PFO Basics

Stroke is currently the leading cause of disability(1) and is projected to pass heart attacks as the leading cause of mortality worldwide. Approximately 40% of all strokes are classified as cryptogenic — meaning that the cause is unknown despite an extensive work up. While patent foramen ovale (PFO) is a common and generally benign condition found on autopsy in 25% of the population, approximately 40% to 50% of patients younger than age 55 with cryptogenic stroke (CS) have PFO on transesophageal echocardiography (TEE). This suggests that paradoxical emboli (PE) (i.e. venous emboli that gain access to the arterial circulation through a PFO) are a major cause of CS. Thus, many physicians will recommend PFO closure in patients, especially younger patients, who have had CS.

However, the benefit of PFO closure in patients with CS has not been demonstrated, and remains unclear. First, for someone presenting with a PFO and a CS, it is not clear for that individual whether the PFO is causally related to the index event or whether it is an incidental finding, since it is found so commonly in the general population. Second, the overall rate of stroke recurrence is relatively low in patients with CS and PFO, ranging in studies from 0% to 12% per year, with an average annualized risk across studies of about 2%. While the risk of recurrence appears to be higher than average in some patients, such as those with atrial septal aneurysm (ASA)6) or other high risk features, it must be lower than the annualized 2% overall risk in others without these features.

How patient-specific factors affect the likelihood that a PFO is related to an index stroke and how patient-specific factors affect the risk of recurrence is not well understood. Prior studies appear to have contradictory and at times paradoxical findings in regard to patient factors associating PFO with CS and with recurrence rate. These studies have been limited by extremely low statistical power for examining factors related to recurrence risk, Further, there have been methodologic problems related to risk modeling – there has been no controlling or accounting for stroke mechanisms other than PE. Given the relatively low overall recurrence risk, the presence of factors likely to affect this recurrence risk, the presence of factors that affect whether the PFO is likely to be causally related to the index event or an incidental finding, and the fact that PFO closure itself is associated with risks, it is apparent that closure may be beneficial in some but not in others.

Based on individual patient data from patients with CS both with and without PFO, patient characteristics may be identifiable that identify those patients presenting with PFO and CS for whom the CS is likely to be causally related to the PFO (i.e. caused by a PE). Further, among patients with CS and PFO, the risk of stroke recurrence, and specifically stroke recurrence from a PE, is heterogeneous and may be predictable from clinical, neuroradiologic and echocardiographic characteristics available at the time of the first stroke.

Patent Foramen Ovale (PFO) in Adults

The foramen ovale is a developmental conduit between the right and left atrium. It is required for normal fetal circulation of blood. At birth the foramen ovale typically closes and the normal post-natal pulmonary circulation is established. In autopsy series, roughly 25% of the adult population has a persistent defect through this anatomical structure, known as a patent foramen ovale (PFO). In subjects without a PFO, particulate matter in the venous circulation, such as small venous clots, is filtered through the small capillaries of the lungs. In the presence of a PFO, venous thromboemboli can cause clinical events if they gain access to the arterial circulation. These are known as paradoxical emboli (PE). When PE enter the cerebral circulation, they can cause strokes.

PFOs range from those that shunt small amounts of blood across the interatrial septum only in the setting of a vigorous Valsalva maneuver (cough, sneezing, straining, etc.) to those that shunt large amounts even when the patient is breathing normally at rest. Some shunts are detectable without contrast agents. Jets of interatrial flow may be seen with Doppler imaging alone. Other cardiac abnormalities (e.g. atrial septal aneurysm [ASA]) occur with increased frequency in patients with PFOs and may increase the risk of PFO-related stroke.

Epidemiology of Stroke, Cryptogenic Stroke (CS) and PFO

There are 750,000 strokes in the United States each year; roughly one third are recurrent. Stroke is the third leading cause of death in the United States and the leading cause of acquired disability. It is typically thought to be a disease of the elderly but 25% of strokes occur in patients who are under age 65. The etiology of stroke remains unknown in 30-40% of patients despite extensive diagnostic testing. In a review by Homma et al, the prevalence of PFOs detectable by transesophageal echocardiography (TEE) in patients with CS is between 40% and 50% in younger patients (less than age 55), and about 20% in older patients (Table 1). The increased prevalence of PFO in younger patients with CS suggests a causal relationship between PFO and CS. This has been attributed to PE (and indeed clots have been identified traversing the PFO, although other mechanisms have also been suggested. Regardless of mechanism, there are about 30,000 patients each year under the age of 65 who have a PFO-associated CS. The best management of these patients is uncertain.

Table 1: Cryptogenic Stroke and PFO Statistics

Attributable Risk in Patients with PFO and CS

The prior literature is surprisingly limited regarding the proportion of PFO-associated CS for which the PFO is likely to be causally-related rather than incidental to the event. While in any given patient, the causal relationship is impossible to determine with a high degree of certainty, the attributable proportion can be calculated if either the odds ratio (OR) or relative risk (RR) for PFO in CS versus control is known and the control rate of PFO is also known. As we discuss further in Section C, if the prevalence of PFO is 25% (as seen in autopsy) and the prevalence of PFO in patients with CS is 40% it can be shown that the proportion of patients with PFO and CS in whom the PFO is causal is 50%. The PFO is incidental (not related to the stroke at all) in the other half of patients, as shown in Figure 1. These rates may be quite surprising to some clinicians, who may presume that a PFO found in the setting of a CS is highly likely to be causal.

Indeed, most studies yield somewhat stronger associations between CS and PFO. This is because TEE studies typically reveal a control PFO rate less than the 25% autopsy rates, while the rate in CS patients tends to be greater than 40%. In stroke patients younger than age 55, the likelihood of finding a PFO is increased 6 fold. In patients aged 55 years or older the risk is doubled.

Numerous factors can potentially affect the degree of association between PFO and CS, and thus the likelihood that a discovered PFO in the setting of CS is likely to be causal. The association decreases when looser definitions of CS are used, and also in patients with other risk factors for stroke—most especially patient age, since the older a patient the more likely a CS is to be caused by other (unidentified) etiologies. This is apparent in Table 1, where differences in PFO rates between CS and control groups attenuate with increasing age. The largest cohort study of CS patients evaluated by TEE, The French PFO/ASA Study, found that, even among CS patients less than age 55, PFO was more likely to be associated with patients who were younger, did not have hypertension (HTN) or hyperlipidemia or smoke. The PFO in CS Study (PICSS) similarly found that conventional stroke risk factors (including those above and diabetes [DM]), as well as lower stroke severity, were less likely to be associated with the presence of PFO. Conversely, multiple studies comparing PFO characteristics of patients with CS versus strokes of known cause have found that larger PFO, greater right-to-left shunt and higher septal wall motion mobility, were found with higher frequency in patients with CS.

Stroke Recurrence Rates in Patients with CS and PFO

The overall likelihood of recurrence in the absence of closure is low. In the review by Homma et al, the annualized rate has ranged from 0% to 12% across 9 studies, but averages about 2% (Table 2).

Treatment of Patients with PFOs

Optimal stroke prevention strategies in patients with PFOs have not been established. Options include antithrombotic medical therapy (antiplatelet drugs like aspirin and anticoagulation like warfarin), implantation of PFO closure devices, and direct surgical closure of the PFO (although this latter approach will likely be replaced with endovascular devices). The only randomized trial was PICSS, which compared warfarin to aspirin and found no statistically significant difference on the primary endpoint (stroke or death), although it was not well powered to detect benefits of moderate size. Both medicine- and device-based therapy carries risks. Antithrombotic medications are associated with an increased risk of gastric ulceration, and gastrointestinal and intracranial bleeding. PFO closure devices are associated with device-specific complications such as fracture of device elements, device embolization or thrombus formation.

Currently, the most commonly used devices in the United States are the Amplatzer PFO Occluder (AGA Medical) and CardioSEAL (NMT Medical) devices.(34;35). The Amplatzer device is currently being evaluated in two randomized clinical trials, the RESPECT Trial (in the US) and the PC-trial in Europe. Both trials have been on-going for several years and are expected to be completed and published in the next couple of years.

Major complications from device implantation include death, major hemorrhage, cardiac tamponade and fatal pulmonary emboli. These may occur in approximately 1.5% of patients. Other complications, including thrombus formation, device fracture, or atrial fibrillation, may occur more frequently. Thus, it is clear that meticulous patient selection is critical, since the risks of the treatment may outweigh the benefits for some (even if the device is shown to improve outcomes overall).

Thus, to advise individual patients about the potential benefits of the interventions, it is vital to understand the natural history of CS in patients with PFO, including factors that influence the risk of stroke recurrence, and most specifically those factors that increase the risk of recurrence from PE. To date, little is known about how to assess this risk, and how to assess whether the risk of recurrence will be modified by treatment such as closure. Indeed, whether the on-going clinical trials yield overall positive or overall negative results, substantial differences among the patients in the trials would suggest that the overall results will not apply uniformly to individual patients.

Current Practice Recommendations

Given the current state of uncertainty, it is not surprising that there is tremendous practice variation and lack of consensus, with some clinicians convinced that PFOs are an insignificant epiphenomenon and others enthusiastically and routinely recommending closure. The American Academy of Neurology (AAN) guidelines for stroke with PFO and ASA and the American Stroke Association guidelines are notable for their absence of firm recommendations.

As for therapy, the ANN guideline states, “Among patients with a CS and atrial septal abnormalities, there is insufficient evidence to determine the superiority of aspirin or warfarin for prevention of recurrent stroke or death (Level U), but the risks of minor bleeding are possibly greater with warfarin (Level C). There is insufficient evidence regarding the effectiveness of either surgical or percutaneous closure of PFO (Level U).”

The current guidelines stress the paucity of methodologically sound evidence regarding the implications of PFO and ASA in CS and the optimal management of these patients. Indeed, because of uncertainty about the benefits of the various PFO treatments, there is tremendous practice variation not only in the management of patients with CS and identified PFO, but in the work-up of patients with otherwise unexplained stroke, and in many settings TEE is not routinely performed in such patients. This uncertainty thus affects the management of many more patients than the subgroup of patients who are eventually identified to have a PFO and a CS.

The guideline authors emphasize the need for research to better characterize the natural history of patients with PFO and ASA and, in particular, to address the clinical and anatomic features that may affect the risk of subsequent stroke. The firmest conclusion from both national societies is that patients with CS and PFO should consider participating in clinical research studies such as the ongoing RESPECT and REDUCE studies.

Risk Factors for Stroke Recurrence

Although little is known about how to assess recurrence risk in individual patients, it is clear that the risks of stroke recurrence in patients with PFO and the benefits of medical therapy or PFO closure are likely to be considerably different depending on other patient specific factors. First, given the factors that affect the probability that a PFO discovered in the setting of CS is likely to be causal versus incidental, even if recurrence risk were uniform in all patients, closure of incidental PFOs does not address the underlying cause of the index stroke and therefore would do nothing to decrease the risk of recurrence (and may increase recurrence risk due to device-related complications). Second, even for patients in whom the index event was likely to be due to a PE, there appear to be identifiable factors which affect the likelihood of recurrence risk, making the recurrence risk (already modest on average) even lower in those without risk factors.

Several studies have sought to identify risk factors for stroke recurrence in patients with CS and PFO. The largest published study to date, the French PFO/Atrial Septal Aneurysm (ASA) Study, included over 581 CS patients, 267 with PFO. As reported by Jean-Luis Mas et al, although patients with a PFO alone were no more likely than patients without a PFO to have a stroke recurrence, the presence of an ASA in addition to the PFO increased stroke recurrence risk approximately 4-fold. In the cohort overall, older age was also associated with a higher recurrence rate.

While it seems self-evident that the size of the PFO and degree of right-to-left shunt (both of which were found in other studies to be associated with CS compared to strokes of known cause) might also be risk factors for stroke recurrence, the French PFO/ASA Study did not find any relationship between these factors and the likelihood of recurrence. However, the study had extremely limited power to examine multiple recurrence risk factors or to detect risk factors of only moderate influence.

Other studies including only CS patients, and not other stroke subtypes, have been inadequately powered to look at recurrence risk. The PFO in Cryptogenic Stroke Study (PICSS) was a sub-study of the Warfarin-Aspirin Recurrent Stroke Study (WARSS), a multicenter, double-blind randomized study of two medical regimens (aspirin and warfarin) for secondary stroke prevention. PICCS included those WARSS patients who were evaluated by TEE, including both those with CS and with stroke of known cause. With regard to recurrence risk, PICSS produced results that many have found puzzling or paradoxical, but which can be understood when one considers that the PICSS population on the whole was at relatively high risk for stroke from causes other than PE. First, the PICSS study failed to confirm the main finding of the French PFO/ASA Study; there was no significant difference in recurrence between patients with isolated PFO and those with PFO and ASA (2-year recurrence rate 14.5% versus 15.9%). More surprising perhaps was the slight trend for a higher recurrence rate among patients with small PFO versus larger PFOs (2-year event rates in no, small and large PFOs were 15.4%, 18.5% and 9.5%;).

This finding may at first seem paradoxical but PICSS included a population of patients with both cryptogenic and non-cryptogenic stroke, and with a high rate of risk factors, who were at high risk of stroke recurrence from other causes of ischemic stroke. Patients who had PFO with high risk features (large size or ASA) were more likely to have PE and less likely to have conventional stroke risk factors. Thus, while the investigators were examining the effect of high risk PFO features on recurrence risk, they were actually comparing the recurrence rates of (mostly) PE versus (mostly) strokes of other causes. Therefore, the high risk PFO-features appeared to be slightly protective. Indeed, the two year stroke recurrence rate of patients with PFO and ASA in PICCS was considerably higher than that found in The French PFO/ASA Study (15.9% versus 4.0%), but this combination did not emerge as a significant risk factor in PICSS because the confounding background risk of stroke was so high and not adequately controlled.

For the purposes of selecting patients for PFO closure what is at issue is not stroke recurrence but recurrence of PE. While the results of PICSS might be misinterpreted to have demonstrated similar recurrent PE risk in those patients with and without an ASA, what was demonstrated is how important it is to take into account how risk factors that may predict the likelihood of PE recurrence also predict the likelihood that the index event was a PE. While patients with both PFO and ASA in PICSS were at similar risk compared to PICSS patients with other causes of stroke, they could well have been at very high risk for recurrent PE, and may well have benefited from closure.

Limitations of Prior Studies Examining Recurrence Risk

The prior studies examining risk factors for recurrence are of limited value when taken individually in selecting patients for PFO closure. The largest study to look at risk factors for recurrence (The French PFO/ASA Study) enrolled 581 CS patients with and without PFO, it had only 36 total recurrent events, and only 21 (12 strokes and 8 transient ischemic attacks) in patients with PFO. Thus, the absence of an association between a risk factor and recurrence could well have been the result of inadequate statistical power, and there was certainly insufficient power to examine the influence of multiple risk factors simultaneously among patients with PFO.

Patent Foramen Ovale Statistics and References:

Strokes per year in the USA

Estimated Number

Source

~800,000

American Heart Association

~550,000

Matcher DB, Duncan PW. Cost of stroke. Stroke Clin Updates 1994; 5: 9-12

~750,000

Williams GR, Jiang JG, Matchar et al. Incidence and occurrence of total (first ever and recurrent) stroke. Stroke 1999; 30: 2523-2528

Frequency of cryptogenic stroke:

Estimated Percentage

Source

~31%

Di Tullio M, Sacco RL, Gopal A et al. Patent foramen ovale as a risk factor for cryptogenic stroke. Ann Intern Med 1992; 117: 461-465

~40%

Sacco RL, Ellenberg JH, Mohr JP et al. Infarcts of undetermined cause: the NINCDS Stroke Data Bank. Ann Neurol 1989; 25: 382-390

Prevalence of PFO in Patients with Cryptogenic stroke

Estimated Percentage

Source

41%

Bogousslavsky J, Garazi S, Jeanrenaud X et al. Stroke recurrence in patient with patent foramen ovale: The Lausanne Study. Neurology1996; 46: 1301-1305

42%

Di Tullio M, Sacco RL, Gopal A et al. Patent foramen ovale as a risk factor for cryptogenic stroke. Ann Intern Med 1992; 117: 461-465

54%

Lechat P, Mas JL, Lascault G et al. Prevalence of patent foramen ovale in patients with stroke. NEJM 1988; 318: 1148-1152

Risk of Recurrent Stroke in Patients with PFO

(treatment not addressed)

Estimated Percentage

Source

1% per year

The French Study Group on Patent Foramen Ovale and Atrial Septal Aneurysm. Recurrent cerebrovascular events in patients with PFO or ASA and cryptogenic stroke or TIA [abstract]. Cerebrovasc Dis 1994;4: 247

2.4% per year

Bogousslavsky J, Garazi S, Jeanrenaud X et al. Stroke recurrence in patient with patent foramen ovale: The Lausanne Study. Neurology1996; 46: 1301-1305

5% or greater

Some subgroups may have particularly high risk of recurrence, according to Libman R, Wein T. “Newer” cardiac sources of embolic stroke. The Neurologist 1999; 5: 231-246

Taking high or low estimates from the above data would yield slightly different numbers with regard to the risk of recurrent stroke in patients with PFO in the USA. There are no reliable estimates on the effect of warfarin on this risk although it is widely presumed to be somewhat effective at lowering it.

Highest estimate
(750,000 x 0.4 x 0.54 x 0.05)

8,640

Medium estimate
(600,000 x 0.35 x 0.46 x 0.025)

2,415

Low estimate
(500,000 x 0.31 x 0.41 x 0.01)

635