- Email: [email protected]
Nosocomial Candida parapsilosis candidaemia: risk factors, antifungal susceptibility and outcome
Journal of Hospital Infection 87 (2014) 54e58 Available online at www.sciencedirect.com
Journal of Hospital Infection journal homepage: www.elsevierhealth.com/journals/jhin
Short report
Nosocomial Candida parapsilosis candidaemia: risk factors, antifungal susceptibility and outcome Y. Hirai a, b, *, S. Asahata a, Y. Ainoda a, A. Goto c, T. Fujita a, K. Totsuka a a
Department of Infectious Diseases, Tokyo Women’s Medical University Hospital, Shinjuku-ku, Tokyo, Japan Department of Haematology, Tokyo Women’s Medical University Hospital, Shinjuku-ku, Tokyo, Japan c Laboratory of Microbiology, Tokyo Women’s Medical University Hospital, Shinjuku-ku, Tokyo, Japan b
A R T I C L E
I N F O
Article history: Received 24 October 2013 Accepted 10 February 2014 Available online 12 March 2014 Keywords: Candida parapsilosis Candidaemia Minimum inhibitory concentration Minimum fungicidal concentration Cardiovascular prosthetic materials
S U M M A R Y
A retrospective analysis was undertaken from 2000 to 2010 to show the risk factors associated with death within 30 days in patients with C. parapsilosis candidaemia (CPC). Fifty-one cases of nosocomial CPC were included in the analysis. All isolates from blood cultures were susceptible to micafungin and fluconazole. The overall mortality rate was 23.5%, and the most severe complications were endocarditis (5.9%) and endophthalmitis (5.9%). On multi-variate analysis, APACHE II score >25 (odds ratio 43.9) and retained cardiovascular prosthetic materials (RCPM) (prosthetic valve or graft) (odds ratio 14.6) were found to be risk factors associated with death. Prompt surgical removal should be considered in CPC patients with RCPM. ª 2014 The Healthcare Infection Society. Published by Elsevier Ltd. All rights reserved.
Introduction Candida spp. remain a major cause of morbidity and mortality in healthcare settings. Candida parapsilosis is often the second or third most commonly isolated Candida spp. from blood cultures.1 Intravenous catheters and parenteral nutrition are well-known risk factors. C. parapsilosis demonstrates an * Corresponding author. Address: Department of Infectious Diseases, Tokyo Women’s Medical University Hospital, 8-1 Kawada-cho, Shinjuku-ku, Tokyo, Japan. Tel.: þ81 3 3353 8111. E-mail address: [email protected] (Y. Hirai).
increased in-vitro minimum inhibitory concentration (MIC) to echinocandins such as caspofungin.1 Recent studies have shown increasing incidence of C. parapsilosis candidaemia (CPC) of 7e24%.2 Although risk factors associated with death attributed to candidaemia have been documented in the literature, any specific risk factors associated with nosocomial CPC are not known. The primary outcome of this study was to establish the risk factors associated with death within 30 days by multivariate analysis. The secondary outcome was to determine antifungal MICs and minimum fungicidal concentrations (MFCs) for C. parapsilosis isolated from blood cultures.
0195-6701/$ e see front matter ª 2014 The Healthcare Infection Society. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jhin.2014.02.006
Y. Hirai et al. / Journal of Hospital Infection 87 (2014) 54e58
Patients and methods Study design
55
Spearman’s rank correlation coefficient was used to show correlation between antifungal MIC and MFC values. P < 0.05 and r > 0.5 were considered significant. Multi-variate analysis was used to determine the independent risk factors associated with mortality using forward stepwise logistic regression. All variables with P < 0.1 on univariate analysis were entered into the multi-variate model. Statistical analyses were performed using R Version 2.14.1 (http://www.r-project.org/) for Microsoft Windows 7.
A retrospective observational study was performed between 1 April 2000 and 31 March 2010. The setting was Tokyo Women’s Medical University (TWMU) Hospital in Japan, a 1423-bedded university hospital that includes the nation’s largest institute of cardiovascular disease. All cases of nosocomial CPC were included in the study. The study protocol was approved by the Ethics Committee at TWMU Hospital.
Results
Definitions
Clinical features
A case of candidaemia was defined as a patient with at least one positive blood culture yielding any Candida spp. Cases with positive blood cultures occurring 48 h after admission were defined as nosocomial candidaemia. Blood samples were drawn under sterile conditions and processed using BACTEC 9240 (Becton Dickinson Diagnostic Instrument Systems, Towson, MD, USA). C. parapsilosis was identified using the API 32C system (bioMe ´rieux, Marcy l’Etoile, France).
In total, 307 cases of nosocomial candidaemia were identified. The incidence of nosocomial candidaemia was 0.074 cases per 1000 patient-admission-days during the study period. C. parapsilosis (N ¼ 89, 29%) was the second most common Candida spp. causing candidaemia after Candida albicans (43%). Fifty-one of the 89 cases of CPC with medical records met the inclusion criteria and were included in this study. The demographic and clinical characteristics of the 51 cases (39 survivors and 12 deaths) of CPC are shown in Table I. Eighteen of these 51 (35.3%) patients had digestive tract or hepatobiliary-pancreatic cancer. Seven cases had RCPM, consisting of prosthetic valves (N ¼ 5) and grafts (N ¼ 2). No significant differences were found between the survivors and the patients who died in terms of receipt of empirical or definitive antifungal treatment, serum C-reactive protein, BDG and white blood cell count. The crude 30-day mortality rate was 23.5% (12/51). APACHE II score >25 [odds ratio (OR) 43.9, 95% confidence interval (CI) 4.77e404; P ¼ 0.000839] and RCPM (OR 14.6, 95% CI 1.05e203; P ¼ 0.0459) were risk factors associated with death within 30 days on multi-variate analysis (Table I). The most severe complications identified in this study were endophthalmitis (N ¼ 3, 5.9%) and endocarditis (N ¼ 3, 5.9%). Two cases developed endophthalmitis during micafungin treatment and one case developed endophthalmitis during fluconazole treatment. Likewise, two cases developed endocarditis during micafungin treatment and one case developed endocarditis during fluconazole treatment.
Study population Patients were observed from the day of diagnosis of CPC until the end of follow-up at 30 days. The following data were obtained for all patients: age; sex; underlying disease; use of intravenous catheters; retained cardiovascular prosthetic materials (RCPM) (prosthetic valve or graft); empirical or definitive antifungal treatment; dose of antifungal agents; complications of candidaemia; Acute Physiology and Chronic Health Evaluation (APACHE) II scores at diagnosis of candidaemia; and crude mortality at 30 days. Hepato-biliary-pancreatic cancer was defined as including hepatocellular carcinoma, pancreatic cancer or bile duct cancer. Blood tests including white blood cell and neutrophil counts, serum C-reactive protein and 1,3-beta-D-glucan (BDG) assay using Fungitec G-test MK (Nissui, Tokyo, Japan; cut-off level 20 pg/mL) were performed within three days of a positive blood culture. Exclusion criteria were: age <16 years; post-haematopoietic stem cell transplantation; end-stage cancer in palliative therapy; human immunodeficiency virus positive; antifungal treatment not initiated; and Candida spp. not stored at 80 C.
Analysis of antifungal susceptibility
Until 2012, micafungin was the only approved echinocandin, and caspofungin was not available during the study period in Japan. The MIC values of micafungin, amphotericin-B, fluconazole, voriconazole and itraconazole against C. parapsilosis isolated from blood culture were measured retrospectively in accordance with the standards of the Clinical and Laboratory Standards Institute (CLSI) M27-A3 using an Eiken DP kit (Eiken, Tokyo, Japan). The MFC was defined as the lowest drug concentration that resulted in less than one colony.
Antifungal MIC and MFC values and Spearman’s rank correlation coefficients between MIC and MFC values are shown in Table II. No reduced susceptibility to antifungal agents (micafungin, fluconazole, itraconazole, voriconazole or amphotericin-B) or resistant strains was found according to breakpoints determined by CLSI and EUCAST. Rank correlations between antifungal MIC and MFC values against C. parapsilosis were observed for amphotericin-B (r ¼ 0.6789, P < 0.0001) and micafungin (r ¼ 0.5449, P < 0.0001) by Spearman’s rank correlation coefficient. In contrast, no correlation between MIC and MFC values was observed for fluconazole (r ¼ 0.2702, P ¼ 0.0504).
Statistical analysis
Discussion
Continuous data were compared using Student’s t-test, and categorical data were compared using Fisher’s exact test.
To the best of the authors’ knowledge, this is the first paper in the English language to report the risk factors associated
Antifungal susceptibility
Table I Demographic and clinical characteristics of 51 cases of Candida parapsilosis candidaemia Characteristics
Overall (N ¼ 51) N (%) or mean SD
Gender (male) Age
Survivors (N ¼ 39) N (%) or mean SD
Deaths (N ¼ 12) N (%) or mean SD
P-value
37 (72.5%) 62.3 15.7
27 (69.2%) 61.7 16.1
10 (83.3%) 64.2 14.9
0.471 0.842
18 (35.3%)
14 (35.9%)
4 (33.3%)
0.134
10 (19.6%)
8 (20.5%)
2 (16.6%)
1
6 (11.7%)
5 (12.8%)
1 (8.3%)
1
5 (9.8%) 7 (13.7%)
4 (10.2%) 3 (7.69%)
1 (8.3%) 4 (58.3%)
1 0.0455
11 (21.5%) 3 (0.6%)
7 (17.9%) 1 (2.56%)
4 (33.3%) 2 (16.6%)
0.262 0.134
Intravenous catheter CLABSI Total parenteral nutrition Removal of intravenous catheter
43 (84.3%) 41 (80.3%) 30 (58.8%)
31 (79.4%) 29 (74.3%) 26 (66.7%)
12 (100%) 12 (100%) 4 (33.3%)
0.173 0.092 0.0445
APACHE II score APACHE II score >25
22.39 5.62 16 (31.4%)
20.74 4.452 6 (15.4%)
27.75 5.879 10 (83.3%)
0.000056 0.000031
30 (58.8%) 371.4 112.5
23 (58.9%) 380 61.57
7 (58.3%) 366.66 81.64
1 0.2273
19 (37.2%) 100 18.89 2 (3.9%)
15 (38.4%) 103.3 22.88 1 (2.56%)
4 (33.3%) 100 1 (8.3%)
1 0.8029 0.96
Definitive antifungal treatment Fluconazole Micafungin Voriconazole Polyene (AMB or L-AMB)
30 (58.8%) 15 (29.1%) 2 (3.9%) 4 (7.8%)
24 (61.5%) 11 (28.2%) 2 (5.12%) 3 (7.69%)
6 4 0 1
0.518 0.7301 1 1
Treatment duration (days)
19.21 14.5
21 15.1 (7e84)
12.9 12.9 (2e29)
0.0863
0.125e0.5 0.5e128 0.25e1 1e8
0.267 0.49 0.383 0.546
8940 5.019 9.53 4.66 372.7 465.6 (N ¼ 12)
0.6496 0.1031 0.1104
Underlying diseases Digestive/hepato-biliarypancreatic cancer Non-digestive/hepatobiliary-pancreatic cancer IBD (ulcerative colitis and Crohn’s disease) Pancreatitis Retained cardiovascular prosthetic materials Diabetes mellitus CKD stage 5 or 5D
Empirical antifungal treatment Fluconazole Maintenance dose of fluconazole (mg/day) Micafungin Dose of micafungin (mg/day) Polyene (AMB or L-AMB)
Antifungal susceptibility against C. parapsilosis (mg/mL) Fluconazole MIC 0.125e1 0.125e1 Fluconazole MFC 0.5e128 0.5e128 Micafungin MIC 0.125e2 0.125e2 Micafungin MFC 0.5e8 0.5e8 Laboratory data White blood cell count (/mL) C-reactive protein (mg/dL) 1,3-beta-D-glucan (pg/mL) Complication Endophthalmitis due to C. parapsilosis Infective endocarditis due to C. parapsilosis Valve replacement during candidaemia
8352 5.060 6.96 6.24 780.1 1048.4
8171 5.124 6.17 6.50 948.6 1176.6 (N ¼ 29)
(50%) (33.3%) (0%) (8.3%)
3 (5.9%)
2 (5.12%)
1 (8.3%)
0.561
3 (5.9%)
1 (2.56%)
2 (16.6%)
0.134
1 (1.96%)
1 (2.56%)
0
1
Multi-variate analysis OR (95% CI), P-value
14.6 (1.05e203), 0.0459
43.9 (4.77e404), 0.000839
IBD, inflammatory bowel disease; CKD, chronic kidney disease; CLABSI, central-line-associated bloodstream infection; APACHE, Acute Physiology and Chronic Health Evaluation; AMB, amphotericin-B; L-AMB, liposomal amphotericin-B; MIC, minimum inhibitory concentration; MFC, minimum fungicidal concentration; SD, standard deviation; OR, odds ratio; CI, confidence interval.
0.1697
13 3 11
6
1
2
1
4
5
1
5
5
3
2
3
5
2 8
7 4 2 10
3
4
3 9
6
9 4 17 22
14
7
1 6 2
20 3 5
14
10
3
6
2
3 4 1 18 10 24
26 2 3 8 25 5 12 2
39 1 49 5 VRCZ
ITCZ
FLCZ
AMB
MIC MFCa MIC MFCa MIC MFCa MIC MFCa MIC MFCa MCFG
MCFG, micafungin; AMB, amphotericin-B; FLCZ, fluconazole; ITCZ, itraconazole; VRCZ, voriconazole; MIC, minimum inhibitory concentration; MFC, minimum fungicidal concentration; r, Spearman’s rank correlation coefficient (1 & r & 1); shaded zone, MIC breakpoint of resistance according to EUCAST (AMB) or from susceptible dose dependent to resistant according to Clinical and Laboratory Standards Institute (MCFG, FLCZ, ITCZ, VRCZ). a MFC breakpoint not established.
0.2245
0.0078
0.2702
0.3619
0.0504
<0.0001 0.6789
<0.0001 0.5449
1 4 0.25 2 0.5 64< 0.03 8< &0.015 8< 0.125e2 0.5e8 0.03e0.5 0.25e8 &0.12e1 0.5e<64 &0.015e0.12 &0.015e16 &0.015e0.03 &0.015e16
P-value
Spearman’s rank correlation coefficient
r MIC90 Range 64< 64 32 16 8 4 2 1 0.5 0.25 0.12 0.06 0.03 0.02
mg/mL
Table II Distribution of antifungal minimum inhibitory concentration (MIC) and minimum fungicidal concentration (MFC)a values, and Spearman’s rank correlation coefficients between MIC and MFC values against 51 strains of Candida parapsilosis isolated from blood culture
Y. Hirai et al. / Journal of Hospital Infection 87 (2014) 54e58
57
with mortality attributed to nosocomial CPC in a tertiary care setting and the importance of RCPM. Although the incidence of CPC is increasing,2 it is still controversial to link this incidence with increased use of echinocandin.3 Intravenous catheter removal improved the mortality rate in this study in cases with CPC (10% vs 31%). In contrast, a similar mortality rate was found with and without catheter removal in neonates with C. albicans candidaemia (35% vs 48%).4 Similar to intravenous catheters, RCPM was a risk factor associated with death in the present study. C. parapsilosis may produce biofilms, and different antifungal resistance in biofilms has been reported. Biofilm formation on cardiovascular prosthetic materials may be more extensive than that on intravascular catheters. An emphasis on prompt removal of prosthetic materials may lead to a more favourable prognosis in cases of CPC; this was recommended in the 2012 guidelines of the European Society of Clinical Microbiology and Infectious Diseases for cases of candida endocarditis (evidence level A-III). The 2009 treatment guidelines of the Infectious Diseases Society of America recommend that cases of invasive candidiasis should receive definitive treatment with fluconazole if C. parapsilosis is identified. However, a meta-analysis found a similar clinical outcome for invasive C. parapsilosis infection treated with echinocandins and other antifungal agents.5 The present results showed significant correlation between MIC and MFC values for micafungin and amphotericin-B, which are categorized as fungicidal agents. In contrast, the MFC values of azoles (fluconazole, itraconazole and voroconazole) were dissociated from the MIC values. The relationship between MIC, MFC and clinical outcome has not been established to date. The present results suggest that antifungal agents, given as empirical or definitive therapy, were not risk factors for mortality. The routine testing of antifungal MIC values against C. parapsilosis is still controversial. The overall mortality rate for C. parapsilosis candidaemia has been reported to range from 4% to 49%.6 Similar to the present results, patients with non-intensive-care-unitacquired candidaemia and APACHE II scores >25 on admission had higher long-term mortality, which reflects the importance of underlying comorbidity for survival.7 Serum BDG is known to support a diagnosis of fungal infection, but iatrogenic false-positive results can occur. BDG tended to be higher among the survivors in this study. It was assumed that elevated levels of BDG resulted from the burden of fragmented fungi destroyed by host immunity and/or antifungal agents. Elevated BDG was not a risk factor associated with mortality in this study. C. parapsilosis is the fifth most common fungal species causing ocular infection. The incidence of endophthalmitis caused by C. parapsilosis (5.0%) has been reported to be lower than that of C. albicans (60%).8 Fundoscopy was not performed for all patients with candidaemia at TWMU Hospital; therefore, the incidence may have been underestimated. Fungal infective endocarditis (IE) accounts for 1.3e6% of all cases of IE. The incidence of IE due to C. parapsilosis has been reported to be associated with 17% of cases of candida IE.9 According to the modified Duke criteria, the incidence of IE due to C. parapsilosis may be underestimated. Similar to RCPM, insufficient source control such as abscesses or intravascular catheters has been found to be independently associated with mortality in patients with septic shock due to candida infection.10
58
Y. Hirai et al. / Journal of Hospital Infection 87 (2014) 54e58
This study had two limitations. First, the study was a retrospective analysis with a small sample size performed at a single centre. Second, antifungal MIC and MFC values were measured retrospectively. No antifungal-resistant strains were found. In conclusion, APACHE II score >25 and RCPM may strongly predict death within 30 days for patients with nosocomial CPC. Prompt surgical removal should be considered in CPC patients with RCPM.
Acknowledgements Results from this study were presented, in part, at the 5th Trends in Medical Mycology, Valencia, Spain, 4th October 2011 (Abstract 429). Conflict of interest statement None declared. Funding sources None.
References 1. Pfaller MA, Moet GJ, Messer SA, Jones RN, Castanheira M. Candida bloodstream infections: comparison of species distributions and antifungal resistance patterns in community-onset and nosocomial isolates in the SENTRY Antimicrobial Surveillance Program, 2008e2009. Antimicrob Agents Chemother 2011;55:561e566.
2. 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;48:1695e1703. 3. Hirai Y, Asahata S, Ainoda Y, et al. The defined daily dose (DDD) of micafungin usage didn’t affect incidence of Candida parapsilosis candidemia at tertiary hospital in Japan. ICAAC Meeting, San Francisco, CA, USA 9the12th September 2012. Poster M-326. 4. Benjamin Jr DK, Stoll BJ, Fanaroff AA, et al. Neonatal candidiasis among extremely low birth weight infants: risk factors, mortality rates, and neurodevelopmental outcomes at 18 to 22 months. Pediatrics 2006;117:84e92. 5. Kale-Pradhan PB, Morgan G, Wilhelm SM, Johnson LB. Comparative efficacy of echinocandins and non-echinocandins for the treatment of Candida parapsilosis infections: a meta-analysis. Pharmacotherapy 2010;30:1207e1213. 6. Asbeck V, Clemons KV, Stevens DA. Candida parapsilosis: a review of its epidemiology, pathogenesis, clinical aspects, typing and antimicrobial susceptibility. Crit Rev Microbiol 2009;35:283e309. 7. Ylipalosaari P, Ala-Kokko TI, Karhu J, et al. Comparison of the epidemiology, risk factors, outcome and degree of organ failures of patients with candidemia acquired before or during ICU treatment. Crit Care 2012;16:R62. 8. Oude Lashof AM, Rothova A, Sobel JD, et al. Ocular manifestations of candidemia. Clin Infect Dis 2011;53:262e268. 9. Smego Jr RA, Ahmad H. The role of fluconazole in the treatment of candida endocarditis: a meta-analysis. Medicine (Baltimore) 2011;90:237e249. 10. Kollef M, Micek S, Hampton N, Doherty JA, Kumar A. Septic shock attributed to candida infection: importance of empiric therapy and source control. Clin Infect Dis 2012;54:1739e1746.
Journal of Hospital Infection journal homepage: www.elsevierhealth.com/journals/jhin
Short report
Nosocomial Candida parapsilosis candidaemia: risk factors, antifungal susceptibility and outcome Y. Hirai a, b, *, S. Asahata a, Y. Ainoda a, A. Goto c, T. Fujita a, K. Totsuka a a
Department of Infectious Diseases, Tokyo Women’s Medical University Hospital, Shinjuku-ku, Tokyo, Japan Department of Haematology, Tokyo Women’s Medical University Hospital, Shinjuku-ku, Tokyo, Japan c Laboratory of Microbiology, Tokyo Women’s Medical University Hospital, Shinjuku-ku, Tokyo, Japan b
A R T I C L E
I N F O
Article history: Received 24 October 2013 Accepted 10 February 2014 Available online 12 March 2014 Keywords: Candida parapsilosis Candidaemia Minimum inhibitory concentration Minimum fungicidal concentration Cardiovascular prosthetic materials
S U M M A R Y
A retrospective analysis was undertaken from 2000 to 2010 to show the risk factors associated with death within 30 days in patients with C. parapsilosis candidaemia (CPC). Fifty-one cases of nosocomial CPC were included in the analysis. All isolates from blood cultures were susceptible to micafungin and fluconazole. The overall mortality rate was 23.5%, and the most severe complications were endocarditis (5.9%) and endophthalmitis (5.9%). On multi-variate analysis, APACHE II score >25 (odds ratio 43.9) and retained cardiovascular prosthetic materials (RCPM) (prosthetic valve or graft) (odds ratio 14.6) were found to be risk factors associated with death. Prompt surgical removal should be considered in CPC patients with RCPM. ª 2014 The Healthcare Infection Society. Published by Elsevier Ltd. All rights reserved.
Introduction Candida spp. remain a major cause of morbidity and mortality in healthcare settings. Candida parapsilosis is often the second or third most commonly isolated Candida spp. from blood cultures.1 Intravenous catheters and parenteral nutrition are well-known risk factors. C. parapsilosis demonstrates an * Corresponding author. Address: Department of Infectious Diseases, Tokyo Women’s Medical University Hospital, 8-1 Kawada-cho, Shinjuku-ku, Tokyo, Japan. Tel.: þ81 3 3353 8111. E-mail address: [email protected] (Y. Hirai).
increased in-vitro minimum inhibitory concentration (MIC) to echinocandins such as caspofungin.1 Recent studies have shown increasing incidence of C. parapsilosis candidaemia (CPC) of 7e24%.2 Although risk factors associated with death attributed to candidaemia have been documented in the literature, any specific risk factors associated with nosocomial CPC are not known. The primary outcome of this study was to establish the risk factors associated with death within 30 days by multivariate analysis. The secondary outcome was to determine antifungal MICs and minimum fungicidal concentrations (MFCs) for C. parapsilosis isolated from blood cultures.
0195-6701/$ e see front matter ª 2014 The Healthcare Infection Society. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jhin.2014.02.006
Y. Hirai et al. / Journal of Hospital Infection 87 (2014) 54e58
Patients and methods Study design
55
Spearman’s rank correlation coefficient was used to show correlation between antifungal MIC and MFC values. P < 0.05 and r > 0.5 were considered significant. Multi-variate analysis was used to determine the independent risk factors associated with mortality using forward stepwise logistic regression. All variables with P < 0.1 on univariate analysis were entered into the multi-variate model. Statistical analyses were performed using R Version 2.14.1 (http://www.r-project.org/) for Microsoft Windows 7.
A retrospective observational study was performed between 1 April 2000 and 31 March 2010. The setting was Tokyo Women’s Medical University (TWMU) Hospital in Japan, a 1423-bedded university hospital that includes the nation’s largest institute of cardiovascular disease. All cases of nosocomial CPC were included in the study. The study protocol was approved by the Ethics Committee at TWMU Hospital.
Results
Definitions
Clinical features
A case of candidaemia was defined as a patient with at least one positive blood culture yielding any Candida spp. Cases with positive blood cultures occurring 48 h after admission were defined as nosocomial candidaemia. Blood samples were drawn under sterile conditions and processed using BACTEC 9240 (Becton Dickinson Diagnostic Instrument Systems, Towson, MD, USA). C. parapsilosis was identified using the API 32C system (bioMe ´rieux, Marcy l’Etoile, France).
In total, 307 cases of nosocomial candidaemia were identified. The incidence of nosocomial candidaemia was 0.074 cases per 1000 patient-admission-days during the study period. C. parapsilosis (N ¼ 89, 29%) was the second most common Candida spp. causing candidaemia after Candida albicans (43%). Fifty-one of the 89 cases of CPC with medical records met the inclusion criteria and were included in this study. The demographic and clinical characteristics of the 51 cases (39 survivors and 12 deaths) of CPC are shown in Table I. Eighteen of these 51 (35.3%) patients had digestive tract or hepatobiliary-pancreatic cancer. Seven cases had RCPM, consisting of prosthetic valves (N ¼ 5) and grafts (N ¼ 2). No significant differences were found between the survivors and the patients who died in terms of receipt of empirical or definitive antifungal treatment, serum C-reactive protein, BDG and white blood cell count. The crude 30-day mortality rate was 23.5% (12/51). APACHE II score >25 [odds ratio (OR) 43.9, 95% confidence interval (CI) 4.77e404; P ¼ 0.000839] and RCPM (OR 14.6, 95% CI 1.05e203; P ¼ 0.0459) were risk factors associated with death within 30 days on multi-variate analysis (Table I). The most severe complications identified in this study were endophthalmitis (N ¼ 3, 5.9%) and endocarditis (N ¼ 3, 5.9%). Two cases developed endophthalmitis during micafungin treatment and one case developed endophthalmitis during fluconazole treatment. Likewise, two cases developed endocarditis during micafungin treatment and one case developed endocarditis during fluconazole treatment.
Study population Patients were observed from the day of diagnosis of CPC until the end of follow-up at 30 days. The following data were obtained for all patients: age; sex; underlying disease; use of intravenous catheters; retained cardiovascular prosthetic materials (RCPM) (prosthetic valve or graft); empirical or definitive antifungal treatment; dose of antifungal agents; complications of candidaemia; Acute Physiology and Chronic Health Evaluation (APACHE) II scores at diagnosis of candidaemia; and crude mortality at 30 days. Hepato-biliary-pancreatic cancer was defined as including hepatocellular carcinoma, pancreatic cancer or bile duct cancer. Blood tests including white blood cell and neutrophil counts, serum C-reactive protein and 1,3-beta-D-glucan (BDG) assay using Fungitec G-test MK (Nissui, Tokyo, Japan; cut-off level 20 pg/mL) were performed within three days of a positive blood culture. Exclusion criteria were: age <16 years; post-haematopoietic stem cell transplantation; end-stage cancer in palliative therapy; human immunodeficiency virus positive; antifungal treatment not initiated; and Candida spp. not stored at 80 C.
Analysis of antifungal susceptibility
Until 2012, micafungin was the only approved echinocandin, and caspofungin was not available during the study period in Japan. The MIC values of micafungin, amphotericin-B, fluconazole, voriconazole and itraconazole against C. parapsilosis isolated from blood culture were measured retrospectively in accordance with the standards of the Clinical and Laboratory Standards Institute (CLSI) M27-A3 using an Eiken DP kit (Eiken, Tokyo, Japan). The MFC was defined as the lowest drug concentration that resulted in less than one colony.
Antifungal MIC and MFC values and Spearman’s rank correlation coefficients between MIC and MFC values are shown in Table II. No reduced susceptibility to antifungal agents (micafungin, fluconazole, itraconazole, voriconazole or amphotericin-B) or resistant strains was found according to breakpoints determined by CLSI and EUCAST. Rank correlations between antifungal MIC and MFC values against C. parapsilosis were observed for amphotericin-B (r ¼ 0.6789, P < 0.0001) and micafungin (r ¼ 0.5449, P < 0.0001) by Spearman’s rank correlation coefficient. In contrast, no correlation between MIC and MFC values was observed for fluconazole (r ¼ 0.2702, P ¼ 0.0504).
Statistical analysis
Discussion
Continuous data were compared using Student’s t-test, and categorical data were compared using Fisher’s exact test.
To the best of the authors’ knowledge, this is the first paper in the English language to report the risk factors associated
Antifungal susceptibility
Table I Demographic and clinical characteristics of 51 cases of Candida parapsilosis candidaemia Characteristics
Overall (N ¼ 51) N (%) or mean SD
Gender (male) Age
Survivors (N ¼ 39) N (%) or mean SD
Deaths (N ¼ 12) N (%) or mean SD
P-value
37 (72.5%) 62.3 15.7
27 (69.2%) 61.7 16.1
10 (83.3%) 64.2 14.9
0.471 0.842
18 (35.3%)
14 (35.9%)
4 (33.3%)
0.134
10 (19.6%)
8 (20.5%)
2 (16.6%)
1
6 (11.7%)
5 (12.8%)
1 (8.3%)
1
5 (9.8%) 7 (13.7%)
4 (10.2%) 3 (7.69%)
1 (8.3%) 4 (58.3%)
1 0.0455
11 (21.5%) 3 (0.6%)
7 (17.9%) 1 (2.56%)
4 (33.3%) 2 (16.6%)
0.262 0.134
Intravenous catheter CLABSI Total parenteral nutrition Removal of intravenous catheter
43 (84.3%) 41 (80.3%) 30 (58.8%)
31 (79.4%) 29 (74.3%) 26 (66.7%)
12 (100%) 12 (100%) 4 (33.3%)
0.173 0.092 0.0445
APACHE II score APACHE II score >25
22.39 5.62 16 (31.4%)
20.74 4.452 6 (15.4%)
27.75 5.879 10 (83.3%)
0.000056 0.000031
30 (58.8%) 371.4 112.5
23 (58.9%) 380 61.57
7 (58.3%) 366.66 81.64
1 0.2273
19 (37.2%) 100 18.89 2 (3.9%)
15 (38.4%) 103.3 22.88 1 (2.56%)
4 (33.3%) 100 1 (8.3%)
1 0.8029 0.96
Definitive antifungal treatment Fluconazole Micafungin Voriconazole Polyene (AMB or L-AMB)
30 (58.8%) 15 (29.1%) 2 (3.9%) 4 (7.8%)
24 (61.5%) 11 (28.2%) 2 (5.12%) 3 (7.69%)
6 4 0 1
0.518 0.7301 1 1
Treatment duration (days)
19.21 14.5
21 15.1 (7e84)
12.9 12.9 (2e29)
0.0863
0.125e0.5 0.5e128 0.25e1 1e8
0.267 0.49 0.383 0.546
8940 5.019 9.53 4.66 372.7 465.6 (N ¼ 12)
0.6496 0.1031 0.1104
Underlying diseases Digestive/hepato-biliarypancreatic cancer Non-digestive/hepatobiliary-pancreatic cancer IBD (ulcerative colitis and Crohn’s disease) Pancreatitis Retained cardiovascular prosthetic materials Diabetes mellitus CKD stage 5 or 5D
Empirical antifungal treatment Fluconazole Maintenance dose of fluconazole (mg/day) Micafungin Dose of micafungin (mg/day) Polyene (AMB or L-AMB)
Antifungal susceptibility against C. parapsilosis (mg/mL) Fluconazole MIC 0.125e1 0.125e1 Fluconazole MFC 0.5e128 0.5e128 Micafungin MIC 0.125e2 0.125e2 Micafungin MFC 0.5e8 0.5e8 Laboratory data White blood cell count (/mL) C-reactive protein (mg/dL) 1,3-beta-D-glucan (pg/mL) Complication Endophthalmitis due to C. parapsilosis Infective endocarditis due to C. parapsilosis Valve replacement during candidaemia
8352 5.060 6.96 6.24 780.1 1048.4
8171 5.124 6.17 6.50 948.6 1176.6 (N ¼ 29)
(50%) (33.3%) (0%) (8.3%)
3 (5.9%)
2 (5.12%)
1 (8.3%)
0.561
3 (5.9%)
1 (2.56%)
2 (16.6%)
0.134
1 (1.96%)
1 (2.56%)
0
1
Multi-variate analysis OR (95% CI), P-value
14.6 (1.05e203), 0.0459
43.9 (4.77e404), 0.000839
IBD, inflammatory bowel disease; CKD, chronic kidney disease; CLABSI, central-line-associated bloodstream infection; APACHE, Acute Physiology and Chronic Health Evaluation; AMB, amphotericin-B; L-AMB, liposomal amphotericin-B; MIC, minimum inhibitory concentration; MFC, minimum fungicidal concentration; SD, standard deviation; OR, odds ratio; CI, confidence interval.
0.1697
13 3 11
6
1
2
1
4
5
1
5
5
3
2
3
5
2 8
7 4 2 10
3
4
3 9
6
9 4 17 22
14
7
1 6 2
20 3 5
14
10
3
6
2
3 4 1 18 10 24
26 2 3 8 25 5 12 2
39 1 49 5 VRCZ
ITCZ
FLCZ
AMB
MIC MFCa MIC MFCa MIC MFCa MIC MFCa MIC MFCa MCFG
MCFG, micafungin; AMB, amphotericin-B; FLCZ, fluconazole; ITCZ, itraconazole; VRCZ, voriconazole; MIC, minimum inhibitory concentration; MFC, minimum fungicidal concentration; r, Spearman’s rank correlation coefficient (1 & r & 1); shaded zone, MIC breakpoint of resistance according to EUCAST (AMB) or from susceptible dose dependent to resistant according to Clinical and Laboratory Standards Institute (MCFG, FLCZ, ITCZ, VRCZ). a MFC breakpoint not established.
0.2245
0.0078
0.2702
0.3619
0.0504
<0.0001 0.6789
<0.0001 0.5449
1 4 0.25 2 0.5 64< 0.03 8< &0.015 8< 0.125e2 0.5e8 0.03e0.5 0.25e8 &0.12e1 0.5e<64 &0.015e0.12 &0.015e16 &0.015e0.03 &0.015e16
P-value
Spearman’s rank correlation coefficient
r MIC90 Range 64< 64 32 16 8 4 2 1 0.5 0.25 0.12 0.06 0.03 0.02
mg/mL
Table II Distribution of antifungal minimum inhibitory concentration (MIC) and minimum fungicidal concentration (MFC)a values, and Spearman’s rank correlation coefficients between MIC and MFC values against 51 strains of Candida parapsilosis isolated from blood culture
Y. Hirai et al. / Journal of Hospital Infection 87 (2014) 54e58
57
with mortality attributed to nosocomial CPC in a tertiary care setting and the importance of RCPM. Although the incidence of CPC is increasing,2 it is still controversial to link this incidence with increased use of echinocandin.3 Intravenous catheter removal improved the mortality rate in this study in cases with CPC (10% vs 31%). In contrast, a similar mortality rate was found with and without catheter removal in neonates with C. albicans candidaemia (35% vs 48%).4 Similar to intravenous catheters, RCPM was a risk factor associated with death in the present study. C. parapsilosis may produce biofilms, and different antifungal resistance in biofilms has been reported. Biofilm formation on cardiovascular prosthetic materials may be more extensive than that on intravascular catheters. An emphasis on prompt removal of prosthetic materials may lead to a more favourable prognosis in cases of CPC; this was recommended in the 2012 guidelines of the European Society of Clinical Microbiology and Infectious Diseases for cases of candida endocarditis (evidence level A-III). The 2009 treatment guidelines of the Infectious Diseases Society of America recommend that cases of invasive candidiasis should receive definitive treatment with fluconazole if C. parapsilosis is identified. However, a meta-analysis found a similar clinical outcome for invasive C. parapsilosis infection treated with echinocandins and other antifungal agents.5 The present results showed significant correlation between MIC and MFC values for micafungin and amphotericin-B, which are categorized as fungicidal agents. In contrast, the MFC values of azoles (fluconazole, itraconazole and voroconazole) were dissociated from the MIC values. The relationship between MIC, MFC and clinical outcome has not been established to date. The present results suggest that antifungal agents, given as empirical or definitive therapy, were not risk factors for mortality. The routine testing of antifungal MIC values against C. parapsilosis is still controversial. The overall mortality rate for C. parapsilosis candidaemia has been reported to range from 4% to 49%.6 Similar to the present results, patients with non-intensive-care-unitacquired candidaemia and APACHE II scores >25 on admission had higher long-term mortality, which reflects the importance of underlying comorbidity for survival.7 Serum BDG is known to support a diagnosis of fungal infection, but iatrogenic false-positive results can occur. BDG tended to be higher among the survivors in this study. It was assumed that elevated levels of BDG resulted from the burden of fragmented fungi destroyed by host immunity and/or antifungal agents. Elevated BDG was not a risk factor associated with mortality in this study. C. parapsilosis is the fifth most common fungal species causing ocular infection. The incidence of endophthalmitis caused by C. parapsilosis (5.0%) has been reported to be lower than that of C. albicans (60%).8 Fundoscopy was not performed for all patients with candidaemia at TWMU Hospital; therefore, the incidence may have been underestimated. Fungal infective endocarditis (IE) accounts for 1.3e6% of all cases of IE. The incidence of IE due to C. parapsilosis has been reported to be associated with 17% of cases of candida IE.9 According to the modified Duke criteria, the incidence of IE due to C. parapsilosis may be underestimated. Similar to RCPM, insufficient source control such as abscesses or intravascular catheters has been found to be independently associated with mortality in patients with septic shock due to candida infection.10
58
Y. Hirai et al. / Journal of Hospital Infection 87 (2014) 54e58
This study had two limitations. First, the study was a retrospective analysis with a small sample size performed at a single centre. Second, antifungal MIC and MFC values were measured retrospectively. No antifungal-resistant strains were found. In conclusion, APACHE II score >25 and RCPM may strongly predict death within 30 days for patients with nosocomial CPC. Prompt surgical removal should be considered in CPC patients with RCPM.
Acknowledgements Results from this study were presented, in part, at the 5th Trends in Medical Mycology, Valencia, Spain, 4th October 2011 (Abstract 429). Conflict of interest statement None declared. Funding sources None.
References 1. Pfaller MA, Moet GJ, Messer SA, Jones RN, Castanheira M. Candida bloodstream infections: comparison of species distributions and antifungal resistance patterns in community-onset and nosocomial isolates in the SENTRY Antimicrobial Surveillance Program, 2008e2009. Antimicrob Agents Chemother 2011;55:561e566.
2. 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;48:1695e1703. 3. Hirai Y, Asahata S, Ainoda Y, et al. The defined daily dose (DDD) of micafungin usage didn’t affect incidence of Candida parapsilosis candidemia at tertiary hospital in Japan. ICAAC Meeting, San Francisco, CA, USA 9the12th September 2012. Poster M-326. 4. Benjamin Jr DK, Stoll BJ, Fanaroff AA, et al. Neonatal candidiasis among extremely low birth weight infants: risk factors, mortality rates, and neurodevelopmental outcomes at 18 to 22 months. Pediatrics 2006;117:84e92. 5. Kale-Pradhan PB, Morgan G, Wilhelm SM, Johnson LB. Comparative efficacy of echinocandins and non-echinocandins for the treatment of Candida parapsilosis infections: a meta-analysis. Pharmacotherapy 2010;30:1207e1213. 6. Asbeck V, Clemons KV, Stevens DA. Candida parapsilosis: a review of its epidemiology, pathogenesis, clinical aspects, typing and antimicrobial susceptibility. Crit Rev Microbiol 2009;35:283e309. 7. Ylipalosaari P, Ala-Kokko TI, Karhu J, et al. Comparison of the epidemiology, risk factors, outcome and degree of organ failures of patients with candidemia acquired before or during ICU treatment. Crit Care 2012;16:R62. 8. Oude Lashof AM, Rothova A, Sobel JD, et al. Ocular manifestations of candidemia. Clin Infect Dis 2011;53:262e268. 9. Smego Jr RA, Ahmad H. The role of fluconazole in the treatment of candida endocarditis: a meta-analysis. Medicine (Baltimore) 2011;90:237e249. 10. Kollef M, Micek S, Hampton N, Doherty JA, Kumar A. Septic shock attributed to candida infection: importance of empiric therapy and source control. Clin Infect Dis 2012;54:1739e1746.