NEW YORK – After announcing a collaboration with Intermountain Health spinout Culmination Bio earlier this week, Cofactor Genomics is “moving aggressively” to advance its OncoPrism test for predicting whether patients with a range of cancers are likely to benefit from immunotherapy.
San Francisco-based Cofactor has been validating OncoPrism, an RNA-sequencing-based T-cell subtyping and immune cell quantification test, within the PREDAPT study using de-identified tumor samples and matched clinical data. The goal, according to Cofactor CEO Jarret Glasscock, is to market OncoPrism as a tool to identify which patients are most likely to respond to PD-1/L1 checkpoint inhibitors such as Merck’s Keytruda (pembrolizumab) or Bristol Myers Squibb’s Opdivo (nivolumab).
The observational, prospective study has a target enrollment of 1,650 patients across 11 cancer types. But the new Culmination partnership is also key to validating the test, Glasscock said, since it allows Cofactor access to a database of 4.5 million patient specimens, each matched to longitudinal clinical outcomes data, and whole-genome, transcriptome, exome sequencing, proteomics, and other omics analyses.
This trove of data, which Culmination calls its Culmination Intelligence Platform, is its offering to partners like Cofactor. Although prospective validation of OncoPrism using samples collected in PREDAPT is critical to demonstrating its predictive performance across multiple cancer types, Glasscock said Culmination’s data gives Cofactor significant advantages, in terms of not just the sheer number of patient samples in the repository but also time.
“We need to know the patient outcomes from these immunotherapy treatments [to validate OncoPrism]. With the prospective data that’s being collected as part of PREDAPT, we have to wait months for the outcome to have the full data,” Glasscock said. “With prospective trials, we’re waiting on patients [to respond], and the value with Culmination Bio is that they have these patient samples and the high-quality patient data already collected and banked. It allows us to move much more quickly with our efforts.”
Cofactor’s OncoPrism test is essentially an immune profiling platform, as Glasscock described it. “We are looking at the most informative gene expression signatures that map to immune cell types and immune cell states that are most predictive of would-be responders and non-responders to immunotherapy,” he said, explaining that the OncoPrism platform essentially produces a measurement of these multiple analytes. “Where the clinical data comes in is in training the model to tell us, ‘What does a profile look like for responders in lung cancer? What does a non-responder profile look like?'”
While the OncoPrism platform is already built, the millions of linked clinical data in the Culmination Intelligence Platform will help Cofactor train its test to predict immunotherapy response across various cancer types, and to do so much more rapidly.
So far, Cofactor has made the most progress validating OncoPrism for predicting which head and neck cancer patients will benefit from immunotherapy, which Glasscock referred to as the firm’s “beachhead indication.” In September 2022, the firm presented data demonstrating that OncoPrism predicted anti-PD-1 benefit among head and neck cancer patients more accurately than immunohistochemistry-based PD-L1 expression, which many oncologists consider a “lousy biomarker” for identifying best responders to immunotherapy. In that study, OncoPrism demonstrated 75 percent accuracy in predicting immune checkpoint inhibitor responders, whereas PD-L1 immunohistochemistry testing had 42 percent accuracy.
With that data readout, Cofactor is “well on its way with head and neck cancer” and is aiming to commercialize OncoPrism in that setting as a lab-developed test in “the very near future,” according to Glasscock. He added that the firm is already working with physicians across the country to prepare for the commercial release in head and neck cancer later this year and is eyeing the lung cancer space shortly thereafter.
“We’re getting to the finish line, but I’m also extremely excited about the next nine cancer indications and how Culmination’s data will really supercharge the timelines for those,” he said, highlighting that Cofactor particularly values the depth and quality of data points that partners can query within Culmination Bio’s repository when trying to identify populations of interest.
The Intermountain Health spinout — which was incorporated as a company in June 2022 and launched publicly in March 2023 — showcases its Culmination Intelligence Platform as a search engine for building precision medicine cohorts. According to the information on its website, partners can query its data repository using factors such as age, sex, race, cancer type, biomarker status, and ICD codes, and home in on the patient cohort they’re looking to study.
“Culmination Bio was built to expand on the success of Intermountain’s precision medicine efforts by creating longitudinal multimodal datasets,” Culmination CEO Lincoln Nadauld said in a statement in March announcing the company’s launch. “Our partnership with Intermountain Health has enabled us to build a platform that will revolutionize the way we identify and implement novel clinical insights to benefit patients.”
Culmination Bio, which hasn’t publicized partnerships other than the one with Cofactor, declined an interview for this article. As such, details about the diversity and geographic representation of its database are not known. Intermountain Health operates 33 hospitals primarily in Utah, Idaho, and Nevada, but has affiliated sites in other areas as well.
“I know [Intermountain] has been building this database for decades, and it’s a rich collection of samples that they’ve been using within their four walls,” Glasscock said. “This commercial launch of Culmination Bio is an effort to bring that rich resource outside of their four walls.”