Endoleaks after endovascular repair of thoracic aortic aneurysms.

OBJECTIVE Endoleaks are one of the unique complications seen after endovascular repair of thoracic aortic aneurysms (TEVAR). This investigation was performed to evaluate the incidence and determinants of endoleaks, as well as the outcomes of secondary interventions in patients with endoleaks, after TEVAR. METHODS Over a 6-year period, 105 patients underwent TEVAR in the context of pivotal Food and Drug Administration trials with the Medtronic Talent (n = 64) and Gore TAG (n = 41) devices. The medical and radiology records of these patients were reviewed for this retrospective study. Of these, 69 patients (30 women and 39 men) had follow-up longer than 1 month and were used for this analysis. The patients were evaluated for the presence of an endoleak, endoleak type, aneurysm expansion, and endoleak intervention. RESULTS The mean follow-up in this patient cohort was 17.3 +/- 14.7 months (range, 3-71 months). Endoleaks were detected in 29% (20/69) of patients, of which 40% (8/20) were type I, 35% (7/20) were type II, 20% (4/20) were type III, and 5% (1/20) had more than one type of endoleak. Patients without endoleaks experienced greater aneurysm sac regression than those with endoleaks (-2.89 +/- 9.1 mm vs -0.13 +/- 7.2 mm), although this difference was not statistically significant (P = .232). All but 2 endoleaks (90%; 18/20) were detected on the initial postoperative computed tomographic scan at 30 days. Two endoleaks (10%; 2/20) developed late. The endoleak group had more extensive aneurysms with significantly larger aneurysms at the time of intervention (69.4 +/- 10.5 mm vs 60.6 +/- 11.0 mm; P = .003). Factors predictive of endoleak included male sex (P = .016), larger aneurysm size (P = .003), the length of aorta treated by stent grafts (P = .0004), and an increasing number of stents used (P < .0001). No open conversions were performed for treatment of endoleaks. Four (50%) of the eight type I endoleaks were successfully repaired by using endovascular techniques. None of the type II endoleaks was treated by secondary intervention. During follow-up, the maximum aneurysm diameter in the type II endoleak patients increased a mean of 2.94 +/- 7.2 mm (range, -4.4 to 17 mm). Spontaneous thrombosis has occurred in 29% (2/7) of the type II endoleaks. Patients with type III endoleaks experienced a decrease in mean maximal aneurysm diameter of 0.78 +/- 3.1 mm during follow-up. CONCLUSIONS Endoleaks are not uncommon after TEVAR. Many type I endoleaks may be treated successfully by endovascular means. Short-term follow-up suggests that observational management of type II endoleaks is associated with continued sac expansion, and these patients should be monitored closely.