Echocardiographic Estimation of Pulmonary Artery Wedge Pressure in Advanced Heart Failure: Head-to-Head Comparison Against Right Heart Catheterization.

Abstract

Several equations have been proposed to non-invasively estimate pulmonary artery wedge pressure (PAWP) and have been validated against right heart catheterization (RHC). However, their accuracy in patients with advanced heart failure (AdvHF) remains uncertain. We assessed the agreement between echocardiographic and invasive PAWP measurements at rest and during vasodilator testing. Consecutive patients referred for AdvHF undergoing paired transthoracic echocardiography (TTE) and RHC within 24 hours were prospectively enrolled. Echocardiographic PAWP was calculated with the Nagueh E/e' equation and with the left atrial volume index (LAVi) and mitral E-based equation reported by Lindow (ePAWP-E). Agreement with invasive PAWP was assessed by Spearman correlation and Blant-Altman analysis, and discrimination for elevated PAWP (≥ 15 mmHg) by receiver operating characteristic (ROC) analysis, overall and in the pre-specified subgroup with left atrial volume index (LAVi) < 48 mL/m². In patients undergoing sodium nitroprusside challenge, invasive and echocardiographic PAWP change (ΔPAWP) was analyzed and compared. Of 176 patients enrolled, 172 (mean age was 52.6 ± 12.1 years and 76.2% were males), had an interpretable PAWP and formed the analytic cohort; 77 (44.8%) had PAWP ≥ 15 mmHg. Correlation with invasive PAWP was ρ = 0.452 for Nagueh (n = 148) and ρ = 0.550 for ePAWP-E (n = 156, both p < 0.001), with AUC 0.759 (95% CI 0.677-0.833) and 0.798 (95% CI 0723-0.864). Both equations overestimated invasive PAWP (bias + 3.2 and + 2.6 mmHg) with wide limits of agreement (-13.3 to + 19.7 and - 11.6 to + 16.8 mmHg). In the LAVi < 48 mL/m² subgroup, agreement improved (ρ = 0.496 vs. 0.655; AUC 0.850 and 0.927) and the limits of agreement for ePAWP-E narrowed to - 8.3 to + 10.3 mmHg. During nitroprusside challenge, invasive ΔPAWP was 9.6 ± 7.6 mmHg whereas echocardiographic ΔePAWP was 2.4 ± 5.6 mmHg (Nagueh, n = 33) and 1.5 ± 6.0 mmHg (ePAWP-E, n = 34; both p < 0.001 vs. invasive), with weak Δ correlations (ρ = 0.414 and 0.389). In AdvHF, echocardiographic equations correlate only moderately with invasive PAWP and are not accurate enough for individual-level quantification. The LAVi- and E-based equation performs best and may be used to rule in or rule out elevated filling pressures at rest when the left atrium is not severely dilated. Neither equation tracks the acute PAWP changes induced by vasodilator challenge, so RHC remains indispensable whenever the clinical question is dynamic.