Development and Validation of a Predictive Model for HBsAg Seroclearance After Peg-IFN-Based Therapy: A Multicentre...

Area of Science:

• Hepatology
• Virology
• Clinical Prediction Modeling

Background:

• Hepatitis B virus (HBV) infection remains a global health concern.
• Predicting HBsAg seroclearance is crucial for optimizing treatment strategies in chronic hepatitis B (CHB).
• Pegylated interferon (Peg-IFN)-based therapy offers a potential treatment option for CHB patients.

Purpose of the Study:

• To develop and validate a predictive model for HBsAg seroclearance.
• To identify baseline parameters associated with HBsAg seroclearance after Peg-IFN therapy.
• To aid in clinical decision-making for Peg-IFN-based treatment in virally suppressed HBeAg-negative CHB patients.

Main Methods:

• Retrospective enrollment of 377 nucleos(t)ide analogue-suppressed, HBeAg-negative CHB patients receiving 48-week Peg-IFN therapy.
• Development of a multivariate Cox regression model in a cohort of 229 patients.
• Validation of the model in an independent cohort of 148 patients.

Main Results:

• HBsAg seroclearance rates were 17.9% in the development cohort and 20.27% in the validation cohort.
• The predictive model incorporated age, baseline HBsAg, and alanine aminotransferase (ALT).
• The model demonstrated strong predictive performance (AUC 0.842 in development, 0.852 in validation) and identified patient subgroups with varying seroclearance incidence.

Conclusions:

• A robust predictive model for HBsAg seroclearance was successfully constructed.
• The model utilizes readily available baseline parameters for clinical application.
• This tool can guide the effective use of Peg-IFN-based therapy in CHB management.