Investigators led by Minetta C. Liu, MD, a medical oncologist with the Mayo Clinic, Rochester, Minn., studied 6,689 participants – 2,482 with cancers of more than 50 types and 4,207 without cancer – drawn from the Circulating Cell-free Genome Atlas Study and the STRIVE Study populations.
The investigators performed bisulfite sequencing that targeted informative methylation regions of plasma cell-free DNA (cfDNA), and developed and validated a molecular classifier using methylation patterns to detect cancer and determine its tissue of origin.
Test performance was assessed both for cancer overall and for a prespecified set of 12 cancers (anus, bladder, colon/rectum, esophagus, head and neck, liver/bile duct, lung, lymphoma, ovary, pancreas, plasma cell neoplasm, stomach) that account for about 63% of U.S. cancer deaths annually.
Results reported this week in the Annals of Oncology showed that the test had a specificity of 99.3% in the validation cohort, corresponding to a false-positive rate of just 0.7%.
Sensitivity for detecting stage I-III disease was 43.9% for cancer overall and 67.3% for the prespecified set of cancers accounting for the majority of U.S. cancer deaths.
Test sensitivity increased with stage both for cancer overall (18%, 43%, 81%, and 93% for stage I, II, III, and IV disease, respectively) and for the prespecified set of cancers (39%, 69%, 83%, and 92%, respectively).
The test was able to predict a tissue of origin in 96% of samples in which a cancerlike signal was detected, and in 93% of cases, that prediction was accurate.
Some of the patients who had cancer were symptomatic and therefore would not be considered a screening population, Dr. Liu and coinvestigators acknowledged. Also, the test’s potential for reducing mortality remains unknown, and 1-year follow-up to verify cancer-free status was not yet available for all of the individuals without cancer.
“Together, these data provide compelling evidence that targeted methylation analysis of cfDNA can detect and localize a broad range of nonmetastatic and metastatic cancer types including many common and deadly cancers that lack effective screening strategies,” they maintained. The test’s “specificity and sensitivity performance approach ... the goal for population-level screening.”
“Considering the potential value of early detection in deadly malignancies, further evaluation of this test is justified in prospective population-level studies,” the investigators conclude. “Clinical validation in intended use populations is ongoing ... and a study has been initiated that is returning results to health care providers and patients ....”
Dr. Liu disclosed that the Mayo Clinic was compensated for her advisory board activities for GRAIL Inc. The study was supported by GRAIL, and by Princess Margaret Cancer Centre’s McCain Genitourinary BioBank in the department of surgical oncology.
SOURCE: Liu MC et al. Ann Oncol. 2020 Mar 31. doi: 10.1016/j.annonc.2020.02.011.