Streptococcus pneumoniae Serotype 19A in Children, South Korea

A single, multidrug-resistant strain was responsible for increased incidence of this serotype before introduction of the pneumococcal 7-valent conjugate vaccine.

S treptococcus pneumoniae is a major cause of invasive infections in young infants and children. Among >90 serotypes, only a limited number account for pneumococcal diseases. Serotype incidence can vary by patient age, geographic region, and time of surveillance. Since the introduction of 7-valent conjugate vaccine (PCV7) in the United States, a decrease in the incidence of invasive pneumococcal disease (IPD) caused by vaccine serotypes has been observed in pediatric and nonpediatric age groups (1,2). However, the incidence of IPD caused by nonvaccine sero-types (including serotype 19A) increased 1.5-fold in 2002 compared to that in 1999 (2,3). To date, replacement for IPD has been observed for serotypes 3, 15, 19A, 22F, 33F, and 35, with the increase in 19A being the most prevalent. (2)(3)(4)(5)(6)(7). Recently, Singleton et al. reported that serotype 19A was responsible for 28.3% of IPD in rural Alaska Native children <2 years of age during 2004-2006 (8). In addition, recent studies from Massachusetts and Texas showed that a multidrug-resistant sequence type of serotype 19A has emerged as an important cause of IPD (9,10).
The direct effects of PCV7 in serotype distribution changes and genetic structures of pneumococcal isolates are not clear. However, capsular switching of vaccine serotypes under selective pressure by PCV7 is one of the mechanisms underlying the expansion of serotype 19A (4,(11)(12)(13).
As part of this surveillance implemented in a tertiary referral center in South Korea, all pneumococcal isolates obtained from sterile body fl uids and from various clinical specimens were subjected to serotype and antimicrobial drug susceptibility pattern determinations (14). Surveillance since 1991 shows that serotype 19A isolates were increasingly recognized among clinical isolates before PCV7 was introduced in South Korea in November 2003. We describe S. pneumoniae serotype distribution changes in Korean children during 1991-2006 with an emphasis on serogroup 19.

Patients and Pneumococcal Isolates
Hospital-wide surveillance was continued to monitor pneumococcal diseases as a part of routine clinical care at Seoul National University Children's Hospital from 1991 through 2006. Only initial isolates were included in the study; repeated isolates from the same patient were excluded. Isolates were classifi ed as invasive (e.g., blood, pleural fl uid, cerebrospinal fl uid, joint fl uid) and noninvasive (pharynx, middle ear fl uid, sputum). Methods of surveillance and of obtaining blood cultures did not change during the study period.

Multilocus Squence Typing Analysis
Multilocus sequence typing (MLST) was performed on all serotype 19A isolates (n = 58) and serotype 19F isolates (n = 68) obtained from children <5 years of age (15)(16)(17). New alleles were verifi ed by resequencing gene fragments of both strands. When sequence types (STs) that had not been associated with a particular serotype in the past were identifi ed, serotyping and sequencing typing were confi rmed by repeating the reactions. STs were divided into sets using eBURST software (Imperial College, London, UK) (18), which is available at the MLST database. eBURST sets meet the requirement that all STs must be single-locus variants (SLVs) of at least 1 other ST within the group. Founder STs were defi ned as described previously (18). Clonal complexes (CCs) consisted of eBURST sets or eBURST sets plus related STs that shared 5 of 7 allelic identities with most other STs included within an eBURST set.
According to eBURST analysis, 1 major CC (CC271) accounted for most of the isolates (n = 116, 92% of serotypes 19A and 19F). CC271 comprised 16 STs, where ST271 was predicted as the founder, and ST320 was the predominant allelic profi le. The 5 ST lineages that were unrelated to CC271 were nonlinked singlets (ST1203, ST1374, ST2394, ST2395, and ST2399) ( Figure 3). ST320 was the only ST found among serotypes 19A and 19F; ST320 was the most common ST (n = 52, 90% of total 19A isolates) among serotype 19A. ST320 was observed in only 9% (n = 6) of serotype 19F isolates. The genetic structure of serotype 19A comprised primarily ST1374 during periods 1 and 2 (1991-1997), but ST1374 isolates were not recovered after 2001. In contrast, ST320 was the most common sequence type from 1998, and all 19A isolates from 2002-2006 were of ST320. The numbers of 19A isolates increased consistently from 1996 through 2003, which suggests that single clonal expansion of ST320 was responsible for the increase of serotype 19A isolates during this period (Figure 4).

Discussion
We found that before PCV7 introduction in South Korea, the proportion of serotype 19A isolates increased from 0% in 1991 to 26% in 2003 but 19F isolates decreased during the same period. Our study also demonstrated that multidrug-resistant ST320 isolates containing mef/erm determinants were responsible for the expansion of serotype 19A.
In the United States, serotype 19A is now the most important cause of IPD by replacement serotypes (4,8,23,24). Contrary to what we report for South Korea, the increase in the United States was documented after widespread use of PCV7 (4)(5)(6). Increase in non-PCV7 serotypes, i.e., serotype replacement, has been noted in carriage studies and the prelicensure clinical trials (11,25,26). After widespread use of PCV7 in young children, replacement of serotypes for IPD has been described for several serotypes in previous studies (4)(5)(6). Of these, an increase in the incidence of IPD cases caused by serotype 19A was quite high. Therefore, the increasing prevalence of IPDs caused by serotype 19A among the vaccine target group is of considerable concern. Our fi ndings of an increase in serotype 19A disease before conjugate vaccine introduction calls into question the role vaccine may play in the emergence of serotype 19A disease and suggests that other factors are important.
Of the factors contributing to the increase in serotype 19A, the MLST fi ndings in our study point to a homogeneous pattern of ST320. ST320 of serotype 19A might have originated from ST271 or ST236 strains that have been prevalent among serotype 19F since 1993 (20) or could have been introduced from other countries. Thus, single clonal expansion of ST320, related to a multidrug-resistant internationally prevalent clone, Taiwan 19F -14 (that also carries mefA/ermB determinants), was most likely responsible for the prevaccine increase observed for serotype 19A. Antimicrobial drug use may provide selective pressure that would give this highly resistant strain an advantage over other strains. A study in the United States demonstrated an increasing prevalence of mefA/ermB; 17% percent of 221 children had been colonized by mefA/ermB containing serogroup 19 pneumococci strains after receiving at 278 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 14, No. 2, February 2008  least 1 dose of PCV7; the major clonal type was related to Taiwan 19F -14 (27). In Alaska, the increase observed in 19A colonization and IPD seems to be related to CC172 clonal expansion (8). In contrast, recent genetic analysis of 19A strains isolated after PCV7 use in the United States showed that diverse mechanisms were involved in the expansion of 19A strains, expansion of preexisting predominant CC199, capsular switching of PCV7 types (4, 6, 14, and 9V), and appearance of multiple unrelated multidrug-resistant CCs among serotype 19A strains (CC271, including ST1451 and ST320, CC156, and CC1296) (4). Capsular switching of PCV7 serotypes under selective pressure by vaccine use is one of the mechanisms underlying the expansion of serotype 19A (4,(11)(12)(13)28). However, there is no evidence of capsular switching as a contributing factor to the increase in serotype 19A in our study. We also found that serotype 19F gradually decreased in proportion and diversifi ed to include several newer descendants that differ from the founder by only 1 or 2 of 7 alleles but ST320 was not a major genetic structure among serotype 19F strains (18). Thus, it appears that ST320 has a selective advantage in serotype 19A strains, whereas ST320 did not seem to have expansion merit in the close serotype 19F strains. This fi nding suggests that the properties of particular clonal or capsular types are likely important determining factors of the potential of pneumococci to infl uence disease type and severity (29,30). Further studies are necessary to explain why certain sequence types exhibit different selective pressures according to serotype, even for close serotypes such as 19A and 19F.
This study has several limitations. First, pneumococcal isolates were collected at a single center, and thus, may not represent the national situation. However, no surveillance system has been established for IPD in South Korea. Nevertheless, at least 2 studies have demonstrated a recent increase in serotype 19A isolates among children in daycare centers (3% in 2002 and 11% in 2004) (14,31). In addition, the number of serotype 19A isolates from children and adults at another tertiary South Korean hospital showed an increase over the same period (32). Second, the number of 19A isolates was relatively small, and it is possible that uncommon strains possessing minor clones were not detected. Thus, our fi nding of a clonal expansion of ST320 among serotype 19A strains may be an overstatement. However, MLST analysis showed that all 7 of the 19A strains from colonized children who were identifi ed from a previous study (14) were of ST320 (H.J. Lee, unpub. data).
The present study has implications for future pneumococcal immunization programs. In particular, the demonstration of an increase in 19A before the use of PCV7 suggests that PCV7 vaccination may not be entirely responsible for the observed increase of serotype 19A. Had the vaccine been introduced in 2000, as in the United States, the increase of serotype 19A would have been attributed to serotype replacement after PCV7 introduction. Nevertheless, whether minor multidrug-resistant clones of serotype 19A (appearing after the introduction of PCV7 in the United States) possess an advantage to increase with time is a concern (23,29).
We demonstrated that multidrug-resistant ST320 strains among serotype 19A have selective advantage in terms of expansion over ST1374, which were less resistant to β-lactams in a country where antimicrobial drug therapy is frequently used. Reports on PCV7 effi cacy indicate negligible cross-protection for serotype 19A (11). In addi- tion, a recent increase in the antimicrobial drug resistance of invasive 19A isolates and the increase in colonization by serogroup 19 strains carrying mef/erm determinants raise the possibility of potential increases in the prevalence of this clone. Thus, potential for colonization because of widespread antimicrobial drug use and resistance may interact and provide the selective advantage necessary for serotype expansion, which may be the situation in South Korea. Similarly, the infl uence of population characteristic dynamics, i.e., HIV infection or poverty and overcrowding, as well as PCV7 introduction, may combine with the necessary factor for serotype replacement and play an important role in serotype expansion, which may be the situation in Alaska. It is too early to determine the effect of PCV7 on expansion of serotype 19A in South Korea. PCV7 was introduced in South Korea in November 2003 when PCV7 serotype coverage was 56% among invasive isolates in children <5 years of age. In 2007, PCV7 coverage is ≈30% of the target group (21). At this time it is diffi cult to predict the effect of low vaccination coverage on serotype expansion/replacement. Therefore, surveillance is essential to monitor antimicrobial drug resistance, serotype expansion, and serotype replacement as early indications of an increase in pneumococcal disease by non-PCV7 or PCV7related serotypes.  (14), ST236 (14), ST283 (7), ST1451 (1), ST1464 (10), ST2395 (3), ST2695 (3), and 1 isolate of each of ST1412, ST1417, ST2396, ST2397, ST2398, ST2694, ST2696, ST2697, and ST2698.