Patients and methods

Results

The anthropometric data and procedural details are given in table1. Device implantation was successful in all 20 patients, with no early or late embolisation. Of the 16 patients with atrial septal defects > 10 mm diameter as assessed by transoesophageal echocardiography, the Qp:Qs ranged from 1.2:1 to 3.3:1 (median 2.5:1). Their defects ranged in size from 11–22 mm (median 18) on transoesophageal echocardiography, with balloon sized diameter (and therefore device size) of 13–28 mm (median 20). For the entire group of 20 patients, the procedure time was 38–78 minutes (median 61), and fluoroscopy time 8.4–24.7 minutes (median 15.2).

In all but three patients, there was complete occlusion of the atrial septal defect by the end of the procedure. The residual shunts in these three patients were trivial (⩽ 1 mm) as assessed by transoesophageal echocardiography. Their defects had measured 20–21 mm on balloon sizing. In one patient this shunt has persisted at one year follow up. In the other two patients there was complete occlusion within 24 hours and six months, respectively. All patients were discharged within 24 hours, with the exception of one patient with multiple previous thromboembolic events, who was positive for factor V Leiden. She was maintained on a heparin infusion for two days, while anticoagulation with warfarin reached a therapeutic level.

Although there were no complications related to device implantation, five patients reported symptoms following device occlusion. In three of them, these had been present before intervention (angina in the presence of coronary artery disease, intermittent atrial flutter, and breathlessness on exertion, respectively). New symptoms following intervention were “glove and stocking” paraesthesia two days after intervention in one (patient 5), and unilateral segmental blurred vision five days after device implantation in the other (patient 8). In both patients the symptoms lasted less than 24 hours. Clinical review by a neurologist revealed no physical abnormalities, and the symptoms in patient 8 were attributed to atypical migraine. In a further patient (patient 6), there was an unexplained deterioration of pre-existing breathlessness with effort intolerance, which improved gradually over the course of three months. All five patients underwent transthoracic echocardiography which confirmed satisfactory device positioning, no obvious thrombus formation, and good ventricular systolic function. Patient 6 also underwent a ventilation:perfusion scan, which was normal.

Although all patients received intravenous heparin, 100 units/kg, during the intervention, their subsequent anticoagulant regimen varied according to the centre in which the intervention was performed, and whether they were perceived as being at high risk. All 20 patients were prescribed aspirin in an antithrombotic dose (300 mg once daily) for at least three months.

Of the 11 Royal Brompton Hospital patients, all received an overnight heparin infusion (one for two days), and three received long term warfarin. Two with large atrial septal defects, and three of the six patients with paradoxical emboli, subsequently received subcutaneous low molecular weight heparin (dalteparin sodium) at a dose of 2500–5000 units once daily, for 5–21 days. Two patients with paradoxical emboli also continued with their preprocedural regimen of warfarin. There have been no postprocedural embolic symptoms in any patients.

Discussion

Although this study reports the experience in a small group of selected patients, the initial results of ASO device occlusion of atrial septal defect in adults is encouraging. The device was successfully deployed in all 20 patients selected by preprocedural transoesophageal echocardiography. There were no procedural complications, and the mean procedure time of 1 hour should become even shorter with more experience. The physiological results were also satisfactory. There were trivial shunts at the end of the procedure (residual defect ⩽ 1 mm) in three patients, but this has persisted in only one (at one year follow up), and is of no haemodynamic significance.

It is too early to be sure whether symptomatic benefit will be achieved in either the group with large atrial septal defects and right heart dilatation, or smaller defects with early stroke. The rationale for surgical repair in the former group remains a subject of debate.2-5 A recent study has shown improved survival and functional status in adults following surgical repair of atrial septal defect, as compared to those managed medically, without closure.9 Our own surgical data suggest that clinical improvement can occur even in patients formerly describing themselves as asymptomatic.7

The closure of patent oval foramens or small atrial septal defects in patients with otherwise unexplained early stroke is even more contentious, and its utility even more difficult to prove. This study cannot in any way address this issue, other than showing that in highly selected patients, who otherwise would have been referred for surgery by their neurologists and physicians, adequate closure of the intracardiac lesion can be achieved using the ASO device. The procedure is performed in the same way as for larger atrial septal defects, with the exception of additional balloon sizing from right to left in case of a valve effect of the remnant of the flap of the oval fossa.

The postprocedural anticoagulant regimen reflected differing institutional preferences. In the Royal Brompton Hospital patients, because of concerns over extrinsic material implanted into the cardiac chambers, and enhanced thrombotic tendency in adults (most of whom were women), an overnight heparin infusion was administered in all cases, following a bolus during the procedure. In all patients who had been receiving warfarin before the procedure, because of paradoxical embolism, this was continued. In practice, those patients at the Royal Brompton Hospital over 35 years old received dalteparin sodium, warfarin, or both. Those under 35 years did not, with the exception of one patient with previous paradoxical embolism, and one patient with previously documented atrioventricular re-entry tachycardia.

The ASO device is particularly useful in adults with atrial septal defect. Defects of up to 34 mm can now be closed, and larger devices are in development. Not all defects are suitable or ideal, however. We excluded patients with multiple defects or fenestrations, preferring to use a non-self centreing device in some of these patients.10 A similar experience has recently been reported by others. Kaulitz et al recently reported successful closure of multiple and aneurysmal defects in six children and one adult using the Cardioseal device.11 A variant of the ASO device, with a larger external to central diameter ratio, is also being developed to cope better with such defects.

In summary, the ASO device can be used successfully to close selected oval fossa defects in adults, with minimal procedural morbidity and excellent early results.