Transarterial embolization of gastrointestinal hemorrhages: retrospective study in Siriraj Hospital.

ABSTRACT

Objective: To determine the technical and clinical results of transarterial embolization of nonvariceal gastrointestinal hemorrhages (GIH) which cannot be managed endoscopically. Methods: A retrospective review of 21 embolizations in 77 patients who underwent arteriography for acute nonvariceal GIH was performed. Gastrointestinal hemorrhage was classified by the site of bleeding as upper, lower, or transpapillar including hemobilia and pancreatic duct bleeding. Clinical parameters and embolized data were assessed for clinical success. In-hospital mortality was also reported. Results: Technical success (bleeding target devascularization) was achieved in all 21 patients (100%). The complete technical success rate was 71.4% (15 out of 21 patients) while the partial technical success rate was 28.6% (6 out of 21 patients). The complication rate was very low (9.5%) including only 2 cases of bowel ischemia. No other complications were found. Eight out of 21 patients (38.1%) had rebleeding within 3 days. Upper GIH seemed to recur more frequently (4 out of 7 patients representing 57.1%) within the first 3 days than did lower (4 out of 11 patients at 36.4%) and transpapillar (0 of 3) GIH. Clinical success (no rebleeding after 30 days) was achieved in 11 of 21 patients (52.4%) including 3 out of 7 patients (42.9%) with UGIH, 7 of 11 patients (63.3%) with LGIH, and all patients with transpapillar hemorrhage. The overall mortality rate was 42.9% (9 out of 21 patients), with 42.9% (3 of 7 patients) for UGIH, 27.3% (3 of 11 patients) for LGIH and none for transpapillar hemorrhage. GIH was the cause of death in 6 of 9 patients (66.7%) while 3 out of 9 patients (33.3%) died from the other causes. Conclusion: Transarterial embolization is an effective treatment modality for lower GIH and transpapillar hemorrhage but less effective in upper GIH.