Transcatheter patent foramen ovale (PFO) closure has been undertaken for many years but only since 2017 has the evidence base become uncontested following several high-quality randomised trials. The most common indication for PFO closure is stroke due to probable paradoxical embolus. The randomised trials suggest a recurrent stroke rate of 1.0–1.5% per year with antiplatelet medical therapy versus 0.5% per year following PFO closure. The average age of patients in these trials was approximately 45 years; therefore, the lifetime cumulative clinical benefit of PFO closure may be considerable, in addition to bringing psychological closure.
In the UK, a commissioned evaluation process was initiated shortly before publication of the anticipated seminal randomised trial results. A total of 973 PFO closure procedures were undertaken in selected centres in the UK over 4 years, with device implantation success of 99%. The recurrent neurological event rate was higher than in the randomised trials but was broadly favourable and PFO services were commissioned in up to 20 UK sites in 2019. In 2018, as a result of the unfortunately timed commissioning exercise, the PFO closure rate in the UK was five per million population, compared with 55 per million in Germany and 52 per million in Italy; the rate is only now starting to increase. The procedure offers lifelong protection from recurrent paradoxical embolisation to young and otherwise healthy individuals.
Indications for Diagnosis
A PFO is only a diagnosis of significant value in the presence of a condition that might warrant its closure. Diagnosing a PFO in otherwise healthy individuals can be unhelpful. A PFO should be sought in the following groups:
- patients aged <60 years with unexplained stroke, transient ischaemic attack (TIA), MI or other systemic embolism;
- patients aged >60 years with recurrent stroke, TIA, MI or systemic embolism that remains unexplained;
- divers with some types of decompression illness; and
- patients with unexplained hypoxic syndromes.
Index of Suspicion for a Patent Foramen Ovale
As a PFO is a common finding, distinguishing between an incidental PFO and a pathological PFO may be difficult. The following phenomena increase the likelihood that a PFO is pathological rather than
incidental:
Anatomical factors:
- floppy interatrial septum;
- resting right-to-left shunt;
- prominent Eustachian valve;
Brain imaging:
- emboli in more than one territory;
- cortical or cerebellar emboli;
- old infarcts with or without clinical correlate;
Haematological:
- prothrombotic conditions;
- deep venous thrombosis;
- pulmonary embolism;
Genetic predisposition:
- familial history of stroke <60 years of age;
- PFO closure in a first-degree relative;
- familial prothrombosis;
Situational:
- recent history of a long journey;
- recent major surgery; or
- Valsalva manoeuvre pre-stroke.
Diagnosis
Diagnosis is usually made with bubble contrast transthoracic echocardiography (BCTTE). Microbubbles are formed by mixing 1 ml air, 2 ml blood and 7 ml saline using two Luer-lock 10 ml syringes and a three-way tap. This mixture is injected into an antecubital fossa vein during a four-chamber echocardiography. The right heart should be fully opacified with this agitated microbubble mix. A resting right-to-left intracardiac shunt is diagnosed if a bolus of bubbles enters the left heart within three heartbeats. Provocative testing is often required. The interatrial septum should be seen to bow to the left during Valsalva release. Several injections may be made with Valsalva release or a sharp sniff, timed to coincide with right atrial opacification. The ‘size’ of the shunt depends much more on patient and echocardiographic co-ordination than it does on the actual size of the PFO. If the index of suspicion is high but the test is negative, femoral venous injections can occasionally reveal an undiagnosed shunt as blood is directly projected at the septum. False positive studies can be obtained if colloid is used as nanobubbles may cross the lung circulation. A pulmonary arteriovenous malformation may be misinterpreted as a PFO: the steady building stream of bubbles to the left heart should alert the operator and can be contrasted with the cluster or cloud of bubbles that characterise a PFO.
Transoesophageal echocardiogram (TOE) is not always required. Valsalva is difficult with a probe in the oesophagus so TOE is less sensitive than BCTTE for PFO diagnosis, but does provide useful additional anatomical information, and can occasionally identify potential alternative embolic sources such as a papillary fibroelastoma. Transcranial Doppler can also be used for diagnosis. This can diagnose a right-to-left shunt but cannot localise it.
The term ‘atrial septal aneurysm’ should be retired as it can cause unnecessary anxiety; it should be replaced by floppy or mobile septum.
Other Indications for Patent Foramen Ovale Closure
Other indications for PFO closure include platypnoea–orthodeoxia syndrome, decompression illness and – rarely – migraine with aura. Platypnoea–orthodeoxia syndrome occurs when deoxygenated blood passes from the inferior vena cava directly through the right atrium and into the left atrium. This phenomenon is usually seen in older age, in association with distortion of the atrial septum and/or mediastinal shift (e.g. because of aortic enlargement, pectus excavatum, scoliosis or pneumonectomy). The deoxygenation can often be positional, such that it occurs on standing but is partially relieved by lying flat. This phenomenon should be differentiated from right-to-left shunting across the atrial septum due to pulmonary vascular disease or tricuspid valve regurgitation. PFO closure for orthodeoxia can be more challenging because of distorted anatomy.
Some types of decompression illness are strongly associated with the presence of a right-to-left intracardiac shunt, most commonly though a PFO. Divers who have suffered a decompression illness should be seen by a diving physician. Post-closure they will need a negative bubble contrast echocardiogram and must pass a diving medical appropriate to the diving being undertaken prior to diving again. Professional divers will need to pass a Health and Safety Executive diving medical.
Migraine with aura is strongly associated with the presence of a PFO. Observational data have shown that PFO closure can improve migraine symptoms in a proportion of individuals undergoing PFO closure for other reasons, but randomised data have not supported routine use of PFO closure for refractory migraine with aura. If PFO closure is undertaken for migraine as the primary indication it should be done only in collaboration with a neurologist with expertise in migraine management and with disclosure of significant uncertainties to the patient.
Patent Foramen Ovale Closure
Pre-procedure
Patients are usually referred to an implanting cardiologist by a stroke physician or another cardiologist. Most patients are referred because of unexplained cerebral embolic events. They have usually had brain imaging (CT and/or MRI), carotid Doppler assessment, cardiac rhythm monitoring and blood tests for thrombophilia, followed by a BCTTE.
The operator should ideally see the patient in person. Telephone assessment is not as good but may be required in geographically large catchment areas. The operator should satisfy themselves that:
- the history and findings are compatible with paradoxical embolisation;
- the patient does have a PFO based on review of the imaging;
- the PFO is unlikely to be incidental; and
- closure of the PFO is warranted on clinical and pathological grounds.
Monitoring for AF can be tailored to the patient’s risk profile. Most patients with probable PFO-related stroke are younger than the age at which anticoagulation would be considered even with persistent AF. Prolonged monitoring should only be done when AF is a realistic and plausible cause of stroke in its own right. Implantation of a loop recorder should be a rarity.
Multidisciplinary team (MDT) meetings were required during the commissioning through evaluation process for discussion of suitable patients for PFO closure. However, it should be noted that the referral process is already multidisciplinary. The MDT can only assess the technical aspects of the case rather than its clinical appropriateness, which requires one-to-one discussion with the patient. An MDT can add value and be useful within a network but should not be considered mandatory for the selection of patients for PFO closure.
Closure Procedure
PFO closures may be undertaken with echocardiographic guidance (TOE or intracardiac echocardiography [ICE]) or with fluoroscopy alone if a TOE has characterised the anatomy accurately. Prophylactic intravenous antibiotics are required. Femoral vein puncture should be undertaken using ultrasound guidance. Pre-closure venous sutures may facilitate same-day discharge. If ICE is used, a second femoral vein puncture should be made with a longer sheath to avoid the ICE catheter preferentially selecting pelvic veins, which can cause discomfort.
After femoral venous access, heparin 70–100 IU/kg should be given prior to crossing the PFO. Crossing the PFO can be done with a J-tipped guidewire or a soft multipurpose catheter, usually in a left anterior oblique fluoroscopic projection. A guidewire is placed posteriorly into the left upper pulmonary vein by preference. The defect is assessed with a sizing balloon to assess the maximum diameter of the defect, its compliance and its location. Balloon sizing should be undertaken unless 3D TOE analysis of the PFO with a stiff wire in situ is made, mimicking balloon palpation. Extremely narrow waisting of the balloon should alert the operator to the possibility that a fenestration has been crossed in error, in which case repeat right-to-left wire passage should be made. Very broad defects (>15 mm width) should be closed with echo guidance. The size of the closure device is chosen based on the echocardiographic, fluoroscopic and balloon interrogation findings. The delivery sheath is passed into the left atrium and is aspirated to ensure de-aired free flow of blood. The device is deployed either side of the interatrial septum. Grasp of the septum secundum is essential to effect closure and is confirmed echocardiographically and/or fluoroscopically. Stability testing is not required if the position of the device is correct. The device is released once correct positioning is confirmed. After sheath removal protamine may be given if desired. Manual compression or suture-based vein closure is usually made.
Procedural intracardiac echo facilitates case scheduling as there is no need for a dedicated imaging specialist or anaesthetist. The expense of the ICE catheter may be offset in this way. An operator should not use ICE-guided PFO closure in their first 50 cases so that they can focus solely on correct device implantation. Prior to relying on ICE, an operator should receive appropriate education and proctoring. Integration of a dedicated structural imaging specialist enhances a PFO closure programme. Many programmes rely on the skills of the imaging specialist for pre-procedural and procedural expertise.
If a PFO is not found on interrogation and contrast fluoroscopy of the oval fossa the preprocedural imaging findings should be reviewed: on-table bubble contrast echocardiography can be repeated and pulmonary angiography may be helpful. Currently, almost all PFO closures are done with double-disc devices. Suture-based devices and in situ closure devices require different implantation methods. Bubble contrast injections to test completeness of closure at the end of the procedure are not recommended.
Procedural complications should be rare with PFO closure. Device embolisation, cardiac tamponade, stroke and air embolism can occur, but together should be seen in ≤1% of cases. However, the ability to deal with complications is a prerequisite for providing a comprehensive PFO closure service. Operators should be familiar with large-bore arterial and venous access pre-closure and sheath insertion, ultrasound-guided urgent pericardiocentesis, and retrieval of an embolised device with a snare or other device, where the retrieval sheath should be at least 50% larger in diameter than the sheath required for implantation, as the device may need to be retrieved without coaxial orientation.
Post-procedure
Patients should have routine observations made for a minimum of 2 hours post-procedure. ECG monitoring is not necessary. Patients are usually discharged on the same day. Prior to discharge, patients should be checked for abnormal clinical symptoms and signs. A pre-discharge echocardiogram is often done but the yield from this is extraordinarily low and it is not considered mandatory.
Patients should receive antiplatelet therapy post-procedure. The usual prescription ranges from a single antiplatelet agent for 6 months, to dual antiplatelets for 6 months and a single antiplatelet agent thereafter. Rarely a patient may be anticoagulated (although this would normally undermine the rationale for the procedure) in which case no additional antiplatelet agent is needed. Beyond 6 months the decision to continue an antiplatelet agent is based on individual risk factors.
Patients should be given contact details and information regarding post-procedure care. Up to 5% of patients may have transient AF in the first 2 months post-implant. Anticoagulation should not be needed except in unusual cases. AF that is other than transient is rare.
Uncomplicated patients should be followed up by telephone or in person. A single echocardiogram post-procedure should be made, either pre-discharge or on follow-up to ensure correct device positioning. Routine bubble contrast echocardiography at 6 months is of uncertain value and is not considered necessary. Some patients will have incomplete closure of the PFO to bubble stress but the physical barrier to thrombus passage remains and there is no evidence that long-term continuation of antiplatelets offers more benefit than harm. Routine follow-up beyond 6 months has a low yield for detecting problems, as late complications are rare.
Standards
Centre and individual procedural volume should be concentrated. Complication rates should be as low as possible as this is a preventative rather than therapeutic procedure. Procedures should take place in a cardiothoracic centre to facilitate good immediate access to a wide range of potential expertise in the event of a serious complication. Centres need to be commissioned by NHS England (or relevant national authority) to undertake the procedure. All cardiothoracic centres in the UK should provide PFO closure as part of a comprehensive regional approach to stroke prevention.
Implanting centres should establish a comprehensive referral network to raise awareness of therapy, educating potential referrers and encouraging equity of access. A centre should conduct a minimum of 25 procedures per year, rising to a minimum of 50 procedures per year after 3 years. An operator should conduct a minimum of 25 procedures per year. An operator who is undertaking ≥50 procedures per year can be a trainer to consultant colleagues or fellows.
Operators should be experienced in managing complications, in particular pericardial effusion, device embolisation and femoral vascular problems. Device manufacturers have established training courses that include theory, procedural simulation, observation of live cases and proctored implants. Operators should attend established training courses before starting a programme. Operators should have onsite proctoring for an initial minimum of 10 cases and should maintain exposure to on-going education at dedicated meetings. Training usually occurs in the setting of a dedicated post-certificate of completion of training fellowship, or at a consultant level.
Implant data should be recorded locally and submitted nationally to the National Institute for Cardiovascular Outcomes Research PFO Closure dataset for audit purposes and to ensure recording of the implant in accordance with the latest regulations. This dataset has been established by the British Cardiovascular Intervention Society, hosted by the National Institute for Cardiovascular Outcomes Research. Local activity should also be presented at audit meetings at least twice annually. A centre that does PFO closure procedures should have formal links with a Level 1 Adult Congenital Heart Disease Service.
Conclusion
Percutaneous PFO closure is effective for prevention of recurrent paradoxical embolisation. Provision in the UK is currently limited, with significant geographical variation. All cardiothoracic centres in the UK should provide a PFO closure service as part of a comprehensive regional stroke prevention strategy.