Left Ventricular Outflow Tract Obstruction

Left Ventricular Outflow Tract Obstruction

William M. DeCampli and Kamal K. Pourmoghadam

The Heart Center at Orlando Health, Arnold Palmer Hospital for Children, Orlando, FL, USA

Left ventricular outflow track obstruction (LVOTO) is broadly defined as physiologically important obstruction to the passage of blood from the left ventricle (LV) to the ascending aorta. LVOTO can occur as an isolated abnormality or in conjunction with other anomalies. Additionally, LVOTO may occur at one or more locations in the left ventricular outflow tract (LVOT) and involve the ventricular septum, mitral valve apparatus, aortic valve, aortic root, or sinotubular junction. As such, the etiology and management are variable, depending on the specific anatomy and associated lesions. When significant LVOTO is present in the neonate, systemic blood flow may depend on an alternate pathway for sufficiently oxygenated blood to reach the aorta, typically via a septal defect within the heart and a patent arterial duct at the great vessel level. This scenario is commonly dubbed critical LVOTO and mandates neonatal intervention. When LVOTO is less severe, a neonate may survive following ductal closure. Frequently, however, obstruction is progressive, leading to increasing ventricular afterload, ventricular hypertrophy, coronary flow insufficiency, and ventricular failure.

The initial approach to the patient with LVOTO requires consideration of all associated anomalies. Often, the first step is to decide whether the initial management will consist of single‐ventricle or biventricular management. Usually, when LVOTO occurs in the setting of a normal or mildly hypoplastic LV and/or mitral valve, a biventricular approach can be taken. This is true whether LVOTO is an isolated anomaly or is in conjunction with other anomalies that do not preclude biventricular repair (such as double‐outlet right ventricle or transposition, interrupted aortic arch, or Shone complex). On the other hand, when LVOTO is present in conjunction with significant LV hypoplasia and/or endocardial fibroelastosis (EFE), or mitral hypoplasia or dysplasia, an initial single‐ventricle approach may be taken. In some of the latter cases, the LV may be able to undergo rehabilitation by resection of EFE or by natural growth, together with serial repairs of the mitral valve, leading ultimately to a biventricular repair after initial single‐ventricle palliation. The rehabilitation strategy is still undergoing development and trials [1, 2]. In the gray zone of LV adequacy, the initial optimal management can be difficult to ascertain. Several studies have attempted to develop single‐value metrics that inform the optimal management (single‐ventricle vs. biventricular pathway), but none has been prospectively validated. While the single‐ventricle pathway is fraught with well‐known morbidity and mortality, the biventricular pathway, in borderline cases, may also lead to multiple reinterventions and a continued hazard for morbidity and mortality [3].

In the remainder of this chapter, we will focus on three types of LVOTO that, in most cases, are remediable along the biventricular pathway. These types correspond roughly to three anatomic levels of obstruction: valvar, subvalvar, and supravalvar obstruction. While each type is frequently associated with other anomalies, we will focus on diagnosis and surgical (or interventional) management of the LVOTO itself. We will cover hypertrophic cardiomyopathy (HCM) in some detail in Chapter 31, as the management protocols, as well as the surgical techniques, have evolved to greater complexity.

Valvar Aortic Stenosis

Aortic valve anomalies, in particular bicuspid aortic valve (BAV), constitute the most common congenital cardiac anomalies. The prevalence of BAV is 1–2% of the general population [4]. These anomalies represent a wide range of pathology, from BAV to leaflet dysplasia, with or without annular hypoplasia. Consequently, the clinical picture is highly variable. With critical aortic stenosis, neonatal intervention is mandated, whereas with typical BAV, patients may not need intervention until adulthood, if ever. In this section, we divide aortic stenosis into two categories: that requiring neonatal or young infant intervention, and that needing intervention beyond infancy.

Critical Aortic Stenosis


The valve leaflets in critical aortic stenosis are often poorly delineated by echocardiography due to leaflet fusion or dysplasia. The leaflets may be so thickened that doming does not occur. Antegrade flow is limited to the effective orifice of the valve, which may be only 1–2 mm in diameter. There may be concomitant valvar and/or annular hypoplasia. Associated left‐sided structures (arch, LVOT, LV) are commonly hypoplastic. The LV may appear small but distended due to the effect of EFE. The mitral valve may or may not be hypoplastic or structurally abnormal.


The newborn baby with critical aortic stenosis has (by definition) ductal‐dependent systemic circulation. On the first day of life, the presence of cyanosis and a murmur should prompt an echocardiogram, which will confirm the diagnosis. The echocardiogram should delineate both the absolute and standardized (z‐score) anatomic dimensions of the left‐sided structures as well as the presence of associated abnormalities. The degree of EFE should be estimated.

In some cases, the diagnosis will have been made during fetal development, as it can be diagnosed at 18–20 weeks. There is some evidence that serial fetal echocardiograms may allow prediction of the initial management strategy [5]. Fetuses that exhibit flow reversal in the transverse arch and foramen ovale, monophasic flow across the mitral valve, and LV dysfunction may be more likely to proceed to single‐ventricle palliation.


The pathophysiology of critical aortic stenosis begins in fetal life. The initiating factor leading to valvar dysplasia remains unknown. As impedance to flow increases, fetal LV hypertrophy will progress to maintain relatively normal myocardial wall stress and thus ejection fraction. As impedance increases further, LV hypertrophy no longer compensates for increased wall stress. As a result, coronary insufficiency results in subendocardial ischemia and EFE. EFE, in turn, impairs diastolic function and impedes or even arrests flow‐related stimulation of LV growth. This scenario explains the spectrum of abnormal findings seen by the time of birth.


The newborn baby with critical aortic stenosis should be managed initially with prostaglandin infusion to maintain ductal patency. As a general guide, the Rhodes score, or the score developed by the Congenital Heart Surgeons Society Data Center, can aid in deciding between a single‐ventricle or biventricular initial management strategy in the neonate with critical valvar aortic stenosis [6, 7]. One should keep in mind, however, that these single‐variable scores have limited predictive power, and should be used only in conjunction with additional patient‐specific data and institutional experience [3, 8]. The critical aortic stenosis calculator and the model used to calculate this score are displayed in Table 30.1. Qualitatively, the presence of high‐grade, severe EFE, severe mitral valve hypoplasia (z≤4), and/or an LV end‐diastolic volume <20 mL/m2 may preclude biventricular management. At the other end of the spectrum, critical aortic stenosis associated with mild hypoplasia of left‐sided structures and only low‐grade or no EFE can be managed with a biventricular pathway. In the latter case, and when effective aortic orifice diameter is borderline, it is reasonable to perform a trial wean of prostaglandin. If the baby maintains adequate systemic perfusion, breathes comfortably, and feeds, then intervention may be delayed. It is generally agreed that intervention is undertaken when the peak Doppler gradient is greater than 30–40 mmHg (after ductal closure). In any case, intervention should certainly be undertaken before LV dysfunction develops, as mid‐term outcomes of valvotomy are not favorable when LV dysfunction is present preoperatively [9]. Initial management consists of either surgical or catheter‐based balloon valvotomy [10]. If the aortic annulus is hypoplastic but borderline, it is reasonable to attempt valvotomy as the initial procedure, realizing that the patient may require earlier reintervention such as the Ross–Konno operation or stage 1 palliation. The latter two options must be considered as the initial intervention in the case of a severely hypoplastic annulus, with or without EFE. The hypoplastic annulus is discussed below in more detail.

Surgical Valvotomy

The technique has been well described in a multimedia format by Hraška, Photiadis, and Arenz [11]. Surgical valvotomy is best performed on cardiopulmonary bypass (CPB) with mild hypothermia. The patent arterial duct is mobilized and temporarily snared. Cardioplegic arrest is achieved. A vent placed through the right superior pulmonary vein, patent foramen ovale, or left atrial appendage facilitates a blood‐free field. The foramen ovale is left patent. An oblique or transverse aortotomy is performed, and stay sutures are placed (Figure 30.1A).

Table 30.1 Incremental risk factors for time‐related death for patients who had an initial procedure indicating an intended biventricular repair pathway.

Covariate Estimate p value
Intercept –.484 <.001
Presence of moderate or severe tricuspid regurgitation –.279 <.001
Z‐score of mitral valve annulus +.030 <.001
Presence of large ventricular septal defect –.312 <.001
Length of apex‐forming ventricle (cm)a +.715 <.001
Minimum diameter of the left ventricular outflow tract (cm)b +.892 <.001
Presence of left ventricular dysfunctionc +.230 <.001
Grade of endocardial fibroelastosisd +.165 <.001
Diameter of the mid‐aortic arch (cm)e –.187 <.001

Risk factors in normal font represent those identified as predictors for death after univentricular repair and those in italicized font represent those identified as predictors for death after biventricular repair.

a Echocardiographic measurement from crux of the heart to the apex, regardless of whether formed by the left or right ventricle. Indexed to height of the patient.

b Minimum diameter measured at any point from the subvalvular region as far distal as the brachiocephalic artery. Indexed to body surface area and entered after inverse transformation.

c Presence of any degree of left ventricular dysfunction, including mild.

d Endocardial fibroelastosis was graded subjectively by the echocardiographic appearance of left ventricular endocardial brightness and thickening as follows: 0 = none; 1 = involvement of papillary muscles only; 2 = papillary muscle with some endocardial surface involvement; 3 = extensive endocardial surface involvement.

e Measured immediately proximal to the left subclavian artery and indexed to the body surface area.

Source: Reproduced by permission from Hickey EJ et al. J Thorac Cardiovasc Surg. 2007;134:1429–1437.

It is important to inspect the valve carefully. The most common anatomy is the appearance of a bicuspid valve with a false raphe, with some degree of commissural fusion and thickening of the leaflets (Figure 30.1B). It is also common to see nodules on various parts of the leaflets, including the hinge points. All of these abnormalities contribute to reducing the effective orifice area. The first maneuver is to divide the fused portion of leaflets where the commissure is well formed using a #11 blade (Figure 30.1C). The incisions are made carefully so as to completely separate the leaflets and enter slightly into the commissural tissue. Care is taken not to lower the height of the commissures, which increases the chance of regurgitation. Next, nodules should be shaved off at all locations, including the leaflet hinge points. Finally, leaflets should be thinned meticulously using a knife. It is tempting to expedite this last step, but instead time should be taken to precisely thin all thickened portions of the leaflets. In some cases, one may appreciate a dissection plane between the fibrosis and true leaflet tissue. In any case, care is taken to avoid injury to the thinned leaflet, as a durable repair of a leaflet injury is very difficult to achieve in a neonate. The orifice should be gently checked with a dilator. This, too, must be done carefully so that the leaflet is not torn at its weakest point. At completion, the aorta is closed (Figure 30.1D) and the patient weaned from bypass in the usual way. The arterial duct is ligated unless cardiac output is inadequate. Transesophageal echocardiography is performed. The goal of the operation is to achieve a peak instantaneous Doppler gradient <30 mmHg with trivial or no aortic regurgitation.

Balloon Valvotomy

Balloon valvotomy is usually performed in the catheterization laboratory, although the procedure has been performed under echocardiographic guidance alone [12]. The surgeon’s role is limited to providing vascular (carotid artery) access, repairing injured vessels, or standing by to provide cardiopulmonary support in very ill neonates. A balloon is gently inflated until the valve is disrupted and the effective orifice is increased. In the ideal case, the gradient is reduced to 35 mmHg or less, with trivial or no regurgitation. The result is considered unacceptable if the gradient is >35 mmHg and/or there is moderate or greater regurgitation [13].

Results of Valvotomy

In neonates, both surgical and balloon valvotomy are palliative procedures, and further interventions are considered to be normal. Both approaches produce comparable early results [10, 14]. In low‐risk patients early mortality averages 5%, whereas in higher‐risk patients mortality may exceed 20%. Indeed, early mortality is mainly associated with coexisting factors such as mitral stenosis, small LV, small aortic annulus, EFE, and depressed LV function [1416]. Late survival is also influenced by these risk factors and ranges from 70% to 90% at 1 and 10 years. In a single‐center study from 2012, survival in babies with isolated aortic stenosis undergoing surgical valvotomy was 91% at 20 years [17]. Of these patients, 93% had a postoperative gradient <30 mmHg and no more than mild regurgitation. In this study patients with trileaflet valves had better freedom from reintervention for recurrent aortic stenosis and from reoperation due to regurgitation (90%) than patients with unicuspid or bicuspid valves (50%). Overall, freedom from aortic valve replacement was 92%, 83%, 68%, and 57% at 5, 10, 15, and 20 years, respectively. In another single‐center study published in 2016, 83 infants undergoing surgical valvotomy were followed for a median of 4.2 years [18]. Freedom from death was 87% and 85% at 5 and 15 years, respectively. Time‐related survival without reintervention was 51%, 35%, and 18% at 5, 10, and 15 years, respectively. Freedom from aortic valve replacement was 67%, 54%, and 39% at 5, 10, and 15 years, respectively. In a 2015 study of 30 infants undergoing catheter‐based balloon valvotomy, early mortality was 13%, with one of these deaths associated with moderate/severe postprocedure regurgitation [12]. Of survivors, 8% were left with moderate or severe regurgitation. At a mean follow‐up of 9 years, 15 of 26 patients required reintervention. The 10‐ and 15‐year survival was 82%.

Schematic illustration of (A) Initial surgical exposure for aortic valvotomy, which shows aortic cross-clamp, cardioplegia (not shown), and intended aortotomy site.

Figure 30.1 (A) Initial surgical exposure for aortic valvotomy, which shows aortic cross‐clamp, cardioplegia (not shown), and intended aortotomy site. (B) Transverse aortotomy reveals a bicuspid aortic valve, which is fused at the commissure. (C) Sharp aortic valvotomy is being performed to enlarge the left ventricular outflow tract. (D) Valvotomy is completed and the aortotomy closure is commenced. Air maneuvers follow before cross‐clamp is removed. Ao, aorta; LCA, left coronary artery; NCC, noncoronary cusp; PA, pulmonary artery; RA, right atrium; RCA, right coronary artery; RV, right ventricle; SVC, superior caval vein. Source: Mavroudis C, Backer CL 2003 / with permission of Elsevier.

Which Initial Palliation Is Better?

There is currently no published prospective, randomized study comparing surgical valvotomy to balloon valvotomy for critical aortic stenosis. Thus, factors such as details of risk adjustment, patient selection, institutional bias, and practitioner experience must be carefully considered when comparing studies. Critics of balloon valvotomy assert that it commonly ruptures the valve at its weakest point, which is frequently a normal part of a leaflet [16]. Thus, the effective orifice area is significantly enlarged, but at the expense of some degree of regurgitation. They further argue that, as opposed to residual aortic stenosis, which can remain stable for a long time and may be intervened on by balloon enlargement, aortic regurgitation is frequently progressive, leading to aortic valve repair or (more commonly) replacement as the only option. Proponents of surgical valvotomy argue that in surgical valvotomy incisions are made only in locations of inappropriate fusion, thus preserving individual leaflet function. Proponents of balloon valvotomy assert that, with increased experience, the effective orifice area can be enlarged without producing more than mild regurgitation, especially with a trileaflet valve [15]. They also argue that balloon valvotomy avoids the risk of CPB and ischemic arrest in what are frequently very ill neonates.

In 2016, Benson reported a contemporaneous comparison of results from two institutions, each committed to one of two strategies [15]. There were 79 neonates, 52 undergoing balloon valvotomy and 27 undergoing surgical valvotomy. Baseline echocardiographic data were generally matched. Early survival was similar at the two institutions, and survival was 91% and 82% for balloon and surgical valvotomy, respectively (p=.15), at both 1 and 5 years. At 5 years, freedom from reintervention was 78% for surgical valvotomy and 52% for balloon valvotomy (p=.09). At 10 years, freedom from reintervention was 16% for surgical and 30% for balloon valvotomy. At 3 years there was a trend for greater need for aortic valve replacement in the balloon valvotomy group (p=.06). For the two groups combined, the hazard function for reintervention had two distinct phases, the first showing a high early risk that declined rapidly over the next 2 years, and the second showing a continuous, gradual rise in hazard from 3 out to 12 years. Like many such comparisons, this study did not definitively point to the superiority of one technique over another.

Critical Aortic Stenosis Associated with Severe Annular Hypoplasia

Critical aortic stenosis may be accompanied by severe annular hypoplasia (annular diameter z‐score <–4). This combination is more commonly found in association with other left‐sided anomalies, such as arch hypoplasia or hypoplastic LV. Valvotomy alone will not be effective in the presence of severe annular hypoplasia. In this case the surgeon must make the decision whether to proceed with initial single‐ventricle palliation, or perform aortic root enlargement and replacement in the neonatal or infant period. In cases where the LV and mitral valve are adequate, the biventricular route is justifiable and consists of the Ross, Ross–Konno, or, in cases where the pulmonary autograft cannot be used, the classic Konno procedure using homograft. Another indication for root replacement in the infant period is symptomatic aortic valve regurgitation resulting from initial valvotomy with or without subsequent balloon dilations.

In the neonate or infant, the procedure is similar to that described for older children with a few modifications. In the infant with annular hypoplasia without diffuse LVOTO, the Konno (septal) incision can be limited to what is needed to effectively enlarge the annulus (Figure 30.2). Generally a running suture can be used. Care is taken to tailor the thickness of the autograft root to avoid obstruction, and also to tailor the length of the infundibular tongue to avoid buckling (too long a tongue) or flattening of the graft (too short a tongue). As long as the LV and mitral valve are adequate, and arch obstruction has also been corrected, perhaps the most significant risk factor for death after infant Ross–Konno is preoperative ventricular dysfunction. Thus, it is important that the decision to proceed with this procedure is made before significant deterioration of LV function has occurred.

Results of Infant Ross/Ross–Konno

In most published series the morbidity and mortality after neonatal/infant Ross–Konno are significantly greater than those in older children. Many of the patients had complex disease and underwent additional procedures at the time of operation. Lo Rito and colleagues published a series of 22 patients <18 months of age undergoing the Ross/Ross–Konno [19]. Three patients died early and all three had poor preoperative LV function. There was one late death at five‐year follow‐up. Among 18 long‐term survivors, 8 patients had mild and 2 patients had moderate autograft regurgitation. Freedom from aortic reoperation was 84% at 10 years. Nelson and coauthors published a series of 240 pediatric Ross operations [20]. There were 43 infants, with an infant operative mortality of 18%, compared to 4% for older children. Long‐term survival in infants was 72% at 15 years, compared to 87% for older children. LVOT reintervention was required in only 2 of 35 survivors. Freedom from right ventricular outflow tract (RVOT) intervention at 15 years was only 19% in infants, compared to 53% for older children.

Schneider and coworkers reported results of 48 patients who underwent the Ross–Konno procedure [21]. Median age was 12.8 months; 22 patients were under 1 year of age. Early and late mortality was 12.5% and 8.3%, respectively. Poor LV function was a risk factor for early mortality. Five patients required aortic reoperation at a median of 14 years postoperatively. Aszyk and colleagues reported mid‐term results of the Ross–Konno procedure in 16 infants [22]. There were no early deaths and one late death, for a Kaplan–Meier survival of 93.3% (61.2–99.0, 95% confidence interval [CI]) at 5 years. At 5 years, no patient had greater than trivial aortic regurgitation. Freedom from RVOT reoperation was 53% at 5 years. Maeda and coauthors reported results of the Ross–Konno in 24 infants [23]. There was one early death and no late mortality at median follow‐up of 81 months. Freedom from mild or greater aortic regurgitation was 74.7% ± 12.9% at 5 years. Two patients required subsequent aortic valvuloplasty. Freedom from reintervention was 36.9% ± 11.3% at 5 years, with 14 of 23 reinterventions being on the RVOT (conduit repair or replacement). Using the Society of Thoracic Surgeons Congenital Heart Surgery Database, Woods and coworkers reported an analysis of 160 infants who underwent aortic valve replacement between 2000 and 2009 at 47 institutions [24]. Hospital mortality was 18% overall, and 28% for neonates. Mortality for those patients undergoing root replacement using homograft was 40%. Postoperative mechanical support was required for 17 (11%) patients. Concomitant aortic arch repair was associated with greater hospital mortality.

Schematic illustration of (A) Neonatal or infant Ross–Konno procedure.

Figure 30.2 (A) Neonatal or infant Ross–Konno procedure. Incision lines are demonstrated. (B) After aortic cross‐clamping, antegrade and retrograde cardioplegic arrest, aortic resection, coronary mobilization, and autograft harvest, the left ventricular outflow tract is enlarged by a localized linear incision without entry into the right ventricular cavity. (C) The neoaortic left ventricular outflow reconstruction. The right ventricular outflow is preserved without a septal incision and without the moiety that is required for the “apron extension.” (D) The completed Ross–Konno procedure after separation from cardiopulmonary bypass. Ao, aorta; PA, pulmonary artery; RV, right ventricle. Source: Mavroudis C, Backer CL 2003 / with permission of Elsevier.

In summary, most, but not all, published series of the Ross–Konno in infants report significant morbidity and mortality. The variability in results most likely reflects differences in risk distribution, including (i) preoperative clinical risks, especially LV dysfunction; (ii) presence of other levels of left‐sided inadequacy that may affect the decision as to which operation to perform, as well as an understanding of the complexity of the operation; and (iii) experience of the clinical team in formulating and performing the optimal management scheme. Overall, the Ross–Konno operation continues to be the most promising approach to the infant in need of aortic valve replacement when preservation of biventricular circulation is preferred.

Fetal Aortic Valvuloplasty

Fetal catheter intervention on the aortic valve is an evolving strategy. The surgeon may be called upon to advise on the indication for fetal intervention or to intervene postnatally on a patient who has had fetal intervention.

The concept was developed over 30 years ago, and the largest single‐institution experience is at Boston Children’s as of 2017. The major indication for the procedure has been severe aortic stenosis in the presence of physiologic aberrations consistent with evolving hypoplastic left heart syndrome (HLHS), thus excluding cases of apparent isolated critical aortic stenosis, where there is a high likelihood of biventricular repair without fetal intervention. This group published their results of 100 consecutive patients in 2014 [25]. Whereas they achieved technical success with the procedure, the number of patients going on to single‐ventricle management remained significant. Additionally, among the 31 patients in whom biventricular circulation was maintained, 27 underwent subsequent aortic valve repair or replacement. While the article is a landmark in the development of this intriguing therapeutic option, the study had no control group; therefore, the effectiveness of the procedure in maintaining biventricular circulation could not be determined. The following year, a report of the International Fetal Cardiac Intervention Registry showed that between January 2001 and June 2014, and among 18 institutions, 245 patients underwent fetal intervention among 370 cases considered for intervention, including 186 aortic valvuloplasties (100 of these were from the original Boston cohort) [26]. Among only live‐born infants, those with a technically successful intervention had a 42.9% (24 of 56) prevalence of discharge with biventricular circulation, whereas those without intervention or with a technically unsuccessful intervention had a 19.4% (6 of 31) prevalence of discharge with biventricular circulation. Thus, this study contained a control group of sorts. However, it was not randomized and was likely biased.

The question has arisen whether fetal valvuloplasty, by potentially augmenting cerebral oxygen delivery, could partially mitigate the detrimental effects in neurodevelopment seen in patients with HLHS. The Boston group has recently reported neurodevelopmental outcomes among 52 patients who underwent fetal aortic valvuloplasty [27]. Patients who underwent the procedure had outcomes similar to patients with HLHS without fetal intervention. Achievement of a biventricular circulation was not associated with better outcomes.

Aortic Valve Stenosis beyond the Newborn Period


The embryology of aortic valve stenosis in this population is similar to that of patients who present in the neonatal period with critical aortic stenosis. A complex abnormal development of the endocardial cushions may lead to either fusion or incomplete formation of the cusps or subsequent fusion of initially separate cusps during fetal development [28]. The resultant abnormality leads to varied aortic valve cusp and commissural development with varying degrees of stenosis and/or insufficiency.


Aortic valve stenosis beyond the neonatal period is congenital in nature, representing one or the other phenotypic forms of BAV in more than 85% of cases [29]. This is especially true if a functionally unicommissural valve with multiple raphes is considered a subcategory of BAV [30]. A pure BAV is formed by having two cusps both functionally and morphologically, without the presence of a raphe. This form is the least common, accounting for 5–7% of cases, with the remainder having either one or two raphes forming a functionally bicuspid aortic valve [30, 31]. The predominant form of BAV shows a raphe between the intercoronary commissure in more than 70% of cases, with the least common showing a single raphe between the left and the noncoronary cusp, occurring in less than 3% of patients. The remaining forms of aortic valve stenosis can be delineated as dysplastic in 11%, and true unicommissural valves in 4% of cases [29].


In patients who present beyond the neonatal period without critical aortic stenosis, LV hypertrophy is gradual and may take years to develop. They generally have an effective aortic valve orifice that allows for enough stroke volume to support the baseline systemic cardiac output. Endocardial fibroelastosis, a hallmark of critical aortic stenosis in neonates resulting from subendocardial ischemia, infarction, and fibrosis, is usually not seen in these patients.

However, if the stenosis is important or progressive, the increasing impedance to the flow of blood from the LV and resultant increase in intraventricular pressure lead to an ongoing stimulus for LV concentric hypertrophy. The increase in LV wall thickness initially compensates for the increase in intracavitary pressure and allows for normalization of the ejection fraction and maintenance of adequate stroke volume. As the severity of stenosis progresses, despite ongoing compensatory LV hypertrophy, the ejection fraction and stroke volume will eventually decrease and fail to meet the necessary cardiac output. This happens as the preload reserve is exhausted and the sarcomeres reach their maximum diastolic length, then further increase in afterload leads to decrease in stroke volume, resulting in an afterload mismatch. Subendocardial perfusion becomes compromised as systolic and diastolic LV intracavitary pressures exceed that of the aorta, decreasing myocardial and especially subendocardial perfusion. Eventually, the combination of ongoing subendocardial hypoperfusion, limited cardiac output, reduced coronary perfusion, and increased myocardial oxygen consumption leads to myocardial ischemia and development of symptoms. These include angina, ventricular dysrhythmias, exercise intolerance, and syncope.

Clinical Features

BAV represents the most common congenital cardiac malformation, with an incidence of 0.9–2% in the general population, and a 65% male predominance [31]. Beyond the newborn period, the degree of aortic valve stenosis and LV hypertrophy determines the pathophysiology of the aortic stenosis and the timing of symptom presentation. Within the first year of life, patients can present with a history of irritability, inadequate feeding, and poor weight gain. Those presenting during the first year of life constitute only 10–15% of patients diagnosed with congenital aortic stenosis.

Beyond infancy, presentation can be similar to that of adults with aortic valve stenosis. Those with a mild to moderate degree of aortic stenosis are generally asymptomatic, with mild LV hypertrophy and normal LV function. Usually, they are referred for evaluation of a systolic ejection murmur by the pediatrician. However, as the aortic stenosis progresses, the classic triad of dyspnea on exertion, angina, and syncope may manifest. Chest pain with exertion may be the only presenting symptom, or it can accompany dyspnea on exertion and/or syncope in 30% of patients with important aortic stenosis.

When evaluating a patient for aortic stenosis with BAV, the presence of associated defects should be assessed. BAV occurs as an isolated lesion in 50–60% of patients [32]. The most common associated defects are the left‐sided obstructive lesions (65%), such as coarctation or interruption of the aorta, and HLHS [31]. Right‐sided obstructive lesions in patients with BAV are rare. Although aortic stenosis and aortic insufficiency are the most commonly associated anomalies with BAV, dilatation of any or all segments of the proximal aorta from the aortic root to the arch can occur in 30–50% of patients. This is referred to as bicuspid aortopathy, and is mild in a majority of patients. Rarely patients develop progressive dilatation that produces complications and places them at risk for sudden death [33].


Patients presenting beyond the newborn period are generally referred for evaluation of a systolic ejection murmur or a click. This is typically heard at the right upper sternal border. A thrill in the suprasternal notch is common and helpful to localize attention to the aortic valve. A precordial thrill is less common, and if present usually indicates significant aortic stenosis. In a low cardiac output state or severe aortic stenosis, the ejection murmur may be absent. Thus, pale and cool periphery with diminished pulses consistent with a narrow pulse pressure may be the presenting signs.

The electrocardiogram is often normal in the presence of a mild to moderate degree of aortic stenosis; in severe circumstances, it can show evidence of LV hypertrophy and possibly LV strain.

Transthoracic echocardiography is the gold standard for the initial diagnosis and subsequent follow‐up evaluation in the progression of aortic stenosis. It can evaluate the aortic valve morphology, the presence of accompanying aortic regurgitation, and LV function, and demonstrate other accompanying congenital cardiac defects. The severity of aortic stenosis can be assessed by directly measuring the aortic stenosis jet velocity from continuous‐wave Doppler tracings through the aortic valve [34]. However, if there is any discrepancy in diagnosis by echocardiography, cardiac catheterization can measure pressure change between the ascending aorta and the LVOT, either by simultaneous measurement or pull‐back technique. It can also confirm echocardiographic findings of associated defects, quantify any intracardiac shunt, and assess valvar competence.

Future Directions in Diagnosis and Evaluation

The role of cardiac magnetic resonance imaging (MRI) in diagnosing aortic stenosis in children is becoming more important [35, 36]. Its appeal may be due to avoidance of radiation exposure, and the ability to obtain both anatomic and hemodynamic data. Although the assessment of maximum peak gradient in patients with aortic stenosis is feasible by MRI, it remains an underestimation when compared to that measured by echocardiography [37]. Computed tomography (CT) scanning plays a limited role in the diagnosis of aortic stenosis, although it can provide the best anatomic assessment of the aortic valve leaflet and annulus in the presence of calcification.

Progress is being made to obtain a more granular evaluation of regional and global LV function by the use of two‐dimensional speckle‐tracking echocardiography (2DSTE) [38]. Studies are underway to evaluate patient myocardial deformation patterns using 2DSTE to better assess more subtle changes in LV remodeling and function when compared to conventional echocardiography. This is gaining interest in follow‐up evaluation of aortic stenosis patients after undergoing aortic balloon valvuloplasty or repair [39, 40].

Indications for Intervention

Beyond infancy, aortic valvuloplasty is indicated regardless of the gradient across the aortic valve if there is isolated valvar aortic stenosis with depressed LV function or if there is resting peak systolic valve gradient of ≥50 mmHg by catheterization. If the patient has symptoms of angina or syncope or ST‐T–wave changes consistent with ischemia at rest or exercise with a peak systolic catheter gradient of ≥40 mmHg, intervention is also indicated [41].

The definition of severe aortic stenosis is considered to be based on the natural history studies of unoperated aortic stenosis patients who show poor prognosis once the peak velocity across the aortic valve is >4.0 m/s2, corresponding to an echocardiographic mean gradient of >40 mmHg [42]. Although guidelines for intervention are based on cardiac catheterization data, the majority of patients in the pediatric population are initially evaluated by echocardiography prior to intervention. As such, in our institution, in the absence of diminished LV function, generally intervention is considered when echocardiographic mean aortic valve gradient is >50 mmHg. This decision is not taken lightly, as the consequences of poor outcome carry significant morbidity and mortality.


Some of the most complex decisions have to be made for newborns presenting with critical aortic stenosis regarding whether a two‐ or single‐ventricle palliation is appropriate. These strategies are discussed in the section on neonatal critical aortic stenosis.

Medical/Intervention beyond the Newborn Period

Patients who present beyond the newborn period with congenital aortic stenosis most often undergo balloon aortic valvuloplasty for their initial palliation. However, once intervention is indicated, there is growing controversy concerning what the optimal therapeutic strategy should be.

A landmark analysis by the Congenital Heart Surgeons’ Society database in 2001 demonstrated no significant difference in mid‐term outcomes for survival and freedom from reintervention between patients undergoing transcatheter balloon valvotomy versus surgical aortic valvuloplasty in 110 neonates from 18 institutions [14]. The potential for catch‐up growth of the aortic valve and the LV after balloon aortic valvuloplasty over time was also demonstrated [43]. Procedural success for balloon valvuloplasty has been shown in multiple single‐center analyses and more recently in a multicenter evaluation demonstrating a better than 70% rate of success, defined as a residual peak systolic gradient ≤35 mmHg with no more than mild aortic regurgitation [44]. In reviewing this data and considering balloon aortic valvuloplasty a relatively low‐risk procedure, many centers have opted for catheter intervention as the primary therapeutic strategy in children presenting with congenital aortic stenosis.

This trend is being challenged for multiple reasons. Catheter intervention is shown as an effective procedure with low mortality for congenital aortic stenosis; however, there is growing evidence of significant long‐term risks of aortic insufficiency and need for reintervention [45, 46]. In a 15‐year follow‐up of more than 150 patients undergoing balloon aortic valvuloplasty for AS, an estimated 32% of neonates and 44% of non‐neonates remained free from reintervention, while an estimated 45% of neonates and 62% of non‐neonates remained free from aortic valve replacement [45].

Predictors of reintervention included higher pre‐ and post‐dilatation gradient, neonatal age, and increasing aortic insufficiency post dilatation. This high rate of late reintervention somewhat diminishes the impact of short‐term success with balloon aortic valvuloplasty.

In addition, surgical techniques of aortic valve repair have improved over the past two decades beyond simple commissurotomy or blind dilation. Recent analyses of long‐term surgical outcomes of aortic valvotomy for congenital aortic stenosis show promising evidence of lower reintervention rates and resultant aortic insufficiency [47]. Freedom from reintervention at 10 years following surgical aortic valvotomy is reported to be over 70%, versus just over 50% in those undergoing balloon aortic valvuloplasty [48]. However, these data originate from single‐center retrospective studies, with their inherent limitations.

A meta‐analysis of the literature compared outcomes after balloon valvuloplasty and surgical valvotomy for congenital aortic stenosis. The analysis included over 2000 patients from 20 studies undergoing balloon aortic valvuloplasty and surgical valvotomy. Kaplan–Meier analysis showed no difference in long‐term survival or freedom from aortic valve replacement. However, consistent with previously reported findings, freedom from reintervention at 10 years was only 46% for balloon aortic valvuloplasty versus 73% for surgical valvotomy [49].

Although the recent reports show similar short‐term results for surgical valvotomy and balloon aortic valvuloplasty, the long‐term results show a significantly higher rate of reintervention after balloon valvuloplasty, overall and in neonates. Yet balloon aortic valvuloplasty is considered a less invasive procedure and the long‐term survival remains similar between the two strategies. At the current time, the reported information supports either approach as the initial therapeutic strategy for congenital aortic stenosis based on institutional preference.

To date, our understanding of the optimal initial strategy for treatment of congenital aortic stenosis is derived from multiple single‐ and multicenter retrospective reviews. Studies are underway to improve patient selection for each strategy, such as evaluating the relation of aortic valve leaflet morphology and resultant aortic insufficiency or residual gradient after balloon aortic valvuloplasty [50]. Improving patient selection may be helpful in improving outcomes; however, the limitations of the current reports support the undertaking of a randomized multicenter trial for comparing the two therapeutic strategies and clarifying the strengths and weaknesses of each approach.

Surgical History

Over the past century many surgeons laid the groundwork for the current level of technical expertise in the care of our patients with aortic stenosis. The first surgical aortic valvotomy is credited by some to Tuffier who, in 1913, reported an attempt to dilate a stenotic aortic valve by invaginating the wall of the aorta from above [51]. Significant progress in aortic valve surgery was made much later in the 1950s and 1960s. Bailey reported on mechanical dilatation of the aortic valve through the LV apex in 1952, and, after gaining experience via this technique, he recommended the use of a blind transaortic approach in 1954 [52, 53]. Using hypothermia and inflow occlusion, Julian and Swan separately reported on successful direct aortic commissurotomy in 1956 [54, 55]. In the same year, Lillehei reported correction of aortic stenosis using complete cardiac bypass with retrograde coronary perfusion [56]. Further progress by Coran and Bernhard enabled treatment of critical aortic stenosis in neonates and infants throughout the 1960s [57].

Aortic valve surgery was firmly established by reporting on implantation of the ball valve prosthesis by Harken and colleagues in 1960 [58], and Starr and colleagues in 1963 [59]. The ball valve was the precursor of many other mechanical and bioprosthetic valves. Barratt‐Boyes and colleagues in New Zealand and Ross and colleagues at Guy’s Hospital in London performed the first orthotopic allograft valve insertions separately, in 1962 [60, 61]. Later, in 1967, Ross and colleagues introduced the pulmonary autograft as a replacement for the aortic valve [62].

Numerous surgical techniques for enlargement of a hypoplastic aortic valve annulus have been described. Nicks and colleagues reported on posterior enlargement of the aortic valve annulus in 1970 [63], while Konno in 1975 described the anterior approach of root enlargement in two patients, one an infant [64]. Another, more complex posterior enlargement technique was first reported performed by Rastan in 1978 [65], and published by Manouguian in 1979 [66]. Due to controversy as to the origin of this concept, it may require combined credit as the Rastan–Manouguian technique [67, 68].

A more recent important contribution was to combine the anterior aortoventriculoplasty of Konno and the implantation of the pulmonary autograft as a single technique for replacing the aortic valve and relieving the LV outflow tract obstruction, the Ross–Konno operation [69].

Parallel to progress in aortic valve surgery, percutaneous balloon aortic valvuloplasty was developed in the 1980s. Its role was initially confined to those who were critically ill and considered to be poor candidates for surgical intervention. However, its indication for use has expanded and its role for neonates and children was described earlier. Fetal intervention to perform balloon aortic valvuloplasty in the hope of achieving biventricular repair in those with critical aortic stenosis is another area of interest and the subject of ongoing study [70]. Its application has been limited, mainly due to the presence of risk to the fetus and mother.

Interest in transcatheter aortic valve implantation has increased since Cribier first described the technique in 2002 [71]. The evolution of this procedure has been rapid and is still ongoing. Currently its use in children is confined to its implantation into a failing bioprosthetic valve. However, an isolated report of its use in the neoaortic position in a Fontan patient with severe aortic insufficiency may be a sign of its evolving indication for use in children [72].

Surgical Techniques

Although balloon aortic valvuloplasty is the procedure of choice for congenital aortic stenosis in most institutions, there is an increasing number of advocates for surgical aortic valvuloplasty as the initial intervention, as surgical techniques of aortic valve repair have improved over the past two decades. However, open valvotomy with inflow occlusion and closed valvotomy will not be discussed in this chapter because they are of historical value only. Consideration and concerns for balloon aortic valvuloplasty are previously discussed in the Medical/Intervention beyond the Newborn Period section.

Surgical Valvotomy and Valvuloplasty

Surgical aortic valvuloplasty usually will include a combination of procedures rather than employment of a single technique. The stenotic aortic valve can create turbulence, with abnormal flow patterns that may lead to leaflet thickening and fibrous nodular formation. Thus, surgeons should be prepared to utilize not only commissurotomy, leaflet thinning, and/or removal of fibrous nodules, but possibly other more complex accompanying repair techniques (Figure 30.3) [47].

Schematic illustration of surgical techniques of aortic valve repair.

Figure 30.3 Surgical techniques of aortic valve repair. Techniques of surgical repair have evolved and always include an extensive debridement of the valve by resection of all nodular dysplasia, thinning of the leaflets, and recreation of an interleaflet triangle by resecting the fibrosis, immobilizing the commissure. Incised portions of stenotic valves are now resuspended with pericardial patches to avoid regurgitation. Source: [47] / with permission of Elsevier.

CPB is established by cannulating the aorta high and obtaining venous drainage through a single right atrial cannulation. Bicaval cannulation is employed if an associated intracardiac lesion is to be repaired. The left heart is drained via an LV vent established through the right superior pulmonary vein/left atrial junction. Single or multidose cardioplegia technique is employed for myocardial arrest with a mild to moderate degree of hypothermia. A hockey‐stick incision is made and carried deep into the noncoronary sinus with extension of the transverse incision to the left for exposure. If the aortic valve is bicuspid, a wide transverse incision can be employed for visualization. In the case of significant accompanying aortic insufficiency, the aorta can be cross‐clamped shortly after establishing CPB, and after the aortotomy direct antegrade cardioplegia can be delivered to each coronary ostium independently via a handheld device.

Morphologic evaluation from transesophageal or three‐dimensional (3D) echocardiography can help guide planning of the procedure. However, anatomic detail is best assessed once the aortic valve is exposed, and commissural traction sutures placed. Valve morphology is carefully examined. If necessary, blade commissurotomy of the fused leaflets can be performed to within 1–2 mm of the aortic wall to prevent loss of leaflet support at the hinge points. Leaflet thinning and removal of fibrous nodules are performed as needed for better free edge coaptation. The goal is not to increase the degree of aortic insufficiency while relieving aortic stenosis, as increasing aortic insufficiency after repair is a risk factor for early valve replacement.

Most stenotic valves are functionally bicuspid. If adequate mobility of the leaflet is not secured via simple commissurotomy, an aggressive commissurotomy can be performed by separating the fused leaflets into the aortic wall with release and mobilization of the subcommissural triangle. This provides a greater degree of mobility to the leaflet and a longer free edge for coaptation. These maneuvers, however, can result in a decrease in leaflet support at the hinge points and a higher likelihood of aortic insufficiency at long‐term follow‐up. Accompanying procedures such as leaflet extension and resuspension may be necessary (Figure 30.4). If debridement of the aortic valve or removal of an immobilizing fibrous nodule at the commissure results in loss of tissue, the commissure can be reconstructed and the interleaflet triangle restored with pericardial patches, as described by Siddiqui and Haydar (Figure 30.3) [47, 73].

A functionally bicuspid aortic valve with significant commissural fusion can be converted to a tricuspid valve with an increase in leaflet mobility and relief of aortic stenosis. This decision should be considered carefully during the preoperative assessment based on the institutional experience, as a complex aortic valve repair may be needed. The fused commissure is debrided and divided to relieve aortic stenosis. Nearly all patients will require leaflet extension and resuspension to prevent aortic insufficiency (Figure 30.4) [74]. Surgical experience is important in these repairs, as sudden death due to ischemia can occur from repaired leaflet prolapse occluding the coronary orifice [75]. In certain bicuspid valves, a rudimentary commissure that does not quite reach the aortic wall may be present and difficult to distinguish from a true fused commissure. This false raphe does not provide a true hinge point and support on the annulus. The surgeon must keep in mind that division of a false raphe, without leaflet extension and resuspension, will lead to aortic insufficiency due to lack of commissural support at the hinge point.

Patients with unicommissural valves are quite uncommon, as stated before. Again, a conservative valvotomy is advisable, especially in the presence of a fused commissure. If necessary, recreation of a new commissure can be performed by the addition of pericardial patches to form a bicuspid valve, as described by Brizard [76].

Schematic illustration of (A–H) Technique of aortic cusp extension with fresh autologous pericardium.

Figure 30.4 (A–H) Technique of aortic cusp extension with fresh autologous pericardium. Source: [74] / with permission of Elsevier.

Aortic Valve Replacement

As discussed earlier, for patients who undergo early intervention for aortic stenosis, freedom from reintervention at 10 years is approximately 50% after balloon aortic valvuloplasty, and 70% after surgical valvuloplasty. Most of these patients will require reintervention after 15 years. A select group of patients may respond well to reintervention with balloon aortic valvuloplasty after an initial balloon aortic valvuloplasty or surgical valvotomy. However, complications such as the need for aortic valve replacement, surgical valvotomy, and death can occur in as many as 40% of these patients [77]. At the same time, by delaying valve replacement using balloon aortic valvuloplasty, more options become available and a larger prosthesis can be implanted.

Bioprosthetic valves in the aortic position have limited durability in children. Although appropriate annular sizes may be available, both porcine and bovine pericardial valves suffer from accelerated degeneration and calcification in the younger age group. Homograft valves in the aortic position routinely calcify and also require frequent reoperations, rendering them less than optimal in this age group.

Mechanical valves do not suffer from limited durability; however, a lifelong commitment to warfarin anticoagulation and surveillance is required. The ongoing thromboembolic and hemorrhagic risk with mechanical valve anticoagulation in a young patient diminishes its advantage in durability. This factor becomes especially important when considering the use of a mechanical valve in young women of child‐bearing age, or children who are still growing. A recent clinical trial (PROACT) demonstrated that reduced anticoagulation (international normalized ratio [INR] 1.5–2.0) is safe with the new On‐X™ aortic valve (On‐X Life Technologies, Austin, TX, USA) [78].

Aortic Annulus Enlargement

Although mechanical and bioprosthetic valves with smaller annular sizes have become available, annular enlargement is frequently still necessary to avoid prosthesis–patient mismatch during aortic valve replacement [79].

Posterior Annular Enlargement: Nicks and Rastan–Manouguian Procedures

Only gold members can continue reading. Log In or Register to continue

May 18, 2023 | Posted by in CARDIOLOGY | Comments Off on Left Ventricular Outflow Tract Obstruction

Full access? Get Clinical Tree

Get Clinical Tree app for offline access