Endoscopic Submucosal Dissection Versus Endoscopic Mucosal Resection in Management of Superficial Squamous Esophageal Neoplasms

Abstract

Background: The efficacy and risks of endoscopic submucosal dissection (ESD) versus endoscopic mucosal resection (EMR) in management of early esophageal cancer has not been compared previously. This study aims to evaluate the outcomes and morbidities of both techniques in management superficial squamous esophageal neoplasms. Methods: A retrospective analysis of all patients with high-grade intraepithelial neoplasms (high grade dysplasia and noninvasive carcinoma) or intramucosal invasive carcinoma resected endoscopically between 2002 and 2007 was performed. Clinical and pathologic outcomes were compared. ESD was performed with a combination of knifes including the insulated tip (IT) knife, hook knife and triangular tip knife (KD 610L, KD 620LR, KD 640L Olympus Co. Ltd., Tokyo, Japan). Submucosal lifting was accomplished by injection of a novel hyaluronic acid, indigocarmine, epinephrine and normal saline mixture. EMR was performed using the suction cap-assisted method. The lesion was resected with the use of a snare loop. Results: ESD was performed in 18 lesions (15 patients) and cap-EMR in 14 lesions (11 patients). The mean (S.D.) follow-up time was 16.46 (15.55) months. ESD resected a significantly larger specimen (5.42 +/− 4.49 cm2 vs 0.91 +/−0.50 cm2, P = 0.002), and was associated with a higher enbloc resection rate (94.44% vs 71.43%, P = 0.032) when compared to cap-EMR. More patients with lesions extending for more than half the circumference of the esophagus were resected by ESD than by EMR (50% and 0% respectively, P = 0.002). There were no significant differences between ESD and EMR in the rate of lateral margin involvement (30% and 14.29%, P = 0.412) and local recurrence (0% and 11.11%, P = 0.492). More procedure-related morbidities occurred in ESD (n = 4) than EMR (n = 1), but the difference was statistically insignificant (P = 0.355). Perforation occurred in one ESD case and was managed conservatively. Esophageal strictures occurred in three ESD treated patients, and they were successfully managed with serial endoscopic dilations. There was no procedure related mortality in the entire series. Conclusion: ESD of early squamous esophageal neoplasms is feasible. ESD increases the extent of resection and rate of enbloc removal when compared to EMR. However, the procedure is associated with higher risk of morbidities without any increase in mortality. Further studies are required to assess the long-term efficacy of the technique.