Expression of a certain protein on two types of immune cells could help predict which cancer patients will respond to a specific immunotherapy, a study shows.

The study, “Host expression of PD-L1 determines efficacy of PD-L1 pathway blockade–mediated tumor regression,” was published in The Journal of Clinical Investigation.

One of the primary immunotherapies used in cancer care are immune checkpoint inhibitors, which block proteins that prevent the immune system from effectively recognizing and killing cancer cells.

Antibodies that target two of these proteins — programmed death–ligand 1 (PD-L1) and programmed death protein 1 (PD-1) — are among these immunotherapies. But this approach only controls tumors in about 25 percent of patients across different types of cancers.

It’s still not clear why some patients respond to treatment and others don’t. Studies have shown that clinical response to PD-L1 and PD-1 blocks is associated with increased expression of the PD-L1 protein on tumor cells. However, this therapy has been successful in patients whose tumor cells do not express PD-L1, and has failed in other patients whose tumor cells do.

There is still a lack of predictive biomarkers that can help identify which patients are more likely to benefit from blocking PD-L1 and PD-1.

In an attempt to determine a predictive biomarker, researchers studied mice with cancer of the colon or lung, ovarian cancer, or melanoma, as well as examining tissue samples from melanoma and ovarian cancer patients treated with immunotherapies.

Researchers discovered that PD-L1 expression on macrophages and dendritic cells was connected to a clinically significant response to treatment. These so-called antigen-presenting cells are types of immune cells found around the tumor and in nearby lymph nodes.

“The antigen presenting cells are the real guide. PD-L1 expression in these cells is responsible for response to therapy. This reshapes our understanding of checkpoint blockade therapies and the biomarkers that may predict clinical efficacy,” Weiping Zou, MD, PhD, the Charles B. de Nancrede Professor of Surgery, Pathology, Immunology and Biology at the University of Michigan, said in a press release.

Tumor tissue is currently widely tested for PD-L1 expression before a patient receives PD-L1 or PD-1 blockade immunotherapy. However, this broad analysis might miss which specific cells are expressing PD-L1. This study suggests the necessity of drilling down to the type of cells with PD-L1 expression, not just the whole tumor.

“Our data suggest that PD-L1⁺ APCs [antigen-presenting cells] may mechanistically shape and therapeutically predict efficacy of PD-L1 and PD-1 blockade,” the investigators concluded.