Geometric, Hemodynamic, and Pathological Study of a Distal Internal Carotid Artery Aneurysm Model in Dogs
Background and Purpose—We created a distal internal carotid artery side-wall aneurysm model in dogs and compared its geometric, hemodynamic, and histological similarities with human models.
Methods—Eight distal internal carotid artery–shaped devices were constructed using rapid prototyping, and 8 aneurysms were created via surgical reconstruction and elastase incubation. The geometric and hemodynamic parameters of the aneurysm and the parent artery of the dog and human models were compared, and histological response was evaluated at 12 weeks.
Results—Eight aneurysms were successfully created with good geometric simulation of the arteries between the dog and human models. Hemodynamic analysis revealed similar changes in the hemodynamic parameters both in the aneurysm sac and in the parent artery of the dog and human models. Histological analysis revealed internal elastic lamina discontinuity, elastic fiber disruption, a thinner muscular layer, increased smooth muscle cell proliferation rate, increased inflammation cell infiltration, and higher matrix metalloproteinase-2 and matrix metalloproteinase-9 expression indices in the medial aneurysm wall.
Conclusions—The distal internal carotid artery aneurysm model in dogs is feasible and exhibited considerable geometric, hemodynamic, and histological similarities with the original human models.
Cerebral aneurysms can cause lethal subarachnoid hemorrhage and occur more frequently in the distal internal carotid artery (ICA).1 However, the course of the distal ICA traverses a relatively compact region of complex anatomic surroundings to form 4 extremely tortuous natural bends, providing a barrier to intracranial interventions,2 and no existing distal ICA aneurysm model has precisely duplicated the geometry of the parent artery and aneurysm and simulated their hemodynamics. Such a model would provide an in vivo platform for studying the performance of flow-diverting devices or covered stents, with potential clinical applications in the near future.
We used a distal ICA-shaped device to reconstruct common carotid arteries in dogs and designed an elastase-induced aneurysm model at a specific site, and compared its geometric, hemodynamic, and histological similarities with the human model.
Material and Methods
Distal ICA Aneurysm Model and Geometric Analysis
A digital distal ICA tube was built using 3-dimensional (3D) digital subtraction angiography data from ruptured (Type A) and unruptured (Type B) postcommunicating artery aneurysms in 2 human subjects. A rapid prototyping printer (Eden350, Objet Geometries, Israel) was used to design a physical model of the distal ICA-shaped device (online-only Data Supplement). The surgical reconstruction of the bilateral common carotid arteries is described in Figure 1. The maximum aneurysm diameter (D); aneurysm height (H); aneurysm neck diameter (N); and 3 nondimensional geometric parameters H/N, D/H, and D/N, were measured as reported previously.3 The spatial angle of each reconstructed bend in the dog model was measured and compared with the human distal ICAs at the same view point.
Computational Fluid Dynamics Analysis
Computational fluid dynamics analysis of the parent artery and aneurysm sac, including the streamline, velocity field, arterial wall surface pressure field, wall shear stress, and wall shear stress grade, was performed 12 weeks after the surgery. The method is described in the online-only Data Supplement.
Reconstructed distal ICA aneurysms were harvested 12 weeks after surgery and the proximal common carotid arteries were harvested for the control, stained with hematoxylin and eosin; Masson’s trichrome; and elastic fiber stain (Van-Gieson); and immunostained for smooth muscle α-smooth muscle actin (α-SMA); antiproliferating cell nuclear antigen (PCNA); CD45; matrix metalloproteinase (MMP)-2; and MMP-9 (online-only Data Supplement). The smooth muscle cell (SMC) proliferation rate in the media and the thickness of the elastic layer and the α-SMA–positive SMC layer were measured. The inflammatory cell infiltration index was defined as the percentage of CD45-positive cells of the medial aneurysm wall. The MMP-2 and MMP-9 expression indices were defined as the percentage positive area in the media.
GraphPad Prism 5.0 software was used for statistical analysis. Data were expressed as mean±SD for continuous variables and as counts or proportions (%) for categorical variables. Fisher exact test was used to compare categorical data. Grouped t tests were used to compare the PCNA proliferation index, elastic layer and media α-SMA–positive layer thickness, inflammation cell infiltration index, and MMP-2 and MMP-9 expression indices. All the tests were 2-tailed. Statistical significance was defined as P<0.05.
We successfully created 8 distal ICA aneurysms in 8 dogs (Type A, n=4; Type B, n=4) by surgical reconstruction. At the 12-week angiographic follow-up, the aneurysms were well opened in 5 dog models, and partial thrombus of the aneurysm sac was observed in 1 Type A dog and 2 Type B dogs.
Geometric and Hemodynamic Comparison of the Parent Artery
The vasculature in the dogs simulated the complex geometry of the human distal ICA well, with <±5° error per bend comparison (Figure 2A; Table in the online-only Data Supplement). Cross-section measurement at each curvature indicated that the lesser curvature received the highest wall shear stress and the lowest blood flow velocity and vice versa.
Geometric and Hemodynamic Comparison of the Aneurysm
At the 12-week angiographic follow-up, geometric measurement revealed that the nondimensional parameters were more similar to the human models than dimensional parameters (Figure 2B; online-only Data Supplement).
Decreased wall shear stress and increased pressure were detected from the aneurysm neck to dome of both dog models, similar to the human models. Velocity field and streamline indicated a complex of turbulent blood flow in the aneurysm sac, with high blood flow velocity at the distal end of the neck and decreased velocity from the neck to dome. An inflow jet directly acting on the distal aneurismal wall was observed in 3 of 4 Type A dogs, similar to the corresponding human model. The wall shear stress grade showed a significant change observed in the aneurysm sac and was highest at the aneurysm dome or side-wall in all 8 models (Figure 2C and 2D).
Elastase treatment and consistent blood flow to the aneurysm wall induced internal elastic lamina discontinuity and elastic fiber disruption/insult in all the models, resulting in thinner walls compared with the control carotid artery (22.09±13.17 μm versus 112.2±14.68 μm; P<0.001).
The number of medial α-SMA–positive SMCs decreased, and the SMC layer of the elastase-treated aneurysm was thinner (32.61±9.26 μm versus 128.5±14.69 μm in control; P<0.001) after consistent complicated blood flow. PCNA staining revealed increased percentage of PCNA-positive medial SMCs in the elastase-treated aneurysm (41.32±8.92% versus 6.38±2.34% in the control; P<0.001).
Increased CD45-positive inflammatory cells infiltrated the media of the aneurysm at 12 weeks, at the infiltration index of 26.70±7.50% (versus 2.52±1.83% in the control; P<0.001). Both MMP-2 and MMP-9 were highly expressed in the media at 12 weeks after modeling, with expression indices were 14.27±4.51% (versus 0.33±0.24% in the control; P<0.001) and 16.03±6.14% (versus 0.30±0.17% in the control; P<0.001; Figure 3).
The complicated spacing angles of the 4 distal ICA curves in humans render it extremely difficult to obtain a precise geometric copy both in vitro and in vivo. Rapid prototyping is a convenient and precise method to fabricate a vessel model using 3D data, based on in vivo computed tomography, magnetic resonance, or digital subtraction angiography data.4
Nondimensional parameters are considered more important to compare the geometric similarity between aneurysms and to predict whether an aneurysm would rupture or be stable. Previous studies have suggested that ruptured aneurysms have higher H/N.3 We successfully simulated this H/N difference between the 2 dog models.
Studies have shown that wall shear stress is closely related to the risk of rupture and aneurysm growth, especially at low levels.5 Our results revealed that wall shear stress was lower in the aneurysm dome, which may explain why aneurysms tend to rupture more frequently there. In most situations, ruptured aneurysms usually have a narrower injector flow than unruptured ones, and the former has a 6.3× greater risk of rupturing than the latter.6 In this study, 3 of 4 Type A dog models simulated this hemodynamic characteristic well to the human model. The other one did not simulate the injector flow because of partial thrombus formation. Nevertheless, computational fluid dynamics only estimates true fluid dynamics by simulating the model in an ideal situation, such as assuming that the physical properties of the arterial wall (ie, density, elasticity, and deformation), blood flow (ie, components, density, and viscosity), and the fluid–solid interaction conditions are all the same. Thus, computational fluid dynamics calculations would only partially reflect the real situations between the dog and humans models.
Histological findings revealed internal elastic lamina discontinuity and elastic fiber disruption/insult of the aneurysm wall, with significant thinning of the smooth muscle layer in the test groups as illustrated by immunohistochemical staining for α-SMA, which is probably a type of arterial wall remodeling that may involve extracellular matrix protein synthesis and SMC apoptosis to counteract the mechanical forces induced by the altered blood flow.7 PCNA-positive cells showed increased proliferation in the artery, indicating that the residual SMCs became hypertrophic to compensate for the decreased number of SMCs. Inflammatory cells were found to infiltrate the media of the aneurysm walls; these results are consistent with observations in animal and human intracranial aneurysms,7–9 indicating that inflammatory cell infiltration plays a critical role in the enlargement of the aneurysms observed here. MMP-2 and MMP-9, which were expressed in the aneurysm wall, possess gelatinase and collagenase activity and can degenerate the important extracellular matrix components of aneurysm walls.7
Our model has limitations because it is difficult to duplicate the morphology precisely of a distal ICA aneurysm surgically and the geometric parameters of the created aneurysm could be affected by vessel diameter, ligation length of the distal arterial cecum, and incision width of common carotid arteries. Meanwhile, elastase damage to the arterial wall and consistent blood flow action could also affect aneurysm size. Importantly, elastase incubation would cause artificial changes to the aneurysm wall, which are similar to but not a real physiological condition of arterial wall degeneration/remodeling. Consequently, our aneurysm model could only partly mimic the clinical situation in humans.
Sources of Funding
This work was supported by China National Natural Science Foundation (No. 81000652), the Shanghai Venus Program Foundation (No. 11QA1405000), the Shanghai Natural Science Foundation (No. 12ZR1422500), and the Med-Technological Crossing Foundation (No. YG2011MS22)
The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.113.002290/-/DC1.
- Received May 25, 2013.
- Accepted June 26, 2013.
- © 2013 American Heart Association, Inc.
- Molyneux A,
- Kerr R,
- Stratton I,
- Sandercock P,
- Clarke M,
- Shrimpton J,
- et al
- Saatci I,
- Cekirge HS,
- Ozturk MH,
- Arat A,
- Ergungor F,
- Sekerci Z,
- et al
- Zeng Z,
- Kallmes DF,
- Durka MJ,
- Ding Y,
- Lewis D,
- Kadirvel R,
- et al
- Boussel L,
- Rayz V,
- McCulloch C,
- Martin A,
- Acevedo-Bolton G,
- Lawton M,
- et al
- Cebral JR,
- Castro MA,
- Burgess JE,
- Pergolizzi RS,
- Sheridan MJ,
- Putman CM
- Aoki T,
- Kataoka H,
- Moriwaki T,
- Nozaki K,
- Hashimoto N
- Norman PE,
- Powell JT
- Tulamo R,
- Frösen J,
- Hernesniemi J,
- Niemelä M