BACKGROUND: In this study we analyzed embolization and stent-graft results. METHODS: Demographics, indications, procedures, and outcomes of patients treated with embolization or stent grafting for late postoperative bleeding after major abdominal surgery were retrospectively recorded. Outcomes were analyzed on an intention-to-treat basis. RESULTS: Between 2004 and 2008, 14 consecutive patients (11 men and 3 women, mean age 64 years) were treated for hemorrhage responsible for shock in 6 patients (43%), occurring after pancreaticoduodenectomy (n=13) or subtotal gastrectomy (n=1). Mean onset occurred at 23 days postoperatively (range 7-75 days). Bleeding site included: the stump of the gastroduodenal artery (n=10), splenic artery (n=2), common hepatic artery (n=1), and right gastric artery (n=1). Initial success was obtained in 13 patients (93%); the only failure of stent-graft deployment required re-laparotomy. Treatment included embolization in 8 patients and stent grafting in 5 patients. In the embolization group, 5 complications (62%) occurred: 4 rebleeding and 1 gastric perforation, compared with no early complications in the stent-graft group. One patient died in each group. The mean follow-up was 25 months (range 6-57 months). CONCLUSIONS: Stent grafting seems to provide definitive hemostasis and fewer complications compared with embolization.
BACKGROUND: In this study we analyzed embolization and stent-graft results. METHODS: Demographics, indications, procedures, and outcomes of patients treated with embolization or stent grafting for late postoperative bleeding after major abdominal surgery were retrospectively recorded. Outcomes were analyzed on an intention-to-treat basis. RESULTS: Between 2004 and 2008, 14 consecutive patients (11 men and 3 women, mean age 64 years) were treated for hemorrhage responsible for shock in 6 patients (43%), occurring after pancreaticoduodenectomy (n=13) or subtotal gastrectomy (n=1). Mean onset occurred at 23 days postoperatively (range 7-75 days). Bleeding site included: the stump of the gastroduodenal artery (n=10), splenic artery (n=2), common hepatic artery (n=1), and right gastric artery (n=1). Initial success was obtained in 13 patients (93%); the only failure of stent-graft deployment required re-laparotomy. Treatment included embolization in 8 patients and stent grafting in 5 patients. In the embolization group, 5 complications (62%) occurred: 4 rebleeding and 1 gastric perforation, compared with no early complications in the stent-graft group. One patient died in each group. The mean follow-up was 25 months (range 6-57 months). CONCLUSIONS: Stent grafting seems to provide definitive hemostasis and fewer complications compared with embolization.