Prevalence of Methicillin-Resistant Staphylococcus aureus Nasal Colonization among Taiwanese Children in 2005 and 2006

MATERIALS AND METHODS

This study was approved by the Institutional Review Board of the Chang Gung Memorial Hospital. From July 2005 to October 2006, all children of ages between 2 months and 5 years who presented for a well-child health care visit to any one of three medical centers in Taiwan were invited to participate in this study. The three medical centers involved were the Chang Gung Children's Hospitals at Linko (hospital A) and Kaohsiung (hospital C) and Taichung Veterans General Hospital (hospital B), which are situated, respectively, in northern, southern, and central parts of Taiwan. In each hospital, around 80 subjects were recruited for study for each month, and the ages of the subjects were evenly distributed in seven separate age ranges, which included >2 to 6 months, >6 to 12 months, >12 to 18 months, >18 to 24 months, >2 to 3 years, >3 to 4 years, and >4 to 5 years. A culture from the anterior nares for the detection of MRSA was obtained from each subject after written consent was obtained from their parents/guardians.

Survey specimens for culture were obtained with a cotton swab, placed in the transport medium (Venturi Transystem; Copan Innovation Ltd., Limerick, Ireland), and then brought to and processed in the microbiological laboratories within 4 hours of the sampling. All S. aureus isolates were sent to Chang Gung Memorial Hospital at Linko for microbiological characterization. Identification of MRSA was confirmed according to Clinical and Laboratory Standards Institute 2005 guidelines (5). Pulsed-field gel electrophoresis (PFGE) with SmaI digestion was used in this study to fingerprint the MRSA isolates and was performed according to procedures described previously (3, 16, 18). The genotypes were designated in alphabetical order, as in our previous studies (3, 15-18); any new genotype, if identified, was designated consecutively. PFGE patterns with fewer than four band differences from an existing genotype were defined as subtypes of that genotype.

SCCmec typing of isolates was done using a multiplex PCR strategy described previously (26). Control strains for SCCmec types I, II, III, and IVa, kindly provided by Keiichi Hiramatsu, were as follows: type I, NCTC10442; type II, N315; type III, 85/2082; and type IVa, JCSC4744. SCCmec typing for type V_T was determined by using a particular primer described elsewhere (1), and strain TSGH-17, kindly provided by Chi-Chien Wang, was used as a control. However, the SCCmec typing method for type V_T yielded inconsistent results; thus, an alternative method was used. The appearance of an isolate with only two bands (414 bp and 243 bp) in the multiplex PCR analysis may have indicated that the isolate contained SCCmec V_T. To confirm their identities, a novel pair of primers, ccrC-5F (5′-CAC TTA ATC CAT TGA CAC AG-3′) and ccrC-5R (5′-AAA GAT TGA GGG ATA AGA CT-3′), was designed according to the published sequence (GenBank accession no. AY894416) of the ccrC gene of a Taiwanese strain, S. aureus TSGH-17. Amplification of a specific 1,081-bp DNA fragment, which was subjected to further sequence analysis for some representative isolates in preliminary experiments, confirmed that the isolates contained SCCmec V_T.

The presence of Panton-Valentine leukocidin (PVL) genes was determined by a PCR strategy described previously (22). Some isolates of representative PFGE patterns were selected and underwent multilocus sequence typing (MLST) as described elsewhere (8). The allelic profiles were assigned through comparison of the sequences at each locus with those of the known alleles in the S. aureus MLST database and were defined as sequence types accordingly.

RESULTS

During the study period, 1,279 subjects were recruited from hospital A, 1,011 subjects from hospital B (from July 2005 to June 2006), and 756 subjects from hospital C (from October 2005 to June 2006). All the children enrolled are Taiwanese. The number of subjects enrolled in each age group ranged from 430 for children of ages >6 to 12 months to 443 for children of ages >2 to 3 years. Of the total of 3,046 subjects enrolled in this study, 713 (23%) were colonized with S. aureus. Of the 713 isolates, 221 (31%) were demonstrated to be MRSA. The details of the nasal MRSA colonization prevalence for subjects in the different parts of Taiwan are shown in Table 1. The MRSA colonization rate in northern Taiwan was significantly higher than that in the central (P < 0.001) and southern (P < 0.039) parts of Taiwan. The nasal MRSA colonization prevalences for the subjects in each age group were 8.4% for the children of ages >2 to 6 months and 6.3%, 3.2%, 3.9%, 9.0%, 9.5%, and 10.1% for children of ages >6 to 12 months, >12 to 18 months, >18 to 24 months, >2 to 3 years, >3 to 4 years, and >4 to 5 years, respectively. For those less than 18 months of age, the carriage rate decreased with increasing age (P = 0.0011; Mantel-Haenszel test for trend), while for those older than 12 months of age, the carriage rate increased with increasing age (P < 0.0001).

Of the 221 MRSA isolates, 212 isolates were available for analysis. All of these 212 isolates were sensitive to vancomycin and teicoplanin. All but two of the isolates identified from hospital A were resistant to penicillin. Most isolates were resistant to erythromycin and clindamycin but sensitive to trimethoprim-sulfamethoxazole (SXT) and doxycycline. The detailed susceptibility distribution of various antibiotics for the isolates is shown in Table 2. No significant difference in antibiotic susceptibility patterns was noted among the isolates from the three different regions of Taiwan.

Table 3 illustrates the detailed distribution of PFGE patterns, SCCmec types, and the presence/absence of PVL genes among these isolates. A total of 10 PFGE patterns were identified. Patterns C and D were the two most common patterns and accounted for 62% and 28% of the isolates analyzed, respectively. The distribution of PFGE patterns among the three regions showed a trend for a difference (P = 0.09 by a log-likelihood contingency test). Four types (types II, III, IV, and V_T) of SCCmec genes were identified among the isolates, with type IV (70%) being the predominant type, followed by type V_T (26%). The distribution of SCCmec types among the three regions was significantly different (P = 0.03). Four isolates of the AF PFGE pattern were untypeable by the methods used in this study. PVL genes were present in 60 isolates (28%). Twenty-five isolates underwent MLST, and eight sequence types were identified. Sequence type 59 (ST59) was the most common sequence type and accounted for 9 of 10 PFGE type C isolates, 4 of 6 PFGE type D isolates, and the isolate of PFGE type AN. The other two isolates of PFGE type D were ST338, which is a single-locus variant of ST59 (a single nucleotide difference in the gmk locus). The remaining isolate of PFGE type C belonged to a new sequence type, which is a single-locus variant of ST59 (a single nucleotide difference in the pta locus). One isolate of PFGE type F also belonged to a new sequence type, which is also a single-locus variant of ST9 (a single nucleotide difference in the gmk locus). The detailed association of PFGE patterns with sequence types and SCCmec types and the presence of the PVL gene of these isolates are shown in Table 4. The MRSA isolates characterized by ST59/PFGE type C/SCCmec IV/absence of PVL genes and ST59/PFGE type D/SCCmec V_T/presence of PVL genes were the two most common clones and accounted for 59% and 25% of the isolates analyzed, respectively.

DISCUSSION

Results from this study indicate that the national prevalence of nasal MRSA colonization among otherwise healthy children in Taiwan was 7.3% during the period from July 2005 to October 2006 inclusively, with values ranging from 4.8% in the central region of Taiwan to 9.5% in the northern region of Taiwan. Compared with those among the healthy children during the period of 2001 to 2002 (1, 17, 23) (Table 5), though the study population was different for these studies, the nasal MRSA colonization prevalence among healthy children in Taiwan increased significantly, from 1.9% in 2001 to 9.5% (P < 0.0001 by chi-square test) during the period of 2005 to 2006 for northern Taiwan and significantly from 3.3% to 6.7% for southern Taiwan (P < 0.001 by chi-square test). This increasing trend of nasal MRSA colonization prevalence might account for the increasing incidence of CA-MRSA infection in children in Taiwan (3, 9, 31). In the United States, where CA-MRSA is also being increasingly reported, the MRSA colonization prevalence for the general population appeared to have been relatively low until the year 2002 (19, 21, 25, 27, 28). In a survey (21) involving 9,622 persons conducted between 2001 and 2002, national S. aureus and MRSA nasal colonization prevalence estimates were 32.4% and 0.8%, respectively. For healthy children, the nasal colonization rates ranged from 0.2% to 2.2% (19, 25, 27, 28), as reported in several pediatric studies; however, an increasing trend in this regard has been noted in certain areas of the United States recently (6). Creech et al. (6) reported that the nasal MRSA colonization rate among healthy children in Nashville, TN, increased significantly from 0.8% in 2001 to 9.2% in 2004, a picture not dissimilar to what we show in the present study from Taiwan.

In the United States, CA-MRSA strains have been recognized as representing a novel pathogen which was genetically different from the nosocomial MRSA strains (14, 24). They have limited antibiotic resistance (except to β-lactams), have two common PFGE patterns (USA 300 and USA 400), possess different exotoxin gene profiles (e.g., PVL), and contain SCCmec DNA (10). In contrast, the CA-MRSA clinical isolates in Taiwan were multiresistant and shared two common PFGE patterns (patterns D and C in this study) (1, 3, 4, 30). In the current study, more than 90% of the MRSA colonization isolates were multiresistant to erythromycin and clindamycin but sensitive to SXT and doxycycline. In addition, most colonization isolates shared common molecular characteristics, and more than 80% of the isolates belonged to one of two major clones, characterized by ST59/PFGE type C/SCCmec IV/absence of PVL genes or ST59/PFGE type D/SCCmec V_T/presence of PVL genes. However, among the clinical isolates, the clone characterized by ST59/PFGE type D/SCCmec V_T/presence of PVL genes was the dominant clone (1, 3, 30), while among the colonized isolates, the clone characterized by ST59/PFGE type C/SCCmec IV/absence of PVL genes was dominant. It seemed that PVL genes, reported to be a virulence factor associated with necrotizing pneumonia and abscesses (22), may be associated with the ability of a PVL-positive clone to cause infection.

There existed several limitations in the current study. First, the demographic characteristics and the risk factors associated with MRSA acquisition were not analyzed and compared between the children with and without CA-MRSA colonization, though all the children were healthy and presented for health care visits. Living with a family member who works in a hospital or clinic and demographic characteristics (e.g., age and gender) were reported to be associated with an increased risk of MRSA colonization (6, 21, 25). Second, the persistence of MRSA carriage in the subjects could not be determined and the incidence of subsequent MRSA infection in the subjects could not be measured in this cross-sectional analysis of MRSA nasal colonization prevalence.

In summary, 7.3% of healthy children in Taiwan were colonized by MRSA in the nares during the period from 2005 to 2006. MRSA carriage in the children may accelerate the spread in the community. Two major CA-MRSA clones were identified and would appear to have spread island-wide. Further studies are needed to determine the host factors of colonization and to develop strategies to disrupt transmission of CA-MRSA to susceptible hosts.