Professional Documents
Culture Documents
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
Fax + 41 61 306 12 34
E-Mail karger@karger.ch
www.karger.com
Vol. 16, Supplement 1, 2003
Contents
1 Introduction
Bousser, M.-G. (Paris); Bogousslavsky, J. (Lausanne)
Introduction
Stroke is the principal cause of disability, dependency botic therapy reduces the risk of serious vascular isch-
and loss of social competence in the western world. The aemic events in AF patients. Oral anticoagulation reduces
majority of strokes are of ischaemic origin, and most isch- the risk of stroke recurrence by approximately 60% com-
aemic strokes are due to atherothrombosis. Clearly, a pared with placebo. Acetylsalicylic acid (ASA) is less
thorough knowledge of prognosis and effective secondary effective than anticoagulation, reducing the risk of stroke
prevention strategies are of paramount importance for cli- by around 20%. Thus, the use of ASA as first-line anti-
nicians who care for stroke patients. thrombotic therapy in AF patients is limited to those at
At the 11th European Stroke Conference, held in low risk. An important topic for future investigation is
Geneva, Switzerland, between May 29 and June 1, 2002, whether dual antiplatelet therapy, for example clopidogrel
delegates had the opportunity to attend two important in addition to ASA, is a suitable alternative to warfarin for
symposia – an Educational Symposium entitled Long- the reduction of thromboembolism in high- or interme-
term outcome after different stroke subtypes and a Satellite diate-risk AF patients. This will be the topic of the
Symposium entitled Future trends in long-term prevention planned ACTIVE trial.
of ischaemic stroke – the role of atherothrombosis manage- Dr Martin Dennis then describes our current knowl-
ment. This supplement to Cerebrovascular Diseases is edge on outcome after primary intracerebral haemorrhage
based on key presentations made during these sessions. (PICH). Dr Dennis highlights the methodological difficul-
In the first article of this supplement, Dr Heinrich ties that hamper the study of PICH prognosis, including
Mattle provides an overview of long-term outcome after the small proportion of all strokes that are due to PICH,
stroke due to atrial fibrillation (AF). AF, which is the most the need for brain imaging (or autopsy) to arrive at a reli-
common cardiac arrhythmia and whose prevalence in- able diagnosis of PICH, differences in definition of PICH
creases with age, places the patient at risk for left atrial according to imaging method and its timing, and the het-
thrombus formation, distal embolism and associated isch- erogeneity of haemorrhage aetiologies and types. Al-
aemic stroke. AF is associated with substantial cardiovas- though individual estimates of 1-month case fatality have
cular mortality and morbidity. Indeed, long-term survival wide confidence intervals, pooled data from unselected
data from the Framingham Heart Study indicates that AF PICH cohorts provide a more precise estimate of about
is independently associated with an approximate dou- 42%. After PICH, greater age and stroke severity are both
bling in mortality in both men and women. Moreover, in associated with increased case fatality and poorer func-
the European Atrial Fibrillation Trial, the annualized rate tional outcomes. Currently, there is no definite evidence
of vascular death, nonfatal stroke, nonfatal myocardial to indicate that the risk of recurrent stroke after PICH
infarction or systemic embolism in patients who received differs from that after ischaemic stroke of equivalent clin-
placebo was 12%. There is clear evidence that antithrom- ical severity. Generation of more precise data on progno-
0
<55 55–59 60–64 65–69 70–74 75–79 80–84 85
Age (years) 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].
Fig. 1. Prevalence of AF by age and gender [reproduced with kind
permission, 2].
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].
Highest Previous stroke or transient ischaemic attack 12% 14 (warfarin vs. placebo)
High Age 1 75 years F6% 28 (warfarin vs. placebo)
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 F3.5% 48 (warfarin vs. placebo)
Diabetes
Coronary heart disease
No high-risk criteria
Low Age ! 65 years F1% F500 (ASA vs. placebo)
No high-risk factors F250 (warfarin vs. ASA)
Prevention of Stroke in AF Patients placebo to 8% per year and the annual rate of all-cause
stroke from 12 to 4%.
Antithrombotic therapy has clearly been shown to A major limitation of oral anticoagulation is its narrow
reduce the risk of serious vascular ischaemic events in therapeutic window. In patients with nonrheumatic AF,
patients with AF [12]. A meta-analysis of six randomized the risk of stroke has been found to rise steeply at INR
trials that compared dose-adjusted warfarin versus place- values !2.0. Thus, at an INR of 1.7, the adjusted odds
bo in a total of 2,900 patients found a pooled relative risk ratio for stroke, as compared with the risk at an INR of
reduction of 62% (95% CI 48–72%) for the endpoint of 2.0, is 2.0 (95% CI 1.6–2.4) [14]. At INR values 14.0, the
all-cause stroke [13]. In these trials, the absolute risk risk of intracranial haemorrhage rises exponentially [15].
reduction for primary prevention was 2.7% per year (data Therefore, in nonrheumatic AF, an INR of 2.0–3.0 gives
from 5 trials) and for secondary prevention it was 8.4% the optimal risk:benefit ratio for the prevention of stroke
per year (data from a single trial (EAFT)). Major extra- [12]. In patients with valvular disease, a higher INR (3.0–
cranial haemorrhage was increased by warfarin therapy 3.5) should be targeted.
(absolute risk increase 0.3% per year, relative risk = 2.4, AF patients can be categorized according to stroke risk,
95% CI 1.2–4.6). Six trials have compared the use of ASA which is influenced by a number of variables and which
versus placebo for stroke prevention in AF patients [13]. should be taken into account when initiating antithrom-
The pooled estimate from these studies gave an overall botic therapy (table 1). AF patients who have suffered a
22% (95% CI 2–38%) reduction in the risk of all-cause previous stroke or transient ischaemic attack have the
stroke. Pooled analysis of data from five trials that com- highest risk of stroke, which is estimated at 12% per year.
pared adjusted-dose warfarin plus ASA versus ASA alone Patients in the lowest risk category (those aged !65 years
showed that combination therapy was superior to ASA with no high-risk factors) have an estimated annual stroke
monotherapy, being associated with a 36% (95% CI 14– risk of 1%, making them potential candidates for ASA
52%) relative risk reduction for all-cause stroke. Major therapy. For warfarin, the number needed to treat (NNT)
extracranial haemorrhage was higher in those who re- to prevent one stroke per year ranges from 14 in the high-
ceived warfarin relative to those who received ASA (rela- est risk category to 48 in moderate-risk patients. In the
tive risk = 2.0, 95% CI 1.2–3.4) [13]. lowest risk category, the NNT for ASA versus placebo is
In EAFT, anticoagulation therapy was more effective about 500 and for warfarin compared with ASA the NNT
than ASA in preventing serious vascular events, reducing is about 250.
the rate of the primary endpoint from 17% per year with
1 Feinberg WM, Blackshear JL, Laupacis A, 9 Hart RG, Pearce LA, McBride R, Rothbart 17 The Atrial Fibrillation Follow-Up Investiga-
Kronmal R, Hart RG: Prevalence, age distribu- RM, Asinger RW on behalf of the Stroke Pre- tion of Rhythm Management (AFFIRM) In-
tion and gender of patients with atrial fibrilla- vention in Atrial Fibrillation (SPAF) Investiga- vestigators: A comparison of rate control and
tion: Analysis and implications. Arch Intern tors: Factors associated with ischemic stroke rhythm control in patients with atrial fibrilla-
Med 1995;155:469–473. during aspirin therapy in atrial fibrillation: tion. N Engl J Med 2002;347:1825–1833.
2 Go AS, Hylek EM, Phillips KA, Chang Y, Hen- Analysis of 2,012 participants in the SPAF I– 18 Van Gelder IC, Hagens VE, Bosker HA, King-
ault LE, Selby JV, Singer DE: Prevalence of III clinical trials: Stroke Prevention in Atrial ma JH, Kamp O, Kingma T, Said SA, Darma-
diagnosed atrial fibrillation in adults: National Fibrillation (SPAF) Investigators. Stroke 1999; nata JI, Timmermans AJM, Tijssen JGP,
implications for rhythm management and 30:1223–1229. Crijns HJGM, for the Rate Control versus
stroke prevention: The AnTicoagulation and 10 Stroke Prevention in Atrial Fibrillation Investi- Electrical Cardioversion for Persistent Atrial
Risk Factors in Atrial Fibrillation (ATRIA) gators Committee on Echocardiography: Fibrillation Study Group: A comparison of rate
Study. JAMA 2001;285:2370–2375. Transesophageal echocardiographic correlates control and rhythm control in patients with
3 International Stroke Trial Collaborative of thromboembolism in high-risk patients with recurrent persistent atrial fibrillation. N Engl J
Group: The International Stroke Trial (IST): A nonvalvular atrial fibrillation. Ann Intern Med Med 2002;347:1834–1840.
randomized trial of aspirin, subcutaneous hep- 1998;128:639–647. 19 The Clopidogrel in Unstable Angina to Prevent
arin, both or neither among 19,435 patients 11 EAFT (European Atrial Fibrillation Trial) Recurrent Events Trial Investigators: Effects of
with acute ischaemic stroke. Lancet 1997;349: Study Group: Secondary prevention in non- clopidogrel in addition to aspirin in patients
1569–1581. rheumatic atrial fibrillation after transient isch- with acute coronary syndromes without ST-
4 Saxena R, Lewis S, Berge E, Sandercock PAG, aemic attack or minor stroke. Lancet 1993;342: segment elevation. N Engl J Med 2001;345:
Koudstaal PJ, for the International Stroke Trial 1255–1262. 494–502.
Collaborative Group: Risk of early death and 12 Albers GW, Dalen JE, Laupacis A, Manning 20 Steinhubl SR, Berger PB, Mann JT III, Fry
recurrent stroke and effect of heparin in 3,169 WJ, Petersen P, Singer DE: Antithrombotic ETA, DeLago A, Wilmer C, Topol EJ, for the
patients with acute ischemic stroke and atrial therapy in atrial fibrillation. Chest 2001;119: CREDO Investigators: Early and sustained
fibrillation in the International Stroke Trial. 194S–206S. dual oral antiplatelet therapy following percu-
Stroke 2001;32:2333–2337. 13 Hart RG, Benavente O, McBride R, Pearce taneous coronary intervention. A randomized
5 Wolf PA, Abbott RD, Kannel WB: Atrial fibril- LA: Antithrombotic therapy to prevent stroke controlled trial. JAMA 2002;288:2411–2420.
lation as an independent risk factor for stroke: in patients with atrial fibrillation: A meta-anal- 21 Leon MB, Baim DS, Popma JJ, Gordon PC,
The Framingham Study. Stroke 1991;22:983– ysis. Ann Intern Med 1999;131:492–501. Cutlip DE, Ho KKL, Giambartolomei A, Div-
988. 14 Hylek EM, Skates SJ, Sheehan MA, Singer DE: er DJ, Lasorda DM, Williams DO, Pocock SJ,
6 Benjamin EJ, Wolf PA, D’Agostino RB, Silber- An analysis of the lowest effective intensity of Kuntz RE: A clinical trial comparing three
shatz H, Kannel WB, Levy D: Impact of atrial prophylactic anticoagulation for patients with antithrombotic-drug regimens after coronary-
fibrillation on the risk of death. The Framing- nonrheumatic atrial fibrillation. N Engl J Med artery stenting. Stent Anticoagulation Resteno-
ham Heart Study. Circulation 1998;98:946– 1996;335:540–546. sis Study Investigators. N Engl J Med 1998;
952. 15 Hylek EM, Singer DE: Risk factors for intracra- 339:1665–1671.
7 Atrial Fibrillation Investigators: Risk factors nial hemorrhage in outpatients taking warfarin. 22 Sievert H, Lesh MD, Trepels T, Omran H, Bar-
for stroke and efficacy of antithrombotic thera- Ann Intern Med 1993;118:511–520. torelli A, Bella PD, Nakai T, Reisman M,
py in atrial fibrillation: Analysis of pooled data 16 Hohnloser SH, Kuck KH, Lilienthal J, for the DiMario C, Block P, Kramer P, Fleschenberg
from five randomized controlled trials. Arch PIAF Investigators: Rhythm or rate control in D, Krumsdorf U, Scherer D: Percutaneous left
Intern Med 1994;154:1449–1457. atrial fibrillation – Pharmacological Interven- atrial appendage transcatheter occlusion to pre-
8 Atrial Fibrillation Investigators: Echocardio- tion in Atrial Fibrillation (PIAF): a randomised vent stroke in high-risk patients with atrial
graphic predictors of stroke in patients with trial. Lancet 2000;356:1789–1794. fibrillation. Circulation 2002;105:1887–1889.
atrial fibrillation: A prospective study of 1,066
patients from three clinical trials. Arch Intern
Med 1998;158:1316–1320.
Study (first author) Location Patient Inclusion criteria Strokes PICH, n (%)
population
relatives to give them an indication of what may occur in ning of the study. This requires either CT scanning within
the future, and to be able to balance risks and benefits of 7 days of stroke onset or a gradient echo MRI (or equiva-
interventions in an informed manner. Second, informa- lent) if later in all cases and an autopsy in fatal cases where
tion on the prognosis of groups of patients is useful for the such scanning was not possible [4]. Patients should be
design of randomized clinical trials, to estimate sample identified as soon after the initial stroke as possible, and
size and decide on the optimal inclusion/exclusion crite- patients should be followed prospectively, with frequent
ria. Finally, data on prognosis of patients can be used in follow-up until death. Stroke recurrences should be char-
the assessment of quality of care, although this requires acterized clinically, and should be investigated with early
adequate adjustment of outcomes for case mix [1, 2]. CT or MRI scans to determine if these are haemorrhagic
Several methodological problems affect studies of the or ischaemic events. Regular measurements should also
prognosis of PICH and their results. First, PICH is rela- be made regarding impairment, disability and handicap,
tively rare, accounting for only 10–20% of all stroke cases. and quality of life.
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, Outcome after PICH
PICH is only reliably identified by brain imaging or
autopsy, and the definition of PICH is dependent on the Overview of Current Data
imaging method and its timing. Third, PICH patients Most of the community-based studies to date have
comprise a heterogeneous group, representing a range of included only small populations – typically 50,000–
haemorrhage aetiologies and types. Finally, outcome after 200,000 – resulting in the analysis of relatively small
PICH is influenced in the different studies by the treat- numbers of first-ever strokes and even smaller numbers of
ment given to patients, and this varies greatly according to PICH. Most have been in White populations, so relatively
geographical location. little is known about the prognosis of PICH in other eth-
The ‘ideal’ study [3] of long-term outcome after PICH nic groups. Hospital-based and less rigorous community-
should provide unbiased and precise estimates of progno- based ones will not identify all stroke patients occurring in
sis. The study should identify all first-ever strokes in a a defined population. In these circumstances, certain
large, well-defined population, including all patients that types of patients will be underrepresented: these include
are not referred to hospital. The study should then identi- patients with minor strokes, who are less likely to be
fy all cases of PICH in the group of first-ever strokes, to referred, patients with major strokes, who may die with-
produce an unselected, representative group at the begin- out diagnosis, and others who are not admitted to hospi-
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
References
1 Davenport RJ, Dennis MS, Warlow CP: Effect ject (ESPro). Incidence and case fatality at 1, 3 prospective community-based L’Aquila regis-
of correcting outcome data for case mix: An and 12 months. Stroke 1998;29:2501–2506. try (1994–1998). First year’s results. Stroke
example from stroke medicine. BMJ 1996;312: 10 Ellekjær H, Holmen J, Indredavik B, Terent A: 1997;28:2500–2506.
1503–1505. Epidemiology of stroke in Innherred, Norway, 18 Counsell C, Dennis M: Systematic review of
2 Weir N, Dennis M: Scottish Stroke Outcomes 1994–1996: Incidence and 30-day case-fatality prognostic models in patients with acute
Group: Towards a national system for monitor- rate. Stroke 1997;28:2180–2184. stroke. Cerebrovasc Dis 2001;12:159–170.
ing the quality of hospital-based stroke ser- 11 Lauria G, Gentile M, Fassetta G, Casetta I, 19 Dennis MS, Burn JPS, Sandercock PAG, Bam-
vices. Stroke 2001;32:1415–1421. Agnoli F, Andreotta G, Barp C, Caneve G, ford JM, Wade DT, Warlow CP: Long-term
3 Sudlow CLM, Warlow CP: Comparing stroke Cavallero A, Cielo R, Mongillo D, Mosca M, survival after first-ever stroke: The Oxfordshire
incidence worldwide: What makes studies Olivieri PG: Incidence and prognosis of stroke Community Stroke Project. Stroke 1993;24:
comparable? Stroke 1996;27:550–558. in the Belluno Province, Italy. First-year results 796–800.
4 Wardlaw JM, Keir SL, Dennis MS: The impact of a community-based study. Stroke 1995;26: 20 Warlow CP, Dennis MS, Van Gijn J, Hankey
of delays in computed tomography of the brain 1787–1793. G, Bamford J, Wardlaw J: A practical approach
on the accuracy of diagnosis and subsequent 12 D’Alessandro G, Di Giovanni M, Roveyaz L, to management of stroke patients; in Stroke: A
management in patients with minor stroke. J Iannizzi L, Pesenti Compagnoni M, Blanc S, Practical Guide to Management, ed 2. Oxford,
Neurol Neurosurg Psychiatry 2003;74:77–81. Bottacchi E: Incidence and prognosis of stroke Blackwell Scientific, 2000, pp 414–441.
5 Keir SL, Wardlaw JM, Warlow CP: Stroke epi- in the Valle d’Aosta, Italy: First-year results of 21 Jorgensen HS, Nakayama H, Raaschou HO,
demiology studies have underestimated the fre- a community-based study. Stroke 1992;23: Olsen TS: Intracerebral hemorrhage versus in-
quency of intracerebral haemorrhage. A sys- 1712–1715. farction: Stroke severity, risk factors and prog-
tematic review of imaging in epidemiological 13 Giroud M, Gras P, Chadan N, Beuriat P, Milan nosis. Ann Neurol 1995;38:45–50.
studies. J Neurol 2002;249:1226–1231. C, Arveux P, Duams R: Cerebral haemorrhage 22 Bailey RD, Hart RG, Benavente O, Pearce LA:
6 Anderson CS, Jamrozik KD, Burvill PW, Cha- in a French prospective population study. J Recurrent brain hemorrhage is more frequent
kera TM, Johnson GA, Stewart-Wynne EG: Neurol Neurosurg Psychiatry 1991;54:595– than ischemic stroke after intracranial hemor-
Determining the incidence of different sub- 598. rhage. Neurology 2001;56:773–777.
types of stroke: Results from the Perth Com- 14 Ricci S, Celani MG, La Rosa F, Vitali R, Duca 23 Burn J, Dennis M, Bamford J, Sandercock P,
munity Stroke Study, 1989–1990. Med J Aust E, Ferraguzzi R, Paolotti M, Seppoloni D, Ca- Wade D, Warlow C: Long-term risk of recur-
1993;158:85–89. puto N, Chiurulla C, Scaroni R, Signorini E: rent stroke after a first-ever stroke. Stroke
7 Counsell C, Boonyakarnkul S, Dennis M, San- SEPIVAC: A community-based study of stroke 1994;25:333–337.
dercock P, Bamford J, Burn J, Warlow C: Pri- incidence in Umbria, Italy. J Neurol Neurosurg 24 Al-Shahi R, Warlow C: A systematic review of
mary intracerebral haemorrhage in the Oxford- Psychiatry 1991;54:695–698. the frequency and prognosis of arteriovenous
shire Community Stroke Project. 2. Prognosis. 15 Ashok PP, Radhakrishnan K, Sridharan R, El- malformations of the brain in adults. Brain
Cerebrovasc Dis 1995;5:26–34. Mangoush MA: Incidence and pattern of cere- 2001;124:1900–1926.
8 Vemmos KN, Bots ML, Tsibouris PK, Zis VP, brovascular disease in Benghazi, Libya. J Neu- 25 PROGRESS Collaborative Group: Random-
Grobbee DE, Stranjalis GS, Stamelopoulos S: rol Neurosurg Psychiatry 1986;49:519–523. ised trial of a perindopril-based blood-pres-
Stroke incidence and case fatality in southern 16 Thrift AG, Dewey HM, Macdonell RAL, sure-lowering regimen among 6,105 individu-
Greece. The Arcadia Stroke Registry. Stroke McNeil JJ, Donnan GA: Stroke incidence on als with previous stroke or transient ischaemic
1999;30:363–370. the East Coast of Australia. The North East attack. Lancet 2001;358:1033–1041.
9 Kolominsky-Rabas PL, Sarti C, Heuschmann Melbourne Stroke Incidence Study (NEME- 26 Burn J, Dennis M, Bamford J, Sandercock P,
PU, Graf C, Siemonsen S, Neundoerfor B, Ka- SIS). Stroke 2000;31:2087–2092. Wade D, Warlow C: Epileptic seizures after a
talinc A, Lang E, Gassmann KG, von Stockert 17 Carolei A, Marini C, Di Naploi M, Di Gianfil- first stroke: The Oxfordshire Community
TR: A prospective community-based study of ippo G, Santalucia P, Baldassarre M, De Mat- Stroke Project. BMJ 1997;315:1582–1587.
stroke in Germany – The Erlangen Stroke Pro- teis G, di Orio F: High stroke incidence in the
Table 3. Overview of community-based studies of stroke, recording stroke and cardiac events over 5 years [10–15]
a Data are for causes of death in the first 30 days after index event.
Predictive Factors for Stroke Prognosis year follow-up data, age at presentation continued to
have predictive power for long-term mortality, whereas
Predictors of Mortality stroke severity did not. However, atrial fibrillation, con-
Current findings on short- and long-term predictors of gestive heart failure, and early- and late-onset ischaemic
death, disability and recurrent stroke are summarized in heart disease at baseline were significant predictors of 10-
table 4. Probably the most reliable data on mortality year mortality.
comes from Olmsted County, Minnesota, USA, where Five-year follow-up data are also available from the
30-day and 5-year follow-up data between 1985 and 1989 Northern Manhattan Stroke Study [17], which analyzed
have been reported for a cohort of 454 patients with first- predictors of outcome in 323 ischaemic stroke patients
ever ischaemic stroke [12], and 10-year follow-up data between 1983 and 1988. In this cohort, the predictors of
are available for a cohort of 1,111 first-ever ischaemic death at 30 days were again stroke severity (as assessed by
stroke patients followed between 1975 and 1989 [16]. depressed consciousness, major hemispheric or basilar
For death at 30 days after ischaemic stroke, the two syndrome, and hyperglycaemia at hospital admission), as
major predictors of early mortality were age (relative risk well as congestive heart failure. At 5 years of follow-up,
(RR)/10 years = 1.3; 95% CI 1.03–1.7) and severe stroke stroke severity and congestive heart failure continued to
as measured by the Rankin scale (RR = 11.6; 95% CI be predictors of mortality, and increasing age also had sta-
2.8–49). For mortality at 5 years, age and severe stroke tistically significant predictive power. Patients who pre-
continued to predict outcome. In addition, congestive sented with lacunar ischaemic stroke syndromes had sig-
heart failure (RR = 1.7; 95% CI 1.2–2.3) and early-onset nificantly lower mortality at 5 years after the index
ischaemic heart disease (RR = 5.0; 95% CI 1.8–13) also stroke.
had statistically significant predictive power. Ischaemic In the Perth Community Stroke Study, data were avail-
stroke caused by large-artery atherosclerosis with 150% able for 1- and 5-year mortality [13, 18]. Predictors of
stenosis was associated with a lower risk of 5-year mortal- death at 1 year were increasing stroke severity and a histo-
ity (RR = 0.5; 95% CI 0.3–0.8). From the study with 10- ry of previous cardiac disease (atrial fibrillation or conges-
Outcome Predictors
short-term long-term
LACS = Lacunar syndrome; TACS = total anterior circulation syndrome; POCS = posterior circulation syn-
drome.
tive heart failure). At 5 years, increasing stroke severity at 3.3–63); other independent predictors were increasing age
baseline was again predictive of mortality, together with and stroke severity [26].
evidence at baseline of multi-bed vascular disease was
also predictive (previous TIA: RR = 1.9 (95% CI 1.3–2.9); Predictors of Stroke Recurrence
peripheral arterial disease: RR = 1.7 (95% CI 1.2–2.5); Around 12% of patients have a recurrent stroke within
cardiac failure: RR = 1.6 (95% CI 1.1–2.3)). These data the first year and 30% experience recurrence over 5 years
indicate that long-term prevention of coronary events is [11, 27]. In Rochester, Minnesota, the major baseline pre-
essential for long-term reduction of mortality in stroke dictor of recurrent stroke at 30 days was atherosclerotic
patients. Other community-based prediction models for ischaemic stroke with 150% stenosis (RR = 3.3, 95% CI
mortality have been reported for shorter follow-up peri- 1.2–9.3), whereas at 5 years, the major predictor was dia-
ods (30 days to 3 years) [19–24]. betes (RR = 1.9, 95% CI 1.2–2.8) [12]. At 10 years in
Rochester, the major predictors were diabetes (RR = 1.7,
Predictors of Disability 95% CI 1.3–2.2) and increasing age (RR/10 years = 1.2,
In most models, the major predictor of early (30 days 95% CI 1.1–1.4) [16].
after stroke) dependency is stroke severity. In the Oxford- In the Northern Manhattan Stroke Study, baseline pre-
shire Community Stroke Project [22, 25], which is the dictors of recurrent stroke at 5 years were high alcohol
only externally validated predictive model, pre-stroke in- consumption (RR = 2.5, 95% CI 1.4–4.4), hypertension at
dependence, and normal Glasgow Coma Scale verbal discharge (RR = 1.6, 95% CI 1.01–2.6), and blood glucose
score, arm power, and ability to walk were significant pre- concentration at hospital admission (RR/10 mg/dl incre-
dictors of survival free of dependency at 6 months, where- ment = 1.4, 95% CI 1.1–1.7) [17].
as increasing age and living alone were associated with a The Perth Community Stroke Study also found older
poorer outcome for this parameter. In the Perth Commu- age, 75–84 years (RR = 2.6, 95% CI 1.1–6.2) and diabetes
nity Stroke Study, the major predictor of new disability at (RR = 2.1, 95% CI 1.0–4.4) to be major predictors of
5 years was recurrent stroke (odds ratio = 12.4, 95% CI recurrent stroke at 5 years, as was intracerebral haemor-
References
1 Sackett DL, Haynes RB, Guyatt GH, Tugwell 10 Scmidt EV, Smirnov VE, Ryabova VS: Results 19 Czlonkowska A, Ryglewicz D, Lechowicz W:
MD: Clinical Epidemiology. A Basic Science of the seven-year prospective study of stroke Basic analytical parameters as the predictive
for Clinical Medicine, ed 2. Boston, Little, patients. Stroke 1988;19:942–949. factors for 30-day case fatality rate in stroke.
Brown, 1991, pp 173–185. 11 Burn J, Dennis M, Bamford J, Sandercock P, Acta Neurol Scand 1997;97:121–124.
2 Counsell C, Dennis M: Systematic review of Wade D, Warlow C: Long-term risk of recur- 20 Carlberg B, Asplund K, Hagg E: The prognostic
prognostic models in patients with acute rent stroke after a first-ever stroke. The Oxford- value of admission blood pressure in patients
stroke. Cerebrovasc Dis 2001;12:159–170. shire Community Stroke Project. Stroke 1994; with acute stroke. Stroke 1993;24:1372–1375.
3 Terént A: Survival after stroke and transient 25:333–337. 21 Nakayama H, Jorgensen HS, Raaschou HO,
ischemic attacks during the 1970s and 1980s. 12 Petty GW, Brown RD Jr, Whisnant JP, Sicks Olsen TS: The influence of age on stroke out-
Stroke 1989;20:1320–1326. JD, O’Fallon WM, Wiebers DO: Ischemic come. The Copenhagen Stroke Study. Stroke
4 Dennis M, Bamford J, Sandercock P, Warlow stroke subtypes. A population-based study of 1994;25:808–813.
C: Prognosis of transient ischemic attacks in functional outcome, survival and recurrence. 22 Counsell C, Dennis M, McDowall M, Warlow
the Oxfordshire Community Stroke Project. Stroke 2000;31:1062–1068. C: Predicting outcome after acute and subacute
Stroke 1990;21:848–853. 13 Hankey GJ, Jamrozik K, Broadhurst RJ, stroke. Development and validation of new
5 Ricci S, Cantisani AT, Righetti E, Duca E, Forbes S, Burvill PW, Anderson CS, Stewart- prognostic models. Stroke 2002;33:1041–
Spizzichino L: Long-term follow-up of TIAs: Wynne EG: Five-year survival after first-ever 1047.
The SEPIVAC Study. Neuroepidemiology stroke and related prognostic factors in the 23 Elneihoum AM, Göransson M, Flake P, Janzon
1998;17:31–54. Perth Community Stroke Study. Stroke 2000; L: Three-year survival and recurrence after
6 Heyman A, Wilkinson WE, Hurwitz BJ, 31:2080–2086. stroke in Malmö, Sweden. An analysis of
Haynes CS, Utley CM, Rosati RA, Burch JG, 14 Hartmann A, Rundek T, Mast H, Paik MC, Stroke Registry Data. Stroke 1998;29:2114–
Gore TB: Risk of ischemic heart disease in Boden-Albala B, Mohr JP, Sacco RL: Mortality 2117.
patients with TIA. Neurology 1984;34:626– and causes of death after first ischemic stroke. 24 Bonita R, Ford MA, Stewart AW: Predicting
630. Neurology 2001;57:2000–2005. survival after stroke: A three-year follow-up.
7 Hankey GJ, Slattery JM, Warlow CP: The 15 Brønnum-Hansen H, Davidsen M, Thorvald- Stroke 1988;19:669–673.
prognosis of hospital-referred transient isch- sen P, for the Danish MONICA Study Group: 25 The FOOD Trial Collaboration: Performance
aemic attacks. J Neurol Neurosurg Psychiatry Long-term survival and causes of death after of a statistical model to predict stroke outcome
1991;54:793–802. stroke. Stroke 2001;32:2131–2136. in the context of a large, simple, randomized,
8 Carolei A, Candelise L, Fiorelli M, Francucci 16 Petty GW, Brown RD Jr, Whisnant JP, Sicks controlled trial of feeding. Stroke 2003;34:127–
BM, Motolese M, Fieschi C: Long-term prog- JD, O’Fallon WM, Wiebers DO: Survival and 133.
nosis of transient ischemic attacks and revers- recurrence after first cerebral infarction: A pop- 26 Hankey GJ, Jamrozik K, Broadhurst RJ,
ible ischemic neurologic deficits: A hospital- ulation-based study in Rochester, Minnesota, Forbes S, Anderson CS: Long-term disability
based study. Cerebrovasc Dis 1992;2:266– 1975 through 1989. Neurology 1998;50:208– after first-ever stroke and related prognostic
272. 216. factors in the Perth Community Stroke Study,
9 Howard G, Evans GW, Rouse JR III, Toole JF, 17 Sacco RL, Shi T, Zamanillo MC, Kargman DE: 1989–1990. Stroke 2002;33:1034–1040.
Ryu JE, Tegeler C, Frye-Pierson J, Mitchell E, Predictors of mortality and recurrence after 27 Hankey GJ, Jamrozik K, Broadhurst RJ,
Sanders L: A prospective reevaluation of tran- hospitalized cerebral infarction in an urban Forbes S, Burvill PW, Anderson CS, Stewart-
sient ischemic attacks as a risk factor for death community: The Northern Manhattan Stroke Wynne EG: Long-term risk of first recurrent
and fatal or nonfatal cardiovascular events. Study. Neurology 1994;44:626–634. stroke in the Perth Community Stroke Study.
Stroke 1994;25:342–345. 18 Anderson CS, Jamrozik KD, Broadhurst RJ, Stroke 1998;29:2491–2500.
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.
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 low-, intermediate- and high-risk strata on the basis of
clopidogrel for the primary endpoint was observed across TIMI risk scores. The results of this study demonstrate a
a range of prespecified subgroups. These included higher- consistent benefit of clopidogrel across the TIMI risk cate-
risk patients such as those with diabetes, those with pre- gories, with the highest absolute benefit in the stratum
vious MI, and those with elevated cardiac enzymes or with the highest risk.
markers at baseline. Of particular interest to the neurolo- Very recently, the CREDO (Clopidogrel for Reduction
gist are the observations that in the 506 CURE patients of Events During Observation) trial has confirmed the sig-
who had a prior history of stroke, the overall incidence of nificant and sustained benefit of long-term clopidogrel
the primary study endpoint was higher than in those with therapy on top of standard therapy including ASA in
no such history (n = 12,056 patients), while the benefit of patients undergoing elective PCI. In CREDO, 2,116 pa-
clopidogrel was consistent between the two patient popu- tients referred for PCI or who were deemed at high likeli-
lations. This finding parallels a recent analysis by Budaj et hood of PCI were randomized to receive a preprocedural
al. [34], in which CURE patients were categorized into clopidogrel loading dose or placebo. After PCI, all pa-
tients received clopidogrel 75 mg/day for 28 days. Be- safety of long-term clopidogrel on top of ASA versus clo-
tween 29 days and 1 year, patients in the loading-dose pidogrel alone in an international, randomized, double-
group received clopidogrel 75 mg/day, and those in the blind trial in patients with recent TIA or ischaemic stroke
control group received placebo. All patients received ASA who have additional factors that are associated with an
for the duration of the study. Long-term data showed that increased risk of atherothrombotic recurrence, i.e. pre-
continuation of clopidogrel for 12 months rather than 1 vious ischaemic stroke, previous MI, prior history of angi-
month after PCI was associated with a 26.9% relative risk na [38], symptomatic peripheral disease [39], or history of
reduction for the composite of death, MI or stroke at diabetes (fig. 3) [40]. Patients are randomized to receive
1 year (p = 0.023) (fig. 2) [35]. As was observed in the ASA 75 mg/day or matching placebo, and all patients
CURE study, long-term clopidogrel therapy in CREDO receive clopidogrel 75 mg/day as part of standard therapy.
reduced the incidence of each component of the primary Patients receive study medication and are followed for 18
endpoint, including a 25.1% reduction for all-cause months after randomization. The primary endpoint for
stroke. Moreover, administration of a clopidogrel loading efficacy is a composite of ischaemic stroke, MI, vascular
dose 66 h before PCI was associated with a 38.6% reduc- death or rehospitalization for acute ischaemia during the
tion in the composite of death, MI, or urgent target vessel treatment period. Patient enrolment into MATCH was
revascularization at 28 days (p = 0.05). These data rein- completed in April 2002, with a total of 7,601 patients
force the benefit of early and sustained clopidogrel ad- recruited in 28 countries. Initial data from this important
ministration that was observed in the PCI-CURE sub- trial are anticipated in 2004.
study of CURE patients who underwent PCI after ran- Additional trials of clopidogrel in neurology are
domization [36]. planned. ATARI will use a factorial design to compare the
use of clopidogrel, low-molecular-weight heparin, both, or
Future Prospects for Clopidogrel in Neurology neither on a background of ASA therapy in patients who
Although CURE has demonstrated the benefit of clopi- have suffered a TIA in the previous 12 h. The primary
dogrel on top of standard therapy (including ASA) in endpoint in ATARI will be a composite of vascular death,
patients with unstable angina/non-Q-wave MI [33], this stroke or MI. The Secondary Prevention of Small Subcor-
regimen has not been proven in patients with recent man- tical Strokes (SPS3) trial will evaluate the efficacy of clopi-
ifestations of cerebrovascular atherothrombosis. This is dogrel in patients with small subcortical strokes (lacunar
being addressed in the ongoing MATCH (Management of infarcts). Although this stroke subtype accounts for
Atherothrombosis with Clopidogrel in High-Risk Patients around 25% of all cerebral infarcts, the effects of anti-
with Recent Transient Ischaemic Attack or Ischaemic platelet therapy and blood pressure lowering have not spe-
Stroke) study [37]. MATCH is comparing the efficacy and cifically been studied in this clinical setting. SPS3 will
References
1 Ruggeri ZM: Platelets in atherothrombosis. 11 Savi P, Labouret C, Delesque N, Guette F, 18 Antiplatelet Trialists’ Collaboration: Collabo-
Nat Med 2002;8:1227–1234. Lupker J, Herbert JM: P2Y12, a new platelet rative overview of randomised trials of anti-
2 Drouet L: Atherothrombosis as a systemic dis- ADP receptor, target of clopidogrel. Biochem platelet therapy. I. Prevention of death, myo-
ease. Cerebrovasc Dis 2002;13(suppl 1):1–6. Biophys Res Commun 2001;283:379–383. cardial infarction and stroke by prolonged anti-
3 Broderick JP, Phillips SJ, O’Fallon WM, Frye 12 Vane JR, Meade TW: Second European Stroke platelet therapy in various categories of pa-
RL, Whisnant JP: Relationship of cardiac dis- Prevention Study (ESPS 2): Clinical and phar- tients. BMJ 1994;308:81–106.
ease to stroke occurrence, recurrence and mor- macological implications. J Neurol Sci 1997; 19 Peters RJG, Zao F, Lewis BS, Fox KAA, Yusuf
tality. Stroke 1992;23:1250–1256. 145:123–125. S on behalf of the CURE Investigators: Aspirin
4 Dennis MS, Burn JP, Sandercock PA, Bamford 13 Antithrombotic Trialists’ Collaboration: Col- dose and bleeding events in the CURE study.
JM, Wade DT, Warlow CP: Long-term surviv- laborative meta-analysis of randomised trials Eur Heart J 2002;23(abstr suppl):510.
al after first-ever stroke: The Oxfordshire Com- of antiplatelet therapy for prevention of death, 20 Algra A, van Gijn J: Cumulative meta-analysis
munity Stroke Project. Stroke 1993;24:796– myocardial infarction and stroke in high-risk of aspirin efficacy after cerebral ischaemia of
800. patients. BMJ 2002;324:71–86. arterial origin. J Neurol Neurosurg Psychiatry
5 Hartmann A, Rundek T, Mast H, Paik MC, 14 CAPRIE Steering Committee: A randomised, 1999;66:255.
Boden-Albala B, Mohr JP, Sacco RL: Mortality blinded, trial of clopidogrel versus aspirin in 21 Hankey GJ, Sudlow CLM, Dunbabin DW:
and causes of death after first ischemic stroke: patients at risk of ischaemic events (CAPRIE). Thienopyridines or aspirin to prevent stroke
The Northern Manhattan Stroke Study. Neu- Lancet 1996;348:1329–1339. and other serious vascular events in patients at
rology 2001;57:2000–2005. 15 Diener H, Cunha L, Forbes C, Sivenius J, high risk of vascular disease? A systematic
6 Sacco RL: Risk factors and outcomes for isch- Smets P, Lowenthal A: European Stroke Pre- review of the evidence from randomized trials.
emic stroke. Neurology 1995;45:S10–S14. vention Study. 2. Dipyridamole and acetylsali- Stroke 2000;31:1779–1784.
7 Hankey GJ: Long-term outcome after isch- cylic acid in the secondary prevention of 22 Cannon C: Effectiveness of clopidogrel versus
aemic stroke/transient ischaemic attack. Cere- stroke. J Neurol Sci 1996;143:1–13. aspirin in preventing acute myocardial infarc-
brovasc Dis 2003;16(suppl 1):14–19. 16 CAST (Chinese Acute Stroke Trial) Collabora- tion in patients with symptomatic athero-
8 Albers GW, Amarenco P, Easton JD, Sacco tive Group: CAST: Randomised placebo-con- thrombosis (CAPRIE trial). Am J Cardiol
RL, Teal P: Antithrombotic and thrombolytic trolled trial of early aspirin use in 20,000 pa- 2002;90:760–762.
therapy for ischemic stroke. Chest 2001;119: tients with acute ischaemic stroke. Lancet 23 Easton JD: Benefit of clopidogrel in patients
300S–320S. 1997;349:1641–1649. with symptomatic cerebrovascular disease.
9 European Stroke Initiative Executive Commit- 17 International Stroke Trial Collaborative Neurology 1998;50(suppl 4):A157.
tee: European Stroke Initiative recommenda- Group: The International Stroke Trial (IST): A 24 Easton JD: Net benefit of clopidogrel over aspi-
tions for stroke management. Cerebrovasc Dis randomized trial of aspirin, subcutaneous hep- rin for the prevention of atherothrombotic
2000;10:335–351. arin, both or neither among 19,435 patients events. Cerebrovasc Dis 1998;8(suppl 4):46.
10 Vane JR: Inhibition of prostaglandin synthesis with acute ischaemic stroke. Lancet 1997;349:
as a mechanism of action for aspirin-like drugs. 1569–1581.
Nat New Biol 1971;231:232–235.