Endoscopic submucosal dissection using gel solution versus glycerol for submucosal injection: a randomized controlled multi-centric trial

Aims Submucosal injection is a crucial step when performing endoscopic submucosal dissection (ESD). Glycerol is the most used solution that has been implemented, however novel formulations are emerging. This study aimed to comparatively evaluate the submucosal injection using a gel solution versus glycerol during ESD procedures in patients with superficial gastric and rectal (pre)malignant lesions.

Methods We conducted a prospective multicenter randomized controlled trial in patients with documented gastric or rectal lesions indicated for ESD. Primary endpoint was dissection speed, defined as the dissected surface (mm2)/ESD duration (min). The enrollment goal was 266 patients randomized in 1:1 ratio assuming the gel solution would increase the dissection speed by 5.13 mm²/min or ~23% compared to glycerol, based in historic data at the coordinating center. Other endpoints included rate of en bloc resection, complete endoscopic resection, total hemostatic time, need for hemostatic forceps during procedure and serious adverse events, evaluated by AGREE classification. ClinicalTrial NCT04977401.

Results We randomized 31 patients (mean age 67, 58% male), 16 to gel group and 15 to glycerol group at 2 centers. The trial was discontinued early because the gel solution was withdrawn from market by the manufacturer. Nine lesions in the stomach and 22 in the rectum were resected. The mean dissection speed in the gel group was 34.4±14.6 vs 25.7±14.0 mm²/min in the glycerol group, which is 25% faster, but the a priori test did not reach statistical significance due to small sample size (p=0.100). Dissection speed was higher in gastric lesions for gel solution than for glycerol (33.1±11.7 vs 17.9±8.4 mm²/min, p=0.065). En bloc resection was similar between the groups with each group having one lesion not dissected en bloc. Complete endoscopic resection was achieved in all lesions that had en bloc resection. The mean number of intraprocedural bleeds per procedure was similar, but we observed less need of hemostatic forceps for bleeding control in the gel group (56% vs 87%, p=0.113). Total hemostatic time was 18.6±31.3 and 23.0±30.1 minutes for gel and glycerol respectively. There were 4 major adverse events (grade IIIa), 3 serious bleeds in the gel group and 1 perforation in the glycerol group (p=0.600). Two bleeds occurred after procedure, day 7 and 10, requiring blood transfusion and all 3 episodes were treated endoscopically. The perforation was treated endoscopically with 3 clips.

Conclusions Although this trial was stopped early, this small, randomized trial showed a potential increase of ESD dissection speed with gel solutions over glycerol, with similar resection success and a trend towards a lower need for hemostatic forceps. Further research into different types of lifting agents is warranted.

 

MissingLinkText

Conflicts of interest

Authors do not have any conflict of interest to disclose.

Publication History

Article published online:
15 April 2024