Isolates received at the French National Reference Center for Mycoses and Antifungals (NRCMA) are identified to the species level by standard mycologic procedures and routinely tested for susceptibility to caspofungin, micafungin, and anidulafungin by using European Committee for Antimicrobial Susceptibility Testing (EUCAST) methods (6) and AM3 medium (7). In addition, RPMI 1640 medium was used here for selected isolates and reference strains. For the clinical isolates with caspofungin MIC >0.5 µg/mL in AM3, nucleotide sequences of hot spot (HS) 1 and 2 regions of the FKS1 gene for C. albicans and C. krusei and of HS1 region of FKS1, FKS2, and FKS3 genes for C. glabrata were determined (7,8).

The resulting protein sequences were aligned with the BioloMics software (BioloMics, BioAware SA, Hannut, Belgium) and compared with reference strains (C. albicans, ATCC32354; C. krusei, ATCC6258; and C. glabrata, ATCC2001). Genetic relatedness of C. albicans and C. glabrata paired isolates was studied by using microsatellite-length polymorphism analysis (9–11). The Wilcoxon signed-rank test was used to compare echinocandin MICs of paired isolates. Surveillance for mycoses by the NRCMA has been approved by the Institut Pasteur Internal Review Board and by the Commission Nationale de l’Informatique et des Libertés.

During September 2004–April 2010, twenty proven infections caused by C. albicans (n = 10), C. glabrata (n = 8), or C. krusei (n = 2) with caspofungin MIC >0.5 µg/mL and a mutation in the target enzyme were reported to the NRCMA (Table 1). Nineteen of the isolates were recovered after caspofungin treatment for a median duration of 27 days (range 10–270 days; 13 of 19 patients received caspofungin at the time the resistant isolate was recovered). Caspofungin was prescribed for 14 patients with proven Candida spp. infection, 1 patient with proven invasive aspergillosis, and 2 patients with febrile neutropenia; for 2 persons with hematologic malignancies, caspofungin was prescribed prophylactically.

The geometric mean MIC for C. glabrata and C. albicans were 2.8 and 1.7 µg/mL for caspofungin, 0.4 and 0.7 µg/mL for micafungin, and 0.2 and 0.09 µg/mL for anidulafungin, respectively (Table 2). Of the 20 mutated isolates found resistant to caspofungin in AM3 by using the EUCAST method, 19 also were resistant to caspofungin (1 intermediate), 18 to micafungin (1 intermediate and 1 susceptible), and 9 to anidulafungin (5 intermediate and 6 susceptible) according to Clinical Laboratory Standards Institute (CLSI) breakpoints and RPMI 1640 medium (Table 2). According to EUCAST breakpoints, 19 isolates also were resistant to anidulafungin, and 1 isolate was almost resistant (MIC 0.03 µg/mL). We thus showed discrepancies between CLSI and EUCAST regarding anidulafungin susceptibility (www.srga.org/eucastwt/MICTAB/EUCAST%20clinical%20MIC%20breakpoints%20-%20antimicrobials%20for%20Candida%20infections.htm [V 3.0 2011–4-27]) (12,13).

Of the 10 caspofungin-resistant C. glabrata isolates, 8 harbored a mutation in Fks2p only, 1 isolate had a mutation in Fks1p, and 1 had mutations in Fks1p and Fks2p (Table 2). Of the 8 caspofungin-resistant C. albicans isolates, 1 had a missense mutation in HS2, and 1 had a combination of 2 heterozygous mutations in HS1. The other 6 isolates harbored 4 different mutations in HS1 (Table 2). Finally, the 2 C. krusei isolates had 2 different mutations in HS1 region. Of the 20 mutated isolates, 6 harbored 7 mutations not yet described in the literature (Table 2) (13).

Prior initial isolates available for 12 patients had the wild-type sequence for the HS regions that were mutated in the paired resistant isolate. All initial isolates were susceptible to anidulafungin and to micafungin and anidulafungin according to EUCAST and CLSI, respectively (data not shown). According to CLSI caspofungin breakpoints, 5 of 6 initial isolates of C. albicans were susceptible, and 1 was intermediate; 4 of 5 C. glabrata isolates were resistant (0.5 µg/mL), and 1 was intermediate; and the C. krusei isolate was resistant (1 µg/mL). For each of the 12 pairs, MICs increased significantly (from 3 to 8 dilutions for caspofungin and micafungin and from 1 to 8 dilutions for anidulafungin) between the wild-type and the mutant isolate (Figure; p<0.001). Genetic relatedness was demonstrated for all C. albicans and C. glabrata paired isolates.

We demonstrated that recent exposure to caspofungin altered the distribution of species causing Candida bloodstream infections (14), and that caspofungin exposure was independently associated with fungemia associated with intrinsically less-susceptible species in hematology (15). Echinocandin resistance in Candida spp. is still uncommon (4,13). Through our surveillance program, we estimated the incidence of decreased susceptibility to caspofungin associated with FKS mutations among C. albicans, C. glabrata, and C. krusei isolates responsible for candidemia in children and adults in Paris at 6 (0.4%) of 1,643 (NRCMA, unpub. data). We report proven caspofungin-resistant Candida spp. infections with none of the isolates belonging to the intrinsically less-susceptible species C. parapsilosis or C. guilliermondii.

We determined antifungal susceptibility testing by the EUCAST technique using AM3 because it enables better discrimination between susceptible wild-type and resistant mutant isolates (7). All isolates with high caspofungin MIC (>0.5 µg/mL) had mutation in the HS1 and/or HS2 region of FKS genes. The mutations were not restricted to a given position but were diverse, especially for C. albicans with 6 different mutations among the 8 resistant isolates; 5 different mutations were observed among the 10 C. glabrata resistant isolates. Most mutations in C. glabrata isolates were in Fks2p. Two mutations in C. albicans, 2 patterns of mutation in C. glabrata, and 1 mutation in C. krusei had not been reported before, highlighting the great mutation diversity that could be responsible for echinocandin resistance (13).

All but 1 patient had received caspofungin (70 mg on day 1, then 50 mg/d) before recovery of the resistant isolate, with a variable duration of exposure (<10 days to >8 months), in agreement with the literature (5 [3] to 420 days). In addition, 13 of 19 patients received caspofungin at the time of recovery of the resistant isolate. Most patients had malignancy, but 7 intensive care unit hospitalizations also were recorded. Echinocandins MICs between the wild-type parent and the subsequent mutant isolate increased by up to 8 log₂ dilutions (Figure). The source of the resistant isolate is not unequivocal; it was acquired from the environment as an already resistant isolate or from the patient’s own flora under drug pressure. Our genotyping results favor the second hypothesis. This study suggests in France the emergence of infections from acquired echinocandin resistance in usually susceptible Candida spp. in patients preexposed to caspofungin, which highlights the need for careful species identification, antifungal drug susceptibility testing, and evaluation of prior drug exposure before antifungal drug prescription.