Current guidelines recommend that all newborns of HBsAg-positive mothers receive hepatitis B immunoglobulin and the first dose of recombinant HBV vaccine within 12 hours of birth, followed by vaccine boosters at one and six months of age. In addition, HBsAg-positive pregnant women with high viral loads are often treated with directly acting antiviral agents during the last trimester. Yet despite these measures, occult HBV infection—defined as the persistence of viral DNA in the absence of surface antigen detectable by current clinical tests—has been reported in newborns worldwide at rates ranging from 3.1 to 64 percent, depending on the sensitivity of the assays used. The mechanism of vertically acquired occult infection in the context of maternal antiviral therapy and neonatal immunoprophylaxis remains unclear.

A new study published in Frontiers in Microbiology by a team led by Liang Xu of Tianjin Second People’s Hospital and Tomasz I. Michalak of Memorial University of Newfoundland set out to reassess the occurrence of occult HBV infection in seemingly HBV-negative infants who received enhanced passive-active immunoprophylaxis. The investigators followed 13 mother–child pairs recruited at the Antenatal Clinic of Tianjin Second People’s Hospital between October 2014 and July 2016. The mothers were divided into three groups: four with consistently low viral loads below 500 genome copies per milliliter who received no antiviral therapy, eight with initially high viral loads who were treated with telbivudine from 24 to 28 weeks of pregnancy until delivery, and one untreated mother with a viral load of 1.5 × 10⁸ genome equivalents per milliliter one month before parturition.

All infants received 100 international units of hepatitis B immunoglobulin intramuscularly within 12 hours of birth, with an additional dose administered 14 days later or as indicated. The recombinant HBV vaccine was given at 10 micrograms per injection within the first 12 hours after birth and at one and six months thereafter. When antibodies to the surface antigen declined below 200 milli-international units per milliliter during follow-up, additional vaccine doses were given at eight months of age or later. The investigators also collected serial blood samples from the infants at two days of life and at two, seven, 12 and 24 months of age, along with maternal peripheral blood plasma and mononuclear cells and umbilical cord blood plasma and mononuclear cells at delivery.

The central technical innovation of the study was the application of a high-sensitivity molecular detection approach that combines polymerase chain reaction with signal detection by nucleic acid hybridization. This assay can detect HBV genomes with a sensitivity below 10 virus genome copies per milliliter of serum or per microgram of total DNA, far exceeding the 500 genome equivalents per milliliter sensitivity of the clinical quantitative PCR used for routine testing. The investigators applied this approach to whole blood clots from the infants, which contain both blood cells and serum, and to maternal and cord blood mononuclear cells isolated by density gradient centrifugation.

The results were striking. All children were serum HBsAg negative at every time point throughout the 24-month follow-up, and HBV DNA was not detected by the available clinical assay. However, when the high-sensitivity approach was applied, HBV DNA was identified in 71.4 percent of infants tested at two months, 77.8 percent at seven months, 60 percent at 12 months, and two of three tested at 24 months. Overall, 11 of 13 children—84.6 percent—were HBV reactive at one or multiple time points during the two-year observation period, at levels of ≤100 genome copies per microgram of DNA. Only two infants, both born to untreated mothers with low viral loads, were HBV DNA negative throughout the entire follow-up. The estimated viral loads ranged from 8.1 × 10¹ to 2.1 × 10³ genome equivalents per milliliter, with a median of 1.35 × 10².

The same prophylactic protocol, however, inhibited the development of clinically overt HBV infection for at least two years after birth. Antibodies to the surface antigen were detected in 100 percent of infants from two months of age onwards, and the surface antigen itself was never detected. The investigators noted that antibodies to the viral core antigen circulated in all newborns at two days and two months, declined to 58.3 percent at 12 months, and then rebounded to 90.9 percent at 24 months of age. This biphasic profile may suggest a decline in anti-HBc transferred from mothers and the initiation of de novo antibody synthesis in response to ongoing low-level infection.

Notably, the treatment of mothers with telbivudine did not influence the incidence of occult infection in their children. Among infants born to treated mothers, 16 of 19 samples collected during follow-up were HBV DNA positive by the high-sensitivity assay, compared with four of eight samples from infants born to untreated mothers. All infants born to telbivudine-treated mothers were found HBV DNA reactive at some point during the observation period. The investigators concluded that gestational antiviral therapy in the last trimester did not halt the establishment of occult infection, possibly because fetal exposure to relatively high viral doses occurred earlier in pregnancy, before the initiation of maternal therapy.

Perhaps the most unexpected finding emerged from sequencing of the 360-base-pair HBV preS-S genomic fragment. In four of the five mother–child pairs examined, the viral sequences identified in the infants were homologous with those in umbilical cord blood mononuclear cells but clearly diverged from the corresponding sequences in maternal plasma or maternal peripheral blood mononuclear cells. In one pair, all clones from the cord blood and from the child at two and seven months of age carried an identical sequence homologous with the HBV C/D subgenotype, which diverged at 28 of 360 nucleotides from the sequence carried in all maternal plasma clones, which belonged to the C2 subgenotype. Among a total of 32 point mutations in the preS-S sequence in the children that differentiated from those in their mothers, 21 were shared by all four mother–child pairs, suggesting a common mechanism underlying this viral sequence dichotomy.

The investigators propose two interpretations of these findings: that circulating lymphomononuclear cells can transmit HBV during fetal development, and that either the placenta, the fetus, or both could be the sites where these cells acquire variant HBV. Prior studies have shown that maternal blood mononuclear cells carrying HBV can cross the placental barrier and facilitate intrauterine transmission, and that HBV-infected peripheral blood mononuclear cells can support active viral replication and serve as a reservoir of infectious virus. Because of the hepadnaviral replication strategy, which involves low-fidelity transcription of viral pregenomic RNA to DNA, HBV is highly prone to mutations, making compartmental sequence diversity between maternal and fetal milieus biologically plausible. The uniformity of HBV variants in newborns and cord blood cells is consistent with bidirectional cell traffic between mother and fetus.

These findings from the woodchuck model of hepatitis B, in which dams with resolved acute hepatitis carrying trace amounts of woodchuck hepatitis virus universally transmitted the virus to their offspring, provide a compelling parallel. In that model, infection in newborns was serum viral DNA positive but negative for surface antigen and antiviral antibodies, while viral genomes and molecular indicators of replication were consistently detected in peripheral blood mononuclear cells and organs of the immune system. This immunovirologically seronegative but molecularly evident infection, designated primary occult infection, has been extensively characterized and shows strong molecular and pathogenic similarities to the HBV profiles identified in children in the current study.

The investigators argue that subclinical HBV infection in infants born to HBsAg-positive mothers is far more frequent than previously reported and that virus transmission to newborns could be an inevitable consequence of maternal infection. They emphasize that there is currently no therapy able to completely eliminate HBV from the infected host, including treatment with tenofovir disoproxil fumarate recommended to prevent vertical transmission in pregnant women. The documented ability of HBV to integrate into the host genome immediately after contact with susceptible cells, and the indefinitely long persistence of prooncogenic integrations, may have serious long-term consequences. The study therefore underscores the tremendous importance of preventive vaccination of HBV-naïve infants and young adults to avert the possibility of vertical viral spread in the future, while the search for more effective antiviral agents and more stringent prevention strategies must continue to further reduce transmission from already infected women.

Subject of Research: Occult hepatitis B virus infection in infants born to HBsAg-positive mothers despite neonatal immunoprophylaxis

Article Title: Hepatitis B Immunoprophylaxis rarely prevents molecularly evident subclinical HBV infection in children born to HBsAg-positive mothers

Article References: Xu, L., Dai, Q., Yang, H., Su, R., Shi, Y., Chen, J., Hao, J., Hao, J., & Michalak, T. I. (2026). Hepatitis B Immunoprophylaxis rarely prevents molecularly evident subclinical HBV infection in children born to HBsAg-positive mothers. Frontiers in Microbiology, 17, Article 1933288. https://doi.org/10.3389/fmicb.2026.1933288

Image Credits: AI Generated

DOI: 10.3389/fmicb.2026.1933288

Keywords: hepatitis B virus, occult HBV infection, mother-to-child transmission, immunoprophylaxis, HBV DNA, peripheral blood mononuclear cells, telbivudine, preS-S sequencing, placental transmission, anti-HBc, hepatocellular carcinoma, woodchuck hepatitis virus model