Procedural risk factors, incidence and timing of reintervention after treatment for native coarctation of the aorta in children: a population-based study.

OBJECTIVES The aim of the present study was to evaluate procedural risk factors, incidence and timing of reintervention because of recurrent aortic coarctation in children. METHODS The study cohort consisted of 304 patients with isolated coarctation: 251 underwent surgery and 53 were treated percutaneously (40 balloon angioplasty, 13 stent) at the Helsinki Children's Hospital in 2000-2012. Characteristics, intervention and reintervention data were retrospectively collected from clinical records until 2014 (median follow-up 7.9years). Age- and sex-matched comparisons between the treatment groups were performed in 86 patients (surgery n=43, percutaneous n=43). RESULTS Forty of the 251 (16%) patients after surgery, 9/40 (23%) patients after balloon angioplasty and 4/13 (31%) patients after stent placement underwent a reintervention after a median time of 3.4, 11.7 and 19.5months (P<0.05), respectively. In the surgery group, all reinterventions occurred in children operated on 12months of age and were related to lower body weight and smaller dimensions of the aorta. In the balloon angioplasty group, a higher post-procedure systolic arm-leg blood pressure gradient was associated with reintervention. After stent placement, three-fourths of the reinterventions were performed in a planned postinterventional catheterization. In the age- and sex-matched comparisons (median 5,7 years, range 0,5-17,6), post-procedure blood pressure gradients were higher (mean 10 vs 4mmHg, P=0.03), and reinterventions were more common (28%, 95% confidence interval 17-43 vs 2%, 95% confidence interval 0-12) in the percutaneous group compared to the surgery group. CONCLUSIONS Reinterventions after surgery in neonates were relatively common. In older children, percutaneous treatment carried a higher risk of reinterventions, which were mainly related to residual coarctation after primary treatment.