Anatomy

The organ of prenatal respiration is the placenta. It is a large, low-resistance circuit that has an enormous influence on the pattern of fetal blood flow. The lungs are almost completely excluded from fetal circulation, and three special shunts (the ductus venosus, the foramen ovale, and the ductus arteriosus) allow the most oxygenated blood to perfuse the heart and brain (Fig. 4-1). Two umbilical arteries originate from the internal iliac arteries and deliver fetal blood to the placenta. One umbilical vein carries oxygenated blood from the placenta to the fetus. When umbilical venous blood approaches the liver, it can take two pathways. It is estimated that 50% to 60% of umbilical venous blood bypasses the liver via the ductus venosus, while the remainder perfuses the left lobe of the liver. Blood flow to the right lobe of the liver is predominantly from the portal circulation. The right and left hepatic veins, along with the ductus venosus, merge into the suprahepatic inferior vena cava (IVC) (Fig. 4-2). Bypassing the high-resistance hepatic microcirculation, umbilical venous blood in the ductus venosus remains not only more oxygenated, but it also flows at a higher velocity. Within the suprahepatic IVC there are now two streams of blood. The stream of blood with higher velocity is derived from the ductus venosus and drainage from the left hepatic vein. The stream of blood with slower velocity consists of drainage from the right hepatic vein mixed with venous return from the abdominal IVC. Upon entering the right atrium, these two streams of blood diverge. The blood with higher velocity is primarily directed across the foramen ovale to the left atrium. This right-to-left shunt is possible because left atrial pressure is low due to minimal pulmonary venous return. The Eustachian valve is a flap of tissue at the junction of the IVC and the right atrium. It functions to help direct the higher-velocity stream of blood across the foramen ovale and into the left atrium (Fig. 4-3). The lower-velocity stream of blood crosses the tricuspid valve and is ejected by the right ventricle. This anatomic arrangement allows the most oxygenated blood from the umbilical vein to bypass the liver and the right side of heart.

FIGURE 4-1 Fetal circulation. Ao, Aorta; DA, ductus arteriosus; DV, ductus venosus; LA, left atrium; LV, left ventricle; PA, pulmonary artery; RA, right atrium; RV, right ventricle.

(From Rudolph AM: The fetal circulation and postnatal adaptation. In Rudolph AM, editor: Congenital diseases of the heart, ed 2, Armonk, NY, 2001, Futura.)

FIGURE 4-2 The course of umbilical venous blood as it reaches the liver. The left lobe of the liver is supplied mainly by the umbilical vein, and the right lobe of the liver is supplied by the portal circulation. IVC, Inferior vena cava; SVC, superior vena cava; LHV, left hepatic vein; RHV, right hepatic vein.

(Modified from Rudolph AM: The fetal circulation and postnatal adaptation. In Rudolph AM, editor: Congenital diseases of the heart, ed 2, Armonk, NY, 2001, Futura.)

FIGURE 4-3 Transesophageal echocardiogram of an adult patient with a prominent Eustachian valve remnant (arrow). The Eustachian valve directs blood from the inferior vena cava across the foramen ovale into the left atrium. The thinnest part of the atrial septum is where the foramen ovale is during fetal life. IVC, Inferior vena cava; RA, right atrium; LA, left atrium; FO, foramen ovale.

(From Sawhney N, Palakodeti V, Raisinghani A et al.: Eustachian valve endocarditis: a case series and analysis of the literature, J Am Soc Echocardiogr 14:11, 2001.)

The majority of blood in the left atrium originates from the higher-velocity stream that crosses the foramen ovale. In the left atrium it mixes with a small amount of pulmonary venous return. The purpose of the ductus venosus and foramen ovale is to allow the most oxygenated blood from the umbilical vein to reach the left ventricle with the least drop in oxygen saturation possible. Oxygen saturation in the umbilical vein is approximately 80%. Inevitable mixing with more deoxygenated blood from the liver, IVC, and superior vena cava (SVC) results in the left ventricle ejecting blood with a saturation of 65% to 70% when the mother is breathing room air. The left ventricle pumps blood primarily to the heart and brain through the ascending aorta and great vessels of the aortic arch. SVC blood enters the right atrium and crosses the tricuspid valve into the right ventricle. Only a small amount of SVC blood and poorly oxygenated IVC blood enters the left atrium via the foramen ovale. High pulmonary vascular resistance (PVR) forces almost all of the right ventricular output to enter the systemic circulation via the ductus arteriosus. The ductus arteriosus is the continuation of the main pulmonary artery and inserts into the aorta at a point immediately distal to the origin of the left subclavian artery. The blood in the descending aorta is a mixture of left and right ventricular outputs with the right ventricle predominating. Consequently, the gut, kidneys, and lower extremities are perfused with blood that has an oxygen saturation of approximately 55%. Two umbilical arteries, branches from the internal iliac arteries, return blood to the placenta.

The key features of fetal circulation are shown in Figure 4-4 and listed below:

FIGURE 4-4 Fetal and neonatal circulation. Circled values are oxygen saturation, with values for pressure (systolic, diastolic, and mean) appearing beside their respective chamber or vessel. A, Fetal circulation near term. B, Transitional circulation at younger than 1 day old. C, Neonatal circulation at several days old. Ao, Aorta; DA, ductus arteriosus; IVC, inferior vena cava; LA, left atrium; LV, left ventricle; PA, pulmonary artery; PV, pulmonary veins; RA, right atrium; RV, right ventricle; SVC, superior vena cava.

(From Rudolph AM: The fetal circulation and Rudolph AM: Changes in the circulation after birth. In Rudolph AM, editor: Congenital diseases of the heart, Chicago, 1974, Mosby.)

Fetal Cardiac Output

The three important fetal shunts create a circulation that is parallel rather than the more efficient postnatal series circulation. Furthermore, the shunts do not function perfectly, which increases the work of the heart. Some of the oxygenated umbilical venous blood enters the right ventricle, crosses the ductus arteriosus, and flows into the descending aorta. Delivering this highly oxygenated blood back to the placenta is the equivalent of a left-to-right shunt, which increases the myocardial work. Additionally, some deoxygenated blood from the SVC and IVC flows across the foramen ovale and is ejected by the left ventricle. This is an effective right-to-left shunt, which decreases the oxygen saturation of the blood ejected by the left ventricle. The fetus and most critically, its very metabolically active heart and brain, grow and develop in a cyanotic environment.

In the fetus, right and left ventricular outputs are not equal. The ventricles, acting in parallel, pump different amounts of blood, and organs receive blood flow from both ventricles. Thus, it is customary to refer to the combined ventricular output (CVO) of fetal circulation. The fetal CVO is estimated at 400 mL/kg per minute, a value that is similar to that of the neonate but almost threefold higher than it is in adults. The ratio of right ventricular to left ventricular output is approximately 1.3:1 (Kenny et al., 1986). The volume load borne by the right ventricle combined with the high fetal PVR results in significant hypertrophy. After birth, the right ventricle remodels over time under the influence of a series circulation and lowered PVR (Fig. 4-5). The demands of a parallel circulation result in increased myocardial work superimposed on the demands of a fetus having to grow and develop in a cyanotic milieu. Nevertheless, the fetal circulation is an efficient arrangement with adequate physiologic reserve. This is demonstrated by the full spectrum of congenital heart lesions that are well tolerated in utero.

FIGURE 4-5 Transesophageal echocardiogram four-chamber view of a healthy newborn, A, and adult B. The ventricles are equal in the newborn heart. Growth of the left ventricle and remodeling of the right ventricle create the usual adult appearance.

Oxygen Delivery

The transfer of oxygen across the placental bed is inefficient when compared with the transfer of oxygen in the lungs. The normal alveolar-arterial gradient for oxygen is minimal for a patient with healthy lungs breathing room air. This is in contrast to the fetal state, in which maternal arterial oxygen tension (Pao₂) is close to 100 mm Hg, and the partial pressure of oxygen (Po₂) in the umbilical vein is no more than 30 to 35 mm Hg. It should be recalled that after umbilical venous blood is diluted with more poorly oxygenated blood, the Po₂ of the blood ejected by the left ventricle is about 25 to 30 mm Hg when the mother is breathing room air. With such a low fetal oxygen tension, any reductions in maternal oxygenation could severely impact the fetus. The large gradient for oxygen across the placenta means that when the mother is breathing 100% oxygen, maternal Pao₂ can be as high as 400 to 500 mm Hg, but the Po₂ in the umbilical vein rises to no higher than 40 mm Hg. In cases of suspected uteroplacental compromise, this small increase in umbilical vein Po₂ may be critical and justifies administering supplemental oxygen to the mother. The modest rise in umbilical vein Po₂ that occurs when the mother is breathing 100% oxygen can result in a large increase in fetal oxygen saturation. This occurs because fetal hemoglobin has a different oxygen-hemoglobin dissociation curve than adult hemoglobin.

Oxygen transport must be achieved in a relatively hypoxic environment (Lister et al., 1979). How does the fetus ensure adequate oxygen delivery? Fetal hemoglobin (HbF) has unique properties that allow the fetus to transport oxygen despite a low Po₂. Approximately 80% of fetal hemoglobin is HbF compared with an adult who has over 90% adult hemoglobin (HbA). The Pao₂ at which Hb is 50% saturated is called the P₅₀. Fetal hemoglobin (HbF P₅₀:19 mm Hg) is shifted to the left in comparison with adult hemoglobin (HbA P₅₀:26 mm Hg). A low level of 2,3-diphosphoglycerate (2,3-DPG) and the decreased affinity of HbF for 2,3-DPG cause the leftward shift of HbF. In vitro, HbF has a sigmoidal oxygen dissociation curve similar to that of HbA. The result is that for any given Po₂, the oxygen saturation is higher for HbF than for HbA.

Assuming a Pao₂ of 30 mm Hg in the blood ejected by the left ventricle, this value falls in the steepest part of the oxygen-hemoglobin dissociation curve. At a Pao₂ of 30 mm Hg, fetal hemoglobin would be approximately 70% saturated, and adult hemoglobin would only be 50% saturated (Fig. 4-6). Using the equation for oxygen content of the blood (Cao₂) demonstrates how the fetus can achieve levels of oxygen transport that are near those of an adult. Oxygen content of blood is defined by the following equation, Sao₂ is the arterial oxygen saturation of hemoglobin, and Hb is the hemoglobin concentration in g/dL:

FIGURE 4-6 Oxygen-hemoglobin dissociation curves for fetal (A) and adult (B) hemoglobin. The dashed lines represent the P₅₀ for fetal and adult hemoglobin, respectively. For any given Po₂, fetal hemoglobin has a higher oxygen saturation than adult hemoglobin.

(From Delivoria-Papadopoulos M, McGowan JE: Oxygen transport and delivery. In Polin RA, Fox WW, Abman SH, editors: Fetal and neonatal physiology, ed 3, Philadelphia, 2004, Saunders.)

At normal levels of Pao₂, when an adult is breathing room air, the amount of oxygen dissolved in blood is negligible, because it is only 0.003 mL O₂/mm Hg. Therefore, the Cao₂ is typically calculated by determining the amount of oxygen that is carried bound to hemoglobin. For example, an adult with a Sao₂ of 100% and a Hb of 11 g/dL would have the following Cao₂:

The fetus maintains Cao₂ through two mechanisms. In addition to the leftward shift of HbF, the fetus is erythrocytotic compared with the adult. Using the equation for Cao₂ for a fetus with a Hb of 17 g/dL and an oxygen saturation of 65% yields the following:

Fetal hemoglobin’s greater affinity for oxygen improves oxygen uptake at the placenta. A greater affinity for oxygen is an advantage for uptake at the placenta but a drawback for the unloading of oxygen at the tissue level. Given that the purpose of hemoglobin is to deliver oxygen to the tissues, this poses a problem. The fetus copes with this problem with another modification to its internal milieu. Oxygen release, or rightward shifting of the oxygen-hemoglobin curve, is increased by acidosis. Fetal pH (normal values 7.25–7.35) is lower than it is in adults, facilitating oxygen release at the tissue level. It is important to realize that the preceding discussion has involved only oxygen content of the blood. Oxygen delivery, which is the goal, can only be achieved when an adequate capacity to carry oxygen is matched with a cardiac output that is sufficient to meet metabolic needs.

Pulmonary Blood Flow and Pulmonary Vascular Resistance

It is clear that the most fundamental and critical transition necessary for postnatal life is the establishment of breathing, accompanied by a fall in PVR. Therapeutic attempts to manipulate PVR occur in certain newborns with congenital heart disease. Deleterious changes in PVR also accompany critical neonatal disease states. Therefore, a brief discussion of pulmonary circulatory physiology is warranted.

At midgestation, PVR is estimated to be tenfold higher than it is 24 hours after an uncomplicated birth. During the last trimester, PVR decreases slightly to levels seven- to eightfold greater than it is 24 hours after delivery (Rudolph, 1979). This reduction results from the physical growth of the pulmonary vasculature that increases the cross-sectional area by more than the corresponding increase in blood flow. Yet, PVR still remains high immediately prior to birth and must fall dramatically in the first postnatal minutes to ensure survival. Pulmonary blood flow increases by an amount corresponding to the decrease in PVR, and pulmonary blood pressure falls by 50% (Fig. 4-7). Thus, the pulmonary vasculature presents a puzzle that flies in the face of the normal principles of cardiovascular embryology. The guiding principle of cardiac and vascular development is that once normal structures form, the correct pathway for blood flow is established. Once flow is established, normal growth ensues. The pulmonary vasculature must grow normally despite greatly reduced flow. The elevated PVR necessary for fetal existence cannot be solely the result of anatomic hypoplasia. If this were the case, the transition to postnatal life would be rocky indeed. Rather, the pulmonary vessels must be of normal size but retain the ability to markedly vasoconstrict. Even more remarkable is that the ductus arteriosus is an outgrowth of the pulmonary arterial system, yet it responds in a completely opposite fashion to the pulmonary vasculature. The very stimuli that raise PVR in utero cause dilation of the ductus arteriosus. The factors that govern pulmonary vasoreactivity have still not been fully elucidated, but our understanding has grown considerably in the last 20 years (Martin et al., 2006).

FIGURE 4-7 A, The changes in pulmonary arterial pressure, blood flow, and PVR from late gestation through the neonatal period in lambs and other species. B, The effect of oxygen tension and acid–base balance on PVR in newborn calves. A low Po₂ results in minimal increase in PVR if pH is maintained at 7.4. Conversely, dramatic rises in PVR occur when acidosis is combined with a low Po₂.

(From Rudolph AM: Prenatal and postnatal pulmonary circulation. In Rudolph AM, editor: Congenital diseases of the heart, ed 2, Armonk, 2001, Futura.)

The primary force driving high fetal PVR is hypoxemia. The estimated Po₂ in the pulmonary arteries is less than 20 mm Hg. This phenomenon of hypoxemia-induced pulmonary vasoconstriction (HPV) persists into adulthood, where it is vital in maintaining oxygen saturation during one-lung anesthesia. In utero, the mechanism of HPV is not fully understood. Oxygen is a potent stimulator of endothelial- derived vasodilating substances such as nitric oxide and prostacyclin. Nitric oxide activates soluble guanylate cyclase, which increases guanylate 3′-5′-cyclic monophosphate (cGMP). Prostacyclin stimulates adenylate cyclase to increase adenylate or adenosine 3′-5′-cyclic monophosphate (cAMP). Both cGMP and cAMP initiate vasodilation in pulmonary vessels. Under conditions of low oxygen tension, the release of nitric oxide and prostacyclin is presumably attenuated, tipping the balance in favor of vasoconstriction. The specific substances inducing pulmonary vasoconstriction are not well understood. Attention is focused on arachidonic acid and its metabolites and the potent vasoconstrictor endothelin (Tod and Cassin, 1984; Hickey et al., 1985a; Ivy et al., 1996).

At delivery, mechanical and biochemical factors lead to the abrupt fall in PVR (Teitel et al., 1990). Aeration of the previously fluid-filled lungs removes the external compressive force on the pulmonary vasculature. Responding to the sudden rise in oxygen tension, the endothelium secretes potent vasodilators, nitric oxide and prostacyclin. Luminal diameter increases as endothelial and smooth muscle cells become thinner. The increase in blood flow further recruits small lumen vessels, leading to an overall increase in the cross-sectional area of the pulmonary vascular bed. Smooth muscle relaxation occurs in the larger pulmonary vessels. This first, rapid phase of pulmonary vasodilation is followed by a period of remodeling that lasts for months. During this time there is maturation of vascular smooth muscles with continuing modest declines in PVR. PVR approximates adult values by about 2 months of age, and the remodeling process is usually complete by 6 months of age. During the first few years of life, new vessels develop to supply the growing lung parenchyma (Fig. 4-8). It is important to understand that in the early postnatal period, PVR is markedly affected by hypoxia and acidosis (Rudolph and Yuan, 1966). More recently, pain has also been associated with increasing PVR.

FIGURE 4-8 Pulmonary angiograms at birth (A) and from an 18-month-old infant (B). Growth of existing vessels and development of intra-acinar arteries create the prominent hazy background appearance.

(From Haworth SG: Pulmonary vascular development. In Long WA, editor: Fetal and neonatal cardiology, Philadelphia, 1990, Saunders.)

Preload

In 1895, the German physiologist Otto Frank published his observations on the relationship of diastolic filling of the heart and the pressure the heart was able to generate during systole (Frank, 1895). Ernest Starling, an English physiologist, conducted the classic experiment in the early 1900s that defined the length-tension relationship for cardiac muscle (Fig. 4-10). Understanding his experiment is important, because in the laboratory he was able to isolate that which is impossible to isolate in vivo. Using a fixed weight to keep afterload constant, he was able to prove that the tension developed was proportional to the length of the muscle strip prior to stimulation. The greater the length of the muscle strip, the greater the tension it was able to develop. The Frank-Starling mechanism, or Starling’s Law of the Heart, is taken from a famous lecture by Dr. Starling himself, in which he stated, “the energy of contraction, however measured, is a function of the length of the muscle fiber” (Starling, 1918).

FIGURE 4-10 A schematic diagram of the classic Starling experiment. The suspended weight (afterload) is held constant, allowing the tension developed to be measured at different lengths (preload) of the muscle strip.

(From Epstein D, Wetzel RC: Cardiovascular physiology and shock. In Nichols DG, Ungerleider RM, Spevak PS, et al., editors: Critical heart disease in infants and children, ed 2, Philadelphia, 2006, Mosby.)

The clinical correlates of length and tension are left ventricular end-diastolic volume (LVEDV) and stroke volume, respectively. In the laboratory, the length of a muscle strip is easily determined because both ends can be fixed. This is very different from the intact heart, where the ventricle is a three-dimensional structure with complex geometry that defies the simple concept of length. Nevertheless, it is clinically appropriate to regard length as LVEDV. The LVEDV represents the loading of the ventricle. It is present before contraction, and therefore has come to be called preload. The force of ventricular contraction increases with increasing preload until a point is reached where the ventricle is over stretched and the force of contraction decreases. This point is the apex of the well-known Starling curve. Force of contraction is not synonymous with contractility, and for all points along a Starling curve, contractility is equal. Changes in the force of contraction, however, occur when preload changes (Fig. 4-11). Laboratory evidence has shown that resting sarcomere distance is 1.6 microns, with optimal conditions occurring at 2.2 microns. Excessive stretch, causing decreased force of contraction, does not occur until sarcomere length reaches 3.5 microns (Sonnenblick, 1974). In the laboratory the force of contraction increases until the sarcomere is stretched to over twice its resting length. If this translated fully to the intact heart, any increase in preload would result in increased force of contraction, because physiologically achieving a doubling of resting sarcomere length is almost impossible. However, well before a doubling of resting sarcomere length is reached, the patient falls on to the descending limb of the Starling curve. The reason is the relationship between volume and pressure, which is known as compliance.

FIGURE 4-11 Starling’s Law of the Heart. Contractility is the same for each curve. Moving from point A to point B requires an increase in LVEDV. Moving to point C or point D can only occur if there is a change in contractility. For the same LVEDV, the stroke volume falls progressively with impaired contractility. The figure demonstrates that the failing ventricle is dependent on preload.

(Modified from Opie LH, Perlroth MG: Ventricular function. In Opie LH, editor: The heart: physiology from cell to circulation. Philadelphia, 1998, Lipppincott-Raven.)

The Starling curve describes the relationship between LVEDV and stroke volume. The problem is that measuring LVEDV in a way that is simple, accurate, and available in real time is not possible. Echocardiography is the best method to quantify LVEDV. However, the two-dimensional images seen echocardiographically may not faithfully represent the volume contained in a three-dimensional ventricle with its elliptical shape. The equations used to objectively measure volume are not practical in most clinical situations and certainly not in the dynamic environment of the operating room. Subjective assessments of LVEDV using echocardiography depend on the skill of the operator and will detect only the extremes of preload conditions. Given the problems in measuring volume, the clinician is forced to use a surrogate measure of LVEDV. A catheter placed in a central vein and connected to a transducer measures pressures simply, accurately and continuously. Using pressure as a guide to volume requires understanding the relationship between the two variables, defined as compliance and described by the following equation:

The change in pressure resulting from a change in volume is compliance. Volume is in the numerator of the equation meaning that if a large increase in volume is met by a relatively small rise in pressure, compliance is high. As a completely empty ventricle is filled, the initial rise in LVEDP is small. In this situation the ventricle is said to be compliant. When the ventricle is full, a small increase in volume results in a large increase in pressure. The ventricle is now poorly compliant or “stiff.” As with the Starling curve, it is important to note that moving to different points on the compliance curve does not mean that the intrinsic compliance of the ventricle has changed. A true intrinsic change in compliance will be reflected in a new compliance curve (Fig. 4-12). In adult cardiac medicine it is readily appreciated that ischemia, infarction, or hypertrophy result in a stiffer, or less compliant, ventricle. However, no ventricle, regardless of its intrinsic compliance, has an infinite ability to accept volume. Eventually the inflection point on the curve will be reached and pressure will climb rapidly for any given increase in volume. It is this sharp increase in pressure beyond the inflection point that is responsible for the descending limb of the Staring curve.

FIGURE 4-12 Diastolic-compliance curves. Each curve represents a different intrinsic compliance. Movement between points A, B, and C reflects a change in preload with no change in compliance. Movement from point A to point D can only occur when the ventricle becomes less compliant.

(From Mark JB: Atlas of cardiovascular monitoring, New York, 1998, Churchill Livingstone.)

Decreases in the force of contraction represented by the descending limb of the Starling curve are due to the result of pressure and not volume or “overstretch.” The reason is that the rise in pressure causes an imbalance in the myocardial oxygen supply/demand ratio that results in decreased force of contraction. The left ventricle is perfused according to the following equation:

When the maximum increase in blood pressure achieved by volume loading has been reached, further volume only serves to increase LVEDP at the expense of coronary perfusion pressure. Thus, myocardial oxygen supply decreases. A key determinant of myocardial oxygen demand is wall stress defined by the modified Law of Laplace:

Wall stress is usually, but erroneously, thought of as similar to afterload, because the value for pressure in the numerator is assumed to be systolic blood pressure. In fact, blood pressure is a key component of wall stress, but they are clearly not the same. Oxygen consumption increases with increasing wall stress, and wall stress has both systolic and diastolic components. During systole the value for pressure in the wall stress equation is the systolic blood pressure. In the diastole the value for pressure is LVEDP. A patient with a value on the steep part of the compliance curve has sharp increases in LVEDP disproportionate to the increase in volume. Concomitantly, the ventricular cavity is fully distended from the volume load. Thus, excessively high LVEDP combined with an enlarged ventricular radius simultaneously causes an increase in myocardial oxygen demand and a decrease in myocardial oxygen supply. This situation, if unchecked, leads to ischemia and a decrease in the force of contraction.

The response of the newborn heart to volume loading (relative insensitivity) has been the source of great confusion (Rudolph, 1974). This confusion stems from experimental work done in the 1970s and is the logical extension of the known structural differences in the newborn heart (Romero and Friedman, 1979; Gilbert, 1980). The response to preload was investigated in fetal sheep for the left and right ventricles. For either ventricle, isolated output rose only slightly with increases in filling pressures. Using a microsphere technique in fetal sheep, alterations in combined cardiac output were measured as blood volume was modulated (Gilbert, 1980). With a decrease of 10% in circulating volume, there was a significant drop in right atrial pressure, as well as in cardiac output. In contrast, cardiac output did not significantly change despite a significant change in right atrial pressure with a 10% increase in circulating blood volume from baseline. These data lead the author to conclude that the fetus operates at the upper end of the Starling curve and possesses limited cardiac reserve.

With only half the amount of sarcomere volume that is in an adult, connective tissue comprises a much greater percentage of the newborn heart. The myofibrils are fewer in number with a disorderly arrangement in the myocyte. There is much more stiff connective tissue. The sum total of these changes makes it plausible that the relatively noncompliant newborn heart does function at the upper end of its Starling curve, where the response to volume loading is blunted. In fact, the original experiments failed to account for an inevitable consequence of increases in preload. An increase in the preload leads to greater stroke volume, and if heart rate is unchanged, cardiac output must increase as well. Blood pressure must now rise unless there is a corresponding drop in SVR. Therefore, an increase in preload leads to an increase in afterload, which then acts to reduce stroke volume. This created the impression that the newborn heart could not increase stroke volume in response to volume loading. Other work, controlling for arterial pressure, shows that the newborn heart is indeed responsive to volume within the limitations of its decreased compliance (Fig. 4-13) (Kirkpatrick et al.; 1976, Hawkins et al., 1989). This has been shown in the laboratory with a sheep model and though echocardiography in the human fetus. The newborn can be likened to the hypertensive adult with diastolic dysfunction. This is not to suggest that the otherwise healthy newborn has diastolic dysfunction, because the newborn heart is not dysfunctional but normal for its stage of development. Rather, the lesson of diastolic dysfunction is that the ventricle is preload dependent. At the low end of the Starling curve, reduced preload is poorly tolerated. The upper end of the Starling curve is flattened, reflecting its reduced compliance. Between these two points is the steepest part of the curve, where increases in LVEDV result in significantly greater stroke volume. Clinical experience observing the response to volume infusion confirms this property of the newborn heart.

FIGURE 4-13 The response of the immature heart to changes in preload and afterload. When afterload is held constant, left ventricular stroke volume (LVSV) increases linearly with increased left atrial (LA) pressure. Increasing afterload creates an inverse linear relationship with LVSV when LA pressure is constant.

(From Hawkins J, Van Hare GF, Schmid KG, et al.: Effects of increasing afterload on left ventricular output in fetal lambs, Circ Res 65:127, 1989.)

Afterload

In addition to preload, there is another loading force that influences myocardial performance. The force that resists the ejection of blood is known as afterload. In isolated muscle strip experiments, the afterload is represented by the weight against which the muscle contracts or shortens. The stretch applied to the isolated muscle strip before contraction is fixed, allowing the experiment to be conducted at a constant length, or preload. Plotting the velocity of myocardial shortening against progressively increasing afterload reveals an inverse relationship. The shape of the curve is exponential with the points of maximal and zero shortening both occurring at physiologically impossible limits. The greatest shortening velocity occurs at zero load. When the load is so great that no muscle fiber shortening can occur, isometric contraction occurs. Between these two extremes, it is intuitive that increasing afterload results in reduced velocity of muscle fiber shortening. Transferring this straightforward laboratory concept to the intact circulation is difficult (Martin et al., 2006). The force generated by the shortening of an isolated muscle strip is directed in only one plane. This is very different from the three-dimensional contraction of the left ventricle with its asymmetric geometry. Additionally, in the laboratory the force that resists the shortening of the muscle is the weight applied to the muscle strip. Physiologically, the force that resists the ejection of blood is a complex interplay between blood pressure, vascular impedance, walls stress, and inertia. Finally, blood pressure and SVR are commonly used measures of afterload, but neither is fully accurate.

In most situations, the systemic blood pressure does provide an appropriate surrogate for afterload. One common exception is in conditions in which there is obstruction to left ventricular outflow. In adults, the most common cause is aortic stenosis, but pediatric patients may also have congenital lesions that cause obstruction at the subvalvular level in the left ventricular outflow tract (LVOT) in addition to aortic valve disease. The left ventricle must generate enough pressure to overcome both the systolic blood pressure and the gradient across the valve or the LVOT obstruction. Assuming that the ventricle must only overcome the systolic blood pressure leads to a significant under estimation of afterload. The use of SVR to approximate afterload is also problematic. The concept of SVR describes resistance, which is defined as the pressure drop across a system divided by the flow across that same system. This is reflected in the equation for SVR:

Multiplication by 80 converts SVR from mm Hg/L/min/m into SI units (dynes•s•cm^–5). The calculated SVR may be similar for patients with very different cardiovascular performance. Consider these two patients:

These two hypothetical patients are very different, yet the SVR is similar. The first patient has normal cardiovascular function and the second patient’s cardiovascular function is compromised. A reduction in contractility has caused a fall in cardiac output despite the attempts of the ventricle to compensate with a higher diastolic filling pressure. This example demonstrates how SVR in isolation does not necessarily reflect large changes in loading conditions and cardiac output. The modified Law of Laplace describes a better approximation of afterload (see Preload section p. 91). As discussed in the section on preload, wall stress was demonstrated to be a key component of myocardial oxygen demand. During left ventricular ejection, the value for pressure is the systolic blood pressure. By including the radius and accounting for the compensatory mechanism of left ventricular hypertrophy the modified Law of Laplace provides a more comprehensive view of the forces that resist ventricular ejection. Importantly, it demonstrates the concept that the dilated ventricle with an increased radius must eject blood under increased afterload conditions. Given the inability to easily measure ventricular radius and thickness, the modified Law of Laplace is not a clinical parameter to guide management. It is presented here to help the reader conceptualize afterload. It still cannot describe all the variables that account for afterload in the intact cardiovascular system.

Studies of the intact heart that assess the effect of changes in afterload on cardiac output are difficult to interpret without controlling for the other variables. The most important additional variable is preload. Preload and afterload are linked. For instance, pharmacologically increasing afterload decreases stroke volume. As the heart ejects less blood with each contraction, the end-diastolic volume must increase if venous retum is held constant. With an increase in LVEDV the Starling curve dictates an enhanced ventricular force of contraction, which acts to return stroke volume to its previous level. The general inverse relationship between afterload and stroke volume holds across the spectrum from fetal life to adulthood (Friedman, 1972). The question is whether the immature myocardium is excessively sensitive to increases in afterload. In the laboratory, afterload can be assessed while strictly controlling for preload. These studies have revealed that the fetal myocardium is indeed more sensitive to increases in afterload than the adult myocardium (Friedman, 1972). Animal experiments with an intact circulation also show a decrease in ventricular output under conditions of increasing afterload (Fig. 4-13) (Gilbert, 1982; Thornburg and Morton, 1986). Van Hare and colleagues (1990) used a balloon occluder in the aortas of fetal sheep to modulate afterload while keeping preload constant. Combined ventricular output was measured by transducing aortic flow. They noted that at physiologic mean atrial pressures, there was an inverse linear relationship between mean arterial pressure and stroke volume. This exists for both right and left ventricles. Despite this, a fetus normally makes a smooth transition to postnatal life when the low-resistance placenta is removed. At birth, pulmonary blood flow dramatically increases, raising left-ventricular preload. Stroke volume increases, which leads to an increase in systemic blood pressure. Despite the rise in blood pressure and the loss of the low-resistance placenta, cardiac performance in the newborn is not compromised. The explanation is that much of the rise in afterload is secondary to increased preload. In clinical practice with newborns and young children, it is rare to encounter conditions of purely increased afterload. A more likely situation is significant hypovolemia that results in hypotension. Despite the low afterload state, overall cardiac performance is impaired, because the decrease in afterload is caused by decreased preload. The correct therapy is to restore intravascular volume. Blood pressure improves, and the increased afterload is well tolerated. This example illustrates the interrelationship between these two variables of cardiac output.

It has been shown that when preload is held relatively constant, cardiac output moves inversely within the physiologic range of afterload. The relevance of afterload in the newborn stage might be questioned, because systemic hypertension is so rare. The right ventricle develops in an environment of raised PVR, yet after birth it is more sensitive to increases in afterload. Although there is limited ability of the systemic circulation to become hypertensive, the pulmonary circulation, under appropriately provocative conditions, can revert to its fetal state. Postnatally, the right ventricle begins to remodel in response to decreasing pulmonary artery pressures. Sustained elevations in pulmonary artery pressures decrease right ventricular output and reduce left-ventricular preload (Thornburg and Morton, 1983). Because both ventricles share the septum, the strain on the right ventricle impairs left-ventricular filling and contraction (Rein et al., 1987). Reductions in left-sided preload and contractile force demonstrate how the inability of the right ventricle to cope with increased afterload can lead to decreases in systemic cardiac output. In the newborn stage, the potential for pulmonary hypertension far exceeds that of systemic hypertension. This is the most common scenario for a significant rise in afterload while preload remains relatively unchanged. The strain on the right side of the heart and ventricular interdependence may lead to biventricular failure.

Contractility

The third parameter of myocardial performance is contractility. Seemingly intuitive, contractility is often confused with changes in stroke volume brought about by alterations in loading conditions. Contractility, synonymous with inotropy, is the intrinsic ability of the myocardium to contract when loading conditions are held constant. The Starling curve is the first source of confusion. As discussed in the section on preload, every point on an individual patient’s Starling curve represents the same inotropic state. Movement along the curve is solely because of changing conditions of preload. In the same way, the demonstration in the laboratory of decreased myocardial shortening velocity in response to increased weight (afterload) is not evidence of a fall in contractility. Contractility has remained constant, but the force of contraction is reduced. Contractility and force of contraction are clearly not the same. The force of contraction, related to loading conditions, can vary widely. Contractility is relatively fixed. It can be increased pharmacologically but is usually either normal or reduced. Further demonstrating misconceptions about contractility, a hypertrophied ventricle is often assumed to have increased contractility. Hypertrophy is a response to increased afterload, but contractility, once corrected for the greater cross-sectional area of the ventricular muscle mass, is normal. Any disease process that injures the myocardium decreases global contractility. In adults, this most commonly occurs secondary to ischemia or infarction. In addition to these causes, pediatric patients may suffer decreased contractility on the basis of genetic disorders, infiltrative diseases, infections, or nutritional deficiencies.

The immature myocardium has a reduced sarcomere concentration combined with a transport system for calcium that is not fully developed. The myofibrillar arrangement is also more disorganized in the infant heart when compared with the adult heart. Mitochondria, the energy powerhouse of the myocyte, are reduced in number in the immature heart (Barth et al., 1992). Based on the above differences, a reduction in contractility is expected. Contractility is measured by the tension developed in the isolated muscle strip. This expected reduction in contractility of the immature myocardium is confirmed in laboratory experiments where preload and afterload are controlled. Across the entire range of loading conditions, fetal cardiac muscle generates less tension than adult myocardium (Fig. 4-14) (Friedman, 1972; Romero et al., 1972). In an intact heart it is not possible to fully separate the influences of different loading conditions. Nevertheless, fractional area shortening of the ventricle measured echocardiographically is used as a measure of contractility. Decreased fractional area shortening has been observed in the fetal heart (St. John Sutton et al., 1984). Effecting a true increase in contractility requires either a greater number of contractile elements or more vigorous action by the contractile elements already present. Increasing the number of contractile elements (hyperplasia and hypertrophy) occurs as a normal part of development. Improving the action of the contractile elements already present requires an understanding of the singular role of calcium. Therapy to improve inotropy exerts its action at the myocyte level by affecting the levels of intracellular calcium.

FIGURE 4-14 Length-tension relationships in adult sheep compared with fetal lambs. The fetal myocardium has greater resting tension but is able to generate less active tension. This represents a state of decreased compliance or a “stiffer” ventricle combined with decreased contractility.

(From Friedman WF: The intrinsic properties of the developing heart, Prog Cardiovasc Dis 15:87, 1972.)

Heart Rate

As mentioned in the section on preload, it has long been considered that the newborn’s heart rate is dependent on and relatively insensitive to changes in end-diastolic volume. Assuming the maintenance of sinus rhythm, heart rate is believed to determine cardiac performance through its influence on preload and myocardial oxygen supply and demand. At very high heart rates, hypotension ensues, because diastolic filling time is severely restricted and preload markedly falls. Coronary perfusion to the left ventricle occurs during diastole, and systolic ejection time is fixed. Thus, tachycardia shortens the time for perfusion of the left ventricle and results in an increased oxygen demand combined with a decreased oxygen supply. The absence of coronary artery disease in newborns offers them some protection in accepting imbalances in myocardial oxygen supply and demand. However, extremes of heart rate are poorly tolerated by adults and newborns alike.

The question is whether the newborn who has a heart rate within normal range is fundamentally different than the adult. When corrected for weight, stroke volume is similar across all ages. The high cardiac output of newborns and infants can only be achieved with a heart rate that is significantly higher than in adults. This has created the idea that the newborn is dependent on heart rate. Here it becomes difficult to sort out the isolated effect of heart rate from its effects on loading conditions and contractility. It is not intuitive as to why changes in heart rate alter contractility. The mechanism is known as the force-frequency relationship. Experiments using atrial pacing, while controlling loading conditions, demonstrate an increase in stroke volume with an increase in heart rate (Anderson et al., 1982).With constant loading conditions, the only explanation for increased stroke volume is that contractility has increased. Thus, an increased heart rate improves contractility. The basis for the force-frequency relationship is that an increase in heart rate is accompanied by enhanced release of intracellular calcium (Parilak et al., 2009). There is a suggestion that the force-frequency relationship has minimal effect in newborns but is present during infancy (Wiegerinck et al., 2008). Spontaneous increases in heart rate also improve cardiac output. In this scenario, the increase in heart rate is the result of a neurohumoral stimulus with an effect that cannot be isolated to producing tachycardia. Both preload and contractility can be expected to increase, as long as the increase in heart rate does not lead to deleterious changes in preload or myocardial oxygen supply and demand. The combination of these factors allows the newborn and infant to use heart rate to significantly augment cardiac output.

Integrating Preload, Afterload, and Contractility

The determinants of myocardial performance, while discussed in isolation, are actually intricately linked. Diagrammatically, the determinants of myocardial performance can be represented by ventricular pressure-volume loops (Suga et al., 1973). The loop shows the pressure and ventricular volume changes that occur during one cardiac cycle. Increases in preload while afterload is held constant result in greater stroke volume (Fig. 4-15). This simple curve does not provide any information about contractility. The end systolic pressure-volume relationship (ESPVR) is a family of curves that is generated by rapidly altering preload or afterload. The curves create a series of points that are connected to become the ESPVR line. The slope of this line represents contractility (Fig. 4-16). The pressure-volume loop illustrates the concept that movement along the ESPVR line reflects changes in loading conditions while contractility remains constant. In Figure 4-16, the increases in afterload result in a series of points that are connected to become the ESPVR line. The slope of the ESPVR line represents contractility. The two figures of ventricular pressure-volume loops represent an idealized situation where preload and afterload can be manipulated independent of each other. In the intact organism, preload and afterload are linked. The interdependence of preload and afterload is shown in a new ventricular pressure-volume loop (Fig. 4-17).

FIGURE 4-15 Ventricular pressure-volume loop. As preload increases from point A to points E and F, the end-diastolic pressure-volume curve represents compliance. The increased preload results in greater stroke volume (SV) when afterload is held constant.

(Modified from Braunwald E, Ross J, Sonnenblick EH: Mechanisms of contraction of the normal and failing heart, Boston, 1976, Little, Brown.)

FIGURE 4-16 Ventricular pressure-volume loop. As afterload is increased, stroke volume falls when preload is held constant. The slope of the line connecting points D, E, and F is the end-systolic pressure-volume relationship and represents contractility. Only alterations in the slope of the line represent changes in contractility. The contractility at points D, E, and F is identical.

(Modified from Braunwald E, Ross J, Sonnenblick EH: Mechanisms of contraction of the normal and failing heart, Boston, 1976, Little, Brown.)

FIGURE 4-17 Ventricular pressure-volume loop. The interdependence of preload and afterload is demonstrated. The increase in preload from point 1 to point 2 results in a rise in stroke volume (SV₁ to SV₂) that is less than anticipated because the increase in preload causes a corresponding increase in afterload.

(Modified from Strobeck JE, Sonnenblick EH: Myocardial contractile properties and ventricular performance. In Fozzard HA, Haber E, Jennings RB, et al., editors: The heart and cardiovascular system, New York, 1986, Raven.)

In states of decreased contractility, the ventricular pressure-volume loop displays the limitations of the failing heart. Analysis of the ventricular pressure-volume loop shows that the failing ventricle is sensitive to changes in loading conditions (Fig. 4-18). The ESPVR line has shifted down and to the right. With preload held constant, stroke volume is reduced. The body’s response to this situation is to increase LVEDV in an attempt to preserve stroke volume. Stroke volume has improved at the expense of a higher end-diastolic pressure. The ventricular pressure-volume loop demonstrates why afterload reduction is the cornerstone of medical therapy for the failing heart. Any increases in afterload come at the expense of stroke volume, with a limited ability for compensatory changes in preload.

FIGURE 4-18 Ventricular-pressure volume loop with decreased contractility. The line representing contractility is shifted down and to the right. Curve 1 is the stroke volume with normal contractility. Curve 2 shows that when contractility is reduced, stroke volume can be maintained through increased preload. Curve 3 demonstrates a significant reduction in stroke volume when afterload is increased in the presence of decreased contractility.

(Modified from Braunwald E, Ross J, Sonnenblick EH: Mechanisms of contraction of the normal and failing heart, Boston, 1976, Little, Brown.)

The reflex of the trainee when confronted with a patient with significantly decreased contractility is to withhold or restrict fluids in a well-meaning attempt to avoid pulmonary edema. The ventricular pressure-volume loop for the patient with heart failure shows the fallacy and danger of this approach. Positive pressure ventilation is associated with decreases in preload. Venous tone is decreased by most anesthetic agents. The combination of these two events means that while the patient’s global volume status is unchanged, he or she has become relatively hypovolemic. He or she is below the optimum LVEDV for his or her failing heart. Far from being on the verge of pulmonary edema, this patient now requires intravascular volume to compensate for the effects of anesthesia and positive pressure ventilation. Although maintaining LVEDV is critical for patients, there are currently limited means to accurately assess preload. Central venous pressure must be interpreted in light of the altered ventricular compliance of the failing heart. Lastly, while the failing heart requires adequate preload, its ability to accept volume rapidly is limited. Cautious infusion of volume with frequent reassessment is necessary lest a rapid bolus of fluid results in pulmonary edema, which is exactly the outcome the trainee sought to avoid by restricting volume in the first place.

Subcellular Structures