Reliability of 2-D virtual colonoscopy under debate

Two-dimensional computed tomography colonoscopy, the most commonly used form of "virtual colonoscopy," is dramatically less reliable than previously believed, according to a recent study that included researchers from UT Southwestern.

Dr. J. Steven Burdick, associate professor of internal medicine and head of endoscopy, and Dr. Cecelia Brewington, associate professor of radiology, and their colleagues found that the accuracy of the procedure that provides doctors with a computer-generated view of the colon varies considerably.

The study of 615 patients at nine facilities, including UT Southwestern, was in the April 14 issue of The Journal of the American Medical Association.

There are two predominant methods for computed tomography (CT) colonoscopy, also called "virtual colonoscopy." One uses a 2-D method in multiple planes. The other is a virtual 3-D "fly-through," which provides a view similar to what the endoscopist would see when conducting a conventional colonoscopy. The 2-D technique was utilized in the JAMA study, with questionable areas then analyzed by the 3-D method.

"The method used for the study was predominately a 2-D review, which does have a learning curve," said Dr. Brewington. "The JAMA study does have a degree of internal bias, as further review of the methodology will demonstrate. I do feel, however, that the 3-D fly-through method can provide greater sensitivity and specificity, and that is how we have performed the exam at UT Southwestern since 1998."

UT Southwestern has been utilizing the latest 3-D scanning technology, available in only about 50 medical centers nationwide, along with the 2D method for virtual colonoscopy.

All participants in the study had a virtual colonoscopy then a traditional colonoscopy the same day. The virtual method detected 55 percent of patients with at least one suspicious polyp at least 10 millimeters in diameter, compared with a 100 percent success rate for traditional colonoscopy. For smaller tumors (at least 6 millimeters in diameter), only 39 percent were detected through virtual colonoscopy, compared to 99 percent for the traditional method.

"This kind of virtual colonoscopy has not in large-scale series been able to replicate the 90 percent detection of 1-centimeter polyps," said Dr. Burdick. "Factors including oral contrast, software and physician training may be involved. But this is important, as missing 50 percent of tumors could be compared to a screening mammography on only one breast."

In a conventional colonoscopy, a long, flexible viewing tube is inserted in the rectum and threaded into the colon. A device on the end of the tube enables doctors to remove suspicious growths that are later tested for cancer.

Virtual colonoscopy is designed to take some of the discomfort out of the exam. A narrower rectal catheter is inserted, and a CT scanner produces images of the colon.

Some physicians believe that patients are frequently less fearful of the virtual colonoscopy and more apt to undergo screening. Many UT Southwestern patients have put off screening for years until being made aware of this option, said Dr. Brewington.

"We should not lose focus of the real issue, which is getting patients screened for colon cancer," said Dr. Burdick. "Only 40 percent of patients currently get screened, and methods such as virtual colonoscopy that encourage people who would not get screened otherwise may be helpful. It is hoped that advances in virtual colonoscopy will improve the detection rate."

He noted that variables such as the radiologist's training which may affect test results need to be examined in further studies. Dr. Burdick also said the technology in this area is advancing rapidly and may improve the accuracy of the tests.


Scott Maier