A Neuro-Dynamic Quantitative Systems Pharmacology (QSP) Model to Describe Pathophysiology of Alzheimer's Disease and Inform Drug Development

Authors

Youfang Cao, PhD – Director of Pharmacometrics, Translational Sciences, Eisai Inc; Brian Willis, PhD – Senior Director, Head of Pharmacometrics, Translational Sciences, Eisai Inc; Pallavi Sachdev, PhD – Executive Director, Translational Sciences, Eisai Inc; Natasha Penner, PhD – Senior Director, Translational Sciences, Eisai Inc; Kristin Wildsmith, PhD – Senior Director, Translational Sciences, Eisai Inc; Arnaud Charil, PhD – Senior Director, Head of Translational Imaging, Translational Sciences, Eisai Inc; Kanta Horie, PhD – Deputy Head, Discovery Concept Validation, Eisai Inc; Akihiko Koyama, PhD – Function Head, Discovery Concept Validation, Eisai Inc; Larisa Reyderman, PhD – VP, Translational Sciences, Eisai Inc

Abstract

Objectives: Lecanemab is an anti-amyloid monoclonal antibody and was approved as a disease-modifying treatment with demonstrated significant clinical benefit in slowing cognitive decline in early Alzheimer's disease (AD). Lecanemab selectively binds soluble amyloid-? (A?) protofibrils, which is the neurotoxic amyloid species that modulates tau pathology. A neuro-dynamic QSP model was developed to provide a computational framework that integrates data on lecanemab mechanism of action with disease biology.

Methods: The QSP model incorporated 3 interlinked modules: A? pathway, tau pathway, and cognitive decline, where A? triggers tau pathology leading to cognitive decline due to neurodegeneration. Mechanisms within each module and their interlinks were informed by literature evidence, preclinical and clinical data. The model was developed using nonlinear mixed-effect modeling of a multidimensional dataset (N=4056) consisting of cognitive scores (CDR-SB and ADAS-Cog), imaging (amyloid and tau PET), and biomarker (plasma A?42/40 ratio and tau phosphorylated at residue 181 [p-tau181]) data from lecanemab Phase 2 and Phase 3 trials, as well as amyloid PET, CDR-SB and ADAS-Cog data from the Alzheimer's Disease Neuroimaging Initiative (ADNI). The model was validated with data from other anti-amyloid antibodies (aducanumab and gantenerumab). Simulations were conducted to investigate long-term lecanemab benefits of monthly (Q4W) maintenance dosing (MD) following initial biweekly (Q2W) treatment.

Results: The final QSP model has 11 differential equations and 74 parameters, which were estimated with reasonable precisions (RSE<=30%). The model describes amyloid pathway as an aggregation cascade from monomers to oligomers, protofibrils, and plaques, and the tau pathway as a multi-staged progression of AD pathology with different levels of misfolded tau aggregates leading to formation of neurofibrillary tangles (NFT). High NFT neurons spread tau seeds, which drive tau pathology cycle. The QSP model was able to describe the pathological progression of imaging and plasma biomarkers and cognitive scores during 3 decades of AD progression. Simulations showed that lecanemab Q4W MD following the 18-month Q2W treatment continues to reduce A? protofibrils and plaques inhibit progression of amyloid and tau pathologies, as measured by plasma p-tau181 and A?42/40 ratio, suppress tau PET, and slow cognitive decline.

Conclusions: A neuro-dynamic QSP model was developed based on clinical data as a modeling framework for AD pathophysiology and clinical outcomes. QSP simulations demonstrated benefits of continuing lecanemab treatment after plaque removal and support Q4W MD following initial Q2W treatment.


Citations: none

Keywords

Neuro-dynamic QSP model, Alzheimer's disease

Date of Conference

November 10-13, 2024

Conference Location

Phoenix, Arizona, USA

DOI

10.70534/YQMT6293