# Non-invasive versus invasive estimation of left ventricular wall stress with cardiac magnetic resonance imaging in severe aortic stenosis

## TL;DR

This study compares non-invasive and invasive methods to estimate heart wall stress in patients with severe aortic stenosis using MRI and finds strong agreement between the two.

## Contribution

A fully non-invasive method for estimating left ventricular wall stress using cuff pressure and echocardiographic data is validated against invasive measurements.

## Key findings

- Non-invasive and invasive LVWS were strongly correlated (r = 0.89, p < 0.001).
- Bland–Altman analysis showed a small positive bias with narrow limits of agreement.
- Lower non-invasive LVWS was associated with better 6MWT performance.

## Abstract

Left ventricular wall stress (LVWS) is a key determinant of myocardial oxygen demand in aortic stenosis (AS), but CMR-based assessment typically requires invasive left ventricular (LV) pressure. We evaluated a fully non-invasive method to quantify systolic LVWS in severe AS.

In 35 TAVR candidates with severe AS who underwent CMR, circumferential LVWS was calculated using a cylindrical LV model across seven short-axis slices. End-systolic LV pressure was estimated as brachial systolic blood pressure plus echocardiographic mean transvalvular gradient. Circumferential LVWS from this cuff-based approach was compared with LVWS calculated using simultaneously acquired invasive LV pressures. Agreement was assessed using Pearson correlation, linear regression, and Bland–Altman analysis. Six-minute walk test (6MWT) distance was compared between patients above and below the median non-invasive LVWS.

Mean age was 80.1 ± 6.2 years and 74.3% were male. Non-invasive and invasive circumferential LVWS were strongly correlated (r = 0.89, p < 0.001) with a regression slope (β) of 0.94. Bland–Altman analysis showed a small positive bias (+0.65 Pa) and 95% limits of agreement from –9.8 to +11.1 Pa. Patients in the lower 50% of non-invasive LVWS had greater 6MWT distance than those in the upper 50% (median [IQR] 352 [224–383] vs 215 [176–279] m, p = 0.046).

Non-invasive CMR-based estimation of circumferential LVWS using cuff pressure and echocardiographic gradient closely approximates invasively measured LVWS and retains functional discriminative value. This pragmatic approach may facilitate broader research into the application of LVWS as a non-invasive marker of early LV decompensation in AS.

ga1