Adenoma number and size both predict later advanced neoplasia on colonoscopy surveillance

Design

Retrospective cohort with inverse probability of censoring weighting and Cox regression

Population

16,870 adults aged 50 to 75 years with at least two colonoscopies (2004 to 2023); median follow-up 5.3 years

Primary outcome

Advanced colorectal neoplasia by number and size of non-advanced adenomas

Effect

5-year incidence 1.3%, 4.0%, 5.3% and 10.1% for 0, 1 to 2, 3 to 4, 5 to 10 adenomas; HR 5.21 (95% CI 2.89 to 9.38) for 5 to 10

This cohort followed 16,870 adults aged 50 to 75 years with at least two colonoscopies between 2004 and 2023. People were grouped by the number of non-advanced adenomas (none, 1 to 2, 3 to 4, 5 to 10) and by size (diminutive up to 5 mm vs small 6 to 9 mm). Advanced colorectal neoplasia was defined as an adenoma of 10 mm or more, tubulovillous or villous histology, high-grade dysplasia or cancer. Median follow-up was 5.3 years.

Five-year cumulative incidence of advanced neoplasia was 1.3% with no adenomas, 4.0% with 1 to 2, 5.3% with 3 to 4 and 10.1% with 5 to 10. Hazard ratios against no adenomas were 1.64, 2.75 and 5.21. At the same count, having at least one small adenoma carried more risk than diminutive adenomas only: 5.5% vs 3.7% with 1 to 2 adenomas, and 8.1% vs 3.0% with 3 to 4.

This was an observational cohort of people who returned for colonoscopy, and quality of the baseline examination affects results. The authors suggest count and size should both inform surveillance intervals, and current guidelines have not changed on this basis. For Indian practice, the finding supports recording the number and size of adenomas clearly in colonoscopy reports.

• Record adenoma number, size and histology in every colonoscopy report.
• Follow your local guideline for surveillance intervals; this study does not change it.
• Consider that three or more adenomas, or a 6 to 9 mm adenoma, may carry higher risk when deciding on follow-up.
• Ensure a complete, well-prepared examination, since missed lesions distort risk.

Why it matters

It suggests the common 1 to 2 versus 3 or more split hides a gradient and a role for size.

Don't overread it

This is an observational cohort of people who had repeat colonoscopies; it does not change guideline surveillance intervals.

The statistics, in plain English

A five-year incidence of 10.1% in the 5 to 10 adenoma group means about 1 in 10 developed advanced neoplasia, against about 1 in 80 with no adenomas. The 5 to 10 group had only 197 people, so the interval is wide (5.8% to 17.1%).

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.