Usefulness of serial FIB-4 score measurement for predicting the risk of hepatocarcinogenesis after hepatitis C virus eradication.

Abstract

There is insufficient information to evaluate the correlation between fibrosis regression and hepatocellular carcinoma (HCC) risk after hepatitis C virus eradication. We analyzed serial changes in fibrosis (FIB)-4 scores after sustained virological response (SVR). The subjects were 717 patients who achieved SVR by interferon (IFN)-based therapy (IFN Group) and 635 patients who achieved SVR by direct-acting antiviral (DAA) therapy (DAA Group). We performed propensity score matching because the baseline characteristics differed between the IFN and DAA groups, and then applied inverse probability weighting (IPW). We compared the changes in FIB-4 scores between the IFN and DAA groups. We also investigated the dynamics of FIB-4 scores, which are useful for predicting hepatocarcinogenesis. Using time-dependent receiver operating characteristic curve analysis and an IPW-adjusted Cox proportional hazards model, we identified an FIB-4 cutoff of 1.50 for predicting hepatocarcinogenesis. The percentages of patients in the IFN and DAA groups who demonstrated IPW-adjusted cumulative reduction and increase in FIB-4 scores indicated no significant differences. No HCC developed during the 5-year follow-up period in 547 of the 1352 patients whose FIB-4 score was <1.50 at SVR or improved from ≥1.50 to <1.50 during follow-up. Only one patient developed HCC, at 7.3 years; this individual had diabetes mellitus and excessive alcohol intake. There was no difference in FIB-4 score reduction between the IFN and DAA groups. Patients whose FIB-4 scores improved to <1.50 or remained at <1.50 during follow-up after SVR had extremely low hepatocarcinogenesis rates.