Andrew Sage - Assistant Professor, Status

Area of Research

The Sage Lab at UHN is pioneering the use of digital twins integrated with ex vivo organ perfusion platforms to model human disease and transform preclinical drug development. By maintaining donated human organs outside the body under physiological conditions, the lab creates a unique testing ground for understanding organ function, disease mechanisms, and therapeutic responses in a setting that closely mimics the human body. These real-world data are used to train and refine computational “digital twins” — dynamic, data-driven virtual replicas of organs — that can simulate complex biological processes and drug interactions.

Primary Research Challenge

A central challenge in biomedical research is the poor translational fidelity of traditional preclinical models. Animal systems and static cell cultures often fail to replicate the nuanced physiology of human organs, resulting in high failure rates when drugs move from lab to clinic. Additionally, understanding how specific diseases manifest and evolve at the organ level — and how individual patients might respond to treatment — remains difficult without tools that integrate biology, engineering, and computational modeling.

Proposed Solution

To address these gaps, the lab combines ex vivo human organ perfusion with the development of high-fidelity digital twins that can simulate disease progression and therapeutic effects. These twins are calibrated using rich data from perfused organs — including functional, molecular, and imaging outputs — enabling a continuous feedback loop between in vitro experimentation and in silico modeling. This integrated platform allows researchers to mimic human disease states, test candidate drugs, and predict therapeutic outcomes with greater accuracy than traditional approaches. It also provides a scalable framework for evaluating interventions across diverse disease conditions and patient populations.

Impact to Date

The lab’s digital twin–enabled ex vivo platform has already demonstrated improved predictive accuracy in modeling disease and drug response. Early work has shown its potential to identify better biomarkers, simulate personalized therapeutic strategies, and reduce uncertainty in preclinical decision-making. This work forms the foundation for more personalized and efficient drug development.

Keywords: ex vivo organ perfusion, ex vivo lung perfusion, biomedical engineering, computer science, artificial intelligence, machine learning, drug development, disease modelling, preclinical drug discovery