The Accuracy of Human Endoscopic Detection of Submucosal Invasive Cancer in Colorectal Polyps

Colorectal cancer (CRC) is a leading cause of death in the Western world. It can be effectively prevented by removal of pre-malignant polyps (polypectomy) during colonoscopy. Large (≥20mm) non-pedunculated colorectal polyps (LNPCPs) represent 2-3% of colorectal polyps, and require special attention prior to treatment. If submucosal invasion (SMI) is suspected careful decision making is required to exclude features which unacceptably increase the risk of lymph node metastases and render local treatment (endoscopic) non-curative. Such patients require a multi-disciplinary approach and consideration of surgery +/- systemic therapy. Recently the endoscopic imaging characteristics which precisely determine the risk of SMI within colon polyps have been elucidated. This suggests endoscopic imaging may be the ideal investigation to stratify the presence and extent of SMI within LNPCP, particularly as it can be applied in real-time at the time of planned endoscopic treatment. Unfortunately, current classification systems are complex, require extensive training and technology not available in the majority of non-tertiary hospitals. They are therefore underused leading to incorrect decision making and negative patient outcome (e.g piecemeal resection without the chance of endoscopic cure or unnecessary further procedures in referral centres with resultant surgery anyway or surgery for benign disease) A simple clinical support tool was created, based on well-established parameters (i.e., presence of a demarcated area within a polyp, size of the polyp, Paris classification, location within the colon and granularity) to identify OVERT (visible on the surface) and COVERT (hidden) submucosal invasion (SMI) within LNPCPs. Crucially this tool only uses what is reproducible in the majority of endoscopy units in the Western world (i.e. standard magnification, no extra chromic dyes etc). predict SMI within LNPCPs and we translated it into a single web-based clinical support tool that