Cost-Effectiveness Analysis of Baseline Testing for Resistance-Associated Polymorphisms to Optimize Treatment Outcome in Genotype 1 Noncirrhotic Treatment-Naïve Patients With Chronic Hepatitis C Virus.

OBJECTIVES: Direct-acting antivirals containing nonstructural protein 5A (NS5A) inhibitors administered over 8 to 12 weeks are effective in ∼95% of patients with hepatitis C virus. Nevertheless, patients resistant to NS5A inhibitors have lower cure rates over 8 weeks (<85%); for these patients, 12 weeks of treatment produces cure rates greater than 95%. We evaluated the lifetime cost-effectiveness of testing for NS5A resistance at baseline and optimizing treatment duration accordingly in genotype 1 noncirrhotic treatment-naïve patients from the perspective of the UK National Health Service. METHODS: A decision-analytic model compared (1) standard 12-week treatment (no testing), (2) shortened 8-week treatment (no testing), and (3) baseline testing with 12-/8-week treatment for those with/without NS5A polymorphisms. Patients who failed first-line therapy were retreated for 12 weeks. Model inputs were derived from published studies. Costs, quality-adjusted life-years, and the probability of cost-effectiveness were calculated. RESULTS: Baseline testing had an incremental net monetary benefit (INMB) of £11 838 versus standard 12 weeks of therapy (no testing) and low probability (31%) of being the most cost-effective, assuming £30 000 willingness to pay. Shortened 8 weeks of treatment (no testing) had an INMB of £12 294 and the highest probability (69%) of being most cost-effective. Scenario analyses showed baseline testing generally had the highest INMB and probability of being most cost-effective if first- and second-line drug prices were low (<£20k). CONCLUSIONS: Optimizing treatment duration based on NS5A polymorphisms for genotype 1 noncirrhotic treatment-naive patients in the United Kingdom is not cost-effective if the drug costs are high; the strategy is generally most cost-effective when drug prices are low (<£20k).