By Ma, H., Campbell, B.C., Parsons, M.W., et al. New England Journal of Medicine, 380(19), pp.1795-1803.
The recently successful thrombectomy trials (DEFUSE 3 and DAWN) have set a paradigm of reperfusion based on tissue viability assessed by perfusion imaging rather than time from symptom onset. However, this has also caused a disparity between the accepted time frame for chemical thrombolysis (up to 4.5 hours) and mechanical thrombectomy (up to 24 hours). This international multicenter phase 3 randomized controlled trial hypothesized that intravenous thrombolysis between 4.5-9 hours could be of functional benefit in patients with a favorable perfusion imaging profile. Adult ischemic stroke patients who awoke with symptoms or those within the 4.5 to 9-hour window were included in the trial. For wake-up strokes, the time of onset was taken as the midpoint of sleep, and these were included if they presented within 9 hours. Inclusion criteria included excellent baseline function (mRS<2), NIHSS 4-26 on presentation, and ischemic penumbra based on MRI or CT perfusion imaging. Core infarct volume on imaging was based on diffusion restriction when MRI was used and <30% relative cerebral blood flow when CT perfusion was used. Hypoperfusion (penumbra+core infarct) was defined as the area involved with a tmax > 6 seconds (delayed arrival on injected tracer). Patients with a mismatch ratio (hypoperfusion to core) > 1.2, penumbra at least > 10 ml, and ischemic core < 70 ml were eligible. Patients being considered for mechanical thrombectomy were excluded. Patients were randomized in a 1:1 ratio to receive standard dose alteplase or placebo. Randomization was stratified based on geographic region and time to intervention (>4.5-6 hours, 6-9 hours, or wake up strokes).
The primary outcome was mRS 0-1 at 90 days adjusted for age and clinical severity (NIHSS) on presentation. Secondary outcomes involved an ordinal shift analysis for mRS 0-6, an mRS score 0-2 (functional independence), >50% and >90% reperfusion (indicted by decreased volume of hypoperfusion) at 24 hours. Tertiary outcomes included recanalization at 24 hours (in the presence of arterial occlusion), and major neurological improvement (reduction in NIHSS >8 points or an NIHSS of 0-1 within 24 hours, 72 hours and 90 days). Safety outcomes included death at 90 days and symptomatic intracranial hemorrhage. The original sample size was calculated as 400 patients, but after a blinded review of observed number of patients with data that could not be evaluated, as well as the number of patients that were lost to follow up, the sample size was reduced to 310 patients.
After the results of the WAKE-UP trial were published in May 2018, the DSMB recommended discontinuation of the trial due to a loss of equipoise. Over a period of 8 years, 225 patients were enrolled (113 alteplase and 112 placebo). The mean age in the alteplase group was 73.7 years (vs 71.0 years in the placebo group), and median NIHSS in the alteplase group was 12 (IQR 8-17) and 10 (IQR 6-16.5) in the placebo group. Of the entire cohort, 10% were randomized in the 4.5-6 hour window, 25% in the 6-9 hour window and 65% that awoke with symptoms. The median time from stroke onset to study intervention was 432 minutes (IQR 374-488) in the alteplase arm and 450 minutes (IQR 374-500) in the placebo arm. The alteplase arm had 69% with a large vessel occlusion (vs. 72.3% in the placebo arm). The median core infarct volumes were 4.6 ml (IQR 0-23.2) and 2.4 ml (IQR 0-19.5) ml in the alteplase and placebo arm respectively. The median volumes of the ischemic areas (core+penumbra) were 74.3 ml (IQR 40.1-134) in the alteplase arm and 78 ml (IQR 47.7-111.8) in the placebo arm.
The primary end point (mRS 0-1) was achieved by 35.4% in the alteplase group and 29.5% in the placebo group (adjusted risk ratio 1.44 [95% CI 1.01-2.06]; p=0.04). Among the secondary outcomes: 1) there was no significant difference between the groups in the ordinal shift analysis; 2) functional independence (mRS 0-2) was attained by 49.6% in the alteplase arm and 42.9% in the placebo arm (adjusted risk ratio 1.36 [95% CI 1.06-1.76]); 3) treatment with alteplase resulted in significantly greater 24 hour reperfusion of at least 50% (71.7% vs. 52.3% [adjust risk ratio 1.35]) and at least 90% (50% vs. 28.4% [adjust risk ratio 1.73]). With regards to tertiary outcomes: 1) recanalization at 24 hours occurred in 67.3% in the alteplase group and 39.4% in the placebo group (adjusted risk ratio 1.68 [95% CI 1.29-2.19]; 2) major neurological improvement at 24 hours was seen in 23.9% of patients in the alteplase group vs. 9.8% patients in the placebo group (adjusted risk ratio 2.76 [95% CI 1.45-5.26]). There were no differences among major neurological improvement at 72 hours and 90 days. There were no significant subgroup interactions regarding presence of large vessel occlusion or time from symptom onset.
Safety: 90-day mortality did not differ significantly among the groups (11.5% alteplase group vs. 8.9% placebo; p=0.67). Symptomatic intracranial hemorrhage occurred in 6.2% of patients in the alteplase group and 0.9% of patients in the placebo group (adjust risk ratio 7.22 [95% CI 0.97-53.54]; p=0.053].
The authors conclude that among ischemic stroke patients with a favorable perfusion imaging profile presenting in the 4.5-9 hour window, intravenous alteplase resulted in a higher percentage of patients with an mRS of 0 or 1 at 90 days. The current study shows marginal benefit in the extended time window for thrombolysis but with higher hemorrhage rates, albeit non-significant. One of the major limitations of the study is its premature termination (less than three fourths of the planned sample size) based on the results of the WAKE-UP trial. The WAKE-UP trial included wake up stroke patients who had ischemic lesions on diffusion weighted imaging but no signal change on FLAIR imaging, suggesting that the stroke had occurred within the previous 4.5 hours. As the current study included patients up to 9 hours, the rationale for termination based on a study with different imaging criteria can be drawn in to question. The initial sample size that was calculated for this study was also changed to a lower number after an interim blinded analysis.
This study’s inability to show significant effect differences in the unadjusted primary and secondary outcomes may have been due to the small sample size. The regression analyses that were done to demonstrate significance in the adjusted primary end point have also been drawn in to question. Could this have been avoided by implementing block randomization that included the adjusted variables? The authors do warn against inferences being made from the adjusted secondary outcomes as there was no initial plan for correction for multiple comparisons. It is also unclear why the current study chose 9 hours as the cut off for thrombolysis when the initial rationale was that of tissue viability as opposed to a time cut off (thrombectomy trials extend to 24 hours if perfusion imaging is favorable). The rates of symptomatic intracranial hemorrhage were higher in the alteplase group (6%) even though it was not significantly higher than placebo. These rates are similar to other perfusion-based reperfusion trials in the extended time window and include the EPITHET, DAWN and DEFUSE-3 trials. This study does support moving away from the archetype of time-based cut offs to an era of tissue viability assessment. Future standardized studies, to assess the utility of chemical thrombolysis in the extended time window and the era of thrombectomy are warranted.
Wazim Mohamed, MD
Assistant Professor of Neurology
Division of Neurotrauma and Critical Care Neurology
Wayne State University