Screening for colorectal cancer, 1996

Finlay A Macrae

Readers may print a single copy for personal use. No further reproduction or distribution of the articles in whole or in part should proceed without the permission of the publisher. For copyright permission, contact the Australasian Medical Publishing Company

Screening average-risk people aged 55 to 70 years for colorectal cancer is now a public health priority in Australia. Pilot studies of faecal occult blood testing are required to find ways of achieving optimal compliance and cost efficiency in the Australian health care setting. Flexible sigmoidoscopy probably should be used as complementary screening but further trials are needed. High-risk groups (family history of colorectal cancer, or previous ulcerative colitis, adenomas or cancer) should already be in surveillance programs.
(MJA 1996; 165: 102-105)

Introduction - Average risk screening - Above average risk - Conclusion - References - Author's Details - Table 1 - Table 2 - Table 3

Home | Issues | MJA shop | Terms of use | MJA Careers | More... | Contact | Topics | Search | RSS

mja.com.au | The Medical Journal of Australia

Introduction

At the same time, techniques to detect colorectal cancer at an early (curable) stage have been refined, and large, long-term randomised studies to prove that these techniques can reduce mortality from colorectal cancer are nearing completion. ^2-6 Our confidence in determining the outcome of these trials has increased with the development of rigorous methods to assess benefit. ⁷

Finally, the leading role that colorectal cell and molecular biology has taken in the science of carcinogenesis ⁸ means that delivery of molecular biological diagnostic services for cancer should already be a clinical service, at least for familial adenomatous polyposis and some other familial colo rectal cancers.

Given the state of our knowledge, what should be done now about colorectal screening?

Average risk screening

Faecal occult blood testing The best chance of cure in colorectal cancer is surgical treatment when the disease is localised to the bowel wall (Australian Clinicopathological Staging System A [ACPS A]) or at least does not involve lymph nodes if it has penetrated through the muscle wall (ACPS B). This means the best chance of diagnosis is before symptoms develop.

The most logical method of diagnosis is population screening to identify people with early-stage or premalignant (adenomatous polyps) tumours, but it is very difficult to prove that this will reduce mortality rate. Such proof is required to attract scarce health dollars. Rigorous controlled trials are needed to counter possible biases (e.g., lead-time bias: screening allowing an earlier diagnosis but not altering date of death or the tumour's natural history; and length bias: screening selecting biologically favourable tumours, such as slow-growing ones) and the requirement to analyse on an intention-to-treat basis ⁹ means recruiting over 100 000 people to adequately test the screening hypothesis.

Three large randomised controlled trials of faecal occult blood testing are in progress ^4-6 and several have been completed (Table 1). ^2,3 The most important completed trial is the Minnesota trial, involving 46 551 subjects studied over 13 years and randomly allocated to annual, biennial or no Hemoccult testing (SmithKline Diagnostics Inc, San Jose, Calif, USA). ² The trial demonstrated a statistically significant 33% relative reduction in mortality from colorectal cancer in subjects having annual testing. This result has withstood close scrutiny by critics, including those who believed that most of the benefit was attributable to chance because of the low specificity of the rehydrated Hemoccult test used, leading to over 30% of the screened population having a colonoscopy. ¹⁰ If the benefit was merely serendipitous, then the chance of a single positive occult blood test versus multiple tests (six were done with each screen) should not be linearly related (which it was), and the odds ratio for tumours in individuals testing positive versus negative to the presence of faecal occult blood would not have reached such high levels, far outstripping almost all other known risk factors. Thus, the Minnesota study shows that faecal occult blood testing does select patients with significant colorectal tumours. The three incomplete European randomised controlled trials are also on track to show a mortality reduction. ^4-6

Supporting data come from case-control studies of faecal occult blood testing (Table 2). ^11-13 The screening histories of people dying from colorectal cancer (cases) are compared with those of control subjects without apparent disease. A deficit of screening in the cases implies protection to the controls by screening. However, rigorous standards need to be applied to case-control studies to counter the subtle biases, which may account for the generally more favourable levels of protection found when compared with the results of randomised controlled trials. ⁷

Cohort studies of screening for colo-rectal cancer are curiously missing from the literature. Large cohort studies of diet and other possible risk factors for colo-rectal cancer have been published, ^14,15 but none have recorded incidence of prior screening in those in the prospective cohort developing cancer versus those not developing cancer. Screening information is not reported in these studies, which seems a valuable opportunity lost. Many of the biases confounding case-control studies would be overcome in a cohort study of screening. New immunochemical tests have already demonstrated improved sensitivity and specificity, which will enhance the cost effectiveness and efficacy of screening programs. ¹⁶

Despite a statistically significant mortality benefit, the absolute number of lives saved per 1000 persons screened for faecal occult blood is small (about three). ² This, of course, is a corollary of well patient screening (most of the people being screened are both free of disease and have negative results). We need to conduct studies to define cost-efficient procedures for screening, using tests with low false positive rates, and targeting particular communities.

Sigmoidoscopy Sigmoidoscopy is likely to be an important adjunct for detection of distal colorectal cancer, as faecal occult blood testing has low sensitivity for rectosigmoid cancers. ¹⁷ Case-control studies of sigmoidoscopy screening show protection to the control groups (Table 2), ^18,19 but it is difficult to identify the reason for the mortality benefit in terms of polyps removed or successful surgery for cancers detected at an early stage. Well accepted randomised controlled trials have not been done. Flexible sigmoidoscopy should detect 50%-70% of colorectal cancers, so its impact in screened populations is likely to be high. Indications from case-control studies are that protection lasts for 10 years, which augurs well for cost-benefit considerations. ¹⁸ Large randomised controlled trials of flexible sigmoidoscopy screening are commencing in Europe. ²⁰

Colonoscopy Even once-off colonoscopy is comparable in cost with breast cancer screening. However, as with flexible sigmoidoscopy, colonoscopy screening does not have enough clinical research proof to be sure of its benefit, although this is probably just a matter of time. ²¹ Compliance in Australian populations at average risk is unknown.

From clinical science to clinical service In summary, faecal occult blood testing is the screening method with the stongest data for deployment. But, should we wait for the results of the additional European controlled trials of screening, or act now on the evidence from the Minnesota trial and other published information? Despite its low sensitivity (about 50% for asymptomatic cancers), ^4,21 repetitive screening programs have shown a reduction in mortality from colorectal cancer. Testing is cheap and simple and reasonably well accepted in Australian populations. ²² Provided specificity is high (and thus false positive results low), follow-up colonoscopy is both feasible and cost-manageable. ²³

Although the scientific basis for implementing faecal occult blood screening for those aged 50-75 years of age is compelling, the best way to do this in Australia is uncertain. Should it be through dedicated centres providing testing and evaluation, as with breast screening in Australia? Should it be through general practice? through a central electoral-roll-resourced organisation? through hospitals? or through State cancer foundations/councils? These logistical questions need to be answered by pilot trials. ²⁴ Mechanisms to evaluate the results on a population basis must be built into whichever system is introduced. Germany has funded faecal occult blood mass screening for over a decade but has a poorly developed evaluation infrastructure, thus denying the world an important opportunity for assessment.

Sigmoidoscopy needs further study before introduction, but sufficient evidence is available to offer sigmoidoscopy to average-risk people in the age group 50 to 75 years who are already in the health care system (case finding). Technical skills and facilities for flexible sigmoidoscopy are limited and their availability must be increased. Issues of cleaning and disinfection, environmental hazards of glutaraldehyde, technical and lesion-recognition training, manpower, support personnel, capital support for equipment for a "type C" (office-based) procedure, all complicate the widespread use of flexible sigmoidoscopy. However, the benefits of being able to examine 50%-70% of the tumour territory of the large bowel make the objective worthwhile.

Above average risk

Family history For individuals with a family history of colorectal cancer the appropriate recommendations are generally less controversial (Table 3). Many consistent studies demonstrate risks two- to sixfold higher in individuals with first degree relatives with colorectal cancer, the risk relating to the earliest age of onset in the family and the number of first degree relatives affected. ^25,26 If there is only one affected first degree relative, age at onset of the cancer is important: if the relative was under 55 years, the risk is six times the average risk (the same as the risk for those with two first degree relatives with onset at any age); if a single affected relative has older-onset cancer, the increased risk to family members is marginal. ²⁵ Thus, average-risk screening is recommended for those families with a single affected relative with older-age onset ( > > 55 years) of colorectal cancer (faecal occult blood testing and five-yearly flexible sigmoidoscopy), but colonoscopy every five years is recommended for those with a first degree relative with early-age onset. ²⁴ If multiple family members are affected in a nuclear family, colonoscopy is recommended for first degree relatives of affected members, either five-yearly or three-yearly, depending on whether two or more relatives are affected. This advice is based on logical grounds of cost benefit: the higher the risk (and therefore prevalence) the greater the positive predictive value of screening, and thus cost-benefit. Protection may be enhanced by annual faecal occult blood testing between colonoscopies.

The significance of a family member with an adenoma (as distinct from a cancer) is less clear in defining risks to individuals in the family. Epidemiological studies suggest that adenomas are equivalent to cancers, ^27,28 but problems of detection bias make it difficult to use this information for surveillance advice. Nevertheless, a relative with a confirmed adenoma should not be ignored in assessing family history, and for risk assessment purposes it should be considered that these lesions may represent early cancer. Hereditary Non-Polyposis Colorectal Cancer syndrome is suspected when three or more first degree relatives have colorectal cancer (especially right-sided or multiple), with at least one affected family member < 50 years of age or when colo rectal cancer is associated with ovarian, endometrial, gastric, small-bowel, pancreatic, biliary, kidney or ureteric tumours. ^29,30 Genotyping for these families is available in Australian research studies, but it may take over 12 months to define the family- specific mutation of the mismatch repair genes responsible. ³¹ Colonoscopy every two years for affected and at-risk family members is important, with faecal occult blood testing between colonoscopies.

Familial adenomatous polyp-osis should also be suspected in families with early-age-onset colo-rectal cancer. Enquiry should be made about polyps or polyposis in affected members and confirmation of the diagnosis sought. Genotyping is available but is still technically difficult due to the large gene responsible ( APC gene); mutations at different sites in the gene are responsible for the condition in different families. ³² Once the mutation in a family is identified in genomic DNA, a family-specific DNA test is easily developed and available for predictive (100% accurate) testing in at-risk family members of any age. Expert genetic counselling is required. Issues of threats to insurance of all types, employment, survivor guilt, prognosis and family planning all need to be handled sensitively and informatively both before and after testing. Screening of APC- mutation carriers needs to proceed as before: annual or biennial flexible sigmoidoscopy screening from 15 to 55 years of age. Those without the mutation are recommended to have average risk screening. They are still at risk for common colorectal cancer.

Previous adenomas or cancer: "polyp follow-up" Patients with previous adenomas or cancers face a similar risk for future cancer. ^33-36 Although no randomised controlled trials have been reported, several large follow-up studies of patients with adenoma have compared outcomes with suitable control groups -- some with adenomas untreated, some with age-sex-matched population control groups, some with patients with polyps detected by barium enema (not removed) before colonoscopy was available -- and demonstrated a convincing reduction in mortality from regular follow-up. ^34,35,37,38 The only exception is patients with small tubular adenoma(s) in the rectosigmoid region, where risk for subsequent cancer is no higher than average and surveillance is unlikely to be of particular value. ³⁵ Intervals of surveillance are controversial but the United States National Polyp Study clearly demonstrated that annual follow-up conferred no benefit over three-yearly follow-up; ³⁹ two- to five-yearly follow-up is recommended, depending on multiplicity, size of index adenoma and, possibly, family history -- all independent risk factors for subsequent neoplastic risk.

Ulcerative colitis Ulcerative colitis poses a cancer risk after eight years for total disease and 12 years for left-sided disease. ⁴⁰ The programs with best results advocate multiple biopsies (18 biopsies from eight colonic sites). Controlled trials have not been performed. ⁴⁰

In summary, controlled trials of surveillance in these above average risk groups would confirm the recommendations but, on current knowledge, ethical considerations would not permit randomisation to no surveillance. Case-control and cohort studies in these above average risk groups providing information on benefit have not been published.

Conclusion

The optimal processes for implementing screening in a cost efficient manner using up-to-date screening tests require studies in pilot public health programs. Australia needs to move forward in this area, to help control its commonest internal cancer (that touches so many Australian families), and its second-largest cause of cancer mortality.

References

1. Coates M, McCredie M, Armstrong B. New South Wales Central Cancer Registry, Cancer in New South Wales, Annual reports 1973-1990. Incidence and mortality. Sydney: NSW Cancer Council, 1991.
2. 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.
3. Winawer SJ, Flehringer BJ, Schottenfeld D, Miller DG. Screening for colorectal cancer with fecal occult blood testing and sigmoidoscopy. J Natl Cancer Inst 1993; 85: 1311-1318.
4. Hardcastle JD, Thomas WM, Chamberlain J, et al. Randomised controlled trial of faecal occult blood screening for colorectal cancer: results of first 107 349 subjects. Lancet 1989; 1: 1160-1164.
5. Kronborg O, Fenger C, Olsen J, et al. Repeated screening for colorectal cancer with faecal occult blood test: a prospective randomized study at Funen, Denmark. Scand J Gastroenterol 1989; 24: 599-606.
6. Kewenter J, Bjork S, Haglind E, et al. Screening and re-screening for colorectal cancer. A controlled trial of fecal occult blood testing in 27 700 subjects. Cancer 1988; 62: 645-651.
7. Gill TM, Horwitz RI. Evaluating the efficacy of cancer screening: clinical distinctions and case-control studies. J Clin Epidemiol 1995; 48: 281-292.
8. Ee HC, Young JP. DNA repair and inherited cancer. J Gastroenterol Hepatol 1995; 10: 108-109.
9. Newell D. Intention-to-treat analysis: implications for quantitative and qualitative research. Int J Epidemiol 1992; 21: 837-841.
10. Lang CA, Ransohoff D. Fecal occult blood screening for colorectal cancer. JAMA 1994; 271: 1011-1013.
11. Selby JV, Friedman GD, Quesenberry CP, Weiss NS. Effect of fecal occult blood testing on mortality from colorectal cancer. A case-control study. Ann Intern Med 1993; 118: 1-6.
12. Wahrendorf J, Robra B-P, Wiebelt H, et al. Effectiveness of colorectal cancer screening: results from a population-based case-control evaluation in Saarland, Germany. Eur J Cancer Prev 1993; 2: 221-227.
13. Saito H, Soma Y, Koeda J, et al. Reduction in risk of mortality from colorectal cancer by fecal occult blood screening with immunochemical hemagglutination test. A case-control study. Int J Cancer 1995; 61: 465-469.
14. Willett W, Stampfer MJ, Colditz GA, et al. Relation of meat, fat and fiber intake to the risk of colon cancer in a prospective study in women. N Engl J Med 1990; 323: 1664-1672.
15. Phillips RL, Snowdon DA. Dietary relationships with fatal colorectal cancer among Seventh-day Adventists. J Natl Cancer Inst 1985; 74: 307-317.
16. St John DJB, Young GP, Alexeyeff MA, et al. Evaluation of new occult blood tests for detection of colorectal neoplasia. Gastroenterology 1993; 104: 1661-1668.
17. Macrae FA, St John DJB. Relationship between patterns of bleeding and Hemoccult sensitivity in patients with colorectal cancers or adenomas. Gastroenterology 1982; 82: 891-898.
18. Selby JV, Friedman GD, Quesenberry CP, Weiss NS. A case-control study of screening sigmoidoscopy and mortality from colorectal cancer. N Engl J Med 1992; 326: 653-657.
19. Newcomb PA, Norfleet RG, Stoner BE, et al. Screening sigmoidoscopy and colorectal cancer mortality. J Natl Cancer Inst 1992; 84: 1572-1575.
20. Atkin WA, Cuzick J, Northover JMA, Whynes DK. Prevention of colorectal cancer by once-only sigmoidoscopy. Lancet 1993; 341: 736-740.
21. Rex DK, Lehman GA, Hawes RH, et al. Screening colonoscopy in asymptomatic average risk persons with negative fecal occult blood tests. Gastroenterology 1991; 100: 64-67.
22. Macrae FA, Hill I, St John DJB. Predicting colon cancer screening behaviour from health beliefs. Prev Med 1984; 13: 115-126.
23. Frommer DJ, Kappario A, Brown MK. Improved screening for colorectal cancer by immunochemical detection of occult blood. BMJ 1988; 296: 1092-1094.
24. Australian Gastroenterology Institute/Australian Cancer Society. Guidelines for early detection screening and surveillance for colorectal cancer, 2nd ed. Sydney: Australian Gastroenterology Institute/Australian Cancer Society, 1994.
25. St John DJB, McDermott FT, Hopper JL, et al. Cancer risk in relatives of patients with common colorectal cancer. Ann Intern Med 1993; 188: 785-790.
26. Lovett E. Family studies in cancer of the colon and rectum. Br J Surg 1976; 63: 13-18.
27. Maire P, Morichau-Beauchant M, Drucker J, et al. PrŽvalence familiale du cancer du c(tm)lon, et du rectum: rŽsultats d'une enqu�te "cas-tŽmoins" de 3 ans . Gastroenterol Clin Biol 1984; 8: 22-27.
28. Winawer SJ, et al. Risk of cancer in relatives of patients with adenomas. N Engl J Med. In press.
29. Lynch HT, Smyrk TC, Watson P, et al. Genetics, natural history, tumour spectrum and pathology of hereditary non-polyposis colorectal cancer: an updated review. Gastroenterology 1993; 104: 1535-1549.
30. Macrae FA. Bowel cancer: watching over the family. J Gastroenterol Hepatol 1995; 10: 337-338.
31. Kohonen-Cornish M, St John DJB, Macrae FA, et al. Genetic defects in hereditary non-polyposis colon cancer. J Hepatol Gastroenterol 1995; 10: A38.
32. Nagase H, Miyoshi Y, Horii A, et al. Correlation between the location of germ-line mutations in the APC gene and the number of colorectal polyps in familial adenomatous polyposis patients. Cancer Res 1992; 52: 4055-4057.
33. Grossman S, Milos ML, Tewaka IS, Jewell NP. Colonoscopic screening of persons with suspected risk factors for colon cancer. II. Past history of colorectal neoplasms. Gastroenterology 1989; 96: 299-306.
34. Williams CB, Macrae FA. The St Mark's neoplastic polyp follow-up study. In: Rozen P, Winawer SJ, editors. Secondary prevention of colorectal cancer: An international perspective. Basel: Karger, 1986: 226-242.
35. Atkin WS, Morsm BC, Cuzick J. Long-term risk of colorectal cancer after excision of rectosigmoid adenomas. N Engl J Med 1992; 326: 658-662.
36. Bulow S, Svendsen LB, Mellemgaard A. Metachronous colorectal cancer. Br J Surg 1990; 77: 502-505.
37. Winawer SJ, Zauber AG, et al. Prevention of colorectal cancer by colonoscopic polypectomy. N Engl J Med 1993; 329: 1977-1981.
38. Murakami R, Tsukuma H, Kanamori S, et al. Natural history of colorectal polyps and the effect of polypectomy on occurrence of subsequent cancer. Int J Cancer 1990; 46: 159-164.
39. Winawer SJ, Zauber AG, O'Brien MJ, et al. Randomized comparison of surveillance intervals after colonoscopic removal of newly diagnosed adenomatous polyps. N Engl J Med 1993; 328: 901-906.
40. Lennard Jones JE, Melville DM, Morson BC, et al. Precancer and cancer in extensive ulcerative colitis: findings among 401 patients over 22 years. Gut 1990; 31: 800-806.
41. US Preventive Services Task Force. Guide to clinical preventive services. 2nd ed. Baltimore: Williams & Wilkins, 1996: 89-103.
42. Solomon MJ, McLeod RS and the Canadian Task Force on the Periodic Health Examination. Periodic health examination, 1994 update: 2. Screening strategies for colorectal cancer. Can Med Assoc J 1994; 150: 1961-1970.

For editorial comment, see Solomon, page 68; see also Olynyk et al., page 74.

Author's details

To top of article - ©MJA; 1996