Screening for Colorectal Cancer
David F. Ransohoff, M.D., and Robert S. Sandler, M.D., M.P.H.
This Journal feature begins with a case vignette highlighting a common clinical problem. Evidence supporting various strategies is then presented, followed by a review of formal guidelines, when they exist. The article ends with the authors' clinical recommendations.
A healthy 50-year-old woman at average risk for colorectal cancer (i.e., age is her only risk factor) is scheduled to undergo a periodic examination. Which screening test for colorectal cancer should be recommended?
The Clinical Problem
Even though colorectal cancer is the second most common cause of cancer-related death in the United States and screening can reduce the rate of death by detecting early cancer or premalignant polyps, the rates of screening are low. Fewer than 30 percent of eligible persons have had a screening test for colorectal cancer, whereas 71 percent of women over the age of 40 years have had a mammogram and 80 percent of such women have had a Papanicolaou smear in the preceding two years. The low rates have been attributed to resistance by physicians, patients, and the health care system. A physician's recommendation that a patient undergo screening strongly influences whether screening is done, and recognition of the need for colorectal-cancer screening constitutes an important clinical and public health challenge.
The choice of which screening test to use is controversial. Colonoscopy has recently received attention because it is the most accurate test, and it is clearly indicated for patients with an increased risk of colorectal cancer because of a personal or family history of colorectal cancer, high-risk adenoma, or long-standing ulcerative colitis. The choice of test for patients at average risk for colorectal cancer requires consideration of many issues, including the diagnostic value of each test, logistic requirements, cost, and the patient's preference.
Strategies and Evidence
Fecal Occult-Blood Testing
A test for fecal occult blood should ideally include an analysis of three serially obtained samples of stool, because this approach has been assessed in randomized, controlled clinical trials. Testing of a single fecal sample for occult blood is commonly done in the United States after a rectal examination, but this approach has not been studied in randomized clinical trials and may be associated with higher rates of false positive and false negative results than is the three-sample test.
The three cards for fecal occult-blood screening can be completed by the patient at home and then mailed to the physician or laboratory. Office procedures are needed to track and process specimens to ensure high-quality testing. False positive results may be caused by nonspecific bleeding (for example, as a result of angiodysplasia), diet, or the rehydration of samples. Rehydration is intended to increase the sensitivity of the test and does so slightly; however, rehydration also increases the false positive rate from about 2 percent to nearly 10 percent in older persons, thus substantially increasing the number of workups for colorectal cancer that are ordered. Physicians may recommend that patients avoid consuming red meat, horseradish, and aspirin before a test. A workup of the whole colon is recommended, usually with colonoscopy, if any of the three test cards for fecal occult blood is positive, because the probability of finding a colorectal cancer or large adenoma ranges from 17 to 46 percent among those with a positive test. Although fecal occult-blood testing itself is relatively inexpensive, the costs in patients with a positive result may be substantial. The most widely used fecal occult-blood test is Hemoccult II (Beckman Coulter, Palo Alto, Calif.), but others have been assessed.
A fecal occult-blood test is the most commonly performed screening test for colorectal cancer in the United States and is the only one demonstrated on the basis of high-quality evidence from randomized clinical trials to reduce the risk of death from colorectal cancer. The reduction in risk ranges from 15 to 33 percent and is similar to the reduction reported in randomized clinical trials of screening mammography, but it depends in part on whether the test is done every year or every other year and on whether the samples are rehydrated (Table 1).
Table 1. Characteristics of Screening Tests for Colorectal Cancer.
Sigmoidoscopy, performed without sedation in a patient who has received a laxative or an enema, can be used to examine about half the colon. The quality and safety of sigmoidoscopy performed as part of routine clinical practice are unknown, although the procedure, when it is done by trained, experienced personnel, including nonphysicians, is safe. Primary care physicians who want to learn how to use a sigmoidoscope face logistic problems of arranging hands-on training, purchasing the equipment, and training personnel to assist them. Low rates of reimbursement for the procedure may dampen the enthusiasm of primary care physicians for solving these logistic problems.
Sigmoidoscopy is highly sensitive and specific for lesions within reach of the instrument but misses the roughly 50 percent of neoplasms that are proximal or right-sided. The capacity of the procedure to detect proximal lesions depends on the threshold used to initiate a colonoscopic workup once a polyp is identified in the distal colon. If the presence of any adenoma is used as a threshold, the sensitivity for the detection of advanced neoplasms (for example, those with high-grade dysplasia or villous histologic features or those larger than 1 cm) anywhere in the colon is about 70 percent. The addition of fecal occult-blood testing to the workup increases the sensitivity by only a few percentage points.
Unlike fecal occult-blood screening, sigmoidoscopic screening has not been assessed in any randomized clinical trials, although two trials are under way in the United States and the United Kingdom. Good-quality evidence from a case–control study suggests that the risk of death from colorectal cancer is reduced by about 60 percent by the finding of a lesion within reach of the instrument, translating to a reduction of 30 percent with respect to the whole colon. Although sigmoidoscopy is not as thorough as colonoscopy, it is safer and easier to perform, is considered to be both efficacious and cost effective, and is recommended by major professional organizations.
Colonoscopy can be used to examine the entire colon and is performed by a trained endoscopist. Patients scheduled to undergo colonoscopy follow a liquid diet and take laxatives beforehand and are lightly sedated for the procedure. As compared with sigmoidoscopy, colonoscopy has a higher rate of perforation (a rate of about 2 per 1000 procedures, as compared with a rate of about 1 per 10,000 for sigmoidoscopic procedures), a higher rate of complications associated with the use of sedatives, and a higher cost, but it examines more of the colon and can be used to remove polyps and early cancers. Furthermore, negative findings on colonoscopic examination can obviate the need for further screening for five years or longer, a feature that may be attractive to patients who are considering their screening options.
Since no randomized clinical trials have been done to assess the effect of colonoscopic screening on the rate of death from colorectal cancer, recommendations are necessarily based on weaker evidence, such as that obtained from case–control studies and mathematical modeling. Even when data from the two ongoing trials of sigmoidoscopy become available in the next decade, some observers think that the results of trials that assess the left side of the colon should not be extended to apply to the entire colon. At issue is whether the behavior of right-sided neoplasms differs from that of left-sided neoplasms in a way that would make endoscopic screening less effective; this might be the case, for example, if right-sided neoplasms grew much faster than left-sided ones, or if they tended to be flatter and more difficult to detect endoscopically. In spite of the lack of evidence from randomized clinical trials, however, the use of colonoscopic screening is supported by common-sense reasons and has received substantial attention from the media as a result of recent studies showing that this strategy is more sensitive than sigmoidoscopy, even when sigmoidoscopy is combined with fecal occult-blood testing. Indeed, Medicare is now providing reimbursement for colonoscopic screening performed at 10-year intervals in average-risk patients. However, the decision whether to choose a test with greater sensitivity must be weighed against other considerations, as discussed below.
Barium enema has been recommended in the past as a possible screening test for colorectal cancer, but its surprisingly low sensitivity for the detection of large polyps (only 48 percent for lesions that are 1 cm or larger) has decreased interest in this approach. Although digital rectal examination has not been shown to be effective and is generally not recommended for screening, digital rectal examination combined with a one-sample fecal occult-blood test performed in the office is probably the most common approach to screening in the United States.
Virtual colonoscopy involves the examination of a computerized image of the colon constructed from data obtained from an abdominal computed tomographic examination. Patients must take laxatives before the procedure. Although the sensitivity is 91 percent for the detection of lesions that are larger than 1 cm, the false positive rate of 17 percent25 is prohibitively high for a screening test if additional preparation with laxatives and colonoscopic workup are then needed. If future developments increase the specificity or eliminate the need for a laxative regimen, then virtual colonoscopy might become an important screening option.
Comparison of Different Tests
The choice among different tests and strategies is made more difficult by the paucity of evidence from randomized clinical trials. Interest in colonoscopy has recently grown because of studies demonstrating that sigmoidoscopy fails to identify about half of all proximal neoplasms. The number of proximal lesions that are found would increase if a finding of distal polyps prompted a colonoscopic workup, depending on the type of polyp that is used as a threshold.
Although the failure of sigmoidoscopy to detect proximal lesions received much attention editorially26 and in the lay press, that finding was not surprising to clinicians or to professional organizations, because sigmoidoscopy was never expected to be a particularly sensitive method of detecting proximal neoplasms. The studies that directly compared sigmoidoscopy and colonoscopy did not take into account the safety, cost, and availability of the tests.
In past as well as in more recent analyses, a variety of tests and strategies have been shown to be cost-effective approaches to colorectal-cancer screening (that is, to cost less than $30,000 per year of life saved), including colonoscopy, fecal occult-blood testing, and sigmoidoscopy.20,29 Because cost-effectiveness analyses do not consistently show that colonoscopic screening is the most effective strategy, let alone the most cost-effective one, colonoscopy cannot be considered the preferred strategy. The relation of colonoscopy to other tests could be clarified by obtaining and assessing additional empirical data on its safety, cost, and availability and by performing analyses that specifically compare the marginal cost effectiveness of various strategies.
Areas of Uncertainty
Unlike the situation 10 years ago, there is now no uncertainty about whether screening can reduce the rates of death from colorectal cancer. What remains uncertain are issues regarding the relative efficacy, cost, and cost effectiveness of the various tests, and specifically whether routine colonoscopic screening should be the test of choice. It also remains unclear how screening can be made routine in the United States. Indeed, the fact that there are so many choices may itself constitute a barrier that causes people simply to make no choice at all.
When sigmoidoscopy is used as a screening test, the threshold that should trigger a colonoscopic workup to examine the proximal colon also remains controversial. The choice is important because 25 percent of patients who undergo sigmoidoscopic screening will have at least one polyp, whereas slightly less than half of these patients will have an adenomatous polyp (the others will have a hyperplastic polyp) and only one fifth of these patients will have a high-risk adenoma. Depending on which of these findings is used as the threshold for performing colonoscopy, the yield and costs associated with a screening approach based on sigmoidoscopy will vary considerably. Because, in the United States, biopsies are not routinely performed for polyps identified by sigmoidoscopic screening, the use of a finding of any polyp as a threshold means that about 25 percent of persons who undergo sigmoidoscopy will subsequently undergo colonoscopy. The inefficiencies of such a strategy could be reduced by an improved understanding of the characteristics that are associated with a significant risk of abnormal proximal findings, such as age, sex, and family history.
Although recommendations for colorectal-cancer screening are still in flux and vary in terms of detail, they now uniformly support the use of screening (Table 2), whereas 10 years ago they did not. The guidelines for persons at average risk recommend annual fecal occult-blood testing alone, sigmoidoscopic screening alone every five years, or a combination of the two tests. The performance of a fecal occult-blood test every other year is not endorsed, although this approach has been shown to be efficacious.
Table 2. Recommended Tests and Screening Intervals in Persons at Average Risk for Colorectal Cancer.
Recommendations regarding colonoscopic screening vary substantially. In 1996, the Preventive Services Task Force concluded that there was insufficient evidence for or against the use of colonoscopic screening, but it is currently revising its recommendations. The Gastrointestinal Consortium recommends that colonoscopic screening be offered, whereas the American College of Gastroenterology considers this approach the "preferred strategy." The recommendations of different groups also vary in other important ways, such as whether a full colonoscopy should be performed after the discovery of a polyp on sigmoidoscopic screening. The guidelines of the Gastrointestinal Consortium state that patients with a large distal polyp (more than 1 cm) should have a full colonic workup, but that patients with a small distal polyp should decide, with their physicians, whether a full colonic workup is necessary. In contrast, the American Cancer Society indicates that the discovery of any distal polyp should lead to colonoscopy.
Conclusions and Recommendations
The variation in recommendations should not obscure the larger message that screening can reduce the rate of death from colorectal cancer. We suggest that the variation be interpreted in a positive way as giving clinicians several choices with respect to colorectal-cancer screening. The variation is due less to a lack of evidence than to different perspectives and goals. The more aggressive recommendations of a specialist physician or organization may reflect the fear that even one case of cancer could be missed. A broader, population-based perspective may reflect trade-offs with respect to availability, cost, safety, and quality. Goals may vary or fail to provide clear guidance because there is no consensus, or even discussion, about the level of absolute risk that is high enough to warrant screening and the magnitude of risk reduction that is reasonable. If the goal were to minimize the likelihood of missing any colorectal cancer, then colonoscopy would be favored; if the goal were to reduce the rate of death from colorectal cancer to a large extent but at a lower cost and a lower risk than those associated with colonoscopy, then sigmoidoscopy or fecal occult-blood testing would be appropriate.
At present physicians should not be dogmatic about which test to use but, rather, should offer patients like the woman in the clinical vignette a choice among fecal occult-blood testing, sigmoidoscopy, and colonoscopy, and they should discuss the features of each test with their patients, who may have their own perspectives and preferences. When asked directly, physicians should recommend the test or tests that are most acceptable to the patient, given concern about cost, safety, discomfort, and fear of cancer. Patients should be told about lifestyle measures that may reduce the risk of colorectal cancer, including avoiding obesity, exercising regularly, not smoking, and limiting their intake of alcohol and red meat. Although no test can eliminate the risk of death from colorectal cancer, screening tests can reduce the risk by a clinically important amount, and the use of any one of these tests is preferable to none at all.
From the Departments of Medicine and Epidemiology, University of North Carolina at Chapel Hill, Chapel Hill.
- Greenlee RT, Hill-Harmon MB, Murray T, Thun M. Cancer statistics, 2001. CA Cancer J Clin 2001;51:15-36. [Erratum, CA Cancer J Clin 2001;51:144.]
- Screening for colorectal cancer -- United States, 1997. MMWR Morb Mortal Wkly Rep 1999;48:116-121.
- Vernon SW. Participation in colorectal cancer screening: a review. J Natl Cancer Inst 1997;89:1406-1422.
- Mandel JS, Bond JH, Church TR, et al. Reducing mortality from colorectal cancer by screening for fecal occult blood. N Engl J Med 1993;328:1365-1371. [Erratum, N Engl J Med 1993;329:672.]
- Kronborg O, Fenger C, Olsen J, Jorgensen OD, Sondergaard O. Randomised study of screening for colorectal cancer with faecal-occult-blood test. Lancet 1996;348:1467-1471.
- Hardcastle JD, Chamberlain JO, Robinson MHE, et al. Randomised controlled trial of faecal-occult-blood screening for colorectal cancer. Lancet 1996;348:1472-1477.
- Ransohoff DF, Lang CA. Screening for colorectal cancer with the fecal occult blood test: a background paper. Ann Intern Med 1997;126:811-822.
- Lang CA, Ransohoff DF. What can we conclude from the randomized controlled trials of fecal occult blood test screening? Eur J Gastroenterol Hepatol 1998;10:199-204.
- Pignone MP, Rich M, Teutsch SM, Berg AO, Lohr KN. Screening for colorectal cancer in adults. Systematic evidence review no. 7. Rockville, Md.: Agency for Healthcare Research and Quality (in press). (AHRQ publication no. 02-S003.)
- Selby JV, Friedman GD, Quesenberry CP Jr, Weiss NS. A case-control study of screening sigmoidoscopy and mortality from colorectal cancer. N Engl J Med 1992;326:653-657.
- Levin TR, Conell C, Shapiro JA, Chazan SG, Nadel M, Selby JV. Complications of screening sigmoidoscopy. Gastroenterology 2001;120:Suppl:A-65.abstract
- Ashley OS, Nadel M, Ransohoff DF. Achieving quality in flexible sigmoidoscopy screening for colorectal cancer. Am J Med 2001;111:643-653.
- Lieberman DA, Weiss DG, Bond JH, Ahnen DJ, Garewal H, Chejfec G. Use of colonoscopy to screen asymptomatic adults for colorectal cancer. N Engl J Med 2000;343:162-168. [Erratum, N Engl J Med 2000;343:1204.]
- Imperiale TF, Wagner DR, Lin CY, Larkin GN, Rogge JD, Ransohoff DF. Risk of advanced proximal neoplasms in asymptomatic adults according to the distal colorectal findings. N Engl J Med 2000;343:169-174.
- Lieberman DA, Weiss DG. One-time screening for colorectal cancer with combined fecal occult-blood testing and examination of the distal colon. N Engl J Med 2001;345:555-560.
- Kramer BS, Gohagan J, Prorok PC, Smart C. A National Cancer Institute sponsored screening trial for prostatic, lung, colorectal, and ovarian cancers. Cancer 1993;71:Suppl:589-593.
- Atkin WS, Edwards R, Wardle J, Northover JM, Cuzick J. UK Flexible Sigmoidoscopy Screening trial: compliance, yield and adverse effects. Gastroenterology 2000;118:Suppl:A187-A187.abstract
- Rex DK, Johnson DA, Lieberman DA, Burt RW, Sonnenberg A. Colorectal cancer prevention 2000: screening recommendations of the American College of Gastroenterology. Am J Gastroenterol 2000;95:868-877.
- Screening for colorectal cancer. In: Preventive Services Task Force. Guide to clinical preventive services. 2nd ed. Baltimore: Williams & Wilkins, 1996:89-103.
- Winawer SJ, Fletcher RH, Miller L, et al. Colorectal cancer screening: clinical guidelines and rationale. Gastroenterology 1997;112:594-642. [Erratum, Gastroenterology 1997;112:1060, 1998;114:625.]
- Smith RA, von Eschenbach AC, Wender R, et al. American Cancer Society guidelines for the early detection of cancer: update of early detection guidelines for prostate, colorectal, and endometrial cancers. Also: update 2001 -- testing for early lung cancer detection. CA Cancer J Clin 2001;51:38-75. [Erratum, CA Cancer J Clin 2001;51:150.]
- Saitoh Y, Waxman I, West AB, et al. Prevalence and distinctive biologic features of flat colorectal adenomas in a North American population. Gastroenterology 2001;120:1657-1665.
- Winawer SJ, Stewart ET, Zauber AG, et al. A comparison of colonoscopy and double-contrast barium enema for surveillance after polypectomy. N Engl J Med 2000;342:1766-1772.
- Herrinton LJ, Selby JV, Friedman GD, Quesenberry CP, Weiss NS. Case-control study of digital-rectal screening in relation to mortality from cancer of the distal rectum. Am J Epidemiol 1995;142:961-964.
- Fenlon HM, Nunes DP, Schroy PC III, Barish MA, Clarke PD, Ferrucci JT. A comparison of virtual and conventional colonoscopy for the detection of colorectal polyps. N Engl J Med 1999;341:1496-1503. [Erratum, N Engl J Med 2000;342:524.]
- Podolsky DK. Going the distance: the case for true colorectal-cancer screening. N Engl J Med 2000;343:207-208.
- Frazier AL, Colditz GA, Fuchs CS, Kuntz KM. Cost-effectiveness of screening for colorectal cancer in the general population. JAMA 2000;284:1954-1961.
- Sonnenberg A, Delco F, Inadomi JM. Cost-effectiveness of colonoscopy in screening for colorectal cancer. Ann Intern Med 2000;133:573-584.
- Wagner JL, Behney CJ, Tunis SR, Ching A. Cost-effectiveness of colorectal cancer screening in average-risk adults. Washington, D.C.: Office of Technology Assessment, 1995. (Report no. BP-H-146.)
- Redelmeier DA, Shafir E. Medical decision making in situations that offer multiple alternatives. JAMA 1995;273:302-305.
- Imperiale TF, Wagner DR, Lin CY, Larkin GN, Rogge JD, Ransohoff DF. A new clinical index to stratify risk for advanced proximal neoplasia in asymptomatic adults. Gastroenterology 2001;120:Suppl:A-15.abstract
- Mandel JS, Church TR, Ederer F, Bond JH. Colorectal cancer mortality: effectiveness of biennial screening for fecal occult blood. J Natl Cancer Inst 1999;91:434-437.
- Byers T, Levin B, Rothenberger D, Dodd GD, Smith RA. American Cancer Society guidelines for screening and surveillance for early detection of colorectal polyps and cancer: update 1997. CA Cancer J Clin 1997;47:154-160.
- Woolf SH. The best screening test for colorectal cancer -- a personal choice. N Engl J Med 2000;343:1641-1643.
- Lieberman T. Matching colon tests to patients, not costs. Los Angeles Times. March 19, 2001:S1, S7.
- Willett WC. Diet and cancer: one view at the start of the millennium. Cancer Epidemiol Biomarkers Prev 2001;10:3-8.