Reevaluating antiviral thresholds in HBV DNA-negative inactive HBsAg carriers: a multicenter histopathological analysis

Abstract

Background: The definition of inactive HBsAg carriers (IHC) varies globally, particularly regarding HBV DNA thresholds. Whether HBV DNA negativity reliably predicts histological quiescence remains uncertain.

Aims: This study evaluated liver pathology in IHC patients to reassess antiviral therapy thresholds.

Methods: This multi-center, retrospective study included 231IHCs(2018-2023) stratified by HBV DNA negativity (< 20IU/mL). Liver biopsies assessed inflammation (G ≥ 2) and fibrosis (F ≥ 2); evident hepatic injury (EHI) was defined as G ≥ 2 and/or F ≥ 2. Multivariable models evaluated predictors.

Results: Among 231 IHC patients(median age:43 years old; 95.2% ≥30 years old), 35.9%(83/231) were HBV DNA negative. The median HBsAg and HBV DNA level were 132 IU/ml and 94 IU/ml, respectively. Notably, EHI prevalence was significantly higher in HBV DNA negative patients than positive ones(44.9% vs. 31%, P = 0.04), driven by fibrosis (F ≥ 2: 42.2% vs. 21.6%, P < 0.001), challenging the assumption that HBV DNA negativity ensures low histological risk. Male sex, HBV DNA negativity, and elevated liver stiffness measurement(LSM) independently predicted EHI (OR = 3.37, AUC = 0.747).

Conclusion: HBV DNA negativity does not guarantee histological quiescence in inactive HBsAg carriers aged ≥ 30 years, with 44.9% exhibiting significant liver injury. In this population, LSM > 6.4 Kpa should prompt consideration of liver biopsy and/or initiation of antiviral therapy.

Keywords: Antiviral thresholds; Chronic hepatitis B; Fibrosis; Inactive HBsAg carriers; Liver histopathology; Non-invasive biomarkers.

Fig. 1

Overall liver tissue pathology of IHC patients according to the age(above 30 years old or not) and HBV DNA grouping: (A) all patients according to the HBV DNA group; (B) all patients grouped by age(above 30 years old or not) and HBV DNA positivity. Note: G: inflammation stage; F: fibrosis stage; EHI (G ≥ 2 and/or F ≥ 2), *: P < 0.05; ***: P < 0.001;ns: non significantly

Fig. 2

(A) Ranking of clinical variables for predicting perioperative complication by the Lasso regression: sex, HBV. DNA_group and LSM; (B) Multiple non-invasive models for diagnosing histologically defined EHI using ROC curves(AUC = 0.747); (C) ROC curves for the model in the training and internal test cohorts, the training cohort (blue curve) achieved an AUC of 0.747 (95% CI: 0.658–0.831), while the internal test cohort (red curve) yielded an AUC of 0.720 (95% CI: 0.579–0.861). (D) Predicted probabilities of EHI in IHC patients over 30 years old: the model predicted distinct probabilities of liver fibrosis across subgroups: male patients aged > 30 years with undetected HBV DNA and LSM ≥ 6.4 kPa exhibited the highest probability (72.8%), significantly surpassing that of female patients with detectable HBV DNA and LSM < 6.4 kPa (6.9%). The probability for other patient subgroups was 40.1%. These differences were statistically significant (χ² = 25.9, P < 0.001). Note: ROC: Receiver operating characteristic; AUC: area under the receiver operating characteristic curve; FIB-4: Fibrosis-4; APRI: aspartate aminotransferase-to-platelet ratio index; LSM: liver stiffness measurement