Single‐beat estimation of the left ventricular end‐systolic pressure–volume relationship in patients with heart failure

Abstract

The end-systolic pressure-volume relationship (ESPVR) constructed from multiple pressure-volume (PV) loops acquired during load intervention is an established method to asses left ventricular (LV) contractility. We tested the accuracy of simplified single-beat (SB) ESPVR estimation in patients with severe heart failure. Nineteen heart failure patients (NYHA III-IV) scheduled for surgical ventricular restoration and/or restrictive mitral annuloplasty and 12 patients with normal LV function scheduled for coronary artery bypass grafting were included. PV signals were obtained before and after cardiac surgery by pressure-conductance catheters and gradual pre-load reductions by vena cava occlusion (VCO). The SB method was applied to the first beat of the VCO run. Accuracy was quantified by the root-mean-square-error (RMSE) between ESPVR(SB) and gold-standard ESPVR(VCO). In addition, we compared slopes (E(ES)) and intercepts (end-systolic volume at multiple pressure levels (70-100 mmHg: ESV(70)-ESV(100)) of ESPVR(SB) vs. ESPVR(VCO) by Bland-Altman analyses. RMSE was 1.7 + or - 1.0 mmHg and was not significantly different between groups and not dependent on end-diastolic volume, indicating equal, high accuracy over a wide volume range. SB-predicted E(ES) had a bias of -0.39 mmHg mL(-1) and limits of agreement (LoA) -2.0 to +1.2 mmHg mL(-1). SB-predicted ESVs at each pressure level showed small bias (range: -10.8 to +9.4 mL) and narrow LoA. Two-way anova indicated that differences between groups were not dependent on the method. Our findings, obtained in hearts spanning a wide range of sizes and conditions, support the use of the SB method. This method ultimately facilitates less invasive ESPVR estimation, particularly when coupled with emerging noninvasive techniques to measure LV pressures and volumes.