Gradient diffusion strips for detecting azole resistance in Aspergillus fumigatus sensu lato.

Abstract

Studies comparing gradient diffusion strips (GDSs) and the EUCAST E.Def 9.4 microdilution method are scarce, thwarted by a low number of isolates, and restricted to selected antifungal agents. We evaluated the performance of GDSs to detect azole resistance in A. fumigatus, including cryptic species. A. fumigatus sensu stricto (n=89) and cryptic species (n=52) were classified as susceptible or resistant to itraconazole, voriconazole, posaconazole and isavuconazole (EUCAST E.Def 9.4; clinical breakpoints v10). A. fumigatus sensu stricto azole-resistant isolates had the following cyp51A gene mutations: TR34 -L98H (n=24), G54R (n=5), TR46 -Y121F-T289A (n=1), F46Y-M172V-N248T-D255E-E427K (n=1), F165L (n=1) and cyp51A gene wild type (n=3). GDSs (ETEST®, bioMèrieux, Marcy-l'Etoile, France and Liofilchem®, Roseto degli Abruzzi, Italy) MICs were obtained by following the manufacturer's guidelines. For A. fumigatus sensu stricto, itraconazole MICs >1.5mg/L, voriconazole >0.38 mg/L, posaconazole >0.75 mg/L, and isavuconazole >0.5mg/L correctly separated resistant from susceptible isolates with two exceptions. Considering the aforementioned cut-off MICs, sensitivity/specificity values of GDSs to detect azole resistance were: itraconazole (97%/100%), voriconazole (97%/100%), posaconazole (97%/100%) and isavuconazole (93.3%/100%). For cryptic species isolates, voriconazole MICs >1mg/L and isavuconazole >0.75 mg/L separated resistant isolates from susceptible isolates with 15 and 27 exceptions, respectively. Considering the aforementioned cut-off MICs, sensitivity/specificity values were as follows: voriconazole (68.1%/100%) and isavuconazole (25%/100%). For itraconazole and posaconazole, it was not possible to establish cut-off values. We set tentative cut-off MIC values to correctly spot resistant Aspergillus fumigatus sensu stricto isolates using GDSs. The performance against cryptic species was poor.