Normal Reference Values for Assessing Diastolic Function Using the Parameterized Diastolic Filling Formalism Method in Patients with Normal Results of Rest and Stress Echocardiography

Mechanistic validation of the 2016 American Society of Echocardiography/European Association of Cardiovascular Imaging Guidelines for the assessment of diastolic dysfunction in heart failure with reduced ejection fraction

Background

The American Society for Echocardiography/European Association of Cardiovascular Imaging (ASE/EACVI) 2016 guidelines for assessment of diastolic dysfunction (DD) are based primarily on the effects of diastolic dysfunction on left ventricular filling hemodynamics. However, these measures do not provide quantifiable mechanistic information about diastolic function. The Parameterized Diastolic Filling (PDF) formalism is a validated theoretical framework that describes DD in terms of the physical properties of left ventricular filling.

Aims

We hypothesized that PDF analysis can provide mechanistic insight into the mechanical properties governing higher grade DD.

Methods

Patients referred for echocardiography showing reduced left ventricular ejection fraction (< 45%) were prospectively classified into DD grade according to 2016 ASE/EACVI guidelines. Serial E-waves acquired during free breathing using pulsed wave Doppler of transmitral blood flow were analyzed using the PDF formalism.

Results

Higher DD grade (grade 2 or 3, n = 20 vs grade 1, n = 30) was associated with increased chamber stiffness (261 ± 71 vs 169 ± 61 g/s², p < 0.001), increased filling energy (2.0 ± 0.9 vs 1.0 ± 0.5 mJ, p < 0.001) and greater peak forces resisting filling (median [interquartile range], 18 [15–24] vs 11 [8–14] mN, p < 0.001). DD grade was unrelated to chamber viscoelasticity (21 ± 4 vs 20 ± 6 g/s, p = 0.32). Stiffness was inversely correlated with ejection fraction (r = − 0.39, p = 0.005).

Conclusions

Higher grade DD was associated with changes in the mechanical properties that determine the physics of poorer left ventricular filling. These findings provide mechanistic insight into, and independent validation of the appropriateness of the 2016 guidelines for assessment of DD.

Alternative diastolic function models of ventricular longitudinal filling velocity are mathematically identical

The spatiotemporal features of normal in vivo cardiac motion are well established. Longitudinal velocity has become a focus of diastolic function (DF) characterization, particularly the tissue Doppler e'-wave, manifesting in early diastole when the left ventricle (LV) is a mechanical suction pump (dP/dV < 0). To characterize DF and elucidate mechanistic features, several models have been proposed and have been previously compared algebraically, numerically, and in their ability to fit physiological velocity data. We analyze two previously noncompared models of early rapid-filling lengthening velocity (Doppler e'-wave): parametrized diastolic filling (PDF) and force balance model (FBM). Our initial numerical experiments sampled FBM-generated e'(t) contours as input to determine PDF model predicted fit. The resulting exact numerical agreement [standard error of regression (SER) = 9.06 × 10^-16] was not anticipated. Therefore, we analyzed all published FBM-generated e'(t) contours and observed identical agreement. We re-expressed FBM's algebraic expressions for e'(t) and observed for the first time that model-based predictions for lengthening velocity by the FBM and the PDF model are mathematically identical: e'(t) = γe^-αtsinh(βt), thereby providing exact algebraic relations between the three PDF parameters and the six FBM parameters. Previous pioneering experiments have independently established the unique determinants of e'(t) to be LV relaxation, restoring forces (stiffness), and load. In light of the exact intermodel agreement, we conclude that the three PDF parameters, relaxation, stiffness (restoring forces), and load, are unique determinants of DF and e'(t). Thus, we show that only the PDF formalism can compute the three unique, independent, physiological determinants of long-axis LV myocardial velocity from e'(t).NEW & NOTEWORTHY We show that two separate, independently derived physiological (kinematic) models predict mathematically identical expressions for LV-lengthening velocity (Doppler e'-wave), indicating that damped harmonic oscillatory motion is a physiologically accurate model of diastolic function. Although both models predict the same "overdamped" velocity contour, only one model solves the "inverse problem" and generates unique, lumped parameters of relaxation, stiffness (restoring force), and load from the e'-wave.