AI Insight
Researchers developed an AI-driven framework that combines high-resolution cellular atlases, public databases, and large language models to identify new targets for CAR T cell therapy. This approach identified GPNMB as a promising target antigen that is expressed across multiple cancer types. GPNMB-targeted CAR T cells demonstrated strong anti-tumor activity in multiple mouse models, suggesting potential as a broad-spectrum cancer immunotherapy.
Why it matters
This work demonstrates how AI can accelerate the discovery of cancer immunotherapy targets and potentially expand CAR T cell treatment beyond blood cancers to solid tumors. The identification of GPNMB as a multi-cancer target could lead to a single therapy effective against several different cancer types, reducing development time and costs.
Understand the Science
An AI-guided framework integrating high-resolution cellular atlases, public knowledge databases, and large language models enables a modular framework for CAR T cell target discovery. GPNMB emerged as a versatile candidate antigen, enabling the development of GPNMB CAR T cells with potent activity in multiple xenograft mouse models of cancer.
Source: AI-driven discovery of GPNMB CAR T cells as a multi-cancer therapy