Computational Simulations to Predict Fenestrated Stent Graft Rotation on Deployment

Abstract

The objective of this study was to use finite element simulations of fenestrated endovascular aneurysm repair to predict whether the stent graft would rotate on deployment, leading to fenestration misalignment and related complications, such as end-organ ischemia. After institutional research ethics approval, preoperative computed tomography scans and stent graft plans were collected for three patients who had undergone fenestrated endovascular aneurysm repair. The aortoiliac geometries were segmented from the common femoral artery to above the visceral vessels, and vessel centerlines were calculated. Geometries of the stent graft, based on the plans, as well as the guidewire and the delivery sheath were created. Realistic material properties, including frictional effects, were assigned for all materials, with the vessel wall properties being those of an elderly man taken from the literature. Using the finite element software LS-DYNA (Livermore Software Technology Corp, Livermore, Calif), vessel deformation in response to the guidewire was simulated by displacing the guidewire to follow the vessel centerline path and then removing this displacement, forcing the guidewire to try to straighten, until the deformations of the guidewire and the vessel wall reached equilibrium. The stent graft model was then compressed into a sheath that was delivered into position in the aorta by following the guidewire path. Once the device was in position, the sheath was removed, allowing the stent graft to expand and potentially rotate. Stent graft rotation was calculated as the angle between the positions of the graft before and after unsheathing. These calculated angles showed excellent agreement with the rotation measured clinically from the intraoperative imaging (7 degrees, 2 degrees, and 12 degrees computationally vs 5 degrees, 2 degrees, and 16 degrees clinically). A finite element model of stent graft delivery and deployment was developed to predict the amount of stent graft rotation that may occur clinically. Preliminary results show excellent agreement with the clinical data.