Introduction Screening colonoscopy has resulted in increasing identification rates of colorectal polyps which can be endoscopically removed. Almost 10% will histologically prove to be pT1 cancers. Factors associated with cancer recurrence post polypectomy have been described (ie lymphovascular invasion, poor differentiation, R1 <1 mm resection margins) and completion surgery is advocated for high risk patients. Since current data are not evidence-based and surgery may implicate significant morbidity, it is of utmost importance to clarify the cost-benefit of each therapeutic option.

Methods Patients with pT1 cancers that had either endotherapy+surgery or endotherapy alone between 2008–2014 were included in the study (n = 61). Endoscopic, histologic and surgery data, including complication rates from both modalities and overall/ cancer-free survival were recorded. Management was compared between two different periods (2008–11, 2012–14) in order to identify changes in practice.

Results 38 (62%) patients were treated with endoscopic resection alone. 23 had a completion surgery, 15 (65%) due to R1 resection. Adenomatous tissue was found in 6 surgical specimens but only one had residual cancer. Major complications occurred in 2 (3%) patients post endoscopy versus 6 (26%) postoperatively (p = 0.044), 2 of whom died. There were no cancer relapses at mean follow up of 41±24 months. 17/31 (55%) patients treated prior to 2012 had surgery compared to 6/30 (20%, p = 0.008) treated in 2012–14. More than one risk factors were apparent in 8 (33%) patients treated with endoscopy only within the last 3 years compared to no one previously (p = 0.017). Out of the overall 38 patients with R1 resection, 14 (74%) had completion surgery between 2008–11 compared to 31% recently treated (p = 0.022).

Conclusion Endoscopic resection of polyp cancers is a safe and successful therapy with minor complications and recurrence rates similar to surgery. Changing our practice recently in favour of endotherapy alone did not lead to increasing cancer relapses, suggesting that presence of risk factors should not be an absolute indication for surgery, but instead an individualised decision should be made based on combination of factors.

Disclosure of Interest None Declared