Effects of prolonged vancomycin administration on methicillin-resistant Staphylococcus aureus (MRSA) in a patient with recurrent bacteraemia

Journal of Antimicrobial Chemotherapy (2006) 57, 699 704 doi:10.1093/jac/dkl030 Advance Access publication 7 February 2006 Effects of prolonged vancomycin administration on methicillin-resistant Staphylococcus aureus (MRSA) in a patient with recurrent bacteraemia George Sakoulas 1 *, Howard S. Gold 2,3, Robert A. Cohen 2,3, Lata Venkataraman 2, Robert C. Moellering 2,3 and George M. Eliopoulos 2,3 1 Westchester Medical Center and New York Medical College, Valhalla, NY, USA; 2 Department of Medicine, Beth Israel Deaconess Medical Center, Boston, MA, USA; 3 Harvard Medical School, Boston, MA, USA Received 14 June 2005; returned 20 September 2005; revised 11 October 2005; accepted 21 January 2006 Objectives: To evaluate microbiological properties of methicillin-resistant Staphylococcus aureus (MRSA) during prolonged vancomycin therapy. Methods: We evaluated vancomycin susceptibility and heteroresistance, accessory gene regulator (agr) function, autolysis, biofilm production and in vitro vancomycin killing in serial MRSA bloodstream isolates obtained over a 30 month period from a patient with a chronic endovascular infection. Results: Despite the fact that the MRSA in this patient had the same genetic background as other clinical glycopeptide intermediate-resistant S. aureus (GISA) isolates, vancomycin administered for 9 months, maintaining serum concentrations >10 mg/l, did not select for GISA. Minimal changes in vancomycin susceptibility were detected using agar dilution and population analysis methods. We noted increases in delta haemolysin production, autolysis and the bactericidal effects of vancomycin in vitro against the MRSA obtained after prolonged vancomycin suppressive therapy was discontinued. Conclusions: Despite the lack of development of detectable resistance, MRSA exposed to vancomycin for prolonged periods may begin to develop vancomycin tolerance and decreased autolysis. In addition, suppression of agr function appears to end after vancomycin is stopped. Whether these changes are prerequisites for attenuated vancomycin efficacy and the development of glycopeptide resistance warrants further study. The development of vancomycin resistance may be more difficult under conditions where vancomycin serum concentrations are maintained >10 mg/l. Keywords: glycopeptides, autolysis, biofilms, agr Introduction Staphylococcus aureus is an increasingly common pathogen in community-acquired and nosocomial infections. The emergence of glycopeptide intermediate-resistant S. aureus (GISA) 1 3 and, most recently, vancomycin-resistant S. aureus 4 further limits therapeutic options for clinicians. In addition, some authors have reported suboptimal clinical efficacy of vancomycin in the treatment of fully vancomycin-susceptible methicillinresistant S. aureus (MRSA) bloodstream infections since their emergence in the clinical mainstream in the mid-1980s. 5 9 The development of GISA may to some extent simply reflect the climax of a sequence of adaptive responses of the organism under increased vancomycin selection pressure. Early changes in the physiology of S. aureus in the pathway towards vancomycin resistance may be undetectable by current methods of susceptibility testing. We were provided with a unique opportunity to study the physiological effects of prolonged vancomycin exposure on MRSA in a patient with a chronic endovascular infection. We noted significant physiological alterations in the organism with minimal changes in vancomycin susceptibility.... *Corresponding author. Tel: +1-914-594-4974; Fax: +1-845-361-1156; E-mail: george_sakoulas@nymc.edu... 699 Ó The Author 2006. Published by Oxford University Press on behalf of the British Society for Antimicrobial Chemotherapy. All rights reserved. For Permissions, please e-mail: journals.permissions@oxfordjournals.org

Sakoulas et al. Table 1. Antimicrobial susceptibility of MRSA bloodstream isolates Methods Bacterial isolates Table 1 demonstrates the temporal relationship of MRSA isolates obtained from a haemodialysis-dependent patient with a persistently bacteraemic endovascular infection that was not amenable to surgical therapy, as described previously. 10 The patient developed the first MRSA bacteraemia with isolate A8821 and responded to 6 weeks of vancomycin therapy. However, the bacteraemia relapsed four times (isolates A8822 A8825) and each was treated with 6 weeks of vancomycin with varying durations of gentamicin, after which it was decided to maintain the patient on vancomycin therapy long term. Vancomycin troughs were monitored after dialysis and maintained above 10 mg/l. Breakthrough bacteraemia after 9 months of vancomycin (A8826) led to a switch to linezolid for ongoing suppression. The organism, which eventually became linezolid-resistant (A8830, A8831), was characterized by our laboratory. 10 In the month between isolation of A8830 and A8831, the patient received both vancomycin and linezolid. All bloodstream isolates from this patient were analysed and found to be indistinguishable by pulsed-field gel electrophoresis (PFGE) using methods described previously. 11 Susceptibility testing for vancomycin, oxacillin and linezolid was performed using agar dilution techniques as described by the NCCLS. 12 Analysis of isolates MIC (mg/l) a Isolate Months b VAN OXA A8821 0 0.5 128 A8822 +1 1.0 128 A8823 +2 1.0 128 A8824 +4 1.0 >128 A8825 +6 1.0 >128 A8826 +9 1.0 >128 A8827 1 1.0 >128 A8828 3 1.0 128 A8829 4 1.0 >128 A8830 19 1.0 128 A8831 +1 1.0 128 VAN, vancomycin; OXA, oxacillin. a All isolates were resistant to erythromycin and quinolones. All isolates were susceptible to clindamycin, gentamicin and tetracycline. b A positive number represents months on VAN therapy. A negative number represents months off VAN therapy. Vancomycin bactericidal assays were performed using a starting inoculum of 10 6 cfu/ml bacteria in brain heart infusion (BHI) broth (Becton Dickinson) containing 16 mg/l vancomycin (vancomycin hydrochloride, American Pharmaceutical Partners, Los Angeles, CA, USA) as described previously. 13,14 Aliquots were obtained at 0, 4, 24 and 72 h for assessment of viable bacterial density as described previously. 13,14 Vancomycin population analyses were performed using a suspension of 10 9 cfu/ml of bacteria that were serially diluted at 10-fold intervals and plated in duplicate (25 ml) on BHI agar containing varying concentrations of vancomycin as described previously. 13,14 Adherence to polystyrene (marker for biofilm production) was performed as described previously 13 by diluting overnight cultures 1:200 in Trypticase soy broth (Difco Laboratories, Detroit, MI, USA) supplemented with 1% glucose and dispensing 200 ml aliquots in triplicate into wells of a sterile 96-well flat-bottom plastic tissue culture plate (Costar Corporation, Cambridge, MA, USA). Bacteria were grown aerobically for 20 h at 35 C. Comparable bacterial growth in each well was confirmed by measuring the optical density at 630 nm (OD 630 ) with an MRXII Microplate Reader (Dynex Technologies, Chantilly, VA, USA). The supernatant was removed, and each well was washed four times with 300 ml of sterile phosphatebuffered saline. The adherent cells were fixed by heating at 65 C for 1 h. Cells were stained with 200 ml of Gram crystal violet (Becton Dickinson), and residual stain was removed with tap water. Plates were air dried, and the OD 540 of the stained adherent bacterial films was measured. A semi-qualitative assessment of agr function using delta haemolysin expression was performed on sheep blood agar plates as described previously. 13 Autolysis in Triton X-100 was determined using a modification of methods described previously. 14,15 Briefly, bacteria were grown in BHI broth to an approximate optical density at 630 nm (OD 630 )of 0.8. Cells were spun down in a microcentrifuge tube for 20 s, washed twice with cold water and resuspended to an OD 630 of 1.0 in 50 mm glycine, 0.01% Triton X-100, ph = 8.0. The suspension was aliquotted in triplicate into a sterile 96-well flat-bottom plastic tissue culture plate (Costar Corporation) and incubated at 35 C with gentle agitation. OD 630 was measured at 4 h using an MRXII Microplate Reader (Dynex Technologies) and expressed as a fraction of the OD 630 at 0 h. Results All MRSA bloodstream isolates from this patient demonstrated SmaI macrorestriction enzyme digest patterns by PFGE similar to the predominant MRSA subtype at our hospital (as demonstrated previously) as well as to clinical GISA. 13,16 All isolates from this patient were susceptible to vancomycin and resistant to oxacillin (Table 1). Vancomycin population analysis of a subset of isolates demonstrated no increase in vancomycin heteroresistance in this patient s MRSA despite repetitive and continuous exposure to vancomycin (Figure 1). Analysis of agr function of the initial bloodstream isolate A8821 demonstrated low baseline activity of delta haemolysin (Figure 2). With the sequential administration of vancomycin over 9 months, delta haemolysin activity remained low but was not abolished (Figure 2b). None of the isolates obtained from vancomycin treatment relapse demonstrated a complete lack of delta haemolysin expression. However, a noticeable increase in delta haemolysin expression was seen in bloodstream isolates obtained after therapy was switched from vancomycin to linezolid (Figure 2c e). Given the findings by us and others 13,17 showing an inverse relationship between agr function and adherence to polystyrene, a marker of biofilm production, we investigated polystyrene adherence of sequential bloodstream isolates from this patient. We found that the polystyrene adherence ability of MRSA remained stable during vancomycin administration. MRSA obtained after the discontinuation of vancomycin and switching to linezolid showed a sharp decline in polystyrene adherence (Figure 3). Sequential bloodstream isolates obtained from this patient during prolonged vancomycin administration demonstrated diminished autolysis in Triton X-100 (Figure 4). Isolates obtained 700

Prolonged vancomycin effects on MRSA 9.6 8.6 7.6 6.6 5.6 4.6 3.6 A8821 A8826 A8830 A5937 A5940 HIP5836 2.6 0 2 4 6 8 10 Vancomycin (mg/l) Figure 1. Population analysis of sequential MRSA bloodstream isolates A8821 (initial, pre-vancomycin), A8826 (after 9 months of vancomycin) and A8830 (after prolonged linezolid). Included for comparison are population analyses for MRSA A5937 (vancomycin MIC = 2 mg/l), 13 hetero-gisa A5940 (vancomycin MIC = 4 mg/l) 13 and GISA HIP5836 (vancomycin MIC = 6 8 mg/l; strain NRS4, www.narsa.net). The limit of detection was 400 cfu/ml (log 10 cfu/ml = 2.6). OD 540 nm 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0 5 10 15 20 25 30 Time (months) Figure 3. Polystyrene adherence of sequential bloodstream isolates. The horizontal axis denotes time after the initial bacteraemia and the vertical axis shows the degree of polystyrene binding. The vertical line denotes the time at which vancomycin therapy was stopped and switched to linezolid. The experiment was performed two times in triplicate and standard deviation is noted by the error bars. a c e b d Fraction of Initial OD 630 nm 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0 5 10 15 20 25 30 Time (months) Figure 4. Autolysis of sequential bloodstream isolates in Triton X-100. The horizontal axis denotes time after the initial bacteraemia and the vertical axis shows the fraction of the initial OD 630 remaining in Triton X-100 after 4 h. The vertical line denotes the time at which vancomycin therapy was stopped and switched to linezolid. The experiment was performed three times in triplicate and standard deviation is noted by the error bars. Figure 2. Delta haemolysin assay of sequential MRSA bloodstream isolates A8821 (a), A8826 (b), A8828 (c), A8829 (d) and A8830 (e). Arrows show representative areas of delta haemolysin activity, which is increased in (c), (d) and (e) compared with (a) and (b). after the switch in therapy from vancomycin to linezolid showed a significant increase in Triton-induced autolysis (Figure 4). To determine whether the prolonged exposure to vancomycin would have an effect on the bactericidal activity of vancomycin in vitro, we performed bactericidal assays on a subset of serial isolates from this patient and found that prolonged vancomycin resulted in attenuated bactericidal activity of vancomycin in A8824 and further attenuated activity in A8826. (Figure 5) This effect was reversed in A8830 after vancomycin administration was stopped. In A8831, the isolate obtained after vancomycin was re-administered, reduced killing was again seen at 72 h of inoculate in the presence of vancomycin (Figure 5). 7 6 5 4 3 2 1 A8821 A8824 A8826 A8830 A8831 0 0 10 20 30 40 50 60 70 80 Time (h) Figure 5. Bactericidal activity of vancomycin (16 mg/l) against MRSA bloodstream isolates. The results shown are of one representative experiment that was performed three times with slightly different starting inocula. 701

Sakoulas et al. (a) (b) 9.6 8.6 7.6 6.6 5.6 4.6 3.6 2.6 0 2 4 6 8 10 9.6 8.6 7.6 6.6 5.6 4.6 3.6 Vancomycin (mg/l) A8821 A8821-V1 A8826 A8826-V16 A8826-V1 2.6 0 2 4 6 8 10 Vancomycin (mg/l) Figure 6. (a) Population analysis of parent isolate A8821 and results of population analysis after growth in vancomycin 1 mg/l. (b) Population analysis of A8826 MRSA isolates and effects of in vitro exposure to either 1 or 16 mg/l vancomycin. The limit of detection was 400 cfu/ml (log 10 cfu/ml = 2.6). Despite in vivo exposure to vancomycin almost continuously for 9 months, the MRSA recovered from this patient (A8826) did not develop vancomycin heteroresistance as described above (Figure 1). In order to determine whether this isolate intrinsically lacked the ability to develop intermediate-level glycopeptide resistance and to study the effects of different vancomycin concentrations in vitro on glycopeptide heteroresistance, we performed vancomycin population analyses on A8821 and A8826 before and after exposure to broth containing 1 or 16 mg/l of vancomycin. For both the initial clinical MRSA A8821 (A8821-V1, Figure 6a) and the isolate obtained after 9 months of vancomycin therapy A8826 (A8826-V1, Figure 6b), growth in BHI broth containing vancomycin 1 mg/l resulted in a significant leftward shift towards vancomycin heteroresistance. However, exposure of bacteria to vancomycin 16 mg/l (cells harvested from 72 h assays of bactericidal activity) did not result in selection of heteroresistant colonies (A8826-V16, Figure 6b). Most notable was the fact that serial passage of both A8821 and A8826 in broth containing incrementally higher concentrations of vancomycin readily yielded GISA. Discussion Although vancomycin still remains the treatment of choice for MRSA bloodstream infections, vancomycin treatment failures in this setting are well documented. 5 9 Clinical experience with vancomycin appears to be heterogeneous, with some patients demonstrating a satisfactory response and others suffering prolonged or refractory bacteraemia despite appropriate serum concentrations of the drug. Several studies have demonstrated characteristics of MRSA that may predict treatment failure. 14,18 20 The isolates from the majority of cases with refractory bacteraemia are most probably fully susceptible by standardized methods of testing since the number of MRSA infections with reduced vancomycin susceptibility remains small. We encountered a unique clinical situation that provided us with the opportunity to determine the effects of vancomycin exposure in vivo on vancomycin susceptibility and bactericidal activity as well as other characteristics. While the findings that are described here are limited to one MRSA strain and therefore should be interpreted with caution, we feel that these studies provide some important general insights. First, it is important to point out that serum vancomycin concentrations in this patient were meticulously maintained at or above 10 mg/l during the 9 months of treatment. This duration of treatment was much longer than in the patients in whom GISA developed in the United States, where treatment ranged from 3.5 to 18 weeks. 3 Exposure to subtherapeutic vancomycin concentrations has been shown to be associated with development of hetero-gisa isolates clinically. 21,22 In prior in vitro studies, we serially passaged several S. aureus (methicillin-susceptible S. aureus and MRSA) in broth media containing incrementally higher concentrations of vancomycin. We observed an inherent difficulty in obtaining bacterial growth in media containing >10 mg/l of vancomycin. 14 Therefore, we hypothesize that the exposure of the organism consistently to vancomycin concentrations >10 mg/l was important in preventing the emergence of intermediate-level glycopeptide resistance in this patient. The theory of the importance of excessively low vancomycin concentrations in selecting for MRSA with reduced vancomycin susceptibility is further supported by other investigators, who applied agr group II S. aureus related to the MRSA in this study by PFGE to an in vitro pharmacodynamic model of vancomycin and found the development of vancomycin heteroresistance to be associated with suboptimal vancomycin dosing, with the agr group II knockout strain demonstrating a stronger predilection to developing glycopeptide heteroresistance. 22 A more precise quantitative determination of optimal vancomycin serum troughs to prevent the development of MRSA with reduced vancomycin susceptibility would require either in vivo studies or similar pharmacodynamic models. The infecting organism in this report was related to clinical GISA and predominating MRSA clones in the United States by PFGE and agr group II genotype (data not shown). Resistance quickly emerged in vitro when the organism was subjected to subinhibitory concentrations of vancomycin. It is important to note that the infection in this patient was felt to be endovascular, so desirable vancomycin concentrations were reliably achievable with appropriate dosing. Exposing organisms in abscess fluid, biliary system, lung tissue and bone to desirable levels of vancomycin would have been more challenging. On the basis of the ability to prevent the emergence of GISA in this patient, 702

Prolonged vancomycin effects on MRSA previously published in vitro data and the adverse clinical outcomes that are observed in patients in whom GISA or hetero- GISA emerge, we feel it is important to monitor vancomycin serum levels when duration of treatment is expected to be prolonged and to maintain trough concentrations >10 mg/l. Despite not selecting for GISA, prolonged vancomycin therapy was ultimately unsuccessful for this patient, necessitating the switch to linezolid. In fact, the patient experienced breakthrough bacteraemia despite an appropriate vancomycin serum concentration. Although this was not completely surprising given the prosthetic material in association with this infection, it highlights the fact that not all MRSA strains that test susceptible to vancomycin in the microbiology laboratory will respond adequately to vancomycin. This may even occur, as in this case, without the development of the hetero-gisa or GISA phenotypes. It is important to point out that the isolates in this study are still considered susceptible to vancomycin according to the new susceptibility breakpoint of 2 mg/l. The breakthrough organism demonstrated decreased in vitro killing in response to vancomycin at 72 h, a characteristic that was accompanied by decreased autolysis in Triton X-100. 13,16 Both phenotypes have been associated directly or indirectly with prolonged bacteraemia on vancomycin therapy. 20,23 These phenotypes appeared to be unstable in vivo because they were lost in subsequent isolates obtained after the patient stopped receiving vancomycin. The decreased bactericidal activity of vancomycin quickly returned when vancomycin was re-administered. This instability of phenotype suggests an effect not due to genetic mutations but rather due to changes in differential gene expression or the emergence and submergence of variable bacterial subpopulations depending on the antibiotic selection pressure. We found it noteworthy that the ability to adhere to polystyrene was influenced by antimicrobial therapy. Specifically, MRSA obtained after vancomycin was replaced by linezolid demonstrated reduced binding to polystyrene as delta haemolysin expression increased. Although we cannot definitively rule out effects of linezolid on decreasing polystyrene adherence, we believe that this phenomenon was due to decreased agr expression induced by vancomycin, resulting in increased biofilm formation in isolates obtained while the patient was receiving vancomycin therapy. We and others have previously shown that agr mutants demonstrate a marked increase in biofilm formation. More recent work by Yao et al. 24 provides evidence that in Staphylococcus epidermidis this occurs through decreased bacterial production of phenol-soluble modulins (PSMs), small peptides that serve to inhibit the formation of thick biofilms. Production of these peptides is under the control of agr, and agr mutants demonstrate severely compromised PSM production. Given the marked increase in implantable biomedical devices and the consequential increasing relevance of biofilms in bacterial pathogenesis and resistance, the study of antimicrobial effects on bacterial biofilm formation is of interest. In summary, this study shows that (i) although this strain showed a propensity to evolve into a hetero-gisa phenotype on exposure in vitro to subinhibitory concentrations of vancomycin, this was not observed in vivo after 9 months on vancomycin. We believe this was due to meticulous maintenance of serum vancomycin concentrations to avoid excessively low serum trough concentrations. (ii) Despite minimal changes in susceptibility by agar dilution and the failure to develop a hetero-gisa phenotype by population analysis, MRSA isolates obtained from clinical failure of vancomycin showed physiological changes when compared with the parent strain, including decreased autolysis, reduced killing by vancomycin in vitro, decreased delta haemolysin expression and increased biofilm production. These phenotypic changes were reversible when vancomycin was discontinued. Whether detecting these physiological changes as a means of more universally predicting vancomycin treatment failure in MRSA bacteraemia warrants prospective study. Transparency declarations H. S. G., R. C. M. and G. M. E. have received financial support through Pfizer Pharmaceuticals, New York, NY, USA. G. S. has received speaking fees from Cubist Pharmaceuticals (Lexington, MA, USA) and Sanofi-Aventis Pharmaceuticals (Bridgewater, NJ, USA). References 1. Geisel R, Schmitz FJ, Fluit AC et al. Emergence, mechanism, and clinical implications of reduced glycopeptide susceptibility in Staphylococcus aureus. Eur J Clin Microbiol Infect Dis 2001; 20: 685 97. 2. Hiramatsu K. Vancomycin-resistant Staphylococcus aureus: a new model of antibiotic resistance. Lancet Infect Dis 2001; 1: 147 55. 3. Srinivasan A, Dick JD, Perl TM. Vancomycin resistance in staphylococci. Clin Microbiol Rev 2002; 15: 430 8. 4. Sievert DM, Boulton ML, Stoltman G et al. Staphylococcus aureus resistant to vancomycin-united States 2002. Morb Mortal Wkly Rpt 2002; 51: 565 7. 5. Levine DP, Cushing RD, Jui J et al. Community-acquired methicillinresistant Staphylococcus aureus endocarditis in the Detroit Medical Center. Ann Intern Med 1982; 97: 330 8. 6. Levine DP, Crane LR, Zervos MJ. Bacteremia in narcotic addicts at the Detroit Medical Center. II. Infectious endocarditis: a prospective comparative study. Rev Infect Dis 1986; 8: 374 96. 7. Levine DP, Fromm BS, Reddy R. Slow response to vancomycin or vancomycin plus rifampin in methicillin-resistant Staphylococcus aureus endocarditis. Ann Intern Med 1991; 115: 674 80. 8. Small PM, Chambers HF. Vancomycin for Staphylococcus aureus endocarditis in intravenous drug users. Antimicrob Agents Chemother 1990; 34: 1227 31. 9. Karchmer AW. Staphylococcus aureus and vancomycin: the sequel. Ann Intern Med 1991; 115: 739 41. 10. Meka VG, Pillai SK, Sakoulas G et al. Linezolid resistance in sequential Staphylococcus aureus isolates associated with a T2500A mutation in the 23S rrna gene and loss of a single copy of rrna. J Infect Dis 2004; 190: 311 7. 11. Maslow J, Slutsky AM, Arbeit RD. The application of pulsed-field gel electrophoresis to molecular epidemiology. In: Persing DH, Smith TF, Tenover FC, White J, eds. Diagnostic Molecular Epidemiology: Principles and Applications. Washington, DC: American Society for Microbiology Press, 1993; 563 72. 12. National Committee for Clinical Laboratory Standards. Methods for Dilution Antimicrobial Susceptibility Tests for Bacteria that Grow Aerobically: Approved Standard M7-A5. NCCLS, Wayne, PA, USA, 2000. 13. Sakoulas G, Eliopoulos GM, Moellering RC Jr. et al. Accessory gene regulator (agr) locus in geographically diverse Staphylococcus aureus isolates with reduced susceptibility to vancomycin. Antimicrob Agents Chemother 2002; 46: 1492 502. 14. Sakoulas G, Eliopoulos GM, Moellering RC Jr et al. Accessory generegulator(agr) group II Staphylococcus aureus: is there a relationship 703

Sakoulas et al. to the development of intermediate-level glycopeptide resistance? J Infect Dis 2003; 187: 929 38. 15. Sieradzki K, Tomasz A. Inhibition of cell wall turnover and autolysis by vancomycin in a highly vancomycin-resistant mutant of Staphylococcus aureus. J Bacteriol 1997; 179: 2557 66. 16. Schwaber MJ, Wright SB, Carmeli Y et al. Clinical implications of varying degrees of vancomycin susceptibility in methicillin-resistant Staphylococcus aureus bacteremia. Emerg Infect Dis 2003; 9: 657 64. 17. Vuong C, Saenz HL, Gotz F et al. Impact of the agr quorum-sensing system on adherence to polystyrene in Staphylococcus aureus. J Infect Dis 2000; 182: 1688 93. 18. Sakoulas G, Moise-Broder PA, Schentag JJ et al. Relationship of MIC and bactericidal activity to efficacy of vancomycin for the treatment of methicillin-resistant Staphylococcus aureus bacteremia. J Clin Microbiol 2004; 42: 2398 402. 19. Moise-Broder PA, Sakoulas G, Eliopoulos GM et al. Accessory gene regulator group II polymorphism in methicillin-resistant Staphylococcus aureus is predictive of failure of vancomycin therapy. Clin Infect Dis 2004; 38: 1700 5. 20. Fowler VG Jr, Sakoulas G, McIntyre LM et al. Persistent bacteremia due to methicillin-resistant Staphylococcus aureus infection is associated with agr dysfunction and low-level in vitro resistance to thrombin-induced platelet microbicidal protein. J Infect Dis 2004; 190: 1140 9. 21. Charles PG, Ward PB, Johnson PD et al. Clinical features associated with bacteremia due to heterogeneous vancomycin-intermediate Staphylococcus aureus. Clin Infect Dis 2004; 38: 448 51. 22. Tsuji B, Rybak M. The influence of Staphylococcus aureus accessory gene regulator (agr) function on the development of vancomycin heteroresistance in an in vitro pharmacodynamic model. In: Abstracts of the Fifteenth European Congress of Clinical Microbiology and Infectious Diseases (ECCMID), Copenhagen, Denmark, 2005. Abstract 1134-03- 359, p. 1590. European Society for Clinical Microbiology and Infectious Disease, Basel, Switzerland. 23. Boyle-Vavra S, Carey RB, Daum RS. Development of vancomycin and lysostaphin resistance in a methicillin-resistant Staphylococcus aureus isolate. J Antimicrob Chemother 2001; 48: 617 25. 24. Yao Y, Sturdevant DE, Otto M. Genomewide analysis of gene expression in Staphylococcus epidermidis biofilms: insights into the pathophysiology of S. epidermidis biofilms and the role of phenolsoluble modulins in formation of biofilms. J Infect Dis 2005; 191: 289 98. 704