ABSTRACT
The objectives of our study were to describe the characteristics of patients with Candida guilliermondii candidemia and to perform an in-depth microbiological characterization of isolates and compare them with those of patients with C. albicans candidemia. We described the risk factors and outcomes of 22 patients with candidemia caused by the C. guilliermondii complex. Incident isolates were identified using molecular techniques, and susceptibility to fluconazole, anidulafungin, and micafungin was studied. Biofilm formation was measured using the crystal violet assay (biomass production) and the XTT reduction assay (metabolic activity), and virulence was studied using the Galleria mellonella model. Biofilm formation was compared with that observed for C. albicans. The main conditions predisposing to infection were malignancy (68%), immunosuppressive therapy (59%), and neutropenia (18%). Clinical presentation of candidemia was less severe in patients infected by the C. guilliermondii complex than in patients infected by C. albicans, and 30-day mortality was lower in C. guilliermondii patients (13.6% versus 33.9%, respectively; P = 0.049). Isolates were identified as C. guilliermondiisensu stricto (n = 17) and Candida fermentati (n = 5). The isolates produced biofilms with low metabolic activity and moderate biomass. The G. mellonella model showed that C. guilliermondii was less virulent than C. albicans (mean of 6 days versus 1 day of survival, respectively; P < 0.001). Patients with candidemia caused by the C. guilliermondii complex had severe and debilitating underlying conditions. Overall, the isolates showed diminished susceptibility to fluconazole and echinocandins, although poor biofilm formation and the low virulence were associated with a favorable outcome.
INTRODUCTION
Candida albicans is the main cause of candidemia, although a shift toward other non-albicans Candida species, such as C. parapsilosis, C. tropicalis, C. glabrata, and C. krusei, has been reported (1–4). The remaining infrequent Candida species account for 8% of episodes (5); however, their epidemiological characteristics remain largely unknown.
C. guilliermondii is a complex comprising several species, of which 3 have been reported as human pathogens (C. guilliermondiisensu stricto [Meyerozyma guilliermondii as the teleomorph form], C. fermentati [M. caribbica as the teleomorph form], and C. carpophila). The proportion of candidemia episodes caused by the C. guilliermondii complex has been poorly studied, although it is thought that this species is involved in up to 2% of cases (3, 6–9). The C. guilliermondii complex is characterized by diminished susceptibility to azoles and echinocandins (9, 10) and has been reported mostly as a cause of candidemia in patients with cancer (11, 12). However, the low number of cases reported means that knowledge of the epidemiology of candidemia caused by this complex is limited (9, 11).
We report 22 cases of candidemia caused by the C. guilliermondii complex and compare them with cases of candidemia caused by C. albicans to identify differential characteristics. We also performed in-depth microbiological characterization of C. guilliermondii isolates based on molecular identification of species, antifungal susceptibility, biofilm formation, and virulence.
(This study was presented in part at the 26th European Congress on Clinical Microbiology and Infectious Diseases in Amsterdam [Abstr P-0350] [13].)
RESULTS
Patient demographics, risk factors, and outcome.Episodes caused by C. albicans in the CANDIPOP study were compared with the overall cohort of C. guilliermondii episodes to identify differential characteristics. To preserve the assumption of independence of observations, only the first episode of candidemia recorded for an individual patient was included in this analysis, and episodes caused simultaneously by different Candida species were excluded. Therefore, 22 C. guilliermondii episodes were finally compared with 337 C. albicans episodes (Table 1). A univariate analysis showed that patients infected by C. guilliermondii were more likely to have underlying malignancy or neutropenia or to have undergone solid organ transplantation. Overall, 59% of these patients had received immunosuppressive therapy within the previous month. Previous exposure to antifungals and long-term central venous catheters (CVCs) were also more common in this group. Conversely, patients with C. albicans infection were older, had more comorbidities (e.g., diabetes), and were more likely to receive total parenteral nutrition.
Comparison of baseline characteristics and clinical data of patients with C. guilliermondii and C. albicans candidemia
Clinical data, therapeutic measures, and outcomes.The clinical presentation of candidemia was less severe in patients infected by C. guilliermondii than in those infected by C. albicans. None of the patients infected by C. guilliermondii had septic shock or severe sepsis at the onset of candidemia. Moreover, the C. guilliermondii infections were less likely to occur during hospitalization in the intensive care unit, and 27.3% of cases were outpatient acquired, with all but one being health care related (Table 1). The therapeutic measures are shown in Table 2. Briefly, early removal of the CVC was less frequent in patients infected by C. guilliermondii than in those infected by C. albicans (15.8% versus 50.2%, respectively; P = 0.004). However, no differences were seen between patients infected by C. guilliermondii and those infected by C. albicans with respect to the frequency of catheter-related candidemia (27.3% versus 30.3%, respectively). Early appropriate antifungal treatment was less common in patients infected by C. guilliermondii than in those infected by C. albicans (36.4% versus 60.5%, respectively; P = 0.026), mainly because therapy was started more than 48 h after diagnosis (8/14 [57.1%]) or because of empirical treatment with fluconazole (5/14 [35.7%]).
Therapeutic measures and outcomes of patients infected by C. guilliermondii complex and C. albicans
As for prognosis, persistent candidemia (blood cultures that were persistently positive for ≥3 days after the incident blood sample) tended to be more common in patients infected by C. guilliermondii than in those infected by C. albicans (38.5% versus 26%, respectively). However, the 30-day mortality was lower in the C. guilliermondii group than in the C. albicans group (13.6% versus 33.9%, respectively; P = 0.049).
Microbiological characterization of the isolates.All cases were caused by C. guilliermondiisensu stricto (n = 17) or C. fermentati (n = 5). In terms of biofilm formation, the isolates were moderate biofilm forming (n = 13) or low biofilm forming (n = 9), with moderate metabolic activity (n = 9), low metabolic activity (n = 12), or high metabolic activity (n = 1). Scanning electron microscopy revealed that the C. guilliermondii complex biofilms were formed mainly by a layer of clamped blastospores without hyphae or pseudohyphae, as seen in cornmeal agar medium (Fig. 1). Table 3 shows the antifungal susceptibility of the isolates in both the planktonic and sessile forms. Up to 10% (2/22) of the isolates had fluconazole MICs above the epidemiological cutoff values (ECOFFs). Differences in susceptibility to all the drugs tested between C. albicans and C. guilliermondii complex isolates were observed in the planktonic form (P < 0.001). According to the susceptibility of the biofilms, no differences reaching statistical significance were observed between anidulafungin and micafungin (sessile MIC50 [SMIC50] geometric mean of 2.46 μg/ml and 8.28 μg/ml, respectively; P = 0.096).
Biofilms formed by C. fermentati (a), C. guilliermondiisensu stricto (b), and C. albicans (c) isolates. ×500 magnification.
Antifungal susceptibilities of C. guilliermondii complex and C. albicans isolates from the CANDIPOP studya
C. guilliermondii complex isolates were significantly less virulent than C. albicans isolates (6 versus 1 mean survival days; P < 0.001) in the Galleria mellonella model (Fig. 2). C. fermentati isolates were more virulent than C. guilliermondiisensu stricto isolates (4 versus 9.5 mean survival days; P < 0.001), although the low number of isolates of this species limited the analysis. We did not find differences in terms of biofilm formation or susceptibility between C. fermentati and C. guilliermondiisensu stricto isolates (data not shown).
Survival curves of G. mellonella infected with the C. guilliermondii complex and C. albicans isolates (of 5 × 106 yeast cells/larva) after 12 days of incubation at 37°C.
DISCUSSION
The C. guilliermondii complex comprises C. guilliermondiisensu stricto, C. fermentati, C. carpophila, and other species (6, 7). In previous studies, C. guilliermondiisensu stricto and C. fermentati accounted for approximately 90% to 95% and 4.5% to 9% of cases, respectively, with the remaining species of the complex playing a marginal role (less than 1%) (9, 14). By contrast, we found a higher frequency of cases caused by C. fermentati (22.7%) and a lower frequency of cases caused by C. guilliermondiisensu stricto (77.2%).
The complex exhibited decreased in vitro susceptibility to all of the antifungal agents studied, with rates of resistance to fluconazole and other azoles of approximately 10% to 15% (9, 12, 14). Up to 10% of the isolates studied here were non-wild type according to the fluconazole ECOFF. In the absence of species-specific breakpoints for C. guilliermondii, the rate of true fluconazole resistance remains unknown. The MICs of micafungin and anidulafungin against C. guilliermondii were 16- and 60-fold higher, respectively, than the MICs against C. albicans reported in the literature (3, 15). We also observed that micafungin and anidulafungin were similarly effective at causing damage to the C. guilliermondii biofilms, although the clinical impact of this observation is unknown.
Candidemia caused by C. guilliermondii commonly originates in the CVC (11, 16). We found that CVCs were involved in a high percentage (86.4%) of patients, as was the case in patients infected by C. albicans. However, C. guilliermondii strains developed biofilms with moderate or low biomass and were significantly different from C. albicans isolates (Fig. 1). This apparent disagreement between the CVC as the source of infection and the scant ability to form biofilms may be due to the fact that many cases of CVC-related candidemia involved long-term CVCs (36.8%), unlike those infected by C. albicans (11.6%). Moreover, the method chosen to study the biomass or metabolic activity implied shorter biofilm maturity (24 h).
C. guilliermondii has been reported to be a cause of infection, mainly in patients with solid tumors and hematological malignancies, patients undergoing chemotherapy, patients with neutropenia, patients with intravascular and vascular catheters, patients receiving total parental nutrition, and patients with previous exposure to antibiotics, corticosteroids, and long-term antifungal therapy (mostly fluconazole, amphotericin B, and caspofungin) (11, 12, 17). We found that patients with candidemia caused by C. guilliermondii were more likely to have solid cancer or hematological malignancies, thus supporting some of the previously reported risk factors associated with this infection.
As for the prognosis, it seems reasonable to expect mortality to be high, considering the severe underlying conditions of the patients and the low susceptibility of the isolates to the antifungal drugs. However, we found that 7-day and 30-day mortalities were lower and clinical presentations were less severe in patients infected by C. guilliermondii than in patients infected by C. albicans. This observation is in agreement with the low virulence of C. guilliermondii isolates compared with that of C. albicans, as shown in the G. mellonella model (P < 0.001). Although this model has been used to study the virulence of several Candida species (18), to our knowledge, this is the first time it has been used for C. guilliermondii. The low virulence of the isolates can also be explained by the limited production of biofilm, as previous studies showed that intense biofilm formation is an independent factor of poor prognosis (19). The G. mellonella model is relatively easy to handle. The correlation between the favorable prognosis of the patients and the high survival of the larvae is encouraging and suggests that this model could be used in the clinical microbiology laboratory to screen for the presence of highly virulent isolates.
In conclusion, the C. guilliermondii complex mainly affected patients with severe debilitating conditions. The isolates were characterized by moderate biofilm production, low susceptibility to azoles and echinocandins, and low virulence. However, the prognoses of the patients infected by this species are favorable.
MATERIALS AND METHODS
Patients and isolates.We described the risk factors and outcome of 22 patients with episodes of monomicrobial candidemia caused by the C. guilliermondii complex who were admitted to Gregorio Marañon Hospital, Madrid (n = 9), or Vall d'Hebron Hospital, Barcelona (n = 5), from 2007 to 2014. We also included cases collected in the remaining hospitals participating in the CANDIPOP project (n = 8), a multicenter population-based study on candidemia conducted from 2010 to 2011 in Spain (3, 20). The incident isolates were identified using amplification and sequencing of the ITS1-5.8S-ITS2 region (3, 21).
Data collection.Cases were identified from the microbiological database of each participating hospital. Data from the CANDIPOP study were collected prospectively, as previously described (3, 20). For cases that were not included in the CANDIPOP study, clinical records were collected from Gregorio Marañon Hospital and Vall d'Hebron Hospital.
Definitions.The definitions used have been reported elsewhere (20, 22). In summary, episodes detected at or within 2 days of hospital admission were considered outpatient acquired. These episodes could also be classified as health care related if patients fulfilled any of the following criteria: hospitalization for ≥2 days within the 90 days prior to candidemia, residence in a long-term-care facility, or chemotherapy or dialysis in the 30 days before diagnosis of candidemia (23). The criteria for proven catheter-related candidemia were those set out in the Infectious Diseases Society of America (IDSA) guidelines (24). Secondary foci required isolation of the same Candida species in the suspected source of infection. Breakthrough episodes were defined as candidemia in patients who had been receiving antifungal drugs for >3 days. The presence of severe sepsis or septic shock was recorded on the day candidemia was diagnosed (25, 26). Adequate early antifungal treatment was defined as the correct dose of antifungal agent for a susceptible Candida isolate administered within the first 48 h after the blood sample collection. Early CVC removal was defined as within the first 48 h after the incident positive blood culture. In patients with multiple CVCs, at least the responsible CVC had to be removed within this time frame.
Antifungal susceptibility testing.The antifungal susceptibility of the planktonic isolates was tested against the following drugs: amphotericin B (Sigma-Aldrich, Madrid, Spain); fluconazole, voriconazole, and anidulafungin (Pfizer Pharmaceutical Group, New York, NY, USA); posaconazole (Merck & Co., Inc., Rahway, NJ, USA); and micafungin (Astellas Pharma, Inc., Tokyo, Japan). Due to the lack of interlaboratory reproducibility, caspofungin was not studied (27).
The procedure used was that of EUCAST EDef 7.2 (28). The antifungal agents were tested at concentrations ranging from 0.015 to 8 μg/ml, except for fluconazole, which was tested at 0.062 to 64 μg/ml. MIC values were determined spectrophotometrically at 530 nm (Multiskan FC microplate photometer; Thermo Scientific, Madrid, Spain) after 24 h of incubation and were defined as the lowest concentration of drug that resulted in inhibition of ≥50% of growth in comparison with a drug-free growth control well for fluconazole, voriconazole, posaconazole, and echinocandins or inhibition of ≥90% for amphotericin B. According to the C. guilliermondii complex ECOFFs, previously reported isolates were considered non-wild type to fluconazole if they had an MIC >16 μg/ml (29).
Biofilm formation and antifungal susceptibility.We studied biofilm formation using crystal violet staining and by measuring the metabolic activity of the biofilm using the XTT reduction assay. We classified the isolates according to their biomass formation and metabolic activity based on our previously reported score (30) as follows: low, moderate, and high biofilm forming (<0.44, 0.44 to 1.17, and >1.17, respectively), and low, moderate, and high metabolic activity (<0.097, 0.097 to 0.2, and >0.2, respectively). The antifungal activity of anidulafungin and micafungin against sessile cells was assessed using the XTT reduction assay at concentrations ranging from 0.015 μg/ml to 512 μg/ml. The sessile MIC50 (SMIC50) was defined as a 50% reduction in the metabolic activity of the biofilm treated with the drug compared with that of the drug-free control well (31).
Scanning electron microscopy.We studied the structures of the biofilms formed by three isolates: C. guilliermondiisensu stricto (n = 1), C. fermentati (n = 1), and C. albicans (n = 1). Biofilms were formed on 50-mm polystyrene discs and prepared for scanning electron microscopy as previously reported (30). The structures of the biofilms were visualized using a scanning electron microscope (JEOL-JSM 6400; Jeol, Tokyo, Japan).
Study of virulence with the Galleria mellonella model.We studied virulence using final-instar larvae of Galleria mellonella (Bichosa, Salceda de Caselas, Spain). Inocula were standardized to enable survival of 50% of the larvae 96 h after infection with 4 strains (5 × 108 yeast cells/ml [5 × 106 yeast cells/larva]). Briefly, inocula were prepared after an overnight incubation of the isolates in yeast extract-peptone-dextrose; the inocula were then washed twice with phosphate-buffered saline (PBS) and adjusted using an automatic cytometer (Moxi flow cytometer; Orflo Technologies, Ketchum, ID, USA). G. mellonella larvae were injected directly into the hemocoel via the last left proleg using a Hamilton syringe and were incubated at 37°C in plastic petri dishes. Dead larvae were counted daily for 12 days. Each experimental group contained 16 randomly chosen larvae of an appropriate weight (330 ± 20 mg). Three control groups containing the same numbers of larvae were included to monitor the insult (PBS was used for injection), the survival of noninjected larvae, and the virulence of 6 randomly chosen C. albicans isolates (5 × 108 yeast cells/ml [5 × 106 yeast cells/larva]).
Data analysis.Quantitative variables are reported as medians and interquartile ranges (IQRs) and qualitative variables are reported as counts (percentages). Categorical variables were compared using the chi-square or Fisher exact tests, as appropriate, and continuous variables were compared with the Mann-Whitney U test. Statistical analyses were performed using Microsoft SPSS-PC+, version 15.0 (SPSS, Chicago, Illinois, USA). P values of < 0.05 were considered significant. The activity of the antifungal drugs is shown as the geometric mean MIC, MIC90, and range of MICs for planktonic and sessile forms; drug activity was compared using the Friedman test. Killing curves were analyzed using the Kaplan-Meier method with GraphPad Prism 5.02 software (GraphPad, La Jolla, CA, USA). Differences in median survival between species were evaluated using the Mann-Whitney U test.
Ethical considerations.This study (protocol no. 39/15) was approved by the ethics committee of Hospital Gregorio Marañón (CEIC-A1). The need for informed consent was waived owing to the retrospective design of the study. The CANDIPOP study was approved by the local institutional review boards, and written consent was obtained from all patients.
ACKNOWLEDGMENTS
We thank Thomas O'Boyle for editing the article.
Members of the CANDIPOP project include Jesús Guinea, Patricia Muñoz, and Belén Padilla (Hospital General Universitario Gregorio Marañón); Manuel Cuenca-Estrella and Óscar Zaragoza (Spanish National Center for Microbiology); José Ramón Paño, Julio García, and Carlos García (Hospital La Paz, Madrid); Jesús Fortún, Pilar Martín, and Elia Gómez (Hospital Ramón y Cajal, Madrid); Pablo Ryan and Carolina Campelo (Hospital Infanta Leonor, Madrid); Ignacio de los Santos and Buenaventura Buendía (Hospital Universitario La Princesa, Madrid); Beatriz Pérez and Mercedes Alonso (Hospital Niño Jesús, Madrid); Francisca Sanz (Hospital 12 de Octubre, Madrid); Paloma Merino and Fernando González (Hospital Clínico, Madrid); Miguel Górgolas and Ignacio Gadea (Fundación Jiménez Díaz, Madrid); Juan Emilio Losa and Alberto Delgado-Iribarren (Hospital de Alcorcón, Madrid); Antonio Ramos, Yolanda Romero, and Isabel Sánchez (Hospital Majadahonda, Madrid); Jesús Rodríguez-Baño and Ana Isabel Suarez (Hospital Universitario Virgen Macarena, Seville); Ana Loza, Estrella Martín-Mazuelos, and Ana Isabel Aller (Hospital Universitario Virgen de Valme, Seville); Maite Ruiz and José Garnacho-Montero (Hospital Universitario Virgen del Rocío, Seville); Carlos Ortiz (Hospital Sagrado Corazón, Seville); Mónica Chávez and Fernando L. Maroto (Hospital San Juan de Dios de Aljarafe, Seville); Miguel Salavert and Javier Pemán (Hospital Universitario la Fe, Valencia); José Blanquer and David Navarro (Hospital Clínico Universitario Valencia); Vicente Abril and Concepción Gimeno (Consorcio Hospital General Universitario de Valencia); Juan José Camarena and Rafael Zaragoza (Hospital Universitario Peset, Valencia); Silvia Hernáez and Guillermo Ezpeleta (Hospital de Basurto, Bilbao); Elena Bereciartua and José L. Hernández (Hospital Universitario de Cruces, Bilbao); Rosa Ana Rivas and Rafael Ayarza (Hospital de Galdakano, Bilbao); Ana Ma Planes (Hospital Universitari Vall d'Hebron, Barcelona); José Mensa and Manel Almela (Hospital Clínic-IDIBAPS Barcelona); Mercè Gurgui and Ferran Sánchez-Reus (Hospital de la Santa Creu i Sant Pau, Barcelona); Joaquín Martinez-Montauti and Montserrat Sierra (Hospital de Barcelona, Barcelona); Juan Pablo Horcajada, Luisa Sorli, and Julià Gómez (Hospital del Mar, Barcelona); Amadeu Gené and Mireia Urrea (Hospital Sant Joan de Déu, Esplugues de Llobregat-Barcelona).
Study collaborators include Mireia Puig-Asensio, Maricela Valerio, Mario Fernández-Ruiz, Ana Díaz-Martín, Francesc Puchades, and Alessandra Mularoni.
This study was partially supported by grants PI14/00740 and MS15/00115 from Fondo de Investigación Sanitaria (FIS; Instituto de Salud Carlos III; Plan Nacional de I+D+I 2013-2016, Fondo Europeo de Desarrollo Regional FEDER Una manera de hacer Europa support) and by Gilead, MSD, Astellas, Pfizer, Fundación SEIMC-GESIDA, and the Spanish Network for Research in Infectious Diseases (REIPI RD12/0015). P.E. (CPI15/00115), J.G. (CPII15/00006), and L.J.M.-Z. (PI14/00740) are supported by the FIS.
J.G. has also received grant support from Gilead Sciences and Scynexis and has been paid for talks on behalf of Gilead Sciences, Merck Sharp and Dohme, Astellas Pharma, and United Medical. M.P.-A. reports personal fees from Pfizer, outside the submitted work. The remaining authors do not report conflicts of interest.
FOOTNOTES
- Received 18 January 2017.
- Returned for modification 11 April 2017.
- Accepted 14 April 2017.
- Accepted manuscript posted online 24 April 2017.
- Copyright © 2017 American Society for Microbiology.
