Meta-analysis of endovascular vs open repair for traumatic descending thoracic aortic rupture

Objective: Traumatic thoracic aortic injuries are associated with high mortality and morbidity. These patients often have multiple injuries, and delayed aortic repair is frequently used. Endoluminal grafts offer an alternative to open surgical repair. We performed a meta-analysis of comparative studies evaluating endovascular vs open repair of these injuries. Methods. A systematic search of studies reporting treatment of traumatic aortic injury was performed using the following databases: Medline/PubMed, CINAHL, Proquest, Up to Date, Database of Abstracts of Reviews of Effects (DARE), ClinicalTrials.gov, the Cochrane Central Register of Controlled Trials and the Cochrane Database of Systematic Reviews. Search terms were thoracic aortic trauma, traumatic thoracic aortic injury, traumatic aortic rupture, stent graft repair, and endovascular repair. Outcomes analyzed were procedure-related mortality, overall 30-day mortality, and paraplegia/paraparesis rate using odds ratios (OR) and 95% confidence intervals (CI). Publication bias was investigated using funnel plots. Assessment of homogeneity was performed using the Q test; statistical heterogeneity was considered present at P < .05. Weighted averages of age, interval to repair, and injury severity score were compared with the Welch t test; P < .05 was considered statistically significant. Results: Seventeen retrospective cohort studies from 2003 to 2007 were included. All were nonrandomized; no prospective randomized trials were found. These studies reported oil 589 patients; 369 were treated with open repair, and 220 underwent thoracic stent graft placement. There was no significant difference in age (mean 38.8 years for both) or interval to repair (mean 1.5 days for endoluminal repair; I day for open repair). Injury severity score was higher for patients undergoing endoluminal repair (mean, 42.4 vs 37.4 for open repair, P < .001). Procedure-related mortality was significantly lower with endoluminal repair (OR, 0.31; 95% CI, 0.15-0.66; P = .002). Overall 30-day mortality was also lower after endoluminal repair (OR, 0.44; 95% CI, 0.25-0.78; P = .005). Sixteen studies reported data for postoperative paraplegia; 215 patients were treated with endograft placement and 333 with open repair. The risk of postoperative paraplegia was significantly less with endoluminal repair (OR, 0.32; 95% CI, 0.1-0.93; P = .037). The Q test did not indicate significant heterogeneity for the outcomes of interest; publication bias was limited. Conclusions: Meta-analysis of retrospective cohort studies indicates that endovascular treatment of descending thoracic aortic trauma is an alternative to open repair and is associated with lower postoperative mortality and ischemic spinal cord complication rates. (J Vasc Surg 2008;48:1343-51.)