Transarterial treatment of acute gastrointestinal bleeding: prediction of treatment failure by clinical and angiographic parameters.

BACKGROUND Acute gastrointestinal bleeding (GIB) is a life-threatening abdominal emergency and can be treated by transarterial embolization (TAE). Rehemorrhage and poor outcome are associated with several clinical factors. This study investigated the clinical and angiographic parameters associated with treatment failure for patients with acute GIB undergoing TAE. METHODS Sixty-seven patients who had angiographic evidence of contrast extravasation and who received subsequent TAE were included in this study. Treatment failure was defined as continuous or recurrent bleeding that required surgery within 7 days after the bleeding episode and/or death within 1 month. Univariate and multivariate logistic regression was applied to analyze the clinical and angiographic parameters affecting treatment failure. RESULTS Patients were divided into two groups: success (n = 35, 52.3%) and failure (n = 32, 47.7%). In the failure group, 22 patients (68.9%) re-bled and then received surgery. With the aid of angiographic localization, 68.2% (15 of 22 patients) survived after surgery. The other 10 patients who did not receive surgery died within 30 days. Several clinical and angiographic parameters analyzed by multivariate analysis were associated with treatment failure (p < 0.05), including presence of coagulopathy [odds ratio (OR), 14.7], number of supplying arteries >1 (OR, 13.2), and a distance of >5 cm (OR, 6.3) during TAE. CONCLUSION Angiographic parameters associated with treatment failure in patients undergoing TAE are established when the number of supplying arteries is >1, and a distance of >5 cm. Patients with these risk factors should be watched carefully for recurrence in the post-procedural period.