Surgical repair of coarctation of the aorta in infancy can now be performed with very low operative mortality, in even the youngest and smallest patients. However, on long-term follow-up, recurrent coarctation may necessitate further aortic arch intervention, in the form of either percutaneous balloon angioplasty or surgery. In the previous surgical era, infants were shown to have the highest risk for reintervention, with a reoperation rate as high as 26% for those undergoing repair during the first year of life. In the modern surgical era, rates of reintervention after repair in infancy have improved to 11% to 14%, with a majority of reinterventions occurring within 1 year of initial repair.

Previous studies have suggested that hypoplasia of the aortic arch and isthmus may be correlated with outcomes in patients undergoing surgical repair. A defining characteristic of the modern era in coarctation surgery has been direct address of the hypoplastic aortic arch with the so-called extended end-to-end anastomosis, as opposed to either subclavian flap angioplasty or coarctectomy with simple end-to-end anastomosis. Given that this newer technique addresses aortic arch hypoplasia at the time of initial operation, the predictive value of preoperative aortic arch hypoplasia may be questioned. We therefore sought to determine whether aortic arch hypoplasia remains a risk factor for recurrent coarctation and to evaluate the role of other preoperative aortic dimensions in correlation with coarctation recurrence.

Methods

A retrospective review of medical records of all patients who underwent surgical repair of coarctation of the aorta during infancy at our institution from July 2002 through February 2009 was performed. This study was approved by the University of Arkansas for Medical Sciences Institutional Review Board. Patients with single-ventricle physiology were excluded. Data collected included patient demographics, associated cardiac diagnoses, and details of cardiac surgery, including the approach and type of repair. The surgical approach to the initial coarctation repair was based on the discretion of the surgeon, with a tendency to use median sternotomy and cardiopulmonary bypass for children with important associated cardiac defects to be simultaneously repaired or worse proximal aortic arch hypoplasia. Recurrent coarctation was defined as clinical (arm-leg systolic blood pressure gradient) and Doppler-derived peak gradient >20 mm Hg, with a diastolic runoff pattern in the descending aorta. Surgical or catheter-based reintervention was performed on all patients with recurrent coarctation. Early reintervention was defined as that performed within 1 year of initial surgery and was based on echocardiographic evidence of recurrent coarctation. Follow-up data regarding timing and type of reintervention were recorded.

An observer blinded to clinical outcomes analyzed preoperative echocardiograms offline. Measurements included the diameters of the pulmonary valve annulus, aortic valve annulus, aortic sinus, sinotubular junction, ascending aorta, proximal and distal transverse aortic arch, aortic isthmus, coarctation segment, and abdominal aorta ( Figure 1 ). The distance between the left subclavian artery and the carotid artery was also measured. These measurements were converted to Z scores on the basis of body surface area using regression equations published previously. The carotid-subclavian index was derived as a ratio of the aortic arch diameter at the left subclavian artery to the distance between the left subclavian artery and left common carotid artery. A novel index, called the aortopulmonary index (API), was calculated as the ratio of the aortic valve diameter to the pulmonary valve diameter. The diameters of the transverse aortic arch and aortic isthmus were also indexed to the diameters of the ascending aorta and abdominal aorta, as done in previous studies.

Schematic of the pulmonary artery, the aorta, and its arch showing the various sites of echocardiographic measurements (A) . Representative echocardiographic images from parasternal long-axis view of the pulmonary valve (B) and aortic valve and sinus (C) , as well as a long-axis view of the aortic arch from the suprasternal notch (D) , with sites of measurement annotated.

Statistical Analysis

Actuarial freedom from early reintervention was determined for the entire sample using nonparametric Kaplan-Meier estimates. Univariate analysis was used to compare risk factors between status groups using nonparametric Wilcoxon and logistic regression analysis for continuous variables or Pearson’s χ ² test for categorical variables. After identifying risk factors that were significantly different between the groups, locally weighted scatterplot smoothing regression analysis was used to determine whether a linear relationship existed. Receiver operating characteristic (ROC) curves were then used to analyze the value of an abnormal Z score (< −2) in correlation with the need for reintervention. For other indices, ROC analysis was also used to identify cutoff values that optimized sensitivity and specificity as a diagnostic test, and the area under the curve (AUC) was calculated. Kaplan-Meier survival analysis was also used to compare groups according to the previously determined cutoff values. Analysis of variance was used to evaluate any correlation between cardiac diagnosis and API or reintervention.