Sage Journals: Discover world-class research

Abstract

Invasive candidiasis (IC) is a leading cause of morbidity and mortality among nonneutropenic ICU patients and these life-threatening nosocomial infections require early diagnosis and prompt treatment. However, none of the predictive tools are sufficiently accurate to identify high-risk patients and the potential interest of IC prophylactic, empirical and preemptive treatment in the nonneutropenic ICU population has not yet been demonstrated. In the case of nosocomial severe sepsis after necrotizing pancreatitis or upper digestive anastomotic leakage, early probabilistic antifungals are probably indicated. In the remaining ICU surgical and medical patients, prophylactic and empirical strategies are highly debated because they may promote antifungal selective pressure through an overuse of these molecules. In this context, non-culture-based methods such as mannan or β-D glucan or polymerase chain reaction tests are promising. However, none of these tests used alone in ICU patients is sufficiently accurate to avoid overuse of empirical/preemptive treatment. The interest of strategies associating predictive clinical scores and non-culture-based methods still needs to be demonstrated by well-conducted randomized, controlled trials. While awaiting these studies, we consider that probabilist treatment should be stopped earlier if IC is not proven.

Keywords

antifungal agents intensive care unit invasive candidiasis

Introduction

Despite the availability of new antifungal therapies and the publication of several guidelines, invasive candidiasis (IC) remains a matter of concern in the intensive care unit (ICU), as it is associated with a large increase in mortality, length of stay and costs. Thus, the diagnostic and treatment strategies are, especially in nonimmunocompromised ICU patients, constantly evolving and the risk–benefit balance of prophylaxis and empiric or preemptive treatment is highly debated. The aim of this narrative review is to provide to ICU practitioners an up-to-date basis for the management strategies of suspected or proven IC cases in nonneutropenic ICU patients. This manuscript will point out the last recommendations concerning diagnosis and targeted treatment of IC and will review the evidence supporting, or not, the use of antifungal prophylaxis and empiric or preemptive treatment in ICU.

Incidence and mortality rate of candidiasis are high

Candidaemia and other forms of IC represent a significant cause of morbidity and mortality, especially in the nosocomial setting. Candida species are responsible for 7–10% of nosocomial bloodstream infections [Vincent et al. 2009; Wisplinghoff et al. 2004; Tabah et al. 2012] and, in a 1-day, point-prevalence study involving more than 7000 patients from 75 countries, Candida was the third most common pathogen and accounted for 17% of infection episodes [Vincent et al. 2009]. A nationwide US surveillance study showed that candidaemia crude mortality was 29% for hospital ward patients and that Candida-related mortality can reach up to 50–70% in ICUs [Wisplinghoff et al. 2004; Bassetti et al. 2014; Lortholary et al. 2014]. Worryingly, although the first decade of 2000s has been marked by many innovations in the field of diagnosis and treatment of invasive fungal infections, the availability of new antifungals and the publication of numerous guidelines, these measures did not prevent an increase of incidence and mortality of candidaemia in ICUs. Indeed, through an active hospital-based surveillance program of incident episodes of candidaemia in 24 tertiary care hospitals between 2002 and 2010, Lortholary and colleagues recently showed an increase in the incidence of Candida bloodstream infections both in the overall population and in ICU overtime [Lortholary et al. 2014]. Worrisomely, the 30 day mortality and early death rate also increased between the first and the last year of study. Thus, management of IC, despite substantial improvement, seems to remain an unmet need.

Early treatment of candidaemia decreases mortality of ICU patients with septic shock

Delay in antimicrobial administration has been associated with a decrease in survival in patients with bloodstream infections, especially in the case of septic shock [Garrouste-Orgeas et al. 2006; Kumar et al. 2006]. Among septicaemia, candidaemia is currently treated later than bacteriaemia [Tabah et al. 2012] and, despite recent advances in microbiological techniques, early diagnosis of IC remains problematic and microbiological documentation occurs late in the course of infection. Several retrospective studies showed that time from first Candida-positive blood culture to initiation of antifungal therapy correlates with mortality increase [Garey et al. 2006; Morrell et al. 2005]. Furthermore, Kollef and colleagues recently included 224 patients with Candida infection-related septic shock and demonstrated that delayed antifungal treatment is a risk factor for 30-day mortality [Kollef et al. 2012]. Although uncontrolled, these studies suggest that initiating empiric treatment may be beneficial. However, as far as the identification of reliable triggers for antifungal treatment is still lacking, such a strategy contributes to a huge financial burden, is also responsible for the overuse of antifungals [Azoulay et al. 2012] and its impact on patient outcome is unclear.

Effective strategies to identify high-risk patients are currently lacking

Thus, predictive models of IC have been developed to stratify patients at high risk of developing IC. Major risk factors for Candida colonization include length of ICU stay, use of parenteral nutrition, broad-spectrum and long-term antibiotics, central lines, and abdominal surgery [Eggimann et al. 2003; Charles et al. 2005]. These markers are very frequent in the ICU and may lead to all ICU patients being treated. In 1994, Pittet and colleagues tried to evaluate the risk of progressing from colonization to infection in surgical patients and developed the Candida colonization index [Pittet et al. 1994]. Indeed, endogenous colonization is responsible for the vast majority of severe candidiasis [Marco et al. 1999; Nucci and Anaissie, 2001] and a continuum exists between Candida colonization and IC, although in contrast to bacterial infections, there is a delay of 7–10 days between exposure to colonization and other risk factors and the development of an IC. In the Pittet colonization index, a ratio of the number of distinct nonblood body sites colonized by the same strain of Candida spp. to the total number of body sites cultured > 0.5 was associated with an increased rate of IC. Several studies performed in surgical ICU patients highlighted the potential usefulness of the Candida colonization index and it became the most widely studied clinical tool for IC prediction [Pittet et al. 1994]. For example, Piarroux and colleagues used the colonization index to assess the intensity of Candida spp. colonization in 478 surgical ICU patients [Piarroux et al. 2004]. Patients with an index above the threshold received empirical antifungal therapy and the incidence of IC among these patients was significantly lower than in an historical cohort of 455 control subjects. However, the low predictive value (less than 9% in the EPCAN study [Leon et al. 2009b]), the very high proportion of treated patients (87% in the study of Piarroux and colleagues [Piarroux et al. 2004]) and the quantity of laboratory work required in this strategy prevents its use among nonimmunocompromised critically ill patients. Nevertheless, the colonization index remains an important way to characterize the dynamic of the colonization of ICU patients, which increases early in patients who will develop IC. Furthermore, its good negative predictive value makes it a useful tool for the identification of low-risk patients [Eggimann and Pittet, 2014].

Then, clinical predictive models combining clinical aspects with Candida colonization have been developed. Dupont and colleagues developed a score aiming at predicting the involvement of Candida in ICU patients with intra-abdominal sepsis, which includes female sex, upper gastrointestinal tract origin of peritonitis, perioperative cardiovascular failure and previous antimicrobial therapy, and exhibits good sensitivity and overall accuracy [Dupont et al. 2003]. Analysing data collected from the EPCAN project (1699 ICU patients), Leon and colleagues identified parenteral nutrition (odds ratio [OR] = 2.48), surgery (OR = 2.71), multifocal colonization (OR = 3.04) and severe sepsis (OR = 7.68) as predictors of IC and developed a ‘Candida score’. The usefulness of the Candida score was assessed in a prospective observational cohort of 1107 colonized patients staying at least 7 days in 36 mixed ICUs [Leon et al. 2009b]. Areas under the receiver operating characteristic (ROC) curve of the Candida score and colonization index were 0.774 and 0.633, respectively. Among colonized patients with a Candida score <3, the rate of IC was less than 5%, reflecting a very good 98% negative predictive value of a Candida score <3. However, the positive value of this predictive model was still very low (14%).

Non-culture-based methods to accelerate diagnosis of IC are promising

In order to improve the early diagnosis of IC and to complement the clinical predictive models in guiding empirical therapy, non-culture-based assays of surrogate markers (detection of antibodies of antigens related to fungal wall components, fungal-related nucleic acids) have been proposed. Before summarizing their characteristics it seems important to remind that they have mainly been evaluated in haematology and in surgical ICU [Mohr et al. 2011; Senn et al. 2008; Ellis et al. 2008] (1-3)-beta-d-glucan (BG) is a cell wall component of Candida spp. and other fungi. It becomes early positive and represents a sensitive but not specific biomarker (positive for IC, invasive aspergillosis and Pneumocystis jiroveccii pneumonia). The recommended positive cutoff value is 80 pg/ml but several causes of false positivity of this test have to be known: dialysis with cellulose membrane, bacteriaemia due to several Gram-positive microorganisms, albumin, antibiotics such as amoxicillin-clavulanic acid, intravenous immunoglobulins [Leon et al. 2014]. Karageorgopoulos and colleagues reported, in a meta-analysis of 11 studies a sensitivity of 57–97% and a specificity of 56–93% for the diagnosis of IC by BG [Karageorgopoulos et al. 2011]. The use of BG in complement of predictive models is now included in recommendations of several learned societies [Bassetti et al. 2013; Cornely et al. 2012; Pappas et al. 2009] and the BG Fungitell assay (Cape Cod, Inc., East Falmouth, MA) is FDA approved. Recently, the FUNGINOS study confirmed the diagnostic accuracy of BG in a particular setting [Tissot et al. 2013]. This study screened 434 patients with abdominal surgery or acute pancreatitis and IC stay 72 hours or longer and compared the effectiveness of BG, Candida score and colonization indexes for intra-abdominal candidiasis (IAC) diagnosis. With a positive predictive value of 72% and a negative predictive value of 80%, global performance of two consecutive measurements of BG greater than or equal to 80 pg/ml in predicting IAC was superior to that of Candida score and colonization index. Furthermore, BG positivity diagnosed IAC earlier (5 days) and correlated with severity of illness and outcome, allowing the authors to recommend the evaluation of BG-driven preemptive therapies. The diagnostic value of BG in prediction candidaemia in other, mainly medical, settings is less convincing. In ICU, a threshold of BG of 80 pg/ml is very sensitive (more that 90%) but unspecific (30–50%) and higher thresholds of 250 pg/ml or more have been proposed [Leon et al. 2014; Poissy et al. 2014].

Mannan is a polysaccharidic antigen from the Candida cell wall. It is specific but not sensitive for IC diagnosis and becomes positive later in the infection course than BG. Mannan antigen (Mn) and anti-mannan antibodies (Anti-Mn) can be measured using enzyme-linked immunosorbent assay (ELISA) techniques and the best results have been obtained with combined Mn/Anti-Mn tests. Mikulska and colleagues reported in a meta-analysis of 14 studies (including seven in non-neutropenic patients) evaluating this combined assay a sensitivity of 83% and a specificity of 86% [Mikulska et al. 2010]. Thus, although clinical usefulness of Mn/Anti-Mn combined test remains to be confirmed by prospective studies, its use is recommended in the last ESCMID guidelines [Cornely et al. 2012].

Finally, detection of Candida DNA using PCR seems to be very promising. A recent study prospectively included 63 ICU patients with suspected IC. The sensitivity, specificity, positive predictive value and negative predictive value for of PCR for the diagnosis of IC were 96.3%, 97.3%, 92.8% and 98.7%, respectively [Fortun et al. 2014]. However, the value of PCR as an early marker of IC remains to be confirmed as far as the sensitivity of this tool at the day of IC suspicion was lower (81%) and only 27 patients with confirmed IC were finally included in the study.

No evidence for effectiveness of empiric/preemptive treatment in ICU patient is currently available

Despite the development of such biomarkers and the multiplication of predictive scores, the identification of populations at high risk of IC that may possibly benefit from early (empiric/preemptive treatment) remains problematic. Thus, recent guidelines could not find the place of empirical antifungal in IC management and failed to provide high-level recommendations about this strategy [Pappas et al. 2009; Cornely et al. 2012]. Indeed, several placebo-controlled studies using clinical predictive models of IC and/or non-culture-based biomarkers failed to demonstrate the clinical usefulness of empirical or preemptive systemic antifungal therapy (SAT) and its impact on survival [Schuster et al. 2008; Ostrosky-Zeichner et al. 2014]. Furthermore, whether or not SAT may be efficient in colonized patients with unresolved sepsis and organ dysfunction remains unknown.

Schuster and colleagues conducted a randomized, controlled trial comparing high-dose fluconazole with placebo in 270 adult patients with fever despite administration of broad-spectrum antibiotics in 26 US ICUs showing no reduction of survival free of invasive fungal infection in the SAT group [Schuster et al. 2008].

Ostrosky-Zeichner and colleagues tested in a randomized, double-blind, placebo-controlled trial the use of caspofungin in 222 adults who stayed in the ICU for at least 3 days, required ventilation, received antibiotics, had a central line, and had one additional risk factor among the following: parenteral nutrition, dialysis, surgery, pancreatitis, systemic steroids, or other immunosuppressive agents. Interestingly, BG was monitored twice weekly [Ostrosky-Zeichner et al. 2014]. The primary endpoint was the incidence of proven or probable IC. Caspofungin treatment was safe and associated with a trend toward a reduction of the incidence of IC (9.8% in the caspofungin arm versus 16.7%, p = 0.14). When all patients who received study drug, including those positive at baseline were analysed, it was 30.4% and 18.8% in patients receiving placebo and caspofungin respectively (p = 0.04). Unfortunately, patients with sepsis and with two consecutive BG samples above 80 pg/ml were classified as probable cases of IC, which allowed the investigators to break the blind and to administer them preemptive therapy with caspofungin.

Two other currently unpublished studies tried to use above-mentioned predictive models to identify high-risk patients and guide preemptive SAT. The Pilot Feasibility Study with Patients Who Are at High Risk for Developing Invasive Candidiasis in a Critical Care Setting (conducted by the Mycoses Study Group [ClinicalTrials.gov identifier: NCT01045798]) aimed to compare caspofungin against placebo. It was terminated early due to low participant enrolment (15 patients in 4 years) without providing any interpretable result. The INTENSE (a Study to Evaluate Pre-emptive Treatment for Invasive Candidiasis in High Risk Surgical Subjects [ClinicalTrials.gov identifier: NCT01122368]) study evaluated, against placebo, the efficacy and the safety of micafungin as a preemptive treatment of IC in high-risk surgical subjects with intra-abdominal infections in a multicentre randomized controlled trial (INTENSE [ClinicalTrials.gov identifier: NCT01122368]). A total of 241 patients were analysed and the proportion of patients developing IC did not differ between arms (8.9% versus 11.1%). There was no difference in mortality, invasive fungal infection-free survival and improvement of organ failures between micafungin and placebo arms. Micafungin reduced significantly the colonization index. These results collectively suggest that the targeted population is not fully understood. The low rate of IC in MSG-01 and INTENSE study is in accordance with this conclusion.

Thus, an ongoing prospective, multicentre, double-blind, randomized, controlled French trial (EMPIRICUS) chose, on the basis of Schuster and colleagues [Schuster et al. 2008] and the EPCAN study group [Leon et al. 2006, 2009a] results to target a new population at risk of IC combining multiple organ failure, sepsis of unknown origin, multiple colonization with Candida, mechanical ventilated patients for more than 4 days and treatment board-spectrum antibacterial [Timsit et al. 2013]. Furthermore, IC prediction strategies combining patient type, clinical scores and biomarkers, as proposed by Leon and colleagues [Leon et al. 2014], should be further implemented and their clinical impact should be evaluated in large, multicentre studies.

Overuse of antifungals modifies fungal ecosystem and promotes antifungal resistance

Despite the lack of clear benefit of preemptive antifungal therapy on survival, the disastrous prognosis of IC caused a great fear for the clinician and a recent survey showed that, in the ICU, at a given point of time, 7.5% of patients are on empirical antifungal therapy whereas two thirds of these patients have no documented invasive fungal infections [Azoulay et al. 2012]. As mentioned above, this large proportion of useless antifungal treatment is favoured by the high frequency of risk factors included in predictive scores in ICU patients and the low predictive value of these tools. As examples, the colonization index positive predictive value is less than 9% in the EPCAN study [Leon et al. 2009b]. In medical ICU patients, 39% developed a colonization index of more than 0.5, while, in the same period, no invasive fungal infections were diagnosed [Charles et al. 2005].

This situation is worrisome because overuse of antifungals has been shown to induce both resistance and emergence of nonalbicans isolates [Perlin, 2014]. Thus, antifungal empirical results in placing the individual patient benefit above the collective interest. Indeed, several studies have underlined the impact of prolonged prior fluconazole or caspofungin exposure on the distribution of Candida species involved in candidaemia [Chow et al. 2008; Lortholary et al. 2011]. Among them, Lortholary and colleagues showed on 2618 isolates collected over 7 years from 2441 patients that both exposures to fluconazole and caspofungin were risk factors (OR = 2.17 and 4.79, respectively) of being infected with an isolate with decreased susceptibility to fluconazole and caspofungin, respectively [Lortholary et al. 2011]. More recently, Lortholary and colleagues also found that fluconazole pre-exposure was an independent factor of bloodstream infection with C. krusei, C. tropicalis, or C. glabrata, whereas caspofungin pre-exposure was an independent factor of infection with C. parapsilosis, C. krusei, C. kefyr, C. glabrata and mixed infections [Lortholary et al. 2014]. The authors went further, demonstrating that pre-exposure to caspofungin is an independent risk factor for 30-day mortality in ICU [Lortholary et al. 2014].

In addition to inducing the emergence of potentially more virulent nonalbicans Candida species, echinocandin exposure has been shown to favour the emergence of echinocandin resistance among usually susceptible Candida species [Alexander et al. 2013; Dannaoui et al. 2012]. Dannaoui and colleagues reported 20 episodes of fungal infections caused by candin-resistant Candida spp. that were harbouring diverse and new resistance mutations. For 12 patients, the initial isolates (low MICs, wild-type FKS gene) and the subsequent isolates (after caspofungin treatment, high MIC, FKS mutation) were genetically identical [Dannaoui et al. 2012]. We also recently described a significant relationship between SAT consumption and MICs of colonizing and infecting fungi in ICU patients [Fournier et al. 2011]. Finally, two studies clearly showed that the pre-exposure to candins is associated with episodes of C. glabrata septicaemia with strains of reduced susceptibility to candins that harboured FKS mutation. Such strains were associated with a higher rate of clinical failure of echinocandin therapy [Alexander et al. 2013; Shields et al. 2013].

Collectively, these data demonstrate that, at present, SAT prescription should follow the same rules as for other antimicrobial agents. It must be effective and safe for the patient, and also for future patients.

If used, prophylaxis needs to be restricted to very specific populations

For this reason, as well as for empiric treatment, the last 2012 ESCMID guidelines could not reasonably recommend not using antifungal prophylaxis in general nonneutropenic ICU population with IC risk factors [Cornely et al. 2012]. Indeed, several antifungal prophylaxis interventions have been tested in ICU populations, mostly surgical, exhibiting well-known risk factors for IC. However, according to meta-analysis comparing azole-prophylaxis with placebo, antifungal prophylaxis seems to decrease the rate of fungal infections without significant improvement of overall survival [Shorr et al. 2005; Vardakas et al. 2006].

Thus, the optimal target population for antifungal prophylaxis, if it exists, remains unknown. Only one specific subset of patients who recently underwent abdominal surgery and had recurrent anastomotic leakages may benefit from this strategy and the last ESCMID guidelines [Cornely et al. 2012] recommends with a moderate strength (B) and a high quality of evidence the use of fluconazole prophylaxis against IC in this population. However, this recommendation is mainly based on two studies. The first exhibited high technical quality (randomized, prospective, double-bind, placebo-controlled) but was limited by the low number of evaluable patients (43) enrolled [Eggimann et al. 1999]. This study showed a lower rate of IAC in the fluconazole prophylaxis group (400 mg/day), as compared with the placebo group. The second study, prospective but noncomparative, showed that caspofungin was able to prevent IAC in 18/19 patients with gastrointestinal perforations or acute necrotizing pancreatitis [Senn et al. 2009].

Echinocandins are now considered as the cornerstone of IC targeted treatment

For many years, fluconazole was considered as the drug of choice for candidaemia [Gafter-Gvili et al. 2008; Pappas et al. 2009; Leroy et al. 2009]. This was based on a great number of clinical trials evaluating fluconazole in this indication [Anaissie et al. 1996; Rex et al. 1994; Reboli et al. 2007]. However, although C. albicans remains the main cause of both candidaemia and IC, a shift toward nonalbicans species such as C. glabrata has been observed over the past two decades in some patients, especially those previously exposed to antifungal therapy [Krcmery and Barnes, 2002; Sendid et al. 2006]. Thus, in 2009, the IDSA therapeutic approach recommended fluconazole for patients who were mild to moderately ill and who had not been exposed to azoles in the past (thus having a measurable risk for fluconazole-resistant Candida infections) and echinocandins or lipid-based polyenes for moderate to severely ill patients or patients with previous azole exposure [Pappas et al. 2009]. Recently, evidence points to the fact that there may be an advantage of initial treatment with echinocandins over azoles [Andes et al. 2012]. Several other reasons are currently supporting the choice of this family as a first intention treatment: broad-spectrum fungicidal activity, biofilm activity [Kuhn et al. 2002], low rate of resistant candida species [Pfaller et al. 2012], low potential for drug–drug interactions, good safety profile [Ostrosky-Zeichner et al. 2014]. Therefore, the latest international guidelines are now recommending initial treatment with echinocandins for all patients and basically all situations with the highest level of evidence (A-1) [Cornely et al. 2012]. However, as mentioned above, previous antifungal treatment can favour the emergence of resistant Candida strains and liposomal amphotericin B, which has the broadest spectrum of activity, may thus be the agent of choice for patients coming under severe antifungal selection pressure (grade B-1) and in the case of C. parapsilosis [Cornely et al. 2012]. On the other hand, amphotericin B deoxycholate, because of his substantial renal and infusion-related toxicity should not be included into targeted treatment possibilities anymore. At this time, the most ‘up to date’ approach seems nevertheless to be treating all patients with echinocandins, whatever the family (caspofungin, micafungin or anidulafungin), and reserving azoles for de-escalation in stable patients with isolates showing susceptibility to the agents. However, the de-escalation strategy remains unclear in case of fluconazole susceptible strain. Indeed, fast Maldi-Tof® species identification theoretically allows early de-escalation but beneficial effects of candin therapy may require waiting longer. Thus, whereas IDSA guidelines [Pappas et al. 2009] recommended in 2009 only 5 days of intravenous echinocandins before de-escalation, ESMID [Cornely et al. 2012] considers in the latest international guidelines that 10 days are necessary before stepping down to oral fluconazole. For patients in whom disseminated, abscesses or end-organ disease has been excluded, treatment duration is generally considered to be 14 days after the end of candidaemia, determined by at least one blood culture per day until negativity, is recommended [Cornely et al. 2012].

In the same way as for antimicrobial therapy, pharmacokinetic and pharmacodynamic data should be used through serum-level determinations to optimize dosing of antifungal agents in critically ill patients, which are at risk of increased volume of distribution. Indeed, some studies pointed out an unpredictable variability of the concentration certainly (voriconazole, posaconazole) or probably (echinocandins, polyenes) associated with treatment failure and toxicity and emergence of less-susceptible strains, probable drug interactions for azoles and echinocandins and great differences in tissue diffusion among antifungal agents [Sinnollareddy et al. 2012].

Pharmacokinetic/pharmacodynamic knowledge of antifungals, especially in the most severe ICU patients, is still poor and requires further studies [Sinnollareddy et al. 2012].

Few clinical trials used combination treatment [Rex et al. 2003; Nivoix et al. 2006; Abele-Horn et al. 1996]. On the basis of their results, there is no significant benefit of combining antifungals for the treatment of IC, except for the mostly anecdotal evidence and common practice of using polyenes and 5-fluorocytosine for severe forms of end-organ disease [Cornely et al. 2012]. Indeed, 5-flurocytosine acts synergistically with polyenes and possess remarkable diffusion properties in difficult to reach tissues such as brain, eyes and cardiac vegetations.

In addition to these pharmacological considerations, it is important to emphasize the critical importance of an adequate candidaemia source control, which has been showed to have a synergistic effect with treatment adequacy on patient survival [Bassetti et al. 2014]. Indeed, the absence of source control within the first 48 hours (OR = 2.99, p = 0.001) was associated with the risk of death in the case of septic shock due to candidaemia. Therefore, all possible central lines should be removed [Andes et al. 2012] and any identified collection should be drained. Similarly, all patients with candidaemia have to undergo a dilated eye exam and an echocardiography to respectively rule out ocular (16% of candidaemia) and cardiac (8% of candidaemia) disseminations [Cornely et al. 2012].

Recently, to introduce the appropriate management of candidaemia into clinical practice, bundles based on key guideline recommendations have been developed [Takesue et al. 2014]. The value of such a strategy was emphasized by the poor status of guideline adherence in this study, as reflected by the only 6.9% compliance rate for achieving all bundle elements. These results are even more important since the compliance with the bundles was an independent predictor of clinical success and mortality.

A particular case, intra-abdominal IC

Patients who have undergone recent intra-abdominal events belong to a subset of general ICU population with a uniquely high risk of IC [Montravers et al. 2006; Dupont et al. 2002]. Indeed, IAC is particular because of its pathophysiology because any perforation or opening of the digestive tract results in contamination of the peritoneum by bowel flora including, especially in the upper digestive tract, Candida. Most of time, surgical cleaning of the abdominal cavity and antibiotics allow full recovery. Nevertheless, in case of recurrent peritonitis following anastomotic leakage or persistent abdominal inflammation (such as pancreatitis), progression of Candida colonization to IAC may occur [Calandra et al. 1989]. Additional factors such as an increasing amount of Candida spp. in a nonfunctioning bowel, prolonged antibiotic therapy and/or requirement for organ support may also play a role in this phenomenon. IAC, which includes Candida peritonitis or intra-abdominal abscesses, occurs in 30–40% of patients with secondary and tertiary peritonitis and is burdened by a mortality reported between 25% and 60% [Dupont et al. 2002; Sandven et al. 2002; Montravers et al. 2011]. However, international guidelines do not address IAC, probably because of the lack of standardized definition and diagnostic criteria. Thus, the Italian Society of Intensive Care Medicine and the International Society of Chemotherapy endorsed a project aimed at producing recommendations for the management of IAC in non-immune-competent patients [Bassetti et al. 2013]. This consensus statement underlines the unmet need in this field and the lack of dedicated studies in this clinical setting. Indeed, none of the recommendations is based on a level 1 quality of evidence, i.e. on at least one properly designed randomized, controlled trial. Nevertheless, concerning diagnosis, the expert panel (EP) strongly recommends, for every purulent and necrotic intra-abdominal specimen obtained during surgery or by percutaneous aspiration in all patients with nonappendicular abdominal infections including secondary and tertiary peritonitis, systematic direct microscopy examination for yeast detection, specific culture for Candida spp. and species identification (grade A2). Blood culture (in specific media if available) is also needed at the time of suspicion or diagnosis of IAC (grade A2). The EP underlines the importance of the quality of the microbiological samples to differentiate infection and colonization and recommends to consider SAT only when adequate (i.e. obtained surgically or within 24 h from external drainage) intra-abdominal specimens were positive for Candida, irrespective of the fungal concentration. Indeed, only surgically obtained positive cultures have been shown to be associated with higher mortality and positive cultures from drains in place for more than 24 h should not be treated [Lee et al. 2002; Prakash et al. 2008]. Concerning treatment, it first seems important to recall that the effect of SAT on IAC was never evaluated and that its use in this field, albeit logical, is based on opinions of respected authorities. Although Candida in the peritoneum is a risk factor of mortality, prophylactic therapy or targeted therapy use did not improve prognosis [Montravers et al. 2006]. IAC targeted therapy recommendations include echinocandins or lipid-based amphotericin B as first rank options and can be simplified by stepping down to an azole was allowed after 5–7 days of fungicidal therapy, in the case of susceptible species and clinically stable patient. First-line azole use is possible in non-critically ill patients without previous exposure to azoles. Duration of treatment is much more difficult to standardize than for candidaemia. The experts only marginally support at least 10–14 days of antifungal treatment with the lowest level of evidence in patients with IAC and clinically ameliorating. Although only based on the expert panel point of view, empirical therapy should be discontinued after 3–5 days in patients without proven Candida infection but clinically improved and immediately in patients without proven Candida infection and not clinically improved.

As mentioned above, the FUNGINOS study demonstrated the high accuracy of BG in predicting IAC in patients with abdominal surgery or acute pancreatitis early and reliably [Tissot et al. 2013]. On the other hand only scarce data are available on the real value of mannan, antimannann and tests in IAC and validating studies are required [Bassetti et al. 2013]. However, as for BG, the EP recommends their use when available and go so far as to support a mannan or antimannann-driven preemptive approach (in patients with intra-abdominal infections with or without specific risk factors for Candida infections).

Finally, as mentioned above (in the prophylaxis part of this review), two small prospective studies, including one placebo-controlled, suggested that antifungal prophylaxis in patients with anastomotic leakage after abdominal surgery may prevent the development of IC [Eggimann et al. 1999; Senn et al. 2009]. Although numbers of patients included were small, this subgroup of high-risk surgical patients is, at present, the only that may benefit for antifungal prophylaxis.

Conclusion

Recent studies are showing that IC management represent an unmet need and still has to be improved. Whereas targeted treatment is now clearly defined, the empirical/preemptive strategy is still controversial because the population that may benefit from this treatment is still poorly understood. Inclusion protocols combining patient type, clinical scores and biomarkers should be used in RCT in order to maximize our chance to demonstrate an effect on the outcome and to limit the overuse of antifungal therapy that promotes resistance. The results of an ongoing randomized, controlled trial including patients with sepsis and multiple organ failures are awaited with great interest.

On the other hand, although clinical predictive models and biomarkers have been developed, more accurate and earlier diagnostic methods are still needed. Recently developed proteomic (MALDI-TOF) and genomic (PCR) microbiological methods that are now frequently used for bacteriaemia should be better implemented for candidaemia.

At present, preemptive treatment introduction should be carefully weighted, knowing the uncertainty of this attitude and its ecological effects.

Footnotes

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Conflict of interest statement

The university of JFT received research grants from Astellas and Merck. JFT received educational grants from Gilead. JFT performed lecture in symposium organized by Astellas Pfizer and Merck. EW has no conflicts of interest to declare.

References

Abele-Horn

Kopp

Sternberg

Ohly

Dauber

Russwurm

(1996) A randomized study comparing fluconazole with amphotericin B/5-flucytosine for the treatment of systemic Candida infections in intensive care patients. Infection 24: 426–432.

Alexander

Johnson

Pfeiffer

Jimenez-Ortigosa

Catania

Booker

(2013) Increasing echinocandin resistance in Candida glabrata: clinical failure correlates with presence of FKS mutations and elevated minimum inhibitory concentrations. Clin Infect Dis 56: 1724–1732.

Anaissie

Darouiche

Abi-Said

Uzun

Mera

Gentry

(1996) Management of invasive candidal infections: results of a prospective, randomized, multicenter study of fluconazole versus amphotericin B and review of the literature. Clin Infect Dis 23: 964–972.

Andes

Safdar

Baddley

Playford

Reboli

Rex

(2012) Impact of treatment strategy on outcomes in patients with candidemia and other forms of invasive candidiasis: a patient-level quantitative review of randomized trials. Clin Infect Dis 54: 1110–1122.

Azoulay

Dupont

Tabah

Lortholary

Stahl

Francais

(2012) Systemic antifungal therapy in critically ill patients without invasive fungal infection. Crit Care Med 40: 813–822.

Bassetti

Marchetti

Chakrabarti

Colizza

Garnacho-Montero

Kett

(2013) A research agenda on the management of intra-abdominal candidiasis: results from a consensus of multinational experts. Intensive Care Med 39: 2092–2106.

Bassetti, M., Righi, E., Ansaldi, F., Merelli, M., Cecilia, T., De Pascale, G. et al. (2014) A multicenter study of septic shock due to candidemia: outcomes and predictors of mortality. Intensive Care Med, in press.

Calandra

Bille

Schneider

Mosimann

Francioli

(1989) Clinical significance of Candida isolated from peritoneum in surgical patients. Lancet 2: 1437–1440.

Charles

Dalle

Aube

Doise

Quenot

Aho

(2005) Candida spp. colonization significance in critically ill medical patients: a prospective study. Intensive Care Med 31: 393–400.

10.

Chow

Golan

Ruthazer

Karchmer

Carmeli

Lichtenberg

(2008) Factors associated with candidemia caused by non-albicans Candida species versus Candida albicans in the intensive care unit. Clin Infect Dis 46: 1206–1213.

11.

Cornely

Bassetti

Calandra

Garbino

Kullberg

Lortholary

(2012) ESCMID* guideline for the diagnosis and management of Candida diseases 2012: non-neutropenic adult patients. Clin Microbiol Infect 18(Suppl. 7): 19–37.

12.

Dannaoui

Desnos-Ollivier

Garcia-Hermoso

Grenouillet

Cassaing

Baixench

(2012) Candida spp. with acquired echinocandin resistance, France, 2004-2010. Emerg Infect Dis 18: 86–90.

13.

Dupont

Bourichon

Paugam-Burtz

Mantz

Desmonts

(2003) Can yeast isolation in peritoneal fluid be predicted in intensive care unit patients with peritonitis? Crit Care Med 31: 752–757.

14.

Dupont

Paugam-Burtz

Muller-Serieys

Fierobe

Chosidow

Marmuse

(2002) Predictive factors of mortality due to polymicrobial peritonitis with Candida isolation in peritoneal fluid in critically ill patients. Arch Surg 137: 1341–1346; discussion 1347.

15.

Eggimann

Francioli

Bille

Schneider

Chapuis

(1999) Fluconazole prophylaxis prevents intra-abdominal candidiasis in high-risk surgical patients. Crit Care Med 27: 1066–1072.

16.

Eggimann

Garbino

Pittet

(2003) Management of Candida species infections in critically ill patients. Lancet Infect Dis 3: 772–785.

17.

Eggimann, P. and Pittet, D. (2014) Candida colonization index and subsequent infection in critically ill surgical patients: 20 years later. Intensive Care Med, in press.

18.

Ellis

Al-Ramadi

Finkelman

Hedstrom

Kristensen

Ali-Zadeh

(2008) Assessment of the clinical utility of serial beta-D-glucan concentrations in patients with persistent neutropenic fever. J Med Microbiol 57: 287–295.

19.

Fortun

Meije

Buitrago

Gago

Bernal-Martinez

Peman

(2014) Clinical validation of a multiplex real-time PCR assay for detection of invasive candidiasis in intensive care unit patients. J Antimicrob Chemother 69: 3134–3141.

20.

Fournier

Schwebel

Maubon

Vesin

Lebeau

Foroni

(2011) Antifungal use influences Candida species distribution and susceptibility in the intensive care unit. J Antimicrob Chemother 66: 2880–2886.

21.

Gafter-Gvili

Vidal

Goldberg

Leibovici

Paul

(2008) Treatment of invasive candidal infections: systematic review and meta-analysis. Mayo Clin Proc 83: 1011–1021.

22.

Garey

Rege

Pai

Mingo

Suda

Turpin

(2006) Time to initiation of fluconazole therapy impacts mortality in patients with candidemia: a multi-institutional study. Clin Infect Dis 43: 25–31.

23.

Garrouste-Orgeas

Timsit

Tafflet

Misset

Zahar

Soufir

(2006) Excess risk of death from intensive care unit-acquired nosocomial bloodstream infections: a reappraisal. Clin Infect Dis 42: 1118–1126.

24.

Karageorgopoulos

Vouloumanou

Ntziora

Michalopoulos

Rafailidis

Falagas

(2011) beta-D-glucan assay for the diagnosis of invasive fungal infections: a meta-analysis. Clin Infect Dis 52: 750–770.

25.

Kollef

Micek

Hampton

Doherty

Kumar

(2012) Septic shock attributed to Candida infection: importance of empiric therapy and source control. Clin Infect Dis 54: 1739–1746.

26.

Krcmery

Barnes

(2002) Non-albicans Candida spp. causing fungaemia: pathogenicity and antifungal resistance. J Hosp Infect 50: 243–260.

27.

Kuhn

D.M.

George

Chandra

Mukherjee

Ghannoum

(2002) Antifungal susceptibility of Candida biofilms: unique efficacy of amphotericin B lipid formulations and echinocandins. Antimicrob Agents Chemother 46: 1773–1780.

28.

Kumar

Roberts

Wood

Light

Parrillo

Sharma

(2006) Duration of hypotension before initiation of effective antimicrobial therapy is the critical determinant of survival in human septic shock. Crit Care Med 34: 1589–1596.

29.

Lee

Fung

Chen

Jwo

Hung

(2002) Candida peritonitis due to peptic ulcer perforation: incidence rate, risk factors, prognosis and susceptibility to fluconazole and amphotericin B. Diagn Microbiol Infect Dis 44: 23–27.

30.

Leon

Alvarez-Lerma

Ruiz-Santana

Leon

Nolla

Jorda

(2009) Fungal colonization and/or infection in non-neutropenic critically ill patients: results of the EPCAN observational study. Eur J Clin Microbiol Infect Dis 28: 233–242.

31.

Leon

Ostrosky-Zeichner

Schuster

(2014) What's new in the clinical and diagnostic management of invasive candidiasis in critically ill patients. Intensive Care Med 40: 808–819.

32.

Leon

Ruiz-Santana

Saavedra

Almirante

Nolla-Salas

Alvarez-Lerma

(2006) A bedside scoring system (“Candida score”) for early antifungal treatment in nonneutropenic critically ill patients with Candida colonization. Crit Care Med 34: 730–737.

33.

Leon

Ruiz-Santana

Saavedra

Galvan

Blanco

Castro

(2009b) Usefulness of the “Candida score” for discriminating between Candida colonization and invasive candidiasis in non-neutropenic critically ill patients: a prospective multicenter study. Crit Care Med 37: 1624–1633.

34.

Leroy

Gangneux

Montravers

Mira

Gouin

Sollet

(2009) Epidemiology, management, and risk factors for death of invasive Candida infections in critical care: a multicenter, prospective, observational study in France (2005–2006). Crit Care Med 37: 1612–1618.

35.

Lortholary

Desnos-Ollivier

Sitbon

Fontanet

Bretagne

Dromer

(2011) Recent exposure to caspofungin or fluconazole influences the epidemiology of candidemia: a prospective multicenter study involving 2,441 patients. Antimicrob Agents Chemother 55: 532–538.

36.

Lortholary

Renaudat

Sitbon

Madec

Denoeud-Ndam

Wolff

. for the French Mycosis Study Group (2014) Worrisome trends in incidence and mortality of candidemia in intensive care units (Paris area, 2002-2010). Intensive Care Med 40: 1303–1312.

37.

Marco

Lockhart

Pfaller

Pujol

Rangel-Frausto

Wiblin

(1999) Elucidating the origins of nosocomial infections with Candida albicans by DNA fingerprinting with the complex probe Ca3. J Clin Microbiol 37: 2817–2828.

38.

Mikulska

Calandra

Sanguinetti

Poulain

Viscoli

for the Third European Conference on Infections in Leukemia Group (2010) The use of mannan antigen and anti-mannan antibodies in the diagnosis of invasive candidiasis: recommendations from the Third European Conference on Infections in Leukemia. Crit Care 14: R222.

39.

Mohr

Sims

Paetznick

Rodriguez

Finkelman

Rex

(2011) Prospective survey of (1–>3)-beta-D-glucan and its relationship to invasive candidiasis in the surgical intensive care unit setting. J Clin Microbiol 49: 58–61.

40.

Montravers

Dupont

Gauzit

Veber

Auboyer

Blin

(2006) Candida as a risk factor for mortality in peritonitis. Crit Care Med 34: 646–652.

41.

Montravers

Mira

Gangneux

Leroy

Lortholary

(2011) A multicentre study of antifungal strategies and outcome of Candida spp. peritonitis in intensive-care units. Clin Microbiol Infect 17: 1061–1067.

42.

Morrell

Fraser

Kollef

(2005) Delaying the empiric treatment of candida bloodstream infection until positive blood culture results are obtained: a potential risk factor for hospital mortality. Antimicrob Agents Chemother 49: 3640–3645.

43.

Nivoix

Zamfir

Lutun

Kara

Remy

Lioure

(2006) Combination of caspofungin and an azole or an amphotericin B formulation in invasive fungal infections. J Infect 52: 67–74.

44.

Nucci

Anaissie

(2001) Revisiting the source of candidemia: skin or gut? Clin Infect Dis 33: 1959–1967.

45.

Ostrosky-Zeichner

Shoham

Vazquez

Reboli

Betts

Barron

(2014) MSG-01: a randomized, double-blind, placebo-controlled trial of caspofungin prophylaxis followed by preemptive therapy for invasive candidiasis in high-risk adults in the critical care setting. Clin Infect Dis 58: 1219–1226.

46.

Pappas

Kauffman

Andes

Benjamin

Jr. Calandra

Edwards

Jr. (2009) Clinical practice guidelines for the management of candidiasis: 2009 update by the Infectious Diseases Society of America. Clin Infect Dis 48: 503–535.

47.

Perlin

(2014) Echinocandin resistance, susceptibility testing and prophylaxis: implications for patient management. Drugs 74: 1573–1585.

48.

Pfaller

Castanheira

Lockhart

Ahlquist

Messer

Jones

(2012) Frequency of decreased susceptibility and resistance to echinocandins among fluconazole-resistant bloodstream isolates of Candida glabrata. J Clin Microbiol 50: 1199–1203.

49.

Piarroux

Grenouillet

Balvay

Tran

Blasco

Millon

(2004) Assessment of preemptive treatment to prevent severe candidiasis in critically ill surgical patients. Crit Care Med 32: 2443–2449.

50.

Pittet

Monod

Suter

Frenk

Auckenthaler

(1994) Candida colonization and subsequent infections in critically ill surgical patients. Ann Surg 220: 751–758.

51.

Poissy

Sendid

Damiens

Ichi Ishibashi

Francois

Kauv

(2014) Presence of Candida cell wall derived polysaccharides in the sera of intensive care unit patients: relation with candidaemia and Candida colonisation. Crit Care 18: R135.

52.

Prakash

Sharma

Saxena

Somashekar

Khare

Mishra

(2008) Effect of Candida infection on outcome in patients with perforation peritonitis. Indian J Gastroenterol 27: 107–109.

53.

Reboli

Rotstein

Pappas

Chapman

Kett

Kumar

. for the Anidulafungin Study Group (2007) Anidulafungin versus fluconazole for invasive candidiasis. N Engl J Med 356: 2472–2482.

54.

Rex

Bennett

Sugar

Pappas

Van Der Horst

Edwards

(1994) A randomized trial comparing fluconazole with amphotericin B for the treatment of candidemia in patients without neutropenia. Candidemia Study Group and the National Institute. N Engl J Med 331: 1325–1330.

55.

Rex

Pappas

Karchmer

Sobel

Edwards

Hadley

. for the American National Institute of Infectious Diseases and Mycoses Study Group (2003) A randomized and blinded multicenter trial of high-dose fluconazole plus placebo versus fluconazole plus amphotericin B as therapy for candidemia and its consequences in nonneutropenic subjects. Clin Infect Dis 36: 1221–1228.

56.

Sandven

Qvist

Skovlund

Giercksky

(2002) Significance of Candida recovered from intraoperative specimens in patients with intra-abdominal perforations. Crit Care Med 30: 541–547.

57.

Schuster

Edwards

Jr. Sobel

Darouiche

Karchmer

Hadley

(2008) Empirical fluconazole versus placebo for intensive care unit patients: a randomized trial. Ann Intern Med 149: 83–90.

58.

Sendid

Cotteau

Francois

D'haveloose

Standaert

Camus

(2006) Candidaemia and antifungal therapy in a French University Hospital: rough trends over a decade and possible links. BMC Infect Dis 6: 80.

59.

Senn

Eggimann

Ksontini

Pascual

Demartines

Bille

(2009) Caspofungin for prevention of intra-abdominal candidiasis in high-risk surgical patients. Intensive Care Med 35: 903–908.

60.

Senn

Robinson

Schmidt

Knaup

Asahi

Satomura

(2008) 1,3-Beta-D-glucan antigenemia for early diagnosis of invasive fungal infections in neutropenic patients with acute leukemia. Clin Infect Dis 46: 878–885.

61.

Shields

Nguyen

Press

Updike

Clancy

(2013) Caspofungin MICs correlate with treatment outcomes among patients with Candida glabrata invasive candidiasis and prior echinocandin exposure. Antimicrob Agents Chemother 57: 3528–3535.

62.

Shorr

Chung

Jackson

Waterman

Kollef

(2005) Fluconazole prophylaxis in critically ill surgical patients: a meta-analysis. Crit Care Med 33: 1928–1935; quiz 1936.

63.

Sinnollareddy

Peake

Roberts

Lipman

Roberts

(2012) Using pharmacokinetics and pharmacodynamics to optimise dosing of antifungal agents in critically ill patients: a systematic review. Int J Antimicrob Agents 39: 1–10.

64.

Tabah

Koulenti

Laupland

Misset

Valles

Bruzzi De Carvalho

(2012) Characteristics and determinants of outcome of hospital-acquired bloodstream infections in intensive care units: the EUROBACT International Cohort Study. Intensive Care Med 38: 1930–1945.

65.

Takesue

Ueda

Mikamo

Oda

Takakura

Kitagawa

. on behalf of the ACTIONs Project (2014) Management bundles for candidaemia: the impact of compliance on clinical outcomes. J Antimicrob Chemother. DOI: 10.1093/jac/dku414.

66.

Timsit

Azoulay

Cornet

Gangneux

Jullien

Vesin

(2013) EMPIRICUS micafungin versus placebo during nosocomial sepsis in Candida multi-colonized ICU patients with multiple organ failures: study protocol for a randomized controlled trial. Trials 14: 399.

67.

Tissot

Lamoth

Hauser

Orasch

Fluckiger

Siegemund

(2013) Beta-glucan antigenemia anticipates diagnosis of blood culture-negative intraabdominal candidiasis. Am J Respir Crit Care Med 188: 1100–1109.

68.

Vardakas

Samonis

Michalopoulos

Soteriades

Falagas

(2006) Antifungal prophylaxis with azoles in high-risk, surgical intensive care unit patients: a meta-analysis of randomized, placebo-controlled trials. Crit Care Med 34: 1216–1224.

69.

Vincent

Rello

Marshall

Silva

Anzueto

Martin

(2009) International study of the prevalence and outcomes of infection in intensive care units. JAMA 302: 2323–2329.

70.

Wisplinghoff

Bischoff

Tallent

Seifert

Wenzel

Edmond

(2004) Nosocomial bloodstream infections in US hospitals: analysis of 24,179 cases from a prospective nationwide surveillance study. Clin Infect Dis 39: 309–317.

Management of invasive candidiasis in nonneutropenic ICU patients

Abstract

Keywords

Introduction

Incidence and mortality rate of candidiasis are high

Early treatment of candidaemia decreases mortality of ICU patients with septic shock

Effective strategies to identify high-risk patients are currently lacking

Non-culture-based methods to accelerate diagnosis of IC are promising

No evidence for effectiveness of empiric/preemptive treatment in ICU patient is currently available

Overuse of antifungals modifies fungal ecosystem and promotes antifungal resistance

If used, prophylaxis needs to be restricted to very specific populations

Echinocandins are now considered as the cornerstone of IC targeted treatment

A particular case, intra-abdominal IC

Conclusion

Footnotes

Funding

Conflict of interest statement

References