Received 21 April 1997/Returned for modification 30 June 1997/Accepted 28 August 1997

JOURNAL OF CLINICAL MICROBIOLOGY, Dec. 1997, p. 3258 3263 Vol. 35, No. 12 0095-1137/97/$04.00 0 Copyright 1997, American Society for Microbiology Comparison of Agar Dilution, Broth Microdilution, E-Test, Disk Diffusion, and Automated Vitek Methods for Testing Susceptibilities of Enterococcus spp. to Vancomycin PEGGY C. KOHNER, 1 ROBIN PATEL, 2 JAMES R. UHL, 1 KAY M. GARIN, 3 MARLENE K. HOPKINS, 1 LEE T. WEGENER, 3 AND FRANKLIN R. COCKERILL III 1,2 * Division of Clinical Microbiology, Department of Laboratory Medicine, 1 and Division of Infectious Diseases, Department of Internal Medicine, 2 Mayo Clinic and Foundation, Rochester, Minnesota 55905, and Department of Laboratory Medicine and Pathology, Mayo Clinic and Foundation, Scottsdale, Arizona 85259 3 Received 21 April 1997/Returned for modification 30 June 1997/Accepted 28 August 1997 An evaluation was undertaken to determine the optimal method for testing the susceptibilities of 100 clinical isolates and two reference strains of Enterococcus spp. to vancomycin in vitro. Six testing methods were studied by using the following media and incubation times: agar screen with the Synergy Quad Plate (Remel, Lenexa, Kans.), an in-house-prepared brain heart infusion () agar plate, and an in-house-prepared Mueller-Hinton () agar plate, all incubated for 24 or 48 h; broth microdilution (Sensititre Just One Strip; AccuMed International, Inc., West Lake, Ohio) with or cation-adjusted broth incubated for 24 or 48 h; agar dilution with or agar incubated for 24 or 48 h; epsilometer test (E test; AB BioDisk, Solna, Sweden) with or agar incubated for 24 or 48 h; disk diffusion with or agar incubated for 24 or 48 h; and the automated Vitek method with the gram-positive susceptibility Staphylococcus aureus card and R02.03 software (biomerieux, Inc., Hazelwood, Mo.). Growth failures occurred with media (n 6) but not with media. One growth failure occurred with the Vitek method. Results for each testing method for each Enterococcus strain were interpreted as susceptible, intermediate, or resistant according to current National Committee for Clinical Laboratory Standards (NCCLS) criteria and compared to the vancomycin resistance genotype (i.e., vana, vanb, vanc-1, or vanc-2/3). For all methods, extension of the incubation time from 24 h to 48 h either produced no difference in the results or gave poorer results. The following methods produced no very major or major interpretive errors: broth microdilution with media incubated for 24 h, agar dilution with media incubated for 24 or 48 h, and E test with media incubated for 24 or 48 h. Unacceptable frequencies of very major errors (>1%) occurred with all methods for which media were used. Minor interpretive errors were frequent with all methods. These minor interpretive errors also occurred most frequently with Enterococcus strains with vanc genes, which encoded low-level vancomycin resistance (MIC < 8 g/ml), as opposed to Enterococcus strains which possessed vana or vanb genes, which encoded higher-level vancomycin resistance (MIC > 64 g/ml). Modification of NCCLS breakpoints, especially for motile Enterococcus spp. (E. casseliflavus, E. flavescens, and E. gallinarum), may resolve this problem; however, in the current study, one E. faecalis strain and one E. faecium strain carried only the vanc gene. The agar screen method may also require reformulation. The current agar screen plate contains 6 g of vancomycin per ml, which may not detect all low-level resistance associated with vanc genotypes. Nevertheless, the clinical significance of this low-level vancomycin resistance remains unknown. * Corresponding author. Mailing address: Division of Clinical Microbiology, Mayo Clinic, Hilton 460C, 200 First St. SW, Rochester, MN 55905. Phone: (507) 284-2901. Fax: (507) 284-4272. E-mail: cock erill.franklin@mayo.edu. Identifying vancomycin-resistant Enterococcus spp. (VRE) is important so that patients receive appropriate antimicrobial therapy and nosocomial outbreaks can be prevented or controlled. We undertook this study to determine the optimal method for testing the susceptibilities of clinical isolates of Enterococcus spp. to vancomycin in vitro. We compared the performance characteristics of three agar screen plates (Synergy Quad Plate [Remel, Lenexa, Kans.] and two agar screen plates prepared in-house), agar dilution, broth microdilution (Sensititre Just One Strip; AccuMed International, Inc., Westlake, Ohio), the epsilometer test (E test; AB BioDisk, Solna, Sweden), disk diffusion, and the automated Vitek method (biomerieux, Inc., Hazelwood, Mo.). For all methods, excluding the Synergy Quad Plate and the Vitek methods, two different media, Mueller-Hinton () or brain heart infusion (), were used. Two different incubation periods, 24 and 48 h, were studied for all methods except the Vitek method. For the Vitek method, the gram-positive susceptibility Staphylococcus aureus (GPS-SA) cards, software version R02.03, and a minimum incubation time of 6 h were used. The susceptibility results for each of the 100 clinical isolates and two reference strains of Enterococcus spp. produced by these methods were interpreted according to 1995 National Committee for Clinical Laboratory Standards (NCCLS) guidelines (6) and compared to the vancomycin resistance genotype. Vancomycin resistance genotypes for each strain were determined by a multiplex PCR method developed in our laboratory (8) to detect the following genes which have been associated with glycopeptide resistance: vana, vanb, vanc1, vanc2, and vanc3. (Data from this paper were presented in part at the 96th General Meeting of the American Society for Microbiology, New Orleans, La., 19 to 23 May 1996. [poster C175]) 3258

VOL. 35, 1997 TESTING SUSCEPTIBILITIES OF ENTEROCOCCI TO VANCOMYCIN 3259 TABLE 1. Identification and van genes and phenotypes of Enterococcus clinical isolates a,b Gene(s) detected c No. of isolates Enterococcus spp. (no. of isolates) MIC ( g/ml) Several studies have evaluated the abilities of commercial and reference antimicrobial susceptibility testing methods to detect vancomycin resistance in Enterococcus isolates (3, 9 16). Those studies which have evaluated current versions of automated or semiautomated commercial systems, such as the Vitek system (biomerieux, Inc.), the MicroScan Rapid system (Baxter Health Care Corp., West Sacramento, Calif.), or the Alamar MIC system (Alamar Biosciences, Sacramento, Calif.), have shown that these methods do not reliably detect vancomycin resistance in enterococci (11, 13, 15, 16). Reasons for the poor performance of these systems likely include the composition of growth media and the relatively short incubation times that are used. Most conventional broth- and agar-based manual methods require a full 24 h of incubation, and cur- Vancomycin Teicoplanin vana 10 E. faecium (8) 256 16 (2) 256 16 vana vanc-1 1 E. gallinarum (1) 256 16 vanb 30 E. faecium (2) 256 16 (1) 256 16 (12) 256 8 (10) 256 8 (3) 128 8 E. faecalis (1) 128 8 (1) 64 8 vanc-1 12 E. gallinarum (5) 8 8 (5) 4 8 E. faecium (1) 8 8 E. faecalis (1) 4 8 vanc-2/3 6 E. casseliflavus/ flavescens (6) PCR amplicon produced but with distinct restriction fragment length pattern d 4 8 4 E. faecalis (1) 256 8 (1) 256 8 E. casseliflavus/ flavescens (2) 4 8 None 37 E. casseliflavus/ 4 8 flavescens (2) E. faecalis (28) 2 8 (1) 4 8 E. faecium (4) 2 8 E. avium (1) 2 8 E. raffinosus (1) 2 8 a Modified from Table 2 in reference 8. b Species designations for all clinical isolates were previously determined by standard biochemical techniques, and MICs were determined by an agar dilution technique (8). c The multiplex PCR used to determine van genes will not discriminate the vanc-2 gene from the vanc-3 gene (8), which is therefore designated vanc-2/3. d DNA sequencing of the amplicons for the two E. faecalis isolates revealed nucleic acid sequences closely related to the published sequence for the vanb gene (1, 2). DNA sequencing of the amplicons for the two E. casseliflavus/ flavescens isolates revealed nucleic acid sequences closely related to the published sequences for vanc-2 and vanc-3 genes (7). MATERIALS AND METHODS One hundred Enterococcus clinical isolates (34 E. faecalis, 43E. faecium, 11 E. gallinarum, 10E. casseliflavus/flavescens, 1E. avium, and 1 E. raffinosus [Table 1]) and 2 reference strains (E. faecalis ATTC 29212 and ATTC 51299) were evaluated for their susceptibilities to vancomycin by different susceptibility test methods. In Table 1, species designations and vancomycin and teicoplanin MICs are shown for the Enterococcus clinical isolates. Species designations for these clinical isolates were previously determined by standard biochemical techniques, and MICs were previously determined by using media and an agar dilution technique (8). For the current study, the test media used, including whether the media were prepared in-house or commercially prepared, inocula, incubation parameters, and susceptibility interpretive guidelines are shown in Table 2. Except for the commercial methods, available NCCLS guidelines were followed for medium and bacterial inoculum preparation for agar dilution ( agar) (5), broth microdilution (cation-adjusted broth [CA broth]) (5) and disk diffusion ( agar) (4). The results for each method were also interpreted according to NCCLS guidelines (6). Therefore, the breakpoints used for interpretive categories were the following: susceptible, 4 g/ml, intermediate, 8 to 16 g/ml, and resistant, 32 g/ml. All Enterococcus isolates were previously surveyed for genes associated with vancomycin resistance by using a multiplex PCR and restriction fragment length polymorphism (RFLP) analysis previously described by us (8). These results are shown in Table 1. By this PCR-RFLP method, the vana, vanb, vanc1, vanc2, and vanc3 genes are identified; however, the vanc2 and vanc3 genes, although amplified, cannot be discriminated from one another. We therefore designate the vanc2 and vanc3 genes vanc-2/3. The susceptibility results for all Enterococcus strains for all methods were compared to the vancomycin genotypes for these strains. RESULTS Table 3 shows the susceptibility testing results compared to the vancomycin genotype for individual Enterococcus species. There were no growth failures with media. Furthermore, there appeared to be more luxuriant growth with media, making the results easier to interpret. Six strains failed to grow on media incubated for either 24 h or 48 h (two with microbroth dilution, one with agar dilution, two with E test, and one with disk diffusion). One growth failure occurred with the Vitek method. For all methods, extension of the incubation time from 24 h to 48 h either produced no difference in the results or gave poorer results. Table 4 summarizes all of the interpretive errors for each method for all of the Enterococcus spp. combined. The following methods produced no very major (resistant strains misinterpreted as susceptible, i.e., false susceptibility) or major (susceptible strains misinterpreted as resistant, i.e., false resistance) interpretive errors when the results were compared to those for the vancomycin resistance genotype: broth microdilution with media incubated for 24 h, agar dilution with media incubated for 24 or 48 h, and E test with media incubated for 24 or 48 h. When the results for the two agar screen methods that used media (Synergy Quad Plate or the in-house-prepared agar plate) were compared to the vancomycin resistance genotype, very major errors ( 2%) occurred only with Enterococcus spp. that carried vanc genes. Unacceptably high frequencies ( 1%) of very major errors occurred when results for disk diffusion with agar or the Vitek method were compared to the vancomycin genotype. Unacceptable frequencies of very major errors ( 1%) occurred with all methods for which media were used. Minor interpretive errors (susceptible or resistant strains misinterpreted as intermediately susceptible) were frequent with all methods. These minor interpretive errors also occurred most frequently with Enterococcus isolates that contained vanc genes which encoded low-level vancomycin resistance (MIC 8 g/ml) as opposed to those isolates which carried the vana or vanb gene, both of which encoded highlevel vancomycin resistance (MIC 64 g/ml) (Tables 1 and 4). DISCUSSION

3260 KOHNER ET AL. J. CLIN. MICROBIOL. rently, the NCCLS recommends media for the agar screen method (5, 6). The reliability of the agar screen method for detecting highlevel vancomycin resistance in Enterococcus spp. has been demonstrated. By using 6 g of vancomycin per ml of agar incubated for a full 24 h, Swenson and colleagues (10) demonstrated that this method was 96 to 99% sensitive and 100% specific for the detection of vancomycin resistance among 158 clinical isolates of Enterococcus spp. tested at eight different medical centers. The reliability of other conventional methods for detecting vancomycin resistance in Enterococcus spp., i.e., disk diffusion, agar dilution, and the recently introduced diffusion gradient method, the E test, is less clear (9, 11). However, problems appear to be most frequent when low-level vancomycin resistance, such as occurs with resistance encoded by the vanc gene group, is encountered. These methods have been studied with media as recommended by the NCCLS (4 6) instead of media as recommended by the NCCLS for the agar screen method (5, 6). One limitation of all of the above studies may be the inconsistent definition for the gold standard or reference susceptibility result. Some investigators have compared test results of the study method to a broth dilution result (11, 13), while others have compared the results of the study method to results obtained by using either agar dilution (9, 14), disk diffusion (15), the Micro Scan Rapid method (15, 16), or E-test method (12) or by comparing results to the vancomycin resistance genotype (3, 10, 11). Furthermore, for two of these studies, evaluations were limited to Enterococcus isolates with high-level vancomycin resistance (3, 12). For the current study, we compared the results of the test methods to the vancomycin resistance genotype. We also compared the performance characteristics of media and the NCCLS-recommended media for detection of vancomycin resistance by the agar dilution or disk diffusion method or CA media (4 6) by the microbroth dilution method. Furthermore, we evaluated the effects of 24 versus 48 h of incubation. The results of our study suggest that the optimal phenotypic method for determining vancomycin resistance in Enterococcus spp. is any one of the following: broth dilution with media incubated for 24 h, agar dilution with media incubated for 24 h, or the E test with media incubated for 24 h. Poorer results occurred for all of the methods when media were used. In all cases in which the incubation periods were extended to 48 h, either no difference or poorer results were obtained. Both the disk diffusion and Vitek methods produced unacceptably high numbers of very major errors (i.e., exceeding 1.0%). Growth failures did not occur with media for any method. In contrast, six growth failures occurred with media incubated for 24 or 48 h (two with microbroth dilution, one with agar dilution, two with E test, and one with disk diffusion). Moreover, the growth of Enterococcus spp. on agar plates (in-house-prepared agar screen plates, agar dilution, E test, and disk diffusion) was more luxuriant and therefore easier to read than agar plates. All of the test methods produced considerable minor errors (minor errors could not be calculated for agar screen plates when results were compared to the vancomycin resistance genotype). These minor errors could be lessened if NCCLS breakpoints for the interpretive categories, susceptible, intermediately susceptible, and resistant, are changed, especially for the Enterococcus spp. with the vanc genes that encode lowerlevel vancomycin resistance. However, the clinical significance of this lower-level resistance remains unknown. One approach might be to identify motile Enterococcus spp. and then apply TABLE 2. Test methods used in this study a Bacterial inoculum Interpretation c Test materials (commercial Vancomycin source when applicable) b concn(s) tested Method Agar screen 1. Synergy Quad Plate ( agar) (Remel) 6 g/ml 10 6 CFU/spot d Resistance defined as 1 colony 2. agar 3. agar 50 l of10 3 dilution of 0.5 Resistant, 32 g/ml; intermediate, McFarland standard d 8 16 g/ml; susceptible, 4 g/ml Enterococcus breakpoint strip: 16, 4 g/ml Broth microdilution 1. Sensititre Just One Strip (AccuMed International, Inc.) with broth 2. Sensititre Just One Strip with CA broth d Agar dilution 1. agar 16 to 2 g/ml 10 4 CFU/spot d Resistant, 32 g/ml; intermediate, 2. agar d 8 16 g/ml; susceptible, 4 g/ml Antimicrobic gradient 1. Epsilometer test (E test; AB BIODISK) with agar 256 to 0.016 g/ml 0.5 McFarland standard Resistant, 32 g/ml; intermediate, 2. E test with agar 8 16 g/ml; susceptible, 4 g/ml Disk diffusion 1. agar 30- g disk 0.5 McFarland standard e Resistant, 14 mm; intermediate, 2. agar e 15 16 mm; susceptible, 17 mm Automated 1. Vitek (biomerieux) with GPS-SA and R02.03 software 32 to 0.5 g/ml a For all methods excluding the Vitek method, incubation was carried out at 35 C without added CO2. b refers to brain heart infusion, refers to Mueller-Hinton, and CA refers to cation-adjusted. All media with the exception of the Synergy Quad Plate and Vitek Card were prepared in-house at the Mayo Clinical Microbiology Laboratory. c For all methods excluding the Vitek method, the results were interpreted at 24 and 48 h, according to NCCLS standard M100-S6 (6). For disk diffusion, the diameters of zones of inhibition are shown. d Media and/or bacterial inoculum prepared according to NCCLS standard M7-A3 (5). e Media and/or bacterial inoculum prepared according to NCCLS standard M2-A5 (4).

TABLE 3. Interpretive discrepancies by the various test methods for different Enterococcus strains with the vancomycin resistance genotype as the reference standard Interpretive discrepancy b by method: Enterococcus species (no. of strains) Type of interpretive discrepancy a Synergy Quad Plate () Agar screen Broth microdilution (Sensititre) Agar dilution E Test Disk diffusion Vitek 24 h 48 h 24 h 48 h 24 h 48 h 24 h 48 h 24 h 48 h 24 h 48 h 24 h 48 h 24 h 48 h 24 h 48 h 24 h 48 h 24 h 48 h E. faecalis (34) Very major 0/5 (0) 0/5 (0) 0/5 (0) 0/5 (0) 0/5 (0) 1/5 (20) 0/5 (0) 0/5 (0) 1/5 (20) 1/5 (20) Major 0/29 (0) 0/29 (0) 0/29 (0) 0/29 (0) 0/29 (0) 0/29 (0) 0/29 (0) 1/29 (3) 0/27 (0) 0/27 (0) Minor NA NA NA NA NA NA 2/34 (6) 4/34 (12) 1/32 (3) 0/32 (0) 0/5 (0) 0/5 (0) 1/5 (20) 1/5 (20) 0/5 (0) 0/5 (0) 1/5 (20) 0/5 (0) 1/5 (20) 0/5 (20) 1/5 (20) 1/5 (20) 1/5 (20) 0/29 (0) 0/29 (0) 0/28 (0) 0/28 (0) 0/29 (0) 0/29 (0) 0/27 (0) 0/27 (0) 0/29 (0) 0/29 (0) 0/28 (0) 0/28 (0) 0/29 (0) 2/34 (6) 2/34 (6) 0/33 (0) 0/33 (0) 5/34 (15) 11/34 (32) 2/32 (6) 2/32 (6) 16/34 (47) 13/34 (38) 0/33 (0) 0/33 (0) 0/32 (0) E. faecium (43) Very major 0/39 (0) 0/39 (0) 0/39 (0) 0/39 (0) 1/39 (3) 1/39 (3) 0/39 (0) 0/39 (0) 1/38 (3) 1/38 (3) Major 0/4 (0) 0/4 (0) 0/4 (0) 0/4 (0) 0/4 (0) 0/4 (0) 0/4 (0) 0/4 (0) 0/4 (0) 0/4 (0) Minor NA NA NA NA NA NA 3/43 (7) 4/43 (9) 0/42 (0) 0/42 (0) 0/39 (0) 0/39 (0) 1/39 (3) 1/39 (3) 0/39 (0) 0/39 (0) 0/39 (0) 0/39 (0) 0/39 (0) 0/39 (0) 1/39 (3) 1/39 (3) 2/38 (5) 0/4 (0) 0/4 (0) 0/4 (0) 0/4 (0) 0/4 (0) 0/4 (0) 0/4 (0) 0/4 (0) 0/4 (0) 0/4 (0) 0/4 (0) 0/4 (0) 0/3 (0) 1/43 (2) 1/43 (2) 0/43 (0) 0/43 (0) 2/43 (5) 1/43 (2) 1/43 (2) 1/43 (2) 2/43 (5) 1/43 (2) 0/43 (0) 0/43 (0) 0/42 (0) E. casseliflavus/ flavescens (10) Very major 0/8 (0) 0/8 (0) 0/8 (0) 0/8 (0) 8/8 (100) 8/8 (100) 0/8 (0) 0/8 (0) 8/8 (100) 7/8 (88) 0/8 (0) 0/8 (0) 8/8 (100) 6/8 (75) 0/8 (0) 0/8 (0) 5/8 (62) 3/8 (38) 0/8 (0) 0/8 (0) 8/8 (100) 8/8 (100) 3/7 (43) Major 2/2 (100) 2/2 (100) 2/2 (100) 2/2 (100) 0/2 (0) 0/2 (0) 0/2 (0) 0/2 (0) 0/2 (0) 0/2 (0) 0/2 (0) 0/2 (0) 0/2 (0) 0/2 (0) 0/2 (0) 0/2 (0) 0/2 (0) 0/2 (0) 0/2 (0) 0/2 (0) 0/2 (0) 0/2 (0) 0/2 (0) Minor NA NA NA NA NA NA 10/10 (100) 10/10 (100) 0/10 (0) 1/10 (10) 10/10 (100) 10/10 (100) 0/10 (100) 2/10 (20) 10/10 (100) 10/10 (100) 4/10 (40) 6/10 (60) 9/10 (90) 9/10 (90) 0/10 (10) 0/10 (0) 6/9 (67) E. gallinarum (11) Very major 1/11 (9) 0/11 (0) 1/11 (9) 1/11 (9) 6/11 (55) 5/11 (45) 0/11 (0) 0/11 (0) 8/11 (73) 7/11 (64) 0/11 (0) 0/11 (0) 5/11 (45) 5/11 (45) 0/11 (0) 0/11 (0) 1/11 (9) 0/11 (0) 4/11 (36) 5/11 (45) 10/11 (91) 9/10 (90) 1/11 (9) Major Minor NA NA NA NA NA NA 10/11 (91) 10/11 (91) 2/11 (18) 3/11 (27) 10/11 (91) 10/11 (91) 5/11 (45) 5/11 (45) 10/11 (91) 10/11 (91) 9/11 (82) 10/11 (91) 6/11 (55) 5/11 (45) 0/11 (0) 0/11 (0) 8/11 (73) a Interpretive discrepancies are as follows: very major, resistant strain misinterpreted as susceptible (false-susceptible result); major, susceptible strain misinterpreted as resistant (false-resistant result) (since all E. gallinarium strains carried van resistance genes, they were defined as resistant, and therefore major errors could not exist); minor, resistant or susceptible strain misinterpreted as intermediately susceptible. b The enterococcal strains were incubated for 24 or 48 h, as indicated. For very major interpretive discrepancies, the number of very major errors is shown before the slash, the total number of resistant strains tested (some strains did not grow or were not tested) is shown after the slash, and the percentage of very major errors is shown in parentheses. For major interpretive discrepancies, the number of major errors is shown before the slash, the total number of susceptible strains tested (some strains did not grow or were not tested) is shown after the slash, and the percentage of major errors is shown in parentheses. For minor interpretive discrepancies, the number of minor errors is shown before the slash, the total number of resistant and susceptible strains (some strains did not grow or were not tested) is shown after the slash, and the percentage of intermediate errors is shown in parentheses. NA, not applicable (the agar screen method produces either a resistant or susceptible interpretive result only, so minor interpretive errors cannot be determined). 3261

3262 KOHNER ET AL. J. CLIN. MICROBIOL. TABLE 4. Summary of interpretive errors for all Enterococcus spp. Method, medium, and genes Interpretive error a Very major Major Minor 24 h 48 h 24 h 48 h 24 h 48 h Agar screen Synergy Quad vana, b vanb b 0/43 (0) 0/43 (0) vanc c 1/20 (5) 0/20 (0) No van genes d 2/37 (5) 2/37 (5) vana b, vanb b 0/43 (0) 0/43 (0) vanc c 1/20 (5) 0/20 (5) No van genes d 2/37 (5) 2/37 (5) vana, b vanb b 0/43 (0) 0/43 (0) vanc c 15/20 (75) 15/20 (75) No van genes d 0/37 (0) 0/37 (0) Broth microdilution vana, b vanb b 0/43 (0) 0/43 (0) 3/43 (7) 3/43 (7) vanc c 0/20 (0) 0/20 (0) 20/20 (100) 20/20 (100) No van genes d 0/37 (0) 1/37 (3) 2/37 (5) 5/37 (14) vana, b vanb b 0/42 (0) 0/42 (0) 1/42 (2) 0/42 (0) vanc c 18/20 (90) 16/20 (80) 2/20 (10) 4/20 (20) No van genes d 0/35 (0) 0/35 (0) 0/35 (0) 0/35 (0) Agar dilution vana, b vanb b 0/43 (0) 0/43 (0) 0/43 (0) 0/43 (0) vanc c 0/20 (0) 0/20 (0) 20/20 (100) 20/20 (100) No van genes d 0/37 (0) 0/37 (0) 3/37 (8) 3/37 (8) vana, b vanb b 0/43 (0) 0/43 (0) 0/43 (0) 0/43 (0) vanc c 15/20 (75) 13/20 (65) 5/20 (25) 7/20 (35) No van genes d 0/36 (0) 0/36 (0) 0/36 (0) 0/36 (0) E test vana, b vanb b 0/43 (0) 0/43 (0) 1/43 (2) 0/43 (0) vanc c 0/20 (0) 0/20 (0) 20/20 (100) 20/20 (20) No van genes d 0/37 (0) 0/37 (0) 6/37 (16) 12/37 (32) vana, b vanb b 1/43 (2) 0/43 (0) 0/43 (0) 0/43 (0) vanc c 7/20 (35) 3/20 (15) 14/20 (70) 17/20 (85) No van genes d 0/35 (0) 0/35 (0) 2/35 (0) 2/35 (0) Disk diffusion vana, b vanb b 0/43 (0) 0/43 (0) 1/43 (2) 0/43 (0) vanc c 5/20 (25) 5/20 (25) 15/20 (75) 15/20 (75) No van genes d 0/37 (0) 0/37 (0) 17/37 (46) 13/37 (51) vana, b vanb b 0/43 (0) 0/43 (0) 0/43 (0) 0/43 (0) vanc c 20/20 (100) 19/19 (100) 0/20 (0) 0/20 (0) No van genes d 0/36 (0) 0/36 (0) 0/36 (0) 0/36 (0) Vitek vana, b vanb b 3/42 (7) 0/42 (0) vanc c 4/19 (21) 12/19 (63) No van genes d 0/35 (0) 2/33 (6) a Interpretive errors are as follows: very major, resistant strain misinterpreted as susceptible (false-susceptible result); major, susceptible strain misinterpreted as resistant (false-resistant result); minor, resistant or susceptible strain misinterpreted as intermediately susceptible. The enterococcal strains were incubated for 24 or 48 h, as indicated, for all methods except Vitek. The number of strains with errors is shown before the slash, the number of strains studied is shown after the slash, and the percentage of strains with errors is shown in parentheses. b One isolate (E. gallinarum) contained both vana and vanc-1 genes. Two isolates (E. faecalis) had PCR amplicons with DNA sequences closely related to the vanb gene. Vancomycin MICs for all isolates with vana or vanb genes were 64 g/ml. c Vancomycin MICs for isolates containing vanc genes ranged from 4 to 8 g/ml. d Vancomycin MICs ranged from 2 to4 g/ml.

VOL. 35, 1997 TESTING SUSCEPTIBILITIES OF ENTEROCOCCI TO VANCOMYCIN 3263 different breakpoint guidelines as motile Enterococcus spp. (E. casseliflavus, E. flavescens, and E. gallinarum) often carry vanc genes which encode low-level vancomycin resistance. However, as we have shown (Table 1), two of the isolates evaluated in the current study were nonmotile, one E. faecium and one E. faecalis, yet carried the vanc-1 gene and had low-level resistance (MICs 8 and 4 g/ml, respectively). Until further clinical information becomes available, we would favor modification of NCCLS breakpoints to resolve this problem. It is generally recommended that resistant enterococci that grow on agar screen plates be confirmed by an MIC method (4). The current study demonstrates that two conventional confirmatory MIC methods, broth dilution and agar dilution if used with, and not as recommended by the NCCLS, adequately detect vancomycin resistance. In settings where the prevalence of vancomycin resistance is high, we feel the current study demonstrates that it is unnecessary to use the agar screen method when microbroth dilution, agar dilution, and the epsilometer tests are established susceptibility testing methods in the clinical microbiology laboratory and these methods are modified to use media. Furthermore, the current formulation of the agar screen which contains 6 g of vancomycin per ml may not detect low-level resistance vanc genotypes. ACKNOWLEDGMENT We thank R. Kondert for help in preparing the manuscript. REFERENCES 1. Evers, S., F. Sahm, and P. Courvalin. 1993. The vanb gene of vancomycinresistant Enterococcus faecalis V583 is structurally related to the genes encoding D-Ala:D-Ala ligases and glycopeptide-resistance proteins vana and vanc. Gene 124:143 144. 2. Evers, S., E. Reynolds, and P. Courvalin. 1994. Sequence of the vanb and ddl genes encoding D-alanine:D-lactate and D-alanine:D-alanine ligases in vancomycin-resistant Enterococcus faecalis V583. Gene 140:97 102. 3. Free, L., and D. F. Sahm. 1995. Investigation of the reformulated Remel Synergy Quad Plate for detection of high-level aminoglycoside and vancomycin resistance among enterococci. J. Clin. Microbiol. 33:1643 1645. 4. National Committee for Clinical Laboratory Standards. 1993. Performance standards for antimicrobial disk susceptibility tests, 5th ed. Approved standard M2-A5. National Committee for Clinical Laboratory Standards, Villanova, Pa. 5. National Committee for Clinical Laboratory Standards. 1993. Methods for dilution antimicrobial susceptibility tests for bacteria that grow aerobically, 3rd ed. Approved standard M7-A3. National Committee for Clinical Laboratory Standards, Villanova, Pa. 6. National Committee for Clinical Laboratory Standards. 1995. Performance standards for antimicrobial susceptibility testing. Sixth informational supplement M100-S6. National Committee for Clinical Laboratory Standards, Villanova, Pa. 7. Navarro, F., and P. Courvalin. 1994. Analysis of genes encoding D-alanine- D-alanine ligase-related enzymes in Enterococcus casseliflavus and Enterococcus flavescens. Antimicrob. Agents Chemother. 38:1788 1793. 8. Patel, R., J. R. Uhl, P. Kohner, M. K. Hopkins, and F. R. Cockerill III. 1997. Multiplex PCR detection of vana, vanb, vanc-1, and vanc-2/3 genes in enterococci. J. Clin. Microbiol. 35:703 707. 9. Schulz, J. E., and D. F. Sahm. 1993. Reliability of the E test for detection of ampicillin, vancomycin, and high-level aminoglycoside resistance in Enterococcus spp. J. Clin. Microbiol. 31:3336 3339. 10. Swenson, J. M., N. C. Clark, M. J. Ferraro, D. F. Sahm, G. Doern, M. A. Pfaller, L. B. Reller, M. P. Weinstein, R. J. Zabransky, and F. C. Tenover. 1994. Development of a standardized screening method for detection of vancomycin-resistant enterococci. J. Clin. Microbiol. 32:1700 1704. 11. Tenover, F. C., J. M. Swenson, C. M. O Hara, and S. A. Stocker. 1995. Ability of commercial and reference antimicrobial susceptibility testing methods to detect vancomycin resistance in enterococci. J. Clin. Microbiol. 33:1524 1527. 12. Van Horn, K. G., C. A. Gedris, K. M. Rodney, and J. B. Mitchell. 1996. Evaluation of commercial vancomycin agar screen plates for detection of vancomycin-resistant enterococci. J. Clin. Microbiol. 34:2042 2044. 13. Willey, B. M., B. N. Kreiswirth, A. E. Simor, G. Williams, S. R. Scriver, A. Phillips, and D. E. Low. 1992. Detection of vancomycin resistance in Enterococcus species. J. Clin. Microbiol. 30:1621 1624. 14. Willey, B. M., B. N. Kreiswirth, A. E. Simor, Y. Faur, M. Patel, G. Williams, and D. E. Low. 1993. Identification and characterization of multiple species of vancomycin-resistant enterococci, including an evaluation of Vitek software version 7.1. J. Clin. Microbiol. 31:2777 2779. 15. Zabransky, R. J., A. R. DiNuzzo, M. B. Huber, and G. L. Woods. 1994. Detection of vancomycin resistance in enterococci by the Vitek AMS System. Diagn. Microbiol. Infect. Dis. 20:113 116. 16. Zabransky, R. J., A. R. Dinuzzo, and G. L. Woods. 1995. Detection of vancomycin resistance in enterococci by the Alamar MIC system. J. Clin. Microbiol. 33:791 793.