ADVANCES IN DIAGNOSIS AND TREATMENT OF
TRANSIENT ISCHEMIC ATTACK
OBJECTIVES:
1. Summarize the diagnostic modalities available for evaluating transient ischemic attack (TIA).
2. Detail the patient-specic treatments available for TIA to prevent stroke and an appropriate
approach for a specic patient.
INTRODUCTION
TIA represents a unique opportunity in
medicine to avert the onset of a devas-
tating disease process. The percentage of
patients that proceed on to acute ischemic
stroke (AIS) after a TIA is very high, but
can be decreased with appropriate diag-
nosis and treatment. Evidence regarding
new diagnostic and treatment modalities
is published at a rapid rate, but keeping
up with the latest science can help guide
therapy to provide the best possible out-
come for a patient seen in the Emergency
Department (ED) with symptoms of a
TIA.
The previous denition of TIA as a sud-
den, focal neurologic decit that lasts less
than 24 hours, is presumed to be of vascu-
lar origin, and is conned to an area of the
brain or eye perfused by a specic artery
was originally established in the 1970s
and was based on the assumption of com-
plete resolution of the ischemia. Newer
imaging techniques have shown that in
many TIAs, while the symptoms resolve,
small cerebral infarcts remain, implying
more urgency to the evaluation and man-
agement. New denitions of TIA have
been proposed that emphasize the fact
that most TIAs resolve within one hour
and that TIAs lasting longer than this fre-
www.emcreg.org
Brian A. Stettler, MD
Department of Emergency Medicine, University of Cincinnati
Cincinnati, Ohio
quently demonstrate an area of persistent
ischemia within the affected vascular ter-
ritory on MRI and are therefore an infarct
and not a TIA. One proposed denition
of TIA is a brief episode of neurolog-
ic dysfunction caused by focal brain or
retinal ischemia, with clinical symptoms
lasting less than one hour, and without
evidence of acute infarction. The authors
The risk of stroke
of this denition compare TIA to unsta-
ble angina to emphasize both the usually after TIA is around ten
short course of the symptoms and also the
percent in the rst 90
danger of progression to a much more se-
vere disease process.1 days in most population-
based studies, although
TIA Evaluation and Diagnosis
The risk of stroke after TIA is around the estimates vary from
ten percent in the rst 90 days in most 4-20% depending on
population-based studies, although the
the population selected.
estimates vary from 4-20% depending on
the population selected.2-6 Importantly,
studies agree that half of these strokes
will occur within the rst 48 hours af-
ter the initial TIA, highlighting the need
for an expeditious evaluation and urgent
treatment to avoid permanent neurologic
decit. The most recent update to the
guidelines for the management of TIA
published by the American Heart Associ-
ation (AHA) Council on Stroke state that
A TIA should be promptly evaluated
55
 ADVANCING THE STANDARD OF CARE:
Cardiovascular and Neurovascular Emergencies
because delaying diagnosis risks preventable stroke.
They further state that Hospitalization is often justi-
ed to expedite evaluation and lessen the possibility
of stroke.7,8
There is no published or agreed-upon time limit under
which the complete evaluation of a TIA should oc-
cur, but given the high short-term stroke risk, the ED
is certainly a pivotal point in the patients care. Rec-
ommendations for initial work-up of a potential TIA
include a complete history and physical examination,
a non-contrast head computed tomography (CT) scan,
blood tests as indicated, and an evaluation of the pa-
tients vascular anatomy, particularly the extracranial
carotid arteries. An echocardiogram is recommended
in some patients but is not mandated for all. While
there are not yet any discrete blood tests that will docu-
ment the presence of cerebral ischemia, the American
Heart Association (AHA) recommends a complete
blood count and chemistry prole, largely to exclude
other possible etiologies for the patients presentation
such as hypoglycemia. An electrocardiogram is also
recommended to evaluate for dysrhythmias.
Although acute imaging, particularly a non-contrast
head CT, has been recommended in the initial evalu-
ation of a TIA, evidence for its utility is sparse. Most
estimates place the ndings of a non-vascular lesion
accounting for neurologic symptoms at about one per-
cent of all scans performed. Douglas performed a ret-
rospective analysis of patients seen in the ED for TIA
who underwent a head CT and had a new infarct seen
despite complete resolution of symptoms. Patients
were then followed for 90 days and the odds ratio for
stroke within 90 days was determined. Of patients with
a new infarct on CT, the odds ratio for a stroke at 90
days was 4 (95% CI 1.16 14.14) after adjustment for
confounding variables.9 While the number of patients
in the study with a new infarct was small, the study
supports placing those patients with a new infarct on
CT, despite being asymptomatic, in a higher-risk cat-
egory and assuring timely evaluation and treatment.
56
There has been much work recently focusing on the
evaluation of TIA using diffusion-weighted imaging
(DWI) of MRI. This imaging modality is attractive
for several reasons. The rst is that a positive lesion
on DWI conrms the diagnosis of TIA in a patient
not currently symptomatic and eliminates many of the
TIA mimics, similar to positive biomarkers in acute
coronary syndrome. This possibility of conrmation
of diagnosis does not exist with CT, as CT rarely
demonstrates a lesion in real-time and is less sensi-
tive for acute ischemia than DWI. In multiple series
of patients evaluated for TIA, around forty percent
of patients show DWI-positivity, with a range from
21 68%.10-15 Diffusion-weighted imaging becomes
positive early in an ischemic insult, and when coupled
with other MRI sequences, can identify which insults
are acute and which are chronic in a much more ac-
curate fashion than CT. Positive DWI tends to occur
much more often in TIAs which last greater than one
hour, with some studies demonstrating a correlation
with duration of TIA and likelihood of positive DWI.
DWI-positivity has been shown in TIAs lasting only
ten minutes, however, and may still have a place in
the diagnosis of TIA with of short duration of symp-
toms. However, it is very unlikely that DWI will be
positive in a patient with only seconds to a few min-
utes of ischemic symptomatology.
The second benet of DWI would be further risk
stratication of the TIA patient with a positive DWI
scan. There are preliminary data which support that
patients with positive DWI are more likely to have
a stroke at 90-day follow-up. This makes sense in-
tuitively, as at least half of DWI hyperintense lesions
will progress to lesions on standard MRI or CT, thus
dening them as infarctions. Coutts et al. described
that patients with a lesion on DWI had a stroke risk
of 10.8% - 32.6% at 90 days, but only 4.3% if there
was no evidence of DWI abnormality.16 The study in-
cluded minor stroke patients, but still provides sup-
port to the notion that patients with DWI lesions have
a worse outcome if not intervened upon urgently.
www.emcreg.org
 ADVANCES IN DIAGNOSIS AND TREATMENT OF
TRANSIENT ISCHEMIC ATTACK
The utility of DWI in the evaluation of TIA is still
being dened. The patient with positive DWI most
likely has a higher risk of subsequent stroke than all
TIAs combined. In addition, the lack of a lesion on
DWI does not refute the diagnosis of TIA, and these
patients still have signicant risk of future ischemia.
At this point, DWI should most likely be regarded as
a specic but insensitive test for transient focal ce-
rebral ischemia, serving to conrm the diagnosis in
some cases. Urgent evaluation is still required in all
patients with a clinical diagnosis of TIA, regardless of
cerebral imaging results.
One test that does help to risk stratify all patients with
TIA and has proved to affect outcomes is evaluation
of the extracranial vasculature. The gold standard of
assessing the vasculature has long been catheter an-
giography, but there are many less invasive tests cur-
rently available that are reliable and reproducible.
Duplex ultrasonography has largely replaced invasive
angiography as a screening test of choice for critical
carotid stenosis, although ultrasonography does tend
to overestimate degree of stenosis and has difculty
determining the difference between very high-grade
stenosis and occlusion. Further, in pooled analysis,
duplex ultrasonography has a lower sensitivity than
some other imaging modalities at 86%.17 Magnetic
resonance imaging (MRA) is another test that has
been used recently as a screening test for vessel ste-
nosis and typically provides good images of the ca-
rotid and vertebral arteries as well as the intracranial
arteries and Circle-of-Willis arteries not accessible
by ultrasound. The sensitivity of MRA for high-grade
stenosis is nearly 95% in pooled analysis, in compari-
son with digital subtraction angiography (DSA).
A newer imaging modality is CT angiography (CTA).
CT angiography takes advantage of a technology
already available in most EDs around-the-clock to
image the cervical vessels with good sensitivity and
specicity. Josephson et al. analyzed a retrospective
case series of consecutive patients with TIA or stroke
that had undergone both CTA and DSA to evaluate
www.emcreg.org
the cervical carotid arteries. Using a 70% cutoff for
high-grade stenosis, CTA and DSA were in agree-
ment in 96% of the vessels analyzed. CTA was 100%
sensitive and 63% specic with a negative predic-
tive value of 100% for determining cervical stenosis
greater than 70%.18 This cut-off value is important as
carotid endarterectomy is recommended for reason-
able surgical candidates who are symptomatic and
have greater than 70% stenosis of the extracranial
carotid artery. The advantages of CTA are that it is
usually available 24 hours per day, does not require
an extra test or transport to be performed as many pa-
tients will undergo a head CT anyway, and has little
risk beyond that of an IV contrast load. It does suffer
from overestimating the degree of stenosis, but ultra-
sound and MRA have similar qualities. Further, it al-
lows another crucial and urgent step in the work-up
of TIA to be performed in real-time in the ED, theo-
retically resulting in improved risk-stratication and
earlier treatment to prevent stroke.
Risk-stratication in a patient who presents with a fo-
cal neurologic decit that is either present transiently
during evaluation or resolved entirely by the time of
evaluation is difcult. Johnston and colleagues de-
scribed factors that made patients in their population
higher risk for stroke after presenting to the ED and
being diagnosed with a TIA. An episode that lasted
greater than ten minutes, weakness during the episode,
speech impairment, history of diabetes, and age great-
er than 60 years were all associated with an increased
risk of stroke by 90 days using multivariate regression
analysis.19 Rothwell et al. further attempted to dene
the stroke risk after TIA using a score derived from
registries of TIA in a population-based study. While
somewhat more complex, a score was derived that did
correlate with 7-day stroke risk in their population.20
Neither of these rules has been prospectively validated
outside their own populations to determine admission
criteria at this point, however, and must be applied
with caution. The emergency physician must still con-
sider a general risk of stroke of 5% at 48 hours after
TIA when determining disposition from the ED.
57
 ADVANCING THE STANDARD OF CARE:
Cardiovascular and Neurovascular Emergencies
TIA Treatment
The treatment of TIA hinges on the presumed cause
of the event. While TIA requires no treatment itself,
the optimal method of preventing a future stroke is the
clinical focus. Antiplatelet agents such as aspirin, clop-
idogrel, and dipyridamole are the cornerstone of treat-
ment for many forms of TIA. Warfarin still plays a role
in the treatment of TIA presumed to be of cardioem-
bolic origin, while other antithrombotic agents remain
controversial. Correction of critical stenosis within the
extracranial carotid artery is of proven benet in some
patients, although there is debate whether surgery or
stent placement is the best option in selected patients.
TIAs that are not thought to be cardioembolic in na-
ture should initially be treated with an antiplatelet
agent, unless immediate carotid surgery is planned.
In patients not currently on an anti-platelet agent, as-
pirin in a dose from 50 325 mg/day is still the rst
line of therapy.8 Aspirin is thought to convey a 13-
15% relative risk reduction versus placebo in second-
ary prevention of ischemic stroke or other vascular
event.21,22 There has not been shown to be a denite
difference in risk reduction across different doses of
aspirin, with low doses appearing to be as effective
as higher doses, but with fewer gastrointestinal and
bleeding side effects.
In patients who have failed aspirin therapy from a re-
current event despite treatment, the second choice of
an antiplatelet agent is less clear. Both clopidogrel and
extended-release dipyridamole/aspirin combinations
are available on the market and have been shown to
have efcacy in secondary prevention, although they
have not yet been compared head-to-head in pub-
lished studies. Two ongoing clinical trials, ProFESS
(Prevention Regimen for Effectively avoiding Second
Strokes) and ESPRIT (European/Australian Stroke
Prevention in Reversible Ischemia Trial) are testing
extended-release dipyridamole/aspirin combinations
against clopidogrel or aspirin alone, respectively.
Current evidence from the CAPRIE trial suggests a
marginal benet for clopidogrel (75mg/day) over as-
pirin (325mg/day) in secondary prevention of stroke,
58
myocardial infarction, or vascular death, with a rela-
tive risk reduction of 8.7% (95% CI 3.0 16.5).23 The
MATCH trial found no benet of the addition of aspi-
rin (75mg/day) to clopidogrel (75mg/day) in secondary
prevention,24 although the usual clinical question of the
addition of clopidogrel to aspirin, instead of the reverse,
was not answered. Clopidogrel is favored as a second-
line antiplatelet agent in those patients who cannot tol-
erate aspirin secondary to gastrointestinal distress.
One major trial, the European Stroke Prevention Study
2, has evaluated aspirin versus an extended-release di-
pyridamole/aspirin combination in patients with TIA
or stroke in the preceding three months. Four treat-
ment groups were randomized, including placebo,
aspirin alone (50 mg/day), extended-release dipyri-
damole alone (400 mg/day), and aspirin plus extend-
ed-release dipyridamole (50/400). The trial showed
a benet to both aspirin and dipyridamole over pla-
cebo in secondary prevention of stroke, but showed
a signicant benet for the combination therapy over
either agent alone. The relative risk reduction for the
combination therapy was 37%, as compared to 18%
for aspirin alone and 16% for dipyridamole alone.25
While these results have yet to be conrmed in other
work, this preliminary result is encouraging.
The patient with atrial brillation and a TIA is usu-
ally presumed to have a cardioembolic source for the
embolus. Guidelines currently recommend systemic
anticoagulation with adjusted-dose warfarin in all pa-
tients who do not have a contraindication.8 Support
for this recommendation is provided by the pooled
analysis authored by Hart and colleagues. This analy-
sis combined 2 randomized trials of patients with atri-
al brillation and TIA who were randomized to either
aspirin or adjusted-dose anticoagulation with warfa-
rin. In patients randomized to warfarin a relative risk
reduction of 56% for ischemic stroke after TIA was
observed versus aspirin. The absolute rate reduction
in stroke was 4% per year for anticoagulation over
aspirin.26 There is strong support for anticoagulation
of the TIA patient with atrial brillation, unless a con-
traindication exists.
www.emcreg.org
 ADVANCES IN DIAGNOSIS AND TREATMENT OF
TRANSIENT ISCHEMIC ATTACK
Oral direct thrombin inhibitors have also been stud-
ied for systemic anticoagulation in the setting of atrial
brillation to reduce stroke risk. The benet of such
agents is that there is no need to monitor the Inter-
national Normalized Ratio (INR) and that there are
few drug-drug or drug-food interactions as seen with
warfarin, thus making them easier and potentially
more cost-effective. One study of ximelagatran (36
mg BID) versus adjusted-dose warfarin (INR 2.0-3.0)
was undertaken in patients with atrial brillation at
risk for stroke, although not all patients had a previ-
ous stroke or TIA. The data supported the efcacy of
the agent in reducing stroke risk, showing no statisti-
cally signicant difference in stroke or embolic events
with a mean 20-month follow-up between the study
drug and warfarin.27 Although this primary outcome
is encouraging, six percent of patients developed an
elevation of liver enzymes, with one documented fatal
case of liver failure in the study group. Future study
will focus on balancing an acceptable safety prole in
addition to proven efcacy for such agents.
Carotid endarterectomy (CEA) has proved benecial in
patients with stenosis of 70 99% of the extracranial
carotid artery who have had a TIA.28,29 While there is
some immediate risk to surgery, the long-term risk fa-
vors surgical treatment over best medical management,
with an absolute risk reduction of 17% over two years
for subsequent ipsilateral stroke in the surgical treat-
ment arm of the trial. The benet in patients who have
had a TIA and have stenosis of 50 - 69% is less robust at
6.5 percent absolute risk reduction over ve years, but
is still clinically relevant in patients who are reasonable
surgical candidates and have good life expectancy.30 It
appears that the benet is realized more in early surgery
after the initial TIA, and delay decreases the overall
benet. This is most likely due to the high incidence of
stroke in the rst 48 hours following TIA, which would
not be prevented by delayed surgery. The benet also
appears to be attenuated somewhat in isolated amau-
rosis fugax and in women, although the latter may also
be a timing effect. There is no proven benet to CEA in
patients with less than 50% stenosis, and these patients
should be treated with an antiplatelet agent.
www.emcreg.org
Stenting of the carotid artery in symptomatic stenosis
is a procedure which has gained momentum in recent
years, particularly in patients who are felt to be a poor
surgical candidate due to concomitant health problems.
While the technology of the devices advances rapidly,
there are no randomized studies that currently demon-
strate their efcacy versus CEA. Newer devices em-
ploy mechanisms to trap potential emboli released by
deployment of the stent attempting to decrease intraop-
erative complications, but large clinical trials are cur-
rently ongoing and results have not yet been reported.
As with all procedures, complication rates are lowest in
experienced hands, and the long-term benet of stent
placement will depend on low rates of intraoperative
complications. Current potential indications for stent-
ing would include patients with symptomatic carotid
stenosis of 70 99% not amenable to open CEA.
There are many groups and subtypes of TIA, includ-
ing those caused by intracranial vessel stenosis, hy-
percoaguable states, and cryptogenic emboli from
patent foramen ovale that are not covered in this
review. Treatment of these patients is typically best
managed in concert with a neurologist and consulta-
tion is warranted. New treatments develop constantly,
but much controversy surrounds some of these less-
common disease states.
The future of diagnosis and treatment of TIA will al-
low emergency physicians to further dene the TIA
population at highest risk for ischemic stroke. There
is great potential for observational medicine to pro-
vide a complete work-up for TIA with neurologic
consultation to begin appropriate treatment and avoid
a hospital admission. Ongoing trials seek to further
dene the optimal antiplatelet agent for TIA and seek
alternatives to warfarin for systemic anticoagulation
when indicated.
SUMMARY
Both currently and in the future, TIAs should be treat-
ed as a medical emergency. The rates of subsequent
stroke after TIA are substantial, many of which are
59
 ADVANCING THE STANDARD OF CARE:
Cardiovascular and Neurovascular Emergencies

preventable with appropriate treatment. Only with a 14. Restrepo L, Jacobs MA, Barker PB, Wityk RJ. Assessment of tran-
sient ischemic attack with diffusion- and perfusion-weighted imag-
thorough evaluation can the appropriate treatment be ing. Am J Neuroradiol 2004;25(10):1645-52.
instituted to attempt to avoid serious complications. If 15. Crisostomo RA, Garcia MM, Tong DC. Detection of diffu-
sion-weighted MRI abnormalities in patients with transient
a complete work-up cannot be obtained in an urgent ischemic attack: correlation with clinical characteristics. Stroke
fashion as an outpatient, admission to the hospital is 2003;34(4):932-7.
16. Coutts SB, Simon JE, Eliasziw M, et al. Triaging transient isch-
warranted. Optimal antiplatelet, antithrombotic, or emic attack and minor stroke patients using acute magnetic reso-
surgical treatment can help avert death or severe dis- nance imaging. Ann Neurology 2005;57(6):848-54.
17. Nederkoorn PJ, van der Graaf Y, Hunink MGM. Duplex ultra-
ability caused by ischemic stroke. sound and magnetic resonance angiography compared with digital
subtraction angiography in carotid artery stenosis: A systematic
review. Stroke 2003;34(5):1324-31.
REFERENCES 18. Josephson SA, Bryant SO, Mak HK, Johnston SC, Dillon WP,
Smith WS. Evaluation of carotid stenosis using CT angiog-
1. Albers GW, Caplan LR, Easton JD, et al. Transient ischemic attack
raphy in the initial evaluation of stroke and TIA. Neurology
- proposal for a new denition. N Engl J Med 2002;347(21):1713-6.
2004;63(3):457-60.
2. Johnston SC, Gress DR, Browner WS, Sidney S. Short-term
19. Johnston SC, Sidney S, Bernstein AL, Gress DR. A comparison of
prognosis after emergency department diagnosis of TIA. JAMA
risk factors for recurrent TIA and stroke in patients diagnosed with
2000;284(22):2901-6.
TIA. Neurology 2003;60(2):280-5.
3. Coull AJ, Lovett JK, Rothwell PM. Population based study of
20. Rothwell P, Giles M, Flossmann E, et al. A simple score (ABCD)
early risk of stroke after transient ischaemic attack or minor stroke:
to identify individuals at high early risk of stroke after transient
implications for public education and organisation of services. Br
ischaemic attack. Lancet 2005;366(9479):29-36.
Med J 2004;328(7435):326-.
21. Johnson ES, Lanes SF, Wentworth III CE, Sattereld MH, Abebe
4. Hill MD, Yiannakoulias N, Jeerakathil T, Tu JV, Svenson LW,
BL, Dicker LW. A metaregression analysis of the dose-response
Schopocher DP. The high risk of stroke immediately after
effect of aspirin on stroke. Arch Intern Med 1999;159(11):1248-53.
transient ischemic attack: A population-based study. Neurology
22. Antithrombotic Trialists Collaboration. Collaborative meta-analy-
2004;62(11):2015-20.
sis of randomised trials of antiplatelet therapy for prevention of
5. Eliasziw M, Kennedy J, Hill MD, Buchan AM, Barnett HJM. Early
death, myocardial infarction, and stroke in high risk patients. Br
risk of stroke after a transient ischemic attack in patients with inter-
Med J 2002;324(7329):71-86.
nal carotid artery disease. Can Med Assoc J 2004;170(7):1105-9.
23. CAPRIE Steering Committee. A randomised, blinded, trial of
6. Lisabeth LD, Ireland JK, Risser JMH, et al. Stroke risk after
clopidogrel versus aspirin in patients at risk of ischaemic events
transient ischemic attack in a population-based setting. Stroke
(CAPRIE). Lancet 1996;348(9038):1329-39.
2004;35(8):1842-6.
24. Diener H-C, Bogousslavsky J, Brass LM, et al. Aspirin and
7. Feinberg W, Albers G, Barnett H, et al. Guidelines for the manage-
clopidogrel compared with clopidogrel alone after recent isch-
ment of transient ischemic attacks. From the Ad Hoc Committee
aemic stroke or transient ischaemic attack in high-risk patients
on guidelines for the management of transient ischemic attacks of
(MATCH): randomised, double-blind, placebo-controlled trial.
the Stroke Council of the American Heart Association. Circulation
Lancet 2004;364(9431):331-7.
1994;89(6):2950-65.
25. Diener HC, Cunha L, Forbes C, Sivenius J, Smets P, Lowenthal
8. Albers GW, Hart RG, Lutsep HL, Newell DW, Sacco RL. Supple-
A. European Stroke Prevention Study 2. Dipyridamole and acetyl-
ment to the Guidelines for the Management of Transient Ischemic
salicylic acid in the secondary prevention of stroke. J Neurol Sci
Attacks : A Statement From the Ad Hoc Committee on Guidelines
1996;143(1-2):1-13.
for the Management of Transient Ischemic Attacks, Stroke Coun-
26. Hart RG, Pearce LA, Koudstaal PJ. Transient ischemic attacks in
cil, American Heart Association. Stroke 1999;30(11):2502-11.
patients with atrial brillation: implications for secondary preven-
9. Douglas VC, Johnston CM, Elkins J, Sidney S, Gress DR, Johnston
tion: The European Atrial Fibrillation Trial and Stroke Prevention
SC. Head computed tomography ndings predict short-term stroke
in Atrial Fibrillation III Trial. Stroke 2004;35(4):948-51.
risk after transient ischemic attack. Stroke 2003;34(12):2894-8.
27. SPORTIF Executive Steering Committee for the SPORTIF V
10. Rovira A, Rovira-Gols A, Pedraza S, Grive E, Molina C, Alvarez-
Investigators. Ximelagatran vs warfarin for stroke prevention in
Sabin J. Diffusion-Weighted MR imaging in the acute phase of
patients with nonvalvular atrial brillation: A randomized trial.
transient ischemic attacks. Am J Neuroradiol 2002;23(1):77-83.
JAMA 2005;293(6):690-8.
11. Kidwell CS, Alger JR, Di Salle F, et al. Diffusion MRI in patients
28. Rothwell P, Eliasziw M, Gutnikov S, et al. Analysis of pooled data
with transient ischemic attacks. Stroke 1999;30(6):1174-80.
from the randomised controlled trials of endarterectomy for symp-
12. Winbeck K, Bruckmaier K, Etgen T, von Einsiedel HG, Rottinger
tomatic carotid stenosis. Lancet 2003;361(9352):107-16.
M, Sander D. Transient ischemic attack and stroke can be dif-
29. North American Symptomatic Carotid Endarterectomy Trial Col-
ferentiated by analyzing early diffusion-weighted imaging signal
laborators. Benecial effect of carotid endarterectomy in symp-
intensity changes. Stroke 2004;35(5):1095-9.
tomatic patients with high-grade carotid stenosis. N Engl J Med
13. Purroy F, Montaner J, Rovira A, Delgado P, Quintana M, Alvarez-
1991;325(7):445-53.
Sabin J. Higher risk of further vascular events among transient
30. Barnett HJM, Taylor DW, Eliasziw M, et al. Benet of carotid
ischemic attack patients with diffusion-weighted imaging acute
endarterectomy in patients with symptomatic moderate or severe
ischemic lesions. Stroke 2004;35(10):2313-9.
stenosis. N Engl J Med 1998;339(20):1415-25.

Copyright EMCREG-International, 2005

60
www.emcreg.org