4707 The use of cyanoacrylate glue as a second line treatment in endoscopic ulcer hemostasis.

Abstract

Background: Endostasis for ulcer hemorrhage is of proven benefit. However, the best treatment for failed endostasis is uncertain. Further attempts at endostasis or surgery are associated with high mortality. The decision to procede to surgery is difficult when severe comorbidity is present. In high risk patients with failed endostasis there may be a role for cyanoacrylate glue injection. Aim: to evaluate cyanoacrylate glue injection in emergency endoscopy for ulcer bleeding when conventional methods of endostasis fail. Patients & methods: A retrospective study of 255 patients who underwent emergency endoscopy for bleeding gastric and duodenal ulcers between April 1996 and September 1999. Initial techniques for hemostasis included epinephrine 1:10,000 and heater probe application. Additional treatment with cyanoacrylate glue was performed in 20 patients (7.84%) in a further attempt to obtain hemostasis as surgical intervention was considered high risk. Results: Of the 20 patients, 12 were male, 8 were female and the median age was 68.5 years (range 33-97 years). There were 13 duodenal ulcers and 7 gastric ulcers. Prior to presentation, 17 patients (85%) were A.S.A. physical status scale ≤3. 15% (n=3) were hospitalised for other reasons prior to bleeding. Initial hemostasis was successful in 17 patients (85%), 3 patients proceeded directly to surgery, one died on day 4. One patient had an early re-bleed (within 24 hours of glue injection) and died 7 days after surgery. Delayed re-bleeding (>24 hours after glue injection) occurred in 3 patients (days 2,3,&9), one died 5 days after surgery, an active decision to withhold further treatment was made in the other 2 and both died. Therefore 5 deaths (25%) were related to bleeding. Four patients died of causes unrelated to further bleeding (days 3,4,7, & 9). Overall mortality was 45% (n=9), Discussion: When traditional methods of endoscopic hemostasis fails, patients usually require surgery. When surgery is considered very high risk cyanoacrylate glue is successful in obtaining initial hemostasis in 85% of patients where epinephrine and heater probe therapy have failed. Delayed re-bleeding after initial successful hemostasis was relatively uncommon (15%). The high mortality in these patients reflect their age and co-morbidity as only high risk patients for surgical intervention were treated with glue injection. Conclusion: Cyanoacrylate tissue glue may have a role to play in the endostasis of bleeding ulcers in conjunction with conventional methods of hemostasis.