Epidemiology and Changes in Patient-Related Factors from 1997 to 2009 in Clinical Yeast Isolates Related to Dermatology, Gynaecology, and Paediatrics

From 1997 to 2009, 1,862 dermatology, gynaecology, and paediatrics (DGP) associated clinical yeast isolates were analysed for species occurrence, specimen origin and type, (multi-) resistance pattern, and testing period. The top seven of the isolated DGP-associated species remained the same as compared to total medical wards, with Candida albicans (45%) as most frequent pathogen. However, the DGP wards and DGP ICUs showed species-specific profiles; that is, the species distribution is clinic-specific similar and however differs in their percentage from ward to ward. By applying the “one fungus one name” principle, respectively, the appropriate current taxonomic species denominations, it has been shown that no trend to emerging species from 1998 to 2008 could be detected. In particular the frequently isolated non-Candida albicans species isolated in the DGP departments have already been detected in or before 1997. As yeasts are part of the cutaneous microbiota and play an important role as opportunistic pathogens for superficial infections, proper identification of the isolates according to the new nomenclature deems to be essential for specific and calculated antifungal therapy for yeast-like DGP-related infectious agents.


Introduction
Superficial fungal infections are often chronic and recurring. It has been estimated that approximately 15% of the population has fungal infections of the skin (tinea pedis or athlete's foot) or nails (onychomycosis) or of the feet. These infections are common in older children and adults [1]. Distal subungual, proximal, subungual, and white superficial onychomycoses are usually caused by dermatophytes, but Candida spp. may be present in all types in less than 1% of these cases [2]. In the past, yeasts are thought to be simply skin contaminants [3]; however, yeasts and nondermatophyte moulds may also cause toenail onychomycosis [4][5][6][7][8]. A higher proportion of yeasts is generally found in onychomycosis, where dermatophytes (68%), yeasts (29%), and moulds (3%) are the most causative fungal pathogens [9]. Some Candida spp. causing onychomycosis were reported to be partly resistant to oral antifungal agents (AFAs). In patients with chronic mucocutaneous infections, the main yeast pathogen is Candida (C.) albicans, but C. tropicalis, C parapsilosis, Issatchenkia (I.) orientalis, and Meyerozyma (M.) guilliermondii may also contribute to these infections [10].
It has been suggested by Clayton and Noble [11] that the spread of yeasts in the hospital ward occurs in a similar way to the spread of Staphylococcus aureus. In addition, the carriage rates of yeasts on the skin in hospital patients appear to be higher than those in the nonhospital population [11]. As cutaneous sites may act as common sources of infection,

Material and Methods
The 1,862 clinical yeast isolates (Table 1) were derived from dermatology, gynaecology, and paediatric wards from German University hospitals in Berlin (Charité), Dresden, Leipzig, Münster, and Munich (Ludwig-Maximilians-Universität München, and Technische Universität München) starting at the end of 1997 until February 2009, in the framework of 4 multicenter studies (MCS) [49][50][51]. The few numbers of strains ( < 40) of the years 1997 and 2009 were added to those isolates of 1998 and 2008, respectively. Therefore, the time period throughout this paper is referenced as 1998 to 2008. In addition, as no MCS were performed from 2005 to 2007, respectively, no DGP clinics participated; therefore isolates from this time period are missing. For comparison and possible trend recognition additionally, and with respect to the number of isolates, two similar test periods (1998-2001 and 2002-2008) have been set up out of the total studyperiod.
Identification and differentiation of the isolates were performed using methods routinely employed at the microbiology/mycology laboratories of the participating test centres. Confirmatory identification was made for unusual or notidentified species by FTIR and/or PCR at the appropriate reference laboratory of the individual multicenter study. As the "one fungus one name" principle is effective since 2013 [52,53], the current valid names for the appropriate species were applied throughout this paper as published in SpeciesFungorum [47], respectively, in MycoBank [48].
Susceptibility testing of these isolates was performed by microdilution against relevant azole antifungal agents, as described in the corresponding paper [46].

Species Distribution.
The distribution according to their isolation frequency of clinical Candida and NCY species isolated during the MCS from 1998 to 2008 of DGP patients is shown in Table 1. Candida and NCY species most frequently recovered are displayed in Figure 1 by their distribution of the year of isolation. As the "one fungus one name" principle is effective from the beginning of 2013 [51], the current valid taxonomic denominations for genera and species [47,53] were listed in parallel to the reported species names by the test centres (Table 1). The "new" genus/species names were used throughout this paper. As consequence of the speciesrenaming, a substantial reduction in species assigned to the genus "Candida" and a raise of "new" species occurred. Thus, of the Candida clade [37] of 19 Candida species reported, nine (47%) had to be renamed, and the namechanges were not only restricted to the Candida species. The "new" ("emerging") taxons, partly called before "not-Candida albicans Candida, " "not-Candida yeasts, " or "cryptic" pathogens [16-18, 34, 54-58], were already widely present at the beginning of and found throughout the German multicenter studies [49][50][51].
Aside from C. albicans, the highest prevalence was found for C. glabrata with 17% (G), 14% (D), 9% (P), 38% (G-ICU), and 17% (P-ICU). In a Hospital-Infection-Surveillance-Study in Germany [71] evaluating nosocomial infections in the ICU, C. albicans was found to be the most frequently pathogen causing vascular catheter associated sepsis (5.6% in all ICUs, 2.8% in paediatric ICUs), and the fourth most agent of urinary catheter associated UTI infections (8.7%). The distribution of the yeast species in the Turkish paediatric ICU was 2%/4%/1% for C. albicans, C. parapsilosis, and D. hansenii. No other species from this ICU had been reported. Although the incidence of candidemia was stable over a ten-year period (0.5 episodes/10.000 patient days per year), it was five times higher in ICUs than in other surgical wards in Switzerland [72]. However, during the recent decades a progressive shift from a predominance of C. albicans towards NCAC/NCY species (including C. glabrata and I. orientalis) has been reported [73], with C. glabrata accounting for 15%-20% of infections in most countries [74][75][76]. These differences in the occurrences of the most important and regularly isolated yeast species in the DGP wards and ICUs of the study are demonstrated in Table 3. C. parapsilosis and M. guilliermondii are most prominent only in the dermatology wards (48%/79%) and found to a significant lesser extent in the paediatrics units (28%/18%) but not at all in the gynaecology wards. Compared to the dermatology units the levels of C. albicans and C. glabrata are about 30% higher in gynaecology and paediatrics, whereas C. parapsilosis occurrence is about 30% higher in dermatology and paediatrics than in gynaecology wards. The occurrence of C. tropicalis (Table 3) is highest in paediatric wards (68%) and significantly lower in gynaecology (14%) and dermatology (5%).
Romeo and Criseo [77] found that 8 out of their 11 C. dubliniensis isolates were derived from oral specimens, and only 2 were found in vaginal and one in gastric fluid. Of the 15 C. dubliniensis strains in this study, one was isolated from blood, 2 from sterile body fluids, 5 from dermatological, and 7 from paediatric swabs. No C. dubliniensis, C. inconspicua, and D. hansenii isolates were derived from DGP ICUs ( Table 2) and none of these strains were found in DGP-specimens, except DGP swabs (Table 2). Gumral et al. [78] reported the lack of C. dubliniensis and C. africana strains in Turkey with vaginal C. albicans isolates, whilst Nnadi et al. [79] found in Nigeria no C. dubliniensis in vulvovaginal samples. As three C. africana isolates appeared in Berlin and two in Munich they were tested during a MCS in 2000, together with the strains from Angola and Madagascar. Although so far considered only as a new subspecies of C. albicans, C. africana should be reconsidered as separate species according to the original proposal of Tietz et al. [80]. This is supported by the results of Forche et al. [81] that rDNS sequences of C. dubliniensis differ significantly from those of C. albicans and that C. africana isolates are phylogenetically different. Moreover, C. africana could clearly be separated by FT-IR [82], probably nowadays by Matrix Assisted Laser-Desorption/Ionisation Time-Of-Flight Mass Spectroscopy (MALDITOF MS) as for C. dubliniensis [83], by pyrosequencing [84], or as described with a specific molecular method [85]. All these methods International Journal of Microbiology 7 demonstrated that they are able to discriminate distinctly between the very closely related species C. albicans, C. africana, and C. dubliniensis. Apart from the reported isolates in the study and those from vaginal specimens from Africa [80,81], none appeared in further German/Austrian MCS, and only a few strains were isolated later in Italy [77,85], Spain [86], Nigeria [79], and Great Britain [84].
Protothecosis is a sporotrichosis-like infection in humans, in both immunocompromised and immunocompetent patients, and in animals. It is caused by achlorophyllic algae of the genus Prototheca, which belongs to the family Chloracellae and is rarely involved in human infections [87,88]. The genus Prototheca (P.) consists currently of 6 species: P. wickerhamii, P. zopfii, P. stagnora, P. ulmea, P. blaschkeae, and P. cutis. P. zopfii contains currently two genotypes [89]. Species of the genus Prototheca exist in the environment as ubiquitous detritus inhabitants and contaminants of various substrates. General protothecosis is caused in humans mainly by P. wickerhamii and in domestic animals by P. zopfii. General symptoms are dermatitis or bovine mastitis, whereas mortal cases are extremely rare. P. wickerhamii and P. zopfii were isolated in 1998 in an outbreak in a children's unit during one of the MCS, where these organisms were transmitted from pet animals to the patients [82].
Of the twelve different species of Malassezia (M.) yeast described [90][91][92] [91], where M. sympodialis (72%) was the most frequently isolated Malassezia species. In addition, Petry et al. [91] reported that the back and the chest of the patients are the most common sites of the lesions, and no statistically significant difference was found between species as a function of gender, age, or the duration of the lesions [92]. Malassezia is strongly associated with dandruff, a common scalp disorder, although not all individuals with Malassezia on their skin develop dandruff. Besides Malassezia spp., which contribute by 5% to the population from dandruff-afflicted scalps, Filobasidium filoforme was reported to be the most isolated basidiomycete, whereas in healthy scalps Cryptococcus spp. (90%), together with Rhodotorula mucilaginosa, are detectable [92]. Whilst during the MCS of 1998 to 2008 Cryptococcus neoformans had been isolated 7 times (0.4% of total isolates), no Rhodotorula spp. infections were observed during the MCS. In addition, due to the limited time frame of the MCS within a study year, none of the Exophiala, Malassezia, and Prototheca species have been isolated anymore during the German/Austrian MCS up to 2009, with the exception of E. dermatitidis of which two strains each have been isolated in the MCS of 1999 and 2000 and of the two separate outbreaks of Malassezia and Prototheca spp. in 1998/99 (Table 1).
The opportunistic yeast pathogen Trichosporon (T.) asahii, which is part of the cutaneous fungal microbiota in humans, was isolated occasionally from 1998 to 2008 (0.2% of total isolates). T. asahii may be one of the routes through which deep-seated trichosporonosis is acquired, whereas environmental T. asahii is not associated with this infection [93].
The DGP-species distribution of urine samples was somewhat different to those isolated in a survey from 2003 to 2004 from urine specimens of 100 hospitalized patients in a Turkish hospital who had nosocomial candiduria [94].
With 80% to 95%, C. albicans is the predominant vaginal colonizing species in premenopausal and pregnant asymptotic and healthy women with acute Candida vaginitis and chronically recurrent vulvovaginal candidosis. NCACspecies, especially C. glabrata, are more frequent in postmenopausal, in diabetic and immunosuppressed women, paralleled by regional differences in the distribution of Candida species [95]. The results of this study illustrate that the yeast spectrum in gynaecological wards and its ICUs did not change significantly. This is in accordance with the findings by Mendling and Brasch [95], who reported at least for Germany no evidence of an increase of NCAC/NCY species in either acute or recurrent vaginal candidosis.
As shown in Table 2, only 4 devices (0.2%) have been sent for determination of associated fungi. Three devices were derived from the dermatology and one device from the paediatric ward. From the isolated species thereof (C. albicans, C. glabrata, and Magnusiomyces capitatus), at least two of them belong to the numerous Candida species (e.g., C. glabrata, C. parapsilosis, C. tropicalis, C. dubliniensis, and I. orientalis), which are reported to form biofilms, catheterrelated blood-stream, and device-related infections [96][97][98]. As it was not mandatory for the in vitro multicenter studies to report detailed epidemiological data, the voluntarily gathered data were insufficient to evaluate more patient-related factors.

Conclusions
About 20-25% of the world's population is affected by skin mycoses, thus being one of the most frequent forms of infection [99]. The epidemiological trend in skin mycoses International Journal of Microbiology worldwide is paralleled to changes of nosocomial and invasive fungal infections. Although a significant shift in the distribution of the infection causing agents is reported for dermatology, gynaecology, and paediatric wards, apart from some local breakouts with Malassezia, Prototheca, and Exophiala species, all the infection-causing agents have been present before and throughout the 10-year study period. Aside from significant differences in the species profiles of the DGP wards, a trend in the distribution of the DGPspecies could not be detected and the overall aetiology has not changed during the time period of the multicenter studies from 1997 to 2009. But, besides the typical skin pathogens like C. albicans, C. glabrata, C. parapsilosis, C. tropicalis, C. dubliniensis, and C. inconspicua, infections with atypical, rare, or "cryptic" yeast isolates, which all have been existent, such as Issatchenkia orientalis, Saccharomyces cerevisiae, Meyerozyma guilliermondii, Kluyveromyces marxianus, Clavispora lusitaniae, Debaromyces hansenii, and Yarrowia lipolytica, tend also to emerge in the DGP wards and DGP ICUs, respectively, as reported for the other wards. This may be markedly amplified by the taxonomic changes which are to be implemented since the beginning of 2013 comprising taxonomic reclassifications and concomitant (partially) renaming of various species according to the "one fungus one name" principle. However, this novel practice based on the phylogenetic mapping of the species may allow in future a better association of different or similar pathogens to clinical entities and characteristics. This may also lead to a better and reliable assessment of in vitro susceptibility data (given in a corresponding paper [46]), which represent the basis not only for specific antifungal therapy, but in particular also for calculated ("empiric") antifungal therapy.