Transient Ischemic Attack: A Threat or an Attack?

Ominous fleeting symptoms of loss of vision, numbness, or paralysis in the territory of carotid arteries were considered by Miller Fisher to be warning episodes preceding strokes.[1] These he dubbed transient monocular blindness and, by default, transient ischemic attack (TIA), a term that took hold and persisted in the front trenches of cerebrovascular clinical science until today. The condition is not inconsequential because estimates of the annual incidence of TIAs that come to medical attention vary from 200,000 to 500,000 in the United States alone.[2] When Fisher coined the term in 1951 [3], clinicians had to rely on duration of symptoms to draw the line between stroke and TIA, and thus, the definition of "a focal neurological event of vascular origin lasting 24 hours or less" was born. The time limit was reinforced by discussions at the Fourth Princeton Conference in 1965 [4] and finally adopted by the National Institutes of Health (NIH) in 1975.[5] Advances in neuroimaging eventually demonstrated the weakness of the 24-hour limit [6], when ischemic lesions of varying extent and location were repeatedly shown in as many as 20% of brain computed tomography scans (CTs) scans and magnetic resonance images (MRIs) of patients whose condition had been diagnosed as TIA. Furthermore, there was the additional complication of availability of treatment for acute ischemic stroke that had to be applied within a time window far shorter than the 24-hour limit. The subsequent clinical clamor culminated in a new definition of TIA published by an ad hoc group of clinicians in 2002.[7] The innovation essentially reduced the 24-hour duration to a new limit of 1 hour based on the knowledge that the majority of TIAs resolve within 60 minutes, and most of these resolve within 30 minutes.[8,9] Soon, it was clear that arguments against the old, time-constrained definition were also applicable here. The response has been a new scientific statement drawn by an American Heart Association (AHA) working group [10] with a revised definition deleting the time limit and relying on a tissue rather than time criterion. The proposed new definition is as follows: "Transient ischemic attack (TIA): a transient episode of neurological dysfunction caused by focal brain, spinal cord, or retinal ischemia, without acute infarction."

Like Philippe Pinel [11], who unchained the insane in the nineteenth century and opened a new era of treatment in psychiatry, the unchaining of TIA to its time limit opens new frontiers in TIA management, but like Pinel, the AHA working group does not assert that restraints could be abandoned. The proposed new definition does away with specific time durations but retains the word transient. The difference now is that clinicians will have to use their clinical judgment to determine how many seconds, minutes, or hours transient means.

Also, the new definition relies heavily on neuroimaging to exclude acute infarction. CT imaging of the brain and, even more so, MR imaging of the brain are luxurious diagnostic tools for half of the planet, and therefore, the definition is not international-sensitive. The AHA working group is cognizant of the fact that stroke and TIA incidence rates will differ depending on whether and when detailed imaging studies are performed. This notion is illustrated by incidence rates coming from Portugal, where the crude overall annual incidence of TIA per 1,000 population was found to be higher in the rural region at 0.96 than in the urban area at 0.61 [12], suggesting reduced availability of MRI resources outside the towns. How can we do comparable research around the world if the incidence of TIA will vary depending on resources? Perhaps we should de-emphasize neuroimaging from our expectations to separate TIA from stroke.

On the other hand, imaging techniques, when available, should be applied immediately because, even if they do not always help in distinguishing TIA from stroke, they can assist in the search of the cause of TIA. Furthermore, neuroimaging can accurately identify intracranial hemorrhage, tumors, and other disorders that cause transient symptoms and are misdiagnosed or masquerade as TIA. The corollary of lowering the diagnostic expectations from neuroimaging is to exclude from the new definition the tail sentence "without acute infarction." The criterion separating TIA from stroke is the evidence of tissue injury that is best shown by MRI. There will never be certainty about the exclusion of brain injury in patients with transient symptoms using the current diagnostic tools. We know that strokes with brain infarction can be associated with focal symptoms of any duration, from 0 (silent strokes) to infinity. The big question is whether all TIAs are in effect mini-ministrokes that were not detected by our, as yet, insufficient diagnostic instruments. It is plausible that as new diagnostic tools evolve, clinicians in advantaged centers will have increasing availability to identify smaller, more subtle, and hidden ischemic brain injuries that will incorporate more ischemic focal brain symptoms into the camp of strokes. On the other hand, could it be that there are indeed focal events genuinely caused by functional disturbances that will never evolve into a stroke? These would be the original, authentic, genuine, bona fide, unmistakable TIAs. I am confident that eventually, we will have at our disposal diagnostic tools capable of identifying functional ischemic injuries without structural lesion that can be diagnosed with certainty as TIAs, but this will be in the future.

We know that TIAs and strokes share the same prognosis, management, pharmacologic, and surgical treatment. In fact, the modern emphasis is to assert that a TIA should be considered as seriously as a stroke, that TIA patients should be hospitalized immediately, and that they should receive the entire neurovascular work-up. Clinicians are being told to adopt the philosophy of the Secret Service: a threat is as dangerous as an attack. Importantly, following a TIA, there is still time to act expeditiously to forestall the stroke that is coming in 20% of instances.[13,14] These considerations lead to the thought that there is no clinical practicality in separating TIAs from strokes, and that we have been struggling in vain by attempting to find the magic time frame that separates TIA from stroke. I sense that this was well understood by both working groups, but if this were the case, why retain the word transient in the definition? Is it loyalty, inertia, or a suspicion that authentic TIAs do really exist?

The severity and the duration of cerebral hypoperfusion determine the outcome of brain injury.[15] Fleeting episodes of ischemia too short in time or too weak in depth, but sufficient to cause transient neuronal dysfunction and symptoms, may never leave a histologic trace. Unfortunately, we cannot, as yet, identify with confidence the border between structural and functional disorder and, in consequence, cannot distinguish with certainty between TIA and stroke. The focus of diagnosis and treatment should be on the causative processes. In other words, what matters is the terrorist behind the threat or the attack, not how long it lasted.

In consequence, there appears to be no clinical advantage in separating TIAs from strokes because our diagnostic tests are insufficient to do so. Impotence should not lead to diagnostic nihilism. Tomás de Iriarte, a Spanish writer of fables of the eighteenth century, wrote a popular fable, "The two rabbits," [16] which has been read by every Spanish-speaking child in grade school. Two rabbits saw dogs coming and began to argue whether these were greyhounds or hunting dogs. So enmeshed were they in their dispute that they failed to run away from the dogs, which finally caught them. As others have repeatedly emphasized, independently of whether TIA or stroke, clinicians should vigorously search for the cause and apply treatment as soon as feasible because the circumstances for a catastrophic event are equally pressing in both instances.

In conclusion, we have finally unchained TIAs from their time constraint. Perhaps the next step should be to de-emphasize the diagnostic quest in search of a line dividing functional from structural lesions until such time when we can do it with confidence. In the meantime, given the uncertainty, the episode could be termed "ischemic attack, probably transient."