Predictive score for hepatocellular carcinoma after hepatitis B e antigen loss in patients treated with entecavir or tenofovir.

ABSTRACT

The risk of developing hepatocellular carcinoma (HCC) after hepatitis B e antigen seroclearance (ESC) remains unclear. We established and validated a new risk prediction model for HCC development after ESC in patients with chronic hepatitis B (CHB) receiving antiviral therapy (AVT). Between 2006 and 2016, 769 patients (training cohort) and 1,061 patients (validation cohort) with CHB who experienced ESC during AVT using entecavir (ETV) or tenofovir disoproxil fumarate (TDF) were recruited. In the multivariate analysis, male sex (hazard ratio [HR] = 2.092; 95% confidence interval [CI] = 1.152-3.800), cirrhosis (HR = 5.141; 95% CI = 2.367-11.167) and fibrosis-4 index (FIB-4) of >3.25 (HR = 2.070; 95% CI = 1.184-3.620) were the independent risk factors for HCC development (all P < .05). Accordingly, a novel HCC-ESCAVT model was developed (1x[sex male = 1, female = 0] + 3x(cirrhosis = 1, noncirrhosis = 0) + 1x(FIB-4 >3.25 = 1, ≤3.25 = 0). The cumulative risk for HCC development was significantly different among the risk groups based on the HCC-ESCAVT category (0-1, 2-4 and 5 for the low-, intermediate- and high-risk groups, respectively) (overall P < .001, log-rank test). The area under the receiver operating characteristic curve (AUC) for predicting HCC development 3, 5 and 10 years after ESC was 0.791, 0.771 and 0.790, respectively (all P < .05). The predictive value of the HCC-ESCAVT model was similar in the validation cohort (AUC = 0.802, 0.774 and 0.776 at 3, 5 and 10 years, respectively; all P < .05). Hence, we have developed and validated a new HCC-ESCAVT model for HCC development, which includes male sex, cirrhosis and FIB-4 of >3.25 as constituent variables.