Activity of a Long-Acting Echinocandin, Rezafungin, and Comparator Antifungal Agents Tested against Contemporary Invasive Fungal Isolates: SENTRY Program 2016-2018.

We evaluated the activity of rezafungin and comparators using Clinical and Laboratory Standards Institute (CLSI) broth microdilution methods against worldwide collection of 2,205 invasive fungal isolates recovered from 2016-2018. Candida spp. (1,904 isolates; 6 species), Cryptococcus neoformans (73), Aspergillus fumigatus (183) and Aspergillus flavus (45) isolates were susceptibility (S) tested for rezafungin as well as the comparators caspofungin, anidulafungin, micafungin, and azoles. Interpretive criteria were applied following CLSI published clinical breakpoints (CBP) and epidemiological cutoff values (ECV). Isolates displaying non-WT echinocandin MIC values were sequenced for fks hot spot (HS) mutations. Rezafungin inhibited 99.8% of Candida albicans isolates (MIC50/90, 0.03/0.06 μg/mL), 95.7% of Candida glabrata (MIC50/90, 0.06/0.12 μg/mL), 97.4% of Candida tropicalis (MIC50/90, 0.03/0.06 μg/mL), 100.0% of Candida krusei (MIC50/90, 0.03/0.06 μg/mL), and 100.0% of Candida dubliniensis (MIC50/90, 0.06/0.12 μg/mL) at ≤0.12 μg/m. All (329/329 [lsqb]100.0%[rsqb]) Candida parapsilosis isolates (MIC50/90,1/2 μg/mL) were inhibited by rezafungin at ≤4 μg/mL. Fluconazole resistance was detected among 8.6% of C. glabrata, 12.5% of C. parapsilosis, 3.2% of C. dubliniensis, and 2.6% of C. tropicalis Rezafungin activity against these 6 Candida spp. was similar to the activity of other echinocandins. Detection of fks HS mutation was performed by sequencing echinocandin resistant or non-WT Candida spp. isolates. Good activity was observed by fluconazole and other azoles against Cr. neoformans, whereas echinocandins, including rezafungin, displayed limited activity. Rezafungin displayed similar activity as other echinocandins against A. fumigatus and A. flavus These in vitro data contribute to accumulating research demonstrating rezafungin potential for preventing and treating invasive fungal infections.


DISCUSSION
This study provides a robust estimate of the WT MIC/MEC distributions of rezafungin for 6 species of Candida as well as A. fumigatus and A. flavus and expands upon our earlier rezafungin activity observations (31)(32)(33). Although establishing definitive ECVs and clinical breakpoints (CBPs) for rezafungin requires multicenter studies involving larger numbers of isolates of each species than the numbers used in this study (39), we suggest that the ECV determined using CLSI BMD methods in the present study is Յ0.12 g/ml for C. albicans, C. tropicalis, C. glabrata, and C. krusei (98.5% of 1,482 isolates; Table 2), Յ0.25 g/ml for C. dubliniensis (100.0% of 93 isolates), Յ4 g/ml for C. parapsilosis (100.0% of 329 isolates), and Յ0.03 g/ml for A. fumigatus (100.0% of 183 isolates) ( Table 2). Notably, these values are far below the peak plasma concentrations of 22 to 30 g/ml achievable at the 400-mg dose (15,26,27) and are equivalent to the ECVs established for these species/species groups and the clinically available echinocandins (41)(42)(43). Additional support for these ECVs is found in a recent multicenter study of rezafungin activity against Candida spp. determined using the EUCAST BMD method and both visual and statistical means of determining possible wild-type upper-limit (WT-UL) values (28). In the four-laboratory study, WT-UL cutoffs were proposed for C. glabrata (0.125 g/ml), C. krusei (0.125 g/ml), and C. parapsilosis (4 g/ml). Although interlaboratory variation precluded proposing cutoffs for C. albicans and C. tropicalis, the WT-UL statistical 97.5% endpoint was 0.063 g/ml for C. albicans and 0.25 g/ml for C. tropicalis (28). These values compare favorably with the ECVs generated by the CLSI BMD method in the present study. Although an essential agreement rate (Ϯ2 dilution steps) of 92.0% for C. albicans and 100.0% for C. glabrata, C. parapsilosis, C. tropicalis, and C. krusei between CLSI and EUCAST methods for rezafungin was observed previously (31), alignment between CLSI and EUCAST susceptibility profiles and breakpoints is yet to be determined, as significant interlaboratory EUCAST MIC variability (likely attributed to nonspecific binding of the drug to plastics) has been identified for rezafungin against a more susceptible collection of Candida spp. (28,44).
As seen in Table 4, the highest rezafungin MIC values for fks mutant strains of C. albicans and C. glabrata were 0.25 g/ml and 2 g/ml, respectively. Both of these MIC values for mutant strains are within the range of concentrations that Bader et al. (20) estimated would achieve percent probabilities of PK-PD target attainment of 100% through week 6, suggesting that weekly regimens of rezafungin can achieve exposures associated with efficacy against some fks mutant Candida isolates (20). In addition, the same study showed that the mutant prevention concentration, the concentration of drug that would inhibit emerging resistant mutants, for both rezafungin and micafungin was 16 g/ml (27). Given that the high plasma drug exposure of rezafungin easily exceeds the mutant prevention concentration for Candida, a possible advantage of rezafungin may be to prevent the development of resistance to the echinocandin class of antifungal agents (20,22,24,27).
Expert panel guidelines from both NA (5) and EUR (12) favor step-down therapy to fluconazole or voriconazole for patients with candidiasis in specific clinical situations, that is, when clinical improvement and the clearance of Candida from the bloodstream are achieved by the initial echinocandin therapy. In addition, the organism must be susceptible to fluconazole (e.g., C. albicans, C. parapsilosis, and C. tropicalis) or voriconazole (e.g., C. krusei). Unfortunately, antifungal susceptibility testing is still not routinely available in many patient care settings. In these circumstances, clinicians are forced to rely on simple identification of the Candida species as a predictor of fluconazole susceptibility (5,12). In most instances, isolates of C. albicans, C. parapsilosis, and C. tropicalis are considered to be reliably susceptible to fluconazole (16), whereas C. glabrata and C. krusei are considered to be intrinsically less susceptible or resistant and are suboptimal targets for using fluconazole (5,12). This approach may be seriously flawed if fluconazole resistance emerges among the traditionally susceptible species. Concern regarding this approach has been raised by Oxman et al. (45), who found that despite the small proportion of C. albicans, C. parapsilosis, and C. tropicalis isolates with resistance/decreased susceptibility to fluconazole, these species comprised 36% of the reduced-susceptibility group (including C. glabrata and C. krusei), potentially compromising therapy with the resultant clinical failure. These concerns are supported by data from the current survey showing that the rate of resistance to fluconazole was 0.4% for C. albicans, 12.5% for C. parapsilosis, and 2.6% for C. tropicalis (Tables 3 and 5). In aggregate, these three normally susceptible species account for 31% of all fluconazoleresistant isolates. Species identification should be used cautiously as the sole criterion for anti-Candida agent selection (5,45).
The increased rate of fluconazole resistance among the C. parapsilosis (12.5% overall) and C. tropicalis (2.6% overall) isolates in the present study is important, as these species are the non-C. albicans species most commonly isolated in LATAM (Table 1). Although less common than C. glabrata in EUR, the rate of fluconazole resistance of 24.8% among C. parapsilosis isolates exceeds the rate observed among C. glabrata (6.0%) isolates and is cause for alarm (Table 5).
This survey has some limitations, as noted elsewhere (16): the SENTRY Surveillance Program is a sentinel surveillance and is not population based; therefore, we may overor underestimate the activity of the tested agents. In addition, we do not collect data concerning antifungal use or outcomes of therapy. The purpose of the SENTRY Program is to identify trends in antifungal resistance and to document the emergence of new species as well as the activity of new and established agents against key fungal pathogens. The broad geographic distribution, the longitudinal nature of the surveillance, and the use of molecular and proteomic identification methods and determination of resistance mechanisms are strengths of the SENTRY Program.
In conclusion, we have provided additional in vitro data demonstrating the activity of rezafungin against a collection of largely echinocandin-WT isolates of Candida spp., C. neoformans, A. fumigatus, and the A. flavus species complex. Given these findings, we suggest that MIC values of Յ0.12 g/ml (C. albicans, C. glabrata, C. tropicalis, and C. krusei), Յ0.25 g/ml (C. dubliniensis), and Յ4 g/ml (C. parapsilosis) and a MEC of Յ0.03 g/ml (A. fumigatus) approximate the ECV/WT-UL MIC/MEC distributions for rezafungin and the common species of Candida and Aspergillus. Further evaluations including at least 100 MIC values per species tested by three different laboratories should be performed to define the ECVs for rezafungin, a fundamental step in establishing clinical breakpoints.
This survey provides new information regarding emerging fluconazole resistance among C. parapsilosis and C. tropicalis clinical isolates from geographic regions beyond NA, in addition to demonstrating evidence of the sustained activity of rezafungin and the other echinocandins against Candida and Aspergillus species. Whereas the highest rates of fluconazole resistance in NA isolates were seen in C. glabrata (13.2%), fluconazole-resistant C. parapsilosis (24.8%) was most prominent in EUR and fluconazole-resistant C. tropicalis was most prominent in APAC (5.0%) and LATAM (4.1%). In all three instances, the rate of fluconazole resistance was highest in species of Candida other than C. glabrata. Species identification should be used cautiously as the sole criterion for selecting antifungal therapy.
Fungal identification methods. Yeast isolates were subcultured and screened using CHROMagar Candida (Becton, Dickinson, Sparks, MD) to ensure purity. Matrix-assisted laser desorption ionizationtime of flight mass spectrometry (MALDI-TOF MS) was applied for the identification of all yeast isolates using a MALDI Biotyper apparatus according to the manufacturer's instructions (Bruker Daltonics, Billerica, MA). Isolates that were not identified by proteomic methods were submitted to the previously described sequencing-based methods (43,46,47).
Molds were cultured and identified by MALDI-TOF MS or DNA sequencing analysis when an acceptable identification was not achieved by MALDI-TOF MS. The sequences of the 28S ribosomal DNA and ␤-tubulin genes of Aspergillus spp. were analyzed (47)(48)(49)(50).
Antifungal susceptibility testing. All isolates were tested by CLSI BMD methods as described in documents M27 (37) and M38 (38). Only systemically active antifungal agents were tested, including rezafungin, anidulafungin, micafungin, caspofungin, itraconazole, fluconazole, voriconazole, posaconazole, and amphotericin B. The ranges of antifungal agent concentrations tested were 0.008 to 4 g/ml for itraconazole, posaconazole, and voriconazole, 0.12 to 2 g/ml for amphotericin B, and 0.12 to 128 g/ml for fluconazole. The echinocandin concentration range tested during 2016 and 2017 was 0.008 to 4 g/ml, whereas this range was expanded to 0.002 to 4 g/ml in 2018. MIC results were determined visually after 24 h (Candida spp.), 48 h (Aspergillus spp.), or 72 h (C. neoformans) of incubation at 35°C. Azole and echinocandin MIC values against yeasts were read as the lowest concentration of drug that resulted in Ն50% inhibition of growth relative to that of the growth control. Complete (100%) inhibition was used to determine itraconazole, posaconazole, and voriconazole MIC values against Aspergillus spp. and amphotericin B MIC values against yeasts and molds. Echinocandin minimum effective concentration (MEC) values, including those of rezafungin, were read against Aspergillus spp. as described in CLSI document M38 (38).
Echinocandin, fluconazole, and voriconazole susceptibility categories were applied for the five most common species of Candida (C. albicans, C. tropicalis, C. parapsilosis, C. glabrata, and C. krusei) following CLSI clinical breakpoints (CBPs) (40). Epidemiological cutoff values (ECVs/ECOFFs) were used to differentiate wild-type (WT) from non-wild-type (NWT) isolates of the species for which there are no CLSI CBPs (39,41). Neither CBPs nor ECVs/ECOFFs have been determined by CLSI methods for rezafungin against Candida, Aspergillus, or Cryptococcus spp. For comparison, we established tentative ECVs for rezafungin and each species using the iterative statistical method recommended by CLSI (28,32,(39)(40)(41). These ECVs must be considered tentative, given the CLSI requirement that ECVs be determined using MIC/MEC data acquired from a minimum of three different laboratories including at least 100 MIC/MEC values from 100 individual isolates, all determined by CLSI reference methods (39).
QC. To ensure proper test conditions and procedures, concurrent quality control (QC) testing was performed. The QC strains recommended by CLSI included C. krusei ATCC 6258, C. parapsilosis ATCC 22019, A. flavus ATCC 204304, and A. fumigatus ATCC MYA-3626.
Screening for 1,3-␤-D-glucan synthase mutations. All Candida isolates that were echinocandin resistant or that showed MIC values higher than the ECV for any echinocandin were submitted to whole-genome sequencing for detecting mutations in the HS regions of fks1 and fks2 (C. glabrata only) as described previously (43,48,50).