A 77-year-old woman with severe myxomatous mitral regurgitation underwent mitral valve replacement. Twenty-six months previously, she had undergone percutaneous closure of an atrial septal defect (ASD) with an Amplatzer device. At the time of surgery inspection of both sides of the septal occluder device revealed bare Nitinol wires suggesting failed endothelialisation (Panel A).

• Download figure
• Open in new tab
• Download powerpoint

Panel A. Intraoperative image of the atrial septal defect closure device, viewed from the left atrium, showing the bare Nitinol mesh.

The use of percutaneous closure devices for the treatment of atrial septal defects is increasing. However, our current knowledge regarding the long-term safety and efficacy of these devices is limited. This case illustrates that endothelialisation of such devices may be significantly delayed or even absent following implantation, raising the possibility of late thromboembolic complications. Presently there is no consensus of opinion regarding antiplatelet therapy following device implantation.1 In the United Kingdom, general guidelines issued in 2004 regarding the percutaneous implantation of ASD closure devices estimated the incidence of device-associated thrombus formation at 0.4–3%, based on data from large non-randomised studies.2 However, these studies differed in the type of device implanted and the antiplatelet agent regimens employed. We routinely perform transoesophageal echocardiography 6 months postprocedure prior to stopping dual antiplatelet therapy, which is continued thereafter only if there has been a neurological event.

Non-endothelialised devices present potent substrates for thrombus formation and the potential for cerebrovascular accidents. This case illustrates a previously unrecognised and potentially serious problem, highlighting the unresolved issue of the appropriate type and duration of antiplatelet therapy, and the need for strategies to accelerate endothelial coverage of such occluder devices.