Hepatitis A virus (HAV), a hepatotrophic pathogen, is a common cause of acute viral hepatitis, the most prevalent infectious liver disease in young children.1 Hepatitis A is asymptomatic in majority of infected children, and in children with symptomatic infection—it is often associated with flu-like or other nonspecific symptoms.1 The illness’ severity increases with age and in patients suffering from chronic liver disease, with an arising risk of fulminant hepatic failure and the case-fatality ratio reaching 2.1% among adults aged ≥40 years.1 Approximately 1.5 million cases of HAV infections are reported annually worldwide. Among several global regions with HAV outbreaks, Africa, Asia and Central and South America have been identified as regions with high endemicity for HAV infections.2
Immunization against HAV is well established as an effective mean of providing long-lasting protection from infection. According to the Centers for Disease Control and Prevention’s 2013 national surveillance data on viral hepatitis, 1781 cases of HAV infections were reported and approximately 3500 (2500–3900) actual cases are suspected in the United States.2 These numbers reflect a decline in the rate of acute hepatitis from 1 case per 100,000 in 2007 to 0.4 cases per 100,000 in 2011.3–5 In Israel, a >95% reduction in reported HAV cases was observed from 1993 to 2004 and was maintained through 2012.6 Such improvements in the epidemiological profiles have been accredited to the universal use of HAV vaccines and enhancements in childhood immunization programs.5
Several inactivated hepatitis A vaccines were developed in the early 1990s and licensed for use in the mid-1990s. Both childhood and adult dosages of these vaccines have an excellent immunogenicity with an acceptable safety record and follow similar immunization schedules.7 Majority of these vaccines are produced by growing HAV strains in human diploid fibroblasts followed by purification of cell extracts, inactivation using formaldehyde and adsorption onto aluminum hydroxide adjuvant.7 Epaxal, an aluminum-free virosomal HAV vaccine, is based on a different adjuvant system wherein the formalin-inactivated HAV are associated with the so-called immunopotentiating reconstituted influenza virosomes.8
Standard doses of Epaxal are shown to achieve protective antibody levels and have acceptable tolerability in several study populations.9–12 Following a 2-dose vaccination regimen, based on mathematical models and extrapolation of data collected 10–12 years after primary vaccination in adults, the standard dose of Epaxal was found to offer long-term seroprotection persisting for about 30 years in at least 95% of the vaccinated subjects.13 Studies on the pediatric dose of Epaxal, Epaxal Junior, have also reported satisfactory immune responses and safety profiles in children aged 1 to16 years14 and in toddlers.15 A long-term follow-up study initiated in the pediatric cohort confirmed that Epaxal Junior elicited real-time seroprotection for at least 5.5 years and that younger children showed lower antibody titers and a faster antibody decline than older ones.16 However, the long-term seroprotection conferred by Epaxal Junior in toddlers along with routine childhood vaccines [measles, mumps and rubella vaccine (MMR), oral polio vaccine (before April 1, 2005) and diphtheria, tetanus, Bordetella pertussis, Hemophilus influenzae type b and inactivated polio vaccine (DTPaHibIPV); routine childhood vaccine] has not been documented to date. Furthermore, antibody persistence following HAV vaccine administration could not be purely studied because of natural boosting by circulating virus HAV could not be ruled out.
The primary objective of the current report was to describe the duration of real time protection of the immunity awarded by Epaxal Junior administered along with RCV as compared with Havrix 720, the vaccine currently being used for hepatitis A immunization in Israel. In addition, the study also aimed to predict the long-term persistence of anti-HAV antibodies in these children using computer-based modeling analysis.
MATERIALS AND METHODS
The study protocol was approved by the appropriate local ethics committees as well as by the Israel National Health Authorities. The study was conducted in accordance with the ethical principles that have their origin in the Declaration of Helsinki and that are consistent with Good Clinical Practices and applicable regulatory requirements. Written informed consent was obtained from parents/legal guardians of participating children before enrolment.
The primary open-labeled, active-controlled, parallel-group phase 2 study was conducted at 2 centers in Israel. The results of the primary study are already published.15 In brief, healthy 12- to 15-month-old children were randomized (1:1:1) to 3 groups: group A, Epaxal Junior (day 1) + RCV (day 1); group B, Epaxal Junior (day 1) + RCV (day 29); and group C, Havrix 720 (day 1) + RCV (day 1). On day 1 (primary vaccination) and at month 6 (booster) children received a single intramuscular dose of either Epaxal Junior (groups A and B) or Havrix 720 (group C). The immunogenicity and safety assessments were carried out from day 1 to month 7.15 During the part of the study being the focus of this article, the follow-up visits were conducted 1.5, 2.5, 3.5, 5.75 and 7.5 years post second dose of study hepatitis A vaccination. No HAV vaccination was carried out during the follow-up period.
Blood samples (approximately 5 mL) were collected at all the follow-up visits (Fig. 1). For the 7.5-year immunogenicity evaluation, the sera from all the follow-up time points were tested in parallel using the same test kit batch. The microparticle enzyme immunoassay HAVAB 2.0 Quantitative for the AxSYM system (Abbott, Wiesbaden, Germany) was used with a detection limit of 6.0 mIU/mL. Levels <6.0 mIU/mL were arbitrarily set to half of the limit (3.0 mIU/mL) for calculation of geometric mean concentrations (GMCs). Antibody concentration of ≥10 mIU/mL was regarded as the cut-off indicative of seroprotection.
The primary objective was to perform a computer-based modeling analysis of long-term seroprotection elicited by Epaxal Junior based on the individual anti-HAV antibody concentrations collected during follow-up visits and compare these values with those obtained with Havrix 720. Prediction of long-term seroprotection was presented as median duration of seroprotection [0.5 quantile; with corresponding 95% confidence interval (CI)] and applied anti-HAV antibody concentration cut-off of ≥10 mIU/mL and a more stringent ≥20 mIU/mL cut-off for sensitivity analysis. The secondary objectives included estimation of GMCs and estimating proportion of children seroprotected in the follow-up periods.
Anti-HAV antibody concentrations were log10-transformed before calculating GMC. The data for GMCs and proportion of seroprotected children (≥10 mIU/mL) was presented by vaccine group and time point with corresponding 95% CI. Anti-HAV antibody results from all follow-up time points (1.5-, 2.5-, 3.5-, 5.75- and 7.5-year time point) were included in the analyses. All statistical analyses were performed using SAS (version 9.3).
Estimates of long-term seroprotection for each participant were made by using a linear mixed model17,18 based on the assumption that the yearly decline rate of concentrations is approximately constant. The time to reach a cut-off concentration (≥10 or ≥20 mIU/mL) was calculated for each child and data based on all actual available timepoints were used to estimate the model that used log-transformed antibody concentrations and incorporated gender effects. The time was defined as the actual number of years after the primary dose. The decline rates were estimated from the model and the prediction of duration of seroprotection to reach the 2 cut-off levels of ≥10 and ≥20 mIU/mL were obtained from Kaplan–Meier analysis.
Of the 291 children who completed the primary trial (group A: n = 99; group B: n = 95 and group C: n = 97), 228 were included in the follow-up study (Fig. 1). The baseline weight, height and BMI were similar across all 3 groups.15
Individual Serum Antibody Concentrations
As expected, the antibody decline was faster in the first 1 to 2 years immediately post second dose when compared with the subsequent follow-up observations (Fig. 2). Of the total 157 children evaluated at the 7.5-year time point, 152 (96.8%) children had seroprotective levels (≥10 mIU/mL); 5 children (group A: n = 1; group B: n = 2; group C: n = 2) had an antibody concentration <10 mIU/mL. Three children in each group had an antibody concentration between ≥10 and <20 mIU/mL; remaining 143 children had anti-HAV antibody concentrations ≥20 mIU/mL.
Geometric Mean Antibody Concentrations
At 1 month after the second dose, GMC values were >1000 mIU/mL in all 3 groups (Table 1). At 1.5 years after the second dose, the GMCs had considerably declined to 257 mIU/mL in group A, 243 mIU/mL in group B and 222 mIU/mL in group C. From this time point onwards, GMCs showed a slow decrease over time. At 7.5 years post second dose, GMCs were 85 mIU/mL in group A, 80 mIU/mL in group B and 61 mIU/mL in group C. Anti-HAV GMCs were similar in groups A and B at all the time points; however, these were lower in group C (Table 1).
The seroprotection rates were 100% at 1 month and 1.5 years after the second dose. They remained at this level up to 2.5 years in group C and up to 3.5 years in groups A and B (group C: 98%). At 7.5 years after the second dose, the seroprotection rates were 98.0% in group A, 96.3% in group B and 96.2% in group C (Table 2).
Prediction of the Long-term Seroprotection
With a cut-off of ≥10 mIU/mL, the median predicted duration of seroprotection based on data up to 7.5 years was 19.1 years (95% CI: 17.0–22.6) in group A, 18.7 years (95% CI: 16.2–23.1) in group B and 17.3 years (95% CI: 16.4–20.6) in group C. According to this prediction, 5% of children were protected for less than 11.3 years in group A, for less than 8.4 years in group B and for less than 8.7 years in group C (Fig. 3A). Applying the more stringent cut-off of ≥20 mIU/mL resulted in a median predicted duration of seroprotection of 15.4 years in group A, 15.0 years in group B and 13.8 years in group C (Fig. 3B).
Validation of the Mathematical Model
Based on the 3.5 years follow-up data, the linear mixed model had estimated an approximate median duration of seroprotection (≥10 mIU/mL) of 15.6 years in children of group A. This estimate was similar to the duration estimated for group B (14.4 years), and slightly higher numerically than the projected duration of seroprotection in group C (13.4 years). This model was validated by graphically comparing the individual projections of the log10 concentrations with the observed 7.5-year follow-up data. The predicted antibody concentrations at 7.5-year time point obtained from the model estimated using up to 3.5 years data generally matched well with the observed antibody concentrations at 7.5-year time point in all 3 groups although most observed values were higher than the predicted values (Fig. 4).
Childhood HAV vaccination courses that provide long-lasting immunity extending into adulthood are recommended as effective strategies for protection and subsequent elimination of hepatitis A in endemic areas.19 In this longest follow-up study after primary immunization with Epaxal Junior in toddlers, all children after 3.5 years post second vaccination maintained seroprotective levels of anti-HAV antibodies (≥10 mIU/mL). At the 7.5-year follow-up, 98% (group A) and 96% (group B) Epaxal Junior immunized children had maintained seroprotective anti-HAV antibody levels (≥10 mIU/mL). The seroprotection rates were similar between Epaxal Junior immunized and Havrix-720 immunized children at both follow-up time points.
High GMCs observed upon immunization with Epaxal Junior in toddlers further confirm the Advisory Committee on Immunization Practices guidelines recommending initiation of anti-HAV vaccination starting at 12 months old.20 The GMCs showed a rapid decline at 1.5 years post second vaccination, after which the antibody decay rate slowed down. The anti-HAV GMCs were similar between both the Epaxal Junior groups at all the time points, suggesting that the concomitant administration of RCV did not have any influence on the long-term immune response to Epaxal Junior. These results further support the outcome of primary study, suggesting no influence of RCV on the immunogenicity and safety of Epaxal Junior.15 Similar to observations on gender differences from previous HAV studies, the GMCs were higher for girls than boys in the Epaxal Junior groups both immediately after vaccination and throughout the follow-up period.14,16 At 7.5 years post second dose, the anti-HAV GMCs were similar in groups A and B (80–85 mIU/mL) but slightly lower in group C (61 mIU/mL).
Using a linear mixed model to estimate the persistence of antibodies over time, the median duration of seroprotection (≥10 mIU/mL) was predicted to be 19.1 years for Epaxal Junior when administered concomitantly with RCV (group A) and 18.7 years when administered separately (group B). Both medians were similar to Havrix-720 groups at 17.3 years.
In this study, the predicted values at 3.5 years post second dose generally matched well with the observed values at 7.5 years in all 3 groups; however, most observed values were higher than the predicted values, thus confirming the fact that follow-up of 3.5 years was too short for reliable estimations. Hence, it is possible that one can predict longer durations of antibody persistence if tested again at 10 or 12 years post second vaccination of Epaxal Junior, like indicates experience with Epaxal from studies in adults.13,21 Also, the real-time persistence of protective antibody levels at 7.5 years validates the mathematical modeling prediction at the 3.5-year follow-up.
At 7.5 years, Epaxal Junior revealed acceptable immunogenic response with median predicted long-term seroprotection ranging 18–19 years. However, if relying solely on the measure of the kinetic of antibody titers decline over time this long-term protection might be shorter in the youngest children than previously estimated based on experience in adults. The same mathematical model as used in Epaxal adult long-term follow-up study21 applied to the 5.5 years follow-up results in children from 1 to 17 years old vaccinated with 2 doses of Epaxal, Epaxal Junior or Havrix Junior, predict a median duration of seroprotection ranging between 18 and 39 years. The median predicted duration of seroprotection was generally lower in children ≤7 years old than in children ≥8 years (with cut-off ≥10 mIU/mL). The results from the youngest age groups [1–4 years old: for Epaxal Junior 20.6 years (95% CI: 18.0–22.3) and for Havrix Junior 22.0 years (95% CI: 19.7–26.6), whereas 5 to 7 years old age group: for Epaxal Junior 20.9 years (95% CI: 16.8–24.0) and for Havrix Junior 18.6 years (95% CI: 13.9–28.2)] were the shortest among all age groups and shorter than the mean results across the entire age range of that study population.16 The above results are analogous to the results from our study. The comparable data after up to 5 years of follow up in adults vaccinated with 2 doses of Epaxal resulted in median predicted duration of seroprotection of 55.5 years (95% CI: 48.7–69.7).21 Although the data do not permit robust conclusions, these results do not provide confirmation of the same duration of predicted seroprotection for children as established for adults.
There is an evidence that inactivated hepatitis A vaccines induce an immune memory, which provides protection beyond the persistence of anti-HAV antibodies, and that robust anamnestic responses are induced after booster vaccination or subsequent HAV encounter, not only in adults but also in toddlers and children. Thus, based on the reported immunological responses, the study suggests that within observation timeframe no booster vaccination is needed in children following the recommended primary course of Epaxal Junior. However, given the observed age-dependent differences in kinetics of antibody decay,16 additional follow-up studies would be needed beyond 7.5 years to further assess the long-term immunogenicity of hepatitis A vaccination in this particular population being target of vaccination programs in many countries.
The study was conducted at 2 centers in Israel where circulation of the HAV is almost eliminated due to widespread immunization against hepatitis A (started since 1993).5,6,22,23 An important advantage in studying long-term antibody persistence in Israel was that the introduction of HAV vaccines in toddlers to the Israeli National Immunization Plan in 1999 sharply reduced virus HAV circulation in the population,6,22,23 permitting pure assessment of antibody persistence with negligible rate of natural boosting. Hence, studying the persistence of antibodies in this population with minimal chances of viral exposure gives a pure approach to study the anti-HAV antibodies, unlike in other populations where presence of even a relatively weak virus circulating activity can make it difficult to assess if the persistence of antibodies is due to viral exposure or an immunogenic response to the vaccine alone.
However, the possibility of accidental viral exposures resulting in anamnestic response until the disease is completely eradicated cannot be overlooked. Another potential study limitation was a significant drop out rate along the follow-up period: from the children randomized to the original study only ~48% attended 7.5 years follow-up visit. Also, it should be noted that some children attending the 3.5- and 7.5-year follow ups missed visits or did not attend the follow-up at all which made the groups heterogeneous at the follow ups. Nevertheless, the baseline characteristics of children participating in the 5.75- or 7.5-year follow-up study visits remained similar to those participating in the primary study.15
The 2-dose vaccination regimen with Epaxal Junior in healthy children provided a real-time seroprotection up to 7.5 years. The median predicted persistence of seroprotective levels (≥10 mIU/mL) of anti-HAV antibodies was estimated to last at least 18.7 years. Additional follow-up studies are needed to further assess the long-term immunogenicity and to conclude on boostability of hepatitis A vaccines in the youngest children.
We acknowledge Shruti Shah, PhD (SIRO Clinpharm Pvt. Ltd.) for writing assistance and Bradford Challis, PhD (Janssen Research & Development, LLC) for additional editorial support for the development of this manuscript.
1. Koff RS. Hepatitis A. Lancet. 1998;351:1643–1649.
2. Alexander J, Manno M. Underuse of analgesia in very young pediatric
patients with isolated painful injuries. Ann Emerg Med. 2003;41:617–622.
3. Vogt TM, Wise ME, Bell BP, et al. Declining hepatitis A mortality in the United States during the era of hepatitis A vaccination. J Infect Dis. 2008;197:1282–1288.
4. Lopalco PL, Prato R, Chironna M, et al. Control of hepatitis A by universal vaccination of adolescents, Puglia, Italy. Emerg Infect Dis. 2008;14:526–528.
5. Dagan R, Leventhal A, Anis E, et al. Incidence of hepatitis A in Israel following universal immunization of toddlers. JAMA. 2005;294:202–210.
6. Levine H KE, Anis E, Givon-Lavi N, Dagan R. The impact of a national routine immunisation programme initiated in 1999 on Hepatitis A incidence in Israel, 1993 to 2012. Euro Surveill. 2015;20:3–10.
7. Weekly epidemiological record. World Health Organisation position paper on hepatitis A Vaccines, No. 28–29, 2012, 87, 261–276. http://www.who.int/wer
8. Just M, Berger R, Dreschsler H, et al. A single vaccination with an inactivated hepatitis A liposome vaccine induces protective antibodies after only two weeks. Vaccine. 1992;10:737–739.
9. Loutan L, Bovier P, Althaus B, et al. Inactivated virosome hepatitis A vaccine
. Lancet. 1994;343:322–324.
10. Poovorawan Y, Theamboonlers A, Chumdermpadetsuk S, et al. Safety, immunogenicity, and kinetics of the immune response to a single dose of virosome-formulated hepatitis A vaccine
in Thais. Vaccine. 1995;13:891–893.
11. Ambrosch F, Wiedermann G, Jonas S, et al. Immunogenicity and protectivity of a new liposomal hepatitis A vaccine
. Vaccine. 1997;15:1209–1213.
12. Usonis V, Bakasénas V, Valentelis R, et al. Antibody titres after primary and booster vaccination of infants and young children with a virosomal hepatitis A vaccine
(Epaxal). Vaccine. 2003;21:4588–4592.
13. Bovier PA, Bock J, Ebengo TF, et al. Predicted 30-year protection after vaccination with an aluminum-free virosomal hepatitis A vaccine
. J Med Virol. 2010;82:1629–1634.
14. Van Der Wielen M, Vertruyen A, Froesner G, et al. Immunogenicity and safety of a pediatric
dose of a virosome-adjuvanted hepatitis A vaccine
: a controlled trial in children aged 1-16 years. Pediatr Infect Dis J. 2007;26:705–710.
15. Dagan R, Amir J, Livni G, et al. Concomitant administration of a virosome-adjuvanted hepatitis a vaccine
with routine childhood vaccines
at age twelve to fifteen months: a randomized controlled trial. Pediatr Infect Dis J. 2007;26:787–793.
16. VanHerck K, Hens A, DeCoster I, et al. Long-term antibody persistence in children after vaccination with the pediatric
formulation of an aluminum-free virosomal hepatitis A vaccine
. Pediatr Infect Dis J. 2014;34;e85–e91.
17. Nuesch E, Rutjes AW, Husni E, Welch V, Juni P. Oral or transdermal opioids for osteoarthritis of the knee or hip. Cochrane Database Syst Rev. 2009:CD003115.
18. Mushtaq S, Choudhary R, Scanzello CR. Non-surgical treatment of osteoarthritis-related pain in the elderly. Curr Rev Musculoskelet Med. 2011;4:113–122.
19. Van Damme P, Banatvala J, Fay O, et al.; International Consensus Group on Hepatitis A Virus Immunity. Hepatitis A booster vaccination: is there a need? Lancet. 2003;362:1065–1071.
20. Pascual ML, Fleming RR, Gana TJ, et al. Open-label study of the safety and effectiveness of long-term therapy with extended-release tramadol in the management of chronic nonmalignant pain. Curr Med Res Opin. 2007;23:2531–2542.
21. Bovier PA, Bock J, Loutan L, et al. Long-term immunogenicity of an inactivated virosome hepatitis A vaccine
. J Med Virol. 2002;68:489–493.
22. Barkai G, Belmaker I, Givon-Lavi N, et al. The effect of universal toddlers-only hepatitis A virus vaccination program on seropositivity rate in unvaccinated toddlers: evidence for reduced virus circulation in the community. Pediatr Infect Dis J. 2009;28:391–393.
23. Belmaker I, Dukhan L, Yosef Y, et al. Elimination of hepatitis a infection outbreaks in day care and school settings in southern Israel after introduction of the national universal toddler hepatitis a immunization program. Pediatr Infect Dis J. 2007;26:36–40.