• antigen quantification;
  • hepatitis B virus;
  • nucleos(t)ide analogues;
  • treatment;
  • virological resistance

Summary.  Serum hepatitis B virus surface antigen (HBsAg) levels have been suggested to predict interferon response in chronic hepatitis B. A few data are available on the role of HBsAg measurement in nucleos(t)ide analogues (NA) treatment. We retrospectively investigated the relation between HBsAg changes and main treatment outcomes during long-term lamivudine treatment in hepatitis e antigen (HBeAg)-negative chronic hepatitis B. A total of 42 HBeAg-negative patients were consecutively enrolled in an open-label study on long-term lamivudine monotherapy (150 mg/die). Serum HBsAg levels were quantified every 6 months by Architect assay (Abbott Diagnostics). HBV-DNA was quantified quarterly by real-time PCR (Roche Diagnostics). The median duration of lamivudine treatment was 66 months (20–153). One patient (2%) was a primary nonresponder, 35 (83%) developed virological breakthrough (VB) and the remaining six patients (14%) were classified as long-term on-treatment responders. During treatment, HBsAg levels decreased only in long-term on-treatment responders, while no changes were observed in resistant patients. Failure to achieve a decrease of 0.7 log10 IU/mL in serum HBsAg at month six of lamivudine had a positive predictive value of developing VB of 90% and a negative predictive value of 100%. These high predictive values were also maintained in the subgroup of patients negative for HBV-DNA at month six. The results of this study with a small sample size suggest a role of on-treatment HBsAg quantification in the management of lamivudine-treated patients. If validated prospectively in a larger patient cohort, HBsAg measurements would be a useful adjunct to optimize antiviral therapy.