$276 Journal of Biomechanics 2006, Vol. 39 (Suppl 1)

6843 Mo, 15:00-15:15 (P11) Investigation of factors influencing the migration of stent-grafts

A. Gebert 1 , H. Imig 2, M. Morlock 1 . 1Biomechanics Section, Hamburg University of Technology, Germany, 2Department of General-, Vascular- and Visceral Surgery, AK-Hamburg-Harburg, Germany

Migration of stent-grafts is one of the major complications of endovascular aneurysm repair. Hypertension and large proximal diameters are speculated to be important factors for migration, but do not correlate with clinical data [1]. Mathematical models and experimental studies have been used to estimate static and pulsatile drag forces acting on stent grafts [2]. Neither of these approaches has been evaluated in a dynamic circulation test bench nor directly compared to clinical outcome. This study estimates migration parameters from static and dynamic measurements for one prosthesis type and compares them to clinically observed histories of the same stent graft. Static (n = 8) and dynamic (n = 3) distraction forces were determined for varying kinking angle, oversizing, proximal fixation length, and systemic flow and pressure parameters in a circulation test bench. Dynamic drag forces were calculated from pressure and flow rate using the principle of momentum. Distraction and drag forces were compared to clinical stent histories (up to 6 years) from a patient database (n=67) documenting hypertension and aneurysm morphology over time. Patient-specific 3D-models were created to investigate the relationship between implantation position and aneurysm morphology. Increasing kinking angle led to slightly increased longitudinal drag forces. This also resulted in stent migration at lower systemic pressures. Calculated longitudinal drag forces for a straight prosthesis were about 3% lower than statically measured distraction forces. This discrepancy increased with kinking angle, implying that transverse forces due to kinking may play an important role for migration. The same tendency was observed in the clinical cases. Higher systemic pressures increase drag forces but also result in increased fixation forces of similar magnitude, depending of the fixation length. Therefore, hypertension alone may not be a good indicator for migration. This study was supported by Aesculap AG & Co. KG.

References [1] Irwin et al, J. EndovascTher. 2002; 9. [2] Morris et al, J. Biomechanics. 2004; 37.

5257 Mo, 15:15-15:30 (P11) Effect of residual aneurysm pressure on the bifurcated stent-graft B.A. Howell 1 , T. Kim 2, A. Cheer 2, H.S. Dwyer 3, T.A.M. Chuter 2. 1Department of Vascular Surgery, UCSF, San Francisco, USA, 2Department of Mathematics, UC Davis, Davis, USA, 3Department of Engineering, UC Davis, Davis, USA

Objective: To calculate the effects of different sac pressures on the magni- tude, direction, location and pulsatile variation in the hemodynamic forces on bifurcated stent-grafts. Method: We used finite volume methods of computational fluid dynamics (CFD) to calculate the hemodynamic forces on 4 geometries, each derived from computed tomography (CT) of a bifurcated stent-graft within an abdominal aortic aneurysm (AAA). We used a pulsatile blood flow and blood pressure model and calculated forces over the cardiac cycle on the trunk, the bifurcation, the limbs, and the stent-graft as a whole. Each analysis was repeated using 4 published values of sac pressures, each representing a different clinical circumstance. Results: Pressure-related forces were far larger than flow-related forces for all 4 stent-grafts in all 4 conditions of sac pressure and were dependent on the relative area and orientation of stent-graft inlet and outlet. The net axial force was directed caudally. The addition of residual sac pressure had a large effect on net axial forces. Higher residual sac pressure reduced the net axial force on the stent-graft. The axially-directed systolic force at the bifurcation was 70% higher with the low sac pressures of a shrinking aneurysm than with high sac pressures of a dilating aneurysm. While the absolute effect of sac pressure reduction on diastolic forces was the same, the proportionate effect was even greater (570% increase). Conclusion: Sac pressure has a large enough effect on net hemodynamic forces to influence the risk of migration. Conditions that raise sac pressure, such as type II endoleak and endotension, may help stabilize an otherwise unstable stent-graft, and changes in a stent-graft that reduce aneurysm pres- sure and diameter may increase the risk of migration.

Oral Presentations

14.2. Cardiac Mechanics and Biology 6795 Th, 08:15-08:30 (P38) Differences in the reflection of pulse wave in the carotid artery between hypertensive and normotensive subjects M. Sugawara 1 , K. Niki 2, T. Okada 3, A. Harada 3. 1Himeji Dokkyo University, Himeji, Japan, 2Tokyo Women's Medical University, Tokyo, Japan, 3 Reseamh Laboratory, Aloka Co. Ltd, Tokyo, Japan

Background: It is generally accepted that nearly all tissues have an intrinsic ability to regulate the local blood flow, which is termed autoregulation. Autoreg- ulation in a region may be manifested as characteristics of pulse wave reflection from that region. There may be differences in the autoregulation, hence in pulse wave reflection, between hypertensive and normotensive subjects. Pulse wave reflection from the head remains poorly reported. Purpose: To characterize pulse wave reflection from the head in hypertensive and normotensive subjects. Methods: We examined the characteristics of carotid arterial wave inten- sity (Wl) noninvasively, using a specially designed combined Doppler and echo-tracking system in 74 hypertensive patients (HT) (systolic pressure >135mmHg or diastolic pressure >85mmHg) and in 59 normal subjects (Norm) (ages matched with HT). Wl is defined as the product of the time derivatives of blood pressure (P) and velocity (U): Wl=(dP/dt)(dU/dt). The negative area (NA) of Wl is attributed to reflections from the head. Results: NA increased with the maximum U in HT (goodness of fit: r 2 =0.22, p<0.0001), but was independent of it in Norm. Conclusion: The association of NA with the maximum U indicates more sensitive autoregulation of blood flow in hypertensive subjects.

6505 Th, 08:30-08:45 (P38) Finite element simulations of the cardiac cycle - comparisons with measured myocardial velocities E.W. Remme 1,2, E. Lyseggen 1 , M.P. Nash 2, O.A. Smiseth 1 . 1Department of Cardiology, Rikshospitalet University Hospital, Oslo, Norway, 2Bioengineering Institute, University of Auckland, Auckland, New Zealand

Measurements of left ventricular wall velocities consistently show unexplained biphasic waves (spikes) just before and after ejection. We hypothesize that the pre- and post-ejection velocity spikes are caused by shortening being interrupted by mitral valve closure and lengthening being interrupted by aortic valve closure, respectively. A finite element model of the ventricular myocardium was used to test our hypothesis. We simulated the cardiac cycle using realistic cavity pressure- volume loading conditions, and investigated the base-apex myocardial motion. If active contraction was elicited prior to mitral valve closure, an initial short- ening occurred that was ceased by the isovolumic cavity constraint imposed following mitral valve closure. This shortening resumed at the time of aortic valve opening. The result was a biphasic spike in the tissue velocity trace prior to the onset of ejection. At the end of systole, relaxation was initiated slightly prior to aortic valve closure. This resulted in a lengthening that was interrupted by aortic valve closure, but that continued at the time of mitral valve opening. The changing deformation pattern was seen as another spike in the tissue velocity trace following ejection. In animal experiments, we have measured and essentially eliminated the pre- or post-ejection tissue velocity spikes by implanting a stent over either the mitral or aortic valve, respectively, to prevent valve closure. In simulations, appropriate pressure-volume loading conditions were prescribed to reproduce either complete mitral or aortic regurgitation. These conditions eliminated the interruption of either the early systolic shortening, or the late systolic lengthening caused by mitral or aortic valve closure, respectively. Furthermore, the respective velocity spikes also disappeared, which is consistent with the experimental results. We conclude that the pre- and post-ejection tissue velocity spikes most likely reflect a changing deformation pattern caused by valve closure occurring shortly after the onset of contraction and relaxation, respectively.

4552 Th, 08:45-09:00 (P38) Cardiovascular rhythms and heart rate chaos L.-L. Shen 1 , D.-K. Tang 3, D.-A. Zheng 3, C.-S. Poon 4, G.-Q. Wu 2. 1DepL of Physiology, 2DepL of Mechanics, 3School of Information Science and Engineering, Fudan University, Shanghai, China, 4Harvard-MIT Division of Health Sciences and Technology, MIT, MA, USA

Conventionally, the implied hypothesis in hemodynamics, either in McDonald- Womersley theory [1] or in windkessel model, is that cardiovascular system beats with a purely periodic rhythm known as the heart rate. Thus, as long as cardiovascular behavior in one cardiac period is solved by bio-fluid dynamics, it can then be extrapolated infinitely. However recent development in