The definition of a transient ischemic attack (TIA) has evolved over the past decade from a clinical diagnosis to a tissue-based definition based on neuroimaging results. TIA shares the same pathophysiology as stroke, which occurs in up to 5% of patients within 48 hours of the TIA and 10% within 90 days. This rate is decreasing, likely due to improved diagnostic and management strategies. Decision support scores have been developed to risk stratify patients, which include clinical and radiological elements. Antiplatelet and anticoagulant therapy, as well as carotid endarterectomy/stenting have been shown to reduce the stroke occurrence after TIA.
Key points
- •
Up to 50% of TIAs using the classic time-based definition show areas of infarction on brain MRI, hence the new definition of TIA is image-based.
- •
TIA shares the same pathophysiology as stroke, which occurs in up to 5% of patients within 48 hours of the TIA and 10% within 90 days.
- •
The use of clinical scores, such as the ABCD2 score, integrated with the clinical and radiological findings, help in stratifying the risk of stroke.
- •
Antiplatelet and anticoagulant therapies, as well as carotid endarterectomy/stenting, when indicated, reduce the risk of stroke after a TIA.
- •
The use of expedited assessment/management protocols for patients with TIA reduces stroke risk by up to 80%.
Introduction
Over the past decade, the diagnosis of a transient ischemic attack (TIA) has evolved from a clinical diagnosis to a tissue-based one based on neuroimaging results. Clinical scoring systems with or without neuroimaging has impacted decision making regarding timing and place of the diagnostic workup. Advances in pharmacotherapeutics have improved outcomes. Because up to 5% of TIAs will be followed by a stroke within 48 hours, facilitated evaluations should be tailored to the individual patient; these evaluations are often best accomplished through proactively designed protocols that provide patients with coordinated, multidisciplinary care. Many patients with TIA initially present to the emergency department (ED); consequently, knowledge of clinical presentations and diagnostic and management strategies is fundamental to improve the outcome of stroke. The class of evidence/graded recommendations, which are presented in this review, are based on the well-described methodology used by the American Stroke Association (ASA).
Introduction
Over the past decade, the diagnosis of a transient ischemic attack (TIA) has evolved from a clinical diagnosis to a tissue-based one based on neuroimaging results. Clinical scoring systems with or without neuroimaging has impacted decision making regarding timing and place of the diagnostic workup. Advances in pharmacotherapeutics have improved outcomes. Because up to 5% of TIAs will be followed by a stroke within 48 hours, facilitated evaluations should be tailored to the individual patient; these evaluations are often best accomplished through proactively designed protocols that provide patients with coordinated, multidisciplinary care. Many patients with TIA initially present to the emergency department (ED); consequently, knowledge of clinical presentations and diagnostic and management strategies is fundamental to improve the outcome of stroke. The class of evidence/graded recommendations, which are presented in this review, are based on the well-described methodology used by the American Stroke Association (ASA).
Definition
TIAs were defined in 1975 as “episodes of temporary and focal dysfunction of vascular origin, which are variable in duration, commonly lasting from 2 to 15 minutes, but occasionally lasting as long as a 24 hours; they leave no persistent neurologic deficit.” Advanced neuroimaging has shown that a subset of patients with TIA actually have evidence of clinically silent infarction. It is estimated that 30% of the events that were previously diagnosed as TIA actually have infarcted brain. Thus, the definition of TIA has evolved into a tissue-based definition: “transient episode of neurologic dysfunction caused by focal brain, spinal cord, or retinal ischemia, without acute infarction.”
The new definition of TIA is dependent on the availability of either head computed tomography (CT) or MRI. CT is not nearly as sensitive as MRI and may require 12 hours after the event to demonstrate injury; diffusion-weighted imaging MRI (DWI MRI) is far more sensitive and more specific for acute injury and turns positive much sooner. In the absence of imaging technology, it is accepted to use the old time-based definition of TIA, accepting that the probability of having a necrotic area is particularly high in patients with symptoms lasting more than 24 hours. This approach is recommended because, although most TIAs last less than 1 hour, up to 50% of patients with TIA lasting 1 to 24 hours have negative DWI MRI, confirming the absence of stroke.
The tissue-based definition of TIA has led to the creation of 2 stroke categories: the “nondisabling stroke,” which is the transitory presence of neurologic symptoms with mild (National Institutes of Health Stroke Scale ≤3) to absence of persistent clinical deficits in the presence of imaging evidence of necrosis, and the “silent stroke,” which is the radiologic finding of cerebral necrosis without neurologic findings. The importance of silent stroke is underscored by its high association with future strokes, cognitive sequelae, and mortality. There is ongoing discussion surrounding silent stroke and whether our current assessments are sensitive enough to identify subtle cognitive and behavioral deficits; there is a need to focus on this disease to better understand its pathophysiology and prognostic implications.
Epidemiology
In 2013, stroke caused 130,000 deaths in the United States, making it the fifth leading cause of mortality after heart disease, cancer, chronic lower respiratory disease, and accidents. The prevalence of stroke in the United States is 3% in patients older than 18 years, whereas the prevalence of silent stroke is estimated at 6% to 28% of the population, depending on age, sex, risk factors, and definition used.
TIA prevalence in the United States is estimated at 2.3%, with an incidence of 0.7 to 0.8 per 1000 people evaluated. The incidence of TIA increases with age and varies by sex and race/ethnicity. Men and African American and Mexican American individuals have higher rates of TIA than women and non-Hispanic white individuals. Approximately 15% of all strokes are preceded by a TIA; this rate has decreased over the past decade, possibly due to improved preventive strategies in patients at risk.
Early outcome studies on the risk of stroke after TIA estimated an incidence of 3% to 10% at 2 days and 9% to 17% at 90 days ; however, recent studies have reported lower rates in the range of 1.5% at 2 days, 1.4% to 3.4% at 90 days, 4.8% at 1 year, and 9.0% at 5 years. This lower incidence is most likely due to improvements in early diagnosis and management. The combined 10-year stroke, myocardial infarction, or vascular death risk in patients who have had a stroke is 43%. Most of the studies done so far, however, have used the old time-based definition of TIA; we expect that there will be a change in the epidemiologic estimates in the next years, when the tissue-based definition will become the standard.
Etiology and pathophysiology
Both TIA and ischemic stroke share the same etiologies and pathophysiology. The final pathway is focal hypoperfusion, oligemia, and impairment of cerebral oxygenation and glucose metabolism. An individual’s ability to tolerate short periods of cerebral hypoperfusion is variable and dependent on multiple factors, including collateral flow and oxygen delivery capacity.
Several systems have been developed to classify stroke etiology ( Table 1 ; often referred to as the TOAST classification). The importance of a standardized classification system is to achieve a framework for study design and outcome analyses that are appropriate for the type of disease process, for example, cardioembolic disease may respond differently to interventions than small vessel occlusion disease.
| Etiology | Example |
|---|---|
| Large artery atherosclerosis | Cervical artery (carotid or vertebral) stenosis; aortic atherosclerosis |
| Cardioembolic | Cardiac dysrhythmia (atrial fibrillation); cardiac thrombosis; valve disease, patent foramen ovale; impaired left ventricular function (such as recent myocardial infarction), dilated cardiomyopathies, infective endocarditis, mechanical valvular prostheses, septal aneurysm, myxomas |
| Small artery occlusion | Intracranial atherosclerosis (hypertension and diabetes) |
| Other defined mechanisms | Hypercoagulable states (eg, protein C disease, inherited thrombophilias, protein S deficiency, antithrombin III deficiency, plasminogen deficiency, activated protein C resistance/factor V Leiden mutation, anticardiolipin antibodies, lupus anticoagulant, ptothrombin mutation, thrombocytosis), arterial dissection, vasculitis, and so forth |
| Undetermined causes |
Cardioembolic etiologies account for approximately 34% of TIAs, followed by small artery occlusion/disease (10%–18%), large artery atherothrombosis (9%–13%), and other causes (3%–6%). These causes must be put in the context of the patient, in which case the percentile may change; for example, in young patients without underlying disease, cervical artery dissection emerges as one of the most common causes of TIA. Nonetheless, in all classification schemes, the proportion of undetermined or cryptogenic causes remains as high as 25% to 50%.
Atheromatous disease of large-sized to medium-sized arteries is often related to hypertension. It usually forms at branch points or in tortuous vessels where endothelial damage can progress to smooth muscle proliferation and fibrolipid plaque formation. Common sites of plaque formation are the aortic arch, proximal subclavian arteries, carotid bifurcation, and vertebral artery origins. An ulcerated plaque can result in emboli or it can directly propagate and obstruct flow to a branching vessel; it can also result in local stenosis to the point of reducing flow to susceptible tissue or to the point of complete occlusion. Individuals with large artery atherosclerosis have a high all-cause mortality at 2 years.
Small vessel disease refers to hyaline arteriosclerosis (lipohyalinosis, which is pathologically distinct from extracranial atherothromboembolic disease) that forms in small penetrating arteries (<0.5 mm in thickness) within the brain. It results in small-volume and lacunar strokes, often at the lenticulo-striate branches of the middle cerebral artery and perforating vessels of the anterior and posterior cerebral arteries. Intracranial vessels are less elastic than their extracranial parent vessels, and they are more prone to narrowing from collagen deposition that occurs with aging. Cortical lacunar strokes and TIAs may present with minor deficits or may be clinically silent.
Cardioembolism has been associated with severely debilitating strokes, and is an independent predictor of mortality. Atrial fibrillation is the most common cause. It has a highly variable average annual risk of first-time stroke in patients who are not anticoagulated, ranging from 0.2% to 23.6%, depending on the clinical characteristic, as defined by CHA 2 DS 2 -VASc criteria ( Table 2 ). This high risk of stroke underscores the importance of expedited management of new-onset atrial fibrillation, although, as discussed in the therapy section, there are limited data on the 1-week and 1-month stroke risk in newly diagnosed atrial fibrillation.
| Comorbidity | Value |
|---|---|
| Congestive heart failure | 1 |
| Hypertension | 1 |
| Age >75 | 2 |
| Diabetes | 1 |
| Prior TIA, stroke, or VTE | 2 |
| Vascular disease (PAD, MI, aortic plaque) | 1 |
| Age 65–74 | 1 |
| Female sex | 1 |
Diagnosis
Differentiating TIAs from alternative diagnoses that cause focal neurologic symptoms is a challenge in emergency medicine. Considerations of high priority include hypoglycemia, stroke, central nervous system (CNS) mass lesions, CNS vasculitis, and CNS infections ( Table 3 ). Although rare, hypokalemia has been reported in case reports to present with focal deficits.
| Diagnosis | Common Characteristics |
|---|---|
| Emergent diagnosis | |
| Transient ischemic attack | Vascular/cardioembolic risk factors. More frequent negative symptoms (numbness, weakness, visual loss), diplopia, does not spread into other sensory modalities. Alteration/loss of consciousness (LOC) rare. Abrupt onset, lasting often <1 h. Headaches may occur, usually during attacks. |
| Ischemic or hemorrhagic stroke | Evidence of cerebral necrosis or bleeding on imaging. In hemorrhagic stroke, sudden onset of severe headache, can have sentinel events in the previous days. |
| Hypoglycemia | History of diabetes mellitus, hypoglycemic medications. Negative symptoms, decreased level of consciousness. Low serum glucose. |
| Hypertensive encephalopathy | History of uncontrolled hypertension, recent suspension of medications. Headache, delirium, cortical blindness, seizure, cerebral edema. Significant hypertension. |
| Central nervous system infections | Immunocompromised patient, intravenous drug use, persistent/worsening symptoms, can have fever, headache, meningismus, increased inflammatory markers, electroencephalographic/radiologic findings. |
| Cerebral venous thrombosis | Procoagulant risk factors; for example, pregnancy; headache, nausea/vomiting, can develop seizures and altered level of consciousness. |
| Chronic or subacute subdural hematoma | Common history of previous head trauma, elderly patients, antiplatelet/anticoagulant medications, subacute onset, and persistent/worsening symptoms. |
| Urgent diagnosis | |
| Central nervous system mass lesions | History of neoplasia (eg, breast or lung neoplasia, melanoma), persistent/worsening symptoms over days, can be associated with seizure/have positive symptoms, alteration on cerebral imaging. |
| Central nervous system vasculitis | Usually young, persistent/worsening negative symptoms, can be associated with headache, can be multifocal, alteration on MRI. |
| Central nervous system inflammatory disease | History of multiple sclerosis, previous episodes, usually young, multiple alterations on MRI. |
| System infections | Elderly, frail patients, variable and fluctuant neurologic findings often associated with confusion/delirium, fever, increased inflammatory markers. This may represent “unmasking” of a prior stroke that is asymptomatic under normal circumstances. |
| Seizure | History of seizure, witnessed seizure activity. Positive symptoms (pain, limb jerking, lip smacking), frequent loss of awareness and amnesia, symptoms usually brief but Todd paresis may persist for hours). Incontinence, tongue bite, muscle pain, postictal period. |
| Less urgent diagnosis | |
| Migraine | History of similar events, visual symptoms most common, symptoms may spread into other sensory modalities, headache follows symptoms, frequent nausea/vomiting, alteration/LOC rare, confusion possible. |
| Transient global amnesia | Paroxysmal, transient loss of memory. Remote and immediate memory spared, attention and visual-spatial skills intact. No neurologic signs on physical examination. |
| Psychogenic (conversion disorder, somatization, hyperventilation) | Usually young, emotional stress. Most frequent isolated sensory symptoms in nonvascular distribution, recurrent. Can have dyspnea, anxiety. In hyperventilation, peripheral bilateral tingling and cramps. Lack of neurologic signs. Diagnosis is rarely made in the emergency department. |
| Syncope | Previous episodes, cardiologic risk factors, emotional stress. Faint, dizziness, light head (presyncope), loss of awareness, blurred vision/may darken, muffled hearing. Lasting seconds to <1 min. Rapid recovery to full alertness, pallor, sweating, palpitation, chest pain, nausea, dyspnea. |
| Acute vestibular syndrome due to peripheral cause | Variable onset of vertigo with fluctuating symptoms, associated with nausea/vomiting, gait impairment, horizontal/torsional nystagmus, absence of dysmetria/neurologic findings, Head Impulse Test positive, test of Skew negative. |
| Peripheral nervous system lesion/compression | Subacute onset, weakness and numbness often preceded by pain/paresthesia, previous episodes with same characteristic, radicular/single nerve distribution, can be persistent. Alteration on medullar MRI, electromyography/electroneurography. |
| Drug toxicity (Lithium, phenytoin, carbamazepine) | History of medication use, tingling/numbness with nonvascular distribution. |
History
TIA almost always produces focal symptoms that can be attributed to a specific vessel. The history should characterize the event, as in most cases the symptoms have resolved by the time the patient is seen. Indeed, if the symptoms have not resolved, TIA cannot be diagnosed and the patient should be assumed to have a stroke, or a stroke mimic. The history focuses on the positive or negative features of the symptoms (see the following paragraph), the outset and time course, comorbidities, risk factors, and medications. Nonfocal symptoms, such as loss of consciousness, confusion, lightheadedness, generalized weakness, or incontinence, are much less often caused by TIAs and tend not to be predictive of future stroke.
The symptoms experienced in a that TIA are predominantly “negative”, are associated with a loss of a function, such as motor (weakness), speech (decreased or altered speech), visual (diminished vision), or sensation (anesthesia). “Positive” symptoms are associated with the presence of something that is not normal, such as motor (involuntary motions), speech (increased volume of incomprehensible speech), visual (flashes of light or scintillating scotoma), or sensation (dysesthesias), and sometimes pain. As a general rule, negative symptoms suggest ischemia or infarction, whereas positive symptoms suggest migraine or seizure-related diagnoses. An important exception is “dizziness,” which, at least in the case of vertigo, is a positive symptom and can be due to posterior circulation ischemia; in particular, even without any other neurologic sign, it has been associated with TIA or stroke in 0.7% of dizzy patients. Taken from the other perspective, in a series of 1141 cases of stroke, a preceding episode of isolated transient vertigo was found in 2% of patients (and 8% in the 275 patients with posterior circulation strokes).
TIA generally occurs suddenly and without prodrome. Stuttering symptoms, over hours to days, are of particular concern and warn of a highly unstable plaque. Although the component of time has been removed from the definition of TIA, the duration of symptoms remains an important aspect of the diagnosis. Most TIAs are brief, most lasting less than 1 hour ; in up to 50% of patients with symptoms lasting more than 1 hour, radiologic evidence of infarction can be found.
Physical Examination
By definition, the physical examination is expected to be normal in a patient with a TIA; the presence of neurologic deficits should be treated as a stroke until proven otherwise. The physical examination must be systematic and comprehensive, addressing strength, sensation, coordination, balance, vision, and cognition. The examination assesses both the neurologic function and also assesses findings that might be associated with TIA risk, including irregular heart rate and carotid bruit, or alternative diagnoses, such as hyperventilation or presence of a heart murmur.
The use of scales, such as the National Institute of Health Stroke Scale (NIHSS) are useful only if the patient has an ongoing deficit; in a patient with TIA, the score will be zero. The NIHSS is of less value in assessing small posterior lesions and lesions on the nondominant side of the brain. Truncal ataxia, decreased visual acuity, Horner syndrome, and memory impairment are just a few focal neurologic deficits that would be missed by the NIHSS. For all these reasons, a full neurologic examination is required before diagnosing a TIA.
Laboratory Tests
There are no laboratory tests that are diagnostic for TIA but there are several tests needed to assess risk for TIA, and to eliminate other processes that may mimic TIA. Serum blood sugar and electrolytes are indicated in all patients. A complete blood count with platelet count evaluates for inflammation, thrombocytosis, and myelodysplastic disease. Evaluation of prothrombin time and international normalized ratio (INR) is useful in the evaluation of patients that are taking warfarin or who have liver dysfunction. Although the 2013 ASA guidelines recommend a troponin level in patients with stroke, its value in the assessment of TIA is not identified.
Other laboratory tests are performed on a case-by-case basis related to pretest risk assessment and suspected etiology of the TIA (eg, pregnancy test, toxicology screening). Diagnostic and prognostic TIA markers have been suggested; for example, C-reactive protein, S100B; however, at this time, none have been identified as useful during clinical practice.
Electrocardiogram and Cardiac Monitoring
An electrocardiogram (ECG) is recommended to assess for a cardioembolic mechanism, including atrial fibrillation, ventricular hypertrophy, or signs of cardiac ischemia (Class of evidence IB). The 2013 stroke American Heart Association (AHA) guidelines recommend 24-hour telemetry to investigate for paroxysmal atrial fibrillation; this time limit can be extended to up to 30 days in case of suspected embolic or cryptogenic cause. Trials have reported paroxysmal atrial fibrillation during 21-day telemetry monitoring in up to 23% of patients thought to have a cryptogenic etiology.
Cardiac Echocardiography
Transthoracic echocardiography (TTE) is used to assess for evidence of cardiac hypertrophy, ventricular hypokinesis or thrombus, mitral stenosis, endocarditis, and valve disease (Class of evidence IIaA). In one retrospective study of 186 individuals with stroke or TIA without obvious cause of cardioembolism, TTE detected a source of embolism in 19% of patients. The transesophageal echocardiography is particularly useful as a secondary examination in the cohort of patients thought to have an embolic cause that is hard to see on the transthoracic examination, such as left atrial appendage thrombosis or patent foramen ovale.
Brain Imaging
The 2013 ASA guidelines state that patients with transient ischemic neurologic symptoms should undergo neuroimaging evaluation within 24 hours of symptom onset or as soon as possible in patients with delayed presentations using MRI (including diffusion-weighted imaging) as the preferred brain diagnostic imaging modality. If MRI is not available, head CT should be performed. (Class of evidence IB). The sensitivity of CT scan in identifying stroke within 12 hours of symptom onset is only 0.39, while the sensitivity of DWI MRI is 0.99. The availability of an MRI is therefore of great importance during the first hours to evaluate for cerebral infarction and hence defining the event as stroke versus TIA ( Fig. 1 ).
The perfusion (for CT and MRI) and fluid-attenuated inversion recovery (for MRI) techniques are helpful to further increase the specificity and sensitivity of MRI and CT and, in the case of a stroke, can identify the area of the brain eventually salvageable by reperfusion therapies (ie, the “ischemic penumbra”).
Magnetic resonance angiography (MRA) or CT angiography (CTA) is recommended to assess the intracranial vessels for intracranial stenosis and guides the need for endovascular interventions (Class of evidence IA). Angiography is indicated when intracranial stenosis is identified.
Cervical Vessel Imaging
The evaluation of patient with suspected TIA includes the noninvasive imaging of cervical vessels (Class of evidence IA) using Doppler-ultrasound, MRA, or CTA ( Fig. 2 , Table 4 ).
| Imaging Technique | Sens and Spec for Detecting Stenosis >70% | Limitations | Cost |
|---|---|---|---|
| Duplex Doppler | Sens 83%–86%, Spec 87%–99% | Limited ability to image vasculature proximal and distal to bifurcation; operator-dependent | Low |
| CTA | Sens 83%, Spec 93% | Radiation exposure Contrast administration | Medium |
| MRA | Sens 93%, Spec 88% Sens 86%–97%, Spec 62%–91% if contrast enhanced | Limited availability, contraindicated in patients with some forms of indwelling hardware, time-consuming | High |
Risk stratification
Many instruments have been developed for risk stratification, including the Stroke Prognosis Instrument, the California score and the ABCD score, and more recently the ABCD2 score, the ABCD2-I score, and the ABCD3-I score. The most used score at this time is the ABCD2 score ( Table 5 ), which stratifies patients as low, moderate, or high risk for stroke at 2, 7, and 90 days following TIA, in an effort to guide appropriate triage and disposition. Despite widespread acceptance of these scores and their incorporation into guidelines, a meta-analysis of the literature failed to find sufficient evidence to validate this instrument. Limitations of the ABCD2 score include the finding that up to 41% of the patients with a high score (>4) have been found to have TIA mimics, and at the same time up to 21% of patients with low score have a high-risk etiology, such as atrial fibrillation or carotid stenosis. Another limitation is that the clinical measurements favor identifying anterior circulation strokes and miss posterior circulation findings, such as dizziness or sensory loss. Regardless, the ABCD2 score does provide a framework for approaching clinical decision making in these patients.
