Development and validation of a model to predict advanced fibrosis in chronic hepatitis B virus-infected patients with high viral load and normal or minimally raised ALT.

Abstract

BACKGROUND/AIMS: Treating a group of chronic hepatitis B virus (HBV)-infected patients with high viral load and advanced fibrosis or cirrhosis, regardless of their serum alanine aminotransferase (ALT) levels, is recommended. The aim of this study was to derive and validate a model to predict advanced fibrosis in a cohort of patients with viral replication and normal or minimally raised ALT [ALT < 2 × upper normal limit (UNL)].

METHODS: Information was collected from 124 patients who underwent liver biopsy. The diagnostic value of predictors was judged using multivariate logistic modeling and area under the receiver operating characteristic curves (ROC area).

RESULTS: Advanced fibrosis (F3 or F4 on METAVIR scoring schema) was diagnosed in 45.7% (95% CI, 34-57.4%) of the patients of the derivation set (70 patients) and in 37% (95% CI, 24.2-49.9%) of the validation set (54 patients). In the derivation set, age and aspartate aminotransferase (AST) were independent clinical predictors of advanced fibrosis. The ROC areas of this Age-AST model were 0.8 (95% CI, 0.7-0.9) in the derivation set and 0.82 (95% CI, 0.71-0.93) in the validation set, respectively. Without missing a single case, this model identified 11 patients (20%) without advanced fibrosis in the validation set.

CONCLUSIONS: A substantial proportion of patients with high viral load and ALT < 2 × UNL have advanced fibrosis. A simple model including age and AST is an easily applicable tool for physicians to guide their decisions on whether or not to perform liver biopsy in individual patients.