Hemodynamic differences by increasing low profile visualized intraluminal support (LVIS) stent local compaction across intracranial aneurysm orifice

Abstract

Background

The Low-profile Visualized Intraluminal Support device (LVIS) has been successfully used to treat cerebral aneurysm, and the push-pull technique has been used clinically to compact the stent across aneurysm orifice. Our aim was to exhibit the hemodynamic effect of the compacted LVIS stent.

Methods

Two patient-specific aneurysm models were constructed from three-dimensional angiographic images. The uniform LVIS stent, compacted LVIS and Pipeline Embolization Device (PED) with or without coil embolization were virtually deployed into aneurysm models to perform hemodynamic analysis. Intra-aneurysmal flow parameters were calculated to assess hemodynamic differences among different models.

Results

The compacted LVIS had the highest metal coverage across the aneurysm orifice (case 1, 46.37%; case 2, 67.01%). However, the PED achieved the highest pore density (case 1, 19.56 pores/mm²; case 2, 18.07 pores/mm²). The compacted LVIS produced a much higher intra-aneurysmal flow reduction than the uniform LVIS. The PED showed a higher intra-aneurysmal flow reduction than the compacted LVIS in case 1, but the results were comparable in case 2. After stent placement, the intra-aneurysmal flow was further reduced as subsequent coil embolization. The compacted LVIS stent with coils produced a similar reduction in intra-aneurysmal flow to that of the PED.

Conclusions

The combined characteristics of stent metal coverage and pore density should be considered when assessing the flow diversion effects of stents. More intra-aneurysmal flow reductions could be introduced by compacted LVIS stent than the uniform one. Compared with PED, compacted LVIS stent may exhibit a flow-diverting effect comparable to that of the PED.

Introduction

Stent-assisted coiling has been widely used as a safe and effective treatment for wide-neck and complex intracranial aneurysms. A stent can not only prevent coils protrusion into the parent artery, but can also produce hemodynamic effects on the cerebral aneurysm that promote aneurysm occlusion.¹ Previous studies have demonstrated that the hemodynamic effect of a stent may be associated with its mesh characteristics.^2,3

The Low-profile Visualized Intraluminal Support device (LVIS; MicroVention-Terumo, Tustin, CA, USA) is a novel, self-expandable, braided stent with greater metal coverage (23%) and a smaller cell size (∼0.9 mm) than other coil-assist stents.^3,4 Because it is braided, the wires of the LVIS stent can slide over one another and generate a spatially varying mesh density, which can increase the mesh density of the LVIS stent at the aneurysm orifice during deployment using the push-pull technique.^5,6 However, no studies have evaluated the hemodynamic effects of the compacted LVIS stent on the aneurysm, and it is unknown whether a compacted LVIS stent can provide a similar hemodynamic effect to a flow diverter. Therefore, in this study, we present a quantitative hemodynamic analysis of the compacted LVIS stent compared with uniformly deployed LVIS stent and flow diverter. Moreover, the finite element method (FEM) technique was applied to simulate the stent and coils in this study. The FEM technique can be used to describe the mechanical properties and provide a higher-fidelity model for accurate post-treatment computational fluid dynamics (CFD) analysis.⁷ This study was a proof-of-concept study for the hemodynamic effect of the LVIS compaction at the aneurysm neck and we hope the findings of this study would provide some beneficial references for physicians.

Methods

This retrospective study was approved by the institutional review board of our hospital. Informed consent was obtained from each study patient or their relatives, and all data were collected anonymously.

Patient population

We chose two representative patient-specific intracranial aneurysms for this computational study (case 1 and case 2). The patient in case 1 presented with transient dizziness. Magnetic resonance angiography was then performed, which showed the aneurysm. The patient in case 2 suffered a subarachnoid hemorrhage, and digital subtraction angiography showed the aneurysm. The two patient-specific, wide-necked internal carotid artery aneurysm models were constructed using three-dimensional (3D) digital subtraction angiography images (Figure 1(a) and (b)). Case 1 involved an aneurysm with a longitudinal diameter of 8.13 mm and a neck width of 4.32 mm. Case 2 involved an aneurysm with a longitudinal diameter of 8.28 mm and a neck width of 7.23 mm. Both aneurysms were treated by uniformly LVIS stent-assisted coiling (Figure 1(c)). The proximal and distal diameter of parent vessel where the stent deployed was 3.69 mm and 3.38 mm for case 1, and 3.54 mm and 3.25 mm for case 2. The diameters of deployed LVIS were 3.5 mm–15mm in both cases. The packing density was 28.06% in case 1 and 28.46% in case 2.

Figure 1.

Computational fluid dynamics model. (a) Three-dimensional digital subtraction angiography images of the aneurysm. (b) A patient-specific aneurysm model was constructed from the three-dimensional digital subtraction angiography images. (c) The aneurysm was treated by LVIS stent-assisted coiling. (d) The stent was virtually deployed in the parent vessel of the aneurysm model and the aneurysm sac was virtually filled with coils using the finite element method.

Stent and coil modeling

The LVIS stent and Pipeline Embolization Device (PED) (Coviden/ev3 Neurovascular, Irvine, CA, USA) were simulated by the FEM technique in the current study. The FEM-based workflow for stent deployment modeling was completed using ABAQUS/Explicit v6.14 (SIMULIA, Providence, RI).^7,8 Briefly, the stent simulation workflow consisted of three steps: crimping, delivery and deployment.⁹ Firstly, the stent was virtually crimped and fit for use in a microcatheter mode, which was modeled by rebar-reinforced shell elements with polymer-like elastic material. Then, the stent was delivered within the microcatheter along the pathway to the target region. The delivery pathway was constructed using the central points of the cross-sections of the blood vessel. When the stent within the microcatheter was delivered across the aneurysm orifice to the target region, the stent began to release virtually. Desired deployment outcomes could be created during the virtual stent release. To generate a uniform stent, the microcatheter was retracted proximally while holding the pusher wire still. If the microcatheter was retracted proximally while advancing the proximal pusher wire distally over the aneurysm orifice, the compacted stent could be produced.¹⁰ In our study, stent compaction was maximized during stent deployment by controlling the movement of the microcatheter and the proximal pusher wire. The distal landing zones for the LVIS and PED were determined according to the location where the LVIS distal end landed in the angiographic images in clinical treatment. Finally, the deployed stent was swept into 3D solid geometry for CFD analysis (Figure 1(d)). The metal coverage and pore density of the stent were calculated. The metal coverage was determined as the ratio of the surface area covered by metal to the total surface area of the stent, and pore density was calculated as the number of cells per unit area of the stent.

Similarly, the coils simulation was also implemented using the general-purpose FEM software ABAQUS 6.14 in Abaqus/Explicit mode.^7,11 Two types of coils were used: helical coil and frame coils. Firstly, the centerline of the coil was extracted in MATLAB (MathWorks, Natwick, MA) and then imported into the FEM code ABAQUS to generate different types of coils. Then, the coils were packaged into a virtual microcatheter to be deployed into the aneurysm sac. Finally, the coils were swept to 3D solid model using Abaqus/CAE according to the actual diameter of the coils.

CFD modeling and hemodynamic analysis

In this study, we not only remodeled the uniformly deployed LVIS stent according to the clinical treatment, but also remodeled the compacted LVIS and PED in the same aneurysm geometry for the further comparison. Seven CFD models were simulated for each patient-specific aneurysm model: pretreatment, uniform LVIS stent (Figure 2(a)), compacted LVIS stent (Figure 2(b)), PED (Figure 2(c)), uniform LVIS stent with coil embolization, compacted LVIS stent with coil embolization and PED with coil embolization. CFD simulations were performed as described previously.^12,13 The surface geometry models (aneurysms merged with the stent and coiling) were imported into the mesh generation software (ICEM CFD, V.14.0; ANSYS Inc, Canonsburg, Pennsylvania, USA) to create finite-volume tetrahedral elements. The largest element was 0.2 mm, and the element size of the stent and coils was set to 0.02 mm for adequate representation of the geometry. After meshing, CFX V.14.0 software (ANSYS, Inc.) was used to simulate blood hemodynamics. The governing equations underlying the calculation were the Navier-Stokes equations under homogenous, incompressible, laminar and Newtonian fluid condition with a density of 1056 kg/m³ and a viscosity of 0.0035 N·s/m². The blood vessel wall was assumed to be rigid with no-slip boundary conditions. A pulsatile velocity profile obtained by transcranial Doppler in a normal subject was used as the inflow boundary condition. The flow waveforms were scaled to achieve a mean internal carotid artery inlet flow rate of 4.6 ml/s under pulsatile conditions. Three cardiac cycle simulations were performed for numerical stability and the last cardiac cycle was recorded.

Figure 2.

Stent models. (a) uniform LVIS stent, (b) compacted LVIS stent, (c) the Pipeline Embolization Device.

The following indexes were quantitatively demonstrated to characterize the aneurysm flow conditions at the systolic peak: wall shear stress (WSS), velocity, high flow volume (HFV) and the streamlines. The WSS is the tangential drag force per unit area of the endothelial surface. WSS was reported as the averaged value of the entire aneurysm dome. The HFV represents the fluid domains with a velocity >0.2 m/s and could influence the thrombosis process in the aneurysm sac, thus predicting aneurysm recurrence after endovascular treatment.¹² The quantitative results of aneurysmal hemodynamics analysis (WSS, velocity and HFV) of the six treatment models were compared with those of the pretreatment model in each case.

Results

The compacted LVIS stent outperformed the uniform LVIS stent in aneurysmal flow reduction

For case 1 (Figures 3 and 4), the uniform LVIS stent showed a relatively obvious intra-aneurysmal flow reduction in WSS (20%), flow velocity (37%) and HFV (16%), and the compacted LVIS stent produced a more obvious intra-aneurysmal flow reduction (WSS 61%, flow velocity 69% and HFV 87%) than the uniform LVIS stent. Similarly, the compacted LVIS stent with coil embolization presented greater intra-aneurysmal flow reduction than that of the uniform LVIS stent with coil embolization (WSS: 85% vs. 69%; flow velocity: 92% vs. 83% and HFV: 99.5% vs. 93%).

Figure 3.

Visualized results of numerical simulations (streamline, high flow volume and WSS) in case 1.

Figure 4.

Percentage of hemodynamic changes (velocity, high flow volume and WSS) for the six endovascular treatment models compared with the pretreatment model for case 1.

For case 2 (Figures 5 and 6), the uniform LVIS stent produced a small intra-aneurysmal flow reduction (WSS 13%, velocity 18% and HFV 20%). Compared with the uniform LVIS stent, the compacted LVIS stent resulted in a greater intra-aneurysmal flow reduction (WSS 37%, velocity 47% and HFV 75%). Meanwhile, the compacted LVIS stent with coil embolization showed a greater intra-aneurysmal flow reduction than that produced by the uniform LVIS stent with coil embolization (WSS: 56% vs. 52%; velocity: 82% vs. 77%; HFV 94% vs. 89%).

Figure 5.

Visualized results of numerical simulations (streamline, high flow volume and WSS) in case 2.

Figure 6.

Percentage of hemodynamic changes (velocity, high flow volume and WSS) for the six endovascular treatment models compared with the pretreatment model for case 2.

The flow diversion effect of the stent was determined by the combination of stent metal coverage and pore density

Among these stent models, the compacted LVIS stent had the highest metal coverage across the aneurysm orifice (case 1, 46.37%; case 2, 67.01%), while the uniform LVIS stent had the lowest metal coverage across the aneurysm orifice (case 1, 23.31%; case 2, 22.77%). The rate of metal coverage of the PED for case 1 and case 2 were 37.65% and 30.50%, respectively. However, the PED achieved the highest pore density for both case 1 (19.56 pores/mm²) and case 2 (18.07 pores/mm²) compared with the uniform LVIS stent (case 1, 2.12 pores/mm²; case 2, 2.67 pores/mm²) and the compacted LVIS stent (case 1, 5.54 pores/mm²; case 2, 5.61 pores/mm²). Meanwhile, the PED produced a greater intra-aneurysmal flow reduction for case 1 (WSS 13%, velocity 18% and HFV 20%), than the uniform LVIS stent and the compacted LVIS stent. However, in case 2, the compacted LVIS stent achieved an intra-aneurysmal flow reduction comparable to that of the PED (WSS: 37% vs. 49%; velocity: 43% vs. 44% and HFV 75% vs. 61%).

Subsequent coil embolization further reduced the intra-aneurysmal flow after stent placement

The PED with coil embolization showed the most evident intra-aneurysmal flow reduction for case 1 (WSS 87%, velocity 94% and HFV 99.8%) and case 2 (WSS 64%, velocity 83% and HFV 95%). Meanwhile, the compacted LVIS stent with coil embolization resulted in an intra-aneurysmal flow reduction comparable to that of the PED for case 1 (WSS: 85% vs. 81%; velocity: 92% vs. 81% and HFV 99.5% vs. 97%) and greater intra-aneurysmal flow reduction than that of the PED for case 2 (WSS: 56% vs. 49%; velocity: 82% vs. 44% and HFV 94% vs. 61%).

Discussion

The LVIS stent has been successfully used to treat intracranial aneurysms, exhibiting a high complete occlusion rate, and the push-pull technique has been used clinically to generate a compacted stent with higher metal coverage across the aneurysm orifice.^5,6 However, less attention has been paid to the hemodynamic effect of the compacted LVIS stent on intracranial aneurysms. In this study, the FEM technique and patient-specific aneurysm models were applied to simulate seven CFD models to quantitatively analyze the hemodynamic differences among the compacted LVIS stent, uniform LVIS stent and PED. An impressive finding was that the flow-diversion effect of the stent was associated with the combination of stent metal coverage and pore density. The compacted LVIS stent produced a greater intra-aneurysmal flow reduction than the uniform LVIS stent, but the results were comparable to those of the PED. Moreover, subsequent coil embolization further reduced the intra-aneurysmal flow after stent placement, and the compacted LVIS stent with coil embolization produced a similar intra-aneurysmal flow reduction to that of the PED. To our knowledge, the present study provides the first hemodynamic data for the compacted LVIS stent and could provide some reference information for physicians regarding how stents can be best used to treat cerebral aneurysms.

The hemodynamic effect plays an important role in the aneurysm recanalization.^14,15 Several previous studies have reported that stents impact intra-aneurysmal hemodynamics because of the flow-diverting effect, and the flow-diverting effect of a stent is correlated with the stent metal coverage and pore density.^9,16 Higher metal coverage could be more effective in flow diversion, which could promote intra-aneurysmal flow stagnation and thrombosis, thus increasing aneurysm occlusion rates.² However, the pore density could also be a critical factor in the flow diversion effect. Sadasivan et al indicated that a stent with less metal coverage and greater pore density could still have a considerable effect on blood flow, which verified that pore density plays an important role in the flow diversion effect.¹⁷ In the current study, the compacted LVIS stent had comparable pore density in case 1 and case 2 (5.54 pores/mm² and 5.61 pores/mm², respectively), but the compacted LVIS stent had higher metal coverage in case 2 than in case 1 (67.01% vs. 46.37%). This may be the reason why the compacted LVIS stent exhibited a smaller intra-aneurysmal flow reduction than the PED in the case 1. However, the compacted LVIS stent achieved an intra-aneurysmal flow reduction comparable to that of the PED in case 2, likely because, in case 2, the combination of the higher metal coverage (67.01%) and lower pore density (5.61pores/mm²) of the compacted LVIS stent had a more profound effect than the combination of the low metal coverage (30.50%) and high pore density (18.07 pores/mm²) for the PED. Compared with the conventional coil assist stents, the LVIS stent has higher metal coverage and pore density.^18,19 Thus, the LVIS stent has a hemodynamic advantage over conventional coil assist stents. Wang et al. reported that a single LVIS stent could produce a greater flow reductions than the double-Enterprise stent.³ Dholakia et al. studied the hemodynamic effect of five different stents using contrast concentration-time curves within the aneurysm, and also found that the LVIS stent performed better than the conventional stents (Neuroform and Enterprise stents) regarding the flow diversion effect.²⁰ Li et al analyzed the hemodynamic changes of 87 medium-sized intracranial aneurysms after stent-assisted coiling and found greater reductions in velocity and WSS in aneurysms treated with LVIS stents than in aneurysms treated with Enterprise stent.²¹ This may be the reason why LVIS stents achieved greater complete or near-complete occlusion rates than the Enterprise stents.²² However, those studies did not analyze the hemodynamic changes of aneurysms after the LVIS stent deployment using the push-pull technique. Moreover, no studies have compared the hemodynamic changes between the compacted LVIS stent and flow diverter.

Theoretically, the compacted LVIS stent could promote intracranial aneurysm occlusion. First, placement of the compacted LVIS stent across the aneurysm orifice could cause it to bulge into the aneurysm sac, thus providing more support for coil embolization and increasing the packing density.²³ Moreover, the dense stent mesh could enhance the flow reconstruction capacity and provide a scaffold for the neointima formation, thus promoting rapid intra-aneurysmal thrombosis and excluding the aneurysm from the circulation.^17,24 In the current study, the results further confirm that the compacted LVIS stent presented higher metal coverage and pore density than the uniform LVIS stent. The LVIS stent could reduce the intra-aneurysmal flow, and the compacted LVIS stent produced a greater intra-aneurysmal flow reduction than the uniform LVIS stent.

Currently, flow diverters, characterized by high metal coverage (approximately 30%-35%), have emerged as a promising option for complex and large aneurysms and produce a higher aneurysmal complete or near-complete occlusion rates than previous intracranial stents.^25,26 Compared with the coil assist stents, flow diverter could also significantly reduce aneurysmal inflow according to the WSS and velocity.^2,3,27 However, because of the increased mesh density in the parent artery, a flow diverter could increase the likelihood of perforator occlusion, resulting in an ischemic event, especially in territories with high densities of perforators.^28,29 Therefore, the LVIS stent has a possible advantage over the flow diverters in peri-aneurysmal regions rich with perforating arteries. We could produce higher metal coverage across the aneurysm orifice to promote aneurysm occlusion and lower metal coverage at perforator regions to protect the perforating arteries using the push-pull technique during the LVIS deployment.⁹ Moreover, the findings of this study showed that the compacted LVIS stent could achieve an intra-aneurysmal flow reduction comparable to that of the PED for some cases, and the compacted LVIS stent with coil embolization could result in an intra-aneurysmal flow reduction comparable to that of the PED. Therefore, the compacted LVIS stent with or without coil embolization may be more appropriate for some selective cases.

This study had some limitations. The sample size of our study was small. It was hardly to make any solid conclusions based on 2 samples only. Although, this study was a proof-of-concept study for the hemodynamic effect of the LVIS compaction at the aneurysm neck, the conclusion should be still verified by further studies with larger sample sizes. It is necessary to explore the hemodynamic effect of compacted LVIS configurations on perforator flow adjacent to the aneurysm with more and proper selected cases in the future. Furthermore, several assumptions, such as a rigid wall, laminar flow, Newtonian blood, and physiological but not patient-specific flow-boundary conditions, were included. In our study, we did not differentiate the tantalum marker wires from the nitinol wires.

Conclusion

The combined characteristics of stent metal coverage and pore density should be considered when assessing the flow diversion effects of stents. More intra-aneurysmal flow reductions could be introduced by compacted LVIS stent than the uniform one. Compared with PED, compacted LVIS stent may exhibit a flow-diverting effect comparable to that of the PED. The further study with larger sample size is needed to provide more evidence.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by National Natural Science Foundation of China (grant numbers: 81220108007, 81801156, 81801158, 81471167, 81671139 and 81960330) and Capital's Funds for Health Improvement and Research (grant number: 2018-4-1077). Beijing Hospitals Authority Youth Programme (code: QML20190503), Beijing Neurosurgical Institute Youth Innovation Fund (grant number: BNI-2019001, BNI-2019002, BNI-2019003, BNI-2019004).

ORCID iDs

Zhongbin Tian https://orcid.org/0000-0002-3635-8915

Xinjian Yang https://orcid.org/0000-0001-7306-0125

Jian Liu https://orcid.org/0000-0002-5454-2847