A panel discusses the science presented during Oral Abstract Session C.
“Increasing mutational burden is generally associated with increasing responses” to immune checkpoint inhibition, Shridar Ganesan, MD, PhD, of the Rutgers Cancer Institute of New Jersey, said during Oral Abstract Session C on biomarkers and inflammatory signatures on January 27. “However, there are some exceptions.” These include malignancies such as Merkel cell carcinoma, as well as RCC.
With Merkel cell, it is understood that exogenous viruses play a role in its responsiveness. With RCC it was less clear, and Dr. Ganesan’s group examined whether endogenous retroviruses (ERVs) may play a role in its response to immunotherapy. These viruses, he said, comprise between 5% and 8% of the human genome and represent an “ancient integration event.” Almost all the retroviruses have mutations that disable key coding genes and are silenced epigenetically in most normal adult tissues. “They are essentially genomic fossils,” Dr. Ganesan said.
There is some inappropriate expression of ERVs in some cancers, and abnormal ERV RNA can activate innate immunity signals. In particular, elevated expression of specific viruses ERV3.2 and ERVK.2 is seen in ER-positive/HER2-negative breast cancer, head and neck squamous cell carcinoma, and colon adenocarcinoma, as well as in RCC. Dr. Ganesan noted that these are “relatively young retroviruses,” with more recent integration into the human genome than various others.
The researchers investigated whether ERV expression correlated with evidence of immune infiltration and upregulation of checkpoint pathways. They did this with 66 annotated retroviruses and found differing levels of expression in various cancers; the strongest signal was in ccRCC.
They found that ERV expression can separate patients with ccRCC into three classes. Comparing high-expressing to low-expressing patients showed “quite striking” results, Dr. Ganesan said. High ERV expression was associated with increased expression of T-cell markers and immune checkpoint genes, including PD-1, CTLA-4, and LAG-3.
Next, they analyzed a cohort of 15 patients with metastatic ccRCC treated with a single-agent immune checkpoint inhibitor regimen who had either a partial response to the therapy or progressive disease. They measured the RNA expression of one retrovirus, ERV3.2, using quantitative reverse transcription polymerase chain reaction; two samples failed, leaving 13 for analysis. The expression of ERV3.2 was significantly higher among the responders to immune checkpoint inhibition than among the nonresponders (p < 0.05).
Dr. Ganesan concluded that expression of ERVs does correlate with immune activation and checkpoint pathway upregulation in ccRCC, as well as in several other solid tumor classes. The expression specifically of ERV3.2 appears to be associated with response to PD-1 blockade in ccRCC.
“Abnormal expression of ERV may be a biomarker of response to immune checkpoint therapy in some cancers with low mutation burden,” Dr. Ganesan said. “The mechanism underlying ERV expression in a subset of ccRCC needs to be investigated and may reflect underlying chromatin alterations.”
Ezra Cohen, MD, PhD, of the University of California, San Diego, was the discussant for the abstract. “Could this be the underlying partial reason for responsiveness to checkpoint blockade?” he asked. “It may be that these retroviruses that are re-expressed because of derangements of transcriptional machinery in cancer cells are responsible for the responses that we see in ccRCC,” he said. “Of course, it probably isn’t the whole story.”
He noted that there are some genomic features that may make some cancers responsive and pointed out that there are other cancers that have relatively high expression of ERVs that are fairly unresponsive to immune checkpoint inhibition.
Dr. Cohen said that this research suggests the need for a change in approach to move the field forward.
“Are we ready as a group, as institutions, as companies, as physicians, and as researchers around the world, to change the paradigm? I think we have to be,” he said. “Instead of the old paradigm, of taking a drug and trying to apply it to patients, we need to start fitting the patient to the drug.”