Extravasation of sclerosant after injection of N-butyl-2-cyanoacrylate for a bleeding gastric Dieulafoy lesion.

cause she had been passing tarry stools for 2 days. Emergency endoscopy re− vealed an actively spurting Dieulafoy le− sion in the gastric fundus (l" Fig. 1). Be− cause of the technical difficulty of apply− ing hemoclips in this case we performed rubber−band ligation, which resulted in immediate hemostasis (l" Fig. 2). Twelve hours later, fresh blood reappeared in the nasogastric aspirate. A repeat endoscopy showed active oozing from the base of the ligated area (l" Fig. 3). Despite the in− jection of 12 mL of diluted epinephrine, the bleeding persisted. After discussing the situation with the patient she agreed to receive a sclerosant injection, and the bleeding stopped after an injection of a mixture of 0.5 mL of N−butyl−2−cyano− acrylate and lipiodol. The patient com− plained of dull epigastric pain after the injection, and a chest radiograph showed radiopaque material along the left dia− phragm (l" Fig. 4). Abdominal computed tomography confirmed that there was ex− travasation of sclerosant along the left subphrenic area (l" Fig. 5). The patient was free of any signs of infection and was discharged on day 6 of her hospital day. She remained well over the following 2 years. Endoscopic therapy is now a mainstay of treatment for Dieulafoy lesions. Several authors have documented hemostasis rates for endoscopic band ligation of 80 % ±100 % in case series of patients with Dieulafoy lesions [1, 2]. However, the optimal rescue therapy after failed endoscopic band ligation is not known. In the present case, injection therapy with diluted epinephrine failed to control the bleeding. Hemoclipping or coagulati− on therapy would have been technically difficult. It was believed that repeated band ligation would increase the risk of gastric perforation [3], and so Histoacryl injection was chosen as the last resort be− fore surgery [4]. Despite the successful hemostasis, the complication of injec− tion−site leakage and resultant transient abdominal pain are concerning. This might have been caused by direct pene− tration of the gastric wall by the injection needle. The short−term outcome of scler− osant extravasation in this case and in an− other reported case [5] was good, but it is still not clear what the long−term effects of this complication are.