Tumor Proteins May Be Barriers to Immunotherapy Success

Tumor samples from a single patient with melanoma provided researchers with new information about gene expression profiles that could inform immunotherapy treatments in the future.

Some immunotherapies may be failing not because of gene alterations, but because proteins are impeding immune responses, according to researchers at the Johns Hopkins Bloomberg~Kimmel Institute and Memorial Sloan-Kettering Cancer Center. The collaborators compared variations in protein expression in tumor samples from a single melanoma patient and discovered some of the mechanisms underlying response or resistance to immunotherapy agents.

The “proof-of-concept” findings, published online in Clinical Cancer Research, point to distinct variations not in the genetic code of each tumor sample, but in the expression levels of certain proteins encoded by normal genes.

Checkpoint inhibitors have shown incredible promise. However, they don’t work consistently in all cancer types or even among different patients with the same type of cancer. Study investigator Suzanne Topalian, MD, a professor in the Department of Surgery at the Johns Hopkins University School of Medicine and an associate director for the Bloomberg~Kimmel Institute in Baltimore, Maryland, said one checkpoint inhibitor (an anti-PD-1) causes cancers to shrink in about 60% of patients with advanced Hodgkin lymphoma, and in about 20% of those with head and neck, lung, and bladder cancers. In addition, responses can vary even among different tumor sites in the same patient.

To investigate these uneven responses, Dr. Topalian and colleagues examined 26 tumors from one 60-year-old patient with melanoma. The patient received anti-PD-1 therapy at Johns Hopkins for 6 months before his sudden death from a non–cancer-related cause. He had been treated for 15 months before his death with a different checkpoint inhibitor (anti-CTLA-4). His physicians received permission to perform a research autopsy soon after his death. Samples from four of the patient’s tumors were collected over a 5-year period prior to his death and the remaining 22 were removed during autopsy.

During the patient’s treatment, tumor surveillance with CT scans and clinical examinations showed that some tumors were shrinking in response to immunotherapy, while others continued to grow. The patient’s history and the unusual access the researchers had to the large number of tumors before and immediately after his death were helpful in providing some important clues.

The researchers looked at the tumors’ gene expression profiles, which show how active each gene is in producing proteins. They discovered clear differences between the regressing and progressing tumors; those that didn’t respond to anti-PD-1 expressed approximately 2,000-fold more activity in the LAMA3 gene, which is responsible for producing a protein molecule that helps cells stick together.

The nonresponding tumors also had about 4-fold more activity in the CXCR2 gene, which is associated with the migration of neutrophils. “We conclude that gene expression analysis is a potentially powerful tool for discovering the Achilles heel for each tumor,” said Dr. Topalian.