CALCULATING THE REAL-WORLD INCREMENTAL COST-EFFECTIVENESS RATIO FOR DARATUMUMAB WITH LENALIDOMIDE AND DEXAMETHASONE IN THE SECOND-LINE SETTING FOR RELAPSED REFRACTORY MYELOMA IN BRITISH COLUMBIA
Author(s)
Jane de Lemos, MSc, PharmD1, Christopher Venner, MD2, Stuart Peacock, BA, MSc, DPhil3, Jelena Mucovic, BSc4, Annie Zeng, BSc5, Brandon Whitmore, BSc6, Fiona Chen, BSc5.
1Pharmacoeconomics Pharmacist, BC Cancer, Vancouver, BC, Canada, 2Cancer Control Research, BC Cancer, Vancouver, BC, Canada, 3Canadian Centre for Applied Research in Cancer Control, Vancouver, BC, Canada, 4BC Cancer, Vancouver, BC, Canada, 5UBC, Vancouver, BC, Canada, 6Interior Health, Kelowna, BC, Canada.
1Pharmacoeconomics Pharmacist, BC Cancer, Vancouver, BC, Canada, 2Cancer Control Research, BC Cancer, Vancouver, BC, Canada, 3Canadian Centre for Applied Research in Cancer Control, Vancouver, BC, Canada, 4BC Cancer, Vancouver, BC, Canada, 5UBC, Vancouver, BC, Canada, 6Interior Health, Kelowna, BC, Canada.
OBJECTIVES: Daratumumab in combination with lenalidomide and dexamethasone (DRd) was introduced as second-line therapy for relapsed/refractory multiple myeloma in 2019. Canada’s Drug Agency estimated incremental cost-effectiveness ratio (ICER) of $165-$594K/QALY based on immature trial data. Real-world evidence indicates similar overall survival (OS) but shorter progression-free survival (PFS), necessitating reassessment of cost-effectiveness in real-world settings.
METHODS: Patients receiving Rd (July 2014-June 2017) and DRd (February 2019-March 2022) in British Columbia were identified. Overlap weighting using propensity scores balanced age at second-line, sex, ECOG (0, 1, ≥2), eGFR (<30, 30-59, ≥60 mL/min), refractory status, prior therapy class, and ASCT. Doubly robust Cox models estimated treatment effects for PFS and OS. Weibull models fitted to Rd survival and extrapolated for DRd. A 3-state Markov model estimated QALYs under three structural assumptions: PFS-benefit only, PFS-driven post-progression survival (PPS) assumptions, and explicit OS modelling. Costs included treatment and adverse event management.
RESULTS: Rd (n=240) and DRd (n=248) patients with median follow-up 44 months (Rd) and 25.7 months (DRd). Incremental QALYs: 0.05 under a PFS-only model, 0.448 with PFS-driven PPS assumptions and 0.434 with explicit OS modelling. ICER approximately $449,000/QALY. Adjustment for ASCT had minimal impact (OS HR 0.89; QALYs 0.435). Greater variation was observed across prognostic subgroups: ISS complete-case analysis showed attenuated benefit (HR 0.91; QALYs 0.343), while cytogenetic complete-case analysis showed greater benefit and a more favourable ICER (HR 0.86; QALYs 0.545; ICER $369,000/QALY). Results were consistent with modest OS benefit (HR 0.89) and substantial PFS effect (HR 0.67).
CONCLUSIONS: Cost-effectiveness estimates were highly sensitive to structural assumptions and prognostic subgroup. Models restricting benefit to PFS substantially underestimated value. Treatment effects were robust to adjustment for ASCT but varied across markers of disease burden (ISS) and disease biology (cytogenetic risk). Further analyses incorporating longer follow-up will address uncertainty through probabilistic sensitivity analysis and will include monitoring costs.
METHODS: Patients receiving Rd (July 2014-June 2017) and DRd (February 2019-March 2022) in British Columbia were identified. Overlap weighting using propensity scores balanced age at second-line, sex, ECOG (0, 1, ≥2), eGFR (<30, 30-59, ≥60 mL/min), refractory status, prior therapy class, and ASCT. Doubly robust Cox models estimated treatment effects for PFS and OS. Weibull models fitted to Rd survival and extrapolated for DRd. A 3-state Markov model estimated QALYs under three structural assumptions: PFS-benefit only, PFS-driven post-progression survival (PPS) assumptions, and explicit OS modelling. Costs included treatment and adverse event management.
RESULTS: Rd (n=240) and DRd (n=248) patients with median follow-up 44 months (Rd) and 25.7 months (DRd). Incremental QALYs: 0.05 under a PFS-only model, 0.448 with PFS-driven PPS assumptions and 0.434 with explicit OS modelling. ICER approximately $449,000/QALY. Adjustment for ASCT had minimal impact (OS HR 0.89; QALYs 0.435). Greater variation was observed across prognostic subgroups: ISS complete-case analysis showed attenuated benefit (HR 0.91; QALYs 0.343), while cytogenetic complete-case analysis showed greater benefit and a more favourable ICER (HR 0.86; QALYs 0.545; ICER $369,000/QALY). Results were consistent with modest OS benefit (HR 0.89) and substantial PFS effect (HR 0.67).
CONCLUSIONS: Cost-effectiveness estimates were highly sensitive to structural assumptions and prognostic subgroup. Models restricting benefit to PFS substantially underestimated value. Treatment effects were robust to adjustment for ASCT but varied across markers of disease burden (ISS) and disease biology (cytogenetic risk). Further analyses incorporating longer follow-up will address uncertainty through probabilistic sensitivity analysis and will include monitoring costs.
Conference/Value in Health Info
2026-11, ISPOR Europe 2026, Vienna, Austria
Value in Health, Volume 29, Issue 12S
Code
EE383
Topic
Clinical Outcomes, Economic Evaluation, Real World Data & Information Systems
Disease
Oncology