Leaflet stresses during crimping simulations of a CoreValve Evolut Pro model using a compression loading system

Abstract

Background: Transcatheter aortic valve implantation is experiencing continued growth as an option for the treatment of aortic stenosis. With larger numbers of procedures being performed on lower risk and younger patients, there is increased scrutiny on valve durability. Leaflet stresses and potential damage have a significant role to play in this regard. Predictions of leaflet stresses have so far focused on either fluid-structure interaction simulations of blood flow through the prosthesis or, crimping simulations using a cylindrical surface. However, in reality, when a compression loading system (CLS) is used in the crimping of self-expanding valves, this could result in different stresses in the valve leaflets relative to those that might occur in crimping with a cylindrical surface.

Method: A full model of a CoreValve Evolut Pro (Medtronic, Minneapolis, MN, USA) device was developed, comprising the frame, skirt and leaflets along with a representative model for the CLS as used in clinical practice. The full device was crimped to a final diameter of 18 Fr using the CLS model and the distribution and intensity of leaflet stresses was assessed. A similar assessment of leaflet stresses was also performed for crimping using radial displacement of a cylindrical surface. Comparison of the predicted leaflet stresses between the two models was undertaken, alongside a comparison of the stresses produced when dynamically loading the leaflets after deployment of the valve.

Results: Both the CLS and cylinder crimping methods produced higher average and peak stresses on the leaflets compared to those produced during leaflet loading. The peak von Mises stresses for CLS crimping, cylinder crimping, and leaflet loading were 3.42 MPa, 3.92 MPa, and 1.77 MPa respectively. The leaflet folding pattern between the CLS crimping and cylinder crimping methods were different, resulting in different high stress locations on the leaflets. However, the average stress magnitude at the final crimped stage between the two crimping methods were similar.

Conclusions: High fidelity simulations of crimping and expansion of a complete CoreValve Evolut Pro model using a compression loading system model have been performed, wherein the results showed that peak leaflet stresses in the crimped valve were approximately twice as high as the maximum leaflet stresses under dynamic loading. This finding has significant implications for device durability due to the high stresses and possible damage they might inflict on the leaflets. It was also found that crimping using a compression loading system versus a simpler cylindrical surface produced different folding patterns and stress distributions. However, for future studies that are not concerned with accurately capturing the leaflet folding patterns and stresses throughout the crimping process, crimping via a cylindrical surface can be used instead of simulating the full CLS model.