(95b) Predictive and Mechanistic Pharmacokinetic Simulation of Drug Heterogeneity

The time, effort, and cost associated with experimentally characterizing new therapeutic, diagnostic, and imaging agents has driven the push for robust computational methods for in vivo drug distribution modeling. However, the complexity associated with modern drug classes (e.g. ‘Beyond Rule of 5’ compounds, antibody drug conjugates, etc.) and differences in physiology of the target tissue have made generalizing pharmacokinetic models difficult. Outside of the life sciences, chemical engineers have used dimensionless numbers, such as the Damköhler number, for decades to simplify complex models in fluid dynamics and reactor design, a strategy that has also proven effective when applied to pharmacokinetic behavior. Here, a general simulation with a graphical user interface was developed to model the distribution of biologic, small molecule, and peptide drugs in tissues. This program leverages theoretical models that have divided drugs into four different classes based on their distribution patterns, and correlations between molecular and drug transport properties such as permeability and diffusivity. The program can be used to predict tumor drug distribution profiles a priori, and design and interpret in vivo experiments with limited information; it can also incorporate specific parameters from experimental data for more accurate simulations. Parameters for two drugs with heterogeneous tissue distribution, trastuzumab and Hoechst 33342, were used for validation, resulting in matching experimental data, including their dependence on permeability and diffusivity respectively. This program is able to guide experimental design, aid in data interpretation from preclinical models, and help scale results to the clinic to help design more effective novel therapeutics.