Ebook: Systolic and Diastolic Function of the Heart

The goal of this book is to improve understanding of what is meant by the terms “systolic function” and “diastolic function” and how these can best be measured and interpreted, both now and in the future..

In Diastole, the ventricular wall must be appropriately compliant to allow rapid, complete filling at low filling pressures. The key questions are:

• What structures and mechanisms determine ventricular compliance? Does compliance primarily reflect the myocytes (titin, tubulin, other elastic intramyocytic proteins)? The collagen extracellular matrix? Pressures, distentions, and flows in fluid compartments (blood, lymph, interstitial fluids)? What is the role of calcium-dependent myofilament interactions (and incomplete relaxation)? Is compliance subject to endothelial modulation? Is it viscoelastic?

• Is elastic systolic energy normally stored in the heart wall to aid filling during diastole (ventricular suction)? If so, where is this energy stored ... in myocytes? in collagen? In what mode and direction ... compression? expansion? torsional deformation, twist, or other shears? Does the contractile state of the myocytes during systole alter diastolic suction? How is such energy storage transmurally distributed? Is it greater nearer the epicardium? the endocardium? uniform throughout the wall? What is the timing of elastic energy release during the cardiac cycle? Is all of the potential energy stored during systole dissipated during relaxation? Is suction normally in evidence or is it a backup mechanism? How does one demonstrate when suction is present? Does energy storage in the mitral valvular-ventricular complex contribute to filling?

• What is the effect of overall chamber geometry (size, shape) on compliance and suction? What are the effects of geometric remodeling (protein turnover with widely different time courses), heart rate, pericardial constraint, ventricular interdependence, asynchrony and asynergy, exercise, hypertrophy, dilatation, heart failure? What is the relationship between diastolic chamber properties and myocardial (muscle) properties?

• Can diastolic performance be measured adequately in the clinical setting and are the measurements clinically meaningful? How important is rapid early diastolic filling? How are regional pressure gradients produced within the ventricle? How should one define ventricular volume in chambers which have valve leaflets opening, closing and deforming? How should relaxation be defined? What are the effects of various clinical interventions and surgical procedures on diastolic performance?

In Systole, the ca. 10⁹ myocardial cells in the ventricular wall must generate adequate coordinated force to provide sufficient pressure to open the outlet valve and then shorten and thicken the wall to pump an appropriate stroke volume. The key questions addressed are:

• Because measured wall thickness increases may not be accounted for solely by cell thickness increase and wall curvature changes, what cellular rearrangements are involved? Fiber slippage? Fiber direction changes? Do sheets of quasiradially oriented fibers contribute to shear-induced wall thickness increase? Do additional fibers run at steep pitch angles from epicardium to endocardium? Does the spiral architecture of the cells lead to systolic torsional deformations which: (1) allow the myocardium to thicken as much as it does within the constraints of sarcomere shortening; and (2) distribute the work of the heart uniformly from endocardium to epicardium by equalizing transmural differences in fiber shortening? Does epicardial shortening “crush” and rearrange the endocardial fibers such that endocardial cross-fiber shortening greatly exceeds fiber shortening?

• How are myocardial cells tethered to one another? Is myocardium stiffer in the fiber direction than in the cross-fiber direction? What are the effects of the spiral-wrapped ventricular wall on transmural dynamics? Why is torsional deformation coupled tightly to chamber volume during systole, but almost uncoupled in diastole? How much regional variation is observed in contractile parameters? Does the volume of the tissue change during contraction? How does the endocardial endothelium affect force generation?

• Can we deduce myocyte contractile properties from measurements at the chamber level (e.g., pressure, volume, torsion) or isolated muscle level (e.g., force, length)? Can we deduce cellular energetics (e.g., oxygen, ATP consumption) from measurements at the chamber level (e.g., pressure-volume area)? How does the force-interval relationship (mechanical restitution curve) relate to calcium kinetics of various cellular compartments?

• How important is the quest for a single index of ventricular “contractility.” How do the concepts of maximal elastance, preload recruitable stroke work, dP/dt, dP/dt_max vs. EDV slope, and maximal power compare in this regard? Does valvular elasticity confound the interpretation of these indices? Is the concept of ‘contractility’ clinically useful? Does an increase in perfusion increase contractility and oxygen consumption?

• How can we define the load on the left ventricle? Is it wall stress? Chamber pressure? Arterial impedance? Various combinations of these, depending on the application?

• Is adrenergic regulation of the force-frequency relation a major determinant of intrinsic contractility in the intact heart (in addition to length-dependent activation, direct beta adrenergic stimulation of the myocardium, and the basal force- frequency relation)? If so, how is it achieved?

• What are the effects of pathophysiological conditions (infarction, hypertrophy, dilatation, stunning, hybernation) on systolic performance?

In a recent editorial [Circulation 91: 1901, 1995], Claude L’ enfant predicted an important role for integrative physiology to help make sense of the important bits and pieces of new knowledge that are so rapidly accumulating from subcellular and molecular biology. He concluded that “...the full realization of the implications of this vi new knowledge will not occur without the concurrent and inspired efforts of scientists who apply this information to the larger organism and use it in the context of the vast knowledge already assembled about living creatures.” It is hoped that the questions addressed in this volume present the most recent views of integrative cardiovascular physiology and focus on the vital issues in this cooperative effort.

Neil B. Ingels, Jr.

George T Daughters

Jan Baan

James W. Covell

Robert S. Reneman

Frank C-P. Yin

Collagen is a stiff structural material with a high tensile strength. While present in the myocardium in relatively small amounts, collagen is the most abundant structural protein of the heart’s connective tissue network. Removal of less than half of the normal amount of collagen results in a dilated ventricle with increased compliance. The hypertrophic response of the myocardium to an abnormal stress includes an increase in collagen concentration, a thickening of existing fibrillar collagen, and the addition of new collagen at all levels of the matrix. The consequence of this remodeling of the collagen matrix is a stiffer left ventricle with diastolic dysfunction. Thus, myocardial collagen is a major factor in determining the left ventricular diastolic pressure-volume relation.

Fibrillar collagen forms the major structural component of the cardiac extracellular matrix. Whereas the central contribution of collagen to the mechanical properties of many connective tissues is well known, the specific role of the cardiac collagen matrix in ventricular mechanics remains unclear and controversial. A review of the cardiac extracellular matrix and the mechanical effects of altered cardiac collagen structure and content suggests that collagen in the myocardium may affect ventricular mechanical function significantly. Mathematical models show that the functional effects of changes in the cardiac collagen matrix during disease probably depend on specific, measurable microstructural parameters.

MRI tagging is a powerful new method for non-invasively studying LV wall deformation. The counterclockwise spiral of epicardial fibers coupled with the clockwise spiral of endocardial fibers, provides a mechanism for extensive thickening. Epicardial deformation is in the direction of its fibers. Endocardium is forced to shorten mostly in the cross-fiber direction, a deformation that results in extensive thickening. Systolic torsion recoils rapidly before diastolic filling begins, likely representing release of restoring forces and augmenting filling. This recoil is diminished in LVH.

Left ventricular (LV) diastolic dysfunction often precedes systolic dysfunction. Thus, the assessment of LV diastolic function is imperative for the early diagnosis of the disease, and many indices have been developed. Most, however, exclude pericardial and right ventricular (RV) effects, both of which alter LV pressure and volume. This study used a previously developed model of the cardiovascular system to examine how the pericardium and the RV affect these indices. By this approach, we could set the intrinsic LV compliance. We compared this to the LV diastolic indices derived from LV pressure and volume measurements. The analysis shows that RV and pericardial pressures affect the measured LV pressures and volumes and hence, the indices of LV diastolic function. Even with completely accurate measurements for pressure and volume, as in the simulation, ignoring the RV and the pericardium resulted in errors ranging from −38% to +37% in derived indices of LV function. Further, ignoring ventricular interdependence minimizes the ability of the indices to separate normal from disease states. To improve accuracy, RV and pericardial effects must be considered when calculating LV diastolic indices.

Measurement of in-vivo diastolic function indexes has required cardiac catheterization. Analysis of Doppler echocardiographic transmitral flow contours using a kinematic paradigm holds promise in elimination of cardiac catheterization as a requirement for quantitative diastolic function characterization. The paradigm suggests that novel, kinematic indexes which characterize diastolic function can be extracted from the Doppler velocity profile (DVP). This article reviews the role of a kinematic modeling strategy. Topics covered include: the role of the heart as a suction pump, automated solution of the ‘inverse problem’ of diastole, the relationship of ‘dissynchronous relaxation’ of the ventricle to the Doppler waveform, prediction of LVEDP from the DVP and the causal relation of the third (S3) and fourth (S4) heart sounds to the Doppler E and A-waves.

In order to study the blood flow patterns in the left atrium, as well as the pulmonary venous inflow, Magnetic Resonance Imaging (MRI) was used to visualize the three-dimensional velocity fields in the transversal plane Left Ventricular Outflow Tract (LVOT) view and sagittal plane LVOT view in humans. A dedicated visualization and analysis software named ‘cardiac MARIAN’™ was developed on a Silicon Graphics Onyx computer. Temporal velocity profiles were measured and vector animations of the flow fields produced throughout the cardiac cycle. This provided a dynamic visualization of the complex flow fields in the left atrium.

At the beginning of systole, flow coming from the pulmonary veins is observed in the transversal and sagittal plane LVOT views. In the transversal view, blood coming from the left and right pulmonary veins flows along the atrial walls. The two streams join as they reach the closed mitral valve to form a vortex in the atrial chamber. In the sagittal view, blood flows from the right upper pulmonary vein into the atrium away from the mitral valve. This flow impinges on the atrial septum and is deflected towards the closed mitral valve. In both views, a vortex is created at peak systole in the left atrium and coincides with the measured peak Swave in the pulmonary venous flow profile. During diastole, the flow no longer forms a vortex in the left atrium, but instead flows directly from the pulmonary veins into the ventricle. During this phase, the atrium acts as a passive conduit from the pulmonary veins to the left ventricle. At the end of the E-wave, reverse flow is also present in the pulmonary veins.

The flow fields obtained from MRI provide information on the complete flow pattern in the left atrium, and are complementary to that obtained by Doppler echocardiography. MRI is unique in its ability to measure the three-dimensional velocity fields and anatomy simultaneously. Therefore, MRI can be used to provide a detailed analysis of cardiac function.

Color Doppler M-mode echocardiography provides a spatiotemporal map of blood distribution (v(s,t)) within the heart, displaying (typically) a temporal resolution of 5 msec, a spatial resolution of 300 μm, and a velocity resolution of 3 cm/sec. M-mode echocardiographic data can be obtained along a streamline from the mid-left atrium to the mid-left ventricle from either the apical transthoracic window or the basal transesophageal window. A key parameter of the color Doppler M-mode is the transmitral propagation velocity, the slope of the leading edge of the M-mode derived E-wave. This propagation velocity is significantly less than the velocities measured within the E-wave by pulsed Doppler echocardiography and appears to provide information about diastolic function which is quite distinct. Published studies have indicated that color M-mode propagation velocity is inversely related to the relaxation time constant and is reduced in settings of acute myocardial ischemia. This manuscript will outline physics and instrumentation issues involved in the use of the color Doppler M-mode, discuss prior applications of this technique, and postulate on its role in the assessment of ventricular diastolic function.

During exercise, the peak LV filling rate in early diastole increases as the duration of diastole decreases with exercise induced tachycardia and stroke volume is maintained or increased. Our studies in conscious instrumented animals indicate that this increased rate of diastolic filling during exercise results from an increased mitral valve pressure gradient produced by a fall in LV early diastolic pressure. This fall in LV early diastolic pressure is associated with a downward shift of the early diastolic portion of the left ventricular pressure volume loop and enhanced rate of left ventricular relaxation during exercise. This is altered after the induction of pacing induced heart failure The early diastole filling rate still increases as does the early diastolic left atrial left ventricular pressure gradient during exercise after CHF. However, this increased pressure gradient results entirely from an increase in left atrial pressure after heart failure. In addition, left ventricular relaxation is not enhanced but in fact is slowed during exercise after heart failure because early diastolic left ventricular pressure does not fall during exercise but actually increases.

Development of an index of the rate of left ventricular isovolumic pressure fall that accurately and consistently reflects alterations in relaxation has been elusive. Considerations with the use of any index include avoiding potential artifacts due to linear data shifts and loss of monoexponentiality. This report details the effects of an intervention that significantly alters the value of the variable asymptote on the calculation of certain indexes of left ventricular isovolumic relaxation. Our results indicate that in the closed-chest canine, in the normal sinus beats immediately following short-coupled extrasystoles, the magnitude and the direction of the trends of relaxation can be a function of the method of index calculation. Our results also emphasize the utility of biexponential approaches in the delineation of subtle trends in the form of isovolumic pressure decay that may reflect the complex interplay between active and passive components of the diastolic process.

Calcium ions play a crucial role in the activation of working ventricular myocardium, however the effect of calcium on relaxation has not been well established. Both the sarcoplasmic reticulum [SR] and the myofilaments play a major role in relaxation. We investigated the effects of isoproterenol, ryanodine, diseased states and two Ca²⁺-sensitizers: EMD 57033 and ORG 30029 on relaxation in isolated myocardium. Isoproterenol, which enhances SR function and desensitizes the myofilaments to Ca lf, shortened relaxation of both the twitch and the Ca²⁺ transient. Ryanodine and disease states such as cardiomyopathy and hypertrophy, which impair SR function, prolonged both the twitch and the Ca²⁺ transient. In this situation, relaxation became directly related to [Ca²⁺]_i cycling. Both EMD 57033 and ORG 30029 increased force of contraction, and diastolic force and prolonged twitch relaxation to a great degree with little effect on the Ca²⁺ transient.This would indicate that interactions at the level of the thin and thick filaments may affect relaxation without significant changes in intracellular calcium mobilization. We conclude that calcium mobilization can affect diastolic function in slowed muscle contractions, but time course changes of Ca²⁺ transients cannot be used alone to elucidate mechanisms of cardiac relaxation.

This summarizes my 25 year long research carried out for a better understanding of the systolic function of the heart. My research began with the discovery of Emax (end-systolic maximum elastance or pressure volume ratio) as an index of ventricular contractility on the basis of the time-varying elastance model of the left ventricle. The same time-varying elastance model evolved PVA (systolic pressure-volume area) as a measure of the total mechanical energy generated by ventricular contraction. We have found experimentally that both Emax and PVA closely correlate with myocardial oxygen consumption and proposed that the oxygen costs of Emax and PVA characterize the abnormality of cardiac mechanoenergetics under various pathophysiological conditions.

The power of the heart is dictated by the force development and velocity of shortening (V) of the cardiac sarcomere. Both depend on the amount of Ca⁺⁺ released by the sarcoplasmic reticulum during the action potential. We have investigated the inter-relationship between force (F) sarcomere length (SL) and V and the intracellular Ca⁺⁺ concentration ([Ca⁺⁺]_i) in trabeculae isolated from the right ventricle of rat heart. Activation of the contractile filaments during a normal heartbeat requires approximately 30 μM Ca⁺⁺ ions. which rapidly bind to cytosolic ligands. Consequently the [Ca⁺⁺]_i transient detected by intracellular probes is less than 2 μM. Length dependent binding of Ca⁺⁺ to Troponin-C is responsible for the shape of the F-SL relationship. Ca⁺⁺ ions are long enough bound to Troponin-C to allow the F-SL relationship, and consequently the end-systolic pressure volume relationship in the intact ventricle, to be largely - but not completely- independent of the loading conditions. V during contraction against a load increases hyperbolically with decreasing load. Stiffness studies revealed that the number of attached cross bridges increases in linear proportion to an increase of the external load. At low external loads the V was large enough to induce a substantial visco-elastic load within the sarcomere itself. The F-V relationship of a single cross-bridge appeared to be linear after correction for the observed visco-elastic properties of the muscle and for load dependence of the number of cross-bridges. Maximal V of sarcomere shortening (Vo) without an external load, depends on the level of activation by Ca⁺⁺ ions because of the internal viscous load. Our studies of the rate of ATP hydrolysis by the actin-activated S₁ fragment of myosin suggest that V_o is limited by the detachment rate of the cross bridge from actin. These studies also suggest that the difference between the fast (V₁) and slow (V₂) myosin isoenzyme can be explained by a difference in the amino acid domain on S₁ involved in binding of the cross bridge to the actin filament.