Immune persistence after hepatitis B vaccination in infancy – Fact or fancy?

abstract

The hepatitis B vaccine has been introduced for more than 3 decades. In Hong Kong, excellent vaccine coverage through an efficient public health care system, together with supplemental programmes and easy availability of the vaccine, meant that most young pregnant women, and university students at entrance, should have been protected. Yet significant correlations in the prevalence of HBV infection with age were found in these groups of subjects, increasing from low to high endemicity rates from late teenage to the early twenties. This can only be attributed to vaccine failure, and there is cumulating evidence that several factors are involved, including the failure to respond to a primary series of hepatitis B vaccination in infancy, the waning of antibody titer with age, and loss of anamnestic response in a significant portion of the vaccinees. The duration of protection conferred by hepatitis B vaccination in infancy should be re-examined and remedial measures undertaken if its long term protection is found to be insufficient. Otherwise, the efforts to control HBV infection, especially in high endemicity regions, with universal vaccination in infancy would be rendered futile.

Keywords:

• adolescents
• age
• anamnestic response
• hepatitis B vaccination
• hepatitis B virus infection
• pregnant women
• vaccine non-responder
• waning immunity

Background

Hepatitis B virus (HBV) infection is a worldwide problem, and the high prevalence in endemic areas is attributed largely to perinatal transmission.¹ To control HBV infection, universal neonatal immunisation with the hepatitis B vaccine was introduced more than 3 decades ago, and its efficacy has been reported in many studies that have shown not only a reduction in the prevalence of childhood HBV infection, but also the presence of an anamnestic response that could maintain immunoprotection for many years.^2-14 Indeed, in Taiwan, following introduction of the vaccine, the incidence of hepatocellular carcinoma (HCC) in children was reduced from 0.54 to 0.20 per 100,000 children aged 6–14 years in those born before versus after the vaccination program.¹⁵

In Hong Kong, selective active-passive hepatitis B immunization was first administered to neonates born to mothers screened positive for hepatitis B surface antigen (HBsAg) from late 1983 to 1988,¹⁶ followed by universal neonatal HBV vaccination from November 1988,¹⁷ and the vaccine has become widely available since.¹⁶ The vaccination program was extended in July 1992 to cover all children born from January 1986 to November 1988 to ensure that all children aged 13 and below and residing in Hong Kong irrespective of place of birth would have been immunized by the end of 1990s.¹⁸ Finally, a Supplemental Hepatitis B Vaccination Program was launched in the 1998–99 school year targeting primary 6 students who had not yet received or completed the 3 dose regimen, so as to provide protection to all local children before they become sexually active.¹⁹ On this basis, the great majority of the young adults in Hong Kong should have been immune to HBV infection and the prevalence of HBV carriers in the population should have fallen progressively. Nevertheless, the notion that 3 decades of hepatitis B vaccination have reduced significantly the prevalence of HBV infection in young adults is no more than an assumption, as no population studies have been conducted to provide evidence of the long-term protection conferred by the HBV vaccine.

Prevalence of HBV infection in Hong Kong

In Hong Kong, pregnant women presenting for antenatal care are the only group in our society to have routine screening of HBsAg status. Pregnant women belong to a low-risk group, in contrast to the high risk group with traditional risk factors for acquiring HBV infection, such as intravenous drug abuse and homosexual practice. Thus pregnant women can be taken as surrogate of the low risk population at large. When the reported prevalence of maternal HBsAg carriage is reviewed, in the pre-vaccination era, it was 6.6% in 1976 ²⁰ and 7.4% in 1981–1983.³ Following the introduction of HBV vaccination, it was 10.0% overall and 8.4% in the locally-born residents in 1996,¹⁶ and 8% in 1998–2001²¹. In our hospital in 1998–2008, it was 10.1% overall, being 10.1% in 1998 and 10.4% in 2008.²² In a prospective survey in 2010, it was 9.1%, and 4.8% of women who had a history of hepatitis B vaccination were HBV carriers.²³ Since older mothers might not have received HBV vaccination before, and the constant influx of pregnant women born outside Hong Kong have most likely raised the prevalence of HBV infection ¹⁶ as most of them were unlikely to have been immunized in infancy, the overall prevalence of HBV infection in the obstetric population could have provided a misleading picture of the efficacy of the immunization program. However, when teenage mothers in Hong Kong were studied, the prevalence of HBV infection increased significantly from 1.2%, 1.5%, 7.1% to 8.3% for those at age ≤16, 17, 18, and 19 years respectively, and which was correlated with (P = 0.002) age.²⁴ Indeed, among young mothers born in Hong Kong who should have been protected by hepatitis B vaccination administered in infancy, HBV infection increased from 2.3% at age ≤16 years to peak at 8.4% at age 25 years.²⁵ In fact, by age 18 years, the 6.9% prevalence of HBV infection in the parturients born in/after 1984, when all high risk newborn infants would have been protected by HBV immunization, was comparable to the figure in 1976,²⁰ while the prevalence of 8.0% in those aged 22 years even exceeded the figure in 1981–1983.³ By age 23 years, the prevalence of 8.4% was the same as the 8.4% in the locally-born resident adult population in 1996.¹⁶

It may be argued that pregnant women are not really low risk after all, since conception would require unprotected sexual intercourse which also provides an opportunity of horizontal transmission of HBV infection from their spouses, as high prevalence of HBV infection in both male and female partners increases the risk of horizontal transmission.^26,27 Yet when first-year university students enrolled in Health Sciences Studies in our university were studied, the prevalence increased significantly from 0.9%, 2.3%, 4.3%, to 5.5% for students tested at age ≤18, 19, 20, and ≥21 years respectively.²⁸ Such finding, which was similar in nature to that of an Italian study conducted before mass hepatitis B vaccination that evidence of HBV infection increased from 1.7% among children aged 3–5 years to 4.5% among teenagers aged 17–19 years,²⁹ was totally unexpected. Furthermore, an age-related increase in HBV infection has also been shown in Pakistan, where HBV infection increased from 13.39% among subjects aged 11–20 years to 34.93% among subjects aged 21–30 years,³⁰ and in Korea where the weighted age-specific prevalence increased from 1.9% in adolescents to 4.9% in those in their 20s.³¹ Although few in number, these studies nevertheless suggested the increasing prevalence of HBV infection with age is likely to be related in part to vaccine failure in immunized subjects.

Is there evidence of HBV vaccine failure?

The concept of vaccine failure includes a number of categories or situations. For the occurrence of specific vaccine-preventable disease such as HBV infection in pregnant women born after the introduction of neonatal vaccination, this would be compatible with the category of confirmed clinical vaccine failure.³² Nevertheless, as the aforementioned studies conducted in the pregnant women in Hong Kong ^22,24,25 were retrospective in design, there was no means of confirmation or documentation of vaccination history. Yet there is evidence that the hepatitis B vaccine may not provide such long term protection as widely held. Long-term protection of a vaccine can be assessed in different ways, such as by the anamnestic response to a booster dose, infection rate in vaccinated populations, and seroepidemiological studies.³³ The evidence for and against the notion that hepatitis B vaccine can provide long-term, if not lifelong, protection, is presented below.

Evidence for the long term protection of HBV vaccine

Some published meta-analyses are in favor of a long-term protective effect. Poorolajal et al ³⁴ assessed 42 separate cohorts studies with 11090 subjects, including 34 cohorts and 9356 subjects in the final meta-analysis, found that overall cumulative incidence of HBV breakthrough infection 5–20 years was 0.007 (varying from 0 to 0.094), and came to the conclusion that protection by the 3 or 4 doses of monovalent HBV vaccine persists for at least 2 decades. But among these studies, only one, 2 and 2 cohorts were followed-up for 16, 18 and 20 years respectively, while the majority (26/40 studies) had follow-up of 10 years or less. FitzSimons et al ¹⁴ reviewed 9 studies in which the number of subjects varied from 46 to 1543, with 8 of the 9 studies having less than 500 subjects and included infants from both carrier and non-carriers mothers. The follow-up period was from 5 to the maximum of 24 years, including 4 with 20 years or more, and one to 18–23 years. However, HBsAg status not tested in 3 studies, was 0% in another 3 studies, was 1% and 2.9% respectively in 2 other studies and 3.8% in the last one. In these studies, offspring of both HBV carrier and non-carrier mothers were included, many were conducted in populations with low endemicity, and duration of follow-up was generally short. These reviews also revealed that while few studies followed up the vaccinees for more than 2 decades, there was already marked decline in anti-HBs with aging. Furthermore, vaccine failure (33.3% to 51.4%) was a major cause of HCC in Taiwan in children aged 6–14 years, and paradoxically a higher risk (risk ratio 2.3–4.5) of HCC was found in HBV carrier children born after the vaccination program.¹⁵ Therefore robust evidence of the long term efficacy of the HBV vaccine remains wanting.

Immune response to a primary series of HBV vaccine

Full vaccination in infancy cannot guarantee complete protection, as up to 31.9% of neonatal vaccinees responded weakly thus rendering them susceptible to infection after childhood.³⁵ Indeed, infants with non-protective anti-HBs levels can be found on postvaccination serological testing (PVST), the incidence being related to the interval of PVST. In one study, it increased from 14.6% to 36.1% when tested at 1–2.5 years' and 5–8 years' intervals respectively.³⁶ In the US, 5.3% of 8654 HBsAg-negative infants born to HBsAg positive mothers failed to respond to a primary series, the incidence increasing from 2% to 21.6% when PVST was done at 1–2 months and 15–16 months respectively.³⁷ A similar observation was reported from Hong Kong, where PVST at age 18–24 months in 160 children born to HBsAg positive mothers revealed that 22.5% had anti-HBs at <10 IU/mL, including one child (0.6%) who was HBsAg positive.³⁸. The remedy is a second series of a full course of vaccination,³⁷ but unless PVST is routinely performed, it is not possible to determine who would require a second series, and the assumption that all infants respond to a primary series would render an unknown proportion of the vaccinees who are non-responders susceptible to HBV infection in the future.

Waning of Anti-HBs titer with advancing age

A number of studies with follow-up of the vaccinees for up to 20 years or more revealed that anti-HBs level in responders can decline fairly rapidly with time. Anti-HBs seropositive rate could decline from 100% at age 2 years to 75% at as early as age 6 years among complete vaccinees.³⁹ In vaccinated children, as many as 22.9% had undetected antibody levels, which was not influenced by gestational age, birthweight or parental origin.³⁶ The decline in anti-HBs is correlated with age. In a study on 640 children born to HBsAg positive mothers and with completed vaccination, protective anti-HBs was 93% afterwards, then falling to 70%, 40% and 25% at 5, 10, and 15 years of age respectively.⁴⁰ Another study on 1095 subjects with early childhood vaccination and assessed at age 1–24 years found that while overall efficacy against infection and carriage were 83.4% and 96.5% respectively, population vaccine efficacy against infection decreased from 92.5% among 1–4 years old to 70.9% among 20–24 years old, and efficacy against carriage decreased from 100% to 91.1%.⁴¹ When tested at specific ages, a significant waning in anti-HBs titer can be found at age 15 to 18 years.^10,42 Two studies from Taiwan, where the national vaccination program began in 1984, are particularly revealing. When 1969 new university entrants in 2005 were analyzed by birth prior to July 1984, July 1984-June 1986, and from July 1986 respectively, the positive rates for HBsAg were 8.7%, 3.2%, 1.7%, that for anti-HBc were 26.5%, 8.0%, 4.7%, that for anti-HBs were 61.2%, 72.6%, 46.5%, and the negative rates for anti-HBc were 12.3%, 19.4%, and 48.8%, while non-protective level of anti-HBs increased from 11.9% for birth-year 1984 to 48.2% for birth-year 1987 students.⁴² A survey in 2000–2006 on first year nursing students found that while HBsAg positivity decreased from 4.9% in 2000 to 2.1% in 2006, anti-HBs also decreased from 77.1% to 39.7%, while both tests being negative increased from 18.0% to 58.2%.⁴³ In Hong Kong, 80.2% of 212 university entrants aged 17–23 years who had received neonatal vaccination were both HBsAg and anti-HBs negative, and HBsAg was found in only 0.9%.⁴⁴ The findings of these studies indicated that decreased HBsAg carriage can be associated simultaneously with decreased seroprotection, so that the former should not be taken as evidence of adequate seroprotection among the population. Indeed, a similar observation was made on 100 subjects born to HBsAg positive mothers during 1984–2004, in whom 2% had anti-HBc and there were no carriers, yet only 63.3% of uninfected had anti-HBs ≥10 mIU/mL, the percentage of protected individuals decreasing with time with significant reduction of anti-HBs GMT after 5 years to reach 10 mIU/mL in about 15 years' time.⁴⁵ The situation was even more striking among Micronesians vaccinated at birth in that only 7.3% of uninfected individuals had positive antibody.⁴⁶ Thus even among complete vaccinees, vaccine-induced immunity should not be taken for granted from the time of adolescence onwards.

Anamnestic response in vaccinees

It has been reported that even though protective antibody concentration was found in only 64% of the children more than 10 years after vaccination in infancy, an anamnestic response was demonstrated in 97% of the remaining children.⁴⁷ Nevertheless, disappearance of an anamnestic effect has been shown in many studies.^{10,40,41,45,48} One study found that following a booster dose in subjects who received 4 doses of vaccine during infancy, 28.7% had undetectable anti-HBs, while HBsAg-specific IFN-γ- or IL-5-secreting peripheral blood mononuclear cells remaining negative in 27.2%, and it was estimated that 10.1% of the total population had lost their vaccine-conferred booster response. ⁴⁹. In the 170 students negative for both HBsAg and anti-HBs in the aforementioned Hong Kong study,⁴⁴ 69 provided blood sample after the first booster dose and anti-HBs was <10 mIU/mL and 10–100 mIU/mL in 14.5% and 29.0% respectively, so that a 3-dose booster was required to induce 97.2% of these subjects to respond with an anti-HBs titer above 100 mIU/mL.⁴⁴ Indeed, by age 15 years, as many as half of the children vaccinated at birth had absent anamnestic responses to booster vaccination.^12,46 Therefore without the challenge of a routine booster, the presence of an adequate anamnestic response in most vaccinees remains an unsubstantiated opinion.

Infection rate in vaccinated populations

In the final analysis, the effectiveness or otherwise of the hepatitis B vaccine rests on whether it can prevent HBV infection. In completed vaccinees, anti-HBc seroconversion has been found in 1.5% from 3–15 years of age ⁴⁰ and in 8% 15 years later,⁴⁶ although none became chronic carriers. However, among complete vaccines in Taiwan, 1.3% were HBsAg positive at age 2–6 years, while the figure was 4.5% among incomplete vaccinees,³⁹ and 1.3% to 3.5% of the immunized infants became HBsAg carriers 15 years later.⁴⁸ A study in the rural villages of Gambia has shown that while vaccination had dramatically markedly reduced the HBV carriage rate, it nevertheless increased from 0% (0/472) at the age of 1–4 years to 2.1% (4/192) at the age of 20–24 years.⁴¹ Among Aboriginal adolescents, active and past infection with HBV was found in 11% and 19% respectively despite complete HBV vaccination in infancy.⁵⁰ But the most important evidence of inadequate vaccine protection was the finding of the increase from 2.3% at age ≤16 years to 8.4% at age 25 years of HBV infection among mothers supposedly protected by vaccination in infancy in Hong Kong.²⁵ Indeed, in the Hong Kong model, the current HBV infection rate among pregnant women appeared even worse than the situation 3 decades ago before the introduction of the immunisation program. The emerging data that increasing age is associated with increasing risk of HBV infection even among vaccinees indicates that only when vaccinees are confirmed to be protected against HBV infection during and beyond the third decade of life would there be genuine evidence of efficacy of hepatitis B vaccination in the long run.

Conclusion

There is scanty data on immune persistence in adulthood after hepatitis B vaccination in infancy. This can be attributed to the relatively short period of implementation of universal hepatitis B vaccination in many parts of the world, and the fact that large-scale seroepidemiological studies have not been conducted in adults born after the implementation of hepatitis B vaccination, especially in the low risk populations. Pregnant women probably represent the only low risk group where data from routine antenatal screening of HBV infection has been available for analysis, and the findings from Hong Kong are not reassuring. The concept of vaccine expiry, including the loss of the anamnestic response, following a full course of hepatitis B vaccination in infancy represents a paradigm shift. Further studies in different populations are warranted to determine if the same observation is replicable in other ethnic and racial groups. It is imperative to clarify the longevity of immunoprotection conferred by hepatitis B vaccination in infancy. Without such information, and recommendation for remedial actions, the efforts of universal hepatitis B vaccination in the control of HBV infection in endemic regions could be rendered futile, as suggested by the data generated from pregnant women in Hong Kong.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.