Abstract
Aortic insufficiency can be a complication of continuous mechanical circulatory support that leads to inefficient circulation and decreased coronary perfusion. In adults, this has been addressed through both surgical and transcatheter approaches. However, this has rarely been reported in children with congenital heart disease. We present the case of a neonate who developed severe neoaortic insufficiency after stage one single ventricle palliation while on central veno-arterial extracorporeal membrane oxygenation (VA-ECMO) that was managed with transcatheter neoaortic occlusion. We used an Amplatzer Vascular Plug II 10 mm device (Abbott Cardiovascular) to successfully improve neoaortic insufficiency and overall hemodynamics. This procedure was technically feasible and may be considered in rare cases to address aortic insufficiency while on mechanical circulatory support.
Highlights
- •
Aortic insufficiency can be a complication of mechanical circulatory support.
- •
We used an AVP II device to occlude the neoaorta and improve AI on VA-ECMO.
- •
This procedure was feasible in a neonate and may be considered in rare cases.
1
Background
Mechanical circulatory support such as ECMO and ventricular assist devices (VADs) are being increasingly utilized in children with congenital heart disease [ ]. Aortic insufficiency (AI) is generally considered a contraindication to mechanical circulatory support as it increases the volume load on the left ventricle and impairs cardiac recovery [ ]. Severe aortic insufficiency can also hinder coronary and systemic perfusion, leading to ischemia. Pre-existing AI has been addressed with surgical valve replacement or closure prior to initiation of support [ ]. However, AI can also develop as a complication of mechanical circulatory support secondary to increased LV afterload and commissural fusion in the setting of inconsistent valve opening or persistent valve closure [ ]. In addition to surgical options, less invasive transcatheter approaches have recently been utilized to replace or occlude the aortic valve. In this case, we discuss the indication for and feasibility of neoaortic valve closure in an infant on mechanical circulatory support after single ventricle stage 1 palliation.
2
Case presentation
Informed patient consent was obtained to publish this report. A 3.5 kg female infant was born at term with a prenatal diagnosis of double inlet left ventricle (DILV), L-malposed great vessels, and hypoplastic aortic arch. There was no valvar stenosis, no restriction at the bulboventricular foramen, mild pulmonary valve regurgitation, mild tricuspid regurgitation, and no mitral or aortic regurgitation. She transitioned well at birth and was started on prostaglandins at 0.01μg/kg/min due to her severely hypoplastic aortic arch. On her third day of life, she went to the operating room for a Norwood procedure with atrial septectomy and a modified Blalock-Thomas-Taussig shunt (mBTTs). After weaning off bypass, there was severe neoaortic insufficiency and a line of demarcation of decreased coronary blood flow to the lateral wall of the ventricle. The DKS had to be taken down and altered multiple times, the aortic root patch augmented, and the neoaortic valve repaired with a commissural pledget stitch. The patient was finally weaned off bypass with a well perfused lateral wall of the ventricle and mild neoaortic insufficiency on echocardiogram. She still had marginal hemodynamics and severely decreased LV posterior wall function, so the decision was made to convert to central VA-ECMO to allow the heart to rest and get volume off. A hemoclip was placed on the 3.5 mm BTTs. In total, she had 4 runs of cardiopulmonary bypass with a total time of 452 min, aortic cross clamp time of 273 min.
The patient was brought back to the cardiac ICU on central VA-ECMO, mechanical ventilation, milrinone 0.5μg/kg/min, epinephrine 0.05μg/kg/min, calcium chloride 10 mg/kg/h, and inhaled nitric oxide 20 ppm. Echocardiogram on postoperative day 2 showed normal LV function on ECMO, mild neoaortic insufficiency, moderate left-sided AV valve regurgitation, and severe right-sided AV valve regurgitation. She was able to wean off ECMO on postoperative day 6, but she had to be urgently placed back on central VA ECMO 3 h later due to persistent hypotension. During the few hours off ECMO, she had an echocardiogram that continued to show normal LV function with mildly reduced free wall motion, mild neoaortic insufficiency, moderate right-sided AV valve regurgitation, and moderate left-sided AVV regurgitation. Blood culture from this time grew Staphylococcus epidermidis , so she remained on cefepime, vancomycin, and fluconazole. Repeat blood cultures were negative.
During this time, she had mediastinal bleeding requiring chest washouts and significant blood product repletion. She failed to improve clinically and required increased flow on ECMO, leading to concern that transthoracic echocardiograms underestimated her degree of AI or missed a coarctation. She underwent diagnostic catheterization on postoperative day 9, which confirmed severe neoaortic valve insufficiency. She had no coarctation, and her coronary anatomy was normal with a patent DKS. It is likely transthoracic echo grossly underestimated her AI in the setting of her open chest and ECMO cannulas. In addition, her need for increased flow on ECMO had been attributed to sepsis but may have correlated with worsening aortic insufficiency over the past few days. She was discussed at a multidisciplinary conference with the consensus to offer single ventricle VAD as a bridge to possible transplant, as all other surgical options were exhausted. In order to optimize her hemodynamics on ECMO and with a VAD, it was necessary to occlude the neoaorta. Given her tenuous clinical status with significant bleeding issues on ECMO and the technical difficulty of closing her neoaortic valve surgically, the decision was made to take her to the catheterization lab for neoaortic occlusion prior to VAD placement. CTA chest was obtained for case planning, which showed the neoaorta measured 10.0 mm.
On postoperative day 12, she was taken to the catheterization lab for neoaortic occlusion. She continued on VA ECMO, general anesthesia, and mechanical ventilation with FiO2 60 %. Access was an 8Fr long sheath in the left femoral vein and a 4Fr sheath in the left femoral artery. We first attempted to place a 12 mm Amplatzer Vascular Plug II via the femoral venous sheath, but injection in the LV through the long sheath demonstrated no antegrade flow through the bulboventricular foramen, so this was removed from the patient. We then placed a 10 mm Amplatzer Vascular Plug II in the neoaorta without change in the native aorta or bulboventricular foramen flow on transesophageal echocardiogram or angiography. This remained stable for 5 min and was released. Injection in the aorta demonstrated the device in a stable position and patent ascending aorta/coronary flow with continued significant neoAI through the device ( Fig. 1 ). Transesophageal echocardiogram showed patent coronary flow, no obstruction to the distal native ascending aorta, an un-restrictive bulboventricular foramen, and mild to moderate insufficiency through the vascular plug. This procedure was technically successful, and she was transferred back to the cardiac intensive care unit for further care. Transthoracic echocardiogram the following day showed resolved insufficiency through the vascular plug. However, she continued to have uncontrolled bleeding on ECMO, requiring further chest washout, and her family chose to withdraw care 2 days later.
