The U.S. Multi-Society Task Force on Colorectal Cancer, the American Cancer Society, and the American College of Radiology have each published colorectal cancer (CRC) screening guidelines before, but Screening and Surveillance for the Early Detection of Colorectal Cancer and Adenomatous Polyps, 2008: A Joint Guideline represents the first such document carrying the imprimatur of all three groups.
Besides adding two newcomers (stool DNA [sDNA] and CT colonography [CTC]) to the list of recommended tests, the participating organizations take for the first time the unambiguous position that preventing CRC, rather than detecting it early, should be the primary goal of screening.
Primary-care providers (PCPs) have an “essential and often underappreciated” role in screening, says Durado Brooks, MD, director of prostate and colorectal cancer in the cancer control department of the American Cancer Society and an author of the Guideline. “When studies have looked at people who haven't been screened for colorectal cancer, a reason often given is ‘my doctor never talked to me about it.' Many PCPs don't realize their patients are waiting for them to bring up the topic.”
While choosing among tests is the patient's prerogative, “there are quite a few tests, and patients may not appreciate the differences among them,” Dr. Brooks comments. The PCP is often in the best position to explain the options, and the Guideline is intended to inform this “shared decision making,” he says.
Setting priorities
Where earlier guidelines presented a menu of options, these prioritize them. “It is clear that all CRC screening tests are not created equal,” Dr. Brooks cautions. They fall into two categories: those highly likely to find cancer (fecal tests) and those that effectively identify precancerous polyps (“structural exams”) as well (see Table 1). Finding and removing such lesions can prevent cancer, he says.
“We've stated emphatically that if resources are available and patients are willing to undergo these somewhat more invasive tests, those that find both polyps and cancer at a high rate are recommended,” he says.
The Guideline also gives new emphasis to questions of quality and factors that influence sensitivity.
Fecal tests
Quality assurance is a particular issue with guaiac fecal occult blood testing (gFOBT): Sensitivity and specificity appear highly dependent on brand, sample collection, preparation, and interpretation. Some older variants (e.g., unrehydrated Hemoccult II) miss most cancers and are not recommended.
Fecal immunochemical testing (FIT) is as sensitive as gFOBT but considerably more specific: False positives related to diet and vitamin C are unlikely. Collection techniques (fewer stool samples, less handling) may make some variants of FIT more acceptable to patients.
When considered several years ago by the organizations that sponsored the Guideline, sDNA testing, like CTC, was “felt to be promising but with insufficient evidence to recommend for screening,” Dr. Brooks notes. Subsequent research and steady improvements in technology “show pretty clearly that both are at least equivalent to existing tests.”
sDNA testing operates on a novel principle: Rather than occult blood, the test detects molecular markers associated with neoplasia. Specificity is high (93%-97%), but since the test can target only a limited panel of abnormalities (21 separate point mutations in the original version), it is bound to miss some lesions. Test sensitivity has actually been comparable to that of gFOBT. The recommended frequency of sDNA testing has not yet been established.
The Guideline emphasizes the importance of proper sample collection and follow-up for stool testing. In a recent national survey, nearly one third of PCPs reported collecting stool samples during digital rectal examination (which reduces gFOBT sensitivity to 9%), and a similar proportion said they followed up positive gFOBT findings by repeating the test or referring for sigmoidoscopy, rather than suggesting colonoscopy, which is recommended.
Structural exams
Colonoscopy “tends to be the standard against which other tests are measured because it is the only one that allows direct visualization of the entire colon wall,” Dr. Brooks says. “But it's a misconception to view it as the ‘gold standard'; that would imply a level of perfection colonoscopy doesn't achieve.” It fails to detect 5% of cancers and 6%-12% of large adenomas.
Clinicians performing the procedure vary considerably in training and skill. When referring patients, PCPs should look for consultants who report factors that can influence accuracy, such as thoroughness (mention of the cecum indicates the entire colon was viewed) and time, Dr. Brooks says. One large well-designed study found significantly more missed abnormalities when withdrawal time was under six minutes.
The polyp detection rate can be another quality indicator. In most screening populations, an incidence of 20%-25% is to be expected.
In experienced hands, the sensitivity of CTC (“virtual colonoscopy”) for cancer and large adenomas appears comparable to that of optical colonoscopy, according to the Guideline.
One drawback of CTC is the need for optical colonoscopy to biopsy or remove lesions that are discovered—a second procedure that will again require full bowel preparation. CTC should be repeated every five years, vs. 10 years for colonoscopy, the Guideline says.
As a practical matter, the availability of CTC is still “extraordinarily limited,” Dr. Brooks notes. As it expands beyond research centers and urban settings, “you want to be certain the individuals performing and reading tests are well trained and use up-to-date equipment.”
Sigmoidoscopy and double-contrast barium enema remain recommended options, although usage of these procedures, both of which are less sensitive than colonoscopy, has declined substantially in recent years.
Non-protruding lesions
A study published in The Journal of the American Medical Association (coincidentally on the day the Guideline was released) suggested that non-polypoid lesions, which are difficult to detect by colonoscopy or CTC, are more common and more likely to be malignant than previously thought. Among patients who were screened in the study, the incidence of malignant, non-polypoid lesions was 0.32% (JAMA. 2008;299:1027-1035).
“If the paper had appeared earlier, it would have been mentioned in the Guideline,” Dr. Brooks declares. “It raises the question of whether colonoscopy needs to be performed differently, at least in certain segments of the population.” This could mean using a special dye, as was done in the study, or different light. At the very least, “the study underlines the importance of awareness of these lesions. I suspect we'll see a much higher rate of detection in coming years,” he says.
The paper also highlights the fact that colonoscopy can miss abnormalities; “if new symptoms develop a year or two after colonoscopy, they should never be ignored,” he says.
Screening and Surveillance 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 was published in CA: A Cancer Journal for Clinicians (2008;58:130-
160).
Mr. Sherman is a medical writer in New York City.