Comparative analysis of noninvasive models to predict early liver fibrosis in hepatitis B e antigen-negative chronic hepatitis B

Background: Portal or bridging fibrosis is an indication for antiviral treatment in patients with chronic hepatitis B (CHB). An early marker predictive of liver fibrosis in hepatitis B e antigen (HBeAg)-negative CHB patients can alert clinicians to plan for treatment before disease progression. GOALS: To predict early and significant liver fibrosis (Ishak score ≥2) in HBeAg-negative CHB by validating several noninvasive markers derived from CHC. STUDY: One hundred seventy-seven consecutive treatment-naive HBeAg-negative CHB patients who underwent liver biopsy were divided into a training group (n=121) and a validation group (n=56). Factors associated with liver fibrosis were analyzed. Results: Multivariate analysis identified Lok's model ≥0.87, cirrhosis discriminant score greater than 4, and positive alanine aminotransferase ratio platelet score as independent factors associated with liver fibrosis in the training group. The area under the receiver operating characteristic curve revealed that Lok's model was better than cirrhosis discriminant score in predicting liver fibrosis in both the training and the validation groups. In patients with hepatitis B virus DNA greater than 2000 IU/mL or greater than 20,000 IU/mL, Lok's model showed equal prediction value (area under the receiver operating characteristic curve 0.709 and 0.704, respectively). Lok's model could also discriminate high and low hepatitis B virus DNA loads. In general, liver biopsy can be avoided in one-third (58 of 177) of patients by Lok's model. Conclusions: Lok's model ≥0.87 can be an early marker of liver fibrosis in HBeAg-negative CHB patients. Lok's model has clinical applications not only for CHC, but also for HBeAg-negative CHB.