Journal of virological Methods

Journal of Virological Methods 56 (1996) 27-31

Determination of hepatitis A antibody response to vaccination by an enzyme immunoassay

Jane N. Zuckerma+* , Lee Powell, Rudolf M. Lequinb, Arie J. Zuckerman

“Academic Unit of Travel Medicine and Vaccines, The WHO Collaboraiing Centre for Reference and Research on Viral Diseases, Royal Free Hospital School of Medicine, Rowland Hill Street, London NW3 2PF, UK

bOrganon Teknika, Boxtel, The Netherlands

Accepted 10 July 1995

Abstract

The quantitative responses to vaccination with hepatitis A vaccine was determined in 113 volunteers using a commercially available enzyme-linked immunosorbent assay for total antibodies to hepatitis A. Administration of vaccine or control preparation was carried out according to two regimens; at 0, 1 and 12 months (regimen I) and at 0,0.5 and 12 months (regimen II). Seroconversion rates (concentrations of HAV antibodies > 50 IUjI) were between 94 and 97% at month 1 for regimen I and regimen II, respectively. The geometric mean titres (GMTs) fell gradually by month 12, and increased rapidly lo- 100 fold 1 month after the booster dose at month 12. The GMTs of the groups receiving the control preparation remained below 50 IU/l. No significant differences were found between the antibody responses after regimen I or regimen II. It is concluded that the antibody test (Hepanostika HAV Antibody) can be used safely and adequately for quantitation of responses to hepatitis A immunisation.

Keywords: Hepatitis A vaccine; HAV antibodies; Quantitation; Enzyme immunoassay

1. Introduction

Infection with hepatitis A virus is particularly prevalent in countries with poor sanitation and hygiene. As a result of improved standards of hygiene and socio-economic conditions, the inci- dence of the infection and the prevalence of anti- bodies to hepatitis A have been falling in developed countries (Hofmann et al., 1992). A shift in age-related seroprevalence from the very

* Corresponding author. Tel.: + 44 0171 794 0500 ext. 5744, fax: +44 0171 830 2268.

young to an older age group (Zuckerman et al., 1991) has occurred so that the average age of exposure to hepatitis A virus increases with a paradoxical increase in morbidity and mortality.

Until recently, high-risk groups including trav- ellers, military personnel, sewage workers, and staff and residents at institutions where personal hygiene may be poor received passive immunisa- tion with human immunoglobulin. This afforded protection for 4-6 months, caused local discom- fort at the site of inoculation and required re- peated injections to ensure continuing protection. Since 1979, the propagation of hepatitis A virus in

0166-0934/96/$15.00 0 1996 Elsevier Science B.V. All rights reserved

SSDI 0166-0934(95)01898-A

28 J.N. Zuckerman et al. / Journal of Virological Methods 56 (1996) 27.-31

tissue culture has led to the development of vac- cine preparations which provide long term protec- tion (Provost and Hilleman, 1979). Hepatitis A vaccine (Havrix) using the HM 175 strain of hepatitis A virus was developed by SmithKline Beecham Biologicals (SB), Belgium, and proved to be safe and well tolerated inducing high titres of neutralising antibodies (Wiedermann et al., 1990).

The quantitative responses of hepatitis A virus (HAV) antibodies in volunteers receiving this vac- cine was assessed using a commercially available enzyme-linked immunosorbent assay for total an- tibodies to HAV (Hepanostika HAV Antibody, Organon Teknika, The Netherlands).

2. Materials and methods

2.1. Study population

One hundred and thirteen healthy adult volun- teers (50 males and 63 females) participated in a vaccine trial at the Royal Free Hospital School of Medicine, London. The mean age of the subjects was 20 years with a range of 19-60 years. Sub- jects were excluded if they were seropositive for hepatitis A antibodies by the screen test (HAVAB EIA, Abbott Laboratories, UK) used at the Med- ical School or had abnormal liver function tests.

2.2. Hepatitis A vaccine

The vaccine (Havrix) and placebo were manu- factured by SB Biologicals, Belgium. The vaccine was prepared using the HM 175 strain of hepatitis A virus cultivated by classical cell culture meth- ods, purified by ultrafiltration and gel permeation and inactivated with formaldehyde (Andre et al., 1990). The preparations were formulated to con- tain not less than 720 Elisa units of HAV antigen protein per 1 ml dose as defined by the manufac- turer, adsorbed onto 0.5 mg of aluminium hy- droxide and stored at 2-8’C. The control group of volunteers received a preparation containing equivalent components to the vaccine apart from the antigen.

2.3. Study design

(A) Blood samples for aspartate aminotrans- ferase (AST) and alanine aminotransferase (ALT) were taken within 2 weeks before entry to the study and at 0, 0.5, 1, 2, 12 and 13 months, respectively. A standard physical examination was also carried out on these occasions if indicated clinically. The subjects were allocated randomly to one of 4 groups; either vaccine or control preparation, with vaccination schedules day 0, month 1 and month 12 (regimen I) or day 0, day 14 and month 12 (regimen II).

(B) Quantitative assessment of anti-HAV total antibodies was carried out at months 1, 2, 12 and 13.

2.4. Serological methods

All volunteers were pre-screened for the pres- ence of antibodies to HAV using an enzyme im- munoassay (HAVAB EIA Abbott Laboratories, UK). The cut-off level of this qualitative test is in the range of 100-200 IU/l. Only volunteers below the cut-off level of this test were considered as seronegative and were enrolled in the study.

Serum samples collected at months 1, 2, 12 and 13 were tested for HAV total antibodies (Hep- anostika HAV Antibody, Organon Teknika, The Netherlands). Quantitation was carried out using the first Reference Preparation Hepatitis A Im- munoglobulin (WHO, 198 1; nominal value of 100 IU per ampoule), supplied by the Central Labora- tory of the Blood Transfusion Service, Amster- dam, The Netherlands. Antibody concentrations are expressed in International Units per litre (IV/

1). The Hepanostika HAV Antibody test (Organon

Teknika, The Netherlands) is an ELISA based on a modified ‘sandwich inhibition’ principle. The IgG fraction of human anti-HAV positive serum is coated on the solid-phase. The test sample or the appropriate controls are incubated simulta- neously with a fixed amount of formaldehyde- inactivated HAV antigen of faecal origin from chimpanzees. After a wash cycle and incubation with anti-HAV (human origin; F(ab), fragment) labelled with horse-radish peroxidase (HRP) as

J.N. Zuckerman et al, /Journal of Virological MethodF 56 (1996) 27-31 29

Composite Standard Line

I

l.WO-

SW-

m-

400-

200.

n IllNn# o=mnn I =1xao

I _” I so loo SW 1.ow Iun.

Fig. 1. Composite standard line (WHO 1st Ref. Prep. 1981 of Hepatitis A Immunoglobulin); each point of the standard curve measured in duplicate; 11 individual runs; mean f S.D; X-axis, log scale (dose); Y-axis, linear scale (absorbance); NC, negative control of test kit.

the conjugate, a wash cycle follows. Colour is developed with tetramethyl-benzidine (TMB/per- oxide substrate), and the reaction is stopped by the addition of sulphuric acid. Absorbances are read at 450 nm. The repeatability and reproduci- bility of this test was found to give a percentage coefficient of variation of 6% and less than 15%, respectively. The limit of detection of this test was found to be 50 IU/l when calculated by reading off the standard curve, the mean absorbance of 20 samples without antibodies to HAV, minutes 2 x S.D. (Fig. 1).

2.5. Biochemical tests of liver functions

ALT and AST were determined using test kits (Boehringer Mannheim, Germany). Both assays were optimised according to the recommendations of the International Federation of Clinical Chem- istry (IFCC), and were run on a Hitachi 717 analyser. The reference interval for both AST and ALT was 5-40 IU/l.

2.4. Analysis of results

Antibodies to hepatitis A were expressed as geometric mean titres (GMTs) and were calcu- lated at each point in time. The non-parametric Wilcoxon signed rank test was used to establish statistically significant differences amongst the means and within the groups. The Wilcoxon rank sum test was used comparing the means between groups.

3. Results

Immunisation with the hepatitis A vaccine in-

duced a seroconversion rate (concentrations > 50 IU/l) of 94% at one month in regimen I and of 97% in regimen II (Table 1). The quantitative responses (GMTs) over the months l-13 are shown in Fig. 2 and Fig. 3.

The GMTs of antibodies to HAV to regimen I were 340 IU/l at month 1 decaying to 188 IU/I at month 12 and increasing to 11300 IU/l at month 13, i.e. one month after administration of the booster dose. With regimen II, the GMTs were 470 IU/l at month 1 falling to 210 IU/l at month 12 and increasing to 5 130 IU/l 1 month following

Table 1 Seroconverion rates after immunisation with Hepatitis A vaccine (HAVRIX)

Month 1 Month 2 Month 12 Month 13

Regimen I (No. samples) Regimen II (No. samples)

94% (31) 100% (23) 79% (28) 100% (32)

97% (33) 100% (16) 88% (26) 100% (33)

Seroconversion: concentrations of antibodies to HAV 2 50 III/l.

J.N. Zuckerman et al. /Journal of Virological Methods 56 (1996) 27-31

RegImea I

I -lC

1

I “mcch

,-

Fig. 2. Geometric mean titres (GMTs; W/l) of vaccination schedule Regimen I in relation to time (month). 0 vaccine group; A control group; Y-axis, log scale.

the booster dose. There were no statistical differ- ences between GMTs using the different regimens.

Sharply increasing antibody concentrations were demonstrated in both groups following the administration of the booster dose. At month 13, all subjects had seroconverted (Table 1).

The GMTs of both control preparation regi- mens were found to be below 50 IU/l (Figs. 2 and 3). No response was demonstrated, i.e. below the limit of detection (50 IV/l) of the antibody test in 76% of the volunteers receiving the control prepa- ration. In 24%, however, measurable concentra- tions, ranging from 70 to 420 IU/l, were found, either occasionally among the volunteers or longi-

lO.cm-

‘_1 ----__I rm

AD-_______-_---_--_---_-_-______--

2 , .r

0 1 2 12 13 -

Fig. 3. Geometric mean titres (GMTs; IV/l) of vaccination schedule Regimen II in relation to time (month). 0 vaccine group; V control group; Y-axis, log scale.

tudinally over the sampling period of 13 months in a given volunteer.

Fifty male and 63 female subjects were included in this study. The possible differences in response between the sexes receiving the two different vac- cination regimens was assessed by the Wilcoxon rank sum test. This demonstrated a significantly lower GMT in male subjects at month 1 in regi- men II (P = 0.01) and slightly lower concentra- tions at month 12 in regimen I in female subjects (P = 0.06). The other GMTs of both regimens did not demonstrate sex differences.

4. Discussion

Protection against infection with hepatitis A antibody depends upon high standards of public health and hygiene and selective passive immuni- sation of those at high risk of infection using human immunoglobulin. The incidence of hepati- tis A antibodies in the general population has declined leading to a reduction in the antibody content of batches of immunoglobulin (Lemon and Stapleton, 1993). The scope for prophylaxis has changed with the introduction of hepatitis A vaccine preparations which offer longer and more effective protection.

The formalin-inactivated hepatitis A vaccine proved to be highly immunogenic with an overall seroconversion rate of 100%. Both vaccination groups mounted a measurable and high response. GMTs of 340-470 IU/l were obtained in both regimens at month 1, well above the presently accepted minimum protective level of lo-20 IU/l. GMTs at month 2 were not significantly different from those at month 1. However, there was a trend of decreasing antibody titres by 12 months which increased rapidly by a factor of lo-100 following the administration of the third dose.

The majority of sera from the control prepara- tion groups did not give a quantitative hepatitis A antibody response. However, in 24% of these volunteers, measurable responses ( > 50 IU/l) up to 500 IU/l were noted. This can be explained by the fact that these subjects may well have had a natural infection with hepatitis A with residual antibody concentrations too low for detection

J.N. Zuckerman et al. /Journal of Virological Methods 56 (1996) 27-31 31

using the screen test (the cut-off value being in the range of 100-200 IU/l).

It is noteworthy that in the literature on HAV vaccine trials, control or placebo groups are men- tioned only occasionally, and that published mea- surement data on those samples are difficult to find.

A gender difference in the response to hepatitis A vaccine was observed at month 1 (regimen I) in this study with a lower response occurring in male subjects. Antibody responses were found to be higher in males than in females in a previous study (Tilzey et al., 1992).

Other studies report findings similar to our observations that vaccination with an inactivated hepatitis A vaccine produces antibody concentra- tions at least 20 times higher than those induced by passive immunisation with human im- munoglobulin (Berger et al., 1993). The duration of protection conferred by the vaccine, and there- fore the need for a booster as well as the response to exposure to a natural hepatitis A infection after vaccination, have yet to be fully established (Wie- dermann et al., 1992).

It is concluded that the Hepanostika HAV An- tibody test can be used adequately and effectively for quantitative determination of antibody titres to hepatitis A after immunisation.

The quantitative results in the control prepara- tion group may be questioned, but we consider that pre-screening of subjects for vaccination should be carried out using a quantitative test with a cut-off level which reflects the presently accepted minimum protective level of lo-20 III/l, while retaining the required specificity.

Acknowledgements

This work was carried out following a compre- hensive collaborative study with SmithKline

Beecham Pharmaceuticals, Welwyn Garden City, UK, on the antigenicity and immunogenicity of the hepatitis A vaccine. We acknowledge with thanks the excellent collaboration with Dr. Ayse Baxter and Miss Helen Tyrrell.

Our thanks are due to the volunteers participat- ing in this vaccination study. The help of Mr. H. Rietbergen, Statistical Services Group, Organon Teknika, Boxtel, is gratefully acknowledged.

References

Andre, F.E., Hepburn, A. and D’Hondt, E. (1990) Inactivated candidate vaccines for hepatitis A. In: J.L. Melnick (Ed), Progress in Medical Virology, Karger, Basel, 37, pp. 72- 95.

Berger, R., Just, M. and Althaeus, B. (1993) Time course of hepatitis A antibody production after active, passive and active/passive immunization; the results are highly depen- dent on the antibody test used. J. Viral. Methods 43, 281-298.

Hofmann, F., Berthold, H. and Wehrle, G. (1992) Immunity to hepatitis A in hospital personnel. Eur. J. Clin. Micro- biol. Inf. Dis. 11, 1195.

Lemon, SM. and Stapleton, J.R. (1993) Section Hepatitis A virus. Ch. 5. Prevention. In: A.J. Zuckerman and H.C. Thomas (Eds), Viral Hepatitis; Scientific Basis and Clinical Management, Churchill Livingstone, Edinburgh, pp. 61- 79.

Provost, P.J. and Hilleman, M.R. (1979) Propagation of hu- man hepatitis A virus in cell culture in vitro. Proc. Sot. Exp. Biol. Med. 160, 213-221.

Tilzey, A.J., Palmer, S.J., Barrow, S., Perry, K.R., Tyrrell, H., Safary, A. and Banatvala, E. (1992) Clinical trial with inactivated hepatitis A vaccine and recommendations for its use. Br. Med. J. 304, 1272-1276.

Wiedermann, G., Ambrosch, F., Kollaritsch, H., Hofmann, H., Kunz, Ch., D’Hondt, E., Delem, A., Andre, F.E., Safary, A. and Stephenne, J. (1990) Safety and immuno- genicity of an inactivated hepatitis A vaccine candidate in healthy adult volunteers. Vaccine 8, 581-584.

Wiedermann, G., Ambrosch, F., Andre, F.E., D’Hondt, E., Delem, A. and Safary, A. (1992) Persistence of vaccine-in- duced antibodies to hepatitis A virus. Vaccine 10 (Suppl. I), 129-131.

Zuckerman, J.N., Cockroft, A. and Griffiths, P. (1991) Hepati- tis A vaccine. Br. Med. J. 303, 241.