A new study shows that nearly half of all strokes that occur after a transient
ischemic attack (TIA) occur within the first 24 hours, highlighting the need for
emergent intervention, the researchers say.

The good news is that the ABCD2 score, a validated risk score, was reliable in
this hyperacute phase, meaning that "appropriately triaged emergency assessment
and treatment are feasible," the researchers, with senior author Peter M.
Rothwell, MD, from the Stroke Prevention Research Unit at Oxford University and
John Radcliffe Hospital, in the United Kingdom, conclude.

This is the first rigorous population-based study of the risk for recurrent
stroke within 24 hours of TIA, Dr. Rothwell told Medscape Neurology. "We found
that nearly half of all the strokes that occur within 30 days after a TIA
actually occur within those first 24 hours, so unless we intervene more quickly
and treat it as a true emergency, rather than a 'see-urgently' problem, we'll
miss the opportunity to prevent some of those early recurrent strokes," he said.

The results, reported on behalf of the Oxford Vascular Study, are published in
the June 2 issue of Neurology.

Risk Underestimated

Over the past few years, Dr. Rothwell's group and others have been examining the
natural history of TIA and minor strokes, looking at extent to which the early
risk for recurrent stroke has been underestimated in the past and trying to
determine how best to prevent recurrent events after the warning signal of TIA
has occurred.

Results of the Early Use of Existing Preventive Strategies for Stroke (EXPRESS)
trial, of which Dr. Rothwell was principal investigator, showed that urgent
aggressive intervention after a TIA or minor stroke cut the 90-day risk for
recurrent stroke by 80%, as well as reducing fatal and nonfatal stroke,
disability, hospital admission days, and costs by the same magnitude (Rothwell
PM et al. Lancet 2007; 370:1398-1400; Luengo-Fernandez R et al. Lancet Neurol
2009;8:218-219).

On the basis of these kinds of findings, clinical guidelines in most countries
have changed significantly, recommending that patients should be assessed within
24 hours of a TIA or minor stroke, down from a recommendation of 7 days only a
year ago.

Still, while 24 hours is better than 7 days, "it's still not quite a medical
emergency," he said. In this paper, they sought to determine the real risk for
recurrence in the 24 hours following a TIA, "to see what the very early risk
really is in the first few hours and what might be gained, therefore, by seeing
patients even earlier, as well as what might be gained by better public
education to get patients to present immediately when they have 1 of these minor
episodes."

The ABCD2 risk score aims to help clinicians identifying those at highest risk
but was derived for prediction of the risk for stroke at 7 days and has not been
examined in this hyperacute phase, Dr. Rothwell noted.

Using data from the Oxford Vascular Study, a prospective, population-based
incidence study of TIA and stroke, they determined the risk for recurrent stroke
at 6, 12, and 24 hours after an index event.

Of 1247 patients with a first TIA or stroke, 35 had recurrent strokes within 24
hours, all of them in the same arterial territory, the authors report. In 25 of
these patients with recurrent strokes, the initial event was a TIA.

Of the 488 patients in total whose initial event was a TIA, 42% of the 25 events
that occurred within the first 30 days actually occurred within the first 24
hours.

"The other thing we were keen to do was make sure that the ABCD2 risk score,
which is now embedded in all the national and international guidelines, actually
worked for the risk of stroke within the first few hours," Dr. Rothwell noted.
"The guidelines say that patients with scores less than 4 needn't be seen quite
so urgently as those with higher scores, but that's only really been looked at
for 7-day risk."

What they found is that basic triage using the ABCD2 score "still seems
reasonable in patients that present within the first few hours," he said. The
12- and 24-hour risks were strongly related to the risk score (P = .02 and
.0003, respectively). However, these findings were still based on small numbers
of outcomes, they caution, and further studies on this would help to confirm
their results.

Of some concern, 16 of the 25 (64%) recurrent stroke patients with TIA as their
initial event did seek medical attention, usually from their family doctor,
after their TIA but did not receive antiplatelet therapy acutely, nor were they
sent to the acute intervention clinic at their institution. "However, the fact
that the majority of patients sought medical attention prior to their recurrence
indicates that emergency triage and treatment are feasible, if front-line
services recognize the need," the authors write.

Dr. Rothwell added that he sees the medical profession and neurologists in
particular as getting the message that patients presenting with a TIA and a high
risk score need to be seen "immediately, rather than tomorrow, which is the
current guideline."

"I think the bigger challenge is to get that message over to the public, because
at the moment only about 50% of patients who have a TIA seek medical attention
within 24 hours, and a lot of patients don't seek medical attention at all," he
said.

TIA: A Medical Emergency

Asked for comment on these findings, Philip B. Gorelick, MD, MPH, professor and
head of the Department of Neurology and Rehabilitation and director of the
Center for Stroke Research at the University of Illinois College of Medicine at
Chicago, said that accumulating data confirm that recent TIA should be treated
as a medical emergency.

A recent change in the definition of TIA by the American Heart
Association/American Stroke Association (Easton JD, et al. Stroke
2009;40:2276-2293) supports this conclusion, Dr. Gorelick noted, suggesting that
neuroimaging and diagnostic workup should be carried out within 24 hours of a
TIA when patients present within this time period, and that it is reasonable to
hospitalize patients who've had such an episode within the previous 72 hours if
they have an ABCD2 score of 3 or more, he added

Importantly, it is critical to rapidly determine the etiology of TIA, for
example, whether it results from a cardiac source embolism or large artery
disease, Dr. Gorelick noted. "Currently, a tissue-based definition of TIA has
been adopted. In aggregate data, it has been estimated that about 39% of
[magnetic resonance imaging] diffusion-weighted image studies in patients with
TIA show a cerebral ischemic injury pattern, and therefore, a cerebral
infarction has actually occurred."

The current study emphasizes again the importance of rapid diagnosis and
treatment, since there was a 5.1% risk for stroke in the first 24 hours after
TIA, with many of the strokes leading to a poor outcome, he said. "Furthermore,
the 7-day stroke rate was close to 10%," he noted. "Although 64% of these early
cases sought urgent medical attention prior to recurrent stroke, none received
antiplatelet therapy acutely."

The lesson from this and other studies is that TIA is not benign and urgent
diagnosis and treatment is indicated, even though this may not occur in
real-world experience, Dr. Gorelick concluded. In this effort, the ABCD2 score
is a reliable clinical tool that assesses the acute risk for stroke in TIA
patients.

"We need to continue to educate the public and healthcare professionals about
the importance of recent TIA as a predictor of stroke and the urgency of
diagnosis and treatment," Dr. Gorelick told Medscape Neurology. "Emergency TIA
assessment and treatment programs have proven to dramatically reduce the risk of
stroke after TIA. Widespread establishment of such programs should be considered
as we need to get TIA patients under the care of those who have experience in
vascular neurology and who can make a difference."

The study was funded by the UK Medical Research Council, the National Institute
of Health Research, the Stroke Association, the Dunhill Medical Trust, and the
Oxford Partnership Comprehensive Biomedical Research Centre. The authors report
no disclosures.

Neurology. 2009;72:1941-1947. Abstract