NEW YORK – Cofactor Genomics has expanded the exploration of potential clinical applications for its immune modeling technology, with new proof-of-concept data published this month from a study that used the firm's predictive immune modeling technology, ImmunoPrism, to analyze a small cohort of high-grade serous ovarian cancer tumors.
Investigators from Women & Infants Hospital of Rhode Island, led by researcher and assistant professor Jennifer Ribeiro, reported in Frontiers in Oncology this month that the platform allowed them to identify a signature combining CTLA-4 and LAG-3 expression with signals associated with the presence of regulatory T cells, and determine that this signature was higher in patients who survived longer without disease progression.
Natalie LaFranzo, Cofactor's VP of market development, said in an interview this week that following recent analytical validation and IP milestones, the company has trained its focus toward internal clinical diagnostic development.
While the firm had previously anticipated that kit dissemination would be a major driver for clinical translation, it has shifted over the last two years to pursuing potential diagnostic applications more directly, collaborating with specialists in different oncology areas and largely testing samples through its in-house lab, LaFranzo said.
"We were able to publish the analytical validation of our platform in the Journal of Molecular Diagnostics last year and [our patent] came through in August [which covers] not just how we're building the RNA models and using them … but also how we leverage that information [across] all immuno-oncology applications," she said, "so this is now our major focus as a company, to take this forward into clinical studies."
As this shift has solidified, the company has expanded its work with academic groups to build new proof-of-concept data across more cancer types and clinical decision points, collecting data in tumor types like sarcoma and non-small cell lung cancer, as well as what it hopes will be its first clinical diagnostic, an immunotherapy response predictor for head and neck cancers.
Investigators from the newly published study believe their findings using ImmunoPrism could be the first step toward development of a biomarker test to predict how ovarian cancer patients respond to standard-of-care chemotherapy. They could also help explain confusing response patterns to immunotherapy and drive new immuno-oncology drug development for this challenging tumor type.
Although future research will be needed to make any "firm conclusion" about the prognostic CTLA-4/LAG-3/Treg signature they identified, the study authors said they believe their report has broken important new ground.
Cofactor's technology uses RNA sequencing to explore gene expression across a set of models the company has developed that encompass the activity of immune cell subsets, immune escape genes, and co-inhibitory/co-stimulatory molecules.
"It's a great platform because it's targeted to the immune features, but it has a lot of different factors that are included, so you have the immune genes, but also the cell types. And then you are looking at that in a multi-dimensional way," Ribeiro said.
In their study, she and her colleagues analyzed ImmunoPrism data for 26 HGSOC tumors split between those with significantly short versus significantly long progression-free survival (PFS).
According to the authors, the results revealed a novel combination: the immunological markers CTLA-4, LAG-3, and Treg immune cells, which were all more abundant in HGSOC patients who had longer PFS when treated with chemotherapy compared to those with poorer PFS.
The group also found that the co-stimulatory receptor ICOS was also significantly increased in patients with a longer PFS and this positively correlated with levels of CTLA-4, PD-1, and infiltration of immune cell subsets.
Checking these observations against data from The Cancer Genome Atlas, the researchers were able to confirm that both ICOS and LAG-3 were also significantly associated with improved overall survival in that larger retrospective cohort.
Apart from providing a potential tool for predicting patient prognoses, Ribeiro said that identifying these new prognostic features also opens opportunities for the development of novel immunotherapeutic approaches.
Although there are currently several EOC clinical trials exploring immunotherapy, patients with these tumors have so far shown relatively low response rates compared to other cancer types like lung cancer or melanoma.
This has been frustrating, Ribeiro said, because "it's very clear from our results that there are immunologic factors [at work]." Prior studies have also shown that patient with more T cells infiltrating their tumor survive longer compared to those with lower numbers. In other words, there is reason to believe that EOC patients could benefit from some form of immunotherapy, but they don't seem to respond well to the drugs currently available.
"We really believe that there are other factors that may be more important or may work synergistically with PD-L1 that we need to be exploring and might serve as better targets for therapy," Ribeiro said.
The Rhode Island group's new data provides a hint at what may be contributing to the observed discrepancy between immunogenicity and immunotherapy response, as well as some suggestion of other areas that could be explored in future drug development.
Authors wrote that the LAG-3 prognostic connection they observed is interesting in light of preclinical studies that have shown efficacy for investigational LAG-3 drugs in combination with immune checkpoint receptors in EOC. In addition, Phase I/II development of a LAG-3 monoclonal antibody now being investigated in a group of solid tumors that includes ovarian cancer.
The contribution of Tregs also raises many questions in light of conflicting data on the contribution of these cells to immunotherapy response or non-response.
"Perhaps one of the most interesting findings from this study is the identification of ICOS's association with improved patient survival in EOC," the authors wrote. ICOS is considered an immune checkpoint agonist, enhancing T cell proliferation, survival, and differentiation. But its function depends on the T cell subset in which it is predominantly expressed.
"On Tregs, it may dampen the immune response, while on cytotoxic T cells, it may promote an immune response," the team wrote.
Results from ImmunoPrism analysis also identified PVRL2 as the most abundantly expressed immune factor. According to Ribeiro and her colleagues, PVRL2 has been shown previously to be overexpressed in ovarian cancer.
Although there was no significant relationship between PVRL2 levels and PFS in the cohort — the marker appeared universally enriched, regardless of patient outcomes — there is reason to hope that it could be a valuable drug target in ovarian cancer, especially for patients whose tumors lack the biomarkers that confer response to existing anti-PD-1 and PD-L1 drugs.
Following up on their current results, Ribeiro said she and her colleagues now plan to zero in on the potential of the signals they identified as chemotherapy response predictors. "One of the things that we really want to look at now is whether these factors be detected in serum. That would make for a better biomarker," she said.
"The other major thing that we plan to follow up on is really just understanding more how chemotherapy changes the immune microenvironment and if that will inform potential novel immunotherapies," she added. "These patients are getting chemotherapy and it works really well in front-line therapy, so what we need to understand is what changes in the tumor and how can we best target after that."
According to LaFranzo, Cofactor plans to continue to explore ImmunoPrism's clinical potential across indications, but its diagnostic development efforts are currently furthest along in head and neck cancer, where collaborators from Washington University and the University of California, San Diego have been working to validate a predictive immune signature they hope can be used to direct first-line treatment with anti-PD-L1 drugs.
There is already an on-label IHC assay in this space, Agilent's PD-L1 IHC 22C3 pharmDx, which the FDA named in 2019 as a companion diagnostic to guide monotherapy use of pembrolizumab (Merck's Keytruda) in these tumors. Despite this, LaFranzo said that there is still an unmet clinical need to get patients on the right treatment in the first line and further minimize unnecessary toxicity from chemotherapy.
"It's a smaller patient population [with] only about 22,000 patients diagnosed every year in the metastatic setting, but we think it's still ethically and scientifically a great space to introduce a new technology," she said.
Although the work is still ongoing, LaFranzo said that the immediate path forward would be to develop an LDT through its CLIA lab. That said, the fact that the company has also developed a research-use kit version of ImmunoPrism means a move to eventually pursue FDA approval would also be very possible.
Cofactor has also noted on its website that it is seeking academic collaborators to direct company-funded research in a variety of other disease areas.