CAR-T cell therapy changed blood cancer treatment. Solid tumors have remained largely out of reach. A new study in Cell00269-2) suggests the problem was the wrong target.
In mouse models, the therapy shrank ovarian, lung, and pancreatic tumors — with durable remission in the ovarian model, where surviving mice resisted new tumors when researchers reintroduced cancer cells weeks later. The MSK team engineered CAR T cells to target uPAR, a surface protein found on both cancer cells and the stromal support cells that surround them. The engineered cells eliminated not just tumor cells but also uPAR-positive fibroblasts and immunosuppressive myeloid cells that make up the protective tumor microenvironment. That dual targeting — hitting both the cancer and its support system simultaneously — is what sets this approach apart from the approved CAR-T paradigm, which targets a single antigen on cancer cells alone.
The uPAR finding is notable because it represents a different strategic angle than existing CAR-T approaches. Approved CAR-T therapies target proteins present on cancer cells themselves — CD19 on B cells, BCMA on plasma cells. Solid tumors have been resistant partly because they lack a similarly accessible and specific surface marker. uPAR is elevated across 12 of 14 cancer types and most strongly associated with p53 and KRAS mutations — the same pathway alterations that drive many of the hardest-to-treat solid tumors. By targeting uPAR on both tumor cells and stromal support cells, the MSK approach sidesteps the antigen heterogeneity problem that has made solid tumors difficult to address with CAR-T.
The cell type that has drawn the most attention is the cancer-associated fibroblast. CAFs are the dominant structural cells in many solid tumors and are partly responsible for the physical barrier that excludes immune cells. uPAR is expressed on CAFs, and the MSK data showed that eliminating uPAR-expressing CAFs disrupted the immunosuppressive microenvironment — the same immunosuppressive niche that has blocked previous CAR-T approaches in solid tumors. In the ovarian cancer model, surviving mice resisted new tumors when researchers reintroduced cancer cells weeks later, suggesting durable immune memory.
This is pre-clinical data. The road from a Cell paper to an approved therapy is measured in years and littered with failures that looked equally compelling at the pre-clinical stage. The MSK team has not yet published data from a human trial. What they have published is a biological mechanism that appears to work differently from existing CAR-T approaches and a pre-clinical dataset that gives the field a reason to test it.
The uPAR finding has a secondary implication worth noting. That breadth — 12 of 14 cancer types expressed uPAR in the MSK screening panel — suggests this is not a strategy for one specific indication but potentially a platform applicable across multiple solid tumor types. The same logic that made CD19 CAR-T a platform rather than a drug for one cancer could apply here.
Whether the Cell data translates is the question the field will spend the next decade answering.
