Coarctation can present at any age. Neonates with ductal dependent or “critical coarctation” often present with heart failure, acidosis, and shock following closure of the ductus arteriosus. Without prompt medical resuscitation and surgical intervention, death may occur rapidly[13,14]. Prenatal diagnosis can prevent these sequelae by allowing for intervention before ductal closure. However, prenatal diagnosis of coarctation is challenging due to the presence of the ductus arteriosus and limited blood flow across the aortic isthmus in utero[13]. In the United States, fewer than 1 in 4 patients with isolated coarctation requiring neonatal intervention are diagnosed prenatally[15,16]. Moreover, approximately 30% of neonates with coarctation remain undiagnosed upon discharge after delivery[17]. For these reasons, many physicians advocate for newborn pulse oximetry screening programs, which increase the likelihood of detecting lesions like coarctation before ductal closure[18,19]. Additionally, coarctation must be suspected in infants with other left-sided obstructive heart lesions and may be diagnosed in infants with chromosomal defects, especially those with Turner syndrome and Jacobsen syndrome[20].

Patients with less severe coarctation may not be diagnosed until later in childhood when a murmur is heard or hypertension noted. In these patients, collateral vessels develop from the internal thoracic and subclavian arteries, thyrocervical trunks, and vertebral and anterior spinal arteries to supply blood to the descending aorta[21,22]. For those who enter adulthood undiagnosed, hypertension is the most common presenting symptom[23]. Others may complain of frequent headaches or claudication of the lower extremities with exertion. In these patients, the most telling exam findings suggestive of coarctation are diminished and/or delayed lower extremity pulses and a systolic pressure gradient between the upper and lower extremities[13,23]. However, for patients with extensive collateral blood flow, femoral pulses and lower extremity blood pressures may only be minimally diminished[24].

Chest X-ray is often nonspecific in young patients. In older patients, an anterior-posterior film may show indentation of the aorta at the site of coarctation with pre- and post-stenotic dilation of the aorta, creating the classic “3 sign”. Notching of the posterior fourth to eighth ribs due to dilated intercostal arteries may also be seen in older patients[24,25]. Electrocardiogram is typically normal in infants, but in older children and adults, left ventricular hypertrophy is common due to ventricular pressure overload[24].

Transthoracic echocardiography can assess the presence and severity of aortic coarctation and any associated cardiac defects and is the diagnostic gold standard in neonates and infants (Figure 1). Although transthoracic echocardiogram remains the initial test of choice for coarctation, in larger children and adults[24], echocardiographic windows may be suboptimal. When this is the case, a computed tomography scan or magnetic resonance imaging (MRI) provides excellent anatomic detail at the site of coarctation, and these modalities are commonly used to create three-dimensional images for interventional planning (Figure 2). MRI has the additional benefit of defining and quantifying collateral vessel flow. Although cardiac catheterization was frequently used for diagnosis of coarctation in the past, it is now typically reserved for therapeutic intervention or in those cases where hemodynamic data is additive to the diagnostic evaluation[24,25].

Open in New Tab   Full Size Figure   Download Figure

Figure 1 Echocardiogram of coarctation. A: Two-dimensional transthoracic echocardiogram image obtained from the suprasternal notch in an 11-day-old infant demonstrating discrete coarctation (arrow); B: Color Doppler of the same image with aliasing of flow at the site of coarctation (arrow).

Open in New Tab   Full Size Figure   Download Figure

Figure 2 Magnetic resonance imaging of coarctation. A: Magnetic resonance image (steady-state free precession) in a sagittal projection demonstrating transverse arch hypoplasia and long segment coarctation of the aorta distal to the left subclavian artery (arrow) in a 12-year-old male; B: Three-dimensional reconstruction of a gated contrasted angiogram for the same patient, which demonstrates transverse arch hypoplasia, coarctation at the aorta at the distal transverse aortic arch and isthmus (arrow), and dilated intercostal arteries (collaterals).

The first surgical technique described for coarctation of the aorta was resection with end-to-end anastomosis by Crafoord in 1944[30] (Figure 3A). Early studies showed recoarctation rates in over half of patients with this technique, and the use of a circumferential suture line was thought to be a major contributor[31,32].

Open in New Tab   Full Size Figure   Download Figure

Figure 3 Surgical techniques in coarctation repair. A: Resection and simple end-to-end anastomosis. The coarctation is resected, and an end-to-end, circumferential anastomosis is created; B: Patch aortoplasty. An incision is extended across the coarctation, and a patch is sutured in place to enlarge the stenotic region; C: Subclavian flap aortoplasty. The left subclavian artery is ligated and divided. A longitudinal incision is extended from the proximal left subclavian artery beyond the area of coarctation, and the proximal left subclavian stump is folded down to enlarge the area of coarctation; D: Resection with extended end-to-end anastomosis. The coarctation is resected using a broad, longitudinal incision, and an oblique anastomosis is constructed between the undersurface of the transverse arch and the descending thoracic aorta. Figures adapted and reprinted with permission from the Journal of Cardiac Surgery[30].

Vosschulte described prosthetic patch aortoplasty as an alternative technique for coarctation repair in 1961. In this approach, the ductal tissue is ligated and divided, a longitudinal incision across the coarctation is made, and a prosthetic patch is sutured in place to enlarge the stenotic region (Figure 3B). This technique can be applied to longer segments of coarctation, avoids a circumferential suture line, and minimizes aortic mobilization and ligation of intercostal arteries[33]. While recoarctation rates of 5%-12%[34] were lower compared to the resection and end-to-end anastomosis technique, aortic aneurysm was a long-term concern, with rates between 18%-51% of patients[35-38]. Using more distensible polytetralfluroethylene patch material instead of Dacron was initially promising[34] but still showed a 7% risk for aortic aneurysm and a 25% risk of recoarctation[39].

Subclavian flap aortoplasty was a modified approach reported by Waldhausen and Nahrwold in 1966. Here, the left subclavian artery is ligated and divided, and a longitudinal incision down the proximal left subclavian artery is extended beyond the area of coarctation. The proximal left subclavian stump is then folded down to enlarge the area of coarctation (Figure 3C). This technique avoids a circumferential suture line and the use of prosthetic material, which may allow for improved growth, and it can be used for repair of long segment coarctation[40,41]. Although still occasionally used by surgeons, one of the main reservations of this approach has been the need to sacrifice the left subclavian artery. This can create a subclavian steal phenomenon, with retrograde flow down the vertebral artery, and it has been associated with decreased length and muscle bulk of the left upper extremity, as well as claudication with exercise[42,43].

Coarctation resection and replacement with an interposition graft was described by Gross[44] in 1951. After aortic cross clamp, the stenotic tissue is excised, and either a homograft or Dacron tube graft is sutured into the aorta. This approach is rarely used in the current era, as it is not ideal for pediatric patients due to growth limitations. However, occasionally it is an appropriate technique for adult patients with coarctation, especially those with aneurysm, long segment coarctation, or recoarctation after primary repair[45].

In 1977, Amato et al[46] described a modification to Crafoord’s resection and end-to-end anastomosis technique, where a broader, longitudinal incision and anastomosis are created across the proximal aorta (Figure 3D). The extended end-to-end anastomosis still avoids the use of prosthetic material and allows resection of the coarctation and residual ductal tissue, but the wider incision is less prone to restenosis and enables enlargement of the transverse aorta, which is particularly helpful in neonates[46,47]. In the present era, extended end-to-end anastomosis is typically the preferred technique for surgical repair, especially in small children, due to low mortality rates and low rates of restenosis, ranging between 4%-11%[47-50].

Surgical therapy was the only treatment option for coarctation until 1982, when the use of balloon angioplasty was described by Lock et al[51]. Several studies since then have shown balloon angioplasty to be a relatively effective acute intervention for native coarctation, with rates of recoarctation ranging from 8%-32%[52-54]. In a report from the prospective, multicenter Congenital Cardiovascular Interventional Study Consortium (CCISC), 34 patients undergoing balloon angioplasty for native coarctation had adequate intermediate follow-up data at 18 to 60 mo post-intervention. In these patients, the rate of recoarctation was 15%[55]. A second concern with native coarctation angioplasty is aneurysm formation. Histologic and intravascular ultrasound studies have demonstrated the mechanism of angioplasty involves tearing of the intima and media[56-59]. Although some of these tears may heal, disruption of vascular integrity is believed to contribute to a relatively high incidence of aneurysm formation. This was demonstrated in a single center randomized trial comparing balloon angioplasty vs surgical repair of coarctation in older children (ages 3 years to 10 years). In this study with mean follow-up of 14 years after intervention, 35% of the balloon angioplasty patients developed aneurysm, compared to none of the surgical patients[60]. Similarly, the 2014 CCISC observational study showed 24% of patients with native coarctation developed aortic aneurysm at intermediate follow-up after balloon angioplasty[55].

In contrast to native coarctation, balloon angioplasty is often the preferred intervention for recurrent coarctation in children[27]. Acute success rates for this procedure range from 80%-93%[61]. Reported rates of aortic wall injury are low (1%-2%), and the longer term risk of aneurysm is believed to be ameliorated by scar tissue at the site of the recoarctation, which limits the degree of vascular disruption. However recoarctation rates remain a concern, with a broad range between 6%-53% described[62,63].

Likely the most fragile patient population to develop recurrent coarctation is children with hypoplastic left heart syndrome or other single right ventricle lesions. These patients are at risk for significant morbidity and mortality with recoarctation due to exacerbation of atrioventricular valve regurgitation and ventricular dysfunction[64]. The Pediatric Heart Network Single Ventricle Reconstruction trial was a large, multicenter, prospective study examining the outcome of infants with single right ventricle lesions after randomization to either right ventricle-pulmonary artery shunt or a modified Blalock-Taussig shunt at the time of Norwood procedure[65]. The incidence and timing of intervention for recoarctation in the first 12 mo after randomization was assessed, and 97 of 549 patients (18%) underwent intervention for recoarctation, which was most commonly performed at the time of pre-stage II cardiac catheterization by balloon angioplasty. Balloon angioplasty achieved adequate short term results, but 39% of patients required reintervention for recoarctation, compared to 5% of the patients who underwent surgical reintervention. Reassuringly, with catheter or surgical intervention, the presence of recoarctation did not affect survival rates in this tenuous patient population[64].

In 1991, the use of endovascular stents for the treatment of coarctation was first reported[66], adding another dimension to the utility of transcatheter treatment for coarctation. Endovascular stents are inserted using balloon catheters but do not require overdilation of the vessel wall. Stents also offer structural support, thereby decreasing the rates of aortic wall injury and restenosis observed with balloon angioplasty alone[5] (Figure 4). Several studies have assessed the utility of endovascular stent placement in the treatment of coarctation[67]. Retrospective analysis of acute procedural data from 17 institutions from 1989 to 2005 showed successful stent placement without a significant residual gradient or serious complication in 553 of 565 (97.9%) patients. The overall complication rate was 14.3%, with aortic wall complications (aneurysm, intimal tear, or dissection) contributing 3.9%[68]. In a subsequent study, acute and long-term data regarding endovascular stent placement for coarctation were obtained from the CCISC. Here, the acute results of stent placement were successful without a significant blood pressure gradient or need for reintervention in 249/260 (96%) of cases. During follow-up spanning between 3 to 60 mo, recoarctation was seen in 20% of the 164 patients with follow-up imaging, and 4% of patients required unplanned repeat interventions. Aortic wall complications consisting of dissection or aneurysm were seen in 1% of the 302 total cases[67].

Open in New Tab   Full Size Figure   Download Figure

Figure 4 Endovascular stent placement for coarctation. A: Angiogram (LAO 30°, caudal 30°) demonstrating a discrete coarctation and intercostal aneurysm (arrow) in a 45-year-old male; B: Angiogram in the same projection after endovascular bare metal stent placement showing no significant residual stenosis. The intercostal aneurysm was successfully occluded with an Amplatzer Vascular Plug II (arrow).

In a multicenter observational study comparing the outcomes of surgical, balloon angioplasty alone, and endovascular stent placement for coarctation using the CCISC registry, patients undergoing stent placement had significantly lower complication rates compared to balloon angioplasty or surgical patients. There was no significant difference among treatment groups for unplanned reintervention rates (4%-7%) at a mean of 1.7 years of follow-up, but those who underwent stent placement were more likely to undergo planned reintervention. Aortic wall injury was more likely to occur in patients who underwent balloon angioplasty alone, at a rate of 9.8% compared to none in the endovascular stent group[69]. In an attempt to more rigorously assess the effectiveness and safety of endovascular stent placement vs surgical repair, a Cochrane review was attempted in 2012, but it was determined that there was insufficient data available to identify the best treatment modality[5].

The Coarctation of the Aorta Stent Trial (COAST) has been an influential prospective study examining the effectiveness and safety of endovascular stent placement in children and adults with coarctation. Currently, no endovascular stent has been granted FDA approval for use in the aorta, and initially biliary stents were used off-label for treatment of coarctation. NuMED (Hopkinton, NY) created the Cheatham Platinum (CP) stent for specific use in the aorta, and in 2007, the COAST trial was designed as a prospective, multicenter, single-arm clinical study to assess the safety and efficacy of this stent for the treatment of coarctation[6].

Results for up to two years from CP stent placement are currently available from the COAST trial. CP stent implantation was attempted in 105 patients ranging from 8 to 52 years of age. All but one implantation was successfully completed, with no significant adverse events acutely. No patient had a significant gradient between the ascending and descending aorta after stent placement, and 99% had a gradient < 20 mmHg at one month. There was 89% follow-up at one year after stent placement and 86% follow-up at two years. At two years, 90% of patients maintained a blood pressure gradient < 20 mmHg between the upper and lower extremities. To date, 23 fractured stents have been identified, though none have led to decreased stent integrity, stent migration, aortic wall injury, or hemodynamic obstruction. Aortic aneurysm was diagnosed in 6 patients, one of which resolved without intervention. No patient has required surgical reintervention, and a total of 19 patients have undergone repeat transcatheter intervention, either due to aortic wall injury or dilation of the initial stent[70].

Thus far, the CP stent is felt to be a safe and effective treatment option for coarctation in older children and adults with native or recurrent coarctation. Stent fracture is common but thus far has been clinically insignificant. Reintervention is also common, either due to planned dilation of smaller stents or aortic wall injury. Follow-up for the COAST trial is planned for up to 60 mo after stent placement and will provide further insight regarding these issues[6].

The use of endovascular stents in small children remains controversial due to the challenges in accommodating for significant somatic growth and the requirement for relatively larger sheath sizes. The ideal stent would be deployed through a small sheath but retain the ability to be dilated to an adult vessel diameter. This technology does not yet exist and still would require multiple interventions for stent dilation throughout a patient’s lifetime. Additionally, neonates often have transverse arch hypoplasia, which does not easily lend itself to endovascular stent placement. Some small, single center studies have had positive short-term results with endovascular stent placement in young children, but further follow-up and investigation is needed[71,72].

Covered endovascular stents represent the latest transcatheter innovation and were first used for the treatment of coarctation in 1999. The fabric within the stent provides additional structural support, creates a protective barrier at the site of stent placement, and can help decrease shear stress. All of these factors theoretically reduce the risk of acute vascular trauma as well as longer term aneurysm formation. When aortic aneurysm or stent fracture occurs with bare stent placement, covered stents are often used as a rescue therapy. Covered stents may also be the initial transcatheter intervention of choice, especially in the setting of complex anatomy of the coarctation or in older patients with more friable and calcified aortic wall tissue. However, covered stents require larger sheath sizes, which limits their use in small children. Additionally, care must be taken to avoid stent occlusion of significant aortic branches, including paraspinal branches off of the descending aorta, which can be difficult to identify[73].

Initial smaller studies examining the use of covered stent placement in aortic coarctation were promising, with no reported aneurysms[73,74]. More recently, a randomized clinical trial was performed comparing bare CP stent vs covered CP stent placement in 120 patients. Both groups had no acute procedural complications, and at an average follow-up of 31.1 mo, the bare CP stent group had a statistically nonsignificant increase in the rate of recoarctation (6.7% vs 0%) and a nonsignificant lower incidence of pseudoaneurysm (0% vs 3.3%), compared to the covered CP stent cohort. The two cases of pseudoaneurysm in the covered stent group occurred at the proximal end of the stent, and both were successfully treated with a second covered stent with no further complications[75]. Further investigation into the safety and efficacy of covered stents is on-going with the COAST II trial. This trial was initiated in 2010, hoping to provide information that will support FDA pre-market approval of the covered CP stent in preventing aortic wall injury in high risk patients with coarctation as well as treatment of existing aortic wall injury related to complications from previous interventions for coarctation. Results are expected in the near future, but at this time, covered CP stents are not available for the treatment of coarctation in the United States outside of use in the COAST II trial[70].