Predictive factors for limb occlusions after endovascular aneurysm repair Elsa M. Faure, MD, Jean-Pierre Becquemin, MD, and Frédéric Cochennec, MD, on behalf of ENGAGE collaborators, Créteil, France Objective: Greater flexibility and smaller sizes for introducer sheaths in the newest stent grafts increase the feasibility of endovascular aneurysm repair but raise concerns about long-term limb patency. The aim of the study was to determine the incidence of and predictive factors for limb occlusion after use of the Endurant stent graft (Medtronic Inc, Minneapolis, Minn) for abdominal aortic aneurysm. Methods: The Endurant Stent Graft Natural Selection Global Postmarket Registry (ENGAGE) prospectively included 1143 patients treated with bifurcated devices who were observed for up to 2 years. Limb occlusions were evidenced by computed tomography, angiography, or ultrasound. To predict stent graft limb occlusion, a two-step model-building technique was applied. We first identified predictors from a total of 47 covariates obtained at baseline and in the periprocedural period. Subsequently, we reduced the set of potential predictors to key factors that are clinically meaningful. To handle large numbers of covariates, we used the Classification And Regression Tree (CART) method. Results: Forty-two stent graft limbs occluded in 39 patients (3.4% of the patients). At 2 years, the rate of freedom from stent graft limb occlusion calculated by Kaplan-Meier plot was 97.9% (standard error [SE], 0.33%). Of the 42 occlusions, 13 (31%) were observed within 30 days and 30 (71%) within 6 months. The strongest independent predictors were distal landing zone on the external iliac artery, external iliac artery diameter #10 mm, and kinking. High-risk vs low-risk patients were identified according to a decision tree based on the strongest predictors. Freedom from stent graft limb occlusion was 96.1% (SE, 0.64%) in high-risk patients vs 99.6% (SE, 0.19%) in low-risk patients. Conclusions: After Endurant stent grafting, the incidence of limb occlusion was low. Classifying patients as high risk vs low risk according to the algorithm used in this study may help define specific strategies to prevent limb occlusion and improve the overall results of endovascular aneurysm repair using the latest generation of stent grafts. (J Vasc Surg 2015;61:1138-45.)

Since the introduction of endovascular aneurysm repair (EVAR) in the 1990s, stent graft designs have evolved with the intent of improving outcomes in patients with abdominal aortic aneurysm (AAA) associated with challenging anatomy. The Endurant stent graft (Medtronic Inc, Minneapolis, Minn), one of the latest generations of endovascular devices, was specifically developed to perform well in complex AAA anatomy.1 The broadening of the indications for EVAR may potentially be associated with an increasing rate of subsequent complications and graft failure. Stent graft designs of lower profile and greater flexibility allow the treatment of more challenging anatomy. This development has raised concerns about long-term limb patency, particularly when the grafts are used in more complex AAA anatomy. To prevent graft failure, it is critical that anatomic and technical factors associated

with stent graft limb occlusion are identified and that mechanisms of failure are elucidated. The Endurant Stent Graft Natural Selection Global Postmarket Registry (ENGAGE) is a large, prospective, consecutive, international patient registry collecting data from 79 high-volume, experienced institutions. This registry was opened in March 2009 and is ongoing.2 Extensive preoperative and postoperative informational forms are centrally collected and intensively monitored to ensure adequate data collection and minimal loss of patients to follow-up. The aim of this study was to determine the incidence of and predictive factors for limb occlusion after Endurant stent graft implantation for AAA in cases recorded in the ENGAGE registry.2 METHODS

From the Department of Vascular Surgery, Henri Mondor Hospital. Author conflict of interest: none. Additional material for this article may be found online at www.jvascsurg.org. Reprint requests: Elsa M. Faure, MD, Service de Chirurgie Vasculaire, Centre Hospitalier Universitaire Henri Mondor, 51 av du Maréchal de Lattre de Tassigny, 94000 Créteil, France (e-mail: elsafaure@hotmail. com). The editors and reviewers of this article have no relevant financial relationships to disclose per the JVS policy that requires reviewers to decline review of any manuscript for which they may have a conflict of interest. 0741-5214 Copyright Ó 2015 by the Society for Vascular Surgery. Published by Elsevier Inc. http://dx.doi.org/10.1016/j.jvs.2014.11.084

1138

The protocol and informed consent were approved by the Institutional Review Board, and all subjects gave informed consent. Patients entered into the ENGAGE registry include elective AAA patients, age >18 years, who have consented to the collection and release of their medical information. All patients underwent intervention with the bifurcated Endurant stent graft. This graft is approved in Europe for patients with a straight proximal neck of >10 mm and