By Steven H. Itzkowitz, Professor of Medicine and Oncological Sciences, Icahn School of Medicine at Mount Sinai New York, USA

Introduction by Dr Jerome Waye, WEO Past President

Throughout the world, the rising incidence and mortality of colon cancer has prompted a movement for screening to discover early cancer and to find and remove adenomas, precursors of bowel cancer. Many parts of the world have not accepted colonoscopy as the tool to use because it is invasive, requires sedation, is expensive and uses costly instruments requiring knowledgeable practitioners. Other screening methods are being investigated and Dr. S. Itzkowitz has pointed out advantages of multi-target stool DNA testing, a non-invasive test that has the potential to provide a risk-free alternative to colonoscopy screening, especially where facilities and trained endoscopists are not widely available.

Over the last 1-2 decades, colonoscopy has been increasingly used as the preferred colorectal cancer (CRC) screening test. No doubt, colonoscopy is a remarkable and extremely important screening test because of its very high sensitivity and specificity, the ability to detect and remove precancerous polyps, and the conveniently long interval before another colonoscopy is needed after a normal exam.


However, we must acknowledge that screening colonoscopy (SC) has its limitations. These include: 1) considerable operator dependence (hence, the enormous international effort to improve colonoscopy quality); 2) high patient costs with some insurance plans; 3) a substantial gap in SC usage among uninsured individuals; 4) a paucity of endoscopists in many regions of the world; 5) the small but finite miss rate for polyps and even cancers; 6) a small but measurable rate of complications. Adding to the risk of colonoscopy as a routine screening test is the fact that many patients referred for SC have multiple serious co-morbidities and/or are taking numerous medications, especially anticoagulants. From the patient’s perspective, limitations include: dislike and inconvenience of a bowel prep, loss of time from work/caretaking, and need for an escort since most procedures are done under sedation.


Thus, as it turns out, at least one-third of screen-eligible individuals are not getting SC or any other CRC screening test. And this is not for lack of trying. For example, in New York City, with considerable endoscopic capacity and systems in place, SC rates seem to be plateauing at approximately 69-70%. As such, a national movement was developed, and is being promoted, to increase CRC screening rates to “80% by 2018” (1). So, how can we get there? In my opinion, we need to offer individuals something different than what we’ve been doing all along.

A new, FDA approved, multi-target stool DNA test (called “Cologuard®”) offers a very viable screening alternative. Unlike fecal immunochemical tests (FIT) that detect only occult blood in the stool, Cologuard® detects abnormal DNA in addition to fecal hemoglobin. In the seminal Deep-C study, the stool DNA test was compared to the most commonly used commercial FIT in over 10,000 asymptomatic average-risk individuals who very closely reflected the general U.S. population in terms of age, gender and ethnicity (2). The stool DNA test demonstrated impressive point sensitivities of 92% for CRC and 69% for adenomas with high-grade dysplasia. And, for essentially all sensitivity parameters by cancer stage and adenoma subtype, the stool DNA test outperformed FIT. However, its specificity was lower than FIT. The specificity of the multi-target stool DNA test for individuals with normal colonoscopies was 90%; i.e. 10% false-positives. In part, this is because epithelial cell DNA undergoes increased methylation with normal aging. Thus, there are more false-positives in older age groups. Importantly, only 6% of individuals 50-65 years old had a false positive stool DNA test, which is close to that of FIT. If the stool DNA test is performed every 3 years, the programmatic specificity and sensitivity should be even higher. Cancers of the upper GI tract cancer rarely cause a false positive Cologuard®, based on preliminary studies (

This new test is an important advance for several reasons. First, many cancers and most adenomas, do not bleed sufficiently or consistently enough to be detected by fecal hemoglobin alone. By contrast, DNA is continuously shed from neoplasms into the lumen due to aberrant cell death. Also, DNA comes from epithelial cells. Since the epithelial layer of adenomas is comprised of tubules representing an extensive invagination of epithelial cells, the effective surface area of a 2 cm adenoma is actually 200 times larger than it appears through the scope (3). And, as a polyp grows, the ability to detect it by multi-target stool DNA testing increases substantially. Second, DNA is amplifiable, so even the very small amounts of human DNA in stool can be detected using sensitive PCR techniques. Third, sessile serrated polyps essentially never bleed but often express hypermethylated DNA. Stool DNA testing detected 45% of sessile serrated polyps ≥1 cm compared to 5% detection by FIT. Fourth, FIT and guaiac FOBTs are notoriously poor at detecting proximal CRC, whereas Cologuard® detects cancers equally well in the proximal and distal colon.

Multi-target stool DNA testing offers convenience from the patient’s perspective. The provider orders the test directly with the lab. The collection kit is mailed from the lab to the patient’s home. The patient collects the stool in a container mounted on the toilet, and mails it back to the lab using a prepaid air-bill. It is safe, and the patient does not have to deal with bowel prep, dietary/medication restrictions, work absence, escort, or returning the test to the clinic. And, a patient navigation system is in place to make sure that the specimen is performed by the patient, received by the lab, and the doctor’s office is notified of the results. Although not yet tested, a 3-year interval for stool DNA testing has been recommended. This may be preferred by patients and physicians to the inconvenience of performing FIT annually (and studies show a marked lack of adherence to annual FIT testing). Finally, compared to not screening, the $600 cost, modeled at 3-year intervals, is cost-effective (4).

We do not yet know how stool DNA testing compares to colonoscopy as a primary screening tool. Will patients with a negative test avoid doing colonoscopy? Of note, if multi-target stool DNA testing is negative, there is only a 0.06% chance that there is cancer. Will endoscopists faced with a positive Cologuard® be more careful when performing a SC, especially in seeking proximal neoplasia? It will be interesting to see how multi-targeted stool DNA testing factors into the landscape of CRC screening.

1. National Colorectal Cancer Roundtable 80% by 2018 fact sheet; 2014.
2. Imperiale TF, Ransohoff DF, Itzkowitz SH, Levin TR, Lavin P, Lidgard GP, Ahlquist DA, Berger BM. Multitarget stool DNA testing for colorectal-cancer screening. N Engl J Med 2014;370:1287-97.
3. Berger BM and Ahlquist DA. Stool DNA screening for colorectal neoplasia: biological and technical basis for high detection rates. Pathology 2012;44:80-8.
4. Berger BM, Schroy III PC, Dinh TA. Screening for colorectal cancer using a multitarget stool DNA test: modeling the effect of the interest interval on clinical effectiveness. Clin Colorectal Cancer, 2015 (in press).