Epidemiology And Antifungal Susceptibility Patterns Of Invasive Fungal Infections From 2012 To 2014 In A Teaching Hospital In Central China.

INTRODUCTION: As participants of the national China Hospital Invasive Fungal Surveillance Net program, we sought to describe the epidemiology and antifungal susceptibility patterns of yeast isolates obtained from patients with invasive fungal infection at the First Affiliated Hospital of Zhengzhou University, China.
METHODS: A total of 434 yeast isolates recovered from blood and other sterile body fluids were identified to species by matrix-assisted laser desorption ionization -time of flight mass spectrometry with or without supplementation by DNA sequencing. Antifungal susceptibilities were determined by Sensititre YeastOneTM YO10 methodology.
RESULTS: Candida albicans was the most common causative species (33.9% of isolates) but significantly decreased in frequency from 37.2% to 27.7% from 2012 to 2014. C. tropicalis was the next most common pathogen (25.1%), followed by C. parapsilosis complex (17.3%), C. glabrata (9%), and C. pelliculosa (6.7%), with other species comprising 8% of isolates. Caspofungin, micafungin, and anidulafungin exhibited potent in vitro activities against the majority of Candida isolates. Azoles demonstrated in vitro activities against C. albicans with a susceptibility rate of >95% and against C. parapsilosis complex, >95% isolates were susceptible. Among C. tropicalis and C. glabrata isolates, resistance rates to fluconazole and voriconazole were 11.9%, 9.1% and 7.7%, 28.2%, respectively. Of note, C. pelliculosa had a high incidence rate in newborns and high rates of resistance to fluconazole and voriconazole of 55.2% and 41.4%, respectively.
CONCLUSION: The present study provided valuable local surveillance data on the epidemiology and antifungal susceptibilities of invasive yeast species, which is essential for guiding antifungal treatment protocol development.

Introduction

Invasive fungal infections (IFIs) are major global health threat, particularly the immunocompromised and critically ill, and are widely recognized as a major cause of substantial morbidity and mortality and excess hospital costs.1,2 Although Candida albicans remains the most predominant species responsible for invasive infections, its dominance has decreased as non-C. albicans yeast species are increasingly encountered.3,4 It is widely appreciated that non-C. albicans yeast species are often less susceptible to antifungal drugs than C. albicans which impact on clinical outcomes.5,6

Epidemiology data including species distribution and antifungal susceptibility profile of the causative yeast species are essential to improve the overall outcomes.5,7 However, species distribution varies considerably with geographic and institution or hospital, and even within any one institution.8,9 Targeted antifungal therapy or antifungal prophylaxis regimens are contingent on the timely diagnosis and antifungal susceptibility testing.8,10 Henan is the third largest province in China with a population of about 96 million, and knowledge of species distribution and their antifungal susceptibility profile of invasive yeast infections in this populous region will have substantial clinical impact.

The national China Hospital Invasive Fungal Surveillance Net (CHIF-NET) program across China has provided updated information on epidemiological data for IFIs.3,5,11 As a participant of this national surveillance program, the present study sought to describe the epidemiology and antifungal susceptibility patterns of yeast isolates obtained from IFI patients who presented to the First Affiliated Hospital of Zhengzhou University in Zhengzhou, China.

Materials And Methods

Study Design And Ethics Statement

Yeast isolates were collected consecutively over the 3-year study period (August 2012 to July 2014) from patients admitted to the First Affiliated Hospital of Zhengzhou University enrolled in the CHIF-NET study. The First Affiliated Hospital of Zhengzhou University is an 8000-bed major tertiary teaching hospital in central China with a range of specialty services. Non-duplicated isolates recovered from blood or other sterile body fluids were included according to the criteria described previously.11 The study was approved by the Human Research Ethics Committee of Peking Union Medical College Hospital (No. S-263).

Species Identification

Yeast isolates were initially identified at the First Affiliated Hospital of Zhengzhou University using chromogenic agar and the VITEK 2® compact system (bioMérieux, Marcy l’Etoile, France). Accurate identification was confirmed to species level at a central laboratory (Peking Union Medical College Hospital) by matrix-assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF MS) using the VITEK MS system (v2.0, IVD database, bioMérieux) supplemented with internal transcribed spacer (ITS) sequencing as previously described.12 DNA extraction and amplification of the ITS region was carried out with primer pairs ITS1/ITS4 as described by Zhang et al.12 The PCR products were sequenced in both directions using corresponding PCR amplification primer pairs at Ruibiotech Co. Ltd. (Beijing, China) using the DNA analyzer ABI 3730XL system (Applied Biosystems, Foster City, CA). Species identification was performed by comparing the obtained sequences against GenBank database with nucleotide Basic Local Alignment Search Tool (BLASTn, ).

Antifungal Susceptibility Testing

The in vitro susceptibility to nine antifungal drugs (fluconazole, voriconazole, itraconazole, posaconazole, caspofungin, micafungin, anidulafungin, amphotericin B, and 5-flucytosine) was determined by Sensititre YeastOneTM YO10 methodology (Thermo Scientific, USA) following the manufacturer’s instructions. Minimum inhibitory concentration (MIC) values for yeast isolates were determined by the Sensititre YeastOne YO10 method and interpreted by species-specific clinical breakpoints (CBPs) as recommended by the CLSI M60 method.13 With regard to species for which there are no CBPs we used epidemiological cutoff values (ECVs) to differentiate wild-type (WT) from non-WT isolates according to CLSI M59 method14 and ECVs from the study by Pfaller et al.15 C. parapsilosis ATCC 22019 and C. krusei ATCC 6258 were the quality control strains.

Statistical Analysis

All statistical analyses were performed using IBM SPSS software (version 24.0; IBM SPSS Inc., New York, USA). Categorical variables were compared using χ2 or Fisher’s exact test. A P value of 0.05 was significant.

Results

Patient Characteristics And Species Distribution

Table 1 summarizes the patient characteristics and species distribution. Overall, a total of 434 non-duplicated yeast isolates from separate patients were collected; there were more isolates from males than females (57.4 vs 42.6%, respectively). Isolates were cultured from 434 patients aged 0 to 96 years (average age 50.8 years). Invasive infection occurred mainly in the age groups 15–49 and over 65 years, accounting for 32% (139/434) and 30.9% (134/434), respectively. For infants (aged 0–1 year), C. pelliculosa was a predominant pathogen with an isolation rate the same as that of C. albicans (both 35.5%) (see species distribution below). The majority of the isolates were from patients in the intensive care unit (ICU) (168, 38.7%), surgical departments (152, 35%), and medical departments (51, 11.8%).

Table 1

Clinical Characteristics Of Patients And Species Distribution Of The Yeast Isolates

Patients Characteristic	Number (%) Of Isolates
Total	Candida Albicans	Candida Tropicalis	Candida Parapsilosis Complexa	Candida Glabrata	Candida Pelliculosa	Cryptococcus Neoformans	Othersb
Gender
Female	185 (42.6)	63 (34.1)	40 (21.6)	26 (14.1)	24 (13.0)	13 (7.0)	6 (3.2)	13 (7.0)
Male	249 (57.4)	84 (33.7)	69 (27.7)	49 (19.7)	15 (6.0)	16 (6.4)	5 (2.0)	11 (4.4)
Age (years)
0–1	31 (7.1)	11 (35.5)	5 (16.1)	1 (3.2)	3 (9.7)	11 (35.5)	-	-
2–14	15 (3.5)	3 (20.0)	7 (46.7)	-	-	1 (6.7)	4 (26.6)	-
15–49	139 (32.0)	41 (29.5)	38 (27.4)	33 (23.7)	10 (7.2)	5 (3.6)	2 (1.4)	10 (7.2)
50–65	115 (26.5)	43 (37.4)	28 (24.3)	14 (12.2)	11 (9.6)	6 (5.2)	5 (4.3)	8 (7.0)
>65	134 (30.9)	49 (36.6)	31 (23.1)	27 (20.1)	15 (11.2)	6 (4.5)	-	6 (4.5)
Ward Type
ICU	168 (38.7)	69 (41.1)	43 (25.6)	28 (16.6)	15 (8.9)	4 (2.4)	4 (2.4)	5 (3.0)
Surgical department	152 (35.0)	56 (36.8)	39 (25.7)	25 (16.4)	13 (8.6)	10 (6.6)	-	9 (5.9)
Medical department	51 (11.8)	(713.7)	12 (23.5)	12 (23.5)	5 (9.8)	3 (5.9)	6 (11.8)	6 (11.8)
Others	63 (14.5)	15 (23.8)	15 (23.8)	10 (15.9)	6 (9.5)	12 (19.0)	1 (1.6)	4 (6.4)
Separated year
2012	129 (29.7)	48 (37.2)	34 (26.4)	31 (24.0)	6 (4.7)	3 (2.3)	1 (0.7)	6 (4.7)
2013	114 (26.3)	46 (40.3)	17 (14.9)	19 (16.7)	17 (14.9)	6 (5.3)	4 (3.5)	5 (4.4)
2014	191 (44.0)	53 (27.7)	58 (30.4)	25 (13.1)	16 (8.4)	20 (10.5)	6 (3.1)	13 (6.8)
Separated sites
Blood	224 (51.6)	72 (32.1)	52 (23.2)	36 (16.1)	27 (12.1)	22 (9.8)	3 (1.3)	12 (5.4)
CVC	66 (15.2)	17 (25.8)	17 (25.8)	20 (30.3)	1 (1.4)	6 (9.1)	-	5 (7.6)
pus	58 (13.4)	22 (37.9)	16 (27.6)	11 (19.0)	5 (8.6)	-	-	4 (6.9)
Ascitic fluid	38 (8.8)	19 (50.0)	10 (26.3)	3 (7.9)	4 (10.5)	-	-	2 (5.3)
Pleural fluid	13 (3.0)	8 (61.5)	2 (15.4)	1 (7.7)	-	-	2 (15.4)	-
CSF	12 (2.7)	3 (25.0)	4 (33.3)	1 (8.4)	-	-	4 (33.3)	-
Bile	7 (1.6)	2 (28.6)	4 (57.1)	-	1 (14.3)	-	-	-
Joint fluid	6 (1.4)	1 (16.6)	1 (16.6)	2 (33.6)	1 (16.6)	-	-	1 (16.6)
Tissue	6 (1.4)	2 (33.3)	2 (33.3)	-	-	1 (16.7)	1 (16.7)	-
Others	4 (0.9)	1 (20.0)	1 (20.0)	1 (20.0)	-	-	1 (20.0)	-
Total	434	147 (33.9)	109 (25.1)	75 (17.3)	39 (9.0)	29 (6.7)	11 (2.5)	24 (5.5)

Notes: aThis includes Candida parapsilosis sensu stricto (69 isolates), Candida metapsilosis (3 isolates), Candida orthopsilosis (3 isolates). bThis included Candida krusei (8 isolates), Candida guilliermondii (4 isolates), Candida haemulonii (4 isolates), Candida lusitaniae (1 isolate), Candida intermedia (1 isolate), Kodamaea ohmeri (3 isolates), Rhodotorula mucilaginosa (2 isolates) and Trichosporon asahii (1 isolate).

Overall, 19 species of yeast were identified among the 434 isolates of which C. albicans was the most common species (147 isolates, 33.9%). Non-albicans Candida species accounted for 270 isolates (62.2%) – among these, C. tropicalis was the most frequent species (25.1%, 109 isolates). C. parapsilosis complex was the third common species (75 isolates, 17.3%), which consisted of C. parapsilosis sensu stricto (69/75 isolates, 15.9%), C. metapsilosis (3/75 isolates, 0.7%), and C. orthopsilosis (3/75 isolates, 0.7%), followed by the C. glabrata (39 isolates, 9%) and C. pelliculosa (29 isolates, 6.7%). In addition, there were small numbers of isolates of other Candida species including C. krusei (8 isolates), C. guilliermondii (4 isolates), C. haemulonii (4 isolates), and each one isolate of C. intermedia and C. lusitaniae. Cryptococcus neoformans (11 isolates, 2.5%) was the most common non-Candida yeast species, followed by Kodamaea ohmeri (3 isolates), Rhodotorula mucilaginosa (2 isolates) and Trichosporon asahii (1 isolate) (Table 1).

Over the 3-year study period, the frequency of all yeast isolates decreased from 29.7% to 26.3% then markedly increased to 44% from 2012 to 2014 (P<0.001), with the frequency of non-Candida species also moderately increasing from 3.1% to 4.7% (P<0.5). With regard to individual species, the isolation rate of C. albicans decreased significantly from 37.2% to 27.7% (P=0.095), whilst the isolation rate of non-albicans Candida reached up to 29.7% in 2014 (compared with 17.7% in 2012, P<0.001). Of these, the proportion of C. tropicalis of 30.4% collected from 2014 exceeded the frequency of 26.4% and 14.9% collected during 2012 and 2013, respectively (P=0.5). The isolation rate of C. glabrata was14.9% and 8.4% of 2013 and 2014, respectively, higher than that in 2012 (4.7%, P<0.5). Of note, the isolation rate of C. pelliculosa rose from 2.3% to 10.5% during the 3 years (P<0.05). Conversely, the frequency of C. parapsilosis complex declined from 24% to 13.1% from 2012 to 2014 (P<0.05).

Table 1 also shows species distribution of the 434 yeast isolates according to the source of specimens. The most common specimen type was blood (51.6%), followed by a central venous catheter (CVC; 15.2%), pus (13.4%), ascitic fluid (8.8%), pleural fluid (3.0%), cerebrospinal fluid (2.8%), bile (1.6%), joint fluid (1.4%), tissue (1.4%). Other than C. metapsilosis and T. asahii, the other 14 species caused bloodstream infections. C. albicans (72/224 isolates, 32.1%) remained the most common pathogen for bloodstream infections, followed by C. tropicalis (52/224 isolates, 23.2%). The proportion of non-albicans Candida isolates recovered from blood cultures (146/434 isolates, 33.6%) was significantly higher than that recovered from other specimen types (124/434 isolates, 28.6%) (P<0.5). The frequency between blood source and non-blood source isolates was similar for C. tropicalis (12% versus 13.1%) and C. parapsilosis complex (8.3% versus 9%). However, there was a significant difference in proportions of blood vs non-blood source for C. glabrata (6.2% versus 2.7%, P<0.05) and C. pelliculosa (5.1% versus 1.6%, P<0.01). There were also a few uncommon yeast species causing bloodstream infections, such as R. mucilaginosa, C. intermedia and K. ohmeri ().

Concordance Of Initial And Final Identification Results

The agreement of identification results for yeast isolates obtained from the First Affiliated Hospital of Zhengzhou University and from the central laboratory is presented in Table 2. Overall concordance was 71.2% (309/434) with the highest observed for C. albicans (97.3%,143/147), while concordance much lower (0% to 76.9%) for other species. In particular, all isolates of C. metapsilosis, C. orthopsilosis, K. ohmeri, R. mucilaginosa, C. intermedia, C. lusitaniae, T. asahii were incorrectly identified or unable to be identified at the First Affiliated Hospital of Zhengzhou University.

Table 2

Comparison Of Identification Results Of The 434 Isolates Between Those Obtained From The First Affiliated Hospital Of Zhengzhou University (Initial identification) And From The Central Laboratory (Final identification)

Final Identification	Number Of Isolates	Concordance Of Initial And Final Identification Results
		Number (%) Of Isolates With Concordance	Number (%) Of Isolates With Identifcation Errors
Major Error	Misidentification (Number)
Candida albicans	147	143(97.3)	4(2.7)	C. parapsilosis (2), C. tropicalis (1), Stephanoascus ciferrii (1)
Candida tropicalis	109	75(68.8)	34(31.2)	C. albicans (7), C. krusei (7), C. famata (6), C. parapsilosis (5), C. glabrata (4), C. laurentii (2), C. guilliermondii (1), C. pelliculosa (1), C. neoformans (1)
Candida parapsilosis complex	75	39(52)	36(48)
Candida parapsilosis sensu stricto	69	39 (56.5)	30(43.5)	C. famata (12), C. laurentii (5), C. albicans (3), C. glabrata (3), C. krusei (2), C. rugosa (2), C. tropicalis (1), C. pelliculosa (1)
Candida metapsilosis	3	–	3(100)	C. krusei (2), C. laurentii (1)
Candida orthopsilosis	3	–	3(100)	C. krusei (2), C. famata (1)
Candida glabrata	39	30(76.9)	9(23.1)	C.parapsilosis (2), C.tropicalis (2), S.ciferrii (2), C.albicans (1), C.laurentii (1), Rhodotorula spp. (1)
Candida pelliculosa	29	8(27.6)	21(72.4)	C. glabrata (11), C. krusei (5), C. parapsilosis (2), C. albicans (1), C. norvegensis (1), C. laurentii (1)
Cryptococcus neoformans	11	8(72.8)	3(27.2)	S. ciferrii (1), C. laurentii (1), C. krusei (1)
Candida krusei	8	4(50.0)	4(50.0)	C. lipolytica (1), C. glabrata (1), C. pelliculosa (1), C. neoformans (1)
Candida guilliermondii	4	1(25.0)	3(75.0)	C. glabrata (2), C. parapsilosis(1)
Candida haemulonii	4	1(25.0)	3(75.0)	C. glabrata (2), C. krusei (1)
Kodamaea ohmeri	3	–	3(100.0)	C. guilliermondii (1), C. pelliculosa (1) C. albicans (1)
Rhodotorula mucilaginosaa	2	–	–	–
Candida intermedia	1	–	1(100.0)	C. glabrata (1)
Candida lusitaniae	1	–	1(100.0)	C. laurentii (1)
Trichosporon asahii	1	–	1(100.0)	C. albicans (1)
Total	434	309(71.2)	123(28.3)	–

Note: aOnly two Rhodotorula mucilaginosa isolates occurred during minor error identification and misidentified to genus level as Rhodotorulaspp.

Further, accurate identification results to species level by conventional methods from the local hospital showed a relatively high error rate for non-C. albicans isolates: C. tropicalis (75/109, 68.8%), C. parapsilosis complex sensu stricto (39/69, 56.5%), C. glabrata (30/39, 76.9%), C. pelliculosa (8/29, 27.6%), C. neoformans (8/11, 72.8%), C. krusei (4/8, 50%), C. guilliermondii (1/4, 25%) and C. haemulonii (1/4, 25%). Other uncommon minority species were prone to incorrect identification. Only two R. mucilaginosa occurred during minor error identification and misidentified to genus level as Rhodotorulaspp. (Table 2).

In Vitro Susceptibilities

The antifungal susceptibilities of the 434 yeast isolates are presented in Table 3. All three echinocandins exhibited potent in vitro activities against the majority of Candida isolates. All 75 isolates of C. parapsilosis complex were susceptible to anidulafungin, micafungin, and caspofungin with MIC50 of 1, 1, and 0.5 μg/mL. All the C. albicans were susceptible to micafungin and caspofungin, except one isolate (0.7%) showed MIC value to anidulafungin of 0.5 μg/mL. Decreased susceptibility to echinocandins was observed among C. tropicalis isolates, two isolates displayed resistant to anidulafungin, micafungin, and caspofungin with MICs of 2 and over 8 μg/mL, 2 and over 8 μg/mL, and 1 and over 8 μg/mL. There was one C. tropicalis isolate that showed MIC value to anidulafungin and caspofungin of 0.5 and 0.5 μg/mL, respectively. Each C. glabrata isolate showed intermediate susceptible to anidulafungin or micafungin with MIC of 0.25 or 0.12 μg/mL, respectively. All echinocandins had MICs of over 8 μg/mL against C. neoformans and T. asahii.

Table 3

In Vitro Susceptibility Results Of 434 Isolates To Nine Antifungal Agents

Species And Agents	MIC (μg/mL)			Number (%) Of Isolates In Each Category
Range	50%	90%	S/WT	I/SDD	R/non-WT
Candida albicans(n=147)
Anidulafungin	≤0.015–0.5	≤0.015	0.12	146(99.3)	1(0.7)
Micafungin	≤0.008–0.25	≤0.008	0.15	147(100)
Caspofungin	0.015–0.25	0.03	0.06	147(100)
Posaconazole	≤0.08–0.5	0.015	0.03	144(98)		3(2)
Voriconazole	≤0.08–0.5	≤0.08	0.15	141(95.9)	6(4.1)
Itraconazole	≤0.015–0.25	0.03	0.06	146(99.3)	1(0.7)
Fluconazole	≤0.12–8	0.25	0.5	146(99.3)		1(0.7)
5-Flucytosine	≤0.06->64	≤0.06	0.12	145(98.6)		2(1.4)
Amphotericin B	≤0.12–2	0.5	1	147(100)
Candida tropicalis (n=109)
Anidulafungin	≤0.015->8	0.12	0.25	106(97.2)	1(0.9)	2(1.8)
Micafungin	≤0.008->8	0.03	0.03	107(98.2)		2(1.8)
Caspofungin	0.015->8	0.03	0.25	106(97.2)	1(0.9)	2(1.8)
Posaconazole	0.015–8	0.25	0.6	50(45.9)		56(54.1)
Voriconazole	≤0.008->8	0.12	0.5	67(61.5)	32(29.4)	10(9.1)
Itraconazole	0.03->16	0.25	0.5	100(91.7)		9(8.3）
Fluconazole	0.25->256	2	16	86(78.9）	10(9.2）	13(11.9）
5-Flucytosine	≤0.06–64	≤0.06	0.12	107(98.2）		2(1.8）
Amphotericin B	≤0.12–4	1	1	107(98.2）		2(1.8）
Candida parapsilosis complex (n=75)a
Anidulafungin	0.03–2	1	2	75(100）
Micafungin	0.06–2	1	1	75(100）
Caspofungin	0.06–1	0.5	0.5	75(100）
Posaconazole	≤0.008–0.5	0.03	0.06	72(96）		3(4）
Voriconazole	≤0.008–8	≤0.008	0.03	71(94.7）		4(5.3）
Itraconazole	≤0.015–0.5	0.03	0.12	75(100）
Fluconazole	≤0.12–256	0.5	2	71(94.7）		4(5.3）
5-Flucytosine	≤0.06->64	≤0.06	0.25	70(93.3）		5(6.7）
Amphotericin B	0.25–2	0.5	1	75(100）
Candida glabrata (n=39)
Anidulafungin	≤0.015–0.25	0.03	0.12	38(97.4）	1(2.6）
Micafungin	≤0.008–0.12	0.015	0.03	38(97.4）	1(2.6）
Caspofungin	0.03–0.12	0.06	0.12	39(100）
Posaconazole	0.03->8	0.5	2	35(89.7）		4(10.3）
Voriconazole	0.015->8	0.25	1	28(71.8）		11(28.2）
Itraconazole	0.03->16	0.5	1	36(92.3）		3(7.7）
Fluconazole	0.5->256	8	32	36(92.3）		3(7.7）
5-Flucytosine	≤0.06->64	≤0.06	≤0.06	38(97.4）		1(2.6）
Amphotericin B	0.25–4	1	1	38(97.4）		1(2.6）
Candida pelliculosa (n=29)
Anidulafungin	≤0.015–0.06	≤0.015	≤0.015
Micafungin	0.03–0.06	0.03	0.06
Caspofungin	≤0.008–0.12	0.06	0.06	29(100）
Posaconazole	0.12->8	0.5	2	28(96.6）		1(3.4）
Voriconazole	0.06->8	0.25	0.5	17(58.6）		12(41.4）
Itraconazole	0.06->16	0.25	0.5
Fluconazole	2->256	8	16	13(44.5）		16(55.2）
5-Flucytosine	≤0.06->64	2	32
Amphotericin B	0.25–1	1	1
Cryptococcus neoformans (n=11)
Anidulafungin	>8	>8	>8
Micafungin	>8	>8	>8
Caspofungin	>8	>8	>8
Posaconazole	0.06–0.25	0.12	0.12
Voriconazole	0.03–0.06	0.03	0.06
Itraconazole	0.03–0.12	0.06	0.12
Fluconazole	2–8	4	4
5-Flucytosine	0.5–4	2	4
Amphotericin B	0.5–1	0.5	1
Other species (n=24)b
Anidulafungin	≤0.015->8	1	2
Micafungin	0.015->8	0.12	0.5
Caspofungin	0.03->8	0.12	2
Posaconazole	0.015–1	0.25	1
Voriconazole	≤0.008->8	0.25	>8
Itraconazole	0.03->16	0.25	1
Fluconazole	0.5->256	16	>256
5-Flucytosine	≤0.06->64	8	64
Amphotericin B	≤0.12–4	0.25	4

Note: aThis includes Candida parapsilosis sensu stricto (69 isolates), Candida metapsilosis (3 isolates), Candida orthopsilosis (3 isolates). bThis included Candida krusei (8 isolates), Candida guilliermondii (4 isolates), Candida haemulonii (4 isolates), Candida lusitaniae (1 isolate), Candida intermedia (1 isolate), Kodamaea ohmeri (3 isolates), Rhodotorula mucilaginosa (2 isolates) and Trichosporon asahii (1 isolate).

Abbreviations: S, Susceptible; SDD, Susceptible-dose dependent; I, Intermediate; WT, Wild-type; R, Resistant.

Of the azoles, C. albicans remained susceptible to all azoles tested during the 3 years (over 95% susceptible). Of these, voriconazole had a susceptibility rate of 95.9% (141/147 isolates) compared with susceptibility rates of 98%, 99.3%, and 99.3%, for posaconazole, itraconazole, and fluconazole, respectively. Azoles also demonstrated potential in vitro activities against C. parapsilosis complex, with susceptibility rate to posaconazole, voriconazole, itraconazole, and fluconazole of 96%, 94.7%, 100%, and 94.7%, respectively. However, two isolates of C. orthopsilosis exhibited high MICs to posaconazole, voriconazole, and fluconazole (0.5, 8, and 256 μg/mL, respectively). There were also one C. parapsilosis sensu stricto isolates resistant to posaconazole, voriconazole, and fluconazole with MIC of 0.51 and 16 μg/mL, and another C. orthopsilosis isolate resistant to voriconazole and fluconazole with MIC of 0.25 and 8 μg/mL, respectively. Significant decreased susceptibility was mostly observed in C. tropicalis, and the resistance rate of this species to posaconazole, voriconazole, itraconazole, and fluconazole of 54.1%, 9.1%, 8.3%, and 11.9%, respectively. In addition, the non-susceptibility of C. tropicalis to both fluconazole and voriconazole clearly increased to 27.5% and 44.8% in 2014, distinctly higher than those of 14.7% and 32.4% in 2012, but no statistical significance (P<0.5) was shown. Similarly, a decreased susceptibility trend was observed among the C. glabrata isolates, especially for the overall voriconazole resistance rate already up to 28.2%, followed by the resistance rate to posaconazole, itraconazole, and fluconazole of 10.3%, 7.7%, and 7.7%, respectively. C. pelliculosa exhibited extremely high rates of resistance to fluconazole and voriconazole that reach up to 55.2% and 41.4%, respectively (Table 3).

Azoles showed a good in vitro antifungal activity against C. neoformans to posaconazole, voriconazole, itraconazole, and fluconazole with the MIC90 of 0.12, 0.06, 0.12, and 4 μg/mL, respectively. Other uncommon yeast species displayed generally high MIC values to four azole agents. It is noteworthy that three of four C. haemulonii isolates showed high MIC values to posaconazole, voriconazole, itraconazole, and fluconazole of 1, >8, >16, and >256 μg/mL, respectively.

All 147 C. albicans isolates showed a WT phenotype to amphotericin B while two isolates of this species (1.4%) were non-WT phenotype to 5-flucytosine. And, 2.6% of C. glabrata (one isolate) and 1.8% of C. tropicalis (two isolates) had non-WT phenotype both to amphotericin B and 5-flucytosine. All isolates of C. parapsilosis complex were susceptible to amphotericin B, but 6.7% of this species (5 isolates) were resistant to 5-flucytosine. C. pelliculosa and other yeast species possessed higher MIC50 to flucytosine of 2 and 8 μg/mL, respectiveluy. The MIC50 value of amphotericin B for all species tested were 0.25 to 1 μg/mL.

Discussion

This retrospective observational study describes a laboratory-based, 3-year surveillance of invasive yeast infections in the First Affiliated Hospital of Zhengzhou University. Although national data have been published, local data of species distribution, patient characteristics, and antifungal susceptibility profiles are essential to inform early and effective treatment for Candida infections.

Although there are many large-scale, global surveillance programs identifying species distribution and resistance trends of IFIs,5,9,15 similar studies in medical institutions within China are few.16–19 Invasive candidiasis remains the most common invasive yeast infection, accounting for 96.1% of the episodes in this study. Generally, although C. albicans remains the dominant species, it accounted for only 33.9% of the isolates collected in this study, lower than that overall in the Asia-Pacific region (46.3%),9 another Chinese study (43–47.4%)5 and more specifically, in Beijing (50.3%)10 during the same period. C. tropicalis was the most common non-albicans Candida species (25.1%), which again differs from the epidemiology reported in the Asia-Pacific region of SENTRY antimicrobial surveillance program, where C. glabrata was second most common species (19.3%)9 and contrasting with the CHIF-NET Study in China.5 However, the relative frequency of C. tropicalis in our study was similar to other single-center studies.16,18 Of interest, we observed a relatively high frequency of C. pelliculosa (6.7%) isolates at our hospital (similar to the frequency of C. glabrata, 9%) with the majority of isolates being from blood (81.5%). The above findings emphasize the necessity to perform locally relevant epidemiological studies.

From the clinical perspective, the frequency of IFIs in the patients older than 65 years deserves attention in keeping with previous observations that infection with some common Candida species, such as C. albicans, C. parapsilosis, C. tropicalis, and C. glabrata, may be associated with older age.10,16,20 Conversely, the isolation of C. pelliculosa in our study was associated with disease in newborns (35.5%). There have been reports of nosocomial transmission of C. pelliculosa fungemia in the neonatal ICU,21,22 and clinicians should be vigilant about the potential presence of this species. Our study also confirmed that patients admitted to ICUs and surgical department pose a high risk of developing IFIs.17,23 MALDI-TOF MS has been extensively proven as a powerful tool in the identification of yeast species compared with traditional identification methods.12,26 To this end, incorrect identification results in our study using the CHROMagar Candida medium and the VITEK 2® compact system not only occurred for uncommon Candida isolates but also for common Candida species (error rates from 2.7% of C. albicans up to 48% of the C. parapsilosis complex). For the common Candida species, conventional identification methods especially exhibited poor identification performance to the C. parapsilosis complex with 48% of the error identification rate. Of these, C. parapsilosis sensu stricto were most indistinguishable from C. famata and C. laurentii while C. metapsilosis and C. orthopsilosis were liable to misidentified as C. krusei. Some studies have highlighted the inferior identification performance of common commercial systems to C. parapsilosis complex, and C. famata has also been reported as the most primary species misidentified from C. parapsilosis sensu stricto.27,28 As is well known, the performance of conventional identification methods to less commonly encountered yeast species is unsatisfactory.29,30 We observed high misidentification rates in the identification of C. pelliculosa (72.4%), C. krusei (50%), C. guilliermondii (75%), C. haemulonii (75%), and K. ohmeri (100%). In consideration of molecular techniques associated with increased costs, longer turn-around time and the need for considerable expertise,29 Zhang et al proposed an algorithm of MALDI-TOF MS supplied with DNA sequencing for yeast identification that can be applied to epidemiological investigation and the routine laboratory identification.12

C. albicans exhibited low antifungal resistance rates as has been shown globally.5,9,10 The majority of the C. parapsilosis complex isolates were susceptible (or WT) to all nine antifungals tested (≥93.3%), with a 5.3% of resistance rate to fluconazole and voriconazole. The findings were in general higher than those obtained from the national CHIF-NET surveillance.5,31 Among the C. parapsilosis complex, C. orthopsilosis was the most resistant species and had very high MICs to azoles (8 or 256 μg/mL for fluconazole). Although C. orthopsilosis is relatively uncommon, it is essential to perform accurate identification and antifungal susceptibility test to detect such species.32 High-level azole resistance was mainly observed in C. tropicalis, with resistance rates to fluconazole of 11.9%, and to voriconazole of 9.1%, both higher than that seen in the SENTRY Antifungal Surveillance Program (1997–2016) (9.2% fluconazole-resistance),9 national level (2010–2012) (both of 5.7%),31 and a teaching hospital in southwest China (2.2% and 0%),16 but similar to that in the national level during a 5-year surveillance (12.8% and 11.4%)33 and Beijing hospitals (both of 9.4%).10 Fan et al have underlined the notable increasing trend in azole non-susceptible invasive C. tropicalis infection in China. Our research data accorded with this epidemical tendency with an even higher voriconazole non-susceptible rate.33 Thus, continuous surveillance, molecular epidemiology, and resistance mechanism study are essential, and empirical therapeutic strategies for C. tropicalis invasive infections may be modified in China.33,34 Moreover, in vitro susceptibility results showed that only 7.7% of C. glabrata isolates were resistant or non-WT to all four azoles similar to results reported in a 5-year multicenter study based on CHIF-NET program35 but contrast with results in the USA.36 Although a rising echinocandin resistance of C. glabrata in the USA poses a serious challenge for clinical therapy,37 the fact that only one isolate (2.6%) herein was considered to be of “intermediate susceptibility” to anidulafungin and micafungin is reassuring. Interestingly, our results showed that a high proportion of C. pelliculosa isolates – 55.2% and 41.4% were non-WT to fluconazole and voriconazole, respectively. Such high resistance rates have not been previously reported11,38. Other yeast species, although uncommon, exhibited high MIC values to the four azole agents. For example, C. haemulonii was highly resistant to the azoles, amphotericin B, and 5-flucytosine but had low MICs for the echinocandins in our study. This resistance pattern is similar to that reported by Hou et al in China and Ramos et al in Brazil.39,40

Two main limitations of this study are mentioned. First, we used the YO10 methodology to perform antifungal susceptibility testing and not a reference broth microdilution method. However, the essential agreement between this methodology and the CLSI and the EUCAST reference procedures had been proven to be very high for yeast species.35,41,42 Importantly, surveillance work was interrupted for a number of reasons beyond our control from 2014 to 2016 limiting the collection of more complete epidemiological data.

In conclusion, the present study provides valuable local surveillance data on the epidemiology and antifungal susceptibilities of invasive yeast species isolated from the First Affiliated Hospital of Zhengzhou University, which can be used to guide the selection of both empiric and targeted antifungal therapy. Although C. albicans remained the most prevalent species, the frequency of C. tropicalis and its notable and increasing trend azole non-susceptibility requires noting. Less common Candida species also exhibited high azole resistance rates, continuously monitoring of local epidemiology of invasive yeast infection and of antifungal resistance is warranted.