Abstract
Left ventricular outflow tract obstruction (LVOTO) is a complex congenital cardiac defect that interferes with the ejection of blood from the left ventricle into the ascending aorta. LVOTO is a heterogeneous defect in which the timing and mode of clinical presentation vary based on multiple factors, including the degree and levels of obstruction, associated hypoplasia of the left ventricle and mitral valve, and the presence of concomitant cardiac and extracardiac anomalies such as aortic arch obstruction, patent ductus arteriosus, and atrial or ventricular septal defect. Subsequently, the timing and type of intervention differ based on those factors, with the initial procedure ranging from an extreme of neonatal single-ventricle palliation to isolated adolescent aortic valve intervention. The classification of LVOTO is typically based on the level of obstruction and includes valvar, subvalvar, and supravalvar stenosis. Valvar stenosis is the most common type and constitutes approximately 65% to 75% of cases, whereas subvalvar and supravalvar stenosis constitute approximately 15% to 20% and 5% to 10% of cases, respectively.
Key Words
Ross procedure, Aortic stenosis, mechanical prosthesis, Aortic valve repair, aortic valvuloplasty, Williams syndrome, hypertrophic obstructive cardiomyopathy
Left ventricular outflow tract obstruction (LVOTO) is a complex congenital cardiac defect that interferes with the ejection of blood from the left ventricle into the ascending aorta. LVOTO is a heterogeneous defect in which the timing and mode of clinical presentation vary based on multiple factors, including the degree and levels of obstruction, associated hypoplasia of the left ventricle and mitral valve, and the presence of concomitant cardiac and extracardiac anomalies such as aortic arch obstruction, patent ductus arteriosus, and atrial or ventricular septal defect. Subsequently, the timing and type of intervention differ based on those factors, with the initial procedure ranging from an extreme of neonatal single-ventricle palliation to isolated adolescent aortic valve intervention.
The classification of LVOTO is typically based on the level of obstruction and includes valvar, subvalvar, and supravalvar stenosis. Valvar stenosis is the most common type and constitutes approximately 65% to 75% of cases, whereas subvalvar and supravalvar stenosis constitute approximately 15% to 20% and 5% to 10% of cases, respectively.
Valvar Aortic Stenosis
Valvar aortic stenosis is the most common form of LVOTO, occurring in 65% to 75% of patients. Bicuspid aortic valve occurs in 1% to 2% of the population and represents 3% to 5% of all congenital heart anomalies, with the incidence in males being three to five times higher than that in females. Valvar aortic stenosis is an anatomic and clinical spectrum that ranges from one extreme of critical neonatal aortic stenosis that is associated with ductal-dependent systemic circulation (almost 10% of those patients) to another extreme of mild aortic stenosis in asymptomatic children with the only finding being an incidental murmur on physical examination.
Critical Neonatal Aortic Stenosis
The aortic valve pathology is most commonly bicuspid associated with thickened dysmorphic cusps and fused commissures. Less commonly, the valve is tricuspid with thickened cusps and fused commissures. When the valve is unicuspid, there might be a single commissure or no commissure, and the small orifice might be central or eccentric.
During fetal development, severe LVOTO exposes the left ventricle to increased afterload that results in ventricular hypertrophy with subsequent systolic and diastolic myocardial dysfunction. Ventricular hypertrophy and higher intracavitary pressure lead to decreased coronary perfusion pressure and chronic in utero subendocardial ischemia with consequent development of endocardial fibroelastosis that further impairs ventricular function. Additionally, reduced in utero antegrade blood flow through the aortic valve leads to underdevelopment of the left heart structures and hypoplasia of the mitral valve, left ventricle, subvalvar area, ascending aorta, and aortic arch. Clinical presentation varies based on the severity of LVOTO and the degree of associated ventricular hypoplasia and dysfunction. Neonates with critical aortic stenosis often have a rapidly developing and dramatic presentation after birth. As the ductus arteriosus closes, they experience decreased systemic and coronary perfusion, acute hemodynamic deterioration with cardiovascular collapse, metabolic acidosis, end-organ injury, and shock. Neonates and infants with lesser degrees of obstruction may gradually develop symptoms secondary to persistent high left ventricular afterload, left atrial hypertension, myocardial dysfunction, and poor systemic cardiac output. Symptoms such as failure to thrive, irritability, and tachypnea and increased work of breathing might necessitate early intervention during the first few months of life.
Diagnostic studies include electrocardiography (ECG), which might demonstrate evidence of left ventricular or biventricular hypertrophy, with common evidence of strain. The chest x-ray examination might show cardiomegaly and evidence of pulmonary congestion in neonates with critical aortic stenosis, but findings might be normal in those with less severe stenosis. Echocardiography is the primary form of diagnosis, and it delineates the level of obstruction, gradient across the LVOT, presence of endocardial fibroelastosis, hypoplasia of left heart structures, mitral valve anomalies, ascending aorta or aortic arch, and associated cardiac anomalies such as atrial and ventricular septal defects and patent ductus arteriosus. The direction of flow across the transverse arch and ductus might help to determine whether or not the left heart is capable of supporting the systemic circulation after biventricular repair. Of note, the LVOT gradient may be underestimated in those neonates due to severe left ventricular dysfunction, frequent presence of mitral regurgitation, and existence of patent ductus arteriosus or distal arch obstruction. Cardiac catheterization for diagnosis is usually not required but can be helpful to measure LVOT gradient directly (in older patients with less severe LVOTO) and to evaluate associated anomalies that may not be apparent on echocardiography. The role of cardiac catheterization in neonatal aortic stenosis is mainly for intervention rather than diagnosis. It is important to mention the role of fetal echocardiography in the prenatal diagnosis of neonates with critical aortic stenosis. Prenatal identification of LVOTO and associated intracardiac and extracardiac anomalies can result in better planning of the care following delivery (both for the medical team and for the family) and avoidance of circulatory collapse in patients with unrecognized pathology as a result of ductal closure and can offer the possibility of fetal intervention with aortic valvuloplasty, which, in selected cases, can be associated with better growth of left heart structures and higher likelihood of achieving biventricular repair.
In neonates with ductal-dependent systemic circulation, urgent neonatal intervention is indicated. In those with less severe stenosis, surgery is indicated when the infants develop symptoms such as failure to thrive, respiratory distress, and irritability, in conjunction with peak Doppler-derived LVOT gradient that is greater than 40 mm Hg or a peak-to-peak catheter-derived LVOT gradient that is greater than 30 mm Hg.
The medical management of neonates with postnatal diagnosis of ductal-dependent critical aortic stenosis involves rapid restoration of ductal patency with the use of prostaglandin E 1 infusion . This will restore systemic perfusion by shunting blood right to left across the ductus and can help reduce pulmonary hypertension seen with severe left ventricular dysfunction ( Fig. 51.1 ). Intubation and mechanical ventilation may also be required to aid in correction of severe acidosis and control of pulmonary hypertension. The use of inotropic drugs improves the poor ventricular contractility. Judicious administration of intravenous fluids and correction of metabolic and electrolyte anomalies are important in the resuscitation of these critically ill patients. In neonates with very severe LVOTO or in those with severely depressed left ventricular function or inadequate left ventricle or mitral valve sizes, balloon atrial septostomy might be needed to allow unobstructed pulmonary venous egress to the right side of the heart, which is supporting the systemic circulation. In the absence of atrial communication, severe pulmonary hypertension and respiratory compromise might be very rapid, necessitating extracorporeal membrane oxygenation (ECMO) support for stabilization before any intervention. On the other hand, when the degree of aortic valve stenosis seems to be less severe, and the left heart structures seem to be well developed, especially in those who seem to have adequate cardiac output generated from the left ventricle with mainly left-to-right shunt across the ductus arteriosus, a trial of discontinuation of prostaglandin can be undertaken, and intervention can be deferred if the child is growing and having no symptoms of poor cardiac output or respiratory compromise.
The most important aspect in the management of infants with critical aortic stenosis is to determine whether the left heart structures are capable of supporting the systemic circulation. Although the decision is obvious in patients with well-developed left heart structures (who should be amenable to a two-ventricle repair pathway) or in patients with severe hypoplasia of the left ventricle or mitral valve (who will require staging toward single-ventricle palliation), the middle of the spectrum is more complex in terms of clinical decision making. The most difficult decisions revolve around patients who have borderline development of the left ventricle and the mitral valve. In those patients, single-ventricle palliative options, although possible, would forfeit the opportunity for appropriate candidates to undergo biventricular repair. Conversely, aggressive attempts to attain biventricular status could come at the relative cost of greater risk of death and higher subsequent morbidity and reoperation. The importance of proper treatment selection is reflected by the higher mortality reported in older series due to failure to address the heterogeneity of this disease and to tailor the treatment to specific patient anatomy. The significance of appropriate initial triage cannot be overestimated because multiple reports have shown universally poor results in patients requiring a crossover between strategies.
Several studies have attempted to identify preoperative predictors of suitability for single-ventricle versus biventricular repair in neonates with critical aortic stenosis. Rhodes identified several clinical risk factors for successful biventricular repair, including mitral valve area less than 4.75 cm 2 /m 2 , long-axis dimension of the left ventricle relative to the long-axis dimension of the heart less than 0.8, diameter of the aortic root less than 3.5 cm/m 2 , and left ventricular mass less than 35 g/m 2 . The presence of more than one of those risk factors predicted high mortality following biventricular repair. Subsequently he proposed a score that is derived from a multivariable regression equation with a discriminating score less than −0.35 predictive of death after a biventricular repair. However, the Rhodes score was based on retrospective data from a small group of 65 patients with critical aortic stenosis who were preselected for biventricular repair, and subsequent investigation from multiple different studies has shown poor discrimination with the Rhodes score when applied to neonates with multiple levels of left-sided obstruction or whose primary pathology is other than critical aortic stenosis.
The Congenital Heart Surgeons’ Society (CHSS) reported a multi-institutional study of 320 neonates with critical aortic stenosis enrolled between 1994 and 2000. Biventricular repair was performed in 116 patients, whereas an initial Norwood procedure was performed in 179 patients. Five-year survival was 70% for neonates who underwent biventricular repair and 60% for those who underwent a Norwood procedure. Complex statistical techniques were then used to model the magnitude and direction of the survival benefit for the Norwood procedure over biventricular repair pathway. Independent factors associated with greater survival benefit with the Norwood pathway included younger age at entry, higher grade of endocardial fibroelastosis, lower z score of the aortic valve at the level of the sinuses of Valsalva, larger ascending aortic diameter, absence of moderate or severe tricuspid regurgitation, and lower z score of the left ventricular length. Subsequently a regression equation was formulated to predict patient’s survival benefit, with a positive number representing improved survival with a Norwood procedure, a negative number representing improved survival with a biventricular repair strategy, and the magnitude of the number representing the degree of predicted survival benefit. Importantly, the CHSS demonstrated that commonly used selection criteria resulted in inappropriate patient triage in a significant number of neonates. Predicted survival benefit favored the Norwood procedure in 50% of patients who had biventricular repair, and it favored biventricular repair in 20% of patients who had the Norwood procedure. Choosing the correct management strategy could have resulted in a substantial survival advantage for patients in the cohort in whom the incorrect management option was chosen. Further analysis of the CHSS data showed that inappropriate pursuit of biventricular repair in borderline candidates was more frequent and more consequential in survival terms than inappropriate pursuit of a single-ventricle palliation strategy. The final updated form of the CHSS equation is available through the CHSS website ( http://www.chssdc.org/content/chss-score-neonatal-critical-aortic-stenosis ).
The single-ventricle palliation strategy using the Norwood operation or the hybrid approach is discussed in Chapter 66 . Heart transplantation can be an alternative strategy, although that is reserved usually for patients who fail initial single-ventricle or biventricular repair strategies.
The focus for the remainder of this chapter will be on biventricular repair options.
Percutaneous balloon valvuloplasty is considered the initial procedure of choice at most centers. However, it is relatively contraindicated in the rare patients with preexistent moderate and severe aortic valve regurgitation. Vascular access is usually obtained with an antegrade (venous) transseptal approach using the umbilical or femoral veins; however, a retrograde approach using the carotid, femoral, or umbilical artery is also common ( Fig. 51.2 ). Advantages of percutaneous valvuloplasty include avoidance of surgical morbidity associated with cardiopulmonary bypass. Disadvantages include vascular access complications, inability to precisely determine where the leaflets will tear with resultant potential for aortic valve insufficiency, and rarely, mitral valve injury. It is important not to overdilate the aortic valve because the goal of intervention in these critically ill patients is improvement, and not complete elimination, of LVOT gradient until the patient is older and bigger, at which point a more precise intervention can be planned. The initial balloon size is usually chosen at 80% to 90% of the diameter of the aortic valve annulus because there is evidence that the incidence of new-onset or worsening aortic regurgitation increases as the balloon to annulus ratio increases. During the procedure the balloon is positioned across the valve with half or more of its length below the annulus. During balloon inflation, observation of the “tight waist” is important. In older patients, ventricular pacing during balloon inflation might be required to avoid balloon malposition. After the first balloon inflation the LVOT gradient is remeasured and angiogram is repeated to assess the degree of aortic regurgitation. It is acceptable to reinflate the balloon or use the next balloon size up for inadequate relief of an LVOT gradient as long as significant iatrogenic aortic regurgitation is nonexistent. Nonetheless, it is still important not to exceed the annular diameter with the larger balloon. At the end of the intervention an exit angiogram and another set of hemodynamic measurements, including cardiac output, should be obtained. Following the percutaneous intervention and removal of the arterial and venous sheaths, hemostasis is obtained while carefully monitoring limb perfusion and distal pulses. Acute occlusive arterial injury is a well-known complication following cardiac catheterization in neonates and small infants, and the reported incidence is 0.6% to 9.6%. The risk of vascular injury increases in smaller children, with the need to perform an intervention via the artery, exchanging arterial catheters and the use of larger catheters, a longer procedural time, and a final activated clotting time of less than 250 seconds. Following percutaneous intervention, patients are transferred back to the intensive care unit. Their convalescence following the intervention varies based on several factors, including preprocedure cardiac function, preexistence of other organ dysfunction, and the degree of residual LVOTO or iatrogenic aortic regurgitation. Although the optimal intervention goal is to achieve adequate relief of LVOTO with minimal iatrogenic aortic regurgitation, this objective is not always attained. In patients with significant residual lesions (LVOTO or aortic regurgitation), cardiac output and systemic perfusion are severely affected, requiring significant inotropic support, mechanical ventilation, and careful fluid management as described later for postoperative care following surgical intervention.
Before the evolution of percutaneous balloon valvuloplasty as the standard first intervention of choice, surgical valvotomy was the mainstay of treatment of critical aortic stenosis in neonates and infants. Different approaches such as transventricular closed aortic valvotomy and open valvotomy with inflow occlusion or with cardiopulmonary bypass were developed. The current surgical technique, when surgery is considered the best option, uses cardiopulmonary bypass and cardioplegic arrest of the heart. In neonates and infants the goal is to perform a valvotomy that will decrease LVOTO and prevent significant aortic insufficiency. An aortotomy that is directed toward the noncoronary aortic sinus is performed to allow adequate exposure of the valve. Aortic valvotomy is then completed by dividing the fused commissures to within 1 to 2 mm of the aortic wall only. In neonates, given that small increases in valve opening are associated with significant reduction in LVOT gradient, conservative valvotomy is recommended to avoid the risk of aortic regurgitation ( Fig. 51.3 ).
The goal of intervention for neonates and infants with critical aortic stenosis is to improve LVOTO by opening the aortic valve as much as possible, without tearing it or creating significant aortic insufficiency. In most patients, aortic valvotomy, whether it is performed percutaneously or surgically under direct vision, will restore enough outflow from the left ventricle to support systemic circulation. However, it is important to emphasize that the condition of these infants can be extremely critical, and it may take some time (several days) for the left ventricle to begin functioning more normally. Many of these patients present for intervention with significant left ventricular dysfunction caused by a combination of intrauterine flow abnormalities as well as from critical LVOTO. Simply opening the aortic valve does not always result in immediate return of normal cardiac output. The left ventricle may still be “stiff” (noncompliant), and, in neonates with significant left ventricular dysfunction and associated pulmonary hypertension, it may sometimes be helpful to maintain ductal patency with prostaglandin infusion for a few days following percutaneous or surgical valvotomy to help support the systemic circulation and decompress pulmonary artery hypertension until the left ventricular compliance and systolic function improve (see Fig. 51.1 ). If the patient is an appropriate candidate for biventricular pathway, left ventricular recovery should occur within a few days following intervention.
Intensive care unit management centers around ensuring that the infant has adequate systemic perfusion. An echocardiogram following valvotomy can indicate the quality of left ventricular function and ensure that there appears to be adequate flow across the LVOT. When left ventricular dysfunction is severe, the infant may be tachycardic (with heart rates approaching 180 beats/min). Because cardiac output = stroke volume × heart rate, when stroke volume is limited by poor left ventricular function, heart rate will generally increase to maintain output. Therefore, in the absence of a primary arrhythmia such as junctional ectopic tachycardia, tachycardia following aortic valvotomy is a sign of compromised left ventricular function. These babies should remain ventilated because adding the work (and increased oxygen demand) of breathing to an already compromised cardiac output might be enough to result in hemodynamic collapse. Perfusion should be monitored by physical signs (such as pulses and skin temperature, near-infrared spectroscopy), as well as by chemical signs (such as lactate level, signs of acidosis on blood gas and mixed venous saturation if available). Babies with decreased perfusion may need to have escalated management with afterload reduction and/or inotropes. As the parameters of perfusion stabilize, the infant can be carefully followed with the expectation that the patient will recover. In extremely compromised patients, ECMO support might be needed to provide adequate systemic perfusion to the kidneys and other vital organs while waiting for the left ventricle function to recover.
When neonates fail to improve, it is important to reevaluate the adequacy of the aortic valve and aortic annulus. If the left ventricle is adequate and there is no significant inflow obstruction (congenital mitral valve abnormality), the neonate may be a candidate for neonatal aortic valve replacement. Although a primary Ross procedure with annular enlargement (Ross-Konno) may be a lifesaving initial operation in selected patients with severe multilevel LVOTO, more commonly it is used later in patients who continue to have persistent poor systemic perfusion and pulmonary hypertension associated with persistent LVOTO or significant aortic regurgitation following percutaneous or surgical intervention. The details of this operation will be described later in this chapter. Persistent postoperative poor systemic perfusion and pulmonary hypertension due to inadequate left ventricle or mitral valve size to support the systemic circulation suggests that a single-ventricle palliation strategy was likely the correct choice for that patient, and usually a crossover to that strategy is associated with high mortality. The alternative option of listing for heart transplantation is restricted by the long time needed until a donor heart is available and the high risk of waiting list mortality.
The choice between percutaneous valvuloplasty and surgical valvotomy is largely institution dependent. Few studies have compared outcomes between the two approaches, although it is worthwhile to note that outcomes with both surgical and percutaneous approaches have improved recently due to technical advances, improved perioperative care, and most importantly improved patient selection and more proper patient triage to undergo single-ventricle palliation versus biventricular repair strategy. There are no prospectively randomized studies comparing outcomes between the two approaches. Numerous reports cited the early and late outcomes of either one approach or the other, but few compared surgical versus percutaneous approaches, and those that did are limited by the lack of adjustments between the two groups. In a retrospective multi-institutional study by the CHSS, 110 patients underwent either surgical aortic valvotomy ( n = 28) or percutaneous balloon valvuloplasty ( n = 82). The study demonstrated that, while controlling for preprocedure morphology, percutaneous balloon valvuloplasty was more effective in relieving stenosis than surgical aortic valvotomy as evidenced by greater mean percentage reduction in systolic gradient (65% versus 41%) and lower residual median gradients (20 mm Hg versus 36 mm Hg). However, percutaneous balloon valvuloplasty was also associated with greater likelihood of important aortic regurgitation (18% versus 3%). Freedom from reintervention was similar for the two groups (91% at 1 month and 48% at 5 years). Significant factors for reintervention included preprocedural use of inotropic agents, the presence of postprocedural moderate to severe aortic regurgitation, and a lower weight at initial intervention. Risk-adjusted freedom from death was also similar between the two groups and was 82% at 1 month and 74% at 1 year. Risk factors for death included preprocedural mechanical ventilation and anatomic factors such as smaller aortic valve diameter at the level of the annulus, sinotubular junction, or subaortic region, indicating that many of those patients might have been inappropriately triaged into the biventricular tract and that they might have been better served by a single-ventricle palliation approach. Both surgical and percutaneous outcomes have improved with better selection, and some centers have reported operative mortality of 5% to 10% and freedom from reintervention of 85% at 5 years from surgical aortic valvotomy in neonates and infants. At the end, careful evaluation of patients’ characteristics along with understanding institutional expertise should help determine the first-line procedure.
Valvar Aortic Stenosis in Older Children
Children with LVOTO that is not critical enough to require care during infancy represent the other end of the spectrum of valvar aortic stenosis. In older patients there is typically adequate development of left heart structures, and their pathology is mainly confined to the aortic valve itself. Aortic valve anatomy in this group of patients is most commonly bicuspid (>70% of patients). In the remaining patients the valve is tricuspid, although a unicuspid valve can be occasionally seen in these older patients.
The pathophysiology of valvar aortic stenosis is related to increased afterload on the left ventricular myocardium with subsequent left ventricle hypertrophy. Ultimately, symptoms may occur related to the associated decreased left ventricular compliance and increased left ventricular end-diastolic pressure. Left ventricular hypertrophy can lead to consequent subendocardial ischemia and predisposition to sudden death from ventricular arrhythmias.
Most children are asymptomatic in the early years of life and have normal growth. The stenosis is often identified during routine physical examination due to the presence of a murmur. Symptoms of congestive heart failure (such as shortness of breath and decreased exercise tolerance), angina, and syncope or even sudden death are indications that the LVOTO is becoming severe. Occasionally, spontaneous endocarditis is the presenting manifestation of valvar aortic stenosis.
The diagnosis is mainly with echocardiography. The correlation between the Doppler-derived and cardiac catheterization gradients is usually reliable in older patients; cardiac catheterization is more often used for those patients in whom balloon valvuloplasty is being considered. Occasionally, diagnostic cardiac catheterization is performed to validate a significant LVOT gradient when echocardiogram data are confusing, to evaluate coronary anatomy and to measure end-diastolic pressures. The role of magnetic resonance imaging for the diagnosis of those patients is increasing as a noninvasive modality that provides valuable information about the size of the aortic annulus, ascending aorta, and aortic arch. It can also delineate coronary origins and allow measurement of systolic function, left ventricular mass, and LVOT gradients. Finally, myocardial perfusion can be assessed during rest and stress induced by adenosine and dobutamine for better understanding of hemodynamic effects of LVOTO. In younger patients, magnetic resonance imaging might require sedation and at times general anesthesia, but in older children that is not required.
Indications for intervention in these older patients include (1) the presence of symptoms such as angina, syncope, or congestive heart failure, ischemic or repolarization changes on rest or exercise ECG, associated with resting peak systolic LVOT gradient of greater than 40 mm Hg; (2) the presence of depressed left ventricular function (even with LVOT gradient <40 mm Hg); or (3) resting peak systolic LVOT gradient of greater than 50 mm Hg. Asymptomatic patients with LVOT gradient of less than 40 mm Hg are generally followed at regular intervals (every 6 to 12 months) by outpatient cardiology.
Both percutaneous balloon valvuloplasty and surgical aortic valvotomy are viable options in this age-group and can be done successfully as initial management strategies. In older patients with significant valvar LVOTO, most centers continue to prefer percutaneous balloon valvuloplasty. However, surgical valvotomy can often be performed with more precision in older patients, with incision in the areas of commissural fusion carefully extended to, but not into, the aortic valve annulus. The risk of aortic insufficiency is less with surgical valvotomy than with balloon valvuloplasty, and surgical options for aortic valve repair are possible. In addition, thinning of the aortic cusps and excision of thickened nodules may increase the mobility of the valve and decrease residual LVOT gradient. Although it is usually recommended not to incise the false raphe, a few groups have reported an improved experience with creation of a tricuspid valve from a bicuspid valve or a bicuspid valve from a unicuspid valve by incising the raphe and resuspending the incised cusps with neocommissures created from triangular patches of glutaraldehyde-treated autologous pericardium. The experience of aortic valve repair is small, and the lowest age limit to allow these techniques is not very clear. Nevertheless, aortic valve repair may become a more common option in the years ahead.
All forms of LVOTO create pressure loading to the left ventricle that can result in varying degrees of left ventricular hypertrophy. In more severe instances this can lead to an increase in left ventricular end-diastolic pressures and a less compliant ventricle. In the preoperative period these patients may be extremely volume sensitive and in fact preload dependent, although they remain at risk of developing pulmonary edema as well. Therefore they require very judicious and careful management of their preload status to preserve their cardiac output. Excessive inotropic therapy may increase cardiac oxygen demand and tachycardia, which may lead to decreased ventricular volume, increased LVOTO, and ventricular ischemia.
Although these patients often have preserved systolic function after relief of their LVOT obstruction, they may have some degree of diastolic dysfunction that can manifest as pulmonary edema or tachycardia (related to low stroke volumes and rate-dependent cardiac output). Because of the challenged cardiac output, particularly in neonates and infants, maintenance of adequate hemoglobin (to increase oxygen delivery capability) is reasonable. Postoperative maintenance of cardiac output can be monitored by measurement of oxygen delivery (mixed venous oxygen saturation, arteriovenous oxygen difference, lactate levels) and intubation to decrease systemic oxygen demand. It is prudent to mechanically support ventilation until oxygen delivery is normal (normal indices of perfusion), with normal heart rate trend without tachycardia and recovering left ventricular function on echocardiogram.
Naturally the postoperative convalescence in older patients is typically less complicated and the risk of death is very low, approaching 1%. Following percutaneous balloon valvuloplasty or surgical aortic valvotomy, almost 35% of patients will require reintervention on the aortic valve within 10 years.
When aortic valvuloplasty or repair fails or is not successful, aortic valve replacement may become necessary. The goal of aortic valvuloplasty or repair is to delay the need for aortic valve replacement, and it is likely that many patients with aortic valve interventions will ultimately require a valve replacement. There are numerous options for aortic valve prostheses, and none of them are “perfect.” An optimal choice for an aortic valve prosthesis would include (1) ready availability in different sizes, (2) durability, (3) excellent (normal) hemodynamic profiles, (4) minimal thromboembolic risk (without the need for lifelong anticoagulation), (5) growth potential, (6) low incidence of structural valve degeneration, and (7) minimal risk for needing reoperation. No such choice is currently available, and all alternatives are associated with important drawbacks.
Mechanical prostheses come in different styles, although the bileaflet design has become the most widely used. Size is limited by the properties of the material, and, although these valves can be made in small sizes (15 to 17 mm) that can be placed in young patients, they are often not suitable for infants and very small children. The hemodynamic profiles vary with size, and smaller prostheses have inferior flow properties.
Annular enlargement techniques could be used to allow placement of larger prostheses in small patients. To enlarge the aortic annulus, it is critical to appreciate the important anatomic constraints imposed by the conduction system, mitral valve, and coronary arteries ( Fig. 51.4 ). Before the more widespread application of the pulmonary autograft procedure in infants and children, annular enlarging procedures that allowed placement of larger prosthetic valves were more common ( Fig. 51.5 ). Those techniques include the Nicks, Manougian, and Konno procedures. In the Nicks procedure the aortic incision is extended to the area between the left noncoronary commissure and the base of the noncoronary cusp into the area of intervalvular fibrosa without cutting into the anterior mitral valve leaflet. In the Manougian procedure the incision is the same as in the Nicks procedure, but the cut is extended across the intervalvular fibrosa into the center of the anterior mitral leaflet. In the Konno procedure the aortic annulus is incised between the right and left coronary cusps extending into the ventricular septum with patch reconstruction of the septum and ascending aorta.
Operative mortality of aortic valve replacement with a mechanical prosthesis in children is 2% to 13%, and survival ranges between 75% and 88% at 15 years. Importantly, lifelong anticoagulation is required, which is challenging due to poor compliance and activity restrictions, in addition to complicated pregnancy in females. Reported freedom from bleeding is 96% to 100% at 10 years, and reported freedom from thromboembolism is 90% to 100%, better than in adults. The different hemodynamic properties in children with faster heart rate and less incidence of arrhythmias, atrial dilation, or myocardial dysfunction might all contribute to the relatively lower thromboembolism risk. Despite the lack of structural valve degeneration, reoperations are not uncommon, and freedom from reoperation is 55% to 90% at 15 years, usually related to pannus formation (fibrous ingrowth across the valve mechanism) or development of patient-prosthesis mismatch as the child grows in the presence of fixed prosthesis sizes, in addition to occasional reoperations for valve thrombosis, perivalvular leak, or endocarditis.
Tissue prostheses most typically are manufactured using bovine or porcine valves mounted on rigid struts. They are unavailable in sizes below 19 mm, and therefore they are not suitable for small children, even using annular enlargement techniques. The hemodynamic profiles vary with size, and smaller prostheses have inferior flow properties, particularly due to the large size of the sewing ring compared to the actual orifice size of the valve opening. Stentless bioprosthetic valves are glutaraldehyde-prepared porcine aortic roots. Although they have the hemodynamic advantage of not needing rigid stents for mounting the valve tissue, and thus have hemodynamic superiority at any comparable size, they are still subject to structural degeneration, calcification, and need for early reoperation, particularly in younger patients. All bioprosthetic valves have low thromboembolic risk and thus do not require anticoagulation. Nonetheless, their use in children is associated with decreased valve longevity due to lack of growth potential and, most importantly, from structural valve degeneration that is faster than that seen in adults and is inversely related to patient age and prosthesis size. For the most part, bioprosthetic valves are not a good choice for aortic valve replacement in young patients. In children and young adults, reported survival is almost 85% at 10 years, whereas reported freedom from reoperation is only 15% to 30% at 10 years.
Homografts (or allografts) are cadaver human valves. Both aortic and pulmonary homografts are available for human implantation. Homografts are harvested from human donors and cryopreserved for implantation. Hospitals using homografts will usually have specially made liquid nitrogen storage containers in the operating room for maintaining a homograft stock, and selected homografts can be thawed in the operating room at the time of implantation. Aortic homografts are most commonly used for aortic valve replacement. Early techniques used a “freestyle” form of placing the homograft valve into the aortic annulus, but more commonly the homograft is now used as a complete aortic root replacement with reimplantation of the coronary arteries ( Fig. 51.6 ). Aortic homografts provide excellent hemodynamic profiles that are similar for larger and smaller sizes; hence they are suitable for small children. However, homograft availability, especially in the smallest ranges, varies due to a limited donor pool. They have a negligible thromboembolic risk and thus do not require anticoagulation. Nonetheless, as is the case for all nonviable tissue prostheses, their use in children is associated with decreased longevity and frequent reoperation due to rapid degeneration, ultimately leading to deterioration of their hemodynamic properties. Homograft longevity varies with type (aortic versus pulmonary) and patient age, and reported freedom from reoperation is 15% to 88% at 10 years. Their use in children is very limited except in very small children unable to undergo the Ross procedure or those with invasive endocarditis.
The Ross procedure (named after Donald Ross, who first performed this procedure in the 1967) uses the pulmonary autograft (the patient’s own pulmonary valve, which is “harvested” during the surgical procedure) that is transplanted into the aortic position and therefore can be applied to all patient ages ( Fig. 51.7 ). Because the pulmonary autograft is the patient’s own living pulmonary valve, the Ross procedure is associated with an excellent hemodynamic profile in all sizes, and growth potential of this living valve replacement allows maintenance of this superior hemodynamic profile throughout life. Furthermore, the Ross procedure is a versatile operation that can be used in patients with various LVOT pathologies. The addition of Konno-type aortoventriculoplasty (Ross-Konno) allows successful management of patients with significant annular hypoplasia or complex multilevel LVOTO. The original Ross-Konno description involved a large incision into the septum, creating a ventricular septal defect that is closed with patch. The modified Ross-Konno involves an incision across the aortic annulus into the septum, extensive septal myectomy and LVOT enlargement without a patch, using a portion of the infundibular muscle from the right ventricular outflow tract that can be harvested with the pulmonary autograft ( Fig. 51.8 ). In children the autograft is implanted most commonly as a full root; alternatively, subcoronary or inclusion techniques could be used in older patients but not in young children with small aortic roots. Anticoagulation is not required after Ross due to negligible thromboembolic risk.
Despite technical complexity, the Ross procedure can be performed safely in experienced hands with operative mortality of less than 2.5%. Infants (patients <1 year old) have a higher mortality risk, approaching 15% to 20%, likely due to the more common association of other important cardiac lesions. Mortality risk can be highest in neonates, particularly those with complex LVOTO needing simultaneous arch or mitral procedures. Mortality risk is also increased in emergency surgery, when longer bypass time is required, or in those patients with preoperative ventricular dysfunction. These findings suggest that prior palliation with surgical or percutaneous aortic valvuloplasty might decrease mortality risk and that neonates with concomitant significant mitral pathology or arch obstruction might benefit from other surgical alternatives, which may include single-ventricle palliation. Time-related survival of children following Ross is stable with meager attrition risk beyond the perioperative period. Long-term survival after Ross is superior to other valve substitutes, likely due to an excellent autograft hemodynamic profile and trivial thromboembolic or bleeding risks.
Despite its numerous advantages, recommendation for Ross has waned due to concerns about late neoaortic root dilation and subsequent autograft regurgitation. In addition, some argue that the Ross procedure creates “two-valve disease” because it requires a replacement for the pulmonary valve in addition to the aortic valve. It is debatable if neoaortic annulus and root dilation following Ross is growth proportional to somatic enlargement or disproportionate pathologic remodeling. In general, neoaortic root dimensions immediately after Ross are larger than in healthy children, the annulus grows in proportion to the child’s somatic growth, whereas the root at sinuses and sinotubular junction levels dilate disproportionately with time. Autograft regurgitation can develop in association with neoaortic dilation, more so with dilation of the sinotubular junction than the sinus.
Several factors increase the risk for late autograft reoperation due to autograft dilation, such as older age, bicuspid aortic valve pathology with predominant regurgitation, dilated sinotubular junction, dilated aortic annulus, and geometric mismatch between semilunar valves (aortic larger than pulmonary). Freedom from autograft reoperation in children with congenital aortic stenosis is 75% to 95% at 10 years, and this freedom from needing reoperation is highest for those with aortic stenosis or mixed (stenosis and insufficiency) disease and worse for patients with primary aortic insufficiency or those with annular-aortic ectasia (often seen as part of bicuspid aortic valve disease).
Children undergoing Ross-Konno are an interesting group because most patients have stenosis and a small aortic annulus and thus are potentially at lower risk of root dilatation. Conversely, annular incision in Ross-Konno and patch placement could result in loss of native annular support to the autograft, with subsequent higher root dilation risk. A recent report showed that, after Ross-Konno, both the neoaortic annulus and the root increased in size proportionately to somatic growth with little risk of late autograft regurgitation or reoperation.
Several technical modifications have been described aiming to reduce autograft dilation risk. These techniques include thinning of the muscle rim below the valve, suturing the autograft within the native aortic annulus, autograft shortening, proximal and distal suture line enforcement, or ascending aorta replacement with Dacron graft. Moreover, some surgeons suggested encasing the entire autograft in Dacron tube to prevent dilation, a technique that is suitable only for patients who will not need autograft growth, and further follow-up is necessary to confirm the hypothetical advantages of this modification.
Management of failing autograft depends on failure mode, cusp status, and neoaortic root size. Patients with reasonably preserved cusps and regurgitation due to dilation with poor coaptation are candidates for aortic valve–sparing root replacement. Right ventricle to pulmonary artery conduit longevity following Ross is higher than for those conduits placed following repair of other congenital anomalies due to several factors, such as anatomic conduit position and infrequent incidence of branch pulmonary stenosis. Reported freedom from conduit reoperation is 90% to 95% at 10 years and 75% to 85% at 15 years. Factors associated with increased conduit reoperation include smaller conduit size, longer follow-up, and fresh or aortic homograft use. Of note, recent experience in percutaneous pulmonary valve replacement has allowed cardiologists to address this problem without surgical intervention with good immediate and mid-term results. Recent use of handmade polytetrafluoroethylene (PTFE) valved conduits in the right ventricle–pulmonary artery position has garnered some enthusiasm by selected providers (see Chapter 60 ).
Valvar Aortic Stenosis
Valvar aortic stenosis is the most common form of LVOTO, occurring in 65% to 75% of patients. Bicuspid aortic valve occurs in 1% to 2% of the population and represents 3% to 5% of all congenital heart anomalies, with the incidence in males being three to five times higher than that in females. Valvar aortic stenosis is an anatomic and clinical spectrum that ranges from one extreme of critical neonatal aortic stenosis that is associated with ductal-dependent systemic circulation (almost 10% of those patients) to another extreme of mild aortic stenosis in asymptomatic children with the only finding being an incidental murmur on physical examination.
Critical Neonatal Aortic Stenosis
The aortic valve pathology is most commonly bicuspid associated with thickened dysmorphic cusps and fused commissures. Less commonly, the valve is tricuspid with thickened cusps and fused commissures. When the valve is unicuspid, there might be a single commissure or no commissure, and the small orifice might be central or eccentric.
During fetal development, severe LVOTO exposes the left ventricle to increased afterload that results in ventricular hypertrophy with subsequent systolic and diastolic myocardial dysfunction. Ventricular hypertrophy and higher intracavitary pressure lead to decreased coronary perfusion pressure and chronic in utero subendocardial ischemia with consequent development of endocardial fibroelastosis that further impairs ventricular function. Additionally, reduced in utero antegrade blood flow through the aortic valve leads to underdevelopment of the left heart structures and hypoplasia of the mitral valve, left ventricle, subvalvar area, ascending aorta, and aortic arch. Clinical presentation varies based on the severity of LVOTO and the degree of associated ventricular hypoplasia and dysfunction. Neonates with critical aortic stenosis often have a rapidly developing and dramatic presentation after birth. As the ductus arteriosus closes, they experience decreased systemic and coronary perfusion, acute hemodynamic deterioration with cardiovascular collapse, metabolic acidosis, end-organ injury, and shock. Neonates and infants with lesser degrees of obstruction may gradually develop symptoms secondary to persistent high left ventricular afterload, left atrial hypertension, myocardial dysfunction, and poor systemic cardiac output. Symptoms such as failure to thrive, irritability, and tachypnea and increased work of breathing might necessitate early intervention during the first few months of life.
Diagnostic studies include electrocardiography (ECG), which might demonstrate evidence of left ventricular or biventricular hypertrophy, with common evidence of strain. The chest x-ray examination might show cardiomegaly and evidence of pulmonary congestion in neonates with critical aortic stenosis, but findings might be normal in those with less severe stenosis. Echocardiography is the primary form of diagnosis, and it delineates the level of obstruction, gradient across the LVOT, presence of endocardial fibroelastosis, hypoplasia of left heart structures, mitral valve anomalies, ascending aorta or aortic arch, and associated cardiac anomalies such as atrial and ventricular septal defects and patent ductus arteriosus. The direction of flow across the transverse arch and ductus might help to determine whether or not the left heart is capable of supporting the systemic circulation after biventricular repair. Of note, the LVOT gradient may be underestimated in those neonates due to severe left ventricular dysfunction, frequent presence of mitral regurgitation, and existence of patent ductus arteriosus or distal arch obstruction. Cardiac catheterization for diagnosis is usually not required but can be helpful to measure LVOT gradient directly (in older patients with less severe LVOTO) and to evaluate associated anomalies that may not be apparent on echocardiography. The role of cardiac catheterization in neonatal aortic stenosis is mainly for intervention rather than diagnosis. It is important to mention the role of fetal echocardiography in the prenatal diagnosis of neonates with critical aortic stenosis. Prenatal identification of LVOTO and associated intracardiac and extracardiac anomalies can result in better planning of the care following delivery (both for the medical team and for the family) and avoidance of circulatory collapse in patients with unrecognized pathology as a result of ductal closure and can offer the possibility of fetal intervention with aortic valvuloplasty, which, in selected cases, can be associated with better growth of left heart structures and higher likelihood of achieving biventricular repair.
In neonates with ductal-dependent systemic circulation, urgent neonatal intervention is indicated. In those with less severe stenosis, surgery is indicated when the infants develop symptoms such as failure to thrive, respiratory distress, and irritability, in conjunction with peak Doppler-derived LVOT gradient that is greater than 40 mm Hg or a peak-to-peak catheter-derived LVOT gradient that is greater than 30 mm Hg.
The medical management of neonates with postnatal diagnosis of ductal-dependent critical aortic stenosis involves rapid restoration of ductal patency with the use of prostaglandin E 1 infusion . This will restore systemic perfusion by shunting blood right to left across the ductus and can help reduce pulmonary hypertension seen with severe left ventricular dysfunction ( Fig. 51.1 ). Intubation and mechanical ventilation may also be required to aid in correction of severe acidosis and control of pulmonary hypertension. The use of inotropic drugs improves the poor ventricular contractility. Judicious administration of intravenous fluids and correction of metabolic and electrolyte anomalies are important in the resuscitation of these critically ill patients. In neonates with very severe LVOTO or in those with severely depressed left ventricular function or inadequate left ventricle or mitral valve sizes, balloon atrial septostomy might be needed to allow unobstructed pulmonary venous egress to the right side of the heart, which is supporting the systemic circulation. In the absence of atrial communication, severe pulmonary hypertension and respiratory compromise might be very rapid, necessitating extracorporeal membrane oxygenation (ECMO) support for stabilization before any intervention. On the other hand, when the degree of aortic valve stenosis seems to be less severe, and the left heart structures seem to be well developed, especially in those who seem to have adequate cardiac output generated from the left ventricle with mainly left-to-right shunt across the ductus arteriosus, a trial of discontinuation of prostaglandin can be undertaken, and intervention can be deferred if the child is growing and having no symptoms of poor cardiac output or respiratory compromise.
