Recombinant hepatitis B vaccines provide an effective means to prevent infections caused by the hepatitis B virus (HBV), including the long-term complications of cirrhosis and liver cancer.1,2 Since the World Health Organization called for global use of HBV vaccines in 1992,2 at least 183 countries have implemented vaccination programs.2 Countries with high endemic rates of HBV infection that were early adopters of routine vaccination have reported marked rate reductions in acute and chronic infection.3–7 Follow-up studies extending several decades have confirmed that immunized individuals enjoy prolonged protection.5,8–10 Breakthrough infections occur occasionally after child and adult vaccination but more frequently after infant vaccination, usually without establishment of chronic infection.5,8,11–13
HBV vaccination programs routinely target infants or adolescents or both. The goal of infant programs is to reduce the high rates of chronic infection commonly associated with infection in early life,2,14 whereas the goal of adolescent programs is to reduce peak rates of acute infection seen in early-to-mid adulthood, when progression to chronic infection is infrequent.2 Available evidence confirms the success of both types of programs.10,14–16 However, an important remaining question is whether infant vaccination can accomplish both goals, including prevention of significant disease well into adulthood. A challenge to evaluating program outcomes is the diversity of schedules recommended for infant HBV immunization. World Health Organization (WHO) guidelines recommend a birth dose followed by 2–3 doses coadministered with other early childhood vaccines according to the local schedule.2 In the United States, Centers for Disease Control and Prevention guidelines also recommend a birth dose for all, followed by doses at 1–2 and 6–18 months or at 2, 4 and 6 months when using a combination vaccine containing HBV antigen.17 Greater diversity is evident in the European Union, where many countries limit the birth dose to infants born to infected women, with 3–4 subsequent doses administered in a variety of schedules (0, 1, 6 months; 2, 3, 4 months; 2, 4, 6 months etc). Some countries substantially delay the final dose after 2–3 initial doses, with recommended ages varying between 10 and 15 months of age.18
In Canada, British Columbia (BC) has provided universal infant HBV vaccination since 2001, after a pilot program in Vancouver in 1998–2000. As BC already had an established program to detect infected pregnant women and administer HBV vaccine to their newborns (0-, 1- and 6-month schedule), a decision was made to provide the vaccine to low-risk infants at 2, 4, 6 months of age, conveniently coadministered with routine infant vaccinations. BC also provided HBV vaccine routinely to adolescents in a school-based program, established in 1992. Graduates of the provincial infant program reached the age at which adolescent immunizations had been given (10–11 years) in 2012, leading to cessation of the adolescent program and reliance on infant vaccination for extended protection. We undertook to study whether the assumption of durable protection was well founded when HBV vaccinations are provided at 2, 4 and 6 months of age. In a meta-analysis of 46 studies15 examining the persistence of anti-HBs after infant vaccination, all but one study provided doses at about 0, 1 and 6 months of age, consistent with the WHO/Centers for Disease Control and Prevention recommendations. Detailed analysis of variation in the interval between doses 2 and 3 confirmed that wider separation (≥4 months) provided superior long-term immunogenicity,15 presumably because the delayed third dose boosted final responses. The compressed 2-, 4- and 6-month schedule might prime infants less effectively, with a reduced series response potentially reducing the duration of protection. This possibility was important to address as compressed schedules have become increasingly widely used to take advantage of combination infant vaccines that include HBV vaccine. An additional factor that might reduce the duration of protection is the absence of natural boosting in low HBV endemicity countries like Canada and the United States.19 This is particularly relevant because natural boosting might have affected the rates of antibody persistence reported from countries with higher HBV endemicity that predominated in the above-mentioned meta-analysis.15
The immune response to HBV vaccination includes production of antibody against HBV surface antigen (anti-HBs),2,11,20 as well as T and B cell responses.12,21,22 A vaccinated individual is considered to have protection if postvaccination serum antibody values are ≥10 mIU/mL.2 This correlates best with protection against chronic infection but does not predict 100% protection against acute or subclinical infection.11,14 Anti-HBs concentrations decline following vaccination, possibly falling below the protective or detection threshold, but immune memory often persists, enabling rapid recall of a protective antibody response following natural exposure or booster vaccination.20,23 Given the lengthy incubation period of HBV infection,2,24 an anamnestic response is likely to provide effective protection against chronic infection. However, antibody persistence and the anamnestic response are variable among individuals and both are proportional to the antibody response obtained after primary vaccination.13,15,23
The stimulus to conduct this study was the lack of data on long-term persistence of antibody following the compressed 2-, 4- and 6-month infant HBV immunization schedule, especially from a low prevalence region without the potential for natural boosting.
MATERIALS AND METHODS
The study was approved by the University of British Columbia Research Ethics Board. A list of randomly chosen infant vaccinees from 2 age groups (10- to 11- and 15- to 16-year olds) was provided by the Vancouver Coastal Health Authority from an immunization registry database, with permission. Only those with 3 HBV immunization dates conforming to the recommended infant schedule were included. Another list of potential participants was generated using a database of our previous study participants who had consented to future contact. In this instance, HBV immunization schedule compliance was verified from parent or provider records. The older age group was the first to receive infant HBV vaccine in Vancouver and the younger age group mimicked the age at which preadolescents were previously offered HBV vaccine, as a benchmark for comparisons. Families were sent an invitation letter describing the study, then telephoned to offer further information. Preliminary eligibility screening was conducted by telephone.
Participants were from the Vancouver area and had to have been born in BC, participated in the routine infant vaccination program at about 2, 4 and 6 months of age and have a record of those immunizations. None had serology testing following series completion. Exclusion criteria included having been immunized on an alternative HBV infant schedule or with additional doses of HBV vaccine; having evidence of prior HBV infection; suffering from conditions associated with immunosuppression during childhood and receipt of blood products within 3 months. For this study, compliance with the routine schedule required that the first dose of the primary HBV series should have been received between 6 and 12 weeks of age, the third dose before 36 weeks of age and the interval between second and third doses was not to exceed 13 weeks. Recombinant HBV vaccine (Recombivax, Merck Frosst Canada Kirkland, Quebec) was routinely administered to both age groups as infants. A proportion of the older participants likely received adult doses of vaccine, although this was not known at study onset (see also “Discussion”).
At participants’ first visit, eligibility was reviewed and informed consent and/or assent was obtained. A baseline blood sample was obtained to determine anti-HBs concentration and to test for HBsAg and anti-HBc to rule out prior infection (eligibility condition). Hepatitis B tests were performed at the British Columbia Centre for Disease Control, the reference laboratory for the province. The assay used was an US Food and Drug Administration-approved chemiluminescence immunoassay-based commercial assay (ADVIA Centaur anti-HBs2 assay, Siemens Healthcare Diagnostics inc, Tarrytown, NY), which expressed results in milli-International Units per milliliter (mIU/mL), with a threshold of ≥10 mIU/mL for protection and a lower limit of detection of 3.1 mIU/mL. Because the assay has a margin of error of 10 ± 2 mIU/mL, we followed the manufacturer’s recommendation to use values ≥12 mIU/mL as more certain to indicate protection without repeated testing of samples.
Participants with anti-HBs titers ≥12 mIU/mL were advised that they were protected and exited the study. Those with titers <12 mIU/mL were invited to return for challenge vaccination with a standard pediatric dose of HBV vaccine (10 mcg Engerix B, GSK, as currently used in the BC program). All doses of vaccine were from the same lot. Participants were monitored for any immediate allergic reaction for 15 minutes after vaccination and encouraged to report any later adverse events of concern.
Continuing participants returned 28 days (±7 days) after challenge vaccination when eligibility and adverse events were reviewed and blood for anti-HBs measurement was obtained. Participants found to have anti-HBs ≥12 mIU/mL exited the study; others were asked to return for a second challenge dose of HBV vaccine, administered as before. These few participants returned 28 days (±7 days) later when blood was obtained to determine their anti-HBs status. Participants with persistently undetectable anti-HBs were possibly primary nonresponders, who might also have failed to respond to their primary series. These individuals were informed of their results and advised to seek additional doses of HBV vaccine. The best evidence of loss of immune memory was provided by participants who lacked anti-HBs before and after initial challenge but who developed ≥12 mIU/mL after the second challenge, ruling out inability to respond.
Statistical Plan and Power Calculations
Based on a meta-analysis of similar follow-up studies,15 we estimated that about 30% of younger and 60% of older participants would lack seroprotection (anti-HBs <12 mIU/mL). We estimated that 25%–30% of older nonprotected participants would not have a recall response to challenge, representing 15%–20% overall of that study population. To detect a rate of 15% with a precision of ±0.05 and 95% confidence intervals, we determined that 220 individuals would need to be tested, allowing for a 10% drop-out rate. Among the younger challenge dose recipients, we projected that 10% would not have a recall response, representing an overall susceptibility rate of 3%. For a precision of ±0.03 and 95% confidence intervals, we determined that 140 participants would need to be assessed, allowing for a 10% drop-out rate.
Statistical analysis was performed using SAS version 9.4 (SAS Institute, Cary, NC), where the PROC FREQ and PROC MEANS were used to compute antibody threshold rates, geometric mean antibody concentration (GMC) and corresponding 95% confidence intervals. P values were computed for proportional data using Fisher’s exact test whereas GMCs were compared by t test. All available data were included in the analyses in the event of participant withdrawal after testing.
Study Population and Participation
Of over 3200 individuals invited by mail to participate, 710 proved not to have a current address or telephone number or could not speak English. Of 1276 individuals who were interviewed by telephone, 917 declined to participate. When the target of 359 volunteers was reached, recruitment efforts ceased, with 1214 mailed invitations not pursued. Subsequent participant flow is outlined in Figure 1. Six enrollees were deemed ineligible after being enrolled as a result of late disclosure of neonatal HBV vaccination (n = 3), ineligible medical condition (n = 2) or additional doses of HBV vaccine (n = 1). After these 6 exclusions and 3 early consent withdrawals, 350 participants (194 males and 156 females) underwent baseline testing. Thereafter, 3 participants withdrew consent (citing needle aversion) and 1 was lost to follow-up, for a net protocol completion rate of 96% (346/359).
Table 1 summarizes the demographic characteristics of study participants. Males predominated in the older group (59:41 ratio; P = 0.08, compared with the younger group). Ethnicity distribution was similar in both age groups, reflecting local population diversity. Health conditions reported by participants were predominantly minor or intermittent ailments, including asthma, skin conditions and anxiety disorders. Mean ages at starting (9 weeks) and completing (6.3 months) infant HBV immunization were uniform between age groups and matched the recommended schedule.
Baseline test results were available for all 350 participants (Fig. 1). None had detectable anti-HBc or HBs antigen present. Anti-HBs concentrations <12 mIU/mL were present in 107/137 (78.1%) younger participants and 137/213 (64.3%) older participants, a significant rate difference (P = 0.006; Table 2). Protection rates did not vary significantly by sex. Among participants with antibody concentrations below 12 mIU/mL, 51.1% of younger and 36.2% of older participants had undetectable values (<3.1 mIU/mL). The geometric mean anti-HBs concentrations in both age groups were low at baseline but somewhat higher in the older group (Table 2). No significant difference in baseline GMC was noted between males and females in either age group (data not shown).
First Challenge Responses
Of participants with <12 mIU/mL anti-HBs at baseline, 106 of the younger and 135 of the older participants returned for HBV vaccine challenge (Fig. 1). No adverse events were documented postvaccination. All but 1 younger participant returned for the follow-up blood test. Postchallenge, anti-HBs values ≥12 mIU/mL were present in 97.1% of younger participants and 91.1% of older participants (P = 0.06; Table 2). Most (91.4% in the younger group and 79.7% in the older group; P = 0.01) had robust responses to challenge, with anti-HBs ≥100 mIU/mL. A significantly greater proportion of younger than older participants developed postchallenge antibody responses ≥1000 mIU/mL (P = 0.02). Furthermore, among participants with undetectable anti-HBs at baseline, most of the younger group responded strongly to challenge dosing compared with a lower proportion of the older group (Fig. 2). Postchallenge GMC in the younger group was significantly greater than in the older group (Table 2). No significant difference in postchallenge GMC was noted between males and females in either age group.
Second Challenge Responses
Participants with anti-HBs <12 mIU/mL after the first challenge dose (n = 15) all returned for a second challenge. No adverse events were documented postvaccination. All 3 rechallenged participants in the younger group (100%) and 9/12 in the older group (75%) developed protective anti-HBs concentrations afterward (Table 2). The response in these participants was less robust than in those who responded to the first challenge, with only 1 of 3 younger and 2 of 9 (22%) older participants developing anti-HBs ≥100 mIU/mL. The GMC of older group responders was just 6% of the GMC of their peers who responded to a first challenge (20 vs. 319 mIU/mL; Table 2). Taken together, this response pattern is consistent with loss of immune memory after infant vaccination and an early primary response to vaccine challenge. Three older group participants (1.4% overall) lacked detectable antibody after both challenges and likely represented primary nonresponders.
This study is one of the first and the largest long-term evaluation of immunity after infant HBV vaccination at 2, 4 and 6 months of age. While this schedule is convenient and favors use of HBV-containing combination vaccines, we found antibody concentrations 10 years later were substantially lower than reported after the 0-, 1- and 6-month schedule. Compared with results of a meta-analysis15 of 45 studies conducted 10 years after HBV vaccination at 0, 1 and 6 months of age when only 30% of children had anti-HBs concentrations <10 mIU/mL, 78% of 10- to 11-year olds that we studied had values <12 mIU/mL, suggesting less durable antibody responses after the 2-, 4- and 6-month schedule. However, neither study included response measurements from infancy to address rates of antibody decay. Lack of natural boosting may have contributed to the low observed anti-HBs concentrations in our study, as in a study from the United States.19 However, despite the relative lack of serum antibodies, none of the younger participants in the current study had evidence of prior HBV infection and nearly all had a recall response to challenge dosing, even those who had no detectable antibody at baseline. The observations indicate that an anamnestic response can be elicited in this group as a result of persistent immune memory, consistent with ongoing protection from infection.
Among the 15- to 16-year olds studied, we estimated that 60% would have anti-HBs concentrations <12 mIU/mL at baseline based on the above meta-analysis15 and found that to be the case in 64%. However, this result was puzzling as the younger cohort had a higher proportion of individuals with <12 mIU/mL anti-HBs after a shorter interval from prior vaccination. Unknown to the investigators at the outset of this study, many of the older participants likely received adult doses of HBV vaccine (Recombivax, Merck) for their primary series, for lack of an available pediatric formulation at the time (pediatric dosage was recommended but providers were urged not to waste half-doses in opened “adult” vials; actual dosages administered were not recorded so the proportion receiving adult vs. pediatric dosages is unknown). By the time the younger cohort was born, pediatric dose formats were available. Prior studies of infant HBV vaccination indicate that persistence of anti-HBs is proportional to the primary series dosage received,8,10–15 which likely explains the higher proportion of individuals with ≥12 mIU/mL anti-HBs in the older than younger cohort.
Despite the frequent absence of HBV antibody concentrations ≥12 mIU/mL in both age groups, the great majority of participants had an anamnestic response to challenge dosing reflecting ongoing immune memory and likely protection. Fewer older than younger participants responded to challenge dosing, consistent with the literature.15 Despite the infant dosage advantage in the older group, a number of observations point to declining response capacity with increasing age. The 15- to 16-year olds had a less robust response to challenge dosing when compared with the younger group, developing a GMC about half that of younger participants. Similarly, older participants were less likely to respond with anti-HBs concentrations ≥100 mIU/mL, compared with the younger group (P = 0.01), particularly those without detectable anti-HBs beforehand (Fig. 2). Responses ≥100 mIU/mL were used by several authors to characterize an anamnestic response in adolescents and adults.23,25–27
We found that 2.2% of younger and 5.6% of older participants (P = 0.06) had <12 mIU/mL anti-HBs at baseline and were nonresponsive to initial vaccine challenge, suggesting lack of immune memory and protection. Most of these individuals responded to a second challenge, demonstrating a capacity to recognize HBs antigen but the observed antibody responses were weak suggesting new primary responses, as can develop in about one-third of naïve vaccinees after a single dose.28 Persistent nonresponsiveness to HBV vaccine was absent among younger participants and present in only 3 older participants (1.4%) consistent with either primary nonresponsiveness29 or a low responder phenotype.30 In the meta-analysis of follow-up studies after infant HBV immunization,15 the modeled rate of persistent protection (including demonstrated immune memory) after 10 years was approximately 95% with adequate dosing and 0-, 1-and 6-month schedule compliance and approximately 90% after 15 years, with similar conditions. We observed similar protection rates, speaking against a notable difference in duration of protection with HBV dosing at 2, 4 and 6 months of age.
This study had several design strengths. Recruitment was community based and participants were generally reflective of the local population. Compliance of participants with the study protocol was excellent, with only 3.6% attrition. Adherence to the infant schedule was verified from provider records, with limited deviation allowed. A single lot of HBV vaccine was used for challenges and laboratory testing was standardized in a reference laboratory. Using a 28-day sampling interval following challenge vaccination maximized the potential to detect immune responsiveness.23
The principal weakness of the study was the limited ability of the challenge vaccination procedure to accurately identify immune memory and anamnestic responses. Other mechanisms might also contribute to protection, including cellular and herd immunity. Strong anti-HBs responses were readily characterized as anamnestic but the methodology used did not allow discrimination between weak anamnestic responses and weak primary responses in those who had lost immune memory. Measurement of anti-HBs responses sooner after challenge would have aided recognition of anamnestic responses, whereas measurement of immunoglobulin M antibody responses to HBs antigen might have identified primary responses but such an assay was not available. Consequently, given that 20% of the older participants had limited responses to initial challenge (<100 mIU/mL), we might have substantially underestimated the proportion who had lost immune memory as all postchallenge responses ≥12 mIU/mL were interpreted as anamnestic, although about one-third of naïve vaccinees can develop anti-HBs ≥10 mIU/mL after a single dose.28 In reality, both anamnestic and rapid primary responses might contribute to protection against chronic HBV infection, making the distinction moot. Breakthrough infections were not encountered as evidence of loss of protection but were not expected in an environment of low endemicity and high population immunity from long-standing immunization programs. Other weaknesses included the vaccine dosage difference between the age groups in infancy (described above) with undefined variability in dosing of the older group, which precluded modeling of age-related changes in immunity. The vaccine used for primary vaccinations was Recombivax, whereas the challenge vaccine provided for our study was Engerix B. Challenge responses might have been affected by this mismatch but mixed vaccine primary immunization schedules had no effect on final responses in other studies.31,32 Documentation of individual responses to primary HBV immunization in infancy was not available. Departing from the usual seroprotection threshold (12 vs. 10 mIU/mL) as required by the available immunoassay did not materially affect study results as few results (eg, 3.7% of baseline samples) fell within this range.
This study indicates that British Columbia’s HBV vaccination program for low-risk infants at 2, 4 and 6 months of age provides satisfactory protection to midadolescence. The question of whether infant HBV vaccination can protect into adulthood lacks a definite answer but mounting evidence points to differences in duration of protection after immunization in early infancy versus later ages. Substantial evidence already exists that individuals immunized as adolescents or young adults enjoy long-term protection, documented 15 to 30 years following vaccination,10,33 well into the period of heightened risk for adults. No booster vaccination is recommended in this context.2,16 In contrast, multiple follow-up studies15 after infant immunization show a progressive loss of detectable immunity (seroprotection and immune memory) in up to 15%–20% of individuals by 15 years after vaccination, with potential development of breakthrough infections13 depending on exposure risks. Follow-up studies after infant vaccination in The Gambia13 demonstrated increasing rates of breakthrough infection (identified by presence of anti-HBc) among young adults, reaching 30% among 25- to 29-year olds although few had chronic infection. While the Gambian experience is exceptional, reflecting high HBV exposure risks and possible population differences, it demonstrates the potential for loss of protection after infant immunization.
Comparison of long-term outcomes after infant and adolescent HBV vaccination highlights the differences between them. For example, preadolescents who were given HBV vaccination in a prospective study33 in Quebec, Canada, and tested 10 years later (age 19) had a residual seroprotection rate (anti-HBs ≥10 mIU/mL) of 86% (61% with ≥1000 mIU/mL) and a GMC of 169 mIU/mL. For the 10- to 11-year olds in our study vaccinated as infants, the residual seroprotection rate was 4-fold lower at 21.9% (none with >1000 mIU/mL) and the GMC was 40-fold lower at 4.2 mIU/mL. Assuming that the HBV booster response reflects residual B cell response capacity, it is noteworthy that the postbooster GMC of 19-year olds in Quebec was 32,477 mIU/mL, whereas that of 15- to 16-year olds in BC was 100-fold lower, at 319 mIU/mL. All teens in the Quebec study had received a booster but only nonseroprotected teens (64% of the total) in the present study were boosted, so the results are not directly comparable.
Given such differences, one can reasonably argue that sufficient uncertainty exists about long-term protection after infant vaccination to warrant reconsideration of the infant priming regimen, favoring schedules with more doses or antigen, use of adjuvants or more delayed final doses, as with the current schedules in Belgium and Germany (doses at 2, 3, 4 and 11–15 months, in addition to a birth dose for high risk infants).18 While WHO does not currently recommend a booster dose for any age group,2,16 booster vaccination in adolescence might eventually prove useful as a means to reinforce and extend protection in those immunized as infants. Whether a booster dose could extend protection of adults immunized as infants to better resemble that of adults immunized as adolescents remains an important unanswered question. The fact that substantial differences exist in measures of residual protection among teenagers after infant or adolescent HBV vaccination warrants close ongoing scrutiny of whether important differences will emerge in long-term protection, with or without booster vaccination.
The authors sincerely thank the following individuals and organizations who helped make this study possible: Ivor Chee and Carol LaJeunesse (Study Coordinators) and the research nurses and staff of the Vaccine Evaluation Center; Vancouver Coastal Health Authority; and British Columbia Centre for Disease Control’s Hepatitis Laboratory. The authors greatly appreciated the excellent cooperation of the families and individuals who participated in the study.
1. McMahon BJ. Natural history of chronic hepatitis B. Clin Liver Dis. 2010;14:381–396.
2. World Health Organization. Hepatitis B vaccines: WHO position paper. Wkly Epidemiol Rec. 2009;84:405–19.
3. Lin HH, Wang LY, Hu CT, et al. Decline of hepatitis B carrier rate in vaccinated and unvaccinated subjects: sixteen years after newborn vaccination program in Taiwan. J Med Virol. 2003;69:471–474.
4. Whittle H, Jaffar S, Wansbrough M, et al. Observational study of vaccine efficacy 14 years after trial of hepatitis B vaccination in Gambian children. BMJ. 2002;325:569.
5. Tosti ME, Alfonsi V, Lacorte E, et al; SEIEVA Collaborating Group. Acute hepatitis B after the implementation of universal vaccination in Italy: results from 22 years of surveillance (1993-2014). Clin Infect Dis. 2016;62:1412–1418.
6. Wiesen E, Diorditsa S, Li X. Progress towards hepatitis B prevention through vaccination in the Western Pacific, 1990-2014. Vaccine. 2016;34:2855–2862.
7. Harpaz R, McMahon BJ, Margolis HS, et al. Elimination of new chronic hepatitis B virus infections: results of the Alaska immunization program. J Infect Dis. 2000;181:413–418.
8. Poorolajal J, Mahmoodi M, Majdzadeh R, et al. Seroprotection of hepatitis B vaccine and need for booster dose: a meta-analysis. Vaccine. 2010;28:623–631.
9. Poorolajal J, Mahmoodi M, Haghdoost A, et al. Booster dose vaccination for preventing hepatitis B. Cochrane database Syst Rev. 2010:CD008256.
10. Bruce MG, Bruden D, Hurlburt D, et al. Antibody levels and protection after hepatitis B vaccine: results of a 30-year follow-up study and response to a booster dose. J Infect Dis. 2016;214:16–22.
11. Jack AD, Hall AJ, Maine N, et al. What level of hepatitis B antibody is protective? J Infect Dis. 1999;179:489–492.
12. Lu CY, Ni YH, Chiang BL, et al. Humoral and cellular immune responses to a hepatitis B vaccine booster 15-18 years after neonatal immunization. J Infect Dis. 2008;197:1419–1426.
13. Mendy M, Peterson I, Hossin S, et al. Observational study of vaccine efficacy 24 years after the start of hepatitis B vaccination in two Gambian villages: no need for a booster dose. PLoS One. 2013;8:e58029.
14. Peto TJ, Mendy ME, Lowe Y, et al. Efficacy and effectiveness of infant vaccination against chronic hepatitis B in the Gambia Hepatitis Intervention Study (1986-90) and in the nationwide immunisation program. BMC Infect Dis. 2014;14:7.
15. Schönberger K, Riedel C, Rückinger S, et al. Determinants of Long-term protection after hepatitis B vaccination in infancy: a meta-analysis. Pediatr Infect Dis J. 2013;32:307–313.
16. Van Damme P. Long-term protection after hepatitis B vaccine. J Infect Dis. 2016;214:1–3.
17. Centers for Disease Control. A Comprehensive Immunization Strategy to Eliminate Transmission of Hepatitis B Virus Infection in the United States. Recommendations of the Advisory Committee on Immunization Practices (ACIP); Part!: Immunization of Infants, Children and Adolescents. MMWR. 2005; 54(RR-16).
19. Boxall EH, A Sira J, El-Shuhkri N, et al. Long-term persistence of immunity to hepatitis B after vaccination during infancy in a country where endemicity is low. J Infect Dis. 2004;190:1264–1269.
20. West DJ, Calandra GB. Vaccine induced immunologic memory for hepatitis B surface antigen: implications for policy on booster vaccination. Vaccine. 1996;14:1019–1027.
21. Carollo M, Palazzo R, Bianco M, et al. Hepatitis B specific T cell immunity induced by primary vaccination persists independently of the protective serum antibody level. Vaccine. 2013;31:506–513.
22. Simons BC, Spradling PR, Bruden DJ, et al. A longitudinal hepatitis B vaccine cohort demonstrates long-lasting hepatitis B virus (HBV) cellular immunity despite loss of antibody against HBV surface antigen. J Infect Dis. 2016;214:273–280.
23. Brunskole Hummel I, Zitzmann A, Erl M, et al. Characteristics of immune memory 10-15 years after primary hepatitis B vaccination. Vaccine. 2016;34:636–642.
24. Lai CL, Ratziu V, Yuen MF, et al. Viral hepatitis B. Lancet. 2003;362:2089–2094.
25. Anderson CL, Remschmidt C, Drobnitzky FP, et al. Hepatitis B immune status in adolescents vaccinated during infancy: a retrospective cohort study from a pediatric practice in Germany. Hum Vaccin Immunother. 2016;12:779–784.
26. Van Der Meeren O, Behre U, Crasta P. Immunity to hepatitis B persists in adolescents 15-16 years of age vaccinated in infancy with three doses of hepatitis B vaccine. Vaccine. 2016;34:2745–2749.
27. Avdicova M, Crasta PD, Hardt K, et al. Lasting immune memory against hepatitis B following challenge 10-11 years after primary vaccination with either three doses of hexavalent DTPa-HBV-IPV/Hib or monovalent hepatitis B vaccine at 3, 5 and 11-12 months of age. Vaccine. 2015;33:2727–2733.
28. Szmuness W, Stevens CE, Harley EJ, et al. Hepatitis B vaccine: demonstration of efficacy in a controlled clinical trial in a high-risk population in the United States. N Engl J Med. 1980;303:833–841.
29. Ko SC, Schillie SF, Walker T, et al. Hepatitis B vaccine response among infants born to hepatitis B surface antigen-positive women. Vaccine. 2014;32:2127–2133.
30. Pichichero ME, Passador S. Administration of combined diphtheria and tetanus toxoids and pertussis vaccine, hepatitis B vaccine, and Haemophilus influenzae type b (Hib) vaccine to infants and response to a booster dose of Hib conjugate vaccine. Clin Infect Dis. 1997;25:1378–1384.
31. Jilg W, Schmidt M, Deinhardt F. Vaccination against hepatitis B: comparison of three different vaccination schedules. J Infect Dis. 1989;160:766–769.
32. Seto D, West DJ, Gilliam RR, et al. Antibody responses of healthy neonates of two mixed regimens of hepatitis B vaccine. Pediatr Infect Dis J. 1999;18:840–842.
33. Gilca V, De Serres G, Boulianne N, et al. Antibody persistence and the effect of a booster dose given 5, 10 or 15 years after vaccinating preadolescents with a recombinant hepatitis B vaccine. Vaccine. 2013;31:448–451.