Use and outcome of empiric echinocandins in critically ill patients

Echinocandins are recommended as a first‐line empiric treatment for fungal infections of patients in an intensive care unit (ICU) with critical illness. The primary aim of the study was to compare outcomes among ICU patients treated with empiric anidulafungin (ANI), caspofungin (CASPO), or micafungin (MICA).


| INTRODUC TI ON
Fungal infections are an increasing problem in intensive care, and contribute significantly to morbidity and mortality as well as costs. [1][2][3][4] Prophylaxis with fluconazole in high-risk, critically ill patients has reduced the incidence of invasive fungal infections and mortality. 5 Blood culture is the "gold standard" for the diagnosis of invasive candidiasis (IC). However, the overall sensitivity of blood culture is only 50% and there may be an identification time lag of up to 5 days. 6,7 The assay time of many blood culture independent methods requiring DNA extraction is only about 2 hours. 8 However, many of these assays have thus far not been validated for diagnosing IC in multi-center studies.
So, the empiric antifungal therapy belongs to everyday practice in the ICU because timely and appropriate empiric antifungal therapy improves survival among patients with septic shock due to candidemia. 9 According to recent guidelines, echinocandins (caspofungin, micafungin, and anidulafungin) are recommended for first-line empiric treatment for critically ill patients in the ICU. 10 They have less severe side effects than amphotericin B, the gold standard of antifungal agents. [11][12][13] The echinocandins are semisynthetic lipopeptides which inhibit the synthesis of the 1,3-beta-D-glucan component of the fungal cell wall. 14 They have rather similar pharmacological and efficacy profiles, as well as few drug-drug interactions. 15 Only a limited number of studies compare the clinical efficacy of echinocandins in patients with critical illness. Anidulafungin has been shown to be non-inferior to fluconazole in the treatment of invasive candidiasis. 16 Recent studies have not shown the superiority of caspofungin or micafungin over fluconazole prophylaxis for highrisk patients in the ICU. [17][18][19][20] Because empirical antifungal therapy with echinocandins is commonly used in the ICU, real-life observations of possible differences between three available echinocandins are warranted. The primary aim of this retrospective study was to compare outcomes among ICU patients treated with empiric anidulafungin (ANI), caspofungin (CASPO), or micafungin (MICA) in an adult mixed ICU.

| Ethics
This non-interventional study using routine clinical data was approved by the Department of Operative Care of the Oulu University Hospital and the Data Protection Ombudsman. The requirement for written informed consent was waived due to the retrospective nature and lack of intervention. Our database guaranteed patient confidentiality.

| Study setting
This was a retrospective single-center observational cohort study performed at the Oulu University Hospital, Finland, which is an academic tertiary-level unit with a 26-bed mixed adult closed intensivist-led ICU. In our multidisciplinary team, intensivists do daily rounds together with an infectious disease specialist, gastroenterology surgeon, pharmacist, cardiologist, and radiologist.
All the patients receiving systemic echinocandins ANI, CASPO, or   MICA during their intensive care unit stay between 2009 and 2014 were included in the study. Patients were allocated to echinocandin groups (ANI, CASPO, and MICA) according to the first echinocandin used.

| Data extraction
Data were extracted from our ICU electronic patient data management system (Clinisoft), the electronic hospital information system, and hospital pharmacy databases. The following data were ob- The total expense of antifungal treatment was calculated by multiplying the actual daily cost of each echinocandin and fluconazole by the days it was used for each patient during their ICU stay. The cost of a daily dose covered the acquisition price, dose of the echinocandin, and was calculated with the aid of the hospital pharmacy.
During the study period, the cheapest echinocandin was used, with the exception that CASPO is a preferred echinocandin among hematological patients in our hospital.

| Statistical analysis
The data were analyzed with descriptive statistics. Summary statistics for continuous or ordinal variables are expressed as medians with 25th and 75th percentiles or 95%CI as appropriate, and the analysis between groups was done by the Kruskal-Wallis test.
Categorical variables were analyzed using Pearson's Chi-square test. Two-tailed P-values less than 0.05 were considered statistically significant. Analyses were performed using the SPSS soft-
The median age of the patients was 60.7 years (Table 1) MICA patients had the lowest SAPS II scores on admission (Table 1). This was the only statistical difference (P = .037), while other severity scores or organ dysfunctions between the echinocandin groups did not differ statistically (Table 1). CASPO patients had the lowest median platelet and leucocyte levels on admission. Either leukemia or lymphoma was detected in 35.7% of patients. MICA patients had the highest frequency of chronic comorbidities (62%) and post-operative admissions (30%).
The foci of infections in the three groups are presented in Table 2. The spectrum of different foci was rather similar in the three groups with three exceptions: Twenty-six patients (7.1%) had a blood culture-positive infection; 16 with Candida species and 10 with bacterial species. Blood culture-positive infections were observed only in the CASPO group (13.1%) and MICA group (7.0%), while none was observed in the ANI group (Table 2). Intra-abdominal infections were observed most often in the MICA group and urinary tract infections in the CASPO group. The focus of the infection was unknown in only 12 cases (3.2%). Renal replacement therapy (RRT) was needed in 25% of the patients and it varied from 24% to 27% in the three echinocandin groups ( Table 4). The duration of RRT was longest in the CASPO group (Table 4). Noradrenalin treatment was least often needed in the CASPO group (Table 4). The resource utilization measured by TISS scores did not differ between the groups.
The total costs of all antifungal treatments during ICU stays did not differ between the three echinocandin groups ( Table 5). The same holds true for fluconazole and echinocandin days. ICU and hospital length of stay or hospital, and 30-day and 1-year mortality did not differ between echinocandin groups (

| D ISCUSS I ON
To the best of our knowledge, this retrospective non-interventional study is the first and one of the largest of its kind comparing real-life ICU data on the empiric use of three echinocandins (ANI, CASPO and MICA). We showed that ICU and hospital stay or mortality (hospital, 30-day, and 1-year) and the costs of empiric antifungal therapy did not differ between the three echinocandin groups.
Echinocandins are recommended as a first-line empiric treatment for critically ill patients in the ICU according to IDSA guidelines. 10 Furthermore, according to European guidelines, empirical treatment of patients with septic shock and suspected candidiasis should be started as soon as possible. 23 We gave empiric echinocandin therapy to patients with surgical intra-abdominal infections with a poor response to fluconazole treatment or to patients with uncontrolled sepsis and long-term, broad-spectrum antibiotic therapy with fluconazole. We did not use Candida scores, which have been shown to be unreliable and costly predictors of candidemia in ICU populations. 24,25 This EMPIRICUS trial is one of the most interesting empiric studies among critically ill patients who acquired severe sepsis with Candida colonization. 25 In this multi-center, double-blinded study, the 28-day mortality in the empirical micafungin group was 30% and the placebo group 29.7%. However, in this study, neutropenic patients were excluded and only 5% of the patients had undergone abdominal surgery.
Our series also included these excluded patient groups. Thus, we consider that our real-life, although retrospective, patient TA B L E 1 Summary of patient characteristics, ICU scores, clinical data, and comorbidities according to echinocandin group  The number represents patients with any comorbidities. Patients may have several comorbidities. Abbreviations: adm, on admission; APACHE II adm, acute physiology and chronic health evaluation II scores on admission; ASO, Arteriosclerosis obliterans; COPD, chronic obstructive pulmonary disease; CRP, C-reactive protein; PTC, procalcitonin; SAPS II, simplified acute physiology score; SOFA adm, sequential organ failure assessment scores on admission; WBC, white blood cell. b Parenthesis, Results presented as medians with 25th and 75th percentiles.  population is rather suitable for the evaluation of empiric echinocandin treatment in an ICU setting. In our series, neither ICU nor hospital stay differed between echinocandin groups. Also, the resource utilization was similar in the three groups. The 30-day mortality, which varied from 27% to 32%, is in harmony with the literature. 25 In addition, 1-year mortality did not differ between the echinocandin groups and the figure of 43% is in concordance with the rate of 59% in earlier literature. [26][27][28] Furthermore, there were no significant differences in 30-day (26%, 32%, and 34%) mortalities in 112 patients with culture-proven Candida infections among our anidulafungin, caspofungin, and micafungin groups.

Foci of infections
According to a recent epidemiologic meta-analysis of candidaemia in Europe, pooled day 30 mortality rate in intensive care units was

37%. 29
In this empiric echinocandin series, blood culture yielded Candida The accurate dosing of echinocandins without known echinocandin concentrations is a challenge in the ICU due to large interindividual variability. 32 Moreover, the concentrations of caspofungin and micafungin were clearly lower than among healthy volunteers.
All three echinocandins require a higher than standard dose for obese or markedly obese patients. 15 with concentration measurements will be needed.

| CON CLUS ION
According to our retrospective non-interventional study of the empiric use of echinocandins, ICU and hospital stay, mortality rates (hospital, 30-day and 1-year), and resource utilization did not differ between patients receiving anidulafungin, caspofungin, or micafungin in a mixed adult ICU. Considering the findings of this research, it is practical to select the least expensive echinocandin for empiric treatment in critically ill patients.

ACK N OWLED G EM ENTS
The help of study nurse S. Sälkiö, RN, in retrieving the data from the electronic archives is highly appreciated.

CO N FLI C T S O F I NTE R E S T S
None declared.

AUTH O R S' CO NTR I B UTI O N S
TA-K, HS, and JK contributed substantially to the study design, acquisition, interpretation, and analysis of data and writing of the manuscript. PY, JJL, and SÄ made substantial contributions to analysis and interpretation of data and writing of the manuscript. SÄ calculated the expenses of echinocandin treatment. All authors read and approved the final manuscript.

Et hic s A p p rova l a n d C o ns e nt to P a r t icipate
The study protocol was approved by the Ethics Committee of Oulu University Hospital. Because the study was epidemiological without any interventions, the requirement for informed consent was waived.

CO N S ENT TO PU B LI S H
Not applicable.

DATA AVA I L A B I L I T Y S TAT E M E N T
All data generated or analyzed during this study are included in this published article. It can also be requested from the corresponding author.