Cancer and Immunity · global
Brief exposure to tumor signals gives natural killer cells stronger anticancer activity in mice
The same signal can weaken immune cells with continuous stimulation, yet brief exposure can help cultivate cells better able to enter tumors. The study offers new clues for cell therapies for solid cancers, but efficacy in humans remains some way from being established.
To fight solid tumors, immune cells must first penetrate tissue occupied by cancer cells and withstand surrounding signals that suppress immune responses. Stanford Medicine and its collaborators have found a way to cultivate natural killer (NK) cells that makes them better able to enter tumors and slows tumor growth in mouse experiments. The research suggests that a key to effective cell activity may lie in how and for how long cells receive signals during cultivation.
The study was published in May 2026 in *Science Translational Medicine*. The team obtained circulating NK cells from human blood donors and attempted to induce them in vitro to become “tissue-resident” cells adapted to the tissue environment. These cells vary in function: some have strong killing activity, while others tend to suppress immune responses. Their ability to reside in tissue alone does not indicate their anticancer capacity.
One source of this difference is transforming growth factor beta (TGF-β). The researchers found that this signal helps induce tissue-resident characteristics, but continuous exposure reduces the cells’ killing function. In contrast, allowing NK cells to briefly make direct contact with human epithelial tumor cells and receive a transient signal from active TGF-β produced cells with stronger killing activity. Simply placing the two cell types near each other was insufficient, suggesting that direct contact also provided other necessary stimuli.
Through multiomics analysis, the team identified a highly cytotoxic population bearing the CD39 marker; these cells also expressed CD49a and CD103. Compared with conventionally activated peripheral blood NK cells, they were better able to kill tumor cells, enter the tumor environment and control tumor growth. The study suggests that these advantages may stem partly from CD103 expression and the cells’ ability to bind to tumor targets; CD39 served here as a marker for identifying this population.
The experiments progressed from tumor organoids in culture dishes to mice bearing human melanoma or head and neck squamous cell carcinoma tumors. The cultivated NK cells could enter the organoids and slow tumor growth in mice. When used with cetuximab, a monoclonal antibody that helps tag cancer cells, a single combined administration suppressed tumors more effectively over a one-month observation period than either therapy alone. However, suppressing growth does not mean eliminating cancer, nor does it directly imply that patients would live longer.
These cells also offer the possibility of batch production. The Stanford team estimated that NK cells from one donor could be cultivated into 20 treatment doses in about two weeks and envisioned cryopreserving them for later use. This remains a production estimate proposed by the research team; whether product quality can remain consistent, whether the cells retain their function after storage and how long they can persist in humans all require further validation.
In a June account of the research, Stanford said the team was planning a phase 1 trial of the combination therapy for advanced squamous cell carcinoma. Launching the trial still requires authorization from the U.S. Food and Drug Administration, and it should not be regarded as already enrolling participants. Although no obvious adverse effects were observed in the mouse experiments, this is insufficient to establish safety in humans. The significance of this work lies in identifying a way to cultivate and expand NK cells with tumor-infiltrating capabilities in vitro; whether it can become a treatment option is a question that human trials must answer next.