Catheter-based techniques, whether palliative or corrective, are the accepted therapy for many congenital cardiac defects. Interventional, or, better termed, therapeutic catheterizations, were initiated by Dotter and Judkins, who first reported the treatment of peripheral vascular lesions during a catheterization in 1964 (1), when they dilated a stenotic peripheral vessel through a cutdown on the vessel. The next major innovative accomplishment and the first intracardiac therapeutic catheterization procedure for pediatric congenital heart disease were the balloon atrial septostomy done by Rashkind and Miller in 1966 (2). That procedure really “set the stage” for all therapeutic catheterization procedures used today. In 1967, Porstmann and colleagues reported the first nonsurgical corrective procedure in the catheterization laboratory with their description of a technique for closure of a patent ductus (3). Vascular occlusion coils were introduced by Gianturco and colleagues in 1975 (4) and in 1976 King and Mills were the first to describe the closure of atrial septal defects (ASDs) in the catheterization laboratory (5). Even though their device has not found widespread use, it set the stage for future development of transcatheter devices. One of the largest contributions to interventional cardiology has probably been made by Gruentzig, a Swiss-native who in 1976 reported on dilation of peripheral vessels with noncompliant balloons. This initiated a rapid innovative spurt within the congenital cardiac community during which narrowed lesions at various locations were treated with balloon angioplasty, frequently initially in a noncontrolled fashion. In 1982, Dr. Jean Kan reported the first successful transcatheter static balloon pulmonary valvuloplasty (6) and Dr. James Lock in 1987 used the clamshell device to occlude a ventricular septal defect (VSD) using a percutaneous approach (7). Dr. Charles Mullins introduced endovascular stents into the management of patients with congenital cardiac lesions (8), and the long list of innovations reached another milestone when Dr. Phillip Bonhoeffer, a German cardiologist working in France in 2000, performed the first transcatheter pulmonary valve replacement in a human (9). Transcatheter valve therapies and other interventional therapies to treat patients with structural heart disease have rapidly increased over the last few years. Another milestone was reached when the Melody valve (Medtronic, Minneapolis, MN) gained FDA HDE approval in January 2010. These therapies are not limited anymore to patients with congenital heart disease. In fact, transcatheter aortic valve implantations (TAVIs) in adults have overtaken transcatheter valve therapies in the congenital population, with thousands of implantation of the CoreValve (Medtronic, Minneapolis, MN) in adults having been performed thus far.

In this section, the most important therapeutic catheterization procedures performed as of this writing are discussed. This chapter is not intended as a complete and exhaustive textbook of interventional techniques, but instead should give the reader a general overview of therapeutic catheterization. Indications for the individual transcatheter interventions were recently published as a collaborative expert consensus statement from the AHA by Feltes and colleagues, and we have not repeated the entire article in this chapter, but strongly recommend to be aware of these recommendations (10).

It should be emphasized that not every pediatric cardiologist, or, for that matter, every center, should offer every therapeutic catheterization procedure. For any procedures to be performed at any particular institution, minimal specific skills are required, special techniques must be mastered and maintained, and a large inventory of specialized and expensive catheters and devices must be stocked to offer the patient an optimal procedure. Absence of appropriate qualifications and equipment can result in unnecessary risk to the patient without a reasonable chance of the therapeutic catheterization procedure being successfully accomplished. In fact, even if the patient is not acutely harmed by the attempt, it is important to be aware of the fact that the next procedure in a more appropriate setting might be compromised by a previously unsuccessful attempt.

For many years, reporting of procedure-related adverse events was limited mostly to single-center retrospective experiences, often without any clearly and consistently applied criteria of what would be considered an adverse event, and how its severity should be defined (11,12,13). Over the years, several multicenter registries have captured procedural outcomes including efficacy and occurrence of adverse events, which included the VACA registry, MAGIC, as well as CCISC (14,15,16). The data derived from these registries often provided the only prospective multicenter outcome data for many procedure types.

A more systematic capture of procedure-related adverse events was accomplished recently through the C3PO multi-institutional registry (17), using definitions for event severity and preventability as defined in the International Pediatric Congenital Cardiology Code nomenclature (18). This registry documented not insignificant rates of adverse events, 10% for diagnostic cases, and 20% for interventional procedures. Higher severity (level 3 to 5) adverse events occurred in 9% of interventional cases, and 5% of diagnostic cases. The incidence of life-threatening adverse events has been reported to be as high as 2.1% (19).

However, to accurately compare adverse event rates and outcome between institutions and operators, an adjustment for case mix and hemodynamic vulnerability is required. Using consensus-based and empirical methods, and derived from the C3PO dataset, Bergersen and colleagues recently reported on procedure-type risk groups, separating different types of diagnostic and interventional procedures into four different risk groups, ranging from diagnostic procedures performed in patients over 1 year of age (Risk Group 1) to balloon angioplasty of four or more pulmonary artery branches (Risk Group 4) (20).

Following the definition of procedure-type risk groups, Bergersen and colleagues reported on hemodynamic variables associated
with vulnerability for AE, which were determined based on an empirical analysis of the C3PO data set. The combination of variables that were identified to offer the best predictive performance for high severity adverse events included systolic main PA pressure ≥45 mm Hg (two-ventricle) or mean PA pressure ≥17 mm Hg (single ventricle), systemic ventricular end-diastolic pressure ≥18 mm Hg, mixed venous saturations <60% (two-ventricle) or mixed venous saturation <50% (single ventricle), and systemic saturations <95% (two-ventricle) or <78% (single ventricle) (21,22). The CHARM model incorporates the procedure-type risk group, hemodynamic vulnerability, as well as age below 1 year as criteria to calculate a standardized adverse event ratio to facilitate an equitable comparison of adverse events between centers and operators (21). In addition, further work originating from the C3PO registry has found a weight of less than 2 kg as an additional independent risk factor for adverse events (23).

However, incidence of adverse events and efficacy of an interventional procedure don’t exclusively depend on patient-specific characteristics and procedure type. In fact, using data from the C3PO registry, Holzer and colleagues found that operators with fewer years in practice had higher risk-adjusted rates of adverse events (22,24,25).

These various quality improvement efforts have culminated in the development of IMPACT Registry, which is an initiative of the American College of Cardiology with support from The Society for Cardiovascular Angiography and Interventions and American Academy of Pediatrics (26). The registry was designed to capture all catheterization procedures performed nationwide within the United States (US) in pediatric and adult patients with congenital heart disease and has now more than 100 participating centers, resulting in important data on procedure-related adverse events (27,28).

The possibility of creating significant aortic regurgitation has always been the main concern when considering balloon dilation of congenitally stenotic aortic valves, especially in infants and small children. In 1984, Lababidi and colleagues reported for the first time on a series of 23 patients with congenital aortic valve stenosis, in whom the procedure was documented to be safe and effective (31). Despite this report, general acceptance of the technique was relatively slow. One of the fundamental problems of the procedure remains the risk of creating significant aortic insufficiency, which then may accelerate the need for any surgical aortic valve procedure. While this is less of a concern in the adolescent, where all other treatment options are available in such a situation, the problems are more significant in the infant who has a moderate degree of aortic valve stenosis, where severe aortic regurgitation may require a surgical procedure to be performed at an age where one would have otherwise preferably waited a little longer for the patient to grow.

Several other centers have demonstrated that the results of balloon aortic valve dilation approximated the results of surgical valvotomy but with less risk and much less morbidity. However, the disease has very wide morphologic variations, ranging from the critical ill neonate with borderline LV and left ventricular outflow tract (LVOT) size and a very dysplastic aortic valve, to the young adult with isolated valvar stenosis and well-formed aortic valve.

The decision when to take a patient with congenital aortic valve stenosis to the catheterization laboratory is not always straightforward. A variety of factors have to be considered including peak and mean Doppler gradients, age and gender, EKG findings, LV function and degree of LVH, symptoms, exercise tolerance and desire to exercise competitively, valve morphology and pre-existing aortic insufficiency, as well as associated lesions such as coarctation or
mitral valve (MV) disease. Guidelines for the treatment of congenital aortic valve stenosis in children are derived from the adult population (63), where a peak-to-peak gradient in excess of 60 mm Hg in asymptomatic patients is considered an indication for transcatheter intervention. However, in symptomatic patient, or with the presence of ischemic or repolarization changes on EKG, a gradient of 50 mm Hg should be used. However, peak systolic gradients are only meaningful if left ventricular function is normal. Documented aortic valve stenosis in the critically ill neonate with a dilated left ventricle and poor left ventricular function, would be considered a candidate for transcatheter intervention irrespective of any obtained transvalvar gradient, and probably represents one of the few true emergency transcatheter interventions in congenital heart disease.

Balloon aortic valvuloplasty is now considered a standard technique performed in virtually any center that offers interventional treatment for congenial cardiac lesions (Fig. 17.3) and the AHA guidelines list critical aortic valve stenosis (regardless of gradient), isolated aortic valve stenosis with a peak-to-peak systolic transcatheter gradient in excess of 50 mm Hg, as well as a gradient in excess of 40 mm Hg when combined with symptoms or EKG changes as class I indications for transcatheter balloon aortic valvuloplasty (10). However, there are proponents of a surgical approach to congenital aortic valve stenosis, especially with newer surgical techniques, and many articles often report comparable results (64).

In contrast to balloon pulmonary valvuloplasty, where the vast majority of patients can be expected not to require any further transcatheter or surgical intervention, aortic valvuloplasty is usually palliative in nature, and not infrequently aimed at delaying an inevitable surgical procedure, be it valve replacement or Ross procedure, until a time when the child has reached close-to-adult size. A surgical series reporting on the results of surgical aortic valvuloplasty documented a freedom from aortic valve replacement (AVR) of 72% at 10 years, and 60% at 18 years (65). This is very similar to the results of transcatheter aortic valvuloplasty where a freedom from AVR of 79% at 10 years and 53% at 20 years has been reported (66).

In general, aortic valve dilation is performed retrograde with a catheter introduced into the femoral artery. While an antegrade approach with transseptal puncture offers a slight advantage in maintaining a centered balloon position across the aortic valve during balloon inflation, the technique is more cumbersome and is associated with risk of injuring the MV apparatus with resulting mitral insufficiency, and is therefore not routinely employed. An end-hole catheter is passed from the femoral artery across the aortic valve to a stable position in the left ventricle. A double-balloon technique with the introduction procedure repeated from both femoral arteries may offer advantages for the aortic valve dilation in selected older patients, even though today, the spectrum of available balloons usually allows a successful employment of a single-balloon technique.

The catheter/wire passage retrograde across the stenotic aortic orifice is the most difficult maneuver in the entire procedure, and therefore should ideally be performed only once during the procedure. Before crossing the aortic valve, an aortogram with 25-degree LAO/cranial angulation of the AP and straight lateral projection should initially be performed to exactly delineate the size of the aortic valve annulus, while at the same time demonstrating any angiographic evidence of pre-existing aortic valve insufficiency. The exact technique for passing the wire or catheter into the left ventricle varies
from operator to operator. A Judkins right coronary catheter curve or multipurpose catheter is used by some operators with success in crossing the aortic valve from this approach. However, the Judkins left coronary catheter may offer advantages in many patients, as the curvature is automatically directed to the leftward and posterior opening of the congenitally stenotic aortic valve. Once the valve is crossed, an end-hole catheter (not Judkins left) is advanced over the wire into the left ventricle, and the wire replaced with an extra stiff exchange-length wire with a long floppy tip, which is looped within the ventricle to protect the ventricular apex from perforation by the catheter tip and to minimize ventricular ectopy. A left ventricular angiogram is optional. If an atrial communication is present, hemodynamic evaluation can be performed by advancing a catheter antegrade into the left ventricle with simultaneous pressure recording in the ascending aorta. The same approach is then also used for left ventricular angiography. In neonates and infants, a floppy-tipped coronary wire with a relatively stiff body may be advanced across the valve and allowed to loop in the left ventricle. Care has to be taken to prevent the wire from being ejected from the left ventricle and therefore, once positioned, wire control should be maintained throughout the procedure. By using a floppy-tipped, high torque guidewire, the wire does not need to be changed, and the first catheter to cross the valve can be the dilation balloon (thus minimizing the period of potential low output). The use of stiffer exchange wires and longer dilation balloons may aid in maintaining an exact position of the balloons across the valve during inflation and, in turn, eliminate the “shear” trauma to the valve from balloon movement during inflation. With the wire secured within the left ventricle, the deflated balloon is manipulated through arterial sheaths and passed retrograde over the wire. We do not believe that direct introduction of the balloon through the skin should have any role to play, as the balloon profile of balloons has considerably decreased, thereby allowing the appropriately sized balloon to be introduced through fairly small sheaths. In addition, it is quite conceivable that the pulling of a deflated balloon directly through the femoral and iliac arteries may cause more harm to the vessel than using the appropriately sized sheath. Once the balloon is positioned across the stenotic valve, the balloon is rapidly inflated to the recommended maximal pressure and then rapidly deflated.

One difficulty of the procedure is to keep the balloon positioned across the aortic valve during inflation. Once it is inflated, a balloon tends naturally to move toward the ascending aorta because of the ejecting forces created by the left ventricle. It is generally difficult to push against these forces (unless using an antegrade approach), but a longer balloon length aids this process. Adenosine has been used to achieve a temporary cardiac standstill, but its timing in relation to balloon inflation is often difficult to predict. A more controllable method of reducing the cardiac output is through rapid right ventricular pacing (67). The rate of pacing can be adjusted prior to balloon inflation to achieve a drop in blood pressure by at least 50% and these settings are then available to be used during the inflation process. An inflation device that can be operated using a single hand is preferential, as this allows the operator to use the other hand to maintain control of the balloon catheter, making very fine adjustments as the balloon is inflated. The balloon is then immediately and rapidly deflated, with the entire process taking no more than 5 to 10 seconds. Arterial pressures should be monitored throughout the procedure. To limit the potential damage to the aortic valve, only one inflation should be performed provided that the operator is assured that (a) the balloon remained properly positioned in the valve; (b) the balloon was of adequate size; and (c) the waist disappeared. Regardless of the technique, a marked drop in systemic pressure, a rise in left ventricular pressure, and resultant bradycardia may transiently result. The double-balloon technique using two balloons placed side by side across the valve may minimize this problem, but more importantly one has to avoid prolonged inflations with any technique when performing aortic balloon dilations. With successful valve dilation, after the balloon is deflated, both the blood pressure and heart rate should return spontaneously to normal.

For a single-balloon technique, the initial balloon is chosen with a diameter of about 80% to 90% of the measured aortic annulus diameter. After each set of inflations, the hemodynamic result and the degree of aortic insufficiency are evaluated. If no or only a mild change in the degree of aortic insufficiency has been observed with still a significant residual gradient (>35 mm Hg), repeat dilation valvuloplasty is done with a balloon sized just 1 to 2 mm above the one previously used. Brown and colleagues recently demonstrated that freedom from AVR was improved when comparing patients with a residual gradient of less than 30 mm Hg to those with a gradient in between 30 and 39 mm Hg (66). Even more importantly, freedom from AVR was better in patients with a gradient of less than 35 mm Hg but moderate to severe aortic regurgitation, when compared to those with more than 35 mm Hg but only mild aortic regurgitation, which suggests that the residual gradient may be more important when compared to aortic insufficiency than previously thought.

When using the double-balloon technique, the combined diameters of the two balloons should approximate 1.2 times the measured diameter of the aortic annulus. Because of the extensive manipulation in the left side of the heart and arteries, all these patients are systemically anticoagulated with heparin at the beginning of the procedure.

In the past, the most common complication of aortic balloon dilation was damage to the femoral arteries by the large balloon dilation catheters. This problem has been minimized by newer lower-profile balloon designs, use of the double-balloon technique where required, and diligent monitoring of ACT levels throughout the procedure. When arterial damage does occur, it usually can be managed medically or, rarely, surgically. In small infants, because of the increased risk of femoral artery injury from the introduction of the dilating balloon catheters into the vessels, several other approaches to aortic valve dilation have been described. The prograde approach, first passing a catheter, then a wire, and finally the balloon from the femoral vein to the right atrium, foramen ovale, left atrium, left ventricle, and prograde across the aortic valve is chosen by some. Although frequently successful, this approach has a high incidence of failure in delivering the balloons and, even more disturbing, a significant incidence of damage to the MV apparatus. Another approach is through a controlled cutdown on the carotid artery. As a result of extensive experience with extracorporeal membrane oxygenation (ECMO) and the safe introduction of cannulae into the carotid arteries, several centers, with the help of pediatric or vascular surgeons, dilate aortic valves in infants from this approach. The approach is direct to the aortic valve and requires less catheter manipulation and less overall time, and has resulted in no reported complications related to the technique. The ideal procedure for the small infant is still to be determined.

With a conservative dilation of the aortic valve, the gradient should be reduced to a gradient equal to or less than 35 mm Hg. This usually can be accomplished without inducing significant aortic insufficiency, no more than that seen after surgical valvotomy. Furthermore, as highlighted above, recent data by Brown and colleagues suggests that the residual gradient may be more important when compared to aortic insufficiency than previously thought, and reducing the gradient to less than 35 mm Hg may be more important, even if it were to come on the expense of moderate aortic insufficiency (66). However, the treatment approach has to be tailored to the individual patient and specifically in infants, gradient reductions to less than 40 mm Hg may be sufficient to delay the early need for aortic valve surgery. The long-term results, like surgical valvotomy, will be palliative; however, the catheter balloon dilation procedure is accomplished without a sternotomy or cardiopulmonary bypass with their inherent risks and morbidity. Balloon dilation of congenital aortic valve stenosis in pediatric patients and young adults is now the standard initial procedure for this lesion in most centers.

With the development of the special, larger dilation balloons, a transcatheter technique for balloon pulmonary valve dilation was first introduced by Kan et al. (6) in 1982. The technique performed acutely was successful and, at the same time, carried little risk over and above the basic risk of a catheterization. By December 1986, 28 centers, voluntarily reporting to a collaborative registry (VACA), demonstrated the successful and safe application of the technique in more than 680 cases of pulmonary valve stenosis (68). With these data and many subsequent reports of successful use (69,70), balloon dilation has been accepted as the standard therapeutic procedure for pulmonary valvar stenosis. It is applicable to patients of all ages from the newborn period throughout adult life. With its excellent results and low rate of procedure-related complications, maximum instantaneous systolic Doppler gradients of as little as 35 mm Hg, when combined with evidence of right ventricular hypertrophy, should be considered an indication for balloon pulmonary valvuloplasty (71).

The degree of pulmonary valve stenosis is documented by accurate hemodynamic measurements in the catheterization laboratory. However, if the pulmonary valve is not easily crossed, then right ventricular angiography should be obtained prior to further attempts at crossing the valve.

The valve anatomy, size, and exact location are visualized angiographically, with standard AP (with some cranial angulation) and lateral being the most appropriate projections. Accurate determination of the valve annulus diameter is obtained using appropriate calibration techniques. With this information available, a long exchange guidewire is passed through an end-hole catheter into a distal pulmonary artery. The left pulmonary artery is preferable for this position because of the straighter course from the valve and main pulmonary artery to the left. However, in neonates with a patent arterial duct the wire may be passed through the duct into the descending aorta. The chosen wire should be fairly stiff to allow the balloon to track over the wire and across the stenosed pulmonary valve. In infants and smaller children, a stiff 0.018 wire with a floppy tip is frequently appropriate for this purpose. However, infants with critical pulmonary stenosis and a closed arterial duct may poorly tolerate placement of a wire or catheter across the valve; therefore the valve should only be crossed when all equipment has been prepared to immediately proceed with balloon pulmonary valvuloplasty.

McCrindle and colleagues documented that the optimum balloon diameter should be between 1.2 and 1.3 times the size of the pulmonary valve annulus for a “single-balloon” dilation (72). Lower balloon to valve annulus ratios are associated with an increased risk of recurrent or residual pulmonary valve stenosis, while ratios in excess of 1.4 are associated with an increased risk of clinically significant pulmonary insufficiency (72).

The choice of balloon catheters that can be used for this procedure is wide and depends to a degree on the individual valve morphology. In general, inflation pressures of more than six atmospheres are rarely necessary in patients with typical valve morphology, and therefore low-pressure balloons, such as Tyshak II (NuMED, Hopkinton, NY) with a lower profile are the preferred initial balloon types to be utilized. However, the maximum rated inflation pressures decline sharply when using the larger varieties of these balloons. Therefore high-pressure balloons, such as ZMed II (NuMED, Hopkinton, NY), or the double-balloon technique should be considered in these situations as the primary approach to balloon valvuloplasty (Fig. 17.4). High-pressure balloons may also be more beneficial when dealing with very dysplastic, thickened pulmonary valves in the older patient, or if there is associated supravalvar narrowing. In neonates with critical pulmonary valve stenosis and a closed arterial duct, a low-profile balloon with fairly rapid deflation characteristics such as the Tyshak II balloon should be used. Very low-profile balloons, such as the mini Tyshak balloons (NuMED, Hopkinton, NY), may cross the valve more readily, but their very slow deflation characteristics make these balloons an inappropriate choice in patients who have critical pulmonary stenosis without a patent arterial duct. If the valve cannot be crossed with the appropriate-sized balloon, smaller coronary balloons can facilitate predilating the valve to allow the larger balloon to be subsequently passed.

With the wire fixed in place in the distal pulmonary artery, the end-hole catheter is removed and the catheter with its deflated balloon is passed over this wire until the center of the balloon length is positioned exactly at the area of the stenotic valve. The balloon is then rapidly inflated to the pressure recommended by the manufacturer and is observed for the appearance of a circumferential indentation or “waist” in the balloon. Full inflation
results in disappearance of the waist. An inflation device that can be operated using a single hand is preferential, as this allows the operator to use the other hand to maintain control of the balloon catheter, making very fine adjustments as the balloon is inflated. The balloon is then immediately and rapidly deflated, with the entire process taking no more than 5 to 10 seconds. Arterial pressures should be monitored throughout the procedure. In contrast to balloon aortic valvuloplasty, more than one inflation is usually performed to assure the operator that (a) the balloon remained properly positioned in the valve; (b) the balloon was of adequate size; and (c) the waist disappeared early and at low pressures during subsequent inflations. When a single balloon is used, there is a significant drop in both systemic blood pressure and heart rate during inflation. With successful valve dilation, after the balloon is deflated, both the blood pressure and heart rate should return spontaneously to normal.

To avoid the marked drop in systemic blood pressure and to reduce the trauma to the peripheral introductory veins, a double-balloon technique was introduced (73). This technique also allows the use of higher inflation pressures in patients with a large pulmonary valve annulus, where a single balloon would provide an inadequate rated burst pressure. The double-balloon technique uses two separate balloon catheters, each on a smaller shaft and with a smaller balloon “profile.” Each is introduced into a separate vein. With this technique, a second exchange wire is introduced from the opposite femoral vein and positioned across the pulmonary valve into a distal pulmonary artery, possibly next to the first wire. Two smaller-diameter balloon dilation catheters are advanced over the separate wires and centered in the valve orifice, and the two balloons are simultaneously inflated. Various formulae have been used to estimate the equivalence of double-balloon to single-balloon technique (71), ultimately resulting in comparison charts that allow choice of sizes of the two smaller balloons, depending on the size that would have been chosen if a single-balloon technique would have been used. However, a combined diameter of 150% to 160% of the pulmonary valve annulus can be used as a guide to choose the appropriate balloon sizes.

Reported success criteria for balloon pulmonary valvuloplasty have been variable. The VACA registry, dating back as much as 15 years, defined a gradient on follow-up of equal to or above 35 mm Hg as procedural failure (72). Holzer and colleagues defined procedural success as either a reduction of the peak systolic RV/MPA gradient to less than 25 mm Hg, or a reduction of the valvar gradient by at least 50%, or a reduction of the RV/systemic pressure ratio by at least 50% (74). However, with pure valve stenosis and a closed PDA, regardless of the initial gradient, one should expect to reduce the pressure gradient across the nondysplastic pulmonary valve to less than 10 mm Hg by balloon valvuloplasty in most patients, with an equivalent reduction in the RV to systemic pressure ratio.

If the initial results are suboptimal, the balloon size should be increased up to 130% to 140% of the pulmonary valve annulus, and high-pressure balloons should be used, after obtaining an intermittent angiographic evaluation, using, for example, a Multi-track catheter that is positioned over the in situ guidewire in the RVOT. However, it is important to recognize that relief of the valvar stenosis may unmask a dynamic infundibular obstruction in some cases resulting in a persistent residual right ventricular outflow gradient (Fig. 17.5). This secondary area of obstruction can be documented by pressure recording during careful catheter withdrawal from the pulmonary artery to the RVOT or with simultaneous pressure recordings from a double-lumen catheter or from separate catheters in each of the two areas. Experience has shown that the infundibular obstruction is dynamic and that it will regress with time. This is particularly notable in adult patients undergoing balloon pulmonary valvuloplasty. Fawzy and colleagues reported an incidence of infundibular gradients in excess of 30 mm Hg in 46% of 93 adult patients undergoing balloon pulmonary valvuloplasty (75,76). All these patients were re-catheterized within 6 to 24 months, documenting a decrease in the mean infundibular gradient from 43 to 25 mm Hg. In less than 10% of patients, a dysplastic pulmonary valve is encountered, with thickened, redundant leaflets resembling a “cauliflower”; frequently supravalve stenosis may coexist. This condition is more common in some genetic conditions, such as Noonan’s syndrome. A higher-pressure balloon is usually required with a gradient reduction frequently less than what would be expected with nondysplastic valves. Within the recent multicenter experience from the C3PO registry reported by Holzer and colleagues, acute procedural success was 77%, with independent risk factors for procedural failure being the presence of a genetic syndrome, complex two-ventricle anatomy, presence
of supravalve pulmonary stenosis, a hemodynamic vulnerability score of 2 or more, as well as the need for balloon inflation above 8 atmospheres (24).

Procedure-related serious adverse events are rare. In a multicenter registry (C3PO) including 290 patients, the incidence of any adverse events was 9%, and high severity adverse events occurred in less than 3% (74). There was no incidence of procedure-related death, and independent predictors of higher level adverse events included among others age below 1 month, complex two-ventricle anatomy, two or more parameters of hemodynamic vulnerability, as well as operator experience of less than 10 years.

The long-term effects of the procedure have not yet been determined. However, studies so far have documented rates of restenosis between 5% and 11% within 10 years after the procedure (69,70). The risk of recurrence or restenosis may be greater in patients who present in infancy or with very dysplastic pulmonary valves, as well as those in whom an undersized balloon was used in the initial procedure. So far, reports have not provided evidence to suggest an increased risk of patients requiring pulmonary valve replacement because of pulmonary insufficiency, secondary to balloon pulmonary valvuloplasty. Dr. Charles Mullins pointed out that in the presence of otherwise normal heart and lungs, the regurgitant fraction is usually small and at low diastolic pressure, due to 80% to 85% of the ejection fraction having “diffused completely into the distal pulmonary capillary bed by the end of systole.” However, we have also learned that progressive right ventricular dilation secondary to pulmonary insufficiency should be considered a potential indication for pulmonary valve replacement.

The diagnosis of pulmonary atresia with intact ventricular septum (PA/IVS) is usually made within the neonatal period. Pulmonary blood flow after birth is maintained through a patent arterial duct until a more definitive source of pulmonary blood supply can be established. An intra-atrial communication allows the systemic venous return to pass into the systemic circulation. The long-term prognosis of patients with PA/IVS can be extremely poor, especially when a diminutive right ventricle is combined with an RV-dependent coronary circulation with multiple coronary abnormalities. In this situation patients may require cardiac transplantation early in life. However, while patients with the diagnosis of PA/IVS and a single-ventricle pathway usually have a very poor long-term outcome, the outlook for those patients with a biventricular or a “one-and-a-half ventricle” circulation is much better.

Perforation of the atretic pulmonary valve plate has an important role to play within the available treatment modalities for these patients (61). Achieving antegrade pulmonary flow not only acutely decompresses the right ventricle, but more importantly serves as an incentive to facilitate further growth of an initially hypoplastic right ventricle. While a variety of sharp instruments as well as laser-guided techniques have been used to perforate the atretic pulmonary valve, these techniques are often poorly controlled and associated with a variably high risk of creating inadvertent injury to surrounding structures, often with disastrous results. Consequently, the use of RF energy was introduced into therapeutic cardiac catheterization in the early 1990s as an alternative to laser-guided perforation of the pulmonary valve plate (77).

The equipment presently most frequently used to achieve perforation of the atretic pulmonary valve with RF energy is the Nykanen RF perforation wire and the Baylis RF puncture generator (Both: Baylis Medical Corporation, Montreal, Quebec, Canada). Suitability for transcatheter RF perforation of the pulmonary valve plate is ascertained by 2-D echocardiography with minimal criteria in most cases being the presence of a tripartite right ventricle as well as a membranous atretic pulmonary valve with a well-formed infundibulum (78). Vascular access is routinely obtained via right femoral venous cannulation. A femoral arterial pressure monitoring line is placed. Initial hemodynamic evaluation includes measurement of right ventricular and systemic arterial pressures, followed by right ventricular angiography with 20-degree cranial angulation of the frontal tubes and standard lateral projection. This allows measurement of the pulmonary valve plate diameter and exclusion of RV-dependent coronary circulation. Further angiography is obtained in the left ventricle in the same projection. The combination of these two ventriculograms allows documentation of the relationship between the blind ending right ventricular infundibulum and main pulmonary artery. A 5-Fr Judkins 2.5 right coronary artery catheter is placed below the pulmonary valve plate within the RVOT using a Touhy Borst adapter to allow passage of the RF wire and simultaneous contrast injections. Once the RF wire and coaxial catheter are loaded, and accurate positioning is confirmed, RF energy is applied while maintaining a gentle push on the RF wire toward the valve membrane. In most patients a power setting of 5 W/sec for 2 seconds should be sufficient to perforate the pulmonary valve plate and it is important for the operator to be particularly suspicious of creating a false track if high-power settings are required. RF energy is discontinued once the wire has advanced through the valve plate. Appropriate position is confirmed using contrast injection through the Touhy Borst adapter. The coaxial catheter is then advanced over the RF wire into the main pulmonary artery and the RF wire is exchanged to a 0.014- or 0.018-inch coronary wire, which can be directed either to a position in a distal branch pulmonary artery or preferably through the PDA into descending aorta. Even though a wire position in the distal pulmonary arteries or preferably across the PDA needs to be established, this should not be attempted with the Nykanen RF wire, as it can easily cause perforation of the main pulmonary artery (61). The use of a coaxial catheter is a significant advantage when comparing the Baylis to the Osypka RF system that is still in widespread use elsewhere. A low-profile balloon valvuloplasty catheter, such as the Mini-Tyshak (NuMED, Hopkinton, NY), with a diameter of about 130% of the valve plate annulus is then advanced over the coronary wire and balloon dilation performed (Fig. 17.6). In cases where the balloon catheter cannot be advanced across the valve, sequential dilatation can be performed starting with a lower-profile 2.5-mm coronary balloon. Trackability can also be improved through arterial fixation or snaring of the coronary wire (79). Balloon valvuloplasty is followed by assessment of the pressure gradient across the pulmonary valve as well as the RV-to-systemic pressure ratio, and a final RV angiography is performed to document the result of the procedure.

A number of studies have reported on the outcome of RF perforation of the pulmonary valve and results have been summarized by Benson and colleagues (80,81,82,83,84,85). Most series are very small and include less than five patients. The overall procedural mortality is about 8% with incidence of procedural complications being about 15%.

It has been advocated that the combination of ductal stenting with RF perforation in one single procedure may reduce the number of transcatheter interventions and prevent prolonged postprocedural care. Approximately 50% of all neonates undergoing successful perforation and balloon valvuloplasty require additional pulmonary blood flow (86), even though it remains difficult to predict this for each individual patient. Some authors have suggested to use the tricuspid valve z-score (83), but this has not consistently been identified of being predictive of the need for pulmonary blood flow augmentation. Regardless, after failing to wean from PGE infusion after 5 to 7 days, one should proceed with either PDA stenting or surgical aortopulmonary shunt placement.

If performed appropriately, this technique allows up to 75% of suitable patients to ultimately sustain a biventricular or one-and-a-half ventricle circulation (61).

Dilation of peripheral arteries was the first therapeutic catheterization procedure and represents yet another area in which the vascular radiologist introduced the techniques. Dotter and Judkins (1) reported on the dilation of atherosclerotic peripheral arteries at the same time that Rashkind and colleagues were working on the atrial septostomy catheter.

The technique for dilation of vessel stenosis uses small, cylindrical, fixed maximal diameter dilating balloons passed over a spring guidewire, positioned across the area of stenosis, and inflated with relatively high pressures. This stretches or tears the area of stenosis up to the predetermined diameter of the balloon. This balloon technique is not only used to treat recurrent or native coarctation of the aorta, but similarly is used for branch pulmonary artery stenoses as well as central and peripheral venous stenoses. As many vessel dilation procedures are associated with immediate recoil and subsequent restenosis, the addition of balloon-expandable stents has further improved upon the available therapeutic interventional armamentarium and is frequently the treatment of choice in adult patients.

In 1979, Sos first experimented with balloon angioplasty of native coarctation in postmortem specimens (90) and this was followed by some fundamental work by Dr. James Lock in the early 80s. He performed balloon angioplasty on excised human coarctation segments as well as experimentally induced coarctation in lambs (91,92). He showed that balloon angioplasty achieved its therapeutic result by creating micro- or macroscopic intimal and medial tears over a variable distance in the vessel. These intimal injuries appeared to have healed completely on late pathologic examination, leaving the ballooned segment with a normal looking intima. The studies also documented that a balloon diameter of less than twice the size of the coarctation segment was unlikely to achieve a successful dilation, while diameters greater than three times appeared to carry a higher risk of deep and extensive tears.

Dr. Ronald Grifka’s and Dr. Charles E. Mullins’ group from Texas Children’s Hospital set the stage for early evaluation of endovascular stent therapy to treat coarctation (Fig. 17.8) (93). In animal experiments, they implanted Palmaz P308, 188 and 108 stents in the abdominal aorta and documented that stents can be safely redilated
after initial stent implantation. Dr. Andrew Redington added to this in 1993 by implanting a self-expandable stent into a 10-week-old girl who had previously had palliative surgery for HLHS (94). While one should generally avoid endovascular stent therapy in small infants, in some very rare circumstances the implantation of a low-profile stent in very sick infants may be justifiable, even though one has to accept that these low-profile stents are not expandable to adult size and therefore will eventually require surgical removal. In 1995, Suarez de Lezo reported the first large series of stent implantation to treat native and recurrent coarctation in humans using the Palmaz stent (95) and in 1999 Cheatham first reported on a new stent design, the CP stent, which is also available in a PTFE-covered variety (96). Holzer and colleagues at Nationwide Children’s Hospital reported a larger experience of stent therapy for complex aortic arch lesions, which further expanded the indications for transcatheter stent therapy with excellent results (Fig. 17.9) (97).