Immune Responses to pneumococcal vaccines in children and adults: Rationale for age-specific vaccination

Abstract

Streptococcus pneumoniae is a significant human pathogen and currently available pneumococcal vaccines are designed to elicit anti-capsule antibodies. The 23-valent polysaccharide vaccine has been used in older adults for many years whereas 7-, 10-, and 13-valent pneumococcal conjugate vaccines have only been used commonly for young children in the last decade. In addition to their high protective efficacy among children, the use of conjugate vaccines in young children has had a number of additional effects, including production of a serotype shift and providing new herd immunity to adults. The immunogenicity of both of these types of vaccines can be determined by using an ELISA assay to measure antibody levels or an opsonophagocytosis assay to assess opsonic function. As these assays have improved over time, awareness of the analytical limitations of older studies has grown. While the 23-valent vaccine is effective among young adults, it is less effective among elderly adults. Aging-associated ineffectiveness may be due to aging-dependent changes in the antibody repertoire and/or a reduction in IgM antibody production associated with aging-dependent changes in B cell subpopulations. The immunologic basis of aging-associated immune defects thus remains an active area of research.

1. Pneumococcal infection epidemiology

First isolated in 1880, Streptococcus pneumoniae was one of the first bacteria to be characterized [1] and it is often simply referred as pneumococcus. S. pneumoniae is a Gram-positive bacterium with a thick polysaccharide (PS) capsule. Because this capsule dramatically increases virulence [2], it has been extensively investigated for its serologic differences. Use of polyclonal rabbit antisera led to the definition of 90 different capsule types by 1995 [3]. More recently, use of monoclonal Abs has led to the identification of several additional serotypes [4].

The pneumococcus is a major pathogen. Prior to the introduction of conjugate vaccines, it is estimated to cause 500,000 cases of pneumonia, 50,000 cases of bacteremia, and 3,000 cases of meningitis per year in the United States alone, resulting in 40,000 deaths per year [5]. The incidence of invasive pneumococcal disease (IPD), which includes bacteremia and meningitis, was 24.3 cases per 100,000 persons in 1998–1999 when pneumococcal conjugate vaccine was not yet available [6]. Most episodes of pneumococcal bacteremia are secondary to pneumococcal pneumonia, with every 3 to 4 episodes of pneumonia leading to bacteremia resulting in an estimated incidence of pneumococcal pneumonia that may exceed 40–60/100,000. Pneumococcal disease incidence has a characteristic age distribution with the highest incidence seen at the extremes of age, the very young and the very old. For instance, in 1990 the incidence of bacteremic pneumococcal disease in South Carolina was found to be 160/100,000 in infants, 5/100,000 in young adults and 70/100,000 in individuals >70 years of age [7].

It is among young children that pneumococcal infections have traditionally had the most devastating impact. For example, it is estimated that 0.7-1 million of the global 1.6 million deaths due to pneumococcal infections per year occur in children [8]. In the pre-conjugate vaccine era, pneumococcus was the most common organism isolated from middle ear fluid of children with otitis media [9], accounting for approximately 50% of cases and a total of 7 million cases of otitis media per year in the United States alone [5]. Unfortunately, although routine vaccination with the conjugate vaccine has significantly reduced the incidence of otitis media caused by vaccine serotypes, pneumococci with capsular serotypes that are not found in the vaccine continue to play a major role as causative agent of otitis media [10–12]. In some regions, increase in IPD prevalence by non-vaccine types has largely replaced the reduction in the prevalence of vaccine types [13, 14].

Among elderly patients, pneumococcusis the most common organism isolated from patients with community acquired pneumonia (CAP) and accounts for at least 30% of all CAP cases [15–19]. Several reports have emphasized the importance of age in increasing the risk of acquiring pneumococcal infection [7, 20–24]. Although age is an important risk factor for IPD, several other factors play an important role in determining IPD risk for adults. These include ethnicity [25], low socioeconomic status, chronic underlying diseases such as chronic obstructive pulmonary disease (COPD), heart disease, diabetes and renal disease, and high risk behaviors such as smoking and alcohol abuse [25–29]. At particularly high risk are those individuals with recent respiratory viral illnesses (i.e., influenza, respiratory syncytial virus). Decreased mucosal clearance [30] and increased pneumococcal adherence [31] results in a significant increase in pneumococcal infections, particularly in the winter months, when these viruses tend to circulate [32]. Indeed, pneumococcal pneumonia following influenza infection was likely responsible for most of the deaths in the historical 1918 influenza epidemic [33].

Perhaps the best recognized population susceptible to pneumococcal infections is persons with immune deficiencies. For instance, children and adults who have a congenital or acquired inability to produce Abs to the capsular PS or deficiencies in early complement components or phagocytes are very susceptible to pneumococcal infections [34]. This category includes patients with selective or generalized hypogammaglobulinemia [35–37], including children with transient hypogammaglobulinemia of infancy, IgG subclass deficiency, impaired PS responsiveness, and selective IgA deficiency [38]. Children and adults with congenital, acquired or functional asplenia, such as sickle cell disease, are at significantly increased risk of fulminant IPD, emphasizing the role of the spleen in clearance of pneumococci [39, 40]. S. pneumoniae is a major cause of morbidity and mortality in human immunodeficiency virus (HIV) infected individuals. It is the most common bacterial respiratory pathogen encountered in the HIV-positive population and the disease is frequently complicated by bacteremia and/or recurrences [41–45]. In the era prior to the availability of highly active anti-retroviral therapy (HAART), the incidence of IPD in HIV-positive individuals was estimated to be a 100-fold higher [41] than in age-matched HIV-negative subjects. Several studies have demonstrated a significant decrease in incidence of pneumococcal disease in HIV-positive individuals in the post-HAART era [46–50]. However, despite the widespread availability of HAART, HIV-infected individuals remain at a 35-fold increased risk of IPD compared to age-matched HIV-negative controls [51].

2. Therapeutic and preventive interventions to control pneumococcal infections

Anti-capsular antisera are highly protective and were successfully used as a serotype-specific therapeutic agent before antibiotics were available [52]. With the development of antibiotics including penicillins and cephalosporins, antibiotics became the mainstay for treating pneumococcal infections. However, the rates of antimicrobial resistance in pneumococci have steadily increased [53, 54]. Moreover, despite the use of appropriate antibiotics and intensive care, for over 50 years the case fatality rate of pneumococcal bacteremia has remained at 15% to 20% in children and young adults and 30 to 40% in the elderly [55]. Furthermore, despite successful antibiotic treatment, pneumococcal meningitis can cause disabling long term sequelae. These factors led to the development of vaccines to prevent pneumococcal infections.

The development of pneumococcal vaccines requires understanding of pathogenic mechanisms involved in pneumococcal infections. The organism colonizes the respiratory tract of healthy children and adults [56, 57] and subsequent invasion of the respiratory tract or bloodstream leads to life threatening infections such as pneumonia or sepsis [58, 59]. Antibodies (Abs) to the capsule can abrogate its shielding effect by fixing complement and opsonizing pneumococci for phagocytosis [60, 61]. This protective mechanism has been clearly illustrated by the various immune deficiencies described above with frequent and severe pneumococcal infections. Based on these observations, all presently available pneumococcal vaccines incorporate protective PS antigens.

Polysaccharide vaccines

The discovery that purified capsular PS was immunogenic led to the development of vaccines containing pooled capsular PSs of various serotypes. Several PS vaccines were developed and tested in young South African gold miners in the early 1970’s. Among the miners, the 13-valent experimental PS vaccine was 82.3% efficacious against bacteremia and 78.5% against pneumococcal pneumonia [62]. Similar results were demonstrated using 6- or 12-valent versions of the vaccine [63]. Experience with these early vaccines led to the development of the 14-valent pneumococcal PS vaccine (PPV14), which was licensed in 1977. It was estimated that 68% of IPDs were caused by the 14 serotypes present in the vaccine [64]. In 1983, the presently available 23-valent PPV (PPV23) was approved, expanding serotype coverage to more than 85% of the organisms that caused IPD at that time [65]. Since its introduction, several efficacy studies have shown that in young adults the PPV is highly effective against IPD [24, 66–68].

In contrast to studies with young adults, several studies have noted a markedly decreased efficacy in the elderly [24, 69–71]. A retrospective cohort study assessed the effectiveness of the pneumococcal vaccine in 47,365 persons over 65 years of age [72]. Immunization with the pneumococcal vaccine was associated with a significant reduction in the risk of pneumococcal bacteremia with an estimated vaccine efficacy of 44%. However, the vaccine failed to reduce the risk of CAP. Several prospective randomized vaccine efficacy trials conducted among older adults have resulted in similar findings [73–76]. Suboptimal clinical efficacy likely reflects the poor immunogenicity of PPV23 in the elderly and the very young, i.e., children <2 years of age, which together represent the age groups at highest risk for IPD.

Pneumococcal conjugate vaccines

The poor immunogenicity of PPV23 in infants and the limited duration of protective antibody levels is associated with the T-independent nature of the pneumococcal capsular PS vaccine. The composition of the antibody and T cell receptor repertoires in the very young differs from that of older children and adults [77, 78]; and anti-PS antibody responses have also been associated with specific splenic B cell subsets that are not fully developed until after the age of two [79]. This latter need for a fully developed spleen may help explain both the increased susceptibility of youth and that of asplenic patients of all ages. Efforts to overcome the inherent problems of the PPV23 have led to the development of the pneumococcal conjugate vaccine (PCV). Covalent coupling of the PS to a protein carrier effectively converts the T-independent Type 2 PS into a T cell dependent antigen [80]. The potential long-term benefits or consequences of inducing a T-dependent response in place of the evolutionarily preferred T-independent process remain unknown.

Early studies demonstrated that PS conjugate vaccines can elicit a protective antibody response, recruit T cell help, and induce immunological memory in infants and young children < 2 years of age [81, 82]. PCV was prepared by conjugating pneumococcal PS to carrier proteins including the non-toxic mutant of diphtheria toxin (CRM197), as well as other proteins such as protein D of non-typeable Haemophilus influenzae. The conjugation process is difficult, requiring each PS to be individually conjugated and thereby limiting the number of serotypes that could be included in PCVs.

It should be mentioned that serotype distribution is both age, geographically, and temporally dependent [83–85]. These factors played an important role in the development and formulation of the conjugate vaccines. The first 7-valent conjugate vaccine (PCV7) was licensed in the United States in 2000 and included capsular PSs of serotypes 4, 6B, 9V, 14, 18C, 19F and 23F individually conjugated to a protein carrier. The selected serotypes represented approximately 80-90% of IPD [86] and 76% of otitis media [10] causing serotypes in young children in the United States. The coverage in Europe and Asia is somewhat less due to the larger contribution of serotypes 1 and 5 as causes of IPD in these parts of the world, particularly in Asia [87, 88]. PCV7 includes the five serotypes (6B, 9V, 14, 19F and 23F) representing 80% of penicillin-non-susceptible isolates at the time of marketing [86]. PCV7 has proven highly efficacious against invasive disease, with an efficacy of 97.4% [89] and it is modestly effective against pneumonia [90] in children <5 years of age. Moreover, PCV7 has reduced the risk for otitis media related physician visits by up to 40% and the need for insertion of pressure equalizing tubes (PETs) by 24.2% [91–94]. The efficacy of PCV7 on the incidence of vaccine-related serotype acute otitis media (AOM) is thought to be 65% [91]. Also, PCV7 was shown to reduce nasopharyngeal carriage of pneumococci in immunized individuals [95–99]. Clearly, introduction of PCV7 has had a major impact on the incidence of pneumococcal disease in infants and children.

In addition, PCV7 demonstrated a phenomenon called “herd immunity” by showing unexpected decrease in IPD among non-vaccinated adults [6, 100]. The decline in disease incidence was age group dependent, with a decrease of 40%, 18% and 37% in individuals between 18–49 years, 50–64 years, and those older than 65 respectively [100]. This herd immunity can be explained by the fact that old adults often acquire pneumococci from young children and that young children carry less pneumococci following vaccination. This explanation is supported by a clear temporal association between a sharp increase in IPD cases and the end-of-the-year holiday season, when the families typically get together [101].

Another interesting observation following the widespread use of the PCV7 was a shift in serotype prevalence. There was a significant decrease in diseases such as IPD and otitis media caused by vaccine serotypes, but there was an increase in diseases by non-vaccine serotypes. Amongst the most prominent of the so called replacement serotypes are 19A and 7F [102, 103]. Post-PCV7, serotype 19A became the most prevalent penicillin-nonsusceptible isolate from clinical specimens [104–107]. In Europe, serotype 1 has recently become common [108]. Another example of serotype shift was shown with a novel pneumococcal serotype 6C, which was identified in 2007 using monoclonal Abs [109]. Serotype 6C was present in the pre-PCV7 era, but had inadvertently been serotyped as 6A because of serologic similarities between serotypes 6A and 6C. Genetic studies demonstrated clear differences between them: a galactosyl transferase gene in the capsule gene locus of 6A has been replaced with a glucosyl transferase gene in 6C [4]. However, several epidemiologic investigations have found a significant increase in both carriage and IPD of 6C following the introduction and widespread use of PCV7; whereas the prevalence of 6A has decreased [110–114].

Because of these serotype shifts, the protective coverage of PCV7 was reduced. Furthermore, PCV7 did not include serotypes 1, 3, and 5 which are common in Europe, Asia and Africa. In view of this, two new PCVs (PCV10 and PCV13) have recently become available. PCV10 is licensed in Europe and contains serotypes 1, 5, and 7F in addition to the 7 serotypes in PCV7. PCV13, which became available in 2010 in the US, contains three additional serotypes: 3, 6A, and 19A [115]. These new PCVs should provide much broader coverage.

Clinical uses of vaccines and vaccination policy

Pneumococcal vaccine policies have been dictated by the epidemiology of pneumococcal infections. Moreover, individuals at highest risk of IPD fall into four general categories: children <2 years of age, individuals >65, children 2–18 years and adults 18–64 with acquired or congenital immunodeficiencies predisposing to IPD.

As young children are highly susceptible to pneumococcal infections, children are an important target population for pneumococcal vaccination. In the pre-conjugate vaccine era, vaccination recommendations with PPV23 were limited to children ≥2 years of age with increased risk for pneumococcal disease. These included children with CSF leaks, chronic cardiovascular disease, anatomical or functional asplenia, agammaglobulinemia, diabetes mellitus, sickle cell anemia and HIV [5]. After approval of the PCV7 in 2000, the Advisory Committee on Immunization Practices (ACIP) recommended routine vaccination with PCV7 of all children 2–23 months of age [86] and in 2007 these were further amended to include routine vaccination with PCV7 of all children 2–59 months of age [116]. The number of vaccine doses varies from 1 to 4 doses and is determined by age at the time of vaccination.

In 2010, PCV7 was replaced by PCV13, increasing coverage to include the so-called replacement serotypes and serotypes prevalent in countries outside of the United States. PCV13 is approved for use in children 6 weeks to 71 months of age for the prevention of IPD and otitis media. The ACIP currently recommends vaccination with PCV13 for all children 2–59 months of age and children 60–71 months of age with underlying medical conditions that increase the risk of pneumococcal disease including cochlear implants [115]. All children two years of age and older at high risk for pneumococcal disease should also receive PPV23 after completing all recommended doses of PCV13, as they may have infections caused by serotypes not included in PCV13. A single dose of PCV13 may be administered for children 6–18 years who have not received PCV13 previously and are at increased risk for invasive pneumococcal disease regardless of whether they have previously received PCV7 or PPV23 [115]. All children with asplenia, sickle cell disease, HIV, and other immunocompromising conditions should also receive a second dose of PPV23 five years after their first dose of PPV23.

Despite the fact that routine use of PCV7 among children has significantly reduced IPD in the non-vaccinated population by herd immunity [100], IPD incidence remains high but controversial in individuals ≥65 years of age and in adults with underlying medical conditions at increased risk of pneumococcal disease. One study demonstrated an increase in IPD from 1998–1999 to 2006–2007 in percentage of adult IPD in the elderly from 69% to 81% and in those with underlying medical conditions aged 18–64 from 52% to 59% [100]. However, conflicting studies demonstrate a significant decrease in IPD in these populations [117, 118]. Based on this age-related incidence of IPD, it is recommended that all individuals ≥65 years of age receive PPV23. In October of 2009, new vaccination guidelines recommend vaccination of those ≥65 years of age, 18–65 years of age with a broad array of underlying medical conditions and expanded recommendations to include all asthmatics and smokers in this age group with PPV23 [119].

Several studies have documented a significant decline in anti-pneumococcal PS antibody concentration, and likely protective efficacy, in elderly and immune compromised adults over time [24, 120–122]. It is therefore recommended that all adults 18–65 years of age with asplenia and/or immune compromise and all those vaccinated before age 65, be revaccinated 5 years after primary PPV23 vaccination. Moreover, repeated immunization with the meningococcal PS vaccine has led to immune tolerance and unresponsiveness [123]. Similar considerations have played a role in revaccination guidelines concerning PPV23. Although some studies have shown reduced responses, the most recent study from CDC found that repeat vaccination with PPV23 is safe, well-tolerated, and not associated with reduced antibody responses [124].

Invasive disease caused by S. pneumoniae is significantly increased in the HIV-positive population and is a major cause of morbidity and mortality. It is therefore recommended that HIV-positive individuals are vaccinated with PPV as close to the time of diagnosis as possible [125]. However, for those individuals with a CD4⁺ cell count <200 at the time of diagnosis, the recommendations are less straight forward. Thus, HIV-infected adults with CD4⁺ cell counts <200 “can be offered” pneumococcal vaccination [126]. Clinical evidence, however, has not confirmed efficacy in this group, although there is some potential benefit if HAART is started immediately [127]. It is currently recommended HIV-positive individuals be revaccinated with PPV23 five years post-primary vaccination [125, 126] although the clinical (or serological) benefit of revaccination is unknown.

Since PCVs are highly immunogenic and induce immunological memory in infants and children, there are attempts to use PCVs among old adults. So far, several studies in adults or the elderly have shown that PCVs do not offer an advantage in immunogenicity over the conventional PPV23 [128–134]. However, additional studies are in progress to investigate various factors such as immunization doses as well as new formulations of PCVs.

3. Monitoring immune responses to vaccination

To evaluate these different vaccines in different target populations, two different serological measures are used as surrogates of protection. One is the amount of Abs against capsular PS. The other is the capacity to opsonize pneumococci as the measure of the functional characteristics of Abs to pneumococcal capsule. While both assays are conceptually simple, the assays were found to be quite complicated and clinical studies based on non-specific assays may have produced misleading conclusions. Consequently, it is important to understand the developmental history of the pneumococcal antibody assays [135].

ELISA

When the general ELISA technology became available in the 1970’s, a pneumococcal antibody ELISA was quickly developed using plastic wells coated with pneumococcal capsular PS with subsequent measurement of the amount of human IgG bound to the plastic wells. Despite the straight forward logic in the assay design, historically there had been a poor correlation between antibody concentration, as measured by ELISA, and functional antibody activity determined by opsonophagocytic assays. The poor correlation was explained when the capsular PS antigen preparations used in ELISA were found to be contaminated with a variety of cell wall components such as cell wall PS (CWPS) and other pneumococcal determinants [136, 137]. The ELISA thus measured non-functional Abs directed at other determinants in addition to the functional anti-pneumococcal PS Abs.

To improve the assay specificity, over a 20 year period the pneumococcal antibody ELISA has undergone two additional generations of improvements [135]. For the second generation ELISA the test sera were pre-absorbed with CWPS alone. For the current third generation assay, the sera are pre-absorbed with CWPS and another capsular PS, such as PS from pneumococcal serotype 22F. This third generation ELISA has a significantly improved correlation between antibody concentration and opsonophagocytic activity [136]. It has been approved by the World Health Organization (WHO) and is now widely used [135]. Nevertheless, when one reads literature published before the development of the third generation ELISA in 2000 [135], one must be aware that the antibody amounts reported may be falsely elevated.

OPA

The opsonophagocytosis assay (OPA) is an inherently more desirable assay than ELISA as a surrogate of protection since anti-capsule Abs provide protection in vivo by opsonization. OPA was clearly preferred when pneumococcal antibody ELISA was non-specific. Even with improvements in specificity, OPA may be preferable to ELISA in certain situations. For instance, PCV7 has been shown to elicit a high level of anti-19A antibody and was expected to provide cross-protection against 19A [138, 139]. But the clinical use of PCV7 demonstrated that it is not protective against 19A [104–107]and also opsonization assays show that PCV7 does not elicit high levels of opsonic antibody to 19A [138, 139]. Perhaps the Abs have enough avidity to bind the PS, but not sufficient to opsonize pneumococci.

For the reasons above, various assays designed to measure opsonic capacity have been developed [140]. These are bioassays typically using human test serum, target bacteria, human phagocytes and rabbit complement. At the end of the assay, surviving pneumococci are enumerated. As these bioassays are technically cumbersome, they have been only rarely used in the past. However, recent improvements have resulted in a more practical OPA, which is 4-fold multiplexed and based on the classical killing type OPA [141]. In fact, the multiplexed OPA is almost as easy and fast as ELISA. In view of these changes, it is likely that there will be an increased use of OPAs in the future.

Clinical uses of ELISA and OPA

In 2005, WHO published new serological criteria to be used in consideration for licensure of new PCVs for use in infants [142, 143]. These criteria included the ability to induce an IgG anti-pneumococcal PS concentration of at least 0.35 μg/ml four weeks post-vaccination. This reference value was based on the results of three clinical efficacy trials of PCV7 conducted in infants and children [89, 142, 144–146] and the second generation ELISA (i.e., absorbed test sera with CWPS alone). The value should not be used to predict protection for an individual since the value represents estimates or threshold levels that predict protection for a group of children. Also, the trials studied protective efficacy against IPD and the value may not be useful in predicting other conditions such as nasopharyngeal carriage. Furthermore, the 0.35 μg/ml may correspond to 0.20 μg/ml if the third generation ELISA is used (i.e., the samples are absorbed with CWPS and a heterologous pneumococcal PS like serotype 22F PS) [147, 148]. The antibody concentration of 0.20–0.35 μg/ml may correspond to an opsonophagocytic titer of 1:8 for young children [142, 143].

Correlates of protection have not been defined for adult populations. The large scale protective efficacy studies necessary to achieve this goal have not been conducted in adults immunized with the PS vaccine. These studies may be difficult to conduct in the future due to the unethical nature of withholding PPV23 from target populations. Many non-vaccinated old adults have high levels of pneumococcal antibody. It is therefore generally accepted that in this population the opsonophagocytic capacity should correlate better with protection than ELISA. Although a numerical value has not been strictly assigned, higher opsonophagocytic antibody titers greater than 1:8 appear more likely to provide protection. For some serotypes, higher thresholds may be necessary.

Pneumococcal antibody ELISA was initially developed to monitor immune responses to newly produced pneumococcal vaccines. However, this assay is frequently used to assess the presence and extent of congenital or acquired immune deficiency [37]. Additional new uses of the ELISA may be to monitor therapeutically-induced immune suppression for patients that have undergone transplantation or are being treated for autoimmune diseases [149]. The blood samples for these clinical uses are generally analyzed by commercial reference laboratories, which commonly use multiplexed immunoassays based on bead array technology [150]. These multiplex assays have been extensively compared by reference laboratories with WHO ELISA using sera from old adults as well as young children. However, clinical studies based on these multiplexed immunoassays are not yet available.

While these tests have well-defined uses for vaccine development, attempts to rationalize the interpretation of the data to make clinical decisions have proved to be complex. Variables include whether to use the PPV23 or the conjugated vaccines, how to administer and utilize them, whether to continue to use ELISAs or switch to opsonophagocytosis assays in order to measure functional responses and how to interpret the data in the context of complex clinical scenarios. As a result, the interpretation of diagnostic vaccination can sometimes lead to more questions than answers. A working group of the American Academy of Allergy Asthma and Immunology (AAAAI) is currently seeking to achieve a consensus on these issues.

4. Problems in vaccinating old adults against pneumococcus

Although efficacious in young adults, the pneumococcal PS vaccine is less effective in the population at highest risk for IPD, namely the elderly [24, 69–76]. The reason for the sub-optimal efficacy of the pneumococcal vaccine in the elderly has been the subject of numerous investigations. First, the elderly antibody response to the vaccine, as measured by RIA or IgG ELISA, has been measured and compared to that in young adults [151–162]. These studies indicate that the post-immunization anti-PS antibody concentrations in the elderly are similar to those in younger adults. Although the decrease in protective efficacy in the elderly was poorly understood based on IgG anti-PS concentrations, several mutually non-exclusive explanations are possible. First, studies performed at the CDC indicated that despite adequate IgG antibody concentrations, the elderly had a significant reduction in opsonophagocytic activity against all serotypes tested [153]. These observations were recently confirmed in two separate studies performed in elderly subjects [158, 163]. As a result, individuals with IgG antibody levels thought to be commensurate with protection may not necessarily be protected against disease. Reduced or absent functional antibody activity, as determined by opsonophagocytic or mouse protection assays, is directly related to low antibody avidity [164, 165]. Specifically, it has been demonstrated that post-vaccination sera from the elderly with low opsonophagocytic activity correlate with low IgG antibody avidity [153]. Moreover, antibody avidity appears to play a critical role in antibody function. Extensive studies with H. influenzae type b (Hib) anti-PS Abs have clearly demonstrated the correlation between antibody avidity, fine specificity, protective efficacy and the expression of particular V regions and clones [166–169].

Studies to define the structure-function relationship of anti-pneumococcal Abs have been performed using human monoclonal Abs [170–173], combinatorial libraries [174–176], and monoclonal Abs derived from a transgenic mouse strain reconstituted with human immunoglobulin loci (the XenoMouse model) [177, 178]. Anti-pneumococcal PS specific human monoclonal Abs of the IgM isotype generated from peripheral blood mononuclear cells (PBMC) of immunized volunteers, or from the XenoMouse model, demonstrate 97 to 100% homology to germline sequences [171, 173, 177, 178]. These mAbs used either VH3 or VH1 gene families.

Sequence analysis of human hybridomas [173] and Fabs isolated from combinatorial libraries [174, 175] of the IgG or IgA isotype indicate that they use the same restricted gene families, namely, VH3 and VH1. In contrast to the IgM Abs, the IgG and IgA Abs demonstrate high levels of somatic mutation in CDR1 and/or CDR2 indicative of a B cell memory response. Somatic mutations within canonical regions of V genes probably affect the conformation of the antigen binding groove resulting in significantly altered avidity and functional activity.

We have characterized the immunoglobulin VH gene usage of the antibody response to capsular PS of serotypes 4 and 14 in 20 young and 20 elderly adults [179]. In all volunteers, the immune response to both PSs consisted predominantly of heavy chains belonging to the VH3 gene family. There were significant differences in the variable gene repertoire between young and elderly adults. Somatic mutation occurred more frequently in sequences derived from young when compared to those derived from the elderly. When compared to young adults, aged individuals demonstarted a loss of oligoclonality in their response to pneumococcal PS of serotypes 4 and 14. The observed differences in VH repertoire, somatic mutation, and loss of oligoclonality may contribute to decreased vaccine efficacy and antibody avidity in the elderly. Studies defining the elderly immune response to pneumococcal PSs performed on a single B cell level are presently under way. These studies will allow direct correlation between antibody structure and functional activity.

A second mechanism potentially responsible for the decrease in opsonophagocytic activity may be related to the observation that the elderly produce significantly less anti-PS IgM Abs than young adults [180–184]. As IgM Abs may be available earlier during the onset of infection and they are more effective in fixing complement and opsonizing ability than IgG, the absence of IgM antibody can be significant. In a recent study, we measured the anti-PS IgG, IgM and IgA responses against serotypes 14, 18C and 19F in elderly and young adults. As previously shown, there was no difference in anti-pneumococcal PS IgG response between these age groups. The opsonophagocytic titer however, was higher in young adults compared to the elderly. Removal of IgM anti-pneumococcal PS Abs increased the correlation coefficient between antibody concentration and opsonophagocytic activity and decreased differences in opsonophagocytic titer between young and elderly [185]. Together, these studies suggest a potentially important role for IgM Abs in generating protective immunity. Thus the current ELISA, which measures only IgG Abs, may be insufficient to truly assess the level of protection and thus ELISA assays for IgM Abs may need to be developed.

This observation is intriguing because there have been observations in the past that suggested that a human IgM memory B cell subpopulation is reduced in older adults [186, 187].The nature of the B cell population responsible for the immune response to capsular PSs has been a controversial issue for many years. Recent studies, however, suggest that peripheral blood IgM memory B cells (CD27⁺IgM⁺IgD^±) may include the circulating counterpart of splenic marginal zone B cells [188, 189]. By virtue of their pre-diversified IgM antigen receptor repertoire, IgM memory B cells are capable of responding immediately, without T cell help, to TI-2 antigens [186, 187]. These IgM memory B cells are thought to form the first line of defense against encapsulated organisms.

It has been shown that persons with decreased or absent circulating IgM memory B cells, including the elderly, splenectomized persons, infants less than 2 years of age, and a subgroup of common variable immunodeficiency patients, all respond poorly to PS vaccines and are highly susceptible to infections with encapsulated organisms [79, 186, 187, 189–191]. In addition, IgM memory B cells show a significant increase in VH3 gene family representation compared to naïve B cells and are often (94.7%) somatically mutated [192]. These observations thus support the concept that IgM memory B cells are important in generating the response to TI-2 antigens.

It is, however, unlikely that IgM memory B cells are solely responsible for anti-PS antibody production. First, switched memory B cells (CD27⁺IgM⁻) secrete higher levels of anti-pneumococcal PS antibody than CD27⁺IgM⁺ memory cells following in vitro stimulation [193]. Second, sequence analysis of anti-pneumococcal PS Abs, 5 days post-vaccination, demonstrate a predominance of IgG and IgA Abs, which are presumably derived from switched memory cells that have undergone somatic hypermutation[174, 176, 179]. This may be due to additional PS responsive B cell subsets and/or the versatile role that IgM memory cells play in pneumococcal antibody responses, as suggested by Bossuyt and colleagues [194]. In mice, the B1b cell subset participates in IgM and isotype-switched Ab production in response to pneumococcal PS [195, 196]. However, whether a human B1b cell counterpart exists is unknown. Although the specific B cell subset(s) involved in the immune response to carbohydrate antigens in humans remains to be elucidated, it is postulated that CD27⁺IgM⁺ memory cells, significantly decreased in the elderly, contribute to the impaired immune response to pneumococcal vaccination in this population.

Although the CD27⁺IgM⁺ memory cells are thought to include effectors of the adaptive response, another subset of B cells, namely B1a cells, could be playing a significant role in the innate immunity to encapsulated organisms. In mice, B1a cells are phenotypically defined by CD5 expression. These cells, compared to conventional B2 cells, are unique in several aspects including phenotype and production of natural Abs [197, 198]. The Abs produced by B1 cells bind autoantigens and are widely cross-reactive binding a variety of bacterial antigens such as PSs and LPS [199]. Studies performed in mice demonstrate that these CD5⁺IgM⁺ B cells are spontaneous producers of natural Abs that appear to play a vital role as the first line of defense against bacterial and viral pathogens including the pneumococcus [200–203]. Moreover, lack of B1a cells increases susceptibility to pneumococcal infections [195]. Limited studies in humans have demonstrated that CD5⁺B cells represent the majority of B cells at birth, but diminish thereafter, with low numbers present in the elderly [204–206]. The existence of a true B1a subset in humans has been hindered by the expression of CD5 on multiple B cell populations. However, a recent study by Rothstein and colleagues [207] demonstrates the human IgM⁺CD20⁺CD27⁺CD43⁺CD5^+/− subset may be the true functional equivalent of murine B1a cells. Notably, this population would have been included in previous studies of the heterogeneous IgM⁺CD27⁺ “memory” population. This recently identified human B1 subset shows skewed specificity for phosphocholine and declines sharply with age. The exact role of B1 cells in increased risk of pneumococcal infection in elderly humans, however, remains to be elucidated.