Predictors of outcome and different management of aortobronchial and aortoesophageal fistulas

Abstract

Aortoesophageal and aortobronchial fistulas are uncommon but life-threatening conditions. The present study aimed to identify potential differences in outcomes, depending on the etiology, type, and management of the fistulas, and to determine mortality predictors. We retrospectively reviewed a series of 26 consecutive patients with thoracic aorta fistulas admitted to our institution from 1998 to 2013 (18 aortobronchial, 7 aortoesophageal, and 1 combined fistula). The mean age was 61.5 ± 13.4 years, with 22 men. Management was thoracic endovascular aortic repair (TEVAR) in 8, open repair in 7, and conservative in 11. The TEVAR and nonoperative patients were significantly older and presented with more comorbidities. Shock developed in 15 patients and sepsis in 9. The most common radiologic findings were intramural hematoma (65.4%), pseudoaneurysm (53.8%), and bronchial compression (46.20%). Active contrast extravasation (23.1%) and ectopic gas (19.2%) were associated with a worse prognosis. In-hospital mortality was 100% in the conservative group, 37.5% in the TEVAR group, and 14.3% in the open repair group (P = .04). Septic shock was the most common cause of death. The risk factors for in-hospital mortality were hemodynamic instability on admission (P = .02), sepsis (P = .04), and conservative management (P < .001). The overall long-term survival in surgical patients at 1 and 5 years was 66% and 58.7%, respectively. Infectious and malignant etiologies resulted in the worst prognosis. The outcomes are ultimately conditioned by the etiology of the fistula. Both open and endovascular management of aortic fistulas can prevent death by exsanguination; however, patients remain at high risk of infectious complications. Failure to treat the underlying cause will result in poor midterm outcomes.