Skip Navigation

Population Dynamics of Nasal Strains of Methicillin-Resistant Staphylococcus aureus—and Their Relation to Community-Associated Disease Activity

  1. Erica S. Pan1,a,b,
  2. Binh An Diep1,4,a,
  3. Edwin D. Charlebois3,
  4. Colette Auerswald2,
  5. Heather A. Carleton1,
  6. George F. Sensabaugh4 and
  7. Françoise Perdreau-Remington1
  1. 1Departments of Medicine and Pediatrics, Division of Infectious Diseases,
  2. 2Department of Pediatrics, Division of Adolescent Medicine, and
  3. 3AIDS Research Institute, University of California, San Francisco, and
  4. 4Program in Infectious Diseases and Immunity, School of Public Health, University of California, Berkeley
  1. Reprints or correspondence: Dr. Françoise Perdreau-Remington, Dept. of Medicine, University of California, San Francisco, Division of Infectious Diseases, San Francisco General Hospital, 1001 Potrero Ave., UCSF 1372, Bldg. 100, Room 301, San Francisco, CA 94110 (fpr{at}epi-center.ucsf.edu)

Abstract

BackgroundNasal carriage of methicillin-resistant Staphylococcus aureus (MRSA) plays a key role in the epidemiology and pathogenesis of disease. The purpose of this study was to determine the characteristics and dynamics of nasal strains of MRSA, as well as their relation to community-associated disease activity

MethodsThis study is a cross-sectional survey and molecular epidemiologic analysis of nasal colonization by S. aureus in homeless and runaway youths, an underserved population at high risk for staphylococcal disease

ResultsOf the 308 study participants, 27.6% carried S. aureus and 6.2% carried MRSA. Subgroups of individuals with increased MRSA carriage rates were also at highest risk for community-associated MRSA infection; these subgroups included individuals with either HIV infection or AIDS, injection drug users, patients with abscesses, and those recently hospitalized. Multilocus sequence typing and pulsed-field gel electrophoresis identified 2 genotypes—ST59:P (USA1000) and ST8:S (USA300)—that accounted for 84.2% (16/19) of the MRSA isolates carried. The genotypes were distinct from nosocomial genotypes endemic in the hospital, although they originated from individuals with prior exposure to health care

ConclusionsComparison of MRSA strains from asymptomatic carriers versus concurrently collected community-associated clinical strains from patients treated at local health-care facilities allowed for the identification of 3 population dynamics of nasal strains of MRSA: (1) endemic clones—for example, ST8:C and ST59:P—sustained asymptomatic carriage and infection over prolonged periods; (2) an epidemic clone, ST8:S, demonstrated enhanced capacity for rapid transmission and widespread infections; and (3) an outbreak clone, ST30:Z (USA1100), was highly infectious but exhibited poor asymptomatic transmission

Hospitals worldwide are traditional reservoirs of methicillin-resistant Staphylococcus aureus (MRSA). Over the past decade, however, MRSA strains have increasingly been found in the community, infecting otherwise healthy individuals [17]. Several studies have demonstrated that community-associated MRSA (CA-MRSA) strains originating from patients with no prior exposure to hospitals are clonally distinct from MRSA strains endemic to hospitals [810]. We have provided further evidence in support of these findings and, moreover, have demonstrated that CA-MRSA strains originating from patients with prior exposure to health care are also clonally unrelated to concurrent nosocomial MRSA genotypes endemic to local hospitals and a long-term–care facility [11]. Although these findings challenge the traditional view that CA-MRSA strains are feral escapees from the hospital reservoir, the vast majority (>90%) of CA-MRSA infections in this geographic area are concentrated among patients with prior exposure to health care

To better define the epidemiology of MRSA in community settings, we conducted a cross-sectional survey of community-based nasal colonization of S. aureus in a high-risk San Francisco population during 2002 and compared the findings to those from 2 similarly designed studies conducted during 1999 and 2000. All 3 studies aimed to assess the population characteristics and dynamics of nasal carriage of S. aureus. The present study targeted homeless and runaway youths, who—because of injection drug use, HIV infection, and/or crowded living conditions with minimal access to sanitation facilities—are at particularly high risk for staphylococcal disease [1214]. Because nasal carriage of S. aureus plays a key role in the epidemiology and pathogenesis of disease [15], we also examined the relation between MRSA genotypes from community-based nasal colonization and concurrent clinical isolates of CA-MRSA identified in local health-care facilities [11]

Subjects, Materials, and Methods

Subjects Subjects were recruited via flyers and in person from the Larkin Street Youth Services (LSYS) drop-in center, shelters, and clinics, between 1 April and 30 September 2002. The LSYS serves ∼3000 12–24-year-old homeless youths annually, providing multidisciplinary services to assist youths in getting off the streets and into housing, as well as on-site professional and peer-administered services such as HIV counseling and testing, substance-abuse referrals, and counseling to prevent other high-risk behaviors. The Larkin Street clinic sees ∼400 patients annually; an additional 40–70 patients are followed at a satellite HIV specialty clinic. Consenting subjects (n=308) were administered a questionnaire considering demographic data, housing status, prior health-care use, prior antibiotic use, prior incarceration, and injection drug use

Subjects were asked to identify their race or ethnicity as “White/Caucasian,” “Black/African-American,” “Latino,” “Asian/Pacific Islander,” “Native American,” “Mixed,” or “Other”; mixed race was self-defined by the subject as including >1 racial category. Subjects were excluded if they reported that they currently used antibiotics. A homeless person was defined as one who lacked access to stable housing; this definition included individuals who, during the preceding 6 months, had slept ⩾30 days either on the streets, in a vehicle, or in a homeless shelter

Nasal culture From both sides of each subject’s anterior nares, a nasal culture obtained by use of a sterile Dacron swab (Becton Dickinson Microbiology Laboratories) was inoculated directly onto a blood agar plate and was streaked for isolation of single colonies

Susceptibility testing S. aureus isolates were tested for phenotypic resistance to oxacillin, by the salt agar method [16]. Susceptibilities to ampicillin, ciprofloxacin, erythromycin, tetracycline, clindamycin, trimethoprim-sulfamethoxazole, gentamicin, vancomycin, and linezolid were tested by use of the disk-diffusion method and were interpreted in accordance with NCCLS guidelines [17]. Inducible clindamycin resistance was determined as described elsewhere [18]

Molecular typing Pulsed-field gel electrophoresis (PFGE) using SmaI [19] and multilocus restriction-fragment typing (MLRFT) [20] was used to characterized all S. aureus isolates. A PFGE group consisted of isolates with a ⩽6-band difference from at least 1 other isolate in the group [21]. MLRFT, which assesses restriction-site variation in 7 housekeeping-gene loci, provided a low-cost method for grouping distantly related PFGE subtypes [20]. To provide a uniform nomenclature for describing MRSA, multilocus sequence typing (MLST) was performed on representative isolates from each concordant PFGE-MLRFT–defined genotype; sequence types were assigned on the basis of reference to the MLST database (available at http://www.mlst.net) [22]. SCCmec types I–IV were identified by use of a multiplex PCR-based protocol described elsewhere [23]

Statistical analysis The prevalence and the 95% confidence interval (CI) were calculated by means of standard equations [24]. Fisher’s exact test was used to test for significant associations between categorical variables [24]. Odds ratios (ORs) and significance for S. aureus and MRSA nasal carriage were calculated by use of STATA (version 8; StataCorp)

Results

Demographic data on subjects, and prevalence of and risk factors for S. aureus and MRSA nasal colonization The 308 subjects in the present study were predominantly male (180 [58.4%]), were of either white (115 [37.3%]) or mixed (79 [25.6%]) race (table 1), and had a median age of 20 years. During the preceding 6 months, 123 (39.9%) of the 308 subjects had spent ⩾30 nights living either on the street, in a vehicle, or in a homeless shelter; 85 (27.6%) were colonized with S. aureus and 19 of these, or 6.2% of the 308 subjects, were resistant to methicillin, a proportion higher than the 2.8% colonization rate reported for homeless adults in a population-based survey conducted 2 years before the present study [19] (table 2)

Figure 1

Longitudinal relationship between nasal carriage and community-associated disease due to methicillin-resistant Staphylococcus aureus (MRSA), in a community setting. The pie charts illustrate the distribution of MRSA genotypes (designated on the basis of multilocus sequence typing combined with pulsed-field gel electrophoresis (MLST:PFGE) in the 3 community-based studies of S. aureus nasal colonization that were conducted in San Francisco and that surveyed different high-risk populations during the time periods indicated [19, 26]. Only 2 of 84 MRSA nasal isolates carried SCCmec type II (ST36:A); the remainder carried SCCmec type IV. The line graph demonstrates dynamic changes of major MRSA genotypes (MLST:PFGE:SCCmec) infecting patients treated at the Community Health Network of San Francisco [11]. The estimated annual numbers of isolates associated with each MRSA genotype have been weighted to account for the nonproportionally stratified random-sampling design, which is described in detail in reference [11]. The PFGE types identified in the present study (for PFGE banding patterns, see reference [25]) correspond to some of the PFGE types described by the Centers for Disease Control and Prevention [32, 46]: ST36:A (SCCmec type II) is similar to pulsed-field type USA200, ST72:B (SCCmec type IV) to USA700, ST5:D (SCCmec type II) to USA100, ST5:D (SCCmec type IV) to USA800, ST1:K (SCCmec type IV) to USA400, ST59:P (SCCmec type IV) to USA1000, ST8:S (SCCmec type IV) to USA300, ST45:X (SCCmec type IV) to USA600, and ST30:Z (SCCmec type IV) to USA1100

Table 1

Subject demographics and risk factors for nasal colonization by methicillin-resistant Staphylococcus aureus (MRSA), in homeless and runaway youths in San Francisco

Table 2

Prevalence of nasal carriage of methicillin-resistant Staphylococcus aureus (MRSA), in 3 community-based settings in San Francisco

Of the 48 subjects who had a history of hospitalization during the preceding 12 months, 9 (18.8%) had nasal carriage of MRSA, compared with only 10 (3.8%) of those who did not have such a history (P=.001) (table 1). Nasal carriage of MRSA was not significantly associated with a history of either outpatient or emergency-department visits occurring during the preceding 12 months

Other univariate risk factors for MRSA carriage included testing positive for HIV (OR, 7.6), a history of injection drug use (OR, 6.7), and abscess during the preceding 12 months (OR, 8.6) (table 1). Neither a history of incarceration nor homelessness was significantly associated with nasal carriage of MRSA (table 1). Of the 19 MRSA isolates, only 2 originated from subjects without 1 of the correlated risk factors; however, both of these subjects had been released from jail during the preceding 12 months, and 1 reported having sought emergency-department care 5 times during the preceding 12 months

Genotype distributions and antimicrobial susceptibilities PFGE and MLRFT differentiated the 19 MRSA isolates into 5 genotypes and the 66 methicillin-susceptible S. aureus (MSSA) isolates into 11 genotypes; 3 genotypes were represented by both MRSA isolates and MSSA isolates. Unique MLRFT-PFGE–defined genotypes were further characterized by MLST, to provide a standardized nomenclature [22]. For brevity, a genotype is identified by its MLST and PFGE types (e.g., ST8:S). Two genotypes, ST8:S and ST59:P, predominated in nasal carriage, accounting for 16 (84.2%) of the 19 MRSA isolates (figure 1)

All MRSA isolates carried the SCCmec type IV element, and 5 of the 8 ST8:S MRSA isolates were resistant to both erythromycin and ciprofloxacin (table 3); 2 other ST8:S isolates also were resistant to clindamycin. The singleton MRSA isolate ST8:C, a strain genetically related to ST8:S [25], was also resistant to erythromycin and ciprofloxacin. In contrast, 7 of the 8 ST59:P MRSA isolates were resistant only to erythromycin. None of the MRSA isolates expressed inducible resistance to clindamycin, according to the results of the D-test

Table 3

Antibiotic susceptibilities and molecular characterization of clinical methicillin-resistant Staphylococcus aureus (MRSA), in homeless and runaway youths

Risk factors and MRSA genotypes of nasal-carriage isolates There was no correlation between risk factors and specific MRSA genotypes. The distribution of genotypes was similar among colonized subjects who did or did not have a history of hospitalization during the preceding 12 months: of the 9 MRSA isolates from subjects with such a history, 4 belonged to ST8:S and 5 belonged to ST59:P; of the remaining 10 MRSA isolates (from subjects who did not have a history of hospitalization during the preceding 12 months), 4 belonged to ST8:S, 3 belonged ST59:P, and 1 each belonged to ST5:D, ST8:C, and ST121:N

Nasal MRSA populations and their relation to CA-MRSA disease activity To assess the relation of the nasal carriage of MRSA to community-associated disease activity over time, we compared the genotypes of carriage isolates from the 3 colonization surveys (table 2) versus concurrent clinical isolates identified between 1998 and 2002 at the Community Health Network of San Francisco (CHN) [11]. The CHN serves the most vulnerable, diverse populations of San Francisco by providing a wide array of health-care services at its 23 affiliated health centers, including San Francisco General Hospital, Jail Health Services, a long-term–care facility, and citywide outpatient clinics

The 3 cross-sectional surveys identified 84 nasal-carriage MRSA isolates (table 2). Of these 84 isolates, only 2 had genotypes that matched those of nosocomial strains endemic to the local hospital; these 2 isolates belonged to ST36:A (SCCmec type II), a clone responsible for an outbreak in the intensive-care unit of San Francisco General Hospital during 1998 (F.P.-R., unpublished observation)

The 3 major nasal-carriage MRSA genotypes—ST59:P, ST8:C, and ST8:S—accounted for 75.3% of the isolates from the 3 cross-sectional studies of colonization by MRSA; these 3 genotypes were also prominent among CA-MRSA disease isolates, accounting for 47.9% (553/1154) of cases of CA-MRSA infection treated at the CHN between 1998 and 2002 (figure 1) [11]. A fourth clone, ST30:Z, was not identified among the nasal-carriage isolates, although it accounted for 28.6% (330/1154) of CA-MRSA infections treated at the CHN between 1998 and 2002

Discussion

The cross-sectional survey described here is the most recent of 3 studies designed to examine risk factors for both the nasal carriage and the population dynamics of MRSA strains from different community-based settings in San Francisco [19, 26]. The prevalence of nasal carriage of MRSA varied between the different high-risk groups—from 2.8% in homeless and marginally housed adults to >6% in injection drug users and in homeless and runaway youths (table 2). These carriage rates are significantly higher than those reported elsewhere [27, 28], a result that may be related to the ongoing CA-MRSA disease epidemics that began in the San Francisco Bay Area during 1998 [11]. Four risk factors were consistently associated with carriage of MRSA: a history of hospitalization during the preceding 12 months, the presence of either HIV infection or AIDS, a history of injection drug use, and a history of abscesses during the preceding 12 months

By comparing MRSA strains from carriers versus concurrent community-associated clinical isolates from patients treated at the CHN, we identified 3 population dynamics of nasal strains and their relation to CA-MRSA disease: (1) endemic clones sustained asymptomatic carriage and infections over prolonged periods; (2) an epidemic clone demonstrated enhanced capacity for rapid transmission and widespread infections; and (3) an outbreak clone was highly infectious but exhibited poor asymptomatic transmission. The differences in carriage patterns are clone specific and are discussed below

The first carriage pattern is represented by 2 endemic clones, ST8:C and ST59:P. These 2 MRSA clones had been continuously present in San Francisco since at least 1996, when routine genotyping of strains became part of MRSA-disease surveillance [11]. Although the ST59:P clone had been found infrequently elsewhere [29, 30], it comprised a significant proportion of isolates in the 3 studies of community-based colonization; among clinical isolates, the ST59:P clone accounted for ∼20% of the CA-MRSA strains (figure 1). Recently, Monk et al. [30] reported that 6 (12.5%) of 48 MSSA isolates from injection drug users’ lesions treated at a hospital in the United Kingdom were ST59:P, and they suggested that, whereas this clone is rarely found in carriage and disease in other populations, it is adept at causing disease in this high-risk group. In San Francisco, ST59:P is common in carriage as well as in disease and is not associated with injection drug use

The other endemic clone, ST8:C, belongs to clonal complex 8, one of the most successful pandemic MRSA lineages [6, 11, 3134]. Although this clone was highly prevalent (frequency, >40%) in the 2 earlier studies of nasal carriage, which considered injection drug users and homeless adults, it represented only 5% of all MRSA carriage isolates from the younger population of homeless and runaway youths considered in the present study (figure 1). Although its community-based carriage rate declined in the 2002 survey, ST8:C still accounted for 15.8% of all cases of CA-MRSA disease treated at the CHN during that year. We note, however, that, because of the small number of MRSA clones recovered, the calculation of the percentage of different MRSA clones in the surveys may be susceptible to sampling variability

The second carriage pattern is represented by the epidemic clone, ST8:S. This CA-MRSA clone has demonstrated enhanced epidemic expansion in at least 13 US states over a relatively short time span. We identified the first ST8:S MRSA clinical isolate in September 2000, in an outpatient with a skin abscess [31], but did not encounter ST8:S during the colonization surveys conducted during 1999 and 1999–2000 (figure 1). In 2002, carriage of ST8:S in homeless and runaway youths accounted for 42% of all cases of MRSA in this population, a finding clearly demonstrating this clone’s capacity for rapid dissemination in the community (figure 1). ST8:S has also been associated with 37 clinical cases of MRSA in 2001 and with 145 cases of MRSA in 2002 (figure 1). We also found extensive spreading of ST8:S in medical facilities in San Mateo [33] and Alameda [34], 2 counties <30 miles from San Francisco, during 2002 and 2003. These data demonstrate the recent emergence of ST8:S in the San Francisco Bay Area

ST8:S is responsible for extensive disease outbreaks in other parts of California, and large-scale outbreaks among jail inmates, neonates in a nursery, men who have sex with men, and university athletes [6] have been associated with the ST8:S clone [31, 35]. On 22 April 2003, concern about the potential epidemic spread of ST8:S in the community led the Los Angeles County Department of Health Services to mandate the reporting of all CA-MRSA disease in hospitalized children <18 years of age [35, 36]. In addition, between 1 February 2003 and 15 April 2004, 14 community-based patients presented at the Harbor-UCLA Medical Center with necrotizing fasciitis, an infection rarely caused by S. aureus and ST8:S was the only pathogen cultured from the patients’ lesions [37]. Multiple studies have also reported the widespread distribution of ST8:S (also classified as USA300) in 13 US states, with notable outbreaks documented among incarcerated individuals [4, 6, 31, 32], athletes [6, 38, 39], military personnel [40, 41], and children [35, 42, 43]

The third carriage pattern is represented by the outbreak clone, ST30:Z. Although ST30:Z has been responsible for major outbreaks of CA-MRSA [3], in the 1999–2000 nasal survey it was represented by only a single isolate. However, the only nasal-carriage ST30:Z isolate has a PFGE subtype, Ze, that is different from the 2 predominant disease-associated subtypes, Za and Zb (F.P.-R., unpublished data). ST30:Za and ST30:Zb accounted for 41% of the cases of CA-MRSA disease occurring between 1998 and 2002 (figure 1). The poor transmissibility of the ST30:Z clonal lineage in community-based settings may explain the sharp decrease in its prevalence in CA-MRSA disease since 2001. The rapid clearance of ST30:Z was also noted in a pediatric hospital in Mexico City [44], a finding attesting to its limited transmission after the outbreak period

ST8:S and ST30:Z, 2 of the major CA-MRSA genotypes that previously had been implicated in the majority of community-associated skin and soft-tissue infections, harbored the genes for Panton-Valentine leukocidin (PVL) [25], which causes leukocytolysis and tissue necrosis. The strong association between PVL and CA-MRSA clones has been observed repeatedly in different geographic areas [5, 8, 9, 25, 45]. On the basis of these observations, it has been suggested that PVL confers a selective advantage on CA-MRSA clones [8, 25]. Whether PVL contributes to the success of CA-MRSA clones is unclear, in light of the 2 distinct population dynamics observed for ST8:S and ST30:Z

The recent National Health and Nutrition Examination Survey of 9662 individuals reported that 0.8% of the general population carried MRSA, and only 0.4% of the general population carried MRSA with SCCmec type IV, which is a marker of CA-MRSA strains [28]. The present study identified subgroups of individuals in the community who had increased MRSA carriage rates; not coincidentally, the same subgroups of individuals are at highest risk for CA-MRSA infection. These individuals constitute the expanding community-based reservoir of MRSA; they include predominantly those with either HIV infection or AIDS, injection drug users, patients with abscesses, and those with a history of recent hospitalization. Further research on the population dynamics of nasal carriage of MRSA may identify novel interventions to prevent acquisition and transmission of these community pathogens in the general population

Footnotes

  • Presented in part: 41st annual meeting of the Infectious Disease Society of America, San Diego, October 2003 (poster 255); International Staphylococci and Staphylococcal Infections meeting, Charleston, SC, October 2004 (oral presentation)

    Potential conflicts of interest: F.P.-R. is a consultant for Pfizer Inc

    Financial support: Pfizer Corporation (unrestricted grant); University of California, San Francisco, Traineeship in AIDS Prevention Studies (grant T32 MH-19105 to E.S.P.); University of California, Berkeley, Eugene Cota-Robles Predoctoral Fellowship and a Genetics Training Grant (to B.A.D.)

  • E.S.P. and B.A.D. contributed equally to this research

  • Present affiliation: San Francisco Department of Public Health, San Francisco, California

  • Received January 19, 2005.
  • Accepted March 25, 2005.

References

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
  41. 41.
  42. 42.
  43. 43.
  44. 44.
  45. 45.
  46. 46.
| Table of Contents