Abstract 12352: A Novel and Strong Predictor for Left Ventricular Filling Pressure Assessed by the Time-volume Curve of the Left Atrium Constructed from Speckle Tracking Echocardiography

Abstract

Background: Left atrial (LA) volume and E/e’ have been proposed to reflect left ventricular (LV) filling pressure. However, noninvasive estimation of pulmonary capillary wedge pressure (PCWP) has not been possible. We hypothesized that the KT index may be a more useful noninvasive method to estimate PCWP in routine clinical practice. The KT index is derived from the LA time-volume curve measured by speckle tracking echocardiography (STE) and defined as 10 x logarithm [100 x (active emptying function / minimum LAV)].

Methods: In the validation study, LA phasic volume and emptying function (EF) were measured in 45 patients (age 69±8, 25 men) by STE. The KT index and E/e’ were obtained just before cardiac catheterization. These parameters were compared with PCWP. In the testing study (n=19), the PCWP measured by the KT index formula were compared with the PCWP measured by cardiac catheterization.

Results: As PCWP increased, maximum and minimum LAV increased, total and active EF decreased and pre-atrial contraction LAV was closer to minimum LAV. Maximum and minimum LAV index were directly correlated with PCWP (r=0.78 and r=0.83, respectively; p<0.01). LA total, passive and active EF were inversely correlated with PCWP (r=-0.80, r=-0.51 and r=-0.71, respectively; p<0.01). The KT index had a strong correlation with PCWP (r=0.84, p<0.01), whereas there was a weak correlation between E/e’ and PCWP (r=0.53, p<0.01). The sensitivity and specificity to predict elevated PCWP (>12 mmHg) were only 63 and 69%, respectively (area under the curve: AUC=0.73) using an E/e’ of 15 as an optimal cutoff value. In contrast, the sensitivity and specificity were 84 and 85%, respectively (AUC=0.94) using a KT index of 18 as an optimal cutoff value. In multivariate regression analysis, only the KT index was associated with PCWP, and the probability of an accurate diagnosis was 84% (PCWP=-1.22 x KT index + 35.1). In the testing study, the PCWP measured by the KT index formula were strongly correlated with the PCWP measured by cardiac catheterization (r=0.93, p<0.01).

Conclusions: For noninvasive estimation of LV diastolic function, the KT index is a novel and more useful predictor of PCWP than LA volume or E/e’ and has incremental value in routine clinical practice.