Diagnostic Microbiology and Infectious Disease 58 (2007) 191 198 www.elsevier.

com/locate/diagmicrobio

Antimicrobial Susceptibility Studies

Laboratory-based evaluation of the colorimetric VITEK-2 Compact system for species identification and of the Advanced Expert System for detection of antimicrobial resistances: VITEK-2 Compact system identification and antimicrobial susceptibility testing
Isamu Nakasonea,4, Tohru Kinjoa, Nobuhisa Yamaneb, Kyoko Kisanukia, Chika M. Shiohirab
a b

Clinical Laboratories, University Hospital of the Ryukyus, University of the Ryukyus, Nishihara-Nakagami, Okinawa 903-0215, Japan Department of Laboratory Medicine, Graduate School and Faculty of Medicine, University of the Ryukyus, Okinawa 903-0215, Japan Received 4 October 2006; accepted 8 December 2006

Abstract The newly redesigned colorimetric VITEK-2 Compact system with updated Advanced Expert System (AES) (bioMerieux, Marcy lEtoile, France) was evaluated for its accuracy and rapidity to identify clinical isolates and to detect several antimicrobial resistances. Overall, the VITEK-2 gave 95.8% of compatibility with the reference API strips (bioMerieux) in the identifications (IDs) of Gram-positive cocci (GPC), Gram-negative rods (GNR), and yeasts. The accuracy was finally estimated to 98.3% through additional confirmatory tests. Also, N 90% of IDs of GPC and GNR were obtained within 7 h of incubations. The VITEK AES correctly detected 97.7% of antimicrobial resistances, including extended-spectrum h-lactamases, oxacillin and inducible clindamycin resistances in staphylococci, vancomycin resistance in enterococci, and penicillin and erythromycin resistances in Streptococcus pneumoniae. The most resistant isolates were identified within 12 h of incubations. In conclusion, the new colorimetric VITEK-2 Compact system with AES greatly improved its accuracy in species ID and detection of antimicrobial resistances, and it will be highly acceptable to clinical microbiology laboratory function. D 2007 Elsevier Inc. All rights reserved.
Keywords: VITEK-2 Compact system; Advanced Expert System; Antimicrobial resistance

1. Introduction A series of the VITEK systems (bioMerieux, Marcy lEtoile, France) has been a fully automated instrument that provides species identification (ID) and antimicrobial susceptibility testing (AST) for a variety of clinical isolates, and are presently used in many clinical microbiology laboratories worldwide. During the past 3 decades, several revisions have been introduced to the system, resulting in a

Presented in part at the 106th General Meeting of American Society for Microbiology, Orlando, FL, May 2006, Abstract Current #C-008. 4 Corresponding author. Tel.: +81-98-895-3331x3332; fax: +81-98895-463. E-mail address: isamu@jim.u-ryukyu.ac.jp (I. Nakasone). 0732-8893/$ see front matter D 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.diagmicrobio.2006.12.008

stepwise improvement of the system performance. Recently, extensive revisions, including reintroduction of colorimetric reading in lieu of fluorescence technology, and addition of several biochemical substrates and taxa covered by the broadened database comparable with the well-established API series (bioMerieux) are created (Funke and FunkeKissling, 2004; Funke and Funke-Kissling, 2005; Aubertine et al., 2006). The efforts have been focused upon the accurate ID, in particular, to solve its inherent weakness in the IDs of glucose-nonfermentative Gram-negative rods (GNR) and members of the family Streptococcaceae (Joyanes et al., 2001; Gavin et al., 2002). In this communication, we describe the results to evaluate the accuracy of ID by the respective VITEK test cards on the VITEK-2 Compact system and to detect several antimicrobial

192

I. Nakasone et al. / Diagnostic Microbiology and Infectious Disease 58 (2007) 191 198 Table 1 (continued) Species ID by API strip (no. of isolates tested) No. of Discrepant tests ID by the agreed VITEK-2 (no. of cases) 5 5 1 5 13 12 11 4 0 Alcaligenes faecalis (2) A. lwoffii (2) incorrect Final ID VITEK-2 was correct or incorrect (no. of cases)

Table 1 Accuracy in species ID by the VITEK-2 Compact system Species ID by API strip (no. of isolates tested) No. of Discrepant tests ID by the agreed VITEK-2 (no. of cases) S. caprae (2)a Final ID VITEK-2 was correct or incorrect (no. of cases) S. caprae (2) correctb

S. aureus (21) 19 S. capitis (14) 14 S. caprae (7) 7 Staphylococcus 2 cohnii (2) Staphylococcus 18 epidermidis (18) Staphylococcus 9 haemolyticus (9) Staphylococcus 1 hominis (1) Staphylococcus 10 lugdunensis (10) Staphylococcus 1 saprophyticus (1) Staphylococcus 3 schleiferi (3) Staphylococcus 2 sciuri (2) S. simulans (1) 0 Staphylococcus 1 warneri (1) Enterococcus avium (8) 8 E. casseliflavus (8) 8 Enterococcus durans (1) 1 E. faecalis (14) 14 E. faecium (10) 10 E. gallinarum (13) 12 Streptococcus 15 agalactiae (16) Streptococcus 14 anginosus (15) S. constellatus (4) 3 S. dysgalactiae (5) Streptococcus mitis/oralis (21) Streptococcus mutans (1) S. parasanguis (2) S. pneumoniae (11) Streptococcus pyogenes (11) Streptococcus salivarius (3) S. sanguis (2) 5 20 1 2 11 11 3 1

Proteus mirabilis (5) Proteus vulgaris (5) Providencia rettgeri (1) Providencia stuartii (5) Salmonella spp. (13) Serratia marcescens (12) Acinetobacter baumannii (11) Acinetobacter junii (4) A. lwoffii (2)

S. capitis (1)

S. capitis (1) correct

E. faecium (1) E. faecium (1) correct S. dysgalactiae (1) S. dysgalactiae (1) correct S. parasanguis (1) S. parasanguis (1) correct Streptococcus S. constellatus gordonii (1) (1) incorrect Streptococcus sanguis (1) S. oralis (1) incorrect

Aeromonas 8 hydrophila (8) Aeromonas 2 sobria (2) A. faecalis (5) 5 Alcaligenes 16 xylosoxidans (16) Burkholderia 2 cepacia (2) Chryseobacterium 4 indologenes (4) Chryseobacterium 4 meningosepticum (4) Ochrobactrum 2 anthropi (4) P. aeruginosa (15) 15 P. fluorescens (2) 0

R. radiobacter (2)

R. radiobacter (2) correct

Pseudomonas putida (4)

P. stutzeri (1) R. radiobacter (3) Sphingobacterium multivorum (1) Candida albicans (24) Candida glabrata (9) Candida intermedia (1) Candida krusei (1) Candida parapsilosis (15) Candida tropicalis (6) Trichosporon asahii (2)
a

0 3 0 24 9 1 1 15 6 2

P. aeruginosa (1) correct, P. fluorescens (1) incorrect P. aeruginosa (1), P. aeruginosa P. fluorescens (2) (1) correct, P. fluorescens (2) incorrect S. paucimobilis (1) P. stutzeri (1) incorrect S. paucimobilis (1) S. paucimobilis (1) correct

P. aeruginosa (1), A. baumannii (1)

S. parasanguis (1) S. parasanguis (1) correct

Citrobacter freundii (7) 7 Citrobacter koseri (5) 5 Enterobacter 5 aerogenes (5) Enterobacter 6 cloacae (6) E. coli (12) 12 Klebsiella 5 oxytoca (5) K. pneumoniae (7) 7 Morganella 5 morganii (5)

Indicates the species identification (no. of cases) by the VITEK-2 but disagreed with the reference API identification. b Indicates the final identification with additional phenotypic testing and whether the ID results by the VITEK-2 was correct or incorrect.

I. Nakasone et al. / Diagnostic Microbiology and Infectious Disease 58 (2007) 191 198

193

resistances by the updated VITEK Advanced Expert System (AES) (Sanders et al., 2000; Barry et al., 2003).

2. Materials and methods 2.1. Isolates and testing by Vitek-2 Compact system A total of 474 clinical isolates comprising 235 of Grampositive cocci (GPC), 181 of GNR, and 58 yeasts were included in the ID study. In addition, a total of 321 clinical isolates including 96 strains of Enterobacteriaceae, 107 of staphylococci, 61 of enterococci, and 57 of S. pneumoniae were used for the detection of specific antimicrobial resistances. All the isolates were the collection of clinical isolates stored 80 8C and were subcultured twice onto the Columbia agar plates supplemented with 5% sheep blood before testing. The inoculum suspension was prepared in 0.45% saline, giving the equivalent of a 0.5-McFarland turbidity. The following respective VITEK test cards were filled with cell suspension according to the manufacturers instruction: GP for ID of GPC, GN for ID of GNR, YE for ID of yeasts, AST-N034 for AST of Enterobacteriaceae, AST-P546 for AST of staphylococci and enterococci, and AST-P518 for AST of S. pneumoniae. In this study, VITEK2 Compact system with the software version V2C 1.01 was used. 2.2. Reference methods All the isolates included in ID study were identified by the respective API test strips as follows: ID 32 STAPH for staphylococci, RAPID ID 32 STREP for streptococci and enterococci, RAPID ID 32 E for Enterobacteriaceae, ID 32 GN for glucose-nonfermentative GNR and members of the genus Aeromonas, and ID 32 C for yeasts. The API test strips were read by the autoreader, mini API (bioMerieux), and its database version 1.3.1. was used. When the VITEK2 gave the discrepant ID result compared with the respective API strip, additional phenotypic tests were performed according to the algorithm established (Ruoff, 2003; Schreckenberger and Wong; 2003). The flowcharts, based on Gram stain characteristics and a selected number of additional easily performed enzymatic, biochemical, and biologic tests, determined which ID results were correct. The specific antimicrobial resistances evaluated include extended-spectrum h-lactamase (ESBL)producing isolates of Escherichia coli and Klebsiella pneumoniae, oxacillin resistance and inducible clindamycin resistance among staphylococci, vancomycin resistance among enterococci, and penicillin resistance and erythromycin resistance among S. pneumoniae. As the reference, ESBL-producing isolates were first screened according to the initial disk screen test described (Clinical and Laboratory Standards Institute [CLSI], 2005, formerly National Committee for Clinical Laboratory Standards), then confirmed and classified on the basis of DNA amplification of the respective target genes by polymerase chain reaction (PCR) and agarose gel electro-

phoresis of the PCR products. Detection of the bla gene sequences coding the TEM, SHV, and CTX-M enzymes were performed as previously described using the specific primer pairs (Yagi et al., 2000). For oxacillin-resistant staphylococci, detection of penicillin-binding protein 2a (PBP2a) by the latex agglutination, MRSA-Screen test (Denka-Seiken, Tokyo, Japan), was used (Cavassini et al., 1999; Horstkotte et al., 2001). Inducible clindamycin resistance was first phenotypically determined by an erythromycin (15-Ag disk) clindamycin (2-Ag disk) double-disk test (D-zone test) and then was genetically confirmed for the presence of specific genes coding ermA and ermC by PCR as described (Khan et al., 1999; Volokhov et al., 2003). Vancomycin resistances among enterococci were determined by significant growth on vancomycin resistance enterococci screening agar plates (CLSI, 2003) and by PCR to detect vanA and vanB genes (Woodford et al., 1995). Penicillin resistance and erythromycin resistance of S. pneumoniae were determined by PCR for the respective genes of pbp1a, pbp2x, pbp2b, mefA, and ermB using the commercially available test reagents, penicillin-resistant S. pneumoniae gene detection version 2.0 (Wakunaga Pharmaceutical, Hiroshima, Japan) (Ubukata et al., 1996; Ubukata et al., 2003).

3. Results 3.1. ID of clinical isolates by the VITEK-2 Compact system Table 1 shows the results when the VITEK-2 and the respective API strips comparatively identified a total of 474 clinical isolates. Of 474 isolates tested, 454 (95.8%) were comparable IDs, resulting in 20 discrepant results comprising 9 of GPC and 11 of GNR. For the isolates belonging to Enterobacteriaceae and yeasts, all the ID results were completely identical to each other. For the

Fig. 1. Cumulative distribution of time to require for final ID by the VITEK-2 Compact system. = staphylococci; EE = enterococci; zz = streptococci; oo = Enterobacteriaceae; 55 = glucosenonfermentative GNR. The isolates of Aeromonas spp. were included in glucose-nonfermentative GNR.

!!

194

I. Nakasone et al. / Diagnostic Microbiology and Infectious Disease 58 (2007) 191 198

Table 2 Accuracy to detect specific resistant isolates by the VITEK AES Antimicrobial resistance and clinical isolates tested Nos. of isolates tested Interpretation by VITEK AES Enterobacteriaceae ESBL Non-ESBL Staphylococci MLSB inducible MLSB inducible negative Modification of PBP Nonmodification of PBP

ESBLs E. coli (TEM type) E. coli (SHV type) E. coli (CTX-M type) K. pneumoniae (TEM type) K. pneumoniae (SHV type) K. pneumoniae (TEM and SHV type) Non-ESBL E. coli K. pneumoniae D-zone testpositive and ermA and/or ermC-positive staphylococci S. aureus S. epidermidis S. haemolyticus S. hominis D-zone testnegative and ermA- and ermC-negative S. aureus Oxacillin-resistant and PBP2a-positive staphylococci S. aureus S. capitis S. epidermidis S. haemolyticus S. hominis S. lugdunensis S. simulans S. warneri Oxacillin-susceptible and PBP2a-negative staphylococci S. aureus S. capitis S. hominis S. lugdunensis S. warneri Antimicrobial resistance and clinical isolates tested

21 2 3 21 1 3

20 2 3 21 1 3

24 21

24 21

33 8 3 2 16

32 8 3 2

16

27 4 30 3 2 1 1 1

27 4 30 3 2 1 1 1

22 8 2 5 1 Nos. of isolates tested Interpretation by VITEK AES Enterococci Vancomycin resistance VanA like VanB like Wild (VanC) S. pneumoniae Modification of PBP High-level resistance Low-level resistance

22 8 2 5 1

Resistant (MLSB)

Wild

vanA-positive enterococci E. casseliflavus E. faecalis E. faecium E. gallinarum vanB-positive enterococci E. casseliflavus E. faecalis E. faecium E. gallinarum vanA- and vanB-positive enterococci

1 1 18 1 3 14 11 1

1 1 18 1 3 14 11 1

I. Nakasone et al. / Diagnostic Microbiology and Infectious Disease 58 (2007) 191 198 Table 2 (continued) Antimicrobial resistance and clinical isolates tested Nos. of isolates tested Interpretation by VITEK AES Enterococci Vancomycin resistance VanA like E. faecium E. gallinarum vanA- and vanB-negative enterococci E. casseliflavus E. gallinarum Penicillin-resistant mutant S. pneumoniae Penicillin-susceptible, wild S. pneumoniae Erythromycin-resistant mutant S. pneumoniae Erythromycin-susceptible, wild S. pneumoniae 1 3 1 3 VanB like Wild (VanC) S. pneumoniae Modification of PBP High-level resistance Low-level resistance Resistant (MLSB)

195

Wild

1 6

1 6

40 17

28

12 2 15

44

42

13

13

discrepant results, additional phenotypic characterizations were performed and determined, in which ID result was correct. Of the 9 discrepant results for GPC, 7 ID results by VITEK-2 were correct, but the API ID 32 STAPH and RAPID ID 32 STREP resulted in misidentifications. Two isolates of Staphylococcus aureus, which were identified by ID 32 STAPH, resulted in the ID as Staphylococcus caprae by VITEK-2, and the IDs of VITEK-2 consisted of negative coagulase, negative clumping factor, and positive urease results. Also, one isolate was identified as Staphylococcus simulans by ID 32 STAPH, but VITEK-2 identified it as Staphylococcus capitis. Additional biochemical testing, acid from lactose, positive urease, and negative h-galactosidase supported the VITEK-2 ID result. One isolate of Enterococcus gallinarum, which was identified by RAPID ID 32 STREP, resulted in Enterococcus faecium by VITEK-2, and nonmotility and not producing yellow pigment simply supported the VITEK-2 ID result. There were 5 discrepant results for streptococci. Of these, 3 results by VITEK-2, 1 of Streptococcus dysgalactiae and 2 of Streptococcus parasanguis, were regarded as being correct IDs by the additional biochemical testing: negative CAMP test and nonfermentation of sorbitol, and positive h-glucosidase and fermentation of trehalose for S. dysgalactiae, and negative VogesProskauer test, and positive h-glucosidase and h-galactosidase for S. parasanguis. Finally, the VITEK-2 correctly identified 233 (99.1%) isolates of GPC, compared with 228 (97.0%) isolates by the reference API strips, ID 32 STAPH and RAPID ID 32 STREP. For GNR, a total of 11 discrepant results were obtained by the VITEK-2, and all the discrepant results came from the isolates of glucose-nonfermentative bacteria. Of these, 5 ID results by VITEK-2 were finally concluded as being correct: 2 isolates of Rhizobium radiobacter with negative

gas production from nitrate, positive nitrate reduction, and positive O-nitrophenyl-h-d-galactopyranoside; 2 isolates of Pseudomonas aeruginosa with significant growth at 428C, positive hydrolysis of acetamide, and gas production from nitrate; and 1 isolate of Sphingomonas paucimobilis with positive motility, positive urease, and susceptibility to polymyxin B. However, there remained 6 misidentifications confirmed, and they were 2 isolates of Acinetobacter lwoffii, 3 isolates of Pseudomonas fluorescens, and 1 isolate of Pseudomonas stutzeri. Overall, the VITEK-2 correctly identified 175 (96.7%) isolates of GNR, compared with 176 (97.2%) isolates by the reference API, RAPID ID 32 E and ID 32 GN. Fig. 1 indicated the cumulative distribution of time to require for final ID by the VITEK-2 Compact system for the respective bacterial groups. For all the groups of the isolates tested, N50% of ID results were obtained within 4 to 6 h, and all the final ID results were completed after 7- to 10-h incubation cycles. 3.2. Detection of specific antimicrobial resistances by the AES The interpretation results by the VITEK AES to detect specific antimicrobial resistances were summarized in Table 2. Of 51 ESBL-producing isolates of E. coli and K. pneumoniae, the VITEK AES correctly identified 50 isolates (98.0%) and missed 1 isolate of TEM-type ESBL-positive E. coli, which was interpreted as an acquired penicillinase plus cephalosporinase-producing isolate. However, all the ESBL-nonproducing isolates were correctly interpreted as being non-ESBL isolates, results indicating 98.0% sensitivity and 100% specificity. Also, the VITEK AES produced highly correlative interpretations to detect positive D-zone test associated with macrolide, lincosamide,

196

I. Nakasone et al. / Diagnostic Microbiology and Infectious Disease 58 (2007) 191 198

incorrectly interpreted as low-level resistance with modification of PBP. Also, all the 13 wild susceptible isolates of S. pneumoniae for the erythromycin resistance determinant genes were correctly identified as bwildQ, but 2 isolates positive for ermB gene resulted in incorrect interpretations, bwildQ. Fig. 2 indicated the cumulative distribution of time to detect antimicrobial resistances by the VITEK AES. More than 50% of all the resistant isolates were interpreted within 7 to 9 h of incubation cycles. The testing of GPC, in general, required longer incubation period; however, most resistant isolates, including vancomycin-resistant enterococci and oxacillin-resistant staphylococci, were correctly determined within 12 h of incubations.
Fig. 2. Cumulative distribution of time to detect the respective antimicrobial resistance by the VITEK AES. = ESBL-producing isolate; nn = inducible clindamycin-resistant staphylococci positive for D-zone test and ermA and/or ermC genes; EE = oxacillin-resistant staphylococci positive for PBP2a; zz = vancomycin-resistant enterococci; oo = penicillin-resistant S. pneumoniae; 55 = erythromycin-resistant S. pneumoniae.

!!

4. Discussion Overall, the evaluation results of the newly redesigned colorimetric VITEK-2 ID impressed us by the performance because more than 98% of the isolates were correctly identified to the species level without any further additional tests. Also, our obtained results indicated that the current VITEK-2 has overcome its inherent weakness in IDs of streptococci and glucose-nonfermentative GNR. Until present, API test strips has been long considered as the bgold standardQ in ID test (Fortin et al., 2003; Aubertine et al., 2006), but the accuracy of the VITEK-2 was finally estimated to be 98.3%, compared with 97.5% by the respective API test strips. Our obtained results were highly consistent with a series of evaluation results recently published for GPC (Funke and Funke-Kissling, 2005), GNR (Funke and Funke-Kissling, 2004), and yeast (Aubertine et al., 2006). When compared with the previous VITEK ID based on fluorescent technology, the current VITEK-2 ID broadened the database concerning the relativity to reaction tests and taxa identified; the expanded database corresponds to GP containing 43 tests for 115 taxa, GN containing 47 tests for 159 taxa, and YE containing 46 tests for 53 species and 14 genera. Although our evaluation study did not cover all the taxa included in the database and very small numbers of the clinical isolates for some species were included, it became apparent that the extension of the database led significantly improved ID accuracy rather than poorer ID results. However, there are still several misidentifications in the results of isolates belonging to the family Streptococcaceae and glucosenonfermentative GNR: 1 isolate each of Streptococcus constellatus and Streptococcus oralis in GP, 2 isolates of A. lwoffii, 3 of P. fluorescens, and 1 of P. stutzeri in GN. Both bacterial groups are taxonomically diverse and are still problematic in phenotypic ID tests. Gene sequence analysis of 16S ribosomal DNA or RNA demonstrates considerable heterogeneity, and the original classification of the genera has undergone extensive revision (Kawamura et al., 1995; Kersterns et al., 1996). Although the current VITEK-2 also

and type B streptogramin (MLSB) resistance coded by ermA and/or ermC genes and oxacillin resistance positive for PBP2a among staphylococci. Of the 46 isolates of staphylococci positive for D-zone test and also positive for ermA and/or ermC genes, 45 (97.8%) were correctly interpreted as bMLSB inducibleQ by VITEK AES, whereas all the 16 isolates, which were negative in the respective reference tests, were reported as bMLSB inducible negativeQ. One remaining isolate positive for both D-zone test and ermA gene was reported as being constitutively resistant to macrolide and streptogramin. A total of 69 oxacillinresistant staphylococcal isolates positive for PBP2a and 38 oxacillin-susceptible isolates negative for PBP2a were tested by VITEK-2, the AES interpretations giving none of discrepant result with the reference tests. Some significant discrepant results by the VITEK AES were demonstrated in detecting vancomycin resistance among enterococci. All the 25 isolates positive for vanA gene, including 4 isolates positive for both vanA and vanB, were correctly identified as bVanA-likeQ. Also, 25 vanBpositive isolates of Enterococcus faecalis and E. faecium gave consistent interpretations. However, 3 isolates of Enterococcus casseliflavus and 1 of E. gallinarum positive for vanB gene were incorrectly identified as bwild (VanC)Q by the VITEK AES. The MICs determined by the VITEK-2 were 32 Ag/mL (2 strains) and 8.0 Ag/mL (1 strain) for E. casseliflavus and 32 Ag/mL for E. gallinarum. For the penicillin resistance among S. pneumoniae, all the isolates with mutations on penicillin-binding protein genes, pbp1a, pbp2x, and/or pbp2b, were correctly identified as bmodification of PBPQ with either high-level or low-level resistances. However, 2 of 17 wild susceptible isolates were

I. Nakasone et al. / Diagnostic Microbiology and Infectious Disease 58 (2007) 191 198

197

has inherent limitations in the ID of the abovementioned groups, addition of biochemical reactions and broadened database enable us to provide more accurate ID results comparable with the up-to-date taxonomy. The VITEK AES has been created to analyze the AST results using the well-established knowledge base of approximately 100 species and 20 000 ranges of MIC to detect more than 2300 phenotypic antimicrobial resistances. VITEK-2 and AES have been evaluated in several countries (Sanders et al., 2000; Barry et al., 2003), the results described indicating that the AES detected and interpreted resistance mechanisms appropriately and would provide aids for therapeutic choice and accurate epidemiologic analysis. In our evaluation, we included 6 clinically important antimicrobial resistances well characterized by genetic and phenotypic reference methods. The accuracy of AES was estimated to be 97.7%, and 10 isolates, comprising 1 isolate each of TEM-type ESBL-producing E. coli and inducible MLSB-resistant S. aureus, 2 isolates each of penicillinsusceptible wild S. pneumoniae and erythromycin-resistant S. pneumoniae, and 4 isolates of VRE, were misidentified. In particular, all the 4 isolates of E. casseliflavus and E. gallinarum positive for vanB gene were interpreted as wild (VanC). AES uses enterococcal ID and MICs against vancomycin and teicoplanin to characterize VRE isolates. The MIC results for the above 4 isolates were interpreted to be resistant to vancomycin and susceptible to teicoplanin similar to the other vanB-positive E. faecalis and E. faecium. However, when the VITEK identifies the test isolate as being E. casseliflavus or E. gallinarum, the VITEK AES automatically reports the message bwildQ (VanC). Both VanA and VanB phenotypes are most commonly detected in E. faecalis and E. faecium but have been found in other species (Clark et al., 1993). Also, discrepancy between VanB phenotype and vanA genotype was recently reported (Song et al., 2006). Thus, refinements in AES algorithm should be urgent to improve accuracy to detect a variety of VRE phenotypes.

VITEK-2 combined with AES will greatly contribute to laboratory function in the field of clinical microbiology. Acknowledgments The authors thank Miyako Higa and Fusako Furugen for their valuable technical assistance and Yukiko Izumi for her help in the preparation of the manuscript.

References
Aubertine CL, Rivera M, Rohan SM, Larone DH (2006) Comparative study of the new colorimetric VITEK 2 yeast identification card versus the older fluorometric card and of CHROMagar Candida as a source medium with the new card. J Clin Microbiol 44:227 228. Barry J, Brown A, Ensor V, Lakhani U, Petts D, Warren C, Winstanley T (2003) Comparative evaluation of the VITEK 2 Advanced Expert System (AES) in five UK hospitals. J Antimicrob Chemother 51: 1191 1202. Cavassini M, Wenger A, Jaton K, Blanc DS, Bille J (1999) Evaluation of MRSA-Screen, a simple anti-PBP 2a slide latex agglutination kit, for a rapid detection of methicillin resistance in Staphylococcus aureus. J Clin Microbiol 37:1591 1594. Clark NC, Cooksey RC, Hill BC, Swenson JM, Tenover FC (1993) Characterization of glycopeptide-resistant enterococci from U.S. hospitals. Antimicrob Agents Chemother 37:2311 2317. Clinical and Laboratory Standards Institute (2003) Methods for dilution antimicrobial susceptibility tests for bacteria that grow aerobically sixth edition M7-A6. Wayne, (PA)7 CLSI. Clinical and Laboratory Standards Institute (2005) Performance standards for antimicrobial susceptibility testing fifteenth edition M100-S15. Wayne, (PA)7 CLSI. Fortin M, Messier S, Pare J, Higgins R (2003) Identification of catalasenegative, nonbeta-hemolytic, Gram-positive cocci isolated from milk samples. J Clin Microbiol 41:106 109. Funke G, Funke-Kissling P (2004) Evaluation of the new VITEK 2 card for identification of clinically relevant Gram-negative rods. J Clin Microbiol 42:4067 4071. Funke G, Funke-Kissling P (2005) Performance of the new VITEK 2 GP card for identification of medically relevant Gram-positive cocci in a routine clinical laboratory. J Clin Microbiol 43:84 88. Gavin PJ, Warren JR, Obias AA, Collins SM, Peterson LR (2002) Evaluation of the Vitek 2 system for rapid identification of clinical isolates of Gram-negative bacilli and members of the family Streptococcaceae. Eur J Clin Microbiol Infect Dis 21:869 874. Horstkotte MA, Knobloch JKM, Rohde H, Mack D (2001) Rapid detection of methicillin resistance in coagulase-negative staphylococci by a penicillin-binding protein 2a-specific latex agglutination test. J Clin Microbiol 39:3700 3702. Joyanes P, del Carmen Conejo M, Martinez-Martinez L, Perea E (2001) Evaluation of the VITEK 2 system for the identification and susceptibility testing of three species of nonfermenting Gram-negative rods frequently isolated from clinical samples. J Clin Microbiol 39:3247 3253. Kawamura Y, Hou X-G, Sultana F, Miura H, Ezaki T (1995) Determination of 16S rRNA sequences of Streptococcus mitis and Streptococcus gordonii and phylogenetic relationships among members of the genus Streptococcus. Int J Syst Bacteriol 48:921 927. Kersterns K, Ludwig W, Vancanneyt M, De Vos P, Gillis M, Schleifer KH (1996) Recent changes in the classification of the pseudomonads: an overview. Syst Appl Microbiol 19:465 477. Khan SA, Nawaz MS, Khan AA, Cerniglia CE (1999) Simultaneous detection of erythromycin-resistant methylase genes ermA and ermC

5. Conclusions The colorimetric VITEK-2 Compact system achieved an excellent performance to provide accurate species ID results. In the evaluation, 466 (98.3%) of 474 clinical isolates, including a variety of species of staphylococci, enterococci, streptococci, Enterobacteriaceae, glucose-nonfermentative GNR, and yeast, were correctly identified. Also, the VITEK AES provided interpretations comparable with phenotypic and genotypic characterizations to determine specific antimicrobial resistances such as ESBL, inducible MLSB- and oxacillin-resistant staphylococci, VRE, and penicillin- and erythromycin-resistant S. pneumoniae. Overall, the accuracy to detect antimicrobial resistances evaluated was estimated to be 431 of 440 (98.0%). Although our study has limitations on the taxa and on the numbers of the isolates included, it can be concluded that the current colorimetric

198

I. Nakasone et al. / Diagnostic Microbiology and Infectious Disease 58 (2007) 191 198 hospitals. Diagn Microbiol Infect Dis [Electronic Publication ahead of print]. Ubukata K, Asahi Y, Yamane A, Konno M (1996) Combinational detection of autolysin and penicillin-binding protein 2B genes of Streptococcus pneumoniae by PCR. J Clin Microbiol 34:592 596. Ubukata K, Iwasa S, Sunakawa K (2003) In vitro activities of new ketolide, telithromycin, and eight other macrolide antibiotics against Streptococcus pneumoniae having mefA and ermB genes that mediate macrolide resistance. J Infect Chemother 9:221 226. Volokhov D, Chizhikov V, Chumakov K, Rasooly A (2003) Microarray analysis of erythromycin resistance determinants. J Appl Microbiol 95:787 798. Woodford N, Johnson AP, Morrison D, Speller DC (1995) Current perspectives on glycopeptide resistance. Clin Microbiol Rev 8:585 615. Yagi T, Kurokawa H, Shibata N, Shibayama K, Arakawa Y (2000) A preliminary survey of extended-spectrum h-lactamases (ESBLs) in clinical isolates of Klebsiella pneumoniae and Escherichia coli in Japan. FEMS Microbiol Lett 184:53 56.

from Staphylococcus spp. by multiplex-PCR. Mol Cell Probes 13:381 387. Ruoff KL (2003) Algorithm for identification of aerobic Gram-positive cocci. In: Manual of Clinical Microbiology. (8th ed.). Eds, PR Murray, EJ Baron, JH Jorgensen, MA Pfaller, and RH Yolken. Washington (DC)7 American Society for Microbiology, pp. 331 333. Sanders CC, Peyret M, Moland ES, Shubert C, Thomson KS, Boeufgras J-M, Sanders WE, Jr (2000) Ability of the VITEK 2 Advanced Expert System to identify h-lactam phenotypes in isolates of Enterobacteriaceae and Pseudomonas aeruginosa. J Clin Microbiol 38:570 574. Schreckenberger PC, Wong JD (2003) Algorithm for identification of aerobic Gram-negative bacteria. In: Manual of Clinical Microbiology. (8th ed.). Eds, PR Murray, EJ Baron, JH Jorgensen, MA Pfaller, and RH Yolken. Washington (DC)7 American Society for Microbiology, pp. 337 342. Song JH, Ko KS, Oh WA, Park S, Heo ST, Kwon KT, Ryu Sy, Peck KR, Lee NY (2006) High frequency of vancomycin-resistant Enterococcus faecium isolates with VanB phenotype and vanA genotype in Korean