Letter: Safety after cessation of nucleos(t)ide analogue therapy in patients with chronic hepatitis B infection—Authors' reply

We thank Liu et al for their interest in our study. We agree that pre-treatment HBeAg status could influence the risk of severe hepatitis flares after withdrawal of nucleos(t)ide analogue (NUCs). However, the limited number of events restricted our ability to explore risk predictors in the eligible sub-cohort, defined by available data to document negative HBeAg and undetectable hepatitis B virus DNA with a period of treatment consolidation prior to treatment cessation. Nonetheless, our data highlight the insufficiency of current stopping rules in safeguarding patients from severe withdrawal flares, irrespective of their pre-treatment HBeAg status. As Liu et al reported, both HBeAg-positive and -negative patients meeting the Asian-Pacific criteria for NUC discontinuation could suffer from liver failure and even death due to severe withdrawal flares.

We also agree that tenofovir disoproxil fumarate (TDF), when compared to entecavir (ETV), was associated with a higher risk of clinical relapse following treatment cessation.¹ However, we did not find an association between the antiviral regimen and hepatic decompensation from withdrawal flares in our study.² If Liu et al. had reviewed our article more thoroughly, they certainly would have noticed in the appendix that this factor was included in our multivariable model.

Our findings align with existing literature. For instance, analyses of the international RETRACT-B consortium showed that, while TDF was linked to an earlier occurrence of virological relapse and a higher rate of clinical relapse,³ it did not differ from ETV in the incidence of hepatic decompensation.⁴ This was corroborated by a recent hospital-based study, in which TDF conferred a higher rate of relapse and an earlier occurrence in HBeAg-positive patients but all of five patients who developed hepatic decompensation were withdrawing from ETV instead of TDF (5/96 vs. 0/70; p = 0.074).⁵ In our view, the discrepancy could be attributed to the timing of re-treatment as an earlier relapse may lead to an earlier resumption of antiviral treatment.⁶

Contrary to the suggestion by Liu et al, our approach, based on the national healthcare database in Taiwan, is less susceptible to attrition bias. Given the universal coverage of Taiwan's national health insurance,⁷ our outcome measurements did not rely on follow-up at a specific hospital. Moreover, we reported the duration of treatment consolidation in the eligible sub-cohort (section 3.4). Our analysis on post-treatment monitoring, with details available in the appendix, indicated that most serious outcomes did not result from poor adherence to medical attention.

We respectfully disagree with the assertion that our data inflated the risk of off-NUC hepatic decompensation. This study was meticulously designed to quantify the risk of severe withdrawal flares in a population where NUC cessation was routinely practised. In fact, the risks estimated in the current study were similar to those reported in large hospital-based cohort studies or meta-analysis of existing literature.^{4, 8} Based on our findings together with available evidence,⁹ we caution against considering NUC cessation as a routine practice for patients with chronic hepatitis B.

Yao-Chun Hsu: Conceptualization (lead); investigation (lead); writing – original draft (lead). Mindie H. Nguyen: Conceptualization (supporting); methodology (supporting); writing – review and editing (supporting). Chun-Ying Wu: Methodology (supporting); supervision (lead); writing – review and editing (lead).

Yao-Chun Hsu: Research support from Gilead Sciences. Advisory committee member for Gilead Sciences. Speaker bureau for Abbvie, Bristol-Myers Squibb, Gilead Sciences, Merck Sharp & Dohme and Novartis. Mindie H. Nguyen: Research support from Pfizer, Gilead Sciences, Vir Biotech, Glycotests, Exact Science, Helio Health, B. K. Kee Foundation and the National Cancer Institute. Advisory board member or consultant for Gilead Sciences, Intercept, Novartis, Eisai, Bayer, Exact Science, Laboratory of Advanced Medicine, Spring Bank, GSK and Janssen. Chun-Ying Wu: nothing to declare.

This article is linked to Hsu et al papers. To view these articles, visit https://doi.org/10.1111/apt.17614 and https://doi.org/10.1111/apt.17657