Visions of TAVR Future: Development and Optimization of a Second Generation Novel Polymeric TAVR

Abstract

Tissue-based transcatheter aortic valve (AV) replacement (TAVR) devices have been a breakthrough approach for treating aortic valve stenosis. However, with the expansion of TAVR to younger and lower risk patients, issues of long-term durability and thrombosis persist. Recent advances in polymeric valve technology facilitate designing more durable valves with minimal in vivo adverse reactions. We introduce our second-generation polymeric transcatheter aortic valve (TAV) device, designed and optimized to address these issues. We present the optimization process of the device, wherein each aspect of device deployment and functionality was optimized for performance, including unique considerations of polymeric technologies for reducing the volume of the polymer material for lower crimped delivery profiles. The stent frame was optimized to generate larger radial forces with lower material volumes, securing robust deployment and anchoring. The leaflet shape, combined with varying leaflets thickness, was optimized for reducing the flexural cyclic stresses and the valve's hydrodynamics. Our first-generation polymeric device already demonstrated that its hydrodynamic performance meets and exceeds tissue devices for both ISO standard and patient-specific in vitro scenarios. The valve already reached 900 × 10⁶ cycles of accelerated durability testing, equivalent to over 20 years in a patient. The optimization framework and technology led to the second generation of polymeric TAV design- currently undergoing in vitro hydrodynamic testing and following in vivo animal trials. As TAVR use is rapidly expanding, our rigorous bio-engineering optimization methodology and advanced polymer technology serve to establish polymeric TAV technology as a viable alternative to the challenges facing existing tissue-based TAV technology.

Introduction

Calcific aortic valve disease (CAVD) is characterized by the development and growth of semirigid calcific deposits within the aortic valve tissue. Mass buildup results in impaired leaflet motion and overall increased stiffness of valvular tissue which induces an additional burden on the left ventricle. This burden, termed aortic stenosis (AS), is presented as an increased pressure gradient to achieve physiological aortic pressures and cardiac outputs, with progression ultimately leading to reduced flow. The progression of CAVD and increase in the aortic stenosis gradient eventually leads to heart failure, which makes CAVD the third highest morbidity of heart diseases [1]. Severe CAVD is estimated in 2–4% of the elderly population (>65 years old) with milder disease severity of AS estimated in 25% of the population [1–3]. About 15,000 deaths annually in North America are attributed to CAVD, with AV replacement the second leading cause of cardiovascular surgery [1] contributing to an estimated annual health care cost burden of $14 billion in the U.S. [4]. Patients with severe AS are estimated to have a survival rate of 40-50% within 2 years and <20% in 5 years, illustrating the urgency for therapeutic intervention [5,6].

The traditional gold standard therapy for AS is a complete surgical aortic valve replacement (SAVR). This is a highly invasive procedure requiring an open-heart approach, with sternotomy and cardiopulmonary bypass. The calcific aortic valve is excised from the patient's heart and a new valve is sutured into the aortic annulus. Surgical aortic valve designs are grouped into two categories: mechanical valves, with pyrolytic carbon bileaflet disc/flaps, and trileaflet bioprosthetic or tissue valves, with bovine/porcine fixed pericardial tissue sutured into a rigid frame. Mechanical valves have functional and hemodynamic advantages of excellent pressure gradients and indefinite device durability, but require complex, lifetime anticoagulation/antithrombotic regimens for the patient, requiring adaptions in patient lifestyle. Often patients will favor tissue valves to avoid the need for long-term pharmacologic regimens, with the expectation that the valve will need replacement between 8.5 and 11 years postimplant [7]. Tissue valves may eventually fail due to structural valve degeneration (SVD), as intrinsic failure due to normal wear and tear, and extrinsic failure due to inflammatory processes, developing new calcific masses within the chemically fixed foreign valvular tissue. However, surgical aortic valve replacement is a highly invasive therapy and is often not appropriate for a large patient cohort deemed high risk to undergo open valve replacement surgery. For this patient population, minimally invasive transcatheter aortic valve replacement (TAVR) was invented in 2002 and approved for use in patients, clinically in the U.S. in 2011 [8].

The Revolution of Transcatheter Aortic Valve Replacement.

Since the advent of TAVR, it has been steadily gaining relevance and procedural uptake in the U.S. and the world. In 2019 the number of TAVR procedures within the U.S. exceeded standard surgical valve replacement for treating AS [9]. Numerous clinical trials have recently also expanded TAVR use in younger, lower risk, and bicuspid aortic valve (BAV) patients. The current U.S. commercial market is focused on two device families with distinct design features the sapien (Edwards LifeSciences Inc., Irvine, CA) balloon-expandable valves, and the CoreValve/Evolut (Medtronic PLC, Dublin, Ireland) self-expandable valves. The current sapien 3 ultrafeatures a cobalt-chromium frame with fixed bovine pericardium tissue leaflets, where the stent is crimped into a 14 F delivery catheter over a balloon. Generally, the delivery catheter gains access to the aortic arch from either femoral or radial arterial access. With the SAPIEN valve, the sheath of the delivery catheter is retracted, the valve on the balloon is brought into the failing calcific aortic valve, and the balloon is expanded to force open the valve and plastically deform the stent valve into place. The current self-expandable evolut pro+ valve is also guided into the aortic valve, but retraction of the sheath exposes and opens the stent, forcing open the native valve and anchoring in place with a constant radial force applied against the calcific leaflets and annulus.

Clinical Complications Associated With TAVR.

With successive generations of each device, the rate and severity of many clinical complications associated with TAVR have decreased. However, persistent complications remain, and the clinical relevance of others has dramatically increased with the inclusion of younger patients. With time since the adoption of TAVR, increasing physician and center experience has yielded more successful procedural outcomes with lower complications and better performing valves [10]. The complication of poor TAVR performance or patient-prothesis mismatch, is generally associated with the selection of a device that did not properly open within the measured patient annulus or the degree of oversizing for the patient was underestimated. Paravalvular leak (PVL) is a ranked degree of regurgitant diastolic flow between the outer skirt of the transcatheter aortic valve (TAV) device and the native annulus/leaflet which can occur in high-velocity jets. The risk of this complication has been gradually reduced with anti-PVL skirts in newer generation devices. Cardiac conduction abnormalities (CCA) occur when the device applies high radial forces against the atrioventricular node or membranous septum below the aortic annulus resulting in conduction disturbances and the need for a permanent pacemaker.

Thrombosis and thromboembolic events are particularly persistent and of great importance with younger/low risk patients. Thromboembolic events like major stroke have remained consistently between 1.0 and 5.5% [11] even with newer generation devices. Thrombosis events such as the termed hypo-attenuated leaflet thickening (HALT) have been increasing as younger TAVR patients age (Fig. 1). In this complication, a thrombus is generated and developed on the aortic surface of the TAVR leaflets, increasing the risk of embolization and restricting the motion of the TAVR leaflets, inducing new stenosis in the TAV device. It is thought that this is a result of poor hemodynamic conditions (stagnation) [15,16] on the aortic leaflet surface as well as unfavorable surface materials promoting thrombus growth [16].

Fig. 1.

(a) Cardiac CT image of HALT from Makkar et al. [12], and (b) reconstruction of the patient showing the presence of thicker leaflets resulting from the growth/development of a stable thrombotic or pannus formation. Images (c) are photos of excised TAV devices with HALT formation on the leaflets from Latib et al. [13]. Adapted images from Dvir et al. [14] showing (d) surgical valve leaflet tear at the commissural connection, (e) surgical and (f) transcatheter calcification formation within the bioprosthetic leaflet tissue causing stenosis of the valve replacement. All images are used with permission.

TAVR has inherent long-term durability concerns as tissue valves experience extensive structural damage during the crimping and expanding steps that are not imparted on their surgical valves counterparts. This damage is hypothesized to limit the durability of the valve in vivo and exacerbate the SVD damage that all tissue valves experience. The actual life span of TAV devices has not been firmly established yet in vivo due to its only recent adoption and the paucity of long-term clinical outcomes studies data. With the expansion of TAVR in younger patients, the lifespan of valves will be determined, while patients face the risk of Valve-in Valve (ViV) rescue operations when present design valves eventually fail [17].

The Current State of Polymeric Transcatheter Aortic Valve Devices.

As an alternative to tissue valves, polymeric surgical valves have been attempted as early as 1958 [18]. The design motivation for polymeric valves has been to combine a physiological mimicking flexible leaflets dynamics, with antithrombotic properties of hemocompatible materials. In general, these early attempts were limited by material selection and design/casting techniques, with many materials unable to withstanding large pressure gradients and continuous functioning demands of an aortic valve. With the advancement in material science, many companies and researchers have begun to develop polymeric valves. The current prominent efforts toward polymeric TAV devices are shown in Fig. 2. The use of newer materials promises resistance to damage (crimping and deployment), predictable and better fatigue (wear and tear) profiles, and resistance to platelet adhesion and activation [19–22].

Fig. 2.

The current state of polymeric TAV devices from the prominent manufacturers and research groups. Each company is organized in the current published or public state of progression from in vitro development to first in human trials. All images are used with manufacturer and publisher permission.

The progress and success of each such effort are related to the characteristics of the selected polymeric material, namely, in its hemocompatibility, fatigue strength, casting technology, and its hydrodynamic response under the demanding aortic valve flow conditions. Figure 3 outlines the four prominent casting techniques for creating polymeric valves, which vary with the thermoset or thermoplastic distinction of each polymer. “Dip coating” appears to be the most common technique, where a central core or mandrel is slowly dipped into a bath of polymer and lifted. Relying on gravity during the curing process, producing a uniform webbing of polymer, as well as the avoidance of bubbles, has led Foldax Inc. (Salt Lake City, UT) to utilize robotic arms to accomplish consistent dip-coated valves. Other traditional plastic molding techniques such as compression or injection molding cannot achieve the same degree of thin webbing but allow casting of any shape that dip coating cannot with a single mold part. Three-dimensional-printing and electrospinning techniques are still in their infancy but would allow directional polymeric material properties to mimic the biomechanical behavior of native valves. All the molding techniques highlight the largest advantage of polymeric valve technology—the repeatable and rapid manufacturing of a final product. Traditional tissue valves need to be hand sutured onto the device frame in a painstaking procedure, increasing product cost as well as the risk of damage and error during manufacturing, resulting in a very high rejection rate of the manufactured valves.

Fig. 3.

Typical molding processes for polymeric TAV devices: compression molding [23], injection molding [24], dip coating [25–27], three-dimensional printing, and electrospinning [28]. Each technique has advantages and challenges but must be appropriate for each type of polymeric material (thermoset versus thermoplastic). All images are used with permission.

In this study, we present the current progress of our novel second-generation polymeric valve by showing the optimization steps of both the stent frame as well as the leaflet shape. Our polymeric valve technology differs from the above-mentioned technologies. Aside from the use of optimization techniques that results in a unique varying thickness leaflet design, our valve uses a proprietary process to produce a crosslinked version of poly(styrene-block-iso-butylene-block-styrene) (xSIBS) [23,29,30] within the molding process. In this study, we further carefully develop the design utilizing advanced numerical techniques to optimize the device crimping and deployment process, as well as scaling it up for animal studies and clinical trials.

Methods

The evolution of our novel polymeric valve design is depicted in Fig. 4, from the initial concept of a surgical valve to the optimized second generation polymeric TAV device [23,29,31]. The initial surgical valve utilized a poly(styrene-block-iso-butylene-block-styrene) (SIBS) material with fiber reinforcement [31]. While initially showing promising hemocompatibility, it proved to be unsuccessful due to the nonoptimized leaflet shape leading to early calcification wear and tear, ultimately resulting in failure in early animal experiments. The next step was to remove the fiber reinforcement and change the approach by cross linking the SIBS polymer (xSIBS) customized for our valve design use, thereby achieving intrinsic leaflet strength [21,29,32]), and optimizing leaflet shape for reduced thrombosis using the device thrombogenicity emulation (DTE) technique [33]. The material was further refined into the current “flexamer” material intramolding cross-linking technology for our first generation TAV device, PolyV-1 [23,30,34,35]. With further technological advances in the casting process (seen in Fig. 4 compression molding), the PolyV-1 was the first design iteration to adjust the nominal or zero-stress configuration seen in Fig. 4, while keeping the DTE optimized leaflets, to reduce leaflet flexural stresses while in motion [23]. In the second generation design, PolyV-2, the leaflets were scaled to the larger size and further optimized for reducing the stresses during valve opening and closing, as well as the crimping stresses experienced by TAV devices before delivery in its crimped state.

Fig. 4.

Development timeline from the initial polymeric surgical concept design to the optimized polymeric PolyV-2 TAV device

The PolyV-2 design approach addresses some of the shortcomings found in the first-generation design, as well as issues involving the scaling up of larger size valves. The PolyV-2 features a unique design with suture-less over molding to create the leaflets and skirt/sleeve in a seamless single manufacturing step. The stent is registered within the mold and the polymer is cast to create the leaflet design and sleeve with wrapped features around the struts of the stent. This drastically reduces the manufacturing time and increases the manufacturing consistency of the final product since the manual suturing work is eliminated. It additionally reduces the risk of calcification and thrombus formation focal points that may occur where sutures are sewn and affixed to the tissue. The continuous wrap of the polymer around the stent and the adhesion of the polymer to the metallic frame allow even distribution of leaflet forces during valve operation.

Optimization of the Stent Frame for Transcatheter Aortic Valve Replacement.

A major challenge in the design and manufacture of polymeric TAV devices is the achievement of the proper deployment size, radial force, and avoidance of plastic deformation during crimping, all while minimizing the volume of the crimped stent and polymer material. Unlike tissue valves, polymeric materials are incompressible, and the total volume must be able to fit in the crimped annular volume (18 F for PolyV-2). Additionally, with the sutureless design, the entire polymer and metal strut needed to be considered so that during crimping, the metal stent joints, and their bending patterns do not place extreme stresses on the polymer.

A series of implicit finite element analysis (FEA) simulations, Abaqus 2020 (Dassault Systemes, Vélizy-Villacoublay, France), were conducted on single strut designs. The strut would be representative of the base cell of the final stent frame with four mirrored struts completing the cell. The strut designs have meshed with C3D8R elements with enhanced hourglass control and a mesh convergence/sensitivity analysis was conducted. A circumferential surface was radially displaced contacting the outer surface and crimping the single strut to below the desired 18 F specification. Radial symmetry and radial contact planes maintained the correct motion and stresses within the strut. The nitinol material of the strut was assumed to be superelastic [36] and generated from the stent manufacturer's material testing. The strut designs were parametrically varied, and optimization was achieved with the design that produced the greatest radial force within the deployment range of the device (between 20 and 25 mm diameter for the 27 mm PolyV-2, see Fig. 5), the lowest crimping strain or at least below the yield limit of 10% strain for nitinol while maintaining the lowest surface area or volume.

Fig. 5.

Optimization of the stent basic strut design from the simple strut of the PolyV-1 (top left) to the final curvilinear strut of PolyV-2 (bottom left). The middle chart shows the optimization of the radial force as a factor of strut area and the right is a rendering of the final design with the labeling of the desired radial force areas.

A final stent shape utilizing the optimal strut and an initial laser cutting profile was created to include the shape and attachment of the leaflets. The laser cut profile is generally cut in a smaller stock nitinol tubing, expanded on a mandrel in multiple steps, and shape set to create the final stent product. This process was simulated in FEA to create intermediate mandrel shapes and achieve a more consistent expanded shape required for the sutureless molding process. This final stent design was further tested in patient-specific and idealized simulations to confirm that deployment, anchoring and cyclic fatigue strains were within the desired limits.

Optimization of the Leaflet Shape and Thickness for Transcatheter Aortic Valve Replacement.

With the creation of the final stent design, the shape of the leaflet attachment region was defined, and the final leaflet shape was determined. Optimization of leaflet shape is a multistep process, similarly utilizing explicit FEA simulations. The optimization process changes the initial free stress profile of the leaflets (cast/molded shape), as well as varying the thickness within the leaflet. This optimization reduces the peak and average stresses within the polymer leaflet material over a typical cardiac cycle with a typical physiological normotensive pressure gradient waveform [23,34,37]. The systolic opening pressure is varied in the complete battery of testing scenarios, with results in this study using an 8 mmHg average systolic gradient to open the valve which is typical for a high-performance valve [37,38]. In this study, a normotensive diastolic gradient is shown, with 100 mmHg peak back pressure, but other pathological conditions were additionally tested in silico. The pressure gradient was applied to the aortic side of the leaflets and the simulation was conducted with all three leaflets contacting during diastole. Mesh and temporal convergence/sensitivity analysis was conducted on the leaflets with the same pressure gradient waveform and the leaflets maintained the same C3D8R elements with enhanced hourglass control with three layers across the thickness of the leaflet. The polymeric material was modeled as an isotropic hyperelastic material (Arruda-Boyce Model) based on uniaxial tensile data. Results are shown in Fig. 6.

Fig. 6.

Maximum absolute principal stress within the three 27 mm leaflet profiles over the cardiac cycle pressure waveform, with the top demonstrating the tensile stresses and the bottom the compressive stresses. The more open profiles have lower systolic stresses and higher diastolic stresses. The more closed profile has lower stresses overall but is harder to manufacture.

The first step was to design a leaflet with an initial configuration similar to typical TAV and surgical valve devices, with leaflets in an almost closed/diastolic configuration. Another consideration for polymeric leaflets is that the molding configuration must be considered in the design. The initial opening and closing cycle were simulated with uniform 300 μm leaflet thickness. The leaflet shape was extracted, and the simulation was restarted after zeroing the stresses, effectively simulating a design with the free stress state of the leaflets. The cyclic stresses were analyzed, and a final free stress profile was selected before varying the leaflet thickness.

In order to vary the thickness of the leaflet, the stresses and mesh were passed into a custom matlab algorithm (MathWorks, Natick, MA) scripted for reducing the stresses between a range of desired upper and lower stress limits. The thickness would be varied along the entire volume of the leaflet, accentuating regions of higher stresses while reducing the overall volume of the leaflets (vital for the polymer crimping). The output design was retested in silico while undergoing the same cardiac cycle conditions, and the resultant stresses were compared. These two steps of optimizing the free stress state and the thickness were iterated—until an acceptable optimized leaflet shape was determined.

Results

The use of in silico simulations allowed us to create an optimized TAV device showing the stent improvement over the previous generation PolyV-1 and similar leaflet performance with a larger size design.

The Optimized 27 mm Stent Frame.

The final stent frame design is shown in Fig. 5. The base optimized strut is highlighted and compared to the strut of the PolyV-1 design (the strut equivalent 27 mm device). The key findings are that with the optimization process, the strut shape became more curvilinear, consisting of variable circumferential thickness along the length of the strut. Overall, the thickness of the optimized strut was far smaller than that of the original with 150–250 μm circumferential and 300 μm radial—as compared to the respective 500 μm and 500 μm of the original strut design. This larger sweeping curved shape allowed the bending motion during crimping to occur over the length of the strut- not constrained to the joint region. The larger bending curvature allows more polymer material to be cast in the joint region without risking increased shearing stresses against the polymer material that may lead to damage during crimping. Importantly, the optimized strut is able to produce higher radial forces as a function of the volume or area of the strut, as compared to the original strut design (as seen in Fig. 5). The open joint angle of the optimized strut allows higher radial force in the desired deployment region between 20 and 25 mm (typical desired range of oversizing for this device). The original strut of the first-generation device reaches a maximum radial force outside the deployment range, making it less effective for proper deployment as compared to the optimized design.

The PolyV-2 design also achieves variable radial force over the length of the device with higher force, aiding in anchoring against the aortic annulus and calcific leaflets while lowering the radial forces in the left ventricular outflow track, thus reducing CCA risk. The crown region contacts the sinotubular junction of the aorta and should also have lower radial forces to avoid damage to the aorta, achieved with three longer contacting joints incorporated into the design. Lower radial force in the left ventricular outflow track section is achieved with longer struts and the additional radial flaring aids in sealing against PVL. Lastly, a single uncoated locking cell is placed in the belly region of the TAVR leaflet which aids in axially locking the device and resisting device migration. This locking feature remains uncoated so that minimal polymer remains behind the leaflets, thus reducing the stagnant flows that may increase HALT rates.

The final stent design was manufactured, and in vitro measurements of the crimping radial force were conducted and compared to the in silico simulation results, confirming the expected optimization results. Additional in silico simulations for fatigue analysis of the stent were conducted by placing a cyclic radial displacement and force due at the leaflet commissures while ensuring that strain amplitudes remained below 0.5%. Complex crimping simulations with the sutureless polymer wrap and leaflets were additionally conducted before finalizing the stent design.

The Optimized 27 mm Leaflets.

The target goal for the optimized leaflets is to maintain peak stress of <2 MPa which was determined to be the fatigue limit of the Flexamer material and peak stress in the PolyV-1 design. A comparison of three opening profiles can be seen in Fig. 6, clearly indicating that there is a tradeoff with this leaflet's size, with more open profiles tending to have higher diastolic stresses and lower systolic stresses, and vice versa for more closed configuration profiles. A more closed profile was selected for the PolyV-2 leaflets since the initial uniform profile was already able to achieve <1.8 MPa stresses. The presented von Mises stresses are applicable to this analysis since the Flexamer material is assumed isotropic and should be compared to the 5 MPa yield stress of the material [33]. Additionally, the larger leaflet sizes benefit from the decreased closing flow achieved by the more closed configuration profile, as the leaflets close faster.

Following the selection of the opening profile, the design was modified to include the variable leaflet thickness seen in Fig. 7 (comparing the nonoptimized and optimized leaflet thickness). The benefit of variable thickness leaflets is clearly seen, achieving a significant reduction in peak principal stress (absolute) at or below +1.2 MPa and above −1.0 MPa and reduction in von Mises stresses below 1 MPa, with a lower, more uniform stress distribution during peak diastole. The optimized leaflets have lower overall stresses while slightly reducing the overall volume of material in the leaflets (additionally benefiting the crimping stage). The same large opening area during systole is maintained with minimal differences in systolic stresses. This unique type of manufacturing of variable thickness leaflets currently is only achievable with compression or injection molding, as two mating mold parts need to define the variable thickness throughout the entire leaflet surface.

Fig. 7.

Stress results from peak systole and peak diastole of the 27 mm PolyV-2 leaflet design. The top row is the original uniform leaflet thickness and the bottom row is after processing into an optimized variable thickness leaflet model.

Validation of in Silico Models.

Interpretation and acceptance of the presented in silico models requires validation of the results with in vitro models. The process of neither manufacturing each design iteration has not cost nor time effective, therefore the validation needs to be conducted at critical design points throughout the process. For example, the final optimized stent design was cut and shape set, and radial force measurements were compared to the in silico results aiding in the acceptance of the crimped strain values previously obtained throughout the optimization process. To validate the leaflet motion in the FEA models, the valve was manufactured and setup in a pulse duplicator (Left Heart Simulator, Vivitro Labs, Victoria, BC), and the motion was recorded with a high-speed camera (1057 FPS, Chronos 1.4, Kron Technologies Inc, Burnaby, BC). Due to the transparent leaflets, the raw images need to be processed to enhance and extract the free edges, and the geometric orifice area (GOA, the projected opening area) can be obtained and compared to the in silico motion. Seen in Fig. 8, there is an excellent agreement between the opening area and profiles of the in vitro and in silico models, with minor discrepancies in the opening and closing speed. Given this large investment in resources to achieve validation, it is critical to establish the optimization protocol with a practical, efficient, and relevant model before proceeding.

Fig. 8.

Validation of the leaflet motion is obtained with the comparison of GOA between the FEA and in vitro model. The in vitro model the high-speed flow loop images are seen in the top row, image processed images are below and the compared FEA profile in the third row. The line graph compares the GOA of each model throughout systole.

Superior Hydrodynamics.

The PolyV-2 design is currently undergoing initial manufacturing and performance testing. We expect similar and improved hydrodynamic performance to that of the PolyV-1 design, as both were optimized in a similar fashion and share a similar in silico performance. The PolyV-1 was previously shown to have excellent performance in both idealized [23] and patient-specific [34] in vitro testing scenarios. Figure 9 highlights the key finding of those studies with the PolyV-1, despite the smaller device size, having performance meeting or exceeding a surgical valve in idealized conditions (Left Heart Simulator, Vivitro Labs, location). The device was tested according to the ISO 5840 (2013) guidelines and exceeded the minimum performance recommend in the standard. Patient-specific testing highlighted the need for improving the stent design for more circular opening profiles and the need for larger device sizes for expanding the testing scenarios with more challenging pathologies [34].

Fig. 9.

Left graph adapted and used with permission from Rotman et al. [23] showing the optimized PolyV-1 in an idealized in vitro simulation exceeds the performance of the gold standard surgical valve. Right is adapted from Kovarovic et al. [34] showing that the PolyV-1 is able to match or exceed the tissue TAV performance in challenge patient-specific models.

Superior Durability.

Previously the PolyV-1 was shown to have long-term durability exceeding the required 200 million cycles recommended by the U.S. Food and Drug Administration and the ISO 5840 (2013) standard [37]. Testing has continued in the accelerated wear tester (AWT, HiCycle, Vivitro Labs, location) at the 10 Hz frequency. Currently, the (n = 4) valves have exceeded 900 M cycles with no loss in performance or noticeable wear and tear (Fig. 10). A target of 1 billion cycles would represent an estimated 25 years of in vivo patient performance and is generally considered the target upper limit for valve devices [39]. One notable feature of the testing is that the closing flow of the leaflets decreased over the testing lifetime as the free stress profile began to close. This reduction helped maintain the performance of the valve by reducing the burden of the stroke volume. This opens the door to further exploit or utilize the fatigue of the leaflets to benefit the valve performance in future designs.

Fig. 10.

The current progress of the durability testing of PolyV-1, extended from Rotman et al. [30] to 800 M. The valves (n = 4) showed minimal fatigue properties with minimal change in effective orifice area, and the semi-open leaflet profile tends to a more closed profile which reduces the closing volume.

Discussion

Transcatheter valves are rapidly becoming the standard of care therapy for all aortic valve replacements. Despite their growth in use, present tissue valve devices are subject to clinical complications and structural valve degeneration that limits their durability compared with surgical valves, hampering their extension for use in younger, lower-risk patients. Recently published studies are beginning to highlight the distinct advantage of polymeric valve technologies and the progression of these promising innovations for valve replacement toward regulatory approval. Advances in material technology have allowed other research groups and companies to progress to first in human trials opening the possibility that these devices may successfully compete for market share versus traditional tissue devices currently employed in TAVR technology.

We have approached the design of our current PolyV-2 with careful consideration and utilization of in silico modeling techniques to achieve a feasible and viable design. Combined with a robust, hemocompatible, and stable polymer—Flexamer, in silico modeling allowed us to study in parallel, the performance of earlier generation device design and to perfect the casting and manufacturing techniques required.

The optimization of the stent frame based on the initial optimization of the strut has yielded a manufacturable, repeatable stent design that meets or exceeds our expectations for the frame. With the incorporation of the sutureless overmold design, it was necessary to design for the additional material within the joint of each cell and to reduce the overall volume of the stent frame. The current design pushes the envelope of current manufacturing technology, with much thinner struts and large opening angles. Additionally, in silico modeling was essential to generate the intermediate shape setting mandrels and create repeatable manufacturability of the stents.

The initial success of the sutureless design shows sufficient adhesion between the polymer and nitinol, as well as ease of manufacture. We believe this will be one of the most significant advantages of polymer TAV technology, as it dramatically reduces the cost, time of manufacture, and greatly reduces the potential for human error and the subsequent high rejection rate that increases costs. Reduction in product manufacturing cost will also allow the specification of tighter quality tolerances. While the long-term fatigue of the overmolded polymer wrap has not been fully established, this design has the advantage of distributing the load across the entire stent as opposed to only along with distinct suture points, supporting that its durability profile will be similar to, or better than, its PolyV-1 design predecessor.

Optimization of the leaflet design achieves lower stresses within the leaflets and reduced peak stresses at critical cycle time points. Reduction in the polymer volume within the leaflets helps reduce the crimping target size as well, potentially easing valve delivery. Selection of the opening profile is necessary for stress reduction and is compatible with our molding technology. This open profile has also been demonstrated to improve the performance of the valve as it fatigues, in long-term durability testing. This optimization process is currently only feasible and useable with polymeric TAV devices as all tissue-based valves use relatively uniform thickness tissues harvested from animals, unable to incorporate variable thickness designs optimized for stress reduction. The success of the PolyV-1 hydrodynamics confirms that this optimization process produces larger opening areas and lower systolic pressure gradients compared to tissues valves, and thus offers the potential for better performance in patients.

Polymeric TAV technology is at the forefront addressing the critical clinical complications of persistent thrombosis events such as HALT and stroke, as well as the limited durability of tissue TAV devices. Polymeric materials such as xSIBS technology additionally offer reduced platelet response and adhesion, with smoother, less-fibrous surfaces and hemocompatible surface chemistries [19–21,23,29,33]. The optimization of leaflet stresses and inclusion of sutureless designs should help reduce the occurrence of wear and tear durability failures associated with tissue-based TAV devices. The material chemistry and inflammatory response will also reduce the occurrence of calcific growth within the TAV leaflets. Polymeric TAV technology is on the precipice of innovation and acceptance for the treatment of AS, as a viable alternative to address the shortcomings of current-tissue-based TAVR technology. The promise of this disruptive technology is a safer extension to younger, lower-risk patients in offering a long-term solution that may obviate the risk of ViV rescue procedures or the need for surgical valve replacement following TAV device failure.

Limitations

This study has several limitations stemming from the use of in silico models to generate design inputs without the resource-intensive direct comparison to in vitro trials. The accuracy of the in silico models relies heavily on our extensive experience with such models and careful consideration for the model approach and setup. More importantly, the results of the models are used as a comparative result to existing and previously in vitro tested PolyV-1 designs, as well as other commercial valves in clinical use. Recent updates to the ISO 5840 (2021) standard have highlighted the underestimation of the in vivo loading in the AWT and require extended (minimum 50 M cycles) real-time wear testing to account for this discrepancy. Previous patient-specific testing [34] also demonstrated the lack of clinical accuracy of these idealized testing scenarios and the ability of the polymeric valve to exceed the performance of current commercial TAV devices. This additional durability testing does not negate the observed extended durability in this study, but rather will reinforce the targeted performance upon eventual successful completion. Additional limitations stem from the use of reference material properties for commercial nitinol and fixed tissues that are used for the comparative studies. In the leaflet FEA studies, the belly edge is assumed to be rigidly fixed and does not account for design features such as fillets and smooth transitions between the leaflet and sleeve that are present in the full design. Therefore, bending stresses on this edge are not entirely representative of the final design, and the goal of the presented studies is to reduce the stress within the leaflet and account for the bending stresses in the final full design model. The polymer is assumed to be isotropic, however, there may be directionally dependent properties during the manufacturing process based on the flow of the polymer within the mold [24].

Future Studies and Trials

The optimized design PolyV-2 27 mm valve is currently undergoing production and further in vitro evaluation according to the ISO 5840 (2021) standard. A similar and expanded battery of patient-specific human CAVD models will be tested to confirm similar robust performance as demonstrated by the PolyV-1design. Additionally, acute and chronic ovine animal trials that currently follow will test the predictions of our simulations and in vitro studies by comparing those to the in vivo outcomes of the optimized PolyV-2 design. We have previously established a similar predictive DTE optimization methodology using it as an in silico predictor of in vitro and in vivo ventricular assist device thrombogenicity [40]. While in vivo results are not available, our DTE modeling experience establishes our confidence that the presented TAVR optimization methodology will also be predictive of our animal experiments outcomes.

Conclusion

This study was designed to demonstrate an optimization process for designing a novel second-generation polymeric TAV device. The results of these studies will help establish a clear R&D methodology outline

Utilizing the in silico modeling of the various aspects of the device, from stent frame crimping to the functioning of the leaflets in a typical cardiac cycle, each result was iteratively optimized to achieve better valve performance, while considering the unique features of our polymeric designs. The stent frame was designed to reduce the crimped volume of the stent compared to its predecessor while generating higher radial forces in the effective range that the device will be used, at the same time avoiding crimping damage. The leaflets were optimized for a larger size device, further reducing the cyclic stresses below the previously achieved fatigue limit in the successful durability tests of our first-generation TAV. While subsequent initial in vivo trials will help determine the ultimate success of this optimized second-generation PolyV-2 TAV, this careful integrative approach studying all aspects of valve deployment and performance before manufacture are likely to ensure the overall success of the device.

Funding Data

• National Institutes of Health (NIH)- National Institute of Biomedical Imaging and Bio-engineering (NIBIB) (Grant No. U01EB026414-01; Funder ID: 10.13039/100000070).

• Small Business Technology Transfer (STTR) (Phase II R42HL134418-03A1, Phase I R41HL134418-02; Funder ID: 10.13039/100007002).

• Center for Biotechnology: A New York State Center for Advanced Technology (Funder ID: 10.13039/100009115).

Nomenclature

AS =

aortic stenosis

AV =

aortic valve

CAVD =

calcific aortic valve disease

CCA =

cardiac conduction abnormality

DTE =

device thrombogenic emulation

FEA =

finite element analysis

HALT =

hypo attenuated leaflet thickening

PVL =

paravalvular leak

SVD =

structural valve degeneration

TAV =

transcatheter aortic valve

TAVR =

transcatheter aortic valve replacement

ViV =

valve in valve