Core principles of cardiovascular physiology and pathophysiology in critical illness

Venous return

It may seem strange that vascular compliance, a static property, is so important for the generation of blood flow in the circulation. The explanation is that pulsatile flow in a tube occurs from a region with a high pressure to an area with a lower pressure. If the closed loop vasculature was made up only of stiff vessels, the pressure wave produced with each stroke volume would be instantaneously transmitted through the system and there would be no pressure difference to drive flow ( Fig. 69.4 ). Thus for flow to occur, the pressure wave and volume pulse have to be transiently taken up in a region that has a higher compliance; that region is the veins and venules, which have a large volume at a low pressure and as such are very compliant (see Fig. 69.4 ). This is analogous to the characteristics of a bathtub (see Fig. 69.2 ). Drainage from a bathtub is determined by the height of the water above the hole at the bottom. Inflow from the tap only increases outflow from the tub by adding volume and raising the height of water in the tub, but the pressure coming out of the taps does not alter flow from the drain. Similarly, the pressure going into the veins and venules does not determine outflow from the veins; only the volume entering per time (i.e., flow) does this.

Based on this discussion, the determinants of the return function are the volume that stretches the vascular walls, the compliance of the walls of the veins and venules, and the resistance to flow between the venous system and the RV (see Fig. 69.2 ). The compliance of arteries and capillaries only add a very small component to the total vascular compliance and can be ignored for simplification.

Guyton provided a useful graphical approach to understanding the return function ( Fig. 69.5 ). He plotted Pra on the x -axis because he considered it to be the variable being controlled, and thus the independent variable. The y -axis is flow (or venous return). The x -intercept is the MSFP. The lower the Pra, the greater the flow. The slope of the venous return line is −1/venous resistance (−1/Rv). This inverted relationship occurs because the inflow pressure, MSFP, remains relatively constant and the outflow pressure, Pra, is lowered by the heart. The function then is plotted against the outflow pressure, which is Pra. When Pra falls below its surrounding pressure, which is atmospheric pressure when breathing spontaneously, the floppy veins collapse where they enter the thoracic cavity, which has a negative pressure relative to atmospheric pressure. This does not stop flow but creates a flow limitation, what is called a vascular waterfall . When that happens, lowering Pra further does not increase venous return. This can be very evident when attempts are made to increase the speed of a mechanical device supporting the heart and there is insufficient upstream pressure to support the flow. When intrathoracic pressure is raised by mechanical ventilation, the collapse of veins occurs when Pra is less than pleural pressure.

It is important to introduce another term, and that is capacitance ( Fig. 69.6 ). By now it should be clear that the volume stretching veins and venules determines flow, rather than total blood volume. The reason is that under resting conditions 70% of blood volume simply rounds out vessel walls but does not stretch them. This is called “unstressed” volume . ^, Thus only about 30% of blood volume—the stressed volume—actually stretches vessel walls. This means that in a 70-kg person, only about 1.3–1.4 L of a total blood volume of 5.5 L actually creates MSFP, the force that is critical for the return of blood to the heart. This needs to be considered when infusing liters of volume to increase cardiac output.

Cardiac function

Output from the heart is determined by the rate of contractions (heart rate) and the amount ejected per beat (stroke volume), which in turn is determined by preload, afterload, and contractility ( Fig. 69.7 ). The fundamental role of preload was demonstrated by Otto Frank and later by Ernest Starling. They appreciated that the greater the initial stretch of cardiac muscle before the onset of contraction, the greater the force produced by the muscle. Starling even correctly hypothesized that this was because stretch of muscle fibers increases overlap of chemically active sites, an insight obtained without knowing about the existence of actin and myosin in sarcomeres, which are the basis of the current sliding filament hypothesis for muscle force development. ^, The significance of this length-tension property is that the exact volume that fills the heart during diastole is ejected on the next beat if all other properties, which are heart rate, contractility, and afterload, stay the same. Although the initial muscle length directly determines the force of contraction and the volume emptied, preload refers to a force, so that preload should be defined as the pressure required to stretch the sarcomere at a given compliance and not by the volume. This “law of the heart” provides perfect matching of what can be called the “stroke return” to the heart and the “stroke output,” or stroke volume. It also ensures that in the steady-state stroke output from the LV perfectly matches that of the RV. The importance of this can be understood by some simple calculations. If the RV pumps 100 mL per beat with a heart rate of 70 beats/min and the LV pumps 99 mL per beat—a 1% difference—and total blood volume is 5.5 L, in an hour and a half the total blood volume would be in the lungs! Thus matching of RV and LV outputs must be perfect over time. This can be a significant clinical problem when mechanical devices are used to bypass the ventricles. Unlike cardiac muscle, these devices have minimal “preload” responsiveness. If RV output is set greater than LV output, pulmonary edema rapidly develops.

Unlike skeletal muscle, cardiac muscle cannot be stretched beyond a defined limit. Thus there is a maximal end-diastolic volume and consequently maximal stroke volume. When this limit is reached, further volume loading only increases the diastolic pressure but does not increase muscle length and thus does not change stroke volume. Most often, cardiac filling is limited by the RV, which is very compliant within the range of normal diastolic volumes but has a sharp decrease in its compliance when filling becomes limited ^, (see Fig. 69.7 ). This limit to RV filling normally occurs partially because of a limit to stretch imposed by the pericardium, but also by the limit to stretch imposed by the myocardial wall. When RV filling is limited, stroke volume of the LV is also affected because the LV only can put out what the RV heart transfers to it. This RV limitation provides a protective mechanism that prevents overfilling of the LV and a consequent increase in pulmonary capillary pressures and pulmonary edema.

Ernest Starling provided a useful graphical approach to help illustrate the preload responsiveness of the heart (see Fig. 69.7 ). Right atrial diastolic pressure is plotted on the x -axis and the output of blood from the LV on the y -axis. (Starling actually plotted the output on the x -axis and Pra on the y -axis, but the real independent variable is Pra, which should be on the x -axis. ) This plot shows that cardiac output in response to an increase in preload rises to a maximum value; beyond that point a sharp plateau appears. No further increase in cardiac output occurs with further increases in Pra/CVP. The plot assumes that contractility, afterload, and heart rate remain unchanged. In this plot, the status of everything between the right atrium and the aortic output are included (see Fig. 69.1 ).

Another other way to understand function of the heart is through an analysis of the pressure-volume relationship of the ventricles ( Fig. 69.8 ). In this analysis, there is a passive diastolic filling curve, a phase of isovolumetric contraction until the ventricular outflow valve opens, ejection of the stroke volume, then isovolumetric relaxation, followed again by the phase of passive filling. The pressure-volume plot gives a clear explanation for why the steady-state stroke volume must match stroke return. It also indicates the roles of preload, afterload, contractility, and heart rate.

Preload is the final pressure before the onset of systole and, as such, determines the final stretch of myocardium before the onset of contraction. The passive filling curve also indicates the limit to ventricular filling ( Figs. 69.8 and 69.9 ).

Afterload refers to the force (pressure) the myocardium faces after it begins to contract. The value is approximated by the pressure at the opening of the outflow valves from the ventricles (pulmonary valve for the RV and aortic valve for the LV). The afterload really should be the generated tension, but because ventricular volume during ejection is falling while the pressure is rising, the tension is quasi-constant. The greater the afterload, the smaller the stroke volume (see Figs. 69.8 and 69.9 ). This restriction relates to the time available for contraction after valve opening, which will become clearer in the discussion that follows on contractility. Under normal conditions, changes in LV afterload only have a small effect on stroke volume, but the effect is much greater when contractility of the heart is decreased.

The third property is contractility. This refers to the speed and extent of shorting of the muscle for a given afterload and preload. This property is best explained by what Suga and Sagawa called “a time varying elastance” ( Fig. 69.10 ). ^{, ,} In very careful studies in isolated hearts from rabbits, they showed that the ventricular muscle becomes progressively stiffer (i.e., increases its elastance) during the contractions phase and then relaxes. On a plot of developed pressure versus ventricular volume, this appears as an elastance line that becomes steeper and steeper during the cardiac contraction cycle and then relaxes. The steepest elastance line gives the maximum possible pressure for any volume in the ventricle. The steepness of the line is determined by how quickly the muscle can stiffen in the time available for myocardial contraction. That time of contraction in turn depends on the amount of time for Ca ²⁺ to be released into the cytoplasm, the amount of Ca ²⁺ released, and how fast it is taken up during each cycle. These factors regulate the turnover of actin and myosin binding sites. Depression of cardiac function means that a lower maximum elastance slope is inscribed during the cycle. The consequence is a lower stroke volume for the same preload, afterload, and heart rate. The normal end-systolic elastance slope for the LV is very steep, which is why changes in afterload only have a small effect on stroke volume in a normal heart (see Fig. 69.9 ). The effect of afterload on the RV, though, is much greater because the maximum elastance of the RV is only about a third of that of the LV. The lesser slope of the elastance line of the RV accounts for the systolic pressure sensitivity of this ventricle.

Another factor that is considered by some to be a determinant of cardiac output is aortic elastance. This value is estimated from a line drawn from its end-diastolic volume and pressure point to the end-systolic volume-pressure point. It thus incorporates stroke volume and the pressure generated in the aorta at the end of systole; the change in pressure for change in volume gives elastance units. However, this is simply a reflection of ventricular afterload, which is already considered in the pressure-volume analysis of the ventricle and is likely better estimated by aortic valve opening. Based on Sagawa’s concept of a time-varying elastance, the pressure generated in the aorta or pulmonary artery by ventricular ejection can never be higher than the pressure generated by the ventricle so that the aortic properties do not provide a load but do affect aortic valve opening. Furthermore, what is being called aortic elastance is a dynamic elastance and is dominated by vascular resistance and the downstream critical closing pressure. Finally, active LV ejection ends in the middle to two-thirds of systole so that the ventricle is relaxing before the end-systolic pressure-volume point. The aortic elastance analysis therefore adds little information to the simple consideration that aortic valve opening gives an estimate of LV coupling with the aorta, and this is easily seen at the end-diastolic pressure.