Introduction

Ventilator-associated pneumonia (VAP) occurs in a considerable proportion of patients undergoing mechanical ventilation and is associated with substantial morbidity, two-fold mortality, and excess cost [1]. Several studies have identified Pseudomonas aeruginosa as the most frequent microorganism in patients with VAP [2]. Tracheobronchial colonization and duration of mechanical ventilation are the two most important risk factors for VAP [3]. Although risk factors for VAP and tracheobronchial colonization are probably similar, outcomes of patients with VAP and those with tracheobronchial colonization are clearly different [3, 4].

According to the results of a recent experimental study, there is a pathogenic interaction between Candida albicans and P. aeruginosa, suggesting that P. aeruginosa could be more virulent in the presence of C. albicans [5]. Other experimental studies identified physical, chemical, and environmental similarities between the two pathogens [6, 7]. Moreover, a recent clinical study identified Candida spp. tracheobronchial colonization as an independent risk factor for P. aeruginosa VAP [8]. Therefore we conducted this retrospective case–control study to determine the impact of antifungal treatment on P. aeruginosa VAP or tracheobronchial colonization in patients with Candida spp. tracheobronchial colonization.

Patients and methods

This observational retrospective case–control study was conducted in a 30-bed medical and surgical intensive care unit (ICU) from January 2004 to January 2005. No informed consent was required by the Institutional Review Board because of the retrospective and noninterventional design of the study.

All patients intubated and ventilated for longer than 48 h who had tracheobronchial colonization by Candida spp. were eligible for this study. Patients with P. aeruginosa VAP or tracheobronchial colonization diagnosed before or at the same time as Candida spp. tracheobronchial colonization were excluded. Patients were identified using the electronic files of mycology laboratory. Data collection was based on retrospective chart review.

All patients were screened on endotracheal aspirate for Candida spp. and P. aeruginosa at ICU admission and weekly thereafter. In addition, other microbiologic examinations were performed according to patient status. Antifungal treatment and its duration were based on medical staff decisions.

Definitions

Tracheobronchial colonization was defined as positive respiratory specimen culture. VAP was defined by the presence of new or progressive radiographic infiltrate associated with two of the following criteria: temperature > 38.5 °C or < 36.5 °C, leukocyte count > 10,000/μl or < 1,500/μl, purulent tracheal aspirate; and positive (≥ 106 cfu/ml) endotracheal aspirate. Only VAP episodes occurring more than 48 h after the commencement of mechanical ventilation were taken into account.

Matching criteria

Cases were patients with P. aeruginosa VAP or tracheobronchial colonization; controls, patients without P. aeruginosa VAP or tracheobronchial colonization. Every case was matched with two controls according to all the following criteria: (1) duration of mechanical ventilation before first positive respiratory specimen for P. aeruginosa (controls ≥ cases), (2) admission category (medical/surgical), (3) immunologic status, and (4) date of ICU admission when more than two potential control patients were available.

Statistical analysis

Qualitative variables were compared using the χ2 test or Fisher's exact test where appropriate. The Mann–Whitney U-test was used to compare quantitative variables.

To determine variables associated with P. aeruginosa VAP or tracheobronchial colonization, cases were compared with controls by univariate and multivariate analyses.

Please see the electronic supplementary material (ESM) for additional information.

Results

One hundred and seventeen patients were eligible. Fifteen patients were excluded because P. aeruginosa VAP or tracheobronchial colonization was diagnosed before or at the same time as Candida spp. tracheobronchial colonization.

Mycology results and antifungal treatment

130 Candida spp. were isolated in the 102 study patients. C. albicans was the most frequently isolated species (67%). Thirty-six patients (35%) received antifungal treatment. The mean duration of antifungal treatment was 13 ± 12 days. Fluconazole was the most frequently used antifungal (66%). Indications for antifungal treatment included: candidemia (n = 3), pneumonia in immunosuppressed patients (n = 8), peritonitis (n = 6), and preemptive therapy (n = 19).

Characteristics of study patients

The characteristics of the study patients are presented in Table 1. The durations of ICU stay and of antifungal treatment were highly colinear (r = 0.716).

Table 1 Patient characteristics

Risk factors for P. aeruginosa VAP or tracheobronchial colonization

Nineteen patients developed a P. aeruginosa VAP or tracheobronchial colonization, including 13 patients who received antifungal treatment (7 patients before antifungal treatment, 4 patients during antifungal treatment, and 2 patients after antifungal treatment), and 6 patients who did not receive antifungal treatment. Among the 19 patients with P. aeruginosa VAP or tracheobronchial colonization, 10 developed at least one VAP episode, and 9 remained colonized.

The 19 patients with P. aeruginosa VAP or tracheobronchial colonization were all successfully matched with 2 control patients each for a total of 38 controls. Results of univariate analysis are presented in Table 2. Multivariate analysis identified antifungal treatment as the only factor independently associated with P. aeruginosa VAP or tracheobronchial colonization (OR = 0.68, 95% CI = 0.49–0.90, p = 0.046).

Table 2 Risk factors for ventilator-associated pneumonia or tracheobronchial colonization related to Pseudomonas aeruginosa in univariate analysis

Please see the ESM for additional results.

Discussion

Our results suggest that antifungal treatment is associated with reduced risk for VAP or tracheobronchial colonization related to P. aeruginosa. To our knowledge, this study is the first to evaluate the impact of antifungal treatment on Candida spp. and P. aeruginosa interaction. Several studies have identified pathogenic interactions between microorganisms, such as herpes simplex virus and human immunodeficiency virus; influenza virus and P. aeruginosa; and C. albicans and P. aeruginosa [5, 9, 10]. Theses interactions have major environmental and medical consequences. A pathogenic interaction between C. albicans and P. aeruginosa has been demonstrated in experimental studies. Recent P. aeruginosa infection has been identified as a risk factor for fatal candidiasis in burned mice [11]. Molecular studies identified phylogenetic similarities between the two pathogens [5, 6]. The morphology and virulence of C. albicans are significantly affected by the presence of P. aeruginosa  [5]. A cell–cell signaling molecule capable of inhibiting C. albicans filamentation is produced by P. aeruginosa  [5]. P. aeruginosa forms a dense biofilm on C. albicans filaments and kills the fungus. In contrast, P. aeruginosa neither binds nor kills yeast-form C. albicans. Several P. aeruginosa virulence factors that are important in disease are involved in killing of C. albicans filaments [5]. Azoulay et al. [8] recently reported the results of the first clinical study suggesting an interaction between C. albicans and P. aeruginosa. The authors identified Candida spp. tracheobronchial colonization as an independent risk factor for P. aeruginosa pneumonia. No cause-and-effect relationship was demonstrated in that study. In addition, Candida spp. tracheobronchial colonization and P. aeruginosa pneumonia could both be a consequence of prior antibiotic treatment. However, the lack of association with Staphylococcus aureus pneumonia, another consequence of antibiotic treatment, indicates that an association between Candida spp. tracheobronchial colonization and P. aeruginosa remains plausible. A more recent experimental study has evaluated the impact of C. albicans tracheobronchial colonization on the occurrence of P. aeruginosa pneumonia [12]. Rate of P. aeruginosa pneumonia was significantly higher in rats with C. albicans tracheobronchial colonization as compared with rats without C. albicans tracheobronchial colonization (33% vs 4%, p < 0.05).

Although antifungal treatment was associated with reduced risk for P. aeruginosa VAP or tracheobronchial colonization, no significant relationship was found between antifungal treatment and P. aeruginosa VAP. However, the small number of patients with VAP precludes a definite conclusion. In addition, a recent meta-analysis outlined the similarity of risk factors for colonization and infection related to multidrug-resistant bacteria [4]. Moreover, our results may provide support for the notion that the interaction between Candida spp. and P. aeruginosa is at bronchial or biofilm level.

Bacterial biofilm has been demonstrated on inner surface of endotracheal tubes removed from mechanically ventilated patients. Bacterial biofilm may play an important role as a persistent source of infectious material for recurrent episodes of VAP [13]. Candida spp. and P. aeruginosa are the most common pathogens retrieved from endotracheal tube biofilm and tracheal secretions in patients with VAP [14]. Although all Candida spp. were taken into account in our study, previous experimental studies [5, 6, 12] were performed exclusively on C. albicans. However, in the study by Azoulay et al. [8] tracheobronchial colonization with any Candida spp. was identified as a risk factor for P. aeruginosa VAP. Further studies are needed to determine whether Candida spp. and P. aeruginosa interaction could be influenced by the nature of Candida spp.

Inclusion of immunosuppressed patients could be a matter of debate. However, immunologic status was a matching criterion. Future randomized interventional studies on the impact of antifungal treatment on P. aeruginosa VAP or tracheobronchial colonization should be conducted in immunocompetent patients.

Our study has several limitations. First, this was a retrospective observational study. Second, some of the trends observed in this study could have reached statistical significance if more patients had been included. In addition, the number of patients needed to demonstrate a beneficial effect of antifungal treatment was not calculated a priori. Third, our study was conducted in a single ICU. Therefore, our results may not be generalizable to other ICUs. Fourth, no information was available on the nature of prior antibiotic treatment or and on the quantity of Candida spp. in respiratory specimens. In addition, invasive methods were not used to diagnose VAP. However, quantitative tracheal aspirate culture was used in all patients with a high threshold (≥ 106 cfu/ml). Postmortem studies showed acceptable overall diagnostic accuracy of quantitative tracheal aspirate compared with bronchoalveolar lavage or protected specimen brush [15]. Finally, antifungal treatment was based on medical staff decisions. Among the 36 patients who received antifungals, 19 patients received preemptive antifungal treatment. This finding is consistent with the results of a recent survey conducted in French ICUs [16]. However, recovery of Candida spp. from the respiratory tract of mechanically ventilated patients without risk factors for immunosuppression is common and frequently reflects a tracheobronchial colonization [17, 18]. Some authors suggest that antifungal treatment should be based on the colonization index in these patients [19, 20].

We conclude that in patients with Candida spp. tracheobronchial colonization, antifungal treatment is associated with reduced risk for P. aeruginosa VAP or tracheobronchial colonization. Prospective randomized studies are required to confirm this result.