Model-based Analysis with Mechanistic Insights to CAR-T Cell Therapy Kinetics: Case Study with Axicabtagene Ciloleucel and Brexucabtagene Autoleucel

In CAR-T cell therapy modeling, drug developers often navigate between two extremes: empirical piecewise-linear models that fit observed clinical data well but lack mechanistic depth for extrapolations and simulations, and complex quantitative systems pharmacology (QSP) models that require granular data rarely collected in late-stage trials. Bridging this gap requires a framework with sufficient mechanistic depth to support extrapolations and simulations while remaining parsimonious enough to be identified using routine total CAR-T cell measurements. To address this challenge, the authors developed a novel mechanism-based population pharmacokinetic (PK) framework.

Key highlights:

• Parsimonious structure: Distilled down to just seven structural parameters, relying only on post-infusion CAR-T cell peripheral blood concentrations and routinely collected clinical covariates.
• Mechanistic foundation: Combines two CAR-T cell compartments with a latent kinetic-pharmacodynamic tumor compartment, replacing empirical piecewise curves with a continuous, biologically driven approach.
• Robust clinical evaluation: Developed and evaluated using a comprehensive database of 473 patients receiving axicabtagene ciloleucel (axi-cel) or brexucabtagene autoleucel (brexu-cel) across multiple lymphoma indications.
• Covariate insights: Identified five significant covariates impacting PK, with product type/mantle cell lymphoma (MCL) disease type showing a clinically relevant impact on predicted CAR-T cell exposure.

This framework provides a practical platform to facilitate extrapolations and simulations (such as varying doses or patient characteristics), better supporting model-informed drug development decisions for future CAR-T cell therapy programs.