Meta-analysis Confirms Late Thrombolysis Benefits in Stroke

May 22, 2019

Further results have confirmed the benefit of thrombolysis up to 9 hours after onset of ischemic stroke, or in cases of wake-up stroke, for selected patients with potentially salvageable brain tissue, as identified by mismatch on CT perfusion imaging.

The recently published EXTEND trial in this population showed a positive primary endpoint (mRS 0–1 at 3 months), but the ordinal shift analysis, a secondary outcome, was not significant.

The current report is a meta-analysis of the EXTEND trial with two previous trials (ECASS4-EXTEND and EPITHET) that evaluated a similar strategy but showed nonsignificant results. The meta-analysis showed significant positive results for all outcomes evaluated — excellent outcome (mRS 0–1), good outcome (mRS 0–2), and mRS shift analysis.

The meta-analysis was presented today at the 5th European Stroke Organisation Conference (ESOC) 2019 by Henry Ma, MD, Monash University, Melbourne, Australia, and was simultaneously published online in the Lancet.

"These data reinforce that thrombolysis in an extended time window out to 9 hours with perfusion imaging selection is effective," Ma commented to Medscape Medical News. "We already had quite good data on this from EXTEND, but the additional patients from the meta-analysis now give enough power to show convincing results on all outcomes."

The first author of the Lancet article, Bruce Campbell, MD, University of Melbourne, Australia, added: "The ECASS4 and EPITHET trials enrolled a broader population of patients, but this meta-analysis allowed us to focus on patients with the designated mismatch on imaging and shows us that this is really where the benefit is. In these patients presenting up to 9 hours after stroke onset, thrombolysis more than doubled the odds of excellent outcome — the ability to return to all normal activities (mRS 0–1)."

Campbell also noted that the latest data from the meta-analysis allow more precise understanding of the risk with tissue plasminogen activator (tPA). "Symptomatic hemorrhage occurred in 4.6% of patients that fit the mismatch criteria and fatal hemorrhage in 2%. This is very consistent with other studies of tPA, so give us reassuring safety data out to this extended time window," he commented.

In the Lancet article, the authors conclude: "The risk of symptomatic intracerebral haemorrhage [ICH] was consistent with that observed in previous trials of stroke thrombolysis with alteplase in the 0–4.5 h treatment window and did not offset the overall benefits in functional outcome."

Campbell reported that most hospitals have the capability to perform CT perfusion imaging. "The mismatch calculation is fully automated, so if hospitals have the software, it is very easy. If you want to treat patients beyond 4.5 hours with thrombolysis and beyond 6 hours with thrombectomy, then CT perfusion is the most practical way of identifying which patients are most likely to benefit."

Ma estimated that expanding thrombolysis to this extended time window in patients meeting the imaging criteria could increase the number of patients who receive tPA by about 50%. "At present, about 10% to 15% of ischemic stroke patients get thrombolysis. We think that could go up to about 20% with these new data," he added.

Campbell noted that for approximately 1 in 10 patients in the extended time window, the perfusion imaging profile is suitable, making them eligible for late thrombolysis. "The EXTEND trial recruited quite slowly, but since we've had the results and we've started treating patients, we've found more patients than we've expected with the suitable imaging results," he said.

Ma reported that 60% of patients in this analysis had a large vessel occlusion so would also be eligible for thrombectomy, which has been shown to be effective out to 24 hours for patients with mismatch on imaging. "But 40% of population did not have a large vessel occlusion, so now we have something to offer some of these patients presenting late."

The meta-analysis involved a total of 414 patients, of whom 213 (51%) were assigned to receive the thrombolytic alteplase (Activase, Genentech), and 201 (49%) were assigned to receive placebo.

Results showed significant improvements in all three efficacy outcomes at 3 months with thrombolysis in the overall population.

Table 1. Outcomes at 3 Months in All Patients

Outcome Placebo (n = 201) Alteplase (n = 213) Odds Ratio (95% Confidence Interval) P Value
mRS 0–1 (Primary Outcome, %) 29 36 1.86 (1.15 – 2.99) .01
Improvement in mRS (Shift) NA NA 1.60 (1.12 – 2.27) .009
mRS 0–2 (%) 44 49 1.74 (1.08 – 2.81) .02
Death at 3 months (%) 9 14 1.55 (0.81 – 2.97) .19
Symptomatic ICH (%) <1 5 9.70 (1.23 – 76.55) .03


The benefits appeared accentuated in patients who had confirmed automated perfusion mismatch.

Table 2. Outcomes in Patients With Confirmed Mismatch

Outcome Placebo (n = 152) Alteplase (n = 152) Odds Ratio (95% Confidence Interval) P Value
mRS 0–1 (Primary Outcome, %) 26 36 2.06 (1.17 – 3.62) 0.012
Improvement in mRS (shift)     1.68 (1.11 – 2.53) 0.014
mRS 0–2 (%) 40 51 2.22 (1.25 – 3.94) 0.006
Death at 3 months (%) 11 13 1.28 (0.60 – 2.73) 0.52
Symptomatic ICH (%) 1 5 7.29 (0.88 – 60.18) 0.07

Is Automated Imaging Necessary?

In a comment that accompanied the Lancet article, Shelagh Coutts, MD, and Bijoy Menon, MD, University of Calgary, Canada, conclude that "the results of this meta-analysis by Campbell and colleagues provide justification for late time window or wakeup thrombolysis in appropriately selected patients."

However, the editorialists question whether perfusion imaging is necessary to identify eligible patients.

"Clear evidence for the use of automated perfusion to identify patients eligible for late window thrombolysis is scarce and the question of which imaging modality to use to select such patients remains unanswered," they say.

They note that they mostly use a combination of noncontrast CT brain, CT angiography of the head and neck and multiphase CT angiography to identify patients who might benefit from late thrombolysis. The noncontrast CT can identify patients for whom there is a discrepancy between the extent of the established ischemia and the clinical examination, but this type of clinical decision making is subjective and thus requires experience, they explain.

"Whether CT perfusion has any further benefit beyond this approach is unclear.... Rather than providing a general solution for treatment, automation is likely to be helpful for less experienced centres, but perhaps not necessary in highly experienced centres," they suggest.

EXTEND was funded by the Australian National Health and Medical Research Council. ECASS4-EXTEND was an investigator-initiated trial supported by a restricted grant from Boehringer Ingelheim. EPITHET was funded by the Australian National Health and Medical Research Council. For all three trials, alteplase was supplied by Boehringer Ingelheim. RAPID imaging software was provided by iSchemaView. Ma, Campbell, Coutts, and Menon have disclosed no relevant financial relationships.

Lancet. Published online May 22, 2019. Abstract, Comment

5th European Stroke Organisation Conference (ESOC) 2019. Presented May 22, 2019.

Follow Medscape on Facebook, Twitter, Instagram, and YouTube


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.