Validation of Vitek version 7.01 software for testing staphylococci against vancomycin☆

Diagnostic Microbiology and Infectious Disease 43 (2002) 135–140

www.elsevier.com/locate/diagmicrobio

Validation of Vitek version 7.01 software for testing staphylococci against vancomycin夞 P.M. Raneya,*, P.P. Williamsb, J.E. McGowan, Jr,b F.C. Tenovera a

Division of Healthcare Quality Promotion, Centers for Disease Control and Prevention, Atlanta, Georgia, USA b Rollins School of Public Health, Emory University, Atlanta, Georgia, USA

Abstract We tested 143 isolates of staphylococci with vancomycin by the National Committee for Clinical Laboratory Standards broth microdilution (BMD) reference method and compared the results to those generated using the Vitek automated system (GPS-105 and GPS-107 cards and version 7.01 software). For ten isolates, the vancomycin MICs by BMD were 8 ␮g/ml. By Vitek, the vancomycin MICs ranged from 2 to 16 ␮g/ml. Vancomycin MICs of ⱖ32 ␮g/ml were reported for two additional isolates by Vitek; however, the MICs decreased to ⱕ0.5 ␮g/ml on retesting. By BMD, the vancomycin MICs for both isolates were 1 ␮g/ml. While the modal vancomycin MIC results by BMD for S. aureus and coagulase-negative staphylococci (CoNS) were both 1 ␮g/ml, Vitek results showed a mode of ⱕ0.5 ␮g/ml for S. aureus, and a mode of 2 ␮g/ml for CoNS. Vitek did not report vancomycin MICs of 1 or 4 ␮g/ml for any of the isolates tested. While the sensitivity of detecting staphylococci with reduced susceptibility to vancomycin appears to be improved with Vitek version 7.01 software, when compared to earlier software versions, laboratories may notice an overall shift in MIC data toward higher vancomycin MICs, although for the most part, this does not affect the categorical interpretations of the results.

1. Introduction Vancomycin remains the primary antimicrobial agent for treatment of infections caused by methicillin-resistant Staphylococcus aureus and coagulase-negative staphylococci (CoNS) (Chambers, 1997; Gilbert et al., 2001). However, there have been several recent reports of S. aureus and CoNS with reduced susceptibility to vancomycin associated with treatment failure (Garrett et al., 1999; Hiramatsu et al., 1997; Kim et al., 2000; Sieradski et al., 1999; Smith et al., 1999). In none of the cases did the vancomycin MICs exceed 16 ␮g/ml, which, by National Committee for Clinical Laboratory Standards (NCCLS) classification, is still considered in the intermediate range. However, concern was raised about the ability of clinical laboratories to detect this emerging resistance problem (Tenover et al. 1998). One of the methods frequently used in laboratories to detect decreased susceptibility to vancomycin among iso-

lates of S. aureus and CoNS is the Vitek automated system1 (bioMe´rieux, Hazelwood, Mo.). Earlier reports indicated that previous Vitek software versions often reported vancomycin MICs in the NCCLS susceptible range (ⱕ0.5– 4␮g/ ml) for isolates that by broth microdilution (BMD) demonstrated vancomycin MICs in the intermediate range (8 –16 ␮g/ml) (Tenover et al., 1998). Vitek recently introduced software version 7.01, which purportedly is more sensitive than earlier software versions in detecting S. aureus and CoNS with reduced susceptibility to vancomycin (Vitek Product Information, 1999). In our study, we evaluated 143 selected isolates of staphylococci to determine whether the new Vitek software improved detection of staphylococcal isolates with decreased susceptibility to vancomycin.

2. Materials and methods 2.1 Bacterial strains

夞The information contained in this manuscript was presented in part as poster D-1624 at the 40th Interscience Conference on Antimicrobial Agents and Chemotherapy, Toronto, Ontario, Canada, Sept. 2000. * Corresponding author: Tel.: ⫹1-404-639-2824; fax: ⫹1-404-6391381 E-mail address: [email protected] (P. M. Raney)

143 isolates of staphylococci (Table 1) from 28 U.S. states and Japan that were submitted to the Centers for Disease Control and Prevention (CDC) or Project ICARE (Intensive Care Antimicrobial Resistance Epidemiology) (Fridkin et al., 1999; Hubert et al., 1999) for confirmation of

0732-8893/02/$ – see front matter © 2002 Elsevier Science Inc. All rights reserved. PII: S 0 7 3 2 - 8 8 9 3 ( 0 2 ) 0 0 3 7 8 - 4

136

Raney et al. / Diagnostic Microbiology and Infectious Disease 43 (2002) 135–140

Table 1 Phase I: Vancomycin MICs generated by broth microdilution and Vitek method Staphylococcus spp.

No. of isolates

MIC Results (␮g/ml) Broth microdilution method

S. aureus S. epidermidis S. hominis S. warneri S. haemolyticus S. saprophyticus S. lugdenensis S. capitis S. simulans Total

84 32 8 7 6 2 2 1 1 143

Vitek (GPS-107)

ⱕ0.06 0.12

0.25

0.5

1

2

4

8

16

ⱖ32

ⱕ0.5

1

2

4

8

16

ⱖ32

— — — — — — — — — 0

1 — — — — — — — — 1

15 — 2 — — 2 — — — 19

37 9 6 6 2 — 2 1 1 64

10 19 — — 1 — — — — 30

16 1 — — 2 — — — — 19

5 3 — 1 1 — — — — 10

— — — — — — — — — 0

— — — — — — — — — 0

43 2 4 1 — 2 2 — 1 55

— — — — — — — — — 0

26 26 4 5 3 — — 1 — 65

— — — — — — — — — 0

1 — — — 1 — — — — 2

12 4 — 1 2 — — — — 19

2 — — — — — — — — 2*

— — — — — — — — — 0

* The vancomycin MICs for these isolates dropped to ⱕ0.5 ␮g/ml on retesting

resistance phenotypes were selected for antimicrobial susceptibility testing. All but four S. aureus isolates and 16 CoNS isolates were oxacillin-resistant by BMD. The isolates were transferred from -70°C storage onto Trypticase Soy agar plates containing 5% defibrinated sheep blood (BD BioSciences, Sparks, MD) and subcultured a second time before testing. The isolates were identified by catalase and coagulase tests (Staphaurex, Murex, UK) and standard biochemical procedures (Kloos and Bannerman, 1999).

a second time to blood agar plates prior to testing on the Vitek systems. The quality control strains used for the BMD method were Enterococcus faecalis ATCC 29212, S. aureus ATCC 29213, and S. aureus ATCC 43300. For the Vitek method, the quality control strains used were E. faecalis ATCC 29212, S. aureus ATCC 29213, Escherichia coli ATCC 35218, and E. faecalis ATCC 51299.

2.2 Study methods

3. Results

The study was conducted in three phases. In Phase I, two methods, the NCCLS BMD method and the Vitek automated system, were used to determine the vancomycin susceptibility of the study isolates. For BMD, 96-well MIC plates were prepared in-house using NCCLS methods (NCCLS, 2000). Vancomycin (Eli Lilly & Co., Indianapolis, IN) was tested at concentrations of 0.06 –128 ␮g/ml. Testing and interpretation were performed according to NCCLS document M7-A5 (NCCLS, 2000). For the Vitek method, GPS-107 cards and the version 7.01 software were used. Testing was performed following manufacturer’s instructions. In Phase II, 139 of the 143 isolates were tested using the Vitek method with both GPS-105 and GPS-107 cards; four isolates were not available for testing. In Phase III, two hospital laboratories in Atlanta (hospitals A and B) that use the Vitek system with software version 7.01 with GPS-106 cards participated in an inter-laboratory MIC comparison using a subset of isolates. Fifteen S. aureus strains for which the broth microdilution MICs were 1 ␮g/ml (seven isolates), 4 ␮g/ml (six), or 8 ␮g/ml (two) were selected. The isolates were subcultured from frozen stocks, transferred to Trypticase Soy agar plates containing 5% defibrinated sheep blood, and incubated at 35°C for 18 h. The isolates were sent to the laboratories where the isolates were subcultured

3.1 Broth microdilution versus Vitek (Phase I ) The vancomycin MIC distributions generated by BMD and the Vitek method using the GPS-107 card are shown in Table 1. For 143 isolates of S. aureus and CoNS, the BMD method yielded a vancomycin MIC range of 0.25– 8 ␮g/ml with a mode of 1 ␮g/ml for both S. aureus and the CoNS. The vancomycin MIC range from the Vitek was ⱕ0.5-ⱖ32 ␮g/ml with a mode of ⱕ0.5 ␮g/ml for S. aureus and 2 ␮g/ml for CoNS. The Vitek did not report vancomycin MICs of 1 or 4 ␮g/ml for any of the study isolates. On initial testing by BMD, five isolates of S. aureus and five of CoNS demonstrated vancomycin MICs of 8 ␮g/ml (Figure 1). However, by Vitek testing, vancomycin MICs of 8 ␮g/ml were reported for only two isolates (one S. aureus and one CoNS), although for 19 isolates (12 S. aureus and seven CoNS) the vancomycin MICs were 16 ␮g/ml. Vancomycin MICs of ⱖ32 ␮g/ml were reported initially for two S. aureus isolates, but the MICs decreased to ⱕ0.5 ␮g/ml on retesting (Figure 1). Of the 23 strains for which the vancomycin MICs were 8 to ⱖ32 ␮g/ml by Vitek, only nine showed MICs of 8 ␮g/ml by BMD; for the remainder MICs were 2– 4 ␮g/ml. The two strains for which the initial Vitek vancomycin

Raney et al. / Diagnostic Microbiology and Infectious Disease 43 (2002) 135–140

137

Fig. 1. Comparison of MIC (␮g/ml) results for broth microdilution (BMD) and Vitek methods for 143 staphylococcal isolated tested in phase I. Solid lines indicate MIC breakpoints for interpretive categories (susceptible, intermediate, resistant). Area between dashed lines indicates MIC results that are ⫹/⫺ 1 dilution from BMD results. Bold underlined numbers represent major errors. Numbers in bold (not underlined) represent minor errors. For the two strains marked with an asterisk, MICs were ⱕ0.5␮g/ml when retested on the Vitek.

MICs were ⱖ32 ␮g/ml had BMD MIC results of 1 ␮g/ml (which is consistent with the MICs of ⱕ0.5 ␮g/ml observed after retesting the strains by Vitek). When the vancomycin MIC results from the Vitek and BMD were compared using BMD as the gold standard, the MIC results for 49 (34.3%) isolates were in absolute agreement, 77 (53.8%) were in essential agreement (⫹/⫺ one dilution), 14 (9.8%) had a two dilution difference, and three (2.1%) differed by three or more dilutions. Of the

17 MIC comparisons (13 S. aureus, two S. haemolyticus, and two S. epidermidis) that varied by two dilutions or more, two (those with initial MICs of ⱖ32 ␮g/ml) constituted major errors, 13 were minor errors, and two did not result in categorical errors. The percentage of MIC results for oxacillin-susceptible staphylococci that were classified as categorical errors (10.0%) was similar to the percentage of MIC results for oxacillin-resistant staphylococci also classified as having categorical errors (10.4%).

138

Raney et al. / Diagnostic Microbiology and Infectious Disease 43 (2002) 135–140

Fig. 2. Comparison of Vitek GPS-107 card results to GPS-105 results for 139 staphylococcal isolates tested in phase II. Solid lines indicate MIC breakpoints for interpretive categories (susceptible, intermediate, resistant). Area between dashed lines indicate MIC results that are ⫹/⫺ 1 dilution from GPS-105 card. Bold numbers represent the number of isolates that had MICs that were two or more dilutions from the GPS-105 results.

3.2 GPS-105 versus GPS-107 cards (Phase II) For 139 isolates, the MIC results obtained using GPS105 cards were compared with those using GPS-107 cards (Figure 2). Absolute agreement was shown for MIC results from 129 (92.8%) of the isolates; results for two (1.4%) additional isolates showed essential agreement. For the latter two isolates the GPS-105 card yielded MICs of 16 ␮g/ml for both isolates while the GPS-107 card reported one isolate as ⱖ32 ␮g/ml and the other as 8 ␮g/ml. Repeat testing of these isolates on the Vitek using GPS-105 and GPS-107 cards demonstrated vancomycin MICs of 16 ␮g/ml for both isolates. (For these isolates, the BMD MICs were 4 and 8 ␮g/ml). For eight (5.8%) isolates (six S. aureus, one S. hominis, and one S. warneri) the MICs reported by the two cards differed by two dilutions. Among these isolates, the MICs reported for the GPS-105 and the GPS-107 cards were ⱕ0.5 and 2 ␮g/ml (four isolates), 2 and 16 ␮g/ml (two), 2 and ⱕ0.5 ␮g/ml (one), and 16 and 2 ␮g/ml (one), respectively. 3.3 Inter-laboratory comparison of Vitek MIC results (Phase III) Fifteen isolates from Phases I and II, for which the vancomycin MICs were 1, 4, or 8 ␮g/ml by BMD, were selected for further testing in hospitals A and B by the Vitek system with software version 7.01 using GPS-106 cards (Table 2). For six of the 15 isolates that demonstrated BMD MICs of 4 ␮g/ml, hospitals A and B reported MICs by Vitek of 2 ␮g/ml (five) or 16 ␮g/ml (one). For the two isolates

with BMD MICs of 8 ␮g/ml, hospital A reported Vitek MICs of 2 ␮g/ml, but hospital B reported MICs of 16 ␮g/ml. Neither hospital reported Vitek MICs for vancomycin of 1 ␮g/ml or 4 ␮g/ml for any of the study isolates, as was true in Phases I and II of this study.

4. Discussion Testing staphylococci with reduced susceptibility to vancomycin using automated MIC methods, such as Vitek and Microscan rapid panels, showed poor sensitivity in earlier studies (Tenover et al., 1998). This study assessed the ability of the new Vitek software Version 7.01 to detect reduced susceptibility to vancomycin in staphylococcal isolates. Our data show that the Vitek was able to detect S. aureus and CoNS isolates with reduced susceptibility to vancomycin; however, MIC results generated by Vitek frequently were higher than those reported by BMD. Vitek vancomycin MICs of ⱖ32 ␮g/ml were not reproducible. We recommend that all isolates for which the vancomycin MICs are ⱖ4 ␮g/ml by Vitek testing should be repeated. If the high MICs from Vitek are reproducible, the isolates should be tested using an alternative NCCLS-approved method, such as BMD using a 24-h incubation period, to confirm the resistance phenotype. Comparison of the vancomycin MIC data between the GPS-105 and GPS-107 cards showed no systematic differences. Therefore, the decision to use the GPS-105 or GPS107 card can be based on other considerations, such as the

Raney et al. / Diagnostic Microbiology and Infectious Disease 43 (2002) 135–140

139

Table 2 Phase III: Vitek vancomycin MIC (␮g/ml) comparison for fifteen Staphylococcus aureus isolates tested at three sites Isolate number

2000 2001 2004 2008 2015 2039 2128 HIP 5173 HIP 7736 HIP 8926 HIP 9143 HIP 9172 HIP 9358 HIP 5827 HIP 5836

Emory/CDC Phase I

Emory/CDC Phase II

Broth microdilution

Vitek GPS-107

Vitek GPS-105

Vitek GPS-107

Hospital A GPS-106

Hospital B GPS-106

1 1 1 1 1 1 1 4 4 4 4 4 4 8 8

ⱖ32, repeat ⱕ0.5 ⱖ32, repeat ⱕ0.5 ⱕ0.5 ⱕ0.5 ⱕ0.5 2 2 16 16 2 16 2 16 16 8

ⱕ0.5 ⱕ0.5 ⱕ0.5 ⱕ0.5 ⱕ0.5 2 2 2 16 2 16 2 16, repeat 16 16 16

ⱕ0.5 ⱕ0.5 ⱕ0.5 ⱕ0.5 2 2 2 2 16 2 16 2 ⱖ32, repeat 16 16 8

ⱕ0.5 ⱕ0.5 2 ⱕ0.5 ⱕ0.5 2 2 2 2 2 2 2 16 2 2

ⱕ0.5 ⱕ0.5 ⱕ0.5 ⱕ0.5 ⱕ0.5 2 2 2 16 2 2 2 2 16 16

variety of other antimicrobial agents that each card has to offer. Because Vitek software version 7.01 did not report vancomycin MICs of 1 or 4␮g/ml, the distribution of vancomycin MICs for the Vitek method was quite different from that of the BMD method, and different from those produced by previous versions of the Vitek software in our laboratory. This may result, over time, in an upward shift of vancomycin MICs in a healthcare facility’s antibiogram. Although this shift may not change the categorical interpretation of the individual susceptibility test results, the increase may erroneously suggest increasing resistance to vancomycin in a healthcare facility.

Notes 1. Use of trade names is for identification purposes only and does not constitute endorsement by the Public Health Service or the U.S. Department of Health and Human Services.

Acknowledgments We thank Robert Jerris and David Lonsway for their testing of selected isolates and Christine Steward for her advice and helpful discussions. This work was supported in part by grants to the Rollins School of Public Health of Emory University by AstraZeneca Pharmaceuticals (Wilmington, DE), Pfizer, Incorporated (New York, NY), and Roche Laboratories (Nutley, NJ) as full sponsors, and Aventis Pharma (formerly RhonePoulenc Rorer), Bridgewater, NJ), the National Foundation for Infectious Diseases (Bethesda, MD), the American So-

Phase III

ciety for Health Systems Pharmacists Research and Educational Foundation (Bethesda, MD), Bayer Corporation, Pharmaceuticals Division (West Haven, CT), Roche Laboratories, Nutley, NJ, and Kimberly Clark Corporation (Roswell, GA) as partial sponsors.

References bioMe´ rieux Vitek, Inc. Vitek Software Product Information–VTK R07.01. Vitek analysis software. Nov. 1999, Centers for Disease Control and Prevention. (2000). Staphylococcus aureus with reduced susceptibility to vancomycin-Illinois, 1999. Morbid Mortal Weekly Rep 48: No 51;1165. Centers for Disease Control and Prevention. (1997). Staphylococcus aureus with reduced susceptibility to vancomycin-United States, 1997. Morbid Mortal Weekly Rep 46, 765–766. Chambers, H. F. (1997). Methicillin resistance in staphylococci: molecular and biochemical basis and clinical implications. Clin Microbiol Rev 10 (4),781–791. Fridkin, S. K., Steward C. D., Edwards, J. R., Pryor, E. R., McGowan, J. E., Archibald, L. K., Gaynes, R. P., Tenover, F. C. Project Intensive Care Antimicrobial Resistance Epidemiology Hospitals. (1999). Surveillance of antimicrobial use and antimicrobial resistance in U.S. hospitals: Project ICARE Phase 2. Clin Infect Dis 29, 245–252. Garrett, D., Jochimsen, E., Murfitt, K., Hill, B., McAllister, S., Nelson, P., Spera, R., Sall, R., Tenover, F., Jarvis, W. (1999). The impending apocalypse, the emergence of vancomycin resistance in Staphylococcus spp. Infect Control Hosp Epidemiol 20, 167–170. Gilbert, D.N., Moellering, R. C., Sande, M. A. (2001). The Sanford guide to antimicrobial therapy. 31st ed. Hyde Park, VT: Antimicrobial Therapy, Inc. Hiramatsu, K., Hanaki, H., Ino, T., Yabuta, K., Oguri, T., Tenover, F.C. (1997). Methicillin-resistant Staphylococcus aureus clinical strain with reduced vancomycin susceptibility. J Antimicrob Chemother 40, 135– 136. Hubert, S. K., Mohammed, J. M., Fridkin, S. K., Gaynes, R. P., McGowan J. E., Tenover, F. C. (1999). Glycopeptide-intermediate Staphylococcus aureus: evaluation of a novel screening method and results of a survey of selected U.S. hospitals. J Clin Microbiol 37, 3590 –3593.

140

Raney et al. / Diagnostic Microbiology and Infectious Disease 43 (2002) 135–140

Kim, M. N., Pai, C. H., Woo, J. H., Ryu, J. S., Hiramatsu, K. (2000). Vancomycin-intermediate Staphylococcus aureus in Korea J Clin Microbiol 38, 3879 –3881. Kloos, W. E., Bannerman, T. L. (1999). Staphylococcus and Micrococcus. In P R. Murray, E. J. Baron, M. A. Pfaller, F. C. Tenover, R. H. Yolken (Eds.), Manual of Clinical Microbiology, 7th ed (pp. 264 –282). Washington, DC: ASM Press. National Committee for Clinical Laboratory Standards. (2000). Methods for Dilution Antimicrobial Susceptibility Tests for Bacteria That Grow Aerobically. NCCLS Approved Standard M7–A5. Wayne, PA: National Committee for Clinical Laboratory Standards. National Committee for Clinical Laboratory Standards. (2000). Performance Standards for Antimicrobial Susceptibility Testing; Ninth Infor-

mational Supplement, M100 –S9. Wayne, PA: National Committee for Clinical Laboratory Standards. Sieradzki, K., Roberts, R. B., Haber, S. W., Tomasz, A. (1999). The development of vancomycin resistance in a patient with methicillinresistant Staphylococcus aureus infection. N Engl J Med 340:517–523. Smith, T., Pearson, M. L., Wilcox, K. R., Cruz, C., Lancaster, M. L., Robinson-Dunn, B. (1999). Emergence of vancomycin resistance in Staphylococcus aureus: epidemiology and clinical significance. N Engl J Med 340, 493–501. Tenover, F. C., Lancaster, M. V., Hill, B.C., Steward, C. D., Stocker, S. A., Hancock, G. A., O’Hara, C. M., Clark, N.C., Hiramatsu, K. (1998). Characterization of staphylococci with reduced susceptibilities to vancomycin and other glycopeptides. J Clin Microbiol 36 (4), 1020 –1027.