News | Published:

To foster screening, new colon cancer tests emphasize convenience

Nature Medicine volume 20, pages 322323 (2014) | Download Citation

MADISON, WISCONSIN — Colon cancer screening saves lives. Yet the most common screening strategy, colonoscopy, requires people to eschew solid foods the day before and down a jug of gag-inducing laxative that keeps them racing to the toilet. No surprise, then, that more than a third of all eligible adults in the US still forgo routine testing, resulting in as many as 44,000 deaths each year that could have been avoided. Two companies developing new noninvasive tests hope to shrink that statistic.

Neither test is meant to replace colonoscopy, a procedure that allows doctors to spot tumors and precancerous polyps. But these alternatives—an at-home stool test and a lab-analyzed blood test—may be attractive to people who dread the procedure. As Nature Medicine went to press, a US Food and Drug Administration (FDA) advisory committee was scheduled to discuss the safety and effectiveness of these new diagnostics over two days in late March. Even if they receive the committee's support, however, it remains to be seen how these tests might fit into the menu of screening options already available.

Colon cancer screening is generally recommended for everyone 50 years and older. But “there are some people who can't get over the idea that someone is going to take this long, lighted instrument and put it in a part of their body that they never talk about and that they don't view as an entry point for anything,” says Durado Brooks, director of prostate and colorectal cancers at the American Cancer Society (ACS) in Atlanta. Even people who accept the necessity of a colonoscopy detest the preparation.

Noninvasive options do exist: the fecal occult blood test (FOBT), which has been available for decades, and a newer, more sensitive stool test called the fecal immunochemical test (FIT) both detect blood not visible to the naked eye, which can come from tumors and polyps. The tests are inexpensive, and both rely on stool samples collected in the home. The performance of these tests varies widely, but some brands claim to detect as much as 90% of colon cancers picked up by colonoscopy. People who test positive on such stool tests are then referred for follow-up confirmation by colonoscopy, reducing the need for an invasive procedure in anyone who gets the all clear by FOBT or FIT.

Yet these stool tests aren't frequently used: a survey published last year by the US Centers for Disease Control and Prevention showed that only 10% of individuals who are up to date with their colon cancer screening use stool tests, whereas 62% rely on colonoscopy1. According to Keith Argenbright, director of the Moncrief Cancer Institute at the University of Texas Southwestern Medical Center in Fort Worth, that may be because many doctors fail to offer stool tests to their patients. “As physicians, we could do a better job of explaining the menu of options,” he says.

Stool sample: Exact Sciences's Tanya Quint runs samples for Cologuard, a new colon cancer test. Image: Shawn Harper

Last year, Argenbright coauthored a study indicating that underserved populations prefer simpler at-home stool tests over colonoscopy2. And although the existing stool tests do a decent job of detecting cancer, a company based here in Madison, Wisconsin, sees room for improvement.

Exacting test

In a nondescript building in Madison's largest research park, Kevin Conroy, chief executive and president of Exact Sciences, demonstrates how easy the Cologuard test kit is to use. He unfolds a plastic frame designed to span a toilet bowl and unscrews the lid of a small bucket that fits inside the frame. “This is the collection device,” he says. “You can't miss.” Next, he unscrews a vial containing a long brush—the FIT component of the test—and swipes it across an invisible piece of feces. He mimes pouring a bottle of DNA preservative over the same imaginary stool and screws the lid back on. If the sample were real, it would be shipped “right here to Madison, Wisconsin,” Conroy says.

Cologuard is designed to detect abnormal DNA methylation patterns and mutational markers from cancer cells or precancerous polyps in the stool. The test promises simplicity: FOBT requires users to avoid red meat, horseradish, vitamin C pills and medications such as aspirin for a few days beforehand, and both FOBT and FIT typically involve sampling from consecutive bowel movements. Cologuard, by comparison, requires only one sample and has no diet restrictions.

“Our goal as a company isn't to dissuade people from getting a colonoscopy,” Conroy says. “Our goal is to say, 'If you're going to get a noninvasive test, use the best noninvasive test.' We think there's room for a number of different options.”

A massive clinical trial published last month offers evidence of Cologuard's potential promise3. In the 10,000-person trial, the Exact Sciences test spotted 92% of colon cancers and 42% of precancerous lesions picked up by colonoscopy. FIT, by comparison, detected just 74% of cancers and 24% of lesions.

Many experts see the improvements provided by Cologuard as only incremental, though. The “transformative technology that everybody is looking for,” says Argenbright, is a blood test. Part of the reason is ease—people who visit the doctor might already be getting blood drawn for other reasons. A blood test could also lessen the 'yuck' factor that comes with stool tests. But most importantly, perhaps, it puts compliance more in the hands of physicians, who can order a blood draw while patients are still in the clinic rather than relying on patients to collect their stools at home.

Many companies are pursuing such blood tests (see 'Blood will tell'), but it's Epigenomics, a German diagnostics company based in Berlin, that has made the most headway. Along with Cologuard, the company's blood test, called the Epi proColon, was one of the two tests slated for discussion at the FDA advisory meeting last month.

Table 1: Blood will tell: Companies with blood-based colon cancer tests in development.

Out for blood

Epi proColon detects abnormal methylation on the SEPT9 gene. Researchers have found that this aberrant methylation occurs in colorectal cancer tissue but not in healthy colon tissue. A large study involving close to 8,000 people in the US and Germany showed that Epi proColon can detect 48% of colorectal cancers identified by colonoscopy4. However, unpublished data involving an updated version of the test puts the sensitivity at 68%.

That's a lower sensitivity than some other tests, including Cologuard. But Uri Ladabaum, director of the gastrointestinal cancer prevention program at Stanford University School of Medicine in California, points out that it's important to examine not only the one-time performance of a screening test but also how the test performs in a screening program. It's not yet clear whether a cancer or polyp that gets missed in a stool or blood test one year will get picked up in a subsequent year. But if at least some of them do, and if people stick to the recommended screening schedule, it's possible that a blood or stool test could perform as well over a few decades as colonoscopy.

Whether such a screening program would be cost-effective is another matter. Ladabaum published a study last year funded by Epigenomics (he currently consults for Exact Sciences) that modeled the economic return on a variety of testing strategies, including the SEPT9 blood test. He and his colleagues found that existing strategies—including FIT and FOBT alone or in combination with a more invasive test—provided better value for money than the blood test, but that Epi proColon was far better than no screening at all5.

Uptake and adherence patterns will ultimately determine what strategy comes out on top. The approval of a convenient blood test could convince large numbers of individuals who normally forgo screening to get tested. But it could also prompt individuals who are using existing techniques such as colonoscopy to switch to less accurate alternatives. “The interesting twists start to come in when you talk about tradeoffs,” Ladabaum says. Still, most experts agree that people need options. “The public should have a right to choose,” Brooks says.


  1. 1.

    , , , & MMWR Weekly 62, 881–888 (2013).

  2. 2.

    et al. JAMA Intern. Med. 173, 1725–1732 (2013).

  3. 3.

    et al. N. Engl. J. Med. 10.1056/NEJMoa1311194 (19 March 2014).

  4. 4.

    et al. Gut 63, 317–325 (2014).

  5. 5.

    , , & Cancer Epidemiol. Biomarkers Prev. 22, 1567–1576 (2013).

Download references

About this article

Publication history




  1. Search for Cassandra Willyard in:

Newsletter Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing