NEW YORK – Eone-Diagnomics Genome Center (EDGC) is on the verge of launching a blood-based multi-cancer early detection test that the firm hopes to ramp up in coming years to detect dozens of cancer types and potentially non-cancer applications such as Alzheimer's progression.
The South Korea-based joint venture between bioinformatics firm Diagnomics and clinical laboratory firm Eone Laboratories says it plans to launch its OncoCatch-E epigenetic liquid biopsy assay by early 2024 as a laboratory-developed test that would compete in the emerging MCED market. It is also in early talks with the US Food and Drug Administration and South Korea's FDA about regulatory pathways for marketing clearance.
Min Lee, CEO of Diagnomics and co-CEO of EDGC, said EDGC has developed a library preparation method for use with next-generation sequencing to identify the presence and source of multiple cancer types though an algorithm-driven analysis of methylation patterns at 65,000 sites in circulating tumor DNA.
He said EDGC plans to launch its OncoCatch-E test for detection of colorectal, lung, breast, and gastric cancers next year and, within the following year, expand screening to include 10 cancer types. With additional research collaborations, he hopes the MCED assay eventually will aid in the detection of 20 or 30 cancer types. By using a deep learning neural network to analyze methylation patterns across tens of thousands of sites, he said the test could be applied to any disease with a distinct methylation-related pattern, and the firm is evaluating its potential for identifying Alzheimer's progression.
EDGC was established in 2013 as a joint venture between San Diego-based Diagnomics and Incheon, South Korea-based Eone Laboratories. The partner companies have been working vigorously in the past five years on epigenetic analysis that could be used for early cancer detection, according to Lee.
EDGC launched its first liquid biopsy assay, OncoCatch-S, for early-stage lung cancer screening in Asian markets in 2019. That assay uses copy number variant-based scores and quantitative cfDNA measurements to identify proteins linked to lung cancer.
With the OncoCatch-E launch, the firm plans to offer the test through its CLIA-certified labs in California and Colorado and is in talks with other cancer centers and labs. Lee said the service would initially be a send-out model through which the patient samples would be sent to EDGC's laboratory in South Korea for analysis. The firm is also in the final stages of obtaining clearance from South Korea's FDA to manufacture and export testing kits in coming months.
EDGC is poised to enter an increasingly competitive market for MCED assays. Grail launched its Galleri assay as an LDT in 2021, while Harbinger Health recently said it plans to bring its' own LDT to market in early 2025. Toronto-based Geneseeq recently gained CE marking for its NGS-based MCED kits to assess multiomic features of circulating free DNA, and Exact Sciences also inked a research agreement with a Texas-based healthcare system to offer testing to 50,000 individuals at primary care clinics over three years.
Lee has previously said its OncoCatch-E test minimizes sequencing by focusing on a subset of methylation sites across a patient's genome and employs a simple and efficient proprietary protocol for enzyme digestion, adapter ligation, and library preparation.
The OncoCatch-E assay uses methylation-sensitive restriction enzyme digestion followed by sequencing (MRE-seq) to identify unmethylated restriction sites. Lee said that, using its proprietary library preparation methods, sequencing primers are added only to the hypomethylated regions that are processed by restriction enzymes, which minimizes background noise from other genomic regions and allows focused analysis of methylation patterns.
"Through this method, we generate a unique hypomethylation profile from 65,000 SacII enzyme sites, which are then suitable for machine-learning analysis," he said. "Our consistent library preparation for both initial pattern discovery and patient testing significantly enhances the accuracy of our AI algorithmic predictions compared to other AI methylation analyses."
In a recent publication in Scientific Reports, a team led by EDGC researchers wrote that they saw promising results in a retrospective study on colorectal and lung cancer detection, and Lee said the firm is close to publishing results of a subsequent study on more cancer types.
In the article, the authors wrote that they analyzed blood samples from 96 patients who had prior diagnoses of colorectal cancer, 95 with lung cancer, and 126 healthy controls. The team used the MRE-seq method to detect colorectal cancer with 78 percent sensitivity, 99 percent specificity, and an area under the curve of 0.978, and lung cancer with 66 percent sensitivity, 99 percent specificity, and an area under the curve of 0.956. The researchers said their analysis also identified the type of cancer in 179 of the 191 samples from cancer patients, with 94 percent accuracy in colorectal cancers and 90 percent in lung cancers.
In a separate unpublished technical paper, the OncoCatch-E assay was able to identify the source of gastric and breast cancers with similar accuracy.
In the colorectal and lung cancer study, the assay delivered stronger performance with more advanced-stage cancers than early-stage cancers. Among stage I cancers, the test performed with 77 percent sensitivity in stage I colorectal cancers and 50 percent in stage I lung cancers.
In the study, the authors extracted cfDNA from plasma with a Revvity Chemagic 360 instrument and performed high-throughput NGS using an Illumina NovaSeq 6000 sequencer.
The OncoCatch-E results will be used to generate reports for physicians on the similarities of a patient's methylation patterns to healthy individuals and those with certain types of cancers. The firm has found that about 10 percent of patients have received high cancer scores that would justify immediate follow-up testing and 40 percent received medium scores that warranted monitoring, lifestyle changes, and retesting six months later.
The testing process, from sample collection to results, takes about one week. Lee estimates that performing the test currently costs a laboratory $500, and he said the firm has not yet set a price tag for its product.
Lee said EDGC plans to launch its LDT in the US first through multiple CLIA labs. The firm is in negotiations with several labs in California that would perform the screenings and send the data to EDGC for analysis. Lee said the firm also has an uncertain path ahead for its gaining regulatory approval through the US FDA, and he said the firm may need to develop standalone tests for individual cancers before it can work on clearance for a multi-cancer test.
The emergence of multi-cancer early detection LDTs has both inspired optimism and raised concerns. Such tests raise hopes of finding cancers in earlier stages when they are more treatable, but healthcare providers see a lack of utility data and an absence of oversight from the US FDA.
Grail's Galleri blood-based MCED test is designed to detect more than 50 cancer types and locate them in tissues or organs through analysis of cfDNA for cancer-associated methylation patterns. The firm just forged a research collaboration in August with the Whitman-Walker Institute and Cancer Support Community to build awareness and use of early cancer screening as well as assess the utility of the MCED in practice with diverse patient populations.
Early results presented in June from Grail's SYMPLIFY study indicated that, in testing on 5,461 patients with signs and symptoms of cancer but no prior cancer diagnoses, the Galleri assay had a positive predictive value of 76 percent and negative predictive value of 98 percent, with overall sensitivity of 66 percent and specificity of 98 percent. The results also indicated Galleri's sensitivity increased with the patient's age and cancer stage, raising questions about its clinical utility.
Lee said that, in addition to the firm's work on OncoCatch-E, the firm is also developing less expensive PCR-based tests that employ far fewer biomarkers for the detection of a single cancer type. A colon cancer screening test with 15 methylation markers provides nearly as much predictive power as the OncoCatch-E for that cancer type, he said. He expects the firm will be able to develop similarly narrow screening methods with 10-20 genetic markers for other cancer types.
If the firm makes a play in Alzheimer's diagnosis, it would be entering a hot market segment with growing competition among firms using cerebrospinal fluid and plasma tests and, more recently, blood tests.
And as the firm explores the potential of applying its methods of recognizing patterns in DNA to Alzheimer's progression testing, Lee envisions possibilities of using the same techniques to develop tests for cardiovascular, neurological, and aging-related diseases.
"Our MRE-seq method targets the same set of methylation markers for both pattern discovery and patient testing, making it broadly applicable to any disease with associated methylation changes," he said.