Controlling serogroup B invasive meningococcal disease: the Canadian perspective

Abstract

With publically funded meningococcal immunization programs established in infants, children and adolescents, Canada is at the forefront of invasive meningococcal disease prevention. The advent of two new serogroup B vaccines that may protect against multiple disease-causing strains offers the potential to reduce endemic disease to very low levels in Canada. Canada likely will be one of the first countries with approval to use recombinant serogroup B vaccine. However, inclusion of these new vaccines into public immunization programs will be decided at the provincial/territorial level, rather than nationally, and may result initially in different immunization schedules throughout the country as we have seen with conjugate meningococcal vaccines. Such heterogeneous use and adoption of new vaccines complicates disease control, but may assist in evaluation of effectiveness. Minimally, it requires regionally specific information. In this article, the authors provide an overview of the Canadian epidemiology, serogroup B vaccine characteristics, potential strain coverage, immunization strategies and remaining postmarketing research questions.

Keywords::

• Canadian epidemiology
• invasive meningococcal disease
• meningococcal immunization strategies
• serogroup B vaccines

Epidemiology

Invasive meningococcal disease (IMD) has been a nationally notifiable disease in Canada since 1924 and it is reportable in every province and territory (PT) in the country. Passive case reporting occurs from laboratories, hospitals and healthcare providers to local public health officials who then forward reports to provincial/territorial public health. Each PT then sends aggregate data to federal public health where national statistics are compiled. In addition, each PT sends Neisseria meningitidis isolates to the National Microbiology Laboratory in Winnipeg (MB, Canada) for identification. Since 2000, cases occurring north of 60° latitude have also been reported to the International Circumpolar Surveillance, a passive surveillance network of hospitals, public health agencies and reference laboratories throughout the Arctic [1]. As ascertainment of cases with passive surveillance is not expected to be comprehensive, since 2002 the Canadian Immunization Monitoring Program Active (IMPACT) has conducted an active, population-based sentinel surveillance for IMD in adults and children across Canada, which includes systematic collection and typing of isolates and covers approximately 50% of the population [2]. The Canadian epidemiology data discussed in this article are derived from IMPACT and the notifiable disease surveillance unless otherwise noted.

Following an outbreak of serogroup C disease from 1999 to 2001, PTs in Canada implemented universal immunization programs for infants, children and adolescents using monovalent C vaccines (Table 1). The initial program has evolved since its start and several PTs now provide a quadravalent vaccine to children or adolescents, usually through school-based immunization programs. Although coverage data are sparse or nonexistent in most Canadian jurisdictions, data from British Columbia are complete and show 86–88% of 2-year-old children and 88–93% of 11-year-old children were fully immunized against serogroup C from 2005 to 2009 [101,102]. Using current immunization schedules and assuming the coverage in British Columbia was representative of the rest of Canada, White et al. estimated that it would take until 2022 for all Canadians 20 years and younger to be immunized with at least one dose of a meningococcal conjugate vaccine [3]. IMPACT shows that N. meningitidis serogroup B (MenB) was responsible for the majority (64%) of IMD in 2010; the remainder was due to serogroup Y (19%), serogroup C (7%), serogroup W135 (3%) and other/untypable/unknown (7%).

Serogroup B causes approximately 110 cases of IMD each year across Canada [103]. From 2002 to 2009, IMPACT data show that the average annual Canadian incidence of MenB was 0.23 per 100,000 and did not change significantly when the disease trends were modeled with Poisson regression [4]. The highest incidence was in infants <1 year of age (a high of 6.2 per 100,000 in 2009), followed by children 1–4 years of age (range: 0.4–1.4 per 100,000) [4]. In Ontario, the most populous Canadian province, over a 12-year period (2000–2011), passive surveillance data from the province’s reportable disease and provincial reference laboratories databases revealed that MenB incidence remained stable ranging from 0.11 to 0.27 per 100,000, with the highest annualized age-specific incidence in infants (3.55 per 100,000) [5,6]. A similar disease pattern was seen in the other provinces with the exception of Quebec [7,8].

Nationally, 48% of Canada’s MenB cases in 2002–2009 occurred in Quebec, a province with 23% of the Canadian population [4]. Over the last 3 years (2009–2011), 88% of all IMD cases within Quebec were caused by MenB [9]. Although the incidence between 1997 and 2010 has remained stable in children <1 year and children 1–4 years of age, the incidence in adolescents (i.e., 15–19 years of age) has increased significantly (from 0.4 to 2.4 per 100,000) [10]. Since 2003, this increase has been attributed to an emerging clone (B:17:P1.19), which accounted for 65% of all MenB disease in the province and 51% of cases in 15- to 24-year-old adolescents and young adults [10].

As in other countries, MenB disease disproportionately affects healthy young infants and children. Data from IMPACT showed 44% of MenB cases occur in children <5 years of age, with 14% of all cases occurring in infants 0–5 months and 6% in infants 6–11 months of age [4]. Surveillance data were similar from Ontario with 38% of cases occurring in children <5 years of age; 21% of total cases occurred in infants <1 year, with 10% occurring in infants 0–4 months of age. Of note, in Ontario, the median age of cases increased from 5 years in 2000 to 20 years in 2010 [5,6].

Disease burden in Canada

For individuals with MenB disease, the mean length of hospitalization nationally was 11.2 days with approximately 61% requiring treatment in the intensive care unit for a mean of 5 days [4]. The overall case–fatality ratio ranged from 5.3% in IMPACT to 10.7% in Ontario’s passive surveillance data, with infants <1 year and adults 65 years and older having the highest case–fatality rates, both in Ontario and nationally [4,5]. This is comparable with rates found in other countries [11]. Among survivors, IMPACT data indicated 19% of infections resulted in sequelae at hospital discharge. Deafness, skin scarring and amputation were the most frequent sequelae [4].

Number needed to vaccinate

Using passive surveillance data from Ontario, Dang et al. calculated the number of infants that would require vaccination to prevent one case of MenB disease [5]. Assuming an annualized age-specific incidence rate of 3.55 per 100,000, a vaccine efficacy ranging from 70 to 80% and only direct protection (no herd effects), the authors crudely calculated that a total of between 35,211 and 40,241 infants would need to be vaccinated to prevent one infant case of MenB disease. It is important to note that this assumes that all cases under 1 year are vaccine preventable. If we assume that cases in infants younger than 4 months of age are not vaccine preventable, the number needed to vaccinate, assuming no herd immunity, increases to between 68,107 and 77,836 infants [5,6]. Although there is no agreed-upon threshold for number needed to vaccinate, these data are relatively simple to calculate and can be useful for decision makers, as long as limitations and assumptions are clear.

Molecular epidemiology

N. meningitidis surface antigens & population biology

Epidemiologic characterization of N. meningitidis is traditionally performed by serological detection/identification of surface antigens (capsular serogroup and subcapsular major outer membrane proteins [MOMPs] specifying serotype and serosubtype) and classification of isolates at the population level by either multilocus enzyme electrophoresis or multilocus sequence typing into clonal groups or lineages [12]. Methods and terms commonly used to characterize N. meningitidis are summarized in Table 2. The population biology of N. meningitidis has been informed by studies of isolates from healthy carriers and invasive disease cases. For example, of the 12 serogroups recognized, most IMD is caused by strains belonging to serogroups A, B, C, Y, W135 [13,14] and more recently X [15]. By contrast, N. meningitidis recovered from healthy carriers usually either do not express their capsular serogroup antigens or do not have the genes essential to synthesize them [16]. A small number of clonal groups or lineages is responsible for the majority of the disease cases, especially for strains associated with outbreaks or epidemics, whereas an almost unlimited number of genotypes may be present among carriage strains [17]. Despite some exceptions, a strong association between clonal lineages and surface antigens (capsule serogroup, MOMPs serotype and serosubtype) has been described in N. meningitides [12,18]. However, surface antigens are subjected to immune selection, and antigenic shifts in MOMPs have been associated with increased meningococcal disease activity [19,20].

Another essential feature of the genetics of N. meningitidis is the plasticity of its genome allowing extensive genetic recombination and horizontal gene transfer between strains of the same or related species [21]. Use of newer genome sequencing technology has shed light on what accounts for virulence in N. meningitidis and why certain clonal lineages are maintained. N. meningitidis is now understood to have a supra-genome with a gene pool larger than the number of genes present in any particular strain [22,23]. This essentially allows strains to pull from this gene pool the genes necessary to thrive in a range of conditions. Comparative genomic studies have also suggested that the virulence trait of meningococci probably depends on a number of independent genes [24]. Furthermore, hyperinvasive clonal complexes are probably maintained by the presence of unique restriction modification systems that only provide a selection of foreign DNA, thus maintaining the genetic structure [18].

External forces can also affect the population biology of meningococci. When vaccines targeting some, but not all, capsular serogroups of meningococcci are used in a population, the vaccine-induced immunity may theoretically impose a selection pressure on the meningococcal population, potentially favoring capsule switching [25–27] and replacement [28,29]. Capsule switching occurs when vaccine strains (i.e., strains with capsule types included in the vaccine) acquire alternate capsular polysaccharide synthesis genes from nonvaccine strains (i.e., strains with capsule types not found in the vaccine) to become resistant to vaccine-induced host immunity. Capsule replacement occurs when nonvaccine strains emerge to cause disease as vaccine strains are eliminated by vaccine use, for example, when serogroup C (MenC) conjugated vaccines eliminate MenC strains from the respiratory tract, allowing non-C serogroups to proliferate. Whether a similar phenomenon can occur when vaccines are based on protein antigens that occur in some, but not all, strains is an intriguing question.

Molecular epidemiology of MenB disease in Canada (2001–2010)

In the MenC conjugate vaccine era, most IMD in Canada is caused by MenB strains [4]. In the past decade (2001–2010), the MenB isolates recovered from IMD cases in Canada’s most populous provinces (Ontario and Québec) were found to be quite different from each other [9,30], possibly reflecting differing disease activity in these adjacent provinces. For example, in the province of Québec that had a population of 7.9 million in 2010, there were 466 culture-positive IMD cases during the period of 2003–2010 and 72% (n = 334) were due to MenB. Of the 344 MenB cases, 54% (n = 180) were due to the ST-269 clonal complex (cc). In Québec, a single ST was responsible for 92% (n = 165) of the cases caused by cc 269 with 85% of the strains (n = 140) expressing the antigenic formula of 17:P1.19 and the PorA genotype of P1.19-1, 15-11, 36. By contrast, in Ontario, which had a population of 13.3 million in 2010, there were 494 culture-confirmed IMD cases between 2001 and 2010 and 39% (n = 193) were due to MenB. The most common cc involved was ST-41/44, which was responsible for 75 cases or 39% of all MenB cases, and extensive strain heterogeneity was detected with 35 different STs. Although cc 269 was the second most common cc responsible for MenB cases in Ontario in 2010, considerable strain diversity was detected with seven different STs and a variety of antigenic profiles found among the 20 cc 269 case isolates.

The emergence of the ST-269 clone that caused outbreaks in Québec, but not other provinces, may be associated with different vaccination policies used to control the MenC ET-15 outbreaks in the early 1990s and 2000s. In Québec, single-dose province-wide vaccination programs were implemented, using the plain polysaccharide vaccine for those 6 months to 20 years old during the first wave of a MenC outbreak in 1992–1993 [31] and the MenC conjugated vaccine for those aged 2 months to 20 years during the second wave of MenC outbreaks in 2001 [32]. By contrast, only small-scale targeted single-dose vaccination programs were used in Ontario for control of both waves of MenC outbreaks until a single-dose publicly funded universal MenC conjugate vaccination program was introduced for children 12 months of age and for grade 7 students (~12 years of age) in 2004/2005 (Table 1). Whether the replacement of MenC with MenB disease in Québec was driven by vaccine pressure is unknown.

The ST-269 clone first emerged in Québec in 2003 associated with localized outbreaks and an increase in incidence of MenB in the age group of 15–24 years old [33]. The clone’s persistence in Québec [9] resulted in an increase in MenB incidence in the younger and older age groups but without significant spread to neighboring provinces [National Microbiology Laboratory, Unpublished Data] [30], which is very different from the rapid dissemination seen with the hyperinvasive MenC clone of ET-15 (ST-11). Shortly after the ET-15 clone was identified as a cause of MenC outbreaks in Ontario [34], this strain quickly spread across the country [35,36] to the USA [37] and globally [38]. The confinement of MenB disease due to a single clone within a specific jurisdiction has also been seen in Oregon, USA due to the ET-5 (cc32) clone [26]. The apparent difference in the epidemiology of IMD due to the hyperinvasive serogroup B cc32 or cc269 clones and the MenC ET-15 clone is still not understood.

The polygenic nature of virulence in N. meningitidis and its various mechanisms of altering its surface antigens makes surveillance and control of meningococcal disease a challenge. Knowledge of the population biology, clonal diversity, genetic and antigenic profile of N. meningitidis will be useful baseline information before introduction of newer MenB vaccines, as well as for an understanding of how the bacteria may change upon introduction of these vaccines.

Vaccine characteristics

Approaches to development of MenB vaccines

Unlike the serogroups A, C, Y and W135, capsular polysaccharides that induce high titers of serum bactericidal antibodies, the MenB capsule is nonimmunogenic due to structural similarity to the neural cell adhesion molecule [39]. Although methods have been developed to circumvent this immunological tolerance [40,41], the possibility of inducing autoimmunity with these capsular candidates is too grave to overcome. Despite this concern, autoimmune diseases have not been described after invasive MenB disease or when antibodies to the MenB capsule are present in humans [42,43]. Nevertheless, most current efforts to develop MenB vaccines have focused on antigens found beneath the capsular polysaccharide. Recent reviews have already described in detail the different approaches to develop MenB vaccines [44,45].

The two most promising MenB recombinant vaccines either incorporate a number of recombinant surface proteins (fHbp, NadA, NHBA, GNA1030 and GNA2091) together with a conventional MenB OMV component prepared from the New Zealand outbreak strain 98/254 [46], or contain two families of a newly identified Neisseria surface protein, fHbp [47].

Safety, immunogenicity & efficacy: 4CMenB

Initial clinical trials compared two vaccine formulations, both containing 50 µg each of fHbp, NadA allele 3 and NHBA. Both also contained OMV, one from the Norwegian outbreak strain and the other from the New Zealand outbreak strain. The latter was selected for further trial as the New Zealand OMV was more immunogenic and the resultant four-component vaccine was called 4CMenB (Novartis, MA, USA) [48–50].

The safety and immunogenicity of 4CMenB vaccine given concurrently with combined diphtheria, tetanus, acellular pertussis, inactivated poliovirus, hepatitis B and Haemophilus influenzae type b conjugate vaccine (DTaP–IPV–Hib–HepB) and 7-valent pneumococcal conjugate vaccine (PCV7) was studied in 1885 infants immunized with 4CMenB and routine vaccines concurrently at 2, 4 and 6 months or 2, 3 and 4 months, or with 4CMenB vaccine separately at 2, 4 and 6 months with the routine vaccines given at 3, 5 and 7 months [51]. In this multicenter European study carried out in six countries, injection site pain and fever were more common after the 4CMenB vaccine given with concomitant vaccines than after the concomitant vaccines were administered separately. Antibody responses to the 4CMenB vaccines were demonstrated with both immunization schedules whether or not other routine vaccines were given concurrently. Increased immunogenicity was generally observed in the 2-, 4-, 6-month schedule compared with the 2-, 3-, 4-month schedule, and when the 4CMenB vaccine was given alone rather than concurrently with other vaccines. Concurrent immunization with 4CMenB did not have an adverse effect on immunogenicity to the routine vaccine antigens except for diminished responses to Bordetella pertussis pertactin and Streptococcus pneumoniae serotype 6B.

Lot-to-lot consistency and concomitant immunization with routine vaccines were also analyzed in a Phase III study of 3630 infants given one of three lots of 4CMenB plus routine vaccines (DTaP–IPV–Hib–HepB and PCV7), routine vaccines alone, or routine vaccines plus MenC vaccine [52]. As with the Phase II study in infants, no effect on immunogenicity was observed with concomitant vaccination and consistency was demonstrated between lots. Some increase in fever was demonstrated for participants receiving 4CMenB vaccine with routine vaccines concurrently. Concomitant immunization with 4CMenB vaccine and measles–mumps–rubella–varicella vaccine (MMRV) was studied in 426 of the 4CMenB vaccine recipients at 12 months of age [53]. Prior to immunization, high antibody levels against fHbp and NadA and lower levels against NHBA and OMV antigens were observed; postimmunization, booster responses were demonstrated against all antigens with no effect of the concomitant MMRV vaccine. In addition, no effect of 4CMenB vaccine on MMRV vaccine responses was detected. No effect on adverse events was demonstrated with administration of 4CMenB and MMRV vaccines; with concurrent vaccination, a biphasic fever response was observed with fever in the first 48 h consistent with the 4CMenB vaccine and fever at 8–10 days consistent with MMRV vaccine.

Concomitant administration of 4CMenB and MMRV vaccines was also explored in 12-month-old toddlers who had not received prior 4CMenB [54]. A total of 285 toddlers were given MMRV vaccine at 12 months and 4CMenB vaccine at 13 and 15 months, and 185 toddlers were given MMRV and 4CMenB vaccines at 12 months and a second dose of 4CMenB vaccine at 13 months of age. Two doses of 4CMenB vaccine elicited high rates of serum bactericidal antibody titers with no difference in the separate and concurrent immunization groups. Adverse events were marginally more frequent in the concomitant compared with the separate vaccination group.

Because of the observed increased frequency of fever associated with 4CMenB vaccine, the use of prophylactic acetaminophen was explored in 367 infants who were part of a larger study of 4CMenB and concomitant pediatric vaccines (DTaP–IPV–Hib–HepB and PCV7) [55]. One dose of acetaminophen was administered prior to and two doses after vaccination, separated by 4–6 h. Use of acetaminophen significantly reduced postvaccination fever without an adverse effect on the proportion of infants with a response to the concurrent vaccines; however, there was a trend towards lower geometric mean antibody titers to most concurrent vaccine antigens that did not reach statistical significance.

Persistence of the antibody response was examined using a serum bactericidal assay with a human complement source in children previously immunized with 4CMenB vaccine as young infants [52,56] or older infants/toddlers [57]. Substantial waning of antibody responses was demonstrated, although the proportion of children with bactericidal antibody levels ≥4 far exceeded age-matched naive controls; an anamnestic response was observed after a booster dose. Waning after infant vaccination occurred at different rates among the components, being greatest for PorA1.4 and NHBA and least for NadA. Implications of this uneven decline in antibodies for vaccine effectiveness (VE) are unknown. To anticipate waning responses after infant vaccination, a booster dose in the second year will probably be a routine requirement, perhaps within a 2-, 4- or 12-month schedule, if validated. Fever reactions were uncommon in children receiving a booster dose, but injection site erythema and other injection site adverse events were common and similar in both primed and boosted and naive participants alike [58].

4CMenB vaccine was also studied in 1631 adolescents 11–17 years of age who received one, two or three doses at 1-, 2- or 6-month intervals [59]. From 93 to 96% of participants achieved bactericidal antibody titers ≥4 against fHbp, NadA and OMV after a single dose that rose to ≥99% after a second dose. Geometric mean antibody titers were higher after two doses than one dose. A third dose did not increase the proportion with antibody titers ≥4, but marginally increased the geometric mean antibody titer if the interval between the doses was increased to 6 months. Reactogenicity did not increase with subsequent doses; rather, there was a slight decrease in reported adverse events with subsequent doses.

The safety and immunogenicity of the three doses of 4CMenB at 0-, 2- and 6-month intervals was examined in 54 laboratory workers 18–50 years of age who were routinely exposed to N. meningitidis under laboratory conditions [60]. Antibody responses were apparent after each of the doses with 80–88% having serum bactericidal antibodies against the three antigens tested after dose 1, 91–100% after dose 2 and 92–100% after dose 3. Geometric mean antibody titers increased after the first two doses against all three of the antigens tested but did not increase further for two of the three antigens after the third dose. Rates of adverse events were similar after the first two doses and tended to be marginally higher after the third dose. A similar study in UK laboratory workers was recently reported [61]; the vaccine was well tolerated and 90–100% achieved seroprotective antibody titers against seven diverse MenB strains.

Potential coverage of the 4CMenB vaccine in Canada

The potential strain coverage of the 4CMenB vaccine was examined in strains circulating in Canada from 2006 to 2009 using the meningococcal antigen typing system (MATS) ELISA. The MATS is used to predict the individual strains that are likely to be covered by the 4CMenB vaccine [62]. Overall, 66% (95% CI: 46–78) of circulating strains were covered by one or more vaccine antigens. By antigen, 26% of strains were covered by one antigen, 29%, by two antigens and 11% by three antigens. No strains were covered by all four antigens. The vaccine appears to provide good coverage (75–90%) for the current prevalent Canadian ccs (cc269 and cc41/44) [63].

Safety, immunogenicity & efficacy: LP2086

A bivalent recombinant lipoprotein 2086 vaccine (rLP2086, Pfizer, NY, USA) comprising subfamily A and B LP2086 (also known as fHbp) proteins is being developed that is predicted to provide broad coverage against diverse MenB strains. In a Phase I, first-in-human study [64], 103 healthy young adults 18–25 years of age received three doses at 0, 2 and 6 months of 20-, 120- or 200 µg of the bivalent rLP2086 vaccine or placebo. Injection site pain was the most commonly reported adverse event with no differences among subjects receiving various dose levels. Induration and erythema were reported more frequently among the 200-µg recipients. Systemic adverse events were reported more frequently after the two higher dose formulations. IgG-binding antibody responses were observed after the first two doses against both LP2086 proteins, declined before the third dose (but remained above the placebo levels) and increased further after the third dose for the two-dose formulations. Bactericidal antibody assay results showed a similar pattern, with the two higher dose formulations eliciting higher antibody levels.

In a similarly designed study [65], 48 participants 18–40 years of age were randomized to receive three doses at 0, 2 and 6–9 months of 60, 120 or 200 µg of the bivalent rLP2086 vaccine or an adult formulation of diphtheria tetanus acellular pertussis vaccine (Tdap; first dose or placebo, second and third doses). Antibody responses were demonstrated against both subfamily A and B LP2086 for all three dose levels without an apparent dose response; however, geometric mean antibody levels increased with subsequent doses. Injection site adverse events tended to be higher in the rLP2086 groups than in the Tdap/placebo group with increased adverse events associated with the higher doses of rLP2086 vaccine. Systemic adverse event rates were higher in the rLP2086 recipients than in the Tdap/placebo recipients.

The rL2086 vaccine was studied in a Phase I trial of 99 healthy toddlers 18–36 months of age given three doses at 0, 1 and 6 months of one of three formulations containing 20, 60 or 200 µg of equal amounts of subfamily A or B rL2086 or hepatitis A vaccine (HAV) [66]. Tenderness was the most frequently reported injection site adverse event occurring more frequently in the rLP2086 recipients than in the HAV recipients. Tenderness did not appear to correlate with antigen quantity or with increasing number of doses. Erythema and induration were also more common after the rLP2086 vaccine than HAV and increased with antigen content but not with the number of doses received. Systemic adverse events were similar between the rL2086 formulations and the HAV, and did not vary by dose number. Serum bactericidal antibodies were elicited against a variety of strains at all vaccine dosages. A fourfold or greater rise in bactericidal antibodies was demonstrated in 61.1–88.9% of participants against strains expressing LP2086 variants homologous or nearly homologous to vaccine antigens and 11.1–44.4% against strains expressing heterologous LP2086 variants.

In a Phase II study, 539 adolescents 11–18 years of age were randomly allocated to receive three doses at 0, 2 and 6 months of 60, 120 or 200 µg of rLP2086 or placebo [67]. Following an initial safety review after immunization of the first 99 participants with the rLP2086 formulations, the remainder of the participants were randomized to receive the 120 or 200 µg formulation or placebo. Against eight diverse MenB strains, bactericidal antibody response ranged from 75.6 to 100% for the 120-µg dose and from 67.9–99% for the 200-µg dose. Injection site adverse events were more common after the LP2086 vaccine than after placebo; systemic adverse events were similar among vaccine and placebo recipients. Adverse events did not increase with increased antigen content or with increasing number of doses.

Potential coverage of the LP2086 vaccine in Canada

To date, no studies have been conducted to examine the potential coverage of the LP2086 vaccine in Canada.

Carriage & herd immunity

N. meningitidis are frequent colonizers of the nasopharynx with carriage rates averaging approximately 10% [68]. However, carriage rates vary by age and geographic location; in developed countries, these rates are lower in infants and children, peak in adolescents and diminish in adults [69]. There is no direct correlation between carriage and disease incidence; however, carriage is an important determinant of transmission. Herd immunity is defined as the ability of immune individuals in a population to protect nonimmune individuals against infection [69,70]. The ability of a meningococcal vaccine to provide herd immunity in large part depends on its ability to prevent carriage acquisition [71]. Meningococcal polysaccharide vaccines have demonstrated short-term herd immunity in military populations but not in open communities [71]. MenC conjugate vaccines have been demonstrated to reduce pharyngeal carriage and produce a herd effect [72,73]. After the implementation of universal MenC conjugate vaccination for all individuals younger than 18 years of age in the UK, meningococcal disease decreased by 67% among nonvaccinated children. Carriage dropped from 0.45 to 0.15% in the vaccinated age group, but remained constant in the older cohorts [73]. Some evidence of a herd effect of MenC conjugate vaccines was also demonstrated in The Netherlands [74] and in Canada [75,76].

Three studies have examined the effect of MenB OMV vaccines on carriage; in Chile [77], Norway [78] and Iceland [79], high vaccine coverage had no effect on rates of meningococcal carriage. A large study in the UK of nearly 3000 young adults immunized with 4CMenB and/or quadravalent meningococcal A, C, W135, Y conjugate vaccines is examining the effect on rates of meningococcal carriage [104]. The results of this study may provide insight into whether an effect on herd immunity could be expected with 4CMenB vaccine.

Immunization strategy

As of January 2013, the 4CMenB vaccine was licensed for use in Europe. The vaccine is currently under review by Health Canada. If authorized for use in Canada, a decision will need to be made regarding whether the vaccine will be publicly funded within jurisdictions as a universal program (to a targeted age cohort), a high-risk program, or only be available for private purchase. This decision will be quite difficult, given the relative rarity of the disease. Multiple factors will need to be considered, including the epidemiology of MenB disease, vaccine characteristics, cost–effectiveness, program feasibility and acceptability, as well as political considerations [80].

The epidemiology of MenB in most of Canada is typical of developed countries, with the highest burden occurring in infants <1 year of age. This epidemiology makes an infant program the most appealing strategy; however, it is noteworthy that a substantial disease burden occurs in very young infants (i.e., those younger than 5 months of age). In Canada, 14% of cases occur among this age group [4], and these cases will probably not be vaccine-preventable using a 2-, 4- and 6-month schedule, which conforms to the current infant immunization schedule used throughout the country. Results from clinical trials suggest that a 2-, 3- or 4-month infant schedule may be acceptable, although further information is needed given the lower immunogenicity that was observed using this schedule [51]. While this accelerated schedule would increase the number of infant cases that would be vaccine preventable, it does not conform to current immunization schedules in Canada and would result in increased visit costs and potential acceptability concerns, as an additional infant visit at 3 months of age would be required. The burden of disease in the province of Quebec, with a relatively high incidence in adolescence, may warrant a different immunization strategy, which would be associated with additional implementation challenges.

Given the current economic climate, cost considerations will likely be paramount. There is little published literature on the cost–effectiveness of 4CMenB vaccine; however, modeling data from the UK presented at scientific meetings suggest that the vaccine is not expected to be cost effective unless the price point is very competitive; unfortunately, we do not know what the per-dose vaccine costs will be. Another important influence on cost–effectiveness will be whether the vaccine results in herd effects and this is also not yet known. If there is associated herd immunity, cost–effectiveness will be improved [81]. The duration of protection and the need for booster doses in early childhood will be another key cost consideration.

Vaccine program success is also dependent on public acceptability and feasibility. Meningococcal disease is highly feared by the public [82]. Parents’ views on the severity of a disease influence their acceptance of a new vaccine [83]. By contrast, parental acceptability is also influenced by vaccine concerns, which include safety, undefined effectiveness, complex immunization schedules and multiple injections at single visits [84]. Currently, the infant schedule (i.e., infants <1 year of age) in Canada is relatively ‘crowded’ with pentavalent vaccines (DTaP–IPV–Hib) and PCVs given in all jurisdictions, and meningococcal C conjugate, rotavirus and hepatitis B vaccines also given in some jurisdictions [105]. Both parents and healthcare providers have expressed discomfort with multiple injections during a single visit, despite a lack of evidence to support this concern [85,86]. Nonetheless, this may impact acceptability, as would safety issues, such as the increased risk of fever that appears to be associated with 4CMenB vaccine. From a parental acceptance perspective, it is unknown whether concerns or fear regarding meningococcal disease will dominate over vaccine safety/scheduling concerns if the vaccine were to become publicly funded.

In general, it would be feasible for MenB vaccines to be introduced during infancy at 2, 4 and 6 months of age, as this is part of the routine vaccination schedule in all jurisdictions. However, as noted earlier, this will add complexity to the infant schedule. Healthcare providers may need to adjust timing of patient appointments to accommodate communication about the risks and benefits of the new MenB vaccine. Political considerations should not be overlooked, especially if there are strong proponents for public immunization programs. Finally, the decision for vaccine resource allocation may rest with federal or provincial governments, and be outside of the control of immunization program decision makers.

Expert commentary

Two recombinant MenB vaccines (4CMenB and rLP2086) have been developed and the 4CMenB vaccine has been approved in Europe and soon will likely be approved for use in Canada. These newer vaccines will enter the market with a number of important unanswered questions. It would be ideal if the first countries to utilize these new MenB vaccines would undertake appropriate studies to address these questions. However, evaluating a vaccine that targets a relatively rare condition poses unusual challenges and may require international collaborations. In the following sections, we list the key questions that will need to be addressed and propose the optimal methods to answer them.

What will the 4CMenB (& any other MenB vaccines) vaccine effectiveness prove to be, especially in infants?

This is a key question for any new vaccine, but particularly for one that is licensed based on immunogenicity data rather than efficacy data and that may not cover all disease strains. The requirements to assess VE will begin with excellent baseline disease incidence data, spanning persons of all ages (to monitor any herd effects after vaccine implementation) and multiple years. The observed rate should be stable and, ideally, relatively high. Data should be based on comprehensive, consistent surveillance, with central collection of isolates for typing. Data should be robust for the under 2 years age group (ideally with rates estimated by month of age under 6 months) as young infants typically have the highest incidence and will be the likely target group for MenB vaccination programs. Understanding the preventable fraction of disease in early infancy will be important for refining VE and cost–effectiveness estimates.

The probability of good uptake rates within the target population(s) will favor measurement of VE. Achieving this may be challenging where other meningococcal vaccines are already provided to the target population, requiring extra injections or clinic visits. Tolerability of the 4CMenB vaccine may become an issue as a result of the appreciable rate of febrile reactions in young vaccinees, although this can be moderated with use of prophylactic antipyretic medication. Public receptivity to new vaccines in general may also limit uptake. Low uptake will require a larger target population size and longer period of observation to determine VE.

Another important requirement for measuring VE is a defined coverage of local strains by 4CMenB vaccine. The strain coverage should be based on testing of an ample, representative collection of disease isolates spanning multiple years and all ages. For 4CMenB vaccine, testing by the MATS assay is the only practical option. Presumably only countries with moderately high coverage of strains affecting the target group will be motivated to implement programs. Regional differences in coverage rates have been detected in some large countries, which could confound national VE estimates. Such rate variations might favor regional implementation of programs for smaller target populations, but would require annual assessment of circulating strains at a regional level to determine the stability of strains that are covered. Population mobility into and out of regions of use will complicate VE estimates. The ideal situation to estimate VE would be a small country with a large population, within which a high coverage rate has been determined with minimal regional differences.

The next requirement for VE estimation is immunization of a suitably large population to recognize true rate reductions in the target group. The ideal situation will be a population large enough to measure disease rates in small age increments for programs aimed at infants. This would enable early recognition of reduced disease rates, as well as any return towards baseline rates at later ages signaling loss of protection. A protective effect may not be evident until after 6 months of age, depending on the number and timing of infant doses needed to impart protection. Since vaccine uptake may be difficult to predict, it would be best to plan the evaluation in terms of numbers needed to vaccinate. This also allows for ready adjustment for the expected strain coverage rate. While multinational studies have been appropriate for other vaccines against uncommon infections, variation in coverage rates and surveillance modes among countries will make a collaborative approach problematic for measuring VE.

Investigation of nonprevented MenB cases will be essential for VE estimation. Recognition of such cases will require comprehensive case surveillance, accurate immunization information and enhanced testing of isolates from MenB cases. Surveillance of cases among unvaccinated members of the target age group will create opportunities for case–control studies and provide ongoing information about strain coverage rates. Such cases will also provide opportunities to explore reasons for nonvaccination. Enhanced testing of MenB isolates from vaccinated cases will be of central importance. Means should exist for central collation and standardized testing of case isolates, including virtually all viable isolates. The objective will be to distinguish between potentially preventable strains (via the MATS assay) and non-preventable strains (not matching the vaccine antigens). Ready access to the MATS assay will be necessary for evaluation of VE. Ideally, each country using the vaccine would have this assay available in a central reference laboratory.

A significant proportion of MenB disease cases may be identified only by PCR assay, a circumstance that will increase the difficulty of discriminating between preventable and nonpreventable cases as isolates are not available for analysis. For PCR-diagnosed cases, additional investigation may be required that may include collection of throat swabs and standardized testing of isolates from these swabs in a central reference laboratory. Direct sequencing of MenB vaccine antigen genes from the clinical specimen of such cases may provide some information. This is current practice in the UK and should be considered by other countries.

Cases occurring in vaccinated persons are expected to result from strains not covered by the 4CMenB vaccine but typing might also reveal differential VE based on which or how many vaccine antigens are expressed by isolates. Host factors should be considered in the investigation of failures.

A relevant additional component of VE estimation will be observing for changes in the incidence rates of other meningococcal serogroups. Since meningococci in other serogroups possess, to some extent, the antigens in 4CMenB vaccine, disease surveillance should be capable of recognizing reduced rates of disease in the target population resulting from non-B serogroups. Reductions will be most readily detected when the target population for 4CMenB vaccine is not also receiving polyvalent meningococcal vaccine (ACYW₁₃₅). Where C conjugate vaccine is also offered, reduction in the incidence of serogroup Y cases might serve to signal crossprotection from 4CMenB vaccine.

What effect will the 4CMenB vaccine (& other MenB vaccines) have on the evolution of N. meningiditis?

As with VE, monitoring the evolution of the target bacterial population will be important. Isolates from vaccinated cases should not only be tested in the MATS assay but also be used to refine the assay cutoffs. The assay was constructed to predict the likelihood of protection, whereas nonprevented case isolates will better inform the discrimination criteria. The ideal approach to this would be a multinational collaboration, with many relevant isolates fed into a single MATS assay location. Ongoing testing of isolates from vaccinated and unvaccinated cases will be important for determining if vaccine use is associated with shifts in the prevalent strains and decreasing strain coverage.

How safe are recombinant MenB vaccines?

The safety of the vaccine will be closely linked to its acceptability and will influence uptake of the vaccine. Active safety surveillance in a multinational collaboration using standardized definitions for adverse events (e.g., Brighton Collaboration Case Definitions) would be the optimal method to monitor the safety of this new vaccine [102]. While fever and the potential for febrile seizure may be expected with this vaccine, other unknown rare events could occur. To establish a threshold for unusual events, background rates for expected events (e.g., seizures) in the immunized age groups should be calculated prior to the start of immunization programs. Therefore, the first countries to use this vaccine should consider enhanced safety surveillance to provide early recognition of any adverse events following immunization. Canada will be well placed with its active surveillance of adverse events through IMPACT, a national surveillance network in 12 tertiary care pediatric hospitals across Canada [2,87,88]. With 21 years of adverse event surveillance data, IMPACT can contribute to estimates for expected events such as febrile seizures. Timely signal investigation that entails the comparison of events in immunized versus unimmunized persons should occur. In jurisdictions with linked health records, such as the province of Manitoba, the self-controlled case series is one option for signal investigation [89].

How acceptable will the vaccine be to Canadian parents & how well will Canadian infants tolerate it?

Parental acceptance of a new vaccine varies greatly by country and even by region and it is influenced by their perception of the risk of the disease and benefits of the vaccine balanced by their concern about vaccine side effects. Meningococcal vaccines have historically been well accepted by the population, particularly in the face of meningococcal outbreaks. Given widespread media attention to all cases of meningococcal disease in a community, specific knowledge, attitudes and beliefs research will be required to gauge the public’s desire for and/or willingness to be part of a universal MenB vaccination program. This research will need to dissect the attitude of the public towards the disease, risk of fever and febrile seizure, as well as the use of acetaminophen prophylaxis in all or high-risk infants. Effective risk communication will be important for acceptance by both parents and health professionals. Given MenB IMD is rare in Canada, effective translation of its risk into terminology the average parent can understand will be necessary. For example, risk estimates may be more accessible if expressed as cases per week rather than as rates per 100,000 population.

Five-year view

The 4CMenB vaccine may soon be approved for use in Canada and will target the leading cause of IMD. The burden of disease from MenB is similar to that seen with MenC when universal immunization programs were implemented. However, the novel technology used in the vaccine as well as the unusually long list of unanswered questions identified above will weigh heavily on immunization advisory committees and decision makers. Who will want to be first to recommend the use of a vaccine of uncertain and possibly short-lived effectiveness, requiring multiple, separate injections associated with increased rates of adverse effects? Who will want to commit to a new program that might require unanticipated additional booster doses to sustain protection through early childhood and beyond? Such decisions will be easiest in the face of outbreaks and epidemics, so it will be helpful if situational uses are well studied to add information about safety and effectiveness. However, the most informative situation would be a national ‘demonstration project’, implemented with appropriate concurrent research on key issues. This would be similar to the UK’s characterization of MenC vaccines. With effectiveness measured, dosing requirements better defined, safety more fully assessed and parental acceptance demonstrated, other countries will find it much easier to implement programs. Without such additional evaluations, it could be anticipated that the 4CMenB vaccine may be recommended in Canada by the national advisory committee but not funded by provincial public programs until additional characterization of vaccine performance in routine use is available.

Table 1. Provincial and territorial meningococcal serogroup C and quadravalent immunization programs.

Province/territory	Infants/children (age in months)	Year started (MenC)	Children/adolescents (age in years)	Year started
British Columbia^†	2, 12	2003	11–12	2003 MenC
Alberta^‡	2, 4, 12	2002	14–16	2010 ACYW135
Saskatchewan	12	2004	11–12	2004 MenC 2011 ACYW135
Manitoba	12	2009	10–11	2005 MenC
Ontario	12	2004	12–14	2005 MenC 2009 ACYW135
Quebec^§	12	2001
New Brunswick	12	2004	14–16	2004 MenC 2007 ACYW135
Prince Edward Island	12	2005	14–16	2005 MenC 2006 ACYW135
Nova Scotia	12	2005	14–16	2005 MenC
Newfoundland	12	2005	10–11	2005 MenC 2009 ACYW135
Northwest Territories^¶	2, 12	2004	14–16	2007 MenC
Nunavut	12	2007	14–16	2007 MenC
Yukon^#	2, 12	2005	14–16	2005 MenC

The following vaccines are authorized for use in Canada and all have been used in provincial and territorial programs: Meningitec^® (Pfizer, NY, USA), Menjugate^® (Novartis, MA, USA), NeisVac-C^® (Baxter, IL, USA), Menactra^® (Sanofi Pasteur, Lyon, France) and Menveo^® (Novartis).

^†British Columbia switched from a one-dose program at 12 months to a two-dose program at 2 and 12 months in 2005. Catch-up program with MenC was offered for adolescents 14–18 years of age from 2004 to 2007.

^‡Alberta switched from a three-dose program at 2, 4 and 6 months to 2, 4 and 12 months in 2007.

^§Quebec mass immunization of 2-month to 20-year-olds in 2001.

^¶The Northwest Territories switched from a two-dose program at 2 and 4 months to a two-dose program at 2 and 12 months in 2007. ACYW₁₃₅ is provided to postsecondary students attending school outside of the Northwest Territories.

^#Yukon switched from a two-dose program at 2 and 6 months to a two-dose program at 2 and 12 months in 2007.

ACYW₁₃₅: Quadravalent serogroups ACYW₁₃₅ conjugate vaccine; MenC: Monovalent conjugate serogroup C vaccine.

Table 2. Terms used to describe typing of Neisseria meningitides.

Typing term	Method of typing	Antigen(s) involved	Ref.
Serogroup	Bacterial agglutination, PCR	Capsular polysaccharide	[90]
Serotype	Indirect whole-cell ELISA	Class 2 or 3 outer membrane protein (PorB)	[91]
Serosubtype	Indirect whole-cell ELISA	Class 1 outer membrane protein (PorA)	[91]
ET	MLEE; starch gel electrophoresis, relative mobility of housekeeping enzymes	Housekeeping enzymes	[92]
ST	MLST	DNA sequencing of housekeeping enzyme genes	[93]
CC	MLEE or MLST	Groups of ET or STs shown to be related by, for example, sharing (being identical) the majority of their housekeeping enzymes or genes	[12]
Expression and seroreactivity of 4CMenB vaccine antigens	MATS	fHbp, NHBA, NadA	[62]

CC: Clonal complex; ET: Electrophoretic type; MATS: Meningococcal antigen typing system; MLEE: Multilocus enzyme electrophoresis; MLST: Multilocus sequence typing; ST: Sequence type.

Key issues

• • Neisseria meningitidis serogroup B (MenB) invasive meningococcal disease is a rare disease in Canada. MenB vaccines are unique in terms of their potential strain coverage and adverse event profile.

• • Meningococcal vaccine implementation can be influenced by public and political forces.

• • Given these issues, a carefully planned implementation and evaluation strategy will be required before the use of any new serogroup B vaccines in Canada.

• • The following are key areas for consideration and further research in Canada:

  • – Effectiveness of MenB vaccines, especially in infants;

  • – Effect of MenB vaccines on the evolution of the Neisseria meningiditis bacterium;

  • – Safety of MenB vaccines;

  • – Parental acceptability and infant tolerability.

Financial & competing interests disclosure

JA Bettinger is supported by a Career Investigator Award from the Michael Smith Foundation for Health Research. JA Bettinger was also involved with ad-hoc Advisory Boards in 2010 (Novartis Vaccines, Canada) and speaker honoraria in 2010 and 2011 (Novartis Vaccines, Pfizer Inc., Baxter Inc.). SL Deeks is an Ontario provincial government employee. SA Halperin was involved with ad-hoc Advisory Board for Novartis Vaccines, Canada and speaker honoraria in the past year (Novartis Vaccines). R Tsang is a Public Health Agency of Canada employee. DW Scheifele was involved with ad-hoc Advisory Board for Novartis Vaccines, Canada. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.