Perspective Chapter: Transcatheter Interventions in the Management of Aortic Valve Stenosis

Abstract

Transcatheter interventions that are useful in the management of valvar aortic stenosis will be reviewed. This chapter focuses on congenital aortic valve stenosis. The procedure of balloon aortic valvuloplasty (BAV) and the results were reviewed; BAV offers good relief of aortic valve obstruction and serves as substitute to surgery and is considered a favored option in the management of aortic stenosis in all age groups. However, BAV in elderly patients with calcific aortic stenosis offers only a temporary relief of aortic valve obstruction and BAV is not recommended for this subgroup of patients. Except for neonates, most patients are discharged home within 24-hours after BAV. While there is conclusive data for provision of pressure gradient relief both acutely and at follow-up as well as deferral of any surgery after BAV, the development of aortic insufficiency (AI) at long-term follow-up is a most important drawback. In neonates, severe AI may develop necessitating surgical intervention. Notwithstanding these drawbacks, BAV is presently believed to be a therapeutic procedure of option in the treatment of valvar aortic stenosis in pediatric and young adult patients. Methodical follow-up to identify reappearance of aortic obstruction and development of substantial AI is suggested.

1. Introduction

The author (PSR) has had interest in the diagnosis and management of aortic stenosis over the years and contributed several original papers, editorials, reviews, letters to the editor, and book chapters [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29] on this subject. The purpose of this chapter is to provide an updated review of transcatheter management of congenital aortic valve stenosis.

Aortic stenosis (AS) is a relatively common congenital heart defect (CHD); mostly seen as an isolated defect though it may be found along with other CHDs such as Shone’s syndrome and aortic coarctation. The incidence of valvar AS is 5–6% of all CHDs. Its occurrence is more frequent in males than in females. The pathology of AS varies from one patient to the next; commissural fusion of bicuspid aortic valve leaflets is the most common pathology. Fusion of tricuspid aortic valve leaflets resulting in AS is seen less frequently. Aortic valve with a single cusp (unicuspid) is observed mostly in the newborn with critical obstruction. Quadricuspid aortic valve is extremely rare. There is concentric hypertrophy of the left ventricle (LV); this is proportional to the degree of obstruction caused by fusion of the aortic valve leaflets. Dilatation of the ascending aorta is also seen; however, the extent of aortic dilatation is independent of the degree of aortic valve obstruction [5, 6, 30, 31]. Clinical features and diagnostic studies used in the assessment of the degree of aortic valve stenosis were previously reviewed elsewhere [5, 14, 16, 19, 22, 27, 30, 31] and will not be discussed in this chapter.

Initially, surgical methodologies were utilized to provide relief of aortic valve obstruction; these include, commissurotomy of the fused aortic valve leaflets via aortotomy [32], plastic repair of aortic valve (neocuspidization) with or without the use of prosthetic material [33, 34, 35, 36], and aortic valve replacement with mechanical [37, 38], bioprosthetic [39, 40] or patient’s own pulmonary valve (Ross procedure) [41, 42], all procedures performed under cardio-pulmonary bypass. Following the use by Kan and her associates of the techniques of Dotter [43] and Gruntzig [44] to open the pulmonary valve [45], Lababidi et al. [46, 47] employed this technique to open the aortic valve. Subsequently, balloon aortic valvuloplasty has become first-line therapy to address AS at most institutions [5, 10, 21, 22]. In this chapter, catheter interventional procedures used in the management of congenital aortic valve stenosis will be reviewed.

2. Indications for balloon aortic valvuloplasty

The indication for transcatheter intervention including AS should be similar to that utilized for surgical therapy. Indications for intervention are largely based on the degree of aortic valve obstruction, as assessed by pressure gradients across the aortic valve immediately preceding balloon aortic valvuloplasty (BAV). A peak-to-peak systolic pressure gradient across the aortic valve greater than 50 mmHg with either symptomatology or ST-T wave changes in the electrocardiogram (ECG) indicative of myocardial ischemia or a peak-to-peak systolic pressure gradient of more than 70 mmHg regardless of symptomatology or ECG abnormalities [2, 5, 7, 10] are indications for BAV. While these criteria are generally agreed upon, it should be noted that catheter interventions in children are commonly performed under general anesthesia at most institutions at the present time and therefore, the trans valvar gradient measurement secured under conscious sedation protocol are not applicable. Consequently, pre-procedure trans valvar gradients measured by Doppler technique are used for determining criteria for BAV. The same 50/70 mmHg gradient criteria alluded to above may be used, but one must ensure that: (1) Doppler recordings from multiple sites in a calm, resting patient should be secured; and (2) correction for pressure recovery phenomenon [27] should be applied.

In neonates with severe/critical AS, the pressure gradients across the aortic valve may not be high if left ventricular function is poor and therefore, the pressure gradient criteria set forth above are not applicable. If clinical signs of congestive heart failure (CHF) are detected; ductal-dependent systemic circulation, requiring administration of prostaglandin E₁ (PGE₁) is present; or poor left ventricular function on echocardiogram is recorded; the described gradient criteria are not necessary for going ahead with BAV [18, 24].

Adolescent and adult AS patients with the above-described trans valvar pressure gradient criteria should also undergo BAV. Because of the enthusiasm expressed by several centers for transcatheter aortic valve replacement (TAVR) [48, 49, 50], it should be pointed out that the TAVR should only be used for calcific AS in the elderly patients and the non-calcific AS in adolescents and adults should be treated with the less aggressive BAV [25].

Patients with recurrent AS following prior BAV or surgical aortic valvotomy are also candidates for BAV subject to meeting pressure gradient criteria listed above. Trivial or mild aortic insufficiency (AI) is not a contraindication, but moderate to severe AI is a contraindication for BAV because of the concern for further increasing AI [3, 5, 10, 22].

3. Technique of balloon aortic valvuloplasty

The most commonly used method of accomplishing BAV is percutaneous femoral arterial route for most children, adolescents and adults and will be described first. Then, other methods used in different age groups will be reviewed.

Cardiac catheterization is performed to confirm clinical and echocardiographic diagnosis of AS after obtaining informed consent as per institutional norms. Most pediatric interventions are performed under general anesthesia at the present time while conscious sedation is used in adult subjects. After securing venous and arterial access, 100 units/kg of heparin (maximum 3000 units) is intravenously administered and activated clotting times checked periodically and kept above 200 s by administering additional doses of heparin as needed.

A #4- to #7-F multipurpose or right coronary artery catheter is positioned in the ascending aorta and advanced into the left ventricle (LV) via the stenotic aortic valve with the assistance of a floppy-tipped coronary guide wire (in infants), a 0.035-inch straight Benston guide wire (Cook) or a similar guide wire. Other types of catheters and guide wires may be utilized, at the discretion of the cardiologist if there is difficulty in crossing the aortic valve. Peak to peak systolic pressure gradient is recorded by pressure pullback across the aortic valve (Figures 1 and 2). If possible, concurrent pressures recording from both the LV and aorta are also documented (Figure 3). But, if there is significant difficulty in crossing the aortic valve, no pressure pullback should be made; in its place, prior recording of the aortic pressure is utilized to calculate the peak-to-peak systolic pressure gradient across the aortic valve (Figure 3).

Figure 1.

Pressure pullback tracing from the left ventricle (LV) to the aorta (Ao) demonstrating a peak-to-peak gradient of 21 mmHg across the aortic valve; this would suggest that the aortic stenosis is mild, provided the cardiac index is within normal range. (Reproduced from reference [22]).

Figure 2.

Simultaneous pressure recordings from the left ventricle (LV) and the aorta (Ao) demonstrating a peak-to-peak gradient of 110 mmHg across the aortic valve suggesting that the aortic stenosis is very severe. (Reproduced from reference [22]).

Figure 3.

Left ventricular (LV) and ascending aortic (AAo) pressures recorded separately showing a peak-to-peak gradient of 55 mmHg across the aortic valve. Pressure pullback was not recorded because of the difficulty in crossing the aortic valve initially. (Reproduced from reference [22]).

Angiograms from the aorta and LV (Figure 4) are secured and the diagnosis is confirmed. Most common cine-angiographic projections used are left anterior oblique and right anterior oblique; these views are likely to highlight the features of AS and associated subvalvar and supravalvar anomalies.

Figure 4.

Selected cine frames from left ventricular (LV) cine angiograms in posterior-anterior view in two neonates with severe aortic stenosis: (A) a pigtail (PG) catheter was introduced into the LV retrogradely; (B) a Berman angiographic (BA) catheter was advanced from the right atrium (RA), across a patent foramen ovale (not marked) into the left atrium (not marked) and from there into the LV. These angiograms demonstrate the aortic valve annulus (arrows in A and B). Note the domed and thickened aortic valve leaflets. (Reproduced from reference [24]).

Once the diagnostic data confirm the indications for BAV, an extra-stiff J-tipped Amplatz guide wire (Cook, Bloomington, IN) or an apex guide wire (Cook) in older children and adults is placed in the LV apex, via the catheter already in place. A balloon valvuloplasty catheter with a balloon diameter that is 80–100% of the annulus of the aortic valve is threaded over the guidewire already in place. The balloon diameter should not go above the annulus of the aortic valve. The aortic valve annulus measurements secured by the echocardiogram prior to the procedure and by the left ventricular angiogram during the procedure are used for the purpose balloon diameter selection. The length of the balloon to be used is largely based on the age and size of the patient. In young babies and neonates, a 2 cm long balloon is used. In older infants and young children, a 3 cm long balloon is preferred. In older children, adolescents and adults, a 4–5.5 cm long balloon is selected. Balloon inflation (Figure 5) with diluted contrast material (1 in 4) to a pressure not to exceed the burst pressure quoted by the manufacturer of the balloon catheter is undertaken. The landmarks of the scout film (Figure 4) at the same camera angulations are used during balloon inflation. I usually perform two to three more balloon inflations, each for a duration of 5 s, 5 minutes in-between.

Figure 5.

Selected cine frames in posterior-anterior projections illustrating a balloon dilatation catheter across the stenosed aortic valve. Waisting of the balloon (arrow) was seen during the early phases of inflation of the balloon (A) which was completely abolished on further inflation of the balloon (B). Ao, aorta; DAo, descending aorta; GW, guide wire; LV, left ventricle; MC, marker catheter. (Reproduced from reference [22]).

If the aortic valve annulus is too big to dilate with a single balloon, a double-balloon method may be used. In this procedure, two balloon catheters are concurrently positioned across the aortic valve (Figures 6 and 7). The effective balloon diameter may be computed by using the formula shown below [51]; this also should not go above the diameter of the aortic valve annulus.

Figure 6.

Selected cineradiographic frames in straight lateral projection demonstrating two balloons placed across the aortic valve; the balloons were positioned retrogradely via both the femoral arteries. Balloon waisting (arrows) during the initial phases of balloon inflation (A) was completely abolished on further inflation of the balloons (B). Ao; aorta; LV, left ventricle. (Reproduced from reference [5]).

Figure 7.

Selected cineradiographic frames in right anterior oblique projection demonstrating two balloons placed across the aortic valve; the balloons were positioned retrogradely via both the femoral arteries. Balloon waisting (arrows) during the initial phases of balloon inflation (A) was completely abolished on further inflation of the balloons (B). Ao; aorta; GWs, guide wires; LV, left ventricle. (Reproduced from reference [22]).

$$\frac{D_1+D_2+\pi \left(\frac{D_1}{2}+\frac{D_2}{2}\right)}{\pi }$$E1

Where D₁ and D₂ indicate diameters of the balloons utilized.

This formulation was made simpler by Narang et al. [52]: Effective balloon diameter = 0.82 ($D_1+D_2$).

A propensity to eject the dilating balloon while inflating the balloon exists. Consequently, we utilize stiff guidewires and long balloons. Other interventionists recommend adenosine-induced transient cardiac standstill [53] or fast right ventricular pacing [54] to attain steady position of the balloon while performing BAV. In the author’s personal experience, use of stiff guide wires and long balloons was found to be satisfactory [1, 3, 5, 7, 10] to successfully accomplish BAV. Nucleus balloons (NuMed) with a “barbell” configuration and hourglass shaped V8 aortic valvuloplasty balloons (Venus Medtech) have a theoretical advantage of keeping the balloon within the aortic valve [29]. While this seems attractive, the bulky nature of these balloon catheters is problematic.

Following the completion of BAV, pressure pullback recording across the aortic valve (Figure 8) is performed and angiograms from the LV and/or aortic root are secured 15 minutes following BAV. The catheters and sheaths are withdrawn, and the procedure concluded. Vascular occlusion devices such as Angio-Seal (St Jude Medical) and others occlusion systems [55, 56] may be used if large balloon catheters are utilized for BAV.

Figure 8.

Pressure pullback tracing across the aortic valve following balloon aortic valvuloplasty, demonstrating a residual peak-to-peak gradient of 18 mmHg, indicating good result of the procedure. Ao, aorta; LV, left ventricle. (Reproduced from reference [22]).

3.1 Balloon aortic valvuloplasty in neonates

Neonatal BAV may be undertaken in a manner similar to that described above [5, 57, 58, 59, 60]. However, injury to the femoral artery [8, 61] is of concern. Consequently, other routes of access, namely, subscapular [62], axillary [63], carotid [64], and umbilical [65] arterial, anterograde femoral venous [66, 67], and umbilical venous [9, 15] routes have been tried. Because of limitations of space, these will not be reviewed in this chapter. The interested reader may find the discussion of these procedures elsewhere [17, 18, 24, 29].

4. Immediate results

Immediate reduction in the peak-to-peak systolic pressure gradients across the aortic valve (Figures 8–11) along with a decrease in the LV peak systolic and end diastolic pressures with no substantial change in cardiac index occurred. There is nearly 60% decline in the systolic gradient when compared to the pre-valvuloplasty values (Figure 11).

Figure 9.

Aortic (Ao) and left ventricular (LV) pressure tracings prior to (A and B) and 15 minutes following (C) balloon aortic valvuloplasty demonstrating almost complete abolition of the peak-to-peak pressure gradient across the aortic valve. (Reproduced from reference [5]).

Figure 10.

Simultaneous pressure recordings from the left ventricle and aorta prior to (PRE—A) and 15 minutes following (POST—B) balloon aortic valvuloplasty demonstrating no residual gradient. There is a slight decrease in aortic diastolic pressure (B) suggesting aortic insufficiency. (Reproduced from reference [22]).

Figure 11.

Bar graph illustrating immediate results of balloon aortic valvuloplasty for aortic valve stenosis. Significant (p = 0.001) decrease in the peak-to-peak systolic pressure gradients (left panel) and percent reduction (right panel) were shown. Mean + standard deviation (SD) is marked. Pre, prior to; post, following balloon aortic valvuloplasty. (Reproduced from reference [22]).

The extent of AI did not worsen (Figure 12; pre vs. post) and there were no patients exhibiting grade 3+ AI. Actually, some improvement of aortic insufficiency was noticed in some patients which suggests restored coaptation of the aortic valve leaflets following BAV. Except for neonates, nearly all patients were sent home within 24 hours of BAV.

Figure 12.

Bar graph demonstrating the prevalence of grade III aortic insufficiency prior to (pre), immediately following (post) balloon aortic valvuloplasty and at late follow-up (FU). No change in aortic insufficiency is seen immediately after balloon valvuloplasty. However, significant increase occurred at late follow-up. (Modified from reference [10]).

Lababidi et al. [47] were the first to document results of BAV in children; they reported the results of 23 consecutive patients with valvar AS. In this series, the peak-to-peak systolic pressure gradient through the aortic valve was reduced from 113 ± 48 to 32 ± 15 mmHg (p < 0.001) following BAV. Mild aortic regurgitation was seen in 10 (43%) patients. Two children needed surgical repair. The author of this chapter presented immediate results of BAV in 16 patients in the late-1980s [1, 3]; the results of larger number of patients (N = 26) became available [5, 7] subsequently. In the initial 16 children, decrease of peak-to-peak systolic pressure gradient through the aortic valve from 72 ± 21 to 28 ± 13 mmHg (p < 0.001) occurred (Figures 8–11). Similarly, LV peak systolic pressures decreased from 162 ± 21 to 124 ± 18 mmHg (p < 0.001) and LV end-diastolic pressures were reduced from 13 ± 5 to 9 ± 6 mmHg (p < 0.01). There was no significant change in cardiac index (3.4 ± 0.5 vs. 3.4 ± 0.4 liters/min/meter²; p > 0.1) [1]. Generally, the gradients decreased by 60% of pre-valvuloplasty values (Figure 11). Similar reduction in peak-to-peak systolic pressure gradients across the aortic were observed in larger cohort (Figure 13) [7].

Figure 13.

Bar graph demonstrating immediate and follow-up results after balloon aortic valvuloplasty. Note significant (p < 0.001) decrease in peak-to-peak systolic pressure gradients across the aortic valve after balloon valvuloplasty (pre, before vs. post, immediately after). Gradient measured during repeat catheterization in 15 patients increased (p < 0.01) at intermediate-term follow-up (ITFU) of mean of 16 months. (Reproduced from reference [7]).

In the second cohort comprising of 26 children [7], the immediate outcome was like that observed by other workers, as tabulated elsewhere (Table I of reference [5]). The occurrence of substantial (3+ or more) AI did not happen for the entire group (Figure 12). Echocardiographic studies revealed no change in the 1. LV end-diastolic dimension (36 ± 9 vs. 35 ± 10 mm; p > 0.1), 2. LV posterior wall thickness in diastole (7.2 ± 2.1 vs. 7.5 ± 1.9 mm; p > 0.1), and 3. LV fractional shortening (50 ± 8 vs. 47 ± 8%; p > 0.1) following BAV (Figure 14). But the Doppler flow velocity magnitudes across the aortic valve (4.0 ± 0.05 vs. 3.0 ± 0.8 m/s; p < 0.001) diminished as were the peak instantaneous Doppler gradients (Figure 15). No child from our study subjects required immediate surgical therapy. Immediate results after BAV documented during the decade of 1983–1992 were tabulated (Table I) in our book [5] for the interested reader. In the ensuing three decades, several interventionalists, too numerous to list, have reported their results of BAV and these are generally similar to those of the first three cohorts described above. However, more recent multi-institutional studies are worthy of review: In the first of these [68], results of BAV in 373 patients from 22 US institutions were examined. Success, defined as residual peak-to peak systolic pressure gradient across the aortic valve ≤35 mmHg and no greater than mild AI, was achieved in 71% patients. Adverse events were seen in 20% of patients. In the second study comprising of 1026 patients from the IMPACT (Improving Pediatric and Adult Congenital Treatments) Registry [69], procedural success was achieved in 71% of non-critical patients and 63% of critical AS patients.

Figure 14.

Bar graph demonstrating left ventricular (LV) end-diastolic dimension (EDD) in mm (left panel), LV posterior wall thickness in diastole (PWTd) in mm (middle panel) and LV shortening fraction (SF) in % (right panel) prior to (pre), on the day after (post) balloon aortic valvuloplasty, and at late follow-up (FU). Mean + standard deviations (SD) are marked. Note that LVEDD, LVPWTd, and LVSF did not change (p > 0.1) immediately after balloon aortic valvuloplasty. At late follow-up the LVEDD increased (p < 0.001) while the LVPWTd and LVSF remain unchanged (p > 0.05). (Reproduced from reference [29]).

Figure 15.

Bar graph showing maximal peak instantaneous Doppler gradients before (pre) and 1 day after (post) balloon aortic valvuloplasty and at intermediate term (ITFU) and late (LTFU) follow-up. There was significant reduction (p < 0.001) in the gradient after balloon aortic valvuloplasty which remained essentially unchanged (p > 0.1) at ITFU (12 ± 5 months) and at LTFU (3–9 years [mean 6 years]). Doppler-derived maximal peak instantaneous gradients at follow-up continued to be lower (p < 0.001) than pre-valvuloplasty gradients. (Reproduced from reference [7]).

5. Short-term follow-up results

At short-term follow up, defined as ≤2 years, peak-to-peak systolic pressure gradients across the aortic valve by cardiac catheterization (Figure 13) as well as by Doppler-calculated peak instantaneous gradients (Figure 15) either were unchanged or slightly increased when compared to acute results. These gradients continue to be pointedly lower than pre-BAV values [7]. Aortic valve gradients measured by Doppler in 26 patients at a follow-up of 16 ± 11 months past BAV were 31 ± 15 mmHg. These data were similar (p > 0.1) to gradients recorded immediately following BAV and remain lower (p < 0.001) than pre-BAV gradients (Figure 15). Nevertheless, when the residual gradient of each patient was assessed, restenosis characterized as a peak gradient of more than 50 mmHg was noticed in 6 (23%) children (Figure 16). Early in our experience, four of these children had aortic valvotomy by surgery and two had repeat BAV at a median period of 9 months after the initial BAV. The extent of AI stayed steady at short-term follow-up [7]. Short-term follow up results described by other researchers were comparable to those of ours as tabulated (Table II) in our book [5].

Figure 16.

Line graph showing aortic valve peak to peak systolic pressure gradients prior to (pre), immediately following (post) and at follow-up (FU) after balloon aortic valvuloplasty. Patients with good results are shown in green while those with poor results are shown in orange. Re-intervention (RI) (balloon valvuloplasty) was performed in some patients and the gradients fell. On further follow-up (2FU), the residual gradients remained low. When severity of the gradients was examined, the severity grade of the stenosis decreased in all patients going from severe to moderate, mild, or trivial and from moderate to mild or trivial. (Reproduced from reference [29]).

5.1 Restenosis and predictors of restenosis

As pointed out in the previous segment, restenosis after BAV seems to occur (Figure 16). The reason why restenosis happens following BAV was examined by analyzing the follow-up outcomes of 16 children [1]. Based on the short-term follow-up results, the 16 patients were split into two groups: Group I who had good results (N = 12) with aortic valve peak gradients less than 50 mmHg at follow-up and Group II who had poor results (N = 4) with peak gradients more than 50 mmHg. In Group I patients, the peak pressure gradient across the aortic valve was reduced from 70 ± 21 to 24 ± 11 mmHg (p < 0.001) at the time of BAV, which stayed unaltered (26 ± 10 mmHg; p > 0.1) at short-term follow-up (Figure 17, left panel). No child in this group needed re-intervention. In Group II patients, the aortic valve peak gradient was reduced (79 ± 20 mmHg vs. 42 ± 13 mmHg; p < 0.001) following BAV. Nevertheless, at short-term re-evaluation, the residual peak gradient increased substantially to 73 ± 5 mmHg (p < 0.001) (Figure 17, right panel). All four children underwent successful re-intervention, two by surgical valvotomy and two by repeat BAV [1].

Figure 17.

Bar graph showing immediate (IMM) and follow-up (FU) results of balloon aortic valvuloplasty in group I with good results (left panel) and in group II with poor results (right panel). In group I with good results, the aortic valve gradient decreased significantly (p < 0.001) immediately after valvuloplasty and remained low (p < 0.001) at follow-up. In group II with poor results, the aortic valve gradient fell (p < 0.01) immediately after valvuloplasty and returned to pre-valvuloplasty values (p > 0.1) at follow-up. Mean + standard error of mean (SEM) is shown. (Reproduced from reference [29]).

Seventeen different variables (Tables I, II, and III of reference [1]) were scrutinized by multivariate stepwise logistic regression testing, as detailed earlier [1, 70, 71] to detect factors that can prognosticate recurrence in Group II subjects. This assessment detected age less than 3 years at the time of BAV and immediate post-BAV peak-to-peak gradient across the aortic valve ≥30 mmHg as prognosticators of recurrent obstruction [1]. In a later study [7, 10], while examining the long-term results of 26 children, the risk factors for recurrence at short-term re-evaluation were precisely identical to those observed in our first report [1]. Furthermore, this analysis [7, 10] indicated that the greater the number of risk factors, the higher the likelihood for restenosis (Figure 18).

Figure 18.

Bar graph demonstrating influence of multiple risk factors on rates of recurrence of aortic stenosis after balloon aortic valvuloplasty. Note that the larger the number of risk factors, the greater is the probability for restenosis. Percentages and actual numbers are shown on the top of each bar. (Reproduced from reference [10]).

Sholler et al. [72] studied the impact of different technological and morphologic issues on the acute outcomes of BAV. However, they were unable to demonstrate any statistically significant role of any factors examined. Other researchers, as reviewed previously [7, 10, 22], explored reasons of reappearance of aortic valve obstruction following BAV; however, they could not discern any factors causing recurrence. A suggestion was made that double balloon BAV may be superior to BAV using one balloon [73]; but thorough assessment of these statistics [74] did not justify such a claim. Balloon/annulus ratios and morphology of the aortic valve may be central elements of restenosis phenomenon; though, the range of variability observed in our study and that of others was not able to establish noteworthy variances; it is likely that investigations involving larger number of patients may unearth other reasons for restenosis [7, 10, 22].

Based on the data presented [1, 5, 7], it was determined that age ≤ 3 years and immediate post-BAV peak-to-peak gradient across the aortic valve ≥30 mmHg may be predictive of aortic valve re-obstruction. It is further surmised that circumventing or reducing risk factors may prevent or lessen the rate of recurrence following BAV. Because the immediate post-BAV gradients across the aortic valve ≥30 mmHg is an alterable risk factor, we advocate use of balloons large enough to decrease the peak-to-peak systolic gradient to <30 mmHg [1, 5, 7].

5.2 Feasibility of repeat BAV for restenosis following prior BAV

As shown in a preceding section, reappearance of aortic obstruction following BAV was detected. We examined the feasibility and effectiveness of repeat BAV in patients who had recurrence after a prior BAV [75]. Twenty-six children with aortic stenosis had BAV between 1983 and 1993; peak gradients across the aortic valve decreased from 71 ± 20 to 26 ± 12 mmHg (p < 0.001). At short-term (10 ± 4 months) evaluation, residual gradients of 34 ± 20 mmHg stayed lesser (p < 0.001) than pre-BAV gradients but have risen (p < 0.01) when compared with acute post-BAV peak gradients. When each patient statistics were examined, six (23%) of the 26 developed re-obstruction, characterized as residual pressure gradients more than 50 mmHg. In our early experience, four patients had successful aortic valvotomy by surgery and two patients underwent a second BAV. Repeat BAV reduced peak gradients from 77 and 66 mmHg to 13 and 6 mmHg, respectively (Figure 19) [75]. Two additional children acquired re-obstruction during long-term follow-up and had repeat BAV successfully at 70 and 107 months after original BAV, respectively. The diameter of the balloons utilized in these 4 patients is a little bigger than that utilized at the time of first BAV.

Figure 19.

Bar graph showing aortic valve peak to peak systolic pressure gradients before (pre), after initial balloon valvuloplasty (1st B), at follow-up (FU), after repeated balloon dilatation (2nd B), and at late follow-up at 6 and 7 years, respectively, in 2 patients with restenosis. Note significant decrease in gradient after each balloon valvuloplasty. Gradients remained low after second balloon valvuloplasty by Doppler (D) and at late follow-up 6 and 7 years later. (Reproduced from reference [75]).

Thus, our experience indicates that repeating BAV is both feasible and effective in managing recurring aortic valve obstruction after prior BAV. Based on these data we recommended that repeat BAV as the therapy of choice for such patients [7, 10, 22, 75]. It should be mentioned that our group of investigators [7, 75] were among the first to demonstrate that repeat BAV is possible and successful in alleviating residual/recurrent aortic stenosis following a previous BAV. In a single institutional study involving 509 patients [76], our findings of feasibility and effectiveness of repeat BAV were validated. These investigators undertook repeat BAV in 115 of 509 patients (23% of initial cohort) who had restenosis following first BAV. A subsequent recurrence occurred in 49 (10% of total). These patients were also effectively managed with a third BAV [76]. In another study of 43 patients [77], the study authors concluded that repeat BAV successfully addresses recurrence of aortic stenosis and delays the need for aortic valve surgery. Accordingly, it is now established that repeat BAV is feasible and effective in alleviating recurrent obstruction following original BAV and in the author’s opinion, repeat BAV is the first choice in the treatment of patients with recurrent AS. The feasibility and effectiveness of repeat balloon dilatation was also demonstrated for other recurrent obstructive lesions such as pulmonary stenosis and coarctation of the aorta [11, 12, 75].

6. Long-term follow-up results

We evaluated long-term, defined as more than 5 years of mean follow-up, results of 26 patients who were restudied 3–10 years (6.7 ± 1.7 years) following BAV. Twenty-two of these patients were reinvestigated longer than 5 years after BAV [7]. In the following paragraphs, these data will be reviewed. Then a review of works of others reporting long-term results will be summarized.

6.1 Residual stenosis

The peak instantaneous Doppler gradients at long-term follow-up were low at 27 ± 17 mmHg (Figure 15). The aortic valve peak gradients were lower than pre-BAV gradients (p < 0.001) and are similar (p > 0.1) to both immediate post-valvuloplasty and short-term follow up values (Figure 15) [7].

6.2 Development of aortic insufficiency

The degree of AI was quantified by the ratio of the jet width on color Doppler of AI to dimension of the LV outflow tract, as described previously [7]. While there was minimal change in the degree of AI both immediately after BAV or at short-term follow-up (Figure 12), the number of patients with 3 + AI increased at long-term follow-up (p < 0.01) (Figures 12 and 20). In these seven (28%) patients with 3 + AI, the left ventricular end-diastolic dimension was at or larger than 90th percentile for the body surface area. Two (8%) of these children had successful Ross operation. The remaining 5 patients were being observed carefully [7]. It was concluded that AI is the most important long-term problem with BAV; this is not too dissimilar to the long-term follow-up outcomes of aortic valve surgery. Additional discussion of AI (probable causes) will be presented in a subsequent part of this chapter.

Figure 20.

Degree of aortic insufficiency by Doppler echocardiography before (pre), the day after (post), and at late follow-up (FU). There is a significant (p = 0.002) increase in aortic insufficiency from pre-valvuloplasty to post-valvuloplasty. Number of patients with grade 3+ aortic insufficiency (0 of 26 vs. 7 of 26) at follow-up (FU) increased (p < 0.02). (Modified from reference [7]).

6.3 Ventricular dimensions and function

At long-term follow-up, the left ventricular end-diastolic diameter (45.4 ± 9.9 mm) was larger (p < 0.01) than both post-BAV (37.2 ± 0.5 mm) and pre-BAV (36.7 ± 8.5 mm) measurements (Figure 14, left panel). To avoid potential impact of growth, standardization of left ventricular measurements to square root of body surface area was made. The resultant values were: 38.5 ± 42 vs. 49.9 ± 5.7 mm/♪m² (p < 0.001); these data continue to show that the LV end-diastolic dimension is larger at late follow-up, presumably related to the adverse effect of AI. Nevertheless, the LV posterior wall thickness in diastole (8.3 ± 1.7 mm) (Figure 14, middle panel) and LV shortening fraction (45 ± 6%) (Figure 14, right panel) at long-term follow-up did not significantly (p > 0.05) change.

6.4 Re-interventions and actuarial event-free rates

Eight (31%) patients, six at the time of short-term follow-up and two during long-term follow-up developed restenosis; these patients were successfully managed with either surgical valvotomy (N = 4) or second BAV (N = 4). One patient had a left ventricular apex-to-descending aortic conduit to circumvent severe left ventricular mid-cavitary obstruction. Seven (27%) patients had severe AI during long-term follow-up (Figures 14 and 20). Two of these children had a successful Ross procedure. Based on these data, event-free rates were calculated (Figure 21). The probability of freedom from re-intervention at 1-, 2-, 5- and 10- year follow-up was 80, 76, 76 and 60%, respectively (Figure 21) [7].

Figure 21.

Actuarial event-free rates after balloon aortic valvuloplasty. Seventy percent confidence limits are marked with dashed lines. Note intervention-free rates at 1, 2, 5, and 9 years are 80, 76, 76, and 76%, respectively. (Modified from reference [7]).

6.5 Long-term results by other investigators

Awasthy et al. [78] compared the results of BAV of adolescents and adults with two other groups, namely, 1. Babies below 1 year of age and 2. Children between 1 and 11 years. The necessity for repeat BAV in 10.3–18.1% patients, occurrence of grade 3 or more of AI in 9–9.6% subjects and need for surgical intervention in 2.4–3.6% at follow-up were similar (p > 0.1) for all three age subsets. In an accompanying editorial published in Indian Heart Journal [25], long-term results of BAV described by several interventional cardiologists were tabulated and this table will not be reproduced here because of limitations of space, but can be found in our editorial [25].

Other studies reporting on long-term outcomes, not included in the table, will now be reviewed. In a single center study involving 78 patients by Sullivan and associates, the estimated freedom from re-intervention was 44% (95% CI: 20–65%) at 15 years and 62% (95% CI: 40–77%) patients remained free of aortic valve replacement. Post-BAV gradients greater than 30 mmHg and acute AI appear to predispose for aortic valve replacement [79]. In another single center study, Pillai et al. [80] followed 92 patients for 5.7 ± 1.3 years following BAV; 85% patients had successful outcome, 10% subjects developed re-stenosis requiring re-intervention, and 2.2% patients developed severe AI. Auld et al. [81] investigated long-term outcomes of sixty patients with a median follow-up of 6.8 years and found freedom from re-intervention in 75% of study subjects. Long-term results after BAV in 57 patients were examined by Godlewski and Werner [82]; they found significant progression of AI and that 90, 77, and 59.5% of patients did not require surgical intervention at 5, 10, and 18 years following BAV. This procedure is also effective in treating rheumatic aortic valve stenosis; Pillai and associates followed 92 patients for a mean of 5.7 years and concluded that BAV is an effective strategy in managing rheumatic aortic valve stenosis [83].

6.6 Summary of long-term results

In brief, the long-term outcome of BAV indicates continued relief of narrowing for the entire group with suggestion for negligible additional re-obstruction, gradual increase of AI, dilatation of the left ventricle and high re-intervention levels [7, 22, 25].

6.7 Causes of aortic insufficiency

As shown in the preceding sections, significant AI was found at long-term evaluation following BAV (Figures 12 and 20). Many other investigations as well as that of ours demonstrate a tendency to increase in intensity of AI as time passes; the lengthier the follow up duration, the greater the degree of AI. Substantial AI was documented in 24–38% subjects with necessity for replacement of the aortic valve in 8–14% patients, as charted elsewhere (Table 4 of reference [10]).

Our study sought to examine the causes of AI [7]. The patients were split into two groups: Group I, 19 children with no significant AI (grade 2+ or less) and Group II, 7 children with 3+ AI. Fifteen biographic, anatomic, physiologic, and technical data (Table II of reference [7]) were assessed by multivariate logistic regression testing to detect factors causing AI [7]. This evaluation detected several factors that were statistically dissimilar between Groups I and II (Table IV of reference [7]). These are Doppler quantified AI both preceding and immediately after BAV and the procedure undertaken during the second half of our experience with BAV. These three items were entered into a multivariate logistic regression model with all feasible groupings. A model that comprises immediate post-BAV Doppler AI fits the information best. Adding pre-BAV Doppler AI and procedural experience to the model that includes post-BAV Doppler AI did not substantially increase its prognostic value [7]. Consequently, it was concluded that the degree of immediate post-BAV grade of AI is prognostic of late onset of substantial AI. The correlation among these two variables is demonstrated in Figure 22. Sullivan [79], Godlewski [82] and their associates also found that the degree of AI at the time of BAV is associated with late AI, confirming our observations.

Figure 22.

Relationship of immediate post-valvuloplasty Doppler-estimated aortic insufficiency (AI) with AI at late follow-up after balloon aortic valvuloplasty (BAV). Note good correlation (R = 0.71) between the two. (Modified from reference [7]).

Intra-operative balloon dilatation with large balloons (1.2–1.5 times the annulus of the aortic valve) both in experimental animal [84] and human [84, 85] models have been shown to result in injury and tears of the aortic valve leaflets producing AI. Hence, we plotted the level of AI at late follow-up along with the balloon/annulus ratio (Figure 23) and noticed no correlation between the size of the balloon and level of AI in our study subjects.

Figure 23.

Relationship of balloon/annulus ratio utilized during balloon aortic valvuloplasty (BAV) with the degree of Doppler-assessed aortic insufficiency (AI) at late follow-up. Note poor correlation (R = 0.36) between these two parameters. Also note grade 3+ AI occurred with wide range of balloon/annulus ratios. (Modified from reference [7]).

The causes for development of AI at late follow-up after BAV are not clearly known. The theories that have previously been advanced are: (1) Doppler-assessed degree of AI both before and immediately after BAV [7, 79, 82], (2) better relief of aortic valve gradient after BAV [84], (3) large balloon/annulus ratio [72, 86, 87], (4) poor morphology of the aortic valve [7] including uni-commissural aortic valve [72], and (5) prolapse of aortic valve leaflets [86]. However, none of these factors appear to have proof in support of their role in producing AI. Our data [7] and that of others [79, 82] suggested that the degree of AI immediately following BAV is prognostic for development of significant AI at long-term follow-up (Figures 12, 20 and 22). We surmised that a mixture of poor aortic valve morphology and large sized balloons [7, 10, 22, 25] are likely to ultimately turn out to be causing AI at long-term follow-up. Further investigations of the above mentioned and additional causes for late AI and formulating techniques to avoid AI were suggested [7, 10, 22, 26].

7. Comparison with surgery

Evaluation of comparative results of BAV vs. surgery is fraught with challenges in that: (1) there are no randomized studies to deal with this issue, (2) difficulties exist for comparison of “older” historical surgical outcomes with “current” BAV results, (3) short duration of follow-up after BAV, and (4) a smaller quantity of BAV subjects accessible for follow up when compared with surgical patients. In the early-1990s, I analyzed the results of surgical therapy reported in 10 papers [5]. The authors of these 10 papers examined the outcomes of 41–179 patients who were followed for 0.3–26 years after surgical intervention. The surgical mortality for children ranged from 0 to 4% while late mortality was 4–22%. In the natural history survey [88], the surgical and late mortality rates were 1.2 and 1.9%, lower than that was reported in above papers. Sixteen to seventy-eight per cent (16–78%) of patients developed restenosis of the aortic valve and 6–65% of patients developed AI. Repeat surgery to alleviate restenosis or to repair/replace regurgitant aortic valve was required in 16–39% patients [5]. Thus, surgical outcomes were not as good as BAV results [5]. Gatzoulis et al. [89] observed no substantial variation in mortality, morbidity, or the need for re-intervention within 12 months of the surgery and BAV. Additional studies, as detailed elsewhere [10, 22] observed no important variation in mortality, morbidity or need for re-intervention among surgical and BAV groups. In addition, the two groups have comparable rates of freedom from re-intervention 5 years after both procedures. A meta-analysis of 2368 patients from 20 studies (1835 in the BAV group and 533 in the surgical group) found no differences in hospital mortality and prevalence of moderate AI between the groups [90]. In addition, they found no differences in long-term survival or freedom from replacement of the aortic valve. However, a greater number of patients required re-intervention in the BAV group. Given the less-invasive BAV, despite higher re-interventions rates, they concluded that a randomized controlled study is necessary. In another meta-analysis of 18 separate investigations consisting of a total of 4078 patients, survival rates, incidence of late AI, and need for aortic valve replacement were similar in both BAV and surgical groups; however, the need for reintervention was higher after BAV than after surgery [91]. These authors suggest comparison of in-hospital days and morbidity associated with both forms of therapy in future studies. However, single institutional studies differ in their conclusions with some suggesting comparable outcomes [92, 93, 94] and others favoring surgery [94, 95, 96]. Therefore, the author favors BAV because of considerable occurrence of mortality, both early and late, universal morbidity and the necessity for re-operation seen with surgical valvotomy. Therefore, BAV is a desirable alternative to surgery [5, 10, 22].

8. Complications seen with BAV

Complications may be observed at the time of BAV or may be detected during follow-up; these will be briefed here and for a more detailed description, the reader is referred to our prior publications [8, 29]. Complications at the time of BAV are transient bradycardia, premature beats, and a drop in arterial pressure during balloon inflation. These abnormalities are restored to normalcy after deflation of the balloon. A short period of balloon inflation (≤ 5 seconds), as suggested previously [5] is likely to lessen such complications. Additional complications are loss of blood necessitating blood transfusion; thrombotic occlusion of the femoral artery needing heparin, streptokinase or thrombectomy [97]; other rhythm abnormalities such as transient left bundle branch block [5], right bundle branch block, transient lengthening of QTc interval [98], short-lived atrioventricular block, supraventricular and ventricular tachycardias [5, 98, 99]; cardiac arrest [100]; perforation of cardiac structures [97, 101]; rupture of the balloon [47, 102]; dislodgement of the balloon [89]; tears of the aortic or mitral valve leaflets [89, 103]; right coronary artery occlusion; transitory ischemia of the myocardium [98]; cerebrovascular accidents [104]; and onset of subvalvar obstruction [105]; however, these complications are infrequent. Tears of the aortic valve were observed in animal models in whom large balloons (1.2–1.5 times the aortic valve annulus) were used [85]. Consequently, large balloons (larger than aortic valve annulus) should not be utilized during BAV. Deaths have been seen in association with BAV [72, 84, 103, 106, 107]; such events are caused by rupture of the aorta, temporary obstruction of severe/critical obstructions, aortic valve cusp perforation or avulsion, exsanguination from iliac/femoral vessel tears, and ventricular fibrillation. Sudden death which is unexplained has also been reported [107] but is very uncommon. Complications seen during follow up were occlusion of the femoral artery [3, 8], development of AI and reappearance of aortic valve obstruction; the latter two were examined in the preceding sections.

9. Miscellaneous issues

Additional issues associated with BAV, such as development of subvalvar obstruction [108, 109], mechanism of valvuloplasty [3, 5, 110, 111, 112], balloon characteristics utilized during BAV [1, 2, 3, 5] will not be reviewed because of limitations of space.

10. Balloon valvuloplasty in specific age groups

In the preceding review, discussion of BAV was primarily centered on infants, children, adolescents, and young adults with congenital AS. The indications, techniques, and outcomes of BAV in the fetuses, neonates, premature infants, and elderly adults with AS are somewhat different; however, will not be examined in this chapter because of limitations of space, but can be found elsewhere [29].

11. Transcatheter aortic valve replacement

Since the description of TAVR [48, 49, 50] in the early 2010s, TAVR has been used extensively to treat elderly patients with calcific AS. Given the enthusiasm with which the TAVR is being used at many institutions, it should be pointed out that the TAVR should only be used for calcific AS of the elderly subjects and the non-calcific AS in adolescents and adults should be addressed with the less invasive BAV [25]. The details of the procedure and results of TAVR are discussed in other chapters in this book and therefore, the discussion of TAVR will not be included in this chapter.

12. Summary and conclusions

After the report by Lababidi et al. of BAV in 1983, this procedure was applied by a number of other cardiologists for alleviation of aortic valve obstruction. This review focuses on congenital aortic valve stenosis. The indications for BAV are peak systolic aortic valve pressure gradients of more than 50 mmHg with symptoms or ECG changes or a peak gradient more than 70 mmHg regardless of the symptoms or ECG abnormalities. One or two balloon valvuloplasty catheters are positioned across the aortic valve, over extra-stiff guide wire(s) and the balloon(s) is/are inflated until the waist of the balloon(s) is eliminated. The recommended balloon/annulus ratio is 0.8–1.0. Femoral arterial access is the most utilized route for BAV; however, other routes of access such as trans-umbilical arterial or venous or trans-venous routes are favored in neonates and young infants to circumvent injury to the femoral artery.

Immediately following BAV, fall in peak systolic pressure gradient across the aortic valve in conjunction with decrease in LV peak systolic and end-diastolic pressures occurs in most patients. Development of AI is rare in children, although it may be seen in the newborn. At short-term follow-up, catheterization-measured and Doppler derived peak aortic valve gradients stay low for the entire cohort. However, when each patient’s data is scrutinized, close to one-fourth of patients developed restenosis, defined as peak-to-peak gradient ≥50 mmHg. When the causes for re-stenosis were investigated, age ≤ 3 years and an immediate post-BAV gradient ≥30 mmHg were found to predict restenosis. Patients with restenosis may be treated with repeat BAV or surgery. Repeating BAV is effective in alleviating restenosis. Long-term follow-up information indicates minimal Doppler gradients, negligible further restenosis beyond what was seen at short-term follow-up and progression of AI in nearly 25% children. Kaplan-Myer event-free rates at 5- and 10-years following BAV are in mid 70s and low 60s respectively. These data indicate good outcomes and avoided or postponed surgical therapy. However, significant AI at long-term follow-up is of concern. Current recommendations favor BAV as first line therapy for alleviation of congenital aortic valve stenosis.

Conflict of interest

The author declares no conflict of interest.