Thrombolysis for wake-up stroke versus non-wake-up unwitnessed stroke: EOS individual patient data meta-analysis

Abstract

Background:

Stroke with unknown time of onset can be categorized into two groups; wake-up stroke (WUS) and unwitnessed stroke with an onset time unavailable for reasons other than wake-up (non-wake-up unwitnessed stroke, non-WUS). We aimed to assess potential differences in the efficacy and safety of intravenous thrombolysis (IVT) between these subgroups.

Methods:

Patients with an unknown onset stroke were evaluated using individual patient-level data of two randomized controlled trials (WAKE-UP, THAWS) comparing IVT with placebo or standard treatment from the Evaluation of Unknown Onset Stroke Thrombolysis trials (EOS) dataset. A favorable outcome was pre-specified as a modified Rankin Scale score of 0–1 at 90 days. Safety outcomes included symptomatic intracranial hemorrhage (sICH) at 22–36 hours and 90-day mortality. The IVT effect was compared between the treatment groups in the WUS and non-WUS with multivariable logistic regression analysis.

Results:

634 patients from two trials were analyzed; 542 had WUS (191 women, 272 receiving alteplase), and 92 had non-WUS (42 women, 43 receiving alteplase). Overall, no significant interaction was noted between the mode of onset and treatment effect (Pinteraction=0.796). In patients with WUS, the frequencies of favorable outcomes were 54.8% and 45.5% in the IVT and control groups, respectively (adjusted odds ratio [aOR], 1.47; 95% CI 1.01–2.16). Death occurred in 4.0% and 1.9%, respectively (p=0.162), and sICH in 1.8% and 0.3%, respectively (p=0.194). In patients with non-WUS, no significant difference was observed in favorable outcomes relative to the control (37.2% vs. 29.2%, aOR 1.76; [0.58–5.37]). One death and one sICH were reported in the IVT group, but none in the control.

Conclusions:

There was no difference in the effect of IVT between patients with WUS and non-WUS. IVT showed a significant benefit in patients with WUS, while there was insufficient statistical power to detect a substantial benefit in the non-WUS subgroup.

Clinical Trial Registration:

PROSPERO CRD42020166903

Graphical Abstract

Introduction

Unknown onset stroke refers to a stroke whose exact onset time is not known. It is common, reported to be 14%–36% in patients with acute ischemic stroke (AIS).^1–3 Patients with unknown onset stroke can be mainly categorized into two groups: the wake-up stroke (WUS), which is a stroke in patients who wake up with stroke symptoms, and unwitnessed stroke or where the symptom onset time cannot be estimated due to aphasia, impaired consciousness, or cognitive impairment (non-wake-up unwitnessed stroke, non-WUS). Few previous studies have analyzed WUS and non-WUS as separate entities, as both groups often are merged in the same category. However, previous studies have shown differences in demographic data, neurological severity, neuroimaging findings, and clinical courses between patients with WUS and those with non-WUS. The differences could impact clinical outcomes.^3–5

Recent advances in neuroimaging techniques have enabled the estimation of the approximate time of stroke onset using magnetic resonance imaging (MRI)⁶ or the identification of potentially salvageable tissue at risk using perfusion-diffusion MRI or computed tomography (CT) perfusion.⁷ Recent randomized controlled trials (RCTs) and meta-analyses have shown that intravenous thrombolysis (IVT) is beneficial for patients with unknown onset stroke selected by these advanced neuroimaging biomarkers.^8–11 The Evaluation of Unknown Onset Stroke thrombolysis trials (EOS) group reported that IVT resulted in better functional outcome at 90 days than placebo or standard care in patients with unknown onset stroke selected by MRI-based tissue-clocking or penumbral imaging.¹⁰ Hence, the current treatment guidelines recommend IVT for eligible patients.^12–14 However, the proportion of IVT was one-third lower among WUS than among known-onset stroke (7.6% vs. 26.4%), and lowest among non-WUS (3.2%), suggesting the paucity of IVT data for unknown onset stroke in a population-based observational study.¹⁵ Furthermore, previous data comparing stroke outcomes in WUS and non-WUS are limited, although the latter might be potentially at higher risk for worse clinical outcomes.⁴ However, functional outcomes after IVT have not been reported, comparing patients with WUS and non-WUS using RCT datasets.

We aimed to compare the efficacy and safety of IVT based on advanced brain imaging for the WUS and non-WUS groups, using two trials selected from the EOS dataset.

Methods

The data supporting the findings of this study are available from the corresponding author upon reasonable request and with approval from the EOS steering committee.

Study population

We analyzed the dataset of EOS, a pooled individual patient dataset composed of participants from four randomized controlled trials that investigated intravenous alteplase for strokes with unknown time of onset guided by advanced brain imaging.¹⁰ We used the patients’ data from the following two trials of the EOS: the WAKE-UP (Efficacy and Safety of MRI-Based Thrombolysis in Wake-Up Stroke)⁸ and the THAWS (Thrombolysis for Acute Wake-Up and Unclear-Onset Strokes With Alteplase at 0.6 mg/kg)¹⁶ trials which enrolled patients with unknown time of onset stroke guided by a mismatch in the visibility of acute ischemic lesions on MRI between diffusion-weighted imaging (DWI) and fluid-attenuated inversion recovery (FLAIR; DWI-FLAIR mismatch). The EOS dataset includes other two trials, but the EXTEND (Extending the Thrombolytic Time Window to 9 Hours for Acute Ischemic Stroke Using Perfusion Imaging Selection)¹⁷ and the ECASS-4 (European Cooperative Acute Stroke Study-4)¹⁸ trials, which used core/perfusion mismatch, were excluded from the current analysis due to the lack of data on the mode of stroke onset. Patients eligible for each trial could perform usual activities of daily living without support before stroke onset (defined as modified Rankin Scale score <2). The individual patient data, meta-analysis design, and primary outcomes have been described elsewhere.^10,19

Ethical approval was obtained from all participating institutions for these included trials by the institutional review board of each site. All patients or their legal representatives provided written informed consent according to national and local regulations, except explicit informed consent in emergencies in some countries.

In the present study, stroke with an unknown time of onset was classified into two groups: the WUS group, in which patients were aware of symptoms upon waking, and the non-WUS group, in which the onset of symptoms had not been witnessed, and the patients were unable to provide information at the time of onset on their own because of aphasia, impaired consciousness, or cognitive impairment.

Clinical data collection

Baseline clinical characteristics included age, sex, atrial fibrillation, vascular risk factors (hypertension, diabetes mellitus, and dyslipidemia), other past medical history (ischemic stroke or transient ischemic attack before the index event), premorbid anticoagulation or antiplatelets, baseline National Institutes of Health Stroke Scale (NIHSS) score, large vessel occlusion, and time intervals (last-known-well to symptom recognition, symptom recognition to hospital arrival, and last-known-well to treatment time). Vessel occlusion was determined using baseline MR angiography or CT angiography readings provided by the individual trials. Large vessel occlusion was defined as occlusion of the internal carotid artery or main stem of the middle cerebral artery. Ischemic lesion volume was measured by DWI or CT perfusion.

Outcomes

The primary efficacy outcome was a favorable outcome, defined as a modified Rankin Scale (mRS) score of 0–1 at 90 days post-stroke onset. This was the same primary outcome as was used in prior EOS analyses.¹⁰ The secondary outcome was a functionally independent outcome defined as an mRS score of 0–2 at 90 days. Safety outcomes included death at 90 days, radiologically defined parenchymal hematoma type 2, and any symptomatic intracranial hemorrhage (sICH). The definition of sICH was based on the Safe Implementation of Thrombolysis in Stroke Monitoring Study (SITS-MOST).²⁰ Additional outcomes were death within 7 days after randomization and severe disability or death (mRS score 4–6) at 90 days.

Statistical analysis

Data between patients with WUS and those with non-WUS are summarized as median (interquartile range [IQR]) for continuous variables and as frequencies and percentages for categorical variables. Statistical differences between the two groups were assessed using the Student’s t-test or Mann-Whitney U test for continuous variables and the χ² or Fisher’s exact test for categorical variables, as appropriate. The effect of intravenous alteplase on each outcome was separately compared between the assigned treatment groups (alteplase and control groups) in patients with WUS and those with non-WUS. Both efficacy and safety outcomes were analyzed with the multivariable logistic regression models with adjustments for trials (WAKE-UP and THAWS), the randomization stratification parameters of age and NIHSS score at baseline were included,¹⁰ atrial fibrillation and vessel occlusion on MRA based on the univariable analysis of baseline characteristics and previous reports.^3–5 We evaluated the adequacy of the models using the Akaike Information Criterion (AIC), where smaller values indicate better model fit. The results were reported as odds ratios (ORs) with 95% confidence intervals (CIs). The interaction between the treatment and the mode of stroke onset was tested by adding an interaction term to the model used to analyze the primary outcome (mRS 0–1 at 90 days). As a secondary analysis, we used proportional-odds logistic regression analysis with the same adjustments to calculate common odds ratios for the categorical shift in the distribution of mRS scores. For further analysis of the impact of the mode of onset in patients treated with IVT, we performed the stratified analysis of efficacy and safety outcomes in the IVT-treated subgroup. Additionally, all reported P values were 2-tailed, with p<0.05 considered statistically significant. Statistical analyses were performed using the JMP version 14 software (SAS Institute, Cary, NC, USA) and Stata (version 16, Stata Corporation, College Station, TX, USA). This study follows the Preferred Reporting Items for a Systematic Review and Meta-analysis of Individual Participant Data (PRISMA-IPD) reporting guideline.²¹

Results

Patient characteristics

Among the 843 patients in the EOS dataset, the data from 634 patients with an unknown-onset time (233 women [36.8%], median age 70 years [IQR, 61–76]) were available for analysis (Supplemental Figure S1). We excluded patients enrolled from the EXTEND (n=146) and ECASS-4 (n=63) trials since the definition of “awoke with stroke symptoms” included both WUS and non-WUS which were not distinguished. Clinical characteristics of the excluded patients and the analyzed patients are summarized in Supplementary Table S1. We found that the excluded patients were significantly older, had higher rates of atrial fibrillation and hypertension, had higher baseline NIHSS score than the analyzed patients.

The patients in the current study were enrolled from the WAKE-UP (n=503) and THAWS (n=131) trials. Of these patients, 542 developed WUS (85.5%), and 92 had non-WUS (14.5%). Clinical characteristics of the WUS and non-WUS groups are presented in Table 1. The patients in the WUS group were younger and had lower rates of atrial fibrillation, diabetes mellitus, history of ischemic stroke/transient ischemic attack, premorbid anticoagulation, vessel occlusion on MRA, as well as smaller ischemic lesion volume and lower baseline NIHSS score than the non-WUS group. The time intervals were insignificantly different between the groups.

Table 1.

Clinical characteristics

	Wake-up stroke (n = 542)	Non-Wake-up stroke (n = 92)	P value
Age, y, median (IQR)	69 (61–75)	72 (63–80)	<0.01
Female, n (%)	191 (35.2)	42 (45.7)	0.056
Medical history
Atrial fibrillation, n (%)	70 (13.0)	21 (23.6)	<0.01
Hypertension, n (%)	302 (55.9)	54 (59.3)	0.54
Diabetes mellitus, n (%)	84 (15.6)	24 (26.7)	0.010
Dyslipidemia, n (%)	193 (36.8)	31 (35.6)	0.84
History of ischemic stroke/TIA, n (%)	70 (12.9)	20 (22.0)	0.022
Premorbid anticoagulation, n (%)	13 (2.4)	7 (7.6)	<0.01
Premorbid antiplatelets, n (%)	159 (29.3)	32 (34.8)	0.29
Findings
Baseline NIHSS score, median (IQR)	6 (4–9)	9 (4–16)	<0.01
Baseline systolic blood pressure, mmHg, median (IQR),	156 (140–170)	150 (135–170)	0.24
Baseline diastolic blood pressure, mmHg, median (IQR)	85 (75–93)	83 (71–90)	0.44
Any vessel occlusion on MRA, n(%)	184 (34.5)	44 (47.8)	0.014
Large vessel occlusion on MRA, n (%)	96 (18.0)	18 (19.6)	0.72
Ischemic core volume, mL, median (IQR)	2.0 (0.7–7.3)	7.5 (1.4–18.3)	<0.01
Time intervals
Time from last-known-well to symptom recognition, min, median (IQR)	420 (300–520)	338 (180–600)	0.14
Time from symptom recognition to hospital arrival, min, median (IQR)	95 (66–140)	86 (57–122)	0.12
Time from last-known-well to treatment, min, (IQR)	626 (507–724)	578 (383–905)	0.24

Abbreviations: IQR, interquartile range; MRA, Magnetic Resonance angiography; NIHSS, National Institutes of Health Stroke Scale; TIA, transient ischemic attack

The peak for symptom recognition in the WUS group occurred in the morning, whereas the peak for the non-WUS group was widely distributed during the daytime hours (Figure 1A). The peak for the last known well time of the WUS group was predominant at night, while the peak for the non-WUS group was more common during the daytime and in the evening (Figure 1B).

Figure 1. Hour of symptoms recognition (A) and last-known well-time (B) in wake-up and non-wake-up unknown-onset strokes.

Hours on the x-axis refers to 24-hour clock time (midnight to midnight, divided into 24 hours, numbered from 0 to 23)

Outcomes for overall patients

The clinical outcomes are summarized in Table 2 and Table 3. The distribution of the mRS scores at 90 days in each group is presented in Figure 2. Overall, no significant interaction for a favorable outcome (mRS score of 0–1 at 90 days) was apparent between the mode of onset and treatment (P for interaction = 0.796, Figure 3). In patients with WUS, the frequency of favorable outcomes was 54.8% in the alteplase group and 45.5% in the control group, showing a significant difference between the groups (adjusted odds ratio [OR], 1.47; 95% CI 1.01–2.16). Similarly, the frequency of functionally independent outcomes (mRS score 0–2 at 90 days) was 73.9% in the alteplase group and 65.0% in the control group (adjusted OR, 1.64; 95% CI 1.05–2.57). There was a significant shift in mRS at 90 days, favoring the alteplase group (adjusted OR, 1.46; 95% CI 1.07–1.99). In patients with non-WUS, no significant difference was observed between the treatment groups in the favorable (37.2% versus 29.2%, adjusted OR 1.76; 95% CI 0.58–5.37) and functionally independent outcomes (55.8% vs. 54.2%, adjusted OR 1.16; 95% CI 0.35–3.86).

Table 2.

Efficacy and safety outcomes of intravenous alteplase in patients with wake-up stroke.

	Alteplase (n=272)	Control (n=257)	Crude odds ratio (95% CI)	P value	Adjusted odds ratio (95% CI)*	P value
Efficacy outcomes
favorable outcome (mRS 0–1) at 90 d, n (%)	149 (54.8)	117 (45.5)	1.47 (1.04–2.07)	0.027	1.47 (1.01–2.16)	0.046
independent outcome (mRS 0–2) at 90 d, n (%)	201 (73.9)	167 (65.0)	1.55 (1.07–2.24)	0.021	1.64 (1.05–2.57)	0.031
modified Rankin Scale at 90 d (IQR)	2 (1–3)	2 (1–3)	1.39 (1.03–1.89)	0.031	1.46 (1.07–1.99)	0.014
Safety outcomes
death at 7 d, n (%)	10 (3.7)	3 (1.2)	3.20 (0.87–11.7)	0.080	3.88 (0.94–16.0)	0.060
death at 90 d, n (%)	11 (4.0)	5 (1.9)	2.12 (0.73–6.20)	0.168	2.23 (0.73–6.88)	0.162
severe disability or death (mRS 4–6) at 90 d, n (%)	38 (14.0)	47 (18.3)	0.73 (0.46–1.16)	0.178	0.71 (0.41–1.21)	0.205
symptomatic intracranial hemorrhage, n (%)	5 (1.8)	1 (0.3)	4.75 (0.55–40.9)	0.156	4.36 (0.47–40.2)	0.194
parenchymal hemorrhage type 2, n (%)	10 (3.7)	3 (1.2)	3.21 (0.87–11.8)	0.079	2.98 (0.78–11.4)	0.110

^*Odds ratios are adjusted for trials (WAKE-UP and THAWS), age, NIHSS score at baseline, atrial fibrillation, and vessel occlusion on MRA.

We used proportional-odds logistic regression model to analyze the outcome on the mRS at 90 days and calculated the common odds ratio for the categorical shift in the distribution of mRS.

Abbreviations: CI, confidence interval; mRS, modified Rankin Scale; NIHSS, National Institutes of Health Stroke Scale; MRA, magnetic resonance angiography

Table 3.

Efficacy and safety outcomes of intravenous alteplase in patients with non-wake-up unwitnessed stroke.

	Alteplase (n=43)	Control (n=48)	Crude odds ratio (95% CI)	P value	Adjusted odds ratio (95% CI)*	P value
Efficacy outcomes
favorable outcome (mRS 0–1) at 90 d, n (%)	16 (37.2)	14 (29.2)	1.44 (0.60–3.46)	0.416	1.76 (0.58–5.37)	0.322
independent outcome (mRS 0–2) at 90 d, n (%)	24 (55.8)	26 (54.2)	1.07 (0.47–2.44)	0.875	1.16 (0.35–3.86)	0.807
modified Rankin Scale at 90 d (IQR)	2 (1–4)	2 (1–4)	1.22 (0.59–2.53)	0.592	1.42 (0.67–3.01)	0.363
Safety outcomes
death at 7 d, n (%)	0 (0.0)	0 (0.0)	NE	...	NE	...
death at 90 d, n (%)	1 (2.3)	0 (0.0)	NE	...	NE	...
severe disability or death (mRS 4–6) at 90 d, n (%)	11 (25.6)	14 (29.2)	0.83 (0.33–2.11)	0.702	0.65 (0.21–2.00)	0.454
symptomatic intracranial hemorrhage, n (%)	1 (2.3)	0 (0.0)	NE	...	NE	...
parenchymal hemorrhage type 2, n (%)	1 (2.3)	0 (0.0)	NE	...	NE	...

^*Odds ratios are adjusted for trials (WAKE-UP and THAWS), age, NIHSS score at baseline, atrial fibrillation, and vessel occlusion on MRA.

We used proportional-odds logistic regression model to analyze the outcome on the mRS at 90 days and calculated the common odds ratio for the categorical shift in the distribution of mRS.

Abbreviations: CI, confidence interval; mRS, modified Rankin Scale; NIHSS, National Institutes of Health Stroke Scale; MRA, magnetic resonance angiography; NE, not estimable

Figure 2. Distribution of modified Rankin scale scores at 90 days after stroke in patients with wake-up and non-wake-up unknown-onset strokes.

Distribution of modified Rankin scale (mRS) scores (range, 0–6) in patients with wake-up and non-wake-up unwitnessed stroke treated with alteplase or placebo/control.

The numbers are % values.

Figure 3. Association of alteplase with favorable outcome.

Forest plots for the primary outcome of favorable outcome (modified Rankin Scale score 0–1 at 90 days) in patients with wake-up stroke or non-wake-up unwitnessed stroke. Odds ratios are adjusted for trials (WAKE-UP and THAWS), age, NIHSS score at baseline, atrial fibrillation, and vessel occlusion on MRA.

Abbreviations: CI, confidence interval; NIHSS, National Institutes of Health Stroke Scale; MRA, magnetic resonance angiography

There were no significant differences in safety outcomes between the respective treatment groups of patients with WUS and those with non-WUS. No significant interaction for severe disability or death at 90 days was apparent between the mode of onset and treatment (P for interaction = 0.967).

Patients treated with IVT

IVT was performed in 272 patients in the WUS group and 43 patients in the non-WUS group. The WUS group was younger, had lower rates of atrial fibrillation and diabetes mellitus, and lower baseline NIHSS scores than the non-WUS group (Supplementary Table S2).

A favorable outcome was observed in 54.8% of patients with WUS and 37.2% of those with non-WUS, showing no significant difference between the groups (adjusted OR 1.20, 95% CI 0.55–2.62). Similarly, there were no significant differences in functional and safety outcomes between the groups in the adjusted analyses (Table S3).

Discussion

The present study is the first to compare differences in the efficacy and safety of IVT in WUS and non-WUS cases using RCT datasets. The current analysis using pooled individual patient data from RCT showed no difference in the treatment effect of IVT among patients with WUS and those with non-WUS. Treatment with IVT has shown significant benefits as measured by adjusted odds ratios in patients with WUS. However, the present study lacked the power to confidently confirm the therapeutic efficacy of IVT for patients with non-WUS, as this group was much smaller, although there was a higher rate of favorable outcomes (8% increase in the proportion patients with an mRS score of 0–1) and positive odds ratio (high point estimate 1.76) after IVT compared with the control treatment.

The main result of the EOS has clearly shown the efficacy of IVT in patients with stroke of unknown-onset time.¹⁰ However, the small sample size of patients with non-WUS in our analysis likely limits our ability to achieve statistical significance to demonstrate a treatment benefit in this subgroup. The proportion of non-WUS in the present study was one-sixth of the WUS (14% vs. 86%), which was lower than the reported non-WUS, representing a nearly similar rate as the WUS in real-world study data.^15,22

Circadian variations in blood pressure, heart rate, hemostatic processes, and atrial fibrillation episodes are prone to inducing acute stroke in the morning and early hours, which also links to incident WUS.^23,24 Previous studies have reported that the neuroimaging findings of many patients with WUS who present within hours of symptom recognition are similar to those of patients with known-onset stroke who present within hours of symptom onset.^1,22,25,26 The onset of WUS might occur shortly before waking up,²⁷ or alternatively, ischemia disrupts the brain during sleep, causing the onset to awaken the patients. Thus, they might potentially arrive at the hospital shortly after actual onset of the stroke.

Previous studies have reported that IVT and mechanical thrombectomy have different efficacies at different times; stroke outcomes are better when these treatments are delivered in the morning hours.^28–30 Circadian variation in systemic phenomena, such as body temperature and penumbral area in the early morning, might explain these differences. Body temperature during the acute phase of ischemic stroke could have a clinical impact on patients with AIS.^31,32 In general, core body temperature falls to its lowest level at night. It rises to its highest level in the evening, suggesting that a lower body temperature in the morning could contribute to a better stroke outcome.³⁰ In addition, an experimental study using a rat model showed that the penumbra area was narrower in the active phase than in the resting phase.³³ Among this cohort, patients with WUS were statistically younger and had a smaller infarct core size than those with non-WUS. Thus, they could have better leptomeningeal collateral circulation, leading to a lower NIHSS score. These findings support the consistent association between IVT and WUS.

In contrast, the arrival time from the true onset of non-WUS patients depends on when symptoms are noticed. The non-WUS group may include individuals who cannot recognize the neurological symptoms or receive assistance due to critical stroke severity, impaired consciousness, aphasia, or living alone. Patients with non-WUS have been reported to have a significantly higher frequency of atrial fibrillation,³ more severe neurological deficits,^3,5 and shorter time from last known well to hospital arrival than those with WUS,^3,4 which is partly consistent with our baseline characteristics. Thus, previous studies have supported the clinical relevance that WUS and non-WUS should theoretically be considered separate entities.^15,22 Nonetheless, this analysis has shown that non-WUS treated with IVT was not associated with an increased likelihood of mortality, disability, or ICH.

Furthermore, there were no significant differences in efficacy and safety outcomes between the WUS and non-WUS groups in patients treated with IVT. In this regard, a meta-analysis of nine RCTs reported that IVT within the therapeutic time window improves functional outcomes of ischemic stroke regardless of stroke severity.³⁴ In addition, functional outcomes adjusted for age and initial NIHSS for patients with WUS and non-WUS treated with IVT were comparable, suggesting that IVT is equally effective in both groups.

The current study is valuable because it is the first to compare differences in the efficacy and safety of IVT by mode of stroke onset using a well-organized, pooled individual patient-level database of multiple RCTs. Nonetheless, our study had some limitations that merit consideration. First, we excluded patients enrolled in the ECASS-4 and EXTEND trials, which used perfusion imaging to determine the indication for IVT which would have introduced significant heterogeneity in the group. This resulted in a smaller sample size for the analysis. In addition, the exclusion of the population with different demographic data or neurological severity (Supplementary Table S1) from the study might have affected the results. Furthermore, the current study included only patients with DWI-FLAIR mismatch, and the effect of the initial imaging modality with perfusion imaging on the outcomes is unknown. Second, the reason for the unknown time of stroke onset in patients with non-WUS was not available. Finally, the analyzed dataset consisted of selected patients enrolled in the RCTs. Thus, the present results might not be generalizable to real-world populations some of whom may have more extensive ischemic lesions, severe neurological deficits, or be selected for mechanical thrombectomy.

In conclusion, our analysis showed that intravenous alteplase was effective in patients with an unknown-onset stroke, with no treatment heterogeneity among patients with WUS and those with non-WUS. Furthermore, treatment with IVT showed a significant benefit in patients with WUS, whereas statistical power did not detect a significant benefit in the subgroup of patients with non-WUS. We suggest that IVT for unknown-onset stroke guided by advanced neuroimaging should be considered, regardless of the mode of onset.

Non-standard Abbreviations and Acronyms

AIS

acute ischemic stroke

DWI

diffusion-weighted imaging

FLAIR

fluid-attenuated inversion recovery

IVT

intravenous thrombolysis

MRI

magnetic resonance imaging

mRS

modified Rankin Scale

NIHSS

National Institute of Health Stroke Scale

RCTs

randomized controlled trials

sICH

symptomatic intracranial hemorrhage

WUS

wake-up stroke

Appendix

Coinvestigators

Bastian Cheng, MD, Universitätsklinikum Hamburg-Eppendorf, Hamburg, Germany; Martin Bendszus, MD, University of Heidelberg, Heidelberg, Germany; Christopher Bladin, MD, PhD, Monash University, Box Hill, VIC, Australia; Leonid Churilov, PhD, The University of Melbourne, Melbourne, VIC, Australia; Brunce Campbell, PhD, The University of Melbourne, Melbourne, VIC, Australia; Mark Parsons, PhD, The University of Melbourne, Melbourne, VIC, Australia; Nawaf Yassi, PhD, The University of Melbourne, Melbourne, The Walter and Eliza Hall Institute of Medical Research, Parkville, VIC, Australia; Martin Ebinger, MD, Charite-Universitaetsmedizin Berlin, Medical Park Berlin Humboldtmühle, Berlin, Germany; Matthias Endres, MD, Charité-Universitätsmedizin Berlin, German Centre of Cardiovascular Research, German Center of Neurodegenerative Diseases, Berlin, Germany; Jochen B. Fiebach, MD, Charité-Universitätsmedizin Berlin, Berlin, Germany; Timothy Kleinig, PhD, Royal Adelaide Hospital, Adelaide, SA, Australia; Lawrence Latour, PhD, National Institutes of Health, Bethesda, MD, USA; Robin Lemmens, MD, University of Leuven, Leuven, Belgium, VIB Center for Brain and Disease Research Leuven, Belgium, University Hospitals Leuven, Leuven, Belgium; Christopher Levi, MD, John Hunter Hospital, University of Newcastle, Newcastle, NSW, Australia; Didier Leys, MD, Université de Lille, Inserm U1171, Lille, France; Carlos Molina, MD, Hospital Universitari Vall d’Hebron, Barcelona, Spain; Keith Muir, MD, University of Glasgow, Glasgow, United Kingdom; Norbert Nighoghossian, MD, Université Claude Bernard Lyon 1, CarMeN Laboratory, INSERM U1060/INRA 1397, Lyon, France; Salvador Pedraza, PhD, Institut de Diagnòstic per la Imatge, Hospital Dr Josep Trueta, Institut d’Investigació Biomèdica de Girona, Girona, Spain; Peter D. Schellinger, MD, Johannes Wesling Klinikum Minden, Ruhr-Universität Bochum, Minden, Germany; Stefan Schwab, MD, University of Erlangen-Nuremberg, Erlangen, Germany; Claus Z. Simonsen MD, Aarhus University Hospital, Aarhus, Denmark; Shlee S. Song, MD, Cedars-Sinai Medical Center, Los Angeles, CA; Vincent Thijs, PhD, University of Melbourne, Melbourne, Victoria, Australia, Austin Health, Heidelberg, VIC, Australia; Danilo Toni, MD, University of Rome “La Sapienza”, Rome, Italy; Chung Y. Hsu, PhD, China Medical University, Taichung, Taiwan; Nils Wahlgren, MD, Karolinska Institutet, Stockholm, Sweden.