Abstract: Objective To investigate the novel biomarkers for predicting the natural course of chronic hepatitis B (CHB) by observing the levels of hepatitis B virus (HBV) RNA, DNA, hepatitis B core-associated antigen (HBcrAg) and other markers in patients with HBV infection. Methods A total of 185 CHB patients who were treatment-naive were enrolled at the First Affiliated Hospital of Hebei North University from Jan 1, 2018 to Jan 1, 2021. The patients were divided into an immune tolerance (IT) group (n=45), an immune clearance (IC) group (n=44), a negative inactive/quiescent carrier (ENQ) group (n=48) and an HBeAg negative hepatitis (ENH) group (n=48) according to the staging criteria of HBV infection. The levels of serum HBV RNA, HBV DNA, HBcrAg and alanine aminotransferase (ALT) were detected by different methods, and the model for predicting CHB staging were established and verified via different statistical software. Results Based on the staging criteria of HBV infection, the levels of HBV DNA, HBV RNA, HBcrAg, HBsAg in IT and IC groups were significantly higher than those in ENQ and ENH groups (all P<0.001); similarily, the concentrations of alanine transaminase (ALT), aspartate transaminase (AST), glutamyl transferase (GGT), controlled attenuation parameter (CAP) and liver stiffness measurements (LSM) in IC and ENH groups were higher than those in IT and ENQ groups (all P<0.05). The HBV RNA concentrations of each stage of infection was significantly different except for IT group when compared with IC group. The proportion of HBV RNA>6 lg copies/mL in the IT group and IC group was 93.34% and 81.82%, respectively (all P<0.05). The concentration distribution of HBcrAg in the four stages was similar to those of HBV RNA. Correlation analysis showed that HBV RNA, HBcrAg, HBV DNA concentrations were positively correlated with each other in HBeAg-positive patients (all P<0.001). In the HBeAg-negative patients, HBV RNA was positively related to HBV DNA (r=0.43) and HBcrAg (r=0.72) (all P<0.001), HBV DNA was positively associated with HBcrAg (r=0.40, P<0.001). On the basis of HBV RNA and HBcrAg, HBeAg-positive and negative patients were compared after ALT stratification, with statistically significant difference (all P<0.01). In this study, baseline data were divided into a training (n=129) and a validation (n=56) group with a ratio of 7:3. The results of C5.0 model for the significance of prediction showed that HBV RNA and ALT were important predictors, with an accuracy of 93.15%. The results of association rules showed that the predictive and association effects of ALT > 38.15, HBcrAg > 7.18, HBV RNA > 4.34 were the best. The results of internal and external validations showed that the actual observed values fit well with the predicted values by this model, and the clinical net benefit of the model was better. The area under curve (AUC) of the training group and validation group were 0.97 (95%CI: 0.94~1.00) and 0.96 (95%CI: 0.91~1.00), respectively, indicating that the training model predicted well. Conclusion Serum HBV RNA and HBcrAg combined with ALT may be helpful in monitoring the progression of CHB.

Key words: Chronic hepatitis B, Hepatitis B virus RNA, Hepatitis B core-associated antigen, Hepatitis B e antigen