Neurology Clinical Trial Database
WAKE-UP
Efficacy and Safety of MRI-Based Thrombolysis in Wake-Up Stroke (WAKE-UP) trial
Year of Publication: 2018
Authors: G. Thomalla, C.Z. Simonsen, F. Boutitie, ..., and C. Gerloff
Journal: The New England Journal of Medicine
Citation: N Engl J Med 2018;379:611-22.
PDF: Clinical Question
Do patients with acute stroke with an unknown time of onset and features suggesting recent cerebral infarction on MRI (diffusion-weighted imaging lesion without parenchymal hyperintensity on FLAIR) benefit from intravenous thrombolysis with alteplase?
Study Overview
Objective
To determine whether IV alteplase improves outcomes in patients with unknown stroke onset time, selected via MRI mismatch between DWI and FLAIR.
Study Summary
- IV alteplase improved functional outcomes at 90 days in patients with unknown-onset stroke selected by DWI-FLAIR mismatch on MRI;
- Alteplase was associated with a higher risk of symptomatic intracranial hemorrhage and mortality.
Intervention
IV alteplase 0.9 mg/kg vs. placebo, administered to patients with MRI-confirmed mismatch pattern.
Patients per Arm
Alteplase: 254, Placebo: 249
Bottom Line
In patients with acute stroke with an unknown time of onset selected by diffusion-weighted imaging (DWI)-FLAIR mismatch on MRI, intravenous alteplase resulted in a significantly better functional outcome at 90 days compared to placebo. However, it was also associated with a numerically higher incidence of intracranial hemorrhages.
Major Points
- WAKE-UP was the first trial to use MRI-based tissue signatures (DWI-FLAIR mismatch) to select stroke patients with UNKNOWN onset time for thrombolysis — pioneering imaging-guided rather than clock-based treatment decisions.
- DWI-FLAIR mismatch concept: a visible DWI lesion (acute ischemia) WITHOUT corresponding FLAIR hyperintensity suggests stroke onset within ~4.5 hours — because FLAIR signal changes take several hours to develop. This imaging 'tissue clock' replaced the traditional 'time clock.'
- Primary outcome: favorable outcome (mRS 0–1) at 90 days in 53.3% alteplase vs 41.8% placebo (adjusted OR 1.61, 95% CI 1.09–2.36, p=0.02) — an 11.5 percentage-point absolute benefit, one of the largest treatment effects in any thrombolysis trial.
- Ordinal mRS analysis also highly significant: common OR 1.62 (95% CI 1.17–2.23, p=0.003). Median mRS was 1 (alteplase) vs 2 (placebo) — a clinically meaningful difference between independence with no significant disability vs slight disability.
- Trial stopped early due to cessation of funding (503 of planned 800 patients enrolled) — this reduced statistical power and may have overestimated the treatment effect. Despite early stopping, the primary endpoint was significant.
- Mortality signal: 10 deaths (4.1%) alteplase vs 3 (1.2%) placebo (adjusted OR 3.38, 95% CI 0.92–12.52, p=0.07) — numerically higher but not statistically significant. This imbalance is concerning and may reflect the risks of treating patients with potentially older strokes.
- sICH (SITS-MOST definition): 2.0% alteplase vs 0.4% placebo (adjusted OR 4.95, 95% CI 0.57–42.87, p=0.15). By ECASS III definition: 2.4% vs 0.4% (OR 6.04, p=0.10). By ECASS II definition: 2.8% vs 1.2% (OR 2.40, p=0.21). Parenchymal hemorrhage type 2 (a separate imaging finding): 4.0% vs 0.4% (OR 10.46, p=0.03).
- 89% of patients presented after waking from nighttime sleep — the classic 'wake-up stroke' scenario where onset is truly unknown. Median time from last-seen-well to treatment was ~10 hours — far beyond any traditional time window.
- Double-blind, placebo-controlled design — the gold standard for thrombolysis trials. This contrasted with the open-label design of IST-3 and ENCHANTED, providing higher-quality evidence.
- Alongside the contemporaneous DAWN and DEFUSE 3 thrombectomy trials — which also used imaging (rather than clock time) to extend the treatment window for unknown-onset and late-presenting strokes — WAKE-UP helped establish the paradigm of IMAGING-BASED patient selection, fundamentally shifting acute stroke treatment from 'time is brain' to 'tissue is brain.'
Design
Study Type: Investigator-initiated, multicenter, randomized, double-blind, placebo-controlled clinical trial
Randomization: 1
Blinding: Double-blind (patients, investigators, central image-reading committee, central safety-adjudication committee).
Enrollment Period: September 24, 2012, to June 30, 2017 (stopped early)
Follow-up Duration: 90 days
Centers: 70
Countries: 8 European countries
Sample Size: 503
Analysis: Intention-to-treat (ITT) population for primary and secondary efficacy outcomes. Unconditional logistic-regression model for primary efficacy (odds ratio, 95% CI). Proportional-odds logistic-regression model for ordinal mRS (common odds ratio). Safety end points analyzed in safety population. Multiple imputation for missing primary outcome data. Two-sided alpha level of 0.05.
Inclusion Criteria
- Patients presented with clinical signs of acute stroke.
- 18 to 80 years of age.
- Able to carry out usual activities in daily life without support before the stroke.
- Patient either recognized stroke symptoms on awakening or could not report the timing of onset (e.g., aphasia or confusion).
- Time elapsed since last known to be well >4.5 hours (no upper limit).
- MRI examination showed a mismatch between the presence of an abnormal signal on MRI diffusion-weighted imaging and no visible signal change on FLAIR in the region of the acute stroke.
- Could be treated with alteplase within 4.5 hours after symptom recognition.
Exclusion Criteria
- MRI showed intracranial hemorrhage or lesions larger than one third of the territory of the middle cerebral artery.
- Thrombectomy was planned.
- Severe stroke (NIHSS score >25).
- Generally recognized contraindications to alteplase (except for unknown time of symptom onset).
- Could not be treated within 4.5 hours after symptom recognition.
Baseline Characteristics
| Characteristic | Control | Active |
|---|---|---|
| Mean age ±SD - yr | 65.2±11.9 | 65.3±11.2 |
| Male sex no. (%) | 160 (64.3) | 165 (65.0) |
| Reason for unknown time of symptom onset - Nighttime sleep no. (%) | 222 (89.2) | 227 (89.4) |
| Reason for unknown time of symptom onset - Daytime sleep no. (%) | 11 (4.4) | 12 (4.7) |
| Reason for unknown time of symptom onset - Aphasia, confusion, or other no. (%) | 16 (6.4) | 15 (5.9) |
| Median interval between last time the patient was known to be well and symptom recognition (IQR) -hr | 7.0 (5.0-9.0) | 7.2 (4.7-8.7) |
| Medical history - Arterial hypertension no. (%) | 131 (52.6) | 135 (53.1) |
| Medical history - Diabetes mellitus no. (%) | 39 (15.7) | 43 (16.9) |
| Medical history - Hypercholesterolemia no. (%) | 85 (34.1) | 93 (36.6) |
| Medical history - Atrial fibrillation no. (%) | 29 (11.6) | 30 (11.8) |
| Medical history - History of ischemic stroke no. (%) | 31 (12.4) | 37 (14.6) |
| Median NIHSS score (IQR) | 6 (4-9) | 6 (4-9) |
| Vessel occlusion on time-of-flight MRA - Any no./total no. (%) | 84/246 (34.1) | 84/249 (33.7) |
| Vessel occlusion on time-of-flight MRA - Intracranial internal carotid artery no. (%) | 11/246 (4.5) | 24/249 (9.6) |
| Vessel occlusion on time-of-flight MRA - Middle cerebral artery main stem no. (%) | 37/246 (15.0) | 35/249 (14.1) |
| Vessel occlusion on time-of-flight MRA - Middle cerebral artery branch no. (%) | 36/246 (14.6) | 32/249 (12.9) |
| Vessel occlusion on time-of-flight MRA - Other no. (%) | 12/246 (4.9) | 12/249 (4.8) |
| Median lesion volume on diffusion-weighted imaging (IQR) - ml. | 2.5 (0.7-8.8) | 2.0 (0.8-7.9) |
| Median time from symptom recognition to MRI (IQR) -hr | 2.6 (2.1-3.3) | 2.6 (1.9-3.3) |
| Median time between end of MRI and treatment initiation (IQR) min | 26 (18-37) | 25 (16-35) |
| Median time from symptom recognition to treatment initiation (IQR)-hr | 3.2 (2.6-3.9) | 3.1 (2.5-3.8) |
| Interval between last time that the patient was last known to be well and treatment initiation (IQR) -hr | 10.4 (8.1-12.1) | 10.3 (8.1-12.0) |
Arms
| Field | Alteplase Group | Control |
|---|---|---|
| Intervention | Intravenous alteplase (0.9 mg per kilogram of body weight, with 10% as bolus and remainder infused over 60 minutes). | Matching placebo. |
| Duration | 60 minutes infusion (followed by 90-day follow-up) | 60 minutes infusion (followed by 90-day follow-up) |
Outcomes
| Outcome | Type | Control | Intervention | HR / OR / RR | P-value |
|---|---|---|---|---|---|
| Favorable clinical outcome, defined as a score of 0 or 1 on the modified Rankin scale of neurologic disability at 90 days. | Primary | 102/244 (41.8%) | 131/246 (53.3%) | 11.5 percentage points | 0.02 |
| Median score on modified Rankin scale at 90 days (ordinal distribution / shift analysis) | Secondary | 2 (IQR 1-3) | 1 (IQR 1-3) | 0.003 | |
| Treatment response at 90 days (deficit-adjusted mRS threshold) - no./total no. (%) | Secondary | 44/244 (18.0%) | 72/246 (29.3%) | 0.004 | |
| Global Outcome Score at 90 days (multidimensional favorable outcome across mRS, NIHSS, Barthel Index, GOS) | Secondary | 0.02 | |||
| Median score on Beck Depression Inventory at 90 days | Secondary | 7.0 (IQR 2.0-14.0) | 6.0 (IQR 2.0-11.0) | 0.69 | |
| Total score on EQ-5D at 90 days | Secondary | 2.4±2.4 | 1.9±2.1 | 0.004 | |
| Score on visual analog scale on EQ-5D at 90 days | Secondary | 64.9±23.8 | 72.6±19.7 | <0.001 | |
| Median infarct volume on diffusion-weighted imaging at 22-36 hr (ml) | Secondary | 3.3 (IQR 1.1-16.6) | 3.0 (IQR 0.8-17.7) | 0.32 | |
| Death or dependency at 90 days (mRS 4-6) | Adverse | 44/241 (18.3%) | 33/244 (13.5%) | 0.17 | |
| Death at 90 days | Adverse | 3 (1.2%) | 10 (4.1%) | 0.07 | |
| Symptomatic intracranial hemorrhage (SITS-MOST definition) | Adverse | 1 (0.4%) | 5 (2.0%) | 0.15 | |
| Symptomatic intracranial hemorrhage (ECASS II definition) | Adverse | 3 (1.2%) | 7 (2.8%) | 0.21 | |
| Symptomatic intracranial hemorrhage (ECASS III definition) | Adverse | 1 (0.4%) | 6 (2.4%) | 0.10 | |
| Symptomatic intracranial hemorrhage (NINDS definition) | Adverse | 12 (4.9%) | 20 (8.0%) | 0.13 | |
| Parenchymal hemorrhage type 2 (imaging finding, distinct from sICH definitions) | Adverse | 1 (0.4%) | 10 (4.0%) | 0.03 | |
| Space-occupying brain infarction or edema with clinical deterioration | Adverse | 2 (0.8%) | 6 (2.4%) | ||
| Recurrent ischemic stroke - Asymptomatic | Adverse | 55 (22.5%) | 58 (23.1%) | ||
| Recurrent ischemic stroke - Symptomatic | Adverse | 8 (3.3%) | 17 (6.8%) | ||
| Major extracranial bleeding | Adverse | 0 | 3 (1.2%) | ||
| Severe anaphylactic reaction | Adverse | 1 (0.4%) | 0 | ||
| Any serious adverse event | Adverse | 52 (21.3%) | 56 (22.3%) | 0.83 | |
| Nervous system disorders (including sICH and recurrent stroke) | Adverse | 8.6% | 13.5% |
Criticisms
- Stopped early due to anticipated funding cessation — only 503 of planned 800 patients enrolled, leaving trial underpowered for secondary endpoints and safety signals including the numerically higher death rate in the alteplase group (4.1% vs 1.2%).
- Excluded patients planned for thrombectomy, limiting generalizability to the modern endovascular era where many unknown-onset patients are thrombectomy candidates.
- Two-thirds of screened patients excluded — primarily for lacking DWI-FLAIR mismatch — limiting applicability to the broader unknown-onset stroke population and raising questions about external validity.
- Parenchymal hemorrhage type 2 (a separate imaging finding, distinct from sICH definitions) was significantly higher with alteplase (4.0% vs 0.4%, adjusted OR 10.46, p=0.03); the various formal sICH definitions all trended higher with alteplase but did not reach statistical significance in this underpowered trial — consistent with known tPA risks but concerning given the population's inherently uncertain symptom onset.
- Infarct volume at 22-36 hours not significantly different between groups despite clinical benefit — unexpected finding that challenges the assumed mechanism of penumbral salvage.
- No CT-based alternative pathway — trial mandated MRI with DWI-FLAIR, excluding hospitals without 24/7 MRI access and limiting real-world implementation, especially in resource-limited settings.
- EU-only enrollment (70 centers in 8 European countries) — ethnic and geographic homogeneity may limit generalizability to non-European populations with different stroke demographics.
- Relatively mild strokes (median NIHSS 6) — treatment effect in severe strokes with unknown onset remains uncertain given very few patients with NIHSS >10.
- No long-term follow-up beyond 90 days — durability of functional benefit and delayed safety outcomes (including post-stroke dementia risk) were not assessed.
Subgroup Analysis
Array
Funding
European Union Seventh Framework Program
Based on: WAKE-UP (The New England Journal of Medicine, 2018)
Authors: G. Thomalla, C.Z. Simonsen, F. Boutitie, ..., and C. Gerloff
Citation: N Engl J Med 2018;379:611-22.
Content summarized and formatted by NeuroTrials.ai.