Methicillin-resistance in Staphylococcus species is encoded via the mecA gene, which results in production of penicillin-binding protein (PBP)2A, a penicillin binding protein with reduced affinity for β-lactams 1. mec is part of a larger genomic element termed the Staphylococcal chromosomal cassette (SCCmec), which contains genes mediating antibiotic resistance. Up to eight types of SCCmec have now been reported in the literature 2 and the differences between these SCCmec types account for the primary differences between various MRSA clones. For example, SCCmec I, II, and III are larger and more difficult to mobilize and are most frequently present in hospital acquired (HA-MRSA) clones (USA 100 and 200). SCCmec IV is a smaller, easier to mobilize genetic element that is frequently present in community-associated MRSA (CA-MRSA; clones USA 300 and 400) 3. It has been observed that CA-MRSA is effectively integrating into the health care environment and it is therefore increasingly less reliable to make this differentiation on the basis of acquisition location 4567. HA-MRSA and CA-MRSA clones are noted to display different resistance patterns as a result of their unique genetic elements. Compared with HA-MRSA, CA-MRSA isolates are more likely to be susceptible to non-β-lactam antibiotics, including tri-methoprim-sulfamethoxazole (TMP-SMX), clindamycin, fluoroquinolones, gentamicin, erythromycin, and tetracyclines with geographic variability 789.

Increasing attention is being paid to the issue of reduced susceptibility and resistance of MRSA to vancomycin. Although vancomycin has long been considered a reliable agent for treatment of MRSA infections, isolates with intermediate (VISA) and full (VRSA) levels of resistance have been reported. The Clinical and Laboratory Standards Institute vancomycin minimum inhibitory concentration (MIC) breakpoints for MRSA were last updated in 2006 and resulted in a lowering of the breakpoints as follows: susceptible, ≤ 2 μg/ml; intermediate, 4 to 8 μg/ml; resistant, ≥ 16 μg/ml.

Vancomycin exerts its antibiotic activity by binding to the D-alanyl-D-alanine portion of cell wall precursors, which subsequently inhibits peptidoglycan polymerization and transpeptidation. High-level resistance is mediated via the vanA gene, which results in production of cell wall precursors (D-Ala-D-lac or D-Ala-D-Ser) with reduced affinity for vancomycin 10. Intermediate level resistance (VISA) is believed to be preceded by the development of heteroresistant vancomycin intermediate S. aureus (hVISA) 11. Heteroresistance is the presence of resistant subpopulations within a population of bacteria determined to be susceptible to the antibiotic tested. It is thought that exposure of such a heteroresistant MRSA population to low concentrations of vancomycin may kill the fully susceptible subpopulations and select for the resistant subpopulations. The mechanisms of heteroresistance are not fully elucidated, but are hypothesized to be due to a thickened cell wall and increased production of false binding sites 11. The accessory gene regulator (agr; discussed in detail below) type and functionality may also play a role in the development of this type of resistance 12.

Reduced susceptibility to glycopeptides may also impact the susceptibility of MRSA to daptomycin. Several reports have found hVISA and VISA isolates to display resistance to daptomycin 131415. Daptomycin is a cyclic lipopetide that works by binding to the cell membrane to subsequently cause destabilization resulting in bactericidal activity. It is hypothesized that the thickened cell wall noted to occur in MRSA isolates with intermediate-level vancomycin resistance may result in sequestration of daptomycin. Additionally, reduced susceptibility has been documented to develop while on prolonged daptomycin therapy 1617.

Linezolid is a synthetic oxazolidinone that inhibits the initiation of protein synthesis by binding to the 23s ribosomal RNA and thereby preventing formation of the 70s initiation complex. Although linezolid has generally remained a reliable antibiotic for MRSA infections, several occurrences of resistance have been observed 1819. The first report of resistance 18 from a clinical isolate was reported in 2001, about 15 months after the drug was introduced to the market. Upon analysis, the organism was found to have mutations in the DNA encoding a portion of the 23s ribosomal RNA (rRNA). Linezolid resistance has been identified more commonly among Staphylococcus epidermidis and Enterococcus species, but the possibility of linezolid resistance among MRSA should be kept in mind.

In vitro studies have reported tigecycline to be highly active against MRSA isolates that have been tested. No reports of resistance to clinical isolates have been reported to our knowledge, but the use of this agent for serious MRSA infections has been very limited. Quinupristin/dalfopristin has similarly been shown to be highly active in vitro against MRSA, but clinical isolates with resistance have been reported 20 and the use of this agent for serious MRSA infections has also been limited.

Virulence factors play an important role in determining the pathogenesis of MRSA infections. Colonization by MRSA is enhanced by biofilm formation, antiphagocytocic microcapsules, and surface adhesions 21. Once an inoculum is established, S. aureus can produce a variety of virulence factors to mediate disease, including exoenzymes and toxins. Exoenzymes include proteases, lipases and hyaluronidases, which can cause tissue destruction and may facilitate spread of infection. The toxins that can be produced are numerous and include hemolysins, leukocidins, exfoliative toxins, Panton-Valentine leukocidin (PVL) toxin, toxic shock syndrome toxin (TSST-1), enterotoxins, and α-toxin 21. S. aureus also has a multitude of mechanisms to further elude and modulate the host immune response. Specific examples include inhibition of neutrophil chemotaxis via a secreted protein called chemotaxis inhibitory protein of staphylococci (CHIPS), resistance to phagocytosis via surface proteins (for example, protein A and clumping factor A (ClfA)), inactivation of complement via Staphylococcus complement inhibitor (SCIN), and production of proteins that confer resistance to lysozyme (for example, O-acetyltransferase) and antimicrobial peptides (for example, modified Dlt proteins and MprF protein) 22.

Various toxins have been associated with different clinical scenarios and clinical presentations 21. For example, α-toxin, enterotoxin, and TSST-1 are believed to lead to extensive cytokine production and a resulting systemic inflammatory response. Epidermolytic toxins A and B cause the manifestations of Staphylococcal scalded skin syndrome. PVL is most frequently associated with CA-MRSA and may play an important role in cavitary pneumonia and necrotizing skin and soft tissue infections, as discussed in the following section.

Expression of virulence factors is largely controlled by the agr 23. Polymorphisms in agr account for the now five different types that have been identified. HA-MRSA isolates are most frequently agr group II, whereas CA-MRSA isolates are most frequently agr groups I and III. Another difference is that agr is functional in a majority of CA-MRSA isolates whereas agr may be dysfunctional in about half of HA-MRSA isolates 24. When agr is active it generally results in upregulation of secreted factors and downregulation of cell surface virulence factors. This pattern of expression has been noted to occur during the stationary growth phase when studied in vitro and in animal models. During an exponential growth phase, upregulation of cell surface factors is increased and production of secreted factors is decreased. A recent study 25 sought to examine virulence gene expression in humans by measuring transcript levels of virulence genes in samples taken directly from children with active CA-MRSA skin and soft tissue infections (superficial and invasive abscesses). This analysis showed that genes encoding secretory toxins, including PVL, were highly expressed during both superficial and invasive CA-MRSA infections whereas surface associated protein A (encoded by spa) was only associated with invasive disease. It was also demonstrated that the virulence gene expression profiles measured from in vivo samples differed from those observed when the clinical isolates were exposed to purified neutrophils in vitro. This study therefore found some differences between in vitro and animal models when compared to this in vivo assessment and supports the hypothesis that the course of an MRSA infection can be altered in recognition of host-specific signals.

The era of MRSA being exclusively a nosocomial pathogen is quickly fading. An epidemiologic study conducted in metropolitan areas throughout the United States found only 27% of MRSA sterile-site infections are of nosocomial origin 26. Taking a closer look, of the 63% of patients presenting from the 'community', the majority had recent healthcare exposures, including hospitalization in the previous 12 months, residence in a nursing care facility, chronic dialysis, and presence of an invasive device at the time of admission. This group of patients deemed to have 'healthcare-associated, community-onset' infection most often harbor strains of MRSA associated with the hospital setting; however, crossover of the CA-MRSA clone into these patients is occurring in many healthcare centers 4567.

Numerous studies have evaluated the impact methicillin resistance has on the outcome of patients infected with S. aureus. A meta-analysis of 31 S. aureus bacteremia studies found a significant increase in mortality associated with MRSA bacteremia compared to methicillin-susceptible S. aureus (MSSA) bacteremia (pooled odds ratio 1.93, 95% confidence interval 1.54 to 2.42; P < 0.001). This finding remained evident when the analysis was limited to studies that were adjusted for potential confounding factors, most notably severity of illness 27. Since this publication, several other investigations comparing MRSA and MSSA bacteremia have yielded similar results 28. The higher attributable mortality associated with MRSA could be explained, in part, by significant delays in the administration of an antibiotic with anti-MRSA activity, particularly in patients presenting from the community. A single-center cohort study found only 22% of MRSA sterile-site infections cultured within the first 48 hours of hospital admission received an anti-MRSA antibiotic within the first 24 hours of culture collection, a factor that was independently associated with hospital mortality 29, and a significant contributor to hospital length of stay and costs 30.

In the majority of hospitals throughout the world, the antibiotic of choice for empiric therapy of suspected MRSA infection is vancomycin. However, just as the era of MRSA occurring only in the hospital setting has ended, so too might the automatic, empiric use of vancomycin in these situations. Increasingly it is being reported that MRSA infections with vancomycin MICs in the higher end of the 'susceptible' range (1.5 to 2 mcg/ml) may be associated with higher rates of treatment failure compared to isolates with a MIC of 1 mcg/ml or less 31. Additionally, a cohort analysis of MRSA bacteremia found vancomycin therapy in isolates with an MIC of 2 mcg/ml was associated with a 6.39-fold increase in the odds of hospital mortality 32.

As the predominant genetic background of MRSA is transitioning from that of the hospital to community architecture (for example, clones USA 100 to USA 300 in hospitalized patients, so too might the severity of infection. Because of its epidemiologic association with CA-MRSA and severe, necrotizing pneumonia, PVL has gained much attention as an important virulence factor. However, the extent of its role in pathogenesis is a matter of significant debate and it is likely that other factors, including expression of adhesion proteins such as staphylococcal protein A, as well as α-toxin and phenol-soluble modulins, are also responsible for increased infection severity 3334. Regardless, the selection of antibiotics in the treatment of MRSA pneumonia characterized by hemoptysis, leukopenia, high fever, and a cavitary picture on chest radiograph 35 as well as other necrotizing infections may be of clinical significance. Secretory toxin production is likely enhanced by beta-lactams such as nafcillin or oxacillin, maintained by vancomycin, and inhibited, even at sub-inhibitory concentrations, by protein-synthesis inhibitors, including clindamycin, rifampin, and linezolid 3637. As such, it may be reasonable to combine these toxin-suppressing agents with beta-lactams or vancomycin in severe MRSA infections.

Timely provision of appropriate antimicrobial coverage in an initial anti-infective treatment regimen results in optimal outcomes for bacterial and fungal infections 293839. This is also true for MRSA infections where it has been shown that antimicrobial regimens not targeting MRSA when it is the cause of serious infection (for example, pneumonia, bacteremia) results in greater mortality and longer lengths of hospitalization 2930. The following represents the antimicrobial agents currently available for serious MRSA infections and those in development (Table 1).

MRSA will continue to be an important infection in the ICU setting for the foreseeable future. Clinicians should be aware of the changing virulence patterns and antimicrobial susceptibility patterns of MRSA in their local areas. This information should be used to develop prevention and treatment strategies aimed at minimizing patient morbidity and healthcare costs related to MRSA infections.

agr: accessory gene regulator; CA-MRSA: community-associated MRSA; CRBSI: catheter-related blood stream infection; HA-MRSA: hospital-acquired MRSA; hVISA: heteroresistant vancomycin intermediate S. aureus; MIC: minimum inhibitory concentration; MRSA: methicillin-resistant Staphylococcus aureus; MSSA: methicillin-susceptible S. aureus; PBP: penicillin-binding protein; PVL: Panton-Valentine leukocidin; SCC: Staphylococcal chromosomal cassette; TSST: toxic shock syndrome toxin; VISA: vancomycin intermediate S. aureus; VRSA: vancomycin-resistant S. aureus.

MHK is on the speakers bureau for the following companies: Pfizer, Bard, Merck, Astrazeneca. MHK is a consultant for the following companies: Pfizer, Bard, Astellas, Orthno-McNeil. LS and SM have no competing interests to report.

This article is part of a review series on Infection, edited by Steven Opal. Other articles in the series can be found online at http://ccforum.com/articles/theme-series.asp?series=CC_Infection