Endovascular stent-graft placement for acute and contained rupture of the descending thoracic aorta

To identify determinants of postinterventional death after endovascular stent-graft placement for acute rupture of the descending thoracic aorta, an emerging therapeutic modality for this highly life-threatening condition.

Between July 1999 and November 2004, 17 patients (14 males; mean age, 65+/-16 (25-83) years) underwent stent-graft repair of the descending thoracic aorta for acute rupture from a thoracic aneurysm (TAA, n=6), acute aortic dissection (AAD, n=6), penetrating aortic ulcer (PAU, n=3), or blunt chest trauma (n=2). Immediate, 30-day, 1-year, and 3-year mortality was assessed. Twenty-one clinical and procedural variables were evaluated in a post-hoc analysis regarding their influence on mortality. Of these, four preprocedural factors with the greatest impact were used to construct a rupture score with a scale from 0 (no adverse prognostic factors present) to 4 (all four adverse factors present).

Stent-graft placement was technically feasible in all patients. Complete exclusion of the ruptured aortic pathology could be achieved in only 11 (65%) patients, despite implantation of 1.6+/-0.9 stent-grafts per patient, with a median length of 130 mm. There was one procedure-related early complication (bleeding at the access site). One patient died immediately following the procedure because of progressive mediastinal hematoma, although the rupture site was effectively sealed. Overall survival rates were (76.5+/-10.3)% at 30 days and (52.9+/-12.1)% at 1 year and remained at (52.9+/-12.1)% at 3 years. The four most important preprocedural denominators of death were (1) TAA or AAD as the underlying etiology of aortic rupture (P=0.024), (2) maximum aortic diameter>5 cm (P=0.024), (3) presence of mediastinal hematoma (P=0.056), and (4) an estimated lesion length requiring >1 stent-graft to be covered (P=0.009). Furthermore, residual leakage at the conclusion of the procedure (P=0.009), postprocedural need for dialysis (P=0.004), and prolonged ventilation (P=0.043) were significantly associated with postprocedural death. Using a threshold of >or=3, the rupture score constructed on the basis of the four preprocedural denominators of death was found to be well suited to discriminate postprocedural death (1-year survival: (20.0+/-12.7)% in patients with a rupture score>or=3 vs. 100% in patients with a rupture score<3, P=0.001).

Endovascular stent-graft placement in patients with acute aortic rupture was technically feasible, albeit still associated with high mortality. A simple risk score constructed in retrospect, on the basis of preprocedural prognostic factors, appeared to provide a useful separation of candidates who are likely to benefit from a straightforward endovascular procedure and should be tested prospectively in future studies.