Percutaneous Closure of Secundum Atrial Septal Defect in Adults — A Single Center Experience with the Amplatzer Septal Occluder


Cezar S. Staniloae, MD, Ziad El-Khally, MD, Reda Ibrahim, MD, Annie Dore, MD, Pierre de Guise, MD, Lise-Andrée Mercier, MD

Atrial septal defect (ASD) is one of the most common congenital cardiac anomalies encountered in adults.1 Surgical ASD closure leads to improved functional status, reduces the risk of progressive right-sided failure and prevents the development of severe pulmonary hypertension.2–4 More recently, percutaneous ASD closure using a number of different devices has emerged as a less invasive therapeutic alternative. Mounting evidence toward the safety and efficacy of percutaneous ASD closure in children has already been accumulated. However, no such data exist in an adult population coming from a single center and using a single device. The purpose of this paper is therefore to report our experience in this population using the Amplatzer Septal Occluder (ASO).


Patients. All patients scheduled for percutaneous ASD closure at the Montreal Heart Institute between May 1998 and April 2002 were prospectively included in our database. The patients were referred because they had symptoms (decreased exercise tolerance, arrhythmias or a history of thromboembolism), echocardiographic features of right ventricular dilatation or pulmonary hypertension, a significant left-to-right shunt (ratio of pulmonary to systemic blood flow larger than 1.5) or a combination of the above. Patients with ASD larger than 34 mm, as measured by transesophageal echocardiography (TEE), were directly referred for surgical closure.
Data collection. All patients underwent a clinical assessment of functional capacity according to the New York Heart Association (NYHA) functional class criteria. Transthoracic echocardiography (TTE) and TEE were obtained prior to the procedure. TTE was used to obtain qualitative and/or quantitative evaluation of right ventricle (RV) size and function, degree of tricuspid regurgitation and systolic pulmonary artery pressure. In a subgroup of 30 patients, quantitative assessment of RV size was reported as the short axis dimension, defined as the maximal dimension from the right septal surface to the free wall, perpendicular on the imaginary line that connects the RV apex with the mid-point of the tricuspid valve annulus. Data measured on TEE included ASD size estimation and measurement of the margins from the defect to: 1) the aortic valve (superior-anterior rim); 2) the coronary sinus (posterior rim); 3) the mitral valve (inferior-anterior rim); 4) the inferior vena cava (inferior-posterior rim); and 5) the superior vena cava (superior-posterior rim). The hemodynamic data obtained at the time of the procedure included: pulmonary artery pressure (PAP); RV pressure; right atrial (RA) pressure; right ventricular diastolic pressure; and shunt (Qp:Qs) measurements. The stretched diameter of the defect was measured both by fluoroscopy and TEE. The patients with a shunt less than 1.5 had the following indications for closure of the ASD: otherwise unexplained dyspnea on exertion (7 patients); RV enlargement (4 patients); dyspnea associated with enlargement of the right cavities (4 patients); and history of a paradoxical embolus (3 patients).
Patients were reevaluated clinically and by TTE at 1 month, 3–6 months, 1, 2 and 3 years after the procedure. In a subset of patients, cardiovascular exercise test and maximal oxygen uptake capacity (VO2 max) were obtained prior to the procedure and 6 months later.
Interventional procedure. The procedure was performed under general anesthesia with TEE guidance. Single-plane fluoroscopy was used. Prior to the procedure, the patients received antibiotic prophylaxis and a bolus of heparin (100 IU/ kg) followed by weight-adjusted heparin for 24 hours. Aspirin was started after the procedure and continued for 18 months. The ASO was implanted using the previously described technique.5,6 TEE evaluation was performed at the end of the procedure to assess device position, mitral and tricuspid valve function, and pericardial space, and to evaluate for eventual residual shunting.
Statistical analysis. Data are reported as mean values and ranges. The t-test was used for paired or unpaired comparisons. Pearson correlation coefficients are reported where appropriate. The differences were considered significant for p < 0.05.


Patients and procedure. One hundred and six (90.6%) of the consecutive 117 patients (28 men, 89 women) had successful deployment of an ASO device. In 2 cases, the device was deployed using a special delivery catheter (the CHECK-FLO® II Hausdorf-Lock Atrial Introducer Set; Cook Cardiology, Bloomington, Indiana) after a first unsuccessful attempt using a conventional delivery catheter. The mean age was 50 ± 16 years (range, 19–79 years; median, 49 years). The left-to-right shunt (Qp:Qs) was calculated in 107 patients. Eighty-four percent of these patients had Qp:Qs >= 1.5. Mean procedure time was 44 minutes (range, 20–112 minutes) and mean fluoroscopy time was 13 minutes (range, 5–39 minutes). There was no residual shunting at the end of the procedure in 75% of the patients. A small defect (< 5 mm) resulting in a minimal left-to-right shunt was present in 25 patients (23.5%). Two patients had a residual defect larger than 5 mm.
Echocardiographic and catheterization data. Echocardiographic and catheterization data are presented in Table 1. The defect diameter on TEE increased by an average of 5.7 mm (from a mean of 18.3–24.0 mm) when the ASD was stretched with the sizing balloon. The largest increase in diameter with the sizing balloon was 14 mm. There was a strong correlation between stretched diameters measured by TEE and by angiography (r = 0.952).
The aortic rim was absent in 22 patients and it was less than 5 mm in 16 patients. The posterior rim was absent in 1 patient and < 5 mm in 3 patients. All other margins were larger than 5 mm in all patients.
When compared with the patients who had successful procedure, the patients in whom the procedure failed had larger defects (mean, 25.0 mm versus 18.3 mm; p < 0.001), larger RV diameter (44.1 mm versus 38.2 mm; p = 0.03) and more significant left-to-right shunts (Qp:Qs, 3.5 versus 2.1; p < 0.001). The absence of an aortic rim did not interfere with the success of the procedure.
Follow-up data. All patients were alive after an average follow-up of 19 months (range, 3–40 months). TTE performed at 1-month follow-up revealed a persistent minimal residual shunt in only 9 (8.5%) out of the 25 (23.5%) patients with small shunts identified immediately after the procedure. The 2 larger shunts (> 5 mm) were still present at 1 month. Color Doppler performed at 1-year follow-up showed a residual shunt in only 1 patient. This patient with persistent shunt had a fenestrated septum and the defect occlusion was known to be incomplete. He was referred for device removal and surgical ASD closure 2 years later because of persistence of symptoms, shunting and enlargement of the right heart cavities. Analysis of the excised device showed complete endothelialization of both atrial buttons.
Among the patients who presented with NYHA functional class 2 or higher, the symptoms improved by at least 1 class in all but 1 patient (Table 2). The improvement was seen as early as 3 months after the procedure. A subset of 30 patients who were in NYHA class I or II underwent baseline and 6-month follow-up exercise testing and maximal oxygen uptake (VO2max) evaluation. As we have reported recently,7 in this selected population, VO2max increased from 23.8 ± 6.7 ml/kg/minute to 27.3 ± 7.0 ml/kg/minute (p < 0.0001) at 6 months.7
Procedural failure. In 13 patients, the ASO could not be successfully deployed at first attempt; in 6 cases, the device could not be positioned parallel to the interatrial septum, in 3 others we were unable to secure a stable device position and in the last 4 cases the stretched defect diameter was larger than the size of the device available at the time of the intervention (mean, 38 mm; range, 36–40 mm).
Reinterventions. Four patients were scheduled for a second attempt at ASD closure after the initial procedure failed because of inability to position the device parallel to the interatrial septum. The stretched ASD diameters in these 4 cases were 28, 28, 30 and 32 mm. Two of these defects were closed with a Hausdorf delivery catheter (Cook Cardiology), which has a 60° angulation. In 1 case, the approach to the interatrial septum was still suboptimal and the device was not deployed. In the fourth case, the stretched diameter of the defect increased from 32 to 34 mm following the initial procedure, presumably because of defect enlargement at the time of the previous sizing balloon inflation, and we were therefore unable to complete the procedure since the Hausdorf catheter delivers devices up to only 32 mm.
Complications. Major complications included 1 patient who had an episode of air embolism in the coronary circulation at the end of the procedure. The patient had to be resuscitated and reintubated, and an intraaortic balloon pump was installed. He recovered completely within 24 hours, without persistent electrocardiographic changes or enzyme elevation. Minor complications included 2 patients who had signs of interstitial edema on the chest radiography routinely performed 12–24 hours after the procedure. Both patients were asymptomatic, and the radiographic signs disappeared without treatment on follow-up chest x-ray. One patient developed pleuritic chest pain immediately after the procedure. He had a normal ventilation-perfusion scan and no pericardial or pleural effusion was identified. A small pericardial fluid collection was seen immediately after the procedure in 2 patients, but disappeared completely at 1-week follow-up. One prosthesis showed unusual protrusion into the left atrium (“mushrooming”) and proximity to the anterior leaflet of the mitral valve after its deployment. Nevertheless, at follow-up, the device was well positioned and the mitral valve was entirely competent. Access complications included 2 hematomas and 2 arteriovenous fistulas. One fistula required surgical closure. Eight patients complained of migraine headache 2–4 weeks after the procedure. All of them were asymptomatic at the 6-month follow-up.


Post new comment

  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.

More information about formatting options

Enter the characters shown in the image.