Yeast of the genus Candida are one of the most frequent causes of bloodstream infections in the US (1, 2). The five most common species of Candida, C. albicans, C. glabrata, C. parapsilosis, C. tropicalis and C. krusei comprise up to 95% of Candida infections (3–6). Other species, such as C. dubliniensis, C. lusitaniae and C. guilliermondii, comprise 1–2% each of isolates, and a few isolates each of 10–15 other Candida species make up the rest of the isolates in any given surveillance (3–5, 7). In recent population-based surveillance in two US cities, C. lusitaniae was the sixth most common species overall, surpassing C. krusei and falling just below C. dubliniensis (5, 8). In a recent report from passive US surveillance, C. lusitaniae accounted for 1.6% of 2,496 isolates (9). Globally, C. lusitaniae tends to be the 6^th or 7^th most common species in most surveillance and ranges roughly from 1–2% of the isolates (10–14).

The Clinical and Laboratory Standards Institute (CLSI) recently established species-specific breakpoint minimum inhibitory concentration (MIC) values for caspofungin, micafungin and anidulafungin against six different Candida species (15). For species/antifungals with no breakpoints, epidemiological cutoff values (ECVs) are being established, but data on ECVs for C. lusitaniae have not been well described (16–20). The ECV is the MIC value that defines the upper limit of the wild type population of MIC values and is meant to distinguish between a wild type isolate and a non-wild type isolate that may have acquired resistance or reduced susceptibility to the given antifungal agent.

Isolates of C. lusitaniae were collected from 2008–2014 as part of a population-based surveillance system in metropolitan Atlanta, GA and Baltimore City and County, MD as described (5) and from 2011–2014 in Knox County, TN and metropolitan Portland area, OR. All isolates were identified as C. lusitaniae by either DNA-based Luminex assay or D1D2 sequencing (5). From 5806 bloodstream isolates of Candida, 106 (1.8%) were identified as C. lusitaniae. The proportion of C. lusitaniae varied by site with TN and GA having the highest proportion at 2.1% of isolates, and the lowest proportion in OR at 1.3%.

The role of susceptibility testing is to detect isolates with elevated MIC values that may not respond to antifungal therapy and is generally only useful when there is a breakpoint MIC value for comparison. MIC values were generated according to the methods of the CLSI guideline M27-A3 (21) using frozen RPMI microbroth trays without indicator dye custom manufactured by TREK Diagnostics (Cleveland, OH). Results were read visually after 24 hours of incubation, and quality control isolates C. parapsilosis ATCC 22019 and C. krusei ATCC 6258 were included on each day of testing. The results for the 106 C. lusitaniae isolates are shown in Table 1.

ECV values for C. lusitaniae have been previously published in both single laboratory and multi-laboratory studies (17, 22–24). For the echinocandins, the C. lusitaniae ECV fall between the lower values seen for C. albicans and C. glabrata and the higher values seen for C. parapsilosis and C. guilliermondii. For anidulafungin, the ECV is 1.0 µg/mL and for micafungin it is 0.5 µg/mL. These values are much higher than those for C. glabrata and C. albicans (0.12 µg/mL and 0.03 µg/mL for anidulafungin and micafungin, respectively), but much lower than the values of 8 µg/mL and 2 µg/mL for C. guilliermondii and 8 µg/mL and 4 µg/mL for C. parapsilosis. We used the statistical program of Turnidge et al. (25) to calculate the ECVs for our cohort. Our ECVs at a 97.5% cutoff were 1.0 µg/mL, 0.5 µg/mL, and 2.0 µg/mL for caspofungin, micafungin, and anidulafungin, respectively. Our ECVs for micafungin and anidulafungin were equal to and one dilution higher, respectively, than those established by a large multicenter study (22). That study did not evaluate caspofungin due to technical issues that have been noted with the testing of that antifungal (26). According to the ECVs generated here, one, five, and zero isolates would be considered non-wild type by the MICs generated for caspofungin, micafungin, and anidulafungin, respectively. Using the anidulafungin ECV of 1 µg/mL, an additional isolate would be considered non-wild type.

Given that there were five isolates that had non-wild type MIC values to the echinocandins, the FKS1 hotspot1 and hotspot2 regions from all 106 C. lusitaniae isolates were sequenced to look for hotspot mutations. FKS sequencing was performed essentially as described by Zimbeck and coworkers (27) with the exception that the primers used were specific for C. lusitaniae as described by Desnos-Ollivier and coworkers (28). There was only a single isolate with an amino acid change in hotspot 1, an F640L substitution. That isolate had anidulafungin and caspofungin MICs of ≥ 1 µg/mL and a micafungin MIC of ≥ 2 µg/mL. The single isolate with all three echinocandin MIC values above the ECV (caspofungin, 8 µg/mL; micafungin 4 µg/mL; anidulafungin 2 µg/mL) had a wild type sequence (no amino acid changes) at FKS1 hotspot1 and hotspot2. Despite having no mutations in the FKS1 hotspots, 5.7% of the isolates would be considered non-wild type using the ECVs of Pfaller and coworkers (22).

Information on prior echinocandin exposure was available for 85 of the C. lusitaniae isolates. Of these, 10 (12%) had prior exposure to micafungin including the isolates with the FKS1 mutation. Only one of the isolates with prior echinocandin exposure had elevated MIC values so there was no clear correlation between MIC value and prior echinocandin exposure.

Although the MIC values for C. lusitaniae against the echinocandins are elevated, the vast majority of isolates in this population-based cohort had MIC values which were below the statistically calculated echinocandin ECV values. For the few isolates that did have MIC values above the ECV, there was no evidence that FKS1 mutations were responsible for this elevation. This would indicate that C. lusitaniae should be considered more along the lines of C. parapsilosis and C. guilliermondii, with elevated echinocandin MIC values as the norm rather than the exception. It is suggested here that isolates of C. lusitaniae that have an MIC value less than or equal to the ECV of 1.0 µg/mL for anidulafungin and 0.5 µg/mL for micafungin are truly wild type in terms of the FKS1 gene and can be treated accordingly.