[SS09] Enhancing Drug Development and Patient Care through Model Informed Drug Development

[SS09-02] Clinical Trial Simulation with Quantitative Systems Pharmacology Virtual Patients Model

*Ryuta Saito¹ (1. Mitsubishi Tanabe Pharma Corporation (Japan))

Keywords:

Model-informed Drug Discovery and Development,Quantitative Systems Pharmacology,Artificial Intelligence,Cardiovascular diseases,Virtual Patients Generation

The pharmaceutical industry has increasingly adopted Model-Informed Drug Discovery and Development (MID3) to enhance productivity in drug discovery and development. Quantitative Systems Pharmacology (QSP), which integrates drug action mechanisms and disease complexities to predict clinical endpoints and biomarkers is central to MID3. QSP approaches are widely applied to address various essential questions in drug discovery and development, such as identification of the mechanism of action of a therapeutic agent, patient stratification, and the mechanistic understanding of the progression of disease. As an example of prospective simulation of late clinical development, a use case for the risk assessment of hyperkalemia in patients with diabetic nephropathy treated with mineralocorticoid receptor antagonists (MRAs, renin-angiotensin-aldosterone system inhibitors) is shown in this presentation. A QSP model for generating virtual patients with diabetic nephropathy was used to quantitatively assess that the nonsteroidal MRAs, finerenone and apararenone, have a lower risk of hyperkalemia than the steroidal MRA, eplerenone. Prospective simulation studies using a QSP model are useful to prioritize pharmaceutical candidates in clinical development and validate mechanism-based pharmacological concepts related to the risk-benefit, before conducting large-scale clinical trials. As a technical study for MID3, a development of artificial intelligence (AI) model for virtual patients (VPs) generation based on QSP modeling is also shown in this presentation. Despite the advantages of clinical trial simulation with QSP modeling, QSP model validation through clinical trial simulations using VPs is challenging because of parameter variability and high computational costs. To address these challenges, we proposed a hybrid approach that combines Bayesian optimization with machine learning for efficient parameter screening. Our approach achieved an acceptance rate of 27.5% in QSP simulations, which is in sharp contrast with the 2.5% rate of conventional random search methods, indicating more than ten-fold improvement in efficiency. By facilitating faster and more diverse virtual-patient generation, this method promises to advance clinical trial simulations and accelerate drug development in pharmaceutical research.