910 Comparison of Safety and Efficacy Between Endoscopic Mucosal Resection and Endoscopic Submucosal Dissection in the Treatment of Superficial Esophageal Cancers - A Meta-Analysis Update

Abstract

INTRODUCTION: From radical esophagectomy, we have advanced towards endoscopic procedures for the treatment of esophageal cancers without lymph node involvement, superficial esophageal cancers (SEC). Endoscopic mucosal resection (EMR) is a widely accepted treatment modality for SEC due to its minimal invasive nature, low cost and excellent patient tolerance but has limitations. Endoscopic submucosal dissection (ESD) was developed to overcome these limitations. However, ESD has several disadvantages owing to longer procedural times and higher complication rates. The use of ESD is still being debated in the United States. We did a systematic review and meta-analysis to compare the efficacy and safety of EMR and ESD for the treatment of superficial esophageal cancers. METHODS: Literature search was done from electronic databases including PubMed, Cochrane, and Embase from inception until November 2018. Data from 9 relevant studies were extracted and meta-analysis with Random effects model was done using RevMan 5.3. RESULTS: There were 692 lesions in the EMR group and 468 lesions in the ESD group. Patient characteristics were available for only 3 out of 9 studies which showed a mean age of 65.66 years [Male-86.3%, Female-13.7%] in the EMR group and a mean age of 65.66 years [Male-83%, Female-17%] in the ESD group. Our primary outcomes were En-bloc resection rate (EBR) and curative resection rate (CRR) which were analyzed in all 9 studies. Bleeding and perforation rates were our secondary outcomes. EBR was achieved in 319 out of 692 lesions by EMR and 455 out of 468 lesions by ESD [OR 50.93, 95%CI (23.29, 111.37), P < 0.00001, I2 = 23%], favoring ESD over EMR. The CRR was significantly better with ESD when compared to EMR [OR 10.96, 95%CI (3.78, 31.74), P < 0.0001, I2 = 72%]. There was no significant difference in the occurrence of bleeding [OR 1.48, 95%CI (0.37, 5.87), P = 0.58, I2 = 0%] but the rate of perforation was higher with ESD [OR 0.40, 95%CI (0.18, 0.91), P = 0.03, I2 = 6%] CONCLUSION: Our Study shows that the EBR and CRR are better achieved with ESD than EMR. Although there was no significant difference in bleeding, the risk of perforation was higher in ESD. ESD is a complex time consuming procedure that requires great technical expertise and is performer dependent. With careful selection of patients and advanced endoscopic skills, ESD might give better overall outcomes than EMR. Non-randomization and usage of data from heterogeneous practice settings are the limitations of our study data.