Thread 1. Computat ional Methods in Biomechanics and Mechanob io logy

7539 Th, 08:30-08:45 (P39) Deep wall in jury due to atherosclerot ic plaque reconf igurat ion dur ing ang iop las ty and in-stent restenosis R. Mongrain 1,2, R. Galaz 1,2, O. Bertrand 3. 1McGill University, Montreal, Canada, 2Montreal Heart Institute, Montreal, Canada, 3 HSpital Laval, Quebec, Canada

Balloon contact stress and stent strut geometry were linked to superficial vascular wall injury during angioplasty as a possible cause of in-stent resteno- sis [1]. Deep wall injury with rupture of the internal elastic lamina (IEL) was also proposed as a possible cause of in-stent restenosis [2] In this work, we propose an hypothesis of deep wall injury related to the reconfigurations of internal plaque constituents without any discharge in the lumen or rupture of the IEL. It is to be noted that such deep wall modifications would be undetectable with conventional clinical modalities. Previous studies have shown the direct relation between high stress concentrations on lipid pool boundaries and an increased susceptibility of plaque rupture at the associated locations [3]. Using 3D reconstructions from in vive intravascular ultrasound (IVUS) clinical data, we present a finite element analysis of the internal plaque structure in order to study the effect of internal plaque redistribution during angioplasty. In particular, we analyzed the morphologies of inclusions before and after angioplasty to investigate possible stress concentration factor on lipid pool for possible propensity toward extrusion and infiltration within the vascular cellular matrix without any discharge in the lumen. We performed internal stress analyses and found that certain inclusion reconfigurations could be responsible of such deep internal injuries that could play a role in in-stent restenosis. Such rearrangement of internal plaque constituents could also affect its global structural integrity (fatigue), lipid extrusion and infiltration may also lead to inflammatory processes.

References [1] Rogers et al. Circ. Res. 1999; 84(4): 378-83. [2] Sullivan et al. J. Vasc. Surg. 2002; 36(1): 143-9. [3] Lee et al. J. Am. Coll. Cardiol. 1993; 21: 777-82.

5297 Th, 08:45-09:00 (P39) Changes in the mechanical env i ronment o f les ions due to stent-artery interact ion - A computat ional analysis D.E. Kiousis 1 , T.C. Gasser 1 , G.A. Holzapfel 1,2. 1Royal Institute of Technology (KTH), School of Engineering Sciences, Stockholm, Sweden, 2 Graz University of Technology, Computational Biomechanics, Graz, Austria

Balloon angioplasty with stenting is a well-established interventional procedure, which aims at the increase of the blood flow through stenotic arteries. However, the related mechanical loading (and disruption) of the arterial tissues may lead to biological responses such as remodelling, neointimal hyperplasia and recoil. Nowadays, clinical studies have identified that device-dependent factors (such as stent design) have a significant impact on the stress-induced tissue growth and related restenosis. Numerical methods may provide deeper insights in the clinically important changes of the mechanical environment of atherosclerotic walls with respect to stent design. The present study aims to investigate the stress evolution in (three) dif- ferent tissue components of a patient-specific stenotic human iliac artery during balloon angioplasty with stenting. The constitutive model of each tissue component captures the observed anisotropy and nonlinearity under supra-physiological loading conditions. The related material parameters are determined from mechanical tests. The 3D balloon-stent-artery interaction problem is modeled by means of a novel contact algorithm based on smooth surface parameterizations, which prevents numerical instabilities of standard facet-based techniques. The computational analysis, based on different stent designs, focuses on the changes of the stress-strain distributions occurring during and after stenting.

5737 Th, 09:00-09:15 (P39) An invest igat ion into the effect o f stent st rut th ickness on restenosis using the f inite e lement method and val idat ion using an in-vitro compl iant ar tery model D. Toner, E Basir, C. Lally. School of Mechanical & Manufacturing Engineering, Dublin City University, Glasnevin, Dublin, Ireland

In-stent restenosis represents the major limitation for stenting procedures [1]. This study focuses on examining the role of strut thickness on in-stent resteno- sis. Finite element analyses have been carried out to investigate the difference in the mechanical loading induced in vessels stented with a thin and thick strut stent. An in-vitro rig is also being developed which can be used to validate the numerical models, thus accurately developing a preclinical testing tool which can be used to optimise stent designs. 3D numerical models of two stents of the same design but different strut thicknesses were developed and deployed in stenotic vessels. The material for the stent, 316L stainless steel, was described using an elasto-plastic material

T1.3 Mechanobiological Effects of Stent-Artery Interaction $403

model. Mooney-Rivlin hyperelastic equations were used to describe the non- linear stress-strain behaviour of the artery and the plaque, obtained from published test data [2,3]. Two different load cases were investigated; stent expansion to achieve (i) the same initial lumen diameter and (ii) the same final lumen diameter. Expansion of the stents to achieve the same initial lumen diameter resulted in considerably more tissue stressed at high levels in the vessel stented with the thicker strut stent, whereas expansion of the stents to achieve the same final lumen diameter resulted in more tissue stressed at high levels in the vessel stented with the thinner strut stent. These results indicate that there may be an optimum diameter for different stent designs whereby minimal stresses in the tissue can be achieved and thereby the stimulus for restenosis reduced. Work is currently being undertaken to validate the numerical models by expanding stents inside mock arteries manufactured from compliant Sylgard elastomer 184 with the deformation recorded using a video extensometer.

References [1] Colombo, A. et al. Selection of coronary stents. JACC 2002; 40: 1021-1033. [2] Lally, C., et al. Elastic behaviour of porcine coronary artery tissue under uniaxial

and equibiaxial tension. Ann Biomed Eng 2004; 32: 1355-1364. [3] Loree et al. Static and circumferential tangential modulus of human atheroscle-

rotic tissue. J Biomech 1994; 27: 195-204.

6693 Th, 09:15-09:30 (P39) Computat ional f luid dynamics est imates o f a l terat ions in wall shear stress in f luencing neoint imal hyperplasia after stent implantat ion

J.E LaDisa 1, L.E. Olson 2, D.C. Warltier 2,3, P.S. Pagel 2,3. 1Department of Pediatrics, Stanford University, 2Department of Biomedical Engineering, Marquette University, 3Department of Anesthsiology, Medical College of Wisconsin, Milwaukee, USA

Stent geometry is as an important predictor of neointimal hyperplasia(NH) and restenosis rates vary with stent design [1]. Stent deployment may also produce acute vascular damage. Recent data further indicates that wall shear stress(WSS) mediates cellular proliferation after vascular injury [2]. This series of investigations tested the general hypothesis that adverse distributions of WSS unique to the geometry of the stented portion of a vessel correlate with NH using computational fluid dynamics(CFD) simulations generated from two in vivo protocols. In one series of experiments, the iliac artery blood flow environment was reconstructed 14 and 21 days after stenting using 3D microfocal CT imaging, and WSS was determined using CFD. NH quantified using histology was localized to the stented area and occurred primarily in areas subjected to low WSS. Time-dependent increases in NH produced compensatory changes in vascular geometry and associated distributions of WSS, leading to progressive elimination of WSS disparity within the stented region. Additional CFD models demonstrated that stent design properties, including deployment diameter, length, scaffolding and strut number, width, and thickness, also produced markedly different patterns of WSS within the stented region. Most recently, we tested the hypothesis that the ability of a stent to conform to the theoretical curvature of a coronary artery influences WSS indices using CFD models generated from measured properties of canine coronary arteries in vivo. Distributions of WSS in curved arteries were sub- stantially different than those observed using cylindrical models. Straightening rather than conforming to the artery curvature introduced WSS alterations that were most pronounced in the proximal and distal portions of the stent. These results agree with clinically reported areas of restenosis [3]. The results of our investigations suggest that the flow environment created by an implanted stent temporally modulates NH by influencing indices of WSS in a selective yet predictable manner.

References [1] McLean DR, Eigler NL. Reviews in Cardiovascular Medicine 2002; 3(Suppl 5):

$16-22. [2] Liu SQ, Goldman J. IEEE Transactions on Biomedical Engineering 2001; 48:

474-483. [3] Wentzel J J, et al. Circulation 2001; 103: 1740-1745.

6797 Th, 09:30-09:45 (P39) Computat ional analysis o f the per formance o f drug-elut ing stents

T. Seo 1,2, A.I. Barakat 1 . 1Department of Mechanical and Aeronautical Engineering, University of California, Davis, USA, 2School of Mechanical Engineering, Andong National University, Andong, Korea

The use of drug-eluting stents has dramatically reduced the incidence of restenosis; however, much remains to be learned about the performance of these stents. We had previously shown that stent design has a pronounced impact on the local flow field in the vicinity of the stent and that the local flow patterns regulate the rate of endothelial cell wound healing following mechanical denudation. In the present study, we tested the hypothesis that the design of drug-eluting stents influences the efficacy of local drug delivery to

$404 Journal of Biomechanics 2006, Vol. 39 (Suppl 1) Oral Presentations

the arterial wall and that this effect depends on both arterial geometry and the prevailing flow conditions. We performed computational simulations in which the coupled Navier-Stokes and advection-diffusion equations were solved to determine the flow field and drug concentration in the vicinity of model drug- eluting stents positioned within either straight or curved arterial segments. Steady and pulsatile flow simulations with Reynolds numbers ranging from 200400 were performed, and stent designs ranging from a simple series of wires to more realistic configurations consisting of multiple intertwined rings were tested. Drug elution was modeled as a constant source emitted from the entire body of the stent. The results demonstrate that for steady flow in straight arterial segments within which stents idealized as a series of wires are deployed, the in-stent drug concentration at the arterial wall increases with stent wire thickness. In curved vessels, drug concentration at the wall de- creases as Reynolds number and vessel curvature increase. In pulsatile flow, the in-stent drug concentrations are lower during flow acceleration than during deceleration. Results using stents modeled as intertwined rings reveal intricate flow field and drug concentration sensitivity to the details of stent design. The present results enhance our understanding of the performance of drug-eluting stents and point to design strategies for optimizing this performance.

6020 Th, 09:45-10:00 (P39) A sequential porohyperelastic-transport framework for simulating drug-eluting stents

P.H. Feenstra 1 , C.A. Taylor 2. 1Mechanical Engineering, Stanford University, Stanford, USA, 2Bioengineering and Surgery, Stanford University, Stanford, USA

Drug-eluting stents (DES) have had a significant impact on the treatment of coronary artery disease by minimizing in-stent restenosis. Due to the complexity of drug transport in diseased coronary arteries, simulation tools have important applications in improving the design of these complex devices. Compared to conventional stents, DES have a dual purpose: the first is mechanical, to provide support to the injured vessel; the second is therapeutic, to deliver drug locally to the vessel wall. The objective of this paper is to present a simulation framework for DES that takes the dual purpose into consideration. Stent deployment is simulated first, a mechanical porohyperelastic (PHE) contact analysis. This analysis calculates the interstitial fluid velocity as the result of pore pressure gradients and me- chanical deformations of the vessel wall. The deformed geometry of stent and vessel, interstitial pore fluid velocity field, and porosity field are extracted and used as input for the drug release simulation: a reaction-advection-diffusion (PAD) transport analysis calculating the spatial and temporal drug distribution. A novel feature of the framework is that stent deployment is simulated prior to simulating drug release. The advantage of this approach is that the deformed geometry and interstitial fluid velocity field are not assumed a-priori, but are actually calculated using a stent deployment simulation. The framework is demonstrated simulating a DES in an idealized, 3D vessel. Varying mechanical and transport properties based on literature review are assigned to each of the three layers in the wall and Kedem-Katchalsky conditions are explicitly modeled to account for varying porosity of the three layers. The stent deployment simulation results in a realistic pressure-radial expansion relationship of the stent and a realistic deformed geometry. The results of the drug release simulation for a period of 13 weeks show that a relative constant concentration field can be sustained. Future work includes simulating drug release with different parameters, and applying the framework to specific stent designs and realistic, image-based vessel geometries.

T1.4 Computational Modelling and Mechanobiology of Vascular Anastomosis 4884 Th, 11:00-11:15 (P42) Transmural stress during bypass surgery: A patient-specific computational analysis

F. Cacho 1,2, M. Doblar 2, G.A. Holzapfel 1,3. 1Graz University of Technology, Computational Biomechanics, Graz, Austria, 2 University of Zaragoza, Group of Structural Mechanics and Materials Modeling (GEMM), Aragon Institute of Engineering Research (13A), Zaragoza, Spain, 3Royal Institute of Technology, School of Engineering Sciences, Stockholm, Sweden

Vascular mechanics at branching locations and junctions such as the carotid sinus or bypass anastomoses have been traditionally studied by researches in the field of fluid mechanics. They mainly focused on the analysis of the complex blood flow patterns and related their alterations to clinical reports (for a recent review see [1]). These studies provide significant contributions for the explanation of, for example, certain long-term anatomical alterations observed in bypassed vessels. Frequently, in these studies the arterial wall is assumed to be rigid. The most sophisticated simulations consider (only) elastic arteries with (too) simplified constitutive models. Given the highly nonlinear behavior

of vascular tissue, and specific cases for bypass geometries in which small geometrical changes may lead to significant flow alterations, it seems to be important to consider appropriate constitutive descriptors for the arterial walls. In this work, we present the simulation of a coronary artery bypass procedure with the focus on the wall mechanics. We consider both the native artery and the venous graft. The complex, multilayered vascular morphology of an individual patient is reconstructed, and the corresponding material behavior is imposed by means of an appropriate constitutive model. The related material parameters are obtained by fitting the model to the experimental data of the (same) specimen from which the geometry was taken. Our results suggest that the transmural stresses depend heavily on the incision length, which is related to the graft insertion angle. In addition, the graft undergoes large displacements and is subject to high stresses under arterial flow conditions at sites where intimal thickening has been reported in the medical literature. Acknowledgement: Financial support for this research was partly provided by the Cooperative Research EU-Project DISHEART; Call Identifier: FP6-2002- SME-I.

References [1] E Migliavacca, G. Dubini. Computational modeling of vascular anastomoses.

Biomech Model Mechanobiol 2005; 3: 235-250.

6261 Th, 11 : 15-11:30 (P42) That the end-to-end anastomosis is superior to the end-to-side anastomosis for peripheral bypass surgery: Observations from computational studies T.P. O'Brien 1 , M.T. Walsh 1 , P. Grace 2, P.D. Devereux 1 , S.M. O'Callaghan 1 , P. Burke 2, T.M. McGIoughlin 8 . 1 Centre forApplied Biomechanical Engineering Research and Materials and Surface Science Institute, University of Limerick, Limerick, Ireland, 2Department of Vascular Surgery, Mid-Western Regional Hospital Limerick, Ireland

Introduction: The end-to-side anastomosis has traditionally been the standard choice amongst surgeons implanting peripheral artery bypass grafts. However, the end-to-side bypass graft has been characterised by poor long-term patency rates resulting in ambitious surgical endeavours to buffer graft/artery material mismatch and attempts to reduce hemodynamic abnormality at the junction. Research to date has suggested that the end-to-end anastomosis may offer significant hemodynamic and wall stress advantages, albeit with a loss in function due to elimination of the host artery proximal to the anastomosis. Materials and Methods: The objective of this study was to assess the performance advantages of the end-to-end anastomosis over that of the end- to-side anastomosis and to develop a surgical technique which would enable end-to-end suturing without loss of function. Studies of both anastomoses were performed using computational fluid dynamics (CFD) and finite element analysis (FEA). The results of these studies were then used to specify design criteria for a novel bypass graft. This novel design was then tested using the same methods to ensure that the desired performance was achieved. Results: Fluid induced wall shear stresses and pressure induced wall stresses were the primary results of interest. The end-to-side anastomosis was found to produce more abnormal hemodynamic patterns, wall shear stresses and wall stresses than the end-to-end anastomosis. Summary results from an in vive study demonstrate how the hemodynamics and wall stresses contribute to dis- ease development in the end-to-side anastomosis and how these contributory factors may be alleviated using the new bypass procedure. Conclusion: The end-to-end anastomosis offers significant hemodynamic and wall stress advantages over the conventional end-to-side anastomosis and it is possible to enable its implementation in a vascular bypass graft without loss of function.

5963 Th, 11:30-11:45 (P42) Non-Newtonian effects of pulsatile blood flow in non-planar distal end-to-side anastomosis W. Wang, Q. Wang. Medical Engineering Division, Department ef Engineering, Queen Mary, University of London, London, UK

A number of factors are known to affect flow patterns and wall shear stress (WSS) distribution in blood vessels. They include the geometry of the vessel, rheological properties of the blood and pulsatility of the circulation. In this study, non-Newtonian effects of blood are examined in a distal end-to-side anastomosis. Shear-thinning behaviour of the blood is depicted by a modified Power-Law (PL) and Carreau-Yasuda models (CY). To explore the effect of non-planarity in vessel geometry, a curved bypass graft is connected to a host tube with out-of-plane layout [1,2]. Pulsatile flow waveform in the femoral artery is applied [3]. Flow velocities, WSS and oscillatory shear index (OSI) are calculated using Fluent and softwares developed in house. Results are given at times of systolic acceleration, peak systole, systolic deceleration, end systole and peak diastole. Effects of rheological properties of the blood is analysed by comparing results to those of a Newtonian flow. Non-planarity serves to