Candida auris: An Overview of How to Screen, Detect, Test and Control This Emerging Pathogen

The multidrug-resistant yeast Candida auris is associated with invasive infections in critically ill patients and has been isolated in different countries worldwide. Ease of spread, prolonged persistence in the environment and antifungal drug resistance pose a significant concern for the prevention of transmission and management of patients with C. auris infections. Early and correct identification of patients colonized with C. auris is critical in containing its spread. However, this may be complicated by C. auris strains being misidentified as other phylogenetically related pathogens. In this review, we offer a brief overview highlighting some of the critical aspects of sample collection, laboratory culture-dependent and independent identification and the susceptibility profile of C. auris.


Introduction
Candida auris, the yeast pathogen firstly isolated from a Japanese patient's external ear canal in 2009, has been involved in invasive healthcare-associated outbreaks and sporadic cases reported in various countries worldwide [1][2][3][4][5][6]. Three hundred forty-nine cases of C. auris were reported solely in the European Union between January 2018 and May 2019: most of these (73.6%) were colonizations and, among the infections, 24.1% were bloodstream infections [7].
This new microorganism has a severe impact on public health, not only because it is often multidrug-resistant, rapidly spreads among patients and persistently colonizes the skin and nosocomial surfaces, but also because it is often difficult to be correctly identified [8].
The European Centre for Disease Prevention and Control (ECDC) C. auris survey collaborative group reported that during 2013-17 several European countries lacked laboratory capacity and/or information on the incidence of cases at a national level [9]. Even if laboratory facilities and awareness seem to have improved since 2018, seven EU/EEA countries do not have a reference laboratory in their country yet [7]. Therefore, cases of C. auris may still be misidentified or unidentified and spread in healthcare settings.
Misidentification of C. auris with other yeasts (e.g., C. haemulonii, C. famata, C. guilliermondii, C. lusitaniae, C. parapsilosis) may occur due to the use of standard biochemical methods and commercially

Phenotypic Methods and Distinctive Characteristics
The morphology of C. auris may resemble other more common Candida spp., thus making in vitro evaluation of colonies appearance impossible as the only laboratory identification method [16]. C. auris forms smooth and white/cream colonies on Sabouraund dextrose/glucose agar. In contrast, when growing on commercial chromogenic Candida agar medium, it forms pink, beige, pale rose or red colonies, which may be difficult to distinguish from C. glabrata [16].
Supplementation of commercial chromogenic media with Pal (sunflower seed extract) agar was reported to be useful, together with higher temperature, for the differentiation of C. auris from the C. haemulonii complex [17].
Recently, Borman and colleagues have also described a chromogenic agar, CHROMagar Candida Plus, for the specific identification of C. auris isolates. On this medium, C. auris colonies appear pale cream but present a distinctive blue halo. Of note, among over 50 different species of Candida spp and related genera that were cultured in parallel, only Candida diddensiae gave a similar appearance [18].
On microscopic examination, C. auris is characterized by oval or elongated yeast cells capable of existing as aggregate or nonaggregate cells. C. auris rarely forms pseudohyphae, depending on specific conditions of the media, such as the presence of NaCl [19]. C. auris shows multiple cellular morphologies when cultivated at 25, 37, and 40 • C using different growth media [20]. On Lee's glucose and Lee's GlcNAc media, C. auris cells exhibit an oval shape at 25 and 37 • C, and a relatively round shape at 40 • C. On Spider and agar plus serum media, cells are round and relatively small. C. auris cells are round on regular Yeast Extract-Peptone-Dextrose (YPD) Medium. At the same time, they take on an elongated shape on YPD plus 10% NaCl. Elongated cells of C. auris resemble opaque cells of C. albicans in shape. A small percentage of highly elongated and pseudohyphal-like cells are observed when grown on YPD plus 10% NaCl. Multiple nuclei are observed with DAPI staining in elongated cells [21]. However, no septin/chitin rings are observed between conjoint cells when Calcofluor white staining is applied [22].
Morphological diversity is a key virulence factor of Candida spp. [23,24]. However, it has been demonstrated that the pathways governing yeast-to-filament transition during morphogenesis and the related signals are different in C. albicans and C. auris [25]. Finally, C. auris can form biofilm, another virulent trait, which seems to be related to the type and phenotypic behavior of the isolates, as sessile and planktonic phenotypes were found to be associated with colonizing and clinical isolates, respectively [26].

Biochemical Methods
Most commercial biochemical tests commonly used may misidentify C. auris, as reported in a list of the most used methodologies for the identification of Candida species compiled and updated by the Centre for Disease Prevention and Control (CDC) ( Table 1) [12]. C. auris may be misidentified as Candida haemulonii, Candida sake, Rhodotorula glutinis or other Candida species by these conventional biochemical systems because it is not always present in databases or because of the overlapping biochemical profiles [27]. In the case of the listed identification/instrument combinations, it is recommended to use the practical identification algorithms available in the CDC website [12].
It has been reported that VITEX 2 XL (bioMérieux version 8.01) is capable of detecting C. auris with high reliability. However, Ambaraghassi and colleagues showed that this software precisely identifies isolates from the South American clade, but had limited power to correctly identify C. auris from the African and East Asian clades [28]. Hence, all isolates identified using this system, as C. auris, C. famata, and species in the C. haemulonii complex should be confirmed by MALDI-TOF or DNA sequencing. Therefore, the methods listed above are currently not precise enough to identify this microorganism. Moreover, the slow turnaround time of enrichment cultures could be a significant limiting factor [29].
Moreover, they were confirmed by testing on large yeast panels and are time-saving as they reduce the overall turnaround time from days to working hours, allowing the rapid identification of colonized patients [29,32,33,37,40,41,46]. Lima and colleagues set up a TaqMan-based real-time PCR assay, which showed to be highly sensitive (PCR linearity of over 5 orders of magnitude with 99% efficiency), had a limit of detection (LoD) of 1 CFU (Ct = 34.3 ± 0.5) and had high specificity [43]. The use of molecular methods from swabs is also simple: commonly 100-200 µL of the total 1 mL of swab medium is used for DNA isolation, while the residual volume should be stored until results are obtained [29].
Currently, a T2 magnetic resonance assay (T2 Biosystems, United States) is available for the rapid (<3 h) and accurate and sensitive (1-3 CFU/mL) detection of specific Candida species (C. albicans/C. tropicalis, C. parapsilosis and C. glabrata/C. krusei) directly from blood specimens [41,42]. Moreover, with the addition of the C. auris panel, it detects this microorganism with a limit of detection < 5 CFU/mL in less than 5 h [41].
Kordalewska and colleagues developed two rapid, accurate, easy-to-perform molecular diagnostic assays based on real-time PCR to discriminate C. auris from other species. While the first assay identifies C. auris only, the second one-thanks to the use of SYBR Green (ThermoFisher Scientific, Waltham, Massachusetts) detection of amplicon and melting-point analysis-distinguishes between C. auris, C. duobushaemulonii, C. haemulonii, and C. lusitaniae [31]. Both PCR methods are highly reliable as they yielded 100% accuracy and concordance with the results of ITS region sequencing [31]. Of note, molecular methods may be costly (less than the MALDI-TOF) and detect DNA from both live and dead C. auris cells; in the case of a positive result from a nonsterile specimen (skin, environment), it could indicate past or present colonization with C. auris.

MALDI-TOF
MALDI-TOF mass spectrometry (MS) can reliably differentiate C. auris from colonies recovered in the enrichment broth, distinguishing them from other Candida spp. only if C. auris spectrum is included in the reference database [47]. Precise identification of C. auris is currently more likely with the FDA-approved MALDI-TOF Biotyper (Bruker-Daltonics), using their updated research use only (RUO) library (Versions 2014 5627 and more recent) or CA System library (Version Claim 4), or with the VITEK (MALDI-TOF) MS (bioMérieux) FDA-approved IVD v3.2 or the RUO Version 4.14 with the Saccharomycetaceae update [12,29].
Some authors reported higher accuracy and efficiency for the Bruker Biotyper over the Vitek 2 MS in the identification of C. auris isolates, even with an upgraded RUO [48].
In addition to these options, the CDC, in collaboration with Bruker, provides an online tool for the accurate classification of C. auris to the species level (https://www.cdc.gov/microbenet/index.html) [49]. Three different MALDI-TOF MS extraction protocols (on-plate, quick and extended tube) are available. Some authors recommended the use of the quick tube extraction method as it scored better confidence levels than the on-plate method, while not differing significantly from the extended method [50]. In spite of this, other authors indicated that the extended tube extraction for MALDI-TOF MS was not better than on-plate extraction [51]. Therefore, MALDI-TOF MS reliably and rapidly identifies C. auris isolates, compared to conventional identification methods. However, it represents a time-consuming method when applied to all colonies recovered from colonized patients and requires a substantial upfront investment [18].

Typing
Multilocus sequence typing (MLST), a proteomic analysis performed with MALDI-TOF MS, amplified fragment length polymorphism (AFLP) and whole genome sequencing (WGS), can be used to evaluate the genetic relatedness of C. auris isolates [47]. For MLST analysis, a set of four genetic loci, namely ITS, D1/D2, RPB1 and RPB2, have been reported to be highly discriminatory for strain differentiation between C. haemulonii complex and C. auris [51].
For AFLP analysis, genomic DNA was subjected to a combined restriction-ligation procedure containing EcoRI and MseI restriction enzymes (New England Biolabs, Beverly, MA, USA) and complementary adaptors [52]. However, whole genome sequencing remains the gold standard for the determination of C. auris isolates clonality as sequence variation between individual strains within a clonal lineage is very small (typically 30-80 SNPs over a whole genome) [52][53][54][55].
Recently, de Groot and colleagues developed a short tandem repeat (STR) typing assay for C. auris. In concordance with WGS analysis, the authors identified five major different C. auris clusters (i.e., South American, South Asian, African, East Asian and Iranian) and most isolates differing by >30 SNPs were determined [56].
In a recent study, Vatanshenassan and colleagues compared different typing techniques (i.e., microsatellite typing, AFLP fingerprinting, ITS sequencing, MALDI-TOF MS and IR Biotyper FTIR spectroscopy) to evaluate their application in typing C. auris [57]. Results indicated microsatellite typing as the tool of choice for C. auris outbreak investigations because only this technique grouped the isolates into four main clusters, in accordance with WGS data. The other typing tools showed poor performances with the highest agreement between microsatellite typing and ITS sequencing with 45% similarity, followed by microsatellite typing and FTIR with 33% similarity. The lowest agreement was observed between FTIR spectroscopy, MALDI-TOF MS and ITS sequencing.

C. auris Resistance Profile
One of the most concerning aspects of C. auris is its ability to develop resistance to all three of the main classes of antifungal drugs (azoles, echinocandins and polyenes), with severely limiting clinical Antibiotics 2020, 9, 778 7 of 14 and therapeutic management options [58][59][60][61]. Multidrug resistance (MDR), i.e., resistance to more than two antifungal classes, is observed in around 40% ≥ of C. auris isolates, with a small proportion (4%) exhibiting resistance to all classes of antifungals. Therefore, when in the presence of an unidentified yeast resistant to more than one antifungal drug, further testing for C. auris identification should be performed [8,14,16,55,62].
Recently, among 801 patients identified with C. auris in New York, three were found to have pan-resistant C. auris that had developed after treatment with antifungal medications, including echinocandins. All three patients had multiple comorbidities, but no recent domestic or foreign travel [63].
Zamith−Miranda and colleagues compared two clinical isolates of C. auris with distinct drug susceptibility profiles with a C. albicans reference strain using a multiomics approach. Their results highlighted that, despite the distinctive drug resistance profile, C. auris isolates were very similar. In contrast, their profile was different from that of C. albicans, both in terms of carbon utilization and in lipid and protein content, supporting a multifactorial mechanism of drug resistance [64].
In general, high levels of resistance to fluconazole are observed in C. auris, especially among isolates from India, where a study of 350 isolates showed around 90% minimal inhibitory concentrations (MICs) of fluconazole greater than 16 µg/mL, and South Africa [65][66][67][68][69][70]. C. auris can also, to a lesser extent, exhibit reduced susceptibility to other triazole antifungals (i.e., voriconazole, posaconazole, itraconazole and isavuconazole) [69]. As regards mechanisms of resistance to antifungals, we have already published a systematic review on this topic, so we will add, if available, only updated information [16]. C. auris commonly exhibits susceptibility to the polyene amphotericin B; however, geographical differences are found. For example, resistance to amphotericin B was detected in 30% of C. auris U.S. isolates [71]. The principal mechanism of amphotericin B resistance has not yet been identified in C. auris. Still, some authors related it to a reduction in ergosterol content in the cellular membrane [8]. Finally, the development of resistance to echinocandins-currently the first-line therapy drugs-has been observed in C. auris isolates from multiple geographic areas in patients initially treated with an echinocandin [8].

Antifungal Susceptibility Testing
Antifungal susceptibility testing (AFST) for C. auris can be performed using CLSI and EUCAST broth microdilution methods, as well as the E-test gradient diffusion method (bioMérieux), Sensititre YeastOne (Thermo Fisher Scientific) and Vitek-2 Yeast susceptibility system (bioMérieux). Nonetheless, these in vitro methodologies may present limitations, such as slow turnaround time (24 h after isolation) and, not lastly, the need of specific know-how to readout C. auris MIC values [29]. In Table 3 we have reported the principal AFST and their respective strengths and limitations.
Currently, no established susceptibility breakpoints are available for C. auris. Some epidemiologic cut-off values (ECVs) have been suggested by Arendrup and colleagues comparing the European Committee on Antimicrobial Susceptibility Testing (EUCAST) and the Clinical and Laboratory Standards Institute (CLSI) reference microdilution MICs, which appeared to have minor differences [65]. However, it must be said that ECVs are tentative breakpoints and isolates with MICs below or above the ECV should not be classified as susceptible or resistant, respectively, especially if considering that MIC distributions can vary substantially for C. auris isolates from different clades [66,72]. Moreover, the CDC provides guidance for C. auris MIC interpretation, based on information gathered for Candida spp. and several expert opinions (www.cdc.gov/fungal/candida-auris/c-auris-antifungal.html).
Talking about the need of specific know-how to readout C. auris MIC values, we must mention the possibility of running into the "eagle" effect or paradoxical growth effect of echinocandins, which is the reduced antifungal behavior at higher doses of the drug in vitro. It is important to underline that the eagle effect does not reflect the potential in vivo response of C. auris to echinocandins, as indicated in an invasive murine candidiasis model [73]. Table 3. Strengths and limitations of AFST used for C. auris.

Strengths and Limitations Reference
Broth Microdilution Methods/Sensititre YeastOne Determined ECVs are valuable in the analysis of MICs of isolates from the South Asian clade. MIC distributions can vary substantially for C. auris isolates from different clades. Are easy to perform. [20] E-test gradient diffusion method Difficulty of interpretation for presence of aggregate directly adjacent to the zone of growth inhibition. The aggregates are present for evaluation of fluconazole, voriconazole, and anidulafungin but not in experiments performed with flucytosine or amphotericin B. [20] VITEK 2 MIC distributions can vary substantially for C. auris isolates from different clades [28] MBT ASTRA MBT ASTRA has a potential to detect echinocandin nonsusceptible C. auris isolates within 6 h. [57] Molecular methods Echinocandin resistance is mediated through limited mutations S639P or S639F in FKS1, and azole resistance through F126L, Y132F, and K143R in ERG11 * [2,5] * To date, these are the only mutations associated with clinical failures due to azole and echinocandin drugs.
As regards automated systems for routine AFST, some authors observed a good (96.7%) agreement between VITEK 2 and CLSI methods in fluconazole AFST, although others have reported suboptimal performance for amphotericin B [74,75].
MALDI-TOF MS has also been increasingly used for AFST. Vella and colleagues have developed a rapid AFST assay based on MALDI-TOF MS analyzing changes in the MS profile spectra induced by antifungals after 3-6 h of incubation, firstly in caspofungin-resistant C. albicans and then in anidulafungin-resistant C. glabrata with known FKS2 mutations, where they obtained less satisfactory results [76,77]. Recently, Vatanshenassa and colleagues used the MALDI Biotyper antibiotic susceptibility test rapid assay (MBT ASTRA) for the rapid detection of caspofungin-resistant C. albicans and C. glabrata and of C. auris [78,79]. The assay showed an accuracy of 100% on both agar plate and blood culture bottles and a sensitivity and specificity of 100% and 98% for anidulafungin and of 100% and 95.5% for micafungin, respectively. A categorical agreement of 98% and 96% was calculated for the two methods. For caspofungin, a sensitivity and specificity of 100% and 73% were found, respectively, with a categorical agreement of 82%. MBT ASTRA has the greatest potential to detect C. auris isolates nonsusceptible against echinocandin antifungals within 6 h, making it a promising candidate for AFST in clinical laboratories in the future [79].
In place of traditional methods, rapid molecular methods can be used to identify resistance-conferring mutations [29]. These methods can be used for high-throughput surveillance and provide important information quicker than standard methods, especially when performed with DNA isolated directly from swabs. For example, to simplify resistant C. auris screening, a duplex ERG11 assay enabling detection of mutations at positions Y132 and K143, and a simplex FKS1 HS1 assay enabling detection of mutations at position S639 were developed to identify known mutations related to resistance to azoles and echinocandins, respectively [80]. Results can be obtained in 2 h and have proven to be 100% concordant with DNA sequencing results [68]. Moreover, allele-specific probe technologies (e.g., Xpert MTB/RIF), molecular beacons, small stem-loop-structured DNA oligonucleotides, used with real-time PCR, have been developed for the detection and evaluation of antifungal resistance and enable distinguishing wild-type (WT) and mutant amplicons based on their melting temperature (T m ). These assays are not only very reproducible but may also be updated each time a new mutation is detected [68].

Infection Control Recommendations
As previously said, C. auris has two characteristics that make it an epidemiologically important microorganism: the ability to spread rapidly and a multidrug resistance phenotype. For these reasons, its detection requires the implementation of basic infection control measures [62]. To contain transmission, once a case of C. auris is detected the ECDC recommends activating the screening of close contact patients, possibly extending contact tracing based on a case-by-case risk evaluation (e.g., type of patient and ward, level of colonization) [11]. Prompt notification to public health authorities and education of healthcare workers on the clinical impact of C. auris are also crucial. Meanwhile, point prevalence surveys should be run to identify colonized patients in hospital units where the index patient is or was present. It is also suggested to review patient records to determine any prior healthcare exposures, mainly overnight stays in healthcare facilities in the month prior to culture positivity [11].
When patients are moved to other healthcare facilities, notification of C. auris colonization/infection status is recommended [11]. For the same reason, screening of patients coming from geographical areas and healthcare facilities with a high incidence of C. auris infection/colonization at the moment of hospital admission is of utmost importance [11,80].
Infection control should include single room isolation or patient cohorting and dedicated nursing staff. Currently, there is no established decolonization protocol, so the above-described measures should be applied promptly. C. auris has proven susceptible to chlorhexidine in vitro in recent works; however, in a cohort of UK patients, despite daily chlorhexidine bathing, colonization with C. auris was not eradicated [6,81]. Further studies are needed to confirm the efficacy of chlorhexidine and other products for decolonization before they can be recommended for use. Moreover, it looks that the sessile/biofilm form of this yeast displays increased tolerance to clinically-relevant concentrations of chlorhexidine and hydrogen peroxide, with eradication achieved only using povidone-iodine [82].
However, it is known that reusable equipment may be a source of transmission of infection of C. auris. It is therefore crucial to have a cleaning protocol in place and strengthen the regular decontamination of the equipment and environment, especially high-touch areas [83].

Conclusions
The emerging pathogen C. auris has been associated with nosocomial outbreaks on five continents in the last ten years. Prevalence of infection is unpredictable due to misidentification and unreported cases. Proper microbiological identification, rigorous epidemiological surveillance, adequate treatment and prevention and containment strategies, combined with higher awareness on the side of physicians, microbiologists and healthcare workers, are indispensable to limit further spreading of this pathogen. Further research should investigate rapid and accurate laboratory identification methods and evaluate the clinical role of new therapeutic options to counteract C. auris antifungal resistance.