BioCentury
ARTICLE | Translation in Brief
BioCentury & Getty Images

Translation in Brief

Unraveling checkpoint responses in PD-L1-negative cancers

City of Hope study suggests NK cells may be behind PD-L1 inhibitor responses in tumors lacking the checkpoint ligand

July 24, 2019 10:17 PM UTC

A new study from City of Hope offers a long-sought explanation for why some patients lacking PD-L1 expression on their tumors respond to antibodies against the target, and uncovers a way to boost NK cell antitumor activity in the process.

One of the top priorities in immuno-oncology is figuring out ways to predict which patients will respond to checkpoint inhibitors, and that’s been particularly challenging for PD-L1.

While many patients expressing PD-L1 on their tumors fail to respond to anti-PD-L1 mAbs, the puzzle is further complicated by the fact that some patients lacking PD-L1 expression on their cancer cells respond well to the therapy.

The new study, published on Wednesday in Cancer Discovery, indicates that PD-L1 expression on another cell type could be to blame.

“PD-L1 expression on NK cells may identify a subset of patients especially susceptible to NK cell therapy.”

Michael Caligiuri, City of Hope

In the paper, Michael Caligiuri and colleagues showed that PD-L1 expression is induced on NK cells when they contact cancer cells, and that PD-L1 expression increases the cells’ antitumor activity. Anti-PD-L1 mAb Tecentriq atezolizumab from Roche (SIX:ROG; OTCQX:RHHBY) further increased PD-L1 expression and antitumor activity of the innate immune cells, even in cancers lacking PD-L1 expression themselves.

In mice with PD-L1 negative lymphoid tumors, NK cell depletion prevented the antitumor effects of Tecentriq. In an orthotopic mouse model of PD-L1 negative human myeloid leukemia, the combination of Tecentriq and NK cell activating cytokines, which also drive up PD-L1 expression on the cells, decreased tumor burden and increased survival.

Based on gene expression profiling studies, the authors proposed that the interaction of activated NK and tumor cells up-regulates NK PD-L1 expression via the PI3K/AKT signaling pathway, and that antibody binding activates a positive feedback loop that continuously activates PD-L1 via p38.

According to Caligiuri, the mechanism they uncovered can be used to help stratify patients and design therapies that boost NK cell function.

Next steps for the team could include using PD-L1 expression on NK cells as a biomarker to identify patients with NK cells that are “alive, awake and active.”

“PD-L1 expression on NK cells may identify a subset of patients especially susceptible to NK cell therapy,” Caligiuri told BioCentury. “If we further activate them with cytokines, anti-PD-L1, or a combination, we might in this fraction of patients see a prolongation of the antitumor response.”

In the paper, the studies focused on tumors lacking PD-L1 expression to prove the point, but selecting patients who express the checkpoint ligand on their tumors and NK cells might be even more effective.

“The whole point of this was to shed some light on PD-L1 negative tumors that seem to respond to anti-PD-L1 antibodies, but if the tumors also expressed PD-L1, you might get a double whammy with use of an antibody,” said Caligiuri. He explained that it would release T cells blunted by the checkpoint and activate PD-L1 positive NK cells.

The results could also apply to adoptive NK cell therapies, he said. Caligiuri explained that treating a patient’s cells ex vivo with an anti-PD-L1 mAb or activating cytokines could increase PD-L1 expression to prime the cells for tumor activation. It may also be possible to induce PD-L1 expression in NK cell therapies using gene editing or gene therapy, he added.

Although other groups have identified PD-L1 expression on immune cells including NK cells, Caligiuri isn’t aware of any other studies that have uncovered a potential explanation for how that expression drives an antitumor immunity in patients with PD-L1 negative tumors.

Targets: AKT (AKT1; PKB; PKBA) - Protein kinase B; PD-L1 (B7-H1; CD274) - Programmed cell death 1 ligand 1; PI3K - phosphoinositide 3-kinase