Risk Stratification Model for Surveillance Interval Extension After Hot Endoscopic Mucosal Resection of Large Non-Pedunculated Colon Polyps

Le résumé fourni par la source

BACKGROUND: Current guidelines recommend uniform 6-month surveillance after endoscopic mucosal resection (EMR) of large non-pedunculated colon polyps (LNPCPs). We aimed to develop and validate a risk stratification model to identify lesions at lower recurrence risk. METHODS: Retrospective multicenter study utilizing prospectively maintained databases across 12 centers (2017-2024) evaluated LNPCPs ≥20 mm removed by hot EMR. Twenty variables underwent multivariate Cox regression with stepwise selection to identify recurrence predictors. Validation was performed in a separate single-center cohort. RESULTS: The derivation cohort included 463 patients with 491 LNPCPs. Multivariate analysis identified four independent predictors: evidence of prior resection (HR 1.97, 95% CI 1.04-3.72, p=0.03), absence of margin ablation (HR 1.96, 95% CI 1.22-3.23, p=0.01), villous histology (HR 1.77, 95% CI 1.09-2.86, p=0.02), and ileocecal valve (ICV) involvement (HR 2.66, 95% CI 1.29-5.45, p=0.01). Validation in a separate cohort (169 patients, 187 polyps) demonstrated effective risk stratification performance (p<0.01). In the validation cohort, low-risk polyps (no prior resection, margin ablation performed, no villous histology, no ICV involvement) had recurrence risks of 2.2%, 5.4%, and 9.7% at 6, 12, and 18 months respectively, compared to 10.6%, 23.1% and 28.1% in high-risk polyps. No patients in either cohort developed high-grade dysplasia or invasive adenocarcinoma during follow-up. CONCLUSIONS: A simple four-factor model incorporating prior resection, margin ablation, villous histology, and ICV involvement stratifies recurrence detection risk after hot EMR. Absence of high-risk features may help identify patients at lower recurrence risk and provides a framework for future prospective studies evaluating risk-adapted surveillance strategies.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.