Transcatheter arterial embolization versus surgery in the treatment of upper gastrointestinal bleeding after therapeutic endoscopy failure.

PURPOSE: To retrospectively compare the outcome of transcatheter arterial embolization (TAE) and surgery as salvage therapy of upper gastrointestinal bleeding after failed endoscopic treatment.
MATERIALS AND METHODS: From January 1998 to December 2005, 658 patients were referred to diagnostic/therapeutic emergency endoscopy and diagnosed with upper gastrointestinal bleeding. Ninety-one of these 658 patients (14%) had repeat bleeding or continued to bleed. Forty of those 91 patients were treated with TAE and 51 were treated with surgery. From the medical records, the following variables were recorded: demographic data, endoscopic diagnoses, comorbidities, lowest hemoglobin levels, total transfusion requirements, lengths of hospitalization stays, postprocedure complications, and mortality rates. The relative survival rate was calculated, and survival probability was calculated with the Kaplan-Meier technique.
RESULTS: Patients treated with TAE were older (mean age, 76 years; age range, 40-94 years) and had slightly more comorbidities compared to patients who underwent surgery (mean age, 71 years; age range, 45-89 years). The 30-day mortality rate in patients treated with TAE was one of 40 (3%) compared to seven of 51 (14%) in patients treated with surgery (P < .07). Most repeat bleeding could be effectively treated with TAE, both in the surgical and TAE groups.
CONCLUSIONS: The results of this study suggest that, after failure of therapeutic endoscopy for upper gastrointestinal bleeding, TAE should be the treatment of choice before surgery and that TAE can also be used to effectively control bleeding after failed surgery or TAE. There was a clear trend to lower 30-day mortality with use of TAE instead of surgery.