Rapid Susceptibility Testing of Medically Importan
http://www.100md.com
微生物临床杂志 2006年第2期
ABSTRACT
The XTT colorimetric assay quantifies fungal growth by measuring fungal metabolism and has been used successfully for susceptibility testing of Aspergillus species after 24 and 48 h of incubation. In the present study using 14 clinical isolates of Zygomycetes (Rhizopus oryzae [5 isolates], Cunninghamella spp. [3 isolates], Mucor spp. [3 isolates], and Absidia corymbifera [3 isolates]), significant metabolic activity was demonstrated before visual or spectrophotometric detection of fungal growth by performing the XTT assay as early as 6 h after inoculation. Testing of susceptibility to amphotericin B, posaconazole, and voriconazole was subsequently performed using the XTT method (100 μg/ml XTT, 25 μM menadione) at 6, 8, or 12 h after inoculation and the CLSI (formerly NCCLS) M38-A method with visual and spectrophotometric MIC determinations at 24 h after inoculation. Concentration-effect curves obtained with the use of the Emax model (a sigmoid curve with variable slope) were comparable between the early XTT and spectrophotometric readings at 24 h. Complete inhibition of early metabolic activity with the azoles was delayed in comparison to that with amphotericin B. Using appropriate cutoff levels, agreement was demonstrated between the early XTT and 24-h spectrophotometric or visual readings. In particular, for MIC-0 (the lowest drug concentration showing absence of visual growth) of amphotericin B, overall agreement levels were 90 to 93% for the 6-h XTT assay and 100% for the 8- and 12-h time points. For MIC-0 of posaconazole, agreement levels were 86% for the 6-h XTT and 93 to 100% for the 8- and 12-h time points. The overall agreement levels for MIC-0 and MIC-2 (the lowest drug concentration showing prominent reduction of growth compared with the control well) of voriconazole (compared with 24-h spectrophotometric readings) were 93 to 98% for the 8- and 12-h XTT assays. These results support the use of the XTT method for rapid MIC determination for Zygomycetes.
INTRODUCTION
Infections caused by Zygomycetes have emerged as an increasingly important cause of morbidity and mortality among severely immunocompromised patients, including those with hematological malignancies and hematopoietic stem cell transplant recipients. The prognosis of zygomycosis in these patients remains extremely poor, despite antifungal treatment (10, 16, 18, 19). For years, amphotericin B has been the drug of choice for these devastating infections. Of the recently introduced second-generation triazoles, voriconazole is not active against the Zygomycetes, while posaconazole has demonstrated activity in vitro, in animal models, and in case reports (4-6, 8, 20-22). As zygomycosis may be rapidly fatal, timely determination of the susceptibilities of clinical isolates to these antifungal agents is important.
A colorimetric method, based on reduction of the tetrazolium salt 2,3-bis{2-methoxy-4-nitro-5-[(sulfenylamino) carbonyl]-2H-tetrazolium-hydroxide} (XTT) by mitochondrial dehydrogenases, has been recently introduced for in vitro susceptibility testing of filamentous fungi. This method quantifies fungal growth by measuring fungal metabolism, using appropriate concentrations of XTT and an electron transfer agent, such as menadione (1, 11, 12). For Aspergillus species, MICs determined by the XTT method at 24 and 48 h were comparable with those obtained by the National Committee for Clinical Laboratory Standards (now known as the Clinical and Laboratory Standards Institute [CLSI]) method (12). The sensitivity of the XTT assay may be increased depending on the menadione concentration (11). High menadione concentrations could increase the rate of XTT conversion and theoretically allow the detection of metabolic activity associated with small, early fungal growth that is not yet detected visually or spectrophotometrically. In this case, it is likely that the XTT method might even be useful for early determination of MICs of antifungal agents. This hypothesis has been previously discussed (12) but has not been investigated.
In preliminary studies using the XTT assay for zygomycete isolates, high metabolic activity was observed even in the presence of small fungal growth (assessed spectrophotometrically) at 12 or 24 h postinoculation and preceded increases in biomass (C. Antachopoulos, J. Meletiadis, E. Roilides, T. Sein, and T. J. Walsh, Abstr. 45th Intersci. Conf. Antimicrob. Agents Chemother., abstr. M-1003, 2005). The present study investigated whether significant metabolic activity with the XTT assay could be demonstrated for the Zygomycetes before hyphal growth is macroscopically or spectrophotometrically detected (at 6 or 8 h postinoculation) and whether changes in this early metabolic activity in the presence of antifungal agents could be used for rapid susceptibility testing of these filamentous fungi.
MATERIALS AND METHODS
Isolates. A collection of 14 clinical zygomycete isolates was used, including the following species: Rhizopus oryzae (5 isolates), Cunninghamella bertholletiae (2 isolates), Cunninghamella echinulata (1 isolate), Mucor circinelloides (2 isolates), Mucor indicus (1 isolate), and Absidia corymbifera (3 isolates). The reference strains Candida krusei 6258 and Candida parapsilosis 22019 were used as quality controls.
Medium. RPMI 1640 medium with L-glutamine but without bicarbonate (Cambrex Bio Science, Inc., Walkersville, MD) buffered to pH 7.0 with 0.165 M 3-N-morpholinopropanesulfonic acid (MOPS) (Sigma-Aldrich, St. Louis, MO) was used as the assay medium.
Inoculum. Sporangiospores were harvested after isolates were subcultured in potato dextrose agar at 35°C for 3 days and were suspended in normal saline containing 0.025% Tween 20. The sporangiospore suspensions were counted with a hemacytometer and then diluted in RPMI in order to achieve the desired concentrations. Inoculum sizes were verified by quantitative colony counts on Sabouraud dextrose agar plates.
XTT and menadione. XTT (Sigma-Aldrich, St. Louis, MO) was dissolved in normal saline at concentrations of 1, 0.5, and 0.25 mg/ml. Menadione (Sigma-Aldrich, St. Louis, MO) was initially dissolved in absolute ethanol at a concentration of 10 mg/ml and subsequently added to the above-mentioned XTT solutions at concentrations of 125, 31.25, and 7.81 μM for each solution. In this way, a total of nine solutions with different XTT-menadione concentrations were prepared.
XTT assay for measurement of early metabolic activity. In the XTT assay for measurement of early metabolic activity, the abilities of the above-mentioned XTT-menadione combinations to detect early metabolic activity of Zygomycetes were investigated. Briefly, 104 sporangiospores/ml of four clinical zygomycete isolates (one each of R. oryzae, C. bertholletiae, M.
circinelloides, and A. corymbifera) were incubated in RPMI 1640 at 37°C in 96-well microtitration plates at a volume of 200 μl/well. After 6 or 8 h of incubation, 50 μl of one of the above-mentioned XTT-menadione solutions was added to each well, as previously described (11, 12). All nine solutions were used, resulting in final concentrations of 200, 100, and 50 μg/ml XTT, each combined with 25, 6.25, or 1.56 μM menadione, respectively. The microtitration plates were further incubated at 37°C for 2 h in order to allow conversion of XTT to its formazan derivative. These plates were then shaken for 1 to 2 min (Wallac Plate Shake 1296-004; Wallac OY, Turku, Finland) until complete dissolution of formazan derivatives was achieved. XTT conversion was measured as optical density (OD) with a microtitration plate spectrophotometric reader (Elx808; Bio-Tek Instruments, Winooski, VT) at 450 nm. For each well, XTT conversion was calculated after subtraction of the background OD, which was the OD of a simultaneously incubated well with 200 μl of RPMI and 50 μl of XTT-menadione solution but no inoculum. Four replicates were used for each species, time point, and XTT-menadione concentration.
Relationship between early XTT conversion and fungal biomass. An assay investigated whether the rate of early XTT conversion, obtained with the different XTT-menadione combinations, in fact correlated with the magnitude of fungal biomass, using a previously described approach (11). Briefly, the XTT assay was performed at 6 or 8 h of incubation, as described above, using 10-fold serially diluted inocula (102 to 106 sporangiospores/ml) for each of the above-mentioned four clinical zygomycete isolates. The relationship between XTT conversion (OD at 450 nm [OD450]) and fungal inoculum (log10) was assessed by linear regression analysis, and the slopes and goodness of fit (r2) were reported for each XTT-menadione combination.
Antifungal-susceptibility testing. MICs were determined for all 14 isolates of Zygomycetes. The broth microdilution method of the CLSI (M38-A) (7, 15) was followed using 96-well microtitration plates. Posaconazole (Schering-Plough Research Institute, Kenilworth, NJ) was dissolved in dimethyl sulfoxide. Amphotericin B (Bristol-Myers Squibb, Princeton, NJ) and voriconazole (Pfizer Inc., New York, NY) were dissolved in sterile distilled water. Serial dilutions of the antifungal agents were prepared to yield concentrations of 0.03 to 32 μg/ml for posaconazole, 0.015 to 16 μg/ml for amphotericin B, and 1 to 1,024 μg/ml for voriconazole at a final volume of 200 μl after inoculation. The trays were maintained at –70°C until the day of testing.
Incubation and MIC determination. After the microtitration trays were defrosted, they were inoculated with a final concentration of 2.5 x 104 sporangiospores/ml from each of the 14 isolates and incubated at 37°C. For each isolate, four different 96-well trays were used, corresponding to four different incubation periods: 6, 8, 12, and 24 h. For each time point, an additional tray was prepared, whose wells contained the above-mentioned drug concentrations at a final volume of 200 μl of RPMI, but no fungal inoculum. These trays served to provide background OD measurements for the XTT and spectrophotometric readings.
(i) XTT assay for early MIC determination. At 6, 8, and 12 h after inoculation, the corresponding trays were removed from the incubator, and 50 μl of saline containing 0.5 mg/ml XTT with 125 μM menadione was added to each well to obtain final concentrations of 100 μg/ml XTT and 25 μM menadione. The trays were further incubated for 2 h at 37°C. After shaking of the trays, XTT conversion was measured at 450 nm. After subtraction of the corresponding background OD, the relative OD450 for each well and drug concentration, in relation to the control well, was calculated as the following fraction: OD450 of the drug-containing well/OD450 of the control well. Due to the lack of previous data regarding the effects of different antifungal agents on the early metabolic activity of Zygomycetes, appropriate cutoff levels for determination of MIC-0 and MIC-2 (see below) and comparison with the CLSI method were employed after the results were obtained and the concentration-effect curves of metabolic activity were analyzed.
(ii) CLSI method. After 24 h of incubation, the corresponding trays were removed from the incubator and fungal growth was assessed visually and spectrophotometrically. For visual assessment, a concave mirror was used. The MIC-0 was determined as the lowest drug concentration showing absence of visual growth and was recorded for all three agents. The MIC-2 was determined as the lowest drug concentration showing prominent reduction of growth compared with the control well and was recorded only for posaconazole and voriconazole. For spectrophotometric assessment, hyphal growth was measured at 405 nm, as previously described (5, 12). After subtraction of the corresponding background OD, the relative OD405 for each well and drug concentration, in relation to the control well, was calculated as follows: OD405 of the drug-containing well/OD405 of the control well. The lowest drug concentrations showing relative ODs of 10% and 50% were determined to be the MIC-0 and MIC-2, respectively. Again, only MIC-0 was recorded for amphotericin B.
Experiments of antifungal susceptibility testing using the early XTT and CLSI methods were repeated in triplicate.
Modeling and data analysis. For four of the zygomycete isolates (one each of R. oryzae, C. bertholletiae, M. circinelloides, and A. corymbifera), for the purpose of modeling, the relative OD values obtained with the XTT assay (6, 8, and 12 h) or spectrophotometric reading at 405 nm (24 h) were analyzed by nonlinear regression analysis, using a four-parameter logistic model (a sigmoid curve with variable slope) known as the Emax model. Deviation from the model was tested by the runs test, and goodness of fit was checked by the r2 value. The best-fit values of the EC50, which is the drug concentration producing 50% of the Emax (maximum relative OD); of the slope m, which describes the steepness of the curve; and of the bottom of the curve were recorded and compared for the early (6-, 8-, and 12-h) and 24-h time points for each isolate. Comparison of best-fit values for these parameters was performed with one-way analysis of variance, followed by Dunn's test for multiple comparisons. The results of comparisons of the EC50 and bottom-of-the-curve values were used to employ appropriate cutoff levels for determination of MIC-2 and MIC-0, respectively, with the early XTT method (6, 8, and 12 h). Statistical analysis was performed using the GraphPad (San Diego, Calif.) Prism software.
For purposes of comparison with CLSI methods, MICs for all 14 isolates obtained by using the above cutoff levels for the early XTT assays were compared with corresponding values from visual and spectrophotometric (405-nm) readings at 24 h. The percentage of relative (within one dilution) agreement between the MICs obtained with different methods and time points for each species and for all isolates was calculated and reported.
Reproducibility. In order to determine the reproducibility of the XTT, visual, and spectrophotometric methods for each time point and antifungal agent, the median MIC of those obtained with the three experiments was calculated for each isolate. The relative interexperimental agreement was subsequently defined as the percentage of MICs (obtained for each method, time point, and antifungal agent) that were within one twofold dilution from the median MIC (12).
RESULTS
Early metabolic activity. Significant metabolic activity was demonstrated with the XTT assay at the 6- and 8-h time points for all four isolates tested, with OD450s ranging from 0.17 to 1.06 for those wells treated with 200, 100, or 50 μg/ml XTT and 25 μM menadione at 6 h (Fig. 1). At this early time point, the use of different XTT concentrations resulted in comparable conversion rates. In contrast, the use of lower menadione concentrations (6.25 and 1.56 μM) resulted in markedly reduced XTT conversion, independently of the concentration of XTT (Fig. 1). Consequently, XTT-menadione combinations containing 25 μM of menadione were used exclusively in further assays of early metabolic activity.
Relationship between early XTT conversion and fungal biomass. A linear relationship was demonstrated between early XTT conversion and fungal inoculum (P < 0.001). Of the three combinations of 200, 100, and 50 μg/ml XTT with 25 μM menadione, a higher median (range) r2 was obtained with the concentrations of 100 μg/ml XTT and 25 μM menadione [0.91 (0.86 to 0.98)]. For the combinations of 200 or 50 μg/ml XTT with 25 μM menadione, the median (range) r2 values were 0.85 (0.76 to 0.95) and 0.88 (0.86 to 0.92), respectively. The slope [median (range)] of the regression line obtained with 100 μg/ml XTT was 0.42 (0.41 to 0.46). Use of 200 or 50 μg/ml XTT resulted in very high (median, 0.81; range, 0.66 to 0.84) or low (median, 0.22; range, 0.20 to 0.23) slopes, respectively. Since we aimed for a linear relationship with intermediate slope (high or low slopes would result in fast or slow XTT conversion, respectively), we employed the combination of 100 μg/ml XTT and 25 μM menadione for early antifungal susceptibility testing of zygomycete isolates.
Antifungal susceptibility testing. In agreement with the previous findings, sufficiently high XTT conversion rates (mean OD450, 0.69; range, 0.19 to 1.67) were obtained when the combination of 100 μg/ml XTT and 25 μM menadione was used for MIC determination at the 6-h time point for all 14 isolates. Examination of microtitration plates under the inverted microscope at the end of the 2-h incubation period following addition of the XTT-menadione solution (i.e., 8 h, in total, after inoculation) demonstrated that XTT conversion increased with progression from germination to the formation of hyphae (Fig. 2).
Modeling of XTT and spectrophotometric measurements. For the four isolates of Zygomycetes (one each of R. oryzae, C. bertholletiae, M. circinelloides, and A. corymbifera) for which data were analyzed by nonlinear regression analysis, the Emax model fit the spectrophotometric (24-h) and XTT (6-, 8-, and 12-h) readings very well; r2 ranged between 0.97 and 0.99, and no statistical deviations from the model were found (P = 0.17 to 0.99; runs test). No statistical differences were found between the best-fit values of EC50 obtained with the spectrophotometric and colorimetric methods for posaconazole and voriconazole (P > 0.05 by one-way analysis of variance) (Table 1). For amphotericin B, P was <0.05 for comparison of the 24-h spectrophotometric and 6- or 8-h XTT readings; however, the differences in EC50 values did not exceed 1 dilution (Table 1). The slopes of the concentration-effect curves were also similar between the different methods (P > 0.05) but varied according to the antifungal agent used: higher absolute m values (steeper curves) were obtained with amphotericin B and lower values were obtained with voriconazole (data not shown). A significant difference, however, between the spectrophotometric (405-nm) and early XTT methods was found in the best-fit values of the bottom of the curve for posaconazole and voriconazole (P < 0.05) (Table 1). For these agents, the bottom values of the curve obtained with the early XTT assays were significantly higher than those obtained with the 24-h spectrophotometric readings. Bottom values were highest at the 6-h time point and decreased gradually at 8 and 12 h (Table 1). This difference in the bottom values of the sigmoid curve between the early XTT and spectrophotometric methods for the azoles, in comparison to amphotericin B, is also illustrated in Fig. 3 and suggests that for the azole agents, significant XTT conversion could occur at the early time points (up to almost 40% of control OD450 for the 6-h time point), even in the presence of drug concentrations equal to or greater than the MIC-0. Thus, suppression of the metabolic activity of Zygomycetes by inhibitory concentrations of the azoles was gradually achieved and was not complete at the early time points.
The comparability of EC50 values obtained with the Emax model suggested that for determination of MIC-2 by the early XTT assays, the cutoff level of 50% (also employed for the spectrophotometric readings at 24 h) would probably provide good agreement between the different methods. However, the differences in the bottom values of the curve suggested that in order to determine MIC-0 for the azoles by the early XTT method, cutoff levels greater than 10% (which was used in the spectrophotometric method) should be employed. Based on the Emax model data (Table 1), these cutoffs could be on the order of 20 to 40% for the 6-h time point and 10 to 20% for the 8- and 12-h time points. Indeed, the MIC-0 values obtained with the above range of cutoff levels were compared with those obtained with spectrophotometric (405-nm) or visual readings, and the cutoffs giving the best agreement are presented in Table 2. For amphotericin B, MIC-0 values obtained using the 10% cutoff for the 6- and 8-h XTT method were in very good agreement with the spectrophotometric method (90 and 95%, respectively). However, the threshold of 5% increased the agreement for the 8-h time point to 100%.
Comparison of MICs. Comparable MICs of zygomycete isolates were obtained with the visual (CLSI), spectrophotometric (405-nm), and early XTT methods using the cutoffs shown in Table 2 for different time points. Differences in median MICs obtained with different methods for zygomycete species did not exceed 1 dilution, with the exception of median MICs of voriconazole for Mucor spp. and A. corymbifera, which were within 2 dilutions (Table 3).
The relative levels of agreement between the different methods are presented in Table 4. The overall agreement levels between the spectrophotometric (405-nm) and visual (CLSI) methods of MIC determination at 24 h ranged between 83 and 100% for different agents, with 83% being the agreement for MIC-0 of voriconazole and 100% being the agreement for MIC-0 of amphotericin B. Lower agreement levels were observed between the two methods in the determination of MICs of posaconazole and voriconazole against Cunninghamella spp. (Table 4). Whenever differences were observed between these two methods, visual MICs tended to be higher than spectrophotometric ones (Table 3).
For amphotericin B, the overall agreement levels of MICs obtained with the early XTT assays compared to the spectrophotometric and visual readings ranged from 90 to 93% for the 6-h time point to 100% for those of 8 and 12 h (Table 4). Lower agreement levels were observed for Cunninghamella spp. with the 6-h XTT assay.
For posaconazole, the overall agreement was 86% for MIC-0 obtained with the 6-h XTT versus spectrophotometric readings and even higher for the 8- and 12-h time points (93 to 100% for MIC-0) (Table 4). Lower agreement levels, however, were observed between the 6-h XTT and visual MIC-0 (76%) and between the 6-h XTT and spectrophotometric MIC-2 (81%). Again, among different species, lower agreement levels were observed in the determination of MICs for Cunninghamella spp., in particular, MIC-2.
For voriconazole, the MIC-0 and MIC-2 obtained by the 8- and 12-h XTT methods showed an overall agreement of 93 to 98% with the spectrophotometric method but showed less with the visual readings (71 to 86%) (Table 4). The overall agreement levels of MIC-2 determined by the 6-h XTT method for voriconazole were 88% compared with the spectrophotometric method and slightly lower (83%) with the visual method. Lower agreement levels, however (67 to 74%), were observed for the MIC-0 values obtained with the 6-h XTT compared to visual or spectrophotometric readings.
Reproducibility. The relative interexperimental agreement levels for MIC-0 and MIC-2 ranged between 98 and 100% for all methods, time points, and antifungal agents, with the exception of MIC-0 determined by the 6-h XTT method for voriconazole and MIC-2 determined by the 24-h spectrophotometric method for posaconazole, for which the agreement was 95% (standard error, 3%).
DISCUSSION
In the present study, we demonstrated that MICs of Zygomycetes can be obtained in a reproducible way with the XTT assay up to 16 h earlier than with current standard methods. This early determination of MICs was based on the demonstration of significant metabolic activity of these fungi before detection of significant growth macroscopically or spectrophotometrically (mean OD405 for zygomycete isolates at 8 h postinoculation, 0.008; range, –0.004 to 0.03).
Several methods for rapid susceptibility testing of filamentous fungi have been developed in the past, including a radiometric assay measuring the inhibition of 14CO2 production (14), a flow cytometry assay of conidial viability (3), a fluorescence-based microplate assay (2), and a turbidometric method assessing the lag phases of fungal growth curves in the presence of antifungal drugs (13). Some of these methods may use expensive equipment or potentially hazardous substances, and they have been evaluated only for amphotericin B. The fluorescence-based microplate assay generated MICs in 16 h for Aspergillus spp., but it remains to be evaluated for other filamentous fungi, including Zygomycetes (2). The use of turbidometric growth curves required up to 12 h of incubation for determination of the amphotericin B resistance of R. oryzae isolates (13). The XTT method presented here is a simple way of measuring fungal metabolism that can be performed in most laboratories.
When menadione is used as an electron transfer agent, the rate of conversion of XTT to its formazan derivative could be affected by the concentrations of both the XTT tetrazolium salt and menadione (11). As shown in Fig. 1, at the early time point of 6 h, the effect of menadione on XTT conversion seemed to be more significant than that of the XTT concentration. A high (25 μM) menadione concentration was required to obtain significant conversion rates. It should be emphasized, however, that while the use of the XTT assay for rapid determination of MICs could probably be evaluated for other fungi as well, the concentrations of XTT and menadione that should be used may need to be modified, due to potential differences in growth properties and metabolic activities among different classes or genera.
Another factor affecting XTT conversion at these early time points is the size of the inoculum, which correlated linearly with the OD450 values obtained. The results from our studies of early metabolic activity (Fig. 1), as well as from the subsequent antifungal susceptibility studies (where the range of viable inoculum, based on quantitative colony counts, was 0.7 x 104 to 3.9 x 104 sporangiospores/ml), suggested that, within the range of inocula suggested by the CLSI guidelines (0.4 x 104 to 5 x 104 sporangiospores/ml) (15), sufficient XTT conversion occurred at the early time points of 6 and 8 h to allow MIC determination. However, deviations beyond the lower limit of the suggested range should probably be avoided, since they may result in significantly reduced XTT conversion (with 103 sporangiospores/ml, an OD450 as low as 0.04 was recorded for some isolates [data not shown]).
Although CLSI recommendations for filamentous fungi suggest reporting only MIC-0 for the azoles (15), others have also used MIC-2 (5, 8). Hence, we determined both MIC-0 and MIC-2 for these agents. For determination of MICs at 24 h, we used both visual and spectrophotometric (405-nm) readings of the trays. Spectrophotometric readings may be more precise and reproducible than visual assessment of fungal growth, which is often subjective (17). For spectrophotometric determination of MIC-0, thresholds of both 10% and 5% have been employed for filamentous fungi (5, 12). We used the 10% threshold because it provided greater agreement with the visual readings than 5% (a 2 to 3% increase in overall agreement for each agent). The levels of agreement obtained between visual and spectrophotometric readings were comparable to those of previous reports (5, 12).
An important finding was that the azole agents at concentrations equal to or greater than MIC-0 do not completely suppress the metabolic activity of Zygomycetes during the first 12 h after inoculation, as manifested by the trend of relative OD450s for the 6-, 8-, and 12-h time points (Table 1 and Fig. 3). The gradual decrease of relative OD450 values observed with inhibitory drug concentrations over these time points was due in part to the increase of the OD450 values of the control wells and to a decrease in the absolute OD450 values of the corresponding wells. In contrast, amphotericin B at inhibitory concentrations almost completely suppressed metabolic activity at the 6- or 8-h time points for all isolates. This differential effect on the early metabolic activity of Zygomycetes was probably the result of different mechanisms of action of the azoles and polyenes against these molds (9) and necessitated the use of cutoff levels for the azole agents different than those for amphotericin B in the early XTT assays (Table 2).
Using appropriate cutoff levels, the MIC-0s determined at the 6-h time point for amphotericin B and posaconazole were in very good agreement with the 24-h spectrophotometric (and visual, for amphotericin B) readings. MICs for all agents obtained at the 8- or 12-h time point were also in very good agreement with the 24-h readings (Table 4). This agreement was usually greater with the spectrophotometric readings than the visual ones, especially for voriconazole. This discrepancy was not surprising, as voriconazole demonstrated lower levels of overall agreement between spectrophotometric and visual readings at 24 h than posaconazole and amphotericin B, in both our study (Table 4) and previous reports (5).
A notable finding in this study was the lower level of agreement between the visual and spectrophotometric readings and between the early-XTT and 24-h MICs for Cunninghamella spp. than for other species. While it is difficult to draw conclusions due to the small number (3) of Cunninghamella isolates, it should be noted that the major "contributor" to disagreement, for most comparisons, was the C. echinulata isolate. When the two C.
bertholletiae isolates were evaluated separately, their levels of agreement between visual and spectrophotometric readings at 24 h were 100% for MIC-0 of posaconazole and voriconazole and 50% for MIC-2 of posaconazole. In addition, the levels of agreement between the 6-h XTT and 24-h spectrophotometric readings were 100% for MIC-0 of amphotericin B and posaconazole, 50% for MIC-2 of posaconazole, and 83% for MIC-0 of voriconazole.
In conclusion, medically important Zygomycetes demonstrate significant metabolic activity with the XTT assay before visual or spectrophotometric detection of fungal growth. This early metabolic activity can be inhibited by antifungal agents, although complete inhibition with the azoles is delayed in comparison to amphotericin B. Using appropriate cutoff levels for each agent and time point, the XTT assay performed at 6, 8, or 12 h after inoculation can determine MICs that are in very good agreement with those obtained with visual or spectrophotometric readings at 24 h. These data now establish a foundation for conducting a multicenter trial for validation of the determination of MICs for Zygomycetes by the early XTT assay. The use of the XTT method for rapid susceptibility testing should also be investigated for other medically important filamentous fungi.
ACKNOWLEDGMENTS
This work was supported in part by the Intramural Research Program of the National Institutes of Health and the National Cancer Institute.
REFERENCES
Ahmed, A. O., W. W. van de Sande, W. van Vianen, A. van Belkum, A. H. Fahal, H. A. Verbrugh, and I. A. Bakker-Woudenberg. 2004. In vitro susceptibilities of Madurella mycetomatis to itraconazole and amphotericin B assessed by a modified NCCLS method and a viability-based 2,3-bis (2methoxy-4-nitro-5-sulfophenyl)-5-[(phenylamino)carbonyl]-2H-tetrazolium hydroxide (XTT) assay. Antimicrob. Agents Chemother. 48:2742-2746.
Balajee, S. A., A. Imhof, J. L. Gribskov, and K. A. Marr. 2005. Determination of antifungal drug susceptibilities of Aspergillus species by a fluorescence-based microplate assay. J. Antimicrob. Chemother. 55:102-105.
Balajee, S. A., and K. A. Marr. 2002. Conidial viability assay for rapid susceptibility testing of Aspergillus species. J. Clin. Microbiol. 40:2741-2745.
Dannaoui, E., J. F. Meis, D. Loebenberg, and P. E. Verweij. 2003. Activity of posaconazole in treatment of experimental disseminated zygomycosis. Antimicrob. Agents Chemother. 47:3647-3650.
Dannaoui, E., J. Meletiadis, J. W. Mouton, J. F. Meis, and P. E. Verweij. 2003. In vitro susceptibilities of zygomycetes to conventional and new antifungals. J. Antimicrob. Chemother. 51:45-52.
Diekema, D. J., S. A. Messer, R. J. Hollis, R. N. Jones, and M. A. Pfaller. 2003. Activities of caspofungin, itraconazole, posaconazole, ravuconazole, voriconazole, and amphotericin B against 448 recent clinical isolates of filamentous fungi. J. Clin. Microbiol. 41:3623-3626.
Espinel-Ingroff, A., M. Bartlett, R. Bowden, N. X. Chin, C. Cooper, Jr., A. Fothergill, M. R. McGinnis, P. Menezes, S. A. Messer, P. W. Nelson, F. C. Odds, L. Pasarell, J. Peter, M. A. Pfaller, J. H. Rex, M. G. Rinaldi, G. S. Shankland, T. J. Walsh, and I. Weitzman. 1997. Multicenter evaluation of proposed standardized procedure for antifungal susceptibility testing of filamentous fungi. J. Clin. Microbiol. 35:139-143.
Gil-Lamaignere, C., R. Hess, S. Salvenmoser, K. Heyn, R. Kappe, and F. M. Muller. 2005. Effect of media composition and in vitro activity of posaconazole, caspofungin and voriconazole against zygomycetes. J. Antimicrob. Chemother. 55:1016-1019.
Groll, A. H., J. C. Gea-Banacloche, A. Glasmacher, G. Just-Nuebling, G. Maschmeyer, and T. J. Walsh. 2003. Clinical pharmacology of antifungal compounds. Infect. Dis. Clin. N. Am. 17:159-191, ix.
Kontoyiannis, D. P., M. S. Lionakis, R. E. Lewis, G. Chamilos, M. Healy, C. Perego, A. Safdar, H. Kantarjian, R. Champlin, T. J. Walsh, and I. I. Raad. 2005. Zygomycosis in a tertiary-care cancer center in the era of Aspergillus-active antifungal therapy: a case-control observational study of 27 recent cases. J. Infect. Dis. 191:1350-1360.
Meletiadis, J., J. W. Mouton, J. F. Meis, B. A. Bouman, J. P. Donnelly, and P. E. Verweij. 2001. Colorimetric assay for antifungal susceptibility testing of Aspergillus species. J. Clin. Microbiol. 39:3402-3408.
Meletiadis, J., J. W. Mouton, J. F. Meis, B. A. Bouman, P. J. Donnelly, and P. E. Verweij. 2001. Comparison of spectrophotometric and visual readings of NCCLS method and evaluation of a colorimetric method based on reduction of a soluble tetrazolium salt, 2,3-bis {2-methoxy-4-nitro-5-[(sulfenylamino) carbonyl]-2H-tetrazolium-hydroxide}, for antifungal susceptibility testing of Aspergillus species. J. Clin. Microbiol. 39:4256-4263.
Meletiadis, J., D. T. te Dorsthorst, and P. E. Verweij. 2003. Use of turbidimetric growth curves for early determination of antifungal drug resistance of filamentous fungi. J. Clin. Microbiol. 41:4718-4725.
Merz, W. G., D. Fay, B. Thumar, and D. Dixon. 1984. Susceptibility testing of filamentous fungi to amphotericin B by a rapid radiometric method. J. Clin. Microbiol. 19:54-56.
NCCLS. 2002. Reference method for broth dilution antifungal susceptibility testing of filamentous fungi; approved standard. NCCLS document M38-A. NCCLS, Wayne, Pa.
Pagano, L., M. Offidani, L. Fianchi, A. Nosari, A. Candoni, M. Piccardi, L. Corvatta, D. D'Antonio, C. Girmenia, P. Martino, and A. Del Favero. 2004. Mucormycosis in hematologic patients. Haematologica 89:207-214.
Pfaller, M. A., and M. G. Rinaldi. 1993. Antifungal susceptibility testing. Current state of technology, limitations, and standardization. Infect. Dis. Clin. N. Am. 7:435-444.
Roden, M. M., T. E. Zaoutis, W. L. Buchanan, T. A. Knudsen, T. A. Sarkisova, R. L. Schaufele, M. Sein, T. Sein, C. C. Chiou, J. H. Chu, D. P. Kontoyiannis, and T. J. Walsh. 2005. Epidemiology and outcome of zygomycosis: a review of 929 reported cases. Clin. Infect. Dis. 41:634-653.
Spellberg, B., J. Edwards, Jr., and A. Ibrahim. 2005. Novel perspectives on mucormycosis: pathophysiology, presentation, and management. Clin. Microbiol. Rev. 18:556-569.
Sun, Q. N., A. W. Fothergill, D. I. McCarthy, M. G. Rinaldi, and J. R. Graybill. 2002. In vitro activities of posaconazole, itraconazole, voriconazole, amphotericin B, and fluconazole against 37 clinical isolates of zygomycetes. Antimicrob Agents Chemother. 46:1581-1582.
Sun, Q. N., L. K. Najvar, R. Bocanegra, D. Loebenberg, and J. R. Graybill. 2002. In vivo activity of posaconazole against Mucor spp. in an immunosuppressed-mouse model. Antimicrob. Agents Chemother. 46:2310-2312.
Tobon, A. M., M. Arango, D. Fernandez, and A. Restrepo. 2003. Mucormycosis (zygomycosis) in a heart-kidney transplant recipient: recovery after posaconazole therapy. Clin. Infect. Dis. 36:1488-1491.
Immunocompromised Host Section, Pediatric Oncology Branch, National Cancer Institute, Bethesda, Maryland
1 Third Department of Pediatrics, Aristotle University, Hippokration Hospital, Thessaloniki, Greece(Charalampos Antachopoulos)
The XTT colorimetric assay quantifies fungal growth by measuring fungal metabolism and has been used successfully for susceptibility testing of Aspergillus species after 24 and 48 h of incubation. In the present study using 14 clinical isolates of Zygomycetes (Rhizopus oryzae [5 isolates], Cunninghamella spp. [3 isolates], Mucor spp. [3 isolates], and Absidia corymbifera [3 isolates]), significant metabolic activity was demonstrated before visual or spectrophotometric detection of fungal growth by performing the XTT assay as early as 6 h after inoculation. Testing of susceptibility to amphotericin B, posaconazole, and voriconazole was subsequently performed using the XTT method (100 μg/ml XTT, 25 μM menadione) at 6, 8, or 12 h after inoculation and the CLSI (formerly NCCLS) M38-A method with visual and spectrophotometric MIC determinations at 24 h after inoculation. Concentration-effect curves obtained with the use of the Emax model (a sigmoid curve with variable slope) were comparable between the early XTT and spectrophotometric readings at 24 h. Complete inhibition of early metabolic activity with the azoles was delayed in comparison to that with amphotericin B. Using appropriate cutoff levels, agreement was demonstrated between the early XTT and 24-h spectrophotometric or visual readings. In particular, for MIC-0 (the lowest drug concentration showing absence of visual growth) of amphotericin B, overall agreement levels were 90 to 93% for the 6-h XTT assay and 100% for the 8- and 12-h time points. For MIC-0 of posaconazole, agreement levels were 86% for the 6-h XTT and 93 to 100% for the 8- and 12-h time points. The overall agreement levels for MIC-0 and MIC-2 (the lowest drug concentration showing prominent reduction of growth compared with the control well) of voriconazole (compared with 24-h spectrophotometric readings) were 93 to 98% for the 8- and 12-h XTT assays. These results support the use of the XTT method for rapid MIC determination for Zygomycetes.
INTRODUCTION
Infections caused by Zygomycetes have emerged as an increasingly important cause of morbidity and mortality among severely immunocompromised patients, including those with hematological malignancies and hematopoietic stem cell transplant recipients. The prognosis of zygomycosis in these patients remains extremely poor, despite antifungal treatment (10, 16, 18, 19). For years, amphotericin B has been the drug of choice for these devastating infections. Of the recently introduced second-generation triazoles, voriconazole is not active against the Zygomycetes, while posaconazole has demonstrated activity in vitro, in animal models, and in case reports (4-6, 8, 20-22). As zygomycosis may be rapidly fatal, timely determination of the susceptibilities of clinical isolates to these antifungal agents is important.
A colorimetric method, based on reduction of the tetrazolium salt 2,3-bis{2-methoxy-4-nitro-5-[(sulfenylamino) carbonyl]-2H-tetrazolium-hydroxide} (XTT) by mitochondrial dehydrogenases, has been recently introduced for in vitro susceptibility testing of filamentous fungi. This method quantifies fungal growth by measuring fungal metabolism, using appropriate concentrations of XTT and an electron transfer agent, such as menadione (1, 11, 12). For Aspergillus species, MICs determined by the XTT method at 24 and 48 h were comparable with those obtained by the National Committee for Clinical Laboratory Standards (now known as the Clinical and Laboratory Standards Institute [CLSI]) method (12). The sensitivity of the XTT assay may be increased depending on the menadione concentration (11). High menadione concentrations could increase the rate of XTT conversion and theoretically allow the detection of metabolic activity associated with small, early fungal growth that is not yet detected visually or spectrophotometrically. In this case, it is likely that the XTT method might even be useful for early determination of MICs of antifungal agents. This hypothesis has been previously discussed (12) but has not been investigated.
In preliminary studies using the XTT assay for zygomycete isolates, high metabolic activity was observed even in the presence of small fungal growth (assessed spectrophotometrically) at 12 or 24 h postinoculation and preceded increases in biomass (C. Antachopoulos, J. Meletiadis, E. Roilides, T. Sein, and T. J. Walsh, Abstr. 45th Intersci. Conf. Antimicrob. Agents Chemother., abstr. M-1003, 2005). The present study investigated whether significant metabolic activity with the XTT assay could be demonstrated for the Zygomycetes before hyphal growth is macroscopically or spectrophotometrically detected (at 6 or 8 h postinoculation) and whether changes in this early metabolic activity in the presence of antifungal agents could be used for rapid susceptibility testing of these filamentous fungi.
MATERIALS AND METHODS
Isolates. A collection of 14 clinical zygomycete isolates was used, including the following species: Rhizopus oryzae (5 isolates), Cunninghamella bertholletiae (2 isolates), Cunninghamella echinulata (1 isolate), Mucor circinelloides (2 isolates), Mucor indicus (1 isolate), and Absidia corymbifera (3 isolates). The reference strains Candida krusei 6258 and Candida parapsilosis 22019 were used as quality controls.
Medium. RPMI 1640 medium with L-glutamine but without bicarbonate (Cambrex Bio Science, Inc., Walkersville, MD) buffered to pH 7.0 with 0.165 M 3-N-morpholinopropanesulfonic acid (MOPS) (Sigma-Aldrich, St. Louis, MO) was used as the assay medium.
Inoculum. Sporangiospores were harvested after isolates were subcultured in potato dextrose agar at 35°C for 3 days and were suspended in normal saline containing 0.025% Tween 20. The sporangiospore suspensions were counted with a hemacytometer and then diluted in RPMI in order to achieve the desired concentrations. Inoculum sizes were verified by quantitative colony counts on Sabouraud dextrose agar plates.
XTT and menadione. XTT (Sigma-Aldrich, St. Louis, MO) was dissolved in normal saline at concentrations of 1, 0.5, and 0.25 mg/ml. Menadione (Sigma-Aldrich, St. Louis, MO) was initially dissolved in absolute ethanol at a concentration of 10 mg/ml and subsequently added to the above-mentioned XTT solutions at concentrations of 125, 31.25, and 7.81 μM for each solution. In this way, a total of nine solutions with different XTT-menadione concentrations were prepared.
XTT assay for measurement of early metabolic activity. In the XTT assay for measurement of early metabolic activity, the abilities of the above-mentioned XTT-menadione combinations to detect early metabolic activity of Zygomycetes were investigated. Briefly, 104 sporangiospores/ml of four clinical zygomycete isolates (one each of R. oryzae, C. bertholletiae, M.
circinelloides, and A. corymbifera) were incubated in RPMI 1640 at 37°C in 96-well microtitration plates at a volume of 200 μl/well. After 6 or 8 h of incubation, 50 μl of one of the above-mentioned XTT-menadione solutions was added to each well, as previously described (11, 12). All nine solutions were used, resulting in final concentrations of 200, 100, and 50 μg/ml XTT, each combined with 25, 6.25, or 1.56 μM menadione, respectively. The microtitration plates were further incubated at 37°C for 2 h in order to allow conversion of XTT to its formazan derivative. These plates were then shaken for 1 to 2 min (Wallac Plate Shake 1296-004; Wallac OY, Turku, Finland) until complete dissolution of formazan derivatives was achieved. XTT conversion was measured as optical density (OD) with a microtitration plate spectrophotometric reader (Elx808; Bio-Tek Instruments, Winooski, VT) at 450 nm. For each well, XTT conversion was calculated after subtraction of the background OD, which was the OD of a simultaneously incubated well with 200 μl of RPMI and 50 μl of XTT-menadione solution but no inoculum. Four replicates were used for each species, time point, and XTT-menadione concentration.
Relationship between early XTT conversion and fungal biomass. An assay investigated whether the rate of early XTT conversion, obtained with the different XTT-menadione combinations, in fact correlated with the magnitude of fungal biomass, using a previously described approach (11). Briefly, the XTT assay was performed at 6 or 8 h of incubation, as described above, using 10-fold serially diluted inocula (102 to 106 sporangiospores/ml) for each of the above-mentioned four clinical zygomycete isolates. The relationship between XTT conversion (OD at 450 nm [OD450]) and fungal inoculum (log10) was assessed by linear regression analysis, and the slopes and goodness of fit (r2) were reported for each XTT-menadione combination.
Antifungal-susceptibility testing. MICs were determined for all 14 isolates of Zygomycetes. The broth microdilution method of the CLSI (M38-A) (7, 15) was followed using 96-well microtitration plates. Posaconazole (Schering-Plough Research Institute, Kenilworth, NJ) was dissolved in dimethyl sulfoxide. Amphotericin B (Bristol-Myers Squibb, Princeton, NJ) and voriconazole (Pfizer Inc., New York, NY) were dissolved in sterile distilled water. Serial dilutions of the antifungal agents were prepared to yield concentrations of 0.03 to 32 μg/ml for posaconazole, 0.015 to 16 μg/ml for amphotericin B, and 1 to 1,024 μg/ml for voriconazole at a final volume of 200 μl after inoculation. The trays were maintained at –70°C until the day of testing.
Incubation and MIC determination. After the microtitration trays were defrosted, they were inoculated with a final concentration of 2.5 x 104 sporangiospores/ml from each of the 14 isolates and incubated at 37°C. For each isolate, four different 96-well trays were used, corresponding to four different incubation periods: 6, 8, 12, and 24 h. For each time point, an additional tray was prepared, whose wells contained the above-mentioned drug concentrations at a final volume of 200 μl of RPMI, but no fungal inoculum. These trays served to provide background OD measurements for the XTT and spectrophotometric readings.
(i) XTT assay for early MIC determination. At 6, 8, and 12 h after inoculation, the corresponding trays were removed from the incubator, and 50 μl of saline containing 0.5 mg/ml XTT with 125 μM menadione was added to each well to obtain final concentrations of 100 μg/ml XTT and 25 μM menadione. The trays were further incubated for 2 h at 37°C. After shaking of the trays, XTT conversion was measured at 450 nm. After subtraction of the corresponding background OD, the relative OD450 for each well and drug concentration, in relation to the control well, was calculated as the following fraction: OD450 of the drug-containing well/OD450 of the control well. Due to the lack of previous data regarding the effects of different antifungal agents on the early metabolic activity of Zygomycetes, appropriate cutoff levels for determination of MIC-0 and MIC-2 (see below) and comparison with the CLSI method were employed after the results were obtained and the concentration-effect curves of metabolic activity were analyzed.
(ii) CLSI method. After 24 h of incubation, the corresponding trays were removed from the incubator and fungal growth was assessed visually and spectrophotometrically. For visual assessment, a concave mirror was used. The MIC-0 was determined as the lowest drug concentration showing absence of visual growth and was recorded for all three agents. The MIC-2 was determined as the lowest drug concentration showing prominent reduction of growth compared with the control well and was recorded only for posaconazole and voriconazole. For spectrophotometric assessment, hyphal growth was measured at 405 nm, as previously described (5, 12). After subtraction of the corresponding background OD, the relative OD405 for each well and drug concentration, in relation to the control well, was calculated as follows: OD405 of the drug-containing well/OD405 of the control well. The lowest drug concentrations showing relative ODs of 10% and 50% were determined to be the MIC-0 and MIC-2, respectively. Again, only MIC-0 was recorded for amphotericin B.
Experiments of antifungal susceptibility testing using the early XTT and CLSI methods were repeated in triplicate.
Modeling and data analysis. For four of the zygomycete isolates (one each of R. oryzae, C. bertholletiae, M. circinelloides, and A. corymbifera), for the purpose of modeling, the relative OD values obtained with the XTT assay (6, 8, and 12 h) or spectrophotometric reading at 405 nm (24 h) were analyzed by nonlinear regression analysis, using a four-parameter logistic model (a sigmoid curve with variable slope) known as the Emax model. Deviation from the model was tested by the runs test, and goodness of fit was checked by the r2 value. The best-fit values of the EC50, which is the drug concentration producing 50% of the Emax (maximum relative OD); of the slope m, which describes the steepness of the curve; and of the bottom of the curve were recorded and compared for the early (6-, 8-, and 12-h) and 24-h time points for each isolate. Comparison of best-fit values for these parameters was performed with one-way analysis of variance, followed by Dunn's test for multiple comparisons. The results of comparisons of the EC50 and bottom-of-the-curve values were used to employ appropriate cutoff levels for determination of MIC-2 and MIC-0, respectively, with the early XTT method (6, 8, and 12 h). Statistical analysis was performed using the GraphPad (San Diego, Calif.) Prism software.
For purposes of comparison with CLSI methods, MICs for all 14 isolates obtained by using the above cutoff levels for the early XTT assays were compared with corresponding values from visual and spectrophotometric (405-nm) readings at 24 h. The percentage of relative (within one dilution) agreement between the MICs obtained with different methods and time points for each species and for all isolates was calculated and reported.
Reproducibility. In order to determine the reproducibility of the XTT, visual, and spectrophotometric methods for each time point and antifungal agent, the median MIC of those obtained with the three experiments was calculated for each isolate. The relative interexperimental agreement was subsequently defined as the percentage of MICs (obtained for each method, time point, and antifungal agent) that were within one twofold dilution from the median MIC (12).
RESULTS
Early metabolic activity. Significant metabolic activity was demonstrated with the XTT assay at the 6- and 8-h time points for all four isolates tested, with OD450s ranging from 0.17 to 1.06 for those wells treated with 200, 100, or 50 μg/ml XTT and 25 μM menadione at 6 h (Fig. 1). At this early time point, the use of different XTT concentrations resulted in comparable conversion rates. In contrast, the use of lower menadione concentrations (6.25 and 1.56 μM) resulted in markedly reduced XTT conversion, independently of the concentration of XTT (Fig. 1). Consequently, XTT-menadione combinations containing 25 μM of menadione were used exclusively in further assays of early metabolic activity.
Relationship between early XTT conversion and fungal biomass. A linear relationship was demonstrated between early XTT conversion and fungal inoculum (P < 0.001). Of the three combinations of 200, 100, and 50 μg/ml XTT with 25 μM menadione, a higher median (range) r2 was obtained with the concentrations of 100 μg/ml XTT and 25 μM menadione [0.91 (0.86 to 0.98)]. For the combinations of 200 or 50 μg/ml XTT with 25 μM menadione, the median (range) r2 values were 0.85 (0.76 to 0.95) and 0.88 (0.86 to 0.92), respectively. The slope [median (range)] of the regression line obtained with 100 μg/ml XTT was 0.42 (0.41 to 0.46). Use of 200 or 50 μg/ml XTT resulted in very high (median, 0.81; range, 0.66 to 0.84) or low (median, 0.22; range, 0.20 to 0.23) slopes, respectively. Since we aimed for a linear relationship with intermediate slope (high or low slopes would result in fast or slow XTT conversion, respectively), we employed the combination of 100 μg/ml XTT and 25 μM menadione for early antifungal susceptibility testing of zygomycete isolates.
Antifungal susceptibility testing. In agreement with the previous findings, sufficiently high XTT conversion rates (mean OD450, 0.69; range, 0.19 to 1.67) were obtained when the combination of 100 μg/ml XTT and 25 μM menadione was used for MIC determination at the 6-h time point for all 14 isolates. Examination of microtitration plates under the inverted microscope at the end of the 2-h incubation period following addition of the XTT-menadione solution (i.e., 8 h, in total, after inoculation) demonstrated that XTT conversion increased with progression from germination to the formation of hyphae (Fig. 2).
Modeling of XTT and spectrophotometric measurements. For the four isolates of Zygomycetes (one each of R. oryzae, C. bertholletiae, M. circinelloides, and A. corymbifera) for which data were analyzed by nonlinear regression analysis, the Emax model fit the spectrophotometric (24-h) and XTT (6-, 8-, and 12-h) readings very well; r2 ranged between 0.97 and 0.99, and no statistical deviations from the model were found (P = 0.17 to 0.99; runs test). No statistical differences were found between the best-fit values of EC50 obtained with the spectrophotometric and colorimetric methods for posaconazole and voriconazole (P > 0.05 by one-way analysis of variance) (Table 1). For amphotericin B, P was <0.05 for comparison of the 24-h spectrophotometric and 6- or 8-h XTT readings; however, the differences in EC50 values did not exceed 1 dilution (Table 1). The slopes of the concentration-effect curves were also similar between the different methods (P > 0.05) but varied according to the antifungal agent used: higher absolute m values (steeper curves) were obtained with amphotericin B and lower values were obtained with voriconazole (data not shown). A significant difference, however, between the spectrophotometric (405-nm) and early XTT methods was found in the best-fit values of the bottom of the curve for posaconazole and voriconazole (P < 0.05) (Table 1). For these agents, the bottom values of the curve obtained with the early XTT assays were significantly higher than those obtained with the 24-h spectrophotometric readings. Bottom values were highest at the 6-h time point and decreased gradually at 8 and 12 h (Table 1). This difference in the bottom values of the sigmoid curve between the early XTT and spectrophotometric methods for the azoles, in comparison to amphotericin B, is also illustrated in Fig. 3 and suggests that for the azole agents, significant XTT conversion could occur at the early time points (up to almost 40% of control OD450 for the 6-h time point), even in the presence of drug concentrations equal to or greater than the MIC-0. Thus, suppression of the metabolic activity of Zygomycetes by inhibitory concentrations of the azoles was gradually achieved and was not complete at the early time points.
The comparability of EC50 values obtained with the Emax model suggested that for determination of MIC-2 by the early XTT assays, the cutoff level of 50% (also employed for the spectrophotometric readings at 24 h) would probably provide good agreement between the different methods. However, the differences in the bottom values of the curve suggested that in order to determine MIC-0 for the azoles by the early XTT method, cutoff levels greater than 10% (which was used in the spectrophotometric method) should be employed. Based on the Emax model data (Table 1), these cutoffs could be on the order of 20 to 40% for the 6-h time point and 10 to 20% for the 8- and 12-h time points. Indeed, the MIC-0 values obtained with the above range of cutoff levels were compared with those obtained with spectrophotometric (405-nm) or visual readings, and the cutoffs giving the best agreement are presented in Table 2. For amphotericin B, MIC-0 values obtained using the 10% cutoff for the 6- and 8-h XTT method were in very good agreement with the spectrophotometric method (90 and 95%, respectively). However, the threshold of 5% increased the agreement for the 8-h time point to 100%.
Comparison of MICs. Comparable MICs of zygomycete isolates were obtained with the visual (CLSI), spectrophotometric (405-nm), and early XTT methods using the cutoffs shown in Table 2 for different time points. Differences in median MICs obtained with different methods for zygomycete species did not exceed 1 dilution, with the exception of median MICs of voriconazole for Mucor spp. and A. corymbifera, which were within 2 dilutions (Table 3).
The relative levels of agreement between the different methods are presented in Table 4. The overall agreement levels between the spectrophotometric (405-nm) and visual (CLSI) methods of MIC determination at 24 h ranged between 83 and 100% for different agents, with 83% being the agreement for MIC-0 of voriconazole and 100% being the agreement for MIC-0 of amphotericin B. Lower agreement levels were observed between the two methods in the determination of MICs of posaconazole and voriconazole against Cunninghamella spp. (Table 4). Whenever differences were observed between these two methods, visual MICs tended to be higher than spectrophotometric ones (Table 3).
For amphotericin B, the overall agreement levels of MICs obtained with the early XTT assays compared to the spectrophotometric and visual readings ranged from 90 to 93% for the 6-h time point to 100% for those of 8 and 12 h (Table 4). Lower agreement levels were observed for Cunninghamella spp. with the 6-h XTT assay.
For posaconazole, the overall agreement was 86% for MIC-0 obtained with the 6-h XTT versus spectrophotometric readings and even higher for the 8- and 12-h time points (93 to 100% for MIC-0) (Table 4). Lower agreement levels, however, were observed between the 6-h XTT and visual MIC-0 (76%) and between the 6-h XTT and spectrophotometric MIC-2 (81%). Again, among different species, lower agreement levels were observed in the determination of MICs for Cunninghamella spp., in particular, MIC-2.
For voriconazole, the MIC-0 and MIC-2 obtained by the 8- and 12-h XTT methods showed an overall agreement of 93 to 98% with the spectrophotometric method but showed less with the visual readings (71 to 86%) (Table 4). The overall agreement levels of MIC-2 determined by the 6-h XTT method for voriconazole were 88% compared with the spectrophotometric method and slightly lower (83%) with the visual method. Lower agreement levels, however (67 to 74%), were observed for the MIC-0 values obtained with the 6-h XTT compared to visual or spectrophotometric readings.
Reproducibility. The relative interexperimental agreement levels for MIC-0 and MIC-2 ranged between 98 and 100% for all methods, time points, and antifungal agents, with the exception of MIC-0 determined by the 6-h XTT method for voriconazole and MIC-2 determined by the 24-h spectrophotometric method for posaconazole, for which the agreement was 95% (standard error, 3%).
DISCUSSION
In the present study, we demonstrated that MICs of Zygomycetes can be obtained in a reproducible way with the XTT assay up to 16 h earlier than with current standard methods. This early determination of MICs was based on the demonstration of significant metabolic activity of these fungi before detection of significant growth macroscopically or spectrophotometrically (mean OD405 for zygomycete isolates at 8 h postinoculation, 0.008; range, –0.004 to 0.03).
Several methods for rapid susceptibility testing of filamentous fungi have been developed in the past, including a radiometric assay measuring the inhibition of 14CO2 production (14), a flow cytometry assay of conidial viability (3), a fluorescence-based microplate assay (2), and a turbidometric method assessing the lag phases of fungal growth curves in the presence of antifungal drugs (13). Some of these methods may use expensive equipment or potentially hazardous substances, and they have been evaluated only for amphotericin B. The fluorescence-based microplate assay generated MICs in 16 h for Aspergillus spp., but it remains to be evaluated for other filamentous fungi, including Zygomycetes (2). The use of turbidometric growth curves required up to 12 h of incubation for determination of the amphotericin B resistance of R. oryzae isolates (13). The XTT method presented here is a simple way of measuring fungal metabolism that can be performed in most laboratories.
When menadione is used as an electron transfer agent, the rate of conversion of XTT to its formazan derivative could be affected by the concentrations of both the XTT tetrazolium salt and menadione (11). As shown in Fig. 1, at the early time point of 6 h, the effect of menadione on XTT conversion seemed to be more significant than that of the XTT concentration. A high (25 μM) menadione concentration was required to obtain significant conversion rates. It should be emphasized, however, that while the use of the XTT assay for rapid determination of MICs could probably be evaluated for other fungi as well, the concentrations of XTT and menadione that should be used may need to be modified, due to potential differences in growth properties and metabolic activities among different classes or genera.
Another factor affecting XTT conversion at these early time points is the size of the inoculum, which correlated linearly with the OD450 values obtained. The results from our studies of early metabolic activity (Fig. 1), as well as from the subsequent antifungal susceptibility studies (where the range of viable inoculum, based on quantitative colony counts, was 0.7 x 104 to 3.9 x 104 sporangiospores/ml), suggested that, within the range of inocula suggested by the CLSI guidelines (0.4 x 104 to 5 x 104 sporangiospores/ml) (15), sufficient XTT conversion occurred at the early time points of 6 and 8 h to allow MIC determination. However, deviations beyond the lower limit of the suggested range should probably be avoided, since they may result in significantly reduced XTT conversion (with 103 sporangiospores/ml, an OD450 as low as 0.04 was recorded for some isolates [data not shown]).
Although CLSI recommendations for filamentous fungi suggest reporting only MIC-0 for the azoles (15), others have also used MIC-2 (5, 8). Hence, we determined both MIC-0 and MIC-2 for these agents. For determination of MICs at 24 h, we used both visual and spectrophotometric (405-nm) readings of the trays. Spectrophotometric readings may be more precise and reproducible than visual assessment of fungal growth, which is often subjective (17). For spectrophotometric determination of MIC-0, thresholds of both 10% and 5% have been employed for filamentous fungi (5, 12). We used the 10% threshold because it provided greater agreement with the visual readings than 5% (a 2 to 3% increase in overall agreement for each agent). The levels of agreement obtained between visual and spectrophotometric readings were comparable to those of previous reports (5, 12).
An important finding was that the azole agents at concentrations equal to or greater than MIC-0 do not completely suppress the metabolic activity of Zygomycetes during the first 12 h after inoculation, as manifested by the trend of relative OD450s for the 6-, 8-, and 12-h time points (Table 1 and Fig. 3). The gradual decrease of relative OD450 values observed with inhibitory drug concentrations over these time points was due in part to the increase of the OD450 values of the control wells and to a decrease in the absolute OD450 values of the corresponding wells. In contrast, amphotericin B at inhibitory concentrations almost completely suppressed metabolic activity at the 6- or 8-h time points for all isolates. This differential effect on the early metabolic activity of Zygomycetes was probably the result of different mechanisms of action of the azoles and polyenes against these molds (9) and necessitated the use of cutoff levels for the azole agents different than those for amphotericin B in the early XTT assays (Table 2).
Using appropriate cutoff levels, the MIC-0s determined at the 6-h time point for amphotericin B and posaconazole were in very good agreement with the 24-h spectrophotometric (and visual, for amphotericin B) readings. MICs for all agents obtained at the 8- or 12-h time point were also in very good agreement with the 24-h readings (Table 4). This agreement was usually greater with the spectrophotometric readings than the visual ones, especially for voriconazole. This discrepancy was not surprising, as voriconazole demonstrated lower levels of overall agreement between spectrophotometric and visual readings at 24 h than posaconazole and amphotericin B, in both our study (Table 4) and previous reports (5).
A notable finding in this study was the lower level of agreement between the visual and spectrophotometric readings and between the early-XTT and 24-h MICs for Cunninghamella spp. than for other species. While it is difficult to draw conclusions due to the small number (3) of Cunninghamella isolates, it should be noted that the major "contributor" to disagreement, for most comparisons, was the C. echinulata isolate. When the two C.
bertholletiae isolates were evaluated separately, their levels of agreement between visual and spectrophotometric readings at 24 h were 100% for MIC-0 of posaconazole and voriconazole and 50% for MIC-2 of posaconazole. In addition, the levels of agreement between the 6-h XTT and 24-h spectrophotometric readings were 100% for MIC-0 of amphotericin B and posaconazole, 50% for MIC-2 of posaconazole, and 83% for MIC-0 of voriconazole.
In conclusion, medically important Zygomycetes demonstrate significant metabolic activity with the XTT assay before visual or spectrophotometric detection of fungal growth. This early metabolic activity can be inhibited by antifungal agents, although complete inhibition with the azoles is delayed in comparison to amphotericin B. Using appropriate cutoff levels for each agent and time point, the XTT assay performed at 6, 8, or 12 h after inoculation can determine MICs that are in very good agreement with those obtained with visual or spectrophotometric readings at 24 h. These data now establish a foundation for conducting a multicenter trial for validation of the determination of MICs for Zygomycetes by the early XTT assay. The use of the XTT method for rapid susceptibility testing should also be investigated for other medically important filamentous fungi.
ACKNOWLEDGMENTS
This work was supported in part by the Intramural Research Program of the National Institutes of Health and the National Cancer Institute.
REFERENCES
Ahmed, A. O., W. W. van de Sande, W. van Vianen, A. van Belkum, A. H. Fahal, H. A. Verbrugh, and I. A. Bakker-Woudenberg. 2004. In vitro susceptibilities of Madurella mycetomatis to itraconazole and amphotericin B assessed by a modified NCCLS method and a viability-based 2,3-bis (2methoxy-4-nitro-5-sulfophenyl)-5-[(phenylamino)carbonyl]-2H-tetrazolium hydroxide (XTT) assay. Antimicrob. Agents Chemother. 48:2742-2746.
Balajee, S. A., A. Imhof, J. L. Gribskov, and K. A. Marr. 2005. Determination of antifungal drug susceptibilities of Aspergillus species by a fluorescence-based microplate assay. J. Antimicrob. Chemother. 55:102-105.
Balajee, S. A., and K. A. Marr. 2002. Conidial viability assay for rapid susceptibility testing of Aspergillus species. J. Clin. Microbiol. 40:2741-2745.
Dannaoui, E., J. F. Meis, D. Loebenberg, and P. E. Verweij. 2003. Activity of posaconazole in treatment of experimental disseminated zygomycosis. Antimicrob. Agents Chemother. 47:3647-3650.
Dannaoui, E., J. Meletiadis, J. W. Mouton, J. F. Meis, and P. E. Verweij. 2003. In vitro susceptibilities of zygomycetes to conventional and new antifungals. J. Antimicrob. Chemother. 51:45-52.
Diekema, D. J., S. A. Messer, R. J. Hollis, R. N. Jones, and M. A. Pfaller. 2003. Activities of caspofungin, itraconazole, posaconazole, ravuconazole, voriconazole, and amphotericin B against 448 recent clinical isolates of filamentous fungi. J. Clin. Microbiol. 41:3623-3626.
Espinel-Ingroff, A., M. Bartlett, R. Bowden, N. X. Chin, C. Cooper, Jr., A. Fothergill, M. R. McGinnis, P. Menezes, S. A. Messer, P. W. Nelson, F. C. Odds, L. Pasarell, J. Peter, M. A. Pfaller, J. H. Rex, M. G. Rinaldi, G. S. Shankland, T. J. Walsh, and I. Weitzman. 1997. Multicenter evaluation of proposed standardized procedure for antifungal susceptibility testing of filamentous fungi. J. Clin. Microbiol. 35:139-143.
Gil-Lamaignere, C., R. Hess, S. Salvenmoser, K. Heyn, R. Kappe, and F. M. Muller. 2005. Effect of media composition and in vitro activity of posaconazole, caspofungin and voriconazole against zygomycetes. J. Antimicrob. Chemother. 55:1016-1019.
Groll, A. H., J. C. Gea-Banacloche, A. Glasmacher, G. Just-Nuebling, G. Maschmeyer, and T. J. Walsh. 2003. Clinical pharmacology of antifungal compounds. Infect. Dis. Clin. N. Am. 17:159-191, ix.
Kontoyiannis, D. P., M. S. Lionakis, R. E. Lewis, G. Chamilos, M. Healy, C. Perego, A. Safdar, H. Kantarjian, R. Champlin, T. J. Walsh, and I. I. Raad. 2005. Zygomycosis in a tertiary-care cancer center in the era of Aspergillus-active antifungal therapy: a case-control observational study of 27 recent cases. J. Infect. Dis. 191:1350-1360.
Meletiadis, J., J. W. Mouton, J. F. Meis, B. A. Bouman, J. P. Donnelly, and P. E. Verweij. 2001. Colorimetric assay for antifungal susceptibility testing of Aspergillus species. J. Clin. Microbiol. 39:3402-3408.
Meletiadis, J., J. W. Mouton, J. F. Meis, B. A. Bouman, P. J. Donnelly, and P. E. Verweij. 2001. Comparison of spectrophotometric and visual readings of NCCLS method and evaluation of a colorimetric method based on reduction of a soluble tetrazolium salt, 2,3-bis {2-methoxy-4-nitro-5-[(sulfenylamino) carbonyl]-2H-tetrazolium-hydroxide}, for antifungal susceptibility testing of Aspergillus species. J. Clin. Microbiol. 39:4256-4263.
Meletiadis, J., D. T. te Dorsthorst, and P. E. Verweij. 2003. Use of turbidimetric growth curves for early determination of antifungal drug resistance of filamentous fungi. J. Clin. Microbiol. 41:4718-4725.
Merz, W. G., D. Fay, B. Thumar, and D. Dixon. 1984. Susceptibility testing of filamentous fungi to amphotericin B by a rapid radiometric method. J. Clin. Microbiol. 19:54-56.
NCCLS. 2002. Reference method for broth dilution antifungal susceptibility testing of filamentous fungi; approved standard. NCCLS document M38-A. NCCLS, Wayne, Pa.
Pagano, L., M. Offidani, L. Fianchi, A. Nosari, A. Candoni, M. Piccardi, L. Corvatta, D. D'Antonio, C. Girmenia, P. Martino, and A. Del Favero. 2004. Mucormycosis in hematologic patients. Haematologica 89:207-214.
Pfaller, M. A., and M. G. Rinaldi. 1993. Antifungal susceptibility testing. Current state of technology, limitations, and standardization. Infect. Dis. Clin. N. Am. 7:435-444.
Roden, M. M., T. E. Zaoutis, W. L. Buchanan, T. A. Knudsen, T. A. Sarkisova, R. L. Schaufele, M. Sein, T. Sein, C. C. Chiou, J. H. Chu, D. P. Kontoyiannis, and T. J. Walsh. 2005. Epidemiology and outcome of zygomycosis: a review of 929 reported cases. Clin. Infect. Dis. 41:634-653.
Spellberg, B., J. Edwards, Jr., and A. Ibrahim. 2005. Novel perspectives on mucormycosis: pathophysiology, presentation, and management. Clin. Microbiol. Rev. 18:556-569.
Sun, Q. N., A. W. Fothergill, D. I. McCarthy, M. G. Rinaldi, and J. R. Graybill. 2002. In vitro activities of posaconazole, itraconazole, voriconazole, amphotericin B, and fluconazole against 37 clinical isolates of zygomycetes. Antimicrob Agents Chemother. 46:1581-1582.
Sun, Q. N., L. K. Najvar, R. Bocanegra, D. Loebenberg, and J. R. Graybill. 2002. In vivo activity of posaconazole against Mucor spp. in an immunosuppressed-mouse model. Antimicrob. Agents Chemother. 46:2310-2312.
Tobon, A. M., M. Arango, D. Fernandez, and A. Restrepo. 2003. Mucormycosis (zygomycosis) in a heart-kidney transplant recipient: recovery after posaconazole therapy. Clin. Infect. Dis. 36:1488-1491.
Immunocompromised Host Section, Pediatric Oncology Branch, National Cancer Institute, Bethesda, Maryland
1 Third Department of Pediatrics, Aristotle University, Hippokration Hospital, Thessaloniki, Greece(Charalampos Antachopoulos)