THE EMERGENT PATIENT: Transient Ischemic Attack

Emerg Med 38(3):41-46, 2006

Newer technologies such as diffusion-weighted magnetic resonance imaging have helped to refine the classic definition of transient ischemic attack. The authors discuss the evolution of the diagnosis to date, the pathophysiology of the phenomenon, the diagnostic pitfalls, and the strategic considerations in treatment and patient disposition.

By Nicholas Hogan, MD, and Ioliene Boenau, MD

Transient ischemic attack (TIA) is a common marker of cerebrovascular disease which, when correctly diagnosed and managed, can lead to the prevention of significant morbidity and mortality. It is estimated that more than 300,000 TIAs occur each year in the United States. Prevalence among men and women aged 65 to 69 is 2.7% and 1.6%, respectively; between ages 75 and 79, it is 3.6% and 4.1%, respectively. Many of these cases of TIA go undiagnosed and become missed opportunities for the prevention of permanent disability.

The overall incidence of an adverse vascular outcome within 90 days of a TIA is 25%. The likelihood of a stroke during this same time period is 11%, with half of these strokes occurring within 48 hours. The vast majority of these events are amenable to treatment that can significantly reduce the likelihood of subsequent strokes.

In this article, we will discuss the definition and pathophysiology of TIA. We will also review appropriate diagnostic and therapeutic strategies and patient disposition.

DEFINING TIA

The definition of TIA continues to evolve. Traditionally, TIA has been defined as the presence of neurologic symptoms of vascular etiology in one area of the brain lasting less than 24 hours. Implicit in this definition is neuronal injury that is not permanent in nature. However, there are several problems with this definition. It is now clear through newer diagnostic modalities such as diffusion-weighted magnetic resonance imaging (MRI) that this time- and symptom-based definition can be clinically misleading. It assumes a complete correlation between the resolution of symptoms and normalization of tissue reperfusion, which often underestimates the potential tissue damage detected by diffusion-weighted MRI. Nearly 50% of patients who meet the classic definition of TIA have in fact suffered subclinical strokes with detectable cerebrovascular infarction.

Transient ischemic attack and stroke lie on opposite ends of a continuum of cerebrovascular ischemia. A tissue-based definition of these conditions has more clinical meaning than a time- and symptom-based definition. Just as patients with permanent perfusion defects of the myocardium are considered to have sustained a myocardial infarction (MI) and not merely angina, those with permanent tissue damage in the brain should be classified as having had a stroke and not a TIA. Moreover, TIAs that do not result in permanent tissue damage produce symptoms that usually resolve within an hour, rather than 24 hours.

A more accurate definition of TIA has been proposed by the Transient Ischemic Attack Working Group formed by Albers and Caplan: Òa brief episode of neurologic dysfunction caused by focal brain or retinal ischemia, with clinical symptoms typically lasting less than one hour, and without evidence of acute infarction.Ó

Although most institutions do not have diffusion-weighted MRI available, an awareness that the patient with transient neurologic deficits may have infarcted brain parenchyma should increase diagnostic vigilance, early workup, and aggressive treatment. Furthermore, given the diagnostic difficulty that TIA poses, some clinicians find it useful to add qualifiers like Òdefinite,Ó Òprobable,Ó or ÒpossibleÓ TIA to prevent overdiagnosis, especially when uncertainty exists. However, no studies have yet validated the use of such qualifiers.

PATHOPHYSIOLOGY OF TIA

To function normally, the adult human brain needs about 50 ml/100 gm/min of blood flow. Once blood flow is reduced to 50% of that rate or lower, neurologic symptoms will develop.

Transient ischemic attacks result from such a decrease in cerebrovascular blood flow, which disrupts neuronal function temporarily to the extent that clinical deficits can be detected. Blood flow is critically dependent on even small reductions in vessel diameter from vasoconstriction or atherosclerotic plaque. Less common causes of pathologically reduced blood flow include hypercoagulable states, arterial dissection, vasculitis, or vasospasm from drugs such as cocaine. More commonly, TIAs result from one of three etiologic pathways: atherosclerosis leading to luminal narrowing of vessels and plaque ulceration; thromboembolism; or a cardioembolic event.

Large arteries, such as the common carotid arteries, are more susceptible to atherosclerotic plaque. Turbulent, high-pressure blood flow (from hypertension, for example) can cause endothelial cell dysfunction and disrupt the normal, nonthrombogenic, vascular environment. Injured endothelium allows intimate contact between the blood and prothrombotic subendothelial connective tissue, triggering a thrombogenic cascade. Platelets are activated; monocytes adhere and emigrate into the intima, enhancing lipid accumulation and stimulating inflammatory mediators. Platelet activation causes release of adenosine diphosphate (ADP) and thromboxane A2, which fuel platelet aggregation.

Simultaneously, the coagulation cascade is initiated by the production of cytokines from dysfunctional endothelium, serving to reinforce the platelet plug with a fibrin meshwork. This clotting pathway lays the groundwork for plaque ulceration and thromboembolism. The resultant plaque can rupture, exposing highly thrombogenic substances, which may encourage thrombus formation or the release of emboli downstream.

Myocardial infarction and cardiomyopathy predispose to thrombus formation in the left ventricle, as can rheumatic heart disease in the atria. Subsequent arrhythmias such as atrial fibrillation can convert these thrombi to emboli. Cardiac thrombi give rise to the majority of arterial emboli. In fact, 15% of ischemic strokes are cardioembolic, underscoring the importance of cardiac disease in the pathology of TIA.

DIAGNOSING TIA

Consumer and physician education with regard to cerebrovascular events has been much less successful than that regarding acute coronary syndrome. It has been estimated that only 9% of the general population is familiar with typical TIA symptoms, and as few as 22% of primary care physicians even know the definition of a TIA. This education gap poses a great challenge for physicians when eliciting a history from patients and families.

It is not surprising then that much variability exists in the diagnostic accuracy of TIA, with either overdiagnosis or underdiagnosis common in practice. The implications of missing a TIA are great, but so are those of mislabeling a benign illness as a TIA. A TIA diagnosis pins a risky prognosis on the patient and commits him or her to a treatment pathway. However, keeping in mind certain key principles regarding pitfalls and risk factors, as well as essential historical points (see box above) and clinical features (see box below), can be helpful in improving diagnostic accuracy.

Mimics of TIA present potential pitfalls to avoid. Migraine with aura, seizure, syncope, BellÕs palsy, hypoglycemia, drug toxicity, electrolyte and metabolic abnormalities, and mass lesions all share common symptoms such as brief neurologic deficits or paresthesias.

Identifying risk factors is crucial for increasing diagnostic efficiency. A scoring system has been suggested that seems highly predictive for stroke within seven days after TIA. Called the ÒABCDÓ system, it focuses on four risk factors: Age, Blood pressure, Clinical features, and Duration. Each is assigned point values, as follows: age 60 or older = 1, younger than 60 = 0; systolic blood pressure above 140 mm Hg and/or diastolic 90 mm Hg or higher = 1; clinical features of unilateral weakness = 2; speech disturbance without weakness = 1, other = 0; duration of symptoms 60 minutes or more = 2, 10 to 59 minutes = 1, and less than 10 minutes = 0. The points for each category are then added (maximum = 6) and a risk assigned. A total of 0 to 3 points suggests a very small risk of stroke within seven days; 4 points, a 2% risk; 5 points, 16%; and 6 points, 36%.

Other possible risk factors to consider include previous stroke, cardiac disease, carotid artery stenosis, hypercoagulable states, and a previous diagnosis of hypertension. Smoking, diabetes, hypercholesterolemia, obesity, excessive alcohol consumption, and physical inactivity are also important risk factors.

In general, negative symptoms are more likely to represent TIAÑthat is, numbness rather than tingling or pain, weakness rather than a shaking limb, visual field cuts instead of scotoma, and unilateral neglect (lack of perception or awareness of objects on the side contralateral to the cerebral ischemia) instead of agitation, are more suggestive. Most symptoms last only about 10 minutes. With such a quick resolution of symptoms, it is very easy for the individual to believe that nothing serious happened and to delay seeking treatment.

The physical examination should concentrate on the neurologic system: cranial nerves, motor and sensory function, coordination, speech and language, and cognition. Although the history will generally be most helpful, correlating the physical deficit with the affected focus in the brain can also be useful in the diagnosis and may suggest an etiology. In a patient with a previous TIA, especially a recent one, involvement of different areas of the brain in the current episode suggests emboli rather than primary thrombosis. Mapping the clinical deficit can help interpret the significance of lesions found on computed tomography (CT) or MRI.

LABORATORY AND ANCILLARY DATA

Immediate evaluation of the patient should focus on ruling out TIA mimics and establishing the diagnosis of TIA. Assuming the patientÕs condition is stable, noncontrast head CT, ECG with cardiac monitoring, and glucose levels are top priorities. A complete blood cell count with platelets, blood urea nitrogen and creatinine, and electrolytes are also helpful. Other laboratory tests and studies should be guided by clinical suspicionÑfor example, prothrombin and partial thromboplastin times and an INR for patients receiving warfarin and a urine drug screen and the fluorescent treponemal antibody absorption test for patients with suspected tertiary syphilis. Elevated homocysteine levels are associated with cerebrovascular disease, but evidence is lacking for the efficacy of treating high levels.

Unless the etiology is known, urgent evaluation with carotid ultrasound is necessary. Carotid duplex ultrasonography evaluates for large artery stenosis, which may indicate a possible source of emboli. Transesophageal echocardiography (TEE) is useful in evaluating for an akinetic left ventricle, mural thrombus, atrial septal defect, patent foramen ovale, and aortic plaque, all of which may contribute to discharging emboli.

Other studies to consider are magnetic resonance angiography and CT angiography. Both of these noninvasive modalities are superior to carotid ultrasound in evaluating extracranial artery stenosis. Magnetic resonance angiography offers the additional benefit of identifying vertebrobasilar stenosis. Computed tomography angiography is useful in finding carotid or vertebral artery dissection. Cerebral angiography remains the gold standard for evaluating extracranial and intracranial vessel pathology. However, cerebral angiography is an invasive study that carries a significant risk of neurologic complications. One of its indications is suspected arterial dissection in spite of negative noninvasive studies.

TREATMENTS FOR TIA

The reason for the rampant undertreatment of patients with TIA is unclear. Up to 40% of patients either are not started on antiplatelet therapy or do not have their therapy augmented after a TIA. The three main treatment regimens are antiplatelet therapy, anticoagulation, and carotid endarterectomy.

Antiplatelet therapy. The drug of choice for antiplatelet therapy remains aspirin, which inhibits the production of thromboxane, thereby decreasing platelet aggregation. Aspirin alone reduces the risk of stroke by 17% over placebo. There exist no data proving the efficacy of one dose over another; the recommended dose is between 50 and 325 mg daily.

Clopidogrel therapy is an alternative for patients who cannot take aspirin. This drug inhibits platelet aggregation through its antagonistic action at the ADP receptor. In the CAPRIE trial (Clopidogrel versus Aspirin in Patients at Risk of Ischemic Events), clopidogrel at a daily dose of 75 mg had a relative risk reduction of 9% for ischemic events compared to aspirin.

Notably, this trial did not include TIA patients. Overall, clopidogrel has not been shown to be superior to aspirin in TIA patients. Aspirin is much cheaper and equally safe, so it remains the first-line recommended therapy in most cases.

A rare adverse effect of clopidogrel is reversible neutropenia.

At 500 mg a day, ticlopidine is another option. Like clopidogrel, this drug inhibits platelets by disrupting the ADP pathway. It demonstrated a 12% relative risk reduction in nonfatal stroke and death compared to aspirin in the TASS (Ticlopidine Aspirin Stroke Study) trial. Side effects include rare instances of neutropenia; also, 1 in 1600 to 5000 patients develops thrombotic thrombocytopenic purpura (TTP). These risks overshadow its modest advantages over the other antiplatelet therapies.

Dipyridamole disrupts platelet function by inhibiting phosphodiesterase activity, which increases cyclic adenosine monophosphate. This drug resulted in a risk reduction of 15% for stroke and death that was in addition to that of aspirin when the two were combined. In the ESPS-2 (European Stroke Prevention Study) trial, the combination of the two drugs (Aggrenox) offered a 23% relative risk reduction over aspirin alone in patients with previous TIA or stroke. This combination therapy is a reasonable option for patients who are on antiplatelet therapy and have had a TIA. The recommended dosage for extended-release dipyridamole and aspirin is 200 mg and 25 mg, respectively, twice daily.

Headache, the chief side effect of dipyridamole, occurs in about 39% of patients.

Anticoagulation therapy. Anticoagulation therapy does not offer additional benefits over antiplatelet therapy in the absence of specific indications. Also, it carries the risk of more bleeding complications than aspirin.

Warfarin inhibits vitamin K epoxide reductase. In so doing, it interrupts the activation of vitamin K-dependent clotting factors II, VII, IX, and X. The anticoagulant effect can last 8 to 12 hours. Anticoagulation with warfarin is appropriate in patients with known or highly suspected cardioembolic sources such as atrial fibrillation, ventricular aneurysm, or recent MI. In patients with atrial fibrillation, warfarin therapy reduces the risk of stroke by 68%. Except possibly in certain cases of atrial fibrillation, bridging therapy with heparin to achieving a therapeutic INR is controversial. (An INR of 2.5 is the therapeutic goal.)

Carotid endarterectomy. This procedure is beneficial in patients with more than 70% stenosis without near-occlusion. In these patients, the procedure showed a 16% relative risk reduction in ipsilateral stroke over five years. If patients with near-occlusion are included, the relative risk reduction approaches 65%.

The greatest benefits are reaped when carotid endarterectomy is performed within two weeks of a TIA. Additionally, the benefits decrease as the degree of blockage decreases. The benefit over medical therapy in moderate carotid disease (50% to 69% stenosis) is small. Careful patient selection is advised.

PATIENT DISPOSITION

Appropriate patient disposition requires rapid identification or exclusion of causes of TIA that are amenable to therapy and identification of high-risk patients. Admission rates vary greatly, from less than 10% to more than 50% in various practice settings. Lower rates most likely reflect the belief that an outpatient workup is safe. This view has been recently challenged, however. It has been found that 5% of TIA patients will have a stroke within 48 hours. Some of these strokes may be prevented if the cause is uncovered and treated expeditiously. Patients and their families should be told about this high-risk time frame, regardless of disposition.

The advantages of an inpatient workup for TIA include rapid evaluation and treatment and patient education. In most practice settings, getting a TEE, carotid ultrasound, or other studies within 24 hours requires inpatient resources. Admission also facilitates timely consultations. Patients and their families need to be educated about risk factor modifications, symptom awareness, and treatment modalities. The cost-effectiveness of an inpatient evaluation is unproven; however, it is well recognized that preventing disability from a stroke can save tens of thousands of dollars.

Risk stratification for subsequent stroke is another priority. JohnsonÕs high-risk criteria (see box) are similar to the previous ABCD risk criteria, with the addition of diabetes. In patients with all five high-risk criteria, the risk for stroke is 1 in 3 within 90 days; in those who do not have any of these criteria, the risk is minimal. Other high-risk patients requiring an expedited workup include those with a suspected cardioembolic source, carotid artery disease, or arterial dissection, or a first TIA, acute TIA (less than 48 hours from onset), crescendo TIA, TIA in spite of maximal medical therapy, or abnormal neuroimaging results.

No studies have identified a patient population that is safe for an outpatient evaluation. Therefore, unless the necessary workup can be accomplished in the emergency department, observation unit, or a stroke clinic, admission should be the rule.

back to top

Printable version of this article

E-mail this article