Screening for colorectal cancer: too early to be dogmatic

The emotive, ethical and economic issues raised by population screening programs remain a challenge

©MJA1996; 165: 68-69

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

Support for screening for colorectal cancer in Australia has been gradually brewing, with reports in the national and international literature, ^1-3 and evidence suggesting that colorectal cancer is increasing in prevalence. Despite improvements in surgical technique and adjuvant therapies, overall survival in the symptomatic phase of the disease has remained static. Now a well performed, randomised controlled trial of repetitive faecal occult blood testing has shown, for the first time, a small survival benefit from screening, ¹ and evidence from case-control studies suggests that flexible sigmoidoscopy screening may have an even greater mortality benefit. ^2,3 In this issue of the Journal, two articles -- a review of screening for colorectal cancer by Macrae ( page 102 ), and a pilot study of flexible sigmoidoscopy screening for colo rectal cancer by Olynyk et al. ( page 74 ) -- highlight this renewed interest in screening for colorectal cancer.

Clinical epidemiology and the assessment of screening have evolved into a science, with epidemiologists, clinicians and statisticians collaborating in government working parties throughout the world to assess preventive practices for cancers. ⁴ In 1996 the momentum for screening prompted an Australian Health Technology Advisory Committee Working Party on colorectal cancer screening to try to sort out the evidence for and against screening, and to establish guidelines for national programs. This is no small task. MEDLINE alone detects over 800 relevant publications on bowel cancer screening, and, as computer searches miss more than 50% of randomised controlled trials, there is obviously a vast ocean of written "evidence" to be found and assessed. ⁵

Despite reviewing the same body of evidence, collaborating on initial interpretations and adopting a scientific approach to the assessment of screening for colorectal cancer, the Task Forces of Canada and the United States have made different recommendations. ^6,7 A stepwise approach to assessing colorectal cancer screening identifies the point in the assessment where the two bodies disagree and the Australian Working Party must adjudicate for the Australian population (Box). ⁴ The major differences relate to the interpretation of one randomised controlled trial of faecal occult blood testing and two case-control studies of flexible sigmoidoscopy, although the much-awaited results of two further randomised controlled trials of faecal occult blood testing should be available in the near future. ^1-3

There is agreement that the Minnesota trial ³ has shown a statistically significant cancer-specific mortality benefit and that this equates to 2.95 lives saved per 1000 people repetitively screened annually with faecal occult blood testing for 13 years. A 10% false positive rate and the 86% quoted compliance would mean an estimated colonoscopy procedural cost alone of over $100 000 per life saved after 13 years in Australia. The United States and Canadian Task Forces have agreed on the efficacy, effectiveness and efficiency of the data (Steps 1-3), but differ in their assessment of the small but expensive cancer-specific mortality benefit of faecal occult blood testing and the allocation of limited resources (Step 4).

Interpretation of case-control studies for flexible sigmoid o scopy has remained at the proof-of-efficacy stage (Step 1). Does an 18% difference in rates of flexible sigmoidoscopy for those who died from colorectal cancer equate to a survival benefit in the control subjects who have not died from this disease? ² The alternative hypothesis is that this difference is due to selection bias inherent in this retrospective study design. ⁷ The lack of detection of any colorectal cancer in the control group, and the detection of adenoma in only 1% of control subjects, imply no mortality benefit from their 24% sigmoidoscopy rate. ² True mortality benefit would have to come from an intervention that detected early cancers or high risk adenomas and resulted in successful treatment. This would suggest that the cases and controls are not comparable and that a randomised controlled trial is still required to answer this question. The Fremantle pilot project reported by Olynyk et al. ( page 74 ) detected colorectal cancer at their estimated cost of $500 per screened patient (unpublished data). Thus, the true cost of detecting a cancer with flexible sigmoidoscopy in Australia may be $85 000, with no evidence as yet of a mortality benefit.

The same data can and will be interpreted differently by different authors, health departments, societies and foundations. To some extent screening can be supported by the current data, but dogmatic conclusions and recommendations remain controversial and raise issues for discussion. Currently, these revolve around the statistical versus clinical significance of the small cancer-specific mortality benefit of annual faecal occult blood screening; the lack of detection of any cancers or significant polyps in the control groups of case-control studies of flexible sigmoidoscopy to explain any mortality benefit from sigmoidoscopy; and the lack of compliance with many of the screening programs. There is also a lack of research into the potential adverse effects of national screening programs, such as physical complications and problems arising from labelling or mis labelling of people screened (which may have psychological effects and also life and medical insurance repercussions). Even the earlier detection of incurable disease may have adverse effects. ⁴ The general population's desire for cancer screening may not be as great as that of its advocates; non-compliance would then impact upon efficacy.

There is little doubt the emotive and ethical issues raised by population screening programs will challenge us for many years to come. Costs need to be compared with number of lives saved, and screening programs balanced against other programs competing for the limited available health care resources. Programs are often promoted by interested parties with more rhetoric than scientific evidence. The exciting advances of recent years -- that colorectal cancer can be detected and treated in the asymptomatic phase with improvement in mortality -- have encouraged clinical researchers, but the best means of achieving this has yet to be established. It may be that the advances and ethical quandaries of genetic testing hold the key to the future and will be the focus of this debate.

Michael J Solomon
Colorectal Surgeon
Director of Research, University of Sydney, and Central Sydney Departments of Colorectal Surgery, Sydney, NSW

1. Mandel JS, Bond JH, Church TR, et al. Reducing mortality from colorectal cancer by screening for faecal occult blood. N Engl J Med 1993; 328: 1365-1371.
2. 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.
3. Newcomb PA, Norfleet RG, Stoner BE, et al. Screening sigmoidoscopy and colorectal cancer mortality. J Natl Cancer Inst 1992; 84: 1572-1575.
4. Hunt J, Solomon MJ. Screening for cancer: the science behind the rhetoric. Int J Surg Sci 1995; 2: 161-167.
5. Solomon MJ, McLeod RS, Laxamana A, Devore L. Randomized controlled trials in surgery. Surgery 1994; 115: 707-712.
6. US Preventive Sevices Task Force. Guide to clinical preventive services. 2nd ed. Baltimore : Williams & Wilkins, 1996: 89-104.
7. 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.

Make a comment - Register to be notified of new articles by email - Current contents list - To top of article - ©MJA; 1996