Underexposure of linezolid in critically ill patients with extracorporeal membrane oxygenation (ECMO): results from a monocentric study including blood samples from 52 patients.

PURPOSE: Linezolid serum concentrations in critically ill patients show high variability. In this retrospective study, we analysed the linezolid plasma through levels (Cmin) in patients on intensive care units (ICUs) under extracorporeal membrane oxygenation (ECMO) support. The aim was to evaluate if patients' clinical or demographic characteristics influence drug concentrations and if these are within the therapeutic range.

METHODS

In total, 156 linezolid trough plasma concentrations from 52 ICU ECMO patients were analysed. Linezolid trough levels were correlated with the following clinical and demographic characteristics: Age, sex, body mass index (BMI), dosage per day, creatinine clearance (CrCl), and requirement for renal replacement therapy (RRT). Drug concentrations were quantified by liquid chromatography with tandem mass spectrometry. The European Committee on Antimicrobial Susceptibility Testing susceptibility breakpoints of 4 mg/L of linezolid for staphylococci and enterococci and of 2 mg/L for streptococci and other Gram-positive rods were used as minimum target concentration.

RESULTS

Patients were treated with standard linezolid dosage (2 × 600 mg iv per day; n = 88 Cmin, 56.4%) or by adjusted dosing regimens (n = 68 Cmin, 43.6%). The total amount of the drug ranged from 600 to 2400 mg per day (median 1200, IQR 1200-1800 mg). The mean of all trough levels measured among the cohort was 2.32 mg/L (median 1.55 mg/L, IQR 0.61-3.07). In total, 84.0% of all measurements were below the 4 mg/L breakpoint (62.2% below the 2 mg/L breakpoint), with 90.4% (78.8%) of patients showing inadequate levels in at least one measurement and 63.5% (36.5%) of patients below the threshold in every measurement. This result was irrespective of a standard or an adjusted dosing regimen. Female sex (p = 0.022), RRT (p = 0.047), increased CrCl (p = 0.012), and BMI (p = 0.023) were significantly correlated with lower Cmin levels. Patients' individual linezolid trough levels and outcomes did not display conclusive patterns, irrespective of dosing or duration of treatment.

CONCLUSIONS

Both standard and increased dosing regimens of linezolid showed potentially inadequate linezolid plasma levels in the large majority of critically ill ECMO patients of our study. Future TDM studies with optimized dosing and application regimens in ECMO patients are warranted.