The effects of increased dose of hepatitis B vaccine on mother-to-child transmission and immune response for infants born to mothers with chronic hepatitis B infection: a prospective, multicenter, large-sample cohort study

This was a prospective, multicenter, large-sample study. Patients were recruited from 4 hospitals in Beijing, Shijiazhuang, Taiyuan, and Tongliao, China. We evaluated and compared the effects of 20 μg HB vaccine on infants born to HBsAg-positive mothers, including immunoprophylaxis failure, immune responses to vaccine, and vaccine safety. The HBsAg-positive mothers were enrolled at 24–28 weeks’ gestation, and peripheral blood samples were collected at parturition for chemical and hematological tests. Demographic information of their infants was recorded at birth, including sex, singleton status, gestational age, birth weight, delivery mode, and 1-min APGAR scores as the baseline data. The infants were divided into two groups according to their mothers’ wishes: 10 μg (0.5 mL) recombinant HB vaccine plus HBIG (10 μg group) and 20 μg (1 mL) recombinant HB vaccine plus HBIG (20 μg group). The details of informed consent are described in Additional file 1: Appendix 1, Methods. All the vaccinations were completed at the corresponding investigational sites, and the adverse events were recorded at each follow-up at 1, 6, and 12 months. At 12 months, peripheral serum samples were taken from infants after standard immunoprophylaxis, and their HBsAg, anti-HBs, HBeAg, anti-HBe, and HB core antibody (anti-HBc) were tested. If the infant was positive for HBsAg, HBV DNA was further tested. The fetal development and infant growth were evaluated at 12 months in both HB vaccine groups. The study protocol was approved by each institutional Ethics Committee and registered at Chinese Clinical Trial Registry (ChiCTR, No. ChiCTR-PRC-09000459).

Patient screening began from January 2009 to September 2010, and the last patient visit was on October 2011. The inclusion criteria for the mothers were as follows: HBsAg positive for > 6 months; age 18–45 years; and willing to cooperate with collection of documented information from 24 to 28 weeks’ gestation, the corresponding intervention measures, follow-up, and detection according to the informed consent. Major exclusion criteria for mothers were (1) infection with hepatitis C/D, human immunodeficiency virus, Treponema pallidum or Toxoplasma gondii; (2) treated with HBIG or antiviral therapy including interferon within 6 months before or during pregnancy; (3) alanine aminotransferase (ALT) ≥ 2× upper limit of normal (ULN) or total bilirubin (TBIL) ≥ 2 mg/dL (34.2 μmol/L), indicating cirrhosis and other liver diseases; (4) malignant tumor or definite disease in the cardiovascular, respiratory, urinary, nervous, digestive, blood, endocrine, or metabolic systems; (5) long-term use of hormones or immunosuppressive agents; and (6) taking part in other studies. The major exclusion criteria for infants were (1) prematurity (born at less than 36 weeks’ gestation), (2) birth weight < 2000 g, (3) congenital malformation, and (4) taking part in other studies.

All infants born to HBsAg-positive mothers had the following prophylaxis schedule: the first dose of 200 IU HBIG (Chengdu Institute of Biological Products, China or Hualan Biological Engineering Inc., China) and the first dose of 10 μg (0.5 mL) or 20 μg (1 mL) recombinant HB vaccine (Hansenula yeast vaccine; Dalian Hissen Biopharm Co., China) were given intramuscularly within 2 h of birth at different sites. The second injection of the same dose of HBIG was administered at 1 month of age. The second and third doses of recombinant HB vaccines were given at 1 and 6 months of age, respectively.

All serum specimens were tested in the hospital central laboratory. The presence of HBsAg, anti-HBs, HBeAg, anti-HBe, and anti-HBc was determined using an electrical chemiluminescence immunoassay (Roche Laboratories, Mannheim, Germany) or chemiluminescent microparticle immunoassay kit (Architect i2000 analyzer; Abbott Diagnostics, Abbott Park, IL, USA). All the serum samples for HBV DNA were tested by real-time polymerase chain reaction with a range of 2–8 log₁₀ IU/mL (Hunan Shengxiang Bio-engineering). ALT was measured by a fully automatic biochemical analyzer (AU5400; Olympus Optical, Tokyo, Japan). ALT > 40 IU/L was considered abnormal.

The primary outcome was immunoprophylaxis failure, which was defined as infants who were HBsAg-positive at age 12 months [15, 16]. The secondary outcome was response status of infants to HB vaccine. All the HBsAg-negative infants (successful immunoprophylaxis) were divided into 4 groups as follows: anti-HBs < 10 IU/L was defined as non-responder, anti-HBs 10–100 IU/L was low-responder, anti-HBs 100–1000 IU/L was medium-responder, and anti-HBs ≥ 1000 IU/L was high-responder [13].

The database was established with EpiData 3.02. Continuous variables values were expressed as the mean ± standard deviation (SD); categorical variables were expressed as percentages. The characteristics of infants who received different doses of the HB vaccine were compared by independent t-test and/or χ² test or Fisher’s exact test. The maternal HBV DNA level, ALT, and TBIL in two groups were compared by the Mann-Whitney U-test. The immunoprophylaxis failure rates (MTCT of HBV) in two doses of HB vaccines were cooperated, and the planned sample size of 955 patients was estimated to provide at least 85% power to detect an absolute difference of 3% in the proportion of infants with HBV infection at 12 months on the basis of two-tailed α = 0.05 and assuming an MTCT rate of 5% in the 10 μg HB vaccine group. A multivariate logistic regression model was fitted with a stepwise method (likelihood ratio test) using significant baseline characteristics (candidate variables such as mother age, HBV DNA ≥ 5 log₁₀ IU/mL, and other clinical indicators with P < 0.20 in Table 1) that had been prefiltered in univariate analysis to identify factors independently associated with vaccine dose grouping. The response rates of different degrees in infants between two doses of groups were compared by χ² test or Fisher’s exact test. The developmental indexes of infants at 12 months between two groups were compared by independent t-test or Mann-Whitney U-test. The data were analyzed using SPSS version 17.0 (SPSS Inc., Chicago, IL, USA) or GraphPad Prism version 5.0 (GraphPad Software Inc., San Diego, CA, USA). A two-sided P < 0.05 was considered statistically significant.

The study was performed at 4 investigational sites in China. A total of 1004 infants born to 1001 HBsAg-positive mothers (3 mothers had twins) were enrolled, and the numbers of patients enrolled from each site were as follows: 668 in Beijing, 108 in Taiyuan, 125 in Shijiazhuang, and 100 in Tongliao. At 12 months, a total of 955 infants completed the study, there were 478 infants who completed the follow-up in the 10 μg group, and 477 infants completed the follow-up in 20 μg HB group (Fig. 1); the groupings of these infants in respective site are showed in Additional file 1: Appendix 2 Table S1. The demographic characteristics of infants at birth and their mothers with chronic HBV infection in each group are shown in Table 1.

As more HBeAg-positive mothers with or without high load of HBV DNA prefer to choose high dose of HB vaccine (20 μg) for their infants, therefore, both the HBeAg-positive rate and HBV DNA levels in mothers in the 20 μg group were higher than those in the 10 μg group (both P < 0.001, Table 1). Further, we used multivariable logistic analysis to screen the possible confounders (in Table 1) that lead to the differences between the high and low vaccine groups. Results showed that the high level of maternal HBV DNA ( ≥5 log₁₀ IU/mL) was the independent factor of differences in baseline characteristics between two dose groups (P < 0.001) (OR = 0.481, 95% CI 0.362–0.639). Therefore, we did stratified analysis in the following investigation according to the maternal HBV DNA level.

Of the 955 infants after the standard immunization schedule, 13 infants were positive for HBsAg at 12 months, who were considered as immunoprophylaxis failure, and 942 were negative for HBsAg, who were considered as immunoprophylaxis success. The total rate of immunoprophylaxis failure was 1.4% (13/955). The characteristics of infants and their mothers between the two outcomes are shown in Additional file 1: Appendix 2 Table S2; high level of maternal HBV DNA and HBeAg-positive were the major differences between immunoprophylaxis failure and success. Among the 13 HBsAg-positive infants, 5 were in the 10 μg group and 8 were in the 20 μg group. The baseline characteristics of the infants at birth and HBV infection status at age 12 months are shown in Table 2. At 12 months, all the 13 HBsAg-positive infants were positive for HBeAg and anti-HBc.

Regarding the influence of high level of maternal HBV DNA, we compared the immunoprophylaxis failure rate and response status of infants between 10 μg and 20 μg groups by stratified analysis. There were 601 infants born to HBsAg-positive mothers with HBV DNA < 5 log₁₀ IU/mL, 350 were in the 10 μg group and 251 were in the 20 μg group. None of the infants was positive for HBsAg in the 10 μg group (0%, 0/350), and 1 was positive in the 20 μg group (0.4%, 1/251); the immunoprophylaxis failure rates between two groups were not significantly different (P = 0.418, Table 3). However, the high-response rate in the 20 μg group (42.2%, 106/251) was evidently higher than that in the 10 μg group (22.9%, 80/350) (P < 0.001); on the contrary, the low-response rate (P = 0.002) and middle-response rate (P = 0.022) in the 20 μg group were both lower than those in the 10 μg group (Table 3). There was no difference in non-response rate between two groups (Table 3).

In the 354 infants born to mothers with high load of HBV DNA (≥ 5 log₁₀ IU/mL), 128 were in the 10 μg group and 226 were in the 20 μg group. Among them, 5 infants (3.9%, 5/128) were HBsAg-positive in the 10 μg group and 7 infants (3.1%, 7/226) were HBsAg-positive in the 20 μg group. The immunoprophylaxis failure rates between two groups were not obviously different (P = 0.922). Interestingly, there were no significant differences in response rates of various levels between two groups, from non-response to high-response (Table 3). These were different from the infants born to mothers with low load of HBV DNA.

Of the 955 infants who finished follow-up, the adverse events were reported in 9 infants: 4 in the 10 μg HB vaccine group (0.8%, 4/478) and 5 in the 20 μg HB vaccine group (1.0%, 5/477). Among the 9 infants, 5 had adverse injection reactions (local swelling and induration) to the first dose of vaccine (2 in the 10 μg group and 3 in the 20 μg group). Additionally, 2 cases developed fever (1 in each group), and 2 had hives (10 μg group). No severe adverse events were reported to vaccination.

We also evaluated the safety of 20 μg HB vaccine by observing the growth indexes of infants at age 12 months. There were no differences in these growth indexes between the two groups, except for body length, which was longer in the 20 μg vaccine group (P = 0.04) (Table 4), but still within the normal range of Chinese children’s growth and development indicators [17]. The results suggested that 20 μg HB vaccine is safe for infants.