GliaDose: An alternative approach for clinical CAR-T dose prediction via PBPK–glioblastoma-on-a-chip integration

Abstract Glioblastoma (GBM) is a highly aggressive brain tumor with poor prognosis and limited treatment options. Chimeric antigen receptor T cell (CAR-T) therapy has emerged as a promising immunotherapy for GBM. However, unlike conventional drugs, CAR-T cells are living therapeutics that exhibit non-linear pharmacokinetics, expansion, and complex tissue trafficking, making conventional animal-to-human dose extrapolation unreliable. Furthermore, species-specific differences and the vasculature limit the translational relevance of preclinical models. To address these challenges, we developed GliaDose, a hybrid framework that integrates physiologically based pharmacokinetic (PBPK) modeling with a vascularized glioblastoma-on-a-chip (GoC) platform to support quantitative first-in-human dose (FIHD) prediction for CAR-T therapy. The PBPK model was parameterized exclusively with human physiological data and incorporates a brain-tumor compartment to explicitly capture CAR-T transport across the vasculature. Complementarily, the GoC platform recapitulates key features of the GBM microenvironment, including three-dimensional tumor architecture, extracellular matrix, and perfusable human vascular networks, enabling assessment of CAR-T transmigration under flow conditions. Integration of GoC-derived efficacy data into the PBPK framework enabled quantitative dose–response predictions and identified vascular barriers as critical determinants of CAR-T tumor access. GliaDose provides a human-relevant, mechanistic alternative to animal-based dose selection and may improve FIHD prediction and personalized CAR-T therapy for GBM.