From January 2003 to June 2010, of a total of 16 patients, 15 patients with PFO (seven males, and eight females) were referred for percutaneous closure of PFO. The primary reasons for referral were stroke {12/16 (75%)}, and TIA {3/16 (18%)}. The symptoms included motor deficit in eight patients (50%), aphasia in one patient (6%), visual disturbances in four patients (25%), headache in four patients (25%), stupor in one patient (6%), and dysarthria in three patients (18%) (Table 3). The patients' conditions were initially diagnosed by neurologists. One patient who had migraine was referred for percutaneous PFO closure from pediatric outpatient department. Multiple thromboembolic events occurred in six patients prior to transcatheter device closure of PFO. The vascular risk factors were infrequent (Table 1). Cryptogenic stroke/TIA was present in eight patients (53%). ASA was present in seven patients (43%). The electrocardiographic findings before the procedure were all non-specific. Cryptogenic stroke was defined by the presence of a transient or permanent neurological deficit with associated magnetic resonance imaging (MRI) evidence of an embolic lesion, in the absence of a clear cause.10) Ischemic stroke was defined as a cerebrovascular event with neurological symptoms lasting more than 24 hours without any sign of a primary cerebral hemorrhage. TIA was defined as a new neurological deficit with complete cessation of symptoms for 24 hours in the absence of lesions in the MRI.11) Migraine was defined as a severe recurring headache lasting from 4 to 24 hours, characterized by sharp pain, and is often accompanied by nausea, vomiting, and visual disturbances.

TEE was performed to detect PFO. An agitated saline bubble study was conducted to evaluate the degree of the right-to-left shunt at rest and after the Valsalva maneuver. The air bubble test was performed in all the patients, and all of them showed positive results except for one patient, wherein the result was not positive because of poor cooperation.

The Amplatzer PFO occluder device was used in this study. It was constructed from the nitinol wires and consisted of two flat disks with a 3-mm long thin connecting waist. The right atrial disk was larger than the left atrial disk, which made it suitable for the occlusion of PFO. In this study, three sizes of the device were used, 18 mm, 25 mm, and 35 mm (Fig. 1).

The full procedure was performed under general anesthesia and was assisted by TEE through the femoral artery and vein. The NIH catheter (Cordis Europa N.V., Roden, The Ne-therlands) was advanced through an 8-Fr sheath, and pulmonary venous angiography was performed in LAO 45° view via the femoral vein. During TEE monitoring, the size of PFO with or without ASA was estimated using a balloon catheter (Boston Medical Corporation, Boston, MA, USA), and the size of the device was selected accordingly. The Amplatzer PFO occluder was implanted in the interatrial septum with great care under the water seal to prevent air embolism.

Before and after the procedure, all the patients received intravenous heparin to prevent thrombus formation, and the transseptal sheaths were removed after the activated coagulation time stabilized (activated clotting time <140 seconds) with manual compression for achieving hemostasis, and an antidote was not used.

TEE was performed before the patient was discharged, mostly at two days after the procedure. After the occlusion of PFO, the air bubble test was performed again and all the patients showed negative results. After the procedure, 100 mg/day of aspirin for six months was prescribed for all the patients.

All the patients returned for regular check-up visits at one, two, six and twelve months after the procedure, and yearly thereafter. Each patient underwent at least one TEE examination during the first six months after the procedure to confirm the adequacy of placement of the device and to document the absence of a residual shunt or thrombus formation on the device. Frequent clinical follow-ups at six- to 12-month intervals were performed with a questionnaire to evaluate the symptoms, and evidence of recurrent thromboembolic events. A detailed history of new neurological incidents was taken, and physical as well as neurological examinations were performed during each follow-up visit.

The clinical variables of the group that underwent PFO closure are listed in Table 1. There were seven male, and nine female patients. Their median age was 43.8 years (range, 10.2-69.8 years).

The procedure results are shown in Table 2. The device was successfully implanted in all the patients under general anesthesia. The sizes of the Amplatzer devices used were 18 mm (n=4), 25 mm (n=8), and 35 mm (n=4). The mean duration of the fluoroscopy time was 19.3 minutes (range, 10.0-35.6 minutes). In all the patients, the PFO was closed completely (Fig. 2A). TEE and chest PA, which were performed before the patients were discharged from the hospital, confirmed the correct placement of the device in the interatrial septum in all the patients. No procedural complications were observed.

All the patients were followed up for a mean period of 54 months (range: 10-88 months). There were no recurrent episodes of cerebrovascular accidents or TIA. Of all the 16 patients, the initial symptoms almost subsided in 10 patients (62.5%), due to which the anticoagulation therapy was discontinued. Three patients were in a minimally assisted gait state, and three patients were in a assisted daily life state, or had lt. hemianopsia, and dysarthria, respectively (Table 3). Three patients had taken anticoagulation therapy to prevent the recurrence of cerebrovascular events. A follow-up TEE was performed during the first six months after the procedure, and no residual shunt was observed. None of the patients documented atrial fibrillation before or after the implantation of the device. No dislocation of the device or fracture of the occluder construction was detected (Fig. 2B). The symptoms in the patient who had migraine subsided after occlusion of the PFO. However, the patient who developed pulmonary embolism was an exceptional case. A 55-year-old woman who had an acute infarction in the left middle cerebral artery was admitted to the authors' hospital due to the development of dysarthria and right-sided weakness, and recurrence of stroke was detected on an MRI scan (Fig. 3). The patient was referred to the Pediatric Department and she underwent PFO closure with an 18 mm Amplatzer PFO occluder. During a follow-up check-up at five months after the PFO closure, the patient had dyspnea and was readmitted to the hospital via the ER. The patient's oxygen saturation was found to be 94-95%, and chest X-ray showed a prominent right hilum (Fig. 4). The 2D echocardiogram revealed severe pulmonary hypertension, right ventricular dysfunction, and tricuspid regurgitation {TR G (III)/VI; TR peak pressure: 82.46 mmHg} (Fig. 5). Neither thromboembolism nor a shunt was observed in this patient treated using the PFO Amplatzer occluder (Fig. 6). However, acute pulmonary embolism was found on the chest CT angiogram (Fig. 7). Hence, 25000 UI bolus of heparin was injected, following which 12 UI/kg/hr of heparin was infused. This protocol was discontinued after the fourth day of heparin administration because the aPTT was 60-90 seconds and the international normalized ratio was 2-3. The patient's symptoms improved after two days of hospital admission. A lower extremity doppler sonography was performed to identify the cause of the pulmonary embolism, but no deep vein thrombosis was detected. No sources for embolism were found either on the abdominal pelvic CT or. Ten days after the pulmonary embolism, the previous RV dysfunction was no longer observed on the 2-D echo, and the patient had slight sequelae.