The patent foramen ovale (PFO) is a common remnant of the intrauterine phase. After birth, the pressure in the right atrium drops with the first breath. The flap–valve-like gap between the cranial muscular septum secundum and the caudal membranous septum primum closes like a door ajar in a draught. Normally, the two septa have sufficient overlap and the septum primum is quite stable, allowing permanent fusion of the gap. However, in about one in four people mechanical fusion fails to occur and the gap remains openable. Facilitating factors are lack of sufficient overlap (the septum primum may even fail to reach the septum secundum, thereby producing one of the various types of secundum atrial septal defects), an extremely thin and mobile septum primum (also called atrial septal aneurysm [ASA]) or a Eustachian valve (a Chiari network bundling the inferior vena cava flow directly onto the site of the foramen ovale). Combinations of these factors compound the chance of a persistent PFO and also render it more dangerous (e.g. opening with virtually every heart beat, even when there is no increased pressure in the right atrium, such as after a Valsalva manoeuvre).
The risk of a PFO remains controversial. Case reports proving that venous thrombi cross the PFO on one side and embolically occluded arterial vessels in otherwise healthy people on the other side suggest that crossing these venous thrombi can cause paradoxical embolism. Opinions vary from the extreme point of view that PFO significantly shortens life by intrinsic selective mortality to the point of view that paradoxical embolisms through the PFO are so rare that one should only bother with PFO closures in the case of a third otherwise unexplained cerebral vascular accident occurring under anticoagulation.
The idea of selective mortality may be corroborated by the fact that the prevalence of PFO decreases with age.1 The only other explanation for this finding is spontaneous closure of the PFO throughout life; this second, conservative point of view is shared by American specialists, who have published the most recent guidelines.2
Even taking an intermediate position, textbooks and definitions will need to be rewritten. The term cryptogenic stroke cannot be used in the presence of a PFO: PFO is a possible cause of stroke, just like atherosclerosis, atrial fibrillation, prior myocardial infarction or spontaneous carotid artery dissection; therefore, the term cryptogenic should be used only if a PFO has already been excluded. Table 1 suggests a modified classification of ischaemic stroke.3 It even makes sense to start looking for PFO in systemic embolism in young people, in whom there is a minimal chance of finding one of the classic causes. PFO should also be looked for in elderly people with one or several competitive causes. Again, textbooks, guidelines and teaching tools will need to be revised to take this into account.
Current percutaneous techniques (particularly when employing the Amplatzer PFO occluder) allow swift closure in a 15-minute procedure, which includes puncture of the femoral vein, a couple of minutes of fluoroscopy, about 10ml of contrast medium and hardware with a price tag of about US$5,000. The patient can resume a completely normal life – including all physical activities – a couple of hours later. Technical failures are virtually absent in cases where the presence of a PFO has been proved unequivocally by documentation of bubbles crossing the PFO on transoesophageal echocardiography (TEE). The only complication (short- or long-term) occurring in more than 0.1% of cases is an arteriovenous fistula created by the venous puncture in the groin. This does not cause symptoms, but should be looked for and corrected when found. If all preventative measures are applied during the venous puncture, the occurrence of this problem is below 1%. There is virtually no need for after-treatment, although acetylsalicylic acid for a few months is standard. Our centre tends to recommend a final TEE at six months, after which the patient is released without any medication or recommendation for prophylaxis against endocarditis for the rest of his or her life. At this follow-up examination, incomplete closure is found in about 10% of cases. If there is still a straight jet from the inferior vena cava through the remaining passage with no impediment caused by the device, a second device (which is even more easy to implant than the first) is required. This will be necessary in about one in 25 cases.
While there are a variety of devices clinically used around the world (Figure 1 shows those with CE approval), the Amplatzer PFO Occluder and its generic copycat the Occlutech PFO Occluder is the best by a large margin. Nevertheless, PFOs can be closed with other devices, albeit at the price of a somewhat more complicated procedure, reduced closure rate and moderately increased risk of complications.
Proof that closing the PFO with a device is safer than not closing it is still lacking; however, as is often the case in medicine, lack of proof does not mean something is not true. The first randomised trial comparing PFO closure with an Amplatzer occluder versus conventional treatment with either acetylsalicylic acid or coumadin was started almost 10 years ago and is still ongoing (see Table 2). Recruitment has been hampered by the ease and appeal of closing a hole in the heart that should not be there rather than waiting for a stroke to occur while swallowing some pills. It may well be that the trials will individually fail to prove the superiority of device closure, although they will certainly prove the safety of such procedures. The spontaneous recurrence of stroke under medical treatment may have been overestimated.4 Therefore, a much longer follow-up (or a larger number of patients randomised) than initially planned may be required to prove the principle; it may be necessary to pool the various trials to accomplish this. Even if pure, level A scientific proof based on several clean, randomised trials remains pending for many years to come, it appears logical that if a patient’s PFO can be closed without any untoward effects, that individual will be better off. Nevertheless, he or she will usually not feel any difference, and will certainly not realise that perhaps 10, 20 or 30 years later a stroke would have occurred, because due to the procedure such an event was avoided.
The best data available so far are from a matched control trial comparing a variety of closure devices with either coumadin or acetylsalicylic treatment. These data showed that device closure is better than acetylsalicylic acid but comparable to coumadin.5
To date, about 70% of patients referred for PFO closure have a background of a cerebral event, be it a full-blown stroke or a transient ischaemic attack. The remainder are people with dangerous professions or hobbies or other situations increasing the possibility of a PFO (see Table 3).
It will likely be possible to prove a significant benefit for symptoms of migraine (a common problem among PFO carriers), particularly aura, long before a significant difference is shown in the stroke trials. Although difficult to explain pathophysiologically, the soothing of migraine is a welcome bonus. In fact, it is the only subjective improvement patients may feel after PFO closure, as a prevented stroke is not ‘felt’ as such. It goes without saying that companies producing PFO closure devices are banking primarily on the migraine trials rather than the stroke trials, which may take decades to be successfully completed.
We certainly have yet to grasp the complexity of the PFO question: we are still grappling with quantifying the individual risk of a PFO based on its anatomical structures; we must still be turning a blind eye to many PFO problems such as strokes or myocardial infarctions in patients with atherosclerosis that may nevertheless be PFO-induced; and, finally, we are still overestimating the complexity or risk potential of having a device implanted to close the PFO.
Nothing has come closer to ‘mechanical vaccination’ than closing a PFO with an Amplatzer occluder (or something even more simple that may appear in the future). If all that was needed was a shot or a pill, as in a true vaccination, nobody would object to at least looking for all of the dangerous PFOs (present in about 4% of the population) or even closing all PFOs (present in about 25% of the population). Looking at an Amplatzer occluder still prompts many physicians or patients to change their mind about the procedure. A single stitch or thermofusion sounds appealing; however, these techniques do not deal successfully with the most dangerous PFOs – those with no septal overlap and a very unstable septum primum. A fan-type multistitch device could work, but such a device does not yet exist. For the time being, we should exploit what we already have to a greater extent, because it is quite spectacular.