Abstract
Hepatitis B surface antigen (HBsAg) seroclearance, an indicator of recovery from hepatitis B virus (HBV) infection, is uncommon in long-term nucleos(t)ide analog (NUC) therapy. We compared the incidence of HBsAg seroclearance in patients with and without NUC discontinuation to identify predictors of HBsAg seroclearance. This retrospective study enrolled adult patients with a chronic HBV infection followed for ≥12 months after NUC discontinuation (finite group) and those treated with NUCs for >3 years (non-finite group). Demographic, clinical, and laboratory data were analyzed. The study cohort included 978 patients, including 509 and 469 patients in the finite and non-finite groups, respectively. Cumulative HBsAg seroclearance incidence was significantly higher in the finite group than in the non-finite group (p = 0.006). The 5- and 10-year cumulative HBsAg seroclearance incidence were 6.6% and 18.9% in the finite group and 3% and 14.6% in the non-finite group, respectively. The likelihood of HBsAg seroclearance was higher in those with end of treatment (EOT) HBsAg levels of <100 IU/mL and in those without clinical relapse (CR). The cumulative 3-year CR incidence was 16.8%. The incidence of liver decompensation and hepatocellular carcinoma were 4.1 and 0.4 per 1000 person-years, respectively. The hepatocellular carcinoma incidence did not significantly differ between the finite and non-finite groups (p = 0.941). In conclusion, higher HBsAg seroclearance incidence in patients receiving finite therapy, and the increased likelihood of HBsAg seroclearance in those with EOT HBsAg levels of <100 IU/mL and in those without CR should be considered during decision-making of treatment options.