Five-Year Colonoscopy After High-Risk Adenomas Meets Interim Noninferiority

Key Takeaways

• Among patients with high-risk adenomas in eight European countries, five-year colorectal cancer incidence was similar whether first surveillance was assigned at five or three years.
• The five-year incidence comparison met the prespecified interim noninferiority criterion; the ten-year primary outcome remains pending.
• Cancer stage at diagnosis did not appear substantially different between surveillance groups.
• Few colorectal cancer deaths were recorded, leaving comparative mortality unestablished.

How long can patients wait for their first surveillance colonoscopy after removal of a high-risk adenoma? In an ongoing randomized colonoscopy-surveillance trial across eight European countries, patients who'd had a high-risk adenoma removed were assigned to a first colonoscopy at five years or to one at three years followed by another at five years. Eligibility was based on adenoma size, high-grade dysplasia, villous growth, or the number of adenomas removed. A prespecified interim analysis assessed cumulative colorectal cancer incidence at five years; cumulative incidence at ten years remains the primary endpoint. Investigators used inverse-probability weighting to account for missing data when participants did not undergo the five-year surveillance colonoscopy.

Five-year cumulative colorectal cancer incidence was 0.77% with first surveillance at five years versus 0.82% with first surveillance at three years, an absolute difference of −0.05 percentage points. Cancer stage at diagnosis did not appear substantially different between groups. Colorectal cancer deaths were uncommon in both groups.

For the interim test, the upper boundary of the one-sided confidence interval for the absolute incidence difference was 0.68 percentage points, below the 0.7-percentage-point noninferiority margin. The criterion was therefore met for cumulative colorectal cancer incidence at five years. The test compares a bound on the possible difference with a prespecified threshold; it does not require identical observed incidence in the two groups.

The interim finding does not resolve the pending ten-year primary outcome or establish equivalent colorectal cancer mortality. Cancer stage distributions were described only qualitatively. Weighting addressed missing five-year surveillance information, but the reported findings do not show how the adjustment affected the estimate. The comparison took place in European countries and does not establish a change in recommended surveillance practice elsewhere.

The investigators concluded that starting surveillance at five rather than three years after high-risk adenoma removal was noninferior for cumulative colorectal cancer incidence at five years.

Clinician Questions

Which adenoma findings qualified patients for the three- versus five-year surveillance comparison?

In the European randomized trial, patients qualified after high-risk adenoma removal if they had at least one adenoma measuring at least 10 mm, at least one with high-grade dysplasia or villous growth, or 3 to 10 adenomas of any kind. These criteria do not encompass all patients after polyp removal.

Does the noninferiority margin for five-year colorectal cancer incidence represent the observed difference between surveillance groups?

No. For patients with high-risk adenomas assigned first surveillance at five versus three years, investigators compared the upper confidence boundary of the absolute between-group difference in five-year colorectal cancer incidence with a prespecified margin. The margin was not the observed difference, a relative-risk threshold, or a test of colorectal cancer mortality.

This content is intended for healthcare professionals for educational purposes. The article and its image were developed with AI assistance and reviewed by members of the ReachMD team prior to publication. Please refer to the original source linked for the complete and most up-to-date information.