The relationship between beat-to-beat variability in left ventricular inflow velocity and left ventricular filling pressure in patients with atrial fibrillation

Background Heart failure complicating atrial fibrillation (AF) is associated with a poor prognosis; thus, early detection of elevated left ventricular filling pressure (LVFP) is crucial. In AF, conventional echocardiographic estimation of LVFP is challenging due to the loss of atrial contraction and irregular cardiac cycles. Normally, peak E-wave velocity varies with fluctuating cycle lengths. However, we hypothesized that when LVFP is elevated, the increased left atrial driving pressure blunts the effect of cycle length variations, resulting in a more constant peak E-wave velocity. Purpose To investigate whether reduced beat-to-beat variability of peak E-wave velocity accurately detects elevated LVFP in patients with AF. Methods We prospectively enrolled patients with non-valvular AF who underwent invasive left atrial pressure (LAP) or pulmonary artery wedge pressure (PAWP) measurement and echocardiography within 24 hours. Peak E-wave velocity was measured for at least 10 consecutive heartbeats to calculate its coefficient of variation (CV: standard deviation/mean × 100). Elevated LVFP was defined as a mean LAP or PAWP ≥ 15 mmHg. Results A total of 98 patients (mean age 70 ± 10 years; 26 women) were included (elevated LVFP group, n=24; non-elevated LVFP group, n=74). The CV of peak E-wave velocity was significantly lower in the elevated LVFP group (4.3 ± 2.8%) than in the non-elevated LVFP group (8.8 ± 3.9%, p<0.0001) and showed a significant negative correlation with invasive LAP/PAWP (r = -0.46, p<0.0001). ROC analysis of E-wave CV demonstrated good predictive ability with an AUC of 0.88 (sensitivity: 75%, specificity: 92%, cutoff: 4.1%). In multivariate logistic regression analysis, a decrease in E-wave CV remained an independent predictor of elevated LVFP (odds ratio per 1% decrease: 1.71, 95% CI: 1.31–2.39) even after adjusting for American Society of Echocardiography guideline-recommended parameters (peak E velocity, E/e’, tricuspid regurgitation velocity, and deceleration time). Furthermore, adding E-wave CV to a baseline model comprising these four parameters significantly improved diagnostic performance (AUC: 0.81 to 0.91, p=0.04). Conclusions Reduced beat-to-beat variability (CV) of peak E-wave velocity is a powerful, independent predictor of elevated LVFP in AF. This novel index provides incremental diagnostic value, effectively complementing current guideline-recommended parameters.