Cost-Effectiveness of Glofitamab vs Chemoimmunotherapy Regimens for Patients With Relapsed/Refractory Diffuse Large B-Cell Lymphoma (RRDLBCL) After at Least Two Lines of Systemic Therapy in the Czech Republic

OBJECTIVES: As patients with DLBCL who have failed two or more prior lines of systemic therapy continue to have a poor prognosis, there is a high unmet need for new treatment options. Glofitamab is a CD20xCD3 bispecific monoclonal antibody, which gained a conditional EMA approval based on results of a phase II study (NCT03075696). Here we aimed to evaluate cost effectiveness of glofitamab compared to R-CEOP and R-GD regimens for patients with relapsed/refractory DLBCL after at least two lines of systemic therapy from payers´ perspective in the Czech Republic.

METHODS: Due to the lack of comparison data, we used rituximab and bendamustine regimen (BR) as a proxy for clinical efficacy of R-CEOP and R-GD. A matching-adjusted indirect comparison (MAIC) vs BR was conducted using data from Hong 2018. As it was impossible to adjust MAIC for imbalances in number of previous treatment lines (2+L in Hong 2018 vs 3+L in NCT03075696) and ECOG status (PS 0-4 in Hong 2018 vs 0-1 in NCT03075696) an additional adjustment of extrapolated survival curves was applied in a glofitamab arm using results from a COTA database (Ip 2022) for these factors. A partitioned survival model including clinical states „without progression“, „progression“ and „death“ was created based on KM curves for PFS and OS from NCT03075696 trial and Hong 2018.

RESULTS: Over a time horizon of 37 years with 3% discount rates, treatment with glofitamab generated 2.80 incremental QALYs compared to R-CEOP or R-GD and incremental costs of 61,266 EUR vs R-CEOP and 61,446 EUR vs R-GD resulting in ICER of 21,882 EUR/QALY or 21,946 EUR/QALY, respectively. A scenario without a COTA database correction resulted in a QALY increment of 2.32.

CONCLUSIONS: Our cost-effectiveness analysis showed a substantial gain in QALYs with acceptable incremental costs. Use of results from COTA database further improved our results.