Outcomes after endovascular or open repair for degenerative descending thoracic aortic aneurysm using linked hospital data

Background: The population-based effectiveness of thoracic endovascular aortic repair (TEVAR) versus open surgery for descending thoracic aortic aneurysm remains in doubt. Methods: Patients aged over 50 years, without a history of aortic dissection, undergoing repair of a thoracic aortic aneurysm between 2006 and 2011 were assessed using mortality-linked individual patient data from Hospital Episode Statistics (England). The principal outcomes were 30-day operative mortality, long-term survival (5 years) and aortic-related reinterventions. TEVAR and open repair were compared using crude and multivariable models that adjusted for age and sex. Results: Overall, 759 patients underwent thoracic aortic aneurysm repair, mainly for intact aneurysms (618, 81.4 per cent). Median ages of TEVAR and open cohorts were 73 and 71 years respectively (P < 0.001), with more men undergoing TEVAR (P = 0.004). For intact aneurysms, the operative mortality rate was similar for TEVAR and open repair (6.5 versus 7.6 per cent; odds ratio 0.79, 95 per cent confidence interval (c.i.) 0.41 to 1.49), but the 5-year survival rate was significantly worse after TEVAR (54.2 versus 65 6 per cent; adjusted hazard ratio 1.45, 95 per cent c.i. 1.08 to 1.94). After 5 years, aortic-related mortality was similar in the two groups, but cardiopulmonary mortality was higher after TEVAR. TEVAR was associated with more aortic-related reinterventions (23.1 versus 14.3 per cent; adjusted HR 1.70, 95 per cent c.i. 1.11 to 2 60). There were 141 procedures for ruptured thoracic aneurysm (97 TEVAR, 44 open), with TEVAR showing no significant advantage in terms of operative mortality. Conclusion: In England, operative mortality for degenerative descending thoracic aneurysm was similar after either TEVAR or open repair. Patients who had TEVAR appeared to have a higher reintervention rate and worse long-term survival, possibly owing to cardiopulmonary morbidity and other selection bias.