OBJECTIVES: To assess the cost effectiveness of liso-cel vs standard-of-care treatment (SOC) i.e., salvage chemotherapy ± high-dose chemotherapy and hematopoietic stem-cell transplantation, and axicabtagene ciloleucel (axi-cel) for 2L treatment of R/R LBCL from the Swiss statutory health insurance system perspective.

METHODS: A partitioned-survival cost-effectiveness model reflecting 3 states (event-free survival [EFS], post-event survival, and death) estimated incremental cost-effectiveness ratios (ICERs), with costs and effects discounted at 3% annually over a 30-year (lifetime) horizon. Parametric mixture-cure modeling was used to extrapolate overall survival (OS) and EFS from TRANSFORM (liso-cel and SOC; NCT03575351) and ZUMA-7 (axi-cel; NCT03391466). Survival was modeled based on the age- and sex-adjusted Swiss population for cured patients, and using parametric distributions for non-cured patients. Axi-cel survival curves were derived from Bucher indirect treatment comparison. EQ-5D-5L utility values for the health states and adverse events (AEs) were taken from TRANSFORM. Costs included pre-treatment, acquisition, CAR-T cell infusion, subsequent treatment, AE management, hospitalization, monitoring, and end-of-life care, derived from the Swiss analysis list (AL), and the outpatient (TARMED) and inpatient (Swiss DRG) tariffs. Uncertainty was examined.

RESULTS: For liso-cel and axi-cel, 5-year OS was predicted to be 56% and 53%, respectively. Liso-cel dominated axi-cel, saving CHF17,839 for 0.50 additional quality-adjusted life-years (QALYs), including savings of CHF9,964 from AE management due to its lower rates of cytokine release syndrome and neurological events. Liso-cel conferred 1.35 more QALYs than SOC at an additional cost of CHF12,000, resulting in an ICER of CHF8,911/QALY. The probability of liso-cel being cost effective at a hypothetical threshold of CHF100,000 was 45.5% vs axi-cel and SOC. Key drivers of ICERs were the OS cure fraction and CAR-T therapy acquisition costs.

CONCLUSIONS: