Introduction Patients with large sessile colorectal polyps can be technically challenging to resect endoscopically and have been subject to colorectal resection in the United Kingdom. Our aims were to determine the safety and efficacy of endoscopic resection of large colorectal lesions at a tertiary referral unit.

Methods A prospective observational study of all patients referred for endoscopic resection to a single endoscopist. Consecutive patients were included in the study from June 2010 to March 2013. All patients underwent magnification chromoendoscopy and NBI for polyp assessment under conscious sedation. ESD was undertaken for lesions that were LST – non granular, flat and pseudodepressed type and those with type Vi pit pattern. Piecemeal EMR was undertaken for the remainder of the lesions.

All patients underwent colonoscopic surveillance at 3 and 12 months by the same endoscopist to check for recurrence at the scar.

Results One hundred and fourteen patients underwent 134 endoscopic resections. There were 54 (47.4%) women and 60 (52.6%) men with a mean age of 71.2 (SD = 10.3 years). 120 lesions underwent EMR (89.6%) and 14 had enbloc resection with ESD (10.5%) with complete resection. The mean size of the lesions was 56 mm (SD 37.1mm). The median lesion size was 50mm (range 25–150 mm).

Histological analyses revealed 8 hyperplastic lesions, 28 tubular adenomas, 90 tubulovillous adenomas, 3 serrated adenomas and 5 early submucosal invasive cancers invading to the upper third of the submucosa (sm1). Endoscopic diagnosis of the colorectal polyps using magnification colonoscopy identified all patients with cancer correctly with 100% sensitivity. All lesions underwent endoscopic resection with curative intent. Overall, there were 2 patients who sustained intra-procedural perforation (perforation rate overall 1.8%) of the bowel, both of which were closed with endoscopic clips without the need for surgery.

13 patients were admitted to hospital post procedure (9.8%). 6 patients were for medical reasons (2 perforation, 3 self limiting abdominal pain, 1 patient with pericolic inflammation on CT scan and abdominal pain) and 7 patients were admitted for social reasons.

Median follow up duration was 8.27 months (range 0.39–34.6 months, IQR 12.04 months). 6 patients had documented recurrence (5.1%) with a median time to detected recurrence being 4.45 months (range 2.83–15.74 months, IQR 11.85 months).

Conclusion Endoscopic resection of large colorectal lesions in a tertiary setting is a safe procedure often performed as a day case. Perforations detected during the procedure can be managed endoscopically without the need for surgical intervention. Meticulous technique utilising magnification chromoendoscopy to examine the scar post resection offers a low incidence of recurrence

Disclosure of Interest None Declared.