Max Taubert (1)*, Michael Zoller (2)*, Barbara Maier (3), Sebastian Frechen (1), Uwe Fuhr (1), Christina Scharf (2), Lorenz Frey (2), Michael Vogeser (3), Johannes Zander (3) * Both authors contributed equally to this work
(1) Department of Pharmacology, Hospital of the University of Cologne, Germany (2) Department of Anesthesiology, Hospital of the Ludwig-Maximilians-University of Munich, Germany (3) Institute of Laboratory Medicine, Hospital of the Ludwig-Maximilians-University of Munich, Germany
Objectives: Linezolid is an antimicrobial agent used in the treatment of severe infections including a set of multi-resistant pathogens. Although it takes on an important role in the treatment of infections in intensive care little is known about its pharmacokinetic properties and the sources of variability in this group of patients. The objective of our analysis was to develop a population pharmacokinetic model and to investigate the sources of variability in critically ill patients.
Methods: Data from a collective of 53 intensive care patients (19 women, 18 on Continuous Renal Replacement Therapy [CRRT], 16 with Acute Respiratory Distress Syndrome [ARDS], 15 with lung transplantation, 8 with liver transplantation, 16 with sepsis, median age 58 years, median weight 75 kg, median APACHE-II score 28, median SOFA score 12) with severe infections was used. Intravenous short-duration infusions of 600mg linezolid were administered every 12 hours for 4 days. Multiple serum samples per day were taken and linezolid was quantified by a liquid chromatography/tandem mass spectrometry method. We developed a population pharmacokinetic model and investigated the influence of a set of covariates including laboratory values, intensive care scores and demographics using NONMEM.
Results: A two-compartment model with first-order elimination was found to be most suitable to describe the data. The central and peripheral volume of distribution were 19 (15, 26) L and 25 L respectively, the elimination clearance was 7.6 (4.4, 12) L/h (median and interquartile range [where applicable] of individual estimates). A complete bioavailability and an absorption constant of 1.7/h were found for oral administration. The body weight was linked to the central volume of distribution and to the elimination clearance. With increasing SOFA score the clearance decreased. The need for CRRT was linked to a central volume of distribution increased by 38% and patients suffering from ARDS had a 79% higher clearance. By including these covariate relationships the observed inter-individual variability could be reduced by 28% for the clearance and by 52% for the central volume of distribution.
Conclusions: The individual pharmacokinetic parameters of linezolid depend on the body weight and various parameters describing the severity of the disease. The individual concentrations required for the optimization of linezolid dosing may be predicted using our model.
Reference: PAGE 24 () Abstr 3480 [www.page-meeting.org/?abstract=3480]
Poster: Drug/Disease modeling - Infection