Screening for Colorectal Cancer
Patrick S. Yachimski
Lawrence S. Friedman
Colorectal cancer is the third leading cause of cancer death among men and women in the United States. Americans face a lifetime risk of colorectal cancer of approximately 1 in 20. The majority of colorectal cancers that present with symptoms or signs have already spread beyond the intestinal wall, resulting in a poor 5-year survival rate; however, detection and removal of early, localized cancers may result in a 5-year survival rate greater than 80%. The long asymptomatic period with a known precursor lesion (adenomatous polyp), availability of tests that can identify localized disease, and improved outcomes with therapy initiated at an early stage have long suggested that screening for colorectal cancer would be effective. In fact, evidence from randomized controlled trials has confirmed that screening can reduce the incidence and mortality of colorectal cancer. Colonoscopy has emerged as the preferred approach to colon cancer screening in the United States, and screening for colorectal cancer has become an important element of cancer screening in primary care practice.
Epidemiology and Risk Factors
The incidence of colorectal cancer is greater in economically developed societies than in developing countries. A high-fat, low-fiber diet prevalent in Western societies has been implicated in the etiology of these cancers. Whether intake of dietary fiber, fruits, and vegetables decreases the risk or red or processed meats increase the risk of colorectal cancer has been harder to prove, but is likely. Obesity, tobacco smoking, heavy alcohol use, and type 2 diabetes mellitus are additional risk factors for colorectal cancer.
Advancing age is an important risk factor, with the incidence of colorectal cancer rising from about 75 in 100,000 in the sixth decade to 300 in 100,000 in the eighth decade. The median age at diagnosis of colorectal cancer is 70 years.
Family history is also a risk factor. Having a first-degree relative with colorectal cancer raises personal risk about twofold, with a greater risk when the cancer has been diagnosed before age 60 years (and especially before age 50 years) and when more than one first-degree relative has been affected. Evidence also suggests that having a first-degree relative with a large adenomatous polyp also increases the risk of colorectal cancer.
Adenomatous polyps of the colon are precursor lesions to colon cancer. The frequency of adenomatous polyps increases with age, occurring in 30% of 50-year-olds, 40% of 60-year-olds, and 50% of 70-year-olds. Adenomatous polyps containing low-grade dysplasia may progress to high-grade dysplasia and, ultimately, carcinoma via a well-established sequence of acquired somatic DNA mutations, including mutations in the K-ras proto-oncogene, as well as the APC and TP53 tumor suppressor genes.
Although most colon cancers arise from sporadic adenomatous polyps, persons with specific inherited polyposis syndromes are at particularly increased risk. Individuals with familial adenomatous polyposis (FAP) possess an autosomal dominant mutation in the APC tumor suppressor gene that confers a risk for colorectal cancer of about 50% by age 40 years. Individuals with Lynch syndrome (also known as hereditary nonpolyposis colorectal cancer) possess mutations in DNA mismatch repair genes and are at risk for both colorectal and noncolorectal malignancies, including endometrial, ovarian, gastric, and pancreatic cancer. Genetic testing is available for both FAP and some forms of Lynch syndrome.
Hyperplastic colonic polyps have historically been considered to be benign colonic polyps without a risk of progression to malignancy. A variant of hyperplastic polyps, however, has undergone histopathologic reclassification and is identified as serrated adenoma, which appears to have a risk of malignant progression similar to that of tubular adenomas. Individuals with multiple large hyperplastic polyps in the right colon, designated as hyperplastic polyposis syndrome, may be at increased risk of proximal colonic neoplasia.
Ulcerative colitis, particularly pancolitis of more than 10 years’ duration, increases a patient’s risk of colorectal cancer 5- to 10-fold, although isolated ulcerative proctitis and proctosigmoiditis are thought to confer no additional risk. Patients with extensive Crohn’s colitis also appear to have a significant increase in risk.
Patients with resected colorectal cancers have a threefold increase in risk of developing metachronous cancer in other locations of the colon. Recurrences of index cancers are usually extraluminal and not amenable to detection by endoscopic surveillance.
The prevalence of synchronous colonic neoplasia is high and important to consider in the evaluation of a patient with a positive screening test result. Synchronous adenomas occur in 40% to 50% of patients with an index polyp, and 3% to 5% with a carcinoma harbor a second carcinoma.
In some observational studies, the use of aspirin or nonsteroidal anti-inflammatory drugs has appeared to be associated with a lower risk of colorectal cancer. In cohort studies, regular aspirin use, especially at doses higher than generally used for cardiovascular prevention and for prolonged durations, has been associated with a reduction in the incidence of colorectal cancer as high as 60%. The gastrointestinal toxicity of these agents at higher doses reduces their utility in colorectal cancer chemoprevention.
Natural History
Adenomas are low-grade dysplastic lesions and on occasion may harbor high-grade dysplasia or malignant foci. The risk of malignancy increases with the size of the adenoma, ranging from 0.5% of colonoscopically removed polyps 0.5 to 0.9 cm in diameter to 25% or more of polyps greater than 3 cm in diameter. Adenomas with a villous morphology on histopathologic assessment have increased malignant potential. The time frame for the evolution of a nonmalignant polyp into an invasive cancer is estimated to average 10 years, as reflected in the recommended colonoscopic surveillance interval of 10 years following an index colonoscopy that demonstrates no polyps.
Symptoms occur late in the course of colorectal cancer growth and may include gastrointestinal blood loss, altered bowel habits, abdominal pain, and colonic obstruction. When cancers are found after symptomatic presentation, 60% have already disseminated to regional nodes or distant organs. Five-year survival rates vary dramatically with the stage of disease at the time of diagnosis. Current cancer staging is classified according to the tumor node metastasis criteria, as endorsed by the American Joint Committee on Cancer. More than one third of patients are diagnosed with localized cancer, confined to the bowel wall, and the 5-year survival rate for early stage disease is approximately 90%. A similar proportion of patients are diagnosed with regional disease and lymph node involvement and have a 5-year survival rate of nearly 70%. The 5-year survival rate for metastatic disease is appreciably lower, at approximately 10%.
SCREENING TESTS
Intermittent occult bleeding occurs with some asymptomatic colorectal cancers and large polyps. Sampling of stool for occult blood can be performed to identify bleeding. Stool sampling is best performed serially (e.g., two samples from a stool on each of 3 days) because of the intermittent nature of the bleeding. A six-test guaiac stool sequence has a sensitivity of 15% to 50% for colorectal cancer. Rehydration of the sample before application of developer can increase the sensitivity, at the cost of a reduction in specificity. False-positive test results may occur with nonmalignant lesions, use of aspirin (at doses >325 mg daily) and nonsteroidal anti-inflammatory drugs, or dietary ingestion of red meat or foods with high peroxidase activity. Results of studies of the ability of iron supplements to cause false-positive guaiac results have been conflicting. High doses of vitamin C increase the false-negative rate. Guaiac testing of single stool samples obtained on office-based digital rectal examinations in truly asymptomatic patients, although widely performed, is not a recommended screening strategy.
Although a single set of guaiac cards has limited sensitivity for colorectal cancer, a program of annual testing may have a sensitivity as high as 80% to 90%. Newer-generation guaiacbased tests, such as Hemoccult Sensa, have superior sensitivity when compared with older guaiac-based tests. Although guaiacbased assays remain an acceptable colorectal cancer screening test, fecal immunohistochemical testing (FIT) has emerged as the recommended method for fecal occult blood test (FOBT). FIT detects human hemoglobin via an antibody-based assay. In randomized comparisons of FIT and guaiac-based testing (using the older-generation Hemoccult II) for population-based screening of individuals at average risk for colorectal cancer, and in which all subjects with a positive FOBT underwent subsequent colonoscopy, FIT had a higher detection rate for advanced adenomas and cancer. In addition, FIT requires fewer stool samples and less stringent dietary and medication restrictions than guaiacbased FOBT; as a result, adherence rates to FOBT screening may be higher with FIT than with guaiac-based testing.
Individuals with a positive FOBT result require a thorough evaluation for colorectal neoplasia, with current guidelines endorsing colonoscopy as the recommended next step in the evaluation of a positive FOBT.
Outcomes
The best outcome measure of screening efficacy is reduced colorectal cancer mortality, as judged from prospective, randomized controlled clinical trials with a long period of follow-up. One such study, the Minnesota Colon Cancer Study, demonstrated a 33% reduction in cumulative colorectal cancer mortality (from 8.3 in 1,000 to 5.9 in 1,000 over 13 years) as a result of annual FOBT. This trial used a six-sample annual testing protocol, followed by colonoscopy for persons with positive test results. Thirty-eight percent of screenees underwent colonoscopy during the study. Compared with the control group, screened patients demonstrated detection of cancers at an earlier stage than nonscreened patients. Rehydration of the guaiac card increased sensitivity, but it also increased the number of colonoscopies needed, the cost of screening, and the false-positive rate. Despite the large relative benefit in terms of mortality, the absolute risk that any one man or woman will die of colorectal cancer is quite low. As a result, data from this trial indicate that about 300 men would need to be screened annually for 13 years to prevent one colorectal cancer death.
Two other trials, conducted in Denmark and the United Kingdom, respectively, used guaiac cards without hydration and conducted screening biennially rather than annually. Only 4% of persons screened in these studies underwent colonoscopy. Even under these circumstances, the trials demonstrated 18% and 15% reductions in colorectal cancer mortality, respectively. A Cochrane meta-analysis of four randomized trials showed an overall reduction in the risk of colorectal cancer mortality of 16% as a result of guaiac-based FOBT screening.
Flexible sigmoidoscopy can potentially detect about 50% of cancers—specifically, cancers in the left colon that are within reach of the sigmoidoscope. The risk of a bowel perforation—the major potential complication of sigmoidoscopy—is low, less than 1 in 10,000 flexible sigmoidoscopic examinations (see Appendix). Adenomatous polyps are found in up to 20% of persons screened. Polyps detected at sigmoidoscopy may be removed via polypectomy; the patient then usually undergoes colonoscopy for examination of the proximal colon.
Outcomes
Sigmoidoscopy can clearly find localized cancers and polyps for removal, but the key issue is the effect of the procedure on colorectal cancer mortality rates when used for periodic screening. In a randomized trial in Italy, in which individuals between ages 55 and 64 were randomized to flexible sigmoidoscopy or no screening, subjects with a negative screening flexible sigmoidoscopy result had a 59% decrease in the incidence of colon cancer after 10 years of follow-up compared with subjects who had not undergone screening. The study documented 65 colorectal cancer-related deaths in the flexible sigmoidoscopy group compared with 83 colorectal cancer-related deaths in the control group. This translated into a 38% mortality reduction (RR 0.62; 95% CI 0.40 to 0.96) in the flexible sigmoidoscopy arm in the per-protocol analysis (with a RR 0.78; 95% CI 0.56 to 1.08 in the intention-to-treat analysis).
TABLE 56-1 Colorectal Cancer Screening: Multisociety Guidelines for Testing Options Beginning at Age 50 | |||||||||||||||||||||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||
In a similar study in Norway, patients were randomized to once-only screening flexible sigmoidoscopy (with or without FOBT) or no screening. Twenty-one percent of patients in the screening group underwent full colonoscopy because of positive findings at sigmoidoscopy or FOBT. Five percent of patients in the screening arm had a prevalent high-risk adenoma or cancer at the time of screening. Over a median of 7 years of follow-up, there was no difference in the incidence of colorectal cancer in the screening and no screening groups. Colorectal cancer mortality was reduced by 27% in the screening group (HR 0.73, 95% CI 0.47 to 1.13). Whether the magnitude of reduction in mortality will achieve statistical significance over an extended duration of follow-up is uncertain.
Flexible sigmoidoscopy clearly offers a protective effect specific for cancers within reach of the sigmoidoscope. Furthermore, detection of adenomatous polyps by flexible sigmoidoscopy is thought to be predictive of the presence of adenomatous polyps in the proximal colon, and flexible sigmoidoscopy may identify patients who will benefit most from a full colonoscopy.
The optimal frequency of surveillance sigmoidoscopy following an initial negative screening examination is unclear. On one hand, a randomized study from the United Kingdom found a 23% reduction in colorectal cancer incidence (HR 0.77, 95% CI 0.70 to 0.84) and 31% reduction in colorectal cancer mortality (HR 0.69, 95% CI 0.59 to 0.82) among subjects randomized to once-only flexible sigmoidoscopy (subjects found to have polyps during flexible sigmoidoscopy then underwent colonoscopy) compared with those randomized to no screening. On the other hand, a repeat examination 3 years after a negative sigmoidoscopy will reveal large adenomas or cancer in about 1% of patients. Practice guidelines offer a range of recommendations for the frequency of flexible sigmoidoscopy screening, with or without FOBT (Table 56-1).
Related posts:
Screening for HIV Infection
Screening for Hyperlipidemia and Associated Coronary Heart Disease Risk Factors
Evaluation of Arterial Insufficiency of the Lower Extremities
Approach to the Patient with Chronic Congestive Heart Failure
Evaluation and Prevention of Occupational and Environmental Lung Disease
Management of Chronic Obstructive Pulmonary Disease
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
