A CAR-T therapy is, at the patent level, a claim to an engineered receptor, and US11564945B2 - "Chimeric antigen receptor and use thereof," issued January 31, 2023 to Nanjing Legend Biotech Co., Ltd. - claims one such construct. Reading it means parsing the modular architecture, not the marketing.
A chimeric antigen receptor is built in modules: an extracellular domain that binds the tumor antigen, a hinge and transmembrane region, and intracellular signaling domains (typically a costimulatory domain plus the CD3-zeta activation domain). The claim recites a specific combination of these modules. The CPC tags - C07K 14/7051 (T-cell receptor / CAR components), A61K 35/17 (T-cell therapy) - place it squarely in the cell-therapy estate.
The claim-scope question is which modules, in which arrangement. 'A CAR' is not patentable - the concept is decades old. What is claimable is the particular antigen-binding domain, the particular costimulatory choice, the particular construct. A competitor's CAR infringes only if its module combination falls within the recited architecture or its equivalents.
This is why describing the patent by its title - 'chimeric antigen receptor' - tells a reader almost nothing about scope. The enforceable boundary is the specific binder and signaling-domain selection. Legend's position (the assignee behind the BCMA-directed cilta-cel program) rests on its particular construct, and the construct details are where any infringement or validity fight lives.
For the landscape, the CAR estate is dense with construct-level claims because the modular design invites endless variation - swap the binder, change the costimulatory domain, and you have arguably a new construct. That modularity is both the field's engineering strength and the reason freedom-to-operate requires construct-by-construct analysis rather than a single platform license.