Mesenteric angiography of patients with gastrointestinal tract hemorrhages: a single center study.

PURPOSE This retrospective study was designed to investigate the transcatheter mesenteric angiography of patients with gastrointestinal (GI) bleeding and to determine the most important variables that should be monitored in patients with GI bleeding prior to transcatheter arteriography. MATERIALS AND METHODS In this study, we evaluated the transcatheter mesenteric angiography results of patients with massive GI bleeding (defined as hypotension, tachycardia, and a greater than 4-unit blood transfusion requirement in 24 h) seen between 2005 and 2009. Detailed clinical follow-up and accessible hospital data from 45 procedures were examined from 42 patients (two procedures were performed in three patients) between 24 and 85 years old (mean age, 57.6 years). The present study included 33 males and 9 females. Angiography was performed for lower GI bleeding in 22 patients, upper GI bleeding in 15 patients, and upper/lower (multiple origins) GI bleeding in five patients. Imaging work-ups, including endoscopic interventions, and follow-ups with patients after the procedure were evaluated in detail. Several variables recorded prior to the procedure, including the clinical status, etiological cause of the bleeding, bleeding parameters (e.g., international normalized ratio, platelets), imaging workup, gender, season, and angiography time, were examined. RESULTS Embolization was performed in 24 (53%) of the 45 procedures. Overall, the technical success rate of the diagnostic arteriograms was 100%, and no major complications occurred. For the embolizations, coils were used in 17 patients (70%), polyvinyl alcohol particles were used in six patients (25%), and n-butyl cyano-acrylate was used in one patient (4%). The detection rate of mesenteric arteriographies to examine GI bleeding performed outside of normal working hours was significantly greater than the detection rate of the arteriographies performed during normal working hours (P = 0.050). Low platelet levels or a prolonged prothrombin time were not associated with the mesenteric arteriography results (P = 1.00). Interestingly, the intermittent nature of GI bleeding was the most challenging part of detection, which made management of the bleeding difficult. Blind embolization of the left gastric artery was only helpful in preventing massive bleeding in three out of eight patients with upper GI bleeding. CONCLUSION Endoscopy for upper gastrointestinal bleeding and scintigraphy for lower gastrointestinal bleeding are important steps in the management and outcome of transcatheter angiography. Computerized tomography angiography is a promising tool for the treatment of both upper and lower GI bleeding, and this procedure has become part of the imaging toolset. In addition, angiography performed outside of working hours had a higher rate of clinical success than the angiographies performed in working hours, most likely secondary to much appropriate timing of arteriogram in terms of critical bleeding intervals.