CELL INTELLIGENCE · 肿瘤细胞治疗研究

Gastrointestinal (GI) malignancies account for approximately one-quarter of new cancer diagnoses and more than one-third of cancer-related deaths worldwide, yet as of August 2026, only one CAR-T therapy has received regulatory approval for a solid tumor indication anywhere in the world. The recent CT041-ST-01 phase II trial of satricabtagene autoleucel, the first randomized CAR-T trial conducted in a solid tumor, demonstrated a significant improvement in progression-free survival for patients with advanced gastric cancer (median 3.25 vs. 1.77 months; hazard ratio (HR) 0.37, p < 0.001). While this landmark study established the clinical feasibility of CAR-T therapy in solid tumors, it also underscored the biological barriers that continue to limit durable responses. GI tumors are characterized by heterogeneous antigen expression, dense desmoplastic stroma, inefficient immune-cell trafficking, profoundly immunosuppressive tumor microenvironments, and progressive T-cell dysfunction, all of which are further compounded by the logistical and economic challenges of autologous cell manufacturing. In this narrative review, we organize these obstacles within a unified four-barrier engineering framework and critically examine the strategies being developed to overcome each of them. We discuss advances in multi-antigen and logic-gated CAR architectures, stromal remodeling through fibroblast activation protein (FAP)-targeted approaches and extracellular matrix-degrading enzymes, chemokine receptor engineering, regional delivery, hypoxia-responsive CARs, cytokine-armored and persistence-enhanced constructs, dominant-negative and switch receptors, metabolic reprogramming, and intrinsic checkpoint disruption. We also review emerging manufacturing platforms, including allogeneic CAR-T and CAR-natural killer (CAR-NK) cells, induced pluripotent stem cell-derived products, CAR-macrophages, and in vivo CAR generation, together with engineering strategies designed to improve safety and scalability. Rather than relying on a single technological advance, the future of CAR-based therapy for GI malignancies will likely depend on integrating multiple engineering approaches to address the diverse biological barriers within the tumor microenvironment. By synthesizing current preclinical and early clinical evidence, this review provides a translational framework for the next generation of CAR-based cellular therapies in gastrointestinal oncology.