The influence of inoculum preparation in EUCAST broth dilution and Etest to detect the coexistence of resistant and susceptible Candida subpopulations (defined as polyresistance [PR]) was evaluated. Cocultures of two echinocandin-resistant and susceptible clinical C. glabrata strains were used to simulate the occurrence of mixed populations in clinical samples, and antifungal susceptibility testing was performed with standard and modified approaches of inoculum preparation. Polyresistant results manifested as microcolonies or double ellipses in Etest and in single reduced optical density (OD) values (dip in OD) in microdilution. The strict inclusion of five distinct colonies of 1:5 and 1:10 resistant and susceptible cocultures led to higher rates of PR and R results compared to including one to two colonies in inoculum preparation (30% and 26% for Etest and broth dilution, respectively). Modifying the inoculum preparation by increasing the turbidity from a 2 to a 4 McFarland standard before redilution to a 0.5 McFarland standard reliably enabled the detection of resistance, with better identification of PR by Etest than by broth dilution (82% versus 32%, respectively) and of resistant minimum inhibitory concentration (MIC) values in 18% of Etests and 67% of microdilutions. The highest identification of PR succeeded with Etest and a modified 3 McFarland standard approach of inoculum preparation. Our data demonstrate that inoculum preparation as recommended and practiced does not reliably identify resistant subpopulations in polyresistant Candida cultures. By increasing the inoculum size for Etest assays from a 2 to a 4 McFarland standard with subsequent redilution, we propose a simple adaptation to increase reliability.
Keywords: Candida; antifungal resistance; antifungal susceptibility testing; inoculum preparation; polyresistance.