, according to a new study.
The cfDNA blood test had 83% sensitivity for CRC, 90% specificity for advanced neoplasia, and 13% sensitivity for advanced precancerous lesions. Other noninvasive screening methods have sensitivity from 67% to 94% for CRC and 22% to 43% for advanced precancerous lesions.
“The results of the study are a promising step toward developing more convenient tools to detect colorectal cancer early while it is more easily treated,” said senior author William M. Grady, MD, AGAF, medical director of the Gastrointestinal Cancer Prevention Program at the Fred Hutchinson Cancer Center in Seattle.
“The test, which has an accuracy rate for colon cancer detection similar to stool tests used for early detection of cancer, could offer an alternative for patients who may otherwise decline current screening options,” he said.
The study was published online on March 14 in The New England Journal of Medicine.
Analyzing the Blood Test’s Accuracy
Dr. Grady and colleagues conducted a multisite clinical trial called ECLIPSE, which compared the sensitivity and specificity of a cfDNA blood test (Shield, Guardant Health) against that obtained with colonoscopy, the gold standard for CRC screening. Guardant led and funded the study.
Dr. William M. Grady
Guardant’s Shield test is designed to detect CRC through genomic alterations, aberrant methylation status, and fragmentomic patterns, which show up as an “abnormal signal detected” result. Similar blood tests are being developed as “liquid biopsy” tests for other emerging cancer screenings as well.
The study included 7861 people with average CRC risk who underwent routine screening with colonoscopy at 265 sites in the United States, including primary care and endoscopy centers in academic and community-based institutions. Eligible people were aged 45-84 years (average age, 60 years), and 53.7% were women. The race and ethnicity characteristics of the participants closely mirrored the demographic distribution in the 2020 US Census.
Overall, 54 of 65 (83.1%) participants with colonoscopy-detected CRC had a positive cfDNA blood test. However, 11 participants (16.9%) with CRC had a negative test.
The cfDNA blood test identified 42 of 48 stage I, II, or III CRCs, indicating a sensitivity of 87.5%, including 65% for stage I cancers, 100% for stage II cancers, and 100% for stage III cancers. The test also identified all 10 of the stage IV CRC cases. There were no substantial differences in sensitivity for CRC based on primary tumor location, tumor histologic grade, or demographic characteristics.
Among participants without advanced colorectal neoplasia on colonoscopy, 89.6% had a negative cfDNA blood test, and 10.4% had a positive test.
Among those with a negative colonoscopy — with no CRC, advanced precancerous lesions, or nonadvanced precancerous lesions — specificity was 89.9%.
Among 1116 participants with advanced precancerous lesions identified as the most advanced lesion on colonoscopy, the cfDNA blood test was positive for 147, indicating a sensitivity for advanced precancerous lesions of 13.2%.
Although the blood test has sensitivity similar to stool-based tests for CRC, the accuracy is lower than it is with colonoscopy, which remains the current gold standard for CRC screening, Dr. Grady said.
“Colorectal cancer is common and very preventable with screening, but only about 50% to 60% of people who are eligible for screening actually take those tests,” he said. “Getting people to be screened for cancer works best when we offer them screening options and then let them choose what works best for them.”