Pneumococcal vaccination: what have we learnt so far and what can we expect in the future?

Individuals <2 years and ≥50 years of age, as well as those with other specific risk factors, are especially vulnerable to invasive pneumococcal disease (IPD). Conjugate vaccines have been developed against encapsulated bacteria such as Streptococcus pneumoniae to provide improved immune responses. The 7-valent pneumococcal conjugate vaccine (PCV7) has significantly reduced the burden of vaccine-type pneumococcal diseases in children, including invasive disease and pneumonia and acute otitis media. There have also been significant declines in antimicrobial resistance in 7-valent vaccine serotypes and carriage of S. pneumoniae in the post-PCV7 era. Two to three years after the introduction of PCV13, there is emerging, global evidence of a reduced burden of pneumococcal diseases in children, including declines in IPD (UK and Germany) and nasopharyngeal carriage of PCV13 serotypes (Portugal and France). The functional immunogenicity of PCV13 in individuals ≥50 years of age has been demonstrated in clinical trials in comparison with the 23-valent pneumococcal polysaccharide vaccine and for children and adults 6 to 49 years of age. Between 2011 and 2013, PCV13 received market authorisation by the European Medicines Agency (EMA) for these additional age groups and is now available in Europe for the prevention of pneumococcal disease in all age groups.

Abstract Individuals <2 years and ≥50 years of age, as well as those with other specific risk factors, are especially vulnerable to invasive pneumococcal disease (IPD). Conjugate vaccines have been developed against encapsulated bacteria such as Streptococcus pneumoniae to provide improved immune responses. The 7-valent pneumococcal conjugate vaccine (PCV7) has significantly reduced the burden of vaccine-type pneumococcal diseases in children, including invasive disease and pneumonia and acute otitis media. There have also been significant declines in antimicrobial resistance in 7-valent vaccine serotypes and carriage of S. pneumoniae in the post-PCV7 era. Two to three years after the introduction of PCV13, there is emerging, global evidence of a reduced burden of pneumococcal diseases in children, including declines in IPD (UK and Germany) and nasopharyngeal carriage of PCV13 serotypes (Portugal and France). The functional immunogenicity of PCV13 in individuals ≥50 years of age has been demonstrated in clinical trials in comparison with the 23valent pneumococcal polysaccharide vaccine and for children and adults 6 to 49 years of age. Between 2011 and 2013, PCV13 received market authorisation by the European Medicines Agency (EMA) for these additional age groups and is now available in Europe for the prevention of pneumococcal disease in all age groups.

Introduction: the burden of pneumococcal disease
The aim of this article is to discuss what could be expected from 13-valent pneumococcal conjugate vaccine (PCV13) use in the population over 5 years of age based current knowledge from the post-marketing experience of childhood vaccination with PCV13 and from pivotal trials performed in children, adolescents and adults. We will also review the characteristics of the immune response to conjugate vaccines.
The burden of invasive pneumococcal disease (IPD) is high, especially in individuals <2 years and ≥65 years of age who have the highest incidence and case fatality rate (CFR) [1]. The overall incidence of IPD in Europe in 2010 was 5.2 cases per 100,000 population, with the most affected age groups being <1 year and ≥65 years old (18.5 and 15.6 cases per 100,000, respectively) [2]. In the USA, CFRs for IPD in adults have not changed significantly over the last several decades [3]. Mortality rates due to IPD have remained high despite the availability of clear guidelines for treatment, with a number of highly effective antibiotics [4].
The clinical presentation of IPD varies according to age. Bacteraemia without source or focus of infection (i.e. nonmeningitis, non-pneumonia IPD) is the most common presentation in children aged 0-5 years, with an estimated global incidence of 87 cases per 100,000 versus 17 cases per 100,000 for pneumococcal meningitis [5]. In the Netherlands, pneumococcal pneumonia was the most frequent presentation in adults during 2004-2006 [6]. Similar findings were observed in a Spanish hospital-based surveillance study (the ODIN study), in which pneumonia accounted for 70 % of IPD cases in adults older than 18 years [7]. Within this study, CFRs were also found to vary according to clinical presentation, with sepsis being associated with the highest rate (40 %), followed by peritonitis (20 %) and complicated pneumonia (18 %).
Certain comorbidities, including chronic heart, hepatic or pulmonary disease, diabetes mellitus, acquired immunodeficiency syndrome (AIDS) or other immunosuppressions, increase the risk of pneumococcal diseases in both children and adults [8]. A retrospective analysis of 22,000 IPD cases in England (2002-2009) reported a higher risk of IPD-related hospitalisation and mortality in individuals with comorbidities versus those without comorbidities in all age groups. The effect of underlying conditions on the risk of hospitalisation for IPD was highest in children aged 2-15 years, with a nearly 12-fold increase in IPD in those with comorbidities compared with those without (versus nearly 8-fold and 3-fold increases in adults aged 16-64 years and adults aged ≥65 years, respectively). Therefore, despite a significant reduction of cases in populations in the post-PCV era, the risk of IPD is generally still higher in comorbid versus healthy populations and in immunocompromised versus immunocompetent patients. [9][10][11][12].
In addition to young and old age and certain comorbidities, environmental, external and behavioural factors may predispose individuals to pneumococcal diseases, as listed in Table 1 [13-15]. Torres et al. suggested that clinical presentations of diseases are different according to the underlying conditions. Pneumonia is common in patients with respiratory diseases and/or in those who smoke, and bacteraemia is common in cancer patients [7].
Pneumococcal pneumonia results in significant morbidity, leading to high rates of hospitalisations, especially in elderly patients. Hospitalisation rates due to pneumococcal pneumonia in Spain were 0.25 per 1,000 in patients aged 50 Conjugate vaccines, comprising a conjugate between an antigenic protein and a polysaccharide, have been developed against a variety of bacterial species, including S. pneumoniae, Neisseria meningitidis and Haemophilus influenzae type b, to overcome the issues associated with the T cell-independent immunological characteristics of pure polysaccharide antigens (reviewed by Blanchard-Rohner and Pollard [24]).
Apart from a few polysaccharides that carry both positive and negative charges, most polysaccharides found in encapsulated bacteria and in polysaccharide vaccines cannot be processed and bound to major histocompatibility complex class II (MHC II) molecules for presentation to T-helper cells and are, therefore, considered T-independent antigens [25,26]. As a result, B-cell activation is incomplete and generally occurs outside germinal centres. There is limited immunoglobulin (Ig) class switch (mostly IgM and IgG2) [27] and somatic hypermutation (a process that is critical for obtaining high-affinity antibodies) [28]. Most importantly, the memory generated by exposure to such polysaccharides is suboptimal compared with that observed after T-dependent antigen stimulation and is mostly supported by long-lived plasma cells [29] and short-lived memory B cells that differ from T-dependent B cells [30].
B-cell response to pure polysaccharide vaccines is limited to specific subsets, principally B1 cells and marginal zone B cells [31]. B1 cells produce short-lived low-affinity antibody responses that provide the first line of defence against pathogen invasion [24]. Marginal zone B cells are sensitive to ageing and inflammation [32,33]. Moreover, they are found only in the spleen, which explains why the response to Tindependent antigens is limited after splenectomy [34]. Accordingly, increased rates of infections from encapsulated bacteria in patients with asplenia or diminished splenic function have been attributed, in part, to the absence of marginal zone B cells in these patients [24]. Infants <2 years of age are also vulnerable to these infections as, before this age, the marginal zone is immature and unable to support the development of marginal zone B cells [35].
Repeated doses of polysaccharide vaccine administered at intervals of <5 years have been shown to result in subsequently lower antibody levels (a phenomenon known as hyporesponsiveness) due to the depletion of polysaccharide-specific B cells [36,37]. It has been shown in a study in neonatal mice receiving meningococcal serotype C polysaccharide booster vaccine that this hyporesponsiveness is due to apoptosis of memory B cells [38], possibly by the inhibition of critical survival pathways for B cells [39].
In contrast with pure polysaccharide vaccines, conjugate vaccines produce an immunological response involving both B cells and T cells (Fig. 1). The chemical conjugation of a carrier protein to the polysaccharide antigen results in activation of the B-cell receptor following polysaccharide binding. The presentation of carrier protein peptides to the T cell in association with MHC II on the B-cell surface also provides signals for the activation of the T-helper cells [24]. This B cell-T cell interaction, which normally takes place during infections, provides the necessary costimulatory signals to B cells to initiate the process of germinal centre reaction [24,40,41]. This results in high titres of opsonising antibodies with Ig class switch and somatic hypermutation [24,28]. New data for conjugate vaccines against S. agalactiae (group B Streptococcus) suggest that part of the carbohydrate molecule also binds and subsequently activates T cells directed against polysaccharide antigens [42]. In view of the crucial role of T cells in mucosal immunity, this observation could help explain the dramatic effect of conjugate vaccines in the reduction of nasopharyngeal colonisation.
A conjugate vaccine, therefore, is expected to have benefits over a polysaccharide vaccine, in terms of booster response, immunological memory and generally improved immune responses, due to the T cell-dependent characteristics of the immune response.

PCV immunisation in infants: what have we learnt so far?
PCV7 has significantly reduced the burden of pneumococcal diseases in children. PCV7 was introduced in the USA in 2000. The Centers for Disease Control and Prevention (CDC) reported a 77 % reduction in overall IPD rates and a 98 % reduction in PCV7 serotype disease in children aged <5 years for 2005 compared with pre-PCV7 years (1998)(1999), based on an analysis of laboratory and population surveillance data [43]. Reductions in overall and/or PCV7 serotype IPD cases have also been documented in children aged <2 or <5 years in many other countries, including Australia, Canada, France, Norway and Spain, following the introduction of PCV7 [44][45][46][47][48][49]. Furthermore, reductions in hospitalisation rates for all-cause pneumonia (39 %) and pneumococcal pneumonia (65 %) in children <2 years of age have been observed in an US analysis of admissions data  [50]. Similarly, a Polish study found that pneumonia admission rates significantly declined following the introduction of PCV7 in children <5 years of age in the city of Kielce [51]. PCV7 has also reduced otitis media in children aged <2 years, as demonstrated by a ≥28 % reduction in recurrent otitis media [52] and a ≥43 % reduction in AOM outpatient visits or prescriptions [53].
Antibiotic resistance in vaccine serotypes and carriage of S. pneumoniae has declined since the introduction of PCV7. Kyaw et al. reported an 81 % decrease in penicillin-resistant IPD (almost all caused by vaccine or vaccine-related serotypes) among children aged <2 years in the USA [54]. In addition, reductions in the carriage of vaccine serotypes and antibiotic-resistant serotypes have been observed in Greece and the USA [55,56].
PCV7 has demonstrated an indirect effect in unvaccinated populations (herd effect), as exemplified by declines in IPD cases within adults aged ≥65 years and infants ≤60 days of age in Canada and the USA [10, 45]. A review of data from 14 countries, most of which were developed countries, also reported consistent and significant declines in both vaccine-type IPD and vaccine-type pneumococcal carriage following PCV introduction in individuals not targeted for PCV vaccination [57]. These decreases were found to be contemporaneous in studies assessing both vaccine-type carriage and vaccine-type IPD, and longitudinal data demonstrated continued declines, with the greatest declines occurring in the first few years following PCV introduction.
Since the availability of PCV7 in 2000/2001, there have been changes in the overall serotype distribution of S. pneumoniae, in particular, a rise in serotype 19A has been observed globally. The estimated proportion of IPD caused by serotype 19A has ranged from 22 % reported in a Spanish study (2001)(2002)(2003)(2004)(2005)  By 2013, 2-3 years after the widespread use of PCV13, there were emerging data concerning the impact of PCV13 on the rate of vaccine serotype-specific IPD in children. There has been a decline in IPD in the UK, due to the six additional serotypes in PCV13 following the introduction of PCV13 in April 2010 (Fig. 2) and a sustained decline in IPD due to shared serotypes in PCV7 and PCV13 [60]. Furthermore, reductions in IPD due to PCV13 serotypes have been reported in children <2 years of age in Germany following the introduction of PCV10 and PCV13 [61]. Temporal trends in nasopharyngeal carriage have also been monitored in several studies following the introduction of PCV13. In a study of 448 healthy children aged ≤6 years attending day-care centres in Portugal, there were reductions of 8 % and 10 %, respectively, in the incidence of nasopharyngeal carriage of serotypes 19A and 6C from 2010 to 2011 following the introduction of PCV13 [62]. Similarly, in a French study involving children <2 years of age with AOM, significant reductions in the nasopharyngeal carriage of PCV13 serotypes were found in PCV13vaccinated children (n = 652) compared with PCV13unvaccinated individuals (n=290) within 1 year of PCV13 introduction (p<0.001) [63]. This reduction in nasopharyngeal carriage is important for the control of pneumococcal diseases in adults, as children in day care are a key source of transmission to adults, as demonstrated by a US casecontrol study, which found this to be an independent risk factor for IPD in adults (odds ratio 3.0; 95 % confidence interval [CI] 1.5-6.2; p=0.003) [64].
Similar to the PCV7 experience, in countries where there has been a high uptake of PCV13 within childhood immunisation programmes, there has been a decline in vaccine-type IPD cases in adults, indicating a possible early herd effect. For example, in Norway, which had a vaccine coverage rate of 92 % for children <2 years of age in 2012, IPD cases caused by PCV13 serotypes in individuals ≥65 years old declined from 27/100,000 in 2010 to 18/ 100,000 in 2012 [65,66]. There have also been declines in the number of IPD cases caused by the additional PCV13 serotypes (1, 3, 5, 6A, 7F and 19A) in individuals aged ≥5 years in the UK and in adults in Germany since the introduction of paediatric PCV13 vaccination [60,67]. These data have led to a debate about whether there will be a reduced requirement for adult vaccination as the indirect effect of PCV13 increases. However, despite this potential herd effect, a significant burden of pneumococcal disease in adults remains and direct vaccination of adults is the optimal way to provide individual protection to those at risk and to significantly reduce the pneumococcal disease burden in the adult population.
There is also evidence that PCV10 has reduced the burden of pneumococcal diseases in children. Palmu et al. reported point estimates for PCV10 efficacy against vaccine-type IPD of 100 (95 % CI 83-100) for the 3+1 schedule and 92  for the 2+1 schedule in a cluster-randomised, double-blind trial, involving Finnish children aged <19 months, vaccinated with PCV10 or hepatitis vaccines (control group) [68]. A PCV10 study in Canada has shown some evidence of reduced vaccinetype IPD in infants, with a reduction in IPD cases caused by the additional three PCV10 serotypes (1, 5 and 7F) observed following the introduction of PCV10 [69]. A reduction in hospitalisation rates due to pneumonia has also been reported in Brazil 1 year after the introduction of PCV10 into the NIP [70]. No available data as of July 2014 have been reported on the impact of PCV10 in nonvaccine eligible populations. PCV10 has been shown to be immunogenic in a number of primary and booster vaccination studies [71][72][73][74][75], and also to reduce vaccine-type pneumococcal carriage [76].

PCV immunisation in other age groups
In Europe, PCV13 was licensed for use in individuals aged 6-49 years and for adults aged ≥50 years in 2011 [23]. Prior to this, PPV23 was the only pneumococcal vaccine licensed for active immunisation against vaccine-specific pneumococcal serotypes in adults in whom there is an increased risk of morbidity and mortality from pneumococcal diseases [77]. However, although PPV23 has been shown to reduce IPD in adults, data concerning the prevention of pneumococcal pneumonia or mortality are less clear [78]. Also, some limitations have been observed with this vaccine in immunodeficient patients, possibly associated with poor or absent immunogenicity [79]. In a randomised, placebocontrolled study (n=1,392) conducted in Uganda, PPV23 was found to be ineffective in preventing a first episode of IPD in individuals with HIV (hazard ratio [HR] 1.47; 95 % CI 0.7-3.3) [80]. In contrast, in a more recent study conducted in Malawi, PCV7 was found to be effective in preventing vaccine-serotype IPD in individuals with HIV (n=439) when compared with placebo (HR 0.26; 95 % CI 0.1-0.7) [81]. The immunogenicity and safety of PCV13 has been demonstrated in a clinical programme that supported licensure. This programme was designed to evaluate the functional immune response to PCV13 compared with PPV23 in two patient populations: adults 60-64 years of age who were naïve to PPV23 and those aged ≥70 years immunised with PPV23 at least 5 years before study enrolment [82,83]. In both these populations, primary vaccination with PCV13 resulted in significantly higher anti-pneumococcal functional antibody responses (as assessed by opsonophagocytic activity [OPA] titres 1 month post-vaccination) than PPV23 for the majority of the PCV13 serotypes [82,83]. Furthermore, unlike PPV23, PCV13 did not negatively affect the responses to a second dose of PCV13 administered 1 year later in the PPV23-preimmunised individuals aged ≥70 years [83]. An extension study conducted in the previously PPV23-naïve individuals also demonstrated that primary PCV13 vaccination can produce an immunological state, which allows recall antipneumococcal responses to subsequent vaccination with either PCV13 or PPV23. In contrast, primary PPV23 vaccination resulted in an immune state in which a subsequent PPV23 dose generally yielded inferior responses compared with the initial response [84]. In an analysis from three PCV13 clinical studies, the functional immune response to PCV13 in high-risk individuals with chronic medical conditions (diabetes mellitus or stable, chronic cardiovascular, pulmonary, liver or renal disease) was similar to that in non-high-risk individuals within these trials [85].
The immunogenicity of PCV13 has also been demonstrated in individuals aged 5-17 years. The immune response to PCV13 (as assessed by anti-pneumococcal IgG geometric mean concentrations) was shown to be non-inferior in children aged 5 to <10 years previously vaccinated with PCV7 compared with a PCV7/PCV13 post-toddler dose from a historical control study. The functional immunogenicity of PCV13 was also demonstrated to be non-inferior in PCV7-naïve children aged 10-17 years compared with that in those aged 5 to <10 years [86]. A similar bridging study was also conducted in the age group 18-49 years.
What can we expect from pneumococcal vaccination?
According to the WHO data, PCVs are used in 87 NIPs worldwide [21]. Within Western Europe, the use of specific PCVs in NIPs varies according to country (Fig. 3). In addition, national vaccination recommendations outside routine infant immunisation programmes differ among EU countries ( Table 2). Some countries have age-based vaccination programmes, while others have risk-based programmes, and some countries have regional variations with respect to recommendations. There are also differences between scientific society guidelines. For example, in France and Spain, the French National Reference Center of Primary Immunodeficiencies (CEREDIH) [100] and the Spanish Society of Preventive Medicine, Public Health and Hygiene (SEMPSPH) [101], respectively, recommend PCV13 for immunocompromised patients. PCV13 is also recommended for renal failure and dialysis, and for smokers in Spain (SEMPSPH and Smoking Working Group of the Spanish Pulmonologist Society [SEPAR] recommendations, respectively) [101,102]. In Germany, the Robert Koch Institute (RKI) and German Society for Hematology and Oncology (DGHO) guidelines recommend PCV13 followed by vaccination with PPV23 for asplenic patients [103,104]. Within Italy, PCV13 is recommended in individuals of all ages within the 'Vaccination calendar for life' approved by the main public health, paediatric and general practitioner organisations [105]. These guidelines, however, may evolve as clinical experience with PCV13 increases, especially in relation to the more recent indications (adults ≥18 years of age and children/ adolescents aged 6-17 years).
The WHO has published guidance for the introduction of new vaccines [106]. Cost-effectiveness, as well as disease burden, efficacy, safety and quality are recommended as key elements that should be considered as part of this process. In Italy, the clinical and economic impact of an adult pneumococcal vaccination programme has been assessed using an ad hoc population model and paediatric efficacy data [107]. This analysis found that, as a consequence of avoided pneumococcal infections, age-based PCV13 immunisation in elderly individuals (aged ≥65 years) resulted in savings ranging from 7 to 19 million Euros, depending on the vaccination strategy. An adult pneumococcal vaccination programme was, therefore, considered to be costeffective from the payer perspective. Similarly, a German cost-effectiveness analysis, which also utilised paediatric data, indicated that adult PCV13 vaccination would provide substantial health and economic benefits relative to PPV23 or no vaccination [108]. However, these findings will need to be verified with adult efficacy data, when available. The efficacy of PCV13 in the prevention of a first episode of vaccine-serotypespecific pneumococcal CAP was studied in the Community-Acquired Pneumonia Immunization Trial in Adults (CAPiTA) [109] that was recently completed. This randomised placebo-controlled trial [110] involved a pneumococcal vaccination-naïve population of 85,000 community-dwelling adults ≥65 years of age from the Netherlands.
Given both data from clinical trials and our experience of paediatric PCV13 vaccination, a beneficial effect could be expected with adult vaccination. A global literature review demonstrated reductions in vaccinetype and all-type IPD in vaccine-eligible children as well as age groups that were not eligible for vaccination following PCV7 introduction [111]. Within the 18 studies assessing the impact of PCV7 in vaccine-eligible children in the post-vaccination period, the median rate of reductions in vaccine-type IPD incidence was 90.1 % (range 39.9-99.1 %). Another literature review reported reductions in hospitalisation rates associated with pneumococcal pneumonia, ranging from 57 to 71 %, in children aged <2 years, as well as a reduction in mortality associated with all-type IPD in children [112]. There is also emerging, global evidence of a reduced burden of pneumococcal diseases (IPD, AOM and meningitis) in children following the introduction of PCV13. In addition to reductions in IPD cases within the UK and Germany (mentioned above), there have also been reports of reductions in the incidence of vaccine-type IPD in the USA and Spain [113,114], as well as reductions in meningitis cases in Greece [115] and vaccine-type AOM cases in the USA and Spain [116,117].

Conclusion
Given the high burden of pneumococcal diseases, pneumococcal vaccination is a key element of global disease prevention. Conjugate vaccines elicit a qualitatively and quantitatively higher level of immune response than polysaccharide vaccines, and, thus, provide greater immunity in children <2 years of age and in those with compromised immunity than polysaccharide vaccines. 13-valent pneumococcal conjugate vaccine  (PCV13) expectations in the more recently indicated populations (adults aged ≥50 years and children and adolescents) can be built on the heritage of PCV7 effectiveness data in children, as well as emerging impact data for PCV13. The Community-Acquired Pneumonia Immunization Trial in Adults (CAPiTA) trial should provide more definitive data on the role of adult PCV13 vaccination in preventing vaccine-serotype-specific pneumococcal community-acquired pneumonia (CAP). Continued surveillance will be important to determine the impact of PCV13 on pneumococcal diseases in these age groups and also to monitor the evolution of causative serotypes.