Discontinuation of empirical antifungal therapy in ICU patients using 1,3-β-D-glucan

Empirical antifungal therapy in high-risk ICU patients is an attractive strategy, but overuse of antifungal agents is a potential problem. We evaluated if ICU patients at high risk to develop candidaemia identified by a prediction rule could discontinue empirical antifungal therapy on the basis of repeatedly negative 1-3-beta-d-glucan (BDG) tests. We conducted a multicentre cohort study in 85 ICU patients receiving antibiotics or with central venous catheter plus two additional factors (dialysis, parenteral nutrition, surgery, pancreatitis or receipt of corticosteroids or other immunosuppressive agents) plus either fever, hypothermia, hypotension, acidosis, elevated C-reactive protein or leucocytosis. Blood cultures (days 1 and 2) and BDG (days 1-3, baseline period) were performed and anidulafungin was given. On day 4, patients with negative blood cultures and BDG discontinued antifungal therapy. Registered in ClinicalTrials.gov (NCT01734525). The incidence of candidaemia was 8.2% in patients selected versus 0.5% in patients without entry criteria (16.9 times higher). Sixty-four patients (75.3%) had baseline positive BDG, including 7 with candidaemia. All 21 patients with baseline negative BDG discontinued anidulafungin on day 4. None developed candidaemia until day 30. Early discontinuation of empirical echinocandin therapy in high-risk ICU patients based on consecutive negative BDG tests may be a reasonable strategy, with great potential to reduce the overuse of echinocandins in ICU patients. Prospective studies with a higher number of patients are needed.