Transcript Stroke prevention, how well are we doing? Professor

From CSRG: thrombolysis for
acute ischaemic stroke
Peter Sandercock, on behalf of Joanna Wardlaw &
Veronica Murray
IST-3 Italian Stroke Forum
Firenze
13th February 2009
Joanna Wardlaw & Veronica Murray
Outline = structure of a review
• Competing interests
• History of this review
• Methods & protocol for update
– Types of studies to include
– Main outcomes
– Planned subgroups
• New data included in the update
• Analyses
• Implications
– For clinical practice
– For future research
Competing interests
• JMW: SITS-MOST Steering and CT
adjudication
• JMW: ECASS 3 CT reading Committee
• JMW & PS IST-3 lead investigators
• VM IST-3 coordinator for Sweden
• IST-3 donation of drug and placebo for first
300 patients from BI
• No funding from any pharmaceutical
company for this review
History of thrombolysis review
• Initiated: (before Cochrane collaboration!)
1990, first published in Cochrane Library
1995
• Inclusion criteria: all randomised controlled
trials of any thrombolytic drug versus control
• Primary outcome: death or dependency
(MRS 3-6) at final follow-up.
• 2003 update: 18 trials, 5675 patients (only
42 patients aged over 80), drugs = rtPA,
streptokinase, uro-kinase, rPro-urokinase,
time = 0-6 hrs, Brain Imaging: CT
Methods for the 2009 update
Included studies
• New trials completed since 2003
• New data from existing trials
Search strategy
• Searches for trials from multiple sources
(including Cochrane Stroke Group
Specialised Register of Trials)
• Two independent reviewers extracted data
Methods – data extracted
Outcomes assessed in previous review:
• Intracranial haemorrhage
• Death early and late,
• Poor functional outcome
• Infarct early swelling,
Subgroups in previous review
• Time to treatment,
• Antithrombotic treatment,
• Stroke severity,
• mRS cut point,
New subgroups :
• Type of imaging, CT or MR
• Presence of ’infarct signs’ on baseline CT,
• Stroke subtype (large artery or lacunar)
New trial data added to review
• 8 trials (1,477 patients)
• Drugs tested:
– 3 rt-PA (ECASS-3, EPITHET, Wang)
– 2 Urokinase (AUST, MELT)
– 3 desmoteplase (DIAS 1&2, DEDAS)
• Route: 2 intra-arterial, 6 intravenous
• Time from onset: 0-6, 3-4.5, 3-9, 0-24 hrs
• Imaging pre randomisation:
–CT: 5
–MR: 3 (+1) DWI/PWI mismatch
• Age over 80: no new data
Summary of effects on main
outcomes. Odds Ratios (95% CI)
SICH
(incl fatal)
Dead
All drugs
n=7152
3.3
2.7 - 4.1
p<0.00001
1.3 *
1.1 - 1.5
p=0.06
rt-PA
n=3977
3.1
2.3 - 4.0
p<0.00001
1.1
1.0 - 1.4
p=0.16
Dead or
dependent
0.8 *
0.7 - 0.9
p<0.0001
0.8 *
0.7 - 0.9
p<0.0001
Significant heterogeneity confounds interpretation:
meta-regression on a variety of factors does not explain it
IV rt-PA < 6hrs only: effect on
death or dependency (mRS 3-6)
= trial completed recently
Odds ratio = 0.78 (0.68-.88)
Heterogeneity (Chi2 p=0.007) I2 = 62%
Test for overall effect p=0.0001
Wardlaw et al 2008
Sensitivity analysis: how robust is the result? Does
it change with the choice of mRS cut-off?
Modified Rankin (mRS):
mRS 2 to 6
3 to 6
IV tPA vs control
Mori
NINDS
ECASS
ECASS 2
ECASS 3
Atlantis A
Atlantis B
rt-PA subtotal
0.1
0.78 1
5
0.1
0.77 1
thrombolysis better
Heterogeneity highly significant :
p=0.007
5
thrombolysis worse
p= 0.006
Secondary outcome: effect of iv rt-PA
on symptomatic cerebral oedema
Odds ratio 0.79 (0.62- 1.01)
p = 0.06
Wardlaw et al 2008
Summary 2008
• No material change in estimates of effect on
major outcomes since 2003.
• i.v. rt-PA:
– Heterogeneity still confounds interpretation of
primary, but not secondary, outcomes
– ECASS 3 consistent with existing rt-PA metaanalysis.
– Interesting effect on symptomatic cerebral
oedema
– Evidence of benefit to at least six hours and
possibly beyond, but in whom?
• Other drugs, other routes: promising but
unproven
IMPLICATIONS FOR PRACTICE:
Even if the EU approval for thrombolysis is
extended to 4.5 hrs, this will still exclude
patients who:
•
•
•
•
•
•
Are aged > 80 years
Have ‘very mild stroke’ or NIHSS > 25
Had prior stroke within the last 3 months
Have a history of prior stroke + Diabetes
Arrive at 4.5 to 6.0 hours
Have other relative contraindications
specified in the licence (e.g. ‘extensive
infarction’, which is not defined in any way)
IMPLICATIONS FOR RESEARCH.
More randomised trial evidence needed on
effects of i.v. rt-PA:
• When used <6hrs (and beyond 6hrs too?)
• In particular categories of patients:
– Aged > 80
– Different subtypes,
– Mild stroke, sever stroke
• On symptomatic massive cerebral oedema
• Clinical and imaging factors that determine
– benefit from treatment
– risk of symptomatic intracranial haemorrhage
– In whom perfusion or angiographic imaging is
necessary?
Grazie
Effect of IV rt-PA < 6hrs on
death at the end of FU
OR 1.14 (95% CI 0.95-1.30)
Primary outcome: Death or dependency at the end of follow-up
IV urokinase
IV streptokinase
0.91 (0.64, 1.42)
0.94 (0.72, 1.24)
IV rt-PA
0.77 (0.47, 0.89)
IV streptokinase
+ aspirin
1.09 (0.49, 1.72)
IA pro-urokinase
0.55 (0.31, 1.00)
IA urokinase
0.57 (0.28, 1.14)
IV desmoteplase
0.85 (0.53, 1.38)
Total
0.82 (0.73, 0.91)