Pre-operative anatomy

In congenital cardiac malformations, anatomy dictates physiology. Various anatomic lesions can produce a functionally univentricular circulation (Table 1). By definition, newborns with this physiology have mixing of the pulmonary and systemic venous returns, but the relative amounts of each vary substantially depending on the underlying anatomy. The most important anatomic issues affecting planned interventions and management of intensive care are the outflows of blood to and from the functionally single ventricle, and the respective pulmonary and systemic circulations. The newborn may have either pulmonary or aortic obstruction, or bilaterally unobstructed outflows. Either the systemic or pulmonary venous return may also be abnormally connected or obstructed in the newborn with a functionally univentricular circulation.

Table 1. Anatomic diagnoses commonly associated with functionally univentricular physiology in the newborn. Very few malformations have an anatomically single ventricle.

Post-operative anatomy

The goal of all initial palliative procedures for the newborn with functionally univentricular anatomy is to establish unobstructed pulmonary and systemic venous return, unobstructed systemic outflow, and to limit the flow of blood to the lungs with normal pulmonary arterial pressures. This is typically accomplished through the first stage Norwood procedure, construction of a modified Blalock-Taussig shunt, or banding of the pulmonary trunk. Although variations on these palliative procedures exist, they represent the spectrum of post-operative anatomy encountered in cardiac intensive care. Since each anatomic arrangement establishes similar physiology, the important differences are the means by which each operation accomplishes the ultimate goal defined above. The Norwood procedure requires cardiopulmonary bypass, cardioplegia, and a period of deep hypothermic circulatory arrest, although newer techniques can limit the period of circulatory arrest.^4–8 This maneuver subjects the tissues to a period of ischemia with subsequent reperfusion, which is often followed by transient post-operative dysfunction, especially in the heart, lungs, kidneys, and vascular endothelium.^9–11 A Blalock-Taussig shunt, either alone or as part of the first stage Norwood procedure, can lower diastolic arterial pressure, which may compromise coronary arterial perfusion. Unlike the Blalock-Taussig shunt, a band on the pulmonary trunk is not associated with diastolic run-off to the lungs, but some investigators have suggested that banding may increase the risk of subaortic obstruction and ventricular hypertrophy,¹² although this has been disputed.¹³ Shunting and banding procedures both carry the risk of distorting the pulmonary arteries, or producing unilateral pulmonary arterial obstruction, which can contribute to cyanosis after either of these procedures.

Sano and others have proposed a modification of the Norwood procedure which entails supplying flow to the lungs directly from the ventricle by insertion of a Goretex conduit from the morphologically right ventricle to the bifurcation of the pulmonary tank, rather than constructing a conventional Blalock-Taussig shunt. Theoretical benefits of this modification include stabilization of the ratio of pulmonary-to-systemic flows and elimination of the aortic diastolic run-off to the lungs, which may improve coronary arterial perfusion. Proponents report excellent results, and assert that postoperative management following this modified procedure is less complicated than after the conventional first stage Norwood palliation. On the other hand, placement of such a conduit requires a ventriculotomy, and introduces an additional volume load on the already volume-loaded funtionally single ventricle, which could increase the risk for ventricular dysfunction or arrhythmias over the long-term.

Functionally univentricular physiology

The balance of pulmonary and systemic flows

Regardless of underlying anatomy, univentricular physiology entails mixing of systemic and pulmonary venous return, and partitioning the total cardiac output to pulmonary and systemic flows based on the amount of anatomic obstruction and the vascular resistance to flow in the respective circulations. It is generally accepted that the arterial oxygen saturation reflects the ratio of flows in the unoperated or shunted newborn with a functionally single ventricle. This is based on a simplification of the Fick principle, which involves two important assumptions. The Fick equations for calculation of pulmonary and systemic flows are:

where VO₂ represents oxygen consumption, CaO₂ is the arterial oxygen content, CmvO₂ is the mixed venous oxygen content, CpvO₂ is the pulmonary venous oxygen content, and CpaO₂ is the pulmonary arterial oxygen content. Since systemic and pulmonary arterial oxygen saturations are identical in functionally univentricular physiology, the equation for the ratio of pulmonary to systemic blood flows (Qp:Qs) becomes:

where SmvO₂ is the oxygen saturation of mixed venous blood, SaO₂ is the oxygen saturation of arterial blood, and SpvO₂ is the oxygen saturation of pulmonary venous blood. Since the lungs in most infants with congenital cardiac malformations are relatively healthy, one can assume that the saturation of oxygen in the pulmonary venous blood is normal, at about 95% in room air. Furthermore, if one also assumes the difference between the systemic arterial and venous oxygen saturation is normal, at approximately 25%, the simplified equation for the ratio of pulmonary to systemic blood flow becomes:

This simplified version of the Fick equation allows estimation of the pulmonary to systemic blood flow ratio based on systemic oxygen saturation alone. Given the ease with which this latter value can be measured in clinical practice, the simplified equation has clear advantages. Specifically, it allows the clinician to estimate the ratio of pulmonary to systemic flows, and systemic delivery of oxygen, simply by looking at the systemic arterial saturation. Theoretically, this allows one to assess the effects of interventions designed to alter flows by observing changes in arterial oxygen saturation. It is important to remember, however, that such estimation of the ratio of pulmonary to systemic blood flows is based on assumptions regarding oxygen saturation of mixed venous and pulmonary venous blood. The assumption regarding the difference in systemic arterial-venous oxygen saturation is only accurate if systemic perfusion is normal. In low output states, which are common in pre- and post-operative neonates with functionally univentricular physiology, the oxygen saturation of mixed venous blood can be quite low, so that arterio-venous oxygen saturation differences will be substantially greater than 25%. If the decrease in systemic flow and mixed venous oxygen saturation is offset by an increase in the amount of well-saturated blood returning from the lungs, the arterial oxygen saturation may remain unchanged, or even increase. Clinical data suggest that the oxygen saturation of mixed venous blood varies considerably in patients with functionally univentricular physiology.^14–21 Many centers have begun to monitor this value routinely following Norwood palliation for hypoplastic left heart syndrome, using a sample from the superior caval vein as representative of mixed venous blood, since there is no true site of systemic mixed-venous blood in the functionally univentricular circulation.^{1, 14–19} Adapted from Tweddell et al.,¹⁸ Figure 1 shows simultaneous mean arterial blood pressure and the oxygen saturations of arterial and mixed venous blood in a patient early after Norwood palliation. The arterial oxygen saturation remains relatively constant, but mixed venous oxygen saturation plummets at about 17:00, accompanied by a slight increase in blood pressure. The only plausible explanation for this event is that systemic vascular resistance and the ratio of pulmonary to systemic flow increase dramatically. The decrease in mixed venous oxygen saturation and systemic flow is offset by a larger amount of well-saturated blood returning to the heart from the lungs, resulting from the increase in the ratio of flows. In this situation, estimation of the ratio of flows from arterial oxygen saturation alone would lead the clinician to the erroneous conclusion that the circulations are well balanced, when in reality systemic flow is critically low. As predicted by the assumptions inherent in equation 4, clinical data indicates that the difference in arterio-venous oxygen saturation averages 25 to 30% in postoperative Norwood patients,^{19, 20} but the range of values varies considerably, limiting the applicability of equation 4 for any particular patient. Although the saturation of oxygen in pulmonary venous blood is likely to be normal in the absence of clinical or X-ray evidence of pulmonary parenchymal disease, there are conditions under which this assumption is incorrect. Using catheters placed in the left lower pulmonary vein and superior caval vein in infants undergoing Norwood palliation, Taeed and colleagues¹⁹ found unexpected pulmonary venous desaturation is common early after Norwood palliation, particularly with fractional inspired oxygen levels less than 0.3 (Fig. 2). Failure to account for pulmonary venous desaturation results in a falsely low estimate of the ratio of pulmonary to systemic flow, using equation 4. As shown in Figure 3, small errors in such assumptions can result in gross inaccuracy in the estimated ratio.²² The clinical relevance of this principle is that maneuvers that decrease pulmonary venous oxygen saturation, rather than the ratio of flows, result in lower arterial oxygen saturation and reduced systemic oxygen delivery, because systemic flow remains unchanged.

Figure 1. Arterial oxygen saturation (SaO2), mean arterial blood pressure (MAP), and mixed venous oxygen saturation (SmvO2) over time in a patient following Norwood palliation. The mixed venous saturation falls precipitously while pressures and systemic saturation do not change significantly. This can be explained by a rise in systemic vascular resistance, associated with an increase in pulmonary blood flow and a decrease in systemic blood flow. (Used with permission from Tweddell et al.64)

Figure 2. Pulmonary venous oxyhemoglobin saturation (SpvO2) plotted as a function of fractional inspired oxygen (FiO2), demonstrates that pulmonary venous desaturation occurs frequently following the Norwood operation. Pulmonary venous saturations of less than (points below the dashed line) most often occur at fractional inspirations of less than 0.21, but can also occur at values as high as 0.40. (Used with permission from Taeed et al.62)

Figure 3. The proportional error in the ratio of pulmonary to systemic blood flows as a function of pulmonary venous oxyhemoglobin saturation (SpvO2). Arterial oxyhemoglobin saturation (SaO2) is assumed to be 77%, and mixed venous oxyhemoglobin saturation is assumed to be 45%. The relationship is plotted for both a 5% and 10% error in pulmonary venous saturations. (Used with permission from Barnea et al.6)

The importance of accurately estimating the ratio of pulmonary to systemic flow is depicted in Figure 4, which demonstrates the relationships between the ratio of flows, systemic oxygen delivery, and total cardiac output.²³ Using mathematical modeling with the oxygen saturation of pulmonary venous blood constant at 96%, Barnea et al.²³ generated a series of curves showing systemic delivery of oxygen as a function of the ratio of pulmonary to systemic flow. Since the total cardiac output of the functionally single ventricle is the combined flows, unless total cardiac output increases, an increase in pulmonary flow is accompanied by a decrease in systemic flow, and vice versa. The dashed line in Figure 4 shows that maximum oxygen delivery occurs at ratios between 0.5 and 1, depending upon total cardiac output. The slope of each isobar for a given cardiac output is steepest on either side of maximal systemic oxygen delivery, suggesting that small changes in the ratio of flows can be associated with large changes in oxygen delivery. The key point demonstrated in Figure 4, however, is that systemic oxygen delivery is increased much more effectively by augmenting total cardiac output than by altering the ratio of flows. One limitation to this type of model is the use of systemic oxygen saturation and systemic flow as interchangeable components of systemic oxygen delivery. Although newborns tolerate cyanosis well, the oxyhemoglobin dissociation curve dictates that, once arterial oxygen saturation becomes critically low, further decreases can no longer be compensated for by increases in systemic flow.²⁴ This is validated in clinical practice by the fact that infants with functionally univentricular physiology may develop acidosis, and become quite unstable as the systemic saturation of oxygen falls below a certain level, usually between 60 and 65%, presumably due to myocardial ischemia. Thus, although optimization of total cardiac output is the best way to assure adequate systemic oxygen delivery, manipulation of the ratio of pulmonary to systemic flow is still very important for improvement of marginal oxygen delivery.

Figure 4. Systemic oxygen availability (mlO2/min) as a function of the ratio of pulmonary to systemic blood flows (Qp/Qs)at different levels of cardiac output. Pulmonary venous oxyhemoglobin saturation (SpvO2) is assumed to be 96%, and oxygen consumption (VO2) is assumed to be 18 mlO2/min. (Used with permission from Barnea et al.5)

Optimizing systemic oxygen delivery in neonates with functionally univentricular physiology

The goal of management in this setting is not to maximize the systemic arterial saturation, but to ensure adequate delivery of oxygen to all tissues. Optimization of systemic oxygen delivery requires maintenance of cardiac inotropy while balancing flows, and maintaining adequate blood pressure and saturations of oxygen. Although systemic delivery of oxygen is best augmented by optimizing total cardiac output, inherent inter-relationships between the various hemodynamic variables can make this task challenging in the patient with a functionally single ventricle. Furthermore, blood pressure and arterial oxygen saturation may not reflect moment-to-moment alterations in underlying physiology, and may remain constant over a wide range of varies of flows.^{18, 19} Although management of this situation in neonates has traditionally focused on manipulation of flows via changes in pulmonary vascular resistance, newer data suggest management of total cardiac output and systemic vascular resistance may be more effective.^14–20 Defects in pulmonary gas exchange resulting in unrecognized pulmonary venous desaturation have also been shown to compromise systemic oxygen delivery.¹⁹ In all instances, maintenance of oxygen-carrying capacity by keeping hemoglobin in the range of 13 to 16 mg/dl is beneficial. In patients with intracardiac shunting, an increased concentration of hemoglobin increases mixed venous and arterial oxygen saturations, and decreases the ratio of pulmonary to systemic flow, by increasing the ratio of pulmonary to systemic resistances.^{28, 29}

Optimizing systemic oxygen delivery by manipulation of pulmonary vascular resistance

Differential manipulation of the resistances in patients with functionally single ventricles has traditionally emphasized the use of oxygen, carbon dioxide, and acid–base status (Table 2).^{30, 31} In infants with unrestricted pulmonary flow, subatmospheric levels of oxygen, or induction of respiratory acidosis, can effectively raise pulmonary resistance and decrease systemic resistance, and thus decrease the ratio between them.³² Subatmospheric oxygen should be used with caution since it may cause pulmonary venous desaturation and reduce systemic oxygen delivery, especially in the postoperative patient.¹⁹ Respiratory acidosis can be induced either through alveolar hypoventilation, or by altering inspired concentrations of carbon dioxide in mechanically ventilated and paralyzed patients. Recent data from both pre-^{32, 33} and post-operative³⁴ newborns with hypoplastic left heart syndrome suggest that inhaled carbon dioxide may be more effective than subatmospheric oxygen in increasing cerebral and/or systemic perfusion in ventilated and paralyzed patients, although these studies did not differentiate whether inspired carbon dioxide improves systemic or cerebral delivery of oxygen, or both. It is important to recognize that these studies did not assess the effect of inspired carbon dioxide in spontaneously ventilated patients. Although both inspired carbon dioxide and subatmospheric oxygen are often used to balance the ratio of flows in spontaneously breathing patients, data on this are lacking.

Table 2. Effects of respiratory maneuvers on pulmonary and systemic vascular resistance.

In infants with low pulmonary vascular resistance and anatomically restricted pulmonary blood flow, it is less clear that manipulation of pulmonary resistance is useful to alter the ratio of flows. Since all palliative procedures are aimed to restrict blood flow to the lungs by mechanical obstruction, it is likely that pulmonary flow is primarily limited by the size of the systemic-to-pulmonary arterial shunt or the tightness of the pulmonary arterial band, so that further decreases in downstream resistance are of minimal consequence. Taeed et al.¹⁹ demonstrated no significant changes in the ratio of flows with subatmospheric oxygen following Norwood palliation. Although the arterial-venous difference in oxygen increased with inspired carbon dioxide in postoperative Norwood patients³⁴, these alterations could also result from direct effects of arterial carbon dioxide and pH on systemic vascular resistance.

Increases in pulmonary resistance can also be induced by increased positive end-expiratory pressure.³¹ When lung compliance is normal, positive airway pressure increases resistance by compressing inter-alveolar pulmonary arterioles. Since the nadir of pulmonary vascular resistance occurs at functional residual capacity rather than zero pressure, the level of pressure must result in an end-expiratory lung volume greater than functional residual capacity. Initial application of positive pressure results in pulmonary recruitment, and aids vascular recruitment through radial traction forces applied to the pulmonary vasculature, which therefore actually decreases resistance. Further increases in pressure above functional residual capacity results in compression of pulmonary vessels, and thus increases resistance. Increased levels of pressure should be used with caution, however, as this may reduce cardiac output by impeding venous return. Optimal lung recruitment with positive pressure may also prevent pulmonary venous desaturation by optimizing pulmonary gas exchange.¹⁹

Not all patients with functionally single ventricles demonstrate pulmonary overcirculation. Elevated pulmonary resistance can easily persist in the newborn with such physiology, and can cause severe cyanosis. When pulmonary flow is very low, with saturations less than 30 Torr, it can effectively increase the pulmonary dead space and impair alveolar ventilation. The occurrence of respiratory acidosis in this setting is of grave concern, because this will further increase pulmonary vascular resistance, limiting the ability to hyperventilate or alkalinize the patient. Treatment of high pulmonary resistance in such patients is much the same as in any other population. Strategies of alveolar recruitment are appropriate when there is atelectasis or pulmonary disease, but airway pressures should be the minimum necessary to maintain the functional residual capacity. High-frequency oscillatory or jet ventilation may be effective to attain hyperventilation at low mean airway pressure,³⁵ a desirable combination in pulmonary hypertension. Use of supplemental inspired oxygen, hyperventilation, and alkalosis may be beneficial. Inhaled nitric oxide and infusion of prostaglandin have also been used in these patients to selectively lower pulmonary vascular resistance.^{36, 37} Raising systemic blood pressure by vasoconstriction may increase pulmonary flow, and thus increase systemic arterial oxygen saturation.

Optimizing systemic oxygen delivery by manipulation of systemic resistance

Another approach to differential regulation of the vascular resistances is pharmacological manipulation of systemic resistance. Although intravenous vasodilators tend to have similar effects on the pulmonary and systemic vasculature, in the setting of poor systemic perfusion with high systemic and low pulmonary resistance, they will have a greater relative effect on the systemic vasculature. Nitroprusside, phenoxybenzamine, amrinone, and milrinone have been advocated to reduce systemic afterload, and to block the α-adrenergic receptor mediated vasoconstriction that occurs with drugs such as epinephrine and norepinephrine. Phenoxybenzamine lowers systemic resistance, decreases the ratio of pulmonary to systemic flow, and improves systemic delivery of oxygen after the Norwood operation, even though it is associated with reductions in systemic blood pressure¹⁸. Beneficial effects reported with use of phenoxybenzamine may also result from increased use of inotropic agents in these patients.^{18, 20} β-adrenergic stimulation of the myocardium, in conjunction with vasodilation, can increase total cardiac output without associated vasoconstriction. Other vasodilating agents can be used to accomplish this same goal, although they involve different receptor mechanisms and cellular pathways. Systemic vasodilation may be particularly valuable after operations that require deep hypothermic circulatory arrest, which are often associated with systemic vasoconstriction, such as the Norwood operation. Both systemic and pulmonary endothelial dysfunction have been observed following cardiopulmonary bypass with or without circulatory arrest, presumably resulting from ischemia-reperfusion injury to the endothelium.^{9, 11, 38–40} Systemic vasoconstriction in this setting as a response to low systemic output is a non-adaptive response. Inappropriate increases in systemic resistance increase pulmonary flow at the expense of systemic flow while maintaining blood pressure and systemic arterial saturation, thus masking potential warning signs of low systemic output. It is important to recognize that data supporting manipulation of systemic vascular resistance to balance flows in neonates with functionally single ventricles is limited to postoperative patients. The utility of manipulating systemic resistance in newborns with unrestricted pulmonary flow has not been studied.

Optimizing total cardiac output

Increasing total cardiac output is among the most effective ways to optimize systemic oxygen delivery. Increasing systemic flow by inotropic support or afterload reduction may also increase systemic arterial saturation simply by increasing the mixed venous saturation. Inotropic agents can improve ventricular contractile function and augment stroke volume for a given preload, but it is important to recognize the systemic vasoconstriction associated with certain inotropic agents. Alterations in vascular tone associated with particular inotropic agents may cause alterations in flows in addition to increases in total cardiac output. In a porcine model of functionally univentricular physiology, Riordan et al.⁴¹ studied the effects of epinephrine, dobutamine and dopamine on pulmonary and systemic flows. They reported that the ratio of pulmonary to systemic flow was increased by dobutamine, decreased by low-dose epinephrine, and not altered significantly by dopamine. While it is likely that these observations are not fully applicable to the clinical situation since human neonates are more likely to show systemic vasoconstriction to epinephrine or high-dose dopamine, these data illustrate the importance of using vasodilator drugs to accompany inotropic agents with prominent vasoconstrictor properties. Furthermore, since the neonatal heart is particularly sensitive to elevated afterload, vasodilator drugs may also augment total cardiac output through afterload reduction. When systemic vasodilators are used in patients with functionally univentricular physiology, it is important to recognize that cardiac output and systemic arterial pressure are particularly sensitive to a reduction in preload, so the clinician must be certain that adequate preload is maintained. Figure 4 demonstrates the importance of combining inotropic and vasodilator agents in patients with univentricular physiology. Systemic oxygen delivery can be improved dramatically by increasing total cardiac output, and further optimized by adjusting the ratio flows, thus the combination of inotropic support and decreasing systemic resistance is often the optimal strategy to maximize systemic oxygen delivery.

Optimizing pulmonary recruitment to avoid pulmonary venous desaturation

The data of Taeed and colleagues¹⁹ suggest that postoperative Norwood patients may have abnormalities in pulmonary gas exchange resulting from cardiopulmonary bypass, circulatory arrest, and the extensive surgical procedure. Ventilatory maneuvers used to manipulate pulmonary vascular resistance in these patients, such as controlled hypoventilation, or inhalation of low concentrations of oxygen, may also predispose patients to microatelectasis and ventilation-perfusion mismatch. The presence of unrecognized pulmonary venous desaturation will compromise systemic oxygen delivery by lowering the systemic arterial oxygen saturation. In each of the patients studied by Taeed et al.,¹⁹ pulmonary venous desaturation normalized with higher inspired oxygen or increased positive end expiratory pressure, which suggests that judicious use of inspired oxygen and/or positive end expiratory pressure may be beneficial in selected patients early after Norwood palliation. These data suggest that optimal alveolar recruitment may be the best way to preserve pulmonary gas exchange and prevent pulmonary venous desaturation in the patient with a functionally single ventricle.¹⁹

Summary

The neonate with functionally univentricular physiology presents unique challenges to the cardiac team. An integrated approach that applies working knowledge of cardiac anatomy, cardiopulmonary physiology, and basic principles of intensive care is essential to guide management of each individual patient. This requires cooperative and constructive involvement of a surgical, medical, nursing and respiratory care team experienced in the management of such patients. In the neonate with this physiology, systemic oxygen delivery is optimized by manipulating pulmonary and systemic resistances, augmenting total cardiac output, and utilizing strategies for ventilation that preserve optimal pulmonary recruitment.