mycoses
Diagnosis,Therapy and Prophylaxis of Fungal Diseases
Original article
Epidemiology of candidaemia in a tertiary care university hospital: 10-year experience with 381 candidaemia episodes between 2001 and 2010 Sehnaz Alp,1 Sevtap Arikan-Akdagli,2 Dolunay Gulmez,2 Sibel Ascioglu,1 Omrum Uzun1 and Murat Akova1 1 Faculty of Medicine, Department of Infectious Diseases and Clinical Microbiology, Hacettepe University, Ankara, Turkey and 2Faculty of Medicine, Department of Medical Microbiology, Hacettepe University, Ankara, Turkey
Summary
Defining the epidemiology of and risk factors for candidaemia is necessary to guide empirical treatment. The objectives of this study were to determine the ranking of Candida among positive blood cultures, to define the epidemiology of candidaemia and to investigate patient characteristics and their relationship with C. albicans vs. non-albicans Candida (NAC) candidaemia. Candidaemia episodes between January 2001 and December 2010 were evaluated retrospectively. Patient characteristics were compared across Candida species. Candida ranked as the fifth most frequently isolated pathogen. Among 381 candidaemia episodes, 58.3% were due to C. albicans, followed by C. parapsilosis (15.2%), C. tropicalis (13.4%) and C. glabrata (6.8%). No statistically significant difference was observed in the distribution of C. albicans vs. NAC (P = 0.432). Patients with NAC had significantly higher rates of haematological disorders (P < 0.001) and neutropenia (P = 0.003), and were older (P = 0.024) than patients with C. albicans, whereas patients with urinary catheters had higher rates of C. albicans (P = 0.007). On species basis, C. tropicalis was more frequently isolated from patients with haematological disorders (P < 0.001) and neutropenia (P = 0.008). Patients with urinary catheters were less likely to have C. parapsilosis (P = 0.043). C. glabrata was most prevalent among patients with solid organ tumours (P = 0.038), but not evident in patients with haematological disorders. Local epidemiological features and risk factors may have important implications for the management of candidaemia.
Key words: Candidaemia, epidemiology, Candida albicans, non-albicans Candida.
Introduction Correspondence: S. Alp, Faculty of Medicine, Department of Infectious Diseases and Clinical Microbiology, Hacettepe University, 06100 SihhiyeAnkara, Turkey. Tel.: +90 312 3051296. Fax: +90 312 3104179. E-mail: [email protected] This study was presented in part at 22nd European Congress of Clinical Microbiology and Infectious Diseases (ECCMID), 31 March–3 April 2012, London, United Kingdom. Abst. No. 3233. Poster No. P777. Clin Microbiol Infect 2012; 18 (Suppl. 3): 153. Submitted for publication 11 December 2014 Revised 6 February 2015 Accepted for publication 26 March 2015
doi:10.1111/myc.12349
The incidence of invasive fungal infections (IFIs) has significantly increased due to the growing number of individuals at risk of developing these life-threatening infections. Candida spp. are the most common pathogens responsible for IFIs in humans.1–4 Available data have shown Candida spp. to be among the top five pathogens causing nosocomial bloodstream infections in the United States and among the top ten in Europe.5–9 The major risk factors for candidaemia mainly involve specific clinical conditions and healthcare-associated factors.3,4 The mortality rate for
© 2015 Blackwell Verlag GmbH Mycoses, 2015, 58, 498–505
Ten-year study of candidaemia
candidaemia is very high, and delay in initiation or inappropriate use of antifungal therapy contributes to increased mortality.10–14 While Candida albicans is still considered the most common species responsible for candidaemia, a gradual decrease in the incidence of C. albicans and a remarkable shift from C. albicans to non-albicans Candida (NAC) species have been reported worldwide.15–18 In addition to reports of azole resistance among NAC species, co-resistance to echinocandins in fluconazole-resistant C. glabrata isolates has also emerged.17–25 Changes in candidaemia epidemiology have a profound impact on the choice of preferred empirical antifungal treatment. The present study was undertaken to (i) determine the ranking of Candida isolates among all positive blood cultures; (ii) define the species distribution of Candida strains responsible for candidaemia and establish whether there is any potential increase in NAC vs. C. albicans and (iii) investigate the demographic characteristics, underlying diseases and predisposing factors of patients with candidaemia, and determine whether any correlation exists between patient characteristics and risk of developing candidaemia due to C. albicans vs. NAC species.
Patients and Methods Collection of patient data
We retrospectively evaluated the demographic and clinical data of adult patients (age ≥18 years) reported to have at least one positive blood culture yielding Candida species according to the records of the mycology laboratory in the Hacettepe University Adult Hospital from January 2001 to December 2010. The setting was a 1000-bedded tertiary care university hospital with medical and surgical wards, intensive care units (ICUs) and an emergency department. Individual patient files were searched and the following data were obtained for all patients with candidaemia: age, sex, underlying diseases, presence of central venous and urinary catheter, prior antimicrobial therapy, total parenteral nutrition (TPN), neutropenia, ICU stay, history of surgery (gastrointestinal surgery, gastrointestinal procedures, cardiac surgery and others) within 30 days prior to the date of candidaemia. In the case of multiple candidaemia episodes during the same hospitalisation period, only the first episode was included even if the patient had subsequent episodes caused by different Candida species. The study protocol was approved by the Non-interventional Clinical Researches Ethics Board of the
© 2015 Blackwell Verlag GmbH Mycoses, 2015, 58, 498–505
Hacettepe University. Written informed consent was not required because of the retrospective nature of this study. Microbiological methods
At our institution, the BD BACTEC 9240 Blood Culture System (Becton Dickinson Diagnostic Instrument Systems, Towson, MD, USA) was used in routine practice during the study period. Blood specimens cultured in BACTEC medium (Becton, Dickinson & Company, Clare, Ireland) were incubated for 7 days. During this period, if the automated alert system signalled any growth in cultivated bottles, Gram stained samples from the bottles were promptly examined microscopically for visualisation of the infecting microorganism. Concomitantly, subcultures on blood, chocolate and eosin methylene blue agar plates were performed and placed in a 35–37 °C incubator (with 5–10% CO2 for blood and chocolate agar) for 48 h. Blind subcultures were also performed from the cultivated bottles at the end of 7 days if the automated system did not signal any growth until that time. Three to five drops of media from each cultivated bottle were subcultured onto chocolate agar plates, and incubated in 5–10% CO2 at 35–37 °C for 48 h. The final report was entered as no growth if there were not any colonies on agar media. If any yeast growth was detected on the agar plates, isolates were identified to a species level using standard mycological methods.26 During the study period, there were no changes to microbiological procedures and the identification process. Definitions of terms used in patient data
Candidaemia was defined as one or more positive blood culture yielding Candida spp. in patients with clinically apparent signs and symptoms of infection such as presence of fever and/or systemic inflammatory response syndrome or sepsis criteria. Neutropenia was defined as an absolute neutrophil count of less than 500 cells ml 1. Peripherally inserted central catheters and port catheters were accepted as central venous catheters. Endoscopic retrograde cholangiopancreatography, insertion of a biliary drainage catheter and percutaneous endoscopic gastrostomy tubes were considered as gastrointestinal procedures other than gastrointestinal surgery. Prior antimicrobial therapy was defined as the administration of any antimicrobial drug for at least 3 days within the month preceding the candidaemia episode.
499
S. Alp et al.
Statistical analyses
Statistical analyses were performed using STATA V.13 (STATA Corp. 2013, College Station, TX, USA). Oneway analysis of variance was used to compare ages across Candida species. Fisher’s exact test was used to compare proportions of categorical variables. A Pvalue of 15%) from the Middle East, South America, East Asia and Southeast Asia.16,18,21,27–34,44–54 C. glabrata accounted for only 6.8% of Candida isolates in our centre; far below the rates of European (>14%) and North American centres (>20%).16,18,21,27–34,47,48 However,
501
S. Alp et al.
Table 2 Demographic data and clinical characteristics of patients with candidaemia according to Candida albicans, C. parapsilosis, C. tropicalis and C. glabrata.
Mean age (Mean SD) Male gender Underlying diseases Solid organ tumours and solid organ transplantation1 Haematological disorders Cardiovascular diseases Diabetes mellitus Chronic renal failure Rheumatological disorders Predisposing factors Prior antimicrobial therapy Presence of urinary catheter Use of central venous catheter Total parenteral nutrition ICU stay History of gastrointestinal surgery or gastrointestinal procedure History of surgery other than gastrointestinal surgery and gastrointestinal procedure Neutropenia
C. albicans candidaemia (n = 222) (%)
C. parapsilosis candidaemia (n = 58) (%)
C. tropicalis candidaemia (n = 51) (%)
C. glabrata candidaemia (n = 26) (%)
55.2 16.6 108 (48.6)
58.8 17.4 31 (53.4)
59.9 16.4 29 (56.9)
60.9 14.7 11 (42.3)
0.106 0.576
85 (38.3)
18 (31.0)
12 (23.5)
14 (53.8)
0.038
23 48 18 9 8
(10.4) (21.6) (8.1) (4.0) (3.6)
11 11 5 1 –
(19.0) (19.0) (8.6) (1.7)
20 8 3 2 4
(39.2) (15.7) (5.9) (3.9) (7.8)
– 4 3 3 1
(15.4) (11.5) (11.5) (3.8)
Diagnosis,Therapy and Prophylaxis of Fungal Diseases
Original article
Epidemiology of candidaemia in a tertiary care university hospital: 10-year experience with 381 candidaemia episodes between 2001 and 2010 Sehnaz Alp,1 Sevtap Arikan-Akdagli,2 Dolunay Gulmez,2 Sibel Ascioglu,1 Omrum Uzun1 and Murat Akova1 1 Faculty of Medicine, Department of Infectious Diseases and Clinical Microbiology, Hacettepe University, Ankara, Turkey and 2Faculty of Medicine, Department of Medical Microbiology, Hacettepe University, Ankara, Turkey
Summary
Defining the epidemiology of and risk factors for candidaemia is necessary to guide empirical treatment. The objectives of this study were to determine the ranking of Candida among positive blood cultures, to define the epidemiology of candidaemia and to investigate patient characteristics and their relationship with C. albicans vs. non-albicans Candida (NAC) candidaemia. Candidaemia episodes between January 2001 and December 2010 were evaluated retrospectively. Patient characteristics were compared across Candida species. Candida ranked as the fifth most frequently isolated pathogen. Among 381 candidaemia episodes, 58.3% were due to C. albicans, followed by C. parapsilosis (15.2%), C. tropicalis (13.4%) and C. glabrata (6.8%). No statistically significant difference was observed in the distribution of C. albicans vs. NAC (P = 0.432). Patients with NAC had significantly higher rates of haematological disorders (P < 0.001) and neutropenia (P = 0.003), and were older (P = 0.024) than patients with C. albicans, whereas patients with urinary catheters had higher rates of C. albicans (P = 0.007). On species basis, C. tropicalis was more frequently isolated from patients with haematological disorders (P < 0.001) and neutropenia (P = 0.008). Patients with urinary catheters were less likely to have C. parapsilosis (P = 0.043). C. glabrata was most prevalent among patients with solid organ tumours (P = 0.038), but not evident in patients with haematological disorders. Local epidemiological features and risk factors may have important implications for the management of candidaemia.
Key words: Candidaemia, epidemiology, Candida albicans, non-albicans Candida.
Introduction Correspondence: S. Alp, Faculty of Medicine, Department of Infectious Diseases and Clinical Microbiology, Hacettepe University, 06100 SihhiyeAnkara, Turkey. Tel.: +90 312 3051296. Fax: +90 312 3104179. E-mail: [email protected] This study was presented in part at 22nd European Congress of Clinical Microbiology and Infectious Diseases (ECCMID), 31 March–3 April 2012, London, United Kingdom. Abst. No. 3233. Poster No. P777. Clin Microbiol Infect 2012; 18 (Suppl. 3): 153. Submitted for publication 11 December 2014 Revised 6 February 2015 Accepted for publication 26 March 2015
doi:10.1111/myc.12349
The incidence of invasive fungal infections (IFIs) has significantly increased due to the growing number of individuals at risk of developing these life-threatening infections. Candida spp. are the most common pathogens responsible for IFIs in humans.1–4 Available data have shown Candida spp. to be among the top five pathogens causing nosocomial bloodstream infections in the United States and among the top ten in Europe.5–9 The major risk factors for candidaemia mainly involve specific clinical conditions and healthcare-associated factors.3,4 The mortality rate for
© 2015 Blackwell Verlag GmbH Mycoses, 2015, 58, 498–505
Ten-year study of candidaemia
candidaemia is very high, and delay in initiation or inappropriate use of antifungal therapy contributes to increased mortality.10–14 While Candida albicans is still considered the most common species responsible for candidaemia, a gradual decrease in the incidence of C. albicans and a remarkable shift from C. albicans to non-albicans Candida (NAC) species have been reported worldwide.15–18 In addition to reports of azole resistance among NAC species, co-resistance to echinocandins in fluconazole-resistant C. glabrata isolates has also emerged.17–25 Changes in candidaemia epidemiology have a profound impact on the choice of preferred empirical antifungal treatment. The present study was undertaken to (i) determine the ranking of Candida isolates among all positive blood cultures; (ii) define the species distribution of Candida strains responsible for candidaemia and establish whether there is any potential increase in NAC vs. C. albicans and (iii) investigate the demographic characteristics, underlying diseases and predisposing factors of patients with candidaemia, and determine whether any correlation exists between patient characteristics and risk of developing candidaemia due to C. albicans vs. NAC species.
Patients and Methods Collection of patient data
We retrospectively evaluated the demographic and clinical data of adult patients (age ≥18 years) reported to have at least one positive blood culture yielding Candida species according to the records of the mycology laboratory in the Hacettepe University Adult Hospital from January 2001 to December 2010. The setting was a 1000-bedded tertiary care university hospital with medical and surgical wards, intensive care units (ICUs) and an emergency department. Individual patient files were searched and the following data were obtained for all patients with candidaemia: age, sex, underlying diseases, presence of central venous and urinary catheter, prior antimicrobial therapy, total parenteral nutrition (TPN), neutropenia, ICU stay, history of surgery (gastrointestinal surgery, gastrointestinal procedures, cardiac surgery and others) within 30 days prior to the date of candidaemia. In the case of multiple candidaemia episodes during the same hospitalisation period, only the first episode was included even if the patient had subsequent episodes caused by different Candida species. The study protocol was approved by the Non-interventional Clinical Researches Ethics Board of the
© 2015 Blackwell Verlag GmbH Mycoses, 2015, 58, 498–505
Hacettepe University. Written informed consent was not required because of the retrospective nature of this study. Microbiological methods
At our institution, the BD BACTEC 9240 Blood Culture System (Becton Dickinson Diagnostic Instrument Systems, Towson, MD, USA) was used in routine practice during the study period. Blood specimens cultured in BACTEC medium (Becton, Dickinson & Company, Clare, Ireland) were incubated for 7 days. During this period, if the automated alert system signalled any growth in cultivated bottles, Gram stained samples from the bottles were promptly examined microscopically for visualisation of the infecting microorganism. Concomitantly, subcultures on blood, chocolate and eosin methylene blue agar plates were performed and placed in a 35–37 °C incubator (with 5–10% CO2 for blood and chocolate agar) for 48 h. Blind subcultures were also performed from the cultivated bottles at the end of 7 days if the automated system did not signal any growth until that time. Three to five drops of media from each cultivated bottle were subcultured onto chocolate agar plates, and incubated in 5–10% CO2 at 35–37 °C for 48 h. The final report was entered as no growth if there were not any colonies on agar media. If any yeast growth was detected on the agar plates, isolates were identified to a species level using standard mycological methods.26 During the study period, there were no changes to microbiological procedures and the identification process. Definitions of terms used in patient data
Candidaemia was defined as one or more positive blood culture yielding Candida spp. in patients with clinically apparent signs and symptoms of infection such as presence of fever and/or systemic inflammatory response syndrome or sepsis criteria. Neutropenia was defined as an absolute neutrophil count of less than 500 cells ml 1. Peripherally inserted central catheters and port catheters were accepted as central venous catheters. Endoscopic retrograde cholangiopancreatography, insertion of a biliary drainage catheter and percutaneous endoscopic gastrostomy tubes were considered as gastrointestinal procedures other than gastrointestinal surgery. Prior antimicrobial therapy was defined as the administration of any antimicrobial drug for at least 3 days within the month preceding the candidaemia episode.
499
S. Alp et al.
Statistical analyses
Statistical analyses were performed using STATA V.13 (STATA Corp. 2013, College Station, TX, USA). Oneway analysis of variance was used to compare ages across Candida species. Fisher’s exact test was used to compare proportions of categorical variables. A Pvalue of 15%) from the Middle East, South America, East Asia and Southeast Asia.16,18,21,27–34,44–54 C. glabrata accounted for only 6.8% of Candida isolates in our centre; far below the rates of European (>14%) and North American centres (>20%).16,18,21,27–34,47,48 However,
501
S. Alp et al.
Table 2 Demographic data and clinical characteristics of patients with candidaemia according to Candida albicans, C. parapsilosis, C. tropicalis and C. glabrata.
Mean age (Mean SD) Male gender Underlying diseases Solid organ tumours and solid organ transplantation1 Haematological disorders Cardiovascular diseases Diabetes mellitus Chronic renal failure Rheumatological disorders Predisposing factors Prior antimicrobial therapy Presence of urinary catheter Use of central venous catheter Total parenteral nutrition ICU stay History of gastrointestinal surgery or gastrointestinal procedure History of surgery other than gastrointestinal surgery and gastrointestinal procedure Neutropenia
C. albicans candidaemia (n = 222) (%)
C. parapsilosis candidaemia (n = 58) (%)
C. tropicalis candidaemia (n = 51) (%)
C. glabrata candidaemia (n = 26) (%)
55.2 16.6 108 (48.6)
58.8 17.4 31 (53.4)
59.9 16.4 29 (56.9)
60.9 14.7 11 (42.3)
0.106 0.576
85 (38.3)
18 (31.0)
12 (23.5)
14 (53.8)
0.038
23 48 18 9 8
(10.4) (21.6) (8.1) (4.0) (3.6)
11 11 5 1 –
(19.0) (19.0) (8.6) (1.7)
20 8 3 2 4
(39.2) (15.7) (5.9) (3.9) (7.8)
– 4 3 3 1
(15.4) (11.5) (11.5) (3.8)
