Universal Diagnostics (Universal DX) announced that the combination of cell-free DNA (cfDNA) methylation, fragmentation, and machine learning to detect early-stage colorectal cancer (CRC) was found to be effective as per an international, observational cohort study.
The research from the US-based bioinformatics and multi-omics firm showed that early-stage (I-II) CRCs detection using the methylation and fragmentation features of cancer-related cfDNA regions in combination with a machine-learning algorithm showed 92% sensitivity at 94% specificity.
The observational cohort study evaluated a patient sample set gathered from the US, Spain, Germany, and Ukraine. Its results will be presented at the ASCO Gastrointestinal Cancers Symposium.
As per the results, the prediction model correctly classified 92% of CRC patients. The sensitivity per cancer stage varied from 91% for stage I, 92% for stage II, 91% for stage III, and 93% for stage IV.
The multi-omics firm said that the specificity of the model was 94%, and it correctly identified 97% non-advanced adenoma (NAA), 93% benign colonoscopy findings of diverticulosis/diverticulitis, haemorrhoids, hyperplastic/inflammatory polyps (BEN), and 94% colonoscopy negative (cNEG).
Universal DX COO Christian Hense said: “This study further validates and reinforces the work we are doing to develop tests that detect cancer in its earliest stages.
“With a completely new sample set, we have again demonstrated highly-accurate early-stage CRC detection, further verifying the robustness of our technology and use of biomarkers to find traces of cancer in a person’s blood.
“At Universal DX, we believe early detection is one of the most powerful tools for improving survival rates, and are encouraged to see these promising results once again.”
Previously, the US-based firm demonstrated that non-invasive blood testing can be employed to detect CRC and pre-cancerous advanced adenomas (AA).
The testing is done via the analysis of cell-free circulating tumour DNA (ctDNA) methylation, fragmentation, and microbiome patterns with single targeted sequencing analysis, combined with advanced computational biology and machine learning algorithms.
Last year, the company expanded early-stage colorectal cancer detection to prognostics and stratification in order to improve outcomes and survival rates.
