Trends in candidemia and antifungal resistance: a nine-year single-center study in Turkey

Abstract

Background

Candidemia is a major global health concern with rising incidence and mortality. Regional surveillance is essential, as species distribution and antifungal resistance vary significantly.

Methods

We analyzed 541 Candida isolates from blood cultures at Cerrahpaşa Medical Faculty Hospital (2015–2023). Identification was performed using phenotypic methods, API 20 C AUX, BD Phoenix YEAST ID, and MALDI-TOF MS. Antifungal susceptibility was tested by gradient test (E-test) on RPMI agar, interpreted according to EUCAST guidelines for azoles, echinocandins, and amphotericin B.

Results

The most frequent isolates were C. albicans (n = 214, 39%; susceptibility tested in 150), C. parapsilosis (n = 164, 29%), C. glabrata (n = 47, 7%), C. tropicalis (n = 42, 7%), and C. krusei (n = 10, 2%). Other species accounted for 16% (n = 64), including cryptic species such as C. auris (n = 9) and C. haemulonii (n = 1). The total cases were 269 in the pre-COVID period and 272 in the COVID and post period, chi-square analysis indicated these differences were not statistically significant (χ²=7.43, df = 6, p = 0.283). Increased MIC values ​​against azoles were significantly detected in C. glabrata isolates, while C. parapsilosis complex generally exhibited a relatively high MIC distribution and a decreased susceptibility pattern. C. krusei demonstrated intrinsic resistance to fluconazole but remained susceptible to voriconazole.

Conclusions

Our findings demonstrate a shift, though not statistically significant, toward non-albicans Candida species with emerging azole resistance, including the detection of multidrug-resistant C. auris. These trends underscore the need for enhanced local surveillance, molecular diagnostics, and evidence-based antifungal stewardship to optimize patient outcomes.

Clinical trial number

Not applicable.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12879-026-13034-x.

Introduction

Invasive fungal infections have emerged as a major global health concern, with reports indicating a significant increase in incidence over recent decades [1,2]. The World Health Organization’s Fungal Priority Pathogens List recognizes this growing threat by classifying several Candida species as critical and high-priority pathogens, with Candida albicans and the multidrug-resistant Candida auris designated as critical-priority pathogens due to their global distribution and treatment challenges [3].

Candidemia, caused by various Candida species, represents one of the most serious forms of invasive fungal infection, characterized by high mortality rates and diagnostic challenges in early stages [4]. Globally, Candida bloodstream infections constitute a significant cause of healthcare-associated morbidity and mortality, particularly in immunocompromised patients and those with multiple comorbidities [5]. The clinical burden is further complicated by the increasing emergence of antifungal resistance, particularly to azole compounds, which has been linked to their widespread prophylactic and empirical use [6].

Fluconazole (FLC) remains a cornerstone in candidemia management, serving both as primary treatment and prophylaxis in low income countries [7]. However, resistance patterns vary significantly among Candida species, with primary resistance observed in C. krusei and acquired resistance frequently developing in C. glabrata and other non-albicans Candida (NAC) species [8, 9]. Therefore, in cases where azole resistance is suspected, species identification and susceptibility test results are crucial [10, 11]. Starting early and effective antifungal treatment has a positive effect on the mortality and morbidity of candidemia [12].

While global data have highlighted major shifts in candidemia epidemiology, regional studies show important differences. In Türkiye, a multicenter study on Candida parapsilosis bloodstream isolates reported an increasing rate of FLC resistance, along with the emergence of echinocandin- and multidrug-resistant strains [13] Furthermore, investigations of Candida tropicalis isolates from 2017 to 2019 revealed the recent emergence of azole and echinocandin resistance, and even multidrug resistance, contrasting with earlier data demonstrating low resistance rates prior to 2017 [14]. However, as a separate study reported low resistance among C. tropicalis isolates in Turkey [15]. Additionally, a Turkish study reported that among candidemia isolates, the non-albicans yeasts accounted for over half—C. parapsilosis complex (28%), C. tropicalis (10%), and C. glabrata (8%). FLC susceptibility rates varied: ~67% in C. albicans, ~ 67% in C. parapsilosis, and lower in C. tropicalis, with some isolates resistant or dose dependent [16].

These findings underscore significant regional variability in species distribution and antifungal susceptibility, reinforcing the importance of ongoing local surveillance to guide empirical therapy and antimicrobial stewardship. This study aimed to evaluate the distribution of Candida species and antifungal resistance patterns in bloodstream isolates over a nine-year period (2015–2023), providing essential long-term epidemiological data for Türkiye to support evidence-based clinical decision-making and contribute to understanding regional and temporal trends in candidemia.

Methods

Clinical isolates

The isolates identified as Candida spp. from blood culture samples received at the Medical Microbiology Mycology Laboratory of Cerrahpaşa Medical Faculty Hospital between 2015 and 2023 were evaluated. In cases where repeated growth was detected, the initial isolate was selected for detailed analysis, and comprehensive characterization was performed using antifungal susceptibility testing after species-level confirmation with MALDI-TOF MS. Information about the patients was obtained from the hospital information system (ISHOP). The study was approved by the Istanbul University-Cerrahpaşa Rectorate Clinical Research Ethics Committee (decision number: 2024-1014569; Annex 1).

Blood culture was evaluated using the BACTEC™ haemoculture (Becton Dickinson, Franklin Lakes, NJ, USA) automated system. When yeast cells and pseudohyphae structures were observed on Gram staining in bottles containing growth signals, Sabouraud dextrose agar (Oxoid, Thermo Fisher, UK) medium and CHROMagar Candida (Becton Dickinson GmbH, Germany) medium were inoculated (35 °C). In the identification of yeast colonies isolated in primary culture by conventional methods; colony morphology, germ tube formation. Yeast colonies grown on Cornmeal-Tween 80 agar (Oxoid, Thermo Fisher, UK) were examined under a microscope for blastoconidium or pseudohyphae structures and identified using the API 20 C AUX (bioMerieux, France) kit and BD Phoenix YEAST ID (Becton, Dickinson and Co., Sparks, USA) for carbohydrate assimilation tests, and confirmed by MALDI-TOF MS (MALDI Biotyper, Bruker Daltonics GmbH).

Antifungal susceptibility tests

Antifungal susceptibility testing was performed on all isolates using the gradient test (E-test). RPMI agar was used as the medium for susceptibility testing. The European Committee on Antimicrobial Susceptibility Testing guidelines (EUCAST; E.Def 7.4, E.Def 9.4, E.Def 11.0), updated, were used to evaluate antifungal susceptibility testing. Epidemiological thresholds and clinical cut-off values were taken into account (Version 4.0, valid from 2023; EUCAST, 2023).

Gradient test (E-test)

Candida spp. isolates were placed in the medium (containing 2% glucose) prepared with RPMI 1640 (Sigma Chemical Co, St Louis, MO, USA) and MOPS (3-N-morpholinopropane sulfonic acid, Sigma Chemical Co, St Louis, MO, USA) buffer at a McFarland turbidity of 0.5. E-test strips (BioMérieux, France) were then placed in the medium. Antifungal susceptibility testing of Candida spp. isolates was performed using the Gradient test (E-test) method, encompassing azole group agents (FLC, VRC), echinocandin group agents (anidulafungin, micafungin), and amphotericin B (AMB), which belongs to the polyene class. The medium was then evaluated after 24–48 h of incubation at 37 °C. Candida lusitaniae was evaluated after 72 h due to insufficient growth. If any azole minimum inhibitory concentration (MIC) value was found to be at or above the resistance breakpoint, antifungal susceptibility testing was repeated for confirmation. MIC50 and MIC90 values were determined for each antifungal. Candida parapsilosis ATCC 22019 and Candida albicans ATCC 10231 control strains with known MIC ranges were used in the anti-fungal susceptibility tests.

Results

A total of 541 Candida isolates were included in the study. The most frequently isolated species were C. albicans (antifungal susceptibility was studied for 150 isolates out of 214) 39%, C. parapsilosis (n = 164) 29%, C. glabrata (Nakaseomyces glabrata; n = 47) 7%, C. tropicalis (n = 42) 7%, C. krusei (Pichia kudriavzevii; n = 10) 2% and other rare species (Candida spp.; n = 64) 16%.

The analysis of Candida species distribution shows changes between the pre-COVID-19 periods (2015–2019) and the COVID-19 and post period (2020–2023) from a total of 541 isolates. C. albicans decreased from 112 cases (41.6%) to 102 cases (37.5%), while C. parapsilosis declined from 87 cases (32.5%) to 77 cases (28.3%). Candida glabrata increased from 22 cases (8.1%) to 25 cases (9.1%), C. tropicalis rose from 17 cases (6.3%) to 25 cases (9.1%), and C. auris increased from 2 cases (0.7%) to 7 cases (2.5%). Candida krusei decreased from 6 cases (2.2%) to 4 cases (1.4%), while unspecified Candida species increased from 23 cases (8.5%) to 32 cases (11.7%). The total cases were 269 in the pre-COVID period and 272 in the COVID and post period, chi-square analysis indicated these differences were not statistically significant (χ²=7.43, df = 6, p = 0.283). The distribution of clinical Candida spp. isolates is shown in Table 1.

Table 1.

Distribution of clinical Candida isolates by species and sample type, 2015–2019 (pre-COVID-19) and 2020–2023 (COVID-19 and post), n (%)

Candida species (= n, %)	Pre-COVID-19 (2015–2019)	COVID-19 and post (2020–2023)
Candida albicans (n = 214) 39,6%	112 (41,6%)	102 (37,5%)
Candida parapsilosis (n = 164) 30,3%	87 (32,5%)	77 (28,3%)
Candida glabrata (n = 47) 8,7%	22 (8,1%)	25 (9,1%)
Candida tropicalis (n = 42) 7,8%	17 (6,3%)	25 (9,1%)
Candida krusei (n = 10) 1,8%	6 (2,2%)	4 (1,4%)
Candida auris (n = 9) 1,66%	2 (0,7%)	7 (2,5%)
Candida spp. (n = 55) 10,1%	23 (8,5%)	32 (11,7%)
Totally (n = 541)	269 (49,7%)	272 (50,3%)

Cryptic Candida species include Candida lusitanıiae (Clavispora lusitaniae; n = 9), Candida lipolytica (Yarrowia lipolytica; n = 7), Candida kefyr (Kluyveromyces marxianus; n = 7), Candida dubliniensis (n = 4), Candida guilliermondii (Meyerozyma guilliermondii; n = 4), Candida pelliculosa (Wickerhamomyces anomalus; n = 3), Candida utilis (Cyberlindnera jadinii; n = 1), Candida inconspicua (n = 4), Candida orthopsilosis (n = 4), Candida blankii (n = 3), Candida auris (Candidozyma auris; n = 9) and Candida haemulonii (Candidozyma haemuli; n = 1). 8 Candida spp. isolates could not be identified at species level with the MALDI-TOF MS system.

When antifungal susceptibility results were evaluated, it was observed that C. parapsilosis complex isolates had reduced susceptibility to azole antifungal drugs frequently used in treatment. C. glabrata and C. parapsilosis complex were identified as species with high risk of resistance to azole treatment. In C. krusei considered intrinsically resistant to FLC among azole group antifungals, the first-line azole option is considered ineffective from the outset; therefore, the determined MIC range for voriconazole (VRC; 0.125–0.75 µg/mL) gains importance as a determining parameter in the selection of treatment strategy and the quantitative evaluation of the resistance phenotype. In addition, some C. albicans and C. glabrata isolates with reduced susceptibility to FLC are generally susceptible to VRC. Table 2 shows the Gradient test (E-test) antifungal susceptibility test results of Candida isolates.

Table 2.

Antifungal susceptibility of Candida isolates determined by gradient diffusion (E-test) method (MIC, µg/mL), including MIC range, MIC₅₀ and MIC₉₀ values

Candida species	Antifungal (= n)	MIC range (µg/mL)	MIC 50 (µg/mL)	MIC 90 (µg/mL)
Candida albicans	FLC (n = 151)	0,016 - >32	0,5 µg/ml	2 µg/ml
	VRC (n = 139)	< 0,002 – >32	0,023 µg/ml	0,064 µg/ml
	AMB (n = 145)	< 0,002–4	0,19 µg/ml	0,5 µg/ml
	AND (n = 82)	< 0,002–1,5	0,004 µg/ml	0,064 µg/ml
	MCF (n = 96)	0,004–1	0,023 µg/ml	0,064 µg/ml
Candida parapsilosis	FLC (n = 160)	0,032 - >256	3 µg/ml	> 32 µg/ml
	VRC (n = 154)	0,004 – >32	0,094 µg/ml	2 µg/ml
	AMB (n = 154)	< 0,002 – >32	0,25 µg/ml	1,5 µg/ml
	AND (n = 84)	< 0,002 – >32	1,5 µg/ml	4 µg/ml
	MCF (n = 120)	0,016 - >32	0,75 µg/ml	2 µg/ml
Candida glabrata	FLC (n = 47)	0,50 - >32	4 µg/ml	> 32 µg/ml
	VRC (n = 43)	0,012 - >32	0,125 µg/ml	8 µg/ml
	AMB (n = 47)	0,016 - >32	0,38 µg/ml	2 µg/ml
	AND (n = 32)	< 0,002–0,125	0,006 µg/ml	0,064 µg/ml
	MCF (n = 28)	0,006 − 1	0,012 µg/ml	0,38 µg/ml
Candida tropicalis	FLC (n = 42)	0,38 − 12	1,5 µg/ml	6 µg/ml
	VRC (n = 42)	0,008 − 2	0,125 µg/ml	0,25 µg/ml
	AMB (n = 42)	< 0,002–4	0,38 µg/ml	0,75 µg/ml
	AND (n = 31)	< 0,002–0,25	0,016 µg/ml	0,125 µg/ml
	MCF (n = 26)	0,016 − 0,25	0,032 µg/ml	0,125 µg/ml
Candida krusei	FLC (n = 10)	16 - >256	> 32 µg/ml	> 256 µg/ml
	VRC (n = 10)	0,125–0,75	0,5 µg/ml	0,5 µg/ml
	AMB (n = 10)	0,023 − 4	0,5 µg/ml	1 µg/ml
	AND (n = 3)	< 0,002–0,064	0,004 µg/ml	0,064 µg/ml
	MCF (n = 8)	0,094 − 0,25	0,125 µg/ml	0,19 µg/ml
Candida auris	FLC (n = 7)	> 32 µg/ml	> 32 µg/ml	> 32 µg/ml
	VRC (n = 7)	0,19 – >256	0,75 µg/ml	1,5 µg/ml
	AMB (n = 7)	0,006–1	0,38 µg/ml	1 µg/ml
	AND (n = 6)	0,25 − 1	0,50 µg/ml	1 µg/ml
	MCF (n = 3)	0,19–0,50	0,19 µg/ml	0,50 µg/ml
Candida spp.	FLC (n = 30)	0,047 - >32	2 µg/ml	> 32 µg/ml
	VRC (n = 29)	0,004 – >32	0,094 µg/ml	2 µg/ml
	AMB (n = 30)	< 0,002–8	0,38 µg/ml	0,75 µg/ml
	AND (n = 18)	< 0,002 - >32	0,032 µg/ml	2 µg/ml
	MCF (n = 20)	0,023 – >32	0,125 µg/ml	2 µg/ml

Candida spp.; E-test, Minimum inhibitory concentration (MIC) µg/mL Fluconazole: FLC, Voriconazole: VRC, Itraconazole: ITR, Amphotericin B: AMB, Anidulafungin: AND and Micafungin: MCF

Discussion

The major findings of our study analyzing 541 Candida isolates (2015–2023) demonstrated C. albicans (39%) and C. parapsilosis (29%) predominance. Notably, high-risk azole resistance emerged in C. glabrata and C. parapsilosis, while multidrug-resistant C. auris isolates were detected. These findings highlight evolving resistance patterns requiring enhanced surveillance and antifungal stewardship strategies.

Invasive Candida infections remain a significant global public health concern. Candidemia is associated with high morbidity and mortality, and despite the availability of antifungal agents, clinical outcomes remain unsatisfactory. Historically, C. albicans has been the predominant pathogen; however, in recent years a clear epidemiological shift toward non-albicans Candida (NAC) species has been observed [17,18]. This trend has been consistently confirmed by both large pan-European studies and local hospital-based investigations [19,20].

Surveillance conducted by the European Confederation of Medical Mycology (ECMM) demonstrated that the proportion of C. albicans declined from 56% in 1997 to 46% in 2018, with marked geographical variability [19]. For instance, across the three ECMM studies, C. albicans prevalence varied significantly by country, decreasing from 56.4% in the initial study (1997–1999) to 45.4% in the most recent study (2018–2022) [19]. These studies revealed persistently high FLC resistance in C. glabrata (12% of isolates in ECMM Candida III), the emergence of FLC-resistant C. parapsilosis particularly in Southern Europe (17% of isolates, primarily from Greece, Italy, and Turkey), and the identification of echinocandin-resistant strains harboring fks gene alterations [19]. Moreover, the global spread of C. auris, first documented in Asia in 2009 and now reported in more than 40 countries with a high mortality rate of 30–60%, has emerged as a critical concern, particularly given the increase in echinocandin-resistant strains despite echinocandins being the first-line treatment for C. auris infections [20].

Local epidemiological data corroborates these findings. In the intensive care units of our single center, a 15-year survey reported C. albicans (32%), C. parapsilosis (28%), C. glabrata (17%), and C. tropicalis (11%) as the leading pathogens [21]. In pediatric populations, C. parapsilosis and C. albicans were most common, whereas NAC species were more frequently isolated in non-neonatal children [22, 23]. Similarly, in Ecuador, C. tropicalis (38%) and C. albicans (37%) were most prevalent, while C. parapsilosis accounted for 48% of candidemia cases [24]. These variations reflect differences in patient demographics, device use, and regional epidemiology.

The distribution of Candida species showed some shifts between the pre-COVID-19 and post-COVID-19 periods, with decreases in C. albicans and C. parapsilosis and increases in C. glabrata, C. tropicalis, and C. auris. However, chi-square analysis indicated these differences were not statistically significant (p = 0.283). Similar stability in species distribution was reported in Korea, [25] whereas other studies observed increased candidemia incidence and a stronger role for non- albicans species during the pandemic [25, 26]. The rise of C. auris remains clinically important given its multidrug resistance and association with outbreaks in COVID-19 intensive care units [27].

The clinical relevance of species distribution lies in antifungal susceptibility. While C. albicans generally remains susceptible to most antifungal agents, alarming resistance trends have emerged among NAC species. More recent reports have shown FLC resistance rates in C. parapsilosis as high as 52% in Greece, 80.6% in Croatia, and 72.6% in Italy. [28, 29] Similarly, in European and national multi-country analyses C. glabrata exhibits high levels of azole resistance in Croatia and in Slovenia. [29, 30] Meanwhile, the rapid global emergence of C. auris since 2021, characterized by 92% FLC resistance and increasing echinocandin resistance, has established it as a critical multidrug-resistant pathogen [31].

These developments highlight the necessity of species-level identification and antifungal susceptibility testing for guiding therapy. Conventional systems such as VITEK 2 have shown significant limitations, with up to 85% inconsistency compared with ITS sequencing [24]. Therefore, the broader use of molecular diagnostic tools and region-specific surveillance is essential for accurate therapy [18, 32].

Despite therapeutic advances, mortality rates remain unacceptably high. ECMM studies reported a crude 30-day mortality of 37.9%, with the highest rates in infections caused by C. glabrata and C. tropicalis [20]. Mortality is further amplified among elderly patients, those with malignancies, and intensive care unit patients. Limited access to antifungal drugs in low- and middle-income countries also negatively impacts outcomes [19].

Our study also identified several rare and cryptic Candida species such as C. lusitaniae, C. kefyr, C. dubliniensis, C. guilliermondii, C. orthopsilosis, C. haemulonii, C. auris, among others. These rare species collectively represented 16% of our isolates. Although less frequent, these species are clinically relevant due to their potential for reduced susceptibility to azoles or AMB, underscoring the need for accurate identification and ongoing surveillance.

In terms of antifungal susceptibility, species-specific differences were observed as detailed in Tables 1 and 2. C. krusei showed intrinsic resistance to FLC but remained susceptible to VRC, while some C. albicans and C. glabrata isolates with reduced FLC susceptibility were generally responsive to VRC. These findings highlight that resistance patterns are not uniform across azoles and support the importance of detailed susceptibility testing for guiding therapy. Finally, the detection of multidrug-resistant C. auris (n = 9), though limited in number, is of particular concern given its global spread and limited treatment options, reinforcing the need for early recognition and strict infection control measures.

Limitations

Several limitations should be acknowledged. As a single-center study, the findings may not be representative of all Turkish healthcare settings. Molecular analysis of resistance mechanisms was not conducted, limiting insights into genetic determinants. Additionally, 8 isolates could not be identified at the species level by MALDI-TOF MS, highlighting current diagnostic limitations that may require molecular identification methods. Antifungal susceptibility testing was performed using a commercial gradient diffusion (E-test) method, and the results were not compared with the EUCAST reference microdilution method. The use of a non-reference AFST method may have led to underestimation of MIC values, especially for rare fungal species.

Conclusions

Our findings demonstrate a shift, though not statistically significant, toward non-albicans Candida species with emerging azole resistance, including the detection of multidrug-resistant C. auris. These trends underscore the need for enhanced local surveillance, molecular diagnostics, and evidence-based antifungal stewardship to optimize patient outcomes. Local surveillance programs and molecular diagnostics remain essential for effective resistance management.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Z.Y.: contributed to the conception, design, fundings, materials, data collection and/or processing, analysis and/or interpretation, literature review, and to the writing of the manuscript. G.A. contributed to the supervision, fundings, data collection and/or processing, analysis and/or interpretation, and critical review.

Funding

The authors received no specific grant from any funding agency.

Data availability

Openly available data.All data generated or analyzed during this study are included in this published article.

Declarations

Ethical approval

The research protocol was approved by the Istanbul University-Cerrahpaşa Rectorate Clinical Research Ethics Committee (decision number: 2024-1014569; Appendix 1). Compliance with the Helsinki Declaration was evaluated by the Istanbul University-Cerrahpaşa Rectorate Clinical Research Ethics Committee. The clinical isolates used in this study are yeast isolates obtained from patient samples sent to the Istanbul University-Cerrahpaşa Faculty of Medicine Hospital Medical Microbiology Mycology Laboratory between 2015 and 2024 and stored at -80°C. Since the study was retrospective and involved yeast (Candida spp.) isolates stored at -80°C, informed consent was not obtained for participation, and the study was evaluated by the Institutional Ethics Committee (IRB). (Informed Consent Forms were not obtained from patients because the study was planned retrospectively and based on archived samples.)

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.