Why is stool such a difficult non-invasive starting material for early colorectal cancer detection?

Fecal Immunohistochemistry Tests (FIT) and guiac-based Fecal Occult Blood tests (gFOBT) are inexpensive and widely available tests for colorectal cancer screening. However there remains a clear need for an accurate blood-based colorectal screening test.

A stool collection vial for colorectal cancer screening

Over 148,000 in the United States will be diagnosed with colorectal cancer (CRC) this year, and over 50,000 will die from this disease. Early detection is vitally important: those diagnosed with localized CRC have a 5-year survival rate of 89.8%; those with advanced metastatic CRC, that 5-year survival rate is only 13.8%.

The current colorectal cancer (CRC) landscape

As colon cancer often goes undetected (the majority of all CRC diagnoses are Stage III and advanced Stage IV CRC) it is estimated that a full 60% of the 50,000 deaths from CRC every year would be prevented if CRC screening were more ubiquitous. That is a full 30,00 deaths every year that could be prevented, if everyone were screened with the current established guidelines.

In 2013 the CRC screening rate compliance in the United States is 58.6% (See Table 7 In this 2016 reference). The US Preventative Services Task Force recommends several methods of CRC screening in two categories: stool-based tests or direct visualization tests. For the first, non-invasive stool tests are the guiac-based fecal occult blood test (gFOBT), the fecal immunohistochemistry test (FIT), or the DNA-FIT test (the Exact Sciences’ ColoGuard test). These are to be taken yearly (or in the case of DNA-FT every three years as recommended by the manufactuer). The direct visualization tests include colonoscopy (every 10 years), Computed Tomography (CT) colonography, Flexible Sigmoidoscopy (these every 5 years), or Flexible Sigmoidoscopy with FIT (sigmoidoscopy every 10 years plus FIT every year).  The current June 2016 USPSTF recommendations for CRC screening can be found here, and the full recommendation statement here.

The colonoscopy is considered both a diagnostic test and a therapeutic; should a pre-cancerous polyp be discovered it is removed simultaneously. This is the great thing about a colonoscopy (called the ‘gold standard’ for CRC diagnostics): if a precancerous polyp is discovered, the patient is effectively cured before cancer develops. For advanced adenoma, a full 85% to 90% of these polyps develop into cancer. The cost of a colonoscopy (covered by insurance in the United States) is about $1,600; however some find the preparation for a colonoscopy a strong deterrent.

Television host Katie Couric, whose husband passed away in 1998 after only 9 months from being diagnosed with metastatic colorectal cancer, co-founded the National Colorectal Cancer Research Alliance, as part of the Entertainment Industry Foundation. She then publicized the importance of colonoscopy screening in a ‘Today Show’ screening that televised her colonoscopy, “from my downing the prep to Dr Ken Forde inspecting every nook and cranny of what I called (blame the drugs) my 'pretty little colon'.” She continues her CRC screening advocacy; even with this and other efforts, the screening rate of less than 60% persists.

The need for frequent fecal occult blood tests

Guiac-based fecal occult blood tests (gFOBT, occult here means ‘hidden’) and fecal immunohistochemistry tests (FIT) look for signs of blood with an inexpensive $20 assay. They need to be re-taken every year per screening guidelines, due to their relatively poor sensitivity and specificity.

The Exact Sciences ColoGuard test is an at-home sample collection with central processing in the company’s laboratories in Wisconsin. Relatively expensive (they cost about $500 but covered by insurance) they need to be re-taken every three years per screening guidelines. Their test tests not only for the presence of blood but also specific DNA markers (mutations in the KRAS gene) and CpG markers of specific genes (eight CpG sites in the promoter of the BMP3 gene, and nine CpG sites in the promoter of the NDRG4 gene, where high methylation status is positively correlated with cancerous cells assayed by real-time PCR). Exact Sciences recommends their test every three years.

Natural reluctance to manipulate stool

The Colorectal Cancer Association of Canada has run an advertising campaign titled “Get Your Butt Seen”  to address the low rates of colorectal cancer screening in that country, estimated to be only 4 to 14 percent of the eligible population.  The reluctance of working with stool and embarrassment to discuss symptoms of CRC with their physician mean needless deaths from a preventable cancer.

Exact Sciences have produced a cheerful animated box in their direct-to-consumer advertising, with the tagline “Get, Go, Gone”, a creative way to look at it. Take a look at a few of their cute commercials for consumers here. The ColoGuard collection kits are provided by physician prescription only, yet a full one out of every three kits that go out (35%) do not come back to the company. Regardless of the education around the need for screening, there’s no getting around the ColoGuard protocol: after the stool sample is deposited into a collection cup, the instructions state to use a brush-like device to gently scrape the outside of the stool and place the brush into a preservative solution. (If you are curious, an instruction video is provided by Exact Sciences here, under ‘watch how to use the kit’.)

The general public’s resistance to manipulate stool is not limited to the United States. In this study in Germany, only 14% of men and 22% of women had a gFOBT test within the past year.

In addition to recommending more frequent stool testing, the stool tests themselves suffer from less than 100% specificity (false-negatives, where the test comes back as cancer-free but the individual has cancer). As an example ColoGuard has a specificity for colorectal cancer of 93%, which means for every 93 individuals who test negative, the test has missed an additional 7 individuals who have colorectal cancer but the test did not detect it.

In addition, ColoGuard has a specificity for advanced adenoma (likely to become cancer) of 70% (i.e. for every 70 individuals who test negative another 30 individuals are missed), and for polyps (that may become cancer) the specificity drops to 43% (missing 57 individuals where 43 are negative).

An incomplete screen for Colorectal Cancer is better than none

Therefore the 65% of the ColoGuard collection kits that eventually get returned will at least pick up CRC 93% of the time, and colonoscopy as the reflex test (to confirm the diagnosis) is well accepted and part of the current USPSTF recommendation to healthcare professionals. Remember these are otherwise healthy, asymptomatic individuals where CRC is detected early.

And the inexpensive fecal tests (gFOBT and FIT), while the performance in specificity is worse, nonetheless can flag CRC as an early-detection screen. As the reflex test (and therapeutic in the case of advanced adenoma or polyps being removed), colonoscopy is very safe and highly effective.

The US PSTF points out that about one-third of eligible adults in the US have never been screened for CRC, and present the screening tests in no particular preferred or ranked order; 'rather, the goal is to maximize the total number of persons who are screened because that will have the largest effect on reducing colorectal cancer deaths'. To repeat, these methods, if adopted more widely, have the power to reduce CRC deaths by 60% if wider screening were in place.

Singlera’s ColonES uses blood as the preferred starting material for CRC screening

The same German study mentioned earlier found a full 97% of those who refused colonoscopy indicated they would be willing to give a blood sample for CRC screening. With the ubiquity of blood draws as part of routine clinical diagnostics, a highly specific (high true-negative rate) and a highly sensitive (high true-positive rate) blood test is preferred, and would achieve much higher CRC screening compliance.

Singlera’s ColonES has demonstrated 99% specificity in a recent study of 1283 enrolled participants (PDF of Product Sheet, manuscript in preparation). This means that for every 99 individuals who test negative, only 1 is positive for CRC that the test did not detect. For Stage I CRC, the Singlera ColonES assay has a sensitivity of 97%, meaning for every 97 individuals who test positive, there are 3 individuals who test positive but do not have CRC.

A Michigan gastroenterologist named Naresh Gunaratnam wrote the following in a piece in Stat News titled “Colorectal cancer screening: Science should trump convenience”

“Every time I diagnose colorectal cancer, I feel the system has failed. Some day we may have a simple blood, saliva or stool test that can detect colorectal cancer and polyps better than colonoscopy.”

With both high accuracy and the convenience of a blood draw, that day could be coming soon.

References

1. Couric K. Am. J. Gastroenterol. 2016 An Unexpected Turn: My Life as a Cancer Advocate. PMID: 27021192 https://www.ncbi.nlm.nih.gov/pubmed/27021192
2. Smith R.A. and Wender R.C. et al 2016 CA Cancer J. Clin. Cancer screening in the United States, 2016: A review of current American Cancer Society guidelines and current issues in cancer screening. PMID: 26797525 https://www.ncbi.nlm.nih.gov/pubmed/26797525/
3. Adler A. and Wiedenmann B. et al 2014 Improving compliance to colorectal cancer screening using blood and stool based tests in patients refusing screening colonoscopy in Germany PMID: 25326034 https://www.ncbi.nlm.nih.gov/pubmed/25326034