Outcomes of Patent Foramen Ovale Transcatheter Closure

Background The risk of erosion of an atrial septal closure device, in particular the Amplatzer Septal Occluder, has been described as higher in patients with a short aortic rim. Similar concern has been applied to patent foramen ovale (PFO) closure devices, but there are only rare reported cases of erosion. It may be that smaller devices are chosen due to fear of device erosion in PFO patients when this is not necessarily an issue. Objectives The authors aimed to assess outcomes after PFO closure with the Amplatzer PFO device in patients with a short (<9 mm) aortic rim. Methods We performed a retrospective analysis of PFO closure for any indication, between 2006 and 2017 at a quaternary center. Preprocedural transesophageal echocardiographic parameters including the aortic rim were remeasured. Long-term outcomes were obtained by linkage to provincial administrative databases. Results Over the study period, 324 patients underwent PFO closure with the Amplatzer PFO device, with a mean age of 49.8 years; 61% had a short aortic rim (<9 mm). The most common indication was cryptogenic stroke (72%); those with longer aortic distance were more likely to have a non-stroke indication for closure, diabetes (15% vs 6.5%, P = 0.04), and heart failure (15.7% vs 4%, P < 0.001). Over a median 7 years of follow-up, there were no cases of device erosion or embolization requiring cardiac surgery. Conclusions In a large cohort with long-term administrative follow-up (1,394 patient-years), implantation of an Amplatzer PFO device was performed safely even in patients with a short aortic rim.

2][3] Less common reasons for closure include platypneaorthodeoxia syndrome, prophylaxis of decompression illness for scuba diving, 4 and prevention of paradoxical embolism in selected patients with permanent indwelling lines or pacemaker wires.
The incidence of procedural complications in clinical trials and registries is very low, with a 0.02% to 0.2% incidence of device erosion 5 or cardiac perforation 6 in long-term follow-up, with an additional 1.4 per 100 person-years incidence of new atrial fibrillation, 7 and 0.58 per 100 person-years incidence of recurrent stroke. 6e Amplatzer PFO Occluder (Abbott Structural Heart) and Gore Cardioform (W.L. Gore) are both approved for use in the United States and Canada.
The initial instructions for use, as included in the Federal Drug Administration approval in 2016, note that the Amplatzer PFO Occluder should only be implanted in patients with adequate aortic and superior vena cava rims, defined as 9 mm on echocardiography. 8The PFO to aortic distance is however shorter than that for many adults. 9These instructions for use were updated in 2019, and now state that the rims have to be 'adequate' without specifying a measurement 10 ; for the Amplatzer Septal Occluder, caution is recommended in patients with deficient rims (<5 mm). 11Unlike for the Amplatzer Septal Occluder, 12 there is a paucity of data to support a higher risk of embolization or erosion of the PFO occluder in patients with a short aortic rim.Theoretically, the structural characteristics of the PFO device, as it is more flexible, not self-centering, and both right and left atrial discs are significantly larger than the defect, may reduce the risks of erosion and embolization compared to the Amplatzer Septal Occluder.In the RESPECT (Randomized Evaluation of Recurrent Stroke Comparing PFO Closure to Established Current Standard of Care Treatment) trial, there were no cases of device embolization reported in the short-or long-term follow-up and only one case of cardiac perforation in the 499 subjects in the device group (further details not available). 6,13[16] We sought to compare the incidence of procedural and long-term adverse events after percutaneous PFO closure with the Amplatzer PFO Occluder by aortic rim distance (ie, <9 mm vs $9 mm).We hypothesize that the Amplatzer PFO Occluder does not require a minimal aortic rim for safe implantation.Demographics, imaging and procedural characteristics, and periprocedural outcomes were abstracted from the chart.Procedural success was defined as device implantation without retrieval or death.Periprocedural outcomes were defined as occurring prior to hospital discharge.

METHODS
ECHOCARDIOGRAPHIC ANALYSIS.Available preprocedural TEE were assessed by 3 reviewers blinded to the clinical outcomes, following the echocardiographic protocol used in the RESPECT trial. 13The aortic distance was measured in the aortic valve short axis view (30 -60 ), and the superior and inferior vena cava distances measured in the bicaval view (90 -120 ) (Figure 1).A short aortic rim was defined as measuring <9 mm.The septum was described as mobile if total septal excursion was 10 to 15 mm; aneurysmal if total excursion was >15 mm or >10 mm in either the right or left atria; and with prominent aneurysm if the excursion was $20 mm.The tunnel height and length were measured in the aortic short axis and bicaval views.Bubble studies were considered positive if more than 1 bubble was seen in the left atrium within the first 3 cardiac cycles, at rest or with a Valsalva maneuver.PROCEDURE.PFO closures were performed under conscious sedation and using fluoroscopic guidance alone with provisional use of intracardiac echo. 17In our practice, an Amplatzer device was chosen except for patients with nickel allergy.The device size  The study was approved by the ethics review board of the University Health Network; a patient consent waiver was granted.

BASELINE DEMOGRAPHICS AND ECHOCARDIOGRAPHIC
PARAMETERS.Over the study period, there were 1,032 patients who underwent PFO closure and were included in the registry; of those, 625 had an Amplatzer PFO device implanted, and 324 (52%) had a preprocedural TEE images available for rim remeasurement (the majority of patients had preprocedural TEE but some were performed at outside institutions).The mean age was 49.8 years, and 56% were male.The most common indication for PFO closure was cryptogenic stroke (247, 72%) (Table 1).Patients with an aortic distance $9 mm were more likely to have a higher body mass index (28.1 vs 26.7 kg/m 2 , P < 0.02), diabetes (15% vs 7.6%, P ¼ 0.035), heart failure (15.7% vs 4%, P < 0.001), and coronary disease (26% vs 16.8%, P ¼ 0.044).They were more likely to have a non-stroke indication for PFO closure, such as platypnea-orthodeoxia or pacemaker/indwelling catheter.
The distribution of aortic distance is represented in Central Illustration.Most patients had a short aortic rim (<9 mm; 197/324, 60.8%), with a mean distance of 5.8 AE 1.8 mm in that group.The patients with a short aortic rim had a numerically higher incidence of atrial septal aneurysms (17% vs 11%) (Table 1), and a longer mean septal aneurysmal excursion (9.5 vs 7.2 mm, P < 0.001).The superior vena cava rim length was similar in both groups, measuring a mean of 17 mm.
Five percent of patients in both aortic distance groups had an associated atrial septal defect.

PROCEDURAL CHARACTERISTICS AND IN-HOSPITAL
OUTCOMES.A 35 mm device was implanted in most patients in both groups (64% in the short vs 59% in the aortic rim $9 mm group) (Table 2).There were no cases of atrial arrhythmias, major vascular complications, or bleeding events after the procedure.Fewer than 6 patients in both groups were treated intraprocedurally for paroxysmal atrial tachyarrhythmias, with antiarrhythmic drugs or cardioversion.
LONG-TERM OUTCOMES.Over a median of 7 years of follow-up (IQR: 4-10 years), there were no cases of device embolization or erosion requiring cardiac surgery in either group (Table 3).There were 9 allcause deaths in each group (4.6% in the short rim group vs 7.1% in the rim $9 mm, P ¼ 0.30).The incidence of stroke overall was 2.2%, and that of stroke or possible transient ischemic attack was similar in both groups (6.3% in those with aortic rim $9 mm vs 5.1% in those with rim <9 mm, P ¼ 0.60; incidence rate ratio of 1.1 (0.3-5), P ¼ 0.90).

DISCUSSION
In this large single-center series of patients undergoing percutaneous PFO closure with the Amplatzer PFO Occluder, the procedure was successful as defined in our study and safe over a median of 7 years of follow-up, regardless of aortic distance on baseline Stefanescu Schmidt et al Transcatheter PFO Closure With Short Aortic Rims M A R C H 2 0 2 3 : 1 0 0 2 5 7  Transcatheter PFO Closure With Short Aortic Rims TEE.More than half of the cohort of patients presenting for PFO closure had an aortic rim <9 mm, and in that group, there were no device erosions requiring cardiac surgery over 1,394 patient-years of follow-up.
Patients with a longer aortic rim were more likely to have a higher body mass index, and more likely to have a history of congestive heart failure; previous autopsy studies have similarly reported an association between longer aortic rim and higher body surface area. 9Conversely, patients with a shorter aortic rim had a trend toward more frequent atrial septal aneurysms.There were no significant differences in all-cause mortality, stroke or transient ischemic attack in long-term follow-up between groups.After adjusting for baseline risk factors (such as diabetes, coronary disease, and pre-existing heart failure), there were no significant differences in risk of new-onset atrial fibrillation or heart failure.The incidence of postprocedural atrial fibrillation in longterm follow-up (16.2/1,000 person-years, 95% CI 11.5, 22.7) is similar to recently published administrative data 19 and meta-analyses. 7,20Previous analyses have suggested that larger devices are associated with an increased risk of postprocedural atrial fibrillation. 21e incidence of all-cause death in this cohort may  Stefanescu Schmidt et al Administration in 2016; the ASO device is stiffer and has a larger self-centering waist, as opposed to the PFO device which has a thin, short, flexible waist that allows the discs to move independently.Only 2 devices were explanted for cardiac perforation, of which one was an Amplatzer PFO 25 mm device (0.01%) late after the procedure, and one a CardioSEAL device (NMT Medical) (0.05%) that was explanted just hours after the procedure.Two additional Amplatzer devices (type not specified) were removed due to pericardial effusion (which may have represented an erosion, procedural complication, or inflammatory reaction).Explantations due to residual shunt were more common, seen in 12 patients (0.09%; 4 with Amplatzer, 0.04%), in which device malposition or an atrial septal injury was noted and required surgical removal rather than another percutaneous closure.
As of 2019, there were over 100,000 implantations of the Amplatzer PFO device. 25In the 3 cases reported in the literature of erosions linked to the Amplatzer PFO device, the aortic rims were normal in 1, 16 and unknown in the other 2, 14,15 with all 3 cases demonstrating erosions through the atrial wall.The size of the atria and the distance from the PFO to the atrial wall (and subsequent choice of device size), as well as the final position of the device, appear to be more relevant to the risk of erosion than the aortic rim.
Reviewing the safety events reports to the Food and Drug Administration Manufacturer and User facility Device Experience (MAUDE) database for Amplatzer PFO devices through March 31, 2021, there were 3 reports of device erosion requiring surgical removal for the 35 mm device.One patient had injury to the aorta, and the second was reported to have atrial and aortic injury repaired with pledgeted sutures but the device was left in place; there are no details for the third.There were no reported device erosions for the 30 mm, and 3 reports of erosion for the 25 mm device (all 3 injuries to the right atrial wall).The 6 patients who had reported erosions presented with dyspnea or chest pain and were found to have pericardial effusions on echocardiogram; atrial or aortic injury was confirmed at the time of operation for device removal.Three patients presented within 5 days from the procedure, and another 2 within a year.In our series with longer follow-up, there were no patients who required cardiac surgery for an atrial septal repair.
We may not, however, have captured patients who suffered sudden cardiac death events or expired before surgery could be performed.
The risk of erosion from septal occluder devices was raised in a review of adverse events after implantation of the Amplatzer ASD Septal Occluder device; in the initial series of 28 patients, all erosions were noted in the dome of the atrium, near the aorta, and 89% had a deficient aortic root on review. 26actice has moved away from oversizing devices more than 2 mm from the balloon-stretched size of the ASD (which was more commonly done at the time of that study), and focused on an adequate aortic rim size.The ASO device in contemporary series has been shown to be implanted safely in patients with deficient aortic rims (<5 mm) acutely (from the American

College of Cardiology National Cardiovascular Data
Registry IMPACT multicenter registry 27 ) and with a high success rate (99%) and no adverse events over 24 months of follow-up in a recent single-center study of 400 patients. 28e original Instructions for Use of the PFO device were based on the initial experience with the Amplatzer Septal Occluder, though the characteristics of the PFO device make erosions significantly less likely.In particular with PFO devices, a higher incidence of erosion has not been demonstrated with larger devices (35 mm), while a higher incidence of embolization (either during or after the procedure) is known, in particular in patients with an aneurysmal atrial septum.In our experience, as previously reported, 17 there was one device embolization during the procedure in the setting of a hypermobile atrial septum, treated with surgical device removal and PFO closure.There were no other cases of device embolization or erosion.Transcatheter PFO Closure With Short Aortic Rims STUDY LIMITATIONS.We present a single-center analysis, with long-term follow-up from an administrative database, introducing the potential for misclassification bias; the robust patient linkage and completeness of data in follow-up have been previously described.Chart abstraction was done retrospectively, and some data is missing; full echocardiographic data were not available for all patients.Echocardiograms in long-term follow-up were not available, nor was it possible to review charts at the time of death to determine if cause of death, or if there was a possible pericardial effusion or device erosion that may not have been detected.Pericardial effusions that may be associated with device erosion (or a non-device related cause, or inflammatory or allergic reaction) were also not captured, and may limit the sensitivity of this study.We choose to focus on severe erosions requiring device surgical removal; device erosions associated with sudden death would not be captured in this registry.As the incidence of erosion is low in the population overall (previously reported at 0.2%, 5 though as low as 0.02% in other series 24 ), we would expect to see only 1 event, or perhaps less, in a cohort of our size.If a short aortic rim carried a clinically relevant higher risk of erosion, we would anticipate a noticeable number of events as more than half of the implanted patients in our cohort had a short aortic rim.Our long duration of follow-up was designed to capture both acute and late events, as has been previously described.

CONCLUSIONS
Percutaneous closure of a PFO using the Amplatzer PFO Occluder was performed safely in a large singlecenter series, including in a majority of patients with a PFO to aortic distance measuring <9 mm, without surgical device removal or repair for erosions in long-term follow-up.Our experience is consistent with the very rare reports of erosion requiring explantation of <1:1,000.

FUNDING SUPPORT AND AUTHOR DISCLOSURES
STUDY POPULATION.A retrospective chart review of all PFO closures performed at the Peter Munk Cardiac Center (Toronto General Hospital, Toronto, Canada) between 2007 and 2017 was performed.Patients were candidates for PFO closure after a cryptogenic stroke or for a non-stroke presentation such as decompression illness during scuba diving, indwelling right heart lines or pacemaker, and platypnea-orthodeoxia syndrome.Patients with cryptogenic stroke were evaluated by a neurologist for other possible causes of stroke, and had magnetic resonance imaging of the brain, imaging of the aorta and head and neck vessels, at least a 48-hour monitor to evaluate for atrial arrhythmias (most patients had at least 2 weeks of monitoring as practice evolved), and a transesophageal echocardiogram (TEE) to evaluate for source of embolism.Presence of a PFO and right to left shunting with a positive bubble study was confirmed by TEE prior to the procedure.

A 3 (
B B R E V I A T I O N S A N D A C R O N Y M S ASD = atrial septal defect PFO = patent foramen ovale TEE = transthoracic echocardiogram Stefanescu Schmidt et al J A C C : A D V A N C E S , V O L . 2 , N O . 2 , 2 0 2 the vast majority 25 mm or 35 mm, based on the size of the right atrial disc) was chosen by the operator based on the preprocedural TEE images, focusing in particular to the mobility of the septum, with larger devices chosen for patients with a mobile interatrial septum, and height of the atrial septum to accommodate the disc size.Balloon sizing was not routinely performed.Patients were discharged on the day of the procedure if there were no complications.They had a follow-up visit with a transthoracic echocardiogram and a bubble study planned for 3 months after the procedure.LINKAGE TO ADMINISTRATIVE DATABASES FOR LONG-TERM OUTCOMES.Long-term outcomes on the study population were obtained using linkage of the study clinical information to Ontario health administrative databases at ICES.The deterministic linkage was performed using Ontario health card numbers.Patients who were referred from other provinces and not followed in the province of Ontario and those with non-linkable records were excluded.ICES databases capture comprehensive information on emergency and inpatient admissions, outpatient physician visits, home and long-term care among others.Complete echocardiograms as part of longterm follow-up were not available in this registry, nor can individual complete medical records be accessed.Event review for etiology ascertainment and adjudication (presence of pericardial effusion at the time of atrial arrhythmia diagnosis, or device

FIGURE 1
FIGURE 1 Measurement of Aortic and Superior Vena Cava Distance, and Atrial Septal Excursion by Transesophageal Echocardiogram

CENTRAL
ILLUSTRATION Long-Term Outcomes After PFO Closure for All Indications Stefanescu Schmidt AC, et al.JACC Adv.2023;2(2):100257.Long-term outcomes in the short (<9 mm) and longer ($9 mm) aortic rim groups.Hazard of new atrial fibrillation diagnosis in long-term follow-up is adjusted for baseline age, sex, pre-existing heart failure and diabetes.Indications for PFO closure in this cohort include cryptogenic stroke, treatment of platypnea orthodeoxia syndrome, prophylaxis of decompression illness for scuba diving, and prevention of paradoxical embolism in selected patients with permanent indwelling lines or pacemaker wires.
are n (%) or n. a 30-mm devices used in 2.2% of all patients.b Exact number cannot be reported due to small cells.J A C C : A D V A N C E S , V O L . 2 , N O . 2

J
A C C : A D V A N C E S , V O L . 2 , N O . 2 , 2 0 2 3Transcatheter PFO Closure With Short Aortic Rims M A R C H 2 0 2 3 : 1 0 0 25 7

TABLE 1
Baseline Characteristics Values are n (%) or mean AE SD. a Information missing in 52 patients (16%) overall, with similar proportion in both aortic rim groups.b Exact number cannot be reported due to small cells.c RoPE score calculated without information on brain imaging as data were not available for all patients; the score presented therefore ranges between 0 and 9. d Missing data in <10 patients.BMI ¼ body mass index; DVT ¼ deep venous thrombosis; PE ¼ pulmonary embolus; PFO ¼ patent foramen ovale; RoPE ¼ Risk of Paradoxical Embolism score.

TABLE 2
Procedural Outcomes -initiated research grant from Abbott on long-term outcomes of PFO transcatheter closure in adults.ICES is supported in part by a grant from the Ontario Ministry of Health and Long-Term Care.The opinions, results, and conclusions are those of the authors and no endorsement by the Ministry of Health and Long-Term Care or by ICES is intended or should be inferred.Parts of this material are based on data and information compiled and provided by Canadian Institute for Health Information (CIHI).However, the analyses, conclusions, opinions, and statements expressed herein are those of the author, and not necessarily those of CIHI.Dr Lee is the Ted Rogers Chair in Heart Function Outcomes, a joint Hospital-University Chair of the University Health Network and the University of Toronto.Dr Horlick is supported by the Peter Munk Chair in Structural Heart Disease Intervention.All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.