BOSTON — New data from two studies are putting virtual colonoscopy in a good light compared with optical colonoscopy.
Interim results from a large military study comparing virtual and optical colonoscopy for colorectal cancer screening suggest that the two methods are comparable in terms of sensitivity and specificity, said Maj. Richard P. Moser III, MC, USA.
If final results of the 8-year screening virtual colonoscopy (VC) trial confirm this trend, they will be seen as a validation of the landmark 2003 trial (N. Engl. J. Med. 2003;349:2191–200) that put VC on the map for colorectal cancer screening, suggested Dr. Moser of Walter Reed Army Medical Center in Washington.
Speaking at an international symposium on virtual colonoscopy sponsored by Boston University, Dr. Moser outlined the trial, which includes 3,000 subjects considered to be at average risk in terms of colorectal cancer screening.
The study's primary goals are to validate the 2003 trial, to evaluate the effectiveness and cost-effectiveness of VC screening in routine clinical practice, and to gather data on the short-term natural history of medium-sized (6–9 mm) polyps.
Patients undergoing VC screening are sent to same-day optical colonoscopy (OC) if they have a polyp measuring 10 mm or more, or if they have three polyps measuring at least 6 mm (medium size), Dr. Moser said. Patients with fewer than three medium-sized polyps are randomized to either same-day colonoscopy or 1-year VC follow-up. And patients with no polyps are randomized to either same-day OC or 5-year VC follow-up.
The interim results suggest that for polyps measuring at least 6 mm, VC has a sensitivity of about 90%, compared with about 97% for OC. The specificity of VC was 73%, slightly less than the 80% specificity found in the 2003 trial, indicating some tendency to identify too many polyps, Dr. Moser suggested.
In a separate presentation, Dr. Joel G. Fletcher of the Mayo Clinic in Rochester, Minn., reported that the miss rate for large advanced neoplasia during optical colonoscopy may be “higher than previously anticipated.”
He based his conclusion on comparisons of virtual colonoscopy findings and optical colonoscopy results in a prospective trial of 452 asymptomatic patients who underwent both procedures on the same day.
Dr. Fletcher's team noted 43 lesions measuring 1 cm or larger, including 5 cancers and 21 large adenomatous lesions. Seven of these lesions were missed on optical colonoscopy but were seen on virtual colonoscopy. Four of these lesions were cancers measuring 1.3–3.2 cm. There was also 1 tubulovillous adenoma measuring 3 cm, 1 hyperplastic polyp measuring 1.3 cm, and 1 probable leiomyoma measuring 4.4 cm.
Four patients were referred for a repeat colonoscopy, two were recommended for close interval surveillance, and one was observed for new symptoms.
Dr. Fletcher said that the lesions missed on index optical colonoscopy were not confined to any particular colonic segment. Two were located on the proximal aspect of a fold, three were flat, and four were polypoid.