on baseline clinical characteristics using 1:1 propensity score matching with a caliper of 0.05, with Auto-HCT and Allo-HCT patients assumed to be clinically distinct populations offered at different lines of therapy and so were matched to CAR T patients separately. Due to limitations of the data, we were not able to match patients by line of therapy.

Baseline characteristics were age, gender, race, census region, dual eligibility status (i.e. Medicare plus Medicaid), ECOG-PS (derived from claims using a validated, published method), Charlson-Deyo Comorbidity Index (CCI) and recent history of DVT/PE or cytopenias. Measures of HCRU were all-cause hospitalizations, outpatient, and emergency department (ED) visits. Costs were total paid amounts. HCRU and cost data were calculated for the 6 months PRE and POST, but do not include the utilization and costs associated with the index procedure itself.

Results: The CAR T/Auto-HCT analysis included 175 patients each, while the CAR T/Allo-HCT analysis included 142 patients each.

All cohorts had a median age of 69-70, with slightly more females, mostly ECOG 0-1, and were predominantly white. Only 10-15% were dual eligible. Although all groups had a median CCI of 4, the CAR T/Auto-HCT patients yielded slighter higher mean comorbidity scores (5.2-5.4) than CAR T/Allo-HCT (4.8).

For the CAR T/Auto-HCT comparison (Table 1a), Auto-HCT utilization was higher than CAR T on all acute care hospitalization measures and mean total medical costs were 35% higher ($85,382 vs. $63,081, respectively) during PRE. During POST, all measures were lower than PRE for both groups including total medical costs ($25,277 and $33,876, respectively). The reductions were greater for Auto-HCT than for CAR T, resulting in Auto-HCT having lower POST utilization and costs than CAR T for all measures except % of patients with outpatient encounters.

Allo-HCT utilization was higher than CAR T on all PRE measures including cost ($92,119 and $70,105, respectively, Table 1b) except ED visits. For POST vs. PRE, CAR T HCRU and costs declined ($34,477) similarly to that in the Auto-HCT comparison. But for Allo-HCT, acute care length of stay increased at POST compared to PRE, and total medical costs were relatively unchanged ($82,847) POST. POST HCRU and cost differences between these groups are not attributable to different survival rates, as no differences were seen.

Conclusions: In this observational descriptive analysis, HCRU and costs for Auto-HCT patients were higher during the 6 months pre-HCT compared to a matched group of CAR T patients but were lower during the 6 months post-HCT. For a matched group of Allo-HCT patients, however, the post-HCT period showed HCRU and costs that were equal to or higher than pre-HCT and were higher than CAR T during the same period. Understanding the relative HCRU and costs associated with these 3 procedures common to relapsed/refractory DLBCL will allow for better HCR planning.