HBsAg Quantification-Based Models for Predicting Hepatocellular Carcinoma Risk in Nucleos(t)ide Analogs-Experienced...

Area of Science:

• Hepatology
• Virology
• Oncology

Background:

• Chronic hepatitis B (CHB) is a primary cause of hepatocellular carcinoma (HCC).
• Nucleos(t)ide analogs (NAs) suppress HBV but do not eliminate HCC risk.
• Current HCC risk models lack accuracy due to omitting key markers like HBsAg levels.

Purpose of the Study:

• Develop and validate a novel risk stratification tool for HCC in NA-treated CHB patients.
• Improve predictive accuracy by incorporating hepatitis B surface antigen (HBsAg) quantification.
• Compare the performance of the new score against existing HCC prediction models.

Main Methods:

• A retrospective cohort of 1190 NA-treated CHB patients was used for Cox regression analysis to identify HCC risk factors.
• A prospective cohort of 506 patients served for external validation of the developed HBsAg-HCC Score.
• Performance was evaluated by comparing AUC values with established tools (PAGE-B, mPAGE-B, aMAP).

Main Results:

• Five independent HCC risk factors identified: HBsAg levels, age, male sex, hypoproteinaemia, and APRI.
• The HBsAg-HCC Score demonstrated strong predictive power with 3-/5-/7-year AUCs of 0.867/0.872/0.871 (training) and 0.784/0.780/0.777 (validation).
• The novel score significantly outperformed existing models in both cohorts, with validated stability and reliability.

Conclusions:

• The HBsAg-HCC Score provides a more accurate and comprehensive method for HCC risk stratification in NA-treated CHB patients.
• Incorporating HBsAg quantification addresses a critical limitation of conventional HCC prediction models.
• This tool aids in better clinical management and monitoring of patients at risk for HCC.