The finding comes from 15 academic and community medical centers that performed both virtual colonoscopy (CT colonography) and traditional colonoscopy on 2,600 patients aged 50 and older.
The study "validates" new guidelines endorsing virtual colonoscopy as a colon cancer screening option, says study leader C. Daniel Johnson, MD, professor of radiology at the Mayo Clinic in Scottsdale, Ariz.
"Up to now, there has been conflicting data on how good this test would be," Johnson tells WebMD. "This is the largest study to date on CT colonography. The performance data shows the sensitivity is very similar to that quoted for colonoscopy."
The study showed that virtual colonoscopy was able to detect 90% of polyps 10 millimeters or more in diameter. That's the same accuracy reported for colonoscopy itself in other studies.
Colonoscopy is, however, able to detect much smaller polyps. As the size of polyps gets smaller, so does the ability of virtual colonoscopy to detect them.
Johnson and colleagues found that CT colonography could detect 78% of polyps as small as 6 millimeters in diameter. Yet the scans caught only 65% of polyps 5 millimeters or smaller.
"It is not clear that leaving small polyps is safe," notes a news release from the American Gastroenterological Association.
Johnson and colleagues note that fewer than 2% of polyps smaller than 5 millimeters have features that suggest they are precancerous.
Virtual Colonoscopy, Real Bowel Prep
A downside of virtual colonoscopy is that it requires the same bowel prep (cleansing of the bowels) as a real colonoscopy.
Patients' biggest complaint about colonoscopy is the prep, not the procedure itself, says internal medicine expert Robert Fletcher, MD, professor emeritus at Harvard University.
"Nobody ever said the prep was easy, and it is just the same for the CT colonography as it is for colonoscopy," Fletcher tells WebMD. "And if the scan finds something in the colon, which it does about 17% of the time, you have to turn right around and get prepped again and have a colonoscopy, unless it can be arranged the same day."
Johnson says that centers performing CT colonography should be equipped to offer same-day colonoscopy for patients who need it. Fletcher says it's important for patients to ask about this before choosing virtual colonoscopy as a screening option.
Virtual colonoscopy has a 14% false-positive rate. That's not bad, considering it's a screening test designed to catch as many polyps as possible. But patients who get a false alarm will have to undergo an unnecessary colonoscopy.
Of greater concern to Fletcher is the fact that CT scans see more of the body than just the bowel. For 16% of patients in the Johnson study, the scans detected non-bowel problems serious enough for further tests or urgent care.
"It's not trivial for your doctor to tell you, 'I found something in your kidney or on your adrenal gland,'" Fletcher says. "The odds of helping people are pretty low for those things outside the colon. And then there is the worry, the bother, of additional tests and of finding additional bad things you never knew about but cannot be changed now that you know."
Johnson has a completely different view of the issue.
"Colonoscopy cannot look outside the colon. CT colonography does," he says. "So there is an opportunity among patients age 50 and older to be able to look for other abnormalities that may be asymptomatic but may cause harm in the future. We can look for abdominal aortic aneurysms and tumors in other body parts."
Is Virtual Colonoscopy for You?
This year, the American Cancer Society, the American College of Radiology, and the U.S. Multi-Society Task Force on Colorectal Cancer added CT colonography to the list of colon cancer screening options.
The American Gastroenterological Association notes that only half of people get recommended colon cancer screening, and that virtual colonoscopy could improve this sorry statistic.
But GI doctors aren't yet ready to embrace CT colonography, says David A. Margolin, MD, director of colorectal surgery research at the Ochsner Clinic Foundation in New Orleans.
"CT colonography will, in the future, play a role in colon cancer screening," Margolin tells WebMD. "At centers of excellence, at a very few centers where the quality of software and the CT engineer and excellent radiologist training all converge, it is useful. But if you expect these kinds of results at every community hospital, you won't get it."
Johnson admits that radiologists require specialized training and experience before they're ready to perform screening CT colonography. But he says that kind of training is readily available.
And Johnson notes that the study purposely included community hospitals and a wide range of different CT equipment and software to simulate what might happen when virtual colonoscopy becomes more common.
"We spend a lot of time training and testing our CT colonography readers. It does require a special skill set," he says. "These are good questions for patients to ask their radiologist: How many CT colonographies have you read? How much training have you had? Radiologists who avail themselves of this training can get the same results we report."
The Johnson study -- and Fletcher's editorial comments -- appear in the Sept. 18 issue of TheNew England Journal of Medicine.