ESD ALLOWS BETTER STRATIFICATION OF METASTATIC RISK IN NON-PEDUNCULATED PT1 COLORECTAL CANCER: A SINGLE CENTRE RETROSPECTIVE STUDY

Aims To compare the effectiveness of endoscopic submucosal dissection (ESD) and endoscopic mucosal resection (EMR) in T1 colorectal cancer (CRC) treatment.

Methods We retrospectively analyzed 267 pT1 colorectal cancers consecutively removed from January 2011 to December 2018 by EMR or polypectomy (n 235, 88%) and ESD (n 32, 12%) in our center.

Results Mean tumors size were 21,1±14,4 mm; 29,9±16,5 mm in ESD group, 19,8±13,6 mm in EMR/polypectomy group. 198/267 lesions (74%) were considered to have a high metastatic risk after histological analysis, one third of which (n 65) due to incomplete or inadequate resection. EMR/polypectomy obtained en bloc resection in 175/235 cases (74%), ESD in 32/32 (100%) tumors. Pedunculated and subpedunculated morphology was independently related to complete resection (OR 5.09, p 0.002) and correct removal for risk stratification (OR 3.19, p 0.027). In multiple logistic regression analysis, compared to EMR/polypectomy, ESD was an independent predictor of radical excision (OR 8.8, p < 0.004) and correct removal that allowed the histologycal stratification of the metastatic risk (OR 4.2, p < 0.001). In non-pedunculated tumors (n 173), univariate comparison confirm ESD was superior to EMR/polypectomy in obtaining complete resection (OR 3.5, p 0.002) and correct removal (OR 9.8, p 0.002) regardless of lesion characteristics such as dimension, morphology and site. Moreover, ESD was significally associated to correct removal (adequate histological evaluation) in non-pedunculated T1 compared to en bloc EMR (OR 2.54, p 0.025).

Conclusions T1 colorectal cancers need often additional surgery for incomplete resection or incorrect removal that doesn’t allow an adequate stratification of metastatic risk. In our retrospective study ESD obtained significantly better results than EMR/polipectomy (even en bloc) in reaching an appropriate removal, expecially in non-peduncolated T1 CRC, and should be preferred in any non-pedunculated lesion with high suspicious of submucosal invasion.