Colorectal Cancer Screening
Carolyn Vachani, RN, MSN, AOCN
The Abramson Cancer Center of the University of Pennsylvania
Last Modified: February 20, 2009
There are approximately 148,000 new cases of colorectal cancer annually in the United States, with approximately 50,000 people who die from the cancer or its complications every year. Over a lifetime, women have a 5% risk of developing colorectal cancer and men carry a 6% risk. This risk is substantially increased if an individual has a family history of colorectal cancer, especially in the well-defined inherited cases of familial adenomatous polyposis (FAP) and hereditary nonpolyposis colorectal cancer (HNPCC).
The vast majority of colorectal cancer cases arise from the progression from normal colonic mucosa (meaning that the lining of the colon is normal) to adenomatous polyp (a precancerous growth) to cancer, a process that takes 10-15 years or longer to occur. Patients may have no symptoms at all or have a combination of symptoms such as blood loss, weight loss or a change in bowel habits. With advancing age, especially individuals in their 50's, 60's, 70's and 80's, there is an increased incidence of polyps and cancer. It is important to detect polyp(s) before they become cancerous and/or to detect cancer in its earliest stage. When colorectal cancers are detected in the earliest stage, the five year survival rate is 90%. This rate drops to 68% when the disease has spread to the lymph nodes, and is only 10% if the cancer is present in other areas of the body (called metastasis).
How can we find these pre-cancerous polyps or early cancers? Screening. Today there are a number of tests available to screen for colorectal cancers, but not all tests are created equal. Patient preference, test availability and cost all play a role in choosing the best test for an individual. The American Cancer Society (ACS), the US Multi-Society Task Force on Colorectal Cancer (USMSTF) and the American College of Radiology have developed guidelines for screening utilizing the available tests including when and how often to perform them. Let’s review what tests are available, how they are performed and how well they work.
Screening with Stool Testing
Stool testing is broken into two groups- those that test for the presence of blood and those that identify abnormal DNA present in cancer or precancerous polyps. One limitation of both of these types of tests is that they are most successful in detecting cancers or larger, advanced polyps (polyps on their way to cancer). They are not good at detecting the smaller, pre-cancerous polyps that other available tests can detect.
Fecal Occult Blood Testing
The least expensive method of colorectal cancer (CRC) screening is the fecal occult blood tests (FOBT). These tests take a sample of stool and using a chemical solution, detect the presence of blood in the stool. Studies have found that, when performed properly, these tests can detect 24 to 40% of colorectal cancers (this does not include the polyps). The test seems simple enough, but there are a few guidelines. The person must refrain from taking aspirin and NSAID products (ibuprophen, naprosyn) for 7 days prior to testing. Dietary restrictions include avoiding vitamin C, iron, eating red meat and some raw vegetables (including broccoli, beets, bananas) for 3 days prior to testing. These restrictions decrease the chance of having a false positive result. In addition, the test must be performed on two or three consecutive bowel movements and repeated every year. If the test is positive, the person must be willing to undergo a colonoscopy to find and treat the cause of the positive test. While some physicians will perform a FOBT at the time of a rectal exam, this is considered ineffective and is not supported by the ACS or the USMSTF. In addition, the reliability of stool blood testing varies greatly from one brand of test to another and is also affected by the procedure for processing the test.
Fecal Immunochemical Test
FIT (fecal immunochemical test) is a test that uses a different technique to detect blood in the stool. Because the test is more specific for human blood, it is not affected by diet and patients do not need to follow the above dietary restrictions. The test is more expensive than FOBT and it is not clear yet if it is superior. While it appears that 2 FIT tests result in better detection than 1, the optimal number of tests per year is not yet known.
Stool testing for blood is certainly better than no screening at all, but the concern is that advanced adenomas or early cancers are unlikely to bleed continuously, therefore not detecting blood in the stool may provide false reassurance when, in fact, a cancer does exist. The tests must be performed more than once and repeated annually in order to have the best chance of detecting a cancer. In addition, these tests are less effective in detecting tumors on the right side of the colon than on the left. This is important to note because studies have found that over 50% of cancers are found on the right side of the colon in women and on the left side in men.
Stool DNA Testing
A third type of stool test, called sDNA, detects the DNA changes that occur in the bowel when cancer or adenomas are present. These cells contain abnormal DNA, which is continuously shed from the lining of the bowel and passed in the stool. Since each CRC can contain different DNA changes, the test looks for a number of abnormalities, but does not detect all the possible DNA changes found in different people, so some tumors going undetected. While the stool blood detection tests use a small sample of stool, sDNA requires the entire stool specimen be submitted for testing using a customized collection kit from the company (only one test is commercially available, with one other in development).
There is one commercially available sDNA test, called ColoSure. How well does it work? Studies have found that it has about 70% sensitivity. This means that it detected cancer in 70 out of 100 people with a cancer. On the flip side, it has a 97% specificity, meaning that 3% of the time the patient will receive false positive result and no cancer will be found on further testing. The significance of a false positive result is not yet known, as this may be the result of a cancer being present, but not detected, on colonoscopy or possibly a cancer in another area of the gastrointestinal tract (small bowel, stomach, etc.). These sensitivity and specificity results are for colorectal cancers and studies have yet to sufficiently evaluate the effectiveness in detecting adenomas (pre-cancers), though it appears to be far less sensitive for these. The test is also less sensitive in detecting tumors on the right side of the colon than on the left. The manufacturer recommends the test be performed every 5 years, but experts feel that, at this time, there is insufficient evidence for this and the optimal interval is unknown.
Cost and availability are two additional issues. The cost is currently $400-500 per test, which can be submitted by the patient to their insurance company, but coverage for this test varies by company and from case to case. Due to regulatory approvals, the test is not available in all states at this point. Despite these limitations, the ACS and USMSTF have added stool DNA testing to its acceptable options for colorectal cancer screening.
Screening with Endoscopy
Endoscopy uses a camera on the end of a thin, flexible tube to navigate the bowel and look for any polyps or tumors. There are two types of endoscopy available, sigmoidoscopy and colonoscopy.
FOBT in conjunction with a flexible sigmoidoscopy has been found to detect about 76% of colon tumors. The sigmoidoscope is a slender, flexible tube that has the ability to view about 1/3 of the colon (the left side). If a polyp or tumor is detected with this test, some physicians are able to perform a biopsy with sigmoidoscopy, while others must refer the patient for a full colonoscopy and biopsy. However, if an adenomatous polyp is found during flexible sigmoidoscopy, then colonoscopy is recommended because of the increased risk of an adenoma on the right side of the colon. It is then recommended that subsequent surveillance be done with colonoscopy. Flexible sigmoidoscopy requires a more limited bowel prep than colonoscopy (2 enemas) and does not require sedation, therefore it may be performed in a physician’s office and does not require recovery time. The test is only as good as the person performing it (their training and adherence to guidelines for the test) and the preparation of the patient (a poorly done bowel prep will limit the sensitivity of the test). The main risk to sigmoidoscopy is perforation of the bowel, which can occur with or without a biopsy. This is quite rare, occurring in fewer than one in 20,000 procedures. Sigmoidoscopy has become less common over the last decade (Medicare data saw a 54% reduction), being replaced by colonoscopy.
When used for CRC screening, the ACS and USMSTF recommend sigmoidoscopy be performed every 5 years.
The colonoscope is similar to the sigmoidoscope, but is longer and thus can view the entire colon (left and right sides). If a polyp is found, the physician can remove it using a cutting tool contained within the scope and send it to a pathology lab to determine if it is adenomatous (precancerous). Colonoscopy is considered the gold standard of colorectal cancer screening, but is only as good as the endoscopist performing the test. Studies have shown the miss rate for large adenomas (>1cm) to be 6 to 12% and about 5% for cancers, so the test is not perfect. Given the slow growth of polyps, it is recommended that a normal exam (in the general population) be repeated every 10 years. Those at higher risk (i.e. family history, personal history of ulcerative colitis, HNPCC or FAP) should be screened more frequently.
To prepare for colonoscopy, the patient must follow a liquid diet for one or more days prior and perform a bowel cleansing, which is done with oral laxative solutions. The bowel is inflated with air to allow the endoscopist to better visualize the colon. The test typically uses sedation, requiring the patient to have a chaperone for transportation and to take the day off from work. Again, the test is only as good as the person performing it, yet no quality assurance programs exist, so the patient is usually unaware of the skill of the endoscopist. Research has shown that endoscopists who take their time withdrawing the scope perform better tests. Studies have found a withdrawl time of at least 6 minutes identifies more polyps than a shorter withdrawl time.
If polyps are found on colonoscopy, when should the next exam be performed? Those considered to be at increased risk for future findings have either 3 or more adenomas, high-grade dysplasia, villous features, or an adenoma 1 cm or larger in size. It is recommended by the USMSTF that they have a 3-year follow-up colonoscopy, with subsequent colonoscopy intervals dependent on those results. Those at lower risk can have follow up in 5 to 10 years, whereas people with hyperplastic polyps only should have a 10-year follow up as average-risk people. More frequent screening is recommended for people with suspected or proven genetic syndromes (HNPCC and FAP). It is unclear how a family history without a genetic syndrome should affect screening intervals.
The main risks of colonoscopy are bleeding after removal of a polyp and perforation of the bowel, which is estimated to occur in 1 out of every 1000 procedures. The risk is higher in those with diverticulitis and increases as people get older.
Additional Screening Tests
Virtual colonoscopy (VC), a method of viewing the colon from outside the body, employs the use of CT scan and was first suggested in the early 1980s. Throughout the procedure, the patient is lying on a table that passes through a donut-like machine that takes pictures from different angles around the body. The 2 dimensional images of the colon are converted by a computer to three-dimensional images and then reviewed by a trained radiologist or gastroenterologist. Some imaging software can interpret the scan into a “fly-through” view, which looks on the screen a lot like what is seen with traditional colonoscopy (TC).
Though the actual time in the scanner is only about 10 minutes, the entire test takes approximately 30 minutes, including reading and interpretation, which does add to the cost of the procedure. In order for these images to be accurate, patients must undergo bowel prep similar to TC (dietary restrictions and oral laxatives), but for the virtual scan the bowel may need to be a bit clearer. This is because in a TC, the physician can clear away any stool remnant that is obscuring their view, but they cannot do this with VC. Some centers are studying the use of oral radio contrast agents that cause any stool remnants to easily differentiated on the CT (known as stool tagging), but this has not become standard yet. Just prior to the CT scan, a tube is inserted into the rectum and air or carbon dioxide is pumped into the colon, expanding it to allow for better visualization of the bowel wall. In some centers, patients are given an intravenous medication called glucagon, causing the bowel walls to relax and improve visualization, but this is becoming less common. No sedation is used, so patients do not require any recovery time, but some report more discomfort than TC.
The technology of VC has improved rapidly and the test appears to be quite good at detecting cancers and larger polyps. An analysis of the many available studies found a 96% sensitivity for cancers and 85 to 93% sensitivity for large polyps, while the sensitivity for smaller polyps (6-9mm) dropped to 70 to 86%. Some physicians have suggested that these smaller polyps are less likely to progress to cancer, and therefore the decreased ability to detect them may not be significant. VC is less able to detect flat lesions (also called non-polypoid lesions). A recent study of U.S. Veterans found that 9% of patients had flat lesions and that these were 5 times more likely than polyps to contain cancer or precancerous cells. So the decreased sensitivity for flat lesions is an important drawback to VC.
As with most colorectal cancer screening tests, it is only as good as the operator and many of the studies were performed with the best equipment and well trained radiologists, which may overestimate how good the test is in general. There are efforts underway to formalize certification of radiologists to give consumers some quality assurance. If polyps are detected, the patient may be referred for TC for polyp removal and biopsy. This is standard for polyps > 10mm or 3 or more polyps > 6mm, but the need is uncertain for fewer small polyps and these people may be followed with surveillance. While some institutions may coordinate the TC to happen the same or next day, avoiding an additional bowel prep, this is a difficult feat in scheduling and the additional test adds another concern to the cost effectiveness of this technology.
There is a very small risk of bowel perforation (estimated to be less than 0.05%) due to the instillation of air into the bowel. There is concern among some experts about the lifetime dose of radiation received during this and other radiology exams, but this risk is not yet well understood. Because the test is a CT scan and includes other areas of the abdomen, findings outside the colon may necessitate further workup. The test has received approval for reimbursement by Medicare in many states, but only for limited indications, and most private insurers have yet to get on the VC bandwagon. The cost for VC is around $500. As a screening method, VC should begin for average risk individuals at age 50, and although the optimal interval between screening has not been studied, experts recommend the test be performed every 5 years.
Double Contrast Barium Enema
The double contrast barium enema (DCBE) is performed by inserting a small tube into the rectum and coating the inside of the colon with barium (a contrast agent) and pumping in air to distend the colon. X-rays are then taken in different positions to evaluate the lining of the colon. The patient does need to undergo bowel preparation, similar to that used with traditional colonoscopy, with diet changes and oral laxatives. No sedation is used for the procedure and patients may experience discomfort during or after the test, which takes 20 to 40 minutes. The test does evaluate the entire colon and appears able to detect most cancers and a majority of significant polyps, though formal clinical studies have not been done. The test may be a good option for people in whom a colonoscopy could not be completed (i.e. due to a blockage) or was contraindicated. Recent studies, however have found barium enema to be far inferior to TC and has led some authorities to remove the test from their recommending screening strategies, including the U.S. Preventative Services Task Force recommendations.
If a polyp greater than 5mm is detected, a colonoscopy is needed to biopsy this finding. As with many colorectal cancer screening tests, the test is only as good as the radiologists performing and reading the test and the patient’s bowel preparation. Because the colon is being examined from the outside, stool remnants can be mistaken for polyps or conceal significant findings. The test is safe, though bowel perforation remains a low risk, estimated at 1 in 25,000 procedures. The use of the DCBE has decreased over the past decade, with the test being largely replaced with colonoscopy. When used for colorectal cancer screening, it is recommended that the DCBE be performed every 5 years in average risk adults over 50 years of age.
The guidelines for colorectal cancer screening are developed by the ACS and the USMSTF. People of average risk should begin screening at age 50. Keeping in mind that the best screening test is one that is completed, patient preference, availability of testing and cost / insurance coverage should play a role in choosing one of the following screening tests.
- flexible sigmoidoscopy, performed every 5 years
- colonoscopy, performed every 10 years
- double contrast barium enema, performed every 5 years
- CT colonography (virtual colonoscopy), performed every 5 years
- fecal occult blood test (FOBT), performed every year
- fecal immunochemical test (FIT), performed every year
- stool DNA test (sDNA), interval for testing uncertain at this time
The interval of further testing and the best test for a particular patient should be addressed if screening detects polyps or cancer. Patients should discuss their personal risk with their physicians, but the following factors increase risk and therefore increase the screening recommendations:
- Personal history of an adenomatous polyp or colorectal cancer
- Presence of inflammatory bowel disease or Crohn’s disease
- If a family member has had colorectal cancer or adenomatous polyps, and if so, depending on how many and at what age. First degree relatives are most important in risk assessment (parent, sibling, child). Screening generally begins 10 years prior to the age of relative’s diagnosis.
- If the family has a known or suspected genetic syndrome (HNPCC or FAP)
Levin, B et al. (2008) Screening and Surveillence for the Early Detection of Colorectal Cancer and Adenomatous Polyps, 2008: A Joint Guideline from the American Cancer Society, the US Multi-Society Task Force on Colorectal Cancer and the American College of Radiology. Gastroenterology 134: 1570-1595.
Rex, DK et al. (2006) Guidelines for Colonoscopy Surveillance After Cancer Resection: A Consensus Update by the American Cancer Society and the US Multi-Society Task Force on Colorectal Cancer. Gastroenterology 130: 1865-1871.
Soetikno, RM et al. (2008) Prevalence of Nonpolypoid Colorectal Neoplasms in Asymptomatic and Symptomatic Adults. JAMA 299(9): 1027-1035.
Whitney, D et al. (2004) Enhanced Retrieval of DNA from Human Fecal Samples Results in Improved Performance of Colorectal Cancer Screening Test. Journal of Molecular Diagnostics 6: 386-395.
Winawer, SJ et al. (2006) Guidelines for Colonoscopy Surveillance after Polypectomy: A Consensus Update by the US Multi-Society Task Force on Colorectal Cancer and the American Cancer Society. CA Cancer J Clin 56: 143-159.