BACKGROUND AND AIM: Intraoperative bleeding is an important determining factor for the technical difficulty and safety of endoscopic submucosal dissection (ESD) for gastric neoplasms, which was previously difficult to predict before ESD. In the present study, we investigated whether endoscopic ultrasound (EUS) could be used to preoperatively predict intraoperative bleeding. METHODS: The study included 106 patients who underwent EUS before ESD. EUS was used to evaluate the submucosal vascular structure. Patients who had at least 10 vascular structures per field of view or a vessel at least 500 µm in diameter were classified into the rich group (Group R), and others were classified into the non-rich group (Group N). The two groups were compared retrospectively with respect to procedure time, degree of anemia, frequency of clip use, and others. RESULTS: There were 24 patients in Group R and 82 patients in Group N. Submucosal caner was found in 54.2% of patients in Group R and 18.3% in Group N. The reduction in hemoglobin was 5.8% in Group R and 3.45% in Group N. The procedure time was 151 min in Group R and 100 min in Group N. The frequency of clip use was 79.2% in Group R and 31.7% in Group N. A multivariate analysis revealed a significant difference in the depth of invasion and frequency of clip use between the two groups. CONCLUSIONS: The results suggest that identification of submucosal vascular structure by EUS might allow prediction of intraoperative bleeding during ESD.
BACKGROUND AND AIM: Intraoperative bleeding is an important determining factor for the technical difficulty and safety of endoscopic submucosal dissection (ESD) for gastric neoplasms, which was previously difficult to predict before ESD. In the present study, we investigated whether endoscopic ultrasound (EUS) could be used to preoperatively predict intraoperative bleeding. METHODS: The study included 106 patients who underwent EUS before ESD. EUS was used to evaluate the submucosal vascular structure. Patients who had at least 10 vascular structures per field of view or a vessel at least 500 µm in diameter were classified into the rich group (Group R), and others were classified into the non-rich group (Group N). The two groups were compared retrospectively with respect to procedure time, degree of anemia, frequency of clip use, and others. RESULTS: There were 24 patients in Group R and 82 patients in Group N. Submucosal caner was found in 54.2% of patients in Group R and 18.3% in Group N. The reduction in hemoglobin was 5.8% in Group R and 3.45% in Group N. The procedure time was 151 min in Group R and 100 min in Group N. The frequency of clip use was 79.2% in Group R and 31.7% in Group N. A multivariate analysis revealed a significant difference in the depth of invasion and frequency of clip use between the two groups. CONCLUSIONS: The results suggest that identification of submucosal vascular structure by EUS might allow prediction of intraoperative bleeding during ESD.