1101

Introduction: Voriconazole (VRC) is essential for the treatment of invasive fungal diseases in critically ill patients; however, data are scarce on VRC concentrations in this population. The aim of this study was to evaluate VRC concentrations and their relationship with clinical variables in a cohort of ICU patients. Methods: We reviewed all patients in whom VRC levels were measured during treatment in the ICU from January 2010 to December 2012. VRC levels were measured just before (trough, C0) drug administration using a high-performance liquid chromatography (HPLC) technique. Optimal C0 was between 1.0 and 5.0 mg/L. Early therapy was defined as a measurement within 4 days of VRC treatment. Results: We obtained 65 VRC levels from 41 patients (median age 56 years; 24 male). Median APACHE II score on ICU admission was 15 [12-16] and median SOFA score on the day of measurement was 5 [3-7]. All patients were also treated with proton pump inhibitors. Median VRC daily dose was 400 [400-600] mg, which corresponded to 6.9 [5.3-8.7] mg/kg per day. Median time from VRC initiation and VRC levels measurement was 6 [4-8] days. Median drug trough concentration was 2.1 [1.4-4] mg/L. C0 were adequate in 38 (58%) of measurements, insufficient in 13 (20%) and excessive in 14 (22%), without significant differences related to timing (early and late). Also, we found no differences in C0 between intravenous and oral administration. Furthermore, drug concentrations were not correlated to daily drug dose (indexed to body weight), even when only the first drug level was analyzed. Intravenous therapy had higher proportion of adequate C0 (23/32) than oral therapy (15/33, p = 0.04). Also, patients with more severe organ dysfunction (SOFA > 5) had higher C0 than those with lower score (3.8 [1.7-6.0] vs. 2.1 [1-3.6] mg/L, p = 0.03). Conclusions: Blood concentrations of VRC are highly variable in critically ill patients. Drug concentrations are higher in patients with more severe organ impairment. Intravenous therapy resulted in a higher proportion of adequate C0 when compared to oral doses.