Risk Factors and Outcomes of Non-albicans Candida Bloodstream Infection in Patients with Candidemia at Siriraj Hospital—Thailand’s Largest National Tertiary Referral Hospital

This study aimed to investigate the risk factors for and the outcomes of patients with candidemia caused by non-albicans Candida. Candidemia patients treated at Siriraj Hospital (Bangkok, Thailand) during January 2016 to December 2017 were enrolled. A total of 156 patients (mean age: 65 years, 56.4% male) were included. The most prevalent underlying conditions were diabetes (32.1%), chronic cardiac disease (28.2%), chronic kidney disease (26.9%), and hematologic malignancies (21.2%). Candida species isolated from patient blood were C. tropicalis (49.4%), C. albicans (28.8%), C. glabrata (16.7%), and C. parapsilosis (5.1%). Fluconazole resistance was significantly increased in C. tropicalis (37.8%). No independent risk factors were associated with patients with non-albicans Candida candidemia compared to those with C. albicans candidemia. There was no significant difference in mortality between patients with non-albicans Candida candidemia and patients with C. albicans candidemia (OR: 1.35, 95% CI: 0.64–2.85). When compared with C. albicans candidemia, multivariate analysis revealed chronic liver disease (OR: 11.39, 95% CI: 1.38–94.02), neutropenia (OR: 4.31, 95% CI: 1.34–13.87), and male gender (OR: 2.34, 95% CI: 1.04–5.29) to be independent risk factors for C. tropicalis candidemia. The observed high resistance of C. tropicalis to fluconazole indicates that fluconazole should not be used for empirical antifungal treatment in these patients.


Introduction
Candida spp. is the most common fungal pathogen that causes nosocomial infection [1,2]. Candidemia, which is a blood stream infection caused by Candida, is a serious public health problem that is associated with high mortality and high medical and economic costs [2]. A previous study reported the incidence of candidemia to be approximately 13.3 per 100,000 patients [3], and the mortality rate ranged from 36% to 59% [4,5]. The incidence of Candida infection increases following advanced medical interventions, such as hematopoietic stem cell transplantation and solid organ transplantation, as well as after invasive medical procedures, such as the use of vascular catheter or mechanical ventilator [4]. Factors reported to be associated with mortality in candidemia included neutropenia, septic shock, intensive care unit (ICU) admission, inappropriate antifungal therapy within 72 h, and renal failure [5].
Previous epidemiological studies demonstrated the same trend toward increasing numbers of patients with candidemia in both neutropenic and non-neutropenic individuals. costeroid equivalent to prednisolone at least 20 mg/day for at least 2 weeks, absolute neutrophil count < 500 cells/mm 3 , yeast in non-sterile body samples, prior antifungal exposure, and comorbidities. Candidemia was categorized into primary fungemia, acute disseminated candidiasis, catheter-related blood stream infection, intra-abdominal infection, hepatosplenic abscess, bone and joint infection, central nervous system infection, endocarditis, and urinary tract infection. All Candida isolated from blood cultures were identified to the species level. The species identification of Candida was performed based on biochemical characteristics using chromogenic medium (Brilliance™ Candida agar) and RapID™ Yeast Plus Panel (Thermo Fisher Scientific, Kent, UK). Antifungal susceptibility was tested by the microbroth dilution method using Thermo Scientific™ Sensititre™ Yeast-One™ YO9 AST Plate (Thermo Fisher Scientific, Waltham, MA, USA). Minimal inhibitory concentration (MIC) breakpoints were interpreted according to the Clinical Laboratory Standards Institute (CLSI). Antifungal susceptibility test results and clinical outcomes were analyzed.

Statistical Analysis
Continuous data are presented as mean ± standard deviation (SD) for normally distributed data, and as median and interquartile range (IQR) for non-normally distributed data. Categorical data are presented as frequency and percentage. The Chi-square test or Fisher's exact test was used for the comparison of categorical variables between two groups. The independent t-test was used to compare normally distributed continuous data, and the Mann-Whitney U test was used to compare non-normally distributed continuous data. Multivariate analysis via multiple logistic regression model (backward stepwise) included all variables with a p-value less than 0.2 from univariate analysis. All statistical analyses were performed using SPSS Statistics (v18.0) software (SPSS, Inc., Chicago, IL, USA), and a 2-sided p-value less than 0.05 was considered to be statistically significant.

Factors Significantly and Independently Associated with Non-albicans Candida Candidemia
Patient characteristics and other factors significantly associated with non-albicans Candida and C. albicans candidemia are shown in Tables 1 and 3. Chronic liver disease and neutropenia were the only two factors found to be significantly associated with non-albicans Candida candidemia in univariate analysis.

Factors Significantly and Independently Associated with Non-albicans Candida Candidemia
Patient characteristics and other factors significantly associated with non-albicans Candida and C. albicans candidemia are shown in Tables 1 and 3. Chronic liver disease and neutropenia were the only two factors found to be significantly associated with non-albicans Candida candidemia in univariate analysis.
Subgroup analysis of candidemia caused by C. tropicalis or C. albicans to identify the factors associated with those two infections is shown in Tables 4 and 5. Multivariate analysis revealed male gender, chronic liver disease, and neutropenia to be factors independently associated with C. tropicalis infection ( Table 6). The mortality rate was not significantly different among patients with candidemia caused by C. albicans, non-albicans Candida, or C. tropicalis.

Discussion
From January 2016 to December 2017, we analyzed data from 156 candidemia patients admitted to Siriraj Hospital, the largest university hospital in Thailand. Most patients had candidemia that was caused by non-albicans Candida (71.2%), which is consistent with a previous study in Thailand, and which reveals an increasing incidence of non-albicans candidemia [4]. C. tropicalis was found to be the most common Candida species in patients with candidemia, and this is similar to the results of previous studies conducted at the same hospital [5,19]. It was proposed that the pathogenicity of C. tropicalis may be associated with a warmer climate [12], which may explain why it was found to be the most common Candida species in our study.
The present study identified neutropenia and chronic liver disease as potential risk factors for non-albicans Candida candidemia. A previous study also reported non-albicans Candida candidemia to be associated with neutropenic patients [6]. However, no previous study has reported chronic liver disease as a risk factor of non-albicans Candida candidemia.
Risk factors associated with C. tropicalis and C. albicans candidemia were also analyzed in this study. The results of that analysis revealed neutropenia, chronic liver disease, and male gender to be independent risk factors for C. tropicalis candidemia. A previous study reported the independent risk factors for C. tropicalis candidemia to be abdomen portal entry, hematologic malignancies, and older age [20]. Our study confirmed that C. tropicalis seems to originate from the intra-abdominal origin, and it tends to infect neutropenic patients.
Candida tropicalis has become increasingly resistant to fluconazole. Two previous studies that were conducted at our center in 2009 and 2013 found rates of C. tropicalis susceptibility to fluconazole of 100% and 70.3%, respectively [5,19]. However, in this study, we found a rate of C. tropicalis susceptibility to fluconazole of only 62%. Therefore, empirical antifungal treatment with echinocandins or amphotericin B product in candidemia patients with risk factors for C. tropicalis infection is strongly recommended. In our study, most of the patients infected with fluconazole-resistant C. tropicalis did not have a history of azole exposure and they may be acquired from the environment [21]. Azoles have been used in agriculture in Thailand and the link between fluconazole-resistant Candida and environmental azole exposure needs to be investigated.
The mortality rate among candidemia patients in this study was approximately 70%, which is higher than in other studies, but the mortality rate was not found to be significantly different between C. albicans and non-albicans Candida candidemia. This extremely high mortality rate in candidemia patients highlights the need to improve candidemia management guidelines in Thailand.
The limitations of this study include its single-center retrospective cohort design, the fact that the epidemiology of Candida species can differ among hospitals, and that antifungal susceptibility testing was not routinely performed in every included study patient.

Conclusions
In candidemia patients, chronic liver disease, neutropenia, and male gender were found to be independently associated with C. tropicalis infection. Due to the observed high resistance of C. tropicalis to fluconazole, fluconazole should not be used for empirical antifungal treatment in these patients.