CT colonography lacks the sensitivity of conventional colonoscopy

A new study published in JAMA on CT colonography or 'virtual' colonoscopy using 2-D imaging finds that this test lacks the sensitivity and specificity of conventional colonoscopy for colorectal cancer screening. Study co-author and American College of Gastroenterology President Douglas K. Rex, M.D. is available for interviews.

Findings of a new multicenter study by Cotton et al. published today in the Journal of the American Medical Association comparing standard colonoscopy with CT colonography for the detection of colorectal cancer reveal that this technology, in the form used most often in the United States, while of significant interest, is not presently a viable option for routine colorectal cancer screening. The study utilized abdominal-pelvic CT scanning and the radiologists relied primarily on 2-dimensional images but 3-dimensional "fly-throughs" of the colon that simulate conventional colonoscopy were also evaluated

The study found that the sensitivity of CT colonography for detecting patients with 1 or more lesions sized at least 6 mm was 39 percent, and for lesions sized at least 10 mm it was 55 percent, this compared with 99 percent and 100 percent respectively when examined using conventional colonoscopy also referred to as complete colonoscopy. Only marginal improvements occurred when the 3-dimensional results were evaluated. These findings contrast sharply with findings by Pickhardt et al. recently published in the New England Journal of Medicine in December 2003.

One of the new study's co-authors, Douglas K. Rex, M.D., FACG, is currently president of the American College of Gastroenterology and director of endoscopy at Indiana University Hospital in Indianapolis, Ind. Dr. Rex is available to discuss the new findings and offer his perspective on how the methodology used by Cotton et al. resulted in findings differing from those by Pickhardt et al., as well as the state of the art of conventional colonoscopy and potential future applicability of CT colonography.

Dr. Rex commented, "This is the third largest study in the literature. Other studies using 2-D imaging had equally poor results. Given these disparate results, we need to see verification of the results obtained in the Pickhardt study using new methods before CT colonography can be considered as appropriate for colorectal cancer screening."

New colorectal cancer screening strategies, including virtual colonoscopy and fecal DNA testing, generally will not match the reliability of complete colonoscopy although they may offer the potential to enhance the acceptability of colorectal cancer screening to some persons who are not candidates for complete colonoscopy. Virtual colonoscopy has not yet been endorsed by any multidisciplinary guideline group as an effective colorectal cancer screening strategy. The principle obstacle to endorsement thus far has been the wide range and frequent lack of effectiveness in clinical trials.