Vascular Cast to Program Antistenotic Hemodynamics and Remodeling of Vein Graft

Abstract The structural stability of medical devices is established by managing stress distribution in response to organ movement. Veins abruptly dilate upon arterial grafting due to the mismatched tissue property, resulting in flow disturbances and consequently stenosis. Vascular cast is designed to wrap the vein‐artery grafts, thereby adjusting the diameter and property mismatches by relying on the elastic fixity. Here, a small bridge connection in the cast structure serves as an essential element to prevent stress concentrations due to the improved elastic fixity. Consequently, the vein dilation is efficiently suppressed, healthy (laminar and helical) flow is induced effectively, and the heathy functions of vein grafting are promoted, as indicated by the flow directional alignment of endothelial cells with arterialization, muscle expansion, and improved contractility. Finally, collaborative effects of the bridge drastically suppress stenosis with patency improvement. As a key technical point, the advantages of the bridge addition are validated via the computational modeling of fluid–structure interaction, followed by a customized ex vivo set‐up and analyses. The calculated effects are verified using a series of cell, rat, and canine models towards translation. The bridge acted like “Little Dutch boy” who saved the big mass using one finger by supporting the cast function.

x% PCL-co-y% PGMA were determined using differential scanning calorimetry (DSC, DSC214; NETZSCH, Germany). The samples were heated from -50 ℃ to 150 ℃ at a rate of 10 ℃ min -1 in a nitrogen atmosphere. The melting temperature (T m ), crystallization temperature (T c ) and melting enthalpy (ΔH m ) were determined using DSC. The crystallinity (X c ) was calculated using the following equation (1): (1) where the melting enthalpy of 100% crystalline PCL is 139.5 J g −1 . [1] A 6-arm 94% PCL-co-06% PGMA was selected to produce the vascular cast owing to the shape recovery near 40 ℃. The mechanical properties of the 6-arm 94% PCL-co-06% PGMA were analyzed using a DMA (dynamic mechanical analyzer, Discovery DMA 850, TA instrument Inc. New Castle, DE, USA) in film form.

SMP degradation
The accelerated aging test was performed to determine SMP degradation following the American Through the calculation, the following values were obtained in correspondence to 12 months (5,6). AAF = 2.0 3.3 = 9.85 (5) AAT = 365/9.85 = 37.01 (6) Following the obtained values, the accelerated aging condition was set by incubating SMP samples in normal saline (pH 7.4) at 70 ± 2 ℃ for 37days. Then, degradation-mediated changes in the thermal properties were determined by DSC analysis. Accordingly, the weight loss (%) was calculated using the following equation (7): m 0 is initial mass, m a is mass after degradation.

Elastic property characterization of rabbit vessels in a customized ex vivo system
The mechanical properties of the vessels were analyzed and entered for computer modeling as the input parameters. A rabbit infrarenal aorta and interior vena cava (IVC) were harvested and cut into 1 cm-long segments. All rabbit experiments and management procedures were approved by the recorded. Subsequently, the relationship between the pressure and wall tension was analyzed by calculating the experimental data using Laplace's law (ς pt/ri , where σ is the wall tension, Δp is the inner pressure-outer pressure, t is the wall thickness, and r is the radius of the cylinder). [2] The hyperelastic property values of the arteries and veins were entered as input data for computational modeling, and curve fitting was conducted using the Mooney-Rivlin 2 parameter model. [3]

Structural modeling
The 2D and 3D models of vascular cast were produced using Fusion 360 CAD programs (version 2018, Autodesk, California, USA). The base design of the strand structure was derived from polymeric stents and altered without (w/o) and with (w/) bridge, as the bridge structure was added to relieve the stress from vein dilation. Structural modeling was conducted using ANSYS Mechanical software (ANSYS 2020R1, ANSYS, Canonsburg, PA, USA) by inputting the mechanical property values of SMP, arteries, and veins that were experimentally obtained, as previously indicated. Structural modeling enabled the simulation of the bridge effect when the cast was deployed to wrap the artery-vein-artery graft by analyzing the structures of vessel and device in response to arterial pressure and pulsative blood flow. The diameter and length of each artery and vein were set to 3 and 30 mm, respectively, while the device length was set to 25 mm to cover the entire anastomosis and the proximal and distal parts of graft. The thickness and length of cast strands were set to 400 μm and 200 μm, respectively. The von-mises stress against the deformation of vascular cast and vein was monitored, and the elastic changes of device were analyzed with increased flow pressure.

Computational fluid dynamics (CFD) modeling
The structure of blood vessels changes in response to blood flow and pressure, which alters the flow profiles. The flow parameters were analyzed by CFD modeling using ANSYS Fluent software (ANSYS 2020R1, ANSYS, Canonsburg, PA, USA). The vessel geometry was constructed by entering the structural modeling data and then discretizing them into mesh structures through a finite element method (FEM). The blood was assumed to be a highly viscous non-Newtonian fluid at a low shear rate, and the viscosity decreases upon increasing the shear rate. Next, the Carreau-Yasuda model was applied (viscosity at zero shear rate = 0.056 Pa·s, viscosity at infinite shear rate = 0.00345 Pa·s, time constant = 1.902 s, power law index = 0.22, Yasuda exponent = 1.25). [4] The blood density was 1.060 kg m -3 , which was assumed to be incompressible. Three groups (no cast, w/o bridge, and w/bridge) were used in the model.
As a boundary condition, the vessel wall was considered to be no-slip, and the blood flow and pressure at the arterial inlet were obtained from the clinical AVF data. [5] Modeling analysis with quantification of the diastolic wall shear stress (WSSd), TAWSS, and OSI. In addition, the helical flow formation was analyzed as a healthy flow characteristic by quantifying the average helicity intensity and secondary velocity, whose vector is perpendicular to the vessel axis as an inducer of helical flow. [6]

Vascular cast fabrication
The computer-aided design (CAD) of vascular cast was first generated using the Solidworks software (Solidworks 2021, Dassault system, Vélizzy-Villacoublay, France). The flat shape (2D type) of cast model was then 3D-printed (IM2, Carima, Seoul, Republic of Korea) to generate a modeling structure in PDMS. SMP (6arm 94%PCL-co-06%PGMA) was dissolved in N-methyl-2pyrrolidone (NMP, Sigma-Aldrich, 1 g mL -1 ), and 1% photo-initiator (Irgacure 2959, Sigma-Aldrich) was added to the solution. The SMP solution was placed in the PDMS mold, pressed with glass, and then subjected to the first crosslinking using an ultraviolet lamp (300-400 nm, UVACUBE 400; Hoenle, Germany) for 30 s. After separating the structure from the glass, a cylindrical shape was formed, followed by a second crosslinking step for 20 min. The cast was washed with distilled water, dried under a vacuum, and stored at room temperature until further use.

Thermomechanical properties
The

Particle flow visualization
The geometry of the blood vessel in the systolic phase was constructed using CFD modeling and 3D printing (Raised3D Pro2, Irvine, CA, USA), which was used to generate the vessel structure in a poly(dimethylsiloxane) (PDMS) mold as a flow chamber after dissolving the 3D printed structure using tetrahydroflurane (THF). The flow chamber was coated with Pluronic F-127 (1 w/v%, P2443, Sigma-Aldrich) to prevent particle adhesion. The flow was visualized by perfusing red polystyrene microspheres (diameter: 4 µm, Invitrogen, Carlsbad, CA) using a peristaltic pump (BT100-1L, LongerPump, Amerham, UK) under a microscope. The particle images were analyzed using the fast Fourier transform (FFT) function of the Image J software (National Institution of Health, Bethesda, MD, USA), followed by determining the angle distribution of the blood flow direction using a plugin option.

Ex vivo system
A 3D ex vivo culture system was constructed to vary the typical arterial pressure (60-100 mmHg) and shear rate (10-15 dyne cm -2 ) from the artery-vein grafting. [5] A chamber (3.3 length × 1.5 width Live ReadyProbes Reagent (Invitrogen, Waltham, MA, USA), followed by confocal imaging (LSM artery between the clamps was excised with saline irrigation. A triangular space was created by widening the space between the hyoid, sternohyoid, and sternomastoid muscles using mosquito forceps. The carotid artery was pulled out to this space for device deployment (cast with or without bridge: 1 mm inner diameter and 5 mm length).
The device groups were sterilized using ethylene oxide (EO) gas and stored at room temperature.
The carotid artery was passed through the vascular cast, the device was moved to the jugular vein, and end-to-end anastomosis was performed using a 10-0 Ethilon suture (Ethicon). The vein was then released from the clamp, the arterial clamp was released to check the patency, and the ability to control the vein dilation was observed. The muscle and subcutaneous tissue layers were sutured using 5-0 Vicryl (Ethicon), and the skin layer was sutured using 5-0 Ethilon (W1661G, Ethicon).
The rats were monitored daily for two weeks until they were euthanized to obtain AV graft specimens.

Canine AV fistula (AVF) model
All

Histological and immunohistochemical analyses
In vivo tissue samples were rinsed with PBS and fixed with 4% paraformaldehyde (CellNest,

Gene expression
In vivo tissues were processed to determine the marker gene expression of arterialization in the ECs (ephrin B2), vein (EphB4, MMP-9, and eNOS), and SMC phenotype (αSMA, MYH-11, KLF4, and vimentin). The total RNA was extracted using an RNA extraction kit (74106, Qiagen, Hilden, Germany) following the manufacturer protocol, and the RNA concentration was determined using a and annealing stage at 60 °C for 1 min. Glyceraldehyde 3-phosphage dehydrogenase (GAPDH) was used as a housekeeping gene, and the relative gene expression was analyzed using the 2 -ΔCCt method.

Statistical Analysis
All statistical analyses were conducted using Excel and Sigmaplot (V12.0, Systat Software, CA, USA); the data are presented as the mean ± standard deviation (SD) using at least three samples.
The statistical significance was determined using an unpaired Student's t-test for two-group comparisons and a one-way analysis of variance (ANOVA) with Bonferroni's and Turkey's post hoc analysis for more than two group comparisons. P-values (*p < 0.05, **p < 0.01, and ***p < 0.001) were considered statistically significant.    Figure S4. Cytotoxicity of SMP was determined using a CCK-8 assay after eluates of the 6-arm 94% PCL-06% PGMA in a series of dilutions (50%, 75%, and 100%) were treated with L929 cells for one day. ** P < 0.005 vs. control with no SMP treatment (N=3/ ns: not significant). Figure S5. Vascular cast production with property tuning. a) The vascular cast is produced using the 6-arm 94% PCL-06% PGMA in a PDMS mold, whose structure is generated using a 3D printed model, followed by two-step crosslinking under UV. b) The original shape of the wrap is programmed to recover from a temporary plate shape, which facilitates deployment with covering end-to-end anastomosis. c) The increased duration of the 2 nd crosslinking reduces T m to recover the shape around the body temperature owing to the reduction of the crystallinity. d) The strand size of the vascular cast is adjusted to 200 μm among the test sizes, as the maximal strain is increased over 150% considering the elastic synchronization with arterial contractility. e) The bridge increases the circumferential tensile strength to suppress vein dilation with the calculated fixity. * P < 0.01 vs. w/o bridge (N=3).

Figure S6.
Ex vivo system to examine vein dilation in response to arterial hemodynamics. The system is set up to generate arterial shear stress and pressure using a peristatic pump with the tube flow control by adjusting the potential energy of media reservoir. The rabbit carotid vein, which functions as a bioreactor is loaded between the silicon tube with cast rapping in a tube. As the height of the media reservoir is elevated, the potential energy increases the venous pressure, which enables the pressure calculation.  WSS indicates the vector of the wall shear stress, μ is the viscosity of blood, u is the directional velocity of the blood flow, y is the normal direction of vessel wall, T is the period of cardiac cycle, v is the velocity vector, ω is the vorticity vector, and is the angle between the velocity and vorticity.