Abstract

The purpose of this study is to develop modeling methodologies to understand how different stents alter the aneurysmal flow. Two porous stents similar to commercial stents (Tristar^TM stent and Wallstent^®) were virtually deployed in an ideal geometry and a patient-specific geometry for Computational Fluid Dynamics (CFD) analysis. An asymmetric stent was manufactured and tested in a canine vein-pouch aneurysm model. CFD results of asymmetric stent were compared with the angiography images of the in vivo studies. The same direct stent modeling techniques were applied to evaluate a patient-specific stent design. We found the stent design affected a cerebral aneurysm hemodynamics. The asymmetric stent, and patient-specific stent efficiently blocked the aneurismal inflow.

Introduction

Stent is a less invasive intervention designed to reduce wall shear stress or induce thrombosis in cerebral aneurysm. However, stenting does not always guarantee affirmative results, rather the results have been inconsistent [1, 2, 4, 5]. The effects of endovascular stenting on aneurysmal flow alterations are poorly understood.

1. Aneurysm Geometry

Typically a saccular aneurysms can be idealized by a spheres on a cylindrical vessel. The stents were deployed in this idealized aneurysm geometry. The diameter of the idealized aneurysm, and vessel were 19mm, and 4.75 mm respectively. The vessel curvature varied from 0(C0) to 1/9mm^-1(C4) (Fig. 1). The patient-specific aneurysm geometry was reconstructed using CT Angoigraphy (CTA) images of a patient’s right anterior communicating artery (ACA) (Fig. 2). Stents were virtually placed across the aneurysm neck.

2. Stent Geometry The effect of stents, similar to commercial stents (TristarTM stent and Wallstent®), on aneurismal hemodynamics was investigated (Fig. 3). We modeled these stent geometries in commercial CAD software, ProEngineer (PTC, Needham, MA). The equivalent diameter of the rectangular struts and the diameter of the circular struts were both 0.1mm. The porosities of these stents were about 82%. The model stents were bent and fitted into the aneurysm model.

Fig. 3: Stent A (simulating TristarTM stent) is constructed by cutting holes through the walls of a tube with a laser, and Stent B (simulating Wallstent®) is constructed by weaving wires in a crisscross pattern to form a mesh. Thus, Stent A struts have a rectangular cross section and Stent B have a circular cross section.   To improve the efficiency of porous stent, a new type of stent, asymmetric (patched) stent, was developed (Fig4). This stent consisted of a low porosity patch to block the aneurysmal inflow at aneurysm neck on a regular porous stent. A vein-pouch aneurysm model was created on the carotid artery of a canine to test the asymmetric stent. Angiographic images were obtained before and after stenting and compared with CFD results. In addition to both commercial stents and asymmetric stent, we designed a patient-specific stent with following criteria. · Block the strong impinging flow at proximal neck. · Interrupt the flow along the centerline of the vessel. · Allow the flow to peripheral vessels (perforators).

Fig. 4: An asymmetric (patched) stent with a 25% porous patch and a 80% porous stent. The low porous patch was clamped and welded on the surface of a stent.   3. CFD Analysis 3.1 Direct Stent Simulation The stented aneurysms were meshed using 0.6~2.4 million tetrahedral volume elements by ICEM-CFD (ICEM CFD Engineering, Berkeley, CA). Mesh independent flow in the models was obtained by increasing the number of meshes approximating the surface of the stents. A finite volume based CFD code, StarCD® (CD-adapco, Melville, NY), was used to solve the Navier-Stokes equations for incompressible flow. Central differencing scheme was selected to obtain the second order accuracy. The Reynolds number (Re) of blood flowing in a cerebral artery is low enough to be considered laminar. The Re at peak systole was 490 and the Womersley number of pulsatile wave was 2.38. Blood was assumed to be Newtonian because the shear rate in the artery was high and the diameter of the artery was large. The viscosity and the density of blood in all models was 3.5cP and 1056kg/m3. The aneurysm and vessel walls were assumed to be non-compliant and rigid as were assumed in other studies.   3.2 Asymmetric (Patched) Stent Simulation Direct simulation of patched stent is costly due to large number of stent wires. Therefore,  The low porosity patch was modeled by specifying pressure drop across the patch [3].

Results 1. Stent effects idealized aneurysm models

Fig. 5: (a) Peak systolic velocity contour on a center plane in the aneurysm. (b) Inflow velocity characteristics on a plane passing the aneurysm neck. Flow enters from left.

Fig. 6: Aneurysmal inflow rate variation in a curved vessel.

Aneurysmal inflow increased as vessel curvature increased. A stent interfered the inflow, however the stent effect diminished with increasing vessel curvature. The aneurysmal inflow rate in varying parent vessel curvature depended on stent design . Stent A in less curved vessel allows more flows to penetrate than stent B, but it is reversed in highly curved vessel.

2. Flow alternations by a stent in a patient-specific aneurysm

Fig. 7: Peak systolic velocity contours on a plane in the aneurysm. Flow comes into the aneurysm from bottom and goes out to the right.     3. Asymmetric stent effects onaneurysmal inflow The inflow entered the aneurysm through distal neck and created a counter clockwise circulation in the aneurysm before treatment. However, intra-aneurysmal flow became stagnant after the asymmetric stent treatment. Both CFD and angiographic images showed that the inflow was completely blocked at the aneurysm neck and the aneurysm was separated from the parent vessel by an asymmetric stent.   Fig. 8: CFD simulation and angiographic images of aneurysmal flow in a canine model.

Conclusions

· The design of porous stents affects aneurysmal flow modification.

· The effect of stent on aneurysm hemodynamics varies in curved vessel.

· Asymmetric (patched) stents may be a viable intervention for treating intracranial aneurysms.

· “Virtual intervention” can provide valuable clinical feedback in treatment planning.

 References

1. Geremia G, Haklin M, Brennecke L: Embolization of experimentally created aneurysms with intravascular stent devices. AJNR AN J Neuroradiol 15: 1223-1231, 1994.

2. Han PP, Albuquerque FC, Ponce FA, MacKay CI, Zabramski JM, Spetzler RF, McDougall CG: Percutaneous intracranial stent placement for aneurysms. J. Neurosurg 99:23-30, 2003.

3. Idelchik, IE: Handbook of hydraulic resistance. CRC Press, Boca Raton, FL, p. 522, 1994.

4. Krings T, et al.: Treatment of Experimentally Induced Aneurysms with Stents. J Neurosurg  56(6): 1347-59, 2005

5. Lanzino G, Wakhloo AK, Fessler RD, Hartney ML, Guterman LR, Hopkins LN: Efficacy and current limitations of intravascular stents for intracranial internal carotid, vertebral, and basilar artery aneurysms. J Neurosurg 91: 538-546, 1999.

Acknowlegements

We acknowledge our collaboration with the imaging and clinical groups.

· Imaging Group: Stephen Rudin, Ph.D., Kenneth Hoffmann, Ph.D., Ciprian Ionita, Ph.D., Rekha Tranquebar

· Clinical Group:  J. Yamamoto, M.D., Lee R. Guterman, Ph.D., M.D., L. N. Hopkins, M.D.

This research is supported by grants from NIH (1R01 EB002873, 1K25 NS047242), NSF (BES-0302389)