A model to predict antiviral treatment in HBeAg negative chronic hepatitis B with alanine aminotransferase≤2 upper limit of normal

Abstract

Background & aims: Liver histological assessment is essential for predicting antiviral therapy in HBeAg negative chronic hepatitis B (CHB) patients with serum alanine aminotransferase (ALT)≤2 upper limit of normal (ULN). The aim was to establish a model to predict antiviral treatment for those patients without liver biopsy.

Methods: Three hundred and one consecutive treatment naive HBeAg negative CHB patients with HBV DNA≥2000 IU/ml and ALT≤2 ULN were retrospectively enrolled, among which 158 patients were for the training set and 143 for validation set. A multivariate logistic regression model was constructed in the training set and validated in the validation set.

Results: Our model identified four independent factors for the timing of treatment: Age (OR 1.050, 95%CI 1.004-1.098), Ln(aspartate aminotransferase) (OR 17.425, 95%CI 5.394-56.292), Log10 [HBV DNA] (OR 0.704, 95%CI 0.514-0.963) and platelet (OR 0.980, 95%CI 0.970-0.990). It showed 94% sensitivity, 88% negative predictive value (NPV) in the training set and 93% sensitivity, 85% NPV in the validation set using the low cut-off point of 5.16. Meanwhile, it showed 92% specificity, 88% positive predictive value (PPV) in the training set and 94% specificity, 92% PPV in the validation set using the high cut-off point of 7.26. It could predict treatment for 179 of 301(59%) patients without biopsy.

Conclusions: We established a model to predict antiviral therapy in HBeAg negative CHB patients with ALT≤2 ULN. Antiviral treatment should be initiated if the model value>7.26 and not if its value≤5.16. Liver biopsy is needed only when its value between the two points.

Keywords: antiviral therapy; chronic hepatitis B; liver biopsy; non-invasive model.