16.1 Early Isolated Heart Preparations and the “Law of the Heart”

The first isolated heart preparation was developed by Cyon in 1866 at the renowned Physiological Institute in Leipzig. This was a recirculation model of a frog’s heart which aptly demonstrated the pumping action of the heart but was not amenable to hemodynamic recordings [1] (Fig. 16.1). Czermak is credited to have developed the first non-recirculating preparation in which direct measurements of chamber pressures were possible [3]. The importance of this preparation lies in the fact that it was further modified by Otto Frank, the youngest of the generation of investigators working under Carl Ludwig at the Leipzig Physiological Institute. Frank constructed a small, elegant preparation of the (single ventricle) frog heart, perfused with ox blood, in which preload and afterload could be altered independently (Fig. 16.2). The improved model allowed for a continuous recording of pressures and volume output. Frank measured maximal pressures generated by the ventricle when the aortic outflow was closed and observed that the heart muscle behaves similarly to the skeletal muscle, in which the force of contraction depends on the initial fiber length (preload) [2]. Frank also noted that the maximal tension (isometric maxima) developed by the ventricle depends on diastolic filling (preload) and was one of several investigators who contributed toward discovery of the “law of the heart” [4, 5]. Finally, Frank was the first to observe that the end-systolic loci of both isovolumic and ejecting contractions fall on the (same) straight line and are independent of the end-diastolic volume. However, he found it difficult to obtain consistent results when altering aortic pressure against which the heart was ejecting (afterload) and thereby recognized the complexity of the ventricular–aortic interaction [6, 7].

H. Newell Martin of Johns Hopkins was the first to construct the mammalian heart-lung preparation in 1881. The whole animal (cat or dog) was placed into a large warm tank. The blood filled the right atrium via superior vena cava by way of a constant pressure reservoir. The left ventricle ejected the oxygenated blood into the aorta fitted with a stopcock, by which ejection pressures could be manually adjusted. Newell Martin observed that rising of venous pressure resulted in increased output by the heart [9].

Ernest Starling resumed investigations on the circulation in 1910 when he turned his attention on the heart in an effort to quantify its mechanical function under varying conditions, since previous works on whole animals by others have given such widely differing results [9]. During the course of former studies on anesthetized, open-chest dogs, Starling noticed a marked weakening of the heart’s action and resolved to construct a model where heart rate, venous inflow, and arterial resistance could be varied independently [10, 11]. Over subsequent years, Starling and his coworkers performed a number of improvements of their original cat heart-lung preparation and finally settled on a dog’s model, in which lower circuit resistance would better represent the heart’s mechanical performance [9] (Fig. 16.3). Several studies followed, in which the effect of changes in heart rate, and of the atrial and aortic pressures were assessed in term of heart size (volume) and its output. It was shown, for example, that for a given venous inflow, cardiac output remains fairly constant over a broad range of arterial pressures and temperatures. When the inflow was increased (by rising of the venous reservoir beyond a certain point), the heart began to distend, exhibited diminished output and failed. Starling came to the conclusion that“…the rise of venous pressure must be regarded as one of the mechanical means which are operative in enabling the heart to maintain an output corresponding to the blood it receives from the venous system” [12]. In subsequent experiments, Patterson and Starling showed that the only factor which consistently varies during altering of venous inflow and aortic pressure is the size (volume) of the heart, i.e., the length of muscle fibers. Thus, in the heart-lung preparation, the right atrial pressure controls the degree of ventricular distension (preload) and is the major determinant of cardiac output. This relationship is curvilinear and shows that performance of the heart (Starling’s curve) increases to an optimal point at peak output, beyond which the heart distends and fails [10] (see Fig. 14.​2).

It is of interest that Starling did not express the “law of the heart” in terms of stroke volume or minute output. He was, of course, well aware of the fact that, “The heart-lung preparation will obviously give us no information as to the amount of blood which can be regarded as normal for that animal at rest,” let alone during exercise, since the heart was essentially “separated” from the animal [10].

The ultimate aim of Starling and his coworkers was to find the “underlying principle, on which the heart’s power of self-regulation may depend,” which would determine the force of its contractions. The analogy was therefore sought between the skeletal muscle where—according to contemporary research—the power of contraction is determined solely by the extending (linear) force and the ventricle, where the volumetric expansion (preload) was considered to play a similar role [13]. Starling later summarized this “regulating principle” in his Linacre lecture: “The law of the heart is thus the same as the law of the muscular tissue generally, that the energy of contraction, however measured, is a function of the length of the muscle fiber” [14].

It is clear that Starling and his collaborators could not quantitate the mechanical function of the heart. Despite the ingenuity of the model, Starling and collaborators were not able to control the multiplicity of variables which they encountered. It was later shown by several investigators that, a stepwise increase in ventricular filling pressure at constant aortic pressure (afterload), results not in a single, but in a family of Starling curves (Fig. 16.4). However, this did not deter Starling, a skilled communicator, from applying his experimental findings to a broad range of clinical scenarios [4]. Starling’s interpretation that the heart responds to increased work load with an increase in size (dilatation), polarized physiologists, and evoked misgivings on the part of clinicians of the early twentieth century, who considered an increased heart size as a sign of failure [4]. It was later shown by Hamilton and others that in acute settings, the heart becomes smaller during acceleration but increases in size when it slows down, because of increased filling time. Thus, changes in rate preclude the application of Starling’s work in the intact animal [15].

As pointed out by Elzinga [17] well over 20 years ago, Frank-Starling’s law has, notwithstanding its original form, undergone “a historical misinterpretation,” such as in this definition by Guyton: “Stating once again, this very important principle, known as the Starling’s ‘law of the heart’: within physiological limits the heart pumps all the blood that comes to it without allowing an excessive rise in peripheral venous pressure” [18]. Evidently, Guyton here adapted “the law” in support of his venous return model of circulation. In fact, most physiology texts to this day depict at least one form of “Starling’s curves,” where cardiac output, stroke volume, or stroke work is represented as a function of right atrial or ventricular filling pressure (see Fig. 14.​2).

16.2 An Obscure Model (Hydraulic Ram)

While Otto Frank performed his classical experiments on the isolated frog heart model, an article appeared in the Viennese Medical Weekly by K. Schmid entitled, “On the apex beat and the pulse waveforms” [19]. At the time, several theories about the origin of the apex impulse were discussed in the literature. Schmid proposed that the apex beat comes about when the contracted wall of the left ventricle strikes against the chest wall, due to a sudden deceleration of the moving blood against the closed mitral valve. This phenomenon, argued Schmid, is equivalent to a pressure surge created in the hydraulic ram when the flow of water is suddenly stopped by closure of the spill valve (Fig. 16.5). It is of note that in the late 1800s and the beginning of the 1900s, before the widespread use of electrically powered pumps, water rams were in common use. At the beginning of the twentieth century, R. Steiner made several references to Schmid’s article, the significance of which goes well beyond the generation of the apex impulse. As already mentioned in Part I, Steiner maintained that the blood moves autonomously and that the heart functions as a damming-up organ whose mechanical function can be compared to a hydraulic ram which is flow activated [20]. This theory, however, was well ahead of its time and was largely unnoticed. Over the years, sporadic studies appeared, such as a paper by Havlicek, who drew a mechanical and morphological analogy between the heart and the hydraulic ram and even constructed a physiological model of the hydraulic ram [21]. The most consistent efforts “to put the heart in its place” came from Manteuffel-Szoege and his collaborators, who systematically pursued the issue of autonomous blood movement from embryological and hemodynamic perspectives. In a review paper on the subject, Manteuffel-Szoege made the following observation: “Is it really true that the heart works like a pump? A pump sucks in fluid from a reservoir, which is a hydrostatic system and not a hydrodynamic one. In the circulation, on the other hand, not only is blood ejected from the heart, but it flows into the heart. The heart is a mechanism inserted into the blood circuit, and so it is a very peculiar kind of pump” [22]. The monograph summarizing his life’s work was published posthumously in 1977 [23]. What then is the function of a hydraulic ram?

The hydraulic ram is a cyclical water pump in which the kinetic energy of water is converted into pressure.¹ The unique feature of the hydro-ram is that the hydraulic power of water does not power the machine parts, such as turbine blades, but works on itself and therefore does not require an external source of power for its operation. The ram is thus an impedance pump which generates pressure, but not flow. (The hydro-ram’s components and work cycle are depicted in Fig. 16.6).

Please note that the pressure vessel (D) is not essential for operation of the ram but increases its efficiency and converts pulsatile flow from the delivery pipe into a steady flow. (The working cycle of such a simplified version of the water ram described by Schmid is depicted in Fig. 16.7.)

Automatic operation of the ram requires careful “tuning” of the pipe diameters and adjustment of the valves. The efficiency (the ratio of delivered vs. wasted water) of the ram varies between 20% and 60% and depends on construction design, on the amount of available water and the desirable working pressure. Like other industrial pumps, each hydro-ram operates at its own working point and can be adjusted to deliver larger flows at smaller pressures or less flow at greater pressures [25].

Basfeld demonstrated on a model ram that by increasing the compliance of the working chamber, i.e., by gradually increasing the ratio of air to water, the amount of water pumped by the ram quickly reaches a peak and remains at optimal level. In terms of efficiency, the ram delivers greater volumes when it operates at lower frequencies and lower working pressures (delivery height). However, when ejecting at higher pressures, the delivered volume is optimal at higher frequencies of operation [26].

Of note is the fact that functionally, the hydraulic ram bears more than a casual resemblance to the isolated heart preparation (see Fig. 16.8).

Thus, the great veins and the atrium represent the reservoir from which the blood empties into the ventricle (atrial systole). Abrupt closure of the mitral (spill) valve causes a steep rise in pressure and opens the aortic (delivery) valve. The air cushion of the ram can be compared to the contractile elements of the heart, i.e., the myocardium. Unlike in the heart where the pressure oscillates between systolic and end-diastolic, the pressure in the ram’s pressure vessel (Windkessel) is high throughout the work cycle. One could object that there is no “spill valve” in the isolated heart preparation, however—the compliant shape of the ventricle, which promotes blood’s vortical movement in diastole, together with systolic ejection—achieves efficient conversion of kinetic energy of flow into pressure, a critical element of the ram’s function. Morphological features of the right ventricle with a thin, highly compliant wall and a long, gently curving outflow tract, suggest that its ram-like function is adapted for high flows at low pressures. The opposite is the case with the left ventricle, where a short, acute-angled outflow tract and a thick, poorly compliant wall is “designed” to generate high pressures (cf. Fig. 13.​13). Last but not least, considering the low energetic efficiency of the heart (13–20%) and of the ram, a mechanical comparison between the two is not too far off the mark. In fact, in a recent editorial on the form and function of the right ventricle, Sengupta and Narula submit: “The most compelling aspect of the analogy is the development of a pulsatile flow from a continuous flow and the striking similarity between the pressure curve tracings obtained from the water ram and the ventricle” [27].

Manteuffel-Szoege was, unfortunately, not familiar with the work done on the isolated heart preparation and did not compare the two; however, he sought out correspondences between the function of the ram and the heart and constructed his own prototype of a model ram [22, 23]. Like the heart, the ram ejects only a part of delivered volume from the working chamber (ejection fraction), and in both, the amount of volume delivered (cardiac output) depends on the height of the reservoir (filling pressure) and loading conditions (aortic pressure).

Further underscoring the similarity between the ram and the heart’s action is the fact that the total heart volume throughout the cardiac cycle remains virtually the same. Employing the cine-MRI and echo planimetry in healthy volunteers at rest, Carlson at al. found that the total heart volume varies between 5% and 13% between diastole and systole, with the largest contribution due to changes at the region at the base of the heart (level of the atrioventricular valve plane) [28]. Up to 64% (range 58–73%), blood volume ejected by the ventricles was replaced by filling of the atrial reservoirs. This number increases up to 80% in faster and smaller hearts [29]. The discrepancy in size is likely to diminish during exercise at higher heart rates and cardiac outputs and reduced diastolic filing times (cf. Sect. 17.​2).

Cyclical performance of the ram can, moreover, be compared with a time-varying elastance model of the ventricle proposed by Suga [30] (see Sect. 16.4). It is significant that, within given design constraints and experimental settings, the ram works at optimal power and efficiency which depend on the hydraulic energy of the driving flow. This is not unlike the heart which works at (hitherto unexplained) optimal power and efficiency. The fundamental discrepancy, known to exist between the cardiac oxygen consumption and cardiac output, can be compared to the operation of the ram in which only a part of the total water energy “driving” the ram is converted into useful work and the rest is “lost” through the spill valve. Thus the comparison, evoked by Steiner [31], between the hydraulic ram and the heart as a damming-up organ is teleologically sound and clarifies a number of unresolved issues with the existing models of the heart.

16.3 Quantification of Ventricular Pump

Between the 1950s and 1970s, numerous indices appeared in the literature which purported to embody various aspects of the mechanical heart performance, such as the ejection fraction, peak systolic pressure, stroke work, maximal velocity of shortening, and peak isovolumic pressure at a given volume (dP/dt_max), to name a few, but, as noted by Elzinga and Westerhof, “… there is no theoretical reason to think that these indices are in any way quantitatively related to the amount of blood the heart can pump against various pressures” [32]. In order to do that, the heart performance would have to be tested against a system in which the mechanical parameters of the arterial circulation could be controlled. Thus’ a model arterial tree was constructed, a three-element Windkessel,² consisting of mechanical modules which would simulate compliance (air chamber, i.e., Windkessel) (Fig. 16.9), resistance and characteristic impedance. It was shown that the impedance values for systemic and pulmonary circulations obtained with this model under experimental conditions were similar to in vivo results in man, cat, and dog [33].

Variations of the three-element Windkessel model have been used extensively as arterial loads for testing of isolated hearts and for modeling of ventriculo–arterial interaction and derivation of CO (for review see [34]). A sophisticated isolated heart preparation was then developed by Elzinga and coworkers to test the performance of the left ventricle against the arterial model on a beat-to-beat basis. Carefully executed experiments on the isolated cat heart during conditions of constant diastolic filling, contractility and fixed heart rate, against different values of aortic impedance, showed that the left ventricle is neither a source of pressure, nor flow, but a pump with a final “source resistance” related to its load [35]. The model was based on the electrical (Ohmic) analog where the driving pressure produced by the pump corresponds to “hydromotive pressure,” i.e., voltage, the flow is equivalent to current, and resistance to “apparent source resistance” which is defined as the ratio of mean aortic pressure and mean outflow. The resistance of the arterial system (peripheral resistance), on the other hand, is represented as the ratio of mean aortic pressure and mean flow [36–38].

According to this model, interaction of the ventricle and the arterial tree is expressed in terms of mean ventricular pressure and flow (cardiac output) and is graphically represented as a “pump function graph,” which closely resembles a mechanical pump head–capacity curve (Fig. 16.10a). An example of such a pressure head–capacity curve for a roller pump is given in Fig. 6.​1. It is evident from the graph that when working against increasing loads (aortic pressure), the isolated heart will generate smaller stroke volumes, eventually reaching the isovolumic state at maximal pressure and no flow. The reverse is the case at decreasing aortic pressures. The graph further shows that the working point of the ventricle falls roughly midway between the maximal flow and pressure. Accounting for the pressure differences, experiments on the right ventricle closely match those of the left [39, 40].

More importantly, they observe that the receptors which would detect power and/or efficiency have not been discovered and would be difficult to imagine since any value, lower than optimal, would have to occur twice [35]. A similar issue has been raised by others [36, 43, 45]. In spite of the limitations of the conventional model to describe the phenomenon, the debate whether the heart is a “flow source” or a “pressure source” continues unabated [46–48].

Is it possible that the isolated model heart does in fact function according to sound design (natural and mechanical) principles emulated, however, by a different mechanical model? It is proposed that this model is a hydraulic ram which can be made to operate at two extremes: at high load, by clamping the delivery pipe, when the ram generates maximum pressures, but no flow. This scenario, as demonstrated on the isolated cat heart preparation, is analogous to isovolumic contractions, where power, the product of pressure and flow, is zero. The reverse situation occurs when the ram pumps at low pressures and generates maximal flows at optimal power and efficiency. This state is comparable to the heart working against low aortic pressures [36, 43]. The often-quoted statement by Starling that, “the heart can only pump as much blood as it receives,” certainly sounds ambiguous when used in the context of the heart as a pressure-propulsion pump, but is an accurate observation, if the heart is considered to function as a hydraulic ram.

16.4 Ventricular Elastance Model

An alternative way to represent the left ventricular pump function is by a time course of the ventricular pressure–volume relationship. What Frank was unable to demonstrate on a frog heart, namely that a unique relationship exists for isovolumic and ejecting beats, was shown some 70 years later by Suga, who had an intuitive insight, that the action of the left ventricle could be modeled by discharge of a capacitor, i.e., by a time-varying function. Suga tested the hypothesis on isolated, perfused, and denervated dog hearts, by measuring ventricular volumes during a graded occlusion of the inferior vena cava, until a volume was reached, when the ventricle could no longer generate pressure (V_d) (see Fig. 10.​1a). In a separate set of experiments, the ascending aorta was occluded in a stepwise fashion, in order to control the pressure against which the ventricle was ejecting [49, 50]. The working cycle of a ventricle can thus be represented on a pressure/volume plane in which a data point moves in counterclockwise direction and describes a loop for a given set of loading conditions and contractility. Suga showed that the relationship between pressure and volume in the LV is described by an elastance curve according to the following equation:

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

1. Perturbation Experiments
2. Heart as a Generator of Pressure
3. 21. Microcirculation
4. and Respiration: Toward a Functional Chronobiology

Stay updated, free articles. Join our Telegram channel

Join