Back to Menu

New test for bowel cancer - not ready to replace the gold standard
(May 2004)

The enthusiasts have been advertising virtual colonoscopy for the last few years, along with total body scans. There are multiple rationales for their enthusiasm - an interesting new technology, an easier test for patients - and than another big reason, money. The best documented approach to prevention of bowel cancer, one of the most important cancers in the United States (150,000 cases a year with 60,000 deaths) is the colonoscopy. A flexible tube inserted in the rectum examines the entire lower bowel (about five feet) to find and remove benign growths (polyps) that can turn cancerous. In recent years, a two-dimensional CT (computed tomography) scan has been developed called virtual colonoscopy. That is what almost all of the heavily advertised commercial centers have been using. It requires the same cleaning out of the bowel that is used with regular colonoscopy, but there are no tubes; it is all done with external x-rays. The problem has been that, when used for general screening purposes, it missed too many intermediate and large sized polyps (the ones most likely to turn cancerous) and there were too many false positives (what appeared on x-ray to be polyps, but were not). A new interesting study analyzes virtual colonoscopy using three-dimensional CT scans. The results are published in the December 4, 2003 issue of The New England Journal of Medicine.

At several centers, 1,233 adults, average age 58 years, who had no symptoms of bowel disease were given two tests on the same day - three-dimensional CT scan (virtual colonoscopy) and standard colonoscopy; 13.6 percent were found to have significant polyps on virtual colonoscopy. Virtual colonoscopy did not detect about 12 percent of polyps 6 millimeters or greater in size. Polyps that size are considered by many experts as the ones that are most likely to become cancerous. The risk of cancer is considerably greater for polyps of 10 millimeters or larger; virtual colonoscopy missed only 6 percent of polyps that large. The false positivity rate was 20 percent for polyps 6 millimeters and larger. That means that, if 6 millimeters is the size that requires further investigation, 20 percent of subjects would either have to have immediate standard colonoscopy (more likely) or be followed with frequent virtual colonoscopies (less likely) for findings that were of no significance.

One-half the subjects had no polyps, and many of the other polyps were not considered significant (5 millimeters or less). But, almost one-third of the subjects would be considered immediate colonoscopy candidates with polyps of 6 millimeters or over in size (including both true positives and false positives).
The authors conclude “our results show that virtual colonoscopy with a three-dimensional emphasis is an effective tool for the detection of colorectal neoplasia in asymptomatic adults with an average risk of colorectal cancer. These findings support the concept of a colon screening center that offers virtual colonoscopy to patients with the opportunity for same-day or next-day colonoscopy”.

Commentary: Despite the authors’ enthusiasm, there is every reason for caution:

First, the advantage of standard colonoscopy is that polyps can be detected and removed at the same time. In contrast, virtual colonoscopy must be followed with a standard colonoscopy in order to remove potentially dangerous polyps.

Second, in the study, the percentage with polyps was smaller than usually found. If false positives are included, it might well be that 40 percent of those over age 50 undergoing virtual colonoscopy would require immediate followup standard colonoscopy with polyp removal.

Third, there is no consensus as to what polyp size on virtual colonoscopy should be an indication for immediate standard colonoscopy. It could end up being 6 millimeters or 8 millimeters or (less likely) even 10 millimeters.

Fourth, if immediate colonoscopy is not done for polyps of 6 or 8 millimeters in size, there is no standard program for followup repeat virtual colonoscopies. Some will undoubtedly recommend yearly follow-ups and, for some entrepreneurs, monetary considerations will likely be a part of the decision making.

Fifth, if the conclusion of this paper is followed and followup colonoscopy is done the next day, a second bowel preparation will be needed - and a lot of people won’t like that.



We need several additional studies in different settings to allow a judgement about this three-dimensional CT procedure. Effectiveness and costs will have to be carefully evaluated before the test can be recommended for health care policy.

And, what about all these centers and entrepreneurs that now urge you to get virtual colonoscopy for $600 or $900 or $1,000 or more? This study and the other recent studies suggest you should be very cautious about such importunings. Interestingly, if they are going to stay in business, they will have to replace the two-dimensional CT with a three-dimensional CT, if additional studies show the three-dimensional CT is really effective.

Of course, Healthful Life recommends screening for bowel cancer starting at age 45 (most experts start at age 50) with either a left-sided bowel examination with a flexible tube every five years (plus annual stool examination for bleeding) or a full colonoscopy every ten years (see Human Tune-Up and articles on bowel cancer in the Archives).

Pickhardt, P.J., et al. Computed tomographic virtual colonoscopy to screen for colorectal neoplasia in asymptomatic adults. The New England Journal of Medicine. Vol 349 (December 4) Pgs 2121-2200. 2003.

After this article was reviewed by the Scientific Advisory Board, a new article (April 2004) appeared that gives strong support to our caution. We will review that article with its editorial comment in the coming months.