Endovascular repair of abdominal aortic aneurysms: device-specific outcome.

OBJECTIVES Endovascular repair of abdominal aortic aneurysms, while advantageous because of its minimally invasive nature, falls short of achieving the long-term durability of traditional open surgical repair. Problems such as device migration, continued sac pressurization from endoleak, and graft limb thrombosis culminate in a high rate of secondary procedures and failure to protect against aneurysm rupture. While prior studies hint at a correlation between these postprocedural events and specific device design, a single comparative analysis that correlates device attributes with clinical outcome has not been performed. METHODS Over 6 years ending in 2002, 703 patients underwent endovascular treatment of infrarenal abdominal aortic aneurysms. During this time, five devices were used, ie, Ancure, AneuRx, Excluder, Talent, and Zenith, and six device-specific groups were analyzed; the Zenith group was subdivided into those placed as part of the multicenter trial (Zenith-MCT) and those under a sponsor-investigator investigational device exemption trial (Zenith-SIT). Results were assessed with the Kaplan-Meier method for censored data, and the log-rank test was used to ascertain differences between device groups. RESULTS While overall survival was diminished in the Zenith-SIT group (P =.046), risk for aneurysm-related death was similar in all groups (P =.336), averaging 2% or less at 12 months. Among the total cohort of patients, freedom from rupture was 98.7% +/- 0.9% at 24 months, without demonstrable differences between groups (P =.533). There were no statistically significant differences in rate of secondary procedures, conversion to open repair, or migration. There were, however, significant differences in risk for graft limb occlusion and rate of endoleak between groups. Limb occlusion occurred most often with Ancure devices (11% +/- 4.6% at 12 months, P =.009). Endoleak of any type was most common with Excluder devices (64% +/- 11% at 12 months, P =.003), a finding directly related to increased frequency of type II leaks in that group (58% +/- 11% at 12 months, P =.001). While there were no differences in frequency of type I or type III endoleak, a trend toward increased risk for microleak was observed with AneuRx devices (4.0% +/- 1.3%, P =.054), and more modular separations were observed with Zenith devices (3.5% +/- 2.3%, P =.032). Shrinkage at 12 months correlated with frequency of endoleak in the device groups, and was most common in the two Zenith groups (54% +/- 7.3% in the Zenith-MCT group and 56% +/- 7.8% in the Zenith-SIT group) and the Talent group (52% +/- 9.7%) and was least in the Excluder group (15% +/- 7.9% at 12 months, P <.001). By contrast, sac growth occurred most often in the Zenith-SIT group (13% +/- 4.5% at 12 months, P =.034), possibly as a result of the challenging aortoiliac anatomy frequently present in these patients. CONCLUSIONS There are significant differences in frequency of limb occlusion and endoleak between groups with different endovascular devices. Knowledge of these and other differences is instructional in development of next-generation endovascular devices, incorporating design features linked to satisfactory outcome while abandoning those associated with device failure.