Folia Microbiol. 53 (2), 153–160 (2008)
http://www.biomed.cas.cz/mbu/folia/
Comparison of Disk Diffusion Test and Etest for Voriconazole and Fluconazole Susceptibility Testing V. BUCHTAa,b *, M. VEJSOVÁa,b, L.A. VALE-SILVAb ** aDepartment of Clinical Microbiology, Teaching Hospital and Faculty of Medicine, Charles University,
500 05 Hradec Králové, Czechia
bDepartment of Biological and Medical Sciences, Faculty of Pharmacy, 500 05 Hradec Králové, Czechia
Received 19 September 2007 Revised version 13 February 2008
ABSTRACT. The distribution for voriconazole and fluconazole susceptibility was determined by Etest and disk diffusion test in 143 clinical isolates. The majority of the strains of Aspergillus spp., Candida krusei, C. inconspicua, C. norvegensis and Saccharomyces cerevisiae displayed resistance or decreased susceptibility to fluconazole in contrast to voriconazole. The absolute categorical agreement for voriconazole and fluconazole susceptibility results by the disk method and Etest was 90.5 and 74.8 %, respectively. The error rate bounding analysis showed only 0.7 % of false susceptible results (very major error) with voriconazole, but 2.8 % with fluconazole. Fluconazole can be used as a surrogate factor to predict voriconazole susceptibility but with lower reliability for susceptible–dose dependent and resistance category, especially in Candida glabrata isolates. The results of the disk method were not substantially influenced by the composition of media (Mueller–Hinton agar vs. Antimycotic Sensitivity Test agar), even if with the latter the results had fewer tendencies to produce false susceptibility of C. glabrata isolates to both of the triazole drugs. Disk test as well as Etest were shown to represent suitable methods for routine evaluation of susceptibility of clinical isolates of pathogenic fungi, including aspergilli, to fluconazole and voriconazole.
Opportunistic fungal infections represent a real threat to life in a continuously growing group of immunocompromised patients, especially those hospitalized at transplant and hematology departments (Richardson 2005). Given the aggressive course of invasive mycoses, as well as their difficult diagnosis and treatment, high mortality of patients in this setting is not surprising. While candidiasis and aspergillosis account for the majority of opportunistic mycoses, other yeasts, including non-albicans Candida species, and new emerging filamentous fungi (e.g., Fusarium, Scedosporium, Mucorales) are being encountered with increasing frequency in immunocompromised patients (Ellis 2002). Since the turn of the millennium, therapeutic options have substantially improved as compared with the previous period because of the growing number of systemic antifungal drugs (voriconazole, posaconazole, echinocandins) suitable for the treatment of invasive systemic mycoses (Cappelletty et al. 2007; Petrikkos and Skiada 2007). Despite the standardization of antifungal susceptibility testing (see, e.g., Růžička et al. 2007), the establishment of interpretive criteria is limited to fluconazole, itraconazole, flucytosin, and most recently to voriconazole (Rex et al. 1997; Pfaller et al. 2006a,b). Voriconazole is a second-generation broad-spectrum triazole derivative, available in both peroral and intravenous forms. This drug is primarily used in the treatment of invasive fungal infections, in particular aspergillosis (Scott and Simpson 2007). Recently, clinical and laboratory mycologists have focused their efforts on the establishment of voriconazole breakpoints, which would further specify the possibilities of voriconazole choice and hence make the therapy more effective (Pfaller et al. 2006a). The goal of this study was to compare the Etest as a reference method and the disk diffusion test as a qualitative method for laboratory evaluation of the susceptibility of fungal isolates to voriconazole in comparison with fluconazole.
MATERIALS AND METHODS Fungal strains. Yeast isolates (n = 121) and isolates of filamentous fungi (n = 28) recovered from patients hospitalized in the University Hospital in Hradec Králové included: Candida albicans (n = 17), C. tro*Corresponding author; fax +420 495 832 019, e-mail
[email protected] . **Present address: Microbiology Service, Faculty of Pharmacy, University of Porto, 4050-047 Portugal.
154
V. BUCHTA et al.
Vol. 53
picalis (12), C. glabrata (12), C. krusei (12), C. parapsilosis (11), C. lusitaniae (12), C. kefyr (9), C. inconspicua (10), C. norvegensis (9), Saccharomyces cerevisiae (10), Trichosporon spp. (4), Blastoschizomyces capitatus (3), Aspergillus fumigatus (13), A. niger (7), A. flavus (6), and A. terreus (2). The testing panel also included 6 strains from the American Type Culture Collection (ATCC): C. albicans ATCC 44859, C. albicans ATCC 90028, C. glabrata ATCC 90030, C. krusei ATCC 6258, C. parapsilosis ATCC 22019 and A. fumigatus ATCC 36607. Yeast isolates were identified by standard laboratory tests (germ tube test, growth on chromogenic and corn-meal agar, auxanograms) and ID 32C (bioMeriéux, Czechia), and Aspergillus isolates by their macroscopic features and microscopy. All yeast and mould isolates were stored as suspensions at –70 °C or in sterile distilled water at room temperature, respectively, until susceptibility tests were performed. Before testing, each isolate of yeast and mould was subcultured on Sabouraud dextrose agar (Difco) and potato dextrose agar (Difco), respectively, and incubated at 35 °C to ensure purity and sporulation. Quality control was assured according to the M27-A2 (NCCLS 2002) and M44-A (NCCLS 2004) Clinical and Laboratory Standards Institute (CLSI) documents using Candida krusei ATCC 6258, C. parapsilosis ATCC 22019, and Aspergillus fumigatus ATCC 36607 as reference strains. Inoculum suspension. Yeast and spore suspensions were prepared in sterile 0.85 % NaCl from fresh colonies grown on SGA at 35 °C for 1 and 4 d, respectively. Turbidity was measured spectrophotometrically (Densi-La-Meter; Pliva–Lachema, Czechia), and the cell concentration was adjusted to correspond to 1.0– 5.0 × 106 yeast cells per mL and 0.5–4.5 × 106 spores per mL. These suspensions were used directly for the inoculation of agar plates. Antifungal agents. Etest strips were obtained from AB Biodisk (Sweden), the drug concentration being from 2 to 32 μg/mL and 16 to 256 μg/mL for voriconazole and fluconazole, respectively. Voriconazole 1-μg disks and fluconazole 25-μg disks were used (ITESTplus, Czechia). Media. Disk diffusion method and Etest were performed on Mueller–Hinton (MH) agar (Difco) with 2 % glucose, 0.5 μg/mL methylene blue (MB); pH of 7.3 ± 0.1. Antimycotic sensitivity test (AST) agar (HiMedia, BioVendor, Czechia) was prepared according to the instructions for use (pH 6.6 ± 0.2). Etest and disk diffusion test. Each fungal strain was tested on MH agar and AST agar. The agar plates were inoculated by overflooding with the fungal inoculum and the excess of the suspension was removed. The inoculated plates were allowed to dry for 15–30 min before the disks or Etest strips were aseptically applied onto the surface, and were subsequently incubated in ambient air at 35 °C, with readings taken after 24 and 48 h. In disk diffusion method, three disks were placed on a plate of MH or AST agar. The diameter of the inhibitory zone was measured after 24 and 48 h at the point where there was an abrupt decrease of growth. Geometrical mean of the diameters of the inhibitory zones determined in triplicate was calculated. Voriconazole susceptibility categories for Etest and the disk test were as follows: S – susceptible (MIC ≤ 1 μg/mL, zone diameter ≥17 mm), SDD – susceptible–dose dependent (MIC 1.5–3.0 μg/mL, zone diameter 14–16 mm), R – resistant (MIC ≥ 4 μg/mL, zone diameter ≤13 mm). Fluconazole susceptibility categories for Etest and disk test were as follows: S – susceptible (MIC ≤ 8 μg/mL, zone diameter ≥19 mm), SDD – susceptible–dose dependent (MIC 12–48 μg/mL, zone diameter 15–18 mm); R – resistant (MIC ≥ 64 μg/mL, zone diameter ≤14 mm). Data analysis. The diameters of the inhibition zones (in mm) surrounding the antimycotic disks after a 24-h incubation were plotted against their respective Etest MIC values (μg/mL) read after a 48-h interval. The overall categorical agreement between the disk diffusion test and Etest MIC results was determined for voriconazole-to-fluconazole using the above MIC interpretive categories. Major errors were classified as resistant by the disk diffusion test and susceptible by Etest, very major errors (false susceptibility) as susceptible by the disk diffusion method and resistant by Etest, and minor errors corresponded to one of the tests having been susceptible or resistant and the other test susceptible–dose dependent. Predictive values (%) were determined for categorical agreement of disk test and Etest as the probability of a result of susceptible by the disk test to correspond with a result of susceptible or susceptible–dose dependent by Etest, and a result of susceptible–dose dependent-to-resistant by disk test to correspond with susceptible–dose dependentto-resistant by Etest (Matar et al. 2003; Pfaller et al. 2007). Quality control isolates C. parapsilosis ATCC 22019, C. krusei ATCC 6258, and ATCC 36607 were included, and the results were within the published limits (NCCLS 2002, 2004). In case of A. fumigatus ATCC 36607, inhibition zones were 18 mm and 14 mm on AST and 22 mm and 17 mm on MH for voriconazole after 24 and 48 h, respectively, and negligible on both media for fluconazole (90 % strains tested were voriconazole susceptible. On AST agar, more voriconazole SDD and resistant strains could be detected; these results correlated better with Etest results after 2 d (not shown). Regardless of the fact that a relatively low number of C. glabrata strains were tested (n = 12), the tendency of both methods to lead to false susceptibility in this case is evident. This susceptibility profile suggests the propensity of C. glabrata for the cross-resistance to triazole antifungals and, therefore, that an increased risk of the resistance development of yeasts to new triazoles can be expected in the future (Pfaller et al. 2003; Burn et al. 2004; Pfaller and Diekema 2004). This fact corresponds to the relatively higher MICs or smaller inhibition-zone diameters obtained for C. glabrata: the values in the category susceptible are close to the limit for the SDD category, especially for fluconazole and less so for voriconazole (Pfaller et al. 2006b; Swine et al. 2004). Regression analysis showed a relatively low correlation coefficient obtained by comparing the values of both tests for a given strain (r = 0.85). This may be due to a variable susceptibility of C. glabrata strains as described above and/or a relatively high rate (17.2 %) of non-albicans Candida isolates in our setting and/or lower numbers of all strains tested. On the other hand, if the values of the inhibition zone were averaged for all isolates with the same MIC value obtained by Etest, the correlation was substantially increased (r = 0.98). The composition of the cultivation medium (MH agar vs. AST agar) did not have an essential influence on the agreement of results of voriconazole susceptibility obtained by both methods as long as the inhibition zones were measured after 24 h (89.9 % overall agreement), with the agreement for C. glabrata being low again (41.7 %). In contrast, the correspondence of the values decreased significantly after 48 h (79.9 %), in particular due to a low degree of agreement for C. glabrata (50.0 %) and also A. niger (42.9 %) and A. flavus (33.3 %). Full agreement was recorded in the disk diffusion test readings for all Aspergillus strains tested after 24 h. These discrepancies of voriconazole testing obtained with aspergilli on AST agar showed that most strains switched the category from susceptible to resistant, which was not observed on MHA (Fig. 1). In the case of fluconazole, the agreement of Etest and disk test results was relatively low (74.3 %), probably due to the resistance or decreased susceptibility of non-albicans Candida species, especially C. glabrata, C. inconspicua, C. norvegensis and, also, partially C. parapsilosis and C. tropicalis. Unlike with voriconazole, the time of the reading (24 vs. 48 h) of the fluconazole inhibition zone had no significant effect on the global agreement. A relatively large proportion of intrinsically resistant species to fluconazole (Aspergillus spp., C. krusei, C. norvegensis) involved in the experiments could contribute to the results. Regression analysis provided a correlation coefficient (r = 0.86) similar to the voriconazole one, including the comparison of the averages of the disk-test values obtained for the strains with identical Etest MIC (r = 0.95). From the point of view of application, the differences in the disk-test values on AST medium are not that important, since the purpose of the test is a presumptive determination of susceptibility which has to be, in the case of need, confirmed by a MIC determination method. The advantage of AST agar is the possibility of obtaining more reliable (less false susceptible) results after 24 h, with better correlation with Etest testing after 48 h. The evaluation of a possible use of fluconazole as a surrogate factor to predict voriconazole susceptibility and resistance showed very good prediction values for the category susceptible but not for the categories SDD and resistant (Tables II and III). This poor agreement for the resistant category and higher number
156
V. BUCHTA et al.
Vol. 53
2008
VORICONAZOLE AND FLUCONAZOLE SUSCEPTIBILITY TESTING
157
158
V. BUCHTA et al.
Vol. 53
2008
VORICONAZOLE AND FLUCONAZOLE SUSCEPTIBILITY TESTING
159
of major errors (false resistance) was associated with resistance or a decreased susceptibility to fluconazole of Aspergillus spp., C. krusei, C. glabrata, C. inconspicua, and C. norvegensis (Table III). As regards fungal species, the disk diffusion test used for testing both fluconazole and voriconazole provided the best results, in terms of agreement as well as predictive value, for the category susceptible. The SDD strains were relatively better detected with Etest for voriconazole susceptibility testing. Even if fluconazole performance as a surrogate marker was improved, merging the susceptible and SDD categories as being susceptible to voriconazole, no considerable effect on predictive values (PS+) was observed, with the exception of fungal species with decreased susceptibility to fluconazole, including C. glabrata. The opposite effect was noticed if the same improvement was made for the SDD category in voriconazole (PSDD+). Fluconazole less susceptible fungal species tested were absolutely not associated with the prediction of cross-resistance between fluconazole and voriconazole, especially when the disk diffusion test is used. One reason could consist in the higher antifungal potency of voriconazole which can be illustrated on the discrepancy in the categories SDD and/or resistant, i.e. strains which are SDD or resistant to fluconazole are less susceptible to voriconazole but still within the category susceptible (although with a relatively higher value of MIC or smaller diameter of inhibition zone). Our results are quite different from those of Pfaller et al. (2007), who report 91.6 % of total categorical agreement and no false susceptible results. This may be explained by the fact that they evaluated Candida species only and included much more isolates from the Global Antifungal Surveillance Program (i.e. no Aspergillus and Saccharomyces species). The rate of non-albicans Candida isolates with resistance and decreased susceptibility to fluconazole (C. krusei, C. glabrata, C. inconspicua and C. norvegensis) was thus much lower (17.1 vs. 28.7 %) than in our study. Voriconazole is considered the drug of first choice in the therapy of invasive aspergillosis, and it constitutes a suitable alternative in the case of mycoses caused by fungal species or strains with decreased susceptibility to fluconazole. With some non-albicans Candida yeasts in particular, a considerable intra- or interspecific variability, and, like with fluconazole, the possibility of resistance development as a result of long term administration should be taken into account. In those cases, data on in vitro susceptibility can help in the choice of an adequate antifungal therapy. The use of fluconazole as a surrogate factor to predict voriconazole susceptibility should be carefully considered, in particular when fungal species with decreased susceptibility to fluconazole are tested. This work was partially supported by grant NI/7464-2 of the Ministry of Health of the Czech Republic. REFERENCES BURN A.K., FOTHERGILL A.W., KIRKPATRICK W.R., COCO B.J., PATTERSON T.F., MCCARTHY D.I., RINALDI M.G., REDDING S.W.: Comparison of antifungal susceptibilities to fluconazole and voriconazole of oral Candida glabrata isolates from head and neck radiation patients. J.Clin.Microbiol. 42, 5846–5848 (2004). CAPPELLETTY D., EISELSTEIN-MC KITRICK K.: The echinocandins. Pharmacotherapy 27, 369–388 (2007). ELLIS M.: Invasive fungal infections: evolving challenges for diagnosis and therapeutics. Mol.Immunol. 38, 947–957 (2002). JOHNSON L.B., KAUFFMAN C.A.: Voriconazole: a new triazole antifungal agent. Clin.Infect.Dis. 36, 630–637 (2003). MATAR M.J., OSTROSKY-ZEICHNER L., PAETZNICK V.L., RODRIGUEZ J.R., CHEN E., REX J.H.: Correlation between E-test, disk diffusion, and microdilution methods for antifungal susceptibility testing of fluconazole and voriconazole. Antimicrob.Agents Chemother. 47, 1647–1651 (2003). NCCLS (National Committee for Clinical Laboratory Standards): Reference Method for Broth Dilution Testing of Yeasts, Approved Standard M27-A2, 2nd ed. NCCLS, Wayne (USA) 2002. NCCLS (National Committee for Clinical Laboratory Standards): Method for Antifungal Disk Diffusion Susceptibility Testing of Yeasts: Approved Guideline M44-A. NCCLS, Wayne (USA) 2004. PETRIKKOS G., SKIADA A.: Recent advances in antifungal chemotherapy. Internat.J.Antimicrob.Agents 30, 108–117 (2007). PFALLER M.A., D IEKEMA D.J.: Rare and emerging opportunistic fungal pathogens: concern for resistance beyond Candida albicans and Aspergillus fumigatus. J.Clin.Microbiol. 42, 4419-4431 (2004). PFALLER M.A., D IEKEMA D.J., BOYKEN L., MESSER S.A., TENDOLKAR S., HOLLIS R.J.: Evaluation of the Etest and disk diffusion methods for determining susceptibilities of 235 bloodstream isolates of Candida glabrata to fluconazole and voriconazole. J.Clin.Microbiol. 41, 1875–1880 (2003). PFALLER M.A., DIEKEMA D.J., REX J.H., ESPINEL-INGROFF A., JOHNSON E.M., ANDES D., CHATURVEDI V., GHANNOUM M.A., ODDS F.C., RINALDI M.G., SHEEHAN D.J., TROKE P., WALSH T.J., WARNOCK D.W.: Correlation of MIC with outcome for Candida species tested against voriconazole: analysis and proposal for interpretive breakpoints. J.Clin.Microbiol. 44, 819–826 (2006a). PFALLER M.A., D IEKEMA D.J., SHEEHAN D.J.: Interpretive breakpoints for fluconazole and Candida revisited: a blueprint for the future of antifungal susceptibility testing. Clin.Microbiol.Rev. 19, 435-447 (2006b). PFALLER M.A., MESSER S.A., BOYKEN L., RICE C., TENDOLKAR S., HOLLIS R.J., D IEKEMA D.J.: 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 (2007).
160
V. BUCHTA et al.
Vol. 53
REX J.H., PFALLER M.A., G ALGIANI J.N., BARTLETT M.S., ESPINEL-I NGROFF A., G HANNOUM M.A., LANCASTER M., ODDS F.C., RINALDI M.G., WALSH T.J., BARRY A.L.: 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. Subcommittee on Antifungal Susceptibility Testing of the National Committee for Clinical Laboratory Standards. Clin.Infect. Dis. 24, 235–247 (1997). RICHARDSON M.D.: Changing patterns and trends in systemic fungal infections. J.Antimicrob.Chemother. 56 (Suppl. 1), i5-i11 (2005). RŮŽIČKA F., H OLÁ V., VOTAVA M., TEJKALOVÁ R.: Importance of biofilm in Candida parapsilosis and evaluation of its susceptibility to antifungal agents by colorimetric method. Folia Microbiol. 52, 209–214 (2007). SCOTT L.J., SIMPSON D.: Voriconazole: a review of its use in the management of invasive fungal infections. Drugs 67, 269–298 (2007). SWINNE D., WATELLE M., VAN DER FLAES M., NOLARD N.: In vitro activities of voriconazole (UK-109, 496), fluconazole, itraconazole and amphotericin B against 132 non-albicans bloodstream yeast isolates (CANARI study). Mycoses 47, 177–183 (2004).
