Pharmacogenetic-guided selection of warfarin versus novel oral anticoagulants for stroke prevention in patients with atrial fibrillation

Abstract

To compare clinical and economic outcomes of two anticoagulation therapy strategies, (i) pharmacogenetic-guided selection (PG-AC) of warfarin versus novel oral anticoagulants (NOACs), and (ii) usual anticoagulation care (usual AC) in patients with atrial fibrillation (AF), from the perspective of US healthcare payers. A Markov model was used to simulate long-term outcomes in a hypothetical cohort of 65-year-old patients with newly diagnosed AF: (i) all usual AC patients received warfarin therapy, and (ii) all PG-AC patients were genotyped. Patients with normal warfarin sensitivity genotypes would receive warfarin. Patients with high or low warfarin sensitivity genotypes would receive NOAC. Model inputs were derived from clinical trials published in the literature. The outcome measure was incremental cost per quality-adjusted life-year (QALY) gained (ICER). PG-AC gained higher QALYs with higher cost (9.912 QALYs and USD94 396) when compared with usual AC (9.721 QALYs and USD93 853) in base-case analysis. The ICER of PG-AC was 2843 USD/QALY. The ICER of PG-AC would exceed 50 000 USD/QALY if the monthly cost of NOAC was more than USD285 or the risk of stroke with NOAC versus warfarin was more than 0.93. In 10 000 Monte Carlo simulations, PG-AC was cost-effective 96.4% of the time and usual AC was cost-effective 3.6% of the time. PG-AC was more costly than usual AC with a mean cost difference of USD1927 (95% confidence interval 1.877-1.977, P<0.001), and gained higher QALYs by 0.209 (95% confidence interval 0.208-0.210, P<0.001). Compared with warfarin therapy with time in therapeutic range of 60%, using genotype to triage AF patients to warfarin or NOAC appears to be highly cost-effective.