With several studies pointing to the sensitivity of CT colonography for detecting colorectal polyps and cancer, many predict the technique will soon compete effectively with other full structural screening exams such as colonoscopy. Before large clinical trials begin, however, CT colonography protocols must be standardized in a way that maximizes the exam's ability to detect colonic neoplasia.
In a prospective study published in the September issue of Radiology, (Vol. 216: pp. 704-711) Dr. Joel Fletcher, Dr. C. Daniel Johnson and colleagues from the Mayo Clinic in Rochester, MN sought to expand the list of currently accepted techniques in CT colonography by evaluating the benefits of prone positioning in addition to supine positioning, and of adding oral iodinated contrast medium to aid detection of colonic polyps. While CT colonography performed far better when prone images were added to the protocol, the authors found that adding oral contrast medium failed to improve the sensitivity of the exam.
Several CT colonography techniques have gained wide acceptance, including thin collimation and reconstruction intervals able to detect smaller polyps, the authors wrote. Prior studies also found that low radiation dosage levels of 70-100 mA do not diminish effectiveness. And it is generally agreed that the best workstations are those that enable interactive review of 2-D transverse images, 2-D multiplanar reformations, and 3-D renderings in order to maximize their ability to interpret CT data, the authors wrote.
Previous studies have concluded that colonic distension (collapse) and the presence of fluid are the two largest sources of correctable nonperceptual errors, according to the authors. Acquiring prone images can mitigate distension by allowing insufflated air to fill in collapsed segments, while iodinated contrast medium can increase the attenuation of colonic fluid and outline dependent colonic walls. The study sought to assess whether improvements resulting from these steps were justified by the additional cost and time necessary to complete them.
One hundred eighty patients underwent CT colonography and colonoscopy, with both exams on the same day. Patients with known or possible polyps, or a family history of polyps or colon cancer, were included. Patients with inflammatory bowel disease, melena, hematochezia, familial polyposis, and other conditions were excluded. Seventy-three randomly chosen patients received iodinated oral contrast along, with standard bowel preparation the night before CT colonography.
CT colonography was performed with patients in supine and prone positions, after subcutaneous administration of glucagon and carbon dioxide, and adequate inflation checked with CT scout images. Helical scanning was performed on a HiSpeed Advantage scanner (GE Medical Systems) with 5-mm collimation, 70 mA, 120 kVp, 3-mm reconstuction intervals, using three or four 20-second breathholds.
Two radiologists blinded to patient histories and colonoscopy results analyzed the CT data using a zoomed axial technique, 3-D volume renderings and 2-D multiplanar reformations on customized software. They recorded the number, size, and location of colonic polyps, as well as confidence levels of their findings. Data from the 73 patients who received the oral contrast were further examined by measuring attenuation of colonic fluid. All of the results were compared with colonoscopy.
The results showed that out of 121 large (1 cm or greater diameter) and 142 smaller (0.5-0.9 mm diameter) polyps discovered at colonoscopy, prone positioning increased sensitivity for identification of larger polyps from 70% to 85% (p=.004) and from 75% to 88% for polyps 0.5 cm or larger (P<.005), with no change in specificity. Use of the oral contrast medium did not increase the detection rate, even when fluid attenuation was markedly increased, the authors wrote.
Adding prone imaging reduced false-negative studies by 50% for patients with both large and small polyps, the authors wrote. Moreover, the combined imaging helped identify colorectal carcinoma in 3 of 14 patients (21%) who were not identified by evaluation of the supine image data alone.
"Combined prone and supine imaging demonstrated additional polyps by helping overcome perceptual errors in every segment for polyps 0.5-0.9 cm in size and in nearly every segment for polyps 1 cm and larger," the authors stated. Additional polyps detected in the sigmoid colon were usually identified because of better luminal distension, whereas those in the right colon were usually identified when perceptual problems were overcome, they wrote.
"We believe that the results of our prospective blinded trial quantify the benefit of prone positioning for the demonstration of colonic polyps at CT colonography and the lack of benefit from the administration of oral iodinated contrast medium," they concluded.
By Eric BarnesAuntMinnie.com staff writer
September 6, 2000
Let AuntMinnie.com know what you think about this story.
Copyright © 2000 AuntMinnie.com
