Introduction
Atrial septal defect is a common CHD in adults, with a prevalence of 0.88 per 1000 individuals.Reference Marelli, Ionescu-Ittu, Mackie, Guo, Dendukuri and Kaouache1 Transcatheter closure is now the preferred approach for secundum atrial septal defect due to its high efficacy, minimal invasiveness, quicker recovery, better cosmetic results, lower morbidity, and low complication rates. Over the past three decades, growing experience and device innovation have enabled closure of increasingly larger defects. The feasibility of transcatheter closure in large atrial septal defects has been reinforced by the availability of larger occluders, progressive improvements in device design, advanced imaging guidance, and refined deployment techniques—often with various procedural assistances—enabling safe and stable implantation even in anatomically challenging cases. Initially limited to 40–42 mm occluders, the availability of even larger devices from select manufacturers has broadened indications for transcatheter closure in complex atrial septal defects.Reference Romanelli, Harper and Mottram2–Reference Lopez, Dalvi and Balzer5 This study evaluates transcatheter closure of large secundum atrial septal defects in adults using septal occluders measuring 40 mm or more.
Materials and methods
Study design and patient population
This retrospective multicentric study included patients from two tertiary-care centres in Eastern India—Manipal Hospitals, EM Bypass (formerly Medica Superspeciality Hospital, Kolkata), and Vivekananda Institute of Medical Sciences (Ramakrishna Mission Seva Pratishthan), Kolkata. It analysed cases of large secundum atrial septal defects with significant left-to-right shunts closed using 40–50 mm occluders between June 2015 and May 2025.
Aims and objective
The aim of this study was to evaluate the efficacy and procedural challenges associated with large atrial septal occluders (40–50 mm) and to assess their mid-term outcomes. The mean follow-up duration was 29.3 ± 21.6 months, as detailed in the Results section.
The procedure
All patients underwent clinical evaluation, chest X-ray, electrocardiogram, and echocardiography. Both transthoracic echocardiography and transesophageal echocardiography were performed to assess defect anatomy and suitability for transcatheter closure. Margins ≥5 mm were considered adequate, 3–5 mm as deficient.Reference Sivakumar, Viswambaran and Bhattacharjya4 Cardiac catheterisation was initiated under local anaesthesia with conscious sedation, transitioning to general anaesthesia when peri-procedural transesophageal echocardiography was required. Dual femoral venous and single femoral arterial access were obtained in all.
Balloon sizing was not performed, per unit protocol, favouring echocardiographic measurements and rim assessment. Haemodynamic assessment and coronary angiography were routinely done. Pulmonary artery pressure and left ventricular end-diastolic pressure were measured in all patients. Pulmonary hypertension was classified by pulmonary artery systolic pressure as mild (35–50 mmHg), moderate (50–70 mmHg), or severe (>70 mmHg).Reference Faqih, Noto-Kadou-Kaza and Abouamrane6 Pulmonary blood flow and resistance were assessed only in moderate-to-severe pulmonary hypertension cases. Left ventricular end diastolic pressure ≥15 mmHg was considered elevated. In suspected left ventricular diastolic dysfunction, test occlusion with balloon/device was followed by repeat left ventricular end-diastolic pressure after 15 minutes; a rise >5 mmHg was deemed significant.Reference Singhi, Mohapatra and Kumar7
Devices used included: Amplatzer™ septal occluder (Abbott, Plymouth, Minnesota)—40 mm; Lifetech septal occluder (Lifetech Scientific, Shenzhen, China)—40–46 mm; Cocoon atrial septal occluder (Vascular Innovations Co. Ltd., Thailand) – 40–42 mm; and Occlunix septal occluder (Rivarp Medical, Bangalore, India)—40–50 mm. Occluder devices exceeding 42 mm and the fenestrated variants from Lifetech Scientific were custom-made by the manufacturer upon specific requests for individual patients, while other device sizes were readily available from the standard product line. Follow-up was conducted on post-procedure day 1, then at 1, 6, and 12 months, and annually thereafter.
Ethical clearance
This retrospective study adhered to the approved protocol of the institutional Clinical Ethics and Research Committee (CERC/2021/Jun/iv). Patient confidentiality was ensured by anonymizing all data used in images and analyses. As the study utilised anonymized data, the ethics committee waived the requirement for individual patient consent. All patients provided informed consent for the procedure and subsequent data collection and analysis, as per the general informed consent protocol of the hospital.
Statistical method
Continuous variables were presented as mean ± standard deviation, and categorical variables as frequencies and percentages. Group comparisons for continuous variables were performed using the independent t-test, assuming a normal distribution. For categorical variables, the Chi-square test or Fisher’s exact test (when expected cell counts were <5) was applied. All statistical tests were two-sided, with a p-value <0.05 considered statistically significant. Analyses were conducted using R software (version 4.3.3).
Results
This study included 62 patients undergoing atrial septal defect closure with large occluders, categorised into Group A (40–42 mm occluder) and Group B (44–50 mm occluder). The classification into two groups was based on device size and availability. Initially, atrial septal occluders were available up to 40 mm (Amplatzer), with 42 mm devices later introduced by Cocoon and Lifetech. Sizes above 42 mm were mostly custom-made, until newer brands such as Rivarp offered devices up to 50 mm. In our practice, occluders measuring 40–42 mm were considered standard large devices (Group A), while those sized 44–50 mm were regarded as very large (Group B). The mean age of the patients was 42.07 years, with 71% females. Mean atrial septal defect diameter was 33.03 mm, and the mean thick-to-thick dimension was 38.71 mm (Table 1).
Demographic parameters and patients caricaturists
SD = standard deviation; mm = millimetre; PA = pulmonary artery; LVEDP = left ventricular end-diastolic pressure.
Device closure was successful in 58 patients (93.5%) (Figure 1a–f). In four patients (6.5%), closure failed due to anatomical challenges—floppy/deficient posteroinferior margins and aneurysmal septum. Three underwent elective retrieval for instability or significant residual shunt and were referred for surgery. One patient (1.6%) had intra-procedural device embolisation (42 mm) requiring emergency surgical retrieval and closure (Figure 2a–d).
(a) TEE with colour Doppler shows a large secundum ASD (orange arrow) with thin, deficient posteroinferior rim. (b) Fluoroscopy shows balloon-assisted (black arrow), TEE-guided deployment. (c) Device in stable position. (d) TEE confirms complete closure with a bulky device (red arrow). (e) Apical four-chamber view shows 46 mm occluder with bulky profile (blue arrow). (f) Follow-up echo shows good device profile (green arrow) and right ventricular remodelling.
(a, b) Apical four-chamber and subcostal views show a large ASD with deficient posteroinferior rim (red arrow). (c, d) Intraoperative images demonstrate retrieval of embolised occluder (green arrow) from the right ventricle.
In Group A, 23 patients (67.6%) received 40 mm and 11 (32.4%) received 42 mm occluders. In Group B, 14 (50%) received 44 mm, eight (28.6%) 46 mm, four (14.3%) 48 mm, and two (7.1%) 50 mm devices.
Procedural assistance was needed in 58 patients (93.5%). Balloon-assisted closure was performed in 34 patients (54.8%) using a 33-mm Equalizer balloon (Boston Scientific) (Figure 1b); device-assisted closureReference Singhi, Mohapatra and De8 in 18 (29%) (Figure 3b); dilator assistance in four (6.5%), and super-stiff wire with catheter support in two (3.2%). Only four patients (6.5%) underwent unassisted closure. Devices were delivered using 14–16F Mullins sheaths or equivalents; limited 16F availability later in the study necessitated 14F systems.
(a) TEE with colour Doppler showing a large ASD (white arrow) with deficient posteroinferior rim. (b, c) Fluoroscopic views during device-assisted (green arrow) ASD device closure (black arrow) ; LAO cranial view reveals a bulky but well-seated device (blue arrow). (d) Post-procedure TEE confirms successful closure (yellow arrow). (e) Parasternal short-axis view shows vegetations on the device’s inferior margin. (f) Abdominal MRI reveals bilateral tubo-ovarian abscess, (left side large -green arrow, right side orange arrow). (g, h) Intraoperative images show explanted device with vegetations. (i) Follow-up echo shows complete surgical patch closure.
The choice of procedural assistance was influenced by operator experience, hardware availability, and procedural era—device closure technique becoming available post-2023. Among balloon-assisted cases, seven patients (11.3%) required additional manoeuvres (Figure 4d), while only one device closure technique-assisted case (1.6%) needed subsequent balloon support. Successful atrial septal defect device closure in a single attempt was achieved in 62.9% of the overall cohort—64.7% in Group A and 57.1% in Group B. All four patients who underwent unassisted device closure had successful outcomes. In retrospect, these cases had more aneurysmal rather than deficient margins, which may have contributed to the procedural success without additional assistance. Multiple attempts (more than two) were required in 21% of patients, with 14.7% in Group A and 28.6% in Group B, indicating a trend towards increased attempts with larger occluders. Outcomes are summarised in Table 2. No significant differences were observed between Groups A and B across evaluated parameters. One patient required device upsizing due to residual shunt.
(a, b) Transthoracic echocardiography showing a large ASD (white arrow) with floppy posteroinferior rim. (c) Fluoroscopy during balloon-assisted (red arrow) device deployment in left pulmonary vein approach (black arrow). (d) TTE showing a well-seated device across the defect. (e, f) Fluoroscopy confirming stable post-deployment position (green arrow). (g) ECG showing atrial flutter. (h) Brain CT revealing intracerebral haemorrhage (red arrow).
Anatomy of the atrial septal defect, procedure, and occluder details
ASD = atrial septal defect; SD = standard deviation; mm = millimetre; DAD = device assisted device closure; F = french.
The cohort’s mean peak systolic and mean pulmonary artery pressures were 35.2 mmHg and 22.33 mmHg, respectively. Mean left ventricular end-diastolic pressure was 12.59 mmHg; two patients had evidence of diastolic dysfunction.
Fenestrated atrial septal defect closure was performed in four patients (6.5%). One patient with atrial septal defect and severe pulmonary stenosis first underwent balloon pulmonary valvotomy, followed by fenestrated device closure. Two patients had significant left ventricular diastolic dysfunction and one had severe pulmonary hypertension with elevated pulmonary vascular resistance (8.04 Wood units·mReference Romanelli, Harper and Mottram2) and increased pulmonary blood flow (Qp:Qs ∼ 1.8). A fenestrated occluder was implanted to permit partial decompression and assess reversibility of pulmonary hypertension. Post-deployment, pulmonary artery pressure declined significantly, with persistent left-to-right shunting through the fenestration, suggesting a flow-related component. On follow-up, the patient was weaned off pulmonary vasodilator therapy and remains clinically stable with mild residual pulmonary hypertension (Figure 6a–g). Among these, one patient received a handmade fenestration in a 40 mm Amplatzer occluder, while the remaining three were treated with custom-made 40 mm Lifetech devices. In one patient with large atrial septal defect, diastolic dysfunction, moderate-to-severe pulmonary hypertension, and obstructive lung disease, pulmonary therapy improved symptoms. Baseline left ventricular end-diastolic pressure was 16 mmHg, rising to 20 mmHg on balloon occlusion. A custom-made 40 mm fenestrated device was implanted, reducing pulmonary artery pressure; left ventricular end-diastolic pressure fell to 18 mmHg but remained elevated. Fenestration showed left-to-right shunting with a mean gradient of 5 mmHg (Figure 5a–h). Follow-up revealed mild pulmonary hypertension in these patients. Custom-made fenestrations remained patent (Figure 6f), while the handmade fenestration was closed at one year (Table 2). None of the patients demonstrated a rise in left ventricular end-diastolic pressure > 5 mmHg following device deployment.
(a, b) TEE showing a large ASD (white arrow) with thin, floppy posterior rim and bidirectional shunt (arrow). (c) RV systolic pressure estimated at 49 mmHg + RA pressure via TR jet. (d) HRCT chest showing obstructive pulmonary features. (e) ASD closed with a fenestrated occluder (green arrow). (f) Post-procedure TEE showing a stable device with left-to-right shunting through fenestration (yellow arrow). (g) 3D TEE highlighting fenestration (red arrow). (h) CW Doppler showing mean gradient of 5 mmHg across fenestration.
(a) Chest X-ray (AP view) showing cardiomegaly with dilated pulmonary arteries and prominent peripheral vasculature (b) Apical four-chamber view showing RV dilatation and contrast in the left atrium suggesting transient right-to-left shunt (green arrow). (c) TEE showing a large ASD (white arrow). (d) High-velocity tricuspid regurgitation (TR) jet indicating elevated RV systolic pressure. (e) Fluoroscopy showing deployment of fenestrated occluder (red arrow). (f) Follow-up echo demonstrating stable device with left-to-right flow through fenestration (green arrow). (g) Doppler showing reduced RV systolic pressure via TR jet.
One patient had a small inferior defect adjacent to the main atrial septal defect, separated by a thin tissue strand. A 46 mm occluder was deployed with balloon assistance under fluoroscopy. Despite apparent stability, significant residual flow persisted at the inferior margin. Transesophageal echocardiography was not feasible due to technical issues, and transthoracic imaging was inconclusive (Figure 7). The device was electively retrieved, and the patient was referred for surgical closure. Device distribution included Lifetech occluders in 42 patients (67.7%), Cocoon and Occlunix in nine each (14.5%), and Amplatzer in two (3.2%) (Table 2).
(a) Subcostal echocardiogram showing large ASDs separated by a tissue strand (black arrow). (b) TEE showing a large defect with a thin posterior rim (orange arrow). (c) Subcostal view showing residual flow beyond the device’s inferior edge (green arrow). (d–f) Fluoroscopy showing balloon-assisted (blue arrow) device deployment with good configuration in AP (e) and lateral (f) views.
Four patients (6.5%) had a bulky device appearance post-release, though all remained stable with effective closure on follow-up (Figure 3c,d). These large devices showed favourable remodelling over time, with improved profiles and normalisation of right heart dimensions (Figure 1d–f). Cobra-shaped deformation of the left atrial disc was seen in six patients (9.7%) and required gentle repositioning or redeployment for optimal stability (Figure 8e–g). Coronary angiography was performed in 60 patients (96.8%); two younger patients in the early study period were not evaluated. One patient had significant coronary artery disease (Figure 9f), while minor coronary artery disease was found in five (8.1%). No significant differences in complications or adverse events were observed between groups (Table 2).
(a) Subcostal echocardiogram with colour Doppler showing two left-to-right shunt jets across the atrial septum (orange arrows). (b) TEE with colour Doppler confirming ASD (white arrow). (c) 3D TEE showing a large defect and a smaller additional ASD (large and small skyblue arrow). (d) Follow-up four-chamber view showing stable device position (blue arrow). (e, f) Fluoroscopic images demonstrating cobra-like deformation of the left atrial disk (red arrow). (g, h) Fluoroscopy showing proper disk alignment after gentle repositioning (green arrow).
(a) TEE showing a large ASD with left-to-right shunting (white arrow). (b) 3D TEE reconstruction delineating the defect (white arrow). (c, d) Fluoroscopic images showing balloon- (black arrow). and TEE-guided device deployment with good positioning (sky-blue arrow). (e) Apical four-chamber view confirming complete occlusion (blue arrow). (f, g) Left coronary angiogram revealing significant proximal left anterior descending artery stenosis (red arrow)., successfully stented (green arrow). with restoration of flow.
Two patients underwent staged interventions for associated conditions. The first had a large secundum atrial septal defect with floppy anterior rim and single-vessel coronary artery disease. Following a multidisciplinary decision, atrial septal defect closure was done first using a large occluder under transesophageal echocardiography and balloon guidance, followed by coronary stenting (Figure 9). The patient remains well on follow-up. The second had severe pulmonary stenosis and a large atrial septal defect with bidirectional shunt. A staged approach was followed: balloon valvotomy first, then fenestrated device closure. Post-procedure, the patient developed transient supraventricular tachycardia, managed medically with beta-blockers. Follow-up showed proper device position, reduced valve gradient, mild left-to-right shunt across the fenestration, and persistence of normal sinus rhythm (Figure 10).
(a) Lateral angiogram showing doming of the pulmonary valve (orange arrow). (b) CW Doppler showing severe gradient across the valve. (c) Balloon valvotomy image with reduction of valve waist (white arrow). (d) Follow-up echo showing reduced gradient. (e, f) TEE showing large ASD with deficient posteroinferior rim (yellow arrow). and bidirectional (predominantly left-to-right) shunting. (g) 3D echo demonstrating the ASD (white arrow). (h, i) Fluoroscopy showing device deployment in right pulmonary vein approach (blue arrow). with optimal positioning (red arrow). (j, m) Apical and subcostal views with colour Doppler confirming effective closure and left-to-right shunt through the fenestration(red arrow). (k, l) ECG showing narrow complex supraventricular tachycardia resolving with treatment.
Dual antiplatelet therapy with aspirin and clopidogrel was the most common regimen, used in 48 patients (77.4%). Five received aspirin with warfarin, two clopidogrel with warfarin, and three apixaban with clopidogrel (Table 1). As per our institutional protocol, dual antiplatelet therapy (aspirin and clopidogrel) was prescribed for one year, unless there were additional indications for prolonged therapy such as concomitant coronary artery disease. In patients with fenestrated occluders, antiplatelets were continued to maintain fenestration patency. For those with atrial fibrillation, one antiplatelet agent along with oral anticoagulation (warfarin or apixaban) was used, guided by individual patient risk profiles. Patients with diastolic dysfunction were treated with beta-blockers and diuretics. No significant medication differences were observed between groups.
Procedure-related complications and follow-up events are summarised in Table 3. No procedure-related deaths occurred. One woman with a large atrial septal defect and floppy posteroinferior rim underwent successful balloon-assisted, transesophageal echocardiography-guided closure but later developed atrial flutter, managed with rate control and warfarin. With a history of irregular medication adherence, she presented two months later with an intracerebral haemorrhage and died (Figure 4). In another case, transesophageal echocardiography was not possible due to failed probe insertion under general anaesthesia. A 42 mm occluder was deployed under transthoracic echocardiography guidance. Although echo showed good positioning, fluoroscopy revealed suboptimal device profile, and post-release embolisation necessitated surgical retrieval and closure (Figure 2).
Complication and events
A late complication occurred in a woman who developed device-related infective endocarditis five months after successful atrial septal defect closure with a 48 mm device and floppy rim. This was associated with a tubo-ovarian abscess from endometriosis. Despite antibiotics and abscess drainage, she required surgical device explantation and atrial septal defect closure. She recovered fully, and follow-up imaging showed satisfactory repair (Figure 3). One young woman developed fever, cough, and leukocytosis three weeks post-closure. Imaging showed left upper lobe collapse and consolidation. High-Resolution Computed Tomography (HRCT) confirmed findings; she recovered with conservative treatment within 10 days (Figure 11).
(a) Posteroanterior chest X-ray post-device closure showing well-positioned device and clear lung fields. (b) X-ray at three weeks showing homogeneous left hemithorax opacity (red arrow). (c) Follow-up X-ray after 10 days showing complete resolution. (d, e) High-resolution CT chest revealing consolidation in the left upper lobe(green arrow).
In another case, super-stiff wire entrapment occurred in the left upper pulmonary vein during device deployment (44 mm). Retrieval was achieved using a 7F coronary guide catheter, and balloon-assisted closure was completed. There was no evidence of pulmonary vein injury, and the patient remained haemodynamically stable throughout the procedure. Post-procedure, the patient developed paroxysmal tachycardia, which resolved within 36 hours with amiodarone and beta-blockers. She remains well (Figure 12).
(a) Apical four-chamber echocardiogram showing a large ASD. (b) 2D and (c) 3D TEE images demonstrating the defect with floppy margins (sky blue arrow). (d) Post-procedure echo showing well-seated but mildly bulky device (orange arrow). (d) Fluoroscopy showing super-stiff wire lodged in the left upper pulmonary vein (red arrow). (e) Wire retrieval using guiding catheter manoeuvres (green arrow). (f) Final fluoroscopic image showing well-deployed device with balloon-assisted placement (black arrow).
Arrhythmia was the most common complication, seen in 12 patients (19.3%) (Table 3). Intra-procedural atrial arrhythmias occurred in three (4.8%), and post-procedural atrial arrhythmias innine (14.5%). One patient developed junctional tachycardia and another, ventricular tachycardia. A patient with pre-existing ventricular ectopy was managed with beta-blockers. Sinus bradycardia with transient moderate pulmonary hypertension in one patient likely reflected underlying diastolic dysfunction; both resolved within a year. No complete heart block was observed.
Bleeding complications occurred in five patients (8.1%). One had a major intracranial haemorrhage (as described). Another had bleeding from a large sheath site, managed with transfusion and compression. Subsequently, all large-bore venous accesses were secured using figure-of-eight sutures. Ecchymosis in three patients (4.8%) resolved after stopping clopidogrel. No significant differences were found between the two occluder groups in terms of anatomical, procedural, haemodynamic, or complication-related parameters. Mean follow-up duration was 29.29 months. Three patients (international/geographically remote) were followed locally beyond one year, with clinical status confirmed via telephonic consultation and local cardiologist validation. The rest continued regular in-person follow-up.
Discussion
Secundum atrial septal defects are the most common CHD in adults. While surgery was traditionally preferred for large or complex atrial septal defects, advances in transcatheter techniques now allow safe percutaneous closure, despite technical challenges. Initially, the 40 mm Amplatzer was the largest occluder available. In a series of 87 patients, a 96.5% success rate was reported with this device, demonstrating favourable safety outcomes.Reference Dalvi, Sheth, Jain and Pinto9 More recently, newer standard and custom occluders up to 50 mm, as well as hybrid or transthoracic approaches, have expanded feasibility, although supporting literature remains limited.Reference Sivakumar, Viswambaran and Bhattacharjya4,Reference Chen, Cao, Zhang, Chen, Xu and Zhang10,Reference Dou, Kan and Guo11
Closure of large or complex atrial septal defects remains technically demanding due to anatomical variations—such as deficient rims, aneurysmal septum, or multiple fenestrations—which can compromise device stability and increase risks of embolisation or residual shunting. The anatomical nuances and operator and unit expertise are important determinants of outcome. In our cohort of 62 patients, 77.4% had thin aneurysmal margins, and 19.4% had deficient posteroinferior rims. While imaging can assess defect size and rim appearance, rim strength—a key determinant of device stability—is difficult to predict. Such complexities demand detailed pre-procedural imaging, careful device selection, and, at times, escalation to surgical closure.Reference Ananthakrishna Pillai, Upadhyay, Gousy and Handa12–Reference Wolf14
Device sizing is particularly challenging in large atrial septal defects with floppy or deficient rims. No universal sizing formula exists. In this study, most patients had thin, pliable margins. We used the thick-to-thick septal dimension, oversizing the device by 2–6 mm. This aligns with Sivakumar et al., who advised 2–10 mm oversizing based on anatomy, especially for floppy posteroinferior rims.Reference Sivakumar, Viswambaran and Bhattacharjya4 Device selection is influenced by anatomical complexity, imaging quality, and operator experience.
Assisted implantation techniques are essential for managing difficult anatomies. Balloon-assisted technique and pulmonary vein–guided deployment are widely used in cases with rim deficiencies or septal malalignment, enhancing success rates.Reference Sivakumar, Viswambaran and Bhattacharjya4,Reference Ananthakrishna Pillai, Upadhyay, Gousy and Handa12 In our series, 93.75% of patients required balloon-assisted closure technique, achieving an 88.7% success rate—comparable to Pillai et al.’s reported 82–90%.Reference Pillai, Rangaswamy Balasubramanian, Selvaraj, Saktheeswaran, Satheesh and Jayaraman15 In the later part of the index study, the device-assisted technique was introduced, proving effective in 18 patients (98.4%) and offering an alternative in challenging cases. Dilator-assisted deployment was used in a few cases when balloon-assisted closure technique or device closure technique hardware was unavailable. The majority of patients achieved successful atrial septal defect device closure in a single attempt, though success was slightly lower in those receiving larger occluders. A notable proportion of cases, particularly in the larger device group, required multiple deployment attempts, suggesting increased technical complexity with increasing device size. As in previous studies, a floppy or deficient posteroinferior rim remained a key predictor of procedural failure.Reference Ananthakrishna Pillai, Upadhyay, Gousy and Handa12,Reference Jung, Kim, Jung and Choi16
Closure of large secundum atrial septal defect using occluders ≥40 mm poses significant technical challenges—not just due to defect size, but also anatomical factors such as deficient inferior vena cava rims, floppy posteroinferior margins, and malaligned septa. Device embolisation often results from undersizing or incomplete rim engagement.Reference Dalvi, Sheth, Jain and Pinto9 In our series, four patients required surgical closure: three electively after device retrieval and one for intraprocedural embolisation. All had floppy or deficient posteroinferior rims. While retroaortic rim deficiency is usually acceptable, mitral and posteroinferior rim deficits carry higher procedural risk.Reference Sivakumar, Viswambaran and Bhattacharjya4,Reference Jung, Kim, Jung and Choi16
Inadequate imaging due to poor transthoracic windows, suboptimal transesophageal echocardiography probe positioning, or lack of access to quality transesophageal echocardiography—can obscure septal anatomy and compromise outcomes. Two failures in our series were attributed to faulty transesophageal echocardiography probe or inability to perform transesophageal echocardiography due to oesophageal issues. These cases emphasise the importance of comprehensive transthoracic echocardiography and transesophageal echocardiography imaging for complete rim assessment and secure device deployment. They also highlight current limitations in reliably assessing rim integrity under physiological conditions.Reference Dalvi, Sheth, Jain and Pinto9
Left ventricular diastolic dysfunction is another key consideration. Chronic left-to-right shunting often masks reduced left ventricular compliance. Atrial septal defect closure—especially in elderly or hypertensive patients—may unmask this dysfunction, leading to elevated left ventricular filling pressures and heart failure symptoms. In this series, left ventricular end-diastolic pressure > 15 mmHg was considered elevated; post-occlusion left ventricular end-diastolic pressure > 18–20 mmHg or a rise >5 mmHg from baseline was clinically significant. Fenestrated devices were used in two such cases, enabling controlled interatrial shunting to buffer the pressure shift.Reference Singhi, Mohapatra and Kumar7,Reference Singhi, Mahesh and Kumar17
Pulmonary pressure assessment is a critical component of pre-procedural evaluation for atrial septal defect device closure, and was performed in all patients in the present series. One patient had severe pulmonary hypertension with increased pulmonary blood flow; closure with a fenestrated occluder led to a significant reduction in flow-driven pressure. Several others had moderate pulmonary hypertension due to volume overload, which improved post-closure. One patient showed post-procedural pressure elevation, highlighting the need for close follow-up. These findings reinforce the value of early atrial septal defect closure—before irreversible pulmonary vascular changes—to optimise outcomes and potentially reverse pulmonary hypertension.Reference Singhi, Mohapatra and Kumar7,Reference Le Gloan, Legendre, Iserin and Ladouceur18
Despite generally favourable outcomes, transcatheter closure of large atrial septal defect carries a small but important risk of complications. A meta-analysis of 28,142 patients reported major complications in 1.4%, with device embolisation, stroke, and thrombosis being most common.Reference Abaci, Unlu, Alsancak, Kaya and Sezenoz19 Berdat et al. noted surgical intervention in ∼ 8% of cases, mainly for residual shunting or device malalignment.Reference Berdat, Chatterjee, Pfammatter, Windecker, Meier and Carrel20 In our series, three patients (4.8%) underwent elective surgery, and one (1.6%) for intraprocedural device embolisation.
Common complications in index study included atrial arrhythmias (25.8%), bleeding, and ecchymosis. Atrial arrhythmias (19.3%) were usually transient and managed conservatively. Prior studies report lower arrhythmia rates (∼6%) with structured follow-up.Reference Sivakumar, Viswambaran and Bhattacharjya4,Reference Bitar, Dorotan-Guevara, Lucas and Snyder21,Reference John Sutton, Tajik and McGoon22 One patient developed life-threatening intracranial haemorrhage, likely from anticoagulation for arrhythmia, emphasising the need for individualised antithrombotic regimens.Reference Bitar, Dorotan-Guevara, Lucas and Snyder21,Reference John Sutton, Tajik and McGoon22 No cases of AV block, erosion, or device thrombus were observed. Access site complications were minimised with figure-of-eight sutures; however, clopidogrel-associated ecchymosis occasionally influenced post-procedure therapy. These findings reinforce the need for vigilant post-procedural monitoring and long-term follow-up.Reference Singhi, Mahapatra, Kumar, Dey, Mishra and De3–Reference Lopez, Dalvi and Balzer5,Reference Dalvi, Sheth, Jain and Pinto9,Reference Jalal, Hascoet and Baruteau13,Reference Jung, Kim, Jung and Choi16
Intraprocedural challenges included device malposition or deformation (e.g., “cobra-shaped” in 9.7%), requiring repositioning or replacement. Stiff wire malfunction was encountered with ethylene oxide–sterilised hardware. In 6.5%, the device appeared bulky but remodelled on follow-up. Variability in septal integrity complicates predicting device–tissue interaction. No cases of air embolism or new valvular regurgitation occurred, though reported elsewhere.Reference Sivakumar, Viswambaran and Bhattacharjya4
Device-related endocarditis, while rare, is serious. One patient developed late-onset infection from a gynaecologic source, necessitating surgical explantation of the device. This highlights the importance of long-term vigilance for bacteraemia-related complications.Reference Nasso, Condello and DAbramo23 No significant difference in outcomes or complications was observed between occluder sizes or device types. While large device closure appears safe and effective, these findings emphasise the importance of continued follow-up. Future improvements should focus on flexible, biocompatible occluder designs and integration of advanced imaging—such as 3D or intracardiac echocardiography—to optimise deployment and reduce complications.
Study limitations
This retrospective, dual-centre study has inherent limitations, including potential documentation gaps and unmeasured confounders. Very large atrial septal defects requiring 40–50 mm occluders are uncommon, making large-scale data collection challenging. The modest sample size limits broader applicability. Selection bias is possible, as patients with complex atrial septal defects were selectively referred for transcatheter closure, all performed under a single lead operator. The 10-year study period involved multiple hospital software transitions, affecting data retrieval consistency. Long-term follow-up was inconsistent; some patients were monitored locally, introducing variability in post-procedure assessment. Additionally, device–tissue interaction and post-deployment remodelling vary widely and may not be fully captured in retrospective analysis. Despite encouraging procedural success and safety outcomes across devices, larger prospective multicenter studies with standardised imaging and follow-up protocols are essential for more generalisable and definitive conclusions.
Conclusion
Transcatheter closure of large atrial septal defects using large (40 –50 mm) occluders is a safe and effective alternative to surgery when supported by careful patient selection, accurate anatomical assessment, and procedural expertise. Despite technical challenges posed by rim deficiencies, aneurysmal septum, pulmonary hypertension, or diastolic dysfunction, favourable outcomes can be achieved with comprehensive anatomical and haemodynamic assessment. The use of fenestrated devices, advanced imaging, and multidisciplinary management further enhances safety in high-risk cases. This study highlights that, with tailored planning and vigilant follow-up, percutaneous closure can be successfully extended to anatomically complex and physiologically vulnerable patients.
Acknowledgements
We extend our sincere appreciation to the outpatient support staff of the Cardiology and Congenital Heart Disease Department of both the organisations, including Ms. Moumita Sil (Cardiology Technician), Ms. Pratima Hembram (Nursing Coordinator), and Ms. Mandira Mondal (OPD Coordinator), for their invaluable assistance with patient assessment and data collection. Mr. Vikas Tiwari and the team of the Department of Information Technology were very helpful in mining the digital data. Sincere thanks to team CTVS of both hospitals for helping during the procedure as backup. Sincere appreciation for Dr Sandip Sardar and Dr Monalisa Dutta for helping with the embolisation case and providing the intraoperative images. Our heartfelt thanks go to Dr. Sivakumar K of the Department of Pediatric Cardiology at Madras Medical Mission, Chennai, India, for his continuous academic and clinical guidance. Finally, we express our deepest gratitude to Finally, we express our deepest gratitude to Manipal Hospital, EM Bypass, and RKMSP Hospital for their unwavering support throughout the study. We are particularly grateful to the Dr. Anil Singhi Foundation for their generous academic support of our scientific research endeavours.
Financial support
Nil.
Competing interests
The authors declare no conflict of interest.
Ethical standard
This scientific work adhered to the approved protocol of the institutional Clinical Ethics and Research Committee (CERC/2021/Jun/iv). Patient confidentiality was ensured by anonymizing all data used in images and analyses. As the study utilised anonymized data, the ethics committee protocol waived the requirement for individual patient consent.
