The outcome of nucleos(t)ide analogues (NAs) discontinuation and retreatment is still uncertain. We evaluated hepatitis B surface antigen (HBsAg) kinetics after NAs discontinuation and during retreatment due to off-treatment clinical relapse among non-cirrhotic HBeAg-positive CHB patients. Four groups were studied: 129 HBeAg-positive patients from a prospective cohort who stopped NAs therapy after achieving sustained response (Group A), 39 patients who received retreatment after off-treatment clinical relapse in the discontinuation group (Group B), 214 patients who maintained treatment after achieving sustained response (Group C) and 291 patients who firstly initiated antiviral treatment (Group D). During a 5-year follow-up, the cumulative incidence of HBsAg loss was significantly higher in Group A than Group C (22.3% vs. 1.6%, p < .001). The quantitative HBsAg (qHBsAg) level at enrolment and NAs discontinuation were independently associated with HBsAg loss. Additionally, patients in Group B showed significantly greater HBsAg loss than those in the Groups C and D, with 5-year cumulative incidences of 9.0%, 1.6% (p = .040) and 0.6% (p < .001), respectively. Moreover, patients in the Group B exhibited better virologic response (100% vs. 98.8%, p < .001) and HBeAg seroconversion (92.6% vs. 69.8%, p < .001) than those in Group D at year 5. Propensity score-matched analysis also showed the similar trend of HBsAg decline. NAs discontinuation with or without subsequent retreatment resulted in a more profound reduction of HBsAg in non-cirrhotic HBeAg-positive patients, suggesting that discontinuation may be a potential cure strategy for those with sustained virological suppression.
Keywords: discontinuation; hepatitis B surface antigen; nucleos(t)ide analogues; retreatment.
© 2021 John Wiley & Sons Ltd.