Prediction model for hepatitis B e antigen seroconversion in chronic hepatitis B with peginterferon-alfa treated based on a response-guided therapy strategy.

Abstract

Background: Models for predicting hepatitis B e antigen (HBeAg) seroconversion in patients with HBeAg-positive chronic hepatitis B (CHB) after nucleos(t)ide analog treatment are rare.

Aim: To establish a simple scoring model based on a response-guided therapy (RGT) strategy for predicting HBeAg seroconversion and hepatitis B surface antigen (HBsAg) clearance.

Methods: In this study, 75 previously treated patients with HBeAg-positive CHB underwent a 52-week peginterferon-alfa (PEG-IFNα) treatment and a 24-wk follow-up. Logistic regression analysis was used to assess parameters at baseline, week 12, and week 24 to predict HBeAg seroconversion at 24 wk post-treatment. The two best predictors at each time point were used to establish a prediction model for PEG-IFNα therapy efficacy. Parameters at each time point that met the corresponding optimal cutoff thresholds were scored as 1 or 0.

Results: The two most meaningful predictors were HBsAg ≤ 1000 IU/mL and HBeAg ≤ 3 S/CO at baseline, HBsAg ≤ 600 IU/mL and HBeAg ≤ 3 S/CO at week 12, and HBsAg ≤ 300 IU/mL and HBeAg ≤ 2 S/CO at week 24. With a total score of 0 vs 2 at baseline, week 12, and week 24, the response rates were 23.8%, 15.2%, and 11.1% vs 81.8%, 80.0%, and 82.4%, respectively, and the HBsAg clearance rates were 2.4%, 3.0%, and 0.0%, vs 54.5%, 40.0%, and 41.2%, respectively.

Conclusion: We successfully established a predictive model and diagnosis-treatment process using the RGT strategy to predict HBeAg and HBsAg seroconversion in patients with HBeAg-positive CHB undergoing PEG-IFNα therapy.