What if colon cancer screening didn’t involve poop?

The current methods for colon cancer screening have an unavoidable “ick” factor: Either get a long scope inserted up your rectum, or collect a sample of your own poop, pack it up, and ship it off to a lab.
But with colon cancer rates rising, especially in adults under 50, spotting irregularities before they become dangerous is more urgent than ever. Enter a third, more palatable option: a blood test designed to screen for cancer, known in clinical terms as a liquid biopsy.
In March, biotech firm Guardant Health reported in the New England Journal of Medicine (NEJM) that its Shield blood test successfully detected cancer in 83 percent of patients whose colon cancer had already been confirmed by a colonoscopy, which is the gold standard for colon cancer screening. After submitting its data to the US Food and Drug Administration earlier this year, the company received approval in late July for the new test.
Freenome, another company developing colon cancer blood tests, announced the preliminary results from their own clinical trial in early April, revealing that their test identified cancer in 79 percent of confirmed cases.
The results come with qualifications, however. An American Gastroenterological Association expert panel published studies in the journals Gastroenterology and Clinical Gastroenterology and Hepatology in March concluding that the new tests are not as sensitive as existing screening techniques, like colonoscopies and specialized stool tests, especially in detecting precancerous lesions. (Sensitivity refers to a test’s ability to correctly detect positive cases of an illness.)
Colonoscopies detect cancers and precancer lesions at the highest sensitivity level, but they also cost the most. While other stool tests are comparatively less sensitive than colonoscopies, they perform similarly to Guardant and Freenome’s blood tests and are also much cheaper.
For now, the current blood tests are less cost-effective than colonoscopies and fecal tests because they aren’t accurate enough at the early stages of disease, which is when detection is more valuable for treatment. A study published last year in JAMA, which measured the cost of different treatment against the years of life gained by patients who use them, reached similar conclusions: These blood tests need to either be cheaper or more accurate if it’s to replace the traditional screening process.
“The biggest difference is that these blood tests aren’t very good at finding precancerous lesions,” said Zainab Aziz, a medical student at Columbia University and the first author on the JAMA paper. “That’s how screening really helps with cancer incidence.”
Without better sensitivity, two of the blood tests furthest along in development lack a strong clinical and cost advantage over the screening methods already available.
“If you substituted a blood test for one of the currently recommended tests, the outcomes would be worse than they are today,” said David Lieberman, a professor of medicine at Oregon Health and Science University and an author on the Gastroenterology paper. There would be more cancers and more deaths than if the patients were electing to have one of the recommended tests.
Still, the blood tests have achieved a new level of accuracy in the recent trials, outperforming their predecessors and clearing the standard set by the federal government for Medicare coverage. Now Guardant’s test has the FDA’s signoff. They hold the potential to make cancer screening much easier for patients.
“A blood test is very attractive for many reasons … A lot of people really don’t like stool tests. They’re messy,” said Chin Hur, a physician specializing in gastroenterology at Columbia University Irving Medical Center, and the senior author on the JAMA paper. “And a colonoscopy is a big deal — it feels like surgery.”
Effective blood tests could significantly expand the number of patients who get screened amid a worrying uptick in cases, but experts stress that they need to be cheaper and more accurate to make a significant difference in improving the outlook for colon cancer patients.
To patients, liquid biopsies look no different from a normal blood test they would get at a doctor’s office to check out cholesterol, vitamin levels, or the makeup of their blood. In their case, liquid biopsies look for specific biological molecules that are left behind by cancer cells or that can otherwise indicate their presence in the body.
In a liquid biopsy, a patient’s blood sample is sent to a lab, where it is analyzed for biomarkers such as proteins, DNA, RNA, and other molecules that can be measured and correlated with possible disease. (Similar tests are being developed for urine and saliva.) Different biomarkers can be mixed and matched to create a signature for a specific cancer.
Screening, though, is a new use of this technology — Guardant and other companies like it started out by using these blood tests not as a way to initially detect and diagnose new cancers, but to track genetic changes in a known tumor in order to match them to targeted cancer treatments, a function known as companion diagnostics.
Guardant’s first liquid biopsy, Guardant 360 CDx, received FDA approval in 2020 as a companion diagnostic for several cancer therapies with specific targets. The test uses a patient’s blood sample to detect mutations in any of the 55 common tumor genes that would make the person a good candidate for particular medications.
For example, patients with advanced or metastatic lung cancer who test positive for the EGFR exon 20 insertion mutation would be eligible for a specialized therapy called ​​amivantamab. Specialization can make a difference in treatment success: In a 2016 review article in Annals of Translational Medicine, researchers found that cancer drugs used in tandem with a companion diagnostic had treatment response rates ranging from 41 percent to 80 percent. For cancer drugs used without a companion diagnostic, the range was 7 percent to 45 percent. (The response rate is the percent of patients who saw their tumor either shrink or disappear after treatment.)
Ultimately, Guardant wants to build blood-based diagnostic tests that can detect multiple cancers early in their development. The race is on for such a breakthrough: Academic researchers and other biotech companies like Grail are working on similar concepts.
Guardant’s new Shield test uses DNA shed by tumors, called circulating tumor DNA (ctDNA), to identify the presence of cancer. Since 1977, researchers have known that cancer patients have such cell-free DNA (which includes ctDNA) in their blood, but at the time, genetic sequencing tools were far too slow and too expensive to be used in patients, and were mostly reserved for preclinical research. By the 2000s, sequencing technology had become advanced enough to prove that these pieces of DNA had the same characteristic mutations as the tumor they originated from. Early-generation blood tests could use those biomarkers to assist doctors in selecting treatments, estimating a patient’s prognosis, and identifying residual disease or risk of relapse.
One of the limitations of the early tests, however, was that ctDNA levels tend to be low in early-stage cancers, and its concentration can be highly variable. That limited their value compared to other screening techniques; even if some of the latter tools are more laborious and expensive up front, they have a greater payoff because they can catch the cancer at an early point when it can be more effectively treated.
The US Preventive Services Task Force, an independent, volunteer panel of national experts in prevention and evidence-based medicine, recommends that healthy adults between 45 and 75 elect one of seven screening methods at regular intervals to reduce the risk of colorectal cancer. The Task Force lists out the pros and cons of each test, but ultimately leaves it up to the physician and patient to discuss which test is best for them.
The fecal immunochemical (FIT) test, a stool test for detecting human hemoglobin, costs about $24 with insurance and is recommended annually for all healthy adults between the ages of 45 and 75. Hemoglobin is a protein found in blood, and blood in stool may be a sign of colorectal cancer since “blood vessels in larger colorectal polyps or cancers are often fragile and easily damaged by the passage of stool,” according to the American Cancer Society.
Guardant recommends its new Shield tests for healthy adults between age 45 and 75 every three years, with an expected price tag of $895 per test, according to the Associated Press. That makes it more expensive than Cologuard, a stool-based DNA detection test for colon cancer that is recommended every 1 to 3 years with an out-of-pocket price range between $500 and $600. On the high end of the spectrum is a colonoscopy, which is recommended every 10 years starting at age 45 and has an average cost of around $2,750. The US Preventive Services Task Force also recommends that abnormal results from the more noninvasive tests should be followed up with a colonoscopy.
“When you look at cost-effectiveness, all of those tests are cost-effective when analyzed with their ability to reduce colon cancer mortality incidence,” Lieberman said. But because blood tests are currently less sensitive than the alternatives, they’re still too expensive to be as cost-effective on a dollar-by-dollar basis. “For a blood test to be equally cost-effective, our modeling suggests that it would probably have to be less than $100. That’s going to be a heavy lift because the development process for these tests was very expensive.”
There is also the risk of false positive tests driving unnecessary use of health care services. It could also cost patients extra medical bills associated with follow-up appointments that turn out to be unnecessary, not to mention the psychological distress that comes with thinking you have cancer. If a large number of patients who don’t actually have cancer receive a false positive test, “those patients are going to go on and get colonoscopies and that drives up the cost and potentially the risk of the program,” Lieberman noted. In their NEJM study, Guardant’s Shield test had a false positive rate of around 10 percent. FIT, by contrast, has a false positive rate of around 5 percent.
Although follow-up colonoscopies when done correctly can be lifesaving when cancer is present, studies have found that unnecessary colonoscopies can expose patients to high out-of-pocket costs from anesthesia and other related services as well as potential clinical harms like bleeding and GI tract perforation. UnitedHealthcare, one of the nation’s largest insurers, recently restricted its coverage of colonoscopies in response to what they say is overutilization of these procedures. The company then modified the policy after protests from patients and physician organizations, instead asking patients and physicians to provide “advanced notification” if, for example, they were scheduling a follow-up colonoscopy after receiving a positive result from a non-invasive stool or blood test.
Colonoscopies remain the primary form of colon cancer screening in the US. They are effective because they function both as a diagnostic and as an early form of treatment. The scope’s camera can clearly see polyps, or an out-of-place protrusion of cells, which can then be snipped off. Most polyps are harmless. But a type of polyp called an adenoma can potentially turn into cancer. Regular colonoscopies decrease colon cancer deaths between 29 to 68 percent, according to various studies.
In many countries including the US, UK, Australia, and India, there are preventive, government-overseen population-based cancer screening programs that look for early signs of common age- or gender-related diseases to improve health outcomes and decrease disease burden on the community. Most countries with such programs for colorectal cancer utilize stool tests performed yearly or every other year. Such tests look for DNA, blood, or proteins that can indicate the presence of a tumor. If a test comes back positive, the patient is supposed to follow up with a colonoscopy.
The FDA reviews new laboratory-developed tests and approves them if they’re safe and effective. The Centers for Medicare and Medicaid Services has a parallel but separate process for coverage of these tests, evaluating not only their quality and safety but also whether they are clinically necessary and cost-effective.
CMS has set a minimal sensitivity criteria for approving a new cancer blood test. Although Epi proColon was the first FDA-approved blood-based screening test for colon cancer, CMS declined to cover it. The test performed inconsistently during clinical trials, with an average sensitivity of 69 percent. (In medical tests, sensitivity is a test’s ability to accurately detect a positive case, while specificity is the ability to accurately detect a negative case.) In a decision memo in 2021, CMS stated that a blood test needs to surpass a sensitivity of 74 percent in order for it to be covered, using the current fecal and stool tests as benchmarks. It also needs a specificity of 90 percent.
From the NEJM data, Guardant’s Shield blood test had an 87.5 percent sensitivity for stage 1, 2, or 3 colorectal cancer, which passes the CMS threshold. But it has only a 13.2 percent sensitivity for advanced precancerous lesions. The fecal immunochemical test can detect these advanced precancerous adenomas with a sensitivity rate of around 20 to 30 percent. Cologuard, which combines FIT with stool DNA detection, has a 42 percent sensitivity for adenomas. Colonoscopies can detect adenomas larger than 6 millimeters with a 75 to 93 percent sensitivity. (Small polyps or outgrowths are usually harmless, and 6 millimeters is the standard lower limit for identifying advanced adenomas that are likely to grow into cancer.)
Detecting and cutting off such advanced adenomas is key to preventing colorectal cancer from developing. “If you can only detect cancer and you can’t detect pre-cancer, then it’s not going to be as effective,” Hur said.
Colon cancer forms on the interior surface, or lumen, of the colon “tube,” Hur explained. “As the stool is created, it sloughs off cancer cells or blood. If you’re trying to look for some type of cancer cell, circulating tumor cells, or DNA from a cell, stool is much better [than blood]. In order for it to be picked up by blood, it has to go into the bloodstream. Early cancer hasn’t invaded through the wall of the colon.”
Early detection matters because the more time a cancerous cluster of cells has to grow and spread, the harder it is to treat and cure. The five-year survival rates for colorectal cancer plummet drastically as it progresses past stage 3, falling from roughly 70 percent in stage 3 to 12 percent in stage 4. According to the American Cancer Society, stage 3 is when the colon cancer has spread to nearby lymph nodes, but not to other parts of the body. Stage 4 is when it is metastatic, and has spread to other tissues and organs in the body.
According to Lieberman, CMS didn’t set any minimum sensitivity rate at which new blood tests need to find advanced precancerous adenomas in order to approve them for coverage. In an email statement to Vox, a CMS spokesperson said that while they acknowledge the utility and potential clinical benefit of screening for advanced adenomas, there is limited available evidence on the accuracy of blood-based tests to detect advanced adenomas. The agency noted the sensitivity of fecal tests for advanced adenoma is also fairly low. For now, to maximize early pre-cancer detection, colonoscopies remain the best option for patients.
Experts believe that blood tests will become more accurate as science advances. But how much they can be improved will dictate their long-term value to medicine.
Some of that depends on the kind of cancer doctors are looking to find. Unsurprisingly, for liquid tumors that affect the blood, bone marrow, or lymphatic system, an early-detection liquid biopsy “totally makes sense,” Hur said. “The tumor is in the bloodstream.” But it’s unclear whether or not a liquid biopsy will ever be sensitively attuned enough that it will be able to consistently detect an adenoma inside a person’s colon.
There could be a scenario where scientists can identify a more specific immune response to a growing polyp that can be targeted with a blood test. But that carries a risk of more false positives, since a wide range of conditions, including something as innocuous as a cold, could trigger a similar immune response. Inflammation is our immune system’s response to trauma or the presence of a pathogen such as a virus, bacteria, or toxin. Scientists have known since the 19th century that cancer can cause an inflammatory response that sends specialized immune cells into disarray. Elevated white blood cells are a telltale sign of all kinds of infection, but they can also be a sign of leukemia.
Despite the hype, experts caution that additional research is needed to improve liquid biopsies. If scientists can learn more about the biology and genomics of tumors, they could identify more specific biomarkers or a combination of biomarkers that could make these tests more accurate and therefore more valuable.
Lieberman acknowledged that scientists are becoming more sophisticated in how they analyze the genomes of cancer cells, which means they could identify more genes that are present in patients at higher risk of colon cancer, and that can be detected before the cancer starts to develop.
“This is still a ripe new area for development,” he said. “I think this first [Guardant] study is exciting in the sense that we can detect cell-free DNA in patients with colon cancer at an early enough stage that it can make a difference for those individuals.”
In the long run, blood might not always be the most revealing substance circulating through the human body when it comes to finding early markers of disease. The accuracy of blood tests will likely vary based on the type of cancer and the organs in which they’re located. For cancers in which blood tests are less effective, the goal is still to develop less invasive diagnostics, but the mechanism might be different.
Urine is also being investigated for its potential to detect circulating tumor DNA in urological and non-urological cancers. It has already shown promising sensitivity rates for non-small cell lung cancer and breast cancer in clinical trials. In an April study in JAMA, a urine test that scans for 18 genes showed a 95 percent sensitivity in patients with more aggressive prostate cancer. In January, a urine-based test using bioengineered bacteria was able to detect colorectal adenomas in mice.
Despite the limited utility of the current tests, the pursuit of better liquid biopsies is unlikely to slow down because they continue to hold enormous theoretical potential. For cancers with no established screening tests at all — like pancreatic cancer, which has an average five-year survival rate of 13 percent — a viable blood test with clinically significant sensitivity for early signs of disease would be a game-changer.
Earlier in April, a new study presented at the American Association for Cancer Research’s annual meeting found that an exosome-based blood test combined with another protein as an added biomarker identified stage 1 and stage 2 pancreatic cancer with 97 percent accuracy. Cells use tiny vesicles called exosomes to communicate with one another, and they’re commonly shed by both cancer cells and healthy cells. Researchers developed a signature associated with pancreatic cancer based on eight microRNAs “uniquely found in exosomes shed from pancreatic cancer” and five cell-free DNA markers found in the blood of patients with pancreatic cancer, according to AACR.
An advantage of these noninvasive blood and stool tests is that they’re a relatively cheap way to get a lot of people screened. However, much of the value of these tests is lost if patients who receive an abnormal result don’t follow up with a colonoscopy to confirm whether or not they have cancer.
“The Achilles’ heel of noninvasive testing is that many patients don’t follow up and get the colonoscopy after they get a positive stool test,” Lieberman said.
A 2023 study in JAMA found that only 56 percent of patients followed up with a colonoscopy after receiving a positive stool-based test. “We don’t really know yet what it will be like with a positive blood test,” he said.
Even with the limitations, there are upsides to having a blood test as another screening option. Prepping the bowels in the lead-up to a colonoscopy is unpleasant and inconvenient, requiring a special preparation for days before the procedure. And too often, people avoid them altogether — the US adherence rate for colorectal cancer screening guidelines is around 65 percent. With the recommended age for colorectal cancer screening recently dropping from 50 to 45, there may be more demand for a less invasive test.
Fast and easy blood tests, along with proper education on the importance of screening for colorectal cancer, could also be beneficial in regard to achieving health equity. Colon cancer deaths, estimated by the American Cancer Society to reach 53,010 in 2024, are highest in Black patients, and screening rates are lower compared to white patients.
“There’s an expression in colon cancer screening: The best colon cancer screening test is the one that gets done,” Hur said. “There’s potential to do good. … People want this … They don’t want to do the scope test. But people just need to know what they’re getting.”
Update, July 29, 10:00 am ET: This story was originally published on June 4 and has been updated to include the recent FDA approval of Guardant Health’s Shield blood test.