Abstract
A fully automated commercial antifungal susceptibility test system (VITEK 2 yeast susceptibility test; bioMerieux, Inc., Hazelwood, Mo.) was compared in three different laboratories with Clinical and Laboratory Standards Institute (CLSI) reference broth microdilution (BMD) method by testing two quality control strains and a total of 426 isolates of Candida spp. (103 to 135 clinical isolates in each laboratory plus 80 challenge isolates in one laboratory) against fluconazole. Reference BMD MIC endpoints were established after 24 and 48 h of incubation. VITEK 2 endpoints were determined spectrophotometrically after 10 to 26 h of incubation (mean, 13 h). Excellent essential agreement (within two dilutions) between the VITEK 2 and the 24- and 48-h BMD MICs was observed. The overall agreement values were 97.9 and 93.7%, respectively. Both intra- and interlaboratory agreement was 100%. The overall categorical agreement between VITEK 2 and BMD was 97.2% at the 24-h BMD time point and 88.3% at the 48-h BMD time point. Decreased categorical agreement at 48 h was attributed to trailing growth observed with Candida glabrata. The VITEK 2 system reliably detected fluconazole resistance among Candida spp. and demonstrated excellent quantitative and qualitative agreement with the reference BMD method.
Standardized broth microdilution (BMD) susceptibility testing of fluconazole against Candida spp. has been available since 1997 (19, 21, 30). The establishment of a panel of quality control (QC) strains and validated, clinically useful interpretive breakpoints (19, 20, 27, 30, 32) has allowed this method to be used worldwide (3, 6-8, 12, 17, 21, 33).
The Clinical and Laboratory Standards Institute (CLSI; formerly National Committee for Clinical Laboratory Standards [NCCLS]) BMD method for testing fluconazole has served as a touchstone for the development of both broth- and agar-based procedures designed to provide simple, flexible, and commercially available alternative susceptibility testing methods for use in the clinical laboratory (8-12, 16, 17, 21, 23-25). The performance of the various commercially available antifungal testing systems has been variable (8, 10, 16, 17), and prior to the present study only two, the Sensititre YeastOne System (Trek, Cleveland, OH) and the Etest (AB BIODISK, Solna Sweden), have been approved by the U.S. Food and Drug Administration (FDA) for in vitro susceptibility testing of fluconazole against Candida spp. (21, 26).
Although spectrophotometric reading of BMD MIC endpoints has been shown to be valid and feasible for use in the clinical laboratory (11, 12, 16, 20, 21), this approach has not been incorporated into a commercially available testing method. Recently, bioMerieux (Hazelwood, MO) has developed a yeast susceptibility test that determines growth spectrophotometrically and allows fully automated antifungal susceptibility testing of fluconazole against Candida using the VITEK 2 microbiology system. The fully automated VITEK 2 system allows for the standardization of all of the critical parameters known for antifungal susceptibility testing: inoculum preparation, filling of the device, duration and temperature of incubation, and endpoint determination. The yeast susceptibility test, coupled with the rapid and accurate yeast identification capabilities already available on the VITEK 2 (4), would allow clinical laboratories to perform both yeast identification and antifungal susceptibility testing using a fully automated and completely standardized format. Preliminary studies by Zambardi et al. (G. Zambardi et al., Abstr. 45th Intersci. Conf. Antimicrob. Agents Chemother., abstr. M-1619, 2005) have shown both essential and categorical agreements of ≥90% in a comparison of VITEK 2 MICs with reference BMD MICs for fluconazole and Candida spp. The VITEK 2 results were available in ≤15 h compared to 48 h for the reference BMD method.
The purpose of the present study was to validate the performance of the VITEK 2 yeast susceptibility test with fluconazole against a broad range of Candida spp. in three independent laboratories. The VITEK 2 results were compared to those from a frozen reference BMD panel performed according to CLSI guidelines. It is notable that after the completion of the present study the VITEK 2 yeast susceptibility test for fluconazole was approved for clinical use by the U.S. FDA (bioMerieux press release, 27 September 2006).
MATERIALS AND METHODS
Study design.
The study was designed to compare the MIC results for fluconazole obtained by the VITEK 2 yeast susceptibility test to those obtained by the M27-A2 BMD method (20) in the three laboratories. Each laboratory tested at least 100 clinical isolates of Candida spp. (range, 103 to 135 isolates) by the VITEK 2 system and the CLSI frozen reference BMD panel (a total of 346 clinical isolates). In addition, a challenge set of 80 well-characterized stock isolates was tested by both methods in one of the laboratories. Intra- and interlaboratory reproducibility was determined by testing a panel of 10 Candida spp. isolates in triplicate on three separate days in each of the participating laboratories. The MIC results obtained with the VITEK 2 system after 10 to 26 h of incubation (depending on the organism growth rate) were compared to those obtained with the reference BMD panel read after both 24 and 48 h of incubation.
Test organisms.
The test organisms included two American Type Culture Collection (ATCC) strains that have been established as QC strains (Candida parapsilosis ATCC 22019 and C. krusei ATCC 6258) by the CLSI (5, 20). A challenge set of 80 isolates of Candida spp. selected to provide strains with on-scale MIC results and to represent both clinically important species and resistance mechanisms were tested in one of the participating laboratories. The challenge set included 32 isolates of C. albicans, 6 isolates of C. dubliniensis, 14 isolates of C. glabrata, 3 isolates of C. guilliermondii, 5 isolates of C. krusei, 5 isolates of C. lusitaniae, 1 isolate of C. norvegensis, 7 isolates of C. parapsilosis, 2 isolates of C. pelliculosa, and 5 isolates of C. tropicalis. An additional 346 clinical isolates of Candida spp. were also tested. The clinical isolates included 166 isolates of C. albicans, 2 isolates of C. dubliniensis, 69 isolates of C. glabrata, 46 isolates of C. krusei, 4 isolates of C. lusitaniae, 36 isolates of C. parapsilosis, and 23 isolates of C. tropicalis. These were all recent clinical isolates and were selected to represent the clinically prevalent species, including fluconazole-resistant strains. Reproducibility within and among laboratories was assessed by using a panel of 10 Candida isolates: C. glabrata strain 304201, C. glabrata strain 304927, C. haemulonii strain 304848, C. krusei strain 304204, C. krusei strain 304845, C. krusei strain 304850, C. lipolytica strain 204856, C. lusitaniae strain 304205, C. norvegensis strain 304852, and C. pelliculosa strain 304847. These isolates were selected to provide on-scale fluconazole MICs ranging from 2 to 32 μg/ml. All isolates were identified by standard methods (14). Prior to testing, each isolate was passaged at least twice on Sabouraud dextrose agar (Remel, Lenexa, KS) to ensure purity and viability.
Antifungal agents and microdilution panels.
The VITEK 2 cards containing serial twofold dilutions of fluconazole (range, 1 to 64 μg/ml) were provided by the manufacturer. The frozen BMD reference panels containing serial twofold dilutions of fluconazole (range, 0.12 to 128 μg/ml) were provided by Trek Diagnostic Systems (Cleveland, OH). The VITEK 2 cards were shipped in sealed packages and stored at 2 to 8°C until testing was performed. The BMD panels were shipped frozen in sealed packages and were stored at −70°C until the day of the test.
Inoculum preparation.
Stock inoculum suspensions of the Candida spp. were obtained from 24-h cultures on Sabouraud dextrose agar at 35°C. The inoculum suspensions for the VITEK 2 were prepared in sterile saline to a turbidity equal to a 2.0 McFarland standard by using the bioMerieux DensiChek instrument. The inoculum suspensions for the reference BMD were prepared by diluting a portion of the 2.0 McFarland suspension prepared for the VITEK 2 to match the turbidity of a 0.5 McFarland.
CLSI broth microdilution method.
Reference BMD testing was performed exactly as outlined in CLSI document M27-A2 (20) with a final inoculum concentration of 1.5 × 103 ± 1.0 × 103 cells/ml and RPMI 1640 medium buffered to pH 7.0 with 0.165 M morpholinepropanesulfonic acid buffer. The panels were incubated in air at 35°C and observed for the presence or absence of growth at 24 and 48 h. The fluconazole MIC was read as the lowest concentration that produced a prominent decrease in turbidity (ca. 50% reduction in growth) relative to the drug-free control (20).
VITEK 2 yeast susceptibility test.
The standardized 2.0 McFarland inoculum suspension was placed into a VITEK 2 cassette along with a sterile polystyrene test tube and a yeast susceptibility test card for each organism. The loaded cassettes were then placed into the VITEK 2 instrument, and the respective yeast suspensions were diluted appropriately, after which the cards were filled, incubated, and read automatically. The time of incubation varied from 10 to 26.1 h, based on the rate of growth in the drug-free control well, and the results were expressed as MICs in micrograms per milliliter.
Quality control.
Quality control was ensured by testing the CLSI-recommended quality control strains C. parapsilosis ATCC 2209 and C. krusei 6258 (5, 20). These isolates were tested between 22 and 29 times in each of the three laboratories (total number of results = 294), and all (100%) MICs were in the respective reference ranges.
Analysis of results.
The MIC results obtained with the VITEK 2 yeast susceptibility test were compared to those of the reference BMD panels read at 24 and 48 h. As with previous studies (9-12, 24, 25), high off-scale MIC results were converted to the next highest concentration, and low off-scale MIC results were left unchanged. Discrepancies among MIC endpoints of more than two dilutions (two wells) were used to calculate the essential agreement (EA). Interlaboratory and intralaboratory agreement, assessed with the 10-isolate reproducibility panel, was defined when MIC results were within a three-dilution range. The CLSI interpretive breakpoints for fluconazole (susceptible [S], ≤8 μg/ml; susceptible dose dependent [SDD], 16 to 32 μg/ml; resistant [R], ≥64 μg/ml) were used to obtain categorical agreement (CA) percentages between the MICs determined by VITEK 2 and the reference BMD (20, 27). Very major errors (VME) were identified when the reference MIC indicated R, and the VITEK 2 MIC indicated S. Major errors (ME) were identified when the isolate was classified as R by the VITEK 2 and S by the reference method. Minor errors were determined when the results of one of the test methods was either S or R and that of the other was SDD. The MIC results obtained for C. krusei were used as such (in micrograms per milliliter) for the purpose of assessing EA but were forced into the R category, as required by the CLSI (20), when determining CA.
RESULTS AND DISCUSSION
Table 1 summarizes the in vitro fluconazole susceptibilities of 426 isolates of Candida spp. as determined by the VITEK 2 system and the reference BMD read at 24 and 48 h. Due to the similarity in results obtained with the VITEK 2 compared to the 24- and 48-h BMDs for both the challenge isolates (80 isolates, 100 and 97.5% EA, respectively) and the clinical isolates (346 isolates, 97.4 and 92.8% EA, respectively), the results for the two organism sets were combined in Table 1. In general, MIC results for fluconazole were typical of each species of Candida (22), with the lowest MICs obtained with both VITEK 2 and BMD observed for C. albicans and the highest MICs observed for C. glabrata and C. krusei. Notably, BMD MICs read at 24 h of incubation tended to be approximately fourfold lower than those read at 48-h for both C. glabrata (MIC50, 4 μg/ml versus 16 μg/ml, respectively) and C. krusei (MIC50, 16 μg/ml versus 64 μg/ml, respectively), suggesting significant trailing with these two species.
TABLE 1.
Fluconazole susceptibilities of 426 isolates of Candida spp. as determined by the VITEK 2 yeast susceptibility test and by CLSI BMD methodsa
| Species | No. tested | Test methodb | MIC (μg/ml)c
|
EA (%)d | ||
|---|---|---|---|---|---|---|
| Range | 50% | 90% | ||||
| C. albicans | 198 | VITEK 2 | ≤1-≥64 | ≤1 | ≤1 | |
| BMD-24 | ≤0.12-≥128 | 0.25 | 0.5 | 97.0 | ||
| BMD-48 | ≤0.12-≥128 | 0.25 | 0.5 | 96.5 | ||
| C. dubliniensis | 8 | VITEK 2 | ≤1-32 | ≤1 | ||
| BMD-24 | ≤0.12-16 | ≤0.12 | 100.0 | |||
| BMD-48 | ≤0.12-16 | 0.25 | 100.0 | |||
| C. glabrata | 83 | VITEK 2 | ≤1-≥64 | 8 | ≥64 | |
| BMD-24 | 0.5-≥128 | 4 | 64 | 98.8 | ||
| BMD-48 | 0.5-≥128 | 16 | ≥128 | 96.4 | ||
| C. guilliermondii | 3 | VITEK 2 | ≤1-4 | 2 | ||
| BMD-24 | 1-4 | 4 | 100.0 | |||
| BMD-48 | 2-4 | 4 | 100.0 | |||
| C. krusei | 51 | VITEK 2 | ≤1-≥64 | 16 | 32 | |
| BMD-24 | ≤1-≥128 | 16 | 32 | 100.0 | ||
| BMD-48 | ≤1-≥128 | 64 | ≥128 | 68.6 | ||
| C. lusitaniae | 9 | VITEK 2 | ≤1-2 | 1 | ||
| BMD-24 | ≤0.12-1 | 0.5 | 100.0 | |||
| BMD-48 | ≤0.12-2 | 0.5 | 100.0 | |||
| C. norvegensis | 1 | VITEK 2 | 8 | |||
| BMD-24 | 16 | 100.0 | ||||
| BMD-48 | 16 | 100.0 | ||||
| C. parapsilosis | 43 | VITEK 2 | ≤1-≥64 | ≤1 | 2 | |
| BMD-24 | 0.25-4 | 0.5 | 1 | 95.3 | ||
| BMD-48 | 0.5-4 | 1 | 2 | 97.7 | ||
| C. pelliculosa | 2 | VITEK 2 | 2-4 | 2 | ||
| BMD-24 | 2-4 | 2 | 100.0 | |||
| BMD-48 | 4 | 4 | 100.0 | |||
| C. tropicalis | 28 | VITEK 2 | ≤1-8 | ≤1 | ≤1 | |
| BMD-24 | ≤0.12-8 | 0.25 | 1 | 100.0 | ||
| BMD-48 | 0.25-16 | 0.5 | 4 | 100.0 | ||
| All Candida spp. | 426 | VITEK 2 | ≤1-≥64 | ≤1 | 16 | |
| BMD-24 | ≤0.12-≥128 | 0.5 | 16 | 97.9 | ||
| BMD-48 | ≤0.12-≥128 | 0.5 | 64 | 93.7 | ||
Isolates included both clinical (n = 346) and challenge (n = 80) sets.
BMD-24 and BMD-48, BMD performed at 24 h and 48 h of incubation, respectively.
50% and 90%, MICs encompassing 50% and 90% of isolates tested, respectively.
% EA is the EA (±2 log2 dilutions) between VITEK 2 and BMD MICs.
The overall EA between the VITEK 2 and the BMD MICs ranged from 97.9% when the 24-h BMD result was used as a reference to 93.7% when the 48-h BMD result was used as reference (95.2 and 83.0%, respectively, for on-scale results). Of the discrepancies noted between the VITEK 2 and 24-h BMD MIC results, the MICs generated by the VITEK 2 were higher than those obtained by BMD in all nine instances (100%). In contrast, of the 27 discrepancies observed between THE VITEK 2 and 48-h BMD MIC results, the MICs generated by the VITEK 2 were lower than those obtained by BMD in 21 instances (77.8%). The latter discrepancies occurred almost exclusively with C. glabrata and C. krusei, emphasizing the impact of trailing on BMD results obtained with these two species.
The mean time to result for the VITEK 2 system was 13 h, with a range from 10 to 26 h. Only one isolate, a clinical isolate of C. parapsilosis, failed to grow in the VITEK 2 system, and all isolates grew sufficiently well in the BMD panel to be read after 24 h of incubation. Similar results were obtained at all three study sites.
Regarding the individual species of Candida, the EA between the VITEK 2 results and either the 24-h or the 48-h BMD MICs was >95% for all species with the exception of C. krusei. Whereas the EA for this species was 100% when VITEK 2 results were compared to the 24-h BMD MICs, it was only 68.6% when the 48-h BMD MICs were used as the reference result. Again, these discrepancies were due to higher MICs obtained with the BMD method than were obtained with the VITEK 2 system for this species. Given the CLSI recommendation that C. krusei should be considered resistant to fluconazole irrespective of the MIC, these discrepancies should not pose a problem clinically (20).
The VITEK 2 fluconazole MIC results were highly reproducible, as determined by replicate testing of a panel of 10 Candida spp. isolates in the three laboratories (Table 2). Both intra- and interlaboratory reproducibility was 100% for all 10 organisms. This high level of reproducibility underscores the excellent level of test standardization achieved with this automated microbiology system.
TABLE 2.
VITEK 2 Fluconazole MIC reproducibility within and among three different laboratories
| Species | Sample | Study site | No. of results at an MIC (μg/ml) of:
|
||||||
|---|---|---|---|---|---|---|---|---|---|
| ≤1 | 2 | 4 | 8 | 16 | 32 | ≥64 | |||
| C. glabrata | 304201 | 1 | 7 | 2 | |||||
| 2 | 9 | ||||||||
| 3 | 8 | 1 | |||||||
| All | 24 | 3 | |||||||
| C. glabrata | 304927 | 1 | 1 | 8 | |||||
| 2 | 9 | ||||||||
| 3 | 9 | ||||||||
| All | 1 | 26 | |||||||
| C. haemulonii | 304848 | 1 | 1 | 7 | 1 | ||||
| 2 | 6 | 3 | |||||||
| 3 | 7 | 2 | |||||||
| All | 1 | 20 | 6 | ||||||
| C. krusei | 304204 | 1 | 8 | 1 | |||||
| 2 | 8 | 1 | |||||||
| 3 | 7 | 2 | |||||||
| All | 23 | 4 | |||||||
| C. krusei | 304845 | 1 | 8 | 1 | |||||
| 2 | 8 | 1 | |||||||
| 3 | 4 | 5 | |||||||
| All | 20 | 7 | |||||||
| C. krusei | 304850 | 1 | 8 | 1 | |||||
| 2 | 9 | ||||||||
| 3 | 7 | 2 | |||||||
| All | 24 | 3 | |||||||
| C. lipolytica | 304856 | 1 | 9 | ||||||
| 2 | 5 | 4 | |||||||
| 3 | 1 | 8 | |||||||
| All | 1 | 22 | 4 | ||||||
| C. lusitaniae | 304205 | 1 | 9 | ||||||
| 2 | 9 | ||||||||
| 3 | 9 | ||||||||
| All | 27 | ||||||||
| C. norvegensis | 304852 | 1 | 9 | ||||||
| 2 | 8 | 1 | |||||||
| 3 | 9 | ||||||||
| All | 26 | 1 | |||||||
| C. pelliculosa | 304847 | 1 | 4 | 5 | |||||
| 2 | 3 | 6 | |||||||
| 3 | 4 | 5 | |||||||
| All | 11 | 16 | |||||||
CA between the VITEK 2 and BMD methods was assessed by combining the data obtained with the clinical and challenge organism collections in all three laboratories (Table 3). Excellent CA was observed for all comparisons with the exception of C. glabrata and C. norvegensis. The overall CA for the comparison of VITEK 2 results with 24-h BMD results was 97.2% with no very major or major errors. Only C. glabrata (89.2%) and C. norvegensis (0.0%) among the 10 species tested showed less than 90% absolute CA with the 24-h BMD results. All of the errors observed with C. glabrata were minor errors and were the result of isolates determined to be susceptible by BMD and SDD by VITEK 2. This is not surprising given the tendency of fluconazole MICs for this species to fall close to the susceptible breakpoint: 82% of MICs determined by VITEK 2 and 78% of MICs determined by 24-h BMD fell between 4 and 16 μg/ml (data not shown). Importantly, all isolates of C. glabrata testing as resistant by the BMD method at 24 h were also resistant with the VITEK 2 system. Only one isolate of C. norvegensis was tested and was found to be susceptible by VITEK 2 (MIC = 8 μg/ml) and SDD by 24-h BMD (MIC = 16 μg/ml).
TABLE 3.
Categorical agreement between VITEK 2 yeast susceptibility test MICs and 24-h and 48-h CLSI BMD fluconazole MICs for 426 isolates of Candida spp. in three laboratoriesa
| Species (no. of strains tested) | Test method | % of MICs by category
|
% CA | % Errors
|
||||
|---|---|---|---|---|---|---|---|---|
| S | SDD | R | VME | ME | Minor | |||
| C. albicans (198) | VITEK 2 | 97.5 | 1.0 | 1.5 | ||||
| BMD-24 | 98.0 | 0.5 | 1.5 | 99.5 | 0.0 | 0.0 | 0.5 | |
| BMD-48 | 97.5 | 1.0 | 1.5 | 99.0 | 0.0 | 0.0 | 1.0 | |
| C. dubliniensis (8) | VITEK 2 | 75.0 | 25.0 | 0.0 | ||||
| BMD-24 | 75.0 | 25.0 | 0.0 | 100.0 | 0.0 | 0.0 | 0.0 | |
| BMD-48 | 75.0 | 25.0 | 0.0 | 100.0 | 0.0 | 0.0 | 0.0 | |
| C. glabrata (83) | VITEK 2 | 75.9 | 13.3 | 10.8 | ||||
| BMD-24 | 84.3 | 4.8 | 10.9 | 89.2 | 0.0 | 0.0 | 10.8 | |
| BMD-48 | 27.7 | 59.0 | 13.3 | 45.8 | 1.2 | 0.0 | 53.0 | |
| C. guilliermondii (3) | VITEK 2 | 100.0 | 0.0 | 0.0 | ||||
| BMD-24 | 100.0 | 0.0 | 0.0 | 100.0 | 0.0 | 0.0 | 0.0 | |
| BMD-48 | 100.0 | 0.0 | 0.0 | 100.0 | 0.0 | 0.0 | 0.0 | |
| C. kruseib (51) | VITEK 2 | 0.0 | 0.0 | 100.0 | ||||
| BMD-24 | 0.0 | 0.0 | 100.0 | 100.0 | 0.0 | 0.0 | 0.0 | |
| BMD-48 | 0.0 | 0.0 | 100.0 | 100.0 | 0.0 | 0.0 | 0.0 | |
| C. lusitaniae (9) | VITEK 2 | 100.0 | 0.0 | 0.0 | ||||
| BMD-24 | 100.0 | 0.0 | 0.0 | 100.0 | 0.0 | 0.0 | 0.0 | |
| BMD-48 | 100.0 | 0.0 | 0.0 | 100.0 | 0.0 | 0.0 | 0.0 | |
| C. norvegensis (1) | VITEK 2 | 100.0 | 0.0 | 0.0 | ||||
| BMD-24 | 0.0 | 100.0 | 0.0 | 0.0 | 0.0 | 0.0 | 100.0 | |
| BMD-48 | 0.0 | 100.0 | 0.0 | 0.0 | 0.0 | 0.0 | 100.0 | |
| C. parapsilosis (43) | VITEK 2 | 97.7 | 2.3 | 0.0 | ||||
| BMD-24 | 100.0 | 0.0 | 0.0 | 97.7 | 0.0 | 0.0 | 2.3 | |
| BMD-48 | 100.0 | 0.0 | 0.0 | 97.7 | 0.0 | 0.0 | 2.3 | |
| C. pelliculosa (2) | VITEK 2 | 100.0 | 0.0 | 0.0 | ||||
| BMD-24 | 100.0 | 0.0 | 0.0 | 100.0 | 0.0 | 0.0 | 0.0 | |
| BMD-48 | 100.0 | 0.0 | 0.0 | 100.0 | 0.0 | 0.0 | 0.0 | |
| C. tropicalis (28) | VITEK 2 | 100.0 | 0.0 | 0.0 | ||||
| BMD-24 | 100.0 | 0.0 | 0.0 | 100.0 | 0.0 | 0.0 | 0.0 | |
| BMD-48 | 96.4 | 3.6 | 0.0 | 96.4 | 0.0 | 0.0 | 3.6 | |
| All Candida spp. (426) | VITEK 2 | 81.5 | 3.8 | 14.7 | ||||
| BMD-24 | 83.3 | 2.0 | 14.7 | 97.2 | 0.0 | 0.0 | 2.8 | |
| BMD-48 | 71.8 | 12.9 | 15.3 | 88.3 | 0.2 | 0.0 | 11.5 | |
Isolates include both clinical (n = 346) and challenge (n = 80) sets.
C. krusei categorical results forced to R regardless of MICs.
Although the overall CA between VITEK 2 and the 48-h BMD results was lower than that seen with the 24-h BMD comparison (88.3% versus 97.2%), the vast majority of errors were minor, and almost all were due to a shift in the MICs for C. glabrata from susceptible at 24 h to SDD at 48 h with the BMD method (Table 3). There were no ME and only one VME (0.2% of all isolates and 1.5% of all resistant isolates), the latter being a C. glabrata result. Again, with the exception of C. glabrata and C. norvegensis, the CA between the VITEK 2 and 48-h BMD results exceeded 95% for all species of Candida. All but one of the errors observed with C. glabrata were minor and can be attributed to clustering of fluconazole MICs around the S-SDD breakpoint (66% of the 48-h BMD results fell in the range from 4 to 16 μg/ml, and 75% were in the range 8 to 32 μg/ml) and the influence of trailing growth seen with this species and fluconazole.
The overall pattern of results shown in Table 3 was also seen with clinical isolates tested in each of the three laboratories (Table 4). In each laboratory the categorical agreement was ≥95% when VITEK 2 results were compared to 24-h BMD results and ranged from 88.0 to 89.6% with the comparison between VITEK 2 and 48-h BMD. A shift from S to SDD was observed in all three laboratories when 24-h BMD results were compared to 48-h BMD results.
TABLE 4.
Agreement between VITEK 2 yeast susceptibility test and reference BMD MIC results for fluconazole against clinical isolates of Candida spp. in each of three laboratories
| Study site (no. of strains tested) | BMD incubation time (h) | % of BMD by category
|
% CA | % Errors
|
||||
|---|---|---|---|---|---|---|---|---|
| S | SDD | R | VME | ME | Minor | |||
| 1 (135) | 24 | 86.0 | 0.7 | 13.3 | 98.5 | 0.0 | 0.0 | 1.5 |
| 48 | 75.6 | 11.1 | 13.3 | 89.6 | 0.0 | 0.0 | 10.4 | |
| 2 (108) | 24 | 81.5 | 0.9 | 17.6 | 97.2 | 0.0 | 0.0 | 2.8 |
| 48 | 69.4 | 11.1 | 19.5 | 88.0 | 0.9 | 0.0 | 11.1 | |
| 3 (103) | 24 | 79.6 | 2.9 | 17.5 | 95.1 | 0.0 | 0.0 | 4.9 |
| 48 | 70.0 | 12.5 | 17.5 | 89.3 | 0.0 | 0.0 | 10.7 | |
The findings of the present study document the excellent degree of standardization and reproducibility that can be achieved with the VITEK 2 yeast susceptibility test. This system is the first automated approach to antifungal susceptibility testing and as such provides the ultimate in test standardization. In addition to providing highly reproducible results, the VITEK 2 system was rapid, with a mean time to result of 13 h. The availability of rapid quantitative antifungal susceptibility data will be a major step in optimizing the therapy of invasive candidal infections (1, 13, 15, 18).
The EA observed in the present study between VITEK 2 and BMD MICs demonstrates excellent quantitative agreement between the methods for all species of Candida tested and is comparable or superior to that reported for other commercial systems (8-11, 16, 17, 24, 25). The highly automated, hands-off nature of the system virtually eliminates the subjectivity that affects all other test systems.
The use of a spectrophotometer to provide objective, and earlier, readings of MICs is well established (8, 11, 12, 16). It is also proposed as a means by which one can mitigate falsely elevated fluconazole MICs due to trailing growth (2, 7, 8, 11, 12). Based on the comparative data presented herein, it appears that the VITEK 2 yeast susceptibility test accomplishes these objectives as well. Thus, although the CA between the VITEK 2 and the 48-h BMD results was lower than that observed for the 24-h BMD comparison, it may be the latter that is more relevant clinically (2, 11, 16). Earlier work has shown that the 24-h fluconazole MIC endpoint correlated better than the 48-h endpoint with sterol quantification (2) and with treatment outcome both clinically (29) and in a murine model of invasive candidiasis (31). These findings suggest that fluconazole results for isolates of Candida spp. with significant trailing should be interpreted on the basis of the lower MIC observed at the earlier (24-h) time point (16, 20). Given the evidence of trailing, especially with C. glabrata and C. krusei, at 48 h shown in Tables 1 and 3, the most appropriate comparator for the VITEK 2 may be the 24-h BMD results. This comparison indicates that the VITEK 2 provides highly accurate quantitative and qualitative results for fluconazole and Candida spp.
The value of a rapid, automated, commercially available antifungal susceptibility testing system such as the VITEK 2 is not limited to the generation of fluconazole susceptibility data. Recently, we have shown that fluconazole may serve as a surrogate marker for voriconazole susceptibility of Candida spp. (28). Specifically, fluconazole MICs of ≤32 μg/ml predict susceptibility and MICs of ≥64 μg/ml predict resistance of Candida spp. to voriconazole with an absolute CA of 97%, 0.1% VME, and 1.4% ME. Thus, clinical laboratories performing antifungal susceptibility testing of fluconazole using the VITEK 2 can reliably use these results as surrogate markers of susceptibility and resistance to voriconazole until commercial FDA-approved voriconazole susceptibility tests become available.
In summary, the MICs of fluconazole can be determined in an automated fashion in less than 15 h for most species of Candida with the VITEK 2 yeast susceptibility test. The VITEK 2 system ensures that each test is performed in a highly standardized manner and provides quantitative MIC results that are reproducible and accurate. The use of spectrophotometry to determine the MIC endpoint eliminates subjectivity and minimizes the effect of trailing growth that compromises the performance of systems relying on visual MIC determination. The VITEK 2 system reliably identifies fluconazole resistance among Candida spp. and demonstrates excellent quantitative and qualitative agreement with the reference BMD method. The introduction of the VITEK 2 system in the clinical laboratory will be an important step toward the optimization of antifungal therapy of candidiasis.
Acknowledgments
This study was supported in part by a grant from bioMerieux, Inc.
We thank Linda Elliott for secretarial assistance in the preparation of the manuscript. We acknowledge the excellent assistance of the technical personnel in the Iowa, Ohio, and Texas laboratories.
Footnotes
Published ahead of print on 10 January 2007.
REFERENCES
- 1.Alexander, B. D., and M. A. Pfaller. 2006. Contemporary tools for the diagnosis and management of invasive mycoses. Clin. Infect. Dis. 43(Suppl. 1):S15-S27. [Google Scholar]
- 2.Arthington-Skaggs, B. A., W. Lee-Yong, M. A. Ciblak, J. P. Frade, M. E. Brandt, R. A. Hajjeh, L. E. Harrison, A. N. Sofair, and D. W. Warnock. 2002. Comparison of visual and spectrophotometric methods of broth microdilution MIC endpoint determination and evaluation of sterol quantitation method for in vitro susceptibility testing of fluconazole and itraconazole against trailing and non-trailing Candida isolates. Antimicrob. Agents Chemother. 46:2477-2481. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Asmundsdottir, L. R., H. Erlendsdottir, and M. Gottfredsson. 2002. Increasing incidence of candidemia: results from a 20-year nationwide study in Iceland. J. Clin. Microbiol. 40:3489-3492. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Aubertine, C. L., M. Rivera, S. M. Rohan, and D. H. Larone. 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. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Barry, A. L., M. A. Pfaller, S. D. Brown, A. Espinel-Ingroff, M. A. Ghannoum, C. Knapp, R. P. Rennie, J. H. Rex, and M. G. Rinaldi. 2000. Quality control limits for broth microdilution susceptibility tests of ten antifungal agents. J. Clin. Microbiol. 38:3457-3459. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Chen, Y. C., S. C. Chang, K. T. Luh, and W. C. Hsieh. 2003. Stable susceptibility of Candida bloodstream isolates to fluconazole despite increasing use during the past 10 years. J. Antimicrob. Chemother. 52:71-77. [DOI] [PubMed] [Google Scholar]
- 7.Cuenca-Estrella, M., W. Lee-Yang, M. A. Ciblak, B. A. Arthington-Skaggs, E. Mellado, D. N. Warnock, and J. L. Rodriguez-Tudela. 2002. Comparative evaluation of NCCLS M27-A and EUCAST broth microdilution procedures for antifungal susceptibility testing of Candida species. Antimicrob. Agents Chemother. 46:3644-3647. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Cuenca-Estrella, M., A. Gomez-Lopez, E. Mellado, and J. L. Rodriguez-Tudela. 2005. Correlation between the procedure for antifungal susceptibility testing for Candida spp. of the European Committee on Antibiotic Susceptibility Testing (EUCAST) and four commercial techniques. Clin. Microbiol. Infect. 11:486-492. [DOI] [PubMed] [Google Scholar]
- 9.Espinel-Ingroff, A., M. Pfaller, S. A. Messer, C. C. Knapp, S. Killian, H. A. Norris, and M. A. Ghannoum. 1999. Multicenter comparison of the Sensititre YeastOne Colorimetric Antifungal Panel with the National Committee for Clinical Laboratory Standards M27-A reference method for testing clinical isolates of common and emerging Candida spp., Cryptococcus spp., and other yeasts and yeast-like organisms. J. Clin. Microbiol. 37:591-595. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Espinel-Ingroff, A., M. Pfaller, S. A. Messer, C. C. Knapp, N. Holliday, and S. B. Killian. 2004. Multicenter comparison of the Sensititire YeastOne Colorimetric Antifungal Panel with the NCCLS M27-A2 reference method for testing new antifungal agents against clinical isolates of Candida spp. J. Clin. Microbiol. 42:718-721. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Espinel-Ingroff, A., F. Barchiesi, M. Cuenca-Estrella, A. Fothergill, M. A. Pfaller, M. Rinaldi, J. L. Rodriguez-Tudela, and P. E. Verweij. 2005. Comparison of visual 24-hour and spectrophotometric 48-hour MICs to CLSI reference microdilution MICs of fluconazole, itraconazole, posaconazole, and voriconazole for Candida spp.: a collaborative study. J. Clin. Microbiol. 43:4535-4540. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Espinel-Ingroff, A., F. Barchiesi, M. Cuenca-Estrella, M. A. Pfaller, M. Rinaldi, J. L. Rodriguez-Tudela, and P. E. Verweij. 2005. International and multicenter comparison of EUCAST and CLSI M27-A2 broth microdilution methods for testing susceptibilities of Candida spp. to fluconazole, itraconazole, posaconazole, and voriconazole. J. Clin. Microbiol. 43:3884-3889. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Garey, K. W., M. Reye, M. P. Pai, D. E. Mingo, K. J. Suda, R. S. Turpin, and D. T. Bearden. 2006. Time to initiation of fluconazole therapy impacts mortality in patients with candidemia: a multi-institutional study. Clin. Infect. Dis. 43:25-31. [DOI] [PubMed] [Google Scholar]
- 14.Hazen, K. C., and S. A. Howell. 2003. Candida, Cryptococcus, and other yeasts of medical importance, p. 1693-1711. In P. R. Murray, E. J. Baron, J. H. Jorgensen, M. A. Pfaller, and R. H. Yolken (ed.), Manual of clinical microbiology, 8th ed. ASM Press, Washington, DC.
- 15.Magill, S. S., C. Shields, C. L. Sears, M. Choti, and W. G. Merz. 2006. Triazole-cross resistance among Candida spp.: occurrence among bloodstream isolates, and implication for antifungal therapy. J. Clin. Microbiol. 44:529-535. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Matar, M. J., L. Ostrosky-Zeichner, V. L. Pretznick, J. R. Rodriguez, E. Chan, and J. H. Rex. 2003. Correlation between E-test, disk diffusion, and microdilution methods for antifungal susceptibility testing of fluconazole and voriconazole. Antimicrob. Agents Chemother. 47:1647-1651. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Morace, G., G. Amato, F. Bistoni, G. Fadda, P. Marone, M. Montagna, S. Oliveri, L. Polonelli, R. Rigoli, I. Mancuso, S. La Face, L. Masucci, L. Romano, C. Napoli, D. Tato, M. G. Buscema, C. M. C. Belli, M. M. Piccirillo, S. Conti, S. Covan, F. Fanti, C. Cavanna, F. D'Alo, and L. Pitzurra. 2002. Multicenter comparative evaluation of six commercial systems and the National Committee for Clinical Laboratory Standards M27-A broth microdilution method for fluconazole susceptibility testing of Candida species. J. Clin. Microbiol. 40:2953-2958. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Morrell, M., V. J. Fraser, and M. J. Kollef. 2005. Delaying empiric treatment of Candida bloodstream infection until positive blood culture results are obtained: a potential risk factor for mortality. Antimicrob. Agents Chemother. 49:3640-3645. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.National Committee for Clinical Laboratory Standards. 1997. Reference method for broth dilution antifungal susceptibility testing of yeasts. M27-A. National Committee for Clinical Laboratory Standards, Villanova, PA.
- 20.National Committee for Clinical Laboratory Standards. 2002. Reference method for broth dilution antifungal susceptibility testing of yeasts: approved standard, 2nd ed., M27-A2. National Committee for Clinical Laboratory Standards, Wayne, PA.
- 21.Pfaller, M. A. 2005. Antifungal susceptibility testing methods. Current Drug Targets 6:929-943. [DOI] [PubMed] [Google Scholar]
- 22.Pfaller, M. A., and D. J. Diekema. 2004. Twelve years of fluconazole in clinical practice: global trends in species distribution and fluconazole susceptibility of bloodstream isolates of Candida. Clin. Microbiol. Infect. 10(Suppl. 1):11-23. [DOI] [PubMed] [Google Scholar]
- 23.Pfaller, M. A., K. C. Hazen, S. A. Messer, L. Boyken, S. Tendolkar, R. J. Hollis, and D. J. Diekema. 2004. Comparison of results of fluconazole disk diffusion testing for Candida species with results from a central reference laboratory in the ARTEMIS Global Antifungal Surveillance Program. J. Clin. Microbiol. 42:3607-3612. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Pfaller, M. A., L. Boyken, R. J. Hollis, S. A. Messer, S. Tendolkar, and D. J. Diekema. 2004. Clinical evaluation of a dried commercially prepared microdilution panel for antifungal susceptibility testing of five antifungal agents against Candida spp. and Cryptococcus neoformans. Diagn. Microbiol. Infect. Dis. 50:113-117. [DOI] [PubMed] [Google Scholar]
- 25.Pfaller, M. A., A. Espinel-Ingroff, and R. N. Jones. 2004. Clinical evaluation of the Sensititre YeastOne colorimetric antifungal plate for antifungal susceptibility testing of the new triazoles voriconazole, posaconazole, and ravuconazole. J. Clin. Microbiol. 42:4577-4580. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Pfaller, M. A, R. N. Jones, the Microbiology Resource Committee of the College of American Pathologists. 2006. Performance accuracy of antibacterial and antifungal susceptibility test methods: report from the College of American Pathologists (CAP) Microbiology Surveys Program (2001-2003). Arch. Pathol. Lab. Med. 130:767-778. [DOI] [PubMed] [Google Scholar]
- 27.Pfaller, M. A., D. J. Diekema, and D. J. Sheehan. 2006. Intepretive breakpoints for fluconazole and Candida revisited: a blueprint for the future of antifungal susceptibility testing. Clin. Microbiol. Rev. 19:435-447. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Pfaller, M. A., S. A. Messer, L. Boyken, C. Rice, S. Tendolkar, R. J. Hollis, and D. J. Diekema. 2007. The use of fluconazole as a “surrogate marker” to predict susceptibility and resistance to voriconazole among 13,338 clinical isolates of Candida spp. tested by Clinical and Laboratory Standards Institute recommended broth microdilution methods. J. Clin. Microbiol. 45:70-75. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Revankar, S. G., W. R. Kirkpatrick, R. K. McAtee, A. W. Fothergill, S. W. Redding, M. G. Rinaldi, and T. F. Patterson. 1998. Interpretation of trailing endpoints in antifungal susceptibility testing by the National Committee for Clinical Laboratory Standards method. J. Clin. Microbiol. 36:153-156. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Rex, J. H., M. A. Pfaller, J. N. Galgiani, M. S. Bartlett, A. Espinel-Ingroff, M. A. Ghannoum, M. Lancaster, F. C. Odds, M. G. Rinaldi, T. J. Walsh, and A. L. Barry. 1997. Development of interpretive breakpoints for antifungal susceptibility testing: conceptual framework and analysis of in vitro-in vivo correlation data for fluconazole, itraconazole, and Candida infections. Clin. Infect. Dis. 24:235-247. [DOI] [PubMed] [Google Scholar]
- 31.Rex, J. H., P. W. Nelson, V. L. Paetznick, M. Lozano-Chiu, A. Espinel-Ingroff, and E. J. Anaissie. 1998. Optimizing the correlation between results of testing in vitro and therapeutic outcome in vivo for fluconazole by testing critical isolates in a murine model of invasive candidiasis. Antimicrob. Agents Chemother. 42:129-134. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Rex, J. H., M. A. Pfaller, T. J. Walsh, V. Chaturvedi, A. Espinel-Ingroff, M. A. Ghannoum, L. L. Gosey, F. C. Odds, M. G. Rinaldi, D. J. Sheehan, and D. W. Warnock. 2001. Antifungal susceptibility testing: practical aspects and current challenges. Clin. Microbiol. Rev. 14:643-658. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Tortorano, A. M., A. L. Rigoni, E. Biraghi, A. Prigatino, M. A. Viviani, et al. 2003. The European Confederation of Medical Mycology (ECMM) survey of candidemia in Italy: antifungal susceptibility patterns of 261 non-albicans Candida isolates from blood. J. Antimicrob. Chemother. 52:679-682. [DOI] [PubMed] [Google Scholar]