Immunotherapy has transformed cancer care, offering a way to harness the body’s own immune system to fight tumors. But even with these advances, many patients still don’t respond. One major barrier to effective cancer immunotherapy is T cell exhaustion. Although immune checkpoint blockade can reinvigorate exhausted T cells, not all patients achieve long-term responses, partly due to the refractory nature of terminally exhausted T cells. However, the environmental drivers of exhaustion remain to be thoroughly characterized.

Now, a new study has discovered a molecular signal that tumors exploit to exhaust the T cells meant to destroy them—and how silencing that signal could revive the body’s immunity. The study shows that tumors not only evade the immune system but can actively suppress immune cells.

“Our findings reveal a completely new way that tumors suppress the immune system,” said Jedd Wolchok, MD, PhD, the Meyer Director of the Sandra and Edward Meyer Cancer Center, professor of medicine at Weill Cornell and an oncologist at New York-Presbyterian/Weill Cornell Medical Center. “By blocking this pathway, we can help exhausted T cells recover their strength and make existing immunotherapies work better for more patients.”

This work is published in Nature Immunology in the paper, “Thrombospondin-1–CD47 signaling contributes to the development of T cell exhaustion in cancer.”

“Our dream is to make immune-based therapies available to every patient. To overcome resistance, we must unlock the power of exhausted T cells, reviving them to destroy cancer,” said Taha Merghoub, PhD, professor, deputy director of the Meyer Cancer Center, and co-director of the Parker Institute of Cancer Immunotherapy at Weill Cornell Medicine. “This discovery moves us closer to a future where the immune system itself defeats tumors.”

T cell exhaustion is a phenomenon triggered by continued exposure to chronic infections or tumors that activate the immune system. The embattled immune cells remain capable of recognizing their foes, but they no longer attack them. “Although such cellular surrender may seem counterproductive, it serves as a brake to protect against out-of-control inflammation and sepsis,” Merghoub said.

The researchers sought to investigate whether CD47 on cancer cells contributed to T cell exhaustion. They identified CD47 upregulation in tumor-infiltrating exhausted CD8+ T cells in both human and murine tumors. “When T cells are activated, they express CD47. And when they get exhausted, they increase CD47 to very high levels,” Merghoub said.

Mice lacking CD47 had delayed tumor growth, suggesting that CD47 on the animal’s immune cells—not the CD47 on the cancer cells—was causing exhaustion. The team hypothesized that eliminating CD47 on T cells could be beneficial. When tested in mice with melanoma, T cells lacking CD47 were better at fighting the cancer than T cells, which had CD47 intact.

The researchers turned their attention to how the cancer cells co-opted T-cell CD47 to promote exhaustion. They focused on thrombospondin-1 that interacts with CD47 and is produced by metastatic cancer cells. When they tested mice lacking thrombospondin-1, they found that T cells were less exhausted. “That was the real eureka moment,” said Merghoub. “It showed us that CD47 and thrombospondin are clearly key players because eliminating either one gives you the same effect.”

More specifically, the authors write that the work reveals “a novel role for the extracellular matrix protein thrombospondin-1 (TSP-1) in engaging CD47 on T cells to promote exhaustion.”

They used the TAX2 peptide, which was designed to selectively disrupt the interaction between CD47 and thrombospondin-1 in their mouse tumor models, as a proof of concept to confirm that disrupting the crosstalk between TSP-1 and CD47 prevents T cell exhaustion in mice with tumors. The results showed that TAX2 preserved T-cell function and slowed down tumor progression in mice with melanoma or colorectal tumors.

The T cells in treated mice stayed more active, produced more immune boosting cytokines, and were better at infiltrating tumors. In addition, TAX2 worked in synergy with PD-1 immunotherapy in controlling colorectal tumor growth.

The interaction, the authors note, “activates calcineurin–NFAT signaling, inducing upregulation of TOX and expression of inhibitory receptors, and impairing effector function during tumor progression.”

Targeting this interaction would be a valuable therapeutic on its own, but it could also help preserve tumor-reactive T cells in patients who would otherwise develop resistance to the currently available T cell-based immune checkpoint inhibitors. Even more promising, experiments in preclinical tumor models suggest that blocking both PD1 and CD47 produces T cells that are more effective at killing cancer cells, Merghoub said. “We plan to explore this therapeutic angle.”