Background: Current stopping rules during pegylated interferon (peg-IFN)/ribavirin (RBV) treatment rely on week 12 HCV RNA response, but earlier identification of non-responders offers clinical and economic advantages.
Aims and methods: To evaluate, among 129 HCV-genotype-1-infected, treatment-naive patients receiving peg-IFN/RBV, the feasibility of predicting treatment failure using receiver operating characteristics (ROC) curves after measuring week 4 HCV RNA decreases, and to assess baseline predictors of not achieving sustained virological response (SVR).
Results: Peg-IFN-alpha2b was used in 84.5% of patients. Fifty-three (41%) reached SVR. The best cutoff value of HCV RNA decrease at week 4 to predict non-SVR corresponded to 1 log10 IU/ml: sensitivity and negative predictive value: 100%; specificity: 64%; positive predictive value: 66%; ROC curve area: 0.91 (95% confidence interval [CI]: 0.86-0.96). By applying this threshold, treatment could have been discontinued at week 4 in 64% of virological non-responders (49/76). By univariate analysis, baseline HCV RNA > 800,000 IU/ml (P = 0.029), older age (P = 0.011), and higher aspartate aminotransferase (AST) levels (P = 0.005) or AST/alanine aminotransferase ratio values (P = 0.04) were associated with failure. After multivariate analysis, only baseline HCV RNA >800,000 IU/ml (odds ratio [OR]: 2.12; 95% CI: 1.005-4.488; P = 0.048) and higher AST levels (OR: 1.01; 95% CI: 1.003-1.024; P = 0.011) remained statistically significant.
Conclusions: The lack of > or = log10 IU/ml decrease in baseline HCV RNA at week 4 was 100% predictive of treatment failure, independently associated with HCV RNA > 800,000 IU/ml and higher AST levels.