Of 236 HBsAg-positive pregnant women screened, 187 were enrolled (details are shown in Fig. 1). Groups A–C comprised 65, 60, and 62 women, respectively. Overall, 27 (14.44%) women were lost to follow-up, primarily due to the COVID-19 pandemic. Therefore, 160 patients were included in the final analysis (Fig. 1). Of these women, nine delivered at 35–36 weeks of gestation, while the others delivered at 37–41 weeks of gestation. These women gave birth to 161 infants (including one set of twins), comprising 86 boys and 75 girls. There were two cases of fetal malformation, one case of syndactyly, and one case of preauricular fistula. Six infants were of low birth weight. All these 161 infants (including those with malformations or low birth weight) were included in the analysis, and all tested negative for HBsAg and positive for anti-HBs, except one who tested negative for anti-HBs after completing the vaccination course (Table S1).
Patients enrollment and study design. HBsAg Hepatitis B surface antigen; HBeAg Hepatitis B e antigen; TDF Tenofovir disoproxil fumarate; HBV Hepatitis B virus; ALT alanine aminotransferase
Mothers’ demographic and clinical characteristics
Maternal demographics and clinical characteristics were comparable across all groups regarding age, multiparous, gestational age at AVT, gestational age at delivery, cesarean section rate, as well as baseline levels of HBsAg, HBeAg, HBV DNA, and ALT (assessed at enrollment prior to TDF initiation), and HBV DNA and ALT levels at delivery (all P > 0.05). All groups demonstrated a > 4 lg IU/mL reduction in HBV DNA levels at delivery compared to baseline (Table 1).
In addition, 6 mothers received antiviral prophylaxis to prevent MTCT of HBV during a previous pregnancy: 2 in Group A (1 stopped antivirals at delivery and the other stopped antivirals at 12 weeks postpartum), 1 in Group B (stopped antivirals at 4 weeks postpartum), 3 in Group C (1 stopped antivirals at delivery and 2 stopped antivirals at 12 weeks postpartum).
Elevated postpartum ALT levels
Incidence and severity of elevated ALT levels
During the first 48 weeks postpartum, 91 (56.88%) of 160 mothers developed elevated ALT levels (> 1 × ULN), and 5 (3.13%) of 160 mothers exhibited ALT exacerbation (> 10 × ULN), with no significant differences among groups (P = 0.338 and P = 0.103, respectively) (Table 2). These findings suggest that TDF therapy duration had limited impact on long-term ALT levels. Within 12 weeks postpartum, the incidence of elevated ALT levels was 57.63% in Group A, 49.02% in Group B, and 40.00% in Group C, with no significant difference among the three groups (P = 0.186) (Table 2). A cross-threshold analysis revealed consistently higher elevated ALT level rates in Group A compared to the combined Groups B and C, though statistically non-significant: >1 × ULN (64.41% vs. 52.48%, P = 0.141), > 5 × ULN (11.86% vs. 4.95%, P = 0.197), and > 10 × ULN (6.78% vs. 0.99%, P = 0.062) (Table S2). All five ALT exacerbations occurred within 12 weeks postpartum, identifying this period as the critical monitoring window.
Among 91 mothers with elevated postpartum ALT levels, further analysis revealed no statistically significant difference in the severity of elevated ALT levels among the three groups (P = 0.558) (Table 2) or between Group A and the combined Groups B and C (P = 0.254) (Table S2). Mild elevated ALT levels was observed in 86.81% (79/91) mothers: 81.58% in Group A, 92.59% in Group B, and 88.46% in Group C. ALT flare was observed in 7.69% (7/91) mothers: 7.89% in Group A, 7.41% in Group B, and 7.69% in Group C. ALT exacerbation was observed in 5.50% (5/91) mothers: 10.53% in Group A, 0.00% in Group B, and 3.85% in Group C (Table 2). While no significant intergroup differences in peak ALT levels were observed (P = 0.401) (Table 2; Fig. 2a), it was noteworthy that all three cases with extreme elevated ALT levels (510–968 U/L) occurred in Group A, suggesting potential association between early TDF withdrawal and severe hepatic flares. No significant bilirubin elevation (> 2 × ULN) was observed in mothers with elevated ALT levels during follow-up, and no acute viral infections were detected concurrent with ALT abnormalities.
(a) Comparison of peak ALT levels among three groups. Peak ALT levels in 91 mothers with elevated postpartum ALT levels. (b) HBV DNA levels at the onset of elevated ALT levels in three groups. HBV DNA levels at the onset time of elevated ALT levels in 91 mothers with elevated postpartum ALT levels
Onset times of elevated ALT levels
The onset times of elevated ALT levels are shown in Table 3. The median onset time was 8 weeks postpartum in all groups, with no significant difference (P = 0.798). Among the 91 mothers with elevated ALT levels, 63 (69.23%) exhibited the initial elevated ALT levels at approximately 6 weeks postpartum (follow-up time point of 6 ± 2 weeks postpartum), 16 (17.58%) at 12 weeks postpartum, 5 (5.49%) at 18 ± 2 weeks postpartum, 4 (4.40%) at 24 weeks postpartum, 3 (3.30%) at 36 weeks postpartum, and none at 48 weeks postpartum. There were no differences in the elevated ALT levels proportions at each follow-up time point among groups (P = 0.561) (Table 3).
Comparison of HBV DNA levels at onset of elevated ALT levels
The median HBV DNA levels at the onset of elevated ALT levels in Groups A–C were 8.03 (7.50–8.21), 5.61 (4.11–7.95), and 3.35 (2.00–4.94) lg IU/mL, respectively (H = 36.98, P = 0.000). The HBV DNA level was significantly lower in Group B than Group A, and in Group C than in Groups A and B (Group A vs. Group B: P = 0.012, Group A vs. Group C: P = 0.000, Group B vs. Group C: P = 0.010, all P values had been adjusted for multiple comparisons) (Fig. 2b).
Risk factors for elevated ALT levels
All mothers were divided into normal ALT group and elevated ALT group according to whether they had elevated postpartum ALT levels (Fig. 1). There were no statistically significant differences between the two groups in age, mode of delivery, breastfeeding, baseline HBsAg, HBeAg, HBV DNA level, elevated ALT during pregnancy, HBV DNA level at delivery, or TDF withdrawal timing (P > 0.05 for all) (Table 4). The elevated ALT group demonstrated significantly higher ALT levels at delivery than the normal ALT group (24 vs. 18 U/L, Z = − 4.386, P = 0.000) (Table 4). Binary logistic regression analysis showed that ALT level at delivery was independently correlated with postpartum ALT elevation (odds ratio = 1.098, 95% confidence interval [CI], 1.039–1.160, P = 0.001). Receiver operating characteristic (ROC) curve analysis demonstrated that delivery ALT level had modest diagnostic utility for predicting postpartum ALT elevation, with an area under the curve of 0.703 (95% CI, 0.622–0.785) (Fig. 3), and the cut-off value was ≥ 23 U/L for the Youden index. An ALT level of ≥ 23 U/L showed a positive predictive value of 81.97% for the occurrence of elevated postpartum ALT levels, and the sensitivity and specificity were 54.95% and 84.06%, respectively (Table 5).
Receiver operating characteristic curve analysis of ALT level at delivery in identifying patients with elevated postpartum ALT levels
Among 61 mothers with ALT levels ≥ 23 U/L at delivery in the three groups, elevated ALT levels were observed in 50 (81.97%) mothers: 22/23 (95.65%) in group A, 14/17 (82.35%) in group B, and 14/21 (66.67%) in group C. A statistically significant difference in the incidence of elevated ALT levels was observed between group A and group C (P = 0.048), while no significant differences were observed between Group A vs. Group B (P = 0.591) or Group B vs. Group C (P = 0.711), all P values had been adjusted for multiple comparisons.
Re-AVT or long-term AVT postpartum
Among the 91 mothers with elevated postpartum ALT levels, 13 (14.29%) received re-AVT or long-term AVT postpartum (8 in Group A, 3 in Group B, and 2 in Group C) (Table S3), with no statistically significant differences among groups (P = 0.179). All developed an elevated ALT level within 12 weeks postpartum, with peak ALT levels (181–968 U/L) occurring within this period in 76.92% (10/13) of cases. 10 mothers (7 in Group A, 2 in Group B, and 1 in Group C) developed ALT > 5 × ULN (HBV DNA 7.13–8.93 lg IU/mL) after TDF withdrawal and hence recommenced AVT. Notably, 2 mothers with postpartum ALT levels < 200 U/L, 1 in Group A (peak ALT 197 U/L, HBV DNA 8.58 lg IU/mL) and the other in Group B (peak ALT 181 U/L, HBV DNA 8.23 lg IU/mL), also received re-AVT due to absent remission. 1 mother in Group C developed ALT exacerbation (peak ALT 463 U/L at 8 weeks postpartum, HBV DNA 3.72 lg IU/mL) while taking TDF and received long-term AVT.
Breastfeeding rates
A total of 47 of 59 mothers in Group A, 13 of 51 in Group B, and 21 of 50 in Group C chose to breastfeed. The breastfeeding rate was significantly higher in Group A (79.66%) than in Group B (25.49%) and Group C (42.00%), both P values were 0.000. There was no significant difference in breastfeeding rate between Groups B and C (P = 0.079).