Barrett's surveillance finds few of the cancers: 90% presented without prior endoscopy

Design

Nationwide retrospective registry cohort, 2008–2018

Population

8,914 Dutch patients treated for Barrett-related dysplasia or adenocarcinoma

Primary outcome

Proportion with vs without prior surveillance endoscopy

Effect

90% de novo; stage ≥II 70% de novo vs 11% with prior endoscopy

A nationwide Dutch cohort in Gut (21 September) identified all 8,914 patients treated with curative intent for Barrett-related dysplasia or oesophageal adenocarcinoma between 2008 and 2018, and checked pathology records for a prior endoscopy showing non-dysplastic Barrett's one to five years earlier.

Ninety per cent (8,025) had no prior endoscopy — they presented de novo. They were far more often diagnosed at stage II or later (70% vs 11% of those under surveillance). Even among patients suitable for organ-preserving endoscopic treatment, 65% presented de novo. In those under surveillance, missed advanced neoplasia was uncommon (11%).

Surveillance works for the patients in it: they are caught early. But almost all Barrett-related cancer arises in people never known to have Barrett's. The gains now lie in finding them — targeted screening of high-risk groups — rather than intensifying surveillance for those already known.

• Consider screening endoscopy in patients with chronic reflux plus several risk factors: male sex, age over 50, white ethnicity, central obesity, smoking, family history
• Do not dismiss new or persistent dysphagia or weight loss in reflux patients
• Continue guideline-based surveillance for known Barrett's — it detects disease early
• Watch for non-endoscopic screening tools such as swallowable cell-collection devices as evidence develops

Why it matters

It shifts attention from how often to survey known Barrett's to how to find the people who have it and do not know.

Don't overread it

The study shows surveillance misses most cases at population level; it does not show that screening would improve survival.

The statistics, in plain English

The 90% figure has a tight interval (89–91%) because the cohort is national and complete. It describes where cancers came from, not whether screening would reduce deaths — that needs a trial.

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.