In microbiological surveillance programs, coagulase-negative staphylococci belong together with enterobacteria and S. aureus to the most frequently isolated pathogens of BSI and to the leading microorganisms causing healthcareassociated BSI with 30% to 37% of all positive blood cultures being obtained in healthcare settings (18, 19, 20 and 21). While in outpatients the isolation of coagulase-negative staphylococci is rarely of clinical significance, coagulase-negative staphylococci have emerged in the healthcare-associated setting as a major cause of healthcare-associated bacteremia, especially in immunocompromised patients who have indwelling or implanted medical devices. As coagulase-negative staphylococci and in particular S. epidermidis belong to the normal microflora of the skin, they are also often found as contaminants of blood culture specimens. The contamination rate of positive blood cultures is approximately 2% to 3%, and most contaminations are indeed due to coagulase-negative staphylococci (22). To determine whether a coagulase-negative staphylococcal isolate represents a true pathogen causing bacteremia or contamination is often difficult and has also financial impacts. Thus, misdiagnosis of bacteremia due to contaminated blood cultures and subsequent unnecessary treatment of patients were shown to prompt estimated additional costs of $1,000 per patient (23). To date, there is no single criterion with sufficient specificity to predict true bacteremia. Instead, several parameters such as more than one positive blood culture bottle, capability of coagulase-negative staphylococci to produce biofilm, time to positivity of samples, laboratory markers, and clinical signs of septicemia were proposed as predictive markers (24, 25, 26 and 27). In this regard, fever or other signs of infection in conjunction with detection of coagulase-negative staphylococci in a blood culture sample are of notoriously unreliable discriminating value as presence of the bacteria might both be due to contamination and infection. Therefore, the time to sample positivity and the number of positive blood culture bottles represent better criteria to determine clinically significant BSI. Although some conflicting data have been obtained suggesting that the predictive value of the number of positive blood culture bottles is low (28), most clinical studies use this criterion to discriminate between infection and contamination. Also, the time to positivity of blood culture samples might be helpful in this respect. Thus, it was found that a medium time to positivity of more than 22 hours had a positive predictive value of 87% for diagnosing a contamination (24). This approach is particularly useful in neonates and preterm infants where it is not feasible to obtain more than one blood culture bottle for culturing. In general, blood cultures should be taken using aseptic techniques and, if a central line is present, paired cultures through the central line and through a peripheral vein should be taken (see below). The site and time of sampling should be recorded. Patients with clinical signs of infection, with multiple positive blood cultures and growth within <24 hours have a high likelihood of true bacteremia due to coagulase-negative staphylococci.

CRBSI are by far the most common complication in immunocompromised, hospitalized patients. In the United States, each year more than 150 million intravascular devices are purchased and about 80,000 CRBSIs occur. Recently, guidelines for the diagnosis and management of these infections have been revised and published (29). The sources of CRBSI are usually the insertion site, the hub, or both and coagulase-negative staphylococci are the leading pathogens causing CRBSIs (30). Diagnosis of true bacteremia versus contamination in the presence of a central line is extremely difficult and requires a thorough workup. A mathematical model was established to calculate the predictive values for true bacteremia in blood cultures positive for coagulase-negative staphylococci in patients with central venous lines (26). The positive predictive value is 98% if both blood cultures were obtained through a peripheral vein, and 96% if one sample is obtained through the catheter and the other by a vein, and only 50% if both bottles were sampled through the catheter. Thus, in the presence of a central line, it is important to draw blood culture through a peripheral vein (optimal 2 × 2, but at least 1 × 2 bottles) in order to diagnose a true BSI. To distinguish between intravascular catheter-associated bacteremia and bacteremia from other sources, the time to positivity from bottles drawn through the catheter and through a peripheral vein can be used. Nowadays, most clinicians take advantage of automated blood cultures systems and use a 2-hour cut-off differential time to positivity of blood cultures drawn from the periphery and the catheter to diagnose CRBSI. The catheter is regarded as the source of positive blood cultures when the catheter blood is positive two or more hours before the peripheral blood (31). If the catheter is removed, a 5-cm segment of the tip should be sent for culture. Growth of more than 15 colony-forming units (CFU) from the catheter tip by semiquantitative (roll plate) culture or growth of more than 102 cfu from a catheter by quantitative (sonication) broth culture suggests catheter colonization. When there are positive blood cultures with a 2-hour differential time to positivity between the central line sample and the peripheral blood culture, the diagnosis of CRBSI can be established.

Coagulase-negative staphylococci are more often found as causative microorganisms in prosthetic valve endocarditis (PVE) than in native valve endocarditis (NVE). However, the number of NVE due to coagulase-negative staphylococci is currently rising, as shown in a recent multicenter study where 8% of NVE were found to be due to these bacteria (32). Patients with a history of healthcare contact are particularly at risk. Coagulase-negative staphylococcal endocarditis on native valves is found in patients receiving long-term hemodialysis or with pacemakers and/or implantable defibrillators, with long-term catheters or with
a history of a recent surgical procedure. Mortality is high (25%), which is probably also associated with the underlying chronic diseases in these patients (32). In PVE, coagulasenegative staphylococci are isolated as causative pathogens in 15% to 40% of the cases. Infections are usually healthcareassociated and occur within 12 months of valve replacement (33, 34 and 35). Patients present with prolonged symptoms (>1 month) of weakness and low-grade fever. The modified Duke criteria are applied to diagnose infective endocarditis (36), which include, among other factors, positive blood cultures and typical echocardiographic findings.

Infections of cardiac electrophysiological devices (pacemakers, defibrillators) occur in up to 4% (reviewed in (37)) and are in 50% to 60% due to coagulase-negative staphylococci (38,39). Patients present often with pocket site inflammation, mostly within 1 month of insertion but also delays of up to 2 years. Systemic symptoms may be absent. Diagnosis is based on tissue cultures of the pocket, cultures of the devices and multiple blood cultures. In many cases, successful treatment is only possible when the infected device is completely removed and a new device is inserted at a new site.

Deep sternal wound infection (DSWI) is an infrequent but severe complication after cardiac surgery with reported incidence rates between 1% and 2% and mortality rates between 10% and 20% (40, 41, 42, 43, 44 and 45). Clinical manifestations of DSWI are variable. Wound discharge, pain, tenderness, and sternal instability are the most common local signs, whereas fever, sepsis, and elevation of inflammatory parameters are less frequently reported. Most common causative microorganisms are coagulase-negative staphylococci and S. aureus, followed by gram-negative bacteria and fungi (46, 47 and 48). In a recently published study, the causative microorganism of DSWI was identified in 86% of superficial swabs, in 94% of deep swabs, and in 88% of sternal biopsies performed before empirical antibiotic treatment was started. In 60 patients from whom results both of superficial and deep swabs were available, agreement between both specimens was observed in 77% of patients with S. aureus and in 68% of those in which coagulase-negative staphylococci were detected (49).

Prosthetic joint replacement is increasingly used to relieve pain of osteoarthritis and to improve mobility. The average infection rate of joint prosthesis is about 2% (reviewed in (37)). Risk factors for infections are previous joint surgery, a perioperative wound complication, and rheumatoid arthritis, the latter being associated with infection rates of nearly 4% (50). The most common microorganisms are coagulase-negative staphylococci, mostly S. epidermidis (30%-34%) and S. aureus (12%-23%), followed by mixed bacterial infections (10%), streptococci, and other microorganisms (51). Infections associated with prosthetic joint replacement are classified as early (<3 months after surgery), delayed (3-24 months after surgery), or late onset (>24 months after surgery) (52,53). Early- and delayed-onset infections are usually acquired during implantation. In contrast, late-onset infections are mainly due to hematogenous spread of virulent strains such as S. aureus, streptococci, or gram-negative bacilli and are rarely caused by coagulase-negative staphylococci. Infections due to coagulase-negative staphylococci can occur early or delayed and manifest with more subtle signs and symptoms. Early infections often present with a history of wound healing complications, purulent secretion at the incision site, and slightly elevated laboratory parameters of inflammation may occur. Delayed infection up to 24 months after surgery may present with persistent joint pain and/or signs of implant loosening, which may be difficult to distinguish from aseptic loosening or with a sinus tract (51,54,55). The diagnosis of prosthetic joint infection is not uniformly established. Detection of coagulase-negative staphylococci may represent either contamination or a true pathogen. Growth of the same microorganism in two or more cultures of synovial fluid or periprosthetic tissues, short time to positivity, a positive Gram stain, the presence of inflammation on histopathological examination, or presence of a sinus tract may help to diagnose an infection (53). Moreover, the recovery of microorganisms can be optimized by sonication of the prosthesis at the time of removal (56) and help to verify a true infection (see Chapter 65).

In general, all types of biomaterial or medical devices inserted across the skin or mucous membranes can become colonized and, thereafter, infected by coagulase-negative staphylococci. Thus, meningitis and encephalitis are the most serious complications associated with cerebrospinal fluid shunt implantation (57, 58 and 59). Other device-related infections that are often caused by coagulase-negative staphylococci and specifically S. epidermidis are peritoneal dialysis catheter-associated infections, infections of genitourinary prostheses, and infections of breast implants (60). Since it is not rare nowadays that a patient has, at the same time, for example, a pacemaker, a hip prosthesis, and a vascular graft, it is anticipated that we will observe in the near future a significant increase in device-related coagulasenegative staphylococcal infections.

Microorganisms of the genus Staphylococcus are nonmotile gram-positive, spheroid bacteria that appear in irregular clusters. They are catalase-positive, lysostaphin susceptible, and have a G + C content ranging between 30% and 40%. Except for a few species that grow exclusively under anaerobic conditions, staphylococci are facultative anaerobes, capable of both aerobic and fermentative metabolism (61). The overall cell wall structure of staphylococci corresponds in general to that of other gram-positive bacteria with some notable characteristics. Thus, in staphylococci, the short peptides that crosslink the heteropolymer glycan chains of the murine, contain glycine residues, which are the targets of the endopeptidase lysostaphin (62). Lysostaphin disrupts the staphylococcal cell wall specifically and the enzyme can therefore be used to distinguish staphylococci from other gram-positive cocci such as micrococci and streptococci (61). The cell wall of many
staphylococcal species is resistant to lysis by lysozyme that normally attacks the β-1,4-glycosidic bonds in the peptidoglycan chains. Lysozyme resistance in staphylococci was found to be based on O-acetylation of the muramic acid of the peptidoglycan, specifically in human pathogens such as S. aureus, S. epidermidis, and Staphylococcus lugdunensis (63,64). Currently, the genus Staphylococcus comprises 47 species and 11 subspecies (Table 30-1). Classification of staphylococci is traditionally based on the production of coagulase, an enzyme that binds fibrinogen and mediates its conversion into fibrin, resulting eventually in blood plasma coagulation. In addition to the major coagulase-positive pathogen S. aureus, six other coagulasepositive species have been described that mainly play a role in veterinary medicine (Table 30-1) (112). Among the 41 coagulase-negative species, S. epidermidis is the most common one with a broad pathogenic potential causing a wide variety of infections. Other coagulase-negative species involved in human disease are Staphylococcus haemolyticus, Staphylococcus saprophyticus, Staphylococcus capitis, S. lugdunensis, Staphylococcus hominis, Staphylococcus warneri, and Staphylococcus schleiferii. The natural reservoir of all staphylococci is the skin and the mucosa of humans and animals where the bacteria mostly reside as harmless and benign commensals. Some species were recovered from processed food or environmental samples and are traditionally used in food industry (e.g., Staphylococcus carnosus, Staphylococcus piscifermentans). In humans, microbial ecology of staphylococcal species varies individually and also depends on the body site. Thus, S. aureus prefers to colonize the anterior nares and the nasopharynx, but only 20% of healthy adults are permanently, and another 60% are transiently colonized by S. aureus (113). In contrast, all human beings are persistently colonized by coagulasenegative staphylococcal species. S. epidermidis is the species that occurs in abundance. It colonizes preferentially the upper part of the body, including the nasopharynx, and constitutes over 50% of the resident staphylococci (114). Other coagulase-negative species have adapted to distinct ecological niches and can be recovered from specific parts of the skin (e.g., S. saprophyticus from the perineum, S. capitis from the scalp, Staphylococcus auricularis from the external ear, S. haemolyticus and S. hominis from the axilla as well as the pubic and the perineal region, respectively).

In the early days of microbiological diagnostics all non-S. aureus staphylococci were referred to as Staphylococcus albus. Later, these isolates were differentiated into S. epidermidis and S. saprophyticus, a diagnosis that was mainly based on the novobiocin resistance of the latter. All other coagulase-negative species were subsumed as S. epidermidis, as the diversity of the genus was largely unknown at that time. Only upon introduction of phenotypic typing schemes in the late 1960s and early 1970s, the situation changed and a large number of novel species were identified in the following years (75,82). However, coagulase-negative staphylococcus species definition was still reserved to specialized laboratories and was rarely performed in routine diagnostics. Nowadays, staphylococcal species determination is based on combinations of various phenotypic and genotypic tests that are widely available to most laboratories. However, depending on the clinical situation, an exact species determination is not always performed. A coagulase-negative species is only reported as such when the appropriate tests have been performed. All other isolates are referred to as “coagulase-negative staphylococci.” Species identification in the routine laboratory
initiates with a variety of phenotypic tests, first of all to differentiate between coagulase-negative staphylococci and S. aureus. Slide-agglutination tests are performed to detect S. aureus-specific clumping factor, and other cell wall-associated proteins of S. aureus. As slide agglutination occasionally provides ambiguous results, the classical coagulase test mentioned above may be performed as well. Putative coagulase-negative staphylococcal species are then further examined for their colony morphology, growth requirements, oxidative and fermentative utilization of carbohydrates, novobiocin susceptibility, and various enzymatic activities (e.g., nitrate reductase, alkaline phosphatase, urease, ornithine decarboxylase, arginine dehydrogenase, etc.) (61). A variety of commercial test kits such as the API 20 Staph and API ID32 Staph systems (bioMerieux), the Staph-Zym (Rosco) or the Vitek system (bioMerieux) are available that combine detection of most of these phenotypic properties and allow for a rapid and convenient identification in the routine diagnostic laboratory. However, phenotypic tests have an inherent weakness as they rely on the expression of the phenotypic characteristic in question, and in coagulase-negative staphylococci these properties may vary within isolates belonging to the same species (115,116,117). Thus, DNA-based genotyping methods become increasingly important for an expression-independent species identification of coagulase-negative staphylococci (118). These methods target highly conserved species-specific DNA loci and genes that are PCR-amplified and subjected to either DNA-fragment pattern comparison or DNA sequencing. PCR-based approaches include ribotyping, amplified-fragment-lengthpolymorphism (AFLP), and tRNA-locus-interspacer-lengthpolymorphism (tDNA-ILP) analyses (Table 30-2). DNA sequencing of amplified 16S rRNA DNA loci is the most common method for species identification across many bacterial genera (122). Due to the close relatedness of some coagulase-negative staphylococcal species, the method may not have, in all cases, sufficient discriminatory power (126). Therefore, DNA sequencing of a range of housekeeping genes has been implemented to complement 16S rRNA locus analysis (see Table 30-2 for details and references). In general, genotyping methods are considered to be superior to mere phenotyping for coagulase-negative Staphylococcus species definition. With the increasing availability and cost effectiveness of molecular techniques in diagnostic laboratories, accurate species identification of coagulase-negative staphylococci can be anticipated to become a routine procedure. This most welcome development in microbiological diagnostics will surely shed more light on the association of specific coagulase-negative staphylococcal species with distinct infection processes.

Coagulase-negative staphylococcal isolates recovered from hospital-acquired infections are notoriously antibiotic resistant, and in many medical facilities multiresistance rates exceed those of S. aureus. Figure 30-1 exemplifies the steady increase of multiresistance rates among coagulase-negative staphylococci in Europe in the period from 1990 to 2007, and similar numbers have been reported from other countries worldwide (10, 11 and 12,132). The classical antistaphylococcal β-lactam antibiotic penicillin G, which inhibits cell wall synthesis by binding to penicillin-binding proteins, is practically no longer suitable to treat staphylococcal infections as approximately 90% of all isolates are nonsusceptible to the antibiotic. Penicillin G resistance is caused by enzymatic destruction of the antibiotic through a staphylococcal β-lactamase that is mostly plasmid encoded and now widespread among staphylococci. But resistance rates toward alternative antibiotics such as methicillin/oxacillin, gentamicin, macrolides, and fluoroquinolones are also alarmingly high (Fig. 30-1). The molecular and genetic basis of resistance against some of these compounds is briefly discussed here.

Methicillin Resistance Methicillin and oxacillin are β-lactam antibiotics that withstand the action of staphylococcal β-lactamases. They were introduced into clinical practice in the early 1960s to overcome β-lactamase-producing Staphylococcus strains. However, methicillin/oxacillin
resistance emerged very soon, and resistance rates are nowadays as high as 70% to 90% among coagulase-negative staphylococci. Methicillin resistance is mediated by the mecA gene complex that is located on a unique molecular vector called the staphylococcal chromosome cassette (SCCmec) (133). mecA encodes the additional low-affinity penicillin-binding protein PBP2a that enables cell wall synthesis in the presence of β-lactam antibiotics. SCCmec cassettes represent large chromosomal DNA fragments that may harbor, in addition to mecA, a great variety of accessory genes, for example restriction modification systems, metabolic genes, integrated plasmids, transposons, insertion sequence (IS) elements, and many more. A characteristic feature of SCCmec cassettes is the presence of recombinase genes that confer mobility and mediate the site-specific integration of the elements into a highly conserved locus of the Staphylococcus chromosome (i.e., orfX). SCCmecs have an independent evolutionary history and eight major SCCmec types have been described to date (133,134). They are considered to be transferred into S. aureus from a coagulase-negative species (135,136). However, the evolutionary origin of SSCmecs, the mechanism of SCCmec acquisition, and the factors that favor or limit their spread are still poorly understood. Interestingly, SCC cassettes cannot only carry mecA and mediate methicillin resistance. A number of SCCs has been identified that are devoid of mecA, but carry other genes instead (135,137). Therefore, SCCs are regarded as effective vectors to spread useful genes among staphylococci (138).

Aminoglycoside Resistance Aminoglycoside antibiotics inhibit protein synthesis by irreversible binding to the bacterial small ribosomal subunit (i.e., the 16S rRNA). Aminoglycoside resistance among coagulase-negative staphylococci is common (40%) and is mostly due to enzymatic inactivation of the antibiotic by acetyltransferases, adenylyltransferases, and phosphotransferases (139). Table 30-3 lists the genes and enzymes that mediate aminoglycoside resistance in staphylococci. These determinants are often located on mobile genetic elements such as plasmids and transposons. The most widespread mechanism is resistance through the bifunctional phosphotransferase-acetyltransferase enzyme AAC(6)-APH(2) that confers resistance to a broad range of aminoglycoside compounds (Table 30-3). The corresponding aacA-aphD gene is located on composite transposon Tn4001 that harbors two IS256 copies at its ends. In S. epidermidis, Tn4001 and IS256 have been shown to be associated with biofilm formation in isolates obtained from device-related healthcare-associated infections (141).

Macrolide-Lincosamide-Streptogramin Resistance Resistance toward the lincosamide clindamycin and the macrolide erythromycin among coagulase-negative staphylococci ranges between 20% and 70% and may vary by region. Although being structurally diverse, macrolide, lincosamide, and group B streptogramin antibiotics share overlapping binding sites on the large subunit of the bacterial ribosome, and are therefore often considered together. In addition to efflux pumps, inactivating enzymes, and point mutations, which are not discussed here, resistance against macrolides, lincosamides, and group B streptogramins is mainly mediated by rRNA methylases that modify the binding sites in the 23S rRNA molecule. In staphylococci, rRNA methylases are encoded by the ermA, ermB, ermC, ermF, and ermQ genes, most of them being located on plasmids and transposons (142,143). The so called MLS_B cross-resistance
phenotype depends on the expression status of the respective erm genes. In staphylococci, ermA and ermC can be expressed either constitutively or inducibly (144,145). Constitutive expression results in cross-resistance against all three antibiotic classes. In contrast, strains with inducible erm expression display in vitro resistance to 14- and 15-membered ring macrolides, which also represent inducer molecules, while retaining susceptibility toward clindamycin and group B streptogramins. Although clindamycin is not able to induce ermA or ermC expression directly, exposure of erm-inducible staphylococcal isolates to clindamycin may result in complete MLS_B cross resistance both in vitro and in vivo. This phenomenon is attributed to the selection of preexisting constitutive erm mutants (144). As clindamycin is an alternative drug for the treatment of some staphylococcal infections, detection of the inducible MLSB resistance phenotype is of clinical relevance and should be performed when required by the D-test (Fig. 30-2) (146).

Glycopeptide Resistance The glycopeptide antibiotics teicoplanin and vancomycin inhibit cell wall synthesis of gram-positive bacteria by interfering with the terminal D-alanine-D-alanine residues in the interpeptide side chains of the peptidoglycan. With the rise of multiresistance, glycopeptides gained significant importance in the treatment of infections due to gram-positive cocci in the late 1980s and 1990s. In enterococci, transmissible high-level glycopeptide resistance is a matter of concern in healthcareassociated isolates, and the resistance mechanism has been elucidated in molecular detail (147,148). Resistance in enterococci is mainly based on the modification of the glycopeptide binding site by replacement of one of the terminal D-alanine residue in the interpeptide side chain by D-lactate. The enzymes and regulators required for that process are transposon and plasmid encoded. Although being transferable from enterococci into staphylococci, high-level glycopeptide resistance through this mechanism is still rare among staphylococci (149, 150, 151 and 152). However, some staphylococcal species exhibit an intrinsically diminished susceptibility toward glycopeptides, and in S. aureus and coagulase-negative staphylococci distinct subpopulations can develop intermediate resistance upon exposure to the antibiotics (153, 154 and 155). These intermediate or heterogeneous resistance phenotypes differ from the resistance mechanism described above for enterococci and are mainly attributed to an altered cell wall synthesis turnover and thickening of the peptidoglycan (156,157). Glycopeptide resistance in coagulase-negative staphylococci was first reported in 1986 in S. haemolyticus (158). S. haemolyticus displays often less susceptibility toward teicoplanin, while vancomycin is still effective. Glycopeptide resistance
has been reported for a number of other coagulase-negative staphylococcal species as well (e.g., S. epidermidis, S. warneri), and it is therefore necessary to perform careful antibiotic resistance testing to avoid treatment failure (153,155,159,160).