A simple-to-use score system for predicting HBsAg clearance to peginterferon alfa-2b in nucleoside analogs-experienced chronic hepatitis B patients.

Objective: Patients with chronic hepatitis B (CHB) often fail to achieve clearance of the hepatitis B surface antigen (HBsAg) with peginterferon treatment. Our study aimed to develop a simple-to-use scoring system to predict the likelihood of HBsAg clearance following treatment with peginterferon alfa-2b(PEG-IFN-α2b) in patients with CHB. Methods: A total of 231 patients were enrolled and divided into HBsAg clearance (n = 37) and non-HBsAg clearance (n = 194) groups. Multifactor logistic models were constructed using univariate and multiple logistic regression analyses. The area under the receiver operating characteristic curve (AUC), calibration curve, and decision curve analysis were used to evaluate the discrimination, calibration, and clinical applicability of the predictive scoring system. Results: Four clinical variables (age, baseline HBsAg level, HBsAg level decline at week 12, and alanine aminotransferase ratio at week 12) were independently associated with HBsAg clearance after PEG-IFN-α2b treatment and, therefore, were used to develop a predictive scoring system ranging from 0 to 13. The optimal cut-off value was >4, with a sensitivity of 86.49%, specificity of 72.16%, positive predictive value of 37.2%, negative predictive value of 96.6%, and an AUC of 0.872. This model exhibited good discrimination, calibration, and clinical applicability. Among patients with scores <4, 4, or > 4 HBsAg clearance was achieved in 0.85, 14.29, and 37.21% of the patients, respectively. Conclusion: The scoring system could effectively predict the predominance of HBsAg clearance after PEG-IFN-α2b treatment in the early stage. This may be helpful when making clinical decisions for the treatment of patients with CHB.