Atrial septal abnormalities are common congenital lesions remaining asymptomatic until adulthood in a great number of patients. The most frequent atrial septal defects in adults are ostium secundum atrial septal defect (ASD) and patent foramen ovale (PFO), both approachable by transcatheter closure using device implantation. The first non-surgical ASD closure was performed in 1975 by Mills and King.1 In 1983, Rashkind developed a double-disc system:2 the ASD is sealed on both sides by the Rashkind umbrella (the left disc deployed on the left side by the delivery catheter after crossing the defect and the right disc deployed on the right side of the defect after removing the delivery catheter through the defect) (see Figure 1). Endothelial covering of both discs will make the closure of the defect complete after several weeks. After the Rashkind umbrella, consisting of a miniature grappling hook device, several systems were introduced from the late 1980s, based on the concept of a double umbrella: Clamshell occluder, buttoned device, ASDOS, Angel Wing, CardioSEAL® and STARFlex® device. All these prostheses consisted of tissue patches attached by metallic arms or hooks, non-recapturable and leading to a risk of perforation and/or fracture. In 1998, Amplatz developed the Amplatzer® septal occluder, a self-centring device consisting of two discs made of nitinol wire mesh that are linked together by a short connecting waist (see Figure 1), easily repositioned and recaptured into smaller delivery catheters. After Bridges published, in 1992, encouraging results of PFO closure in prevention of cryptogenic stroke in young patients,3 and because the procedure has been dramatically simplified using later generations of devices, a great number of ASD and PFO closure procedures have been performed and published since the beginning of the 21st century. Currently, the technique of transcatheter interatrial septal defect closure has become recognised as a common alternative to conventional surgical technique.
Implantation Procedure
The procedure is performed under fluoroscopic and echocardiographic guidance in a still patient. Transoesophageal echo requires general anaesthesia whereas intracardiac echocardiography guidance allows a more simple procedure under sedation. Antibiotics should be administered at the beginning of the procedure.
The femoral vein is punctured, allowing a multipurpose catheter (5 or 6 French) to be advanced to cross the defect by orientation of the tip to the left and posterior. After pressure measurement, the catheter is advanced in the left upper pulmonary vein. Heparin should be given after crossing the defect and obtaining a safe and stable position of the catheter on the left side of the heart. A sizing balloon is advanced on a stiff exchange wire. Sizing of the defect can be performed by echocardiography and/or fluoroscopic balloon sizing. The safest approach is probably to use both and to measure the waist on the balloon to select the appropriate size of the device. Echocardiography has the advantage of providing the ‘stop flow’ diameter, using colour Doppler to evaluate the size of the waist when the shunt disappears; this technique decreases the risk of oversizing as compared with the stretched diameter. Device selection is performed depending on the sizing and the characteristics of the septum. For ASD closure, a device size 2 mm greater than the stop flow diameter is usual; for PFO, the size of the right disc of the device should be twice the size of the right opening of the tunnel (see Figure 2). Once the device has been chosen, the delivery sheath is advanced over the stiff exchange wire into the left atrium. The wire and the dilator are removed very slowly from the sheath, positioned in a dependent position, allowing free flow of blood coming out from the sheath, in order to reduce the risk of air embolism. The device is advanced into the delivery catheter to deploy the left disc on the left atrium; the sheath is pulled back in the right atrium allowing the right disc of the device to be deployed in the right atrium. The correct device position is assessed by both echo and fluoroscopy (on a left oblique cranial projection): the discs should be parallel to each other and separated from each other by the atrial septum. Echo should also evaluate the non-interference of the device with the surrounding structures and the absence of pericardial effusion, and exclude a significant residual shunt using colour Doppler. For some devices, a stability manoeuvre should be performed before release of the device. After device release, the position of the device is again checked by echo and fluoroscopy. After the procedure, aspirin (low-dose <500 mg/day) and endocarditis prophylaxis are recommended for six months. Chest X-ray, electrocardiography and transthoracic echocardiography are performed at day 1, at 1, 6, and 12 months, then yearly thereafter.
Devices
Several devices are currently available (see Figure 3); we describe those most frequently used in clinical practice.
The Amplatzer Septal Occluder (AGA Medical Corporation: Plymouth, Minnesota). The device is self-expanding, made of nitinol wire mesh, which forms two discs connected by a 3–4 mm thick waist, with Dacron® polyester patches inside the discs. The easy manipulation, repositioning and recapture of this ‘superelastic’ device have made the Amplatzer occluder the most popular occlusion device delivered using small-diameter sheaths. The main indication for this device is large and complex ASD because it is easy to retrieve and replace it in case of complex anatomy. In case of ‘Swiss cheese’ or multifenestrated septum with several holes, small and close to each other, a cribriform septal occluder is a good option.
The HELEX® Septal Occluder (Gore: Flagstaff, Arizona). The device is a non-self-centring double-disc device of nitinol with a helical pattern and a curtain of expanded polytetrafluoroethylene (ePTFE). This device is very flexible and atraumatic, very interesting in the case of a large-sized defect in small cavities, with a smooth contact between the discs and the cardiac structures. Patches of ePTFE have excellent biocompatibility with rapid endothelialisation.
The STARFlex device (NMT Medical: Boston, Massachusetts). This is a double umbrella of polyester fabric supported by four wire spring arms and a centring nitinol coil spring, delivered by a single-operator and simple delivery catheter. The BioSTAR® septal occlusion system is a septal implant which is bioresorbable by incorporation of biological material on the STARFlex framework, with a decrease in thrombogenicity and reduced obstruction access across the interatrial septum.
The Premere™ PFO occluder (St Jude Medical: Maple Grove, Minnesota). This self-expanding dual-anchor arm occlusion device is specifically designed for PFO closure in the case of a long tunnel. The distance between the left atrial anchor (without fabric, for less thrombogenicity on the left side) and the right anchor is adjustable in order to adjust to the anatomy, because the length of the PFO tunnel can vary from 2 to 20 mm.
The Cardia Intrasept™ device (Cardia: Eagan, Minnesota). This has been designed to accommodate anatomical variations in interatrial septum thickness and mobility, thanks to the articulation made in titanium between the sails attached to a nitinol frame. The unique design of the centre post allows each sail to articulate independently through 360°. This device is particularly indicated in the case of a thick septum or an aneurysmal, hypermobile septum.
The Occlutech Figulla® device (Occlutech: Helsingborg, Sweden). This device has two nitinol discs connected by a 3 mm waist, very similar to the Amplatzer device; it has only one right atrial side central pin, reducing the risk of trauma or clot in the left atrium, and reduced meshwork material as compared with the Amplatzer device.
Clinical Results
During the last decades, a considerable number of transcatheter PFO and ASD closures were performed and published with high procedural success (>95 %), effective shunt closure by device implantation, a low rate of procedural complications (<3 %) and very rare major procedural complications.4–6 Transcatheter ASD closure is now considered a valuable alternative to surgical treatment, with a similar closure success rate, lower complication rate and a shorter length of hospital stay.7 Several studies demonstrated clinical efficacy via relief of heart failure symptoms and haemodynamic improvement observed in more than 90 % of cases after ASD closure.8–11 In the series of PFO closures as a preventive treatment for recurrent cryptogenic stroke, a low incidence of recurrent thromboembolic events was reported, ranging between 0 and 3 % per year.12–14 During long-term follow-up, several late complications were reported: thrombus on the left disc was reported with an incidence ranging from 0.5 to 2 %, usually resolved with anticoagulant therapy;6–15 late aortic erosion was observed after Amplatzer device implantation with an incidence of 0.1 % and death in 0.01 % of cases;16 endocarditis, embolisation and pericardial effusions are also described. If these complications are very rare, surgical treatment for complications of transcatheter closure of interatrial septal defect resulted in mortality ranging from 2.6 to 5.4 %.17,18
Current Indications for Transcatheter Atrial Septal Defect and Patent Foramen Ovale Closure
Atrial Septal Defect
Transcatheter closure of ASD is indicated in the case of significant intracardiac shunt with signs and/or symptoms of right heart failure, pulmonary hypertension or presumed paradoxical embolism.19 Exclusion criteria are size of defect >40 mm, Eisenmenger syndrome, pulmonary vascular resistance >5 Wood units, not reversible, and >2/3 peripheral vascular resistance. European Society of Cardiology guidelines state that device closure is the method of choice in case of ostium secundum ASD closure when applicable (class I, level of evidence C).
Patent Foramen Ovale
The main indication for PFO closure is the prevention of recurrent paradoxical embolic cryptogenic stroke in younger patients (<55 years) after an extensive work-up including a complete neurological examination and screening blood tests for thrombophilia. Cryptogenic stroke is diagnosed by a neurologist, in the case of ischaemic stroke, and in the absence of abnormal results of several tests including cerebral magnetic resonance imaging (except for ischaemic lesions), angio-magnetic resonance imaging of the circle of Willis, echo Doppler of extracranial cervical arteries, 12-lead electrocardiography, 24-hour electrocardiographic Holter monitoring and transoesophageal echocardiography (except for PFO).20 PFO closure was also proposed in the prevention of decompression illness in divers and to treat platypnea-orthodeoxia syndrome. Finally, prevention of migraine by PFO closure was investigated in a randomised study21 without clinical evidence of benefit as compared with medical treatment.
Since the early report of Bridges,3 interest in treatment of paradoxical embolism by transcatheter PFO closure has increased. Several observations confirmed the concept of a thrombus trapped in the PFO tunnel and able to embolise to the brain, as in the first report by Cohnheim in 1877. Retrospective analysis suggested an association between PFO, cryptogenic stroke and young age. Some questions are unresolved: cryptogenic stroke is a complex diagnosis, not just related to PFO; a high percentage of the young and healthy population has a PFO without stroke; the association of both pathologies, PFO and cryptogenic stroke, may be incidental and is no proof that PFO is the cause of stroke.
Controversial data are available in the literature about PFO closure for cryptogenic stroke prevention. There is a need for randomised trials comparing medical treatment and PFO closure in the setting of recurrent cryptogenic stroke prevention. But patient selection criteria, duration of follow-up, definition of cryptogenic stroke, residual shunt after the procedure, different properties of the devices and antithrombotic treatment remain unclear and make the job very complicated. Despite all these difficulties, five randomised trials will compare PFO closure and medical therapy to provide some answers (see Table 1). The first of these was presented during the last American Heart Association meeting: Furlan presented the results of the CLOSURE I study comparing the STARFlex device for PFO closure with medical therapy (aspirin and/or warfarin) in preventing recurrent stroke or transient ischaemic attack (TIA) among 900 patients. This study (not yet published) showed a higher rate of complications in the intervention group with a similar efficacy in preventing stroke or TIA in both groups. But after this first randomised trial, there are more questions than answers: why was TIA included instead of a restriction to cryptogenic stroke? The definition of TIA remains controversial; why was there a high rate of atrial fibrillation in the device group? Is the design of the STARFlex responsible for the clots, atrial fibrillation and residual shunts observed after the procedure? How can we explain this unacceptable and unusually high rate of procedural complications?
In 2011, there is a lack of robust scientific evidence that PFO closure is a preventive therapeutic option for recurrent cryptogenic stroke. The guidelines of the American Heart Association/American Stroke Association22 state that there are insufficient data to make a recommendation regarding PFO closure in patients with stroke and PFO (class IIb, level of evidence C).
There is considerable evidence supporting an association between PFO and cryptogenic stroke in young patients. This is consistent with a causal relationship, but does not prove causality. It is the responsibility of cardiologists to perform the procedures of transcatheter interatrial septal defect closure using an experienced team and skilled operators in order to reduce procedural complications and improve procedural results in terms of residual shunt and device selection. It is the responsibility of neurologists to identify patients with a high probability of causal relationship between PFO and cryptogenic stroke. Finally, it is the responsibility of both neurologists and cardiologists to include these patients in randomised trials to obtain high statistical power on a large sample of patients and elucidate the question.
Conclusions
Transcatheter interatrial septal defect closure is a safe and effective therapeutic option in patients with recognised indications, when performed by an experienced team and skilled operators. Results will be improved if the operator has knowledge of the characteristics of several devices and takes into account the anatomy of the heart for device shape and size selection. A good imaging technique (transoesophageal or intracardiac echocardiography) is helpful in optimising the result. Follow-up and treatment after the procedure are of paramount importance for early detection and avoidance of late complications. If the indications for transcatheter ASD closure are now well accepted, there is a need for large randomised trials conducted to identify the type of cryptogenic stroke to be treated and prevented by PFO closure.