Predictors of spontaneous closure of isolated secundum atrial septal defect in children: A longitudinal study

OBJECTIVES. The goals were to assess the frequency of spontaneous closure of isolated secundum atrial septal defect in children and to identify predictors of spontaneous atrial septal defect closure. METHODS. A retrospective cohort study was performed in a tertiary care pediatric cardiology center. Consecutive patients (n = 200) diagnosed as having isolated atrial septal defects (no multiple or fenestrated atrial septal defects, no additional congenital heart disease, and no syndromes) were monitored for > 6 months with serial 2-dimensional echocardiography, according to a standardized protocol. RESULTS. The median age at diagnosis was 5 months (minimum: 0 months; maximum: 13.9 years). The atrial septal defect diameter at diagnosis was 4 to 5 mm in 40% of cases, 6 to 7 mm in 28% of cases, 8 to 10 mm in 21% of cases, and > 10 mm in 11% of cases. The median age at the final follow-up evaluation was 4.5 years (range: 6.8 months to 16.2 years). Thirty-four percent of atrial septal defects showed spontaneous closure, and 28% decreased to a diameter of <= 3 mm. Logistic regression analysis revealed atrial septal defect diameter and age at diagnosis as independent predictors of spontaneous closure or regression to <= 3-mm defect size. Of atrial septal defects with a diameter of 4 to 5 mm at diagnosis, 56% showed spontaneous closure, 30% regressed to a diameter of <= 3 mm, and none required surgical closure. Of atrial septal defects with a diameter of > 10 mm at diagnosis, none closed spontaneously, whereas 77% required surgical or device closure. Gender and observation time were not associated with spontaneous atrial septal defect closure or regression to <= 3 mm. CONCLUSIONS. In the present study population of children with atrial septal defects, 62% showed spontaneous closure (34%) or regression to <= 3 mm (28%). Initial atrial septal defect diameter was the main predictor of spontaneous closure.