NEW YORK – Researchers from Roswell Park Comprehensive Cancer Center have identified an inflammation-related biomarker that could be combined with current biomarkers like PD-L1 and microsatellite instability to identify gastroesophageal cancer patients who may benefit from checkpoint inhibitors.
The researchers, led by Sarbajit Mukherjee, assistant professor of oncology at Roswell Park, began their study exploring whether obese patients with metastatic gastroesophageal cancer shared a common biomarker that facilitated a better response to checkpoint inhibitors, going off of a hypothesis that immunotherapy could mediate the immune dysregulation caused by obesity. "It is well known that obesity leads to inflammation, and inflammation is something that can not only drive cancer but also leads to immune dysregulation," Mukherjee said.
While the researchers didn't identify a biomarker predictive of immunotherapy benefit specific to obese patients, they did find that gastroesophageal cancer patients who responded to checkpoint inhibitors had certain types of inflammation biomarkers in common.
The researchers used an RNA-seq assay from OmniSeq, a Buffalo, New York-based commercial lab that spun out from Roswell Park in 2015. Using unsupervised clustering of RNA-seq data, they developed an inflammation signature that stratified patients into three groups: inflamed or hot tumors, moderate tumors, and non-inflamed or cold tumors. They presented the results at the virtual ESMO World Congress on Gastrointestinal Cancer earlier this month.
The patients with inflamed tumors, Mukherjee said, were more immunogenic and were characterized by high expression of more than 250 immune-related genes. Cold tumors, he added, had less immune dysregulation, higher levels of cell proliferation, and overexpression of cancer/testis antigens, which are antigens that are expressed normally in germ cells in the testis but not in tumor tissues.
The study was small and included 46 patients with advanced gastroesophageal cancer. In a subgroup analysis of 23 patients who were treated with a checkpoint inhibitor, most of whom were treated with pembrolizumab (Merck's Keytruda), Mukherjee said those with inflamed tumors saw better outcomes, though he noted that the analysis was not statistically significant due to the small cohort size.
"We saw that in immune checkpoint inhibitor-treated patients, inflammation status was also prognostic," he said. "There was a trend towards improved overall survival and progression-free survival for these patients with hot tumors" compared to those with cold tumors.
Mukherjee's team embarked on this research based on a 2019 study, in which he and his colleagues explored why overweight patients seemed to be more sensitive to checkpoint inhibitors. In that study, published in Nature Medicine in 2019, researchers found obesity led to increased immune aging, tumor progression, and PD-1-mediated T-cell dysfunction, but it was also associated with increased efficacy of PD-1/PD-L1 blockade in both mouse models and clinical cancer patients.
Based on that research, Mukherjee and his team set out to explore how obesity-related inflammation made these patients more sensitive to immune checkpoint inhibitors. To do so, overweight or obese patients, defined as a BMI of 25 or higher, made up 60 percent of the study participants.
"We tried to dive into the mechanisms, such as what is it about obesity that makes patients more responsive to immune checkpoint inhibitors?" he said. "We hypothesized that this immune dysregulation, that is partly driven by obesity, could possibly be reinforced by immune checkpoint inhibitors, and when that happens, maybe obese patients can have a better response to new checkpoint inhibitors."
While the latest study began with a hypothesis that obese patients were more likely to have inflamed tumors, the researchers found that inflammation occurred across both obese and non-obese patients in the study. Overweight patients made up about the same proportion of the hot tumor and cold tumor groups, suggesting that the inflammation signature is not specific to obese patients.
Mukherjee and the researchers at Roswell Park sent samples to OmniSeq's lab for RNA sequencing using its immune response assay that is part of OmniSeq INSIGHT. The researchers analyzed 395 genes. Currently, OmniSeq offers several tests to patients through a partnership with LabCorp, including next-generation sequencing assays like the OmniSeq Comprehensive, Immune Report Card, OmniSeq Advance, and OmniSeq MSI. The lab's INSIGHT cancer genomic and immune profiling test became available this year and was also approved by the New York State Department of Health's Clinical Laboratory Evaluation Program, making it available to patients in the state.
OmniSeq's Chief Scientific Officer Jeffrey Conroy said in an emailed statement the company hopes to "include these novel signatures in future iterations of our testing to improve patient treatment options and to advance drug development opportunities with our pharma partners."
The Roswell Park researchers have already begun a larger study in gastroesophageal cancer to confirm these findings and look further into the prognostic value of the inflammation signature. They will present the results at a medical meeting later this year. Mukherjee is also interested in researching the mechanisms underlying the inflammation biomarker and learning more about immune response and immune resistance in these patients.
In practice, Mukherjee doesn't envision the inflammation signature replacing other validated biomarkers, like PD-L1 or microsatellite instability testing using immunohistochemistry, commonly used to predict immunotherapy response. Instead, he sees the inflammation test as additive.
If validated, physicians could use the inflammation signature, alongside PD-L1 expression and microsatellite instability testing, to home in on the patients who should receive checkpoint inhibitors. In the US, Merck's checkpoint inhibitor pembrolizumab (Keytruda) is approved for metastatic esophageal or gastroesophageal junction cancer patients who cannot have surgery or chemoradiation, if they've had prior systemic therapy and have PD-L1 positive tumors. Patients with metastatic gastric, gastroesophageal junction, and esophageal cancer can also receive Bristol Myers Squibb's nivolumab (Opdivo) with chemotherapy.
While many patients with gastroesophageal cancer may be eligible for checkpoint inhibitors, not all patients will respond, Mukherjee said. While IHC-based PD-L1 testing is a useful tool for predicting response, it only considers one pathway associated with response. Patients' ability to respond to immunotherapy can depend on other dysregulated pathways, potentially muddling test results, Mukherjee noted. Further, the link between PD-L1 or MSI status and immunotherapy response can vary between patients, he has found, making it necessary for other precise tools.
"PD-L1 will still be important because it's something that's been validated in several clinical trials, but at the same time we have to think about how we can combine other tests with PD-L1 to better predict immune response," Mukherjee said. "We believe that this biomarker selection should be more refined in the future so that we could avoid unnecessary treatment in patients who are not going to benefit, select [treatment] for those who are really going to benefit, and move more towards precision medicine.