Rest and exercise-stress estimated pulmonary capillary wedge pressure using real-time free-breathing cardiovascular magnetic resonance imaging.

BACKGROUND: Identification of increased pulmonary capillary wedge pressure (PCWP) by right heart catheterisation (RHC) is the reference standard for the diagnosis of heart failure with preserved ejection fraction (HFpEF). Recently, cardiovascular magnetic resonance (CMR) imaging estimation of PCWP at rest was introduced as a non-invasive alternative. Since many patients are only identified during physiological exercise-stress we hypothesised that novel exercise-stress CMR derived PCWP emerges superior compared to its assessment at rest.

METHODS: The HFpEF-Stress Trial prospectively recruited 75 patients with exertional dyspnoea and diastolic dysfunction who then underwent rest and exercise-stress RHC and CMR. HFpEF was defined according to PCWP (overt HFpEF ≥15mmHg at rest, masked HFpEF ≥25mmHg during exercise-stress). CMR derived PCWP was calculated based on previously published formulae using LV mass (LVM) and either biplane left atrial volume (LAV) or monoplane left atrial area (LAA).

RESULTS: LAV (rest/stress: r=0.50/r=0.55, p<0.001) and LAA PCWP (rest/stress: r=0.50/r=0.48, p<0.001) correlated significantly with RHC-derived PCWP whilst numerically overestimating PCWP at rest and underestimating PCWP during exercise-stress. LAV and LAA PCWP showed good diagnostic accuracy to detect HFpEF (AUC LAV rest 0.73, stress 0.81; LAA rest 0.72, stress 0.77) with incremental diagnostic value for the detection of masked HFpEF using exercise-stress (AUC LAV rest 0.54 vs stress 0.67, p=0.019, LAA rest 0.52 vs stress 0.66, p=0.012). LAV but not LAA PCWP during exercise-stress was a predictor for 24 months hospitalisation independent of a medical history for atrial fibrillation (HR 1.26, 95% CI 1.02-1.55, p=0.032).

CONCLUSIONS: Non-invasive PCWP correlates well with the invasive reference at rest and during exercise stress. There is overall good diagnostic accuracy for HFpEF assessment using CMR-derived estimated PCWP despite deviations in absolute agreement. Non-invasive exercise derived PCWP may particularly facilitate detection of masked HFpEF in the future.