Duodenal endoscopic mucosal resection (EMR) has a high risk of bleeding. In a historical consecutive series of duodenal EMR in our hospital we found in 13/34 cases (38%) active bleeding or high-risk stigma of bleeding (visible vessel after removal of a blood clot) on next day gastroscopy (EGD). Prophylactic application of a haemostatic gel (3D-Matrix 621-062) to the EMR field may be an easy way to prevent delayed bleeding¹.

The aim of the study was to assess the efficacy of 3D-Matrix 621-062 in reducing delayed bleeding following duodenal EMR.

We conducted a single-arm, multicenter observational trial of patients undergoing hot-snare EMR of duodenal non-ampullary lesions of ≥ 10 mm.

Following EMR, 3D-Matrix 621-062 was applied directly to the EMR defect. The next day, an EGD was performed to detect active bleeding or high-risk stigma of bleeding (visible vessel after removal of a blood clot).

We hypothesized a 50% decrease in bleeding stigma from 38 to 19%; sample size was calculated for power of 90% and Alpha of 0.05. The number of subjects to be enrolled was 59 patients.

From October 2020 until August 2025, 59 patients from 5 different centers were included. Mean lesion size was 21.73 mm. 3D-Matrix 621-062 application was feasible in all patients. On EGD the next day, active bleeding (10 patients; 17%) or visible vessel after removal of a blood clot (6 patients; 10%) was seen in 16 patients (27%) (Fisher’s exact P=0.35, NS). During hospital stay, melena and hematemesis were observed in respectively 4 (7%) and 1 patient (2%). Hypotension was not observed. Treatment by placement of a hemoclip was performed in only 9 patients (15%) and relevant blood loss of ≥ 1 g/dL was seen in 3 patients (5%).

Application of 3D-Matrix 621-062 following duodenal EMR did not significantly decrease the rate of bleeding or bleeding stigmata on next day EGD in comparison to a historical cohort.