Secondary Prevention of Cerebral Ischemia in Patent Foramen Ovale: Systematic Review and Meta-analysis.

ABSTRACT

Background. Therapy for patent foramen ovale (PFO), found with increased frequency in the setting of stroke in the young, is controversial. We compiled a synthesis of the literature on therapeutic options for PFO and stroke.

Methods. Searching MEDLINE (1966 to 1999), published abstracts, and bibliographies of primary and review articles, we identified five retrospective cohort studies that compared at least two different treatment options (antiplatelet therapy, warfarin, or surgery). Outcomes data on mode of therapy and recurrent stroke were abstracted, and pooled analysis was done using a random effects model.

Results. Warfarin was superior to antiplatelet therapy in preventing recurrent ischemic events (odds ratio [OR] 0.37; 95% confidence interval [CI], 0.23 to 0.60). Surgical PFO closure was comparable to warfarin treatment (OR 1.19; 95% CI, 0.62 to 2.27).

Conclusion. Available data suggest that warfarin is superior to antiplatelet therapy and comparable to surgical PFO closure for the prevention of recurrent cerebral ischemic events from paradoxical embolism.

YOUNG PATIENTS with stroke are frequently found to have a patent foramen ovale (PFO). [1-3] However, because of the difficulty in proving a causal relationship between the PFO and a cerebral ischemic event, the diagnosis of paradoxical cerebral embolism is often presumptive, and therapeutic decision making is complicated. The currently available active treatment options for PFO and paradoxical embolism include antiplatelet therapy, anticoagulation, and surgery. To date, no published, randomized, controlled clinical trials have compared clinical outcomes among the therapeutic modalities used for PFO and ischemic stroke. The objective of our report is to qualitatively and quantitatively synthesize the available literature on therapy to prevent recurrent events in young patients with ischemic stroke and PFO.

METHODS

We searched MEDLINE (1966 to 1999) using the terms patent foramen ovale, stroke, and treatment. We also searched the bibliographies of reports and review articles for additional references missed by the computerized database search. Unpublished study results were searched using the National Institutes of Health registry of funded studies. Five retrospective cohort studies were identified that compared at least two different treatment options for the secondary prevention of stroke from PFO. [4-8] All five studies compared warfarin with antiplatelet therapy. Four studies also compared medical therapy with surgical closure of the PFO. [4-7]

Recurrent stroke or transient ischemic attack (TIA) as defined within each study for the therapeutic categories of warfarin, antiplatelet therapy, and surgery was abstracted as the outcome of interest. In one study [1] that found no difference in outcome between warfarin, antiplatelet therapy, and surgery but did not report specific event data within each subgroup, recurrent events were evenly distributed across treatment groups. Based on qualitative similarities in patient populations (mean age and sex distribution), the duration and method of follow-up, and the criteria used to determine recurrent cerebral ischemic events, as well as quantitative evidence of a lack of significant heterogeneity ([[chi].sup.2] test for heterogeneity, [[chi].sup.2] = 4.41), [9] the data were pooled for meta-analysis.

Contingency tables were constructed from abstracted data from individual studies. A summary odds ratio (OR) and 95% confidence interval (CI) was calculated using a random effects model [9] to provide a conservative estimate of the effect size for the outcome of interest.

RESULTS

All five studies were retrospective cohort studies among predominantly young, male stroke patients (mean age ranging from 38 to 48 years) (Table). The primary diagnosis in all five studies was PFO. A single study [8] included patients with an atrial septal aneurysm (with or without PFO) in the report of therapy and outcomes. Therapies included anticoagulation (warfarin and antiplatelet therapy, five studies) and surgical PFO closure (four studies). Two studies [4,5,8] also reported the use of no active therapy (n = 14 patients). Therapies were non-randomly assigned in all five studies, but two studies indicated that warfarin or surgery was primarily used for patients considered at higher risk of recurrent stroke. [4,6] The primary method of outcome assessment (four studies) was indirect follow-up (surveys and record review). [5-8] Follow-up rates ranged from 86% to 100%.

Hanna et a1 [6] reported the first experience in secondary prevention of stroke associated with PFO. A total of 15 patients received nonrandomized therapies: surgery (n = 1), transvenous closure device (n = 1), warfarin (n 7; prothrombin time 1.3 to 1.5 times control), and aspirin (n = 6; 325 mg/day). No recurrent cerebral ischemic events were reported after a mean of 28 months of therapy.

Hausmann et a1 [7] reported on the relationship between treatment and recurrent events in 44 patients with stroke and PFO. Three recurrent events were identified during a mean 59-month follow-up. Recurrent events (1 TIA, 1 stroke) occurred in 2 patients on antiplatelet therapy (n = 17) and 1 patient (TIA) on warfarin (n = 24). There were no recurrent events in patients who had surgical closure (n = 3).

Mas et al [8] reported on 92 stroke patients with either PFO or atrial septal aneurysm who received warfarin therapy (international normalized ratio [INRI, 2 to 3), antiplatelet therapy, or no therapy. Over 23 months, recurrent cerebral ischemic events (4 TIAs, 2 strokes) occurred in 6 patients, all of whom received either antiplatelet therapy (n =5) or no therapy (n = 1).

Bogousslavsky et a1 [4] studied 140 patients under the age of 60 years who were admitted to a population-based, primary care stroke center from 1988 to 1994 with PFO and stroke. Therapy for the secondary prevention of stroke was assigned based on clinical characteristics, including the presence of an atrial septal aneurysm, a major right to left shunt using saline contrast echocardiography, a history of multiple events (clinical or silent), or a stroke precipitated by a valsalva maneuver. Surgery was done in patients with two clinical risk factors (n = 11), warfarin was used in patients who declined surgery or who had one clinical risk factor (n = 37; target INR, 3 to 4), and all remaining patients (n = 92) received aspirin (250 mg/day). Recurrent events (8 TIAs, 8 strokes; identified in regular neurology clinic follow-up) occurred in 16 patients over 36 months, and mode of therapy was unrelated to outcome.

Cujec et al [5] used the echocardiography data base of a university-based neurology referral center to identity 52 patients with PFO and stoke. Therapies used in these patients (with some cross-over among therapies) were surgical closure (n 14), warfarin (n = 38), antiplatelet therapy (n = 36), or no therapy (n = 8). Over a mean follow-up period of 42 months, 19 confirmed recurrent cerebral ischemic events (12 TIAs, 7 strokes) occurred, including 13 with antiplatelet therapy and 6 with oral anticoagulation. There were no recurrent events in patients treated with surgical closure or who received no therapy.

A total of 378 patients have been studied in the five reports previously summarized, including 139 treated with warfarin, 210 treated with antiplatelet therapy, and 29 treated with surgical closure. A total of 44 recurrent cerebral ischemic events have been reported after an average follow-up of 36 months (3.9% recurrent events/year), including an estimated annual recurrent cerebral ischemic event rate of 1.0% for surgical closure, 2.5% for warfarin, and 5.1% for antiplatelet therapy.

The ORs for recurrent cerebral ischemic events varied greatly among the five studies that compared adjusted-dose warfarin and antiplatelet therapy (Fig 1) and for the four studies that compared medical and surgical therapy (Fig 2). There were no statistically significant differences for these therapies in any of the individual studies. Meta-analysis showed that the OR for recurrent cerebral ischemic events for adjusted-dose warfarin versus antiplatelet therapy was 0.37 (95% CI, 0.23 to 0.60). Compared with medical therapy, surgical PFO closure was significantly protective of recurrent cerebral ischemia (OR 0.27; 95% CI, 0.11 to 0.66). Analysis within subgroups of medical therapy had limited statistical power. There was a nonsignificant trend for PFO closure to be superior to aspirin (OR 0.36; 95% CI, 0.04 to 3.09). Surgical PFO closure was comparable to warfarin (OR 1.19; 95% CI, 0.62 to 2.27).

DISCUSSION

Our meta-analysis suggests that warfarin is superior to antiplatelet therapy and is comparable to surgical PFO closure for the prevention of recurrent cerebral ischemic events related to PFO. Pending the completion of randomized clinical trials, these findings, when integrated with specific patient and therapeutic characteristics, can help guide clinical therapy for patients with cerebral ischemic events and PFO.

Multiple variables influence risks and benefits of therapy for stroke in the setting of a PFO. These include characteristics of the PFO, patient age, and the efficacy and risk of a particular therapy. The risk of recurrent cerebral ischemic events in the setting of PFO may be increased when the PFO is associated with atrial septal aneurysm. [1, 8] Additionally, large PFO size, measured by the extent of bubble transit on contrast echocardiography, may also predict a greater risk of recurrent cerebral ischemic events. [1, 10, 11] A recent Markov decision analysis by Nendaz et al [12] indicates that, as treatment complication rates increase, the threshold for active therapeutic intervention must also increase. This analysis concluded that, for all patients except those with the lowest risk of a recurrent event, surgery or anticoagulation appear to be reasonable therapeutic options. Our meta-analysis suggests that, when anticoagulation therapy is chosen for the secondary prevention of stroke from PFO, full antico agulation is more effective than antiplatelet therapy. This conclusion seems plausible when stroke from PFO is considered as an atrial-based thromboembolic event, particularly in light of the known 36% reduction in thromboembolic complications of atrial fibrillation for warfarin (maintained at an INR of 2 to 3) versus aspirin therapy. [13]

The choice between surgical PFO closure and warfarin anticoagulation is less clear. Many factors, such as patient age and predicted surgical or anticoagulalion complication rates, influence this decision. The potential benefit of warfarin anticoagulation for the secondary prevention of stroke from PFO must be counterbalanced by an attendant increased risk of major bleeding complications, a risk that is age and patient dependent. Despite the intuitive appeal of PFO closure as a means to eliminate the risk of a paradoxical cerebral embolism, the apparent benefit of a surgical procedure will be reduced by surgical complications, incomplete efficacy of surgical closure, and the presence of false-positive diagnoses of paradoxical embolism. As suggested by Nendaz et al, [12] surgery is less likely to be beneficial in older patients (age [greater than]50 years) because of the increased likelihood of a false-positive diagnosis of paradoxical embolism and the shorter duration of exposure to complications of life-long systemic anticoagulation.

In contrast, younger patients are at lower risk for anticoagulation complications, but they are exposed to these potential complications for a longer duration, are likely at lower risk during surgical closure of a PFO, and are more likely to have had a true paradoxical embolism as a cause of the cerebral ischemic event. Alternative surgical approaches, such as percutaneous PFO closure, are promising but require further development and evaluation in clinical trials. Two recent reports in patients with paradoxical embolism and PFO reported an annual risk of 2.5% [14] and 3.2% [15] for recurrent TIA or stroke after insertion of a percutaneous PFO closure device. Although these figures are roughly comparable to the pooled rate of recurrent cerebral ischemic events for surgery (1.0%) and warfarin (2.5%) from existing studies, complications related to device insertion occurred in up to 10% of patients. [14] Thus, well-conducted, randomized trials are needed to establish accurate estimates of the risk of recurrent ischemic events in the setting of PFO, as well as the long-term efficacy and complication rates of available therapies.

Our meta-analysis has several limitations that arise from inherent limitations in the reported experience on PFO and recurrent cerebral ischemic events. First, all the studies were retrospective, observational studies involving few patients treated with surgical PFO closure. Any differential outcomes observed among interventions may well be explained by the biases inherent in observational studies, though the magnitude of discrepancy between randomized trials and observational studies has recently been questioned. [16-18] None of the studies controlled for potentially confounding factors, so our meta-analysis is an unadjusted, pooled synthesis. Thus, there is a great need for randomized, controlled trials to establish an unbiased assessment of the relative efficacy of surgical and medical therapies for the secondary prevention of stroke from PFO.

Based on the available clinical data derived from young to middle-aged patients with an overall intermediate to high risk (3.8% per year) of recurrent cerebral ischemic events, warfarin appears to be superior to antiplatelet therapy and comparable to surgical PFO closure in the prophylaxis of recurrent cerebral ischemic events. The decision between these modalities should be individualized, taking into consideration patient age and the balance between surgical and long-term anticoagulation risk. The completion of randomized clinical trials, such as the PFO in Cryptogenic Stroke Study, a multicenter, controlled clinical trial, will provide important data to further clarify therapeutic options in this area.

References

(1.) Cabanes L, Mas JL, Cohen A, et al: Atrial septal aneurysm and patent foramen ovale as risk factors for cryptogenic stroke in patients less than 55 years of age. a study using transesophageal echocardiography. Stroke 1993; 24:1865-1873

(2.) Lechat P, Mas JL, Lascault G, et al: Prevalence of patent foramen ovale in patients with stroke. N Engl J Med 1988; 318:1148-1152

(3.) Webster MW, Chancellor AM, Smith HJ, et al: Patent foramen ovale in young stroke patients. Lancet 1988; 2:11-12

(4.) Bogousslavsky J, Garazi S, Jeanrenaud X, et al: Stroke recurrence in patients with patent foramen ovale: the Lausanne Study. Lausanne Stroke with Paradoxal Embolism Study Group. Neurology 1996; 46:1301-1305

(5.) Cujec B, Mainra R, Johnson DH: Prevention of recurrent cerebral ischemic events in patients with patent foramen ovale and cryptogenic strokes or transient ischemic attacks. Can J Cardiol 1999; 15:57-64

(6.) Hanna JP, Sun JP, Furlan AJ, et al: Patent foramen ovale and brain infarct. echocardiographic predictors, recurrence, and prevention. Stroke 1994; 25:782-786

(7.) Hausmann D, Mugge A, Daniel WG: Identification of patent foramen ovale permitting paradoxic embolism. J Am Coll Cardiol 1995; 26:1030-1038

(8.) Mas JL, Zuber M: Recurrent cerebrovascular events in patients with patent foramen ovale, atrial septal aneurysm, or both and cryptogenic stroke or transient ischemic attack. French Study Group on Patent Foramen Ovale and Atrial Septal Aneurysm. Am Heart J l995; 130:1083-1088

(9.) DerSimonian R, Laird N: Meta-analysis in clinical trials. Control Clin Trials 1986; 7:177-188

(10.) Homma S, Di Tullio MR. Sacco RL, et al: Characteristics of patent foramen ovale associated with cryptogenic stroke. a biplane transesophageal echocardiographic study. Stroke 1994; 25:582-586

(11.) Steiner MM, Di Tullio MR, Rundek T, et al: Patent foramen ovale size and embolic brain imaging findings among patients with ischemic stroke. Stroke 1998; 29:944-948

(12.) Nendaz MR, Sarasin FP, Junod AF, et al: Preventing stroke recurrence in patients with patent foramen ovale: antithrombotic therapy, foramen closure, or therapeutic abstention? a decision analytic perspective. Am Heart J 1998; 135:532-541

(13.) Hart RG, Benavente O, McBride R, et al: Antithrombotic therapy to prevent stroke in patients with atrial fibrillation: a meta-analysis. Ann Intern Med 1999; 131:492-501

(14.) Windecker S, Wahl A, Chatterjee T, et al: Percutaneous closure of patent foramen ovale in patients with paradoxical embolism: long-term risk of recurrent thromboembolic events. Circulation 2000; 101:893-898

(15.) Hung J, Landzberg MJ, Jenkins KJ, et al: Closure of patent foramen ovale for paradoxical emboli: intermediate-term risk of recurrent neurological events following transcatheter device placemcnt. J Am Coil Cardiol 2000; 35:1311-1316

(16.) Pocock SJ, Elboume DR: Randomized trials or observational tribulations? (Editorial). N Engl J Med 2000; 342:1907-1909

(17.) Concato J, Shah N, Horwitz RJ: Randomized, controlled trials, observational studies, and the hierarchy of research designs. N Engl J Med 2000; 342:1887-1892

(18.) Benson K, Hartz AJ: A comparison of observational studies and randomized, controlled trials. N Engl J Med 2000; 342: 1878-1886

TABLE.
Characteristics of Studies of Patent Foramen Ovale (PFO) and
Recurrent Cerebral Ischemic Events
                           No.     Mean Age  Follow-up  No. Events
Study                    Subjects    (yr)      (mo)       Surgery
Bogousslavsky et al [4]    140        --        36       1/11 (9.1%)
Cujec et al [5]             88        38        42       0/14 (0%)
between
Hanna et al [6]             15        43        28        0/1 (0%)
Hausmann et al [7]          44        48        59        0/3 (0%)
Mass and Zuber [8]          92        40        23          NA
Study                      Warfarin    Antiplatelet Therapy
Bogousslavsky et al [4]  4/37 (10.8%)     11/92 (12.0%)
Cujec et al [5]          6/38 (15.8%)     13/36 (36.1%)
between
Hanna et al [6]           0/7 (0%)         0/6  (0%)
Hausmann et al [7]       1/24 (4.2%)       2/17 (11.8%)
Mass and Zuber [8]       0/33 (0%)         5/59 (8.4%)
Study                         Comments
Bogousslavsky et al [4]  Estimated recurrence risk:
                          surgery=warfarin=aspirin
Cujec et al [5]          Includes cross-overs
between
                          anticoagulation therapies
Hanna et al [6]          Warfarin group includes
                          aspirin failures
Hausmann et al [7]       Only 86% follow-up
Mass and Zuber [8]       Included patients with PFO
                          and atrial septal aneurysm
NA = Not applicable,
PFO = patent foramen ovale.

KEY POINTS

* Warfarin is superior to antiplatelet therapy and is comparable to surgical PF0 closure in preventing recurrent cerebral ischemic events related to PFO.

* Multiple variables influence the risks and benefits of therapy for stroke in the setting of a PFO.

* There is an important need for randomized, controlled trials to establish an unbiased assessment of the relative efficacy of surgical and medical therapies for the secondary prevention of stroke from PFO.