Advantages of Endoscopic Submucosal Dissection Over Conventional Endoscopic Mucosal Resection for Intramucosal Gastric Neoplasms

Abstract

Background: Endoscopic mucosal resection (EMR) has been widely used to treat intramucosal gastric neoplasms in Japan. Conventional EMR has predominantly been performed using strip biopsy, but local recurrence sometimes occurs due to such piecemeal resection. Endoscopic submucosal dissection (ESD) for achieving reliable en bloc resection has recently been developed in Japan, using new devices such as an insulation-tip diathermic knif. Conversely, a high frequency of complications such as severe bleeding and perforation has been reported for ESD. Therefore, the efficacy and problems associated with ESD were evaluated by comparison with conventional EMR in the present study. Methods: Treatment comprised conventional EMR for 48 lesions from January 1999 to October 2002, and ESD for 157 lesions from November 2002 to November 2006 at Nagoya City University Hospital and Iwata Municipal Hospital. Informed consent was provided by all patients. ESD was performed using an insulation-tip diathermic knife and flex and hook knives, as appropriate. Results: Histological diagnosis was adenoma in 16 lesions and adenocarcinoma in 32 lesions for the EMR group, and adenoma in 50 lesions and adenocarcinoma in 107 lesions for the ESD group. No significant differences were noted between the EMR and ESD groups for gender ratio, mean age, histological diagnosis, median tumor size or tumor location. For lesions ≧11 mm in size, en bloc resection rates (EMR:ESD, 16.2%:92.9%) and median treatment time (EMR:ESD, 25 min:91 min) were significantly higher and recurrence rates rates (EMR:ESD, 40.5%:2.3%) were significantly lower with ESD than with EMR (p < 0.001). En bloc resection rates were higher with ESD than with conventional EMR in any area of the stomach (p < 0.005). Median treatment time was significantly longer with ESD than with conventional EMR in all areas (p < 0.001). No significant differences in frequency of complications were noted between groups according to tumor size and tumor location. In the ESD group, no significant differences were noted for en bloc resection rates and frequency of complications according to tumor size and tumor location, but median treatment time was significantly longer for lesions ≧31 mm in size or in the upper third of the stomach. However, treatment time with ESD could be shortened by improvements in skill following experience. Conclusions: ESD can take a long time, but is superior to conventional endoscopic mucosal resection for treating intramucosal gastric neoplasms. These results show that ESD may remove size limitations from EMR for treating intramucosal gastric neoplasms.