Direct-acting antivirals (DAAs) have significantly improved the efficacy and tolerability of the treatment of hepatitis C virus (HCV). However, studies conducted on actual patients with the aim of predicting the risk associated with treatment failure are lacking.
Our study enrolled 334 new HCV patients and assessed the effectiveness of treatment and predicted the risk of failure to achieve sustained virological response (SVR) by developing a multiple logistic model. Our study compared the variables between the two groups, those who did (group 0, ) and did not achieve SVR (group 1, ).
The cure rate of HCV at 12th week in our study was 71.56%. We found that advanced cirrhosis, HCV genotype, HBV coinfection, rapid virological response (RVR), fibrosis index (FIB-4) score, serum levels of AST, ALP, hemoglobin, and viral load before treatment were prognostic factors associated with rate of failure to achieve SVR at week 12 of DAA therapy. In the multiple logistic model, eight significant predictors including advanced cirrhosis status, HCV genotype, RVR, AST/ALP levels, FIB-4 score, and viral load before treatment predicted the risk of failure with excellent model performance (area under the receiver operating characteristic curve (AUCROC) [95% CI] =0.986 (0.971-0.999)). RVR and advanced cirrhosis were the two strongest predictors with odd ratios (95% CI) =9.72 (2.8, 39.28) and 51.54 (6.39, 139.82), respectively.
The multiple logistic regression model included significant factors to estimate the probability of failure to achieve SVR, which could improve HCV treatment strategy.