CTC surveillance edges optical coloscopy for management of small polyps

Using surveillance of 6 to 9-mm polyps with CT colonography instead of referring them for immediate colonoscopy and polypectomy provides a significant cost-effectiveness edge, according to a study presented Sunday.

Death rates might increase by a small amount with a CTC surveillance strategy, but it would avoid nearly 10,000 colonoscopy referrals and, with them, 10 bowel perforations, possibly including one death, producing an incremental cost-effectiveness ratio of $372,853 per life saved, according to the study led by Dr. Perry J. Pickhardt of the University of Wisconsin-Madison.

The study was among several presented at the session addressing CTC. Two others based on American College of Radiology Imaging Network (ACRIN) 6664 trial data pinpointed which reading styles affect performance and found that polyp size estimates with traditional colonoscopy are consistently greater than those based on CTC.

Pickhardt's study was based on a concentrated hypothetical population of 100,000 60-year-olds with a single 6 to 9-mm polyp. The study included a host of assumptions, but these were collectively conservative in that they tended to overestimate the risk of CTC for surveillance of the small polyps, he said.

The study estimated the five-year risk of death for CTC versus optical colonoscopy in the concentrated cohort as .03 versus .02. In a general screening population, the figures were .0024 for CTC versus .0015 for optical colonoscopy.

Exclusion of large polyps at CTC confers a very low risk for cancer, and aggressive management for the 6 to 9-mm polyps may not be indicated, Pickhardt said. Immediate polypectomy is associated with high costs, additional complications, and low yield, while CTC for polyp identification allows for selective identification of small polyps at risk.

The study examining CTC performance was based on a survey of 15 radiologists who together had read more than 2500 studies for the ACRIN 6664 trial, said Dr. Amy Hara of the Mayo Clinic, who presented the findings.

The study broke them into superior (eight) and average (seven) performers and then looked at the reading characteristics of each. The superior performing group had sensitivity and specificity figures for polyps 10 mm and larger of 96% and 89%, respectively, while the average group had comparable figures of 83% and 82%. For polyps of 5 mm or larger, the sensitivity and specificity figures for the superior group dropped to 84% and 91%, respectively, and for the average group to 71% and 84%.

The superior readers shared other characteristics:

• No preference for primary 2D or 3D reads
• An evaluation of prone and supine images in both directions
• The use of entire monitor to interpret 2D axials
• Slightly longer interpretation times (23.5 minutes per study versus 21.2)

The researchers also concluded that the decision to use 2D or 3D for the primary read does not affect performance nor does the level of experience.

A study at Brown University evaluated polyp size estimates obtained with CTC, optical colonoscopy, and pathology. After adjusting for multiple polyps per patient, the colonoscopy measure was consistently 12% greater than the measure for CTC. The pathology measure was consistently 25% smaller than the one for CTC. The study did not, however, account for shrinkage of the polyp due to dehydration after it was removed from the colon.