Imaging

A 35-year-old man presented to the emergency department because of night sweats, fever, chills, and shortness of breath. He also had an acute onset of blue discoloration of his right fourth finger. His symptoms (except for the finger discoloration) had begun about 6 months previously and had rapidly progressed despite several courses of different antibiotics of different types, given both intravenously in the hospital and orally at home. He had lost 20 lb during this time. Previously, he had been healthy.

About 1 month after his symptoms began, he had consulted his primary care physician, who detected a new grade 4/6 systolic and diastolic murmur. Transthoracic echocardiography about 2 months after that demonstrated mild aortic and mitral insufficiency but no echocardiographic features supporting infective endocarditis. Of note, the patient had no risk factors for endocarditis such as illicit drug use or poor dental health.

In the emergency department, his temperature was 99.4°F (37.4°C), pulse 109 beats per minute, and blood pressure 126/60 mm Hg. He had a grade 3/6 harsh holosystolic murmur best heard at the right upper sternal border, a grade 3/4 holodiastolic murmur audible across the precordium, and a grade 3/4 holosystolic blowing murmur best heard at the cardiac apex. Other findings included signs of aortic insufficiency—the Duroziez sign (a diastolic murmur heard over the femoral artery when compressed), Watson’s water-hammer pulse (indicating a wide pulse pressure), and the Müller sign (pulsation of the uvula)—and small Janeway lesions on the inner aspect of his right arm and palm.

Electrocardiography showed normal sinus rhythm, PR interval 128 ms, QRS complex 100 ms, QT interval 360 ms, and corrected QT interval 473 ms.

Blood cultures grew Streptococcus sanguinis. Both transthoracic and transesophageal echocardiography were done promptly and revealed multiple mobile echodensities attached to a trileaflet aortic valve, consistent with vegetations and valve leaflet destruction; severe (4+) aortic regurgitation with flow reversal in the abdominal aorta; mild mitral regurgitation; and a mitral valve aneurysm with mild mitral regurgitation (Figure 1).

INFECTIVE ENDOCARDITIS: WORTH CONSIDERING

S sanguinis is a member of the group of viridans streptococci. As a normal inhabitant of the healthy human mouth, it is found in dental plaque. It may enter the bloodstream during dental cleaning and may colonize the heart valves, particularly the mitral and aortic valves, where it is the most common cause of subacute bacterial endocarditis.

Infective endocarditis is often diagnosed clinically with the Duke criteria (www.med-calc.com/endocarditis.html).¹ However, the variability of the clinical presentation and the nonspecific nature of the initial workup often create a diagnostic challenge for the evaluating physician.^1,2

In cases of recurrent persistent fever and a new heart murmur, infective endocarditis must always be considered. Blood cultures should be ordered early and repeatedly. If blood cultures are positive, transesophageal echocardiography should be done without delay if transthoracic echocardiography was unremarkable. Prompt diagnosis and surgical intervention prevent complications.

MITRAL VALVE ANEURYSM IN AORTIC VALVE ENDOCARDITIS

Aortic valve endocarditis often also involves the mitral valve; mitral valve endocarditis is seen in 17% of patients undergoing surgery for aortic valve endocarditis.³ Proposed mechanisms for this association include jet lesions from aortic regurgitation, vegetation prolapse with direct contact between the aortic valve and anterior mitral leaflet (“kissing lesions”), and direct local spread of infection.^4–7

One of every five patients with mitral valve involvement in aortic valve endocarditis has a mitral valve aneurysm.³ This is a serious finding, as it can lead to septic embolization. Also, the weakened lining of the mitral valve aneurysm can rupture, resulting in severe mitral regurgitation, acute pulmonary edema, and precipitous cardiopulmonary decompensation.⁵

Transesophageal echocardiography is more sensitive than transthoracic echocardiography for detecting mitral valve aneurysm.⁸ On two-dimensional echocardiography, the lesion appears as a narrow-necked, saccular echolucency with systolic protrusion into the left atrium. Color Doppler imaging often shows turbulent, high-velocity flow.

Differential diagnosis of mitral valve aneurysm

Differential diagnostic considerations include a valvular blood cyst, a congenital cardiac diverticulum, and mitral valve prolapse.

Valvular blood cysts are extremely rare in adults.⁹ These benign, congenital tumors are most often found on the atrioventricular valves in infants, in whom the reported incidence is between 25% and 100%. In almost all cases, these cysts are believed to regress spontaneously with time.

In almost all reported cases, the cyst involved the valvular apparatus or papillary muscle of the tricuspid, pulmonary, or mitral valve.¹⁰ Cysts consist of a benign diverticulum lined with flattened, cobblestone-shaped endothelium and are filled with blood. They can cause heart murmurs in otherwise asymptomatic patients.

On echocardiography, a blood cyst appears as an oval mass (often at the interatrial septum), often with normal cardiac function. In the rare case in which a blood cyst is found incidentally during echocardiography, the hemodynamic impact, if any, should be determined by Doppler techniques.

When benign, a valvular blood cyst can be safely monitored with echocardiographic follow-up.¹¹ Treatment involves surgical resection of the mass in symptomatic patients in whom cardiac function is impaired by the presence of the cyst.

Congenital cardiac diverticuli are extremely rare, most often seen in children, and associated with a midline thoracoabdominal defect. Echocardiography can differentiate a ventricular diverticulum from an aneurysm or a pseudoaneurysm.

A ventricular diverticulum has a fibrous, narrow neck connecting with the ventricle, and a small circular echo-free space that communicates with the ventricle via this narrow neck.² Doppler imaging shows systolic flow from the diverticulum to the ventricle, and systolic contractility may also be seen during cardiac catheterization. Congenital diverticulum is typically confused with ventricular aneurysm and, to a lesser degree, with mitral valve aneurysm.

Mitral valve prolapse is characterized by interchordal ballooning or hooding of the mitral valve leaflets that occurs when one or both floppy, enlarged leaflets prolapse into the left atrium during systole.

BACK TO OUR PATIENT

The patient underwent open heart surgery, with successful repair of the aortic root, replacement of the aortic valve, and repair of the mitral valve. An abscess was found within the aneurysmal cavity.

A 35-year-old man presented to the emergency department because of night sweats, fever, chills, and shortness of breath. He also had an acute onset of blue discoloration of his right fourth finger. His symptoms (except for the finger discoloration) had begun about 6 months previously and had rapidly progressed despite several courses of different antibiotics of different types, given both intravenously in the hospital and orally at home. He had lost 20 lb during this time. Previously, he had been healthy.

About 1 month after his symptoms began, he had consulted his primary care physician, who detected a new grade 4/6 systolic and diastolic murmur. Transthoracic echocardiography about 2 months after that demonstrated mild aortic and mitral insufficiency but no echocardiographic features supporting infective endocarditis. Of note, the patient had no risk factors for endocarditis such as illicit drug use or poor dental health.

In the emergency department, his temperature was 99.4°F (37.4°C), pulse 109 beats per minute, and blood pressure 126/60 mm Hg. He had a grade 3/6 harsh holosystolic murmur best heard at the right upper sternal border, a grade 3/4 holodiastolic murmur audible across the precordium, and a grade 3/4 holosystolic blowing murmur best heard at the cardiac apex. Other findings included signs of aortic insufficiency—the Duroziez sign (a diastolic murmur heard over the femoral artery when compressed), Watson’s water-hammer pulse (indicating a wide pulse pressure), and the Müller sign (pulsation of the uvula)—and small Janeway lesions on the inner aspect of his right arm and palm.

Electrocardiography showed normal sinus rhythm, PR interval 128 ms, QRS complex 100 ms, QT interval 360 ms, and corrected QT interval 473 ms.

Blood cultures grew Streptococcus sanguinis. Both transthoracic and transesophageal echocardiography were done promptly and revealed multiple mobile echodensities attached to a trileaflet aortic valve, consistent with vegetations and valve leaflet destruction; severe (4+) aortic regurgitation with flow reversal in the abdominal aorta; mild mitral regurgitation; and a mitral valve aneurysm with mild mitral regurgitation (Figure 1).

INFECTIVE ENDOCARDITIS: WORTH CONSIDERING

S sanguinis is a member of the group of viridans streptococci. As a normal inhabitant of the healthy human mouth, it is found in dental plaque. It may enter the bloodstream during dental cleaning and may colonize the heart valves, particularly the mitral and aortic valves, where it is the most common cause of subacute bacterial endocarditis.

Infective endocarditis is often diagnosed clinically with the Duke criteria (www.med-calc.com/endocarditis.html).¹ However, the variability of the clinical presentation and the nonspecific nature of the initial workup often create a diagnostic challenge for the evaluating physician.^1,2

In cases of recurrent persistent fever and a new heart murmur, infective endocarditis must always be considered. Blood cultures should be ordered early and repeatedly. If blood cultures are positive, transesophageal echocardiography should be done without delay if transthoracic echocardiography was unremarkable. Prompt diagnosis and surgical intervention prevent complications.

MITRAL VALVE ANEURYSM IN AORTIC VALVE ENDOCARDITIS

Aortic valve endocarditis often also involves the mitral valve; mitral valve endocarditis is seen in 17% of patients undergoing surgery for aortic valve endocarditis.³ Proposed mechanisms for this association include jet lesions from aortic regurgitation, vegetation prolapse with direct contact between the aortic valve and anterior mitral leaflet (“kissing lesions”), and direct local spread of infection.^4–7

One of every five patients with mitral valve involvement in aortic valve endocarditis has a mitral valve aneurysm.³ This is a serious finding, as it can lead to septic embolization. Also, the weakened lining of the mitral valve aneurysm can rupture, resulting in severe mitral regurgitation, acute pulmonary edema, and precipitous cardiopulmonary decompensation.⁵

Transesophageal echocardiography is more sensitive than transthoracic echocardiography for detecting mitral valve aneurysm.⁸ On two-dimensional echocardiography, the lesion appears as a narrow-necked, saccular echolucency with systolic protrusion into the left atrium. Color Doppler imaging often shows turbulent, high-velocity flow.

Differential diagnosis of mitral valve aneurysm

Differential diagnostic considerations include a valvular blood cyst, a congenital cardiac diverticulum, and mitral valve prolapse.

Valvular blood cysts are extremely rare in adults.⁹ These benign, congenital tumors are most often found on the atrioventricular valves in infants, in whom the reported incidence is between 25% and 100%. In almost all cases, these cysts are believed to regress spontaneously with time.

In almost all reported cases, the cyst involved the valvular apparatus or papillary muscle of the tricuspid, pulmonary, or mitral valve.¹⁰ Cysts consist of a benign diverticulum lined with flattened, cobblestone-shaped endothelium and are filled with blood. They can cause heart murmurs in otherwise asymptomatic patients.

On echocardiography, a blood cyst appears as an oval mass (often at the interatrial septum), often with normal cardiac function. In the rare case in which a blood cyst is found incidentally during echocardiography, the hemodynamic impact, if any, should be determined by Doppler techniques.

When benign, a valvular blood cyst can be safely monitored with echocardiographic follow-up.¹¹ Treatment involves surgical resection of the mass in symptomatic patients in whom cardiac function is impaired by the presence of the cyst.

Congenital cardiac diverticuli are extremely rare, most often seen in children, and associated with a midline thoracoabdominal defect. Echocardiography can differentiate a ventricular diverticulum from an aneurysm or a pseudoaneurysm.

A ventricular diverticulum has a fibrous, narrow neck connecting with the ventricle, and a small circular echo-free space that communicates with the ventricle via this narrow neck.² Doppler imaging shows systolic flow from the diverticulum to the ventricle, and systolic contractility may also be seen during cardiac catheterization. Congenital diverticulum is typically confused with ventricular aneurysm and, to a lesser degree, with mitral valve aneurysm.

Mitral valve prolapse is characterized by interchordal ballooning or hooding of the mitral valve leaflets that occurs when one or both floppy, enlarged leaflets prolapse into the left atrium during systole.

BACK TO OUR PATIENT

The patient underwent open heart surgery, with successful repair of the aortic root, replacement of the aortic valve, and repair of the mitral valve. An abscess was found within the aneurysmal cavity.

A 35-year-old man presented to the emergency department because of night sweats, fever, chills, and shortness of breath. He also had an acute onset of blue discoloration of his right fourth finger. His symptoms (except for the finger discoloration) had begun about 6 months previously and had rapidly progressed despite several courses of different antibiotics of different types, given both intravenously in the hospital and orally at home. He had lost 20 lb during this time. Previously, he had been healthy.

About 1 month after his symptoms began, he had consulted his primary care physician, who detected a new grade 4/6 systolic and diastolic murmur. Transthoracic echocardiography about 2 months after that demonstrated mild aortic and mitral insufficiency but no echocardiographic features supporting infective endocarditis. Of note, the patient had no risk factors for endocarditis such as illicit drug use or poor dental health.

In the emergency department, his temperature was 99.4°F (37.4°C), pulse 109 beats per minute, and blood pressure 126/60 mm Hg. He had a grade 3/6 harsh holosystolic murmur best heard at the right upper sternal border, a grade 3/4 holodiastolic murmur audible across the precordium, and a grade 3/4 holosystolic blowing murmur best heard at the cardiac apex. Other findings included signs of aortic insufficiency—the Duroziez sign (a diastolic murmur heard over the femoral artery when compressed), Watson’s water-hammer pulse (indicating a wide pulse pressure), and the Müller sign (pulsation of the uvula)—and small Janeway lesions on the inner aspect of his right arm and palm.

Electrocardiography showed normal sinus rhythm, PR interval 128 ms, QRS complex 100 ms, QT interval 360 ms, and corrected QT interval 473 ms.

Blood cultures grew Streptococcus sanguinis. Both transthoracic and transesophageal echocardiography were done promptly and revealed multiple mobile echodensities attached to a trileaflet aortic valve, consistent with vegetations and valve leaflet destruction; severe (4+) aortic regurgitation with flow reversal in the abdominal aorta; mild mitral regurgitation; and a mitral valve aneurysm with mild mitral regurgitation (Figure 1).

INFECTIVE ENDOCARDITIS: WORTH CONSIDERING

S sanguinis is a member of the group of viridans streptococci. As a normal inhabitant of the healthy human mouth, it is found in dental plaque. It may enter the bloodstream during dental cleaning and may colonize the heart valves, particularly the mitral and aortic valves, where it is the most common cause of subacute bacterial endocarditis.

Infective endocarditis is often diagnosed clinically with the Duke criteria (www.med-calc.com/endocarditis.html).¹ However, the variability of the clinical presentation and the nonspecific nature of the initial workup often create a diagnostic challenge for the evaluating physician.^1,2

In cases of recurrent persistent fever and a new heart murmur, infective endocarditis must always be considered. Blood cultures should be ordered early and repeatedly. If blood cultures are positive, transesophageal echocardiography should be done without delay if transthoracic echocardiography was unremarkable. Prompt diagnosis and surgical intervention prevent complications.

MITRAL VALVE ANEURYSM IN AORTIC VALVE ENDOCARDITIS

Aortic valve endocarditis often also involves the mitral valve; mitral valve endocarditis is seen in 17% of patients undergoing surgery for aortic valve endocarditis.³ Proposed mechanisms for this association include jet lesions from aortic regurgitation, vegetation prolapse with direct contact between the aortic valve and anterior mitral leaflet (“kissing lesions”), and direct local spread of infection.^4–7

One of every five patients with mitral valve involvement in aortic valve endocarditis has a mitral valve aneurysm.³ This is a serious finding, as it can lead to septic embolization. Also, the weakened lining of the mitral valve aneurysm can rupture, resulting in severe mitral regurgitation, acute pulmonary edema, and precipitous cardiopulmonary decompensation.⁵

Transesophageal echocardiography is more sensitive than transthoracic echocardiography for detecting mitral valve aneurysm.⁸ On two-dimensional echocardiography, the lesion appears as a narrow-necked, saccular echolucency with systolic protrusion into the left atrium. Color Doppler imaging often shows turbulent, high-velocity flow.

Differential diagnosis of mitral valve aneurysm

Differential diagnostic considerations include a valvular blood cyst, a congenital cardiac diverticulum, and mitral valve prolapse.

Valvular blood cysts are extremely rare in adults.⁹ These benign, congenital tumors are most often found on the atrioventricular valves in infants, in whom the reported incidence is between 25% and 100%. In almost all cases, these cysts are believed to regress spontaneously with time.

In almost all reported cases, the cyst involved the valvular apparatus or papillary muscle of the tricuspid, pulmonary, or mitral valve.¹⁰ Cysts consist of a benign diverticulum lined with flattened, cobblestone-shaped endothelium and are filled with blood. They can cause heart murmurs in otherwise asymptomatic patients.

On echocardiography, a blood cyst appears as an oval mass (often at the interatrial septum), often with normal cardiac function. In the rare case in which a blood cyst is found incidentally during echocardiography, the hemodynamic impact, if any, should be determined by Doppler techniques.

When benign, a valvular blood cyst can be safely monitored with echocardiographic follow-up.¹¹ Treatment involves surgical resection of the mass in symptomatic patients in whom cardiac function is impaired by the presence of the cyst.

Congenital cardiac diverticuli are extremely rare, most often seen in children, and associated with a midline thoracoabdominal defect. Echocardiography can differentiate a ventricular diverticulum from an aneurysm or a pseudoaneurysm.

A ventricular diverticulum has a fibrous, narrow neck connecting with the ventricle, and a small circular echo-free space that communicates with the ventricle via this narrow neck.² Doppler imaging shows systolic flow from the diverticulum to the ventricle, and systolic contractility may also be seen during cardiac catheterization. Congenital diverticulum is typically confused with ventricular aneurysm and, to a lesser degree, with mitral valve aneurysm.

Mitral valve prolapse is characterized by interchordal ballooning or hooding of the mitral valve leaflets that occurs when one or both floppy, enlarged leaflets prolapse into the left atrium during systole.

BACK TO OUR PATIENT

The patient underwent open heart surgery, with successful repair of the aortic root, replacement of the aortic valve, and repair of the mitral valve. An abscess was found within the aneurysmal cavity.

A transient ischemic attack (TIA), like an episode of unstable angina, is an ominous portent of future morbidity and death even though, by definition, the event leaves no residual neurologic deficit.

But there is a positive side. When a patient presents with a TIA, the physician has the rare opportunity to reduce the risk of a disabling outcome—in this case, stroke. Therefore, patients deserve a rapid and thorough evaluation and appropriate stroke-preventive treatment.

MANY ‘TIAs’ ARE ACTUALLY STROKES

TIA has traditionally been described as a sudden focal neurologic deficit that lasts less than 24 hours, is presumed to be of vascular origin, and is confined to an area of the brain, spinal cord, or eye perfused by a specific artery. This symptom-based definition was based on the arbitrary and inaccurate assumption that brief symptoms would not be associated with damage to brain parenchyma.

The definition has since been updated and made more rational based on new concepts of brain ischemia informed by imaging, especially diffusion-weighted magnetic resonance imaging (MRI).¹ One-third of episodes characterized as a TIA according to the classic definition would be considered an infarction on the basis of diffusion-weighted MRI.² The new tissue-based definition characterizes TIA as a brief episode of neurologic dysfunction caused by focal ischemia of the brain, spinal cord, or retina, with clinical symptoms lasting less than 24 hours and without evidence of acute infarction.³

AN OPPORTUNITY TO INTERVENE

Most TIAs resolve in less than 30 minutes. The US National Institute of Neurological Disorders and Stroke trial of tissue plasminogen activator found that if symptoms of cerebral ischemia had not resolved by 1 hour or had not rapidly improved within 3 hours, complete resolution was rare (only 2% at 24 hours).⁴ Hence, physicians evaluating and treating patients with TIAs should treat these episodes with the urgency they deserve.

Moreover, half of the strokes that follow TIAs occur within 48 hours.⁵ A rapid and thorough evaluation and the initiation of secondary preventive treatments have been shown to reduce the early occurrence of stroke by up to 80%.⁶ Hence, the correct diagnosis of TIA gives the clinician the best opportunity to prevent stroke and its personal, social, and sometimes fatal consequences.

STROKES OUTNUMBER TIAs, BUT TIAs ARE UNDERREPORTED

According to 2012 statistics, nearly 795,000 strokes occur in the United States each year, 610,000 of which are first attacks and 185,000 are recurrences. Every 40 seconds, someone in the United States has a stroke.⁷

In comparison, the incidence of TIA in the United States is estimated at 200,000 to 500,000 per year, though the true number is difficult to know because of underreporting.^8,9 About half of patients who experience a TIA fail to report it to their health care provider—a lost opportunity for intervention and stroke prevention.^10,11

A meta-analysis showed that the risk of stroke after TIA was 9.9% at 2 days, 13.4% at 30 days, and 17.3% at 90 days.¹²

Interestingly, the risk of stroke after TIA exceeds the risk of recurrent stroke after a first stroke. This was shown in a study that found that patients who had made a substantial recovery within 24 hours (ie, patients with a TIA) were more likely to suffer neurologic deterioration in the next 3 months than were those who did not have significant early improvement.¹³

RISK FACTORS FOR TIA ARE THE SAME AS FOR STROKE

The risk of cerebrovascular disease increases with age and is higher in men¹⁴ and in blacks and Hispanics.¹⁵

The risk factors and clinical presentation do not differ between TIA and stroke, so the evaluation and treatment should not differ either. These two events represent a continuum of the same disease entity.

Some risk factors for TIA are modifiable, others are not.

Nonmodifiable risk factors

Nonmodifiable risk factors for TIA include older age, male sex, African American race, low birth weight, Hispanic ethnicity, and family history. If the patient has nonmodifiable risk factors, we should try all the harder to correct the modifiable ones.

Older age. The risk of ischemic stroke and intracranial hemorrhage doubles with each decade after age 55 in both sexes.¹⁶

Sex. Men have a significantly higher incidence of TIA than women,¹¹ whereas the opposite is true for stroke: women have a higher lifetime risk of stroke than men.¹⁷

African Americans have an incidence of stroke (all types) 38% higher than that of whites,¹⁸ and an incidence of TIA (inpatient and out-of-hospital) 40% higher than the overall age- and sex-adjusted rate in the white population.¹¹

Low birth weight. The odds of stroke are more than twice as high in people who weighed less than 2,500 g at birth compared with those who weighed 4,000 g or more, probably because of a correlation between low birth weight and hypertension.¹⁹

A family history of stroke increases the risk of stroke by nearly 30%, the association being stronger with large-vessel and smallvessel stroke than with cardioembolic stroke.²⁰

Modifiable risk factors

Modifiable risk factors include cigarette smoking, hypertension, diabetes, lipid abnormalities, atrial fibrillation, carotid stenosis, and dietary and hormonal factors. Detecting these factors, which often coexist, is the first step in trying to modify them and reduce the patient’s risk.

Cigarette smoking approximately doubles the risk of ischemic stroke.^21–23

Hypertension has a relationship with stroke risk that is strong, continuous, graded, consistent, and significant.²⁴

Diabetes increases stroke risk nearly six times.²⁵

Lipid abnormalities. Most studies have found an association between lipid levels (total cholesterol and low-density lipoprotein cholesterol) and the risk of death from ischemic stroke,^26–28 and an inverse relationship between high-density lipoprotein cholesterol levels and stroke risk.²⁹

Atrial fibrillation increases the risk of ischemic stroke up to fivefold, even in the absence of cardiac valvular disease. The mechanism is embolism of stasis-induced thrombi that form in the left atrial appendage.³⁰

Carotid stenosis. Asymptomatic carotid atherosclerotic stenotic lesions in the extracranial internal carotid artery or carotid bulb are associated with a higher risk of stroke.^24,31

Lifestyle factors. Diets that lower blood pressure have been found to decrease stroke risk.²⁴ Exercise in men and women reduces the risk of stroke or death by 25% to 30% compared with inactive people.³² Weight reduction has been found to lower blood pressure and reduce stroke risk.²⁴

Other potentially modifiable risk factors include migraine with aura, metabolic syndrome, excess alcohol consumption (and, paradoxically, complete abstinence from alcohol), drug abuse, sleep-disordered breathing, hyperhomocysteinemia, high lipoprotein (a) levels, hypercoagulability, infection with organisms such as Chlamydia pneumoniae, cytomegalovirus, and Helicobacter pylori, and acute infections such as respiratory and urinary infections.²⁶

Conditions in certain demographic groups

Patients in certain demographic groups present with rarer conditions associated with stroke and TIA.

Sickle cell disease. Eleven percent of patients with sickle cell disease have clinical strokes, and a substantial number have “silent” infarcts identified on neuroimaging.^33,34

Postmenopausal hormone replacement therapy with any product containing conjugated equine estrogen carries a risk of cerebrovascular events,³⁵ and the higher the dose, the higher the risk.³⁶ Also, oral contraceptives may be harmful in women who have additional risk factors such as cigarette smoking, prior thromboembolic events, or migraine with aura.^37,38

THREE CAUSES OF STROKE AND TIA

Stroke and TIA should not be considered diagnoses in themselves, but rather the end point of many other diseases. The diagnosis lies in identifying the mechanism of the cerebrovascular event. The three main mechanisms are thrombosis, embolism, and decreased perfusion.

Thrombosis is caused by obstruction of blood flow within one or more blood vessels, the most common cause being atherosclerosis. Large-artery atherosclerosis, such as in the carotid bifurcation or extracranial internal carotid, causes TIAs that occur over a period of weeks or months with a variety of presentations in that vascular territory, from years of gradual accumulation of atherosclerotic plaque.³⁹

In patients with small-artery or penetrating artery disease, hypertension is the primary risk factor and the pathology, specific to small arterioles, is lipohyalinosis rather than atherosclerosis. These patients may present with a stuttering clinical course, and episodes are more stereotypical.

Less common obstructive vascular pathologies include fibromuscular dysplasia, arteritides, and dissection.

Embolism can occur from a proximal source such as the heart or from proximal vessels such as the aorta, carotid, or vertebral arteries. The embolic particle may form on heart valves or lesions within the heart (eg, clot, tumor), or in the venous circulation and paradoxically cross over to the arterial side through an intracardiac or transpulmonary shunt. Embolism may also be due to a hypercoagulable state.⁴⁰ Embolic stroke is suspected when multiple vascular territories within the brain are clinically or radiographically affected.

Decreased systemic perfusion caused by severe heart failure or systemic hypotension can cause ischemia to the brain diffusely and bilaterally, limiting the ability of the blood-stream to wash out microemboli, especially in the border zones (also known as “watershed areas”), thus leading to ischemia or infarction.⁴¹ Decreased perfusion can also be local, due to a fixed vessel stenosis.

Using another classification system, a study in Rochester, MN, found the following incident rates of stroke subtypes, adjusted for age and sex, per 100,000 population⁴²:

• Large-vessel cervical or intracranial atherosclerosis with more than 50% stenosis—27
• Cardioembolism—40
• Lacunar, small-vessel disease—25
• Uncertain cause—52
• Other identifiable cause—4.

THREE CLINICAL FEATURES SUGGEST TIA

TIAs can be hard to distinguish from nonischemic neurologic events in the acute setting such as an emergency room. Up to 60% of patients suspected of having a TIA actually have a nonischemic cause of their symptoms.⁴³

Three clinical features suggest a TIA during the emergency room evaluation:

• Rapid onset of symptoms—“like lightning” or “in seconds,” in contrast to migraine and seizures, which develop over minutes
• No history of similar episodes in the past
• Absence of nonspecific symptoms—eg, stomach upset or tightness in the chest.

CLINICAL DIAGNOSIS

Because most TIA symptoms and signs have already resolved by the time of evaluation, the diagnosis depends on a careful history with special attention to the pace of onset and resolution, the duration and nature of the symptoms, circumstances at the time of symptom onset, previous similar episodes, associated features, vascular risk factors, and family history (Table 1).^44,45

A detailed neurologic examination is imperative and should include fundoscopy. A cardiovascular assessment should include cardiac rhythm, bruits in the neck, orbits, and groin, peripheral pulses, and electrocardiography.

Do neurologists do a better job at diagnosing TIA and stroke?

Primary care physicians, internists, and emergency department physicians are often the ones to carry out the clinical assessment of possible TIA.

Determining if transient neurologic symptoms are caused by ischemia can be a challenge. When in doubt, referral to a neurologist with subspecialty training in cerebrovascular disease should be considered.

But do neurologists really do a better job? A recent study sought to compare the accuracy of diagnosis of TIA made by general practitioners, emergency physicians, and neurologists. The nonneurologists considered “confusion” and “unexplained fall” suggestive of TIA and “lower facial palsy” and “monocular blindness” less suggestive of TIA—whereas the opposite is true. This shows that nonneurologists often label minor strokes and several nonvascular transient neurologic disturbances as TIAs, and up to half of patients could be mislabeled as a result.⁴⁶

Differences in diagnosing cerebrovascular events between emergency room physicians and attending neurologists have been tested,⁴⁷ with an accuracy of diagnosis as low as 38% by emergency department physicians in one study.⁴⁸ However, other studies did not show such a trend.^49,50

A study at a university-based teaching hospital found the sensitivity of emergency room physician diagnosis to be 98.6% with a positive predictive value of 94.8%,⁴⁹ showing that at a large teaching hospital with a comprehensive stroke intervention program, emergency physicians could identify patients with stroke, particularly hemorrhagic stroke, very accurately.

Improving the diagnosis of stroke and TIA

Routine use of imaging and involvement of a neurologist increase the sensitivity and accuracy of diagnosis. Education and written guidelines for acute stroke treatment both in the emergency department and in out-of-hospital settings seem to dramatically improve the rates of diagnostic accuracy and appropriate treatment.⁵⁰

Emergency medical service personnel use two screening tools in the field to identify TIA and stroke symptoms:

• The Cincinnati Prehospital Stroke Scale, a three-item scale based on three signs: facial droop, arm drift, and slurring of speech⁵¹
• The Los Angeles Prehospital Stroke Screen, which uses screening questions and asymmetry in the face, hand grip strength, and arm drift.⁵²

Knowing that the patient is having a minor stroke or TIA is important. Urgent treatment of these conditions decreases the risk of stroke in the next 90 days, which was 10.5% in one study.⁵ Urgent assessment and early intervention could reduce this risk of subsequent stroke down to 2%.⁶

ASSESSING RISK OF STROKE AFTER TIA

There is a practical need for prediction of stroke during the first few days after the event. The ABCD and ABCD2 scores were developed to stratify the short-term risk of stroke in patients with recent TIA.

The ABCD score

The ABCD score⁵³ was derived to allow primary care physicians and other physicians to identify which patients with a suspected diagnosis of TIA should be referred for emergency assessment, to allow secondary-care physicians to determine which patients with probable or definite TIA need emergency investigation and treatment, to allow public education about the need for medical attention after a TIA, and to identify people at high risk.

The ABCD2 score

The ABCD2 score predicts the short-term risk of stroke following a TIA.⁵⁴ Points are assigned as follows:

• Age > 60 years: 1 point
• Blood pressure (systolic) > 140 mm Hg or diastolic blood pressure > 90 mm Hg: 1 point
• Clinical factors: unilateral weakness with or without speech impairment: 2 points (1 point for speech impairment without weakness)
• Duration of symptoms > 60 minutes: 2 points (1 point for 10–59 minutes)
• Diabetes: 1 point.

Thus, the possible total ranges from 0 to 7 points. Higher scores indicate a greater risk of stroke at 2, 7, 30, and 90 days:

• Total score 0, 1, 2, or 3: 2-day stroke risk 1.0% (low risk)
• Total score 4 or 5: 2-day stroke risk 4.1% (moderate risk)
• Total score 6 or 7: 2-day stroke risk 8.1% (high risk).

WHO SHOULD BE HOSPITALIZED?

It has been suggested that the ABCD2 score can help in triaging patients to hospital admission or outpatient care, though no randomized trial has actually evaluated the utility of the ABCD2 score in this way.³

A study of consecutive TIA patients admitted over 12 months⁵⁵ found that patients with an ABCD2 score of 3 or less had the same chance of requiring hospitalization (based on positive diffusion-weighted MRI studies, risk factor identification, and treatment initiation) as those with a score of 4 to 7. Hence, admitting TIA patients on the basis of the ABCD2 score alone requires further study. However, such decisions, though informed by clinical data, depend heavily on societal input (eg, from insurance companies, national health protocols) and may be outside the purview of clinical investigation.

The benefits of hospitalization include the ability to rapidly carry out tests such as cardiac monitoring for atrial fibrillation; to detect atherosclerosis, aortic arch atheroma, and paradoxical emboli; and to quickly start secondary prevention treatments and education about the importance of adhering to them. Early endarterectomy in the case of carotid stenosis can be offered. Additionally, if stroke symptoms recur, thrombolytic drug therapy can be started quickly.

Nguyen-Huynh et al⁵⁶ analyzed the cost utility of 24-hour hospitalization for patients diagnosed with a recent TIA who were candidates for tissue plasminogen activator if a stroke occurred. They found hospitalization to be borderline cost-effective on the whole, with definite cost-effectiveness found in patients with higher stroke risk.

If patients come to medical attention several days after the TIA, then assessing risk with the ABCD2 score may no longer be reliable.⁵⁷

INVESTIGATIONS

Parenchymal neuroimaging

Computed tomography (CT) without contrast is the most widely used neuroimaging test in the acute setting, since it is widely available, fast, and relatively low-cost. It will not show any abnormality in TIA or early ischemic stroke. However, it is helpful as a screening tool to rule out intracranial lesions such as hemorrhage or tumor. It may also show evidence of established infarction, which would indicate that the ischemia probably had been present for at least 6 to 12 hours.

MRI is clearly superior to noncontrast CT for detecting small areas of ischemia in patients with TIA, and it should be used unless the patient has a contraindication to it. Roughly one-third of TIA patients have lesions detectable on diffusion-weighted imaging, which helps to confirm that the episode was caused by cerebral ischemia, but nearly half of the diffusion MRI changes may be fully reversible.⁵⁸ Evidence of prior stroke, leukoaraiosis, or white matter disease on fluid-attenuated inversion recovery and T2 sequences and microhemorrhages (on gradient echo sequences) help to determine a mechanistic diagnosis.

Subcategorizing TIA patients on the basis of the findings on diffusion-weighted MRI and the ABCD² score is prognostically helpful.⁵⁹ It can help to determine which patients need hospitalization and aggressive treatment, and in the case of identified diffusion-weighted MRI-positive stroke, it helps to localize and elucidate the mechanism of stroke. Hence, MRI is the preferred neuroimaging study for evaluating patients with TIA.³

Vascular imaging

Establishing the status of both intracranial and extracranial vessels is important for understanding the etiology, estimating the risk of future ischemic events, and formulating a treatment plan—eg, carotid endarterectomy in cases of significant stenosis (70% to 99%), which reduces the risk of ipsilateral stroke.⁶⁰ Imaging studies include CT angiography, magnetic resonance angiography, extracranial and transcranial ultrasonography, and conventional catheter-based angiography.

CT angiography has higher spatial resolution, but vessels may be obscured by calcification associated with atherosclerotic plaque. It has the advantage of wide availability, low cost, short scanning time, and excellent patient tolerability.

Magnetic resonance angiography with gadolinium enhancement offers good quality imaging from the great vessels in the chest to the medium-sized vessels distal to the circle of Willis.

The contrast agents used in MRI and CT can have negative consequences in patients with renal disease. MRI contrast has been associated with nephrogenic fibrosing dermopathy, ⁶¹ and CT contrast can cause contrast-induced nephropathy.⁶²

Carotid ultrasonography and transcranial Doppler ultrasonography are noninvasive and are not associated with significant adverse events. They can be used safely in patients with renal dysfunction, and they provide physiologic information that cannot be obtained with MRI and CT, which are static imaging techniques. Detecting microemboli on transcranial Doppler is an independent predictor of recurrent ischemic events.^63,64

Catheter-based angiography is occasionally needed in confusing or more complicated cases, but it is invasive and occasionally is associated with iatrogenic stroke and other vascular complications.

Cardiac and aortic imaging

Echocardiography is used to detect lesions that can be sources of embolism such as regional wall-motion abnormalities, cardiac thrombus or mass, endocarditis, aortic arch atheroma, and patent foramen ovale. In patients with cryptogenic TIA or stroke, those with patent foramen ovale alone were found to have a lower risk of recurrent stroke than those who had both atrial septal aneurysm and patent foramen ovale.⁶⁵

Transesophageal echocardiography is more sensitive than transthoracic echocardiography for detecting cardioembolic lesions, especially patent foramen ovale.⁶⁶ In patients with cerebral ischemia and normal transthoracic findings, cardiac sources of embolism may be detected in about 40% of patients with transesophageal echocardiography.⁶⁷

Cardiac rhythm monitoring

Electrocardiography and prolonged telemetry are recommended in patients with cryptogenic TIA to detect cardiac ischemia and paroxysmal atrial fibrillation. In one study, Holter monitoring detected atrial fibrillation in 6% of patients hospitalized with ischemic stroke or TIA.⁶⁸ In another study, atrial fibrillation was detected after a median of 21 days of outpatient cardiac monitoring in 23% of patients.⁶⁹

The optimal duration of outpatient telemetry has not yet been established, but studies have found significant increases in detection of paroxysms of atrial fibrillation with monitoring for 7 or longer.⁷⁰

Laboratory tests in the acute setting

These include lipid profile, hemoglobin A_1c, and cardiac enzymes. The advantages of hospitalization are early detection of these modifiable risk factors and early initiation of treatment.

Tests for rarer disorders

Tests for rarer disorders are sometimes indicated in unusual cases, such as ischemic symptoms occurring in young patients without other common risk factors. This includes testing for prothrombotic states, toxicology, blood cultures, inflammatory markers, hemoglobin electrophoresis, and lumbar puncture. The benefit of routine testing for thrombophilic disorders in cerebrovascular disease remains uncertain, with no clear association demonstrated with arterial stroke, but testing is more relevant in the case of venous (and paradoxical) thromboembolism.⁷¹

TREAT THE UNDERLYING DISORDER

Treatment depends on the mechanism that is thought to be responsible for the ischemic event. Vascular risk factors are important to identify and modify for all stroke subtypes.

Illustrating the importance of treating TIA and minor stroke, one study⁷² found that for antiplatelet therapy (aspirin, dipyridamole, or aspirin plus dipyridamole), the number needed to treat for 2 years was around 18.

Anticoagulation for cardioembolism

Atrial fibrillation, especially following a cerebrovascular ischemic event, should be treated with long-term anticoagulation with warfarin (Coumadin), dabigatran (Pradaxa), rivaroxaban (Xarelto), or apixaban (Eliquis).⁷³ If the patient cannot tolerate anticoagulation, aspirin is recommended, and if he or she cannot tolerate aspirin, clopidogrel (Plavix) is recommended.

Antiplatelet therapy for large-vessel atherosclerosis and small-vessel disease

In the acute phase, aspirin 81 mg to 325 mg orally can be given. If the patient is allergic to aspirin, a loading dose of clopidogrel 300 mg and then 75 mg daily may be given.

A pilot study of loading with aspirin 325 mg or clopidogrel 375 mg in acute ischemic stroke and TIA patients showed that these treatments were safe when given within 36 hours and decreased the risk of neurologic deterioration.⁷⁴ The patient should continue on aspirin 81 mg or clopidogrel 75 mg, as suggested by the Fast Assessment of Stroke and Transient Ischaemic Attack to Prevent Early Recurrence (FASTER) trial.⁷⁵ In the long term, an antiplatelet drug such as aspirin or clopidogrel or the combination of aspirin and extended-release dipyridamole is reasonable.⁷⁶

Cilostazol (Pletal) is not inferior and is possibly superior to aspirin in preventing noncardioembolic ischemic stroke. It is used off-label for secondary prevention of stroke of noncardioembolic origin.⁷⁷

Statins

In the Stroke Prevention by Aggressive Reduction in Cholesterol Levels (SPARCL) trial, high-dose atorvastatin (Lipitor)—80 mg daily—was found to reduce the risk of subsequent stroke and other cardiovascular events in patients with recent stroke irrespective of low-density lipoprotein cholesterol (LDL-C) level, but there was a small increase in the risk of hemorrhagic stroke.⁷⁸

In patients with hyperlipidemia, current recommendations suggest a target LDL-C level lower than 100 mg/dL in patients with atherosclerotic stroke or TIA, and lower than 70 mg/dL in those with concomitant diabetes.⁷⁹

Antihypertensive therapy

In the acute period, ie, the first 24 hours after symptoms, guidelines have advocated allowing high blood pressure to remain high (“permissive hypertension”) unless the systolic pressure is greater than 200 mm Hg or the diastolic pressure is greater than 120 mm Hg or the patient is receiving thrombolytic therapy.⁸⁰ However, this has recently been challenged by findings in randomized trials.⁸¹ Permissive hypertension and avoidance of dehydration with intravenous normal saline may improve cerebral perfusion, which is especially important in patients with high-grade intracranial or extracranial stenosis. Within the parameters outlined above, we recommend against aggressively treating high blood pressure in the acute phase.

In the long term, antihypertensive therapy reduces the risk of recurrent stroke or TIA.⁸² The goal is to keep blood pressure lower than 140/90 mm Hg, or lower than 130/80 mm Hg in patients with diabetes. A study of patients with ischemic noncardioembolic stroke showed a higher risk of recurrent stroke if the systolic blood pressure was lower than 120 or higher than 140 mm Hg.⁸³

Some classes of antihypertensive medication may be more beneficial than others. There is some evidence that angiotensin-converting enzyme (ACE) inhibitors alone or in combination with a diuretic or an angiotensin receptor blocker are superior to other regimens, possibly because of neuroprotective mechanisms.⁸⁴ A recent meta-analysis found angiotensin receptor blockers to be more effective than either ACE inhibitors or beta-blockers in stroke prevention; however, calcium channel blockers were superior to renin-angiotensin system blockers (ACE inhibitors and angiotensin receptor blockers).⁸⁵

Lifestyle modifications

Smoking cessation and cardiovascular exercise for more than 10 minutes more than 3 times per week is strongly recommended.

For patients with diabetes, the goal is to keep the fasting blood glucose level lower than 126 mg/dL.

Moderate alcohol intake has been shown to decrease stroke risk compared with excessive intake or none at all.⁸⁶

Carotid endarterectomy

Carotid endarterectomy has been recommended within 2 weeks of cerebral or retinal TIA in those cases attributable to high-grade internal carotid artery stenosis in patients who have low surgical risk.⁸⁷ This risk can be estimated on the basis of patient factors, surgeon factors, and hospital volume. The specific recommendations are as follows:

• 70% to 99% carotid stenosis: carotid endarterectomy recommended
• 50% to 69% carotid stenosis: carotid endarterectomy recommended in select patients with a perioperative complication rate < 6%
• < 50% carotid stenosis: carotid endarterectomy not routinely recommended.

Carotid artery angioplasty and stenting with distal embolic protection device

Data from the Carotid Revascularization Endarterectomy Versus Stenting Trial (CREST) and European stenting trials indicate that in patients over age 70, carotid endarterectomy appears to be superior to carotid artery stenting, whereas in younger patients the periprocedural risks of stroke and death are similar. Hence, carotid artery stenting performed by an interventionist with a low complication rate is a reasonable alternative to carotid endarterectomy.^88,89

A transient ischemic attack (TIA), like an episode of unstable angina, is an ominous portent of future morbidity and death even though, by definition, the event leaves no residual neurologic deficit.

But there is a positive side. When a patient presents with a TIA, the physician has the rare opportunity to reduce the risk of a disabling outcome—in this case, stroke. Therefore, patients deserve a rapid and thorough evaluation and appropriate stroke-preventive treatment.

MANY ‘TIAs’ ARE ACTUALLY STROKES

TIA has traditionally been described as a sudden focal neurologic deficit that lasts less than 24 hours, is presumed to be of vascular origin, and is confined to an area of the brain, spinal cord, or eye perfused by a specific artery. This symptom-based definition was based on the arbitrary and inaccurate assumption that brief symptoms would not be associated with damage to brain parenchyma.

The definition has since been updated and made more rational based on new concepts of brain ischemia informed by imaging, especially diffusion-weighted magnetic resonance imaging (MRI).¹ One-third of episodes characterized as a TIA according to the classic definition would be considered an infarction on the basis of diffusion-weighted MRI.² The new tissue-based definition characterizes TIA as a brief episode of neurologic dysfunction caused by focal ischemia of the brain, spinal cord, or retina, with clinical symptoms lasting less than 24 hours and without evidence of acute infarction.³

AN OPPORTUNITY TO INTERVENE

Most TIAs resolve in less than 30 minutes. The US National Institute of Neurological Disorders and Stroke trial of tissue plasminogen activator found that if symptoms of cerebral ischemia had not resolved by 1 hour or had not rapidly improved within 3 hours, complete resolution was rare (only 2% at 24 hours).⁴ Hence, physicians evaluating and treating patients with TIAs should treat these episodes with the urgency they deserve.

Moreover, half of the strokes that follow TIAs occur within 48 hours.⁵ A rapid and thorough evaluation and the initiation of secondary preventive treatments have been shown to reduce the early occurrence of stroke by up to 80%.⁶ Hence, the correct diagnosis of TIA gives the clinician the best opportunity to prevent stroke and its personal, social, and sometimes fatal consequences.

STROKES OUTNUMBER TIAs, BUT TIAs ARE UNDERREPORTED

According to 2012 statistics, nearly 795,000 strokes occur in the United States each year, 610,000 of which are first attacks and 185,000 are recurrences. Every 40 seconds, someone in the United States has a stroke.⁷

In comparison, the incidence of TIA in the United States is estimated at 200,000 to 500,000 per year, though the true number is difficult to know because of underreporting.^8,9 About half of patients who experience a TIA fail to report it to their health care provider—a lost opportunity for intervention and stroke prevention.^10,11

A meta-analysis showed that the risk of stroke after TIA was 9.9% at 2 days, 13.4% at 30 days, and 17.3% at 90 days.¹²

Interestingly, the risk of stroke after TIA exceeds the risk of recurrent stroke after a first stroke. This was shown in a study that found that patients who had made a substantial recovery within 24 hours (ie, patients with a TIA) were more likely to suffer neurologic deterioration in the next 3 months than were those who did not have significant early improvement.¹³

RISK FACTORS FOR TIA ARE THE SAME AS FOR STROKE

The risk of cerebrovascular disease increases with age and is higher in men¹⁴ and in blacks and Hispanics.¹⁵

The risk factors and clinical presentation do not differ between TIA and stroke, so the evaluation and treatment should not differ either. These two events represent a continuum of the same disease entity.

Some risk factors for TIA are modifiable, others are not.

Nonmodifiable risk factors

Nonmodifiable risk factors for TIA include older age, male sex, African American race, low birth weight, Hispanic ethnicity, and family history. If the patient has nonmodifiable risk factors, we should try all the harder to correct the modifiable ones.

Older age. The risk of ischemic stroke and intracranial hemorrhage doubles with each decade after age 55 in both sexes.¹⁶

Sex. Men have a significantly higher incidence of TIA than women,¹¹ whereas the opposite is true for stroke: women have a higher lifetime risk of stroke than men.¹⁷

African Americans have an incidence of stroke (all types) 38% higher than that of whites,¹⁸ and an incidence of TIA (inpatient and out-of-hospital) 40% higher than the overall age- and sex-adjusted rate in the white population.¹¹

Low birth weight. The odds of stroke are more than twice as high in people who weighed less than 2,500 g at birth compared with those who weighed 4,000 g or more, probably because of a correlation between low birth weight and hypertension.¹⁹

A family history of stroke increases the risk of stroke by nearly 30%, the association being stronger with large-vessel and smallvessel stroke than with cardioembolic stroke.²⁰

Modifiable risk factors

Modifiable risk factors include cigarette smoking, hypertension, diabetes, lipid abnormalities, atrial fibrillation, carotid stenosis, and dietary and hormonal factors. Detecting these factors, which often coexist, is the first step in trying to modify them and reduce the patient’s risk.

Cigarette smoking approximately doubles the risk of ischemic stroke.^21–23

Hypertension has a relationship with stroke risk that is strong, continuous, graded, consistent, and significant.²⁴

Diabetes increases stroke risk nearly six times.²⁵

Lipid abnormalities. Most studies have found an association between lipid levels (total cholesterol and low-density lipoprotein cholesterol) and the risk of death from ischemic stroke,^26–28 and an inverse relationship between high-density lipoprotein cholesterol levels and stroke risk.²⁹

Atrial fibrillation increases the risk of ischemic stroke up to fivefold, even in the absence of cardiac valvular disease. The mechanism is embolism of stasis-induced thrombi that form in the left atrial appendage.³⁰

Carotid stenosis. Asymptomatic carotid atherosclerotic stenotic lesions in the extracranial internal carotid artery or carotid bulb are associated with a higher risk of stroke.^24,31

Lifestyle factors. Diets that lower blood pressure have been found to decrease stroke risk.²⁴ Exercise in men and women reduces the risk of stroke or death by 25% to 30% compared with inactive people.³² Weight reduction has been found to lower blood pressure and reduce stroke risk.²⁴

Other potentially modifiable risk factors include migraine with aura, metabolic syndrome, excess alcohol consumption (and, paradoxically, complete abstinence from alcohol), drug abuse, sleep-disordered breathing, hyperhomocysteinemia, high lipoprotein (a) levels, hypercoagulability, infection with organisms such as Chlamydia pneumoniae, cytomegalovirus, and Helicobacter pylori, and acute infections such as respiratory and urinary infections.²⁶

Conditions in certain demographic groups

Patients in certain demographic groups present with rarer conditions associated with stroke and TIA.

Sickle cell disease. Eleven percent of patients with sickle cell disease have clinical strokes, and a substantial number have “silent” infarcts identified on neuroimaging.^33,34

Postmenopausal hormone replacement therapy with any product containing conjugated equine estrogen carries a risk of cerebrovascular events,³⁵ and the higher the dose, the higher the risk.³⁶ Also, oral contraceptives may be harmful in women who have additional risk factors such as cigarette smoking, prior thromboembolic events, or migraine with aura.^37,38

THREE CAUSES OF STROKE AND TIA

Stroke and TIA should not be considered diagnoses in themselves, but rather the end point of many other diseases. The diagnosis lies in identifying the mechanism of the cerebrovascular event. The three main mechanisms are thrombosis, embolism, and decreased perfusion.

Thrombosis is caused by obstruction of blood flow within one or more blood vessels, the most common cause being atherosclerosis. Large-artery atherosclerosis, such as in the carotid bifurcation or extracranial internal carotid, causes TIAs that occur over a period of weeks or months with a variety of presentations in that vascular territory, from years of gradual accumulation of atherosclerotic plaque.³⁹

In patients with small-artery or penetrating artery disease, hypertension is the primary risk factor and the pathology, specific to small arterioles, is lipohyalinosis rather than atherosclerosis. These patients may present with a stuttering clinical course, and episodes are more stereotypical.

Less common obstructive vascular pathologies include fibromuscular dysplasia, arteritides, and dissection.

Embolism can occur from a proximal source such as the heart or from proximal vessels such as the aorta, carotid, or vertebral arteries. The embolic particle may form on heart valves or lesions within the heart (eg, clot, tumor), or in the venous circulation and paradoxically cross over to the arterial side through an intracardiac or transpulmonary shunt. Embolism may also be due to a hypercoagulable state.⁴⁰ Embolic stroke is suspected when multiple vascular territories within the brain are clinically or radiographically affected.

Decreased systemic perfusion caused by severe heart failure or systemic hypotension can cause ischemia to the brain diffusely and bilaterally, limiting the ability of the blood-stream to wash out microemboli, especially in the border zones (also known as “watershed areas”), thus leading to ischemia or infarction.⁴¹ Decreased perfusion can also be local, due to a fixed vessel stenosis.

Using another classification system, a study in Rochester, MN, found the following incident rates of stroke subtypes, adjusted for age and sex, per 100,000 population⁴²:

• Large-vessel cervical or intracranial atherosclerosis with more than 50% stenosis—27
• Cardioembolism—40
• Lacunar, small-vessel disease—25
• Uncertain cause—52
• Other identifiable cause—4.

THREE CLINICAL FEATURES SUGGEST TIA

TIAs can be hard to distinguish from nonischemic neurologic events in the acute setting such as an emergency room. Up to 60% of patients suspected of having a TIA actually have a nonischemic cause of their symptoms.⁴³

Three clinical features suggest a TIA during the emergency room evaluation:

• Rapid onset of symptoms—“like lightning” or “in seconds,” in contrast to migraine and seizures, which develop over minutes
• No history of similar episodes in the past
• Absence of nonspecific symptoms—eg, stomach upset or tightness in the chest.

CLINICAL DIAGNOSIS

Because most TIA symptoms and signs have already resolved by the time of evaluation, the diagnosis depends on a careful history with special attention to the pace of onset and resolution, the duration and nature of the symptoms, circumstances at the time of symptom onset, previous similar episodes, associated features, vascular risk factors, and family history (Table 1).^44,45

A detailed neurologic examination is imperative and should include fundoscopy. A cardiovascular assessment should include cardiac rhythm, bruits in the neck, orbits, and groin, peripheral pulses, and electrocardiography.

Do neurologists do a better job at diagnosing TIA and stroke?

Primary care physicians, internists, and emergency department physicians are often the ones to carry out the clinical assessment of possible TIA.

Determining if transient neurologic symptoms are caused by ischemia can be a challenge. When in doubt, referral to a neurologist with subspecialty training in cerebrovascular disease should be considered.

But do neurologists really do a better job? A recent study sought to compare the accuracy of diagnosis of TIA made by general practitioners, emergency physicians, and neurologists. The nonneurologists considered “confusion” and “unexplained fall” suggestive of TIA and “lower facial palsy” and “monocular blindness” less suggestive of TIA—whereas the opposite is true. This shows that nonneurologists often label minor strokes and several nonvascular transient neurologic disturbances as TIAs, and up to half of patients could be mislabeled as a result.⁴⁶

Differences in diagnosing cerebrovascular events between emergency room physicians and attending neurologists have been tested,⁴⁷ with an accuracy of diagnosis as low as 38% by emergency department physicians in one study.⁴⁸ However, other studies did not show such a trend.^49,50

A study at a university-based teaching hospital found the sensitivity of emergency room physician diagnosis to be 98.6% with a positive predictive value of 94.8%,⁴⁹ showing that at a large teaching hospital with a comprehensive stroke intervention program, emergency physicians could identify patients with stroke, particularly hemorrhagic stroke, very accurately.

Improving the diagnosis of stroke and TIA

Routine use of imaging and involvement of a neurologist increase the sensitivity and accuracy of diagnosis. Education and written guidelines for acute stroke treatment both in the emergency department and in out-of-hospital settings seem to dramatically improve the rates of diagnostic accuracy and appropriate treatment.⁵⁰

Emergency medical service personnel use two screening tools in the field to identify TIA and stroke symptoms:

• The Cincinnati Prehospital Stroke Scale, a three-item scale based on three signs: facial droop, arm drift, and slurring of speech⁵¹
• The Los Angeles Prehospital Stroke Screen, which uses screening questions and asymmetry in the face, hand grip strength, and arm drift.⁵²

Knowing that the patient is having a minor stroke or TIA is important. Urgent treatment of these conditions decreases the risk of stroke in the next 90 days, which was 10.5% in one study.⁵ Urgent assessment and early intervention could reduce this risk of subsequent stroke down to 2%.⁶

ASSESSING RISK OF STROKE AFTER TIA

There is a practical need for prediction of stroke during the first few days after the event. The ABCD and ABCD2 scores were developed to stratify the short-term risk of stroke in patients with recent TIA.

The ABCD score

The ABCD score⁵³ was derived to allow primary care physicians and other physicians to identify which patients with a suspected diagnosis of TIA should be referred for emergency assessment, to allow secondary-care physicians to determine which patients with probable or definite TIA need emergency investigation and treatment, to allow public education about the need for medical attention after a TIA, and to identify people at high risk.

The ABCD2 score

The ABCD2 score predicts the short-term risk of stroke following a TIA.⁵⁴ Points are assigned as follows:

• Age > 60 years: 1 point
• Blood pressure (systolic) > 140 mm Hg or diastolic blood pressure > 90 mm Hg: 1 point
• Clinical factors: unilateral weakness with or without speech impairment: 2 points (1 point for speech impairment without weakness)
• Duration of symptoms > 60 minutes: 2 points (1 point for 10–59 minutes)
• Diabetes: 1 point.

Thus, the possible total ranges from 0 to 7 points. Higher scores indicate a greater risk of stroke at 2, 7, 30, and 90 days:

• Total score 0, 1, 2, or 3: 2-day stroke risk 1.0% (low risk)
• Total score 4 or 5: 2-day stroke risk 4.1% (moderate risk)
• Total score 6 or 7: 2-day stroke risk 8.1% (high risk).

WHO SHOULD BE HOSPITALIZED?

It has been suggested that the ABCD2 score can help in triaging patients to hospital admission or outpatient care, though no randomized trial has actually evaluated the utility of the ABCD2 score in this way.³

A study of consecutive TIA patients admitted over 12 months⁵⁵ found that patients with an ABCD2 score of 3 or less had the same chance of requiring hospitalization (based on positive diffusion-weighted MRI studies, risk factor identification, and treatment initiation) as those with a score of 4 to 7. Hence, admitting TIA patients on the basis of the ABCD2 score alone requires further study. However, such decisions, though informed by clinical data, depend heavily on societal input (eg, from insurance companies, national health protocols) and may be outside the purview of clinical investigation.

The benefits of hospitalization include the ability to rapidly carry out tests such as cardiac monitoring for atrial fibrillation; to detect atherosclerosis, aortic arch atheroma, and paradoxical emboli; and to quickly start secondary prevention treatments and education about the importance of adhering to them. Early endarterectomy in the case of carotid stenosis can be offered. Additionally, if stroke symptoms recur, thrombolytic drug therapy can be started quickly.

Nguyen-Huynh et al⁵⁶ analyzed the cost utility of 24-hour hospitalization for patients diagnosed with a recent TIA who were candidates for tissue plasminogen activator if a stroke occurred. They found hospitalization to be borderline cost-effective on the whole, with definite cost-effectiveness found in patients with higher stroke risk.

If patients come to medical attention several days after the TIA, then assessing risk with the ABCD2 score may no longer be reliable.⁵⁷

INVESTIGATIONS

Parenchymal neuroimaging

Computed tomography (CT) without contrast is the most widely used neuroimaging test in the acute setting, since it is widely available, fast, and relatively low-cost. It will not show any abnormality in TIA or early ischemic stroke. However, it is helpful as a screening tool to rule out intracranial lesions such as hemorrhage or tumor. It may also show evidence of established infarction, which would indicate that the ischemia probably had been present for at least 6 to 12 hours.

MRI is clearly superior to noncontrast CT for detecting small areas of ischemia in patients with TIA, and it should be used unless the patient has a contraindication to it. Roughly one-third of TIA patients have lesions detectable on diffusion-weighted imaging, which helps to confirm that the episode was caused by cerebral ischemia, but nearly half of the diffusion MRI changes may be fully reversible.⁵⁸ Evidence of prior stroke, leukoaraiosis, or white matter disease on fluid-attenuated inversion recovery and T2 sequences and microhemorrhages (on gradient echo sequences) help to determine a mechanistic diagnosis.

Subcategorizing TIA patients on the basis of the findings on diffusion-weighted MRI and the ABCD² score is prognostically helpful.⁵⁹ It can help to determine which patients need hospitalization and aggressive treatment, and in the case of identified diffusion-weighted MRI-positive stroke, it helps to localize and elucidate the mechanism of stroke. Hence, MRI is the preferred neuroimaging study for evaluating patients with TIA.³

Vascular imaging

Establishing the status of both intracranial and extracranial vessels is important for understanding the etiology, estimating the risk of future ischemic events, and formulating a treatment plan—eg, carotid endarterectomy in cases of significant stenosis (70% to 99%), which reduces the risk of ipsilateral stroke.⁶⁰ Imaging studies include CT angiography, magnetic resonance angiography, extracranial and transcranial ultrasonography, and conventional catheter-based angiography.

CT angiography has higher spatial resolution, but vessels may be obscured by calcification associated with atherosclerotic plaque. It has the advantage of wide availability, low cost, short scanning time, and excellent patient tolerability.

Magnetic resonance angiography with gadolinium enhancement offers good quality imaging from the great vessels in the chest to the medium-sized vessels distal to the circle of Willis.

The contrast agents used in MRI and CT can have negative consequences in patients with renal disease. MRI contrast has been associated with nephrogenic fibrosing dermopathy, ⁶¹ and CT contrast can cause contrast-induced nephropathy.⁶²

Carotid ultrasonography and transcranial Doppler ultrasonography are noninvasive and are not associated with significant adverse events. They can be used safely in patients with renal dysfunction, and they provide physiologic information that cannot be obtained with MRI and CT, which are static imaging techniques. Detecting microemboli on transcranial Doppler is an independent predictor of recurrent ischemic events.^63,64

Catheter-based angiography is occasionally needed in confusing or more complicated cases, but it is invasive and occasionally is associated with iatrogenic stroke and other vascular complications.

Cardiac and aortic imaging

Echocardiography is used to detect lesions that can be sources of embolism such as regional wall-motion abnormalities, cardiac thrombus or mass, endocarditis, aortic arch atheroma, and patent foramen ovale. In patients with cryptogenic TIA or stroke, those with patent foramen ovale alone were found to have a lower risk of recurrent stroke than those who had both atrial septal aneurysm and patent foramen ovale.⁶⁵

Transesophageal echocardiography is more sensitive than transthoracic echocardiography for detecting cardioembolic lesions, especially patent foramen ovale.⁶⁶ In patients with cerebral ischemia and normal transthoracic findings, cardiac sources of embolism may be detected in about 40% of patients with transesophageal echocardiography.⁶⁷

Cardiac rhythm monitoring

Electrocardiography and prolonged telemetry are recommended in patients with cryptogenic TIA to detect cardiac ischemia and paroxysmal atrial fibrillation. In one study, Holter monitoring detected atrial fibrillation in 6% of patients hospitalized with ischemic stroke or TIA.⁶⁸ In another study, atrial fibrillation was detected after a median of 21 days of outpatient cardiac monitoring in 23% of patients.⁶⁹

The optimal duration of outpatient telemetry has not yet been established, but studies have found significant increases in detection of paroxysms of atrial fibrillation with monitoring for 7 or longer.⁷⁰

Laboratory tests in the acute setting

These include lipid profile, hemoglobin A_1c, and cardiac enzymes. The advantages of hospitalization are early detection of these modifiable risk factors and early initiation of treatment.

Tests for rarer disorders

Tests for rarer disorders are sometimes indicated in unusual cases, such as ischemic symptoms occurring in young patients without other common risk factors. This includes testing for prothrombotic states, toxicology, blood cultures, inflammatory markers, hemoglobin electrophoresis, and lumbar puncture. The benefit of routine testing for thrombophilic disorders in cerebrovascular disease remains uncertain, with no clear association demonstrated with arterial stroke, but testing is more relevant in the case of venous (and paradoxical) thromboembolism.⁷¹

TREAT THE UNDERLYING DISORDER

Treatment depends on the mechanism that is thought to be responsible for the ischemic event. Vascular risk factors are important to identify and modify for all stroke subtypes.

Illustrating the importance of treating TIA and minor stroke, one study⁷² found that for antiplatelet therapy (aspirin, dipyridamole, or aspirin plus dipyridamole), the number needed to treat for 2 years was around 18.

Anticoagulation for cardioembolism

Atrial fibrillation, especially following a cerebrovascular ischemic event, should be treated with long-term anticoagulation with warfarin (Coumadin), dabigatran (Pradaxa), rivaroxaban (Xarelto), or apixaban (Eliquis).⁷³ If the patient cannot tolerate anticoagulation, aspirin is recommended, and if he or she cannot tolerate aspirin, clopidogrel (Plavix) is recommended.

Antiplatelet therapy for large-vessel atherosclerosis and small-vessel disease

In the acute phase, aspirin 81 mg to 325 mg orally can be given. If the patient is allergic to aspirin, a loading dose of clopidogrel 300 mg and then 75 mg daily may be given.

A pilot study of loading with aspirin 325 mg or clopidogrel 375 mg in acute ischemic stroke and TIA patients showed that these treatments were safe when given within 36 hours and decreased the risk of neurologic deterioration.⁷⁴ The patient should continue on aspirin 81 mg or clopidogrel 75 mg, as suggested by the Fast Assessment of Stroke and Transient Ischaemic Attack to Prevent Early Recurrence (FASTER) trial.⁷⁵ In the long term, an antiplatelet drug such as aspirin or clopidogrel or the combination of aspirin and extended-release dipyridamole is reasonable.⁷⁶

Cilostazol (Pletal) is not inferior and is possibly superior to aspirin in preventing noncardioembolic ischemic stroke. It is used off-label for secondary prevention of stroke of noncardioembolic origin.⁷⁷

Statins

In the Stroke Prevention by Aggressive Reduction in Cholesterol Levels (SPARCL) trial, high-dose atorvastatin (Lipitor)—80 mg daily—was found to reduce the risk of subsequent stroke and other cardiovascular events in patients with recent stroke irrespective of low-density lipoprotein cholesterol (LDL-C) level, but there was a small increase in the risk of hemorrhagic stroke.⁷⁸

In patients with hyperlipidemia, current recommendations suggest a target LDL-C level lower than 100 mg/dL in patients with atherosclerotic stroke or TIA, and lower than 70 mg/dL in those with concomitant diabetes.⁷⁹

Antihypertensive therapy

In the acute period, ie, the first 24 hours after symptoms, guidelines have advocated allowing high blood pressure to remain high (“permissive hypertension”) unless the systolic pressure is greater than 200 mm Hg or the diastolic pressure is greater than 120 mm Hg or the patient is receiving thrombolytic therapy.⁸⁰ However, this has recently been challenged by findings in randomized trials.⁸¹ Permissive hypertension and avoidance of dehydration with intravenous normal saline may improve cerebral perfusion, which is especially important in patients with high-grade intracranial or extracranial stenosis. Within the parameters outlined above, we recommend against aggressively treating high blood pressure in the acute phase.

In the long term, antihypertensive therapy reduces the risk of recurrent stroke or TIA.⁸² The goal is to keep blood pressure lower than 140/90 mm Hg, or lower than 130/80 mm Hg in patients with diabetes. A study of patients with ischemic noncardioembolic stroke showed a higher risk of recurrent stroke if the systolic blood pressure was lower than 120 or higher than 140 mm Hg.⁸³

Some classes of antihypertensive medication may be more beneficial than others. There is some evidence that angiotensin-converting enzyme (ACE) inhibitors alone or in combination with a diuretic or an angiotensin receptor blocker are superior to other regimens, possibly because of neuroprotective mechanisms.⁸⁴ A recent meta-analysis found angiotensin receptor blockers to be more effective than either ACE inhibitors or beta-blockers in stroke prevention; however, calcium channel blockers were superior to renin-angiotensin system blockers (ACE inhibitors and angiotensin receptor blockers).⁸⁵

Lifestyle modifications

Smoking cessation and cardiovascular exercise for more than 10 minutes more than 3 times per week is strongly recommended.

For patients with diabetes, the goal is to keep the fasting blood glucose level lower than 126 mg/dL.

Moderate alcohol intake has been shown to decrease stroke risk compared with excessive intake or none at all.⁸⁶

Carotid endarterectomy

Carotid endarterectomy has been recommended within 2 weeks of cerebral or retinal TIA in those cases attributable to high-grade internal carotid artery stenosis in patients who have low surgical risk.⁸⁷ This risk can be estimated on the basis of patient factors, surgeon factors, and hospital volume. The specific recommendations are as follows:

• 70% to 99% carotid stenosis: carotid endarterectomy recommended
• 50% to 69% carotid stenosis: carotid endarterectomy recommended in select patients with a perioperative complication rate < 6%
• < 50% carotid stenosis: carotid endarterectomy not routinely recommended.

Carotid artery angioplasty and stenting with distal embolic protection device

Data from the Carotid Revascularization Endarterectomy Versus Stenting Trial (CREST) and European stenting trials indicate that in patients over age 70, carotid endarterectomy appears to be superior to carotid artery stenting, whereas in younger patients the periprocedural risks of stroke and death are similar. Hence, carotid artery stenting performed by an interventionist with a low complication rate is a reasonable alternative to carotid endarterectomy.^88,89

A transient ischemic attack (TIA), like an episode of unstable angina, is an ominous portent of future morbidity and death even though, by definition, the event leaves no residual neurologic deficit.

But there is a positive side. When a patient presents with a TIA, the physician has the rare opportunity to reduce the risk of a disabling outcome—in this case, stroke. Therefore, patients deserve a rapid and thorough evaluation and appropriate stroke-preventive treatment.

MANY ‘TIAs’ ARE ACTUALLY STROKES

TIA has traditionally been described as a sudden focal neurologic deficit that lasts less than 24 hours, is presumed to be of vascular origin, and is confined to an area of the brain, spinal cord, or eye perfused by a specific artery. This symptom-based definition was based on the arbitrary and inaccurate assumption that brief symptoms would not be associated with damage to brain parenchyma.

The definition has since been updated and made more rational based on new concepts of brain ischemia informed by imaging, especially diffusion-weighted magnetic resonance imaging (MRI).¹ One-third of episodes characterized as a TIA according to the classic definition would be considered an infarction on the basis of diffusion-weighted MRI.² The new tissue-based definition characterizes TIA as a brief episode of neurologic dysfunction caused by focal ischemia of the brain, spinal cord, or retina, with clinical symptoms lasting less than 24 hours and without evidence of acute infarction.³

AN OPPORTUNITY TO INTERVENE

Most TIAs resolve in less than 30 minutes. The US National Institute of Neurological Disorders and Stroke trial of tissue plasminogen activator found that if symptoms of cerebral ischemia had not resolved by 1 hour or had not rapidly improved within 3 hours, complete resolution was rare (only 2% at 24 hours).⁴ Hence, physicians evaluating and treating patients with TIAs should treat these episodes with the urgency they deserve.

Moreover, half of the strokes that follow TIAs occur within 48 hours.⁵ A rapid and thorough evaluation and the initiation of secondary preventive treatments have been shown to reduce the early occurrence of stroke by up to 80%.⁶ Hence, the correct diagnosis of TIA gives the clinician the best opportunity to prevent stroke and its personal, social, and sometimes fatal consequences.

STROKES OUTNUMBER TIAs, BUT TIAs ARE UNDERREPORTED

According to 2012 statistics, nearly 795,000 strokes occur in the United States each year, 610,000 of which are first attacks and 185,000 are recurrences. Every 40 seconds, someone in the United States has a stroke.⁷

In comparison, the incidence of TIA in the United States is estimated at 200,000 to 500,000 per year, though the true number is difficult to know because of underreporting.^8,9 About half of patients who experience a TIA fail to report it to their health care provider—a lost opportunity for intervention and stroke prevention.^10,11

A meta-analysis showed that the risk of stroke after TIA was 9.9% at 2 days, 13.4% at 30 days, and 17.3% at 90 days.¹²

Interestingly, the risk of stroke after TIA exceeds the risk of recurrent stroke after a first stroke. This was shown in a study that found that patients who had made a substantial recovery within 24 hours (ie, patients with a TIA) were more likely to suffer neurologic deterioration in the next 3 months than were those who did not have significant early improvement.¹³

RISK FACTORS FOR TIA ARE THE SAME AS FOR STROKE

The risk of cerebrovascular disease increases with age and is higher in men¹⁴ and in blacks and Hispanics.¹⁵

The risk factors and clinical presentation do not differ between TIA and stroke, so the evaluation and treatment should not differ either. These two events represent a continuum of the same disease entity.

Some risk factors for TIA are modifiable, others are not.

Nonmodifiable risk factors

Nonmodifiable risk factors for TIA include older age, male sex, African American race, low birth weight, Hispanic ethnicity, and family history. If the patient has nonmodifiable risk factors, we should try all the harder to correct the modifiable ones.

Older age. The risk of ischemic stroke and intracranial hemorrhage doubles with each decade after age 55 in both sexes.¹⁶

Sex. Men have a significantly higher incidence of TIA than women,¹¹ whereas the opposite is true for stroke: women have a higher lifetime risk of stroke than men.¹⁷

African Americans have an incidence of stroke (all types) 38% higher than that of whites,¹⁸ and an incidence of TIA (inpatient and out-of-hospital) 40% higher than the overall age- and sex-adjusted rate in the white population.¹¹

Low birth weight. The odds of stroke are more than twice as high in people who weighed less than 2,500 g at birth compared with those who weighed 4,000 g or more, probably because of a correlation between low birth weight and hypertension.¹⁹

A family history of stroke increases the risk of stroke by nearly 30%, the association being stronger with large-vessel and smallvessel stroke than with cardioembolic stroke.²⁰

Modifiable risk factors

Modifiable risk factors include cigarette smoking, hypertension, diabetes, lipid abnormalities, atrial fibrillation, carotid stenosis, and dietary and hormonal factors. Detecting these factors, which often coexist, is the first step in trying to modify them and reduce the patient’s risk.

Cigarette smoking approximately doubles the risk of ischemic stroke.^21–23

Hypertension has a relationship with stroke risk that is strong, continuous, graded, consistent, and significant.²⁴

Diabetes increases stroke risk nearly six times.²⁵

Lipid abnormalities. Most studies have found an association between lipid levels (total cholesterol and low-density lipoprotein cholesterol) and the risk of death from ischemic stroke,^26–28 and an inverse relationship between high-density lipoprotein cholesterol levels and stroke risk.²⁹

Atrial fibrillation increases the risk of ischemic stroke up to fivefold, even in the absence of cardiac valvular disease. The mechanism is embolism of stasis-induced thrombi that form in the left atrial appendage.³⁰

Carotid stenosis. Asymptomatic carotid atherosclerotic stenotic lesions in the extracranial internal carotid artery or carotid bulb are associated with a higher risk of stroke.^24,31

Lifestyle factors. Diets that lower blood pressure have been found to decrease stroke risk.²⁴ Exercise in men and women reduces the risk of stroke or death by 25% to 30% compared with inactive people.³² Weight reduction has been found to lower blood pressure and reduce stroke risk.²⁴

Other potentially modifiable risk factors include migraine with aura, metabolic syndrome, excess alcohol consumption (and, paradoxically, complete abstinence from alcohol), drug abuse, sleep-disordered breathing, hyperhomocysteinemia, high lipoprotein (a) levels, hypercoagulability, infection with organisms such as Chlamydia pneumoniae, cytomegalovirus, and Helicobacter pylori, and acute infections such as respiratory and urinary infections.²⁶

Conditions in certain demographic groups

Patients in certain demographic groups present with rarer conditions associated with stroke and TIA.

Sickle cell disease. Eleven percent of patients with sickle cell disease have clinical strokes, and a substantial number have “silent” infarcts identified on neuroimaging.^33,34

Postmenopausal hormone replacement therapy with any product containing conjugated equine estrogen carries a risk of cerebrovascular events,³⁵ and the higher the dose, the higher the risk.³⁶ Also, oral contraceptives may be harmful in women who have additional risk factors such as cigarette smoking, prior thromboembolic events, or migraine with aura.^37,38

THREE CAUSES OF STROKE AND TIA

Stroke and TIA should not be considered diagnoses in themselves, but rather the end point of many other diseases. The diagnosis lies in identifying the mechanism of the cerebrovascular event. The three main mechanisms are thrombosis, embolism, and decreased perfusion.

Thrombosis is caused by obstruction of blood flow within one or more blood vessels, the most common cause being atherosclerosis. Large-artery atherosclerosis, such as in the carotid bifurcation or extracranial internal carotid, causes TIAs that occur over a period of weeks or months with a variety of presentations in that vascular territory, from years of gradual accumulation of atherosclerotic plaque.³⁹

In patients with small-artery or penetrating artery disease, hypertension is the primary risk factor and the pathology, specific to small arterioles, is lipohyalinosis rather than atherosclerosis. These patients may present with a stuttering clinical course, and episodes are more stereotypical.

Less common obstructive vascular pathologies include fibromuscular dysplasia, arteritides, and dissection.

Embolism can occur from a proximal source such as the heart or from proximal vessels such as the aorta, carotid, or vertebral arteries. The embolic particle may form on heart valves or lesions within the heart (eg, clot, tumor), or in the venous circulation and paradoxically cross over to the arterial side through an intracardiac or transpulmonary shunt. Embolism may also be due to a hypercoagulable state.⁴⁰ Embolic stroke is suspected when multiple vascular territories within the brain are clinically or radiographically affected.

Decreased systemic perfusion caused by severe heart failure or systemic hypotension can cause ischemia to the brain diffusely and bilaterally, limiting the ability of the blood-stream to wash out microemboli, especially in the border zones (also known as “watershed areas”), thus leading to ischemia or infarction.⁴¹ Decreased perfusion can also be local, due to a fixed vessel stenosis.

Using another classification system, a study in Rochester, MN, found the following incident rates of stroke subtypes, adjusted for age and sex, per 100,000 population⁴²:

• Large-vessel cervical or intracranial atherosclerosis with more than 50% stenosis—27
• Cardioembolism—40
• Lacunar, small-vessel disease—25
• Uncertain cause—52
• Other identifiable cause—4.

THREE CLINICAL FEATURES SUGGEST TIA

TIAs can be hard to distinguish from nonischemic neurologic events in the acute setting such as an emergency room. Up to 60% of patients suspected of having a TIA actually have a nonischemic cause of their symptoms.⁴³

Three clinical features suggest a TIA during the emergency room evaluation:

• Rapid onset of symptoms—“like lightning” or “in seconds,” in contrast to migraine and seizures, which develop over minutes
• No history of similar episodes in the past
• Absence of nonspecific symptoms—eg, stomach upset or tightness in the chest.

CLINICAL DIAGNOSIS

Because most TIA symptoms and signs have already resolved by the time of evaluation, the diagnosis depends on a careful history with special attention to the pace of onset and resolution, the duration and nature of the symptoms, circumstances at the time of symptom onset, previous similar episodes, associated features, vascular risk factors, and family history (Table 1).^44,45

A detailed neurologic examination is imperative and should include fundoscopy. A cardiovascular assessment should include cardiac rhythm, bruits in the neck, orbits, and groin, peripheral pulses, and electrocardiography.

Do neurologists do a better job at diagnosing TIA and stroke?

Primary care physicians, internists, and emergency department physicians are often the ones to carry out the clinical assessment of possible TIA.

Determining if transient neurologic symptoms are caused by ischemia can be a challenge. When in doubt, referral to a neurologist with subspecialty training in cerebrovascular disease should be considered.

But do neurologists really do a better job? A recent study sought to compare the accuracy of diagnosis of TIA made by general practitioners, emergency physicians, and neurologists. The nonneurologists considered “confusion” and “unexplained fall” suggestive of TIA and “lower facial palsy” and “monocular blindness” less suggestive of TIA—whereas the opposite is true. This shows that nonneurologists often label minor strokes and several nonvascular transient neurologic disturbances as TIAs, and up to half of patients could be mislabeled as a result.⁴⁶

Differences in diagnosing cerebrovascular events between emergency room physicians and attending neurologists have been tested,⁴⁷ with an accuracy of diagnosis as low as 38% by emergency department physicians in one study.⁴⁸ However, other studies did not show such a trend.^49,50

A study at a university-based teaching hospital found the sensitivity of emergency room physician diagnosis to be 98.6% with a positive predictive value of 94.8%,⁴⁹ showing that at a large teaching hospital with a comprehensive stroke intervention program, emergency physicians could identify patients with stroke, particularly hemorrhagic stroke, very accurately.

Improving the diagnosis of stroke and TIA

Routine use of imaging and involvement of a neurologist increase the sensitivity and accuracy of diagnosis. Education and written guidelines for acute stroke treatment both in the emergency department and in out-of-hospital settings seem to dramatically improve the rates of diagnostic accuracy and appropriate treatment.⁵⁰

Emergency medical service personnel use two screening tools in the field to identify TIA and stroke symptoms:

• The Cincinnati Prehospital Stroke Scale, a three-item scale based on three signs: facial droop, arm drift, and slurring of speech⁵¹
• The Los Angeles Prehospital Stroke Screen, which uses screening questions and asymmetry in the face, hand grip strength, and arm drift.⁵²

Knowing that the patient is having a minor stroke or TIA is important. Urgent treatment of these conditions decreases the risk of stroke in the next 90 days, which was 10.5% in one study.⁵ Urgent assessment and early intervention could reduce this risk of subsequent stroke down to 2%.⁶

ASSESSING RISK OF STROKE AFTER TIA

There is a practical need for prediction of stroke during the first few days after the event. The ABCD and ABCD2 scores were developed to stratify the short-term risk of stroke in patients with recent TIA.

The ABCD score

The ABCD score⁵³ was derived to allow primary care physicians and other physicians to identify which patients with a suspected diagnosis of TIA should be referred for emergency assessment, to allow secondary-care physicians to determine which patients with probable or definite TIA need emergency investigation and treatment, to allow public education about the need for medical attention after a TIA, and to identify people at high risk.

The ABCD2 score

The ABCD2 score predicts the short-term risk of stroke following a TIA.⁵⁴ Points are assigned as follows:

• Age > 60 years: 1 point
• Blood pressure (systolic) > 140 mm Hg or diastolic blood pressure > 90 mm Hg: 1 point
• Clinical factors: unilateral weakness with or without speech impairment: 2 points (1 point for speech impairment without weakness)
• Duration of symptoms > 60 minutes: 2 points (1 point for 10–59 minutes)
• Diabetes: 1 point.

Thus, the possible total ranges from 0 to 7 points. Higher scores indicate a greater risk of stroke at 2, 7, 30, and 90 days:

• Total score 0, 1, 2, or 3: 2-day stroke risk 1.0% (low risk)
• Total score 4 or 5: 2-day stroke risk 4.1% (moderate risk)
• Total score 6 or 7: 2-day stroke risk 8.1% (high risk).

WHO SHOULD BE HOSPITALIZED?

It has been suggested that the ABCD2 score can help in triaging patients to hospital admission or outpatient care, though no randomized trial has actually evaluated the utility of the ABCD2 score in this way.³

A study of consecutive TIA patients admitted over 12 months⁵⁵ found that patients with an ABCD2 score of 3 or less had the same chance of requiring hospitalization (based on positive diffusion-weighted MRI studies, risk factor identification, and treatment initiation) as those with a score of 4 to 7. Hence, admitting TIA patients on the basis of the ABCD2 score alone requires further study. However, such decisions, though informed by clinical data, depend heavily on societal input (eg, from insurance companies, national health protocols) and may be outside the purview of clinical investigation.

The benefits of hospitalization include the ability to rapidly carry out tests such as cardiac monitoring for atrial fibrillation; to detect atherosclerosis, aortic arch atheroma, and paradoxical emboli; and to quickly start secondary prevention treatments and education about the importance of adhering to them. Early endarterectomy in the case of carotid stenosis can be offered. Additionally, if stroke symptoms recur, thrombolytic drug therapy can be started quickly.

Nguyen-Huynh et al⁵⁶ analyzed the cost utility of 24-hour hospitalization for patients diagnosed with a recent TIA who were candidates for tissue plasminogen activator if a stroke occurred. They found hospitalization to be borderline cost-effective on the whole, with definite cost-effectiveness found in patients with higher stroke risk.

If patients come to medical attention several days after the TIA, then assessing risk with the ABCD2 score may no longer be reliable.⁵⁷

INVESTIGATIONS

Parenchymal neuroimaging

Computed tomography (CT) without contrast is the most widely used neuroimaging test in the acute setting, since it is widely available, fast, and relatively low-cost. It will not show any abnormality in TIA or early ischemic stroke. However, it is helpful as a screening tool to rule out intracranial lesions such as hemorrhage or tumor. It may also show evidence of established infarction, which would indicate that the ischemia probably had been present for at least 6 to 12 hours.

MRI is clearly superior to noncontrast CT for detecting small areas of ischemia in patients with TIA, and it should be used unless the patient has a contraindication to it. Roughly one-third of TIA patients have lesions detectable on diffusion-weighted imaging, which helps to confirm that the episode was caused by cerebral ischemia, but nearly half of the diffusion MRI changes may be fully reversible.⁵⁸ Evidence of prior stroke, leukoaraiosis, or white matter disease on fluid-attenuated inversion recovery and T2 sequences and microhemorrhages (on gradient echo sequences) help to determine a mechanistic diagnosis.

Subcategorizing TIA patients on the basis of the findings on diffusion-weighted MRI and the ABCD² score is prognostically helpful.⁵⁹ It can help to determine which patients need hospitalization and aggressive treatment, and in the case of identified diffusion-weighted MRI-positive stroke, it helps to localize and elucidate the mechanism of stroke. Hence, MRI is the preferred neuroimaging study for evaluating patients with TIA.³

Vascular imaging

Establishing the status of both intracranial and extracranial vessels is important for understanding the etiology, estimating the risk of future ischemic events, and formulating a treatment plan—eg, carotid endarterectomy in cases of significant stenosis (70% to 99%), which reduces the risk of ipsilateral stroke.⁶⁰ Imaging studies include CT angiography, magnetic resonance angiography, extracranial and transcranial ultrasonography, and conventional catheter-based angiography.

CT angiography has higher spatial resolution, but vessels may be obscured by calcification associated with atherosclerotic plaque. It has the advantage of wide availability, low cost, short scanning time, and excellent patient tolerability.

Magnetic resonance angiography with gadolinium enhancement offers good quality imaging from the great vessels in the chest to the medium-sized vessels distal to the circle of Willis.

The contrast agents used in MRI and CT can have negative consequences in patients with renal disease. MRI contrast has been associated with nephrogenic fibrosing dermopathy, ⁶¹ and CT contrast can cause contrast-induced nephropathy.⁶²

Carotid ultrasonography and transcranial Doppler ultrasonography are noninvasive and are not associated with significant adverse events. They can be used safely in patients with renal dysfunction, and they provide physiologic information that cannot be obtained with MRI and CT, which are static imaging techniques. Detecting microemboli on transcranial Doppler is an independent predictor of recurrent ischemic events.^63,64

Catheter-based angiography is occasionally needed in confusing or more complicated cases, but it is invasive and occasionally is associated with iatrogenic stroke and other vascular complications.

Cardiac and aortic imaging

Echocardiography is used to detect lesions that can be sources of embolism such as regional wall-motion abnormalities, cardiac thrombus or mass, endocarditis, aortic arch atheroma, and patent foramen ovale. In patients with cryptogenic TIA or stroke, those with patent foramen ovale alone were found to have a lower risk of recurrent stroke than those who had both atrial septal aneurysm and patent foramen ovale.⁶⁵

Transesophageal echocardiography is more sensitive than transthoracic echocardiography for detecting cardioembolic lesions, especially patent foramen ovale.⁶⁶ In patients with cerebral ischemia and normal transthoracic findings, cardiac sources of embolism may be detected in about 40% of patients with transesophageal echocardiography.⁶⁷

Cardiac rhythm monitoring

Electrocardiography and prolonged telemetry are recommended in patients with cryptogenic TIA to detect cardiac ischemia and paroxysmal atrial fibrillation. In one study, Holter monitoring detected atrial fibrillation in 6% of patients hospitalized with ischemic stroke or TIA.⁶⁸ In another study, atrial fibrillation was detected after a median of 21 days of outpatient cardiac monitoring in 23% of patients.⁶⁹

The optimal duration of outpatient telemetry has not yet been established, but studies have found significant increases in detection of paroxysms of atrial fibrillation with monitoring for 7 or longer.⁷⁰

Laboratory tests in the acute setting

These include lipid profile, hemoglobin A_1c, and cardiac enzymes. The advantages of hospitalization are early detection of these modifiable risk factors and early initiation of treatment.

Tests for rarer disorders

Tests for rarer disorders are sometimes indicated in unusual cases, such as ischemic symptoms occurring in young patients without other common risk factors. This includes testing for prothrombotic states, toxicology, blood cultures, inflammatory markers, hemoglobin electrophoresis, and lumbar puncture. The benefit of routine testing for thrombophilic disorders in cerebrovascular disease remains uncertain, with no clear association demonstrated with arterial stroke, but testing is more relevant in the case of venous (and paradoxical) thromboembolism.⁷¹

TREAT THE UNDERLYING DISORDER

Treatment depends on the mechanism that is thought to be responsible for the ischemic event. Vascular risk factors are important to identify and modify for all stroke subtypes.

Illustrating the importance of treating TIA and minor stroke, one study⁷² found that for antiplatelet therapy (aspirin, dipyridamole, or aspirin plus dipyridamole), the number needed to treat for 2 years was around 18.

Anticoagulation for cardioembolism

Atrial fibrillation, especially following a cerebrovascular ischemic event, should be treated with long-term anticoagulation with warfarin (Coumadin), dabigatran (Pradaxa), rivaroxaban (Xarelto), or apixaban (Eliquis).⁷³ If the patient cannot tolerate anticoagulation, aspirin is recommended, and if he or she cannot tolerate aspirin, clopidogrel (Plavix) is recommended.

Antiplatelet therapy for large-vessel atherosclerosis and small-vessel disease

In the acute phase, aspirin 81 mg to 325 mg orally can be given. If the patient is allergic to aspirin, a loading dose of clopidogrel 300 mg and then 75 mg daily may be given.

A pilot study of loading with aspirin 325 mg or clopidogrel 375 mg in acute ischemic stroke and TIA patients showed that these treatments were safe when given within 36 hours and decreased the risk of neurologic deterioration.⁷⁴ The patient should continue on aspirin 81 mg or clopidogrel 75 mg, as suggested by the Fast Assessment of Stroke and Transient Ischaemic Attack to Prevent Early Recurrence (FASTER) trial.⁷⁵ In the long term, an antiplatelet drug such as aspirin or clopidogrel or the combination of aspirin and extended-release dipyridamole is reasonable.⁷⁶

Cilostazol (Pletal) is not inferior and is possibly superior to aspirin in preventing noncardioembolic ischemic stroke. It is used off-label for secondary prevention of stroke of noncardioembolic origin.⁷⁷

Statins

In the Stroke Prevention by Aggressive Reduction in Cholesterol Levels (SPARCL) trial, high-dose atorvastatin (Lipitor)—80 mg daily—was found to reduce the risk of subsequent stroke and other cardiovascular events in patients with recent stroke irrespective of low-density lipoprotein cholesterol (LDL-C) level, but there was a small increase in the risk of hemorrhagic stroke.⁷⁸

In patients with hyperlipidemia, current recommendations suggest a target LDL-C level lower than 100 mg/dL in patients with atherosclerotic stroke or TIA, and lower than 70 mg/dL in those with concomitant diabetes.⁷⁹

Antihypertensive therapy

In the acute period, ie, the first 24 hours after symptoms, guidelines have advocated allowing high blood pressure to remain high (“permissive hypertension”) unless the systolic pressure is greater than 200 mm Hg or the diastolic pressure is greater than 120 mm Hg or the patient is receiving thrombolytic therapy.⁸⁰ However, this has recently been challenged by findings in randomized trials.⁸¹ Permissive hypertension and avoidance of dehydration with intravenous normal saline may improve cerebral perfusion, which is especially important in patients with high-grade intracranial or extracranial stenosis. Within the parameters outlined above, we recommend against aggressively treating high blood pressure in the acute phase.

In the long term, antihypertensive therapy reduces the risk of recurrent stroke or TIA.⁸² The goal is to keep blood pressure lower than 140/90 mm Hg, or lower than 130/80 mm Hg in patients with diabetes. A study of patients with ischemic noncardioembolic stroke showed a higher risk of recurrent stroke if the systolic blood pressure was lower than 120 or higher than 140 mm Hg.⁸³

Some classes of antihypertensive medication may be more beneficial than others. There is some evidence that angiotensin-converting enzyme (ACE) inhibitors alone or in combination with a diuretic or an angiotensin receptor blocker are superior to other regimens, possibly because of neuroprotective mechanisms.⁸⁴ A recent meta-analysis found angiotensin receptor blockers to be more effective than either ACE inhibitors or beta-blockers in stroke prevention; however, calcium channel blockers were superior to renin-angiotensin system blockers (ACE inhibitors and angiotensin receptor blockers).⁸⁵

Lifestyle modifications

Smoking cessation and cardiovascular exercise for more than 10 minutes more than 3 times per week is strongly recommended.

For patients with diabetes, the goal is to keep the fasting blood glucose level lower than 126 mg/dL.

Moderate alcohol intake has been shown to decrease stroke risk compared with excessive intake or none at all.⁸⁶

Carotid endarterectomy

Carotid endarterectomy has been recommended within 2 weeks of cerebral or retinal TIA in those cases attributable to high-grade internal carotid artery stenosis in patients who have low surgical risk.⁸⁷ This risk can be estimated on the basis of patient factors, surgeon factors, and hospital volume. The specific recommendations are as follows:

• 70% to 99% carotid stenosis: carotid endarterectomy recommended
• 50% to 69% carotid stenosis: carotid endarterectomy recommended in select patients with a perioperative complication rate < 6%
• < 50% carotid stenosis: carotid endarterectomy not routinely recommended.

Carotid artery angioplasty and stenting with distal embolic protection device

Data from the Carotid Revascularization Endarterectomy Versus Stenting Trial (CREST) and European stenting trials indicate that in patients over age 70, carotid endarterectomy appears to be superior to carotid artery stenting, whereas in younger patients the periprocedural risks of stroke and death are similar. Hence, carotid artery stenting performed by an interventionist with a low complication rate is a reasonable alternative to carotid endarterectomy.^88,89