Early phenotypic detection of fluconazole- and anidulafungin-resistant Candida glabrata isolates

Abstract

Background: Increased fluconazole and echinocandin resistance in Candida glabrata requires prompt detection in routine settings. A phenotypic test based on the EUCAST E.DEF 7.3.2 protocol was developed for the detection of fluconazole- and anidulafungin-resistant isolates utilizing the colorimetric dye XTT.

Methods: Thirty-one clinical C. glabrata isolates, 11 anidulafungin resistant and 14 fluconazole resistant, were tested. After optimization studies, 0.5-2.5 × 105 cfu/mL of each isolate in RPMI 1640 + 2% d-glucose medium containing 100 mg/L XTT + 0.78 μΜ menadione and 0.06 mg/L anidulafungin (S breakpoint) or 16 mg/L fluconazole (I breakpoint) in 96-well flat-bottom microtitration plates were incubated at 37°C for 18 h; we also included drug-free wells. XTT absorbance was measured at 450 nm every 15 min. Differences between the drug-free and the drug-treated wells were assessed using Student's t-test at different timepoints. ROC curves were used in order to identify the best timepoint and cut-off.

Results: The XTT absorbance differences between fluconazole-containing and drug-free wells were significantly lower for the resistant isolates compared with susceptible increased exposure isolates (0.08 ± 0.05 versus 0.25 ± 0.06, respectively, P = 0.005) at 7.5 h, with a difference of <0.157 corresponding to 100% sensitivity and 94% specificity for detection of resistance. The XTT absorbance differences between anidulafungin-containing and drug-free wells were significantly lower for the resistant isolates compared with susceptible isolates (0.08 ± 0.07 versus 0.200 ± 0.03, respectively, P < 0.001) at 5 h, with a difference of <0.145 corresponding to 91% sensitivity and 100% specificity, irrespective of underlying mutations.

Conclusions: A simple, cheap and fast phenotypic test was developed for detection of fluconazole- and anidulafungin-resistant C. glabrata isolates.

Figure 1.

Metabolic curves generated with 100 mg/L XTT/0.78 μΜ menadione for an anidulafungin- and fluconazole-resistant C. glabrata isolate (SSI-3203) (left-hand panels) and an anidulafungin-susceptible and fluconazole-susceptible increased exposure C. glabrata isolate (SSI-7965) (right-hand panels). An increase around 4 h was observed for all of the curves, probably because of multiplication of Candida cells detected by the spectrophotometer. ANI, anidulafungin; FLC, fluconazole; S, susceptible; I, susceptible increased exposure; R, resistant. This figure appears in colour in the online version of JAC and in black and white in the print version of JAC.

Figure 2.

ROC curves and deltaXTT-ABS for resistant and susceptible C. glabrata isolates for anidulafungin (a and b) and fluconazole (c and d). Horizontal broken lines correspond to optimal deltaXTT-ABS of 0.145 and 0.157 for detecting resistance to anidulafungin and fluconazole, respectively. IE, increased exposure.

Figure 3.

Examples of metabolic curves that showed false resistance to fluconazole, due to delayed metabolic activity of the drug-free growth control (a), or false susceptibility to anidulafungin, due to delayed metabolic activity in the drug-containing well (b). ANI, anidulafungin; FLC, fluconazole; GC, growth control; I, susceptible increased exposure; R, resistant. This figure appears in colour in the online version of JAC and in black and white in the print version of JAC.