Introduction Endoscopic mucosal resection (EMR) can successfully treat large polyps endoscopically, however adequate information regarding margins of excision, or in the malignant cases, submucosal invasion are crucial for deciding future management. En bloc resection with clear margins allows confident further management but compromised margins or unexpected malignancy in excised polyps creates prognostic and management uncertainty. Therefore it is crucial to accurately assess for malignancy and feasibility of en-bloc excision prior to commencing EMR.

Method A retrospective case note review of all large (≥20 mm) colonic polyps identified on colonoscopy between the study period August 2013 to August 2014. We collected data on lesion characteristics, procedural, clinical and histological outcomes.

Results 92 Patients were identified as having a large colonic polyp, mean age 64.5 years (range 19–90). Mean polyp size was 29.0 mm (range 20–100). During the study period there was one adverse event of EMR which was a caecal perforation treated with an emergency right hemi-colectomy.

18 cases of invasive adenocarcinoma were identified (19.5%). 12 were correctly recognised and had biopsy only at endoscopic procedure (EMR not attempted). Malignancy was found in 38% (13/31) of rectal polyps, 11.7% (4/34) and 75% (3/4) of polyps identified as Pit Pattern IV and V respectively. 6 patients had piecemeal EMR of malignancies resulting in inadequate histology to predict prognosis. 3 of these patients underwent subsequent surgical resection, no residual malignancy was found in these 3 cases suggesting en bloc resection may have prevented major surgery.

Conclusion EMR can lead to inadequate tissue sampling for diagnosis and treatment of invasive carcinoma. We propose that EMR should not be attempted in rectal polyps ≥20 mm without specialist assessment.

Disclosure of interest None Declared.