A computational framework for investigating the positional stability of aortic endografts

Abstract

Endovascular aneurysm repair (Greenhalgh in N Engl J Med 362(20):1863-1871, 2010) techniques have revolutionized the treatment of thoracic and abdominal aortic aneurysm disease, greatly reducing the perioperative mortality and morbidity associated with open surgical repair techniques. However, EVAR is not free of important complications such as late device migration, endoleak formation and fracture of device components that may result in adverse events such as aneurysm enlargement, need for long-term imaging surveillance and secondary interventions or even death. These complications result from the device inability to withstand the hemodynamics of blood flow and to keep its originally intended post-operative position over time. Understanding the in vivo biomechanical working environment experienced by endografts is a critical factor in improving their long-term performance. To date, no study has investigated the mechanics of contact between device and aorta in a three-dimensional setting. In this work, we developed a comprehensive Computational Solid Mechanics and Computational Fluid Dynamics framework to investigate the mechanics of endograft positional stability. The main building blocks of this framework are: (1) Three-dimensional non-planar aortic and stent-graft geometrical models, (2) Realistic multi-material constitutive laws for aorta, stent, and graft, (3) Physiological values for blood flow and pressure, and (4) Frictional model to describe the contact between the endograft and the aorta. We introduce a new metric for numerical quantification of the positional stability of the endograft. Lastly, in the results section, we test the framework by investigating the impact of several factors that are clinically known to affect endograft stability.

Figure 1

Components of stent design showing (a) planar base strut, (b) cylindrical stent ring and (c) stent assembly.

Figure 2

Left: Anterior and lateral views of the idealized baseline aortic model labeled including dimensions (in mm). Right: Short neck and higher curvature models adopted for the study of neck fixation length and neck angulation.

Figure 3

Tensile strain-stress relationship for the aortic wall tissue based on uniaxial test data (Raghavan, Webster et al. 1996).

Figure 4

(a) Deformed spatial configuration Ω_t and unloaded material configuration Ω₀ obtained via incremental inverse stress analysis. (b) Pre-stress in deformed spatial configuration Ω_t under mean blood pressure.

Figure 5

Schematic representation of components and steps of the deployment of the endograft within the aortic model. (a) Initial spatial configuration of pre-stressed aortic model under mean aortic pressure MAP, unloaded stent, and bent graft following the main curvature of the abdominal aorta. No contact is activated at this time. (b) The stent is crimped, bent, and expanded until contact is established with the graft. At this point, the junctions between stent and graft are enforced by tied contact constraints. (c) The proximal and distal landing zones are expanded via an incremental pressure P to make room for the stent-graft. (d) The incremental pressure P is released and contact between the aortic wall and stent-graft is established. The aneurysm wall contracts slightly in the sac region since it is no longer loaded by the MAP.

Figure 6

(a) mixed (triangular and quadrilateral) mesh of the solid domain. (b) Triangulated surface mesh of the fluid domain. (c) Volumetric mesh of the fluid domain.

Figure 7

Sequential coupling strategy between the solid mechanics and fluid mechanics components of the framework: The CSM provides the equilibrium configuration of the deployed endograft subjected to certain hemodynamic loads. This equilibrium configuration is passed to the CFD module, where a new computation of the loads acting on the fixed CSM geometry is performed. The CSM module shows the Von Mises stress on the wall, including the concentration of stresses in the landing zone of the endograft, whereas the CFD module shows a volume render of blood velocity, pressure, and wall shear stress.

Figure 8

Left: Plot of contact pressure P_c, contact shear: τ_eq, and instability index II at proximal the proximal fixation zone for the baseline endograft. Left: Plot of instability index for the Baseline model in lateral and anterior views.

Figure 9

Average instability index for the baseline (left panel) and short-neck (right panel) aortic models. Significantly reduced proximal neck in the SN model leads to an 18% increase in the proximal average II.

Figure 10

Average instability index for the baseline (left panel) and higher curvature (right panel) aortic models.

Figure 11

Average instability index II for the baseline model under different friction coefficients between device and aortic wall (μ = 0.05, 0.30, 0.75). With the low friction coefficient μ = 0.05 the proximal and distal II increased significantly by 21% and 13% respectively compared to the baseline model, indicating lower device stability. With the high friction coefficient μ = 0.75 the proximal and distal II decreased by 26% and 21% respectively compared to the baseline model, indicating improved device stability.

Figure 12

Instability index for the baseline model with a 10% oversized device (left), and a 15% oversized device (right). Device oversizing leads to a 6% improvement in device stability in the proximal zone.

Figure 13

Maximum circumferential stress in the aortic neck is a function of endograft oversizing: larger degrees of oversizing increase the circumferential stress more and more and may result in adverse events such as neck dilation.