Family Practice Vol. 18, No. 4, 457-460
© Oxford University Press 2001
Selections from Current Literature |
Colorectal cancer screening
Department of Family Medicine, Health Sciences Center, L-4, SUNY at Stony Brook School of Medicine, Stony Brook, NY 11794, USA.
Pesce A. Colorectal cancer screening. Family Practice 2001; 18: 457–460.
Introduction
Colon cancer is the second leading cause of cancer deaths in the USA. Annually, ~129 000 new cases of colorectal cancer are diagnosed, with 59 000 associated deaths.1 It is widely believed that most colorectal cancers arise from adenomatous polyps. The probability of an adenomatous polyp becoming a cancer depends on the gross appearance of the lesion, its histological features and its size.2 Sessile (flat-based) polyps are more likely to develop into cancer than are pedunculated (stalked) lesions. Histologically, villous adenomas are more likely to develop into carcinoma than are tubular lesions. The main approach to colorectal cancer prevention is the detection and removal of pre-cancerous lesions (primary prevention). In addition, cancer found in its early stages may be more amenable to treatment (secondary prevention).
There have been no randomized controlled clinical trials suggesting that screening sigmoidoscopies decrease mortality, although other studies suggest this association. In one retrospective study, only 9% of controls who died from cancer of the distal 20 cm of the colon had ever had rigid sigmoidoscopy compared with 24% in the cancer-free control group.3 Neither the optimal frequency of sigmoidoscopic screening nor the rate of complications have been delineated. It is believed that the rate of perforation from rigid sigmoidoscopy is 1 in 1000–10 000.4 The risk from flexible sigmoidoscopy is believed to be less than or equal to that of rigid sigmoidoscopy.
Compared with sigmoidoscopy, colonoscopy visualizes the entire colon and is, therefore, more complete. A further advantage is that polyps can be removed during the procedure for histological study. Colonoscopy, however, carries increased risk and patient burden. Such burdens include increased cost, a full day of missed work, requiring assistance driving the patient to and from the procedure, bowel preparation (usually involves drinking a laxative solution on the eve of the procedure; aspirin avoidance for at least 1 week previously) and sedation (with its inherent risks). The perforation risk is ~1 in 500–3000 examinations. Serious bleeding occurs in ~1 in 1000.5 There are no randomized controlled clinical trials in the literature documenting a decrease in colorectal cancer mortality from screening colonoscopy.
There is good evidence that faecal occult blood testing (FOBT) reduces mortality from colorectal cancer in asymptomatic average-risk adults. In 1993, a randomized controlled clinical trial involving 46 000 volunteers in Minnesota demonstrated that annual FOBT decreased colorectal cancer mortality by 33% after 13 years.5 Other randomized controlled trials and observational data have corroborated these results.4 Unfortunately, non-neoplastic conditions such as gastritis, diverticulosis and haemorrhoids can produce false-positive results for neoplasia. Additionally, foods containing peroxidases, and gastric irritants such as salicylates can cause false-positive results. This is an important consequence of testing since these false-positive results can lead to anxiety, increased cost, discomfort and complications from the follow-up diagnostic tests.
Although these screening tests can be uncomfortable and unpleasant, acceptance rates for each of the available tests are >80% when recommended to the patient by their physician.6 Research-based guidelines offered by the United States Preventive Services Task Force4 state that there is fair evidence that annual FOBT and/or flexible sigmoidoscopy every 3–5 years be considered in a periodic health exam. The guide suggests that colonoscopy may have greater risks than potential benefits when screening average-risk adults. Screening colonoscopy receives a ‘C’ recommendation (defined as: "There is insufficient evidence to recommend for or against the inclusion of the condition in a periodic health exam, but recommendations may be made on other grounds"). On the other hand, recommendations from expert consensus groups tend to be more aggressive. The American Gastroenterological Association and The American Cancer Society recommend annual digital rectal exam for all adults beginning at age 40, annual FOBT beginning at age 50 and sigmoidoscopy every 5 years or a colonoscopy every 10 years beginning at age 50.
Mandel JS et al. Colorectal cancer mortality: effectiveness of biennial screening for fecal occult blood. J Natl Cancer Inst 1999; 91: 434–437.
Although it has been well established that annual FOBT decreases the mortality from colorectal cancer, the optimal frequency of screening is less clear. Two studies have shown a statistically significant mortality reduction with biennial testing.7,8 The 1993 randomized clinical trial from Minnesota which documented a 33% reduction in mortality with annual FOBT found only a 6% reduction in mortality with biennial screening (not statistically significant).5 In this follow-up study, the participants were followed for an additional 5 years to clarify further the benefit to those tested every 2 years. A total of 46 000 men and women, from a university setting in Minnesota, aged 50–80 years were randomized into three groups (control group; annual FOBT; biennial FOBT). Persons in the screened group were asked to prepare six slides consisting of two samples obtained from each of three consecutive bowel movements. Those with one or more positive specimens were invited for an extensive diagnostic work-up that included colonoscopy. Compliance with annual and biennial screening was 75 and 78%, respectively. Compliance with the diagnostic work-up for those with a positive screen was 95%. In the 18 years of follow-up, there were >15 000 deaths from all causes, with colorectal cancer accounting for 2.9% of them. All-cause mortality rates were similar in each of the three study groups. Death from colorectal cancer was 33% lower in those screened annually when compared with those in the control group.
Comment. There was a statistically significant 21% reduction in mortality in those screened biennially. This was consistent with larger studies done previously.7,8 It is important to remember that the benefit may be greater if the compliance rate is greater than the 78% seen in this study for biennial screening, if the follow-up period is longer and if there is no contamination by screening in the control group (i.e. usual care group).
Mandel JS et al. The effect of fecal occult blood screening on the incidence of colorectal cancer. N Engl J Med 2000; 343: 1603–1607.
This study, using the same participants and methods as in the previous manuscript, had as its goal to study the impact that FOBT screening had on colorectal cancer incidence. During the 18 years of follow-up, there were 417 new cancers in the annually screened group, 435 in the biennially screened and 507 in the usual care group. The differences in each of the two screened groups compared with the control group were statistically significant.
Comment. These data demonstrate a significant reduction in the incidence of colorectal cancer (not simply a decrease in mortality) in those screened with FOBT. Additionally, polyps visualized during colonoscopy were removed, corroborating the widely believed polyp–cancer link.
These data also imply a correlation between the number of slides positive for faecal occult blood and the polyp size. Larger polyps have been more closely associated with the development of colorectal cancer. This study suggests that it is the sensitivity of FOBT in detecting colorectal cancer precursors rather than chance finding during colonoscopy for the work-up of false-positive FOBT that is responsible for the success of FOBT programmes.
The appropriate work-up for patients found to have polyps on screening flexible sigmoidoscopies has not been clearly established. The value of a distal colonic polyp as a predictor of significant proximal disease remains uncertain. Colonoscopy may be a more thorough and accurate screening test for colorectal cancer but at an increased risk to the patient. The following studies will help to determine the appropriate work-up of polyps found on screening sigmoidoscopy and may add to the debate as to whether flexible sigmoidoscopic screening should be replaced by full colonic exam.
Levin TR et al. Predicting advanced proximal colonic neoplasia with screening sigmoidoscopy. J Am Med Assoc 1999; 281: 1611–1617.
Utilizing demographics and data from flexible sigmoidoscopic findings, this manuscript attempts to determine who is at higher risk for advanced proximal neoplasia (in this study defined as adenocarcinoma, adenomas >1 cm or with villous features or severe dysplasia located proximal to the splenic flexure). A total of 2972 asymptomatic subjects over 50 years of age underwent colonoscopy as follow-up to screening flexible sigmoidoscopy. Those with a first-degree relative diagnosed with colon cancer before the age of 50 were excluded. Five hundred and forty-four (18%) of the subjects were self-selected volunteers who had no adenoma found on routine flexible sigmoidoscopy. It was found that the risk of advanced proximal neoplasia was similar among those with a tubular adenoma of any size and those with no adenomas at sigmoidoscopy (5.5% incidence). This high rate of proximal neoplasms in those without sigmoidoscopic findings must be viewed with caution because these patients were not selected randomly. These findings suggest that tubular adenomas removed via sigmoidoscopy need not be followed-up with a complete exam of the colon. Randomized studies should be performed to clarify this result before one should consider changing their practice.
Subjects with tubulovillous adenoma and those with villous adenoma in this study had the greatest likelihood of advanced proximal neoplasia (12.1% incidence). The implication is that patients with these findings should receive diagnostic colonoscopy. Other significant predictors of advanced proximal neoplasia included age over 65 years, having more than one adenoma and a positive family history of colon cancer (defined as a first-degree relative who was over 50 years old when diagnosed or age unknown). These results should be useful when discussing various screening options with individual patients.
Imperiale TF et al. Risk of advanced proximal neoplasms in asymptomatic adults according to the distal colorectal findings. N Engl J Med 2000; 343: 169–174.
As with the previous study, this article may be important in determining both the optimal screening strategy, and patient selection for colonoscopy after sigmoidoscopy screening. Both studies, unlike those cited previously, contain control groups without distal abnormalities. This was a cross-sectional study of asymptomatic adults over the age of 50 years who underwent colonoscopy screening in a university setting in Indianapolis, Indiana. Distal colon was defined as that distal to the splenic flexure, and advanced neoplasm included polyps with villous features, high-grade dysplasia or cancer. Ninety-seven per cent of the 1994 colonoscopic exams were completed successfully to the caecum.
There were a total of 61 advanced neoplasms in the distal colon (these would have been picked up by sigmoidoscopy). There were 50 advanced lesions in the proximal colon. Of the 50 proximal lesions, seven had distal neoplasms, 20 had benign distal findings (eight hyperplastic polyps; 12 tubular lesions) and 27 had no findings. If the physician were to refer all those with any distal findings for colonoscopy, then 22% of the patients would have received colonoscopy following screening sigmoidoscopy and 79% of the total advanced neoplasms (including proximal and distal) would have been detected. If only those with advanced neoplasms found on sigmoidoscopy were referred, then 68 of the 111 (61%) advanced neoplasms would have been detected. With this less aggressive approach, only 68 of the 1994 (3.4%) patients would have received colonoscopy.
Benign lesions found on sigmoidoscopy were associated with an increased risk of advanced proximal lesions. Other risk factors for proximal lesions included male sex and advanced age (specifically, for every 5 years over the age of 50, the risk increased by 32%). In the 1994 patients studied, there were no deaths and only one had perforation that required medical management only. Three patients had bleeding complications from polypectomy, none requiring transfusion. It is important to note that it is unclear whether the proximal lesions would have been detected by FOBT if that was incorporated into the screening protocol.
Lieberman DA et al. Use of colonoscopy to screen asymptomatic adults for colorectal cancer. N Engl J Med 2000; 343: 162–168.
The investigators studied 3121 asymptomatic persons who underwent screening colonoscopy with 98% completed to the caecum. Of these, 10.5% were found to have advanced neoplasms. This was a higher percentage than in the previous study because adenoma of >1 cm was included in the advanced neoplasm category and these were higher risk patients (a VA population of mostly men, and those with a family history were not excluded). Again, many persons without distal polyps were found to have neoplasms in the proximal colon (2.7%). As previously found, those with polyps in the distal colon were more likely to have advanced proximal neoplasia than those without. Fifty-two per cent of those with advanced proximal neoplasia had no distal adenomas.
In the hands of these experienced endoscopists, only 10 patients (0.3%) had serious complications. This included six patients who were hospitalized for gastrointestinal bleeding, one myocardial infarction, one cerebrovascular accident, one Fornier's gangrene and one thrombophlebitis. There were no perforations or deaths attributed to this procedure.
Conclusion
In recent years, the debate over whether or not to screen for colon cancer has changed to a discussion of what screening test is best. If you prefer reliable data found with randomized controlled trials when making recommendations, then FOBT screening may be most attractive. The study by Mendel et al. added to the evidence that biennial FOBT is effective and should be included in any discussion about screening options. Flexible sigmoidoscopy has been shown in this review and in many other studies to be an effective screening tool. Simoidoscopies are safer, less expensive and more convenient than colonoscopy; therefore, patients are more likely to get the test. The studies in this review have added to the evidence that colonoscopies are more sensitive for detecting colonic neoplasia than flexible sigmoidoscopy.
The indications for full colonic exam following sigmoidoscopy findings remain unclear. The best evidence suggests that there is a greater risk of advanced neoplasm in those with benign-appearing distal lesions. The studies have also demonstrated an increased risk with advancing age and in those with a family history of colon cancer. If a full colonic exam in all of those with benign distal lesions is too aggressive, then possibly only those with other risk factors should receive this full evaluation.
With so few Americans receiving any of the recommended screening evaluations, physicians, the media and educators must be more diligent in recommending screening for early detection of colorectal cancer. The appropriate amount of risk and inconvenience must be balanced by the efficacy of the screening test. The literature does not support one screening test as superior to the rest. Using the patient's values and perspectives, physicians must help guide the patient in their selection. Physicians should not be rigid in their recommendations. With today's controversies regarding colorectal cancer screening, "redefining the best test as the one the patient wants may save the most lives".9
References
1
Landis Sh, Murray T, Bolden S, Wingo PA. Cancer statistics, 1999. CA Cancer J Clin 1999; 49: 8–13.
2 Harrison's Principles of Internal Medicine, 14th edn. McGraw-Hill Companies Inc., 1998.
3 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.[Abstract]
4 United States Preventive Services Task Force. Guide to Clinical Preventive Services: Report of the US Preventive Services Task Force. Baltimore: Williams and Wilkins, 1996.
5
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.
6 Leard LE, Savides TJ, Ganiats TG. Patient preferences for colorectal cancer screening. J Fam Pract 1997; 45: 211–218.[Web of Science][Medline]
7 Hardcastle JD, Chamberlain JO, Robinson MH et al. Randomized controlled trial of faecal-occult-blood screening for colorectal cancer. Lancet 1996; 348: 1472–1477.[Web of Science][Medline]
8 Kronberg O, Fenger C, Olsen J, Jorgenson OD, Sandergard O. Randomized study for screening for colorectal cancer with faecal-occult-blood test. Lancet 1996; 348: 1367–1371.
9
Woolf SH. The best screening test for colorectal cancer—a personal choice. N Engl J Med 2000; 343: 1641–1642.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||