More, and larger, non-advanced adenomas meant more advanced neoplasia at five years

Design

Retrospective cohort with repeat colonoscopy, Cox regression with censoring weights

Population

16,870 adults aged 50–75 with at least two colonoscopies

Primary outcome

Five-year advanced colorectal neoplasia

Effect

1.3%, 4.0%, 5.3%, 10.1% by 0, 1–2, 3–4, 5–10 adenomas; 8.1% vs 3.0% with a 6–9 mm lesion (3–4 adenomas)

A cohort of 16,870 adults aged 50 to 75 who had at least two colonoscopies between 2004 and 2023 was grouped by the number of non-advanced adenomas at the first, and by whether any were 6–9 mm rather than all 5 mm or smaller. Median follow-up was 5.3 years. It was published in May.

Five-year advanced colorectal neoplasia was 1.3% with no adenomas, 4.0% with one or two, 5.3% with three or four and 10.1% with five to ten (hazard ratios 1.64, 2.75 and 5.21). Within count groups, having at least one 6–9 mm adenoma raised risk: 5.5% against 3.7% for one or two, and 8.1% against 3.0% for three or four.

Most surveillance guidance stratifies non-advanced adenomas by count. This study suggests size within the non-advanced range adds information, and that three or four adenomas carry more risk than one or two. It is observational, from a single health system, and a guideline change will need more data, but it supports shorter intervals for people with several adenomas or any 6–9 mm lesion within current guideline ranges.

• Record the size of each non-advanced adenoma, not just the number, in colonoscopy reports.
• Consider the shorter end of guideline surveillance intervals when any adenoma is 6–9 mm.
• Treat three or four non-advanced adenomas as higher risk than one or two.
• Keep surveillance within current guideline ranges until guidelines are updated.

Why it matters

It challenges count-only rules for setting the next colonoscopy.

Don't overread it

This was observational, from one health system; it does not by itself justify changing guideline intervals.

The statistics, in plain English

A hazard ratio of 5.21 means people with five to ten adenomas developed advanced neoplasia about five times as fast as those with none. The interval for that group is wide (2.89 to 9.38) because only 197 people were in it.

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.