Letter: Safety and Methodological Considerations for Hepatitis B Immunoglobulin Withdrawal in Hepatitis B Virus Mono‐Infected Liver Transplant Recipients

We read with great interest the multicentre prospective study by Viganò et al. [1] on hepatitis B immunoglobulin (HBIG) withdrawal in HBV mono-infected liver transplant (LT) recipients on high–genetic barrier nucleos(t)ide analogue (NA) monotherapy. The authors report low HBsAg reappearance (5.4%; 1, 2, and 3-year rates: 4.08%, 5.37%, and 6.89%), persistently undetectable HBV-DNA, routine liver function tests, and no HBV-related hepatitis or hepatocellular carcinoma (HCC), supporting the safety of HBIG discontinuation. We believe that several clarifications would strengthen the interpretation and generalisability. First, eligibility and exclusions merit greater specificity. The cohort excluded donors with HDV coinfection and those who were HBsAg-positive, and participants were required to undergo regular liver tests and maintain undetectable HBsAg/HBV-DNA levels at enrolment. However, essential modifiers—such as prior hepatic or biliary surgery post-LT, recurrent/chronic rejection episodes, significant biliary strictures, large ascites or pleural effusions, renal impairment at baseline (beyond eGFR-based dosing), and ongoing mTOR inhibitor or MMF intensification—may influence HBsAg kinetics, NA exposure, and surveillance intensity [2]. Explicit reporting or adjustment for these factors would help readers assess the external validity across centres with different graft/biliary complications, as well as immunosuppression practices [3]. Second, time-dependent processes warrant time-aware methods. HBIG was withdrawn a median of 11.6 years after LT, and HBsAg reappeared at a median of 9 months thereafter. Because exposure status (on vs. off HBIG; NA switches from LAM to ETV/TDF/TAF), anti-HBs titers, and immunosuppression evolve over follow-up, conventional Cox models with only baseline covariates may be vulnerable to time-related bias or non-proportional hazards. Landmark analyses at 3, 6, and 12 months post-withdrawal, or Cox models with time-dependent covariates (e.g., anti-HBs titre bands, NA type, immunosuppression phase) could better accommodate state transitions and clarify whether early vs. late HBsAg reappearance differs in risk or significance [4]. Third, outcome definitions could be expanded to capture clinically relevant events beyond the qualitative detection of HBsAg. The study elegantly shows that isolated low-level HBsAg positivity without HBV-DNA is clinically benign on third-generation NAs. Still, reporting quantitative HBsAg trajectories (slopes), anti-HBs waning patterns, and any low-level HBV-DNA “blips” (even those below the limit of detection by ultrasensitive assays, where available) would help contextualise biological noise versus meaningful reactivation [5]. In addition, competing risks (death, retransplantation) may modestly distort the cumulative incidence; Fine–Grey models could complement Kaplan–Meier estimates, especially in historical cohorts with long post-LT intervals [6]. Lastly, subgroup detail would enhance applicability. While univariate analyses did not identify predictors, stratified descriptions by: (a) pre-LT HCC versus non-HCC; (b) donor anti-HBc status; (c) duration of prior HBIG (< 3, 3–5, 5–10, > 10 years) and anti-HBs levels at withdrawal; and (d) NA backbone (ETV vs. TDF vs. TAF) could help centres implement earlier withdrawal policies with confidence. Renal safety data after switching from LAM to TDF/TAF are reassuring; providing absolute eGFR changes and proteinuria trends would further support the selection of NA in CKD [7]. In summary, Viganò et al. [1] provide compelling real-world evidence that HBIG can be safely discontinued in adherent HBV mono-infected LT recipients who are maintained on high–genetic–barrier NAs, with very low HBsAg reappearance and no virological or clinical relapse. Clarifying eligibility details, adopting time-dependent analyses, and broadening outcome granularity would further consolidate these practice-changing findings and facilitate wider guideline-aligned de-implementation of lifelong HBIG. Wenying Huang: writing – original draft, writing – review and editing. Wen Zhang: writing – original draft, writing – review and editing. Chang Hu: writing – original draft, writing – review and editing. The authors have nothing to report. The authors declare no conflicts of interest. This article is linked to Viganò et al. paper. To view this article, visit https://doi.org/10.1111/apt.70348. The authors have nothing to report.