Haemodynamic Analysis of Single And Double Bare-metal Stents For Coronary Artery Aneurysm


 Background The aim of this study was to investigate a novel method to treat coronary artery aneurysms(CAAs).Methods Three CAAs patients who underwent single or double bare-metal stent implantation were recruited. The CAAs and parent artery diameters were measured for model construction. Single- and double-stent implantation were simulated, and the changes in the CAAs haemodynamics after stenting were analysed. Results In Case 1, the flow velocity in the single-stent model was significantly lower than that in the double-stent model (0.0046±0.013 vs 0.0050±0.011, p<0.001), while the pressure (127.33±18.03 vs 121.19±26.92, p<0.001) and wall shear stress (WSS, 0.28±1.19 vs 0.22±1.13, p<0.001) were significantly higher. In Case 2, the flow velocity (0.005±0.011 vs 0.007±0.01, p<0.001) and WSS (0.17±0.82 vs 0.23±0.88, p<0.001) in the double-stent model were significantly lower than those in the single-stent model, while the pressure was significantly higher(117.70±10.07 vs 110.64±6.34, p<0.001). The same tendency was also observed in Case 3. All CAAs occluded during the follow-up period without obvious in-stent restenosis.Conclusion Application of the single or double bare-metal stent technique, according to the neck diameter of the coronary artery aneurysms, can effectively change the flow in vessels and aneurysm haemodynamics to achieve occlusion.Trail registration: Clinicaltrails, NCT04265989. Registered 12 February 2020, https://clinicaltrials.gov/ct2/show/NCT04265989?term=NCT04265989&draw=2&rank=1

therapeutic regimen relies on doctors' experience. Treatments include conservation therapy, interventional therapy and surgery. Interventional therapy includes the use of polytetra uoroethylene-covered stents and stent-assisted coiling to seal the CAA [2] . However, these two interventions require operators with high skill levels; moreover, some disadvantages, such as the high incidence of in-stent restenosis (ISR) [13] and high cost, should also be considered. Therefore, this study proposed the use of a single or double-layer baremetal stent (SBMSs and DBMSs, respectively) to cover the neck of a CAA, combined with haemodynamic evaluation and patient follow-up to verify the feasibility of this technique.

Patient selection and CAA model construction
To investigate the range of application of this technique in CAAs with different neck diameters. We selected 3 different CAA neck sizes from patients who had an SBMS or DBMS implanted. The neck size of the CAA, maximum length and transverse length of the aneurysm, and parent artery diameter were measured, and the data were input into SolidWorks 2016 software (Dassault Systèmes SolidWorks Corporation, Waltham, MA, USA) for model construction.
Written informed consent was obtained from each patient included in the study. The study protocol conformed to the ethical guidelines of the 1975 Declaration of Helsinki, and the study protocol was approved by Fuwai Hospital's ethics committee on research in humans.

Stent model construction
The H-stent (Lepu Medical, Beijing, China) used in this study was made from 316L stainless steel with an elasticity of 200 GPa, Poisson's ratio of 0.3, and thickness of 89 μm. The stent mesh shape was square, covering an area of 91 μm*89 μm. The metal coverage (the ratio of the area covered by metal to the total area after the stent is expanded) was between 8 and 20%, the radial retraction rate was not greater than 10%, and the rate of axial shortening/elongation was less than 10%. The metal coverage rate was 15%, and the above parameters were input into SolidWorks 2016 software to construct a single-layer stent model (Fig. 1A). After successful construction, a double-layer stent model was constructed by adding a second layer on top of the single-layer strut . Common stacking situations were simulated. The two layers of struts were staggered, and the metal coverage rate and porosity of the double-layer stents were calculated as 28.7% and 71.3%, respectively (Fig. 1B).

Construction and meshing of CAA with the stent
On the basis of the above work, Boolean subtraction calculations related to the parent artery and stent model were performed in SolidWorks 2016 to obtain a scaffold model (Fig. 1C), which was output in *x_b format. Models of SBMS and DBMS implantation in the parent artery were simulated. The model les were entered into ANSYS ICEM CFD 14 software for gridding.

Boundary condition setting
Research shows that when the artery diameter is greater than 0.5 mm, the ow eld error is less than 2% when a Newtonian uid and non-Newtonian uid model is adopted. Therefore, in this study, blood ow was simulated as adiabatic, incompressible and steady Newtonian laminar ow, and the blood density was 1056 kg/m3. Blood viscosity was 0.035 g/cm *(s) [15] . The average peak normal coronary ow velocity was between 0.2 and 0.4 m/s [16] , and the inlet velocity was set to 0.4 m/s to verify the maximum effect of the stent on coronary blood ow in the resting state. The velocity was evenly distributed, the outlet pressure was set at 0, and the vessel wall was set as rigid without slip conditions. The blood ow followed the law of conservation of mass and momentum and conformed to the Navier-Stokes and continuity equations. The above settings were input to ANSYS ICEM CFD 14 software(Ansys Inc.,Berkeley,CA) to analyse the haemodynamics of stent-free, single-stent-loaded, and double-stentloaded CAAs. The portion size was between 0.02 mm and 0.25 mm.

Statistical analysis
Continuous variables are presented as the mean±standard deviation. Between-group comparisons were conducted using one-way analysis of variance for continuous variables. Statistical analysis was performed using SPSS 22.0 Software (SPSS Inc. Chicago, Illinois, USA). A 2-tailed P<0.05 was considered signi cant. (Table 1) Patient information is presented in Table 1. Case 1 underwent SBMS implantation, and the other two cases underwent DBMS implantation. All aneurysms were occluded without obvious ISR 12 months after the intervention.  -up showed also that the aneurysm was also occluded and that the stent surface was covered by endothelial cells without obvious ISR. Figure 4 shows the operation process and follow-up for Case 3. Saccular aneurysms distal to the right coronary artery (RCA) were observed before stenting (Fig. 4A). Stents of 3.5 mm*16 mm and *4.0 mm*24 mm were implanted in turn ( Fig. 4B-C). Angiography 1 year after stenting showed that the aneurysm had occluded without ISR (Fig. 4D), and coronary computed tomography angiography indicated that the aneurysm had a low density ( Fig. 4E-F).

Computational uid dynamics in 3 cases
Computational uid dynamics (CFD) analysis was performed on stent-free, single-stent and double-stent models for each case ( Table 2)

Discussion
The main ndings of this study are as follows: 1) single or double bare-metal stents covering the neck of aneurysms can signi cantly change the ow in the parent artery and the haemodynamics of the aneurysm to achieve occlusion; 2) CAAs with a neck diameter of less than 5 mm and without adjacent stenosis can be treated with a single bare-metal stent; 3) CAAs with a neck diameter of 5-12 mm and no adjacent stenosis can be treated with double metal stents; and 4) no obvious ISR was found during follow-up. CFD analysis can be used to simulate and analyse convective physicochemical problems, which helps to understand the haemodynamic changes in arteries resulting from abnormal structures. Therefore, it is widely used in cases of cerebral aneurysms [15,17] and thoracic and abdominal aortic aneurysms [18] , while it is rarely used in the study of coronary aneurysms.
In this cerebral aneurysm intervention study, it was found that ow velocity, pressure and WSS were the most important parameters affecting the development of cerebral aneurysms [19] . Flow velocity affected the formation of intracranial thrombi, and pressure and WSS were closely related to structural changes in the aneurysm wall. This study found that after stent implantation, the ow velocity and WSS in the CAA gradually declined, while the wall pressure increased( Table 2). The reason for this is due to the existence of the stent; the permeability of the ow inlet of the CAA was reduced, while a double stent makes this permeability even lower. In Case 1, after implantation of the second stent, the ow velocity in the CAA increased rather than decreased. This abnormal change may have been due to the small range in aneurysm neck diameter (<5 mm) and the change in permeability of the double-layer stents compared with the single-layer stents, which had a limited in uence on the haemodynamics in the CAA. This indicates that for a CAA with a neck diameter of less than 5 mm, a single stent can effectively change the haemodynamics of the CAA and reduce the ow velocity.
In this study, the CAA pressure in the double-stent model was higher than that in the single-stent and stent-free models in Cases 2 and 3. Contrast retention was also found in the CAA after stent release. The reason for the increase in pressure was also due to the increased resistance of the outlet ow from the CAA due to stent implantation. This is consistent with the studies in cerebral aneurysms [20] .
WSS is the tangential friction between the ow and the vessel wall per unit area. In Case 2, the WSS in the CAA after implantation of a single stent showed no signi cant difference compared with the stent-free condition, while after implantation of the second stent, the WSS was signi cantly reduced. This suggests that double-stent placement could in uence the haemodynamics in the CAA signi cantly more than single-stent placement. In Case 3, a signi cant reduction in WSS was observed after the implantation of a single stent. After implantation of the second stent, the WSS decreased further. As shown in Figure 5, with stent implantation, the low WSS area in the CAA of the 3 cases was signi cantly increased. Malek [21] found that WSS is an important indicator of CAA growth. When the WSS in the vessel wall is lower than 1.5 Pa, a macrophage-related in ammatory response can be induced, as can endothelial cell degeneration and apoptosis. By regulating the expression level of metalloproteinases, elastic bres and collagen bres in the artery wall are broken and absent, causing aneurysm growth and rupture.
Stent design and porosity are important factors in the outcome of aneurysm intervention [15,22] . Kim et al [15] found that the "square mesh" stent shape was more effective in changing the velocity and WSS in cerebral aneurysms than the "ring mesh" stent shape. Lieber's [22] study found that the haemodynamic changes in aneurysms were most pronounced when the porosity dropped to 76% but not noticeable when the porosity was lower than 70%. The range of metal coverage used for aneurysm treatment is generally between 60% and 86%, and the lower limit of this range is usually achieved by overlapping 2-layer stents [23,24] . However, metal coverage< 60% will lead to poor biocompatibility and high operational di culty when stents are delivered through curved arteries [22] . The H-stent had a square mesh design with metal coverage of 10-20%. When double-layer stents were superimposed, the metal coverage was approximately 71.3%, which was just within the range that has the greatest in uence on haemodynamics.
In terms of material, the bare-metal stents were made of 316L stainless steel, which is suitable for the growth of new endothelial cells due to the absence of dug coatings. However, with the process of epithelialization, the permeability of the inlet is further reduced, and thrombosis in the CAA is promoted. OCT follow-up in Case 2 suggested that the stent surface was completely covered by neointima without excessive proliferation. Commercially covered stents contain poly uortetraethylene (PTFE), which has poor biocompatibility and could affect epithelialization, leading to ISR. The PTFE study [13] suggests that the incidence of subacute stent thrombosis in covered stents was 5.7%, while the incidence of ISR was as high as 31.6% (29.8% at the stent edge, 8.8% in the middle). RECOVERS studies [25] have found similar results, and the incidence of subacute thrombosis within 30 days of PTFE stent placement was higher than that of traditional stent placement. Hachinohe et al [26] performed a retrospective study of 190 patients implanted with PTFE-covered stents over 20 years and found that target vessel myocardial infarction, occlusion, revascularization and in-stent thrombosis events occurred at the early stage after stent implantation, and their incidence increased gradually in the long-term follow-up.
Stent compliance is one of the key factors of operation di culty. Commercially covered stents, such as Graftmaster stents, generally have a sandwich structure, that is composed of a 50-μm thick PTFE layer between 2 stainless steel stents. Due to the poor compliance of covered stents, the operation di culty is increased if there is obvious bending or narrowing of the vessels during delivery, and there are even reports that the operational strategy has been changed accordingly [27][28] . Covered stents are not suitable for segments with branches. Sequential bare-metal stent release can not only maintain good stent compliance without increasing the di culty of the operation but also enable strategic adjustment according to the location of the branches to preserve the branch ow as much as possible. Finally, the longest covered stent is 28 mm, and the longest bare-metal stents can reach 30 mm, indicating a slight advantage of bare-metal stents in terms of stent length.
In terms of antiplatelet therapy, because covered stents are associated with a high incidence of ISR, physicians tend to choose a dual antiplatelet combination with a small dose of oral anticoagulants (such as warfarin), which increases the risk of bleeding. However, if patients are implanted with bare-metal stents without complications that they must use oral anticoagulants for (atrial brillation, deep venous thrombosis, etc.), they need to take dual antiplatelet drugs for only 6-12 months, which could reduce the risk of bleeding compared to covered stent therapy.

Limitations
This study has some limitations. First, the core of the DBMS technique is to reduce the porosity of the inlet area of the CAA. However, in practice, the second implanted stent cannot be accurately positioned.
There is a very small chance of complete overlap of the struts of the two stents, which prohibits the stent porosity from being further reduced and delays CAA occlusion. Second, this study used a CAA model without adjacent stenosis. The main objective was to investigate the effects of SBMSs and DBMSs on the haemodynamics of CAAs. The use of simpli ed models is more conducive to nding common features of diseases and treatments. CAAs with adjacent stenosis will be included in future studies. Third, in previous interventions, the longest CAA neck was closed with DBMSs of 22 mm [29] . The maximum neck length applicable to this technique remains to be further studied.

Conclusion
The application of SBMSs and DBMSs to cover the neck of CAAs can effectively change the ow in vessels and the haemodynamic state in the aneurysm, causing thermogenesis in the CAA. The SBMS and DBMS techniques were selected according to the CAA neck diameter. The proposed technique is an effective method for interventional treatment of CAAs that does not increase the incidence of ISR or the di culty of the operation.

Consent for publication Not applicable
Availability of data and materials The datasets used and analysed during the current study are available from the corresponding author on reasonable request Competing interests The authors declare that they have no competing interests Funding: Clinical and Translational Foundation of The Chinese Academy of Medical Sciences (2019XK320064) has provide nancial support to the research.
Authors' contributions: ZXY and SBQ contributed to drafting and revising the manuscript. JSY contributed to the Statistical analysis and coronary angiography operation. JGC contributed to follow up and drafting results. WYF and AKQ contributed to the stent and coronary aneurysm model construction and computational uid dynamics. All the authors critically revised the manuscript and gave nal approval and agreed to be accountable for all aspects of the work, ensuring both its integrity and accuracy.