In Central Greece over the 7-year study period, MRSA strains have widely spread in the community and since 2005 account for >50% of all S. aureus isolates recovered from children with community-associated infections. These MRSA isolates are often MDR, exhibiting resistance to fusidic acid, tetracycline, erythromycin, and clindamycin. It appears to be a sustained phenomenon rather than a transient situation. In Greece, the excessive use of antibiotics may have contributed to this high frequency of MRSA isolates.
In our area, 85% of the tested MRSA isolates belonged to the European ST80. This finding suggests that the increase in MRSA infections is associated with the predominance and spread of a single clone in the community, which is the European ST80 [18, 19]. Recent reports document the identification of the European ST80 clone also in Algeria , Tunisia , and Southern Israel .
It is not known why ST80 is so capable of rapidly spreading in a community or why it causes infections seemingly more readily than other S. aureus clones. The molecular analysis revealed that 93.6% of the MRSA isolates from Central Greece were PVL-positive. The suggested role of PVL in enhancing the organism's ability to spread rapidly or to increase severity of invasive infections is under investigation [7, 22]. In addition, other factors are currently also investigated. A recent study described a class of secreted staphylococcal peptides that have a remarkable ability to recruit, activate and subsequently lyse human neutrophils, thus eliminating the main cellular defense against S. aureus infection .
As a consequence of the wide spread of such isolates in the community, the number of cases of SSTIs and invasive infections due to MRSA has increased across all ages. Of the 309 infections, 11.3% occurred during the first month of life. This is in accordance with the literature from areas with high rate of CA-MRSA [24, 25].
Among the studied children with SSTIs, the proportion of cases admitted to the hospital was significantly higher in the group of patients with an MRSA infection than in those infected by an MSSA isolate. The difference in hospitalization frequencies is consistent with the CA-MRSA infections being more severe than the MSSA infections. The identification and antibiotic susceptibilities of the isolates were not known at the time of admission.
Fusidic acid is considered to be a very active agent for S. aureus and is available in intravenous, oral and/or topical (skin and ophthalmic) preparations. It has been a good choice for parenteral therapy, preferably as combination treatment, in moderate/severe suspected staphylococcal infections. However, in the present 7-year study, fusidic acid-resistant isolates, MRSA or MSSA, accounted for invasive infections and SSTIs. Resistance to this antibiotic has also been detected in CA-MSSA [26–28] as well as CA-MRSA strains [18, 29–32] from other European countries. Both clonal  and non-clonal  dissemination of fusidic acid-resistant S. aureus have been well documented to occur in the community.
In contrast to CA-MRSA isolates from other continents, those belonging to the European ST80 clone have been reported to frequently express reduced susceptibility to fusidic acid and commonly carry the plasmid-located far-1 (also known as fusB) gene encoding fusidic acid resistance [18, 32]. It is notable that in some previous reports from Europe these isolates have been reported as intermediate to fusidic acid . This is related to the fact that due to the nonexistence of CLSI breakpoints for fusidic acid the authors had applied the previous breakpoints of the French Society for Microbiology . According to these breakpoints, MICs of 4 to 16 μg/ml were considered indicative of an isolate intermediate to fusidic acid, whereas the breakpoint for resistance was ≥32 μg/ml. In 2010, the EUCAST breakpoints have been accepted in Europe and MICs of ≥2 μg/ml are considered representing resistance to fusidic acid . Recently, Jones et al. have evaluated the activity of fusidic acid against S. aureus isolates using broth microdilution, disk diffusion, and Etest methods and proposed ≤1 μg/ml (disk zone ≥22 mm) as breakpoints for susceptibility and ≥4 μg/ml (disk zone ≤19 mm) for resistance .
In Greece, despite the high prevalence of fusidic acid-resistant S. aureus isolates in the community, the use of oral and/or topical (skin and ophthalmic) preparations of fusidic acid continues to be extensive (authors' unpublished data). Parenterally administered fusidic acid is used at a lower scale, but yet attention should be paid to this, as it usually concerns patients with severe infections. The prolonged use of fusidic acid as topical monotherapy for chronic skin conditions appears to have resulted in the emergence of resistance among S. aureus isolates in some countries [35, 36], making this agent less active both for topical and systemic therapy . In areas with high rate of fusidic acid resistance, topical and oral or systemic fusidic acid monotherapy should be restricted, in order to prevent the further spread of fusidic acid-resistant S. aureus isolates . Topical use of a valuable systemically active agent is best to be avoided . High frequency of fusidic acid resistance may persist despite such restrictions and may represent the development of a fusidic acid-resistant S. aureus reservoir in the community.
Since the time that the increase of CA-MRSA infections was well appreciated, clindamycin, an antibiotic that is considered to be active against a high percentage of CA-MRSA and MSSA isolates, has been extensively used empirically, particularly in pediatric infections [8, 37]. In areas with low rate of clindamycin resistance, clindamycin is the recommended empirical antistaphylococcal therapy for hospitalized pediatric patients with clinical entities likely caused by CA-MRSA, such as cutaneous abscesses or pneumonia with empyema. In parallel to the antistaphylococcal activity, clindamycin possesses the ability to suppress production of PVL by S. aureus
in vitro at the translational (ribosomal) level . In the present study, PVL was produced by most of the MRSA isolates.
According to our recent data, the rate of clindamycin resistance in MRSA isolates from 2007 to 2009 has been 23.8-31.2% and that of MSSA isolates 15.4-38.1%. With the increased rates of clindamycin resistance among CA-MRSA and CA-MSSA isolates in our area, clindamycin is not useful for empiric monotherapy treatment for suspected S. aureus invasive infections and severe SSTIs. Thus, one should reconsider the empirical treatment regimen, which, in our area, until recently consisted primarily of clindamycin or vancomycin. Empirical clindamycin treatment for suspected staphylococcal infections is not recommended in areas where the proportion of CA-MRSA isolates exceeds 10% to 15% . Among staphylococcal infections due to isolates exhibiting inducible clindamycin resistance [39, 40], the risk of treatment failure during clindamycin therapy is increased when there is a high bacterial inoculum .
In Greece, tetracycline resistance is common among the CA-MRSA isolates. This is a significant difference between the European ST80 clone and the USA300 . Actually in the United States, oral tetracyclines, such as minocycline or doxycycline, are used as an alternative treatment for suspected S. aureus SSTIs in children 8 years of age and older . Tetracyclines are particularly prescribed by the dermatologists. In Central Greece, trimethoprim-sulfamethoxazole and ciprofloxacin resistance appeared to be rare, whilst vancomycin, rifampin, and linezolid susceptibility was universal, reinforcing the utility of these compounds for the time being at least.
The increasing rate of CA-MRSA infections has important implications for patient management and for the selection of appropriate antimicrobial therapy. If possible, single treatment which is active against both MSSA and MRSA should be considered as an empirical therapy in patients with staphylococcal infections. However, caution should be paid in monotherapy, as well as in combination therapy, using antibiotics recently recognized as having reduced activity against S. aureus. Currently in Central Greece, vancomycin, linezolid, or daptomycin, as monotherapy or in combination with rifampin or gentamicin, appear to be choices for empiric therapy of severe or difficult to treat possible staphylococcal infections. Vancomycin is inferior to nafcillin for the treatment of invasive MSSA infections including bacteremia and endocarditis. Some experts have proposed nafcillin or another penicillinase-resistant penicillin in combination with an agent active against MRSA isolates, such as vancomycin, for critically ill patients . If a penicillinase-resistant penicillin is used as an empiric combination therapy, attention should be paid to the selection of the second antibiotic. Trimethoprim-sulfamethoxazole is another treatment choice, particularly in SSTIs.
Mild/moderate SSTIs can be empirically treated with clindamycin or trimethoprim-sulfamethoxazole [41, 42], either alone or in combination with rifampin. However, optimal management for pediatric CA-MRSA SSTIs has yet to be determined because of limited clinical studies and efficacy data. Drainage of the abscesses is the key to treatment and is helpful in isolating the causative pathogen. A number of studies have shown primarily good outcomes in patients whose abscess is drained regardless of the antibiotics administered [43, 44].
There were several possible limitations to our study. This was a retrospective study with its inherent limitations in data availability. We did not have detailed information on maternal infection history or that of other family members, especially for the very young infants. Additionally, although we tried to classify the infections as accurately as possible, a small number of patients with SSTIs which was mainly described in the charts as pyodermatitis, was classified as "unspecified pustular lesions". However, this does not influence the statistical analysis for impetigo. Infections classified as impetigo were clearly described with this diagnosis in the medical records. Finally, the molelular analysis of the isolates was based on MLST typing of representative isolates. Pulse-field gel electrophoresis could have provided additional information on the clonal relationship of the isolates.
Because antimicrobial resistance continues to evolve, it is important to continue monitoring S. aureus infections in our area. This provides valuable data on resistance trends and contributes to more effective treatment recommendations for regional and national use. Physicians caring for children throughout Europe and the world are likely to face the same problem with CA-MRSA that we have described in this report, as the ST80 or other CA-MRSA clones are introduced into their communities.