Clinical epidemiology

A literature search, restricted to publications in English, including publications up to 8 December 2017, was performed by using Medline/Pubmed. At that moment, published clinical reports of C. auris infections originated from Japan, South Korea, India, Kuwait, Oman, South Africa, Venezuela, Panama, Colombia, Pakistan, Israel, Spain, UK, Canada and the USA describing 109 patients with candidaemia, 19 patients with chronic otitis media or otitis externa and 28 patients with other infections (including candiduria/urinary tract infections (n=16) and wound and soft tissue infections (n=6)).1–4 8–21 Real-time data from the Centers for Disease Control and Prevention show that new cases are reported on an almost weekly basis with a total of 243 clinical cases of C. auris infection in the USA.22 Public Health England shows a more episodic pattern in reported clinical cases (56 in total) with high increases during outbreaks.23 No paediatric patients have been reported in the UK (Dr E M Johnson, Public Health England, personal communication).

Ten out of the 18 publications described infections in adult patients only. Two publications did include paediatric patients but no details were given separately for the patients <18 years of age19 21 or age was not mentioned.13 Twenty-three paediatric patients have been reported in five case series published, of which 20 neonates and children suffered from a blood stream infection (18% of the total population).3 8 9 12 Three other paediatric patients suffered from chronic otitis media with positive ear swabs for C. auris.2 Table 1 summarises the clinical characteristics of those infections. Paediatric patients have only been reported in Asia and South America. Underlying conditions and risk factors are comparable to those known to render paediatric patients at risk for developing candidaemia and invasive candidiasis. Of the 20 paediatric patients with candidaemia due to C. auris, 14 were neonates and/or born prematurely. Older infants and children developed C. auris candidaemia during intensive care unit (ICU) admissions, postsurgery or with an underlying haematological malignancy. Antifungal regimens prescribed varied hugely with half of the patients receiving combination antifungal therapy with two or three antifungals. Mortality was 30% in paediatric patients with C. auris blood stream infections and lower compared with adult patients with mortality rates ranging from 30% to 60%. In contrast, during the C. auris outbreak in a large UK hospital, no attributable deaths were observed due to C. auris infections.4

As C. auris has been commonly misidentified as Candida haemulonii, a rare encountered Candida species causing invasive infections.3 8 10 11 13 Paediatric patients with C. haemulonii blood stream infections reported in the literature have been summarised in table 2. The clinical characteristics are comparable to those described for C. auris fungaemia and neonates and children with well-known underlying conditions and risk factors to develop invasive fungal disease are affected. Thirteen paediatric patients from Kuwait, Korea and Brazil have been described, including five premature neonates.2 24–26 C. haemulonii infections have been more often described in infants and children (62%) compared with C. auris infections affecting mostly neonates (70%). Half of the patients with C. haemulonii infections were treated with combination antifungal therapy and the mortality rate was 30%.

Due to the low numbers of paediatric patients described, it is difficult to draw meaningful comparisons with Candida infections caused by other species, specific patient groups affected, geographic patterns and the outcome of the infections.

Identification and susceptibility

C. auris isolates have been misidentified as a range of other Candida species by using conventional phenotypic and biochemical methods. Most commonly, these isolates have been misidentified as C. haemulonii, as C. auris is phylogenetically closely related to the C. haemulonii species complex.3 8 10 11 13 C. haemulonii complex species are less frequently detected than C. auris, although inaccuracies with the molecular identification of less common Candida species prevents a robust insight into these infections.5 It is also possible that some of the reported isolates of C. haemulonii are misidentified as C. auris. It has been suggested that chromogenic agar is a low-cost method to differentiate between C. auris and C. haemulonii isolates.5 Molecular techniques are recommended to identify C. auris to the species level. Matrix-assisted laser desorption ionisation–time of flight mass spectrometry is capable of providing an accurate identification to the species level once spectra for C. auris are present in its database.5 The development of C. auris-specific PCR assays allows for rapid identification and are of particular use during outbreak settings.5 Sequencing of genetic loci and internal transcribed spacer domains of the rRNA provides the ability to differentiate between geographic clades.5 21 Genome sequencing has been shown to be useful in the identification of C. auris but is often not an available technique in most laboratories and is less feasible for routine identification purposes.

Intrinsic susceptibility patterns for C. auris show elevated minimum inhibitory concentrations for all three classes of antifungals, the polyenes, the azoles and the echinocandins.27 Exact breakpoints as have been established for the common Candida species using the Clinical and Laboratory Standards Institute and the European Committee on Antimicrobial Susceptibility Testing methodologies have not been established.28 Antifungal susceptibility data for 54 C. auris isolates showed that 93% were resistant to fluconazole (≥32 mg/L), 54% to voriconazole (≥2 mg/L), 35% to amphotericin B (≥2 mg/L) and 7% to echinocandins (≥8 mg/L).3 Forty-one per cent of those isolates were resistant to greater than or equal to two classes of antifungals. Resistance to all three classes of antifungals was observed in two Indian isolates.21 Of the 13 C. haemulonii infections in paediatric patients, eight isolates (62%) showed amphotericin B resistance, two isolates (15%) showed resistance for fluconazole and two (15%) for echinocandins.2 24–26

Emergence of C. auris

Results from whole-genome sequencing studies suggest that C. auris has emerged near simultaneously in four or more locations rather than spreading from a single source.21 Sequencing of C. auris isolates in the UK have clearly shown that those isolates have several geographic origins and belong to at least three different clades.29 This suggest that C. auris isolates have been introduced into the UK from different locations.

But what has driven the emergence of this new Candida species? One of the explanations could be that C. auris has just not been recognised before, but this is not supported by retrospective reviews of large collections of Candida isolates. A thorough review of over 15 000 Candida isolates from four continents (SENTRY isolate collection) did not detect any C. auris isolate before 2009.21 Increasing antifungal selection pressures either in humans and/or animals and/or the environment may cause the emergence of a new multidrug-resistant Candida species. The availability of antifungals with an improved toxicity profile has contributed to an expansion in the prescriptions of antifungals for prophylactic and empirical use. Increased antifungal use in agriculture and in the clinical environment has led to well-recognised emerging resistance among Aspergillus fumigatus (azole resistance) and C. glabrata (echinocandin resistance).30 However, selection by antifungal pressure alone seems to be less likely as non-albicans Candida species have been increasing since the introduction of fluconazole in 1990. Changes to the ecological niches of C. auris and intrinsic characteristics as numerous virulence attributes, thermotolerance and salt tolerance and aggregation into difficult-to-disperse clusters, may have allowed the emergence of this newly recognised species causing human infections and persistence in hospital environments.31 A number of molecular resistance mechanisms are well described for causing resistance to particular antifungals in several Candida species.28 As yet, molecular mechanisms resulting in antifungal resistance in C. auris need to be elucidated.

Clinical management

Based on susceptibility data, echinocandins seem to be the drug of choice as less than 10% resistance is reported.27 28 Micafungin showed increased efficacy in a murine study of C. auris candidaemia compared with fluconazole and amphotericin B.32 For the paediatric population, micafungin and caspofungin can be used while awaiting susceptibility testing results.33 34 Depending on the site of infection, alternative choices might be considered.5 28 33 34 Echinocandins have limited penetration in the cerebrospinal fluid and urine, and therefore other antifungal compounds should be used to treat central nervous system or renal tract infections.35 No recommendations can be made with regard to treatment with combinations of antifungals, as data are not available.

Consideration should be given to determine Candida isolates from mucosal surfaces and non-sterile sites to the species level and perform susceptibility testing. Particularly, in paediatric patients at high risk for developing invasive Candida infections, as these results may guide early and targeted treatment if a Candida infection is suspected based on clinical signs and symptoms. Even more so in premature neonates where a high percentage of false-negative blood cultures for Candida species has been described.36

An additional reason to be informed about colonising C. auris isolates in intensive care settings is to prevent transmission and to instal proper infection prevention and control measures. Several countries including the UK have released guidance documents with recommendations regarding screening policies, isolation of patients, contact precautions and cleaning of equipment and clinical environments.5 21

Summary

C. auris infections have emerged as an important challenge in the management of already vulnerable paediatric patients including premature neonates, infants and children admitted to ICUs and those with underlying malignancies. Proper identification is challenging with conventional methodologies failing. Great concern exists how to control this emerging new Candida species showing multidrug resistance and being very well capable to persist in hospital environments. The increased number of cases detected worldwide is causing a global concern and research is urgently needed. Current research is addressing the many unanswered questions related to the emergence of this Candida species, its molecular resistance and virulence mechanisms and improved management options to minimise its impact on patient outcomes.