| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
 |
ABSTRACT |
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Our population-based, age-stratified serosurvey of HBV markers in Vietnam2 indicated a very high seroprevalence of HBV infection commencing early in life. It confirmed the findings of other published3–5 and unpublished surveys, and highlighted the need for a program of immunization beginning at birth. However, delivery of vaccines to newborns in Vietnam is not easy because the rate of home birth is high (average = 21.3%, and up to 56.1% in the northern highlands6) and access of health workers in rural areas can be very difficult. Furthermore, the scarcity of refrigeration at local health stations makes it difficult for these workers to provide heat-sensitive vaccines on a daily basis. Similar problems exist throughout the Mekong region.
After a preliminary report from China on the performance of hepatitis B vaccine (HepB vaccine) stored at ambient temperature,7 studies in Indonesia have confirmed the feasibility8 and success (in terms of immunogenicity)9 of storing it outside the cold chain (OCC) for administration within a few days of birth to infants born at home. These field studies have made use of the thermostability of HepB vaccine, as demonstrated in controlled animal10 and human studies,11,12 and as verified by WHO.13 However, published field studies7,9 lack detail on the temperatures at which the vaccine was exposed, the average duration of OCC storage prior to use, and in one case9 failed to allow for maternal antibody amongst the young infants assessed.
From 1998 to 2001, we implemented a project to evaluate three alternative strategies for the introduction of HepB vaccine in rural Vietnam. Two of these strategies included the administration of a birth dose after storage of the vaccine at ambient temperature for up to one month. We used a locally produced, low-dose vaccine, at that time the only option for the Expanded Program on Immunization (EPI) for Vietnam.
The project was evaluated by assessing the operational success of the different strategies used and their scientific outcome in terms of the immunogenicity and protective efficacy of the vaccine used. We add to the existing knowledge on field use of HepB vaccines after storage at ambient temperature by comparing the immunogenicity of three doses of the local vaccine after storage of the first dose OCC or refrigerated.
|
METHODS |
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Project activities began in nine geographically dispersed communes in each district in early 1998. The population of these communes was approximately 107,000 and the prevailing population growth rate was approximately 1.5%.
Vaccine. The HepB vaccine used was produced by the National Institute of Hygiene and Epidemiology (NIHE) in Hanoi. It is a plasma-derived vaccine that has been assessed by the national licensing authority as safe, immunogenic, and heat stable. The vaccine is sold to the EPI at a dosage of 2.5 µg/0.5 mL of hepatitis B surface antigen (HBsAg) protein in two-dose vials. No information was available regarding the lot-to-lot variation of this vaccine or whether the vaccine used came from the same or different lots. Independent testing for the HBsAg content of the vaccine was not possible, but the same national licensing authority was involved in vaccine quality assurance during the project period.
Vaccination strategies. In Vietnam the EPI is conducted as a monthly activity and reliable refrigeration is not available beyond district level in most areas. In project areas, the first EPI contact of many infants was not until two or more months after birth. To improve this, alternate strategies of HepB vaccination were introduced. In the first strategy, birth doses were available using vials stored at ambient temperature (OCC) in commune health centers for up to one month, followed by two additional doses stored inside the cold chain (ICC) of the EPI and given at ages two and three months. In the second strategy, all doses of HepB vaccine were stored ICC, and were thus only available on the monthly EPI day.
The OCC strategy was introduced in six communes and the ICC strategy in three communes in each of the two districts. Since the capacity of certain communes to administer a birth dose of HepB vaccine was anticipated to be limited, allocation of these 18 communes to their respective strategies was geographically dispersed, but not random, to ensure that an adequate number of infants received a birth dose for our research purposes.
Vaccination commenced in these communes in November 1998. Doses were provided free of charge and at the time these were the only rural communes in the country to have introduced free, routine HepB vaccination.
Serologic and vaccination survey. The OCC and ICC groups were compared by means of a serologic survey conducted over an eight-day period in mid-October 2000. Venous blood was collected from 1,106 infants 9–18 months of age, the entire reported birth cohort. Data pertaining to the birth dates of infants and each dose of HepB vaccine were obtained from EPI registers held at the health center of each commune.
Informed oral consent for participation of their infants was given by mothers after a community-wide publicity campaign before this survey. All project activities including this survey were reviewed and approved by the NIHE, the seat of the Vietnam EPI, leaders of the participating communities, and the Vietnam Ministry of Health. Mothers and infants who participated were given a small, non-monetary gift in return for their cooperation.
Analysis of samples. After separation of the samples on the evening of collection, serum was frozen for later analysis at NIHE. The NIHE participates in the WHO quality assurance program and conducted serologic testing for this study according to standard laboratory procedures. All sera were coded with a unique identifier, but laboratory staff were not blind as to the origin of each sample. All sera were assessed by enzyme-linked immunosorbent assay (Sanofi Diagnostics Pasteur, Marnesla-Coquette, France) for HBsAg. The HBsAg-negative samples were tested for antibody to HBsAg (anti-HBs). Quality assurance testing was not done in this survey because laboratory capacity was previously deemed adequate.2
Data analysis. Data were entered by two different persons using Epi-Info version 6.04d (Centers for Disease Control and Prevention, Atlanta, GA) and analyzed using Stata release 6 (Stata Corporation, College Station, TX). Only infants who had received three doses of vaccine were included in analyses of anti-HBs titers. The geometric mean titer (GMT) was used to compare anti-HBs in infants grouped by district and first-dose strategy (OCC or ICC). Because of the skewed distribution of anti-HBs levels, for comparison among infants within the OCC group the titer of each infant was transformed by taking its natural logarithm. Comparisons between groups of infants were performed using the chi-square test for proportions and the t-test for means. Multiple linear regression was used to examine the associations between independent variables and anti-HBs titers, simultaneously adjusting for potential confounders. Variables were selected into the multivariate model based on prior knowledge, and included the quadratic terms of two continuous variables known to influence anti-HBs titer, the delay between doses 2 and 314–16 and that between dose 3 and sampling,17 and the delay between birth and dose 1. It also included the two-way interaction between district and strategy.
To further assess the validity of the calculated GMTs, predicted anti-HBs levels were also assessed for each group of infants based on the mathematical model available at http://www2.stat.unibo.it/palareti/vaccine.htm.17 This model enables the prediction of anti-HBs levels on any date in the future based on the titer at baseline and the delay until it is measured again. To make these estimates, the GMT for each group was compared with the actual level from the group with the shortest delay between dose 3 and sampling.
|
RESULTS |
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Most infants (98.4% in Quang Xuong and 77% in Ngoc Lac, total = 973) had received all three doses of vaccine at the time of sampling. There were wide differences in the delay between doses 2 and 3. In Quang Xuong, in each case the mean delay in this interval for OCC, ICC, and all communes was 31 days, compared with 58, 83, and 60 days, respectively, in Ngoc Lac. In each case these differences were significantly longer in Ngoc Lac (P < 10–4 for each comparison).
As expected, the mean delay between dose 3 and the date of sampling varied widely. For infants in Quang Xuong (324 days) it was much longer than those in Ngoc Lac (257 days) (t = 11.4, P < 10–5). When assessed by strategy, the mean delay was also longer among infants in OCC (300 days) than ICC (283 days) communes (t = 2.5, P = 0.01). However, when broken down by district, there was no difference in this delay between infants in OCC (324 days) and ICC communes (323 days) in Quang Xuong, but a large difference between Ngoc Lac OCC (272 days) and ICC communes (205 days) (t = 5.4, P < 10–5).
Prevalence of HBsAg. The overall proportion of infants who were HBsAg positive was 6.5% (n = 72), a decrease of 6% from the 12.5% in our published baseline survey.2 The current proportion ranged from 5% in Quang Xuong (n = 28), a decrease of 3.3% from baseline, to 8.2% (n = 44) in Ngoc Lac, a decrease of 8.3% from baseline.
Prevalence of anti-HBs.
Overall, anti-HBs was detected among 81.5% of 1,105 infants able to be tested (95% confidence interval [CI] = 79.1–83.8). A protective level of anti-HBs (
10 mIU/mL) was present in 80.3% (95% CI = 77.8–82.7) of 1,068 infants for whom sufficient serum was available for quantitative testing. Among the 996 HBsAg-negative infants for whom quantitative anti-HBs levels were measured, 86.1% (95% CI = 83.8–88.2) were protected.
Table 1
shows the prevalence of a protective level of anti-HBs after three doses of vaccine by district and strategy and, in the relevant communes, by the number of days dose 1 was OCC. Prevalence did not differ between any groups of infants (by chi-square test).
|
shows the GMTs among infants who had received three doses of vaccine by district and strategy. The titer was slightly higher in ICC than OCC communes, but this difference was not significant on crude analysis (t = 1.19, P = 0.23). Infants in Ngoc Lac had significantly higher titers than those in Quang Xuong both overall (t = 2.92, P < 0.01) and in the OCC group (t = 2.50, P = 0.01), but not in the ICC group (t = 1.34, P = 0.18).
|
also shows GMTs differentiated by duration of storage of dose 1 OCC. In Quang Xuong, where mean dose intervals did not vary across communes, there was no difference between the titers of infants in ICC communes (GMT = 114 mIU/mL) and those in communes storing dose 1 OCC for 15–31 days (94 mIU/mL) (t = 0.84, P = 0.40). A similar comparison was not attempted for infants in Ngoc Lac because of the widely differing intervals between doses 2 and 3, and thus delays between dose 3 and sampling.
Figure 1 shows a dot plot of the natural logarithms of anti-HBs titers of infants in the OCC communes according to the number of days the vial containing their first dose had been kept at ambient temperature. The line connects the medians of these logarithms for all infants receiving dose-1 on each day. The flat slope of this line suggests no relationship between anti-HBs titers and the duration for which dose 1 was stored OCC.
|
Anti-HBs decay.
Table 2 also shows the GMTs of anti-HBs predicted by extrapolating from the group with the shortest delay between dose 3 and sampling (Ngoc Lac ICC group) using a published mathematical model of anti-HBs decay.17 Although there was overall support for the decay model, in each case in Quang Xuong the actual GMT was lower than predicted.
Temperature exposure.
Table 3 shows the average maximum and minimum temperatures in the two project districts. The distinct seasonal changes noted are typical of lowland northern Vietnam and also of many other tropical countries north of the Tropic of Cancer. However, night-time temperatures during the summer months are higher than in many tropical areas and have an impact on the average temperature to which vaccine OCC was exposed.
|
|
DISCUSSION |
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
It is possible that maternal antibody was still present in some of the youngest infants sampled, but not in more than a few percent16 and presumably equally across the groups. It is also possible that one or even two of the doses in any individual infant or group was impotent, but that the anti-HBs response was still adequate to provide protection. However, if this was more common among the infants in OCC communes and due to heat spoilage, one might expect the GMT to be lower in this group, which it was not.
We were careful to allow for programmatic influences on the titers of antibody among infants in the different groups, and the analyses suggest that these variables did influence anti-HBs titer. In Ngoc Lac, there were longer intervals between doses 2 and 3 of vaccine, which has usually been shown to result in higher antibody titers.14–16,18 This applied overall, as well as in ICC and OCC communes, when assessed separately. A shorter interval between the last dose of vaccine and the collection of blood allows less time for antibody decay.17 This interval was shorter in Ngoc Lac, but it was also much shorter in Ngoc Lac ICC than OCC communes. This difference did not apply in Quang Xuong; thus, the overall shorter delay in ICC than OCC communes must have been due to the influence of Ngoc Lac. We conclude that the differences in the titers of anti-HBs observed between the different groups of infants may be explained by differences in the timing of doses between the two districts and between ICC and OCC communes in Ngoc Lac, and not by any other variable assessed.
Ngoc Lac ICC communes clearly performed relatively poorly in vaccinating infants on time. As described, the choice of vaccine storage for the nine communes in each district was not random. Of necessity for our research on the efficacy of the vaccine used, in some cases this choice included consideration of whether staff in those communes would likely be able to deliver a birth dose of vaccine to all infants, as required by the OCC strategy. Some Ngoc Lac ICC communes were more remote or had a weaker preventive health service than others. This obviously affected their capacity to deliver vaccines in a timely way.
The relatively low rate of a protective level of anti-HBs (80.3%) in this study warrants comment. The dosage of HBsAg protein in the vaccine assessed is lower than that used in many commercially available HepB vaccines. In another unpublished study, we evaluated the immunogenicity of the same locally produced, plasma-derived HepB vaccine in two different doses and compared it with that of two 10 µg/dose Korean recombinant HepB vaccines. In that trial, after vaccine storage in experimental conditions at 2–8°C and administration of all doses to each infant by the same nurse or doctor, the 2.5 µg/dose vaccine induced a protective level of antibody in 85.9% of 71 eight-month old infants whose mothers were seronegative for all HBV markers. This was not as high as the percentages among those given 5 µg doses of this vaccine or the Korean vaccines, but compared well to another field trial of a HepB vaccine of similar HBsAg content.19
In addition, the rate of 80.3% was achieved under field conditions, with the associated risks of vaccine freezing (which can render it impotent10), variable standard injection techniques, and a wide variation in scheduling. Infants in this study were also older than those in the controlled study, which allowed time for antibody decay. Finally, overall rates of birth dosing were rather low in this cohort of infants, which may have influenced rates of perinatal infection and thus HBsAg prevalence. The background rate of HBsAg positivity among mothers in the two districts surveyed was 15% in our previous survey.2 Thus, it seems likely that a number of infants were infected by their mothers very early in life, which affected the rate of response to the vaccine. The higher rate of protection among HBsAg-negative infants (86.1%) almost certainly includes a higher proportion of infants born to HBsAg-negative mothers, and is therefore reassuring with regard to the overall immunogenicity of the vaccine. Our studies suggest that higher antigen content in the local vaccine might yield a higher rate of protection, but in most infants the existing formulation, which will be phased out in 2006, is immunogenic and protective into the second year of life.
The GMTs calculated by the decay model were lower in Quang Xuong than predicted. Our use of the model assumed that infants were selected from the same population and represented by the group with the shortest delay between dose 3 and sampling. However, this was not the case because the infants in Quang Xuong were vaccinated with a shorter interval between doses 2 and 3, which is a key factor in the highest level of anti-HBs achieved before commencement of decay, than those in the reference group (Ngoc Lac ICC communes). Thus, although there was support for the model in the Ngoc Lac group, it was not appropriate to use it for infants who were vaccinated according to a different schedule. This is further indirect evidence for the influence of the interval between doses 2 and 3 on anti-HBs levels.
The prevalence rates of HBsAg among the infants surveyed were not as low as in most field trials of HepB vaccine.18 A discussion of these findings is beyond the scope of this report, but they may relate to the low dose of vaccine used,20 vaccination technique, the timing of the first dose (which was often late in Ngoc Lac), or possibly freezing of the vaccine before it was taken OCC.
This report adds to the existing literature on use of HepB vaccine OCC and expands upon certain parameters not previously discussed. In particular, as well as providing information on the range of ambient temperatures to which dose 1 of the vaccine was exposed, we have shown that the prevalence of anti-HBs or the GMT induced did not vary according to the duration of this exposure up to one month. This concurs with the existing scientific information on heat stability of HepB vaccine,7,9,11–13 and provides further support for its storage OCC in rural areas of developing countries that lack refrigeration. Moreover, the recent application of high-temperature vaccine vial monitors on HepB vaccines of many manufacturers and the assertion of WHO that these may be used to monitor vaccines stored OCC21 is also reassuring on the safety and effectiveness of HepB vaccines stored OCC.
Received April 26, 2005. Accepted for publication September 9, 2005.
Acknowledgments: We gratefully acknowledge the support of the scientists and administrative staff at the National Institute of Hygiene and Epidemiology in Hanoi, the staff of the District and Commune Health Services in Quang Xuong and Ngoc Lac, Mirella Ozols (University of Melbourne), and the families of infants who participated in the study.
Financial support: This study was supported by the Australian Agency for International Development as part of the broader primary health care project "Strengthening Immunization and Malaria Control in Vietnam."
* Address correspondence to Beverley-Ann Biggs, Department of Medicine (Royal Melbourne Hospital/Western Hospital), University of Melbourne, Royal Melbourne Hospital, Parkville 3050, Australia. E-mail: babiggs{at}unimelb.edu.au 
Authors’ addresses: David B. Hipgrave, United Nations Children’s Fund Indonesia, Wisma Metropolitan II, Jalan Jend. Sudirman, Kav. 31, Jakarta 12920, Indonesia. Trung Nam Tran, Department of Medical Epidemiology and Biostatistics, Karolinska Institute, Box 281, SE 171 77, Stockholm, Sweden, E-mail: dhipgrave{at}unicef.org. Vu Minh Huong and James E. Maynard, Program for Appropriate Technology in Health, Box 252, International Post Office, Hanoi, Vietnam. Do Tuan Dat, Nguyen Tuyet Nga, Hoang Thuy Long, and Nguyen Thu Van, National Institute of Hygiene and Epidemiology, 1 Yersin Street, Hanoi 10000, Vietnam. Beverley-Ann Biggs, Department of Medicine (Royal Melbourne Hospital/Western Hospital), University of Melbourne, Royal Melbourne Hospital, Parkville 3050, Australia, Telephone: 61-3-8344-3257, Fax: 61-3-9347-1863, E-mail: babiggs{at}unimelb.edu.au.
|
REFERENCES |
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  D. Davaalkham, T. Ojima, S. Wiersma, T. Lkhagvasuren, P. Nymadawa, R. Uehara, M. Watanabe, I. Oki, and Y. Nakamura Administration of hepatitis B vaccine in winter as a significant predictor of the poor effectiveness of vaccination in rural Mongolia: evidence from a nationwide survey J. Epidemiol. Community Health, July 1, 2007; 61(7): 578 - 584. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
