Determinants of HBsAg Loss After Nucleos(t)ide Discontinuation in a Prospective Cohort of HBeAg‐Negative Caucasian Patients

Introduction Nucleos(t)ide analogues (NAs) discontinuation in HBeAg‐negative chronic hepatitis B (CHB) aims at increasing functional cure rates. However, clinical outcomes and predictive factors for HBsAg loss or re‐treatment are affected by the heterogeneity of studies. We aimed to analyse baseline factors associated with outcomes after NAs discontinuation in a cohort with homogeneous re‐treatment criteria. Methods Prospective multicenter study of 149 HBeAg‐negative CHB with complete viral suppression and absence of cirrhosis recruited from 2016 to 2021. Re‐treatment criteria after NAs discontinuation were homogeneous based on liver tests and HBV‐DNA levels. Results In this predominantly Caucasian cohort, 71% received tenofovir with a median treatment duration of 10 (7–13) years. Forty patients (27%) achieved HBsAg loss after 44 (28–54) months. Lower qHBsAg at end of treatment (EOT) and longer treatment duration were independent predictors of HBsAg loss. Combining qHBsAg at EOT (1000 IU/mL) and month 3 after NA interruption (100 IU/mL) stratified patients into high (> 70%) or low (< 10%) likelihood of HBsAg loss. On the other hand, 43 (29%) patients were retreated; HBV‐DNA < 100 IU/mL at month 3 was significantly associated with remaining off‐therapy. Overall, safety was favourable, with mild or transient ALT flares occurring in 64 patients (43%). However, one patient (0.6%) developed a severe hepatitis and required liver transplantation. Conclusions NA withdrawal results in functional cure rates exceeding 25%, particularly in patients with long‐term antiviral treatment and low qHBsAg levels. Combining qHBsAg levels at EOT and at 3 months accurately stratifies patients according to the probability of HBsAg loss. Close monitoring to detect the need for treatment re‐introduction is mandatory.