Ventricular Diastolic Filling and Function
Ventricular emptying and filling are complex interdependent processes with the cardiac cycle conceptually divided into systole and diastole to allow clinical measurements of disease severity. Diastolic ventricular dysfunction plays a key role in the clinical manifestations of disease in patients with a wide range of cardiac disorders. In patients with clinical heart failure who have a preserved ejection fraction (HFpEF), diastolic dysfunction is the predominant cause of symptoms. Diastolic dysfunction also may be an early sign of cardiac diseases (as in hypertension), often antedating clinical or echocardiographic evidence of systolic dysfunction. In addition, in patients with heart failure with reduced ejection fraction (HFrEF), the degree of diastolic dysfunction may explain differences in clinical symptoms between patients with similar ejection fractions.
Echocardiographic techniques allow evaluation of right and left ventricular (RV and LV) diastolic filling patterns, the velocity of myocardial motion, and right and left atrial (RA and LA) filling patterns. Newer approaches to the evaluation of diastolic function include strain and torsion echocardiography. The relationship between these noninvasive measures and ventricular diastolic function and the utility of these measures in patient evaluation are discussed in this chapter.
Diastole is the interval from aortic valve closure (end-systole) to mitral valve closure (end-diastole) (Fig. 7-1). The isovolumic contraction period, from mitral valve closure to aortic valve opening, is part of systole.
Figure 7–1 Diastolic pressure curves.
The relationship among LV, LA, and aortic (Ao) pressures and M-mode tracings of the aortic and mitral valve are shown. The isovolumic relaxation time (IVRT) is the interval from aortic valve closure to mitral valve opening. During this interval, LV pressure declines rapidly. A rapid rise in LV pressure occurs during the isovolumic contraction time (IVCT), the interval between mitral valve closure and aortic valve opening.
Isovolumic relaxation starts with aortic valve closure followed by a rapid decline in LV pressure. When LV pressure falls below LA pressure, the mitral valve opens, ending the isovolumic relaxation period. Maximal opening of the mitral leaflets occurs rapidly, within 100 ± 10 ms of valve opening, in normal individuals. Mitral valve opening is followed by rapid early-diastolic filling, with the rate and time course of LA to LV flow determined by several factors, including the pressure difference along the flow stream, ventricular relaxation, and the relative compliances of the two chambers.
As the ventricle fills, pressures in the atrium and ventricle equalize, resulting in a period of diastasis, during which there is little movement of blood between the chambers and the mitral leaflets remain in a semiopen position. The duration of diastasis is dependent on heart rate, being longer at slow heart rates and entirely absent at faster heart rates. With atrial contraction, LA pressure again exceeds LV pressure, resulting in further mitral leaflet opening and a second pulse of LV filling. In normal individuals this atrial contribution accounts for only about 20% of total ventricular filling (Fig. 7-2).
Figure 7–2 Diastolic filling curves.
The relationship between LV volume and the diastolic LV Doppler filling pattern is shown. Early rapid filling coincides with the E velocity, followed by diastasis (with little or no flow from the LA to the LV), and atrial contraction (which coincides with the late diastolic A velocity). The Doppler velocity curve, in effect, is the first derivative of the LV volume curve.
The phases of diastole for the RV are analogous to those described for the LV, with the difference that the total duration of diastole is slightly shorter in normal individuals because of a slightly longer RV systolic ejection period.
There are several physiologic parameters that can be used to describe different aspects of diastolic function, but there is no single measure of overall diastolic function. The most clinically relevant parameters of diastolic function are:
Additional parameters of interest include elastic recoil of the ventricle and the effect of pericardial constraint, but the importance of these factors in normal diastolic ventricular function remains controversial.
LV relaxation, occurring during isovolumic relaxation and the early-diastolic filling period, is an active process involving the utilization of energy by the myocardium. Factors affecting isovolumic relaxation include internal loading forces (cardiac fiber length), external loading conditions (wall stress, arterial impedance), inactivation of myocardial contraction (metabolic, neurohumoral, and pharmacologic), and nonuniformity in the spatial and temporal patterns of these factors. Abnormal relaxation results in prolongation of the isovolumic relaxation time, a slower rate of decline in ventricular pressure, and a consequent reduction in the early peak filling rate (due to a smaller pressure difference between the atrium and the ventricle when the atrioventricular valve opens). Measures of LV relaxation include the isovolumic relaxation time (IVRT), the maximum rate of pressure decline (−dP/dt), and the time constant of relaxation (tau or τ). There are several different mathematical approaches to the calculation of τ, but basically it reflects the rate of pressure decline from the point of maximum −dP/dt to mitral valve opening. While peak rapid filling rate is affected by ventricular relaxation, it is only an indirect measure of this physiologic parameter because several other factors also affect peak filling (Fig. 7-3).
Compliance is the ratio of change in volume to change in pressure (dV/dP). Stiffness is the inverse of compliance: the ratio of change in pressure to change in volume (dP/dV). Conceptually, compliance can be divided into myocardial (the characteristics of the isolated myocardium) and chamber (the characteristics of the entire ventricle) components. Chamber compliance is influenced by ventricular size and shape, in addition to the characteristics of the myocardium. Extrinsic factors also may affect measurement of compliance, including the pericardium, RV volume, and pleural pressure. Evaluation of ventricular compliance is based on diastolic passive pressure-volume curves showing the degree to which pressure and volume change in relation to each other over the physiologic range (Fig. 7-4).
Figure 7–4 Reduced diastolic compliance.
The passive pressure-volume relationship of the LV is steeper than normal. As LV volume increases in diastole, pressure rises rapidly, resulting in an initial high LA-LV pressure gradient with a rapid decrease in the filling gradient during diastole. The Doppler velocity curve shows a decreased isovolumic relaxation time (IVRT), steep deceleration slope, and reduced A velocity. Note that even with normal compliance, reduced systolic function results in a rightward shift along the normal pressure-volume relationship, resulting in a pattern of diastolic filling similar to decreased compliance.
Clinically, evaluation of diastolic pressures alone often is used in patient management. Diastolic “filling” pressures include LV end-diastolic pressure and mean LA pressure. LV end-diastolic pressure reflects ventricular pressure after filling is complete, and LA pressure reflects the average pressure in the LA during diastole. Clinically, LA pressure is estimated by the pulmonary wedge pressure either at a single time point in the cardiac catheterization laboratory or at many time points with an indwelling right heart (Swan-Ganz) catheter in the intensive care unit.
Another clinically available measure related to diastolic function is the time course of ventricular filling. In theory, an LV volume curve could be generated by multiplying mitral annulus area by the integral of the Doppler velocity curve for each time point in diastole, but this is not accurate or practical in clinical applications. LV filling curves also can be generated from frame-by-frame measurements of ventricular volumes using three-dimensional (3D) echocardiography, but the clinical value of this data has not been evaluated.
Unfortunately, while ventricular diastolic function is one of the major factors affecting the pattern of diastolic filling, these two concepts are not identical. Several physiologic parameters other than diastolic function affect diastolic filling. Given no change in diastolic function (i.e., relaxation, compliance, etc.), the peak early-diastolic filling rate will be affected by:
The importance of considering how these factors impact the Doppler pattern of diastolic filling is discussed in more detail in the following sections. In addition, it is obvious that the utility of ventricular diastolic filling patterns for assessing diastolic function is valid only in the absence of mitral valve disease because, with mitral stenosis, LV filling velocity and timing are predominantly affected by the severity of valve obstruction, whereas with mitral regurgitation, the transmitral volume flow rate is increased, altering the LV inflow velocity curve. In patients with rhythms other than normal sinus rhythm (e.g., atrial fibrillation) evaluation of diastolic function with Doppler is more challenging because of the absence of atrial contraction and the varying length of the diastolic filling period.
Another component in the evaluation of ventricular diastolic function is the measurement of atrial filling patterns and pressures. The atrium serves as a “conduit” for flow from the venous circulation to the ventricle, especially in early diastole when the atrium is not contracting. In addition, elevations in ventricular diastolic pressures will be reflected in elevated pressures in the atrium (Fig. 7-5).
Figure 7–5 RA (hepatic vein) and LA (pulmonary vein) filling patterns.
Atrial filling patterns are similar to the pattern of jugular venous pulsations seen on physical examination. Pulmonary and hepatic vein patterns appear “opposite” in direction, because the direction of flow in the hepatic vein using a transthoracic subcostal view is away from the transducer (into the RA), while the direction of flow from a transthoracic apical view of the pulmonary vein is toward the transducer (into the LA).
These filling phases are reflected in the patterns of jugular venous pulsation familiar to the clinician: the a-wave following atrial contraction, the x-descent corresponding to atrial filling during ventricular systole, the v-wave at end-systole, and the y-descent corresponding to atrial filling during ventricular diastole. Disease processes affect the jugular venous pulsations and the Doppler pattern of RA filling in similar ways.
In normal individuals, the systolic and diastolic filling phases are approximately equal in volume. Normal LA pressure is low (5 to 10 mm Hg), corresponding to the normal LV end-diastolic pressure, with slight increases in pressure following atrial (a-wave) and ventricular (v-wave) contraction.
Normal LV and RV diastolic filling show respiratory variation. With inspiration, negative intrapleural pressure results in an increase in systemic venous return into the thorax and, thus, into the RA. This increase in RA volume and pressure results in a transient increase in RV diastolic filling volumes and velocities, with a normal magnitude of increase of up to 20% compared with end-expiratory values.
LA filling does not increase with inspiration because pulmonary venous return is entirely intrathoracic and thus not affected significantly by respiratory changes in intrathoracic pressure. In fact, LA and, consequently, LV diastolic filling is slightly higher at end-expiration than during inspiration. The mechanism of the observation remains controversial. Some postulate a delay in transit of the increased RV filling to the left side of the heart. Others suggest a decrease in LA filling during inspiration because of an increased volume (or “pooling”) in the pulmonary venous bed. Less likely in normal individuals is impaired LV diastolic filling due to an increase in RV diastolic volume within a fixed-volume pericardium. This last mechanism may become important in patients with pericardial disease (e.g., constriction, tamponade) and may partly account for the exaggerated respiratory changes in RV and LV diastolic filling seen in these conditions.
While diastolic dysfunction can be seen with a wide range of cardiac disorders, the four basic mechanisms of disease (Table 7-1) that lead to diastolic dysfunction are:
Evaluation of ventricular chamber dimensions and wall thickness is an integral part of the echocardiographic evaluation of diastolic function. The relative degree of systolic and diastolic dysfunction in patients with heart failure ranges from severe diastolic dysfunction with a normal ejection fraction to severe systolic dysfunction with normal filling pressures. However, most patients with systolic dysfunction have some degree of diastolic dysfunction, and most patients with diastolic dysfunction have anatomic cardiac changes evident on echocardiographic imaging. Typically, diastolic heart failure (HFpEF) occurs in patients with a thick walled small ventricle due to either restrictive cardiomyopathy or hypertensive heart disease. The presence and severity of LA enlargement reflect chronically elevated filling pressures so that measurement of LA size or volume is integral to evaluation of diastolic function (see Fig. 2-16).
In patients with heart failure due primarily to systolic ventricular dysfunction (HFrEF), typical imaging findings include a dilated LV with global or regional dysfunction and a reduced ejection fraction. Diastolic dysfunction usually accompanies systolic dysfunction, and measures of diastolic function and LV filling pressures are important for patient management and prognosis.
Other imaging findings that raise the question of diastolic dysfunction include pericardial thickening (as in constrictive pericarditis), the pattern of ventricular septal motion with respiration (especially with tamponade physiology), and dilation of the inferior vena cava and hepatic veins (consistent with elevated RA pressures).
Doppler recordings of LV diastolic filling velocities correspond closely with ventricular filling parameters measured by other techniques. The normal Doppler ventricular inflow pattern is characterized by a brief time interval between aortic valve closure and the onset of ventricular filling (the isovolumic relaxation time). Immediately following mitral valve opening, there is rapid acceleration of blood flow from the LA to the ventricle with an early peak filling velocity of 0.6 to 0.8 m/s occurring 90 to 110 ms after the onset of flow in young, healthy individuals (Table 7-2). This early maximum filling velocity (E velocity) occurs simultaneously with the maximum pressure gradient between the atrium and the ventricle. After this maximum velocity, flow decelerates rapidly (i.e., with a steep slope) in normal individuals with a normal deceleration slope of 4.3 to 6.7 m/s2. Deceleration time, defined as the time interval from the E peak to where a line following the deceleration slope intersects with the zero baseline, ranges from 140 to 200 ms. Early-diastolic filling is followed by a variable period of minimal flow (diastasis), depending on the total duration of diastole. With atrial contraction, LA pressure again exceeds ventricular pressure, resulting in a second velocity peak (late diastolic or atrial velocity), which typically ranges from 0.19 to 0.35 m/s in young, normal individuals (Fig. 7-6).
Data from Tebbe et al: Clin Cardiol 3:19, 1980; Shapiro et al: Br Heart J 51:637, 1984; Pearson et al: Am Heart J 113:1417, 1987; Snider et al: Am J Cardiol 56:921, 1985; Garcia-Fernandez et al: Eur Heart J 20:496, 1999.
Figure 7–6 Normal pattern of LV diastolic filling recorded with pulsed Doppler in an apical four-chamber view at the mitral leaflet tips (left) and at the mitral annulus level (right).
The recording at the mitral tips level is used to measure E velocity, A velocity, and the deceleration time (arrow). The annular flow signal is used for measurement of atrial flow duration. If transmitral stroke volume is calculated, the annular flow signal is used for the velocity-time integral.
Figure 7–7 Schematic diagram of quantitative measurements that can be made from the Doppler LV filling curve.
DFP, diastolic filling period; DT, deceleration time; IVRT, isovolumic relaxation time; VTI, velocity-timeintegral.
In order to convert the Doppler ventricular inflow velocity curve to a volume curve, the cross-sectional area of flow must be taken into account. Volumetric flow rates can be calculated as the product of velocity and cross-sectional area in regions where flow is laminar with a spatially symmetric flow pattern (Fig. 7-8). Thus the instantaneous volume flow rate across the mitral valve can be calculated as instantaneous velocity times the flow cross-sectional area (CSA). Similarly, transmitral stroke volume (SV) can be determined from the integral of the flow velocity curve (VTI) over the diastolic filling period:
Figure 7–8 Transmitral volume flow rate.
Volumetric flow rates across the mitral annulus can be calculated as shown. Mitral annular diameter can be measured from both apical four-chamber and parasternal long-axis views to calculate an elliptical cross-sectional area. Alternatively, a circular cross-sectional area is used as an approximation using a single annular diameter measurement.
The standard approach to determining the cross-sectional area of flow across the mitral valve is to calculate the cross-sectional area of flow at the mitral annulus level. Motion of the mitral leaflets is a passive process, with the degree of motion reflecting flow across the valve (in the absence of mitral stenosis). Although flow area tapers from the annulus to the leaflet tips, the more rigid mitral annulus is a preferable site for flow measurement rather than the flexible, mobile leaflets. Even though the shape of the mitral annulus is complex in 3D, in clinical practice assuming either a circular or elliptical geometry is a reasonable approximation, based on a diameter measurement in the apical four-chamber or parasternal long-axis or both (see Fig. 6-13).
LV inflow can be recorded in nearly all patients from an apical approach in either a four-chamber or long-axis view on transthoracic (TTE) imaging. This window allows parallel alignment between the ultrasound beam and the direction of LV filling. On transesophageal (TEE) echocardiography, LV inflow can be recorded from a high esophageal position, taking care to align the Doppler beam parallel to the inflow stream (Fig. 7-9). In some patients, a transgastric apical view also may allow recording of LV inflow, although caution is needed in order to avoid a nonparallel intercept angle (with resultant underestimation of velocities) and a foreshortening of the ventricle from this window.
Figure 7–9 TEE transmitral flow velocities.
The typical pattern of LV diastolic filling is seen with a parallel alignment between the ultrasound beam and flow direction from a TEE four-chamber view. The only difference from a TTE flow recording is that flow is directed away from the transducer on TEE.
Inflow velocities should be recorded using pulsed Doppler with a 2 to 3 mm sample volume positioned either at the mitral leaflet tips (for evaluation of diastolic function) or at the mitral annulus level (for measurement of volume flow rates and the duration of atrial filling) (see Fig. 7-6). For diastolic function evaluation, with the beam aligned parallel to the flow stream, the sample volume is moved slowly along the length of the ultrasound beam to identify the site of maximal velocity, usually at the mitral leaflet tip level. The velocity range is adjusted to maximize the display of the velocity of interest and avoid signal aliasing. The sweep speed of the spectral display is maximized (100 cm/s) and wall filters are reduced (as allowed by signal quality) so that the velocities approach the baseline, allowing accurate time interval measurements. Flows are recorded at end-expiration during normal breathing.