NEW YORK – A pair of newly published studies suggest DNA isolated from blood or stool samples offers relatively noninvasive avenues for colorectal cancer (CRC) detection in average-risk members of the population.
For the first of the studies, published in the New England Journal of Medicine on Wednesday, researchers from Fred Hutchinson Cancer Center, Massachusetts General Hospital, and other centers focused on CRC detection with cell-free DNA (cfDNA) in blood samples.
As part of Guardant Health-funded, multisite trials known as the ECLIPSE study, the team compared Guardant Shield blood-based cfDNA testing with colonoscopy-based screening for CRC in symptom-free participants between 45 and 84 years old who were deemed to be at average risk of the disease.
The researchers initially enrolled 10,258 in the study, ultimately focusing on 7,861 participants who met the study's eligibility conditions. There, they found that the cfDNA-based test correctly picked up just over 83 percent of the CRC cases found using colonoscopy, with a false-negative rate of nearly 17 percent.
The blood-based approach appeared particularly effective for finding early-stage CRCs, the team reported. Indeed, the sensitivity reached 87.5 percent for individuals diagnosed with stage I to III CRC, but dipped to 13.2 percent for participants with advanced precancerous lesions at risk of progressing to cancer. The cfDNA test had a false-positive rate of 10.4 percent in the group of individuals who were deemed negative for any form of advanced colorectal neoplasia by colonoscopy.
"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," senior and corresponding author William Grady, a gastroenterologist affiliated with the Fred Hutchinson Cancer Center and University of Washington School of Medicine and medical director of the Fred Hutch Gastrointestinal Cancer Prevention Program, said in a statement.
The test's performance raises the possibility of offering a blood-based CRC screening alternative for individuals who currently forgo conventional screening options, Grady added.
He and his colleagues noted that colonoscopy screening is failing to reach up to half of those eligible for it — a gap that may be particularly pertinent as CRC diagnoses become more common in younger adults. CRC rates have increased roughly 1 percent annually in individuals under 55, according to data from the American Cancer Society, and regular CRC screening is recommended for average-risk individuals starting at 45.
"We continue to see younger people getting colorectal cancer, and it's now the third most common cancer for people under the age of 50," Grady said. "Having a blood-based test for people to take during routine doctor's visits could be an opportunity to help more people be screened."
Last spring, Guardant told investors that it had submitted data from ECLIPSE to the US Food and Drug Administration to support its application for premarket approval of the Guardant Shield test. Investigators also shared ECLIPSE data at the Digestive Disease Week conference last year.
In another NEJM study — funded by Exact Sciences, the company behind the noninvasive, stool-based CRC test Cologuard — an Indiana University-led team took a look at a next-generation sequencing-based screening strategy being developed by Exact Sciences that centers on multitarget profiling of stool-based DNA.
For an effort dubbed the BLUE-C study, the team prospectively enrolled 20,176 average-risk individuals over the age of 40 at 186 US sites, including 98 individuals diagnosed with CRC by colonoscopy and 2,144 individuals with advanced precancerous lesions. The group also included 6,973 individuals with non-advanced adenomas and 10,961 participants with negative or non-neoplastic colonoscopy results.
With the multitarget stool-based screening strategy, the investigators successfully picked up 93.9 percent of CRC cases, with nearly 91 percent specificity for advanced neoplasias. The sensitivity dipped to 43.4 percent for finding advanced precancerous lesions, they reported, noting that the test had 92.7 percent specificity when it came to the colonoscopy-negative/non-neoplastic individuals.
In contrast, the team saw 67.3 percent and 23.3 percent sensitivity for detecting CRC and advanced precancerous lesions, respectively, with fecal immunochemical testing (FIT). The specificity came in at almost 95 percent in the advanced neoplasias found by FIT, with almost 96 percent specificity for the non-neoplastic/colonoscopy-negative participants screened with FIT.
"The next-generation multitarget stool DNA test showed higher sensitivity for colorectal cancer and advanced precancerous lesions than FIT, but also showed lower specificity," first and corresponding author Thomas Imperiale, a researcher affiliated with Indiana University School of Medicine and the Regenstrief Institute, and his colleagues wrote.
Though he emphasized that the current study did not directly compare the next-gen test to the current version of Cologuard, Imperiale noted that the team hopes to see further improvements in the next-gen test's specificity, particularly a reduced false-positive rate for finding advanced neoplasias, as the test is rolled out and implemented more broadly.
"[T]he expectation is that the specificity of the next-gen test for advanced neoplasia will improve when the test is implemented — meaning fewer false-positive tests — while retaining high sensitivity for colorectal cancer and good sensitivity for the most advanced of the advanced, precancerous polyps (those with high-grade dysplasia)," Imperiale said in an email.
In November, Exact Sciences offered an early look at BLUE-C study data and outlined its plans to bring the next-gen Cologuard test to market in 2025.