Analysis of discordance between transient elastography and liver biopsy for assessing liver fibrosis in chronic hepatitis B virus infection.

Abstract

Transient elastography (TE) is used to assess liver fibrosis in chronic hepatitis B virus (CHBV) infection. However, factors affecting liver stiffness (LS) values and discordance between TE and liver biopsy in CHBV infection remain to be evaluated. The aim is to define the optimal cutoff values of LS for significant fibrosis (≥F2) and cirrhosis (F4) and to study the clinical and histological variables associated with LS values and discordance between TE and liver biopsy in assessing liver fibrosis in CHBV-infected subjects. Patients with CHBV infection (n=200; 159 male; age 37.6±3.7years) underwent liver biopsy concomitantly with TE. Liver biopsy was scored for activity (Ishak score), fibrosis (METAVIR score), steatosis, cholestasis, and congestion. Hepatic fibrosis percentage was estimated by morphometry. Liver stiffness values were significantly correlated with histological activity index (HAI) score, F score, and fibrosis percentage. Optimal cutoff values for prediction of significant fibrosis and cirrhosis were 7.05kPa [sensitivity 81.2%; specificity 74%; area under the receiver operating characteristic curve (AUROC) 0.850] and 10.85kPa (sensitivity 87%; specificity 85.3%; AUROC 0.907), respectively. A total of 47 (23.5%) [overestimation of actual fibrosis by TE, 34 (17%); underestimation, 13 (6.5%)] and 28 (14%) [overestimation, 25 (12.5%); underestimation, 3 (1.5%)] patients showed discrepant results for diagnosis of significant fibrosis and cirrhosis, respectively. HAI and interquartile range (IQR) were the factors predictive of overestimation in cirrhosis. Fibrosis and necroinflammatory activity are the main determinants of TE in CHBV infection. Overestimation of actual fibrosis stage by TE is common and is influenced by necroinflammatory activity and IQR for estimation of cirrhosis.