An affordable sequencing-based method for the early detection of cancer using cell-free DNA (cfDNA) is reported in Nature Communications this week, overcoming key hurdles associated with other liquid biopsy approaches. Despite the promise of using cfDNA methylation to detect both cancer and its tissue or origin, its use is hampered by the low fraction of tumor cell-free DNA, molecular heterogeneity of cancer, and sample sizes that are not sufficient to reflect a diverse patient population. To address such challenges, a team led by University of California, Los Angeles researchers developed an integrated cancer-detection system consisting of an assay — called cfMethyl-Seq — for genome-wide methylation profiling of cfDNA, along with a computational method that extracts four types of cfDNA methylation features and performs ensemble learning for detecting and locating cancer. "Applying our approach to 408 colon, liver, lung, and stomach cancer patients and controls, at 97.9 percent specificity we achieve 80.7 percent and 74.5 percent sensitivity in detecting all-stage and early-stage cancer, and 89.1 percent and 85.0 percent accuracy for locating tissue-of-origin of all-stage and early-stage cancer, respectively," the researchers write. Notably, the detection power of the method continues to increase as training sample sizes increase, facilitating a big data approach for cancer detection. The system, they add, "uniquely and cost-effectively retains the genome-wide epigenetic profiles of cancer abnormalities, thereby permitting the classification models to learn and exploit newly significant features as training cohorts grow, as well as expanding their scope to other cancer types."
UCLA Team Reports Cost-Effective Liquid Biopsy Approach for Cancer Detection
Sep 29, 2022
What's Popular?