Abstract
Data on arrhythmia outcome following device closure of atrial septal defect (ASD) are lacking. This study provides medium-term follow-up data on atrial arrhythmias in patients who were ≥40 years of age at the time of transcatheter ASD closure.
It is a retrospective review. Mean age of the 159 patients was 57 years. Median follow-up was 3.6 years (range 6 months–10.9 years). Patients were classified, according to arrhythmia status prior to ASD closure, into Group I, no history of atrial arrhythmia (n = 119, mean age 55.5 years); Group II, paroxysmal atrial arrhythmia (n = 18, mean age 55.7 years); and Group III, persistent atrial fibrillation (n = 22, mean age 65.7 years). Group III patients were significantly older, had larger left atrial size, and had higher mean pulmonary arterial pressure than Group I and II patients (P < 0.001). Prior to closure, radiofrequency ablation was carried out in 12/18 (66%) of Group II and 3/22 (14%) of Group III. After device closure, 7 patients (6%) of Group I developed new atrial fibrillation. Fifty per cent (9/18) of Group II but only 9% (2/22) of Group III were in sinus rhythm on follow-up.
Device closure alone in patients with persistent atrial arrhythmia is not likely to restore sinus rhythm in the medium term. New atrial arrhythmia occurred in 6% of patients who were in sinus rhythm prior to device closure. At least 50% of the patients with paroxysmal atrial arrhythmia continue to have significant atrial arrhythmia following device closure, and the role of ablation prior to closure in patients with a history of arrhythmia requires refinement.
Medium term data on arrhythmia outcome after device closure of ASD are lacking.
About 6% of patients (≥40 years), with no prior history of atrial arrhythmia, develop atrial fibrillation on medium term follow-up.
In patients known to have persistent atrial fibrillation, prior to ASD closure, device closure alone is not likely to result in sinus rhythm.
Ablation procedures prior to ASD can restore sinus rhythm in some patients with a history of arrhythmia, but those requiring more than one ablation procedure have a high rate of recurrence, and this group requires further study.
There is a risk of new atrial fibrillation in patients over the age of 40 years at the time of device closure of ASD is ∼6% and suggests that routine electrophysiological study prior to closure is not justified.
Introduction
Atrial tachyarrhythmia is found in 15–20% of adults over the age of 40 years referred for management of haemodynamically significant atrial septal defect (ASD).1,2 These arrhythmias may be associated with symptoms and also are an important risk factor for embolic events such as cerebrovascular accidents. As percutaneous device closure of ASDs has evolved to become the standard of care, it is clear that arrhythmia management needs to be specifically addressed, particularly in the older age group, as part of the ASD care package. The causative effect of ASD on atrial arrhythmia behaviour is multifactorial, and it has been shown that closing the ASD alone, by surgery, does not result in sinus rhythm or prevent development of new atrial arrhythmia.3,4 At present, there is a gap in knowledge on arrhythmia outcome and management in patients who have device closure of haemodynamically significant ASDs.
We report the medium term arrhythmia outcomes in a group of consecutive patients who were over 40 years of age at the time of ASD closure. The main aims were (i) to describe arrhythmia status prior to device closure, (ii) describe the patient characteristics of those with persistent or paroxysmal atrial arrhythmia, (iii) to investigate whether new arrhythmia developed and identify factors associated with the onset of new arrhythmia, and (iv) to report the outcome of current ablation strategies, in patients with a history of atrial arrhythmia.
Methods
All patients who underwent device closure of their ASD in the period 1997–2014 were identified from the departmental database. Ethical approval to conduct this retrospective analysis was obtained from the local research committee.
The study was confined to patients ≥40 years at the time of ASD closure. Patients with cryptogenic stroke who had device closure of patent foramen ovale or small ASDs were excluded as the study was confined to patients with significant left to right shunts. Patients with significant left ventricular dysfunction were not offered ASD closure.
Atrial septal defect closure
All procedures were carried out under general anaesthetic, and a haemodynamic study was carried out prior to device closure. Trans-oesophageal echocardiography (TOE) was used to assess the defect and to monitor the procedure. Heparin was given prior to placing a guide wire in the pulmonary vein. Balloon sizing of the defect was done in all cases. Generally, a device was chosen based on the balloon diameter at the point colour flow was abolished. Stability was tested by gently pushing and pulling the device under fluoroscopy and TOE.
Data collection
Clinical information was retrieved from the notes and the catheter reports. The echocardiograms recorded at the time of ASD closure were stored on the departmental echocardiographic database (Medcon, Medcon UK Ltd, London, UK) and were reviewed by one of the team (P.D.). Data on left atrial (LA) volume could not be retrospectively calculated from archived studies. Left atrial longitudinal diameter was measured using apical four-chamber view at the end of ventricular systole in all patients. Catheter ablation for those with a history of atrial tachyarrhythmia prior to ASD closure was determined by the outcome of a multi-disciplinary team discussion between adult congenital heart disease and electrophysiology specialists before any intervention. Using the international consensus on atrial fibrillation (AF) nomenclature,5 patients were subdivided into Group I—those with no history of arrhythmia; Group II—patients with a history of paroxysmal arrhythmia (self-limiting episodes of <7 days); and Group III—patients with persistent AF prior to device closure. Rhythm status following device closure was assessed at regular follow-up intervals (3 months, 6 months, and 12 monthly thereafter) by symptoms and standard 12-lead ECG.
The primary outcome was time to atrial arrhythmia, defined as documented atrial flutter or fibrillation that occurred or persisted at >3 months following transcatheter closure. Peri-procedural atrial arrhythmia that resolved within 3 months of closure and did not recur during follow-up was considered temporary and did not contribute to the outcome measures. The minimum follow-up period was 12 months. Twenty-four-hour ECG (Holter) monitoring was carried out routinely at 12–18 months post-device closure and in patients with symptoms suggestive of arrhythmia. The most recent clinical contact was reviewed, and all patients who were discharged to peripheral units were contacted by letter or telephone interview using a standard questionnaire.
Statistical analysis
Kaplan–Meier analysis was undertaken to estimate freedom from new-onset atrial arrhythmia in Group I. The baseline characteristics between groups were compared using Mann–Whitney U and Kruskal–Wallis tests. Multi-variable regression was used to identify risk factors for the development of arrhythmia following device closure in patients in Group I. A value of P < 0.05 was considered significant. Data analysis was carried out using Microsoft Excel 2011 and MedCalc (Ostend, Belgium).
Results
In the period January1997 to December 2014, 159 consecutive patients ≥40 years underwent ASD device closure. The mean age of the total group was 57 years (±11 years). The median follow-up was 3.6 years (range 6 months–10.9 years). Device closure was successful in all cases, and one patient had two devices placed. There were two major complications with one patient developing transient pulmonary oedema requiring ventilation for 24 h and one patient had a recurrent pericardial effusion drained surgically on the evening of the procedure. Eleven patients (7%) had minor complications—bruising and/or haematoma in groin.
Groups and patient characteristics
The majority of patients were in sinus rhythm [Group I = 119/159 (75%)]. There were 40 patients (25%) with a prior history of atrial arrhythmia (Groups II and III) (Table 1). In 18 of these, the arrhythmia was classified as paroxysmal (Group II), and in 22 patients, persistent (Group III) prior to ASD closure. Twelve patients (67%) of Group II and 3 patients (14%) of Group III underwent electrophysiological study with catheter ablation prior to device closure.
Breakdown of study cohorts and the individual characteristics based on history of arrhythmia
| Group I n = 119 | Group II n = 18 | Group III n = 22 | P-value | |
|---|---|---|---|---|
| Age (years) | 55.5 (10.6) | 55.7 (10.5) | 65.7 (9.2) | <0.001 |
| PA pressure (mmHg) | 20 (5.8) | 20 (6.5) | 29 (8) | <0.001 |
| Device size (mm) | 24 (6.9) | 24 (5.0) | 26 (7.9) | 0.167 |
| LA length (cm) | 5.4 (0.8) | 6.1 (1.1) | 7.2 (1.6) | <0.001 |
| Group I n = 119 | Group II n = 18 | Group III n = 22 | P-value | |
|---|---|---|---|---|
| Age (years) | 55.5 (10.6) | 55.7 (10.5) | 65.7 (9.2) | <0.001 |
| PA pressure (mmHg) | 20 (5.8) | 20 (6.5) | 29 (8) | <0.001 |
| Device size (mm) | 24 (6.9) | 24 (5.0) | 26 (7.9) | 0.167 |
| LA length (cm) | 5.4 (0.8) | 6.1 (1.1) | 7.2 (1.6) | <0.001 |
Group I (n = 119), patients with no history of arrhythmia; Group II (n = 18), patients with paroxysmal arrhythmia; Group III (n = 22), patients with persistent arrhythmia. Values are expressed as mean or median with standard deviation. PA, pulmonary artery; LA, left atrium.
Breakdown of study cohorts and the individual characteristics based on history of arrhythmia
| Group I n = 119 | Group II n = 18 | Group III n = 22 | P-value | |
|---|---|---|---|---|
| Age (years) | 55.5 (10.6) | 55.7 (10.5) | 65.7 (9.2) | <0.001 |
| PA pressure (mmHg) | 20 (5.8) | 20 (6.5) | 29 (8) | <0.001 |
| Device size (mm) | 24 (6.9) | 24 (5.0) | 26 (7.9) | 0.167 |
| LA length (cm) | 5.4 (0.8) | 6.1 (1.1) | 7.2 (1.6) | <0.001 |
| Group I n = 119 | Group II n = 18 | Group III n = 22 | P-value | |
|---|---|---|---|---|
| Age (years) | 55.5 (10.6) | 55.7 (10.5) | 65.7 (9.2) | <0.001 |
| PA pressure (mmHg) | 20 (5.8) | 20 (6.5) | 29 (8) | <0.001 |
| Device size (mm) | 24 (6.9) | 24 (5.0) | 26 (7.9) | 0.167 |
| LA length (cm) | 5.4 (0.8) | 6.1 (1.1) | 7.2 (1.6) | <0.001 |
Group I (n = 119), patients with no history of arrhythmia; Group II (n = 18), patients with paroxysmal arrhythmia; Group III (n = 22), patients with persistent arrhythmia. Values are expressed as mean or median with standard deviation. PA, pulmonary artery; LA, left atrium.
Table 1 summarizes the differences between the groups. Patients with persistent arrhythmias (Group III) had statistically significant higher PA pressure, were older, and had larger LA dimensions (all P < 0.001) than the other groups. The patients in the paroxysmal arrhythmia group (Group II) had mean LA length significantly larger than those without arrhythmia (Group I) (P = 0.03). There was no significant difference in device size between the groups (P= 0.17). There was no statistical difference between patients with paroxysmal arrhythmias (Group II) and those with no arrhythmias (Group I) in terms of age (P = 0.47) and mean PA pressure (P = 0.26).
Arrhythmia outcome
Figure 1 shows the arrhythmia outcome for each subgroup. In all patients, the arrhythmia that either developed or recurred was atrial fibrillation. In those with persistent atrial fibrillation prior to ASD closure (Group III, n= 22), sinus rhythm was initially maintained in three, all of whom had an ablation procedure prior to device closure, and in two of these sinus rhythm persisted on medium term follow-up. β-Blocker was the only medication these patients were taking prior to device implantation and was discontinued at 6-month post-procedure. Cardioversion had been attempted in a further five patients from Group III but was unsuccessful.
Summary of study cohort of patients with ASD, subdivided into arrhythmia history and electrophysiological intervention prior to device closure and subsequent arrhythmia outcome, see text for details on definition.
In the paroxysmal atrial arrhythmia group (Group II, n = 18), 12 had an electrophysiological ablation (5 of these had a repeat procedure) prior to ASD closure. The numbers are relatively small but the proportions who maintained sinus rhythm (off medication) was similar between those who had an ablation procedure (6/12) and those who did not (3/6). Interestingly, the five patients who had repeated ablations prior to ASD closure continued to have arrhythmia in the medium term. In patients from Group II who had an ablation (n = 12), the focus of arrhythmia was in the left atrium in 6 patients, in the right atrium in 3 patients, and in both atria in 3 patients.
Seven (6%) of the group with no prior arrhythmia (Group I, n = 119) developed atrial fibrillation. These new atrial arrhythmias, which to be included had to persist beyond 3 months after device closure, started at various times (9 days to 6 years) following the device closure. It is interesting to note however that most arrhythmias developed relatively early, and this is shown graphically in Figure 2 (Kaplan–Meier curve). The patients who developed atrial fibrillation had a trend towards larger devices (34 vs. 24 mm); however, this did not reach statistical significance (P= 0.21). There was also a trend (Table 2) towards increased LA longitudinal diameter and elevated pulmonary pressures in the patients who went on to develop atrial arrhythmia, but these did not reach statistical significance, [(LA length 6.8 vs. 5.4 cm, P = 0.61); (mean PA pressure 22 vs. 20 mmHg, P = 0.23), Table 2]. Cardioversion had been carried out in three of these patients (3 months, 18 months, and 9 years post-device closure) which had been initially successful in all (confirmed by Holter monitoring), but atrial fibrillation recurred in one of these three patients within 2 years.
Characteristics of patients in Group I (no arrhythmia at the time of referral for ASD closure) who developed arrhythmia after ASD closure (7 patients), compared with those who remain in sinus rhythm (112)
| SR n = 112 | New arrhythmia n = 7 | P-value | OR (95% CI) | |
|---|---|---|---|---|
| Age (years) | 55.6 (10.7) | 54.9 (8.4) | 0.641 | 1.12 (0.70–1.77) |
| Mean PA pressure (mmHg) | 20 (5.8) | 22.5 (4.2) | 0.232 | 1.86 (0.67–5.19) |
| Device size (mm) | 24 (6.6) | 34 (7.5) | 0.211 | 1.94 (0.69–5.45) |
| LA length (cm) | 5.4 (0.7) | 6.8 (0.9) | 0.611 | 2.61 (0.06–105.20) |
| SR n = 112 | New arrhythmia n = 7 | P-value | OR (95% CI) | |
|---|---|---|---|---|
| Age (years) | 55.6 (10.7) | 54.9 (8.4) | 0.641 | 1.12 (0.70–1.77) |
| Mean PA pressure (mmHg) | 20 (5.8) | 22.5 (4.2) | 0.232 | 1.86 (0.67–5.19) |
| Device size (mm) | 24 (6.6) | 34 (7.5) | 0.211 | 1.94 (0.69–5.45) |
| LA length (cm) | 5.4 (0.7) | 6.8 (0.9) | 0.611 | 2.61 (0.06–105.20) |
PA, pulmonary artery; LA, left atrium; SR, sinus rhythm.
Characteristics of patients in Group I (no arrhythmia at the time of referral for ASD closure) who developed arrhythmia after ASD closure (7 patients), compared with those who remain in sinus rhythm (112)
| SR n = 112 | New arrhythmia n = 7 | P-value | OR (95% CI) | |
|---|---|---|---|---|
| Age (years) | 55.6 (10.7) | 54.9 (8.4) | 0.641 | 1.12 (0.70–1.77) |
| Mean PA pressure (mmHg) | 20 (5.8) | 22.5 (4.2) | 0.232 | 1.86 (0.67–5.19) |
| Device size (mm) | 24 (6.6) | 34 (7.5) | 0.211 | 1.94 (0.69–5.45) |
| LA length (cm) | 5.4 (0.7) | 6.8 (0.9) | 0.611 | 2.61 (0.06–105.20) |
| SR n = 112 | New arrhythmia n = 7 | P-value | OR (95% CI) | |
|---|---|---|---|---|
| Age (years) | 55.6 (10.7) | 54.9 (8.4) | 0.641 | 1.12 (0.70–1.77) |
| Mean PA pressure (mmHg) | 20 (5.8) | 22.5 (4.2) | 0.232 | 1.86 (0.67–5.19) |
| Device size (mm) | 24 (6.6) | 34 (7.5) | 0.211 | 1.94 (0.69–5.45) |
| LA length (cm) | 5.4 (0.7) | 6.8 (0.9) | 0.611 | 2.61 (0.06–105.20) |
PA, pulmonary artery; LA, left atrium; SR, sinus rhythm.
Kaplan–Meier curve representing freedom from developing arrhythmia in those without history of arrhythmia (Group I, n = 119) prior to transcatheter ASD closure.
Discussion
The association between atrial arrhythmia and ASD is well recognized, and older patients are at the highest risk. A previous study on device closure in adults identified age over 40 years as an important risk factor for arrhythmia,6 and that age group was the focus of our study. We report that in this group (patients ≥40 years at the time of ASD closure), 25% had a history of either paroxysmal or persistent atrial arrhythmia, prior to transcatheter device closure. The mechanism for the development of arrhythmia is not completely understood, but we demonstrate a clear relation to LA size. It has been suggested that removing the left to right shunt, which should of course decrease LA size, would remove the risk of future atrial arrhythmia. Surgical results of ASD closure in adults over the age of 40 years are disappointing in terms of arrhythmia control.7–13 It had been suggested that closure of the ASD by surgery before the age of 21 years would result in a risk of late atrial arrhythmia similar to the normal population.14 However, a recent study showed that surgical ASD closure in childhood carries a significant long-term risk of late arrhythmia.15 It is likely therefore that the surgical scar itself may be important in terms of arrhythmia risk, raising the possibility that a transcatheter procedure may be associated with an improved arrhythmia outcome. This study provides some much needed data in this area.
We report a 6% incidence of new atrial arrhythmia after a median follow-up of 2.5 years in patients treated by device, and to our knowledge, this is the longest follow-up study of this group published.16–18 There are some interesting trends. The onset of the new arrhythmia is generally in the first 6 months following device closure. Our data showed a trend to increased device size, larger LA dimensions, and higher pulmonary pressure in those who develop arrhythmia, but these did not reach statistical significance. It is possible that measurement of LA volume would have been more predictive of later arrhythmia, but we were not able to measure that in this study. It is also possible that asymptomatic arrhythmias were present prior to device closure in these patients but were not detected as we do not routinely carry out prolonged monitoring in asymptomatic patients prior to ASD closure.
We have shown that the majority of atrial arrhythmia circuits in these patients involve the left atrium which may, of course, now be more difficult to access after placing a device. There are reports of successfully crossing the device after perforation of the septum around the device or indeed going through the device,19,20 but this probably increases the risk of the procedure and may limit the success of the ablation procedure itself. It should also be remembered, however, that 94% of those with no history of arrhythmia remain in sinus rhythm in the medium term, making it difficult to justify a generally more aggressive approach prior to device closure in terms of electrophysiological testing. The challenge is to try and identify the subgroup of patients who are in sinus rhythm prior to intervention but have a high risk of developing atrial arrhythmia post-closure. The possibility that echo-Doppler parameters and/or prolonged ECG monitoring in patients with no history of arrhythmia may help identify such patients needs further consideration.
In this age group (≥40 years), at the time of device closure, a significant proportion of patients already have a history of atrial arrhythmia prior to referral for closure (25% in our study) and the main group in our report were those in persistent atrial fibrillation. It is generally accepted that the risk of atrial fibrillation increases with age,16,17 and therefore, it is not surprising that those with an ASD and persistent atrial fibrillation were older, had higher pulmonary artery pressure, and had increased LA dimensions compared with the sinus rhythm group. It is clear from our data that device closure alone generally is not associated with restoration of sinus rhythm, and this is similar to other reports.16 Only 2 out of 22 patients in our study with persistent arrhythmia maintained sinus rhythm at a median follow-up of 2.6 years, and both of these patients had EP ablation prior to device closure. It is possible that ablation prior to device closure may be beneficial to a subgroup of patients with persistent atrial fibrillation, but the criteria for a more aggressive approach remain to be defined.
Those with paroxysmal atrial arrhythmia formed the smallest group in our study (11%), and it would appear that ∼50% of these patients continue to have arrhythmia after device closure and this is in keeping with other reports.
We were not able to demonstrate an improved outcome in the subgroup of 12 patients with paroxysmal arrhythmia who had ablation procedures prior to device closure despite the fact that some of these patients having more than one ablation procedure. In fact, it is interesting to note that those who required only one ablation, most (6 out of 7 patients) were still in sinus rhythm on medium term follow-up. There is therefore a subgroup that does well, and the challenge is to identify and possibly adopt a different approach to those who require more than one procedure. The poor outcome (in terms of arrhythmia) of this subgroup could be explained by the difficulty in identifying specific pathways and those identified may turn out to be the epiphenomenon which do not correspond to the actual pathology. It is important to remember that the subgroups reported here are relatively small and difficult to draw firm conclusions on an ablation strategy. It seems that a strategy of ablation prior to device closure, as currently practiced, particularly in those with paroxysmal arrhythmia, has a variable outcome and needs refinement.
Conclusion
The optimal arrhythmia management of adult patients, ≥40 years, with haemodynamically significant secundum ASDs, suitable for device closure remains unresolved if the goal is to maintain or achieve sinus rhythm in the medium to long term. There remains a significant risk (6%) of developing new atrial arrhythmia in the medium term with device closure alone, and this tends to occur relatively early. The incidence of new arrhythmia is broadly similar to reported surgical data. Our data show a non-significant trend towards larger devices in that subgroup, but further study is required to identify these patients, who may benefit from electrophysiological study, prior to device closure.
Our data would support the hypothesis that atrial arrhythmia due to chronic left to right shunts is related to chronic changes in the left atrium which are not reversed with abolition of the left to right shunt alone. Unfortunately, identifying these circuits prior to device closure does not appear to be straightforward. Ablation procedures have inconsistent medium-term results in patients with documented atrial arrhythmia prior to device closure with ∼50% having symptomatic arrhythmia on follow-up, and clearly, a better understanding of the arrhythmia circuits and refinement of ablation strategy is needed.
Conflict of interest: none declared.
