The Role of Vancomycin in the Treatment Paradigm

Clinical Infectious Diseases, Jan 2006

Vancomycin was introduced in the United States in 1956 as a possible treatment for infections due to penicillin-resistant Staphylococcus aureus, but it was not used widely because of toxicity and the nearly simultaneous development of semisynthetic antibiotics and cephalosporins. Thus, its main indication was the treatment of serious gram-positive infections in penicillin-allergic patients. For susceptible strains of S. aureus, vancomycin was more rapidly bactericidal than penicillin, nafcillin, or cefazolin, and, in a rabbit model of S. aureus endocarditis, sterilization of vegetations was more rapid with vancomycin. In clinical practice, however, nafcillin remained the treatment of choice for staphylococcal bacteremia, largely because it had failure rates of only 4%. With the appearance of methicillin-resistant S. aureus and coagulase-negative staphylococci, vancomycin became the drug of choice for these infections. Recently, the efficacy of vancomycin has been questioned because of vancomycin's increasing minimum inhibitory concentrations among staphylococci, poor tissue penetration, and apparently slower bacterial killing than previously was recognized.

A PDF file should load here. If you do not see its contents the file may be temporarily unavailable at the journal website or you do not have a PDF plug-in installed and enabled in your browser.

Alternatively, you can download the file locally and open with any standalone PDF reader:

http://cid.oxfordjournals.org/content/42/Supplement_1/S51.full.pdf

The Role of Vancomycin in the Treatment Paradigm

0 Veterans Affairs Medical Center , 500 W. Fort St., Boise, Idaho 83702 (dlsteven @mindspring.com). Clinical Infectious Diseases 2006 ; 42:S51-7 2005 by the Infectious Diseases Society of America. All rights reserved. 1058-4838/2006/4201S1-0008$15.00 1 Dennis L. Stevens Infectious Disease Section, Veterans Affairs Medical Center , Boise, Idaho Vancomycin was introduced in the United States in 1956 as a possible treatment for infections due to penicillinresistant Staphylococcus aureus, but it was not used widely because of toxicity and the nearly simultaneous development of semisynthetic antibiotics and cephalosporins. Thus, its main indication was the treatment of serious gram-positive infections in penicillin-allergic patients. For susceptible strains of S. aureus, vancomycin was more rapidly bactericidal than penicillin, nafcillin, or cefazolin, and, in a rabbit model of S. aureus endocarditis, sterilization of vegetations was more rapid with vancomycin. In clinical practice, however, nafcillin remained the treatment of choice for staphylococcal bacteremia, largely because it had failure rates of only 4%. With the appearance of methicillin-resistant S. aureus and coagulase-negative staphylococci, vancomycin became the drug of choice for these infections. Recently, the efficacy of vancomycin has been questioned because of vancomycin's increasing minimum inhibitory concentrations among staphylococci, poor tissue penetration, and apparently slower bacterial killing than previously was recognized. - FAILURE OF VANCOMYCIN FOR THE TREATMENT OF STAPHYLOCOCCAL BACTEREMIA AND ENDOCARDITIS For susceptible strains of Staphylococcus aureus, vancomycin was more rapidly bactericidal than were penicillin, nafcillin, or cefazolin [1], and, in a rabbit model of S. aureus endocarditis, sterilization of vegetations was more rapid with vancomycin [1]. In clinical practice, however, nafcillin remained the treatment of choice for staphylococcal bacteremia, largely because it had failure rates of only 4% [2]. In 1977, Esposito and Gleckman [3] reported cure rates of 63% but partial responses or treatment failures in 37% of patients treated with vancomycin for staphylococcal endocarditis. In 1981, Geraci and Wilson [4] reported a 42% failure rate in patients treated with vancomycin for endocarditis due to methicillin-susceptible S. aureus (MSSA). Similarly, in 1990, Small and Chambers [5] reported failure rates of 38% among patients treated with vancomycin for MSSA endocarditis, compared with failure rates of only 1.4% for patients receiving nafcillin. Finally, in the treatment of endocarditis caused by MSSA, Gentry et al. [6] demonstrated complete responses in 74% of patients treated with nafcillin, compared with 50% of patients receiving vancomycin. The duration of bacteremia was 2 days in the nafcillin group and 5 days in the vancomycin group. Among patients with MSSA bacteremia, the failure rate was also higher in patients receiving vancomycin (20%) than in patients receiving nafcillin (4%) [7]. Bacteremia persisted for 13 days in 20% of patients and for 17 days in 12% of patients receiving vancomycin [7]. In patients treated with vancomycin for methicillin-resistant S. aureus (MRSA) endocarditis, the duration of bacteremia was even more prolonged (mean, 7 days) [5]. IN VITRO SUSCEPTIBILITY, MINIMUM BACTERICIDAL ACTIVITY, AND BACTERIAL KILLING Although susceptibility does not guarantee success, resistance certainly predicts failure, and several factors contribute to the failure of vancomycin in treating staphylococcal bacteremia and endocarditis. First, as described above, in patients with endocarditis due to MSSA infection, the rate of bacterial clearance is slower for vancomycin than for nafcillin. Recently, slower bacterial killing has also been demonstrated in vitro. For example, the rate of in vitro bacterial killing of MSSA isolates was comparatively lower for vancomycin than for nafcillin (figure 1) [5], a finding that differs from those of studies performed with strains obtained during the 1960s and 1970s [1]. Second, most strains remain susceptible to vancomycin; however, minimum bactericidal concentrations (MBCs) have been increasing over the course of several decades. The MBC and its relationship to the MIC determine whether an agent is bactericidal or bacteriostatic, and an antibiotic with an MBC: MIC ratio of 1 against a specific pathogen is defined as bactericidal for that agent. Generally, agents that are bacteriostatic have MBC:MIC ratios of 48. For antibiotics that are generally considered to be bactericidal against such pathogens as staphylococcus (e.g., b-lactams), an MBC:MIC ratio of 132 is defined as denoting tolerance. Remarkably, MBC:MIC ratios of 416 have been reported for vancomycin [8, 9]. The emergence of glycopeptide-intermediate S. aureus (GISA), vancomycinintermediate S. aureus (VISA), and heterogenous GISA and VISA isolates among patients with serious infections has become problematic, as suggested by recent reports [10, 11]. Heterogenous VISA infections are associated with high rates of treatment failure, prolonged bacteremia, high bacterial loads, and relatively lower vancomycin trough levels (table 1) [12]. These strains behave as if they are tolerant to vancomycin. In the United States, GISA may be related to accessory gene regulator (agr) group II polymorphism in MRSA [13]. Thus, as MICs of vancomycin approach 2 mg/mL, vancomycin treatment failure is more common (figure 2) [8], and the mechanism could be either agr group II polymorphism [13] or heterogenous GISA [12] or both. Finally, since the late 1990s, vancomycin-resistant strains (e.g., SAVR [S. aureus vancomycin resistant]) have been described in patients with previous exposure to vancomycin (e.g., patients undergoing renal dialysis) [14, 15]. Tigecycline, a substitute minocycline compound, and daptomycin demonstrated good in vitro and in vivo activity against GISA strains [16]. In addition, at least one patient with VISAassociated endocarditis responded successfully to linezolid [10]. ANTIBIOTIC SERUM CONCENTRATIONS AND PROTEIN BINDING DURING THERAPY FOR INTRAVASCULAR INFECTION For both intravascular and extravascular infections, antimicrobial efficacy is clearly determined by the concentration of free drug in serum. For example, peak blood concentrations of 25, 18, and 77 mg/mL may be obtained for vancomycin, linezolid, and daptomycin, which are 50%, 30%, and 90% protein bound, respectively. This would yield peak levels of 1215, 78, and 78 mg/mL free drug for these 3 agents, respectively (figure 3). If the organism is S. aureus with an MIC of 0.52 mg/mL, then free drug concentrations in serum would exceed the MIC for this putative MRSA strain by a factor of at least 4, and suppression of growth would occur for each drug. When the serum level of free drug decreases to !2 mg/mL, then a bactericidal effect may be lost. If serum trough levels of 10 mg/mL are maintained for vancomycin, then the free drug concentration remains greater than the MIC for the entire dosing interval. When inadequate dosing results in trough levels of !10 mg/mL or when the MIC for an MRSA strain is 12 mg/mL, free drug Table 1. Clinical features of bacteremia caused by heterogenous vancomycin-intermediate Staphylococcus aureus (hVISA) and vancomycin-susceptible, methicillin-resistant S. aureus (vsMRSA). Clinical feature Treatment failure, no. of episodes/ total cases of bacteremia assessed (%) Duration of bacteremia, mean, days Low serum trough levels of vancomycin,a no. of episodes/total cases of bacteremia assessed (%) Bacteremia due to hVISA Bacteremia due to vsMRSA 5/5 (100) 39 NOTE. Adapted from Charles et al. [12]. a Low serum trough levels are defined as serum vancomycin trough concentrations of !10 mg/mL during the first 7 days of therapy after the onset of bacteremia. The denominator is the no. of cases in each group treated with vancomycin for 11 week and for which trough levels were assessed. Figure 2. Relationship of MIC to vancomycin treatment failure in patients with methicillin-resistant Staphylococcus aureus infections. Data points denote the percentage of patients for whom treatment failed. Adapted from Moise-Broder et al. [13]. concentrations in serum might not exceed the MIC for that particular strain. Thus, careful attention to monitoring blood levels of vancomycin is crucial in the treatment of aggressive S. aureus infections, such as endocarditis. This may explain, in part, the slower rate of bacterial eradication and the higher failure rate associated with vancomycin, compared with those for such drugs as nafcillin, in the treatment of S. aureus endocarditis [5, 12]. CURRENT TREATMENT OF MRSA BACTEREMIA AND ENDOCARDITIS At present, the optimal treatment for endocarditis caused by MRSA does not exist. Although the Food and Drug Administration has not approved a drug with proven superiority to vancomycin, this does not mean that vancomycin is the best treatment. Newer drugs need to be developed, and approved drugs need to be studied in the treatment of endocarditis. One such study, which excluded patients with endocarditis, documented a clinical cure rate for MRSA bacteremia of 50% for vancomycin and 56% for linezolid [17]. Daptomycin is currently being studied for the treatment of endocarditis caused by MRSA because of its rapid bactericidal effects [18], long postantibiotic effect [19], and relatively high plasma levels of drug [18, 19]. This drug has been approved for the treatment of skin and skin-structure infections at a dose of 4 mg/kg given twice daily [20]. At this dose, no apparent increase in creatine phosphokinase or clinical myopathy was realized. An endocarditis clinical trial comparing vancomycin with daptomycin at a dose of 6 mg/kg is currently under way. On the basis of bactericidal data, the efficacy of daptomycin should be greater than that of vancomycin against MRSA endocarditis; however, it will be critically important to monitor myopathic symptoms clinically and chemically. COMBINATION THERAPIES In an attempt to improve the bactericidal capacity in patients treated with vancomycin or nafcillin for staphylococcal endocarditis, combination treatment has been advocated. Combinations of gentamicin and nafcillin showed increased efficacy in experimental models of staphylococcal endocarditis [21] and resulted in more-rapid clearance of MSSA from the blood of patients with right-side endocarditis (2.6 vs. 3.6 days) [5]. Combinations of vancomycin and gentamicin resulted in morerapid clearing of organisms in animal models [21], in vitro [22], and in some in vitro models of endocardial vegetations [23] but not others [24]. In contrast, treatment with combinations of vancomycin and rifampin for humans with MRSA endocarditis prolonged the duration of bacteremia (7 days for vancomycin vs. 9 days for vancomycin plus rifampin) [25]. In vitro antagonism has also been demonstrated when rifampin is added to vancomycin [22]. Anecdotally, a single patient with MRSA endocarditis who experienced vancomycin therapy failure responded to combination therapy with linezolid and amikacin [10]. STAPHYLOCOCCAL PNEUMONIA: CONCERNS FOR THE FUTURE The clinical spectrum of MRSA pneumonia had been largely nosocomial and associated with the use of ventilators. Although 2 studies reported comparable cure rates for vancomycin and linezolid in the treatment of nosocomial MRSA pneumonia (52%65% [17] vs. 54%78% [26]), a retrospective meta-analysis demonstrated markedly greater efficacy in groups receiving linezolid, compared with vancomycin, for ventilator-associated MRSA pneumonia (62% vs. 21%, respectively [27]). Additional prospective, randomized studies are needed to determine the optimal treatment of MRSA pneumonia, largely because the prevalence of MRSA is increasing in most regions of the world. Recently, community-acquired MRSA pneumonia with high mortality has been reported in the United States. In fact, community-acquired MRSA pneumonia will likely become increasingly problematic for physicians for 3 reasons. First, there is an increasing prevalence of community-acquired MRSA strains bearing the SCCmec IV genotype. Second, hospital strains containing the larger SCCmec I, II, and III cassettes containing antibiotic resistance genes will mix with community-acquired strains bearing SCCmec IV and genes for PantonValentine leukocidin, resulting in greater resistance among community-acquired strains and potentially greater virulence in hospital-acquired strains. Third, secondary staphylococcal pneumonia among patients with influenza is associated with high mortality, and there is a great likelihood of an increasing incidence of influenza due to vaccine shortage or human-tohuman spread of the potentially pandemic H5N1 avian influenza virus. Thus, in 2006, nosocomial and community-acquired postinfluenza pneumonia caused by MRSA may become more common, more severe (by virtue of Panton-Valentine leukocidin production [28]), and more resistant to conventional agents. Of great concern is the likelihood that contemporary guidelines for empirical treatment of community-acquired pneumonia will be ineffective for treating MRSA pneumonia. Thus, delays in effective anti-MRSA treatment may be associated with poorer outcomes. Although it is presumptuous to recommend changes in empirical treatment of community-acquired pneumonia, clinicians should be cognizant of the prevalence of MRSA in their own locales and should vigorously attempt to make an early and definitive bacteriological diagnosis. DRUG LEVELS IN TISSUE AND DRUG-PROTEIN BINDING IN ANTIBIOTIC TREATMENT OF MRSA PNEUMONIA If analysis of an ongoing clinical trial verifies the previously reported meta-analysis conclusion [27] that linezolid is superior to vancomycin for the treatment of MRSA pneumonia, there may be several potential reasons for this discrepancy, including differences in tissue penetration, the degree of protein binding, and inadequate dosing of vancomycin for infections of the lung. For example, Lamer et al. [29] measured concentrations of vancomycin in endothelial lining fluid (ELF) and plasma and found vancomycin concentrations of !3.0 mg/mL in the ELF of patients who had plasma concentrations of 10 mg/mL and concentrations of 4.5 mg/mL in the ELF of patients who had plasma concentrations of 25 mg/mL. Similarly, after intravenous administration of a 1-g dose, Cruciani et al. [30] found ELF S54 CID 2006:42 (Suppl 1) Stevens vancomycin concentrations of 4.0 mg/mL at 4 h, 2.0 mg/mL at 6 h, and 2.5 mg/mL at 12 h (figure 4). Note that these authors excluded 1 of 6 patients at the 6-h time point and 3 of 7 patients at the 10-h time point; they also excluded 3 of 7 patients at the 12-h time point because their vancomycin levels were zero. Thus, in reality, their mean vancomycin concentrations were even lower than reported [30]. Although plasma vancomycin levels of 2 mg/mL would be sufficient for MRSA strains with MICs of 0.5 mg/mL, problems could occur with MRSA strains with higher MICs or with GISA or VISA strains. If one considers that the rate of protein binding is 50% for vancomycin, then peak free drug concentrations in ELF are marginal at best (figure 4). In patients with S. aureus pneumonia, bacterial eradication rates were lower for vancomycin than for other antistaphylococcal agents [31]. Among patients receiving vancomycin, clinical and bacteriological responses were best when the 24-h area under the curve value divided by the MIC was 1400 mg/mL [31]. In a second study of staphylococcal pneumonia that involved 70 patients, including 35 with MRSA infection, the success rate of vancomycin was 76% for patients with area under the inhibition curve values of 1345 mg/mL and 22% for patients with area under the inhibition curve values of !345 mg/mL [32]. These data suggest that current vancomycin dosing recommendations and nomograms may be inadequate for patients with MRSA pneumonia, particularly those with infection caused by strains with vancomycin MICs of 14 to 8 mg/mL [32]. Although specific recommendations cannot be made with certainty, it may be necessary to maintain serum trough levels of 115 mg/mL or even 20 mg/mL. Some researchers advocate continuous infusion of vancomycin for ventilator-associated MRSA pneumonia according to the following protocol: 15 mg/ kg for 1 h, followed by 30 mg/kg infused for 24 h with a volumetric infusion pump [3335]. The dose is adjusted to maintain a plateau serum vancomycin concentration of 2030 mg/mL [33, 34]. Linezolid is slowly bactericidal for streptococcus but, like vancomycin, is slowly bacteriostatic for S. aureus and has a moderate postantibiotic effect of 34 h [36]. Unlike vancomycin, linezolid penetrates into ELF bronchoalveolar lavage fluid impressively. Thus, although plasma concentrations following an intravenous or oral dose are 1520 mg/L, bronchoalveolar lavage concentrations exceed 64 mg/mL [37]. Because linezolid is 30% protein bound (figure 4), this provides free drug concentrations of 140 mg/mL in ELF (figure 3) [38]. Thus, there is a distinct advantage for linezolid, compared with vancomycin, in terms of pharmacokinetics. Ongoing clinical trials comparing vancomycin and linezolid for MRSA pneumonia will provide a definitive answer regarding efficacy. Daptomycin should not be used for primary staphylococcal pneumonia because of its failure in treating pneumococcal pneumonia. Such failure is due, in part, to a high degree of protein binding (figure 4) and its unique inactivation by surfactant [39]. MRSA SKIN AND SOFT-TISSUE INFECTIONS The prevalence of MRSA skin and soft-tissue infections has increased rapidly during the past few years and has resulted in outbreaks among prisoners, high school football and fencing teams, and professional football teams. Although many infections have been mild (e.g., impetigo or furunculosis), large abscesses and necrotizing fasciitis have also been reported. Most of these infections are caused by a single clone of MRSA bearing SCCmec IV with a high prevalence of Panton-Valentine leukocidin. Unlike hospital-acquired strains, many of these strains remain susceptible to tetracycline and trimethoprim-sulfamethoxazole, although resistance to erythromycin is common and inducible resistance to clindamycin is increasing in certain geographic regions. Several agents with in vitro activity against MRSA have recently been compared in clinical trials. For example, linezolid, a protein synthesis inhibitor, is an effective treatment for complicated soft-tissue infections when given intravenously or orally [40]. A recent trial comparing linezolid with vancomycin for the treatment of soft-tissue infections due to MRSA documented clinical cure rates of 79.4% and 73.3%, respectively [17]. For all other types of MRSA infections, there is no clear difference in treatment efficacy between antimicrobial agents classified as bactericidal versus bacteriostatic. Interestingly, a second larger study demonstrated markedly better efficacy for linezolid (a protein synthesis inhibitor), compared with vancomycin (a cell wallactive agent) [41]. Daptomycin is 190% protein bound [42] and, as stated earlier, has a long postantibiotic effect [36]. Daptomycin reached concentrations of 27.6 mg/mL in inflammatory tissues, such as skin blister fluid, after intravenous infusion of 4 mg/kg/day. At 18 and 24 h after the dose was administered, the inflammatory tissue concentrations were 10 and 8 mg/mL, respectively; thus, the free drug concentrations were 1 and 0.8 mg/mL, respectively. At these time points, the free drug concentrations were below the MIC of some strains of S. aureus. Despite these marginal free drug concentrations at the end of the dosing interval, the long postantibiotic effect provides antibacterial activity for most of the 24-h dosing interval. Dalbavancin, a semisynthetic glycopeptide for intravenous infusion, inhibits peptidyl cross-linking and has a very long half-life of 912 days. It is active against both MSSA and MRSA (MICs of 0.13 and 0.5 mg/mL, respectively) [43]. Like most antibiotics with long serum half-lives, dalbavancin has extremely high levels of protein binding (195%) [44]. Clearly, plasma concentrations of 135 mg/mL are realized for 7 days, and, at 8 days, concentrations in serum are 3031 mg/mL. This translates to free drug plasma concentrations of 1.51.75 mg/ mL. Thus, for bloodstream infections, the MIC of most organisms would be exceeded by a factor of 34 for the entire dosing interval. Although tissue concentrations of dalbavancin in humans are not yet available, a small, proof-of-concept trial showed the drug to be an effective treatment for skin and softtissue infections cause by S. aureus, including MRSA [45]. In the past, there has been less emphasis on tissue levels and protein binding of antibiotics, largely because of the very low b-lactam antibiotic MICs of gram-positive pathogens, such as pneumococcus and S. aureus.. As the MICs of nafcillin and cephalosporins have increased dramatically for these bacteria, additional agents have been developed. Vancomycin and several newer agents are effective against MRSA and have relatively high MICs against MRSA isolates (0.52.0 mg/mL). There is wide variation in the dynamics of tissue penetration and the degree of protein binding among these agents. Because free drug concentrations are the determinant of antibacterial activity of all antibiotics, it is imperative that we understand these aspects of pharmacology if we are to choose agents effective against resistant pathogens that infect different organs and tissues. It is easy to understand why highly protein-bound antibiotics with MICs of several micrograms per milliliter may fail in the treatment of microbes. Problems associated with vancomycin treatment are now emerging. It is more important than ever to (1) establish a definitive diagnosis, (2) obtain antibiotic susceptibility information, and (3) be aware of resistance patterns of communityand hospital-acquired pathogens. In addition, if vancomycin is selected as treatment for MRSA infections, careful monitoring of serum levels is critical to assure adequate dosing, irrespective of whether twice-daily dosing or continuous infusion is used. For practicing physicians, the next decade will prove to be challenging as antibiotic resistance continues to accelerate. We are more dependent than ever on pharmaceutical industry research to provide us with newer, more effective treatments. Acknowledgments Potential conflicts of interest. D.L.S. has received grant and research support from Pfizer and is a consultant for Arpida, Basilia, and Pfizer. References


This is a preview of a remote PDF: http://cid.oxfordjournals.org/content/42/Supplement_1/S51.full.pdf

Dennis L. Stevens. The Role of Vancomycin in the Treatment Paradigm, Clinical Infectious Diseases, 2006, S51-S57, DOI: 10.1086/491714