Online newspaper of professor Yasser Metwally

The author: Professor Yasser Metwally

http://yassermetwally.com

ATRIAL FIBRILLATION, ITS ROLE IN STROKE

Atrial fibrillation (AF) is a common cardiac arrhythmia that is found in approximately 0.4% of adults. 53 64 It is actually the most common chronic cardiac arrhythmia, and it affects 2% to 4% of subjects greater than 60 years of age . 67 It is the only common cardiac arrhythmia associated with a rapid and irregular- irregular ventricular rate. AF is seen in association with rheumatic heart disease (RHD), hypertension, hypertrophic cardiomyopathy, mitral valve prolapse, dilated cardiomyopathy, atrial septal defect, and thyrotoxicosis. 89 “Lone” atrial fibrillation refers to AF without coexistent cardiac disease,9 and it is associated with a relatively benign prognosis in terms of thromboembolism. Of note, lone AF is seen in 5% to 8% of patients with paroxysmal AF and approximately 0.6% of patients with chronic AF. 44, 105

AF is associated with a relatively high risk of thromboembolism. In association with RHD, there is a 17-fold increased risk of stroke. 105 AF without RHD, termed nonvalvular AF (NVAF), is associated with a 5.6-fold increased risk of stroke even after age, sex, and coexistent hypertension are taken into account. This excess risk is independent of associated cardiac failure and coronary artery disease as well. 106 In one study, there was a distinct clustering of stroke events at the time of onset of AF, and the recurrence rate for stroke was most pronounced within the first 6 months of the initial stroke in those with AF (47%) compared to those without AF (20%). 106

CHRONIC VERSUS PAROXYSMAL ATRIAL FIBRILLATION

AF often occurs paroxysmally before it becomes the established rhythm. 89 It can be precipitated by chest trauma, surgery, pericarditis, or excessive alcohol intake in subjects with normal hearts. 89 According to one study, paroxysmal AF occurred in 35% of 1212 patients hospitalized with AF. 30 This percentage can be difficult to corroborate, however, and it depends on the methodology used to establish the nature of the arrhythmia. 58 The paroxysmal form of AF appears more likely to occur in younger subjects. 89

FACTORS ASSOCIATED WITH RISK OF THROMBOEMBOLISM IN ATRIAL FIBRILLATION

The most important aspect of medical intervention in AF is to identify the higher risk patient. As mentioned previously, lone AF, especially in someone less than 60 years of age, is associated with a relatively low risk of thrombo-embolism. 52 In one multivariate analysis of stroke risk, AF in and of itself was not found to be a significant risk factor. 18 This has led to an intensive search for coexistent factors that compound the risk of thromboembolism in AF. 47 One such schema is outlined in Table 1. 34

Table 1. STRATIFICATION OF RISK OF THROMBOEMBOLISM IN PATIENTS WITH ATRIAL FIBRILLATION

Not only is the presence or absence of coexistent factors important, but of equal importance is the presumptive stroke mechanism. Certainly not all strokes that occur in patients with AF are cardioembolic in nature. It has been estimated that between 19% and 75% of strokes in patients with AF are cardioembolic. 11, 36, 63 Although review of the literature suggests that 15% to 20% of all ischemic strokes are directly related to AF, 31,34 it can be difficult to discern whether the AF was the precipitating factor or a sequela of the stroke. 65

• Left Atrial Thrombus

The demonstration of a thrombus within the left atrial appendage in a stroke patient with AF certainly supports a cardioembolic mechanism. In one study, 62% of subjects with chronic AF had thrombi of the left atrial appendage and only a minority had coexistent RHD. 17 The detection of such thrombus formation has been limited by the sensitivity of routine transthoracic echocardiography. This sensitivity appears to be significantly enhanced with the use of transesophageal echocardiography. 3, 111 The enthusiasm for this procedure, in terms of its increased sensitivity for the detection of potential intracardiac sources of cerebral emboli, must be tempered by the persistent relatively low yield and concerns about the clinical significance when abnormalities are found. 76

• Left Atrial Enlargement

There have been reports of an association between left atrial enlargement in AF and stroke risk. 13, 60 In one study, 18 of the 20 stroke patients with AF (90%) had left atrial enlargement, compared with 20% of controls who had AF without stroke. 13 This study has been criticized for its retrospective nature and failure to take into account duration of the AF. 98 A correlation has been suggested between left atrial enlargement and the duration of AF. 37, 96 This issue was addressed in a prospective fashion by Petersen et al. These authors compared echocardiographic findings in subjects with long and short duration AF to controls in normal sinus rhythm. There was a positive correlation between left atrial enlargement and the presence and duration of AF. Furthermore, a significant difference between subjects with short-duration versus longer- duration AF, in terms of left atrial size, remained at 6-month follow-up. In addition, there was an increase in left atrial size in both groups of AF patients over time. Thus, left atrial enlargement appears to be a sequela of AF and not necessarily a cause of it. This suggestion has been supported by others, 50, 77 and may have implications for the success of cardioversion of AF to normal sinus rhythm. Of note, a recent study that addressed this issue found no predictive value of echocardiography for success in cardioversion. 19

• Hypertension, Congestive Heart Failure, and Left Ventricular Dysfunction

The Framingham Study, 106 found AF, in and of itself, as well as systolic hypertension, ischemic cardiac disease, age, and congestive heart failure (CHF) to be independent risk factors for stroke. It therefore appears warranted to assume that coexistence of at least some of these risk factors compounds the risk of stroke. The relationship between AF, hypertension, and increased stroke risk has been reported. 26 A trend toward increased stroke risk with a combination of AF and CHF was observed in another study, 68 but the relationship was not statistically significant. Ruocco and Most, 76 reported an increased risk of thromboembolism in AF with echocardiographic evidence of left ventricular dysfunction and segmental wall motion abnormalities.

• Prior Thromboembolism

Previous thromboembolism, including prior clinical stroke, appears to be a primary determinant of future stroke risk in AF. It has been found that the risk of recurrent embolism, in patients with AF who present with stroke, is as high as 20% within 11 days if anticoagulant therapy is not immediately instituted, 36 although this relatively high figure is in dispute. 80 A literature review reported a recurrence rate for stroke of between 15% and 20% in the first year alone. 34 One study found a rather low early risk of recurrent embolism of 4% within the first month, but these subjects received platelet antiaggregate therapy and low-dose subcutaneous heparin. 6

• Silent Cerebral lnfarction

A relatively high frequency of asymptomatic (silent) cerebral infarction has been reported in subjects with AF. 22, 49,61 Feinberg et al reported that 8 of 24 subjects with chronic AF had silent infarcts by CT brain scan, but only 3 (13%) were compatible with cardiac emboli. 22 In a study of 30 elderly patients with chronic AF compared with 30 age- and sex-matched controls, 48% of the AF patients and 28% of controls had low-density lesions by CT brain scan. 72 Although these findings were not statistically significant, there were more than twice as many lesions in study subjects as in controls and this was significant. It is important to note that the frequency of ischemic lesions found in healthy elderly subjects without AF (28%) was unexpectedly high. Despite the fact that the CT scans were read blindly by two neuroradiologists, this study raises concern about the definition of discrete tissue loss versus localized atrophy in elderly subjects.

In a limited study of silent cerebral infarction in paroxysmal AF versus controls with normal sinus rhythm, no significant difference was found. 73 This lessened risk of silent cerebral infarction in subjects with paroxysmal AF compared with subjects with chronic AF was also reported by Sasaki et al. 83 This study did find, however, that there was an incidence of 38% in paroxysmal AF compared with 13% in normal sinus rhythm, and the incidence was 58% in chronic AF. The authors also reported an association between silent infarction and advanced age, duration of AF, and diuretic therapy. TABLE 2

Table 2. INCIDENCE OF SILENT CEREBRAL INFARCTION IN ATRIAL FIBRILLATION

• Reduction in Cerebral Blood Flow

It is possible that the reduction in regional cerebral blood flow found in chronic AF makes the subject more susceptible to cerebral ischemic insult. AF is associated with a 25% to 33% reduction in cardiac output from loss of atrial systole. 29 There is a concomitant reduction in regional cerebral blood flow in AF of between 5.5% and 17.5% which is most prominent in the younger age group (35 to 50 years) when compared with age-matched controls in normal sinus rhythm. 45 Furthermore, an increase in cerebral blood flow over time after cardioversion of AF has been reported. 71

• Carotid Artery Stenosis

A low prevalence of carotid artery stenosis ipsilateral to hemispheric infarction has been reported in patients with AF. 34 Tegler et al, 97 found that 17% of asymptomatic NVAF patients had at least 50% carotid artery stenosis. Weinberger et al, 100 reported that 20% to 30% of symptomatic patients with NVAF had carotid stenosis, which might explain their symptoms. In another study that addressed this issue, 67% of 159 patients with nonvalvular AF and stroke had ipsilateral internal carotid artery stenosis. 38 Of note, the degree of stenosis was 50% or greater in only I1 % of subjects. An interesting clinical finding, in reference to this, was reported by Harrison and Marshall, 35 who observed that TIAs in association with AF tended to be of shorter duration (<60 minutes)unless there was a coexistent carotid disease.

• Thyrotoxicosis

AF complicates thyrotoxicosis in 10% to 15% of patients and is most commonly seen in patients greater than 60 years of age, presumably reflective of underlying cardiac disease. 27 The prevalence of hyperthyroidism in AF is estimated to be 2% to 5%. 65 Several studies have reported a substantial risk for thromboembolism in patients with coexistent AF and thyrotoxicosis. 4, 43, 90, 110 Embolic risk appears to be highest for older patients, especially those with left ventricular dysfunction and CHF. 43A retrospective study of thromboembolism in thyrotoxic AF found a risk rate of 6% per year 69 which was interpreted as comparable to the risk seen with other causes of AF. 65

• Valvular Cardiac Disease

It has been reported that up to 50% of patients with RHD and mitral stenosis develop thromboembolism and that 60% of these ischemic events are cerebral. 20 Coexistent AF obviously compounds this risk, resulting in a 17-fold increased risk of stroke. 105 It is generally thought that mitral stenosis is the valve abnormality most likely to be associated with increased embolic risk, 95 although mitral regurgitation also appears to be important. 16 It has been estimated that up to 40% of subjects with mitral stenosis have coexistent AF. 85 Compilation of a number of studies suggests that the risk of thromboembolism is increased by as much as sevenfold when AF is superimposed on mitral stenosis. 59

Prosthetic heart valves are either mechanical, e.g., Starr-Edwards, Bjork-Shiley, or bioprosthetic, which are typically porcine valves. It is generally thought that mechanical valves are thrombogenic, and anticoagulant therapy is routinely instituted if there are no absolute contraindications. Bioprosthetic valves, on the other hand, do not necessarily require anticoagulant therapy. For patients with mechanical values treated with anticoagulant therapy, the risk of embolus is about 3% per year for mitral valves and 1.5% per year for aortic valves. 14 Coexistent AF can increase the yearly risk of embolus to up to 22.6% despite anticoagulant therapy. 93 The risk of embolus for bioprosthetic valves without coexistent AF and without anticoagulant therapy is reported to be 2% to 4% per year. 14 Patients with coexistent AF and bioprosthetic valves are usually treated with anticoagulant therapy because of the increased risk of thromboembolism. 5, 62

Mitral valve prolapse (MVP) can be associated with AF 84 and presumably represents an increased risk of thromboembolism. The absolute risk of stroke with isolated MVP, in a patient less than 40, is only about 1 in 6000 per year. 107 In a prospective case-control study of cerebral ischemia in MVP, a relationship was found between AF with CHF and cerebral ischemia but not between AF and MVP. 48

Mitral annulus calcification (MAC) has been reported to be associated with an increased risk of embolism, 107 but at least one study concluded that MAC was primarily a marker of atherosclerotic disease. 28 In a controlled, prospective trial of anticoagulant therapy in NVAF, however, there was a significantly higher incidence of MAC in subjects who suffered stroke (67%) than in those who did not suffer stroke (29%), and this was significant at P 0.003. 7

• Age

Flegel and Hanley, 25 looked at factors associated with stroke in patients with NVAF in a retrospective fashion. By univariate analysis, they found age (75 years, hazard ratio = 2.5) and systolic hypertension (160 mm Hg, hazard ratio = 6.4) to be significant. They also found that prior thromboembolism was a significant factor. The impact of chronic NVAF on stroke risk in the elderly was addressed by Wolf et al. 104 The authors found that the 2-year initial stroke incidence increased with age, from 6.7% in subjects 50 to 59 years, to 8.1% in subjects 60 to 69 years, 21.3% for subjects 70 to 79 years, and 36.2% for subjects 80 to 89 years. Risk factors for stroke associated with an elderly group of patients with chronic AF (mean age = 82 years) was the subject of a study by Aronow et al. 2 These authors found that 44 of the 110 patients (40%) had strokes, including 4 with rheumatic mitral stenosis. Factors that were found to be significantly associated with stroke risk, by logistic regression analysis, included prior myocardial infarction, systolic hypertension, left ventricular hypertrophy and left atrial enlargement.

• Hematologic Factors

Gustafsson et al, 32 investigated the possible relationship of disturbances in either the blood coagulation or fibrinolytic systems in NVAF compared to subjects in sinus rhythm. Hemostatic function was found to be altered in NVAF patients, with or without prior stroke, but not in subjects in sinus rhythm, with or without stroke. The authors concluded that disturbance of hemostatic function in NVAF may contribute to the increased stroke risk. Perhaps reflective of this relationship between AF and hematologic status, Lowe et al, 57 found a significantly higher mortality in stroke patients with AF whose hematocrit was 50% or higher.

RISK OF THROMBOEMBOLISM IN NONVALVULAR ATRIAL FIBRILLATION

In view of the declining incidence of RHD, NVAF has taken on increasing importance in stroke prevention. It is estimated that one third of patients with AF eventually suffer a stroke. 34 Furthermore, it is believed that 15% to 20% of all ischemic strokes have AF as the primary factor. 65 There has been some variability in stroke risk reported among various studies, however. One study of thromboembolic complications in NVAF, with a median patient age of 74 years, reported an annualized stroke rate of 5.5% without prophylactic therapy. 66 A very similar figure of 5. 1 % per year in untreated patients was reported in the Stroke Prevention in Atrial Fibrillation Study. 92 There have been reports of a somewhat lower risk, however. In the Boston Area Anticoagulation Trial, the incidence of stroke was 2.98% per year in untreated subjects. In the British Whitehall and Regional Heart Studies, 26 the annual stroke rate was 1.5% and 0.4%, respectively, although the relative risks were 6.9 and 2.3, respectively. A summary of pertinent results from studies that addressed the annual stroke incidence in NVAF is provided in Table 3.

Table 3. STROKE INCIDENCE IN NONVALVULAR ATRIAL FIBRILLATION (UNTREATED)

*The control subjects in the Boston Area Trial were not restricted from taking aspirin-containing compounds.

THE EFFICACY OF ANTITHROMBOTIC THERAPY FOR STROKE PREVENTION IN ATRIAL FIBRILLATION

• Anticoagulant Therapy

The use of anticoagulant therapy in patients with coexistent RHD is well established. It is estimated that the use of long-term anticoagulant therapy can reduce the risk of thromboembolism, which is on the order of 50%, 20 by 30% to 40%. 102

With a lower stroke risk in NVAF, one must place even greater emphasis on risks versus benefits of anticoagulant therapy. In a literature review of long- term anticoagulant therapy for ischemic cerebrovascular disease, Levine and Hirsh, 55 found a risk for major bleeding of between 2% and 22% per year. Furthermore, the risk for fatal bleeds was between 2% and 9% per year. A major concern is the finding that the risk of intracerebral hemorrhage in patients on anticoagulant therapy increases by 10-fold in patients over age 50. 103 Also of note in this study, 80% of subjects who bled had hypertension, and the risk of bleeding correlated with an increasing intensity of anticoagulant therapy. These findings are especially pertinent when one considers the increasing incidence of NVAF-related stroke with age. 114

Two major developments in the use and monitoring of anticoagulant therapy bode well for an increased degree of safety. The prothrombin time (PT) has been the primary method of monitoring the effect of sodium warfarin in North America. The PT makes use of thromboplastin reagents derived from rabbit brain. The International Normalized Ratio (INR) is based upon human brain thromboplastin reagents, which are much more sensitive to reductions in vitamin K-dependent clotting factors. 40 The INR has been adopted throughout most of Europe, and its use is being strongly promoted in North America. It is generally thought in the United Kingdom that a lower dose of oral anticoagulant is efficacious and that the historic precedent of using a PT of 2 to 2.5 times control was based on less sensitive rabbit brain thromboplastin reagents. 39

The use of a lower dose of sodium warfarin has been found to be efficacious in a number of clinical conditions. The first study involved secondary prevention after venous thrombosis, 42 and found that an, INR of 2.0 to 2.5, which corresponds to a PT of between 1.3 and 1.4 times control, was as effective as higher doses. 42 In a study of different intensities of warfarin in patients with prosthetic heart valves, 82 moderate intensity anticoagulation (PT ratio = 1.5, INR = 2.65) was found to offer the same protection as high-intensity therapy (PT ratio = 2.5, INR = 9) and was significantly safer. The importance of accurate INR equivalents for the conventional North American thromboplastins has recently been stressed. These equivalents can either be calculated from the international sensitivity index (ISI) as follows: INR = (prothrombin time ratio) ISI,74 or can be obtained from a published monogram. 75 An accurate INR equivalent is justifiably important because recent consensus meetings now recommend an INR range of 2 to 3 in AF.74

Kitchens and Flegel, 51 estimate that the relative risk of ischemic stroke in patients with NVAF versus the risk of intracranial hemorrhage secondary to anticoagulant therapy is 2.15. They also estimate that the relative risk of stroke- related death in NVAF versus anticoagulant-induced death is 2.1. Based upon these relative risks, they conclude that the benefits of anticoagulant therapy justify the risks in selected circumstances. In a retrospective study of 134 patients with NVAF 78 the rate of embolus was reduced from 5.9% per year without anticoagulant therapy to 0.7% per year with anticoagulant therapy.

• Randomized, Controlled Trials of Anticoagulant Therapy for Stroke Prevention in Atrial Fibrillation

Three published reports evaluated the risks versus benefits of anticoagulant therapy in NVAF in a prospective fashion. All three studies evaluated sodium warfarin versus placebo, and two prospectively evaluated a possible beneficial effect of aspirin therapy. 66 In the Copenhagen AFASAK study, 1007 patients were entered, with 335 randomized to warfarin, 336 to aspirin, and 336 to placebo. The end points for this study were thromboembolism (stroke, TIA, or systemic embolism), with a secondary end point of death.

The target point for warfarin therapy in the AFASAK study was an INR of 2.4 to 4.2, which corresponds to a PT of 1.5 to 1.9 times control. Over a study period of 2 years, there were 5 embolic events in the warfarin group (1.5%) versus 21 embolic events in the placebo group (6.2%). By efficacy analysis, this was statistically significant at P <0.05. In this study 21 patients in the warfarin group were withdrawn because of nonfatal haemorrhage, compared with none in the placebo group.

The SPAF study was composed of two arms: Group 1, patients who were warfarin-eligible, and Group 2, patients who were warfarin-ineligible. 91 This study also included an aspirin group in both arms. At the time of the initial report, 92 a total of 1244 patients were entered, with 588 in Group 1 and 656 in Group 2. The dose of warfarin was selected for a PT range of 1.3 to 1.8 times control, which corresponds to an INR of roughly 2.0 to 3.5.

The primary end points for this study were ischemic stroke and systemic embolism. Withdrawal of medication for reasons not related to end point events occurred at a rate of 10.9% per year for warfarin, 4.2% per year for aspirin, and 6.3% per year for placebo. Serious hemorrhagic events occurred at a rate of 1.7% per year in patients assigned to warfarin, with a 1.2% per year rate of intracranial hemorrhage. In the placebo group, there was a 1.2% per year rate of hemorrhage and a 0.2% per year rate of intracranial hemorrhage.

The preliminary intention to treat analysis for SPAF found an 81% reduction in primary events (P <00005). The Boston Area Trial, 7 looked at warfarin versus placebo, and the control subjects were allowed to take aspirin-containing compounds. A total of 420 patients were entered, with 212 in the warfarin group and 208 in the control group. Over an average follow-up of 2.2 years, 10% of study patients had permanent discontinuation of warfarin. In this study, the target range for the PT was 1.2 to 1.5 times control, which corresponds to an INR of 1.5 to 2.7 for the thromboplastin reagents most commonly used.

The Boston Area Trial, 7 looked at warfarin versus placebo, and the control subjects were allowed to take aspirin-containing compounds. A total of 420 patients were entered, with 212 in the warfarin group and 208 in the control group. Over an average follow-up of 2.2 years, 10% of study patients had permanent discontinuation of warfarin. In this study, the target range for the PT was 1.2 to 1.5 times control, which corresponds to an INR of 1.5 to 2.7 for the thromboplastin reagents most commonly used.

This study reported a dramatic reduction in ischemic stroke of 86% with warfarin (0.41% per year incidence in the study group compared with 2.98% per year incidence in the control group, P = 0.0022). The death rate was also significantly lower in the warfarin group: 2.25% per year compared with 5.97% per year (P = 0.005). The risk of serious hemorrhage was 1.9% in the warfarin group and 1% in the control group. The risk of minor hemorrhage was 18% in the warfarin group and 10% in the control group.

• Antiplatelet Therapy

Antiplatelet therapy, specifically aspirin, has shown some benefit in preventing stroke in stroke-prone individuals. Review of the literature indicates that aspirin, at a total dose of 1300 mg per day, reduces the risk of stroke in TIA patients by 25% to 30% . I,8, 10, 12, 23, 24 It is possible that a lower dose of aspirin, such as 325 mg per day, is of equivalent efficacy to 1300 mg per day. 99 In the European Stroke Prevention Study, 21 aspirin, 330 mg three times a day, combined with dipyridamole, 75 mg three times per day, was found to be superior to placebo.

Theoretically, a smaller dose of aspirin, 20 to 40 mg per day, should have effective platelet antiaggregant activity. Despite the fact that low-dose aspirin, e.g., 40 mg per day, can result in platelet antiaggregation equivalent to 325 mg per day, 38 this might not effectively extrapolate to the clinical situation. 94

Dipyridamole is of questionable efficacy for stroke prevention in patients with TIA. 1, 21 On the other hand, when combined with warfarin, this antiplatelet agent protects against cardiogenic thromboembolism. 93 It is thus possible for antiplatelet therapy to have a protective effect against cardiac disease which can promote thromboembolism, such as AF. Until recently, aspirin had not been adequately evaluated in this regard, but the combination of aspirin and warfarin was found to be excessively dangerous in prophylaxis for prosthetic cardiac valves. 15

• Randomized, Controlled Trials of Aspirin Therapy for Stroke Prevention in Atrial Fibrillation

In the AFASAK study, 66 336 of the 1007 subjects with chronic NVAF received 75 mg of aspirin per day. Twenty of these subjects (6%) suffered an embolic event, and the stroke rate of 5.5% per year was the same as that of the placebo group. The authors concluded that this dose of aspirin was ineffective.

The SPAF study, 92 reported a beneficial effect of aspirin in patients less than 76 years of age who were given a dose of 325 mg, enteric-coated, per day. For the 517 patients given aspirin, the rate of primary events (ischemic stroke and systemic embolism) was 3.2% per year, compared with 6.3% per year for the placebo patients. This was a 49% reduction, which was significant at P = 0.014. The risk reduction was more dramatic, at 65% (P = 0.0042) when patients greater than 75 years of age were excluded (Fig. 1). In the AFASAK study, there were two patients on aspirin with nonfatal bleeding complications and none in the placebo group. In the SPAF study, hemorrhagic events occurred at a rate of 0.9% per year in the aspirin group and 1.2% per year in the placebo group.

The explanation for the lack of efficacy of aspirin in the AFASAK study might be related to several differences between this study and the SPAF study. For one, the dose was 75 mg per day in the AFASAK study, whereas it was 325 mg per day in the SPAF study. In addition, the mean age of 74 years in the AFASAK study, compared with 67 years in the SPAF study, means that a larger proportion of patients in the AFASAK study were beyond age 75, and such patients did not respond to aspirin in the SPAF study. There were also other differences in the study populations that might be pertinent: (1) AF was chronic in all AFASAK patients but in only 68% of SPAF patients; (2) the percentage of females was 46% in AFASAK and 30% in SPAF; and (3) the prevalence of significant coexistent cardiac disease was approximately twice as high in the AFASAK study. The Boston Trial did not assess aspirin therapy in a prospective, controlled fashion. The authors did report no apparent efficacy in those subjects who regularly took aspirin when they addressed this issue in a retrospective fashion.

SECONDARY PREVENTION OF ISCHEMIC STROKE IN PATIENTS WITH ATRIAL FIBRILLATION

The risk of recurrent stroke within the first year of the initial stroke in patients with AF is estimated to be 20%. 87 Hart et al, 36 reported that the risk of recurrent embolism in NVAF was as high as 20% within 11 days of the initial event if anticoagulant therapy was not initiated. In a retrospective study of AF in which 20% of the 150 patients had valvular heart disease, 50% had multiple events and one quarter of the recurrences took place within 2 weeks of the initial event. 86 Sage and Van, 81 Uitert reported that for the 59 subjects, of the initial 144 patients with stroke and NVAF in their study, who were available for follow-up for up to 9 years without anticoagulant therapy, the risk of recurrent cerebral ischemia remained constant at 20% per year. Of note, the first recurrence was seen at day 12 in this retrospective study.

In the Framingham Study,106 Wolf et al reported only a slightly higher recurrence rate for stroke in AF patients (25%) than in non-AF. patients (20%). The stroke recurrence rate was more than twice as high in the AF group (47%) as in the non-AF group (20%) within 6 months of the initial stroke, however. Gustafsson and Britton, 33 reported that the risk of recurrent embolism was 11% within the first month in stroke patients with NVAF, compared with 2% in stroke patients in sinus rhythm. This excessive recurrence rate was seen only for the first month over a 5-year follow-up period for survivors. Thus, the risk of recurrence appears to be quite variable, tends to lessen over time, and perhaps reflects the criteria for recurrence as well as other factors. In a recent study that addressed this issue 6 for example, only 4% of patients with NVAF had a recurrent embolic event within 1 month of the initial stroke.

An initial stroke in a patient with NVAF appears to identify an individual at relatively high risk for a recurrent event. Unfortunately, this has not been adequately addressed in a convincing fashion from the standpoint of prophylaxis. Lodder et al, 56 reported no benefit of long-term anticoagulant therapy in patients with ischemic stroke and NVAF compared with a control group over a mean follow-up of 27 months. Yamanouchi et al, 108 reported a rate of recurrence of 4% over a mean follow-up period of 3.8 years in 23 patients with NVAF and initial stroke treated with anticoagulant therapy. An untreated control group from an autopsy series had a recurrence rate of 26% over a mean period of 1.3 years.

A major concern with anticoagulant therapy in acute ischemic stroke with a cardiogenic source is the potential for spontaneous hemorrhagic conversion of the infarct. It has been reported that up to 31% of cerebral embolic infarcts may spontaneously become hemorrhagic on serial CT scan. 101 Anticoagulant therapy can promote catastrophic consequences in such a clinical setting. This has prompted guidelines for the judicious use of anticoagulant therapy in acute embolic strokes of cardiac origin. 109

SUMMARY

Antithrombotic therapy is clearly indicated in patients with atrial fibrillation who have associated factors that put them at significant risk for thromboembolism. This does not include subjects with lone atrial fibrillation who are less than 60 years of age. High-risk patients include those with valvular heart disease, recent congestive heart failure, severe left ventricular dysfunction by echocardiography, prior thromboembolism, demonstration of a cardiac thrombus by echocardiography, and thyrotoxicosis. Anticoagulant therapy appears to be the most efficacious means of preventing thromboembolism in atrial fibrillation. Potential bleeding complications with sodium warfarin mandate judicious selection of patients for long-term anticoagulant therapy. The risk of anticoagulant therapy certainly appears justified in subjects who are at high risk for thromboembolism and can be monitored with a reasonable degree of safety. Aspirin therapy is a reasonable alternative for those subjects at relatively lower risk of thromboembolism, especially subjects who are not suitable candidates for anticoagulation. The efficacy of aspirin has not been established in patients with atrial fibrillation who are greater than 75 years of age.

---

References

1. American-Canadian Cooperative Study Group: Persantine-aspirin trial in cerebral ischemia. Part 11: End-point results. Stroke 16:406, 1985

2. Aronow WS, Gutstein H, Hsieh FY: Risk factors for thromboembolic stroke in elderly patients with chronic atria] fibrillation. Am j Cardiol 63:366, 1989

3. Aschenberg W, Schluter M, Kremer P, et al: Transesophageal two-dimensional echocardiography for the detection of left atrial appendage thrombus. j Am coll Cardiol 7:163, 1986

4. Bar-Sela S Ehrenfeld M, Gliakim M: Arterial embolism in thyrotoxicosis with atrial fibrillation. Arch Intern Med 141:1191, 1981

5. Bloomfield P, et al :a prospective evaluation of the Starr-Edwards, Bjork-Shiley heart valve prostheses,circulation 73:1213 1986

6. Bogousslavsky J, Van Melle G, Regli F, et al: Pathogenesis of anterior circulation stroke in patients with nonvalvular atrial fibrillation. Neurology 40:1046, 1990

7. Boston Area Anticoagulation Trial for Atrial Fibrillation Investigators: The effect of low-dose warfarin on the risk of stroke in patients with nonrheumatic atrial fibrillation. N Engl J Med 323:1505, 1990

8. Bousser MG, Eschwege G, Haguenaum, at al: “AICLA” controlled trial of aspirin and dipyridamole in the secondary prevention of athero-thrombotic cerebral ischemia. Stroke 14:7, 1983

9. Brand IN, Abbott RD, Kannel WB, et al: Characteristics and prognosis of lone atrial fibrillation. 30-year follow up in the Framingham Study. JAMA 254:3449, 1985

10. Brittain M, Helmers C, Samuelsson K: High-dose acetylsalicyclic acid after cerebral infarction: A Swedish cooperative study. Stroke 17:132, 1986

11. Britton M, Gustafson C: Nonrheumatic atria] fibrillation as a risk factor for stroke. Stroke 16:182, 1985

12. Canadian Cooperative Study Group: A randomized trial of aspirin and sulfinpyra- zone in threatened stroke. N Engl J Med 299:53, 1978

13. Caplan LR, D’Cruz 1, Hier DB, et al: Atrial size, atrial fibrillation and stroke. Ann Neurol 19:158, 1986

14. Cerebral Embolism Task Force: Cardiogenic brain embolism. The second report of the cerebral embolism task force. Arch Neurol 46:727, 1989

15. Chesebro JH, Fuster V, Elveback LR, et al: Trial of warfarin plus dipyridamole or aspirin therapy in prosthetic heart valve replacement: Danger of aspirin compared with dipyridamole. Am j Cardiol 51:1537, 1983

16. Coulshed N, Epstein EJ, Mackendrick CS, at al: Systemic embolism in mitral valve disease. Br Heart J 32:26, 1970

17. Davies Mj, Pomerance A: Pathology of atrial fibrillation in man. Br Heart j 34:520, 1972

18. Davis PH, Dambrosia JM, Schoenberg BS, et al: Risk factors for ischemic stroke. A prospective study in Rochester, Minnesota. Ann Neurol27:319, 1987

19. Dittrich HC, Erickson JS, Schneiderman T, et al: Echocardiographic and clinical predictors for outcome of elective cardioversion of atrial fibrillation. Am j Cardiol 63:193, 1989

20. Easton JD, Sherman DG: Management of cerebral embolism of cardiac origin. Stroke 11:433, 1980

21. ESPS Group: European Stroke Prevention Study. Stroke 21:1122, 1990

22. Feinberg WM, Seeger JF, Carmody RF, et al: Asymptomatic cerebral infarction in patients with atrial fibrillation [abstract]. Circulation 78(Suppl 11):600, 1988

23. Fields WS, Lemak NA, Frankowski RF, et al: Controlled trial of aspirin in cerebral ischemia. Stroke 8:301, 1977

24. Fields WS, Lemak NA, Frankowski RF, et al: Controlled trial of aspirin in cerebral ischemia. Part 11, Surgical group. Stroke 9:309, 1978

25. Flegel KM, Hanley J: Risk factors for stroke and other embolic events in patients with nonrheumatic atrial fibrillation. Stroke 10:1000, 1989

26. Flegel KM, Shipley Mj, Rose G: Risk of stroke in nonrheumatic atrial fibrillation. Lancet 1:526, 1987

27. Forfar JC, Toft AD: Thyrotoxic atrial fibrillation: An underdiagnosed condition. Br Med j 285:909, 1984

28. Furlan Aj, Craciun AR, Salcedo EE, et al: Risk of stroke in patients with mitral annulus calcification. Stroke 15:801, 1984

29. Goldberger E: Supraventricular tachyarrhythmias. In Treatment of Cardiac Emergencies. St. Louis, CV Mosby, 1972, p 86

30. Godtfredsen J: Atrial Fibrillation. Etiology, Course and Prognosis. A Follow up Study of 1,212 Cases. Copenhagen, Munksgaard, 1975, p 17

31. Godtfredsen J, Petersen P: Thromboembolic complications in atrial fibrillation. In Refsum H, Suig IA, Rasmussen K (eds): Heart Brain and Brain Heart. Heidelberg, Springer-Verlag, 1989, pp 225-229

32. Gustafsson C, Blomback M, Britton M, et al: Coagulation factors and the increased risk of stroke in nonvalvular atrial fibrillation . Stroke 21:47, 1990

33. Gustaf sson C, Britton M: Prognosis after brain infarction in patients with nonvalvular atrial fibrillation compared with sinus rhythm abstracts. Acta Neurol Scand (Suppl) 73:520, 1986

34. Halperin ]L, Hart RG: Atrial fibrillation and stroke: New ideas, persisting dilemmas. Stroke 19:937, 1988

35. Harrison MJG, Marshall J: Atrial fibrillation, TIAs and completed strokes. Stroke 15:441, 1984

36. Hart RG, Coull BM, Hart D: Early recurrent embolism associated with nonvalvular atrial fibrillation: A retrospective study. Stroke 14:688, 1983

37. Henry WL, Morganroth i Pearlman AS, et al: Relation between echocardiographically determined left atrial size and atrial fibrillation. Circulation 53:273, 1976

38. Hirsh J: Progress review: The relationship between dose of aspirin, side effects and antithrombotic effectiveness. Stroke 16:1, 1985

39. Hirsh J, Levine MN: The optimal intensity of oral anticoagulant therapy. JAMA 258:2723, 1987

40. Hirsh J, Levine M: Therapeutic range for the control of oral anticoagulant therapy. Arch Neurol 43:1162, 1986

41. Hoglund C, Rosenhamber G: Echocardiographic left atrial dimension as a predictor of maintaining sinus rhythm after conversion of atrial fibrillation. Acta Med Scand 217:411, 1985

42. Hull R, Hirsh J, Jay R, et al: Different intensities of oral anticoagulant therapy in the treatment of proximal-vein thrombosis. N Engl J Med 307:1676, -1982

43. Hurley DM, Hunter AN, Hewett Mj, et al: Atrial fibrillation and arterial embolism in hyperthyroidism. Aust NZ j Med 11:391, 1981

44. Hurst JW,Paulk EA, Proctor HD, et al: Management of patients with atria] fibrillation. Am I Med 37:728, 1964

45. lespersen CM, Egeblad H: Mitral annulus calcification and embolism. Acta Med Scand 222:37, 1987

46. Kaplan EL, Meier P: Nonparametric estimation from incomplete observations. j Am Stat Assoc 53:467, 1958

47. Kelley RE, Berger JR, Alter M, et al: Cerebral ischemia and atrial fibrillation: Prospective study. Neurology 34:1285, 1984

48. Kelley RE, Pina I, Lee S-C: Cerebral ischemia and mitral valve prolapse. Stroke 19:443, 1988

49. Kempster PA, Gerraty RP, Gates PC: Asymptomatic cerebral infarction in patients with chronic fibrillation. Stroke 19:955, 1988

50. Keren G, Etzion T, Sherez J, et al: Atrial fibrillation and atrial enlargement in patients with mitral stenosis. Am Heart J 114:1146, 1987

51. Kitchens JM, Flegel KM: Atrial fibrillation, stroke and anticoagulation: What is to be done? J Gen Intern Med 1:126, 1986

52. Kopecky SL, Gersh Bj, McGoon MD, et al: The natural history of lone atrial fibrillation. N Engl J Med 317:669, 1987

53. Kulbertus HE, Leval-Ruttan FD, Bartsch P, et al: Atrial fibrillation in elderly, ambulatory patients. In Kulbertus HE, Olsson SB, Schlepper M (eds): Atrial Fibrillation. Kiruna, Sweden, Molndal, AB Hassle, 1982, pp 148-157

54. Lavy S, Stern S, Melamed E, et al: Effect of chronic atrial fibrillation on regional cerebral blood flow. Stroke 11:35, 1980

55. Levine M, Hirsh J: Hemorrhagic complications of long-term anticoagulant therapy for ischemic cerebral vascular disease. Stroke 17:111, 1986

56. Lodder J, Dennis MS, Van Raak L, et al: Cooperative study on the value of long term anticoagulation in patients with stroke and non-rheumatic atrial fibrillation. Br Med J 296:1435, 1988

57. Lowe GDO, Forbes CD, Jaap AJ: Relation of atrial fibrillation and high hematocrit to mortality in acute stroke. Lancet 1:784, 1983

58. Martin A, Benbow Lj, Butrous GS, et al: Five-year follow-up of 101 elderly subjects by means of long-term ambulatory cardiac monitoring. Eur Heart j 5:592, 1984

59. Mohr JP: Neurologic complications of cardiac valvular disease and cardiac surgery including systemic hypotension. In Vinken Pj, Bruyn GW (eds): Handbook of Clinical Neurology. Neurological Manifestations of Systemic Disease. Vol 38, Part 1. New York, North Holland Publ Co, 1979, p 143

60. Moss AJ: Atrial fibrillation and cerebral embolism [editorial]. Arch Neurol 41:707, 1984

61. Norrving B, Nilsson B: Cerebral embolism of cardiac origin: The limited possibilities of secondary prevention [abstract]. Acta Neurol Scand 73:520, 1986

62. Olesen KH, Rygg IH, Wennevold A, Nyboe J: Long-term follow-up in 185 patients after mitral valve replacement with the Lillehei-Kaster prosthesis: Overall results and prosthesis-related complications. Eur Heart j 8:680, 1987

63. Olsen TS, Shriver EB, Herning M: Cause of cerebral infarction in the carotid territory. Its relation to the size and the location of the infarct and to the underlying vascular lesion. Stroke 16:459, 1985

64. Ostrander LD Jr, Brandt RL, Kjelsberg MO, et al: Electrocardiographic findings among the adult population of a total natural community. Tecumseh, Michigan. Circulation 31:888, 1965

65. Petersen P: Thromboembolic complications in atrial fibrillation. Stroke 21:4, 1990

66. Petersen P, Boysen G, Godtfredsen J, et al: Placebo-controlled, randomized trial of warfarin and aspirin for prevention of thromboembolic complications in chronic atrial fibrillation. The Copenhagen AFASAK study. Lancet 1:175, 1989

67. Petersen P, Godtfredsen J: Atrial fibrillation review of course and prognosis. Acta Med Scand 16:5, 1984

68. Petersen P, Godtfredsen J: Risk factors for stroke in chronic atrial fibrillation. Eur Heart j 9:291, 1988

69. Petersen P, Hansen JH: Stroke in thyrotoxicosis with atrial fibrillation. Stroke 19:15, 1988

70. Petersen P, Kastrup J, Brinch K, et al: Relation between left atrial dimension and duration of atrial fibrillation. Am j Cardiol 60:382, 1987

71. Petersen P, Kastrup J, Videback R, et al: Cerebral blood flow before and after cardioversion of atrial fibrillation. J Cereb Blood Flow Metab 9:4?9, 1989

72. Petersen P, Madsen EB, Brun B, et al: Silent cerebral infarction in chronic atrial fibrillation. Stroke 18:1098, 1987

73. Petersen P Pedersen F, Johnsen A, et al: Cerebral computed tomography in paroxysmal atrial fibrillation. Acta Neurol Scand 79:482, 1989

74. Poller L: Response to SPAF study [letter]. N Engl J Med 323:483, 1990

75. Poller L, Hirsh J: Special report: A simple system for the derivation of International Normalized Ratios for the reporting of prothrombin time results with North American thromboplastin reagents. Am J Clin Pathol 92:124, 1989

76. Pop G, Sutherland GR, Koudstaal Pj, at al: Transesophageal echocardiography of intracardiac embolic sources in patients with transient ischemic attacks. Stroke 21:560, 1990

77. Presti CF, Asinger RW, Goldman ME: Comparative measurements of the left atrium in patients with constant vs. intermittent nonvalvulopathic atrial fibrillation [ab- stract]. Circulation 78 (Suppl 11):600, 1988 P- gulant therapy in the

78. Roy 0, Marchand E, Gagne P, et al: Usefulness of anticoagulation in prevention of embolic complications of atrial fibrillation. Am Heart j 112:1039, 1986

79. Ruocco NA, Most AS: Clinical and echocardiographic risk factors for systemic embolization in patients with atrial fibrillation in the absence of mitral stenosis [abstract]. j Am coll Cardiol 7:165A, 1986 ed with non-valvular atrial fibrillation:

80. Sage JI: Prospective study of stroke associat Ann Neurol 20:155, 1986 Low frequency of early recurrence [abstract].

81. Sage JI, Van Uitert RL: Risk of recurrent stroke in patients with atrial fibrillation and non-valvular heart disease. Stroke 14:537, 1983

82. Saour JN, Sieck JO, Mamo LAR, et al: Trial of different intensities of anticoagulation in patients with prosthetic heart valves. N Engl J Med 3)):428, 1990

83. Sasaki W, Yanagisawa S, Maki K, et al: High incidence of silent small cerebral infarction in patients with atrial fibrillation [abstract]. Circulation 76 (Suppl IV):104, 1987

84. Schwartz MH, Teicholz LE, Donoso E: Mitral valve prolapse. A review of associated arrhythmias. Am j Med 62:377, 1977

85. Selzer A: Effects of atrial fibrillation on circulation of patients with mitral stenosis. Am Heart J 59:518, 1960

86. Sherman DG, Goldman L, Whiting RB, et al: Thromboembolism in patients with atrial fibrillation. Arch Neurol 41:708, 1984

87. Sherman DG, Hart RG, Easton JD: The secondary prevention of stroke in patients with atrial fibrillation. Arch Neurol 43:68, 1986

88. Shields RW Jr, Laureno R, Lachman T, et al: Anticoagulant-related hemorrhage in acute cerebral embolism. Stroke 15:426, 1984

89. Sokolow M, Massie B: Heart and great vessels. In Schroeder SA, Krupp MA, Tierney LM Jr, et al (eds): Current Medical Diagnosis and Treatment. Norwalk, CT, Appleton & Lange, 1989, p 240

90. Staffurth JS, Gibed MC, Tang Fui SNGT: Arterial embolism in thyrotoxicosis with atrial fibrillation. Br Med J 2:688, 1977

91. Stroke Prevention in Atrial Fibrillation Investigators: Design of a multicenter randomized trial for the Stroke Prevention in Atrial Fibrillation Study. Stroke 21:538, 1990

92. Stroke Prevention in Atrial Fibrillation Investigators: Preliminary report of the Stroke Prevention in Atrial Fibrillation Study. N Engl J Med 322:863, 1990

93. Sullivan JM, Harken DE, Gorlin R: Pharmacologic control of thromboembolic complications of cardiac-valve replacement. N Engl J Med 284:1391, 1971

94. Svensson J, Samualsson K: Inhibition of platelet function by low-dose acetylsalicyclic acid in patients with cerebrovascular disease. Thromb Res 31:499, 1983

95. Szekely P: Systemic embolism and anticoagulant prophylaxis in rheumatic heart disease. Lancet 1:1209, 1964

96. Takahashi N, Imatakak, Seki A, et al: Left atrial enlargement in patients with paroxysmal atrial fibrillation. Jour Heart j 23:677, 1982

97. Tegler CH, Stroke Prevention in Atrial Fibrillation Study: Carotid stenosis in atrial fibrillation [abstract]. Neurology (Suppl 1) 39:159, 1989

98. Tegeler CH, Hart RG: Atrial size, atrial fibrillation and stroke [letter]. Ann Neurol 21:315, 1987

99. UK-TIA Study Group: The UK-TIA Aspirin Trial: The interim results. Br Med J 296:316, 1988

100. Weinberger J, Rotlauf E, Materese E, et al: Noninvasive evaluation of the extracranial carotid arteries in patients with cerebrovascular events and atrial fibrillation. Arch Intern Med 148:1785, 1988

101. Weisberg LA: Nonseptic cardiogenic cerebral embolic stroke: Clinical-CT correlations. Neurology 35:896, 1985

102. Weksler BB, Lewin M: Anticoagulation in cerebral ischemia. Stroke 1:658, 1983

103. Wintzen AR, de Jonge H, Loeliger EA, et al: The risk of intracerebral hemorrhage during oral anticoagulant treatment: A population study. Ann Neurol 16:553, 1984

104. Wolf PA, Abbott RD, Kannel WB: Atrial fibrillation: A major contributor to stroke in the elderly. The Framingham Study. Arch Intern Med 147:1561, 1987

105. Wolf PA, Dawber TR, Thomas HE Jr, et al: Epidemiological assessment of chronic atrial fibrillation and risk of stroke: The Framingham Study- Neurology 28:973, 1978

106. Wolf PA, Kannel WB, McGee DL, at al: Duration of atrial fibrillation and imminence of stroke: The Framingham Study. Neurology 14:664, 1983

107. Wolf PA, Sila CA: Cerebral ischemia with mitral valve prolapse. Am Heart j 113:1308, 1987

108. Yamanouchi H, Nagura H, Ohkaway, et al: Anticoagulant therapy in recurrent cerebral embolism: A retrospective study in non-valvular atrial fibrillation. j Neurol 235:407, 1988

109. Yatsu FM, Hart RG, Mohr JP, et al: Anticoagulation of embolic strokes of cardiac origin: An update. Neurology 38:314, 1988

110. Yuen RWM, Gutteridle DH, Thompson PL, et al: Embolism in thyrotoxic atrial fibrillation. Med j Aust 1:630, 1979 -

111. Zenker G, Erbel R, Kramer G, et al: Transesophageal two-dimensional echocardiography in young patients with cerebral ischemic events. Stroke 19:345, 1988