- Identifying patients in advance of treatment who will benefit from immune checkpoint blockade drugs.
- Extending effective immunotherapy to more patients.
MD Anderson researchers have identified a pair of factors that separately thwart cancer immunotherapy, providing new targets to improve treatment. One of those is a genetic mutation that could help guide treatment decisions.
“These findings are a major step on the path to more personalized immunotherapy, where we identify targets for specific groups of patients, and tailor treatment strategies for them, just as we did with molecular targeted therapy,” says John Heymach, M.D., Ph.D., professor and chair of Thoracic/Head and Neck Medical Oncology.
How immunotherapy is blocked
Both research papers reveal factors that thwart treatment with anti-PD1 or anti-PD-L1 drugs that block the PD1 off-switch on immune system T cells, freeing them to attack tumors.
Researchers found a genetic mutation that produces a “cold” tumor environment, with very little penetration of the tumor by T cells. Another team discovered a protein on tumors that shuts down immune response in a “hot” tumor that has attracted an immune system attack.
Both teams identified drugs that can potentially target these suppressors of immune response when they’re combined with anti-PD1 immunotherapies.
CD38 – shutting down T cell assault
One team found that treating with anti-PD1 immunotherapy triggers a counterattack by a surface protein found on the tumors of some patients that stifles the immune system's assault.
Their research shows in detail how the protein, called CD38, reaches out to disable attacking immune system T cells and points to a variety of drugs that could counter CD38's defenses.
"We've shown that CD38, historically best known as a surface protein on immune system cells and a therapeutic target in multiple myeloma, plays an active role on solid tumors, shutting down immune response in immunologically 'hot' tumors that are undergoing immune attack," says senior author Don Gibbons, M.D., Ph.D., associate professor of Thoracic/Head and Neck Medical Oncology.
"There are a number of ways we might translate these findings in the clinic, because there are three categories of drugs under development that aim at targets we've identified," Gibbons says. "We're working on that now."
The team analyzed gene expression among mice treated with anti-PD-L1 antibodies whose tumors developed resistance after five to seven weeks and found that only CD38 was prominently expressed.
They found attacking T cells and the resulting inflammatory environment produce substances that activate CD38 on the tumors.
CD38 drives adenosine production, which connects to receptors on the T cells to suppress their activity.
The researchers found CD38 present on human tumors in 15 percent and 23 percent of early-stage lung cancer patients in two cohorts comprising 793 patients.
Immunotherapy combination overcomes resistance
Treating mice with a combination of anti-PD-L1 and an anti-CD38 antibody stifled both the original tumor and those that had spread to other organs. Adding anti-CD38 to mice after their tumors developed resistance to anti-PD-L1 agent enhanced the activity of attacking T cells and reduced two types of cell that inhibit immune response.
Their data suggests the CD38 resistance mechanism could be present in other cancers.
"We're studying other tumor types now," Gibbons says.
One Phase 1 clinical trial underway at MD Anderson pairs a CD38 inhibitor with an anti-PD1 antibody for advanced lung or prostate cancer patients whose tumors have previously responded but progressed on immunotherapy.
STK11/LKB1 mutations chill immune response
Another team found that a gene called STK11, which is mutated or deleted in a third of non-small cell lung cancer patients, fosters an immunologically "cold” tumor microenvironment, with minimal penetration of tumors by T cells, rendering anti-PD1/PDL1 drugs ineffective.
“We’ve identified what we think is the most prevalent genomic driver of a cold tumor microenvironment and primary resistance to anti-PD1 immunotherapy in non-small cell lung cancer,” says lead author Ferdinandos Skoulidis, M.D., Ph.D., assistant professor of Thoracic/Head and Neck Medical Oncology.
“Our results suggest that a single molecular mechanism downstream from STK11/LKB1 mutations accounts for a very large percentage of patients whose tumors resist immunotherapy,” Skoulidis says.
Multi-institution effort to advance lung cancer treatment
Researchers previously identified patients with mutations in both the tumor-suppressing STK11/LKB1 and the cancer-promoting KRAS genes as a distinct group. Working with colleagues at other cancer centers through a Stand Up to Cancer dream team grant, they tracked down the impact of the mutations.
Analysis of 174 patients treated at MD Anderson, Memorial Sloan-Kettering Cancer Center and Dana-Farber Cancer Center showed that only 7.4 percent of those with both STK11/LKB1 and KRAS mutations responded to anti-PD1 therapy. Response rates in other groups ranged from 28.6 to 35.7 percent.
Median overall survival for the STK11/LKB1 and KRAS group was only 6.4 months, compared to 16 months each for the other two groups.
The team conducted an unbiased genomic analysis of 924 tumors from Foundation Medicine to identify genes that drive the absence of PD-L1 expression in tumors. Absence of PD-L1 is an indicator of a “cold” tumor.
They found STK11/LKB1 to be the only significantly enriched gene in tumors that both lacked PD-L1 and also had a high burden of mutated genes — usually a sign of vulnerability to immunotherapy because abundant mutations provide multiple targets for T cells.
A follow-up analysis of 66 patients treated with PD-1/PD-L1 inhibitors at MD Anderson confirmed the impact of STK11/LKB1 on response to treatment.
The team is working to better understand how the mutation specifically causes cold tumors and is exploring drug combinations to overcome resistance.
Skoulidis notes that STK11/LKB1 mutation status could become a biomarker for guiding treatment, used in combination with tumor mutational burden and PD-L1 status, both of which suggest vulnerability to immunotherapy, but are not conclusive indicators of patient response.