Correlation between in vivo humoral and in vitro cellular immune responses following immunization with hepatitis B surface antigen (HBsAg) vaccines

Correlation between in vivo humeral and in vitro cellular immune responses following immunization with hepatitis B surface antigen (HBsAg) vaccines Geert Leroux-Roels *°~, Els Van Hecke*, Walter Michielsen t, Pierre Voet ~, Pierre Hauser * and Jean Patre * To study the regulation of the human immune response to hepatitis' B surface antigen (HBsAg) we have carefully monitored the in vivo humoral and in vitro cellular immune responses to HBsA9 in 50 subjects receivin9 .four doses of hepatitis B vaccine accordin 9 to a O, 1, 2, 12 month vaccination scheme. Twenty-three subjects were 9iven a plasma-derived vaccine (Hevac B) and 27 received a recombinant HBsAy vaccine (yeast-derived; Engerix-B). The humoral and cellular immune responses were measured before vaccination (day 0); 6 days after the second dose (day 36); 6 days (day 66), 2 months (day 120) and 10 months (day 365) after the third dose and 1 month after the jourth dose (day 395). Based on the kinetics of the humoral immune responses, the vaccinees could be classified into fast, intermediate and slow/non-responders. Based on the magnitude of the immune response (anti-HBs titre) on day 395, the vaccinees could be divided into high ( >~2000 U l t) and low ( <~2000 U l 1) responders. A close correlation between the kinetics and the magnitude ( f the humoral immune response was observed. The in vivo anti-HBs response was measured usin9 commercially available immunoradiometric assays. The in vitro cellular immune response was measured usin9 an HBsA,q-specific lymphoproliferation assay. Because of interassay variability the results were considered as dichotomous variables (proliferation versus non-proliferation)for further data analysis. A statistically significant correlation was observed between the kinetics and magnitude o[ the humoral immune response on the one hand and the in vitro anti-HBs response on the other hand. Although this study was not designed to compare the immunofeniciO' of a plasma-derived versus a recombinant vaccine, the data show that both vaccines are equally immunogenic. They induced similar seroconversion rates and 9eometric mean titres at all time points evaluated. We ,further demonstrate that female vaccine recipients display a swifter and more vigorous anti-HBs response that male vaccinees. Keywords"Hepatitis B surface antigen;

in vivo humoral response; in vitro cellular response

Considerable indirect evidence suggests that the immune response to hepatitis B virus (HBV) envelope antigens may be involved in the pathogenesis of hepatocellular injury in HBV infection 1'2. Clearly, the humoral immune response to the T-cell dependent hepatitis B surface antigen (HBsAg) is the paramount protective mechanism against reinfection in convalescent patients and against primary infection in HBsAg vaccine recipients. Evidence is accumulating that the variability of the humoral immune response to HBsAg in vaccine recipients is based on a genetic regulation of the immune response to HBsAg Departments of *Clinical Chemistry and +Internal Medicine, University of Gent, University Hospital, De Pintelaan 185, B-9000 Gent, Belgium. ++SmithKline Beecham Biologicals, Rixensart, Belgium. °CTo whom correspondence should be addressed. (Received 2 August 1993; revised 11 October 1993; accepted 25 October 1993) 0264-410x/94/09/0812-07 ~ 1994 Butterworth-HeinemannLtd 812

Vaccine 1994 Volume 12 Number 9

in humans 3 8, as has been demonstrated in mice 9'1°. Despite the opportunity provided by the widespread administration of HBsAg vaccines, knowledge of the regulation of the human immune response to HBsAg has not kept up with the progress made in the murine system. Apart from ethical and genetic (outbred population) constraints, the study of the human immune response to HBsAg has been hampered by the lack of an adequate in vitro system to monitor the HBsAg-specific T-cell response. No studies are available where the in vivo humoral and in vitro cellular immune responses to HBsAg of a large series of vaccinees have been monitored concomitantly in order to estimate the role of the T-cell response on anti-HBs production. We have developed an HBsAg-specific lymphoproliferative assay and have used it to monitor the evolution of the cellular immune response to HBsAg at several fixed time points during the course of a four-dose

Humoral and cellular immune response to HBsAg: G. Leroux-Roels et al.

vaccination schedule with either a plasma-derived or a recombinant vaccine in 50 volunteers. The strong correlation between the outcome of in vitro T-cell responses and the in vivo antibody production provides direct evidence for the hypothesis that the variability of the B-cell responses reflects the variability of the T-cell response. The latter may be based on genetic differences between the vaccinees.

MATERIALS AND METHODS

Subjects and immunization schedule Fifty health-care workers were given four doses of a commercial plasma-derived (Hevac-B, Pasteur, Paris, France; lot 86E06) or recombinant (Engerix-B, SmithKline Beecham Biochemicals, Rixensart, Belgium; lot ENG 137A4) hepatitis B vaccine in the deltoid muscle following a 0, 1, 2, 12 month schedule. All subjects were in good health and were seronegative for hepatitis B surface antigen (HBsAg), antibodies to HBsAg (anti-HBs) and antibodies to hepatitis B core antigen (anti-HBc) at the start of the study. They were rescreened for anti-HBc 1 month after the fourth vaccine dose. Approval for the study was received from the institutional ethics committee and informed written consent was obtained from all participants.

Assessment of hepatitis B virus serology and anti-HBs responses Blood samples for serological and cellular assays were collected from the antecubital vein on the following occasions: (i) before administration of the first vaccine dose to evaluate the prevaccination status; (2) 6 days after the second dose (day 36); (3) 6 days after the third dose (day 66); (4) 2 months after the third dose (day 120); (5) immediately before the administration of the fourth dose (day 365); and (6) 1 month after the fourth dose (day 395). HBsAg, anti-HBs and anti-HBc were measured using radioimmunoassaysfrom Abbott Laboratories (N. Chicago, IL). Anti-HBs titres were expressed in IU 1- ~ using the AUSAB Quantitation Panel (Abbott).

Lymphocyte proliferation assay Peripheral blood mononuclear cells (PBMCs) were prepared from heparinized peripheral venous blood by isopycnic density centrifugation on Ficoll Hypaque (Lymphoprep, Nyegaard, Denmark) following a standard protocol and suspended in RPMI 1640 medium supplemented with 25 mM HEPES, 2 mM L-glutamine, 50 U ml-1 penicillin and 50 #g ml- ~ streptomycin (all from Gibco Europe, Gent, Belgium), 5x10 -5 M 2mercapto-ethanol (Sigma, St Louis, MO) and 10% heat-inactivated pooled human AB serum. This human AB serum was obtained from healthy blood donors with blood group AB + and was only used when serological markers of HBV infection or vaccination were absent. This 'complemented' RPMI 1640 medium is termed 'complete medium'. During the assay, PBMC were either kept in complete medium only (unstimulated control cultures) or stimulated with varying concentrations of antigens for 5 days at 37°C in an atmosphere of 5% CO 2 in air, at which time 0.5 #Ci [3H]-thymidine was added to each microculture. After 16 to 20 h the cultures were harvested onto glass-fibre filters using a multichannel

cell harvester (PHD, Cambridge, MA) and incorporation of [3H]-thymidine was measured by liquid scintillation counting in an LKB-WalIac 8100 counter (LKB, Bromma, Sweden). Results are expressed as stimulation index (SI: mean counts min- 1 of antigen-stimulated cultures/mean counts min- 1 of control cultures) or as Acounts min- 1 (mean counts min-1 of antigen-stimulated culturesmean counts min -1 of control cultures) when more appropriate. Standard deviations of the mean counts min-x of triplicate cultures were consistently below 10%. The antigens used for lymphocyte stimulation were two recombinant yeast-derived HBV envelope preparations. The first contained only the small protein (HBsAg or S particles) and was identical to the recombinant HBsAg vaccine used. The second consisted of recombinant particles containing the pre-S2-S protein only (pre-S2-S particles) ix. Tetanus toxoid (TT), used as a control antigen in all assays, was obtained from Statens Seruminstitutet (WHO, Copenhagen, Denmark).

Statistical methods ~2 and Mann-Whitney U test methods were used for statistical comparison. RESULTS

Study population The subjects in the study were 50 healthy adult volunteers aged 21-51 (mean 30.3) years without evidence of medical illness, as determined by history taking and physical examination. Twenty-seven volunteers were given Engerix-B and 23 received Hevac-B. Both vaccine groups were comparable with regard to the age and sex of their members (Table 1).

Anti-HBs response of the vaccine groups The anti-HBs responses of the vaccinees were monitored on days 36, 66, 120, 365 and 395. The seroconversion rates (anti-HBs~> 10 IU 1-1) and the geometric mean titres (GMTs) of seroconverters in both groups did not differ significantly(Table 2). When the anti-HBs responses of female and male participants were compared (Table 3), a slower seroconversion of the male group and significantly (p < 0.05) lower titres immediately before (day 365) and 1 month after the fourth dose (day 395) were observed. Based on the kinetics of the immune response and selecting an anti-HBs titre of 10 IU 1-1 (seroconversion) as the threshold value, the vaccinees could be classified into three groups: fast responders (n=25) reached a anti-HBs titre~>10 IU 1-1 on day 36; intermediate Table 1

Demographic characteristics of the vaccine groups Vaccine group

Number of subjects M:F ratio Age (years) Women Range Mean __s.d. Men Range Mean ±s.d.

Hevac-B

Engerix-B

23 13:10

27 14:13

21-45 31.7 + 7.6

22-51 29.4_+ 10.5

22-39 30.1 --+4,8

23-43 30.1 __+5.6

Vaccine 1994 Volume 12 Number 9

813

H u m o r a l and cellular i m m u n e response to HBsAg: G. Leroux-Roels et al. Table 2 Seroconversion rate at a 10 IU I 1 threshold and geometric mean titres (GMTs) of anti-HBs in seroconverters in Hevac-B and Engerix-B vaccine groups

Day 36 a Seroconversion rate (SR) (%) Hevac-B Engerix-B

47.8 51.8

GMT (IU I 1) Hevac-B Engerix-B

71 76

68b

120c

82.6 81.5 93 230

95.7 92.6 258 452

365 d

395e

91.3 88.9

100 100

110 168

5497 2844

aSix days after dose 2 CSix days after dose 3 CTwo months after dose 3 dprior to dose 4 eOne month after dose 4 (booster)

Table 3 Seroconversion rates at a 10 IU I 1 threshold and geometric mean titres (GMTs) of anti-HBs in women and men

Day 36a Seroconversion rate (SR) (%) Women Men

66.7f 30.4

GMT(IUI Women Men

86 50

66b

120c

85.2 78.3

92.6 95.7

365d

85.2 95.7

395e

100 100

HBsAg-specific lymphoproliferation assay To study the regulation of the human immune response to HBsAg in general and the role of T lymphocytes in particular, we developed an HBsAg-specific lymphoproliferative assay. The definitive assay conditions described in the Materials and methods section were determined after preliminary experiments that explored the influence of many variables on the outcome of the assay. These variables were: the geometry of the culture wells, the qualities of the culture media and in particular the quality of the serum added, the numbers of PBMC added per microculture, the duration of the culture, etc. The factors determining the success of the procedure were the quality of the human AB + serum and the number and freshness of the PBMC employed. We were unable to obtain good assay results with non-human serum and with 'frozen-stored' cells. All lymphoproliferation assays were performed in triplicate with HBsAg and pre-S2-S particles at final concentrations of 3, I, 0.3 and 0.1 /~g ml- 1 and with TT at 20, 6 and 2 #g ml- 1. The antigen dose at which a maximal proliferative response was observed differed from vaccinee to vaccinee, and it often differed from assay to assay within one subject. To

Vaccine dose

1

2

3

4

1) 224 96

533 207

1931 97

l

10000

8769I 1466

1000

aSix days after dose 2 bSix days after dose 3 CTwo months after dose 3 aPrior to dose 4 eOne month after dose 4 (booster) tDifferences are statistically significant at p < 0.05

responders (n=16) reached the 10 IU 1-1 threshold on day 66; and slow/non-responders (n = 9) had not reached the 10 IU 1-1 level on day 66. The male:female ratios in these fast, intermediate and slow/non-responder groups were 7:18, 11:5 and 5:4, respectively, whereas the Engerix-B:Hevac-B ratios in these groups were 14:11, 8:8 and 5:4. These differences did not reach statistical significance. A graphical display of the evolution of the anti-HBs response (expressed as the GMT) of the three subgroups along the vaccination scheme (Figure 1) clearly demonstrates that the kinetics and the magnitude of the anti-HBs response are closely correlated. This correlation is further evidenced by a comparison of the kinetic data with the magnitude of the anti-HBs response measured 1 month after the fourth vaccine dose. Based on an arbitrarily chosen cut-off value of 2000 IU 1-1 reached 1 month after the fourth dose, two-thirds (32) of the subjects could be designated high responders, and those not reaching this titre were called low responders. The male:female ratios in the high- and low-responder categories were 9:23 and 14:4, respectively (p < 0.05). The Engerix-B:Hevac-B ratios in these groups were 16:16 and 11:7, respectively. Table 4 shows that 24 out of 25 fast responders and six out of 16 intermediate responders reached a titre of ~>2000 IU 1-1 . Two out of nine slow/non-responders had to be classified as high responders since, following a fourth vaccine dose, they produced anti-HBs titres of 11 100 and 14 100 IU 1-1.

814

V a c c i n e 1994 V o l u m e

12 N u m b e r

9

100 ..r. •- '

10

.

.

.

.

0

i

.

60

.

.

.

i

.

.

.

.

i

.

.

.

.

120

i , 360

DAY

Figure 1 Correlation between kinetics and magnitude of the humoral anti-HBs response. (E]) Fast responders; ( A ) intermediate responders and ( x ) slow/non-responders are defined based on the time point at which they reach the 10 IU I 1 threshold. The points in the graph represent the geometric mean titres (GMTs) of each subgroup reached at the different time points: days 0, 36, 66, 120, 365 and 395

Table 4 Comparison of the magnitude and kinetics of the in vivo anti-HBs response

Magnitude of the in vivo anti-HBs response measured on day 395

Kinetics of the in vivo anti-HBs response

Fast responder (/>101UI 1 on day 36) Intermediate responder (<10 IU 1-1 on day 36 and />101UI 1onday66) Slow/non-responder (<101UI l o n d a y 6 6 )

High responder (anti-HBs />2000 IU I 1)

Low responder (anti-HBs <2000 IU I 1)

24

1

6

10

2

7

Humoral and cellular immune response to HBsAg: G. Leroux-Roels et al.

14

14

a

b 12

12

lO

ISI =

10

ISI = 11.05

.

8

g

~

,

~

4

4-

2

2

0 0.1

'

'

•

I

0

1

0.1

"

"

=

= ' ' "

1

preS2 (~.g/ml)

H B s A g (lag/rnl)

Figure 2 Lymphoproliferation assays with HBsAg (a) and pre-S2-S particles (b) as stimulating antigens at concentrations of 0.1, 0.3, 1 and 3 #g ml-1. The integrated stimulation index (ISI) is the surface of the polygon defined by the Sis measured at each antigen concentration and the baseline at SI = 1

/ 1:1 12"" b .....

a 1 oooo

lOOO

m

-~

m |

6-

1oo 4'

m

1o

0

I

0

100

200

I

300

0

400

DAY

9P

0

"

I

100

|

200

"

I

300

"

400

DAY

Figure 3 Kineticsof the in vivo anti-HBs response (a) and the in vitro lymphoproliferative responses (b) in a selected subject. PBMC were stimulated with four concentrations of HBsAg (/k) or pre-S2-S particles ( x ) and the ISis were determined as shown in Figure 2

standardize the expression of our data we have chosen to express the assay results as 'integrated stimulation indices' (ISI). The ISI is the surface of the polygon defined by the stimulation indices reached at the different antigen concentrations of HBsAg and TT employed and by the baseline at the SI value of 1. Figure 2 provides a graphical example of this calculation. Experimental data not presented here have demonstrated that the cells proliferating in these assays were mainly T lymphocytes. Their proliferation was antigen-dose dependent and required the presence of functional antigen-presenting cells (monocytes).

Envelope-specific lymphoproliferative responses in vaccine recipients Once the definitive conditions were established, we employed the HBsAg-specific lymphoproliferative assay

to examine the cellular immune response of the 50 vaccinees each time their anti-HBs titre was measured. Figure 3 shows the quantitative and temporal relationship between the in vivo anti-HBs production and the outcome of the in vitro proliferative assays in one selected vaccine recipient. Fi#ure 3a shows the evolution of the anti-HBs response along the vaccination schedule, while Figure 3b shows the progression of the in vitro lymphoproliferative responses. Although the curves of humoral and cellular immune responses are not exactly overlapping, a correlation is discernible. Similar courses of serological and cellular immune responses were seen in 23 other subjects (data not shown). Six subjects did not display this prototypic course of events but had ISI >/2 in at least two assays performed during the course of this follow-up. Data not shown here revealed that the lymphoproliferative assays were highly susceptible to biological influences (inter- and intraindividual) and analytical

Vaccine 1994 V o l u m e 12 N u m b e r 9

815

H u m o r a l and cellular i m m u n e response to HBsAg: G. Leroux-Roels et al. Table 5 Occurrence of in vitro envelope-specific proliferative responses according to kinetics and magnitude of the in vivo anti-HBs responses Proliferative response a

No proliferative response a

24 5 1

1b 11° 8b

27 3

5c 15c

Kinetics Fast responder Intermediate responder SIow/non-responder

Magnitude High responder Low responder

"The proliferative response is considered significant if a subject displays an ISI ~>2 in at least two in vitro lymphoproliferation assays during the entire follow-up bThe difference between proliferation and non-proliferation is significant at the p<0.001 level CThe difference between proliferation and non-proliferation is significant at the p<0.01 level

variability (interassay). Unidentified qualities of the human AB serum and uncontrollable variations in the subjects' condition (common cold, medication, etc.) influenced the background proliferation in the unstimulated control cultures and hence the Sis of the different cultures, leading to varying outcomes of the assays. In view of this variability we have chosen to consider the assay results as a dichotomous variable (proliferative response versus no proliferative response) in further data analysis. Of the 30 subjects showing a consistent proliferative response 20 were women and ten were men. Seventeen had received Engerix-B and 13 had received Hevac-B. Table 5, displaying the relationship between in vitro lymphoproliferation and the kinetics and magnitude of the in rive antibody response to HBsAg, reveals a correlation between the kinetics and magnitude of the in vivo anti-HBs response and the in vitro lymphoproliferative response. Indeed, 24 out of 25 fast responders displayed a proliferative response, whereas only five out of 16 intermediate responders did so. The single slow/nonresponder that showed a proliferative response was one of the two subjects who turned out to be a high responder (11 100 IU 1-1). Table 5 also demonstrates that an in vitro proliferative response is more frequently observed in high responders than in low responders. DISCUSSION Using an in-house HBsAg-specific lymphoproliferative assay and a commercially available immunoassay we have monitored the cellular and humoral anti-HBs responses in 50 vaccine recipients. A clear correlation between these responses was observed. Cellular immune assays to detect HBsAg-specific T-cell responses have long been suboptimal for a variety of reasons, not least of which has been the necessity to rely upon highly variable clinical material to develop assay conditions a2-19. Despite the availability of highly purified plasma-derived or recombinant HBsAg of pharmaceutical quality, the progress in lymphocyte culture techniques and the advent and widespread use of an HBsAg vaccine that should allow the establishment of optimal parameters of an HBsAg-specific lymphoproliferative assay using PBMC from healthy vaccine recipients, the development of such an assay remained a difficult procedure.

816

V a c c i n e 1994 V o l u m e

12 N u m b e r

9

In the past decade several reports on m vitro lymphoproliferative responses to HBsAg have been published20 2~ Their results were conflicting and suggested that HBsAg-specific responses were difficult to detect in vitro. A few groups have been able to induce HBsAg-specific lymphoproliferation in cultures of fresh PBMC from some vaccine recipients and naturally infected subjects 2°'2'*'26"2"7.Others only observed HBsAgspecific proliferation when PBMC from vaccine recipients were stimulated with HBsAg in the presence of IL-222. In earlier studies we were unable to detect lymphoproliferation upon stimulation of fresh PBMC of vaccinees with HBsAg 1°. Antigen-specific proliferation could only be revealed after a second round of in vitro stimulation with HBsAg. Comparable results were obtained by Celis et al. 2° who found an HBsAg-specific response in only one out of ten vaccinees if fresh PBMC were used and in three out of these ten subjects when PBMC were first stimulated in vitro with HBsAg, maintained in IL-2 and rechallenged with HBsAg. This difficulty in inducing HBsAg-specific proliferative responses in cultures of fresh PBMC is most probably due to the low frequency of HBsAg-specific T cells in circulation. Recent studies using limiting dilution analysis of the T-cell responses to HBV envelope antigens show precursor frequencies of 1:35335 and 1:59311 in two vaccinees 28 and between 1:10 000 and 1:20 000 in one high-responding vaccinee (Van Hecke et al., unpublished results). These results may explain why proliferation was so difficult to detect in microcultures seeded at low cell densities ( ~< 105 cells per well). Speculating that the low precursor frequency might be the major cause of assay failure, we have chosen to stimulate 4 x 105 unfractionated PBMC per well. The serum used in the assays was another critical factor determining success or failure. Several sera of different origins were tested and only human serum (from A B + blood donors) gave satisfactory results. To reduce random variability we performed all assays at four antigen concentrations (3, 1, 0.3 and 0.1 #g ml-1). Since proliferation maxima were observed at different concentrations from one subject to another and from one time point to another within the same subject, we have chosen to take all assay results into consideration and to express the data as ISI. Despite great efforts to standardize most assay conditions and ingredients, an uncontrollable (most likely biological) random variability persisted, preventing exact quantitative comparisons of assay results. The results were, however, sufficiently clear and consistent to allow a classification of subjects into in vitro responder and non-responder categories. We have used this dichotomous classification to analyse the study results. To our knowledge this is the first hepatitis B vaccination study that simultaneously monitors the in vivo humoral and the in vitro cellular immune response in 50 subjects. The results demonstrate that such a project is feasible once a simple and reliable lymphoproliferative assay is available. This study also shows that the in vivo humoral and in vitro cellular immune responses to HBsAg are closely correlated. Indeed, 95% of the fast responders in vivo show clear HBsAg-specific proliferative responses, whereas most in vivo slow or non-responders do not. The in vitro proliferative responses to HBsAg correlate better with the kinetics than with the magnitude of the in vivo anti-HBs response. This observation, apart

H u m o r a l a n d c e l l u l a r i m m u n e r e s p o n s e to H B s A g : G, L e r o u x - R o e l s et al.

from the assay-related issues discussed before, may explain why some investigators considering the magnitude (titre) of the anti-HBs response rather than its kinetics have found so few in vitro lymphoproliferative responses in apparently high responder vaccinees 2°'2s. A correlation between in vivo and in vitro anti-HBs responses was also observed by Wismans et al. 29'3°. Using a spot-ELISA, a sensitive in vitro method that quantifies anti-HBs-producing B lymphocytes rather than proliferating T-cells, these investigators found a good correlation between the in vitro IgG anti-HBs spots and 'immune memory' in naturally infected subjects and vaccinees 29. Such a correlation between in vivo and in vitro anti-HBs production was also seen in a prospective vaccination study of insulin-dependent diabetic patients and matched control subjects 3°. In numerous vaccine recipients a clear lymphoproliferative response was observed as late as 10 months after the third vaccine dose. The Sis were generally lower than those of the preceding (6 days or 2 months after third vaccine dose) or following (1 month after fourth dose) assays. The timing of blood sampling after the last boost thus seems less critical than suggested by Degrassi et al. 22 who could induce in vitro lymphoproliferation only within a limited time window (approximately 1 week) after vaccine administration. Although it was not the prime goal of this study to compare the immunogenicity of the recombinant yeastderived vaccine Engerix-B with that of the plasma-derived vaccine Hevac-B, our data show that both products induce similar seroconversion rates and geometric mean titres at all time points examined. This confirms the observations of numerous controlled clinical studies 31. Our data also show that both vaccines induce comparable in vitro lymphoproliferative responses. PBMC from subjects immunized with the plasma-derived vaccine containing pre-S2 antigen did not display an enhanced in vitro proliferative response regardless of whether antigen preparations containing pre-S2 or HBsAg alone were used. Since Hevac-B contains HBsAg of both ad and ay subtypes 32 while the in vitro pre-S2-S particle was of the ad subtype, it is unlikely that the lack of pre-S2specific response is due to subtypic differences 33. In agreement with previous observations, women displayed significantly faster and higher immune responses to HBsAg than men (10<0.05) 22,34-36. We have clearly demonstrated that the varibility of humoral immune responses to HBsAg in vaccine recipients is reflected by a concurrent variability of in vivo cellular immune responses to this antigen. The intriguing questions now are why HBsAg is recognized less efficiently and induces less vigorous T-cell responses in some individuals and how this selective deficiency can be overcome. Our data suggest that the addition of pre-S2 sequences to the HBsAg vaccine will not solve the problem.

REFERENCES

10

11

12 13

14

15

16

17 18 19 20

ACKNOWLEDGEMENTS The authors are grateful to Mrs A.-M. Devreeze, G. Verheye, M. Scheiris and L. Laute for their excellent technical assistance. This work was supported in part by a grant from the National Research Foundation (Belgium). E. Van Hecke was supported by a grant from the IWONL-IRSIA (Belgium).

21

22

23

Milich, D.R. Immune response to hepatitis B virus proteins: relevance of the murine model. Semin. Liver Dis. 1991,11,93-112 Van Hecke, E., Paradijs, J., Molitor, C., Bestin, C., Pala, P., Slaoui, M. and Leroux-Roels, G. Hepatitis B virus (HBV)-specific cytotoxic T lymphocyte responses in patients with acute and chronic HBV infection. ,1. Hepatol. (in press) Walker, M., Szmuness, W., Stevens, C. and Rubinstein, P. Genetics of anti-HBs responsiveness. I. HLA-DR7 and non-responsiveness to hepatitis vaccination. Transfusion 1981, 21,601 Usonis, V., K~hnl, P., Brede, H. and Doerr, H. Humoral immune response after hepatitis B vaccination: kinetics of anti-HBs antibodies and demonstration of HLA antigens. Zentralbl. Bakterio/. Mikrobio/. Hyg. 1986, 262, 377-384 Craven, D., Awdeh, Z., Kunches, L., Yunis, E., Dienstag, J., Werner, B. eta/. Nonresponsiveness to hepatitis B vaccine in health care workers. Ann. Intern. Med. 1986, 105, 356-360 Alper, C., Kruskall, M., Marcus-Bagley, D., Craven, D., Katz, A., Brink, S. eta/. Genetic prediction of nonresponse to hepatitis B vaccine. N. Engl. J. Med. 1989, 321,708-712 Kruskall, M., Alper, C., Awdeh, Z., Yunis, E. and Marcus-Bagley, D. The immune response to hepatitis B vaccine in humans: inheritance patterns in families. J. Exp. Med. 1992, 175, 495-502 Watanabe, H., Okumura, M., Hirayama, K. and Sasazuki, T. HLA-Bw54-DR4-DRw53-DQw4 haplotype controls nonresponsivehess to hepatitis-B surface antigen via CD6-positive suppressor T cells. Tissue Antigens 1990, 36, 69-74 Milich, D.R. and Chisari, F.V. Genetic regulation of the immune response to hepatitis B surface antigen (HBsAg). I. H-2 restriction of the murine humoral immune response to the a and d determinants of HBsAg. J. Immunol. 1982, 126, 320-325 Milich, D.R., Leroux-Roels, G.G., Louie, R.E. and Chisari, F.V. Genetic regulation of the immune response to hepatitis B surface antigen (HBsAg). IV. Distinct H-2 linked Ir genes control antibody responses to different HBsAg determinants on the same molecule and map to the I-A and I-C subregions. J. Exp. Meal. 1984, 158, 41-56 Rutgers, T., Cabezon, T., Hartford, N., Vanderbrugge, D., AIscurieux, M., Van Opstal, O. et al. Expression of different forms of hepatitis B virus envelope proteins in yeast. In: Viral Hepatitis and Liver Disease (Ed. Zuckerman, A.J.) Alan R. Liss, New York, 1988, pp. 304-308 de Gast, G.C., Houwen, B. and Nieweg, H.O. Specific lymphocyte stimulation by purified, heat-inactivated hepatitis B antigen. BMJ 1973, Iv, 707-709 de Moura, M.C., Vernace, S.J. and Paronetto, F. Cell-mediated immune reactivity to hepatitis B surface antigen in liver diseases. Gastroenterology 1975, 60, 310-317 Lee, W.M., Reed, W.D., Mitchell, C.G., Woolf, I.L., Dymock, I.W., Eddleston, A.L.W.F. and Williams, R. Cell-mediated immunity to hepatitis B surface antigen in blood donors with persistent antigenaemia. Gut 1975, 16, 416-420 Ohta, G., Nonomura, A. and Nishimura, I. Leucocyte migration inhibition with inner and outer membranes of mitochondria and insoluble hepatocyte surface membranes prepared from rat liver in patients with chronic hepatitis and cirrhosis. C/in. Exp. Immunol. 1976, 26, 491--504 1-iku, M.L., Beutner, K.R., Tiku, K. and Ogra, P.L. Cell-mediated immune response to liver tissue antigen and hepatitis B surface antigen after infection with hepatitis B virus in humans. J. Infect. Dis. 1978, 138, 587-596 Tong, MJ., Wallace, A.M., Peters, R.L. and Reynolds, T.B. Lymphocyte stimulation in hepatitis B infections. N. Engl. J. Med. 1975, 263, 318-322 Yeung Laiwah, A.A.C. Lymphocyte transformation by Australia antigen. Lancet 1971, II, 470-471 Yeung Laiwah, A.A.C., Chaudhuri, A.K.R. and Anderson, J.R. Lymphocyte transformation and leukocyte migration-inhibition by Australia antigen. C/in. Exp. Immunol. 1973, 15, 27-34 Celis, E., Kung, P. and Chang, T.W. Hepatitis B virus-reactive human T lymphocyte clones: antigen specificity and helper function for antibody synthesis. J. Immunol. 1984, 132, 1511-1516 Cupps, T.R., Gerin, J.L., Purcell, R.H., Goldsmith, P.K. and Fauci, A.S. In vitro antigen-induced antibody responses to hepatitis B surface antigen in man. Kinetic and cellular requirements. J. Clin. Invest. 1984, 74, 1204-1213 Degrassi, A., Mariani, E., Honorah, M.C., Roda, P., Miniero, R., Capelli, M. and Facchini, A. Cellular response and anti-HBs synthesis in vitro after vaccination with yeast-derived recombinant hepatitis vaccine. Vaccine 1992, 10, 617-622 Femam, A., Cayzer, C.J.R. and Cooksley, W.G.E. HBsAg-induced

V a c c i n e 1994 V o l u m e 12 N u m b e r 9

817

H u m o r a l a n d cellular i m m u n e r e s p o n s e to HBsAg: G. Leroux-Roels et al.

24

25

26

27

28

29

30

antigen-specific T and B lymphocyte responses in chronic hepatitis B virus carriers as immune individuals. Clin. Exp. Immunol. 1989, 76, 222-226 Hellstr6m, U., Sylvan, S. and Lundbergh, P. Regulatory functions of T- and accessory-cells for hepatitis B surface antigen induced specific antibody production and proliferation of human peripheral blood lymphocytes, in vitro. J. Clin. Lab. Immunol. 1985,18, 173-181 Jin, Y., Shih, W.K., Alter, H.J. and Berkower, I. An oligoclonal human T-cell line specific for HBsAg is restricted to MHC Class I antigens and responds to soluble antigen and to endogenous HBsAg. In: Viral Hepatitis and Liver Disease (Ed. Zuckerman, A.J.) Alan R. Liss, New York, 1988, pp. 678--683 Steward, M.W., Sisley, B.M., Stanley, C., Brown, S.E. and Howard, C.R. Immunity to hepatitis B. Analysis of antibody and cellular responses in recipients of a plasma-derived vaccine using synthetic peptides mimicking S and pre-S regions. Clin. Exp. Immunol. 1988, 71, 19-25 Sylvan, S.P.E., Hellstr~m, U.B. and Lundbergh, P.R. Detection of cellular and humoral immunity to hepatitis B surface antigen (HBsAg) in asymptomatic HBsAg carriers. C/in. Exp./mmuno/. 1985, 62, 288-295 Ferrari, C., Penna, A., Bertoletti, A., Cavalli, A., Valli, A., Schianchi, C. and Fiaccadori, F. The preS1 antigen of hepatitis B virus is highly immunogenic at the T cell level in man. J. C/in. /nvest. 1989, 84, 1314-1319 Wismans, P, van Hattum, J., de Gast, (3., Endeman, H., Poel, J., Stolk, B. eta/. The spot-ELISA: a sensitive in vitro method to study the immune response to hepatitis B surface antigen. C/in. Exp. /mmuno/. 1989, 78, 75--78 Wismans, P., van Hattum, J., de Gast, G., Bouter, K., Diepersloot,

818

Vaccine 1994 V o l u m e 12 N u m b e r 9

R., Maikoe, T. and Mudde, G. A prospective study of in vitro anti-HBs producing B cells (spot-ELISA) following primary and supplementary vaccination with recombinant hepatitis B vaccine in insulindependent diabetic patients and matched controls. J. Med. Virol. 1991, 35, 216-222 31 Stephenne, J. Recombinant versus plasma-derived hepatitis B vaccines: issues of safety, immunogenicity and cost-effectiveness. Vaccine 1988, 6, 299-303 32 Adamowicz, P., Chabanier, G., Hyafil, F., Lucas, G., Prunet, P., Reculard, P. and Vinas, R. Elimination of serum proteins and potential virus contaminants during hepatitis B vaccine preparation. Vaccine 1984, 2, 209-214 33 Milich, D.R., Hughes, J.L., McLachlan, A., Langley, K.E, Thornton, G.B. and Jones, J.E. Importance of subtype in the immune response to the pre-S(2)region of the hepatitis B surface antigen. I. T cell fine specificity. J. Immunol. 1990, 144, 3535-3543 34 Andre, F.E. Clinical results with a hepatitis B vaccine produced by recombinant DNA technology. In: Engerix-B, a New Hepatitis B Vaccine Produced in Yeast. Asian Pacific Association for the Study of the Liver, SmithKline Beecham Biologicals Satellite Symposium, Hong Kong, 9 January 1986, pp. 11-18 35 Hammond, G.W., Parker, J., Mimms, L., Tate, R., SeMa, L. and Minuk, G. Comparison of immunogenicity of two yeast-derived recombinant hepatitis B vaccines. Vaccine 1991, 9, 97-100 36 Minuk, G.Y., Bohme, C.E. and Bowen, T.J. Passive and active immunization against hepatitis B virus infection: optimal scheduling in healthy young adults. Clin. Invest. Med. 1989, 12, 175-180