Pauline Thémans (1), Joseph J. Winkin (1) and Flora T. Musuamba (2)
(1) Namur Institute for Complex System (naXys) and Department of Mathematics, University of Namur, Belgium, (2) U850 INSERM, Université de Limoges, France
Objectives: To develop a reduced physiologically-based pharmacokinetic (PBPK) model using a retrograde approach based on a previously developed compartmental pharmacokinetic (PK) model in patients with severe nosocomial pneumonia and to propose an approach for individualized drug dosing. A control strategy was designed to achieve and maintain plasma and epithelial lining fluid (ELF) concentrations at target levels.
Methods: Steady-state PK data were obtained from 60 adult patients diagnosed with severe lung infection [1]. They included rich plasma and sparse ELF samples. Previously, a two-compartment model with an additional compartment for the site of effect (ELF) was successfully fitted to the data [6]. A reduced PBPK model was developed using NONMEM version 7.3. Model qualification was based on successful numerical convergence, adequate precision in parameter estimates (as assessed by bootstraps), acceptable goodness of fit plots with no indication of bias, and acceptable performance of visual predictive checks (VPCs). External validation was performed by fitting the model to external independent data. Data from two previously published studies on meropenem were used for external validation [3], [2]. Graphical analysis was performed using R version 3.0.2 and MATLAB 2014b.
The analytical expression of the asymptotic response (i.e. when time goes to infinity) was determined and used to derive a closed-form formula designed to estimate the effective dosing regimen, given the patient’s characteristics and the target minimal concentration (open-loop control method). The system asymptotic response corresponds to the pharmacokinetic steady-state. This method was studied and validated with MATLAB 2014b.
Results: A reduced PBPK model with six compartments (arterial and venous pools, lungs, liver, kidneys and rest of the body) provided an adequate fit to the data. The model was successfully qualified internally and externally as described above. Numerical simulations showed that the system output trajectory converges exponentially towards the asymptotic response, so that the asymptotic response (steady-state) is a good approximation of the actual response after few dosing intervals. Numerical simulations were performed to produce pharmacokinetic profiles using the open-loop dosing approach. Simulations of concentrations in ELF for average virtual patient receiving the recommended maintenance dose (1.47 grams) for the susceptibility breakpoint of 2 mg/L show that 50% of the simulated patients are above the target concentration for 100% of the dosing interval; 75% reach the target concentration for at least 80% of the dosing interval (this value is an approximation, as exact value changes at each iteration).
Conclusions: A dosing strategy based on a formula in a closed-form was proposed which displayed acceptable and reliable results. The established formula can be readily generalised to any n-compartment model.
However, its open-loop nature does not take into account unexplained interindividual variability and only provides a result for the average patient. The proposed PBPK model can alternatively be integrated in other existing tools enabling dosing optimization such as Maximum A posteriori Bayesian estimations [4], [5].
References:
[1] F. Frippiat, F. T. Musuamba, et al. Modelled target attainment after meropenem infusion in patients with severe nosocomial pneumonia: The PROMESSE study. Journal of Antimicrobial Chemotherapy, 70(1):207-216, Sep 2015.
[2] J. Goncalves-Pereira et al. Assessment of pharmacokinetic changes of meropenem during therapy in septic critically ill patients. BMC Pharmacology and Toxicology, 15(1):15-21, 2014.
[3] J. Karjagin et al. Pharmacokinetics of meropenem determined by microdialysis in the peritoneal fluid of patients with severe peritonitis associated with septic shock. Clinical Pharmacology & Therapeutics, 83(3):452-459, 2008.
[4] D. R. Mould, and R. N. Upton. Basic concepts in population modeling, simulation, and model-based drug development-Part 2: Introduction to pharmacokinetic modeling methods. CPT: pharmacometrics & systems pharmacology, 2(4):1-14, April 2013.
[5] THT Nguyen et al. Model evaluation of continuous data pharmacometric models: metrics and graphics. CPT: pharmacometrics & systems pharmacology, 6(2):87-109, 2017.
[6] P. Thémans, F. T. Musuamba, and J. Winkin. Modeling, analysis and control of pharmacokinetic systems: application to meropenem. Submitted
Reference: PAGE 27 (2018) Abstr 8434 [www.page-meeting.org/?abstract=8434]
Poster: Drug/Disease Modelling - Infection