19: Approach to the Fetus, Premature, and Full‐Term Neonate

Approach to the Fetus, Premature, and Full‐Term Neonate

Lee P. Ferguson1, Annette Y. Schure2, Peter C. Laussen3, and Kirsten C. Odegard4

1 Associate in Cardiac Anesthesia, Department of Anesthesiology, Critical Care and Pain Medicine, Boston’s Children’s Hospital, Harvard Medical School, Boston, MA, USA

2 Senior Associate in Cardiac Anesthesia, Department of Anesthesiology, Critical Care and Pain Medicine, Boston’s Children’s Hospital, Harvard Medical School, Boston, MA, USA

3 Executive Vice President of Health Affairs, Boston’s Children’s Hospital, Harvard Medical School, Boston, MA, USA

4 Senior Associate in Cardiac Anesthesia, Department of Anesthesiology, Critical Care and Pain Medicine, Boston Children’s Hospital, Harvard Medical School, Boston, MA, USA


The advances in pediatric cardiology and cardiac surgery over the past 35 years have resulted in a substantial decrease in morbidity and mortality associated with congenital heart disease (CHD). Most congenital heart lesions are now amenable to either anatomical or physiological repair early in infancy. Opinion regarding the optimal timing of corrective surgery for infants with symptomatic CHD regardless of their age or weight has undergone radical change over the past several decades. Rather than the previous strategy of initial palliation followed by correction in early childhood, the current approach favors complete repair within days to weeks of birth, if feasible. Advances in diagnostic and interventional cardiology, the evolution of surgical techniques, management of cardiopulmonary bypass (CPB), as well as refinements in postoperative care have all contributed to the successful strategy of early corrective two‐ventricle repair. Most recently, this approach has been extended to include premature and low birthweight (LBW) neonates also. However, the low mortality achieved with two‐ventricle repairs has not been the experience in LBW neonates undergoing palliation for single‐ventricle defects, such as hypoplastic left heart syndrome (HLHS).

Cardiac surgery in the premature and very LBW neonate presents additional challenges for the anesthesiologist. As for any pediatric cardiac procedure, a thorough understanding of the pathophysiology of the heart defect and planned surgical procedure and anticipation of specific postoperative problems are essential. There are further considerations, however, when providing anesthesia to the premature and LBW neonates: immaturity of the airway, lungs, cardiovascular system, liver, kidney, and central nervous system makes these infants more susceptible not only to surgical but also to anesthetic complications. Finally, as the limits for managing CHD continue to be extended, fetal cardiac interventions (FCIs) emerge as the next challenging frontier.This chapter will describe general principles relevant to anesthesia for the newborn, including the premature and LBW neonate with CHD. The impact of prematurity, the outcome of cardiac surgery in the premature and LBW neonate, and new directions with transcatheter ductus arteriosus stenting and FCIs will be discussed.

Approach to treatment in the neonate

Early palliation

In the early days of cardiac surgery for neonates, particularly those born prematurely, initial palliation or medical management of congenital cardiac defects was preferred. With technical limitations for various surgical techniques and relatively high risks associated with CPB in small babies, a successful repair often required a certain size or weight. Although the actual “target” weight that a neonate should achieve before the repair was never well documented, newborns with a weight of less than 2.5 kg were considered at higher risk.

The goal of palliative procedures is to control pulmonary blood flow sufficiently to allow for growth, whilst avoiding excessive blood flow to the pulmonary circulation that impairs systemic perfusion and volume overloads in the systemic ventricle. Nevertheless, palliation with a pulmonary artery (PA) band or a modified systemic‐to‐pulmonary artery shunt, such as a modified Blalock–Taussig (BT) shunt, can be technically demanding procedures in neonates and tend to be even more problematic in those with LBW (see Table 19.1).

Systemic‐to‐pulmonary artery shunt

While the originally described BT shunt directly connected the subclavian to the pulmonary artery, in contemporary practice a modified version of the procedure is used with an interposition polytetrafluoroethylene (PTFE) (Gore‐Tex) graft between the subclavian or innominate artery and pulmonary artery. The size and length of the tube graft as well as the site selected for the proximal anastomosis are critical. A relatively large shunt may lead to excessive pulmonary blood flow, congestive heart failure (CHF), and possible pulmonary vascular obstructive disease (PVOD). Conversely, a small shunt can result in inadequate pulmonary blood flow, lower arterial oxygen saturation (SaO2), possible shunt thrombosis, and poor growth of the pulmonary arteries making further repair more challenging. Assuming a normal cardiac output and hematocrit, absence of pulmonary venous desaturation and unrestricted mixing of systemic and pulmonary venous return in the atrium, an ideal SaO2 between 80 and 85% indicates a relatively balanced circulation with a pulmonary to systemic blood flow ratio (Qp: Qs) close to 1:1.

Table 19.1 Complications of palliative surgery

Early complications Late complications
Systemic‐to‐pulmonary artery shunt Excessive pulmonary blood flow
Heart failure
Distortion of pulmonary arteries

Inadequate pulmonary blood flow
Shunt obstruction:
Asymmetrical growth of the pulmonary arteries
Pulmonary vascular obstructive disease
Pulmonary artery banding Band too loose: excessive pulmonary blood flow Complications at band site:
Distortion and residual stenosis after repair

Band too tight: inadequate pulmonary blood flow Complications proximal to band site:
Right ventricular hypertrophy
Subaortic stenosis
Pulmonary valve stenosis
Complications distal to the band: pulmonary artery stenosis

A further problem of a small shunt is the likelihood of outgrowing the shunt size causing progressive cyanosis and requiring earlier surgical intervention. On the other hand, if a larger shunt size (i.e., ≥4.0 mm) is used in a newborn, the excessive pulmonary blood flow may compromise systemic perfusion, cause ventricular volume overload, heart failure, and prolonged postoperative recovery. In our experience, a shunt of 3.5 mm is the optimal size to use in a neonate weighing between 3 and 4 kg. If a 3.0 mm long shunt is used, the risk for acute thrombosis in the early postoperative period is increased, even in LBW neonates. The introduction of anticoagulation with heparin is important once hemostasis has been secured after surgery. Early platelet inhibition with an intravenous glycoprotein IIb/IIIa inhibitor may help reduce the risk of early thrombosis [1]. The mortality rate after neonatal modified BT shunt remains high, particularly for infants weighing less than 3 kg. A large multicenter study from the Society of Thoracic Surgeons Congenital Heart Surgery Database reported an in‐hospital mortality rate of 7.2% with a 13% rate of serious morbidity in neonates undergoing modified BT shunt placement without concomitant procedures [2]. Patients weighing <2.5 kg had a four‐fold increase in odds of death.

Transcatheter ductus arteriosus stent

Recent advances in transcatheter techniques and device technology have allowed stenting of the ductus arteriosus to be a successful alternative to surgically placed modified BT shunt in patients with ductus‐dependent cyanotic CHD. Although patients undergoing ductal stenting have a higher reintervention rate, retrospective multicenter cohort studies support an early survival advantage and shorter ICU stay for infants with ductal stenting [3, 4]. Ductal stent additionally avoids potential operative complications of modified BT shunt placement including vocal cord paralysis, chylothorax, and surgical site infection.

Attention to the ductal morphology is critically important in planning ductal stent placement. CT imaging, in addition to echocardiography, is helpful to delineate ductal anatomy and plan the access site to achieve the straightest trajectory to the ductus. A ductus off the underside of the aortic arch may be successfully approached with left carotid or axillary artery access. While in the past access to these vessels would usually occur via surgical cutdown, there are increasing reports of the safety of percutaneous access [5, 6]. Technical success can be achieved even with a highly tortuous ductus [7]. Prostaglandin E1 is typically discontinued 6–12 hours prior to the procedure to ensure the ductus is restrictive prior to stenting. Ductal stenting is not without some risk. Complications include ductal spasm, dissection, thrombosis, and stent embolization. Appropriate surgical and extracorporeal membrane oxygenation (ECMO) backup is essential. Drug‐eluting stents may decrease reintervention rates and in‐stent stenosis [8]. Although immunosuppressive levels of sirolimus may occur for a prolonged period following neonatal stent implantation, negative outcomes related to immunosuppression have not been reported yet [9]. Dual antiplatelet therapy is commonly used to reduce the risk of stent thrombosis.

Banding of the pulmonary artery

Banding of the main PA to reduce pulmonary blood flow can also be a challenging palliative procedure. If the band is too tight, severe cyanosis may occur, and if the band is too loose, the increase in pulmonary blood flow will contribute to CHF and possible PVOD. Distortion of the pulmonary artery secondary to migration of the band may contribute to both proximal and distal artery stenoses and complicate later repair, but can also lead to right ventricular hypertrophy, subaortic stenosis and pulmonary valve stenosis, depending on the relationship of the great arteries to the ventricular outflow tract. Determining the correct size of a band at the time of surgery is difficult. There are no accurate formulas for band size, and the hemodynamic changes at the time of band placement must be closely observed. Ideally, the banded PA will result in an increase in systemic systolic blood pressure of approximately 20%, and, depending on the underlying pathology, a fall in SaO2 to around 85% breathing room air. The pressure gradient across the band can also be directly measured. Usually, a pressure difference of approximately 50% proximal to the distal across the band is sufficient. Monitoring hemodynamic changes at the time of PA band placement is essential, and anesthetic techniques that could decrease ventricular function or cardiac output are best avoided. Therefore, an opioid technique is most often necessary, and extubation should be delayed until the hemodynamic effect of the band is determined as the patient emerges from anesthesia and starts to wean from mechanical ventilation.

The case for early complete repair

As noted previously, whenever possible, early two‐ventricle repair of congenital cardiac defects is the preferred approach in the modern era. Avoiding the long‐term consequences of excessive volume and pressure overload on the ventricles and pulmonary vasculature as well as the potentially detrimental effect of chronic hypoxia, early repair allows for more normal growth and development. The considerable advances in cardiac surgery and CPB techniques have contributed to a dramatic reduction in mortality following cardiac surgery in newborns [10–12], but these patients nevertheless remain at risk for significant end‐organ impairment, particularly neurological injury [13].

Medical management alone, with the goal of controlling pulmonary blood flow and volume overload on an immature myocardium, is often extremely difficult and unsuccessful. Long‐term effects associated with pulmonary overcirculation, chronic volume and/or pressure load on the ventricles, and cyanosis may substantially alter growth and development and lead to myocardial and pulmonary injury which will influence the outcome of subsequent repairs. In most newborns with a large left‐to‐right shunt, the imbalance between pulmonary and systemic blood flow will increase as pulmonary vascular resistance (PVR) falls in the first few weeks of life and the physiologic nadir in hematocrit is reached. The clinical manifestations of an infant with CHF are shown in Box 19.1. Tachypnea and the additional work of breathing, secondary to an increase in pulmonary blood flow and total lung water, raise the metabolic demand and the percentage of the total cardiac output directed toward respiratory muscle work (most notably, the diaphragm). This essentially diverts cardiac output from other metabolically active functions, particularly from the splanchnic circulation and absorption of enteral nutrition. The abnormal circulatory physiology is unable to meet metabolic needs and patients fail to achieve normal growth.

Special considerations for the neonate

Although the risk for early mortality in neonates undergoing cardiac surgery may be increased, randomized and prospective studies comparing early surgery versus medical management in hopes of weight gain in neonates with critical lesions have not been performed. Such has been the nature of many of the advances in CHD management. For example, infants with an increased pulmonary blood flow who undergo delayed surgical intervention often fail to thrive and are at risk of recurrent respiratory infections. Their work of breathing and energy expenditure is significantly increased, and cardiomegaly with hyperinflated lung fields is evident on chest radiograph. Cardiac surgery and CPB may be delayed because of concerns for intercurrent infection and the risk of exacerbation or reactivation of inflammatory lung processes, which in turn may cause intrapulmonary shunting and severe hypoxemia, pulmonary hypertension, and prolonged mechanical ventilation. The early repair or palliation of defects to limit pulmonary overcirculation and volume load on the systemic ventricle will often avoid these complications.

The optimal timing of delivery of neonates with CHD is now felt to be at 39–40 weeks’ gestation. The former practice of elective delivery at earlier gestational ages (e.g., 36–37 weeks) is associated with higher mortality and should be avoided unless there are clear maternal or fetal indications [14, 15].

The management of neonates in the immediate postoperative period after a two‐ventricle repair can be a challenge, but with substantially improved outcomes, mortality alone is no longer a reliable index against which to measure or compare new or alternative treatments in this group of patients. The focus should rather shift to long‐term quality of life, ongoing need for reintervention, and hospital admissions. Survival after palliative procedures in patients with complex single‐ventricle defects remains lower, although there has been a steady improvement in mortality and longer‐term outcomes [16].

Limited physiologic reserve

Care of the critically ill neonate requires an appreciation of the special structural and functional features of immature organs. The neonate appears to respond more quickly and extremely to physiologically stressful circumstances; this may be expressed in terms of rapid changes in, for example, pH, lactic acid, glucose, and temperature [17].

The physiology of the preterm and full‐term neonate is characterized by a high metabolic rate and oxygen demand (two‐ to three‐fold increase compared with adults), which may be compromised at times of stress because of limited cardiac and respiratory reserve. The myocardium in the neonate is immature: contractile tissue constitutes only 30% of the myocardial mass, compared with 60% in mature myocardium. In addition, neonates have a lower velocity of shortening, a diminished length–tension relationship, and a reduced ability to respond to afterload stress [18, 19]. Because the compliance of the myocardium is reduced, the stroke volume is relatively fixed, cardiac output is heart rate‐dependent, and the Frank–Starling relationship is functional only within a narrow range of left ventricular end‐diastolic pressure compared with the mature myocardium. The cytoplasmic reticulum and T‐tubular system are underdeveloped, and the neonatal heart is dependent on the trans‐sarcolemmal flux of extracellular calcium both to initiate and sustain contraction. It is important to note that much of this information is derived from animal data.

Cardiorespiratory interactions are important in neonates and infants. In simple terms, ventricular interdependence refers to the fact that a relative increase in ventricular end‐diastolic volume and pressure causes a shift of the ventricular septum and diminishes the diastolic compliance of the opposing ventricle [20]. This effect is particularly prominent in the immature myocardium. Therefore, a volume load from an intracardiac shunt or valve regurgitation, and a pressure load from ventricular outflow obstruction or increased vascular resistance, may lead to biventricular dysfunction. For example, in neonates with tetralogy of Fallot and severe outflow obstruction, hypertrophy of the ventricular septum may contribute to diastolic dysfunction of the left ventricle and an increase in end‐diastolic pressure. This does not improve immediately after repair in the neonate as it takes some time for the myocardium to remodel. A persistent volume load to the left ventricle following surgery, such as from a residual ventricular septal defect (VSD), may further exacerbate this situation.

The mechanical disadvantage of an increased chest wall compliance and reliance on the diaphragm as the main muscle of respiration limits ventilatory capacity in the neonate. The diaphragm and intercostal muscles have fewer type I muscle fibers (slow‐contracting, high‐oxidative fibers for sustained activity) and this contributes to early fatigue when the work of breathing is increased. In the newborn, only 25% of fibers in the diaphragm are type I, reaching a mature proportion of 55% by 8–9 months of age [21]. The diaphragmatic function may be significantly compromised by raised intra‐abdominal pressure such as from gastric distension, hepatic congestion, and ascites.

The tidal volume of full‐term neonates is between 4 and 6 mL/kg and, because of the above mechanical limitations, minute ventilation is dependent on respiratory rate. The resting respiratory rate of the newborn infant is between 30 and 40 breaths/min, which provides the optimal alveolar ventilation to overcome the work of breathing and match the compliance and resistance of the respiratory system. When the work of breathing increases, such as with parenchymal lung disease, airway obstruction, cardiac failure, or increased pulmonary blood flow, a larger proportion of total energy expenditure is required to maintain adequate ventilation. Infants, therefore, fatigue readily and fail to thrive.

The neonate has a reduced functional residual capacity (FRC); FRC is determined by the balance between the chest wall and lung compliance. Closing capacity is increased in newborns with airway closure occurring during normal tidal ventilation [22]. Neonates have reduced oxygen reserve, and in conjunction with the increased basal metabolic rate and oxygen consumption that is two to three times adult levels, are at risk for hypoxemia. Atelectasis and hypoxemia do not occur in the healthy neonate because FRC is maintained by dynamic factors including tachypnea, breath stacking (early inspiration), expiratory breaking (expiratory flow interrupted before zero flow occurs), and from laryngeal breaking (auto positive end‐expiratory pressure). The propensity of the neonatal capillary system to leak fluid out of the intravascular space is especially pronounced in the neonatal lung [23], in which the pulmonary vascular bed is almost fully recruited at rest and the increases in lymphatic flow required to handle elevated mean capillary pressures (due to augmented pulmonary blood flow) are limited [24].

The glomerular filtration rate is generally low at birth but normalizes over the first few months of life. Urinary sodium excretion increases slowly during the first 2 years of life, and the inability of immature kidneys to concentrate urine and to excrete acute water and sodium loads makes fluid management in neonates, especially preterm infants, difficult. Urinary acidification capability is limited in neonates and the bicarbonate threshold is reduced. Thus, premature infants have decreased serum bicarbonate levels and lower serum pH (a non‐anion gap acidosis). Neonates tolerate fluid restriction poorly, so fasting should be kept to a minimum and intravenous (IV) fluid started early; however, excessive fluid administration (as after CPB) is also poorly tolerated.

Drug pharmacodynamics and kinetics may be different in the newborn because of immature hepatic and renal function. In addition to altered drug metabolism, protein binding and clearance, the drug volume of distribution is affected by the increase in total body water of the neonate compared with the older patient. The dosing of drugs that mainly depends on renal excretion will have to be reduced and, if possible, the plasma concentration should be closely assessed to avoid accumulation and side effects. Larger doses of furosemide are needed to induce diuresis in neonates compared with adults.

The caloric requirement for neonates, especially preterm neonates, is high (100–150 kcal/kg/24 hours) [25]. Supplying adequate nutrition can be a difficult task, especially if the total amount of fluid, administered either parenterally or enterally, must be restricted, as is the case in premature neonates with CHD. Hyperosmolar feedings have been associated with an increased risk of necrotizing enterocolitis (NEC) in the preterm neonate or in the full‐term neonate who has decreased splanchnic blood flow of any cause (e.g., left‐sided obstructive lesions).

Stress response

In general terms, the “stress response” is a systemic reaction to injury, with hemodynamic, endocrinologic, and immunologic effects (Box 19.2). Stress and adverse postoperative outcomes have been linked in critically ill newborns and infants. This is not surprising given their precarious balance of limited metabolic reserve and increased resting energy demand. Metabolic derangements, such as altered glucose homeostasis, acidosis, salt and water retention, and a catabolic state contributing to protein breakdown and lipolysis, are commonly seen following major stress in sick neonates and infants [17]. This complex of maladaptive processes may be associated with prolonged mechanical ventilation and intensive care unit (ICU) stay as well as with increased morbidity and mortality.

The neuroendocrine stress response is activated by afferent neuronal impulses from the site of injury, traveling via sensory nerves through the dorsal root of the spinal cord to the medulla and hypothalamus. Anesthesia can therefore have a substantial modulating effect on this neuroendocrine pathway by virtue of providing analgesia and loss of consciousness. Outcomes after major surgery in neonates and infants may be improved when the stress response is attenuated. This was initially reported in two controlled, randomized trials comparing N2O/O2/curare anesthesia with or without fentanyl in neonates undergoing patent ductus arteriosus (PDA) ligation [17], and with or without halothane in neonates undergoing general surgery [26]. Fentanyl doses as low as 10 μg/kg may be sufficient for effective baseline anesthesia in neonates, although larger doses are necessary for prolonged anesthesia [27].

It is important to distinguish between suppression of the endocrine response and attenuation of hemodynamic response to stress. Because of their direct effects on the myocardium and vascular tone, anesthetic agents can readily suppress the hemodynamic side‐effects of the endocrine stress response. However, the postoperative consequences of the endocrine stress response, in particular fluid retention and increased catabolism, remain unabated. Relying on hemodynamic variables to assess the level of “stress” is therefore often inaccurate. Metabolic indices such as hyperglycemia and lactic acidosis are also indirect markers of “stress”, particularly as they are influenced by other factors such as cardiac output and catecholamine levels.

The effect of surgical stress has been specifically evaluated in neonates and infants undergoing cardiac surgery. Wood et al. first demonstrated a substantial increase in epinephrine and norepinephrine levels in response to profound hypothermia and circulatory arrest in infants [28]. The hormonal and metabolic response was further characterized by Anand et al. and noted such response to be more extreme and distinct from that seen in adults [29]. In addition to an increase in catecholamine, glucagon, endorphin, and insulin levels, hyperglycemia and hyperlactatemia persisted into the postoperative period. In an important subsequent study, Anand and Hickey compared a high dose sufentanil technique with halothane/morphine anesthesia in 45 neonates undergoing cardiac surgery and deep hypothermic CPB [30]. They reported a significant attenuation of hormonal and metabolic responses to surgery and bypass in the sufentanil group, with less postoperative morbidity and mortality. A conclusion from these studies supported the notion that reducing the stress response with high‐dose opioid anesthesia and extending this into the immediate postoperative period is important to reduce the morbidity and mortality associated with congenital heart surgery in neonates.

These studies were performed over two decades ago. During the intervening period, there have been substantial changes in the perioperative care of infants with heart disease as well as in the management of CPB in general; along with these changes, outcomes have considerably improved. Changes in surgical practice, especially the timing of surgery, have altered the perioperative course of numerous defects and reduced the incidence of certain pathophysiologic consequences. Currently, it is less likely to see neonates in the immediate post‐bypass period with extensive peripheral edema, anasarca, or other associated complications such as impaired ventricular function, reactive pulmonary hypertension, and substantial alterations in lung compliance and airway resistance. For example, just one or two decades ago, postoperative pulmonary hypertensive crises were relatively common events in infants who had been exposed to weeks or months of high pulmonary pressure and flow, such as truncus arteriosus, complete atrioventricular canal defects, and the transposition of the great arteries with VSDs. High‐dose opioids were an important management component for patients at risk for pulmonary hypertensive crises. In current practice, however, these patients are operated upon at an earlier age and are therefore less likely to have significant or irreversible changes in the pulmonary vascular bed. Consequently, a strategy of high‐dose opioid anesthesia to blunt the stress response may be a less critical determinant of outcome.

Further, it has been well demonstrated that high‐dose opioid anesthetic techniques do not consistently block the endocrine stress response to cardiac surgery. The dose of sufentanil used by Anand and Hickey was extremely high and difficult to translate to the more common practice of fentanyl‐based anesthesia. More recent studies in neonates, infants, and older children undergoing cardiac surgery have demonstrated attenuation of the pre‐bypass endocrine and hemodynamic response to surgical stimulation with a variety of anesthetic techniques. These studies have included high‐dose fentanyl (50 μg/kg) either by bolus or infusion [31, 32], high‐dose bolus fentanyl (25–150 μg/kg) with or without low‐dose isoflurane [33], remifentanil infusions at various rates (0.25–5 μg/kg/min) [34], and more recently low‐dose fentanyl (10‐25 μg/kg) with dexmedetomidine [35, 36]. Based upon the lack of significant stress responses reported in these studies, it is reasonable to conclude that there was appropriate neuraxial inhibition in these patients and that they were adequately anesthetized during the surgery. It is not possible to conclude, however, that one technique is superior to another. No specific dose‐response between plasma opioid level and level of hormone or metabolic stress response has been established, nor has a specific benefit been demonstrated for the method or route of opioid administration, i.e., bolus or continuous infusion.

The use of lower doses of opioids may reduce the duration of mechanical ventilation and ICU length of stay by avoiding the resulting respiratory depression from a high‐dose opioid technique [37]. Recent studies have described successful early extubation of neonates undergoing heart surgery involving CPB with general anesthesia using low‐dose fentanyl, volatile anesthetic agent and dexmedetomidine, administered as a continuous infusion without loading [37–39].

May 17, 2023 | Posted by in ANESTHESIA | Comments Off on 19: Approach to the Fetus, Premature, and Full‐Term Neonate

Full access? Get Clinical Tree

Get Clinical Tree app for offline access