Prevalence, Distribution, and Antifungal Susceptibility Profiles of Candida parapsilosis, C. orthopsilosis, and C. metapsilosis in a Tertiary Care Hospital

Abstract

Candida parapsilosis, an emergent agent of nosocomial infections, was previously made up of a complex of three genetically distinct groups (groups I, II, and III). Recently, the C. parapsilosis groups have been renamed as distinct species: C. parapsilosis sensu stricto, C. orthopsilosis, and C. metapsilosis. In Portugal, no data pertaining to the distribution and antifungal susceptibility of these Candida species are yet available. In the present report, we describe the incidence and distribution of C. parapsilosis sensu stricto, C. orthopsilosis, and C. metapsilosis among 175 clinical and environmental isolates previously identified by conventional methods as C. parapsilosis. We also evaluated the in vitro susceptibilities of the isolates to fluconazole, voriconazole, posaconazole, amphotericin B, and two echinocandins, caspofungin and anidulafungin. Of the 175 isolates tested, 160 (91.4%) were identified as C. parapsilosis sensu stricto, 4 (2.3%) were identified as C. orthopsilosis, and 5 (2.9%) were identified as C. metapsilosis. Six isolates corresponded to species other than the C. parapsilosis group. Interestingly, all isolates from blood cultures corresponded to C. parapsilosis sensu stricto. Evaluation of the antifungal susceptibility profile showed that only nine (5.6%) C. parapsilosis sensu stricto strains were susceptible-dose dependent or resistant to fluconazole, and a single strain displayed a multiazole-resistant phenotype; two (1.3%) C. parapsilosis sensu stricto strains were amphotericin B resistant. All C. orthopsilosis and C. metapsilosis isolates were susceptible to azoles and amphotericin B. A high number of strains were nonsusceptible to the echinocandins (caspofungin and anidulafungin).

At present, Candida parapsilosis stands out as the second most common yeast species isolated from patients with bloodstream infections in Latin America and Asia (38, 44), but it is also commonly found in European surveys (1, 32, 44). Similar results were found in Portugal, according to a prospective, observational study of bloodstream fungal infections (6).

This pathogenic fungus is responsible for a broad variety of clinical manifestations that generally occur in individuals with impaired immune systems, neutropenia, or burns and patients admitted to neonatal or surgical intensive care units (32) and that also commonly occur in pediatric units (11, 39).

Isolates of C. parapsilosis are phenotypically indistinguishable but are genetically heterogeneous. On the basis of randomly amplified polymorphic DNA analysis (20), multilocus enzyme electrophoresis (22), analysis of the internal transcribed spacer sequences of DNA encoding ribosomes (22), DNA relatedness analysis (41), morphotyping (2), analysis of mitochondrial DNA sequence differences (25), DNA topoisomerase II gene sequence analysis (14), and analysis with an oligonucleotide probe used for fingerprinting of C. parapsilosis isolates (9), it was found that C. parapsilosis forms a complex composed of three genetically distinct groups (groups I, II, and III). Tavanti et al. (47) proposed the replacement of the existing designations of C. parapsilosis group II and group III by C. orthopsilosis and C. metapsilosis, respectively, and retention of the designation C. parapsilosis for the former group I isolates. In order to distinguish between the three species, Tavanti et al. (47) proposed the analysis of the restriction polymorphism of the SADH gene, which encodes a secondary alcohol desidrogenase and which is common to all three species.

Group I, or C. parapsilosis sensu stricto, represents the predominant species among clinical isolates, which may partially be explained by its enhanced ability to form biofilms (47). Recent studies have shown that some clinical isolates formerly identified as C. parapsilosis in fact correspond to isolates of the closely related species C. orthopsilosis and C. metapsilosis (15, 22, 23, 47) and account for about 10% of the total number of former C. parapsilosis isolates (23). These two new species may inhabit important niches in certain patient populations, thus stressing the need for continued surveillance in each country to monitor the prevalence and distribution of these three species.

Although C. parapsilosis is not considered to be particularly prone to the development of antifungal resistance (19, 31, 34, 48), several recent reports suggested that its decreased susceptibility to azoles and echinocandins might become a cause for clinical concern (3, 34, 36, 45, 48). The emergence of C. parapsilosis strains that have decreased susceptibility to fluconazole (FLU) and that may be transmitted throughout the hospital environment may be related to the extensive use of FLU in combination with suboptimal hand hygiene and central venous catheter care in a population of seriously ill patients (1).

Few studies have yet addressed the global epidemiology and antifungal susceptibility profile of C. parapsilosis (1, 34, 49). Induction studies showed an increase in the level of resistance after exposure to azoles (46). This fact, in association with the lower susceptibility of C. parapsilosis to echinocandins, urged the collection of all possible additional information regarding this opportunistic fungal pathogen. Moreover, the definition of the antifungal susceptibility profile of the new species C. orthopsilosis and C. metapsilosis is of clinical relevance. The differences in susceptibility between these three species may influence the therapeutic choices that are made.

Due to the medical importance of C. parapsilosis in Portugal (6) and the lack of availability of local epidemiological data, we aimed to provide insight into the incidence and distribution of C. parapsilosis sensu stricto, C. metapsilosis, and C. orthopsilosis in a large (1,350-bed) university hospital in the north of Portugal. Additionally, we evaluated the susceptibilities of such isolates to six commonly used antifungal agents, namely, FLU, voriconazole (VOR), posaconazole (POS), amphotericin B (AMB), and the echinocandins caspofungin (CAS) and anidulafungin (ANI).

MATERIALS AND METHODS

Strains.

One hundred sixty-seven strains were collected from patients admitted to different care units of the Hospital S. João over a 4-year period; the assumed phenotypic and genetic diversity of this set of strains is based on the distinct date and anatomic site of isolation. Isolates obtained from environmental studies were also included: five isolates were collected from the air by filtration with a MAS-100 Eco instrument (Merck Eurolab, Switzerland), and three isolates were collected from food administered to patients and isolated in Cooke Rose Bengal agar (Difco Laboratories, Sparks, MD). All isolates were initially identified as C. parapsilosis by the use of Vitek YBC identification cards (bioMérieux, Paris, France). Type strains ATCC 22019, ATCC 96139, and ATCC 96144, from the American Type Culture Collection, were used as controls for C. parapsilosis, C. orthopsilosis, and C. metapsilosis, respectively. Details regarding the 175 C. parapsilosis isolates are depicted in Table 1.

TABLE 1.

Distribution of C. parapsilosis, C. orthopsilosis, and C. metapsilosis

Site of isolation	No. (%) of isolates
	C. parapsilosis sensu stricto	C. orthopsilosis	C. metapsilosis	Total
Urine	9	4	0	13
Respiratory tract	2	0	1	3
Central venous catheter	6	0	0	6
Blood	60	0	0	64
Mucosal surface	65	0	4	71
Peritoneal fluid	1	0	0	1
Bile fluid	2	0	0	2
Spinal fluid	1	0	0	1
Hospital environment	8	0	0	8
Unknown	6	0	0	6
Total	160 (91.4)	4 (2.3)	5 (2.9)	175a (100)

^aThe total number of isolates is higher than the sum of the numbers of isolates of C. parapsilosis sensu stricto, C. orthopsilosis, and C. metapsilosis since the molecular analysis of six isolates showed that they corresponded to distinct Candida species.

DNA extraction.

Yeast genomic DNA was extracted as described by Hoffman and Winston (13), with some adaptations. Briefly, yeast cells grown overnight in YPD medium (2% glucose, 2% peptone, 1% yeast extract) at 30°C with agitation were harvested and lysed by vortexing for 10 min with 0.3 g of glass beads (diameter, 0.45 to 0.52 mm; Sigma, St. Louis, MO), 200 μl of lysis buffer (100 mM Tris-HCl, pH 8.0, 2% Triton X-100, 1% sodium dodecyl sulfate, 1 mM EDTA), and 200 μl of phenol-chloroform-isoamyl alcohol (25:24:1; Fluka Biochemika). After the mixture was vortexed, 200 μl of TE (1 mM EDTA, 10 mM Tris-HCl, pH 8.0) was added to the lysate, the mixture was microcentrifuged at 20,000 × g for 10 min, and the aqueous phase was transferred to a new vial. DNA was precipitated following the addition of 1 ml of ethanol and afterwards was treated with RNase (10 mg/ml; Sigma) for 1 h at 37°C; the DNA was reprecipitated and resuspended in 100 μl of TE.

SADH gene restriction profile.

Amplification of the SADH gene was performed by PCR with the primers previously described by Tavanti et al. (47). The amplification conditions were as follows: a first cycle of 2 min at 94°C, followed by 35 cycles at 94°C for 30 s, 50°C for 30 s, and 72°C for 30 s and a final step of 10 min at 72°C. PCRs were performed in a Realplex Mastercycler instrument (Eppendorf, Madrid, Spain). The PCR product was then digested with the BanI enzyme (New England Biolabs, Hitchin, United Kingdom) in a 40-μl volume containing 20 μl of the PCR product and 40 U of BanI. The digestion products were separated on a 2% agarose gel. The clinical isolates were discriminated as C. parapsilosis sensu stricto, C. orthopsilosis, and C. metapsilosis according to the SADH restriction profile. Furthermore, the amplified fragments of the C. orthopsilosis strains were sequenced to exclude the hypothesis that a point mutation exists in the restriction site of C. parapsilosis sensu stricto.

Antifungals.

Standard antifungal powder of FLU (Pfizer, Groton, CT), VOR (Pfizer), POS (Schering-Plough, Kenilworth, NJ), AMB (Bristol-Meyers Squibb, New York, NY), CAS (Merck, Rahway, NJ), and ANI (Pfizer) were obtained from the respective manufacturers. Stock solutions of FLU and CAS were initially prepared in distilled water; VOR, AMB, POS, and ANI were prepared in dimethyl sulfoxide. All antifungals were kept frozen at −70°C until use. The antifungal agents were then diluted with RPMI 1640 medium (Sigma) and buffered to pH 7.0 with 0.165 M morpholinepropanesulfonic acid buffer (Sigma).

Antifungal susceptibility testing.

The MICs of each antifungal drug were determined according to the M27-A3 protocol and the M27-S3 supplement of the Clinical and Laboratory Standards Institute (CLSI) (4, 5). The MICs were registered after 24 h of incubation for the echinocandins and after 48 h for all the azoles and AMB. The susceptibility breakpoints were those of the CLSI (4). For FLU, the MIC for susceptibility was ≤8 μg/ml, the MIC for susceptible-dose dependent was 16 to 32 μg/ml, and the MIC for resistance was ≥64 μg/ml; for VOR, the MIC for susceptibility was ≤1 μg/ml and the MIC for resistance was ≥4 μg/ml. The interpretive criteria for the echinocandins (CAS and ANI) were based upon the MIC-2 (the MIC-2 is the lowest concentration at which a visual reduction in growth of approximately 50% was registered according to the M27-A3 protocol of the CLSI [4]), and susceptibility was considered an MIC of ≤2 μg/ml. Standard MIC₅₀s and MIC₉₀s were calculated as MIC-2's. For POS and AMB, strains inhibited by ≤1 μg/ml of each antifungal were considered to be susceptible. The C. parapsilosis ATCC 22019 type strain was used for quality control of antifungal susceptibility testing.

RESULTS

Reidentification of C. parapsilosis.

Amplification of the SADH gene fragment was performed by using the genomic DNA of the isolates as the template. Isolates displaying a fragment with a size other than 716 bp were excluded from this study. Six strains (3.4%), corresponding to four blood isolates and two mucosal isolates, were not confirmed to belong to C. parapsilosis groups (Fig. 1a, lane 3). For the C. parapsilosis groups, a fragment of the SADH gene (716 bp) was amplified by PCR (Fig. 1a, lanes 1, 2, 4, 5, and 6). According to the BanI restriction pattern of the fragment described above, isolates were molecularly discriminated as C. parapsilosis sensu stricto, C. orthopsilosis, and C. metapsilosis (Fig. 1b) as follows: C. parapsilosis sensu stricto has one BanI restriction site (at position 196), C. orthopsilosis has no restriction site, and C. metapsilosis has three BanI restriction sites (at positions 96, 469, and 529). The SADH fragments of the previously identified C. orthopsilosis isolates were sequenced; all four sequenced isolates were in fact C. orthopsilosis. Therefore, the possibility of misidentification due to a possible point mutation in C. parapsilosis sensu stricto was excluded (data not shown).

FIG. 1.

Discrimination of C. parapsilosis (group I), C. orthopsilosis, and C. metapsilosis isolates according to their SADH gene restriction profiles. (a) Amplification of a 716-bp SADH gene fragment from the genomic DNA of isolates provided genetic confirmation of the C. parapsilosis groups (lanes 1, 2, 4, 5, and 6); the amplification of smaller SADH gene fragments corresponds to misidentified C. parapsilosis species (lane 3). (b) Representative SADH gene restriction profile for C. metapsilosis (lane 1), C. orthopsilosis (lane 2), and C. parapsilosis (lane 3) type strains. Lanes M, 100-bp ladder.

Incidence and distribution of C. parapsilosis, C. orthopsilosis, and C. metapsilosis.

The incidence and distribution of the three species are detailed in Table 1; 160 isolates (91.4%) were identified as C. parapsilosis sensu stricto, 4 (2.3%) were identified as C. orthopsilosis, and 5 (2.9%) were identified as C. metapsilosis; 6 isolates corresponded to other Candida species. C. orthopsilosis and C. metapsilosis were found in biological products other than blood; C. parapsilosis sensu stricto species were found in all types of biological samples tested. All eight environmental strains were identified as C. parapsilosis sensu stricto.

Antifungal susceptibility profile.

The profiles of susceptibility to the azoles and AMB are summarized in Table 2. The vast majority of the C. parapsilosis sensu stricto isolates (94.4%) were susceptible to FLU; 8 isolates (5%) were susceptible-dose dependent to FLU, and a single (0.6%) isolate was resistant to FLU. No FLU-resistant or FLU-susceptible-dose-dependent strains were detected among the C. orthopsilosis and C. metapsilosis isolates; the highest MIC was 4 μg/ml. A single isolate of C. parapsilosis sensu stricto was resistant to VOR and POS. All isolates of C. orthopsilosis and C. metapsilosis were susceptible to VOR and POS. The AMB MICs were equal to or less than 1 μg/ml for all isolates tested, with the exception of two C. parapsilosis sensu stricto isolates, which were resistant. The AMB MICs for C. orthopsilosis and C. metapsilosis were lower than the AMB MICs for C. parapsilosis sensu stricto. The patterns of susceptibility to the echinocandins are detailed in Table 3. A high number of C. parapsilosis sensu stricto isolates nonsusceptible to CAS (38%), as well as a wide range of MICs (4 to 32 μg/ml), were found, while all C. orthopsilosis and C. metapsilosis isolates were susceptible to CAS. Similar findings were registered for ANI: all C. orthopsilosis and C. metapsilosis isolates were susceptible to ANI, while 29% of the C. parapsilosis sensu stricto isolates were nonsusceptible to ANI but the ANI MICs (4 to 16 μg/ml) were less than those of CAS. For C. parapsilosis sensu strict, the standard MIC₅₀ and MIC₉₀ of both echinocandins were 2 and 4 μg/ml, respectively. For C. orthopsilosis, the standard MIC₅₀ of CAS was 2 μg/ml and that of ANI was 1 μg/ml. For C. metapsilosis, the standard MIC₅₀s of both echinocandins were 1 μg/ml. The MIC₉₀ was not calculated because the number of C. orthopsilosis and C. metapsilosis isolates was less than 10.

TABLE 2.

Susceptibilities of Candida parapsilosis (group I), C. orthopsilosis, and C. metapsilosis to the tested azoles and AMB

Species (no. of isolates)	Antifungal	MIC range (μg/ml)	% of isolatesa
			S	S-DD
C. parapsilosis (160)	FLU	0.125-64	94.4	5
	VOR	≤0.015-8	99.4
	POS	≤0.03-2	99.4
	AMB	0.06-2	98.8
C. orthopsilosis (4)	FLU	0.5-1	100	0
	VOR	≤0.015-0.03	100
	POS	≤0.03	100
	AMB	0.5	100
C. metapsilosis (5)	FLU	1-4	100	0
	VOR	≤0.015-0.125	100
	POS	<0.03-0.125	100
	AMB	0.125-1	100

^aS, susceptible; S-DD, susceptible-dose dependent.

TABLE 3.

Susceptibilities of Candida parapsilosis, C. orthopsilosis, and C. metapsilosis to echinocandins

Species (no. of isolates)	Antifungal	Phenotype (% of isolates)a	MIC-2 range (μg/ml)
C. parapsilosis (160)	CAS	S (62)	0.06-2
	CAS	NS (38)	4-32
	ANI	S (71)	0.125-2
	ANI	NS (29)	4-16
C. orthopsilosis (4)	CAS	S (100)	1-2
	ANI	S (100)	1
C. metapsilosis (5)	CAS	S (100)	1-2
	ANI	S (100)	0.5-1

^aS, susceptible (MIC ≤ 2μg/ml); NS, nonsusceptible.

DISCUSSION

Three distinct groups of C. parapsilosis (groups I, II, and III) have previously been recognized on a molecular basis. Subsequent studies focusing on the levels of heterozygosity and on the linear and circular forms of the mitochondrial genome in C. parapsilosis support the existence of these distinct species (43). In 2005, after sequencing analysis of several genes, groups II and III were reclassified as two new species separate from C. parapsilosis and were named C. orthopsilosis and C. metapsilosis, respectively (47).

After molecular analysis of a large set of C. parapsilosis isolates from a national university hospital, 91.4% were found to be C. parapsilosis sensu stricto, while 2.3% and 2.9% corresponded to C. orthopsilosis and C. metapsilosis, respectively. Discrepancies between the results of biochemical identification and molecular analysis were found since six isolates (3.4%) corresponded to other distinct Candida species. This fact highlights the relevance of complementing traditional biochemistry-based identification methods with DNA-based molecular assays such as PCR, since such tests provide higher discriminatory powers. These results improve the knowledge about the national distribution of C. parapsilosis sensu stricto, C. orthopsilosis, and C. metapsilosis. Lockhart et al. (23) identified among 67 Portuguese yeast isolates 1 C. orthopsilosis isolate and not a single C. metapsilosis isolate. The values found in our study are in accordance with data reported from international studies, namely, 91.3% C. parapsilosis sensu stricto isolates, 6.1% C. orthopsilosis isolates, and 1.8% C. metapsilosis isolates (23).

This study reinforces the assumption that C. parapsilosis sensu stricto is predominant among clinical isolates, namely, isolates from patients with hematogenous infections. Interestingly, C. orthopsilosis and C. metapsilosis isolates were obtained from other biological products, like urine, respiratory tract samples, and mucosal surfaces. The higher prevalence of C. parapsilosis sensu stricto may be related to its ubiquitous nature, since it is commonly isolated from different environmental sources. Those sources may represent potential routes for nosocomial transmission to patients (16, 21) and/or may be related to a larger set of virulence attributes compared to the amounts of virulence attributes of the other two species (7, 10, 42). However, the last assumption still needs to be fully confirmed. In fact, the strains isolated from the hospital air and patient food corresponded to C. parapsilosis sensu stricto, which supports the first assumption.

We evaluated the in vitro activities of FLU, VOR, POS, CAS, ANI, and AMB against C. parapsilosis sensu stricto, C. orthopsilosis, and C. metapsilosis. The establishment of susceptibility profiles provides information relevant to the development of recommendations for prophylactic, empirical, or targeted antifungal therapeutic strategies (17, 30, 35), particularly at the local level.

Valid statistical comparisons of the antifungal susceptibility patterns of the three species were seriously compromised due to the reduced number of strains of C. orthopsilosis and C. metapsilosis tested. Although the MICs of the echinocandins were high for all three species, the C. parapsilosis sensu stricto isolates were less susceptible to CAS and ANI in vitro than the C. orthopsilosis and C. metapsilosis isolates. Previous studies by our team (6) had already described high CAS MICs (for all C. parapsilosis isolates, the CAS MICs were ≥2 μg/ml). The profiles of susceptibility to both echinocandins (CAS and ANI) of the C. parapsilosis isolates from our hospital are quite different from the ones described in other surveys, like the study of Pfaller et al. (37). This finding deserves further attention. It has repeatedly been shown that echinocandin MICs are consistently higher for the C. parapsilosis group than for C. albicans or C. glabrata when they are assessed by broth microdilution methods (6, 26, 33, 36, 37, 40), thus suggesting that the C. parapsilosis group is less susceptible to this class of antifungal agents. The reduced susceptibility of C. parapsilosis to echinocandins (ANI, CAS, and micafungin) has been related to polymorphisms in the FSK1 gene (27, 28, 29). van Asbeck et al. (48) suggested that interspecies differences in CAS susceptibility result from the combination of structural differences in the cell wall components, a reduced affinity for the glucan synthase protein complex, or variations in the regulatory network of this complex. Azole resistance was detected only among C. parapsilosis sensu stricto isolates and not among C. orthopsilosis or C. metapsilosis isolates. The differences in FLU susceptibility among the three species reflect the distinct affinity of azoles for the key ergosterol-synthesizing enzyme, 14-demethylase, or other enzymes involved in this metabolic pathway (48). Only two C. parapsilosis sensu stricto isolates were resistant to AMB. The levels of nonsusceptibility were similar between isolates from different biological products (data not shown).

We detected a multiechinocandin-, multiazole-resistant phenotype of C. parapsilosis sensu stricto, as reported by Moudgal et al. (24) and Vazquez and colleagues (12). The development and subsequent nosocomial expansion of echinocandin- and azole-resistant C. parapsilosis sensu stricto have important clinical implications. Continuous monitoring for the emergence of this multidrug-resistant phenotype of C. parapsilosis sensu stricto represents an important component of the effective prevention of nosocomial infections.

Although FLU showed good activity against C. parapsilosis sensu stricto, C. orthopsilosis, and C. metapsilosis, local routine monitoring of the susceptibility profiles of these organisms may be highly advisable. The reduced echinocandin susceptibility profile reported in this study also represents a problematic issue, at least in Portugal. The combination of decreased susceptibility to FLU and the ability to form extensive biofilms on central venous catheters and other medical indwelling devices may become problematic (8, 18, 19, 21). In over 50% of the episodes of C. parapsilosis fungemia, the source is a vascular catheter, and such infections commonly develop in patients who had previously received antifungal treatment (1). Thus, an adequate response to FLU or to any other antifungal may not be fully obtained. The administration of the antifungal should be coupled with the prompt removal of the catheter to ensure an optimal response (27).

The data reported here represent the most comprehensive data from a study of the incidence and antifungal susceptibility profiles of C. parapsilosis sensu stricto, C. orthopsilosis, and C. metapsilosis conducted to date in Portugal.

We demonstrated the low incidence of C. orthopsilosis and C. metapsilosis among clinical isolates, especially isolates from blood cultures. The differences in antifungal susceptibility described here are not sufficiently distinct to warrant discrimination of the species in the clinical laboratory.