Methicillin-resistant Staphylococcus aureus (MRSA) is an important pathogen associated with complicated skin and soft-tissue infections (cSSTIs). MRSA isolates are classified according to the type of staphylococcal cassette chromosome mec (SCCmec), with SCCmec types I, II, and III associated with healthcare-associated MRSA and SCCmec type IV associated with community-associated MRSA (CA-MRSA). Most CA-MRSA isolates obtained from cSSTIs in the United States carry the gene encoding Panton-Valentine leukocidin (PVL). We sought to determine the impact of SCCmec type on clinical presentation, severity of the cSSTI, and healthcare resource utilization among patients that were enrolled in an international Phase 4 study designed to compare the clinical and microbiological outcomes and safety of linezolid and vancomycin for the treatment of culture-confirmed MRSA-cSSTIs [1] . (This study was presented at 111th General meeting of the American Society of Microbiology, New Orleans, LA, May 21-24, 2011).

The parent study was approved by the local Institutional Review Board at each site and informed consent was obtained from each study participant. A total of 545 MRSA isolates were evaluated from specimens obtained by fine-needle aspiration, tissue biopsy, or collection of debridement tissue from patients with cSSTIs. Superficial swabs of infection or wound sites were not allowed. MRSA was identified by local laboratories using routine methods, and isolates were submitted to a central laboratory for SCCmec typing (I–IV) and PVL-encoding genes screen using a multiplex polymerase chain reaction (PCR) as previously described [2] .

We defined CA-MRSA as SCCmec type IV, PVL positive strain [3] . Patients were treated with linezolid 600 mg orally or intravenously every 12 hours or vancomycin 15 mg/kg of body-weight intravenously every 12 hours, with dose adjustment based on trough levels and creatinine clearance, for up to 28 days. Baseline variables were collected at screening to calculate Wilson severity score, a validated test to assess patient risk and predict clinical outcome (cure or failure) [4] . Length of hospital stay (LOS) analysis used Kaplan Meier survival techniques. It was based on the daily hospital location data and was censored at end of study. Duration of intravenous therapy was compared using the Student t test. Statistical analyses were performed using SAS v.8.2 (SAS Institute, Cary, NC, www.sas.com).

SCCmec type IV accounted for 69.0% (376/545) of cSSTIs MRSA isolates. In total, 54.1% (295/545) of isolates were CA-MRSA as defined by SCCmec type IV, PVL+. Most were from the United States and subsequently characterized as USA300 sequence type 8 (ST8) [2] . Thirty-one percent (169/545) of MRSA isolates were SCCmec type I, II, or III. CA-MRSA was most commonly associated with an abscess presentation (Table 1). SCCmec types I to III were most commonly seen with surgical wound infections (Table 1).

SCCmec type IV PVL+ had lower Wilson severity scores (23.8) compared with patients with SCCmec types I (38.0), II (45.5), III (33.7) and IV, PVL– (40.6) (p < 0.05 for each pairwise comparison). In the linezolid treatment group, SCCmec type IV, PVL+ was associated with a shorter mean intravenous antibiotic treatment duration and shorter length of hospital stay which were statistically significant when compared to SCCmec type I to III and type IV, PVL– (Table 2). In the vancomycin treatment group, patients with SCCmec type IV, PVL+ had statistically significant shorter length of hospital stay compared with SCCmec types I to III and type IV, PVL–. Compared to SCCmec type I, there was a statistically significant shorter mean intravenous antibiotic treatment duration for SCCmec type IV, PVL+ in patients treated with vancomycin (Table 2).

Although some studies have shown that SCCmec II is associated with a higher mortality rate compared to the other SCCmec types [5] , we have previously shown no significant difference between the type of SCCmec and clinical cure and mortality among patients with MRSA cSSTIs [6] . This has also been confirmed by other studies [7] [8] . However, an association between length of stay (LOS) and SCCmec type has been reported by other authors. Davis et al. showed that the mean length of stay was significantly longer for SCCmec II/III (18 days) than for SCCmec IV (10 days) [8] . We found similar results with MRSA SCCmec type IV, PVL+ being associated with a shorter length of stay compared with SCCmec types I to III and type IV, PVL–. The long LOS

MRSA, methicillin-resistant Staphylococcus aureus; SCCmec, staphylococcal cassette chromosome mec; PVL, Panton-Valentine leukocidin. ^*p < 0.05 versus IV PVL+; Fisher exact test. Because of rounding, some values may not add up to 100%. ^**Includes infected skin ulcer, trauma wound infection, decubitus ulcer and infected burn.

^aPatients who started on IV linezolid study therapy are included in this analysis. ^bThree patients who started on oral therapy and received IV during their course of therapy are included in this analysis. ^*p < 0.05 versus IV PVL+; one-way analysis of variance. IV, intravenous; MRSA, methicillin-resistant Staphylococcus aureus; SCCmec, staphylococcal cassette chromosome mec; PVL, Panton-Valentine leukocidin; SE, standard error.

for SCCmec IV reported by Davis et al. might have been due to the presence of PVL– strains which have a longer LOS compared to the PVL+ strains as seen in our study. This is consistent with a lower Wilson severity score found in SCCmec type IV, PVL+ compared the other SCCmec types and may be partially explained by MRSA SCCmec type IV, PVL+ being associated with abscesses, which are often less severe than other types of cSSTIs.

PVL has been linked to cSSTIs [9] , however its presence in these infections has not been associated with worse outcomes. Bae et al. reported that patients with cSSTIs caused by PVL+ MRSA strain were more likely to achieve cure than patients with PVL− MRSA strain infections (91.6% vs. 80.7%) [10] . We obtained similar results in that SCCmec type IV, PVL+ strains were associated with a lower severity score and shorter LOS than PVL− strains. The higher Wilson severity score seen with non-SCCmec type IV might be related with the higher involvement of surgical wound infection which is an independent risk factor for worse outcome in this scoring system [4] . Similarly, the longer LOS of non-SCCmec type IV is most likely due to the higher Wilson severity scores in these patients.

Although CA-MRSA cSSTIs are on the rise, the data from this prospective Phase IV clinical trial suggest these infections are associated with a less severe type of cSSTI and shorter LOS compared with the non-CA-MRSA infection.

The study was sponsored by Pfizer Inc. Statistical support was provided by Arlene Reisman and Michele Wible of Pfizer Inc.

EG and TC were not paid for their contributions to this manuscript. PAH and DEM are employees and shareholders of Pfizer Inc. DBH, formerly of Pfizer, was an employee and shareholder of Pfizer Inc at the time the study was conducted.