Transcatheter device closure (TCDC) and intraoperative device closure (IODC) have emerged as minimally invasive methods in the treatment of secundum atrial septal defects (ASDs), but the long-term safety and efficacy remains uncertain for the large ASDs.
TCDC may be as safe and efficacious as IODC for closure of large ASDs in terms of long-term clinical outcomes.
Ninety-two patients who had ASDs with a defect diameter of ≥30 mm were included in this study. The patients received either TCDC (n = 42) or IODC (n = 50). An Amplatzer septal occluder was used in both groups. The dumbbell-like device deploying technique was introduced in the TCDC group. Physical exams, electrocardiography, and echocardiography were performed preprocedurally and postprocedurally at the index follow-up visits.
The procedural immediate success rate was 97.6% for TCDC and 98.0% for IODC (P = 0.328). The rate of periprocedural complications was 9.5% for TCDC and 28.0% for IODC (P = 0.026). The mean hospital stay was 7.5 ± 2.7 days for TCDC and 11.9 ± 3.8 days for IODC (P < 0.001). For the mean follow-up of 5.4 ± 0.5 years, there were no cardiac deaths and late complications in either group. No significant residual shunts were documented, and symptoms were significantly improved in both groups. Right and left ventricular diameter, pulmonary artery diameter, and pulmonary systolic pressure were all significantly decreased in both groups (P < 0.05).
The present study confirmed the long-term safety and efficacy for closing a large ASD either by TCDC or IODC. Either of them could become an effective alternative to the surgery for large ASD closure.
The authors have no funding, financial relationships, or conflicts of interest to disclose.
Atrial septal defects (ASDs) are 1 of the most common congenital heart diseases. Surgical suture of the defect has remained the gold standard therapy for the past 5 decades.1 Recently, transcatheter device closure (TCDC) and intraoperative device closure (IODC) have emerged as new minimally invasive alternatives to the classic surgery,2–4 and the Amplatzer septal occluder (ASO) has became the most popular device systems.5–11 However, device closure of a large ASD with an ASO remains technically challenging,12–16 and the long-term clinical outcomes have not yet been validated.
The aim of this study was to observe the long-term clinical outcomes of using and ASO for closure of a large ASD and to compare the safety and efficacy between TCDC and IODC, and to especially address periprocedural and late complications.
The inclusion criteria were: (1) patient's age of at least 14 years, (2) patients with a large isolated secundum ASD (the minimal defect diameter ≥30 mm as measured preprocedurally by echocardiography),17 and (3) patients with ASD anatomically adequate for device therapies (eg, 3–5-mm rims except aortic side to enable secure fixation of the device). Exclusion criteria were: (1) patients with a primum ASD, (2) patients with any other cardiac comorbidity that needed to be corrected surgically, (3) patients with right-to-left shunting due to severe pulmonary hypertension, and (4) patients with any other contraindications to device therapies.
From January 2003 to December 2009, 92 eligible patients were included and were assigned to receive either TCDC (the TCDC group, n = 42) or IODC (the IODC group, n = 50) based on patients' and/or their parents' preference.
The study protocol was reviewed and approved by the institutional board and the ethics committee of Union Hospital, Fujian Medical University and conformed to the principles outlined in the Declaration of Helsinki. Written informed consent was obtained from the patients or their parents.
TCDC and IODC Procedure
An Amplatzer atrial septal occluder (Xianjian Medical Apparatus Co. Ltd, Shenzhen, China) was used for ASD closure. The procedure was performed under sedation and local anesthesia. The ASO was selected to be 4 to 6-mm larger than the largest diameter of the ASD measured by transthoracic echocardiography (TTE). Under TTE guidance, a special (so-called dumbbell-like) ASO deploying technique was used for the procedure. The delivery catheter (14F) was positioned inside the left upper pulmonary vein. The ASO was advanced up to the ostium of the left upper pulmonary vein, and then the catheter was pulled back slowly to let the distal disc wedge itself into the ostium of the pulmonary vein, with the waist positioned approximately in the defect and the proximal disc in the right atrium so that the whole ASO became dumbbell-like. By lightly pulling and shaking the delivery cable, the ASO quickly recovered its memorized shape and repairedd the defect. TTE was then examined to confirm the device position and to ensure that there was no encroachment of the adjacent cardiac structures. Once its position was optimal, the device was released and TTE was performed again to assess the result of closure.
The technique of IODC has been described in previous studies.18,19 Briefly, a 2.5 to 3 cm incision was made in the right anterior fourth intercostal space of the right sternal border after general anesthesia. Right atrial puncture was made within a purse string suture placed on the right atrium and the sheath was inserted. Under TTE guidance, it was advanced through the ASD into the left atrium. The left atrial disc was extruded first by advancing the delivery cable. The waist and the right disc were deployed subsequently. After release, TTE was performed for further demonstration of the device position and residual shunts. The sheath was then withdrawn with the purse-string snugly tied. The minithoracotomy was closed without a drainage tube.
A dose of an appropriate antibiotic was given during the procedure, and 2 doses at 8-hour intervals were given after. Routinely, 200 mg of aspirin were initiated 48 hours before closure, and 100 to 200 mg per day continued for 6 months.
Clinical Follow-up and Protocol
All patients underwent routine pre- and periprocedural evaluations during hospitalization. In addition to demographic data and clinical characteristics, periprocedural evaluation was addressed to document immediate technical success, recovery time (return to normal activities) and hospital stay time, and to record acute complications including all-cause death, cardiac perforation, severe infection, pleural effusion, major bleeding, thrombosis and embolism, hemolysis, cardiac arrhythmia, residual shunt, and ASO dislodge. The immediate successful ASD closure was defined as absence of significant residual shunt assessed by color Doppler flow imaging (CDFI) immediately after the procedure.
Postprocedural follow-up was scheduled at 1, 3, 6, and 12 months in the first year and yearly thereafter. A clinic visit with history taking, physical examination, chest radiography, electrocardiography, and echocardiography was required. Clinical evaluation addressed the procedure-related chronic complications such as cardiac death and infective endocarditis.
Two-dimensional echocardiography with continuous-wave Doppler imaging and CDFI was performed pre-, peri-, and postprocedurally for each follow-up visit by using a commercially available echocardiographic system (Vivid-7; GE Medical Systems, Milwaukee, WI). Pulmonary hypertension (PHT) was defined as pulmonary artery systolic pressure (PASP) ≥30 mm Hg, and persistent residual shunt was defined as trivial (shunt width ≤1 mm), small (shunt width 1–2 mm), moderate (shunt width 2–4 mm), or large (shunt width ≥4 mm).20
SPSS 13.0 (SPSS Inc., Chicago, IL) was used for the analyses. Continuous variables are expressed as mean ± standard deviation. Differences between the 2 groups were analyzed using independent-samples t test for continuous variables and χ2 test for categorical variables. Paired-samples t test was used to compare differences of variables pre- and postprocedurally. A P value <0.05 was considered to be statistically significant.
As summarized in Table 1, both groups were similar in the demographic data and preoperative clinical characteristics. The TCDC group had PASP range from 25 to 85 mm Hg with mild-moderate PHT in 35 and moderate-severe PHT in 7 patients. The IODC group had a PASP range from 21 to 98 mm Hg with mild-moderate PHT in 41 patients and moderate-severe PHT in 9 patients.
Table 1. Baseline Characteristics Between the Two Groups
TCDC Group (n = 42)
IODC Group (n = 50)
Abbreviations: ASD, atrial septal defect; IODC, intraoperative device closure; LAD, left atrial diameter; LVEDD, left ventricular end-diastolic diameter; LVEF, left ventricular ejection fraction; LVESD, left ventricular end-systolic diameter; PAD, pulmonary artery diameter; PASP, pulmonary artery systolic pressure; RVEDD, right ventricular end-diastolic diameter; TCDC, transcatheter device closure.
Data are presented as mean ± standard deviation. P > 0.05 for all comparisons.
33.6 ± 16.4
38.1 ± 18.3
54.6 ± 11.1
51.4 ± 13.6
160.8 ± 9.8
156.1 ± 18.8
ASD size (mm)
31.5 ± 3.5
30.1 ± 2.9
Device size (mm)
36.2 ± 3.4
36.1 ± 3.6
30.7 ± 5.9
30.3 ± 3.9
38.9 ± 2.4
36.4 ± 4.5
24.4 ± 2.6
23.4 ± 2.3
67.6 ± 5.4
63.4 ± 10.0
40.5 ± 6.1
39.5 ± 4.9
25.5 ± 2.9
27.1 ± 5.8
PASP (mm Hg)
52.4 ± 16.6
53.2 ± 19.1
Immediate and Short-Term Outcomes
The immediate and short-term outcomes are listed in Table 2. The preprocedural defects measured echocardiographically were 31.5 ± 3.5 mm in the TCDC group and 30.1 ± 2.9 mm in the IODC group (P = 0.185), and the implanted occluders were 36.2 ± 3.4 mm and 36.1 ± 3.6 mm, respectively (P = 0.930). All patients for IODC required general anesthesia with intubation and ventilation, whereas no TCDC patient had such a requirement (P < 0.001). Compared with the IODC group, the recovery time (P < 0.001) and hospital stay (P < 0.001) were significantly shorter in the TCDC group.
Table 2. The Immediate and Short-Term Clinical Outcomes in the Two Groups
The immediate success for ASO implantation was 97.6% for TCDC and 98.0% for IODC (P = 0.328), with 1 failed case in each group. Among them, 1 had dislodgement of the released occluder in the TCDC group, and the ASO embolized in the right ventricle and was irretrievable using a snare, therefore emergency surgery was required for retrieval with concomitant ASD closure. The other in the IODC group experienced unsuccessful closure due to lack of appropriate ASO for the huge (40 mm) ASD and was switched to surgical repair. Although not significant, a trivial residual shunt was documented immediately after the procedure in 4 patients in the IODC group and 3 patients in the TCDC group (9.5% vs 6.0%, P = 1.000).
The rate of procedure-related complications was 9.5% for TCDC and 28.0% for IODC (P = 0.026). There was no death or tamponade due to adjacent cardiac structure perforation, infective endocarditis, thromboembolism, or hemolysis in both groups. Hydrothorax was noted in 10 patients in the IODC group requiring drainage in 2 patients, whereas no hydrothorax was observed in the TCDC group (P < 0.001). Periprocedural pneumonia occurred in 6 patients in the IODC group, but no pneumonia was documented in the TCDC group (P = 0.030).
Temporary cardiac arrhythmias, but no permanent arrhythmias, were observed in 5 patients periprocedurally in the TCDC group. Among them, a first-degree atrioventricular block (AVB), frequent ventricular premature beats, and atrial flutter were noted in 1 patient respectively. Atrial fibrillation (AF) was observed in 2 patients. Most of them were temporary and did not require therapy; only 1 persistent AF needed be converted by amiodarone. In the IODC group, cardiac arrhythmias were documented in 8 patients periprocedurally, including AF in 3 patients, 1 atrial tachycardia, 1 atrial flutter, 2 frequent ventricular premature beats, and 1 second-degree AVB. Similarly, only 1 persistent AF needed antiarrhythmic agents.
There were 41 patients (97.6%) in the TCDC group and 47 patients (94.0%) in the IODC group who received a mean 5.4 ± 0.5 years' follow-up. The number of patients lost to follow-up in the TCDC group and the IODC group were 1 and 3, respectively. During the follow-up period for both groups, effective closure was achieved in 100% patients; there was no cardiac death, infective endocarditis, pericarditis, pericardial constriction, or strokes observed clinically. No occluder dislodged, and there was no deformation or rupture, or erosion of the adjacent structures such as the atrial wall and ascending aorta by the device, and no obstruction of the atrioventricular valves, the caval veins, the pulmonary veins, and the coronary sinus by the ASO were documented by TTE.
All symptomatic patients improved significantly in the TCDC group, whereas 3 patients with severe PHT before IODC complained of exertional dyspnea or fatigue in the IODC group.
As shown in Figure 1 and Figure 2, at a mean 5.4 years follow-up, echocardiography-measured structure and hemodynamic parameters were significantly improved as compared with those at baseline. Right ventricular end-diastolic diameter decreased from 40.5 ± 6.1 mm to 20.0 ± 4.6 mm in the TCDC group (P < 0.001) and 39.5 ± 4.9 mm to 26.5 ± 8.5 mm in the IODC group (P < 0.001). Left ventricular end-diastolic diameter (LVEDD) increased from 38.9 ± 2.4 mm to 46.2 ± 1.8 mm in the TCDC group (P < 0.001) and 36.4 ± 4.5 mm to 43.9 ± 2.7 mm in the IODC group (P = 0.016). Left ventricular end-systolic diameter (LVESD) increased from 24.4 ± 2.6 mm to 28.5 ± 2.3 mm in the TCDC group (P = 0.007) and 23.4 ± 2.3 mm to 30.3 ± 6.1 mm in the IODC group (P = 0.013). Left ventricular ejection fraction (LVEF) increased from 67.6 ± 5.4% to 71.8 ± 7.1% in the TCDC group (P = 0.072) and 63.4 ± 10.0% to 65.8 ± 18.2% in the IODC group (P = 0.879), pulmonary artery diameter decreased from 25.5 ± 2.9 mm to 21.4 ± 2.9 mm in the TCDC group (P < 0.001) and 27.1 ± 5.8 mm to 21.0 ± 2.8 mm in the IODC group (P = 0.007). PASP declined from 52.4 ± 16.6 mm Hg to 35.1 ± 7.5 mm Hg in the TCDC group (P = 0.008) and 53.2 ± 19.1 mm Hg to 31.7 ± 17.6 mmHg in the IODC group (P < 0.001).
The present study demonstrates that closing large ASDs using ASO either by TCDC or IODC is not only procedurally feasible but also safe and efficacious in the long-term follow-up. To the best of our knowledge, for device closure of large ASDs, this is the first study to provide long-term outcomes of up to 5 years and to compare TCDC and IODC in the treatment of such challenging patients.
Immediate Success and Procedural Complications:
As for device closure of the small to medium size ASDs, the immediate technical success was high with few procedural complications.21,22 However, closing large ASDs in particular remains a technical challenge.13,23 For our patient cohort, all occluders implanted were very big (36.2 ± 3.4 mm for TCDC and 36.1 ± 3.6 mm for IODC). In the treatment of such challenging patients, both procedures had only 1 failed case in each group. The immediate success rate for ASO implantation was similar for both approaches (97.6% for TCDC vs 98.0% for IODC), with no significant residual shunt in all patients. Neither death nor life-threatening complications occurred in either group. Therefore, both approaches for closing large ASDs are technically feasible.
Safety and Efficacy:
The long-term outcomes, especially the safety of device closure of a very large ASD, is a major concern, but there are no data available specifically aimed at the late or very late device-related major adverse cardiac events. In the general population, for those who underwent device closure of ASDs, early or late device-related complications did occur; thrombosis and thromboembolism have been reported with an overall incidence of 2% to 27%,24–26 and may be more frequently encountered with big occluders because of delayed or incomplete endothelialization. Late or very late cardiac perforation, with or without pericardial effusion or tamponade, was reported to occur predominantly in the anterosuperior atrial wall and/or adjacent aorta,8,27 and erosion of atrioventricular valves, with or without regurgitations, was rare but has been documented in the literature.28 Hemodynamic influences, such as obstruction of adjacent venous drainage or functional interference of atrioventricular valves, are all theoretically possible especially for device closure of large ASDs. Recent registry data suggested device oversizing and deficient anterosuperior rims as the risk factors for adjacent cardiac structure erosion or perforation.29 Fortunately, our data showed no such late or very late complications.
Long-term efficacy has been validated for the general population but not for patients with large ASDs who received device closure.30–33 The main expectation is to achieve completely anatomic closure of ASDs. However, device malposition or breach in structural integrity and residual shunting could not be completely avoided, particularly for closing large ASDs. Our echocardiographic data revealed that neither device malposition/breach nor significant residual shunting was documented in follow-up of up to 5 years.
After ASD closure, apparent clinical improvement and an increase in effort capacity were observed in most patients, which might be the result of remodeling of the right ventricle and the decline in PHT, but not the improvement of left ventricular function. Although both enlargement of LVESD and LVEDD were observed after occlusion of the defect; LVEF, however, did not increase in our series. Such findings were similar to the study by Teo et al34 but different from the studies by Wilson et al28 and Yew and Wilson.30 Differences in race and study design may in part explain these inconsistent data.
Above all, our data demonstrated excellent long-term safety and efficacy for closing large ASDs either by the TCDC or IODC approach.
Technical Comparison of 2 Approaches
As described previously in our and other studies,18,19 both TCDC and IODC had similar long-term outcomes in terms of safety and efficacy. However, it is noteworthy to mention that the 2 procedures are technically different. For TCDC, the difficulty is that the self-centering property of the ASO no longer works in the setting of very large defects, which may easily lead to device malposition, fixation inability, or even a disastrous device dislodgement event. Such situations are more frequently encountered by TCDC and less by IODC, because the delivery catheter of the latter approach is more perpendicular to the septum, facilitating position of the occluder. Moreover, additional sutures may be necessary for prevention of device dislodgment in case of uncertainty of device fixation. In this regard, IODC seems to be more reliable for closing very large ASDs.4,35 However, with the introduction of the dumbbell-like device deploying technique, TCDC in this study was able to achieve a high immediate success rate of 97.6%, which is not different than that of 98% for IODC.
It is noteworthy to mention that IODC, a new minimally invasive technique, is inherently more traumatic and has much more acute procedure-related complications than TCDC, as documented by overall an incidence of 28.0% for IODC vs 9.5% for TCDC in the present study, regardless of no life-threatening events.
In comparison, both approaches of TCDC and IODC are technically feasible for closing very large ASDs, with the possible superiority of IODC for some particularly difficult patients.
Despite the encouraging results, the study had several limitations. First, it was a single-center study performed on selected cases. Second, the present study was retrospective with a relatively small population. Therefore, our study needs to be further validated by a multicenter, prospective, randomized trial.
Closing large ASDs via a TCDC or IODC approach is technically feasible and associated with excellent long-term safety and efficacy. Either of those approaches could become an effective alternative to surgery for closure of large ASDs.
The authors thank Zi-Wen Zhao, MD, PhD, for editorial support. The authors are also grateful to Ling-Zhen Wu, MD, PhD, and Zhen-Dong Cheng, MD, for assistance with the study.