Potentials and limitations of protein vaccines in infants

PII: SO264-410X(98)00105-4

Potentials and limitations vaccines in infants Jean-Louis

Vaccine, Vol. 16, No. 14/15. pp. 1439-1443, 1998 0 1998 Elsevier Science Ltd. All rights reserved Printed in-Great Britain 0264-410X/98 $19+0.00

of protein

Excler

Immunization of infants represents a difJicult challenge since immune responses in infants differ qualitatively from those of adults (immature immune and bias towards Th2 rather than a Thl immune responses). while protein vaccines are immunogenic in infants earlier than polysacchatides in the very first months of life, requiring repeated administration. Protein vaccines such as diphtheria and tetanus toxoids and hepatitis B antigens are safe and immunogenic in infants, able to induce long-lasting memory, and effective. HB vaccines are the first vaccines administered at birth against a sexually transmitted disease and that prevent cancel: These success stories must however be tempered by the existence of immunologic interferences that may occur when combining vaccines in a single injection. Moreover; very little is known whether inappropriate protein conformation and glycosylation play roles in the impairment of the functional immunogenicity of protein vaccines in infants. Although much is known about how to produce and purifi proteins, we know much less about how urotein characteristics are related to $.m$onal immunogenic@. 0 1998 Elsevier Scie&e Ltd. All rights reserved. Keywords: vaccines;

proteins; infants; hepatitis B; immunization; glycosylation

Infant immunization represents the best, if not the only, way to achieve wide vaccine coverage, and is required to protect against pathogens to which infants are exposed during the first months of life. However, immunization of infants represents a difficult challenge because exposure to pa.thogens increases much faster after birth than the rate of development of defence mechanisms, resulting in an age-dependent risk of severe infectious diseases in infancy. In murine models, the neonatal immune responses differ qualitatively from those of adults in having vaccine-specific antibodies of low IgG2,a/IgGl ratio, the secretion of significantly higher IL-5 and lower INFy levels by vaccine-specific T cells, and, finally, an impaired induction of cytotoxic T cell precursors. There is therefore a neonatal bias towards a Th2 rather than Thl response. This bias might be beneficial when the goal of a protein vaccine is to induce protective antibodies, rather than a cell-mediated immune response’. The deficiencies in the immature immune response account for the susceptibility of neonates to infections with encapsulated bacteria and intracellular pathogens. The immune response typically seen in neonates includes: reduced T cell-independent B cell activation by polysaccharides and reduced maturation into plasma cells secreting sugar-specific antibodies; poor antibodydependent opsonization and phagocytosis of bacteria; limited B cell responses to polysaccharides (up to 8-24 months of age); T cell responses with a different Pasteur Mdrieux Connaught, Marnes-la-Coquette, France

3 Avenue

Pasteur,

92430

cytokine profile from that seen in adults, such as low levels of TNFa, INFy and other Thl-dominant cytokines; a decrease in inflammatory reactions and macrophage activation; and slow maturation of highavidity antibodies. The susceptibility of infants to infections with intracellular agents could be explained by a decreased function of NK cells, a reduction in IL-12-dependent induction of Thl-like CD4+ T cells, and a reduction in INFy, impairing macrophage killing and CDS+ CTL activation’. While protein antigens are immunogenic in infants earlier than polysaccharides, proteins still may be poorly immunogenic in the very first months of life, requiring repeated administration. Among other theoretical limitations, infants may respond poorly to a heavily glycosylated protein, whether bacterial or viral.

HOW DO THESE CLASSICAL ASSUMPTIONS APPLY TO THE CURRENTLY USED PROTEIN VACCINES IN INFANTS? The protein vaccines currently recommended for children in the United States and Europe include the well-known diphtheria and tetanus toxoids, as well as the acellular pertussis and hepatitis B vaccines (whether plasma-, yeast- or mammalian-cell-derived) (Table I), Furthermore, HIV-l envelope glycoproteins (gp120 derived from the HIV-l strains MN and SF2) have been recently administered to infants born to HIV-infected mothers. Very preliminary results seem to indicate that gp120 is safe and immunogenic in

Vaccine

1998 Volume

16 Number

14/15

1439

Potentials and limitations of protein vaccines in infants: J.-L. Excler Table 1

Protein vaccines currently used in infants

Diphtheria toxoid Tetanus toxoid Acellular pertussis:

pertussis toxoid, filamentous hemagglutinin, agglutinogens, pertactin (69 kD), plasma-derived recombinant proteins (S. pre-Sl , and preS2) yeast-derived mammalian cell-derived

Hepatitis B:

infants, and is able to induce both humoral and cellmediated immune responses (unpublished). Diphtheria and tetanus toxoid vaccines are wellknown antigens with proven efficacy that have been used for decades in infants. Although these protein vaccines will not be discussed in this presentation, we must recall that they are poorly immunogenic if given at birth and so must be administered later in infancy (6 weeks)‘. For this reason, the eradication of neonatal tetanus will only be possible by expanded immunization of women, especially mothers, in order to ensure protection of the newborn through passive transfer of tetanus antibodies.

POTENTIALS: THE UNPRECEDENTED ‘SUCCESS STORY’ OF HEPATITIS B VACCINES There is probably no better example of a success story of an infant protein vaccine than that of the hepatitis B (HB) vaccines, all adjuvanted in alum. The first HB vaccines used in infants were plasma-derived, and their development was largely empirical, except that the immunogenicity of this protein vaccine was known to be related to its presentation in a pseudo-particular form. Since then, a new generation of recombinant HB vaccines has been developed, derived either from yeast or mammalian cells, that include either the S protein alone or both S and pre-S2 proteins, with varying degrees of glycosylation. For example, the yeastderived vaccines contain the S protein as a 24 kD nonglycosylated polypeptide, whereas the Chinese hamster ovary (CHO) cell-derived vaccine is a mixture of 36 kD diglycosylated S+pre-S2 polypeptides, and 24 kD nonglycosylated and 27 kD glycosylated (ratio 4:l) S polypeptides3. Table 2 shows that three injections of HB vaccine are needed for optimal immunization, with the first

Table 2 Immunization schedules recommended by the American Academy of Pediatrics and the World Health Organization for administration of hepatitis B vaccine to infants born to HBsAgseronegative women or by the Expanded Program of Immunization for infants in developing countries Schedule Schedule Dose 1 Dose 2 Dose 3

Age I l-2 months 4 months 6- 18 months

Schedule 2a Dose 1 Dose 2 Dose 3

Birth (before discharge from hospital) l-2 months (4-l 2 weeks post-dose 1) 6-18 months (4-12 weeks post-dose 2)

“Expanded Program of Immunization in countries with moderate to high incidence of infection (HBs antigen carrier rate > 2%).

one given at birth, and then the second and third injections given between 4 and 12 weeks, and between 4 and 12 months of age, respectively. This great flexibility in the immunization schedule, due to the strong immunogenic@ of the S and pre-S2 proteins in infants, is illustrated in Table 3. Infants from the Australes Archipelago, French Polynesia, have benefited from an intervention program of systematic HB immunizations with a CHO-cell-derived HB vaccine expressing both S and pre-S2 proteins. A single injection was able to induce protective levels of anti-HB S protein antibodies in 60% of vaccinated infants (compared to about l-20% in adults who received the same vaccine and dosage). Several schedules with different intervals between infant vaccinations showed equivalent efficiency in inducing antibodies to HB S and pre-S2 proteins. This intervention program was conducted in field conditions, with great variations in the intervals between immunizations and related blood samplings, giving even more robustness to these results4. Table 4 displays some examples of results of efficacy trials on various HB vaccines conducted with infants from infected mothers in developed and developing countries. A high efficacy rate is constantly demonstrated either with or without the concomitant administration of hepatitis B immunoglobulin (HBIG) (mean protective rate of efficacy is 90% with HBIG and 75% without HBIG; the study conducted in Bangkok and mentioned in Table 4 showed an exceptionally high efficacy rate), whatever the immunization schedule and the type of vaccine (plasma-derived or recombinant) used’. In developed countries, the current recommend-

Table 3 Seroprotection rates and geometric mean titres (GMT) of anti-HBs and anti-pre-S2 antibodies GenHevac B (Australes archipelago, French Polynesia)4

%

(n/N)

0 0. 0, 0, 0, 0,

60 80 95 92 93 99

(27/45) (167/210) (179/l 89) (283/307) (84/90) (1931195)

1 I,2 1, 6 1, 12 1,2, 12

1440

with

Seroprotection rates for both anti-HBs and anti-pre-S2 (240 mU ml-‘)

Seroprotection rates (anti-HBs 2 10 mlU ml-‘) Immunization schedule (months)

among infants vaccinated

GMT HBs (mlU ml-‘)

GMT preS2 (mu ml-‘)

%

19 68 217 389 344 1228

102 253 280 252 257 350

88 97 100 96 97 100

Vaccine 1998 Volume 16 Number 14/15

(n/N) (22/25) (1821187) (180/l 80) (279/291) (64/66) (187f187)

Interval (days) between last dose and sampling 34 133 156 171 252 206

Potentials and limitations Table 4

Efficacy of hepatitis 13immunoglobulin

(HBIG) and recombinant

of protein vaccines in infants: J.-L. Excler

hepatitis B vaccine in infants of HBsAg/HBeAg-positive

mothers

Study site

Vaccine used

Antigen content

HBIG

Number of subjects

HBIG

HB vaccine schedule (months)

HBsAg carrier rate (%)

Protective efficacy (%)

United States

plasma (MSD) recombinant (MSD) plasma (MSD) recombinant (MSD) recombinant (SK)

10 119 5 pg 10 ,ig 5 /tg lO/(g

Yes Yes Yes Yes No

39 83 20 20 55

+ + + + _

0, 0, 0, 0, 0,

10.2 4.8 5 10 3.6

ND > 90 (estimated) 94.6 89.2 95

Bangkok Bangkok

MSD: Merck Sharp & Dohme; SK: SmithKline

Beecham Biologicals;

+: HBIG given at birth.

ation is to immunize all infants from HBV infected mothers at birth with HB’IG and HB vaccine. VACCINE-INDUCED IMMUNOLOGIC MEMORY FOR HEPATITIS B SURFACE ANTIGEN HB vaccines induce active synthesis of antibodies to the hepatitis B virus surface antigen (HBsAg) accompanied by immunologic memory for HBsAg. A persistent memory has been demonstrated over several years as well as a rapid increase in HBsAg antibody titres post-booster (reviewed in ref. 5). For example, long-term follow-up for HBsAg antibody persistence has been documented in four different studies performed with plasma-derived vaccines in three countries, the Gambia, Senegal and Venezuela’-” (Table 5). These data are supported by the observation that the quantity of memory B cells that can produce HBsAg antibodies (ELISPOT) does not diminish as the level of serum antibody declines’, and that an effective humoral immune response is still present over 5 years later in vaccinees frequently exposed to HBV. Overall, there is a good retention of immunologic memory in healthy vaccinees over periods of 5-12 years. Findings such as these led to the recommendation that it is unnecessary to give booster vaccinations even when HBsAg antibody titres fall below the ‘seropositive’ threshold of 10 mIU ml ’ 5. HB MASS VACCINATION DECREASES CHILDHOOD HEPATO-CELLULAR CARCINOMA HB mass vaccination in Taiwan was followed by a decrease in the annual incidence of childhood hepatoTable 5

Long-term

persistence

Time after immunization

of anti-HBsAg

antibodies

cellular carcinoma: in 1981-1991 there were 4.48-7.13 cases/100000 children (6-14 years) and by 1992 the incidence had fallen to 1.74 cases/100000 children (6-14 years). In this regard, the HB vaccines are the first vaccines administered that prevent cancer3.“‘. LIMITATIONS OF PROTEIN VACCINES IN INFANTS The simultaneous use of various protein vaccines administered at different injection sites generally does not impair the immune responses to the various antigens. This is well demonstrated with the classical diphtheria-tetanus-pertussis (whole cell or acellular) combination vaccine administered either alone or simultaneously with a recombinant HB vaccine. However, when all antigens are combined in a single vaccine, in some cases, interference might occur. For example, after administration of a combined vaccine containing HBsAg and diphtheria and tetanus toxoids (RDT), the corresponding geometric mean titre values for HBsAg antibodies were significantly lower in infants immunized with the combination vaccine; however, the percentage of infants having HBsAg antibody titres above the protective threshold was not different from that obtained when the hepatitis B and diphtheria-tetanus vaccines were administered simultaneously at different sites (R+DT) (Table 6)l’. Mass HB immunization programs may exert a selective immunologic pressure leading to the emergence of HBV variants in immunized populations. For example, in the Gambia, a lysine to glutamic acid mutation at amino acid (aa) 141 was present in 8% of immunized children. In Singapore, variants (aa 145, 129, 126, 133, 140) are detected in almost 15% of the virus populations. However, to ascertain the role of immunization in the generation of escape mutants, the

in immunized

infants

Vaccine

(year)

Country

Type of vaccine

Content

Schedule

1

Gambia Gambia Senegal Venezuela Gambia Senegal Gambia Senegal Gambia Senegal Senegal Senegal

plasma plasma plasma plasma plasma plasma plasma plasma plasma plasma plasma plasma

20 pg 10 1’9 5 ,1g 101(g 10 i19 5 ,cg t0pg 5 Fg 20 1’9 5 /1g 5 119 5 !Jg

0, 0, 0, 0,

2

1, 6 1,6 1, 6 1, 6 1, 2, 12

(MSD) (MSD) (PM) (MSD) (MSD) (PM) (MSD) (PM) (MSD) (PM) (PM) (PM)

2, 4 2, 4, 1,2, 1, 6 0, 2, 4, 0, 1,2, 0, 2, 4, 0, 1,2, 0, 2, 4 0, 1.2, 0, 1,2, 0, 1,2,

(months)

9 12 9 12 9 12 12 12 12

Seroconversion 100 98 99 99 95 97 95 96 95 100 93 90

(%)

GMT (mlU ml-‘) 926 1920 1458 767 524 2143 376 2134 75 1525 826 60

MSD: Merck Sharp & Dohme; PM: Pasteur Merieux.

Vaccine 1998 Volume 16 Number 14/15

1441

Potentials and limitations

of protein vaccines in infants: J.-L. Excler

Table 6 Antibody responses to hepatitis 6 surface antigen (HBsAg) (GMT and seroprotection immunization and one month after third immunization according to the mode of administration Administration

mode

rates) two months after the second of the vaccines (RDT vs R+DT) RDT

R+DT Post dose 2

Post dose 3

Post dose 2

Post dose 3

69 321.5 [233-4441 98.6 91.3 [82.0-96.71

63 3068” [2291-41071 98.4 98.4 [91.5-99.91

203 103.7 [87.3-123.31 95.6 87.7 [83.2-92.21

187 1309” [1079-15381 100 98.4 [95.3-1001

HBsAg (RIA-mlU ml-‘) ~MT [95% Cl] %a2 mlU ml-’ %>lO mlU ml-’ [95% Cl]

R: HBs antigen; D: Diphtheria toxoid; T: Tetanus toxoid; RDT: combination vaccine. “Inequivalence of the two modes of administration at post dose 3 cannot be rejected.

percentage of virus mutants must be compared in immunized versus non immunized subjects (i.e. adults vs infants of infected mothers). Recent evidence suggests that is more likely to be concomitantly administered HBIG, rather than the HB vaccine itself, that is responsible for the selection of mutants (i.e. selection of mutants not neutralized by HBIG)“. These findings might be of concern for the future and should be carefully monitored. One could well imagine the scenario where after years, these escape mutants could spread among a population that had become immune to the vaccine virus type, causing a renewed increase in HBV infection and disease”.

IS GLYCOSYLATION IMPORTANT FOR HB VACCINES IN INFANTS? The lack of mammalian patterns of protein glycosylation of the antigens in yeast-derived vaccines (YDV) does not alter any antigenic or immunogenic property when compared to the licensed plasma-derived vaccines (PDV). The negligible importance of correct glycosylation for immunogenicity of HB vaccines is further indicated by the observation that the non-glycosylated YDVs are no less immunogenic than a glycosylated mammalian cell-derived vaccine (GenHevac BTM)3. The nature of the glycosylation, however, may be important. In the pre-S2+ S vaccine expressed in yeast, some of the glycans are hyperglycosylated at variable lengths of up to 50 kD or more. This vaccine is poorly immunogenic when compared to the classic YDVs. The same vaccine without hyperglycosylation was as immunogenic as classical YDVs in adults (unpublished). Glycosylation of the pre-S2 region does not appear to be an issue for the CHO-cell-derived vaccine, since all infants immunized with GenHevac BTM develop early and strong pre-S2 antibody responses4. However, while these findings are suggestive, they do not help to predict the immune response of new proteins, whether glycosylated or not, when administered to infants. Glycosylation may be required for proper antigen conformation, and so the induction of adequate neutralizing antibodies, but may also be deleterious to immunogenicity, by limiting antibody access to the protein by steric hindrance. Both of these processes are antigen-dependent. This paradox is not specific to infant immunogenicity, which is similar to that of adults in this respect, and is’to be distinguished

1442

Vaccine 1998 Volume 16 Number 14/15

from the poor capacity of infants to respond to polysaccharides. This has important implications, for example, in the design of antigens suitable for inclusion in HIV-l preventive or immunotherapeutic vaccines. Recently, indications about the role of carbohydrates in the immune response to HIV have been reported. Rhesus monkeys were infected with mutant forms of simian immunodeficiency virus (SIV) lacking sites for N-linked glycosylation in the external envelope of the virus. Sera from monkeys infected with the mutant viruses exhibited markedly increased neutralizing activity compared to sera from monkeys infected with the parental virus, suggesting a role for N-linked glycosylation of the external envelope in limiting the neutralizing antibody response to SW and in shielding the virus from immune recognition13. Likewise, mutations in gp120 V2-C2 glycosylation sites were shown to enhance the neutralizing antibody response in HIV-l infected patients14. How these findings apply to immune recognition of HIV-l recombinant gp120 (where carbohydrate represents 50% of the total mass of gp120) in infants remain to be explored.

CONCLUSION Protein vaccines currently administered in infants are safe, immunogenic (able to induce a long-lasting immune memory) and effective. The best example is provided by the hepatitis B vaccine, the first vaccine to be administered at birth that protects against a sexually transmitted disease and a cancer. However, this success story must be tempered by the existence of immunologic interferences that may occur when combining vaccines in a single injection (such as HBsAg antigen with diphtheria and tetanus toxoids). The emergence of HBV escape mutants should in any case delay the implementation of the worldwide policy of mass immunization against hepatitis B. However, viral surveillance systems in populations benefiting from mass vaccination campaigns might be helpful for the detection of escape mutants. In this regard, more attention should be paid to the design of future HB vaccines, should ever the spread of these escape mutants become significant. Very little is known whether inappropriate protein conformation and glycosylation play roles in the impairment of the functional immunogenicity of protein vaccines in infants. This might merit further investigation to help better predict the immune response generated with a given protein. Although

Potentials and limitations of protein vaccines in infants: J.-L. Excler much is known about how to produce and purify proteins, we know much less about how protein characteristics are related to functional immunogenicity.

9

REFERENCES Siegriest, C.A. Potential advantages and risks of nucleic acid vaccines for infant immunization. Vaccine 1997, 15, 798800 Plotkin, S.A. and Mortimer, E.A. (Eds) Vaccines; 2nd edn. Saunders, Philadelphia, I994 Ellis, R.W. (Ed.) Hepatitis 6 Vaccines in Clinical Practice. Marcel Dekker, New York, 1993 Mouliat-Pelat, J.P., Spiegel, A. and Excler, J.L. et al. Lutte contre I’hepatite I3 par un programme de vaccination systematique des nouveau-&s avec le vaccin GenHevac 6. Cahiers Sant6 1996, 6,11-15 West, D.J. and Calandw, G.B. Vaccine induced immunologic memory for hepatitis B surface antigen: implications for policy on booster vaccination. Vaccine 1996, 14, t 019-l 027 Whittle. H.C.. Inskia H.. Hall. A.J.. Mendv. M., Downes. R. and Hoare, S. Vaccinatiun against hepatitis B and protection against chronic viral carriage in the Gambia. Lancer 1991, 337,747-750 Hadler, S.C., de Monzon, M.A., Lugo, DR. and Perez, M. Effect of timing of hepatitis B vaccine doses on response to vaccine in Ycpa indians. Vaccine 1989, 7, 106-l 10 Coursaget, P., Yvonnet, B., Chotard, J., Sarr, M., Samb, A., N’Doye, I?., Diop-Mar, I. and Chiron, J.P. Long-term efficacy of hepatitis B vaccine in infants from an endemic area. In:

10

11

12

13

14

Progress in Hepatitis B Immunization (Coursaget, P. and Tong, M.J.). Colloque INSERM/John Libbey Eurotext, Paris, 1990, pp. 287-300 van Hattum, J., Maikoe, T., Poel, J. and deGast, G.C. In vitro anti-HBs production by individual B cells of responders to hepatitis B vaccine who subsequently lost antibody. In: Viral Hepatitis and Liver Disease (Eds Hollinger, F.B., Lemon, SM. and Margolis, H.S.). Williams and Wilkins, Baltimore, 1991, pp. 774-776 Chang, M.H., Chen, C.J. and Lai, MS. et al. For the Taiwan childhood hepatoma study group. N. Engf. J. Med. 1997, 336, 1855-l 859 Zanetti, A., Capasso, D.A., Cordone, A. et a/. Studio multicentrico, randomizzato, controllato di fase II sulla sicurezza e I’immunogenicita di un vaccine combinato contra I’epatite B, la difterite ed il tetano (RDT) somministrato a lattanti. 37th Congress0 Nazionale L’lgiene e la Sanita Pubblica al/e soglie de/ 2000, Napoli, 25-28 September 1996 He, J.W., Lu, Q., Zhu, Q.R., Duan, S.C. and Wen, Y.M. Mutations in the ‘a’ determinant of hepatitis B surface antigen among Chinese infants receiving active postexposure hepatitis B immunization. Vaccine 1998, 16, 170-l 73 Pantaleo, G. Mechanisms of virus escape from the immune response during human immunodeficiency virus (HIV) infection. Xlth Cent Gardes meeting on Retroviruses of Human AIDS and Related Animal Diseases, Marnes-la-Coquette, France, 27-29 October 1997 Desrosiers, R.C. A role for carbohydrates in immune evasion in AIDS. X/t/t Cent Gardes meeting on Retroviruses of Human AIDS and Related Animal Diseases, Marnes-la-Coquette, France, 27-29 October 1997

Vaccine 1998 Volume 16 Number 14/l 5

1443