An epidemiologic analysis of Candida spp. urinary infections in intensive care unit

Uma análise epidemiológica das infecções urinárias por Candida spp. em unidade de terapia intensiva

Un análisis epidemiológico de las infecciones urinarias por Candida spp. en la unidad de cuidados intensivos

Esta obra está bajo una Licencia Creative Commons Atribución 4.0 Internacional.

Candida species are members of microbioata at various sites in the human body¹. However, they are capable of colonizing mucocutaneous tissues via attaching to the superficial mucosal cells and are accepted as opportunistic pathogens². The finding of Candida species in urine is a usual clinical situation and is called candiduria³. In urine cultures, Candida albicans and non-albicans species are predominantly isolated among fungi, and in the last two decades, there was a remarkable increase in urinary tract infections (UTIs) caused by opportunistic fungi, especially among hospitalized patients which create considerable public health predicaments.^1,4 Also, there is an increase in the frequency of UTIs caused by fungi, especially Candida species in critically ill patients⁵.

Candida species may cause UTIs via both antegrade pathway by entering the upper urinary tract from the systemic circulation and retrograde pathway by ascending the urinary tract from a colonization site around the urethra⁶. Several reports indicate that Candida species are responsible for at least 10-15% of nosocomial UTIs.^3,7UTIs caused by Candida is an emergent problematic issue for immunocompromised and critically ill patients, among the hospitalized patients, candiduria is a frequent finding particularly in intensive care units (ICUs) and in adult surgical ICUs candiduria more frequently go along with UTIs.^3,7-10

There are well defined independent risk factors for candiduria and Candida UTIs including; age >65 years, female sex, prolonged hospitalization, ICU admission, diabetes mellitus, disturbance in microbiome caused by broad-spectrum antimicrobials, female sex, total parenteral nutrition, bladder dysfunction, congenital abnormalities of the urinary tract, renal transplantation, urinary stasis, nephrolithiasis, concomitant bacteriuria, genitourinary tuberculosis, neutropenia, urinary tract instrumentation, chronic renal failure, mechanical ventilation and immunosuppressive therapy.^2,3,5,8,9-11

Even though C. albicans is often reported as the predominant species responsible for UTIs, all common Candida species can cause UTIs, and non-albicans species emerge with better adaptation to the urinary tract system because many studies worldwide stating that half of the candiduria isolates are non-albicans.^7-9,12 Frequent use of antifungal prophylaxis and treatment results in infections with non-albicans species showing resistance to antifungals.¹³An increase in non-albicans species appears as a significant problem due to decreased susceptibility of non-albicans species to antifungals, which may result in complexities or failures in the management of UTIs.¹⁴This study aimed to determine the epidemiological, clinical, and mycological characteristics of Candida urinary infections in intensive care unit (ICU) patients and antifungal susceptibilities of Candida species.

In our study, urine cultures of ICU patients were investigated duplicate samples and patients who were taking prior antifungal therapy were excluded. A total of 394 nonrepetitive patients with clinical suspicion of UTIs from anesthesiology and reanimation ICU (50.7%) and internal diseases ICU (49.3%) were evaluated. Since urinary catheterization is a standard practice in ICUs, all patients included in the study had an indwelling urinary catheter. The demographic information and laboratory findings of the patients including age, sex, length of stay, existence of concomitant bacteriuria, existence of concurrent candidemia and average days for detection of candiduria after admittance to th ICU were recorded. Urine samples were transferred with sterile urine containers and inoculated onto Sabouraud Dextrose Agar (SDA; Salubris, Turkey) medium. After 24-48 hours of incubation at 25 °C and 37 °C, colonies appeared to grow as yeast, were morphologically examined by Gram staining. Candida strains that grew 10⁴ ≥ CFU/mL in urine cultures were accepted as candiduria and included in our study.^15,16

For the identification of Candida species an automated identification system Phoenix (Becton Dickinson, Germany) and chromogenic agar (Chromagar; Salubris, Turkey), as well as classical methods like germ-tube formation was used. Color change of colonies at chromogenic agar was observed after 48 hours of incubation; C. albicans was observed as green, C. tropicalis as blue, C. glabrata, and C. kefyr as pink-purple

The susceptibilities of the Candida strains to amphotericin B, itraconazole, fluconazole, voriconazole, flucytosine, and caspofungin were investigated using the reference broth microdilution method in the Clinical and Laboratory Standards Institute (CLSI) M27-A3, M27-S3, and M27-S4.^17-19 For broth microdilution susceptibility experiments, caspofungin (Sigma, China), amphotericin B (Sigma, Israel), fluconazole (Sigma, USA), flucytosine (Sigma, UK), voriconazole (Sigma, USA), and itraconazole (Sigma, USA) were used as antifungals. Distilled water was used for fluconazole and flucytosine, DMSO (dimethyl sulfoxide) (Merck, USA) was used for water-insoluble caspofungin, amphotericin B, voriconazole, and itraconazole as a solvent. Stock solutions were prepared at 1280 μg / mL for fluconazole, 1600 μg / mL for amphotericin B, 1600 μg / mL for voriconazole, 1600 μg / mL for itraconazole, 1600 μg / mL for flucytosine, and 640 μg / mL for caspofungin. Prepared antifungal stock solutions were passed through a membrane filter, divided into 1 mL volumes, placed in sterile Eppendorf tubes, and stored at -80 ° C until use. Amphotericin B was coated, protected from light.

An inoculum concentration adjusted to 1.5×10³±1.0×10³ cells/ml with using RPMI 1640 medium (Sigma, USA), were tested with two-folds increasing antifungal concentrations of amphotericin B (0.0313-16 μg/mL), flucytosine (0.125-64 μg/mL), itraconazole (0.0313-16 μg/mL), fluconazole (0.125-64 μg/mL), voriconazole (0.0313 - 16 μg/mL), caspofungin (0.015-8 μg/mL) by broth microdilution method. After incubation at 35 °C for 48 h (24 hours for caspofungin), minimum inhibitory concentrations (MICs) were defined as the lowest concentration that inhibited visual fungal growth compared with the drug-free controls. C. parapsilosis ATCC 22019 and C. krusei ATCC 6258 strains were used for control.

Breakpoints for antifungal susceptibility were evaluated according to CLSI guidelines but, CLSI has not determined breakpoints for amphotericin.¹⁹ The isolates inhibited by amphotericin B at ≤1 µg /ml were considered susceptible, resistant isolates were defined as isolates with MIC >1 µg/ml. Also, since there is no new update in M27-S4 for C. kefyr, the values in M27-S3 were taken into consideration while evaluating the MIC breakpoints.^18,19Also for C. lusitaniae, since there is no update for M27-S4 fluconazole and voriconazole, the MIC breakpoints specified in M27-S3 were taken into consideration.

SPSS 16.0 (Statistical Package for Social Sciences) Package program was used for the analysis of the data obtained from the study. Mean, standard deviation, and percentage distributions were given as descriptive statistics. In addition, the Chi-Square test was used for the comparison of non-numerical variables. Mann Whitney-U test was used to compare binary variables with numerical data. The results obtained were evaluated at the 95% (P <0.05) significance level. In order to carry out this study, ethics committee approval was obtained from Gaziantep University Clinical Research Ethics Committee (Code:2016/298).

Urine samples from 394 ICU patients were examined; there was candiduria in 54 (13.7%) patients, candiduria and concomitant bacteriuria in 46 (11.6%) patients, bacteriuria in 69 (17.5%) patients, and 235 (59.6%) patients had negative urine cultures. A total of 100 Candida strains were evaluated. The distribution of the isolated urinary Candida strains were as, 54 C. albicans, 34 C. glabrata, 7 C. tropicalis, 2 C. kefyr, 2 C. lusitaniae, and 1 as C. parapsilosis. Concurrent candidemia was detected in 14 of patients with candiduria. The distribution of Candida species isolated from blood cultures was 10 (71.4%) C. albicans, 2 (14.2%) C. parapsilosis, 1 (7.1%) C. glabrata, and 1 (7.1%) C. lusitaniae.

Determined by broth microdilution; 91 Candida species were susceptible to caspofungin, 8 were moderately susceptible, and 1 was resistant, 23 Candida species were susceptible to fluconazole, 37 were dose-related susceptible, and 40 were resistant, 13 Candida species were susceptible to itraconazole, 18 were dose-dependent susceptible, and 69 were resistant. There were 17 (25.8%) voriconazole susceptible strains, 13 dose-dependent susceptible (19.7%) strains, and 36 (54.5%) resistant Candidastrains (C. glabrata not included due to CLSI statement: current data are insufficient to demonstrate a correlation between in vitro susceptibility testing and clinical outcome). Among the 100 Candida strains examined, resistant strains against amphotericin B (≥2 µg/mL) and flucytosine (≥32 µg/mL) were not detected. MIC range, MIC_50, and MIC₉₀ values of antifungal drugs for Candida species are given in Table 1. Also, detailed antifungal susceptibility results of different Candida species were given in Table 2.

Table 1
MIC ranges, MIC₅₀ and MIC₉₀ values of antifungals for different Candida species

Table 2
Antifungal susceptibility results of different Candida species

Abbreviations: S,susceptible; I,intermediate/susceptible-dose dependent; R,resistant.

†There is no CLSI defined breakpoints for C. kefyr and C. lusitaniae, breakpoints for C. tropicalis was applied

‡There is no CLSI defined voriconazole breakpoints for C.glabrata, breakpoints for C. krusei was applied

In our study, 60 (60%) of the patients with candiduria were female and 40 (40%) were male. A significant difference was found between the two groups in terms of sex distribution (p=0.046). When Candida species were analyzed according to sex, C. albicans was higher (53.7%) among males and non-albicans species were higher (58.3%) among females. Non-albicans species were isolated more frequently in females and a statistically significant difference was found (p=0.002). Of the 100 patients whose Candida strains were isolated, 9 were in the age range of 20-40, 16 were 41-60, and 75 were 61 and over. The length of stay of the patients in the ICUs was varying between 6 and 120 days (mean 33.8 days). After admittance to ICU, candiduria was detected (mean 9.7 days) within 1-9 days in 64 patients, 10-19 days in 23 patients, 20-29 days in 8 patients and >30 days in 22 patients. While the mortality rate was 41.1% among patients included in our study (n:349), the overall mortality rate among patients with candiduria was 69%, and among patients with both candiduria and candidemia was 92.8%. There was a significant difference in mortality rate between patients with candiduria and without candiduria (p <0.001). No significant difference was found between the causative agent of candiduria and the mortality rate.

The detection of candiduria manifests as a diagnostic and therapeutic challenge for all levels of health care settings and may be frustrating for physicians from primary care or infectious diseases, along with intensive medicine and surgery⁶. Urinary Candida may be related to a number of conditions ranging from sample contamination to UTIs, including invasive candidiasis, therefore, require detailed analysis.^1,6 Obtaining new urine samples and confirming whether candiduria persists, can usually help for differentiation of contamination from colonization or UTI². If there is growth in the second culture repeated, but the patient is asymptomatic, predisposing factors should be reviewed, the urinary catheter should be removed, and antibiotic therapy should be terminated⁵. Urinary tract imaging is recommended in patients with diabetes mellitus and patients with known urinary tract abnormalities, it may be guiding for appropriate treatment⁵.

In our study, female sex and advanced age were detected as risk factors for the development of candiduria. Although females are twice as likely to develop nosocomial candiduria when compared to males, possibly due to the anatomical differences of their genitals and vaginal colonization, females with candiduria was linked to a reduced risk of candidemia when compared to males.²⁰ Candida species are more common in the urine of the elderly, especially after broad-spectrum antibiotic treatments, advanced age, normal physical changes, and/or various metabolic disorders or neoplastic diseases that cause disruption of the mucosal and cutaneous barriers and make the person vulnerable to Candida infections.²¹Interestingly we did not find a correlation between long-stay in ICU and candiduria while 64% of our patients develop candiduria in 9 days after admittance to ICU, in a similar study from France, the time between admission to the ICU and the development of candiduria was reported as 17.2±1.1 days⁴.

There are no defined standard diagnostic criteria for diagnosis of Candida UTIs and their differention of from asymptomatic candiduria, and differentiation of upper from lower UTIs.²²Also, there is no consensus in diagnostic evaluation colony counts(CFU/ml) and urine collection technique for neonatal candiduria unlike in bacteriuria.¹¹Although the clinical significance of candiduria is still contradictory, various researchers suggest that colony counts greater than 10.–10.CFU/mL are more likely related to primary or disseminated candidiasis, rather than sample contamination or colonization.¹⁶

Candiduria frequency among ICU patients increased in recent years, especially among patients requiring urinary instrumentation or receiving broad-spectrum antibiotics and risk of occurence is as high as 22.89% in ICU patients.²³ The finding of Candida in the urinary samples is associated with higher mortality, particularly in ICU patients with accompanying comorbidities. However, higher mortality rates in patients with candiduria are not often directly attributable to invasive candidiasis. Nevertheless, candiduria may be an indicator of severe underlying diseases.7 In a clinical study in-hospital mortality was 48.8% in patients with candiduria compared to 36.6% in those without candiduria (p <0.001), they also found significant differences for ICU mortality (38.% vs. 28.1%, p <0.001).⁸ Researchers found, candiduria detected at any time in the surgical ICU was independently associated with mortality.²⁴Our study also revealed candiduria as an independent risk factor for mortality ( p <0.001). The incidence of concurrent candidemia is infrequent and has been encountered in 1-8% of patients with candiduria, even so, ICU patients constitute the high-risk group.¹⁰Our results showing 14% candiduria with concurrent candidemia in ICU patients also indicate physicians should be more alarmed about invasive candidiasis in critically ill patients with candiduria. Long hospital-stay and malignancy are predictors for developing candidemia in patients with candiduria; however, the patient characteristics linked to concomitant candidemia in the presence of candiduria remain unknown.²⁰

Management of candiduria is still contradictory because the finding of Candida spp. in urinary specimens may indicate asymptomatic infection, lower UTI, upper UTI with a potential for ascending pyelonephritis, renal candidiasis leading to invasive and disseminated candidiasis, which not only results in considerable morbidity and mortality but also prolonged hospitalization and growing cost². Candiduria may be an indicator of disseminated candidiasis in neutropenic, low birth weight infants, patients undergoing urological procedures, and renal transplant recipients patient groups⁵. Candiduria in critically ill patients whether symptomatic or not should initially be considered as a clue of disseminated candidiasis and antifungal drug prophylaxis appears to be warranted since the kidney is affected by disseminated candidiasis in 80% of patients². Detection of candiduria may be the only evidence that the patient has a serious infection. In these patients, systemic therapy with fluconazole or another azole derivative is recommended. An echinocandin, such as caspofungin, is selected if the patient has had recent exposure to fluconazole, which is the drug of choice⁵. In the treatment of cystitis and pyelonephritis, oral fluconazole is used in susceptible strains, and flucytosine and amphotericin B are used in those with fluconazole resistance. Bladder irrigation with amphotericin B may be beneficial in cystitis caused by fluconazole-resistant strains such as C. glabrata and C. krusei⁵. Voriconazole is stated as an effective antifungal that can be used in isolates resistant to fluconazole. However, a significant portion of fluconazole-resistant Candida isolates become resistant to voriconazole as well as itraconazole as a result of cross-resistance.¹⁷The most important features of caspofungin are that it is effective against azole and amphotericin B resistant Candida strains. Since there is no cross-resistance between azole antifungals, caspofungin can be a good option for Candida species resistant to azole antifungals¹.

Although C. albicans is the most prevalent species reported in urine culture, other species such as C. glabrata, C. parapsilosis, C. tropicalis, C. kefyr, C. lusitanae, C. guilhermondi, and C. dubliniensis can also be isolated¹.The distribution of causative agents of Candida UTIs is shifting, non-albicans species are detected in more than half of the urinary samples, which also bring along antifungal resistance issues⁷ . These non-albicans Candida may not only show better adaptation to the kidney and collecting system but also more challenging to eradicate than C. albicans⁹. The detection of candiduria in an ICU patient should be regarded as an indicator of poor prognosis and the accurate identification of Candida spp., as well as the investigating the antifungal susceptibility, will be beneficial in terms of the effectiveness of the treatment and the prevention of resistance development.