Secondary Logo

Journal Logo

Original Studies

Comparative Long Term Immunogenicity of Two Recombinant Hepatitis B Vaccines and the Effect of a Booster Dose Given After Five Years in a Low Endemicity Country

Duval, Bernard MD*; Gîlca, Vladimir PhD; Boulianne, Nicole MSc*; De Wals, Philippe PhD§; Massé, Richard MD*; Trudeau, Gisele MD; De Serres, Gaston PhD§

Author Information
The Pediatric Infectious Disease Journal: March 2005 - Volume 24 - Issue 3 - p 213-218
doi: 10.1097/01.inf.0000154329.00361.39
  • Free


Licensure of safe and effective hepatitis B (HB) vaccines and universal immunization programs in many countries have provided health benefits that are beginning to be seen.1–3 The question remains: how long will HB vaccine immunity last? In many countries, the vaccine is administered to infants. In some low endemicity countries, the vaccine is offered to preteenagers. In these countries, the highest rates of infection are observed between the ages of 20 and 40 years. This question about the duration of protection is important for public health authorities who must plan the immunization programs and eventual booster doses.

It is technically difficult to assess the persistence of protection. Antibody titers decrease with time in all vaccinees and eventually disappear in a large proportion of them.4–7 The absence of antibodies does not mean the absence of protection because of remaining immune memory,1 so seroprevalence surveys underestimate the proportion of people who are protected. Alternatively it is possible that the immune memory eventually disappears with time. There is no standardized and validated laboratory test for reliable measurement of cellular immunity against HB. It is difficult to detect secondary vaccine failures (cases in previously vaccinated and protected people) because of the relatively low incidence rates of HB, the often asymptomatic course of the disease, the lack of hepatitis B surveillance in many countries and ignorance of serology status after primary immunization. It has been shown that people protected by the vaccine can become infected by the virus and develop anti-hepatitis core antibody (anti-HBc) although they rarely remain carriers or show symptoms.8–11

To our knowledge, no prospective study has monitored over a period of time the titers obtained after a booster dose in teenagers previously vaccinated with pediatric doses. This information might be useful in deciding whether booster doses are needed for long term protection.

To assess the duration of protection and the need to introduce a booster dose in the immunization program, we designed a 15-year study to follow 2 large groups of preteenagers who had been immunized with 1 of 2 recombinant vaccines. We report the results after 5 and 6 years.


Study Design.

The province of Quebec (Canada) introduced a school-based HB virus universal immunization in grade 4 (8- to 10-year-old children) in 1994. In 1995–1996 and 1996–1997, 2 cohorts of children received Engerix-B 10 μg (EB) and Recombivax-HB 2.5 μg (RB) vaccines, respectively. Antibodies were measured one month after the third dose. These results were presented in a previous publication.12 Each cohort was randomized in 3 groups to be tested and boosted at year 5, 10 or 15 after the primary immunization (Fig. 1). The persistence of immune memory is assessed by measurement of the antibodies before, one month and one year after administration of a booster.

Design and participation in the study.

Study Population.

At the beginning, 1129 8-10-year-old children received EB and 1126 received RB following a 0-, 1-, 6-month calendar. Subjects were actively followed with an annual mail-based contact. During the 4 years after primary vaccination, 151 subjects (13.4%) in the EB group and 142 (12.6%) in the RB group were lost to follow-up or excluded. Most of them refused to continue to participate in the study (90.5%), a few moved out of the province (3.4%) and some were excluded because they had not reached a protective titer of 10 mIU/mL and were offered supplementary doses or violated the protocol requirements during the initial phase of the study (6.1%). At the beginning of year 5, the remaining subjects, 978 in the EB and 984 in the RB group, were randomized. There were 326 and 328 subjects in each of the 3 EB and RB groups respectively (Fig. 1). EB and RB group A subjects were asked to participate in the year 5 and 6 procedures. All these subjects had an anti-HB surface antibody (HBs) response of >10 mIU/mL 1 month after primary vaccination.

Booster Vaccine Administration.

Subjects were boosted with the vaccine from the same manufacturer as for the primary vaccination course administered, intramuscularly in the deltoid by the study nurses. We used the dosage recommended for this age group in Canada: EB 0.5 mL [10 μg of hepatitis B surface antigen (HBsAg)]; RB 0.5 mL (5 μg of HBsAg). The same lots of vaccines were used in all vaccinees.

Data Collection.

All procedures were done at the participants’ residence by a trained nurse. Blood samples and relevant medical, demographic and anthropometric data were collected immediately before booster dose administration. A second blood sample was obtained 28–60 days later with a similar distribution of intervals in the 2 groups. Adverse events occurring within 1 month after administration of the vaccine were recorded, irrespective of severity or whether or not they were considered vaccine-related. A third blood sample was obtained 1 year after the booster dose.

Serologic Procedures.

Anti-HBs titers were assessed by the AUSAB EIA/Abbott Laboratories Diagnostic Division Test, anti-HBc by the ORTHO HBc Elisa Test System and HBsAg by the HBsAg EIA 2.0/Genetic System. Anti-HBs titers were expressed in international units (IU) according to the manufacturer's instructions, with the World Health Organization standard reference serum provided in each kit. The same internal control serum was also tested at the beginning and end of each batch to ensure reproducibility of the assay between batches. All tests were done at the Laboratoire de Santé Publique du Quebec.


All the subjects with pre- and postbooster serologic results were included in the analysis of immunogenicity. Statistical analyses were performed with Statistical Analysis System SAS 8.02 on Windows 2000 5.00.2195. We estimated the proportion of participants with detectable (≥1 mIU/mL) and seroprotective anti-HBs (≥10 mIU/mL) and the geometric mean titers (GMTs) in the samples collected before and after administration of the booster dose with 95% confidence intervals using the exact binomial proportion method. GMTs were compared with the F statistics from the analysis of variance. Multivariate analysis for the examination of potential confounding variables was performed using the Kruskall-Wallis nonparametric rank test. Polynomial regression was fitted to verify the potential exponential association between the anti-HBs titers at different moments of time. Where appropriate, independent variables were categorized in quartiles to provide a sufficient number of subjects in each category.


The study protocol was initially approved by the ethics committee of the Centre Hospitalier Universitaire de Québec and reapproved for the year 5 and 6 procedures. Informed consent was signed by the parents at the beginning of the study and by the subjects themselves before the booster dose.


Characteristics of Subjects.

From the 326 EB- and 328 RB-eligible subjects, pre- and postbooster serologic results were available for 301 (95.1%) EB subjects and 272 (82.9%) RB subjects. In both groups, 53% were girls, and 41 and 59% were, respectively, 14 and 15 years old. The body mass index (kilograms/m2) was also identical in both groups (EB 21.3 ± 3.9, RB 21.6 ± 3.9). Overall 98% of the teenagers in each group (EB 97.7%, RB 98.5%) followed the protocol schedule with the interval between booster dose administration and the blood sample set between 28 and 60 days. The data of 11 (1.9%) participants who had their blood drawn 61–86 days after the booster were included in the analyses as they did not modify the results. One year after the booster dose, the blood samples of 292 of 301 (97%) subjects in the EB group and 268 of 272 (98.5%) subjects in the RB group were available for analysis.

Prebooster Serologic Results.

Five years after the primary vaccination, 285 (94.7%) subjects in the EB group and 259 (95.2%) in the RB group (P = 0.85) still had detectable level of anti-HBs. Respectively, 263 (87.4%) and 222 (81.6%) (P = 0.06) had a titer ≥10 mIU/mL. The GMT was significantly higher in the EB group, 252 mIU/mL versus 66 mIU/mL in the RB group (P < 0.0001) (Table 1). There was a 29-fold decrease from the GMT baseline 5 years after the primary vaccination in the EB group and a 56-fold decrease in the RB group. None of the subjects had higher anti-HBs titers or became HBsAg or anti-HBc positive during this 5-year period.

Serologic Profiles and Geometric Mean Titers of Anti-HBs Antibodies After Primary Vaccination and Before and After Administration of Booster Dose

The relationship between the postprimary vaccination titer anti-HBs and the titer 5 years later is presented in Table 2. The 11 subjects who had a titer <100 mIU/mL after the primary vaccination dropped below 10 mIU/mL 5 years later, and 9 of them (82%) had no detectable anti-HBs.

Concordance of Anti-HBs Titers 1 Month After Primary Vaccination, Before Booster Dose and 1 Month and 1 Year After Booster Dose

Immunogenicity of the Booster Dose.

One month after administration of the booster dose, >99% of the subjects in both cohorts had anti-HBs titers ≥10 mIU/mL. The GMTs were 113,201 mIU/mL in the EB group and 16,623 mIU/mL in the RB group (P < 0.0001) (Table 1). The GMTs rose from the postprimary vaccination baseline 15.7- and 4.5-fold, respectively, in the EB and RB groups. No severe adverse event was observed. Fifty-seven (9.9%) subjects had postbooster titers lower than their postprimary vaccination titer, 17 (5.6%) in the EB group and 40 (14.7%) in the RB group (P < 0.001). Three subjects had a titer between 1 and 10 mIU/mL, and 2 had no detectable anti-HBs after booster administration. In these 5 subjects, the titer was between 10 and 99 mIU/mL after the primary vaccination, and they had nondetectable anti-HBs before administration of the booster dose. They were all in good health. The GMT rose 449-fold over the prebooster baseline in the EB and 252-fold in the RB group. The increase in the anti-HBs titer between the pre- and postbooster results was at least 4-fold in all subjects who reached 10 mIU/mL or more after booster dose administration.

One Year After Booster Results.

One year after administration of the booster dose, >99% of the subjects in both cohorts had anti-HBs titers ≥1 mIU/mL, and 97.9% in the EB group and 98.5% in the RB group had anti-HBs ≥10 mIU/mL. The GMTs were 14,028 mIU/mL in the EB group and 3437 mIU/mL in the RB group (P < 0.0001) (Table 1). During the first year after the booster dose, an 8.1-fold decrease in the GMT was observed in the EB group and a 4.8-fold decrease in the RB group. The GMT was 2 times higher than after primary vaccination in the EB group and similar to that obtained 1 month after primary vaccination in the RB group. The majority of subjects had a titer ≥100 mIU/mL 1 year after the booster dose, 92.8% in the EB group and 89.2% in the RB group.

Eight subjects had a titer between 1 and 9 mIU/mL, and 2 had nondetectable anti-HBs 1 year after booster administration. In these 10 subjects, the titer was between 10 and 999 mIU/mL after the primary vaccination and 1 month after the booster dose. All 10 had no detectable anti-HBs before administration of the booster dose. No health problems were reported in these subjects.

The relationships between the postprimary vaccination, prebooster and 2 postbooster anti-HBs titers are presented in Table 2. There was a strong correlation between the postprimary, prebooster and postbooster anti-HBs titers (r = 0.75–0.88 in EB group and r = 0.73–0.83 in RB group).

In the uni- and multivariate analyses of the relationships between the host variables, vaccine-related variables and immune responses to the booster vaccination, no association was found.


These results confirm that pediatric doses of recombinant hepatitis B vaccines confer protection in preteenagers. Both vaccines performed very well, but statistically significant differences in GMTs obtained in the EB and RB groups were observed at 4 different times. Five years after primary immunization, a significant decrease in the antibody titers was observed, but only 5% of the vaccinees had no detectable antibodies. Persisting protection was demonstrated by the response to the booster dose (at least 4-fold increase) in >99% of the subjects. No definition or consensus about anamnestic response to HB vaccine exists. In previous studies, different criteria were used: 4-fold increase in anti-HBs titers,13 ≥10 mIU/mL 1 month after vaccination,14 ≥10 mIU/mL anti-HBs 10–14 days postbooster in persons with baseline anti-HBs <10 mIU/mL,15 >20 mIU/mL,16 ≥50 mIU/mL and >100 mIU/mL17 4 weeks after the initial dose of vaccine; the criteria of a 2-fold increase and at least 100 or 200 mIU/mL were also used.18

Strong correlation among post-primary vaccination, prebooster and post-booster antibody titers in our subjects confirms the results of Alaska studies about the importance of post vaccination response in long-lasting immune memory and protection.10

There are few long term immunogenicity studies, all were conducted in high endemicity areas where natural boosters might influence the persistence of protection5,7,19–22 and most involved infants. The number of subjects was often small, the subjects’ initial immune response was unknown and the observation period was short. In 1996, West and Calandra23 published a review of these studies and concluded that routine booster vaccination should not be needed to sustain immunologic memory and protection within 5 years and perhaps longer after the primary vaccination series. This view was reaffirmed by the European Consensus Group on Hepatitis B Immunity in 2000.24

A few more studies have been published since. Williams et al7 examined the persistence of antibodies after HB immunization at birth and the response to a subsequent challenge with a booster dose vaccine. One group (n = 70) received a recombinant vaccine and was followed for 5 years. The other (n = 42) received a plasma-derived vaccine and was followed for 9 years. Before the booster dose, 41 and 39% of the 5- and 9-year-old children, respectively, had a protective titer. All children at 5 years and 93% at 9 years had an anamnestic response. Seto et al25 tested 42 healthy 6-year-old children who had been immunized at birth. All children had titers >10 mIU/mL after immunization. Six years later, nearly one-half became seronegative, and only one-fifth retained a titer of at least 10 mIU/mL. All regained a protective anti-HBs titer of at least 10 mIU/mL 30 days after a booster dose. Dentico et al26 performed a study with 64 healthy 18- to 50-year-old individuals who had a seroprotective titer after primary vaccination and no protective titer of anti-HBs before the booster dose (interval between primary vaccination and booster not mentioned). Four dosages of crude nonadsorbed HBsAg were used. Between 75 and 100% of the subjects had a seroprotective titer after the booster dose. In an observational study of vaccine efficacy 14 years after a trial of various HB vaccines in Gambia, Whittle et al19 found the efficacy to be 94% against chronic carriage and 80% against infection. Among the uninfected participants, 36% had no detectable anti-HBs. Low peak antibody response was a risk factor for chronic carriage. From this type of data, it is not easy to state the proportion of vaccinees who will remain protected and for how long.

In our study, fewer than 1% of subjects lost their immune memory despite having had initial titers in excess of 10 mIU/mL. The clinical implication of the absence of anamnestic response is unknown. All the children were weak responders initially (10-99 mIU/mL). Because the weak responders were a small proportion of the preteenagers, it is unlikely to have had any serious impact on the effectiveness of the universal immunization programs at the population level. This might be different in infants because they have lower initial titers. Immune memory may not completely correlate with the ability to mount an anamnestic response to a booster dose. Because the incubation for HBV is long, cellular memory may play a role, and exposed persons may become infected but not develop disease or become chronic carriers. This may explain why many studies show vaccinated persons developing “breakthrough” anti-HBc-positive status without clinical hepatitis or chronic carriage.

In the following phases of this study, it will be important to observe whether the proportion of subjects losing their immune memory increases significantly after 10 or 15 years. Booster doses are useful only if there is an immune memory left; otherwise the primary immunization must be done again. Furthermore the importance of the initial level of antibodies titers and the effect of a booster dose on the duration of immune memory will be of interest.


We thank the Laboratoire de Santé Publique du Québec, particularly Jean Rochefort, for serologic tests and Louis Rochette for statistical support; Colette Couture, Manon Loisel, Sophie Auger, France Lavoie and the research nurses and technicians for their assistance; and Pierre Van Damme for helpful comments.


1. Koff RS. Immunogenicity of hepatitis B vaccines: implications of immune memory. Vaccine. 2002;20:3695–3701.
2. de Franchis R, Hadengue A, Lau G, et al. EASL International Consensus Conference on Hepatitis B: September 13–14, 2002, Geneva, Switzerland: consensus statement (long version). J Hepatol. 2003;39(suppl 1):S3–S25.
3. Global progress toward universal childhood hepatitis B vaccination. MMWR. 2003;52:868–870.
4. Jilg W, Schmidt M, Deinhardt F. Immune response to hepatitis B revaccination. J Med Virol. 1988;24:377–384.
5. Berlioz-Arthaud A, Perolat P, Buisson Y. Ten year assessment of infant hepatitis B vaccination program, in the Loyalty Islands, New Caledonia. Vaccine. 2003;21:2737–2742.
6. Keating GM, Noble S. Recombinant hepatitis B vaccine (Engerix-B): a review of its immunogenicity and protective efficacy against hepatitis B. Drugs. 2003;63:1021–1051.
7. Williams IT, Goldstein ST, Tufa J, Tauillii S, Margolis HS, Mahoney FJ. Long term antibody response to hepatitis B vaccination beginning at birth and to subsequent booster vaccination. Pediatr Infect Dis J. 2003;22:157–163.
8. Al-Faleh FZ, Al-Jeffri M, Ramia S, et al. Seroepidemiology of hepatitis B virus infection in Saudi children 8 years after a mass hepatitis B vaccination programme. J Infect Dis. 1999;38:167–170.
9. Lin YC, Chang MH, Ni YH, Hsu HY, Chen DS. Long-term immunogenicity and efficacy of universal hepatitis B virus vaccination in Taiwan. J Infect Dis. 2003;187:134–138.
10. Wainwright RB, Bulkow LR, Parkinson AJ, Zanis C, McMahon BJ. Protection provided by hepatitis B vaccine in a Upik Eskimo population: results of a 10-year study. J Infect Dis. 1997;175:674–677.
11. Lee PI, Lee CY, Huang LM, Chen JM, Chang MH. A follow-up study of combined vaccination with plasma-derived and recombinant hepatitis B vaccines in infants. Vaccine. 1995;13:1685–1689.
12. Duval B, Boulianne N, De Serres G, Laflamme N, et al. Comparative immunogenicity under field conditions of two recombinant hepatitis B vaccines in 8-10-year-old children. Vaccine. 2000;18:1467–1472.
13. Watson BM, Di Prinzio S, West D. Persistence of hepatitis B antibody and anamnestic response to a booster dose in healthy adolescents and adults. Paper presented at: 32nd National Immunization Conference; July 21–24, 1998; Atlanta, GA. Abstract 32.
14. Lai CL, Lau JY, Yeoh EK, Chang WK, Lin HJ. Significance of isolated anti-HBc seropositivity by ELISA: implications and the role of radioimmunoassay. J Med Virol. 1992;36:180–183.
15. Williams JL, Christensen CJ, McMahon BJ, Bulkow LR, et al. Evaluation of the response to a booster dose of hepatitis B vaccine in previously immunized healthcare workers. Vaccine. 2001;19:4081–4085.
16. Aoki SK, Finegold D, Kuramoto IK, Douville C, et al. Significance of antibody to hepatitis B core antigen in blood donors as determined by their serologic response to hepatitis B vaccine. Transfusion. 1993;33:362–367.
17. McIntyre A, Nimmo GR, Wood GM, Tinniswood RD, Kerlin P. Isolated hepatitis B core antibody - can response to hepatitis B vaccine help elucidate the cause? Aust N Z J Med. 1992;22:19–22.
18. Lai CL, Wong BC, Yeoh EK, Lim WL, Chang WK, Lin HJ. Five-year follow-up of a prospective randomized trial of hepatitis B recombinant DNA yeast vaccine vs. plasma-derived vaccine in children: immunogenicity and anamnestic response. Hepatology. 1993;18:763–767.
19. Whittle H, Jaffar S, Wansbrough M, Mendy M, et al. Observational study of vaccine efficacy 14 years after trial of hepatitis B vaccination in Gambian children. BMJ. 2002;325:569–573.
20. Poovorawan Y, Sanpavat S, Theamboonlers A, Safary A. Long term follow-up (11 to 13 years) of high-risk neonates vaccinated against hepatitis B. Paper presented at: Tenth International Symposium on Viral Hepatitis and Liver Disease; April 9–13, 2000; Atlanta, GA. Abstract 24.
21. Yuen MF, Lim WL, Cheng CC, Lam SK, Lai CL. Twelve-year follow-up of a prospective randomized trial of hepatitis B recombinant DNA yeast vaccine versus plasma-derived vaccine without booster doses in children. Hepatology. 1999;29:924–927.
22. Wu JS, Hwang LY, Goodman KJ, Beasley RP. Hepatitis B vaccination in high-risk infants: 10-year follow-up. J Infect Dis. 1999;179:1319–1325.
23. West DJ, Calandra GB. Vaccine induced immunologic memory for hepatitis B surface antigen: implications for policy on booster vaccination. Vaccine. 1996;14:1019–1027.
24. European Consensus Group. Are booster immunisations needed for lifelong hepatitis B immunity? Lancet. 2000;355:561–565.
25. Seto D, West DJ, Ioli VA. Persistence of antibody and immunologic memory in children immunized with hepatitis B vaccine at birth. Pediatr Infect Dis J. 2002;21:793–795.
26. Dentico P, Crovari P, Lai PL, Ponzio F, Safary A. Anamnestic response to administration of purified non-adsorbed hepatitis B surface antigen in healthy responders to hepatitis B vaccine with long-term non-protective antibody titers. Vaccine. 2002;20:3725–3730.

hepatitis B vaccination; booster; long term immunity

© 2005 Lippincott Williams & Wilkins, Inc.