OBJECTIVES

: Anti-neutrophil cytoplasmic antibody-associated vasculitis (AAV) is a rare and often life-threatening disease. Immunosuppressants and glucocorticoids (GCs) improve prognosis, but with associated toxicity. Avacopan is an orally administered treatment to induce and sustain remission in AAV. Cost-effectiveness needs to be demonstrated for reimbursement, however there are limited data on long-term outcomes. A model was developed utilising data from Clinical Practice Research Datalink (CPRD) and validated against external clinical sources.

METHODS

: A state-transition model compared avacopan in combination with rituximab (RTX) or cyclophosphamide (CYC) against GCs with RTX/CYC. Health states were: Active disease, Remission, Relapse, ESRD and Death, with tunnel states to track 6-month induction period (three in total). GC-related adverse events (AEs) were also modelled (infections, cardiovascular events (CV), ocular disease, bone disease). Transition probabilities to remission and relapse in RTX/CYC arms were estimated using CPRD data, with GC dosage (>10mg/day) used as a proxy for AAV relapse and a treatment effect applied for avacopan (ADVOCATE trial). AEs rates for no/low/high dose GC (>30mg/day) were estimated from CPRD using generalised linear models, adjusting for baseline covariates. Unit costs were sourced from NHS Reference Costs 2018/19. Modelled survival and AEs in the RTX/CYC arms were validated using two studies of UK patients.

RESULTS

: Survival in RTX/CYC+GC arms and Wallace (2016) was between 90-95% (one-year) and 85-90% (5-year). Patients experiencing events in the model compared to organ damage reported in Robson (2015) at 7.3 years were: 47.2% and 47.9% CV events, 16.0% and 14.1% bone disease; 33.9% and 31.7% renal disease (GFR<50 mL/min); 12.2% and 9.0% ESRD (GFR<15 mL/min); 11.9% and 8.5% ocular disease, respectively.

CONCLUSIONS

: The cost-effectiveness model appears to validate reasonably against other available clinical data sources and gives some confidence in populating with further data (treatment effect, QoL, costs), including from ADVOCATE trial and subsequent use for reimbursement processes.