- Email: [email protected]
Predictors of candidaemia caused by non-albicans Candida species: results of a population-based surveillance in Barcelona, Spain
ORIGINAL ARTICLE
MYCOLOGY
Predictors of candidaemia caused by non-albicans Candida species: results of a population-based surveillance in Barcelona, Spain D. Rodrı´guez1, B. Almirante1, M. Cuenca-Estrella2, J. L. Rodrı´guez-Tudela2, J. Mensa3, J. Ayats4, F. Sanchez5, A. Pahissa1 and the Barcelona Candidemia Project Study Group* 1) Infectious Diseases Division, Hospital Universitari Vall d¢Hebron, Universitat Auto`noma de Barcelona, Barcelona, 2) Mycology Department, Instituto de Salud Carlos III, Madrid, 3) Infectious Diseases Division, Hospital Clinic, IDIBAPS, Barcelona, 4) Microbiology Department, Hospital Universitari de Bellvitge, Hospitalet de Llobregat, Barcelona and 5) Microbiology Department, Hospital de la Santa Creu i Sant Pau, Barcelona, Spain
Abstract Although Candida albicans (CA) is the most common cause of Candida bloodstream infections (BSIs), recent studies have observed an increasing percentage of candidaemias caused by non-albicans Candida species (NAC). In the present study, we attempted to identify the predictors of candidaemia due to NAC compared to CA. We analyzed data from an active population-based surveillance in Barcelona (Spain) from January 2002 to December 2003. Factors associated with NAC fungaemia were determined by multivariate analysis. A total of 339 episodes of Candida BSI, in 336 patients (median age 63 years, interquartile range: 41–72 years), were included. CA was the most commonly isolated (52%), followed by Candida parapsilosis (23%), Candida tropicalis (10%), Candida glabrata (8.6%), Candida krusei (3.4%) and other NAC spp. (3%).Overall, 48% of cases were due to NAC spp. Multivariate logistic regression analysis identified factors associated with a risk of BSI due to NAC spp.: having received a haematologic transplant (OR 10.8; 95% CI 1.31–90.01; p 0.027), previous fluconazole exposure (OR 4.47; 95% CI 2.12–9.43; p <0.001) and neonatal age (OR 4.42; 95% CI 1.63–12.04; p 0.004). Conversely, previous CA colonization (OR 0.33; 95% CI 0.19–0.57; p 0.001) and previous antibiotic use (OR 0.42; 95% CI 0.21–0.85; p 0.017) were associated with CA fungaemia compared to NAC. In conclusion, NAC candidaemia comprised 48% of cases in our series. Predictors of NAC include having received a haematologic transplant, neonatal age and previous fluconazole use. Keywords: Candida albicans, candidaemia, non-albicans candida spp., predictors, prognosis Original Submission: 24 November 2009; Revised Submission: 18 January 2010; Accepted: 25 February 2010 Editor: E. Roilides Article published online: 6 March 2010 Clin Microbiol Infect 2010; 16: 1676–1682 10.1111/j.1469-0691.2010.03208.x Corresponding author: D. Rodrı´guez, Infectious Diseases Division, Hospital Universitari Vall d¢Hebron, Pg. Vall d¢Hebron 119-129, 08035 Barcelona, Spain E-mail: [email protected] *Members of the Barcelona Candidaemia project study group members are listed in the Appendix.
Introduction Candida species accounted for 8–10% of all nosocomial bloodstream infections (BSIs) in the USA during the 1990s and were considered the fourth leading cause of nosocomial BSI [1,2]. Although Candida albicans (CA) continues to be the most common cause of Candida BSI, longitudinal studies have detected an increase of other Candida species in this condition [3–6]. Non-albicans Candida species (NAC) currently account for approximately half of all cases of candida-
emia, although there is little published information on the factors associated with the increase in NAC candidaemia [4,5,7–10]. The changing epidemiology of Candida BSI has generated concern about the emergence of azole drug resistance and its clinical relevance [5,9,11]. Although CA is generally susceptible to fluconazole, Candida glabrata, the second cause of candidaemia in the USA, shows decreased susceptibility to this agent in up to 65% of cases [10–12]. The emergence of azole-resistant strains and the discovery of new antifungal drugs (new triazoles and echinocandins) have raised important questions about the use of fluconazole as a first-line drug. Therefore, it is of great value to identify the risk factors that distinguish CA infections from those due to NAC to aid clinicians in choosing an empiric anti-Candida therapy [13]. Various studies in immunocompromised and intensive care unit (ICU) patients with candidaemia have examined the clinical features that differentiate CA infection from NAC infec-
ª2010 The Authors Journal Compilation ª2010 European Society of Clinical Microbiology and Infectious Diseases
CMI
Rodrı´guez et al. Predictors of non-albicans candidaemia
tion; however, the predictors of NAC BSI have not been prospectively studied in a general patient population [3,14– 16]. The present study aimed to identify the risk factors associated with fungaemia due to NAC species compared to CA in a general patient population in the hope that this information can be used to guide the initial empiric therapy.
Materials and Methods Study and study population
We conducted a prospective, population-based active surveillance for Candida BSI in the Barcelona (Spain) area (population, 3.9 million), between 1 January 2002 and 31 December 2003. Fourteen major institutions participated, ranging in size from 214 to 1200 beds. A case was defined as the incident isolation of any Candida species from the blood of a surveillance area resident. Candidaemia episodes that occurred >30 days after the initial case were considered a new episode. Patients with candidaemia caused simultaneously by different species of Candida were excluded from the analysis because they had both CA and NAC species. Data collection
The case report was laboratory-based. All blood cultures in which a Candida species had been isolated were reported to the study coordinator (DR). The coordinator visited the hospital to confirm the infection, completed the case report form, and recorded the outcome. The clinical laboratories were periodically audited to ensure that all cases of candidaemia were reported. A standardized case report form was used to abstract the medical records. The data collected included patient demographic characteristics; comorbid conditions; concomitant infections; previous Candida colonization; exposure to antibiotic, antifungal and immunosuppressive therapy; total parenteral nutrition received; surgical procedures; central venous catheter (CVC) use; haemodialysis; mechanical ventilation; treatment of candidaemia and outcome. To measure disease severity, we used the Acute Physiology and Chronic Health Evaluation II score [17] for adult patients admitted to ICUs and the Karnofsky performance status scale for adults outside the ICU [18]. No standardized paediatric severity of illness score was used. Risk factors were assessed within 30 days prior to the diagnosis by positive sample. Microbiological methods
Candida detection and species identification were performed at the participating laboratories according to their standard
1677
protocols. Isolates were sent to the Mycology Reference Laboratory (MRL), National Center for Microbiology, Madrid, Spain, for species confirmation and antifungal susceptibility testing, using standard morphological and physiological methods, including fermentation of and growth on carbon sources, growth on nitrogen sources and growth at various temperatures. Candida parapsilosis ATCC 22019 and Candida krusei ATCC 6258 were used as quality control organisms for antifungal drug susceptibility testing [12]. When MRL and submitting laboratory identifications differed, the MRL identification was used for the study. MICs of fluconazole and voriconazole were determined by the broth microdilution method as recommended by the European Committee on Antimicrobial Susceptibility Testing (EUCAST). Interpretive breakpoints proposed by EUCAST for fluconazole and voriconazole resistance are >4 and >0.12 mg/L, respectively [19]. Statistical analysis
Population incidences of CA and NAC were calculated using denominator data obtained from the 2001 local census. Statistical analyses were performed using SPSS software, version 15.0 (SPSS Inc., Chicago, IL, USA). The results for categorical variables are expressed as a percentage, and numerical data are expressed as the mean (SD), median, and interquartile range for ages. The chi-square test or Fisher’s exact test (two-tailed) were used to compare categorical variables, and an unpaired Student’s t-test was used for continuous variables. Multivariate stepwise logistic regression analysis was carried out to identify predictors of candidaemia due to NAC. p <0.05 was considered statistically significant in the multivariate modeling.
Results During the study period, 345 cases of Candida BSI were detected. Six episodes of polyfungal candidaemia were excluded, leaving 339 episodes in 336 patients (median age 63 years, interquartile range: 41–72 years). A total of 176 (52%) patients had fungaemia due to C. albicans and 163 (48%) had infection due to NAC. The average annual incidence of candidaemia was 4.3 cases per 100 000 population (2.2 cases per 100 000 population for CA and 2.1 cases per 100 000 population for NAC). Demographics, clinical characteristics, underlying comorbid conditions and the outcome of patients with NAC BSI compared to those with CA BSI are summarized in Table 1. The NAC species isolated included 78 C. parapsilosis (23%), 34 Candida tropicalis (10%), 29 Candida glabrata (8.6%), 12 C. krusei (3.6%), three Candida kefyr (0.9%), two Candida
ª2010 The Authors Journal Compilation ª2010 European Society of Clinical Microbiology and Infectious Diseases, CMI, 16, 1676–1682
1678
Clinical Microbiology and Infection, Volume 16 Number 11, November 2010
CMI
TABLE 1. Demographics, epidemiological data and clinical
TABLE 2. Species distribution and in vitro susceptibilities to
characteristics of the patients included in the present study
fluconazole and voriconazole (European Committee on
Characteristics (n = 339 episodes in 336 patients)
C. albicans, Non-albicans n = 176 Candida spp., (52%) n = 163 (48%) p
Age (years), mean (SD) 57.9 (24.07) Elderly (>65 years) 95 (54) Male sex 101 (57.4) Nosocomial candidaemia 158 (89.8) Prior hospitalizationa 86 (48.9) APACHE II (severity high)b 29 (26.9) Length of stay before 30.1 (35.3) candidaemia (days), mean (SD) Location of the patients Medical ward 49 (27.8) Surgical ward 44 (25) ICU hospitalization (except neonates) 52 (29.5) c Neonatal ICU hospitalization 7 (4) Pediatric ward (excluding neonates) 7 (4) Outpatient 17 (9.7) d Predisposing factors Diabetes mellitus 44 (25) Chronic renal failure 73 (41.5) Solid tumour 45 (25.6) Haematological malignancy 16 (9.1) Neutropaenia 11 (6.3) Solid organ transplant recipient 2 (1.1) Haematological transplant recipient 1 (0.6) Surgery in previous 3 months 85 (48.3) e Prior immunosuppressive drugs 66 (37.5) Prior antibiotics (1 previous month) 158 (89.8) Prior fluconazole therapy 12 (6.8) CVC placement 130 (73.9) Total parenteral nutrition 57 (32.4) Previous Candida colonizationf 74 (42.0) Candiduria due to Candida albicans 36 (20.5) Portal of entry Primary candidaemia 105 (59.7) CVC-related candidaemia 54 (30.7) Secondary candidaemia 17 (9.7) g Optimal antifungal treatment 82 (52.2) Outcome Shock 46 (26.1) Mechanical ventilation 55 (31.3) ICU admission during outcome 15 (8.5) Acute renal failure 37 (21.0) Early mortality (<7 days) 47 (26.7) Overall mortality (<30 days) 75 (42.6)
47.7 66 98 149 77 20 29.7
(27.45) (40.5) (60.1) (91.4) (47.2) (19) (33.6)
0.011 0.013 0.659 0.573 0.770 0.176 0.752
64 22 37 17 9 14
(39.3) (13.5) (22.7) (10.4) (5.5) (8.6)
0.026 0.008 0.152 0.032 0.611 0.851
26 44 27 32 28 8 15 69 66 137 41 138 67 34 12
(16) (27.0) (16.6) (19.6) (17.2) (4.9) (9.2) (4236) (40.5) (84.0) (25.2) (84.7) (41.1) (20.9) (7.4)
0.037 0.006 0.043 0.005 0.002 0.054 <0.001 0.271 0.479 0.146 <0.001 0.015 0.096 <0.001 0.001
95 53 15 86
(58.3) (32.5) (9.2) (52.8)
0.797 0.717 1 0.36
43 46 10 26 35 60
(26.4) (28.2) (6.1) (16.0) (21.5) (36.8)
0.967 0.592 0.533 0.242 0.261 0.275
Categorical variables are expressed as total number and percentages, and numerical data are expressed as the mean and standard deviation (SD). Patients hospitalized within 3 months prior to candidaemia. APACHE, Acute Physiology and Chronic Health Evaluation score system. b Includes patients with APACHE >20 (only patients admitted to ICU). ICU, intensive care unit. c NAC in the neonatal ICU were due to Candida parapsilosis in 16 cases and Candida glabrata in one case. d Some patients had more than one predisposing factor. e Includes corticoids, chemotherapy, and other immunosuppressive drugs. CVC, central venous catheter. f Previous Candida colonization due to the same Candida spp. that caused candidaemia (includes candiduria). g Defined as at least 5 days of adequate antifungal treatment. a
Antimicrobial Susceptibility Testing breakpoint for Candida susceptibilitya)
Species
Fluconazole resistant isolates/n total isolates (%)
Voriconazole resistant isolates/n total isolates (%)
Candida Candida Candida Candida Candida Candida Candida Candida Candida Candida Candida Total
2/176 2/78 1/34 13/29 12/12 0/3 0/2 0/2 1/1 1/1 1/1 33/339
1/176 2/78 1/34 19/29 12/12 0/3 0/2 0/2 1/1 0/1 1/1 37/339
albicans parapsilosis tropicalis glabrata krusei kefyr lusitaniae guillierrmondii famata inconspicua norvegensis
(1.1) (2.6) (2.9) (44.8) (100) (0) (0) (0) (100) (100) (100) (9.7)
(0.6) (2.6) (2.9) (65.5) (100) (0) (0) (0) (100) (0) (100) (11)
a
Fluconazole-resistant isolates: MIC >4 mg/L. Voriconazole-resistant isolates: MIC >0.12 mg/L.
renal failure (41.5% CA vs. 27.2% NAC; p 0.006) and had been previously colonized with C. albicans (42.3% CA vs. 20.9% NAC; p <0.001). Cases with C. parapsilosis were more likely to be infants (28% vs. 5%; p <0.001) and have an indwelling CVC (97% vs. 87%; p 0.006), and less likely to die within 30 days (23% vs. 45%; p 0.001) compared to the other cases. C. krusei cases had received previous fluconazole (75% vs. 14%; p <0.001) and immunosuppressive therapy (75% vs. 38%; p 0.01) more often, were more frequently neutropaenic (67% vs. 9%; p <0.001) and had haematological malignancies (58% vs. 12%; p <0.001) more often than non-krusei cases. C. tropicalis cases were also likely to be neutropaenic (24% vs. 10%; p 0.04), have haematological malignancies (29% vs. 11%; p 0.007), and die within 30 days (56% vs. 38%), although this latter difference did not reach statistical significance (p 0.06). Overall mortality tended to be lower in patients with NAC BSI compared to CA BSI, although this trend did not reach statistical significance (36.8% NAC vs. 42.6% CA; p 0.27). Factors associated with NAC fungaemia were determined by univariate analysis (Table 1) and multivariate analysis TABLE 3. Multivariate analysis of risk factors associated with non-albicans Candida infection among all patients with
lusitaniae (0.6%), two Candida guilliermondii (0.6%), one Candida famata (0.3%), one Candida inconspicua (0.3%) and one Candida norvegensis (0.3%). Percentages of in vitro fluconazole and voriconazole resistance among the 339 incident bloodstream isolates of Candida spp. are shown in Table 2. Subgroup analysis for specific Candida spp. showed that patients with C. albicans fungaemia were admitted more often to surgical wards (25% CA vs. 13.5% NAC; p 0.008), had diabetes mellitus (25% CA vs. 16% NAC; p 0.03) or chronic
candidaemia Variable
OR
95% CI
p
Haematological transplant recipient Previous fluconazole usea Neonatal intensive care unit admission Previous antibiotic useb Previous Candida albicans colonization
10.8 4.47 4.42 0.42 0.33
1.31–90.01 2.12–9.43 1.63–12.04 0.21–0.85 0.19–0.57
0.027 <0.001 0.004 0.017 <0.001
a
At least 3 days of fluconazole treatment within 30 days prior to candidaemia. At least 3 days of any antibiotic treatment within 30 days prior to candidaemia.
b
ª2010 The Authors Journal Compilation ª2010 European Society of Clinical Microbiology and Infectious Diseases, CMI, 16, 1676–1682
CMI
Rodrı´guez et al. Predictors of non-albicans candidaemia
(Table 3). In the multivariate analysis, the fact of having undergone haematological transplantation (OR 10.8; 95% CI 1.31–90.1; p 0.027) or previous exposure to fluconazole (OR 4.47; 95% CI 2.12–9.43; p <0.001), and neonatal age (OR 4.42; 95% CI 1.63–12.04; p 0.004) were significantly associated with NAC fungaemia, whereas previous CA colonization (OR 0.33; 95% CI 0.19–0.57; p <0.001) and previous antibiotic use (OR 0.42; 95% CI 0.21–0.85; p 0.017) were associated with CA candidaemia.
Discussion In the general population of Barcelona, 48% of all cases of candidaemia were caused by NAC. The distribution of Candida species was generally similar to those reported from other European countries, such as France and Italy [20,21], with C. parapsilosis being the second most common Candida spp. isolated. An exception is the incidence of C. glabrata, which was the fourth most common species in the present study, but the second most common in England and Wales [22], Switzerland [23], Denmark [24] and the USA [10,25,26]. An increase in the percentage of C. glabrata BSI was first reported in the USA during the 1990s [6,25,26]. However, this trend is not universal, and C. glabrata is less common in other areas, including South America, Canada and some European countries (e.g. Italy and France) which, again, is in keeping with our results [20,21,27]. The changing pattern of candidaemia may be attributable in part to the large number of immunocompromised hosts and the widespread use of prophylactic or empiric antifungal therapy [9], although the role of widespread azole use in the emergence of species other than CA as causes of candidaemia remains controversial. Knowledge of the local epidemiological trends in Candida species isolated in blood cultures is important to guide the choice of empiric therapy. Previous studies of risk factors for candidaemia due to NAC compared to CA have focused on critically ill ICU patients or patients with a haematological malignancy [14– 16,28] and have investigated a single species, such as C. parapsilosis or C. glabrata [8,14,29,30]. To our knowledge, the risk factors for NAC vs. CA have not been prospectively described in a general patient population. The results obtained in the present study and those of Chow et al. [15] contrast with the results reported by Shorr et al. [16] who found no differences in the clinical characteristics of ICU patients with BSI caused by NAC or CA. These authors theorized that clinical variables do not allow one to successfully predict the microbiology of fungaemia in a particular patient because NAC species are essentially endemic
1679
[16]. Nonetheless, we found several differences in patients with NAC fungaemia compared to those with CA: they were significantly younger, had a higher incidence of haematological malignancies, were neutropaenic or had undergone haematological transplantation, or had been exposed to fluconazole previously. NAC was not associated with an increase in mortality or length of hospital stay compared to CA, probably because C. parapsilosis accounted for a considerable percentage of NAC isolates in our series, and this species is known to be less virulent, more frequently associated with a CVC infectious origin, and uniformly susceptible to fluconazole [8,30]. Subsequent to multivariate analysis, NAC fungaemia was found to be independently associated with having undergone haematological transplantation (p 0.027). The potential risk factors for developing NAC BSI identified in previous studies have included neutropaenia and an underlying haematological disease and, in these cases, C. tropicalis, C. krusei and even C. glabrata were more prevalent [7,14,28]. Because it is a general practice to use fluconazole prophylaxis in high-risk haematological patients, it is reasonable to conclude that undergoing haematological transplantation could be a confounding risk factor, with previous fluconazole exposure being the determining predictor in these cases. Although fluconazole prophylaxis is associated with a reduction in the incidence of candidaemia and attributable mortality, several studies have investigated the way in which fluconazole can influence the development of azole resistance [9,31]. In our series, the predominant Candida species varied according to the hospital care unit. C. parapsilosis was the most prevalent species in the neonatal ICU (16 of 24 [67%] cases), whereas CA was the most prevalent in surgical areas. In our series, all except one case of NAC among neonates were caused by C. parapsilosis. Considering that multivariate analysis identified neonatal age as a risk factor for NAC, we were able to determine that neonatal age is a predictor for C. parapsilosis fungaemia. The association between neonatal fungaemia and C. parapsilosis has been extensively described, and some studies suggest that it may reflect the intensive use of intravascular devices to treat neonates and children [32,33]. We also found that fluconazole exposure is a risk factor for the development of NAC BSI. The results obtained in the present study agree with the findings reported in previous studies postulating that widespread fluconazole use would result in selection of yeast species that are less sensitive to this antifungal agent, such as C. krusei, C. glabrata and C. tropicalis [9,15,31]. Fluconazole use and selection of fluconazole-resistant Candida spp. has been extensively described in studies performed in ICU patients and in patients with
ª2010 The Authors Journal Compilation ª2010 European Society of Clinical Microbiology and Infectious Diseases, CMI, 16, 1676–1682
1680
Clinical Microbiology and Infection, Volume 16 Number 11, November 2010
cancer or haematological malignancies [3,30,31]. In patients without cancer, the association between previous fluconazole exposure and NAC fungaemia is not as clear, with some studies suggesting an independent relationship and others failing to note this association [15,16,34]. Moreover, the important percentage of C. parapsilosis, a yeast species that is almost always susceptible to fluconazole, is not explained by the increase of fluconazole use. It is likely that fungemia due to C. parapsilosis reflects acquisition associated with the widespread use of parenteral feeding and intravascular devices [30–32]. We also found that patients were more likely to develop CA fungemia if they had been treated previously with antibiotics at the onset of candidaemia or if they had been colonized previously by CA. Previous antibiotic exposure has been associated with CA in non-neutropaenic patients [34] and previous Candida colonization by the same Candida spp. has also been described as a risk factor for candidaemia [35]. The observations made in the present study are subject to limitations. First, the data were obtained 6 years ago. Several new treatment options subsequently have been developed (e.g. echinocandins and new triazoles) but, in all probability, the epidemiology and species distribution of Candida has not changed substantially, and these data are still relevant. Second, only the presence or absence of risk factor exposure was considered, and not the duration of exposure. Because the study was not designed to quantify the length of exposure, this variable is not available and any associated bias cannot be assessed. Third, it may not have been appropriate to pool all NAC species together because certain NAC species are uniformly susceptible to fluconazole. To address this limitation, other studies [16] have analyzed C. krusei and C. glabrata together vs. C. albicans, C. tropicalis and C. parapsilosis. However, the main goal of the present study was not to investigate potential fluconazole resistance but, instead, to identify predictors of candidaemia due to NAC. Nonetheless, the small number of fluconazole-resistant isolates precluded a risk factor analysis for individual Candida species. On the other hand, some strengths of the present study include the fact that risk factors for NAC fungaemia were examined in a general population, the sample size is relatively large, and the study design is prospective. In summary, a number of factors were found to be independently associated with an increased risk of candidaemia due to NAC compared to CA. Neonatal age, having undergone haematological transplantation, and prior fluconazole use were associated with NAC BSI. Physicians can take these specific characteristics into consideration when empiric treatment with an antifungal agent must be started while waiting for Candida species to be identified.
CMI
Appendix Other members of the Barcelona Candidaemia Project Study Group
S. Fridkin, R. Hajjeh, B. Park, J. Morgan, D. W. Warnock (Centers for Disease Control and Prevention, Atlanta, GA); A. M. Planes (Hospital de Vall d’Hebron, Barcelona, Spain); M. Almela, F. M. Reverter, C. M. Soler (Hospital Clinic, IDIBAPS, Barcelona, Spain); M. Salvado´, P. Saballs (Hospital del Mar, Barcelona, Spain); A. Gener (Hospital Sant Joan de Deu, Esplugues de Llobregat, Barcelona, Spain); D. Fontanals (Hospital Parc Taulıı´, Sabadell, Barcelona, Spain); M. Xercavins (Hospital Mutua de Terrassa, Terrassa, Barcelona, Spain); L. Falgueras, M. T. Torroella, M. de Ramon (Hospital General de Catalunya, Sant Cugat del Valles, Barcelona, Spain); Carles Alonso (Hospital Creu Roja, Hospitalet de Llobregat, Barcelona, Spain); J. de Otero (Hospital Creu Roja, Barcelona, Spain); M. Sierra and J. Martinez-Montauti (Hospital de Barcelona, Barcelona, Spain); M. A. Morera (Hospital de Terrassa, Terrassa, Barcelona, Spain); and M. Gime´nez (Hospital Universitari Germans Trias i Pujol, Badalona, Barcelona, Spain).
Author Contributions D. Rodriguez, B. Almirante and A. Pahissa made a substantial contribution to the conception and design of the manuscript; acquisition, analysis and interpretation of data; drafting the article and revising it critically for important intellectual content; and giving final approval of the version to be published. M. Cuenca-Estrella and J. L. Rodriguez-Tudela performed all microbiological studies in the Mycology Reference Laboratory and made a substantial contribution to the acquisition and interpretation of data; revising the manuscript critically for important intellectual content; and giving final approval of the version to be published. J. Mensa, J. Ayats and F. Sanchez made a substantial contribution to the acquisition and interpretation of data; revising the manuscript critically for important intellectual content; and giving final approval of the version to be published.
Acknowledgements This study was presented, in part, as a poster (M-2141a) at the 48th Annual Meeting of the Interscience Conference on Antimicrobial Agents and Chemotherapy/Infectious Diseases Society of America, Washington, DC, 25–28 October 2008. We thank C. Cavallo for assistance with the English language.
ª2010 The Authors Journal Compilation ª2010 European Society of Clinical Microbiology and Infectious Diseases, CMI, 16, 1676–1682
CMI
Rodrı´guez et al. Predictors of non-albicans candidaemia
Transparency Declaration 13.
This study was supported by research grants from Pfizer, Inc., Gilead Sciences S.A., and the Sociedad Espan˜ola de Enfermedades Infecciosas y Microbiologia Clinica (SEIMC, Spanish Society of Infectious Diseases and Clinical Microbiology) and by a medical research grant from the Ministerio de Sanidad y Consumo, Instituto de Salud Carlos III, Spanish Network for the Research in Infectious Diseases (REIPI RD 06/ 0008). None of the authors has any conflicts of interest.
14.
15.
16.
17.
References 1. Edmond MB, Wallace SE, McClish DK, Pfaller MA, Jones RN, Wenzel RP. Nosocomial bloodstream infections in United States hospitals: a three-year analysis. Clin Infect Dis 1999; 29: 239–244. 2. Pfaller MA, Jones RN, Messer SA, Edmond MB, Wenzel RP. National surveillance of nosocomial blood stream infection due to Candida albicans: frequency of occurrence and antifungal susceptibility in the SCOPE Program. Diagn Microbiol Infect Dis 1998; 31: 327– 332. 3. Bassetti M, Righi E, Costa A et al. Epidemiological trends in nosocomial candidemia in intensive care. BMC Infect Dis 2006; 6: 21. 4. Krcmery V, Barnes AJ. Non-albicans Candida spp. causing fungaemia: pathogenicity and antifungal resistance. J Hosp Infect 2002; 50: 243–260. 5. Pfaller MA, Diekema DJ, Jones RN et al. International surveillance of bloodstream infections due to Candida species: frequency of occurrence and in vitro susceptibilities to fluconazole, ravuconazole, and voriconazole of isolates collected from 1997 through 1999 in the SENTRY antimicrobial surveillance program. J Clin Microbiol 2001; 39: 3254–3259. 6. Trick WE, Fridkin SK, Edwards JR, Hajjeh RA, Gaynes RP. Secular trend of hospital-acquired candidemia among intensive care unit patients in the United States during 1989-1999. Clin Infect Dis 2002; 35: 627–630. 7. Almirante B, Rodriguez D, Park BJ et al. Epidemiology and predictors of mortality in cases of Candida bloodstream infection: results from population-based surveillance, Barcelona, Spain, from 2002 to 2003. J Clin Microbiol 2005; 43: 1829–1835. 8. Almirante B, Rodriguez D, Cuenca-Estrella M et al. Epidemiology, risk factors, and prognosis of Candida parapsilosis bloodstream infections: case-control population-based surveillance study of patients in Barcelona, Spain, from 2002 to 2003. J Clin Microbiol 2006; 44: 1681– 1685. 9. Blot S, Janssens R, Claeys G et al. Effect of fluconazole consumption on long-term trends in candidal ecology. J Antimicrob Chemother 2006; 58: 474–477. 10. Horn DL, Neofytos D, Anaissie EJ et al. Epidemiology and outcomes of candidemia in 2019 patients: data from the prospective antifungal therapy alliance registry. Clin Infect Dis 2009; 48: 1695–1703. 11. Pfaller MA, Messer SA, Boyken L. Geographic variation in the susceptibilities of invasive isolates of Candida glabrata to seven systemically active antifungal agents: a global assessment from the ARTEMIS Antifungal Surveillance Program conducted in 2001 and 2002. J Clin Microbiol 2004; 42: 3142–3146. 12. Cuenca-Estrella M, Rodriguez D, Almirante B et al. In vitro susceptibilities of bloodstream isolates of Candida species to six antifungal agents: results from a population-based active surveillance
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
1681
programme, Barcelona, Spain, 2002-2003. J Antimicrob Chemother 2005; 55: 194–199. Charlier C, Hart E, Lefort A et al. Fluconazole for the management of invasive candidiasis: where do we stand after 15 years? J Antimicrob Chemother 2006; 57: 384–410. Bodey GP, Mardani M, Hanna HA et al. The epidemiology of Candida glabrata and Candida albicans fungemia in immunocompromised patients with cancer. Am J Med 2002; 112: 380–385. Chow JK, Golan Y, Ruthazer R et al. Factors associated with candidemia caused by non-albicans Candida species versus Candida albicans in the intensive care unit. Clin Infect Dis 2008; 46: 1206–1213. Shorr AF, Lazarus DR, Sherner JH et al. Do clinical features allow for accurate prediction of fungal pathogenesis in bloodstream infections? Potential implications of the increasing prevalence of non-albicans candidemia. Crit Care Med 2007; 35: 1077–1083. Chang RW, Jacobs S, Lee B. Predicting outcome among intensive care unit patients using computerised trend analysis of daily Apache II scores corrected for organ system failure. Intensive Care Med 1988; 14: 558–566. Motzer RJ, Mazumdar M, Bacik J, Berg W, Amsterdam A, Ferrara J. Survival and prognostic stratification of 670 patients with advanced renal cell carcinoma. J Clin Oncol 1999; 17: 2530–2540. Rodriguez-Tudela JL, Arendrup MC, Barchiesi F et al. EUCAST definitive document EDef 7.1: method for the determination of broth dilution MICs of antifungal agents for fermentative yeasts. Clin Microbiol Infect 2008; 14: 398–405. Richet H, Roux P, Des CC, Esnault Y, Andremont A. Candidemia in French hospitals: incidence rates and characteristics. Clin Microbiol Infect 2002; 8: 405–412. Tortorano AM, Biraghi E, Astolfi A et al. European Confederation of Medical Mycology (ECMM) prospective survey of candidaemia: report from one Italian region. J Hosp Infect 2002; 51: 297–304. Kibbler CC, Seaton S, Barnes RA et al. Management and outcome of bloodstream infections due to Candida species in England and Wales. J Hosp Infect 2003; 54: 18–24. Marchetti O, Bille J, Fluckiger U et al. Epidemiology of candidemia in Swiss tertiary care hospitals: secular trends, 1991-2000. Clin Infect Dis 2004; 38: 311–320. Arendrup MC, Fuursted K, Gahrn-Hansen B et al. Semi-national surveillance of fungaemia in Denmark 2004–2006: increasing incidence of fungaemia and numbers of isolates with reduced azole susceptibility. Clin Microbiol Infect 2008; 14: 487–494. Diekema DJ, Messer SA, Brueggemann AB et al. Epidemiology of candidemia: 3-year results from the emerging infections and the epidemiology of Iowa organisms study. J Clin Microbiol 2002; 40: 1298–1302. Hajjeh RA, Sofair AN, Harrison LH et al. Incidence of bloodstream infections due to Candida species and in vitro susceptibilities of isolates collected from 1998 to 2000 in a population-based active surveillance program. J Clin Microbiol 2004; 42: 1519–1527. Pfaller MA, Jones RN, Doern GV, Sader HS, Hollis RJ, Messer SA. International surveillance of bloodstream infections due to Candida species: frequency of occurrence and antifungal susceptibilities of isolates collected in 1997 in the United States, Canada, and South America for the SENTRY Program. The SENTRY Participant Group. J Clin Microbiol 1998; 36: 1886–1889. Viscoli C, Girmenia C, Marinus A et al. Candidemia in cancer patients: a prospective, multicenter surveillance study by the Invasive Fungal Infection Group (IFIG) of the European Organization for Research and Treatment of Cancer (EORTC). Clin Infect Dis 1999; 28: 1071–1079. Fairchild KD, Tomkoria S, Sharp EC, Mena FV. Neonatal Candida glabrata sepsis: clinical and laboratory features compared with other Candida species. Pediatr Infect Dis J 2002; 21: 39–43.
ª2010 The Authors Journal Compilation ª2010 European Society of Clinical Microbiology and Infectious Diseases, CMI, 16, 1676–1682
1682
Clinical Microbiology and Infection, Volume 16 Number 11, November 2010
30. Weems JJ Jr. Candida parapsilosis: epidemiology, pathogenicity, clinical manifestations, and antimicrobial susceptibility. Clin Infect Dis 1992; 14: 756–766. 31. Marr KA, Seidel K, White TC, Bowden RA. Candidemia in allogeneic blood and marrow transplant recipients: evolution of risk factors after the adoption of prophylactic fluconazole. J Infect Dis 2000; 181: 309–316. 32. Levin AS, Costa SF, Mussi NS et al. Candida parapsilosis fungemia associated with implantable and semi-implantable central venous catheters
CMI
and the hands of healthcare workers. Diagn Microbiol Infect Dis 1998; 30: 243–249. 33. Rodriguez D, Almirante B, Park BJ et al. Candidemia in neonatal intensive care units: Barcelona, Spain. Pediatr Infect Dis J 2006; 25: 224–229. 34. Davis SL, Vazquez JA, McKinnon PS. Epidemiology, risk factors, and outcomes of Candida albicans versus non-albicans candidemia in nonneutropenic patients. Ann Pharmacother 2007; 41: 568–573. 35. Ostrosky-Zeichner L. New approaches to the risk of Candida in the intensive care unit. Curr Opin Infect Dis 2003; 16: 533–537.
ª2010 The Authors Journal Compilation ª2010 European Society of Clinical Microbiology and Infectious Diseases, CMI, 16, 1676–1682
MYCOLOGY
Predictors of candidaemia caused by non-albicans Candida species: results of a population-based surveillance in Barcelona, Spain D. Rodrı´guez1, B. Almirante1, M. Cuenca-Estrella2, J. L. Rodrı´guez-Tudela2, J. Mensa3, J. Ayats4, F. Sanchez5, A. Pahissa1 and the Barcelona Candidemia Project Study Group* 1) Infectious Diseases Division, Hospital Universitari Vall d¢Hebron, Universitat Auto`noma de Barcelona, Barcelona, 2) Mycology Department, Instituto de Salud Carlos III, Madrid, 3) Infectious Diseases Division, Hospital Clinic, IDIBAPS, Barcelona, 4) Microbiology Department, Hospital Universitari de Bellvitge, Hospitalet de Llobregat, Barcelona and 5) Microbiology Department, Hospital de la Santa Creu i Sant Pau, Barcelona, Spain
Abstract Although Candida albicans (CA) is the most common cause of Candida bloodstream infections (BSIs), recent studies have observed an increasing percentage of candidaemias caused by non-albicans Candida species (NAC). In the present study, we attempted to identify the predictors of candidaemia due to NAC compared to CA. We analyzed data from an active population-based surveillance in Barcelona (Spain) from January 2002 to December 2003. Factors associated with NAC fungaemia were determined by multivariate analysis. A total of 339 episodes of Candida BSI, in 336 patients (median age 63 years, interquartile range: 41–72 years), were included. CA was the most commonly isolated (52%), followed by Candida parapsilosis (23%), Candida tropicalis (10%), Candida glabrata (8.6%), Candida krusei (3.4%) and other NAC spp. (3%).Overall, 48% of cases were due to NAC spp. Multivariate logistic regression analysis identified factors associated with a risk of BSI due to NAC spp.: having received a haematologic transplant (OR 10.8; 95% CI 1.31–90.01; p 0.027), previous fluconazole exposure (OR 4.47; 95% CI 2.12–9.43; p <0.001) and neonatal age (OR 4.42; 95% CI 1.63–12.04; p 0.004). Conversely, previous CA colonization (OR 0.33; 95% CI 0.19–0.57; p 0.001) and previous antibiotic use (OR 0.42; 95% CI 0.21–0.85; p 0.017) were associated with CA fungaemia compared to NAC. In conclusion, NAC candidaemia comprised 48% of cases in our series. Predictors of NAC include having received a haematologic transplant, neonatal age and previous fluconazole use. Keywords: Candida albicans, candidaemia, non-albicans candida spp., predictors, prognosis Original Submission: 24 November 2009; Revised Submission: 18 January 2010; Accepted: 25 February 2010 Editor: E. Roilides Article published online: 6 March 2010 Clin Microbiol Infect 2010; 16: 1676–1682 10.1111/j.1469-0691.2010.03208.x Corresponding author: D. Rodrı´guez, Infectious Diseases Division, Hospital Universitari Vall d¢Hebron, Pg. Vall d¢Hebron 119-129, 08035 Barcelona, Spain E-mail: [email protected] *Members of the Barcelona Candidaemia project study group members are listed in the Appendix.
Introduction Candida species accounted for 8–10% of all nosocomial bloodstream infections (BSIs) in the USA during the 1990s and were considered the fourth leading cause of nosocomial BSI [1,2]. Although Candida albicans (CA) continues to be the most common cause of Candida BSI, longitudinal studies have detected an increase of other Candida species in this condition [3–6]. Non-albicans Candida species (NAC) currently account for approximately half of all cases of candida-
emia, although there is little published information on the factors associated with the increase in NAC candidaemia [4,5,7–10]. The changing epidemiology of Candida BSI has generated concern about the emergence of azole drug resistance and its clinical relevance [5,9,11]. Although CA is generally susceptible to fluconazole, Candida glabrata, the second cause of candidaemia in the USA, shows decreased susceptibility to this agent in up to 65% of cases [10–12]. The emergence of azole-resistant strains and the discovery of new antifungal drugs (new triazoles and echinocandins) have raised important questions about the use of fluconazole as a first-line drug. Therefore, it is of great value to identify the risk factors that distinguish CA infections from those due to NAC to aid clinicians in choosing an empiric anti-Candida therapy [13]. Various studies in immunocompromised and intensive care unit (ICU) patients with candidaemia have examined the clinical features that differentiate CA infection from NAC infec-
ª2010 The Authors Journal Compilation ª2010 European Society of Clinical Microbiology and Infectious Diseases
CMI
Rodrı´guez et al. Predictors of non-albicans candidaemia
tion; however, the predictors of NAC BSI have not been prospectively studied in a general patient population [3,14– 16]. The present study aimed to identify the risk factors associated with fungaemia due to NAC species compared to CA in a general patient population in the hope that this information can be used to guide the initial empiric therapy.
Materials and Methods Study and study population
We conducted a prospective, population-based active surveillance for Candida BSI in the Barcelona (Spain) area (population, 3.9 million), between 1 January 2002 and 31 December 2003. Fourteen major institutions participated, ranging in size from 214 to 1200 beds. A case was defined as the incident isolation of any Candida species from the blood of a surveillance area resident. Candidaemia episodes that occurred >30 days after the initial case were considered a new episode. Patients with candidaemia caused simultaneously by different species of Candida were excluded from the analysis because they had both CA and NAC species. Data collection
The case report was laboratory-based. All blood cultures in which a Candida species had been isolated were reported to the study coordinator (DR). The coordinator visited the hospital to confirm the infection, completed the case report form, and recorded the outcome. The clinical laboratories were periodically audited to ensure that all cases of candidaemia were reported. A standardized case report form was used to abstract the medical records. The data collected included patient demographic characteristics; comorbid conditions; concomitant infections; previous Candida colonization; exposure to antibiotic, antifungal and immunosuppressive therapy; total parenteral nutrition received; surgical procedures; central venous catheter (CVC) use; haemodialysis; mechanical ventilation; treatment of candidaemia and outcome. To measure disease severity, we used the Acute Physiology and Chronic Health Evaluation II score [17] for adult patients admitted to ICUs and the Karnofsky performance status scale for adults outside the ICU [18]. No standardized paediatric severity of illness score was used. Risk factors were assessed within 30 days prior to the diagnosis by positive sample. Microbiological methods
Candida detection and species identification were performed at the participating laboratories according to their standard
1677
protocols. Isolates were sent to the Mycology Reference Laboratory (MRL), National Center for Microbiology, Madrid, Spain, for species confirmation and antifungal susceptibility testing, using standard morphological and physiological methods, including fermentation of and growth on carbon sources, growth on nitrogen sources and growth at various temperatures. Candida parapsilosis ATCC 22019 and Candida krusei ATCC 6258 were used as quality control organisms for antifungal drug susceptibility testing [12]. When MRL and submitting laboratory identifications differed, the MRL identification was used for the study. MICs of fluconazole and voriconazole were determined by the broth microdilution method as recommended by the European Committee on Antimicrobial Susceptibility Testing (EUCAST). Interpretive breakpoints proposed by EUCAST for fluconazole and voriconazole resistance are >4 and >0.12 mg/L, respectively [19]. Statistical analysis
Population incidences of CA and NAC were calculated using denominator data obtained from the 2001 local census. Statistical analyses were performed using SPSS software, version 15.0 (SPSS Inc., Chicago, IL, USA). The results for categorical variables are expressed as a percentage, and numerical data are expressed as the mean (SD), median, and interquartile range for ages. The chi-square test or Fisher’s exact test (two-tailed) were used to compare categorical variables, and an unpaired Student’s t-test was used for continuous variables. Multivariate stepwise logistic regression analysis was carried out to identify predictors of candidaemia due to NAC. p <0.05 was considered statistically significant in the multivariate modeling.
Results During the study period, 345 cases of Candida BSI were detected. Six episodes of polyfungal candidaemia were excluded, leaving 339 episodes in 336 patients (median age 63 years, interquartile range: 41–72 years). A total of 176 (52%) patients had fungaemia due to C. albicans and 163 (48%) had infection due to NAC. The average annual incidence of candidaemia was 4.3 cases per 100 000 population (2.2 cases per 100 000 population for CA and 2.1 cases per 100 000 population for NAC). Demographics, clinical characteristics, underlying comorbid conditions and the outcome of patients with NAC BSI compared to those with CA BSI are summarized in Table 1. The NAC species isolated included 78 C. parapsilosis (23%), 34 Candida tropicalis (10%), 29 Candida glabrata (8.6%), 12 C. krusei (3.6%), three Candida kefyr (0.9%), two Candida
ª2010 The Authors Journal Compilation ª2010 European Society of Clinical Microbiology and Infectious Diseases, CMI, 16, 1676–1682
1678
Clinical Microbiology and Infection, Volume 16 Number 11, November 2010
CMI
TABLE 1. Demographics, epidemiological data and clinical
TABLE 2. Species distribution and in vitro susceptibilities to
characteristics of the patients included in the present study
fluconazole and voriconazole (European Committee on
Characteristics (n = 339 episodes in 336 patients)
C. albicans, Non-albicans n = 176 Candida spp., (52%) n = 163 (48%) p
Age (years), mean (SD) 57.9 (24.07) Elderly (>65 years) 95 (54) Male sex 101 (57.4) Nosocomial candidaemia 158 (89.8) Prior hospitalizationa 86 (48.9) APACHE II (severity high)b 29 (26.9) Length of stay before 30.1 (35.3) candidaemia (days), mean (SD) Location of the patients Medical ward 49 (27.8) Surgical ward 44 (25) ICU hospitalization (except neonates) 52 (29.5) c Neonatal ICU hospitalization 7 (4) Pediatric ward (excluding neonates) 7 (4) Outpatient 17 (9.7) d Predisposing factors Diabetes mellitus 44 (25) Chronic renal failure 73 (41.5) Solid tumour 45 (25.6) Haematological malignancy 16 (9.1) Neutropaenia 11 (6.3) Solid organ transplant recipient 2 (1.1) Haematological transplant recipient 1 (0.6) Surgery in previous 3 months 85 (48.3) e Prior immunosuppressive drugs 66 (37.5) Prior antibiotics (1 previous month) 158 (89.8) Prior fluconazole therapy 12 (6.8) CVC placement 130 (73.9) Total parenteral nutrition 57 (32.4) Previous Candida colonizationf 74 (42.0) Candiduria due to Candida albicans 36 (20.5) Portal of entry Primary candidaemia 105 (59.7) CVC-related candidaemia 54 (30.7) Secondary candidaemia 17 (9.7) g Optimal antifungal treatment 82 (52.2) Outcome Shock 46 (26.1) Mechanical ventilation 55 (31.3) ICU admission during outcome 15 (8.5) Acute renal failure 37 (21.0) Early mortality (<7 days) 47 (26.7) Overall mortality (<30 days) 75 (42.6)
47.7 66 98 149 77 20 29.7
(27.45) (40.5) (60.1) (91.4) (47.2) (19) (33.6)
0.011 0.013 0.659 0.573 0.770 0.176 0.752
64 22 37 17 9 14
(39.3) (13.5) (22.7) (10.4) (5.5) (8.6)
0.026 0.008 0.152 0.032 0.611 0.851
26 44 27 32 28 8 15 69 66 137 41 138 67 34 12
(16) (27.0) (16.6) (19.6) (17.2) (4.9) (9.2) (4236) (40.5) (84.0) (25.2) (84.7) (41.1) (20.9) (7.4)
0.037 0.006 0.043 0.005 0.002 0.054 <0.001 0.271 0.479 0.146 <0.001 0.015 0.096 <0.001 0.001
95 53 15 86
(58.3) (32.5) (9.2) (52.8)
0.797 0.717 1 0.36
43 46 10 26 35 60
(26.4) (28.2) (6.1) (16.0) (21.5) (36.8)
0.967 0.592 0.533 0.242 0.261 0.275
Categorical variables are expressed as total number and percentages, and numerical data are expressed as the mean and standard deviation (SD). Patients hospitalized within 3 months prior to candidaemia. APACHE, Acute Physiology and Chronic Health Evaluation score system. b Includes patients with APACHE >20 (only patients admitted to ICU). ICU, intensive care unit. c NAC in the neonatal ICU were due to Candida parapsilosis in 16 cases and Candida glabrata in one case. d Some patients had more than one predisposing factor. e Includes corticoids, chemotherapy, and other immunosuppressive drugs. CVC, central venous catheter. f Previous Candida colonization due to the same Candida spp. that caused candidaemia (includes candiduria). g Defined as at least 5 days of adequate antifungal treatment. a
Antimicrobial Susceptibility Testing breakpoint for Candida susceptibilitya)
Species
Fluconazole resistant isolates/n total isolates (%)
Voriconazole resistant isolates/n total isolates (%)
Candida Candida Candida Candida Candida Candida Candida Candida Candida Candida Candida Total
2/176 2/78 1/34 13/29 12/12 0/3 0/2 0/2 1/1 1/1 1/1 33/339
1/176 2/78 1/34 19/29 12/12 0/3 0/2 0/2 1/1 0/1 1/1 37/339
albicans parapsilosis tropicalis glabrata krusei kefyr lusitaniae guillierrmondii famata inconspicua norvegensis
(1.1) (2.6) (2.9) (44.8) (100) (0) (0) (0) (100) (100) (100) (9.7)
(0.6) (2.6) (2.9) (65.5) (100) (0) (0) (0) (100) (0) (100) (11)
a
Fluconazole-resistant isolates: MIC >4 mg/L. Voriconazole-resistant isolates: MIC >0.12 mg/L.
renal failure (41.5% CA vs. 27.2% NAC; p 0.006) and had been previously colonized with C. albicans (42.3% CA vs. 20.9% NAC; p <0.001). Cases with C. parapsilosis were more likely to be infants (28% vs. 5%; p <0.001) and have an indwelling CVC (97% vs. 87%; p 0.006), and less likely to die within 30 days (23% vs. 45%; p 0.001) compared to the other cases. C. krusei cases had received previous fluconazole (75% vs. 14%; p <0.001) and immunosuppressive therapy (75% vs. 38%; p 0.01) more often, were more frequently neutropaenic (67% vs. 9%; p <0.001) and had haematological malignancies (58% vs. 12%; p <0.001) more often than non-krusei cases. C. tropicalis cases were also likely to be neutropaenic (24% vs. 10%; p 0.04), have haematological malignancies (29% vs. 11%; p 0.007), and die within 30 days (56% vs. 38%), although this latter difference did not reach statistical significance (p 0.06). Overall mortality tended to be lower in patients with NAC BSI compared to CA BSI, although this trend did not reach statistical significance (36.8% NAC vs. 42.6% CA; p 0.27). Factors associated with NAC fungaemia were determined by univariate analysis (Table 1) and multivariate analysis TABLE 3. Multivariate analysis of risk factors associated with non-albicans Candida infection among all patients with
lusitaniae (0.6%), two Candida guilliermondii (0.6%), one Candida famata (0.3%), one Candida inconspicua (0.3%) and one Candida norvegensis (0.3%). Percentages of in vitro fluconazole and voriconazole resistance among the 339 incident bloodstream isolates of Candida spp. are shown in Table 2. Subgroup analysis for specific Candida spp. showed that patients with C. albicans fungaemia were admitted more often to surgical wards (25% CA vs. 13.5% NAC; p 0.008), had diabetes mellitus (25% CA vs. 16% NAC; p 0.03) or chronic
candidaemia Variable
OR
95% CI
p
Haematological transplant recipient Previous fluconazole usea Neonatal intensive care unit admission Previous antibiotic useb Previous Candida albicans colonization
10.8 4.47 4.42 0.42 0.33
1.31–90.01 2.12–9.43 1.63–12.04 0.21–0.85 0.19–0.57
0.027 <0.001 0.004 0.017 <0.001
a
At least 3 days of fluconazole treatment within 30 days prior to candidaemia. At least 3 days of any antibiotic treatment within 30 days prior to candidaemia.
b
ª2010 The Authors Journal Compilation ª2010 European Society of Clinical Microbiology and Infectious Diseases, CMI, 16, 1676–1682
CMI
Rodrı´guez et al. Predictors of non-albicans candidaemia
(Table 3). In the multivariate analysis, the fact of having undergone haematological transplantation (OR 10.8; 95% CI 1.31–90.1; p 0.027) or previous exposure to fluconazole (OR 4.47; 95% CI 2.12–9.43; p <0.001), and neonatal age (OR 4.42; 95% CI 1.63–12.04; p 0.004) were significantly associated with NAC fungaemia, whereas previous CA colonization (OR 0.33; 95% CI 0.19–0.57; p <0.001) and previous antibiotic use (OR 0.42; 95% CI 0.21–0.85; p 0.017) were associated with CA candidaemia.
Discussion In the general population of Barcelona, 48% of all cases of candidaemia were caused by NAC. The distribution of Candida species was generally similar to those reported from other European countries, such as France and Italy [20,21], with C. parapsilosis being the second most common Candida spp. isolated. An exception is the incidence of C. glabrata, which was the fourth most common species in the present study, but the second most common in England and Wales [22], Switzerland [23], Denmark [24] and the USA [10,25,26]. An increase in the percentage of C. glabrata BSI was first reported in the USA during the 1990s [6,25,26]. However, this trend is not universal, and C. glabrata is less common in other areas, including South America, Canada and some European countries (e.g. Italy and France) which, again, is in keeping with our results [20,21,27]. The changing pattern of candidaemia may be attributable in part to the large number of immunocompromised hosts and the widespread use of prophylactic or empiric antifungal therapy [9], although the role of widespread azole use in the emergence of species other than CA as causes of candidaemia remains controversial. Knowledge of the local epidemiological trends in Candida species isolated in blood cultures is important to guide the choice of empiric therapy. Previous studies of risk factors for candidaemia due to NAC compared to CA have focused on critically ill ICU patients or patients with a haematological malignancy [14– 16,28] and have investigated a single species, such as C. parapsilosis or C. glabrata [8,14,29,30]. To our knowledge, the risk factors for NAC vs. CA have not been prospectively described in a general patient population. The results obtained in the present study and those of Chow et al. [15] contrast with the results reported by Shorr et al. [16] who found no differences in the clinical characteristics of ICU patients with BSI caused by NAC or CA. These authors theorized that clinical variables do not allow one to successfully predict the microbiology of fungaemia in a particular patient because NAC species are essentially endemic
1679
[16]. Nonetheless, we found several differences in patients with NAC fungaemia compared to those with CA: they were significantly younger, had a higher incidence of haematological malignancies, were neutropaenic or had undergone haematological transplantation, or had been exposed to fluconazole previously. NAC was not associated with an increase in mortality or length of hospital stay compared to CA, probably because C. parapsilosis accounted for a considerable percentage of NAC isolates in our series, and this species is known to be less virulent, more frequently associated with a CVC infectious origin, and uniformly susceptible to fluconazole [8,30]. Subsequent to multivariate analysis, NAC fungaemia was found to be independently associated with having undergone haematological transplantation (p 0.027). The potential risk factors for developing NAC BSI identified in previous studies have included neutropaenia and an underlying haematological disease and, in these cases, C. tropicalis, C. krusei and even C. glabrata were more prevalent [7,14,28]. Because it is a general practice to use fluconazole prophylaxis in high-risk haematological patients, it is reasonable to conclude that undergoing haematological transplantation could be a confounding risk factor, with previous fluconazole exposure being the determining predictor in these cases. Although fluconazole prophylaxis is associated with a reduction in the incidence of candidaemia and attributable mortality, several studies have investigated the way in which fluconazole can influence the development of azole resistance [9,31]. In our series, the predominant Candida species varied according to the hospital care unit. C. parapsilosis was the most prevalent species in the neonatal ICU (16 of 24 [67%] cases), whereas CA was the most prevalent in surgical areas. In our series, all except one case of NAC among neonates were caused by C. parapsilosis. Considering that multivariate analysis identified neonatal age as a risk factor for NAC, we were able to determine that neonatal age is a predictor for C. parapsilosis fungaemia. The association between neonatal fungaemia and C. parapsilosis has been extensively described, and some studies suggest that it may reflect the intensive use of intravascular devices to treat neonates and children [32,33]. We also found that fluconazole exposure is a risk factor for the development of NAC BSI. The results obtained in the present study agree with the findings reported in previous studies postulating that widespread fluconazole use would result in selection of yeast species that are less sensitive to this antifungal agent, such as C. krusei, C. glabrata and C. tropicalis [9,15,31]. Fluconazole use and selection of fluconazole-resistant Candida spp. has been extensively described in studies performed in ICU patients and in patients with
ª2010 The Authors Journal Compilation ª2010 European Society of Clinical Microbiology and Infectious Diseases, CMI, 16, 1676–1682
1680
Clinical Microbiology and Infection, Volume 16 Number 11, November 2010
cancer or haematological malignancies [3,30,31]. In patients without cancer, the association between previous fluconazole exposure and NAC fungaemia is not as clear, with some studies suggesting an independent relationship and others failing to note this association [15,16,34]. Moreover, the important percentage of C. parapsilosis, a yeast species that is almost always susceptible to fluconazole, is not explained by the increase of fluconazole use. It is likely that fungemia due to C. parapsilosis reflects acquisition associated with the widespread use of parenteral feeding and intravascular devices [30–32]. We also found that patients were more likely to develop CA fungemia if they had been treated previously with antibiotics at the onset of candidaemia or if they had been colonized previously by CA. Previous antibiotic exposure has been associated with CA in non-neutropaenic patients [34] and previous Candida colonization by the same Candida spp. has also been described as a risk factor for candidaemia [35]. The observations made in the present study are subject to limitations. First, the data were obtained 6 years ago. Several new treatment options subsequently have been developed (e.g. echinocandins and new triazoles) but, in all probability, the epidemiology and species distribution of Candida has not changed substantially, and these data are still relevant. Second, only the presence or absence of risk factor exposure was considered, and not the duration of exposure. Because the study was not designed to quantify the length of exposure, this variable is not available and any associated bias cannot be assessed. Third, it may not have been appropriate to pool all NAC species together because certain NAC species are uniformly susceptible to fluconazole. To address this limitation, other studies [16] have analyzed C. krusei and C. glabrata together vs. C. albicans, C. tropicalis and C. parapsilosis. However, the main goal of the present study was not to investigate potential fluconazole resistance but, instead, to identify predictors of candidaemia due to NAC. Nonetheless, the small number of fluconazole-resistant isolates precluded a risk factor analysis for individual Candida species. On the other hand, some strengths of the present study include the fact that risk factors for NAC fungaemia were examined in a general population, the sample size is relatively large, and the study design is prospective. In summary, a number of factors were found to be independently associated with an increased risk of candidaemia due to NAC compared to CA. Neonatal age, having undergone haematological transplantation, and prior fluconazole use were associated with NAC BSI. Physicians can take these specific characteristics into consideration when empiric treatment with an antifungal agent must be started while waiting for Candida species to be identified.
CMI
Appendix Other members of the Barcelona Candidaemia Project Study Group
S. Fridkin, R. Hajjeh, B. Park, J. Morgan, D. W. Warnock (Centers for Disease Control and Prevention, Atlanta, GA); A. M. Planes (Hospital de Vall d’Hebron, Barcelona, Spain); M. Almela, F. M. Reverter, C. M. Soler (Hospital Clinic, IDIBAPS, Barcelona, Spain); M. Salvado´, P. Saballs (Hospital del Mar, Barcelona, Spain); A. Gener (Hospital Sant Joan de Deu, Esplugues de Llobregat, Barcelona, Spain); D. Fontanals (Hospital Parc Taulıı´, Sabadell, Barcelona, Spain); M. Xercavins (Hospital Mutua de Terrassa, Terrassa, Barcelona, Spain); L. Falgueras, M. T. Torroella, M. de Ramon (Hospital General de Catalunya, Sant Cugat del Valles, Barcelona, Spain); Carles Alonso (Hospital Creu Roja, Hospitalet de Llobregat, Barcelona, Spain); J. de Otero (Hospital Creu Roja, Barcelona, Spain); M. Sierra and J. Martinez-Montauti (Hospital de Barcelona, Barcelona, Spain); M. A. Morera (Hospital de Terrassa, Terrassa, Barcelona, Spain); and M. Gime´nez (Hospital Universitari Germans Trias i Pujol, Badalona, Barcelona, Spain).
Author Contributions D. Rodriguez, B. Almirante and A. Pahissa made a substantial contribution to the conception and design of the manuscript; acquisition, analysis and interpretation of data; drafting the article and revising it critically for important intellectual content; and giving final approval of the version to be published. M. Cuenca-Estrella and J. L. Rodriguez-Tudela performed all microbiological studies in the Mycology Reference Laboratory and made a substantial contribution to the acquisition and interpretation of data; revising the manuscript critically for important intellectual content; and giving final approval of the version to be published. J. Mensa, J. Ayats and F. Sanchez made a substantial contribution to the acquisition and interpretation of data; revising the manuscript critically for important intellectual content; and giving final approval of the version to be published.
Acknowledgements This study was presented, in part, as a poster (M-2141a) at the 48th Annual Meeting of the Interscience Conference on Antimicrobial Agents and Chemotherapy/Infectious Diseases Society of America, Washington, DC, 25–28 October 2008. We thank C. Cavallo for assistance with the English language.
ª2010 The Authors Journal Compilation ª2010 European Society of Clinical Microbiology and Infectious Diseases, CMI, 16, 1676–1682
CMI
Rodrı´guez et al. Predictors of non-albicans candidaemia
Transparency Declaration 13.
This study was supported by research grants from Pfizer, Inc., Gilead Sciences S.A., and the Sociedad Espan˜ola de Enfermedades Infecciosas y Microbiologia Clinica (SEIMC, Spanish Society of Infectious Diseases and Clinical Microbiology) and by a medical research grant from the Ministerio de Sanidad y Consumo, Instituto de Salud Carlos III, Spanish Network for the Research in Infectious Diseases (REIPI RD 06/ 0008). None of the authors has any conflicts of interest.
14.
15.
16.
17.
References 1. Edmond MB, Wallace SE, McClish DK, Pfaller MA, Jones RN, Wenzel RP. Nosocomial bloodstream infections in United States hospitals: a three-year analysis. Clin Infect Dis 1999; 29: 239–244. 2. Pfaller MA, Jones RN, Messer SA, Edmond MB, Wenzel RP. National surveillance of nosocomial blood stream infection due to Candida albicans: frequency of occurrence and antifungal susceptibility in the SCOPE Program. Diagn Microbiol Infect Dis 1998; 31: 327– 332. 3. Bassetti M, Righi E, Costa A et al. Epidemiological trends in nosocomial candidemia in intensive care. BMC Infect Dis 2006; 6: 21. 4. Krcmery V, Barnes AJ. Non-albicans Candida spp. causing fungaemia: pathogenicity and antifungal resistance. J Hosp Infect 2002; 50: 243–260. 5. Pfaller MA, Diekema DJ, Jones RN et al. International surveillance of bloodstream infections due to Candida species: frequency of occurrence and in vitro susceptibilities to fluconazole, ravuconazole, and voriconazole of isolates collected from 1997 through 1999 in the SENTRY antimicrobial surveillance program. J Clin Microbiol 2001; 39: 3254–3259. 6. Trick WE, Fridkin SK, Edwards JR, Hajjeh RA, Gaynes RP. Secular trend of hospital-acquired candidemia among intensive care unit patients in the United States during 1989-1999. Clin Infect Dis 2002; 35: 627–630. 7. Almirante B, Rodriguez D, Park BJ et al. Epidemiology and predictors of mortality in cases of Candida bloodstream infection: results from population-based surveillance, Barcelona, Spain, from 2002 to 2003. J Clin Microbiol 2005; 43: 1829–1835. 8. Almirante B, Rodriguez D, Cuenca-Estrella M et al. Epidemiology, risk factors, and prognosis of Candida parapsilosis bloodstream infections: case-control population-based surveillance study of patients in Barcelona, Spain, from 2002 to 2003. J Clin Microbiol 2006; 44: 1681– 1685. 9. Blot S, Janssens R, Claeys G et al. Effect of fluconazole consumption on long-term trends in candidal ecology. J Antimicrob Chemother 2006; 58: 474–477. 10. Horn DL, Neofytos D, Anaissie EJ et al. Epidemiology and outcomes of candidemia in 2019 patients: data from the prospective antifungal therapy alliance registry. Clin Infect Dis 2009; 48: 1695–1703. 11. Pfaller MA, Messer SA, Boyken L. Geographic variation in the susceptibilities of invasive isolates of Candida glabrata to seven systemically active antifungal agents: a global assessment from the ARTEMIS Antifungal Surveillance Program conducted in 2001 and 2002. J Clin Microbiol 2004; 42: 3142–3146. 12. Cuenca-Estrella M, Rodriguez D, Almirante B et al. In vitro susceptibilities of bloodstream isolates of Candida species to six antifungal agents: results from a population-based active surveillance
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
1681
programme, Barcelona, Spain, 2002-2003. J Antimicrob Chemother 2005; 55: 194–199. Charlier C, Hart E, Lefort A et al. Fluconazole for the management of invasive candidiasis: where do we stand after 15 years? J Antimicrob Chemother 2006; 57: 384–410. Bodey GP, Mardani M, Hanna HA et al. The epidemiology of Candida glabrata and Candida albicans fungemia in immunocompromised patients with cancer. Am J Med 2002; 112: 380–385. Chow JK, Golan Y, Ruthazer R et al. Factors associated with candidemia caused by non-albicans Candida species versus Candida albicans in the intensive care unit. Clin Infect Dis 2008; 46: 1206–1213. Shorr AF, Lazarus DR, Sherner JH et al. Do clinical features allow for accurate prediction of fungal pathogenesis in bloodstream infections? Potential implications of the increasing prevalence of non-albicans candidemia. Crit Care Med 2007; 35: 1077–1083. Chang RW, Jacobs S, Lee B. Predicting outcome among intensive care unit patients using computerised trend analysis of daily Apache II scores corrected for organ system failure. Intensive Care Med 1988; 14: 558–566. Motzer RJ, Mazumdar M, Bacik J, Berg W, Amsterdam A, Ferrara J. Survival and prognostic stratification of 670 patients with advanced renal cell carcinoma. J Clin Oncol 1999; 17: 2530–2540. Rodriguez-Tudela JL, Arendrup MC, Barchiesi F et al. EUCAST definitive document EDef 7.1: method for the determination of broth dilution MICs of antifungal agents for fermentative yeasts. Clin Microbiol Infect 2008; 14: 398–405. Richet H, Roux P, Des CC, Esnault Y, Andremont A. Candidemia in French hospitals: incidence rates and characteristics. Clin Microbiol Infect 2002; 8: 405–412. Tortorano AM, Biraghi E, Astolfi A et al. European Confederation of Medical Mycology (ECMM) prospective survey of candidaemia: report from one Italian region. J Hosp Infect 2002; 51: 297–304. Kibbler CC, Seaton S, Barnes RA et al. Management and outcome of bloodstream infections due to Candida species in England and Wales. J Hosp Infect 2003; 54: 18–24. Marchetti O, Bille J, Fluckiger U et al. Epidemiology of candidemia in Swiss tertiary care hospitals: secular trends, 1991-2000. Clin Infect Dis 2004; 38: 311–320. Arendrup MC, Fuursted K, Gahrn-Hansen B et al. Semi-national surveillance of fungaemia in Denmark 2004–2006: increasing incidence of fungaemia and numbers of isolates with reduced azole susceptibility. Clin Microbiol Infect 2008; 14: 487–494. Diekema DJ, Messer SA, Brueggemann AB et al. Epidemiology of candidemia: 3-year results from the emerging infections and the epidemiology of Iowa organisms study. J Clin Microbiol 2002; 40: 1298–1302. Hajjeh RA, Sofair AN, Harrison LH et al. Incidence of bloodstream infections due to Candida species and in vitro susceptibilities of isolates collected from 1998 to 2000 in a population-based active surveillance program. J Clin Microbiol 2004; 42: 1519–1527. Pfaller MA, Jones RN, Doern GV, Sader HS, Hollis RJ, Messer SA. International surveillance of bloodstream infections due to Candida species: frequency of occurrence and antifungal susceptibilities of isolates collected in 1997 in the United States, Canada, and South America for the SENTRY Program. The SENTRY Participant Group. J Clin Microbiol 1998; 36: 1886–1889. Viscoli C, Girmenia C, Marinus A et al. Candidemia in cancer patients: a prospective, multicenter surveillance study by the Invasive Fungal Infection Group (IFIG) of the European Organization for Research and Treatment of Cancer (EORTC). Clin Infect Dis 1999; 28: 1071–1079. Fairchild KD, Tomkoria S, Sharp EC, Mena FV. Neonatal Candida glabrata sepsis: clinical and laboratory features compared with other Candida species. Pediatr Infect Dis J 2002; 21: 39–43.
ª2010 The Authors Journal Compilation ª2010 European Society of Clinical Microbiology and Infectious Diseases, CMI, 16, 1676–1682
1682
Clinical Microbiology and Infection, Volume 16 Number 11, November 2010
30. Weems JJ Jr. Candida parapsilosis: epidemiology, pathogenicity, clinical manifestations, and antimicrobial susceptibility. Clin Infect Dis 1992; 14: 756–766. 31. Marr KA, Seidel K, White TC, Bowden RA. Candidemia in allogeneic blood and marrow transplant recipients: evolution of risk factors after the adoption of prophylactic fluconazole. J Infect Dis 2000; 181: 309–316. 32. Levin AS, Costa SF, Mussi NS et al. Candida parapsilosis fungemia associated with implantable and semi-implantable central venous catheters
CMI
and the hands of healthcare workers. Diagn Microbiol Infect Dis 1998; 30: 243–249. 33. Rodriguez D, Almirante B, Park BJ et al. Candidemia in neonatal intensive care units: Barcelona, Spain. Pediatr Infect Dis J 2006; 25: 224–229. 34. Davis SL, Vazquez JA, McKinnon PS. Epidemiology, risk factors, and outcomes of Candida albicans versus non-albicans candidemia in nonneutropenic patients. Ann Pharmacother 2007; 41: 568–573. 35. Ostrosky-Zeichner L. New approaches to the risk of Candida in the intensive care unit. Curr Opin Infect Dis 2003; 16: 533–537.
ª2010 The Authors Journal Compilation ª2010 European Society of Clinical Microbiology and Infectious Diseases, CMI, 16, 1676–1682