Protruding Umbrella into the Left Atrium Associated with<br />
Recurrence of a Stroke: A Possible Late Complication of the CardioSEAL

ABSTRACT: Percutaneous closure of a patent foramen ovale or atrial septal defect using various closure systems is now considered an alternative to surgical therapy in selected patients with cryptogenic strokes. We report the case of a late protrusion of the umbrella of a CardioSEAL Occluder into the left atrium associated with occurrence of a stroke. No recurrence of symptoms was seen at six-month follow-up after reinstitution of clopidogrel in addition to aspirin.

Key words: late complication, patent foramen ovale, stroke

Non-surgical closure of patent foramen ovale (PFO) or atrial septal defects (ASD) has become a widely applied procedure in selected patients. Late dislodgement and embolization of the occluders utilized in these procedures have been infrequently reported.^1–5 We present a case of late protrusion of the umbrella of a CardioSeal Occluder (NMT Medical, Boston, Massachusetts) into the left atrium associated with recurrence of a stroke. Case Report. A 68-year-old male was referred to our center for percutaneous closure of a PFO diagnosed after he sustained a stroke confirmed by magnetic resonance imaging (MRI). He underwent a detailed neurological evaluation, including carotid ultrasound, MRI angiography of the cerebral circulation and a hypercoagulable screening. All studies were unremarkable. He was a non-smoker, but had a history of well-controlled hypertension and diabetes mellitus. He had no history of atrial fibrillation or known cardiac disease. A transesophageal echocardiogram (TEE) was performed and revealed no intracardiac source of emboli except for the PFO, which was confirmed by saline bubble contrast. The patient was placed on coumadin, but had difficulty maintaining an INR (International Normalized Ratio) in the therapeutic range. He had no further strokes or transient ischemic attacks (TIA). He wished to be off coumadin and wanted to pursue percutaneous closure of the PFO. After explaining the risks and benefits of the procedure, consent was obtained from the patient and percutaneous closure of the PFO was carried out using the CardioSEAL Occluder. PFO size was determined with a sizing balloon and measured at 10 mm. The inter-atrial septum was redundant, and therefore the Occluder was sized to cover the area of redundancy. A 33 mm CardioSEAL Occluder was successfully deployed using the right common femoral vein approach (Figure 1A). There were no complications or residual shunt immediately post-procedure (tested with Valsalva and saline bubble contrast). The patient was placed on clopidogrel 75 mg per day and aspirin 324 mg per day for three months followed by aspirin alone. Three months after placement of the Occluder, a transthoracic echocardiogram was performed, which revealed continued complete sealing of the PFO and no obvious anomalies at the interatrial septum. Six months after the procedure, the patient presented to the emergency room with a 2–3 day history of intermittent weakness of the left arm and some confusion. An MRI confirmed the presence of a new stroke. Carotid Doppler was repeated and was unremarkable. TEE was repeated and showed normal left ventricular function, with no vegetations or mobile thrombi seen. The left atrial umbrella of the Occluder now protruded toward the center of the left atrium (Figure 1B). It appeared fixed and no thrombus was seen on the Occluder. A chest x-ray was performed and showed no indication of fracture of the Occluder. A saline bubble study showed no shunt across the inter-atrial septum despite cough and Valsalva. The patient was placed back on clopidogrel in addition to his aspirin. At six-month follow-up, he had no further recurrence of strokes or TIA-like symptoms. Discussion. Closure of a PFO or ASD with a percutaneous system is now considered an alternative to surgical therapy in selected patients with cryptogenic strokes. Late dislodgement and protrusion of these Occluders have been infrequently reported in the literature.^1–5 Celiker et al.¹ reported a case of protrusion of the umbrella of a CardioSEAL Occluder into the right atrium after 18 months of deployment. Similarly, among 91 asymptomatic patients followed by serial TEE, Kauliz et al.³ reported protrusion (n = 5) or malposition (n = 7) of one arm of the CardioSeal Occluder in twelve patients without evidence of thrombus formation, atrial arrhythmia or infective endocarditis. Single fatigue fracture was seen in seven patients (7.7%) within the first six months after implantation. No fracture was seen in our case, as evaluated by chest radiography. Furthermore, Carminati and colleagues⁵ reported one case of malposition and one case of embolization on follow-up of 334 patients who received the CardioSEAL or Starflex double-umbrella Occluders, both required surgical removal. The association of protrusion or dislodgement of the CardioSEAL with subsequent thrombus formation and recurrence of strokes or TIAs has been previously reported.² In our case, there was no indication of a thrombus on the protruded left atrial umbrella of the Occluder. Blood turbulence, however, around the protruded umbrella could have created a transient thrombus with subsequent embolization. In our case, clopidogrel added to aspirin appeared effective at 6-month follow-up in preventing a subsequent cerebrovascular event. It is unclear how the Occluder arm protruded following deployment into the left atrium. We had selected relatively large Occluder size with the intention to stabilize the redundancy of the interatrial septum. It is possible that the device was too large, resulting in suboptimal positioning. Whether a larger size Occluder has a higher chance of late protrusion compared to a smaller one remains unknown at this time. This report has its limitations. We cannot exclude the possibility that the protruding arm of the Occluder could have been missed by the tranthoracic echocardiogram at three-month following examination. Also, the discovery of a protruding left arm by TEE at six months after implantation and following an evaluation of a recurrent stroke could have been coincidental. It is clear that our case does not prove with certainty that the first and second strokes were related to the PFO or the protrusion of the umbrella of the Occluder. Nevertheless, the coincidental occurrence of two infrequent late complications — stroke after PFO closure with the CardioSeal Occluder and the protrusion of its left atrial umbrella — makes it possible that an association exists between these two events.