Who should be treated in the ICU?

Y. Cohen


Invasive fungal infections (IFIs) in non-neutropenic patients have become, in a few decades, a major risk of infection. Invasive candidiasis (IC) represent 90% of IFI, and are the fourth leading cause of nosocomial bloodstream infections (1).

These ICs Occur in 30-50% of cases in intensive care (2, 3, 4). The prognosis is a major way impacted by the initiation of treatment period (5) particularly since the diagnosis by blood culture is insensitive and often late.

Faced with this challenge, ICU patients can receive antifungal therapy based on different therapeutic strategies.They take into account the terrain, the risk factors and the existence or not of a Candida colonization, thus defining different treatments:


Prophylactic treatment:

Priced in patients with complicated abdominal surgery (perforation, anastomotic dehiscence) , in surgical patients with length of stay in a superior resuscitation at least 3 days or even among patients on mechanical ventilation for 48 hours and having difficulty withdrawal within 3 days (6, 7, 8), prophylactic treatment in these studies give conflicting results, little concluantspour confirm that attitude.


Treatment "Preemptive"

Starting treatment is decided either in these patients heavily colonized with Candida or on clinical scores (also taking into account the existence of a settlement)

Strong colonization may be evaluated by the colonization index described by Pittet, et al (9) in surgical patients.It is defined by the ratio between the number of sites colonized by Candidaet the total number of test sites. When it is greater than or equal to 50%, it defines patients with high risk of CHD establishing an antifungal treatment.This has been evaluated by several studies with heterogeneous methodology and conflicting conclusions.

Clinical scores, either one of the factors of "very high risk" Ostrovsky et al. (10) or Candida score De Leon et al.(11) distinguish patients at major risk CI:


Factors very "high risk" include a septic patient, with at least one positive sample site Candida, the following criteria: a residence time in the upper resuscitation to 3 days, the presence of mechanical ventilation, antibiotics broad spectrum, a central catheter 1 and factors such parenteral nutrition, hemodialysis, a surgery, acute pancreatitis, steroid therapy or the use of an immunosuppressant.

A blind experiment with a complex methodology, has évaluél'intérêt of caspofungin versus placebo in 222 ICU patients with factors of high risk associated with 1,3ß-D-Glucan (instead of colonization ). The results found themselves no improvement on the occurrence of CI or mortality (12).


Candida score includes the existence of a surgery, parenteral nutrition, multifocal colonization (at least 2 sites) Candida and the existence of severe sepsis. The first 3 items are worth 1 point each and the last two points.Candida score is considered positive if the sum is greater than 2.5.

Leroy et al. showed in a prospective study, including a small number of patients, the benefit of the negative predictive value of this score for not putting in antifungal treatment of patients whose score is less than 4 (13).Further studies are needed to confirm these results.

Finally, in some studies the colonization is replaced by the determination of 1,3ß-D-glucan (see previous statement)


Empirical treatment:

This is a treatment used in patients with risk factors CI, septic remaining despite appropriate antibiotic therapy.

Schuster was evaluated in a prospective randomized double-blind study in 270 non-neutropenic patients remaining febrile resuscitation under broad-spectrum antibiotics for at least 4 days, the interest of fluconazole versus placebo (14). The result of this study showed , between the 2 groups, the lack of improvement of sepsis, lack of decrease in CI and mortality.

Note that in this study a significant points that can explain the negative result of empirical therapy is the low colonization of included patients (20%), essential criterion for the occurrence of CI.

Currently a multicenter French study (Empiricus) randomized, double-blind, taking into account the essential parameter is underway. It assesses Micafungin therapy versus placebo in patients at high risk of CHD, broken more than 4 days presenting organ failure and sepsis of unknown origin under broad spectrum antibiotic with at least one website extradigestif colonized with Candida .

The results will be to improve the management of these critically ill patients.

To improve the prognosis of patients with IC overtreatment and outside of a more efficient direct diagnosis, it is essential to better target patients at risk. A study measuring the benefit of the combination of different criteria such as the ground, the impact of certain colonization sites, different current scores, the 1,3ß-D-Glucan, the PCT, and the place of each of they would certainly complex but could better define the indications of early antifungal treatment of non-neutropenic intensive care patients.


Full conference title: 

Réunion Interdisciplinaire de Chimiothérapie Anti Infectieuse