What the Anesthesiologist Should Know before the Operative Procedure

Anesthesia for pediatric cardiac catheterization provides multiple challenges.

  • The location is remote from the operating room.
  • Children have acquired or have congenital heart disease (CHD), covering the spectrum from relatively healthy to critically ill.
  • Children with single ventricle physiology or pulmonary hypertension (PHTN) are high-risk patients with unique cardiovascular limitations.
  • Cardiovascular effects of anesthetic drugs, the mode of ventilation, inspired oxygen content (FiO2), arterial carbon dioxide (PaCO2), and acid-base status must all be integrated in such a way that diagnostic studies truly reflect the child’s cardiac status.
  • Higher-risk interventional procedures now outnumber diagnostic studies by 2 to 1.

1. What is the urgency of the surgery?

What is the risk of delay in order to obtain additional preoperative information?


1. Preoperative evaluation

I. Diagnostic catherization

Hemodynamic catheterization quantifies the gradients, shunts, pulmonary blood flow, pulmonary vascular resistance (PVR), systemic blood flow, and systemic vascular resistance (SVR) on which clinical decision making is based. The following indications exist for diagnostic studies.

Continue Reading

Algorithm 1

Shunt fraction: Qp/Qs

A simpler method to calculate Qp:Qs is the following. It assumes the child is breathing room air, the VO2 is normal, and that the child is not severely anemic. Qp and Qs are not actually measured.

Algorithm 2

Pressures are measured at the relevant locations in the pulmonary and systemic circulations for derivation of PVR and SVR.

Algorithm 3

In children, Qp and Qs are indexed to weight (cardiac index, not the measured cardiac output), and the units for PVR and SVR are Wood units.

Abbreviations: VO2: oxygen consumption; SpvO2: pulmonary venous oxygen saturation; SpaO2: pulmonary artery oxygen saturation; SaoO2: aortic oxygen saturation; SsvcO2: superior vena cava oxygen saturation; mean PAP: mean pulmonary artery pressure; PVP: pulmonary venous pressure; MAP: mean arterial pressure; CVP: central venous pressure.

A. Hemodynamic assessment

The clinician, especially a trainee, should pay close attention to the calculations performed because it enhances the pathophysiologic understanding of CHD. Understanding the hemodynamic calculations makes one a physician who comprehends the full spectrum of CHD rather than merely a technician. The measurement of pulmonary blood flow (Qp) and systemic blood flow (Qs) via the Fick equation allows the calculation of shunt fraction. Oxygen consumption (VO2) can be directly measured or estimated from tables based on age and weight.

The most common scenario for diagnostic catheterization is prior to the Glenn and Fontan procedures.

B. Examination of complex anatomy

The complexity of certain CHD conditions and/or poor echocardiographic windows may necessitate a diagnostic catheterization.

C. Evaluation of myocardial function and pulmonary hypertension

Hemodynamic indices are often measured in children with heart failure. Heart failure during infancy can be due to anomalous coronary anatomy. This is amenable to surgical correction but may require coronary angiography for a definitive diagnosis. Patients with idiopathic PHTN require measurement of their pulmonary artery (PA) pressure and PVR and evaluation of pulmonary vascular responsiveness to vasodilating drugs (oxygen, nitric oxide) in planning therapy. Untreated left to right shunts resulting in PHTN and possible elevated PVR need to be assessed to determine whether surgery is appropriate and if so, whether specific therapies for PHTN should be considered preoperatively (sildenafil) or perioperatively (nitric oxide).

D. Evaluation after heart transplant

Standard follow-up requires endomyocardial biopsies at predetermined intervals after the transplant. Children range from those with excellent heart function, thus having an elective procedure, to those who present with severe heart failure from rejection.

E. Electrophysiologic studies

These are usually initial diagnostic studies to induce and map the arrhythmia, followed by therapeutic intervention, if suitable.

II. Catherization in the laboratory environment

The catheterization laboratory is an unfamiliar location, remote from the operating room. The environment is cramped and access to the anesthesia machine, drug carts, intravenous poles, monitors, and, most critical, the patient is restricted. Even in the new so-called “hybrid labs” designed for combined percutaneous and surgical procedures, the extra space is entirely devoted to the surgical scrub table, along with room for a cardiopulmonary bypass (CPB) circuit or extracorporeal membrane oxygenator (ECMO) circuit.

The lesson here is that it is crucial to spend extra time to organize the “workspace.” Monitoring and intravenous access, along with the anesthesia machine and breathing circuit, need to be organized, labeled, and easily accessible. Other equipment should not restrict access to the child unless absolutely necessary.

All standard emergency cardiac medications should be available in doses appropriate for the child and placed in an easily accessible location. Placing medications on the anesthesia cart – the usual procedure – is not sensible if the anesthesia cart is trapped behind other equipment. Trying to manage a resuscitation cannot be done competently when one cannot locate the injection port of the intravenous setup.

Last, in an anticipated difficult airway, my preference is to induce anesthesia and manage the airway in the operating room, where equipment and other anesthesia personnel are readily available. After the airway is secured, the child can be transferred to the catheterization laboratory.

c. Intraoperative management: What are the options for anesthetic management, and to determine the best technique?

Anesthetic goals

During the early days of pediatric cardiac catheterization, the “lytic cocktail” or “CM3” (chlorpromazine, meperidine, and promethazine) was given by intramuscular injection. This produced a child with stable cardiorespiratory parameters at the expense of prolonged deep sedation and lack of titratability.

Today, the wide array of sedatives and anesthetics come with far greater risk of cardiorespiratory depression but can be titrated to effect and are relatively short acting. Light or “conscious sedation” as used for most adults has minimal cardiorespiratory effects when compared to general anesthesia but is not feasible for most children. Almost all children receive general anesthesia or deep sedation, with the difference between the two determined by whether the airway has been instrumented. This is a semantic point because an anesthetized child can breathe spontaneously with a natural airway.

A hemodynamic study must accurately reflect the true physiological state of the child. Given that there are no hemodynamically neutral anesthetic drugs, this requirement is challenging. Furthermore, there is no surgical stimulation to offset the effect of anesthetic drugs, making the maintenance of baseline hemodynamics even more difficult. (The cardiovascular effects of anesthetic drugs in children with CHD are discussed in a later section.) Different ventilatory modes affect venous return and the loading conditions for the heart. The ventilatory mode, FiO2, PaCO2, temperature, and acid-base status all critically effect hemodynamic measurements.

The presence of an anesthesiologist ranges from near universal to rare, depending on institutional practice and the comfort of the cardiologist in managing sedation. (A combination of ketamine and midazolam given in low doses has been suggested as a safe option in the hands of the pediatric cardiologist.) The guidelines of the American Academy of Pediatrics stress aspects of care that are a routine part of anesthesiology, such as an appropriate fasting interval, a focused history, and physical examination with careful attention to the airway, continuous monitoring by a dedicated individual, and a medically supervised duration of recovery to full consciousness.

The issue of who manages sedation and anesthesia for cardiac catheterization is relevant in light of the recent findings of the Pediatric Perioperative Cardiac Arrest Registry. Fully one-third of cardiac arrests happened in children with CHD, and most of these occurred during noncardiac surgery (50%) or cardiac catheterization (20%). A convincing argument has been made that all pediatric cardiac catherizations should be attended by an anesthesiologist. In addition, higher-risk cases require a pediatric anesthesiologist with subspecialized experience or training in CHD.

3. What are the most common intraoperative complications and how can they be avoided and/or treated?

a. What are the most common intraoperative complications and how can they be avoided and/or treated?

Retrospective series have defined a complication rate of 7% to 8%, a major complication rate of 2%, and a mortality of 0.1% to 0.2%.

The most likely complication is vascular injury, which can occur at the access site, allowing direct compression to stop bleeding. Bleeding may also involve the great vessels of the heart, resulting in life-threatening hemorrhage or cardiac tamponade. Blood must be available and interventional procedures that use bigger vascular access sheaths and have a higher rate of complications should only be performed in centers with cardiac surgery and ECMO capabilities. Given the alternative of near certain death, ECMO can result in survival with good neurologic outcome.

Great vessel bleeding during balloon angioplasty can be temporized by reinflation of the balloon. Resuscitation, rapid transfusion, and arterial monitoring are best done through the vascular access sheaths. Cardiac tamponade can be treated by needle pericardiocentesis, guided fluoroscopically or echocardiographically.

Major vascular injuries most often will require surgical repair or ECMO. Arrhythmias are usually iatrogenic and respond to catheter withdrawal. Rarely does heart block ensue and temporary pacing necessary. Other rare events include thromboembolism, air embolus, and embolism of device parts, such as coiling wires or stent material.

Hemodynamic compromise may result from the cardio-respiratory effects of anesthesia. The highest risk groups are single ventricle infants and children with PHTN.

Effect of anesthetic drugs in children with CHD

There is a dearth of good data on anesthetic effects in CHD. Following is brief summary of the effect of anesthetic drugs in children with CHD. How do anesthetics alter the balance between SVR and PVR, thereby changing the nature and direction of shunt flow? How do anesthetics affect the key determinants of cardiac output which are preload, afterload, heart rate, contractility, and shunt flow?

A. Propofol

Propofol is commonly used for the induction and maintenance of anesthesia. Frequently administered via infusion to a spontaneously breathing child with a natural airway, this popular technique avoids airway instrumentation. The rapid clearance of propofol enables a prompt, smooth awakening that is usually free from nausea, vomiting, and emergence agitation.

Children can tolerate the respiratory depressant effects of propofol far better than adults. Infusion rates in excess of 200 mcg/kg/min may be required. In healthy children without CHD, propofol causes a decrease in blood pressure of 30% due to decreases in SVR (15%) and heart rate (10-20%). Data in children with CHD demonstrate that propofol’s primary effect of decreased blood pressure is through a decrease in SVR. Systemic cardiac output increased without a change in heart rate or PVR.

In children without shunts, there was a small decrease in PaO2, presumably due to decreased respiratory drive but no increase in PVR. Importantly, in children with shunts, the decrease in SVR was consequential. Left-to-right shunting decreased and right-to-left shunting increased. The effects are dose-dependent, with usual infusions rates between 100 to 200 mcg/kg/min in most studies.

While most children tolerate propofol well, it should be used with caution as the sole agent in the setting of right-to-left shunts and particularly in those patients for whom a decrease in systemic afterload is dangerous (aortic stenosis, hypertrophic cardiomyopathy, severe ventricular dysfunction).

B. Ketamine

Ketamine combines anesthesia, analgesia, cardiovascular stability, and lack of respiratory depression with maintenance of airway reflexes. Drawbacks to ketamine are prolonged action, emergence phenomena, and a dissociative anesthetic state that may result in patient movement during the procedure.

Anesthetic doses of ketamine (50-75 mcg/kg/min) are much higher than analgesic doses (5-10 mcg/kg/min). In children with CHD, ketamine (50-75 mcg/kg/min) resulted in the maintenance of the relationship between SVR and PVR. Systemic blood pressure increased through an increase in cardiac output with little change in heart rate.

Increased inotropy may be beneficial for children with significant ventricular dysfunction. A common question is whether ketamine’s sympathomimetic action raises PVR in children with PHTN. Evidence suggests that in the setting of normocarbia with supplemental oxygen administered, PVR is not increased. Combined with 0.5 minimal alveolar concentration (MAC) sevoflurane in spontaneously breathing children with severe PHTN, ketamine did not raise PVR and overall hemodynamics were stable.

C. Etomidate

Etomidate is little studied in children with or without CHD. Bolus dosing of 0.3 mg/kg was well tolerated with maintenance of systemic blood pressure and preservation of the balance between SVR and PVR. Transient adrenal suppression does occur even with bolus dosing and could be consequential in the sickest children. Other drawbacks of etomidate are pain on injection and emesis.

D. Volatile anesthetics

Halothane, no longer available in the United States, will not be addressed here other than to state that at equivalent MAC levels to isoflurane and sevoflurane, there is greater myocardial depression and suppression of the baroreceptor-mediated increase in heart rate.

The Qp:Qs ratio is unchanged with halothane, sevoflurane, and isoflurane when ventilation is controlled, even with a high FiO2. Cardiac output with isoflurane was maintained even at 1.5 MAC as the decrease in SVR and increase in heart rate offset reduced inotropy. Sevoflurane was similar, but an overall decrease in cardiac output resulted from the lack of a compensatory increase in heart rate combined with reduced inotropy. Children with significant ventricular dysfunction may not hemodynamically tolerate MAC levels of volatile anesthesia.

E. Fentanyl/midazolam

Combining a synthetic opioid with a short-acting benzodiazepine has long been favored by pediatric cardiac anesthesiologists. When the combination is the only anesthetic or sedative, its duration of action usually limits its use to sick children who will remain intubated at the conclusion of the procedure.

The ratio of Qp:Qs is unchanged as long as ventilation is controlled. Fentanyl at doses of 25 mcg/kg maintains SVR, PVR, and systemic blood pressure. The caveat here is that significant bradycardia is avoided. Therefore, the general tone of the sympathetic nervous system greatly influences the blood pressure, and patients can easily become hypotensive due to changes in the loading conditions of the heart while intropy is preserved. Particularly sympatholytic is the combination of synthetic opioids with volatile anesthetics, midazolam or propofol.

F. Dexemedetomidine

Dexemedetomidine is a selective alpha-2 agonist with an emerging role in pediatric anesthesia and for children with CHD). It is an analgesic and sedative, with minimal respiratory depression. The sedation of dexmedetomidine has been likened to a natural sleep state. Decreased sympathetic outflow is usually marked by relative bradycardia and stable blood pressure.

However, side effects detrimental to children with CHD are hypertension (peripheral alpha 2 agonist effect), bradycardia, and hypotension. The pros and cons of dexmedetomidine for CHD surgery have recently been debated. Dexmedetomidine may be poorly tolerated in heart-rate-dependent neonates and infants.

What is the author’s preferred anesthesia technique and why?

1. Techniques for anesthetic induction and maintenance

Any of the usual array of anesthetic induction and maintenance techniques can be used, taking into account the procedure, the type of CHD, and the child’s cardiovascular reserve.

A. Premedication

Children with CHD are familiar with hospitals and are often anxious. Parents desire to limit the psychological stress and anxiety for their child. They may have also developed a “routine” for induction based on past experiences and one should be reluctant to deviate from this.

Premedication should be used liberally and parental presence at induction is not a substitute for premedication. Parental presence at induction precludes separation anxiety but premedication is better than parental presence at lessening anxiety during mask induction.

A most distressing aspect of pediatric anesthesia is a forceful induction of the previously happy, unpremedicated child who completely refuses the mask despite the presence of their parent. The best prevention for this scenario is oral midazolam (0.5-0.75 mg/kg). Higher doses of oral midazolam provide more reliable sedation with faster onset and without an increased risk of airway obstruction or respiratory depression. Discharge is not delayed with higher doses of midazolam.

B. Mask induction with volatile anesthesia

This is well tolerated in the majority of children with CHD because they do not have ventricular dysfunction or severe obstructive lesions. Volatile anesthetics are used for maintenance of anesthesia with a child either breathing spontaneously or being manually ventilated via an endotracheal tube (ETT) or laryngeal mask airway (LMA). For children with significant cardiac compromise, controlled ventilation, neuromuscular blockade, and reduced-dose volatile anesthesia may be necessary during the maintenance of anesthesia.

C. Intravenous induction

Available agents are propofol, ketamine, etomidate, opioids, benzodiazepines, and dexmedetomidine. Propofol, ketamine, and dexmedetomidine all offer the potential for maintenance of anesthesia. Combinations of opioids and benzodiazepines sufficient for anesthesia usually have a longer duration of action and are used in children who will remain intubated after the procedure. An intravenous induction allows the depth of anesthesia to be titrated to the hemodynamic response combined with the ability to support circulating blood volume with a fluid bolus. The appropriate drug for induction is determined by the child’s cardiovascular status.

D. Intramuscular induction

In the hands of pediatric anesthesiologists intramuscular (IM) induction with ketamine is usually reserved for two scenarios. One is the older child, usually developmentally delayed, who refuses premedication, an intravenous catheter or mask induction. The other is the hemodynamically compromised child who does not have an intravenous catheter and has poor venous access. A dose of 3 to 4 mg/kg will produce an anesthetized, hemodynamically stable child within 10 to15 minutes. An increase in oral secretions may occur. Glycopyrrolate may be added to the IM injection. The child must be observed and monitored after IM ketamine.

Here, a brief mention of the subject of emergence is necessary. Large sheaths would have been placed in the femoral artery/vein or the internal jugular vein. Pressure is held at the site at the conclusion of the procedure and a compressive dressing is applied. Hemostasis at the site is tenuous and can easily be disrupted by rises in arterial pressure from agitation or increased central venous pressure from coughing and straining during emergence. If bleeding resumes, then pressure must be re-applied, which is poorly tolerated in the awake child.

It is my practice to routinely extubate under deep volatile anesthesia, unless a compelling case can be made that it is unsafe. However, “deep extubation” does not preclude emergence agitation; it simply transfers it to an event that will occur in the recovery area instead of the catheterization laboratory. Therefore, I supplement a deep extubation with further intravenous agents to reduce the incidence of emergence agitation and prolong the time to full awakening.

Multiple drugs have been successfully used to prevent emergence agitation; but one is clearly not better than the other. However, given the lack of post-procedure pain, opioids are a poor choice and not recommended. It is my practice to administer small doses of propofol (0.5-1 mg/kg), which ensures a quiet child for approximately the first 15 to 30 minutes in PACU. Dexemedetomidine is another attractive choice in this setting, and can be given as a bolus or low-dose infusion throughout the case.

2. Diagnostic procedures

This section is a review of common CHD conditions that require diagnostic catheterization. It is not possible to cover every potential scenario, as many CHD conditions that do not routinely require cardiac catheterization may need diagnostic studies in certain unique scenarios.

For diagnostic studies, it is critical to maintain cardio-respiratory homeostasis to ensure that the cardiologist collects clinically relevant data. Readers are referred to the earlier section on the cardiovascular effects of commonly used anesthetics and sedatives. Of equal importance, particularly when determining the PVR and the Qp:Qs ratio, is the ventilatory status. Hypoventilation with significant hypercarbia will alter hemodynamic calculations. Due to the pulmonary vasodilating effect of oxygen, hemodynamic calculations are done while the patient is breathing room air.

The simple guideline is that spontaneous ventilation under anesthesia is acceptable only if hypercarbia is avoided and oxygen saturation can be maintained at approximate baseline levels without significant supplemental oxygen. If not, the child will require assisted ventilation of some type.

A. Tetralogy of Fallot with pulmonary atresia (TOF/PA) and major aortopulmonary collateral arteries (MAPCAs)

TOF/PA with MAPCAs is a unique, rare variant of tetralogy of Fallot, where there is no antegrade blood flow from the heart to the lungs because the pulmonary valve is atretic. All systemic venous return is shunted across the VSD to the left ventricle. Therefore, air bubbles in the intravenous line will travel systemically. The source of pulmonary blood flow is the MAPCAs that originate from the aorta. They often flow tortuously and do not insert into a normal branch pulmonary artery configuration.

The volume of pulmonary blood flow can be estimated by the baseline room air oxygen saturation. Oxygen saturation higher than 80% suggests a Qp:Qs >1 and an oxygen saturation of less than 80% indicates a Qp:Qs <1.

A high oxygen saturation reflects pulmonary overcirculation and possible congestive heart failure. Conversely, some children may be extremely cyanotic. Ventricular function is usually normal and children tolerate a standard mask induction with volatile anesthesia. Without antegrade blood flow from the right heart there can be no hypercyanotic or “Tet spell,” which is reassuring.

Severe hypotension for any reason will also increase cyanosis because the driving force for pulmonary blood flow is systemic arterial pressure. Therefore, both a fluid bolus and early recourse to vasopressors are needed to prevent the dangerous situations, hypotension and hypoxia.

The purpose of the cardiac catheterization is to define the MAPCA anatomy in preparation for eventual surgery. Hemodynamic calculations are done but are not the focus of the procedure and, therefore, maintenance of baseline hemodynamics is not as important.

B. Left ventricular failure

Infants with structurally normal cardiac anatomy who present with heart failure in the first few months of life often require diagnostic catheterization. The differential diagnoses are dilated cardiomyopathy, postviral cardiomyopathy, anomalous left coronary artery from the PA (ALCAPA,) and rare nutritional deficiencies manifesting as heart failure.

The primary reason for diagnostic study is to rule out ALCAPA if noninvasive studies have not been able to do so. ALCAPA occurs when the left main coronary artery originates from the PA. A common misconception is that the myocardial injury occurs because the majority of the left ventricle is supplied with deoxygenated PA blood. If this is the case, presentation would be in the newborn period. However, infants present at a few months of age, when the natural fall in PVR results in decreased flow through the anomalous left main coronary artery. The combination of decreased perfusion pressure with systemic venous blood results in severe myocardial ischemia and heart failure.

ALCAPA is a potentially surgically treatable condition, creating urgency for diagnostic catheterization. If ALCAPA has been ruled out (with echocardiography), cardiac catheterization may still be required as part of pretransplant evaluation and/or to assess hemodynamics. (Beyond infancy, the most likely diagnoses is dilated cardiomyopathy or postviral cardiomyopathy; these children may also require catheterization to assess hemodynamics and suitability for transplant. Anesthesia for children with cardiomyopathy recently has been well reviewed.)

These infants have profound decreases in left ventricular function. The echocardiographic appearance is usually more severe than the clinical presentation suggests. This is because infants are, by definition, sedentary. When they have left ventricular dysfunction severe enough to be tachypneic at rest and diaphoretic with feeds, the ejection fraction is 20%, at best. Their degree of cardiovascular reserve is minimal. Usually, these infants have already been medically stabilized with milrinone +/- inotropes and have sufficient intravenous access and, usually, invasive monitoring. If already intubated, the anesthetic is “plug and play,” consisting of vigilance, +/- neuromuscular blockade and continuation of sedative infusions (usually fentanyl/midazolam).

Infants not already intubated present a major challenge. Spontaneous ventilation with a natural airway is an attractive option. Ketamine and midazolam given as cautious intermittent boluses or infusion is a common method employed and is very cardiovascularly stable. Propofol’s effect on SVR and dexmedetomidine’s vagotonic effect on heart rate are major limitations of these drugs.

Generally, these patients will not hemodynamically tolerate the depth of volatile anesthesia required to breath spontaneously through an LMA or ETT. Therefore, if the airway must be secured, controlled ventilation with an ETT is usually chosen and combined with neuromuscular blockade and low-dose volatile anesthesia or some combination of ketamine/midazolam/fentanyl.

Options for induction are ketamine, etomidate, or fentanyl/midazolam. Critical points to remember are that the sicker the heart, the less it can tolerate hypoxia. Therefore, at all cost, the anesthesiologist must not lose the airway. The consequences of laryngospasm or airway obstruction, regardless of the type of anesthesia chosen, can be severe. Aggressive positive pressure ventilation will decrease preload and is poorly tolerated.

Last, ventricular dilation makes the heart prone to arrhythmia, which may occur spontaneously or in response to catheter manipulation. Aggressive diuresis may result in abnormal electrolytes (potassium, magnesium) and increase the propensity for arrhythmia. Defibrillation equipment must be immediately available, along with cardiac resuscitation medications, inotrope infusions, and antiarrhythmic drugs.

C. Pre-Glenn catheterization

Infants with single ventricle lesions require hemodynamic catheterization to assess suitability for the Glenn procedure. The Glenn operation, also termed superior cavo-pulmonary anastomosis, is generally done at 4 to 6 months of age. Assessment of PA pressures and PVR is mandatory prior to Glenn surgery. It is not possible to cover the complete anatomic and physiologic spectrum, but the children will fall into one of three categories.

1. Hypoplastic left heart syndrome (HLHS) with either a Blalock-Taussig shunt (BTS) or right ventricle-pulmonary artery connection (RV-PA) to provide pulmonary blood flow. The RV-PA is often referred to as a “Sano” connection. Although the subject remains controversial, the contemporary trend is in favor of the Sano connection over a BTS because of better outcomes perioperatively and in the period prior to Glenn surgery.

2. Single ventricle lesions, usually of left ventricular morphology (tricuspid atresia, double inlet LV, pulmonary atresia), with a BTS for pulmonary blood flow.

3. Single ventricle anatomy with intracardiac mixing and enough pulmonary stenosis to appropriately limit pulmonary blood flow. These infants did not require an operation in the newborn period.

Institutional practice varies widely in the involvement of anesthesiologists in these studies. Baseline oxygen saturations higher than 80% equate to a Qp:Qs >1, with the converse being true when the oxygen saturation is less than 80%. Those infants with BTS may be “outgrowing” their shunt and may have oxygen saturations of less than 70%.

Multiple anesthetic techniques can be used from ketamine/midazolam/fentanyl sedation and spontaneous ventilation with a natural airway to endotracheal intubation, positive pressure ventilation, and maintenance with volatile anesthetic. Ventricular function is usually good, providing the infant adequate physiological reserve.

The key factor with spontaneous ventilation is the maintenance of normocarbia without the need for significant oxygen supplementation. If the infant cannot ventilate adequately while sedated or anesthetized, it is best to control ventilation to ensure the collection of valid data. HLHS infants are only a few months removed from the enormous physiological stress of the Norwood operation. Frequent problems encountered are poor peripheral venous access, residual respiratory difficulties (phrenic nerve palsy, chronic aspiration from gastroesophageal reflux), airway problems (subglottic stenosis from prolonged intubation), and various other congenital noncardiac anomalies that accompany CHD.

To accurately measure PA pressure in the setting of a BTS, a catheter must be advanced into the pulmonary artery. A BTS is usually 3 or 3. 5 mm in diameter, and crossing the shunt with a catheter may temporarily restrict pulmonary blood flow, resulting in sudden desaturation. Injury or damage to the BTS from the catheter can result in a catastrophic reduction in pulmonary blood flow. If the problem cannot be rapidly corrected by withdrawal of the catheter, ECMO or emergency surgery may be required because it is unlikely that the cardiologist can restore pulmonary blood flow with a percutaneous repair. For this reason, some institutions will not advance a catheter across a BTS, preferring to measure a wedge pressure from a pulmonary vein as a surrogate for PA pressure. Because the RV-PA connection (“Sano”) is much larger, it is very rare to have a disruption of pulmonary blood flow as the catheter is advanced into the PA.

Knowledge of single ventricle physiology is enhanced if the clinician makes an attempt to understand the calculations done by the pediatric cardiologist. An introduction to these calculations was covered earlier:

Algorithm 4

Shunt fraction: Qp/Qs

The Qp:Qs may also be estimated from the following simple equation. Assumptions are FiO2 0.21 and normal Hgb:

Algorithm 5

After the Glenn operation, the “Glenn pressure” is the CVP measured in the SVC and is synonymous with PA pressure. High Glenn pressures lead to venous congestion of the head and neck.

Sample calculation:

Atrial pressure 6, PA pressure 13, measured Qp:Qs 1.1

Algorithm 6

The Qp:Qs, after the Glenn operation, is estimated to be 0.6, which reflects all upper body blood flow returning to lungs. The ratio is higher in an infant than in an adult because cerebral blood flow is proportionately much greater in infancy. The above calculation provides a rough approximation only.

D. Pre-Fontan catheterization

The same infants with single ventricle physiology and a prior Glenn operation will eventually require Fontan surgery as the final stage in single ventricle surgical palliation. The Fontan operation is usually done between 2 to 3 years of age. The Fontan consists of a connection between the inferior vena cava (IVC) and PA. This completes a total cavo-pulmonary anastomosis. The child now has separation of systemic and pulmonary blood flow and is no longer cyanotic. The price to be paid is completely passive pulmonary blood that mandates central venous hypertension as the driving force for pulmonary blood flow. Catheterization prior to the Fontan operation is necessary to measure hemodynamics and assess suitability for this unique physiological state.

Similar to the pre-Glenn catheterization, anesthesia can be administered in numerous ways from intravenous agents and spontaneous ventilation with a natural airway to endotracheal intubation and positive pressure ventilation with volatile anesthetics. No method is definitively better than another.

Regardless of the method chosen, children are required to be normocarbic, with an FiO2 <0.3 in order to accurately assess PA pressures and PVR. Preprocedure sedation is recommended, even in the setting of parental presence at induction. Intravenous access will be required in the internal jugular vein to measure Glenn pressures along with femoral access. Procedural complications are rare.

After the Fontan operation, central venous pressure measured in either the IVC or SVC, is equivalent to PA pressure, and is also referred to as Fontan pressure. The Fontan pressure is ideally the lowest possible pressure that produces sufficient pulmonary blood flow. When the Fontan pressure is high (>20-25 mm Hg), lymphatic return into the venous system is impaired and effusions develop (ascites, pleural, pericardial). To predict suitability for the Fontan operation, calculations similar to those done before the Glenn operation are required. With all IVC blood now directed into pulmonary vasculature, the expected post-Fontan Qp:Qs is 1.

Sample calculation:

Atrial pressure 6, Glenn (SVC) pressure 13, measured Qp:Qs 0.6

Algorithm 7

E. “Failing Fontan”

The “failing Fontan” is a descriptive term for an extremely sick group of patients. Their Fontan physiology is “failing” because of some combination of cyanosis, effort intolerance, arrhythmia, ventricular dysfunction, thrombosis, plastic bronchitis, or the poorly understood phenomenon of protein-losing enteropathy. Not all of these problems are amenable to intervention.

The purpose of catheterization is to assess the hemodynamic state and determine whether any therapeutic options exist. Cyanosis may be due to venoarterial collaterals that create a right-to-left shunt and could be coil occluded. Elevated PVR may require creation of a fenestration that functions as a small right-to-left shunt, preserving ventricular preload at the expense of systemic desaturation. Relief of mechanical obstructions in the Fontan pathway with dilation and stenting may be another therapeutic option.

Physiologically, these patients are fragile. Those with an older version known as the “atriopulmonary Fontan” have an anastomosis between the right atrium and PA. (The modern, total cavopulmonary anastomosis consists of bidirectional Glenn and either an extracardiac or lateral tunnel Fontan. The entire spectrum of failing Fontan physiology can occur in any patient with Fontan circulation, although the incidence is lower with extracardiac or lateral tunnel operations.) Over time, the atrio-pulmonary Fontan leads to a very dilated atrium in which blood swirls slowly before it enters the pulmonary circulation. The dilated atrium provides little, if any, propulsive force for pulmonary blood flow, but it is a ready source of thrombus. Atrial arrhythmias are common, including sinus node dysfunction. Induction with intravenous agents is prolonged. If the underlying ventricular function is poor, standard doses of propofol or volatile anesthesia may not be tolerated.

Atrial tachyarrhythmias and sinus node dysfunction require the ability to perform pharmacologic or electrical cardioversion, as well as transcutaneous pacing. The two key variables in the anesthetic plan are the mode of ventilation and the age of the patient. Fontan physiology is improved with spontaneous ventilation because negative intrathoracic pressure augments venous return and enhances pulmonary blood flow and cardiac output. Therefore, the best representation of the patient’s true physiologic state is catheterization performed with spontaneous ventilation. The crucial caveat is that if spontaneous ventilation occurs at a depth of anesthesia that results in significant hypercarbia, its benefit may be entirely lost because of an increase in PVR.

“Failing Fontan” circulation usually takes years to develop, with the consequence that most patients are teenagers or young adults who can tolerate the procedure with light sedation. Younger children who require general anesthesia provide more of a challenge.

Options with a natural airway are intermittent dosing with a combination of ketamine/midazolam/fentanyl and/or a propofol infusion. The more drugs with respiratory depressant properties used, the greater the risk of hypoventilation. Spontaneous ventilation using a laryngeal mask airway (LMA) and volatile anesthesia requires a deeper plane of anesthesia but usually results in adequate respiratory drive. Spontaneous ventilation through an endotracheal tube is possible but requires an even greater depth of anesthesia such that at least a small degree of hypoventilation and hypercarbia will likely occur. The lack of respiratory depression with dexmedetomidine makes it theoretically attractive but its use is limited by concerns about the propensity for bradycardia in patients with potential sinus node dysfunction.

F. Post-transplant catheterization

A regular schedule of cardiac catheterizations is required after heart transplant. Once the early post-transplant period has passed, the procedure is performed every 6 to 12 months. The guidelines are under frequent review and are subject to change. A left heart catheterization is done to demonstrate coronary artery anatomy and a right heart study is required for a myocardial biopsy to assess potential rejection. Perforation of the RV with cardiac tamponade can occur but is very rare. Full hemodynamic calculations are usually not required and the procedure is relatively brief.

Most children have a structurally normal heart with good ventricular function. Techniques range from deep sedation with a natural airway to endotracheal intubation and volatile anesthesia. No technique is clearly superior to another. The more difficult scenario is a child with a clinical suspicion of rejection because of symptoms of heart failure, who requires hemodynamic catheterization and endomyocardial biopsy. These children are already in hospitals and most often have intravenous access established. They will have had a recent echocardiogram that provides an assessment of ventricular function. When there is significant ventricular dysfunction, the same concerns outlined in the subsection “Left ventricular failure” apply.

G. Pulmonary hypertension

This is the highest-risk subset of children with CHD and has been recently well reviewed. When presenting for hemodynamic catheterization, children have either idiopathic PHTN or PHTN secondary to left-to-right cardiac shunts. An estimate of the degree of PHTN can be gained echocardiographically and should be classed as subsystemic, systemic, or suprasystemic.

Children with idiopathic PHTN require a diagnostic study to determine the response to pulmonary vasodilators to guide future therapy. The vasodilators used are oxygen and inhaled nitric oxide (iNO). In children with PHTN secondary to CHD, the findings of the diagnostic catheterization are needed to determine the appropriateness and timing of future surgery.

The most feared complication is a pulmonary hypertensive crisis, resulting in severe hypoxemia and right heart failure. Predictably, patients with suprasystemic PHTN are at highest risk. In a study involving a large review of 145 patients from a single center, there were 6 pulmonary hypertensive crises, which included two deaths from intractable hypoxemia and right heart failure. Their case series consisted of cardiac catheterizations and noncardiac surgery, but all pulmonary hypertensive events only occurred during hemodynamic studies. This suggests that the provocative maneuvers of limiting the fraction of inspired oxygen (FiO2) may be harmful.

The anesthetic principles for PHTN are to control the physiological variables that affect PVR and to prevent rises in PVR from potent stimuli. The physiological abnormalities that elevate PVR are acidosis, hypercarbia, hypoxia, hypothermia, and significant atelectasis. Hypercarbia, if accompanied by an adequate compensatory metabolic response to prevent acidosis, does not significantly elevate PVR. However, most children do not retain CO2 at baseline and hypercarbia causes acidosis. The most potent stimuli are laryngoscopy and endotracheal intubation. This raises the question of why endotracheal intubation is the most common form of airway management. There does not appear to be a good reason other than that most anesthesiologists seem nervous with anything less than a fully secured airway in the child with PHTN.

While it may be difficult to administer iNO and different levels of FiO2 through a tight fitting facemask, a LMA provides an excellent airway with the ability to vary FiO2 and deliver iNO. PVR is lowest when patients ventilate to normal functional residual capacity (FRC). However, the preservation of FRC does not mandate intubation and positive pressure ventilation. A physiological amount of positive end expiratory pressure (PEEP) can be added to spontaneous ventilation with an LMA with excellent preservation of lung volumes. Positive pressure ventilation can also be delivered through an LMA with effective tidal volume and low peak pressures. Spontaneous ventilation augments venous return when compared to positive pressure ventilation. This is important for the hypertrophic right ventricle, which is preload-dependent.

Last, at the end of the case, an LMA can be smoothly removed at a light plane of anesthesia, avoiding the elevations in PVR secondary to coughing and straining on the endotracheal tube. In summary, an LMA is an underused method of airway management in patients with PHTN. It offers the following benefits: avoidance of stimulating laryngoscopy, a smooth emergence, the ability to alter FiO2 and administer iNO, spontaneous ventilation at a depth of anesthesia that usually prevents hypercarbia, preservation of right ventricular preload, and the ability to deliver positive pressure ventilation if necessary. What is most critical for children with PHTN is meticulous airway management.

Based on preoperative echocardiography, the severity of PHTN should have been estimated as subsystemic, systemic, or suprasystemic. General experience has found that when pulmonary artery (PA) pressures are less than 50% systemic, children are low risk and tolerate anesthesia well. As PA pressures approach systemic, risk increases considerably.

No ideal drug for PHTN exists. Most commonly used are subanesthetic doses of multiple drugs in the attempt to minimize the hemodynamic and respiratory effects of one drug given in full anesthetic dose. The crucial cardiovascular determinant is the status of the right ventricle. Therefore, a recent echocardiogram must have been done and reviewed. In the setting of PHTN, the right ventricle develops compensatory hypertrophy with preserved function. Only at the end stage of PHTN does the right ventricle dilate and fail. Thus, most children with PHTN will have a preload-dependent, stiff, and noncompliant right ventricle with relatively preserved function.

As long as venous return can be supported with intravenous fluids, children will tolerate a cautious titrated induction of anesthesia with any combination of the available agents. The preceding statement only precludes a mask induction with deep levels of volatile anesthesia, followed by placement of an intravenous catheter.

In young children in whom intravenous access is presumed to be difficult, intramuscular ketamine is the preferred option, followed by securing intravenous access. A frequent question regarding ketamine is whether its sympathomimetic effect will exacerbate pulmonary hypertension. It appears that as long as hypercarbia and hypoxia are avoided, ketamine does not have deleterious effects on PVR.

A management plan for a pulmonary hypertensive crisis must be formed prior to the induction of anesthesia. The best prevention for trouble is skilled airway management. Laryngospasm, breath holding, airway obstruction, and bronchospasm are all potentially lethal events in children with PHTN.

As mentioned earlier, skilled airway management is not synonymous with endotracheal intubation. One must remember that children with PHTN can become hypoxic for all the same reasons that healthy children become hypoxic. Therefore, before frantically calling for iNO, ensure that there is an appropriate FiO2, confirm that the circuit or endotracheal tube is not kinked or blocked with secretions, make sure that the LMA has not become dislodged, and check that the carbon dioxide absorbent is not depleted.

If a pulmonary hypertensive crisis develops, the management principles are to reverse the inciting event should one be identified, administer pulmonary vasodilators, and support right heart function. In the catheterization laboratory, the inciting event may be something as simple as allowing the child to breath room air while hemodynamic parameters are assessed. If there is an intracardiac shunt, at least initially, systemic blood pressure will be maintained, but profound cyanosis may occur.

If the heart is structurally normal, there will be both hypotension and cyanosis. The child should immediately be switched to enriched oxygen (FiO2 1.0) and ventilation controlled. Depending on the rapidity at which the pulmonary hypertensive crisis develops, options are intravenous milrinone or iNO. Both should be combined with epinephrine to support right heart function. Rapid echocardiographic assessment will reveal the intracardiac effects (septal bowing to the left, pulmonary insufficiency, tricuspid regurgitation) of the sudden rise in PA pressure.

Epinephrine will usually offset systemic vasodilation induced by milrinone. Initially, epinephrine can be given through a peripheral intravenous in small bolus doses (1 mcg/kg). If necessary, an infusion can be started though a central catheter already in place for the procedure. Unless the pulmonary hypertensive crisis is due to a reversible respiratory event (laryngospasm, airway obstruction) it usually does not completely resolve. Therefore, children remain intubated with ventilation controlled, FiO2 1.0, and iNO. Vasoactive support is usually required. When all standard resuscitative measures fail, ECMO is an option of last resort, although the chance of decannulation with good neurologic outcome in the setting of a pulmonary hypertensive crisis is unknown.

3. Interventional procedures

While not universal, pediatric cardiologists are more likely to request anesthesia for interventional procedures because these have more potential for hemodynamic disturbance and complications. Children range from those with well-compensated conditions to those who are critically ill. Given the possible hemodynamic disturbance, many anesthesiologists opt to secure the airway, although for most procedures, this is not strictly necessary. If one thinks along the lines of the “ABCs” of trauma resuscitation, when the airway is secured with an endotracheal tube, “A” and “B” are taken care of, in case the circulation or “C” deteriorates.

A. Pulmonary stenosis

Balloon valvuloplasty for valvar pulmonary stenosis (PS) was the first intervention technique described for CHD and now is a frequent procedure. Children usually maintain good functional status with PS. The echocardiogram will determine the gradient across the pulmonary valve. Intervention is indicated when the gradient across the valve exceeds 50 mm Hg. Compensatory right ventricular hypertrophy is common, with frank right heart failure being rare. The one exception is presentation in the newborn period with critical PS and insufficient pulmonary blood flow. Shunting will occur from right to left across the atrial septum (ASD or PFO). Prostaglandin (PGE1) is administered to maintain ductal patency and provide adequate pulmonary blood flow.

The majority of these infants will be intubated and mechanically ventilated. They have functionally single ventricle physiology even though, anatomically, the right ventricle is sufficiently developed to justify pulmonary valvuloplasty.

Induction: Children with good functional status and no echocardiographic evidence of significant right ventricular dysfunction will tolerate a standard mask induction with volatile anesthesia. Premedication may be given, depending on the age of the child. The high-risk subset is newborns with critical PS. As mentioned earlier, the majority of these infants will be intubated and mechanically ventilated. Anesthetic management consists of neuromuscular blockade and low-dose volatile anesthetic or continuation of the intravenous sedation (usually fentanyl and midazolam) already being used.

Occasionally, a child or infant presents with worrisome signs (right heart failure, cyanosis, failure to thrive). These children are best induced intravenously or with intramuscular ketamine if intravenous access cannot be secured. Any combination of the available agents can be used in a titrated and cautious intravenous induction. Muscle relaxants are given to control ventilation at a lighter plane of anesthesia. Forceful positive pressure should be avoided because of its negative effect on venous return in the setting of a preload dependent right ventricle. Continued neuromuscular blockade and low dose volatile anesthesia is usually used for maintenance.

Airway: Although not mandatory, it is my practice to intubate these children. Trans-esophageal echocardiography (TEE) is frequently used, which necessitates intubation.

Procedure: Relatively large access sheaths are used, which may cause vascular injury. Arrhythmia may occur as catheters are moved through the right ventricle. Right ventricular injury or perforation will result in tamponade and sudden hemodynamic deterioration. The diagnosis can be confirmed with TEE, surface echocardiography, or fluoroscopy, which may demonstrate a globular heart.

When the balloon is inflated across the pulmonary valve, there will be no forward output from the right ventricle and greatly reduced left ventricular preload. Blood pressure will decrease abruptly, but it is usually transient and recovers promptly when the balloon is deflated. A reduction in right ventricular pressure to less than 50% systemic is the goal.

More aggressive balloon valvuloplasty to eliminate the gradient completely creates a risk of severe pulmonary insufficiency. While not ideal, severe pulmonary insufficiency is generally well tolerated although there will likely be some immediate right ventricular dysfunction from the sudden volume overload.

In newborn critical PS with ductal patency, there will be maintenance of pulmonary blood flow and left ventricular preload during balloon dilation. Hemodynamic disturbance during balloon dilation is rare. However, even with good relief of PS, the right ventricle needs time to remodel and functional normally. Consequently, the neonate should be left intubated. Later in the ICU, if echocardiography confirms good antegrade pulmonary blood flow, a trial stoppage of PGE1 can be done.

B. Aortic stenosis

Balloon valvuloplasty for aortic stenosis (AS) is an infrequent intervention but one that gets the attention of the anesthesiologist because of its high risk nature. The procedure is palliative. The hope is that the stenosis can be relieved enough to defer surgery until late childhood or early adulthood. Surgical options are valve repair, valve replacement (mechanical valve with life-long anticoagulation), or Ross operation. The procedure is done when the gradient exceeds 70 mmHg or 50 mmHg, with symptoms or EKG ischemic findings.

Neonates with critical AS may have lower gradients because of poor left ventricular output. Children with AS are at risk for heart failure, arrhythmia, ischemia, and sudden death. Only two factors affect the gradient across the valve: left ventricular function and the severity of the aortic stenosis. Both of these factors would have been assessed prior to the procedure using echocardiography. A high gradient provides reassurance that the left ventricular function is good. Most concerning are children with low gradients, which signifies decreased left ventricular function in the setting of significant aortic stenosis.

The anesthetic principles for AS in children are not different from those in adults: maintenance of preload (sinus rhythm, normal heart rate appropriate for age, fluid bolus to offset anesthetic induced venodilation, and decreased venous return from positive pressure ventilation), preservation of systemic vascular resistance (SVR) to ensure sufficient coronary perfusion pressure, and avoidance of negative inotropy.

Induction:Induction should be done intravenously. Intravenous access allows a cautious, titrated induction and the ability to support intravascular volume with a fluid bolus. If the ventricular function is good, induction can be safely accomplished with any of the available agents. Muscle relaxants are given and ventilation is controlled. Care should be taken to avoid large decreases in preload with positive pressure ventilation.

When intravenous access is not achievable prior to induction, intramuscular ketamine is the best option. Children with poor ventricular function demand extra vigilance. Traditional teaching in this scenario recommends ketamine or pure synthetic opioid such as fentanyl or sufentanil. While synthetic opioids are free of negative inotropy, they do produce cardiovascular depression; it is a myth that they do not.

Intense narcotization decreases central sympathetic nervous system output and can result in hypotension even in the absence of myocardial depression. For the highest-risk children, I prefer ketamine; for those with good left ventricular function, I induce cautiously with small doses of propofol and volatile anesthesia as tolerated. Because the procedure is not painful or stimulating, the children can usually only maintain an acceptable blood pressure when there is a light plane of anesthesia combined with neuromuscular blockade and controlled ventilation.

Airway: The likely possibility of hemodynamic disturbance during the procedure makes intubation the safest option. TEE is often required.

Procedure: The stressed left ventricle is prone to catheter-induced arrhythmias, which may be poorly tolerated. The catheter should be immediately withdrawn. If the arrhythmia does not promptly resolve, consideration should be given to electrical cardioversion. An antegrade approach with access to the left ventricle through an atrial septal puncture often results in less hemodynamic disturbance. Major vascular injury or cardiac tamponade are rare but real possibilities.

During balloon dilation, there will be no forward cardiac output and no coronary perfusion because the coronary ostia are occluded by the balloon. In some centers, the heart is brought to a standstill with overdrive pacing prior to balloon dilation. These measures all stress the heart considerably. It may be necessary to keep the blood pressure in the high normal range (fluid boluses, light anesthesia, calcium bolus) prior to balloon dilation and overdrive pacing. Starting with a higher blood pressure attenuates the degree of hypotension that occurs during balloon dilation and more readily reestablishes coronary perfusion once the balloon is deflated. After balloon valvuloplasty, three scenarios may occur.

a. Technically good result with reduction in gradient and minimal aortic insufficiency (AI)

b. Inadequate reduction in AS

c. Correction of AS but significant AI

Children having a good relief of AS may in fact be hypertensive at the conclusion of the procedure because the gradient across the aortic valve has been reduced. Those with an insignificant reduction in AS need vigilance during emergence and extubation. They are not good candidates for deep extubation because of the hypotension generally associated with that technique.

The most dangerous scenario occurs when there is significant AI after the procedure. The previously pressure-loaded left ventricle with hypertrophy and poor compliance now becomes a volume-loaded chamber. The propensity for pulmonary edema is high. The children require observation in a monitored setting and echocardiography to assess the left ventricular tolerance of the sudden change in physiology from AS to AI. Depending on the degree of AI created, it may be prudent to leave the child intubated because of high likelihood of significant pulmonary edema.

C. Coarctation

Balloon angioplasty, with or without stenting for coarctation, can be done as a primary intervention or for re-stenosis occurring after surgical repair.

Neonatal coarctation is best treated surgically because the rates of re-stenosis or aneurysm are lower compared with balloon angioplasty. Older children with native coarctation are better candidates for balloon angioplasty, although this is not universally accepted.

The most common indication is for re-stenosis after surgery because further operations in this scenario carry considerable risk. In the presence of normal cardiovascular anatomy, the procedure is straightforward and the children usually healthy. One unique scenario is that children with hypoplastic left heart syndrome after the Norwood operation may develop coarctation at the site of the distal aortic arch reconstruction. Procedural concerns are the same, but obviously these children come with the limitations of single ventricle physiology.

Induction: Unless the child has severe cardiovascular compromise, a standard mask induction followed by intravenous access is usual. This includes single ventricle children, assuming they have good ventricular function. Premedication is given as needed.

Airway: Given the rare but possible occurrence of hemodynamically significant vascular injury, intubation is usually preferred.

Procedure: A relatively large vascular access sheath is needed, which may result in femoral arterial injury. Injury to the aorta may occur at the time of balloon angioplasty. The injury may be either catastrophic rupture or dissection. The risk of rupture is higher when there is aneurysmal aorta around the coarctation. A significant aneurysm is a relative contraindication to balloon angioplasty. Dissection may be treated by deploying a stent at the site of the intimal flap. Coarctations of long-standing duration have extensive collateralization that minimizes hemodynamic disturbance during balloon dilation.

Hypertension proximal to the coarctation will occur during balloon dilation. Maintaining blood pressure in the low normal range prior to angioplasty is prudent. After an uncomplicated procedure, children are extubated. Target goals for blood pressure should be discussed with the cardiologist because almost all children will have some degree of hypertension.

My preference for pharmacologic treatment of blood pressure is nicardipine given by infusion. The dose is usually 0.5 to 2 mcg/kg/min. The onset is gradual, without the precipitous swings in blood pressure associated with sodium nitroprusside. Blood pressure monitoring using a noninvasive cuff is usually sufficient.

D. Branch pulmonary artery stenosis

Children with CHD characterized by decreased pulmonary blood flow often have branch PA stenosis. Stenosis may also occur at the site of previous surgery involving the branch PAs, such as a BTS.

Balloon angioplasty, with or without stenting, is the preferred method of treatment. Immediate results are good, but long-term re-stenosis frequently occurs. It is impossible to generalize about the cardiovascular status of children with branch PA stenosis because their underlying CHD is so diverse. Obviously, an individualized management plan is required, but most children have adequate physiologic reserve to tolerate a standard anesthetic. The key preprocedure investigation is echocardiography to determine the status of the right ventricle in the setting of increased afterload from branch PA stenosis.

The high-risk subgroup is children with Williams syndrome, a condition marked by supravalvar AS, branch PA stenosis, and coronary stenoses. A recent review describes the challenge of these children. There is a clear risk of perioperative death, which is believed to be due to coronary ischemia. The adult corollary of these children would be a patient with good ventricular function, AS, and significant coronary artery disease (CAD).

The physiological challenge is delivering an anesthetic that combines the management principles of AS while optimizing the balance of myocardial oxygen supply and demand.

Aortic stenosis: maintain preload, preserve afterload to maintain coronary perfusion pressure, avoid negative inotropes

Coronary artery disease:

Increase myocardial oxygen supply: relative bradycardia, adequate oxygen content of the blood, maintain coronary perfusion pressure

Decrease myocardial oxygen demand: relative bradycardia, decrease wall stress, decrease contractility

Induction: In the absence of Williams syndrome, children with good physiologic reserve can be premedicated if appropriate and will tolerate a mask induction with intravenous access achieved after induction. Children with Williams syndrome present a clinical dilemma. These children do not routinely get coronary angiograms so the nature of their CAD is unknown, but it can be presumed to be significant. The severity of the supravalvar AS and branch PA stenosis should be known from the most recent echocardiogram.

The key physiological goals are maintenance of blood pressure and heart rate control. An intravenous induction is preferred because it offers the ability to titrate anesthetic drugs, control heart rate with beta blocker, and support intravascular volume with fluid.

Multiple attempts at securing intravenous access in a small child with poor peripheral veins may do more harm than good. The child may become very irritable and tachycardic and while future intravenous sites are being lost due to failed attempts. In this scenario, there are two unattractive possibilities: IM ketamine or mask induction with volatile anesthesia and a placement of an intravenous at the earliest opportunity. Ketamine will maintain blood pressure but may induce tachycardia, whereas volatile anesthesia in the absence of intravenous access may result in considerable hypotension.

Airway: The procedure carries the risk of vessel injury and respiratory complications. It is usual to intubate these children.

Procedure: Pulmonary blood flow, already compromised, may deteriorate because of large catheters traversing the right heart and being introduced into the pulmonary arteries. Communication with the cardiologist is essential. Balloon angioplasty may result in vessel trauma with subsequent hemoptysis. Balloon dilation may induce coughing at a crucial period in the procedure. Unilateral pulmonary edema may occur because of a sudden increase in blood flow to a chronically underperfused lung. Barring serious procedural complications, children are extubated after the procedure. Children with Williams syndrome must be admitted overnight for observation to a monitored setting because of their risk for sudden cardiac death. The etiology is believed to be an ischemia-induced arrhythmia.

E. Right ventricle to pulmonary artery stenosis

This condition results from stenosis of a surgically created connection between the right venticle and pulmonary artery (RV-PA). It occurs in the following scenarios:

Tetralogy of Fallot. A RV-PA connection is required when coronary artery anatomy precludes a trans-annular patch repair

Ross procedure

Truncus arteriosus repair

Hypoplastic left heart syndrome (HLHS) after the Norwood operation with a RV-PA connection known as a “Sano”.

In the first three conditions, the surgeon creates a connection between the right ventricle and the pulmonary artery using homograft tissue. The valve in the homograft tissue will eventually fail, leading to free pulmonary insufficiency. The homograft may also calcify over time, particularly at the anastomotic sites leading to stenosis. Assuming regular follow up of the child, a gradient across the RV-PA connection can be quantified echocardiographically.

In HLHS, a nonvalved synthetic conduit is placed between the right ventricle and the pulmonary artery. This arrangement only lasts until it is taken down during Glenn surgery at approximately 4 to 6 months of age. Prior to Glenn surgery, intervention may be required because of RV-PA stenosis. In the setting of single ventricle physiology, the obstruction to pulmonary blood flow presents as an oxygen saturation lower than expected. Children with HLHS receiving anesthesia during the period between the Norwood and Glenn procedures are one of the highest risk subsets in CHD.

Induction: The two ventricle children have the functional equivalent of pulmonary stenosis that is usually well tolerated clinically. Therefore, anesthesia can be safely induced by a variety of methods. Children with HLHS and RV-PA stenosis are less than 6 months old, the highest risk age for intervention in single ventricle children. However a standard mask induction remains well tolerated in most circumstances.

Intravenous access must be obtained at the earliest possible moment. These single ventricle children have a surgically created “double outlet right ventricle”. With obstruction to pulmonary blood flow because of the RV-PA stenosis and excessive systemic vasodilation from deep volatile anesthesia, a situation may occur where the majority of ventricular output is ejected systemically.

“Balancing the circulation” after the Norwood operation requires maintaining the correct relationship between PVR and SVR. The anatomic stenosis of the RV-PA conduit creates the physiologic equivalence of a high relatively fixed PVR. To prevent the ventricle from choosing the path of least resistance and ejecting the majority of its output systemically, significant decreases in SVR must be avoided.

Airway: Intubation is usually preferred because of the cardiorespiratory disruptions that can occur during balloon inflation.

Procedure: A relatively large catheter is required to dilate the stenosis and deploy the stent. Vascular injury at the access site is possible. The right ventricle will have compensatory hypertrophy and possibly some degree of dilation with decreased function if the RV-PA stenosis is longstanding. Catheters and wires passed through an irritable right ventricle may induce arrhythmias. These usually terminate spontaneously when the catheter is withdrawn. During the time of balloon dilation, pulmonary blood flow will cease.

Two ventricle children may also become hypotensive because there is no left ventricular preload. Prompt resolution upon balloon deflation is usual. Single ventricle children, paradoxically, may tolerate balloon dilation better because they can continue to eject blood systemically. The period of balloon dilation is usually brief and well tolerated, but if it is prolonged or if the right ventricle becomes significantly distended and dilated, inotropic support may be necessary. Bolus epinephrine and calcium are the preferred choices. A volume bolus should not be give because the right ventricle will already be distended.

F. Patent ductus arteriosus (PDA) coil occlusion

Coil occlusion of PDA is usually a straightforward short procedure. The key determinant of the child’s physiologic state is the duration of the PDA and the degree of PHTN that has ensued.

In a structurally normal heart, the degree of PHTN can be estimated from the tricuspid regurgitant jet velocity (4 x velocity squared added to the estimated central venous pressure). It is rare to encounter a child with a long-standing undiagnosed PDA. Most children will have some degree of well-tolerated PHTN without elevations in PVR because the higher pulmonary blood flow creates correspondingly high PA pressures. Reducing pulmonary blood flow by coil occlusion of the PDA removes the volume overload, allowing the pulmonary vasculature to remodel over time.

Children who are not fully saturated breathing room air should be suspected of having near systemic PA pressures with bidirectional shunting through the PDA. The anesthetic principles of PHTN previously discussed should guide management.

Rare complications are embolization of the coils into the main or branch PAs. The degree of obstruction to pulmonary blood flow after embolization of the coils determines the urgency of the need for surgery. Temporizing measures are ventilation with FiO2 1.0, and support of the right heart and systemic circulation.

Epinephrine is the drug of choice because of its inotropic and vasoconstrictive properties. Pulmonary vasodilators, such as milrinone, have no role because the obstruction to pulmonary blood flow is mechanical. As long as systemic perfusion is maintained, moderate or even severe cyanosis can be tolerated long enough to rapidly transfer to the operating room.

In the setting of cardiovascular collapse, unresponsive to standard resuscitative measures, the only recourse is ECMO.

G. Patent ductus arteriosus stenting

Single ventricle lesions in the newborn period almost always require a PDA and have either ductal-dependent systemic or pulmonary blood flow. Tricuspid atresia and pulmonary atresia are two of the classic right-sided, single ventricle lesions with ductal-dependent pulmonary blood flow.

The PDA is temporarily maintained with a PGE1 infusion, but the neonate requires a more secure source of pulmonary blood flow and cessation of PGE1. This is either a surgical BTS or stenting of the PDA. Institutional practice can vary widely in this regard. The echocardiogram should determine whether PDA stenting is a viable option, based on the anatomic relationship between the aorta and PA.

Pulmonary blood flow requires left-to-right flow through the PDA. If breathing room air or an FiO2 of less than 0.3, a quick estimation of the Qp:Qs can be made from the systemic oxygen saturation. Assuming a normal newborn hemoglobin and normal oxygen consumption (30-40%), when the SpO2 is higher than 80%, the Qp:Qs is greater than 1.

An SpO2 above 90% suggests a Qp:Qs of at least 3:1. The price to be paid for this high SpO2 is greatly increased myocardial work. If the Qp:Qs is 3:1, the heart must work three times as hard to meet the systemic oxygen requirements because three times as much blood is being pumped to lungs compared to the body. In the short term, this increased myocardial work is tolerated but leaves the heart with little physiologic reserve.

When the SpO2 is less than 80%, the Qp:Qs is less than 1. With a normal neonatal hemoglobin of 15 to 18 gm/dL, systemic oxygen delivery is easily maintained. Oxygen delivery may suffer when there is marked cyanosis (less than 60%), but myocardial work remains normal. Given the high oxygen extraction ratio of the heart (50-70%), when the systemic oxygen saturation is very low, the myocardial oxygen supply/demand balance becomes very tenuous.

The PDA is effectively maintained by low-dose PGE1 (0.01-0.03 mcg/kg/min). Higher doses cause apnea and some neonates will be intubated for this reason. When already intubated, care is relatively straightforward, consisting of continued ventilation with sedative infusions usually supplemented by neuromuscular blockade. After successful PDA stenting, the PGE1 infusion can be stopped but the neonate is left intubated until the apneic effects of PGE1 have fully resolved.

The neonate with a natural airway requires intubation. With intravenous access already established, the neonate can be induced with any of the available anesthetics or low-dose volatile anesthesia combined with neuromuscular blockade. The stent will be deployed over an inflated balloon. During inflation, pulmonary blood flow will cease but returns promptly upon balloon deflation.

The neonate’s physiology does not change after PDA stenting. Flow through the stented ductus remains dependent on adequate systemic perfusion pressure. Vessel injury is possible, which may be temporized by reinflation of the balloon as long as it does not completely occlude all pulmonary blood flow.

Embolization of the stent may occur into the pulmonary vasculature or systemically. Systemic embolization will likely require surgical retrieval but not emergently because the stent would migrate distally into the thoracic or abdominal aorta. (If embolization occurs into the pulmonary vasculature, management is described in the section on PDA coil embolization.)

H. Balloon atrial septostomy for d-transposition

Transposition of the great arteries (d-TGA) as an isolated condition results in the pulmonary and systemic circulations occurring in parallel rather than in series. Deoxygenated venous blood is delivered systemically from the right ventricle to the aorta and oxygenated pulmonary venous is returned to the pulmonary artery by the left ventricle. For this arrangement to be compatible with life, there must a site or sites where blood from the two parallel circulations can mix. Sites of mixing are either PDA, ASD, or VSD. A sufficient ASD or VSD are the most stable locations for mixing. A PDA is not always a reliable site for mixing. Flow through the PDA must be bidirectional to achieve adequate mixing.

In the setting of d-TGA with inadequate mixing, the newborn is critically ill with deeply compromised oxygen delivery, most importantly to the heart and brain. Although the PDA is not providing a site of sufficient mixing, the small amount of bidirectional shunting may be the only the thing keeping the newborn alive. Therefore, PGE1 to maintain some ductal patency is a vital temporizing measure. Although not the first percutaneous intervention, balloon atrial septostomy (BAS) ushered in the era of interventional procedures for CHD.

The BAS may be done at the bedside under echocardiographic guidance or in the catheterization laboratory. Institutional practice varies. When brought to the catheterization laboratory, the newborn may be accompanied by the intensive care staff or an anesthesiologist. Once again, institutional practice varies. One might ask: What can be done when the newborn in intubated, paralyzed, and sedated? Although the situation is critical, it is not futile.

Optimization of oxygen delivery is the goal that involves increasing supply and decreasing demand. The newborn is sedated with full neuromuscular blockade to decrease oxygen consumption. Ventilation with FiO2 allows for the highest dissolved fraction of oxygen. The severe hypoxia induces a progressive cardiovascular deterioration, which is the scenario the clinician must avoid. Hypoperfusion combined with severe hypoxia is lethal. Therefore, support of the circulation with cautious fluid boluses (usually albumin) and vasopressor support (usually dopamine) are required.

In a severely acidotic environment, cardiovascular function will deteriorate. Sodium bicarbonate should be given. The newborn normally has a hematocrit well over 40%, but if for any reason, the hematocrit is below this level, transfusion is justified to increase oxygen carrying capacity.

I. Septal device closure

Closure of a secundum ASD with a percutaneous device is now common procedure. The procedure is limited to small and moderate-sized secundum ASDs because of the need for an adequate rim of surrounding atrial tissue to “land” the device. These children usually have well compensated pulmonary overcirculation. In fact, even untreated, an isolated secundum ASD rarely gives rise to significantly elevated PVR until well into adulthood. Thus, the anesthetic is straightforward. Intubation is required because of the need for TEE. Complications are rare.

Two situations that bear noting are embolization of device and aortic insufficiency. Management of device embolization depends on the severity and location of the embolization. Specific management strategies obviously depend on the nature and location of the embolization. Good communication with the cardiologist is essential. Aortic insufficiency is caused by mechanical impingement of the device on the aortic annulus. To guard against this, a trial deployment is done and TEE used to assess any degree of aortic insufficiency.

Muscular VSDs are also suitable for device closure, but this is a rare event, as most of these lesions close spontaneously. (The Food and Drug Administration has not approved device closure of perimembranous VSDs because of concern over the late development of heart block.) The anesthetic principles of managing compensated pulmonary overcirculation apply. The degree of pulmonary overcirculation is usually higher with a VSD, and the child may require treatment for heart failure. However, unlike adults, the etiology of the heart failure pulmonary overcirculation rather ventricular dysfunction or valvular heart disease, and thus anesthesia, is generally well tolerated. Positioning the deployment catheters within the confines of the right ventricle is technically challenging. The procedure can be long and may involve hemodynamic instability due to arrhythmia, heart block, tricuspid or mitral regurgitation, and blood loss.

J. Percutaneous pulmonary valve replacement

Free pulmonary insufficiency (PI) is a consequence of most TOF repairs. Once thought to be physiologically benign, the condition is now known to lead to progressive RV dilation, dysfunction, and eventual failure. If accompanied by significant tricuspid regurgitation (TR), the volume overload can be severe, with an accelerated progression to RV failure.

Patients are usually adults because in the absence of severe TR, the timeframe for the development f RV dysfunction is many years. Pulmonary valve interventions are now possible across a wide spectrum of age and disease. At this time, a percutaneous valve can only be deployed if the patient has RV-PA conduit. Those patients with a transannular patch repair develop an aneurysmal RVOT in which it is not technically possible to deploy the valve.

It is rare for patients to have frank RV failure. A standard induction is well tolerated. Intubation is usually required because of frequent use of TEE. Free PI leads to a chronically volume-loaded RV. Anesthesia-induced venodilation combined with preload reduction due to positive pressure ventilation may require cautious fluid boluses and a reduced depth of anesthesia. The valve is deployed over an inflated balloon. During balloon inflation, there is no RV output and, therefore, no LV preload. The period is usually brief, with prompt recovery. However, if prolonged or occurring in the setting of a compromised RV, the ventricle may rapidly dilate and fail.

Similar to balloon dilation of severe PS, the treatment is not a fluid bolus because the RV is already distended. Inotropic support is needed, with epinephrine being the drug of choice. It is advisable to have the blood pressure in the high normal range to ensure adequate coronary perfusion pressure during the period of balloon dilation when there is reduced LV output.

K. Electrophysiological (EP) studies

It is difficult to make generalizations about anesthetic management for children having EP studies. The most common diagnoses are teenagers with preexcitation (Wolff-Parkinson White) or atrioventricular nodal reentry tachycardia (AVNRT). Both of these occur in the setting of a structurally normal heart. There are also children with CHD who may develop arrhythmia and require EPS. Given the diversity of CHD diagnoses, management is guided by the specific lesion and physiologic reserve.

Ideally, EP mapping and potential ablation would occur using only light or conscious sedation. This would prevent suppression of the arrhythmia by anesthetic agents. However, the procedures are long, the X-ray table uncomfortable, and the induced tachycardia unpleasant, at best. Thus, anesthesia or deep sedation is the rule for all children and mature teenagers. While most anesthetics do not have a direct antiarrhythmic effect, arrhythmias may be more difficult to induce under general anesthesia because of suppression of accessory pathways, the AV node, or an ectopic foci.

Data is sparse on the effects of anesthetics in suppressing arrhythmia. Propofol appears safe as do volatile agents at less than 1 MAC, when comparing their effects on the AV node or reentry pathways. Increased automaticity of an ectopic focus is an electrophysiologically distinct entity from WPW or AVNRT, and there is a possibility that propofol should be avoided. The one drug that would seem ideal for sedation with maintenance of a natural airway is dexmedetomidine, but its negative effects on the SA and AV node seem to preclude its use for EPS.

A standard induction is well tolerated because most children have a structurally normal heart; however, there are a few key points to remember. A reduced depth of anesthesia causes the least potential suppression of the arrhythmia and may be requested by the electrophysiologist. In this scenario, neuromuscular blockade may be required to ensure an immobile patient. I prefer volatile agents because the anesthetic depth can be quantified.

The MAC-awake level is usually described as 0.3 MAC, allowing an adequate margin of safety if the level of volatile anesthetic is not allowed to drop below 0.5 MAC. This is a scenario in which a level of consciousness monitor may be useful. Because the procedure is not painful, narcotics are rarely indicated as these most often lead to hypotension.

Blood pressure should be kept in the high normal range by appropriately titrating the depth of anesthesia. This will allow the child to better tolerate the hemodynamic consequences of inducing the tachyarrhythmia. Intravenous volume should be restricted, especially during long procedures as the continual flushing of the EP catheters delivers significant fluid. The volume delivered is kept track of by the EP technician and should be noted.

Transcutaneous patches are routinely placed on all patients for cardioversion or defibrillation. Ablation is accomplished using either radiofrequency or cold freezing (cryoablation) techniques. The anesthetic technique and choice of drugs is not affected by the type of ablation used. Complications with ablation are rare, but it may lead to complete heart block when used in the vicinity of the AV node, for example, with AVNRT. Pacing can be achieved using the existing catheters if necessary.

What’s the evidence?

Burch, TM, McGowan, FX, Kussman, BD, Powell, AJ, DiNardo, JA. “Congenital supravalvular aortic stenosis and sudden death associated with anesthesia: what’s the mystery?”. Anesth Analg. vol. 107. 2008. pp. 1848-54.

Carmosino, MJ, Friesen, RH, Doran, A, Ivy, DD. “Perioperative complications in children with pulmonary hypertension undergoing noncardiac surgery or cardiac catheterization”. Anesth Analg. vol. 104. 2007. pp. 521-7.

Crystal, MA, Ing, FF. “Pediatric interventional cardiology: 2009”. Curr Opin Pediatr. vol. 22. 2010. pp. 567-72.

Drover, DR, Cao, H, Jackson, E, Williams, GD, Ramamoorthy, C, Van Hare, GF, Niksch, A, Dubin, AM. “The effects of dexmedetomidine on cardiac electrophysiology in children”. Anesth Analg. vol. 106. 2008. pp. 79-83.

Friesen, RH, Williams, GD. “Anesthetic management of children with pulmonary arterial hypertension”. Paediatr Anaesth. vol. 18. 2008. pp. 208-16.

Goldberg, DJ, Shaddy, RE, Ravishankar, C, Rychik, J. “The failing Fontan: etiology, diagnosis and management”. Exp Rev Cardiovasc Ther. vol. 9. 2011. pp. 785-93.

Hickey, PR, Hansen, DD, Cramolini, GM, Vincent, RN, Lang, P. “Pulmonary and systemic hemodynamic responses to ketamine in infants with normal and elevated pulmonary vascular resistance”. Anesthesiology. vol. 62. 1985. pp. 287-93.

Mehta, R, Lee, KJ, Chaturvedi, R, Benson, L. “Complications of pediatric cardiac catheterization: a review in the current era”. Cathet Cardiovasc Interv. vol. 72. 2008. pp. 278-85.

Ramamoorthy, C, Haberkern, CM, Bhananker, SM, Domino, KB, Posner, KL, Campos, JS, Morray, JP. “Anesthesia-related cardiac arrest in children with heart disease: data from the Pediatric Perioperative Cardiac Arrest (POCA) registry”. Anesth Analg. vol. 110. 2010. pp. 1376-82.

Rosenthal, DN, Hammer, GB. “Cardiomyopathy and heart failure in children: anesthetic implications”. Paediatr Anaesth. vol. 21. 2011. pp. 577-84.

Sarkar, M, Laussen, PC, Zurakowski, D, Shukla, A, Kussman, B, Odegard, KC. “Hemodynamic responses to etomidate on induction of anesthesia in pediatric patients”. Anesth Analg. vol. 101. 2005. pp. 645-50.

Vitiello, R, McCrindle, BW, Nykanen, D, Freedom, RM, Benson, LN. “Complications associated with pediatric cardiac catheterization”. J Am Coll Cardiol. vol. 32. 1998. pp. 1433-40.

Williams, GD, Jones, TK, Hanson, KA, Morray, JP. “The hemodynamic effects of propofol in children with congenital heart disease”. Anesth Analg. vol. 89. 1999. pp. 1411-6.

Williams, GD, Philip, BM, Chu, LF, Boltz, MG, Kamra, K, Terwey, H, Hammer, GB, Perry, SB, Feinstein, JA, Ramamoorthy, C. “Ketamine does not increase pulmonary vascular resistance in children with pulmonary hypertension undergoing sevoflurane anesthesia and spontaneous ventilation”. Anesth Analg. vol. 105. 2007. pp. 1578-84.