Determinants of survival following repair of interrupted aortic arch in infancy.

Between January 1971 and March 1987, surgery was performed in 26 infants with interrupted aortic arch. At operation the 14 boys and 12 girls weighted between 1.71 and 4.23 kg (mean +/- SD = 3.1 +/- 0.63 kg) and ranged in age from 2 to 90 days (13 +/- 18 days). The interruption was distal to the left subclavian artery in 4 (15%), between the left carotid and subclavian arteries in 20 (77%) and between the brachiocephalic (innominate) and left carotid arteries in 2 (8%). Associated complex cardiac lesions in 8 patients included complete transposition (2), common arterial trunk (2), aortopulmonary window (2), double inlet left ventricle (1) and tricuspid atresia (1). The remaining patients had an isolated ventricular septal defect. The arch was reconstructed with a prosthetic conduit in 14 patients; by a direct anastomosis in 6; using the subclavian artery in 3; and with the pulmonary trunk and the arterial duct in 2. Twenty patients (77%) underwent palliative surgery as the first stage of management, and banding of the pulmonary trunk was also performed in 16 of these. Five patients (19%) underwent primary complete repair of the interruption and intracardiac anomalies. One patient (4%) died soon after thoracotomy for palliative surgery. Of the 15 (57%, 70% confidence limits CL = 46-69%) early deaths, 7 occurred in patients with complex associated defects and 4 occurred when single stage repair was attempted. Survival following first-stage palliative surgery for arch interruption with isolated ventricular septal defect was 64% (9/14) [70% CL = 47-79%]. All of these patients subsequently underwent complete repair. Chi-squared and t-tests showed the year of operation and the type of operation (two-stage repair) to be associated with improved survival. It is concluded that a two-stage repair of interrupted aortic arch offers a reasonable alternative to primary complete correction and will lead to satisfactory subsequent repair in most cases.