Identification and antifungal susceptibility of Candida species isolated from the urine of patients in a university hospital in Brazil.

The aim of this study was to identify Candida spp. isolated from candiduria episodes at a tertiary hospital in the Midwest region of Brazil, and to determine their susceptibility profiles to antifungal compounds. From May 2011 to April 2012, Candida spp. isolated from 106 adult patients with candiduria admitted to the University Hospital of the Federal University of Mato Grosso do Sul were evaluated. Both, species identification and susceptibility testing with fluconazole-FLC, voriconazole-VRC, and amphotericin B-AmB were carried out using the Vitek 2. To discriminate species of the C. parapsilosis complex, a RAPD-PCR technique using the RPO2 primer was performed. From the total of 106 isolates, 42 (39.6%) C. albicans and 64 (60.4%) Candida non-albicans (CNA) - 33 C. tropicalis, 18 C. glabrata, 5 C. krusei, 4 C. parapsilosis sensu stricto, 2 C. kefyr, 1 C. lusitaniae, and 1 C. guilliermondii were identified. All isolates were susceptible to AmB and VRC, whereas all C. glabrata isolates presented either resistance (5.6%) or dose-dependent susceptibility (94.4%) to FLC. The study of Candida spp. and their resistance profiles may help in tailoring more efficient therapeutic strategies for candiduria.

INTRODUCTION

The isolation of Candida spp. from urine cultures may indicate colonization or urinary tract infection (candiduria), but it may also be a sign of severe systemic candidiasis or candidemia - . Candida spp. can reach the urinary tract via the ascending route, from the urethra to the bladder, or by hematogenous spread, as Candida spp. is filtered by the kidneys and excreted in the urine.

The ability to form biofilm and the production of hydrolytic enzymes, which disrupt cell membranes, facilitate the spread of yeasts in the host, causing infection - . Lesions in the renal pelvis, tubules, and ureters, and formation of a “fungus ball”, which blocks and causes injury to the urinary system, are the main complications associated with candiduria, while development of candidemia is unusual , , .

The source of candidiasis in humans is mostly endogenous, as Candida spp. are commensals in the digestive tract of a vast range of healthy people. Some conditions allow these commensal yeasts to become opportunistic, resulting in candidiasis in different sites of the body. Previous studies have shown that advanced age, female sex, urinary tract abnormalities, drainage catheters used in hospitalized patients, diabetes, malignancies, as well as the use of broad-spectrum antibiotics, corticosteroids and immunosuppressive agents are high-risk conditions for developing urinary tract infections , , . Candiduria is frequently found in critically ill patients, in those who present with metabolic and immunosuppressive diseases, as well as in patients undergoing surgical procedures or facing long hospital stays , .

A study conducted in Spain revealed that 22% of critically ill patients hospitalized for more than seven days in intensive care units (ICUs) developed candiduria, and 10-15% urinary tract infections in the ICU were caused by the genus Candida . In the United States, from years 1995-2001, it was estimated that there was an average of 25,000 cases of candiduria per year. Around one third of hospitalized patients with positive urine cultures for Candida spp were from the ICU, where the use of urinary catheters is frequent , . Moreover, it was found that a patient in a given ICU who is exposed to four different antibiotics has about a 35% risk of developing candidemia, and if Candida is isolated from another site, such as in urine, the risk increases to 80% .

In severe systemic episodes of Candida infections, management of clinical conditions and enhanced patient survival depend on rapid interventions. Thus, correct identification of the pathogen and administration of specific antifungal therapies are crucial for patient recovery , , .

The etiology of candiduria varies according to the geographic region, study period, and type of hospital. Candida albicans is the most prevalent candiduria agent; however, Candida non-albicans (CNA) species have been reported worldwide , .

The emergence of CNA species with reduced susceptibility or intrinsic resistance to antifungal compounds, mainly the azoles, is a major problem due to the increasing use of fluconazole (FLC) in candidiasis therapy , . In addition, studies conducted in ICU settings have reported that patients with candiduria have increased mortality rates when compared with similar patients without candiduria . Therefore, early and accurate species identification, as well as investigations of antifungal susceptibility profiles of isolates, is indispensable for determining appropriate therapies for treating episodes of recurrent candiduria and/or persistent candidemia related to candiduria.

Therefore, the monitoring of epidemiological data in hospitals across different regions is important for establishing efficient control measures. The aim of this study was to identify the various species of Candida isolated from candiduria episodes in a tertiary hospital in the Midwest region of Brazil, as well as to determine the in vitro susceptibility profiles of these species to various antifungal compounds.

PATIENTS AND METHODS

Patients

This study was conducted at the University Hospital, Federal University of Mato Grosso do Sul, Campo Grande, MS, from May 2011 to April 2012. It was approved by the institutional Ethics Committee in Research with Human Beings (process N° 86663, date of approval 08/30/2012). Adult female and male patients who were admitted to any hospital ward and presented with positive urine cultures for Candida spp. at least 1,000 CFU/mL were included in the study.

Isolation and phenotypic identification

Urine sample sediments were initially seeded in Chromogenic Candida Selective Agar (CHROMagar, DIFCO; Becton Dickinson France S.A., Le Pont de Claix, France) for identification of Candida species. After inoculation, plates were incubated for 24-48 h at 30 °C and colonies were identified based on the chromogenic reaction produced, according to the standard instruction from the manufacturers.

The confirmation of Candida species identification and the antifungal susceptibility was performed using the automated Vitek 2 system (bioMérieux, Marcy-l'Etoile, France). From each sample, a yeast suspension was prepared in sterile saline (0.9%), using the densitometer (Densichektm Vitek 2). Yeast suspensions were added and incubated at 35.5 °C for 18 to 24 h. The equipment classified the final reading and biochemical differentiation reactions as excellent.

Differentiation of the C. parapsilosis complex species was performed by random amplified polymorphic DNA-polymerase chain reaction (RAPD-PCR) with the RPO2 primer (5′-GCGATCCCCA-3′), according to Rycovska et al. .

C. parapsilosis isolates were re-suspended in saline with turbidity equivalent to 0.5 in the McFarland scale, centrifuged at 5000 × g and the pellets were subjected to DNA extraction. The final volume of reactions was 50 µL containing 20 ng of genomic DNA of each sample, 1.5 U of GoTaq DNA polymerase (Promega Co., USA), 10 µL of 5x concentrated buffer (1.5 mM MgCl2, 200 µM dNTP and 1 µM of primer RPO2). Amplifications were performed in a Veriti Thermocycler (Life Technologies, USA). After a first cycle of denaturation for 2 min at 94 °C, 35 cycles of denaturation at 94 °C for 1 min, annealing at 36 °C for 1 min, and extension at 72 °C for 75 s were performed, with a final extension step of 10 min at 72 °C.

Reference strains ATCC (American Type Culture Collection) were used as positive controls [C. parapsilosis (ATCC 22019), C. orthopsilosis (ATCC 96141) and C. metapsilosis (ATCC 96143)]. Amplified DNA products were separated by electrophoresis in a 1.5% agarose gel (Invitrogen, USA) with Tris-EDTA pH 8.0 buffer, stained with GelRed (Biotium, USA).

Antifungal susceptibility testing

Determination of the minimum inhibitory concentration (MIC) was performed using the automated Vitek 2 (bioMérieux) system with the following antifungal drugs: amphotericin B (AmB), FLC, and voriconazole (VRC).

The minimum inhibitory concentration (MIC) breakpoints used for fluconazole and voriconazole were those suggested by Laboratory Standards Institute's M27-S4 document , for C. albicans, C. tropicalis and C. parapsilosis ≤ 2 µg/mL were susceptible, 4 µg/mL were susceptible dose-dependent (SDD) and ≥ 8 µg/mL were resistant. For C. glabrata isolates, ≤ 32 µg/mL were SDD and ≥ 64 µg/mL were resistant. Due to a lack of consensus about the definition of MIC breakpoints for amphotericin B, we considered values suggested by previous publications , MICs ≤ 1 µ g/mL as susceptible and MICs ≥ 2 µ g/mL as resistant.

From the growth on Sabouraud dextrose agar for 24 h, a yeast suspension with turbidity equivalent to 0.5 in the MacFarland was prepared to test the performance of the broth microdilution test.

The MIC values of FLC (Pfizer) were confirmed by the microdilution broth technique using the RPMI-1640 culture medium (Vitrocell). The stock solution of FLC (5120 µg/mL) was thawed at room temperature and diluted in a twofold dilution ratio.

Flat-bottom microdilution plates containing 100 µL of the twofold serial dilutions of FLC were inoculated with 100 µL of Candida suspensions. The final inoculum concentration was 0.5 to 2.5×10 CFU/mL and the final ranges of the tested FLC dilutions were 0.25 to 128 µg/mL. Microdilution plates were incubated at 37 °C for 24 h. Each sample was tested in triplicate.

The quality control strains C. krusei (ATCC 6258) and C. parapsilosis (ATCC 22019) were provided by the Central Laboratory of Mato Grosso do Sul (LACEN-MS).

Statistical analysis

Statistical analysis was performed using the Minitab software version 14 (Minitab Inc., State College, PA, USA). To determine the association between variables, the chisquared (x 2) or the Fisher's exact test were used; a p-value <0.05 was considered significant.

RESULTS

Patients’ epidemiological characteristics

During the study period, 106 adult patients were diagnosed with candiduria; 58 (54.7%) patients were female and 48 (45.3%) were male. The patients’ ages ranged from 19 to 93 years, with a median of 68 years. Twelve (11.3%) patients were aged between 18 and 30 years, 25 (23.6%) between 31 and 60 years, and 69 (65.1%) were older than 60 years. Regarding the hospital setting, 33 patients (31.1%) were admitted to the ICU, 33 (31.1%) to emergency medical care, 20 (19.0%) to internal medical clinical wards, eight (7.5%) to the coronary unit, five (4.7%) to the surgical units, four (3.8%) to the maternity ward, two (1.9%) to the infectious diseases ward, and one patient (0.9%) to the renal unit.

Candida species

The CNA species predominated, occurring in 64 (60.4%) patients. There was no identification of more than one species in the same patient. CNA was more likely to be observed in patients older than 60 years when compared to those younger than 60 years (68.1% versus 48.9%, respectively; P=0.026). C. albicans, C. tropicalis, and C. glabrata species were the most frequently isolated, accounting for 87.7% of all infections. The frequency of C. albicans (39.6%) and C. tropicalis (31.1%) cases did not differ from one another, and they were more frequently found than C. glabrata (17.0%) ([C. albicans = C. tropicalis] > C. glabrata; P<0.05).

The four samples phenotypically identified as C. parapsilosis (sensu lato) weremolecularly characterized as C. parapsilosis (sensu stricto) after performing the RAPD assay (Figure 1).

Figure 1

Agarose gel showing the RAPD products obtained from Candida parapsilosis complex isolates and ATCC strains. Lane L - 100 bp molecular weight marker (GE Healthcare); lanes 1-4 - Candida parapsilosis complex isolates, all identified as Candida parapsilosis sensu stricto; lane 5 - C. parapsilosis sensu stricto (ATCC 22019); lane 6 - C. orthopsilosis (ATCC 96141); lane 7 - C. metapsilosis (ATCC 96143). RAPD= random amplified polymorphic DNA; ATCC=American Type Culture Collection

Antifungal susceptibility

All isolates showed MIC = 1 µg/mL for AmB and an MIC = 0.012 µg/mL for VRC; therefore, they were considered susceptible. Regarding FLC, C. krusei and all C. glabrata isolates showed MIC values that were compatible with the SDD or R categories (Table 1). C. kefyr, C. guilliermondii, and C. lusitaniae were 100% susceptible (MIC = 2 µg/mL).

Table 1

Candida species and antifungal susceptibility in urine isolates (n=106)

Candida species (n; %)	Amphotericin B (%)			Fluconazole (%)			Voriconazole (%)
	R	SDD	S*	R	SDD	S**	R	SDD	S***
Candida albicans (42; 39.6)	0	0	100	0	0	100	0	0	100
Candida tropicalis (33; 31.1)	0	0	100	0	0	100	0	0	100
Candida glabrata (18; 17.0)	0	0	100	5.6(a)	94.4 (b)	0	0	0	100
Candida krusei (5; 4.7)	0	0	100	100	0	0	0	0	100
Candida parapsilosis (4; 3.8)	0	0	100	0	0	100	0	0	100
Candida kefyr (2; 1.9)	0	0	100	0	0	100	0	0	100
Candida guilliermondii (1; 0.9)	0	0	100	0	0	100	0	0	100
Candida lusitaniae (1; 0.9)	0	0	100	0	0	100	0	0	100

R = resistant; SDD = susceptible dose dependent; S = susceptible; MIC = minimal inhibitory concentration.

(MIC ≤ 1 μg/mL);

(MIC ≤ 2 μg/mL),

(MIC ≤ 0.12 μg/mL);

(MIC > 64 μg/mL);

(MIC ≤ 32 μg/mL).

We did not observe associations among any of the studied variables with FLC resistance or SDD to FLC (Table 2). However, the mean number of C. glabrata isolated from ICU patients was higher than the numbers obtained from patients admitted to all the other hospital wards (22.7% versus 7.5%, respectively; P=0.043).

Table 2

Demographic variables and susceptibility to fluconazole (n=106)

Variables		Resistant or SDD (n=23)	Susceptible (n=83)	p-value
		n (%)	n (%)
Inpatient unit	ER/ICU	18 (78.3)	48 (57.8)	0.074
	Other	5 (21.7)	35 (42.2)
Age ≥60 years	No	6 (26.1)	31 (37.3)	0.316
	Yes	17 (73.9)	52 (62.7)
Sex	Female	14 (60.9)	44 (53.0)	0.503
	Male	9 (39.1)	39 (47.0)
CFU/mL	>100.000	22 (95.7)	70 (84.3)	0.294*
	≤100.000	1 (4.3)	13 (15.7)
Leukocytes >5/field	No	2 (9.1)	13 (15.7)	0.732
	Yes	20 (90.9)	70 (84.3)

Fisher's exact test.

N= number of patients; SDD= susceptible dose-dependent; ER= emergency room; ICU= intensive care unit; CFU= colony-forming units.

DISCUSSION

The correct identification of yeasts isolated from the urine cultures of hospitalized patients, as well as the yeasts’ antifungal susceptibility profiles, are relevant for determining appropriate treatments and urinary tract infection management , . This study revealed a greater proportion of CNA isolates from urine samples, which points to the emergence of strains resistant to FLC in our hospital.

There was a higher frequency of candiduria in women. Previous studies have shown that up to 30% of healthy women may experience persistent vulvovaginal colonization by Candida spp. This colonization, facilitated by the female anatomy, can ascend to the bladder and kidneys, causing urinary tract infections - .

Similar to other reports, most patients in our study were over 60 years of age. In general, older adults tend to exhibit natural modifications of the immune system, resulting in longer hospital stays in critical care units and the need of urinary catheters , , - .

Most patients in our study that tested positive for candiduria were hospitalized in ICU or emergency units, where most of them are generally submitted to invasive procedures and monitored with bladder catheters. These conditions may lead to the rupture of anatomical barriers, resulting in changes of the native host microbiota , , , .

It is generally accepted that urine cultures with at least 100,000 CFU/mL, with or without pyuria, can serve as an alert to a possible urinary tract infection, which is initially presumed as a bacterial infection , , . There is no consensus among the authors regarding the cutoff point for interpreting urine cultures containing fungi; however, a minimum count of 1,000 CFU/mL in symptomatic patients has been considered a positive infection , . The presence of candiduria in critically ill patients has been regarded as an indicator of invasive candidiasis .

The increased prevalence of CNA as a cause of candidiasis in hospital settings has been observed around the world at any site of infection, including the urinary tract , , . The importance of this fact is that CNA species are likely to be resistant to antifungal agents, which reinforces the need to identify various Candida species, as well as to assess their susceptibility profile to antifungal agents in hospitalized patients , , .

C. albicans, C. tropicalis, and C. glabrata were the main isolated species in our investigation. The comparison of the Candida species prevalence found in the present investigation with other eight Brazilian studies , - on candiduria revealed a predominance of C. albicans in seven of them , - and, in the second position, C. tropicalis in three studies , , , the same behavior we observed in our study. In one study C. tropicalis was the most frequent . The prevalence of C. parapsilosis, which was 4.7% in our study, in the other studies ranged from zero to 17.4% , - . The evaluation of studies carried out in different regions of the country revealed a great similarity in the prevalence of Candida species among studies performed in the Midwest region , , and those from the Southeast among themselves , , . C. albicans is the most frequently found species in the digestive tracts of healthy people, and it has greater pathogenic mechanisms when compared to other Candida spp. , . Emerging C. tropicalis causes urinary tract infections primarily in patients presenting chronic degenerative diseases and/or trauma , , .

The prevalence of C. glabrata in our study (17%) was higher than the one found in other earlier Brazilian reports, which ranged from 0 to 12.5 % , - , . Our data corroborate previous findings, which have shown that C. glabrata has emerged in tertiary hospitals in recent years, both in Brazil and in other countries , .

The C. parapsilosis complex has emerged as a primary agent of urinary tract infections in hospitalized patients due to its capacity to form biofilm , . Not all members of the C. parapsilosis complex are virulent; among them, C. metapsilosis appears to present the lowest virulence, but this evidence is still limited . This reinforces the importance of differentiating C. parapsilosis in clinical studies, including isolates from other anatomic sites . In our study, the four (3.8%) isolates that belong to the C. parapsilosis complex were identified as C. parapsilosis (sensu stricto).

Several studies have shown the emergence of antifungal resistance, especially FLC resistance in C. albicans ; however, all of the C. albicans isolates were susceptible to AmB, FLC, and VRC. Moreover, the non-albicans species were also susceptible to the tested drugs, except C. krusei, which is intrinsically resistant to FLC. Furthermore, all isolates of C. glabrata were also resistant or SDD to this azole. Resistance to FLC or AmB has already been described in C. glabrata and C. tropicalis .

In Brazil, five studies from different regions evaluated the susceptibility to FLC of Candida species isolated from urine , - , . Using the broth microdilution method to evaluate C. albicans isolates, one study showed no resistance to FLC , as did ours, and another showed 15.0% of resistance . Using the disk diffusion, 7.5% and 75% of resistance to FLC were observed. The isolates of C. tropicalis were 100% sensitive to FLC, using the broth microdilution method, in two studies , and 12.5% of resistance in one , confirming our findings. Using the disk diffusion method, high rates of FLC resistance were observed , . Regarding C. parapsilosis, there was no resistance to FLC , as in the present study. We have also observed variations in the susceptibility profile of Candida species to the most common antifungal agents of hospital use, especially to FLC, depending on the Candida species, methodology used, number of cases studied and year of publication.

The increasing prevalence rates of these species, which feature high MIC values for FLC, have already been reported in North America and Europe , , , . Some reports have demonstrated that this change in the susceptibility profile to FLC may be a consequence of the indiscriminate use of this antifungal agent in nosocomial settings , , . In this study, the patients’ previous use of FLC was not investigated. In fact, all C. parapsilosis (sensu stricto) isolates were susceptible to AmB, FLC, and VRC, corroborating the findings of other authors , .

In serious and systemic infections, the patient's survival depends on the early intervention and specific antifungal therapy. Delays to initiate an effective antifungal therapy among hospitalized patients is associated with high mortality . Candiduria in seriously ill patients should be carefully evaluated, as this medical condition may be the only indicator of invasive candidiasis. Its detection increases the chances of successful treatment and enhances the patients’ survival , .