Flawed assumptions could be spoiling cancer immunotherapy research

Cancer immunotherapy drugs called immune checkpoint inhibitors take the brakes off the body’s immune system, allowing it to find and attack cancer cells. One such checkpoint protein, programmed cell death protein 1 (PD-1), has proved to be a successful immunotherapy target. But this therapy is effective in only some patients, and scientists haven’t known why.

A recent paper shows how preclinical mouse studies may be confounding researchers. The researchers found that a small sequence of amino acids, or motif, that plays a major role in the PD-1 mechanism is present in humans but absent in mice. This difference makes the immune response interactions of PD-1 in mice significantly weaker. Therefore, mice are more responsive to immunotherapy drugs. (Sci. Immunol. 2025, DOI: 10.1126/sciimmunol.ads6295)

“Our study suggests that, at least for the PD-1 pathway, the mouse models cannot accurately predict the human PD-1 [pathway],” says study author Enfu Hui of the University of California San Diego.

The researchers used quantitative assays to compare mouse and human PD-1 biochemistry and function in the same cellular background. They found that mouse PD-1 has weaker interactions than human PD-1, both inside and outside the cell. The extracellular domain is responsible for binding with PD-1’s partner protein on the tumor cell, while the intracellular one binds to an effector molecule that regulates immune cell function.

When the researchers replaced rodent PD-1 in mouse models with human PD-1, they observed that the immune system was less able to attack tumor cells.

“Preclinical studies [of immunotherapy drugs] in mice often show very striking results,” Hui says. “But the translation from mouse to human has a very low success rate. . . . The models we’re relying on right now might have limited power to predict the [outcome for] human patients.”

This study tested only functional significance in a single tumor model, the melanoma. To better understand the full implications of the difference in human and mouse PD-1, the researchers would like to test other cancer models.