Non-invasive cardiovascular magnetic resonance assessment of pressure recovery distance after aortic valve stenosis

Abstract

Background: Decisions in the management of aortic stenosis are based on the peak pressure drop, captured by Doppler echocardiography, whereas gold standard catheterization measurements assess the net pressure drop but are limited by associated risks. The relationship between these two measurements, peak and net pressure drop, is dictated by the pressure recovery along the ascending aorta which is mainly caused by turbulence energy dissipation. Currently, pressure recovery is considered to occur within the first 40-50 mm distally from the aortic valve, albeit there is inconsistency across interventionist centers on where/how to position the catheter to capture the net pressure drop.

Methods: We developed a non-invasive method to assess the pressure recovery distance based on blood flow momentum via 4D Flow cardiovascular magnetic resonance (CMR). Multi-center acquisitions included physical flow phantoms with different stenotic valve configurations to validate this method, first against reference measurements and then against turbulent energy dissipation (respectively n = 8 and n = 28 acquisitions) and to investigate the relationship between peak and net pressure drops. Finally, we explored the potential errors of cardiac catheterisation pressure recordings as a result of neglecting the pressure recovery distance in a clinical bicuspid aortic valve (BAV) cohort of n = 32 patients.

Results: In-vitro assessment of pressure recovery distance based on flow momentum achieved an average error of 1.8 ± 8.4 mm when compared to reference pressure sensors in the first phantom workbench. The momentum pressure recovery distance and the turbulent energy dissipation distance showed no statistical difference (mean difference of 2.8 ± 5.4 mm, R2 = 0.93) in the second phantom workbench. A linear correlation was observed between peak and net pressure drops, however, with strong dependences on the valvular morphology. Finally, in the BAV cohort the pressure recovery distance was 78.8 ± 34.3 mm from vena contracta, which is significantly longer than currently accepted in clinical practise (40-50 mm), and 37.5% of patients displayed a pressure recovery distance beyond the end of the ascending aorta.

Conclusion: The non-invasive assessment of the distance to pressure recovery is possible by tracking momentum via 4D Flow CMR. Recovery is not always complete at the ascending aorta, and catheterised recordings will overestimate the net pressure drop in those situations. There is a need to re-evaluate the methods that characterise the haemodynamic burden caused by aortic stenosis as currently clinically accepted pressure recovery distance is an underestimation.

Keywords: 4D Flow MRI; Aortic stenosis; Flow momentum; Non-invasive pressure drop; Pressure recovery; Turbulence.

Conflict of interest statement

P. Lamata is member of the Scientific Advisory Board of Ultromics Ltd, UK, and receives a compensation for it. P. Lamata and A. Faraci are shareholders at Congenita Ltd, UK. P. Mortier is the CTO and V. Galli was an employee at FEops NV. No other author have any other financial interest.

© 2023. The Author(s).

Figures

Fig. 1

Definition of the pressure concepts in the context of aortic stenosis. Blood flows from left ventricular outflow tract (LVOT) into the ascending aorta (AAo) across the aortic valve that defines the geometric orifice area (GOA). When crossing the valve, a portion of the potential energy of the fluid is converted into kinetic energy, which is maximal at the effective orifice area (EOA) defining the peak pressure drop (ΔPpeak) where the ejection jet is narrowest. Downstream along the AAo the kinetic energy is transformed back to potential energy until a fully developed laminar flow profile where is the correct point to measure the net pressure drop (ΔPnet). The pressure recovery (ΔPrec) is the difference between ΔPpeak and ΔPnet. The pressure recovery distance (PRecDist) is defined as the distance necessary for the blood flow to go from the ΔPpeak to ΔPnet

Fig. 2

Illustration of the definition of the pressure recovery distance based on momentum (PRecDist-M; accessed via simplified advective work–energy relative pressure formulation—SAW) or based on turbulent dissipation (PRecDist-T) relative to the effective orifice area (EOA), based on an exemplary experimental result in our phantom. The distance is defined from the point of the effective orifice area (when momentum is greatest), until the recovery of 95% either the momentum created (in PRecDist-M) or until the accumulation of 95% of turbulent dissipation (in PRecDist-T )

Fig. 3

Workbench 1—flow phantom with 8 pressure sensors (channel 1 to 8, CH1-CH8), with their location relative to the aortic valve (AV point at X location = 0), and the respective measurement of pressure recovery distance (PRecDist). A A slice of the magnitude of the CMR image, showing most of the locations of the eight pressure ports. B Measurement of the pressure recovery distance from the measured pressure acquired (ΔPmeasured) and interpolated from the eight sensors. C Advective pressure component (ΔPSAW) from the 4D flow CMR data and from simulation (computed fluid dynamics, CFD) along the centerline of the phantom with 100 ml/s constant flow. Note that sensor 8 is located at 500 mm after aortic valve (AV) and is not being presented in the diagram

Fig. 4

Workbench 2, a rigid pipe with constant flow and different 3D-printed heart valves to simulate aortic valve stenosis (AS). A The valve geometries considered—tricuspid aortic valve (TAV), two different configurations of bicuspid aortic valves (BAV1 and BAV2), two circular valves with different diameters (Circ 1 and Circ2), and two malfunctioning prosthetic heart valves (PHV1 and PHV2) [29]. B The respective cross-sectional blood flow profile at vena contracta (VC) cross-plane. C The respective streamlines distal from the valve

Fig. 5

Agreement of pressure recovery distance (PRecDist) estimated from momentum computation (PRecDist-M) and measured PRecDist in workbench 1 including 3 constant (blue) and 5 pulsatile flow regimes (red) (total n = 8). A Linear regression analysis; B Bland–Altman plot for the agreement

Fig. 6

Agreement of pressure recovery distance (PRecDist) estimated from momentum computation (PRecDist-M) and turbulence-based (PRecDist-T) in workbench 2 including 4 constant flow regimes across 7 different valves (n = 28). A Linear regression analysis; B Bland–Altman plot for the agreement

Fig. 7

Relationship between pressure recovery (i.e., reduction of advection) and energy loss (i.e., turbulence dissipation) across the 28 experimental conditions of workbench 2, colour coded accordingly to their stenotic level (Peak advective pressure drop; ΔPSAW). Each line is built by the amount of advection (by ΔPSAW, in mmHg, Y axis) and the dissipation accumulated (by turbulence, in mmHg, X axis) at each point along the centreline of the vessel phantom. The low stenotic lines (i.e. weaker momentum in the jet) correspond to the more irregular (i.e. noisy) relationships. On opposition, the larger the stenotic level, the better the agreement. Furthermore the agreement is better as more momentum has been recovered

Fig. 8

Relationship between peak and net pressure drop (ΔP), given by the comparison between the peak advective pressure drop (SAW) at the EOA and the total turbulence dissipation given by work–energy relative pressure including the turbulent dissipation component (WERP-T) for all the valves (left panel) and each valve undergoing four different flow conditions (right panels). The overall agreement (left panel) is good, but each valve (right panels) reports quite different linear regression coefficients between peak advection and net pressure drop caused by turbulent dissipation, indicating that each valve has its own peak versus net pressure drop signature. The axis of the individual valve plots has the same units as the aggregate plot

Fig. 9

Relationship between pressure recovery distance via momentum recovery (PRecDist-M) and studied factors in 32 bicuspid aortic valve subjects. A Linear relationship with the magnitude of aortic stenotic burden (assessed by the simplified advective work energy relative pressure, ΔPSAW), and identification of two subgroups that do not fit well the model. B Lack of relationship with remaining factors considered and that have been reported to influence net pressure drop: radius, eccentricity angle (Ecc Angle) and eccentricity displacement (Ecc disp.) analysed at the effective and geometric orifice area location (respectively, EOA and GOA locations). Note that the patients presenting a single plateau of pressure recovery are presented in blue and those with double pressure recovery plateau are presented in red

Fig. 10

Example of a double plateau case (A), as well as a mild coarctation (CoA) located at plane c (B). Both cases display the advective momentum (ΔPSAW) along the centerline, the velocity profiles in 3 cross-section planes: (a) effective orifice area (EOA), (b) distal ascending aorta and (c) aortic arch; and the 3D geometry with the streamlines seeded from the same 3 cross-section planes