Association of HBeAg decline rate from mid-pregnancy to delivery with HBeAg seroconversion after delivery in hepatitis B virus-infected mothers

Abstract

There is still controversy about whether to continue antiviral therapy (AVT) after delivery, especially for pregnant women in the immune tolerance (IT) phase. In this study, a retrospective cohort study was conducted to explore the relationship between hepatitis B e antigen (HBeAg) decline rate (%) from mid-pregnancy to delivery and HBeAg seroconversion postpartum among patients using nucleos(t)ide analogs (NAs) to prevent mother-to-child transmission (MTCT), with the goal of identifying the ideal candidates for postpartum AVT continuation. This retrospective cohort study included 151 postpartum women. Univariate and multivariable logistic regression analyses were conducted to assess the association between the HBeAg decline rate (%) from mid-pregnancy to delivery and HBeAg seroconversion postpartum. Receiver operating characteristic (ROC) analysis was utilized to evaluate the predictive capacity of the HBeAg decline rate (%) and determine the optimal cut-off point. The univariate analysis revealed a significant association between the HBeAg decline rate (%) and HBeAg seroconversion postpartum (OR 1.068, 95% CI: 1.034–1.103, p < .001). In the multivariate regression analysis, adjusting for age, hepatitis B surface antigen (HBsAg) titre (log10 IU/mL) at mid-pregnancy, HBeAg titre (log10 S/CO) at mid-pregnancy, and hepatitis B virus (HBV) DNA load decline rate (%) from mid-pregnancy to delivery, the HBeAg decline rate(%) remained significantly associated with HBeAg seroconversion postpartum (OR 1.050, 95% CI: 1.015–1.093, p = .009). Then HBeAg decline rate (%) was treated as a categorical variable (tertiles) for sensitivity analysis. In the three distinct models, taking Tertile1 as a reference, women in Tertile3 still had a 4.201-fold (OR 4.201, 95% CI: 1.382–12.773, p = .011) higher risk of developing HBeAg seroconversion (p for trend <.05) after adjusting above covariates. The area under the curve (AUC) was 0.723 (95% CI: 0.627–0.819). The optimal cut-off value was 5.43%, with a sensitivity of 0.561, specificity of 0.791, and Youden's index of 0.352.A higher HBeAg decline rate (%) from mid-pregnancy to delivery independently correlated with an increased risk of HBeAg seroconversion postpartum. This decline rate can serve as a valuable clinical indicator for predicting HBeAg seroconversion.