Real-World Patent Foramen Ovale Closure in Japan　― Results From 1-Year Follow-up of the Amplatzer^TM PFO Occluder Japan Post-Marketing Surveillance Study ―

Abstract

Background: The Amplatzer^TM PFO Occluder was approved for marketing in Japan in May 2019, and the Amplatzer PFO Occluder Japan Post-Marketing Surveillance (PFO Japan PMS) study started in December 2019. This analysis presents clinical outcomes of study patients through 1 year of follow-up.

Methods and Results: PFO Japan PMS is a prospective single-arm multicenter clinical study. Eligible patients were indicated for patent foramen ovale (PFO) closure and underwent an implant attempt with the Amplatzer^TM PFO Occluder, with no age restrictions. PFO closure was evaluated at 1 year via a bubble study, and patients will be followed for 3 years. From December 2019 to July 2021, 500 patients were enrolled across 53 sites. The mean (±SD) patient age was 52.7±15.4 years, with 29.8% of patients aged >60 years. Low adverse event rates were observed through 1 year of follow-up, including atrial fibrillation (2.4%; predominantly transient and within the first 30 days) and ischemic stroke (0.6%). Among patients in whom a 1-year bubble study was performed, a high rate (91.5%) of clinically relevant PFO closure (<20 bubbles) was achieved.

Conclusions: Through 1 year of follow-up in this real-world Japanese study with 30% of patients aged >60 years, a high degree of closure was achieved with the Amplatzer^TM PFO Occluder, along with low rates of atrial fibrillation, ischemic stroke, and overall adverse events.

Randomized controlled studies have demonstrated that transcatheter patent foramen ovale (PFO) closure using the Amplatzer^TM PFO Occluder is more effective than conventional antithrombotic (AT) therapy in reducing the risk of recurrent stroke.^1,2 The Amplatzer PFO Occluder Japan Post-Marketing Surveillance (PFO Japan PMS) study was initiated in December 2019 following Pharmaceutical and Medical Devices Agency approval of the device in May 2019. The PFO Japan PMS study has completed enrollment, and 30-day outcomes have been reported.³ This analysis presents outcomes of PFO Japan PMS patients through 1 year of follow-up.

Methods

Study Design and Patient Population

PFO Japan PMS is a prospective single-arm non-randomized multicenter clinical study designed to assess the safety and effectiveness of transcatheter PFO closure with the Amplatzer^TM PFO Occluder to reduce the risk of recurrent ischemic stroke (IS) in patients with PFO-associated stroke. The target sample size for the study is 500 patients, with a follow-up duration of 3 years. The study is being performed in accordance with the Declaration of Helsinki and was approved by the ethics committees of each participating institution. The design and methods of this study have been described in detail previously;³ additional information is also provided in the Supplementary Methods. Of particular significance to the present analysis, a transthoracic echocardiogram (TTE) or transesophageal echocardiogram (TEE) was performed at the 1-year follow-up visit, during which the degree of PFO closure was evaluated by agitated saline microbubble assessment (“bubble study”), with shunt grade assessed and reported by the sites. Although there were no study requirements or recommendations regarding AT therapy throughout follow-up, changes to AT medications since the most recent study visit were to be documented, as were the occurrences of any adverse events.

Endpoints

The primary safety endpoint of this study is predefined device- and/or procedure-related serious adverse events through 30 days. The other short-term endpoints are technical success and procedural success (for further details, see Supplementary Methods or Akagi et al.³). The long-term study endpoint, to be evaluated upon completion of the 3-year follow-up, comprises the rates of pulmonary embolism, deep vein thrombosis, IS, and atrial fibrillation (AF) between 30 days and 3 years. The intermediate study endpoint, as reported in this analysis, is effective closure of the PFO at 1 year. Complete closure is defined in the protocol as a Grade 0 (0 bubbles) maximal shunt through the PFO, as assessed by a bubble study at rest and/or during a Valsalva maneuver. Effective closure is defined in the protocol as a Grade 0 or Grade I (1–5 bubbles) maximal shunt. An additional measure of PFO closure was defined for post hoc analysis as a Grade 0, Grade I, or Grade II (6–19 bubbles) maximal shunt; this measure of residual shunt size (<20 bubbles) aligns closely with what has been considered a “small” shunt⁴ and is referred to herein as “clinically relevant closure.”

Statistical Analysis

All enrolled patients were included in the analysis. Baseline patient clinical characteristics were summarized using descriptive statistics. Continuous variables are summarized as the mean±SD and categorical variables are summarized as proportions (n/N). Risk of Paradoxical Embolism (RoPE) score analysis⁵ and PFO-Associated Stroke Causal Likelihood (PASCAL) classification⁴ were performed on all patients with sufficient data available (for further details on these statistical methods, see Akagi et al.³ or the Supplementary Methods).

To be included in the calculation of PFO closure rates at 1 year (complete closure, effective closure, and clinically relevant closure), a bubble study had to have been performed during a Valsalva maneuver via TEE or TTE. If a bubble study was also performed at rest, the maximum shunt grade (at rest and during a Valsalva maneuver) was used to characterize the shunt.

Results

All study results through the 30-day visit have been reported previously;³ the 30-day results considered most pertinent to the present analysis are summarized below, along with reporting of the 1-year results.

Patient Enrollment and Follow-up

In all, 500 patients were enrolled at 53 Japanese sites from December 2019 to July 2021 and these patients will be followed through a 3-year visit (Supplementary Table 1). The number of patients enrolled per site varied from 1 to 50, with a median of 7 enrollments. At the 30-day visit time point, 494 patients were active in the study (5 patients had been lost to follow-up, and 1 patient had been withdrawn due to unsuccessful implant), of whom 483 completed the visit (Figure 1). At the 1-year visit time point, 474 patients were active (an additional 19 patients had been lost to follow-up, and 1 patient had died after the 30-day visit time point), of whom 460 completed the visit.

Figure 1.

Patient flowchart for the Amplatzer^TM PFO Occluder Japan Post-marketing Surveillance (PFO Japan PMS) study. In all, 500 patients were enrolled in PFO Japan PMS (patients were considered enrolled upon insertion of the Amplatzer^TM PFO Occluder delivery system into the body). Of these patients, 483 completed the 30-day follow-up and 460 completed the 1-year follow-up. Reasons for patients not completing follow-up visits are provided within the figure. ^AA device deformation occurred when the left atrial disc of a 30-mm Amplatzer^TM PFO Occluder was opened in the left atrium. The device was withdrawn, and an Amplatzer^TM Multi-Fenestrated Septal Occluder “Cribriform” was subsequently implanted successfully.

Patient Characteristics and Baseline PFO Assessments

Among the study cohort, 29.8% of patients were aged >60 years at enrollment (Table 1). Among the 398 patients with sufficient data to allow for calculation of the RoPE score, the median RoPE score was 6 (interquartile range 5–8); further, 229 (57.5%) of these patients had a RoPE score <7. The maximum baseline shunt grade obtained via TEE and/or TTE (performed at rest and/or during a Valsalva maneuver) was a Grade III (≥20 bubbles) or Grade IV (Opacification: only classified using TTE) in 81.0% of patients (Supplementary Table 2). The mean PFO height was 3.0±1.9 mm, and the mean PFO tunnel length was 9.7±4.4 mm (approximately half the patients (49.0%) had a PFO tunnel length ≥10 mm). An atrial septal aneurysm (ASA) was identified in just over one-quarter of patients (25.7%). Among 394 patients with sufficient data to allow for PASCAL classification, 57.6% had a low RoPE score (<7); however, 85.5% had one or more high-risk PFO features (Supplementary Figure 1). As a result, for the vast majority of these patients, a causal role of the PFO in the patient’s IS was classified as either probable (35.8%) or possible (56.3%); in only 7.9% of patients was causality categorized as unlikely.

Table 1.

Baseline Characteristics (n=500)

Age (years)
Mean±SD	52.7±15.4
Median	53.0
Range	14.0–89.0
Age group (years)
≤30	8.6 (43)
30<age≤40	12.8 (64)
40<age≤50	22.2 (111)
50<age≤60	26.6 (133)
60<age≤70	16.2 (81)
>70	13.6 (68)
Male sex	62.4 (312)
Medical history
None	0.2 (1)
Cerebrovascular accident	98.8 (493)
Deep vein thrombosis	16.6 (83)
Migraine headaches	9.6 (48)
Pulmonary embolism	8.2 (41)
Coronary artery disease	3.0 (15)
Palpitations	1.8 (9)
Chronic obstructive pulmonary disease	1.6 (8)
Other sources of right-to-left shuntA	1.0 (5)
Sinus bradycardia	0.8 (4)
Sinus tachycardia	0.8 (4)
Peripheral vascular disease	0.6 (3)
Major bleeding	0.6 (3)
Dilated cardiomyopathy	0.6 (3)
Congestive heart failure	0.4 (2)
MI within the past 1 month	0.4 (2)
Unstable angina	0.2 (1)
Rheumatic mitral/aortic valvular disease	0.2 (1)
Congestive heart failure with LVEF <30%	0.2 (1)
Mitral or aortic valve vegetation or prosthesis	0.2 (1)
Old MI with LVEF <28%	0.0 (0)
Non-infective thrombotic endocarditis	0.0 (0)
Infective endocarditis	0.0 (0)
Stroke risk factors
None	31.0 (155)
Hypertension	34.8 (174)
Former tobacco user	23.6 (118)
Hypercholesterolemia	22.6 (113)
Hyperlipidemia	11.8 (59)
Current tobacco user	10.8 (54)
Family history of cerebral infarctionB	7.2 (36)
Obesity	7.2 (36)
Diabetes	6.6 (33)
Chronic, persistent or paroxysmal AF/Afl	1.2 (6)
Birth control/hormone replacement therapy	0.8 (4)
RoPE score
Mean±SD (n)	6.1±2.0 (398)
Median [IQR]	6.0 [5.0–8.0]
Range	1.0–10.0
RoPE scores (n=398)
0	0.0 (0)
1	1.3 (5)
2	2.3 (9)
3	5.8 (23)
4	14.6 (58)
5	17.8 (71)
6	15.8 (63)
7	16.3 (65)
8	13.1 (52)
9	8.8 (35)
10	4.3 (17)

Unless indicated otherwise, data are given as % (n). ^AAtrial septal defect and/or fenestrated septum. ^BParent or sibling. AF, atrial fibrillation; Afl, atrial flutter; IQR, interquartile range; LVEF, left ventricular ejection fraction; MI, myocardial infarction; RoPE, Risk of Paradoxical Embolism.

Technical/Procedural Success and 30-Day Primary Safety Endpoint

The rate of technical success was 99.8%, and the rate of procedural success was 98.8% (Supplementary Table 3). Device- and/or procedure-related adverse events meeting primary safety endpoint criteria occurred in only 2 patients in the study (0.4%; 95% confidence interval [CI] 0.05–1.44%). Both events were episodes of asymptomatic transient AF, occurring 1 and 26 days after the procedure (for further details on these events, see Supplementary Results).

Safety Through 1-Year Follow-up

Through 1 year of follow-up, 48 adverse events were reported in 37 (7.4%) patients; of these, 20 were procedure-/device-related adverse events in 17 (3.4%) patients (Table 2). There were no events of device thrombus, device erosion, or device embolization. Fewer patients had adverse events between Days 31–365 (18 patients; 3.6%) compared with Days 0–30 (21 patients; 4.2%; Supplementary Table 4), and much fewer patients had procedure-/device-related adverse events between Days 31–365 (3 patients; 0.6%) compared with Days 0–30 (14 patients; 2.8%). The most common adverse event through 1 year was AF (12 patients; 2.4%), which predominantly occurred within the first 30 days (9/12 patients), and more frequently in patients aged >60 years (7/12 patients; Supplementary Table 5). Ten of the 12 total AF events were confirmed to be paroxysmal and self-limiting, and no patient experienced recurrent events. Finally, all 3 patients who experienced AF after 30 days were aged >60 years.

Table 2.

Adverse Events Through 365 Days

Adverse eventA	No. subjects with eventsB (n=500)	No. events	No. subjects with SAEB (n=500)	SAEs (n)	Procedure- related events (n)	Device- related events (n)	Non-procedure and non-device related events (n)
Air embolism	1 (0.2)	1	1 (0.2)	1	1	0	0
Anemia	1 (0.2)	1	1 (0.2)	1	1	0	0
Atrial fibrillation	12 (2.4)	12	5 (1.0)	5	6	9	2
Atrial flutter	2 (0.4)	2	2 (0.4)	2	0	1	1
BleedingC	1 (0.2)	1	1 (0.2)	1	0	0	1
DVT	1 (0.2)	1	1 (0.2)	1	0	0	1
Edema	1 (0.2)	1	1 (0.2)	1	0	0	1
Hemorrhagic stroke	2 (0.4)	2	2 (0.4)	2	0	0	2
Ischemic stroke	3 (0.6)	3	3 (0.6)	3	0	0	3
TIA	1 (0.2)	1	1 (0.2)	1	1	0	0
Vascular AV fistula	1 (0.2)	1	1 (0.2)	1	1	0	0
Vascular access site complications	1 (0.2)	1	1 (0.2)	1	1	0	0
Acute hepatitis	1 (0.2)	1	1 (0.2)	1	0	0	1
Arteriosclerosis obliterans	1 (0.2)	1	1 (0.2)	1	0	0	1
Asymptomatic cerebral infarction	2 (0.4)	2	2 (0.4)	2	1	1	1
COVID-19	1 (0.2)	1	1 (0.2)	1	0	0	1
Death (cancer)	1 (0.2)	1	1 (0.2)	1	0	0	1
Epidural hematoma	1 (0.2)	1	1 (0.2)	1	0	0	1
Hypopharyngeal perforation	1 (0.2)	1	1 (0.2)	1	1	0	0
Ischemic enteritis	1 (0.2)	1	1 (0.2)	1	0	0	1
Left hemiparesisD	1 (0.2)	1	1 (0.2)	1	1	0	0
Left secondary pneumothorax	1 (0.2)	1	1 (0.2)	1	0	0	1
Left ureteral stone	1 (0.2)	1	1 (0.2)	1	0	0	1
Lumbar compression fracture	2 (0.4)	2	2 (0.4)	2	0	0	2
Microscopic polyangiitis	1 (0.2)	1	1 (0.2)	1	0	0	1
Periapical periodontitis	1 (0.2)	1	1 (0.2)	1	0	0	1
Radicular cyst	1 (0.2)	1	1 (0.2)	1	0	0	1
Seizure	1 (0.2)	1	0 (0.0)	0	1	0	0
Spondylosis	1 (0.2)	1	1 (0.2)	1	0	0	1
Subcutaneous hematoma	1 (0.2)	1	1 (0.2)	1	0	0	1
Urination disorder	1 (0.2)	1	1 (0.2)	1	0	0	1
Total	37 (7.4)	48	31 (6.2)	40	15	11	28

Unless indicated otherwise, data are presented as n (%). ^ANo adverse events were related to study device deficiency/malfunction. One event was related to COVID-19; this event was COVID-19 itself. ^BPatients may have had multiple events. ^CThe site of bleeding was the colon. ^DThis event was believed to be related to an intraprocedural air embolism that occurred during delivery sheath removal. Computed tomography and magnetic resonance imaging on the day of procedure showed no new cerebral infarction. The exact duration of hemiparesis is unknown, but it was confirmed to have resolved within 19 days of the procedure. A serious adverse event (SAE) was defined as an event that: led to death; potentially led to death; required in-patient hospitalization or prolongation of existing hospitalization for treatment; potentially led to a persistent or significant disability or incapacity; led to a congenital disease or anomaly; or another event considered serious per the attending physician’s judgment. AV, arteriovenous; DVT, deep vein thrombosis; TIA, transient ischemic attack.

Three (0.6%) patients experienced IS in their first year of follow-up (Supplementary Table 6). All 3 patients had several IS risk factors, including: males aged >60 years at the time of the procedure; current or former tobacco use; and controlled hypercholerestolemia. For 1 patient, a single episode of self-limiting asymptomatic AF had been observed on an insertable cardiac monitor 36 days after closure. The treating physician decided not to administer anticoagulants (ACs) due to concern over elevated bleeding risk. This patient had an IS 133 days after closure. Considering the low AF burden observed for this patient, along with the presence of several other pre-existing stroke risk factors (male aged >60 years, former tobacco user, hypercholesterolemia, and hyperlipidemia), it is unclear whether the AF was causally related to the stroke.

Echocardiography performed after the recurrent stroke events in all 3 patients demonstrated absence of device-related thrombus; the shunt grade assessed via bubble study during a Valsalva maneuver in 2 patients was Grade 0 and Grade I. Although a bubble study was not performed for the third patient, the site determined the cause of stroke to have been atherosclerosis based on the infarct location as observed on magnetic resonance imaging. Finally, each IS event was determined by the treating physician to have been unrelated to either the device or the procedure. (For further detail on patients with IS, see Supplementary Results.)

In the first year of follow-up, 4 (0.8%) patients experienced hemorrhagic stroke (HS). Of these patients, 3 were on ACs (all direct oral ACs) at the time of the HS, and 1 was on single antiplatelet therapy (SAPT). Two of the patients on ACs were also taking antiplatelet (AP) therapy. Following the HS events, 1 patient on AP+AC stopped their AP medication, another patient on AP+AC discontinued both medications, and the patient on AC only continued with their medication.

All 4 patients who experienced HS had a history of deep vein thrombosis (for whom continued anticoagulation after PFO closure is recommended⁶) and/or multiple ISs (for whom continued anticoagulation may be considered reasonable).⁶ In addition, all patients were aged >60 years with several HS risk factors, including hypertension (2 patients), current or former tobacco use (2), and diabetes (2).

Finally, there were no instances of bleeding that required transfusion or surgical/endovascular intervention through 1 year of follow-up.

There was 1 death that occurred within 1 year of follow-up, with the cause of death reported as cancer (for further details on this patient and event, see Supplementary Results).

PFO Closure at 1 Year

Of the 460 patients who completed the 1-year visit, 424 had an echocardiogram performed, with bubble studies done at rest and/or during a Valsalva maneuver for 241 patients (Supplementary Figure 2). PFO closure rates were derived from the 200 patients for whom a bubble study was done during a Valsalva maneuver; among these patients, clinically relevant closure was achieved in 91.5%, effective closure was achieved in 81.5%, and complete closure was achieved in 65.5% (Table 3).

Table 3.

Residual Shunt and Closure Rates at 1-Year Echocardiography (n=200)

	Rate (n)
Residual shunt
Grade 0	65.5% (131)
Grade I	16.0% (32)
Grade II	10.0% (20)
Grade III/IV	8.5% (17)
Closure
CompleteA	65.5% (131)
EffectiveB	81.5% (163)
Clinically relevantC	91.5% (183)

^AGrade 0 maximal shunt at rest and/or Valsalva. ^BGrade 0 or I maximal shunt at rest and/or Valsalva. ^CGrade 0, I, or II maximal shunt at rest and/or Valsalva.

Association were observed between severity of residual shunt and proportions of ASA and long PFO tunnel (tunnel length ≥10 mm at baseline; Figure 2), as well as larger device size (Table 4). Closure rates were higher for patients without than with ASA (complete closure: 72.6% vs. 44.0%, respectively; effective closure: 87.0% vs. 66.0%, respectively; clinically relevant closure: 94.5% vs. 82.0% respectively), as well as for patients without than with a long PFO tunnel (complete closure: 71.4% vs. 60.8%, respectively; effective closure: 86.7% vs. 78.4%, respectively; clinically relevant closure: 94.9% vs. 89.7% respectively). In addition, closure rates were higher for patients with 25-mm devices than for those with 35-mm devices (complete closure: 68.5% vs. 51.9%, respectively; effective closure: 85.7% vs. 59.3%, respectively; clinically relevant closure: 94.0% vs. 81.5% respectively). This association of larger device size with lower closure rates may be highly influenced by the more challenging PFO anatomy with which the larger devices are intended to be used (large, complex PFOs).

Figure 2.

Relationships between rates of patent foramen ovale (PFO) closure (complete, effective, and clinically relevant) and atrial septal aneurysm (ASA; Left) and PFO tunnel length (Right). Correlations were observed between PFO closure rates and proportions of ASA and a long PFO tunnel (tunnel length ≥10 mm at baseline).

Table 4.

Relationship Between Closure Rate and Device Size

Closure rate (definition)	Device size
	18 mm (n=1)	25 mm (n=168)	30 mm (n=5)	35 mm (n=27)
Complete (0 bubbles)	100.0 (1)	68.5 (115)	20.0 (1)	51.9 (14)
Effective (<6 bubbles)	100.0 (1)	85.7 (144)	40.0 (2)	59.3 (16)
Clinically relevant (<20 bubbles)	100.0 (1)	94.0 (158)	40.0 (2)	81.5 (22)

Data are given as % (n).

There were 4 patients with a 30-mm device who underwent a 1-year bubble study; 2 of these patients had a Grade III residual shunt, but a larger sample size would be needed to make confident inferences regarding this relationship. There was only 1 patient with an 18-mm device who underwent a 1-year bubble study; this patient had a Grade 0 shunt.

AT Medication Use

AT therapy after PFO closure was determined by the heart–brain team at each site, as were any medication changes during follow-up. The medication regimens reported at each study visit through 1 year are summarized in Table 5. Overall, a high proportion of patients were receiving ACs: 61.1% at discharge, 58.6% at 1 month, and 38.8% at 1 year. The use of dual antiplatelet therapy (DAPT) decreased throughout follow-up (32.1% of patients at discharge, 20.1% at 1 month, 1.6% at 1 year), as SAPT increased (6.4% of patients at discharge, 20.3% at 1 month, 29.1% at 1 year). Notably, 30.5% of patients were not on any AT medications at 1 year. Discontinuation of AT medications occurred more often in younger patients: of the 307 patients aged ≤60 years at the time of the procedure who completed the 1-year follow-up visit, 112 (36.5%) were not on AT medications; in comparison, of the 136 patients aged >60 years at the time of the procedure who completed the 1-year follow-up visit, 23 (16.9%) were not on AT medications.

Table 5.

Antithrombotic Medication Usage (Discharge Through 1 Year)

	Discharge (n=499)	1 Month (n=483)	1 Year (n=443)
None	0.4 (2)	1.0 (5)	30.5 (135)
Antiplatelet alone	38.5 (192)	40.4 (195)	30.7 (136)
Single antiplatelet therapy	6.4 (32)	20.3 (98)	29.1 (129)
Dual antiplatelet therapy	32.1 (160)	20.1 (97)	1.6 (7)
Anticoagulant alone	21.2 (106)	27.5 (133)	35.2 (156)
Heparin	0.4 (2)	0.2 (1)	0.0 (0)
Warfarin	2.8 (14)	3.7 (18)	3.8 (17)
DOAC	18.0 (90)	23.6 (114)	31.4 (139)
Other anticoagulant	0.4 (2)	0.0 (0)	0.0 (0)
Antiplatelet plus anticoagulant	39.9 (199)	31.1 (150)	3.6 (16)

Data are given as % (n). DOAC, direct oral anticoagulant.

Important observations can also be made by examining patient-level medication changes at the 30-day visit (Supplementary Table 7) and 1-year visit (Supplementary Table 8). Medication changes were relatively infrequent at 30 days; among the patients who completed the 30-day visit, 72.4% remained on their discharge medication regimen. The most frequent changes at this time point were from DAPT to SAPT (11.6%), and from combined AP/AC to AC alone (8.3%). Medication changes were more common at the 1-year visit; among patients who completed both the 30-day and 1-year visit, 29.5% stopped AT medication entirely at 1 year, 10.5% changed from DAPT to SAPT, and 7.6% changed from AC to SAPT.

Discussion

The PFO Japan PMS study enrolled 500 patients across 53 Japanese sites from December 2019 to July 2021. A previous analysis of 30-day outcomes found that: (1) although almost one-third of study patients were aged >60 years, the vast majority of all patients exhibited high-risk PFO features and were thereby categorized as likely to benefit from PFO closure according to the PASCAL classification; and 2) the procedure was performed safely and successfully, with a low incidence of procedure-related atrial arrhythmias.³

The present analysis reports outcomes through 1 year of follow-up of PFO Japan PMS patients. During this period, low adverse event rates were observed, including a 2.4% rate of AF, and a 0.6% rate of IS. There were no events of device thrombus, device erosion, or device embolization in any patient through 1 year of follow-up. Among patients who underwent 1-year echocardiography with a bubble study during a Valsalva maneuver, a high degree of clinically relevant closure was achieved (91.5%), with effective closure and complete closure observed in 81.5% and 65.5% of these patients, respectively. Notable patterns of AT medication use after PFO closure were also observed. Although medication changes were infrequent at 30 days, there were substantial changes at 1 year, including approximately 30% of patients stopping all AT medications at this time. The use of oral ACs decreased during follow-up but was relatively common at all visit time points (61% of patients at discharge, 59% at 30 days, 39% at 1 year).

The ultimate objective of PFO closure in patients with PFO-associated stroke is to reduce the risk of recurrent IS. In this study, 3 patients experienced an IS within 1 year of closure (0.6%). In the RESPECT (Randomized Evaluation of Recurrent Stroke Comparing PFO Closure to Established Current Standard of Care Treatment) trial, which established the safety and effectiveness of the Amplatzer^TM PFO Occluder, 5 of 474 as-treated device arm patients had an IS within 1 year of closure (1.1%).¹ Thus, although the RESPECT trial excluded patients aged >60 years and the PFO Japan PMS study had ~30% of patients aged >60 years at the time of the procedure, the recurrent IS rate at 1 year was numerically lower in the PFO Japan PMS study. Further, there is low probability of any of the IS events in the present study being associated with PFO. For 1 patient the cause was determined to be atherosclerosis, and follow-up shunt assessments for the other 2 patients revealed complete closure or minimal residual shunt (<6 bubbles). Meanwhile, all 3 patients had several IS risk factors unrelated to PFO (males aged >60 years, current or former tobacco use, and controlled hypercholerestolemia). Each IS event was also determined by the treating physician to have been unrelated to either the device or the procedure.

Another important measure of the performance of a PFO occluder is the ability to mechanistically close the PFO. The guideline-recommended method to quantify the degree of shunt through a PFO is a bubble study performed along with a Valsalva maneuver during an echocardiogram.⁷ Complete closure of the PFO (i.e., no shunt or Grade 0 shunt) is universally defined as 0 bubbles visualized in the left atrium during the first 3 cardiac cycles. Various definitions and criteria have been implemented in different PFO studies to classify the degree of residual shunt in the absence of complete closure; however, the threshold for a “large” shunt is typically ≥20 bubbles.^4,8,9

In the PFO Japan PMS study, complete closure was protocol-defined as Grade 0, and effective closure as Grade 0 or Grade I (in accordance with guidance from Japanese clinical societies, Grade I shunt was defined as 1–5 bubbles). An additional classification was adopted post hoc to describe freedom from large shunt (<20 bubbles), termed “clinically relevant” closure. The resulting closure rates at 1 year in the PFO Japan PMS study were 65.5% complete closure (0 bubbles), 81.5% effective closure (<6 bubbles), and 91.5% clinically relevant closure (<20 bubbles).

In a single-center study, Nakayama et al. reported shunt grades at baseline and 1-year after closure among 106 patients who received the Amplatzer^TM PFO Occluder, with shunts assessed via bubble study during a Valsalva maneuver during TTE.⁸ The proportion of patients with a large shunt at baseline was comparable between the study of Nakayama et al. and the PFO Japan PMS study (80.2% and 81.0%, respectively). At 1 year, the rates of effective closure were also similar in the Nakayama et al. and PFO Japan PMS studies (81.1% and 81.5%, respectively), whereas the rates of complete closure (73.6% and 65.5%) and clinically relevant closure (96.2% vs. 91.5%) were somewhat higher in the study of Nakayama et al.⁸ Possible contributors to the slight differences in closure rates include a relatively small sample size in the study of Nakayama et al., along with greater variability in implanter experience in the PFO Japan PMS study with patients enrolled across 53 clinical sites.

Nakayama et al. also investigated the relationship between residual shunt and various baseline PFO morphological characteristics (presence of ASA, PFO tunnel length, and PFO height) along with implanted device size, and found a significant correlation between the presence of ASA and greater residual shunt.⁸ Similar analyses performed on PFO Japan PMS data revealed that patients with ASA or a long tunnel had greater residual shunt than patients without such features; in addition, patients with a 35-mm device had greater residual shunt than patients with 25-mm devices. It is important to note that the relationship between the use of 35-mm devices and greater residual shunt may be highly influenced by the more challenging PFO anatomy with which this device size is intended (large, complex PFOs).

Excellent safety through 30 days after closure in the PFO Japan PMS study has been demonstrated previously,³ with low rates of procedure-/device-related events (2.8%) and AF (1.8%); through 1 year of follow-up, these rates increased only marginally to 3.4% for procedure-/device-related events and 2.4% for AF. There were also no instances of device thrombus, embolization, and erosion through 1 year. These findings indicate that the safety risk associated with PFO closure with the Amplatzer^TM PFO Occluder is very low, both in the acute post-closure period and through interim follow-up.

In randomized clinical trials and observational studies, PFO closure has long been associated with an increased short-term risk of AF.¹⁰ However, recent analyses have characterized post-closure AF as predictably early (within 30–45 days), transient, and non-recurrent, and found no impact of PFO closure on the long-term risk of AF.^11–15

In the PFO Japan PMS study, a low rate of AF was observed within 30 days, occurring in 9 of 500 (1.8%) patients. Between Days 31 and 365 after closure, only 3 additional patients experienced AF, resulting in a 1-year AF rate of 2.4%. Ten of the 12 total AF events were transient. Notably, although approximately 30% of study patients were aged >60 years, approximately 60% of AF patients belonged to this age group, including all 3 patients with AF between Days 31 and 365; a higher rate of post-closure AF among older patients was also reported by Krishnamurthy et al.¹⁶ Thus, the PFO Japan PMS study demonstrated a low rate of AF through 1 year and provides further evidence of post-closure AF as early, transient, and without recurrence. In fact, this study did not need to consider additional AC therapy or catheter ablation for AF that occurred after PFO closure. However, because the incidence and persistence of AF in the general population increases with age,¹² careful follow-up, including the indication for insertable cardiac monitoring, is needed after PFO closure, especially in the elderly population. Recent research has reported that PFO morphology may also change with age, and PFO closure of the elderly should be examined from various perspectives in the future.¹⁷

The manufacturer recommendation for AT therapy after PFO closure with the Amplatzer^TM PFO Occluder (for patients with no history of venous thromboembolism) is aspirin plus clopidogrel for 1 month after closure, followed by aspirin monotherapy for 5 months; the decision to continue AT therapy beyond 6 months is at the physician’s discretion.¹ (For patients with a history of venous thromboembolism, it is recommended to consider chronic AC instead.¹)

There is no consensus among guidelines on PFO management regarding AT therapy recommendations beyond 6 months after closure. On this matter, guidelines from the Japan Stroke Society, Japanese Circulation Society, and Japanese Association of Cardiovascular Intervention and Therapeutics state that “it is necessary to consider long-term continuation of AT therapy based on the current evidence.”¹⁸ In the device arm of the RESPECT trial, cessation of AT therapy at any point during follow-up was extremely rare, because patients remained on some form of AT therapy for 94% of the entire follow-up duration of this group.¹

Further, the relative safety and effectiveness of AC therapy compared with AP therapy in reducing the risk of recurrent stroke remains unclear. The RE-SPECT ESUS (Randomized, Double-Blind, Evaluation in Secondary Stroke Prevention Comparing the Efficacy and Safety of the Oral Thrombin Inhibitor Dabigatran Etexilate versus Acetylsalicylic Acid in Patients with Embolic Stroke of Undetermined Source) study found that in patients with a recent history of embolic stroke of undetermined cause (ESUS; a stroke classification among which PFO-associated stroke is a major category), dabigatran was not superior to aspirin in preventing recurrent stroke.¹⁸ In addition, although the incidence of major bleeding was not greater in the dabigatran group, there were more clinically relevant non-major bleeding events among patients receiving dabigatran.¹⁹ However, a predefined subanalysis of Japanese study patients found that dabigatran was associated with a lower relative risk of recurrent stroke compared with aspirin in this population.²⁰ For this reason, it is necessary to consider the benefits of continuing direct oral ACs for this cohort as well. In the future, it is likely that tailor-made treatment options will be considered that take into consideration the age and lifestyle of the patient.

Taking this all into consideration, important observations can be made regarding the AT medication regimens observed through 1 year of follow-up in the PFO Japan PMS study. First, AC use was very common throughout this period, with over 60% of patients on AC at discharge, and almost 40% of patients on AC at 1 year. This may be largely due to the substantial number of older study patients, who are more likely to have a history of venous thromboembolism as well as other medical conditions that require AC treatment independently. As noted previously, 3 patients on AC experienced HS within the 1-year follow-up, as did 1 patient on AP therapy.

Second, medication changes at 30 days were very rare, with less than 25% of regimens altered at this time point. Finally, approximately 30% of patients had stopped AT therapy entirely at 1 year, with discontinuation occurring more frequently among younger patients. With such a high rate of cessation at this time point, it will be especially impactful to monitor the outcomes of study patients through the complete 3-year follow-up with respect to their AT medication regimens.

Another important investigation will be of the relationship between recurrent IS and varying degrees of residual shunt. In a single-center study of 1,078 patients with PFO-associated stroke who underwent closure, Deng et al. found an approximate 5-fold increase in the risk of recurrent IS in patients with moderate/large shunt (>10 bubbles) compared with no shunt, but did not observe an increased risk associated with small shunt (1–9 bubbles) compared with no shunt.⁹ The application of such an analysis to PFO Japan PMS data and shunt criteria may provide further valuable insights into the predictive power of residual shunt size with respect to stroke recurrence. In patients with residual shunts identified by a bubble study at 1 year, appropriate AT treatment has not yet been clearly defined. In general, it is thought that AT treatment should be continued in patients with residual shunts, but a consensus on the requirement for this has not yet been established.

Study Limitations

This study did not have an independent clinical events committee to adjudicate the seriousness and relatedness (procedure and/or device) of adverse events, nor did it have an independent core laboratory to assess residual shunt at 1-year echocardiography. These functions were instead performed by the site investigators. Finally, there was poor compliance with regard to the protocol-required shunt assessment at 1 year; this was performed in only 200 of 460 (43.5%) patients who completed the 1-year follow-up visit.

Conclusions

PFO Japan PMS data through 1 year of follow-up indicate that patients treated with the Amplatzer^TM PFO Occluder experience low adverse event rates, including a 2.4% rate of AF and a 0.6% rate of IS. A high degree of clinically relevant closure at 1-year was achieved (91.5%), and relationships were established between residual shunt and the presence of ASA or long-tunnel PFO. Notable patterns in post-closure medication regimens included the cessation of AT therapy in approximately 30% of patients by the 1-year mark. The upcoming 3-year results will provide further insights into the long-term safety and efficacy of the Amplatzer^TM PFO Occluder, the relationship between residual shunt and stroke recurrence, and the long-term impact of different AT medication regimens.

Acknowledgments

The authors thank all the investigators and institutions participating in PFO Japan PMS, as well as Nils Peter Borgstrom, PhD (Abbott), for his contributions to data review and manuscript preparation and Deepika Morishetti (Abbott), Devon Sauerer (Abbott), Joshua Rapkin (Abbott), and Justin Liu (Abbott) for their contributions to data analysis.

Sources of Funding

Abbott designed and funded PFO Japan PMS. Further, Abbott was responsible for selecting and monitoring sites, as well as data management and analysis. The authors had full access to the data and attest to the integrity of the study and the accuracy and completeness of the reported data. No funding was provided for this analysis.

Disclosures

H.H. is a member of Circulation Journal’s Editorial Team. The remaining authors have no conflicts of interest to disclose.

IRB Information

This study was approved by Okayama University Hospital Institutional Review Board (Reference no. 24092024).

Data Availability

The deidentified patient data will not be shared.

Supplementary Files

Please find supplementary file(s);

https://doi.org/10.1253/circj.CJ-25-0115

References

• 1.   Saver JL, Carroll JD, Thaler DE, Smalling RW, MacDonald LA, Marks DS, et al. Long-term outcomes of patent foramen ovale closure or medical therapy after stroke. N Engl J Med 2017; 377: 1022–1032.
• 2.   Lee PH, Song JK, Kim JS, Heo R, Lee S, Kim DH, et al. Cryptogenic stroke and high-risk patent foramen ovale: The DEFENSE-PFO trial. J Am Coll Cardiol 2018; 71: 2335–2342.
• 3.   Akagi T, Hara H, Kanazawa H, Fukui S, Hashimoto Y, Iguchi Y, et al. Real-world patent foramen ovale (PFO) closure in Japan: 30-day clinical outcomes from the Amplatzer^TM PFO Occluder Japan Post-Marketing Surveillance Study. Circ J 2024; 88: 1391–1397.
• 4.   Kent DM, Saver JL, Kasner SE, Nelson J, Carroll JD, Chatellier G, et al. Heterogeneity of treatment effects in an analysis of pooled individual patient data from randomized trials of device closure of patent foramen ovale after stroke. JAMA 2021; 22: 2277–2286.
• 5.   Kent DM, Ruthazer R, Weimar C, Mas JL, Serena J, Homma S, et al. An index to identify stroke-related vs incidental patent foramen ovale in cryptogenic stroke. Neurology 2013; 81: 619–625.
• 6.   Kavinsky CJ, Szerlip M, Goldsweig AM, Amin Z, Boudoulas KD, Carroll JD, et al. SCAI guidelines for the management of patent foramen ovale. J Soc Cardiovasc Angiogr Interv 2022; 1: 100039.
• 7.   Messe SR, Gronseth GS, Kent DM, Kizer JR, Homma S, Rosterman L, et al. Practice advisory update summary: Patent foramen ovale and secondary stroke prevention. Report of the Guideline Subcommittee of the American Academy of Neurology. Neurology 2020; 94: 876–885.
• 8.   Nakayama R, Takaya Y, Akagi T, Takemoto R, Haruna M, Nakashima M, et al. Relationship between patent foramen ovale anatomical features and residual shunt after patent foramen ovale closure. Cardiovasc Interv Ther 2024; 39: 200–206.
• 9.   Deng W, Yin S, McMullin D, Inglessis-Azuaje I, Hung J, Lo EH, et al. Residual shunt after patent foramen ovale closure and long term stroke recurrence: A prospective cohort study. Ann Intern Med 2020; 172: 717–725.
• 10.   Vukadinovic D, Scheller B, Ukena C, Ewen S, Mahfoud F, Bohm M. Device related risk of atrial fibrillation after closure of patent foramen ovale: A systematic review and meta analysis. Clin Res Cardiol 2022; 111: 583–587.
• 11.   Reddy SA, Peverelli M, MacCarthy T, Vibhishanan J, Agarwal S, Duehmke R, et al. Influence of device choice on atrial arrhythmia incidence following percutaneous patent foramen ovale closure. JACC Cardiovasc Interv 2024; 17: 2320–2321.
• 12.   Ravellette KS, Gornbein J, Tobis JM. Incidence of atrial fibrillation or arrhythmias after patent foramen ovale closure. J Soc Cardiovasc Angiogr Interv 2023; 3: 101173.
• 13.   Apostolos A, Tsiachris D, Drakopoulou M, Trantalis G, Oikonomou G, Ktenopoulos N, et al. Atrial fibrillation after patent foramen ovale closure: Incidence, pathophysiology, and management. J Am Heart Assoc 2024; 13: e034249.
• 14.   Skibsted CV, Korsholm K, Pedersen L, Bonnesen K, Nielsen-Kudsk JE, Schmidt M. Long-term risk of atrial fibrillation or flutter after transcatheter patent foramen ovale closure: A nationwide Danish study. Eur Heart J 2023; 44: 3469–3477.
• 15.   Chen JZ, Thijs VN. Atrial fibrillation following patent foramen ovale closure. Stroke 2021; 52: 1653–1661.
• 16.   Krishnamurthy Y, Ben-Ami J, Robbins BT, Sommer RJ. Incidence and time course of atrial fibrillation following patent foramen ovale closure. Catheter Cardiovasc Interv 2022; 100: 219–224.
• 17.   Nakashima M, Takaya Y, Nakayama R, Tsuji M, Akagi T, Miki T, et al. Morphological features of patent foramen ovale compared between older and young patients with cryptogenic ischemic stroke. Circ J 2024; 88: 1398–1405.
• 18.   Iguchi Y, Iwama T, Oki K, Kataoka H, Kawano H, Yamagami H, et al. Guidance on percutaneous closure of patent foramen ovale (PFO) in cryptogenic stroke patients. The Japan Stroke Society, The Japanese Circulation Society, and Japanese Association of Cardiovascular Intervention and Therapeutics, 2nd edition, 2023 chrome-extension. https://pfo-council.jp/docs/publications/guidance_02_240328.pdf (in Japanese) (accessed April 16, 2025).
• 19.   Diener HC, Sacco RL, Easton JD, Granger CB, Bernstein RA, Uchiyama S, et al. Dabigatran for prevention of stroke after embolic stroke of undetermined source. N Engl J Med 2019; 380: 1906–1917.
• 20.   Toyoda K, Uchiyama S, Hagihara Y, Kuwashiro T, Mori T, Kamiyama K, et al. Dabigatran vs. Aspirin for secondary prevention after embolic stroke of undetermined source: Japanese subanalysis of the RE-SPECT ESUS randomized controlled trial. Circ J 2020; 84: 2286–2295.