Vaccination with the pneumococcal 7-valent conjugate: a successful experiment but the species is adapting

Streptococcus pneumoniae exacts a tremendous death toll and is currently the most common cause of pneumonia and meningitis in young children, older adults and immunocompromised populations. A 23-valent polysaccharide vaccine is protective against invasive pneumococcal disease (IPD) in adults and children over 2 years of age; however, it does not affect colonization, nor is it effective in young children. A vaccine consisting of seven capsular polysaccharides (comprising serotypes 4, 6B, 9V, 14, 18C, 19F and 23F) conjugated to diphtheria toxoid protein CRM197 was introduced in the USA during the year 2000 for infants and young children. This vaccine targeted the serotypes that cause more than 80% of IPD in children and the serotypes accounting for 83% of IPD caused by penicillin-nonsusceptible pneumococci in children under 5 years of age. This 7-valent conjugate vaccine (PCV7) has demonstrated spectacular success, not only in reducing IPD in children [1,2], but also in reducing colonization in this target population with resultant decreased IPD in older-age groups [1,3].

The pneumococcus is a human commensal that has a well-documented history of rapid change in response to immunologic and antibiotic pressures in its nasopharyngeal environment. Until the use of penicillin was routine, this opportunistic pathogen was a predominant cause of death [4]. The emergence of strains resistant to penicillin and other antibiotics during the past 20 years exacerbated the pneumococcal disease burden and demonstrated the remarkable capability of this organism to adapt to environmental selection through the mechanism of genetic recombination. In virtually all penicillin-resistant pneumococcal clinical isolates examined, there are obvious indicators of recombination within multiple genes encoding penicillin-binding proteins (pbps). In these strains at least three mosaic pbp gene exhibit blocks of nonpneumococcal gene segments that indicate past recombination events with related nonpneumococcal species [5]. Even though the original nonpneumococcal pbp gene segment donors are not naturally penicillin resistant, within the mosaic segments contributed by these species are a limited number of key substitutions that are critical for decreased pbp binding affinity to penicillin. Besides pbp genes, mosaicism has also been seen within other determinants of this species, including the dhf gene, which confers resistance to trimethoprim [6,7], giving further evidence of the ability of this species to adapt to selective pressures by undergoing recombination events [6,7].

Recombination also potentially serves as an immune escape mechanism for the survival of pneumococcal strains. Capsular switching events that have occurred owing to recombinational replacement of the capsular biosynthetic locus in the pre-PCV7 era have been characterized at the molecular level and were seen to involve PCV7 serotype donors and recipients [8]. Recent data are also suggestive of capsular switching in the post-PCV7 era, involving non-PCV7 serotype donors and PCV7 serotype recipients [9–11]. A capsular switch event is initially suggested without DNA sequence analysis of crossover regions when a strain of a given serotype exhibits a genetic type that is normally highly associated with a different serotype. Regardless of whether switching events occurred before or after vaccine introduction, evidence suggests that in the post-PCV7 era, serotype 19A, a serotype for which the vaccine provides no protection, has become more genetically diverse owing to the expansion of clonal complexes formerly highly associated with PCV7 serotypes [9,10]. In a relatively short period after PCV7 introduction, serotype 19A has become the prevalent invasive serotype within the USA [9,11,12].

Although the overall impact of vaccine-induced replacement of prevailing pneumococcal serotypes is currently small, there have been significant increases in IPD due to individual nonvaccine serotypes (NVTs), such as 3, 19A, 22F and 33F [12]. Between the years 2000 (the year of PCV7 introduction) and 2005, the incidence of invasive disease due to serotype 19A increased in all age groups accounting for an overall increase of more than 330% [11]. Serotype 19A accounted for 2.5% of invasive pneumococcal isolates recovered from children under 2 years of age during 1998/1999, whereas it accounted for 36% during 2005. The percentage of invasive type 19A isolates fully resistant to penicillin skyrocketed from 6.4% among isolates collected during 1998/1999 to 34.5% during 2005. These dramatic changes in the incidence of serotype 19A IPD and proportion of type 19A isolates that are resistant reflect marked changes in the overall genetic structure of invasive serotype 19A pneumococci. The increased incidence of type 19A IPD together with the appearance of serotype 19A variants with highly related or identical multilocus sequence types (STs) to STs from well-known antibiotic-resistant clones, most often linked to PCV7 serotypes, suggest that vaccine and/or antibiotic selective pressure effected changes in the 19A genetic structure. The most striking clonal emergence within serotype 19A in the post-PCV7 period has been the rise in clonal complex 320 (CC320) [9,11]. This highly penicillin-resistant, multiresistant complex is now the second most common genotype found among 19A isolates in the USA, comprising more than 20% of serotype 19A invasive isolates, based upon multilocus sequence typing of 528 invasive serotype 19A isolates recovered during 2005 [11]. The primary multilocus ST within this complex, ST320, represents a double locus variant of the widespread Taiwan^19F-14 clone, which comprises the majority of the currently declining multiresistant invasive serotype 19F isolates recovered in the USA [13,14]. It is important to note that CC320 was not detected during extensive genotypic surveillance of year 1999 serotype 19A invasive pneumococci [13]. Other clonal complexes, including complexes highly related or genetically identical to PCV7-targeted type pneumococcal clones Spain^9V-3, North Carolina^6A-23 and Spain^23F-1, also contribute to the high-level antibiotic resistance that has emerged within serotype 19A in the post-PCV7 era [9,11].

A very interesting serotype 19A serotype switch variant recovered in the USA during the postvaccine era possesses the multilocus ST (ST695) that identifies the major antibiotic-susceptible invasive serotype 4 clone [9–11]. It is interesting that the ST695 type 19A strain, unlike the putative serotype 4 parental strain, is intermediately penicillin resistant [11]. This is apparently due to a single double-crossover event, resulting in recombinational replacement of the cps locus, as well as the flanking pbp2x and pbp1a genes in the ST695 type 4 clone, with the cps locus and mosaic pbp gene counterparts from the major serotype 19A clone donor [10]. These current data suggest that the immune escape mechanism of serotype switching may contribute to the emergence of antibiotic resistance within new clonal types owing to the hitch-hiking effect of the flanking pbp genes. Analogous cotransformation of the cps locus, pbp1a and pbp2x, has been demonstrated under laboratory conditions, with simultaneous conversions in serotype and penicillin resistance [15].

Although recent surveillance suggests that serotype switching is a relatively rare event, the type 19A ST695 variant now constitutes the third largest genotype among US invasive serotype 19A isolates [11]. Regardless of whether or not the frequency of serotype switching recombination events are rare, certain switch variants have become prevalent. Indeed, the major serotype 19A clone shares multilocus ST199 with the clone accounting for the majority of invasive type 15B and 15C isolates, which currently account for 2–3% of invasive isolates collected in the USA and are generally penicillin-sensitive [13,14]. At this point, it is not known which strain is the precursor of the other. The major ST199 19A strain, which is typically intermediately penicillin resistant, may have arisen from a serotype switch event in a type 15B recipient. It is also possible that a penicillin-sensitive ST199 serotype 15B strain served as a donor of the cps locus and ‘sensitive’ pbp alleles introduced into a serotype 19A strain. We suspect that we have detected an analogous serotype switch variant that corresponds to this latter hypothetical example. Within our serotype 19A genetic surveillance, we have detected two putative serotype switch variants with the same multilocus ST as the Taiwan^19F-14 clone (ST236); however, these two 19A variants are unique within this ST (and within the entire large genetic complex of ST320 and other STs highly related to ST236) in that they exhibit only low level intermediate penicillin resistance, typical of the major ST199 19A strain. Analysis of the regions flanking the cps locus is ongoing in order to reveal the two chromosomal crossover points between the likely cps locus (and pbp gene) donor (the major type 19A clone [ST199]) and the Taiwan^19F-14 recipient resulting in type 19A ST236 progeny).

Owing to the relatively short length of vaccine clinical trials relative to the duration required for the pneumococcus to adapt to changes in its environment, the introduction and widespread use of a new vaccine is a highly complex experiment. Although PCV7 has demonstrated a dramatic positive impact, a small, incremental, yearly increase of replacement invasive disease caused by certain NVTs not included in PCV7 has been observed [12]. This observation coincides with reports of increased colonization and otitis media caused by NVTs [3,16,17,18]. Although there are reports that PCV7 has successfully decreased nasopharyngeal carriage of antibiotic-resistant strains [19], a more recent carriage study reported increases in previously recognized penicillin-nonsusceptible clones of NVTs 19A, 15A and 35B [17]. This information is concordant with the observation that much of the IPD in the USA currently caused by penicillin-nonsusceptible pneumococci is due to these same NVTs [12].

It is perhaps not a surprising observation that removal of pneumococcal strains covered by PCV7 from the nasopharyngeal environment could provide a more favorable niche for NVT strains, with the resultant increase in invasive disease caused by these replacement NVT strains. One intriguing possible mechanism for serotype replacement could involve the pneumococcal blp locus, which appears to encode bacteriocins that target heterologous pneumococcal strains [20]. A mutant serotype 19A strain lacking both blp locus putative bacteriocin and immunity genes was unable to compete with its parental strain, as well as a serotype 4 isolate in mouse nasopharynx co-colonization experiments. It is also possible that the increase in serotype 19A disease may have been facilitated by the high prevalence of antibiotic nonsusceptibility within this serotype, and also by the serotype being one of the most common carriage NVTs in the pre-PCV7 era. Recent data have demonstrated that post-PCV7 expansion of serotypes 19A, 15A and 35B in carriage studies is due to the preferential increase of penicillin-nonsusceptible clones [11,17]. Over time, these continued expansions may lessen the impact of PCV7 in decreasing IPD caused by penicillin-resistant strains.

It is already apparent that conjugate vaccines designed to target more serotypes are needed. In view of the increased disease burden due to serotype 19A, it is fortunate that a 13-valent conjugated vaccine (PCV13) is in a late stage of development for use in children and adults. PCV13 contains all the PCV7 serotypes, as well as serotypes 19A, 1, 3, 5, 6A and 7F. Unfortunately, certain serotypes not included within PCV13 are increasing among both carriage and IPD isolates since the introduction of PCV7. This is evident for types 15A, 15B, 22F, 23A, 33F and 35B [11,12,17]. This is of particular concern for serotypes such as 15A, 23A and 35B, which are comprised of successful widely disseminated penicillin-nonsusceptible strains [14,21,22].

PCV7 has proven efficacy in reducing IPD and its overall positive impact continues to be vast. Nonetheless, there are disturbing indications of increased replacement IPD, shifts in nasopharyngeal carriage serotype distribution and limited protection against acute otitis media (AOM) [11,17,21,23]. Another negative issue is the prohibitive cost of PCV7, which is among the most expensive vaccines of the childhood schedule in the USA [24]. The cost is problematic for developing countries, where serotype distributions often differ such that PCV7 may not provide the ideal formulation [25].

A vaccine consisting of a common pneumococcal protein or combination of proteins that would efficiently target all pneumococcal strains, if effective, may be a preferred strategy. There are several candidates, and it has been suggested that the ideal vaccine would consist of one or more proteins that protect against both nasopharyngeal colonization and systemic challenge [26]. Whether or not a protein vaccine will be preferable to conjugate vaccines is a question that will only be answered by clinical trials and long-term implementation.

Certainly, there are unknown consequences of eliminating normal flora species, or specific strains of a species, from the nasopharynx through the use of any vaccine. Recently, it was shown that AOM in mice could be induced in pneumococcal-colonized mice by subsequent influenza virus infection. AOM in this novel model could be prevented through intranasal phage-lysin treatment that eliminated or greatly reduced the colonizing pneumococci [27]. As these authors suggested, a positive effect of eliminating pneumococcal colonization through effective vaccination could include the prevention of secondary bacterial complications of influenza. In a 9-valent conjugate vaccine (PCV9) trial, a significant reduction in viral pneumonias was found in children who had received PCV9 [28]. As quoted from my colleague’s presentation [10], an implemented pneumococcal vaccine is inevitably “nature’s experiment”.

Acknowlegement

I am indebted to the Active Bacterial Core surveillance personnel (see www.cdc.gov/ncidod/dbmd/abcs), participating hospitals and laboratories, the Emerging Infections Program sites, the entire CDC Respiratory Disease Branch (RDB) Streptococcus Laboratory, and members of the RDB Epidemiology section.