Isolation of Candida species on media with and without added fluconazole reveals high variability in relative growth susceptibility phenotypes.

Mouthwashes from human immunodeficiency virus-positive individuals were sampled for yeasts by direct plating on a differential agar medium with and without added fluconazole and via enrichment broths with and without added fluconazole. The colonies of the yeasts isolated were tested for relative growth in the presence of single concentrations of itraconazole and fluconazole. Among 258 culture plates containing yeasts obtained via different isolation routes from 86 yeast-positive samples, 33 (12.7%) of the plates showed unexpectedly high colony-to-colony variation in relative growth. Intercolony variation was seen in 41 (47.7%) of the 86 isolates when relative growth data were analyzed for all colonies of an isolate tested, regardless of the medium used for isolation. The prevalence of relative growth variability with the azoles was highest for Candida glabrata (100% of 13 isolates), followed by Candida krusei (60% of 5 isolates) and Candida albicans (40% of 53 isolates), and the visual patterns of variability seen in scatter plots of the data showed species specificity. Relative growth phenotypes generally tended to be stable for each yeast colony in subcultures, whether or not the medium used for subculture contained antifungal agents. DNA fingerprinting of stable and variable C. albicans isolates showed changes in band patterns detected with the probe Ca3, suggesting that the variability may have resulted from selection of different subtypes of the yeasts during the isolation procedure. These findings suggest that the yeasts isolated from single clinical samples were often not clonal in nature. The relative growth test revealed colony variability more readily than conventional susceptibility testing.