Interatrial Septal Closure Devices and Aortic Perforation: A Note of Caution

ABSTRACT: Percutaneous defect closure is well established in patients with a symptomatic patent foramen ovale (PFO). Despite a safe interventional procedure, severe complications have been described during device implantation or follow up. Two patients are described whose cases may serve to warn of an unusual complication associated with a closure device. Several weeks after percutaneous PFO closure, perforation of a device strut into the aortic root was evident in both asymptomatic patients. The perforations were repaired surgically and both patients had further uneventful follow-ups. Procedures which could decrease the risk of device perforation should be adhered to and the indication for closure should consider this potential complication.
Closure of a patent foramen ovale (PFO) or an atrial septal defect (ASD) with a transcatheter double-disc device is a valuable treatment alternative to lifelong medical treatment or cardiac surgery. Closure of a PFO should be discussed in patients with paradoxical embolism and PFO, especially when associated with an interatrial septal aneurysm or further risk factors. However, we present two patients whose cases may serve to warn of an unusual complication associated with a closure device. During regular follow-up visits, we found that in both patients, the device struts had perforated into the aortic root after percutaneous PFO closure. Patient 1. A 31-year-old female with a history of recurrent transient ischemic attacks underwent transesophageal echocardiography (TEE), which revealed a 3 mm PFO and broad movement of contrast agent from the right to the left atrium during the Valsalva maneuver. Aside from the PFO, no other reason for these events could be found: electrocardiography (ECG), electroencephalography (EEG) and blood tests for clotting disorders showed normal parameters. Duplex ultrasound scanning of the carotid arteries excluded a vascular cause. As a prophylaxis for recurrent neurological events, the patient underwent percutaneous closure of the PFO with a Cardia Star 30 mm double-disc device (Cardia Inc., Eagan, Minnesota). The procedure was performed in accordance with the directions for use of the closure system. Peri-procedural TEE showed that the device was properly positioned, without residual shunts. No peri-interventional complications occurred. Four weeks later, the asymptomatic patient underwent a routine control of the device by TEE. Surprisingly, perforation of the aortic root with a small shunt from the aortic root to the right atrium was evident. Cardiac magnetic resonance imaging (MRI) confirmed the diagnosis of device malposition, with perforation of the aortic wall by a device strut, leading to left-to-right shunting between the aortic root and the right atrium. Cardiac surgery was therefore performed. Surgical inspection of the aortic root confirmed the TEE and MRI findings. The struts were not broken or damaged. After removal of the single perforating strut, the bleeding lesion was closed with a felt patch. The main part of the PFO device was knitted with the atrial septum and was left in situ, as this part does not cause any further harm. The patient remained free of any additional complications during the postoperative course and was discharged from the hospital 7 days later. Patient 2. A 25-year-old male presented with a history of cryptogenic cerebral thrombo-embolization. TEE revealed an atrial septal aneurysm (ASA) with a long PFO tunnel. Again, aside from the PFO, no other reason for cerebrovascular events could be found. The patient underwent percutaneous closure of the defect with a Cardia Star 35 mm device. Periprocedural TEE showed a properly positioned device without residual shunts. No complications occurred during the procedure. TTE on the following day showed the device in a typical position with the discs applied parallel to the interatrial septum in the right and left atria. No residual shunt was detectable. Three months after implantation, the patient underwent a routine control of the device without any adverse clinical symptoms. As in patient 1, perforation of the aortic root by a strut of the device was suspected, resulting in left-to-right shunting between the aorta and right atrium (Figure 1). MRI confirmed the diagnosis of device angulation and perforation of the a coronary sinus wall (Figure 2). An emergency cardiac operation was performed. The surgical findings confirmed the diagnosis, and it was found that a strut on the left side of the disc was perforating the aortic root (Figure 3). The struts were not broken or damaged. The 35 mm device and the adherent wall of the interatrial septum were removed. A pericardial patch was used to repair the interatrial septum. The right atrium and the aortotomy were closed directly. The patient had an uncomplicated postoperative recovery and was discharged 6 days later.

Discussion

The purpose of this article is to alert the medical community to the recognition of a rare but potentially serious adverse event following a percutaneous interatrial defect closure. Up to now, erosion and perforation of double-disc umbrella devices has been described as rare complications after implantation, but only for the most widely-used Amplatzer Septal Occluder (ASO) system. Amin et al postulated that the risk of device perforation with ASO is low (≤ 0.1%), and complications can be decreased by identifying high-risk patients such as individuals with a deficient aortic rim and/or a superior rim. Additionally, oversized ASOs may increase the risk of erosion.² To our knowledge, perforation of devices has not been described for closure systems other than the ASO system. There are surprising reports in the literature of delayed heart perforation by an occluder device up to 16 months after implantation.^3,4 Hence, these patients require long-term echocardiographic follow-up to allow early diagnosis of device failure and to prevent emergency surgical intervention. The indication for closure should consider this potential complication, even though the interventional procedure is otherwise safe.⁵ In both presented cases, a Cardia PFO device (Intrasept) was used, which is available and well-known in Europe, but has no FDA approval. This device is a double-umbrella occluder that consists of 2 discs made of Ivalon (polyvinyl alcohol). Each disc has a diameter between 15 and 45 mm and is expanded by 6 radial nitinol arms, resulting in a hexagonal umbrella. In both cases, the perforation was caused by protrusion of a device strut into the aorta. Point pressure on the aortic wall from the adjoining struts is most likely the driving force for perforation. In contrast to the point pressure of the Cardia device strut, the design of the ASO seems more likely to cause friction on the adjacent structures. Oversizing probably contributed to these aortic perforations. However, both devices were selected and implanted according to the directions for use as follows: the lengths of the septum primum in the short and long axes were measured via TEE. The width of the closure system was then selected by doubling the larger of the 2 septum primum measurements. A PFO device within a range of ± 2 mm around this result was chosen. The length of the device was selected by the thickness of the septum secundum in a long-axis TEE view. If the septum secundum was thinner than 5 mm, a 3 mm device was selected; otherwise a 5 mm device was used. Peri-interventional TEE showed that the implanted devices were properly aligned. The presented cases demonstrate that life-threatening aortic perforation of occlusion devices may be detected as an incidental finding in asymptomatic patients and laceration of the aortic wall is not limited to the ASO system. In two cases, perforation of the aortic root by a strut of a Cardia device was evident. The implantation procedures were uneventful in both cases and the patients were asymptomatic at their follow-up visits. The first patient had a properly-located PFO at the center of the interatrial septum without ASA. No anatomical or procedural reasons for the later severe adverse event could be identified. In the second case, a somewhat oversized device was used to cover the PFO which was associated with an ASA. The oversized device in a PFO with ASA was probably the reason for the erosion and perforation of the aorta. The risk of aortic perforation with a Cardia device at our institution is low (≤ 0.5%). Aortic perforations by device struts have not been specifically described for other atrial septal occluder devices, so this should encourage operators to be vigilant, without being an indictment of the class of devices. In general, careful patient and device selection are essential before transcatheter closure of interatrial septal defects. Before the device is implanted, the surrounding anatomic structures should be imaged and safety margins to these structures exactly adhered to. Otherwise, it is possible that the struts of the device may lacerate the aortic wall and lead to life-threatening consequences.

From the Department of Cardiology, Heart Centre, University of Dresden, Dresden, Germany. The authors report no conflicts of interest regarding the content herein. Manuscript submitted September 29, 2008, provisional acceptance given October 22, 2008, final version accepted December 11, 2008. Address for correspondence: Dr. med S. P. Schoen, Department of Cardiology, Heart Centre, University of Dresden, Fetscherstr. 76 - 01307 Dresden, Germany. E-mail: cardiology@email.de

__________________________________________

J INVASIVE CARDIOL 2009;21:E39–E41

References

1. Meier B. Closure of patent foramen ovale: Technique, pitfalls, complications, and follow up. Heart 2005;91:444–448. 2. Amin Z, Hijazi ZM, Bass JL, et al. Erosion of Amplatzer septal occluder device after closure of secundum atrial septal defects: Review of registry of complications and recommendations to minimize future risk. Catheter Cardiovasc Interv 2004;63:496–502. 3. Teoh E, Wilton S, Brecker, Jahangiri M. Simultaneous removal of an Amplatzer device from an atrial septal defect and the descending aorta. J Thorac Cardiovasc Surg 2006;131:909–910. 4. Cecconi M, Quarti A, Bianchini F, et al. Late cardiac perforation after transcatheter closure of patent foramen ovale. Ann Thorac Surg 2006;81:e29–e30. 5. Braun M, Gliech V, Boscheri A, et al. Transcatheter closure of patent foramen ovale (PFO) in patients with paradoxical embolism. Periprocedural safety and mid-term follow-up results of three different device occluder systems. Eur Heart J 2004;25:424–430.