Stroke is the third leading cause of death, and the leading cause of long term morbidity in the United States. Prospective observational studies have established stroke risk factors and determined their relative importance. Large randomized trials have demonstrated that significant risk reductions are possible when modifiable risk factors are targeted. But despite our increasingly sophisticated knowledge about stroke prevention, stroke risk factors are controlled in only 25 percent of patients known to have them. As the prevalence of stroke increases with the ageing population, improving adherence to proven prevention strategies takes on increasing urgency, and may be more important than the development of ‘newer and better’ treatments.

Nonmodifiable risk factors for stroke are age, gender, ethnicity, and prior stroke or TIA. Modifiable risk factors include hypertension, atrial fibrillation, cardiovascular disease (MI), diabetes, hyperlipidemia, dyslipidemia, carotid stenosis, cigarette smoking, and alcohol abuse. Physical inactivity, homocysteine, and chronic infection may soon be added to this standard list. Northern Manhattan Stroke Study (NOMASS) cohort data reveal higher population attributable risks for hypertension and inactivity, among blacks and Hispanics and for atrial fibrillation, among whites.

Approximately 80 70 percent of first strokes are ischemic. Of these, most 15 percent are atherothrombotic, producing either flow failure infarcts from large artery stenosis or occlusion (10 percent) or thromboembolic, from atheromatous plaque (5 percent). with Twenty-five percent are small vessel disease, (~25 percent), thromboembolism from atheromatous plaque (~10 percent), and 20 percent are due to cardiac embolism. making up about 55 percent; tThe rest are of undetermined origin; most are presumed to be atherothromboticdue to occult embolism. A small percentage may be due to hypercoagulable states, arteritis, infection or migraine.^,, This chapter will focus on ischemic stroke.

Hypertension is the most important modifiable determinant of stroke risk. The association between hypertension and stroke increases with age relative to most other risk factors. Systolic and diastolic hypertension are each strongly and independently associated with stroke risk. Data from the Northern Manhattan Stroke Study (NOMASS) show that hypertension carries a higher population attributable risk for blacks and Caribbean Hispanics than for whites -- 37 percent and 32 percent, versus 25 percent.

Isolated systolic hypertension is highly prevalent and highly associated with stroke in the elderly population.^, Even controlling for age, stroke risk is proportional to systolic hypertension. Treatment of isolated systolic hypertension produces a large stroke risk reduction, on the order of 35-40 percent.^,

For every 7.5 mmHg increase in diastolic pressure, there is a 46 percent increase in stroke incidence.^, Metaanalysis of observational studies involving over 400,000 subjects demonstrates a continuous increase in stroke risk directly related to diastolic blood pressure, even in the normotensive range (starting at 69 mmHg). Conversely, an average 5.8 mmHg blood pressure reduction corresponds to a 42 percent reduction in stroke incidence.^, In another study, a 7.5 mmHg decrease in diastolic pressure was associated with a 46 percent decrease in stroke incidence.

There is overwhelming evidence to support the benefit of treating systolic and/or diastolic hypertension, particularly in the elderly. Large randomized controlled trials (RTCs), metaanalyses of RTCs, and observational studies all demonstrate a marked impact of treatment of hypertension on the incidence of stroke (Table 1).^,, Newer studies include UKPDS and HOPE, which confirm the significant stroke risk reductions in diabetics treated for hypertension.^,

The evidence supporting a correlation between total cholesterol level and stroke remains weak, although some trials suggest a link in middle age.^,,, Framingham data imply that total cholesterol is an independent risk factor for stroke before age 55. The Multiple Risk Factor Intervention Trial (MRFIT), which enrolled only middle-aged men and did not look at non-fatal stroke, found an increased risk of death due to non-hemorrhagic stroke varying directly with total cholesterol levels above 200mg/dL.

LDL levels generally do not predict stroke, although a recent study in patients with CHD showed that high LDL levels were associated with increased incidence of stroke and TIA. In contradistinction,LDL cholesterol-lowering agents (statins, niacin and fibrates) have, however, been shown to reduce stroke in patients with CAD.^,,,, The Long-Term Intervention with Pravastatin in Ischaemic Disease (LIPID) study showed all-cause and non-hemorrhagic stroke risk reductions of 16 and 21 percent, respectively, with high dose pravastatin (40 mg) in high risk patients. Baseline total and LDL cholesterol levels did not correlate with treatment effect (though HDL levels did). The Heart Protection Study results show an overwhelming stroke benefit for statins in high risk patients: high dose simvastatin (40 mg) reduced stroke by approximately 27 percent over five years, versus antioxidant vitamins or placebo. Unlike many previous statin studies, this large British trial enrolled significant numbers of elderly patients, women, and diabetics. One-third of patients had LDL below 116 at baseline. Results were similar across all subgroups, and may herald the elimination of target LDL levels for statin therapy.

Low HDL is an independent stroke risk factor and elevated HDL is an independent protective factor for stroke. ^,, NOMASS observational data reveal an independent, dose-dependent, stroke-protective effect of high HDL (> 35) in patients 75 or older. The protective effect is significantly stronger in atherosclerotic versus nonatherosclerotic stroke, and holds across ethnic groups.

The data for high triglycerides as a stroke risk factor are mixed.^,, However, in a recent prospective study of lipids and stroke risk in patients with established CAD, triglycerides > 200 mg/dL were an independent risk factor for ischemic stroke/TIA, even after controlling for usual risk factors and other cholesterol fractions.

The benefits of HMG-CoA reductase inhibitors ("statins") in stroke prevention for patients with risk factors are clearly established, and these risk factors alone, e.g., age, diabetes, prior cerebrobvascular disease, CAD, low HDL, may soon be indications for their use without reference to LDL.^,

Fibrates, niacin, and antioxidants are of less certain value in stroke prevention. The recently published Veterans Affairs HDL Intervention Trial (VA-HIT) showed that 1200mg/day of gemfibrozil reduced stroke by 29 percent versus placebo in men with CAD and low HDL. Other major trials and/or metaanalyses have not demonstrated clear stroke prevention benefits of fibrates, though fibrates clearly raise HDL levels.^,,, Niacin (2000-3000 mg/day) lowers LDL and raises HDL, in a dose dependent fashion (plateauing at about 2500 mg), and its effects are additive to those of statins.^, Its role in stroke prevention per se has yet to be elucidated. Antioxidants have been shown not to reduce stroke risk, though they appear to be safe.^, Of note, in a recent randomized trial of antioxidant vitamins in patients with CAD and low HDL, antioxidants, when added to statin-niacin treatment, reversed much of the demonstrated clinical benefits of that treatment. This was a small study looking at coronary disease, but the results have potential implications for lipid modulation aimed at stroke prevention.

Diabetes is an independent risk factor for stroke.^,, This risk exceeds that attributable to atherogenic diabetic comorbidities – hypertension and dyslipidemia – and likely derives from associated micro- and macrovascular disease.^, There is no evidence that tight glucose control helps in primary or secondary prevention of stroke in diabetics.^,, As above, there is compelling evidence that hypertension is an independent predictor of stroke in diabetics and that tight blood pressure control ( < 130/85) significantly reduces stroke risk, irrespective of glucose control.

Severe (greater than 70 percent) carotid artery stenosis (CAS) confers an annual risk of stroke of two to 13 percent, depending on the presence or absence of symptoms and/or prior events.^,, In patients with symptomatic severe CAS, the benefit of carotid endarterectomy (CEA) is unequivocal. However, for symptomatic patients with moderate disease (50-69 percent stenosis), and asymptomatic patients with moderate to severe disease, careful risk stratification is required to determine whether medical or surgical treatment is optimal therapy. Optimal medical treatment involves antiplatelet therapy and stroke risk factor control. Endovascular treatment (carotid artery PTCA with stenting) remains investigational, though it may offer benefits in higher risk patients with asymptomatic disease.^, A large randomized trial (CREST) comparing CEA to stenting in patients with symptomatic carotid stenosis is underway.

Three large trials have confirmed both a stroke and mortality benefit of CEA over aspirin in patients with severe symptomatic CAS, with absolute risk reductions of 17 percent over two years, implying that five patients would need to have surgery to prevent one stroke during this time.^,, One trial also demonstrated a modest benefit in patients with moderate disease (50-69 percent stenosis). The benefit was smaller – 6.5 percent absolute risk reduction over five years – and aggregated with more severe stenosis, age > 75, male sex, recent stroke (rather than TIA), and hemispheric TIA.^, For patients with < 50 percent stenosis, there is no significant benefit with CEA.

Asymptomatic CAS of 50-60 percent confers a stroke risk of one to three percent per year.^,,,,, There is no advantage to CEA in asymptomatic patients with CAS < 60 percent. In contrast, several large RCTs have shown a benefit of CEA over medical treatment in subgroups of asymptomatic patients with moderate to severe stenosis (> 60 percent).^,, The largest of these trials, the Asymptomatic Carotid Atherosclerosis Study (ACAS), demonstrated a six percent absolute risk reduction for ipsilateral stroke or death over a mean of 2.7 years (number needed to treat = 17). Of note, the following factors predict poor CEA outcome in these patients: surgical risk > three percent, age > 79, unstable cardiac disease, and lack of an experienced surgeon. Thus, asymptomatic patients with CAS > 60 percent should be carefully risk stratified, and the surgeon and center closely considered, before proceeding with CEA. Progression of degree of stenosis appears to be associated with increased risk of stroke in patients with asymptomatic carotid artery stenosis, emphasizing the importance of closely monitoring moderate stenosis.

Medical treatment of CAS is aimed at preventing atherothrombus formation, stabilizing and reducing plaque, and controlling hypertension, diabetes, smoking, obesity, high cholesterol, and excessive alcohol use. Anti-platelet medication is the mainstay of thrombus prevention (see below). Statins cause plaque regression in high risk patients with subclinical carotid atherosclerosis. ACE inhibitors have been shown to decrease carotid artery intima-media thickness, and reduce stroke in high risk patients.^,

Myocardial infarction (MI) is a risk factor for stroke, especially within the first month after the event, when the rate is about three percent. Before aspirin became standard acute MI and post-MI therapy, three large randomized trials demonstrated that warfarin or hydroxycoumarin significantly decreased mortality and stroke over two to three years.^,,, Complication rates were high, but the mortality benefit remained. Aspirin is now standard peri-MI therapy, rendering the warfarin data outdated.

According to the Antiplatelet Trialists’ Collaboration, one month of antiplatelet therapy following acute MI results in 38 fewer serious vascular events per 1000 treated. Most of this effect reflects reduced non-fatal reinfarction events, though it includes a small but significant reduction in non-fatal stroke. The Clopidogrel in Unstable Angina to Prevent Recurrent Events Trial (CURE), a randomized trial comparing combination clopidogrel/aspirin to aspirin alone in patients with acute coronary syndromes found that combination therapy significantly reduced recurrent combined MI, stroke and vascular death compared to aspirin alone, but not did not effect stroke as a lone endpoint. It remains to be seen whether clopidogrel offers stroke protection over and above aspirin.

There is a dose-dependent relationship of cigarette smoking to stroke risk, demonstrated in both the Framingham and Nurses Health studies.^, People who smoke one to two packs per day have double the risk of stroke of those who smoke a half-pack or less per day. Within two years of smoking cessation, stroke risk decreases substantially, and returns within five years to that of people who never smoked. All smokers should routinely be advised of the health benefits of quitting (see Chapter 3).

Light to moderate alcohol intake protects against ischemic stroke. Prospective case-control data from NOMASS show a risk reduction of 50 percent among those who drink two alcoholic drinks per day, and an increased risk in those who drink more than five drinks per day. Hemorrhagic stroke risk is elevated with any alcohol consumption. Physicians’ Health Study data confirm a protective effect for non-hemorrhagic stroke, showing a smaller, but significant 20 percent risk reduction with one or more alcoholic drinks per week. In this study there was no change in risk with heavier consumption, but the prevalence of heavy drinking in the cohort was low. Proposed mechanisms of this protective effect include increased HDL and nitric oxide levels, reduced circulating Lp(a) lipoprotein and inhibition of platelet aggregation and fibrinogen levels.

Prospective cohort data show an association between moderate elevations of serum homocysteine and risk for ischemic stroke and carotid artery stenosis.^, A causal relationship remains unclear. Homocysteine may cause endothelial oxidative damage, or may be an indirect marker of oxidative tissue damage caused by other stroke risk factors. Two large randomized trials designed to test the effect of vitamin treatment to lower homocysteine on ischemic vascular outcomes, including stroke, are underway.^,

Exercise appears to protect against stroke. The Physicians Health Study demonstrated an inverse relation between physical activity and stroke; vigorous exercise at least once-weekly reduced stroke risk by 21 percent. In this study the significance of the risk reduction was eliminated after controlling for body mass index, hypertension, high cholesterol and diabetes, suggesting exercise exerts its effect indirectly, through these variables. On the other hand, Framingham, Honolulu Heart Study, and NOMASS data, which cumulatively describe a more heterogeneous and probably more representative population, show a consistent inverse relation between exercise and stroke risk that remains significant after controlling for the same factors.^,,

Antiplatelet therapy in high risk patients confers an approximate 25 percent risk reduction compared to placebo. While aspirin’s benefit in secondary prevention in patients with prior TIA/stroke is well established, it has now been shown that patients at intermediate to high risk for vascular events per se – patients with prior or acute MI, stable or unstable angina, peripheral arterial disease, or diabetes – whose annual risk for subsequent vascular events is approximately two to three percent – benefit from longer term antiplatelet therapy. (reference for thi Also, reconcile with statement in the next section that ASA has no role in primary prevention.) The major antiplatelet drug classes are aspirin, dipyridamole, the thienopyridine derivatives (ticlopidine and clopidogrel), and glycoprotein IIb/IIIa inhibitors (e.g., abciximab).

Aspirin is the mainstay of secondary stroke prevention, and consistently shows a moderate13-15 percent reduction in recurrent stroke/TIA. Low-dose aspirin (81-325 mg) is as effective as high-dose, with fewer bleeding events. Aspirin may safely be given in acute ischemic stroke or TIA, and confers an approximate one percent reduction in early death, recurrent stroke and dependence – a meaningful improvement given the prevalence of ischemic stroke and its high rate of early recurrence.^,,, Although warfarin is superior for cardioembolic stroke prevention, the recent WARRS study demonstrated that aspirin is as effective as warfarin in prevention of ischemic stroke. There is no role for aspirin in primary stroke prevention in patients without established stroke risk factors.

Dipyridamole; Aspirin/Dipyridamole (Aggrenox™)

Dipyridamole is a platelet aggregation inhibitor. It is a slightly less effective alternative to aspirin for secondary stroke prevention, conferring an approximate 13 percent risk reduction. Combination dipyridamole/aspirin may provide a significant benefit over aspirin, but its use is limited by conflicting RCT and metaanalysis results make its role unclearadverse reactions. The Second European Stroke Prevention Study (ESPS-2) showed a 37 percent stroke reduction with combination therapy (versus placebo), compared to 16 and 18 percent reductions, for dipyridamole and aspirin alone. Combination treatmentThe dipyridamole was poorly tolerated, however, in ESPS-2, and the study had a high withdrawal rate. The ongoing European-Australian Stroke Prevention in Ischemia Trial (ESPRIT), compares aspirin, aspirin/dipyridamole, and anticoagulation in patients with prior ischemic TIA/stroke, and may clarify the role of aspirin/dipyridamole. At present, aspirin/extended-release dipyridamole (Aggrenox™) is reserved a viable alternative to aspirin for secondary prevention in patients who can tolerate it.cannot tolerate aspirin or a thienopyridine derivative (see below).

Ticlopidine (Ticlid™) and clopidogrel (Plavix™) are thienopyridine derivatives. They inhibit platelet function by interfering with membrane glycoprotein IIb/IIIa receptor conformation, effectively blocking fibrinogen binding. Ticlopidine has been associated with rare but serious adverse outcomes (neutropenia, TTP), which has limited its use. Until recently cClopidogrel hasd been shown to confer only marginal benefit over aspirin in secondary stroke prevention.^,,

Warfarin’s role in primary and secondary cardioembolic stroke prevention in patients with atrial fibrillation is well documented (see Chapter 9). Data on coumarin derivatives in secondary prophylaxis for stroke in general show an increase in stroke and death from hemorrhagic complications, although data from studies done after 1974 (when better coumarin derivatives became available) trend towards an insignificant benefit. The recent Warfarin vs. Aspirin in Recurrent Stroke Study (WARSS) compared warfarin to aspirin in preventing recurrent ischemic stroke. Warfarin was equivalent, but not superior to, aspirin in safety and efficacy. Notably, subgroup analysis revealed trends toward superior efficacy of warfarin in cryptogenic stroke (thought to reflect occult cardioembolic disease), and of aspirin in large vessel and lacunar stroke.

· HTN treatment with thiazides, beta-blockers, or ACEIs, even in the very old

• Tight BP control in diabetics ( < 130/80)

· Smoking cessation

· Statins in high risk patients

· Moderate alcohol consumption, from once a year to twice a day

· CEA

- after TIA or stroke when stenosis > 70%

- in symptomatic stenosis 50 – 69 % after careful risk stratification

- in asymptomatic stenosis > 70% after careful risk stratification

· Aspirin after TIA or stroke

- Low dose for chronic prophylaxis

- Loading dose (> 150 mg) may be needed in acute stroke

· Dipyridamole for those who cannot tolerate aspirin or clopidogrel

· Exercise: 30 – 60 minutes per day, at least once per week, up to 4 – 5 times per week

We thank Dr. Randolph Marshall for helpful comments and suggestions.