The general objective of this study is to compare in vitro and in vivo the hemodynamic performance of different types of aortic valve prostheses and possible surgical implantation techniques.

The study involves the collection of data deriving from tests and procedures foreseen by normal clinical practice, including a transthoracic color Doppler echocardiogram, to which is added a 4DFlow cardiac MRI (magnetic resonance imaging) without contrast medium, to be performed between 1 and 3 months after the operation.

The study will last 24 months and the patients involvement will end approximately 1-3 months after the intervention.

There are no direct benefits expected from participation in the study. However, the instrumental data obtained by echocardiogram and Magnetic Resonance Imaging will be useful for monitoring the medium-term surgical outcome of patients after surgery.

The goal of the study is testing hydrodynamic performances of various surgical and transcatheter valve prostheses, also according to the technique of implantation. A dedicated Magnetic Resonance Imaging (MRI) compatible mock circulation simulator will be implemented, including a circuit filled with blood-alike fluid, a pulsatile pump, and 3D printed silicon phantoms of ascending aorta.

Data collection will focus on structural changes of prostheses in response to flow, including leaflet motion, gradients and flows across the valve, taken either by direct pressure measurements and 4D Flow MRI. As second step, trans-valvular gradients and valve performance in patients implanted with the same tested valves will be measured by echocardiogram and 4D Flow MRI.

Degenerative aortic valve stenosis is a common disease in people >65 years, with an estimated rate approaching 10% in Western countries. Despite progress in diagnosis and therapy, invasive treatments still represent the only definitive option, including surgical aortic valve replacement (SAVR) and more recently less-invasive transcatheter aortic valve replacement (TAVR).

Patients with small aortic annulus constitute a challenging subset, where the need to minimize both procedural and longterm risk entwine with that of avoiding patient-prosthesis mismatch (PPM) and structural valve deterioration (SVD). The common denominator appears to be blood flow, which has demonstrated to play a pivotal role in late complications. In fact, PPM is associated with earlier structural valve dysfunction and higher long-term mortality. To obtain larger effective orifice area, surgeons must consider many different procedures (annular enlargement), valve models (stented valves, subcoronary stentless valves, or rapid deployment valves) or different techniques of valve suturing. However, in the literature, there is no clear indication to the ideal strategy of treatment so far. Heart valve prostheses generate turbulent flow in the aortic root and the ascending aorta. Hydrodynamic instabilities and turbulent flow may produce clinically adverse events. Therefore, a thorough evaluation of hemodynamic and flow conditions should be tested, to provide the most effective treatment strategy for each patient.

Both units regularly perform surgical as well as percutaneous aortic valve replacements: patients from both centres will be enrolled to evaluate post-implant valve performances by echocardiogram and MR imaging.

Both units will have an equal contribution with regard to study design, collection of patients, data, analysis and interpretation of data, drawing of manuscript.

In addition, Unit 1 will set up and run the mock circulation system for in-vitro tests.

Valve performance can be tested using advanced cardiovascular imaging techniques combined with in vitro testing platforms. On the one hand, three-dimensional time-resolved MRI with three-directional velocity encoding (4D Flow MRI) offers the possibility to assess blood flow patterns enabling full volumetric coverage of the region of interest, a detailed 3D visualization of the flow pattern and post hoc quantification of several hemodynamic metrics. On the other hand, the in vitro setting is the gold-standard approach to consistently investigate the blood pressure variability (BPV) hemodynamics in a fully controlled environment, through ad hoc system sensorization, under a standardized setup and protocol of measurements, with no influence by confounding in vivo factors. The combination of these two techniques allows for the characterization of selected heart valves prostheses.

The 3D and Computer Simulation Laboratory (in short C3DLab) is equipped with facilities specifically devoted to 3D medical image processing, including dedicated commercial softwares for 3D anatomic reconstructions (e.g., Mimics Innovation Suite, Materialise). Dedicated workstations (Fujitsu CELSIUS J550/2, Intel® Core i7-7700 CPU 3.60 GHz with 32 GB RAM) are available for pre- and post-processing. C3DLab facilities also include a high-end 3D printer (MJP2500 Plus, by 3D Systems) to obtain deformable and suturable phantoms of cardiovascular organs and a highly accurate 3D scanner (Artec Micro, by Artec 3D) to obtain the digital model of implantable devices or small cardiovascular structures. MRI, and specifically 4D Flow MRI, can be performed at IRCCS Policlinico San Donato using a 1.5 T MAGNETOM Aera machine (Siemens Healthcare, Erlangen, Germany) scanner.

Specific aim 1 To classify and compare valves according to actual measured size and haemodynamic performance (pressure gradients and derived measures), rather than label size.

In the surgical group, the two most common techniques of implantation will be compared: intra-annular vs supra-annular positioning, obtained by placing interrupted everting and non-everting pledgeted sutures, respectively. Valve prostheses will be divided in three groups and tested accordingly:

• Tissue valves: stented pericardial.
• Sutureless/rapid deployment.
• Bi-leaflet mechanical valves. In the Transcatheter Aortic Valve Implantation (TAVI) group, valve position depends on the prosthesis of choice. Moreover, at the time of valve deployment, commissural alignment should be granted to avoid flow impairment to coronary ostia, due to major overlap of the prosthetic commissures. This could be a near fatal event and leads to a redo-TAVI procedure. Furthermore, recent evidence suggests that commissural alignment after TAVI may have an impact on valve hemodynamics and durability.

Transcatheter valve prostheses will be divided in two major groups:

• Self-expandables (SE)
• Balloon-expandables (BE)
• Both valve groups will be evaluated in the valve-in-valve position also.

Specific aim 2 The investigator's assumption is that hemodynamic performance may not be the only factor to consider in valve choice, but others may influence long-term durability and risk of device-related complications such as early degeneration, thrombo-embolic events, and endocarditis.

Therefore, analysis and classification of tested prosthetic valves will be carried out, by flow patterns, shear forces and turbulences acting through the valve opening, structures and on the leaflets.

Three-dimensional time-resolved MRI with three-directional velocity encoding (4D Flow MRI) offers the possibility to assess blood flow patterns enabling full volumetric coverage of the region of interest.

The 4D Flow MRI sequence includes a prototype time-resolved 3D gradient echo sequence with three-directional velocity encoding. Several fluid dynamic parameters can be extracted from 4D Flow MRI, using advanced numerical schemes, including pressure gradients, kinetic energy and viscous energy loss. Additionally, streamlines and pathlines can be used to visually assess blood flow trajectories.

Specific aim 3 In-vivo validation of results. Measurements will be taken from an adequate number of patients implanted with the same valve types and models that were tested in-vitro.

Transthoracic echocardiography is an easy, cheap, and widely adopted technique, representing the standard of care in postoperative follow-up of patients with prosthetic valves. The critical aspect for complete prosthetic evaluation are:

• Morphology (flail/avulsed, thickening, calcification, pannus) and/or mobility (excessive, reduced) of the bioprosthetic valve leaflets and correct prosthetic stent position.
• Hemodynamic data: integrate evaluation of peak and mean transvalvular gradients, effective orifice area (EOA) by the continuity equation and Doppler velocity index (DVI) to evaluate correct prosthetic function and differentiate between valve dysfunction and PPM in case of elevated trans-valvular gradient prosthetic regurgitation: distinguish between intra-valvular jet (rare) and paravalvular leak (PVL - more frequent) combining a number of Doppler parameter: number, location and circumferential extent of PVL velocity waveform density and jet deceleration rate (CWD), diastolic flow reversal (PWD) in descending aorta/ abdominal aorta, vena contracta width and area (2D/3D CD)
• Quantitative analysis: regurgitant volume, regurgitant fraction and Effective regurgitant orifice area (EROA).

Regarding 4D Flow MRI, collected parameters will be similar to those analysed in-vitro: pressure gradients, kinetic energy, viscous energy loss, as well as streamlines and pathlines.

The investigators hypothesise that the system allows to predict flow dynamics and its alterations, better than those already existing, and more closely reflects the haemodynamic performance of in-vivo implants.

The aim is to give an unbiased insight on which treatment is best in different subsets of patients, to help guide clinical decisions.

Inclusion Criteria:

• patients who underwent SAVR or TAVI for severe aortic stenosis
• implanted with valve models and sizes among those which we evaluated in vitro

Exclusion Criteria:

• treated or untreated aortic root/ascending aorta aneurysm
• preoperative aortic regurgitation more than moderate
• recent (<12 weeks)/ concomitant revascularization
• recent (<12 weeks) ACS
• PPI (permanent pacemaker implantation)
• other than sinus rythm
• Paravalvular leak (PVL) > moderate.