Quantitative approaches to drug development

This publication is a collaboration between pharma, consultancies and academia, and outlines a pivotal shift from Model-Informed Drug Development (MIDD) to Model-Integrated Evidence (MIE). In this emerging paradigm, rigorously validated computational models generate “digital evidence” that carries evidentiary weight comparable to empirical clinical data for regulatory purposes, including approvals, labeling, and lifecycle management. The vision is for MIE to be recognized as primary evidence of safety and effectiveness, accelerating R&D and improving treatment options for patients.

Traditional Visual Predictive Checks (VPCs) can fall short when applied to real-world data shaped by adaptive clinical decisions. This new CPT publication from InsightRX and Pharmetheus outlines why standard and prediction-corrected VPCs may mislead, and offers a path forward for more accurate pharmacometric modeling.

This research highlights the model-informed drug development (MIDD) that secured FDA and EMA approval for elafibranor. By integrating data from 17 clinical trials, the team used population PK/PD modeling to confirm that an 80 mg/day dose provides near-maximum efficacy without the need for dose adjustments across patient subpopulations. This work marks a critical advancement for patients with primary biliary cholangitis who require effective second-line therapies.

This publication highlight key advancements from the second Open Systems Pharmacology (OSP) Community Conference, including application of PBBM modeling to bridge in vitro dissolution and in vivio pharmacokinetics, and hybrid approaches combining deep learning with mechanistic PBPK to predict human PK directly from chemical structures. OSP is a community-driven initiative that provides transparent, open-source solutions for model-informed drug development (MIDD).