NEWS

By DAVID WILLIAMSON
UNC News Services

CHAPEL HILL – "Virtual colonoscopy," the promising new technology that might one day offer a safer, non-invasive way of screening for colon cancer, is not ready for routine use in clinics, a new national study concludes.

Researchers at nine major hospitals led by Dr. Peter B. Cotton of the Medical University of South Carolina in Charleston found that conventional colonoscopy remained significantly more accurate than virtual colonoscopy in detecting potentially cancerous lesions.

Cotton and colleagues found that the experimental method, also known as computed tomographic colonography, was only 39 percent as effective as standard colonoscopies in revealing lesions of at least 6 millimeters in diameter and only 55 percent as effective in showing those measuring at least 10 millimeters.

A report on the findings appears in the April 14 issue of the Journal of the American Medical Association.

"Computed tomographic colonography…is not yet ready for widespread clinical application," the authors wrote. "Techniques and training need to be improved."

In an accompanying editorial, Dr. David F. Ransohoff, professor of medicine at the University of North Carolina at Chapel Hill, agreed that the test’s ability to detect colonic lesions needed to be improved and wrote that technical issues affecting virtual colonoscopy’s sensitivity likely will be identified and addressed over time. The result will be at some point "a clearly useful option for colorectal cancer screening."

"If virtual colonoscopy technology can evolve to the point where satisfactory images can be obtained without patients having to undergo laxative bowel cleansing (one of the most unpleasant parts of virtual colonoscopy and of standard colonoscopy), its potential use may be even greater," he wrote.

In their study, Cotton and colleagues found 827 lesions such as polyps in 308 of the 600 patients who underwent CT and then standard colonoscopy immediately afterward. In 104 participants at the nine centers, doctors identified lesions of at least six millimeters. CT colonography missed two of eight active cancers.

"The accuracy of CTC varied considerably between centers and did not improve as the study progressed," the authors wrote. "Participants expressed no clear preference for either technique."

All patients were age 50 or older.

An earlier study using more advanced techniques showed the new method could reveal polyps of a centimeter in diameter in about 95 percent of cases. That finding prompted some clinicians to believe it could soon be considered for colon cancer screening. As a result, the new work is a disappointment, Cotton and colleagues said.

Ransohoff, a member of UNC’s Lineberger Comprehensive Cancer Center, said results of the previous study revealed what the advanced technology, including bowel preparation, imaging software, method of interpretation and training, can achieve in ideal circumstances. The current federally-funded study demonstrates what the technology likely would achieve if implemented in community practice, which is less rigorous. Reasons for the differing success remain unclear.

"If sensitivity is actually as low as the 55 percent reported …, virtual colonoscopy could not be recommended as an infrequent stand-alone test but would need to be performed repeatedly over time or in combination with other tests," he wrote.

Among key issues needing to be addressed is clarifying the best "target" of colorectal cancer screening, Ransohoff said. Can some lesions not be removed but rather followed over time since unlike conventional colonoscopy, the new CT version does not allow for their immediate removal?

Large polyps are found in only 1 percent to 2 percent of people undergoing the standard tests, while small ones can be found in 30 or more percent of patients screened. Most polyps do not progress to life-threatening colorectal cancer.

"The level of sensitivity and specificity that virtual colonoscopy can achieve, in some specialized situations, is known," he wrote. "Yet the differences between what (it) can do and what it will do if applied in ordinary practice circumstances are so great that physicians must be cautious. There are many important steps to be taken in learning how to implement this new technology appropriately."

Besides the Medical College of South Carolina, other centers involved were Wake Forest University School of Medicine, Emory University Hospital, Indiana University Hospital, the University of Texas Southwestern in Dallas, Virginia Commonwealth University Medical Center, St. Mary’s Hospital in London, M.D. Anderson Cancer Center in Houston and University Hospitals of Cleveland.

The Office of Naval Research supported the multicenter comparative study.

- 30 -

Note: Ransohoff can be reached at (919) 966-1256 or via e-mail at ransohof@med.unc.edu. To contact Cotton, call Ellen Bank, (843) 792-2626.

UNC Media Contact: David Williamson, (919) 962-8596.