Introduction Little is known about the risk of stenosis and outcomes following endoscopic resection of colorectal lesions which leave extensive mucosal defects. A limited number of studies suggest significant stenosis rates, although reported outcomes and suggested management are conflicting. We determined the risk of stenosis and outcomes of endoscopic resection leaving mucosal defects≥75% of the circumference.

Method Patients who underwent endoscopic resection using endoscopic mucosal resection (EMR), endoscopic submucosal dissection (ESD) and hybrid techniques of colorectal lesions≥2 cm were included. Patients were grouped according to circumferential extent of the mucosal defect. Surveillance colonoscopy was performed at 3 and 12 months. Clinicopathological characteristics and outcomes were compared between groups.

Results 435 lesions≥2 cm were resected using EMR (n=342), ESD (n=45) or hybrid techniques (n=48). Circumferential extent of the mucosal defect was ≥75% in 41 patients. 8 were fully circumferential: 1 caecal lesion and the rest in the recto-sigmoid and rectum. 3 circumferential lesions contained deeply invasive adenocarcinoma and 1 benign lesion ultimately required surgery. 41 lesions with a mucosal defect ≥75% of the circumference had a mean size of 100.5 mm vs 49.0 mm for other lesions (p<0.001). These patients had significantly more complications (16.7% vs 4.7%, p<0.001), including a higher rate of perforation (8.3% vs 2.3%, p=0.02), although none required surgery, and a significantly higher rate of recurrence (44.8% vs 9.2%, p<0.001). 79% of patients without cancer were free from recurrence and had avoided surgery at last follow up compared to 97% with mucosal defects<75% (p<0.001). Stenosis occurred in 7 patients: 4 lesions extensively involving the rectum and recto-sigmoid and 2 lesions involving the sigmoid extending to the rectosigmoid. 1 of these involved a mucosal defect of only 50% of the circumference and 3 were fully circumferential. 1 patient had a symptomatic anorectal stenosis requiring dilatation under anaesthesia, 1 patient was asymptomatic but underwent early dilatation after the first surveillance endoscopy. The remaining patients were asymptomatic and managed expectantly. In all these latter cases spontaneous improvement in the stricture was noted at the subsequent surveillance colonoscopy.

Conclusion The majority of patients with these extensive complex lesions can successfully be treated with endoscopic resection and avoid surgery. However, these patients have a significantly greater risk of complications and recurrence and should be managed in a tertiary institution. Although there is a significant risk of stenosis, it appears that most cases are asymptomatic and spontaneously improve with expectant management.

Disclosure of Interest None Declared