Panton-Valentine Leukocidin Is Not the Primary Determinant of Outcome for Staphylococcus aureus Skin Infections: Evaluation from the CANVAS Studies

The impact of Panton-Valentine leukocidin (PVL) on the severity of complicated skin and skin structure infections (cSSSI) caused by Staphylococcus aureus is controversial. We evaluated potential associations between clinical outcome and PVL presence in both methicillin-resistant S. aureus (MRSA) and methicillin-susceptible S. aureus (MSSA) isolates from patients enrolled in two large, multinational phase three clinical trials assessing ceftaroline fosamil for the treatment of cSSSI (the CANVAS 1 and 2 programs). Isolates from all microbiologically evaluable patients with monomicrobial MRSA or MSSA infections (n = 473) were genotyped by PCR for pvl and underwent pulsed-field gel electrophoresis (PFGE). Genes encoding pvl were present in 266/473 (56.2%) isolates. Infections caused by pvl-positive S. aureus were associated with younger patient age, North American acquisition, and presence of major abscesses (P<0.001 for each). Cure rates of patients infected with pvl-positive and pvl-negative S. aureus were similar overall (93.6% versus 92.8%; P = 0.72), and within MRSA-infected (94.5% vs. 93.1%; P = 0.67) and MSSA-infected patients (92.2% vs. 92.7%; P = 1.00). This finding persisted after adjustment for multiple patient characteristics. Outcomes were also similar when USA300 PVL+ and non-USA300 PVL+ infections were compared. The results of this contemporary, international study suggest that pvl presence was not the primary determinant of outcome in patients with cSSSI due to either MRSA or MSSA.

The role of PVL in determining the outcome and severity of S. aureus cSSSI remains one of the most controversial topics in the field of staphylococcal research [9]. The presence of pvl in S. aureus strains has been associated with severe infections in some [10][11][12][13][14][15] but not all [9,16,17] previous studies. For example, our group has reported that pvl presence was paradoxically associated with a better clinical outcome than cSSSI infections caused by pvlnegative MRSA [17,18]. For this reason, the role of pvl in cSSSI is unresolved.
The goal of the current study was to evaluate the impact of pvl presence on the outcome of patients with cSSSI caused by both MRSA and MSSA. To accomplish this goal, we used clinically well-characterized S. aureus isolates collected from patients with cSSSI who were enrolled in two phase 3 multinational clinical trials. The objectives of the current study were: 1) to confirm our previous observations that pvl is not the primary determinant of clinical outcome in patients with cSSSI due to MRSA, and 2) to test whether this same finding is also encountered in MSSAinfected patients. In addition, we aimed to evaluate whether outcomes of infections caused by USA300 PVL+ isolates differed from those caused by non-USA300 PVL+ isolates.

Methods
Patients and settings CANVAS 1 (NCT00424190) and CANVAS 2 (NCT00423657; Ceftaroline versus Vancomycin in Skin and Skin-Structure Infection) were two methodologically identical, randomized, double-blind, multinational phase 3 clinical trials that compared the efficacy and safety of intravenous ceftaroline fosamil monotherapy with intravenous vancomycin plus aztreonam combination therapy for the treatment of adults with cSSSI caused by gram-positive and gram-negative organisms. The study designs of the CANVAS trials have been published in detail [19][20][21].
A total of 1378 patients from 111 participating centers in 12 countries were enrolled from February 2007 through December 2007 [19,20]. Men and non-pregnant women aged $18 years and diagnosed with cSSSI severe enough to require initial hospitalization or treatment in an emergency department and $5 days of intravenous antimicrobial therapy and that involved either (1) deep soft tissue or required deep surgical intervention, or (2) cellulitis or abscess of a lower extremity in patients with diabetes mellitus (DM) or well-documented peripheral vascular disease (PVD), were eligible for the study. Purulent or seropurulent drainage and/or collection or the presence of at least three of the following signs or symptoms was also required for participation: erythema, fluctuance, heat and/or localized warmth, pain and/or tenderness to palpation, fever (temperature .38uC) or hypothermia, white blood cell (WBC) count of .10,000 / mm 3 , or .10% immature neutrophils irrespective of WBC count.
Patients were randomized to receive either ceftaroline fosamil (600 mg every 12 h) followed by normal saline placebo or vancomycin followed by aztreonam (1 g each every 12 h) for 5-14 days. Bacterial isolates were obtained from all patients at baseline by needle aspiration or a surgical procedure. Test-of-cure evaluation was performed 8 to 15 days after administration of the last dose of the study drug. The clinical response at the test of cure was classified as cure, failure, or indeterminate [21]. Clinical cure was defined as total resolution of all signs and symptoms of the baseline infection or improvement such that no further antimicrobial therapy was necessary. The microbiologically evaluable (ME) population was defined as clinically evaluable patients with $1 bacterial pathogen isolated from blood or the cSSSI site at baseline, excluding patients whose baseline cultures revealed monomicrobial Pseudomonas aeruginosa or anaerobic infection. In the present investigation, the analysis population was defined as microbiologically evaluable patients with cSSSI due to monomicrobial S. aureus and a clinical response at test of cure evaluation. The investigation was approved by the Duke University Medical Center Institutional Review Board.

Pulsed Field Gel Electrophoresis
Pulsed-field gel electrophoresis (PFGE) with the restriction enzyme SmaI was performed as previously described [22]. The PFGE profiles were analyzed using BioNumerics software (Applied Maths, Kortrijk, Belgium). A similarity coefficient of 80% was used to define pulsed-field type clusters. Dice coefficients (pairwise similarity) were calculated for each pair of isolates, and a dendrogram was constructed using an optimization value of 0.50% and a position tolerance ranging from 1.25% to 1.35% (the end of the fingerprint). PFGE profiles were interpreted according to a published typing schema [22].
PCR assays for genotyping S. aureus genomic DNA was extracted using an ultraclean microbial DNA kit (MO BIO Laboratories, Inc., Carlsbad, CA) in accordance with the manufacturer's instructions. PCR assays were used to screen the S. aureus genome for the putative bacterial virulence toxin pvl for all S. aureus isolates using previously described methods [17,18].

Statistical methods
Cure rate by pvl status was stratified by markers of severity, including abscess versus non-abscess infection type, presence of fever, white blood cells (.10,000/ mm 3 ), infection size (baseline infection area of .100 cm 2 ), presence of diabetes, patient age, and study medication. Infection size was calculated as the product of the length and width of the primary infection site at its longest and widest axes, respectively. Continuous variables were compared between groups by using the two-sample t test. Categorical variables were analyzed using Fisher's exact test, its 2-by-K extension, or the Cochran-Mantel-Haenszel test, as appropriate, for stratified analyses. Our a priori hypothesis, that the presence of pvl in MRSA cSSSI is not associated with a worse clinical outcome, was tested using Fisher's exact test. All reported P values are two sided. P values of ,0.05 were deemed statistically significant. Since subjects were not randomized on the basis of pvl status, all P values calculated should be considered descriptive and not inferential. Results were obtained using SAS software, version 9.1.3 (SAS Institute Inc, Cary, NC, USA).

Study population and baseline characteristics
A total of 473 patients from the CANVAS studies with cSSSI due to monomicrobial S. aureus (280 MSSA, 59.2%; 193 MRSA, 40.8%) and for whom clinical outcome were available, were included in the current study. Among these 473 study patients, 253 (53.5%) received ceftaroline fosamil and 220 (46.5%) received vancomycin plus aztreonam. Patients were predominantly white, male, and North American, with an average age of 45 years.
A comparison of infections caused by USA300 PVL+ with non-USA300 PVL+ isolates (Table 3) demonstrated that USA300 PVL+ infections were associated with non-Caucasian ethnicity, North America, prior antibiotic use, MRSA, and the absence of hospitalization as a risk factor (all P,0.001). Neither the presence of a deep abscess, the receipt of surgical intervention nor the antibiotic regimen received differed between the two groups. Outcomes from USA300 PVL+ and non-USA300 PVL+ infections were similar overall (P = 0.80) and when stratified across multiple sub-groups including methicillin-susceptibility (Table 4).

Discussion
The significance of PVL in the outcome of staphylococcal infections remains unresolved. The present study used a large, contemporary, multinational collection of MRSA and MSSA isolates to evaluate the impact of pvl presence on clinical outcome of patients with cSSSI. Despite adjustment for a number of clinically relevant variables, pvl presence was not associated with a higher risk of failure. This finding has several implications.
This analysis of the isolates collected in the CANVAS 1 and 2 studies add to an increasing body of evidence [17,18,23] that the simple presence of pvl is not the primary outcome determinant in all S. aureus infections. This finding persisted after adjusting for patient comorbidities, infection type, infection severity, and patient age. The results of this study are consistent with two previous reports by our group evaluating potential associations between pvl and cSSSI caused by MRSA [17,18]. However, the current investigation extends our understanding of the role of pvl in staphylococcal pathogenesis by considering its relevance in MSSA infections. While pvl is currently considered primarily in the setting of MRSA infections [1][2][3][4], its presence in MSSA is also well established [3,6,8,[24][25][26]. Collectively, these results underscore the recent conclusions of Otto [9] that factors other than the mere presence of pvl are the primary determinants of clinical outcome in patients with S. aureus cSSSI.
However, the infections process is a continuum from onset to resolution and while outcomes did not differ, the presence of pvl was associated with important clinical characteristics in patients with S. aureus cSSSI. Patients with pvl-positive cSSSI were younger, more likely to have major abscess, more likely to be North American and less likely to have health care contact, findings that have been previously noted [17,26]. In contrast, patients with pvlnegative strains tended to be older and have comorbidities such as diabetes or peripheral vascular disease. These differences persisted in both the MRSA and MSSA subgroups, and may reflect in part the epidemiologic settings in which the infections were acquired. Healthcare-associated S. aureus infections are less likely to contain pvl [2] and are also more likely to cause cSSSI in older patients with more comorbid conditions. By contrast, presence of pvl is often associated with community-acquired infection. Thus, the presence of pvl is associated with a different spectrum of disease in a distinct subgroup of patients and also influences the treatment received in that pvl+ cSSSI more frequently required surgical management. However, the need for surgical intervention is a clinical decision and prior knowledge of pvl status is unlikely to be of assistance at the time this decision is made.
Following the provision of appropriate management, the current investigation underscores the fact that pvl is not the principle determinant of clinical outcome in patients with either MRSA or MSSA cSSSI. Factors other than pvl should be Table 2. Clinical outcome of cure for 473 patients with methicillin-resistant and methicillin-susceptible S. aureus complicated skin and skin structure infections by pvl gene status, stratified by infection severity, diabetes, patient age, and study medication.  considered for future therapeutic and diagnostic strategies in combating S. aureus cSSSI. For example, other characteristics including alpha hemolysin [27], a new bicomponent leukotoxin [28], the antagonistic effects of antibody to control proliferation of PVL-producing S. aureus [29], and host species specificity of the neutrophil lysis properties of PVL [30], may also be associated with the clinical outcome of S. aureus cSSSI. It is also possible that the genes coding for PVL may simply be markers of particular bacterial clones and are not themselves directly involved in the pathogenesis of more severe infection.
Due to the predominance of USA300 in North America as previously described and also in this study, most reports regarding PVL from this region focus on this strain of S. aureus. The multinational nature of the CANVAS studies allowed us to compare infections caused by USA300 PVL+ with non-USA300 PVL+ isolates. Despite some differences in epidemiology regarding the likelihood of community-acquisition, deep abscesses were the main clinical manifestation of both groups and, importantly, the need for surgery and outcomes were equivalent. This suggests that findings relating to the clinical manifestations and outcomes of cSSSI due to USA300 are likely to be generalizable to other PVL+ infections. However, the epidemic spread of USA300 in North America compared to PVL+ clones elsewhere in the world remains unexplained and our data indicate that factors other than PVL may play a more important role in the success of USA300. The S. aureus isolates collections used in this study are large, contemporary, multinational, cSSSI specific, and clinically characterized to the level of rigor of a U.S. Food and Drug Administration (FDA) registrational trial. Limitations include the fact that we did not evaluate quantitative pvl expression or the presence of pvl polymorphisms, as these properties may influence the function of gene products [31,32], although pvl polymorphisms have not been found to be of clinical importance [26]. By design, this investigation focused only upon infections in which the causative pathogen was available for culture. We were also unable to consider potential relationships between vancomycin trough levels and outcome.
In summary, this study has demonstrated that presence of pvl in S. aureus strains is not associated with a worse outcome in patients with cSSSI. This finding persisted regardless of methicillinresistance status. These findings indicate that factors other than the presence of pvl are the primary determinants of clinical outcome in S. aureus cSSSI. Future efforts are necessary for a more complete understanding of the role of pvl in the pathogenesis of serious infections due to S. aureus.