Candida species isolated from clinical samples in a tertiary hospital in Honduras: Where is Candida auris?

Background and Purpose: Infections by emerging and multiresistant Candida species are becoming more frequent throughout the world. This study aimed to describe Candida species in different wards of a tertiary hospital in Honduras. Materials and Methods: The prevalence of species within the C. albicans complex was estimated using a molecular approach, and C. auris was investigated using a yeast pool-based DNA extraction method. In total, 328 yeast isolates were identified using phenotypic approaches. For the identification of species within the C. albicans complex, a molecular approach based on the size polymorphisms of the hpw1 gene was used. In addition, a technique was optimized based on DNA extraction in pools for the rapid identification of C. auris. Results: A total of 11 species of Candida were identified in the hospital wards. C. albicans showed the highest number of isolates (52.4%). Within the C. albicans complex, C. albicans sensu stricto was the most common, followed by C. dubliniensis. However, C. auris was not found. Conclusion: Reports on the distribution of Candida species in Honduras are limited; accordingly, the data from this study are of importance for a better understanding of their epidemiology. Moreover, a simple method was offered for the detection of C. auris that could help in its detection in low-resource settings.

Introduction he incidence of fungal infections has increased significantly in recent decades, particularly those caused by Candida species [1,2]. Candidiasis range from mild superficial to life-threatening invasive infections [3], with an annual prevalence of tens of millions of cases and mortality rates above 40% [4,5]. More than 300 species belong to the genus Candida [6]. In the 1990s, human Candida infections were limited to five species, namely C. albicans, C. glabrata, C. parapsilosis, C. tropicalis, and C. krusei [7,8]; however, thanks to advances in molecular diagnostic tools, more than 37 species have been reported [9].
C. albicans remains the most frequently isolated species [10,11]; however, the epidemiology of Candida seems to have undergone recent changes [12]. Infections caused by species other than C. albicans have increased markedly, accounting for more than 50% of isolates recovered from blood cultures [11]. This increase could be due to the widespread use of broad-spectrum antibiotics, immunosuppressive agents, prolonged hospitalization, chemotherapy, and more recently, patients infected with SARS-CoV-2 [4,11,13].
C. glabrata, C. parapsilosis, C. tropicalis, C. krusei, C. guilliermondii, and C. lusitaniae are the most common causes of infections caused by nonalbicans species [9,11,14]. C. auris, a multiresistant pathogen described in 2009, has been added to this list of non-albicans yeasts that are most frequently isolated in the hospital environment [9,13,14] due to its high capacity to indefinitely colonize various anatomical locations, which facilitates its transmission between patients and the appearance of long-term outbreaks. On the other hand, it is estimated that around 25% of patients colonized by C. auris may develop candidemia and/or invasive candidiasis. In addition, the epidemiological relevance of this species is due to the high resistance to antifungals and disinfectants and the difficulties to identify it by the phenotypic methods available in routine laboratories [15,16]. This is a worrying scenario for public health, mainly due to the difference in the sensitivity profiles of these species against the available therapies [17], making it difficult to manage patients, resulting in high mortality rates and high financial costs [2,18]. Sustained surveillance and monitoring of the epidemiology of clinically important yeasts should be a priority, particularly in countries considered low-and middle-income where yeast identification methods are limited. Knowledge of the local epidemiology of Candida species is essential for adequate prevention and clinical management since in many cases the identification of species allows predicting their potential susceptibility to antifungals.
The main objective of this study was to describe, through a phenotypic approach, the Candida species circulating in the wards of a hospital in Honduras. In addition, the prevalence of species within the C. albicans complex was estimated, and the presence of C. auris was investigated using molecular methods and a DNA extraction methodology based on yeast pools.
Regarding ethical considerations, the study was conducted on anonymous biological samples. Furthermore, this research did not involve any personal data that could directly or indirectly identify a specific individual. Therefore, a statement of consent from the ethics committee of the hospital or university was not necessary. Clinical samples were collected according to ethical standards and due to the retrospective nature of the study, the consent form was not applicable to the study.

Yeast Identification
Clinical With the isolated colonies, a suspension was made with a density equivalent to the McFarland standard Nº2. Suspensions were then inoculated onto yeast identification panels and loaded onto the BD Phoenix TM M50 instrument (Becton, Dickinson and Company, NJ, USA). Results were registered after 4 to 8 h. The prevalence of Candida species was determined according to the hospital ward and gender of patients.

Detection of Candida species from yeast pools
To assess the presence of three species of the C. albicans complex (C. albicans sensu stricto, C. dubliniensis, and C. africana), the approach proposed by Theill et al. was used [19] with slight modifications. This method is based on the combination (pooling) of individual axenic cultures for the detection of a target species. This approach was also used to detect C. auris.
To optimize the pooling method and demonstrate its sensitivity in detecting C. auris, 10 Candida species, including a reference strain of C. auris, were cultured individually in YPD liquid medium (1% yeast extract, 2% peptone, and 2% dextrose) at 30°C for 24 h under constant stirring at 200 rpm. Individual cultures were arranged in pools that included 300 µL of each culture. Six different pools were tested ( Figure 1). Once the pools were prepared, DNA extraction was carried out using a previously described approach based on phenol chloroform [20].

Molecular identification of species within the albicans complex
Isolates identified as C. albicans were grown individually in 4 mL YPD broth at 30°C for 24 h with constant shaking at 200 rpm. 300 µL of each culture were taken and mixed in pools of 10. DNA was extracted from each pool according to a previously published protocol [20].
The identification of the species within the albicans complex was based on a size polymorphism in the hyphal wall protein 1 gene (hwp1) [21]. DNA was amplified in a volume of 50

Molecular detection of C. auris
Briefly, pools were prepared using 10 yeasts previously identified as non-albicans species or as C. tropicalis by the BD Phoenix Yeast Identification System or CHROMagar™ Candida Medium (Becton, Dickinson and Company, NJ, USA). This decision was based on previous reports of white to mauve colonies of C. auris on CHROMagar TM (Becton, Dickinson and Company, NJ, USA), and because identification has not yet been optimized for any clade of C. auris in the BD Phoenix Yeast Identification System (Becton, Dickinson and Company, NJ, USA), this yeast could be misidentified as C. catenulata, C. haemulonii, or C. parapsilosis [15,22,23].
The identification of C. auris was carried out by amplifying the gene glycosylphosphatidylinositol (gpi) [24]. A mix was prepared with 12.5 µL of 2× Master Mix (Promega Corp. Madison, WI, USA), 9 µL of nuclease-free water, 1.5 µL of DNA (40 ng/µL), and 1 µL of each primer 10 µM: 03410_F: 5′-GCCGCTAGATTGATCACCGT-3′ and 03410_R: The PCR program consisted of 1 cycle at 94°C for 3 min, 35 cycles at 94°C for 30 s, 60°C for 30 s, 72°C for 30 s, and a cycle at 72°C for 3 min. A band size of 137 bp was indicative of C. auris, and no amplification product was expected for the rest of the species. Positive and negative controls were included in each PCR reaction.
Yeasts were most frequently isolated from urine (44.5%), sputum (23.7%), and blood (11.5%). C. albicans was the most frequent species in most of the samples with the exception of rectal exudates/feces and blood, in which the non-albicans species were more   Figure 2). Since 80% of the isolates were from urine, sputum, or blood, the distribution of Candida species in these samples was analyzed according to the hospital ward. The wards with the highest number of isolates in urine were internal medicine (26%), emergency (21%), and gynecology (14.3%). The most frequent non-albicans species in urine were C. tropicalis, C. glabrata, and C. lusitaniae. The internal medicine (44%), emergency (30%), and COVID-19 (10.3%) wards contributed more isolates from sputum. The non-albicans species most frequently isolated from sputum were C. krusei, C. tropicalis, and C. glabrata.
Internal medicine (29%), emergency (21%), and pediatrics (18.4%) had the highest number of isolates from blood cultures. Non-albicans species (C. glabrata, C. krusei, and C. parapsilosis) accounted for 73.6% of bloodstream infections. Two cases of mixed candidemia were found, one caused by C. krusei / C. lusitaniae in a patient in the COVID-19 ward, and another case caused by C. glabrata / C. parapsilosis from pediatrics. Cases of mixed candidemia accounted for 5.3% of isolates from blood cultures. The prevalence of candidemia in the ICU was 10.5%.
Regarding C. auris, a simple and inexpensive method based on DNA extraction from mixed cultures was optimized for the detection of suspect isolates. In the six pools prepared to optimize the method, C. auris DNA was successfully amplified (Figure 3). A total of 112 yeast isolates suspected of being C. auris were evaluated; however, it was not demonstrated in any case.

Discussion
C. albicans was the most frequently isolated species. This result is consistent with a previous study carried out in the same hospital in 2019 [25], as well as with several studies from other countries [3,[26][27][28][29]. Although the incidence of non-albicans species seems to be increasing [9,11,30], our results reveal that C. albicans remains the most frequently isolated species from clinical samples in Honduras. C. tropicalis, C. krusei, and C. glabrata were the most commonly isolated yeasts after C. albicans. The incidence of these species agrees with a previous study carried out in Honduras where 167 cultures were evaluated, with C. albicans, C. tropicalis, and C. glabrata being the most commonly isolated species [25]. Similar results have been reported in Mexico and India [27,31,32].
C. krusei represented 11.3% of isolates and was the third most frequent species. This represents a two-fold increase in the number of isolates, compared to the 2019 study when 4.5% of isolates were reported from the same hospital [25]. This difference could be attributed to the two-year difference between both studies, or it could also be influenced by the indiscriminate use of azoles, particularly fluconazole, which has been related to the increase in C. krusei due to selective pressure exerted on sensitive strains [33]. The scope of the present study was not intended to confirm this hypothesis, and future studies would probably delve into the reasons for these changes in the epidemiology of non-albicans species.
Urine, sputum, and blood were the samples in which yeasts were isolated most frequently. Our results coincide with two available studies on the epidemiology of Candida in Honduras [20,25] in which C. albicans is described as the predominant yeast in urine and sputum. Similarly, our results are also similar to those previously reported elsewhere [32,34,35]. The clinical significance of yeasts in urine and sputum is still not entirely clear, as it could indicate contamination, commensalism, colonization, or even a localized infection that could later evolve into a disseminated infection [20,36]. Whatever the meaning of isolating yeasts in urine or sputum, it is important to report their presence so that the clinician has a complete picture of the microbiota in the patient.
Two studies have evaluated species causing candidemia in Honduras. In a laboratory-based survey conducted at 20 centers in seven Latin American countries, nine species of Candida were reported to cause candidemia among 135 cases in Honduras [37], while Montes et al. reported five species among 12 cases [25]. Here, we identify seven species isolated from the bloodstream. In three studies, the nonalbicans species were the most frequent in blood, widely surpassing C. albicans. These data support those reported by Pappas et al., where non-albicans species could cause more than 50% of candidemia cases worldwide [11].
C. glabrata, C. krusei, and C. parapsilosis were found in equal proportion, disputing the second place as causative agents of candidemia. Infections with C. glabrata, C. krusei, and C. parapsilosis have been associated with mortality rates ranging from 10% to 40% [33]. C. glabrata causes 5%-20% of candidemia cases worldwide and is the second leading cause of candidemia in the United States and Northern Europe [38,39]. In addition, this yeast is associated with elderly patients and transplant recipients [33,40].
In recent years, the increased prevalence of C. glabrata resistant to azoles and echinocandins has been reported [37,[40][41][42]. C. krusei is commonly associated with infections in patients with hematologic malignancies who are undergoing treatment with fluconazole, echinocandins, or antibacterial agents, specifically vancomycin or piperacillin-tazobactam. It also has intrinsic resistance to fluconazole; and in vitro high minimal inhibitory concentrations against amphotericin B have been observed [33,37,43].
C. parapsilosis is the second most frequently isolated species in Asia, Latin America, and in some hospitals in southern Europe [37,40,41,44], whereas in North America, it has been considered the third most frequent [45]. Infections by C. parapsilosis predominate in low-birth-weight newborns, critically ill or cancer patients [44,46]. Although in most reports C. parapsilosis tend to be sensitive to available antifungals, it has reduced susceptibility to echinocandins [47,48]. Secondary resistance to azoles in C. parapsilosis, particularly fluconazole is also becoming frequent [49][50][51]. Our results highlight the need for a comprehensive effort to assess the distribution of Candida species and risk factors associated with systemic candidiasis in Honduras. Antifungal susceptibility profiles need to be routinely assessed as well, as non-albicans species could be associated with increased resistance [17].
Two cases of mixed candidemia were demonstrated. Most cases of candidemia are usually monomicrobial, while cases of mixed candidemia are rare [52,53]. Within the candidemia cases, the estimated prevalence of mixed candidemia is usually less than 10% [53,54], which makes this finding an unusual event.
In this study, the prevalence of the three species of the C. albicans complex was determined. Historically, C. albicans has been considered the main causal agent of superficial and invasive candidiasis [9,41]; however, isolates identified as C. albicans had phenotypic and genetic inconsistencies at the time of their identification, which is why they were called atypical isolates of C. albicans [55,56]. Due to advances in molecular diagnostic tools, atypical strains were later considered new species within a complex [55,56]. Hence, the epidemiology of C. albicans sensu stricto may have been historically overestimated. Within the isolates classified in the albicans complex, 2.8% were C. dubliniensis, while C. africana was not found. Fontecha  Since C. dubliniensis was first identified in Ireland as the cause of oropharyngeal candidiasis in patients living with HIV, it has been identified in many other countries [19,28,[57][58][59][60][61]. The pathogenesis and antifungal susceptibility of C. dubliniensis are not clearly defined. Several studies suggest that C. albicans and C. dubliniensis share the same in vitro antifungal susceptibility patterns, as well as clinical breakpoints and epidemiological Cut-Off values [19,58,[62][63]; however, Ayadi et al. studied the in vitro reaction of C. albicans and C. dubliniensis against fluconazole and two echinocandins and observed that both yeasts behave differently [64]. Likewise, the comparative analysis of the genomic and transcriptomic data has revealed important differences in their infective capacity [65].
A technique was optimized based on pools in order to offer an easy and economical method to identify C. auris. This method was then applied to screen 112 suspected isolates of C. auris, and as reported in 2017 and 2019, it was not possible to demonstrate the presence of C. auris in the country yet [20,57]. Consequently, the evidence seems to suggest that C. auris is not yet found in Honduras, or at least in the hospital where this study was carried out. However, the report of C. auris in other Latin American countries [66], plus the recent association of outbreaks by C. auris in the context of the COVID-19 pandemic [66], makes its active search a priority for the country. Having tools that allow the timely identification of C. auris when the first cases appear will guarantee quick decision-making for its management and control.
The main limitation of this study is that the data were collected from a single hospital in Tegucigalpa; therefore, our results cannot necessarily be extrapolated to the epidemiological situation of candidiasis in the country. Another limitation is the lack of identification of cryptic species of the C. parapsilosis and C. glabrata complexes.

Conclusion
This study updated the distribution of Candida species isolated from clinical samples and provided data to better understand the epidemiology of Candida species in Honduras. In addition, a simple and inexpensive technique was offered for the detection of C. auris that could help in the rapid detection of this pathogen in low-resource laboratories. It is suggested that decision-makers remain alert to the possible appearance of C. auris in Honduras.