Hafedh Marouani PhD(1), Claire Contargyris MD(2), Stamatios Anastopoulos(1), Ioannis Kartsonakis(1), Christian Woloch PhD(1), Athanassios Iliadis Pr(1)
(1) Dpt. of Pharmacokinetics, UMR INSERM 911 CRO2, Aix-Marseille University, France. (2) Intensive care Dpt., Laveran Military Hospital, 13 Bd Laveran, 13013 Marseille, France.
Objectives: Amikacin is concentration-dependent antibiotic used against severe gram-negative infections. The use of amikacin is difficult because of its narrow therapeutic index (renal and auditory toxicities) and its wide pharmacokinetic variability. The aim of this work is to propose, in a population of critically ill patients, a rapid and simple bedside tool which allows precisely reaching the amikacin efficiency target concentrations of 80 mg/L without exceeding the toxic threshold of 2.5 mg/L for residual concentrations.
Methods: 1) Population PK study: sparse, retrospective therapeutic drug monitoring data for amikacin were obtained from 91 critically ill septic patients during the first 24 to 96 hours after treatment. All patients received 30 min intravenous infusion of doses ranging from 750 mg to 4 g of amikacin in a once daily regimen. Population modeling was performed using Monolix software [1]. Covariate analysis included weight, gender, age, total proteins and renal clearance (RC). 2) Kinetic nomograms (KN) [2]: based on the population analysis, three groups of patients were identified according to the renal impairment status. For each group, kinetic nomograms were obtained for the individualization of amounts and schedules in the regimens. All calculations were performed with the MATLAB software [3].
Results: A two-compartment model with first-order elimination best fitted the amikacin concentrations. RC was revealed significant covariate in the final model; it allows identification of 3 groups of patients: RC<20, 20<RC<90 and 120<RC mL/min. Population analysis confirmed the wide interindividual PK variability (51.5% for amikacin clearance) and subsequent need for individual dosage adjustment. For each group, the first KN uses the assayed amikacin concentration from samples drawn between 0 and 6h to adjust the amount of the next administration, and the second KN uses assayed drug concentration between 18 and 30h to adjust the timing for the next administration.
Conclusions: KN were successfully developed for amikacin therapy in critically ill septic patients. They allow reliable dosage and schedule adjustment immediately for the second administration of amikacin. Based on only two samples, dose and schedule of further amikacin administration were individualized, thus maximizing bactericidal killing while minimizing the risk for adaptive resistance and avoiding severe drug toxicity.
References:
[1] MONOLIX 2.4 User Guide, vol. http://software.monolix.org. Accessed Oct 2011.
[2] Marouani H, Zografidis A, Iliadis A. Kinetic nomograms assist individualization of drug regimens. Clin Pharmacokinet. 2011 Dec 1;50(12):773-9.
[3] MATLAB. High-performance Numeric Computation and Visualization Software. In. 7.0 ed. Natick MA: The Math Works; 2004.
Reference: PAGE 22 (2013) Abstr 2809 [www.page-meeting.org/?abstract=2809]
Poster: Other Modelling Applications