Association of week-12 HBsAg Decline with HBsAg Clearance in Patients with Hepatitis B-Related Cirrhosis Receiving Pegylated Interferon Combined with Nucleos(T)ide Analogue Therapy: A Retrospective Cohort Study.

PURPOSE: This study aimed to evaluate the association between the relative decline rate of HBsAg at week 12 (HBsAg R) and achieving HBsAg clearance in patients with compensated HBV-related cirrhosis. PATIENTS AND METHODS: This study included 174 patients with compensated hepatitis B-related cirrhosis receiving pegylated interferon-alpha and nucleos(t)ide analogue therapy. HBsAgR was defined as (week-12 HBsAg - baseline HBsAg) / baseline HBsAg, where a more negative value indicates greater decline. To identify stable predictors, we employed the Boruta machine learning algorithm for feature selection. The association between HBsAgR and HBsAg clearance at 48 weeks was assessed using multivariable logistic regression and ROC analysis. Model stability was internally validated using Bootstrap resampling with 1,000 iterations. RESULTS: HBsAgR was identified as a stable factor associated with HBsAg clearance. After adjustment, each one-unit increase in HBsAgR was associated with a 53% lower odds of clearance (aOR=0.47; 95% CI: 0.28-0.79), indicating that greater HBsAg decline (more negative HBsAgR) was associated with higher clearance probability. ROC analysis showed an AUC of 0.746. The optimal cutoff value of HBsAgR was -0.56585, which corresponds to a 56.6% decline in HBsAg from baseline at week 12. Patients with HBsAgR less than -0.56585 (ie, a decline greater than 56.6%) were associated with a higher probability of HBsAg clearance, yielding a sensitivity of 76.19% and specificity of 73.33%. Bootstrap validation showed minimal optimism (difference: -0.001). CONCLUSION: HBsAgR is strongly associated with HBsAg clearance at week 48 in patients with hepatitis B-related cirrhosis, and may serve as a useful modicification of treatment strategies for this population.