Relapses are observed in most HBeAg-negative chronic hepatitis B patients who discontinue nucleos(t)ide analogues (NAs), but the rates of relapse vary widely among studies and whether all patients with relapse need retreatment is unclear. The aim of this study was to assess the impact of different definitions on the rates of post-treatment relapse and therefore on the probability of retreatment in such patients who discontinued effective long-term NAs therapy. In total, 130 non-cirrhotic HBeAg-negative chronic hepatitis Bpatients before NAs were included. All had on-therapy virological remission for ≥24 months and close follow-up for ≥12 months afterstopping NAs or until retreatment, which started on stringent predefined criteria. Relapses rates based on several predetermineddefinitions of virological and perhaps biochemical criteria were assessed. The median duration of therapy was 60 and of on-therapy virological remission 43 months. During a median off-NAs follow-up of 15 months, no patient experienced liver decompensation or died. Cumulative relapse rates ranged between 2%-49% at 3, 4%-73% at 6, 11%-82% at 12 and 16%-90% at 24 months, whereas cumulative retreatment rates were 15%, 22% and 40% at 6, 12 and 24 months after NAs discontinuation. No patient characteristic was independently associated with the probability of relapse based on at least two definitions or of retreatment. In conclusion, in HBeAg-negative chronic hepatitis B patients who discontinue NAs, the definition of relapse has great impact on the off-NAs relapse rates and potentially on the probability of retreatment. Regardless of definition, off-NAs relapses cannot be easily predicted by the patients' characteristics. A substantial proportion of such patients may not require retreatment if stringent criteria are adopted.
This article is protected by copyright. All rights reserved.