Superselective embolization for lower gastrointestinal hemorrhage: an institutional review over 7 years.

INTRODUCTION: Superselective embolization of visceral arterial branches has become integral in the management of acute lower gastrointestinal (GI) hemorrhage. The present study aimed to evaluate the success of superselective embolization as a primary therapeutic modality in the control of lower GI hemorrhage and to identify factors associated with rebleeding and surgical intervention after the procedure.
METHODS: We performed a retrospective review of all cases of superselective embolization for acute lower GI bleeding during a 7-year period (December 2000-October 2007) in a single 1,300-bed hospital in Singapore. Hemostasis was achieved with microcoils, polyvinyl alcohol particles, gelfoam, or by selective vasopressin infusion. Various clinical and hematologic factors were analyzed against rebleeding and surgical intervention after the procedure.
RESULTS: A total of 265 patients underwent mesenteric angiography for GI hemorrhage. Superselective embolization of visceral vessels for lower GI hemorrhage was performed in 32 patients (12%) whose median age was 66 years (range: 34-82 years). The group was of similar gender distribution, and the median follow-up was 8 months (range: 1-32 months). Location was the small bowel in 19% and the colon in 81%. The underlying etiologies included diverticular disease (59%), angiodysplasia (19%), ulcers (19%), and malignancy (3%). In 31 patients (97%) technical success was achieved, with immediate cessation of hemorrhage in every case. Clinical success was achieved in 20 patients (63%), all of whom were discharged well with no further intervention. Seven patients rebled, and 9 underwent surgery: 1 for incomplete hemostasis, 4 for rebleeding, 1 for infarcted bowel postembolization, and 3 on the basis of the surgeon's decision. There were 2 anastomotic leaks; 1 after surgery for postembolization ischemia and 1 after surgery for rebleeding. Overall mortality in this series was 9%. Rebleeding was more likely to occur if the site of bleeding was located in the small bowel compared to the colon (OR: 8.33, 95% CI 1.03-66.67). It was also more likely in patients with a hematocrit level </=20.0% (OR: 7.52, 95% CI: 1.14-50.00) and a platelets level </=140 x 10(9)/l (OR: 9.35, 95% CI: 1.36-62.5) just before the procedure. Surgical resection was also more likely in patients with a hematocrit level </=20.0% just before embolization (OR: 12.66, 95% CI: 1.96-83.33), and it appeared to be more likely if the underlying cause was diverticular disease (OR 8.70, 95% CI: 0.93-83.33).
CONCLUSIONS: The use of superselective mesenteric embolization for the treatment of lower GI bleeding is highly successful and relatively safe-97% technical success and 3% postembolization ischemia in our series. In 63% of cases it is definitive without any further intervention. Postembolization ischemia and surgery may be associated with a higher risk of anastomotic leak. Greater vigilance must be adopted in treating patients who have active hemorrhage from the small bowel and in those with a hematocrit </=20.0%.