Abstract
Objective: To assess the applicability of an ultrafiltration method to determine free plasma concentrations of beta-lactam antibiotics in ICU patients.
Patients and methods: Eligible patients included adult ICU patients treated with ceftazidime (CAZ), meropenem (MEM), piperacillin/tazobactam (PIP/TAZ), or flucloxacillin (FXN) by continuous infusion. Up to two arterial blood samples were drawn at ...
Abstract
Objective: To assess the applicability of an ultrafiltration method to determine free plasma concentrations of beta-lactam antibiotics in ICU patients.
Patients and methods: Eligible patients included adult ICU patients treated with ceftazidime (CAZ), meropenem (MEM), piperacillin/tazobactam (PIP/TAZ), or flucloxacillin (FXN) by continuous infusion. Up to two arterial blood samples were drawn at steady state. Patients could be included more than once if they received another antibiotic. Free drug concentrations were determined by HPLC-UV after ultrafiltration, using a method that maintained physiological conditions (pH 7.4/37 °C). Total drug concentrations were determined to calculate the unbound fraction. In a post-hoc analysis, free concentrations were compared with the target value of 4× the epidemiological cut-off value (ECOFF) for Pseudomonas aeruginosa as a worst-case scenario for empirical therapy with CAZ, MEM, or PIP/TAZ, and against methicillin-sensitive Staphylococcus aureus (MSSA) for targeted therapy with FXN.
Results: Fifty different antibiotic treatment periods in 38 patients were evaluated. The concentrations of the antibiotics showed a wide range due to the fixed dosing regimen in a mixed population with variable kidney function. The mean unbound fractions (fu) of CAZ, MEM, and PIP were 102.5%, 98.4%, and 95.7%, with inter-patient variability of <6%. The mean fu of FXN was 11.6%, with inter-patient variability of 39%. It was observed that 2 out of 12 free concentrations of CAZ, 1 of 40 concentrations of MEM, and 11 of 23 concentrations of PIP were below the applied target concentration of 4×ECOFF for P. aeruginosa. All concentrations of FXN (9 samples from 6 patients) were >8×ECOFF for MSSA.
Conclusions: For therapeutic drug monitoring purposes, measuring total or free concentrations of CAZ, MEM, or PIP is seemingly adequate. For highly protein-bound beta-lactams such as FXN, free concentrations should be favored in ICU patients with prevalent hypoalbuminemia.
Altmetric