This is a pilot study to investigate the usefulness of closure of mucosal defect after gastric endoscopic submucosal dissection (ESD) in patients that are at high risk of post-procedural haemorrhage. Delayed haemorrhage is still an important adverse event of ESD, occurring more frequently for gastric lesions. Risk factors identified for delayed haemorrhage include chronic kidney disease, use of antithrombotic agent, lesion size >20mm, specimen size >30mm. Current established methods to prevent this complications could not completely eliminate the chance of bleeding, especially among high-risk cases. Closure of ESD defect may prevent ongoing exposure of submucosal vessels to gastric acid, and further reduce the risk of delayed haemorrhage. The investigators conduct this study to investigate the effect of closing the defect with endoscopic clips and loop. 30 patients who are undergoing gastric ESD deemed high risk of delayed haemorrhage would be recruited, with closure of defect after resection. The rate of delayed haemorrhage would be compared with historical cohort of patients.

Endoscopic submucosal dissection (ESD) is an endoscopic technique aiming to achieve en-bloc resection of mucosal neoplastic lesion in the gastrointestinal tract. Since the first report of ESD in 2002, the indication of the procedure has been expanding. Over the years, large-scale prospective study of ESD in the stomach has been reported. It is now considered as the standard of treatment for early gastric cancer confined to the mucosa, achieving an excellent overall survival comparable to that of surgical resection.

As compared to conventional endoscopic mucosal resection (EMR), ESD is technically more challenging and also carries higher procedural risks. Important adverse events associated with gastric ESD include hemorrhage (intraoperative or delayed) and perforation. The reported incidence of delayed hemorrhage of gastric ESD is generally higher than that of esophageal or colorectal ESD5. A recent systematic review and meta-analysis identified risk factors for post-ESD delayed hemorrhage as follow: chronic kidney disease (OR 3.38), use of antithrombotic agent (OR 1.63), lesion size >20mm (OR 2.70), specimen size >30mm (OR 2.85).

Established methods of preventing post-ESD hemorrhage include the use of proton pump inhibitor (PPI) and prophylactic coagulation of visible vessels after ESD7-9. However, both methods could not completely eliminate the chance of bleeding, especially among high-risk cases. Additional strategies may be required to further reduce the risk of delayed hemorrhage. Recently, Kataoka Y et al investigated the use of polyglycolic acid (PGA) sheets for covering the ESD defect in a randomized controlled trial. Unfortunately the endoscopic "shielding" method did not decrease the rate of post-ESD hemorrhage. The study had an unexpectedly low rate of bleeding in the control arm, which maybe due to the suboptimal inclusion criteria. On the other hand, with recent advances in endoscopic closure methods such as the use of OverStitchTM (Apollo Endosurgery, Inc), loop-clip purse string closure and line-assisted closure, complete closure of post-resection defect (even full thickness defects) is technically feasible. It is unclear whether achieving complete closure of the mucosal defect after gastric ESD could reduce the risk of delayed hemorrhage. Therefore the investigators decided to perform this pilot study to investigate the efficacy of this technique.

The purpose of this study is to investigate the clinical usefulness of endoscopic closure of mucosal defect after gastric ESD in patients at high risk of post-ESD hemorrhage. Technical success, clinical success and the incidence of post-ESD adverse events would be recorded. Comparative study would be conducted with historical control.

Inclusion Criteria:

1. Patients undergoing elective gastric endoscopic submucosal dissection for epithelial lesions

2. Procedure deemed at high risk of post-procedural hemorrhage due to the following:

   1. End stage renal disease (Estimated GFR <15ml/min)
   2. Patients on anti-thrombotic agents (Double antiplatelet, warfarin or direct oral anticoagulants)
   3. Post ESD mucosal defect size >4cm

3. Target subjects receiving sufficient briefing from the attending physician regarding the content of this study and providing informed consent for participation

4. Over 20 years of age

Exclusion Criteria:

1. Recurrent / remnant lesion after previous endoscopic resection
2. Lesions arising from surgical anastomotic site, such as gastrojejunostomy / gastroduodenostomy.
3. Marked electrolyte abnormalities
4. Allergic to components of injection solutions: Epinephrine, hyaluronic acid etc
5. Other cases deemed by the examining physician as unsuitable for safe treatment
6. Patients who refused to participate