Virtually viewing the large bowel: the future of colorectal cancer screening?

The problem is that all current screening tools are imperfect. FOBT is most widely advocated as the only test shown to reduce mortality from colorectal cancer (by 15%-33%) when used for mass population screening.¹ It is also cheap, safe and can be administered by the general practitioner. Accuracy depends on the type of FOBT used and the frequency of testing, but, on an individual basis, FOBT misses between 21% and 63% of cancers and most adenomas, and has a false-positive rate of 2%-13%.³

Flexible sigmoidoscopy is under trial both in Australia and overseas as a tool for population screening. The rationale is that most neoplasms occur within reach of the flexible sigmoidoscope, and that distal adenomas may be predictors of proximal lesions. In this issue of the Journal, Nicholson and colleagues⁴ show, as have others,⁵ that, among screened subjects with adenomas, 25% have proximal adenomas only (defined by Nicholson et al as proximal to the splenic flexure). These are beyond the reach of flexible sigmoidoscopy. This study illustrates one of the problems of flexible sigmoidoscopy screening -- accuracy for cancer and polyp detection. However, other important issues must be considered in assessing a screening test, such as acceptability, compliance, availability, safety and cost. Indeed, in the Australian context, initial participation rates in flexible sigmoidoscopy screening have been disappointing (12%),⁶ although recent data indicate that these rates have increased to around 40% (Associate Professor John Olynyk, Department of Gastroenterology, Fremantle Hospital, Fremantle, WA, personal communication).

The findings of Nicholson and colleagues support the need for imaging the whole colon in colorectal neoplasm screening. Methods advocated for this include double-contrast barium enema and colonoscopy. Both have their supporters. However, data on use of these methods for population screening of average-risk individuals are limited, and both have drawbacks that make them unlikely to be widely accepted for mass screening. Setting aside considerations of compliance, double-contrast barium enema is probably not sufficiently accurate without concomitant flexible sigmoidoscopy (which would increase costs and almost certainly decrease compliance), and imposes a significant radiation dose. Total colonoscopy has the advantages of accuracy and ability to combine screening with therapy (polypectomy) but carries a small but significant risk.⁷ A certain level of competence is required to achieve adequate rates of caecal intubation, and, although it is difficult to determine exact completion rates, outside specialist centres they may be only 80%-90%^8,9 or less. Taking into account the need for sedation, consequent bed fees and cost of time off work, colonoscopy is relatively expensive.

A recent contender for screening is virtual colonoscopy (computed tomography [CT] colography). After bowel preparation, the colon is insufflated with air or carbon dioxide, and a spiral CT scan performed, preferably in supine and prone positions. Because of the volumetric nature of data acquisition, sagittal and coronal reformatted images can be viewed, as well as the source axial images, and endoluminal images can be reconstructed, simulating an endoscopic view. Navigation using these images can be achieved by manual manoeuvres or "fly-through" techniques that automatically centre on the bowel lumen. While not yet as accurate as colonoscopy for polyp detection, virtual colonoscopy is likely to become significantly more accurate with expected developments in hardware and software. Currently, virtual colonoscopy is more accurate than FOBT and can probably compete with flexible sigmoidoscopy with regard to larger polyps. A study from Boston has reported sensitivities of 91%, 82% and 55% for polyps of diameter 10 mm or more, 6-9 mm, and 5 mm or less, respectively.¹⁰

Virtual colonoscopy also has several potential advantages as a screening tool: it is minimally invasive and quick for the patient (the scan takes only a few minutes); no sedation is required; and initial studies have shown that it is highly acceptable to patients.¹¹ Its "high-tech", virtual reality profile makes it potentially attractive to the lay public. While using ionising radiation, dosages are considerably less than for double-contrast barium enema when low-dosage protocols are used. Using the current scanning protocol in our institution, total effective radiation dose has been calculated to be less than 5 mSv, even when supine and prone scans are performed (compared with about 8 mSv for conventional double-contrast barium enema). In addition, early studies hold out the possibility that magnetic-resonance (MR) virtual colonoscopy may eventually supersede CT virtual colonoscopy, eliminating ionising radiation.¹² Lastly, there is the potential to detect incidental extracolonic disease, such as asymptomatic aortic aneurysms and renal carcinoma.

Problems currently limiting the application of virtual colonoscopy as a screening tool include its lack of sensitivity for small polyps, particularly those 5 mm or less in diameter.¹⁰ Does this matter in the context of a screening program? Probably not: the chances of a 5 mm lesion being malignant are negligible; if screening takes place every five years the dwell time for such a small lesion allows an enlarging lesion to be picked up on subsequent examinations. A further limitation is the need for bowel preparation, which is likely to be a significant factor in reducing participation rates. However, the use of faecal tagging to allow software to differentiate faeces and polyps may eventually minimise, or even eliminate, the need for bowel preparation.

In addition, other factors, such as availability, operator experience and cost, need to be evaluated in assessing the potential role of virtual colonoscopy as a screening tool for colorectal cancer. Much of the present cost is related to the time required for image processing and reading of the images by the radiologist, which is as long as 30-45 minutes with current commercially available technology. However, this time will inevitably be reduced significantly by further technological advances, such as faster computer processing and automated polyp detection software. Finally, it would be preferable that images are read promptly so that individuals with abnormalities have the opportunity of proceeding to same-day colonoscopy to avoid the need for a second bowel preparation.

So, is virtual colonoscopy a viable option as a screening tool for colorectal cancer in the average-risk individual? There is little doubt that, in its current state of development, it is not ready for widespread use. In addition to the limitations already discussed, the excellent sensitivity data reported by some centres^10,13 have not been widely replicated.¹⁴ Equally, there is a high probability that, at its rate of evolution, in the not-too-distant future CT (or MR) virtual colonoscopy will become an accepted (or even the accepted) modality for colorectal cancer screening. In the meantime, while waiting for the technology to catch up, feasibility studies of virtual colonoscopy are needed to examine issues such as participation rates, factors affecting recruitment into screening programs, acceptability and cost.

Richard M Mendelson
Radiologist

Geoffrey M Forbes
Gastroenterologist, and Clinical Senior Lecturer
University of Western Australia, Royal Perth Hospital, WA

Disclosure statement: The authors are active in clinical research into virtual colonoscopy and are planning a feasibility study of the technique in colorectal cancer screening.

1. Australian Health Technology Advisory Committee. Colorectal cancer screening. Canberra: AGPS, 1997.
2. National Health and Medical Research Council. Guidelines for the prevention, early detection and management of colorectal cancer. Canberra: NHMRC, 1999.
3. Allison JE, Tekawa IS, Ransom LJ, Adrain AL. A comparison of fecal occult blood tests for colorectal-cancer screening. N Engl J Med 1996; 334: 155-159.
4. Nicholson FB, Korman MG, Stern AI, Hansky J. Distribution of colorectal adenomas: implications for bowel cancer screening. Med J Aust 2000; 172: 428-430.
5. Kadakia SC, Wrobleski CS, Kadakia AS, Meier NJ. Prevalence of proximal colonic polyps in average-risk asymptomatic patients with negative fecal occult blood tests and flexible sigmoidoscopy. Gastrointest Endosc 1996; 44: 112-117.
6. Olynyk JK, Aquilia S, Fletcher DR, Dickinson JA. Flexible sigmoidoscopy screening for colorectal cancer in average-risk subjects: a community-based pilot project. Med J Aust 1996; 165: 74-76.
7. Waye J, Kahn O, Auerbach M. Complications of colonoscopy and flexible sigmoidoscopy. Gastrointest Endosc Clin N Am 1996; 6: 343-377.
8. Thiis-Evensen E, Hoff GS, Sauar J, et al. Flexible sigmoidoscopy or colonoscopy as a screening modality for colorectal adenomas in older age groups? Findings in a cohort of normal population aged 63 to 72 years. Gut 1999; 45: 834-839.
9. Freeman B, Engel JJ, Fine MS, DiVita DP. Colonoscopy to the cecum: How often do we get there? Experience in a community hospital. Am J Gastroenterol 1993; 88: 789.
10. Fenlon HM, Nunes DP, Schroy P, et al. A comparison of virtual and conventional colonoscopy for the detection of colorectal polyps. N Engl J Med 1999; 341: 1496-1503.
11. Forbes GM, Mendelson RM. Patient acceptance of virtual colonoscopy [letter]. Endoscopy 2000; 32: 274.
12. Debatin JF, Luboldt W, Bauerfeind P. Virtual colonoscopy in 1999: computed tomography or magnetic resonance imaging? Endoscopy 1999; 31: 174-179.
13. Kay CL, Kulling D, Hawes RH, et al. Virtual endoscopy -- comparison with colonoscopy in the detection of space-occupying lesions of the colon. Endoscopy 2000; 32: 226-232.
14. Rex DK, Vining D, Kopecky KK. An initial experience with screening for colon polyps using spiral CT with and without CT colography (virtual colonoscopy). Gastrointest Endosc 1999; 50: 309-313.

Make a comment