Article Text

Download PDFPDF

Nucleos(t)ide analogues in patients with chronic hepatitis B: to stop or not to stop?
  1. Jia-Horng Kao1,
  2. Thomas Berg2
  1. 1 Graduate Institute of Clinical Medicine, National Taiwan University College of Medicine and Hospital, Taipei, Taiwan
  2. 2 Clinic and Polyclinic of Gastroenterology, Hepatology, Infectiology and Pneumology, University Hospital Leipzig, Leipzig, Germany
  1. Correspondence to Prof. Dr Thomas Berg, Clinic and Polyclinic for Gastroenterology, Hepatology, Infectiology and Pneumology, Univeristy Hospital Leipzig, Liebigstr. 20; 04103 Leipzig, Germany; thomas.berg{at}

Statistics from

Chronic HBV infection is a treatable but not curable disease. Long-term treatment with potent and safe nucleos(t)ide analogues (NA) with a high genetic barrier to drug resistance helps to reduce cirrhosis, hepatocellular carcinoma and liver-related mortality by profound viral suppression.1 Nevertheless, several controversial issues of chronic hepatitis B (CHB) management still exist and deserve further studies, including the selection of best candidates for a finite approach of NA therapy (box 1).

Box 1

Controversial issues and unmet needs of chronic hepatitis B management

  • A finite approach of NA treatment.

  • Risk stratification of HCC under long-term NA therapy.

  • Combination therapy with NA and interferon.

  • Treatment of immune-tolerant or HBeAg-positive patients with chronic HBV infection.

  • Novel biomarkers to optimise chronic hepatitis B management.

  • HBeAg, hepatitis B e antigen; HCC, hepatocellular carcinoma; NA, nucleos(t)ide analogue.

International practising guidelines recommend that NA therapy should continue until HBsAg loss, at least in patients with HBeAg-negative CHB. However, a finite NA therapy has been practised in resource-limited regions, especially in Asia. The recent European Association for the Study of the Liver (EASL) guideline suggested NA may be stopped in HBeAg-positive non-cirrhotic patients who achieve HBeAg seroconversion with undetectable HBV DNA for at least 12 months, or in HBeAg-negative non-cirrhotic patients with persistent virological remission for >/=3 years.2 The pros of stopping NA are a finite treatment duration, with improved adherence, reduced costs and minimisation of renal and bone toxicity. The cons are the risk of reactivation of suppressed HBV, resulting in an unpredictable worsening of disease and possible development of severe hepatitis or even acute-on-chronic liver failure. The off-therapy relapse is common and up to 70% of patients have virological relapse (HBV DNA >2000 IU/mL), whereas up to 50% of patients experience clinical relapse (virological relapse with serum alanine aminotransferase (ALT) level >2× upper limit of normal (ULN)), and some may develop hepatic decompensation or even mortality, especially in patients with cirrhosis.3 In contrast, previous retrospective studies, small prospective studies and systematic reviews claimed that stopping NA-related hepatitis flares may lead to long-term virological remission with a relatively high rate of HBsAg loss or ‘functional cure’.4 5 However, these results remain debatable and need to be validated by larger prospective studies.

In Gut, Liem et al conducted a randomised controlled trial (Toronto STOP study) to evaluate a finite treatment approach in HBeAg-positive and HBeAg-negative patients with CHB who met the stopping rules suggested by EASL guideline.6 In total, 67 patients, almost of Asian origin, were enrolled and randomised 2:1 to stop or continue NA therapy until 72 weeks. Sustained disease remission defined as HBeAg negativity, HBV DNA level <2000 IU/mL and normal ALT was evaluated at 72 weeks after stopping NA therapy. They found that stop group had a lower sustained disease remission than continue group (29% vs 82%), and HBsAg loss occurred in only one HBeAg-negative patient in the group who stopped therapy but also in one of the control patient who entered the study with nearly undetectable HBsAg levels. The meian HBsAg decline was similar between these two groups (0.2 vs 0.1 log IU/mL in stop vs continue patients). Of stop group patients, 33% had virological or biochemical relapse and 38% were retreated. Of note, 21% of them developed ALT >10× ULN and another 10% had ALT >5× ULN. No hepatic decompensation or mortality occurred. In addition, pretreatment HBeAg-positive patients required more retreatment than HBeAg-negative patients (61% vs 22%). On the basis of these data, they made a different conclusion from previous studies that stopping NA therapy conferred little benefit in the CHB population of mainly Asian patients.

What can we learn from this randomised controlled trial? First, their results consistently confirmed that stopping NA therapy in all patients with CHB without any selection criteria universally leads to a high rate of virological and/or clinical relapse in need of retreatment, even after a profound and prolonged viral suppression. Recently, Lampertico and Berg divided the typical courses of HBeAg-negative patients with CHB after stopping long-term NA therapy into four different chronological phases (treatment phase, lag phase, reactivation phase and consolidation phase).7 Following the consolidation phase, four different long-term outcomes may be defined: HBsAg loss in 20% after 2–3 years of follow-up; sustained virological response with true healthy carrier state in 20%–30%; indeterminate state not fulfilling immediate retreatment criteria in 10%–20%; and CHB requiring retreatment in 40%. In line with these outcomes, 38% of stop patients in this prospective study required retreatment. Of particular note is that the lower rate of HBsAg loss (2%) in this study compared with previous ones (~20%) may be reasoned by different ethnicities (Asian vs Caucasian), HBV genotypes (B and C vs A and D) and retreatment criteria (virological relapse vs clinical relapse). Further larger comparative and matched studies are needed to address these interesting and important issues.

Second, since virological and/or clinical relapse after stopping long-term NA therapy is not uncommon, it is thus clinically imperative to identify possible host, virus and treatment factors predictive of relapse after stopping NA to discern the best candidates to stop NA therapy (table 1). In the meantime, these factors may also distinguish those who can stop NA to induce a durable remission of HBV infection and reach the goal of functional cure over time. Among these hepatitis B viral factors, serum HBsAg, HBV RNA or HBcrAg level has been proposed to guide physicians the stopping of NA therapy.8 In general, the lower the HBsAg, HBcrAg or HBV RNA level at the end of therapy, the lower the risk of virological or clinical relapse. However, HBcrAg and HBV RNA assays are not yet approved by the European Medicines Agency or Food and Drug Administration, and remain research tools. Accordingly, quantitative HBsAg level may serve as a more convenient serological marker in our clinical practice.9 For example, a previous prospective study and a recent systematic review showed that HBeAg-negative patients with CHB with end-of-therapy HBsAg level <100 IU/mL had lower rates of off-therapy virological relapse (9%–19%) and clinical relapse (15%–30%), but a higher rate of off-therapy HBsAg loss rates (21%–59%), implying a finite NA therapy remains an option in selected candidates.10 11 In addition, prolonged on-treatment viral suppression is expected to offer higher chances of off-therapy viral remission. However, the option of optimal consolidation duration is still open. The timing and retreatment criteria are another unsettled issue.12 Although transient hepatitis flares may be beneficial to induce disease remission and even HBsAg loss, severe hepatitis flare-associated liver decompensation resulting in liver transplantation or mortality is not rare (cumulative decompensation rate of 3% and mortality rate of 1% among patients with cirrhosis).13 Considering the small but significant risk of adverse outcomes after stopping NA, close monitoring after treatment discontinuation and timely re-retreatment is of paramount importance. Lastly, in our clinical practice, assessment of liver fibrosis and exclusion of cirrhosis is difficult in patients under NA therapy (outside of clinical studies with carefully selected patients) and current measurements at the time of viral suppression may miss the presence of cirrhosis.

Table 1

Possible factors predictive of clinical relapse after stopping NA therapy

In summary, compared with long-term NA therapy, a finite approach of NA treatment with a satisfactory rate of functional cure is possible in highly selected population of non-cirrhotic HBeAg-negative patients who can have a close follow-up. However, further studies are needed to define potential predictors and retreatment criteria to maximise the benefits of HBsAg loss and minimise the adverse outcomes of severe hepatitis flare after stopping NA therapy.6


View Abstract


  • Contributors JHK wrote the manuscript and TB provided critical comments.

  • Funding This work was supported by grants from the Ministry of Science and Technology, Taiwan; National Taiwan University Hospital; and the Liver Disease Prevention and Treatment Research Foundation, Taipei, Taiwan.

  • Competing interests None declared.

  • Patient consent for publication Not required.

  • Provenance and peer review Commissioned; internally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles