Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Congenital aortic stenosis is a cardiac anomaly in which narrowing at valvar, subvalvar, supravalvar, or combined (multiple) levels results in a systolic pressure gradient between the inflow portion of the left ventricle (LV) and the aorta beyond the obstruction. A spectrum of defects involves the aortic root, with some overlap of abnormalities. Congenital aortic stenosis in neonates and infants may be part of the constellation of hypoplastic left heart physiology (see “Coarctation as Part of Hypoplastic Left Heart Physiology” under Morphology in Section I of Chapter 48 ). This association is highly relevant to therapy.
Congenital valvar aortic stenosis includes defects in which the major malformation involves the aortic valve cusps. An obstruction at valve level is caused by imperfect cusp development with cusp thickening and fusion. Cusp abnormalities can be severe in early life; when they are not, important obstruction may not develop until later in life when calcification occurs. This chapter discusses congenital valvar aortic stenosis only in the age range from birth to young adult life.
Congenital valvar aortic stenosis has been long recognized by morphologists. Initial efforts to find a surgical solution were made by Carrel and Jeger, who independently attempted experimentally to place conduits between left ventricular apex and aorta. In 1955, Marquis and Logan reported surgical treatment using dilators introduced through the LV apex, as did Downing in 1956. Also in 1956, valvotomy was performed by an open technique during inflow stasis with moderate hypothermia induced by surface cooling. The first report of its treatment by accurate valvotomy during cardiopulmonary bypass (CPB) was by Spencer and colleagues in 1958, although this had been performed at the Mayo Clinic in 1956 and was reported by Ellis and Kirklin in 1962.
The prevalences and precise nature of the various types of morphology of severely stenotic aortic valves and coexisting anomalies are incompletely understood for several reasons. First, it is difficult to obtain enough cases to constitute a reasonable sample of the spectrum of severe congenital valvar stenosis in neonates, infants, and children. Second, sources of data range from autopsy studies and surgical studies to echocardiographic and cineangiographic imaging in patients undergoing balloon valvotomy. Third, differing terminology contributes to incompleteness of morphologic data; ideally, not only should the morphologic nature of the cusps be defined, but also that of the sinuses of Valsalva, commissures (upper points of attachment of cusps to the aortic wall), and interleaflet triangles (fibrous or muscular tissue interposed between the sinuses in the subcusp LV outflow tract), as described by Angelini and colleagues for bicuspid valves.
In patients with stenosis severe enough to require operation in infancy or childhood, the valve is bicuspid in about 65% ( Table 47-1 ). The valve usually consists of thickened right and left cusps in association with anterior and posterior commissures and a slitlike orifice with its long axis in the sagittal plane ( Fig. 47-1, A ). The left cusp is frequently the larger and may contain a transversely placed central thickened ridge, or buttress, representing a rudimentary commissure between normal right and left cusps. Less often, the two cusps are anterior and posterior, and the orifice is then oriented in a coronal plane. There is usually fusion peripherally of one commissure and occasionally of both. However, severe stenosis can occur without fusion, resulting only from thickened cusps and a bicuspid configuration. If free edges of both thickened cusps are taut, they are then equal in length to the diameter of the aortic root and cannot open ( Fig. 47-1, B ). Most bicuspid valves will show three intercusp triangles on their ventricular side, indicating that three cusps were present in the developing valve. A bicuspid valve with only two definitive cusps is uncommon and usually is not stenotic early in life, rather presenting later in life with obstruction or regurgitation. A full discussion of the genetics, morphology, natural history, and therapeutic options for adult patients with a congenitally bicuspid aortic valve is found in Chapter 12 .
Valve | No. | % of 290 |
---|---|---|
Bicuspid | 186 | 64 |
Tricuspid | 89 | 31 |
Unicuspid a | 15 | 5 |
T otal | 290 | 100 |
a Frequency may be underestimated because “bicuspid” valve with rudimentary commissures might better be termed unicuspid.
In about 30% of patients the valve is tricuspid, with three thickened cusps of approximately equal size and three recognizable commissures that are fused peripherally to varying degrees, creating a dome with a central stenotic orifice. This type of valve is more favorable for valvotomy because all three commissures can usually be opened.
Less often (5%) the valve may have a unicuspid configuration with only one commissure ( Fig. 47-2 ). This variety is more common in infants presenting with severe stenosis. Occasionally, however, the stenosis is not severe, and signs and symptoms develop in later life as the valve thickens and calcifies. A thickened unicuspid valve is inherently stenotic, whether the commissure is fused or not, unless the cusp is particularly redundant.
The cusps are approximately equal in size in only a small percentage of congenitally bicuspid or stenotic tricuspid valves. Likewise, the raphae are variable in thickness and length. Number of sinuses may not be the same as number of cusps; most congenitally bicuspid and unicuspid (unicommissural) valves have three sinuses and three intercusp triangles.
Diffuse cusp thickening, most marked at the free cusp edges, contributes importantly to valvar stenosis. Thickening is more extreme in symptomatic neonates and infants, and cusps may be irregular and myxomatous or dysplastic in appearance. In addition, particularly in infants, the aortic “anulus” may be small and stenotic, especially with unicuspid valves, and frequently in association with other components of hypoplastic left heart physiology such as endocardial fibroelastosis of a hypoplastic left ventricle ( Fig. 47-3 ), coarctation of the aorta, patent ductus arteriosus, and mitral regurgitation or stenosis.
The LV is always concentrically hypertrophied in children with severe aortic stenosis, but in infants hypertrophy may be extreme, with a tiny cavity and extensive fibrosis in the wall ( Fig. 47-4 ). Fibrosis is primarily in the subendocardial region. Extensive endocardial fibroelastosis may also be present, possibly the result of ischemia of subendocardial layers. In these hearts, the ventricle may be dilated ( Fig. 47-5 ).
Congenital valvar aortic stenosis may be associated with a fibrous subvalvar or supravalvar stenosis, as well as with coarctation of the aorta.
The various coexisting components of hypoplastic left heart physiology include LV hypoplasia of varying degrees, extreme LV hypertrophy with small cavity size, endocardial fibroelastosis, congenital mitral stenosis or regurgitation, severe coarctation, and subaortic stenosis caused by mitral valve abnormalities ( Table 47-2 ). Patent ductus arteriosus or ventricular septal defect (VSD) as well as pulmonary atresia may also be present.
Type of Stenosis | Associated Anomaly | No. of Cases |
---|---|---|
Valvar ( n = 78) | Isolated PDA | 8 |
PDA + ASD | 3 | |
PDA + coarctation | 3 | |
Isolated coarctation | 3 | |
Isolated VSD | 1 | |
PDA + coarctation + VSD | 1 | |
Left SVC | 1 | |
T otal | 20 (26%) | |
Subvalvar (n = 41) | Isolated PDA | 2 |
PDA + coarctation | 1 | |
Coarctation + congenital mitral stenosis | 1 | |
VSD + important PS | 3 | |
Unroofed coronary sinus syndrome | 1 | |
Left SVC + single coronary orifice | 1 | |
AR (3 mild, 2 important) | 5 | |
VSD | 4 | |
T otal | 18 (44%) | |
Supravalvar (n = 10) | Pulmonary artery stenosis | 1 |
Left SVC | 1 | |
T otal | 2 (12%) | |
Combined ( n = 7) | PDA + coarctation | 1 |
Congenital mitral stenosis | 1 | |
VSD | 1 | |
T otal | 3 (43%) |
a Data from 142 patients, UAB experience 1967-1982. Categories are mutually exclusive. Numbers in parentheses are percentages of the n of the type.
Neonates and infants with severe valvar aortic stenosis usually present with pallor, perspiration, and inability to feed. Shortness of breath and cyanosis may be present. In children and young adults, even important stenosis may be without symptoms. However, effort dyspnea, effort angina, or effort syncope, singly or in combination, usually indicates a severe lesion. Dyspnea may be present with moderate stenosis.
In neonates and infants with severe valvar aortic stenosis, the most striking feature is small pulse volume with pallor, dyspnea, and at times cyanosis. Both the murmur and gradient across the valve may be unimpressive because of a low cardiac output. There also may be a hyperactive right ventricular impulse.
Clinical signs in children and young adults include an ejection systolic murmur (and thrill) at the base radiating to the carotid vessels, accompanied by a systolic ejection click. An aortic diastolic murmur is uncommon, particularly when compared with patients with discrete subvalvar stenosis. A severe lesion is characterized by palpable pulse of low volume and slow upstroke, single or reversed splitting of second heart sound, apical fourth and sometimes third heart sound, and thrusting LV impulse.
Many investigators have concluded that physical signs are unreliable in assessing severity of valvar stenosis in children. However, physical signs can be used to differentiate among mild, moderate, and severe lesions in most patients, and severe lesions can always be distinguished from mild ones.
The electrocardiogram (ECG) usually shows severe LV hypertrophy but can be near normal. Right ventricular hypertrophy on the ECG may be associated with a left-to-right shunt at atrial level through a stretched patent foramen ovale and rarely a reversed shunt at ductus level.
The ascending aorta frequently is prominent in older children but is small in neonates and infants. Increased heart size is seldom seen except in neonates and infants in heart failure, in whom it may be marked. Radiologically demonstrable valvar calcification is rare in patients younger than age 25.
Two-dimensional echocardiography has become particularly important as a diagnostic tool. In neonates and infants, morphology and severity of narrowing of the valve and size, wall thickness, and contractility of the LV can be assessed. The congenitally stenotic aortic valve can be continuously reevaluated by Doppler ultrasound measurement of flow velocity across stenotic valves, which can be used to quantify transvalvar pressure gradient. Echocardiographic markers are useful for managing fetuses with important aortic stenosis. Serial measurements of fetal cardiac size and function may predict postnatal outcome.
Estimation of subendocardial oxygen requirements may be helpful in assessing severity of stenosis.
A systolic gradient across the aortic valve can be demonstrated at cardiac catheterization, usually through a retrograde aortic approach if possible or otherwise a transseptal approach. Cardiac output can also be measured so that valve area can be calculated. Systolic gradient greater than 75 mmHg or valve area less than 0.5 cm 2 · m −2 is indicative of severe stenosis (see “ Summary ” in text that follows). Measuring gradient and valve area at cardiac catheterization has become less important as echocardiographic measurements have become more accurate. A raised LV end-diastolic pressure indicates LV failure or a fibrotic and noncompliant ventricle.
Angiography demonstrates thickened leaflets that form a dome in systole, with a localized jet of contrast entering the aorta ( Fig. 47-6 ). Although this type of study is not reliable in assessing severity of stenosis, it can assess size of the aortic “anulus” and LV. An aortic root injection allows quantification of aortic regurgitation if present.
By a combination of clinical and hemodynamic assessments (echocardiography, cardiac catheterization), patients with congenital valvar aortic stenosis can be categorized as having mild, moderate, or severe obstruction. Mild implies that pulse volume and contour are normal, as is the second heart sound. Patients with these findings have an LV-aortic systolic pressure difference less than 40 mmHg at rest, with a mean of 20 mmHg. Patients with moderate stenosis have an abnormally small pulse volume on palpation and abnormal contour, and narrow inspiratory splitting of the second heart sound may be present. Such patients generally have systolic gradients less than 75 mmHg, with a mean of 20 to 50 mmHg. Patients with severe stenosis have a systolic gradient in excess of 75 mmHg and an abnormal pulse volume and contour, as well as a single second heart sound or reverse splitting. These patients have a mean calculated aortic valve area index of less than 0.5 cm 2 · m −2 .
Congenital valvar aortic stenosis is three to four times more common in males than in females and occurs in about 5% of Caucasians with congenital heart disease.
When neonates and infants present with valvar stenosis, the lesion is typically severe, with rapidly progressive heart failure and death within a few days to a few weeks of birth. Thus, most neonates and young infants come to intervention (currently with percutaneous balloon valvotomy) critically ill and in New York Heart Association (NYHA) class IV or V ( Table 47-3 ). Many have other anomalies associated with the spectrum of the hypoplastic left heart physiology. Ten Harkel and colleagues noted a 5-year survival rate of 73% among patients presenting in infancy.
Age at Operation | NYHA Functional Class (mean ± SD) b | |||||
---|---|---|---|---|---|---|
≤ | Age | < | Valvar (n = 78) | Subvalvar (n = 41) | Supravalvar (n = 16) | Combined (n = 7) |
Weeks | ||||||
1 | 4.8 ± 0.45 | |||||
1 | 4 | 4.1 ± 0.94 | ||||
Months | 4.0 | |||||
1 | 3 | 4.5 ± 0.71 | ||||
3 | 12 | 3.0 ± 1.26 | ||||
12 | 48 | 2.2 ± 0.98 | 3.2 ± 0.96 | 1.0 | 3.0 | |
Years | ||||||
4 | 12 | 1.7 ± 0.76 | 1.8 ± 0.94 | 1.7 ± 0.52 | 2.0 ± 1.00 | |
12 | 20 | 1.7 ± 0.73 | 1.8 ± 1.19 | 1.7 ± 0.58 | 2.5 ± 0.71 | |
20 | 1.9 ± 0.74 | 2.8 ± 0.84 |
a Data from 142 patients, UAB experience 1967-1982.
b Patients were categorized into NYHA classes I to V, with class V indicating those undergoing emergency operation because of shock or metabolic acidosis. In the current era, isolated valvar aortic stenosis in newborns and infants is generally treated by percutaneous balloon valvotomy.
When symptoms are delayed beyond age 1 year, heart failure is rare, and survival without treatment generally is prolonged. Also, associated anomalies are less common. The Second Natural History Study of Congenital Heart Defects includes data on many patients treated for valvar aortic stenosis and followed for 25 years. Patients were 2 years or older at entry into the study, and 40% managed medically subsequently required surgical management. For patients presenting with LV-aortic pressure gradient greater than 50 mm Hg, 70% required surgical intervention. Almost 40% of patients required a second operation.
Survival is related to (1) sudden death in untreated children and (2) rate of progression of stenosis.
Occurrence of sudden death varies between 1% and 19% of patients. Of 58 patients younger than 35 years old who died suddenly and were found to have congenital heart disease, three (5%) had aortic valve stenosis. Analysis of the literature and of a series of 218 patients with congenital valvar stenosis indicates that sudden death directly attributable to aortic stenosis is virtually confined to patients with a severe lesion. Sudden death in patients with no symptoms and normal physical findings except for the murmur of aortic stenosis has not been documented. Sudden death may occur in patients with a normal ECG, but this finding is not incompatible with severe stenosis. Thus, the true prevalence of sudden death in children and adolescents in whom surgery is deferred until the lesion is considered severe on clinical grounds is probably about 1%.
When congenital aortic valvar deformities are nonobstructive in infancy and childhood, less than 10% progress to mild obstruction within about 10 years. Leech, Mills, and colleagues obtained information on 26 patients aged 1 week to 29 years when first seen, and in whom diagnosis of nonobstructive aortic valve deformity was made based on an isolated aortic ejection sound. During a 5- to 16-year follow-up, two patients (7%; CL 2%-16%) developed signs of mild stenosis after 7 and 15 years. As more years pass, an undetermined time-related proportion of patients with deformed (usually congenitally bicuspid) aortic valves develop progressive thickening and calcification and ultimately important stenosis. Vollebergh and Becker suggest that minor inequality of size of tricuspid valves present from birth may lead to formation of senile or degenerative type of aortic valve stenosis presenting in the seventh or eighth decade of life.
When mild stenosis is present at first evaluation in childhood, progression is more rapid. Moderate or severe stenosis develops in about 20% of patients within 10 years and in 45% within about 20 years ( Fig. 47-7 ). Even after this long interval, therefore, 55% of the mild lesions remain mild.
When moderate stenosis is present initially, the lesion becomes severe within 10 years in about 60% of patients ( Fig. 47-8 ).
Spontaneously occurring infective endocarditis appears in less than 1% of patients. The reported incidence is 1.8 to 2.7 episodes per 1000 patient-years. Infective endocarditis may produce aortic regurgitation and may be a cause of death.
Percutaneous balloon valvotomy is often used for treating congenital valvar aortic stenosis, but a description of this technique is beyond the scope of this text. Its place in treating neonates is discussed under Special Situations and Controversies later in this section.
Closed techniques and surface cooling for hypothermic circulatory arrest have largely been replaced by aortic valvotomy on CPB using cold cardioplegic myocardial management.
Anesthetic and supportive management must be precise (see Section II of Chapter 4 ). Drifting downward of body temperature to 32°C to 34°C is probably advantageous. Preparation and median sternotomy are described under “Preparation for Cardiopulmonary Bypass” in Section III of Chapter 2 . As the pericardium is being opened, care is taken to touch the heart as seldom as possible because ventricular fibrillation is easily provoked. The purse-string suture is placed for the aortic cannula, the patient heparinized, and the cannula inserted and connected to the arterial tubing. Only then is a purse-string suture placed around the right atrial appendage; if the heart fibrillates, CPB can be established in less than a minute. A single venous cannula is inserted, and CPB is begun with the perfusate at 34°C; the ductus arteriosus is ligated, perfusate taken to 20°C to 28°C, aorta clamped, cold cardioplegia administered, and the perfusate temperature is then taken to 34°C to 36°C. In the presence of aortic regurgitation to a degree that interferes with cardioplegia delivery, cardioplegia can be delivered via small olive tip catheters directly into the coronary ostia, or coronary sinus cardioplegia can be used.
A transverse aortotomy is made ( Fig 47-9, A ). Two stay sutures are placed on the upstream side of the aortotomy for exposure. The aortic valve is inspected to determine which of the commissures to incise. Only partially formed commissures should be incised; commissurotomy should not be performed where there is only a rudimentary raphe ( Fig. 47-9, B ).
Valvotomy is performed by dividing fused commissures with a knife to within 1 mm of the aortic wall; in neonates the cusps are often gelatinous, but every effort still should be made to identify these commissures. It is important that even tension be placed on the two adjoining cusps so that incision is precise ( Fig. 47-9, C ). Only commissures with adequate cusp/commissural attachment to the aortic wall are opened, because division of rudimentary commissures produces regurgitation. Incisions are deepened in stages, and cusps on each side are evaluated for competence and lack of prolapse before each further incision. If further incising of the commissure might cause cusp prolapse, the incision is carried no further. Occasionally, myxomatous nodules can be excised from the cusp's free edge, or fibrous thickening can be shaved off the ventricular aspect of one or more cusps. The aortotomy is then closed with a continuous suture. If there is any degree of aortic regurgitation by saline filling of the aortic root, a 6-0 polypropylene suture (Frater stitch) can be placed through the midpoint of each cusp and brought out through the closed aortotomy. The suture is removed when LV contraction begins.
The remainder of the procedure is completed as usual (see “Completing Cardiopulmonary Bypass” in Section III of Chapter 2 ). A left atrial catheter should be positioned before discontinuing CPB. If the neonate is of suitable size for placing a transesophageal echocardiography (TEE) probe, the repair is evaluated before and after discontinuing CPB. LV and aortic pressures are measured and recorded before closing the chest.
Preparation for operation, the incision, and preparations for CPB are described under “Preparation for Cardiopulmonary Bypass” in Section III of Chapter 2 . Using a single venous cannula or caval cannulation, CPB is established at 28°C. A left atrial or LV vent is used. External cardiac cooling may be applied. The aorta is clamped and cold cardioplegic solution infused. The transverse aortotomy is made, and stay sutures applied to edges of the incision for exposure. Commissurotomy is performed as described.
The aortotomy is closed by continuous stitches, and the rest of the operation is completed as usual (see “Completing Cardiopulmonary Bypass” in Section III of Chapter 2 ).
Pericardial cusp extension valvuloplasty procedures (as described by Duran) to compensate for deficiency of valvar tissue and increase the coaptation surface area have been applied primarily to regurgitant aortic valves (see Chapter 12 ). More recently, encouraging midterm results with these techniques applied to congenital aortic stenosis support their inclusion as surgical options. Aortic cusp extension valvuloplasty may be considered as an adjunctive procedure to primary open valvotomy or in reoperative situations with recurrent aortic stenosis or regurgitation following previous valvotomy. Before proceeding with valve reconstruction, preoperative echocardiographic studies should ascertain the absence of LV subaortic obstruction.
Following a median sternotomy, autologous pericardium is harvested, thoroughly cleaned of all fatty tissue and adhesions, treated with 0.625% glutaraldehyde solution for 3 to 5 minutes, and kept moist with normal saline. CPB, LV venting, and myocardial management are performed as described under “Valvotomy in Older Infants, Children, and Adults.”
An oblique aortotomy is made, and the aortic valve evaluated for presence of complete but fused commissures and a raphe in congenitally bicuspid aortic valves. Each cusp is examined for thickness and mobility, free-edge irregularities, and tissue deficiency.
The valve is prepared for cusp extension by first thinning the thickened cusp edges. Fused commissures are incised out to the aortic wall, and subcommissural fusion or scar tissue is released to maximize cusp mobility. Bicuspid valves with a rudimentary raphe are tricupidized by incising through the fused cusp at the raphe all the way to the aortic wall ( Fig. 47-10 ). The pericardial patches are each cut to a length determined by the diameter of the aorta, supplemented with an additional 15% to 20% to account for later pericardial shrinkage. Height of each patch is chosen to extend the line of coaptation of the repaired cusps about 5 mm higher than the highest cusp and to bring the extended cusps into a coaptation point in the center of the valve orifice. The pericardial extensions are sutured to each cusp with continuous 5-0 polypropylene, beginning in the center of the cusp and working toward the commissures. The extensions are attached to the aortic wall, creating neocommissures at the level of the sinutubular junction. Ilbawi and colleagues recommend leaving a little excess pericardial patch at the commissural level, secured with a pledgeted mattress suture through the aortic wall. With all the patch extensions in place, the newly constructed extensions are trimmed to provide a uniform cusp height and symmetric coaptation surface (see Fig. 47-10 ). Adequacy of the valve opening is examined, and initial valve competence is assessed by filling the aortic root with saline. Aortotomy closure and discontinuation of CPB are conducted as usual. Evaluation of valve function by TEE is performed before and following discontinuation of CPB. If aortic regurgitation is more than mild or if peak transvalvar gradient by TEE exceeds 30 mmHg, consideration should be given to reestablishing CPB and revising the valve repair or, if improvement is not feasible, proceeding with valve replacement.
When viewed at operation, the congenitally stenotic aortic valve may be too extensively deformed to be opened and remain reasonably competent. However, this situation is rare in primary operations in patients younger than age 10 and uncommon in those younger than 20. It is more common when multiple prior balloon valvotomies have been performed, and particularly when progressive or recurrent stenosis is accompanied by moderate or worse aortic regurgitation.
Aortic valve replacement in older children may be done in a standard fashion using a mechanical prosthesis (see “Isolated Aortic Valve Replacement” under Technique of Operation in Chapter 12 ). It may also be performed in the standard freehand manner using an aortic valve allograft (see “Allograft Aortic Valve” under Technique of Operation in Chapter 12 ) or a pulmonary valve autograft (see “Autograft Pulmonary Valve” under Technique of Operation in Chapter 12 ). Because the aortic root and LV-aortic junction may be quite small in young children who require aortic valve replacement, aortic root enlargement (see “Root-Enlarging Technique” under Technique of Operation in Chapter 12 ) or replacement (see “Replacement of Aortic Valve and Ascending Aorta, En Bloc” under Technique of Operation in Chapter 12 ) may be advantageous. An aortic allograft can be used as the replacement device, or a pulmonary autograft may be preferred. The autograft has the advantage of remaining unchanged and uncompromised by host reaction, and it also may grow.
Postoperative care after aortic valvotomy or other procedures discussed in this section are conducted in the manner generally used after intracardiac operations (see Chapter 5 ).
Whenever valvotomy is performed for congenital valvar aortic stenosis, long-term follow-up is indicated because of possible recurrence of stenosis requiring reoperation.
Hospital mortality for surgical treatment of congenital valvar aortic stenosis in heterogeneous groups of patients younger than 20 to 25 years of age is largely an unhelpful value because of the important role of incremental risk factors for death and the selection processes by which treatments (or no treatments) are chosen.
Mortality varies widely among patient subsets, with few or no deaths after valvotomy in children and young adults. Mortality is higher in neonates, but the potential safety of an open approach in neonates with severe congenital aortic stenosis has been demonstrated. Still, the current preference in most centers is initial percutaneous balloon valvotomy. Multiple centers have achieved hospital mortalities of 15% or less in neonates. Again, however, mortality figures in heterogeneous groups of patients, even if all are neonates, are difficult to interpret, as demonstrated by Gaynor and colleagues and a Congenital Heart Surgeon's Society (CHSS) analysis. In contrast to many situations in cardiac surgery, nearly all deaths after operation for congenital valvar aortic stenosis occur early postoperatively, most within 48 hours.
Success in salvaging such patients with emergency temporary extracorporeal membrane oxygenator (ECMO) or left ventricular assist device (LVAD) support has not been fully evaluated, but such support is advisable in the face of progressive circulatory failure (see “Treatment of Low Cardiac Output” in Chapter 5 ).
Hospital mortality after operations for congenital valvar aortic stenosis in patients older than age 1 year approaches zero.
Overall survival up to 40 years is good after the primary operation for congenital valvar aortic stenosis in older infants and children. In very ill neonates and young infants, however, survival is compromised, primarily by high early risks.
A CHSS study of 320 neonates with critical aortic stenosis noted 1- and 5-year survival of 72% and 70%, respectively, among those receiving an initial procedure aimed at biventricular repair ( Fig. 47-11 ).
Almost all early deaths are in acute cardiac failure, and theoretically most should be preventable by (1) stabilization of critically ill neonates and others (with ECMO or LVAD support if needed) so that operation is not performed in NYHA class V patients as it was in the past, and (2) proper myocardial management. In neonates and young infants, however, many deaths result from (1) failure to appreciate the importance of coexisting components of the spectrum of the hypoplastic left heart physiology (see “Coarctation as Part of Hypoplastic Left Heart Physiology” under Morphology in Section I of Chapter 48 ), and (2) nonoptimal selection, in particular of a biventricular pathway rather than a single-ventricle pathway, at least in the present state of knowledge (see Indications for Operation later in this section).
Deaths occurring late after operation are in various modes, and inferences are made with difficulty. Thus “sudden death” has been reported as the mode of death in 12% of patients included in one long-term follow-up study, but the majority had severe residual or recurrent stenosis or severe aortic regurgitation. Among neonates for whom staging of repair is necessary for a univentricular pathway, few late deaths now occur between the cavopulmonary shunt stage and completed Fontan.
Left-sided cardiac defects (components of the spectrum of the hypoplastic left heart physiology, such as small aortic valve diameter, aortic hypoplasia, severe endocardial fibroelastosis, LV hypoplasia, extreme LV hypertrophy with small cavity size, and congenital mitral valve disease ) are associated with high mortality after operation ( Table 47-4 ). These coexisting major cardiac anomalies, poor preoperative functional class, and young age at admission tend to occur together, and all are risk factors.
Risk Factor | Coefficient ± SE a | P | |
---|---|---|---|
Intended Biventricular Repair | |||
Higher | Grade of endocardial fibroelastosis b | 0.53 ± 0.23 | .02 |
Lower | Aortic valve diameter z score at level of sinuses of Valsalva | 0.36 ± 0.109 | <.001 |
Younger | Age at entry c | 1.49 ± 0.53 | .005 |
Initial Norwood Procedure | |||
Smaller | Diameter of ascending aorta d | 0.95 ± 0.40 | .02 |
Presence of moderate or severe tricuspid regurgitation | 0.86 ± 0.43 | .05 |
a Single early hazard phase (see Fig. 47-11 , B ).
b Graded subjectively by echocardiographic appearance of left ventricular endocardial brightness and thickness: 0, None; 1, involvement of papillary muscles only; 2, papillary muscle with some endocardial surface involvement; 3, extensive endocardial surface involvement.
The important study by Karl and colleagues from Melbourne, Australia, emphasized the major role of these coexisting important cardiac anomalies in the early postsurgical mortality in neonates. No deaths (0%; CL 0%-19%) occurred after open valvotomy in neonates with no coexisting anomaly or only a patent ductus, whereas early mortality was 47% (CL 39%-62%) among those with important coexisting cardiac anomalies.
Advanced symptoms, or NYHA class IV and particularly class V, are associated with a considerably increased risk of death early after operation. Thus, for patients preoperatively in NYHA class I or II (most older infants and children), 15-year survival (all deaths, including those in hospital) after the primary valve operation is about 90%. In preoperatively very ill neonates and young infants, 10-year survival is about 30%. However, the risk of death in the constant hazard phase (after about 5 years postoperatively) is no greater in this group than in older patients.
These ideas came from an era when critically ill neonates and young infants were not resuscitated preoperatively by the infusion of prostaglandin E 1 (PGE 1 ). This risk factor can be neutralized at a cost of only about 5% mortality among neonates by management that includes stabilization on PGE 1 and usually low-dose inotropic support.
In a few patients, a truly unicuspid or severely dysplastic bicuspid valve may be essentially uncorrectable. In a few patients, the very small aortic anulus may prevent a satisfactory outcome. Again, these situations usually are found in very sick neonates and young infants. Currently such patients typically are managed by a staged protocol leading to a univentricular repair.
Very young age at operation is associated with a high risk of early death postoperatively. However, in the past, most patients coming to operation as neonates have been in NYHA functional class IV or V. It is important to recall, however, that with contemporary medical management, survival after operation is possible in critically ill neonates and young infants.
Most surviving patients, including those who have had reoperations, are in NYHA class I or II ( Table 47-5 ). Objective evidence of improvement in functional capacity is provided by Whitmer and colleagues, who demonstrated marked regression of exercise-induced ST depression 1 year after operation, as well as an increase in mean total work and peak exercise systolic blood pressure.
ECG evidence of LV hypertrophy may persist after valvotomy or valve replacement either because of residual stenosis or regurgitation or a progressive secondary cardiomyopathy. Intraoperative damage to the LV or preexisting ischemic myocardial fibrosis exacerbated by delaying operation can contribute to or cause this condition. Usually, however, LV hypertrophy is reversible.
Preoperative inordinate LV hypertrophy and wall thickness often found in children with congenital valvar aortic stenosis often regresses after successful valvotomy or valve replacement, which reduces LV afterload and increases systolic function.
Pressure gradient usually is substantially reduced after valvotomy and persists for 5 to 10 years. Thereafter, the gradient tends to rise steadily, occurring earlier and more frequently when valvotomy was necessary during the neonatal period or in infancy. In patients with a good initial result, the later rise in gradient is mainly the result of progressive cusp immobility and calcification. Recurrence and progression of LV-aortic pressure gradient is usually an indication for reintervention with either percutaneous balloon valvotomy or operation.
Following aortic cusp extension procedures with autologous pericardium, long-term durability of repair has been incompletely studied. Alsoufi and colleagues have emphasized the importance of satisfactory relief of aortic stenosis at the time of operation. Among 22 children who underwent this procedure, those with postoperative peak echocardiographic gradients of less than 30 mmHg had stabilization of their peak gradient over the next 2 to 3 years. However, progressive worsening of aortic stenosis was noted in those with early gradients exceeding 30 mmHg (moderate or greater aortic stenosis).
Important aortic valve regurgitation is uncommon after valvotomy when the operation has been performed as described (see Technique of Operation earlier in this section). Moderate to severe regurgitation without residual stenosis is present at late follow-up in about 10% of patients, but some regurgitation is combined with moderate or severe residual or recurrent stenosis in an additional 15% to 20%. Postoperative regurgitation occurs more frequently when valvotomy is radical, and particularly when an attempt is made to convert a bicuspid into a tricuspid valve.
The incidence of endocarditis is not lessened by valvotomy and may even be somewhat higher than in the natural history.
Rarely, patients with severe valvar aortic stenosis who undergo surgical or balloon valvotomy as neonates or in infancy develop severe diastolic heart failure years later. Robinson and colleagues at Boston Children's Hospital reported four such patients who presented 14 to 19 years after balloon valvotomy with heart failure and severe diastolic dysfunction. All had evidence of a confluent layer of LV subendocardial hyperenhancement demonstrated by gadolinium-enhanced magnetic resonance imaging that was documented by histopathology in two patients to be endocardial fibroelastosis (EFE). One patient experienced clinical improvement following aortic valve replacement and extensive EFE resection. Robinson and colleagues hypothesize that EFE may result from early (possibly in utero) and irreversible myocardial damage induced by subendocardial ischemia secondary to persistent pressure overload with decreased ventricular flow, which may gradually progress irrespective of relief of LV outflow obstruction.
As with time-related freedom from death, time-related depictions of freedom from reintervention and aortic valve replacement are of limited value when they are derived from a heterogeneous population.
In general, however, about 85% to 95% of children and young adults (excluding neonates and infants) are free of reintervention (usually valve replacement) for at least 10 years after the initial operation ( Figs. 47-12 through 47-14 ). Then, although constant in the intermediate term, the hazard function (rate of reintervention) begins to rise. By 20 years after initial operation, only 60% of patients will be free of reintervention, and by 40 years only 10% will be free. The older the patient, the more likely it is that the reintervention will be valve replacement.
Reintervention appears to be required at a shorter interval and in greater prevalence when the initial intervention has been performed in neonatal life or infancy, and is more likely to consist of valvotomy than valve replacement. Greater frequency of reintervention may be related to generally higher residual gradients in these cases. Reintervention appears to be required more frequently when initial valvotomy has been performed by some method other than an open operation using CPB. Reintervention rate increases 15 years after operation from 0.73% per year to 2.3% per year ( P < .0001).
Procedures done at reoperation are generally more varied than at initial operation ( Table 47-6 ). A satisfactory repeat valvotomy is sometimes possible, especially when the initial operation has been done in infancy. At times, an overlooked second level of obstruction is found that requires treatment such as patch graft supravalvar enlargement, a Konno procedure (see Technique of Operation in Section II ), or an aortic root replacement (see “ Aortic Valve Replacement in Children ” under Technique of Operation earlier in this section). These reoperations carry a low risk, but generally carry greater risk than primary operation (see Table 47-6 ).
Category | Procedure | Prior Procedure | n | Hospital Deaths |
---|---|---|---|---|
Valvar | Aortic valvotomy | Aortic valvotomy | 2 | 0 |
AVR | Aortic valvotomy | 2 | 0 | |
AVR + patch ascending aorta | Aortic valvotomy | 3 | 0 | |
AVR + Konno | Aortic valvotomy | 1 | 0 | |
AVR + Manougian aortic root enlargement | Aortic valvotomy | 2 | 0 | |
10 | 0 | |||
Subvalvar | Excision | Excision | 1 | 0 |
Excision + myotomy + modified Konno | Excision | 1 | 0 | |
Myotomy + modified Konno | Excision | 1 | 0 | |
Excision + patch ascending aorta | Excision | 1 | 0 | |
AVR + excision + patch ascending aorta | Excision | 1 | 0 | |
LV-Ao conduit | Excision | 1 | 0 | |
6 | 0 | |||
Supravalvar | Excision of subvalvar stenosis | Ascending aortic patch | 1 | 0 |
Aortic valvotomy + myotomy + ascending aortic patch | Ascending aortic patch | 1 | 0 | |
2 | 0 | |||
Combined valvar + supravalvar | Excision of subvalvar stenosis | Ascending aortic patch | 1 | 1 |
AVR + ascending aortic patch | Valvotomy + ascending aortic patch | 1 | 1 | |
2 | 1 | |||
Combined valvar + subvalvar | Valvotomy + excision of accessory mitral tissue | Valvotomy | 1 | 0 |
AVR + Konno | Valvotomy + excision of subaortic stenosis | 1 | 0 | |
AVR + excision of fibromuscular stenosis and accessory mitral tissue + Konno | Valvotomy + excision of subaortic stenosis | 1 | 0 | |
3 | 0 | |||
Combined valvar + subvalvar + supravalvar | Patch ascending aorta + LV-Ao conduit | Excision of subaortic stenosis | 1 | 0 |
a Data from 24 patients undergoing reoperation anywhere after initial operation at UAB, 1967-1982.
Freedom from further reoperation following aortic valve cusp extension procedures has been variable, with freedom from subsequent aortic valve repair or replacement of 60% to 80% at 5 years and about 50% at 15 years. Durability of repair appears greater if a tricuspid valve can be created.
In neonates and young infants with severe congenital valvar aortic stenosis, medical treatment is begun on an emergency basis. When the diagnosis is suspected before transport of a neonate to a cardiac surgical center in the first week or two of life, or as soon as such a patient, usually moribund or in metabolic acidosis, is admitted, prostaglandin E 1 is begun (see Indications for Operation in Chapter 49 ). This substance usually opens the ductus arteriosus, particularly if the neonate is just a few days old, improves systemic oxygenation, and relieves metabolic acidosis because the right ventricle can support both systemic and pulmonary circulations. The child's condition should be stable and good before operation is begun.
Before intervention, care must be taken to distinguish the neonates or very young infants with isolated severe congenital valvar aortic stenosis from those whose anomaly is part of the spectrum of hypoplastic left heart physiology. When the anomaly is hypoplastic left heart physiology class III (see Table 48-1 in Chapter 48 ), the Norwood operation rather than aortic valvotomy is indicated; simple aortic valvotomy is futile. The criteria for using the more extensive operation are (1) mitral valve area less than 4.75 cm 2 · m −2 ; (2) LV inflow dimension less than 25 mm; (3) small LV, evidenced by a ratio between the apex-to-base dimension of the LV and that of the right ventricle of less than 0.8; or (4) transverse cavitary and aortic “anular” dimension of 6 mm or less.
In the Congenital Heart Surgeons multi-institutional study of decision making based on 362 neonates, greater intermediate-term survival was obtained by a strategy of an initial Norwood procedure vs. two-ventricle strategy if the arch was small, LV dysfunction was present, or LV outflow tract was small, particularly when less than 4 mm. Colan and colleagues have also developed and subsequently revalidated a scoring algorithm for decision making in neonates with aortic stenosis and a mitral valve z value of greater than −2.
In those patients in whom aortic valvotomy alone is indicated, the decision to use percutaneous balloon aortic valvotomy (see Special Situations and Controversies later in this section) or surgical valvotomy remains controversial. Surgical valvotomy may be accomplished by closed transventricular valvotomy, open surgical valvotomy with CPB, or using hypothermic or normothermic circulatory arrest. The preference is for surgical valvotomy using CPB, as well as more sophisticated methods of myocardial management than have been generally used. However, a large trial may be the only way to determine comparative outcomes in this complex setting.
Severe congenital valvar aortic stenosis is an indication for operation in older infants and children. Symptoms of angina or syncope always indicate severe stenosis and thus are indications for operation. Conversely, severe stenosis requiring operation frequently occurs without symptoms, but in such circumstances there will usually be physical signs, particularly in the pulse and behavior of the second heart sound. Also, the ECG will usually show an LV hypertrophy pattern; an ECG that shows severe hypertrophy (important ST-T depression) is an indication for operation even if the gradient is less than 50 mmHg.
Mild congenital aortic stenosis is not an indication for operation. Because of the natural history, these patients require long-term periodic noninvasive reevaluation and invasive study and operation if indicated.
Older infants and children with moderate stenosis are a controversial group. Many recommend operation, and others recommend periodic reevaluation of LV-aortic gradient, subendocardial oxygen requirement, or valve area. Those against possibly premature operation argue that (1) sudden death is rare in children whose systolic gradient is 50 to 75 mmHg, (2) operation and probable valve replacement will still be necessary, and (3) valve replacement cannot be delayed by early operation. Therefore, operation usually is not recommended in this group, but is advised if stenosis becomes severe on repeated noninvasive follow-up.
Become a Clinical Tree membership for Full access and enjoy Unlimited articles
If you are a member. Log in here