A Cost-Effectiveness Analysis of Glecaprevir/Pibrentasvir Versus Existing Direct-Acting Antivirals to Treat Chronic Hepatitis C in Japan.

INTRODUCTION

The objective of the study was to evaluate the cost-effectiveness of glecaprevir/pibrentasvir versus other direct-acting antivirals (DAAs) for treating chronic hepatitis C virus (HCV) infections in Japan.

METHODS

We developed a health state transition model to capture the natural history of HCV. A cost-effectiveness analysis of DAAs from the perspective of a public healthcare payer in Japan with a lifetime horizon over annual cycles was performed. Treatment attributes, baseline demographics, transition probabilities, health-state utilities, and costs data were extracted from publications. Costs and outcomes were discounted at 2% per annum. In the base case we focused on genotype 1 (GT1) treatment-naïve patients without cirrhosis. The scenario analysis examined a pan-genotype treatment in GT1-3 (i.e., portfolio), treatment-naïve, and treatment-experienced patients. The portfolio cost-effectiveness of DAAs was derived by calculating a weighted average of patient segments defined by treatment history, cirrhosis status, and genotype.

RESULTS

The base case results indicated that glecaprevir/pibrentasvir was dominant (i.e., generating higher quality-adjusted life years [QALYs] and lower lifetime costs) compared to all other DAAs. The predicted lifetime risk of hepatocellular carcinoma was 3.66% for glecaprevir/pibrentasvir and sofosbuvir/ledipasvir, 4.99% for elbasvir/grazoprevir, and 5.27% for daclatasvir/asunaprevir/beclabuvir. In scenario analysis the glecaprevir/pibrentasvir (GLE/PIB) portfolio dominated the sofosbuvir (SOF)-based portfolio (namely sofosbuvir/ledipasvir in GT1-2 and sofosbuvir + ribavirin in GT3). The base case probabilistic sensitivity analysis (PSA) showed that glecaprevir/pibrentasvir was cost-effective in 93.4% of the simulations for a willingness-to-pay/QALY range of Japanese yen (JPY) 1.6-20 million. The PSA for the portfolio scenario indicated that the GLE/PIB portfolio was cost-effective in 100% of simulations until the willingness-to-pay/QALY reached JPY 5.2 million; this proportion decreased to 69.4% at a willingness-to-pay/QALY of JPY 20 million. Results were also robust in deterministic sensitivity analyses.

CONCLUSION

In GT1 treatment-naïve non-cirrhotic patients GLE/PIB was a cost-effective strategy compared to other DAAs. When a pan-genotypic framework was used, the GLE/PIB portfolio dominated the SOF-based portfolio.