NEW YORK – Arima Genomics is making a concerted push to expand its business into cancer diagnostics, following the demonstration of its sequencing-based chromosome conformation assay kits to identify cancer driver mutations.
To help with the effort, the Carlsbad, California-based firm is creating a clinical advisory board, led by New York University Langone Health molecular pathologist Matija Snuderl. His lab has partnered with Arima to develop a second-line diagnostic test to find cancer gene fusions and rearrangements missed by sequencing-based DNA and RNA assays. He will be joined by Darren Sigal, director of gastrointestinal oncology at MD Anderson Cancer Center, and Ken Young, director of hematopathology at Duke University.
More than half of patients fail to get a diagnostic result from DNA-based cancer panels, Snuderl said, and Hi-C assays can provide useful information — contributing to diagnosis, prognosis, or therapy selection — in about half of its use cases, a number based on not-yet-published results from a 220-patient cohort profiled at NYU and Memorial Sloan Kettering Cancer Center.
In at least one pediatric case at NYU, Snuderl was able to identify a novel rearrangement in the PDL-1 gene in a brain tumor, which led to the patient's treatment with immunotherapy.
Arima has also partnered with Protean Biodiagnostics' CLIA lab to start a clinical service for whole-genome Hi-C-based analysis. Though the test is still being validated, the firm hopes to have the service running by July, beginning with tens of samples a month and ramping up to 100 per month near the end of the year. "This is us priming the pump for our reagents," said Chris Roberts, Arima's senior VP of corporate strategy.
The Hi-C-based approach does two important things for cancer diagnostics. First, it helps find interactions between genes and regulatory regions of DNA — missed by RNA-seq or exome sequencing — that create new drivers of cancer. And, perhaps most importantly, it works with formalin-fixed paraffin-embedded (FFPE) samples, letting it slip in easily with standard pathology lab workflows.
"We've been following what our customers have been doing with our tech," Roberts said. "They've taken our kits and shown they can ID novel structural variants with clinical significance."
Snuderl's relationship with Arima began about three years ago, as he sought new technologies to find cancer drivers that weren't being picked up by large sequencing panels. He had developed a 607-gene DNA panel called Profiling of Actionable Cancer Targets (PACT), which obtained 510(k) clearance from the US Food and Drug Administration in 2021, and an RNA-seq gene fusion panel, but was still coming up empty-handed in about half of cases. Some in the field turned to whole-exome sequencing, but he felt that wasn't the answer.
Having performed work showing the effects of methylation on cancer, he had a hunch that epigenetics might yield results.
When a colleague told him Arima might be able to do Hi-C, a proximity ligation assay that uses sequencing to provide information on long-range interactions in the genome and 3D structure, Snuderl said his initial reaction was, "That's not possible, that's crazy talk."
But by the end of 2021, he became convinced that Hi-C could help him test his hypothesis that regulatory regions were getting rearranged with oncogenes to drive tumor progression and in 2022 began "extensive collaboration." Snuderl disclosed that he will receive stock options for his participation in the clinical advisory board.
At first, he sent samples to Arima, but eventually he brought the firm's whole-genome Hi-C kit in-house. "It worked really well; the success is really high," Snuderl said. "It depends on how you define what is useful, but we can find some useful information in about 50 percent to 70 percent of cases."
The prime example was the child whose brain cancer was driven by a novel PDL-1 rearrangement. His lab has also found tumors with NTRK1, PLAG1, and MYBL1 fusions, among others. In general, Roberts said that about one-third of cases analyzed by Arima Hi-C result in an alteration in a biomarker that is the target of an FDA-approved therapy. Another 8 percent return a result that can qualify the patient for a clinical trial, and about 9 percent result in a marker with either diagnostic or prognostic value, based on NCCN guidelines for that tumor.
NYU and Arima are also conducting validation and concordance studies. Roberts noted that Protean is in the process of validating a pan-solid tumor assay, which is expected to be completed by the end of June, but declined to disclose financial details of the deal. The company is also considering hematologic panels and some smaller panels for specific cancers, such as lung cancer. Additionally, NYU Langone is in contact with the New York State Department of Health and Mental Hygiene to launch a laboratory-developed test using the Arima kit.
Snuderl is also preparing a manuscript comparing his lab's Hi-C assay to RNA-seq for structural variants. "We can find everything RNA can find," he said, as well as many variants it can't.
In one "crazy" case, he said, RNA-seq identified three gene fusions covering six genes, all on the same chromosome. "I wondered what else is happening there, so I used Hi-C," Snuderl said. "There weren't just three, there were a hundred-plus events. The chromosome had completely shattered and been stitched back together."
Arima isn't abandoning its other application areas, and it isn't alone in pursuing proximity-ligation-based technology for cancer genomics. Phase Genomics, a Seattle-based competitor, for example, is starting to see similar results, with some customers "moving towards developing an LDT," Phase CEO and Cofounder Ivan Liachko said.
"Every time we process a solid tumor sample, we find new fusions," he said. "In the vast majority of solid tumor samples we work with, there are new rearrangements and new fusions, as well as known ones."
Dovetail Genomics, a subsidiary of Cantata Bio that also offers proximity ligation assays, sells kits for cancer research, as well.
While Hi-C can provide an interesting look at cancer genomics, how to stratify and report these findings remains to be worked out. Snuderl suggested a tripartite strategy. First and easiest to deal with are clear rearrangements between genes. Then, there are rearrangements between genes and noncoding regions. Events such as enhancer hijacking could be straightforward to report, but these events may need to be integrated with a gene expression profile to clarify their effects.
Last and trickiest is spatial information that can be captured with the technology. "We'll be able to see [topologically associated domains and chromatin loops] that are interacting and changing regulation, but I don't think anybody has any idea what the significance of that will be," Snuderl said.
These tiers will need to be addressed in bioinformatics, and Snuderl proposed a consortium to build pipelines over several years. Beyond simply categorizing them, establishing the clinical significance of each call will take even more effort.
But Hi-C should be valuable even in the short term. Labs that don't have their own RNA fusion panels might seek to go straight to Hi-C, especially as validation data emerge. The assay could also be useful in lymphomas, where karyotyping is difficult, and in brain cancers, which don't normally get karyotyped. "If you do Hi-C correctly, you can do cytogenetics," Snuderl said.
Finally, the ability to detect fusions makes the assay potentially valuable for cancer therapy R&D. Fusions make good drug targets, but why patients respond or not isn't clear. Hi-C data could help provide insight into this question. It could also provide new data on the effects of drugging multiple fusions at the same time.
"Nobody uses two fusion drugs at the same time, because people don't usually see cases like that," Snuderl said.