CHAPTER 34 Methicillin-Resistant Staphylococci
ETIOLOGY OF STAPHYLOCOCCAL INFECTIONS
Bacteria of the genus Staphylococcus are gram positive, facultatively anaerobic cocci that exist as part of the normal cutaneous and mucosal microflora of mammals and birds. Many staphylococcal species also are opportunistic pathogens capable of causing serious infections of the skin and other body tissues and cavities.1,2 When cutaneous or systemic disease disrupts the skin’s surface defense mechanisms, skin infection (pyoderma) or otitis externa may result from these same staphylococcal species. Invasive infections involving the genitourinary tract, respiratory tract, joints, and body cavities also may result either from ascension along epithelial tracts, introduction via penetrating wounds, or hematogenous spread.
Although the coagulase-negative staphylococci are receiving renewed attention with regard to their medical importance,3,4 it is the coagulase-positive species that have long been considered to be the important pathogens in veterinary medicine. Although both groups have been isolated from healthy feline skin and from cats with skin lesions, reports of pathogenic coagulase-negative staphylococci in cats remain rare.5
The coagulase-positive species that normally colonize the skin of the domestic cat are Staphylococcus intermedius and Staphylococcus aureus, although there is disagreement in the veterinary literature regarding which is dominant, and this factor may vary by geographical region.6–9 S. aureus also is the primary species colonizing human skin, and some investigators have suggested that feline carriage of S. aureus is reflective of human-to-cat transmission.
S. intermedius was first recognized as a species distinct from S. aureus in the mid-1980s, and the phylogenetic structure and nomenclature of S. intermedius has changed again as a result of advances in molecular characterization.10,11 The S. intermedius group (SIG) is now comprised of three genetically demonstrable species: S. intermedius, S. pseudintermedius, and S. delphinii, each of which occupy distinct ecological niches.10 The primary canine pathogen is now known to be S. pseudintermedius, and this is likely to be the case for cats as well. However, for discussion purposes the author uses the nomenclature common to the feline literature, keeping in mind that most reports on S. intermedius prior to late 2007 are likely to refer in actuality to S. pseudintermedius.
Finally, Staphylococcus hyicus and Staphylococcus schleiferi also have been isolated from the skin of healthy cats and from cats with skin lesions.5 S. schleiferi is a coagulase-variable species, with both coagulase-positive and negative subspecies. S. schleiferi subsp. coagulans is thought to be the primary subspecies occurring in dogs, in whom it causes infections that are commonly associated with prior antibiotic use.12,13 In human beings it is the coagulase-negative variant (S. schleiferi subsp. schleiferi) that has been shown to be most commonly pathogenic, causing postsurgical skin and soft-tissue infections.14 Isolation of either subspecies of S. schleiferi from cats remains exceedingly rare5 and will not be considered further in this text. However, principles discussed for other staphylococcal species apply to S. schleiferi as well.
ETIOLOGY OF METHICILLIN RESISTANCE
Since the inception of antibiotic use in the practice of modern medicine, staphylococci have evolved in response to the presence of antimicrobial drugs in their environments. Currently all staphylococcal species that infect human beings and domestic animals exhibit some degree of antimicrobial resistance.15,16 Even nonpathogenic staphylococci harbor drug resistance factors that can be transferred to pathogenic species. In human medicine, methicillin resistance (MR) in S. aureus has contributed to the scope of multiple drug resistance (MDR) since the early 1960s, whereas MR in staphylococci of feline origin has been recognized as a serious and widespread problem only within the past decade.
Methicillin and oxacillin are members of a class of antibacterial agents known as the semisynthetic penicillinase-resistant penicillins (SSPRP). Because of its superior stability in vitro, oxacillin is now used by most microbiology laboratories as the surrogate for testing the susceptibility of bacteria to this entire class of antibiotics.16 Even so, the term methicillin resistant has persisted in the common vernacular and in most scientific publications. The SSPRP class was developed to circumvent staphylococcal resistance to the first generation penicillins, which is mediated by bacterial production of penicillinase enzymes. Although the SSPRP class is unaffected by penicillinases, it is susceptible to an acquired penicillin-binding protein (PBP), known as PBP2a or PBP2′. This staphylococcal PBP is encoded by the mecA gene, which confers an intrinsic resistance to all beta-lactam antibiotics and their derivatives (including all classes and generations of penicillins and cephalosporins).17 Methicillin-resistant staphylococcal strains may express coresistance to any combination of other drug classes, including aminoglycosides, fluoroquinolones, macrolides, tetracyclines, fucidic acid, and mupirocin, in which case they are referred to as MDR. However, the mechanism of resistance is distinct for each of these antimicrobial classes (i.e., owing to mechanisms other than PBP).
EPIDEMIOLOGY
METHICILLIN-RESISTANT STAPHYLOCOCCUS AUREUS (MRSA)
Since the early 1960s the incidence of MR has escalated within human hospital strains of S. aureus, and hospital-acquired methicillin-resistant S. aureus (HA-MRSA) now has become the most prevalent pathogen causing nosocomial infections of people throughout the world.18 More recently MRSA strains that cause skin and soft tissue infections in people with no known nosocomial risk factors have arisen de novo within the community.18,19 The proliferation of these community-acquired (CA-MRSA) strains has been global, and although variations exist according to geographic area and ethnic populations, at least half of the persons colonized with MRSA in the United States now carry CA-MRSA strains.20
The population biology of MRSA has been studied extensively, and it has been determined that several international epidemic clones exist, resulting from the horizontal transmission of the mecA gene, which encodes methicillin resistance.20 The mecA gene is a small part of a much larger genetic element that is integrated precisely into the S. aureus chromosome. This staphylococcal chromosome cassette mec (SCCmec) varies in size but always contains the mecA gene. Five prototypic forms of SCCmec (I-V) have now been defined, and although the same types often are associated with divergent clonal lineages, particular clones have been shown to be associated with single SCCmec elements.21 SCCmec types I, II, and III typically are associated with HA-MRSA strains, whereas SCCmec types IV and V typically are identified in CA-MRSA cases.22 Recently CA-MRSA strains have been displacing the “traditional” nosocomial strains within some hospitals, and HA-MRSA strains have migrated into community settings.18 Therefore the term health-care associated (HCA)–MRSA may be used to describe hospital strains that have spread into the community and caused subsequent community-onset infections. These may be mistaken for “true” CA-MRSA based on patient demographics and inadequate molecular and epidemiological data.23
Although the prevalence of nasal colonization by MRSA in the U.S. population has been estimated most recently to be 1.5 per cent,23a the proportion of S. aureus infections that are methicillin resistant has been reported to be as high as 72 per cent among local community-onset cases.20 Risk factors associated with transmission of CA-MRSA include crowded living conditions and shared bathing facilities (e.g., military ships, prisons, day cares, sports teams, residential facilities). Risk factors for HA-MRSA transmission include immunosuppressive conditions, invasive medical instrumentation, surgery, and hospitalization. These epidemiological definitions do not appear to be globally applicable to animal MRSA cases (see the following).
Although rates of human MRSA infection have been monitored closely for several decades, the prevalence of MRSA infections in domestic animals is difficult to estimate. A retrospective analysis of pets presented to the author’s institution during the 24-month period of January 1, 2003 through December 30, 2004, showed that MRSA infection occurred with equal prevalence (one case per 1000 admissions) in dogs and cats.24 Overall the proportion of feline S. aureus infection isolates exhibiting MR was 28 per cent. However, it should be noted that the prevalence of resistance reported in this study is likely to be higher than in more general patient populations, as a result of referral bias and case selection bias (for bacterial culture) by specialist clinicians.
METHICILLIN-RESISTANT S. INTERMEDIUS (MRSI)/S. PSEUDINTERMEDIUS (MRSP)
In recent years, MRSI has emerged as a clinically important pathogen that causes treatment-resistant infections of dogs and cats.24,25 Like MRSA strains, phenotypic resistance of S. intermedius isolates to methicillin has been shown to be mediated by penicillin-binding protein 2a, which is encoded by a mecA gene.26 Also, as observed with HA-MRSA strains (see the following), most MRSI isolates coexpress resistance to other classes of antimicrobials, such as the fluoroquinolones, macrolides, tetracyclines, and aminoglycosides.24,25 For the first-generation fluoroquinolones in particular, the disparity in resistance between MRSI and methicillin-susceptible S. intermedius (MSSI) is striking, where only 55 to 57 per cent of MRSI isolates were susceptible to enrofloxacin and marbofloxacin respectively, whereas 98.5 per cent of MSSI isolates maintained susceptibility to both.24
Because human MRSA isolates have a clonal population structure and global dissemination has occurred, it has been hypothesized that MRSI isolates also would be highly clonal. A recent study of the population genetic structure of S. intermedius isolates by multilocus sequence typing showed that all previously identified MRSI were actually of the S. pseudintermedius phylotype, and were highly clonal.10 These isolates had been obtained from several countries, including the United States, Canada, Japan, United Kingdom, and several other European Union nations. Additionally, sequencing of the mecA gene revealed a high degree of homology (95 to 100 per cent) with the mecA gene of S. aureus, suggesting horizontal transfer of the gene. The structure of the MR-S. pseudintermedius phylogenetic tree suggests that the mecA gene has been received by this staphylococcal species on multiple occasions on several different continents.10
PATHOGENESIS
In human beings most CA-MRSA strains have maintained antimicrobial susceptibility profiles comparable to methicillin-sensitive S. aureus (MSSA), with the obvious exception of resistance to beta lactams. However, HA-MRSA strains commonly express resistance to several other classes of antimicrobial agents, which likely has resulted from selective pressure exerted by antimicrobial use in health care facilities.18,26 The broad antimicrobial resistance patterns inherent to HA-MRSA contribute significantly to the morbidity and mortality associated with human nosocomial MRSA infection. Meta-analyses of S. aureus bacteremia and surgical site infections in people have demonstrated that methicillin resistance is associated independently with increased length of hospital stay, hospital charges, and mortality, compared with MSSA infections.27 Although CA-MRSA strains generally exhibit less multidrug resistance than health care–associated (HA-MRSA) strains, some CA-MRSA strains express virulence factors, such as necrolytic toxins and super antigen production, that HA-MRSA do not.22 These virulence factors increase their pathogenicity in otherwise healthy persons, and may produce characteristic clinical signs, such as cutaneous abscesses and necrotizing hemorrhagic pneumonia.
An association between MRSA infection, subcutaneous abscesses, and lymphadenitis has been reported in a series of feline surgical biopsy specimens.28 Ozaki and colleagues reported a histological pattern that included central abscesses with colonies of bacterial cocci, surrounded by collagenous granulation tissue and prominent eosinophilic inflammation. In 15 of 17 cases tested, gram-positive cocci were positive for PBP2′ using an immunohistochemical technique. A retrospective study by the author and colleagues hypothesized that MRSA infection would affect a different population of cats, produce more severe clinical signs, and carry a less favorable prognosis than MSSA infection.29 However, the results did not indicate a discernible difference in signalment, clinical signs, or outcomes between the two groups, despite all isolates being SCCmec Type II (HA-MRSA) strains. The authors also documented three cases with lymph node pathology similar to that reported by Ozaki, but all three were associated with culture-confirmed MSSA infections, suggesting that this inflammatory pattern may represent an unusual response to S. aureus in general, rather than a specific response to MRSA. Caution is advised in interpreting these data because of the relatively low number of MRSA cases and lack of complete records for several animals.