Abstract
Neisseria meningitidis is a major cause of meningitis and septicemia globally. Vaccines directed against N. meningitidis serogroup B (MenB) have been used to control sporadic and sustained disease in industrialized and non-industrialized countries. Early outer membrane vesicle (OMV) vaccines effectively reduced MenB disease in countries such as Norway, New Zealand, and France; however, these vaccines were highly specific for their targeted outbreak strain, did not elicit a durable immune response, and were ineffective for widespread use due to the diversity of MenB-disease-causing isolates. Recently developed recombinant protein-based MenB vaccines that target conserved surface proteins have the potential to induce a broader immune response against the diversity of disease-causing strains. Given the deleterious consequences and sporadic nature of MenB disease, the use of optimal vaccination strategies is crucial for prevention. Reactive vaccination strategies used in the past have significant limitations, including delayed implementation, substantial use of resources, and time constraints. The broad coverage potential of recombinant protein-based MenB vaccines suggests that routine use could result in a reduced burden of disease. Despite this, routine use of MenB vaccines is currently limited in practice.
Keywords
Introduction
Invasive meningococcal disease (IMD) resulting from infection with Neisseria meningitidis most often presents in clinical cases as meningitis or septicemia. Though disease incidence is currently low (estimated at 0.16 cases per 100,000 individuals in the United States 1 and 0.7 per 100,000 individuals in Europe 2 ) each case is life threatening.
Meningococci are transmitted by aerosol droplets and can colonize the nasopharynx without the appearance of symptoms, a process known as carriage. However, in a small percentage of cases, bacteria enter the bloodstream and cause IMD. 3 Symptoms of IMD include headache, stiff neck, fever, nausea, vomiting, photophobia, and altered mental status. 3 Severe illness can ensue within 15 to 24 hours, 4 with fatality occurring in 10–15% of patients, even among those receiving appropriate treatment. 3 Recently, a subset of IMD cases caused by serogroup W has been characterized by an atypical gastrointestinal presentation that includes nausea, vomiting, and diarrhea; most of these cases occurred in adolescents and were associated with a high mortality rate (33% within 24 h of hospitalization). 5 Globally, N. meningitidis causes disease in industrialized and non-industrialized countries and is responsible for low-level endemic disease, as well as outbreaks and epidemics (Figure 1).6–12

Global prevalence of Neisseria meningitidis serogroup B invasive meningococcal disease and incidence rates of selected outbreaks.
Across all ages, infants and young children have the highest incidence of IMD, with a second peak in incidence occurring among adolescents and young adults, and a third peak in adults ⩾65 years of age. 13 However, acquisition of N. meningitidis does not always lead to disease, and a state of carriage occurs when the bacterium asymptomatically colonizes the upper respiratory tract. 14 Adolescents have the highest incidence of carriage among all age groups in many countries, 14 likely due to the propensity for high-risk behaviors, such as living in dormitories and other activities that involve close personal contact.15,16 Decreasing meningococcal carriage rates in adolescents will likely decrease subsequent transmission in all age groups, including other high-risk age groups such as infants and the elderly. 14 Therefore, the most effective strategies to control meningococcal infection would target adolescents because this group has the highest propensity to transmit the bacterium.
N. meningitidis is categorized into 12 serogroups according to the biochemical composition of the capsular polysaccharide; however, only 6 serogroups (A, B, C, W, X, and Y) are responsible for the majority of disease. 17 Polysaccharide-based vaccines targeting serogroups A, C, W, and Y have been used in both monovalent and polyvalent forms to successfully prevent disease due to these serogroups. In contrast, development of a polysaccharide meningococcal serogroup B (MenB) vaccine is not feasible due to the structural similarity of the serogroup B polysaccharide to α(2–8)-linked N-acetyl-neuraminic acid on human neuronal cells. 18 The reduced immunogenicity of the serogroup B capsular polysaccharide, coupled with concerns associated with immunizing with a self antigen necessitated alternative strategies for developing MenB vaccines. These strategies include the development of highly strain-specific outer membrane vesicle (OMV)-based vaccines (discussed in detail below) and, more recently, recombinant protein vaccines that elicit a broadly protective immune response across diverse MenB strains.
Two novel protein-based vaccines have been developed and are currently in use. MenB-4C (Bexsero®, 4CMenB; GlaxoSmithKline Vaccines, Siena, Italy) 19 includes three recombinant bacterial proteins, including the bacterial lipoprotein termed factor H binding protein (FHbp; subfamily B only), neisserial adhesin A, neisserial heparin-binding antigen, as well as an OMV derived from an epidemic MenB strain from New Zealand. 20 MenB-4C is licensed in Europe, Canada, and Australia in individuals ⩾2 months of age and was approved in the United States in individuals 10 to 25 years of age in 2015. 19 MenB-FHbp (Trumenba®, Bivalent rLP2086; Pfizer Inc., Philadelphia, PA) 21 includes both subfamilies (subfamily A and subfamily B) of the surface-expressed FHbp, which is expressed as one of these subfamilies on virtually all invasive MenB isolates. 22 In 2017, MenB-FHbp received approval in the United States and Canada in individuals 10 to 25 years of age23,24 and in the European Union, 25 Iceland, Liechtenstein, Norway,25,26 and Australia 27 in individuals 10 years of age and older.
Meningococcal B vaccines: modalities for use
Historically, the first nonpolysaccharide-based MenB vaccines used in outbreak control were protein-based OMV vaccines isolated from the strain causing the majority of disease. 28 The immune response to OMV vaccines is largely directed against porin A (PorA) that is highly heterogeneous across different N. meningitidis MenB strains. Therefore, an OMV vaccine provides protection against only those MenB strains that express the same PorA variant and would have little utility against epidemics or outbreaks caused by divergent strains.29,30 As such, OMV vaccines were developed primarily in response to regional epidemics, including those in Norway, 31 New Zealand, 7 and France. 8 The newer recombinant protein MenB vaccines, which are designed to have a broader activity against diverse strains, are discussed in relation to use in response to outbreaks in Quebec, Canada, 9 select university campuses in the United States, 10 as well as use outside of outbreaks as part of routine immunization.
Use of Neisseria meningitidis serogroup B vaccines in epidemics
Use of MenBvac® in Norway
Beginning in 1974, Norway experienced an increase in MenB incidence, the majority of which was due to the emergence of a single, hypervirulent strain. 32 In response to the epidemic, the Norwegian National Institute of Public Health developed a novel OMV vaccine (MenBvac®, Norwegian Institute of Public Health, Oslo, Norway) that targeted the outbreak strain. To assess the efficacy of the vaccine, a placebo-controlled study was conducted in 171,800 students enrolled in 1335 secondary schools in Norway, with approximately 50% of students receiving vaccine and 50% receiving placebo. Cases of MenB were assessed from 2 weeks after the second dose and students were followed for up to 29 months. Vaccine efficacy was determined to be 57%. A subsequent study conducted in adolescents who received a three-dose primary series found that a booster dose administered 10 months after dose three induced a protective immune response in 93% of individuals at 6 weeks postvaccination and in 64% after 1 year. 33 MenBvac® effectiveness was also demonstrated during an outbreak in Normandy, France, caused by a strain with a high degree of homology to the Norwegian strain targeted by MenBvac®. 8
Use of MenBvac® in France
Generally, the incidence of IMD in France is low, with an annual incidence rate of 0.7 to 1.6 cases per 100,000 person-years; 8 however, in 2003, an outbreak occurred in Normandy that yielded an annual incidence rate of 2 cases per 100,000 person-years in the Seine-Maritime region (Table 1). 8 The strain responsible for the outbreak (B:14:P1.7,16/ST-32) had been endemic in France since the 1990s. During the outbreak, Dieppe, the city at the epicenter of the outbreak, accounted for 60% of cases in the region. Although no specific OMV vaccine was in existence to combat the outbreak strain, a genetically related OMV vaccine developed 20 years prior in Norway (MenBvac®) was similar and considered to match closely enough to warrant use. In 2006, a vaccination campaign was implemented that initially targeted individuals 1 to 5 years of age using a 3+1 schedule. 8 In 2008, infants <1 year of age were included using the same schedule. Also in 2008, an amended 2+1 dosing schedule was implemented for individuals 1 to 19 years of age due to an inadequate supply of the vaccine. In the 3 years before the start of the immunization campaign, 80% of typeable meningococcal isolates were MenB, and of these, 66% had the genotype of the outbreak strain. Young children and teenagers had the highest burden of disease. While the vaccination campaign initially targeted children 1 to 5 years of age residing in the Dieppe, vaccination was expanded to include children aged 1 to 5 years outside of Dieppe, and finally to include those aged 6 to 19 years within Dieppe. In total, >26,000 individuals were vaccinated by March 2010 and, of these, 75% received at least two doses, and 64% received a booster dose. 8
Summary of select Neisseria meningitidis serogroup B epidemics and outbreaks.
MeNZB administered in a three-dose series until June 2006. A booster dose was added in January 2006 for infants aged 10 months who were aged <6 months at the start of the primary series.
Vaccine schedule change due to shortage of vaccine and did not apply for those aged <1 year.
⩾2 months between doses.
Included students, faculty, and staff.
Percentage of students receiving MenB-FHbp, which was the primary vaccine administered.
Attack rate (frequency of infection).
MenB, Neisseria meningitidis serogroup B; NA, not available.
Between July 2008 and August 2010, three cases of IMD due to the outbreak strain were detected in the epicenter among individuals aged <20 years. This was an 83% reduction among the total population in the epicenter; a 23% reduction was observed among the unvaccinated population in the region. 8 By 2012, the incidence of MenB disease was considered relatively stable in France compared with previous years; however, localized regional outbreaks remained a concern. 34 In 2014, the French Haute Autorité de Santé outlined the need for a vaccine targeting diverse MenB strains and recommended vaccination for high-risk individuals and those living in areas experiencing outbreaks or clusters of IMD. 34
Use of the MeNZB vaccine in New Zealand
An OMV vaccine (MeNZB®, Chiron/New Zealand Ministry of Health, Wellington, New Zealand) was developed in response to an epidemic that occurred in New Zealand in the early 1990s that primarily affected individuals aged <20 years (Table 1).7–11,35 In 2001, which was the peak of the epidemic, the annual IMD incidence in New Zealand was 24.7 cases per 100,000 population. 7 Routine vaccination with MeNZB began in 2004 for children and young adults and continued until 2008, but those considered to be at high risk for IMD continued to be vaccinated until 2011. 36 MeNZB was administered in a three-dose schedule, with a booster dose administered to children who received their first dose before age 6 months.37–39 However, the immune response to MeNZB waned quickly, particularly among infants. At 7 months postvaccination for those receiving the three-dose infant series only, approximately 30% demonstrated protective antibody titers and, as such, booster dosing was required to achieve immunity. 35 The immune response was slightly better in older children (aged 8 to 12 years), among whom 36% had protective immunity at 14 months after the three-dose primary series. 35 Vaccination efforts could not overcome the lack of a durable immune response. As a result, in 2013, 53% of all meningococcal infections in New Zealand were due to MenB, with 40% of the MenB cases due to the strain targeted by MeNZB. 40
Use of the MenB-4C vaccine in Quebec
Beginning in 2003, IMD due to MenB increased in the Canadian province of Quebec (Table 1). Between 2006 and 2013, the average incidence rate of IMD reached 3.4 cases per 100,000 person-years in the Saguenay–Lac-Saint-Jean region and was five times the average rate in the province. 9 IMD incidence was highest in infants aged <1 year and children and adolescents 1 to 14 years of age who were economically deprived. 41 This prolonged outbreak was due to a hypervirulent MenB strain (ST269) that, although not genetically identical, shared enough sequence homology with the FHbp antigen in MenB-4C to predict crossreactivity. A short-term mass vaccination campaign targeting individuals aged 2 months to 20 years was initiated in May 2014 in the Saguenay–Lac-Saint-Jean region. Infants aged 2 to 5 months received a four-dose schedule; those aged 6 to 11 months received a three-dose schedule; and children aged ⩾1 year received a two-dose schedule. All doses were administered at least 2 months apart, and approximately 82% of the target population was vaccinated. After implementation of the mass immunization campaign, the incidence rate of IMD in the epicenter region decreased and reached undetectable levels in 2015 and 2016. Moreover, IMD due to MenB has not been detected in vaccinees for >2 years postvaccination or in unvaccinated individuals in the target population (defined by age or geographic location). In other regions proximal to Quebec but outside of the epicenter, MenB incidence decreased by approximately 50% across all ages. 9
Use of MenB-FHbp/MenB-4C vaccines in the United States
The overall incidence of IMD is low in the United States, with an annual incidence of 0.11/100,000 individuals in 2015. 42 However, outbreaks of MenB disease are an ongoing concern on college and university campuses due to factors such as crowded living conditions and behaviors that increase the risk of transmission such as kissing, smoking, and sharing of personal items. 43 Recently, MenB has been the cause of several outbreaks on US campuses, including nine cases at Princeton University in New Jersey in 2013–2014 with an additional case at Drexel University in Pennsylvania; five cases at the University of California at Santa Barbara beginning in November of 2013; seven cases at the University of Oregon and two cases at Providence College in Rhode Island in 2015; three cases each at Rutgers University in New Jersey, Santa Clara University in California, and the University of Wisconsin in 2016; and six cases at Oregon State University in 2016–2017.44,45 In all outbreaks, the universities responded with mass immunization clinics for students, faculty, and staff; however, vaccination was not universally required. Some institutions chose an optional ‘opt-in’ strategy, resulting in an overall lower vaccination rate compared with an opt-out strategy (i.e. requiring vaccination for all students). 10 A clear example of the discordance in the success of different vaccination strategies employed during campus-based outbreaks is the vaccine uptake achieved during the MenB outbreaks in 2015 at the University of Oregon and Providence College in Rhode Island. The University of Oregon, with seven cases in total, chose an ‘opt-in’ strategy; Providence College, with two cases in total, chose an ‘opt-out’ policy.
At Providence College, an outbreak was declared by the Rhode Island Department of Health after two cases of MenB were confirmed in two students within 4 days (Table 1). 46 Among the 3745 individuals eligible for vaccination, 3061 participants were vaccinated within 48 h and an additional 464 participants were vaccinated within 1 week at makeup clinics. Vaccine coverage was 94%, 47 and no additional cases were reported. At the University of Oregon, MenB was identified in mid-January 2015 (Table 1). 48 Of the seven cases reported, one was a fatality and the seventh occurred in a parent visiting the campus. The University of Oregon responded with a series of opt-in vaccination clinics in which MenB-FHbp was available through the university; MenB-FHbp and MenB-4C were both available at off-campus pharmacies. 10 Approximately 22,000 students, faculty, and staff were eligible for vaccination during the first mass vaccination clinic in March, and of these, it is estimated that 46% received dose one and 29% received dose two. These two outbreaks underscore the importance of having a vaccination strategy in place to mobilize students quickly and maximize vaccine uptake during an outbreak.
Use of Neisseria meningitidis serogroup B vaccines outside of outbreaks
Currently, MenB vaccines are used in high-risk groups in many countries and are now starting to be considered for use as a prophylactic versus reactive strategy for the prevention of IMD. Examples below consider the prophylactic use of MenB vaccines in the United Kingdom and Italy and current recommendations in the United States, Canada, Australia, and parts of Europe.
The overall incidence of IMD in the United Kingdom is low, with approximately 935 cases diagnosed in 2015; 49 however, MenB was responsible for a disproportionately high rate of disease compared with other disease-causing meningococcal serogroups from 2002 to 2012. 50 In 2014, due to the predominance of MenB IMD and based on recommendations from the Joint Committee on Vaccination and Immunization, the United Kingdom became the first country to include MenB-4C in the national childhood immunization schedule. The vaccine was subsidized by the government and offered free of charge. MenB-4C was administered using a 2- and 4-month vaccination schedule with a booster dose administered at age 12 to 13 months. 50 This vaccination strategy represents the first prophylactic use of a MenB vaccine to protect a target population in the absence of an outbreak. Vaccine effectiveness was estimated to be 83% for infants aged <1 year and, compared with the 4 previous years, MenB cases were reduced by 50% during the first year of the program. A reduction in MenB incidence was also observed in nonvaccinated individuals (14%). After adjusting for the decreasing trend, the overall net reduction in MenB during the first year for those receiving vaccine was 42%. 51
In 2016 in Ireland, the Heath Service Executive amended the routine infant vaccination schedule to include MenB for infants born on or after 1 October 2016; MenB is administered as a two-dose infant series at 2 and 4 months of age with a booster dose at 12 months. 52 In 2017, Italy expanded the routine infant vaccine program to include MenB administered as a three-dose infant series at 3, 4, and 6 months of age with a booster dose at 13 months.53,54 Studies reporting vaccine effectiveness and MenB incidence in response to these vaccination programs are expected to follow. In addition to the abovementioned countries, Austria and the Czech Republic both recommend MenB use for infants, but the vaccine is not part of a government-supported routine infant immunization program. 53
Clinical recommendations for the use of Neisseria meningitidis serogroup B vaccines in adolescents in the United States, Canada, Australia, and the Czech Republic
Because MenB outbreaks are unpredictable and adolescents have a high incidence rate, prophylactic vaccination is considered the best strategy for preventing disease in this population. In the United States, the Advisory Committee on Immunization Practices issued a category B recommendation for MenB for the immunization of adolescents and young adults 16 to 23 years of age, meaning that vaccination may be considered for individuals based on consultation with a healthcare provider. 55 Although the recommended age for vaccination is 16 to 23 years, the preferred age for vaccination is 16 to 18 years 55 to establish immunity before the observed peak in IMD incidence occurs among those aged 18 to 22 years. 42 MenB vaccination is also recommended for individuals aged ⩾10 years who have an increased risk, including those with deficiencies in the complement pathway, those with functional or anatomic asplenia, and those likely to be exposed during outbreaks or work-related activities. 56 MenB vaccination is recommended in both two-dose (MenB-FHbp, MenB-4C) and three-dose (MenB-FHbp) schedules. 57
MenB immunization is recommended on an individual basis for children aged ⩾2 months 58 and for adolescents aged 15 to 19 years in Australia. 59 In the Czech Republic, MenB vaccination is recommended for those aged 12 to 19 years. 60
Neisseria meningitidis serogroup B carriage studies
Data regarding the extent to which vaccination may affect transmission of meningococci are important to policy decisions regarding routine MenB vaccination in adolescents. Studies have been performed or are under way to determine the ability of MenB vaccines to reduce acquisition and carriage of meningococci. Specifically, in a study of MenB-4C conducted in university students in England, MenB carriage was not significantly reduced at 1 month postvaccination; however, at 3 months postvaccination, carriage prevalence among MenB-4C recipients significantly decreased for all N. meningitidis isolates collectively and for some non-MenB isolates, though not for MenB isolates. 61 The broad activity of MenB-4C in reducing carriage of non-MenB isolates likely resulted from crossreacting bacterial protein components of the vaccine. 61 Another large meningococcal carriage study is currently being conducted as part of a phase IV, randomized, double-blind study in 45,000 Australian students aged ⩾14 years vaccinated with two doses of MenB-4C. This study will examine carriage prevalence of all serogroups, including MenB, within 1 year of vaccination, and is expected to close in June 2019. A smaller carriage study is being conducted at the University of Oregon in conjunction with a mass vaccination campaign using MenB-FHbp and MenB-4C following the 2015 outbreak on that campus. 48 The information provided by these studies pertaining to the ability of MenB vaccines to interrupt transmission through the reduced acquisition of carriage is expected to have implications on future vaccine recommendations.
Cost considerations
The introduction of MenB vaccines into national routine immunization schedules requires careful consideration of a number of parameters, including, but not limited to, the incidence of IMD within a given country, vaccine efficacy, real-world effectiveness, anticipated compliance, vaccine incorporation into existing immunization schedules, and cost.
In the United Kingdom, initial cost-effectiveness studies were conducted before MenB-4C was adopted into routine infant schedules in 2014. 50 These analyses determined that routine infant MenB vaccination would be cost effective with or without catch-up vaccination and would be expected to reduce MenB cases by 71% after 10 years of use. 62 This observation was later confirmed using updated cost-effectiveness models (assuming a cost ranging from £3 to £22 per dose, depending on the scenario) for a vaccine administered as a three-dose infant series plus toddler dose at 12 months; cost effectiveness was also demonstrated for a one-dose adolescent schedule at age 13 years and a combination of both infant and adolescent immunization. 63 This updated modeling noted that adolescent vaccination as a sole strategy, although economical, would depend on reducing transmission (through carriage reduction), and without an infant program, would require a significant amount of time to have a measurable impact on IMD incidence. 63
In 2017, Italy adopted a routine infant MenB vaccination program 54 despite different cost-modeling analyses that reported conflicting results. For example, one model evaluated MenB-4C administered as a three-dose infant series (at 2, 4, and 6 months of age) followed by a toddler dose at 12 months and a booster at 11 years of age. 64 In a scenario that evaluated the official IMD incidence as part of the cost-effectiveness calculation, vaccination was not considered cost effective; however, a second scenario that replaced official incidence rate with estimated rates (which were higher) was determined to be cost effective and routine vaccination was considered advisable. Of note, IMD is considered to be greatly underestimated in Italy (due to screening methods, culture versus molecular typing, and use of antibiotics prior to screening). A second study conducted in Italy employed a model that evaluated IMD incidence based on national surveillance notification and hospital discharge records from two regions (Piemonte and Lombardia). 65 Using this model, which assumed a three-dose infant series and toddler dose, estimated vaccine costs, strain coverage, and vaccine efficacy, routine MenB vaccination was not considered cost effective.
Cost-effectiveness studies similar to those conducted in the United Kingdom have been adapted for Germany, but given the low incidence of IMD and high associated cost, routine vaccination has not been adopted. 66 Similarly, cost-effectiveness studies conducted in the Netherlands 67 and France 68 have not favored routine MenB immunization from a cost standpoint.
Notably, cost-effectiveness predictions often take into account multiple key factors, including IMD incidence in a given area. In fact, as was evident in Italy, low IMD incidence is often a critical parameter in determining the cost effectiveness of a vaccine program 65 and has likely been a precluding factor for routine immunization in many of the countries noted above. However, current low overall incidence does not take into account the unpredictable nature of MenB outbreaks, historical changes in MenB epidemiology, or other factors such as carriage reduction and MenB outbreak management that may also provide additional and potentially more useful data for cost-effectiveness estimates. In particular, carriage reduction after MenB vaccination is currently being assessed in a large study of 4C-MenB in Australia. 69 Moreover, recent mass vaccination programs in response to outbreaks at universities and colleges in the United States 45 are likely associated with a significant financial impact and resource expenditure. Therefore, cost-effectiveness analyses based on current incidence rates alone may not be an ideal indicator when considering the implementation of routine MenB immunization programs.
Discussion and conclusion
The recent availability of recombinant, broadly protective MenB vaccines offers the opportunity to move beyond the traditional methods of disease control that have centered on reactive strategies (i.e. large-scale immunization after the occurrence of an outbreak and targeted vaccination in an area where MenB was hyperendemic). Although past vaccination efforts had reasonable efficacy at quelling MenB outbreaks like those in New Zealand, 7 France, 8 Canada, 9 and on US university campuses,10,11 this strategy does not prevent the emergence of new outbreaks. Proactive strategies that can help reduce the overall incidence of disease through preventive versus reactive management of meningococcal infection are optimal.
MenB outbreaks have been reported in countries and regions with both high rates of endemic disease, such as New Zealand, 7 Normandy, 8 and Quebec, 9 and low rates of disease, such as the United States.10,11 Regardless of the country of origin, meningococcal outbreaks are costly in terms of financial resource expenditures and resources required for management. Reactive vaccination strategies require extensive logistics for vaccine identification and acquisition, implementation of immunization clinics, and partnerships with vaccine manufacturers. These efforts require extensive support from many groups, including governments (federal, regional, and state), local health experts, and pharmaceutical partners, under the onus that all logistics must be developed, coordinated, and executed in a short period of time, therefore taking away resources from other routine tasks. Preventive disease management could, in many cases, eliminate this type of emergency response. Cost analyses that will inform efforts to help balance the cost of vaccination against the cost of disease are needed.
MenB vaccines have been included in routine infant immunization programs in the United Kingdom, Italy, and Ireland, 53 and data are forthcoming regarding the effect of the vaccine on overall disease incidence within this and other countries. In the near future, more information on the effectiveness of MenB vaccines will become available, as will additional safety data. The breadth of coverage that these vaccines will provide against disease caused by the diversity of circulating strains and their effect on carriage reduction will be determined. These data will potentially support the widespread use of MenB vaccines to control endemic disease.
Footnotes
Acknowledgements
All authors have contributed to the development of this manuscript. Editorial and medical writing assistance was provided by Susan E DeRocco, PhD, of Complete Healthcare Communications, LLC (West Chester, PA), a CHC Group company, and was funded by Pfizer Inc.
Funding
This manuscript was funded by Pfizer Inc.
Conflict of interest statement
Drs Balmer and York are employees of Pfizer Inc and may hold stocks and/or stock options.
