Exploratory cost-utility and cost-effectiveness analysis of venetoclax plus azacitidine vs. azacitidine monotherapy in Bulgarian patients with AML ineligible for intensive chemotherapy: a real-world-data-influenced state-occupancy model

Background : Older patients with newly diagnosed acute myeloid leukemia (AML) who are ineligible for intensive chemotherapy have poor outcomes, and local pharmacoeconomic evidence from Bulgaria is limited. An exploratory economic model comparing venetoclax plus azacitidine (VEN+AZA) versus azacitidine monotherapy (AZA) was developed using local real-world data. Objective : To estimate the incremental costs, life-years gained (LYG), and quality-adjusted life-years (QALYs) of VEN+AZA versus AZA from the Bulgarian healthcare payer perspective. Methods : A 1-year exploratory state-occupancy model with monthly cycles was developed. Health states were event-free survival (EFS), progressed/relapsed disease (PD/RL), and death. The analysis used local data from 69 patients in total, including 49 patients treated with VEN+AZA and 20 treated with AZA. Overall survival (OS) and reconstructed EFS informed state occupancy. Monthly PD/RL occupancy was derived as the difference between OS and EFS, and death was derived as one minus OS. Direct medical costs included drug acquisition, disease management, hospitalization, monitoring, and supportive and palliative care. QALYs were calculated by applying monthly health-state utility weights. A deterministic tornado diagram and probabilistic Monte Carlo sensitivity analyses were performed. Results : In the base case, VEN+AZA was associated with higher costs and greater health benefits than AZA, with an incremental cost-effectiveness ratio (ICER) of €289,309.59 per QALY and €235,736.28 per LYG. The model estimated higher QALYs and life-years for VEN+AZA, with incremental results driven primarily by improved survival and event-free occupancy. Probabilistic sensitivity analysis revealed that the model was sensitive to changes in efficacy. Conclusion : In this exploratory local-data model, VEN+AZA improved health outcomes compared with AZA at increased cost. The model should be interpreted as hypothesis-generating because of the limited sample size, incomplete progression ascertainment, short time horizon, and use of reconstructed state occupancy.