Henning Schmidt
Novartis Pharma AG
Objectives: The main goal of quantitative approaches to drug development is the integration of all available information in order to achieve high confidence in compound and dose selection for Phase III and beyond.
Pharmacokinetic and pharmacodynamic modeling and simulation aim at supporting drug development by characterizing the important relationships between dose, concentration, and clinically relevant response levels. Once models are available and adequately capture the signature of the clinical data, they can be used to efficiently address various questions, e.g., about optimal dose and dosing regimen, profiling against competitors, and study design. Traditional nonlinear mixed-effect modeling, however, is often time consuming and anecdotal evidence indicates that lack of meeting the time constraints often limits the extent of pharmacometric involvement in the industry [1].
Methods: In this work, we propose a methodology for efficient characterization of dose-concentration-response relationships on a population level. Instead of considering individual level patient data, summary level data are used. Inter-individual variability can be taken into account by bootstrapping and covariates can be considered based on stratification of the individual level patient data. The methodology is implemented in the form of a module within the SBPOP package [2], providing a user-friendly and well-documented framework for model building and trial simulation.
Results and Conclusions: The approach was applied in support of a Phase IV study design, dose-concentration-response analysis of summary level competitor data, and support of a submission project in which a multitude of different endpoints had to be assessed. Available examples show that parameter estimates for fixed effects essentially agree with the ones obtained from nonlinear mixed-effect modeling, but time for model development was only a small fraction of typical NLME modeling efforts.
References:
[1] Gobburu JVS (2010) Pharmacometrics 2020, J Clin Pharmacol, 50:1-51S–157S
[2] Schmidt H (2013) SBPOP Package: Efficient support for model based drug development – from mechanistic models to complex trial simulation, PAGE meeting, Glasgow, UK, [http://www.page-meeting.eu/default.asp?abstract=2670], [http://www.sbtoolbox2.org]
Reference: PAGE 24 (2015) Abstr 3444 [www.page-meeting.org/?abstract=3444]
Poster: Methodology - New Modelling Approaches