Advances in

Stroke Prevention
Plenary Symposium at the 11th European Stroke Conference
Geneva, Switzerland, May 29–June 1, 2002
Supported by an Educational Grant from Sanofi-Synthelabo

Editors
Julien Bogousslavsky, Lausanne
Marie-Germaine Bousser, Paris

13 figures and 6 tables, 2003

Basel 폷 Freiburg 폷 Paris 폷 London 폷 New York 폷

Bangalore 폷 Bangkok 폷 Singapore 폷 Tokyo 폷 Sydney
S. Karger
Medical and Scientific Publishers
Basel 폷 Freiburg 폷 Paris 폷 London
New York 폷 Bangalore 폷 Bangkok
Singapore 폷 Tokyo 폷 Sydney
Fax + 41 61 306 12 34
E-Mail karger@karger.ch
www.karger.com
Drug Dosage
The authors and the publisher have exerted every effort to en-
sure that drug selection and dosage set forth in this text are in
accord with current recommendations and practice at the time
of publication. However, in view of ongoing research, changes
in government regulations, and the constant flow of informa-
tion relating to drug therapy and drug reactions, the reader is
urged to check the package insert for each drug for any change
in indications and dosage and for added warnings and precau-
tions. This is particularly important when the recommended
agent is a new and/or infrequently employed drug.
All rights reserved.
No part of this publication may be translated into other
languages, reproduced or utilized in any form or by any means,
electronic or mechanical, including photocopying, recording,
microcopying, or by any information storage and retrieval
system, without permission in writing from the publisher or, in
the case of photocopying, direct payment of a specified fee to
the Copyright Clearance Center (see ‘General Information’).
© Copyright 2003 by S. Karger AG,
P.O. Box, CH–4009 Basel (Switzerland)
Printed in Switzerland on acid-free paper by
Reinhardt Druck, Basel
ISBN 3–8055–7583–1
 Vol. 16, Supplement 1, 2003

Contents

1 Introduction
Bousser, M.-G. (Paris); Bogousslavsky, J. (Lausanne)

3 Long-Term Outcome after Stroke due to Atrial Fibrillation

Mattle, H.P. (Bern)

9 Outcome after Brain Haemorrhage

Dennis, M.S. (Edinburgh)

14 Long-Term Outcome after Ischaemic Stroke/Transient Ischaemic Attack

Hankey, G.J. (Perth)

20 Evidence with Antiplatelet Therapy and ADP-Receptor Antagonists

Easton, J.D. (Providence, R.I.)

© 2003 S. Karger AG, Basel

Fax + 41 61 306 12 34 Access to full text and tables of contents,

E-Mail karger@karger.ch including tentative ones for forthcoming issues:
www.karger.com www.karger.com/ced_issues
 Cerebrovasc Dis 2003;16(suppl 1):1–2
DOI: 10.1159/000069933
Introduction
Stroke is the principal cause of disability, dependency
and loss of social competence in the western world. The
majority of strokes are of ischaemic origin, and most isch-
aemic strokes are due to atherothrombosis. Clearly, a
thorough knowledge of prognosis and effective secondary
prevention strategies are of paramount importance for cli-
nicians who care for stroke patients.
At the 11th European Stroke Conference, held in
Geneva, Switzerland, between May 29 and June 1, 2002,
delegates had the opportunity to attend two important
symposia – an Educational Symposium entitled Long-
term outcome after different stroke subtypes and a Satellite
Symposium entitled Future trends in long-term prevention
of ischaemic stroke – the role of atherothrombosis manage-
ment. This supplement to Cerebrovascular Diseases is
based on key presentations made during these sessions.
In the first article of this supplement, Dr Heinrich
Mattle provides an overview of long-term outcome after
stroke due to atrial fibrillation (AF). AF, which is the most
common cardiac arrhythmia and whose prevalence in-
creases with age, places the patient at risk for left atrial
thrombus formation, distal embolism and associated isch-
aemic stroke. AF is associated with substantial cardiovas-
cular mortality and morbidity. Indeed, long-term survival
data from the Framingham Heart Study indicates that AF
is independently associated with an approximate dou-
bling in mortality in both men and women. Moreover, in
the European Atrial Fibrillation Trial, the annualized rate
of vascular death, nonfatal stroke, nonfatal myocardial
infarction or systemic embolism in patients who received
placebo was 12%. There is clear evidence that antithrom-
© 2003 S. Karger AG, Basel
ABC 1015–9770/03/0165–0001$19.50/0
Fax + 41 61 306 12 34
E-Mail karger@karger.ch Accessible online at:
www.karger.com www.karger.com/ced
botic therapy reduces the risk of serious vascular isch-
aemic events in AF patients. Oral anticoagulation reduces
the risk of stroke recurrence by approximately 60% com-
pared with placebo. Acetylsalicylic acid (ASA) is less
effective than anticoagulation, reducing the risk of stroke
by around 20%. Thus, the use of ASA as first-line anti-
thrombotic therapy in AF patients is limited to those at
low risk. An important topic for future investigation is
whether dual antiplatelet therapy, for example clopidogrel
in addition to ASA, is a suitable alternative to warfarin for
the reduction of thromboembolism in high- or interme-
diate-risk AF patients. This will be the topic of the
planned ACTIVE trial.
Dr Martin Dennis then describes our current knowl-
edge on outcome after primary intracerebral haemorrhage
(PICH). Dr Dennis highlights the methodological difficul-
ties that hamper the study of PICH prognosis, including
the small proportion of all strokes that are due to PICH,
the need for brain imaging (or autopsy) to arrive at a reli-
able diagnosis of PICH, differences in definition of PICH
according to imaging method and its timing, and the het-
erogeneity of haemorrhage aetiologies and types. Al-
though individual estimates of 1-month case fatality have
wide confidence intervals, pooled data from unselected
PICH cohorts provide a more precise estimate of about
42%. After PICH, greater age and stroke severity are both
associated with increased case fatality and poorer func-
tional outcomes. Currently, there is no definite evidence
to indicate that the risk of recurrent stroke after PICH
differs from that after ischaemic stroke of equivalent clin-
ical severity. Generation of more precise data on progno-
sis after PICH will require pooling of data from communi-
ty-based studies that have used consistent definitions and
methodology. It is already well established that blood-
pressure lowering dramatically reduces the risk of recur-
rent intracerebral haemorrhage.
In the first of two articles relating to the generalized
nature of atherothrombosis and its consequences, Dr
Graeme Hankey reviews the evidence from high-quality
studies on short- and long-term predictors of post-stroke
outcome and studies on prognosis in patients with tran-
sient ischaemic attack (TIA) or ischaemic stroke. This
article demonstrates that the most consistent predictor of
short-term (1-month) mortality is stroke severity, whereas
over the longer term (1–5 years), the strongest predictor of
death is increasing age, closely followed by cardiac failure.
Additional predictive factors for death over this time peri-
od include a history of previous symptomatic athero-
thrombosis and risk factors for atherothrombotic disease.
Furthermore, over time, the annual risk of recurrent cere-
brovascular events decreases but the risk of cardiovascu-
lar events increases, emphasizing that patients with TIA
and ischaemic stroke are at risk of recurrent ischaemic
events of both the brain and the heart. On the basis of
these findings, Dr Hankey concludes that strategies for
secondary prevention after TIA or stroke should include
removal of symptomatic disease, control of risk factors for
atherogenesis, and prevention of atherothrombosis in all
vascular beds.
Finally, Dr J. Donald Easton provides an overview of
current evidence and future prospects for ADP-receptor
antagonist therapy for the prevention of vascular isch-
aemic events in atherothrombotic patients, with a partic-
ular focus on patients with symptomatic cerebrovascular
disease. As shown in meta-analyses by the Antithrombot-
2 Cerebrovasc Dis 2003;16(suppl 1):1–2
ic Trialists’ Collaboration and the Cochrane Stroke
Group, the ADP-receptor antagonists clopidogrel and
ticlopidine are associated with superior protection against
major atherothrombotic events (stroke, MI or vascular
death) compared with ASA. The Cochrane Group also
confirmed that compared with ticlopidine, clopidogrel
offers a more favourable safety/tolerability profile.
The CURE trial, which provided ‘proof of principle’
for the use of clopidogrel on top of ASA to provide incre-
mental benefit in high-risk atherothrombtic patients, has
stimulated interest in advanced antiplatelet strategies in
patients with ischaemic stroke. Since Geneva, this inter-
est has been reinforced by publication of the results of the
CREDO trial, which showed that long-term use of clopi-
dogrel on top of ASA provides sustained, incremental
benefit in patients who undergo percutaneous coronary
intervention. The ongoing MATCH study is comparing
clopidogrel on top of ASA versus clopidogrel alone in an
international, randomized, double-blind trial in patients
with recent TIA or ischaemic stroke who are at high risk
of atherothrombotic recurrence. The ongoing CHARIS-
MA trial compares clopidogrel and placebo on top of
usual treatment based on low-dose ASA in high-risk ath-
erothrombotic patients. Additional trials of clopidogrel
on top of standard therapy including ASA are planned in
neurology, including SPS3, in patients with small subcor-
tical strokes, and ATARI, in patients who have recently
recovered from a TIA.
The articles assembled here represent a current refer-
ence source on outcome after stroke and its management.
We hope that you find them useful and informative.
Marie-Germaine Bousser, MD
Julien Bogousslavsky, MD
Introduction
 Cerebrovasc Dis 2003;16(suppl 1):3–8
DOI: 10.1159/000069934
Long-Term Outcome after Stroke due to
Atrial Fibrillation
Heinrich P. Mattle
Department of Neurology, Inselspital, University of Bern, Bern, Switzerland
Key Words
Atrial fibrillation W Prognosis W Stroke
Abstract
Atrial fibrillation (AF) is the most common cardiac ar-
rhythmia. AF is paroxysmal or persistent and becomes
permanent when it does not convert to sinus rhythm
spontaneously or when attempted cardioversion fails.
The prevalence of AF is 0.4% in the general population
and increases with age up to 6–8% in octogenarians. In
men, the age-adjusted prevalence is generally higher
than in women. During AF, synchronous mechanical
atrial activity is disturbed, resulting in haemodynamic
impairment. This can give rise to thrombus formation
and embolism to the systemic circulation. Thrombus
associated with AF arises most frequently in the left atrial
appendage. Cerebrovascular emboli in AF patients most
often manifest as transient ischaemic attacks or isch-
aemic strokes. The overall rate of ischaemic stroke
among patients with nonrheumatic AF averages 5% per
year, but the rate increases with age. Patients with AF are
at higher risk of cerebrovascular events from all causes.
Of all strokes, one in every six occurs in patients with AF.
Including transient ischaemic attacks and silent strokes
detected radiographically, the overall rate of all cerebro-
© 2003 S. Karger AG, Basel
ABC 1015–9770/03/0165–0003$19.50/0
Fax + 41 61 306 12 34
E-Mail karger@karger.ch Accessible online at:
www.karger.com www.karger.com/ced
vascular events in AF patients rises to more than 7% per
year, although approximately one third of these are due
to causes that are only secondarily or incidentally associ-
ated with AF or related anticoagulant therapy. Antiar-
rhythmic therapy is useful to improve cardiac rate and
function in AF. However, to reduce first or recurrent
emboli, antithrombotic therapy is of paramount impor-
tance. Results from several randomized clinical trials of
antithrombotic therapies have shown that adjusted-dose
warfarin reduces first or recurrent stroke by about 60%
compared with placebo. When patients with nonvalvular
AF are anticoagulated, the odds against ischaemic stroke
and intracranial bleeding favour an INR between 2.0 and
3.0. Acetylsalicylic acid is less efficacious than warfarin
in AF patients, reducing the risk of stroke by about 20%.
Therefore, this antiplatelet agent should be used only for
AF patients at low risk. Anticoagulation is the current
treatment modality in AF patients at high or intermediate
risk, i.e. patients with history of transient ischaemic
attack or stroke, those aged 1 65 years, those with a his-
tory of hypertension, diabetes, heart failure or structural
heart disease, valvular disease or significant systolic dys-
function. The benefit of dual antiplatelet regimens in AF
patients is unknown, and combining antiplatelet agents
with different mechanisms of action is an important topic
for future investigation.
Copyright © 2003 S. Karger AG, Basel
Heinrich Mattle, MD
Department of Neurology
Inselspital, University of Bern
CH–3010 Bern (Switzerland)
Tel. +41 31 6329784, Fax +41 31 6320321, E-Mail heinrich.mattle@insel.ch
 12
Women
10 Men
Prevalence (%)
0
<55 55–59 60–64 65–69 70–74 75–79 80–84
Age (years)
Fig. 1. Prevalence of AF by age and gender [reproduced with kind
permission, 2].
Introduction
Atrial fibrillation (AF) is the most common cardiac
arrhythmia. It occurs more often concomitantly with
structural heart disease than alone. Approximately 16%
of stroke patients have AF, and the disease itself accounts
for 10% of all strokes. In the USA, it is estimated that
2.2–2.3 million people have AF [1, 2]. A crude estimate,
assuming a 0.95% prevalence of AF in Europe, is that 8.3
million of the 874 million people living in Eastern and
Western Europe have AF.
The incidence and prevalence of AF rises with age
(fig. 1) [2]. Indeed, after the age of 55, the prevalence dou-
bles with each decade. The prevalence of AF is 0.4% in
the general population and increases with age up to 6–8%
in octogenarians. In men, the age-adjusted prevalence is
generally higher than in women. However, because of
their greater longevity, there are more women than men
with AF. The incidence and prevalence of AF are ex-
pected to rise substantially in future decades due to the
growing proportion of elderly individuals, which in the
USA is expected to increase 2.5-fold over the next 50
years.
Data on the typical characteristics of the stroke patient
in AF are available from the International Stroke Trial
[3]. IST was a large, randomized trial of antithrombotic
therapy in over 19,000 patients with acute ischaemic
stroke, of which 3,169 had AF. In this study, stroke
4 Cerebrovasc Dis 2003;16(suppl 1):3–8
85
Fig. 2. Left atrial appendage thrombus in a 73-year-old female
patient with AF [reproduced with kind permission from Prof. Tho-
mas Schaffner, Department of Pathology, University of Bern].
patients who had AF were more likely to be older (odds
ratio (OR) = 7.2; 95% CI 6.9–7.6), female (OR = 1.6; 95%
CI 1.5–1.7), have depressed consciousness (OR = 2.4;
95% CI 2.2–2.6), and have a larger stroke (total anterior
circulation syndrome) (OR = 2.1; 95% CI 2.0–2.3) than
those in sinus rhythm [4].
Patients with AF carry a substantial risk of thromboem-
bolism. In AF, the contraction of the left atrium is dis-
turbed, resulting in haemodynamic impairment of both
the atrium and the ventricle. This leads to stasis and
thrombus formation, especially in the left atrial appendage
(fig. 2). The end result is a risk of distal embolization to the
systemic circulation, and ultimately a silent cerebral in-
farction, transient ischaemic attack, or ischaemic stroke.
Prognosis of AF
AF is an independent predictor of first-ever stroke,
increasing the risk by up to 5-fold [5]. Based on long-term
survival data from the Framingham Heart Study, the
mortality of men and women with AF was significantly
higher than that of subjects without AF [6]. After 10 years
of follow-up, 61.5% of men aged 55–74 years had died
compared with 30.0% of men without AF. In women,
57.6% of those with AF had died at 10 years compared
with 20.9% of women without AF (fig. 3). Comparable
results were observed in subjects aged 75–94 years. More-
over, in a secondary multivariate analysis that was limited
to those subjects who were initially free of cardiovascular
Mattle
 45
No AF
40
AF
35

30

Incidence (%)
25

20

15

10

0
Fig. 3. Cumulative mortality during long-term follow-up in subjects
Death at 2 weeks Death at 6 months
aged 55–74 years with and without AF in the Framingham Heart
Study [reproduced with kind permission, 6].

Fig. 4. Stroke fatality at 2 weeks and 6 months in AF patients

enrolled in the International Stroke Trial [reproduced with kind per-
mission, 4].
disease and valvular heart disease, AF was independently
associated with an approximate doubling in mortality in
both men and women (OR for men = 2.4, 95% CI 1.8–3.3;
OR for women = 2.2, 95% CI 1.6–3.1) [6].
A number of independent predictors of ischaemic
stroke have been identified in patients with AF. The
Atrial Fibrillation Investigators group analysed pooled
data from the control groups of five randomized primary
prevention trials and found the following risk factors on
multivariate analysis: previous stroke or transient isch-
aemic attack, OR = 2.5; hypertension, OR = 1.6; diabetes,
OR = 1.7; age, OR = 1.6 per 10 years [7]. Analysis of echo-
cardiograms from three of these trials subsequently found
that moderate-to-severe left ventricular dysfunction was
also predictive of stroke (OR = 2.5) [8]. Broadly similar
findings were reported by Hart et al. [9] based on analysis
of data from patients receiving acetylsalicylic acid (ASA)
in the SPAF (Stroke Prevention in Atrial Fibrillation) I, II
Fig. 5. Survival analysis for primary outcome event (vascular death,
and III trials. Transoesophageal echocardiography has nonfatal stroke, nonfatal myocardial infarction or systemic embo-
also identified predictors of stroke in AF patients. These lism) in the European Atrial Fibrillation Trial [11]. Anticoagulants
include thrombus of the left atrial appendage, sponta- reduced the annualized rate of primary outcome events from 17 to
neous echo contrast, and aortic arch plaques [10]. 8%. The annual stroke rate (not shown) was reduced from 12 to 4%
[reproduced with kind permission, 11].
In the IST trial, stroke patients with AF were slightly
more likely to experience a recurrent ischaemic stroke or
an intracranial haemorrhage, and were more than twice as
likely to die within 2 weeks as those without AF (fig. 4). A Overall, the annualized rate of the primary endpoint (vas-
similar effect of AF on mortality was also observed at 6 cular death, nonfatal stroke, nonfatal MI or systemic
months. embolism) in the placebo arm of this study was 17%. The
Data on longer-term outcome have been reported in annualized rates of all-cause stroke and of death were 12
the European Atrial Fibrillation Trial (EAFT) [11] (fig. 5). and 9%, respectively.

Prognosis of Atrial Fibrillation Cerebrovasc Dis 2003;16(suppl 1):3–8 5

 Table 1. Risk categories in patients with AF
Risk Risk factors
category
Highest Previous stroke or transient ischaemic attack
High Age 1 75 years
Hypertension
Left ventricular dysfunction
Risk factor detected by transoesophageal echocardiograpy
Rheumatic or valvular heart disease
1 1 moderate risk factor
Moderate Age 65–75 years
Diabetes
Coronary heart disease
No high-risk criteria
Low Age ! 65 years
No high-risk factors
Prevention of Stroke in AF Patients
Antithrombotic therapy has clearly been shown to
reduce the risk of serious vascular ischaemic events in
patients with AF [12]. A meta-analysis of six randomized
trials that compared dose-adjusted warfarin versus place-
bo in a total of 2,900 patients found a pooled relative risk
reduction of 62% (95% CI 48–72%) for the endpoint of
all-cause stroke [13]. In these trials, the absolute risk
reduction for primary prevention was 2.7% per year (data
from 5 trials) and for secondary prevention it was 8.4%
per year (data from a single trial (EAFT)). Major extra-
cranial haemorrhage was increased by warfarin therapy
(absolute risk increase 0.3% per year, relative risk = 2.4,
95% CI 1.2–4.6). Six trials have compared the use of ASA
versus placebo for stroke prevention in AF patients [13].
The pooled estimate from these studies gave an overall
22% (95% CI 2–38%) reduction in the risk of all-cause
stroke. Pooled analysis of data from five trials that com-
pared adjusted-dose warfarin plus ASA versus ASA alone
showed that combination therapy was superior to ASA
monotherapy, being associated with a 36% (95% CI 14–
52%) relative risk reduction for all-cause stroke. Major
extracranial haemorrhage was higher in those who re-
ceived warfarin relative to those who received ASA (rela-
tive risk = 2.0, 95% CI 1.2–3.4) [13].
In EAFT, anticoagulation therapy was more effective
than ASA in preventing serious vascular events, reducing
the rate of the primary endpoint from 17% per year with
6 Cerebrovasc Dis 2003;16(suppl 1):3–8
Annual risk Number needed to treat
of stroke
12% 14 (warfarin vs. placebo)
F6% 28 (warfarin vs. placebo)
F3.5% 48 (warfarin vs. placebo)
F1% F500 (ASA vs. placebo)
F250 (warfarin vs. ASA)
placebo to 8% per year and the annual rate of all-cause
stroke from 12 to 4%.
A major limitation of oral anticoagulation is its narrow
therapeutic window. In patients with nonrheumatic AF,
the risk of stroke has been found to rise steeply at INR
values !2.0. Thus, at an INR of 1.7, the adjusted odds
ratio for stroke, as compared with the risk at an INR of
2.0, is 2.0 (95% CI 1.6–2.4) [14]. At INR values 14.0, the
risk of intracranial haemorrhage rises exponentially [15].
Therefore, in nonrheumatic AF, an INR of 2.0–3.0 gives
the optimal risk:benefit ratio for the prevention of stroke
[12]. In patients with valvular disease, a higher INR (3.0–
3.5) should be targeted.
AF patients can be categorized according to stroke risk,
which is influenced by a number of variables and which
should be taken into account when initiating antithrom-
botic therapy (table 1). AF patients who have suffered a
previous stroke or transient ischaemic attack have the
highest risk of stroke, which is estimated at 12% per year.
Patients in the lowest risk category (those aged !65 years
with no high-risk factors) have an estimated annual stroke
risk of 1%, making them potential candidates for ASA
therapy. For warfarin, the number needed to treat (NNT)
to prevent one stroke per year ranges from 14 in the high-
est risk category to 48 in moderate-risk patients. In the
lowest risk category, the NNT for ASA versus placebo is
about 500 and for warfarin compared with ASA the NNT
is about 250.
Mattle
 The risk of thromboembolism with atrial flutter is of
the same order as the risk with AF. Therefore, anticoagu-
lants should be used with atrial flutter in the same way as
with AF.
Another issue in AF is rate and rhythm control. Several
trials have shown that restoration of sinus rhythm im-
proves cardiac function and exercise tolerance, but does
not reduce the risk of vascular death or thromboembolism
substantially [16–18]. For younger patients with structur-
ally normal hearts, restoration and maintenance of sinus
rhythm is still the goal, but in the great majority of AF
patients, rhythm control does not bring any advantage
over rate control. In older patients, patients with conges-
tive heart failure and patients with coronary artery dis-
ease, rhythm control was even associated with a higher
risk of death [17]. In the AFFIRM study, during a mean
follow-up of 3.5 years, the overall ischaemic stroke rate
was 6.3%. Mortality was 26.3%. The implications from
the rate and rhythm control trials are that rate control is
the primary approach to the treatment of AF, that rhythm
control, if used, can be abandoned early if it is not fully
satisfactory, and that continuous anticoagulation is war-
ranted in all patients with AF and risk factors for stroke,
even when sinus rhythm appears to be restored and main-
tained.
Unanswered Questions and Future Prospects
Patients with AF would clearly benefit from anti-
thrombotic therapy that is superior to ASA, equally effica-
cious as oral anticoagulation, easy to use and has an excel-
lent safety profile. Therefore, the question arises of wheth-
er dual antiplatelet regimens would be an alternative or
even superior to anticoagulation. There is a strong ratio-
nale to evaluate the use of clopidogrel on top of ASA in AF
patients, because: (1) the typical patient with AF who
requires antithrombotic therapy also has multiple risk fac-
tors for arterial vascular disease as well as for left atrial
thrombus; (2) based on the results of the CURE [19] and
CREDO trials [20], it is justifiable to expect that clopido-
grel plus ASA would be superior to ASA in this patient
population, and (3) in the only trial (in coronary stenting)
that has compared an ADP-receptor antagonist plus
thromboxane-inhibiting ASA versus warfarin plus ASA,
ticlopidine plus ASA significantly reduced the risk of
death or MI by 49% compared with warfarin plus ASA,
with a substantially lower risk of serious bleeding [21].
The answer to the question of whether a dual antiplatelet
regimen will be superior or comparable to anticoagulation
Prognosis of Atrial Fibrillation
will most likely come from the ACTIVE trial. ACTIVE is
planned to compare clopidogrel on top of ASA with war-
farin in AF patients.
Another unresolved issue is whether warfarin has a
favourable benefit:risk ratio in very elderly AF patients
(those aged 180 years), bearing in mind the increased risk
of haemorrhage in this population. In the absence of reli-
able data on this topic, a reasonable approach is to treat
these very elderly patients in the same way as slightly
younger patients.
Mechanical devices are under development for endo-
vascular occlusion of the left atrial appendage to prevent
thrombus formation and embolization. In a recent pilot
study, 15 high-risk AF patients who were poor candidates
for warfarin therapy underwent successful implantation
of an expandable cage for occlusion of the left atrial
appendage [22]. However, further studies are required to
demonstrate the long-term efficacy and safety of this
approach for the prevention of atrial thrombus formation
in patients with AF.
Conclusions
AF is an important risk factor for stroke and other vas-
cular events and is a predictor of increased stroke severity
and mortality. Thus, antithrombotic therapy is of para-
mount importance to prevent first or recurrent embolism
in patients with AF. In higher-risk patients, warfarin has a
superior benefit:risk ratio compared with ASA. In con-
trast, in patients with the lowest risk of stroke (F1% per
year) there is little, if any, net benefit of warfarin therapy
and therefore ASA is the current preferred option. Evalu-
ation of dual antiplatelet therapy, for example clopidogrel
on top of ASA, in AF patients may provide evidence of a
new alternative to warfarin for the reduction of thrombo-
embolism in high- or intermediate-risk AF patients.
Cerebrovasc Dis 2003;16(suppl 1):3–8 7
 References
1 Feinberg WM, Blackshear JL, Laupacis A,
Kronmal R, Hart RG: Prevalence, age distribu-
tion and gender of patients with atrial fibrilla-
tion: Analysis and implications. Arch Intern
Med 1995;155:469–473.
2 Go AS, Hylek EM, Phillips KA, Chang Y, Hen-
ault LE, Selby JV, Singer DE: Prevalence of
diagnosed atrial fibrillation in adults: National
implications for rhythm management and
stroke prevention: The AnTicoagulation and
Risk Factors in Atrial Fibrillation (ATRIA)
Study. JAMA 2001;285:2370–2375.
3 International Stroke Trial Collaborative
Group: The International Stroke Trial (IST): A
randomized trial of aspirin, subcutaneous hep-
arin, both or neither among 19,435 patients
with acute ischaemic stroke. Lancet 1997;349:
1569–1581.
4 Saxena R, Lewis S, Berge E, Sandercock PAG,
Koudstaal PJ, for the International Stroke Trial
Collaborative Group: Risk of early death and
recurrent stroke and effect of heparin in 3,169
patients with acute ischemic stroke and atrial
fibrillation in the International Stroke Trial.
Stroke 2001;32:2333–2337.
5 Wolf PA, Abbott RD, Kannel WB: Atrial fibril-
lation as an independent risk factor for stroke:
The Framingham Study. Stroke 1991;22:983–
988.
6 Benjamin EJ, Wolf PA, D’Agostino RB, Silber-
shatz H, Kannel WB, Levy D: Impact of atrial
fibrillation on the risk of death. The Framing-
ham Heart Study. Circulation 1998;98:946–
952.
7 Atrial Fibrillation Investigators: Risk factors
for stroke and efficacy of antithrombotic thera-
py in atrial fibrillation: Analysis of pooled data
from five randomized controlled trials. Arch
Intern Med 1994;154:1449–1457.
8 Atrial Fibrillation Investigators: Echocardio-
graphic predictors of stroke in patients with
atrial fibrillation: A prospective study of 1,066
patients from three clinical trials. Arch Intern
Med 1998;158:1316–1320.
8 Cerebrovasc Dis 2003;16(suppl 1):3–8
9 Hart RG, Pearce LA, McBride R, Rothbart
RM, Asinger RW on behalf of the Stroke Pre-
vention in Atrial Fibrillation (SPAF) Investiga-
tors: Factors associated with ischemic stroke
during aspirin therapy in atrial fibrillation:
Analysis of 2,012 participants in the SPAF I–
III clinical trials: Stroke Prevention in Atrial
Fibrillation (SPAF) Investigators. Stroke 1999;
30:1223–1229.
10 Stroke Prevention in Atrial Fibrillation Investi-
gators Committee on Echocardiography:
Transesophageal echocardiographic correlates
of thromboembolism in high-risk patients with
nonvalvular atrial fibrillation. Ann Intern Med
1998;128:639–647.
11 EAFT (European Atrial Fibrillation Trial)
Study Group: Secondary prevention in non-
rheumatic atrial fibrillation after transient isch-
aemic attack or minor stroke. Lancet 1993;342:
1255–1262.
12 Albers GW, Dalen JE, Laupacis A, Manning
WJ, Petersen P, Singer DE: Antithrombotic
therapy in atrial fibrillation. Chest 2001;119:
194S–206S.
13 Hart RG, Benavente O, McBride R, Pearce
LA: Antithrombotic therapy to prevent stroke
in patients with atrial fibrillation: A meta-anal-
ysis. Ann Intern Med 1999;131:492–501.
14 Hylek EM, Skates SJ, Sheehan MA, Singer DE:
An analysis of the lowest effective intensity of
prophylactic anticoagulation for patients with
nonrheumatic atrial fibrillation. N Engl J Med
1996;335:540–546.
15 Hylek EM, Singer DE: Risk factors for intracra-
nial hemorrhage in outpatients taking warfarin.
Ann Intern Med 1993;118:511–520.
16 Hohnloser SH, Kuck KH, Lilienthal J, for the
PIAF Investigators: Rhythm or rate control in
atrial fibrillation – Pharmacological Interven-
tion in Atrial Fibrillation (PIAF): a randomised
trial. Lancet 2000;356:1789–1794.
17 The Atrial Fibrillation Follow-Up Investiga-
tion of Rhythm Management (AFFIRM) In-
vestigators: A comparison of rate control and
rhythm control in patients with atrial fibrilla-
tion. N Engl J Med 2002;347:1825–1833.
18 Van Gelder IC, Hagens VE, Bosker HA, King-
ma JH, Kamp O, Kingma T, Said SA, Darma-
nata JI, Timmermans AJM, Tijssen JGP,
Crijns HJGM, for the Rate Control versus
Electrical Cardioversion for Persistent Atrial
Fibrillation Study Group: A comparison of rate
control and rhythm control in patients with
recurrent persistent atrial fibrillation. N Engl J
Med 2002;347:1834–1840.
19 The Clopidogrel in Unstable Angina to Prevent
Recurrent Events Trial Investigators: Effects of
clopidogrel in addition to aspirin in patients
with acute coronary syndromes without ST-
segment elevation. N Engl J Med 2001;345:
494–502.
20 Steinhubl SR, Berger PB, Mann JT III, Fry
ETA, DeLago A, Wilmer C, Topol EJ, for the
CREDO Investigators: Early and sustained
dual oral antiplatelet therapy following percu-
taneous coronary intervention. A randomized
controlled trial. JAMA 2002;288:2411–2420.
21 Leon MB, Baim DS, Popma JJ, Gordon PC,
Cutlip DE, Ho KKL, Giambartolomei A, Div-
er DJ, Lasorda DM, Williams DO, Pocock SJ,
Kuntz RE: A clinical trial comparing three
antithrombotic-drug regimens after coronary-
artery stenting. Stent Anticoagulation Resteno-
sis Study Investigators. N Engl J Med 1998;
339:1665–1671.
22 Sievert H, Lesh MD, Trepels T, Omran H, Bar-
torelli A, Bella PD, Nakai T, Reisman M,
DiMario C, Block P, Kramer P, Fleschenberg
D, Krumsdorf U, Scherer D: Percutaneous left
atrial appendage transcatheter occlusion to pre-
vent stroke in high-risk patients with atrial
fibrillation. Circulation 2002;105:1887–1889.
Mattle
 Cerebrovasc Dis 2003;16(suppl 1):9–13
DOI: 10.1159/000069935
Outcome after Brain Haemorrhage
Martin S. Dennis
Department of Clinical Neurosciences, Western General Hospital, Edinburgh, UK
Key Words
Intracerebral haemorrhage W Prognosis
Abstract
Between 10 and 20% of strokes are due to intracerebral
haemorrhage. The 1-month case fatality is about 42% in
unselected cohorts. This relatively low incidence (com-
pared with ischaemic stroke) and high early case fatality
means that relatively few patients are available for long-
term follow-up and therefore the available data on prog-
nosis are imprecise. Moreover, improvements in diag-
nostic methods, such as the introduction of gradient
echo MRI, which is very sensitive to intracerebral haem-
orrhage, are altering the types of patients being entered
into studies of prognosis. Despite these methodological
difficulties, it does appear that the overall prognosis with
respect to survival and residual disability is similar to
that for ischaemic stroke of equivalent clinical severity.
Greater age and stroke severity, whether graded by neu-
rological score or extent of haemorrhage on imaging, are
both associated with increased case fatality and poorer
functional outcomes. There is no definite evidence of dif-
ferential recovery between ischaemic and haemorrhagic
stroke. Epileptic seizures occur more commonly after
haemorrhagic stroke (about 8 per 100 patient-years)
compared with ischaemic stroke and more commonly in
lobar rather than basal ganglia haemorrhage. There is no
reliable evidence to indicate that the risk of recurrent
stroke after haemorrhage differs from that after isch-
© 2003 S. Karger AG, Basel
ABC 1015–9770/03/0165–0009$19.50/0
Fax + 41 61 306 12 34
E-Mail karger@karger.ch Accessible online at:
www.karger.com www.karger.com/ced
aemic stroke. However, strokes due to haemorrhage, like
those due to infarction, are heterogeneous not only in
terms of severity but also in their causes. The causes
(e.g. amyloid angiopathy, hypertension, coagulation def-
icits) are likely to influence the risk of subsequent stroke.
Pooling of data from community-based studies of haem-
orrhagic stroke that have used consistent definitions and
methods represents the only feasible way to obtain more
precise data on prognosis after intracerebral haemor-
rhage.
Copyright © 2003 S. Karger AG, Basel
Introduction
In general, the term primary intracerebral haemor-
rhage (PICH) refers to stroke due to intracranial haemor-
rhage and excludes those that are due to subarachnoid or
subdural haemorrhage, traumatic haematoma, that are
secondary to arterial or venous infarction, or that are due
to bleeding into a tumour. It is unclear whether published
studies on this topic have excluded cases of haemorrhage
due to vasculitis, underlying arteriovenous malformation,
or saccular aneurysm, and if so, how aggressively the
patient was investigated to exclude these aetiologies.
Therefore, as in lacunar disease, there is a difficulty in
defining homogeneous populations of patients with
PICH.
There are several reasons for studying prognosis in
PICH. First, physicians wish to talk to patients and their
Martin Dennis, MD
Department of Clinical Neurosciences
Bramwell Dott Building, Crewe Road
Edinburgh EH4 2XU (UK)
Tel. +44 131 5371000, Fax +44 131 3325150, E-Mail msd@skull.dcn.ed.ac.uk
Table 1. Overview of community-based studies of 1-month case fatality after PICH [6–17]

Study (first author) Location Patient Inclusion criteria Strokes PICH, n (%)
population

Anderson [6] Perth, Australia 138,708 Stroke or transient 536 37 (6.9)

ischaemic attack
Counsell [7] Oxfordshire, UK F105,000 First-ever stroke 675 66 (9.8)
Vemmos [8] Arcadia, Greece 80,774 First-ever stroke 555 77 (13.9)
Kolominsky-Rabas [9] Erlangen, Germany 101,450 First-ever stroke 354 48 (13.6)
Ellekjær [10] Innherred, Norway 69,295 First-ever and 432 first-ever strokes 45 (10.4)
recurrent stroke 161 recurrent strokes
Lauria [11] Belluno, Italy 211,389 First-ever stroke 474 93 (19.6)
D’Alessandro [12] Valle d’Aosta, Italy 114,325 First-ever stroke 254 33 (13.0)
Giroud [13] Dijon, France 150,000 First-ever stroke 984 87 (8.8)
Ricci [14] Umbria, Italy 49,218 First-ever stroke 375 37 (9.9)
Ashok [15] Benghazi, Libya 518,745 First-ever stroke 329 63 (19.1)
Thrift [16] Melbourne, Australia F3,500,000 Consecutive cases of PICH 370 40 (10.8)
Carolei [17] L’Aquila, Italy 297,838 First-ever stroke 819 122 (14.9)

relatives to give them an indication of what may occur in
the future, and to be able to balance risks and benefits of
interventions in an informed manner. Second, informa-
tion on the prognosis of groups of patients is useful for the
design of randomized clinical trials, to estimate sample
size and decide on the optimal inclusion/exclusion crite-
ria. Finally, data on prognosis of patients can be used in
the assessment of quality of care, although this requires
adequate adjustment of outcomes for case mix [1, 2].
Several methodological problems affect studies of the
prognosis of PICH and their results. First, PICH is rela-
tively rare, accounting for only 10–20% of all stroke cases.
Therefore, most studies have generally contained too few
patients to give precise estimates of prognosis or to reli-
ably identify the factors that predict outcome. Second,
PICH is only reliably identified by brain imaging or
autopsy, and the definition of PICH is dependent on the
imaging method and its timing. Third, PICH patients
comprise a heterogeneous group, representing a range of
haemorrhage aetiologies and types. Finally, outcome after
PICH is influenced in the different studies by the treat-
ment given to patients, and this varies greatly according to
geographical location.
The ‘ideal’ study [3] of long-term outcome after PICH
should provide unbiased and precise estimates of progno-
sis. The study should identify all first-ever strokes in a
large, well-defined population, including all patients that
are not referred to hospital. The study should then identi-
fy all cases of PICH in the group of first-ever strokes, to
produce an unselected, representative group at the begin-
ning of the study. This requires either CT scanning within
7 days of stroke onset or a gradient echo MRI (or equiva-
lent) if later in all cases and an autopsy in fatal cases where
such scanning was not possible [4]. Patients should be
identified as soon after the initial stroke as possible, and
patients should be followed prospectively, with frequent
follow-up until death. Stroke recurrences should be char-
acterized clinically, and should be investigated with early
CT or MRI scans to determine if these are haemorrhagic
or ischaemic events. Regular measurements should also
be made regarding impairment, disability and handicap,
and quality of life.
Outcome after PICH
Overview of Current Data
Most of the community-based studies to date have
included only small populations – typically 50,000–
200,000 – resulting in the analysis of relatively small
numbers of first-ever strokes and even smaller numbers of
PICH. Most have been in White populations, so relatively
little is known about the prognosis of PICH in other eth-
nic groups. Hospital-based and less rigorous community-
based ones will not identify all stroke patients occurring in
a defined population. In these circumstances, certain
types of patients will be underrepresented: these include
patients with minor strokes, who are less likely to be
referred, patients with major strokes, who may die with-
out diagnosis, and others who are not admitted to hospi-

10 Cerebrovasc Dis 2003;16(suppl 1):9–13 Dennis

 Study n/N

Perth 13/37
Oxfordshire 34/66
Arcadia 37/77
Erlangen 20/48
Innherred 17/45
Belluno 32/93
Aosta 15/33
Dijon 47/87
Umbria 14/37
Libya 32/63
Melbourne 18/40
L’Aquila 63/122

All 331/748

20 30 40 50 60 Fig. 2. Long-term survival after PICH or cerebral infarction in the

Case fatality(%)
Fig. 1. One-month case fatality in PICH identified in community-
based studies [6–17]. The squares indicate the point estimate and the
horizontal lines indicate the 95% confidence intervals of the esti-
mates. The size of the squares reflects the number of patients with
PICH included in the study. The diamond indicates the pooled esti-
mate.
tal. In the available studies, many did not perform a CT
scan within 7 days of stroke onset, and none routinely had
good access to gradient echo MRI, which is probably the
most sensitive and reliable way of identifying previous
haemorrhage. This may lead to certain types of haemor-
rhage being underrepresented in these study cohorts. Mis-
diagnosis may result from delays. These may be attribut-
able to patient delay in reporting symptoms, delay in
referral to hospital, and delay in imaging to arrive at the
diagnosis. The latter type of delay may not occur in hospi-
tal-based practice, but certainly occurred in the communi-
ty-based stroke registers, which provide the most un-
biased information on prognosis in PICH [5]. Other prob-
lems that potentially affect the studies of prognosis of
PICH include losses to follow-up in some studies, small
numbers of survivors due to high early mortality, lack of
uniform adoption of standard definitions, an absence of
treatment description or recurrence characterization, and
the fact that assessments looking at predictive factors are
often not blinded to patient outcome.
Survival Data
Twelve community-based studies reporting 1-month
case fatality after PICH have been published (table 1) [6–
Oxfordshire Community Stroke Project. Kaplan-Meier plot showing
proportion of patients surviving at increasing intervals after a first-
ever stroke due to PICH or cerebral infarction [reprinted with per-
mission, 19].
17]. The majority (9) of these studies included patients
with first-ever stroke, while one each studied outcome in
patients with stroke or transient ischaemic attack [6],
first-ever and recurrent stroke [10] or in consecutive cases
of PICH [16]. Study populations ranged in size from
F50,000 to 3.5 million patients, and the number of
patients with PICH in these cohorts ranged from 33 to
122. Although the estimates of 1-month case fatality from
the individual studies have wide confidence intervals,
reflecting the small number of PICHs in the individual
cohorts, the pooled estimate is more precise, and gives a
value of 42% for 1-month case fatality (fig. 1).
Several factors have been identified in studies as being
associated with a high early case fatality [18]. These are:
increasing age; clinical severity of PICH; poor pre-stroke
function; volume of PICH; presence of intraventricular
blood, and anticoagulant therapy. There is, of course, a
degree of interrelation between these factors, since, for
example, the volume of PICH and the presence of intra-
ventricular blood are related to the clinical stroke severi-
ty. Moreover, studies which have suggested that patients
receiving anticoagulants die more frequently may not
have adjusted for volume of PICH.
Kaplan-Meier analysis of survival in the Oxfordshire
Community Stroke Project (OCSP) confirms the high ear-
ly mortality in PICH patients compared with ischaemic
stroke patients over the time period (fig. 2) [19]. Over lon-
ger-term (30-day to 5-year) follow-up, the survival curves

Outcome after Brain Haemorrhage Cerebrovasc Dis 2003;16(suppl 1):9–13 11


Fig. 3. Stroke recurrence after PICH in the Oxfordshire Community
Stroke Project [reprinted with permission, 7].
for PICH patients and those with ischaemic stroke appear
to converge. However, this is probably due to a survival
effect, with long-term survivors of PICH being younger
and fitter than the survivors of ischaemic stroke several
years after the index event.
Functional Outcome
Data on 1-year outcomes from the OCSP indicate that
around 25% of PICH patients are independent in their
activities of daily living, and around 10–15% survive but
are dependent. In the same study, patients within the total
anterior circulation infarction (TACI) category had a sim-
ilar case fatality at 1 year, but the majority of survivors
were dependent in everyday activities [20]. The most like-
ly explanation for this is that TACI patients form a homo-
geneous group with clinically severe stroke and high
stroke scale scores, whilst PICH patients are more hetero-
geneous in terms of severity.
In a prospective study of 1,000 unselected patients
with acute stroke, Jorgensen et al. [21] analysed the out-
comes of intracerebral haemorrhage versus infarction,
with adjustment for initial stroke severity. In this analysis,
the relative frequency of PICH rose exponentially with
increasing stroke severity. However, on multivariate anal-
ysis, pathological type (i.e., cerebral infarction versus
PICH) did not independently predict mortality, neurolog-
ical outcome, functional outcome, or the time course of
recovery. These data suggest that the poorer prognosis in
patients with PICH is due to the increase in frequency of
intracerebral haemorrhage with increasing stroke severi-
ty.
12 Cerebrovasc Dis 2003;16(suppl 1):9–13
Recurrence after PICH
Very few data are available on the risk of recurrence
after a first PICH. In the OCSP, the annual risk of recur-
rent stroke between 30 days and 5 years of follow-up was
7% (fig. 3) [7]. However, in this study the sample size was
small, with only 66 patients with PICH at the start of fol-
low-up, leading to wide 95% confidence intervals.
A systematic review of studies reporting recurrent
stroke after PICH has recently been reported by Bailey et
al. [22]. This overview analysed data from three commu-
nity-based populations (total sample size = 146 patients)
and seven hospital-based cohorts (total sample size =
1,734 patients). From these pooled data, the overall stroke
recurrence rate was 4.3% per patient-year, with rates of
6.2% per patient-year in community-based studies and
4.0% per patient-year in hospital-based studies (p = 0.04).
These recurrence rates are very similar to those reported
for recurrence after ischaemic stroke [23]. In the same
overview, over half (59%) of the recurrent strokes were
haemorrhagic, 26% were ischaemic, and 15% were of
unknown aetiology. This finding may have implications
for our approach to secondary prevention and in particu-
lar the use of antithrombotic medication.
Although the overall risk of recurrence in PICH pa-
tients is probably between 4 and 7% per year, it is probable
that certain factors will indicate a higher or lower than
average risk. It seems likely that the site of the first haem-
orrhage, which to some extent reflects the underlying cause
of the haemorrhagic stroke, will be important. For exam-
ple, in the review by Bailey et al. [22], recurrence was more
likely in those with lobar haemorrhage than in those with a
deep hemispheric PICH (4.4 vs. 2.1% per patient-year; p =
0.002). Lobar haemorrhage often results from amyloid
angiopathy which seems to carry a higher risk of recur-
rence. Indeed, many patients may have several haemor-
rhages within a few days or weeks of each other.
Furthermore, it seems likely that patients with abnor-
mal coagulation and those with vascular malformations
(e.g. arteriovenous fistula) will have a higher risk of recur-
rence. In a systematic review by Al Shahi and Warlow
[24], patients who had an arteriovenous malformation
and who presented with a haemorrhage had an 18%
annual risk of recurrent bleed, compared with lower rates
amongst patients with other presentations, such as epilep-
tic seizures. Although not yet formally studied, it is likely
that patients who present with a first stroke due to an
intracerebral haemorrhage but who on gradient echo T2
have many microhaemorrhages will have a higher than
average risk of recurrence. The results of the PROGRESS
trial showed that lowering blood pressure, even within the
Dennis
‘normal’ range, reduced the risk of recurrent stroke [25].
This was particularly marked in those who entered the
study following a haemorrhagic stroke. Thus, blood pres-
sure during follow-up is likely to be a major factor in
determining an individual’s risk of recurrence.
Risk of Epilepsy
Very few data are available on the risk of epileptic sei-
zures after first-ever stroke. In the OCSP, Kaplan-Meier
analysis showed that patients with PICH have a higher risk
of developing seizures than do ischaemic stroke patients
(hazard ratio = 10.2; 95% CI 3.7–27.9) [26]. The risk of
epilepsy is higher in the first year after the index event and
overall, corresponds to about 8 per 100 patient-years. The
risk is probably higher in those with lobar haemorrhage,
which more commonly involves the cerebral cortex.
Conclusions
Current data from community-based studies indicate
that the 1-month case fatality after PICH is around 40–
45% and that haemorrhage severity is the major predictor
of mortality. Functional outcome is similar to that of cere-
bral infarction and again depends on the initial severity of
the stroke. The recurrence rate is perhaps between 4 and
7%, i.e. very similar to that of ischaemic stroke, and prob-
ably depends on the cause and also on blood pressure. The
seizure rate is estimated at 8%, and is probably dependent
on the location of the haemorrhage. Pooling of data from
community-based studies of haemorrhagic stroke that
have used consistent definitions and methods represents
the only feasible way to obtain more precise data on prog-
nosis after PICH.

References

1 Davenport RJ, Dennis MS, Warlow CP: Effect
of correcting outcome data for case mix: An
example from stroke medicine. BMJ 1996;312:
1503–1505.
2 Weir N, Dennis M: Scottish Stroke Outcomes
Group: Towards a national system for monitor-
ing the quality of hospital-based stroke ser-
vices. Stroke 2001;32:1415–1421.
3 Sudlow CLM, Warlow CP: Comparing stroke
incidence worldwide: What makes studies
comparable? Stroke 1996;27:550–558.
4 Wardlaw JM, Keir SL, Dennis MS: The impact
of delays in computed tomography of the brain
on the accuracy of diagnosis and subsequent
management in patients with minor stroke. J
Neurol Neurosurg Psychiatry 2003;74:77–81.
5 Keir SL, Wardlaw JM, Warlow CP: Stroke epi-
demiology studies have underestimated the fre-
quency of intracerebral haemorrhage. A sys-
tematic review of imaging in epidemiological
studies. J Neurol 2002;249:1226–1231.
6 Anderson CS, Jamrozik KD, Burvill PW, Cha-
kera TM, Johnson GA, Stewart-Wynne EG:
Determining the incidence of different sub-
types of stroke: Results from the Perth Com-
munity Stroke Study, 1989–1990. Med J Aust
1993;158:85–89.
7 Counsell C, Boonyakarnkul S, Dennis M, San-
dercock P, Bamford J, Burn J, Warlow C: Pri-
mary intracerebral haemorrhage in the Oxford-
shire Community Stroke Project. 2. Prognosis.
Cerebrovasc Dis 1995;5:26–34.
8 Vemmos KN, Bots ML, Tsibouris PK, Zis VP,
Grobbee DE, Stranjalis GS, Stamelopoulos S:
Stroke incidence and case fatality in southern
Greece. The Arcadia Stroke Registry. Stroke
1999;30:363–370.
9 Kolominsky-Rabas PL, Sarti C, Heuschmann
PU, Graf C, Siemonsen S, Neundoerfor B, Ka-
talinc A, Lang E, Gassmann KG, von Stockert
TR: A prospective community-based study of
stroke in Germany – The Erlangen Stroke Pro-
ject (ESPro). Incidence and case fatality at 1, 3
and 12 months. Stroke 1998;29:2501–2506.
10 Ellekjær H, Holmen J, Indredavik B, Terent A:
Epidemiology of stroke in Innherred, Norway,
1994–1996: Incidence and 30-day case-fatality
rate. Stroke 1997;28:2180–2184.
11 Lauria G, Gentile M, Fassetta G, Casetta I,
Agnoli F, Andreotta G, Barp C, Caneve G,
Cavallero A, Cielo R, Mongillo D, Mosca M,
Olivieri PG: Incidence and prognosis of stroke
in the Belluno Province, Italy. First-year results
of a community-based study. Stroke 1995;26:
1787–1793.
12 D’Alessandro G, Di Giovanni M, Roveyaz L,
Iannizzi L, Pesenti Compagnoni M, Blanc S,
Bottacchi E: Incidence and prognosis of stroke
in the Valle d’Aosta, Italy: First-year results of
a community-based study. Stroke 1992;23:
1712–1715.
13 Giroud M, Gras P, Chadan N, Beuriat P, Milan
C, Arveux P, Duams R: Cerebral haemorrhage
in a French prospective population study. J
Neurol Neurosurg Psychiatry 1991;54:595–
598.
14 Ricci S, Celani MG, La Rosa F, Vitali R, Duca
E, Ferraguzzi R, Paolotti M, Seppoloni D, Ca-
puto N, Chiurulla C, Scaroni R, Signorini E:
SEPIVAC: A community-based study of stroke
incidence in Umbria, Italy. J Neurol Neurosurg
Psychiatry 1991;54:695–698.
15 Ashok PP, Radhakrishnan K, Sridharan R, El-
Mangoush MA: Incidence and pattern of cere-
brovascular disease in Benghazi, Libya. J Neu-
rol Neurosurg Psychiatry 1986;49:519–523.
16 Thrift AG, Dewey HM, Macdonell RAL,
McNeil JJ, Donnan GA: Stroke incidence on
the East Coast of Australia. The North East
Melbourne Stroke Incidence Study (NEME-
SIS). Stroke 2000;31:2087–2092.
17 Carolei A, Marini C, Di Naploi M, Di Gianfil-
ippo G, Santalucia P, Baldassarre M, De Mat-
teis G, di Orio F: High stroke incidence in the
prospective community-based L’Aquila regis-
try (1994–1998). First year’s results. Stroke
1997;28:2500–2506.
18 Counsell C, Dennis M: Systematic review of
prognostic models in patients with acute
stroke. Cerebrovasc Dis 2001;12:159–170.
19 Dennis MS, Burn JPS, Sandercock PAG, Bam-
ford JM, Wade DT, Warlow CP: Long-term
survival after first-ever stroke: The Oxfordshire
Community Stroke Project. Stroke 1993;24:
796–800.
20 Warlow CP, Dennis MS, Van Gijn J, Hankey
G, Bamford J, Wardlaw J: A practical approach
to management of stroke patients; in Stroke: A
Practical Guide to Management, ed 2. Oxford,
Blackwell Scientific, 2000, pp 414–441.
21 Jorgensen HS, Nakayama H, Raaschou HO,
Olsen TS: Intracerebral hemorrhage versus in-
farction: Stroke severity, risk factors and prog-
nosis. Ann Neurol 1995;38:45–50.
22 Bailey RD, Hart RG, Benavente O, Pearce LA:
Recurrent brain hemorrhage is more frequent
than ischemic stroke after intracranial hemor-
rhage. Neurology 2001;56:773–777.
23 Burn J, Dennis M, Bamford J, Sandercock P,
Wade D, Warlow C: Long-term risk of recur-
rent stroke after a first-ever stroke. Stroke
1994;25:333–337.
24 Al-Shahi R, Warlow C: A systematic review of
the frequency and prognosis of arteriovenous
malformations of the brain in adults. Brain
2001;124:1900–1926.
25 PROGRESS Collaborative Group: Random-
ised trial of a perindopril-based blood-pres-
sure-lowering regimen among 6,105 individu-
als with previous stroke or transient ischaemic
attack. Lancet 2001;358:1033–1041.
26 Burn J, Dennis M, Bamford J, Sandercock P,
Wade D, Warlow C: Epileptic seizures after a
first stroke: The Oxfordshire Community
Stroke Project. BMJ 1997;315:1582–1587.

Outcome after Brain Haemorrhage Cerebrovasc Dis 2003;16(suppl 1):9–13 13

 Cerebrovasc Dis 2003;16(suppl 1):14–19
DOI: 10.1159/000069936
Long-Term Outcome after Ischaemic
Stroke/Transient Ischaemic Attack
Graeme J. Hankey
Stroke Unit, Royal Perth Hospital and Department of Medicine, University of Western Australia, Perth, WA, Australia
Key Words
Ischaemic stroke W Prognosis W Transient ischaemic
attack
Abstract
During the first 30 days after a stroke, the case fatality is
about 25% and the major cause of death is the index
stroke and its sequelae. The most consistent predictor of
30-day mortality after stroke is stroke severity. Other pre-
dictors include increasing age, a history of previous
stroke, cardiac failure, and a high blood glucose concen-
tration and white blood cell count. Other less common,
but important, causes of early mortality are recurrent
ischaemic stroke and a coronary event. The risk of a
recurrent cerebrovascular event is highest in the first
month (4%) and year (12%) after a stroke and transient
ischaemic attack (TIA), probably reflecting the presence
of active, unstable atherosclerotic plaque. Thereafter, the
risk of a recurrent cerebrovascular event falls to about
5% per year, similar to the risk of a coronary event. Dur-
ing years 1–5 after a TIA and ischaemic stroke, cardiovas-
cular disease increasingly becomes the major cause of
death, reflecting the generalized nature of atherothrom-
bosis, the most common cause of the index stroke. The
most robust predictor of death within 1–5 years after
stroke is increasing age, closely followed by cardiac fail-
© 2003 S. Karger AG, Basel
ABC 1015–9770/03/0165–0014$19.50/0
Fax + 41 61 306 12 34
E-Mail karger@karger.ch Accessible online at:
www.karger.com www.karger.com/ced
ure. Additional baseline predictors of longer-term mor-
tality include a history of previous symptomatic athero-
thrombosis (TIA, ischaemic stroke, peripheral arterial
disease, and early-onset ischaemic heart disease), risk
factors for atherothrombosis (smoking), other heart dis-
eases (cardiac failure, atrial fibrillation) and increasing
stroke severity. Lacunar syndromes can be predictive of
relative longevity. At 5 years after stroke, survival is
about 40%, and about half of survivors are disabled and
dependent. The most robust predictors of disability at 5
years after stroke are increasing age, stroke severity, and
recurrent stroke. The most powerful predictor of early
recurrent stroke (within 30 days after stroke) is an athero-
sclerotic ischaemic stroke caused by large-artery athero-
sclerosis with 150% stenosis, whereas the strongest pre-
dictor of stroke recurrence over 5 years is diabetes. Other
predictors of recurrent stroke include increasing age,
previous TIA, atrial fibrillation, high alcohol consump-
tion, haemorrhagic index stroke, and hypertension at
discharge. The clinical implication of these findings is
that strategies for optimizing long-term outcome after
TIA and stroke should be directed toward reducing the
high risk of recurrent stroke and coronary events by
removing/recanalizing the symptomatic atherosclerotic
plaque, controlling the underlying causal vascular risk
factors, and administering long-term, effective antiplate-
let therapy.
Copyright © 2003 S. Karger AG, Basel
Graeme J. Hankey, MD
Stroke Unit, Department of Neurology
Royal Perth Hospital, Wellington Street
Perth, WA 6001 (Australia)
Tel. +61 8 9224 2598, Fax +61 8 9224 3323, E-Mail gjhankey@cyllene.uwa.edu.au
 Introduction
The long-term prognosis for survival free of a recurrent
stroke and coronary event after a stroke is of importance
to patients and their carers, relatives and clinicians. For
the clinician, reliable information on long-term prognosis
helps to guide patient management. It facilitates: (1) the
selection of appropriate and treatments (particularly
those that are risky or expensive); (2) provision of reliable
information to patients and their relatives; (3) compari-
sons of patient outcome with the outcome of other
patients (by allowing adjustment for case mix); (4) im-
proved design and analysis of clinical trials (e.g. risk strat-
ification and sample size calculation), and (5) improved
understanding of the underlying causal disease, which is
atherothrombosis in most cases of stroke.
The ideal study of the prognosis after transient isch-
aemic attack (TIA) and stroke is one that is characterized
by: (1) complete ascertainment of all cases of first-ever
stroke and TIA from large community-based inception
cohorts in which the standard diagnostic criteria and
details of disease severity, co-morbidity and socio-demo-
graphic factors are described at inception; (2) patients are
followed up at regular intervals prospectively and com-
pletely; (3) valid, reliable and objective and standardized
measures of important outcomes which are recorded pro-
spectively and blinded to the study hypotheses, and (4)
actuarial methods of survival analyses are used, with
adjustment for extraneous prognostic factors [1].
Since 1966, more than 250 published studies have
described prognostic models of outcome after stroke, but
the majority (around three-quarters) of these are not inter-
nally valid, and all but 3% (12% of the remainder) are
prone to hospital-referral selection bias [2]. The number
of robust studies on long-term outcome after TIA or
Table 1. Community-based studies of long-term prognosis after TIA [3–5]
Study Location Time period
Terént [3] Söderhamn, Sweden 1975–1979/
1983–1987
Dennis et al. [4] Oxfordshire, UK 1981–1986
Ricci et al. [5] Umbria, Italy 1986–1989
a Data are for two successive 3-year periods.
Outcome after Ischaemic Stroke/TIA
stroke is similarly limited. The aim of this article is to
review current evidence from the small proportion of
high-quality studies on the prognosis of patients with TIA
or ischaemic stroke and the predictors of short- and long-
term mortality, disability and recurrent stroke.
Long-Term Stroke and Cardiac Risk in
Stroke/TIA Patients
Prospective, Community-Based Studies of Prognosis
of TIA
There have been three prospective, community-based
studies of the long-term prognosis after TIA (table 1).
These were undertaken in Söderhamn, Sweden [3], Ox-
fordshire, UK [4] and Perugia, Italy [5] over two succes-
sive 3-year periods, 5 years and 9 years, respectively. The
average annual risk of stroke was 5.3% (95% CI 3.1–
7.4%) over each of the two 3-year periods in Söderhamn,
6.7% (95% CI 4.7–8.9%) over 5 years in Oxfordshire, and
2.4% (95% CI 0.7–4.7%) over 9 years in Perugia. The risk
of stroke was highest within the first month (4.4% (95%
CI 1.5–7.3%) in Oxfordshire) and first year (11.6% (95%
CI 6.9–16.3%) in Oxfordshire) after TIA and fell thereaf-
ter to about 2–3% per year during subsequent years. The
average annual risk of death after MI was 2.5% in both
Oxfordshire and Perugia (no data were reported for Sö-
derhamn). In contrast to stroke, there was no excess early
risk of MI. The average annual risk of death was 4.9%
(95% CI 0.1–10.3%) and 7.3% (95% CI 1.1–14.4%) over
two 3-year periods in Söderhamn, 7.2% (95% CI 5.2–
9.5%) over 5 years in Oxfordshire, and 7.0% over 9 years
in Perugia. The proportion of fatalities due to cardiac
causes was similar to the proportion caused by stroke in
Oxfordshire (35 vs. 31%) and Perugia (22 vs. 20%).
Patients Follow-up Annual event rate, %
years
stroke MI death
97 3 5.3, 5.3a – 4.9, 7.3a
184 5 6.7 (4.7–8.9) 2.5 (1.2–4.0) 7.2 (5.2–9.5)
94 9 2.4 2.5 7.0
Cerebrovasc Dis 2003;16(suppl 1):14–19 15

Fig. 1. Event rate curves for survival free of stroke (a) and coronary
events (b) in hospital-referred TIA patients [reproduced with permis-
sion, 7].
Prospective, Hospital-Based Studies of Prognosis of TIA
The scant information from community-based studies
of TIA prognosis is supplemented by data from four major
hospital-based studies recording stroke and cardiac events
after TIA (table 2) [6–9]. In these studies, which followed
larger cohorts of younger patients compared with the com-
munity-based cohorts, the annual rate of stroke was 2.2–
5.0% over 4–5 years, which is similar to the risk of MI
(1.1–4.6%) over the same period. In the study in Oxford-
16 Cerebrovasc Dis 2003;16(suppl 1):14–19
Fig. 2. Proportion of patients dying from different causes during dif-
ferent time intervals from the onset of their first-ever stroke in the
Perth Community Stroke Study [reproduced with permission, 13].
shire [7], Kaplan-Meier analysis for stroke-free survival
reveals a high risk of stroke in the first 1–6 months after
stroke (fig. 1), which probably reflects active, unstable ath-
erosclerotic plaque. In contrast, the Kaplan-Meier curve
for coronary events exhibits a long-term linear trend after
TIA (fig. 1). In prospective hospital-based studies of TIA,
the most important cause of death is observed to be car-
diac disease (42%), followed by stroke (23%).
Prospective Studies of Stroke
There have been 11 community-based studies of the
long-term prognosis after all stroke, first-ever stroke, or
first-ever ischaemic stroke. Six of these recorded stroke
and cardiac events over 5 years (table 3) [10–15]. At 5
years of follow-up, the proportion of deaths due to cardiac
causes was equal to, or greater than, the proportion due to
stroke (the index event and recurrent stroke). In the Perth
Community Stroke Study, the proportion of death due to
stroke decreased with the passage of time after the index
stroke, and the proportion of deaths due to coronary
events and non-vascular events increased over time
(fig. 2). The average annual risk of recurrent stroke was
about 6% (95% CI 4–8%) over 5 years. The risk of stroke
was highest within the first 6 months (9%) and first year
(13–14%) after stroke, and fell thereafter to about 4–5%
per year during subsequent years.
Hankey
Table 2. Hospital-based studies of long-
Study Time period Patients Mean age Follow-up Annual event rate, %
term prognosis after TIA [6–9]
years years
stroke MI

Heyman et al. [6] 1974–1978 390 61.7 5 5.0 4.6

Hankey et al. [7] 1977–1986 469 62.1 5 3.4 3.1
Carolei et al. [8] 1977–1980 712 55.7 4 2.2 1.1
Howard et al. [9] 1987–1991 280 64 4 3.0 4.2

Table 3. Overview of community-based studies of stroke, recording stroke and cardiac events over 5 years [10–15]

Study Location Stroke type Time Patients Mortality at Causes of death, %

period 5 years, %
stroke cardiac

Scmidt et al. [10] Moscow, Russia First stroke 1972–1974 941 72 23 42

Burn et al. [11]a Oxfordshire, UK First stroke 1981–1986 675 45 36 34
Petty et al. [12] Rochester, USA First ischaemic stroke 1975–1989 1,111 53 27 24
Hankey et al. [13] Perth, Australia First stroke 1989–1995 370 60 27 31
Hartmann et al. [14]a Northern Manhattan, USA First ischaemic stroke 1990–1997 980 41 15 29
Brønnum-Hansen Copenhagen County, All stroke 1982–1991 4,162 60 32 23
et al. [15] Denmark

a Data are for causes of death in the first 30 days after index event.

Predictive Factors for Stroke Prognosis
Predictors of Mortality
Current findings on short- and long-term predictors of
death, disability and recurrent stroke are summarized in
table 4. Probably the most reliable data on mortality
comes from Olmsted County, Minnesota, USA, where
30-day and 5-year follow-up data between 1985 and 1989
have been reported for a cohort of 454 patients with first-
ever ischaemic stroke [12], and 10-year follow-up data
are available for a cohort of 1,111 first-ever ischaemic
stroke patients followed between 1975 and 1989 [16].
For death at 30 days after ischaemic stroke, the two
major predictors of early mortality were age (relative risk
(RR)/10 years = 1.3; 95% CI 1.03–1.7) and severe stroke
as measured by the Rankin scale (RR = 11.6; 95% CI
2.8–49). For mortality at 5 years, age and severe stroke
continued to predict outcome. In addition, congestive
heart failure (RR = 1.7; 95% CI 1.2–2.3) and early-onset
ischaemic heart disease (RR = 5.0; 95% CI 1.8–13) also
had statistically significant predictive power. Ischaemic
stroke caused by large-artery atherosclerosis with 150%
stenosis was associated with a lower risk of 5-year mortal-
ity (RR = 0.5; 95% CI 0.3–0.8). From the study with 10-
year follow-up data, age at presentation continued to
have predictive power for long-term mortality, whereas
stroke severity did not. However, atrial fibrillation, con-
gestive heart failure, and early- and late-onset ischaemic
heart disease at baseline were significant predictors of 10-
year mortality.
Five-year follow-up data are also available from the
Northern Manhattan Stroke Study [17], which analyzed
predictors of outcome in 323 ischaemic stroke patients
between 1983 and 1988. In this cohort, the predictors of
death at 30 days were again stroke severity (as assessed by
depressed consciousness, major hemispheric or basilar
syndrome, and hyperglycaemia at hospital admission), as
well as congestive heart failure. At 5 years of follow-up,
stroke severity and congestive heart failure continued to
be predictors of mortality, and increasing age also had sta-
tistically significant predictive power. Patients who pre-
sented with lacunar ischaemic stroke syndromes had sig-
nificantly lower mortality at 5 years after the index
stroke.
In the Perth Community Stroke Study, data were avail-
able for 1- and 5-year mortality [13, 18]. Predictors of
death at 1 year were increasing stroke severity and a histo-
ry of previous cardiac disease (atrial fibrillation or conges-

Outcome after Ischaemic Stroke/TIA Cerebrovasc Dis 2003;16(suppl 1):14–19 17

 Table 4. Short- and long-term predictors of death, disability and stroke recurrence

Outcome Predictors
short-term long-term

Death Severe stroke Increasing age

Impairments (coma, weakness, incontinence) Cardiac disease
Syndromes (TACS, POCS) Cardiac failure
Laboratory findings (blood glucose level, white cell count) Ischaemic heart disease
Atrial fibrillation
Syndromes (LACS protective)
Disability Severe stroke Increasing age
Impairments (coma, weakness, incontinence) Stroke severity
Syndromes (TACS, POCS) Recurrent stroke
Laboratory findings (blood glucose level, white cell count)
Pre-stroke dependency and not living alone
Recurrent stroke Large-artery atherosclerotic ischaemic stroke Diabetes mellitus
Previous TIA
Atrial fibrillation
Hypertension at discharge
High alcohol consumption
Intracerebral haemorrhage
Increasing age

LACS = Lacunar syndrome; TACS = total anterior circulation syndrome; POCS = posterior circulation syn-
drome.

tive heart failure). At 5 years, increasing stroke severity at
baseline was again predictive of mortality, together with
evidence at baseline of multi-bed vascular disease was
also predictive (previous TIA: RR = 1.9 (95% CI 1.3–2.9);
peripheral arterial disease: RR = 1.7 (95% CI 1.2–2.5);
cardiac failure: RR = 1.6 (95% CI 1.1–2.3)). These data
indicate that long-term prevention of coronary events is
essential for long-term reduction of mortality in stroke
patients. Other community-based prediction models for
mortality have been reported for shorter follow-up peri-
ods (30 days to 3 years) [19–24].
Predictors of Disability
In most models, the major predictor of early (30 days
after stroke) dependency is stroke severity. In the Oxford-
shire Community Stroke Project [22, 25], which is the
only externally validated predictive model, pre-stroke in-
dependence, and normal Glasgow Coma Scale verbal
score, arm power, and ability to walk were significant pre-
dictors of survival free of dependency at 6 months, where-
as increasing age and living alone were associated with a
poorer outcome for this parameter. In the Perth Commu-
nity Stroke Study, the major predictor of new disability at
5 years was recurrent stroke (odds ratio = 12.4, 95% CI
3.3–63); other independent predictors were increasing age
and stroke severity [26].
Predictors of Stroke Recurrence
Around 12% of patients have a recurrent stroke within
the first year and 30% experience recurrence over 5 years
[11, 27]. In Rochester, Minnesota, the major baseline pre-
dictor of recurrent stroke at 30 days was atherosclerotic
ischaemic stroke with 150% stenosis (RR = 3.3, 95% CI
1.2–9.3), whereas at 5 years, the major predictor was dia-
betes (RR = 1.9, 95% CI 1.2–2.8) [12]. At 10 years in
Rochester, the major predictors were diabetes (RR = 1.7,
95% CI 1.3–2.2) and increasing age (RR/10 years = 1.2,
95% CI 1.1–1.4) [16].
In the Northern Manhattan Stroke Study, baseline pre-
dictors of recurrent stroke at 5 years were high alcohol
consumption (RR = 2.5, 95% CI 1.4–4.4), hypertension at
discharge (RR = 1.6, 95% CI 1.01–2.6), and blood glucose
concentration at hospital admission (RR/10 mg/dl incre-
ment = 1.4, 95% CI 1.1–1.7) [17].
The Perth Community Stroke Study also found older
age, 75–84 years (RR = 2.6, 95% CI 1.1–6.2) and diabetes
(RR = 2.1, 95% CI 1.0–4.4) to be major predictors of
recurrent stroke at 5 years, as was intracerebral haemor-

18 Cerebrovasc Dis 2003;16(suppl 1):14–19 Hankey

rhage (RR = 2.1, 95% 1.0–4.4) [27]. Baseline predictors of
stroke recurrence at 5 years in the Malmö Stroke Registry
were diabetes (RR = 1.6, 95% CI 1.1–2.5), TIA (RR = 1.6,
95% CI 1.1–2.4) and atrial fibrillation (RR = 1.8, 95% CI
1.1–2.9) [23].
Conclusions
Data from prospective, community-based studies indi-
cate that patients with TIA and ischaemic stroke are at
risk of a recurrent event affecting the brain and ischaemic
events involving the coronary arteries. The risk of a recur-
rent cerebrovascular event is higher within the first month
(and year) after TIA and stroke, but thereafter the risk of a
cardiac event becomes equal, if not greater. Furthermore,
non-cerebral cardiovascular disease becomes the major,
and ever-increasing, cause of death amongst patients with
TIA and ischaemic stroke with the passage of time. The
major predictors of long-term mortality and morbidity
are ischaemic heart disease and its complications, and
causal risk factors for atherothrombosis. Atherothrombo-
sis is a multifocal, systemic disease, which represents the
major cause of index and recurrent stroke and the major
cause of cardiac events. Strategies for secondary pre-
vention after TIA/stroke should be directed toward re-
moving the symptomatic disease, which is usually athero-
thrombosis, control of major causal risk factors for ath-
erogenesis, and optimal antiplatelet therapy.

References

1 Sackett DL, Haynes RB, Guyatt GH, Tugwell
MD: Clinical Epidemiology. A Basic Science
for Clinical Medicine, ed 2. Boston, Little,
Brown, 1991, pp 173–185.
2 Counsell C, Dennis M: Systematic review of
prognostic models in patients with acute
stroke. Cerebrovasc Dis 2001;12:159–170.
3 Terént A: Survival after stroke and transient
ischemic attacks during the 1970s and 1980s.
Stroke 1989;20:1320–1326.
4 Dennis M, Bamford J, Sandercock P, Warlow
C: Prognosis of transient ischemic attacks in
the Oxfordshire Community Stroke Project.
Stroke 1990;21:848–853.
5 Ricci S, Cantisani AT, Righetti E, Duca E,
Spizzichino L: Long-term follow-up of TIAs:
The SEPIVAC Study. Neuroepidemiology
1998;17:31–54.
6 Heyman A, Wilkinson WE, Hurwitz BJ,
Haynes CS, Utley CM, Rosati RA, Burch JG,
Gore TB: Risk of ischemic heart disease in
patients with TIA. Neurology 1984;34:626–
630.
7 Hankey GJ, Slattery JM, Warlow CP: The
prognosis of hospital-referred transient isch-
aemic attacks. J Neurol Neurosurg Psychiatry
1991;54:793–802.
8 Carolei A, Candelise L, Fiorelli M, Francucci
BM, Motolese M, Fieschi C: Long-term prog-
nosis of transient ischemic attacks and revers-
ible ischemic neurologic deficits: A hospital-
based study. Cerebrovasc Dis 1992;2:266–
272.
9 Howard G, Evans GW, Rouse JR III, Toole JF,
Ryu JE, Tegeler C, Frye-Pierson J, Mitchell E,
Sanders L: A prospective reevaluation of tran-
sient ischemic attacks as a risk factor for death
and fatal or nonfatal cardiovascular events.
Stroke 1994;25:342–345.
10 Scmidt EV, Smirnov VE, Ryabova VS: Results
of the seven-year prospective study of stroke
patients. Stroke 1988;19:942–949.
11 Burn J, Dennis M, Bamford J, Sandercock P,
Wade D, Warlow C: Long-term risk of recur-
rent stroke after a first-ever stroke. The Oxford-
shire Community Stroke Project. Stroke 1994;
25:333–337.
12 Petty GW, Brown RD Jr, Whisnant JP, Sicks
JD, O’Fallon WM, Wiebers DO: Ischemic
stroke subtypes. A population-based study of
functional outcome, survival and recurrence.
Stroke 2000;31:1062–1068.
13 Hankey GJ, Jamrozik K, Broadhurst RJ,
Forbes S, Burvill PW, Anderson CS, Stewart-
Wynne EG: Five-year survival after first-ever
stroke and related prognostic factors in the
Perth Community Stroke Study. Stroke 2000;
31:2080–2086.
14 Hartmann A, Rundek T, Mast H, Paik MC,
Boden-Albala B, Mohr JP, Sacco RL: Mortality
and causes of death after first ischemic stroke.
Neurology 2001;57:2000–2005.
15 Brønnum-Hansen H, Davidsen M, Thorvald-
sen P, for the Danish MONICA Study Group:
Long-term survival and causes of death after
stroke. Stroke 2001;32:2131–2136.
16 Petty GW, Brown RD Jr, Whisnant JP, Sicks
JD, O’Fallon WM, Wiebers DO: Survival and
recurrence after first cerebral infarction: A pop-
ulation-based study in Rochester, Minnesota,
1975 through 1989. Neurology 1998;50:208–
216.
17 Sacco RL, Shi T, Zamanillo MC, Kargman DE:
Predictors of mortality and recurrence after
hospitalized cerebral infarction in an urban
community: The Northern Manhattan Stroke
Study. Neurology 1994;44:626–634.
18 Anderson CS, Jamrozik KD, Broadhurst RJ,
Stewart-Wynne EG: Predicting survival for 1
year among different subtypes of stroke. Re-
sults from the Perth Community Stroke Study.
Stroke 1994;25:1935–1944.
19 Czlonkowska A, Ryglewicz D, Lechowicz W:
Basic analytical parameters as the predictive
factors for 30-day case fatality rate in stroke.
Acta Neurol Scand 1997;97:121–124.
20 Carlberg B, Asplund K, Hagg E: The prognostic
value of admission blood pressure in patients
with acute stroke. Stroke 1993;24:1372–1375.
21 Nakayama H, Jorgensen HS, Raaschou HO,
Olsen TS: The influence of age on stroke out-
come. The Copenhagen Stroke Study. Stroke
1994;25:808–813.
22 Counsell C, Dennis M, McDowall M, Warlow
C: Predicting outcome after acute and subacute
stroke. Development and validation of new
prognostic models. Stroke 2002;33:1041–
1047.
23 Elneihoum AM, Göransson M, Flake P, Janzon
L: Three-year survival and recurrence after
stroke in Malmö, Sweden. An analysis of
Stroke Registry Data. Stroke 1998;29:2114–
2117.
24 Bonita R, Ford MA, Stewart AW: Predicting
survival after stroke: A three-year follow-up.
Stroke 1988;19:669–673.
25 The FOOD Trial Collaboration: Performance
of a statistical model to predict stroke outcome
in the context of a large, simple, randomized,
controlled trial of feeding. Stroke 2003;34:127–
133.
26 Hankey GJ, Jamrozik K, Broadhurst RJ,
Forbes S, Anderson CS: Long-term disability
after first-ever stroke and related prognostic
factors in the Perth Community Stroke Study,
1989–1990. Stroke 2002;33:1034–1040.
27 Hankey GJ, Jamrozik K, Broadhurst RJ,
Forbes S, Burvill PW, Anderson CS, Stewart-
Wynne EG: Long-term risk of first recurrent
stroke in the Perth Community Stroke Study.
Stroke 1998;29:2491–2500.

Outcome after Ischaemic Stroke/TIA Cerebrovasc Dis 2003;16(suppl 1):14–19 19

 Cerebrovasc Dis 2003;16(suppl 1):20–26
DOI: 10.1159/000069937
Evidence with Antiplatelet Therapy and
ADP-Receptor Antagonists
J. Donald Easton
Department of Neurology, Brown University, Providence, R.I., USA
Key Words
Acetylsalicylic acid W Antiplatelet therapy W Clopidogrel W
Stroke
Abstract
Antiplatelet drugs have been shown to prevent a range
of atherothrombotic events, including transient isch-
aemic attack (TIA) and ischaemic stroke. Clopidogrel and
ticlopidine are adenosine diphosphate (ADP)-receptor
antagonists that inhibit ADP-induced fibrinogen binding
to platelets, a necessary step in the platelet aggregation
process. The Antithrombotic Trialists’ Collaboration re-
cently published a major meta-analysis that assessed the
effect of antiplatelet therapy in patients with various
manifestations of atherosclerosis. In total, this analysis
included 135,000 patients in comparisons of antiplatelet
agents versus control and 77,000 patients in compari-
sons of different antiplatelet regimens. This meta-analy-
sis found that overall, antiplatelet therapy reduces the
combined odds of stroke, myocardial infarction (MI) or
vascular death by 22%, and that antiplatelet agents
reduce the odds of a non-fatal stroke by 25% over a wide
range of patients with or without a history of cerebrovas-
cular disease. In the CAPRIE trial of clopidogrel versus
acetylsalicylic acid (ASA), there was a 10% odds reduc-
tion for stroke, MI or vascular death in favour of clopido-
grel (p = 0.03). In a meta-analysis performed by the Coch-
© 2003 S. Karger AG, Basel
ABC 1015–9770/03/0165–0020$19.50/0
Fax + 41 61 306 12 34
E-Mail karger@karger.ch Accessible online at:
www.karger.com www.karger.com/ced
rane Stroke Group, ADP-receptor antagonist therapy sig-
nificantly reduced the odds of a serious vascular event
(stroke, MI or vascular death) by 9% (2p = 0.01) and of
any stroke by 12%. The safety/tolerability profile of clopi-
dogrel was superior to that of ticlopidine, and at least as
good as that of ASA. In CURE, a long-term benefit was
observed with the use of clopidogrel on top of standard
therapy (including ASA in all patients), with a 20% rela-
tive risk reduction for the primary endpoint of cardiovas-
cular death, MI or stroke (p ! 0.001) in patients with
unstable angina and non-Q-wave MI. A consistent bene-
fit was seen across all patient subgroups, including
patients with a previous history of stroke. More recently,
CREDO has demonstrated the incremental benefit of pro-
longed use of clopidogrel on top of ASA in patients
undergoing elective PCI, with a 27% reduction in the
combined risk of death, MI or stroke after 12 months of
therapy (p = 0.02) and a 25% reduction in stroke over the
same time period. The MATCH trial is currently being
conducted to test the hypothesis that long-term adminis-
tration of clopidogrel on top of ASA is superior to clopi-
dogrel alone for the reduction of major ischaemic events
in patients with recent TIA or ischaemic stroke who are at
high risk of atherothrombotic recurrence. Further trials of
clopidogrel on top of standard therapy (including ASA)
are planned in neurology; these include SPS3, in patients
with small subcortical strokes, and ATARI, in patients
who have recently recovered from a TIA.
Copyright © 2003 S. Karger AG, Basel
J. Donald Easton, MD
Rhode Island Hospital
110 Lockwood Street, Suite 324
Providence, RI 02903 (USA)
Tel. +1 401 4448795, Fax +1 401 4448781, E-Mail j_easton@brown.edu
 Introduction
Atherothrombosis is the acute pathophysiological pro-
cess whereby platelet activation and aggregation occurs at
the site of a ruptured or eroded atherosclerotic plaque [1].
The clinical manifestations of atherothrombosis include
transient ischaemic attack (TIA), ischaemic stroke, unsta-
ble angina, myocardial infarction (MI) and intermittent
claudication [2], and patients with symptomatic disease
in one vascular bed are liable to have diffuse disease, plac-
ing them at risk of subsequent events in additional vascu-
lar territories. This is illustrated in the case of stroke
patients, in whom the long-term risk of cardiac disease
matches or exceeds that of recurrent stroke [3–7]. Anti-
platelet therapy has been shown to prevent a broad range
of atherothrombotic events, and international guidelines
recommend that all patients who have experienced stroke
or TIA should receive regular antiplatelet therapy to
reduce the risk of recurrent stroke and other vascular isch-
aemic events [8, 9].
The pivotal role of the platelet in chronic atherogenesis
and its acute complications has led to the development
and evaluation of a range of antiplatelet agents. Currently
available oral antiplatelet agents for stroke prevention
include acetylsalicylic acid (ASA), the adenosine diphos-
phate (ADP)-receptor antagonists clopidogrel and ticlo-
pidine, and dipyridamole (either alone or in combination
with low-dose ASA). In contrast to ASA, which inhibits
cyclooxygenase activity and ultimately the thromboxane
A2 pathway of platelet activation [10], clopidogrel, like
ticlopidine, is a specific inhibitor of the P2Y12 ADP
receptor on human platelets, resulting in irreversible inhi-
bition of binding of ADP to the platelet membrane [11].
The mechanism of action of dipyridamole is less well
defined, and this agent has been reported to exhibit both
antiplatelet and vasodilatory effects [12].
Antithrombotic Trialists’ Collaboration
Meta-Analysis
The Antithrombotic Trialists’ Collaboration recently
published a major updated meta-analysis that assessed
the effects of antiplatelet drugs in patients with various
manifestations of atherosclerosis, including TIA and isch-
aemic stroke, as well as many other types of patient who
are at risk for atherothrombotic events [13]. The final
analysis included 287 randomized studies involving
135,000 patients in comparisons of antiplatelet therapy
versus control and 77,000 patients in randomized, head-
Evidence with ADP-Receptor Antagonists
to-head comparisons of different antiplatelet regimens.
Data from recent large trials such as CAPRIE [14], ESPS2
[15], CAST [16] and IST [17] were included.
The latest findings reinforced the messages of the pre-
vious meta-analysis [18]. Moreover, the update incorpo-
rated substantial new data on patients with a history of
stroke or TIA, those treated early after an acute stroke,
and those with stable angina, atrial fibrillation, peripheral
arterial disease, or diabetes, extending the range of patient
types for whom antiplatelet therapy is of clear benefit.
Overall, antiplatelet therapy reduced the overall odds of a
serious vascular event (stroke, MI or vascular death) by
22%, with odds reductions of 25% for non-fatal stroke,
34% for non-fatal MI and 15% for vascular mortality [13].
Based on trials comparing ASA versus control, the risk of
bleeding appeared to be dose-related, but there was no
evidence of heterogeneity with regards to efficacy accord-
ing to ASA dose ranges. Interestingly, a similar effect was
observed in the recent CURE trial, in which all patients
received daily ASA, but the precise dose (75–325 mg) was
at the discretion of the investigator [19].
Based on data from the CAPRIE trial [14], the meta-
analysis found a statistically significant 10% (B4%) odds
reduction (10.1 vs. 11.1%) in favour of clopidogrel versus
ASA for the composite endpoint of stroke (all-cause), MI
or vascular death (p = 0.03). This is consistent with the
12% (B7%) odds reduction (21.1 vs. 23.2%) for the same
endpoint obtained from pooled data of four trials of ticlo-
pidine versus ASA in patients with different clinical man-
ifestations of atherothrombosis [13].
Trials of ASA in Stroke Patients
In 1999, Algra and van Gijn [20] published a meta-
analysis of 11 placebo-controlled trials of ASA treatment
in patients with a previous TIA or ischaemic stroke.
Pooled data from over 10,000 patients yielded a relative
risk reduction of 13% (95% CI 6–19%) in the combined
endpoint of vascular death, MI or stroke. As noted by the
authors, these data underscore the need for more effica-
cious antiplatelet regimens.
Cochrane Meta-Analysis of ADP-Receptor
Antagonist Data
A systematic review of all unconfounded, randomized
trials that directly compared an ADP-receptor antagonist
with ASA was recently reported by the Cochrane Stroke
Cerebrovasc Dis 2003;16(suppl 1):20–26 21
Group [21]. This analysis identified four trials that en-
rolled a total of 22,566 high-risk patients with a history of
vascular disease, including 9,840 patients in whom the
qualifying event was symptomatic cerebrovascular dis-
ease. Overall, the meta-analysis showed that ADP-recep-
tor antagonist therapy was superior to ASA, reducing the
odds of a serious vascular event, i.e. stroke (all-cause), MI
or vascular death by 9% (odds ratio (OR) = 0.91, 95% CI
0.84–0.98; 2p = 0.01), corresponding to the prevention of
11 (95% CI 2–19) events per 1,000 patients treated for
F2 years. Compared with ASA, ADP-receptor antagonist
treatment had a consistent, favourable effect on a range of
secondary effectiveness outcomes, with a significant re-
duction in all-cause stroke (OR = 0.88, 95% CI 0.79–
0.98), and a trend towards a significant reduction in isch-
aemic stroke/stroke of unknown aetiology (OR = 0.90,
95% CI 0.81–1.01), MI (OR = 0.88, 95% CI 0.76–1.01),
vascular death/death of unknown cause (OR = 0.93, 95%
CI 0.82–1.06), and all-cause mortality (OR = 0.95, 95%
CI 0.85–1.05).
Comparison of adverse event data for clopidogrel and
ticlopidine revealed the superior safety/tolerability profile
of clopidogrel compared with ticlopidine. Whereas clopi-
dogrel was associated with a slight increase in the inci-
dence of skin rash compared with ASA (OR = 1.32, 95%
CI 1.17–1.50), the corresponding odds ratio for ticlopid-
ine was 2.23 (OR = 2.23, 95% CI 1.74–2.86). The differ-
ence between clopidogrel and ticlopidine was highly sig-
nificant (2p = 0.0003). Moreover, there was no excess of
neutropenia (OR = 0.63, 95% CI 0.29–1.36) or thrombo-
cytopenia (OR = 1.00, 95% CI 0.57–1.74) for clopidogrel
compared with ASA [21].
CAPRIE
Evidence for the efficacy of clopidogrel was provided
by the CAPRIE trial, which compared clopidogrel with
ASA in 19,185 patients with symptomatic atherothrom-
bosis [14]. In CAPRIE, clopidogrel was associated with an
8.7% relative risk reduction (or 10% odds reduction) in
the combined risk of ischaemic stroke, MI or vascular
death, the primary study endpoint (p = 0.043). Further
analyses of the CAPRIE data set have demonstrated the
consistent benefit of clopidogrel over ASA for the individ-
ual components of the primary endpoint [14, 22, 23] and
for a range of composite endpoints related to vascular
ischaemia [24–26]. The largest benefit of clopidogrel com-
pared with ASA was seen for fatal or non-fatal MI, with a
relative risk reduction of 19.2% (p = 0.008) [22].
22 Cerebrovasc Dis 2003;16(suppl 1):20–26
Dual Antiplatelet Therapy
Existing Data
Mechanistic studies have revealed the involvement of
multiple agonists and signalling pathways in the activa-
tion of the glycoprotein (GP) IIb/IIIa receptor and subse-
quent platelet activation [27]. This realization led to the
hypothesis that by inhibiting the platelet P2Y12 ADP-
receptor antagonist on a background of thromboxane A2
inhibition, a potentiation of platelet inhibition would be
obtained. Indeed, a synergistic antithrombotic effect is
observed when clopidogrel is used with ASA, and this
effect has been demonstrated in a range of animal models
[28–30] and in post-MI patients [31].
The recent Antithrombotic Trialists’ Collaboration
meta-analysis also contains updated information from
randomized comparisons of dual antiplatelet regimens
compared with ASA alone. In the latest meta-analysis,
which incorporates data from the ESPS2 trial [15], the
combination of dipyridamole and ASA was associated
with a non-significant 6% (B6%) reduction in the odds of
vascular death, MI or stroke compared with ASA alone
(fig. 2) [13]. Analysis of each of the components of this
endpoint reveals that only non-fatal stroke was numeri-
cally lower in patients treated with dipyridamole and
ASA, and that this finding was driven by the results of a
single, large trial, i.e. ESPS2 [13]. Fifteen trials have com-
pared the combination of short-term intravenous GP IIb/
IIIa antagonist therapy and ASA versus ASA alone. The
pooled data generate a 19% (B4%) odds reduction for
this combination (p ! 0.0001). One trial has compared the
use of ticlopidine on top of ASA versus ASA alone, in cor-
onary stent patients [32]. In this study, the use of an ADP-
receptor antagonist on top of ASA was associated with an
incremental 20% reduction in vascular death, stroke or
MI (p = n.s.).
Subsequent to the preparation of the latest Antithrom-
botic Trialists’ Collaboration analysis, two randomized
trials have provided evidence to support the long-term use
of clopidogrel on top of ASA in patients with coronary
manifestations of atherothrombosis. CURE has demon-
strated the benefit of clopidogrel in patients with unstable
angina and non-Q-wave MI [33]. In CURE, clopidogrel
therapy was associated with a 20% relative risk reduction
for the primary endpoint of cardiovascular death, stroke
or MI (p ! 0.001) (fig. 1) [33]. Importantly, long-term clo-
pidogrel therapy in CURE (maximum = 12 months) was
associated with a sustained benefit over placebo, and the
relative benefit of clopidogrel was consistent for each of
the components of the primary endpoint, including a 14%
Easton
 % odds
Comparison Odds ratio p
reduction

Antithrombotic Trialists’ Collaboration

Dipyridamole 6 NS
Ticlopidine 20 NS
IV GP IIb/IIIa inhibitor 19 <0.0001
Subtotal 15 <0.0001

CURE
Clopidogrel 20* 0.00009

Fig. 1. Overview of the effects of different 0.0 0.5 1.0 1.5 2.0
dual antiplatelet regimens for the prevention Dual antiplatelet
ASA alone better
regimen better
of major atherothrombotic events [13, 33].
* Relative risk reduction.

15
Combined endpoint occurrence (%)

11.5%
27% RRR*
10 p = 0.02
8.5%

Placebo
Fig. 2. Kaplan-Meier curves for the compos- Clopidogrel
ite endpoint death, MI or stroke after long-
term treatment with clopidogrel or placebo 0
in the CREDO trial. * The relative risk re-
0 3 6 9 12
duction for all-cause stroke with long-term
Months from randomization
clopidogrel was 25.1% [reproduced with
kind permission, 35].

relative risk reduction for all-cause stroke. The benefit of
clopidogrel for the primary endpoint was observed across
a range of prespecified subgroups. These included higher-
risk patients such as those with diabetes, those with pre-
vious MI, and those with elevated cardiac enzymes or
markers at baseline. Of particular interest to the neurolo-
gist are the observations that in the 506 CURE patients
who had a prior history of stroke, the overall incidence of
the primary study endpoint was higher than in those with
no such history (n = 12,056 patients), while the benefit of
clopidogrel was consistent between the two patient popu-
lations. This finding parallels a recent analysis by Budaj et
al. [34], in which CURE patients were categorized into
low-, intermediate- and high-risk strata on the basis of
TIMI risk scores. The results of this study demonstrate a
consistent benefit of clopidogrel across the TIMI risk cate-
gories, with the highest absolute benefit in the stratum
with the highest risk.
Very recently, the CREDO (Clopidogrel for Reduction
of Events During Observation) trial has confirmed the sig-
nificant and sustained benefit of long-term clopidogrel
therapy on top of standard therapy including ASA in
patients undergoing elective PCI. In CREDO, 2,116 pa-
tients referred for PCI or who were deemed at high likeli-
hood of PCI were randomized to receive a preprocedural
clopidogrel loading dose or placebo. After PCI, all pa-

Evidence with ADP-Receptor Antagonists Cerebrovasc Dis 2003;16(suppl 1):20–26 23


Annual event rate (%)
Fig. 3. Elevated event rates and amplified
benefit of clopidogrel over ASA in CAPRIE
patients with a history of diabetes at study
entry. Data are shown for the composite of
vascular death, MI, stroke, or rehospitaliza-
tion for ischaemia or bleeding. Numbers
within the arrows indicate the number of
events prevented per 1,000 patient-years for
this endpoint [reproduced with kind permis-
sion, 40].
tients received clopidogrel 75 mg/day for 28 days. Be-
tween 29 days and 1 year, patients in the loading-dose
group received clopidogrel 75 mg/day, and those in the
control group received placebo. All patients received ASA
for the duration of the study. Long-term data showed that
continuation of clopidogrel for 12 months rather than 1
month after PCI was associated with a 26.9% relative risk
reduction for the composite of death, MI or stroke at
1 year (p = 0.023) (fig. 2) [35]. As was observed in the
CURE study, long-term clopidogrel therapy in CREDO
reduced the incidence of each component of the primary
endpoint, including a 25.1% reduction for all-cause
stroke. Moreover, administration of a clopidogrel loading
dose 66 h before PCI was associated with a 38.6% reduc-
tion in the composite of death, MI, or urgent target vessel
revascularization at 28 days (p = 0.05). These data rein-
force the benefit of early and sustained clopidogrel ad-
ministration that was observed in the PCI-CURE sub-
study of CURE patients who underwent PCI after ran-
domization [36].
Future Prospects for Clopidogrel in Neurology
Although CURE has demonstrated the benefit of clopi-
dogrel on top of standard therapy (including ASA) in
patients with unstable angina/non-Q-wave MI [33], this
regimen has not been proven in patients with recent man-
ifestations of cerebrovascular atherothrombosis. This is
being addressed in the ongoing MATCH (Management of
Atherothrombosis with Clopidogrel in High-Risk Patients
with Recent Transient Ischaemic Attack or Ischaemic
Stroke) study [37]. MATCH is comparing the efficacy and
24 Cerebrovasc Dis 2003;16(suppl 1):20–26
ASA
Clopidogrel
25 38
20 21
21.5%
9 17.7% 17.7%
15
15.6%
12.7%
10 11.8%
5
0
Non diabetic All diabetic patients Treated with insulin
safety of long-term clopidogrel on top of ASA versus clo-
pidogrel alone in an international, randomized, double-
blind trial in patients with recent TIA or ischaemic stroke
who have additional factors that are associated with an
increased risk of atherothrombotic recurrence, i.e. pre-
vious ischaemic stroke, previous MI, prior history of angi-
na [38], symptomatic peripheral disease [39], or history of
diabetes (fig. 3) [40]. Patients are randomized to receive
ASA 75 mg/day or matching placebo, and all patients
receive clopidogrel 75 mg/day as part of standard therapy.
Patients receive study medication and are followed for 18
months after randomization. The primary endpoint for
efficacy is a composite of ischaemic stroke, MI, vascular
death or rehospitalization for acute ischaemia during the
treatment period. Patient enrolment into MATCH was
completed in April 2002, with a total of 7,601 patients
recruited in 28 countries. Initial data from this important
trial are anticipated in 2004.
Additional trials of clopidogrel in neurology are
planned. ATARI will use a factorial design to compare the
use of clopidogrel, low-molecular-weight heparin, both, or
neither on a background of ASA therapy in patients who
have suffered a TIA in the previous 12 h. The primary
endpoint in ATARI will be a composite of vascular death,
stroke or MI. The Secondary Prevention of Small Subcor-
tical Strokes (SPS3) trial will evaluate the efficacy of clopi-
dogrel in patients with small subcortical strokes (lacunar
infarcts). Although this stroke subtype accounts for
around 25% of all cerebral infarcts, the effects of anti-
platelet therapy and blood pressure lowering have not spe-
cifically been studied in this clinical setting. SPS3 will
Easton
therefore compare the use of clopidogrel on top of ASA
versus ASA alone and usual blood pressure treatment ver-
sus intensive blood pressure reduction in the prevention
of subsequent stroke and vascular events in patients with
subcortical strokes.
A further major study is CHARISMA (Clopidogrel for
High Atherothrombotic Risk and Ischaemic Stabiliza-
tion, Management and Avoidance), a double-blind, ran-
domized, placebo-controlled trial that will evaluate the
long-term efficacy and safety of clopidogrel in a broad
spectrum of patients at high risk of atherothrombosis.
CHARISMA will compare clopidogrel versus placebo on
top of standard therapy (including low-dose ASA) in over
15,000 high-risk patients (qualifying with a combination
of atherothrombotic risk factors and/or documented cere-
brovascular disease and/or documented coronary disease
and/or symptomatic PAD) [40].
Conclusions
The ADP-receptor antagonists clopidogrel and ticlo-
pidine are associated with superior protection against
major atherothrombotic events (stroke, MI or vascular
death) compared with ASA. However, compared with
ticlopidine, clopidogrel offers a more favourable safety/
tolerability profile. The landmark CURE trial represents
‘proof of principle’ for the use of clopidogrel on top of
ASA to provide incremental benefit in high-risk athero-
thrombotic patients. Based on a consideration of the
modes of action of clopidogrel and ASA, and the patho-
physiology of atherothrombosis, there is a sound basis to
hypothesize that the dual antiplatelet approach will be of
value in patients with cerebrovascular manifestations of
the disease. This is the subject of the ongoing MATCH
trial, and the planned SPS3 and ATARI studies. These
important new trials form part of an extensive pro-
gramme whose aim is to enable a more extensive assess-
ment of the efficacy and safety of clopidogrel on top of
ASA in a range of clinical settings.

References

1 Ruggeri ZM: Platelets in atherothrombosis.
Nat Med 2002;8:1227–1234.
2 Drouet L: Atherothrombosis as a systemic dis-
ease. Cerebrovasc Dis 2002;13(suppl 1):1–6.
3 Broderick JP, Phillips SJ, O’Fallon WM, Frye
RL, Whisnant JP: Relationship of cardiac dis-
ease to stroke occurrence, recurrence and mor-
tality. Stroke 1992;23:1250–1256.
4 Dennis MS, Burn JP, Sandercock PA, Bamford
JM, Wade DT, Warlow CP: Long-term surviv-
al after first-ever stroke: The Oxfordshire Com-
munity Stroke Project. Stroke 1993;24:796–
800.
5 Hartmann A, Rundek T, Mast H, Paik MC,
Boden-Albala B, Mohr JP, Sacco RL: Mortality
and causes of death after first ischemic stroke:
The Northern Manhattan Stroke Study. Neu-
rology 2001;57:2000–2005.
6 Sacco RL: Risk factors and outcomes for isch-
emic stroke. Neurology 1995;45:S10–S14.
7 Hankey GJ: Long-term outcome after isch-
aemic stroke/transient ischaemic attack. Cere-
brovasc Dis 2003;16(suppl 1):14–19.
8 Albers GW, Amarenco P, Easton JD, Sacco
RL, Teal P: Antithrombotic and thrombolytic
therapy for ischemic stroke. Chest 2001;119:
300S–320S.
9 European Stroke Initiative Executive Commit-
tee: European Stroke Initiative recommenda-
tions for stroke management. Cerebrovasc Dis
2000;10:335–351.
10 Vane JR: Inhibition of prostaglandin synthesis
as a mechanism of action for aspirin-like drugs.
Nat New Biol 1971;231:232–235.
11 Savi P, Labouret C, Delesque N, Guette F,
Lupker J, Herbert JM: P2Y12, a new platelet
ADP receptor, target of clopidogrel. Biochem
Biophys Res Commun 2001;283:379–383.
12 Vane JR, Meade TW: Second European Stroke
Prevention Study (ESPS 2): Clinical and phar-
macological implications. J Neurol Sci 1997;
145:123–125.
13 Antithrombotic Trialists’ Collaboration: Col-
laborative meta-analysis of randomised trials
of antiplatelet therapy for prevention of death,
myocardial infarction and stroke in high-risk
patients. BMJ 2002;324:71–86.
14 CAPRIE Steering Committee: A randomised,
blinded, trial of clopidogrel versus aspirin in
patients at risk of ischaemic events (CAPRIE).
Lancet 1996;348:1329–1339.
15 Diener H, Cunha L, Forbes C, Sivenius J,
Smets P, Lowenthal A: European Stroke Pre-
vention Study. 2. Dipyridamole and acetylsali-
cylic acid in the secondary prevention of
stroke. J Neurol Sci 1996;143:1–13.
16 CAST (Chinese Acute Stroke Trial) Collabora-
tive Group: CAST: Randomised placebo-con-
trolled trial of early aspirin use in 20,000 pa-
tients with acute ischaemic stroke. Lancet
1997;349:1641–1649.
17 International Stroke Trial Collaborative
Group: The International Stroke Trial (IST): A
randomized trial of aspirin, subcutaneous hep-
arin, both or neither among 19,435 patients
with acute ischaemic stroke. Lancet 1997;349:
1569–1581.
18 Antiplatelet Trialists’ Collaboration: Collabo-
rative overview of randomised trials of anti-
platelet therapy. I. Prevention of death, myo-
cardial infarction and stroke by prolonged anti-
platelet therapy in various categories of pa-
tients. BMJ 1994;308:81–106.
19 Peters RJG, Zao F, Lewis BS, Fox KAA, Yusuf
S on behalf of the CURE Investigators: Aspirin
dose and bleeding events in the CURE study.
Eur Heart J 2002;23(abstr suppl):510.
20 Algra A, van Gijn J: Cumulative meta-analysis
of aspirin efficacy after cerebral ischaemia of
arterial origin. J Neurol Neurosurg Psychiatry
1999;66:255.
21 Hankey GJ, Sudlow CLM, Dunbabin DW:
Thienopyridines or aspirin to prevent stroke
and other serious vascular events in patients at
high risk of vascular disease? A systematic
review of the evidence from randomized trials.
Stroke 2000;31:1779–1784.
22 Cannon C: Effectiveness of clopidogrel versus
aspirin in preventing acute myocardial infarc-
tion in patients with symptomatic athero-
thrombosis (CAPRIE trial). Am J Cardiol
2002;90:760–762.
23 Easton JD: Benefit of clopidogrel in patients
with symptomatic cerebrovascular disease.
Neurology 1998;50(suppl 4):A157.
24 Easton JD: Net benefit of clopidogrel over aspi-
rin for the prevention of atherothrombotic
events. Cerebrovasc Dis 1998;8(suppl 4):46.

Evidence with ADP-Receptor Antagonists Cerebrovasc Dis 2003;16(suppl 1):20–26 25

25 Bhatt DL, Hirsch AT, Ringleb PA, Hacke W,
Topol EJ on behalf of the CAPRIE Investiga-
tors: Reduction in the need for hospitalization
for recurrent ischemic events and bleeding with
clopidogrel instead of aspirin. Am Heart J
2000;140:67–73.
26 Rupprecht HJ: Consistency of the benefit of
clopidogrel across a range of vascular-related
endpoints: Results from CAPRIE. Eur Heart J
1998;19(suppl):52.
27 Schafer AI: Antiplatelet therapy. Am J Med
1996;101:199–209.
28 Harker LA, Marzec UM, Kelly AB, Chronos
NRF, Sundell IB, Hanson SR, Herbert JM:
Clopidogrel inhibition of stent, graft, and vas-
cular thrombogenesis with antithrombotic en-
hancement by aspirin in nonhuman primates.
Circulation 1998;22:2461–2469.
29 Herbert JM, Dol F, Bernat A, Falotico R, Lalé
A, Savi P: The antiaggregating and antithrom-
botic activity of clopidogrel is potentiated by
aspirin in several experimental models in the
rabbit. Thromb Haemost 1998;80:512–518.
30 Makkar RR, Eigler S, Kaul S, Frimerman A,
Nakamura M, Shah PK, Forrester JS, Herbert
JM, Litvack F: Effects of clopidogrel, aspirin
and combined therapy in a porcine ex vivo
model of high-shear induced stent thrombosis.
Eur Heart J 1998;19:1538–1546.
26 Cerebrovasc Dis 2003;16(suppl 1):20–26
31 Moshfegh K, Redondo M, Julmy F, Wuillemin
WA, Gebauer MU, Haeberli A, Meyer BJ: An-
tiplatelet effects of clopidogrel compared with
aspirin after myocardial infarction: Enhanced
inhibitory effects of combination therapy. J
Am Coll Cardiol 2000;36:699–705.
32 Leon MB, Baim DS, Popma JJ, Gordon PC,
Cutlip DE, Ho KK, Giambartolemei A, Diver
DJ, Lasorda DM, Williams DO, Pocock SJ,
Kuntz RE: A clinical trial comparing three
antithrombotic-drug regimens after coronary-
artery stenting. N Engl J Med 1998;339:1665–
1671.
33 The Clopidogrel in Unstable Angina to Prevent
Recurrent Events Trial Investigators: Effects of
clopidogrel in addition to aspirin in patients
with acute coronary syndromes without ST-
segment elevation. N Engl J Med 2001;345:
494–502.
34 Budaj A, Yusuf S, Mehta S, Fox KAA, Tognoni
G, Zhao F, Chrolavicius S, Hunt D, Keltai M,
Grazia Franzosi M for the Clopidogrel in Un-
stable angina to prevent Recurrent Events
(CURE) Trial Investigators: Benefit of clopido-
grel in patients with acute coronary syndromes
without ST-segment elevation in various risk
groups. Circulation 2002;106:1622–1626.
35 Steinhubl SR, Berger PB, Mann JT III, Fry
ETA, DeLago A, Wilmer C, Topol EJ, for the
CREDO Investigators: Early and sustained
dual oral antiplatelet therapy following percu-
taneous coronary intervention. A randomized
controlled trial. JAMA 2002;288:2411–2420.
36 Mehta SR, Yusuf S, Peters RJG, Bertrand ME,
Lewis BS, Natarajan MK, Malmberg K, Rup-
precht HJ, Zhao F, Chrolavicius S, Copland I,
Fox KAA, for the Clopidogrel in Unstable An-
gina to Prevent Recurrent Events Trial
(CURE) Investigators: Effects of pretreatment
with clopidogrel and aspirin followed by long-
term therapy in patients undergoing percuta-
neous coronary intervention: The PCI-CURE
study. Lancet 2001;358:527–533.
37 Anonymous: Major ongoing stroke trials.
Stroke 2002;33:1728–1737.
38 Hacke W, Hirsch AT, Topol EJ: The benefit of
clopidogrel over aspirin is amplified in high-
risk subgroups with a prior history of ischaemic
events. Eur Heart J 1999;20(abstr suppl):666.
39 Hirsch AT, Hacke W, Topol EJ, Criqui MH,
Hiatt WR, Creager MA: The morbidity of pe-
ripheral arterial disease: A powerful and consis-
tent predictor of clinical cardiovascular isch-
aemic risk. Eur Heart J 1999;20(abstr
suppl):415.
40 Bhatt DL, Marso SP, Hirsch AT, Ringleb PA,
Hacke W, Topol EJ: Amplified benefit of clopi-
dogrel versus aspirin in patients with diabetes
mellitus. Am J Cardiol 2002;90:625–628.
Easton