Abstract
This study addressed the application of anidulafungin as a surrogate marker to predict the susceptibility of Candida to caspofungin due to unacceptably high interlaboratory variation of caspofungin MIC values. CLSI reference broth microdilution methods and species-specific interpretive criteria were used to test 4,290 strains of Candida (eight species), including 71 strains with documented fks mutations. Caspofungin MIC values were compared with those of anidulafungin to determine the percentage of categorical agreement (CA) and very major (VME), major (ME), and minor error rates, as well as the ability to detect fks mutants. For all 4,290 isolates the CA was 97.1% (0.2% VME and ME, 2.5% minor errors) using anidulafungin as the surrogate. Among the 62 isolates of Candida albicans (4 isolates), C. tropicalis (5 isolates), C. krusei (4 isolates), C. kefyr (2 isolates), and C. glabrata (47 isolates) that were nonsusceptible (NS; either intermediate [I] or resistant [R]) to both caspofungin and anidulafungin, 52 (83.8%) contained a mutation in fks1 or fks2. Eight mutants of C. glabrata, two of C. albicans, and one each of C. tropicalis and C. krusei were classified as susceptible (S) to both antifungal agents. The remaining 7 mutants (2 C. albicans and 5 C. glabrata) were susceptible to one of the agents and either intermediate or resistant to the other. Using the epidemiological cutoff value (ECV) of 0.12 μg/ml for both caspofungin and anidulafungin to differentiate wild-type (WT) from non-WT strains of C. glabrata, 42 of the 55 (76.4%) C. glabrata mutants were non-WT and 8 of the 55 (14.5%) were WT for both agents (90.9% concordance). Anidulafungin can accurately serve as a surrogate marker to predict S and R of Candida to caspofungin.