Circulation Journal
Online ISSN : 1347-4820
Print ISSN : 1346-9843
ISSN-L : 1346-9843
Reviews
Cardiological Aspects of Stroke Prevention
Marc FisherMark McAllister
Author information
JOURNAL FREE ACCESS FULL-TEXT HTML

2015 Volume 79 Issue 2 Pages 271-277

Details
Abstract

Many cardiac disorders and their treatment are associated with an increased risk for ischemic or hemorrhagic stroke, so it is important for cardiologists to be aware of recent advances in the field of stroke prevention. Atrial fibrillation (AF) is the most common cardiac disorder associated with a substantial risk for ischemic stroke (IS). The availability of implantable cardiac monitoring devices has substantially increased the detection rate of occult AF after IS. The 4 new oral anticoagulants have advantages when compared with warfarin, the standard therapy in AF to prevent IS, demonstrating a reduced risk for IS or intracerebral hemorrhage. Patients with cardiomyopathy, cardiac valve replacement, recent myocardial infarction, larger aortic arch atheroma and patent foramen ovale all have some level of increased risk for IS. The best approach for IS prevention in these disorders remains unsettled and varying approaches are recommended. (Circ J 2015; 79: 271–277)

Ischemic stroke (IS) is an untoward consequence of many cardiology disorders and intracerebral hemorrhage is a feared side effect of therapies used to prevent IS. It is incumbent on cardiologists to be aware of the IS risks associated with disorders they commonly encounter and to choose the safest and most effective therapies to prevent IS. This review will focus on the most common cardiology disorder associated with a substantial IS risk, atrial fibrillation (AF) (Table 1). Cardiac conditions for which anticoagulation is indicated are cardiomyopathy, valvular heart disease, recent myocardial infarction (MI), aortic arch disease and patent foramen ovale (PFO). For each condition the risk of IS will be provided, and data about therapeutic approaches to IS prevention surveyed together with the risks of side effects associated with these therapies, with the intention of enabling cardiologists to approach the management of these patients rationally and decisively.

Table 1. Overview of Cardiac Conditions and Anticoagulation
Indicated Unclear whether indicated Not indicated
Mitral stenosis and history of embolism, AF,
or LA thrombus
Cardiomyopathy Mitral regurgitation
Mechanical valves Mitral stenosis and enlarged LA Aortic valve disease
AF and elevated risk scores MI with LV thrombus or apical hypokinesis Bioprosthetic valves >3 months after surgery
  Aortic arch disease and embolism Bacterial endocarditis
  PFO-related stroke  

AF, atrial fibrillation; LA, left atrium; MI, myocardial infarction; LV, left ventricle.

AF

The risk of IS in AF patients is substantial and several risk stratification scores based on age, sex, and the presence of concomitant disorders have been developed. The most widely used risk score is the CHADS2 score, where C stands for congestive heart failure, H for hypertension, A for age ≥75 years, D for diabetes and S for prior stroke or transient ischemic attack (TIA).1 Each component when present is given a score of 1, except for prior stroke/TIA, which is scored 2 (Table 2). The annual risk of IS is 1.9% for a score of 0, 2.8% for a score of 1, 4.0% for a score of 2, 5.9% for a score of 3, 8.5% for a score of 4, 12.5% for a score of 5 and 18.2% for a score of 6. The IS risk with the CHADS2 score has been validated in many different studies and populations.2,3 A similar but enhanced risk score, known as the CHA2DS2-VASC, may increase IS risk prediction.4 In this score an additional point is given for age of 65–74 and age ≥75 years is scored 2. Female sex and vascular disease other than prior IS or TIA are each scored 1. The IS risk with a lower score is less than with the CHADS2 score. Both scores may underestimate the risk of subsequent stroke in IS and TIA patients without other risk factors. Another important risk stratification in AF patients is the risk for bleeding if oral anticoagulation is initiated. The HAS-BLED is a score that predicts bleeding risk (Table 2) and has been validated in a large cohort of AF patients treated with warfarin.5 Both the IS risk score and the bleeding risk score should be considered when making decisions about oral anticoagulation therapy.

Table 2. Components of the CHADS2, CHA2DS2-VASC and HAS-BLED Scores (Points)
CHADS2 CHA2DS2-VASC HAS-BLED
CHF (1) CHF (1) Hypertension (1)
Hypertension (1) Hypertension (1) Abnormal renal/liver function (1–2)
Age ≥75 (1) Age ≥75 (2) Prior stroke (1)
Diabetes (1) Diabetes (1) Bleeding history (1)
Prior stroke (2) Prior stroke (2) Labile INR (1)
  Other vascular disease (1) Age >65 (1)
  Age 65–74 (1) Drugs and/or ETOH (1–2)
  Female sex (1)  

CHF, congestive heart failure; ETOH, alcohol; INR, international normalized ratio.

The pathophysiological relationship of AF and IS is relatively obvious when patients are known to have AF prior to IS onset or the arrhythmia is observed at presentation to the hospital or shortly thereafter. However, it must be acknowledged that AF in IS patients does not necessarily imply that AF was the source of the IS. Patients may have alternative etiology for their stroke, such as large vessel atherosclerosis or small vessel disease in penetrating intracranial arteries.6 In AF patients who have stenosis or occlusion in a vessel supplying the territory involved with the infarct it can be problematic to determine if the local vascular disease was the cause or the AF. In patients with lacunar stroke that is typically ascribed to small vessel atherosclerosis concomitant AF is likely not the cause of IS and is an incidental finding. It must be acknowledged that lacunar stroke may in unusual circumstances be caused by AF, especially in patients under 75 years of age without known vascular risk factors. IS patients without known AF and who have no readily identifiable source for their stroke, which includes head and neck vessel imaging and evaluation for a cardiac source of embolization, are said to have cryptogenic stroke.7 Such cryptogenic stroke patients with a stroke in a large vessel territory and no obvious source are presumed in many cases to have undetected AF. To determine if AF is present in cryptogenic IS patients, various approaches have been used. Initially, cardiac rhythm is monitored for ≥24 h during the hospitalization, as suggested by current management guidelines.8 This rhythm monitoring approach leads to AF detection in approximately 6% of unselected patients and 13% in patients selected for a higher risk of AF detection.9 The yield increases with the length of monitoring. External loop recorders used for up to 7 days and re-applied 3–6 months later increased the yield of previously unknown AF to 14%.10 The heterogeneity and selection bias in the studies analyzed makes generalization of the results of this meta-analysis problematic. Recently, 2 randomized trials were reported that evaluated AF detection in cryptogenic stroke patients undergoing prolonged ECG monitoring vs. short-term monitoring. In the 30-day Cardiac Event Monitoring Belt for Recording Atrial Fibrillation After a Cardiac Ischemic Event (EMBRACE) study,11 572 patient with cryptogenic stroke were randomly assigned to 24-h monitoring or a 30-day event triggered monitor. The mean age of the patients was 70 years and monitoring began on average 75 days after the index stroke. AF lasting ≥30 s was detected in 16.1% of the 30-day event monitor group and 3.2% of the 24-h monitoring group. The Cryptogenic Stroke and Underlying AF (CRYSTAL AF) trial compared the time to detection of AF in 441 IS or TIA patients determined to be cryptogenic after an extensive baseline evaluation for potential sources with 2 cardiac rhythm monitoring strategies.12 The control group had ECG monitoring performed at the discretion of the investigator and the other group had an implantable cardiac monitoring device. The mean time to randomization was 38 days from the ischemic event and the mean age of the patients included was 61.5 years. AF at 6 months was detected in 8.9% of the patients assigned to the inserted cardiac monitor and in 1.4% of the control group, and at 12 months the rates were 12.4% and 2.0%, respectively. The median time to AF detection was 84 and 55 days, respectively, in the 2 groups and most of the detected AF episodes were asymptomatic. The median duration of the detected AF episodes was 4.3 min. The higher rates of AF detection with the implantable device led to an increase in the initiation of oral anticoagulation. The low rate of AF detection in the control group of the CRYSTAL AF trial as compared with the EMBRACE trial may reflect the age difference of the subjects in the 2 studies or other uncertain factors. Our personal experience with 30-day event monitors in cryptogenic IS and TIA patients had a much lower yield than observed in the EMBRACE trial and has led to use of implantable cardiac monitors for AF detection in cryptogenic patients deemed to be likely to have AF as the source of their ischemic event. An important question raised by increasing detection of AF by better monitoring is what duration of AF is clinically meaningful. It has been suggested that <1 min has uncertain significance and the duration needs to be several minutes at least to support the likelihood of cryptogenic IS or TIA being related to device-detected AF.13

In AF patients who are deemed to be appropriate candidates for oral anticoagulation therapy, warfarin has been the mainstay of treatment for many years because of the clear evidence from multiple clinical trials of efficacy for both primary and secondary IS prevention.14 Evaluation of 5 clinical trials demonstrated a relative risk reduction for primary prevention of 68% (confidence interval (CI) 50–79%) of warfarin compared with placebo. One trial of warfarin vs. placebo demonstrated a similar relative risk reduction for secondary prevention. The optimal target international normalized ratio (INR) is between 2 and 3 for reducing IS risk with an acceptable bleeding risk. Aspirin has also been evaluated for prevention of IS in AF and a combined analysis of 3 trials demonstrated a relative risk reduction of 21% (CI 0–38%) compared with placebo, which was not statistically significant.15 The combination of aspirin plus clopidogrel vs. aspirin in AF patients who were not anticoagulation candidates was evaluated in the Atrial Fibrillation Clopidogrel with Irbesartan for Prevention of Vascular Events trial (ACTIVE A).16 IS outcomes were significantly lower with the combination therapy (2.4% per year vs. 3.3%, relative risk reduction 0.72, CI 0.62–0.83), but major bleeding risk was increased (2% per year vs. 1.3%), negating the benefit for ischemic event outcome. A companion trial, ACTIVE W, compared the combination of aspirin plus clopidogrel with warfarin monotherapy in patients with at least 1 stroke risk factor, but was terminated prematurely because of clear superiority in the warfarin arm.17

A major development for primary and secondary prevention in AF patients has been the development of a new generation of anticoagulants (rivaroxaban, apixaban and edoxaban) that exert their effects on the coagulation cascade by directly inhibiting thrombin, dabigatran, or factor Xa. Dabigatran, rivaroxaban and apixiban are approved for use in the USA and many other countries. Dabigatran was the first drug to report trial results for the 110-mg and 150-mg twice daily doses.18 When compared with warfarin, the 150-mg dose significantly reduced IS and systemic embolism outcomes and intracranial hemorrhage (P<0.001 for superiority). Major gastrointestinal hemorrhage was significantly increased with this dose of dabigatran, as was dyspepsia. A small but significant increase in MI was also observed with dabigatran 150-mg twice daily, but in a revised analysis that included previously unrecognized MI outcomes the difference was no longer significant.19 The 110-mg dose of dabigatran twice daily did not significantly reduce IS outcomes but there were significantly fewer intracranial hemorrhages. In the USA, only the 150-mg dose of dabigatran was approved by regulatory authorities, but curiously a 75-mg dose was approved for patients with impaired renal function. Rivaroxaban 20-mg once daily (or 15 mg in patients with reduced creatinine clearance) was also compared with warfarin, using the same outcome and safety measures as in the dabigatran trial.20 Noninferiority was demonstrated in the intention-to-treat analysis when rivaroxaban was compared with warfarin (P<0.001 for noninferiority), but the rate of IS outcomes was not significantly different. The incidence of intracranial hemorrhage was significantly lower with rivaroxaban, but not the overall major bleeding side effects. Major gastrointestinal bleeding was significantly increased in the rivaroxaban group. For the warfarin arm of the study, the median percentage of time the INR was in the target range was 58%, the lowest percentage among the trials of the new anticoagulants (Table 3). Apixaban was studied in 2 clinical trials. In the first,21 apixaban 5 mg twice daily was compared with aspirin (81–324 mg/day) in AF patients who were considered not to be candidates for warfarin. The primary outcome measure was the same as in the other trials; cases of IS, system embolism or hemorrhagic stroke were significantly fewer with apixaban than with aspirin (1.6% vs. 3.7% per year, P<0.001) with no difference in major bleeding risk. The risk of both IS and systemic embolism was significantly lower with apixaban. The study was stopped after the first interim analysis because of the clear superiority of apixaban. In the second study,22 apixaban was compared with warfarin in AF patients deemed appropriate for anticoagulation. The same primary outcome measure was used and was significantly reduced with apixaban (1.27% vs. 1.60% per year, P=0.01 for superiority). The superiority of apixaban to warfarin was driven by a significant 49% reduction in hemorrhagic stroke (P<0.001), but no difference in IS outcomes was observed. The rate of major bleeding was significantly lower with apixaban (P=0.0002). In both apixaban studies and the rivaroxaban study, no significant differences in the rates of MI were seen. The last drug to have clinical trial results reported was edoxaban. The primary outcome was the same as in the trials of the other new oral anticoagulants and patients were randomized to low- or high-dose edoxaban (30 or 60 mg) once daily vs. warfarin.23 The edoxaban dose was halved if the creatinine clearance was 30–50 ml/min, the patient weighed <60 kg or was taking verapamil, quinidine or dronedarone. The primary outcome occurred in 1.57% of the high-dose edoxaban patients vs. 1.80% of the warfarin patients in the intention-to-treat population (P=0.08). In a modified intention-to-treat analysis that included patients receiving at least 1 dose of the study drug, the primary outcome occurred in 1.18% of high-dose edoxaban patients vs. 1.50% per of the warfarin patients (P=0.02 for superiority). In both analyses, no difference in IS outcomes was observed, but hemorrhagic stroke was significantly reduced in the high-dose edoxaban group. In the low-dose edoxaban group, the intention-to-treat analysis demonstrated a trend towards more primary outcome events with edoxaban, which was significant in the modified intention-to-treat analysis. More IS occurred with low-dose edoxaban than with warfarin, but fewer hemorrhagic strokes. Major bleeding side effects were significantly reduced with both doses of edoxaban and lower in the low-dose group. Edoxaban was recently approved for stroke prevention in Japan and an FDA advisory panel in the USA recommended approval.

Table 3. Overview of the Clinical Trial Data for New Anticoagulant Drugs vs. Warfarin
  Dabigitran
150 mg
Rivaroxaban
20 mg
Apixaban
5 mg
Edoxaban
60 mg
Mean age (years) 71.5/71.6 73/73 70/70 72/72
Mean CHADS2 2.2/2.1 3.5/3.5 2.1/2.1 2.8/2.8
Prior stroke or TIA (%) 20/20 55/55 19/18 28/28
Prior warfarin (%) 50/49 62/63 57/57 59/59
Median time in therapeutic range (%) 67 58 66 68

TIA, transient ischemic attack.

In addition to anticoagulation, devices have been used to attempt to isolate the left atrial appendage, the main source of emboli to the brain, and to isolate the region of pulmonary vein insertion into the left atrium (LA) by radiofrequency ablation. In the Watchman Left Atrial Appendage System for Embolic Protection in Patients with Atrial Fibrillation open-label trial,24 AF patients were randomized to device closure of the LA or warfarin. Warfarin was stopped after 45 days in the closure group. Device closure was found to be noninferior to warfarin for the primary endpoint of stroke, systemic embolism or cardiac death, but had significantly more complications, primarily pericardial effusion, procedure-related stroke and device embolization. With experienced operators, the complication rate may be substantially reduced, so LA appendage closure may be considered in AF patients refractory to medical therapy or who cannot be anticoagulated.25 Another device approach to AF is catheter-based, radiofrequency ablation, especially in symptomatic patients who do not respond to antiarrhythmic drugs. The efficacy of ablation therapy in clinical trials for maintaining sinus rhythm at 12 months ranges from 66% to 86% vs. 9–58% with medication.26,27 The main side effects of the procedure are cardiac tamponade, stroke and air embolism. In a large single-center case series, in which almost 80% of AF patients were arrhythmia-free off medication, anticoagulation could be discontinued in 76.5% with CHADS2 score ≤2.28 Despite these encouraging results of ablation, its efficacy in reducing IS risk remains unclear.

The clinical trial results for the 4 new oral anticoagulants provide a comparison for each one to warfarin, but no data are available on how each of the new agents compare with one another. Extrapolation and clinical judgment are necessary when making treatment decisions for individual AF patients. Cost may be an issue for some patients, which is substantially higher than for warfarin for many patients in the USA and may influence patients’ decision on medication for stroke prophylaxis. In addition to cost, other factors will influence the physician’s recommendations for individual AF patients. In AF patients with the major concern of preventing IS (ie, those with a high CHADS2 or CHA2DS2-VASC score), we would recommend that dabigatran be used because it is the only one of the new oral anticoagulants that significantly reduced the risk of IS when compared with warfarin. For AF patients in whom intracranial hemorrhage or other bleeding risk is the primary concern (ie, a patient who had a bleeding event while on warfarin), we would recommend that apixaban be used because it appears to have the lowest bleeding risk according to the clinical trial data. Although both dabigatran and rivoroxaban reduced the risk of intracranial hemorrhage, they both had an increased risk of major gastrointestinal hemorrhage when compared with warfarin, so their use in patients with a bleeding concern may be problematic.

Cardiomyopathy

Cardiomyopathy increases the risk of embolic IS, even in patients in sinus rhythm. The rates of IS in those with cardiomyopathy without AF range from 1% to 2.2% annually.2934 The best approach to antithrombotic treatment has yet to be established, despite intensive inquiry. The Heart Failure Long-term Antithrombotic Study (HELAS) study failed to demonstrate a difference in the rate of embolization in patients with either ischemic or dilated cardiomyopathy treated with aspirin or warfarin.30 The Warfarin/Aspirin Study in Heart failure (WASH) study similarly showed no difference in embolization rates in patients assigned to aspirin, warfarin, or no treatment.32 The Warfarin and Antiplatelet Therapy in Chronic Heart Failure (WATCH) trial compared aspirin, clopidigrel, and warfarin and also found no difference among these therapies for the combined primary endpoint of death, MI, or IS.33 By far the largest study was Warfarin vs. Aspirin in Reduced Cardiac Ejection Fraction (WARCEF),34 in which 2,305 patients were randomized to receive aspirin or warfarin, with no significant difference between groups in terms of the combined primary endpoint of IS, intracerebral hemorrhage, or death. However, in keeping with previous studies, the rate of IS was lower in the warfarin group, but with an increased rate of major hemorrhage. The results of these trials are summarized in Table 4. A meta-analysis of these randomized trials supported the conclusion that treatment of non-AF cardiomyopathy patients with warfarin resulted in fewer strokes with a relative risk of 0.59 (95% CI 0.41–0.85). However, there were significantly more major hemorrhages, with a relative risk of 1.95 (95% CI 1.37–2.76). Mortality was not significantly different between the aspirin and warfarin groups.35

Table 4. Overview of the Clinical Trial Data for Anticoagulation in Cardiomyopathy
  HELAS WASH WATCH WARCEF
No. of patients 197 279 1,587 2,305
Aspirin dose (mg) 325 300 162 325
INR target 2.0–3.0 2.0–3.0 2.5–3.0 2.0–3.5
Stroke RR (95% CI) 0.66 (0.10–4.58) 0.2 (0.01–4.1) 0.58 (0.36–1.32) 0.61 (0.40–0.92)
Major hemorrhage RR (95% CI) 10 (0.58–171.95) 4.09 (0.47–35.88) 1.43 (0.81–2.52) 2.17 (1.43–3.30)
Mortality RR (95% CI) 0.96 (0.44–2.10) 0.83 (0.50–1.38) 0.95 (0.73–1.23) 1.04 (0.89–1.20)

Relative risk (RR) calculated as warfarin condition compared with aspirin condition. CI, confidence interval; INR, international normalized ratio.

The optimal antithrombotic regimen for patients with cardiomyopathy remains unclear. Presently, the treatment choice of antithrombotic drugs should be individualized with respect to the perceived risks of IS and hemorrhage. The role of novel oral anticoagulants in patients with cardiomyopathy is undetermined.

Valvular Heart Disease

The risks of IS in patients with valvular heart disease varies according to which valve is affected and disease etiology. Patients with mitral stenosis who have a history of IS or other embolic event, a demonstrated atrial thrombus, or concurrent AF should be anticoagulated. Mitral stenosis with enlarged LA (≥55 mm in diameter) is associated with increased IS risk and anticoagulation should be considered.36 In contrast to mitral valve stenosis, the risk of IS in patients with mitral valve prolapse is less clear.37,38 Mitral regurgitation in patients without AF probably does not significantly increase the risk of IS.39 There have been conflicting results regarding whether mitral annular calcification increases the risk of IS.40,41 Diseases of the aortic valve, including aortic regurgitation, stenosis, sclerosis, and annular calcification, do not appear to independently increase the risk of IS.42

Valve replacement increases the risk of IS via thromboembolism and this risk is greater with mechanical than bioprosthetic valves. Most patients with mechanical valves should be treated with warfarin and aspirin, with a target INR of 2.0–3.0 for low-risk aortic valve patients, or of 2.5–3.5 for patients with additional risk factors or mitral valves.43 The Randomized, Phase II Study to Evaluate the Safety and Pharmacokinetics of Oral Dabigatran Etexilate in Patients after Heart Valve Replacement (RE-ALIGN) trial compared warfarin and dabigatran in patients with mechanical heart valves and was terminated early because of excessive bleeding without additional benefit for dabigatran.44 Presently, there is no role for the novel oral anticoagulants in the treatment of valvular disease. Though patients with bioprosthetic valves are at increased risk of IS, anticoagulation is not indicated beyond the first 3 months. There is no role for treatment of bacterial endocarditis with anticoagulation absent another indication; instead appropriate antibiotic therapy is the definitive treatment.36

Acute MI

Acute MI is a risk factor for IS, adding to disability and mortality; within 30 days of an MI, 12.2 patients per 1,000 will suffer IS.45 Acute MI can cause IS via ventricular hypokinesis, arrhythmia, increased sympathetic activity, and less commonly, hypoperfusion because of pump failure. Additionally, treatment for acute MI can lead to IS via atheroembolism and intracerebral hemorrhage when treatment with anticoagulation or fibrinolysis is initiated. Unfortunately, the prognosis after IS and acute MI is poor.46

Patients with acute MI and left ventricular (LV) thrombus are at especially high risk of IS. A meta-analysis of studies of anterior MI demonstrated a rate of embolization of approximately 11% in patients with LV thrombus.47 In 2 large meta-analyses, each with more than 24,000 patients, treatment with warfarin was associated with a decreased risk of IS after MI, but a greater risk of significant bleeding.48,49 Current AHA guideline recommendations are that anticoagulation is reasonable in patients with LV thrombus, as well as in patients with anterior apical akinesis or dyskinesis.50 Patients with an anterior MI should be evaluated with echocardiography to assess for LV thrombus. The use of novel oral anticoagulants in the context of LV thrombus has not been investigated and their potential role in treatment remains unclear.

Aortic Arch Disease

Aortic atherosclerosis may lead to IS by thromboembolism or atheroembolism, particularly if the atheroma is instrumented. Thromboembolism is much more common. The prospective French Study of Aortic Plaques in Stroke study demonstrated that aortic wall thickness ≥4 mm identified on transesophageal echocardiogram was strongly predictive of IS.51 Ulcerated and “soft” plaque morphologies without calcification appear to confer higher risk.52 As the aortic arch is insufficiently visualized by transthoracic echocardiography, transesophageal echocardiography should be considered in the evaluation of IS in selected patients in whom an etiology is not evident.

The recently published Aortic Arch Related Cerebral Hazard (ARCH) trial53 was terminated early because of lack of funding. Consequently, the trial was under-powered, randomizing only 349 of the planned 744 patients. The primary combined outcome occurred in 7.6% of patients treated with aspirin plus clopidigrel compared with 11.3% of warfarin-treated patients (P=0.2). Though ARCH represents the only randomized trial for treatment of aortic plaque to date, the lack of sufficient power caused the results to be inconclusive. A small, nonrandomized study demonstrated decreased risk of recurrent IS in patients with aortic plaque presenting with systemic embolism treated with warfarin.54 A larger, also nonrandomized study showed a decreased rate of embolization in warfarin-treated patients.55 Subgroup analyses of patients with aortic plaque in trials investigating treatment with anticoagulation for AF demonstrate conflicting results.56,57 The role of the novel oral anticoagulants for prevention of IS in patients with aortic plaque remains undefined at this time. Given the lack of data, selection of therapy based on individual patient characteristics, including risk factors for thromboembolism and individualized hemorrhage risk, is reasonable.

Statin therapy in patients with aortic arch disease was associated with a markedly reduced risk of IS in a retrospective study, with an odds ratio of 0.3 (95% CI 0.2–0.6) compared with placebo for IS, TIA, or peripheral emboli.58 As with all patients with IS and atherosclerosis, patients with aortic plaque should be treated with statins. Because of the high associated risks, surgical therapy of aortic endarterectomy or stenting should not be performed to reduce IS risk.59

PFO

The relationship of IS with PFO has been established in many case-control studies, especially in patients younger than 55 years.60 Stroke risk with PFO appears to be greater when there is an associated interatrial septal aneurysm (IAS).61 Because PFO is so commonly found in healthy adults, a possible association with IS is only made when other potential explanations have been excluded. When IS is ascribed to a PFO, proximal and distal venous clots in the legs/pelvis should be sought by ultrasound and MRI techniques, but it must be acknowledged that the yield is low.62 In patients with PFO-related IS, the risk of recurrent stroke is low, approximating 2.5% annually, and the recurrence risk is apparently greater when PFO and IAS coexist. Both medical therapy and device closure of the PFO were evaluated in clinical trials. In a substudy of the Warfarin Aspirin Reinfarction Stroke Study, no significant outcome difference was observed in cryptogenic IS patients who were treated with warfarin compared with aspirin.63 Three randomized clinical trials6466 compared device closure of the PFO (Starflex or Amplatzer PFO closure devices) to medical therapy that could be either antiplatelet therapy or warfarin at the discretion of the investigator. No significant differences in the primary outcome measures of fatal or nonfatal IS and death (TIA was included in the Starflex trial) were observed. A meta-analysis of the 3 trials looking at IS as the outcome demonstrated a trend towards a better outcome with device closure of the PFO, but it was not significant.67 In the larger Amplatzer device trial, patients with an IAS and PFO did significantly better with device closure of the PFO than with medical therapy, but this subgroup benefit was not seen in the other 2 trials. The rate of recurrent IS was very low in all 3 trials with medical therapy, reinforcing the contention that PFO-related stroke is a low-risk condition. Currently, medical therapy with either antiplatelet drugs or warfarin is recommended, unless a venous thrombosis is identified and anticoagulation is necessary. Device closure remains of unproven value.

Disclosures

Dr Fisher is compensated by the American Heart Association as the Editor-in-Chief of Stroke.

References
 
© 2015 THE JAPANESE CIRCULATION SOCIETY
feedback
Top