Letter by Heidrich et al regarding article, "A word of caution: risk of device erosion after percutaneous treatment of atrial septal defect in patients with dilated aortic root".

We appreciate the interest of Dr Heidrich and his colleagues in our recent report. We agree overall with their comments, and we think that, until large retrospective or perspective studies clarify this issue, bicuspid aortic valve as opposed to tricuspid aortic valve should be a relative contraindication to percutaneous device closure of atrial septal defects because of the high risk of aortic root dilation creating friction with the device and therefore the risk of aortic rupture or dissection. The prevalence of bicuspid aortic valve in the general population is estimated to be between 0.5% and 2.0%, and aortic dilatation occurs in up to 50% of adults with bicuspid aortic valve. Marfan disease and other fibrillinopathies could also be considered a contraindication to percutaneous device closure, given that the percentages of patients with an ascending aortic dilatation and the presence of an aortic event (dissection or prophylactic surgery) are 96% and 74%, respectively, until 60 years of age.