The morphologic and functional spectrum of the left ventricle in severe aortic stenosis ranges from the smaller, hypertrophied “borderline” left ventricle to the dilated and thin-walled left ventricle. In addition, several studies have demonstrated that aortic stenosis with left-ventricular dysfunction presenting in fetal life may lead to hypoplastic left heart syndrome at birth, with varying degrees of chamber hypoplasia and endocardial fibroelastosis (EFE) [6, 8, 10]. Neonates with severe aortic stenosis generally require either surgical valvotomy or transcatheter balloon valvuloplasty [7], whereas those with hypoplastic left heart are generally managed with univentricular palliation (Norwood or hybrid transcatheter-surgical techniques) or cardiac transplantation.

A 3.48-kg male infant was delivered by way of planned repeat Cesarean section at 39 2/7 weeks’ gestation to a 36 year-old gravida 3, para 2 mother. The pregnancy was complicated by a diagnosis of fetal aortic stenosis made at 22 3/7 weeks’ gestation (Fig. 1). At 23 4/7 weeks’ gestation, fetal balloon aortic valvuloplasty was performed. After the procedure, as demonstrated by color Doppler, aortic arch flow had gone from retrograde to antegrade across the aortic valve and in the ascending and transverse aortic arch; there was mild aortic insufficiency. Postnatally the left ventricle and mitral valve were of normal size; the aortic valve was mildly hypoplastic at 5 mm with a peak Doppler-derived gradient of 60 mmHg; and there was significant left-ventricular hypertrophy (diastolic left ventricular posterior wall dimension, LVPWd 6 mm) with poor systolic function (LVEF% < 20%). Doppler flow in the transverse arch and isthmus was antegrade. Cardiac catheterization showed increased left atrial pressure of 25 mm Hg. The aortic valve was dilated by way of a retrograde approach with a 5-mm Slalom balloon catheter (Cordis Corporation, Miami, FL, USA) with a decrease in peak-to-peak gradient from 60 to 45 mm Hg. After valvuloplasty, a 5-mm stent was placed in the interatrial septum. The neonate was maintained on prostaglandin infusion.

On day of life 14, with no change in clinical status or LVEF%, the child underwent stage I hybrid palliation with placement of an 18-mm ductal stent and bilateral pulmonary artery banding. Repeat dilation of the aortic valve was performed at that time, resulting in a decrease in peak-to-peak gradient to 11 mm Hg and mild to moderate aortic insufficiency.

The LVEF% improved gradually to 60% during the subsequent months. Cardiac catheterization at 9 months of age showed a 38 mm Hg peak-to-peak gradient across the aortic valve and mild to moderate aortic insufficiency. Left atrial pressure was 11 mm Hg; left ventricular end diastolic pressure was 14 mm Hg; and estimated pulmonary vascular resistance was 4 Woods units. After recovery from a severe intercurrent pulmonary illness, the child underwent a Ross-Konno procedure, placement of a 16-mm right ventricle-to-pulmonary artery homograft, ductus arteriosus and atrial stent removal, aortic arch reconstruction, pulmonary artery debanding, resection of extensive EFE (Fig. 2), and closure of an atrial septal defect. He was successfully extubated on postoperative day 7 and weaned from inotropic and vasodilatory support. Follow-up echocardiography showed no evidence of mitral inflow obstruction or regurgitation, normal left-ventricular systolic function, and a well-functioning autograft with no aortic arch obstruction. At catheterization 1 month later, the cardiac index was 3.3 ml/min/M2 by thermodilution; wedge pressure mean was 10 mm Hg; and pulmonary vascular resistance was 6 Woods units (decreasing to 2.6 Woods units in response to inhaled nitric oxide).

At 1 year of life, left heart dimensions are normal (mitral valve 13.7 mm [Z score +0.7], LV end diastolic diameter 24.9 mm [Z score +0.3], and LVPWd 6 mm [Z score +2.4]), and his systolic function is normal with an LVEF% of 69% by Simpson’s biplane method (Fig. 3).

Experimental models involving surgically created aortic obstruction do not recapitulate the human phenotype [3]; therefore, study of the pathophysiology of this disease has been difficult. However, natural history studies have shown that in the presence of certain physiologic findings (specifically left-to-right atrial shunting and retrograde aortic arch perfusion in fetal life), progression to hypoplastic left heart syndrome at term can be reliably predicted [6]. The rationale for fetal aortic valvuloplasty is to relieve aortic obstruction and promote antegrade flow through the left heart and aorta. With restoration of antegrade flow, we observed continued growth of the left heart structures through the remainder of gestation; however, ventricular function did not improve appreciably. The reason for this is speculative, but it has been proposed that the fetal left ventricle responds to obstruction with increase in mass without increase in capillary density, which may result in a hypertrophied myocardium with increased susceptibility to ischemia [9, 12]. The extensive EFE in this patient as well as the observed hypertrophy might support this theory. Fetal valvuloplasty partially relieved the obstruction, and after birth we undertook a combined approach that may have further mitigated the effect of long-standing obstruction and ischemic insult through relief of obstruction (by way of additional valvuloplasties) and decrease in preload (left atrial decompression with atrial stenting). Facilitating antegrade perfusion of the aorta and coronary circulation with normally saturated blood, along with improvement in myocardial perfusion, may have contributed to the slow improvement in the patient’s systolic and diastolic function.

Previous reports have also documented the possibility of reversible severe left-ventricular dysfunction and ventricular growth in the neonate [1, 2, 5]. Emani et al. [4] reported a series of patients with borderline left heart disease who underwent “primary left-ventricular rehabilitation” at a median age of 5.6 months, with aortic and/or mitral valve surgery and EFE resection, with good short-term results and no deaths at a median follow-up of 25 months. Our case, however, represents the first report to our knowledge of a patient initially palliated with a hybrid transcatheter-surgical approach after fetal intervention followed by successful surgical takedown of the hybrid and conversion to biventricular circulation by adding Ross-Konno procedure and aortic arch reconstruction to the EFE resection.

Long-term outcome for patients undergoing extensive EFE resection as part of left-ventricular rehabilitation are unknown. Medium-term results from patients undergoing neonatal Ross-Konno procedures have been mixed, and particularly poor results in patients requiring concomitant aortic arch reconstruction have been noted. However, a recent large single-center experience [11] suggested that overall survival and functional status may represent improvement compared with single-ventricle palliation or neonatal transplantation. Certainly long-term follow-up data, including evaluation of functional and neurodevelopmental status and comorbidities (including pulmonary vascular disease) will be needed before there is widespread adoption of this type of approach. However, feasibility of induction of significant left heart structure growth and significant myocardial recovery allowing restoration of biventricular circulation is certainly suggested by our case and deserves further study.