Early history

Balloon aortic valvuloplasty was first described in 1984 in a series of 23 patients who were 2–17 years of age at the time of their procedure.Reference Lababidi, Jiunn-Renn and Walls¹ The catheters were 9-Fr with a balloon length of 40 mm, and therefore not indicated for use in neonates or infants. Over the next several years, as lower profile balloons became available, the procedure was subsequently reported in neonates. Surgical valvotomy had been desribed earlier, but generally involved rigid valve dilation, and transcatheter balloon aortic valvuloplasty was sought as a less invasive alternative. Kasten-Sportes et al,Reference Kasten-Sportes, Piechaud and Sidi² reported their series using a transfemoral approach where they used the antegrade approach to the aortic valve by advancing the valvuloplasty balloon across a patent foramen ovale if present. Otherwise, they would access the aortic valve through a retrograde femoral arterial approach. They reported a 50% incidence of temporary femoral artery occlusion. Subsequent series reported using surgical carotid artery cutdown approachReference Fischer, Ettedgui and Park³ and the umbilical artery.Reference Beekman, Rocchini and Andes⁴ These initial series demonstrated that balloon aortic valvuloplasty was feasible and resulted in acute gradient reduction with improved left ventricular systolic function. There was, however, a high rate of mortality in the early era of this procedure.

Technical aspects

Vascular access is often obtained in the femoral vessels and the valve can be approached retrograde through the femoral artery, or antegrade through the femoral vein. Many centres prefer crossing the valve retrograde through the femoral artery. The majority of balloons used for valvuloplasty go through a 3- or 4-Fr vascular sheath, which is generally well tolerated in term gestation neonates. If there is an atrial septal communication, the valve can also be crossed antegrade through a femoral venous sheath, and other centres prefer this approach. Some operators believe this may result in more stable balloon position during inflation, which may decrease the risk for avulsing an aortic valve leaflet, resulting in significant aortic insufficiency. However, navigating the balloon through a smaller left ventricular cavity can be challenging and may result in compromised cardiac output in a patient with left ventricular systolic dysfunction. It is also imperative to be sure that the valvuloplasty balloon is free from the mitral valve chordal apparatus before inflation. Surgical cutdown can also be performed to access the carotid artery. This gives a fairly straight course to the aortic valve and can potentially minimise the time required to cross the valve with a guidewire that is particularly important in patients with critical aortic stenosis and left ventricular dysfunction. It also minimises the risk of vascular access complications, particularly in smaller or premature neonates. It has been reported that the rate of femoral arterial access complications is much higher in patients <2.5 kg (6%) compared with those who weigh 2.5–3.5 kg (0.9%).Reference Kobayashi, Sallaam and Aggarwal⁵ There is evidence that this does not appear to affect long-term patency of the right carotid artery.Reference Borghi, Agnoletti and Poggiani⁶ Balloon aortic valvulopasty in neonates has also been performed at the bedside with carotid artery access and transoesophageal echocardiographic guidance without fluoroscopy.Reference Weber, Mart and Kupferschmid⁷

When choosing the diameter of the valvuloplasty balloon, we typically start with a balloon to the annulus ration of 0.8–0.9. The balloon diameter can be increased by 1 mm, or a higher pressure balloon can be used, if there is a significant gradient. While it is not clear that larger diameter balloons cause higher grade aortic insufficiency, most operators will keep the final balloon to annulus ratio at 0.9–1.0, particularly in neonates. One study showed that balloon size does not correlate with the degree of aortic insufficiency immediately following the procedure or at 6-week and 1-year follow-up.Reference Hamidi-Manesh, Tibby and Herman⁸ The authors used a balloon to annulus range of 0.7–1.6; however, the average was 0.9 and no long-term follow-up was reported.

Rapid right ventricular pacing may also be used during balloon inflation. This is done to decrease the left ventricular stroke volume and improve balloon stability, which may reduce the risk of trauma to the aortic valve leaflets and decrease the number of inflations to achieve the desired result. We initiate pacing at a rate of 160–180 beats/minute and increase by 10–20 beats/minute until the systolic blood pressure decreases by 50% or the mean arterial pressure decreases by 25%. This has been shown to improve gradient reduction and decrease the incidence of higher grade aortic valve insufficiency.Reference Wizke, Don and Cubeddu⁹

Outcomes

The goal of balloon aortic valvuloplasty is to effectively reduce the degree of stenosis as much as possible without creating significant aortic valve insufficiency. This is particularly challenging in neonates owing to several factors. Other left heart structures including the mitral valve, left ventricle, aortic arch, and the aortic valve annulus itself are often hypoplastic. Many neonates have critical aortic stenosis and are acutely ill with left ventricular dysfunction, and are brought to the catheterisation laboratory on inotropic suport and prostaglandin infusion. The valve morphology in neonates may also be different than in older patients with aortic stenosis. The most common valve morphology in all patients with aortic stenosis is the functional bicuspid valve, which is the least common valve type to require reintervention or surgical valve replacement. Neonates have a higher incidence of a functionally unicuspid valve than older patients, and this valve type may be associated with a higher need for repeat balloon valvuloplasty. Neonates also have a higher incidence of dysplastic valves that do not often respond favourably to balloon aortic valvuloplasty and commonly require surgical intervention. Valve morphology was not been shown to be associated with increase risk of significant aortic valve insufficiency following balloon aortic valvuloplasty.Reference Maskatia, Justino and Ing¹⁰

In-hospital mortality following balloon aortic valvuloplasty has decreased significantly since the procedure was described in neonates. Early series for both surgical and balloon valvuloplasty reported early mortality of up to 59%.Reference McCrindle, Blackstone and Williams¹¹ As the procedure evolved and patient selection criteria improved, the mortality rates decreased to 11–13%, with a 10–20% surgical mortality in this era.Reference Egito, Moore and O’Sullivan¹² More recent series have reported an early mortality rate of 6–8% with acute procedural mortality of 2%.Reference Petit, Ing and Mattamal¹³ The presence of endomyocardial fibroelastosis is an independent predictor of time-related mortality.Reference Han, Gurofsky and Lee¹⁴ Long-term survival following neonatal aortic valvuloplasty is reasonably high with a reported 10-year survival of 87% in patients who survived the early period.Reference McElhinney, Lock and Keane¹⁵ Other adverse events associated with the procedure include stroke, occlusion at the arterial access site, and acute severe aortic insufficieny requiring urgent surgical valve repair or replacement. The mechanism of severe aortic insufficieny may be owing to leaflet avulsion, cusp prolapse, or disruption of the annulus.

The need for reintervention following balloon aortic valvuloplasty is high, with neonates having a higher rate of reintervention than those who undergo the procedure after 1 month of age. Over two-thirds of neonates undergoing balloon aortic valvuloplasty require either repeat balloon valvuloplasty or surgical repair or replacement of the aortic valve at 10 years after the procedure.Reference McElhinney, Lock and Keane^15, Reference Brown, Dipilato and Chong¹⁶ Risk factors that are associated with the need for earlier reintervention include higher residual gradient after balloon aortic valvuloplasty, left ventricular dysfunction before the procedure, and the presence of left ventricular endomyocardial fibroelastosis. Repeat balloon aortic valvulopasty is performed in over half of these patients within 10 years of their initial procedure owing to recurrent stenosis, with many of these occuring within the first year of life. It is likely that rapid somatic growth that occurs in infancy contributes to a high rate of recurrence during the first year. Decreased left ventricular systolic function before valvuloplastyReference Petit, Ing and Mattamal¹³ and an increased residual gradient immediately after the procedure are associated with an increase risk of reintervention. The Boston Children’s Hospital group reported that a residual gradient <35 mm mercury post valvuloplasty was asssociated with a longer freedom from reintervention and aortic valve replacement.Reference Brown, Dipilato and Chong¹⁶ Repeat balloon valvuloplasty can be performed safely, results in significant gradient reduction, and can delay the need for eventual surgical intervention. The degree of aortic insufficiency is often increased following repeat balloon valvuloplasty; however, this has not been shown to increase the risk of surgical subsequent intervention.Reference Petit, Maskatia and Justino¹⁷ Surgical reintervention following neonatal balloon aortic valvuloplasty is also common with predictors of shorter freedom from reintervention, including increased post valvuloplasty gradient, increased aortic regurgitation, and the presence of other left heart obstructive lesions. The aortic valve can be replaced surgically with the Ross operation, aortic homograft, or mechanical valve in patients with a large enough aortic annulus. The reported freedom from aortic valve replacement is 87, 75, and 59% at 1, 5, and 10 years, respectively.Reference Fratz, Gildein and Balling¹⁸

The majority of neonates have either no or mild aortic insufficiency immediately after balloon aortic valvuloplasty, with a 2% incidence of severe insufficiency in more recent series.Reference Petit, Ing and Mattamal^13, Reference Brown, Dipilato and Chong¹⁶ However, aortic insufficiency is generally progressive during long-term follow-up and progression to moderate-to-severe insufficiency occurs in over one-third of the patients by 10 years after balloon valvuloplasty.

Although many of these patients have associated hypoplasia of other left heart structures at the time of diagnosis, successful balloon aortic valvuloplasty has been shown to result in the growth of the aortic valve annulus. The majority of patients with aortic valve annulus z-score <1 have normalisation of their annulus size within 1–2 years. Patients with low left ventricular end-diastolic diameter z-scores also showed normalisation of their left ventricular dimensions during this time. The mitral valve annulus z-scores generally remain low.Reference McElhinney, Lock and Keane¹⁵

Comparison with surgery

There have been no prospective randomised studies comparing balloon aortic valvuloplasty with surgical valvotomy to date. There are numerous retrospective studies comparing outcomes between these two treatment modalities; however, the majority of these are from single centres with a clear bias in which procedure they favour. In 2001, McCrindle et al, Reference McCrindle, Blackstone and Williams¹¹ published the outcomes from 18 centres who were evaluated by the Congenital Heart Surgeons Society. There were 110 neonates in this series, 82 of whom underwent balloon aortic valvuloplasty. The procedural risks, including early mortality, and the rate of reintervention were similar in both the surgical and balloon valvuloplasty groups. They did find that neonates who underwent balloon aortic valvuloplasty had a lower immediate post-procedural gradient, higher degree of aortic valve insufficiency, and a shorter duration of hospitalisation compared with the surgical group. Since this publication, there have been significant improvements in both surgical and transcatheter techniques. Prior surgical series reported patients who had undergone rigid aortic valve dilation in the operating room, including earlier series using a transapical approach. Since then, more refined surgical techniques have been described including resection of nodular dysplasia, thinning of the aortic valve leaflets, and creation of neo-commisures. Catheter technology has also improved during this time as well with the development of lower profile balloons, soft tip guidewires for crossing the aortic valve, and the introduction of rapid right ventricular pacing. Recently, Siddiqui et al,Reference Siddiqui, Brizard and Galata¹⁹ published a comparison of surgical valvotomy with balloon valvuloplasty from three centres in Australia. They reported 129 patients, 67 of which were neonates. Of the total number of patients, 86 in their series underwent surgical valvotomy. They reported that patients who underwent balloon aortic valvuloplasty had a higher rate of reintervention than the surgical valvotomy group. The median time to reintervention for the balloon valvuloplasty group was 11 months compared with 5 years for the surgical group. However, they also reported that the mean residual gradient post valvuloplasty was 40 mmHg, versus 30 for the surgical group, which has been shown to be a risk factor for earlier intervention. They also stated in their manuscript that they predict patients who undergo balloon aortic valvuloplasty will ultimately need surgical valve replacement owing to the “destructive nature” of the procedure. However, they did not show a significant difference in the need for surgical valve replacement between the two groups in their study. Interestingly, the reported incidence of surgical valve replacement 10 years after sugical commisurotomy was 57% in another recent publication, which is very similar to that reported following balloon aortic valvuloplasty.Reference Rehnstrom, Malm and Jogi²⁰