Extending the tPA Window to 4.5 – 9 Hours in Acute Ischemic Stroke (AIS)?

Background: No matter which side of the debate you sit on in regard to systemic thrombolysis in acute ischemic stroke (AIS), there is one truth: systems have undergone major changes to ensure tPA is offered to patients in the ≤4.5-hour window.  The debate surrounding tPA in AIS lies in the equipoise surrounding benefits while there are very real harms.  Advocates of tPA in AIS hang on to two trials that have never been replicated (i.e. NINDS and ECASS-III), and both have major methodological issues. Skeptics of tPA in AIS appropriately argue that there are 11 other randomized clinical trials which have shown almost no benefit, but come at the cost of early increased early mortality and symptomatic intracranial hemorrhage (sICH) (Nice breakdown of individual trials of thrombolysis in stroke can be found at First10EM).  Now there is a push to extend the window of tPA out to 9 hours in AIS with newer imaging modalities such as MRI diffusion-weighted studies in patients with unknown onset of symptoms. The push for this stems from the fact that patients with a visible ischemic lesion on diffusion-weighted imaging, combined with the absence of a clearly visible hyperintense signal in the same region on fluid-attenuated inversion recovery (FLAIR) is predictive of symptom onset within 4.5 hours before imaging.

Visible Ischemic Lesion on Diffusion-Weighted Imaging, Combined with Absence of a Clearly Visible Hyperintense signal in the Same Region on FLAIR?

  • To be fair, I am not a neuroradiologist and had to look this up to understand what this actually meant in plain English.
  • DWI shows high signal intensity within the first few minutes after an ischemic stroke, allowing for early diagnosis (i.e. Minutes after a stroke, cytotoxic edema can develop causing a net decrease in water diffusion, which is visualized by an increased signal on DWI).
  • FLAIR signal changes tend to be delayed (i.e. Hours after stroke, vasogenic edema gradually develops causing a visible hyperintensity on FLAIR).
  • Therefore, the DWI-FLAIR mismatch of an ischemic stroke that occurred in ≤4.5 hrs would show a positive increased signal on DWI and a negative, non-hyperintense FLAIR signal
  • Plain English Translation: DWI says yes there is an ischemic stroke, FLAIR says if the stroke is new or old (i.e. ischemic vs infarcted, respectively)


 What They Did:

  • Multicenter, randomized, double-blind, placebo-controlled clinical trial of adult patients with unknown time of onset of stroke (with signs of the stroke onset < 4.5 hours based on advanced MR imaging) to receive either IV tPA or placebo


  • Primary: favorable outcome (defined as mRs 0 – 1) at 90d
  • Secondary:
    • Shift analysis of whether alteplase would lead to lower ordinal scores on mRs at 90d
    • Proportion of patients with a treatment response at 90d (defined as a mRs = 0 if NIHSS score ≤7, a score of 0 or 1 for patients with NIHSS score of 8 – 14, and a score of 0 – 2 for patients with NIHSS score ≥14)
    • Global outcome score at 90d (defined as a good outcome on 4 scales: mRs 0 – 1, Barthel index 95 – 100, GCS = 5, 90d Beck Depression Inventory, 90d score on EQ-5D, or VAS)
  • Safety:
    • Death or dependence (mRs 4 – 6) at 90d
  • Secondary Safety:
    • Symptomatic ICH causing deterioration in neurologic symptoms
    • Incidence of parenchymal hematoma on MRI 22 – 36hrs after randomization


  • Ischemic lesion visible on MRI diffusion-weighted imaging but no parenchymal hyperintensity on fluid-attenuated inversion recovery (FLAIR)
  • Measurable disabling neurological deficit (defined as an impairment of one or more of the following: language, motor function, cognition, gaze, vision, neglect)
  • Age 18 – 80 years
  • Written informed consent from patient or proxy


  • Planned endovascular thrombectomy
  • Pre-stroke disability (inability to carry out all daily activities, requiring some help or supervision)
  • Severe stroke by clinical assessment (NIHSS >25)
  • Hypersensitivity to alteplase
  • Pregnancy or lactating
  • Known hemorrhagic diathesis
  • Known history of or suspected intracranial hemorrhage
  • Suspected SAH (even if CT is normal) or condition after SAH from aneurysm
  • History of CNS damage (i.e. neoplasm, aneurysm, intracranial or spinal surgery)
  • Current use of anticoagulants
  • Platelet count <100k
  • Blood glucose <50 or >400mg/dL
  • Severe uncontrolled HTN (SBP >185mmHg or DBP > 100mmHg) or requiring aggressive medication to maintain blood pressure within these limits)
  • Manifestations of severe liver disease including hepatic failure, cirrhosis, portal hypertension, and active hepatitis
  • Major surgery or significant trauma in past 3 months
  • Life expectancy ≤6 months
  • Any contraindication to MRI


  • 503 patients enrolled (Trial stopped early due to lack of funding but anticipated 800 patients)
    • tPA = 254 patients
    • Placebo = 249 patients
    • Majority of patients with an unknown time of stroke symptoms was due to was wake up stroke (89% of patients)
    • Median time to medication administration was 3.1hrs in tPA arm and 3.2hrs in placebo arm
  • Favorable Outcome at 90d
    • tPA: 53.3%
    • Placebo: 41.8%
    • aOR 1.61
    • 95% CI 1.09 – 2.36
    • P = 0.02
  • Median mRs at 90d
    • tPA = 1
    • Placebo = 2
    • aOR 1.62
    • 95% CI 1.17 – 2.23
    • P = 0.003
  • Deaths:
    • tPA = 10 (4.1%)
    • Placebo = 3 (1.2%)
    • OR 3.38
    • 95% CI 0.92 – 12.52
    • P = 0.07
  • ICH:
    • tPA = 4.0%
    • Placebo 0.4%
    • OR 10.46
    • 95% CI 1.32 – 82.77
    • P = 0.03


  • Multicenter, randomized, double-blind, placebo-controlled trial
  • First study of its kind
  • MRI imaging results and symptomatic ICH results were adjudicated in a blinded manner
  • Further advancement of understanding of ischemic stroke in terms of time vs perfusion


  • Trial was stopped early owing to cessation of funding after the enrollment of 503 of the anticipated 800 patients. Trial required 370 pts per group to provide a power of 80% to show a 10% between-group difference, which was not met.
    • The confidence interval in this study already approaches 1, so completion of enrollment may have resulted in the CI crossing. 1.
    • Smaller trials stopped early have been shown to grossly overestimate the benefitof a drug/intervention.
  • More patients in the tPA arm had intracranial occlusion of the internal carotid artery, which typically undergoes endovascular therapy (9.6% vs 4.5%)
  • Small number of patients did not allow for subgroup analysis
  • All investigators involved in image reading had to pass a software-based image training and certification before they were allowed to enroll patients. MRI or neuro-radiologists to read the MRI may not be available at many institutions


  • One aspect of this study I find very interesting was that of the 1362 patients screened, 455 were excluded due to a marked FLAIR lesion (no DWI-FLAIR mismatch) and 137 patients were excluded due to not having DWI lesions. I don’t have MRI readily available at each of the institutions I work at but may be feasible in stroke centers to exclude stroke mimics.
  • Over 5 years 1362 stroke patients had to be screened to find 500 patients that met indications for randomization. This equates to about 9 patients per month. Sounds very similar to endovascular therapy for stroke, as in this is a very select group of patients and not all stroke comers.
  • The authors even admit, “The trial was stopped early because of a discontinuation of funding and enrolled fewer patients than planned. This factor limits the interpretation of the safety results because the observed trend toward a higher rate of death in the alteplase group may have become significant with a larger sample size.”
  • In this trial, the authors used a dichotomous outcome of good (mRs 0 – 1) vs not good (mRs 2 – 6) neurologic outcomes. It’s interesting that a good neurologic outcome was defined as a modified rankin score of 0 -1 at 90 days and not 0 – 2.  Most of the stroke literature uses a score of 0 – 2.
    • mRs 0 = No symptoms
    • mRs 1 = No significant disability despite symptoms; able to perform all previous activities and activities of daily living
    • mRs 2 = Slight disability; unable to perform all previous activities but able to look after own affairs without assistance

Author Conclusion: “In patients with acute stroke with an unknown time of onset, intravenous alteplase guided by a mismatch between diffusion-weighted imaging and FLAIR in the region of ischemia resulted in a significantly better functional outcome and numerically more intracranial hemorrhages than placebo at 90 days.” 

Clinical Take Home Point: In this multicenter randomized trial of patients with strokes of unknown time onset, the use of tPA resulted in a 11% improvement in neurologic outcomes at 90days, and although not statistically significant also resulted in more death  and more ICH.


What They Did:

  • Unfortunately, at this time we only have the abstract which was presented at the international stroke conference 2019, and not full publication, but is worth covering in this post as it is similar to the WAKE UP trial
  • This was a multicenter randomized double-blind, placebo-controlled trial of alteplase in AIS in patients presenting within 4.5 – 9 hours from onset or those with wake-up strokes


  • Primary: mRs 0 – 1 at 3mo
  • Secondary:
    • Independent functional status (mRs 0 – 2)
    • Early reperfusion
    • Clinical improvement with NIHSS reduction of 8 points or reaching 0 – 1 at 2hrs
    • Death
    • Symptomatic ICH


  • Acute hemispheric ischemic stroke
  • Informed consent
  • Age ≥18 years
  • Treatment onset to commence after 3hrs and up to and including 9 hours after stroke onset
  • Wake up stroke (defined as having no symptoms at sleep onset, but stroke symptoms on waking)
  • NIHSS ≥4 – 26 with clinical signs of hemispheric infarction
  • Automated perfusion imaging software showing salvageable brain tissue
    • Penumbral mismatch: a penumbra to core lesion volume ratio of greater than 1:2 and an absolute difference greater than 10mL (Using a MR or CT Tmax >6s delay perfusion lesion and MR-DWI or CT-CBF core lesion) AND an infarct core lesion volume of ≤70mL using MR-DWI or CT-CBF 


  • ICH
  • Rapidly improving symptoms
  • Prestroke mRs ≥2
  • Contraindication to contrast
  • Infarct core > 1/3 MCA territory qualitatively on CT
  • Terminal illness with expeted survival ≤1 year
  • Pregnancy
  • Stroke within previous 3 months
  • Current use of oral anticoagulation
  • Use of glycoprotein IIb-IIIa inhibitors within past 72hr
  • Clinically significant hypoglycemia
  • Uncontrolled HTN (defined as SBP > 185mmHg or DPB > 110mmHg)
  • Hereditary or acquired hemorrhagic diathesis
  • GIB or Hematuria within previous 21days
  • Major surgery within preceding 14 days
  • Exposure to a thrombolytic agent within the previous 72hrs


  • Another study stopped early (225 patients of planned 310 patients had been randomized)
    • tPA = 113
    • Placebo = 112
  • mRs 0 – 1 at 3mo:
    • tPA: 35%
    • Placebo 29%
    • aRR 1.44
    • p = 0.042
  • mRs 0 – 2 at 3mo:
    • tPA: 50%
    • Placebo: 43%
    • aRR 1.36
    • p = 0.017
  • Clinical Improvement (NIHSS reduction) at 24hrs:
    • tPA: 22%
    • Placebo: 10%
    • aRR = 2.51
    • p = 0.006
  • Mortality:
    • tPA: 12%
    • Placebo: 9%
    • aRR 1.17
    • p = 0.665
  • sICH:
    • tPA: 6%
    • Placebo: 1%
    • aRR: 7.22
    • p = 0.053


  • According to the authors the study was terminated early due to loss of clinical equipoise based on other trial results.
  • It is difficult to discuss strengths and limitations when only the abstract is available, and so not done for this study

Author Conclusion: “Ischemic stroke patients with salvageable brain tissue presenting 4.5 – 9 hours from onset or with WUS who received alteplase achieved better functional outcomes, reperfusion and early neurological improvement.  Mortality was comparable despite numerically more sICH. Further per-protocol and subgroup analysis results will be presented.” 

Clinical Take Home Point: A second multicenter randomized trial of patients with strokes of unknown onset, the use of tPA resulted in a modest 6% improvement in neurologic outcomes at 90days, but more ICH, and although not statistically significant also resulted in more death.

Clinical Bottom Line:  Both of these studies are examples of trials stopped early that likely suffer from overestimation of benefit. The authors of these studies clearly recommend expansion of tPA use to a broader stroke population, at the expense of changing our current health care systems (How many EDs in the community have the ability to perform MR DWI/FLAIR or CT-CBF advanced imaging?).

These studies do, however, further confirm moving away from a time is brain paradigm to a perfusion-based strategy to see if there is salvageable tissue instead of time cutoffs.  This may or may not be beneficial to patients with strokes with unknown onset of symptoms, but tPA is certainly not a benign medication with trends toward increased early mortality and increased sICH.  Let’s learn from our past mistakes and not change systems until these studies are replicated with proven benefit and no increase in harms.


  1. Thomalla G et al. MRI-Guided Thrombolysis for Stroke with Unknown Time of Onset. NEJM 2018. PMID: 29766770
  2. Ma H et al. Abstract LB21 – Extending the Thrombolytic Time Window to 9 Hours for Acute Ischemic Stroke Using Perfusion Imaging Selection – The Final Result. International Stroke Conference Feb 2019. [Epub Ahead of Print]
  3. Ma H et al. A Multicentre, Randomized, Double-Blinded, Placebo-Controlled Phase III Study to Investigate Extending the time for Thrombolysis in Emergency Neurological Deficits (EXTEND). Int J of Stroke 2012. PMID: 22188854

Post Peer Reviewed By: Anand Swaminathan, MD (Twitter: @EMSwami)

Cite this article as: Salim Rezaie, "Extending the tPA Window to 4.5 – 9 Hours in Acute Ischemic Stroke (AIS)?", REBEL EM blog, April 1, 2019. Available at: https://rebelem.com/extending-the-tpa-window-to-4-5-9-hours-in-acute-ischemic-stroke-ais/.

Like this article?

Share on Facebook
Share on Twitter
Share on Linkdin
Share via Email

Want to support rebelem?