Study design and patient population

We conducted a retrospective study in all patients with PFO-related stroke underwent PFO closure between January 2018 and December 2023, in the interventional cardiology unit of the Dijon University Hospital.

We included patients fulfilling the European definition of PFO-related stroke, corresponding to a selected subgroup of ESUS patients in whom no alternative stroke etiology (no identifiable cause after excluding carotid or intracranial artery stenosis, atrial fibrillation, intracardiac thrombus, and atheromatous plaque at the aorta) was identified and PFO was considered as the most plausible mechanism.

PFO closure was subsequently validated by a multidisciplinary team comprising an interventional cardiologist, a cardiologist specialized in echocardiography, a neurologist, and a radiologist, based on clinical and echocardiographic data, as described below. Obviously, all patients had a PFO with closure criteria corresponding to the guidelines to previous studies with an associated atrial septal aneurysm (ASA) or large interatrial shunt.

Preprocedural diagnostic work-up

All data were extracted from electronic medical records of the University Hospital of Dijon. The following baseline patient characteristics were collected: demographics (age, sex), cardiovascular risk factors (smoking, hypertension, overweight, diabetes, dyslipidemia), comorbidities (previous stroke, MI, deep vein thrombosis or pulmonary embolism, migraine), and medications such as estrogen-progestin contraceptives. We also collected data on the type of qualifying event (stroke or transient ischemic attack (TIA)), NIH Stroke Scale Score (NIHSS), RoPE score and CHA₂DS₂-VASc (calculated before the stroke event) score.

Screening for PFO was systematically performed as part of the standardized etiological work-up of ischemic stroke, particularly in patients younger than 60 years. Initial evaluation relied on transthoracic echocardiography (TTE) with contrast (agitated saline), including a Valsalva maneuver when feasible. In cases of suboptimal acoustic window or when a more detailed anatomical characterization was required, TEE was performed. PFO morphology and complexity were assessed based on established echocardiographic criteria, including the length and width of the PFO tunnel, the presence of an ASA or associated atrial septal defect (ASD), the thickness of the septum secundum, and associated anatomical features such as enlargement of the ascending aorta. In accordance with the CLOSE trial criteria [8], shunt severity was assessed by visual counting of microbubbles on contrast echocardiography and was considered significant when more than 30 microbubbles appeared in the left atrium within three cardiac cycles after opacification of the right atrium, whether the shunt was spontaneous or induced by a Valsalva maneuver, in accordance with previous studies. Atrial septal aneurysm was diagnosed on the basis of a septum primum excursion greater than 10 mm. Others standard echocardiographic parameters were collected, including left ventricular ejection fraction, left atrial volume index or potential valvulopathy. However, transcranial Doppler was not routinely used in our diagnostic pathway.

Procedural workflow and procedure description

Percutaneous PFO closure were performed in Philips® cath lab and ICE imaging was performed using the 9-F ViewFlex® Xtra ICE catheter (Abbott) with the Zonare ViewMate® Ultrasound Console (Abbott).

The Amplatzer® PFO Occluder (Abbott), GORE® Septal Occluder (Gore Medical and Associates Inc., Newark, DE, USA), and CARDIA Ultrasep® PFO Occluder (Cardia) were the potential devices implanted in all patients.

All ICE-guided PFO closure procedures were performed under local anesthesia. Sometimes, slight sedation by benzodiazepines, hypnosis or virtual reality headset has been used to improve patient comfort. The procedures were performed by two experienced interventional cardiologists, one for deploying the device and the other for handling ICE.

The right and sometimes left femoral veins were punctured to introduce sheaths for the PFO closure device and ICE catheter. Most of the procedures in our center have been realized with a single venous access to limit the number of blood punctures, using the "Two-In-One Technique", minimizing vascular punctures and facilitating early discharge, as previously described [17]. Briefly, this venous access strategy involves ultrasound-guided puncture of the common femoral vein, followed by a small skin incision and pre-closure using a suture-based vascular closure device. An initial introducer sheath is inserted, through which two 0.035-inch guidewires are advanced. The sheath is then removed, allowing placement of the required introducer sheaths over each guidewire, typically a 10-F sheath for the ICE catheter and an 8- or 9-F sheath for the PFO delivery system. In cases of pre-closure failure, an additional suture-based closure device or manual compression was used to achieve hemostasis. After femoral venous access was obtained using the selected technique, intravenous heparin was administered to achieve an activated clotting time greater than 250 s, in accordance with current recommendations.

ICE setup and catheter advancement into the right atrium were systematically performed before introduction of the PFO delivery sheath, allowing early optimization of image quality and definition of a stable septal reference plane. The ICE catheter was positioned in the mid right atrium (under fluoroscopic guidance) with slight posterior flexion and clockwise rotation to obtain the optimal “PFO view,” enabling detailed visualization of the interatrial septum and PFO anatomy. Particular attention was paid to septal elongation during imaging in order to avoid underestimation of anatomical dimensions (for the sizing of the device).

Under ICE and fluoroscopy guidance, the PFO was then crossed using a 0.035-inch guidewire and a multipurpose catheter, with the guidewire advanced into the left upper pulmonary vein to provide stable support. The delivery sheath was subsequently advanced over the guidewire in parallel to the ICE catheter.

Device deployment was performed predominantly under ICE guidance. The left atrial disc was first released and positioned against the septum, followed by gentle traction under continuous ICE visualization to ensure appropriate septal capture before releasing the right atrial disc. Device stability was assessed using standardized push-and-pull maneuvers, with ICE confirmation of correct disc alignment on both sides of the septum prior to final release. ICE was also used to verify final device position and exclude immediate complications. Before final release, all tension on the delivery system was systematically released to allow assessment of the device’s final resting position and stability under physiological conditions. At the end of the procedure, the catheter is removed and vascular closure was achieved using a suture-based closure device (Perclose ProGlide®, Abbott Cardiovascular) or manual compression. The step-by-step procedural workflow is illustrated in Fig. 1, combining representative angiographic and ICE views.

Fig. 1

Representative angiographic and intracardiac echocardiography views illustrating the step-by-step workflow of minimalist ICE-guided PFO closure. Panel A. Angiographic views. A1. Initial angiographic position of the ICE catheter in the right atrium, serving as a reference for subsequent procedural steps. A2. Fluoroscopic visualization of PFO crossing using a multipurpose catheter and a 0.035-inch guidewire, with advancement and stabilization of the guidewire into the left upper pulmonary vein. A3. Positioning of the delivery sheath across the interatrial septum prior to device advancement, with guidewire withdrawal testing to confirm appropriate left atrial positioning under ICE guidance. A4. Device deployment under fluoroscopy, performed by gradual sheath retraction to allow controlled release of the device. A5. Stability assessment using standardized push-and-pull maneuvers to evaluate device anchoring and mobility before final detachment. A6. Final angiographic appearance after complete device release, confirming stable device position across the interatrial septum. Panel B. Intracardiac echocardiography views. B1. Initial ICE visualization of the patent foramen ovale, providing detailed assessment of interatrial septal anatomy and baseline PFO morphology. B2. ICE-guided visualization of guidewire passage across the PFO, confirming correct traversal of the interatrial septum. B3. Intraprocedural ICE-based anatomical measurements for device sizing, including assessment of PFO tunnel length, septal thickness, and the presence of atrial septal aneurysm. B4. ICE-guided device deployment with visualization of left atrial disc release and subsequent septal capture. B5. ICE confirmation of device stability during push-and-pull maneuvers, ensuring appropriate disc alignment and anchoring on both sides of the septum. B6. Final ICE assessment after device release, demonstrating correct device position, adequate septal apposition, and absence of immediate complications

Device selection and sizing

Device size selection was primarily guided by real-time intraprocedural ICE, which provided direct visualization of the interatrial septum and allowed dynamic assessment of PFO anatomy, enabling individualized device sizing rather than relying solely on preprocedural measurements. The choice of device size was based on a combination of anatomical criteria, including the length and width of the PFO tunnel, the presence and extent of an ASA, the thickness and width of the septum secundum, the presence of potential multiple septal fenestrations, and septal mobility. During the procedure, ICE was used to simulate device positioning and to assess disc alignment, septal capture, and stability prior to final release. The selected device aimed to ensure optimal anchoring, procedural safety, and the absence of significant residual interatrial shunt.

Practical considerations and radiation minimization

ICE played a central role throughout the procedure, allowing device deployment with minimal reliance on fluoroscopy. Fluoroscopy was primarily and only used for catheter advancement, guidewire manipulation, and final confirmation of device stability prior to detachment. Several practical points were identified during adoption of this minimalist strategy. Adequate septal elongation during ICE imaging was systematically sought to avoid underestimation of PFO dimensions. Excessive device mobility after deployment was considered suggestive of insufficient anchoring and prompted reconsideration of device size. Device recapture and repositioning were safely performed under combined ICE and fluoroscopic guidance when needed, using controlled push-and-pull maneuvers and systematic release of tension to assess final device position. Radiation exposure was minimized by prioritizing ICE for anatomical guidance and device deployment, resulting in limited fluoroscopy times despite the absence of TEE.

Postprocedural management and follow-up

All patients received already aspirin before device implantation, due to preexisting stroke. All patients received the addition of clopidogrel 75mg from the device implantation, to obtain dual antiplatelet therapy for a minimum duration of 3 months.

A 12-lead electrocardiogram and TTE was obtained after the index procedure, allowing to check the correct position of the device and to ensure the absence of pericardial effusion, before hospital discharge.

During the in-hospital period, the following safety criteria were collected: periprocedural complications, including access-related complications (bleeding, false aneurysm, deep vein thrombosis, arteriovenous fistula), pericardial effusion, device migration, supraventricular tachycardia and vital status. Early migration or pericardial effusion was systematically checked on pre-discharge echocardiography. Supraventricular arrhythmias were detected during intraprocedural monitoring, on the discharge electrocardiogram, or in the presence of symptoms.

Vascular complications included bleeding, graded according to the Bleeding Academic Research Consortium (BARC) classification, pseudoaneurysm, deep vein thrombosis, and arteriovenous fistula. Doppler ultrasound was performed only in the presence of abnormal findings on pre-discharge clinical examination.

Patients were contacted by phone at 3 months. Data was collected regarding vital status, current treatments, cardiovascular events (ischemic stroke recurrence, myocardial infarction, heart failure hospitalization, atrial fibrillation), vascular complication, or any hemorrhagic event. A questionnaire was used to ensure a complete satisfaction of the patient and evaluate the feeling of the use of ICE catheter over TTE. Patient comfort was assessed by asking patients to rate their comfort from 0 (unbearable) to 10 (no discomfort) after the procedure.

Study endpoints

Procedural success was defined as successful implantation of the PFO closure device with adequate stability and without device embolization. The presence of a moderate residual shunt immediately after the procedure was not considered procedural failure, as spontaneous resolution after device endothelialization may occur.

Study endpoints included early discharge defined as discharge within 24 h of admission, outpatient care defined as same-day discharge, patient-reported procedural comfort, and clinical outcomes assessed at three months.

Safety endpoints included in-hospital periprocedural complications such as access-related vascular complications, pericardial effusion requiring intervention, device migration, supraventricular arrhythmias, and all-cause mortality.

Statistical analysis

Statistical analyses were performed using SPSS software (Statistical Package for the Social Sciences, version 12.0.1, IBM Inc, USA). Continuous variables were expressed as mean ± SD or median with interquartile range (IQR). Categorical variables were reported as counts and percentages.

Ethics

This study was a retrospective observational analysis based exclusively on routinely collected clinical data and involved no additional intervention or modification of patient management. According to French public health law, such non-interventional retrospective studies are exempt from Institutional Review Board approval. This exemption is consistent with institutional policy at Dijon University Hospital. The study was conducted in accordance with the principles of the Declaration of Helsinki. All patients were informed that their anonymized clinical data could be used for research purposes, and no patient objected to such use. In addition, verbal consent was obtained during follow-up telephone contact for participation in the patient-reported outcome questionnaire and for publication of the results.