BOSTON—A transient ischemic attack (TIA) often precedes stroke, but new research suggests that some patient groups have a high risk of recurrent TIA but a low risk of future stroke. By looking at the adverse events immediately following the diagnosis of TIA, researchers were able to differentiate patients who were likely to have a stroke from those who were likely to experience a second TIA. The two groups were surprisingly different, according to S. Claiborne Johnston, MD, MPH, Assistant Professor of Neurology at the University of California, San Francisco.

Not all patients who are diagnosed with TIA are the same, Dr. Johnston said at the 125th Annual Meeting of the American Neurological Association. Part of the problem lies in the recognition of TIA. "What some [clinicians] call a TIA isn't what others call a TIA," he said. Spells of temporary neurologic symptoms may be identical to those produced by other causes, such as seizure or migraine, he pointed out. According to Dr. Johnston, one clear way to identify a "true TIA" is to see that it was followed by a stroke soon afterwards.

In an attempt to categorize patients given the diagnosis of TIA by their outcomes, Dr. Johnston and his colleagues reviewed the adverse events—including stroke, recurrent TIA, cardiovascular events, and death—that occurred within 90 days following a diagnosis of TIA in 1,707 patients admitted to the emergency departments of the Northern California Kaiser Permanente system between March 1997 and February 1998. The median duration of a TIA was 70 minutes, and all the outcomes were reviewed by at least one neurologist, Dr. Johnston said. One fourth of the patients returned with an adverse event. About 10% had stroke and 13% had recurrent TIA. Less than 3% were hospitalized for subsequent cardiovascular events—including myocardial infarction, ventricular arrhythmia, unstable angina, and congestive heart failure—and 2.6% of the patients died. More than half of the events occurred within the first two days after their initial presentation, he pointed out.

WHICH PATIENTS CARRY A HIGH STROKE RISK?

A total of 180 patients had a stroke following their initial TIA presentation, Dr. Johnston said. These patients were likely to be older than age 60 and to have diabetes—and the TIA in their initial presentation usually lasted longer than 10 minutes. Weakness and speech difficulty with the spell were also associated with an increased risk of future stroke.

A total of 225 patients had recurrent TIA following their initial TIA presentation, Dr. Johnston said. The risk factors for recurrent TIA included age greater than 60, numbness at the time of evaluation, and TIA duration less than 10 minutes. Patients whose TIA lasted less than 10 minutes were more likely to have a recurrent TIA but less likely to have a stroke, he explained. A history of multiple TIAs, he pointed out, also increased the risk of another TIA. Patients with more than one TIA in the past were "more likely to show up again with a TIA, but not more likely to show up with a stroke."

A NEW CLINICAL ENTITY?

Because of these observations, Dr. Johnston proposed that a subgroup of patients diagnosed with recurrent TIA have different clinical characteristics, a lower risk of stroke, and that their TIAs are probably caused by something other than ischemia. Some of the risk factors, such as speech difficulty and diabetes, were clearly indicative of stroke but not of recurrent TIA, he reiterated. Likewise, multiple TIAs, numbness, and short-duration spells were risk factors for recurrent TIA but not for stroke. According to a small ad hoc analysis, a patient having four independent risk factors for recurrent TIA was associated with a "60% risk of going on to have a recurrent TIA, and none of those went on to have a stroke," Dr. Johnston said.

"I struggle with whether even to put a name on this," he said. There appears to be a subset of TIA patients who have a benign short-term course with brief, recurrent TIAs that are frequently characterized by sensory symptoms. These events, he continued, may be due to nonischemic events such as those related to migraine. In other words, the TIAs that lead to either stroke or recurrent TIA could represent different types of pathophysiologies, he proposed.

TIA REDEFINED

"The whole designation of TIA needs to be reworked," Dr. Johnston asserted. Currently, TIAs are defined as neurologic symptoms that are due to focal cerebral or retinal ischemia but last less than 24 hours. In practice, however, agreement between neurologists when they are evaluating the same patient is only about 25% above what would be expected by chance alone, he noted.

One problem is that the diagnosis of TIA is based on history alone, and histories aren't particularly reliable, Dr. Johnston said. Moreover, the manner in which clinicians interpret TIA symptoms are not completely reliable either. It is difficult, if not impossible, to differentiate an ischemic event from a seizure or migraine, he said. In this study, about 5% of the TIAs were thought to be either transient global amnesias or spells related to migraines; however, a few of these patients did go on to have a stroke, he said.

How important is the 24-hour time limit in the diagnosis of TIA? In this study, about 10% of the patients had no clear documentation that their symptoms were completely gone within 24 hours, Dr. Johnston pointed out. "If you think about it, that makes sense," he said. "The ER doctors aren't necessarily going to wait for all the symptoms to go away before they call it TIA."

Because of the results of this study, Dr. Johnston proposed that the frequency and outcome of TIAs is suggestive of distinct subsets of patients. But whether the clustering of benign events can represent a new clinical entity remains to be seen.

NR