Stroke Prevention in Atrial Fibrillation

Atrial fibrillation is a relatively common clinical problem, affecting an estimated 2.3 million adults in the United States.¹ The prevalence of atrial fibrillation increases with age; nearly 4% of persons aged 60 years or older and 9% of those aged 80 years or older have atrial fibrillation,¹ and about half the patients in the United States with atrial fibrillation are older than 75 years.² By the year 2050, the prevalence of atrial fibrillation is expected to increase about 2.5 fold, with most of that increase explained by the growing proportion of individuals living into their 80s and beyond.¹

Until a therapy or therapies are found that reliably, effectively, and safely prevent its recurrence, the 2 most important clinical problems associated with atrial fibrillation are prevention of tachycardia-mediated cardiomyopathy and prevention of ischemic stroke. Although prevention of tachycardia-mediated cardiomyopathy should almost always be possible, prevention of ischemic stroke remains a major challenge.

Compared with patients with sinus rhythm, those with atrial fibrillation have a 4-fold to 5-fold increase in risk of ischemic stroke,³ with comparable risk for patients with paroxysmal and chronic atrial fibrillation.⁴ This risk of stroke may be stratified by several factors, including previous thromboembolic stroke, hypertension, diabetes, ventricular dysfunction, age, or rheumatic mitral valve disease,⁵ and may be stratified further into mild, moderate, or severe categories. Other risks such as coronary artery disease, sex, degree of hypertension, thyrotoxicosis, and hypertrophic cardiomyopathy are important to consider.⁵ Patients with atrial fibrillation but without structural heart disease (including no hypertension) are at relatively low risk of ischemic stroke, especially if they are younger than 65 years.⁵

The article by Wang et al⁶ in this issue of THE JOURNAL examines the predictors of stroke as well as stroke or death in patients with newly recognized atrial fibrillation in a community-based cohort (Framingham) and provides quantitative risk scores for these outcomes. The investigators sought to devise this score because previously published risk stratification schemes were based on either clinical trial cohorts or administrative databases. An absolute estimate of risk was derived in the Framingham cohort based on 5 risk factors, namely, advancing age, female sex, increasing systolic blood pressure, prior stroke or transient ischemic attack, and diabetes mellitus. The ability to estimate the absolute risk of stroke outcomes in an individual patient should prove helpful to both the clinician and the patient in making treatment decisions. Although the authors identify the strengths and limitations of their scoring system, another important consideration is that their patient cohorts were from a different era of clinical care (ie, no statins and, for some, no angiotensin-converting enzyme inhibitors or β-blockers), which may affect outcomes, especially risk of death. Nevertheless, this scoring system should make an important contribution to patient care.

Determining which patients with atrial fibrillation should receive oral anticoagulation will always be a critical question. A major problem has been that all too often, patients with clear indications for warfarin therapy do not receive it. Compared with placebo, warfarin therapy reduces the ischemic stroke risk substantially (68% risk reduction by intention-to-treat analysis,⁷ 83% risk reduction by on-treatment analysis⁸ ) in patients with atrial fibrillation. Furthermore, in patients with atrial fibrillation who are at risk of ischemic stroke, warfarin is significantly better than aspirin for preventing ischemic stroke.^{5 ,9} Despite this evidence, only one third to a little more than half of patients with atrial fibrillation and who would be good candidates for receiving warfarin therapy actually receive it, with elderly patients the least likely to be treated with warfarin.^{10 - 12} This is of great concern because both the prevalence of atrial fibrillation and the risk of ischemic stroke associated with atrial fibrillation increase with age. In short, not enough patients with atrial fibrillation at recognized risk of ischemic stroke receive warfarin therapy, and patients with the greatest risk of ischemic stroke in the face of atrial fibrillation are the ones who receive it least.

Several factors may be involved in the apparent underuse of warfarin, especially among elderly patients.¹³ The most noteworthy concern is the risk of bleeding, particularly due to falls, frailty, or forgetfulness. Also, it is difficult to maintain the international normalized ratio (INR) in the therapeutic range (ie, 2-3), in part because of the many interactions of warfarin with other drugs and foods.^{14 - 15} Although the relative risk of ischemic stroke increases by 1.4 per decade beginning at age 65 years, so does the relative risk of intracranial bleeding while taking warfarin.¹⁶ However, because the base rate of ischemic stroke is considerably greater, the risk of ischemic stroke in the absence of warfarin therapy is greater than the risk of intracranial bleeding while receiving warfarin. Careful attention to maintaining the INR between 2 and 3 maximizes the benefit of warfarin and minimizes its risk.⁵ Attention to maintaining the INR in the therapeutic range is important because if the INR dips below 2, the odds ratio for stroke increases steeply, doubling with an INR of 1.7 and tripling with an INR of 1.5.¹⁷ Moreover, the increased risk of intracranial hemorrhage is largely stable until the INR reaches 4.0.¹⁷

Thus, for each patient, clinicians must strike an acceptable balance between their patients' risk of ischemic stroke and the risk of bleeding. In the absence of an absolute or important relative contraindication, the data seem compelling that warfarin therapy should be offered to patients with atrial fibrillation at risk of stroke. The difficulty is to know what threshold of stroke risk is low enough so that the potential risks of warfarin therapy outweigh its potential benefits. The risk scoring system of Wang et al⁶ should be most helpful in determining benefit vs risk.

Several potential alternatives to warfarin therapy are currently under investigation. One approach to prevent ischemic stroke in patients with atrial fibrillation is to occlude the left atrial appendage using catheter techniques because the left atrial appendage is where most clots occur.¹⁸ A similar but more aggressive approach involves surgical excision of the left atrial appendage.¹⁹ Both of these approaches are probably most appropriate to consider for patients with contraindications to anticoagulation therapy. Perhaps the most practical and important alternative strategy for stroke prevention involves new oral anticoagulants that directly inhibit thrombin. Preliminary data from one recently reported large clinical trial of patients with atrial fibrillation at risk of ischemic stroke demonstrated that use of this type of drug is as effective as warfarin in preventing stroke, has less bleeding, does not require monitoring for anticoagulation efficacy, and, to date, has no documented interactions with foods or drugs.²⁰ If the efficacy and safety of oral direct thrombin inhibitors are supported by subsequent studies, these agents should help reduce the burdens of warfarin therapy for both the patient and the physician and make achieving effective oral anticoagulation easier and, hopefully, more fully applied.

For patients with atrial fibrillation at risk of stroke, every effort should be made to prescribe warfarin and maintain the INR between 2 and 3, with a target of 2.5. For most patients, the potential benefits of stroke prevention will outweigh the potential risks of bleeding secondary to warfarin. Considering that accurate risk stratification is essential to identify patients who have the best chance of benefit from oral anticoagulant therapy, clinicians should find the Framingham risk scoring system a useful aid for clinical decision making.