New developments in endovascular therapy or sfoktr...
Transcript of New developments in endovascular therapy or sfoktr...
New developments in endovascular therapy
for stroke
Dr Bruce Campbell Consultant Neurologist and Head of Hyperacute Stroke Royal Melbourne Hospital, University of Melbourne
1
Acute Stroke interventions: evidence base Interven�on Outcome RRR ARR NNT
All stroke types and severi�es
Stroke Unit care Cochrane 2007
Death/Dependency 17 3.6 28
Ischemic Stroke
tPA <4.5hr Emberson et al, 2014
mRS 0-‐1 20 6.8 15
thombectomy <6hr MR-‐CLEAN ,EXTEND-‐IA,
ESCAPE, SWIFT-‐PRIME, 2015
Death/Dependency 70 30 3
Aspirin IST, CAST 1997
Recurrent stroke/Death 10 0.9 111
Hemicraniectomy Vahedi K et al, 2007
Death 69 49 2
Intracerebral hemorrhage
BP lowering in ICH Anderson et al, 2013
Death/Dependency 6.5 3.6 28 2
Number Needed to Treat 4.5 9 14.1
tPA -‐ Time is BRAIN!
Our aim: shi� pa�ents up the curve!
Emberson Lancet 2014
Excellent outcome (mRS 0-‐1) n=6756
3
4
USA – “TARGET-‐STROKE”
Fonarow JAMA 2014
15min reduction in DTN from 74min to 59min based on prenotification and direct to CT associated with:
5
Generalized benefit of tPA
Emberson Lancet 2014 6
Do we need be�er than tPA?
7
Saver et al UCLA Stroke Center
WHY??
Non-‐contrast CT “selec�on” includes: -‐ already reperfused -‐ no penumbra remaining more advanced imaging selec�on?
tPA doesn’t always open the artery be�er reperfusion strategies
Effect of IV tPA on outcome 0-‐3hr post onset
8
Site of occlusion Success at 2 hours
Caro�d terminus 5%
MCA M1 30%
MCA M2 42%
Basilar 11%
Overall 30%
Early Recanaliza�on a�er IV tPA
9
Distribution of reperfusion Distribution of change in NIHSS
Selection = visually assessed CT perfusion mismatch + arterial occlusion tPA 0.1mg/kg TNK 0.25mg/kg tPA 0.1mg/kg TNK 0.25mg/kg
10
Desmoteplase?
11
Endovascular Therapy
12
13
SYNTHESIS
Italian trial n=362 – compared IV tPA with “direct to endovascular” – ~1hr delay between when tPA was started and when IA was started
– no requirement for proven vessel occlusion and loose clinical criteria
– mostly intra-‐arterial tPA and a few of the older devices – no data released on reperfusion rates
14
SYNTHESIS
Ciccone NEJM 2013
Italian n=362 IV vs IA (no bridging) Selec�on based on tPA criteria only No requirement for vessel occlusion
6% sICH both groups
15
MR-‐RESCUE
USA trial n=118 – compared endovascular to standard care (which included tPA in
%) – MRI or CT perfusion imaging prior but not used for selec�on – Different defini�on of “penumbra” to the EPITHET-‐DEFUSE
groups, allowed core volume up to 90mL as “penumbral” (we limit to 70mL)
– older devices – 27% TICI 2b/3 (>50% reperfusion) – only 6 “penumbral” pa�ents using EPITHET-‐DEFUSE criteria had
reperfusion – suspected selec�on bias due to lack of equipoise in inves�gators
(glacial recruitment – 22 sites, 7 years)
16
MR-‐RESCUE
Kidwell NEJM 2013 17
IMS-‐3
large NIH funded trial stopped for fu�lity a�er n=656 – compared IV tPA versus IV tPA (mostly 2/3 of standard dose) + endovascular
– no requirement to prove arterial occlusion
– older devices – 40% TICI 2b/3 rate (ie >50% reperfusion) – suspected selec�on bias due to lack of equipoise in inves�gators
18
OVERALL
IV-‐IA
IV only
NIHSS 8-‐19
IV-‐IA
IV only
NIHSS ≥20
IV-‐IA
IV only
IMS-‐3
Broderick NEJM 2013
p=0.25 for shi�
p=0.83 for shi�
p=0.06 for shi�
19
IMS-‐3 IA success at end of procedure:
TICI 2-‐3 TICI2b-‐3
Overall 74% 40%
Extracranial ICA 83% 33%
Intracranial ICA 65% 38%
ie subop�mal IA strategy (1st genera�on devices & IA tPA)
0
20
40
60
80
100
0 1 2a 2b 3
Good outcome (mRS 0-‐2)
20
IMS-‐3
Time?
mean �me to recanaliza�on = 5hr25min (range 3-‐7hr)
~3.5hr delay to consent, randomise, transport and do IA Rx
every 30min delay a/w 14% RRR in independent outcome (mRS 0-‐2)
21
so how do we do be�er?
be faster increase recanaliza�on success – new devices (and raise the bar for defining success – TICI 2b/3) stop “an�-‐selec�ng” ie trea�ng the best candidates open label!
Exclude pa�ents likely to be fu�le/risky (large core)
22
Excellent outcome (mRS 0-‐1) n=6756
IV tPA meta-‐analysis
Emberson Lancet 2014
IV tPA -‐ Time is BRAIN!
23
IA therapy – �me is s�ll brain (in unselected pa�ents)
Khatri Neurology 2009
IMS-‐3 median onset-‐recan
24
Collaterals ma�er imaging selec�on
Ribo Stroke 2011
Good Collaterals
25
More effec�ve devices
26
SWIFT & TREVO-‐2
TICI 2b/3 68% vs 44% 27
STAR Solitaire registry Pa�ent characteris�cs (n=188)
combined IV-‐IA 59%
Age 72 (18-‐85) NIHSS 17 (8-‐30)
ICA 18% M1 67% M2 14%
0-‐3hr 26% 3-‐4.5hr 38% >4.5hr 36%
Procedural characteris�cs Mean ± SD
number of passes 1.5 ± 0.7
onset to groin 251 ± 99min
Median groin to comple�on 44min
Symptoma�c hemorrhage 1.5%
Success (TICI 2b/3) = 79% core lab adjudicated (85% a�er “rescue”) mRS 0-‐2 58%
Pereira ISC 2013
28
Selec�on beyond occlusion? ASPECTS NCCT
CTP mismatch
DWI –MRA or diff-‐perf mismatch Collateral score (mul�phasic CTA)
Clot length
29
30
MR-‐CLEAN
n=500 Dutch trial – 19 centres, universal enrolment due to government funding restricted to trial pa�ents
– “endovascular <6hr vs standard care” but 90% were tPA vs tPA + IA
– vessel occlusion documented on non-‐invasive imaging, no penumbral selec�on
31
MR-‐CLEAN
Age (~65) and NIHSS (~17) well matched ~60% M1, 25% ICA + a few M2s Although protocol was IA vs best medical therapy ~90% had tPA as “medical therapy” Although protocol was liberal ~97% of pa�ents were treated with “stentrievers” – the newer more effec�ve devices TICI 2b/3 (>50% reperfusion) was 58% (versus 40% IMS3 and 27% MR-‐RESCUE)
32
MR-‐CLEAN
Timing: onset to tPA ~90min
onset to randomiza�on ~200min
onset to procedure ~260min ie tPA to groin ~2hr50min
33
MR-‐CLEAN
!"# $!"# %!"# &!"# '!"# (!"# )!"# *!"# +!"# ,!"# $!!"#
-./0123#4#.5/67281932:#
-./0123#;<.:2=>#
!# $# %# &# '# (# )#
mRS IA vs medical (unadj p-‐value)
0-‐1 (back to all ac�vi�es) 11.6 vs 6% p=0.037
0-‐2 (Independent) 32.6% vs 19.1% p=0.001 NNT 7
0-‐3 (Ambula�ng independently, some assistance with domes�c ADLs)
51.1 vs 35.6% p=0.001
6 (death) 21.0% vs 22.0% p=0.77
ORDINAL SHIFT OR 1.67 (95% CI 1.21-‐2.30) 34
MR-‐CLEAN
Secondary endpoints NIHSS reduc�on at 24hr and 7 days significant recanaliza�on at 24hr -‐ 33% medical vs 75% IA symptoma�c hemorrhage 11% IA vs 10% medical (PH2 – large parenchymal haematoma 6% vs 5%)
subgroups – no heterogeneity in treatment effect across age (>/<80) or NIHSS strata
35
MR-‐CLEAN
well conducted RCT no penumbral selec�on in protocol (but many had CTP, only a small number had ASPECTS 0-‐4 non-‐contrast CT scans) clear posi�ve result for func�onal outcome and consistent with secondary outcomes of recanaliza�on and early NIHSS response.
36
IMS-‐3 + MR-‐CLEAN
The two “bridging” IV vs IV-‐IA trials
!"# $!"# %!"# &!"# '!"# (!"# )!"# *!"# +!"# ,!"# $!!"#
-./01.2#3/4'+$5#
6/017871091:72#3/4)'+5#
6;#
6;86<
#
!# $# %# &# '# (# )#
mRS 0-1 – p=0.001 mRS 0-2 – p<0.001 mRS 0-3 – p<0.001 Death – p=0.45 Shift analysis p<0.001
! " # $ % & '()*+,)-./*0%1"2 "3 && '$ 43 """ %4 "!45*+,676,+8,96-./*0'%12 &3 3! "!% ""% ""' $$ "$#
37
MR-‐CLEAN
ISC 2015 38
MR-‐CLEAN: �me to reperfusion
ISC 2015 39
Extending the �me for Thrombolysis in Emergency Neurological Deficits – Intra-‐Arterial
A randomized controlled trial of endovascular thrombectomy a�er standard dose intravenous t-‐PA within 4.5 hours of stroke onset
u�lizing dual target imaging selec�on
Acknowledging support from: Solitaire FR™ device supplied free of charge by
Bruce Campbell Peter Mitchell Co-‐PI and Medical Coordinator Co-‐PI and Head of Neurointerven�on
Stephen Davis and Geoffrey Donnan Co-‐chairs
ClinicalTrials.gov NCT01492725
40
41
Ra�onale
In 2011 (before IMS-‐3, SYNTHESIS, MR-‐RESCUE neutral) – Fu�le recanaliza�on noted in single arm studies (need be�er selec�on)
– Major improvement in procedural success observed in pa�ents ineligible for IMS-‐3 treated with stentrievers versus our pa�ents eligible for IMS-‐3
Aimed to select pa�ents with best chance of response to reperfusion -‐ “dual target” – proven major vessel occlusion and
– salvageable �ssue downstream with ischemic core <70mL (CT perfusion)
Treat as fast as possible (no wai�ng to assess “tPA failure”)
Use the most effec�ve device (stentriever)
42
CT perfusion
Clinically useful to improve diagnos�c certainty1
With appropriate thresholding provides similar informa�on on the presence of penumbra to MRI2,3
Fast acquisi�on and processing ~6min (and need IV access and contrast for CT angiogram anyway)4
1Campbell et al JNNP 2013;84:613 2,3Campbell et al Stroke 2011;42:3435, 2012;43:2648 4Campbell et al Int J Stroke 2015;10:51-54
43
CT perfusion MRI
RAPID for CT and MRI Fast, standardized, fully automated, quan�ta�ve, thresholded mismatch
CT relCBF / Diffusion MRI
RAPID ischemic core segmenta�on CT relCBF<30%
Tmax
RAPID Tmax>6sec segmenta�on
Ischemic core: 6mL Perfusion lesion: 58mL Mismatch ra�o = 9.7 Absolute mismatch = 52mL
Randomize pa�ent
Ischemic core: 7mL Perfusion lesion: 55mL Mismatch ra�o = 7.6 Absolute mismatch = 48mL
Randomize pa�ent
Straka et al JMRI 2010 44
Example – mismatch
Ischemic core volume: 24mL Perfusion (Tmax>6s) lesion: 119mL
45
70mL ischemic core
Ischemic core volume: 73mL Perfusion (Tmax>6s) lesion: 88mL
46
Example – large ischemic core
Ischemic core volume: 146mL Perfusion (Tmax>6s) lesion: 215mL 47
Inclusion criteria:
Age ≥18 years (no upper limit), No NIHSS restric�ons
Anterior circula�on ischemic stroke eligible for intravenous tPA within 4.5 hours of stroke onset
Imaging “dual target”
– Major vessel occlusion – ICA, M1, M2 amenable to clot retrieval
– RAPID “Mismatch” CT or MRI Mismatch ra�o > 1.2 AND Absolute mismatch >10mL AND Ischemic core <70mL
Able to commence intra-‐arterial therapy within 6 hours of onset
Informed consent obtained from pa�ent or person responsible
Exclusion criteria:
Premorbid disability (mRS≥2)
Inability to access the cerebral vasculature in the opinion of the neurointerven�onal team
Contra-‐indica�on to imaging with contrast agents
Major neurological recovery (to NIHSS 0-‐1 or inves�gator’s discre�on) prior to randomiza�on.
ie BEFORE RANDOMIZATION neurologists had to check that the neurointerven�onist was available and had assessed proximal access
* Campbell et al Stroke 2011 Bivard et al Cerebrovasc Dis 2011 Kamalian et al Stroke 2011
48
10.15am Solitaire FR device 10:35am
Eligible Pa�ents
TRIAL DESIGN -‐ PROBE design, planned 100 pa�ents
24hr MRI reperfusion* (recan/growth/ICH) 24hr NIHSS
3 day NIHSS*
90 day NIHSS & mRS *co-‐primary outcome
Blinded outcomes
IV tPA only
Randomise 50:50 (web-‐based)
0.9mg/kg IV tPA + Solitaire FR clot retrieval -‐ start asap (<6hr)
49
2 4 7 9 11 13
16 17 17 19
25 27 29 30
35 38 40
43 45 47 49
55 60
63 66 68
70
0
20
40
60
80
100
120
No. Par�cipants
EXTEND-‐IA Recruitment
No. Par�cipants Randomised -‐ Expected
No. Par�cipants Randomised -‐ Actual
14 centres in Australia and New Zealand
MR-‐CLEAN results DSMB halted EXTEND-‐IA recruitment (Haybi�le-‐Peto boundary for efficacy)
Planned n=100
IMS3/SYNTHESIS/MR-‐RESCUE NEJM
50
Generalizability
Acute Ischemic stroke patients
n= 7798
Patients treated with alteplase n= 1044 (13%)
EXTEND-IA n= 70
Listed reason for exclusion:
No major vessel occlusion (n=495, 47%)
Out of operating hours (n=225, 22%)
Poor pre-morbid function (n=171, 16%)
Large ischemic core (n=42, 4%)
Cervical arterial dissection or poor vascular access (n=18, 1.7%)
Inability to obtain multimodal CT imaging (n=16, 1.5%)
Inability to obtain consent (n=7, 0.7%)
85%
51
Generalizability -‐ RMH
In largest center 21 randomized + 7 with occlusion and clinical eligibility excluded for large core ie ~25% (95%CI 11-45%) excluded by CTP
Endovascular appropriate for ~37-40% of tPA patients + tPA ineligible
Acute Ischemic stroke patients
N=954
Patients treated with alteplase N=178 (19%)
EXTEND-IA n= 21
(12% of tPA)
Hierarchical reason for exclusion:
No major vessel occlusion (n=106, 60%)
Poor pre-morbid function (n=28, 16%)
Out of operating hours (n=13, 7%)
Large ischemic core (n=7, 4% BUT 25% of otherwise eligible)
Cervical arterial dissection or poor vascular access (n=1, 0.5%)
Inability to obtain multimodal CT imaging (n=0, 0%)
Inability to obtain consent (n=2, 1%)
83%
52
CONSORT trial profile
Randomized (n=70)
Alteplase + Endovascular (n=35) Alteplase only (n=35)
Did not receive Angiogram (n=2)
Received tPA only (n=35)
Received angiogram (n=33)
1 – major improvement (temporary) 1 – major deterioration (unrecognized 2nd embolism)
4 – already recanalized by tPA 1 – ICA stent sufficient to restore flow 1 – wire perforation pre-deployment
Solitaire device not deployed (n=6)
Solitaire device deployed (n=27)
All analyses = intention to treat 53
Demographics Characteristic IV tPA only IV tPA +
endovascular Number 35 35
Age – yr: Mean (SD) 70.2 (11.8) 68.6 (12.3)
Male sex – no. (%) 17 (48.6%) 17 (48.6%)
NIHSS score: Median (IQR) 13 (9-19) 17 (13-20)
Onset to tPA – min Median (IQR) 145 (105-180) 127 (93-162)
Door-to-needle – min Median (IQR) 46 (35-70) 43 (19-61)
tPA to randomization – min Median (IQR) 36 (18-55) 29 (23-46)
54
Ini�al imaging Characteristic IV tPA only IV tPA +
endovascular Site of vessel occlusion ICA MCA-M1 MCA-M2
11 (31%) 18 (51%) 6 (17%)
11 (31%) 20 (57%) 4 (11%)
Baseline Ischemic core – mL Mean, Median (IQR)
19.6, 18 (4-29)
18.9, 12.3 (4-32)
Baseline Perfusion lesion – mL Mean, Median (IQR)
116, 115 (72-158)
105, 106 (76-137)
55
Procedural Characteris�cs
ie s�ll room for improvement as becomes standard care
Characteristic median, (IQR) IV tPA + endovascular
ONSET to groin puncture – min 210 (166-251)
CT to groin puncture – min 93 (71-138)
Groin puncture to TICI 2b/3 or completion 43 (24-53)
ONSET to mTICI 2b/3 or completion 248 (204-277)
* MR-CLEAN onset to groin = 260min = 50 min slower * IMS-3 onset to groin = 208min - similar * IMS-3 onset to recanalization = 325min = 77min slower
56
Procedural Characteris�cs Characteristic median, (IQR) IV tPA + endovascular
Final mTICI (core lab adjudicated) mTICI 2b/3 25/29 (86%)
mTICI 3 14/29 (48%)
mTICI 2b/3 ICA
M1
M2
10/11 (91%)
14/15 (93%)
1/3 (33%)
57
Co-‐primary outcome
58
Reperfusion at 24hr (Inten�on to treat)
median 37% vs 100% p<0.0001
tPA only tPA + endovascular Treatment Group
% Reperfusion at 24 hours
59
>90% Reperfusion at 24hr without SICH
Pa�e
nts (%
)
p<0.001
60
Reperfusion and mRS:
7
30
4
28
19
14
7
19
19
5
11
2
33
2
0% 20% 40% 60% 80% 100%
Reperfusion <90%
Reperfusion>90%
0 1 2 3 4 5 6
70% chance of death/dependence
Patients (%)
Modified Rankin Scale Score No symptoms Death
61
Early neurological recovery Reduc�on of ≥8 NIHSS points or reaching 0-‐1 by day 3
Pa�e
nts (%
)
p=0.002
62
NIHSS recovery
P=0.001 P=0.007 P=0.03
63
17
26
11
26
11
20
11
17
17
3
11 20
9
0% 20% 40% 60% 80% 100%
0 1 2 3 4 5 6
Combined Intravenous t-PA and Endovascular Therapy
Intravenous t-PA alone
Patients (%)
Ordinal p=0.006 (unadj), p=0.02 (adj) NNT 3 for ≥1 point better on mRS
mRS 0-2 p=0.01 71% vs 40% - NNT 3.2 for independence
mRS 0-1 p=0.09
Day 90 mRS
64
Recanaliza�on at 24hr
Characteristic
median, (IQR)
IV tPA only IV tPA +
endovascular
TIMI 2-3 (overall) 15/35 (47%) 33/35 (94%)
p<0.001
ICA M1 M2
2/11 (18%) 9/18 (50%) 4/6 (67%)
10/11 (91%) 19/20 (95%) 4/4 (100%)
65
Infarct volume & growth
Characteristic
median, (IQR)
IV tPA only IV tPA +
endovascular
Baseline Ischemic core 18 (4-29) 12.3 (4-32)
24hr Ischemic core 49 (17-82) 18 (12-52)
Infarct Growth 35 (6-73) 11 (0-24)
p=0.007 (adj)
66
Adverse events
Adverse Event IV tPA only IV tPA + endovascular p value
Deaths 7/35 (20%) 3/35 (9%) 0.18
SICH* 2/35 (6%) 0/35 (0%) 0.49
PH§ 3/35 (9%) 4/35 (11%) 0.99
Wire perforation - 1/35 (2.9%) -
Emboli - 2/35 (5.7%) -
* pre-specified SITS definition = PH2 + ≥4 point increase NIHSS § PH = parenchymal hematoma
67
Outcome of mTICI 2b vs 3
10/11 (91%) of mTICI 2b had 100% reperfusion by 24h perfusion imaging
10/11 (91%) had major early neurological recovery versus 13/14 (93%) TICI 3 pa�ents 8/11 (73%) had mRS 0-‐2 at 90 days versus 10/14 (71%) TICI 3 pa�ents
(and the 4 extra cases that were 2b at ini�al angiogram (due to tPA) also proceeded to 100% reperfusion by 24h)
68
Limita�ons
Sample size 70 -‐ unable to interrogate subgroups (planned individual pa�ent meta-‐analysis) Cannot exclude some benefit of endovascular therapy in pa�ents excluded from this trial on the basis of large ischemic core/no mismatch.
Purely volume-‐based criteria do not account for the loca�on of the ischemic core, which is also relevant to clinical outcome. Early termina�on of the trial does create poten�al for overes�ma�on of the effect size.
69
Conclusions
early mechanical stent-‐thrombectomy a�er tPA using Solitaire FR led to: – Faster and more complete reperfusion
In this popula�on selected for vessel occlusion and salvageable �ssue this translated to: – Improved early neurological recovery – Improved func�onal outcome at 3 months – No safety concerns
70
Implica�ons
tPA + mechanical stent-‐thrombectomy should be the new standard of care
Systems re-‐organiza�on and transfer protocols
Details of selec�on paradigm remain a key discussion – EXTEND-‐IA indicates a popula�on with high probability of major clinical response but others may derive some benefit – requires further study
71
Acknowledgements
Recrui�ng Sites Royal Melbourne Hospital (21) B.C.V. Campbell, P.J. Mitchell, S.M. Davis, B. Yan, R.J. Dowling,
N. Yassi, T.J. Oxley, T.Y. Wu, G. Silver, A. McDonald, R. McCoy; Royal Adelaide Hospital (12) T.J. Kleinig, R. Scroop; Aus�n Hospital (10) H.M. Dewey, M. Simpson, M. Brooks, B. Coulton; Royal North Shore Hospital (6) M. Krause, T.J. Harrington, B. Steinfort, K. Faulder, M. Priglinger, S. Day; Monash Medical Centre (4) T. Phan, W. Chong, M. Holt, R.V. Chandra, H. Ma, D. Young; Western Hospital (4) T. Wijeratne, H. Tu, E. Mackay; Box Hill Hospital (3) C.F. Bladin, P.S. Loh, A. Gilligan, Z. Ross, S. Coote, T Frost; John Hunter Hospital (3) M.W. Parsons, F. Miteff, C.R. Levi, T. Ang, N. Spra�; Gold Coast University Hospital (3) M. Badve, H. Rice, L. de Villiers. New Zealand: Auckland City Hospital (4) P.A. Barber, B. McGuinness, A. Hope, M. Moriarty.
Neuroscience Trials Australia – E Cowley, R McCoy
CSIRO (eCRF) – S McBride, K Harrap, C Stanbridge
Stanford Stroke Centre (RAPID) – G Albers, R Bammer
Pa�ents and families
72
73
Methods
22 centres in Canada (11), US (6), Korea (3), UK (1), Ireland (1) Acute ischemic stroke pa�ents within a 12-‐hour window from onset, good func�onal status, with no age limit
tPA given when pa�ent eligible (no wai�ng for tPA response) Imaging must have shown: small core, proximal intracranial artery occlusion, moderate-‐good collaterals using CT, mCTA (use of MRI discouraged)
Intensive quality improvement program with personalized site visits
74
Inclusion and exclusion criteria
> 5 NIHSS < 12 hours from symptom onset Adult; No age limit Good pre-‐morbid status
CT head: ASPECTS > 5 (exclude large core) CTA: ICA + M1 or M1 or func�onal M1 (all M2s) CTA (preferably mul�phase): moderate to good collaterals
75
Workflow and Interven�on
Randomiza�on: standard of care vs. standard of care + endovascular treatment
Focus on speed – �me targets: -‐ CT head to groin puncture: 60 min* -‐ CT head to first recanaliza�on: 90 min*
Use of retrievable stents recommended
Use of balloon guide catheter recommended *Time from first slice NCCT Head chosen to encourage fast imaging acquisition and interpretation
76
Trial designed for speed
Deferral of consent where applicable Internet based, rapid randomiza�on process using randomized minimiza�on (MSB algorithm) Site signed agreements – document speed; commit to serial randomiza�on
Communica�on and feedback network to promote enrolment Quality improvement and follow-‐up program
77
Target Recruitment was 500 prior to the study being halted
316
Just below pace for 1.5 subjects/site/month
78
At 11 of 22 sites this was an approved mechanism of consent
17.8% of pa�ents were enrolled using this approach
The qualita�ve assessment from inves�gators was that this very much facilitated enrolment at those 11 sites and the lack of available surrogate decision maker was a major factor in slower enrolment at sites where this was not available.
Deferral of Consent
79
Quality Improvement
We reviewed each sites performance on treatment times and on imaging selection internally and then with each site by teleconference (times) and webinar (imaging) to promote achievement of time targets and to promote appropriate patient selection by imaging.
80
Low % Imaging Protocol Devia�ons
Sites interpreta�on: all appropriate Core Lab:
ASPECTS CTA occlusion Collaterals 3.6% 4.5% 6.5%
ASPECTS 5 n=5 M2-‐MCA n=9 Poor collaterals (<50% of the MCA territory)
ASPECTS 4 n=3
ASPECTS 3 n=2 ICA, no MCA
n=5
ASPECTS 3 n=1
81
Comparing Interval Times
ESCAPE MR CLEAN IMS III SYNTHESIS
Onset Randomiza�on
171 min (IQR 118-‐285)
204 min (IQR 152-‐251)
~135 min ~146 min
Onset IV tPA 114 min (IQR 82-‐160)
85 min (IQR 67-‐110)
~111 min 165 min (IQR 140-‐200)
Onset Groin Puncture
200 min (IQR 144-‐315)
260 min (IQR 210-‐313)
~196 min ~225 (?200) min**
Onset Reperfusion
241 min (IQR 176-‐359)
-‐-‐-‐ ~321 min -‐-‐-‐
NEJM. 2015; 372(1): 11-20.; NEJM 2013: 368(25): 2433-34; Circulation. 2014;130:265-272
**Onset-to-treatment (eg. infusion of intra-arterial tPA)
82
Comparing Interval Times
ESCAPE IMS III CT Groin Puncture “Picture to puncture”
51 min ~107 min.
CT Reperfusion “picture to perfusion”
84 min ~232 min
Circulation. 2014;130:265-272
Picture = first slice non-contrast CT ESCAPE patients were first encountered at similar times to prior trials but were
effectively treated much faster
83
Reperfusion
Interven�on Control
IV tPA No IV tPA IV tPA No IV tPA
TICI 2b/3 70.5% 77.3% -‐-‐-‐ -‐-‐-‐
mAOL 2-‐3 -‐-‐-‐ -‐-‐-‐ 37.3% 7.1%
mAOL assessed on CTA done at 2-8h post randomization
84
Comparing Reperfusion
ESCAPE MR CLEAN IMS-‐III
TICI 2b/3 72.6% 58.7% 45.2%
Higher rates of reperfusion with current technique and devices
Early assessment of the control group (2-8h in ESCAPE) shows only modest rates of reperfusion in the control group; 24h assessment time is too late
85
Results
A�er MR CLEAN results were presented at the WSC, the trial steering commi�ee suspended recruitment in the trial The planned interim analysis was conducted several weeks early
The trial was then halted at the recommenda�on of the DSMB because the efficacy boundary had been crossed
86
87
Baseline Characteris�cs Demographics Intervention (N=165) Control (N=150)
Age yr – median (IQR) 71 (60-81) 70 (60-81)
Female sex 52.1% 52.7%
Caucasian 87.3% 87.3%
Baseline NIHSS – median
(IQR)
16 (13-20) 17 (12-20)
Hypertension 63.4% 72.0%
Diabetes Mellitus 20.0% 26.0%
Atrial fibrillation 37.0% 40.0%
88
Baseline Imaging
Imaging characteristics – (%)
Intervention (N=165)
Control (N=150)
CT ASPECTS median (IQR) 9 (8-10) 9 (8-10)
CTA occlusion location –(%)
ICA ‘T’ or ‘L’ 27.6% 26.5%
M1-MCA or both/all M2-MCA 68.1% 71.4%
M2-MCA 3.7% 2.0%
Ipsilateral cervical carotid
occlusion
12.7% 12.7%
89
Process times min – median (IQR)
Intervention [N=165]
Control [N=150]
Symptom onset to
randomization (N=315)
169 (117-285) 172.5 (119-284)
Onset to IV alteplase (N=237) 110 (80-142) 125 (89-183)
CT to groin puncture 51 (39-68) ---
CT to first reperfusion 84 (65-115) ---
Onset to first reperfusion 241 (176-359) ---
Treatment with IV alteplase 72.7% 78.6%
Treatment Time Intervals
90
Safety Outcomes
Intervention [n=165]
Control [n=150]
RR (CI95) Adjusted§
RR (CI95)
Death [N=311] 10.4% 19.0% 0.5 (0.3-0.95)
0.5 (0.3-0.8)
Large MCA/malignant MCA
stroke
4.9% 10.7% 0.5 (0.2-1.0) 0.3 (0.1-0.7)
sICH (clinically determined at
site)
3.6% 2.7% 1.4 (0.4-4.7) 1.2 (0.3-4.6)
Access site hematoma 1.8% 0% --- ---
MCA perforation 0.6% 0% --- ---
91
Angiographic Outcomes
Intervention Control
Final Reperfusion TICI 2b/3 [Angio Core lab determined]
72.4% ---
mAOL 2-3 (at 2-8h CTA) [CT Core lab determined]
--- 31.2%
Retrievable Stent Use 86.1%
92
Outcomes (NNT = 4)
Clinical Intervention [n=165]
Control [n=150]
RR or cOR (CI95)
Adj RR or cOR (CI95)
mRS primary
outcome (“shift
analysis”) [n=311]
--- --- 2.6 (1.7-3.8) 3.1 (2.0-4.7)
mRS 0-2 at 90d
[n=311] 53.0% 29.3% 1.8 (1.4-2.4) 1.7 (1.3-2.2)
EQ-VAS at 90d
(median, iqr) 80 (30) 65 (30) P<0.001 (rank sum
test) 93
14.6 20.7 17.7 16.5 13.4 6.7 10.4
7.5 10.2 11.6 15.0 24.5 12.2 19.0
0 29.3 53 100percent
intervention (N=165)
control (N=150)
cOR adj = 3.1 (95%CI 2.0-4.7)Shift on 90-d mRS
Overall ESCAPE trial results
mRS 0 mRS 1 mRS 2 mRS 3mRS 4 mRS 5 Death
94
12.6 21.0 17.6 17.6 17.6 6.7 6.7
7.8 12.1 9.5 16.4 20.7 12.9 20.7
0 20 40 60 80 100percent
intervention (N=119)
control (N=116)
Shift on 90-d mRS
N=238, 3 lost to follow-up. P for interaction = 0.889
IV tPA group [N=235]
mRS 0 mRS 1 mRS 2 mRS 3mRS 4 mRS 5 Death
95
20.0 20.0 17.8 13.3 2.2 6.7 20.0
6.5 3.2 19.4 9.7 38.7 9.7 12.9
0 20 40 60 80 100percent
intervention (N=45)
control (N=31)
Shift on 90-d mRS
N=77, 1 lost to follow-up. P for interaction = 0.889
Non-IV tPA group [N=76]
mRS 0 mRS 1 mRS 2 mRS 3mRS 4 mRS 5 Death
96
0.0
00.2
50.5
00.7
51.0
0M
ort
alit
y
0 30 60 90Days from Stroke Onset
90-day K-M Curve
Kaplan Meier Curve of Mortality
97
98
10.9 15.2 10.9 15.2 17.4 10.9 19.6
2.6 10.3 5.1 10.3 12.8 15.4 43.6
16.1 22.9 20.3 16.9 11.9 5.1 6.8
9.3 10.2 13.9 16.7 28.7 11.1 10.2
0 20 40 60 80 100percent
> 80 yrs
<= 80 yrs
Intervention (n=46)
Control (n=39)
Intervention (n=118)
Control (n=108)
Shift on 90 day mRS
mRS 0 mRS 1 mRS 2 mRS 3mRS 4 mRS 5 Death
99
15.3 16.5 16.5 16.5 18.8 7.1 9.4
3.8 9.0 14.1 17.9 23.1 12.8 19.2
13.9 25.3 19.0 16.5 7.6 6.3 11.4
11.6 11.6 8.7 11.6 26.1 11.6 18.8
0 20 40 60 80 100percent
Female
Male
Intervention (n=85)
Control (n=78)
Intervention (n=79)
Control (n=69)
Shift on 90 day mRS
mRS 0 mRS 1 mRS 2 mRS 3mRS 4 mRS 5 Death
100
15.3 24.6 18.6 15.3 14.4 5.1 6.8
9.4 11.3 14.2 12.3 26.4 12.3 14.2
14.0 11.6 16.3 16.3 11.6 11.6 18.6
2.5 7.5 5.0 22.5 20.0 12.5 30.0
0 20 40 60 80 100percent
NIHSS<20
NIHSS>=20
Intervention (n=118)
Control (n=106)
Intervention (n=43)
Control (n=40)
Shift on 90 day mRS
mRS 0 mRS 1 mRS 2 mRS 3mRS 4 mRS 5 Death
101
Imaging-‐Defined Sub-‐groups All Benefit
102
16.4 22.7 16.4 15.6 14.1 6.2 8.6
9.0 10.7 11.5 14.8 24.6 13.1 16.4
10.3 13.8 17.2 20.7 13.8 10.3 13.8
0.0 8.0 12.0 16.0 24.0 8.0 32.0
0 20 40 60 80 100percent
ASPECTS>7
ASPECTS<=7
Intervention (n=128)
Control (n=122)
Intervention (n=25)
Control (n=29)
Shift on 90 day mRS
mRS 0 mRS 1 mRS 2 mRS 3mRS 4 mRS 5 Death
Supplementary Figure 7b: Figure shows benefit in the intervention group when compared to the control group across the 90-day mRS distribution in subjects with baseline non-contrast CT ASPECTS ≤ 7 vs. > 7. Nine subjects had CT ASPECTS 0-5 (protocol violators) and are included in the ASPECTS < 7 group for this analysis. Threshold for ASPECTS was pre-defined. There is no evidence of heterogeneity of treatment effect between these subgroups. (pinteraction=0.914, Wald test).
103
19.0 23.8 19.0 23.8 0.00.0 14.3
0.0 10.5 5.3 10.5 21.1 31.6 21.1
14.0 20.3 17.5 15.4 15.4 7.7 9.8
8.6 10.2 12.5 15.6 25.0 9.4 18.8
0 20 40 60 80 100percent
Yes
No
Intervention (n=21)
Control (n=19)
Intervention (n=143)
Control (n=128)
Shift on 90 day mRS
mRS 0 mRS 1 mRS 2 mRS 3mRS 4 mRS 5 Death
15.8%
61.8%
104
14.5 21.8 20.9 15.5 13.6 6.4 7.3
7.8 9.7 12.6 17.5 26.2 11.7 14.6
13.3 15.6 8.9 22.2 13.3 8.9 17.8
2.6 10.5 10.5 7.9 21.1 15.8 31.6
0 20 40 60 80 100percent
MCA
ICA
Intervention (n=110)
Control (n=103)
Intervention (n=45)
Control (n=38)
Shift on 90 day mRS
mRS 0 mRS 1 mRS 2 mRS 3mRS 4 mRS 5 Death
31.5%
60%
105
Effect size for Interven�on
common OR* (“shi�”) 3.1 (2.0-‐4.7)
mRS 0-‐2 29.3% 53.0% NNT = 4
Death HR* 19.0% 10.4% 0.4 (0.2-‐0.8)
*Adjusted for age, sex, baseline NIHSS score, baseline ASPECTS score, IV alteplase use, baseline occlusion loca�on
106
Conclusion
Endovascular thrombectomy is a safe, highly effec�ve procedure that saves lives and drama�cally reduces disability WHEN: – Pa�ents are carefully selected by imaging to iden�fy proximal occlusions, and exclude large core and exclude pa�ents with absent collaterals
– Treatment is extremely fast with target first slice imaging to groin puncture < 60 min and
imaging to reperfusion < 90 min
– Safe effec�ve technology (retrievable stents) is used 107
ESCAPE: Key Messages
Select pa�ents with imaging – measure the physiology -‐ www.aspectsinstroke.com – Good scan (exclude the large core pa�ents), proximal artery occlusion, moderate-‐good collaterals on mCTA
Act very fast on that informa�on – Picture-‐to-‐puncture (First slice CTgroin puncture) < 60 minutes
– Picture-‐to-‐perfusion (First slice CTreperfusion) < 90 minutes
Achieve reperfusion – TICI 2b/3 Work as a team! 108
109
Key Inclusion Criteria
Age 18 – 80
Pre-‐stroke Modified Rankin Score ≤ 1
NIHSS 8 – 29 at randomiza�on
Received IV t-‐PA within 4.5 hours of stroke onset
CTA or MRA confirma�on of large vessel occlusion (intracranial ICA, M1 or caro�d terminus)
Groin puncture within 6 hours of stroke onset and within 90 minutes* of qualifying imaging
*op�mal target: within 70 mins
110
Key Exclusion Criteria CT/MRI evidence of hemorrhage
CT hypodensity or MRI hyperintensity > 1/3 of the MCA territory (or in other territories, >100 cc of �ssue)
Caro�d dissec�on or complete cervical caro�d occlusion requiring sten�ng at the �me of the mechanical thrombectomy procedure ASPECTS < 6*
Protocol Rev F “Small to moderate core”
Protocol B-D “Target mismatch”
a) MRI- or CT-assessed core infarct lesion greater than 50 cc
b) Severe hypoperfusion lesion (10 sec or more Tmax lesion larger than 100 cc)
c) Ischemic penumbra < 15 cc and mismatch ratio ≤1.8
a) ASPECTS < 6
*Implemented after 1st 71 patients were enrolled
Imaging entry criteria revised to accommodate sites with limited perfusion imaging capability and ensure accelerated treatment delivery
111
Final Assessment Available (n=93) ¨ Attended 90d follow-up (n=78) ¨ Died prior to 90d (n=12) ¨ LOCF from 7-10d or 30d follow-up (n=3)
Final Assessment Unavailable (n=5) ¨ Investigator withdrew after entry criteria deviation (n=2) ¨ Subject withdrew consent after randomization(n=1) ¨ Subject withdrew consent after 27h visit (n=1) ¨ Subject withdrew consent after 7-10d & requested deletion of all data (n=1)
Allocated to Solitaire + IV t-PA (n=98)
Received Solitaire (n=87)
Randomiza�on & Follow-‐up
Intention to Treat (n=98) Intention to Treat with data (n=98) Modified Intention to Treat (n=98)
Final Assessment Available (n=98)
¨ Attended 90d visit (n=88) ¨ Died prior to 90d (n=9) ¨ LOCF from 7-10d or 30d follow-up (n=1)
Allocated to IV tPA only (n=98)
Intention to Treat (n=98) Intention to Treat with data (n=97) Modified Intention to Treat (n=93)
Randomized (n=196)
112
Solitaire + IV t-PA IV t-PA P
value Age – yr – mean (SD) N=98 65.0 (±12.5) N=95 66.3 (±11.3) 0.44
Male N=98 54 (55.1%) N=96 45 (46.9%) 0.31
Race N=90 N=92 0.94
White 79 (88.8%) 83 (90.2%)
Black 10 (11.2%) 8 (8.7%)
Asian 0 (0%) 1 (1.1%)
Other 1 (1.1%) 0 (0%)
Ethnicity, Hispanic N=90 8 (8.9%) N=92 7 (7.6%) 0.79
Prestroke mRS – median (IQR)
N=98 0 (0-0) N=94 0 (0-0) 0.76
NIHSS Score – median (IQR)
N=98 17.0 (13-20) N=94 17.0 (13-19) 0.76
Baseline Characteris�cs: Demographics and Severity
113
Solitaire + IV t-PA IV t-PA P value
Medical History N= 98 N= 97
Hypertension 66 (67.4%) 56 (57.7%) 0.18 Diabetes Mellitus 12 (12.2%) 15 (15.5%) 0.54 Hyperlipidemia 24 (24.5%) 22 (22.7%) 0.87 Current or Ex-Smoker 15 (15.3%) 14 (14.4%) 1.00 Atrial Fibrillation 35 (35.7%) 38 (39.2%) 0.66 Myocardial Infarction 8 (8.2%) 11 (11.3%) 0.48 Peripheral Arterial Disease 7 (7.1%) 5 (5.2%) 0.77 Neurological History N= 98 N= 97
Prior Ischemic Stroke 3 (3.1%) 1 (1.0%) 0.62 Hemorrhagic Stroke 0 (0%) 0 (0%) 1.00 Transient Ischemic Attack 3 (3.1%) 5 (5.2%) 0.50
Baseline Characteris�cs: Medical History
114
Solitaire + IV t-PA IV t-PA P
Systolic blood pressure (mmHg) – median (IQR)
N=98 150.0 (135 - 166) N=94 148.5 (135 -
165) 0.32
Serum glucose (mmol / liter) - mean (SD)
N=97 7.3 (±2.5) N=94 7.3 (±2.6) 0.93
ASPECTS – median (IQR) N=97 9.0 (7-10) N=96 9.0 (8-10) 0.68
Site of occlusion* N=93 N=94 0.57
Intracranial ICA 4 (4.3%) 4 (4.3%) Carotid terminus 13 (14.0%) 11 (11.7%) M1 MCA 63 (67.7%) 73 (77.7%) M2 MCA 13 (14.0%) 6 (6.4%)
Side of occlusion – Left N=96 39 (41.9%) N=97 48 (51.1%) 0.24
Penumbral Imaging performed N=98 85 (86.7%) N=97 76 (78.4%) 0.13
Target mismatch profile** 69 (70.4%) 64 (66.0%) 0.54 Malignant profile*** 13 (13.3%) 7 (7.2%) Time from onset to IV t-PA (mins) – median (IQR)
N=98 110.5 (85 - 156) N=94 117 (80 - 155)
Site of IV t-PA – outside hospital N=98 31 (31.6%) N=97 35 (37.2%) 0.45
Baseline Characteris�cs: Physiologic and Imaging
*Site of occlusion was assessed by the core lab **Target mismatch profile: MRI- or CT-assessed core infarct lesion ≤50 cc, Tmax>10s lesion ≤100cc, mismatch volume ≥15cc and mismatch ratio >1.8 ***Malignant profile: MRI or CT-assessed core infarct >50cc and Tmax>10s lesion more than 100cc
115
Core Lab Assessed Reperfusion Outcomes
TICI 2B/3 rate is 88.0% N=83 pts Based on all patients with final TICI data
1Reperfusion measured by reperfusion ratio assessed by core lab: reperfusion volume at 27hrs ÷ hypoperfusion lesion volume (Tmax >6s) at baseline
Solitaire + IV t-PA IV t-PA Odds Ratio P value Successful reperfusion1 (≥90% reperfusion) at 27hrs 53 (82.8%) 21 (40.4%) 7.11 [3.03,
16.70] <.0001
Modified TICI scale
TICI 2B: Perfusion of half or greater of the vascular distribution of the occluded artery
TICI 3: Full perfusion with filling of all distal branches
116
8.6
17.3
10.8
25.5
16.1
17.3
17.2
12.2
21.5
15.3
25.8
12.2
0 1 2 3 4 5 & 6
Primary Endpoint
IV t-PA
(n=93)
Solitaire + IV t-PA
(N=98)
Modified Rankin Scale Score p = 0.0002 (Cochran-Mantel-Haenszel p value)
Subjects (%)
117
Conclusions
We have a new standard of care tPA is s�ll the star�ng point for all eligible pa�ents <4.5hr
Stent-‐thrombectomy works <6hr if large vessel occlusion regardless of age or severity CT angiography needs to be standard (minimum – ideally also CT perfusion)
Need to work out how to implement imaging and transfer outside metropolitan regions
118