Optimal Design for the improvement of sampling schedules of microdialysis studies.
M. Scholz (1), N. Plock (1,2), C. Kloft (1)
(1) Martin-Luther-Universitaet Halle-Wittenberg, Dept. Clinical Pharmacy, Halle, Germany (2) Nycomed, Pharmacometrics, Constance, Germany.
Objectives: The methods of Optimal Design have been explored for various applications, but not for the design of sampling schedules in microdialysis (µD) studies. In contrast to plasma sampling, µD allows to determine the concentration of drugs or PD markers at the site of action. Microdialysate is continuously sampled over a longer time period which is divided in usually identical collection time intervals. Optimal Design usually focusses on sampling time points rather than time intervals, thus for µD studies a new approach has to be taken. In a first step the experimental basis for optimisation was laid by obtaining data from a clinical trial. From the developed population PK model, optimal sampling time points were determined. Finally methodological investigations due to the special settings in µD studies were performed.
Methods: 34 individuals were included in a clinical trial with linezolid, 600 mg bid [1]. Samples were taken after single dose and at steady state over a period of 8 h. 1176 unbound plasma and 2325 µD concentrations were determined. Population PK analysis was carried out using NONMEM. POPT and PopED were used to retrospectively optimise the sampling design as optimal time points for plasma and µD samples. Based on simulated concentration-time profiles in µD the impact of placement of time points as in the traditional sampling approach on bias is assessed.
Results: Unbound linezolid PK in plasma was described by a two-compartment model including a time-dependent nonlinear influence on clearance implemented by an additional inhibition compartment [1]. Microdialysate concentrations were incorporated into the model using two additional compartments. The optimised design reduced the number of plasma samples from 40 to 6 and of µD from 20 to 4 per indvidual [2]. Investigations of the impact on bias for different scenarios of placement of sampling time intervals around the optimised time points are performed. One scenario includes the analysis of differently long, fixed intervals where the measured concentration is placed at the midinterval as used in the conducted trial and as typically used in µD studies.
Conclusions: Prerequisites for methodological investigations of Optimal Design for µD studies were generated and the special characteristics of this sampling technique explored. These analyses give useful information about the utilisation of µD as an attractive tool to monitor target-site exposure in patients.
References:
[1] Plock N, Buerger C, Kuester K, Joukhadar C, Kljucar S, Kloft C. A Population Pharmacokinetic Model for the Simultaneous Description of Linezolid Tissue and Plasma Disposition in Healthy Volunteers and Septic Patients, PAGE 15 (2006) Abstr 886 [www.page-meeting.org/?abstract=886].
[2] Plock N, Kloft C. unpublished
