Methicillin-resistant Staphylococcus aureus infections are becoming more frequent and less easily treated by means of currently recommended agents. Vancomycin has been associated with decreased susceptibility in staphylococci and with treatment failures. Daptomycin is rapidly bactericidal; a dosage of 4 mg/kg daily is approved for treatment of skin and soft-tissue infections, and a dosage of 6 mg/kg daily is approved for treatment of patients with S. aureus bacteremia and right-sided endocarditis. Findings of in vitro studies suggest a correlation between the minimum inhibitory concentrations of daptomycin and vancomycin. Clinical failure was associated with increasing minimum inhibitory concentrations in case reports and in a randomized study of persons with S. aureus bacteremia and endocarditis. Patients who did not respond to therapy had deep-seated infections that required but could not be or were not managed with adjunctive surgical therapy. No definitive resistance mechanism has been identified, although genetic mutations have been described. Clinically, prior vancomycin therapy has not been associated with failure of daptomycin therapy. Although clinical practitioners must monitor for daptomycin resistance, the available data support the use of daptomycin in the treatment of methicillin-resistant S. aureus bacteremia and endocarditis.
Methicillin-resistant Staphylococcus aureus (MRSA) infections are becoming more frequent and less easily treated with currently available agents, even with prompt diagnosis and aggressive adjunctive therapies, including debridement and removal of prostheses [1,2–3]. MRSA has emerged as an important nosocomial pathogen, accounting for ∼60% of clinical S. aureus isolates recovered in US intensive care units [4]. Infections caused by MRSA are associated with longer hospital stays, longer durations of antibiotic use, higher costs, and greater mortality rates than comparable infections caused by methicillin-susceptible S. aureus (MSSA) [5,6–7]. The emergence of S. aureus isolates intermediately susceptible to vancomycin (VISA), first isolated in Japan in 1997, and vancomycin-resistant S. aureus (VRSA), initially recovered in the United States in 2002, have spearheaded a clinical demand for new antistaphylococcal drugs [8,9,10–11]. However, we are just beginning to understand that the emergence of VISA and VRSA represent the tip of the iceberg of a more microbiologically subtle evolution, highlighted by vancomycin treatment failure in serious MRSA infections, that has occurred in staphylococci under vancomycin selection pressure.
Daptomycin is a cyclic lipopeptide in a class of antibiotics derived from the fermentation of Streptomyces roseosporus. The mechanism of daptomycin action is unique: the drug kills bacteria in a concentration-dependent manner by binding preferentially to gram-positive bacterial membranes, inserting into the membrane, and causing rapid membrane depolarization and bacterial cell death due to disruption of critical metabolic functions, such as protein, DNA, and RNA synthesis. Unlike antimicrobial agents active at the cell wall, agents that kill by means of membrane depolarization fail to lyse cells [12,13,14–15]. In vitro studies demonstrated that daptomycin had bactericidal activity equal to or greater than that of vancomycin, linezolid, and quinupristin-dalfopristin [16]. In vivo, daptomycin was effective in models of endocarditis [17]. Use of daptomycin was approved in the United States and Europe at a dosage of 4 mg/kg daily for treatment of skin and soft-tissue infections and was approved in the United States at a dosage of 6 mg/kg daily for treatment of S. aureus bacteremia and right-side infective endocarditis [18]. Clinical studies showed that daptomycin for the treatment of pneumonia failed, and it was shown that daptomycin is inactivated by surfactant [19].
Other options for treating S. aureus infections include antistaphylococcal penicillins and cefazolin (for MSSA), trimethoprim-sulfamethoxazole, clindamycin, linezolid, quinupristin-dalfopristin, tetracyclines, and tigecycline. Although many clinicians routinely use these agents for the management of soft-tissue infection, many drugs developed earlier are not approved by the US Food and Drug Administration for this purpose. Only vancomycin and linezolid are recommended for the treatment of MRSA pneumonia. Several investigational agents, including dalvabancin, televancin, and PBP-2a—targeted β-lactams (e.g., ceftobiprole), will likely become available in the near future for treatment of soft-tissue infection. The role of combination therapy, with the intent of potentiating the regimen or reducing the likelihood of resistance, remains debatable, with fewer clinical data than in vitro data.
Failure of vancomycin to treat glycopeptide-"susceptible" MRSA; concern for the development of staphylococcal resistance to daptomycin, linezolid, and other gram-positive agents; the increased cost of the newer antibiotics; and the surge in the rate of methicillin resistance among community-acquired S. aureus strains have all come together to create a situation in which there are fewer clinical data than microbiological observations. In this review, we summarize the interface of available clinical and microbiological data on daptomycin resistance and provide perspective for clinicians who treat patients with MRSA infection.
Although vancomycin was introduced in 1958, its use was not significant until the mid-1980s, when the prevalence of methicillin resistance among S. aureus isolates surged [20]. However, another decade would pass before the first VISA isolate was identified [21] and still longer before the first VRSA isolate was recovered [11]. The history of the development of S. aureus resistance to antimicrobials places vancomycin in a unique light, given that it has generally taken only 2 or 3 years for resistance to be detected after clinical introduction of an antibiotic (table 1) [22].
Several reasons for this anomaly are proposed. First, one must look at the definition of microbiologic susceptibility, which, for vancomycin, was originally set at ⩽4 µg/mL. The recent change in the Clinical and Laboratory Standards Institute breakpoint to a vancomycin MIC of ⩽2 µg/mL leads many to conclude in retrospect that the initial breakpoint was misleading because it was based on the correlation between clinical responses and MICs in a small number of patients treated soon after vancomycin was first available in clinical settings (table 2). Second, some susceptibility testing methods, particularly automated systems used by clinical laboratories, are not sensitive enough to detect all isolates with reduced susceptibility to glycopeptides, prompting the recommendation to confirm susceptibility by means of agar-based methods, such as the Etest (AB Biodisk) [23, 24]. Nevertheless, many standardized methods, including the Etest, are limited in their ability to detect heteroresistant VISA (hVISA) (table 2), perhaps because of problems with inoculum sizes. Only population analysis, the gold standard for the detection of glycopeptide heteroresistance, can detect the ⩽1 in 1,000,000 organisms in a population able to grow in media containing vancomycin concentrations greater than the susceptible range. However, this method is too cumbersome to be practical in a clinical laboratory. Third, unlike the mechanisms of resistance for most antimicrobials, reduced susceptibility to glycopeptides in S. aureus can be unstable phenotypically [8]. To date, although the thickened cell wall phenotype of S. aureus isolates with intermediate susceptibility to glycopeptides appears to be consistent across all such strains, there remains to be identified a consistent profile of genetic mutations that are present across all VISA. This phenotype is likely the end result of a common pathway that consists of multiple permutations of genetic mutation and transcriptional regulation. It is becoming clear that the expression of intermediate susceptibility to glycopeptides is the end result of altered cell wall metabolism [8, 25]. These physiologic changes appear to result in increased numbers of D-alanyl-D-alanine residues that serve as dead-end binding sites for vancomycin, leading to a reduced vancomycin diffusion coefficient, sequestration of vancomycin within the cell wall by these false targets, and prevention of vancomycin from reaching its site of action [26]. In addition, studies of S. aureus with reduced vancomycin susceptibility and of isogenic vancomycin-susceptible progenitors showed cell walls with reduced peptigodglycan cross-linking, cell wall turnover, and autolysis that investigators suggested may be due to alterations in teichoic acid structure and metabolism [27]. These metabolic changes result in considerable cell wall thickening.
The notion that the changes leading to glycopeptide resistance emerge before S. aureus demonstrates a glycopeptide resistance phenotype, as defined by the Clinical and Laboratory Standards Institute, is only now becoming fully appreciated. Jones [28] evaluated the prevalence and significance of tolerance to vancomycin by analyzing S. aureus isolates from the global SENTRY surveillance project. Of >35,000 S. aureus isolates recovered between 1997 and 2003, there was no evidence of increasing vancomycin MICs (sometimes referred to as “MIC creep”), suggesting that the prevalence of vancomycin resistance was not increasing [28]. The percentage of isolates with vancomycin MICs of >2 mg/L ranged from 0% to 0.1% per year, and no isolates had vancomycin MICs of >4 mg/L [28].
However, retrospective evaluation of vancomycin MICs of MRSA clinical isolates at individual institutions revealed subtle but significant increases. One institutional study from New York showed that, between 1994 and 1999, the percentage of isolates with vancomycin MICs of >1.6 µg/mL increased from 25% to >50% and that the percentage with vancomycin minimum bactericidal concentrations (MBCs) of >1.6 µg/mL increased from <10% to >50% [29]. Another study of isolates from patients with cancer at a large institution in Texas showed that the vancomycin MIC90 for MRSA increased from 0.25 µg/mL in 1985 to 2.0 µg/mL in 2004 [30]. Finally, a study from a large center in Boston, Massachusetts, also showed a significant increase in vancomycin MICs among MRSA bloodstream isolates between 2002 and 2005 [31].
It is clear from these reports that differences seen at individual health care centers are not being found in global surveys. This suggests that the magnitude of increases in vancomycin MICs over time may vary between centers and, therefore, that institutions where the effect is less pronounced may “dilute” the data from centers where vancomycin MIC creep is more pronounced.
A number of studies have demonstrated increased clinical failure of vancomycin therapy in patients infected with MRSA isolates for which MICs are increased yet still within the susceptible range. In one study, vancomycin was <10% successful for the treatment of bacteremia due to MRSA with vancomycin MICs of 1–2 µg/mL, compared with a success rate of 56% when the vancomycin MIC was ⩽0.5 µg/mL. In the same study, an increase in the percentage of MRSA killed by vancomycin in vitro was associated with a higher likelihood of treatment success [32]. A more recent study reported statistically significantly decreased rates of treatment success when the vancomycin MIC was 2 µg/mL, compared with <2 µg/mL [35]. These findings might be associated with the observation that the percentage of S. aureus isolates that are hVISA increases as the vancomycin MIC increases. For example, a recent survey showed that the percentage of hVISA isolates increased from 57% to 81% to 100% as the vancomycin MIC increased from 1 to 2 to 4 µg/mL, respectively [33, 34].
Furthermore, consistent with reports suggesting improved clinical outcome after treatment with bactericidal agents [36], the efficacy of vancomycin in the treatment of bacteremia was associated with increased killing in vitro, independent of the MIC [32]. Jones [28] analyzed 17 VISA strains, 88 hVISA strains, 3 VRSA strains, and 105 wild-type MRSA strains. Vancomycin tolerance was defined as a ratio of MBC to MIC of ⩾1 : 32 or a ratio of MBC to MIC of ⩾1 : 16 with a vancomycin MBC of ⩾32 mg/L. In this analysis, 15% of wild-type MRSA strains met the definition of tolerance, compared with 74% of hVISA strains and 100% of VISA and VRSA strains, providing additional support that vancomycin is suboptimal treatment for hVISA bacteremia [28].
Because reduced susceptibility to glycopeptides has been linked to suboptimal clinical response in patients with MRSA bacteremia, outcomes of treatment with newer antistaphylococcal agents might be superior to those of vancomycin for soft-tissue infection and pneumonia [37,38,39,40,41–42]. Clinical resolution of soft-tissue infection due to MRSA is more rapid with daptomycin treatment than with vancomycin treatment (5 days vs. 7 days; P < .05), which has translated into shorter antibiotic-related lengths of stay and lower hospital costs [43].
The microbiological definition of resistance to daptomycin has not been established for S. aureus; organisms with a daptomycin MIC of ⩽1 µg/mL are considered susceptible, and those with a daptomycin MIC of >1 µg/mL are considered nonsusceptible. Wild-type S. aureus with daptomycin MICs greater than the susceptible range are rare but have been recovered from patients who received vancomycin, patients who never received daptomycin, and patients who were antibiotic naive [44, 45]. Spontaneous resistance is uncommon, emerging in vitro at a rate of <1 × 10-10, but resistance can be induced by serial passage in increasing concentrations of daptomycin [46].
Mutations in mprF (which encodes lysylphosphatidylglycerol synthetase), yycG (which encodes sensor histidine kinase), and rpoB and rpoC (which encode the β and β′ subunits, respectively, of RNA polymerase) have been found in S. aureus isolates with daptomycin MICs greater than the susceptible range. Mutations in mprF appear to occur early in the selection process, whereas mutations in rpoB and rpoC occur later. Given that these mutations are not present in several daptomycin-nonsusceptible clinical S. aureus strains, more work is needed to define the mechanism by which these and perhaps other mutations lead to decreased microbiological activity of daptomycin. Further understanding of decreased daptomycin susceptibility should enhance our understanding of the mechanism of interaction between daptomycin and the bacterial cell membrane [47].
Exposure of S. aureus to vancomycin influences the global physiology of the organism, resulting in significant metabolic and physiologic alterations [8]. Such adaptations may allow for attenuated activity of other antimicrobial molecules such as platelet microbicidal peptides, endogenous molecules whose bactericidal activity is attenuated when function of the accessory gene regulator (agr) is decreased and when the organism is exposed to vancomycin in vitro and in vivo [48]. Therefore, the reduced activity of other antimicrobials, such as daptomycin, on glycopeptide-exposed S. aureus may be another consequence of these global effects.
Several reports have linked increases in vancomycin MICs to increases in daptomycin MICs [26, 49, 50], although no definitive mechanism has been elucidated. Although the mechanism is unclear, the report by Cui et al. [26] suggests that thick cell walls induced by the glycopeptide impair the diffusion of daptomycin, a molecule with a molecular weight of >1600 D, to target sites in the cell membrane. Of interest, a recent study that used whole-genome sequencing detected mutations in rpoC and yycH genes in MRSA treated with vancomycin and rifampin [51]. Mutations at these loci have been associated with daptomycin resistance, even though the patient never received daptomycin.
Knowledge of clinical resistance to daptomycin originated in case reports from diverse geographic locations, including the United States and Europe, in which clinical failure was associated with increasing daptomycin MICs [52, 53]. These patients had staphylococcal or enterococcal infections associated with septic arthritis [53, 54], osteomyelitis [53, 55, 56], septic thrombophlebitis [57], and/or endocarditis. Many infections also occurred in the presence of intravascular catheters and other prosthetic devices [52, 58, 59]. For these patients, the etiologic agents of infection were present in large numbers, requiring surgical intervention. Several other reports included patients with underlying immunosuppression [58, 60, 61].
In a randomized study of daptomycin versus standard therapy (i.e., vancomycin or semisynthetic antistaphylococcal penicillin, both administered with initial low-dose gentamicin) for S. aureus bacteremia and endocarditis, Fowler et al. [62, 63] reported on the development of increasing MICs in 7 (5.8%) of 120 patients treated with daptomycin and in 7 (13.2%) of 53 patients treated with vancomycin. This study provided a unique prospective dataset of >1200 serial clinical S. aureus isolates. Six patients in each group were assessed as failures and had S. aureus isolates with possibly elevated daptomycin or vancomycin MICs. In a neutropenic mouse model, isolates obtained before and after initiation of daptomycin therapy responded to a daptomycin concentration less than that of the clinical daptomycin exposure, which was calculated using each patient's pharmacokinetic data [64].
As discussed above, several reports have linked increasing vancomycin MICs to increasing daptomycin MICs [26, 49, 50], although no definitive mechanism for this association has been elucidated. Some reports involved isolates from highly treatment-experienced patients, and others involved groups of isolates in which the prevalence of reduced susceptibility to vancomycin was high [65].
In clinical practice, most patients receive initial therapy with vancomycin while awaiting speciation and susceptibility reports from the microbiology laboratory. To date, most patients have received daptomycin as salvage therapy following vancomycin failure [66, 67]. Similarly, in the study reported by Fowler et al. [62], most patients received vancomycin therapy for a mean of 2 days before starting either daptomycin treatment or standard therapy. In a post hoc analysis of MRSA-infected patients from their study, Rehm et al. [68] reported that a past history of vancomycin therapy did not impact the clinical outcome of daptomycin-treated patients (table 3). Although these findings involved a sample of patients treated in a contemporary randomized trial, the analysis was not prespecified and was limited by the relatively short duration of previous vancomycin exposure.
Where does the above discussion leave the clinician treating patients with MRSA infection? Vancomycin MICs are increasing slowly, and certain hospitals may observe more-pronounced increases in vancomycin MICs. In addition, automated systems may not detect elevated MICs. Vancomycin-susceptible S. aureus isolates with higher MICs (range, 1–2 µg/mL) show increased rates of vancomycin-intermediate heteroresistance and vancomycin treatment failure.
Bactericidal activity in vitro is associated with superior clinical response in the management of bacteremia. Vancomycin is bacteriostatic against hVISA and VISA isolates. Several recent reports have associated elevated vancomycin MICs with elevated daptomycin MICs. The clinical significance of this microbiological observation is not known. Daptomycin is the most potent in vitro bactericidal agent against S. aureus isolates, including hVISA and VISA [69], and was proven to be effective for treating patients with S. aureus bacteremia and right-side endocarditis, most of whom received prior vancomycin therapy. Previous vancomycin exposure alone has yet to be proven to influence daptomycin nonsusceptibility.
Daptomycin is associated with myopathy that is easily monitored by measuring serum creatinine phosphokinase levels. In the randomized trial by Fowler et al. [62] of daptomycin for treatment S. aureus bacteremia and endocarditis, 3 (2.5%) of 120 patients discontinued therapy because of increased creatinine phosphokinase levels. In addition, daptomycin was also associated with a lower rate of nephrotoxicity and was easier to administer, compared with vancomycin given with initial low-dose gentamicin [62].
Taking these facts together, for patients in whom bactericidal therapy is medically desirable, including patients with MRSA bacteremia (especially endocarditis) and perhaps patients with MRSA infections and neutropenia, daptomycin may the preferred agent.
When treating patients with S. aureus bacteremia, clinicians should obtain a precise vancomycin MIC, supplementing automated methods with quantitative methods, if necessary. Consideration should also be given to additional methods for detecting hVISA. If daptomycin is administered, daptomycin MICs should be tested for all S. aureus isolates recovered from blood or deep tissue specimens.
In patients with nonrespiratory sepsis, the choice of empirical therapy against MRSA should be based on the likelihood of MRSA bacteremia, local vancomycin susceptibility patterns, host factors, and severity of illness. Treatment may include daptomycin if there is a strong possibility, based on local microbiological data or a recent history of vancomycin treatment in the infected patient, that the isolates have a vancomycin MIC of >1 µg/mL. This recommendation is based on collective data showing poor clinical response to vancomycin in patients for whom alternative therapy may be superior.
Failure of initial therapy or clinical worsening of the patient's condition should prompt orders for follow-up cultures and MIC tests, as well as careful evaluation for metastatic foci of infection. Failure to recognize and address sources and foci of infection are common causes of clinical failure in patients infected with S. aureus; aggressive drainage and debridement are critical to successful management.
Development of antimicrobial therapy effective against S. aureus will continue to challenge researchers and clinicians. Attempts to predict the success of vancomycin treatment for MRSA infections on the basis of microbiological definitions of susceptibility have been unsuccessful. Hopefully, with recent data, laboratory definitions of susceptibility can be better aligned with clinical success in treating patients with MRSA infections.
In summary, daptomycin has been clinically available since September 2003 and has been extensively studied in the laboratory and in clinical trials, leading to US Food and Drug Administration approval for treatment of skin and soft-tissue infections and of S. aureus bacteremia and right-side endocarditis but not for treatment of pneumonia [19, 41, 62]. Mutations associated with increased daptomycin MICs have been identified, enhancing our understanding of the role genetic mechanisms in the development of resistance. In vitro studies suggest a correlation between increasing vancomycin MICs and increasing daptomycin MICs, although the clinical relevance of this relationship is unclear. Clinically, treatment failure has been associated with increasing MICs, as reported in individual case studies and a randomized study of S. aureus bacteremia and endocarditis. The long-term impact of this observation on the usefulness of daptomycin therapy cannot yet be defined. Careful attention to results of follow-up cultures and susceptibility patterns are critical in the care of patients with deep-seated S. aureus infections. Much remains unknown, but there is clinical and microbiological evidence for alternative options to vancomycin in the management of MRSA bacteremia. Although microbiologists and clinicians must monitor isolates for resistance to daptomycin, the available data support its use in the treatment of MRSA bacteremia and endocarditis, both as initial and salvage therapy.
Potential conflicts of interest. H.W.B. serves as an advisor and/or consultant to Cubist, Johnson & Johnson, Pfizer, Schering-Plough, and Targanta; as a speaker for Cubist, Pfizer, and Schering-Plough; and owns or has owned shares of Pfizer and Cubist. G.S. is on the speaker bureau of Cubist, Pfizer, Merck, and Wyeth Pharmaceuticals; has received research grant support from Cubist Pharmaceuticals; and has served as advisor and/or consultant to Cubist and Pfizer.
IDSA Members: For your free access to this journal, log in via the IDSA members area.
Open access options for authors visit Oxford Open
This journal enables compliance with the NIH Public Access Policy