Retrospective study of necrotizing fasciitis and characterization of its associated Methicillin-resistant Staphylococcus aureusin Taiwan

Necrotizing fasciitis (NF) is a rare and life-threatening infection that involves deep soft tissue and is characterized by widespread fascial necrosis, with a mortality of 25.3-73% [1–3]. Necrotizing fasciitis typically is polymicrobial infection caused by aerobic and anaerobic organism that participate in synergistic infection; whereas Group A Streptococcus, Clostridium spp., Vibrio spp. and Klebsiella spp. have been the cause of monomicrobial infection. Although Staphylococcus aureus is the most prevalent pathogen in both hospitals and communities and MRSA has become a common isolate associated with skin and soft tissue infections globally over the past few years, monomicrobial MRSA necrotizing fasciitis has been reported only in a few studies [4–6]. In Taiwan, MRSA accounts for 53-83% of S. aureus isolates in most major hospitals [7, 8]. Community-acquired MRSA (CA-MRSA) has become increasingly endemic, accounting for 74% and 28% of all MRSA in two studies that were performed in northern and central Taiwan, respectively [9]. Recently, MRSA has been identified as an important cause of necrotizing fasciitis in Taiwan [10].

Several virulence factors are reportedly associated with the pathogenicity of skin and soft tissue infections. Panton-Valentine leukocidin (PVL), a pore-forming exotoxin, and γ-hemolysin (Hlg) are associated with pus-forming diseases. CA-MRSA isolates with the PVL genes generally cause skin and soft tissue infections and necrotizing fasciitis [4, 11]. Toxic shock syndrome toxin 1 (TSST-1) can cause in part a desquamative skin rash and toxemic syndrome [12]. Exfoliatin A (ETA) and exfoliatin B (ETB) are associated with numerous blistering skin diseases and staphylococcal scalded-skin syndrome [13]. The variable clinical presentations of S. aureus skin and soft tissue infections are believed to be associated with the particular combinations of virulence factors and regulatory genes given an appropriate genetic background [14].

In staphylococcal species, low-affinity penicillin-binding protein 2a, which is responsible for methicillin resistance, is produced by the mecA gene in an exogenous mobile staphylococcal chromosomal cassette mec (SCCmec) [15, 16]. Hence a methicillin-susceptible S. aureus (MSSA) isolate may acquire the SCCmec elements and thereby become an MRSA isolate [17]. Based on variations in the ccr operon and mec complexes of the SCCmec, at least seven SCCmec types have been identified as having some clinical implications [18]. The SCCmec IV element (~24 kb) is most commonly identified in CA-MRSA isolates, whereas the SCCmec II and III elements are often found in HA-MRSA isolates [19, 20].

The management of NF absolutely depends on the prompt administering of suitable antibiotics and adequate surgical debridement. Increasing prevalence of MRSA, the choice of appropriate antibiotics for NF patients in the community must now take into account the increasing global prevalence of MRSA and the patterns of antimicrobial susceptibility [21]. Therefore, a study population of 247 NF cases in 2004-2008 and their associated MRSA isolates were investigated to understand the epidemiological or clinical effects of MRSA on NF cases and clinical and genetic differences between CA- and HA-MRSA. Further, antimicrobial susceptibility, SCCmec types, pulsed field gel electrophoresis (PFGE) patterns, virulence factors, and MLST types of MRSA strains that were isolated from 16 NF patients in 2008 were examined.

The mainstay treatments of necrotizing fasciitis are urgent operation for decompression and drainage, subsequently repeated debridement, and appropriate antibiotics. The selection of appropriate antimicrobial agents for any suspected necrotizing fasciitis must take into account the nature of patient's exposure and local epidemiologic data. However, there are no reliable epidemiological or clinical risk factors that could distinguish patients infected with MRSA from those infected with MSSA or non-S. aureus [33]. With the explosion of MRSA infections, recent reports have established that MRSA is responsible for 3.6% to 39% of NF disease [1, 4–6]. In this study, the prevalence of MRSA in 247 NF cases over a four-year period was 19.8% (Table 1 and Figure 1). The variation in the prevalence of MRSA necrotizing fasciitis around the world may reflect variations in genetic background and local epidemiology. MRSA is gradually becoming the main monomicrobial pathogen to cause NF in Taiwan, reflecting changes in the microbial flora in this area. In this investigation, the observed mortality rate did not significantly vary among the four groups, even though NF patients with HA-MRSA infection had a higher incidence of comorbidity on admission than those infected by CA-MRSA, MSSA or non-S. aureus group. However, the amputation rate and reconstruction rate for NF patients with MRSA infection were 18.4% and 30.6% respectively - significantly higher than those infected by non-S. aureus. These analytical findings further highlight the role of MRSA as an important pathogen in causing a destructive and deep-seated infection.

Recent reports have demonstrated that the CA-MRSA strains with specific virulence factors have a shorter doubling time than HA-MRSA strains and tend to be more virulent [34]. PVL has been associated with epidemic CA-MRSA strains causing lethal necrotizing fasciitis [35]. Contrary to these findings, in the present investigation, all of the patients who were infected by CA-MRSA survived. Miller [4] reported no deaths in their series, hypothesizing that necrotizing fasciitis that is caused by CA-MRSA may be less virulent than similar infections that are caused by other organisms. According to the findings herein, patients who were infected by CA-MRSA tended to have less severe clinical characteristics in terms of comorbidity, amputation rate and the involvement of lower extremities than those with HA-MRSA infections, suggesting that it may be associated with limited elevation of C-reactive protein in patients as well as lower virulence of CA-MRSA strains which differed distinctly from the highly virulent PVL-positive CA-MRSA strains spreading elsewhere.

In the present study, the PVL genes were only identified in two CA- and one HA-MRSA isolates. Genes for the virulence factors TSST-1 and ETA were present only in a few MRSA isolates. However, patients who were infected by pulsotype C isolates with the ETA gene all underwent amputation (Table 3). Unlike the Hlg genes, which were present in all isolates, the ETB gene was absent from all isolates. Other series has also reported a lower frequency of toxin genes carriage. Miller reported that no TSST-1, ETA and ETB genes were identified in any of the CA-MRSA strains [4]. The severity of PVL-specific pathology remains controversial. The PVL genes have been detected in approximately 50% of S. aureus isolates that are responsible for subcutaneous abscesses and cellulitis, but not in isolates that are associated with deep-seated infections [36]. A study of isogenic PVL strains of pulsotypes USA300 and USA400 also found that PVL was not the major virulence factor for necrotizing fasciitis [37]. Conversely, the PVL gene has been detected in five NF-associated ST8/pulsotype USA 300/SCCmec IV CA-MRSA isolates [4]. A recent investigation stated that PVL tends to damage the muscle rather than skin in young mice [38]. These results are consistent with the clinical characteristics of necrotizing fasciitis, which typically damages deep tissues but may spare skin. Thromboangiitis obliterans followed by the occlusion of cutaneous perforators may be the leading cause of NF cases with extensive skin necrosis - rather than direct cutaneous invasion.

In Taiwan, ST239/pulsotype A type is the most prevalent clone of HA-MRSA and the ST59/pulsotype C and D isolates are the dominant clones of CA-MRSA [30, 39–41]. Moreover, SCCmec VII and multi-drug resistant CA-MRSA isolates have been identified in pediatric patients with skin and soft tissue infections in northern Taiwan [29]. This investigation identified various NF-associated MRSA strains, which have been the prevalent clones in Taiwan. These strains differ distinctly from the particular ST8/pulsotype USA300 CA-MRSA clone, which has been identified as a dominant cause of necrotizing fasciitis in the USA [4, 6]. The results herein verified that the SCCmec III type was found only in HA-MRSA and the SCCmec IV element was present in both CA- and HA-MRSA isolates with ST59/pulsotype C and pulsotype D (Table 3), suggesting transmission of the community strain of MRSA into the hospital setting, and that the community strain had became a hospital-associated pathogen in Taiwan [9].

Empiric antibiotic therapy for NF patients must be modified for the broad coverage of the regional MRSA strains. Earlier, antimicrobial susceptibility patterns could be used to differentiate CA-MRSA from HA-MRSA isolates; however, this approach is no longer reliable because CA-MRSA may also develop resistance to non-β-lactam antibiotics [34]. Clindamycin has been a common prescribed medication for MRSA infection. In this investigation, both CA- and HA-MRSA isolates exhibited extremely high resistance (>90%) to erythromycin and clindamycin (Table 1), which were contradicting findings obtained from different countries, but supporting other findings from Taiwan [5, 7, 42]. Trimethoprim-sulfamethoxazole was the only drug associated with different susceptibility between CA- and HA-MRSA strains, to which CA-MRSA isolates were significantly more susceptible than HA-MRSA isolates [9]. For complex skin and soft tissue infections associated with considerable comorbidities, treatment with appropriate antimicrobials is required to reduce clinical failures. Delayed treatments are associated with the morbidity and mortality of MRSA infection [43]. The most common antibiotic prescribing error in this investigation was inadequate coverage of the pathogen MRSA. Even though vancomycin was administered to 46.9% of MRSA patients after the susceptibility test, 83.7% of the patients with MRSA infections herein underwent empiric therapy with antibiotics that were not active against the MRSA strains at the time of hospital admission. This fact may explain the higher amputation and reconstruction rates of patients who were infected by MRSA than those of patients who were infected by MSSA or non-S. aureus. An effective empiric antimicrobial therapy may be critical to ensure a favorable surgical outcome for NF patients with MRSA infection. The empiric use of vancomycin should be considered for NF patients based on epidemiologic data and nature of patient's exposure in an area where high prevalence of methicillin resistance was noted such as Taiwan. Additionally, to prevent the emergence of the resistant strains and prolong the effectiveness of currently available antimicrobials, empiric therapy using agents such as vancomycin should not be prescribed for immuno-competent patients with mild skin and soft-tissue infections [44].

MRSA has emerged as an important cause of necrotizing fasciitis in Taiwan. In this investigation, the prevalence of MRSA in 247 NF cases over a four-year period was 19.8%. NF patients with MRSA infection presented more severe clinical outcomes in terms of amputation rate and reconstruction rate than those with MSSA or non-S. aureus infection. In contrast to CA-MRSA infection, NF patients with HA-MRSA infection had a significantly higher amputation rate, comorbidity, C-reactive protein level, and involvement of lower extremity. Further, HA-MRSA isolates were more resistant to trimethoprim-sulfamethoxazole. The ST239/pulsotype A, ST59/pulsotype C and D MRSA are the major pathogenic clones to cause necrotizing fasciitis. The ST59/pulsotype D/SCCmec V/PVL MRSA isolates were found in both CA- and HA-MRSA; the ETA gene was found in ST59/pulsotype C/SCCmec type IV CA- and HA-MRSA. As well as surgical debridement, initial empiric antimicrobials with a broad coverage for MRSA should be considered in the treatment of NF patients in an endemic area.