1. Background
Candida species are opportunistic pathogenic fungi that colonize in human body surfaces and the mucous membrane of the oral cavity and digestive tract. They may cause diseases ranging from non-life-threatening mucosal Candida infections to life-threatening invasive candidiasis, when the host possesses certain risk factors, such as the aggressive use of immunosuppressive agents, cytotoxic therapies, treatment with broad-spectrum antifungal agents, prolonged central venous catheterization, total parenteral nutrition, AIDS, diabetes, and drug abuse. Candidiasis is one of the most common invasive fungal nosocomial infections associated with a high attributable mortality (1). Over the past several decades, the frequency of Candida infections has risen progressively, especially among immunocompromised patients (2, 3).
The ARTEMIS DISK global antifungal surveillance project found that 90% of fungal infections were caused by Candida albicans, C. glabrata, C. parapsilosis, and C. tropicalis (4). Furthermore, the proportions of them in Candida infections are significantly associated with geographical differences. Retrospective studies in intensive care units (ICUs) showed the prevalence rate of C. albicans was between 29% and 67% of Candida infections in China (5, 6), while it accounted for 50 - 70% in the US (7, 8).
Candida albicans has traditionally been the leading cause of candidemia worldwide. However, the non- albicans species have been increasing in frequency to become the major pathogens in parts of Asia (including China), South America, and southern Europe in recent years (4, 9, 10) although C. albicans is still dominant. The increased incidence of the acquired resistance to azoles and the episode shift from C. albicans to non-albicans species may pose serious problems in the management of infections by such Candida spp., and it is necessary to identify the Candida species quickly in the clinical settings and monitor the antifungal resistance for providing appropriate treatments for candidiasis.
2. Objectives
In this study, we retrospectively investigated 149 isolates (from over 3,000 Candida) of the clinical Candida species from sterile fluids over a 3-year period at a tertiary hospital. The results would not only help us monitor the distribution and anti-drug susceptibility of the invasive Candida infections but also provide sufficient evidence for Candida treatment.
3. Methods
3.1. Ethics Statement
This study was approved by the ethics committee of the study hospital (code number: 2014036).
3.2. Isolates
A total of 149 clinical isolates taken from aseptic body fluids were collected by the clinical laboratory of the first affiliated hospital of Nanchang university between August 2008 and July 2010. Each positive isolate was recovered from a unique clinical specimen, purified with thrice streaking on Sabouraud dextrose agar (SDA), and incubated at 35°C for 48 hours.
3.3. Identification
Two traditional clinical identification methods involving CHROMagar Candida medium (CHROMagar, Paris, France) and Vitek2 YST (bioMe’rieux) were performed in combination for individual isolates. Multiple PCR, a method based on nucleic acid molecule identification, was used, as well (11). PCR reaction was performed with primers Its1 (5’-TCCGTAGGTGAACCTGCGG-3’), Its4 (5’-TCCTCCGCTTATTGATATGC-3’), Ca1 (5’-GGTTTGCTTGAAAGACGGTAG-3’), Ca2 (5’-AGTTTGAAGATATACGTGGTAG-3’), Ct1 (5’-CAATCCTACCGCCAGAGGTTAT-3’), and Ct2 (5’-TGGCCACTAGCAAAATAAGCGT-3’) (11-13). PCR reactions were carried out in a GeneAmp PCR system 2400 (PerkinElmer, US) and performed in 0.2 mL microcentrifuge tubes with a final reaction mixture containing 2 μL of the prepared template, 12.5 μL of 2 × Taq PCR master mix (Tiangen Biotech (Beijing) Co., Ltd., China), 0.5 μL of each primer (10 mM)(BGI, China), and 7.5 μL ddH2O. Amplification conditions were as follows: 5 minutes initial denaturation at 96°C, repeated for 40 cycles of 30 seconds denaturation at 94°C, 30 seconds of primer annealing at 53°C, and elongation at 72°C for 1 minute. The final elongation at 72°C was extended to last 10 minutes.
For each experiment, the size of the DNA fragments amplified by Multi-PCR was determined by direct comparison with the DNA marker. Control tubes without template were included in each run, and reproducibility was verified for each reaction. The PCR products were separated using electrophoresis in agarose gels (1.5%) at 100 V for approximately 45 min at room temperature in the 1 × TAE buffer. Reaction products were detected by ethidium bromide and visualized in UV light. Candida albicans included two bands, 150 bp and 550 bp (faint). Bands approximately at 900 bp, 350 bp, and 550 bp represented C. glabrata, C. tropicalis, and C. krusei, respectively.
3.4. Antifungal Susceptibility Testing
Broth microdilution testing was performed in accordance with the guidelines of CLSI document M27-A3 (14) using RPMI 1640 medium with 0.2% glucose an inoculum of 5 × 102 to 2.5 × 103 cells/mL, followed by incubation at 35°C. The minimal inhibitory concentration (MIC) values were determined visually after 24 hour of incubation (CLSI document M27-A3) as the lowest concentration of drug that caused a significant diminution (≥ 50% or 90% inhibition) in growth relative to the growth of the control (14). In all instances, MIC trays were prepared using reagent-grade powder, as directed by CLSI. All of the drugs (including amphotericin B flucytosine, fluconazole, itraconazole, and ketoconazole) were bought from Sigma Aldrich, Germany.
3.5. Interpretive Criteria for Susceptibility to Antifungal Agents
We adopted the clinical breakpoints (CBPs) that have been newly revised to provide species-specific interpretive criteria in this study (16). Considering the absence of the CBPs for some species, we applied the epidemiological cut-off values (ECVs) instead to detect the emergence of potential resistance to antifungal agents (reviewed by Pfaller et al. (16)). The strains with a MIC ≥ CBP were considered resistant (R). We also considered MIC ≤ ECV as the wild-type strains (WT) and MIC > ECV as the non-WT strains. Interpretive criteria for susceptibility to antifungal agents were as follow: for amphotericin B, the ECV was 2 μg/mL for each species of Candida. For flucytosine, the ECV was 0.5 μg/mL for Candida species except for 32 μg/mL for C. krusei and 1 μg/mL for C. guilliermondii. For fluconazole, MICs ≤ 2 μg/mL were considered S for C. albicans, C. parapsilosis and C. tropicalis, MICs ≥ 8 μg/mL were R, and MICs = 4 μg/mL were susceptible dose-dependent (SDD). For C. glabrata, MIC ≤ 32 μg/mL was considered SDD while MIC ≥ 64 μg/mL was considered R. For C. krusei and C. guilliermondii, the ECV was 64 μg/mL and 8 μg/mL, respectively. For itraconazole, the MIC ≤ 0.125 μg/mL was considered S for C. albicans, and MIC ≥ 1 μg/mL was R, 0.25 - 0.5 μg/mL was SDD, while the ECV was 2 μg/mL for C. glabrata, 0.5 μg/mL for C. parapsilosis, and C. tropicalis, and 1 μg/mL for C. krusei and C. guilliermondii. For ketoconazole, referring to M27-A2 and based on Sanguinetti et al. (17) and White et al. (18), MICs ≤ 0.125 μg/mL, 0.25 - 0.5 μg/mL, and ≥ 1 μg/mL were considered S, SDD, and R, respectively for each species of Candida.
4. Results
Distributions of Candida strains: baseline characteristics in the study population were analyzed (Table 1). Briefly, we obtained 149 Candida strains, mainly containing C. albicans (71, 47.7%), C. glabrata (40, 26.8%), C. parapsilosis (13, 8.7%), C. tropicalis (20, 13.4%), and C. krusei (3, 2.0%). Most Candida strains were isolated from urine (113, 75.8%), followed by blood (25, 16.8%). The remaining 7.4% of the strains were isolated from drainage (6, 4.0%), hydrothorax and ascites (3, 2.0%), succus prostaticus (1, 0.7%), and cerebrospinal fluid (1, 0.7%). The highest infection rate was found in ICU (52, 34.9%), followed by the departments of Urology (45, 30.2%) and Burn (15, 10.1%). Other departments, including chest surgery, gastroenterology, orthopedics, and gynaecology, separately represented a small proportion of the isolates. C. albicans was most predominant in ICU (53.8%), urology (46.7%), and burn (53.3%) departments. C. glabrata was the second one, with 19.2%, 31.1%, and 26.7% rates at the above three departments, respectively.
Table 1.Distributions of the Isolates in the Current Study According to Candida Species in Nanchang, China (2008 - 2010)a
| Characteristic | Candida species | |||||||
|---|---|---|---|---|---|---|---|---|
| All Isolates (n = 149) | C. albicans (n = 71) | C. glabrata (n = 40) | C. tropicalis (n = 20) | C. parapsilosis (n = 13) | C. krusei (n = 3) | C. guilliermondii (n = 1) | C. intermedia (n = 1) | |
| Departments | ||||||||
| Intensive care unit | 52 (34.9) | 28 (53.8) | 10 (19.2) | 6 (11.5) | 6 (11.5) | 2 (3.8) | Δ | |
| Urology | 45 (30.2) | 21 (46.7) | 14 (31.1) | 6 (13.3) | 4 (8.9) | Δ | Δ | Δ |
| Burn | 15 (10.1) | 8 (53.3) | 4 (26.7) | 2 (13.3) | Δ | 1 (2.2) | Δ | Δ |
| Cerebral surgery | 9 (6.0) | 3 (33.3) | 3 (33.3) | 2 (22.2) | Δ | Δ | Δ | 1 (11.1) |
| Emergency | 11 (7.4) | 5 (45.5) | 3 (27.3) | Δ | 3 (27.3) | Δ | Δ | Δ |
| Others | 17 (11.4) | 6 (35.3) | 6 (35.3) | 4 (23.5) | Δ | Δ | 1 (5.9) | Δ |
| Sources | ||||||||
| Urine | 113 (75.8) | 55 (48.7) | 29 (25.7) | 16 (14.2) | 10 (8.8) | 1 (0.9) | 1 (0.9) | 1 (0.9) |
| Blood | 25 (16.8) | 7 (28.0) | 10 (40.0) | 3 (12.0) | 3 (12.0) | 2 (8.0) | Δ | Δ |
| Drainage liquid | 6 (4.0) | 4 (66.7) | 1 (16.7) | 1 (16.7) | Δ | Δ | Δ | Δ |
| Hydrothorax and ascite | 3 (2.0) | 3 (100.0) | Δ | Δ | Δ | Δ | Δ | Δ |
| Cerebrospinal fluid | 1 (0.7) | 1 (100.0) | Δ | Δ | Δ | Δ | Δ | Δ |
| Succus prostaticus | 1 (0.7) | 1 (100.0) | Δ | Δ | Δ | Δ | Δ | Δ |
| Years | ||||||||
| 2008 | 34 | 17 (50.0) | 10 (29.4) | 3 (8.8) | 3 (8.8) | 1 (2.9) | Δ | Δ |
| 2009 | 47 | 19 (40.4) | 13 (27.7) | 8 (17.0) | 4 (8.5) | 2 (4.3) | Δ | 1 (2.1) |
| 2010 | 68 | 35 (51.5) | 17 (25.0) | 9 (13.2) | 6 (8.8) | Δ | 1 (1.5) | Δ |
a Δ, data unavailable.
In Table 1, we analyzed the constituent ratio of the Candida species in each year from 2008 to 2010. The percentage of C. albicans or C. glabrata changed slightly, while the increasing trend of C. tropicalis was distinct as the second major species in non-albicans Candida.
Antifungal susceptibility testing: The frequency of resistance to some antifungal agents for the five most common Candida species is shown in Table 2. We counted the percentages of strains that were susceptible or resistant to each agent, except the two agents that have no interpretive criteria. Generally, all isolates were susceptible to amphotericin B and flucytosine, except one flucytosine-resistant C. albicans isolate. This is while azoles resistance was evidently uncommon with 27.5% to ketoconazole (32.4% of C. albicans, 15.0% of C. glabrata, 30.0% of C. tropicalis and 38.5% of C. parapsilosis), 22.1% to itraconazole (36.6% of C. albicans, 2.5% of C. glabrata, 20.0% of C. tropicalis, and 30.8% of C. parapsilosis) and 17.4% to fluconazole (28.2% of C. albicans, 15.0% of C. tropicalis, and 23.1% of C. parapsilosis). Notably, 16.1% (24/149) of the isolates exhibited a cross-resistance to fluconazole, itraconazole, and ketoconazole. For clotrimazole and nystatin, we analyzed the MIC ranges as listed in Table 2.
Table 2.Susceptibilities of 149 Candida Isolates to Antifungal Agents.
| Antifungal Agent | Isolates, No. (%) | MIC (mg/L)a | CBPb | ECVc | |||||
|---|---|---|---|---|---|---|---|---|---|
| Range | 50% | 90% | S, No. (%) | SDD, No. (%) | R, No. (%) | WT, No. (%) | NWT, No. (%) | ||
| C. albicans | 71 (47.7) | ||||||||
| Amphotericin B | 0.25 - 1 | 1 | 1 | 71 (100.0) | 0 (0.0) | ||||
| Flucytosine | 0.125 - 1 | 0.125 | 0.5 | 70 (98.6) | 1 (1.4) | ||||
| Fluconazole | 0.125 - 64 | 2 | 16 | 39 (54.9) | 12 (16.9) | 20 (28.2) | 16 (22.5) | 55 (77.5) | |
| Itraconazole | 0.03125 - 16 | 0.5 | 16 | 18 (25.4) | 27 (38.0) | 26 (36.6) | 18 (25.4) | 53 (75.6) | |
| C. glabrata | 40 (26.8) | ||||||||
| Amphotericin B | 0.25 - 1 | 0.5 | 1 | 40 (100.0) | 0 (0.0) | ||||
| Flucytosine | 0.125 - 0.5 | 0.125 | 0.5 | 40 (100.0) | 0 (0.0) | ||||
| Fluconazole | 0.125 - 16 | 1 | 8 | NAd | 40 (100.0) | 0 (0.0) | 40 (100.0) | 0 (0.0) | |
| Itraconazole | 0.03125 - 4 | 0.5 | 1 | 39 (97.5) | 1 (2.5) | ||||
| C. tropicalis | 20 (13.4) | ||||||||
| Amphotericin B | 0.25 - 1 | 0.5 | 1 | 20 (100.0) | 0 (0.0) | ||||
| Flucytosine | 0.125 - 0.5 | 0.25 | 0.5 | 20 (100.0) | 0 (0.0) | ||||
| Fluconazole | 0.125-32 | 2 | 8 | 12 (60.0) | 5 (25.0) | 3 (15.0) | 12 (60.0) | 8 (40.0) | |
| Itraconazole | 0.03125 - 16 | 0.25 | 1 | 16 (80.0) | 4 (20.0) | ||||
| C. parapsilosis | 13 (8.7) | ||||||||
| Amphotericin B | 0.25 - 1 | 0.5 | 1 | 13 (100.0) | 0 (0.0) | ||||
| Flucytosine | 0.125 - 0.5 | 0.125 | 0.25 | 13 (100.0) | 0 (0.0) | ||||
| Fluconazole | 0.25 - 32 | 2 | 8 | 8 (61.5) | 2 (15.4) | 3 (23.1) | 8 (61.5) | 5 (38.5) | |
| Itraconazole | 0.03125 - 16 | 0.25 | 4 | 9 (69.2) | 4 (30.8) | ||||
| C. krusei | 3 (2.0) | ||||||||
| Amphotericin B | 0.5 - 1 | 1 | 1 | 3 (100.0) | 0 (0.0) | ||||
| Flucytosine | 0.125 - 0.25 | 0.125 | 0.25 | 3 (100.0) | 0 (0.0) | ||||
| Fluconazole | 1 - > 64 | 2 | > 64 | 2 (66.7) | 1 (33.3) | ||||
| Itraconazole | 1 - 2 | 1 | 2 | 2 (66.7) | 1 (33.3) | ||||
| C. guilliermondii | 1 (0.7) | ||||||||
| Amphotericin B | 0.5 | NA | NA | 1 (100.0) | 0 (0.0) | ||||
| Flucytosine | 0.125 | NA | NA | 1 (100.0) | 0 (0.0) | ||||
| Fluconazole | 4 | NA | NA | 1 (100.0) | 0 (0.0) | ||||
| Itraconazole | 0.0625 | NA | NA | 1 (100.0) | 0 (0.0) | ||||
| C. intermedia | 1 (0.7) | ||||||||
| Amphotericin B | 0.5 | NA | NA | NA | NA | ||||
| Flucytosine | 0.125 | NA | NA | NA | NA | ||||
| Fluconazole | 2 | NA | NA | NA | NA | ||||
| Itraconazole | 0.5 | NA | NA | NA | NA | ||||
a 50% and 90%, MIC value encompassing 50% and 90% of isolates tested, respectively.
b Categories of susceptible (S), susceptible-dose dependent (SDD), and resistant (R) according to CLSI (2012) and Pfaller and Diekema (2012).
c Categories of wild-type (WT) and non-wild type (NWT) according to Pfaller and Diekema (2012).
d NA, data not available or not applicable.
5. Discussion
In this study, we described the distribution of species and determined in vitro antifungal susceptibility of Candida strains isolated from aseptic body fluids in our hospital. Consequently, we found the number of Candida infections was increasing from 2008 to 2010. It indicated that we should pay more attention to monitoring the candidiasis frequently because it is an intractable disease, especially candidemia. Compelling evidence has demonstrated that more than 90% of Candida infections were caused by C. albicans, C. glabrata, C. parapsilosis, and C. tropicalis, and these four Candida held 96.6% of all isolates in our study. Traditionally, C. albicans was considered the predominant pathogen in candidiasis worldwide. In the current study, the most frequent isolates were still C. albicans. The proportion concurs with recent reports in ICUs in China (6, 19), but is considerably lower than that in the ARTEMIS DISK global antifungal surveillance study between 2001 and 2007 (64%~67%) (4). The reasons for this discrepancy may include different specimen types and locations.
In addition, it is noteworthy that more than 50% of the Candida isolates in the current study were non-albicans. In the current study, C. glabrata (26.8%) was the most prevalent non-albicans Candida, which is in agreement with the results of the studies conducted by Hazen et al. (20) and Schmalreck et al. (21). In contrast, C. parapsilosis was the foremost non-albicans Candida species in Latin America (25%), Canada (16%), and Europe (17%) (22) and C. tropicalis was the most prevalent in Taiwan (23). It indicated that the distribution of species is significantly associated with the geographical area. Furthermore, our data showed that the prevalence of non-albicans Candida, especially C. glabrata and C. tropicalis, has been increasing recently. Owing to the non-albicans treatment that is slightly different from C. albicans treatment, the results remind us that we should be more concerned about the non-albicans infections.
Among the 149 clinical isolates, Candida was frequently isolated from urine specimens (75.8%), followed by blood (16.8%). Similar results were observed in Taiwan (urine, 45.2%; blood, 19.7%) as a survey reported in 2010 (23). In urine, C. albicans had the largest proportion (41.4%), followed by C. glabrata (21.8%) and C. tropicalis (12.0%). On the one hand, it may be developed from the Candida species that colonize on the mucous membrane of the urinary tract; on the other hand, it was more likely to be due to surgery and intubation. Notably, C. glabrata surpassed C. albicans to be the most common species isolated from the bloodstream, accounting for 40.0% of the isolates. In contrast, C. albicans was the most common cause of candidemia (35.9%) and C. glabrata was reported third (13.0%) in a retrospective analysis of 270 cases of candidemia occurring from 2000 to 2009 at a teaching hospital in China (24). Furthermore, clinical works should pay more attention to C. albicans and C. parapsilosis in urine because approximately 70.0% of them (39/55 of C. albicans and 7/10 of C. parapsilosis) were resistant (data not shown). In blood, however, C. albicans should be focused on because of its high resistance rate (57.1%, 4/7) (data not shown).
Candida infection rates in the departments of ICU and Urology were most predominant (34.9% and 30.2%, respectively), which is consistent with the distributions reported in the US (25). This is possible because the patients in ICUs may be immunocompromised or critically ill, and they were more likely to be infected with Candida. In addition, the high resistance rates of C. albicans in ICU and Urology departments reveal that more attention should be paid to ICU or urology patients to prevent them from contracting Candida (data not shown).
Of all the 149 isolates, 27.5% were resistant to ketoconazole. In contrast, only 7.7% of the Candida strains isolated from HIV/AIDS patients were identified to be resistant to ketoconazole in a study conducted by Mulu et al. (26). This contradiction may be because of the divergence of the methods and interpretive criteria to ketoconazole or due to the sources of strains. We found that a large percentage of C. albicans (32.4%), C. glabrata (15.0%), C. parapsilosis (38.5%), and C. tropical (30.0%) were ketoconazole-resistant. It was significantly higher than the results obtained by Badiee and Alborzi (9.4% of C. albicans, 15% of C. glabrata, and 3.5% of C. parapsilosis) (27). These may prompt the reconsideration of ketoconazole therapy.
With regard to itraconazole, 22.1% of the isolates were resistant and the rate was slightly above a recent report in a tertiary care hospital in south India (14%) (28). 36.6% of C. albicans exhibited resistance to itraconazole, followed by C. parapsilosis (30.8%) and C. tropicalis (20.0%), unlike the results of the China-SCAN (16.0% of C. albicans, 39.8% of C. parapsilosis, and 31.3% of C. tropicalis) (19). For fluconazole, 18.1% of the isolates were resistant, and this rate was similar to the result of a population-based surveillance in India (14.8%) (28). However, the proportion of ketoconazole- or itraconazole-resistant isolates was significantly higher compared to fluconazole-resistant isolates. It may be because of prescribing habits in clinics; fluconazole was not commonly used to treat the infections of the bloodstream and other sterile sites.
Some surveillance had shown that resistance to fluconazole was highly predictive of resistance to voriconazole (25, 29). In this regard, we found that 16 out of 20 fluconazole-resistant C. albicans were cross-resistant to itraconazole. Furthermore, six of them were also resistant to ketoconazole. This phenomenon highlights the importance of cross-resistance among azole agents. One of the three C. krusei was resistant to itraconazole, ketoconazole, and fluconazole. However, C. krusei was considered inherently resistant to triazoles (16). The high prevalence of azoles resistance among Candida spp. may be correlated with the increased use of these drugs in the area. It appears to have a major impact in selecting azole-resistant Candida spp. when strains are continuously exposed to azoles (30).
Amphotericin B was reported to be the first systemic antifungal agent for the treatment of invasive fungal infection in many studies; however, it has limited use due to nephrotoxicity in up to 80% of the patients. In this study, its ECV value was 2 μg/mL for Candida spp. and the MICs of all isolates were below 2 μg/mL. Flucytosine was used to treat systemic severe Candida infections. With regard to flucytosine, all 149 isolates were susceptible except one C. albicans isolated from urine, contrary to the result (1.7%) of a study conducted by Pfaller et al. (31) whose strains came from nine cities across China. The findings of antifungal susceptibility of Candida spp. in the current study and previous reports revealed that the prevalence of antifungal resistance in Candida isolates differs from area to area and time to time.
5.1. Conclusions
In conclusion, we described the distributions and in vitro antifungal susceptibilities of invasive Candida strains isolated from sterile fluids. The trends of species distribution and increasing resistance of Candida spp. (especially the occurrence of cross-resistant isolates) confirm the importance of continuous epidemiological surveillance and the rational use of fungal agents. The mechanisms of the cross-resistance of antifungal agents in our study are not clear, while this problem may be interpreted in further studies.
