Measuring Stress on a Third-Generation Balloon-Expandable Aortic Valve During Expansion

Summary

The following protocol outlines simulations conducted to evaluate the influence of systemic pressure loading on the structural integrity of a 26-mm SAPIEN 3 valve. The stress level was determined by utilizing finite element analyses, which indicated that the highest stresses during the crimping phase were concentrated at the commissural tips.

Abstract

This study aimed to determine the mechanical stresses experienced by third-generation, balloon-expandable transcatheter aortic valves in the presence of aortic calcification, defined by an elevated Agatston score index for calcium. The device under investigation was the Edwards SAPIEN 3 valve (referred to as S3-TAV henceforth), a 26-mm commercial device. Five steps were undertaken. The first was the reconstruction of the three-dimensional (3D) computer-aided design (CAD) model of the device. The second was a simulation of the crimping and reopening process. The third was evaluating the crimping effect on leaflet stress and a comparison to an ideal condition. The fourth was a simulation of real clinical cases. Finally, the subsequent phase of the study entails the post-processing of the simulation results and a comparison with the follow-up data. A high-resolution micro-computed tomography scan was used to develop an accurate 3D geometric mesh of the stent and the valve. The material properties of the leaflets were derived from surgical bioprostheses; the material properties of the stents were based on those of cobalt-chromium. A series of simulations were conducted to assess the impact of systemic pressure loading on the structure. Stress was quantified through the application of finite element analyses. A stress analysis of the valve, conducted using exact geometry derived from high-resolution scans, revealed that peak stresses during the crimping phase were concentrated at the commissural tips, where the leaflets were attached. These regions were identified as the most likely sites for thrombosis and degeneration. The persistence of bulky calcifications following the device expansion results in the formation of paravalvular leakages (PVLs), which act as precursors to thrombosis and structural valve degeneration.

Introduction

Transcatheter aortic valve implantation (TAVI) has been demonstrated to be an effective treatment for severe aortic stenosis (AS), particularly in patients who are at intermediate or high risk of undergoing surgery or in those who are not eligible for surgical intervention^1,2,3,4. The SAPIEN 3 represents the third generation of transcatheter balloon expandable aortic valves (TAVs). It was initially released for provisional commercial use in Europe in January 2014 and subsequently in the United States in June 2015. The valve was the pioneering TAVI device to be approved by the United States Food and Drug Administration (FDA) for an expanded indication, namely intermediate-risk surgical patients with severe symptomatic AS. A noteworthy enhancement in design compared to the preceding SAPIEN XT (referred to as S-XT henceforth) iteration was the integration of an outer polyethylene terephthalate (PET) sealing cuff at the basal aspect of the stent. This innovation markedly reduced the occurrence of paravalvular leakage (PVL). Furthermore, in order to accommodate more convenient procedures in patients with smaller annuluses and provide superior performance in valve-in-valve implantations, a 20-mm diameter S3-TAV was developed in addition to the standard 23-mm, 26-mm, and 29-mm sizes. A significant restructuring of the S3-TAV stent configuration has resulted in notable changes to its clinical implications. The crimped valve exhibited a markedly diminished profile, thereby facilitating its insertion within a 14F or 16F expandable sheath. Furthermore, the stent included a greater number of cells at its distal end, which facilitated subsequent percutaneous coronary interventions via the struts. Finally, the overall length of the stent was increased⁵. The advent of stent configurations that permit routine transfemoral percutaneous access represents a significant advancement in vascular access for TAVI.

In the original clinical trials involving high-risk and inoperable individuals, who were typically octogenarians, concerns regarding TAVI durability were not a significant issue^{6,7,8,9,10,11,12}. Nevertheless, as patients of a lower risk and younger age are included in the study, the issue of TAVI durability becomes increasingly pertinent. The Placement of Aortic Transcatheter Valves (PARTNER) 3 trial commenced recruitment in April 2016 with the objective of evaluating the safety and efficacy of S3-TAV in low-risk surgical patients with aortic stenosis (AS)¹³. The durability of a valve is directly influenced by its design. It has been determined that heightened leaflet stresses can be found in regions that are in close proximity to areas exhibiting calcific degeneration and tearing, which have the potential to be implicated in thrombosis complications^{14,15,16,17,18,19,20,21,22}. Consequently, an appreciation of the characteristics associated with TAV leaflet stresses is essential in order for a meaningful comparison of the relative durability of TAVs to be conducted in comparison with other devices or bioprosthesis. In previous studies, the methodology of finite element analysis (FEA) was employed for the assessment of mechanical stresses on the leaflets of both balloon-expandable transcatheter heart valves (TEHVs) and self-expanding TEHVs^20,21,23,24.

FEA constitutes a firmly established methodology for the determination of indispensable data pertaining to sophisticated biological structures, which would remain unfeasible to directly measure in vivo were it not for this approach. FEA is a highly valuable tool in physiologic studies when employed to determine device durability in silico by means of stress estimation and failure mode determination. In order to create accurate finite element models, it is necessary to have a precise 3D representation of the geometry in a zero-pressure state, a comprehensive understanding of the assembly, the material properties, and the physiologic loading conditions. Given the substantial revisions to the design of the valve, the stress distribution on the most recent third-generation, balloon-expandable TAV, when present, is yet to be established in the context of aortic valve calcification^25,26.

The objective of this study was to ascertain the stresses experienced by TAV stents and leaflets in a 26-mm S3-TAV. The second step entailed the utilization of a computational biomechanical model to examine the prospective structural modifications that may occur to the leaflets and stent following transcatheter heart valve (THV) implantation in two patients with aortic valve stenosis who underwent a transcatheter aortic valve implantation procedure with the S3-TAV 26 mm system. Subsequently, the potential complications that may have developed with the device were evaluated in two clinical case studies.

In this study, calcium plaques are modeled based on the data obtained from micro-computed tomography (Micro-CT) reconstruction. There are four specific objectives of the study: (1) developing a three-dimensional computer-aided design (3D CAD) model of the device reconstruction. (2) Evaluating the impact of crimping and the simulation of reopening with and without prosthetic leaflets. (3) Evaluating the effect of crimping on the stress experienced by the leaflets. (4) Replicating the complete clinical procedure. The subsequent phase of the study necessitates the post-processing of the simulation results and a comparison with the follow-up data^3,4 (Figure 1).

A systematic approach is proposed for the realistic simulation of TAVI, which is aligned with clinical practice. The objective is to predict the postoperative performance of the prosthesis relative to the specific anatomical features in question. The investigational device was a 26-mm S3-TAV obtained from a patient and measuring 25.8 mm in external diameter and 20 mm in height.

The TAV assembly consists of four components: bovine pericardial leaflets, cobalt-chromium stent, Dacron covering, and outer polyethylene terephthalate (PET) sealing skirt. The physical dimensions and the suture junctions between the various components are recorded to facilitate the precise modeling of the assembly with its connections. A 3D mesh representing the apparatus is developed, followed by the application of an FEA using a finite element explicit solver. Post-processing of the FEA and data analysis determines the stresses experienced by the leaflets and stent (Figure 2).

The example provided highlights two patients who underwent TAVI. The two patients presented with severe AS and were considered to be at higher risk for complications. Both patients underwent TAVI via transfemoral access. Pre-operative planning was initiated in both cases using CT data, which represents the current standard methodology for device selection based on the assessment of the virtual ring and the evaluation of the ideal approach in accordance with the guidelines^3,4.

Protocol

The use of these data for research was approved by the Institutional Review Board, with patient consent waived. IRB ID: 202201057, Approval Number assigned by the IRB: IRB-MTP_2022_03_202201057; ClinicalTrials.gov ID: NCT05261204.

NOTE: A summary of the protocol is provided in Table 1.

1. Investigated device biomodeling study

1. Computed biomodelling study
   NOTE: One patient exhibited a malfunction attributable to a PVL. The device was subsequently removed and utilized for the biomechanical study.
   1. Perform a micro CT scan for the patient in question.
   2. Compare CT imaging data obtained from the finite element model (FEM) using a similar model from two control patients. In order to ascertain the accuracy and precision of the device simulation data, undertake a comparative analysis with the data obtained from a control CT scan. Create a 3D CAD model of the device using the actual device explanted as a reference (Figure 2A).
   3. Conduct imaging of the 26 mm S3-TAV with a desktop cone-beam micro-computed tomography scanner (micro CT -40) in various orientations, including longitudinal sections through the short axis to visualize the valve leaflets and the long axis to facilitate reconstruction of the stent structure.
   4. Additionally, employ different intensities to facilitate differentiation between the stent and the valve leaflet geometries. Use the following settings for the scan: An energy of 45 kVp, a current of 200 mA, a filter of 0.5 mm aluminum, a field of view of 50 mm, a voxel size of 50 mm, and an integration time of 200 ms for the X-ray examination.
   5. Import the digital radiographic images, with a voxel size of 50 mm, into Abaqus (commercial finite element solver) 2017, an open-source software program for surface reconstruction.
   6. Obtain the geometrical model through two distinct processes.
      1. First, geometrically reconstruct the stent frame, prosthetic sheets, and inner skirt using CAD software Rhinoceros 5.0 commands.
         NOTE: The CAD models of the stent frames were created using Rhinoceros 5.0 software, and the crimping phases of both the devices and their catheters were reproduced using the commercial finite element solver/Explicit software from MatLab. This process involved gradually reducing the initial catheter diameters of 27 mm until they reached their final size of 4.7 mm. The behavior of the balloon-expandable valve within the patient's aortic root was replicated using a standardized radial displacement applied to the nodes of a fixed cylindrical model designed to replicate the inner expanding balloon. This process commenced with the crimped configuration of the transcatheter frame.
      2. Subsequently, define the properties of the FEM.
         NOTE: The elastoplastic behavior comprises a von Mises plasticity model with isotropic hardening, an isotropic hyperplastic model (strain-energy potential: Ogden, 1st order), an isotropic elastic model, and finally, a mesh comprising 89,718 elements. These are subdivided into leaflet and skirt elements (41,154 S4R elements) and frame elements (48,564 C3D8R elements) (Figure 2B).
   7. Perform pre-operative CT examinations (Centre Cardiologique du Nord, Saint Denis, France) using a dual-source computed tomography scanner. To obtain contrast-enhanced images, inject an iodinated contrast agent. The main scan parameters for cardiac CT include mixed axial gated and helical ungated, 120 kV, smart mA, dose-length product (DLP ) 475 mGy.cm, body mass index (BMI) 26, 70-96 beats per minute (bpm) and acquisition time 6.8 s (Supplementary Figure 1).
   8. Use the commercial finite element solver (https://www.suse.com/pcsc/viewVersionPage?versionID=17782) to create an aortic valve FEM as reported in steps 1.1.3-1.1.7. (Figure 2, Figure 3, and Figure 4).
      NOTE: Table 2 presents a summary of the performed tests of crimping and reopening simulation.
   9. Use these models in simulations involving stent crimping and prosthesis implantations within the native roots.
   10. To create patient-specific geometrical models of the aortic valve complex, utilize a retrieved TAV, ensuring precise material properties.
       NOTE: The TAVI simulation included the following steps: (i) Pre-processing the medical images; (ii) Identifying appropriate models for analysis; (iii) Simulation of the procedure based on the acquired data; (iv) Post-processing of the simulation results.

2. Native aortic root model

1. Use ITK-SNAP 3.6 software (accessible via 20, 2021) to identify the principal anatomical characteristics of the patient's aortic root from pre-operative DICOM images and generate three-dimensional reconstructions (Segmentation).
   1. Process the CT data sets using ITK-Snap v2.4.
   2. Extract a confined region of interest (i.e., the aortic root from the left ventricular outflow to the sinotubular junction) from the whole reconstructed body, exploiting the contrast enhancement, cropping, and segmentation capabilities of the software.
      1. Employ the ITK-Snap v.2.4 software for the extraction of the stereolithography (STL) representation of the aortic root and calcific deposits from CT images.
      2. Process the STL file and generate the finite element mesh of the aortic root using a bespoke in-house code. Model the native leaflets with shell elements under the assumption of a uniform thickness of 2.5 mm^23,25.
         NOTE: Refer to section 1 of Supplementary File 1^{12,20,23,25,26}for more details.
   3. After segmentation, use an in-house Matlab code (v.R2018b) to generate a suitable mesh of the aortic wall.
2. Assume a constant thickness of 2.5 mm for simplicity.
3. Employ Rhinoceros 5.0 commands to construct a definitive 3D CAD geometry of the aortic root as described previously in step 1.1.6.
4. Export the aortic root volumetric reconstruction to the commercial finite element solver for discretization using C3D4 tetrahedral elements. Refer to section 2 of Supplementary File 1.

3. Calcifications

NOTE: To accurately re-enact TAVI, include calcifications close to the leaflet. These may significantly affect the dynamics of stent expansion alongside the technical complexity and efficacy of the procedure.

1. For the commercial finite element solver model, treat each calcification as a standalone entity.
2. Use a kinematic coupling constraint technique, connecting the surface nodes of the calcium deposits to the designated reference nodes of the leaflets to ensure precise alignment.
3. Refer to section 3 of Supplementary File 1, Supplementary Figure 2, and Supplementary Figure 3.

4. Prosthetic model

NOTE: High-resolution micro-CT images of the 26-mm TAV devices, captured following expansion, have been employed to generate precise geometrical models of the devices through the creation of accurate 3D replicas using CAD software.

1. Construct CAD models of the stent frames using Rhinoceros 5.0 and Matlab.
2. Obtain the geometric model using two distinct processes.
   1. First stage: Use CAD software Rhinoceros 5.0 commands, enabling geometrical reconstruction of the stent frame, prosthetic sheets, and inner skirt.
   2. Second stage: Perform the analysis and stress distribution on element mesh mapping by in-house Matlab code (https://www.mathworks.com/help/matlab/matlab_prog/create-scripts.html) (Figure 2).
3. Use the commercial finite element solver/Explicit to replicate the crimping stages of the device within the catheters, with the initial diameter of 27 mm gradually reduced to the final diameter of 4.7 mm.
4. Import the digital radiographic images, with a voxel size of 50 mm, into the commercial finite element solver for surface reconstruction (Figure 3, Figure 4, Figure 5, and Figure 6). Investigate the contact definitions between the leaflets and the stent to choose the most accurate representation of the overall behavior based on this data.
5. Connect the leaflet nodes to the stent using a tie contact, with a coefficient of friction value of 0.1 (t = 0.1 s).
6. Affix TAV leaflet geometries to Dacron at the bottom and secure it to the stent geometry at the top.
7. Divide the TAV leaflet mesh into four distinct regions to study the stress distribution due to pressure loading: (i) upper commissure, (ii) lower commissure, (iii) upper free leaflet region, (iv) lower leaflet belly region.
8. In order to replicate the balloon-expandable valve behavior within the patient's aortic root, implement a uniform radial displacement at the nodes of a rigid cylindrical surface, which serves to emulate the inner expanding balloon.
   NOTE: This was done starting from the crimped configuration of the transcatheter frame.

5. Material models

1. Model the native leaflet tissues.
   1. Use simplified isotropic St. Venant-Kirchhoff material properties. These properties include Young's modulus E with a value of 8 MPa and a Poisson's ratio v of 0.45.
   2. Use a six-order reduced polynomial constitutive model to describe the hyperplastic material¹¹, which represents the nearly incompressible behavior of cardiac root tissue.
   3. For the aortic wall and leaflets, assume an E = 8 MPa, v = 0.45, and ρ = 1.1 × 10^-9 tons/mm³.
   4. For the calcified tissues, use the following values^12,24: E = 10 MPa, v = 0.35, and ρ = 2 × 10^-9 tons/mm³.
   5. Use von Mises plasticity model with isotropic hardening to describe the elastoplastic behavior of the stent.
   6. Use the following values for the relevant parameters: E = 233 GPa; v = 0.35; 414 MPa, 933 MPa, and 44.5% in terms of yield stress, ultimate stress, and deformation at break, respectively²⁴.

6. Simulation details

NOTE: In the context of TAVI simulations, the ratio between kinetic energy and internal energy has remained below 10% throughout, reflecting the essentially quasi-static nature of stent deployment.

1. In order to ensure an accurate representation of the in vivo environment observed in aortic roots, impose preliminary boundary conditions on both extremities which were restrained to a normal plan aligned with the axis of the stents. Employ this approach to prevent excessive movements. Refer to Supplementary File 1.
2. Furthermore, obstruct the nodes at the base of the stents to impede the longitudinal translation of the prosthetic implants during the prosthesis reopening phase. Fix the simulation time period at 0.8 s, thereby enabling the deployment of stents within the aortic root of patients. Refer to Supplementary File 1, Supplementary Figure 2, and Supplementary Figure 3.

Results

Evaluation of the crimping effect on leaflet stress: general consideration
The stress values on prosthetic leaflets in the diastolic phase were obtained following the completion of the following steps: (a) Leaflet closure was achieved by applying uniform physiological pressure (0.005 MPa ≈80 mmHg) to their external surface. (b) Leaflet apposition was then facilitated by re-expansion of the stent frame. (c) Finally, the leaflets were included in a systolic configuration throughout the crimping phase (...

Discussion

The biomechanical performance of S3-TAV has been evaluated in relation to the device's technological advances, which represent a third generation of such devices. However, this analysis has also taken into account the ongoing challenges that remain in the implantation region and the geometry of the aortic annulus and root or solid bulky calcification. The present study demonstrates that the maximum and minimum principal stresses occurring in the 26 mm S3-TAV are localized to the cells situated immediately adjacent to...

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
Abaqus	Dassault Systèmes (Simulia, Providence, RI, USA)	2017	commercial finite element solver
Edwards SAPIEN 3 valve	Edwards Lifesciences, Inc. of Irvine, California	https://www.edwards.com/healthcare-professionals/products-services/transcatheter-heart/transcatheter-sapien-3-ultra	Sapien valve studied
ITK-SNAP	www.itksnap.org	Version 3.6	Identifying anatomy on pre-operative DICOM images
Matlab	Mathworks Inc., Natick, MA, USA)	v.R2018b	Generate a suitable mesh of the aortic wall
MicroCT-40	Scanco Medical AG, Basel, Switzerland		Desktop cone-beam micro-computed tomography scanner
Revolution Apex	GE Healthcare, Chicago, Illinois, USA		Dual-source computed tomography scanner
Rhinoceros	McNee and Associates, Seattle, WA, USA	Version 5.0	Construct a definitive 3D Computer-aided design