ESD versus EMR in non-ampullary superficial duodenal tumors: a systematic review and meta-analysis.

Endoscopic submucosal dissection (ESD) has been developed as an option for treatment of esophageal, gastric and colorectal lesions. However, there is no consensus on the role of ESD in duodenal tumors. This systematic review and meta-analysis compared ESD and endoscopic mucosal resection (EMR) in sporadic non-ampullary superficial duodenal tumors (NASDTs), including local experience. We conducted a search in PubMed, Scopus and the Cochrane library up to August 2017 to identify studies that compared both techniques reporting at least one main outcome (en-bloc/complete resection, local recurrence). Pooled outcomes were calculated under fixed and random-effect models. Subgroup analyses were conducted. A total of 753 patients presenting with 784 NASDTs (242 ESD, 542 EMR) in 14 studies were included. Tumor size (MD: 5.88, [CI95 %: 2.15, 9.62], = 0.002, I = 79 %) and procedure time (MD: 65.65, [CI95 %: 40.39, 90.92], < 0.00001, I = 88 %) were greater in the ESD group. En-bloc resection rate was significantly higher in Asian studies (OR: 2.16 [CI95 %: 1.15, 4.08], = 0.02, I : 46 %). ESD provided a higher complete resection rate (OR: 1.63 [I95 %: 1.06, 2.50], = 0.03, I : 59 %), but there was no risk difference in the risk of local recurrence (RD: - 0.03 [CI95 %: - 0.07, 0.01], = 0.15, I : 0 %) or delayed bleeding. ESD was associated with an increased number of intraoperative perforations [RD: 0.12 (CI95 %: 0.04, 0.20), = 0.002, I : 56 %] and emergency surgery for delayed perforations. The inclusion of eligible studies was limited to retrospective series with inequalities in comparative groups. Duodenal ESD for NASDTs may achieve higher en-bloc and complete resections at the expense of a greater perforation rate compared to EMR. The impact on local recurrence remains uncertain