Candidiasis, which includes both superficial infections and invasive disease, is the
most common cause of fungal infection worldwide. Candida bloodstream infections (BSI)
cause significant mortality and elicit a major threat to intensive care unit (ICU)
patients [1]. The annual global burden of Candida spp. BSIs is about 400,000 cases,
with most cases reported from the developed world. Although Candida albicans remains
the most frequently isolated Candida species in the clinical setting, in some countries,
a marked shift towards species of Candida that have increased resistance to azoles
such as fluconazole (FLU), the standard antifungal drug of choice in many countries,
and to the recently introduced antifungals known as echinocandins, is reported. Several
species of non-albicans Candida, such as C. tropicalis, C. glabrata, and C. parapsilosis,
are well-recognized pathogens in BSIs in different geographic locations. More recently,
Candida auris, a multidrug-resistant (MDR) yeast that exhibits resistance to FLU and
markedly variable susceptibility to other azoles, amphotericin B (AMB), and echinocandins,
has globally emerged as a nosocomial pathogen (Fig 1) [2–20]. Alarmingly, in a span
of only 7 years, this yeast, which is difficult to treat and displays clonal inter-
and intra-hospital transmission, has become widespread across several countries, causing
a broad range of healthcare-associated invasive infections [4, 5, 10, 12, 21, 22].
10.1371/journal.ppat.1006290.g001
Fig 1
A global map depicting rapid emergence of multidrug-resistant clinical Candida auris
strains in 5 continents.
The value in parentheses denotes the year of report of C. auris from the respective
country or state.
Why is C. auris often misidentified in the routine microbiology laboratory?
In 2009, a novel Candida species, C. auris, in the C. haemulonii complex (Metchnikowiaceae),
was first described from a patient in Japan after its isolation from the external
ear canal [23]. The species exhibits a close phylogenetic relationship to C. haemulonii
and is differentiated based on sequence analysis of the D1/D2 domain of the large
ribosomal subunit (LSU) of 26S rRNA gene and the internal transcribed spacer (ITS)
regions of the nuclear rRNA gene operon [23]. The first 3 cases of nosocomial fungemia
due to C. auris reported in 2011 from South Korea highlighted the fact that this yeast
is commonly misidentified as C. haemulonii and Rhodotorula glutinis by the commercial
identification systems VITEK (BioMérieux, Marcy l’Etoile, France) and API-20C AUX
(BioMérieux), respectively [3]. These systems involve precast panels of assimilation/growth
tests using sets of carbon and nitrogen compounds and are still widely used for routine
identification of yeasts. A comprehensive study from India investigated C. auris prevalence
among 102 clinical isolates previously identified as C. haemulonii or C. famata with
the VITEK system and found that 88.2% of the isolates were C. auris, as confirmed
by ITS sequencing [9]. It is evident from several studies published recently that
C. auris in routine microbiology laboratories remains an unnoticed pathogen, as 90%
of the isolates characterized by commercial biochemical identification systems are
misidentified primarily because of a lack of the yeast in their databases [3–9, 12,
16–19, 24, 25]. Different biochemical systems are used in microbiology laboratories,
and the majority of them listed in Table 1 misidentify C. auris. A recent study on
validating the identification of C. auris with 4 biochemical identification platforms
found that all C. auris isolates were misidentified as R. glutinis by API-20C AUX,
as C. haemulonii (except 1, as C. catenulata) by Phoenix (BD-Diagnostics, Sparks,
MD), as C. haemulonii by VITEK, and as C. famata, C. lusitaniae, C. guilliermondii,
or C. parapsilosis by MicroScan (Beckman Coulter, Pasadena, CA) [25] (Table 1). However,
Matrix-assisted laser desorption ionization–time of flight mass spectrometry (MALDI-TOF
MS) is considered a more rapid and robust diagnostic technique for C. auris identification
[9, 10, 13, 16]. Currently, the MALDI-TOF MS approach is commercialized by mainly
2 manufacturers, namely MALDI Biotyper (Bruker-Daltonics, Bremen, Germany) and VITEK
MS (BioMérieux). The MALDI Biotyper (Bruker-Daltonics) has a database library that
contains spectra of 3 strains of C. auris: 2 from Korea and 1 from Japan. Although
both the Bruker-Biotyper and VITEK-MS MALDI-TOF systems lack C. auris entries in the
FDA-approved libraries, the research-use-only libraries contain the C. auris database
in both MALDI-TOF MS systems [25]. Due to the fact that this yeast is MDR, it is important
to identify these species correctly in order to provide optimal patient care.
10.1371/journal.ppat.1006290.t001
Table 1
Worldwide reports of Candida auris along with their misidentification using commercial
systems and frequency of antifungal resistance.
Country
Number of Candida auris isolates
Sample (number)
Biochemical misidentification (System)
Molecular/MALDI-TOF MS identification
Number of isolates
Year of publication [References]
FLU(≥32 μg/ml)
ITC(≥1 μg/ml)
VRC(≥2 μg/ml)
Echinocandins(≥8 μg/ml)
AMB(>1 μg/ml)
Japan
1
Ear discharge
ND
ITS, D1D2
ND
ND
ND
ND
ND
2009 [23]
Korea
15
Ear discharge
ND
ITS, D1D2
8
8
2
none
9
2009 [2]
South Korea
6
Blood
C. haemulonii (VITEK), Rhodotorula glutinis (API20C-AUX)
ITS, D1D2
4
2
none
none
3
2011 [3]
India
12
Blood
C. haemulonii, C. famata (VITEK); C. sake (API20C-AUX)
ITS, D1D2
10
none
none
none
none
2013 [4]
15
Blood (7), CVC tip (3), Excised tissue (3), BAL (1), pus (1)
C. haemulonii (VITEK)
ITS
15
none
7
none
none
2014 [5]
4
Blood (1), urine (1), pericardial fluid (1), BAL (1)
C. haemulonii (VITEK); C. sake (API20C-AUX)
ITS, D1D2
4
none
none
none
none
2014 [6]
102
Blood (78), tissue (4), pleural fluid (6), peritoneal fluid (7), urine (4), sputum
(3)
C. haemulonii/C. famata (VITEK)
ITS, MALDI-TOF MS
80
none
32
none
14
2015 [9]
51
Blood
Not mentioned
ITS, D1D2
49
3
9
none
10
2017 [19]
India, South Africa, Korea, Japan, Brazil
104: 90 India (I), 6 South Africa (SA), 5 Brazil (B), 2 Korea (K), 1 Japan (J)
Blood (n = 89; 78 I, 6 SA, 5 B), peritoneal and pleural fluid (5), invasive infections
(4), urine (1), sputum (2)
C. haemulonii (VITEK)
ITS, D1D2, MALDI-TOF MS
5 (SA); 5 (B); 1 (K); none (J)
None (SA); none (B); 1 (K); none (J)
1 (SA); 5 (B); 1 (K); none (J)
none (SA); none (B); none (K); none (J)
none (SA); 3 (B); none (K); none (J)
2016 [10]
a
Kuwait
1
Blood
C. haemulonii (VITEK)
ITS, D1D2
1
ND
none
none
none
2015 [8]
Israel
6
Blood (5), urine (1)
C. haemulonii (VITEK)
ITS, D1D2
6
none
none
none
6
2017 [18]
Spain
8
Blood (4), catheter tip (4)
Saccharomyces cerevisiae (AuxaColor 2); C. sake (API20C-AUX); C. lusitaniae, C. haemulonii
(VITEK)
ITS
8
none
8
none
none
2017 [17]
UK
12
Blood, sputum, CSF, pleural fluid, arterial line, pustule swab, wound swab, femoral
line
ITS, D1D2, MALDI-TOF MS
5
ND
1
none
none
2016 [11]
50
Blood (16), wound (3), urinary catheter (1), unknown site with invasive candidiasis
(2), colonization (28)
b
MALDI-TOF MS
50
ND
ND
none
Range 0.5–2 μg/ml
2016 [13]
Kenya
21
Blood
C. haemulonii (VITEK)
ITS
-
-
-
-
-
2014 [24]
South Africa
4
Blood
C. haemulonii (VITEK) and R. glutinis (API20C-AUX)
ITS, D1D2
4
none
1
none
none
2014 [7]
US
7
Blood (5), urine (1), external ear canal (1)
Whole genome sequencing
5
c
ND
ND
1
c
1
c
2016 [14]
CDC Collaborative Project [Pakistan (n = 18), India (n = 19), South Africa, (n = 10),
Venezuela (n = 5), Japan (n = 1)]
54
Blood (27), urine (10), soft tissue (5), other sites (12)
D1D2, Whole genome sequencing
50
Range 0.125–2 μg/ml
29
4
19
2017 [15]
US
10
NA
R. glutinis (API20C-AUX); C. haemulonii, C. catenulata (BD Phoenix); C. haemulonii
(VITEK); C. famata, C. lusitaniae, C. guilliermondii, or C. parapsilosis (MicroScan)
ITS, D1D2, MALDI-TOF MS
ND
ND
ND
ND
ND
2017 [25]
US, tested strains from Germany (n = 2), India (n = 11), Korea (n = 2), Japan (n =
1)
16
Blood (15), ear (1)
Unidentified (API20C-AUX)
ITS
8
5
5
d
none
12
e
, 16
d
2017 [20]
Venezuela
18
Blood
C. haemulonii (VITEK)
ITS
18
ND
18
none
Range 1–2 μg/ml
2016 [12]
Colombia
17
Blood (13) peritoneal fluid (1), CSF (1), bone (1), urine (1)
C. haemulonii (VITEK, Phoenix); C. tropicalis (MicroScan Walkaway); C. famata (API
Candida); C. albicans (MicroScanautoSCAN); C. tropicalis (MicroScan Walkaway)/ C.
famata (API Candida); C. albicans (MicroScanAutoSCAN)
MALDI-TOF MS
10
ND
4
none
11
2017 [16]
Abbreviations: -, not clear in the abstract; AMB, amphotericin B; BAL, bronchoalveolar
lavage; CDC, US Centers for Disease Control and Prevention; CSF, cerebral spinal fluid;
CVC tip, central venous catheter tip; FLU, fluconazole; ITC, itraconazole; ITS, internal
transcribed spacer; MALDI-TOF MS, Matrix- assisted laser desorption ionization–time
of flight mass spectrometry; MIC, minimum inhibitory concentration; ND, not done;
VRC, voriconazole.
aAntifungal susceptibility testing data of Indian isolates is same as reported by
Kathuria et al., 2015.
b Colonization with C. auris was defined as culture-positive skin, oropharynx, vascular
line exit site, respiratory, and urinary tract without clinical signs of Candida infection.
cMIC value not given.
dMICs read after 48 hours.
eMICs read after 24 hours.
Does genetic predisposition make C. auris virulent?
A recently published draft genome of C. auris shows that it has a genome size of approximately
12.3 Mb [26, 27]. A significant percentage of genes in C. auris are devoted to central
metabolism, a property that is common to pathogenic Candida and crucial for adaptation
to divergent environments. In addition, C. auris shares numerous virulence attributes
with C. albicans, including genes and pathways involved in cell wall modelling and
nutrient acquisition, histidine kinase-2 component systems, iron acquisition, tissue
invasion, enzyme secretion, and multidrug efflux [21, 26, 27]. However, in vitro results
in a single study that tested the production of phospholipase and secreted proteinase
in multiple isolates of C. auris from different geographical regions showed that both
secreted proteinase and phospholipase production was strain dependent. The phospholipase
activity and secreted proteinase were detected in 37.5% and 64% of the tested isolates,
respectively [20]. In general, the tested C. auris strains tended to have weak phospholipase
activity, with the majority of isolates being non-phospholipase producers [20]. Furthermore,
a significant portion of the C. auris genome encodes the ATP-binding cassette (ABC)
and major facilitator superfamily (MFS) transporter families along with drug transporters
that may explain the exceptional multidrug resistance in this pathogen [21, 27]. ABC-type
efflux activity by Rhodamine 6G transport was significantly greater among C. auris
than C. glabrata isolates, suggesting the intrinsic resistance of C. auris to azoles
[18].
Interestingly, comparison of whole genome sequencing (WGS) data shows C. auris to
be a close phylogenetic relative of C. lusitaniae, a species recognized for intrinsic
antifungal resistance [21, 27]. C. auris also demonstrates thermotolerance, growing
optimally at 37°C and maintaining viability at up to 42°C, salt tolerance, and cell
aggregation into large, difficult-to-disperse clusters, which may help some strains
to persist in the hospital environment [11, 23]. In a Galleria mellonella model, the
aggregate-forming isolates exhibit significantly less pathogenicity than their non-aggregating
counterparts [11]. Importantly, the non-aggregating isolates exhibited pathogenicity
comparable to that of C. albicans, which is the most pathogenic member of the genus
[11]. However, it is important to mention here that the observations made in this
study are yet to be correlated with clinical cases and thus, assuming the same results
in patients, need further experimentation. Furthermore, the virulence of C. auris
tested in a mouse model of hematogenous-disseminated candidiasis showed distinct yeast
cell aggregates in the kidneys of mice, with lethal C. auris infection suggesting
that aggregation might be a mode of immune evasion and persistence in tissue [18].
Another significant factor involved in C. auris virulence is its ability to differentially
adhere to polymeric surfaces, form biofilms, and resist antifungal agents that are
active against its planktonic counterparts [28]. However, a more recent study reported
that C. auris biofilms were significantly thinner, i.e., exhibited 50% thickness compared
to C. albicans biofilm [20]. Also, C. auris exhibits minimal ability to adhere to
silicone elastomer (a representative catheter material) relative to C. albicans [20].
C. auris’s weak adherence ability suggests that it is likely to play some role in
catheter-associated candidiasis but not a large one, in contrast to C. albicans and
C. parapsilosis, which are known to cause such infections [20]. Although, C. auris
expresses several virulence factors, albeit to a lesser extent than C. albicans and
in a strain-dependent manner [20].
The past and present of C. auris: Is the emergence of C. auris a menace to public
health?
In 2009, 15 isolates of C. auris were recovered from the ear canals of patients suffering
from chronic otitis media in South Korea [2]. Most of these isolates showed a reduced
susceptibility to AMB and azole antifungals. This report was followed by the first
3 cases of nosocomial fungemia caused by C. auris from South Korea [3]. The latter
study reported that the earliest isolate of C. auris was found in 1996 in the Korean
isolate collection [3]. All 3 patients had persistent fungemia for 10 to 31 days,
and 2 patients who received FLU therapy followed by AMB showed therapeutic failure
and had fatal outcome. Subsequently, 2 larger series of candidemia and deep-seated
infections from India in 2013 and 2014 clearly showed that clonal strains of MDR C.
auris had emerged in 3 hospitals [4, 5]. The isolates were resistant to FLU and 5-flucytosine
(FC) and had elevated minimum inhibitory concentrations (MICs) of voriconazole (VRC)
and caspofungin (CFG) [4, 5]. The most worrisome findings were persistent candidemia
and high attributable mortality rates [4, 5]. C. auris accounted for >5% of candidemia
in a national ICUs survey and up to 30% of candidemia at individual hospitals in India
[4, 19]. In the subsequent 2 years, several reports of hospital-associated infections
emerged from South Africa, United Kingdom, Venezuela, Colombia, United States, Pakistan,
Israel, Kenya, and Spain [7, 11–18, 24]. Table 1 lists several countries reporting
C. auris infection published so far across 5 continents. A collaborative project undertaken
by the US Centers for Disease Control and Prevention (CDC) to understand the global
emergence and epidemiology of C. auris reported that isolates from 54 patients with
C. auris infection from Pakistan, India, South Africa, and Venezuela showed that 93%
of isolates were resistant to FLU, 35% to AMB, and 7% to echinocandins; 41% were resistant
to 2 antifungal classes, and 4% were resistant to 3 classes [15]. The fact that this
yeast exhibits MDR clonal strains that are nosocomially transmitted is unusual in
other Candida species [3, 5, 21]. Therefore, the possible threat of rapid spread in
affected countries and its emergence in unaffected countries will not only challenge
clinicians for its effective therapeutic management but will also bring high economic
burden, especially to countries in resource-limited settings where modern identification
facilities and access to antifungals other than FLU are limited.
What are the drivers of clonal transmission and nosocomial outbreaks of C. auris?
There is increasing evidence that suggests likely transmission of C. auris in healthcare
settings. Recent reports highlight the persistent colonization by C. auris of hospital
environments and multiple body-sites of patients, leading to high transmissibility
and protracted outbreaks [13, 14]. A large outbreak of 50 C. auris cases in a London
cardio-thoracic center between April 2015 and July 2016 showed persistent presence
of the yeast around bed-space areas [13]. Genotyping with amplified fragment length
polymorphism (AFLP) demonstrated that C. auris isolates clustered. Similarly, the
investigation of the first 7 cases of C. auris infection identified in the US, which
occurred between May 2013 and August 2016, showed colonization with C. auris on skin
and other body sites weeks to months after their initial infection, which could possibly
lead to contamination of the healthcare environment and pose a risk of continuous
transmission [14]. Furthermore, C. auris was isolated from samples taken from the
mattress, bedside table, bed rail, chair, and windowsill [14]. WGS results demonstrate
that isolates from patients admitted to the same hospital in New Jersey were nearly
identical, as were isolates from patients admitted to the same Illinois hospital [14].
Also, in the London outbreak, a healthcare worker caring for a heavily C. auris–colonized
patient had a C. auris–positive nose swab [13]. Effective implementation of strict
infection-prevention control measures are required to prevent transmission of C. auris.
These include isolation of patients and their contacts, wearing of personal protective
clothing by healthcare workers, screening of patients on affected wards, skin decontamination
with chlorhexidine, environmental cleaning with chlorine-based reagents, and terminal
decontamination with hydrogen peroxide vapor or ultraviolet (UV) light [13, 29]. Enhanced
terminal cleaning with UV light has recently been shown to reduce infections with
many nosocomial pathogens and might also be of use for preventing C. auris transmission
[30].
Previously, several geographically related clusters have been reported from South
Korea [2, 3], India [4, 5, 10], South Africa [10], Pakistan [15], and hospitals in
Latin America [12, 16]. Clonality within C. auris has been shown using AFLP, multilocus
sequence typing, and MALDI-TOF MS among strains in India, South Africa, and Brazil
[10]. A recent study applying WGS demonstrated highly related C. auris isolates in
4 unrelated and geographically separated Indian hospitals, suggesting that this pathogen
exhibits a low diversity [21]. A large-scale application of WGS analysis suggests
recent independent and nearly simultaneous emergence of different clonal populations
on 3 continents, demonstrating highly related C. auris isolates in the same geographic
areas [15]. So far, no reservoir of C. auris has been identified, although future
studies on its isolation from animals, plants, and water sources are warranted.
Is antifungal resistance in C. auris a therapeutic challenge?
Patients with C. auris infections have risk factors similar to those of other Candida
spp. infections, including abdominal surgery (25%–77%), broad-spectrum antibiotics
(25%–100%), ICU admission (58%), diabetes mellitus (18%), presence of central venous
catheters (25%–94%), and malignancies (11%–43%) [3–5, 7, 12, 14–16]. The overall crude
in-hospital mortality rate of C. auris candidemia ranges from 30% to 60%, and infections
typically occur several weeks (10‒50 days) after admission [4, 5, 10, 12, 13]. C.
auris invasive infections represent a therapeutic challenge, and no consensus exists
for optimal treatment. A few studies report breakthrough fungemia while on FLU, and
this correlates with commonly reported high MICs (>32 μg/ml), suggesting intrinsic
resistance against this drug [3–5]. Although epidemiological cutoff values (ECVs)
or clinical breakpoints are not yet defined for C. auris, newer azoles such as posaconazole
(range, 0.06–1 μg/ml) and isavuconazole (range, <0.015–0.5 μg/ml) show excellent in
vitro activity against C. auris [4, 5, 7, 15, 19]. Analysis of antifungal data published
in various studies and depicted in Table 1 clearly shows that about 90% of strains
tested are resistant to FLU. Regarding VRC, elevated MICs are reported in 50% of isolates
in 2 large series published from India and the CDC [9, 15]. Furthermore, variable
susceptibility has been seen with AMB: 15%–30% of the isolates exhibit high (>2 μg/ml)
MICs [9, 15]. Up till now, echinocandin resistance is noted in fewer isolates (2%–8%)
[9, 14, 15], but almost half of isolates are MDR (resistant to ≥2 antifungal classes),
and a low number (4%) exhibit resistance to all classes of antifungals [2, 9, 12,
15, 16, 19]. Echinocandins remain the first-line therapy for C. auris infections,
provided that specific susceptibility testing is undertaken at the earliest opportunity.
Although CFG is normally highly effective against Candida biofilms, a recent report
demonstrated that CFG was predominately inactive against C. auris biofilms [29]. FC
(MIC50, 0.125–1 μg/ml) is a treatment option in renal tract or urinary tract infections,
as the echinocandins fail to achieve therapeutic concentrations in urine [4, 5, 7,
9, 11–13, 15, 18]. Also, a novel drug, SCY-078, which is the first orally bioavailable
1, 3-β-D-glucan synthesis inhibitor, has been shown to possess potent activity against
various Candida spp. and exhibit potent antifungal activity against C. auris isolates
[20]. Furthermore, SCY-078 showed growth-inhibition and anti-biofilm activity and
could be an important antifungal to treat this MDR species [20]. At present, the mechanism
of antifungal resistance in C. auris is unclear. The recently published draft genome
of C. auris revealed the presence of single copies of ERG3, ERG11, FKS1, FKS2, and
FKS3 genes [21]. Detection of azole-resistant mutations by comparing ERG11 amino acid
sequences between C. albicans and C. auris showed that alterations at azole-resistance
codons in C. albicans were present in C. auris isolates [15]. These substitutions
were strongly associated with country-wise–specific geographic clades [15]. Resistance
is probably inducible under antifungal pressure, resulting in rapid mutational changes.
However, future studies with emphasis on several molecular mechanisms, including efflux
and transporters, could provide insight on C. auris resistance.
What are the important things that we still need to learn about C. auris?
We are just beginning to know the epidemiology and behavior of C. auris, but at the
present, far more gaps exist in our knowledge. The earliest findings of C. auris are
from 1996. The pertinent question remains whether this pathogen existed far earlier
than 1996, and we were just unable to identify it. The latter is less plausible because
many centers have reviewed archived isolate collections that have not shown any isolates
of C. auris before 1996. We also do not know why C. auris is independently, almost
simultaneously, emerging in so many places worldwide. It has been shown that there
is a profound phylo-geographic structure with large genetic differences among geographic
clades and high clonality within the geographic clades. However, a common characteristic
is the high level of antifungal resistance, which is rare in other Candida spp. C.
auris is the only species in which several isolates have been identified with resistance
to all 4 classes of human antifungal drugs. It seems reasonable to opine that changes
or misuse of antifungal drugs is one of the factors, although no specific risk factors
for acquiring C. auris seem to exist. What we do know is that environmental factors
probably play a role in outbreaks in healthcare settings that include prolonged survival
in healthcare environments, probably due to skin colonization of patients and asymptomatic
carriers. It is obvious that future research is warranted on multiple aspects of C.
auris, which seems to have the typical characteristics of well-known, healthcare-associated
pathogens such as carbapenemase-producing gram-negatives, Clostridium difficile, vancomycin-resistant
Enterococcus (VRE), and methicillin-resistant Staphylococcus aureus (MRSA). Given
the behavior of the latter 4, a further spread of C. auris in healthcare settings
on a worldwide scale is expected. C. auris worldwide emergence has prompted the CDC,
(http://www.cdc.gov/fungal/diseases/candidiasis/candida-auris-alert.html [last accessed
February 2017]), Public Health England (PHE), London (https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/534174/Guidance_Candida__auris.pdf
[last accessed February 2017]), and the European Centre for Disease Prevention and
Control (ECDC), Europe (http://ecdc.europa.eu/en/publications/Publications/Candida-in-healthcare-settings_19-Dec-2016.pdf)
to issue health alerts for strict vigilance of C. auris cases. International collaborative
consortia and timely efforts by the medical community are indispensable in controlling
this super bug before it adapts in our healthcare facilities. Furthermore, more intensive
efforts are required, and one such crucial step is the support from funding agencies
to initiate multidisciplinary research to better understand its ecology, evolution,
and resistance mechanisms, which will go a long way for its treatment and prevention.