Ilia, L. (1), Busse, D. (1,2), Michelet, R. (1), Kloft, C. (1)
(1) Dept. of Clinical Pharmacy and Biochemistry, Institute of Pharmacy, Freie Universitaet Berlin, Germany, (2) and Graduate Research Training program PharMetrX, Germany
Objectives:
Determination of drug concentrations at target sites such as interstitial space fluid (ISF) is gaining increasing attention to predict probability of pharmacokinetic/pharmacodynamics target attainment (PTA) which can be assessed using the minimally invasive microdialysis (µD) sampling technique [1], based on passive diffusion of drug molecules across a semipermeable membrane at the tip of a µD catheter. The catheter is continuously perfused with a drug-free solution (perfusate) closely resembles unbound drug concentration in the ISF. Loss of drug molecules (delivery) related to the sampling technique is accounted for by a catheter-specific calibration, e.g. via retrodialysis (recovery) and represented in the relative recovery value (RR, %).
In a clinical µD study, target-site PK of linezolid (LIN) were investigated in (non-)obese patients [2]. However, the high observed µD technique variability prompted investigations to identify and quantify its sources to optimise study designs of future clinical trials and ultimately support the interpretation of performed PTAs. Hence, inter- and intracatheter variability was compared between an in vitro dynamic µD system (dIVMS) in different matrices and using an ex vivo approach in human subcutaneous ISF of abdominal tissue with different localisations of µD catheter to quantify their impact on the observed in vivo variability of the µD technique.
Methods:
To investigate the impact of different matrices on the observed µD variability, the typical in vivo LIN concentration-time profile (C(t) profile) in ISF [2] was mimicked in a dIVMS using Ringer’s solution (RS) and mimicked artificial ISF (pooled plasma + NaCl 0.9%, 1+1) as in vitro matrices, respectively. Samples were taken from the flask of the dIVMS, which represents the ISF, collected from three different µD catheters (CMA 60, 20 kDa cut-off) simultaneously and retrodialysis was consecutively performed. The experiment was repeated on four different occasions using the same catheters.
To investigate the impact of different localisation of µD catheter only retrodialysis of LIN was performed ex vivo in human subcutaneous ISF of abdominal tissue on two different occasions.
Analysis 1: Using all ISF, µD and retrodialysis in vitro data NLME modelling was performed using NONMEM® (version 7.4.3) to characterise the LIN C(t) profile in the dIVMS and to quantify the different levels of variability. An integrated ISF and micro-/retrodialysis modelling approach was applied to evaluate data from three available sources simultaneously [3].
Analysis 2: For ex vivo retrodialysate data LME modelling was performed using R-package lmer (version 1.1-21) to quantify intercatheter and residual variability (including intracatheter variability).
All quantified variabilities derived from in vitro and ex vivo settings were compared to variability of in vivo clinical data [2].
Results:
The LIN C(t) profile in the dIVMS was successfully described using a one compartmental model with linear clearance for the flask and three parallel µD measurement compartments. Estimated ISF-compartment volume and clearance were 91.1 mL (85.1 – 97.0 mL) and 155 µL/min (133 – 167 µL/min), reflecting the experimental in vitro settings with a flask volume of 100 mL and pump rate of 145 µL/min. Method related variability in the in vitro setting was split up in intra- and intercatheter variability being 2.60%CV (1.97 – 3.23%CV) and 2.80%CV (2.34 – 3.26%CV) in RS and 2.55%CV (2.12 – 2.97%CV) and 4.10%CV (3.40 – 4.80%CV) in artificial ISF, respectively, compared to 27.2%CV (21.8 – 32.0%CV) and 26.1%CV (16.7 – 33.8%CV) in vivo [2].
In the ex vivo setting, delivery related variability was split in intercatheter and residual variability (including intracatheter variability), which were RSD=25.3%CV and 12.3%CV comparable to in vivo (RSD=16.0%CV and 18.1%CV).
Conclusions: The integrated in vitro, ex vivo and model-based analysis approach provided quantitative and qualitative insights into the observed µD technique variability in a clinical µD study. Changing matrices from RS to artificial ISF had a negligible impact on the µD technique variability, whereas different localisations of the inserted µD catheters in the ex vivo setting were in line with the observed in vivo variability. Therefore, inserting the µD catheter at consistently same localisation is crucial to reduce the method related variability for future clinical µD studies.
References:
[1] Committee for Human Medicinal Products (CHMP). Guideline on the use of pharmacokinetics and pharmacodynamics in the development of antibacterial medicinal products. Eur. Med. Agency (2016).
[2] Ehmann L, 26th PAGE, Budapest, Hungary (2017).
[3] Minichmayr IK, Schaeftlein A, Kuti JL, Zeitlinger M, Kloft C. Clinical Determinants of Target Non-Attainment of Linezolid in Plasma and Interstitial Space Fluid: A Pooled Population Pharmacokinetic Analysis with Focus on Critically Ill Patients. Clin. Pharmacokinet. (2016).
Reference: PAGE () Abstr 9457 [www.page-meeting.org/?abstract=9457]
Poster: Drug/Disease Modelling - Infection