We would like to thank Zhou for their correspondence on our Review (Terrault, N. A. et al. Viral hepatitis and pregnancy. Nat. Rev. Gastroenterol. Hepatol. 18, 117–130 (2021))1 and for an opportunity to expand on in utero transmission of hepatitis B virus (HBV) (Zhou, Y.-H. Evidence against in utero transmission of hepatitis B virus. Nat. Rev. Gastroenterol. Hepatol. https://doi.org/10.1038/s41575-021-00455-z (2021))2. Although the estimate of 10% incidence might be high, we believe there is ample evidence to support in utero HBV infection.

In utero infection via germline is possible. HBV DNA could be detected in ova of women with HBV infection3 and in embryos of hepatitis B surface antigen (HBsAg)-discordant couples4. Placental infection is another mechanism of in utero infection, with HBV (detected by in situ hybridization) found in a gradient of infected placental cell layers from the maternal to the fetal side and presence of placental HBV DNA, particularly in endothelial cells, associated with fetal infection in newborns5. Fetal contamination by maternal blood during invasive prenatal tests could also be a cause of in utero infection, with maternal HBV DNA levels greater than 6-log IU/mL or presence of hepatitis B e antigen (HBeAg) defining a higher risk of in utero infection6.

Clinical studies also support the plausibility of intrauterine infection. In a large, randomized trial of tenofovir disoproxil fumarate (TDF) versus placebo, two of three infants who acquired HBV infection, all in the placebo arm, had a very high HBV DNA load on the day of birth. This observation is difficult to explain unless there was established infection prior to birth7. Similarly, another study found that infant birth blood HBV DNA positivity was significantly lower in TDF-treated mothers than in untreated mothers (6% versus 31%)8. These studies highlight the role of maternal antiviral therapy in reducing intrauterine infection.

It is important to emphasize that exposure to HBV in pregnancy does not equate with infection. HBsAg, HBeAg, anti-HBe, anti-HBc and HBV DNA are detectable in amniotic fluid from 16-23 weeks and in fetal blood from 24-32 weeks of gestation, but only a minority have post-natal infection9. Embryonically, liver development with production of bile is evident only after 12 weeks gestation; when fetal hepatocytes can support HBV replication is unknown. Studies have shown that fetal HBV exposure through the placenta can induce innate immune cell maturation with a predominant T helper 1 cell profile and suppression of proinflammatory cells10. Prolonged fetal exposure to maternal HBeAg through the placenta coupled with the high cross-reactivity of HBeAg and HBcAg, in terms of T helper cell recognition, might account for the lack of anti-HBc immunoglobulin M in newborns. Further, the altered immune milieu in utero (compared with adult-acquired infection) might account for the absence of symptomatic acute HBV infection in newborns.

Provision of third trimester maternal TDF therapy and timely neonatal immunization have reduced mother-to-child transmission (MTCT) to very low rates (<1%)1. The rare cases of prophylaxis failure might be the consequence of in utero infections when fetal hepatocytes are susceptible to HBV and/or with compromise of the placental barrier. Given the safety of TDF in pregnancy, we believe consideration of earlier antiviral treatment during pregnancy in certain high-risk women is warranted.

Thus, while the predominant risk of MTCT of HBV occurs peripartum, there is strong evidence for the existence of in utero infection and the continued efforts to improve prophylactic measures, including consideration of maternal antiviral therapy earlier than the third trimester, might reduce MTCT to zero.