Invasive validation of a novel approach to determine aortic valve area with phase-contrast cardiac magnetic resonance imaging

Abstract

Abstract Funding Acknowledgements Type of funding sources: Public grant(s) – National budget only. Main funding source(s): Austrian Society of Cardiology Background. Echocardiography is considered the standard method for screening and diagnosing aortic valve stenosis. However, inaccuracies in the determination of stroke-volumes by the continuity equation might particularly make the evaluation of patients with low-flow states difficult. Phase-contrast cardiac magnetic resonance (PC-CMR) is a promising tool in overcoming these limitations by the simultaneous determination of flow volumes and velocities across the stenotic valve. Purpose The aim of this study is to validate a novel approach based on PC-CMR against the invasive determination of the aortic valve area (AVA). Methods. PC-CMR was performed in 50 patients with moderate or severe AS (n = 52; age 72 years [interquartile range (IQR) 66 - 78], 38% of patients with low-flow states). All of them were referred to invasive evaluation of aortic stenosis by cardiac catheterization. Additionally, transthoracic echocardiography (TTE) was performed. Aortic valve area (AVA) was determined by PC-CMR (AVAPC-CMR) via plotting momentary flow across the valve against momentary flow velocity. AVAPC-CMR at different time points over the entire cardiac cycle was compared to invasively determined AVA, calculated according to the Gorlin-formula. Stroke volumes (SV) were determined by the Fick-principle, pressure gradients according to the modified Bernoulli-equation. Results. Mean AVA during the whole systolic phase showed a good correlation (r: 0.544, p < 0.001) with invasive AVA with a small bias (AVACMR: 0.78 cm², IQR: [0.60-0.96] versus AVAINVASIVE: 0.70 cm², IQR: [0.52-0.87], bias: 0.08 cm², p = 0.017). Intermethodical correlation and bias of AVA as measured by TTE (AVATTE) and AVAINVASIVE were similar to AVAPC-CMR (AVATTE: 0.81 cm²; IQR: [0.64-0.96] versus AVAINVASIVE: 0.70 cm², IQR: [0.52-0.87] r: 0.580, p < 0.001, bias 0.11 cm², p < 0.001). SV by PC-CMR showed a good correlation with Cine-CMR with no significant bias (r: 0.730, p < 0.001; SVPC-CMR: 86 ± 31 ml; SVCine: 85 ± 19 ml). Maximum gradients determined by PC‑CMR were 65 ± 2 9mmHg and showed a good inverse correlation with AVAPC-CMR (r: ‑0.371; p = 0.008). Conclusion. PC-CMR with continuous determination of flow volumes and flow velocities is able to determine AVA in patients with severe aortic stenosis with a tendency to overestimate AVA compared to invasively determined AVA.