Blood Assay Predicts Response to Pembrolizumab

Although some patients with advanced melanoma have dramatic responses to treatment with PD-1 inhibitors, most do not. Identifying which patients are most likely to respond to these immunotherapies has proven to be a significant challenge for researchers.

Eager to determine which patients might fare best with anti–PD-1 therapy, Alexander Huang, MD, and colleagues at the University of Pennsylvania in Philadelphia and Memorial Sloan Kettering Cancer Center in New York, NY, investigated whether tracking the effect of pembrolizumab (Keytruda; Merck) on immune cells in blood samples might provide an answer. Huang presented the team's findings at the Second CRI-CIMT-EATI-AACR International Cancer Immunotherapy Conference in New York, NY, held September 25–28.

Researchers analyzed blood samples from 29 patients with stage IV melanoma taken before and 3, 6, 9, and 12 weeks after beginning treatment with pembrolizumab. Exhausted phenotype CD8+ T cells express high levels of PD-1 on their surface, Huang explained, meaning that treatment with pembrolizumab could reenergize them.

The researchers assessed this effect and “found a modest but significant increase in the proliferative marker Ki67,” Huang said. In fact, Ki67 levels were significantly higher in PD-1–expressing CD8+ T cells in posttreatment blood samples, compared with pretreatment samples, in 78% of the patients, indicating a strong, on-target immunologic response. “Yet only a fraction of the patients had significant tumor shrinkage—38% of the cohort,” Huang noted, meaning that Ki67 alone did not predict clinical outcomes.

Instead, the researchers found that the magnitude of reinvigoration of exhausted T cells—the peak level of Ki67 after treatment—compared with tumor burden (the volume of all tumors added together) prior to treatment correlated with clinical response. In fact, all patients who had a ratio of reinvigorated T cells to pretreatment tumor burden greater than 1.94 were alive after 11 months, compared with 50% of patients with a ratio less than 1.94. In a second cohort of 18 patients, 75% of patients with a ratio greater than 1.94 were alive after 24 months, compared with 29% of patients with a ratio less than 1.94.

The findings suggest that the failure of pembrolizumab may be due not to the inability to reinvigorate exhausted T cells, but rather to an imbalance between CD8+ T-cell reinvigoration and tumor burden.

“These are still small trials and have to be validated in a larger [cohort],” said Huang. “But it raises the possibility of a blood draw during therapy, where clinicians can intervene early and, if necessary, add another immunotherapy” to augment the immune response.

Although a few studies have concluded that high expression of the PD-1 ligand, PD-L1, in tumor tissue predicts a response to pembrolizumab, its use as a biomarker is far from perfect: Some patients with high PD-L1 expression do not respond whereas some with low expression do. Huang said his team doesn't know whether their predictive ratio might be better at determining the likelihood of response than PD-L1—or whether assessing and combining both markers would prove more accurate than either method alone, a question that merits additional research.

“One of the limitations is whether we can get biopsies on a routine basis from these patients to compare with the blood, and if so, I think it's absolutely necessary to do that,” Huang explained. –Suzanne Rose

For more news on cancer research, visit Cancer Discovery online at http://cancerdiscovery.aacrjournals.org/content/early/by/section.