Candidiasis is the most common fungal infection of blood [1, 2]. Candidemia in hospitalized patients is associated with a higher attributable mortality, ranging from 5%–71%, a significantly increased cost of healthcare, and an increased length of hospital stay [3, 4]. Despite advances in healthcare, invasive candidiasis-associated mortality has remained stable for the past 2 decades, although a decreased mortality rate has been observed for other fungal infections e.g. invasive aspergillosis [5].
Fluconazole remains the initial choice of therapy for candidemia as recommended by the Infectious Disease Society of America [7]. However, fluconazole-resistant strains, particularly of non-
The broth microdilution (BMD) method is the standard approved reference method for
The aim of the present study was to evaluate the performance of DD and ET assays compared with the standard BMD method, and to validate the accuracy of these agar-based methods for initial routine laboratory screening for antifungal susceptibilities of local strains in Thailand.
A total of 63 isolates of
All isolates were cultured on chromogenic agar (CHROMagar, Paris, France) for presumptive identification and to exclude contaminants. Species identity of isolates was confirmed using an API 20C Aux yeast identification system (bioM rieux, Montalieu, France) [13]. Additional tests to confirm the identity of species included morphological examinations such as the germ tube test, chlamydospore production test, and colony morphology observation.
All
Three laboratory technicians determined MICs independently at 24 h and 48 h after incubation at 35°C Because MIC values at 24 h and 48 h were found to be comparable (data not shown) we used only the 48 h readings for this analysis. The MIC values were based on consensus of at least 2 of the 3 technicians, otherwise the susceptibility test was repeated until a consensus was established. Interpretation of susceptibility was performed by applying the clinical breakpoints (CBPs) defined by CLSI document M27-A3 [9].
An inoculum of 1–5 × 106 cells/ml was used and applied to the agar plates (RPMI 1640 with 2% glucose) according to the manufacturer’s instructions. E-test strips (bioM rieux) were then applied. MICs were determined after 24 h and 48 h of incubation and interpreted according to CLSI document M27-A3 [9].
An inoculum of 1–5 × 106 cells/ml was used and disks applied to Mueller-Hinton agar with 2% glucose according to the manufacturer’s instructions (Bio-Rad Life Science, Marnes-la-Coquette, France). The plates were incubated at 35°C and read at 24 h and 48 h. Inhibitory zone diameters were interpreted according to CLSI document M44-A at the transitional point where growth abruptly decreased [14].
All methods were validated using quality control strains of
All statistical analyses were performed using IBM SPSS Statistics 21.0 for Mac (IBM Corp, Armonk, NY, USA). Descriptive statistics were used to present the data. The MIC values were reported as a range.
For the three-way comparison of BMD, ET, and DD, correlations were determined as recommended by the CLSI: (i) the definition of the occurrence of a very major error (VME) is when one strain is susceptible by the test methods (DD or ET), but is resistant by the BMD reference method; (ii) major errors (MEs) are defined as occurring when an isolate is found to be resistant by the test methods (ET or DD), but is susceptible by the BMD reference method; (iii) a minor error (mE) is indicated when an isolate is susceptible and dose dependent by one method, but susceptible or resistant by other methods. The total percentage agreement was calculated by the proportion of all test isolates that are in agreement, minus any errors, divided by the total number of test isolates.
A total of 63
Antifungal susceptibilities of 63
Antifungal agent | MIC range (μg/mL) | No. (%) of isolates indicated susceptibility category by CBPs | |||
---|---|---|---|---|---|
S | S-DD | R | |||
Fluconazole | 0.06–2 | 32(100) | 0 | 0 | |
0.06–64 | 13(93) | 0 | 1(7) | ||
0.125–4 | 12(100) | 0 | 0 | ||
0.375–4 | 4(100) | 0 | 0 | ||
1–1 | 1 (100) | 0 | 0 | ||
Voriconazole | 0.06–2 | 29(91) | 3(9) | 0 | |
0.03–1 | 11(79) | 2(14) | 1(7) | ||
0.06–1 | 12(100) | 0 | 0 | ||
1–4 | 2(50) | 1(25) | 1(25) | ||
0.5–0.5 | 1 (100) | 0 | 0 |
MIC, minimum inhibitory concentration; CBPs, clinical breakpoints defined by CLSI document M27-A3 [9]: S, susceptible; S-DD, susceptible-dose dependent; R, resistant.
The BMD-MIC values showed good correlation with ET-MIC values for fluconazole and voriconazole (r = 0.95;
The total percentage agreement of the ET and DD determinations, compared with the standard reference BMD-MIC values for the two antifungal agents, exceeded 90% as summarized in
Errors with the E-test and disk diffusion methods compared with broth microdilutions for fluconazole and voriconazole
Method | Category | Fluconazole by BMD | No. of strains | Total agreement (%) | Voriconazole by BMD | No. of strains | Total agreement (%) | ||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
S | I | R | VME | ME | mE | S | I | R | VME | ME | mE | ||||
S | 61 | 0 | 0 | 0 | 0 | 1 | 98 | 54 | 2 | 0 | 0 | 0 | 5 | 92 | |
I | 1 | 0 | 0 | 1 | 4 | 2 | |||||||||
R | 0 | 0 | 1 | 0 | 0 | 0 | |||||||||
S | 57 | 0 | 0 | 0 | 1 | 4 | 92 | 55 | 1 | 0 | 0 | 0 | 1 | 98 | |
I | 4 | 0 | 0 | 0 | 5 | 0 | |||||||||
R | 1 | 0 | 1 | 0 | 0 | 2 |
BMD, broth microdilution; ET, E-test; DD, disk diffusion; VME, very major error; ME, major error; mE, minor error; S, susceptible; I, susceptible-dose dependent; R, resistant.
The percentage agreement for the ET assays ranged between 92% to 98% and for DD between 92% to 98%. No VMEs were detected and only one ME was identified for a C.
Determining the species distribution and susceptibility profiles of the yeast isolated from cases of invasive candidiasis is essential to guide physicians in the choice of antifungal agents in any empirical treatment, because these susceptibility profiles vary geographically and not all hospitals perform antifungal susceptibility tests routinely. Therefore, this information is crucial and currently lacking for Thailand and other places in South-east Asia. The last published data from Thailand were in 2002 [15]. The present study provides an incidence and susceptibility profile for
The predominant species was
BMD as the standard antifungal testing method recommended by the CLSI can be difficult to perform in routine laboratory practice because of its complexity, the need of experienced laboratory technicians, and labor-intensive nature. By contrast, agar-based testing methods, (in particular DD, which has a ten-fold lower cost than ET) are commercially available, less expensive than BMD, easy to perform, and widely available, making them feasible as cost-effective screening methods. Our study demonstrated that BMD-MIC values showed good correlations with both ET-MIC and DD-MIC values. For the ET results, no VME was observed and the ME rates (92%–98%) were within the CLSI limits recommended in their document M-27A3 [9]. The DD test results also demonstrated an excellent performance (92%–98%) and no VME was observed. Only 1 ME was found for fluconazole against C.
We found that the agar-based ET and DD assays are reliable alternatives to the standard reference BMD method for isolates that are susceptible to fluconazole and voriconazole, because they showed >90% agreement with the BMD method. The present study validates these agar-based methods for use as initial screening tools in Thailand [11, 12, 20]. A limitation of our study is that we only selected
In summary, DD is a valid alternative method to BMD for testing