CLN Article
Last July, when the Food and Drug Administration (FDA) approved a blood-based test for colorectal cancer screening, the test joined the ranks of growing options available — from traditional colonoscopy to stool-based testing — to try to detect these cancers and their precursors. While the test’s sensitivity lags behind some other screening options, it could prove to be a more attractive option for people not currently engaging in screening, experts say.
Guardant Health’s Shield test, designed to detect fragments of DNA released by colorectal tumors into the blood (known as cell-free or cfDNA), was approved following publication of the ECLIPSE (Evaluation of the ctDNA LUNAR Test in Average Patient Screening Episode) trial (N Engl J Med 2024; doi: 10.1056/NEJMoa2304714). The clinical validation study of 7,861 people found a sensitivity of 83.1% for detection of colorectal cancers, including an 87.5% sensitivity for stage I, II, or III colorectal cancers and a 13.2% sensitivity for advanced precancerous lesions. Medicare and the U.S. Department of Veterans Affairs now cover the test for people at average risk for developing colorectal cancer with no personal or family history of the disease.
For years, colonoscopy has been the gold standard as the most sensitive test for cancers and precancers, said gastroenterologist John Kisiel, MD, a professor of medicine at the Mayo Clinic College of Medicine in Rochester, Minnesota, “but that requires that the colonoscopy be done very well. We know from countless studies that clinicians’ colonoscopy performance varies. When a colonoscopy is done to evaluate a patient with a positive stool-based test, detection rates are much higher, and we assume that’s probably due to some differences in endoscopist behavior when they know that the patient is being referred in.”
The Shield test has the same sensitivity for colon cancer as the fecal immunochemical test (FIT), said gastroenterologist and molecular biologist William “Bill” Grady, MD, senior author of the New England Journal of Medicine (NEJM) study and a professor in the translational science and therapeutics division of the Fred Hutchinson Cancer Center in Seattle. “In fact, it’s probably a little bit better. But it’s not as good as the Cologuard [at-home] stool test or colonoscopy.” Cologuard has a sensitivity of 92%.
Sensitivity for advanced cancers across all the tests “is quite high,” Kisiel said, including 100% for stage II–IV cancers in the Shield study. “But that’s not what matters, because advanced cancers are not what we want to find in a screening program. What we really want to find, ideally, are stage one or two cancers, which are considered curable. Even more important is finding advanced precancers.”
Detecting precancers is still a challenge for blood-based testing, however, as the amount of mutated DNA shed by tumors circulating in peripheral blood is extremely small, said gastroenterologist Aasma Shaukat, MD, MPH, the Robert M. and Mary H. Glickman Professor of Medicine at New York University’s Grossman School of Medicine.
When targeting advanced polyps, colonoscopy will miss 5–10 out of 100 times, Grady explained, whereas Cologuard will miss about 55 out of 100 times and Shield will miss 87 times out of 100. “So, it’s clear that colonoscopy is much better than Cologuard or the FIT or Shield tests.”
If the Shield test is not as accurate as colonoscopy or a stool-based test, why use it? “Right now, anywhere from 25%–35% of people who should be doing colon cancer screening are not doing it,” Grady said. As people are more likely to agree to a blood test than those other tests, it presents a potentially compelling option.
Uptake for the blood-based test by patients and providers is still “largely unknown,” Shaukat noted. However, there was a “pretty high” uptake for tests that screen for prostate cancer by measuring prostate-specific antigen, she said. In studies where patients have been called or interviewed to ask if they would complete blood tests if they were available to screen for cancer, “the answer is largely yes,” she said.
“The enthusiasm for a blood-based test is higher than for colonoscopy or for stool-based tests, the reason being colonoscopy is a very invasive test, there’s prep involved, time off, and there’s risk of complications,” Shaukat said. While stool tests are convenient and can be done at home, some patients don’t like handling stool, she added. Others may take the test home and forget to do it or misplace it.“With a blood test, the potential is that it can be coupled with other blood draws that patients are used to getting for cholesterol or other routine things,” Shaukat said. “It can be a convenient option, because as a physician, if I tell a patient to do it, then I can send them to the lab right away. I don’t lose momentum by their taking it home and forgetting about it.”
However, Grady said he worries that because the blood test is easy to order and complete, more people will opt for that as their first choice, rather than going to more accurate tests. “That we don’t want to have happen … because more people would die from colon cancer than if we stick with our current program of people not getting any colon cancer screening. We don’t want to completely replace it. We want that 30% of people who aren’t doing anything to do something.”
Recommendations for individuals at average risk include:
Whether physicians and guideline authors will embrace this option also remains to be seen, he said. The Shield test is not yet readily available at a lot of hospitals, and Grady has heard from people who say their doctor isn’t aware of the test, which he said is typical with anything new.
The NEJM study is definitive enough for guidelines organizations such as the U.S. Preventive Services Task Force to potentially include in its recommendations, according to Grady. “I think they’ll include it,” he said, “but I think it’s going to be probably a tier two option, where if you’re going to do screening, the tier one options are going to be FIT testing, colonoscopy, or Cologuard, and then tier two would be the blood test.”
For now, test results for blood and stool tests are returned solely as positive or negative, “and the FDA wants it that way,” Shaukat said. “The FDA gets very nervous when there’s interpretation involved, because they don’t want the consumer or the patient trying to do that.” Manufacturers of some blood tests in development are looking to see if they could provide a range of risk and interpretation around that range, she said.
Going forward, next-generation sequencing could be incorporated into these tests, Shaukat said, as could technology to detect signals coming from the stromal and immune cells surrounding tumor cells to help identify preneoplastic lesions.
Meanwhile, other manufacturers also are aiming to get into the game. Freenome in 2024 announced results from the PREEMPT clinical study of its blood test for colorectal cancer, which demonstrated a nearly 80% sensitivity in detecting colorectal cancer and a 92% specificity for nonadvanced colorectal neoplasia. Universal Diagnostics initiated a clinical validation study of its own blood test for colorectal cancer, called Signal-C. Exact Sciences (manufacturer of Cologuard and other tests) and CellMaxLife also are developing blood-based screening tests for colorectal cancers.
Beyond colon cancer, companies are adapting blood-based tests toward detection of multiple cancers, Kisiel said. Screening tests are available for breast, cervical, colon, prostate, and lung cancers, he said, but not for ovarian or pancreatic cancer, “which are much less common, but when people get them, they’re highly fatal.” Multicancer early detection testing could assure that while someone is getting blood drawn to screen for colon cancer, laboratories also could study the blood for presence of these other malignancies. “That’s a rapidly exploding field that 5 to 10 years ago was science fiction, and is now being commercialized,” he said.
For example, GRAIL’s Galleri test screens the blood for cell-free methylated DNA from more than 50 types of cancer. The test is not approved by the FDA but is available from CLIA-certified labs as a laboratory developed test. However, it has shown an overall sensitivity of only 27.5% for earlier stage 1 and 2 cancers (Lancet 2024; doi: 10.1016/S0140-6736(23)02830-1).
“I don’t think they’re quite ready yet,” said Grady, but clinical trials are evaluating blood-based multicancer tests to see if they are good enough for use in the clinic. The National Institutes of Health in February 2024 launched the Cancer Screening Research Network to evaluate these types of trials, he said. “We need to see what those trials show before we figure out whether these tests really should be used in the clinic … but if it turned out we had a test that could screen for 15 cancers at once in a blood test? Wow, that would be amazing.”
Kisiel is an inventor of the Cologuard at-home stool screening test and receives royalties from sales of the product outside of the Mayo Clinic. Grady is a member of the scientific advisory board for Guardant Health, manufacturer of the Shield blood-based screening test, has consulted for Freenome, and consults for Karius.
Karen Blum is a freelance medical and science writer in Owings Mills, Maryland. +Email: [email protected]