New versus old meningococcal Group B vaccines: How the new ones may benefit infants & toddlers

Epidemiology of meningococcal disease

One of the defining characteristics of invasive meningococcal disease is substantial cyclical fluctuation in its epidemiology. The high variability of the disease is in accordance with its geographical and serogroup distribution13. The traditional approach to classifying Nm is based on serological typing into at least thirteen groups [A, B, C, E-29, H, I, K, L, W-135, X, Y, Z and Z’ (29E)] with distinct features in terms of immunological reactivity and structure of the capsular polysaccharide; however, only serogroups A, B, C, W-135, X and Y can cause life-threatening disease14.

The world's highest incidence of meningococcal disease occurs in the Sahel and sub-Sahel African regions (the so-called “meningitidis belt”); in this biogeographic zone a large number of epidemics have been caused by serougroup A, accounting for nearly 200,000 cases in 199615. To contain this dramatic epidemiologic situation, a group A conjugate vaccine specifically designed for Africa (MenAfriVac) has been developed16. The vaccine is highly efficacious; no cases caused by serogroup A have been recorded among vaccinees16 and group A carriage among both vaccinated and unvaccinated people has disappeared, which indicates vaccine-induced herd immunity17. However, other serogroups also play an important role in the epidemiology of meningococcal disease in Africa, with outbreaks caused by serogroup C, W-135 and, more recently, X181920. In Asia, serogroups A and C are responsible for the majority of cases; in some countries, other serogroups are playing an increasing role21.

Endemic meningococcal infection in industrialized countries displays a relatively stable and sporadic background of incidence22. However, prolonged epidemics have also been described2324. In Europe, the overall incidence of IMD diminished from 1.9 per 100,000 in 1999 to 0.73 per 100,000 in 2010, thanks to the widespread introduction of the Nm serogroup C (NmC) conjugate vaccine; Nm serogroup B (NmB) has, therefore, become by far the most frequent causative agent of the disease25. A similar serogroup distribution has also been reported in Australia and New Zealand5. Serogroup Y strains are relatively common in the United States (US) (accounting for more than 30% of cases) and other countries of the American continent26. Moreover, cases caused by this serogroup have increased in some parts of Europe27.

Importance of carriage in the transmission of the disease

Nm has its own unique survival niche in humans28; it is considered a normal commensal of the upper respiratory tract, although it can lead to serious invasive disease. The reservoir of Nm is substantially constituted by healthy carriers. The proportion of carriers varies in the different stages of life; it is low among infants and school children, and then increases during adolescence and early adulthood. The relationship between carrier status and the development of IMD is a subject of research and is not yet fully understood29.

Evaluating carriage is relevant to comprehending both the dynamics of carriage and disease and the potential effect of vaccination on the transmission of Nm. For example, the NmC Conjugate Vaccination Programme in the United Kingdom (UK) was successful because the vaccine not only protected against the disease, but was also able to prevent the acquisition of carriage, thereby enhancing herd immunity3031. This explains how strengthening the immune system by means of appropriate vaccines can lead to marked reductions in the incidence of meningococcal disease2830.

Dutch OMV vaccines

A Dutch hexavalent OMV-based vaccine (HexaMen) consisted of OMV of two recombinant engineered strains, each of which expressed three different PorA subtypes (P1.5-2,10; P1.12-1,13; P1.7-2,4; P1.19,15-1; P1.7,16; and P1.5-1,2-2)6768. In a UK study, 103 infants received HexaMen at 2, 3 and 4 months, together with routine vaccines, and a booster dose was administered at 12–18 months. Good immune responses against two of the six NmB strains which expressed PorA contained in the vaccine were observed after the three-dose course. After the booster dose, higher h-SBA responses were observed, suggesting that the primary course had primed memory lymphocytes and that revaccination stimulated a booster response39. In another Dutch study, which involved toddlers aged 2-3 yr, HexaMen was administered through a three-dose scheme at different vaccine doses. The percentage of subjects showing a four-fold increase of h-SBA titres against the specific serosubtype varied from 28 to 98 per cent, with no statistically significant difference between higher and lower doses of the vaccine68. In both studies, HexaMen was seen to be well tolerated and safe3968.

The theoretical coverage of HexaMen, based on the exact match of vaccine subtypes, was estimated to be 50 per cent of NmB in the Netherlands69. To ensure a broader level of protection, a third recombinant OMV was added; the result was a nonavalent vaccine (NonaMen)67. On the basis of European PorA subtype data from 1999 to 2004, NonaMen was estimated to have a potential coverage of 80 per cent70.

One problem of OMV-based vaccines lies in the ability of meningococcal OMPs to undergo antigenic shift or gene deletion, as seen with PorA, thus rendering the vaccines ineffective71. Another limitation of OMV vaccines is that the immunity elicited rapidly wanes. To overcome the limitations of old and new OMV vaccines, an alternative strategy (i.e. reverse vaccinology) has been undertaken.

Meningococcal B vaccine by means of reverse vaccinology

Thanks to the steady progress of bioinformatics, an innovative approach to the development of an NmB vaccine has been implemented and optimized. Unlike conventional methods, this approach begins by defining the genome sequence of the pathogen and continues with the computer-assisted prediction of more promising antigens for the new vaccine72. The first vaccine designed by means of the reverse vaccinology approach was rMenB73. Initial scrutiny of the NmB genome (MC58 virulent strain) turned up about 600 antigens, approximately 350 of which were expressed in Escherichia coli; these were used to immunize mice. Subsequent analysis of serum from immunized mice uncovered 91 previously unknown surface proteins that were able to induce antibodies in vivo, 29 of which induced bactericidal antibodies in vitro. Later research identified five antigens, four of which were expressed as fusion proteins [Genome-derived Neisseria Antigen 1030 (GNA1030) with GNA2132 and GNA2091 with GNA1870], while the fifth, Neisseria adhesin A (NadA), was not fused747576.

GNA1870 (factor H binding protein - fHbp) is a surface-exposed lipoprotein that binds factor H. As this is an effective inhibitor of the alternative complement pathway, it protects the pathogen from complement-mediated killing77. Early research found that all hypervirulent B strains contained fHbp7879. More recently, however, Lucidarme et al80 have described isolates from patients with IMD which do not express fHbp. As fHbp can differ from strain to strain, two approaches to classifying this antigen have been developed. The first approach divides fHbp into two subfamilies: A and B; within a subfamily, the amino acid sequence displays 83 per cent homology or more, while between subfamilies, homology is approximately 60-75 per cent81. The second approach, which was proposed by Masignani et al79, identifies three variants of fHbp (1, 2 and 3), with homology of 91.6-100 per cent within variants and 62.8 per cent between variants. Subfamily A corresponds to variants 2 and 3, while subfamily B corresponds to variant 182.

NadA is a member of non-fimbrial adhesins, or rather oligomeric coiled-coil adhesins; these antigens are all trimeric autotransporter adhesins. NadA is of vital importance to Nm, as it mediates binding and the subsequent invasion of human epithelial cells; indeed, antibody binding to NadA results in the killing of Nm, even in the presence of its polysaccharide capsule83. The recombinant NadA contained in the meningococcal B vaccine conserves the functional features of native NadA and is able to induce high levels of bactericidal antibodies in various models; moreover, it is recognized by serum from convalescent children84. It has been also shown that NadA is mostly associated with disease isolates rather than with isolates from carriers85.

GNA2132 (Neisseria Heparin Binding Antigen - NHBA) is another surface-exposed lipoprotein discovered by reverse vaccinology, and is able to induce bactericidal antibodies in humans. This antigen is an important virulence factor that binds heparin, thereby promoting Nm resistance in blood. Genetically diverse B strains present variable segments of NHBA; however, C- and N-terminal regions are highly conserved8687.

The new vaccine was called 5CVMB (five component vaccine against NmB) and included 20 μg of each of the two protein-protein fusions (GNA1030-GNA2132 and GNA2091-GNA1870) and 20 μg of NadA, adsorbed to Al(OH)₃88. In a preclinical study, the bactericidal activity of 5CVMB was tested against 85 different strains of NmB; 5CVMB was also compared with two OMV vaccines in terms of the bactericidal activity. Sera from mice vaccinated with OMV vaccines made from the Norwegian strain H44/76 and the New Zealand strain NZ98/254 were able to kill 20 and 21.2 per cent of the strains, respectively, while sera from mice immunized with 5CVMB killed 77.7 per cent of the strains88. The final formulation was called four component meningococcal serogroup B (4CMenB) vaccine and consisted of recombinant NmB NHBA fusion protein (50 μg), recombinant NmB NadA protein (50 μg), recombinant NmB fHbp fusion protein (50 μg), and OMV from NmB strain NZ98/254, measured as the amount of total protein containing PorA P1.4 (25 μg). OMV-NZ was added to achieve broader strain coverage and to reduce the risk of escape mutants. The vaccine contained 0.5 mg of Al(OH)₃ as an adjuvant; other excipients were sodium chloride, histidine, sucrose, and water for injection89. On January 14, 2013, the Committee for Medicinal Products for Human Use of the European Medicines Agency (EMA) authorized 4CMenB, commercially named Bexsero, for use in subjects from two months of age90.

Clinical trials in infants and toddlers

The successful results of preclinical studies and phase 1 studies on adults were followed by a series of clinical trials in infants and toddlers. To evaluate immunogenicity in infants, a phase II, single-blind, randomized trial was conducted in the UK. Infants aged 6-8 months were randomized in a 1:1 ratio to receive recombinant meningococcal serogroup B (rMenB) or rMenB+OMV vaccines on day 0, day 60 and at the age of 12 months. h-SBA was evaluated against seven different NmB strains. After three doses of rMenB+OMV, h-SBA titres ≥4 against five NmB strains were found among ≥90 per cent of participants, and 70 per cent of participants also showed h-SBA titres ≥4 against the sixth NmB strain. When the infants who received rMenB alone were evaluated, it was found that 88 per cent of them showed an h-SBA titre ≥4 against only three of the seven strains tested. Both vaccines were found to have acceptable safety and tolerability profiles91.

Another phase II clinical trial demonstrated good immunogenicity in infants after three vaccine doses at 2, 4 and 6 months of age (rMenB or rMenB+OMV). rMenB+OM Velicited h-SBA titres ≥4 against 5/99 (anti-NadA response), 44/76-SL (anti-fHbp response), NZ 98/254 (anti-PorA response), and M00 242922 (anti-PorA response) strains in 95, 87, 85 and 63 per cent of subjects, respectively. The rMenB vaccine alone displayed comparable immunogenicity to that of rMenB+OMV, except against NZ 98/254 and M00 242922. Furthermore, the fourth dose of rMenB+OMV at 12 months of age elicited a good anamnestic response, with h-SBA titres ≥4 against 44/76-SL, NZ 98/254, 5/99, and M00 242922 strains in 100, 93, 96 and 78 per cent of subjects, respectively. Both vaccine formulations were well-tolerated; however, local reactions of induration and tenderness were more frequent after rMenB+OMV than after rMenB92.

A Phase 2b, multicenter, open-label, parallel-group, randomized controlled study of 1885 infants was conducted in Europe between 2008 and 2010. Infants aged 2 months were randomized in a 2:2:1:1 ratio to receive: (i) 4CMenB at 2, 4, and 6 months together with their routine vaccines; (ii) 4CMenB at 2, 4, and 6 months and routine vaccines at 3, 5, and 7 months; (iii) 4CMenB administered together with routine vaccines at 2, 3, and 4 months; and (iv) a control group in which only routine vaccines were administered at 2, 3, and 4 months. In all three groups receiving three doses of 4CMenB, h-SBA titres ≥ 5 against 44/76-SL (anti-fHbp response) and 5/99 (anti-NadA response) were found in 99.1-100 per cent of the infants. h-SBA titres ≥ 5 against NZ98/254 (anti-PorA response) ranged from 79.0 to 86.1 per cent across three groups. No clinically significant interaction with routine vaccines was found. Local reactions of erythema, swelling or induration were observed in less than 1 per cent of all infants. Fever was noted in 80, 71, 76 and 51 per cent of infants in the first, second, third and control groups, respectively93.

The first phase III clinical trial evaluated the immunogenicity and safety of 4CMenB when administered with routine vaccinations at 2, 4 and 6 months of age. After the third dose, 100 per cent of infants had h-SBA titres ≥5 against 44/76-SL (anti-fHbp response) and 5/99 (anti-NadA response) strains, and 84 per cent had h-SBA titres ≥5 against NZ98/254 (anti-PorA response) and M10713 (anti-NHBA response) strains. After a booster dose, h-SBA titres ≥5 against all four strains were achieved in 95-100 per cent of infants. This trial also found that 4CMenB did not interfere with routinely administered vaccines. Concomitant vaccination was associated with increased reactogenicity. In particular, concomitant vaccination more frequently produced fever; in the majority of cases, however, this resolved in one day94.

Bivalent recombinant lipoprotein 2086 (rLP2086)

Another vaccine currently under clinical development is based on fHbp [also called lipoprotein 2086 (LP2086)]73. According to the above-described division of NmB strains into two subfamilies on the basis of fHbp8182, this vaccine (rLP2086) is bivalent, as it is composed of a representative variant of each subfamily (A05 and B01). The immunogenicity, safety and tolerability of this vaccine were investigated in a randomized controlled trial in infants aged 18-36 months. Specifically, 99 healthy toddlers were subdivided into three dose cohorts – dose cohort 1: 20 μg rLP2086, dose cohort 2: 60 μg rLP2086 and dose cohort 3: 200 μg rLP2086. Each cohort was matched with a control group of subjects, who received hepatitis A virus (HAV) vaccine in a 2-dose schedule and a saline placebo administered at the second of the 3 vaccination time-points (0, 1 and 6 months). After dose 3, seroconversion (h-SBA ≥ 4-fold rise from baseline) against NmB strains expressing LP2086 variants homologous to the vaccine antigens was found in 61.1-88.9 per cent of toddlers (rate dependent on dose-level) and against NmB strains expressing heterologous LP2086 variants in 11.1-44.4 per cent. This study indicated that the rLP2086 vaccine had an acceptable safety profile and was well tolerated95.

Immunization schedules of 4CMenB (Bexsero) in infants and toddlers

Bexsero has been approved for active immunization against disease caused by NmB in subjects aged ≥ 2 months. In those aged 2-5 months, the primary immunization schedule comprises 3 doses, with an interval of at least 1 month between doses. The primary schedule may be boosted at 12-23 months of age. The schedule for unvaccinated infants aged 6-11 months comprises 2 doses, with an interval of at least 2 months, and a subsequent booster dose in the second year of life. In toddlers aged 12-23 months, the primary schedule recommends two doses, administered at least two months apart; a booster dose may be administered 12-23 months after the primary course89.

4CMenB (Bexsero) may be co-administered with one or more of the infants (monovalent or combination) vaccines, such as a cellular pertussis, diphtheria, Haemophilus influenzae type b, hepatitis B, heptavalent pneumococcal conjugate, inactivated poliomyelitis, measles, mumps, rubella, tetanus and varicella8994.

Potential coverage of 4CMenB (Bexsero) against invasive NmB disease

The potential coverage of 4CMenB has recently been assessed in seven European countries by means of the Meningococcal Antigen Typing System (MATS)96. MATS is a standardized and reproducible vaccine antigen-specific ELISA, which has been developed to measure the amount of each antigen expressed by a strain and its immunological cross-reactivity with the antigen present in the 4CMenB vaccine97. In this large epidemiological survey (1052 hypervirulent strains collected) MATS analysis showed that 4CMenB could cover a significant proportion of European strains that cause invasive disease. Overall predicted coverage was 78 per cent, with some variation between single countries (73% in England/Wales, 85% in France, 82% in Germany, 87% in Italy, 85% in Norway, 74% in the Czech Republic and 69% in Spain)96. In a recent study, the potential coverage of 4CMenB against Canadian hypervirulent strains circulating from 2006 to 2009 was estimated by applying MATS analysis to 157 isolates from adults and children. Overall, MATS predicted a strain coverage of 66 per cent (95% CI: 46-78%), with 26, 29 and 11 per cent of strains covered by one, two and three vaccine antigens, respectively98.

Potential efficacy of 4CMenB (Bexsero) against serogroups other than B

Given that the 4CMenB antigens may be present in the external membrane of all pathogenic Nm, this vaccine has the potential to prevent disease caused by different serogroups. In this regard, interesting results have recently been published by Hong et al99, who estimated the potential coverage of 4CMenB against serogroup X isolates from several African countries. Using MATS, the authors concluded that 4CMenB could cover African isolates, since the strains tested expressed at least one vaccine antigen (in particular fHbp).

Potential coverage of rLP2086 against NmB disease and against serogroups other than B

Although the bivalent rLP2086 vaccine has not completed all clinical phases, studies on the potential efficacy of this vaccine against NmB and other serogroups have been carried out. In a recent study, Anderson et al100 reported that rLP2086 was able to provide broad protection against invasive NmB strains. They tested the sera of vaccinees against different isolates from Europe and the US. The proportion of vaccinees with h-SBA titres ≥4 against hypervirulent NmB strains with different variants of fHbp ranged from 75 to 100 per cent.

As fHbp is expressed by other Nm serogroups, the anti-fHbp antibodies elicited by rLP2086 might exert a bactericidal effect on meningococci, regardless of the serogroup. On the basis of this assumption, Harris et al101 tested some invasive NmC isolates in a preclinical study. They demonstrated that all NmC isolates had the fhbp gene, and the non-human serum showed bactericidal antibody activity against the NmC tested.