Tu1108 FEASIBILITY OF ENDOLOOP CLOSURE AFTER GASTRIC ENDOSCOPIC SUBMUCOSAL DISSECTION TO PREVENT DELAYED BLEEDING IN PATIENTS RECEIVING ANTITHROMBOTIC THERAPY

Abstract

Endoscopic submucosal dissection (ESD) has been established as a standard treatment for early gastric cancer without the possibility of lymph node metastasis. However, adverse events, such as post-ESD bleeding occurs approximately 5%. In particular, antithrombotic therapy increases the rate of post-ESD bleeding to about 21-38%. Therefore, effective methods for prevention of post-ESD bleeding are desirable. The aim of this study was to assess the efficacy of endoloop closure (ELC) using endoloops and clips in preventing post-ESD bleeding among patients receiving antithrombotic therapy. This study was a nonrandomized, retrospective, single-center study of 1634 early gastric cancers among 1163 patients who were treated by ESD from March 2016 to January 2019. Patients who were taking antithrombotic agents were reviewed. Antithrombotic agents were defined as antiplatelet agents (low-dose aspirin, cilostazol, thienopyridine derivatives) or anticoagulants (warfarin, direct oral anticoagulants) . Exclusion criteria were follows; (1) other upper gastrointestinal endoscopic intervention within 28 days; (2) with an advanced gastric cancer; (3) who underwent ESD for multiple lesions; (4) another prevention after ESD (covering using polyglycolic acid sheets or clip closure). We divided the patients into two groups; ELC group and control group (no preventive treatment). We compared the post-ESD bleeding rates between the two groups. A total of the 178 patients were included, 37 patients underwent ELC and 141 patients were control. Post-ESD bleeding occurred in 8.1% (3/37) and 22.7% (32/141) in the ELC and control groups, respectively; there was no significant difference between the two groups (p=0.06). However, subgroup analysis showed significant difference. Post-ESD bleeding in the resected specimen less than 40 mm occurred in 0% (0/26) lesions in the ELC group and in 16% (12/75) lesions in the control group (p=0.03). Post-ESD bleeding in multiple antithrombotic agents occurred in 10% (1/10) lesions in the ELC group and 70% (7/10) lesions in the control group (p=0.02). Closure using endoloop and endoclips after gastric ESD would prevent post-ESD bleeding in patients receiving antithrombotic therapy, particularly in those with resected specimen <40 mm and among patients undergoing treatment with multiple antithrombotic agents.