Simulation of Balloon Post-Dilatation following Self-Expanding Transcatheter Aortic Valve Implantation in Challenging Cases of Bicuspid Aortic Valve Stenosis

Introduction: : The bicuspid aortic valve (BAV) is a common congenital disease in which the aortic valve presents with two functional cups instead of the typical three. Recent studies report that up to 50% of younger patients undergoing aortic valve replacement present with BAV morphology and transcatheter aortic valve implantation (TAVI) is increasingly being used as the preferred treatment [1]. However, TAVI in BAV patients poses significant challenges due to anatomical features such as asymmetrical and heavily calcified cusps, commissural fusions, and concomitant aortopathy [2]. Accurate device sizing is necessary to ensure optimal prosthesis-anatomy interaction and prevent paravalvular leakage (PVL) [3]. Although PVL rates have reduced in recent years [3], balloon post-dilatation is frequently required to resolve incomplete valve apposition [2]. While several BAV-specific sizing strategies have been proposed [2-4], standardized sizing protocols remained poorly defined and not widely adopted.
Patient-specific computational modelling, including finite element (FE) and computational fluid dynamics (CFD) approaches, have shown promise in optimizing TAV sizing and positioning in the BAV cohorts [5,6]. These tools offer the potential to systemically evaluate various sizing strategies and assess the impact of balloon post-dilatation on the occurrence of PVL; however, such comprehensive frameworks have yet to be fully developed.
The objective of this study is to develop, verify, and validate a computational framework for simulating TAVI in BAV utilising clinical and in vitro data. Using this validated approach, we investigate the impact of balloon post-dilatation and device sizing on the occurrence of PVL following TAVI in BAV cohorts.

Materials and

Methods: : A retrospective, patient-specific model of a type-1 BAV with severe aortic stenosis was reconstructed from pre-TAVI CT. The aortic root was segmented and prestress applied via incremental internal pressure until diastolic conditions were achieved. Native leaflets were generated based on reference points along the aortic lumen and modeled as a hyperelastic-isotropic material. Calcifications on both the native leaflets and aortic root (total volume = 1800 mm3) were assumed to be an elasto-plastic material.
To simulate TAVI, we created a FE model of the self-expanding ACURATE Prime transcatheter aortic valve (TAV) (Boston Scientific) using Abaqus/Explicit. The device consists of porcine pericardium, modeled as a hyperelastic-anisotropic material, and nitinol stent frame, a phase-transforming, superelastic alloy [7]. The simulation workflow included: (1) crimping to 10 mm, (2) ‘top-down’ deployment inside the patient-specific model by sequential displacement of an upper and lower catheter (Figure 1a), and (3) hypertensive cardiac cycle conditions.
To assess the use of post-dilatation, we created a FE model of a non-compliant balloon, which was validated against bench testing [8]. To simulate post-dilatation, we applied an internal surface pressure on the balloon (3 atm), which was positioned inside the deployed TAV (Figure 1c). Hard-penalty contact simulated interactions between the TAV device, native anatomy, and balloon during inflation, after which balloon contact was removed to allow for device recoil.
Model validation was performed using post-TAVI clinical data and in vitro testing. Specifically, hydrodynamic performance was evaluated in a pulse duplicator, where the device was deployed within a patient-specific 3D-printed model (Figure 1b).

Results, Conclusions, and Discussions:: Our models were validated against in vitro testing using a 3D-printed, patient-specific annulus and predicted similar geometric orifice area (1.68 cm2 vs. 1.75±0.075 cm2) and post-stent deflection (P1: 1.71 mm vs. 1.73±0.26 mm, P2: 1.13 mm vs. 1.10 ± 0.08 mm, P3: 0.86 mm vs. 1.12±0.05 mm).
Results show that our model predicts a more eccentric deployment for direct-TAVI (without balloon post-dilatation), compared to the post-procedural CT data at the inflow, annular, and outflow levels respectively (48.5% vs. 26.1%, 45.2% vs. 23.7%, 36.7% vs. 30.4%). The predicted leakage gap area, located between the native anatomy and the TAV, was 62 mm2. In contrast, with balloon post-dilatation, our model predicted less eccentric deployment than direct-TAVI, which more closely aligned with recorded post-procedural CT (32.3% inflow, 28.2% annular, and 28.5% outflow, Figure 1d). The leakage gap area post-balloon inflation was substantially reduced to ~14 mm2, representing a noticeable reduction (~4x) when compared to direct-TAVI (Figure 1e). This reduction is consistent with clinical observations, where balloon post-dilatation has been associated with decreased PVL, and improved short- and long-term outcomes [9].
In conclusion, here we developed and validated a novel framework for predicting the performance of TAVI in a patient-specific case of severe BAV stenosis. We reported that balloon post-dilatation reduced leakage gap area by nearly 80% in this anatomically challenging case of BAV stenosis potentially lowering the risk of post-TAVI PVL. While we assumed that leakage gap area is proportional PVL risk, our ongoing work is assessing this using CFD. We are also investigating the predictive value of BAV-specific sizing strategies [2], and future efforts will focus on integrating deep learning-based methods for anatomical analysis and reconstruction within this modeling framework [10].

Acknowledgements and/or References (Optional):: Acknowledgements
Funded by the IRC Government of Ireland Postgraduate Scholarship (GOIPG/2022/2032).

References
[1] Roberts et al., Circulation, 2005. [2] Xiong et al., Nat Rev Cardiol, 2023. [3] Tchetche et al., Circ-Cardiovasc Inte, 2019. [4] Petronio et al., J Cardiovasc Comput, 2020. [5] McGee at al., Biomech Model Mechan, 2019. [6] Lavon et al., Med Biol Eng Comput, 2019. [7] Armfield et al, JMBBM, 2024. [8] Sathananthan et al, Catheter Cardio Inte, 2020. [9] Deneault et al., Circulation: Cardiov Interv, 2013. [10] Rouhollahi et al., Comput Med Imag Grap, 2023.