New Model to Predict Recurrence After Endoscopic Mucosal Resection of Non-pedunculated Colonic Polyps ≥ 20 mm.

Area of Science:

• Gastroenterology
• Colorectal Surgery
• Oncology

Background:

• Polyp recurrence after endoscopic mucosal resection (EMR) for large non-pedunculated colonic polyps (≥20 mm) is a significant clinical challenge.
• Existing prediction models, such as the Sydney EMR recurrence tool (SERT) and the size, morphology, colonic site, and access (SMSA) score, lack validation in diverse U.S. patient populations.
• There is a need for accurate, real-world validated tools to predict recurrence risk and guide surveillance strategies.

Purpose of the Study:

• To evaluate the external validity of the SERT and SMSA models in a U.S. cohort of patients undergoing EMR for large colonic polyps.
• To develop and validate a novel, simpler model for predicting polyp recurrence after EMR in this specific patient population.
• To compare the performance of the new model against existing tools.

Main Methods:

• A retrospective cohort study included 461 patients with non-pedunculated colonic polyps (≥20 mm) treated with EMR between January 2012 and June 2020.
• Univariate and multivariate analyses identified predictors of polyp recurrence.
• Receiver Operating Characteristic (ROC) curves were generated to compare the predictive accuracy (Area Under the ROC curve - AUC) of the new model, SERT, and a modified SMSA (mSMSA).

Main Results:

• The study analyzed 461 polyps (average size 29.1 mm) with a 29.0% recurrence rate at a median follow-up of 15.6 months.
• A new predictive model was developed incorporating four key variables: polyp size >40 mm, tubulovillous adenoma histology, right colon location, and piecemeal resection.
• The new model demonstrated superior predictive performance (AUC 0.618) compared to SERT (AUC 0.538) and mSMSA (AUC 0.550).

Conclusions:

• The SERT score and mSMSA exhibit poor external validity for predicting recurrence after EMR of large colonic polyps in a U.S. cohort.
• The newly developed model is simpler and shows improved performance in predicting polyp recurrence.
• This new model offers a more effective tool for risk stratification and personalized surveillance strategies in a multiethnic, non-referral U.S. population.