ANTIMICROBIAL AGENTS AND CHEMOTHERAPY, Sept. 2008, p. 3216–3220 0066-4804/08/$08.00⫹0 doi:10.1128/AAC.00358-08 Copyright © 2008, American Society for Microbiology. All Rights Reserved.

Vol. 52, No. 9

Temporal and Spatial Distribution of Clonal Complexes of Streptococcus pneumoniae Isolates Resistant to Multiple Classes of Antibiotics in Belgium, 1997 to 2004䌤 Heather Amrine-Madsen,1* Johan Van Eldere,2 Robertino M. Mera,1 Linda A. Miller,3 James A. Poupard,4 Elizabeth S. Thomas,3 Wendy S. Halsey,3 Julie A. Becker,3 and F. Patrick O’Hara3 GlaxoSmithKline, 5 Moore Drive, Research Triangle Park, North Carolina 277091; University Hospital Gasthuisberg, Herestraat 49, B-3000 Leuven, Belgium2; GlaxoSmithKline, 1250 S. Collegeville Road, Collegeville, Pennsylvania 194263; and Pharma Institute of Philadelphia, 3612 Earlham St., Philadelphia, Pennsylvania 191294 Received 14 March 2008/Returned for modification 16 April 2008/Accepted 17 June 2008

We performed multilocus sequence typing on 203 invasive disease isolates of Streptococcus pneumoniae to assess the clonal compositions of isolates from two provinces in Belgium and to determine the relationship between clones and antibiotic nonsusceptibility, particularly nonsusceptibility to two or more classes of antibiotics. The frequency of multiclass nonsusceptibility (MCNS) was higher in the province of West Flanders (38%) than in Limburg (21%). This difference was largely attributable to five clonal complexes (CC156, CC81, CC143, CC193, and CC1848), which contained high proportions of isolates with MCNS (>47%) and which were circulating at higher frequencies in West Flanders. The S. pneumoniae population changed over time, as CC156 and CC81 declined in frequency from 1997 to 1999 to 2001 to 2004. Over the same time period, the frequency of pneumococcal conjugate vaccine 7 (PCV7) serotypes dropped from 69% to 41%. In contrast, the nonvaccine serotype 19A increased in frequency from 2.1% to 6.6%. None of these changes can be attributed to PCV7 vaccine, as it was not in use in Belgium during the time period studied. There was evidence that MCNS clones flowed from West Flanders to Limburg. opportunity to study temporal trends in a country before the widespread use of the PCV7 vaccine. The present study used multilocus sequence typing (MLST) on a random sample of S. pneumoniae invasive disease isolates from West Flanders and Limburg to determine the clonal composition of the population over the time period 1997 to 2004. We focus on multiclass-nonsusceptible (MCNS) isolates, isolates that are nonsusceptible to two or more classes of antibiotics. These data were combined with information on serotypes and geographic origins of isolates to address several questions about the S. pneumoniae population.

Streptococcus pneumoniae is a significant human pathogen, and antibiotic-resistant strains are increasingly problematic (10, 12). Evolution of S. pneumoniae clones has become an especially important topic since the introduction of the heptavalent pneumococcal conjugate vaccine 7 (PCV7) in the United States in 2000. Particular attention has been given to the rise of replacement clones that increase in frequency after targeted serogroups decline (2) and to capsular switching (7), when clones recombine to change their capsular serotypes. In particular, the incidence of disease due to serotype 19A, which is not targeted by PCV7, appears to be increasing in the United States (9). In 1985, an ongoing S. pneumoniae surveillance study began in Belgium that captured information on serotype, antibiotic susceptibility, geographic location, and patient characteristics. Studying the clonal structure of S. pneumoniae isolates in Belgium offers a unique opportunity because Belgium is located among countries with very high nonsusceptibility to two or more antibiotic classes, such as France (52%); countries with an intermediate resistance pattern, such as Luxembourg (14.8%); and countries with very low multiple resistance, such as Germany (8.1%) and The Netherlands (1.3%). The provinces of West Flanders, which borders France, and Limburg, which borders The Netherlands, were selected because they differed in the prevalence of risk factors for developing resistance to multiple classes of antibiotics including antibiotic consumption and population density (15). Belgium also offers the

MATERIALS AND METHODS The Belgian S. pneumoniae national reference laboratory provided the isolates, which had been stored at ⫺80°C. Two provinces in Belgium and four time points were selected for the study. The selected provinces are not contiguous and border France (West Flanders) and The Netherlands (Limburg). They represent extremes of penicillin and macrolide resistance in the country. Isolates were randomly selected proportional to the sampling distribution of location and time in the surveillance database. A total of 203 isolates were chosen: 42 isolates were from 1997, 55 from 1999, 53 from 2001, and 53 from 2004. All of the S. pneumoniae isolates were from invasive infections, including blood, other sterile body fluids, and middle ear fluid. Over the period 1997 to 2004, 72% of isolates were collected from blood, 18% from middle ear fluid, 5% from cerebrospinal fluid, 1% from pleural fluid, and 3% from other sources, including peritoneal fluid, joints, etc. Over the time period included, the percentage of isolates collected from blood increased while the percentage from middle ear fluid decreased. Upon arrival at the reference laboratory, all isolates were reidentified and serogrouped according to the Kopenhagen scheme. Susceptibility to the following antibiotics was determined via agar diffusion: erythromycin, oxacillin, tetracycline, cefotaxime, and ofloxacin. For each isolate, in addition to serotype and antibiotic susceptibility, the following data were available: sample origin and collection date and patient data (age, sex, and address). MCNS was defined as nonsusceptibility to two or more antibiotic classes (␤-lactams, macrolides, tetra-

* Corresponding author. Mailing address: 5 Moore Drive, 3-3244C, Research Triangle Park, NC 27709. Phone: (919) 483-1542. Fax: (919) 315-032. E-mail: [email protected]. 䌤 Published ahead of print on 23 June 2008. 3216

VOL. 52, 2008

CLONAL COMPOSITION OF S. PNEUMONIAE IN BELGIUM

cyclines, cephalosporins, or quinolones). The nonsusceptibility designation was based on Clinical and Laboratory Standards Institute breakpoints according to the most recent document available at that time. Since breakpoints for the antibiotics tested and testing methods have not changed over the years, the data remain comparable. MLST. MLST was performed according to the protocol of Enright and Spratt (4), with some modifications of the PCR conditions. The following primers were used in PCR and sequencing: aroe1F (CACTGCGGATGTGACTGGTTCGA), aroe2R (C CCTCAATAATAGCTGTTAGACGGGG), aroe3F (GATGGCTATACACGTTT AGCTGCAG), aroe3R (CCAAGTAGTCTTTCATCTCTTCCAGA), aroeint1F (GACCTTACCTAGACAAGTTACAGG), aroeint1R (CCATCTTGCGCTTCAC CTGACAAG), ddl5F (GCMCAAGTTCCTTATGTGGCTATCG), ddl5R (GTA GTGGGTACATAGACCACTGGG), ddlint1F (GAGGATAGTAGAGATGTG GCAGC), ddlint1R (CGTGCTCTTGACATCGTAGTTACC), gdh2F (TGTTAC AATTTTCGGTGCGAGTG), gdh2R (TCTAAGCGACCATCTTGACGATA), gdhint3F (GGTGTAGAAGAACGTGGTGG), gdhint2R (CCACCACGTTCTTC TACAAC), gki1F (CACGCAAACCTTTGCATAAGTGA), gki2R (ACAAGTGA TGCTGCTCCGATAAC), gki2F (TATTGGGATTGACCTTGGTGGAA), gkiint1F (CTTCGTATTCATCGCCATAGC), gkiint1R (CGCAGAAGGCAAA TTGCTTCA), recP2F (CTGCCAATACTTTTGGTGCTGGG), recP1R (CTTT GGAGGATTTCCGTATGTTG), recP4F (GGAAGTTGGCGGTAAGTACAC TTATCC), recP4R (CAAGGCCATCATACCTACAAGCATG), recP5F (CTT AGTAGAATACCATCATCCACGTGG), recPint3R (CTGTGACTTATGTCT TTACCC), spi1F (GATAGAAGAAGAGGCTGAGATTGGT), spi1R (CAAT CTCACGGCTGAGCTGAGTT), spi3F (GTTCCGATACGGGTGATTGGC CA), spi3R (CAAGTATCACACTCACACCAAGCG), spiint1F (GATATAGT CTGCTAGATAAGGCTCG), spiint1R (CAGATGAGGTGGGAAATGGTTC CT), xpt1F (GTCTATGATACCACTACAACGGGA), xpt1R (CGGCATTGA GGACAATAGCGAGT), xpt2F (GATAAGACTCGCTTCGCTCGTAAT), xptint1R (CGCGTGCTGTCGATTGGATCTTTT), xpt3F (GAAATTATTAGA AGAGCGCATCCTC), and xpt3R (CTGTCGGCATGCTGACAAAGAGAT). The PCR products were purified using the QIAquick PCR purification kit (Qiagen, Valencia, CA) following the manufacturer’s instructions. Direct sequencing of purified PCR products was performed with BigDye terminator cycle sequencing kit and a Genetic analyzer 3730XL (both from Applied Biosystems, Foster City, CA). Contigs were assembled using Sequencher 4.5 (Gene Codes, Ann Arbor, MI), homologous DNA sequence alignments were generated by ClustalX 1.83 (14), and sequences were analyzed and manually edited in GeneDoc 2.6 (11). After alignment and analysis, the sequences for each locus were compared to alleles on the S. pneumoniae MLST public database (http: //www.mlst.net) and given a corresponding allele number. Thus, for each isolate, a seven-number allelic profile was generated and compared to the public profiles. Following the MLST scheme, a sequence type (ST) was then given to every unique allelic profile. Clonal complexes were determined using eBURST V3, with the minimum number of alleles shared set to 6. Statistics. All categorical data were analyzed with a chi-square test or Fisher’s exact test, where appropriate. Serotyping. Initial serogrouping was performed at the reference laboratory. Isolates within the PCV7-related serogroups 9, 18, 19, 6, and 23 were chosen for subtyping, because some subtypes within these serogroups are targets of PCV7, while other subtypes are not. Subtypes were determined using the Quellung reaction with antibodies obtained from MiraVista Diagnostics (Indianapolis, IN).

RESULTS The 203 isolates analyzed using MLST comprised 97 STs, including 25 STs not found in the MLST database. All of the novel STs were single- or double-locus variants of STs in the database. The eBURST algorithm described 18 clonal complexes consisting of 55 STs and 129 isolates. We included the three largest singleton clones as additional clonal complexes, bringing the total to 21 clonal complexes encompassing 150 isolates. The remaining 53 isolates were singletons. The clonal complex with the largest number of isolates in this study was CC156, which accounted for 16.7% of all isolates. The 10 largest clonal complexes in our study and their associated serogroups/serotypes are shown in Table 1. Overall, 49.3% (n ⫽ 100) of the sampled isolates were nonsusceptible to at least one of five antibiotics (oxacillin, eryth-

3217

TABLE 1. The 10 most commonly observed clonal complexes in this study and their associated serotypes Clonal complex

Serotype(s) (n)

CC15 .....................................................14 (11), 19F (1) CC53 .....................................................8 (9) CC81 .....................................................23F (16) CC156 ...................................................9V (16), 14 (16), 19F (1), 38 (1) CC177 ...................................................19F (5) CC180 ...................................................3 (10) CC191 ...................................................7 (9) CC304 ...................................................1 (5) CC350 ...................................................1 (7) CC439 ...................................................23F (7), 23B (1)

romycin, tetracycline, cefotaxime, or ofloxacin). In both Limburg and West Flanders, the majority of isolates that were nonsusceptible to antibiotics were MCNS isolates nonsusceptible to two or more classes of antibiotics. Forty-five isolates were nonsusceptible to two classes; most of these (n ⫽ 24) were nonsusceptible to erythromycin and tetracycline. Nineteen isolates were nonsusceptible to three classes; most of these (n ⫽ 17) were nonsusceptible to erythromycin, tetracycline, and oxacillin. Five isolates were nonsusceptible to four classes; four of these were nonsusceptible to cefotaxime, oxacillin, erythromycin, and tetracycline. The one isolate that was resistant to ofloxacin was also nonsusceptible to cefotaxime, oxacillin, and erythromycin. None of the sampled isolates was nonsusceptible to all five classes. Of 127 isolates sampled in West Flanders, 56 (44%) were susceptible, 23 (18%) were nonsusceptible to one class, and 48 (38%) were MCNS. Of 76 isolates from Limburg, 47 (62%) were susceptible, 13 (17%) were nonsusceptible to one class, and 16 (21%) were MCNS. The frequency of MCNS isolates was significantly higher in West Flanders than in Limburg (P ⫽ 0.01). MCNS isolates were found in 31 different STs and 9 clonal complexes (Table 2). Half of all MCNS isolates were found in two clonal complexes: CC156 and CC81. Of the 16 MCNS isolates from CC156, 10 were serotype 14 and 6 were serotype 9V. All 16 MCNS isolates from CC81 were serotype 23F. Of all MCNS isolates, 82.8% were PCV7 serotypes (Table 3), 14.1% were PCV7-related serotypes (serotypes 19A and 9L/N), and 3.1% were non-PCV7-related serogroups (serogroup 1). The PCV7 serotypes 23F and 14 were most frequently observed, with 23F accounting for 26.6% of MCNS isolates and serogroup 14 accounting for 25%. Among nonPCV7 serotypes, 19A was the most commonly observed, accounting for 9.2% of MCNS isolates. The increased frequency of MCNS in West Flanders versus Limburg is largely attributable to the clonal complexes circulating in the two provinces (Table 4). While 18 of 21 clonal complexes were found in both provinces, they were found at different frequencies. Five clonal complexes (CC156, CC81, CC143, CC193, and CC1848) had high levels of MCNS (⬎47%). The combined frequency of those clonal complexes was 37.0% in West Flanders but only 19.7% in Limburg (P ⫽ 0.012). Among the 203 sampled isolates, PCV7 serotypes were more frequent in West Flanders (56.7%) than in Limburg (51.3%), but the difference was not statistically significant (P ⫽ 0.47).

3218

AMRINE-MADSEN ET AL.

ANTIMICROB. AGENTS CHEMOTHER.

TABLE 2. Distribution of MCNS isolates among clonal complexes

TABLE 4. Distribution of MCNS clonal complexes by province No. (%) of clonal complexes

Clonal complex

Total no. of isolates

% of MCNS isolates within clonal complex

% of MCNS isolates

Type of clonal complex

CC156 CC81 CC15 CC350 CC177 CC143 CC1848 CC193 CC66 CC180 CC53 CC191 CC439 CC304 CC247 CC205 CC62 CC97 CC289 CC113 CC235

34 16 12 7 5 4 4 4 4 10 9 9 7 5 4 4 3 3 2 2 2

47.1 100 16.7 28.6 40 100 75 75 25 0 0 0 0 0 0 0 0 0 0 0 0

24.6 24.6 3.1 3.1 3.1 6.2 4.6 4.6 1.5 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0

High levels of MCNS (CC156, CC193, CC1848, CC143, and CC81) Low levels of MCNS (CC177, CC15, CC66, and CC350) No MCNS

Changes in the S. pneumoniae population over time. The clonal composition of the S. pneumoniae population changed over time (Fig. 1). The two largest clonal complexes decreased significantly in frequency between 1997 to 1999 and 2001 to 2004. CC156 and CC81 accounted for 25.8% and 13.4%, respectively, of isolates sampled in 1997 to 1999. By 2001 to 2004, the frequencies had dropped to 8.5% and 2.8% (P ⫽ 0.001 and P ⫽ 0.008, respectively). The decrease in frequency of CC156 and CC81 resulted in a shift in the population structure of clones in Belgium, as these high-frequency clones were replaced with more-moderate-frequency clones, while the overall number of clonal complexes increased from 17 to 20. The change in the S. pneumoniae population was also seen in the distribution of serotypes. PCV7 and PCV7-related serotypes accounted for 69.1% of all isolates in 1997 to 1999, but the frequency dropped to 40.6% in 2001 to 2004 (P ⬍ 0.005). In contrast, serotype 19A increased in frequency, from 2.1% in

TABLE 3. Serotypes with MCNS isolates No. of isolates with serotype

% of MCNS isolates within serotype

% of total MCNS isolates

PCV7 14 19F 23F 6B 9V

33 11 28 11 16

48.5 54.5 60.7 72.7 37.5

26.2 9.2 26.2 12.3 9.2

PCV7 related 9L/N 19A

4 12

25.0 75.0

1.5 12.3

Non-PCV7 related 1

15

13.3

3.1

No MCNS

73

0

Serotype

0

West Flanders

Limburg

Total

47 (37.0)

15 (19.7)

62 (30.5)

14 (11.0)

14 (18.4)

28 (13.8)

66 (52.0)

47 (61.8)

113 (55.7)

1997 to 1999 to 6.6% in 2001 to 2004. These changes in clonal composition occurred in similar fashions in both Limburg and West Flanders. These changes in the S. pneumoniae population occurred in both the pediatric (under age 5 years) and nonpediatric populations. In 1997 to 1999, CC156 and CC81 accounted for 51.5% (n ⫽ 33) of isolates collected from the pediatric population and 32.8% (n ⫽ 64) of isolates collected from the nonpediatric population. Those frequencies had declined to 20.8% (n ⫽ 24) and 8.6% (n ⫽ 81) in 2001 to 2004 (P ⫽ 0.03 and P ⫽ 0.0003, respectively). Similarly, the frequency of PCV7-related serogroups declined from 93.9% to 70.8% in the pediatric population (P ⫽ 0.03) and from 69.2% to 50.6% in the nonpediatric population (P ⫽ 0.03). Age data was missing for one patient. The frequencies of CC156 and CC81 and of PCV7 serogroups declined in all sources of infection. In blood isolates, the frequency of CC156 and CC81 declined significantly from 34.4% (n ⫽ 61) in 1997 to 1999 to 9.4% (n ⫽ 86) in 2001 to 2004 (P ⫽ 0.0003), and the frequency of PCV7 serogroups declined significantly from 75.4% to 52.3% (P ⫽ 0.006). In middle ear fluid and other sources of infection, the frequencies of CC156 and CC81 and PCV7 serogroups declined over time, but the results were not statistically significant, due in part to small sample sizes. In middle ear fluid, the frequency of CC156 and CC81 declined from 56.5% (n ⫽ 23) to 23.1% (n ⫽ 13), while in other sources the frequency declined from 30.7% (n ⫽ 13) to 14.2% (n ⫽ 7). In middle ear fluid, the frequency of PCV7 serogroups declined from 87.0% (n ⫽ 23) to 76.9% (n ⫽ 13). In other sources, the frequency of PCV7-related serogroups declined from 69.2% (n ⫽ 13) to 57.1% (n ⫽ 7). Despite the changes in the clonal composition observed over time, the frequency of MCNS isolates remained stable. In 1997

FIG. 1. Changes in the clonal composition of isolates from Limburg and West Flanders over time.

VOL. 52, 2008

CLONAL COMPOSITION OF S. PNEUMONIAE IN BELGIUM

TABLE 5. Years in which clonal complexes harboring MCNS first appeared in West Flanders and Limburga Clonal complex

CC15 CC81 CC156 CC177 CC350 CC66 CC193 CC1848 CC143

Yr clonal complex first appeared in:

Yr MCNS isolate first appeared in:

Limburg

West Flanders

Limburg

West Flanders

1999 1999 1997 1997 2001 1999 1999 2001 NA

1997 1997 1997 2004 1997 1997 1999 2001 1999

2004 1999 1999 2001 NA NA 2004 NA NA

2001 1997 1997 NA 2004 1997 1999 2001 1999

a Also shown are years in which MCNS isolates within those clonal complexes first appeared in West Flanders and Limburg. The earlier year in which an MCNS clonal complex or isolate was sampled is shown in italics. NA, not applicable because the clonal complex or MCNS isolate was never found in that province.

to 1999, 34.0% of all isolates (n ⫽ 97) were MCNS. In 2001 to 2004, 29.2% were MCNS (n ⫽ 106, P ⫽ 0.54). In West Flanders, the frequencies were 41.3% (n ⫽ 63) in 1997 to 1999 and 34.3% (n ⫽ 64) in 2001 to 2004 (P ⫽ 0.37). In Limburg, the frequencies were 20.6% (n ⫽ 34) in 1997 to 1999 and 21.4% (n ⫽ 42) in 2001 to 2004 (P ⫽ 1.0). There is evidence that clonal complexes and MCNS isolates flowed from West Flanders to Limburg (Table 5). Of the nine clonal complexes that harbored MCNS isolates, five were first isolated in West Flanders, while only one was first isolated in Limburg. Within those clonal complexes, MCNS isolates were first observed in West Flanders eight times, compared to just one time in Limburg. DISCUSSION It is essential to genetically characterize pneumococcal strains within specific geographic regions to determine the relative importance of various clones and patterns of change within the pneumococcal population. The S. pneumoniae population in Belgium is highly dynamic, and its clonal composition varies over time and place. A striking observation is the substantial decrease in the frequency of PCV7 serogroups in 2001 to 2004 compared to 1997 to 1999 in the absence of any significant vaccination. This change was accompanied by a decrease in the frequencies of CC156 and CC81, which are associated with the international antibiotic-resistant clones Spain9V-3 and Spain23F-1, respectively (7). This cannot be attributed to the introduction of the PCV7 vaccine, because the vaccine was not introduced in Belgium until October 2004. The vaccine was not introduced in neighboring Germany or The Netherlands until 2005. In France, gradual introduction of the vaccine in at-risk children started in 2002, with approximately 41% of these at-risk children having received three doses by the end of 2004 (4). Indeed, the observed change in Belgium is remarkably similar to the observed change in the population of S. pneumoniae in the United States, where the PCV7 vaccine was introduced in 2000. For example, among children with invasive pneumococcal disease in Utah, disease in 73% was caused by PCV7 serogroups in 1997 to 2000,

3219

compared with 50% in 2001 to 2003 (3). In Belgium, 76% of isolates were from PCV7 serogroups in 1997 to 1999, compared to 57% in 2001 to 2004. An analysis of bacteremic isolates from 1994 to 2004 showed a decrease in PCV7 serogroups in Belgium (6), although in that case, the decline was limited to nonpediatric patients. In the absence of a vaccine effect in Belgium, the observed changes in the pneumococcal population could have several possible explanations. Vaccine effects in other countries could be spreading across national borders more rapidly than anticipated. Here it is notable that changes in the pneumococcal population in Belgium were observed among pediatric patients (less than 5 years of age) as well as nonpediatric patients. While PCV7 vaccination is targeted to children under the age of five, a herd immunity effect of PCV7 vaccination of children appears to reduce the disease load among adults in the United States (16). Alternatively, genetic drift or natural selection driven by forces other than human antibacterial strategies could change the population structure. Other public health strategies could have influenced the S. pneumoniae population in Belgium. For instance, public health initiatives in Belgium resulted in a decline in antibiotic sales between 2000 and 2002 (1). However, this hypothesis does not fit our observation that the frequency of MCNS did not decrease over time. Whatever the causes of changes in the S. pneumoniae population, they could affect the interpretation of vaccine effects after a conjugate vaccine is introduced in a region. The presence of significant numbers of isolates from serotype 19A is particularly striking. Its appearance in the United States had been considered a response to declining numbers of disease incidents due to PCV7 serotypes. Again, the frequency change in Limburg and West Flanders, where serotype 19A accounted for 2.1% of infections in 1997 to 1999 and increased to 6.6% in 2001 to 2004, resembles, to a degree, changes observed in the United States over a similar time period. For example, Hicks et al. (9) report that serotype 19A accounted for 3.3% of infections in 1998 to 1999 and increased to 17.1% in 2004. Its presence in Belgium and the fact that it has increased in frequency between 1997 and 2004 have some important ramifications. It may be more evidence that the S. pneumoniae population is prone to changes in its structure for reasons that are not apparent. It may be evidence that replacement serotypes that arise due to selection pressure in other countries can rapidly spread worldwide. In any case, the presence of a genetically diverse, MCNS population of serotype 19A in Belgium should be considered when planning public health strategies there. It is interesting to note that the level of MCNS did not drastically fall over time, despite the changes in the S. pneumoniae population in Belgium. In 1997 to 1999, CC156 and CC81 combined to account for 68% of all MCNS isolates. When those clonal complexes declined in frequency, the frequency of MCNS observed in other clonal complexes increased. By 2001 to 2004, CC156 and CC81 accounted for only 32% of all MCNS isolates. Thus, while the frequency of MCNS remained relatively stable, the clonal complexes harboring MCNS isolates were in flux. Belgium borders countries with different levels of antibiotic resistance and MCNS strains. This helps to explain the different levels of MCNS observed in West Flanders, which borders

3220

AMRINE-MADSEN ET AL.

France, and Limburg, which borders The Netherlands. Since 18 of 21 clonal complexes in Belgium are found in both provinces, it does not appear that the difference is explained by different clones entering the provinces from bordering nations. Rather, the frequency of clones with high levels of MCNS differs between the two provinces. Whether serotypes or clones are better suited as the unit of pneumococcal epidemiology analysis remains a matter of debate (13). Data from different studies support the existence of differences in epidemiology between pneumococcal serotypes. These differences are apparent in relation to age, carriage, and disease but also in relation to antibiotic resistance (8). In this study, clones were a better indicator of geographic differences in the frequency of MCNS than serotypes. In conclusion, the clonal composition of the S. pneumoniae population in Belgium appears to be dynamic and changing over time. The decrease in the frequency of PCV7 serotypes, as well as the increase in the frequency of replacement serotype 19A, resembles changes that occurred in the United States after the use of the PCV7 vaccine became widespread. Also, this study indicates that MLST clonal complexes are useful markers of the prevalence of MCNS S. pneumoniae isolates within a region.

ANTIMICROB. AGENTS CHEMOTHER.

4.

5.

6.

7.

8. 9.

10.

11. 12.

ACKNOWLEDGMENTS We thank the GlaxoSmithKline United States-based sequencing facility, under the management of Ganesh Sathe, for its efforts on this project. We also thank Jan Verhaegen from the Belgian National Reference Laboratory for assistance with this project.

13.

14.

REFERENCES 1. Bauraind, I., J. M. Lopez-Lozano, A. Beyaert, J. L. Marchal, B. Seys, F. Yane, E. Hendrickx, H. Goossens, P. M. Tulkens, and L. Verbist. 2004. Association between antibiotic sales and public campaigns for their appropriate use. JAMA 292:2468–2470. 2. Beall, B., M. C. McEllistrem, R. E. Gertz, Jr., S. Wedel, D. J. Boxrud, A. L. Gonzalez, M.-J. Medina, R. Pai, T. A. Thompson, L. H. Harrison, L. McGee, and C. G. Whitney for the Active Bacterial Core Surveillance Team. 2006. Pre- and postvaccination clonal compositions of invasive pneumococcal serotypes for isolates collected in the United States in 1999, 2001, and 2002. J. Clin. Microbiol. 44:999–1017. 3. Byington, C. L., M. H. Samore, G. J. Stoddard, S. Barlow, J. Daly, K.

15.

16.

Korgenski, S. Firth, D. Glover, J. Jensen, E. O. Mason, C. K. Shutt, and A. T. Pavia. 2005. Temporal trends of invasive disease due to Streptococcus pneumoniae among children in the intermountain west: emergence of nonvaccine serogroups. Clin. Infect. Dis. 41:21–29. Dubos, F., I. Marechal, M. O. Husson, C. Courouble, M. Aurel, and A. Martinot for the Hospital Network for Evaluating the Management of Common Childhood Diseases. 2007. Decline in pneumococcal meningitis after the introduction of the heptavalent-pneumococcal conjugate vaccine in northern France. Arch. Dis. Child. 92:1009–1012. Enright, M. C., and B. G. Spratt. 1998. A multilocus sequence typing scheme for Streptococcus pneumoniae: identification of clones associated with serious invasive disease. Microbiology 144:3049–3060. Flamaing, J., J. Verhaegen, J. Vandeven, N. Verbiest, and W. E. Peetermans. 2008. Pneumococcal bacteraemia in Belgium (1994–2004): the pre-conjugate vaccine era. J. Antimicrob. Chemother. 61:143–149. Gherardi, G., L. Fallico, M. Del Grosso, F. Bonanni, F. D’Ambrosio, R. Manganelli, G. Palu `, G. Dicuonzo, and A. Pantosti. 2007. Antibiotic-resistant invasive pneumococcal clones in Italy. J. Clin. Microbiol. 45:306–312. Hausdorff, W. P., D. R. Feikin, and K. P. Klugman. 2005. Epidemiological differences among pneumococcal serotypes. Lancet Infect. Dis. 5:83–93. Hicks, L. A., L. H. Harrison, B. Flannery, J. L. Hadler, W. Schaffner, A. S. Craig, D. Jackson, A. Thomas, B. Beall, R. Lynfield, A. Reingold, M. M. Farley, and C. G. Whitney. 2007. Incidence of pneumococcal disease due to non-pneumococcal conjugate vaccine (PCV7) serotypes in the United States during the era of widespread PCV7 vaccination, 1998–2004. J. Infect. Dis. 196:1346–1354. Musher, D. M. 1992. Infections caused by Streptococcus pneumoniae: clinical spectrum, pathogenesis, immunity, and treatment. Clin. Infect. Dis. 14:801– 807. Nicholas, K. B., H. B. Nicholas, and D. W. Deerfield II. 1997. Genedoc: analysis and visualization of genetic variation. embnet.news 4:14. Poole, M. D. 1995. Otitis media complications and treatment failures: implications of pneumococcal resistance. Pediatr. Infect. Dis. J. 14:S23–S26. Sandgren, A., K. Sjostrom, B. Olsson-Liljequist, B. Christensson, A. Samuelsson, G. Kronvall, and B. Henriques Normark. 2004. Effect of clonal and serotype-specific properties on the invasive capacity of Streptococcus pneumoniae. J. Infect. Dis. 189:785–796. Thompson, J. D., D. G. Higgins, and T. J. Gibson. 1994. CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position specific gap penalties and weight matrix choice. Nucleic Acids Res. 22:4673–4680. Van Eldere, J., R. M. Mera, L. A. Miller, J. A. Poupard, and H. AmrineMadsen. 2007. Risk factors for development of multiple-class resistance to Streptococcus pneumoniae strains in Belgium over a 10-year period: antimicrobial consumption, population density, and geographic location. Antimicrob. Agents Chemother. 51:3491–3497. Whitney, C. G., M. M. Farley, J. Hadler, L. H. Harrison, N. M. Bennett, R. Lynfield, A. Reingold, P. R. Cieslak, T. Pilishvili, D. Jackson, R. R. Facklam, J. H. Jorgensen, and A. Schuchat for the Active Bacterial Core Surveillance of the Emerging Infections Program Network. 2003. Decline in invasive pneumococcal disease after the introduction of protein-polysaccharide conjugate vaccine. N. Engl. J. Med. 348:1737–1746.