2021 Guideline for the Prevention of Stroke in Patients With Stroke and Transient Ischemic Attack: A Guideline From the American Heart Association/American Stroke Association

Diuretics, angiotensin-converting enzyme inhibitors, and angiotensin receptor blockers have demonstrated benefit in RCTs or systematic reviews of RCTs.^{185–188,206} Although calcium channel blockers are recommended for the treatment of hypertension, there are limited data on their efficacy for secondary stroke prevention. However, the use of calcium channel blockers is reasonable for patients with stroke who require additional medication options.¹⁸⁵

Data from 4 RCTs and recent meta-analyses support a benefit of treating patients with prior stroke or TIA to achieve a goal BP of <130/80 mm Hg. The RESPECT (Randomized Evaluation of Recurrent Stroke Comparing PFO Closure to Established Current Standard of Care Treatment),¹⁹⁰ PAST-BP (Prevention After Stroke–Blood Pressure),¹⁹¹ SPS3 (Secondary Prevention of Small Subcortical Strokes),¹⁹² and PODCAST (Prevention of Decline in Cognition after Stroke Trial)¹⁹³ RCTs all compared intensive control of BP (SBP targets ranging from <120–<130 mm Hg) with standard BP control (SBP targets ranging from <140–<150 mm Hg) in patients with prior cerebrovascular disease. These trials all reported nonsignificant tendencies toward lower recurrent stroke rates in the intensive treatment groups. However, a meta-analysis of these 4 trials showed a significant reduction in recurrent stroke risk with an intensive versus standard target (RR, 0.78 [95% CI, 0.64–0.96]). An independent Cochrane analysis¹⁸⁵ of SPS3, PAST-BP, and PODCAST (done before RESPECT publication) reported a trend toward benefit of intensive BP targets (pooled RR for recurrent stroke, 0.80 [95% CI, 0.63–1.00]). In addition, the largest meta-analysis to date including >40 000 patients from 14 RCTs showed a significantly lower rate of recurrent stroke in patients with an achieved SBP of <130 mg Hg compared with higher SBP groups.¹⁹⁴ It should be noted that for patients with intracranial large artery atherosclerosis, a higher BP target may be appropriate (see Section 5.1.1).

The magnitude of BP lowering appears to be more important for risk reduction than the class of antihypertensive agent used.^{188,189,195,196} Therefore, individual patient characteristics that may affect the safety and efficacy of treatment (eg, T2D, chronic kidney disease, AF) should be taken into account in the selection of antihypertensive agents.

The recommended threshold BP of >130/80 mm Hg for starting antihypertensive medications is informed by the baseline BPs of patients with cerebrovascular disease studied in trials of BP treatment. Among the 4 RCTs comparing intensive and standard BP targets in patients with prior cerebrovascular disease, the RESPECT¹⁹⁰, PAST-BP,¹⁹¹ and PODCAST¹⁹³ trials included patients with baseline SBPs as low as 125 mm Hg. In PAST-BP,¹⁹¹ ≈50% of patients had baseline SBP <140 mm Hg. Similarly, in the PRoFESS trial (Prevention Regimen for Effectively Avoiding Second Strokes) of >20 000 patients with ischemic stroke,¹⁹⁷ ≈33% of patients had baseline SBP <135 mm Hg. The large number of subjects with prior stroke and SBP <140 mm Hg included in these trials supports the safety and efficacy of the use of antihypertensive medications in patients with SBP >130 mm Hg, which is the threshold recommended for secondary prevention of vascular events in the AHA/ACC hypertension guideline.²⁹

What is the optimal timing to begin BP lowering after acute stroke?

What is the optimal time during the day for BP medication administration to prevent recurrent stroke?

Do lower BP targets improve or worsen cognition in patients with prior stroke?

What is the optimal BP target for very elderly patients with stroke?

What is the optimal BP target for patients with diabetes with stroke?

The SPARCL trial included adults who had an ischemic or hemorrhagic stroke (or TIA presumably owing to atherosclerotic causes) in the prior 1 to 6 months and an LDL-C level of 100 to 190 mg/dL. The main exclusion criteria were coronary heart disease, peripheral vascular disease, AF, a prosthetic heart valve, clinically significant mitral stenosis, or sinus node dysfunction. There was no exclusion based on ischemic stroke subtype, although practically, most patients with cardioembolic stroke would be excluded on the basis of the exclusion criteria related to cardiac disease. Eligible patients were randomized to atorvastatin 80 mg or placebo. During a median follow-up of 4.9 years, the primary end point of stroke occurred in 11.2% of patients receiving atorvastatin versus 13.1% of patients receiving placebo (adjusted HR, 0.84 [95% CI, 0.71–0.99]).²⁰⁸

The TST trial included adults with cerebral infarction in the prior 3 months or high-risk TIA (that included at least arm and leg motor deficit or speech disturbance lasting >10 minutes) in the prior 15 days, documented atherosclerotic disease (defined as carotid, aortic, intracranial, or coronary atherosclerotic disease), and a clear indication for statin therapy. The main exclusion criteria were a cardioembolic stroke/TIA without documented atherosclerotic disease, a baseline LDL-C <100 mg/dL while not taking a statin, inability to intensify statin therapy (already on maximum dose), or a history of symptomatic hemorrhagic stroke. Eligible patients were randomly assigned to an LDL-C target of <70 mg/dL (lower-target group) versus 90 to 110 mg/dL (higher-target group). LDL-C targets were pursued by intensification of statin therapy and the addition of ezetimibe if needed. During a median 3.5 years of follow-up, the primary composite outcome of major cardiovascular events occurred in 8.5% of those in the lower-target group versus 10.9% in the higher-target group (HR, 0.78 [95% CI, 0.61–0.98]).²¹⁰ In addition, a secondary analysis of the IMPROVE-IT trial (Improved Reduction of Outcomes: Vytorin Efficacy International Trial) found a significantly lower risk of ischemic stroke with ezetimibe treatment (in addition to simvastatin) among patients with previous MI.²¹⁶

Very high-risk patients include those with history of multiple major ASCVD events or 1 major ASCVD event and multiple high-risk conditions. In these patients, additional benefit from LDL-C lowering when LDL-C is > 70 mg/dL (1.8 mmol/L) or non–high-density lipoprotein >100 mg/dL (2.6 mmol/L) by ezetimibe and 2 PCSK9 inhibitors (evolocumab and alirocumab) was demonstrated by 3 RCTs.^211–213 This guideline strongly recommends (COR 1) that ezetimibe be added to maximally tolerated statin as a first step to further lower LDL-C. Although no RCT tested the strategy of ezetimibe first and then a PCSK9 inhibitor, ezetimibe was allowed at entry with statin therapy in both PCSK9 inhibitor trials (FOURIER [Further Cardiovascular Outcomes Research With PCSK9 Inhibition in Subjects With Elevated Risk] and ODYSSEY OUTCOMES [Evaluation of Cardiovascular Outcomes After an Acute Coronary Syndrome During Treatment With Alirocumab]), but only 3% and 5% received ezetimibe during these trials. Ezetimibe use before PCSK9 inhibitor is recommended because ezetimibe is available as a generic drug and has proven safety.²¹¹ This approach is supported by 2 simulation studies from large populations of very high-risk patients showing that ezetimibe plus statin therapy lowers LDL-C to <70 mg/dL (1.8 mmol/L) in most patients, leaving a minority eligible for a PCSK9 inhibitor.^217,218 In addition, the TST trial of patients with ischemic stroke used ezetimibe as second-line therapy to achieve the LDL-C target of <70 mg/dL (1.8 mmol/L),²¹⁰ supporting the use of ezetimibe before PCSK-9 inhibitors in patients with ischemic stroke who are not at their LDL-C targets.

Goals and clinical efficacy for LDL-C lowering are defined and monitored by percentage LDL-C reductions relative to baseline levels. Baseline LDL-C is estimated by pretreatment measurements, chart reviews, or measurement after drug therapy is interrupted. Without a baseline level, response to therapy is difficult to evaluate. Adherence to LDL-lowering diets reduces LDL-C levels 10% to >15%.²¹⁴ Moderate-intensity statins reduce LDL-C levels by another 30% to 49%; high-intensity statins, by ≥50%. Adding ezetimibe or bile acid sequestrants to statin therapy reduces LDL-C by an additional 15% to 25%. Adding a PCSK9 inhibitor to statin plus ezetimibe causes greater reductions. Lifestyle changes and statin therapy are commonly introduced together. The maximum percentage change occurs 4 to 12 weeks after therapy is started, at which time drug efficacy or initial adherence to therapy can be evaluated. Periodic remeasurements can confirm adherence to therapy. Because recommended intensities of drug therapies vary in adolescents, young adults, adults 40 to 75 years, those with severe hypercholesterolemia, and those treated for secondary prevention, the recommended LDL-C levels to achieve also vary. Given the modest differences in LDL-C levels associated with the postprandial state, a nonfasting sample is effective to document baseline lipid levels before initiation of statin therapy.

The REDUCE-IT trial (Reduction of Cardiovascular Events With Icosapent Ethyl–Intervention Trial)²¹⁹ randomized 8179 patients with ASCVD including history of ischemic stroke or TIA (70%) or diabetes with other risk factors (30%) to IPE 2 g twice daily plus statin versus statin alone. Enrollment criteria included fasting triglycerides of 135 to 499 mg/dL and LDL-C of 41 to 100 mg/dL on statin dose for ≥4 weeks. Follow-up for a median of 4.9 years revealed a 25% reduction (17.2% IPE versus 22.0% control; HR, 0.75 [95% CI, 0.68–0.83]; P<0.001) in the primary end point of MACEs. Results for nonfatal stroke and TIA were the same for MACEs. No difference occurred in hemorrhagic stroke. A small, significant increase in AF occurred in those treated with IPE (5.3% versus 3.9%). Because benefits were similar across baseline triglycerides and unrelated to triglyceride levels attained, the mechanism by which IPE reduces MACEs, including secondary prevention of stroke, is unknown, possibly related to factors other than lowering triglycerides. The JELIS trial (Japan EPA Lipid Intervention Study)²²⁰ found a 20% relative reduction of stroke in hypercholesterolemic patients treated with IPE, an ester derivative of eicosapentaenoic acid (EPA), and low-dose statin. Two studies, the STRENGTH trial (Outcomes Study to Assess Statin Residual Risk Reduction With Epanova in High CV Risk Patients With Hypertriglyceridemia)²²⁴ and the OMEMI trial (Omega-3 Fatty Acids in Elderly Patients With Acute Myocardial Infarction)²²⁵ (underpowered), found no benefit from combined EPA/docosahexaenoic acid (DHA) in patients with high CVD risk, diabetes, or CVD. Treatment with EPA/DHA resulted in lower median blood EPA levels in STRENGTH than those achieved by treatment with IPE in REDUCE-IT, which had a 25% reduction in the primary end point. Because triglyceride levels were similar, higher EPA levels may account for the results seen in REDUCE-IT. In addition, the differing effects of EPA and DHA on membrane stabilization may contribute to the lack of effect with EPA/DHA compared with EPA alone or IPE.

Most patients with triglycerides ≥500 mg/dL (≥5.6 mmol/L) have increased VLDL and chylomicrons. Increased chylomicrons occur when triglycerides are ≥500 mg/dL (≥5.6 mmol/L). With increasing triglyceride concentrations, chylomicronemia may intensify and cause acute pancreatitis. Higher triglycerides levels convey greater risk.²²² Patients with triglycerides from 500 to 999 mg/dL are at risk of developing unrecognized marked increases in triglycerides, leading to pancreatitis. Most cases of severe hypertriglyceridemia have a genetic component, but the hallmark of hypertriglyceridemic pancreatitis is the combination of both genetic and acquired causes of elevated triglycerides.²²⁶ To prevent acute pancreatitis, it is reasonable to reduce triglycerides whenever levels exceed 500 mg/dL. This reduction can be achieved by addressing and eliminating the underlying factors by implementing a very low-fat diet²²³ and by adding fibrates or omega-3 fatty acids for patients with persistently elevated severe hypertriglyceridemia.²²¹ These are the most reliable therapies to reduce triglycerides to a safer level. If a fibrate is necessary in a patient being treated with a statin, fenofibrate is safer than gemfibrozil because of a lower risk of severe myopathy.²²⁷ Severe hypertriglyceridemia during pregnancy is best managed in consultation with a lipid specialist.²²⁸

Combination therapy (combined EPA/DHA compared with IPE).

The effect that placebo composition might have on lipids and outcomes (eg, mineral oil versus corn oil).

Intensive control of blood sugar for patients with both type 1 diabetes and T2D has been shown to reduce risk for microvascular complications, including retinopathy, nephropathy, and peripheral neuropathy. The first evidence emerged from the DCCT (Diabetes Control and Complication Trial), which enrolled patients with type 1 diabetes between 13 and 39 years of age.²²⁹ Participants were randomized to conventional glucose control (goal to avoid symptoms of hyperglycemia, to avoid ketonuria, and to preserve normal growth) or intensive control to achieve near-normal glucose. After a mean of 6.5 years, intensive control delayed the onset or progression of nephropathy, retinopathy, and neuropathy. The findings were confirmed for patients with T2D in the UKPDS (United Kingdom Prospective Diabetes Study).²³⁰ As in DCCT, however, intensive control was associated with increased risk for hypoglycemia. The American Diabetes Association recommends a target HbA1c <7% for most adult patients.²⁶² The American Diabetes Association, however, advocates for less stringent goals (eg, 7%–8%) for patients with a limited life expectancy, history of hypoglycemia, long-standing disease, or advanced microvascular or macrovascular disease when the risk and inconvenience of intensive control outweigh the potential benefit.

In response to the development of new classes of glucose-lowering medications that also prevent clinical vascular disease, the American Diabetes Association, the European Association for the Study of Diabetes, and other professional organizations have revised their algorithms for the management of T2D. An evidence-based consensus report by the American Diabetes Association and European Association for the Study of Diabetes recommends metformin and comprehensive lifestyle optimization as first-line therapy.²⁶³ In patients with established ASCVD, including ischemic stroke, when prevention of further vascular events is the priority, GLP-1 receptor agonist therapy should be added to metformin independently of baseline HbA1c. When concern for heart failure or chronic kidney disease predominates, addition of a sodium-glucose cotransporter 2 inhibitor to metformin is recommended. These developments in management have implications for the care of patients with diabetes and ischemic stroke. Clinicians should now engage patients in a discussion of the new therapies and alternatives. Through shared decision-making, clinicians should help patients decide if a GLP-1 receptor agonist or a sodium-glucose cotransporter 2 inhibitor is right for them. Risk for future vascular disease is only one consideration in selecting among available diabetes medications. Cost, side effects, desire for weight loss, aversion to injection therapy, and desire to reduce risk for hypoglycemia are also important.

Optimal management of T2D is achieved with a multidimensional approach that includes (1) medical nutritional therapy; (2) lifestyle counseling (for physical activity, weight loss, smoking cessation, etc); (3) diabetes self-management, education, and support; and (4) medication therapy.^263,264 For patients with overweight or obesity, it can be helpful to include intensive behavioral counseling for weight loss in medical nutritional therapy. T2D in most patients is highly responsive to weight optimization; weight loss can result in diabetes remission for some patients and reduced medication use in most.

Multifactorial interventions to simultaneously improve diet quality, increase physical activity, and reduce body weight reduce risk for progression to diabetes among patients with impaired glucose tolerance by 43% to 58% during 3 to 5 years of treatment.^237,238 The benefit achieved is in proportion to patients’ ability to achieve lifestyle goals.

Approximately 20% of patients with acute ischemic stroke will be found to have undiagnosed diabetes after testing with an HbA1c or oral glucose tolerance test.²⁵⁵ These patients are at risk for progressing to symptomatic disease and for developing complications related to their diabetes, including microvascular and macrovascular disease. With the development of drugs that both control glucose and reduce risk for major adverse cardiovascular events, it has become more important to identify all patients with diabetes after an ischemic stroke so that they can be offered appropriate therapy. Available methods to screen for diabetes include fasting blood sugar, oral glucose tolerance test, and HbA1c. Of these, HbA1c has advantages in that it is more convenient (ie, it does not require fasting), has less variability day to day, and is less likely to be perturbed by medications, stress, or illness.²⁶⁵ In general, HbA1c may be more accurate than other screening tests in the immediate postevent period. It is probably the preferred method of diagnosis for patients hospitalized with an acute stroke.

Compared with evidence for a benefit in microvascular disease, there is less evidence for a benefit of intensive glucose control on macrovascular end points such as stroke. No macrovascular benefit was demonstrated in the main publications from DCCT, the UKPDS, or 3 other more recent trials designed to test intensive control for prevention of macrovascular disease.^239–241 However, a benefit of intensive control emerged during long-term follow-up of the DCCT²⁴² and the UKPDS. Two meta-analysis of 5 trials that included the UKPDS reported a benefit of intensive control for some cardiovascular outcomes but not for stroke.^243,244 No trials or meta-analyses have reported the effect of intensive control on patients with a history of stroke. Many experts conclude that the benefit of intensive control on macrovascular disease is likely restricted to younger patients with recent-onset diabetes and without established vascular disease. In this regard, however, some professional organizations recommend targeting an HbA1c of 7% to 8% or even 8% to 9% to reduce the risk for hypoglycemia in elderly individuals with limited life expectancy or significant comorbid illness.^266,267

Although lifestyle change is the safest and most efficacious method to prevent progression from prediabetes to diabetes, several drugs have demonstrated benefit. In the Diabetes Prevention Program trial, patients with impaired glucose tolerance were randomized to standard lifestyle recommendations plus metformin 850 mg twice a day, standard lifestyle recommendation plus placebo, or intensive lifestyle counseling. The primary outcome, progression to diabetes, was observed in 11.0%, 7.8%, and 4.8% of patients in the placebo, metformin, and intensive counseling groups. The rate of progression was 58% lower (95% CI, 48–66) with intensive counseling and 31% (95% CI, 17–43) lower with metformin compared with standard lifestyle counseling. Metformin is well tolerated and inexpensive. Other drugs that have been shown to reduce progression to diabetes include pioglitazone,²⁴⁵ acarbose,²⁴⁶ liraglutide,²⁴⁷ and dapagliflozin.²⁶⁸ These are not as well tolerated as metformin. Liraglutide is delivered by injection.

Approximately 30% of patients with ischemic stroke have prediabetes as defined by HbA1c and other measures of glucose metabolism.²⁵⁵ Most of these patients will have insulin resistance as a contributing cause, but insulin resistance can occur before prediabetes develops. Thus, ≈50% of patients without diabetes with ischemic stroke have insulin resistance.^269,270 Both conditions have been associated with increased risk for first ischemic stroke,^{252,253,271–275} but no research has been designed to test their effect on risk for recurrent vascular events. One large clinical trial has examined the effect of a specific intervention for patients with ischemic stroke and insulin resistance. The IRIS trial (Insulin Resistance Intervention After Stroke) examined the effect of the insulin-sensitizing agent pioglitazone compared with placebo among patients without diabetes who had a recent ischemic stroke and insulin resistance. Patients with heart failure were excluded. After a mean of 3.8 years, pioglitazone reduced the risk of recurrent stroke or MI by 24% (RR, 0.76 [95% CI, 0.62–0.93]), from 11.8% among placebo recipients to 9.0% among pioglitazone recipients. Active treatment, however, was associated with weight gain and increased bone fracture risk. These adverse events have restrained clinical use of pioglitazone.

It is plausible that interventions for prediabetes, established diabetes, and glucose management will reduce risk for neurocognitive outcomes and subclinical ischemic damage and discrete stroke events. The IRIS trial demonstrated that 1 glucose-lowering intervention, pioglitazone, prevented stroke and MI in patients with insulin resistance without diabetes. The benefit of other interventions for prediabetes, including GLP-1 receptor agonists, dietary improvement, and weight reduction, has not been investigated but would logically follow from the IRIS trial.

The same interventions could be examined in patients with established diabetes.

Some early investigations have explored the effectiveness of poststroke behavioral counseling and physical training, modeled on cardiac rehabilitation, for improving lifestyle after stroke.

Because diabetes disproportionately affects Black and Hispanic communities, effective community-based research for diabetes prevention and management may yield interventions to reduce disparities in stroke risk and recovery.

Because patients with ischemic stroke or TIA tend to be older and to have established vascular disease, the optimal goal for glycemic control remains uncertain. In this regard, however, some professional organizations recommend targeting an HbA1c of 7% to 8% or even 8% to 9% to reduce the risk for hypoglycemia in elderly individuals with limited life expectancy or significant comorbid illness.^266,267

Our recommendation for weight loss in patients with overweight or obesity after stroke is consistent with the 2013 AHA guideline on obesity,²⁷ the 2014 AHA guideline on primary prevention,¹⁸ and other professional guidelines on management of obesity.²⁹⁹ Even a modest weight loss of 5% to 10% is associated with important improvements in conventional cardiovascular risk factors.²⁷⁹ In particular, weight loss can improve glucose control in patients with diabetes.²⁷⁷ The DiRECT trial (Diabetes Remission Clinical Trial) compared food substitution followed by stepped food reintroduction and structured support with best practice by guideline among patients with diabetes and BMI of 27 to 45 kg/m². After 12 months, 24% of participants in the intervention group lost ≥15 kg compared with none in the control group. Diabetes remission was achieved in 46% of patients in the intervention group and 4% in the control group. Studies of metabolic surgery confirm that weight loss is associated with improvements in glucose control, BP, indicators of inflammation, and lipid metabolism.^259,276,278

Available strategies for helping patients achieve meaningful weight loss include intensive behavioral counseling, drugs to reduce appetite, and bariatric surgery. Of these, intensive behavioral counseling is recommended as the first option in evidence-based guidelines because of its greater effectiveness and safety compared with pharmacotherapy and greater safety and lower cost compared with metabolic surgery.^{27,281,299,300} Components of multicomponent behavioral programs typically include goal setting, feedback, problem solving, coaching for physical activity, and frequent individual or group meetings (at least 12 sessions in the first year).³⁰⁰ Examples include the Diabetes Prevention Program,²³⁸ Weight Watchers,²⁸⁰ and the Counterweight Program.²⁵⁸ On average, behavior-based programs achieve ≈2.4 kg greater weight loss than control intervention over up to 24 month, and participants are more likely to lose 5% of their body weight.²⁸¹ Ideally, patients with obesity, according to need, are offered care in a comprehensive bariatric program in which all strategies for weight reduction are offered.

In 2012, the US Preventive Services Task Force issued a Grade B recommendation to screen all adults for obesity.³⁰¹ The recommendation is consistent with the AHA guideline on obesity from 2013 and more recent guidance from other professional organizations.^27,299 Although there is no direct evidence that weight loss interventions among patients with acute ischemic stroke reduce risk for recurrent stroke or otherwise improve future brain health, there is clear evidence that intensive counseling can help patients achieve meaningful weight loss.²⁸¹ The US Preventive Services Task Force determined that the risks for intensive counseling were small to none.³⁰⁰ For selected patients, pharmacotherapy and bariatric surgery also may be effective, although surgical trials typically have excluded patients with stroke. Specific benefits of weight loss among patients with ischemic stroke include improved glucose metabolism, BP, and lipid metabolism. It also has a favorable effect on obstructive sleep apnea (OSA), AF, and vascular inflammation. Screening is necessary to identify adults for referral to effective weight loss therapy.

Our recommendation to treat patients with OSA using CPAP is consistent with a guideline from the American College of Physicians that recommends CPAP as the initial therapy.³³⁰ In studies of various populations, treatment with CPAP effectively reduced the AHI and improved measures of sleepiness, BP control, sleep-related quality of life, and physical functioning.^302,303 As described in the above synopsis, limited data from small trials specifically among patients with stroke show that treatment with CPAP is safe, reduces the AHI, improves sleepiness, and may improve neurological function.^304–313 Evidence from 1 trial suggests that treatment may improve mood.³¹⁴ In a planned ancillary analysis in the SAVE trial, which enrolled patients with either coronary or cerebrovascular disease, CPAP reduced the risk for stroke in patients with better CPAP adherence.³⁰³ The American College of Physicians recommends mandibular advancement devices as an alternative for patients who do not tolerate CPAP or who prefer mandibular advancement.³³⁰ These devices, however, have not been specifically tested in patients with stroke.

OSA is common among patients with stroke, but most cases are undiagnosed.³¹⁵ The combined prevalence of known and unknown moderate to severe OSA approaches 40%. Arguing against evaluating patients for undiagnosed OSA is the absence of evidence that treatment improves stroke-related outcomes, including recurrence. Since the last version of this guideline, a trial of 2717 patients with moderate to severe OSA and coronary or cerebrovascular disease failed to show a benefit of CPAP on recurrent vascular events.³⁰³ In favor of evaluating stroke patients is the high prevalence of undiagnosed OSA, combined with evidence that treatment improves daytime sleepiness, BP, sleep-related quality of life, and physical functioning.^302,303 If an evaluation is undertaken, options include questionnaires to identify patients at high risk for OSA who can be selectively referred for polysomnography.^332,333 However, available questionnaires show inconsistent performance in patients with stroke.³³² In addition, the prevalence of OSA after stroke is sufficiently high to justify omitting prescreening before polysomnography when detection of asymptomatic OSA is clinically warranted. Facility-based (ie, sleep laboratory–based), multichannel polysomnography is the reference standard for diagnosing OSA. Home (ie, out of sleep laboratory) monitors are appropriate in selected patients.^316,317 Although the American Academy of Sleep Medicine recommends against the use of home testing in patients with stroke,³³⁴ recent research suggests that home testing can be effective.³¹⁵

The selection of patients who might benefit from CPAP,

Timing of testing and treatment in relation to stroke onset,

Type of testing (in home or in facility), and

Dose/type of CPAP.

Sleep SMART (Sleep for Stroke Management and Recovery Trial), a randomized trial to determine whether early treatment of OSA with CPAP after ischemic stroke or TIA reduces the risk for 2 primary outcome measures: recurrent stroke, acute coronary syndrome, and all-cause mortality 6 months after the event and functional status at 3 months (ClinicalTrials.gov identifier NCT03812653).

The RISE-UP trial (Recovery in Stroke Using PAP) focused on optimal timing of CPAP initiation after stroke (NCT04130503).

ASAP (Addressing Sleep Apnea Post Stroke/TIA), which will test a quality improvement initiative in the Veterans Affairs Medical Administration and include a secondary aim of recurrent vascular disease (NCT04322162).

Within the realm of surgery, there is considerable controversy as to what constitutes the best surgical option, direct superficial temporal artery to middle cerebral artery bypass or an indirect bypass procedure such as encephaloduroarteriosynangiosis or encephalomyosynagiosis. The differences in surgical technique even within the broad categories of direct and indirect bypass, the frequent use of combined procedures, and the heterogeneity of the patient population make it difficult to draw robust conclusions about the superiority of one procedure over another. Again, there are no prospective RCTs to inform recommendations in this area. Although several meta-analyses and case series favor direct over indirect bypass,^391–395 all but 1 of these studies combined both hemorrhagic and ischemic patients. A propensity score–matched analysis on 220 adult patients with ischemic-type moyamoya disease showed that direct bypass was superior to indirect bypass for the prevention of recurrent stroke.³⁹¹ In this series, however, the primary end point of recurrent stroke was defined as both ischemic and hemorrhagic events. In contrast, another meta-analysis³⁹⁶ and a retrospective multicenter series³⁹⁷ reported no difference between direct and indirect modalities.

Although traditionally thought to result from poor perfusion, there is some evidence that many instances of ischemia in moyamoya disease actually result from thromboembolic phenomenon.⁴⁰⁶ In a propensity score–matched analysis, Onozuka et al⁴⁰⁰ found that prehospital antiplatelet use was significantly associated with good functional status on hospital admission for patients with nonhemorrhagic moyamoya disease in Japan. Several meta-analyses and case series have compared antiplatelet therapy with surgical bypass or observation alone with mixed results.^{393,397–399} Although some found a benefit or no difference with antiplatelet therapy,^394,397,398 others found surgery to be superior.^393,399 However, many studies include patients with both ischemic and hemorrhagic moyamoya disease, and conclusions are often based on subgroup analyses. Antiplatelet therapy may also be of benefit in conjunction with surgical revascularization.⁴⁰¹ An international survey of perceived experts in the treatment of moyamoya disease reported that the majority of non-Asian respondents recommended antiplatelet therapy, in contrast to their Asian counterparts, perhaps reflecting a difference in the typical pattern of disease seen in the 2 geographic regions.⁴⁰⁷

To better understand the natural history of moyamoya disease, including ethnic and geographic variability.

To examine the efficacy of medical management options, including antiplatelet therapy and anticoagulation, with or without surgical intervention for ischemic moyamoya disease.

To determine the most effective surgical intervention for moyamoya disease.

To establish the optimal timing for surgical intervention after an acute clinical event in a patient with moyamoya disease.

Cilostazol has been studied as an alternative to aspirin therapy for ischemic stroke related to small vessel disease. In the CSPS trial (Cilostazol for Prevention of Secondary Stroke), cilostazol was compared with placebo in 1095 Japanese patients with noncardioembolic ischemic stroke (74% small vessel stroke).⁴¹⁶ There was a significant reduction in the risk of recurrent stroke in the cilostazol treatment group and in participants with lacunar stroke in the subgroup analysis. Cilostazol was compared with aspirin for the prevention of ischemic stroke in the CSPS II trial, which enrolled 2575 Japanese participants, 64% of whom had small vessel stroke as the qualifying event.⁴¹⁷ Cilostazol was associated with a significantly reduced risk of first occurrence of ischemic or hemorrhagic stroke in the cilostazol group compared with the aspirin group. There were significantly fewer hemorrhagic events in the cilostazol group. There was a nonsignificant reduction in recurrent stroke risk in participants with small vessel stroke in the subgroup analysis. There were more reported side effects, including headache, dizziness, palpitations, diarrhea, tachycardia, and constipation, in the cilostazol group. These findings have not yet been duplicated, and the inclusion of a single ethnic group may limit generalizability.

The role of cilostazol in secondary prevention after stroke related to small vessel disease.

Targeted strategies for secondary prevention after small vessel stroke that also reduce the risk of vascular dementia.

The Fontan circulation is a palliative approach to complex single-ventricle physiology and is associated with complex hemodynamics related to an obligatory chronic elevation in central venous pressure and reduced cardiac output.³⁰ Consideration for antithrombotic therapy in Fontan patients should take into account the high prevalence of thrombus formation and potentially catastrophic impact of pulmonary or systemic thromboembolism. The physiology associated with the Fontan circulation may lead to stroke/TIA via different mechanisms. The associated ventricular dysfunction and abnormal ventricular morphology, the progressive atrial scarring from the Fontan surgery, and the elevated intra-atrial pressures may trigger atrial arrhythmias, increasing the risk of thromboembolic events.⁵⁸¹ The role of DOACs in patients with Fontan is less defined, and thus far, the published evidence is limited to small observational series.^582,583 Given the lack of conclusive evidence on the safety of DOACs in the presence of the hepatic dysfunction and altered coagulation in patients who underwent Fontan palliation, warfarin should be considered the preferred oral anticoagulant regimen.

Complex congenital heart disease is associated with an increased risk of stroke in adults.^574–576 Mandalenakis et al⁵⁷⁵ conducted a large prospective registry in Sweden and documented an almost 11-fold increase in the risk of ischemic stroke in patients with congenital heart disease compared with control subjects matched by age, with the highest risk among patients with complex lesions. In a retrospective large database including >29 000 adult (age >18 years) patients with congenital heart disease, Lanz et al⁵⁷⁴ reported an overall risk of ischemic stroke of 6.1% in women and 7.7% in men. In addition, in this study, patients with complex congenital heart disease had the highest risk (8.9% [95% CI, 6.0–11.5]). The mechanisms underlying such an increased risk of stroke are different and multifactorial and are related to the high prevalence of atrial arrhythmias, risk for paradoxical embolization (eg, cyanotic congenital heart disease), atrial mechanical dysfunction resulting from the presence of scar, and associated hyperviscosity/hypercoagulability caused by chronic hypoxemia. In patients with cyanotic and other complex congenital heart disease and ischemic stroke or TIA of presumed cardioembolic origin, treatment with warfarin is reasonable in the absence of contraindications to oral anticoagulation (eg, increased risk of bleeding).

A subset of patients with Fontan palliative surgery have a fenestration between the Fontan pathway and the pulmonary venous atrium to provide a controlled right-to-left shunt to augment ventricular preload and to partially offload systemic venous (Fontan) hypertension. Right-to-left shunt from a fenestration can be a potential source of thromboembolism and stroke.⁵⁸⁴ Catheter closure of Fontan fenestration along with anticoagulation has been performed in some patients with a patent fenestration presenting with stroke,⁵⁸⁵ although more data are needed to determine whether this approach is of real incremental benefit to prevent ischemic stroke or TIA compared with oral anticoagulation therapy alone.

DOACs have been demonstrated to be noninferior or superior to dose-adjusted warfarin therapy to prevent thromboembolism and bleeding in patients with nonvalvular AF, and they have rapidly become the standard of care for these patients. Important advantages of DOAC therapy compared with warfarin include a more predictable pharmacokinetic profile requiring no monitoring, fewer interactions with other drugs, and a rapid onset/offset of action. Thus far, data on the efficacy and safety of DOACs to prevent recurrent thromboembolism in patients with congenital heart disease and history of ischemic stroke or TIA are limited to small observational series. Larger-scale multicenter observational data will be necessary to achieve adequate statistical power and to clarify whether DOACs can be used safely in patients with congenital heart disease. Multicenter registries such as that for non-VKA anticoagulants for thromboembolic prevention^585a are currently ongoing and will provide important and needed data on the safety and efficacy of DOACs for thromboembolic prevention in patients with congenital heart disease.

Retrospective studies have shown increased risk of stroke with left-sided cardiac myxoma and papillary fibroelastoma.⁵⁸⁶ Atrial myxomas can occur in any cardiac chamber but are noted most frequently in the left atrium adherent to the interatrial septum. Broad morphological variation of atrial myxomas has been described, and more villous tumors are felt to have greater embolic and thromboembolic potential. Papillary fibroelastomas are fibrinous mobile tumors occurring most frequently adherent to a cardiac valve with nearly equal distribution across the cardiac valves.⁵⁸⁷ In a single-center study of 323 patients after excision of a cardiac tumor, the OR of embolism associated with left atrial tumors was 1.95 and higher (4.17) with aortic valve tumors.⁵⁸⁶ In another series of 725 patients with papillary fibroelastoma, tumor mobility and aortic valve location were independent predictors of embolism.⁵⁸⁸ Less common tumors and metastatic lesions in the heart have been anecdotally associated with stroke; however, there is not enough evidence to make a consistent recommendation for those circumstances. For patients with stroke or TIA whose evaluation reveals a left-sided cardiac tumor, consideration should be given to tumor excision to reduce the risk of recurrent stroke.

Most cardiac tumors are found incidentally, carrying the threat of stroke or recurrent stroke, and open heart surgery is the only established treatment option. Better understanding of the risk modification from antiplatelet or anticoagulation in particular tumor subtypes or morphologies may provide alternative therapeutic options for patients. Specifically, pooled experiences including national and multi-institutional data sets may allow a better understanding of tumor subtypes, fractionation by morphology, and natural history.

As technology advances, minimally invasive and percutaneous treatment strategies should be investigated. Similarly, advances in imaging technology will allow greater delineation of tumor subtypes and morphologies. The prognostic implications of these findings should be investigated.

The recommendation of using antithrombotic agents is based on expert opinion. Although arterial dissection is a hemorrhagic process within the arterial wall, the most common mechanism for stroke is artery-to-artery embolism⁵⁹⁸ from an intraluminal thrombus. This is the rationale for using antithrombotic agents to decrease the rate of ischemic stroke.⁵⁹⁹

CADISS (Cervical Artery Dissection in Stroke Study) was an open-label trial that randomized 250 patients with extracranial dissection within 7 days from symptom onset to receive either antiplatelets or anticoagulants in the following 3 months.⁵⁹³ The specific agents and treatment regimens were left to the treating physician. Stroke or death occurred in 3 patients (2%) assigned to antiplatelets versus 1 patient (1%) in the anticoagulation group (OR, 0.335 [95% CI, 0.006–4.233]; P=0.63).⁵⁹³ At the 1-year follow-up, there was no difference in the primary end point of stroke or death between the 2 groups: 3.2% in the antiplatelet group versus 1.6% in the anticoagulant group (OR, 0.56 [95% CI, 0.10–3.21]; P=0.51).⁵⁹⁴ This trial supports that there is equipoise between antiplatelets and anticoagulants in the first 3 months after a cervical artery dissection.

Endovascular procedures are used to reconstruct cervical arterial dissections refractory to medical treatment and to treat complications such as pseudoaneurysms.^595,596 There is a lack of randomized studies to support benefit of endovascular interventions in dissections, and most of the data stem from small series. Still, reviews of the literature suggest a reasonably low rate of complications from these procedures.^595,596

Early identification of patients at risk for early deterioration or complications who might benefit from endovascular therapy.

Whether some antithrombotic agents are more efficacious if administered early in the course (eg, is anticoagulation more efficacious than antiplatelets in the acute phase?).

The benefit of long-term antithrombotic agents.

Management of dissections complicated with intraluminal thrombus.

Management of intracranial dissections and risks/benefits of anticoagulation.

The determination of what physical activities are safe and which ones need to be restricted.

Optimal monitoring/management of pseudoaneurysms associated with dissection needs further study.

Results from VITATOPS (Vitamins to Prevent Stroke)⁶²⁵ and VISP (Vitamin Intervention for Stroke Prevention)⁶²⁶ did not show a significant reduction in risk of secondary stroke among patients with recent stroke who were randomized to receive various doses of folic acid and vitamin B supplementation.

A subgroup analysis in the ROCKET AF trial comparing rivaroxaban and warfarin in patients with AF and history of cancer showed similar rates of stroke and systemic embolism 0.52 (95% CI, 0.22–1.21) and nonmajor bleeding events 1.09 (95% CI, 0.82–1.44).⁶⁴⁷ Similarly, the efficacy (1.09 [95% CI, 0.53–2.26]) and safety (0.76 [95% CI, 0.45–1.29]) of apixaban were comparable to those of warfarin in patients with AF and history of cancer in the ARISTOTLE trial.⁶⁴⁸ This was supported by claims data analysis looking at bleeding events of rivaroxaban (1.09 [95% CI, 0.79–1.50]), dabigatran (0.96 [95% CI, 0.72–1.27]), and apixaban (0.37 [95% CI, 0.17–0.79]) compared with warfarin.⁶⁴⁹ A literature search of 31 660 patients in 5 trials in patients with AF and cancer DOACs showed a trend for more efficacy and better safety compared with warfarin.⁶⁵⁰

Patients who had a stroke attributable to hypercoagulability from cancer may be at a particularly high risk of bleeding with anticoagulation. In these patients, the benefit of anticoagulants for secondary stroke prevention is not well established, and further research is needed.

The particular need of anticoagulation to prevent stroke in different type of cancers is not known and should be a matter of further research.

In patients who had a stroke attributable to hypercoagulability from cancer who are anticoagulated, there is a paucity of data on the best treatment regimen for these patients.

Although low-molecular-weight heparin is often used in patients with cancer and stroke to prevent thromboembolic complications, the potential benefit on stroke prevention is unknown.

The recommendations for treatment of patients with SCD are based on stroke primary prevention studies in a pediatric population and a secondary stroke prevention study in a population of pediatric patients with SCD with silent cerebral infarctions (SIT [Silent Cerebral Infarct Transfusion] multi-center clinical trial).⁶⁷¹ The STOP trial was a randomized, placebo-controlled trial that showed that a long-term prophylactic transfusion strategy was effective for the primary prevention of stroke in children with SCD and high TCD velocities.⁶⁵⁵ Blood transfusion was used because it was demonstrated effective in reducing recurrent clinical ischemic stroke and progression of arterial stenosis in a series of children with SCD.⁶⁵⁶ This was associated with a reduction in the rate of recurrent stroke during a mean follow-up of 3 years compared with historical control subjects.⁶⁵⁷ Most of the patients in the series were children, and it is unclear whether adults have the same untreated risk or benefit from treatment.⁶⁵⁸ Regular transfusions are associated with long-term complications, especially iron overload, typically requiring iron chelation therapy. The SIT trial showed reduced recurrent stroke in children with SCD and silent cerebral infarct who underwent transfusion compared with observation.⁶⁷¹

SWiTCH (Stroke With Transfusions Changing to Hydroxyurea) was a randomized secondary stroke prevention trial in a pediatric SCD population that found no recurrent strokes with long-term transfusion but 10% with hydroxyurea.⁶⁵⁹ In situations in which transfusion is not available, a nonrandomized study of patients with an initial stroke suggested that patients who do not receive hydroxyurea have a higher recurrent stroke risk (HR, 9.4 [95% CI,1.5–70.6]).⁶⁶⁰ TWiTCH (TCD With Transfusions Changing to Hydroxyurea) was a primary stroke prevention randomized trial of children with SCD and abnormal TCD velocities (>200 cm/s) but no severe vasculopathy on MRA⁶⁶¹ who had received 1 year of transfusions. Children were subsequently randomized to transfusion or hydroxyurea. Noninferiority was shown: 3 TIAs in each group and no new MRI stroke in either group. Hematopoietic cell transplantation can be curative for SCD⁶⁶² but is usually undertaken in children refractory to other treatment; it results in survival without SCD in 80% to 90% of patients. Clinical and subclinical infarctions have been reported to be arrested by this procedure.⁶⁶³ Surgical bypass has been reported with improved outcomes in patients with SCD with moyamoya, but no randomized data are available.^664,665 Given the lack of experience with antithrombotics, antihypertensive agents, and statins for secondary stroke prevention in SCD, specific recommendations cannot be stated outside of general treatment recommendations.

An ecological study of a newborn screening program in Ireland supports the effectiveness of interventions to drive free homocysteine to ≤11 mmol/L to prevent thromboembolic events.⁷³⁹ A study of treatment of severe hyperhomocysteine performed in Australia, the Netherlands, and Ireland on patients who had cystathionine β-synthase deficiency (mean age, 27.8 years; range, 2.5–70 years) supported a dramatic reduction of risk of vascular events relative to a historical cohort study (RR, 0.091 [95% CI, 0.043–0.190]; P<0.001).⁷⁴⁰

A systematic review of RCTs of agalsidase alfa or beta in patients with Anderson-Fabry disease found that the enzyme replacement therapy improved pain-related quality of life but that their effects on morbidity or mortality required further study.⁷⁴¹

Specific treatments are needed for several single-gene disorders associated with high risk of stroke, including mitochondrial myopathy, encephalopathy, lactic acidosis, stroke-like episodes, cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy, cerebral autosomal recessive arteriopathy with subcortical infarcts and leukoencephalopathy, cathepsin-A–related arteriopathy with strokes and leukoencephalopathy, retinal vasculopathy with cerebral leukodystrophy, and cases of pathogenic mutations in the COL4A1/COL4A2 genes.⁷⁴⁵ Because of the low prevalence of these conditions, generating high-quality evidence for therapies for single-gene disorders that cause stroke will be challenging but not impossible.

For patients with mitochondrial myopathy, encephalopathy, lactic acidosis, and stroke-like episodes, prospective multicenter single-arm studies suggest that oral l-arginine may extend the interictal phase of stroke-like spells and intravenous l-arginine may improve headache, nausea/vomiting, impaired consciousness, and visual disturbances. However, the intravenous treatment may result in fevers and lower hemoglobin.⁷⁴⁶ l-Arginine may work by acting as a nitric oxide precursor.⁷⁴⁷ Although l-citrulline raises nitric oxide production more than l-arginine, it has not been systematically studied in humans.⁷⁴⁷ Future RCTs are needed to clarify dosage, timing of initiation relative to onset of an attack, duration of therapy, and safety for nitric oxide precursors for treating patients with mitochondrial myopathy, encephalopathy, lactic acidosis, and stroke-like episodes before treatment recommendations can be made.

The optimal management of symptomatic carotid web is unknown. Medical management with antithrombotic therapy is first-line treatment; however, it is not known whether SAPT, short-term DAPT, or anticoagulant therapy is superior. In the absence of such data, it is recommended to treat patients with antiplatelet therapy first line or to follow antithrombotic recommendations in this guideline.

There is a high risk of recurrent stroke or TIA in patients with symptomatic carotid web on medical management, estimated in 29% to 56% of patients.^749,750 Carotid stenting or CEA is a good alternative treatment for patients with symptomatic carotid web, with published series revealing no recurrent stroke risk.^748,749

High BP is common in patients with FMD, from either essential hypertension or renovascular hypertension related to FMD.⁷⁵² The optimal BP target for patients with FMD is unknown. In the US FMD registry, most patients were on antihypertensive medicine.⁷⁵³ Smoking cessation should be encouraged for all patients with FMD who smoke because of its health benefits in the general population. The recommendation of using antiplatelet therapy for secondary stroke prevention is based on current practice and expert opinion.^752,753,755 In the US FMD registry, 73% of patients were prescribed antiplatelet therapy; aspirin was the most common agent.⁷⁵³ There are no placebo-controlled randomized trials of patients with symptomatic or asymptomatic FMD comparing aspirin with placebo. Statins are not routinely prescribed for isolated FMD in the absence of another indication.

Dissection is a common manifestation associated with FMD, present in 11% to 15% of patients with spontaneous carotid or vertebral artery dissection.^757,758 In the US FMD registry, ischemic stroke was present in 19% of patients presenting with cervical artery dissections.⁷⁵⁹ The CADISS study demonstrated equivalent efficacy for recurrent stroke risk in patients with spontaneous, symptomatic cervical carotid dissection randomized to antiplatelet versus anticoagulant for 3 months (3% versus 1%; P=0.66) (see Section 5.7, Dissection).

This recommendation has been adapted from a prior guideline by Brott et al²⁵ on carotid angioplasty with or without stenting in symptomatic patients with patients to COR 2b. There are no comparative data evaluating medical management versus angioplasty or stenting for patients with FMD and recurrent ischemic stroke. A case series of 7 patients with symptomatic FMD revealed no complications with balloon angioplasty.⁷⁵⁴

There are no prospective RCTs comparing antithrombotic therapies with observation alone. However, case series tend to suggest that antithrombotic therapy lowers the risk of recurrent ischemic events compared with the natural history.^761,762 There are also no prospective RCTs comparing antiplatelet strategies (either aspirin or second-generation antiplatelet drugs) with anticoagulation. The case series reviewed demonstrated no clear or consistent benefit for 1 form of antithrombotic therapy over another.^760–763 Although 1 small case series favored anticoagulation, not all of the patients in that series presented with ischemic symptoms, and the series consisted of patients with fusiform aneurysms rather than true vertebrobasilar dolichoectasia.⁷⁶² Although the risks of hemorrhage appear to be low in patients presenting with ischemic symptoms,⁷⁶⁴ there are insufficient data to provide a recommendation for anticoagulation over antiplatelet therapy.

Development of an agreed-on definition for vertebrobasilar dolichoectasia that is distinguishable from the definition of a fusiform aneurysm.

Accumulation of more robust natural history data with detailed clinical outcomes.

Clarification of the type, dosing regimens, and efficacy of antiplatelet therapy versus anticoagulation for primary and secondary ischemic stroke prevention.

Improved understanding of the role of reconstructive procedures, whether endovascular or open surgical, in the management of patients with vertebrobasilar dolichoectasia.

The NAVIGATE ESUS and RESPECT ESUS (Dabigatran Etexilate for Secondary Stroke Prevention in Patients With Embolic Stroke of Undetermined Source) randomized trials found no reduction in secondary stroke risk among patients with ischemic stroke or TIA who were treated with a direct oral anticoagulant.^779,780

A post hoc subgroup analysis of subjects with ESUS enrolled in the SOCRATES trial (4329 subjects, 32.8%) found no treatment effect by ESUS category of ticagrelor on recurrent vascular event risk.⁷⁸¹

What markers of atrial dysfunction are most strongly associated with recurrent stroke risk?

Are there clinical factors that help identify patients with ESUS who may benefit from anticoagulation rather than antiplatelet treatment for secondary prevention?

The use of different strategies for long-term arrhythmia detection and their impact on ESUS diagnosis criteria.

The use of additional cardiac markers of atrial pathology, including atrial size, and their impact on ESUS diagnosis criteria, secondary stroke risk, and response to anticoagulation.

The WARSS study randomized 2206 patients with stroke not attributable to cardioembolism or high-grade carotid stenosis to adjusted-dose warfarin (INR, 1.4–2.8) versus aspirin 325 mg.⁷⁹⁰ The primary end point of 2-year recurrent ischemic stroke or death was no different between the 2 groups (17.8% warfarin versus 16% acetylsalicylic acid; HR, 1.13 [95% CI, 0.92–1.38]). Similarly, a subsequent meta-analysis of 8 other trials of VKA among 5762 patients with TIA or nondisabling noncardioembolic stroke found no benefit in secondary stroke prevention with the use of oral anticoagulation of various intensities compared with antiplatelet agents (medium-intensity anticoagulation: RR, 0.80 [95% CI, 0.56–1.14]; high-intensity anticoagulation: RR, 1.02 [95% CI, 0.49–2.13]); both medium-intensity anticoagulation and high-intensity anticoagulation were associated with a significantly increased risk of bleeding.⁷⁸⁹

A series of single antiplatelet trials conducted in patients with noncardioembolic stroke did not use clinical eligibility criteria in the same manner as the more recent DAPT trials. Thus, some patients in the general noncardioembolic stroke pathogenesis group may qualify for the more specific recommendation for early DAPT. For those who do not meet DAPT clinical criteria, SAPT is recommended. The PRoFESS trial found no difference in secondary stroke prevention after noncardioembolic stroke for aspirin-dipyridamole versus clopidogrel.⁷⁹⁴ The ESPRIT trial (European/Australasian Stroke Prevention in Reversible Ischaemia Trial)⁷⁹² suggested that aspirin-dipyridamole may be slightly more effective than aspirin alone for vascular event prevention, as did the ESPS2 trial (Second European Stroke Prevention Study).⁷⁹¹ ESPS2 also showed that aspirin alone is more effective than placebo in reducing recurrent stroke. The CAPRIE trial (Clopidogrel Versus Aspirin in Patients at Risk of Ischaemic Events) was not focused strictly on secondary prevention but found fewer events among patients treated with clopidogrel compared with those treated with aspirin, although no benefit was seen in stroke reduction among the stroke subgroup. A pooled analysis of trials in patients with lacunar stroke suggested that any of the evaluated antiplatelet agents were similarly effective compared with placebo.⁷⁹³ Aspirin dosing may be guided by a recent patient-level pooled analysis of data from primary prevention RCTs that showed weight-based dosing (eg, aspirin 75–100 mg for patients weighing <70 kg and higher doses for those >70 kg) is more effective for the prevention of vascular events.⁸⁰³ However, similar studies have not been performed for secondary stroke prevention.

Recommendations for the short-term (21–90 days) use of DAPT with aspirin and clopidogrel are informed by 2 major RCTs,^384,410 4 meta-analyses,^{802,804–806} and a systematic review by an AHA Evidence Review Committee.¹⁵ Both the CHANCE and POINT trials compared DAPT with SAPT with aspirin initiated early after onset (<24 and <12 hours, respectively) of minor stroke (NIHSS score ≤3) or high-risk TIA (ABCD² score ≥4) but varied in duration of DAPT treatment (21 and 90 days, respectively). DAPT dosing was also slightly different in the 2 trials: POINT used a 600-mg clopidogrel load (then 75 mg daily) and an aspirin regimen of 50 to 325 mg daily, whereas CHANCE used a 300-mg clopidogrel load (then 75 mg daily) and an aspirin load of 75 to 300 mg followed by 75 mg daily. Both trials demonstrated a reduction in recurrent ischemic stroke at 90 days. Further analyses of POINT revealed that the benefit of DAPT was seen primarily in the first 21 days.³⁶⁵ Meta-analyses that included other DAPT trials also have shown a reduction in recurrent stroke from DAPT compared with aspirin monotherapy, but the benefit is limited if not started early (<7 days) after the index event. Although most studies showed some increase in bleeding risk with DAPT, this was offset by the stroke prevention benefit if limited to short-term use.^795,796,806

The use of ticagrelor (180-mg loading dose, then 90 mg twice daily) plus aspirin (300- to 325-mg loading does, then 75–100 mg daily) for 30 days was shown in the THALES trial to be slightly superior to aspirin alone in preventing recurrent stroke (recurrent stroke rate, 5% versus 6.3%; P=0.004) but was also associated with significantly increased risk of severe bleeding (0.5% versus 0.1%; P=0.001).⁷⁹⁷ ICH was also significantly increased among the DAPT group (0.4% versus 0.1%; P=0.01), and significantly more patients in the DAPT group discontinued treatment because of bleeding (2.8% versus 0.6%; P<0.001); the number needed to treat to prevent 1 primary outcome event was 92, and the number needed to harm was 263 for severe bleeding. Further study of this DAPT combination is needed to fully evaluate benefit and risk and to determine whether there is a specific subgroup most likely to benefit. The use of ticagrelor alone compared with aspirin was not associated with any reduction in cardiovascular events.⁸⁰⁷ The use of ticagrelor plus aspirin for the subgroup of patients with ipsilateral carotid artery stenosis and intracranial stenosis is discussed in Section 5.1.1, Intracranial Large Artery Atherosclerosis.⁸⁰⁸

No randomized trials have focused on the question of changing long-term antiplatelet therapy for the prevention of recurrent stroke. A meta-analysis suggested that, among the 4 studies that examined stroke outcomes, there was a reduced risk of recurrent stroke (HR, 0.70 [95% CI, 0.54–0.92]) among patients who switched from 1 antiplatelet agent to another or to DAPT.⁸⁰⁰ However, both of the RCTs that specifically examined recurrent stroke used DAPT rather than switching antiplatelets. In a subgroup analysis of the SPS3 trial, there was no benefit from the addition of clopidogrel to aspirin among patients with ischemic stroke who were already taking aspirin (risk of recurrent stroke, 3.1% versus 3.3%; HR, 0.91 [95% CI, 0.61–1.37]).⁷⁹⁸ In a subgroup of patients already on aspirin in the CHANCE trial, there was a reduction in recurrent stroke among patients with clopidogrel added to aspirin versus aspirin alone (12.3%–9.0%; HR, 0.66 [95% CI, 0.47–0.92]).⁴¹⁰ A prospective registry study found that maintaining aspirin therapy was associated with higher recurrent stroke rates (8.0%) compared with switching to another antiplatelet (6.9%) or adding another antiplatelet (6.6%).⁷⁹⁹

Long-term use of DAPT with aspirin and clopidogrel has been shown in 2 secondary stroke prevention RCTs^381,382 to have no benefit over SAPT for recurrent stroke prevention and to have a significantly increased risk of ICH and major bleeding. The exact duration of DAPT at which the risk of hemorrhage begins to outweigh the benefit of stroke prevention is unknown, but meta-analyses report as early as 21,⁸⁰² 30,⁸⁰⁵ or 90 days.⁸⁰⁶ Older patients⁸⁰⁴ and those with more severe stroke⁸⁰⁶ appear to be at higher risk of ICH with DAPT. Triple antiplatelet therapy with aspirin, clopidogrel, and dipyridamole was compared with standard antiplatelet therapy in the TARDIS trial (Triple Antiplatelets for Reducing Dependency after Ischaemic Stroke), which found no difference in stroke outcomes and a significantly increased risk of bleeding with triple antiplatelet therapy.⁸⁰¹ There may be other non–stroke-related indications for DAPT beyond 90 days after stroke (eg, recent drug-eluting cardiac stent placement), but these indications should be clarified to ensure that DAPT is not continued indefinitely beyond the clinically recommended time frame.

Optimal combination of medications, timing of initiation, and duration of DAPT.

Effectiveness and potential harm of DAPT among specific subgroups of patients according to stroke characteristics, laboratory or genetic tests, or other factors.

Effectiveness/selection of a given antiplatelet agent over another in specific subgroups of patients with noncardioembolic stroke.

Benefit of switching antiplatelet agent for patients already taking 1 antiplatelet medication at the time of stroke.

Effectiveness of DOACs compared with or in combination with antiplatelet therapy for secondary stroke prevention among patients with noncardioembolic ischemic stroke.

Adherence to evidence-based guidelines for secondary stroke prevention is affect by health system–level, hospital-level, provider-level, and patient- and community-specific factors. Regular hospital and outpatient-based programs for adherence and risk factor monitoring provide the opportunity to identify areas of need and to optimize strategies to improve care quality, improve risk factor control, and ultimately reduce the risk of recurrent stroke and TIA. The Get With The Guidelines registry is an example of an inpatient registry used for quality monitoring and improvement.⁸³¹

Interventions that include a multidisciplinary team–based approach to care may be beneficial for risk factor control and secondary prevention after ischemic stroke and TIA. In the ICARUSS study (Integrated Care for the Reduction of Secondary Stroke) conducted in Australia, an integrated shared-care model intervention was associated with a significant decrease in SBP at 12 months compared with usual care.^810,811 Several additional studies, including NAILED Stroke (Nurse Based Age Independent Intervention to Limit Evolution of Disease After Stroke; Sweden) and PREVENTION (Preventing Recurrent Vascular Events in Patients With Stroke or Transient Ischemic Attack; Canada), showed significant benefit for BP lowering when nurses or pharmacists were involved in BP management.^812–814 In the SUSTAIN study (Systemic Use of Stroke Averting Interventions), investigators tested an intervention that included a nurse practitioner/physician assistant care manager, group clinics, self-management support, and care coordination to enhance risk factor control. There was no between-group difference in the primary outcome of mean SBP; however, the intervention group was more likely to achieve LDL goals compared with the usual care group.⁸¹⁵ Additional evidence for the benefit of team-based interventions for secondary prevention of stroke is derived from nonrandomized data. In Ontario, Canada, Stroke Prevention Clinics evolved as an interdisciplinary approach to stroke assessment, prevention, and risk factor management.⁸³² Patients are referred to these clinics from their primary doctors after TIA or after emergency department visit for TIA or stroke. Care in these clinics has been associated with increased use of evidence-based strategies for stroke prevention and decreased risk of mortality at 1 year compared with patients who are not referred to such clinics.⁸¹⁶ In the COMPASS pragmatic trial (United States), there was no benefit of a stroke transitional care model on secondary behavioral outcomes for risk factor control; however, implementation of the model into real-world practice was a major barrier. Additional ongoing and recently completed randomized trials are evaluating the impact of team-based interventions on risk factor control and secondary prevention after ischemic stroke and TIA.^833–838

In the FASTEST study (Efficacy and Safety of a Transient Ischaemic Attack [TIA] Electronic Support Tool) conducted in New Zealand, use of an electronic decision support tool for TIA and stroke management by general practitioners was compared with usual care for patients presenting to their general practitioner as first care contact after stroke or TIA.⁸¹⁸ The tool was designed to improve diagnostic accuracy, to support appropriate triage, and to guide providers in appropriate medication and nonpharmacological (lifestyle counseling) management. TIA and stroke recurrence at 90 days was significantly lower in the intervention group compared with the usual care group, and patients receiving the intervention were more likely to receive guideline-adherent care. This study provides evidence for the use of an electronic decision support tool to guide primary care providers in the management of TIA or minor stroke.

Effective transitions of care from the acute setting to the community affect access to postacute care for secondary prevention. Research on the most effective transitions of care strategies for secondary stroke prevention is needed.

Many patients have multiple risk factors for recurrent stroke. Health systems–based interventions that address simultaneous control of multiple risk factors for secondary stroke prevention are needed.

Patients who are uninsured and underinsured may have limited access to poststroke care. Interventions to improve care access and to mitigate the impact of lack of insurance on poststroke risk factor control and secondary stroke prevention are needed.

Additional studies are needed to evaluate the impact of decision support tools on risk factor control and secondary stroke prevention after stroke hospitalization.

The impact of telemedicine on acute stroke care is well established.⁸³⁹ Although telemedicine offers many potential benefits over in-person care, patients with stroke may face multiple barriers to engaging in telemedicine, including cognitive impairment and physical disability. Research is needed to evaluate the impact of telemedicine on poststroke risk factor control and secondary stroke prevention.

Once proven effective, interventions must be implemented across hospital and health systems in order to achieve desired outcomes. Research that examines the implementation of interventions across health systems is needed.

Longer-term follow-up in behavioral interventions after stroke addressing recurrent events is lacking.^111,112,840 However, meta-analysis (from 4 identified studies; N=4053; I²=0) demonstrates that multimodal interventions addressing active education about risk factors, medications, and medication compliance and interventions to modify ≥1 lifestyle risk factors decrease the odds of recurrent cardiac events (OR, 0.38 [95% CI, 0.16–0.88]), although no significant difference in the odds of recurrent TIA/stroke was reported in this review.⁸⁴⁰ Two small multimodal RCTs show promise for reducing cardiovascular end points. However, the small trial numbers dictate that results must be interpreted with caution pending larger, more definitive trials. A 24-week Japanese RCT (N=70)¹³⁴ comprising exercise, salt restriction, and nutrition advice compared with usual care terminated at a median follow-up of 2.9 years when the prespecified efficacy point for the intervention was reached for the primary end point (composite stroke death, cardiac death, and hospitalization for stroke recurrence, MI, angina pectoris, or peripheral artery disease; (adjusted HR, 0.194 [95% CI, 0.121–0.737]). Long-term effects (3.5 years) of an exercise and education program, compared with usual care in TIA or minor stroke patients (N=60), found that within the intervention group, recurrent stroke, minor stroke, or TIA events were reduced compared with the usual care group (RR, 0.23 [95% CI, 0.07–0.72]; number needed to treat, 3) and a reduction in mortality associated with the intervention (RR, 0.11 [95% CI, 0.01–1.98]; number needed to treat, 8).¹³¹

Adhering to pharmacological secondary prevention strategies is an important self-care behavior in stroke secondary prevention. However, nonadherence to medication regimens is reported in up to 40% of individuals with stroke.⁸⁵³ Multimodal interventions have been shown in a meta-analysis to improve compliance with antithrombotic medications (OR, 1.45 [95% CI, 1.21–1.75]; I²=0%; N=2792) and statins (OR, 2.53 [95% CI, 2.15–2.97]; I²=0%; N=2636).⁸⁴⁰ Self-management interventions have similarly shown a moderate effect size in adherence to prescribed secondary prevention medications after stroke (standardized MD, 0.31 [95% CI, 0.07–0.56]; I²=0%; N=802).⁸⁴¹ In the MIST phase III RCT (Motivational Interviewing in Stroke; N=386), improving adherence to secondary stroke prevention strategies through motivational interviewing demonstrated positive effects on self-reported medication adherence 9 months after stroke (RR, 4.295 [95% CI, 1.56–11.84]; P=0.0049) but no associated reduction in recurrent stroke (RR, 0.67 [95% CI, 0.19–2.33]).⁸⁴² Assistive technology may play an emerging role. The SMS4Stroke (Short Messaging Service for Stroke) trial (N=200) showed an adjusted MD of 0.54 (95% CI, 0.22–0.85) in the Morisky Medication Adherence Scale score after 2 months of receiving short messaging service reminders for each dose of medication.⁸⁴³

Systematic reviews of lifestyle interventions compared with usual care after stroke identify favorable effects on reducing BP.^111,112 A mean SBP reduction of −3.6 mm Hg (95% CI, −5.6 to −1.6; I²=33%, N=650) has been noted.¹¹² Subgroup analysis identified nonsignificant effects for behavior interventions alone, significant effects for cardiovascular exercise interventions (MD, −3.9 mm Hg [95% CI, −6.5 to −1.3]; I²=19; N=70), and larger effects for combined exercise and lifestyle interventions (MD, −5.3 mm Hg [95% CI, −9.0 to −1.6]; I²=46%; N=228). A systematic review of exercise-based interventions on cardiovascular risk factors after stroke¹¹³ reported similar reductions in SBP in addition to other physiological outcomes (fasting insulin and glucose and high-density lipoprotein cholesterol). Mean SBP reductions in exercise interventions with or without other interventions (MD, −5.32 [95% CI, −9.46 to −1.18]) were greater than those observed with exercise interventions alone (MD, −2.51 [95% CI, −4.72 to −0.30]).¹¹³ A review of educational/behavioral interventions (excluding exercise) showed no effect on SBP.⁸²⁹

Cardiac rehabilitation reduces cardiovascular mortality in coronary heart disease.⁸⁵⁴ Emerging evidence supports its broad application in TIA/nondisabling stroke populations. One small RCT (N=22) demonstrated superiority to usual care in risk profiles, perceived physical functioning, and mental health when patients with TIA/nondisabling stroke access existing programs.¹³³ A cohort study (N=80) of participants with TIA attending a 6-month cardiac rehabilitation program demonstrated a significant (26%) increase in the lowest mortality risk category of the Duke treadmill score and significant improvements in fitness, total cholesterol/high density lipoprotein ratios, waist circumference, and BMI.¹³² Two RCTs demonstrated promising longer-term reductions in cardiovascular events. A mixed exercise program combined with education and lifestyle counseling (N=60) demonstrated reduced recurrent stroke or TIA events in the intervention group compared with the usual care control group (RR, 0.23 [95% CI, 0.07–0.72]) and a reduction in mortality (RR, 0.11 [95% CI, 0.01–1.98]) at 3.5 years.¹³¹ A tailored intervention of exercise, salt restriction, and nutrition advice for 24 weeks compared with usual care (N=70) was also associated with a reduction in the primary composite endpoint of cardiovascular outcome (HR, 0.194 [95% CI, 0.121–0.737]) at 2.9 years.¹³⁴

The Cochrane nonpharmacological interventions (based on cardiac rehabilitation) review for preventing vascular events after stroke/TIA⁸⁵⁵ identified 1 pilot RCT of cardiac rehabilitation compared with usual stroke care (N=48), demonstrating promising and statistically significant gains in vascular risk profiles and cardiovascular fitness. A systematic review of lifestyle interventions in stroke identified this study and 2 others broadly matching cardiac rehabilitation in their format and concluded that, although meta-analysis was not possible, all studies were associated with significant improvements in health profiles and management of risk,¹¹¹ with this approach endorsed in the AHA/ASA physical activity and exercise recommendations for stroke survivors.³⁴ A recent prospective cohort study with a matched subgroup comparison (N=609) evaluated a Stroke Recovery Programme with 3 components: physician visits, outpatient therapy, and modified cardiac rehabilitation consisting of group therapy of 4 to 5 patients with stroke directed by a specially trained physical therapist and physical therapy assistant compared with usual rehabilitation. Nonrandomized subgroup analysis comparing Stroke Recovery Programme participants (n=76) and matched pairs of nonparticipants (n=66), in addition to functional gains, identified an age-adjusted HR of 0.11 (SE, 1.07 [95% CI, 0.01–0.90]; P=0.039) for the effect of participant group, suggesting that nonparticipants had a 9 times higher hazard of mortality.⁸⁴⁴

Although acknowledging that educating patients with stroke or TIA about their condition and the causes of their stroke is an important aspect of stroke care, evidence does not support its role in modifying health risk behaviors. The Cochrane review of interventions for improving modifiable risk factor control in the secondary prevention of stroke identified that patient education alone does not lead to improvements in modifiable risk factor control or the prevention of recurrent cardiovascular events.⁸²⁹ The recent STANDFIRM trial (Shared Team Approach Between Nurses and Doctors for Improved Risk Factor Management for Stroke Patients; N=563) of specialist review and nurse education found no effect on attainment of cardiometabolic targets for secondary prevention.⁸⁴⁵ Similarly, the ExStroke trial (N=314 ambulatory participants with stroke) identified that repeated encouragement with verbal instruction on being physically active over a 2-year period after stroke was not an effective means of increasing physical activity participation.¹²⁹

The optimal behavior change intervention to reduce recurrent stroke is unknown at present, as is the optimal window to deliver interventions of this nature.

Multiple systematic reviews of lifestyle-based behavioral interventions identify an existing knowledge gap relating to the potential of the interventions reported to date to affect outcomes of recurrent stroke and other cardiovascular end points in the longer term, beyond the intervention period.^111,112,840

Many behavioral interventions in published trials after stroke fail to identify whether a behavior change theory underpinned the intervention delivered, making it difficult to ascertain successful and unsuccessful components of complex behavior change interventions.⁸⁴⁹

A significant knowledge gap exists with respect to the specific or allied role of family caregivers and dyads in stroke secondary prevention in both the promotion and maintenance of health behaviors after stroke.

A clearly identified behavior change theory as a basis for intervention

A documented time point after stroke for targeted intervention delivery

A detailed intervention description that identifies the frequency and duration of the intervention, the personnel required to deliver the intervention, the intervention setting (eg, subacute care, community care, mixed), and the skill level and training in behavioral counseling of the interventionist

Data collection that permits disaggregation by sex and other relevant characteristics to identify responders/nonresponders to intervention.

Socioeconomic inequalities are strong predictors of cardiovascular risk.⁸⁷³ Interventions should be tailored to patients’ socioeconomic and educational status, as well as cultural, work, and home environments.^19,875 Examples of upstream social determinants of health that affect treatment adherence and ASCVD health outcomes include comorbid mental illness, lack of health literacy, exposure to adversity (eg, home/community violence, trauma exposures, safety concerns), financial strain, inadequate housing conditions, lack of food security (ie, access to affordable and nutritious food), structural and individual discrimination, and inadequate social support.⁸⁷⁶ The Centers for Medicare & Medicaid Services has developed a screening tool to assess 5 domains of non–health-related measures that affect health outcomes: housing instability, food insecurity, transportation difficulties, utility assistance needs, and interpersonal safety.⁸⁷⁷

In 2010, the National Institute of Neurological Disease and Stroke convened an expert panel to provide recommendations on key priorities for health disparities research in stroke. One recommendation was to use population-based surveillance to assess achievement of evidence-based guidelines.⁸⁷⁸ Alliances with the federal government through the National Institute of Neurological Disease and Stroke, Centers for Disease Control and Prevention, nonprofit organizations such as the AHA/ASA, and medical specialty groups such as the American Academy of Neurology and the Brain Attack Coalition are needed to coordinate, develop, and optimize implementation of evidence-based stroke prevention recommendations.⁹

The AHRQ Universal Precautions Toolkit for Health Literacy⁸⁷⁹ is a publicly available document that includes a 21-step approach to ensuring that written and oral instructions to patients are clear, understandable, and sensitive to health literacy. Healthcare providers who listen, speak slowly, use nonmedical language, encourage questions, apply teach-back methods, address language differences, are culturally competent, and incorporate graphics in their teaching promote a culture sensitive to health literacy.⁸⁸⁰

Several RCTs have tested secondary stroke prevention interventions in predominantly Black, Hispanic, and/or low-socioeconomic-status populations.^{815,856–859} In all trials, primary outcomes improved in the intervention and control groups without significant differences between the 2 arms. A few of the trials showed improvement in secondary outcomes or subpopulations. The SUSTAIN trial (Systemic Use of Stroke Averting Interventions) of a culturally tailored Chronic Care Model–based intervention (incorporating advance practice provider delivery of evidence-based care, protocols, real-time decision support, self-management support, and education) in a multiracial/multiethnic population in a safety-net setting showed better LDL control (secondary outcome) in the intervention arm; intervention participants were more likely to lower LDL <100 md/dL compared with control subjects (OR, 2.0 [95% CI, 1.1–3.5]).⁸¹⁵ In the PRAISE trial (Prevent Recurrence of All Inner-City Strokes Through Education) of peer-led community-based poststroke self-management workshops in the inner city, the primary composite outcome of BP, LDL, and antithrombotic use at 6 months was not different in the intervention and control arms, but a higher proportion of those randomized to the intervention had controlled BP at 6 months (76% versus 67%; P=0.02).⁸⁵⁸ Finally, a subgroup analysis of Hispanic individuals enrolled in the DESERVE trial (Discharge Educational Strategies for Reduction of Vascular Events), evaluating the efficacy of a culturally tailored skills-based intervention, revealed a 9.9–mm Hg greater reduction in SBP compared with control.⁸⁵⁷

A more thorough understanding of which populations have inequities in risk factor control after stroke.

A more nuanced understanding of the drivers of inequities in risk factor control after stroke, including social determinants of health and structural racism.

Strategies for improving risk factor control among groups at risk for inequities.

Strategies for addressing social determinants of health among stroke survivors.