Investigating thrombosis formation in replacement aortic valves
Project reference: DTP_SIE_17
First supervisor: Jack Lee
Second supervisor: Tiffany Patterson, GSTT
Start date: February 2022 or October 2022
Project summary: Aortic stenosis affects 2% of the population over 65 years of age, that rises to >10% in over-75s, and carries a poor 50% mortality rate at two years.
Project description: The introduction of Transcatheter Aortic Valve Implantation (TAVI) in the past decade has provided a revolutionary alternative to Surgical Aortic Valve Replacement (SAVR) and has seen a rapid growth in the number of procedures. Following advancements in device design and procedural optimisation, the clinical outcomes of TAVI has acquired an excellent outlook, with risk of stroke and mortality rates reaching that of the normal population 3 months after the procedure. However, recent reports indicate that in around 15% of the TAVI recipients, leaflet thrombosis and reduced mobility is found, as reliably detected by HALT (Hypo-Attenuation and Leaflet Thickening) in CT imaging. Valve thrombosis increases risk of major cardiovascular events, and even if sub-clinical in the short term, can adversely affect the long-term durability of the implanted valves which would limit the expansion trans-catheter therapies to younger, lower surgical risk population. To this end, FDA have released a statement calling attention to this problem. Though oral anti-coagulation therapies are an option, currently there is no way to gauge the risk of valve thrombosis in patients, potentially raising the risk of bleeding and adverse outcomes.
Hypothesis: High transvalvular gradient at post-TAVI procedure is correlated with the development of valve thrombosis, modulated by a multifactorial interplay between the patient anatomy, bioprosthetic valve geometry, and haemodynamics.
To predict patients at risk of leaflet thrombosis from standard clinical imaging data
To identify optimal valve configuration to minimise thrombosis formation
This project will be conducted alongside a clinical investigation in which markers of leaflet thrombosis will be temporally recorded in patients of differing transvalvular hemodynamics. Haemodynamic modelling will enable a deeper quantitative analysis of data and predictive simulations. The clinical proposal is currently in review with the BHF.
The clinical data will be augmented through building ensembles of haemodynamic simulations, by permuting from available patient anatomies and known valve geometries (annulus diameter 23-29mm), deployed in various orientations. The routine clinical data includes pre-TAVI contrast enhanced CT and post-TAVI echocardiograms (n>50) from which the anatomy and boundary conditions can be estimated. Segmentation and meshing pipeline will be set up based on our previous work in biventricular image analysis, and CFD simulations will be performed with our in-house code. Wall shear stress, turbulence and kinetic energy indices will be determined from the output of simulations.
A comprehensive collection of features will be quantified from the full modelling pipeline, including anatomical (annular size, annular geometry, calcification, distance between aortic annulus and coronary ostia, dimensions of sinus of Valsalva, LVOT size and orientation) and derived haemodynamic parameters. These parameters will be correlated with the degree of HALT observed in longitudinal CT scans performed in TAVI recipients. Spatial localisation of HALT on leaflets with respect to the features will also be investigated.