Neonatal immunization with hepatitis B (HB) vaccine is highly effective; however, more needs to be learned about the duration of protection and indications for boosters. We measured antibody to HB core antigen (anti-HBc), HB surface antigen (HBsAg), and pre- and postbooster titers of HBsAg antibody (anti-HBs) 15 years after primary neonatal immunization with plasma-derived HB vaccines in 2 cohorts of 15-year-old children. Group A consisted of 78 children who were born to HB e antigen–positive HBsAg carrier mothers and had developed protective levels of anti-HBs antibodies (≥10 mIU/mL) following HB immunization. Group B consisted of 113 apparently healthy children whose anti-HBs titers after vaccination were unknown. Anti-HBs was undetectable (antibody titer <10 mIU/mL) in 29.9% in group A and 62.4% in group B (P < .001). Anti-HBc was detected in 33.3 % in group A and 4.4 % in group B (P < .001). After a single booster dose of HB vaccine, 2.7% in group A and 3.3% in group B remained anti-HBs–negative. A blunted serological response was noted in approximately 20% in both groups. One HBsAg carrier was detected in group A (1.3%) and 4 in group B (3.5%). Fifteen years after neonatal immunization with plasma-derived HB vaccine, a large proportion of children exhibited waning immunity. This poses the risk of breakthrough infection. A single booster augmented the serological response to the vaccine in most but not all subjects. In conclusion, our findings suggest that one or more booster immunizations are needed in seronegative subjects by at least 15 years following neonatal immunization with plasma-derived HB vaccine. (HEPATOLOGY 2004;40:1415–1420.)
Although substantial progress has been made in preventing hepatitis B (HB) infection and its complications through the use of HB vaccines,1, 2 there are still 400 to 500 million HB carriers worldwide.3 The large number of HB carriers presents a major threat to new generations. Thus worldwide access to HB immunization and appropriate use of HB vaccines continues to be critical to further reduce the burden of this disease.
HB vaccines became available in the early 1980s. They are highly immunogenic and efficacious.4–8 Most vaccines induce protective antibody against HB surface antigen (anti-HBs) after a primary series of immunizations. This is followed by a gradual decline in titer.9 Vaccinees who have low or undetectable anti-HBs antibodies are reported to exhibit a robust anamnestic response up to 10 years after primary immunization.10–13 Currently no vaccinee has become a carrier 5 to 10 years after primary immunization; thus many investigators consider boosters to be unnecessary at 10 years of age.10–16 In 2000, the European Consensus Group on Hepatitis B Immunity went one step further. They issued a statement recommending against the use of boosters of HB vaccine in immunocompetent individuals 15 years after primary immunization.17 Some investigators continue to recommend boosters18–20 because of the progressive decline of anti-HBs over time and the associated potential risk of development of HB infections.
Children born to mothers positive for HB surface antigen (HBsAg) and HB e antigen (HBeAg) are at a high risk of vertical transmission of HB virus (HBV).21 Their siblings and other family members are also more likely to be positive for HB through vertical and/or horizontal spread of HBV. There is also an increased risk of spread of HBV through sexual contact as the children mature. For these reasons, it is important to determine the need for a booster dose of HB vaccine in this high-risk group.
In this report, we describe the seromarkers and immune response to HBV before and after booster injections in two cohorts of children born during the nationwide HB vaccination program begun in Taiwan in 1984.22 One cohort consisted of high-risk children born to HBeAg-positive and HBsAg-positive mothers. The other cohort consisted of apparently healthy children representative of the general population. All of the children had received 4 doses of plasma-derived HB vaccine 15 years prior to follow-up. We provide evidence of waning immunity in a substantial proportion of these cohorts. In addition, we report several breakthrough HB infections, including one in an individual proven to be positive for anti-HBs after vaccination. We also note a high risk for acquisition of the antibody against HB core antigen (anti-HBc), and a blunted response to one or two boosters in some subjects.
HB, hepatitis B; anti-HBc, antibody to hepatitis B core antigen; HBsAg, hepatitis B surface antigen; anti-HBs, hepatitis B surface antigen antibody; HBV, hepatitis B virus.
Patients and Methods
Two groups of children were recruited. Group A consisted of 78 15-year-old adolescents (40 male and 38 female) who were born to HBeAg-positive and HBsAg-positive mothers. They were documented to have received 4 doses of plasma-derived (5 μg/dose) HB vaccine (Hevac B; Pasteur, Paris, France) at 0, 1, 2, and 12 months during 1988. All of them also received 0.5 mL (145 IU) of hepatitis B immunoglobulin (Hyper-Hep, Miles, PA) within 24 hours after birth. All were shown to be negative for HBsAg and positive for anti-HBs at the age of 18 months. Group B included 113 15-year-old volunteer students (57 male and 56 female) who had completed 4 doses of plasma-derived HB vaccination (Hevac B; Pasteur) during infancy. Their serological status against HBV prior to the study was unknown.
After obtaining informed consent, all children underwent blood tests for HBsAg, anti-HBs, and anti-HBc. A booster dose of 20 μg of a recombinant DNA HB vaccine (Engerix-B; SmithKline Beecham, Rixensart, Belgium) was given to all subjects in group A and 63 children in group B who were HBsAg-negative and whose anti-HBs titer was less than 100 mIU/mL. Another blood sample was taken 4 weeks after booster vaccination.
Subjects in group B were offered a second dose of booster if they remained anti-HBs–seronegative 4 weeks after the first dose of HBV vaccine booster. A third blood sample was taken 4 weeks later in this subgroup.
Radioimmunoassays (Ausab, Ausria II, and Corab; Abbott Laboratories, North Chicago, IL) were used to assay anti-HBs, HBsAg, and anti-HBc. Anti-HBs concentrations greater than 10 mIU/mL were considered protective. Concentrations between 10 and 100 mIU/mL were considered low titers.16, 17, 23, 24 A carrier was defined as an individual who was HBsAg-positive for more than 6 months.
Genotyping for Hepatitis B Virus.
Genotype analysis for HBV was conducted by both restriction fragment length polymorphism analysis25 and direct sequencing of products of nested polymerase chain reaction using type-specific primers targeting pre-S/S region.26
A chi-square test was used to compare frequencies between groups. Fisher's exact test was used when the expected value was less than 5. A P value of less than .05 was considered significant.
Serological Studies in Children Born to Mothers Doubly Positive for HBsAg and HBeAg (Group A).
Group A consisted of 78 children who were born 15 years previously to HBV-infected mothers and were immunized with HB vaccine at birth. The anti-HBs titers at 18 months of age were available for 61 of the 78 children; 25 (41%) had an anti-HBs titer less than 100 mIU/mL. The distribution of anti-HBs titers at 18 months of age before and after a booster dose of recombinant vaccine at 15 years of age is shown in Table 1. The geometric mean titers of anti-HBs before and after a booster dose, together with those at 18 months of age, are shown in Fig. 1.
Table 1. Distribution of Anti-HBs Titers After a Booster Dose of Recombinant Vaccine at the Age of 15 Years and After the Neonatal Vaccineation at the Age of 18 Months in 78 Children Born to Mothers Doubly Positive for HBsAg and HBeAg (Group A)
NOTE. This group included one child (1.3%) who was HBsAg-positive and 26 children (33.3%) who were anti-HBc-positive. The HBsAg-positive child was excluded from the booster study. Numbers in parentheses denote case numbers.
Subjects were grouped by the prebooster anti-HBs titers at the age of 15 years.
One child was positive for HBsAg. This child was also documented to have received 4 doses of plasma-derived HB vaccines and acquired an anti-HBs titer of 21 mIU/mL at the age of 18 months. He remained HBsAg-negative at 7 years of age. HBV genotyping via restriction fragment length polymorphism analysis and nested polymerase chain reaction sequencing for the child and his mother revealed mixed genotypes B and C in the child, but only genotype B in the mother. There was no glycine 145 arginine mutation of the HB surface protein in the child. This child was excluded from the booster study.
Anti-HBc was detected in 26 (33.3%) of the 78 children. The long-term overall effectiveness of HB neonatal vaccination in maintaining HB disease-free and carrier-free at the age of 15 years were 98.7% (77/78) for all children and 96.2% (25/26) for children with evidence of HBV infection.
Anti-HBs was undetectable (<10 mIU/mL) in 23 (29.9%) of the 77 children who were HBsAg-negative. All but 2 of these 23 children (91.3%) exhibited a serological response to the booster. Although children with moderate or high antibody titers responded well to the booster, the increase of anti-HBs titers was blunted in children with low pre-booster anti-HBs titers (Fig. 1). The geometric mean titers for the group increased 15-fold to 353 mIU/mL following 1 dose of HBV vaccine booster. The geometric mean titers for children with a prebooster titer of less than 10 increased only sevenfold from 9 to 65 mIU/mL. A similar correlation between the anti-HBs titers at 18 months of age and booster responses at 15 years was noted. The lower the anti-HBs titers at the age of 18 months were, the lower the antibody responses to a booster dose at the age of 15 years would be (Fig. 1). Among the 46 children with negative anti-HBc and anti-HBs titers (≤100 mIU/mL) in this group, 21.7% had a less than 2-fold increase of anti-HBs titers after a booster vaccination.
The effect of the presence of anti-HBc on the titers of anti-HBs before and after a booster is shown in Table 2. Children who were seropositive for anti-HBc were significantly more likely to be seropositive to anti-HBs. The positive rate for anit-HBs was 88.0% (22/25) in anti-HBc (+) children and 61.5% (32/52) in anti-HBc (−) children (P = .018). The only 2 children who were unresponsive to a booster were both anti-HBs and anti-HBc negative.
Table 2. Distribution of Prebooster Anti-HBs Titers in 77 Children Born to HBsAg- and HBeAg-Positive Mothers With Respect to Presence of Anti-HBc at 15 Years of Age (Group A)
NOTE. Two children who were anti-HBc–negative and had an initial titer of less than 10 failed to demonstrate a booster response.
Prebooster anti-HBs titers at 15 years of age were used to categorize children.
Serological Studies in Unselected Healthy Children With Unknown Responses to Neonatal HB Vaccination (Group B).
Group B consisted of 113 children who were immunized with HB vaccine during infancy. The frequency of HBsAg, anti-HBc, and anti-HBs titers before and after a booster dose of recombinant vaccine is shown in Table 3. Four of these children (3.5%) were positive for HBsAg. Tracing of the maternal history revealed that 3 of the 4 were born to HBeAg-positive mothers and 2 of them received HB immunoglobulin after birth. Genotyping of the HBV was performed in 3 of the 4 child–mother pairs. HBV DNA could not be detected via nested polymerase chain reaction in 1 child and 2 mothers. The only child–mother pair with available genotyping showed both had genotype B HBV. The serological status after the primary vaccination was unknown; therefore, it is impossible to tell whether they became carriers because of vaccine failure or breakthrough infections. These HBsAg-positive children were excluded from the booster study. Anti-HBc was detected in 5 (4.4%) of the 113 children. This proportion is significantly lower than that of group A (P < .001). Four of these 5 anti-HBc–positive children were also positive for HBsAg.
Table 3. Hepatitis B Serological Markers and Anti-HBs Titers Before and After a Booster Dose of Recombinant Vaccines at the Age of 15 Years in 113 Volunteer Students (group B)
Before Booster (n = 109)
After First Booster (n = 63)
After Second Booster (n = 15)
NOTE. This group included 4 children (3.5%) who were HBsAg-positive and 5 children (4.4%) who were anti-HBc-positive. The HBsAg-positive children were excluded from the booster study.
Anti-HBs was undetectable in 68 (62.4%) of the 109 children who were HBsAg negative. This proportion was significantly higher than that of group A (P < .001). All but 2 of the 63 children (96.8%) exhibited a serological response to the first booster, and only 1 out of 15 failed to have a response to the second booster. All the children with moderate or high antibody titers responded well to the first or second booster.
HB vaccination programs are highly effective and have led to marked declines in carrier rates and the incidence of hepatocellular carcinoma in countries such as Taiwan that are highly endemic for HBV.1, 2, 6, 27 However, our results clearly show that natural exposures to HBV continued in the general population. Without reliable long-term immunity, HBV infection is unlikely to be eradicated. The current priority is to continue to protect vaccinated children who are exposed to HBV carriers in the household and community. To accomplish this, we must determine how long immunity persists, whether boosters are needed, and, if so, when and in whom they should be administered.
In a prior study of infants who had received recombinant HB vaccine in early infancy,13 we demonstrated that in anti-HBs–seronegative subjects at 10 years of age, one booster dose of HB vaccine induced a 100% anti-HBs seroconversion. Antibody titers rose to no less than 1,000 mIU/mL with a 900-fold increase in geometric mean titer. In the current study, we found unacceptably high rates of undetectable anti-HBs among children who had received 4 doses of plasma-derived HB vaccine as the neonatal immunization 15 years previously. Four (3.1%) of 128 children with anti-HBs levels of 100 mIU/mL or less did not respond to a HB booster vaccine. The increase of anti-HBs titers after booster at 15 years of age was often blunted, especially for children with low anti-HBs titers after the neonatal vaccination. These low responders tend to have low anti-HBs titers and respond inadequately to booster vaccinations at the age of 15, and hence carry the greatest risks of getting breakthrough HB infections.
The most alarming finding, however, was the detection of 1 HBsAg carrier among 78 children (1.3%) who had responded to the neonatal immunization. The child remained seronegative until age 7. At the age of 15, he was proved to have 2 different genotypes of HBV, though only 1 of them was similar to his mother's. Therefore, horizontal infections from one or more sources accounted for the seroconversion that happened between 7 and 15 years of age. Besides, no glycine 145 arginine mutation was detected in the HB surface protein of the child. Breakthrough infection due to waning and eventual loss of the vaccine protectiveness—rather than infection by escaped mutants—is likely.
Children living in a household with HB carriers are more likely to receive a natural booster for HBV. We found that children born to mothers who were doubly positive for HBsAg and HBeAg were much more likely to exhibit anti-HBc, a marker of a natural booster effect, than nonselected children representative of the general population. However, both groups continued to have high rates of undetectable antibody 15 years after primary immunization. In previous studies, we found that in Taiwanese children born to mothers who were doubly positive for HBsAg and HBeAg,13, 28 the anti-HBc–positive rates at 5 and 10 years were 12%. In the current study, the rate was 33.3% at 15 years of age. Although these data were derived from different cohorts, the risk of HB infection appears to rise between 10 and 15 years. The reasons for this phenomenon are not clear, but they may be related to the intimate mother–child interaction during the first 5 years of life and changes in lifestyle and sexual activity in older children.
Another concern is that batch-to-batch variation in immunogenicity might present in plasma-derived HB vaccines.29, 30 Positive anti-HBs rates in neonates using plasma-derived HB vaccine have been reported to be as low as 75%.31 Even if the children acquired “protective” antibodies after the vaccination, the antibody titers might be low. In the present study, 41% of children in group A had anti-HBs titers less than 100 mIU/mL. Without evidence of technical problems in vaccination or antibody determination, suboptimal batches of vaccine were more likely to be blamed for the high proportion of children with low anti-HBs titers at 18 months of age. Especially for children with low initial anti-HBs titers, the duration of protection against HB infection remained uncertain. Our results suggest that immune memory was well preserved in most of the children even though waning of anti-HBs was significant. However, for children who had a low response to HB vaccine initially, breakthrough infections might occur 10 to 15 years later.
Based on our previous and current observations, it seems reasonable to determine the anti-HBs status of children at 10 to 15 years after receiving plasma-derived HB vaccine and provide a booster to those whose immunity has waned. Similar considerations should be given to children who received recombinant vaccine. Alternatively, poor responders after the neonatal vaccinations should be identified, and additional doses of vaccine should be provided early on.
The authors are grateful to Dr. Calvin M. Kunin for his critical review of this manuscript. We thank the children and parents who volunteered to participate in the study and the residents and nurses in the Hepatitis Studying Group at the Department of Pediatrics, National Taiwan University Hospital. We also thank the Department of Health, Taiwan, for their administrative help.