Virtual colonoscopy: Is it ready for the real world?

(This article is printed in the new publication from Harvard Health Publications, The Year in Medicine 2004. For more information or to order, please follow this link.)

"Virtual" colonoscopy using CT scanners offers an alternative to the traditional screening test. But is it accurate? While the technology shows promise, relying on it as a screening tool may still be a few years off.

Of all routine cancer screening tests, colonoscopy may be the most onerous. You must stop eating 24 hours beforehand. To further clean out your system, you need to either drink about a half gallon of an unpalatable solution or swallow a lot of pills — and then be prepared to spend a great deal of time on the toilet.

The test itself involves some high-tech spelunking. The doctor peers at the lining of your rectum and colon through a slender endoscope inserted through the rectum, looking for early cancers or adenomatous polyps. Although these polyps are benign, almost all colon cancers arise from them, so finding and removing them is a way to prevent colon cancer.

The procedure takes less than half an hour, although if you include the recovery time it’s 90 minutes or more. And it is safe, although not entirely without risk. About 1 in every 1,000 times, the endoscope pokes a hole in the colon. If that happens, the patient must be hospitalized, and surgery may be necessary. Intravenous painkillers and sedatives go a long way toward taking the edge off the whole experience, but they also mean that you need someone to drive you home afterward.

Flying through the colon

Virtual colonoscopy (more prosaically referred to as CT colonography) doesn’t depend on an endoscope. Instead, a CT scanner that is completely outside the body takes hundreds of wafer-thin “pictures” of the inside of your colon and rectum that a computer then assembles into a complete image. Doctors “fly through” the image on a computer screen, looking for abnormalities. This is colonoscopy as video game rather than cave exploration. There’s no endoscope, so there’s no risk of perforation. Because it’s a lot less invasive, people would presumably be less apprehensive and more likely to have the test.

Researchers started publishing studies of virtual colonoscopy over six years ago, but it’s remained a research project so far. Turf is an issue. Colonoscopy has been the province of gastroenterologists. Virtual colonoscopies are done by radiologists. But the biggest obstacle has been doubts about virtual colonoscopy’s accuracy. It has missed polyps found by conventional colonoscopy — over half of them, in some studies. In others, it has been quite accurate.

Why results vary

Some screening tests are simple to perform and interpret, for example, tests that involve measurements of substances in the blood such as cholesterol. The results are highly reliable. Other screening tests are more complicated and, therefore, less reliable. For example, imaging studies such as x-rays require the expertise of a trained eye. Training takes time. The use of CT (computed tomographic) scanners to take pictures of the colon is very new, and even the doctors with the most experience interpreting the images do not have much experience. This may explain some differences in results.

The study of virtual colonoscopy published in the April 14, 2004, issue of the Journal of the American Medical Association (JAMA), tested the accuracy of results at nine major medical centers. The study followed more than 600 participants over age 50 who were referred for routine clinically indicated colonoscopy. Virtual colonoscopy was performed before a conventional colonoscopy. In this study, the virtual colonoscopies missed two of eight cancers and also detected fewer polyps — some of which were of the type that can become cancerous. The conclusion: CT colonography by these methods is not yet ready for widespread clinical application.

An earlier study, published in the Dec. 4, 2003, New England Journal of Medicine raised hope for virtual colonoscopy. It was a large study (1,200 subjects) of people with an average colorectal cancer risk. A virtual colonoscopy was done first and a conventional one almost immediately afterward. The detection rates (sensitivity) of the two were very close.

The equipment and software used for this study produced vivid, three-dimensional images. Previous studies had used equipment producing largely two-dimensional images. The pictures might also have been clearer because two cups of barium and half a cup of a solution called Gastrografin were added to the bowel preparation.

Virtual colonoscopy still requires inflation of the colon. In conventional colonoscopy, that’s done by the doctor through the endoscope. In this study, a small flexible catheter about the diameter of the pinky finger was inserted about an inch into the rectum. The patient controlled the amount of air pumped into the colon by squeezing a small rubber bulb attached to the catheter.

Hurdles ahead

In light of varying results, researchers in the latest trial believe virtual colonoscopy still needs work. The study published in JAMA cited that techniques and training needed to be improved.

In addition, conventional colonoscopy has one clear advantage compared to virtual colonoscopy, even when the virtual colonoscopy accurately spots cancers and precancerous polyps. With conventional colonoscopy, once a cancer or a polyp is discovered it can be removed right then and there. The doctor snakes a thin, wire-like device through the endoscope and snares the polyp. Virtual colonoscopy, though, is purely diagnostic, so if a polyp is discovered you still need a separate — and it’s still endoscopic — procedure to have it removed.

Whether virtual colonoscopy becomes popular may depend on the rival camps, radiologists and gastroenterologists, cooperating so that if a polyp or cancer is found by virtual colonoscopy, a gastroenterologist is right there, ready to remove it via an endoscope. Another exciting but unproven possibility is so-called subtraction software that lets the CT scanner “see” the colon lining through the contents of the bowel, so all that bowel preparation won’t be needed.

Dr. Perry Pickhardt, the University of Wisconsin researcher who led the 2003 study, believes changing the threshold for removing polyps is also necessary if virtual colonoscopy is to enter the mainstream. With conventional colonoscopy, even small polyps tend to be removed. But Pickhardt believes — and there’s research to support his view — that small polyps pose a very slight cancer risk and that the risks of removing them (perforation and bleeding) outweigh any risk of leaving them alone. With virtual colonoscopy and a shift toward removing only large polyps, over 90% of asymptomatic adults could be “cleared” without any further examination, he says. People with intermediate-sized polyps (6–9 millimeters) would get another virtual colonoscopy in two to three years.

Surprisingly, about 40% of the people in Pickhardt’s study said that if they were to have another colonoscopy, they’d choose a conventional one. The medication people get for conventional colonoscopy apparently offsets the discomfort. Pickhardt says virtual colonoscopy patients could be offered a mild, oral sedative.

Conventional colonoscopy has the most data to support it. So, your doctor will likely recommend it over a virtual colonoscopy. It’s best to talk with your doctor about your individual risk factors, and when you should be screened. Although colorectal cancer can occur at any age, most people who develop it are over the age of 50.