OBJECTIVES: Failure-free survival (FFS) in chronic graft versus host disease (cGVHD) is a single, clinically recognised composite end point defined as the absence of treatment change, the absence of underlying malignancy and non-relapse mortality. This study describes an economic analysis model that utilises FFS data to estimate the mean costs and benefits of cGVHD treatments.

METHODS: A partitioned survival model with a 4-week cycle and three health states (failure-free [FF], failure and death) was developed in Microsoft Excel^Ò to estimate the clinical and economic outcomes of cGVHD treatments over a lifetime horizon. Patients enter the model in the FF state and may transition to the failure state after experiencing a recurrence of their malignancy or the initiation of a new cGVHD systemic therapy. Patients may transition to the death state either from the FF or the failure state. The model is populated by FFS and overall survival curves for each treatment and offers the flexibility to account or not account for response status (complete response, partial response or a lack of response) within the FF state using time to response and duration of response data. Health-related quality of life (HRQoL) associated with these health states and the impact on caregivers’ HRQoL were also assessed and incorporated into the model.

RESULTS: The model was populated with data from pooled analyses of the ROCKstar and Phase 2a clinical trials for belumosudil and the REACH3 trial for best available therapy (BAT) for cGVHD and estimated lifetime clinical outcomes. The life years and quality-adjusted life years (discounted at 3.5% per annum) were 6.77 and 3.58 for belumosudil versus 5.52 and 2.10 with BAT, respectively.

CONCLUSIONS: Modelling FFS outcomes is a clinically and payer-relevant approach to cost-effectiveness analysis, which captures the patient journey and value of cGVHD treatments.