{"entity":{"id":"paper-cancerguard-mp-classifier-cebp-2026","kind":"paper","name":"Performance of a multi-biomarker class, multi-cancer early detection (MCED) blood test in a prospectively collected cohort","aka":[],"tldr":"Exact Sciences' development study of the methylation and protein classifier behind Cancerguard: it found about half of cancers overall but about one in seven stage I cancers.","summary":"Gainullin and colleagues assessed a methylation and protein (MP) classifier in a prospectively collected multi-centre case-control set of 6,352 samples (1,438 cancer, 4,914 non-cancer) across 21 tumour organ types, split between training and test sets. At 98.5% specificity, overall sensitivity was 50.9%, with 15.4% at stage I, 26.1% at stages I and II combined, 38.0% at stage II, 67.8% at stage III and 85.5% at stage IV. Excluding breast and prostate cancers, overall sensitivity was 56.8% and stage I sensitivity 17.2%. A somatic mutation reflex for borderline samples was explored.","asOf":"2026-10-29","links":[{"label":"CEBP 2026: methylation and protein classifier","url":"https://doi.org/10.1158/1055-9965.EPI-26-0541"},{"label":"PubMed","url":"https://pubmed.ncbi.nlm.nih.gov/42826075/"}],"tags":[],"related":[],"cancers":[],"sections":["early-detection"],"technologies":["mced","cfdna-methylation-testing","proteomics"],"targets":[],"drugs":["cancerguard"],"companies":["exact-sciences"],"institutions":[],"pathways":[],"terms":[],"trials":[],"people":[],"bottlenecks":[],"keyPapers":[],"journals":["cancer-epidemiology-biomarkers-prevention"],"dependsOn":[],"notes":[],"journal":"Cancer Epidemiology, Biomarkers & Prevention","year":2026,"doi":"10.1158/1055-9965.EPI-26-0541","pmid":"42826075","authors":"Gainullin VG, Manesse M, Zhuang F, et al.","paperType":"observational","findings":["Overall sensitivity 50.9% at 98.5% specificity.","Stage I 15.4%, stage II 38.0%, stage III 67.8%, stage IV 85.5%.","Without breast and prostate: overall 56.8%, stage I 17.2%."],"whatItMeans":"It is the only peer-reviewed stage-by-stage figure for the Cancerguard approach, and it describes a development classifier at 98.5% specificity, not the launched test, which Exact Sciences says runs at 97.4% specificity.","caveats":["Case-control; no positive predictive value in a screening population.","Development classifier, not the commercial product's threshold.","Authors employed by Exact Sciences."],"changedPractice":false,"participants":6352},"route":"/key-papers/paper-cancerguard-mp-classifier-cebp-2026/","neighbours":{"section":[{"id":"early-detection","kind":"section","name":"Early Detection & Screening","route":"/fronts/early-detection/"}],"technology":[{"id":"cfdna-methylation-testing","kind":"technology","name":"Cell-free DNA methylation tests","route":"/technologies/cfdna-methylation-testing/"},{"id":"mced","kind":"technology","name":"Multi-cancer early detection (MCED)","route":"/technologies/mced/"},{"id":"proteomics","kind":"technology","name":"Proteomics & phosphoproteomics","route":"/technologies/proteomics/"}],"drug":[{"id":"cancerguard","kind":"drug","name":"Cancerguard","route":"/drugs/cancerguard/"}],"company":[{"id":"exact-sciences","kind":"company","name":"Exact Sciences (Abbott)","route":"/companies/exact-sciences/"}],"journal":[{"id":"cancer-epidemiology-biomarkers-prevention","kind":"journal","name":"Cancer Epidemiology, Biomarkers & Prevention","route":"/journals/cancer-epidemiology-biomarkers-prevention/"}]}}