14 Diastolic Dysfunction and Echocardiographic Hemodynamics

Based on spectral Doppler, echocardiography is able to establish pressure gradients with considerable accuracy. Determination of absolute pressure by echocardiography is less accurate, because it entails the addition of an estimate of pressure of variable accuracy and error magnitude. Valve disease can be managed clinically largely on the basis of gradients, whereas myopathic disease is managed more on the basis of absolute filling pressures.

Determination of valve areas is an attractive concept, because it involves less flow dependence, but the larger number of parameters involved in most area calculations augments the error involved in valve area calculations.

Using several different techniques, different volumes and flows can be determined by echocardiography, with variable ease and accuracy. Volumes are better hemodynamic descriptors than are dimensional measurements.

The “Holy Grail” of echocardiography is to provide the same robustness of hemodynamic information provided by the venerable pulmonary artery (PA) catheter. The PA catheter, indiscriminantly used for decades in the critical care arena, has been shown overall not to reduce mortality, and its usage has diminished. Stroke volume, cardiac index, and PA pressure are well determined by echocardiography. Reliably accurate determination of left atrial pressure by echocardiography is the remaining challenge. Although echocardiography is being used as a partial or direct substitute to fill the void of information left by reduced use of PA catheters, it is unclear whether the same information is available from echocardiography as previously was obtained through catheters, and too many assumptions have been made to be true all the time. For example, the presence of normal systolic function and no significant left-sided valve disease does not eliminate the possibility of left-sided heart failure, which may still be present due to either diastolic failure or volume overload.

Echocardiography has generated many formulae, equations, and methods—far more than are manageable. Most have more limitations than robustness, and it is prudent to practice using the best methods, with full knowledge of their limitations, and simply move on from the rest. An inaccurate equation, even if it may be readily calculated, is still an inaccurate equation.

Basic Equations and Their Assumptions

Doppler Shift Equation

where V = blood velocity, F_d = Doppler shift, cos_θ = cosine of the Doppler angle, and K= speed of ultrasound in tissue 2× carrier frequency.

Misalignment by ≤20 degrees introduces an error of ≤6%.

where P = density of fluid, ds = the “path element,” V = velocity, ΔP = pressure drop across the orifice, and R = resistance due to viscous friction. V₁ is usually much less than 1.5 m/sec and is generally ignored.

Three factors that affect pressure drop across an orifice are

Convective acceleration

• Acceleration due to pressure difference

• This is the principal component of the equation.

Flow acceleration

• Inertial changes at onset and termination of flow, which account for temporal differences between pressure changes and velocity changes

• Relevant only to valve opening and closure; therefore, of minimal importance in echocardiography

• This component usually is ignored.

Viscous friction

• Energy loss from interactions among moving particles and from particles to conduit—of minimal importance in echocardiography

• This component usually is ignored.

“Simplified” (Modified) Bernouilli Equation

Pressure gradient (ΔP) = 4 × V²

Mean pressure gradient = 4 ×

Peak instantaneous gradient = 4 × , where V_max = peak velocity

Assumptions

• Laminar flow

• Ignores viscous and frictional forces

• Omits upstream velocities and nonuniform pressure recovery

Pros

• Most assumptions are generally valid

• Generally not significantly affected by hematocrit

• Linear relationship over a wide range of velocities

Cons

• Viscous forces become significant with long (tubular) stenoses, e.g., long coarctation.

• Nonlinear relationship for orifices of different sizes—applies less well to orifices of ~7 mm in diameter, and tends to underestimate the gradient across these

Diastolic Function, Dysfunction, and Heart Failure with Normal Ejection Fraction

Decades of research in diastology have produced many conceptual advances in pathophysiology. However, frustratingly few clear-cut criteria directed toward the challenge of the clinical diagnosis of diastolic heart failure have emerged, and no direct patient care treatments have grown out of echocardiographic diastology findings, despite the number of measurements that have been made.

The Four Phases of Diastole

The basic problem with analysis of Doppler-derived filling patterns is the many, often oppositely directed, factors that determine the amplitude and characteristics of each parameter.

1. Isovolumic relaxation time (IVRT)

The IVRT is calculated from aortic valve closure to mitral valve opening. IVRT is therefore

Increased by

• Higher aortic pressure

• Lesser dP (rate of change in pressure)/dt (rate of change in time)

• Lower atrial pressure

Decreased by

• Lower aortic pressure

• Greater dP/dt

• Higher atrial pressure

• Marked tachycardia

• Greater LV elastic recoil

2. Early rapid filling

Early rapid filling is denoted by the “E-wave.” From the mitral valve opening to the end of the early diastolic, rapid filling normally accounts for 70% to 80% of left ventricular filling. Rapid filling is affected by the following:

Myocardial relaxation, which is itself affected by

• Left ventricular hypertrophy

• Ischemia

• Infarction

• Sympathetic activation

• Age

Myocardial elastic recoil

Pericardial elastic recoil

(Reduced) mitral valve area (stenosis)

Hence, the E-wave is not synonymous with myocardial relaxation.

3. Diastasis (“the apart”)

Diastasis is the interval from the end of early diastolic rapid filling to the beginning of atrial systole (the interval between the E and A waves). The diastasis interval may be quite long at low heart rates as long as there is no first-degree atrioventricular bock. Diastasis contributes less than 5% to left ventricular filling. Diastasis, the expendable phase of diastole, is shorter, with faster heart rates, and is abolished above 100 beats per minute.

4. Atrial systole

Atrial systole is denoted by the “A-wave,” from the onset of atrial systole to mitral valve closure. Atrial systole contributes approximately 10% to 20% to normal filling, but up to 25% in some disease states as long as atrial mechanical function is preserved. The effect of left atrial (LA) systole on left ventricular (LV) inflow is affected by

LA contractility: In advanced congestive heart failure, the LA contribution to LV filling decreases as LA systolic function fails.¹

LA pressure (preload)

Intrinsic LV stiffness

The left ventricular end-diastolic pressure (atrial afterload)

RV forces: In patients with RV dilation and RV systolic pressures >40 mm Hg, there is increased atrial systolic filling of the LV.²

The autonomic nervous system

Diastolic Function, Dysfunction, and Heart Failure

The function of diastole is to provide adequate volume load to the ventricles for the systolic output needs, at filling pressures physiologically acceptable to the atria and venous drainage. Traditionally, diastolic properties of the left ventricle were held to be the principal determinant of exertional tolerance in cases of systolic dysfunction. “Diastolic dysfunction” is a variably used term, but generally means that the pattern and properties of diastolic filling are abnormal, reflecting abnormal diastolic properties. Exercise intolerance in patients with HFNSF is attributed to failure of the Frank-Starling mechanism (Fig. 14-1).

Figure 14-1 Plot of pulmonary capillary wedge pressure versus left ventricular end-diastolic volume indicating the directional change from rest to peak exercise in patients (blue circles) and normal subjects (green circles).

(From Kitzman DW, Higginbotham MB, Cobb FR, et al. Exercise intolerance in patients with heart failure and preserved left ventricular systolic function: failure of the Frank-Starling mechanism. J Am Coll Cardiol. 1991;17[5]:1065–1072.)

Diastolic heart failure is the clinical manifestation of severe diastolic dysfunction occurring in the absence of systolic dysfunction (usually ejection fraction [EF%] >55, although definitions differ). “Diastolic heart failure” has become a controversial term, increasingly replaced by “heart failure with normal systolic function” (HFNSF)—a term that is solely descriptive and does not speculate about the etiology.³ In the absence of systolic dysfunction, the presence of heart failure (after excluding valvular disease and coronary artery disease) is not invariably due to diastolic failure, as volume excess may be the principal cause of heart failure.

The most accurate means by which to evaluate diastolic dysfunction of the heart has evolved over decades, to become increasingly complex and invasive, rather than less complex and invasive. Hemodynamic studies that describe the stiffness of the heart through the end-diastolic pressure volume relationship (EDPVR), by reducing the venous inflow to the heart and simultaneously measuring the diastolic pressure and volume in the left ventricle with a conductance catheter, are held to be the best means to establish diastolic failure of the heart. A stiff ventricle has an upward-shifted EDPVR.³ The complexity and logistics of this catheterization technique simply prohibit its routine use. The findings of patients with HFNSF are upshifted EDPVR at reduced, normal, or even increased volumes (Figs. 14-2 and 14-3).³ No single adequate explanation of HFNSF exists, and the topic is becoming more, not less, complicated.

Figure 14-2 Left: Pressure-volume representation of prevailing paradigm of heart failure with normal systolic function (HFNEF), showing elevated end-diastolic pressure volume relationship (EDPVR) with no significant effect on end-systolic pressure-volume relationship (ESPVR). Respective end-diastolic pressure-volume point is shown by filled circle. Upper right: When the entire EDPVR cannot be measured, an alternate means of indexing diastolic properties for purposes of comparing heart sizes is via capacitance, the volume at a specified filling pressure. Lower right: EDPVR is nonlinear, so stiffness (the slope of the relationship, ∆P/∆V) depends on filling pressure, as indicated by the tangent line at each level of end-diastolic pressure (EDP). EDV, end-diastolic volume.

(From Burkhoff D, Maurer MS, Packer M. Heart failure with a normal ejection fraction: is it really a disorder of diastolic function? Circulation. 2003;107:656–658.)

Figure 14-3 Data re-plotted from Table 1 of Grossman et al.24 showing that diastolic stiffness (∆P/∆V) varies directly with filling pressure. Data from 10 hypertrophic hearts (solid squares) span high filling pressures and stiffness values. Data from one patient of the original publication with aberrant data (end-diastolic pressure [EDP] of 52 mm Hg and stiffness of 0.9 mm Hg/mL) was excluded. Line of linear regression (±95% confidence interval) was determined for data from these hypertrophic hearts. Data from 12 nonhypertrophic patients (open circles) have lower filling pressures and lower values of stiffness; a majority of these data fell within the 95% confidence intervals of the regression line determined from the patients with hypertrophy.

(From Burkhoff D, Maurer MS, Packer M. Heart failure with a normal ejection fraction: is it really a disorder of diastolic function? Circulation. 2003;107:656–658.)

LV inflow, pulmonary venous, indicies, and velocity of propagation Doppler echocardiography indices of diastolic dysfunction have not provided the same precision or the correlation to clinical status that catheter indices provided. LV inflow and pulmonary venous patterns are influenced by volume effects. Tissue Doppler indices, which may be afterload influenced, appear to offer more correlation. Many individuals have abnormal Doppler patterns of diastolic filling, but do not have heart failure. The Doppler diastolic parameters of patients with HFNSF are variable.⁴

Conventional Doppler diastolic indicies correlate only moderately with catheter-derived indicies of diastolic failure⁵:

• E/A: r = –0.36; IVRT: r = 0.31; E´/A´_lateral (<1): r = –0.37; E/E´_lateral (≥8): r = 0.53*

Diastolic dysfunction is detected in only 70% of patients by LV inflow patterns, but in 81% of cases by E´/A´_lateral, and in 86% of cases by E/E´_lateral.

Using either E´/A´_lateral or E/E´_lateral detected 93% of cases.

In all patients LV inflow profiles and tissue Doppler are suitable for analysis, but only 60% of pulmonary venous profiles are suitable.⁵

Vasan and Levy ⁶ proposed criteria for diagnosing diastolic heart failure/HFSNF:

Definite diastolic heart failure/HFSNF

• Definitive clinical evidence of heart failure

• Objective evidence of normal systolic function (>50%) within 72 hours of the heart failure event

• Objective evidence of LV diastolic dysfunction on catheterization (abnormal LV relaxation/filling/distensibility)

Probable diastolic heart failure/HFSNF

• Definitive clinical evidence of heart failure

• Objective evidence of normal systolic function (>50%) within 72 hours of the heart failure event

• No conclusive objective evidence of LV diastolic dysfunction on catheterization (abnormal LV relaxation/filling/distensibility)

Possible diastolic heart failure/HFSNF

• Definitive clinical evidence of heart failure

• Objective evidence of normal systolic function (>50%) outside of a 72-hour window of time from the heart failure event

• Objective evidence of LV diastolic dysfunction on catheterization (abnormal LV relaxation/filling/distensibility) is lacking

As can be seen, the only of these criteria that echocardiography contributes to is the punctual assessment of LV systolic function. Echocardiographic indices of diastolic function are not criteria according to Vasan and Levy, but are included in the European Study Group on Diastolic Heart Failure report.⁷

The simplified and standardizing notion of HFNSF, and the fact that no treatment strategies have been validated based on echocardiographic diastolic filling indices, have resulted in many questions regarding whether the gamut of diastolic indices should be recorded on routine studies, which, in fact, most laboratories do not do. If echocardiographic parameters are not used for the most severe entity of diastolic physiology—HFNSF—clinicians may question whether they have any clinical use at all.

Most patients with HSNSF are elderly women with mild left ventricular hypertrophy and chronic advanced heart failure symptoms.⁸ Borderline or mildly increased wall thickness is usual, but severe left ventricular hypertrophy is unusual.⁴ The left atrium is almost invariably enlarged (volume appears superior to dimension); hence, a normal-sized left atrium can be held against the diagnosis of diastolic heart failure.^9,¹⁰

Traditionally, echocardiography assesses diastolic function by three means: