Candida auris: A systematic review and meta‐analysis of current updates on an emerging multidrug‐resistant pathogen

Abstract From 2009, Candida auris has emerged as a multidrug‐resistant ascomycete yeast pathogen with the capacity for easy transmission between patients and hospitals, as well as persistence on environmental surfaces. Its association with high mortalities, breakthrough and persistent candidaemia, inconsistencies in susceptibility testing results, misidentification by available commercial identification systems and treatment failure, complicates its management and detection. Within the last nine years, C. auris has been increasingly reported from far‐Eastern Asia, the Middle East, Africa, Europe, South and North America with substantial fatalities and misidentification. Herein, I provide a systematic and thorough review of this emerging pathogen. Meta‐analysis showed that at least 742 C. auris isolates have been reported in 16 countries, with most of these being from India (≥243), USA (≥232) and UK (≥103) (p‐value = .0355) within 2013–2017. Most isolates were from males (64.76%) (p‐value = .0329) and blood (67.48%) (p‐value < .0001), with substantial crude mortality (29.75%) (p‐value = .0488). Affected patients presented with other comorbidities: diabetes (≥52), sepsis (≥48), lung diseases (≥39), kidney diseases (≥32) etc. (p‐value < .0001). Resistance to fluconazole (44.29%), amphotericin B (15.46%), voriconazole (12.67%), caspofungin (3.48%) etc. were common (p‐value = .0059). Commonly used diagnostic tools included PCR (30.38%), Bruker MALDI‐TOF MS (14.00%), Vitek 2 YST ID (11.93%), AFLP (11.55%) and WGS (10.04%) (p‐value = .002). Multidrug resistance, high attributable mortality and persistence are associated with C. auris infections. Two novel drugs, SCY‐078 and VT‐1598, are currently in the pipeline. Contact precautions, strict infection control, periodic surveillance and cleaning with chlorine‐based detergents, efficient, faster and cheaper detection tools are necessary for prevention, containment and early diagnosis of C. auris infections.

canal or discharges of patients with otitis media, latter reports have shown their involvement in candidaemia/fungemia and other deepseated invasive infections with very high associated mortalities and co-morbidities (Azar, Turbett, Fishman, & Pierce, 2017;Ben-Ami et al., 2017). Unlike other yeasts, they can be transmitted within and between hospitals, patients and the environment. Furthermore, their resistance to at least one antifungal drug such as the azoles (particularly fluconazole and/or voriconazole), polyenes (amphotericin B), flucytosine, and the echinocandins (caspofungin, micafungin and anidulafungin) is well documented (European Centre for Disease Prevention and Control, 2016;Rudramurthy et al., 2017;Schelenz et al., 2016;Tsay et al., 2017). Various studies have established their persistence in clinical environments, including the air and bedding materials, and even in patients undergoing antifungal treatment (Schelenz et al., 2016;Vallabhaneni et al., 2016). As well, their virulence and pathogenicity have been investigated and found to be almost equal to or a little lesser than that of Candida albicans (Ben-Ami et al., 2017;Borman, Szekely, & Johnson, 2016;Larkin et al., 2017;Sherry et al., 2017); notably, Sherry et al. (2017) found aggregative C. auris to be more virulent than C. albicans in Galleria mellonella larvae (Sherry et al., 2017). Currently, C. auris has been reported in 16 countries on five continents: North America (Canada and USA), South America (Colombia and Venezuela), Europe (Germany, Norway, Spain, UK), Africa (South Africa), Asia (India, Israel, Japan, Kuwait, Oman, Pakistan, South Korea) (Chowdhary, Sharma, & Meis, 2017).
Early detection of C. auris infections has been shown to be beneficial as earlier initiation of appropriate antifungal therapy saved many lives Todd, 2017). However, the inability of several available commercial identification systems/ platforms to quickly diagnose C. auris remains a challenge to early therapy (European Centre for Disease Prevention and Control, 2016; Kordalewska et al., 2017). While the MALDI-TOF MS and PCR are currently aiding in this regard with their faster turnaround times, the cost and skill involved in their procurement and operation, respectively, is still a hurdle for most under-resourced mycology laboratories (Kathuria et al., 2015;Kordalewska et al., 2017;Prakash et al., 2016). There are currently no official therapeutic guidelines, dosage or Clinical Laboratory Standards Institute (CLSI)/European Committee on Antimicrobial Susceptibility Testing (EUCAST) minimum inhibitory concentration (MIC) breakpoints for C. auris infections, and studies evaluating these are few (Arendrup, Prakash, Meletiadis, Sharma, & Chowdhary, 2017;Lepak et al., 2017). The sensitivities and specificities of all the diagnostic tools, kits, and media used for detecting this new pathogen are discussed herein.
Microscopic and molecular/genomic analysis have established the presence of phenotypic and genetic/genomic differences between different C. auris strains from the same or different regions Tsay et al., 2017). These include the ability to exist as aggregates or nonaggregate cells, biofilm formation ability, clonality of outbreak strains, and genetic variations between strains from different geographical locations (Borman et al., 2016;Sherry et al., 2017). The virulence characteristics of aggregating and nonaggregating cellular morphologies have been investigated by at least two studies (Borman et al., 2016;Sherry et al., 2017). However, there is much to be done to answer several pending questions about this pathogen and these loopholes are highlighted below. There are currently two novel antifungal drugs that have 100% efficacy against C. auris: SCY-078 from

| Included articles
The literature search yielded 163 published articles in addition to reports from the CDC, PHE, and the ECDC. Further screening and exclusion reduced these to 48 articles that were used for the write-up; 38 articles were used for the statistical analysis ( Figure 1).

| PHENOTYPIC FEATURES
Microscopy has been instrumental in providing pictorial images of the shapes, color, size, and population structure ( Figure 4) of C. auris strains growing on different culture media such as Sabouraud's dextrose agar (SDA), CHROMagar, Brilliance Candida agar, GYPA culture plates, CS4 agar medium and cornmeal agar at different temperatures and incubation times (Table 1). Particularly on CHROMagar, which is the most common media used, C. auris appear as pale purple or pink smooth colonies occurring as single, paired and/or grouped ovoid, ellipsoidal to elongate budding cells (Kathuria et al., 2015;Mohsin et al., 2017;Satoh et al., 2009); on SDA, they appear as smooth white to cream-colored colonies (Prakash et al., 2016).
However, Kumar, Banerjee, Pratap, and Tilak (2015)   saw no characteristic color on CHROMagar with their C. auris strains, which could be due to the conditions used. The size [(2.0-3.0) × (2.5 × 5.0) μm] and growth rate of C. auris is comparable to Candida glabrata than to C. albicans (Borman et al., 2016), although its growth patterns are similar to C. albicans (Larkin et al., 2017).
The thermoresistance of C. auris that allows it to grow between 30 F I G U R E 2 Frequency of Candida auris isolated per country between 1996 and 2007 (a), comorbidities presented by C. auris-infected patients (b) and crude mortality rates per country (c). Total number of reported isolates, comorbities, and mortalities per study were collated per country and used to calculate the frequencies. GraphPad was used to calculate the p-values F I G U R E 3 Frequency of males and females infected with Candida auris per country (a), specimen sources (b), and antifungal resistance rates (c). Total number of reported cases per male and female patients, specimen sources and antifungal resistance per study were collated per country and used to calculate the frequencies. GraphPad was used to calculate the p-values and 42°C, albeit slowly and weakly at 42°C (Satoh et al., 2009), is a unique characteristic that is unseen in other species of Candida. This characteristic can be used in the easy identification of this pathogen from other species and has been cited as a possible reason for the high survival of this pathogen in humans and its potential to survive in avian species (Borman et al., 2016;Chatterjee et al., 2015;Chowdhary et al., 2014;Satoh et al., 2009). Evidently, this thermoresistance will also enhance persistence in the host, aiding in the dissemination of this pathogen in the environment (Piedrahita et al., 2017;Schelenz et al., 2016;Welsh et al., 2017).
In determining the species of this novel Candida pathogen, Satoh et al. (2009) determined the sugar fermentation and assimilation characteristics of C. auris, which has been confirmed by other authors (Table 1) (Satoh et al., 2009). The differences between sugar fermentation and assimilation, nitrogen sources utilization, and high salt tolerance in C. auris and other species of Candida, has further been used by Welsh et al. (2017) to formulate a highly sensitive and specific Salt Sabouraud dextrose/dulcitol/mannitol and Salt Yeast Nitrogen Base dulcitol/mannitol broths that can easily isolate C. auris from clinical and environmental specimens . Moreover, the inability of C. auris to grow on cycloheximide-containing medium (0.1%-0.01%) ( Brazil on the other hand in terms of N-acetyl glucosamine (NAG) utilization (Table 1). This difference has not been fully investigated to ascertain the underlying genetic and/or phenotypic mechanism. Further research should be undertaken to characterize the genetic basis for these differences to aid in a better typing and description of different C. auris strains in future.
The inability of C. auris to grow pseudohyphae, germ tube, chlamydoconidia, and chlamydospores on cornmeal agar has been established by several researchers (Table 1). However, Borman et al. (2016) and Sherry et al. (2017), respectively, found the formation of rudimentary and occasional pseudohyphae in C. auris, suggesting that pseudohyphae formation might be strain-specific or condition-specific (Borman et al., 2016;Sherry et al., 2017); further investigations with a larger number of strains will be necessary to comprehensively characterize these differences between strains, the underlying genetic and epigenetic mechanisms or factors and environmental conditions inducing these differences in pseudohyphae formation. The formation of hyphae, pseudohyphae, and germ tube in species such as C.  Kumar et al., 2015). Thus, the absence of germ tubes, chlamydoconidia/chlamydospores in strains that grow at 42°C, but are unable to grow on NAG-containing medium should be indicative of C. auris.
Furthermore, the higher virulence characteristics of C. auris even in the absence of pseudohyphae and germ tube formation remains a mystery yet to be unraveled. Borman et al. (2016), Ben-Ami et al. (2017), and Sherry et al. (2017) have reported of the presence of at least two cellular morphologies of C. auris: aggregating and nonaggregating cells (Ben-Ami et al., 2017;Borman et al., 2016;Sherry et al., 2017). Borman et al. (2016) showed that aggregating C. auris strains could not be separated by mechanical action using vigorous shaking/vortexing and/or chemical treatment with detergents. Thus, it is argued that the aggregating cells are not due to flocculation or encapsulation of cells in biofilms but rather, to the inability of daughter cells to separate after budding. Through G.
mellonella infection model studies, it has been established that nonaggregating cells are more virulent and pathogenic than aggregating cells and equally, highly or a little less virulent than C. albicans (Borman et al., 2016;Sherry et al., 2017). Moreover, nonaggregating C. auris cells formed a greater biofilm mass than aggregating ones and C. glabrata, and a lower biofilm mass than C. albicans (Sherry et al., 2017).
Besides the G. mellonella infection model studies (Borman et al., 2016;Sherry et al., 2017), no study has shown a higher pathogenicity for Hyphae formation is negative. Some strains form pseudohyphae occasionally, but most strains do not. No germ tube formed on cornmeal agar. Little adherence to catheter material (compared to Candida albicans). Phospholipases (P z ) and proteinases production were strain-dependent, at different degrees (0.78-1 and 0.0-5.3, respectively) and relatively lower than C. albicans (P z = 0.66)  with murine infection models are described below (Larkin et al., 2017).
Furthermore, the finding of a higher virulence/pathogenicity among nonaggregating cells than in aggregating cells has only been established in G. mellonella models. Thus, additional studies are necessary to establish the relative pathogenicity of these two cellular morphologies in different infection models.
Summing up, C. auris has a complicated phenotypic plasticity in terms of cellular morphology, nitrogen and carbon source assimilation and utilization, virulence and pathogenicity, which can be cellular morphology type-, strain-and/or country of origin-specific. Nevertheless, their ability to grow at 40-42°C has been confirmed worldwide.
Within the C. auris genome, orthologs of several C. albicans efflux genes belonging particularly to the major facilitator superfamily (MFS) and the ATP-binding cassette (ABC) transporter families were identified, suggesting that efflux is a potential resistance mechanism mediating multidrug resistance (MDR) against azoles, polyenes, and echinocandins in this pathogen (Chatterjee et al., 2015;Sharma et al., 2016). This was phenotypically confirmed by Ben-Ami et al. (2017) with a rhodamine-based efflux assay (Ben-Ami et al., 2017). Further, the zinc (II) 2 cys 6 transcription factor family, of which four members are key regulators of MDR1, an efflux pump gene whose upregulation leads to MDR, was enriched in the C. auris genome (Chatterjee et al., 2015).
Orthologous genes of C. albicans virulence proteins such as STErelated proteins, MADS-box, Ste12p, mannosyl transferases, adhesins, and integrins as well as orthologs of C. albicans kinases involved in virulence and antifungal stress response such as Hog1 protein kinase, 2-component histidine kinase etc., were discovered in the C.
auris genome. Functional annotation of most C. auris genes remain to be undertaken and this will be necessary to comprehend the genetic mechanisms of this pathogens' MDR and virulence/pathogenicity (Grahl, Demers, Crocker, & Hogan, 2017).
Thus, the C. auris genome is still not fully characterized and bears little resemblance to the genomes of other species of Candida. Several orthologous efflux and virulence genes are present in the genome, but its actual sexual cycle remains a mystery.
Although susceptible C. auris strains, specifically to FLZ, have been described (Vallabhaneni et al., 2016), most C. auris strains have been reported to be resistant to FLZ and/or to other azoles such as VRZ and to AMB, with a minority being resistant to FCN, other azoles and the echinocandins (Table 2; Figure 2). In several cases, MDR to FLZ and AMB or to all three antifungal drug classes (azoles, polyenes and echinocandins) have been reported (Table 2) Lockhart et al., 2017). The order of resistance as shown in Figure 2 namely, FLZ > AMB > echinocandins, is the same in most of the studies reported so far in most countries (Arendrup et al., 2017 Todd, 2017). The echinocandins can then be maintained or changed based on the susceptibility results (Lepak et al., 2017;Todd, 2017); it should, however, be noted that some patients have died even while on echinocandins (Azar et al., 2017;Ruiz Gaitán et al., 2017;Schelenz et al., 2016). Early initiation of echinocandin therapy has been advised to cut down on C. auris-mediated mortalities Lee et al., 2011).
NS Blood (7), sputum (2), groin swab (2) (2) Resistance to other azoles such as ITZ, ISA and PSZ has been variable ( Table 2). Arendrup et al. (2017) suggested that the variable resistance to other azoles besides FLZ might be due to a mixed population of resistant and susceptible (or wild-type and nonwild-type) C.
auris strains or the presence of different resistance mechanisms within the population being tested. In other words, the collective resistance mechanism(s) found in the various strains making up the population can affect the final MIC (Arendrup et al., 2017). Caution should be exercised in interpreting MIC data for AMB and CFG generated by Vitek 2 as substantial discrepancies (higher AMB and lower CFG MICs) has been reported between MICs generated by the CLSI's microbroth dilution (MBD) method and the Vitek 2 instrument (Kathuria et al., 2015;Khillan et al., 2014). The direct role of efflux pumps in C. auris antifungal resistance is yet to be comprehensively characterized although Ben-Ami et al.
(2017) used rhodamine, an efflux substrate, to show that C. auris expressed a higher ABC-type efflux pump activity than C. glabrata and C.
haemulonii (Ben-Ami et al., 2017). This higher efflux activity suggested that efflux pumps play an important role in C. auris MDR mechanisms, which is corroborated by the several MFS and ABC-type efflux pumps' orthologous genes identified by Chatterjee et al. (2015) in C. auris genomes (Chatterjee et al., 2015).
Furthermore, the role of mutations in ERG3 and ERG11 genes in conferring resistance to azoles in C. auris has been investigated (Chatterjee et al., 2015;Lockhart et al., 2017;Sharma et al., 2016) by aligning orthologs of these genes in C. auris to that of susceptible C. auris and/or C. albicans and calling SNPs. The presence of known resistance-conferring mutations and/or hotspots in C. albicans' ERG11 in C. auris orthologs have been inferred as a possible resistance mechanism . However, transcomplementation or functional studies of these mutated genes have not been undertaken to establish the effect of these mutations in C. auris, specifically in terms of MIC effects. In addition, no known resistance-conferring mutations in the FKS gene have been identified to date and the ERG11 mutations identified by Lockhart et al. (2017), that is, F126T in South African strains, Y132F in Venezuelan strains, and Y132F or K143R in Indian and Pakistani strains, were found to be clonally and geographically related . A comprehensive study on the resistance mechanisms of C. auris is required to decipher the MDR nature of this pathogen.
Hence, it is obvious that efflux, mutations in the ERG and FKS, and biofilm formation are potential C. auris resistance mechanisms. In addition, C. auris is generally resistant to FLZ, moderately resistant to AMB, and variably resistant to other azoles, flucytosine and echinocandins. Kumar et al. (2015) first undertook phospholipase, proteinase, and hemolysin activity assays in C. auris to evaluate their virulence in vitro.

| VIRULENCE AND PERSISTENCE
Phospholipases, proteinases, and hemolysins are important enzymes that are used by fungi to invade and infect the host Larkin et al., 2017). In that report, substantial phospholipase activity (P z value = 0.72), proteinase activity (P rz value = 0.66) and hemolysin activity (H z value = 0.74) were recorded in the single C. auris isolate; a P z , P rz or H z value of 1 represents no activity Larkin et al., 2017). The presence of several virulence genes in C. auris genomes has also been attested to (Chatterjee et al., 2015;Sharma et al., 2016). As already noted, no germ tubes were formed by the isolate on corn meal agar. These findings were recently followed up by Larkin et al. (2017) with a larger number (n = 16) of isolates in which they observed that not all C. auris strains expressed phospholipases and proteinases, and none produced germ tubes (germinated) upon incubation with fetal bovine serum. Moreover, even among strains that did express the virulence proteins, the degree of activity was not the same but strain-specific, showing that not all C. auris strains are virulent/pathogenic or equally virulent/pathogenic. As well, two representative strains showed relatively poorer adherence to catheters than C. albicans, suggesting that adherence to catheters could not be a major means/cause of invasive C. auris infections and persistence in patients and hospitals (Larkin et al., 2017). However, there are reports on the clearance of C. auris candidaemia upon removal of urinary or central venous catheters from patients Lee et al., 2011;Ruiz Gaitán et al., 2017).
It is currently agreed that C. auris forms relatively less biofilms in terms of biomass and metabolic activity than C. albicans, with nonaggregating C. auris strains forming more biofilm mass than aggregating ones (Larkin et al., 2017;Sherry et al., 2017). Whereas C. auris biofilms have been shown to be resistant to FLZ, VRZ, echinocandins, and AMB (Sherry et al., 2017), the biofilms were found to be composed of very limited extracellular matrix, relatively thin and composed mainly of yeast cells (Larkin et al., 2017). Notably, orthologous biofilm-  (Borman et al., 2016). This was seconded by Sherry et al. (2017) that nonaggregating cells can be more virulent and pathogenic than C. albicans (Sherry et al., 2017).
Further, Borman et al. (2016) showed that hyphae and pseudohyphae formation are important virulent factors in Candida in that nonhyphae and nonpseudohyphae-forming species such as C. glabrata, Candida kefyr, C. krusei, and Saccharomyces cerevisiae were less virulent and pathogenic than hyphae-forming ones such as C. albicans and C.
tropicalis and the rudimentary pseudohyphae-forming pathogen, C.
auris (Borman et al., 2016). This was evident from the survival times recorded in G. mellonella infection models. Larkin et al. (2017) however, contend that the use of murine infection models indicates that C. auris is far less virulent than C. albicans and that the MDR nature of C. auris is a fitness cost for its reduced virulence compared to C. albicans. Moreover, they asserted that C. auris could not effectively infect and disseminate in mice unless they were immunocompromised, and a larger C. auris inoculum size (3 × 10 7 yeast cells/animal) was administered (Larkin et al., 2017). In contrast, a higher virulence and patho- Enterobacteriaceae (CRE) (Piedrahita et al., 2017). However, C. auris was recovered at a higher rate than C. albicans, but significantly less than Candida parapsilosis (Piedrahita et al., 2017). Further, Welsh et al.
(2017) also evaluated the persistence of C. auris vis-à-vis C. parapsilosis on plastic surfaces and found that C. auris can persist for at least 2 weeks on culture and 1 month when their esterase activity (viability) is measured with a solid-phase cytometer .
The higher sensitivity of the esterase activity test, which can identify single cells, makes it ideal for testing the sterility of sterile products and determining the presence of C. auris in hospital environments; it should thus be used alongside culture-based surveillance.
This is because C. auris failed to grow on culture after 2 weeks on plastic surfaces while the esterase activity test continually remained positive for an additional 2 weeks. Furthermore, while the cultured C. auris isolates from plastic surfaces grew for 2 weeks, C. parapsilosis grew for 1 month; however, the esterase activity test showed that C.
auris persisted for a least a month and was more viable than C. parapsilosis . Persistence times between resistant and susceptible C. auris strains need further investigation. And the potential of culture-negative but esterase activity-positive (viable) strains to cause infection and hospital spread should be interrogated.
Candida auris can colonize, persist and recur in patients several months after first detection, allowing it to be distributed or spread to other patients and in hospitals; even more worrying is the persistent presence of a susceptible C. auris strain in the urine of a patient on FLZ treatment (Vallabhaneni et al., 2016). It is estimated that ≥4 hr is the minimum contact period for acquisition of C. auris from an infected person or surface (Schelenz et al., 2016). Moreover, C. auris can colonize and be shed from the skin at a rate of approximately 10 6 cells/hr, leading to prolonged outbreaks and transmissions in hospitals (Schelenz et al., 2016;Welsh et al., 2017 auris infections (Schelenz et al., 2016). Candida auris has been isolated from the axilla and groins of patients and swabbing of these regions are recommended for C. auris surveillance (Vallabhaneni et al., 2016;Welsh et al., 2017). In all, these show the ability of C. auris to inhabit and persist in various niches, and corroborates the need to periodically surveil and disinfect healthcare settings previously infected with C. auris.
In conclusion, C. auris persists in a viable form on dried or moist surfaces for several weeks longer than C. albicans and C. parapsilosis.
It forms lesser biofilm mass than C. albicans, has poorer adherence to catheters, produces no germ tubes and has strain-specific expression of hemolysins, proteinases and phospholipases virulence factors. Out of 316 patients, 94 were recorded as demised, which translated into 29.75% crude mortality rate (p-value = .0488). Crude mortality per country showed that C. auris infections resulted in 33.33% to 100% crude mortality worldwide, with the least (33%) being recorded in South Africa and Israel; p-value = .1789 ( Figure 2c). Table 2  As geriatrics are more prone to be hospitalized in acute-care hospitals or long-term care facilities, it is more likely that they will be exposed to auris as a lethal pathogen and nosocomial threat. The true prevalence of C. auris-mediated candidaemia could be higher if they are rightly detected.

As shown in
In summary, C. auris has been isolated from both sexes in 16 countries and five continents worldwide, with risk factors ranging from the presence of invasive devices to immunocompromised conditions. This is a serious observation as the Vitek 2 is a commonly used instrument for measuring the MICs of various antifungals against C. auris (Tables 3-4). Although Shin et al. (2012) have argued that the Vitek 2 was better than the CLSI and EUCAST MBD protocols in detecting AMB resistance, particularly as the latter two methods yield very narrow AMB MICs that are unable to efficiently discriminate between AMB susceptible and resistant isolates, the Vitek 2 should not be used alone to report on the susceptibility of C. auris strains (Shin et al., 2012). This is particularly important as wrong susceptibility results can result in fatal consequences Kathuria et al., 2015;Kumar et al., 2015;Ruiz Gaitán et al., 2017;Vallabhaneni et al., 2016). For now the gold standard for C. auris MICs is the CLSI MBD protocol, which is the most widely used (Arendrup et al., 2017). Welsh et al. (2017) recently reported of two novel in-house diagnostic broths they designed to efficiently screen for and detect C. auris from clinical and environmental specimens with relative ease, 100% specificity and sensitivity, and low cost. These broths, consisting of 10% salt, gentamicin, chloramphenicol and either dulcitol, mannitol or dextrose in Sabouraud broth or Yeast Nitrogen base (YNB), could inhibit the growth of all other species when cultivated at 42°C. However, when the Sabouraud broth with dextrose was used and cultured at a lower temperature, C. glabrata could also grow as it has high salinity tolerance. This easy-to-prepare and cheaper broth has been useful in controlling the spread of C. auris in the US and other countries . Thus, its adoption in other laboratories will facilitate the easy and quicker detection of this problematic pathogen.

Kumar et al. (2017) combined two culture media, CHROMagar
Candida media supplemented with Pal's medium, to perfectly distinguish between C. auris and C. haemulonii. Pal's medium was originally designed for the identification of Cryptococcus neoformans and has been useful in distinguishing C. albicans from Candida dubliniensis. On this merged medium, C. auris produced no pseudohyphae, grew at 42°C, and had confluent growth of white-cream colored smooth colonies while C. haemulonii did not grow at 42°C, had pseudohyphae, and showed poor growth of smooth light-pink colonies. This method, while very sensitive and specific (100%) if used for only these two species, is limited by the fact that an initial identification by available commercial systems to rule out other nonalbicans species of Candida is required (Kumar et al., 2017).
In an earlier work, Shin et al. (2012) (Shin et al., 2012). Further tests will be necessary to confirm this preliminary finding.

| Diagnostic tools: MALDI-TOF MS
The inefficiencies of available diagnostic tools in detecting or misidentifying C. auris are already mentioned above (Tables 3-4). Using an updated research use only (RUO) library or database, which can be updated in-house, the two available MALDI-TOF MS platforms, the commonly used Bruker Biotyper™ and the lesser used Vitek MS, can detect C. auris with 100% sensitivity and specificity within a few minutes (Table 3). The Bruker Biotyper™ database 3.1 has spectra of three C. auris strains (Kathuria et al., 2015). Grenfell et al. (2016) showed that adding ClinProTools to the Flex Analysis provided higher discriminatory power in detecting biomarker peaks (Grenfell et al., 2016).
Several researchers have also reported of the higher efficiency of the Bruker Biotyper over the Vitek 2 MS in detecting C. auris and other nonalbicans species of Candida, even with an updated database (Ghosh et al., 2015;Grenfell et al., 2016;Kim, Kweon, Kim, & Lee, 2016). The In terms of specimen preparation protocols, Ghosh et al. (2015) showed that the on-plate formic acid extraction method is the most cost and time efficient (Ghosh et al., 2015). Mizusawa et al. (2017) also observed that the direct extraction method enabled the perfect detection of C. auris on the Vitek MS system while the full-length or partial extraction method was necessary for 100% identification by the Bruker MS system (Mizusawa et al., 2017). Using the direct on-plate extraction method resulted in only 50% identification of C. auris with low score match, with 50% being unidentified (Mizusawa et al., 2017). Girard et al. (2016) also used the direct smear protocol to identify C.

| Diagnostic tools: PCR, real-time PCR and whole genome sequencing (WGS)
The use of conventional PCR to amplify the ITS and/or D1/D2 DNA sequences, followed by sequencing of the amplicons is currently the gold-standard and most commonly used technique to identify, confirm the identity and type C. auris strains (Tables 3-4) with 100% specificity and sensitivity, and shorter turnaround time (Kordalewska et al., 2017). Recently, Kordalewska et al. (2017) developed a conventional PCR and real-time assay that could respectively identify C. auris as well as C. auris, C. duobushaemulonii and C. lusitaniae with 100% sensitivity and specificity, and shorter turnaround time of 2.5 and 2 hr respectively. This protocol was also used in direct colony PCR to achieve the same optimum results. Either gel electrophoresis (for conventional PCR) or melting temperature (Tm) analysis (real-time PCR) was used for final confirmation or differentiation of the results respectively. The amplicons covered a fragment of 5.8S, ITS2 and a part of 28S ribosomal DNA using CauF/R primers, which yielded a 163 bp long (conventional) PCR amplicon for C. auris. Further, CauRe1R primers (real-time PCR) selectively amplified regions in either C. auris, C. duobushaemulonii, C. haemulonii or C. lusitaniae. The limit of detection of these assays were 10 CFU/reaction (Ct = 28.61 ± 0.25) for C. auris-specific assays and 1,000 CFU/reaction (Ct = 27.83 ± 0.87) for C. auris-related species (Kordalewska et al., 2017).
Besides using the sequenced amplicons to identify an isolate as C.
auris by comparing the sequence to available sequences at GenBank, they can also be used in phylogenetic analysis to draw evolutionary or phylogenetic trees. These phylogenetic dendrograms has been instrumental in tracing the sources and clonality of the isolates in relation to other isolates from the same or different hospital, region, or country.
Other PCR-based typing tools such as AFLP and MLST have been used to identify and type C. auris strains (Tables 2-4). As well, other molecular but non-PCR-based or restriction enzyme-based techniques such as PFGE and REAG-N have been used occasionally to aid in the typing of C. auris (Oh et al., 2011). However, these above-mentioned (PCRbased and non-PCR-based) typing tools are labor intensive with longer turnaround times.
WGS is increasingly being used to aid in the simultaneous identification and typing or evolutionary analysis of C. auris cases (Chatterjee et al., 2015;Lockhart et al., 2017;Sharma et al., 2016;Tsay et al., 2017;Vallabhaneni et al., 2016). Due to its higher resolution, it can provide a better evolutionary and epidemiological analysis of C. auris cases than all other methods within a comparatively short turnaround time (8-72 hr), except that it is more expensive and requires higher skill and data processing capacity (Table 3-4) . Arendrup et al. (2017)  T A B L E 4 Usage characteristics and analyzed sample sizes per diagnostic tool finding. Thus, although the MIC differences from both protocols are relatively minor, researchers should be mindful of the specifics when comparing MICs obtained from both methods (Arendrup et al., 2017).

| CLSI and EUCAST MIC protocols
Notwithstanding these little differences in MIC values, it is expected that the CLSI protocol will continue to hold preeminence among researchers because of the large number of available MIC data generated from this protocol, which will facilitate easy comparison with already available data from other works.
The authors observed that the collective MIC values from any population will be influenced by the presence of wild-type and nonwildtype colonies as well as by the collective resistance mechanisms of the various strains (Arendrup et al., 2017). This might explain the variable resistance of C. auris to the other azoles besides FLZ. A low acquired resistance to AMB and echinocandins was recorded in the C. auris strains.
Summing it up, PCR and MALDI-TOF are the commonly used diagnostic tools and the CLSI MBD remains the most commonly used protocol for MIC determination.

| MOLECULAR EPIDEMIOLOGY
While the above-mentioned methods have enabled the easy description of the molecular epidemiology and phylogenetic relationship between strains of the same or different hospitals and/or countries, their resolution power is relatively weaker than that of WGS, which has recently been used by Lockhart et al. (2017)  Due to misidentification of C. auris by most commercial identification systems and the nonspecies identification of many species of Candida in many mycology laboratories, the true prevalence and epidemiology of C. auris infections in most countries and the world is not known and is likely to be underestimated than overestimated (Kordalewska et al., 2017;Todd, 2017). Moreover, blood, fluid and tissue cultures for detecting C. auris grow slowly and they could be falsely negative in cases of low-level or intermittent candidaemia (Todd, 2017). The molecular epidemiology of all reported C. auris cases are described below under their continents and countries according to the order of detection.

| Far East Asia: Japan and South Korea
The earliest C. auris case was misidentified and undetected as far back as 1996 in South Korea , prior to the first reported case of C. auris by Satoh et al. (2009), which was isolated from a 70- year old female Japanese patient. Satoh et al. (2009) were thus the first to describe and name the new pathogen as C. auris due to its closer phylogenetic, phenotypic and genotypic (Table 1) relationship to the Candida genus and its isolation from the ear. Using the D1/ D2 and ITS sequences, they showed that this new pathogen phylogenetically clustered in the Metschnikowiaceae clade (Satoh et al., 2009).
Thus far, this first work by Satoh et al. (2009) is the only reported case of C. auris in Japan to date. Later in the same year, Kim et al. (2009) also reported of a novel yeast species with close phenotypic similarity to C. haemulonii from the ear of 15 otitis media patients who visited five hospitals in South Korea between 2004 and 2006. These historical isolates, some of which were later found to be clonally related (Oh et al., 2011), were actually C. auris, with elevated FLZ, VRZ and AMB MICs or resistance.
Thus, it is obvious that C. auris first appeared in South Korea as early as 1996 (Table 2), but was misidentified and undescribed until Satoh et al. (2009) did so in Japan. Moreover, the Japanese isolate was later found to be very closely related phylogenetically to some of the isolates from South Korea (Ben-Ami et al., 2017;Mohsin et al., 2017;Schelenz et al., 2016) and they all assimilated NAG while those from other countries did not (Table 1) (Prakash et al., 2016); thus, the possibility of transfer from South Korea to Japan or otherwise, should be investigated further. It is notable that almost all the isolates recovered from Japan and South Korea were from the ear (Table 2), except a few (n = 6) that were obtained from blood; at least two patients with candidaemia demised Shin et al., 2012). Fortunately, no C. auris cases, either from the ear or blood (fungemia), have been reported in South Korea since 2013. Chowdhary et al. (2013) were the first to report on a clonal out-  (Table 2), including MDR isolates Chowdhary et al., 2014;Prakash et al., 2016). There is a higher prevalence of C. auris infections in the public sector than private sector hospitals in India due to overcrowding and possible compromise in infection control (Rudramurthy et al., 2017), with C.

| South Asia: India and Pakistan
auris prevalence ranging from 5% to 30% of all candidaemia cases in certain hospitals (Chowdhary et al., 2013;Rudramurthy et al., 2017). WGS, AFLP, MLST and MALDI-TOF MS typing of several Indian strains showed their closer evolutionary or phylogenetic relationship and wider evolutionary or phylogenetic distance from those of other countries Prakash et al., 2016). Lockhart et al. (2017) showed that the genomes of isolates from India differed from that of other countries by >10,000 SNPs, indicating the independent emergence of C. auris in this country . However, strains from the Pakistan, USA and UK have very close phylogenetic relationship with those from India, suggesting that they were possibly imported from India (Borman, Szekely, & Johnson, 2017;Vallabhaneni et al., 2016). Further, the first C. auris case in Canada was in a patient previously hospitalized in India (Schwartz & Hammond, 2017).
The higher prevalence of C. auris cases ( Figure 2) and clonal outbreaks in India is very concerning, particularly as many were multidrug resistant and can spread to other countries as already reported Schwartz & Hammond, 2017;Vallabhaneni et al., 2016). In one study, there was interhospital and intrahospital spread of clonal C. auris strains, even though there were no exchange of healthcare personnel between these hospitals and wards (Chowdhary et al., 2013). Resistance to FLZ has been found to be mediated by known mutations (Y132F and K143R) in ERG11 . Recommended infection control protocols should be instituted and strictly followed to reduce the incidence of this MDR  and Control, 2016;Schelenz et al., 2016;Todd, 2017).
The incidence of C. auris in Pakistan (n = 19) was first reported by Lockhart et al. (2017) and characterized using WGS ( Table 2). The Pakistani isolates were collected between 2012 and 2015 and were found to be very closely related to those from India, with <60 SNPs between isolates. Notably, they resulted in very high crude mortalities (72%; 13/18) and also shared the same FLZ resistance mechanism (Y132F and K143R in ERG11) as that of the Indian strains . There are no reports of C. auris infections in Pakistan besides this, but the higher mortality rate is worrying. Further surveillance and prompt report of C. auris cases are necessary to detect cases as early as possible.

| Middle East: Israel, Kuwait and Oman
Only a single report of C. auris candidaemia in six patients from two hospitals in Tel-Aviv has been published to date in Israel (Ben-Ami et al., 2017). These strains were collected between May 2014 and May 2015 and were phylogenetically different from those from East Asia, Africa, and the Middle East. They formed aggregates in the kidneys of mice infection models and were less virulent than C.
albicans, but more virulent than C. haemulonii. The formation of aggregates by these Israeli strains is akin to that reported by Borman et al. (2016) and Sherry et al. (2017), and corroborates the assertion that some C. auris cells form aggregates, which are less virulent/pathogenic than nonaggregating ones and C. albicans (Ben-Ami et al., 2017;Borman et al., 2016;Sherry et al., 2017). In addition, these strains were found to have higher ABC efflux activity than C.
glabrata and C. haemulonii, which agrees with the enriched efflux genes reported by Chatterjee et al. (2015) in the C. auris genome and explains the MDR nature of this pathogenic yeast (Chatterjee et al., 2015). The phylogeny of these strains indicates that they emerged independently in Israel and were not imported as they did not have a close relationship with other isolates from the Middle East, South or East Asia (Ben-Ami et al., 2017).
The first and only C. auris candidaemia case in Kuwait was reported by Emara et al. (2015). This case was in a 27-year old woman with chronic renal failure who was admitted to the ICU in May 2014.
Two different research groups simultaneously reported of separate C. auris incidence at different hospitals in Oman in the same year (Al-Siyabi et al., 2017;Mohsin et al., 2017), one of which described two clonal strains from two old (70 and 77 years) patients from the same hospital; one patient died (50% mortality) (Mohsin et al., 2017).
Al-Siyabi reported of five C. auris candidaemia cases involving mostly old patients in another hospital, of which three died (Al-Siyabi et al., 2017). All the C. auris candidaemia cases were detected between August 2015 and February 2017, and the isolates expressed high resistance to FLZ; some patients died even though they were on ANF therapy. The onset of infection after hospitalization ranged from 22 to 62 days, showing that these candidaemia cases were nosocomially acquired. The phylogenetic relationship between the isolates from these two reports has not been undertaken, albeit this is necessary to show if the isolates from the two reports are clonally related, and if these cases were locally acquired or imported. The two clonally related isolates (Mohsin et al., 2017) however seem to have been locally acquired as the patients had never traveled outside Oman; notwithstanding, they clustered phylogenetically between isolates from India and UK.
Further investigations might be necessary to show whether they had contacts with persons from some of these countries. However, no further reports of C. auris have been published from Oman.

| Africa: South Africa
Candida auris candidaemia was detected in four male South African patients between October 2012 and October 2013, with high resistance to FLZ (Magobo, Corcoran, Seetharam, & Govender, 2014). Except for one patient aged 27, the ages of the patients were between 60 and 85 years. Lockhart et al. (2017) subsequently reported of an additional 10 isolates collected from South Africa between 2012 and 2014, which were closely related to each other with <70 SNPs, but very distant phylogenetically to those from Pakistan, India and Venezuela . Borman et al. (2017) showed that isolates from the UK had very close sequence similarity with those from South Africa, the first and only African country to report of a C. auris mediated candidaemia . Prakash et al. (2016) also reported that the South African strains clustered with other isolates of diverse geographical origin (Prakash et al., 2016).

| Europe: Germany, Norway, Spain and UK
Reports of C. auris fungemia and colonization have been rare in continental Europe, with most C. auris cases being reported in the UK, which was the first country in Europe to report of C. auris incidence as well as a clonal outbreak involving 50 patients in a cardiothoracic center in London (Borman et al., 2016Schelenz et al., 2016).
The first reported case (candidaemia) of C. auris in the UK was in 2013 from three unrelated patients in distant geographical localities (Borman et al., 2016

| South America: Colombia and Venezuela
Three invasive C. auris reports, one from Colombia (n = 17) (Morales-Lopez et al., 2017) and another two involving an outbreak case (n = 18) and additional cases (n = 5) from Venezuela (Calvo et al., 2016;Lockhart et al., 2017), have been published from South America. In the report from Colombia, most patients had a CVC (n = 16), a urinary catheter (n = 15) and a mechanical ventila- Interestingly, only two C. auris cases had been reported by 2015 in the US, but this number shot up afterwards, suggesting a recent and rapid emergence or higher detection of this menace possibly due to increased awareness and education on detection methods. The minimum time from hospital admission to first isolation of C. auris was 18 days in the first seven cases; and five out of the seven patients died. In one case, a C. auris candidaemia that was susceptible to FLZ persisted in the patient even though the patient was on the same drug.
In two cases, the C. auris candidaemia recurred for 3-4 months while some patients remained colonized months after first detection (Tsay et al., 2017). This shows that C. auris can easily spread through hospitals and patients from colonized or infected persons.
Only a single C. auris case has been detected in Canada in a patient who was initially admitted in a hospital in India (Schwartz & Hammond, 2017). The isolate was continually obtained from repeated swabbing of the same ear of the patient over a 6-week period. No other C. auris isolate has been reported in Canada afterwards.

| MANAGEMENT AND CLINICAL OUTCOMES (MORTALITIES)
An official management protocol for C. auris infections is yet to be concluded, an optimal antifungal agent(s) or dosing regimen for C.
auris infections is not defined and CLSI/EUCAST breakpoints for this pathogen is wanting (Lepak et al., 2017), making researchers resort to that established for closely related species of Candida (Arendrup et al., 2017;Lockhart et al., 2017). Interestingly, the efficacy of this approach was recently established by Lepak et al. (2017) for FLZ and AMB using neutropenic disseminated candidiasis murine models infected with C. auris (Lepak et al., 2017). Pharmacokinetic and pharmacodynamic (PK/PD) data showed that the MICs breakpoints of these two antifungals for other species of Candida, will hold for C.
auris as the concentrations (exposure target) associated with optimal outcomes were similar. Furthermore, there was a strong relationship between the PK/PD parameters and treatment outcome for each drug (including MCF) and the dose-effect against C. auris for each drug was proportional to the MIC (Lepak et al., 2017). A tentative breakpoint for some selected antifungals has however been proposed by the CDC and was used in analyzing some of the data in this review (Table 2; Figure 2) (Centers for Disease Control and Prevention, 2017b; Schwartz & Hammond, 2017). Clinicians currently advise on the use of echinocandins empirical therapy when C. auris is expected as it is the current antifungal with the most effective activity against this pathogen. However, this can be changed when subsequent susceptibility results show other antifungals to be potent against the patient's isolate (Lepak et al., 2017;Todd, 2017).
Liposomal AMB has also been shown to be very effective against C. auris including the inhibition of biofilm formation and potency against C. auris biofilms. It is also used in combination therapies with an echinocandin (Azar et al., 2017;Emara et al., 2015;Ruiz Gaitán et al., 2017;Sherry et al., 2017;Vallabhaneni et al., 2016).  . Particularly, SCY-078 is not affected by common mutations in protein targets, is orally bioavailable and active against echinocandinresistant strains . It is not advisable however, to offer antifungal therapy to colonized patients (Todd, 2017).

| INFECTION CONTROL AND PREVENTION
The persistence and recurrence of C. auris in the hospital environment in the face of rigorous decontamination, disinfection and decolonization protocols, as occurred in the UK over a one-year period, should be a wake-up call to all infection control personnel in all healthcare centers (Schelenz et al., 2016). Due to issues with misidentification, it is necessary that all microbiology (mycology) laboratories update their commercial identification softwares to enable them to easily and efficiently identify C. auris cases. Specifically, C. famata, C. haeumolonii, C. sake, C. krusei, R. glutinis, etc. strains should be further analyzed with PCR or MALDI-TOF to confirm they are not misidentified. Or where such systems are unavailable, to quickly transport such specimens to local or foreign reference laboratories, isolate the patient under contact precautions and start empirical echinocandin therapy while awaiting the outcome of the laboratory tests (Azar et al., 2017; Centers for Disease Control and Prevention, 2017a, b; Lepak et al., 2017;Todd, 2017). In cases of organ transplants, donor organs should be scrutinized for sterility from C. auris prior to transplantation (Azar et al., 2017). Antifungal stewardship is necessary and prophylactic an-  (Schelenz et al., 2016). Patients or healthcare workers coming in close contact with infected persons should also be placed under strict contact precautions until they consistently provide negative cultures over 3 weeks (Schelenz et al., 2016). The wards of patients found to be colonized or infected with C. auris should be thoroughly disinfected as described (Schelenz et al., 2016).
MICs should ideally be measured using the CLSI MBD protocol to ensure accurate susceptibility results, which can inform correct therapeutic choices (Khillan et al., 2014;Lepak et al., 2017). Clinicians should also consider removing CVCs and other catheters where possible as certain studies have found such options useful in resolving persistent candidaemia (Calvo et al., 2016;Lee et al., 2011;Ruiz Gaitán et al., 2017).
Besides culture-based methods in surveillance studies, esterase activity as measured by a solid-phase cytometer should be considered to enhance the detection of viable but nonculturable strains . Hospital wards, bedding materials, beds, invasive and noninvasive medical devices, clothing of patients, skin and surface wounds etc. should be decontaminated, using chlorine-based detergents such as chlorhexidine (0.2%-4%) and hydrogen peroxide vapor (Schelenz et al., 2016;Sherry et al., 2017). As well, chlorhexidine-impregnated protective discs for all CVC exit sites can aid reduce line-associated C. auris BSIs (Schelenz et al., 2016). Oral nystatin plus nasal ointments have proved effective in decolonizing healthy nasal carriers (Schelenz et al., 2016). Soap and handwashing followed by alcohol-based hand sanitizer is recommended by PHE.

| CONCLUSIONS, FUTURE PERSPECTIVES AND STUDY LIMITATIONS
It is evident from this review that C. auris infections are more commonly reported in India, the USA and the UK, with fewer or isolated cases in South America, Africa, and continental Europe. Phylogenetic data show the independent emergence of C. auris in several countries. Misidentification, intrahospital transmission, poor treatment outcomes and higher crude mortalities between 33.33% and 100% are associated with C. auris infections worldwide. An official treatment guideline for C. auris infections is lacking and empirical treatment involving an echinocandin is advised. Contact precautions and effective disinfection with chlorine-based agents are advised for hospitals with C. auris cases. PCR and MALDI-TOF MS remain the most efficient and commonly used diagnostic tools. Todd (2017) has suggested that the variations in C. aurisassociated mortalities could emulate those of other emerging infections in which initial cases are most severe, but tend to drop in severity over time (Todd, 2017). It would be welcoming should this be the case with C. auris. Efficient identification tools such as the MALDI-TOF MS, PCR and WGS are still beyond the reach of many mycology laboratories worldwide, defeating efficient and prompt detection, earlier initiation of therapy and effective surveillance of C. auris in hospitals. Without efficient detection, the true prevalence of this menace will never be known, effective management of potential cases will be elusive and mortalities will continually rise.
Thus, designing a simple, low-cost detection technique, kit or tool with a shorter turnaround time is the key to defeating this deadly pathogen. Furthermore, the possibility of this yeast also spreading into the community, farms and the general environment should not be lost sight of. Evidence from CRE and MCR-1-positive bacteria should advise mycologists of the potential of this yeast to also infect livestock.
This review was limited by the fact that several studies failed to detail the year, sex, age(s), mortality, antifungal MICs, total number of isolates and patients, comorbidities, and specimen of the reported C. auris infections. This made the meta-analysis challenging as such studies had to be excluded.

AVAILABILITY OF MATERIALS AND DATA
Supplementary data is included in this manuscript.

AUTHOR CONTRIBUTIONS
JOS designed and undertook the meta-analysis, systematically reviewed the literature and wrote the paper.

ETHICAL APPROVAL
Not applicable.

TRANSPARENCY DECLARATION
The author declares no conflict of interest in the publication of this manuscript.