Rapid PCR/ESI-MS-based molecular genotyping of Staphylococcus aureusfrom nasal swabs of emergency department patients

Staphylococcus aureus is a major cause of healthcare-associated infections [1]. Although several epidemiological studies have provided information on the prevalence of methicillin-resistant Staphylococcus aureus (MRSA) colonization in different hospital settings [2‐4] and in emergency department personnel [5, 6], limited data exist on MRSA colonization rates for patients seen in emergency departments. The cost benefit of screening for MRSA colonization is not overwhelming and the subject is controversial [7]. A limitation of both culture-based and molecular methods of MRSA screening is that current tests determine only the presence or absence of colonization with no information on the colonizing strain type [8]. As only small subsets of MRSA clones have been responsible for the major epidemics in the United States [9] screening might have greater value if future studies demonstrate that certain clones are more likely to cause serious infections than others.

We have developed a rapid molecular surveillance assay based on PCR coupled to electrospray ionization mass spectrometry (PCR/ESI-MS) [10‐12] that identifies and provides detailed information about S. aureus strains. A previous version of this assay has been shown to accurately genotype S. aureus strains and to identify the clonal complex of a series of isolates from the Centers for Disease Control (CDC). Excellent agreement with multilocus sequence typing (MLST) and pulsed-field gel electrophoresis (PFGE) results [13] was reached using in the PCR/ESI-MS assay a set of eight primer pairs that target the same genes as those for MLST. This set of primer pairs was designed to optimize the information content of the amplicons and distinguished 99% of 710 distinct S. aureus sequence types [13]. Using a second set of eight primer pairs, PCR/ESI-MS has also been used to characterize virulence factors, toxin-encoding genes and antibiotic resistance determinants on isolated colonies [14] and to detect MRSA in nasal swabs with sensitivity and specificity similar to culture methods [15]. In this work, elements from the previous studies were consolidated into a single PCR/ESI-MS research application to directly screen nasal swabs from sample of 200 patients who presented to an urban hospital emergency department.

While some Emergency Departments have a high prevalence of MRSA isolated from patients with sporadic skin and soft-tissue infections [24‐26], limited research has explored nasal MRSA carriage in this population. The mechanisms leading to S. aureus nasal carriage are multifactorial, and an optimal fit between host and bacteria seems to be essential [27]. This study demonstrated a significant prevalence of S. aureus, MRSA and drug-resistant CoNS nasal colonization in a random sample of individuals seeking care from the Emergency Department at The Johns Hopkins Hospital. The prevalence of MRSA (13.5%) among nasal isolates was somewhat greater than those estimated from other studies that had a similar objective and sample population [28]. Three studies performed outside healthcare facilities (i.e., isolated, semi-closed rural population [29], urban poor in San Francisco [30], or university students in Portugal [31]) that include a total of 4,452 people, showed a pooled MRSA colonization prevalence of only 0.76% [32]. The prevalence of MRSA colonization in an inpatient setting in a tertiary hospital was reported to be 0.18%–7.2% [32, 33]. Our findings were similar to the prevalence noted in a study investigating the prevalence of MRSA among newly arrested men in Baltimore (29.5% and 13.5% in our study vs. 40.4% and 15.8% for S. aureus and MRSA prevalence, respectively) [34]. Other studies have identified three risk factors for MRSA colonization: hospitalization within the past 12 months [16], previous MRSA infection [17] and injection drug use [18]. Because these risk factors had already been identified, they were exclusion criteria in our study. Although we evaluated patient demographics, primary complaints, chronic co-morbidities, antibiotic prescriptions and providers’ admission decisions, none were independent risk factors for MRSA colonization.

Our data confirms recent epidemiologic data indicating that nosocomial MRSA strains are present in the community. Interestingly, there was substantial genotypic diversity of S. aureus and MRSA types in this the patient population. There was a significant frequency of co-colonization of S. aureus and CoNS (24 patients, 12%) and at least one third of patients were colonized with a CoNS that had a high-level mupirocin resistance gene (65 patients or 32%, non- counting the S. aureus co-infections). This is concerning because it has previously been reported that the gene for high-level mupirocin resistance can be found on conjugative plasmids that carry multiple resistance determinants for other antimicrobial agents [21]. Moreover, this resistance has been shown to be transferable to S. aureus [35], which has implications for mupirocin decolonization strategies.

The PCR/ESI-MS methodology operates on an automated, high-throughput platform; patient specimens are screened without culture and results are obtained in a time frame similar to existing molecular tests. As data from more extensive analysis of the S. aureus strains that result in the hospital infections become available, additional targets for genetic analysis of S. aureus can be added. While a goal of this assay was to add targets that provide information that have value for infection control and tracking outbreaks in hospital settings, the limitation to a current assay is the lack of resolution compared to other technologies. We acknowledge that the inclusion of typing for TSST or PVL, would probably not aid in a clinical scenario, but might aid a research study where predictors of subsequent infection in hospitalized patients were being studied.

This study was performed in one location, the Johns Hopkins Hospital Emergency Department and, therefore, these results may not be applicable to other areas of the United States. Although the ethnic diversity of the study participants was similar to that of the population of Johns Hopkins Hospital patients, of which 80% are African American, these results may not be applicable to more ethnically diverse areas of the United States. It also should be noted that literature cites significant heterogeneity in the risk factors for nasal S. aureus and MRSA colonization, as well as the fact that only surveying nasal swabs likely underestimates the prevalence of S. aureus and MRSA carriage [4, 36]. Several other, previously described risk factors for S. aureus and MRSA colonization (i.e. previous admission to hospital, or previous antibiotic use) were also investigated in this study. However, there are other known factors for which we do not have data, that might have significantly contributed to the relatively high burden of MRSA encountered in our study, including HIV status, diabetes, men who have sex with men, recent incarceration, high-risk sexual behavior, household members with SSTI, and hospital-associated exposures [4]. It should be noted that this was not a longitudinal study that would distinguish at least three S aureus nasal carriage patterns in healthy individuals: persistent carriage, intermittent carriage, and non-carriage. This distinction is important because persistent carriers have higher S. aureus loads and a higher risk of acquiring S. aureus infection [37, 38]. A further important limitation is that in the significant proportion of patients (~40%) with both S. aureus and CNS, this assay alone cannot distinguish for the presence of MRSA or other resistance determinants being carried by S. aureus. Thus, there remains the need for plating samples on ChromAgar, which adds significantly to the turn-around time for infection control purposes.

In conclusion, the prevalence of S. aureus and MRSA colonization in a convenience sample of patients seeking care in Emergency Department was 29.5% and 13.5%, respectively. A substantial fraction of the S. aureus-colonized patients were co-colonized with CoNS and high-level mupirocin-resistant CoNS. These findings highlight the importance of active surveillance studies to detect MRSA colonization and/or infection among emergency department patients. The molecular genotypes were directly identified from nasal swabs by PCR/ESI-MS. The ability to identify unique molecular signatures from MRSA may allow infection control resources to be effectively focused and will enable tracking origins of outbreaks in hospital settings. Determining the molecular genotype of S. aureus during intake screening may prove valuable if certain molecular genotypes become associated with increased infection risk.