Patent foramen ovale (PFO) and cryptogenic stroke

Julius Cohnheim, a German pathologist, is credited, in 1877, as the first to hypothesize that the failed closure after birth of the foramen ovale (the interatrial opening that allows the oxygenated blood from the placenta to bypass the non‐functioning pulmonary circulation and enter the fetal systemic circulation) might underlie the spreading of venous emboli to the brain. However, although commonly accepted as a possible cause of stroke, paradoxical embolism was believed to merely represent a medical curiosity up until the end of the 1980s when, due to the advancement of echocardiography, it was recognized that the patency of the foramen ovale (PFO) was two to three times more frequent in patients with strokes of otherwise unknown cause (i.e. cryptogenic) than in healthy controls or in patients with stroke of known etiology [1Overell J.R. Bone I. Lees K.R. Interatrial septal abnormalities and stroke A meta‐analysis of case‐control studies.Neurology. 2000; 55: 1172-9Crossref PubMed Google Scholar]. The echocardiographic demonstration, in single case reports, of thrombi trapped across the foramen provided the evidence that emboli of venous origin may be able to bypass the pulmonary filter and enter the systemic circulation across the PFO, and so represent a risk factor for stroke [2Dörr M. Hummel A. Paradoxical embolism – thrombus in a patent foramen ovale.N Engl J Med. 2007; 357: 22Crossref Scopus (18) Google Scholar]. Recent meta‐analyses of case‐control studies have confirmed that the likelihood of cryptogenic stroke patients with a PFO is about three times that of controls [1Overell J.R. Bone I. Lees K.R. Interatrial septal abnormalities and stroke A meta‐analysis of case‐control studies.Neurology. 2000; 55: 1172-9Crossref PubMed Google Scholar, 3Amarenco P. Patent foramen ovale and the risk of stroke: smoking gun guilty by association?.Heart. 2005; 91: 441-3Crossref PubMed Scopus (0) Google Scholar, 4Handke M. Harloff A. Olschewski M. Hetzel A. Geibel A. Patent foramen ovale and cryptogenic stroke in older patients.N Engl J Med. 2007; 357: 2262-8Crossref PubMed Scopus (479) Google Scholar] (Table 1). More recently, Alsheik et al. applied the Bayes theorem to the available data to ascertain the probability of PFO being incidental in a patient with cryptogenic stroke. They came to the conclusion that PFO may represent an incidental finding in 33–48% of cryptogenic strokes when in isolation, but only in 11% when associated with bulging of the interatrial wall (so called atrial septal aneurism – ASA). These figures drop to 20% and 9%, respectively, in patients less than 55 years old [5Alsheikh‐Ali A.A. Thaler D.E. Kent D.M. Patent foramen ovale in cryptogenic stroke incidental or pathogenic?.Stroke. 2009; 40: 2349-55Crossref PubMed Scopus (0) Google Scholar].Table 1Summary table of studies comparing incidence of PFO in cryptogenic stroke patients vs. healthy controls [1Overell J.R. Bone I. Lees K.R. Interatrial septal abnormalities and stroke A meta‐analysis of case‐control studies.Neurology. 2000; 55: 1172-9Crossref PubMed Google Scholar, 3Amarenco P. Patent foramen ovale and the risk of stroke: smoking gun guilty by association?.Heart. 2005; 91: 441-3Crossref PubMed Scopus (0) Google Scholar, 4Handke M. Harloff A. Olschewski M. Hetzel A. Geibel A. Patent foramen ovale and cryptogenic stroke in older patients.N Engl J Med. 2007; 357: 2262-8Crossref PubMed Scopus (479) Google Scholar]StudyCryptogenicControlOR (95% CI)n.PFO+/n.totalLechat, 198814/2610/10010.50 (3.82–28.85)Webster, 198819/346/407.18 (2.39–21.58)Vella, 19911/380/332.68 (0.11–68.05)Hausmann, 199214/6525/1161.00 (0.48–2.09)10de Belder, 19929/353/9410.50 (2.65–41.63)Di Tullio, 199210/211/2481.82 (9.54–701.61)Labovitz, 199338/27039/7723.08 (1.92–4.93)Cabanes, 199336/649/505.86 (2.44–14.04)Van Camp, 19939/294/282.70 (0.72–10.10)Job, 199427/4127/632.57 (1.14–5.81)Jones, 199414/7131/2021.35 (0.67–2.72)Zahn, 199550/11815/813.24 (1.66–6.32)Roijer, 199717/6715/681.20 (0.54–2.66)Serena, 199830/5332/1002.77 (1.39–5.51)Handke, 200777/22734/2763.65 (2.32–5.74)Total (95% CI)365/1159251/20473.29 (2.74–3.94) Open table in a new tab Similarly, a prospective study conducted in France on the risk of recurrence in young patients with cryptogenic stroke [6Mas J.L. Arquizan C. Lamy C. Zuber M. Cabanes L. Derumeaux G. Coste J. Recurrent cerebrovascular events associated with patent foramen ovale, atrial septal aneurysm, or both.N Engl J Med. 2001; 345: 1740-6Crossref PubMed Scopus (1215) Google Scholar] showed that, compared with patents with no atrial septal abnormalities, the PFO‐ASA association conveyed a 4‐fold risk of relapse (hazard ratio 4.17, 95% CI 1.47–11.84). That PFO may enhance the risk of getting brain emboli has been further elegantly shown by a recent prospective study that identified, by logistic regression analysis, PFO was found to be an independent predictor of silent brain infarction with an OR of 34.9 (95% CI 31–394.3, P = 0.004) [7Clergeau M.R. Hamon M. Morello R. Saloux E. Viader F. Hamon M. Silent cerebral infarcts in patients with pulmonary embolism and a patent foramen ovale: a prospective diffusion‐weighted MRI study.Stroke. 2009; 40: 3758-62Crossref PubMed Scopus (0) Google Scholar]. The difficulty in identifying those with truly cryptogenic stroke amongst aged patients (due to the frequent co‐occurrence of vascular risk factors) probably explains why PFO was only marginally associated with stroke in patients aged over 55 years in older studies [1Overell J.R. Bone I. Lees K.R. Interatrial septal abnormalities and stroke A meta‐analysis of case‐control studies.Neurology. 2000; 55: 1172-9Crossref PubMed Google Scholar]. Recent evidence, however, strongly suggests that PFO is over‐represented in cryptogenic strokes irrespective of the age of occurrence [4Handke M. Harloff A. Olschewski M. Hetzel A. Geibel A. Patent foramen ovale and cryptogenic stroke in older patients.N Engl J Med. 2007; 357: 2262-8Crossref PubMed Scopus (479) Google Scholar]. To mitigate the perception that PFO is a dangerous mate, two population‐based studies both failed to identify PFO as an independent predictor of a first‐in‐a‐lifetime stroke during a median follow‐up of 5 years after adjustment for age and comorbidity, although in both studies there was a trend towards a 1.5 higher risk in PFO carriers [8Meissner I. Khandheria B.K. Heit J.A. Petty G.W. Sheps S.G. Schwartz G.L. Whisnant J.P. Wiebers D.O. Covalt J.L. Petterson T.M. Christianson T.J.H. Agmon Y. Patent foramen ovale: innocent or guilty? Evidence from a prospective population‐based study.J Am Coll Cardiol. 2006; 47: 440-5Crossref PubMed Scopus (0) Google Scholar, 9Di T.u.l.l.i.o.M.R. Sacco R.L. Sciacca R.R. Jin Z. Homma S. Patent foramen ovale and the risk of ischemic stroke in a multiethnic population.J Am Coll Cardiol. 2007; 49: 797-802Crossref PubMed Scopus (272) Google Scholar].Likewise, and somehow contradicting the French study, the PICCS study, which evaluated a subpopulation of the bigger WARS study undergoing echocardiography, failed to show an increased risk of recurrent stroke in patients with PFO, independent of the association with ASA [10Homma S. Sacco R.L. Di Tullio M.R. Sciacca R.R. Mohr J.P. PFO in Cryptogenic Stroke Study (PICSS) InvestigatorsEffect of medical treatment in stroke patients with patent foramen ovale: patent foramen ovale in cryptogenic stroke study.Circulation. 2002; 105: 2625-31Crossref PubMed Scopus (894) Google Scholar]. The negative findings of these studies may in part be explained by the exclusion of younger patients, the enrollment of patients with multiple risk factors, the relatively short follow‐up and the relatively small sample size, but undoubtedly they stand as a reminder that ‘cryptogenic stroke with PFO is not equal to paradoxical embolism’ [11Thaler D.E. Saver J.L. Cryptogenic stroke and patent foramen ovale.Curr Opin Cardiol. 2008; 23: 537-44Crossref PubMed Scopus (42) Google Scholar] and this is both because PFO might be found in up to 25% of healthy subjects and strokes also occur without apparent cause in patients without PFO. Therefore, there is an urgent need to identify among PFO‐associated cryptogenic strokes those in whom PFO is causative either as a conduit for paradoxical embolism, a site of thrombus formation or as an anatomical marker of an increased propensity to develop major arrhythmias. It should be pointed out that the latter two mechanisms are rather speculative and much less well documented [11Thaler D.E. Saver J.L. Cryptogenic stroke and patent foramen ovale.Curr Opin Cardiol. 2008; 23: 537-44Crossref PubMed Scopus (42) Google Scholar]. Finding clues in favour of the diagnosis of paradoxical embolism requires a number of procedures. First, a thorough diagnostic work‐up is necessary to exclude the co‐occurrence of vascular risk factors. Secondly, an accurate history should be taken to look for conditions that may have either facilitated, or be a concomitant consequence of, venous thrombosis (DVT). In a well‐designed case‐control study comparing cryptogenic stroke patients with and without PFO, a recent DVT or pulmonary embolism, a Valsalva manoeuvre preceding the onset of neurological deficits, recent prolonged travel and waking up with stroke were all independently associated with PFO [12Ozdemir A.O. Tamayo A. Munoz C. Dias B. Spence J.D. Cryptogenic stroke and patent foramen ovale: Clinical clues to paradoxical embolism.J Neurol Sci. 2008; 275: 121-7Abstract Full Text Full Text PDF PubMed Scopus (90) Google Scholar]. Stroke upon awakening has been recently related to the suggested high incidence of PFO in patients with obstructive sleep apnoeas [13Ozdemir O. Beletsky V. Hachinski V. Spence J.D. Cerebrovascular events on awakening, patent foramen ovale and obstructive sleep apnea syndrome.J Neurol Sci. 2008; 268: 193-4Abstract Full Text Full Text PDF PubMed Scopus (0) Google Scholar]. Among PFO carriers, those symptomatic for cerebrovascular disease have been found to exhibit a wider PFO (3.9 ± 1.6 vs. 2.9 ± 1.4 mm), a significantly higher proportion of ASA (45% vs. 21%), of greater or equal to 1 cm long tunnels (78% vs. 55%) and of severe right to left shunt (16% vs. 5%), compared with asymptomatic subjects, thus confirming earlier reports [14Goel S.S. Tuzcu E.M. Shishehbor M.H. de Oliveira E.I. Borek P.P. Krasuski R.A. Rodriguez L.L. Kapadia S.R. Morphology of the patent foramen ovale in asymptomatic versus symptomatic (stroke or transient ischemic attack) patients.Am J Cardiol. 2009; 103: 124-9Abstract Full Text Full Text PDF PubMed Scopus (101) Google Scholar]. Of the above‐mentioned features, the association with ASA appears to be the strongest indicator of the possible pathogenic role of PFO. Searching for occult venous thrombosis may be rewarding provided it is done early and the investigation is extended to pelvic veins. Magnetic resonance imaging (MRI) venography might prove the gold standard in this respect in the near future [15Cramer S.C. Rordorf G. Maki J.H. Kramer L.A. Grotta J.C. Scott Burgin W. Hinchey J.A. Benesch C. Furie K.L. Lutsep H.L. Kelly E. Longstreth Jr W.T. Increased pelvic vein thrombi in cryptogenic stroke: results of the Paradoxical Emboli from Large Veins in Ischemic Stroke (PELVIS) study.Stroke. 2004; 35: 46-50Crossref PubMed Scopus (205) Google Scholar]. Further support for paradoxical embolism lies in the demonstration of the prothrombotic genotype associated with factor V Leiden and prothrombin G20210A mutations with an O.R of 4.7 (95%CI 1.4–16.1; [16Botto N. Spadoni I. Giusti S. Ait‐Ali L. Sicari R. Andreassi M.G. Prothrombotic mutations as risk factors for cryptogenic ischemic cerebrovascular events in young subjects with patent foramen ovale.Stroke. 2007; 38: 2070-3Crossref PubMed Scopus (92) Google Scholar]). Up to 45% of patients with PFO‐associated cryptogenic stroke may show asymptomatic pulmonary embolism as assessed by perfusion/ventilation lung scintigraphy, according to an as yet unpublished study carried out by Kaps and associates (M. Kaps, personal communication). Another way to confirm the paradoxical embolism hypothesis is to look for occult lesions in other organs, especially in the heart, where late contrast MRI may be able to detect clinically silent infarcts [17Meijs M.F.L. Bots M.L. Cramer M.J.M. Vonken E.‐.J.A. Velthuis B.K. van der Graaf Y. Visseren F.L. Mali W.P.T.M. Doevendans P.A. on behalf of the SMART Study groupUnrecognised myocardial infarction in subjects at high vascular risk: prevalence and determinants.Heart. 2009; 95: 728-32Crossref PubMed Scopus (0) Google Scholar]. We are still far from being able to identify with certainty those, among PFO‐associated strokes, that occurred because a venous embolus crossed the interatrial wall and ended up in a brain vessel. However, from the results of the above‐mentioned studies, we now have additional pathways to investigate to identify causative PFO. After excluding stroke due to atrial fibrilation, carotid disease, aortic arch atheroma, lacunes and rare but documented ‘strong’ causes (e.g. arterial dissection, intracranial atherosclerosis or less frequent genetically determined vasculopathies such as CADASIL or Fabry’s disease), ∼20% of ischemic strokes could reasonably remain of undetermined cause. Among these, about 40% of patients (i.e. ∼8% of all strokes) are expected to have PFO. By conservatively excluding small shunts as irrelevant, it can be assumed that 3–6% of all ischemic strokes may be worthy of further testing. In Italy, with an estimated annual incidence of about 120 000 ischemic strokes, this would mean a figure of 3600–7200 cases per year. The diagnostic work‐up of cryptogenic PFO‐related stroke should become routinely comprehensive and involve, beside the obvious exclusion of atherothrombotic risk factors (diabetes, hypertension, and carotid and aortic arch plaques), the assessment of the local anatomy (atrial septal aneurysm, PFO size, and amount of right to left shunt), and the search for prothrombotic mutations and for venous thrombi in unconventional locations (pelvic or iliac veins), as well as asymptomatic pulmonary and systemic embolism (at least in the heart). This should not be restricted to young patients. Indeed, as the vast majority of strokes occur in older people, the recurrence rate of cryptogenic strokes in such individuals is not negligible, the incidence of deep vein thrombosis increases with age and from a post hoc analysis of the PICSS data, PFO appears to be an independent predictor of further stroke in patients aged over 65 years (27.6% in those with vs. 8.1% in those without PFO at 2 years – [18Homma S. Di Tullio M.R. Sacco R.L. Sciacca R.R. Mohr J. PPICSS InvestigatorsAge as a determinant of adverse events in medically treated cryptogenic stroke patients with patent foramen ovale.Stroke. 2004; 35: 2145-9Crossref PubMed Scopus (78) Google Scholar]), the systematic search for PFO might paradoxically be more rewarding in aged than in young stroke patients [18Homma S. Di Tullio M.R. Sacco R.L. Sciacca R.R. Mohr J. PPICSS InvestigatorsAge as a determinant of adverse events in medically treated cryptogenic stroke patients with patent foramen ovale.Stroke. 2004; 35: 2145-9Crossref PubMed Scopus (78) Google Scholar, 19Handke M. Harloff A. Bode C. Geibel A. Patent foramen ovale and cryptogenic stroke: a matter of age?.Semin Thromb Hemost. 2009; 35: 505-14Crossref PubMed Scopus (40) Google Scholar]. In the authors’ opinion this approach is the only way to select, amongst PFO‐related strokes, those with a reasonable convincing diagnosis of paradoxical embolism amenable to randomized trials comparing medical vs. interventional treatment. For the time being, the optimal treatment to prevent the recurrence of PFO‐associated cryptogenic strokes is uncertain. Overall, the pooled absolute rate of recurrent ischemic stroke or transient ischemic attack (TIA) in patients with PFO was 4.0 events per 100 person‐years (95% CI 3.0–5.1) while the rate of recurrent ischemic stroke was 1.6 events per 100 person‐years (95% CI 1.1–2.1) in a recent meta‐analysis [20Almekhlafi M.A. Wilton S.B. Rabi D.M. Ghali W.A. Lorenzetti D.L. Hill M.D. Recurrent cerebral ischemia in medically treated patent foramen ovale: a meta‐analysis.Neurology. 2009; 73: 89-97Crossref PubMed Scopus (0) Google Scholar]. However, these figures are of little help inasmuch as they are derived from the pooled analysis of reports in which patients of different ages were included, and no attempt was made to stratify for the presence of those factors that may increase the likelihood of PFO being the culprit lesion. This heterogeneity is further emphasized by the huge variability in the outcomes of individual reports, which ranges from 0% to 4.4% for recurrent stroke and from 0% to 12% for recurrent TIA/stroke at 1 year [21Wöhrle J. Closure of patent foramen ovale after cryptogenic stroke.Lancet. 2006; 368: 350-2Abstract Full Text Full Text PDF PubMed Scopus (0) Google Scholar]. Another uncertainty arises from the lack of appropriately designed randomized trials that compare antiplatelet vs. anticoagulant treatment. The only prospective study that addressed such a comparison in non‐cardioembolic stroke, the PICSS trial, was not sufficiently powered to assess their effect in PFO‐associated cryptogenic strokes. Yet, although non‐statistically significant, the results showed a clear trend for warfarin to better prevent recurrence when compared with aspirin (2‐year rate of recurrent stroke or death 9.5% in warfarin vs. 17.9% in aspirin‐treated patients [10Homma S. Sacco R.L. Di Tullio M.R. Sciacca R.R. Mohr J.P. PFO in Cryptogenic Stroke Study (PICSS) InvestigatorsEffect of medical treatment in stroke patients with patent foramen ovale: patent foramen ovale in cryptogenic stroke study.Circulation. 2002; 105: 2625-31Crossref PubMed Scopus (894) Google Scholar]). Further non‐randomized studies have confirmed that treatment with anticoagulants tends to half the recurrence of ischemic cerebral events in comparison to aspirin [21Wöhrle J. Closure of patent foramen ovale after cryptogenic stroke.Lancet. 2006; 368: 350-2Abstract Full Text Full Text PDF PubMed Scopus (0) Google Scholar]. These findings are not unexpected inasmuch as they affirm the concept that an as yet undefined proportion of strokes is due to paradoxical embolism from venous sources in patients with cryptogenic stroke and PFO, and that the recurrence of these is prevented by anticoagulants. In the remaining population, PFO may therefore be regarded as an innocent bystander, as stroke is likely to have occurred because of other mechanisms. Consequently, optimal protection against recurrence is possibly based on antiplatelet treatment. What may be the benefit of percutaneous PFO closure? Pending the results of ongoing randomized trials, which, for the same reasons as those regarding the controversy between antiplatelet and anticoagulant treatments, are not expected to solve the problem, transcatheter closure of PFO may be achieved with a very low complication rate, both in the short and long terms [22Holmes Jr, D.R. Cohen H.A. Ruiz C. Patent foramen ovale, systemic embolization, and closure.Curr Probl Cardiol. 2009; 34: 483-530Crossref PubMed Scopus (0) Google Scholar]. PFO closure should be considered as an alternative to anticoagulation unless there is a documented high risk of deep vein thrombosis, which imposes lifelong anticoagulant therapy, in which case PFO closure should be regarded as an add‐on treatment. The type of treatment that best fits the patient’s needs should be determined by first establishing the contribution of PFO to the index event, and second, and only when the PFO has been deemed causative, from the careful balance of pros and cons of anticoagulant vs. percutaneous treatment. The authors state that they have no conflict of interest.