TORONTO, June 13, 2022 – Early cancer detection can significantly benefit patients with more effective treatments and better prognosis; the earlier one catches the disease, the better the chances of overall survival. But there are only a few existing early screening tests, especially for multi-cancer detection. Combining Geneseeq’s early screening technology MERCURY and a multi-dimensional fragmentomics model, we have published a series of early screening studies under the DECIPHER (Detecting Early Cancer by Inspecting ctDNA Features) program and showed excellent performance in different cancer types.
Today, we are excited to unveil another DECIPHER study published in Molecular Cancer, exploring the clinical utility of a novel multi-dimensional fragmentomics approach in the multi-cancer population through collaborations with multiple clinical facilities. This ultrasensitive model again demonstrated promising performance in the multi-cancer population, validating the performance in other cancer populations Geneseeq previously studied and published for.
In this study, a total of 1,214 participants were enrolled, including 381 primary liver cancer (PLC), 298 colorectal cancer (CRC), 292 lung adenocarcinoma (LUAD), and 243 healthy individuals. All participants were randomly split into training and testing datasets in a 1:1 ratio. Plasma samples were collected from the participants followed by cell-free DNA (cfDNA) extraction and low-coverage whole-genome sequencing. Five cfDNA fragmentomics features covering Fragment Size Coverage (FSC), Fragment Size Distribution (FSD), EnD Motif (EDM), BreakPoint Motif (BPM), and Copy Number Variation (CNV) were then extracted from the training dataset and implemented in five machine learning algorithms to build the ensemble stacked model. The model was then evaluated in the testing dataset and true-positive cases were selected to validate the cancer origin model.
Our model showed an area under the curve (AUC) of 0.983 for differentiating cancer patients from healthy individuals. At 95% specificity, the sensitivities for detecting all cancer reached 95.5%, while 100%, 94.6%, and 90.4% for PLC, CRC, and LUAD, individually. The cancer origin model demonstrated an overall 93.1% accuracy for predicting cancer origin in the test cohort (97.4%, 94.3%, and 85.6% for PLC, CRC, and LUAD, respectively). Finally, the performance of the model is consistent (cancer detection ≥ 91.5% sensitivity at 95.0% specificity, cancer origin ≥ 91.6% accuracy) when sequencing depth was down-sampled to 1X coverage.
“This proof-of-concept study showed that our early detection model held great potential for developing accurate and affordable early detection assays for clinical practice. Our next step is to target a broader population and more cancer types including the less prevalent ones”, says Dr. Hua Bao, author and the Associate Dean of Geneseeq Research Institute.