reperfusion after 4·5 hours reduces infarct growth and improves clinical outcomes

Reperfusion after 4·5 hours reduces infarct growth and improves clinical outcomes

Miguel R. Picanço1, Søren Christensen2, Bruce C. V. Campbell1, Leonid Churilov3,Mark W. Parsons4, Patricia M. Desmond2, P. Alan Barber5, Christopher R. Levi4,Christopher F. Bladin6, Geoffrey A. Donnan3, and Stephen M. Davis1*; for theEPITHET Investigators

Background The currently proven time window for throm-bolysis in ischemic stroke is 4·5 h. Beyond this, the risks andbenefits of thrombolysis are uncertain.Aims To determine whether thrombolysis and reperfusionwere beneficial after 4·5 h, we examined clinical and radio-logical outcomes in patients treated with tissue plasminogenactivator or placebo within 4·5–6 h, using data from the Echo-planar Imaging Thrombolytic Evaluation Trial.Methods In the Echoplanar Imaging Thrombolytic EvaluationTrial, ischemic stroke patients presenting three to six-hoursafter stroke onset were randomized to tissue plasminogenactivator or placebo, without knowledge of magnetic reso-nance imaging results. This analysis was restricted to patientstreated between 4·5 and 6 h. The effect of tissue plasminogenactivator and reperfusion on infarct growth between baselinediffusion-weighted imaging and day 90 T2 imaging wasassessed, along with good neurological outcome (≥8 pointreduction or reaching 0–1 at 90 days on National Institutes ofHealth Stroke Scale) and functional outcome (modified Rankinscale). The effect of tissue plasminogen activator on reperfu-sion was also analyzed.Results Sixty-nine patients were treated 4·5–6 h after onset,and infarct growth was assessed in 63. Tissue plasminogen

activator was associated with lower relative growth (94% vs.168%, P = 0·03) and a trend to lower absolute growth(−0·17 ml versus 9·6 ml, P = 0·07). Reperfusion was increased inthe tissue plasminogen activator group (58% versus 25%,P = 0·03) and was associated with increased rates of goodneurological (86% versus 28% P < 0·001) and functional (modi-fied Rankin scale 0–2 73% versus 34%, P = 0·01) outcomes.Reperfusion was strongly associated with lower relative (80%versus 189%, P < 0·001) and absolute (−2·5 ml versus 40 ml,P < 0·001) infarct growth.Conclusions Thrombolysis 4·5–6 h after stroke onset reducedinfarct growth and increased the rate of reperfusion, which wasassociated with good neurological and functional outcome.Key words: MRI, stroke, thrombolysis, tPA

Introduction

Intravenous tissue plasminogen activator (tPA) has been shown

to reduce disability when administered within 4·5 h of stroke

onset (1,2). However, pooled analysis of the major tPA trials (3)

was unable to show significant benefit of tPA beyond 4·5 h, and

there was a trend toward increased risk.

The International Stroke Trial 3 (4) further investigated the role

of tPA up to six-hours after stroke onset. The prespecified primary

end-point was neutral. Although secondary analysis of ordinal

shift in modified Rankin scale (mRS) demonstrated a modest,

statistically significant benefit, this was strongly driven by patients

treated within three-hours. More recently, the Safe Implementa-

tion of Treatment in Stroke (SITS) International Thrombolysis

Registry (5) update showed no difference in functional outcome

or intracranial hemorrhage rates when patients were treated with

tPA 4·5 to 6 h after onset compared with patients in the 3–4·5 h

and 0–3 h windows. Our aim was to assess the impact of tPA and

reperfusion on clinical and radiological outcomes exclusively in

the 4·5–6 h window. We analyzed the Echoplanar Imaging

Thrombolytic Evaluation Trial (EPITHET) randomized trial data

to assess the effect of treatment and reperfusion on clinical

outcome and attenuation of infarct growth in patients treated

4·5–6 h after stroke onset.

Methods

Echoplanar imaging thrombolytic evaluation trial was a phase II

prospective, randomized, double-blind trial of tPA versus placebo

in acute stroke patients administered three to six-hours after

stroke onset. The methodology has previously been described

(6). In brief, patients with acute hemispheric ischemic stroke

three to six-hours after symptom onset, aged >18 with National

Correspondence: Stephen M. Davis*, Department of Neurology, RoyalMelbourne Hospital, Grattan Street, Parkville 3050, Australia.E-mail: [email protected] of Medicine and Neurology, Melbourne Brain Centre atthe Royal Melbourne Hospital, University of Melbourne, Melbourne,Australia2Department of Radiology, University of Melbourne, Melbourne,Australia3Florey Institute of Neuroscience of Mental Health, University ofMelbourne, Melbourne, Australia4Priority Research Centre for Brain and Mental Health Research, JohnHunter Hospital, University of Newcastle, Newcastle, Australia5Department of Neurology, University of Auckland, Auckland, NewZealand6Box Hill Hospital, Monash University, Melbourne, Australia

Received: 09 September 2013; Accepted: 09 September 2013; Publishedonline 21 November 2013

Subjects codes: [44] Acute Cerebral Infarction, [58] Computerizedtomography and Magnetic Resonance Imaging, [73] Thrombolysis.

Conflict of interest:

M. R. P., S. C., B. C. V. C., and L. C. have no disclosures.

C. R. L., M. W. P., and C. F. B. have accepted honoraria or consultancy andtravel grants from Boehinger-Ingelheim and Sanofi-Aventis.

G. A. D. is a member of Boehinger-Ingelheim, PAION, Servier, and Sanofi-Aventis advisory boards.

S. M. D. is a member of Servier, Novo Nordisk, and Sanofi-Aventis advi-sory boards.

DOI: 10.1111/ijs.12209

Research

© 2013 The Authors.International Journal of Stroke © 2013 World Stroke Organization

266 Vol 9, April 2014, 266–269

Institutes of Health Stroke Scale (NIHSS) score ≥4, and premor-

bid mRS ≤2 were eligible. Magnetic resonance imaging (MRI)

was performed after informed consent and before treatment, but

was not used to select patients. Diffusion and perfusion MRI was

obtained at baseline and day 3–5. At day 90, T2-weighted images

were obtained to measure final infarct volume. Infarct volumes

were manually outlined by two stroke neurologists, and the

volumes were averaged. For patients without day 90 imaging, the

last observation was carried forward with the day 3–5 infarct

volume being substituted as per EPITHET methodology. A sec-

ondary analysis was performed using infarct growth from base-

line to day 3–5 (previously shown to strongly correlate with day

90 volume) (7).

Reperfusion was defined as >90% reduction in Tmax ≥ 2s

lesion volume between baseline and day 3–5. Mismatch was

defined as MR perfusion/diffusion-weighted imaging (MRP/

DWI) ratio >1·2 and absolute difference MRP–DWI >10 ml as in

EPITHET but modified to use a Tmax > 6s perfusion threshold, as

the original EPITHET Tmax ≥ 2s definition included significant

benign oligemia (8–10). Good functional outcome was defined as

day 90 mRS 0–2, excellent functional outcome as mRS 0–1, and

good neurological outcome by a reduction in NIHSS ≥8 points or

reaching 0–1 at day 90. Both National Institute of Neurological

Disorders and Stroke (NINDS) (1) and SITS-Monitoring Study

(SITS-MOST) (11) definitions of symptomatic intracerebral

hemorrhage (sICH) were applied.

Statistical analysis was performed using Stata (v.12, Stata-

Corp, College Station, TX, USA). Infarct growth was assessed

using nonparametric testing (Mann–Whitney). All categorical

variables were analyzed using Fisher’s exact test. Median regres-

sion was used to assess the interaction between treatment and

time to treatment on infarct growth in the entire three to six-

hours EPITHET cohort.

Results

Sixty-nine of 100 randomized patients in EPITHET received

treatment after 4·5 h (Fig. 1). Except for increased hypertension

in the tPA group, baseline characteristics were similar between

placebo and tPA groups (Table 1).

Intravenous tPA was associated with significantly less relative

growth and trends toward reduced absolute growth and improved

clinical outcomes compared with placebo (Table 2). Tissue plas-

minogen activator was associated with significantly increased rep-

erfusion compared with placebo (58% versus 25%, P = 0·03).

Reperfusion was strongly associated with reduced infarct growth,

improved neurological outcome, and more frequent good and

excellent functional outcomes (Table 2).

In the subgroup with mismatch (n = 45), tPA was associated

with significantly less relative growth (median 96% versus 203%,

P = 0·02) and a trend toward reduced absolute (median –0·25 ml

versus 31·0 ml, P = 0·06). Reperfusion was strongly associated

with reduced absolute (median −3·5 ml versus 44·3 ml, P < 0·001)

and relative (median 76% versus 217%, P < 0·001) growth.

Repeating the analysis using infarct growth between baseline

and days 3–5, the strong effect of reperfusion on infarct growth

was maintained (median relative growth, 145% versus 267%,

P < 0·001; median absolute growth, 6·1 ml versus 61 ml,

P < 0·001). However, tPA treatment was no longer significant

(median relative growth, 160% versus 205%, P = 0·08; median

absolute growth, 7·0 ml versus 19 ml, P = 0·27).

In the full three- to six-hour trial population, there was no

significant interaction between time to treatment, and either tPA

or reperfusion with absolute (tPA P = 0·89, reperfusion P = 0·80)

or relative (tPA P = 0·31, reperfusion P = 0·21) infarct growth

using median regression analysis. The incidence of sICH and

mortality did not differ significantly between the 3–4·5 h and

Fig. 1 Analysis flow chart. EPITHET, echoplanar imaging thrombolytic evaluation trial; LOCF, last observation carried forward; tPA, tissue plasminogenactivator.

ResearchM. R. Picanço et al.


Vol 9, April 2014, 266–269 267

4·5–6 h groups (SITS-MOST criteria 14·3% versus 5·3%, P = 0·29;

NINDS criteria 14·3% versus 7·9%, P = 0·60; mortality 36%

versus 21%, P = 0·30).

Discussion

This study has demonstrated that reperfusion 4·5–6 h after stroke

onset is associated with reduced infarct growth and favorable

clinical outcomes. Tissue plasminogen activator treatment was

associated with significantly increased reperfusion and reduced

relative infarct growth. Although tPA did not reach significance

for absolute growth or clinical outcomes in this sample of 63

patients, there were strong trends.

The main limitation of this study was the small sample size. In

contrast to the primary EPITHET analysis that examined mean

geometric growth restricted to mismatch patients, this sub-study

analyzed all patients treated 4·5–6 h and used nonparametric sta-

tistics to assess growth due to skew distribution. Loss to imaging

follow-up may create bias because of imbalances between tPA and

placebo groups, and substitution of days 3–5 volumes when day

90 was unavailable may have introduced heterogeneity because of

infarct edema at days 3–5 versus infarct atrophy at day 90.

However, the secondary analysis of infarct growth to days 3–5

gave similar results.

Despite the extension of the time window for tPA from 3 to

4·5 h after stroke onset, many patients worldwide still arrive in

hospital too late for tPA. Ongoing trials including DIAS-3/4

(ClinicalTrials.gov NCT00790920 and NCT00856661) and

EXTEND (12) continue to investigate delayed therapy. Our results

are encouraging as, even in this relatively small sample, tPA had

biological benefits with increased reperfusion and reduced infarct

growth. The unambiguous response to reperfusion in both

imaging and clinical outcomes is consistent with significant

amounts of penumbral tissue being present in the 4·5–6 h

window.

References1 The National Institute of Neurological Disorders and Stroke rt-PA

Stroke Study Group. Tissue plasminogen activator for acute ischemicstroke. N Engl J Med 1995; 333:1581–7.

2 Hacke W, Kaste M, Bluhmki E et al. Thrombolysis with alteplase 3 to4.5 hours after acute ischemic stroke. N Engl J Med 2008; 359:1317–29.

3 Lees KR, Bluhmki E, von Kummer R et al. Time to treatment withintravenous alteplase and outcome in stroke: an updated pooledanalysis of ECASS, ATLANTIS, NINDS, and EPITHET trials. Lancet2010; 375:1695–703.

4 Sandercock P, Wardlaw JM, Lindley RI et al. The benefits and harms ofintravenous thrombolysis with recombinant tissue plasminogen acti-vator within 6 h of acute ischaemic stroke (the third internationalstroke trial [IST-3]): a randomised controlled trial. Lancet 2012;379:2352–63.

Table 1 Baseline characteristics for patients treated 4·5–6 h after stroke onset

NumbertPA Placebo

PReperfusion No reperfusion

P38 31 22 32

Age, year, median (IQR) 75 (65–80) 73 (59–78) 0·29 77 (73–81) 71 (60–78) 0·05Male 23 (60%) 19 (69%) 0·99 10 (45%) 22 (69%) 0·10Hypertension 33 (87%) 19 (61%) 0·02 14 (63%) 25 (78%) 0·35Diabetes mellitus 7 (18%) 8 (26%) 0·52 5 (23%) 8 (25%) 0·99Hyperlipidemia 18 (47%) 15 (48%) 0·99 11 (50%) 15 (47%) 0·99Atrial fibrillation 18 (47%) 11 (35%) 0·34 7 (32%) 15 (47%) 0·40Smoking 16 (42%) 16 (52%) 0·47 6 (27%) 18 (56%) 0·52Baseline NIHSS, median (IQR) 12 (7–18) 11 (8–17) 0·07 13 (8–17) 11 (8–17) 0·99Time to treatment, min, median (IQR) 310 (300–341) 330 (310–350) 0·17 314 (297–330) 327 (307–353) 0·12Baseline DWI volume, ml, median (IQR) 17·3 (8·9–51·1) 20·4 (8·1–33·6) 0·87 15·6 (8·5–29·2) 21·6 (9·1–56·3) 0·24Baseline MRP volume, ml, median (IQR) 65·6 (40·3–146·3) 94·4 (52·1–149·3) 0·33 67·2 (37·0–104·1) 90·1 (47·3–170·6) 0·28Perfusion-diffusion mismatch 28 (76%) 21 (68%) 0·59 18 (82%) 25 (78%) 0·99

DWI, diffusion-weighted imaging; IQR, interquartile range; NIHSS, National Institutes of Health stroke scale; MRP, MR perfusion imaging; tPA, tissueplasminogen activator.

Table 2 Infarct growth and clinical outcomes

Received tPA Achieved reperfusion

Yes No P Yes No P

Infarct growth n = 33 n = 30 n = 22 n = 32Median relative growth (%) 94 168 0·03 80 189 <0·001Median absolute growth (ml) −0·17 9·6 0·07 –2·5 40 <0·001Clinical outcomes n = 37 n = 31 n = 22 n = 32Good neurological outcome 40·5% 29·0% 0·45 86·4% 28·1% <0·001mRS 0–2 48·6% 45·2% 0·81 72·7% 34·4% 0·01mRS 0–1 40·5% 29·0% 0·45 54·5% 21·9% 0·02

mRS, modified Rankin scale; tPA, tissue plasminogen activator.

Research M. R. Picanço et al.


268 Vol 9, April 2014, 266–269

5 Ahmed N, Kellert L, Lees KR, Mikulik R, Tatlisumak T, Toni D. Resultsof intravenous thrombolysis within 4.5 to 6 hours and updated resultswithin 3 to 4.5 hours of onset of acute ischemic stroke recorded in thesafe implementation of treatment in stroke international strokethrombolysis register (SITS-ISTR): an observational study. JAMANeurology 2013; 70:837–44.

6 Davis SM, Donnan GA, Parsons MW et al. Effects of alteplase beyond3 h after stroke in the echoplanar imaging thrombolytic evaluationtrial (EPITHET): a placebo-controlled randomised trial. Lancet Neurol2008; 7:299–309.

7 Ebinger M, Christensen S, De Silva DA et al. Expediting MRI-basedproof-of-concept stroke trials using an earlier imaging end point.Stroke 2009; 40:1353–8.

8 Christensen S, Campbell BCV, Perez de la Ossa N et al. Optimal per-fusion thresholds for prediction of tissue destined for infarction in thecombined EPITHET and DEFUSE dataset. Stroke 2010; 41:e297.

9 Olivot JM, Mlynash M, Thijs VN et al. Optimal Tmax threshold forpredicting penumbral tissue in acute stroke. Stroke 2009; 40:469–75.

10 Zaro-Weber O, Moeller-Hartmann W, Heiss WD, Sobesky J. Maps oftime to maximum and time to peak for mismatch definition in clinicalstroke studies validated with positron emission tomography. Stroke2010; 41:2817–21.

11 Wahlgren N, Ahmed N, Davalos A et al. Thrombolysis with alteplasefor acute ischaemic stroke in the Safe Implementation of Thromboly-sis in Stroke-Monitoring Study (SITS-MOST): an observational study.Lancet 2007; 369:275–82.

12 Ma H, Parsons MW, Christensen S et al. A multicentre, randomized,double blinded, placebo controlled phase 3 study to investigateEXtending the time for Thrombolysis in Emergency NeurologicalDeficits (EXTEND). Int J Stroke 2012; 7:74–80.

ResearchM. R. Picanço et al.


Vol 9, April 2014, 266–269 269

reperfusion after 4·5 hours reduces infarct growth and improves clinical outcomes

Documents