Whole-Body Physiologically Based Pharmacokinetic Modeling Framework for Tissue Target Engagement of CD3 Bispecific Antibodies

Background : T-cell-engaging bispecific (TCB) antibodies represent a promising therapy that utilizes T-cells to eliminate cancer cells independently of the major histocompatibility complex. Despite their success in hematologic cancers, challenges such as cytokine release syndrome (CRS), off-tumor toxicity, and resistance limit their efficacy in solid tumors. Optimizing biodistribution is key to overcoming these challenges. Methods : A physiologically based pharmacokinetic (PBPK) model was developed that incorporates T-cell transmigration, retention, receptor binding, receptor turnover, and cellular engagement. Preclinical biodistribution data were modeled using two TCB formats: one lacking tumor target binding and another with target arm binding, each with varying CD3 affinities in a transgenic tumor-bearing mouse model. Results : The PBPK model successfully described the distribution of activated T-cells and various TCB formats. It accurately predicted preclinical biodistribution patterns, demonstrating that higher CD3 affinity leads to faster clearance from the blood and increased accumulation in T-cell-rich organs, often reducing tumor exposure. Simulations of HER2-CD3 TCB doses (0.1 µg to 100 mg) revealed monotonic increases in synapse AUC within the tumor. A bell-shaped dose-Cmax relationship for synapse formation was observed, and Tmax was delayed at higher doses. Blood PK was a reasonable surrogate for tumor synapse at low doses but less predictive at higher doses. Conclusions : We developed a whole-body PBPK model to simulate the biodistribution of T-cells and TCB molecules. The insights from this model provide a comprehensive understanding of the factors affecting PK, synapse formation, and TCB activity, aiding in dose optimization and the design of effective therapeutic strategies.