We congratulate Ning et al. 1 for their recent tour-de-force (JGH, 2025;00:1–12) elucidating how the microbiota-derived bile acid deoxycholic acid (DCA) stabilizes the HBs–HBc interaction via the TM2 domain of the small surface protein, thereby facilitating Dane-particle assembly and sustaining low-level viremia (LLV) in patients receiving nucleos(t)ide analogues (NAs). The work elegantly closes a long-standing knowledge gap—namely, why 20%–30% of fully adherent NA-treated individuals continue to harbor 10–2000 IU mL−1 HBV DNA despite profound suppression of viral transcription. We would like to offer three considerations that we believe are critical if this mechanistic insight is to be converted into clinical benefit. The authors' ROC analyses (AUC 0.68–0.70) were generated in a single-center cohort (n = 111). Although modest, these values exceed the discriminatory power of HBsAg titre alone in LLV contexts (AUC ≈ 0.55). We urge the authors to validate a serum DCA threshold (e.g., ≥ 2 μM) in a prospective, multi-ethnic NA-treated cohort with paired cccDNA quantification. If confirmed, DCA could be incorporated into existing “LLV-risk calculators” together with HBcrAg and HBV-RNA, potentially guiding the decision to intensify therapy (add-on peg-IFN or switch to tenofovir alafenamide) before histological progression ensues 2. The demonstration that 7-MeCDC, a 7α-dehydroxylase inhibitor, simultaneously lowered hepatic DCA and plasma HBV DNA provides an immediately testable intervention. However, 7-MeCDC is not clinically approved. An alternative, already used in cholestatic disorders, is the non-absorbable FXR agonist obeticholic acid (OCA). Paradoxically, OCA might raise endogenous DCA by inhibiting CYP7A1, yet FXR activation also represses NTCP, thereby reducing HBV entry. We therefore propose a randomized, placebo-controlled pilot study of OCA 5–10 mg day−1 for 24 weeks in NA-treated LLV patients, with serial single-cell RNA-seq of liver biopsies to map FXR- versus DCA-mediated effects on cccDNA transcription and particle egress. Such data would clarify whether FXR agonism is friend or foe in the DCA-centric model 3. Broad-spectrum antibiotics (polymyxin B + vancomycin) reduced DCA and HBV DNA, but at the cost of collateral dysbiosis. A more sustainable strategy may be precision suppression of 7α-dehydroxylase-expressing taxa (e.g., Clostridium scindens) by CRISPR-antimicrobial peptides delivered via food-grade Lactobacillus, or competitive engraftment of DCA-low Akkermansia consortia. Fecal-microbiota transplantation from HBV-uninfected, low-DCA donors into HBV-transgenic mice with established LLV would formally test causality without the confounders of antibiotic-induced bile-acid pool depletion 4. Finally, we note that TM2 mutations disrupted virion morphogenesis without altering HBsAg secretion. This dissociation suggests that therapeutic targeting of the DCA-binding pocket might purge infectious particles while preserving circulating HBsAg, an advantageous profile for vaccines or monoclonal antibodies currently in development. In conclusion, Ning et al. have repositioned a host-derived metabolite as an obligate cofactor for HBV persistence. The onus now falls on the hepatology community to determine whether intercepting the gut–bile-acid–HBV axis can finally eradicate LLV and, by extension, reduce the unacceptably high incidence of HCC that continues to plague NA-suppressed patients. Yours faithfully. The authors declare no conflicts of interest. The authors will be able to provide data as needed.
Ge et al. (Wed,) studied this question.