Model-Informed Drug Development (MIDD) is having a profound impact on contemporary pharmacological innovation. It provides essential tools for optimising clinical trial design, selecting appropriate dosing, and evaluating the balance of benefits and risks, all with an inherent focus on sustainability.
One significant sustainability benefit lies in the reduction of resource consumption and waste. Advanced modelling techniques help prioritise drug candidates more effectively, allowing fewer compounds to move forward into clinical trials, which are notoriously resource-intensive and the most significant expense in developing a new drug. MIDD also addresses the ethical and environmental concerns surrounding animal testing. Through in silico simulations, researchers can evaluate drug behaviour, safety, and efficacy with greater precision and less reliance on animal models. These approaches align with evolving regulatory perspectives, such as the FDA’s encouragement in leveraging non-animal testing methods, particularly in developing biologics like monoclonal antibodies [PMID: 40135941].
Importantly, models can help scientists understand the underlying ADME characteristics of a drug, shedding light on mechanisms that are hard to test experimentally or that yield fuzzy results. This is particularly valid for mRNA-encoded monoclonal antibodies. Indeed, despite significant achievements in the field, the underlying biological steps that take the mRNA from being complexed with lipid nanoparticles (LNP) to become a functional antibody, remain poorly understood.
To address this gap, we developed an ODE-based model that describes the principal events occurring after intravenous (IV) injection in mice of mRNA-LNPs based on the most recent findings in the literature, namely their adsorption in the liver, the complex uptake process at the hepatocyte level, their escape from the endosomes and finally, their translation. This mechanistic layer is also equipped with a Physiologically Based Pharmacokinetic (PBPK) model, based on the work of Sepp et al. [PMID: 31079322], which describes the kinetics of mRNA-mAbs throughout 15 different organs and tissues, following an approach that proved to be successful in [PMID: 40677727]. Our mechanistic approach allows us to pinpoint bottlenecks and the most crucial steps to optimise, while developing these therapeutics.
To fit the unknown parameters in the model, we leveraged three preclinical studies in mice that report the concentration time-profile of mRNA-mAbs of different sizes, with or without the Fc region, all delivered with LNPs [PMID: 36403209, PMID: 35069904, PMID: 38776391] and targeting cancer. Our multi-scale PBPK model accurately predicts the concentration-time profiles of both the mRNA-encoded therapeutics and the relative recombinant proteins.
The model was also validated on unseen data presented in the reference literature [PMID: 36403209, PMID: 35069904, PMID: 38776391], demonstrating high accuracy across different dosing schedules and dosages.
Our model can predict mAbs disposition in remote tissues, which experimentally would require extensive animal testing (and ultimately, sacrifice), and their kinetics in different animal models. Moreover, its inherent modularity enables the exploration of various routes of administration and tropisms of the mRNA-LNPs therapeutic of interest, offering valid support to developing these ground-breaking therapies.