In a validation study reported in Annals of Oncology, Klein et al found that a targeted methylation-based multicancer early detection (MCED) assay using cell-free DNA (cfDNA) sequencing had high specificity for cancer signal detection and high accuracy in predicting cancer signal origin across a wide range of cancers.
As stated by the investigators: “…An MCED test used to complement existing screening could increase the number of cancers detected through population screening, potentially improving clinical outcomes. The Circulating Cell-free Genome Atlas study (CCGA)…was a prospective, case-controlled, observational study and demonstrated that a blood-based MCED test utilizing cfDNA sequencing in combination with machine learning could detect cancer signals across multiple cancer types and predict [cell signal origin] with high accuracy. The objective of this third and final CCGA substudy was to validate an MCED test version further refined for use as a screening tool.”
Study Details
The CCGA study enrolled 15,254 participants (8,584 with cancer and 6,670 without cancer) from 142 sites in North America between August 2016 and February 2019. The current prespecified substudy included 4,077 participants; of them, 2,823 had cancer and 1,254 did not have cancer, with absence of cancer confirmed at 1-year follow-up. In the cancer group, 55% of patients had stage I to II disease.
Key Findings
Cancer signals were detected across more than 50 cancer types. Specificity for cancer signal detection was 99.5%, and overall sensitivity across cancer classes and stages was 51.5%. The positive and negative predictive values for cancer signal detection adjusted for Surveillance, Epidemiology, and End Results (SEER) cancer incidence and clinical stage distribution in those aged 50 to 79 were 44.4% and 99.4%, respectively.
Across cancer classes, sensitivity of cancer signal detection was 51.9% for stages I to IV disease and 40.7% for stages I to III disease. Sensitivity increased with increasing stage: 16.8% for stage I, 40.4% for stage II, 77.0% for stage III, and 90.1% for stage IV.
Among the 12 prespecified cancers that account for two-thirds of annual cancer deaths in the United States (anal, bladder, colon/rectal, esophageal, head and neck, liver/bile duct, lung, lymphoma, ovarian, pancreatic, plasma cell neoplasm, and stomach), sensitivity of cancer signal detection was 67.6% for stages I to III disease and 76.3% for stages I to IV disease. The overall accuracy of cell signal origin prediction in true-positive cases was 88.7%.
The investigators concluded: “In this prespecified, large-scale, clinical validation substudy, the MCED test demonstrated high specificity and accuracy of [cell signal origin] prediction and detected cancer signals across a wide diversity of cancers. These results support the feasibility of this blood-based MCED test as a complement to existing single-cancer screening tests.”
Eric A. Klein, MD, of Glickman Urological and Kidney Institute, Cleveland Clinic, is the corresponding author of the Annals of Oncology article.
Disclosure: The study was supported by GRAIL, Inc. For full disclosures of the study authors, visit www.sciencedirect.com/journal/annals-of-oncology.
