Transcatheter Aortic Valve Implantation (TAVI) is a minimally invasive technique for the treatment of severe aortic valve disease which consists in the deployment via catheter of a bio-prosthetic valve in place of the native aortic valve. Introduced twenty years ago as an alternative to surgery for elderly patients at high-risk for open-heart surgery, TAVI has been recently extended to lower-risk, younger patients thanks to comparable performances with the surgical gold-standard. In this context, it is particularly relevant to assess the long-term durability of the TAVI bio-prosthetic valves which is limited by Structural Valve Deterioration (SVD). SVD is an inevitable process manifested as intrinsic permanent changes in the prosthesis (calcification, wear and tear) which ultimately leads to the hemodynamic failure of the implanted valve, but its pathogenesis is still not completely understood. In particular, SVD has been associated with an increased infiltration of host immune cells in the bio-prosthetic leaflets and clinical risk factors for SVD suggest a possible influence of aortic hemodynamics on its development. Accordingly, in this thesis we aim to propose computational-based hemodynamic indices able to identify a premature onset of SVD. To this aim, we develop a highly personalized Computational Fluid Dynamics (CFD) model to efficiently simulate realistic early post-TAVI hemodynamics, i.e. when the implanted valve is still fully functioning, in order to study its relationship with SVD. Specifically, we consider a cohort of fourteen patients with and without SVD at long-term follow-up exam and, starting from pre-operative clinical images, we build post-TAVI virtual geometries, providing a realistic representation of the implanted TAVI valve. Personalized flow rate conditions, derived from Echo Doppler data, are used to perform CFD simulations exploiting an ad-hoc algebraic preconditioner proposed in this thesis. From a post-processing of the CFD results we propose novel computational hemodynamic indices and scores that show statistically significant differences between patients with and without SVD, proving a correlation between aortic hemodynamics features and SVD. Furthermore, based on the hemodynamic scores, we propose a synthetic scoring system that clearly separates the SVD from the non-SVD sub-group of patients, suggesting the our model captures some key biomechanical drivers of SVD. This study represents a preliminary proof of the influence of aortic hemodynamics on a premature onset of SVD, which has been poorly investigated by clinicians and researchers, showcasing the power of computational modeling in providing a detailed description of complex features that would be impossible to capture otherwise. Finally, the synthetic score could assist clinicians in a patient-specific planning of follow-up exams in order to closely monitor those patients at high-risk of prematurely developing SVD.
L'impianto valvolare aortico transcatetere (TAVI; dall'inglese) è una tecnica minimamente invasiva per il trattamento di una patologia severa della valvola aortica che consiste nel rilascio tramite catetere di una valvola bio-prostetica al posto della valvola aortica nativa. Introdotta venti anni fa come alternativa all'intervento chirurgico per pazienti anziani ad alto rischio per una chirurgia a cuore aperto, la TAVI è stata recentemente estesa a pazienti a minor rischio e più giovani grazie a prestazioni comparabili al gold-standard chirurgico. In questo contesto è particolarmente importante valutare la durabilità a lungo termine delle valvole bio-prostetiche TAVI che è limitata dal deterioramento strutturale della valvola (SVD; dall'inglese). SVD è un processo inevitabile che si manifesta con modifiche intrinseche e permanenti nella protesi (calcificazione, usura) risultanti nel fallimento emodinamico della valvola impiantata, ma la sua patogenesi non è stata ancora completamente capita. In particolare, SVD è stata associata con un'elevata infiltrazione di cellule immunitarie nei lembi bio-prostetici e i fattori clinici di rischio per SVD suggeriscono una possibile influenza dell'emodinamica aortica sul suo sviluppo. Di conseguenza in questa tesi miriamo a proporre indici emodinamici computazionali capaci di identificare una manifestazione prematura di SVD. A questo scopo sviluppiamo un modello di fluido-dinamica computazionale (CFD; dall'inglese) altamente personalizzato per simulare efficientemente un'emodinamica realistica immediatamente dopo la TAVI, i.e. quando la valvola impiantata è ancora pienamente funzionante, al fine di studiare la sua relazione con la SVD. Specificatamente, consideriamo un campione di quattordici pazienti con e senza SVD all'esame follow-up a lungo termine e, partendo da immagini cliniche pre-operative, costruiamo geometrie virtuali post-TAVI, fornendo una rappresentazione realistica della valvola TAVI impiantata. Condizioni di flusso personalizzate, derivate da dati Eco Doppler, sono utilizzate per eseguire simulazioni CFD sfruttando un precondizionatore algebrico ad-hoc proposto in questa tesi. Da un'analisi dei risultati CFD proponiamo nuovi indici e score emodinamici computazionali che mostrano differenze statisticamente rilevanti tra pazienti con e senza SVD, provando una correlazione tra le caratteristiche dell'emodinamica aortica e la SVD. Inoltre, basandoci sugli score emodinamici, proponiamo un sistema di score sintetico che separa chiaramente i sottogruppi di pazienti con e senza SVD, suggerendo che il nostro modello catturi alcuni fattori biomeccanici chiave della SVD. Questo studio rappresenta una prova preliminare dell'influenza dell'emodinamica aortica su una manifestazione prematura di SVD, che non è stata propriamente investigata da clinici e ricercatori, mettendo in luce l'abilità dei modelli computazionali nel fornire una descrizione dettagliata di fenomeni complessi che sarebbe impossibile catturare altrimenti. Infine, lo score sintetico potrebbe assistere i medici in una pianificazione paziente specifica degli esami follow-up al fine di monitorare strettamente quei pazienti che si trovano ad alto rischio di sviluppare prematuramente la SVD.
Personalized computational study of hemodynamics in Transcatheter Aortic Valve Implantation: assessment of long-term degeneration
CRUGNOLA, LUCA
2026
Abstract
Transcatheter Aortic Valve Implantation (TAVI) is a minimally invasive technique for the treatment of severe aortic valve disease which consists in the deployment via catheter of a bio-prosthetic valve in place of the native aortic valve. Introduced twenty years ago as an alternative to surgery for elderly patients at high-risk for open-heart surgery, TAVI has been recently extended to lower-risk, younger patients thanks to comparable performances with the surgical gold-standard. In this context, it is particularly relevant to assess the long-term durability of the TAVI bio-prosthetic valves which is limited by Structural Valve Deterioration (SVD). SVD is an inevitable process manifested as intrinsic permanent changes in the prosthesis (calcification, wear and tear) which ultimately leads to the hemodynamic failure of the implanted valve, but its pathogenesis is still not completely understood. In particular, SVD has been associated with an increased infiltration of host immune cells in the bio-prosthetic leaflets and clinical risk factors for SVD suggest a possible influence of aortic hemodynamics on its development. Accordingly, in this thesis we aim to propose computational-based hemodynamic indices able to identify a premature onset of SVD. To this aim, we develop a highly personalized Computational Fluid Dynamics (CFD) model to efficiently simulate realistic early post-TAVI hemodynamics, i.e. when the implanted valve is still fully functioning, in order to study its relationship with SVD. Specifically, we consider a cohort of fourteen patients with and without SVD at long-term follow-up exam and, starting from pre-operative clinical images, we build post-TAVI virtual geometries, providing a realistic representation of the implanted TAVI valve. Personalized flow rate conditions, derived from Echo Doppler data, are used to perform CFD simulations exploiting an ad-hoc algebraic preconditioner proposed in this thesis. From a post-processing of the CFD results we propose novel computational hemodynamic indices and scores that show statistically significant differences between patients with and without SVD, proving a correlation between aortic hemodynamics features and SVD. Furthermore, based on the hemodynamic scores, we propose a synthetic scoring system that clearly separates the SVD from the non-SVD sub-group of patients, suggesting the our model captures some key biomechanical drivers of SVD. This study represents a preliminary proof of the influence of aortic hemodynamics on a premature onset of SVD, which has been poorly investigated by clinicians and researchers, showcasing the power of computational modeling in providing a detailed description of complex features that would be impossible to capture otherwise. Finally, the synthetic score could assist clinicians in a patient-specific planning of follow-up exams in order to closely monitor those patients at high-risk of prematurely developing SVD.| File | Dimensione | Formato | |
|---|---|---|---|
|
PhD_thesis_LucaCrugnola.pdf
accesso aperto
Licenza:
Tutti i diritti riservati
Dimensione
30 MB
Formato
Adobe PDF
|
30 MB | Adobe PDF | Visualizza/Apri |
I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/20.500.14242/376608
URN:NBN:IT:POLIMI-376608