The accuracy of baseline viral load for predicting the efficacy of elbasvir/grazoprevir in participants with hepatitis C virus genotype 1a infection: An integrated analysis.

European treatment guidelines for hepatitis C virus (HCV) infection recommend that people with genotype (GT) 1a infection and baseline viral load ≤800 000 IU/mL receive elbasvir/grazoprevir (EBR/GZR) for 12 weeks, and those with baseline viral load >800 000 IU/mL receive EBR/GZR plus ribavirin for 16 weeks. This analysis was conducted to clarify whether baseline viral load can serve as an accurate, sensitive or specific stratification factor for defining EBR/GZR regimens. In this post hoc, integrated analysis, participants with GT1a infection who received EBR 50 mg/GZR 100 mg for 12 weeks were stratified according to baseline viral load. Sustained virologic response at 12 weeks post-treatment was achieved by 95.2% (911/957) of participants and was higher among participants with baseline viral load ≤800 000 IU/mL vs >800 000 IU/mL (98.5% vs 93.9%). The 800 000 IU/mL threshold had a positive predictive value of 98.5%, a negative predictive value of 6.1%, a specificity of 91.3%, a sensitivity of 28.4% and an overall accuracy of 31.5%. A baseline viral load cutpoint of 800 000 IU/mL had high positive predictive value and specificity but poor negative predictive value, sensitivity and accuracy in predicting treatment outcomes in this population. Baseline NS5A resistance-associated substitutions (RASs) were detected in 25% (1/4) of virologic failures with baseline viral load ≤800 000 IU/mL and 59.5% (25/42) of those with baseline viral load >800 000 IU/mL. Overall, these data suggest that, compared with the use of a baseline viral load cutpoint, baseline testing for NS5A RASs enables more individuals to receive the 12-week EBR/GZR regimen without compromising the opportunity for SVR.